{"schema_version":5,"study_role":"current_base_and_references","formal_base":{"status":"selected","method":"hnm","version":"hnm-v3","reason":"雑音PSDの台形積分とFFT端点配分を揃える候補Dを採用。固定120録音で帯域・平均包絡・音量誤差と予測MOSが小幅改善し、制御と保存互換を確認。明瞭性・スペクトル誤差の小さな系統的悪化を許容した開発上の選択で、知覚品質の優越は未確証。","comparison_reference":"hnm","comparison_reference_meaning":"実験開始時点の唯一の自作ベース。過去結果は当時の版を保持する。既存比較対象はWORLD（Harvest）。","previous_version":"hnm-v2","selection_report":"experiments/2026-10-12/noise-stability/results/run-001/index.html"},"comparison_reference":"hnm","selected_methods":["hnm","world_harvest","facodec","hifiglot","golf","sifigan","llsm2","magphase"],"optional_omissions":["dac"],"method_labels":{"identity":"原音","hnm":"HNM（現ベース: hnm-v3）","world_harvest":"WORLD（Harvest）","facodec":"FACodec","hifiglot":"HiFi-Glot","golf":"GOLF-ff","sifigan":"SiFi-GAN","llsm2":"LLSM2","magphase":"MagPhase"},"native_sample_rates":{"hnm":16000,"world_harvest":16000,"facodec":16000,"hifiglot":44100,"golf":24000,"sifigan":24000,"llsm2":24000,"magphase":48000},"index_methods":["identity","hnm","world_harvest","facodec","hifiglot","golf","sifigan","llsm2","magphase"],"generated_at":"2026-10-13T08:46:11.624578+00:00","scope":"固定100音声以上の現ベース・既存方式比較","planned_cases":120,"contract":{"schema_version":1,"methods":["hnm","world_harvest","facodec","hifiglot","golf","sifigan","llsm2","magphase"],"inputs":[{"id":"librispeech_1089-134686-0008","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1089-134686-0008.wav","source_file_sha256":"2c90ecd16bfcd8dfa5b27b40ead74ef401b6a538b844be4d7e6782a0eb24b1d4","original_source_sha256":"1bdf7eede41ca2cc8883defb66bbd2e59ec93bf2d587327cb051d0ae427495ae","decoded_pcm_sha256":"40f044b90ed06d7cffc4145524ea3e1b8bd36b6599e24e1e905262fbb3e50d8d","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1089","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232"},{"id":"librispeech_1089-134686-0011","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1089-134686-0011.wav","source_file_sha256":"39ef8d5c98fd2bf998cd7a808086b625bf1ce9439d3f2a706cfd0fb0ca5ef055","original_source_sha256":"c65e16be3ef92c11a31a9c2c4c570542eb2ed66153150a7c4cca00c01b30a4e2","decoded_pcm_sha256":"3db35e4a6ed6a12218e0ab1bfe72f1810bad8f201648ec0e2a9af7dae7aa32eb","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1089","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254"},{"id":"librispeech_1089-134686-0032","sample_rate":16000,"samples":65440,"source_file":"audio/librispeech_1089-134686-0032.wav","source_file_sha256":"ad66d8c38a7752bb14d57a12a0136be632673d6e29087b6752939ac518230141","original_source_sha256":"ba1d2185ad72cd3142e6309865859a8e9e5dfdc91ab046adfcc07338c5a0f0f0","decoded_pcm_sha256":"643f787b905d5160255daf650b50e8a272ea88d166a6e7aca2a7ee357cf232fb","excerpt_start_sample":0,"excerpt_samples":65440,"speaker":"librispeech_1089","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d"},{"id":"librispeech_1188-133604-0003","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1188-133604-0003.wav","source_file_sha256":"05590026475894458d72c914ab7238a677196195b9c6cee28855fe9ab4fb84f4","original_source_sha256":"a013400c991a40062cb4f458d247cfd6dac24fa0016bf6e2eac7a3e8e08b8ae8","decoded_pcm_sha256":"81283c3eb08ea201c5ec2d52e3a8e0ed4d0258d76164b522d1ee5ec43edebe81","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1188","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4"},{"id":"librispeech_1188-133604-0031","sample_rate":16000,"samples":68000,"source_file":"audio/librispeech_1188-133604-0031.wav","source_file_sha256":"84a9492ededa3502789e0ce010d0a75be9fa6498a829e0fab2536e7bfd44ddcc","original_source_sha256":"08584f8b87cd859e9212f1730979ed141623f35e38995a88ba50747d04cdc222","decoded_pcm_sha256":"b739cb4e3f2719b98dd98461d6859b487da84f7d1b9f40d280b7f0eb538a6458","excerpt_start_sample":0,"excerpt_samples":68000,"speaker":"librispeech_1188","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b"},{"id":"librispeech_1188-133604-0037","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1188-133604-0037.wav","source_file_sha256":"94ea87763e43bc015868132dbaded57c5c542092378372090a6b2074b6d35287","original_source_sha256":"94c0626f16e39947864688eb2721a57f838b43caeb1857767a9b4381807368c8","decoded_pcm_sha256":"55ff59cb08cac135da664331bc418f7f430ca702cefbb6986e7121d9fe94ca8b","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1188","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9"},{"id":"librispeech_121-121726-0012","sample_rate":16000,"samples":64720,"source_file":"audio/librispeech_121-121726-0012.wav","source_file_sha256":"8ef403805c49abfe19e8e77de836676737356b045e42fce431585140b7ed94bf","original_source_sha256":"6d1fab44fd2e6930a69e1f6f7e52a141ed5290b9a45f56df5e0c3177fa05c6ba","decoded_pcm_sha256":"9c9a8032f3e173a5a354dda74e6cb24f7034610d2cd1b6d15dff6a794afd035e","excerpt_start_sample":0,"excerpt_samples":64720,"speaker":"librispeech_121","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8"},{"id":"librispeech_121-123852-0000","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_121-123852-0000.wav","source_file_sha256":"f25f2203604e14b0e20e4868fc955027532ed005e619778a15e18d6a40a058d9","original_source_sha256":"58d6350b52675d6d19616bc0347afc52a1535f8ea8be82aabb7275e6919a2598","decoded_pcm_sha256":"085897f5f5cc183ef4daa8868cb997e49f924fa58877dcfe167b6419325ce812","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_121","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a"},{"id":"librispeech_121-127105-0035","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_121-127105-0035.wav","source_file_sha256":"2d6fd79d5f61197efa5a679ef5c7afef60e9f5ae3cad9a2112e2b3d4e5a09750","original_source_sha256":"7e293a1a5f64d336563f850f69929dd6b112e3ce37913f4422f2c29e9c55042d","decoded_pcm_sha256":"48306d9610578521a036cf3d52224eb0076f9d14ddb6850ac09278e898fc7303","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_121","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0"},{"id":"librispeech_1221-135766-0005","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1221-135766-0005.wav","source_file_sha256":"6f17d76a753d74d05f4c6bbe31ca3e19a5426dd782db392b1a2a556af4ce1a1b","original_source_sha256":"57ad0fad287d98afb900ccaa898a6e75348b80a3388201cd3e1feee0a0afc6b1","decoded_pcm_sha256":"ae61e68356f4a720a3d9f729ddcb9fa3e695cdbc220194eb938e92185005daf4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1221","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da"},{"id":"librispeech_1221-135767-0006","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1221-135767-0006.wav","source_file_sha256":"c5efedd2f843baa1d0cf0957e2b8a1185a949b0704a25d8f2a8152e92c9f0c53","original_source_sha256":"d70493003bb4a6833163a79d2df440c45bad786d8b50b5c194b896f5a4afd905","decoded_pcm_sha256":"1bf36f00ad8a7e2e9cd6b41efd7f023aab063334b9ad526bda961251493c65cb","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1221","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6"},{"id":"librispeech_1221-135767-0019","sample_rate":16000,"samples":60480,"source_file":"audio/librispeech_1221-135767-0019.wav","source_file_sha256":"ae3b6b6fe861f08a51ef52e51868d2e8210fbc770c6f3aa2377bd61b846765f0","original_source_sha256":"245fd9a32160c5fa9dd1a76624a71c7e1398de763f8e9ab88a7dd25aae1002f4","decoded_pcm_sha256":"0ed57e185be961c721604990dce623ab138f9937e79722c597c3874498c02937","excerpt_start_sample":0,"excerpt_samples":60480,"speaker":"librispeech_1221","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d"},{"id":"librispeech_1284-1180-0020","sample_rate":16000,"samples":93920,"source_file":"audio/librispeech_1284-1180-0020.wav","source_file_sha256":"62b58cb82e8bcd45c378fe5bbfdaf301bde4198a2e998a4a95d67ae07dc44c12","original_source_sha256":"b67810620f4f39a1bca2214b4f6475b98f19452b1cd0a4cc67efee2a8a805016","decoded_pcm_sha256":"0bce114d9020d75b12c3d76b3c232bd8cd315dc052825ca64d389fcb87293422","excerpt_start_sample":0,"excerpt_samples":93920,"speaker":"librispeech_1284","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78"},{"id":"librispeech_1284-1180-0027","sample_rate":16000,"samples":52320,"source_file":"audio/librispeech_1284-1180-0027.wav","source_file_sha256":"d5a15bf2be42bddb14ae3b60a9491ef1223496d85f0be4801c4a72fbc9dce154","original_source_sha256":"9c4d5811c87e0273494850f3401459de8eafd6731f43a615c07f9ff944262a95","decoded_pcm_sha256":"96e7e523ecd8732e2ed1542dd9b6cbde8714adfeef42d0c1bfb37fb1a7c1cf83","excerpt_start_sample":0,"excerpt_samples":52320,"speaker":"librispeech_1284","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581"},{"id":"librispeech_1284-1180-0028","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1284-1180-0028.wav","source_file_sha256":"d144ef3e776a8b3d9055f376aca2bb08127cd682903d5d6a8dcdc44fdd25bb0e","original_source_sha256":"1ee65c4ba89a7fe76bef27dade9b0ab41fe2109b77c8e68d5b80e34194031ee4","decoded_pcm_sha256":"0bfea5c92c4d2ee444f3984a4d2b53407bcff068a4ef104efccbb0731c86faf4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1284","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b"},{"id":"librispeech_1320-122617-0019","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1320-122617-0019.wav","source_file_sha256":"e48b2a34c6417b658e12d8e8a5139b12b81676a3fa42dbfac6037f099b90b1f5","original_source_sha256":"3f0ea5a1e33f502dca87255f51487aa60fcb4f12605e9feba2adeb14df1d89f7","decoded_pcm_sha256":"8e4bd7f99207ae6f91c9d166201c4877840bfbd64affa6d81029b74db79e47d6","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1320","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2"},{"id":"librispeech_1320-122617-0023","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1320-122617-0023.wav","source_file_sha256":"d525891a3974fe432ea2357061f350966460bf770f6f7cac30ddd42207131539","original_source_sha256":"553fc6594343774831ac876fee34c4ec2b47b53d451f313b10e64eea5cadd8c2","decoded_pcm_sha256":"b3661757a6ed833feab8313fa697ecafb680dc14c418cd468619e460e550745e","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1320","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2"},{"id":"librispeech_1320-122617-0039","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1320-122617-0039.wav","source_file_sha256":"798fc523e613ab0e3376effff3d55c6cae2be184cd29e158f6692c3e62d044eb","original_source_sha256":"264a53a8c31bd50758ba7eeeb3b5efa9c5bc7908757d44a635bec4c9a4f2c3c0","decoded_pcm_sha256":"9cf3ed8985625af93ca84569280377a536aa4892a8aefbabb265fc0a052a863c","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1320","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e"},{"id":"librispeech_1580-141083-0010","sample_rate":16000,"samples":85120,"source_file":"audio/librispeech_1580-141083-0010.wav","source_file_sha256":"5a32c325118ac1003afb40a432e94614a627b8dbbe5d8927e227f8d63af0cb7c","original_source_sha256":"15b52d8b6b3ab1e0d95b51380278481420b17cdd350b6fe62c29aca97d46a6a7","decoded_pcm_sha256":"d43cef78fec64a97f03dd03591e102c104a281997670664d424ccf19276cb4a0","excerpt_start_sample":0,"excerpt_samples":85120,"speaker":"librispeech_1580","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7"},{"id":"librispeech_1580-141084-0016","sample_rate":16000,"samples":95360,"source_file":"audio/librispeech_1580-141084-0016.wav","source_file_sha256":"38b92a767a1ad3e3e13c6fe5e48ec69c697a08ef308c65254cb22a81451b4ed8","original_source_sha256":"86f57be846b5ad8817f2752d90d5f2aba1e97ec394b2f4ccf28bc5dc39a181ef","decoded_pcm_sha256":"0698365674d63379a8ffa4e1a7bae8ef5f6ed226e14107811f5311d9de011ca9","excerpt_start_sample":0,"excerpt_samples":95360,"speaker":"librispeech_1580","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba"},{"id":"librispeech_1580-141084-0047","sample_rate":16000,"samples":84000,"source_file":"audio/librispeech_1580-141084-0047.wav","source_file_sha256":"2224f20789b5ec071ee8447cea9283180ec32feddfd092a06f8e23d4e930ed24","original_source_sha256":"0e78a5bcc7c122a563c24dc2d7a65f1be8d851e0d45c8caa4226571676cc5d4c","decoded_pcm_sha256":"ca185084384947ca6386941bb27f596abb096a46a774b4716da42a2e461e4cab","excerpt_start_sample":0,"excerpt_samples":84000,"speaker":"librispeech_1580","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723"},{"id":"librispeech_1995-1836-0000","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1995-1836-0000.wav","source_file_sha256":"7f367cdfcead3a8df60b8c12b534b1bd6a1b44e49274ab0d3f01750bd5374395","original_source_sha256":"8120c55c7231b13d3d497b04b2fe3d999cfdac785a0134eb0fc9e41aa90ce591","decoded_pcm_sha256":"3feb5dcb82941d2a5e9f89efbf0e32e0138e5eec693c118bb9e8031d23d08925","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1995","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26"},{"id":"librispeech_1995-1836-0014","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1995-1836-0014.wav","source_file_sha256":"06a364bfb12fdbf27e8c22222660542c04f7462d9cfdbf9f1a9c9b94e92063e4","original_source_sha256":"c90576861bc6bc0b6837f0b1c340ecc276ac0339ee0a25442a4784f5fc4a4821","decoded_pcm_sha256":"af7f0befa255d17d0806307d2e5a0582ff2d024a50adfd0a534360c964de3b82","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1995","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6"},{"id":"librispeech_1995-1837-0023","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_1995-1837-0023.wav","source_file_sha256":"61753cb8551c9914d0cc152a02c1b7e38df1dd0cf1914ae624de66ff4033f523","original_source_sha256":"7f584164eef2f5976f874ee9c027933c892ba7f920f7233bfd2a219d0ebffcd4","decoded_pcm_sha256":"2fff476dd7493d67e46c49e37f104d20891e347d86560cf73c09f434075020cb","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_1995","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd"},{"id":"librispeech_2094-142345-0032","sample_rate":16000,"samples":51840,"source_file":"audio/librispeech_2094-142345-0032.wav","source_file_sha256":"5782c94353e36728855a2c0380c90d1344bb6a52b8af06c3217c5124fcc92f8e","original_source_sha256":"b92bf752cc5264a2a47e0153c236b90edb48dd32b819cfd12d0e5250ca3c7a1e","decoded_pcm_sha256":"973bdb09bc9497c63b4dfbb3bc77778c87a6ced88d08e7f2f6e22739dd6d5444","excerpt_start_sample":0,"excerpt_samples":51840,"speaker":"librispeech_2094","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53"},{"id":"librispeech_2094-142345-0039","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_2094-142345-0039.wav","source_file_sha256":"ab022a45def58f8bdddaa158088e2eedf83eb4cf0a3259088662c2eb206f239c","original_source_sha256":"21d7d00350a4ca2b0109da67cfad724f36979e7c1b2e098f9327bc8c4f433435","decoded_pcm_sha256":"c411d502f97ab6fc22d81b52c4ba95e5e562514c255c2379c8837148877e0ea1","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_2094","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b"},{"id":"librispeech_2094-142345-0040","sample_rate":16000,"samples":38160,"source_file":"audio/librispeech_2094-142345-0040.wav","source_file_sha256":"8d136be714c564f76e87ca6fb1a10f8ef8bf4df7a3e056c0803a06854b098fe2","original_source_sha256":"8b3f5a025e4853fabdb000f2820a651f4a66a5ef234d912581ae29fafea72e48","decoded_pcm_sha256":"9cbc49d67edc85ca6ba00d13dae38ec545b7aaf549562d1af9d9e72881814d19","excerpt_start_sample":0,"excerpt_samples":38160,"speaker":"librispeech_2094","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd"},{"id":"librispeech_2300-131720-0022","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_2300-131720-0022.wav","source_file_sha256":"521b559ec12971c556226c0bbb186563df37e99537a0d0ca2bcfdd96bc78e538","original_source_sha256":"93212060c39673d7a3ee80b52a8f2612e70e3c121fd588c690e3d51be10e6963","decoded_pcm_sha256":"b7f1420236bac4a5c404bc851efe87bc6142a76f2a2935e40197b946d8182998","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_2300","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f"},{"id":"librispeech_2300-131720-0028","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_2300-131720-0028.wav","source_file_sha256":"d3f454e289791b711d49008dfe821cf4c9944922cafd6a5dc3a3f28ad1285c96","original_source_sha256":"72f44e8aee815c88d0ad1b8f6d9a15fea55954dbd8e4d996b97112fe34bda0a1","decoded_pcm_sha256":"9eaf55b481a213cff53140518cebb83838d1187d3e2187282fcf13a3699505c4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_2300","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1"},{"id":"librispeech_2300-131720-0041","sample_rate":16000,"samples":60000,"source_file":"audio/librispeech_2300-131720-0041.wav","source_file_sha256":"c3941af709f54603bb04f691cf75af18cbf399e2d4e6a66af8390e53fb199137","original_source_sha256":"4ede50e44cbe5dfb258311f80fd68b7efe2e080265efa5e6bc07aea531a22e28","decoded_pcm_sha256":"976092b4e8947717eb460c0f8df2d08babdc9b088a74a4efec5fca50e34afea2","excerpt_start_sample":0,"excerpt_samples":60000,"speaker":"librispeech_2300","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1"},{"id":"librispeech_237-126133-0003","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_237-126133-0003.wav","source_file_sha256":"be1753ee940eacd39316a50f4f44a5c8c2a2add8753d47ff98918a1bf090423a","original_source_sha256":"18c38673f71d4012a9c023d7548bff3c0f78172415f67c69abf403590c4e84bc","decoded_pcm_sha256":"15627325ed4198a6a91f3a9dd72c16085a4986165c018e492c56a9076ad969f4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_237","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa"},{"id":"librispeech_237-134500-0029","sample_rate":16000,"samples":33040,"source_file":"audio/librispeech_237-134500-0029.wav","source_file_sha256":"e65170b1f0140c829be8d76f0ea22de10705c116c3990aaf43083c69b4e44954","original_source_sha256":"a71914689325868d93f40677ab17bdfbb8c8209801864e31a4f4303cf7fd0fa0","decoded_pcm_sha256":"8562d47701848db95c8610733017a69a5928db4edb524e0f205ab4182cf7b4f7","excerpt_start_sample":0,"excerpt_samples":33040,"speaker":"librispeech_237","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd"},{"id":"librispeech_237-134500-0031","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_237-134500-0031.wav","source_file_sha256":"14d36d955bcc1b83a5e97a4d5b31b211e9963b601b8ae957f5a113f1412ea101","original_source_sha256":"7456acb116b3a8c6d43c44ebf79a8f7b59f6ed546ffb63ce78a6dc58bb6d5abe","decoded_pcm_sha256":"66fcbe0885d7123cd5c5c14ff8b544f94ed65d51857cb70d13de972dd00f9421","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_237","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94"},{"id":"librispeech_260-123286-0008","sample_rate":16000,"samples":59760,"source_file":"audio/librispeech_260-123286-0008.wav","source_file_sha256":"387f69e03283db57cbc29a7b48adb0c4038d0d72095c13e509e89cbadbae4e26","original_source_sha256":"735769549e4dbbfb3f4ba43f89866cde8085933d8b16e67550cc65264bf84fc3","decoded_pcm_sha256":"931193da803cf63a31b439032de16719f19aca2c3130d571b4b8044d43a1413c","excerpt_start_sample":0,"excerpt_samples":59760,"speaker":"librispeech_260","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a"},{"id":"librispeech_260-123288-0003","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_260-123288-0003.wav","source_file_sha256":"df2faf1f0223442b55f77c221ae6875598e81be415b8c0c128ec31539dff25e0","original_source_sha256":"5e8ca953fe1b2dd885f0cfbf4d706040e55637eee883bea8f0dc6e08afac3240","decoded_pcm_sha256":"d480f00d02d595a6352c610e02536f099ff03473f6693b5628683647e9479766","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_260","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc"},{"id":"librispeech_260-123440-0020","sample_rate":16000,"samples":79920,"source_file":"audio/librispeech_260-123440-0020.wav","source_file_sha256":"28dfa70fcf9350194cc1b8a250dfc97be7ac06dd3875c812cd01caa93ed803b8","original_source_sha256":"44be2bb2c15059058798a25f87de6310601074e1bc4b68bda0ad3ad1860a9040","decoded_pcm_sha256":"c10a9cdd8fd4960bfb88b1611c0123695ebcd137816c47ff78925a8699616669","excerpt_start_sample":0,"excerpt_samples":79920,"speaker":"librispeech_260","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2"},{"id":"librispeech_2830-3980-0000","sample_rate":16000,"samples":59680,"source_file":"audio/librispeech_2830-3980-0000.wav","source_file_sha256":"0c85f1d38569a847df48f56c9a056afc5b892dfd3b129170c8269b485334be9f","original_source_sha256":"a73adb84b9560fdfc8f6c7f5801e567267ac8beb6fd51df4f5dec7217eb43ad8","decoded_pcm_sha256":"8f71058a78243905dd373f5302bc82f25ff2a7e7631413b5ce924104f0f14bea","excerpt_start_sample":0,"excerpt_samples":59680,"speaker":"librispeech_2830","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496"},{"id":"librispeech_2830-3980-0038","sample_rate":16000,"samples":88640,"source_file":"audio/librispeech_2830-3980-0038.wav","source_file_sha256":"fe0db7c2f0544f554e2fc2cfca5c7063632e71768f8fb1117c5a3947cec2f9ee","original_source_sha256":"2c8978e75dd10716d066062bb2fbfd1a002ac3190e42b0f251f15d0aa2d6d272","decoded_pcm_sha256":"77d64124560df236e40a3ec37cc728b327f805fee264fb8362c33791320222e1","excerpt_start_sample":0,"excerpt_samples":88640,"speaker":"librispeech_2830","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9"},{"id":"librispeech_2830-3980-0042","sample_rate":16000,"samples":48320,"source_file":"audio/librispeech_2830-3980-0042.wav","source_file_sha256":"1ff7af7ba8275fdd191f6a105fb066b79b8aa226a7f899aaddf0c12016c2bc21","original_source_sha256":"ff30195a24b3f6c24a6ad4748f2ab49f4a036975c0b62f5b1c2f4203488ee9ae","decoded_pcm_sha256":"3b4a08b0741e0ca396487185085d64365ab56b16d5868e61a4550a5dc24466f2","excerpt_start_sample":0,"excerpt_samples":48320,"speaker":"librispeech_2830","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70"},{"id":"librispeech_2961-960-0010","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_2961-960-0010.wav","source_file_sha256":"b0f1c819a380cb1d5ed8992119735708677bb6b436203f4e972eb1a9accccc42","original_source_sha256":"d93e37326bfa987a0cf66c0edd79dfa034ea013e5967c8f2ba9edf54cde8d2be","decoded_pcm_sha256":"ab9593d38d133538cf3475c5a7bf30b265737c427a5c4a501ecf3384b1e2aee8","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_2961","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19"},{"id":"librispeech_2961-960-0019","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_2961-960-0019.wav","source_file_sha256":"395f9b6c5e4cca0116fda31ec17844a1ccc1c2f86a13d96e1bb769ff09217364","original_source_sha256":"f5052549b9cbc3b8433ae187d0e876b214f5250ddb4fbb637dea220e8a4726cf","decoded_pcm_sha256":"966508e2ccc362392d8ec80fc3d05a564460cb6f0dd69f613c7b86089cb21caa","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_2961","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09"},{"id":"librispeech_2961-961-0011","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_2961-961-0011.wav","source_file_sha256":"615aaa8347b39864569d9a08b03be7ac1d6d4b1bf4ad43926f37f5c30950eebd","original_source_sha256":"f859855c45fc07e8d561ed7a846fd14ca4ae1515749b60c4e8d4ce825df30180","decoded_pcm_sha256":"00bd4132c58a261e6169cd035b93bf1134d0bd1b082f44ceb99cf14faa9b7062","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_2961","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269"},{"id":"librispeech_3570-5694-0001","sample_rate":16000,"samples":90800,"source_file":"audio/librispeech_3570-5694-0001.wav","source_file_sha256":"9c431a66d7f2d4bfb2dfd819f2f2f38c03fed56b8fb5cc0ad15a2a509754f29f","original_source_sha256":"83d42febda01d467c35c0b90dc7bb5d5d80b18ed95d65748a831afca53f2e3cd","decoded_pcm_sha256":"d7ebe73ddd2095245c2316a87024f8d15117a6bca41809d6c43a3d2cfe68c3a9","excerpt_start_sample":0,"excerpt_samples":90800,"speaker":"librispeech_3570","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de"},{"id":"librispeech_3570-5694-0004","sample_rate":16000,"samples":85280,"source_file":"audio/librispeech_3570-5694-0004.wav","source_file_sha256":"03e14afbc15ed13fc22b31c48c55d982de799c1e9424b711c83a36988c6c96a9","original_source_sha256":"e7c4c68060bd5da1df857aa90f0fb6601cfee83e0001d75b64164988b070e7d8","decoded_pcm_sha256":"86a8946d8572399656832193f585d37dfcbd327696eb662dcf0e41d4b67d0bd1","excerpt_start_sample":0,"excerpt_samples":85280,"speaker":"librispeech_3570","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88"},{"id":"librispeech_3570-5696-0010","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_3570-5696-0010.wav","source_file_sha256":"d61b6ff35e8fc55f36b0f2e01ce1faddd28c2c5419c859fb12828276eeb28574","original_source_sha256":"426a986fec027f14754878775f6b605d58d61a0fd0d77bf122657dc1cef2ce9d","decoded_pcm_sha256":"0efec23fe6eeee5dfee017e6be4b06b64a152a981c62a71c533addc819701d30","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_3570","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f"},{"id":"librispeech_3575-170457-0034","sample_rate":16000,"samples":55920,"source_file":"audio/librispeech_3575-170457-0034.wav","source_file_sha256":"e3c2614b784bb39cc4e1529e4792c7ff4d8db18f982059dc3b9e56777c712c38","original_source_sha256":"7379f6afffbfc11aded5aae60cf94ff2f5aa5a2dc317f61b2cd70c9c4d4847df","decoded_pcm_sha256":"b1c60dfc008f07f56156adc48b00db15f16da435fb212895c0516ecbd9e5de30","excerpt_start_sample":0,"excerpt_samples":55920,"speaker":"librispeech_3575","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3"},{"id":"librispeech_3575-170457-0041","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_3575-170457-0041.wav","source_file_sha256":"53e8ec669f1f356f4ff4e4c4bb08474703995fd940542ba42b9cccea6f9fc4ef","original_source_sha256":"90d1dd17384a0960d3acaf50b3f7f40b4699288aa74a04c25a221270e104ce55","decoded_pcm_sha256":"ce9155c8469c627681fdbab9602fc52b17f46ce8849d53a2be4d450d58cfb2c4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_3575","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630"},{"id":"librispeech_3575-170457-0054","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_3575-170457-0054.wav","source_file_sha256":"e3b4f2201808714c1a2a293e0a482bb754beddc4c5bf9617bddfe762663396ca","original_source_sha256":"638053737df40babfd92ea956f1083964cfe38bf547b96334228a8f67b2cb78b","decoded_pcm_sha256":"5ff0a2aa3e7e6a57cda2e60dda3bc719388bba59551820dff2445caca9cf7c6a","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_3575","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8"},{"id":"librispeech_3729-6852-0000","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_3729-6852-0000.wav","source_file_sha256":"7ca2b3e40ceb9cbcc21cf23410fb69486e85d9ff72b3428ce5c9a57b4a33ee7e","original_source_sha256":"3fa96f8e0276d79ec027bb1595215a8bbeb488cefcc1702ff798e91ec8504403","decoded_pcm_sha256":"e1524c8eb759928cd93fd86e98694c9a5efccdfa2a0fd948935f558dbffb8843","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_3729","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a"},{"id":"librispeech_3729-6852-0007","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_3729-6852-0007.wav","source_file_sha256":"5f0b275d90aac419fc4594e557570e79629b14905e85a01c05b0b6f6fe069164","original_source_sha256":"9d9bbbeeff1bfb647e2698db36f186cb439a8b2894772ca3cdfe9425c3e55a1a","decoded_pcm_sha256":"d081e3d6cdc89188a9f3d229a815f9784efd7e72dfcfa5317b4ac90fa9d86ff1","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_3729","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a"},{"id":"librispeech_3729-6852-0017","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_3729-6852-0017.wav","source_file_sha256":"b5aaf417e4a9ed7ec0ea0155bd259da3e7e8e060c9adf56f1ec177ef5c71605f","original_source_sha256":"12a9659e29190ce17dd89b5209d4ccaebc746bfb35093185d5970a91878e42b9","decoded_pcm_sha256":"ba156f183286a1f4044b0b7794b3982603585d0e122bdb35978c337419392c08","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_3729","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c"},{"id":"librispeech_4077-13751-0001","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4077-13751-0001.wav","source_file_sha256":"e214cdf66f38d5ff7aaa62fdda033f0efb71989942844ccd21cbb6008ac5a300","original_source_sha256":"1f85c5ad8dfb8fcf3c8d86e54afc055f756f147e7b426ff8b7f48780d8d43c0e","decoded_pcm_sha256":"ff5a7ca9328a329a2ca9ba953b0550653b86fe7bbf7bda2cae73908a1fbfde2a","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4077","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727"},{"id":"librispeech_4077-13754-0002","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4077-13754-0002.wav","source_file_sha256":"1d6db9cffedf1980251f6002f51b33778a8cd31fbedf9aaa094e7a74147dfc39","original_source_sha256":"58ff360d46e4fc6bcaf04da1a6ead97dc191c54426528cfa13efe7e780349ff5","decoded_pcm_sha256":"e4ef84f14b8958b7a78512946198f223e31d6190beadfe364d4772e4a9963ced","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4077","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81"},{"id":"librispeech_4077-13754-0012","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4077-13754-0012.wav","source_file_sha256":"047625b1a8702b3b799110165a267b4b9f3d72c44f0bc718b86ca459a52106a6","original_source_sha256":"65e48d9b9257be487e80c6938e34856d742870d390ad5d3d1b04d5f95f955d4a","decoded_pcm_sha256":"43e800e9f69ae92ed54ca36d54a78bd7c56d8ff0657f75ee376d88adab6b3764","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4077","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed"},{"id":"librispeech_4446-2271-0003","sample_rate":16000,"samples":59200,"source_file":"audio/librispeech_4446-2271-0003.wav","source_file_sha256":"ef62b1d7b8e854066c8a34f2fe8730e01e13176767e11878cfb138338cea5984","original_source_sha256":"b8fecefeca8a7ae57ad53900dbba581245a09ad79d46ae1fc06140e0e5a5b281","decoded_pcm_sha256":"257baf7b0200414f4ef5d6ab5db2362889681a1c0de114633c54f50f9b628549","excerpt_start_sample":0,"excerpt_samples":59200,"speaker":"librispeech_4446","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68"},{"id":"librispeech_4446-2271-0022","sample_rate":16000,"samples":37200,"source_file":"audio/librispeech_4446-2271-0022.wav","source_file_sha256":"e113d0c7213890d71aee6ee9f66e740262a4dfc16988c059f2628ecba02452b8","original_source_sha256":"ed740f49d11cc5c58f71f1925854cc9c1f30170b81143873aa8c58cd35430583","decoded_pcm_sha256":"36c74fdf4f12eb94aa845b16c467769326a4db667ab611a8845b7f72c2cae5b1","excerpt_start_sample":0,"excerpt_samples":37200,"speaker":"librispeech_4446","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a"},{"id":"librispeech_4446-2273-0010","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4446-2273-0010.wav","source_file_sha256":"93cfb1377861cfa7014bfa159ec6b9409183e0c8c4b0bf020dadd18397901560","original_source_sha256":"ee6acd6b5c33e2d7e114405b48aec2700f720d68dd8d59676172e81ca63317dc","decoded_pcm_sha256":"263c284aa66f0a837f0cdfcf012d15e0d4edfeb7a8a0717217e17395942a83fd","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4446","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f"},{"id":"librispeech_4507-16021-0016","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4507-16021-0016.wav","source_file_sha256":"b1b65b7b62d666f4aa4fc10e2b6164e6b9e988f301494abde1f459c839b1cc3e","original_source_sha256":"e3cf855f3ed817d96563b7914780df19ad9175cd7deb4a04880dfb6093dcccf7","decoded_pcm_sha256":"0f759c362b7bb3f7ab6547020834b6f41d471a7cfe5c2018d0a8cc506e042103","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4507","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a"},{"id":"librispeech_4507-16021-0032","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4507-16021-0032.wav","source_file_sha256":"d54cfb67fc1814dd0dfbf13adf8cf7f849a96d00270124bc506980de13a270fe","original_source_sha256":"cf872613e0d590501ff569c17e5f9d525920c4babeda14ac94577f60cfc54401","decoded_pcm_sha256":"7576336f86f07936b03cf91b64bf80c306a8a013d42a9e0eb9b1dae02be902c2","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4507","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad"},{"id":"librispeech_4507-16021-0045","sample_rate":16000,"samples":78240,"source_file":"audio/librispeech_4507-16021-0045.wav","source_file_sha256":"7f1752bebd45b4bb6d9f05107f15543d5f3caf6fad16945c0061cbad41288a5e","original_source_sha256":"86e41fe1ab9a5fdf5d806ab2728428f4b8d2720df04ce6431b65f617878b924e","decoded_pcm_sha256":"ef91d0d7b9ebfc4c08837a0dcebf79c72e3e4ff97d381f356a678c2a085d6e6c","excerpt_start_sample":0,"excerpt_samples":78240,"speaker":"librispeech_4507","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463"},{"id":"librispeech_4970-29093-0004","sample_rate":16000,"samples":60000,"source_file":"audio/librispeech_4970-29093-0004.wav","source_file_sha256":"560787fb75d65682ec423601d38fbf4913601648283cc88fc70ef26b7931b23e","original_source_sha256":"48c89f99c44d1be73c92460e6d61d4c1090efe126701f7492184112d37575229","decoded_pcm_sha256":"222e257a563c92e3e220ca509960e218a09ffd8f995b3f1148bba432a24fc5cf","excerpt_start_sample":0,"excerpt_samples":60000,"speaker":"librispeech_4970","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25"},{"id":"librispeech_4970-29093-0005","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4970-29093-0005.wav","source_file_sha256":"6d3951aa443b1751a0352c47cc11de27767730226074c34a517bd7d565532fcc","original_source_sha256":"f466147cca3169748874913da56b587fd3d7fed00a285a1c9124dac56a8323ba","decoded_pcm_sha256":"e5b971d7dd1c4efc72382481eb6d7cbe047355a62af1c39714d28b6cb739b74e","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4970","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7"},{"id":"librispeech_4970-29095-0008","sample_rate":16000,"samples":48640,"source_file":"audio/librispeech_4970-29095-0008.wav","source_file_sha256":"d8b3795ad9c18b024d18c5c9878b817cf4a929b1285bab7254fea0887a57681d","original_source_sha256":"cd2a3b352b93beb33caced744406acf8dace3f1bbb5fc63c8feb4c8a0ed55bb4","decoded_pcm_sha256":"44ec7bab382a51da3a9f07cbc8ffdb491029ea5b6b84d63bfceb71c60641803b","excerpt_start_sample":0,"excerpt_samples":48640,"speaker":"librispeech_4970","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4"},{"id":"librispeech_4992-23283-0016","sample_rate":16000,"samples":71920,"source_file":"audio/librispeech_4992-23283-0016.wav","source_file_sha256":"cc2c4387ab000575766a8c38139e5c4dbedbbc41089848ab66e300a1fc82e490","original_source_sha256":"e7442a8f1bbd834ad37ee3c9cc5fa26b1cfb61446f827b54298716f59b010002","decoded_pcm_sha256":"4af74b1b8682fb324d2e81cb2f0cc0dc50b77d4a3b44fd510cfb4df09ae57d6d","excerpt_start_sample":0,"excerpt_samples":71920,"speaker":"librispeech_4992","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788"},{"id":"librispeech_4992-41797-0000","sample_rate":16000,"samples":87760,"source_file":"audio/librispeech_4992-41797-0000.wav","source_file_sha256":"dcf4cf49fd10acb3fc555b99d81b83725063938be4428632f1b3d7c8db683d41","original_source_sha256":"f7dbbd42eb06171dccf937d53a6c9144e51b88ea5b1b2ee69978900b87216fb5","decoded_pcm_sha256":"70049242cb761e6a91df355dd7a9dbdfb974c89071ee6b5dd83506a6ce8390e3","excerpt_start_sample":0,"excerpt_samples":87760,"speaker":"librispeech_4992","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1"},{"id":"librispeech_4992-41806-0000","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_4992-41806-0000.wav","source_file_sha256":"c80465377971cbc3794c7aa181ffdb6dfd537344fc2142eed4e4e98a33a7054d","original_source_sha256":"ce4dd04d130b3b575f1826c6505b5519c09a7301eee9b76e58eeb1f591131035","decoded_pcm_sha256":"9e17e42f2de437c240d5850615912f710137bb8bb0f2e42082d93bc758515234","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_4992","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149"},{"id":"librispeech_5105-28240-0011","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_5105-28240-0011.wav","source_file_sha256":"d9054a56be6d8aa2c82601a041821f097896f29370337bbc16e2f2a8aef2943b","original_source_sha256":"2e162ff81e45f909e947ebfae4f39e339d73d5fc19afafb1c903db96f7f8532d","decoded_pcm_sha256":"b9bc3828aa0284c0660c9ec7acd1980026fcc21a26b62a3c5c68458711b73ba2","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_5105","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16"},{"id":"librispeech_5105-28241-0004","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_5105-28241-0004.wav","source_file_sha256":"b7fb218e5a8c5a06c4ace6cfdc94d6da236d9a9cca5b2fa66fabb9a917f514ec","original_source_sha256":"70c67685f8e24f7a27f6de3b63d5b3ecbde28f017d1b7aac9304f6c6034fa1bc","decoded_pcm_sha256":"6704bc85da6844b255174407f722499ab1188299d4775a37e25cc238c7fb903e","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_5105","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696"},{"id":"librispeech_5105-28241-0019","sample_rate":16000,"samples":84640,"source_file":"audio/librispeech_5105-28241-0019.wav","source_file_sha256":"0558e278e1b22e4c0947ea5f8a5bf3806a82dac8042ef2f18d2b184c96da6a64","original_source_sha256":"96d523790a28d90ffd89ea67603421a3881a96d21cb523495d1a066c4c5ceecd","decoded_pcm_sha256":"3d0b0ee3168cffdc7dff07188417fdc9c6c2c31369b906ef6a97ab560d3e8f4b","excerpt_start_sample":0,"excerpt_samples":84640,"speaker":"librispeech_5105","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864"},{"id":"librispeech_5142-33396-0019","sample_rate":16000,"samples":79760,"source_file":"audio/librispeech_5142-33396-0019.wav","source_file_sha256":"06367c6cb3079127c93fc4a6fffbbfb8c3c4ed3f6c741637f2ab5480e7d1b40d","original_source_sha256":"2025e96df477d9ebc46b91bae88c601221bbbb2ab12d9fa83ceced80b0f9b72f","decoded_pcm_sha256":"f3e12cfe88aca2fb306552ad3fe2c213b91a4bfc2f0aa979df8a1f01006ba85d","excerpt_start_sample":0,"excerpt_samples":79760,"speaker":"librispeech_5142","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb"},{"id":"librispeech_5142-33396-0030","sample_rate":16000,"samples":44240,"source_file":"audio/librispeech_5142-33396-0030.wav","source_file_sha256":"2ee1bc4f1d3a77404c9400e73aa144540d074b19be626ddd611d3dc304da9aa5","original_source_sha256":"87b13e7bd9a84ee5f9ba237f84dc992d0fbbad061a6df1750b87a47756642b33","decoded_pcm_sha256":"cdc8b64c805f94ff6dd7f431b45e924a4edf992efb5d08cda66456142b76481e","excerpt_start_sample":0,"excerpt_samples":44240,"speaker":"librispeech_5142","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7"},{"id":"librispeech_5142-33396-0065","sample_rate":16000,"samples":83120,"source_file":"audio/librispeech_5142-33396-0065.wav","source_file_sha256":"716e93df6d4c3ffff814b4fcaa46a3a7f31b9d4ca81db41240c09b7c06aa7060","original_source_sha256":"67800641844bf8c9d654fc9a6099941ba97ff31b0ac0a579209378598c4ea09b","decoded_pcm_sha256":"0b650128a600d9d1c16e5bf3b64c63018c80e6b3463b2d6fb23e3aead6248f75","excerpt_start_sample":0,"excerpt_samples":83120,"speaker":"librispeech_5142","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291"},{"id":"librispeech_5639-40744-0013","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_5639-40744-0013.wav","source_file_sha256":"0188b213089d4bb2539294730de6f9f1670e04dd5f7e91884fb74c3120baa71b","original_source_sha256":"a2688cb66fd448ee0d578c6d82312afea0f9ad95b24350f251eb79ca5a4f759e","decoded_pcm_sha256":"ed7f1ce67f421ec6afa8566971a1c9c8ec8d332f4950f7baa3aacb009b96f5ef","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_5639","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd"},{"id":"librispeech_5639-40744-0023","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_5639-40744-0023.wav","source_file_sha256":"92a0351de2e57e7925cc2113b8ec9b404a11cfe047eaa2954ee3de5ebb673859","original_source_sha256":"dd277d4b42ef738b14b59fa1d835988e71ec031e2f01d024279202edc26ad9c0","decoded_pcm_sha256":"da6b45c38da036a67d46d62fc3e6e54310d46e3e14d8f0a3eff1b445bb1755f4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_5639","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d"},{"id":"librispeech_5639-40744-0030","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_5639-40744-0030.wav","source_file_sha256":"ef3ca02fb369c8c43fb91d72c92144b36cb9942934a774dd18899f187e80362e","original_source_sha256":"c7f95b2f012ab8f4629d15c8c78e78bf17b3445c96e5a9b0c525b0743e50d9c5","decoded_pcm_sha256":"76c45c30d066507ae3cbb52fafb188ebe36da864c7c54942db2ea8356726ce3a","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_5639","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1"},{"id":"librispeech_5683-32865-0017","sample_rate":16000,"samples":87280,"source_file":"audio/librispeech_5683-32865-0017.wav","source_file_sha256":"efcfbc29d0a21169b4a87f29a08a17b985733cb5cfae848f6a03eeddc29dbd99","original_source_sha256":"d0027b49e5e0c29b3b59d1d782b8d4d9753f90e4343f243a15698e787c8e64ca","decoded_pcm_sha256":"44477e160959a9526d3545567baa9b096405cf9d914595d737ffdf3d46ce1e6f","excerpt_start_sample":0,"excerpt_samples":87280,"speaker":"librispeech_5683","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff"},{"id":"librispeech_5683-32879-0009","sample_rate":16000,"samples":82160,"source_file":"audio/librispeech_5683-32879-0009.wav","source_file_sha256":"7e131dcc24fb3acb39b942544291e958984d331bb65bf3a6a2dda4ee23a77392","original_source_sha256":"c66196c0436cce295ac3ae7695f4904f87672f1d63d89b8810937a146378411a","decoded_pcm_sha256":"c0eff412ac130cf1ccdb29b7cbb8149ed148614513d6a355a1677d911d2c6b5f","excerpt_start_sample":0,"excerpt_samples":82160,"speaker":"librispeech_5683","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473"},{"id":"librispeech_5683-32879-0019","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_5683-32879-0019.wav","source_file_sha256":"94cca041d58d55b61cd2e7d92edb13560c9406e92ee6a61cb1626326c7a32d1c","original_source_sha256":"0c9375f8c28b1576aaa73e07006a7fc838f49b250df910885e24a16abef0c19f","decoded_pcm_sha256":"cf57296e7013c02425f574c6c3f30623e8057ab1e199fc8fcd761f1b2ff63d52","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_5683","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e"},{"id":"librispeech_61-70968-0042","sample_rate":16000,"samples":44560,"source_file":"audio/librispeech_61-70968-0042.wav","source_file_sha256":"95a00794dc691d3fd435660998edc22cb01aeb3e2b0be248166f3e6b10c95097","original_source_sha256":"9aa0aa0746140383c3458b53991ccd9ebfc2c0ded6c6817ec0748aa70e6b8088","decoded_pcm_sha256":"4d1c295afd61c34ca0f59a70b2baace0777ff6e94127ed717b67dc664c9d9ee4","excerpt_start_sample":0,"excerpt_samples":44560,"speaker":"librispeech_61","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f"},{"id":"librispeech_61-70968-0043","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_61-70968-0043.wav","source_file_sha256":"5ccd87f12173ce731b81ea8fdd47a9882e4032382dc1f9d8dadbded184e153bf","original_source_sha256":"92cde1bfb6bab32cd1428505868f1ca27db04ed17bec9e47e7cbb24e34dc09c3","decoded_pcm_sha256":"1731c4205d3ab140aea21c6e2df47e27d59a45b1eaa6b80c407cddcad99de7af","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_61","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd"},{"id":"librispeech_61-70968-0054","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_61-70968-0054.wav","source_file_sha256":"f96060003b37e55a94719dc0c8bc7e31f44bfe3675659a89fefd26a4482d80db","original_source_sha256":"7b58da65dd63a6b09d43ee20c6aa4b5a8853199af7c350a87ab692fb4d0d874c","decoded_pcm_sha256":"0c30dfc3b23081f52641ba6b1793c8bc5c9ea236aea529b533de133f00169e95","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_61","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734"},{"id":"librispeech_672-122797-0003","sample_rate":16000,"samples":76160,"source_file":"audio/librispeech_672-122797-0003.wav","source_file_sha256":"18161f622ffad5f72bfb78f277ffe15c2bb6e0b258126945429cbae71f9e2ec8","original_source_sha256":"b00cbf4705fcd94f9ea9130da79993f83b7e3a83313328314b3492d19295af94","decoded_pcm_sha256":"4d9f339e6dbd2372c553d9d5f6718849a37e2770b2e7e3a43beccd5922c086ec","excerpt_start_sample":0,"excerpt_samples":76160,"speaker":"librispeech_672","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff"},{"id":"librispeech_672-122797-0044","sample_rate":16000,"samples":59840,"source_file":"audio/librispeech_672-122797-0044.wav","source_file_sha256":"6ffc2e9415b90ff339b9b514b9e5487623d101240b622a32f00d0ba7c82e7203","original_source_sha256":"30dfd88ccc8b641d250263094e829de48032160f0e8f954c0487a8ecd0da85cc","decoded_pcm_sha256":"432f9a1ae86511f6d52017a356d8d76c16009ab41a0b0e212b50e3c7fe63423f","excerpt_start_sample":0,"excerpt_samples":59840,"speaker":"librispeech_672","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250"},{"id":"librispeech_672-122797-0056","sample_rate":16000,"samples":83600,"source_file":"audio/librispeech_672-122797-0056.wav","source_file_sha256":"90c7f0491bb4d2a989243e5c771a35b629c3f8d7dc4b3cd6389fb6305b86f9ff","original_source_sha256":"e58678d63f84c218fb28da622351b90e344380dbfc13763dd1123e71faf708d1","decoded_pcm_sha256":"1ee86311a9a7b6ee7bf0ef8ccad1d6c7f831fa6001565687e3d9e0eb90cd44f4","excerpt_start_sample":0,"excerpt_samples":83600,"speaker":"librispeech_672","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41"},{"id":"librispeech_6829-68769-0014","sample_rate":16000,"samples":58480,"source_file":"audio/librispeech_6829-68769-0014.wav","source_file_sha256":"78c56d595feff01ca9d811b7d4be231d4cf81e7f1063ed3d30f1c909a5c9b4cf","original_source_sha256":"c228d6cd297ffb4916b952d7d5861d8ac8225058219ec05b587d8dd4bd041179","decoded_pcm_sha256":"f8c838b07c2dd578b611a15c028e07c92a592c03766bbc5d9020f61b8ed5195e","excerpt_start_sample":0,"excerpt_samples":58480,"speaker":"librispeech_6829","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f"},{"id":"librispeech_6829-68769-0038","sample_rate":16000,"samples":40080,"source_file":"audio/librispeech_6829-68769-0038.wav","source_file_sha256":"5f50faccf72a597b600e7ecc50008cb3e9f8cdbd9741d8768e2d56e595e4d45b","original_source_sha256":"0f141cbbe137a2927bbf38106a3b14d2b2f186b86e1a54bcbd354d8c2a8981c4","decoded_pcm_sha256":"2491d28a0087423aef64b260b9a3bef96a9ffd38bdd5460aeb1d8818a5a4e0a8","excerpt_start_sample":0,"excerpt_samples":40080,"speaker":"librispeech_6829","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57"},{"id":"librispeech_6829-68769-0041","sample_rate":16000,"samples":42560,"source_file":"audio/librispeech_6829-68769-0041.wav","source_file_sha256":"091bdcd2fea02212fe7c7f184311a2fbf92fc8f284c50a4010b14938dba5cc5a","original_source_sha256":"e0e3964567d7a515d2443c6e1eff105f88da0e87ef44cc2954b36f1b9d93096c","decoded_pcm_sha256":"d14d8bc17d20f03134a1360b706d2d290d10240225382c5adf8b0919adb152e4","excerpt_start_sample":0,"excerpt_samples":42560,"speaker":"librispeech_6829","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821"},{"id":"librispeech_6930-75918-0013","sample_rate":16000,"samples":47040,"source_file":"audio/librispeech_6930-75918-0013.wav","source_file_sha256":"b49bcb9919d3803824dc69418b744b75c8c1474f36515b3ba592f73492fa7afe","original_source_sha256":"83f99c357f446d1fddffc61c09f59eb8a41c82c5e4261db42c71396c537c409e","decoded_pcm_sha256":"3f54b1539585ee3c865737cd0781dd64a40895a5d1930b99c37db78e450618a7","excerpt_start_sample":0,"excerpt_samples":47040,"speaker":"librispeech_6930","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064"},{"id":"librispeech_6930-76324-0024","sample_rate":16000,"samples":64800,"source_file":"audio/librispeech_6930-76324-0024.wav","source_file_sha256":"ced34d42f7fef433876666f6ac2b6200e6fd9f5181696ca8ee4577d66dfbfbce","original_source_sha256":"5d8bd7d094fe646881a4848026dbf1a3748ec90f170b87d2f0836adcfd9b7f7f","decoded_pcm_sha256":"d8dd1ce759e26c72605c90e837ae65e87caae1e3095da47a3f97063f55664b19","excerpt_start_sample":0,"excerpt_samples":64800,"speaker":"librispeech_6930","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef"},{"id":"librispeech_6930-81414-0026","sample_rate":16000,"samples":49200,"source_file":"audio/librispeech_6930-81414-0026.wav","source_file_sha256":"4278f90e4e39fd055bfe6f5968e8304ba8e18cacc7c7bb55083dc13c98a3c9d5","original_source_sha256":"84048bb63a1968e9856f38f430d293fd5036cc1b8c31312f3aff96d087732145","decoded_pcm_sha256":"6ea614ae159242a4cb64e926932b9c1cf0c8b84fb4912932e606a31b6d48c909","excerpt_start_sample":0,"excerpt_samples":49200,"speaker":"librispeech_6930","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50"},{"id":"librispeech_7021-79740-0002","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7021-79740-0002.wav","source_file_sha256":"4e9e43f7685d0916f4ad5813d805e7b03541d6e36bd194082be28518e086bcf8","original_source_sha256":"df09e0bfaf8000a8183b367c2fbc6ce3404f4152ef14d9340515b37edb7155ba","decoded_pcm_sha256":"32223cc598805a0301d5fd306db57f06a7d2c624ffce3ec523d6a45df0f1b192","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7021","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8"},{"id":"librispeech_7021-79740-0011","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7021-79740-0011.wav","source_file_sha256":"466e8b94d92d17e81ab0cf454452e8cfa849149b97191b6d988e488ca1533324","original_source_sha256":"af3d9a4d3df12f73bc4432b8be4762df32f7381a8c20009735ca8e982084c2a6","decoded_pcm_sha256":"6cda455d70e2130c6b8f8269b688ec0381f52c4c07376a3f0f4bbbf73ca24fc7","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7021","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317"},{"id":"librispeech_7021-85628-0025","sample_rate":16000,"samples":44400,"source_file":"audio/librispeech_7021-85628-0025.wav","source_file_sha256":"4d55ea3e61e7af0b05b9c98461786a6f20a32c9c58bd0dc59e5323e1f2537924","original_source_sha256":"477a753670424f1fd835f77fdd15e54570114ec1a4f99c206e555e94137688c6","decoded_pcm_sha256":"f8114fb1a86ed60fb5b4e5ba870142361984ad8304c63676d1f9c029942602ee","excerpt_start_sample":0,"excerpt_samples":44400,"speaker":"librispeech_7021","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4"},{"id":"librispeech_7127-75946-0002","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7127-75946-0002.wav","source_file_sha256":"bdb7ada34b81c81f3db2ce085123c35f2be08e981addd1afcf1ebf3a4bffbf50","original_source_sha256":"8a4077cfaf005d616cab7fe18bc2a434b627b24de056d78cafc33f09faa1cf6c","decoded_pcm_sha256":"93c54b2cc025d80c3059dc9d273712d4259e0ddbef769ef56e3f3d16805ad734","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7127","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd"},{"id":"librispeech_7127-75946-0029","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7127-75946-0029.wav","source_file_sha256":"4c5f2e0b5ab06e75c53a7dc8dc6a8bb6dccbec4239c8889e78850cdf0c2e82c0","original_source_sha256":"e26b6b1a8cceb2a05b3aae15719c785d4d24a2a09d91dd6292da8a4b4e32401a","decoded_pcm_sha256":"92db6e9b9692ce0f9a8c0ad4504635e4a6e649b6d8760fca3400f3ea75d63e54","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7127","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6"},{"id":"librispeech_7127-75947-0011","sample_rate":16000,"samples":57920,"source_file":"audio/librispeech_7127-75947-0011.wav","source_file_sha256":"a143bac430698a98e42d67116e56d11f8f153d739cf225647eaa8948c1054138","original_source_sha256":"e1bbd2d0c4d8800ab59fe4f4a1f027601dc96c787ad8bf50fd97e051270ee72b","decoded_pcm_sha256":"aebe57a65af629b724d8ca28eb599d83587cc213e6e303682e713453ab9de80a","excerpt_start_sample":0,"excerpt_samples":57920,"speaker":"librispeech_7127","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792"},{"id":"librispeech_7176-88083-0002","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7176-88083-0002.wav","source_file_sha256":"98d2f049c490b6b39660c19c901f040ad3343fe2be421980f6294bb3979cd7ec","original_source_sha256":"3c042266b649c153544bdb4b6c17ac6faf27e9a2b589a3d673225c7ea91180cf","decoded_pcm_sha256":"22dedd61331f17693761f502975e31bf01741aced72867d6ade55e38b2e09d33","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7176","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a"},{"id":"librispeech_7176-88083-0004","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7176-88083-0004.wav","source_file_sha256":"f53a216eaeb2c1b293abd9205bc2498d63ceeedad13e10b2632b8f49ef2bd2b0","original_source_sha256":"a0879f1d42b65a0bc10e3b725393bac049736d153e7a1c1b5d35e48bda4e3778","decoded_pcm_sha256":"43bbb6d389351da5d47cefcacfede3e848cb81497c672b951f8b533d7fbedfed","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7176","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4"},{"id":"librispeech_7176-88083-0026","sample_rate":16000,"samples":88480,"source_file":"audio/librispeech_7176-88083-0026.wav","source_file_sha256":"7491216ca57634802637da88e3a26e3fbbd7d7ce83bd4daf307c2633fdd90dda","original_source_sha256":"c637736d1dbd70e42d9ad3e319b273baf390cbc3f3ec8b31239ea7dca8e9d5d6","decoded_pcm_sha256":"bc9284b0dea002361e46bae8e07d7c2c20a9984ef0ce6f6e2c9fda86e70a579e","excerpt_start_sample":0,"excerpt_samples":88480,"speaker":"librispeech_7176","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325"},{"id":"librispeech_7729-102255-0028","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7729-102255-0028.wav","source_file_sha256":"9c4f547a596f4436fadb2010b0131456586372e80b9986b219330892c77f067f","original_source_sha256":"4654ddc035bcd2a1c3def2c04709f8f93eff6f9d9da968a1f044f65edf3a5b15","decoded_pcm_sha256":"7a4ab1cb22b23d73175ff25f3f2557bc2724e5cd7a3952f1cc5626d2a9bfbd3f","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7729","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b"},{"id":"librispeech_7729-102255-0038","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7729-102255-0038.wav","source_file_sha256":"45e58a4ed331fd014914979b00576d6c7d413f1ba510b936f7b5249b1f6f209b","original_source_sha256":"7bad471a79798adfc54e63acefeaf2a9579bc1f06779fd3aa49cc060df7c47f7","decoded_pcm_sha256":"4488be03955742e6e44fcdb33f0ae71643965b6765824a166c40ca31fbd9f66a","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7729","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e"},{"id":"librispeech_7729-102255-0042","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_7729-102255-0042.wav","source_file_sha256":"eeccaa551cfccdeaac8c6e94873f0209dc803be8608aba2884d6b7b6786d475a","original_source_sha256":"258a5f83b294976c312b375ef48ed79028fb41459458f6ece40b90fab7bce317","decoded_pcm_sha256":"3e552debe1c0d524847e6550ec6f2d9550fc271ad1a9ac47e3543e154db1f552","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_7729","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d"},{"id":"librispeech_8224-274381-0010","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8224-274381-0010.wav","source_file_sha256":"d24e32f261a5f138c2d6da847f9d524494b6e47ed6f61799de7255ca83a36af5","original_source_sha256":"e9f711f02a8f7cfe4527d6d16f4ad04dc95369de860a6c4f56a645dc7da5a02b","decoded_pcm_sha256":"34d9175c2a6517466fd56565f6c4abc5a8a75b16d60e1e96d0615d4383e19b19","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8224","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845"},{"id":"librispeech_8224-274381-0013","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8224-274381-0013.wav","source_file_sha256":"12076a3ca2c4e4f6ad74a06f495b7e3c8a9bd3dfcf902ce7ce10c7bc18c7d9ee","original_source_sha256":"cd0ba8a197cc480b016a6386531b3e1ee14c9ff6ddc85f19abde2abc4f00070a","decoded_pcm_sha256":"31d48338892252220a9e3042986f534b91479b75a07438f2a0515b00bc8fe055","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8224","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945"},{"id":"librispeech_8224-274384-0010","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8224-274384-0010.wav","source_file_sha256":"2435bb6510d6eae3f7a1d44177cdaef74143eccb6973b3c64d7410d790a92928","original_source_sha256":"677a2c48f5ddd5929e5c68e682b66998c48322833dac3baa41a086f87ef15077","decoded_pcm_sha256":"bfff7d3cf9d00e04b5e5f471ec3e7e23254f82646da0069c2f8d619c084a00f4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8224","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878"},{"id":"librispeech_8230-279154-0018","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8230-279154-0018.wav","source_file_sha256":"3cab56d8e446d70ddf0c42deac013fe8e42d57fc277e580ba2cb962175f7bfb4","original_source_sha256":"6fb6d05dd68dfe641a0f58b800e9d2e86662dab952355bc7573505bc35efb2fc","decoded_pcm_sha256":"2becab040601a06405a880e563913b87f3eb167e942e97f7225c55643949f142","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8230","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8"},{"id":"librispeech_8230-279154-0032","sample_rate":16000,"samples":62080,"source_file":"audio/librispeech_8230-279154-0032.wav","source_file_sha256":"db74628554756ee19b7ea21a4f8d96a3e84c6ec96261b0e608fbcc67c2aee9fa","original_source_sha256":"d60da130333b0e1adba450d171b7f753f423dce0f66a00448dfb036d43a7231d","decoded_pcm_sha256":"d59c4c75b338357fc4c79c9980c56076c2ae54441e0d223325fd8514435d4576","excerpt_start_sample":0,"excerpt_samples":62080,"speaker":"librispeech_8230","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6"},{"id":"librispeech_8230-279154-0033","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8230-279154-0033.wav","source_file_sha256":"5245398137201ec9e4fc43229b5036d433433699accfe32b9ac3d88c2907ab4a","original_source_sha256":"1bcf539698173ec1996fa39e5b4c5faf7a8ddf5b5f9636d269ca583befc17eb6","decoded_pcm_sha256":"bd287f891d96a4cd6b7a04337bf6be93b04db8cbec30940ae4945f8b4efde0f4","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8230","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed"},{"id":"librispeech_8455-210777-0010","sample_rate":16000,"samples":49600,"source_file":"audio/librispeech_8455-210777-0010.wav","source_file_sha256":"9dc0240032da656260916bd7364e269bea7181a2e7b41600dee73c8bbe1376e6","original_source_sha256":"28029242eb37055c0eb70761eebf5b7be98a5da6c30127b3b282bf7e11663a68","decoded_pcm_sha256":"10ec70ae1221ff8684ca0b1b6aa22fd7ae9343c0b701f90c315735ff27f3d5ad","excerpt_start_sample":0,"excerpt_samples":49600,"speaker":"librispeech_8455","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e"},{"id":"librispeech_8455-210777-0049","sample_rate":16000,"samples":65760,"source_file":"audio/librispeech_8455-210777-0049.wav","source_file_sha256":"d129eb1ea69ea47aef64c6ce3231befee5551b9993ce39b96c348fd6443e9d27","original_source_sha256":"2c4c3f4fa85e1e234b108aeada818530820e7a515dc7d91f2a2c075af7105fed","decoded_pcm_sha256":"6bab2226481307e41ffb714c528db253ee0755e0775a2890f32c07e3f5832d55","excerpt_start_sample":0,"excerpt_samples":65760,"speaker":"librispeech_8455","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e"},{"id":"librispeech_8455-210777-0060","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8455-210777-0060.wav","source_file_sha256":"f49c9fe046b2c43f52537d1784414f5f81b0a8915ae11f37840a512c004859eb","original_source_sha256":"ef57db07976e02027d19da092bac4747c7928a93199441010b967c12ce13e724","decoded_pcm_sha256":"71c64b069cc8f43d09cfadd1defcda3beccfdf73919c8518a13182fa2c801f93","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8455","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a"},{"id":"librispeech_8463-294828-0003","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8463-294828-0003.wav","source_file_sha256":"7659348492b7d95316e0272d6e19529c0145bc4669eeb7e5cbe4df68386c4cac","original_source_sha256":"3790b43d50403978fd6ba4d90f598e8e6afddfb630a3bf5b4110bd432517f1e9","decoded_pcm_sha256":"cd223f40e942ab9ac205813cbe938225779742060f5fe247eb3e2fd12df52eba","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8463","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b"},{"id":"librispeech_8463-294828-0014","sample_rate":16000,"samples":91600,"source_file":"audio/librispeech_8463-294828-0014.wav","source_file_sha256":"b5bd05a6e7eea30e14a01e15f2980b0a9712c84d9cdef9665a938e20e751e9cc","original_source_sha256":"b61f11aa0c2b3d1adca49e6199328637c6ecc84c003dab03fc8c97b95cab4c63","decoded_pcm_sha256":"c608aedb889e9a3c3821520a449396c5b4cd1264e1d081a38be40b0b66f8b285","excerpt_start_sample":0,"excerpt_samples":91600,"speaker":"librispeech_8463","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412"},{"id":"librispeech_8463-294828-0022","sample_rate":16000,"samples":37680,"source_file":"audio/librispeech_8463-294828-0022.wav","source_file_sha256":"10cf3ff66d8250a6fec86a1f43780412a98a30f4c52d69014e7c78f858135bf9","original_source_sha256":"4b0b7f62860ac9f04945b08f79183b52c26f80eedd14ff829e2ff2aaa5143e3f","decoded_pcm_sha256":"276e5d5f0e0b54af2b651a409e0f7fbcd7b42508e69037a0becabc956b358b07","excerpt_start_sample":0,"excerpt_samples":37680,"speaker":"librispeech_8463","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f"},{"id":"librispeech_8555-284447-0007","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8555-284447-0007.wav","source_file_sha256":"442dbe50a41ace8dc2a58b2b8b989591e9fc77071276186f3375a3e69fa9dee0","original_source_sha256":"73d10187e60efe5f030a76fa33d1eab2abae7e38817b356dd61e30b743edf22f","decoded_pcm_sha256":"b2e9b095c084b8ff6903b45eed724068bad7267cf8c1f692479d68f2b6adbaed","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8555","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a"},{"id":"librispeech_8555-284449-0020","sample_rate":16000,"samples":81520,"source_file":"audio/librispeech_8555-284449-0020.wav","source_file_sha256":"5a52c84f8ee2291235ad5ecfe444691a1c6544fb65cb9659631463a9c337713c","original_source_sha256":"5703c01afc67c6eb161bd9fe7fe92f5b373e9331629f60d1c163e5d893c97d7a","decoded_pcm_sha256":"fea5f67e0d192edbb42257fa9560e233f643cc196059371f1de72f17987367e7","excerpt_start_sample":0,"excerpt_samples":81520,"speaker":"librispeech_8555","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd"},{"id":"librispeech_8555-292519-0001","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_8555-292519-0001.wav","source_file_sha256":"81565a433405c276e6e739ab2d05024a761fb5113ba469c02b6781d81896b6a2","original_source_sha256":"b082de228dbad73be95d916e5149c02035758a390482845def17215c0eed0fa9","decoded_pcm_sha256":"c2e62e46b5a18b9f8f30423189b31cc00703ba64084bb8bca7ffe75ae8d6977a","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_8555","language":"en","style":"normal","speaker_gender_as_distributed":"F","pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af"},{"id":"librispeech_908-157963-0018","sample_rate":16000,"samples":68080,"source_file":"audio/librispeech_908-157963-0018.wav","source_file_sha256":"8b673b841c10fd96664240ac3016693e75c85c78eedc9b403d8b3595a9872c3e","original_source_sha256":"53efff3282f4691482a4e735887fab9130c78ec10f3bdb79bf68f8c212898134","decoded_pcm_sha256":"8ae8f9234bac8c63b967d3c7b5075657f9771ac253a59a383e523ecff9c3b006","excerpt_start_sample":0,"excerpt_samples":68080,"speaker":"librispeech_908","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412"},{"id":"librispeech_908-31957-0002","sample_rate":16000,"samples":76640,"source_file":"audio/librispeech_908-31957-0002.wav","source_file_sha256":"15ab98084665c3d06e930fc929c6d7004dcc07b27ea5bc18150529184f9f1ca7","original_source_sha256":"77ad707e976d262395c265cdfda5b9a3da7439f329aac0f5216a404acefe3fe7","decoded_pcm_sha256":"5e0a5e522c82eea67f456aec58b62300d4c743b9610bb9f4f869f3b8bd8e7f17","excerpt_start_sample":0,"excerpt_samples":76640,"speaker":"librispeech_908","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4"},{"id":"librispeech_908-31957-0016","sample_rate":16000,"samples":96000,"source_file":"audio/librispeech_908-31957-0016.wav","source_file_sha256":"cb3776a40852c01ee89c57b7b4197b532c19c4341313a5294de44074055e2d25","original_source_sha256":"2a57896e33e6a87debcb741f00196495aa665bb2f0d7c205ac980a8417976d12","decoded_pcm_sha256":"a1a0fc0c8f705774ce9810918acda74ca07e00f2107a855ab611aa7ee1735b94","excerpt_start_sample":0,"excerpt_samples":96000,"speaker":"librispeech_908","language":"en","style":"normal","speaker_gender_as_distributed":"M","pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c"}],"scope":"fixed_corpus","seed":0,"threads":1,"device":"cpu","mos_requested":true,"source_hashes":{"BASELINE.json":"da49356e381abd9a4538ef48a578b5474373efa1a9618d62e74e08fa45b302f4","uv.lock":"c8b2bc28ebb1523bf8cfe0f4627d06bd5d03e0ba6a2bb5dc0905be77a3dc8823","vocoderbench/__init__.py":"73738d70706824f85fdfac302306af505437a50b0ca4cdc68fd57aec211a308d","vocoderbench/audio_report.py":"f46ff0cf69e7cb956ea94ee0c7013d44052c0400c5ffc64d7f9f90b4489aa980","vocoderbench/audio_samples.py":"b9b975c1f305cdaff47909b99bd2a3dce64996e90774d2a8305adc1c67ed6c17","vocoderbench/benchmark.py":"7fdb6b4dfa4f9cf64bc8d009dd4b0b1e267ab18a5e87a9e1a7004ae45d84a244","vocoderbench/benchmark_io.py":"d2fa3e56f52da225573939c92056005be47568c531b7a317a5e893e10a5d9d5c","vocoderbench/benchmark_pipeline.py":"a049034dcd272bbafda47cbf5719cc6b1cd8dae6f80682200cd3fc7b90e522cd","vocoderbench/benchmark_report.py":"2a8061e8acc89dc4897d77efa206922720d6aa9843c340daea8ec78c2cb0b759","vocoderbench/benchmark_summary.py":"8221086fe08bfc70cbfb9308a3a01d6fea976afb7de86af8cbacd7999e8281da","vocoderbench/build_world.py":"7d29d81c9a48d633edfeae0425c5cc055bc0a10d230232a51eab5720b2cc6974","vocoderbench/calibrate_pitch.py":"282c917bf87c6f549f32467647802852907e4b0b88128eb8017157b84566adec","vocoderbench/control.py":"bc41e73ff34b5f9cb52280a6f2317c989abc53f8b15f8f82db5ee3102b4b58bd","vocoderbench/corpus.py":"bc14cee2c3dd1a651cf2e5f0e9fe871fbd3a712d1a90d7c18a18547a625802e8","vocoderbench/extended_metrics.py":"ab3537d2c8e5d568a0cd2fa8aa36309de09c3bce7cf2f849214331a22fe3624f","vocoderbench/fixtures.py":"2b2ba922ed611abc91a29f1d270c9a91c0743bc9a8a6a131c7dcd184fcf60dc9","vocoderbench/hnm.py":"58f0378793b7433f80bf854972882298649799d8e3d0868674816e9bf6b63f74","vocoderbench/hnm_analysis.py":"3f45703667237333d76c4145c1513e4eb3342bd40f01a714c10303082440e26a","vocoderbench/hnm_control.py":"b2f5efbbb1e550569c231a08f12997d94b0bacb130bcc816fe6bd458b17a98bc","vocoderbench/hnm_diagnostics.py":"076edd5a785e83540a0cf55fea06075769ef35503f210eb977efc1703dd0081e","vocoderbench/hnm_parameters.py":"635bf222bac8ce41897c0977677e67be4366724f4239a9e85e153169fcea2cf0","vocoderbench/hnm_synthesis.py":"079057d2fba2b4ca0d189e050449aa7bc3fa20033afb0b03893300950ae13593","vocoderbench/logging_utils.py":"c80eb5d9647516a4bc03980cb178f8b4bb130ac15877d1b3289797e7ad81e381","vocoderbench/methods.py":"99a040b960139aa842a9a46623b34717db9c71195df21ae969fb2a9e9b9155b0","vocoderbench/metrics.py":"b8767feb49494058bd9ec08fd95ae2fb50ce7897fcdd6cebe319ff8a0ac51f22","vocoderbench/mos.py":"fb632aa5440245525289962372db4f3eb007cd2b3b493b2d05ecccb878e85f03","vocoderbench/mos_batch.py":"c7a732dfe52b08fa155201a7134d2f5561a10b16d1866e9f0526f61b2b244f64","vocoderbench/noise_screen.py":"8eb38cfb49b315dd09f394a1a6753cd416704442b78051096c168da70bdc41f1","vocoderbench/noise_stability.py":"624dc434add0086c53c41b9fb7fd0e65b2c51b16c08d4485e4147a77f5534213","vocoderbench/reference.py":"345fe4c8ae0fb107f01e7ce973daaccdc5e60e1624753de61928b3718b23672f","vocoderbench/references/__init__.py":"d96022954c3cd493c7e6d7baf35e8bd33eafea192a9f9154c8470b7937d05732","vocoderbench/references/adapters_dsp.py":"ab5a0cf63085524842a0614d2ebc63adbe11521852f53bf9d561a131c37ab285","vocoderbench/references/adapters_neural.py":"cfe845d99cf0257556da99e5a1f950c9d9ac5ac490fb7d8e8cfb9b46f2c03a08","vocoderbench/references/adapters_source.py":"28b70e6680a3c8090fa5d16f696eae5c28d6b55bee5eee9e90ef4b80735f2b99","vocoderbench/references/catalog.py":"ff03fb9f0de8ccd11d48d76773611cebbed0c6f05e45e8ff8a31f6b108c09e81","vocoderbench/references/common.py":"2d8f4418414b1abf0cfd7f15e00093b3ba1e9c637ddcbb7aef2c965d0f6e5579","vocoderbench/references/current_base.py":"8a9161607166ab93378f6957a7fc39011e01689076f5e53094d91e7856849704","vocoderbench/references/evaluate.py":"f37d09a45d8f801fb68d2e9eef31c0c2aca421a64bfdc42be4d233ccc815f917","vocoderbench/references/evaluation_cache.py":"4ff288c1f4f839db9004960333e1692c47f3e28cc19b2c8781a73e65cf453e0a","vocoderbench/references/inputs.py":"11acdfe0b4611075f3d232161008f098f59daf2a9345d7e1e75a4e1680ada1c8","vocoderbench/references/integration.py":"d64e036964d4b6ef782d69b927d5d3b1ca62aedca22e2979255d0eea861a1930","vocoderbench/references/native.py":"a6581f389587b85f547ba66d068ce974d8fe6d08252500f0f28788a7014780fb","vocoderbench/references/prepare.py":"82092abb7f8e950b0c09b08568580627e6696796d054a716824e038ed80699d4","vocoderbench/references/publish.py":"72c2dd1dd0251cd4a623f9724610d581547298ba2e6105e0bd7a99a6957319b2","vocoderbench/references/report.py":"011afb9c6453d9b215f7bc687d9e2dbafbdaddfdd9864e8925ce01ab9cb88fcc","vocoderbench/references/report_formal.py":"3eabc1892e804ff05e11c93025616cb530e4eac91eda92660058044701c6132b","vocoderbench/references/report_metrics.py":"f4e3e1424b091433d34aae77383cbb51c6063e7803f3c5381375a6fd8b5f33f3","vocoderbench/references/report_summary.py":"c92f65aca61793b5907470dacd31ef0afb2d6ca310878417b3e4ce0beb5c2bd9","vocoderbench/references/requirements.txt":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","vocoderbench/references/sources.lock.json":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","vocoderbench/references/worker.py":"5705f551bcd4eec6a44f204e3089b0cf4ddfde3408b613d7a6b2e1917e3a0115","vocoderbench/report.py":"659d8406e8c1d09e29e120d6f0ef13ddc3047565915c125588a2704fb25c6737","vocoderbench/run.py":"ac2b6c3a90dea0b9b563ddf30c94e58b8cd9b2e16fb368546ad4efd206a729a6","vocoderbench/site_summary.py":"a33996e31664281ccfc556b944b9013856369c4e1fdf1ff88b1af86f615ae91f","vocoderbench/supplement.py":"8b72b63b7059f3e94ecb13f575ab93060e00fee0cb89c5abcaa51719061b890d","vocoderbench/supplement_contract.py":"748bff65db3bd3709cdd472b02358af35518c5d1a96fbf212f6b02ba5fa45cd4","vocoderbench/supplement_report.py":"9146b635745c6bb10c733c090696c564d463180507809588ac377ef682c668ad","vocoderbench/temporal_metrics.py":"ce21b6a4767071a7b98ea0cdbfe54bdba15b596bf8d6b6127281a97095e7be91","vocoderbench/world.lock.json":"44b8bc3adb62802ed060834d9e06ed7f5f3cc4d28583b8a44066c5959a42462b","vocoderbench/world.py":"d14558eed772e07031a9183f57facc55acb1c27deccb3dd4fa098ffcbe0cde3a"},"prepared_methods":{"hnm":{"baseline":{"status":"selected","method":"hnm","version":"hnm-v3","reason":"雑音PSDの台形積分とFFT端点配分を揃える候補Dを採用。固定120録音で帯域・平均包絡・音量誤差と予測MOSが小幅改善し、制御と保存互換を確認。明瞭性・スペクトル誤差の小さな系統的悪化を許容した開発上の選択で、知覚品質の優越は未確証。","comparison_reference":"hnm","comparison_reference_meaning":"実験開始時点の唯一の自作ベース。過去結果は当時の版を保持する。既存比較対象はWORLD（Harvest）。","previous_version":"hnm-v2","selection_report":"experiments/2026-10-12/noise-stability/results/run-001/index.html"},"source_hashes":{"BASELINE.json":"da49356e381abd9a4538ef48a578b5474373efa1a9618d62e74e08fa45b302f4","vocoderbench/hnm.py":"58f0378793b7433f80bf854972882298649799d8e3d0868674816e9bf6b63f74","vocoderbench/hnm_analysis.py":"3f45703667237333d76c4145c1513e4eb3342bd40f01a714c10303082440e26a","vocoderbench/hnm_control.py":"b2f5efbbb1e550569c231a08f12997d94b0bacb130bcc816fe6bd458b17a98bc","vocoderbench/hnm_diagnostics.py":"076edd5a785e83540a0cf55fea06075769ef35503f210eb977efc1703dd0081e","vocoderbench/hnm_parameters.py":"635bf222bac8ce41897c0977677e67be4366724f4239a9e85e153169fcea2cf0","vocoderbench/hnm_synthesis.py":"079057d2fba2b4ca0d189e050449aa7bc3fa20033afb0b03893300950ae13593"},"python":"3.12.14","packages":{"numpy":"2.3.5","scipy":"1.17.0","torch":"2.8.0+cpu"},"noise_seeds":[0],"role":"current_development_base_not_new_candidate"},"world_harvest":{"schema_version":1,"method":"world_harvest","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"../../vocoderbench/.cache/world/c794856b537758a3d300757db2e5e31685a822c2/libworld.so":"27f782b0194b1017cc6d828a2729fd6faf11acb405dc22b7c8b7ab5344642d4e"},"sources":{},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"facodec":{"schema_version":1,"method":"facodec","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/naturalspeech3_facodec/ns3_facodec_encoder.bin":"c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","models/naturalspeech3_facodec/ns3_facodec_decoder.bin":"4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536","src/facodec/ns3_codec/__init__.py":"6b86c557f111bb6c0c1c43fe77bf34ac16ad973411d652fda63449f112cb24c5","src/facodec/ns3_codec/alias_free_torch/__init__.py":"7e8bdb5ab07a79de67d1490600964f60b0a995c6691217f4fcfb76e175a0b79a","src/facodec/ns3_codec/alias_free_torch/act.py":"2add4414a8dec8b5ffe55e55dc26c17aeba31224b5a5fb3959cc82804454b558","src/facodec/ns3_codec/alias_free_torch/filter.py":"674e5fafffd4bc46b58a7de3590b3bc6c2228930e70b50f4ceea014367438dce","src/facodec/ns3_codec/alias_free_torch/resample.py":"9341fe6756d3c8222ecbd4155366ee12aefab8c98ab0cd9badc34a0c6d7eb755","src/facodec/ns3_codec/facodec.py":"086962bbe3b0913491bf1d5874b9005afc8f3a6b7f2707f385e5711f0df84140","src/facodec/ns3_codec/gradient_reversal.py":"8d727bbb012ee5275ad4c823ac747a86db88a69faf393b04b9a46c7408c85ebf","src/facodec/ns3_codec/melspec.py":"6a14da2a4d531f1e8fdb3f83398bfb80c379f5fa41162a28742d6b9dd31b0c3f","src/facodec/ns3_codec/quantize/__init__.py":"69964f5c0ecf93a00b2b4767cc2952cad19e7eebbc7bbd154214c01fc34c2943","src/facodec/ns3_codec/quantize/fvq.py":"40a1efbbcedd5b09fff3db1d91d831114dba49a63d4f58215892d3fe5dbf1c10","src/facodec/ns3_codec/quantize/rvq.py":"e761c3b3f6130e91546830ff4308111ebd23a9e9a7c7443204a358399d98cfd4","src/facodec/ns3_codec/transformer.py":"1cf20b2262cd3dd754b4884f9a134ced9b0ce416332382c00e75b41ef6401f12","src/facodec/setup.py":"1cc676afa26edc2e6d2d0babf2864f787f8940e7d510f74e537f21b45bd1b0e9","src/facodec/test.py":"a4e50e05663a335615c07a93e0586d36a64eaa7fe8cccd317ac95caea882d90a"},"sources":{"facodec":{"repository":"https://github.com/lifeiteng/naturalspeech3_facodec.git","revision":"aac16b7268b1b96e70b9e3526b13e216984e80d0"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"hifiglot":{"schema_version":1,"method":"hifiglot","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"golf":{"schema_version":1,"method":"golf","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"src/golf/ckpts/interspeech24/golf-ff/checkpoints/epoch=539-step=996840-val_loss=3.073.ckpt":"1eba39620c64d1c59b0a6acb4149200c4791f8e22ec0435bf13bfe5fd86468ab","src/golf/autoencode.py":"2a2f12303a23d09fd44d94bd45054709439c4d109d7f5b28c57d1957b9a06025","src/golf/biquads.py":"adec860a081743399fcba8852d6f7292c26d6716ba6990d047bbe829c8854da0","src/golf/cfg/ae/decoder/ddsp.yaml":"9e8eb0a9b276db7bb8fc72fba4abb34758c708c76e2632eeb5b93bed7fc26961","src/golf/cfg/ae/decoder/golf-precise.yaml":"327e2a1c08e689185e58f6c138697470a4b52fb1254a0bc3d8568dfdb4e2c449","src/golf/cfg/ae/decoder/golf-v1.yaml":"e1222ec194a949fc6658de1eff9c2a972777f97954703368d0cd768bef64a60c","src/golf/cfg/ae/decoder/golf.yaml":"a0f4938cafabf211af968a3a39ad2f0588d6831d619924c10893a39fe55e3d60","src/golf/cfg/ae/decoder/mlsa-taylor.yaml":"7d965b156e559c9e59f3bdf1ee0bda68ca30c7aa29d1a04a162cf2bb845e17da","src/golf/cfg/ae/decoder/mlsa.yaml":"11b3aeaf4ad4e5b0763f5447bda495c60da606c1cba8cc6724da8e0157d9a98a","src/golf/cfg/ae/decoder/nhv.yaml":"e0c9b88eb3f3c1847e79763a91d35d0de336160874c922383f7b9b9c2e5d36bf","src/golf/cfg/ae/decoder/world.yaml":"96354d851ba54bca2da610bf070be6bad34957ab287a42e2e353ea00a777118f","src/golf/cfg/ae/pyworld.yaml":"3030b4fab5766c9cea1678eb61ecfeaac94f6ceeeb999c80d0d384221eb00103","src/golf/cfg/ae/vctk.yaml":"d60df8ffc9b089985f900e650e47c1647d7565c5f1b10fcd6b677012dd5d43c2","src/golf/ckpts/interspeech24/ddsp/config.yaml":"0191313884d522ba81fec1f5283c05785995beeae7d3ddc0aa1e15ebfcbec992","src/golf/ckpts/interspeech24/golf-ff/config.yaml":"a80c16801d77ac2b7541621fe0b68839fb0894988c3725d4b5f6fbb5f57c566d","src/golf/ckpts/interspeech24/golf-ss/config.yaml":"a73e5acc91035b294703cbbeaf03c458a6a31533841732c1853ee25592cef498","src/golf/ckpts/interspeech24/golf-v1/config.yaml":"b115807cb4641e5c189d345f2fff6923ca354de44a0e0b52656dd0176e2d964a","src/golf/ckpts/interspeech24/mlsa/config.yaml":"09894c5ff25733047ee6f2e9d3adc73e5dd9c7cb26d7e6bc05f980e9ec871bc8","src/golf/ckpts/interspeech24/nhv/config.yaml":"a892ce1923fce68ab98d0f3e4875bbeff00f8e6be2d30f223d29baa1a8963930","src/golf/ckpts/interspeech24/world/config.yaml":"5254514fe16427e33d722a30a8280858f3bcfd3c1fa822b4b49dd1968805127c","src/golf/ckpts/ismir23/ddsp_f1/config.yaml":"f6ea30c95a2648e0c699d9e0165fac116bd46fd6c73a419f2f347224321aee22","src/golf/ckpts/ismir23/ddsp_m1/config.yaml":"424d3a7608b88d32b2276de288adf2775c32c8ffacfac9b3babfdea82be7423f","src/golf/ckpts/ismir23/glottal_d_f1/config.yaml":"2b085ec1ad9a544c582267d2621c2ecd908969d0557805f81b8a37c817141824","src/golf/ckpts/ismir23/glottal_d_m1/config.yaml":"496ecf006e2ac4601f1d7b4db539c28583f186a014b4d44cefb0dcf7a560a3bb","src/golf/ckpts/ismir23/pulse_f1/config.yaml":"9206730b004983bef907598f4f1c465fda3f72337df2ba4d1b50ca50732ef7a8","src/golf/ckpts/ismir23/pulse_m1/config.yaml":"62da88b5d6fdfa25e330608dbdd34f77b78a923a4e224c711c146aa400acf20a","src/golf/ckpts/ismir23/sawsing_f1/config.yaml":"adc7d851f4d5dc4f3aa66e386484a4490770108801962cea7e8409447192f223","src/golf/ckpts/ismir23/sawsing_m1/config.yaml":"e70a279a8fa5c3b674773ea87658c9d165ef1d0c8fc5521b55367089e6014ddf","src/golf/convert2v2.py":"7725e1d2751f49f9dc48b425c2016d5cd35e63c93022f149e05d490b3a2c30b7","src/golf/datasets/__init__.py":"32517fc25fc15e7bbb19506c56da5587e5a94035749dd86471a68d41ebec10a2","src/golf/datasets/autoregressive_wav.py":"9ee9a500c467d5dc8a8f1b86868e15552128a7ef3f36edc3e258e4ecc83ab63b","src/golf/datasets/fixed_sr_random_wav.py":"9b27a411235103c1c6301c3088c426e8fc89a3a980c09ae23613896f91f4af08","src/golf/datasets/mir1k.py":"5658b61b63a2df7b1b13a14a073332dc68f3ec5909b602a1b110dd3fb4708d3c","src/golf/datasets/mpop600.py":"da9bd7ecab718b33b65f6c103d55022dd42156e4d20723c1cdde09c945da008f","src/golf/datasets/random_wav.py":"b7818dec321c9bcd6f35fd06327a3fc7ee90745dda4cb447eb0c6fc2e548a8e2","src/golf/datasets/speaker.py":"77ff1ea3468d1778657f5e194d08a7503e121bb802b6d5d6a9de154b6eb25dc6","src/golf/datasets/wav.py":"81187f5e3a20fd00fded9d98f408ecbd4adea14beea01418d4ca039f0c073629","src/golf/eval_pesq.py":"a45fdf7e30ce8693c32871a9c77dc74a97abe59583a2b52bc793886ed074c3e9","src/golf/fad.py":"b94b4e9491e710e23f853edff1c18c372303daed65c37aca0fb6c3ee8a27dec4","src/golf/harm_and_noise.py":"04d4a00566e936f7fcd19f24a9cfccc07990531b71fbd630833b90504efdaead","src/golf/loss/spec.py":"5c1273c069342c4883e299fd3641f0178b327e90ec6d80d390d0ad4accdda722","src/golf/ltng/ae.py":"ee35f49f9d13d6772fd74cf4b150dad2ef10c0b7cd8237551e87a1d675195fb5","src/golf/ltng/cli.py":"6c42cfc3f43c54607572ede66cfa6bb2229cc125c81910bb405cd95cf9c853cb","src/golf/ltng/data.py":"16e85b554969b06d2d7b7549f81b80abe09e5574fdf1859cb1dc4ae6f610aadc","src/golf/ltng/lpcnet.py":"0a17284f0426cc65e80e0670f35ed22cf2ecb5423a8d5759ee93f08c6106f606","src/golf/ltng/vocoder.py":"ed7e3a3074803d89588a25b47b3799084940471839c3446389b7de2f427a100f","src/golf/ltng/world_ae.py":"70e2849a6fa26d4daf44e4958125572bf3aac6e4b5486ab8e8899919b61e12a7","src/golf/main.py":"500b8944c1331a49ef4ce72e103fa8d1e9b965d93df9147b35432cd7724757dd","src/golf/models/audiotensor/__init__.py":"fab3af586887478666f1e1454b8d0a46cfe805dfada05641e0e636b3540401d4","src/golf/models/audiotensor/audiotensor.py":"1ac461dd57554457f325939814ec1f6120d9b84cbd1e576ae450a11f5a7527c2","src/golf/models/crepe.py":"6118fb1b3f43cbadffb6db0502d103f9ad714ec9e0654780dc76a988c2fb34d6","src/golf/models/ctrl.py":"35ca67a52299b1452c741ee197ac3449d2363d02f2ee5aad54da3c8906c6a11c","src/golf/models/enc.py":"288c3729e88f64a16b38018bf19ba5ab9906edbc58e5b43909b68a7064fc2d47","src/golf/models/filters.py":"6414a4ef443cf147d4d30fd3a5eead58bef1d9c399fbed512e31eec9b4701bb7","src/golf/models/hpn.py":"62aca27a29790f6f158bbf66f8bb319ba80424edc0b669469c1a7734346885d5","src/golf/models/lpc.py":"6c72b47e460ca610ad8d2ff2fbf62f9aecef95ff54325b75bdf181086cdaf839","src/golf/models/lpcnet.py":"f992f99d5fe6bb000f08cc391457a8899b7d9fabec28f2935585d2caa1dc5dd4","src/golf/models/lru/__init__.py":"900fbf7b8089409a4358dbf34522d82f7fa4e38d06c17738269d8f09a5d3ced3","src/golf/models/lru/lru.py":"295fadd12f67eb8e900952a217ebb8f9fe6d86e17c2cf37a5943a66d0bb90e07","src/golf/models/lru/parallel_scan.py":"259ebf28f6496ee5fa22174a39b3e3e16bd35ec4bae53ba2fa3f0b28a503d843","src/golf/models/lru/recurrence.py":"5350766b079db0fb8deb3d117ca31777f109930ca9b4a464572ebc4d10c9707e","src/golf/models/lru/test.py":"eb03bde88a02f3b0291a548fcc1d6590af7522c273724e907ec97f5f2337bb6a","src/golf/models/mel.py":"0ea64cf409a11a5490a73e8691e351b2a1d9aec8649ccb418f64f1db0523871b","src/golf/models/noise.py":"1449921f428a194eb6efbce2495875532c388b1b6db7cfe220230b7ea3475a67","src/golf/models/sf.py":"a72c4398df77042f0e1061e4014fa1460df3ee608852491e74d8c9658277f17d","src/golf/models/synth.py":"25dfc2ee890e477be13af55419e0671ead5a7b199b2bfb7b3b074d294e926c10","src/golf/models/tspn.py":"ab26963db50f47f05f9aef55eae8ff98b83511764b12bc4b6598afcab929deb5","src/golf/models/unet.py":"c645b21f802328f366a4d0bb0f3b8f7e854294158cd8a47e2473605124c50c2d","src/golf/models/utils.py":"e11d97b27d42504d31974a1404fc6bde58a71af89b5024e7c91f10f5fd9ea1e6","src/golf/scripts/resample_dir.py":"a7154048c5ea19c0138dfc8077a5fc45f2f009b11061e049ca0fc7a727350974","src/golf/scripts/wav2f0.py":"f86aedc473ba6ceecf856798b3163c9f8cf773f727214ac71abe8328e4538467","src/golf/test_rtf.py":"a7bef08d57a5942b4fc67cabf21ce46b2b749ec53983906121ebd93e15c32ea6","src/golf/tests/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/golf/tests/test_glottal.py":"3ac3a12e0d875d748ff6207b5d88b9b65b5abad8bdfa029d57abef20d71f3cb3","src/golf/tests/test_time_tensor.py":"253bf48388fd00fa0f97d3cf3d869263b99adbf344e33d5f960601a1e56e5449"},"sources":{"golf":{"repository":"https://github.com/iamycy/golf.git","revision":"dfe4f4628b8d59d05c9988ae9341a544b1ff30c4"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"sifigan":{"schema_version":1,"method":"sifigan","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"llsm2":{"schema_version":1,"method":"llsm2","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"src/llsm/buffer.h":"cbbc9b886bd7f20176e6f5798bf357fd18b54f1b9596a96facd07bbd0c9734d7","src/llsm/coder.c":"870cbbff5d99547ad8def1bac8615319b113d84eea78c4db81b151204c887aea","src/llsm/constants.h":"e17e4f921753f9608e79a118e06f5a433ed6579e69344aa852f2b042423c4c8d","src/llsm/container.c":"0064ed944ea38439e9cbad3cb142f61ea3bed09804d819f0b4c9d7116b515ff3","src/llsm/dsputils.c":"00bfef0a8b55a7619ccc88fc3c852da393e908316b3efec3f760465eb7cdb698","src/llsm/dsputils.h":"8994e39a248784cc3e68c79b1ad231a61e020cfd246332f0a8a921753ec88a08","src/llsm/filter-coef.h":"bc7e3c6f2164c8cd4effa881e0fd24deea67d8283efaac6f6f2e0b63b7d60ca0","src/llsm/frame.c":"431139f767172a8f1707ac12ca6e11a6bbf328ed120016d0de18f1df09e3a205","src/llsm/layer0.c":"40e44ae44e7b37a2d66e50c2320794f0fbacf9c4df78034194d9a48c299b2dee","src/llsm/layer1.c":"e8a5c2df95e3c240427f1300886e8b0d129051f7d0e6dc7e2f66651a40bc0d1e","src/llsm/llsm.h":"ebbe38a332b94f4d5dcf9fa0170cfbd76190bb9df049364e95990ea122079bed","src/llsm/llsmrt.c":"53a6db74429167f4ae7af6b34c205737181c971f3d19c643065a01695f742136","src/llsm/llsmrt.h":"82ed9d3dbc970ad766ffc1d5f40801b6f0233add8a096f4b7a4250ee1d932f08","src/llsm/llsmutils.c":"853ef4972b3eb9f01ab8afa919455de652fa89a2b64e3ab3e166dd1e5f3bc3ba","src/llsm/llsmutils.h":"f8b002d50e21bc4d52af2bfffeb724a00c9a6d53f8943268e947e9b095e00820","src/llsm/test/demo-stretch.c":"7cd8ead474c73cbdf9281d9ddbbece49e4385d50b0d8b1d4461e48b50a6562fe","src/llsm/test/test-coder.c":"753a8036ddfec1dbe990e0a28b635dfdabe23bb887bd43c46f6415f2073791ce","src/llsm/test/test-dsputils.c":"d5282947347a2d91f5b686b74d646993fc1c85fa23c105fb50ec006ed3bda655","src/llsm/test/test-harmonic.c":"b0ee2f7e19175da2bdb673cd0c14046e902749792501975fef73a630956ea294","src/llsm/test/test-layer0-anasynth.c":"d55fbeb221460afdd7733804af5afc562f5509215457a0a7f890669c7d6cecb3","src/llsm/test/test-layer0-edgecase.c":"2ea95422798bdd4f4762f821459581ffbd2ee657f306311e35e1243ab4245507","src/llsm/test/test-layer1-anasynth.c":"0c0a86b673d8910edf9258c7889bfa75f9ec4e58947c0cada0efc8df082d85d3","src/llsm/test/test-llsmrt.c":"ef3ef6d92cc49e69f362231b94090af6f613bd724d5547817be41e9bbf9d07bc","src/llsm/test/test-pbpeffects.c":"cf00074e01b9fcd132bc618763baea5fe1ce92c22fcbdea60c90b1536e11e7d6","src/llsm/test/test-structs.c":"ff8bc187136ac6ab1a70523b6c24f45399287241d1a470e236bd9eda7de12c63","src/llsm/test/verify-utils.h":"541e7cbad0c9bdc5f2541014b2cd46d97d0204d6907277676a2d6f9dc06d6625","src/ciglet/ciglet.c":"41d3f5b84dbe32f447675756007cac6ba6fe836cfbf9b6d122b393d47fd77b46","src/ciglet/ciglet.h":"1290f7e14e6d0fc2f1849f31f75ef610ae7dd2abdf1e1f470ac93b58bf56331e","src/ciglet/external/fast_median.c":"08e803cea8cec1e0d8a36a4df4224650cf667260277829c98f2eaa184f89c596","src/ciglet/external/fastapprox-all.h":"568b4b09def0f5539b3487ed01a9f616955fb8f11b4d90c1c7ffc0f34c9fc543","src/ciglet/external/fftsg_h.c":"bfb6457551b5d071e865bff6b7998635040e9fd6fea9d45891553e3ac9013678","src/ciglet/external/wavfile.c":"c791eaa06d3b626224a2c97356f41177fb8637064c4856d912443431afeedaa4","src/ciglet/test/test-fft.c":"36c41142f5e28e896806132c3f29487f335ef59f1048af737ac3268f9f6668a5","src/ciglet/test/test.c":"ce19053bca09affb155bce3b8c38f3f46acfbcaf22922f509a34a85239a5c8fd","native/build.json":"782908e15d75aaad6ab792066e1722a1bf4eed9d4e8e392b0575ab5b5830135c","native/include/ciglet/ciglet.h":"1290f7e14e6d0fc2f1849f31f75ef610ae7dd2abdf1e1f470ac93b58bf56331e","native/include/ciglet/external/fastapprox-all.h":"568b4b09def0f5539b3487ed01a9f616955fb8f11b4d90c1c7ffc0f34c9fc543","native/libllsm_speed.so":"d8da444d9c3298899065de57d5efbea69574d38ec1c4d9f42a50c2cb1a730c2a","native/wrapper.c":"6a2abdf3b87f4af1dc402dbaec4340ad9907a1f35f27222641f7f6dda40724b9"},"sources":{"llsm":{"repository":"https://github.com/Sleepwalking/libllsm2.git","revision":"d18cdf02b31819670e3bc23ead933ad9467da06d"},"ciglet":{"repository":"https://github.com/Sleepwalking/ciglet.git","revision":"895ba9b1c0eabee83d3544208bbc82420efa3206"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"magphase":{"schema_version":1,"method":"magphase","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"src/magphase/demos/demo_copy_synthesis_lossless.py":"d3aa2abd2f7facff5dc821e5db9851b75216dc507f607bd8a77bf27df736ce68","src/magphase/demos/demo_copy_synthesis_low_dim.py":"9aad8c21dd0cfce4f390008d8da3204df6dec154ad0de849e3243167d8bae25d","src/magphase/development/compare_mags.py":"382f9a6ada1cd00edec29da28acda9de990e364e27af59a9f0afa48b4c97a6cd","src/magphase/development/run_merlin_experiment.py":"870158543e6e6244c363ad1bbeda31184feb46f2ead86d244e205480530c4c9d","src/magphase/scripts/batch_convert_label_state_aligned_to_variable_frame_rate.py":"6705bdb715d6c87caf7da424458c4ba0d726764dd23eab5fc598487dc8bcda4d","src/magphase/scripts/batch_feature_extraction_for_tts.py":"e5e901d8c8807e48d002ecb8751d3beb7d3fbb856db9b223f9ff9b9f66ffff34","src/magphase/scripts/batch_waveform_generation.py":"c0df72751851ab044bb980b006df6f0e5c000bc849d444fc80a275e1e19faaa3","src/magphase/src/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/magphase/src/libaudio.py":"40fb05431d31fc762129a9b92f70a8f60200e564105b49f4f1e0c2b37fd43690","src/magphase/src/libplot.py":"348db4785d0d6d33565a3c9192ec02cf2a3c028900cd998e1d842c6723da6b11","src/magphase/src/libutils.py":"305810cc60870fa8bb6241f0bf7046477c8015c82d31a23c8e2a33b72ac69ee4","src/magphase/src/magphase.py":"4dd367e608f58fde1b2699dd57f8d33092fb8626ac493697d04561e556dd92bc","src/reaper/build/CMakeFiles/3.31.10/CompilerIdC/CMakeCCompilerId.c":"60d0dc2f177f0317302cef41d454988d9d49759ad3e011365bd6f6c544dfd126","src/reaper/build/CMakeFiles/CMakeConfigureLog.yaml":"10d42118b8f326e2d5dfaa0ce977b3a3ba82b586e909ec0a0999071e64c361d0","src/reaper/core/file_resource.h":"46b11710b474e34b15a5b5abbf648016e49f79d8825501712e2d1b6d06359c78","src/reaper/core/float_matrix-inl.h":"1a2c36ff15eb5bbc20ea0d918855c2b61edeac44d99900885c76967198432117","src/reaper/core/float_matrix.h":"9027d8d94753f74a2edb422acbe73cb881bf30affa8b6f74b612543c140679b0","src/reaper/core/track.h":"072c9c901156fa2d7a21bf688235c53bc25cf6d40c42e9bb81dda47d93cb5f81","src/reaper/epoch_tracker/epoch_tracker.h":"074f3ee3bde6fd346e28b9b4f8dbb5281870d3424d58c99baef73b6e315af80f","src/reaper/epoch_tracker/fd_filter.h":"4b6e6bdb34305e2ba726a9735b2850bdff62de8b223ca13824cc56c55086f167","src/reaper/epoch_tracker/fft.h":"0a6d8a9e3b45f1c9ce3356fff157a5cd543ccd39f0f90099afc19654f754174b","src/reaper/epoch_tracker/lpc_analyzer.h":"a2009fad56bf64dd8cff42c75e0724b9df5c5fdf2ab92486c48b91acb2f5b945","src/reaper/wave/codec_api-inl.h":"4fa4a9c669e3330c1b2cc46d31333c0a0a5e5d9bd7ab752159ad903b84d1e163","src/reaper/wave/codec_api.h":"85d3e07e807e8bf6e47ab13226b4151e670095accd04b4226ddcf8d2c8c32915","src/reaper/wave/codec_riff.h":"e7ea52ab3b1ffabd7b099ea2a18d051a9787b2ca9ba40fc180b44af1bfe45d70","src/reaper/wave/wave.h":"0f96bece1e3313dd0cf55d1933161070a3c2ce778253a2250d175f8fc42e87b1","src/reaper/wave/wave_io-inl.h":"fe5f7375d75f0b0892a367cffe405efb113173a0aa16e9ab3d8f2cd730d2c8a6","src/reaper/wave/wave_io.h":"04245f9d0b38c24c9c5c8e17df758d350e550c31023eef73a93dd77a6db04294","magphase-port/demos/demo_copy_synthesis_lossless.py":"d3aa2abd2f7facff5dc821e5db9851b75216dc507f607bd8a77bf27df736ce68","magphase-port/demos/demo_copy_synthesis_low_dim.py":"9aad8c21dd0cfce4f390008d8da3204df6dec154ad0de849e3243167d8bae25d","magphase-port/development/compare_mags.py":"382f9a6ada1cd00edec29da28acda9de990e364e27af59a9f0afa48b4c97a6cd","magphase-port/development/run_merlin_experiment.py":"870158543e6e6244c363ad1bbeda31184feb46f2ead86d244e205480530c4c9d","magphase-port/scripts/batch_convert_label_state_aligned_to_variable_frame_rate.py":"6705bdb715d6c87caf7da424458c4ba0d726764dd23eab5fc598487dc8bcda4d","magphase-port/scripts/batch_feature_extraction_for_tts.py":"e5e901d8c8807e48d002ecb8751d3beb7d3fbb856db9b223f9ff9b9f66ffff34","magphase-port/scripts/batch_waveform_generation.py":"c0df72751851ab044bb980b006df6f0e5c000bc849d444fc80a275e1e19faaa3","magphase-port/src/__init__.py":"2e262bac3afb815e57a323ef5b5a60df817d61eca72fb3b86620dd1b30b7e899","magphase-port/src/libaudio.py":"f93e14c379b87719a015017a56032a5920a5a4164034e7a3aaa5816dca4ce96a","magphase-port/src/libplot.py":"c4f77b57d88edb05e13107f471aa2269d0cf3f713d3cbaa5593a8c13b3b7a7c5","magphase-port/src/libutils.py":"6d4b040230702f765f6ba8cc71ed386d17ff61b79c6fe5eba0cc81c6722c8139","magphase-port/src/magphase.py":"2ce23c57862dfca7b106570aeecf988e8e60cd7cbbf8ba583fe16591db548ba5","magphase-port/tools/bin/acep":"6bcba02c7409c6499df60c1b6bfa6c9a9767269ad07ff2d51f9c8d9cdf0a06b1","magphase-port/tools/bin/acorr":"40b345c29a76af572173fbee18ca04291a421079172501c9bf405d744baca922","magphase-port/tools/bin/agcep":"621366418f75d67118177a2cdfc6fad510f4a38e235308b24bc254f7c6c1f06a","magphase-port/tools/bin/amcep":"dfc779388b2512664dd11eabd5aeb0eb198188cb860a9304769709f14b9b4bdc","magphase-port/tools/bin/average":"6265ba18044907b5f566b4f9bcbeac7ee2d2e24aff463ec1d84f9ecb83142f83","magphase-port/tools/bin/b2mc":"48730786cd37183749507a65a827d32bcac99ce307c13eacfb4c4485fa978770","magphase-port/tools/bin/bcp":"249d8fd4f9d55a6b7e1f2d7477be2cbd056facefa2a208fbe4873ca36c4528c0","magphase-port/tools/bin/bcut":"3fd5ea37b6927fefb248b4714d118d15133cc4d8b8973bcf908cf9ad0eabf0d5","magphase-port/tools/bin/bell":"daa10b3db0d317c45ba972cc043f10edc755a512502d533a800a64ed73cfb631","magphase-port/tools/bin/c2acr":"3559fb8f2e74c0d54484a43186c97f00abce3eccbfeeb43cd85bbb0d572ba9c9","magphase-port/tools/bin/c2ir":"c0931aeea115f1c9d49b10fd0d350399009dd74d50442e5448087688af39fe30","magphase-port/tools/bin/c2ndps":"b534dd109102f91c69e1c80eedfddf5b3c38686aed50bacaffc0c9232ba073d3","magphase-port/tools/bin/c2sp":"e7140b7e16d9dd2832ffc0d595e82c0f99c457ef53ed3f73f8d819ef3f70354d","magphase-port/tools/bin/cat2":"ff6d7f8c12c709f0d9efe58aee6728e2598f10307b92e6f9f67018991585ac41","magphase-port/tools/bin/cdist":"d0651a134a49ba2eec4539655c9b2dfd9dc41c03e4eb90e82ed63dd174e8c4f0","magphase-port/tools/bin/clip":"a1fb3c48021a3aed319bfc4f4d1a8a7676292062a61e32b54cae65cfde82d955","magphase-port/tools/bin/da":"67acd139295ea8ba0a3e680461a089b9e602fb64546c60c7274207c58e6b3f3d","magphase-port/tools/bin/dawrite":"2bed1e27e651704640b6a0f4f1a1639cb6c92df68b4094c46266e702923abbf0","magphase-port/tools/bin/dct":"1d3360f55e43b4cd4059d8f3ab74b42dedbac2b1971d8a117a1bf6244e4c3ef3","magphase-port/tools/bin/decimate":"288a66781105f49619f6fb47f4f48f58dea87d98ba44e5d9c27684ed499e8863","magphase-port/tools/bin/delay":"94021e97270e99364668f983b9b2ffe900f35d89c1f706ff043fb18e197f4919","magphase-port/tools/bin/delta":"bf7b0471feefdd19b660561a8e9fdbeb786f57614c24005d13bddafaf9491209","magphase-port/tools/bin/df2":"4fa2b4d4ca61e03c687ffe4b189fdc2624be0ed6373afe4f9b7b37cc07f42320","magphase-port/tools/bin/dfs":"343bd14685c1ca6451fbaf62307ece3b0c1a4b46d3c221ba0fc3165e12406e0f","magphase-port/tools/bin/dmp":"f514bf1b75a620449af5309dcb0ca971c7950ddded5c2f7734b3cac1d5815b16","magphase-port/tools/bin/ds":"99d2f4e9d4478cf3a85a3f4d825e86d68b61d26856fd58a639c6f65da3aff2ff","magphase-port/tools/bin/dtw":"8e933a52dc8bb05389f5aa10c446e187d9ecd482bb7b2ac04a3713d4e1756be7","magphase-port/tools/bin/echo2":"0f279df08e94546cb5bae8d6d499ba8bd007ad34c7b75a38e54f288da06a3327","magphase-port/tools/bin/excite":"8b53d5be5f9be08216ca8293ea196a1af1d97bd3cebd5afd4e85a4b65962e67f","magphase-port/tools/bin/extract":"e986de89fc2f727b8895b69707b81f477f08ff5dcdeff2dcb4d5be13cf0deabe","magphase-port/tools/bin/fd":"c8b52cd777408091d037951f0309c5882187a3e2c50c8303a17a017e5614c2b2","magphase-port/tools/bin/fdrw":"be2eed0f57b714b332b652d69b35c43e1cab95e1c4f1993bddbccdba24bb60f0","magphase-port/tools/bin/fft":"892a2240dd43a28eb689b05a5875b8b3127b0a2cf977ecb3a79cccbdfb08d7df","magphase-port/tools/bin/fft2":"c0818406f80b2904cf2293f59990770d8c58d522ac652208a8d09e36a4bf117f","magphase-port/tools/bin/fftcep":"f404224105498c6475697fe76bf4c625444b525c72d70c2e1690aa09c7f3e6c1","magphase-port/tools/bin/fftr":"ccb25a9345b4f5e02a63c035897a015711213f4c20f54e3f6bae2fc5d5d30642","magphase-port/tools/bin/fftr2":"d23f189867b6473a6a509b5e1110effe61771ce0a4992bdd18ea74463fe0a901","magphase-port/tools/bin/fig":"7972e185428f8b845cc6fabaca8f6b2f28eff5eae1884b635cbd4aa7eb0059bd","magphase-port/tools/bin/frame":"e9901c8bb27daa2ab35af8c283e7149d4ef20df6a436330b740fea8e13824242","magphase-port/tools/bin/freqt":"d02eb0b8076cb097d728bccc94cd0d22ef8abe65fc856348a432f934f32ff236","magphase-port/tools/bin/gc2gc":"cae28a86b522fa38fb482c468e6dc9ace2d33a6cff3d4775dfec98cab62ede57","magphase-port/tools/bin/gcep":"0ab3acabbb379400d6b096d79aedbfc53a67a67e8520df52b0b5254a1b039169","magphase-port/tools/bin/glogsp":"60452cb49b78101b4deafd3f3efa3f49b19c6100939bdce427b5f935c938715f","magphase-port/tools/bin/glsadf":"39945c96e336fb8785a4082955853cf8526305f8f730da9632fada40886d79ef","magphase-port/tools/bin/gmm":"2cbb37fdc061b8a7eba7054816b571ba183cf6509ae654d6d7c37bd37f52a559","magphase-port/tools/bin/gmmp":"f83797aa26512ec5ad7d05ba0c9d8de125aebb40fdaa93fa06aa4cec62e69582","magphase-port/tools/bin/gnorm":"304b7440255ddc1c3036c8f6129fd746c700f945ffecf9d804565769e61d56e6","magphase-port/tools/bin/grlogsp":"9e17bb41a8ec045f8976f1e1027575c82627af17a1845699c27475c549cf8f19","magphase-port/tools/bin/grpdelay":"7854bec7ac1df335616a5213689f7be069e415a7ceeffba72b6c79847f362ef7","magphase-port/tools/bin/gseries":"288bae1d4b2c451f355878c10babeeb6dc1ffdb8966d35de8490f79695368ffa","magphase-port/tools/bin/gwave":"6be98a889a70b98c39432d6aee76a7762aad614ccc9ab72740c51f48018fdf56","magphase-port/tools/bin/histogram":"9c4ddf19fbe7927c04d14992ca552e1650e6a0e193d8b1d85ab6e309d2f970ca","magphase-port/tools/bin/idct":"021774e33cbab425ae994bd8c769280cec375c9683e523a8a11aa233e7dcff0c","magphase-port/tools/bin/ifft":"4935440ef5fad4cbba0fc423524ced3f12fe4be8bff15bd8f84836ff0756a68b","magphase-port/tools/bin/ifft2":"98bf0f89f21b2d9fc62c8d8aabdacdf017c04349209ee3722a1cb5bfbd505485","magphase-port/tools/bin/ifftr":"cb41b4d6677c0d2e2e6dcb687f3bf2f65aa916a387082594541cc0994faebca3","magphase-port/tools/bin/ignorm":"6ddb1d433e5a0c59ab30bdf0c681edd0231cdaacfd2e50b8e536c44bcc69fab6","magphase-port/tools/bin/impulse":"683fb21e27418aee9dcc3c8b0e46a21702d145af686808f61fde40a2fb729d95","magphase-port/tools/bin/imsvq":"007546566a66ea0e0498304152a40db62ce3eb1ac7c97153022bfd1cabd4b167","magphase-port/tools/bin/interpolate":"cdafdf7d41eee498ce6075071df8969540af5b844112671f37fd44fe321036ae","magphase-port/tools/bin/ivq":"da812cfc837e7e6277b9449643f078bca5055767c444a1b06e0544b24525c145","magphase-port/tools/bin/lbg":"2f1e4f3af19590c8edab91fa41f30f0ac733cf9917bf36ea351eebb31a33440d","magphase-port/tools/bin/levdur":"de69974a33aed931c89242da162a81f9e1e3ff51b9e80149eaad45c8ddc7bee3","magphase-port/tools/bin/linear_intpl":"985153e7226cedc3476e91ce12456ad2c5c44670aab10a737f0636625fc33d81","magphase-port/tools/bin/lmadf":"42afa48e1d7e1839f3f521814a07cd444e49a5d96d02f85db25056f89a160be3","magphase-port/tools/bin/lpc":"bc914d7d60724674ab04bf4c6461d97b349f6f75e2a5f52c44a510476fceea7d","magphase-port/tools/bin/lpc2c":"405bd26dae50fc8fd632597928414d4ad17b942749d2989689beb47f8f0eeb5b","magphase-port/tools/bin/lpc2lsp":"acedb18128bd515337f1f3f5c64566fb7871126deb3958acbcff8d82fedecddb","magphase-port/tools/bin/lpc2par":"5ee3b40438da56b88adbeff80bf5252ca30ff30b9c39f57d3edf4cc1fa5cc1ab","magphase-port/tools/bin/lsp2lpc":"20c1843629b90ab5c131960980045d25a9fa96045ef6c89435c77795242696dd","magphase-port/tools/bin/lsp2sp":"7e24a823b6f45c44c29b76aff9f6bb51d6e67d33883abbacd97d85e799f9103e","magphase-port/tools/bin/lspcheck":"768f0243aacc28e2e3f548f7fbe591145d9f2ddbcfd3856cce581bb696fc5e88","magphase-port/tools/bin/lspdf":"c5e56c4ff1dc5fcb07b84ff656785efbe7d9eae70945a2edda9217a90a0c78f6","magphase-port/tools/bin/ltcdf":"adce4383d4149d748fef71147fa7050af1899416963234776b06b3451c39d29b","magphase-port/tools/bin/mc2b":"ea3329ae30e040fadc71fdacd486841986c363f17afd20bdcc9596c085b27d39","magphase-port/tools/bin/mcep":"8bb766fba07543853ed9cc9ed2d808e0cba40eda43469d40f07a291a2b030224","magphase-port/tools/bin/merge":"7de4a5f2e97e649ba29345dbb5ee2c64740addf06776b69d281986e36f5aa129","magphase-port/tools/bin/mfcc":"0e98f6ba2fcf9910ce354c4624e4a933ad3f08112fcc36d887621aacd94cce8a","magphase-port/tools/bin/mgc2mgc":"e00033c602b548e85042cf236f64b9ada4f8d3b1c9e128b4990cb2712a4f3aa6","magphase-port/tools/bin/mgc2mgclsp":"3dfa7f5377e2ea86644a70273db5bbae348628172f456d4ca71f7d8f9722e450","magphase-port/tools/bin/mgc2sp":"ccd56338898385c8a677c847c4478d67747e204d1931f55b13e8756a1200c494","magphase-port/tools/bin/mgcep":"6dfca6c328b44206cf9e69ca50b809d6e553d8292f57d13a881cd4dc48777bcb","magphase-port/tools/bin/mgclsp2mgc":"14eacd029f883d307b583f7d30b59772477f9a32010dff4f02f65314eabed271","magphase-port/tools/bin/mgclsp2sp":"9c31655dd4a7b807b02f05ab7a4b423b0a10d7c2efaee85c8d4620d1806cf026","magphase-port/tools/bin/mglsadf":"1fac23824e8aa92e805eb91c8620a9711bc2943d9627b878d1d2559fa06a9033","magphase-port/tools/bin/minmax":"8dd56b7248b302ca009999be3bf5a64c689aa616079d20c135bfd1ba14b20d75","magphase-port/tools/bin/mlpg":"c6d579ac39dcce2657882d0836c6656cdec354198c30275047115faf5053ec79","magphase-port/tools/bin/mlsacheck":"97890ada7a2a7baf8566fef8b545c399ca6fca92e6eb60d11d76a277c44e5a88","magphase-port/tools/bin/mlsadf":"c66eae544340244cb1d14292ac751bc34e2ba5fbb86c74e3caa0c0b5967b9436","magphase-port/tools/bin/msvq":"094fb197c1d4d45206e0501e652d15817e7b2fa49805db4a787c2c060f95db91","magphase-port/tools/bin/nacorr":"bc5eb674420acb78099714046867939072d64908b8f40034edd1aad01ba00d25","magphase-port/tools/bin/nan":"6bf4aeeda60dbfa1d7f996bc4f58b81af335f7224d55597b48b6e8b66104b903","magphase-port/tools/bin/ndps2c":"c9277ea5338a6b3af8d78593fed5080c5aa6ea9c0f4c4828c2ba285a23d22ad1","magphase-port/tools/bin/norm0":"fe2c25bd553a5ede6cd7c5ab706cb3b6345d0c8a87bc6a982850b5f74b819db0","magphase-port/tools/bin/nrand":"5987592a4b47967594bd952966b4c3156cf1f14ad089a3c4833f49b5de00f770","magphase-port/tools/bin/par2lpc":"d8df839a4a454848527e91f1990f880d536ebc06d43ec7ae8e46fb14c15ab1d2","magphase-port/tools/bin/pca":"6dd0ea09ce617f96727be49c612e3db9d8f1d1487f2e3972b433fa067b6c57d5","magphase-port/tools/bin/pcas":"0941a3b3c8afbe11a338c11a0784de21586cd638bc3740235e8284b40b1644b6","magphase-port/tools/bin/phase":"c0feab34a986833ba22e0b0092d7a78e58e190762939d530ac709848b5252539","magphase-port/tools/bin/pitch":"5c6ae7bcbee5423d708aa6f5fa75d34ca3ce4b9f99fb2d55833635fdc19bf370","magphase-port/tools/bin/poledf":"631538e18ab03546f9c4575f743bc1037f97054aae5538650389f7269a455247","magphase-port/tools/bin/psgr":"8a035b8de87f10421adca12dd7a30fb151dd2985573222890ba2c1f8ceff11d1","magphase-port/tools/bin/ramp":"fc6ab6bb7fb5e555cacf5234e6eeeeadb4afc3171209ed148e7e5659c12a6c0b","magphase-port/tools/bin/raw2wav":"b01300b07cf5052d39c708eb9a09e353044a2aae10439105c4fe70b3804ea525","magphase-port/tools/bin/rawtowav":"729ef2071819f5232edcf1b8a9d3efe73390554f8586dc18b0601a195357543f","magphase-port/tools/bin/reaper":"11103df7710bfed4b44cbb53ef2560a5471800c830772dbd3520d333d48e086e","magphase-port/tools/bin/reverse":"231fe73aea63ac05bb2056a575fc0cc5746183629f6d10c1788ae05d2d9844b8","magphase-port/tools/bin/rmse":"593d676f92e852a51ee41b22c168203c53388c86e54850d41ec800de35945619","magphase-port/tools/bin/root_pol":"cbe47e53eaa0b994b8d2712fd9941826ac20fc3870694e42220779b6c0c3af14","magphase-port/tools/bin/sin":"3ccc70750e641c5560b9934256af321b98e48bfbc2ba73c0046ca7d1d8b78518","magphase-port/tools/bin/smcep":"8e309d75e6e657f1e71b5a4dbc53d0df74812e607c33261f898d0ab0241686a7","magphase-port/tools/bin/snr":"0bc4a054fa24199751294a2f0020cea3fce29df2d96f5dc6dbd772e00a53d272","magphase-port/tools/bin/sopr":"e89c76923d14118ac18d4438c42f7c5a265c1e184bcfb492396300c22afbf76a","magphase-port/tools/bin/spec":"1c13dc06c80f84b67ba7a30ca049f5d26098854dd08aab16df586599171fc840","magphase-port/tools/bin/step":"5d265e09ba6594cd4822b5445826f056c606afab3046bd1142f471f3e404eccb","magphase-port/tools/bin/str":"eeb793b137b828eb5d4ffc66d9d23355cab3f15089765e83fb58f19f86cd1d26","magphase-port/tools/bin/swab":"db1c32e17b1d95617c99ebadac0bf1228ea6c624e40475c26b917ef75e32c63f","magphase-port/tools/bin/symmetrize":"a9df430df81911f32d01f9715b671b60b086991e21f524005b24b32e219ddd01","magphase-port/tools/bin/train":"a738bbe949adfe80b9d7e6b3f5f4e4a5a55aa7df10250b99b5a31794639aff48","magphase-port/tools/bin/transpose":"d97a4799b535cbca9a937f721215307e04c17e0047faa683fd178191d22a3493","magphase-port/tools/bin/uels":"64164e3af1e95028430f301dcdb1be49445d94feb4ca7f918f290ab7ec2f5a6f","magphase-port/tools/bin/ulaw":"690bfd424ac0254660fce7a5f8f377d2c9f3742ca0745314f9af2939aa9f1ade","magphase-port/tools/bin/us":"d3854dfa640244a42b06588bf37344138a4b29e30f966bbde8db538e09f7ad6f","magphase-port/tools/bin/us16":"68f3b772ff4a32b4455f0ac493c3d2c21c89323b738482a8a5cdc0fd8920e504","magphase-port/tools/bin/uscd":"6abcabe445a388256572a24a38f3bd585c419fc35de4e22d922982454029a5ca","magphase-port/tools/bin/vc":"6edc937a55bf76f268e860f2e6bdbc36330433b05b43a1ca9f9eca228ca03bd1","magphase-port/tools/bin/vopr":"7035d7c9eb1924ec2d815fd73b9a8d97ba4627a2c27b7fbc19d70c41fdb96494","magphase-port/tools/bin/vq":"20dda03fc4dcc500d8684da80e37c5eb071e61efc1947d0f69c99bfb3fb56d39","magphase-port/tools/bin/vstat":"bf88bbbc29ffdb213043ebe0e41f743933344455b77d6175726608b58ca61413","magphase-port/tools/bin/vsum":"c09fcd9864764bda701c2862b5c9de4c40ff83b5118b391e40ed08a3f26acc85","magphase-port/tools/bin/wav2raw":"b8527b413a56256bac2835925b274f9f57d6700fb439385ad7bcd32bfea06687","magphase-port/tools/bin/wavjoin":"255c0999124f882812ddf0ce8efa1776066e928c26fad4ff9233d70cd8811722","magphase-port/tools/bin/wavsplit":"5ecc5215401773b91d650be51a23d60acb24584059b55b0d626f431ec9845415","magphase-port/tools/bin/window":"31ddcabf1746e21b778966cd452c165f30b864d1b863b15c199a895acc7d830a","magphase-port/tools/bin/x2x":"56dcf49403dd043807e0ef3eabd3c4c20c1b5e26f1d7b2f4e4c8325a0e7c8f11","magphase-port/tools/bin/xgr":"845bded24ed8edab0d4232f493969af12cf10f198edb3dc50b8a80d0b85de7be","magphase-port/tools/bin/zcross":"2c4d95b8675684707d419e44d2fedbfe4cb2cd9e5c1117acb17959fb1bffc21e","magphase-port/tools/bin/zerodf":"ee9f37f440779406851d01ae42d7e651676d1bb573f7d225d8a8d069110bfdac"},"sources":{"magphase":{"repository":"https://github.com/CSTR-Edinburgh/magphase.git","revision":"ec6ae285c3128d3f416857f05d2a7e4be0bd5707"},"reaper":{"repository":"https://github.com/google/REAPER.git","revision":"1d6e9b95e6b08b500fccbc9a043989dbda747276"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"}},"native_sample_rates":{"hnm":16000,"world_harvest":16000,"facodec":16000,"hifiglot":44100,"golf":24000,"sifigan":24000,"llsm2":24000,"magphase":48000},"evaluation":"native-full-MOS + strict-length + original-rate-common-prefix-v1","runtime_python":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","host_python":"3.12.14","host_environment":{"platform":"Linux-6.18.44-x86_64-with-glibc2.41","machine":"x86_64","cpu_count":9,"cpu_model":"INTEL(R) XEON(R) PLATINUM 8573C","packages":{"numpy":"2.3.5","scipy":"1.17.0","pystoi":"0.4.1","torch":"2.8.0+cpu","torchaudio":"2.8.0+cpu","torchvision":"0.23.0+cpu","transformers":"4.51.3","timm":"1.0.15","librosa":"0.11.0","utmosv2":"1.3.1.dev0"}},"manifest_sha256":"644f4d57483366bb20d8dbcfdaa254e354aed4af8318a6a2984bb2e8b3ddc9c1","current_base":{"baseline":{"status":"selected","method":"hnm","version":"hnm-v3","reason":"雑音PSDの台形積分とFFT端点配分を揃える候補Dを採用。固定120録音で帯域・平均包絡・音量誤差と予測MOSが小幅改善し、制御と保存互換を確認。明瞭性・スペクトル誤差の小さな系統的悪化を許容した開発上の選択で、知覚品質の優越は未確証。","comparison_reference":"hnm","comparison_reference_meaning":"実験開始時点の唯一の自作ベース。過去結果は当時の版を保持する。既存比較対象はWORLD（Harvest）。","previous_version":"hnm-v2","selection_report":"experiments/2026-10-12/noise-stability/results/run-001/index.html"},"source_hashes":{"BASELINE.json":"da49356e381abd9a4538ef48a578b5474373efa1a9618d62e74e08fa45b302f4","vocoderbench/hnm.py":"58f0378793b7433f80bf854972882298649799d8e3d0868674816e9bf6b63f74","vocoderbench/hnm_analysis.py":"3f45703667237333d76c4145c1513e4eb3342bd40f01a714c10303082440e26a","vocoderbench/hnm_control.py":"b2f5efbbb1e550569c231a08f12997d94b0bacb130bcc816fe6bd458b17a98bc","vocoderbench/hnm_diagnostics.py":"076edd5a785e83540a0cf55fea06075769ef35503f210eb977efc1703dd0081e","vocoderbench/hnm_parameters.py":"635bf222bac8ce41897c0977677e67be4366724f4239a9e85e153169fcea2cf0","vocoderbench/hnm_synthesis.py":"079057d2fba2b4ca0d189e050449aa7bc3fa20033afb0b03893300950ae13593"},"python":"3.12.14","packages":{"numpy":"2.3.5","scipy":"1.17.0","torch":"2.8.0+cpu"},"noise_seeds":[0],"role":"current_development_base_not_new_candidate"},"synthesis_repetitions":1,"noise_seeds":[0],"timing_protocol":"single encode/decode; no warmup/repetitions; diagnostic only","mos_protocol":{"fold":0,"extraction_seed":20261012,"repetitions":1,"cpu_threads":4,"include_originals":true},"scope_limit":"今回HNMもseed 0。過去の8 seed音響研究の再現ではない"},"execution":{"started_at":"2026-10-13T04:37:48.919662+00:00","repository_commit":"c00edae9ed07a258d43c7c3e8bf8bb9d6ea7c595","local_source_commit":"c00edae9ed07a258d43c7c3e8bf8bb9d6ea7c595","remote_source_commit":"2308accfd3069e19fed852e1334b8269da761b5a","source_tree_sha":"53cb6143cf00340b8a5b1b328571aefd9aabbc07","remote_source_verification":"Exact whole Git tree equality after fetching remote branch; original local execution commit retained."},"resynthesis":[{"case":"librispeech_1089-134686-0008","method":"hnm","status":"measured","mos":{"status":"measured","value":2.425536632537842,"predicted_mos":2.425536632537842,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"93a0b9d54173e3e8d45679c3447f63e46bd208b319c3e39b069793f8a9fd7b70","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.402614669001196},"started_at":"2026-10-13T04:37:53.786140+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3766000847681426e-05,"native_input_samples":96000,"native_input_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","output_sha256":"fb0de2e771eb44c1aca53337caf01155e4b5b8c8d34df49556db48037cd303b2","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4964093649996357,"synthesis_s":0.5370818889996372,"total_s":2.033491253999273,"rtf":0.3389152089998788,"output_peak":0.3883210433659479,"output_rms":0.042113093996853423,"finished_at":"2026-10-13T04:37:55.827334+00:00","evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"fb0de2e771eb44c1aca53337caf01155e4b5b8c8d34df49556db48037cd303b2","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.393300978191057,"spectral_convergence":0.3549098768296807,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.666793491143492,"spectral_convergence":0.3293487986418764,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.094949864609703,"spectral_convergence":0.38042465147280186,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.00032587002735268167,"delta_db":0.10396745723255543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.000842232073479996,"delta_db":-0.3606420653479739,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.00043552332787888794,"delta_db":-0.2517164718982998,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00011028422739288325,"delta_db":-0.12034357431544501,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":3.9952211995856145e-05,"delta_db":0.13044650489315612,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.965081788750654e-05,"delta_db":1.5135318269636349,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.41344131677517754,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20877726563894836,"pitch":null,"output_rms_dbfs":-27.51165700659871,"output_peak":0.3883210433659479,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.051681444648085},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.393300978191057,"spectral_convergence":0.3549098768296807,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.666793491143492,"spectral_convergence":0.3293487986418764,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.094949864609703,"spectral_convergence":0.38042465147280186,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.00032587002735268167,"delta_db":0.10396745723255543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.000842232073479996,"delta_db":-0.3606420653479739,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.00043552332787888794,"delta_db":-0.2517164718982998,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00011028422739288325,"delta_db":-0.12034357431544501,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":3.9952211995856145e-05,"delta_db":0.13044650489315612,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.965081788750654e-05,"delta_db":1.5135318269636349,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.41344131677517754,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20877726563894836,"pitch":null,"output_rms_dbfs":-27.51165700659871,"output_peak":0.3883210433659479,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.051681444648085},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8605243696289566,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04313760336580714,"rms_dbfs":-27.30287974095976,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.449441334882454,"mean":-9.814103444417318e-05,"zero_energy":false},"output":{"rms":0.042113093996853423,"rms_dbfs":-27.51165700659871,"peak":0.3883210433659479,"peak_dbfs":-8.216181482404684,"crest_factor_db":19.295475524194025,"mean":-9.738938799563281e-05,"zero_energy":false},"rms_difference_db":-0.20877726563895038},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":2.3352066621161184e-06,"output_rms_dbfs":-56.316746790186116,"output_zero_energy":false,"output_peak":0.03135356331169534,"output_energy_fraction":0.0002479808902260979,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":161,"both_estimated_voiced_frames":143,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9662162162162162,"absolute_difference_cents_median":0.7601093922979201,"absolute_difference_cents_p95":4.8398545601154055,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016375851184643225,"log_rms_mae_db":1.1267817830356814,"rms_envelope_nrmse":0.10953294584064513,"envelope_correlation":0.9888973505049091,"rms_gain_error_db":-0.20876852419846925,"reference_mean_power":0.0018623503698243853,"output_mean_power":0.0017749435161334075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.48847287729811e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002826666045552883,"output_power":0.00027850534693776456,"reference_active":true,"delta_db":-0.06440964633100756},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.155794557866922e-05,"output_power":5.9337146600989084e-05,"reference_active":true,"delta_db":-0.15957459181234093},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.62934396131159e-06,"output_power":9.831817408275758e-06,"reference_active":true,"delta_db":0.566560243538333}],"active_band_count":3,"active_band_mae_db":0.2635148272272272}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014981717790874295,"log_rms_mae_db":0.5968187586410607,"rms_envelope_nrmse":0.07448098587179383,"envelope_correlation":0.9948466614972954,"rms_gain_error_db":-0.20880157779660544,"reference_mean_power":0.0018713077892111094,"output_mean_power":0.0017834669575677556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.792361294396735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025909451544035513,"output_power":0.0002557576212475342,"reference_active":true,"delta_db":-0.05629635865915429},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9672561350575066e-05,"output_power":2.8645102113652143e-05,"reference_active":true,"delta_db":-0.15304661700154742},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579324289571115e-07,"output_power":7.743140500232653e-07,"reference_active":true,"delta_db":0.09286650440004414}],"active_band_count":3,"active_band_mae_db":0.10073649335358194}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014318840727902982,"log_rms_mae_db":1.8503525527640854,"rms_envelope_nrmse":0.15345018398974,"envelope_correlation":0.9835791043188611,"rms_gain_error_db":-0.470084440344433,"reference_mean_power":0.0010573416450716808,"output_mean_power":0.000948870388743474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.91879361710777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021964908416761717,"output_power":0.00020729372699824922,"reference_active":true,"delta_db":-0.25143236876710295},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090309412479679e-05,"output_power":3.4151086549252003e-05,"reference_active":true,"delta_db":-0.7835163583603427},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.52813465615998e-06,"output_power":6.0278013154305425e-06,"reference_active":true,"delta_db":1.242395956619642}],"active_band_count":3,"active_band_mae_db":0.7591148945823626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013022613407710025,"log_rms_mae_db":1.2251272926255705,"rms_envelope_nrmse":0.11704625784440395,"envelope_correlation":0.9909925269203403,"rms_gain_error_db":-0.47010047406461375,"reference_mean_power":0.001062644583935512,"output_mean_power":0.0009536257855482872},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.561667844043168e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020949277098297487,"output_power":0.0001968981088900344,"reference_active":true,"delta_db":-0.2692749628537069},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8871602692351808e-05,"output_power":1.587679885103812e-05,"reference_active":true,"delta_db":-0.7504584210202674},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293974094910672e-07,"output_power":4.2046634148486873e-07,"reference_active":true,"delta_db":-0.09128183994170247}],"active_band_count":3,"active_band_mae_db":0.37033840793855893}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":1.908144328774027,"rms_envelope_nrmse":0.1947502927568842,"envelope_correlation":0.9708998246634316,"rms_gain_error_db":-0.10653106382685085,"reference_mean_power":0.00011346719055744517,"output_mean_power":0.00011071773572541476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5656809677774876e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.231325141923398e-05,"output_power":2.3166231323517644e-05,"reference_active":true,"delta_db":0.1629252965473732},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9000473344355792e-06,"output_power":3.6720703645107228e-06,"reference_active":true,"delta_db":1.0250590810427145},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0009489505392986e-06,"output_power":1.8540639306189253e-06,"reference_active":true,"delta_db":2.6771277656093533}],"active_band_count":3,"active_band_mae_db":1.2883707143998138}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":1.1236596030591932,"rms_envelope_nrmse":0.11262946926741536,"envelope_correlation":0.9897530611525162,"rms_gain_error_db":-0.10672934464762066,"reference_mean_power":0.00011403279016460418,"output_mean_power":0.00011126455013912915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.16052039626869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1030857564945275e-05,"output_power":2.267649330032667e-05,"reference_active":true,"delta_db":0.3271891392508516},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3790687257240347e-06,"output_power":1.8564678519278751e-06,"reference_active":true,"delta_db":1.2910152331062257},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.278096740186842e-08,"output_power":1.1893987352253393e-07,"reference_active":true,"delta_db":2.1330964843199602}],"active_band_count":3,"active_band_mae_db":1.2504336188923457}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":2.343635273266763,"rms_envelope_nrmse":0.35047765357365385,"envelope_correlation":0.9253045220295367,"rms_gain_error_db":0.4381043774547478,"reference_mean_power":2.180480752557149e-05,"output_mean_power":2.4119188693773135e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7142973306691606e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0022096022650048e-05,"output_power":9.569943219432422e-06,"reference_active":true,"delta_db":-0.2004919842262752},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6747496282854544e-06,"output_power":2.8650910360121752e-06,"reference_active":true,"delta_db":0.29855290051417965},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.037734768320821e-07,"output_power":9.061476600934742e-07,"reference_active":true,"delta_db":0.5206530221877177}],"active_band_count":3,"active_band_mae_db":0.3398993023093908}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":1.8277212237492757,"rms_envelope_nrmse":0.2990801709865258,"envelope_correlation":0.9422869724679457,"rms_gain_error_db":0.43806374534587195,"reference_mean_power":2.191432140518644e-05,"output_mean_power":2.4240099682743622e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5673668617618725e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.014084411974247e-05,"output_power":9.603083781659823e-06,"reference_active":true,"delta_db":-0.23663389135695756},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.343672608691746e-06,"output_power":1.5007469624975918e-06,"reference_active":true,"delta_db":0.48014009033151306},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.915436539179563e-08,"output_power":6.700676763939551e-08,"reference_active":true,"delta_db":1.34556574864137}],"active_band_count":3,"active_band_mae_db":0.6874465767766136}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0008","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.521312713623047,"predicted_mos":3.521312713623047,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ecff9b73a78c2d00420351d11a6aca59f0017d16b5e01eb9a6220806c364e44a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.479579421999006},"started_at":"2026-10-13T04:40:44.583009+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2632999389315955e-05,"native_input_samples":96000,"native_input_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","output_sha256":"b9642f806433f5df51fe8afdae3f61875accd58d811322c5149ac4a0e21327b6","output_samples":96000,"length_delta_samples":0,"analysis_s":1.824730891999934,"synthesis_s":0.07957798000006733,"total_s":1.9043088720000014,"rtf":0.31738481200000024,"output_peak":0.532400631371262,"output_rms":0.04670554336652409,"finished_at":"2026-10-13T04:40:46.494869+00:00","evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"b9642f806433f5df51fe8afdae3f61875accd58d811322c5149ac4a0e21327b6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.465250015980082,"spectral_convergence":0.7739551464958658,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.24890908644032,"spectral_convergence":0.3008773731906827,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.496434271645142,"spectral_convergence":0.2557851796757117,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.00033436625086758355,"delta_db":0.21574764285838763,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0011355866984160985,"delta_db":0.9372431428838958,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.000524224566150803,"delta_db":0.5533429973018997,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.0001264705646566574,"delta_db":0.47441695095996983,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":4.507912276379589e-05,"delta_db":0.6547927712665703,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.568057830732348e-05,"delta_db":0.5333462745473655,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5614816299696815,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.690248320427418,"pitch":null,"output_rms_dbfs":-26.612631420532338,"output_peak":0.532400631371262,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.070197791355182},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.465250015980082,"spectral_convergence":0.7739551464958658,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.24890908644032,"spectral_convergence":0.3008773731906827,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.496434271645142,"spectral_convergence":0.2557851796757117,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.00033436625086758355,"delta_db":0.21574764285838763,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0011355866984160985,"delta_db":0.9372431428838958,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.000524224566150803,"delta_db":0.5533429973018997,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.0001264705646566574,"delta_db":0.47441695095996983,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":4.507912276379589e-05,"delta_db":0.6547927712665703,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.568057830732348e-05,"delta_db":0.5333462745473655,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5614816299696815,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.690248320427418,"pitch":null,"output_rms_dbfs":-26.612631420532338,"output_peak":0.532400631371262,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.070197791355182},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9336871648363211,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04313760336580714,"rms_dbfs":-27.30287974095976,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.449441334882454,"mean":-9.814103444417318e-05,"zero_energy":false},"output":{"rms":0.04670554336652409,"rms_dbfs":-26.612631420532338,"peak":0.532400631371262,"peak_dbfs":-5.4752287633878005,"crest_factor_db":21.13740265714454,"mean":5.241233886015747e-08,"zero_energy":false},"rms_difference_db":0.6902483204274219},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":1.513771811131965e-06,"output_rms_dbfs":-58.199395862858765,"output_zero_energy":false,"output_peak":0.025771172254009416,"output_energy_fraction":0.00013069252505582476,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":182,"both_estimated_voiced_frames":145,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9797297297297297,"absolute_difference_cents_median":2.5021695708090315,"absolute_difference_cents_p95":11.55020835646315,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016375851184643225,"log_rms_mae_db":2.2409897546644753,"rms_envelope_nrmse":0.1804830840846601,"envelope_correlation":0.9774595186767175,"rms_gain_error_db":0.6903750223186029,"reference_mean_power":0.0018623503698243853,"output_mean_power":0.002183226990587238},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.48847287729811e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002826666045552883,"output_power":0.00032850291687209403,"reference_active":true,"delta_db":0.6526472803088637},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.155794557866922e-05,"output_power":6.325962722374451e-05,"reference_active":true,"delta_db":0.11842511024424364},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.62934396131159e-06,"output_power":6.5851411355233085e-06,"reference_active":true,"delta_db":-1.1741269271373653}],"active_band_count":3,"active_band_mae_db":0.6483997725634909}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014981717790874295,"log_rms_mae_db":1.5046031967203988,"rms_envelope_nrmse":0.11917599811398764,"envelope_correlation":0.9930377346622896,"rms_gain_error_db":0.6908583827930141,"reference_mean_power":0.0018713077892111094,"output_mean_power":0.0021939719135980903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.792361294396735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025909451544035513,"output_power":0.00030059682696628314,"reference_active":true,"delta_db":0.6452617209960207},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9672561350575066e-05,"output_power":3.078773952989079e-05,"reference_active":true,"delta_db":0.16022767234058788},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579324289571115e-07,"output_power":5.62746651823846e-07,"reference_active":true,"delta_db":-1.2931756946506978}],"active_band_count":3,"active_band_mae_db":0.6995550293291021}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014318840727902982,"log_rms_mae_db":2.4972380456921495,"rms_envelope_nrmse":0.212895948967307,"envelope_correlation":0.9784549512898931,"rms_gain_error_db":0.9518818849977507,"reference_mean_power":0.0010573416450716808,"output_mean_power":0.0013164474464043194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.91879361710777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021964908416761717,"output_power":0.00026662131041943845,"reference_active":true,"delta_db":0.8416546180673129},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090309412479679e-05,"output_power":4.226228425141226e-05,"reference_active":true,"delta_db":0.14196805145199357},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.52813465615998e-06,"output_power":2.8009906856854283e-06,"reference_active":true,"delta_db":-2.086076686382961}],"active_band_count":3,"active_band_mae_db":1.0232331186340893}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013022613407710025,"log_rms_mae_db":1.9357289682985022,"rms_envelope_nrmse":0.15985578522252447,"envelope_correlation":0.9913124569946283,"rms_gain_error_db":0.9518870474403128,"reference_mean_power":0.001062644583935512,"output_mean_power":0.001323051464363206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.561667844043168e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020949277098297487,"output_power":0.00025244096843281933,"reference_active":true,"delta_db":0.8099079632114414},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8871602692351808e-05,"output_power":1.944694357681447e-05,"reference_active":true,"delta_db":0.1304256945882532},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293974094910672e-07,"output_power":2.6071754366248803e-07,"reference_active":true,"delta_db":-2.1668916426341718}],"active_band_count":3,"active_band_mae_db":1.0357417668112887}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":1.7994694753813048,"rms_envelope_nrmse":0.18511664318843607,"envelope_correlation":0.9766635613884935,"rms_gain_error_db":0.3930815348670972,"reference_mean_power":0.00011346719055744517,"output_mean_power":0.00012421626181368948},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5656809677774876e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.231325141923398e-05,"output_power":2.424768434626446e-05,"reference_active":true,"delta_db":0.36107410829645703},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9000473344355792e-06,"output_power":2.926856994731759e-06,"reference_active":true,"delta_db":0.039964169760745696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0009489505392986e-06,"output_power":6.381064967200102e-07,"reference_active":true,"delta_db":-1.955187622594006}],"active_band_count":3,"active_band_mae_db":0.7854086335504028}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":1.0114314013169539,"rms_envelope_nrmse":0.09230155124739112,"envelope_correlation":0.9950606552066227,"rms_gain_error_db":0.3928795488933437,"reference_mean_power":0.00011403279016460418,"output_mean_power":0.00012482963645291947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.16052039626869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1030857564945275e-05,"output_power":2.287390187741513e-05,"reference_active":true,"delta_db":0.36483271715496374},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3790687257240347e-06,"output_power":1.432874438487717e-06,"reference_active":true,"delta_db":0.1662222549719984},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.278096740186842e-08,"output_power":4.6207395573280715e-08,"reference_active":true,"delta_db":-1.9730633321498636}],"active_band_count":3,"active_band_mae_db":0.8347061014256086}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":2.1077184295003186,"rms_envelope_nrmse":0.28373988915239673,"envelope_correlation":0.9601814318078213,"rms_gain_error_db":0.9038865620238677,"reference_mean_power":2.180480752557149e-05,"output_mean_power":2.6849791257382325e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7142973306691606e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0022096022650048e-05,"output_power":1.259640812450089e-05,"reference_active":true,"delta_db":0.9928816392103237},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6747496282854544e-06,"output_power":3.827378577515185e-06,"reference_active":true,"delta_db":1.5561828588478672},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.037734768320821e-07,"output_power":7.934930212818937e-07,"reference_active":true,"delta_db":-0.055905595909604894}],"active_band_count":3,"active_band_mae_db":0.8683233646559319}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":1.2060140234307937,"rms_envelope_nrmse":0.19098721408982558,"envelope_correlation":0.9850604939333156,"rms_gain_error_db":0.9038225874139689,"reference_mean_power":2.191432140518644e-05,"output_mean_power":2.6984245890805507e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5673668617618725e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.014084411974247e-05,"output_power":1.3030445351674688e-05,"reference_active":true,"delta_db":1.0888515218872268},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.343672608691746e-06,"output_power":1.8901435571148537e-06,"reference_active":true,"delta_db":1.4820132626898568},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.915436539179563e-08,"output_power":4.933282738295553e-08,"reference_active":true,"delta_db":0.015739131278913354}],"active_band_count":3,"active_band_mae_db":0.8622013052853323}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0008","method":"facodec","status":"measured","mos":{"status":"measured","value":3.583048105239868,"predicted_mos":3.583048105239868,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"db65ad0b489bbe4021d5c89486cb8cf2ad01d8581baf5527247dd1d3e6b4a7eb","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.1005812250004965},"started_at":"2026-10-13T04:44:07.684994+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.117200023756595e-05,"native_input_samples":96000,"native_input_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","output_sha256":"54c433b3606273b2954094c4357cb2fbec5e685d1ba0bfb4928aff376072c353","output_samples":96000,"length_delta_samples":0,"analysis_s":2.3596668569998656,"synthesis_s":15.08092965800006,"total_s":17.440596514999925,"rtf":2.9067660858333206,"output_peak":0.3907930552959442,"output_rms":0.04153907252673629,"finished_at":"2026-10-13T04:44:25.136978+00:00","evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"54c433b3606273b2954094c4357cb2fbec5e685d1ba0bfb4928aff376072c353","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.2215740096342085,"spectral_convergence":0.1966252407172143,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.180356128298943,"spectral_convergence":0.19542825590381951,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.43071329061716,"spectral_convergence":0.21311081803374343,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.0003190616307452553,"delta_db":0.012269092301030035,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0008252094046391849,"delta_db":-0.4493181368904965,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.00043241066553958236,"delta_db":-0.2828667179703377,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00010084903030669596,"delta_db":-0.5087603577456491,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":3.669324364010195e-05,"delta_db":-0.2391007441249916,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.1270571510636762e-05,"delta_db":-0.9008150927463531,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3988550236298096,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3279843491064328,"pitch":null,"output_rms_dbfs":-27.63086409006619,"output_peak":0.3907930552959442,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.277547809516771},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.2215740096342085,"spectral_convergence":0.1966252407172143,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.180356128298943,"spectral_convergence":0.19542825590381951,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.43071329061716,"spectral_convergence":0.21311081803374343,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.0003190616307452553,"delta_db":0.012269092301030035,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0008252094046391849,"delta_db":-0.4493181368904965,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.00043241066553958236,"delta_db":-0.2828667179703377,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00010084903030669596,"delta_db":-0.5087603577456491,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":3.669324364010195e-05,"delta_db":-0.2391007441249916,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.1270571510636762e-05,"delta_db":-0.9008150927463531,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3988550236298096,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3279843491064328,"pitch":null,"output_rms_dbfs":-27.63086409006619,"output_peak":0.3907930552959442,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.277547809516771},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9261520705087799,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04313760336580714,"rms_dbfs":-27.30287974095976,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.449441334882454,"mean":-9.814103444417318e-05,"zero_energy":false},"output":{"rms":0.04153907252673629,"rms_dbfs":-27.63086409006619,"peak":0.3907930552959442,"peak_dbfs":-8.161063252875348,"crest_factor_db":19.469800837190846,"mean":1.7588983189568807e-05,"zero_energy":false},"rms_difference_db":-0.3279843491064298},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":1.0289021507668496e-06,"output_rms_dbfs":-59.87625924948568,"output_zero_energy":false,"output_peak":0.007259069010615349,"output_energy_fraction":0.00011230204820648458,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":140,"both_estimated_voiced_frames":136,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.918918918918919,"absolute_difference_cents_median":2.359151232361109,"absolute_difference_cents_p95":10.893110190810907,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016375851184643225,"log_rms_mae_db":1.16145058980538,"rms_envelope_nrmse":0.1000917793113361,"envelope_correlation":0.9917888922227545,"rms_gain_error_db":-0.3280482720460892,"reference_mean_power":0.0018623503698243853,"output_mean_power":0.0017268577432513756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.48847287729811e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002826666045552883,"output_power":0.0002776254179759194,"reference_active":true,"delta_db":-0.07815276703587261},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.155794557866922e-05,"output_power":5.593175412860505e-05,"reference_active":true,"delta_db":-0.4162567749126528},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.62934396131159e-06,"output_power":8.483777801465879e-06,"reference_active":true,"delta_db":-0.07388494702542944}],"active_band_count":3,"active_band_mae_db":0.18943149632465162}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014981717790874295,"log_rms_mae_db":0.8221619003027709,"rms_envelope_nrmse":0.09102366849309282,"envelope_correlation":0.9930098360485584,"rms_gain_error_db":-0.3281815023613949,"reference_mean_power":0.0018713077892111094,"output_mean_power":0.0017351102489431118},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.792361294396735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025909451544035513,"output_power":0.0002551865787615584,"reference_active":true,"delta_db":-0.06600390417370239},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9672561350575066e-05,"output_power":2.6768439594153978e-05,"reference_active":true,"delta_db":-0.447319807484079},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579324289571115e-07,"output_power":6.955842517795142e-07,"reference_active":true,"delta_db":-0.3728074832719188}],"active_band_count":3,"active_band_mae_db":0.29537706497656674}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014318840727902982,"log_rms_mae_db":1.5893808521914219,"rms_envelope_nrmse":0.11198789474880838,"envelope_correlation":0.9920738537402952,"rms_gain_error_db":-0.41012568175805036,"reference_mean_power":0.0010573416450716808,"output_mean_power":0.0009620613547780458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.91879361710777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021964908416761717,"output_power":0.00020258566853556282,"reference_active":true,"delta_db":-0.35120677872230327},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090309412479679e-05,"output_power":3.630861373507853e-05,"reference_active":true,"delta_db":-0.5174649372319664},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.52813465615998e-06,"output_power":4.567941372684739e-06,"reference_active":true,"delta_db":0.03801188037336917}],"active_band_count":3,"active_band_mae_db":0.30222786544254626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013022613407710025,"log_rms_mae_db":1.303998894508933,"rms_envelope_nrmse":0.10071119187861191,"envelope_correlation":0.9936493854140934,"rms_gain_error_db":-0.41017529420112525,"reference_mean_power":0.001062644583935512,"output_mean_power":0.0009668753842528969},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.561667844043168e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020949277098297487,"output_power":0.00019344457891967609,"reference_active":true,"delta_db":-0.34612477674650644},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8871602692351808e-05,"output_power":1.6422175245950197e-05,"reference_active":true,"delta_db":-0.6037810237970853},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293974094910672e-07,"output_power":4.3722393092769474e-07,"reference_active":true,"delta_db":0.07844504460347682}],"active_band_count":3,"active_band_mae_db":0.3427836150490229}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":1.5470321788629469,"rms_envelope_nrmse":0.15723602619703986,"envelope_correlation":0.9823153005278715,"rms_gain_error_db":-0.45550318226086633,"reference_mean_power":0.00011346719055744517,"output_mean_power":0.00010216919784833012},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5656809677774876e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.231325141923398e-05,"output_power":2.4843916626167644e-05,"reference_active":true,"delta_db":0.4665720415672303},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9000473344355792e-06,"output_power":2.9019862095328644e-06,"reference_active":true,"delta_db":0.002902578142002731},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0009489505392986e-06,"output_power":1.349286967788665e-06,"reference_active":true,"delta_db":1.2969239714450227}],"active_band_count":3,"active_band_mae_db":0.5887995303847519}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":1.168687764381478,"rms_envelope_nrmse":0.13590508521306674,"envelope_correlation":0.9865104449003119,"rms_gain_error_db":-0.45552307668109915,"reference_mean_power":0.00011403279016460418,"output_mean_power":0.00010267801002624575},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.16052039626869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1030857564945275e-05,"output_power":2.3512932783767478e-05,"reference_active":true,"delta_db":0.48449820283458706},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3790687257240347e-06,"output_power":1.4137393720464837e-06,"reference_active":true,"delta_db":0.10783443357059408},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.278096740186842e-08,"output_power":9.16554744256062e-08,"reference_active":true,"delta_db":1.0014058582929506}],"active_band_count":3,"active_band_mae_db":0.5312461648993773}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":1.9893513866800738,"rms_envelope_nrmse":0.22727203477229513,"envelope_correlation":0.9676453752702072,"rms_gain_error_db":-0.6824576157197035,"reference_mean_power":2.180480752557149e-05,"output_mean_power":1.8634017364925112e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7142973306691606e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0022096022650048e-05,"output_power":9.793795862381725e-06,"reference_active":true,"delta_db":-0.10007511911524096},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6747496282854544e-06,"output_power":2.6842923251011254e-06,"reference_active":true,"delta_db":0.015466738218735742},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.037734768320821e-07,"output_power":8.416609862649002e-07,"reference_active":true,"delta_db":0.2000352556738473}],"active_band_count":3,"active_band_mae_db":0.10519237100260799}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":1.5248711157732084,"rms_envelope_nrmse":0.18511334260585194,"envelope_correlation":0.9785262956034766,"rms_gain_error_db":-0.6824825035683851,"reference_mean_power":2.191432140518644e-05,"output_mean_power":1.8727498741533162e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5673668617618725e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.014084411974247e-05,"output_power":9.858603252672199e-06,"reference_active":true,"delta_db":-0.1225871767820167},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.343672608691746e-06,"output_power":1.3230744463769571e-06,"reference_active":true,"delta_db":-0.06709182352834375},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.915436539179563e-08,"output_power":5.4932941939802336e-08,"reference_active":true,"delta_db":0.4827076482483107}],"active_band_count":3,"active_band_mae_db":0.22412888285289037}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0008","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.459876537322998,"predicted_mos":3.459876537322998,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"80a203cf8efdb39625608b2206a060d7b1602bf5e1549d07b87ed913c00d45a1","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.055388773000232},"started_at":"2026-10-13T04:56:49.748097+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008515696999893407,"native_input_samples":264600,"native_input_sha256":"734d00264cfa195eb1b7a784ad9f5187fb910b17123ecdc840ec955ebcce8a60","output_sha256":"9128a38a97728f80fee59db760e745225f875bdf53016d72cae16cb27bf2158e","output_samples":264704,"length_delta_samples":104,"analysis_s":0.22596667899961176,"synthesis_s":35.59502101700036,"total_s":35.820987695999975,"rtf":5.9701646159999955,"output_peak":0.40726590156555176,"output_rms":0.039487858088305884,"finished_at":"2026-10-13T04:57:25.592328+00:00","evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"9128a38a97728f80fee59db760e745225f875bdf53016d72cae16cb27bf2158e","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.065587840448725,"output_peak":0.4074434854800976},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.35919169654694,"spectral_convergence":0.5088041935298112,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.950543668614943,"spectral_convergence":0.3845538910525353,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.786140023360824,"spectral_convergence":0.4614119466399052,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.0003426533391770393,"delta_db":0.32207297057206785,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0007482861329048551,"delta_db":-0.874282854936807,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.0003257345557102626,"delta_db":-1.513192358627176,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00010224444180514515,"delta_db":-0.4490805419457772,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":3.498023992040729e-05,"delta_db":-0.44673395126721627,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":7.843723761408894e-06,"delta_db":-2.475051578053113,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0134023759003596,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7610261545954295,"pitch":null,"output_rms_dbfs":-28.06390589555519,"output_peak":0.4074434854800976,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.031958462840903},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7757477258771858,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04313760336580714,"rms_dbfs":-27.30287974095976,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.449441334882454,"mean":-9.814103444417318e-05,"zero_energy":false},"output":{"rms":0.03951888704504618,"rms_dbfs":-28.06390589555519,"peak":0.4074434854800976,"peak_dbfs":-7.798652432774489,"crest_factor_db":20.265253462780702,"mean":-5.8745100798911116e-05,"zero_energy":false},"rms_difference_db":-0.7610261545954309},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":8.45499623701014e-07,"output_rms_dbfs":-60.72886581353772,"output_zero_energy":false,"output_peak":0.009502450496192028,"output_energy_fraction":0.00010196032269504964,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":123,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":42,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.7162162162162162,"absolute_difference_cents_median":8.026400317447102,"absolute_difference_cents_p95":52.61775782841704,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016375851184643225,"log_rms_mae_db":2.0449304599003173,"rms_envelope_nrmse":0.14997845881162855,"envelope_correlation":0.9847538984631122,"rms_gain_error_db":-0.7609361017842948,"reference_mean_power":0.0018623503698243853,"output_mean_power":0.0015630316757455232},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.48847287729811e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002826666045552883,"output_power":0.00023126511187361297,"reference_active":true,"delta_db":-0.8716438106923098},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.155794557866922e-05,"output_power":4.437915937088383e-05,"reference_active":true,"delta_db":-1.4210504577598393},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.62934396131159e-06,"output_power":5.8598283195893225e-06,"reference_active":true,"delta_db":-1.6809288778690543}],"active_band_count":3,"active_band_mae_db":1.3245410487737344}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014981717790874295,"log_rms_mae_db":1.6120453793518021,"rms_envelope_nrmse":0.11936260179804381,"envelope_correlation":0.9926882101958365,"rms_gain_error_db":-0.7605469580526434,"reference_mean_power":0.0018713077892111094,"output_mean_power":0.0015706901835837546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.792361294396735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025909451544035513,"output_power":0.00021309925280967588,"reference_active":true,"delta_db":-0.8487629290956216},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9672561350575066e-05,"output_power":2.0868290079295973e-05,"reference_active":true,"delta_db":-1.5286817146029534},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579324289571115e-07,"output_power":5.372032632723966e-07,"reference_active":true,"delta_db":-1.4949184706986032}],"active_band_count":3,"active_band_mae_db":1.2907877047990592}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014318840727902982,"log_rms_mae_db":2.8312493160746,"rms_envelope_nrmse":0.18110168462151605,"envelope_correlation":0.9807825295090499,"rms_gain_error_db":-0.9080594859694692,"reference_mean_power":0.0010573416450716808,"output_mean_power":0.0008578461160676733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.91879361710777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021964908416761717,"output_power":0.00017099227006936273,"reference_active":true,"delta_db":-1.0875291882113522},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090309412479679e-05,"output_power":3.0970704008031427e-05,"reference_active":true,"delta_db":-1.2080508396621583},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.52813465615998e-06,"output_power":3.6386451788585598e-06,"reference_active":true,"delta_db":-0.9497962569198768}],"active_band_count":3,"active_band_mae_db":1.081792094931129}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013022613407710025,"log_rms_mae_db":2.160919819326525,"rms_envelope_nrmse":0.1525716630476447,"envelope_correlation":0.9880688676983859,"rms_gain_error_db":-0.9080704749715474,"reference_mean_power":0.001062644583935512,"output_mean_power":0.0008621463334449847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.561667844043168e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020949277098297487,"output_power":0.00016362798731649268,"reference_active":true,"delta_db":-1.073114528173321},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8871602692351808e-05,"output_power":1.4101153305288846e-05,"reference_active":true,"delta_db":-1.2655415049458132},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293974094910672e-07,"output_power":3.543734254357445e-07,"reference_active":true,"delta_db":-0.8339827348667077}],"active_band_count":3,"active_band_mae_db":1.0575462559952806}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":3.153258759818782,"rms_envelope_nrmse":0.25672522175089674,"envelope_correlation":0.9500513911286008,"rms_gain_error_db":-0.407093354740598,"reference_mean_power":0.00011346719055744517,"output_mean_power":0.00010331442582129119},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5656809677774876e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.231325141923398e-05,"output_power":2.4340255013979582e-05,"reference_active":true,"delta_db":0.37762265073225376},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9000473344355792e-06,"output_power":3.1125511819940807e-06,"reference_active":true,"delta_db":0.30711415098865325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0009489505392986e-06,"output_power":9.308655866541055e-07,"reference_active":true,"delta_db":-0.3152495347802111}],"active_band_count":3,"active_band_mae_db":0.33332877883370604}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":2.49186280507833,"rms_envelope_nrmse":0.1886116049196811,"envelope_correlation":0.972059342697607,"rms_gain_error_db":-0.40693712861145026,"reference_mean_power":0.00011403279016460418,"output_mean_power":0.00010383315202874195},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.16052039626869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1030857564945275e-05,"output_power":2.291938715217907e-05,"reference_active":true,"delta_db":0.37346018670468606},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3790687257240347e-06,"output_power":1.4431063810490997e-06,"reference_active":true,"delta_db":0.19712437303874936},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.278096740186842e-08,"output_power":7.195796667026459e-08,"reference_active":true,"delta_db":-0.04938941008007215}],"active_band_count":3,"active_band_mae_db":0.20665798994116916}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":3.7703887178133817,"rms_envelope_nrmse":0.3176157437698995,"envelope_correlation":0.9603352908289929,"rms_gain_error_db":-2.302119490411972,"reference_mean_power":2.180480752557149e-05,"output_mean_power":1.2833357966441224e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7142973306691606e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0022096022650048e-05,"output_power":6.156676560895275e-06,"reference_active":true,"delta_db":-2.1161222077883406},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6747496282854544e-06,"output_power":1.5085794887596526e-06,"reference_active":true,"delta_db":-2.4871493725510048},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.037734768320821e-07,"output_power":5.892476235754063e-07,"reference_active":true,"delta_db":-1.3483583140926392}],"active_band_count":3,"active_band_mae_db":1.9838766314773284}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":3.353340534982731,"rms_envelope_nrmse":0.28329973234467853,"envelope_correlation":0.9765910781219573,"rms_gain_error_db":-2.3021411003733174,"reference_mean_power":2.191432140518644e-05,"output_mean_power":1.2897748878108328e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5673668617618725e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.014084411974247e-05,"output_power":6.112494359106e-06,"reference_active":true,"delta_db":-2.1985563565339743},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.343672608691746e-06,"output_power":7.29113453575437e-07,"reference_active":true,"delta_db":-2.654983520907845},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.915436539179563e-08,"output_power":4.6406253473600205e-08,"reference_active":true,"delta_db":-0.24985585667408938}],"active_band_count":3,"active_band_mae_db":1.7011319113719694}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0008","method":"golf","status":"measured","mos":{"status":"measured","value":2.483652114868164,"predicted_mos":2.483652114868164,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"e063e706c93e64fe4a090f1579af7b0fe9848c08d4f393be7db7603fbfea5b02","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.167095471000721},"started_at":"2026-10-13T05:01:47.903871+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004776212999786367,"native_input_samples":144000,"native_input_sha256":"c3ec24287ce5221a8f7341530bc84ef9b4145b0bca75d1a9c8bca35d214daad8","output_sha256":"9440b9944951c57c8d96b02841521abc40deb66dcd2c15bf838574a05bd944c6","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.781775442000253,"synthesis_s":0.2129700760006017,"total_s":0.9947455180008546,"rtf":0.16579091966680912,"output_peak":0.3779416084289551,"output_rms":0.03732813695503369,"finished_at":"2026-10-13T05:01:48.911815+00:00","analysis_stages":{"external_f0_s":0.09211771799982671},"evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"9440b9944951c57c8d96b02841521abc40deb66dcd2c15bf838574a05bd944c6","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.55461251762454,"output_peak":0.37479817643673435},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.721675132366528,"spectral_convergence":0.7442540399402032,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.404028603371511,"spectral_convergence":0.3300183616844286,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.677192660551974,"spectral_convergence":0.2914694348186892,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031853654548682055,"output_power":0.00027024008106314495,"delta_db":-0.7140950334700168,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000916613571267429,"output_power":0.0006934279047475076,"delta_db":-1.2118496862793149,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046228310413536203,"output_power":0.00032207992923608697,"delta_db":-1.5694435868624967,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011356955058880547,"output_power":7.26932193849009e-05,"delta_db":-1.9376800421723914,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.883339980496069e-05,"output_power":2.6072152539563236e-05,"delta_db":-1.7302852565217393,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.389104439995238e-05,"output_power":1.0372818409462538e-05,"delta_db":-1.2683812441508109,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4052891415761284,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.2584359894732118,"pitch":null,"output_rms_dbfs":-28.55461251762454,"output_peak":0.37479817643673435,"reference_rms_dbfs":-27.296176528151328,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.267632132096671},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8834410953574674,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043170907051894684,"rms_dbfs":-27.296176528151328,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.442738122074026,"mean":-0.00010490194584969089,"zero_energy":false},"output":{"rms":0.037348174056848704,"rms_dbfs":-28.55461251762454,"peak":0.37479817643673435,"peak_dbfs":-8.524050616378373,"crest_factor_db":20.03056190124617,"mean":0.00037464189864251434,"zero_energy":false},"rms_difference_db":-1.2584359894732131},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":1.2221380286249737e-06,"output_rms_dbfs":-59.128797419781314,"output_zero_energy":false,"output_peak":0.005835488275465054,"output_energy_fraction":0.00016528488154661899,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":173,"both_estimated_voiced_frames":140,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":33,"paired_voiced_fraction":0.9459459459459459,"absolute_difference_cents_median":3.9372418344616937,"absolute_difference_cents_p95":15.769540424202127,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00016375851184643225,"log_rms_mae_db":2.1737447250726447,"rms_envelope_nrmse":0.19353659629552777,"envelope_correlation":0.9789849296235908,"rms_gain_error_db":-1.2583135773613932,"reference_mean_power":0.001865217017158461,"output_mean_power":0.0013960404797276169},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.494658224416831e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028289899966995425,"output_power":0.0001952178859249553,"reference_active":true,"delta_db":-1.6111180644318206},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.167352050588308e-05,"output_power":4.6840740769558547e-05,"reference_active":true,"delta_db":-1.1947498527333542},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.6469513403528e-06,"output_power":7.30910201952395e-06,"reference_active":true,"delta_db":-0.7299899129024584}],"active_band_count":3,"active_band_mae_db":1.1786192766892112}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014981717790874295,"log_rms_mae_db":1.546416802207027,"rms_envelope_nrmse":0.15531699779670918,"envelope_correlation":0.9929830238716343,"rms_gain_error_db":-1.2586193758447073,"reference_mean_power":0.0018741181017081256,"output_mean_power":0.0014026038215872677},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.797514743247486e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002593012564299891,"output_power":0.0001788937140299906,"reference_active":true,"delta_db":-1.6120954053504803},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.972383619764963e-05,"output_power":2.281715334800832e-05,"reference_active":true,"delta_db":-1.1484339813870366},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.576921855843926e-07,"output_power":6.295638852655123e-07,"reference_active":true,"delta_db":-0.8045300130185228}],"active_band_count":3,"active_band_mae_db":1.1883531332520132}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014318840727902982,"log_rms_mae_db":2.1114886126948904,"rms_envelope_nrmse":0.20781984490406769,"envelope_correlation":0.9812960135085808,"rms_gain_error_db":-1.3860992917214754,"reference_mean_power":0.0010591011572958088,"output_mean_power":0.0007697106814875624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.92383736639673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021992210709032183,"output_power":0.0001435892056559416,"reference_active":true,"delta_db":-1.8514709489507288},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.098613181709365e-05,"output_power":2.8093457427173447e-05,"reference_active":true,"delta_db":-1.640317418948244},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.53577169298986e-06,"output_power":3.519008186078134e-06,"reference_active":true,"delta_db":-1.1023090921145546}],"active_band_count":3,"active_band_mae_db":1.531365820004509}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013022613407710025,"log_rms_mae_db":1.614190557538622,"rms_envelope_nrmse":0.17263009610730673,"envelope_correlation":0.9926029996810185,"rms_gain_error_db":-1.3861802900038136,"reference_mean_power":0.0010644025499925768,"output_mean_power":0.0007735490860110016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.5662418607459644e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020977816334470865,"output_power":0.0001370296823182384,"reference_active":true,"delta_db":-1.8494562786109425},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8889943180592067e-05,"output_power":1.3456969393698844e-05,"reference_active":true,"delta_db":-1.4728338694292074},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293207185912101e-07,"output_power":3.0215747637963985e-07,"reference_active":true,"delta_db":-1.5254850299976375}],"active_band_count":3,"active_band_mae_db":1.6159250593459291}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":2.22507854646307,"rms_envelope_nrmse":0.27285976611058854,"envelope_correlation":0.9636087017170645,"rms_gain_error_db":-1.8109782694209888,"reference_mean_power":0.00011365476805829702,"output_mean_power":7.490138237340286e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.5732596317060364e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.236775893101711e-05,"output_power":1.3607317169688872e-05,"reference_active":true,"delta_db":-2.1584996570194974},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.896023132692235e-06,"output_power":1.9144743496728534e-06,"reference_active":true,"delta_db":-1.7975247457234023},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.002182612707456e-06,"output_power":8.828226744503379e-07,"reference_active":true,"delta_db":-0.5507338464361623}],"active_band_count":3,"active_band_mae_db":1.502252749726354}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":1.7795881001449734,"rms_envelope_nrmse":0.22013217691785672,"envelope_correlation":0.9869929233742237,"rms_gain_error_db":-1.811171239731963,"reference_mean_power":0.00011420722341620208,"output_mean_power":7.526212030280278e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.166207501027228e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1074310941155912e-05,"output_power":1.2973685564407375e-05,"reference_active":true,"delta_db":-2.106900155834345},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3762062171604796e-06,"output_power":9.63386295207984e-07,"reference_active":true,"delta_db":-1.548830515841432},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.273730303412874e-08,"output_power":5.833836720602918e-08,"reference_active":true,"delta_db":-0.9580292439904734}],"active_band_count":3,"active_band_mae_db":1.5379199718887502}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":2.6869018148284676,"rms_envelope_nrmse":0.2579228309086577,"envelope_correlation":0.9691527385317468,"rms_gain_error_db":-1.5101120754388484,"reference_mean_power":2.184113579005984e-05,"output_mean_power":1.5426379510283802e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.718883175560183e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0059937656551993e-05,"output_power":7.3873839059224495e-06,"reference_active":true,"delta_db":-1.3410462039197226},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.678574829561279e-06,"output_power":1.871068105037201e-06,"reference_active":true,"delta_db":-1.558141877947363},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.028335881301012e-07,"output_power":7.024928305401483e-07,"reference_active":true,"delta_db":-0.5798363746089427}],"active_band_count":3,"active_band_mae_db":1.1596748188253427}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":2.389339853984391,"rms_envelope_nrmse":0.21327891173518568,"envelope_correlation":0.9841651122602411,"rms_gain_error_db":-1.5101288990380655,"reference_mean_power":2.1950397287763947e-05,"output_mean_power":1.550349077540307e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5714726514935125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0172705008993022e-05,"output_power":7.4815614121250075e-06,"reference_active":true,"delta_db":-1.3344420570709365},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3463006927983342e-06,"output_power":9.466691541928206e-07,"reference_active":true,"delta_db":-1.5294384276990804},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884588713096465e-08,"output_power":4.7833392581003895e-08,"reference_active":true,"delta_db":-0.09096817316745899}],"active_band_count":3,"active_band_mae_db":0.9849495526458254}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0008","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.510982036590576,"predicted_mos":3.510982036590576,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"e5371d76351e4844d42c67e40a7f47f7b612435e544f7200c0847df63779c286","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.972312188001524},"started_at":"2026-10-13T05:03:22.742426+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009725375000016356,"native_input_samples":144000,"native_input_sha256":"c3ec24287ce5221a8f7341530bc84ef9b4145b0bca75d1a9c8bca35d214daad8","output_sha256":"a5116013baec909c42cdeb9aed5a6360d666ebbd826051a427318ee7e1af7446","output_samples":144120,"length_delta_samples":120,"analysis_s":2.6489300310004182,"synthesis_s":15.960100782000154,"total_s":18.609030813000572,"rtf":3.1015051355000955,"output_peak":0.4264642596244812,"output_rms":0.043730898368012576,"finished_at":"2026-10-13T05:03:41.377489+00:00","analysis_stages":{"external_f0_s":1.7218790360002458},"evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"a5116013baec909c42cdeb9aed5a6360d666ebbd826051a427318ee7e1af7446","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.17996638563143,"output_peak":0.42264523968445494},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.995798347624888,"spectral_convergence":0.6470443946537844,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.337176367370257,"spectral_convergence":0.33976719270377753,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.466728487480336,"spectral_convergence":0.3140396906691245,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.0004661248082727362,"delta_db":1.6585454639242239,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0008640679474027403,"delta_db":-0.24948087247329948,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.0004230059745630194,"delta_db":-0.37836567793928577,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00011070289730562313,"delta_db":-0.10388775132277017,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":3.935338846759477e-05,"delta_db":0.0648595362133784,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.2283420207502926e-05,"delta_db":-0.5270813915722266,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.49703678224086406,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.1257884483451213,"pitch":null,"output_rms_dbfs":-27.177091292614634,"output_peak":0.42264523968445494,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.93323440082516},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8985928313979544,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04313760336580714,"rms_dbfs":-27.30287974095976,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.449441334882454,"mean":-9.814103444417318e-05,"zero_energy":false},"output":{"rms":0.043766864592054304,"rms_dbfs":-27.177091292614634,"peak":0.42264523968445494,"peak_dbfs":-7.480480363669164,"crest_factor_db":19.69661092894547,"mean":-1.2077322010169505e-05,"zero_energy":false},"rms_difference_db":0.12578844834512637},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":2.9345416261927072e-06,"output_rms_dbfs":-55.32459725680966,"output_zero_energy":false,"output_peak":0.05819688639508631,"output_energy_fraction":0.00028852044721019873,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":174,"both_estimated_voiced_frames":137,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9256756756756757,"absolute_difference_cents_median":5.133281149325519,"absolute_difference_cents_p95":19.321754934925305,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016375851184643225,"log_rms_mae_db":2.8818315854536785,"rms_envelope_nrmse":0.15767390317794397,"envelope_correlation":0.9768403748439517,"rms_gain_error_db":0.12573443325527142,"reference_mean_power":0.0018623503698243853,"output_mean_power":0.0019170561474526333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.48847287729811e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002826666045552883,"output_power":0.00029853273093338815,"reference_active":true,"delta_db":0.23717451730910355},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.155794557866922e-05,"output_power":5.678491983862306e-05,"reference_active":true,"delta_db":-0.35051100311213734},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.62934396131159e-06,"output_power":6.748776483337247e-06,"reference_active":true,"delta_db":-1.0675273536642358}],"active_band_count":3,"active_band_mae_db":0.5517376246951589}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014981717790874295,"log_rms_mae_db":2.0928787755132414,"rms_envelope_nrmse":0.09654391764757914,"envelope_correlation":0.9907334862590874,"rms_gain_error_db":0.1252484594119862,"reference_mean_power":0.0018713077892111094,"output_mean_power":0.0019260611498010832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.792361294396735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025909451544035513,"output_power":0.000273462041735904,"reference_active":true,"delta_db":0.23438832168179854},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9672561350575066e-05,"output_power":2.6529891483261746e-05,"reference_active":true,"delta_db":-0.4861956288902134},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579324289571115e-07,"output_power":5.57348824122717e-07,"reference_active":true,"delta_db":-1.3350340001733276}],"active_band_count":3,"active_band_mae_db":0.6852059835817799}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014318840727902982,"log_rms_mae_db":2.419198179199329,"rms_envelope_nrmse":0.16949876615339904,"envelope_correlation":0.9787830211934485,"rms_gain_error_db":-0.36102597437417233,"reference_mean_power":0.0010573416450716808,"output_mean_power":0.0009729997764345496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.91879361710777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021964908416761717,"output_power":0.00020130143354607835,"reference_active":true,"delta_db":-0.3788252912316273},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090309412479679e-05,"output_power":3.304028890696162e-05,"reference_active":true,"delta_db":-0.9271232539372767},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.52813465615998e-06,"output_power":3.2857928233237947e-06,"reference_active":true,"delta_db":-1.39279156564865}],"active_band_count":3,"active_band_mae_db":0.8995800369391848}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013022613407710025,"log_rms_mae_db":1.7395999849474528,"rms_envelope_nrmse":0.1091960932174776,"envelope_correlation":0.9914678189549653,"rms_gain_error_db":-0.3610078243822079,"reference_mean_power":0.001062644583935512,"output_mean_power":0.0009778837980258187},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.561667844043168e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020949277098297487,"output_power":0.00019132384689250136,"reference_active":true,"delta_db":-0.39399936738036656},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8871602692351808e-05,"output_power":1.5209618987669077e-05,"reference_active":true,"delta_db":-0.9369044989699982},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293974094910672e-07,"output_power":2.957004878604723e-07,"reference_active":true,"delta_db":-1.6200737900965905}],"active_band_count":3,"active_band_mae_db":0.9836592188156518}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":2.2676008431456944,"rms_envelope_nrmse":0.2180101849087167,"envelope_correlation":0.9639670467600056,"rms_gain_error_db":-0.037459572001415316,"reference_mean_power":0.00011346719055744517,"output_mean_power":0.00011249270104572883},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5656809677774876e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.231325141923398e-05,"output_power":2.3452722350364367e-05,"reference_active":true,"delta_db":0.2163040314140011},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9000473344355792e-06,"output_power":2.7645730821656506e-06,"reference_active":true,"delta_db":-0.2077701138345736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0009489505392986e-06,"output_power":1.0507816035753878e-06,"reference_active":true,"delta_db":0.21100532261662003}],"active_band_count":3,"active_band_mae_db":0.2116931559550649}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":1.3967030880727178,"rms_envelope_nrmse":0.11341273630317793,"envelope_correlation":0.9895924604658012,"rms_gain_error_db":-0.03740450794599324,"reference_mean_power":0.00011403279016460418,"output_mean_power":0.00011305487652757225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.16052039626869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1030857564945275e-05,"output_power":2.220899603512519e-05,"reference_active":true,"delta_db":0.2367194450909722},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3790687257240347e-06,"output_power":1.3089710936083663e-06,"reference_active":true,"delta_db":-0.22655853723870267},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.278096740186842e-08,"output_power":8.278183861148248e-08,"reference_active":true,"delta_db":0.5591724407872612}],"active_band_count":3,"active_band_mae_db":0.34081680770564543}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":2.339015351708774,"rms_envelope_nrmse":0.2695825112923994,"envelope_correlation":0.9517582221772104,"rms_gain_error_db":-0.39101795137590467,"reference_mean_power":2.180480752557149e-05,"output_mean_power":1.99273919188147e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7142973306691606e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0022096022650048e-05,"output_power":9.115800188070044e-06,"reference_active":true,"delta_db":-0.4116376225988986},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6747496282854544e-06,"output_power":2.1496492640909774e-06,"reference_active":true,"delta_db":-0.9491552945376381},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.037734768320821e-07,"output_power":5.728455497281127e-07,"reference_active":true,"delta_db":-1.4709612757416237}],"active_band_count":3,"active_band_mae_db":0.9439180642927202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":1.4045200747861386,"rms_envelope_nrmse":0.15577960497402688,"envelope_correlation":0.9837614416109791,"rms_gain_error_db":-0.3910246266987039,"reference_mean_power":2.191432140518644e-05,"output_mean_power":2.0027445761307067e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5673668617618725e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.014084411974247e-05,"output_power":9.011727768022232e-06,"reference_active":true,"delta_db":-0.512660431908313},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.343672608691746e-06,"output_power":1.0272515563453945e-06,"reference_active":true,"delta_db":-1.1661665607073903},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.915436539179563e-08,"output_power":3.3585482774113046e-08,"reference_active":true,"delta_db":-1.6541049806821917}],"active_band_count":3,"active_band_mae_db":1.1109773244326318}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0008","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.599886655807495,"predicted_mos":2.599886655807495,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"ce50198d3592304229d8cbd1cf55f58e0dfb172abee3b821424aead0865583b2","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.98618662500121},"started_at":"2026-10-13T05:09:52.193482+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005784803999631549,"native_input_samples":144000,"native_input_sha256":"c3ec24287ce5221a8f7341530bc84ef9b4145b0bca75d1a9c8bca35d214daad8","output_sha256":"aecdf1f9fc1024df41780ac0dc77448689f9e91693b60457e36325680df32cee","output_samples":144240,"length_delta_samples":240,"analysis_s":1.3422592360002454,"synthesis_s":0.14344884199999797,"total_s":1.4857080780002434,"rtf":0.24761801300004058,"output_peak":0.4055539071559906,"output_rms":0.04207628729733093,"finished_at":"2026-10-13T05:09:53.699834+00:00","analysis_stages":{"external_f0_s":0.19047539299936034},"evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"aecdf1f9fc1024df41780ac0dc77448689f9e91693b60457e36325680df32cee","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.514510471763664,"output_peak":0.4003161001990831},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.088200820395201,"spectral_convergence":0.19678893825780003,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.456661296201002,"spectral_convergence":0.17956924907920976,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.751360776029057,"spectral_convergence":0.19486542218820624,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.0003114888971163498,"delta_db":-0.09205100305422703,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0008824077486958395,"delta_db":-0.15826669405935687,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.00042807582362641797,"delta_db":-0.3266236827179955,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00010727509906440584,"delta_db":-0.2404883821426183,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":3.3715922741476935e-05,"delta_db":-0.6066112899343887,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.2333356586830537e-05,"delta_db":-0.5094616000641293,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3222504419954526,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20441043951679558,"pitch":null,"output_rms_dbfs":-27.507290180476556,"output_peak":0.4003161001990831,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.765407630875086},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9381748488518014,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04313760336580714,"rms_dbfs":-27.30287974095976,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.449441334882454,"mean":-9.814103444417318e-05,"zero_energy":false},"output":{"rms":0.042134271654216605,"rms_dbfs":-27.507290180476556,"peak":0.4003161001990831,"peak_dbfs":-7.951938855554244,"crest_factor_db":19.555351324922313,"mean":7.364839409335604e-05,"zero_energy":false},"rms_difference_db":-0.2044104395167956},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":1.8934035253268145e-06,"output_rms_dbfs":-57.2275681860527,"output_zero_energy":false,"output_peak":0.018221655193770236,"output_energy_fraction":0.0002008627445632488,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":134,"both_estimated_voiced_frames":127,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.8581081081081081,"absolute_difference_cents_median":0.6174165102027334,"absolute_difference_cents_p95":4.70123310859422,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016375851184643225,"log_rms_mae_db":1.1879260421711586,"rms_envelope_nrmse":0.08120825217440968,"envelope_correlation":0.994111953524513,"rms_gain_error_db":-0.20432182812387148,"reference_mean_power":0.0018623503698243853,"output_mean_power":0.0017767617930555019},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.48847287729811e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002826666045552883,"output_power":0.00027398421004502365,"reference_active":true,"delta_db":-0.13548967345620108},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.155794557866922e-05,"output_power":4.951336562318277e-05,"reference_active":true,"delta_db":-0.9456166927022353},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.62934396131159e-06,"output_power":6.53078850056456e-06,"reference_active":true,"delta_db":-1.2101216072218492}],"active_band_count":3,"active_band_mae_db":0.7637426577934286}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014981717790874295,"log_rms_mae_db":0.6869154704384722,"rms_envelope_nrmse":0.0506288898589673,"envelope_correlation":0.9978713364060952,"rms_gain_error_db":-0.20415043013818898,"reference_mean_power":0.0018713077892111094,"output_mean_power":0.0017853780137792486},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.792361294396735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025909451544035513,"output_power":0.0002516365088207196,"reference_active":true,"delta_db":-0.12684568661467438},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9672561350575066e-05,"output_power":2.4192564079480613e-05,"reference_active":true,"delta_db":-0.8867313640448136},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579324289571115e-07,"output_power":4.2452933764458713e-07,"reference_active":true,"delta_db":-2.517227811271279}],"active_band_count":3,"active_band_mae_db":1.1769349539769223}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014318840727902982,"log_rms_mae_db":1.5130749732464137,"rms_envelope_nrmse":0.10319790913672111,"envelope_correlation":0.9923355048118536,"rms_gain_error_db":-0.24007977010305231,"reference_mean_power":0.0010573416450716808,"output_mean_power":0.0010004775801406107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.91879361710777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021964908416761717,"output_power":0.00021580825757870062,"reference_active":true,"delta_db":-0.07661338512746088},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090309412479679e-05,"output_power":2.9971668824710792e-05,"reference_active":true,"delta_db":-1.3504523639741182},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.52813465615998e-06,"output_power":4.724806249066988e-06,"reference_active":true,"delta_db":0.18464670745902204}],"active_band_count":3,"active_band_mae_db":0.5372374855202003}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013022613407710025,"log_rms_mae_db":0.9279015680198629,"rms_envelope_nrmse":0.06762200928707057,"envelope_correlation":0.996853564636889,"rms_gain_error_db":-0.24008776505912505,"reference_mean_power":0.001062644583935512,"output_mean_power":0.0010054934747711543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.561667844043168e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020949277098297487,"output_power":0.00020501311527911984,"reference_active":true,"delta_db":-0.09387396244974712},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8871602692351808e-05,"output_power":1.350811472067692e-05,"reference_active":true,"delta_db":-1.4521404436290144},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293974094910672e-07,"output_power":3.293631680095853e-07,"reference_active":true,"delta_db":-1.151843887451923}],"active_band_count":3,"active_band_mae_db":0.8992860978435614}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":1.5676573622566392,"rms_envelope_nrmse":0.12719423948075584,"envelope_correlation":0.9886002749429217,"rms_gain_error_db":-0.3592140850114379,"reference_mean_power":0.00011346719055744517,"output_mean_power":0.00010445972865630838},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5656809677774876e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.231325141923398e-05,"output_power":1.829083348154729e-05,"reference_active":true,"delta_db":-0.8632936299469458},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9000473344355792e-06,"output_power":2.1390819731799214e-06,"reference_active":true,"delta_db":-1.3217765871332792},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0009489505392986e-06,"output_power":8.947671712731203e-07,"reference_active":true,"delta_db":-0.4870188697523265}],"active_band_count":3,"active_band_mae_db":0.8906963622775171}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":0.9275154966299815,"rms_envelope_nrmse":0.08473895835423813,"envelope_correlation":0.9954472749035062,"rms_gain_error_db":-0.3592513423381438,"reference_mean_power":0.00011403279016460418,"output_mean_power":0.00010497952818778199},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.16052039626869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1030857564945275e-05,"output_power":1.7282779861002732e-05,"reference_active":true,"delta_db":-0.8524338389429241},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3790687257240347e-06,"output_power":1.0273943903237481e-06,"reference_active":true,"delta_db":-1.2784871962498134},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.278096740186842e-08,"output_power":5.855240768901322e-08,"reference_active":true,"delta_db":-0.9447306584585955}],"active_band_count":3,"active_band_mae_db":1.025217231217111}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":1.7887358450407562,"rms_envelope_nrmse":0.13493953704588546,"envelope_correlation":0.9889187052978876,"rms_gain_error_db":-0.3876489342927648,"reference_mean_power":2.180480752557149e-05,"output_mean_power":1.9942856488001354e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7142973306691606e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0022096022650048e-05,"output_power":9.008492294642164e-06,"reference_active":true,"delta_db":-0.46306448031928926},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6747496282854544e-06,"output_power":2.097783123032592e-06,"reference_active":true,"delta_db":-1.0552254870861408},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.037734768320821e-07,"output_power":4.616968542717445e-07,"reference_active":true,"delta_db":-2.407767555712095}],"active_band_count":3,"active_band_mae_db":1.308685841039175}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":1.1339688950858562,"rms_envelope_nrmse":0.08271306337276813,"envelope_correlation":0.9964230755350235,"rms_gain_error_db":-0.3876955482624693,"reference_mean_power":2.191432140518644e-05,"output_mean_power":2.00428036581148e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5673668617618725e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.014084411974247e-05,"output_power":8.93514915169194e-06,"reference_active":true,"delta_db":-0.5497230055196363},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.343672608691746e-06,"output_power":1.0039815336084014e-06,"reference_active":true,"delta_db":-1.2656773915118213},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.915436539179563e-08,"output_power":2.6170314285376408e-08,"reference_active":true,"delta_db":-2.7375315529414084}],"active_band_count":3,"active_band_mae_db":1.5176439833242885}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0008","method":"magphase","status":"measured","mos":{"status":"measured","value":3.4394965171813965,"predicted_mos":3.4394965171813965,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":291025,"input_sample_rate":48000,"input_seconds":6.063020833333334,"input_float32_sha256":"953c5896bea71866fdaff3b36944ba29b075d9a6ef98b2305242af792ead080e","internal_samples":97009,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.516730456000005},"started_at":"2026-10-13T05:11:47.644269+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.013587679000920616,"native_input_samples":288000,"native_input_sha256":"d351c22a986aa5cde9097af758a12fac70a661cea60f94122eddb607230ad2b2","output_sha256":"40fccb70a6530f1b50d9c543a4a208570c0b0abd8a0861ac17c2bccb3b135e14","output_samples":291025,"length_delta_samples":3025,"analysis_s":2.9725437340002827,"synthesis_s":1.1500163799992151,"total_s":4.122560113999498,"rtf":0.6870933523332496,"output_peak":0.46946272109766585,"output_rms":0.04165982363326483,"finished_at":"2026-10-13T05:11:51.793312+00:00","evaluation_input":{"reference":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"40fccb70a6530f1b50d9c543a4a208570c0b0abd8a0861ac17c2bccb3b135e14","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":97009,"length_delta_samples":1009,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.60089211723185,"output_peak":0.46731955922069124},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9895989031945489,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":11.338668046167236,"spectral_convergence":0.8542004667233347,"spectral_convergence_status":"ok","amplitude_floor":1.4536470931091146e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":10.252782022211244,"spectral_convergence":0.6437076014961385,"spectral_convergence_status":"ok","amplitude_floor":9.268471977763907e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.464804146315858,"spectral_convergence":0.6680255217189102,"spectral_convergence_status":"ok","amplitude_floor":5.673364127394706e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031816153358243563,"output_power":0.00028168153628666724,"delta_db":-0.5288929063836048,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009151577935124143,"output_power":0.0008570445437873968,"delta_db":-0.28492588122540785,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046151207587702967,"output_power":0.0004470076195437244,"delta_db":-0.13868143180073395,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011338295265990407,"output_power":0.00011168648151095324,"delta_db":-0.06547153014775695,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877003374326731e-05,"output_power":4.0110104655452545e-05,"delta_db":0.14757616153117695,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3868434770644032e-05,"output_power":1.444256281205994e-05,"delta_db":0.17616816825075216,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2236193465565721,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26184654549241865,"pitch":null,"output_rms_dbfs":-27.564726286452178,"output_peak":0.46731955922069124,"reference_rms_dbfs":-27.30287974095976,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":10.352084738231445},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.37946428458223436,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04313760336580714,"rms_dbfs":-27.30287974095976,"peak":0.45428466796875,"peak_dbfs":-6.853438406077304,"crest_factor_db":20.449441334882454,"mean":-9.814103444417318e-05,"zero_energy":false},"output":{"rms":0.041856574735592664,"rms_dbfs":-27.564726286452178,"peak":0.46731955922069124,"peak_dbfs":-6.607720832162986,"crest_factor_db":20.95700545428919,"mean":-0.0002864876314828518,"zero_energy":false},"rms_difference_db":-0.2618465454924177},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.681685020215809e-06,"threshold_dbfs":-55.71592233947322,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":1.0167871832649792e-06,"output_mean_power":0.00014092949096434853,"output_rms_dbfs":-38.50998116735142,"output_zero_energy":false,"output_peak":0.13454588320507505,"output_energy_fraction":0.015149618796633665,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":148,"output_estimated_voiced_frames":147,"both_estimated_voiced_frames":97,"reference_voiced_output_unvoiced_frames":51,"reference_unvoiced_output_voiced_frames":50,"paired_voiced_fraction":0.6554054054054054,"absolute_difference_cents_median":16.69406439678167,"absolute_difference_cents_p95":115.82319680915968,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016375851184643225,"log_rms_mae_db":8.015679463152944,"rms_envelope_nrmse":0.48775450317668634,"envelope_correlation":0.7708482515607503,"rms_gain_error_db":-0.2617299812386039,"reference_mean_power":0.0018623503698243853,"output_mean_power":0.001753429832958464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.48847287729811e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002826666045552883,"output_power":0.00026923504535104106,"reference_active":true,"delta_db":-0.21142912336212194},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.155794557866922e-05,"output_power":5.818940149684933e-05,"reference_active":true,"delta_db":-0.24440227042934315},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.62934396131159e-06,"output_power":6.83102082205416e-06,"reference_active":true,"delta_db":-1.0149217095735705}],"active_band_count":3,"active_band_mae_db":0.4902510344550119}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014981717790874295,"log_rms_mae_db":6.56080086152148,"rms_envelope_nrmse":0.4191205769081868,"envelope_correlation":0.8166760563209103,"rms_gain_error_db":-0.2609708118211617,"reference_mean_power":0.0018713077892111094,"output_mean_power":0.0017621713828122423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.792361294396735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025909451544035513,"output_power":0.0002478030858529768,"reference_active":true,"delta_db":-0.1935150958403637},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9672561350575066e-05,"output_power":2.8089433137781953e-05,"reference_active":true,"delta_db":-0.23812061497873954},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579324289571115e-07,"output_power":6.879044085741552e-07,"reference_active":true,"delta_db":-0.42102396499277317}],"active_band_count":3,"active_band_mae_db":0.28421989193729214}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014318840727902982,"log_rms_mae_db":10.292091946839372,"rms_envelope_nrmse":0.5272931982051035,"envelope_correlation":0.7856690581819793,"rms_gain_error_db":-0.2709626419424398,"reference_mean_power":0.0010573416450716808,"output_mean_power":0.0009933883757102704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.91879361710777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021964908416761717,"output_power":0.00020777869262373052,"reference_active":true,"delta_db":-0.24128387510946672},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090309412479679e-05,"output_power":3.718860008824373e-05,"reference_active":true,"delta_db":-0.4134633129385121},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.52813465615998e-06,"output_power":4.115904193113068e-06,"reference_active":true,"delta_db":-0.41454076245419247}],"active_band_count":3,"active_band_mae_db":0.3564293168340571}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013022613407710025,"log_rms_mae_db":9.024137349848685,"rms_envelope_nrmse":0.45929469693651187,"envelope_correlation":0.8274879351953965,"rms_gain_error_db":-0.27093423257595606,"reference_mean_power":0.001062644583935512,"output_mean_power":0.0009983770973785422},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.561667844043168e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020949277098297487,"output_power":0.0001983517412652561,"reference_active":true,"delta_db":-0.2373302390104174},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8871602692351808e-05,"output_power":1.7559454488645438e-05,"reference_active":true,"delta_db":-0.31297764928443433},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.293974094910672e-07,"output_power":4.1751125993220507e-07,"reference_active":true,"delta_db":-0.12191227516346222}],"active_band_count":3,"active_band_mae_db":0.224073387819438}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.227046231131277e-05,"log_rms_mae_db":8.7138384414567,"rms_envelope_nrmse":0.5737438085386322,"envelope_correlation":0.7536154782909763,"rms_gain_error_db":0.033669502992257426,"reference_mean_power":0.00011346719055744517,"output_mean_power":0.00011435028521821825},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5656809677774876e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.231325141923398e-05,"output_power":2.248302014361878e-05,"reference_active":true,"delta_db":0.032917906161683996},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9000473344355792e-06,"output_power":2.9837343620626866e-06,"reference_active":true,"delta_db":0.12355069365571233},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0009489505392986e-06,"output_power":1.0213337863160015e-06,"reference_active":true,"delta_db":0.0875577029573846}],"active_band_count":3,"active_band_mae_db":0.08134210092492698}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.58929337283286e-05,"log_rms_mae_db":7.4754090354878775,"rms_envelope_nrmse":0.488346234982934,"envelope_correlation":0.8092249416943483,"rms_gain_error_db":0.03350305288859667,"reference_mean_power":0.00011403279016460418,"output_mean_power":0.00011491588237098755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.16052039626869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1030857564945275e-05,"output_power":2.139552013802254e-05,"reference_active":true,"delta_db":0.07465866849479276},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3790687257240347e-06,"output_power":1.3704574584989226e-06,"reference_active":true,"delta_db":-0.02720350937116287},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.278096740186842e-08,"output_power":7.104470722058593e-08,"reference_active":true,"delta_db":-0.10486095071263017}],"active_band_count":3,"active_band_mae_db":0.06890770952619527}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9795267482338986e-05,"log_rms_mae_db":8.052923062747988,"rms_envelope_nrmse":0.761200403754682,"envelope_correlation":0.6349261586358872,"rms_gain_error_db":0.28673466169029516,"reference_mean_power":2.180480752557149e-05,"output_mean_power":2.3293015925293604e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7142973306691606e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0022096022650048e-05,"output_power":1.0792139899041644e-05,"reference_active":true,"delta_db":0.3214900703821517},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6747496282854544e-06,"output_power":2.9762306017125943e-06,"reference_active":true,"delta_db":0.4638344196858066},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.037734768320821e-07,"output_power":8.490858778610589e-07,"reference_active":true,"delta_db":0.23817946574454937}],"active_band_count":3,"active_band_mae_db":0.34116798527083586}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7747549191119275e-05,"log_rms_mae_db":6.867055056615679,"rms_envelope_nrmse":0.6073375705552723,"envelope_correlation":0.7515104603439368,"rms_gain_error_db":0.286377041671237,"reference_mean_power":2.191432140518644e-05,"output_mean_power":2.3408076660638144e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5673668617618725e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.014084411974247e-05,"output_power":1.1005419386025635e-05,"reference_active":true,"delta_db":0.3553248996542461},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.343672608691746e-06,"output_power":1.4416953084241761e-06,"reference_active":true,"delta_db":0.3058002118754318},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.915436539179563e-08,"output_power":5.672875796729162e-08,"reference_active":true,"delta_db":0.6224118161518757}],"active_band_count":3,"active_band_mae_db":0.4278456425605179}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"hnm","status":"measured","mos":{"status":"measured","value":2.261254072189331,"predicted_mos":2.261254072189331,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7582b41c3ab7b1f23aac94fce12946eabee336d8df3dafa01531e1b524b33c94","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.145232961000147},"started_at":"2026-10-13T04:37:55.828923+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.571099983266322e-05,"native_input_samples":96000,"native_input_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","output_sha256":"ad0df8d418a97b4cdbf7656f289c4079ac04f41be20c3a0e1ef04a6005d1302c","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2668584510001892,"synthesis_s":0.5577611449998585,"total_s":1.8246195960000477,"rtf":0.304103266000008,"output_peak":0.500779564938914,"output_rms":0.04971192125295543,"finished_at":"2026-10-13T04:37:57.661090+00:00","evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"ad0df8d418a97b4cdbf7656f289c4079ac04f41be20c3a0e1ef04a6005d1302c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.318521514609894,"spectral_convergence":0.24859356003857938,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.507437954435053,"spectral_convergence":0.24368293809344063,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.726527028988703,"spectral_convergence":0.25842325394677523,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.0003723777885167991,"delta_db":0.09302617541909329,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0008339140348844994,"delta_db":-0.08563880051048343,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.000988596029523961,"delta_db":-0.2752988027277143,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0002164794907061045,"delta_db":-0.06650868463484282,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.669166506530922e-05,"delta_db":-0.23414251531562735,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.3216105963367823e-05,"delta_db":-0.2193907807375475,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1623342932242181,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13834094153203133,"pitch":null,"output_rms_dbfs":-26.07078904080761,"output_peak":0.500779564938914,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.850828832677883},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.318521514609894,"spectral_convergence":0.24859356003857938,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.507437954435053,"spectral_convergence":0.24368293809344063,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.726527028988703,"spectral_convergence":0.25842325394677523,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.0003723777885167991,"delta_db":0.09302617541909329,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0008339140348844994,"delta_db":-0.08563880051048343,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.000988596029523961,"delta_db":-0.2752988027277143,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0002164794907061045,"delta_db":-0.06650868463484282,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.669166506530922e-05,"delta_db":-0.23414251531562735,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.3216105963367823e-05,"delta_db":-0.2193907807375475,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1623342932242181,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13834094153203133,"pitch":null,"output_rms_dbfs":-26.07078904080761,"output_peak":0.500779564938914,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.850828832677883},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8767845500971785,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050510026345542525,"rms_dbfs":-25.932448099275575,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.160600805855687,"mean":-9.429613749186198e-05,"zero_energy":false},"output":{"rms":0.04971192125295543,"rms_dbfs":-26.07078904080761,"peak":0.500779564938914,"peak_dbfs":-6.007068029472503,"crest_factor_db":20.063721011335105,"mean":1.608694773646245e-06,"zero_energy":false},"rms_difference_db":-0.1383409415320358},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":1.6381455900939396e-06,"output_rms_dbfs":-57.856475029635135,"output_zero_energy":false,"output_peak":0.015750131229225767,"output_energy_fraction":9.943119527907902e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":160,"both_estimated_voiced_frames":126,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9767441860465116,"absolute_difference_cents_median":1.5005193572476756,"absolute_difference_cents_p95":7.380151574313434,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019142337741905198,"log_rms_mae_db":1.0048357453418697,"rms_envelope_nrmse":0.0791220850682927,"envelope_correlation":0.9941571276135458,"rms_gain_error_db":-0.1383541082348617,"reference_mean_power":0.0025533227171819564,"output_mean_power":0.0024732629879980835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1748402838347296e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007147512141697323,"output_power":0.00070844463313905,"reference_active":true,"delta_db":-0.038489871566495776},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.175142428480134e-05,"output_power":8.174617936768401e-05,"reference_active":true,"delta_db":-0.0002786387704765861},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4312039590506752e-05,"output_power":1.3329056679291634e-05,"reference_active":true,"delta_db":-0.30902114203271885}],"active_band_count":3,"active_band_mae_db":0.11592988412323041}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016640454830055276,"log_rms_mae_db":0.5461995477716274,"rms_envelope_nrmse":0.04586001236183127,"envelope_correlation":0.9980469574233886,"rms_gain_error_db":-0.1380374870750621,"reference_mean_power":0.002564981061322552,"output_mean_power":0.0024847369258468605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0626539543337418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006440432662381922,"output_power":0.0006378044151576633,"reference_active":true,"delta_db":-0.04227522555001907},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1296244809426825e-05,"output_power":4.0961940426399714e-05,"reference_active":true,"delta_db":-0.035300401475053644},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.585289215752229e-07,"output_power":8.844363336186094e-07,"reference_active":true,"delta_db":0.12911645589633497}],"active_band_count":3,"active_band_mae_db":0.06889736097380256}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016776043588712122,"log_rms_mae_db":1.8023615760724023,"rms_envelope_nrmse":0.10764112532015786,"envelope_correlation":0.9915429392310849,"rms_gain_error_db":-0.20840609512966962,"reference_mean_power":0.0015150798695545982,"output_mean_power":0.0014440921871597308},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.503421266374217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005710315291680691,"output_power":0.0005460287737760272,"reference_active":true,"delta_db":-0.19444559137111256},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462070521392158e-05,"output_power":5.8078129395635166e-05,"reference_active":true,"delta_db":-0.4635907323668843},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.823790922205639e-06,"output_power":7.474093690015189e-06,"reference_active":true,"delta_db":-1.1872057333224564}],"active_band_count":3,"active_band_mae_db":0.6150806856868177}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015135634256581334,"log_rms_mae_db":1.1137416307790764,"rms_envelope_nrmse":0.06292507639085029,"envelope_correlation":0.9971986104386356,"rms_gain_error_db":-0.2083096038299171,"reference_mean_power":0.0015225180811750088,"output_mean_power":0.001451214130808158},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.758019890177841e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005253930575206644,"output_power":0.0005036654569650498,"reference_active":true,"delta_db":-0.18342163240100334},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0513708673138778e-05,"output_power":2.7880875581404565e-05,"reference_active":true,"delta_db":-0.3918858714232384},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.334657094067361e-07,"output_power":5.073378382958354e-07,"reference_active":true,"delta_db":-2.1559048261343237}],"active_band_count":3,"active_band_mae_db":0.9104041099861885}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":1.8236613213450097,"rms_envelope_nrmse":0.14136358315412803,"envelope_correlation":0.9854484095266136,"rms_gain_error_db":-0.16147072546468147,"reference_mean_power":0.00023682053730462446,"output_mean_power":0.0002281772226150329},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.163166626171004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014015200472863973,"output_power":0.00013383282474555604,"reference_active":true,"delta_db":-0.20036669767173926},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.313645741402795e-05,"output_power":1.4197690399180172e-05,"reference_active":true,"delta_db":0.33739439275079275},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2309159151114036e-06,"output_power":2.659275680091509e-06,"reference_active":true,"delta_db":-0.8456229378208295}],"active_band_count":3,"active_band_mae_db":0.46112800941445387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":0.9776601996483805,"rms_envelope_nrmse":0.08700782473431574,"envelope_correlation":0.9942915075616824,"rms_gain_error_db":-0.16182535088627828,"reference_mean_power":0.0002379882918616052,"output_mean_power":0.00022928363422246021},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.020973483325393e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013180546055965978,"output_power":0.00012648666654380304,"reference_active":true,"delta_db":-0.17888655749969534},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.132933188997126e-06,"output_power":7.02260737866907e-06,"reference_active":true,"delta_db":0.5883015488055927},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2319303983633038e-07,"output_power":1.756700430444417e-07,"reference_active":true,"delta_db":-1.0398293950789}],"active_band_count":3,"active_band_mae_db":0.6023391671280627}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":2.309787978793165,"rms_envelope_nrmse":0.18913176662114028,"envelope_correlation":0.9746448081413418,"rms_gain_error_db":-0.24391258848236763,"reference_mean_power":3.051701314781676e-05,"output_mean_power":2.8850328725811193e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.8263265386055575e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1311662288856238e-05,"output_power":1.0395638897433235e-05,"reference_active":true,"delta_db":-0.366752452760165},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.547549874304093e-06,"output_power":2.2749391667607317e-06,"reference_active":true,"delta_db":-0.4915290724844531},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.732573657220465e-07,"output_power":8.680335636761186e-07,"reference_active":true,"delta_db":1.8018687443608614}],"active_band_count":3,"active_band_mae_db":0.8867167565351598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":1.46734805401051,"rms_envelope_nrmse":0.12040579318245671,"envelope_correlation":0.9894759448421049,"rms_gain_error_db":-0.24400972599891171,"reference_mean_power":3.0668913834731314e-05,"output_mean_power":2.8993284875404938e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6363060360497885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1094596811060145e-05,"output_power":1.0250280980998163e-05,"reference_active":true,"delta_db":-0.34375753703495804},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.190115655321781e-06,"output_power":1.2731546887203595e-06,"reference_active":true,"delta_db":0.2929200564604888},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.417829557129175e-08,"output_power":6.879332036432777e-08,"reference_active":true,"delta_db":1.9233731504214355}],"active_band_count":3,"active_band_mae_db":0.8533502479722941}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.2870359420776367,"predicted_mos":3.2870359420776367,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"fc76c1d818bc42abc914b8cd5fffa1c22a5078bf198beee5c80e1c76c88f3dd6","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.151550289001534},"started_at":"2026-10-13T04:40:46.495774+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5206999453075696e-05,"native_input_samples":96000,"native_input_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","output_sha256":"df848eb287bde95da589c7047fe25d6c87cb2fa341f5aa2409108172e163e694","output_samples":96000,"length_delta_samples":0,"analysis_s":1.8106553629995688,"synthesis_s":0.0736490720000802,"total_s":1.884304434999649,"rtf":0.31405073916660814,"output_peak":0.6085349550348448,"output_rms":0.05466858862754457,"finished_at":"2026-10-13T04:40:48.384440+00:00","evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"df848eb287bde95da589c7047fe25d6c87cb2fa341f5aa2409108172e163e694","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.589808601697905,"spectral_convergence":0.7311184870419607,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.262162617634435,"spectral_convergence":0.3014193340821015,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.550950474632851,"spectral_convergence":0.2704386758912277,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.00037721408956055657,"delta_db":0.14906755106791883,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0010189034002856335,"delta_db":0.7844784838781209,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.001278512551611137,"delta_db":0.8415625385237778,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0002472292130335133,"delta_db":0.5103216045116645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":4.302951330304038e-05,"delta_db":0.4578476191574492,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.3765814733814e-05,"delta_db":-0.11775748124044313,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.476839213063229,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6872053374207926,"pitch":null,"output_rms_dbfs":-25.245242761854783,"output_peak":0.6085349550348448,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.13430723132173},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.589808601697905,"spectral_convergence":0.7311184870419607,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.262162617634435,"spectral_convergence":0.3014193340821015,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.550950474632851,"spectral_convergence":0.2704386758912277,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.00037721408956055657,"delta_db":0.14906755106791883,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0010189034002856335,"delta_db":0.7844784838781209,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.001278512551611137,"delta_db":0.8415625385237778,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0002472292130335133,"delta_db":0.5103216045116645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":4.302951330304038e-05,"delta_db":0.4578476191574492,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.3765814733814e-05,"delta_db":-0.11775748124044313,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.476839213063229,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6872053374207926,"pitch":null,"output_rms_dbfs":-25.245242761854783,"output_peak":0.6085349550348448,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.13430723132173},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9081680881436537,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050510026345542525,"rms_dbfs":-25.932448099275575,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.160600805855687,"mean":-9.429613749186198e-05,"zero_energy":false},"output":{"rms":0.05466858862754457,"rms_dbfs":-25.245242761854783,"peak":0.6085349550348448,"peak_dbfs":-4.314289405613815,"crest_factor_db":20.93095335624097,"mean":2.0443175315755697e-06,"zero_energy":false},"rms_difference_db":0.6872053374207923},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":1.0080493684008912e-06,"output_rms_dbfs":-59.96518198149357,"output_zero_energy":false,"output_peak":0.011527860720178891,"output_energy_fraction":5.059380436405198e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":150,"both_estimated_voiced_frames":119,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.9224806201550387,"absolute_difference_cents_median":3.9111977900847217,"absolute_difference_cents_p95":12.710784203990734,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019142337741905198,"log_rms_mae_db":2.0587356125328315,"rms_envelope_nrmse":0.1500258719024504,"envelope_correlation":0.9861820427866407,"rms_gain_error_db":0.687315048275378,"reference_mean_power":0.0025533227171819564,"output_mean_power":0.0029911432606695857},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1748402838347296e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007147512141697323,"output_power":0.0008228861198736005,"reference_active":true,"delta_db":0.6118483509381883},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.175142428480134e-05,"output_power":8.934822065407367e-05,"reference_active":true,"delta_db":0.385905803315621},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4312039590506752e-05,"output_power":1.0274350182657155e-05,"reference_active":true,"delta_db":-1.4394716513975419}],"active_band_count":3,"active_band_mae_db":0.8124086018837837}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016640454830055276,"log_rms_mae_db":1.5371463198860156,"rms_envelope_nrmse":0.10864340784582448,"envelope_correlation":0.9952701334879644,"rms_gain_error_db":0.6878713845495045,"reference_mean_power":0.002564981061322552,"output_mean_power":0.003005185614679098},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0626539543337418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006440432662381922,"output_power":0.0007419053769754433,"reference_active":true,"delta_db":0.6143347480477376},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1296244809426825e-05,"output_power":4.663080857770113e-05,"reference_active":true,"delta_db":0.5276238442053965},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.585289215752229e-07,"output_power":6.731381619648386e-07,"reference_active":true,"delta_db":-1.0565071743316437}],"active_band_count":3,"active_band_mae_db":0.7328219221949258}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016776043588712122,"log_rms_mae_db":2.3906976316559168,"rms_envelope_nrmse":0.18394532989174392,"envelope_correlation":0.984815673499722,"rms_gain_error_db":0.884573478221053,"reference_mean_power":0.0015150798695545982,"output_mean_power":0.0018573462603462497},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.503421266374217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005710315291680691,"output_power":0.0006731824768132035,"reference_active":true,"delta_db":0.7147271438548382},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462070521392158e-05,"output_power":7.17926578390656e-05,"reference_active":true,"delta_db":0.4570833844731871},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.823790922205639e-06,"output_power":5.174049554410636e-06,"reference_active":true,"delta_db":-2.7844852704036875}],"active_band_count":3,"active_band_mae_db":1.3187652662439042}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015135634256581334,"log_rms_mae_db":1.818623746397872,"rms_envelope_nrmse":0.14443280511274453,"envelope_correlation":0.9931486943246243,"rms_gain_error_db":0.8845788189254232,"reference_mean_power":0.0015225180811750088,"output_mean_power":0.001866467107594018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.758019890177841e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005253930575206644,"output_power":0.0006173753632245636,"reference_active":true,"delta_db":0.7006496490702708},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0513708673138778e-05,"output_power":3.3933783575700177e-05,"reference_active":true,"delta_db":0.46137290124557406},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.334657094067361e-07,"output_power":4.823485000658648e-07,"reference_active":true,"delta_db":-2.3752680435203892}],"active_band_count":3,"active_band_mae_db":1.179096864612078}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":1.839567367410513,"rms_envelope_nrmse":0.15624504729670324,"envelope_correlation":0.9859590293816859,"rms_gain_error_db":0.5246239847703539,"reference_mean_power":0.00023682053730462446,"output_mean_power":0.00026722787649150533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.163166626171004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014015200472863973,"output_power":0.00015693967203235732,"reference_active":true,"delta_db":0.49133426427527077},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.313645741402795e-05,"output_power":1.328615728044549e-05,"reference_active":true,"delta_db":0.04921127020758879},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2309159151114036e-06,"output_power":2.1412830547892376e-06,"reference_active":true,"delta_db":-1.7865157542102506}],"active_band_count":3,"active_band_mae_db":0.7756870962310366}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":1.1010576057294343,"rms_envelope_nrmse":0.09651466652273274,"envelope_correlation":0.9960447402861773,"rms_gain_error_db":0.5247581174215905,"reference_mean_power":0.0002379882918616052,"output_mean_power":0.00026855386288089964},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.020973483325393e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013180546055965978,"output_power":0.00014737062887586934,"reference_active":true,"delta_db":0.48477533804002926},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.132933188997126e-06,"output_power":6.5397246631893996e-06,"reference_active":true,"delta_db":0.27891230362926295},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2319303983633038e-07,"output_power":1.25618839211493e-07,"reference_active":true,"delta_db":-2.4962587125773488}],"active_band_count":3,"active_band_mae_db":1.0866487847488804}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":2.1587871350241365,"rms_envelope_nrmse":0.20222650911390422,"envelope_correlation":0.9712715306565816,"rms_gain_error_db":0.03534545999244237,"reference_mean_power":3.051701314781676e-05,"output_mean_power":3.0766392115666715e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.8263265386055575e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1311662288856238e-05,"output_power":1.2149864511862197e-05,"reference_active":true,"delta_db":0.3104500423656822},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.547549874304093e-06,"output_power":2.599450690012853e-06,"reference_active":true,"delta_db":0.0875888844582928},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.732573657220465e-07,"output_power":5.136501532569961e-07,"reference_active":true,"delta_db":-0.4768222161082202}],"active_band_count":3,"active_band_mae_db":0.2916203809773984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":1.2801312279782304,"rms_envelope_nrmse":0.12586206443616846,"envelope_correlation":0.9881186041343192,"rms_gain_error_db":0.03542971423384379,"reference_mean_power":3.0668913834731314e-05,"output_mean_power":3.092013395750925e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6363060360497885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1094596811060145e-05,"output_power":1.1947047964242913e-05,"reference_active":true,"delta_db":0.3214908312279987},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.190115655321781e-06,"output_power":1.2147275000104196e-06,"reference_active":true,"delta_db":0.08889695380466912},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.417829557129175e-08,"output_power":3.521529432779381e-08,"reference_active":true,"delta_db":-0.9847763392533749}],"active_band_count":3,"active_band_mae_db":0.46505470809534755}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"facodec","status":"measured","mos":{"status":"measured","value":3.937617063522339,"predicted_mos":3.937617063522339,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"362ab7edb09035ea720a849faf18f5bdf4c4f8a44798feab32667a21035f0743","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.029854278998755},"started_at":"2026-10-13T04:44:25.137742+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1546000425587408e-05,"native_input_samples":96000,"native_input_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","output_sha256":"a68ecf32e544fd5685846b03ac7f95f4b3e580e84f769d0e5cdb232d733f3882","output_samples":96000,"length_delta_samples":0,"analysis_s":2.7411877120002828,"synthesis_s":11.85379195799942,"total_s":14.594979669999702,"rtf":2.432496611666617,"output_peak":0.5266932845115662,"output_rms":0.049039854393836936,"finished_at":"2026-10-13T04:44:39.738443+00:00","evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"a68ecf32e544fd5685846b03ac7f95f4b3e580e84f769d0e5cdb232d733f3882","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.301660881171016,"spectral_convergence":0.2045804220755592,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.267237343455254,"spectral_convergence":0.21350534461550474,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.450184848375086,"spectral_convergence":0.23788003027369148,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.00037456119477280827,"delta_db":0.11841630796781981,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0007659487320954152,"delta_db":-0.4548546158361783,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.000976754351465596,"delta_db":-0.3276338745970768,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0002269566974356769,"delta_db":0.13875377201563852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.774783324276562e-05,"delta_db":-0.11089644175523705,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.293850995646585e-05,"delta_db":-0.27163251878122097,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2370312551588619,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.25656846009597417,"pitch":null,"output_rms_dbfs":-26.18901655937155,"output_peak":0.5266932845115662,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.339694357667118},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.301660881171016,"spectral_convergence":0.2045804220755592,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.267237343455254,"spectral_convergence":0.21350534461550474,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.450184848375086,"spectral_convergence":0.23788003027369148,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.00037456119477280827,"delta_db":0.11841630796781981,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0007659487320954152,"delta_db":-0.4548546158361783,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.000976754351465596,"delta_db":-0.3276338745970768,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0002269566974356769,"delta_db":0.13875377201563852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.774783324276562e-05,"delta_db":-0.11089644175523705,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.293850995646585e-05,"delta_db":-0.27163251878122097,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2370312551588619,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.25656846009597417,"pitch":null,"output_rms_dbfs":-26.18901655937155,"output_peak":0.5266932845115662,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.339694357667118},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8993729833735308,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050510026345542525,"rms_dbfs":-25.932448099275575,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.160600805855687,"mean":-9.429613749186198e-05,"zero_energy":false},"output":{"rms":0.049039854393836936,"rms_dbfs":-26.18901655937155,"peak":0.5266932845115662,"peak_dbfs":-5.568844379695385,"crest_factor_db":20.620172179676164,"mean":1.0272705412395074e-05,"zero_energy":false},"rms_difference_db":-0.25656846009597345},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":9.151228565574067e-07,"output_rms_dbfs":-60.38520597361713,"output_zero_energy":false,"output_peak":0.009482362307608128,"output_energy_fraction":5.707846926195661e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":128,"both_estimated_voiced_frames":117,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9069767441860465,"absolute_difference_cents_median":3.058381821348301,"absolute_difference_cents_p95":9.893269231843952,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019142337741905198,"log_rms_mae_db":1.3498769664419425,"rms_envelope_nrmse":0.09127953181786576,"envelope_correlation":0.9927012532732138,"rms_gain_error_db":-0.2566469046586609,"reference_mean_power":0.0025533227171819564,"output_mean_power":0.0024068056301545008},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1748402838347296e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007147512141697323,"output_power":0.0007317369954894585,"reference_active":true,"delta_db":0.1020011106504792},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.175142428480134e-05,"output_power":7.975559334836522e-05,"reference_active":true,"delta_db":-0.10734177385953932},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4312039590506752e-05,"output_power":1.5210779916330972e-05,"reference_active":true,"delta_db":0.2644995368210206}],"active_band_count":3,"active_band_mae_db":0.15794747377701304}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016640454830055276,"log_rms_mae_db":1.1158489388655912,"rms_envelope_nrmse":0.0803268225044087,"envelope_correlation":0.9941201218403986,"rms_gain_error_db":-0.25755424551520795,"reference_mean_power":0.002564981061322552,"output_mean_power":0.0024172899044807013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0626539543337418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006440432662381922,"output_power":0.0006630929695531056,"reference_active":true,"delta_db":0.12659379480414543},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1296244809426825e-05,"output_power":3.9120019997721264e-05,"reference_active":true,"delta_db":-0.23511493638567654},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.585289215752229e-07,"output_power":1.0276816762611097e-06,"reference_active":true,"delta_db":0.7810368307323361}],"active_band_count":3,"active_band_mae_db":0.380915187307386}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016776043588712122,"log_rms_mae_db":1.5647063238036307,"rms_envelope_nrmse":0.10030099548092113,"envelope_correlation":0.9934281626864582,"rms_gain_error_db":-0.3351810759237492,"reference_mean_power":0.0015150798695545982,"output_mean_power":0.001402546989372114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.503421266374217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005710315291680691,"output_power":0.000559086039986671,"reference_active":true,"delta_db":-0.09181439876329174},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462070521392158e-05,"output_power":6.089016537165984e-05,"reference_active":true,"delta_db":-0.25824539661112467},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.823790922205639e-06,"output_power":1.0736968586906588e-05,"reference_active":true,"delta_db":0.3860257160630593}],"active_band_count":3,"active_band_mae_db":0.24536183714582524}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015135634256581334,"log_rms_mae_db":1.340478844805102,"rms_envelope_nrmse":0.08763867281801746,"envelope_correlation":0.9949725777226295,"rms_gain_error_db":-0.33560341484225054,"reference_mean_power":0.0015225180811750088,"output_mean_power":0.0014092956696163236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.758019890177841e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005253930575206644,"output_power":0.000514713203214026,"reference_active":true,"delta_db":-0.0891902103531409},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0513708673138778e-05,"output_power":2.892591943709343e-05,"reference_active":true,"delta_db":-0.23207823233346966},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.334657094067361e-07,"output_power":9.541916029140446e-07,"reference_active":true,"delta_db":0.5874785364594484}],"active_band_count":3,"active_band_mae_db":0.302915659715353}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":1.4829526329432847,"rms_envelope_nrmse":0.10377411371948973,"envelope_correlation":0.992367409084992,"rms_gain_error_db":0.08647337177520167,"reference_mean_power":0.00023682053730462446,"output_mean_power":0.00024158318313544498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.163166626171004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014015200472863973,"output_power":0.00014684596476521717,"reference_active":true,"delta_db":0.20262702611946595},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.313645741402795e-05,"output_power":1.493965023653313e-05,"reference_active":true,"delta_db":0.5586216774351898},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2309159151114036e-06,"output_power":3.5688129167387867e-06,"reference_active":true,"delta_db":0.431981264845105}],"active_band_count":3,"active_band_mae_db":0.39774332279992025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":1.1298871823084273,"rms_envelope_nrmse":0.0772176369942411,"envelope_correlation":0.9955148070071129,"rms_gain_error_db":0.08646228594802437,"reference_mean_power":0.0002379882918616052,"output_mean_power":0.00024277380244127044},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.020973483325393e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013180546055965978,"output_power":0.00013892317650778264,"reference_active":true,"delta_db":0.22841301999430366},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.132933188997126e-06,"output_power":6.9881298551522555e-06,"reference_active":true,"delta_db":0.5669273296091595},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2319303983633038e-07,"output_power":2.5683926167768304e-07,"reference_active":true,"delta_db":0.6098076558293095}],"active_band_count":3,"active_band_mae_db":0.46838266847759086}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":1.935711036071042,"rms_envelope_nrmse":0.24018470590790095,"envelope_correlation":0.9596775566151458,"rms_gain_error_db":-0.5816205363179509,"reference_mean_power":3.051701314781676e-05,"output_mean_power":2.6691929626889186e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.8263265386055575e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1311662288856238e-05,"output_power":1.1020178697099306e-05,"reference_active":true,"delta_db":-0.113377938772626},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.547549874304093e-06,"output_power":2.2171898768014328e-06,"reference_active":true,"delta_db":-0.6031980806923041},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.732573657220465e-07,"output_power":7.707712742984108e-07,"reference_active":true,"delta_db":1.2857587702069626}],"active_band_count":3,"active_band_mae_db":0.6674449298906309}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":1.5266189433499484,"rms_envelope_nrmse":0.19877565562810667,"envelope_correlation":0.9711649651273353,"rms_gain_error_db":-0.5817803080574161,"reference_mean_power":3.0668913834731314e-05,"output_mean_power":2.6823803844283997e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6363060360497885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1094596811060145e-05,"output_power":1.0803894473597666e-05,"reference_active":true,"delta_db":-0.11531190560934752},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.190115655321781e-06,"output_power":1.06371917246513e-06,"reference_active":true,"delta_db":-0.48762181106480956},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.417829557129175e-08,"output_power":5.3723874147794224e-08,"reference_active":true,"delta_db":0.8495836658289713}],"active_band_count":3,"active_band_mae_db":0.4841724608343761}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.7053587436676025,"predicted_mos":3.7053587436676025,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"0d19db53cf07b06a4a198566af3e09c8498a15e7e6e10da841f82fe260c438d9","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2865621500004636},"started_at":"2026-10-13T04:57:25.592938+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006024889000400435,"native_input_samples":264600,"native_input_sha256":"9169f01d3c7c3383c6e44714b4e4c55199f2240113804c10228d9de7bd6275ff","output_sha256":"11b792de55e36c07939d42e4b51ac9013b3c30226e538ea9f59d8aa59c95c6a7","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13339803899998515,"synthesis_s":3.9891060290001406,"total_s":4.122504068000126,"rtf":0.6870840113333543,"output_peak":0.4341559410095215,"output_rms":0.046889588624071894,"finished_at":"2026-10-13T04:57:29.732009+00:00","evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"11b792de55e36c07939d42e4b51ac9013b3c30226e538ea9f59d8aa59c95c6a7","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.572533794338653,"output_peak":0.4321118640162255},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.934608481857097,"spectral_convergence":0.43505181719332464,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.40335653073764,"spectral_convergence":0.34153603767105845,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.144596746900804,"spectral_convergence":0.41136743326386727,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.0003619677577404013,"delta_db":-0.030112633705146094,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0007025908644734584,"delta_db":-0.8298266465951597,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.0009006773723161818,"delta_db":-0.6797949003701189,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0001841793320684026,"delta_db":-0.7682673233317932,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.536222163179396e-05,"delta_db":-0.3944213023760216,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":1.7696521331896104e-05,"delta_db":-1.3984055144517562,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6834713868049992,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6384438604732201,"pitch":null,"output_rms_dbfs":-26.570891959748796,"output_peak":0.4321118640162255,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.49418725316518},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7515116553387734,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050510026345542525,"rms_dbfs":-25.932448099275575,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.160600805855687,"mean":-9.429613749186198e-05,"zero_energy":false},"output":{"rms":0.046930523857742455,"rms_dbfs":-26.570891959748796,"peak":0.4321118640162255,"peak_dbfs":-7.288076191662629,"crest_factor_db":19.282815768086166,"mean":-0.00026057340562639396,"zero_energy":false},"rms_difference_db":-0.6384438604732203},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":8.363070222780605e-07,"output_rms_dbfs":-60.776342565288445,"output_zero_energy":false,"output_peak":0.0051033898186097,"output_energy_fraction":5.6956880934652064e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":115,"both_estimated_voiced_frames":84,"reference_voiced_output_unvoiced_frames":45,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.6511627906976745,"absolute_difference_cents_median":12.905069584593033,"absolute_difference_cents_p95":50.10895952744036,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019142337741905198,"log_rms_mae_db":1.8161699127042545,"rms_envelope_nrmse":0.12221986134709985,"envelope_correlation":0.9905530693019966,"rms_gain_error_db":-0.6385577676426293,"reference_mean_power":0.0025533227171819564,"output_mean_power":0.002204194591693036},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1748402838347296e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007147512141697323,"output_power":0.0006727780408699761,"reference_active":true,"delta_db":-0.262830939078122},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.175142428480134e-05,"output_power":7.353753860569678e-05,"reference_active":true,"delta_db":-0.4598623834648255},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4312039590506752e-05,"output_power":1.018778025744404e-05,"reference_active":true,"delta_db":-1.4762195992697882}],"active_band_count":3,"active_band_mae_db":0.7329709739375785}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016640454830055276,"log_rms_mae_db":1.412373064574507,"rms_envelope_nrmse":0.0992495745858092,"envelope_correlation":0.9952661080242827,"rms_gain_error_db":-0.6379153334629482,"reference_mean_power":0.002564981061322552,"output_mean_power":0.002214586404770039},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0626539543337418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006440432662381922,"output_power":0.0006096852708089996,"reference_active":true,"delta_db":-0.23809340678851648},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1296244809426825e-05,"output_power":3.6420598278011896e-05,"reference_active":true,"delta_db":-0.5456348612540269},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.585289215752229e-07,"output_power":7.2717921304577e-07,"reference_active":true,"delta_db":-0.721134742159123}],"active_band_count":3,"active_band_mae_db":0.5016210034005555}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016776043588712122,"log_rms_mae_db":2.4307739853390773,"rms_envelope_nrmse":0.1428007173167824,"envelope_correlation":0.9866407709846298,"rms_gain_error_db":-0.5461526073000423,"reference_mean_power":0.0015150798695545982,"output_mean_power":0.0013360424807036168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.503421266374217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005710315291680691,"output_power":0.0004801167692829771,"reference_active":true,"delta_db":-0.7531321316209022},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462070521392158e-05,"output_power":5.368489128713232e-05,"reference_active":true,"delta_db":-0.8051961522558483},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.823790922205639e-06,"output_power":8.039656987052518e-06,"reference_active":true,"delta_db":-0.8704159097551023}],"active_band_count":3,"active_band_mae_db":0.8095813978772842}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015135634256581334,"log_rms_mae_db":1.7245845328125269,"rms_envelope_nrmse":0.11926028656302348,"envelope_correlation":0.9909357960207134,"rms_gain_error_db":-0.5463682016762869,"reference_mean_power":0.0015225180811750088,"output_mean_power":0.0013425350685193473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.758019890177841e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005253930575206644,"output_power":0.0004449414595783699,"reference_active":true,"delta_db":-0.7218145464589226},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0513708673138778e-05,"output_power":2.5717759582372754e-05,"reference_active":true,"delta_db":-0.7426186344524632},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.334657094067361e-07,"output_power":6.852174400549861e-07,"reference_active":true,"delta_db":-0.8505932867701772}],"active_band_count":3,"active_band_mae_db":0.7716754892271878}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":2.8711464168614653,"rms_envelope_nrmse":0.23123481509774282,"envelope_correlation":0.9639566020523435,"rms_gain_error_db":-0.7251625717441268,"reference_mean_power":0.00023682053730462446,"output_mean_power":0.00020040248612797866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.163166626171004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014015200472863973,"output_power":0.0001263349439483905,"reference_active":true,"delta_db":-0.4507582226767013},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.313645741402795e-05,"output_power":1.2827049414959372e-05,"reference_active":true,"delta_db":-0.10351494438790199},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2309159151114036e-06,"output_power":2.6721702272121094e-06,"reference_active":true,"delta_db":-0.8246153474682855}],"active_band_count":3,"active_band_mae_db":0.4596295048442962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":2.445542198497467,"rms_envelope_nrmse":0.20147867682181167,"envelope_correlation":0.9723898458330501,"rms_gain_error_db":-0.7249567288850804,"reference_mean_power":0.0002379882918616052,"output_mean_power":0.00020140020998347585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.020973483325393e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013180546055965978,"output_power":0.00011831505756541135,"reference_active":true,"delta_db":-0.4689338362264891},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.132933188997126e-06,"output_power":6.624468239258535e-06,"reference_active":true,"delta_db":0.3348278863566505},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2319303983633038e-07,"output_power":1.6328558331543128e-07,"reference_active":true,"delta_db":-1.3573280518337316}],"active_band_count":3,"active_band_mae_db":0.7203632581389571}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":3.7248084746830195,"rms_envelope_nrmse":0.27634010304733725,"envelope_correlation":0.9523547992277169,"rms_gain_error_db":-1.2198078057819917,"reference_mean_power":3.051701314781676e-05,"output_mean_power":2.3044179222183255e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.8263265386055575e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1311662288856238e-05,"output_power":7.3446555715312185e-06,"reference_active":true,"delta_db":-1.8755499641426163},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.547549874304093e-06,"output_power":1.921125233933572e-06,"reference_active":true,"delta_db":-1.2256701861395154},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.732573657220465e-07,"output_power":2.781282327343482e-07,"reference_active":true,"delta_db":-3.141045674037473}],"active_band_count":3,"active_band_mae_db":2.0807552747732014}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":3.0469635175275385,"rms_envelope_nrmse":0.22996949111821777,"envelope_correlation":0.968950954764823,"rms_gain_error_db":-1.219663423275593,"reference_mean_power":3.0668913834731314e-05,"output_mean_power":2.3159653263131398e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6363060360497885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1094596811060145e-05,"output_power":7.117026273157613e-06,"reference_active":true,"delta_db":-1.9281295508337435},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.190115655321781e-06,"output_power":8.897605736059957e-07,"reference_active":true,"delta_db":-1.2631601042670928},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.417829557129175e-08,"output_power":2.4974006790337235e-08,"reference_active":true,"delta_db":-2.4772073101749346}],"active_band_count":3,"active_band_mae_db":1.8894989884252569}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"golf","status":"measured","mos":{"status":"measured","value":2.494191884994507,"predicted_mos":2.494191884994507,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"6061f86e003e4d8f2d1a4d866c83f351e4f2367090555677c521fa54574a62f2","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.124975396998707},"started_at":"2026-10-13T05:01:48.912567+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002614021000226785,"native_input_samples":144000,"native_input_sha256":"3dfc1833d7d9b747b3655cd21f3408448686978fdc631dd1e90e0337b558fd88","output_sha256":"f28130f40ec954f4f5886b92facba82f73b8cb034b691dd52abaf0993cc13dc9","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.79786952700033,"synthesis_s":0.1127914819999205,"total_s":0.9106610090002505,"rtf":0.1517768348333751,"output_peak":0.3999296724796295,"output_rms":0.04353257671850164,"finished_at":"2026-10-13T05:01:49.833350+00:00","analysis_stages":{"external_f0_s":0.08684278200053086},"evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"f28130f40ec954f4f5886b92facba82f73b8cb034b691dd52abaf0993cc13dc9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.218911220425518,"output_peak":0.40028096061389074},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.736141521041546,"spectral_convergence":0.708197376931781,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.392384638754208,"spectral_convergence":0.32804217790021345,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.630325077265818,"spectral_convergence":0.29645359978460295,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036449544146104193,"output_power":0.0003100867252293328,"delta_db":-0.7020892666437166,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008518912745091038,"output_power":0.0006574286151017907,"delta_db":-1.1253556683995933,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010550481268421655,"output_power":0.0007322759934178895,"delta_db":-1.5859747409674814,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022018989989863153,"output_power":0.00015068505411192415,"delta_db":-1.647272156210065,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.877691756336759e-05,"output_power":2.6869066168219812e-05,"delta_db":-1.5932071054749568,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.446006010926411e-05,"output_power":1.9836032411093213e-05,"delta_db":-0.9100271072129161,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2606543408181214,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.2925852134712656,"pitch":null,"output_rms_dbfs":-27.218911220425518,"output_peak":0.40028096061389074,"reference_rms_dbfs":-25.926326006954255,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.252950412353857},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8781261315315703,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050545639974023227,"rms_dbfs":-25.926326006954255,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.154478713534367,"mean":-9.380620787657163e-05,"zero_energy":false},"output":{"rms":0.043556646868649684,"rms_dbfs":-27.218911220425518,"peak":0.40028096061389074,"peak_dbfs":-7.9527013328953275,"crest_factor_db":19.26620988753019,"mean":0.0006086706330146601,"zero_energy":false},"rms_difference_db":-1.2925852134712628},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":1.3629122451996784e-06,"output_rms_dbfs":-58.65532106492889,"output_zero_energy":false,"output_peak":0.006714267722002361,"output_energy_fraction":0.00010793808348490335,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":151,"both_estimated_voiced_frames":119,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9224806201550387,"absolute_difference_cents_median":5.216065545448607,"absolute_difference_cents_p95":17.14356943146254,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019142337741905198,"log_rms_mae_db":2.017282273706641,"rms_envelope_nrmse":0.17932137573421825,"envelope_correlation":0.9855346521837143,"rms_gain_error_db":-1.2926323304152174,"reference_mean_power":0.002556883262135026,"output_mean_power":0.0018986620379548398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.175498553700518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007141851126548946,"output_power":0.0005410825031281271,"reference_active":true,"delta_db":-1.2054730251781356},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160309002864643e-05,"output_power":7.243022111054179e-05,"reference_active":true,"delta_db":-0.5178679331477837},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4221235162217439e-05,"output_power":1.1904934054466736e-05,"reference_active":true,"delta_db":-0.7721032393104237}],"active_band_count":3,"active_band_mae_db":0.8318147325454476}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016640454830055276,"log_rms_mae_db":1.4987798043979088,"rms_envelope_nrmse":0.15620152709119942,"envelope_correlation":0.9936542216817108,"rms_gain_error_db":-1.293248417003706,"reference_mean_power":0.0025683591585647586,"output_mean_power":0.001906913149629155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0633150758463186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006432386843032706,"output_power":0.000488482746652516,"reference_active":true,"delta_db":-1.195229260249813},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1128873185169776e-05,"output_power":3.5901401034756734e-05,"reference_active":true,"delta_db":-0.5903541419089153},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.542299519192527e-07,"output_power":8.0149774079662e-07,"reference_active":true,"delta_db":-0.27672492518601577}],"active_band_count":3,"active_band_mae_db":0.6874361091149147}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00016776043588712122,"log_rms_mae_db":1.9715898932904965,"rms_envelope_nrmse":0.19432397534109652,"envelope_correlation":0.9856153090391543,"rms_gain_error_db":-1.3729885184921742,"reference_mean_power":0.0015175373878160273,"output_mean_power":0.0011062175591687754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.51174535124836e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005706349575679751,"output_power":0.0004301198718101197,"reference_active":true,"delta_db":-1.2276886567669378},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.445821954299549e-05,"output_power":5.370453372641424e-05,"reference_active":true,"delta_db":-0.792673550094075},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.768779821788585e-06,"output_power":7.518686059191328e-06,"reference_active":true,"delta_db":-1.1369836981479302}],"active_band_count":3,"active_band_mae_db":1.052448635002981}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015135634256581334,"log_rms_mae_db":1.539356201435109,"rms_envelope_nrmse":0.1701206084640788,"envelope_correlation":0.9928397827425433,"rms_gain_error_db":-1.3726807396498648,"reference_mean_power":0.0015248347919477414,"output_mean_power":0.0011116158196255355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.766486755659543e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005247306554695345,"output_power":0.00039753359906124114,"reference_active":true,"delta_db":-1.205625964157761},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0360220613177104e-05,"output_power":2.573620999879913e-05,"reference_active":true,"delta_db":-0.7176033143986753},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294411731558717e-07,"output_power":5.86320024432478e-07,"reference_active":true,"delta_db":-1.5065086299953316}],"active_band_count":3,"active_band_mae_db":1.1432459695172559}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":2.210849288850177,"rms_envelope_nrmse":0.23089013225555863,"envelope_correlation":0.9812142028171895,"rms_gain_error_db":-1.7130141633309603,"reference_mean_power":0.00023720444556869473,"output_mean_power":0.00015989003874953246},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.1636103690177146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014002111865811292,"output_power":9.348392708060057e-05,"reference_active":true,"delta_db":-1.7545659501738962},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3103971123855928e-05,"output_power":8.631969632267879e-06,"reference_active":true,"delta_db":-1.8129302359208306},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2255937358391144e-06,"output_power":1.8768857106375086e-06,"reference_active":true,"delta_db":-2.351718391271904}],"active_band_count":3,"active_band_mae_db":1.9730715257888771}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":1.7754297895135303,"rms_envelope_nrmse":0.19940908007274874,"envelope_correlation":0.9927091543342916,"rms_gain_error_db":-1.71287383962349,"reference_mean_power":0.0002383594926577856,"output_mean_power":0.00016067380124349585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.0213541418927988e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013163447495927977,"output_power":8.777749738439509e-05,"reference_active":true,"delta_db":-1.7598645097961099},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.095298755536558e-06,"output_power":3.962781007907303e-06,"reference_active":true,"delta_db":-1.8699492410769727},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2198941445803914e-07,"output_power":1.3944374417908345e-07,"reference_active":true,"delta_db":-2.019332302354285}],"active_band_count":3,"active_band_mae_db":1.8830486844091225}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":2.778506611570752,"rms_envelope_nrmse":0.23080327733800157,"envelope_correlation":0.9677660761717968,"rms_gain_error_db":-0.90191612087551,"reference_mean_power":3.056731194925443e-05,"output_mean_power":2.4835084214855506e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.8263386858754374e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1323852605483402e-05,"output_power":8.729015678090728e-06,"reference_active":true,"delta_db":-1.1302893430306666},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.543766818379685e-06,"output_power":2.33970112505602e-06,"reference_active":true,"delta_db":-0.3631691406856177},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.721076988550209e-07,"output_power":5.965112812334465e-07,"reference_active":true,"delta_db":0.18140869362937267}],"active_band_count":3,"active_band_mae_db":0.558289059115219}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":2.2791970961210763,"rms_envelope_nrmse":0.1627836257019782,"envelope_correlation":0.9872529216766963,"rms_gain_error_db":-0.9018598591404925,"reference_mean_power":3.071919647262595e-05,"output_mean_power":2.4958809465932824e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.636369058484133e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1104324576615646e-05,"output_power":8.73554904218626e-06,"reference_active":true,"delta_db":-1.042019413870847},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1874425468708628e-06,"output_power":1.1692548068806842e-06,"reference_active":true,"delta_db":-0.06703441937941913},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.380746444577275e-08,"output_power":4.2545853535779645e-08,"reference_active":true,"delta_db":-0.12690876456472455}],"active_band_count":3,"active_band_mae_db":0.4119875326049969}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.094534397125244,"predicted_mos":3.094534397125244,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"ba254e11574f748f5694c21cb40fb678060ce956a3b4d3cf259b1a7c0c7da5d2","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9818087940002442},"started_at":"2026-10-13T05:03:41.378737+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032323400000677793,"native_input_samples":144000,"native_input_sha256":"3dfc1833d7d9b747b3655cd21f3408448686978fdc631dd1e90e0337b558fd88","output_sha256":"0bba1f6868ce56081a167e88097e9d1902c9f4c1e1efbf40f16b8f6a3c064173","output_samples":144120,"length_delta_samples":120,"analysis_s":2.67406774400024,"synthesis_s":4.17402206099996,"total_s":6.8480898050002,"rtf":1.1413483008333667,"output_peak":0.5074725151062012,"output_rms":0.050775436001642894,"finished_at":"2026-10-13T05:03:48.241381+00:00","analysis_stages":{"external_f0_s":1.737016107999807},"evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"0bba1f6868ce56081a167e88097e9d1902c9f4c1e1efbf40f16b8f6a3c064173","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.882403404726343,"output_peak":0.5068729787019272},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.86404436795075,"spectral_convergence":0.7987384350030682,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.390104931006685,"spectral_convergence":0.36818305608692536,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.391003291731289,"spectral_convergence":0.3017595632086923,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.0005546034480228509,"delta_db":1.8230141443307402,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0008134461876897388,"delta_db":-0.1935633505907266,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.0009487586781877593,"delta_db":-0.45392980789901444,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.00020641095417015575,"delta_db":-0.27334884538116677,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.840179547753232e-05,"delta_db":-0.036301416906157534,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.1235752998082753e-05,"delta_db":-0.606607901487214,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.56446091109917,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.05345141185078881,"pitch":null,"output_rms_dbfs":-25.878996687424785,"output_peak":0.5068729787019272,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.215050863562908},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8553715802844556,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050510026345542525,"rms_dbfs":-25.932448099275575,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.160600805855687,"mean":-9.429613749186198e-05,"zero_energy":false},"output":{"rms":0.05082181437676266,"rms_dbfs":-25.878996687424785,"peak":0.5068729787019272,"peak_dbfs":-5.902017206245792,"crest_factor_db":19.976979481178994,"mean":-0.0001680533115239423,"zero_energy":false},"rms_difference_db":0.05345141185079072},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":2.1543448510483703e-06,"output_rms_dbfs":-56.666847769340556,"output_zero_energy":false,"output_peak":0.029303993318464304,"output_energy_fraction":0.00012511406965616642,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":154,"both_estimated_voiced_frames":118,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.9147286821705426,"absolute_difference_cents_median":6.216470757328695,"absolute_difference_cents_p95":18.246411934472818,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019142337741905198,"log_rms_mae_db":2.811216109430247,"rms_envelope_nrmse":0.1604395854182222,"envelope_correlation":0.9754466523065671,"rms_gain_error_db":0.05331524419931179,"reference_mean_power":0.0025533227171819564,"output_mean_power":0.0025848612357700766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1748402838347296e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007147512141697323,"output_power":0.0006996431116670628,"reference_active":true,"delta_db":-0.0927833913478294},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.175142428480134e-05,"output_power":8.0171353956983e-05,"reference_active":true,"delta_db":-0.08476109587746652},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4312039590506752e-05,"output_power":1.2505925466995957e-05,"reference_active":true,"delta_db":-0.5858569288531572}],"active_band_count":3,"active_band_mae_db":0.25446713869281773}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016640454830055276,"log_rms_mae_db":2.2036044681847433,"rms_envelope_nrmse":0.1060716450256626,"envelope_correlation":0.988425534655294,"rms_gain_error_db":0.05297068420125555,"reference_mean_power":0.002564981061322552,"output_mean_power":0.002596457577623473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0626539543337418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006440432662381922,"output_power":0.0006265035963618333,"reference_active":true,"delta_db":-0.11991475498434385},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1296244809426825e-05,"output_power":4.052596468387071e-05,"reference_active":true,"delta_db":-0.08177200130474355},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.585289215752229e-07,"output_power":8.182703827801168e-07,"reference_active":true,"delta_db":-0.20858097766662778}],"active_band_count":3,"active_band_mae_db":0.1367559113185717}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016776043588712122,"log_rms_mae_db":2.3553462553084374,"rms_envelope_nrmse":0.18690909609833647,"envelope_correlation":0.974304314828803,"rms_gain_error_db":-0.44097535819314837,"reference_mean_power":0.0015150798695545982,"output_mean_power":0.0013687936823708865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.503421266374217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005710315291680691,"output_power":0.0005219734147873979,"reference_active":true,"delta_db":-0.390117041939001},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462070521392158e-05,"output_power":5.5655070903406676e-05,"reference_active":true,"delta_db":-0.6486695284853763},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.823790922205639e-06,"output_power":8.630073262431133e-06,"reference_active":true,"delta_db":-0.5626462834786359}],"active_band_count":3,"active_band_mae_db":0.5338109513010044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015135634256581334,"log_rms_mae_db":1.737747299132764,"rms_envelope_nrmse":0.12895527958456385,"envelope_correlation":0.9879166623493408,"rms_gain_error_db":-0.4408035139646075,"reference_mean_power":0.0015225180811750088,"output_mean_power":0.0013755681372265576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.758019890177841e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005253930575206644,"output_power":0.0004810137718840237,"reference_active":true,"delta_db":-0.38326818956084135},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0513708673138778e-05,"output_power":2.615045687266604e-05,"reference_active":true,"delta_db":-0.6701571473235903},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.334657094067361e-07,"output_power":7.657494522597246e-07,"reference_active":true,"delta_db":-0.3680104188332487}],"active_band_count":3,"active_band_mae_db":0.47381191857256016}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":2.3594103608366286,"rms_envelope_nrmse":0.21038004019768158,"envelope_correlation":0.9674582157360778,"rms_gain_error_db":-0.2167288104758075,"reference_mean_power":0.00023682053730462446,"output_mean_power":0.00022529236864844755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.163166626171004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014015200472863973,"output_power":0.00013143182313760327,"reference_active":true,"delta_db":-0.2789878207856871},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.313645741402795e-05,"output_power":1.226438843576736e-05,"reference_active":true,"delta_db":-0.29832365287666673},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2309159151114036e-06,"output_power":2.6435253536782686e-06,"reference_active":true,"delta_db":-0.8714217559210702}],"active_band_count":3,"active_band_mae_db":0.482911076527808}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":1.4661263364748396,"rms_envelope_nrmse":0.1320740026025727,"envelope_correlation":0.9866065357326567,"rms_gain_error_db":-0.2165793039226799,"reference_mean_power":0.0002379882918616052,"output_mean_power":0.0002264110722738215},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.020973483325393e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013180546055965978,"output_power":0.00012348813569156028,"reference_active":true,"delta_db":-0.2830816888373388},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.132933188997126e-06,"output_power":5.829997480193178e-06,"reference_active":true,"delta_db":-0.21999866541242075},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2319303983633038e-07,"output_power":1.8396505931656744e-07,"reference_active":true,"delta_db":-0.8394530240036829}],"active_band_count":3,"active_band_mae_db":0.4475111260844808}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":2.5275204449029443,"rms_envelope_nrmse":0.23855650704850165,"envelope_correlation":0.9592238464186307,"rms_gain_error_db":-0.3610458795726166,"reference_mean_power":3.051701314781676e-05,"output_mean_power":2.8082607940034072e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.8263265386055575e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1311662288856238e-05,"output_power":1.01609756019207e-05,"reference_active":true,"delta_db":-0.4659102217122723},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.547549874304093e-06,"output_power":2.264565628307033e-06,"reference_active":true,"delta_db":-0.5113778381699632},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.732573657220465e-07,"output_power":4.773843513586944e-07,"reference_active":true,"delta_db":-0.7948146489369512}],"active_band_count":3,"active_band_mae_db":0.5907009029397289}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":1.6003613067142664,"rms_envelope_nrmse":0.155515210091147,"envelope_correlation":0.9820647875446186,"rms_gain_error_db":-0.36096008178166633,"reference_mean_power":3.0668913834731314e-05,"output_mean_power":2.822294875301824e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6363060360497885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1094596811060145e-05,"output_power":9.903102225260364e-06,"reference_active":true,"delta_db":-0.493402620520289},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.190115655321781e-06,"output_power":1.0826787317446636e-06,"reference_active":true,"delta_db":-0.4108956260192351},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.417829557129175e-08,"output_power":3.6682947919235926e-08,"reference_active":true,"delta_db":-0.8074472725896152}],"active_band_count":3,"active_band_mae_db":0.5705818397097131}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.9345715045928955,"predicted_mos":2.9345715045928955,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"e8f49f1707d7ea3baaa2ace6fbbddd312fb15a5cc9376bd94d0972ac1c14180b","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0098258320012974},"started_at":"2026-10-13T05:09:53.702248+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029031949998170603,"native_input_samples":144000,"native_input_sha256":"3dfc1833d7d9b747b3655cd21f3408448686978fdc631dd1e90e0337b558fd88","output_sha256":"e544cfb025a56035af7266a90f4a46cc021e038b010d912134798ef632226895","output_samples":144240,"length_delta_samples":240,"analysis_s":1.2529164360003051,"synthesis_s":0.14709949800089817,"total_s":1.4000159340012033,"rtf":0.23333598900020056,"output_peak":0.5094049572944641,"output_rms":0.049156976309799046,"finished_at":"2026-10-13T05:09:55.114938+00:00","analysis_stages":{"external_f0_s":0.170736845000647},"evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"e544cfb025a56035af7266a90f4a46cc021e038b010d912134798ef632226895","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.16320266473318,"output_peak":0.5051777562913853},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.989982249294416,"spectral_convergence":0.1889383428971063,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.327727623603536,"spectral_convergence":0.1774405930109197,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.700579304825704,"spectral_convergence":0.19184452928451987,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.0003480776713407403,"delta_db":-0.20004986027834298,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0008090430387815229,"delta_db":-0.21713537729093776,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.0010148300721462534,"delta_db":-0.16155414099164372,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.0001965029444056377,"delta_db":-0.4869856374760426,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.320545958025462e-05,"delta_db":-0.6677217651884566,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.1614150907185494e-05,"delta_db":-0.5299027605670732,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3772249236320828,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2235278620082642,"pitch":null,"output_rms_dbfs":-26.155975961283843,"output_peak":0.5051777562913853,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.672763059241219},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9317272457761304,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050510026345542525,"rms_dbfs":-25.932448099275575,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.160600805855687,"mean":-9.429613749186198e-05,"zero_energy":false},"output":{"rms":0.04922675428221522,"rms_dbfs":-26.155975961283843,"peak":0.5051777562913853,"peak_dbfs":-5.931115606221083,"crest_factor_db":20.22486035506276,"mean":1.6418373599727758e-05,"zero_energy":false},"rms_difference_db":-0.22352786200826813},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":1.257027976576889e-06,"output_rms_dbfs":-59.00655056492437,"output_zero_energy":false,"output_peak":0.010010643118074999,"output_energy_fraction":7.780971035953741e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":138,"both_estimated_voiced_frames":119,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9224806201550387,"absolute_difference_cents_median":1.99957135146643,"absolute_difference_cents_p95":6.247446416518449,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019142337741905198,"log_rms_mae_db":0.9410487334819715,"rms_envelope_nrmse":0.07545163327359478,"envelope_correlation":0.9950460261777514,"rms_gain_error_db":-0.22355148267403385,"reference_mean_power":0.0025533227171819564,"output_mean_power":0.0024252167604853127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1748402838347296e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007147512141697323,"output_power":0.00068546653238066,"reference_active":true,"delta_db":-0.18168646384881856},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.175142428480134e-05,"output_power":6.870021630452762e-05,"reference_active":true,"delta_db":-0.7553722328785597},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4312039590506752e-05,"output_power":9.737396053054368e-06,"reference_active":true,"delta_db":-1.6725869432190212}],"active_band_count":3,"active_band_mae_db":0.8698818799821332}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016640454830055276,"log_rms_mae_db":0.5274061075156733,"rms_envelope_nrmse":0.04643418123051671,"envelope_correlation":0.9983943573229421,"rms_gain_error_db":-0.22313112236496085,"reference_mean_power":0.002564981061322552,"output_mean_power":0.0024365260029053535},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0626539543337418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006440432662381922,"output_power":0.000618888150332779,"reference_active":true,"delta_db":-0.17302876381118548},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1296244809426825e-05,"output_power":3.508225450730959e-05,"reference_active":true,"delta_db":-0.7082306690074115},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.585289215752229e-07,"output_power":5.529614843952484e-07,"reference_active":true,"delta_db":-1.9106004767469256}],"active_band_count":3,"active_band_mae_db":0.9306199698551741}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016776043588712122,"log_rms_mae_db":1.2953057683708127,"rms_envelope_nrmse":0.09435741527818294,"envelope_correlation":0.9935961462494769,"rms_gain_error_db":-0.21092403602544807,"reference_mean_power":0.0015150798695545982,"output_mean_power":0.0014432551779286482},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.503421266374217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005710315291680691,"output_power":0.0005486592204937736,"reference_active":true,"delta_db":-0.1735740608679909},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462070521392158e-05,"output_power":5.269342324058661e-05,"reference_active":true,"delta_db":-0.8861527968598238},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.823790922205639e-06,"output_power":8.151628475135312e-06,"reference_active":true,"delta_db":-0.8103473318071696}],"active_band_count":3,"active_band_mae_db":0.6233580631783281}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015135634256581334,"log_rms_mae_db":0.7227004489249794,"rms_envelope_nrmse":0.05916042043453854,"envelope_correlation":0.9975757177839127,"rms_gain_error_db":-0.21094579165083313,"reference_mean_power":0.0015225180811750088,"output_mean_power":0.0014503335043413875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.758019890177841e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005253930575206644,"output_power":0.0005044104126159739,"reference_active":true,"delta_db":-0.17700286577608582},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0513708673138778e-05,"output_power":2.5092016816184214e-05,"reference_active":true,"delta_db":-0.8495942560702292},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.334657094067361e-07,"output_power":5.436569535319092e-07,"reference_active":true,"delta_db":-1.8556278955418004}],"active_band_count":3,"active_band_mae_db":0.9607416724627051}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":1.335503778725799,"rms_envelope_nrmse":0.11733436837136854,"envelope_correlation":0.9918677568940217,"rms_gain_error_db":-0.5068841815932217,"reference_mean_power":0.00023682053730462446,"output_mean_power":0.00021073222078586506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.163166626171004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014015200472863973,"output_power":0.00012311214877560738,"reference_active":true,"delta_db":-0.5629840280347993},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.313645741402795e-05,"output_power":1.0568536255016281e-05,"reference_active":true,"delta_db":-0.9446342068260737},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2309159151114036e-06,"output_power":2.1539411708818446e-06,"reference_active":true,"delta_db":-1.760918180736706}],"active_band_count":3,"active_band_mae_db":1.0895121385325262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":0.8220718848271943,"rms_envelope_nrmse":0.08081528569808206,"envelope_correlation":0.9972779522699212,"rms_gain_error_db":-0.5068309920103895,"reference_mean_power":0.0002379882918616052,"output_mean_power":0.00021177392833606742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.020973483325393e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013180546055965978,"output_power":0.00011605962051612548,"reference_active":true,"delta_db":-0.5525225678371577},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.132933188997126e-06,"output_power":5.209479279835367e-06,"reference_active":true,"delta_db":-0.7087391858359156},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2319303983633038e-07,"output_power":1.350495448829079e-07,"reference_active":true,"delta_db":-2.1818752226668816}],"active_band_count":3,"active_band_mae_db":1.1477123254466515}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":1.5970151200280098,"rms_envelope_nrmse":0.12207837416617158,"envelope_correlation":0.9902965278465646,"rms_gain_error_db":-0.367528258087519,"reference_mean_power":3.051701314781676e-05,"output_mean_power":2.8040722466142042e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.8263265386055575e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1311662288856238e-05,"output_power":1.0653840708432779e-05,"reference_active":true,"delta_db":-0.2602023161214069},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.547549874304093e-06,"output_power":2.190778382950643e-06,"reference_active":true,"delta_db":-0.6552424813408354},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.732573657220465e-07,"output_power":5.023890917297874e-07,"reference_active":true,"delta_db":-0.5730944255532369}],"active_band_count":3,"active_band_mae_db":0.49617974100515977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":1.01242082217291,"rms_envelope_nrmse":0.0721082862447916,"envelope_correlation":0.9972505817188292,"rms_gain_error_db":-0.3675167440123944,"reference_mean_power":3.0668913834731314e-05,"output_mean_power":2.8180371945382193e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6363060360497885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1094596811060145e-05,"output_power":1.0454664257651937e-05,"reference_active":true,"delta_db":-0.2580143375901954},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.190115655321781e-06,"output_power":1.034761482242281e-06,"reference_active":true,"delta_db":-0.6074891388514565},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.417829557129175e-08,"output_power":3.892982859608942e-08,"reference_active":true,"delta_db":-0.5492646491217565}],"active_band_count":3,"active_band_mae_db":0.4715893751878028}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0011","method":"magphase","status":"measured","mos":{"status":"measured","value":3.034853219985962,"predicted_mos":3.034853219985962,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289919,"input_sample_rate":48000,"input_seconds":6.039979166666667,"input_float32_sha256":"243565f105501e3c6c0af8f3c1e94c51086798adb0c9b198c1dd2a25cb19fdc9","internal_samples":96640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.031486672000028},"started_at":"2026-10-13T05:11:51.795411+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00988081900140969,"native_input_samples":288000,"native_input_sha256":"fc77304b714105de39d05267ca819bc64c81c18e1c887f2f1389c86c1b6160f4","output_sha256":"99a999d71f996d0e0bbe073584b6b1694a5fde551f0106472c856baa21eb9411","output_samples":289919,"length_delta_samples":1919,"analysis_s":2.7261447480013885,"synthesis_s":0.8447189089984022,"total_s":3.5708636569997907,"rtf":0.5951439428332984,"output_peak":0.5397496460271357,"output_rms":0.04882987267723104,"finished_at":"2026-10-13T05:11:55.389671+00:00","evaluation_input":{"reference":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"99a999d71f996d0e0bbe073584b6b1694a5fde551f0106472c856baa21eb9411","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96640,"length_delta_samples":640,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.22127710707769,"output_peak":0.5335993443623145},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9933774834437086,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.516951619595613,"spectral_convergence":0.8753068275824724,"spectral_convergence_status":"ok","amplitude_floor":1.4412887028114478e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.945198880514983,"spectral_convergence":0.6578666138193592,"spectral_convergence_status":"ok","amplitude_floor":8.082487629299873e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.040316373332347,"spectral_convergence":0.6338534418654975,"spectral_convergence_status":"ok","amplitude_floor":6.705293681874596e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00036448625115270087,"output_power":0.0003316428351154585,"delta_db":-0.41010531725112365,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008505212390156939,"output_power":0.0007993512639934495,"delta_db":-0.26947497147576355,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010532918945599765,"output_power":0.0009829061967392598,"delta_db":-0.30036669064731,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002198202138287696,"output_power":0.00021845140259681338,"delta_db":-0.027127884874452973,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.872413022951283e-05,"output_power":3.964128070801755e-05,"delta_db":0.10166004990461883,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.441903264074617e-05,"output_power":2.6817335009212497e-05,"delta_db":0.40687161965667495,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25260108896832395,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2675928256715418,"pitch":null,"output_rms_dbfs":-26.200040924947118,"output_peak":0.5335993443623145,"reference_rms_dbfs":-25.93244809927558,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.167488957814314},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.5454751203413323,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.050510026345542525,"rms_dbfs":-25.932448099275575,"peak":0.5145263671875,"peak_dbfs":-5.771847293419889,"crest_factor_db":20.160600805855687,"mean":-9.429613749186198e-05,"zero_energy":false},"output":{"rms":0.04897765117032678,"rms_dbfs":-26.200040924947118,"peak":0.5335993443623145,"peak_dbfs":-5.455694254244088,"crest_factor_db":20.74434667070303,"mean":-0.0002178246988117393,"zero_energy":false},"rms_difference_db":-0.2675928256715423},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.977121322765015e-06,"threshold_dbfs":-55.26203466752751,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.1715363426547912e-06,"output_mean_power":2.8410451509511725e-05,"output_rms_dbfs":-45.46521864252027,"output_zero_energy":false,"output_peak":0.08852169677216694,"output_energy_fraction":0.0017765338514959311,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":129,"output_estimated_voiced_frames":132,"both_estimated_voiced_frames":95,"reference_voiced_output_unvoiced_frames":34,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.7364341085271318,"absolute_difference_cents_median":6.79404463222267,"absolute_difference_cents_p95":34.09152591681406,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019142337741905198,"log_rms_mae_db":5.258426225393876,"rms_envelope_nrmse":0.5040305737494805,"envelope_correlation":0.7531583023355833,"rms_gain_error_db":-0.26755553020761874,"reference_mean_power":0.0025533227171819564,"output_mean_power":0.002400767792594955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1748402838347296e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007147512141697323,"output_power":0.0006869453223026544,"reference_active":true,"delta_db":-0.17232731193862785},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.175142428480134e-05,"output_power":8.436753319185742e-05,"reference_active":true,"delta_db":0.1368002329165532},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4312039590506752e-05,"output_power":1.3328263161475801e-05,"reference_active":true,"delta_db":-0.309279697964724}],"active_band_count":3,"active_band_mae_db":0.206135747606635}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016640454830055276,"log_rms_mae_db":4.353639773485667,"rms_envelope_nrmse":0.43229445007583367,"envelope_correlation":0.8030668740590989,"rms_gain_error_db":-0.27178255807887375,"reference_mean_power":0.002564981061322552,"output_mean_power":0.0024093833621097282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0626539543337418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006440432662381922,"output_power":0.0006195489573812845,"reference_active":true,"delta_db":-0.16839413391812325},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1296244809426825e-05,"output_power":4.3080649594888955e-05,"reference_active":true,"delta_db":0.18371681457349004},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.585289215752229e-07,"output_power":8.368697988883811e-07,"reference_active":true,"delta_db":-0.11097034737120426}],"active_band_count":3,"active_band_mae_db":0.15436043195427251}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016776043588712122,"log_rms_mae_db":7.1623805152061975,"rms_envelope_nrmse":0.5692454534952699,"envelope_correlation":0.750207379825558,"rms_gain_error_db":-0.2933169254223538,"reference_mean_power":0.0015150798695545982,"output_mean_power":0.0014161323221472343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.503421266374217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005710315291680691,"output_power":0.0005390994548525635,"reference_active":true,"delta_db":-0.24991195558602505},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462070521392158e-05,"output_power":6.259369113542268e-05,"reference_active":true,"delta_db":-0.13841130664915005},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.823790922205639e-06,"output_power":8.87499480691897e-06,"reference_active":true,"delta_db":-0.4411100328633314}],"active_band_count":3,"active_band_mae_db":0.2764777650328355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015135634256581334,"log_rms_mae_db":6.066373291198541,"rms_envelope_nrmse":0.487920691162802,"envelope_correlation":0.8027301413614814,"rms_gain_error_db":-0.2992484036741994,"reference_mean_power":0.0015225180811750088,"output_mean_power":0.0014211424709932709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.758019890177841e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005253930575206644,"output_power":0.0004970242664147727,"reference_active":true,"delta_db":-0.24106736789666755},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0513708673138778e-05,"output_power":3.017474922450974e-05,"reference_active":true,"delta_db":-0.04851326066633383},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.334657094067361e-07,"output_power":7.287285185122711e-07,"reference_active":true,"delta_db":-0.5832197090531873}],"active_band_count":3,"active_band_mae_db":0.2909334458720629}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.52680657351613e-05,"log_rms_mae_db":6.347170019816786,"rms_envelope_nrmse":0.5465735863305697,"envelope_correlation":0.7833757699146796,"rms_gain_error_db":-0.0425069238817425,"reference_mean_power":0.00023682053730462446,"output_mean_power":0.00023451394353062995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.163166626171004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014015200472863973,"output_power":0.00013879324516485438,"reference_active":true,"delta_db":-0.042309840245703015},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.313645741402795e-05,"output_power":1.3210506636421002e-05,"reference_active":true,"delta_db":0.024412113129960895},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2309159151114036e-06,"output_power":2.9343395276972495e-06,"reference_active":true,"delta_db":-0.41815291680979}],"active_band_count":3,"active_band_mae_db":0.16162495672848465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.30167762571354e-05,"log_rms_mae_db":4.927869136671681,"rms_envelope_nrmse":0.45401571875119845,"envelope_correlation":0.8405588602607744,"rms_gain_error_db":-0.05050682753575953,"reference_mean_power":0.0002379882918616052,"output_mean_power":0.0002352366083795393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.020973483325393e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013180546055965978,"output_power":0.00013071696350288217,"reference_active":true,"delta_db":-0.03601452150697715},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.132933188997126e-06,"output_power":6.170296531135744e-06,"reference_active":true,"delta_db":0.02637802198241446},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2319303983633038e-07,"output_power":2.172467164601446e-07,"reference_active":true,"delta_db":-0.11727426122204178}],"active_band_count":3,"active_band_mae_db":0.05988893490381112}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.803319177623984e-05,"log_rms_mae_db":6.698683283045671,"rms_envelope_nrmse":0.7384480033865273,"envelope_correlation":0.6487543167147044,"rms_gain_error_db":0.6213188906813041,"reference_mean_power":3.051701314781676e-05,"output_mean_power":3.52106395821586e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.8263265386055575e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1311662288856238e-05,"output_power":1.2415465912726321e-05,"reference_active":true,"delta_db":0.40436591150934686},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.547549874304093e-06,"output_power":2.683385100432509e-06,"reference_active":true,"delta_db":0.22560308912528468},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.732573657220465e-07,"output_power":8.818114760945475e-07,"reference_active":true,"delta_db":1.8702610289319845}],"active_band_count":3,"active_band_mae_db":0.8334100098555387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4591770468872877e-05,"log_rms_mae_db":5.41684820817423,"rms_envelope_nrmse":0.6144951283521317,"envelope_correlation":0.7383781986429281,"rms_gain_error_db":0.6154142674735346,"reference_mean_power":3.0668913834731314e-05,"output_mean_power":3.5337825520871e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6363060360497885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1094596811060145e-05,"output_power":1.2280886412903931e-05,"reference_active":true,"delta_db":0.4411819040134987},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.190115655321781e-06,"output_power":1.3095424589457946e-06,"reference_active":true,"delta_db":0.4153041586543315},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.417829557129175e-08,"output_power":6.904406396611491e-08,"reference_active":true,"delta_db":1.9391738975523856}],"active_band_count":3,"active_band_mae_db":0.9318866534067386}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"hnm","status":"measured","mos":{"status":"measured","value":3.0739176273345947,"predicted_mos":3.0739176273345947,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65440,"input_sample_rate":16000,"input_seconds":4.09,"input_float32_sha256":"94d40c5166df923073d285e081db3baf330ae5cd19bdbb19942ba58e88e41efc","internal_samples":65440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.1954971739996836},"started_at":"2026-10-13T04:37:57.663831+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4881999959470704e-05,"native_input_samples":65440,"native_input_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","output_sha256":"0be0c02eec6224b11a3fc6206e0182efdd59b1896c5e6496623229d33bca111a","output_samples":65440,"length_delta_samples":0,"analysis_s":1.1144753420003326,"synthesis_s":0.35441092800010665,"total_s":1.4688862700004393,"rtf":0.3591408973106209,"output_peak":0.4027166768176805,"output_rms":0.04638579944852163,"finished_at":"2026-10-13T04:37:59.137830+00:00","evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"0be0c02eec6224b11a3fc6206e0182efdd59b1896c5e6496623229d33bca111a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.307369409989189,"spectral_convergence":0.2845713449996821,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.419722571029543,"spectral_convergence":0.2522289149274661,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8526978425275775,"spectral_convergence":0.2826776818454911,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.0002994041213247997,"delta_db":0.1123062584011107,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0009645056676986042,"delta_db":0.09234038715714063,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0007834776517845299,"delta_db":-0.14191237575712315,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":6.518156452939091e-05,"delta_db":-0.3961762920968924,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":2.9324104172592867e-05,"delta_db":0.11482915527956333,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":9.749273708628632e-06,"delta_db":0.5297692615542148,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2312222883743408,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.004904831674896172,"pitch":null,"output_rms_dbfs":-26.6722990807418,"output_peak":0.4027166768176805,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.859929941182102},"common_prefix":{"samples":65440,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.307369409989189,"spectral_convergence":0.2845713449996821,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.419722571029543,"spectral_convergence":0.2522289149274661,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8526978425275775,"spectral_convergence":0.2826776818454911,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.0002994041213247997,"delta_db":0.1123062584011107,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0009645056676986042,"delta_db":0.09234038715714063,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0007834776517845299,"delta_db":-0.14191237575712315,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":6.518156452939091e-05,"delta_db":-0.3961762920968924,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":2.9324104172592867e-05,"delta_db":0.11482915527956333,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":9.749273708628632e-06,"delta_db":0.5297692615542148,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2312222883743408,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.004904831674896172,"pitch":null,"output_rms_dbfs":-26.6722990807418,"output_peak":0.4027166768176805,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.859929941182102},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65440,"output_samples":65440,"intelligibility":{"estoi":{"value":0.8516054824060347,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046412000424763004,"rms_dbfs":-26.667394249066906,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.473615807431017,"mean":-9.563744505343636e-05,"zero_energy":false},"output":{"rms":0.04638579944852163,"rms_dbfs":-26.6722990807418,"peak":0.4027166768176805,"peak_dbfs":-7.9000077108098825,"crest_factor_db":18.77229136993192,"mean":-0.00014401200131943965,"zero_energy":false},"rms_difference_db":-0.00490483167489586},"full_scale":{"reference":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":409,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":6.057582671993074e-07,"output_mean_power":6.933760521828222e-07,"output_rms_dbfs":-61.590311620803384,"output_zero_energy":false,"output_peak":0.00921747438701376,"output_energy_fraction":8.903358921670414e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":402,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":92,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9770114942528736,"absolute_difference_cents_median":1.3553226717603986,"absolute_difference_cents_p95":6.577228977942179,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00016079151656388168,"log_rms_mae_db":1.246923711888013,"rms_envelope_nrmse":0.08503330065774942,"envelope_correlation":0.9936343097612488,"rms_gain_error_db":-0.004905661911683199,"reference_mean_power":0.0021567095381351967,"output_mean_power":0.0021542747582683867},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.6620838184249172e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000592954618765589,"output_power":0.0005923933371120278,"reference_active":true,"delta_db":-0.004112911182661807},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014246995030046006,"output_power":0.0001448810255377999,"reference_active":true,"delta_db":0.07288238661323178},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9396326507323893e-05,"output_power":2.410597728266037e-05,"reference_active":true,"delta_db":0.9440525676414933}],"active_band_count":3,"active_band_mae_db":0.340349288479129}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00012458719833187296,"log_rms_mae_db":0.6722753583319012,"rms_envelope_nrmse":0.04330466332144277,"envelope_correlation":0.9982438744040437,"rms_gain_error_db":-0.004911071071003548,"reference_mean_power":0.0021726617564687806,"output_mean_power":0.002170206264617111},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.5403145620848265e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005607371759141369,"output_power":0.0005611803185653966,"reference_active":true,"delta_db":0.003430812320755952},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.84281540251884e-05,"output_power":7.100189967526015e-05,"reference_active":true,"delta_db":0.16035144277175314},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9268319788025274e-06,"output_power":1.966789108811884e-06,"reference_active":true,"delta_db":0.08913949197319629}],"active_band_count":3,"active_band_mae_db":0.08430724902190179}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00013740128966162354,"log_rms_mae_db":1.3738835089678332,"rms_envelope_nrmse":0.1095710439351135,"envelope_correlation":0.991275230509578,"rms_gain_error_db":-0.024384802970663977,"reference_mean_power":0.001344215530062727,"output_mean_power":0.0013366891667548127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.2437342579426783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004237422220408671,"output_power":0.00041383705291962237,"reference_active":true,"delta_db":-0.10272366987130918},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010753440950326828,"output_power":0.00011201427433272035,"reference_active":true,"delta_db":0.17725915078984997},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.157936749643158e-05,"output_power":1.6396335598660705e-05,"reference_active":true,"delta_db":1.5106196140407036}],"active_band_count":3,"active_band_mae_db":0.5968674782339543}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00011336523032200158,"log_rms_mae_db":0.9973334403499283,"rms_envelope_nrmse":0.05665188328239605,"envelope_correlation":0.997556331443211,"rms_gain_error_db":-0.02438494087178222,"reference_mean_power":0.0013541602801659725,"output_mean_power":0.001346578192700534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.1658618871138857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004110855381392752,"output_power":0.00040540988078356416,"reference_active":true,"delta_db":-0.060378694503899114},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.54169136496196e-05,"output_power":4.9909691913044335e-05,"reference_active":true,"delta_db":0.40967271038980463},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3416624074278085e-06,"output_power":1.4945015826916245e-06,"reference_active":true,"delta_db":0.46853128021598683}],"active_band_count":3,"active_band_mae_db":0.3128608950365635}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":1.8834156683192522,"rms_envelope_nrmse":0.23318790050402313,"envelope_correlation":0.9569456839842427,"rms_gain_error_db":-0.1522746881785807,"reference_mean_power":7.742789466080947e-05,"output_mean_power":7.476011860770797e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.015275941663668e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8994572701205544e-05,"output_power":3.550694148525385e-05,"reference_active":true,"delta_db":-0.40690901309357563},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.452200984753589e-06,"output_power":6.5826658303128555e-06,"reference_active":true,"delta_db":-0.5388275059082299},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.182412963771586e-06,"output_power":2.53490324409368e-06,"reference_active":true,"delta_db":3.3119220459014316}],"active_band_count":3,"active_band_mae_db":1.4192195216344123}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":1.1832301872499706,"rms_envelope_nrmse":0.14369507621613248,"envelope_correlation":0.9826406796775469,"rms_gain_error_db":-0.15227594163310443,"reference_mean_power":7.800066882867518e-05,"output_mean_power":7.5313136120067e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":6.50533884720925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.845568227980426e-05,"output_power":3.53802934358954e-05,"reference_active":true,"delta_db":-0.36199090476055895},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.058614755593347e-06,"output_power":2.7721648658927937e-06,"reference_active":true,"delta_db":-0.42705724329842437},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905753913869999e-08,"output_power":2.128248755610282e-07,"reference_active":true,"delta_db":3.3213485341907685}],"active_band_count":3,"active_band_mae_db":1.370132227416584}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":1.9476814776617704,"rms_envelope_nrmse":0.3297344733159204,"envelope_correlation":0.9243498271258311,"rms_gain_error_db":0.10112708468845272,"reference_mean_power":1.4177032937223138e-05,"output_mean_power":1.4511023880479553e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.931109844663439e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.831691262181725e-06,"output_power":4.225006684661345e-06,"reference_active":true,"delta_db":0.4243689179389478},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0928531053800198e-06,"output_power":1.4909211453573369e-06,"reference_active":true,"delta_db":1.3489288360808336},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4875100105629565e-07,"output_power":1.164537791700596e-06,"reference_active":true,"delta_db":8.936936897672256}],"active_band_count":3,"active_band_mae_db":3.570078217230679}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":1.4068648710279854,"rms_envelope_nrmse":0.261063119000551,"envelope_correlation":0.9491797278968143,"rms_gain_error_db":0.10112638677285213,"reference_mean_power":1.428188481146513e-05,"output_mean_power":1.4618343568127539e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":7.397985811629504e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7144994114751925e-06,"output_power":4.769231590374191e-06,"reference_active":true,"delta_db":1.085481180910062},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.071953545728326e-07,"output_power":8.385441774102032e-07,"reference_active":true,"delta_db":2.183506805142103},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3628863245634317e-08,"output_power":1.5976536818205707e-07,"reference_active":true,"delta_db":10.690230108105807}],"active_band_count":3,"active_band_mae_db":4.653072698052657}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.3701260089874268,"predicted_mos":3.3701260089874268,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65440,"input_sample_rate":16000,"input_seconds":4.09,"input_float32_sha256":"7e77f4bf4c9b035a1daebbde3868d451fa5761cbb2318a315a5225b55d926524","internal_samples":65440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.1403049809996446},"started_at":"2026-10-13T04:40:48.384980+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.031500025623245e-05,"native_input_samples":65440,"native_input_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","output_sha256":"6394df1140b6777934311a8dc28ba7822d5a141b547a43f24c38dd1235b1e086","output_samples":65440,"length_delta_samples":0,"analysis_s":1.3243961360003595,"synthesis_s":0.057936055999562086,"total_s":1.3823321919999216,"rtf":0.3379785310513256,"output_peak":0.44440785869589405,"output_rms":0.050380883789996465,"finished_at":"2026-10-13T04:40:49.770824+00:00","evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"6394df1140b6777934311a8dc28ba7822d5a141b547a43f24c38dd1235b1e086","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.539302461401627,"spectral_convergence":0.7945504544560746,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.225106761223228,"spectral_convergence":0.30479525423879955,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.613695197065775,"spectral_convergence":0.2553169787623434,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.0003078454928888591,"delta_db":0.2330565123485801,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0011418791956395876,"delta_db":0.825494151356355,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0009672568218154285,"delta_db":0.7732395079669779,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":8.147021530095487e-05,"delta_db":0.5725645434370039,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":3.0474540968347456e-05,"delta_db":0.281953345315994,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":9.3071585752328e-06,"delta_db":0.3282177635469644,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5024209706619792,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7127098829924107,"pitch":null,"output_rms_dbfs":-25.954684366074492,"output_peak":0.44440785869589405,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.126034806563543},"common_prefix":{"samples":65440,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.539302461401627,"spectral_convergence":0.7945504544560746,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.225106761223228,"spectral_convergence":0.30479525423879955,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.613695197065775,"spectral_convergence":0.2553169787623434,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.0003078454928888591,"delta_db":0.2330565123485801,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0011418791956395876,"delta_db":0.825494151356355,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0009672568218154285,"delta_db":0.7732395079669779,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":8.147021530095487e-05,"delta_db":0.5725645434370039,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":3.0474540968347456e-05,"delta_db":0.281953345315994,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":9.3071585752328e-06,"delta_db":0.3282177635469644,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5024209706619792,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7127098829924107,"pitch":null,"output_rms_dbfs":-25.954684366074492,"output_peak":0.44440785869589405,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.126034806563543},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65440,"output_samples":65440,"intelligibility":{"estoi":{"value":0.9282090581218657,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046412000424763004,"rms_dbfs":-26.667394249066906,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.473615807431017,"mean":-9.563744505343636e-05,"zero_energy":false},"output":{"rms":0.050380883789996465,"rms_dbfs":-25.954684366074492,"peak":0.44440785869589405,"peak_dbfs":-7.044365396149739,"crest_factor_db":18.910318969924756,"mean":2.287307141580476e-09,"zero_energy":false},"rms_difference_db":0.7127098829924137},"full_scale":{"reference":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":409,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":6.057582671993074e-07,"output_mean_power":6.067690612123029e-07,"output_rms_dbfs":-62.16976571733903,"output_zero_energy":false,"output_peak":0.014952132224651825,"output_energy_fraction":6.604607301931978e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":402,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":107,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9770114942528736,"absolute_difference_cents_median":3.30018712135585,"absolute_difference_cents_p95":14.152256521495973,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00016079151656388168,"log_rms_mae_db":3.056133323721249,"rms_envelope_nrmse":0.1816856307239993,"envelope_correlation":0.9791827019659032,"rms_gain_error_db":0.712711451417072,"reference_mean_power":0.0021567095381351967,"output_mean_power":0.0025413401870800927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.6620838184249172e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000592954618765589,"output_power":0.0007146953937635795,"reference_active":true,"delta_db":0.8109952675845784},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014246995030046006,"output_power":0.00012670918979876036,"reference_active":true,"delta_db":-0.5091515883229693},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9396326507323893e-05,"output_power":1.2982770416853868e-05,"reference_active":true,"delta_db":-1.7435210894610331}],"active_band_count":3,"active_band_mae_db":1.0212226484561937}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00012458719833187296,"log_rms_mae_db":2.501003503209954,"rms_envelope_nrmse":0.1266714472068684,"envelope_correlation":0.9925166029958478,"rms_gain_error_db":0.7127177751419801,"reference_mean_power":0.0021726617564687806,"output_mean_power":0.002560141074534939},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.5403145620848265e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005607371759141369,"output_power":0.0006762758789334859,"reference_active":true,"delta_db":0.8136454747219464},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.84281540251884e-05,"output_power":5.6926160230099185e-05,"reference_active":true,"delta_db":-0.7992293339432959},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9268319788025274e-06,"output_power":9.531099593173809e-07,"reference_active":true,"delta_db":-3.0570083786390487}],"active_band_count":3,"active_band_mae_db":1.5566277291014303}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00013740128966162354,"log_rms_mae_db":2.3208421626375553,"rms_envelope_nrmse":0.19223489447032988,"envelope_correlation":0.9826668637344919,"rms_gain_error_db":0.8517245844911324,"reference_mean_power":0.001344215530062727,"output_mean_power":0.001635465424319207},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.2437342579426783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004237422220408671,"output_power":0.0005225988476957666,"reference_active":true,"delta_db":0.9106670816063751},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010753440950326828,"output_power":0.00011050716725995355,"reference_active":true,"delta_db":0.11842991716355325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.157936749643158e-05,"output_power":8.316099372413104e-06,"reference_active":true,"delta_db":-1.4376516718041104}],"active_band_count":3,"active_band_mae_db":0.8222495568580129}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00011336523032200158,"log_rms_mae_db":2.140798587240498,"rms_envelope_nrmse":0.13595962585544447,"envelope_correlation":0.9945816895968941,"rms_gain_error_db":0.8517243182947765,"reference_mean_power":0.0013541602801659725,"output_mean_power":0.0016475647929375067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.1658618871138857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004110855381392752,"output_power":0.0004985945724240708,"reference_active":true,"delta_db":0.8381534786729955},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.54169136496196e-05,"output_power":4.76785138817232e-05,"reference_active":true,"delta_db":0.21105092191366207},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3416624074278085e-06,"output_power":7.605935349892639e-07,"reference_active":true,"delta_db":-2.464906216682516}],"active_band_count":3,"active_band_mae_db":1.1713702057563913}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":1.6130831469466056,"rms_envelope_nrmse":0.19560405984210122,"envelope_correlation":0.9745305242610994,"rms_gain_error_db":0.5108795171074247,"reference_mean_power":7.742789466080947e-05,"output_mean_power":8.709343151281437e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.015275941663668e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8994572701205544e-05,"output_power":4.3586057248469256e-05,"reference_active":true,"delta_db":0.48343419200120874},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.452200984753589e-06,"output_power":8.055140273663013e-06,"reference_active":true,"delta_db":0.3378854846356811},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.182412963771586e-06,"output_power":1.0372417924020855e-06,"reference_active":true,"delta_db":-0.5688917562383994}],"active_band_count":3,"active_band_mae_db":0.46340381095842975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":0.8888607826729513,"rms_envelope_nrmse":0.10229024011931548,"envelope_correlation":0.9946443261723285,"rms_gain_error_db":0.510879972980006,"reference_mean_power":7.800066882867518e-05,"output_mean_power":8.773771586366109e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":6.50533884720925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.845568227980426e-05,"output_power":4.2686675224996614e-05,"reference_active":true,"delta_db":0.4533180883923538},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.058614755593347e-06,"output_power":3.4984486569137077e-06,"reference_active":true,"delta_db":0.5835072525455155},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905753913869999e-08,"output_power":8.232772218991668e-08,"reference_active":true,"delta_db":-0.8034143501901589}],"active_band_count":3,"active_band_mae_db":0.6134132303760095}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":1.737234328373204,"rms_envelope_nrmse":0.2525527499192752,"envelope_correlation":0.9574983618975296,"rms_gain_error_db":0.30492558504918843,"reference_mean_power":1.4177032937223138e-05,"output_mean_power":1.5208203209028913e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.931109844663439e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.831691262181725e-06,"output_power":4.2041684862567075e-06,"reference_active":true,"delta_db":0.40289603872411045},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0928531053800198e-06,"output_power":1.0802405820841503e-06,"reference_active":true,"delta_db":-0.050413020003247065},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4875100105629565e-07,"output_power":1.8876407303641835e-07,"reference_active":true,"delta_db":1.034594426128256}],"active_band_count":3,"active_band_mae_db":0.4959678282852045}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":0.8697521851543131,"rms_envelope_nrmse":0.12662609573426842,"envelope_correlation":0.9893038921940455,"rms_gain_error_db":0.30492651034921214,"reference_mean_power":1.428188481146513e-05,"output_mean_power":1.5320684776505884e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":7.397985811629504e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7144994114751925e-06,"output_power":4.095512584308837e-06,"reference_active":true,"delta_db":0.4240797085508535},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.071953545728326e-07,"output_power":5.380169678836353e-07,"reference_active":true,"delta_db":0.25620705417260164},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3628863245634317e-08,"output_power":1.3900384608875677e-08,"reference_active":true,"delta_db":0.08567183116924607}],"active_band_count":3,"active_band_mae_db":0.25531953129756707}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"facodec","status":"measured","mos":{"status":"measured","value":4.185952186584473,"predicted_mos":4.185952186584473,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65600,"input_sample_rate":16000,"input_seconds":4.1,"input_float32_sha256":"7507686925893946c6eda1c73be0ce493e9315518f5fbbba0cb9a4d970ecd44d","internal_samples":65600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2750019849991077},"started_at":"2026-10-13T04:44:39.739478+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1439999727590475e-05,"native_input_samples":65440,"native_input_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","output_sha256":"9a5eb45ec742e01085c7e9e8f57dacac1ba37540d3b332af2999577fa37616e1","output_samples":65600,"length_delta_samples":160,"analysis_s":1.3175110510001105,"synthesis_s":11.536801480000577,"total_s":12.854312531000687,"rtf":3.1428636995111705,"output_peak":0.4102873206138611,"output_rms":0.044536610119173814,"finished_at":"2026-10-13T04:44:52.599865+00:00","evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"9a5eb45ec742e01085c7e9e8f57dacac1ba37540d3b332af2999577fa37616e1","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":65440,"output_samples":65600,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.025656842835218,"output_peak":0.4102873206138611},"common_prefix":{"samples":65440,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975609756097561,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.204417691732248,"spectral_convergence":0.1865511364411807,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.086898325466797,"spectral_convergence":0.18235174167210394,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.3451681131079285,"spectral_convergence":0.20318997895111424,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.0003162585309297838,"delta_db":0.35015100926786946,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.000850096329817499,"delta_db":-0.4560260413412941,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0007245350142552309,"delta_db":-0.4815847232055669,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":6.407777692538582e-05,"delta_db":-0.4703497348205585,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":2.5467320276778296e-05,"delta_db":-0.49758588101713436,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":7.92430049148543e-06,"delta_db":-0.3703440138856253,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.43767356725634143,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3476571613246746,"pitch":null,"output_rms_dbfs":-27.015051410391578,"output_peak":0.4102873206138611,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.21216137676899},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65440,"output_samples":65440,"intelligibility":{"estoi":{"value":0.9237241777265291,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046412000424763004,"rms_dbfs":-26.667394249066906,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.473615807431017,"mean":-9.563744505343636e-05,"zero_energy":false},"output":{"rms":0.044591022332935874,"rms_dbfs":-27.01505141039158,"peak":0.4102873206138611,"peak_dbfs":-7.738238082427288,"crest_factor_db":19.27681332796429,"mean":1.4799946253302341e-05,"zero_energy":false},"rms_difference_db":-0.34765716132467617},"full_scale":{"reference":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":409,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":6.057582671993074e-07,"output_mean_power":5.733408486666487e-07,"output_rms_dbfs":-62.415871150567696,"output_zero_energy":false,"output_peak":0.00561551796272397,"output_energy_fraction":7.96660274255331e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":402,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":94,"both_estimated_voiced_frames":87,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":2.8079873562812323,"absolute_difference_cents_p95":7.713135265951112,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00016079151656388168,"log_rms_mae_db":1.4563475415737732,"rms_envelope_nrmse":0.09861433641221318,"envelope_correlation":0.9926979457743017,"rms_gain_error_db":-0.3476557330012386,"reference_mean_power":0.0021567095381351967,"output_mean_power":0.0019907929115572406},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.6620838184249172e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000592954618765589,"output_power":0.00046888838404929436,"reference_active":true,"delta_db":-1.0195198237950378},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014246995030046006,"output_power":0.00011998691668564093,"reference_active":true,"delta_db":-0.7458937942882687},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9396326507323893e-05,"output_power":1.4201855723502544e-05,"reference_active":true,"delta_db":-1.3537438984241783}],"active_band_count":3,"active_band_mae_db":1.0397191721691617}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00012458719833187296,"log_rms_mae_db":1.1392014538152653,"rms_envelope_nrmse":0.09093083004034361,"envelope_correlation":0.9936627967399132,"rms_gain_error_db":-0.3476538771571398,"reference_mean_power":0.0021726617564687806,"output_mean_power":0.002005518775626261},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.5403145620848265e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005607371759141369,"output_power":0.000445809006618492,"reference_active":true,"delta_db":-0.9961051180816716},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.84281540251884e-05,"output_power":5.6168903507238286e-05,"reference_active":true,"delta_db":-0.8573887832028966},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9268319788025274e-06,"output_power":1.4659880313832887e-06,"reference_active":true,"delta_db":-1.1871342084774408}],"active_band_count":3,"active_band_mae_db":1.013542703254003}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00013740128966162354,"log_rms_mae_db":1.5152299145629295,"rms_envelope_nrmse":0.08613869716194791,"envelope_correlation":0.995966500421705,"rms_gain_error_db":-0.38096283385621754,"reference_mean_power":0.001344215530062727,"output_mean_power":0.0012313247959659427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.2437342579426783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004237422220408671,"output_power":0.0003635503282833959,"reference_active":true,"delta_db":-0.6653719848925543},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010753440950326828,"output_power":9.526151889153084e-05,"reference_active":true,"delta_db":-0.5262995282223413},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.157936749643158e-05,"output_power":9.770015813595469e-06,"reference_active":true,"delta_db":-0.7378957076850008}],"active_band_count":3,"active_band_mae_db":0.6431890735999655}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00011336523032200158,"log_rms_mae_db":1.4180202061609977,"rms_envelope_nrmse":0.07423087709806739,"envelope_correlation":0.9973286142481227,"rms_gain_error_db":-0.380963342718599,"reference_mean_power":0.0013541602801659725,"output_mean_power":0.0012404342146176278},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.1658618871138857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004110855381392752,"output_power":0.0003527136841351427,"reference_active":true,"delta_db":-0.6650988950786563},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.54169136496196e-05,"output_power":4.080880326969935e-05,"reference_active":true,"delta_db":-0.4646375883898965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3416624074278085e-06,"output_power":1.0994253765208335e-06,"reference_active":true,"delta_db":-0.8647749430957932}],"active_band_count":3,"active_band_mae_db":0.6648371421881153}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":1.5776027949698666,"rms_envelope_nrmse":0.13421226309899972,"envelope_correlation":0.9870703687017421,"rms_gain_error_db":-0.4197850422658861,"reference_mean_power":7.742789466080947e-05,"output_mean_power":7.029411156207393e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.015275941663668e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8994572701205544e-05,"output_power":3.6759120417187e-05,"reference_active":true,"delta_db":-0.25639054830984725},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.452200984753589e-06,"output_power":6.04706024150228e-06,"reference_active":true,"delta_db":-0.9074026412127688},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.182412963771586e-06,"output_power":1.0491458466769554e-06,"reference_active":true,"delta_db":-0.519333169066455}],"active_band_count":3,"active_band_mae_db":0.5610421195296903}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":1.2452224485201566,"rms_envelope_nrmse":0.11114776680940672,"envelope_correlation":0.991096018357363,"rms_gain_error_db":-0.4197887338658696,"reference_mean_power":7.800066882867518e-05,"output_mean_power":7.081405325456455e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":6.50533884720925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.845568227980426e-05,"output_power":3.587020395176519e-05,"reference_active":true,"delta_db":-0.302266749955886},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.058614755593347e-06,"output_power":2.4931165763918687e-06,"reference_active":true,"delta_db":-0.8878219306714169},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905753913869999e-08,"output_power":9.150764573976661e-08,"reference_active":true,"delta_db":-0.344301525167262}],"active_band_count":3,"active_band_mae_db":0.5114634019315215}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":2.0373476697051403,"rms_envelope_nrmse":0.24937834471377623,"envelope_correlation":0.959150761666507,"rms_gain_error_db":-0.9244770054277747,"reference_mean_power":1.4177032937223138e-05,"output_mean_power":1.1458760644878836e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.931109844663439e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.831691262181725e-06,"output_power":3.7030703590352005e-06,"reference_active":true,"delta_db":-0.1482854481336646},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0928531053800198e-06,"output_power":8.330555063583516e-07,"reference_active":true,"delta_db":-1.1788785132425144},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4875100105629565e-07,"output_power":1.5703969560402855e-07,"reference_active":true,"delta_db":0.23549547681862637}],"active_band_count":3,"active_band_mae_db":0.5208864793982685}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":1.54593597947905,"rms_envelope_nrmse":0.2140152473713943,"envelope_correlation":0.9703916617700785,"rms_gain_error_db":-0.9244838863447846,"reference_mean_power":1.428188481146513e-05,"output_mean_power":1.154349016804351e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":7.397985811629504e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7144994114751925e-06,"output_power":3.5412670304259004e-06,"reference_active":true,"delta_db":-0.20741617761501796},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.071953545728326e-07,"output_power":3.9118326200294944e-07,"reference_active":true,"delta_db":-1.1279500329946548},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3628863245634317e-08,"output_power":1.1920668451638523e-08,"reference_active":true,"delta_db":-0.5815902461461021}],"active_band_count":3,"active_band_mae_db":0.6389854855852583}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.237227439880371,"predicted_mos":3.237227439880371,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":180736,"input_sample_rate":44100,"input_seconds":4.098321995464852,"input_float32_sha256":"1278f500dab3167c1dbbcf46f715d14f66ab15d739e0357efd5a391c9fc0c19c","internal_samples":65574,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2097118889996636},"started_at":"2026-10-13T04:57:29.732552+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004853569999795582,"native_input_samples":180369,"native_input_sha256":"50e26abf4613a10ff529b5221e04dfab1d313491efa68359f21f802b340af297","output_sha256":"b40848ee91f48fe9ecfc065610ad1f42fb9252b169c202a868b2c6b1878d198d","output_samples":180736,"length_delta_samples":367,"analysis_s":0.11494023299928813,"synthesis_s":20.656419084000845,"total_s":20.771359317000133,"rtf":5.078571960146732,"output_peak":0.42995449900627136,"output_rms":0.042083861401654064,"finished_at":"2026-10-13T04:57:50.517250+00:00","evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"b40848ee91f48fe9ecfc065610ad1f42fb9252b169c202a868b2c6b1878d198d","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":65440,"output_samples":65574,"length_delta_samples":134,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.51144773461472,"output_peak":0.4255705024195256},"common_prefix":{"samples":65440,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.997956507152225,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.343489729587208,"spectral_convergence":0.4699797029339651,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.7950597379014,"spectral_convergence":0.3431284055382566,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.565112364967858,"spectral_convergence":0.4333894349545896,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.00024244076099249636,"delta_db":-0.8042150977242825,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0007613109417054816,"delta_db":-0.9350867370365796,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0006887300322296353,"delta_db":-0.701688296676559,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":5.33819910867548e-05,"delta_db":-1.2634764123911957,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":2.5539863919733732e-05,"delta_db":-0.4852325917418424,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":5.8260359208101214e-06,"delta_db":-1.7062217915300257,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9826534878500808,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8351703051085178,"pitch":null,"output_rms_dbfs":-27.50256455417542,"output_peak":0.4255705024195256,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.901220610818823},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65440,"output_samples":65440,"intelligibility":{"estoi":{"value":0.7785236140840234,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046412000424763004,"rms_dbfs":-26.667394249066906,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.473615807431017,"mean":-9.563744505343636e-05,"zero_energy":false},"output":{"rms":0.042157201371762265,"rms_dbfs":-27.50256455417542,"peak":0.4255705024195256,"peak_dbfs":-7.4205696387146425,"crest_factor_db":20.08199491546078,"mean":-0.00024624427511787874,"zero_energy":false},"rms_difference_db":-0.835170305108516},"full_scale":{"reference":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":409,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":6.057582671993074e-07,"output_mean_power":4.285118886451062e-07,"output_rms_dbfs":-63.680371244961094,"output_zero_energy":false,"output_peak":0.004784906441904804,"output_energy_fraction":6.66153733700812e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":402,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":78,"both_estimated_voiced_frames":57,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.6551724137931034,"absolute_difference_cents_median":10.22107793315772,"absolute_difference_cents_p95":46.58782431840708,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00016079151656388168,"log_rms_mae_db":2.0347608335738943,"rms_envelope_nrmse":0.14513298224976764,"envelope_correlation":0.9883133344997631,"rms_gain_error_db":-0.8351703934845551,"reference_mean_power":0.0021567095381351967,"output_mean_power":0.0017794042342345992},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.6620838184249172e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000592954618765589,"output_power":0.0005252593674854216,"reference_active":true,"delta_db":-0.5264764994584069},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014246995030046006,"output_power":0.0001139300646226852,"reference_active":true,"delta_db":-0.970849290839664},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9396326507323893e-05,"output_power":1.6016246071351728e-05,"reference_active":true,"delta_db":-0.8315875355881666}],"active_band_count":3,"active_band_mae_db":0.7763044419620792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00012458719833187296,"log_rms_mae_db":1.6784366307621785,"rms_envelope_nrmse":0.11702039305385817,"envelope_correlation":0.995030409194381,"rms_gain_error_db":-0.8351694567239709,"reference_mean_power":0.0021726617564687806,"output_mean_power":0.00179256608020446},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.5403145620848265e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005607371759141369,"output_power":0.0004952930763601268,"reference_active":true,"delta_db":-0.5389709295641231},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.84281540251884e-05,"output_power":5.486560619638006e-05,"reference_active":true,"delta_db":-0.9593464229156736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9268319788025274e-06,"output_power":1.4156767968499177e-06,"reference_active":true,"delta_db":-1.3387973153037285}],"active_band_count":3,"active_band_mae_db":0.945704889261175}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00013740128966162354,"log_rms_mae_db":2.0926688144605654,"rms_envelope_nrmse":0.1345435792128374,"envelope_correlation":0.9899149648948328,"rms_gain_error_db":-0.6454611050907161,"reference_mean_power":0.001344215530062727,"output_mean_power":0.0011585713856266475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.2437342579426783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004237422220408671,"output_power":0.00037389104201136146,"reference_active":true,"delta_db":-0.5435667953917843},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010753440950326828,"output_power":9.441662737905223e-05,"reference_active":true,"delta_db":-0.564989715331911},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.157936749643158e-05,"output_power":1.0372083401087807e-05,"reference_active":true,"delta_db":-0.4781883719180039}],"active_band_count":3,"active_band_mae_db":0.5289149608805664}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00011336523032200158,"log_rms_mae_db":1.685200089182584,"rms_envelope_nrmse":0.10193104070358436,"envelope_correlation":0.9957021426736304,"rms_gain_error_db":-0.6454639716380954,"reference_mean_power":0.0013541602801659725,"output_mean_power":0.0011671419363364116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.1658618871138857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004110855381392752,"output_power":0.00036300609480766256,"reference_active":true,"delta_db":-0.5401828188066744},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.54169136496196e-05,"output_power":4.007043816146494e-05,"reference_active":true,"delta_db":-0.5439352661036598},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3416624074278085e-06,"output_power":1.1092096264622973e-06,"reference_active":true,"delta_db":-0.8262962125611611}],"active_band_count":3,"active_band_mae_db":0.6368047658238317}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":2.6054741445651532,"rms_envelope_nrmse":0.24736219693833275,"envelope_correlation":0.9559025269748733,"rms_gain_error_db":-0.8631970937594212,"reference_mean_power":7.742789466080947e-05,"output_mean_power":6.347135081110009e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.015275941663668e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8994572701205544e-05,"output_power":3.639815267988407e-05,"reference_active":true,"delta_db":-0.29924823341243945},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.452200984753589e-06,"output_power":4.914057849293141e-06,"reference_active":true,"delta_db":-1.8084429446228014},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.182412963771586e-06,"output_power":1.0426917849430749e-06,"reference_active":true,"delta_db":-0.5461323072859776}],"active_band_count":3,"active_band_mae_db":0.8846078284404061}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":2.08692516496831,"rms_envelope_nrmse":0.19569451811234317,"envelope_correlation":0.9737658489587481,"rms_gain_error_db":-0.8631988594848181,"reference_mean_power":7.800066882867518e-05,"output_mean_power":6.39408552117756e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":6.50533884720925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.845568227980426e-05,"output_power":3.521700983120068e-05,"reference_active":true,"delta_db":-0.38208042423839544},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.058614755593347e-06,"output_power":1.977269969300999e-06,"reference_active":true,"delta_db":-1.8945880899039906},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905753913869999e-08,"output_power":7.923960087741802e-08,"reference_active":true,"delta_db":-0.9694525534222252}],"active_band_count":3,"active_band_mae_db":1.0820403558548704}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":4.071890141802082,"rms_envelope_nrmse":0.303302369549553,"envelope_correlation":0.9406989489084322,"rms_gain_error_db":-1.080892260243532,"reference_mean_power":1.4177032937223138e-05,"output_mean_power":1.1053406001089203e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.931109844663439e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.831691262181725e-06,"output_power":2.5950437477984324e-06,"reference_active":true,"delta_db":-1.6924582494349565},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0928531053800198e-06,"output_power":8.94418256998783e-07,"reference_active":true,"delta_db":-0.8702113515212712},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4875100105629565e-07,"output_power":1.5322949488925965e-07,"reference_active":true,"delta_db":0.1288247286044124}],"active_band_count":3,"active_band_mae_db":0.8971647765202132}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":3.460295975761891,"rms_envelope_nrmse":0.21660683655380675,"envelope_correlation":0.9740361720414981,"rms_gain_error_db":-1.080887013573352,"reference_mean_power":1.428188481146513e-05,"output_mean_power":1.1135169304777125e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":7.397985811629504e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7144994114751925e-06,"output_power":2.5512959451758695e-06,"reference_active":true,"delta_db":-1.63139455060711},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.071953545728326e-07,"output_power":4.5283290454217987e-07,"reference_active":true,"delta_db":-0.49237290369927417},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3628863245634317e-08,"output_power":1.0743039311545277e-08,"reference_active":true,"delta_db":-1.0333246884063945}],"active_band_count":3,"active_band_mae_db":1.0523640475709262}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"golf","status":"measured","mos":{"status":"measured","value":3.4072506427764893,"predicted_mos":3.4072506427764893,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":97920,"input_sample_rate":24000,"input_seconds":4.08,"input_float32_sha256":"9e41b638f9461f27c5dcfc1d73291a3aad2d48f0a98cc74c8ed81d1e3ed381ff","internal_samples":65280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.829267218001405},"started_at":"2026-10-13T05:01:49.833928+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0020987710004192195,"native_input_samples":98160,"native_input_sha256":"adf9b8f686f3858487380ec7e58f35e84ea45deab9ed3afe90f6e57d93b241ec","output_sha256":"0c0b52f7f1929cedf8722b7b130742d85a104016c1c7e7662f3afd07b91c9c2c","output_samples":97920,"length_delta_samples":-240,"analysis_s":0.4958353249994616,"synthesis_s":0.19524542199997086,"total_s":0.6910807469994324,"rtf":0.16896839779937226,"output_peak":0.3669404983520508,"output_rms":0.03886190329155098,"finished_at":"2026-10-13T05:01:50.534228+00:00","analysis_stages":{"external_f0_s":0.045521666999775334},"evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"0c0b52f7f1929cedf8722b7b130742d85a104016c1c7e7662f3afd07b91c9c2c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":65440,"output_samples":65280,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.204147466576913,"output_peak":0.36586687330935586},"common_prefix":{"samples":65280,"sample_rate":16000,"reference_coverage":0.9975550122249389,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65280,"output_samples":65280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.9768265511555665,"spectral_convergence":0.508736398574136,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.363734899666922,"spectral_convergence":0.2929195364570726,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.664830788002208,"spectral_convergence":0.2895611897103046,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002924707813686955,"output_power":0.00023264190091671253,"delta_db":-0.9939454756509731,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009465320480587567,"output_power":0.0006791869561074401,"delta_db":-1.4414598644215335,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008114863396667601,"output_power":0.0005199583874291624,"delta_db":-1.9331262524573476,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.158258627318859e-05,"output_power":5.602279373576319e-05,"delta_db":-1.064426230070971,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8628940354482847e-05,"output_power":1.8928178091227626e-05,"delta_db":-1.7969646025068742,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.651263812210498e-06,"output_power":5.378287067349086e-06,"delta_db":-2.064355765469739,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5490463650962394,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.5473818177828178,"pitch":null,"output_rms_dbfs":-28.204147466576913,"output_peak":0.36586687330935586,"reference_rms_dbfs":-26.656765648794092,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.001797412941566},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65280,"output_samples":65280,"intelligibility":{"estoi":{"value":0.9044398013418194,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046468827826125485,"rms_dbfs":-26.656765648794092,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.462987207158204,"mean":-9.594384361715878e-05,"zero_energy":false},"output":{"rms":0.038885942232992826,"rms_dbfs":-28.204147466576913,"peak":0.36586687330935586,"peak_dbfs":-8.733538221322247,"crest_factor_db":19.470609245254668,"mean":0.0006851736901661495,"zero_energy":false},"rms_difference_db":-1.5473818177828207},"full_scale":{"reference":{"sample_count":65280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":408,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":6.060067140164652e-07,"output_mean_power":8.421814587939322e-07,"output_rms_dbfs":-60.74594324103265,"output_zero_energy":false,"output_peak":0.004144648566881956,"output_energy_fraction":0.00015288971888613142,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":401,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":103,"both_estimated_voiced_frames":81,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9310344827586207,"absolute_difference_cents_median":5.078790412546925,"absolute_difference_cents_p95":19.662006222410287,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016079151656388168,"log_rms_mae_db":2.345107392112114,"rms_envelope_nrmse":0.1952582561255604,"envelope_correlation":0.9878654320939213,"rms_gain_error_db":-1.5473813489922894,"reference_mean_power":0.0021620005113072694,"output_mean_power":0.001513971352306133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.6651551444095693e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005904947329463013,"output_power":0.00042855871100798496,"reference_active":true,"delta_db":-1.3920570105414654},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014190577217244028,"output_power":0.00011352336005051488,"reference_active":true,"delta_db":-0.9691482438509947},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.890466773101126e-05,"output_power":1.108341242633717e-05,"reference_active":true,"delta_db":-2.3189555446780856}],"active_band_count":3,"active_band_mae_db":1.5600535996901819}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00012458719833187296,"log_rms_mae_db":1.7756984013075372,"rms_envelope_nrmse":0.17706606952792178,"envelope_correlation":0.9946983396650311,"rms_gain_error_db":-1.5473903288478246,"reference_mean_power":0.002178031765009068,"output_mean_power":0.001525194310196599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":1.5433531623879135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005580428626596856,"output_power":0.00041046952043256803,"reference_active":true,"delta_db":-1.33386643947251},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.783926605175337e-05,"output_power":5.220758721897292e-05,"reference_active":true,"delta_db":-1.1374751797903973},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9269006141489923e-06,"output_power":1.2142786580234406e-06,"reference_active":true,"delta_db":-2.0054095316344602}],"active_band_count":3,"active_band_mae_db":1.4922503836324559}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00013740128966162354,"log_rms_mae_db":2.1266515637790686,"rms_envelope_nrmse":0.22049000662644097,"envelope_correlation":0.9871181692813388,"rms_gain_error_db":-1.7679211785393365,"reference_mean_power":0.001347514194398345,"output_mean_power":0.0008968942314723137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.2474285164183978e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042281465946871546,"output_power":0.00029654153226353906,"reference_active":true,"delta_db":-1.5406450910601834},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001066201531269166,"output_power":8.280071405457323e-05,"reference_active":true,"delta_db":-1.098052198611552},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1378448818963377e-05,"output_power":7.8084421378026e-06,"reference_active":true,"delta_db":-1.6351866393203296}],"active_band_count":3,"active_band_mae_db":1.4246279763306884}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00011336523032200158,"log_rms_mae_db":1.962088577696766,"rms_envelope_nrmse":0.20441160788678078,"envelope_correlation":0.9927821025710273,"rms_gain_error_db":-1.7679213544284855,"reference_mean_power":0.0013575079795100859,"output_mean_power":0.0009035459750637171},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":1.1694383736308564e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004095528046779407,"output_power":0.00028872955966419996,"reference_active":true,"delta_db":-1.5181865586961485},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4838241661745966e-05,"output_power":3.53225927960802e-05,"reference_active":true,"delta_db":-1.0359599859621715},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3398822953982881e-06,"output_power":9.690400260186717e-07,"reference_active":true,"delta_db":-1.4072493271118405}],"active_band_count":3,"active_band_mae_db":1.3204652905900536}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":2.100527688041068,"rms_envelope_nrmse":0.2431260664552969,"envelope_correlation":0.9662656027697263,"rms_gain_error_db":-1.3885905975322643,"reference_mean_power":7.761788133075413e-05,"output_mean_power":5.637709895511058e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":7.026387282627106e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8954844881996105e-05,"output_power":2.7792944375525914e-05,"reference_active":true,"delta_db":-1.4662692109850766},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.4081584948175165e-06,"output_power":7.465166704621668e-06,"reference_active":true,"delta_db":0.03329245055007688},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1568867421083e-06,"output_power":8.824500400156477e-07,"reference_active":true,"delta_db":-1.1760071696199512}],"active_band_count":3,"active_band_mae_db":0.8918562770517017}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":1.5621201095475823,"rms_envelope_nrmse":0.18292799402401982,"envelope_correlation":0.988708703878701,"rms_gain_error_db":-1.3885923753845186,"reference_mean_power":7.819348554807946e-05,"output_mean_power":5.6795161014365334e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":6.516175589334703e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8373890349877084e-05,"output_power":2.8184067310209062e-05,"reference_active":true,"delta_db":-1.3403216288780513},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.023907084252208e-06,"output_power":3.12264161068484e-06,"reference_active":true,"delta_db":0.13953700147434528},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.87815775629872e-08,"output_power":9.804200121199052e-08,"reference_active":true,"delta_db":-0.03263790262009362}],"active_band_count":3,"active_band_mae_db":0.5041655109908301}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":2.9178288046779537,"rms_envelope_nrmse":0.2855948008422622,"envelope_correlation":0.957298960991987,"rms_gain_error_db":-1.6349978376117904,"reference_mean_power":1.4211812763052752e-05,"output_mean_power":9.753257567635487e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":7.927827092057373e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8004002596497758e-06,"output_power":2.9850385943283365e-06,"reference_active":true,"delta_db":-1.0487938848502163},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0964768555021817e-06,"output_power":1.0049751920530701e-06,"reference_active":true,"delta_db":-0.3784412772538094},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4676605109073828e-07,"output_power":2.075997794826753e-07,"reference_active":true,"delta_db":1.5060127866204789}],"active_band_count":3,"active_band_mae_db":0.9777493162415016}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65280,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":2.5727115492006556,"rms_envelope_nrmse":0.2066837099085965,"envelope_correlation":0.9888342939022786,"rms_gain_error_db":-1.6349984537389772,"reference_mean_power":1.4317180563714502e-05,"output_mean_power":9.82556780039213e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.396017520370497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6819007828539625e-06,"output_power":2.9709517986753375e-06,"reference_active":true,"delta_db":-0.9317647549932301},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.085079625934018e-07,"output_power":4.865242681561623e-07,"reference_active":true,"delta_db":-0.19193249771762885},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3793360383351659e-08,"output_power":1.8697877144394646e-08,"reference_active":true,"delta_db":1.321222185929087}],"active_band_count":3,"active_band_mae_db":0.8149731462133153}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.6052236557006836,"predicted_mos":3.6052236557006836,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":98280,"input_sample_rate":24000,"input_seconds":4.095,"input_float32_sha256":"53c9f7d3bdc844f5f4a02c3080651054989fdb4e3b3e5eb3168aefc37131c0fb","internal_samples":65520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.239666262001265},"started_at":"2026-10-13T05:03:48.243012+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027598670003499137,"native_input_samples":98160,"native_input_sha256":"adf9b8f686f3858487380ec7e58f35e84ea45deab9ed3afe90f6e57d93b241ec","output_sha256":"09e6ca452aa5cc6a2af99326bcdf9c7baa6f8f27b93c8a5e93add8d2f84b3583","output_samples":98280,"length_delta_samples":120,"analysis_s":1.987361604000398,"synthesis_s":8.548409730999992,"total_s":10.53577133500039,"rtf":2.575983211491538,"output_peak":0.3777529001235962,"output_rms":0.04634920579519881,"finished_at":"2026-10-13T05:03:58.788872+00:00","analysis_stages":{"external_f0_s":1.3270155189993602},"evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"09e6ca452aa5cc6a2af99326bcdf9c7baa6f8f27b93c8a5e93add8d2f84b3583","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":65440,"output_samples":65520,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.674051812303258,"output_peak":0.3761948084500339},"common_prefix":{"samples":65440,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987789987789988,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.829460791138255,"spectral_convergence":0.7877141001595652,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.421219038375399,"spectral_convergence":0.3408668882129887,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.457201320630362,"spectral_convergence":0.2833885549881295,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.0004033791221481636,"delta_db":1.4068626819728314,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0009110543475803305,"delta_db":-0.15526460373080986,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0007384354339555781,"delta_db":-0.3990532235970032,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":6.438674572192813e-05,"delta_db":-0.4494593443250976,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":2.8858942894075423e-05,"delta_db":0.04538580678206009,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":7.2886009255812855e-06,"delta_db":-0.7335116526894542,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5315895521828761,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.001351650117853133,"pitch":null,"output_rms_dbfs":-26.668745899184756,"output_peak":0.3761948084500339,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.235960383381339},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65440,"output_samples":65440,"intelligibility":{"estoi":{"value":0.8922966832769194,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046412000424763004,"rms_dbfs":-26.667394249066906,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.473615807431017,"mean":-9.563744505343636e-05,"zero_energy":false},"output":{"rms":0.04640477860781088,"rms_dbfs":-26.668745899184756,"peak":0.3761948084500339,"peak_dbfs":-8.49174404158521,"crest_factor_db":18.177001857599546,"mean":-8.505972247358934e-05,"zero_energy":false},"rms_difference_db":-0.0013516501178507667},"full_scale":{"reference":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":409,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":6.057582671993074e-07,"output_mean_power":3.184538444965186e-07,"output_rms_dbfs":-64.96953503777706,"output_zero_energy":false,"output_peak":0.02120533450272205,"output_energy_fraction":4.085791697766047e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":402,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":105,"both_estimated_voiced_frames":80,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.9195402298850575,"absolute_difference_cents_median":6.644230760173402,"absolute_difference_cents_p95":25.438187461621848,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00016079151656388168,"log_rms_mae_db":3.9367368969875134,"rms_envelope_nrmse":0.17771872685054446,"envelope_correlation":0.9721756160366172,"rms_gain_error_db":-0.0013501688768312128,"reference_mean_power":0.0021567095381351967,"output_mean_power":0.00215603914750802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.6620838184249172e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000592954618765589,"output_power":0.000610847396380924,"reference_active":true,"delta_db":0.12911270811805414},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014246995030046006,"output_power":0.0001164946402385691,"reference_active":true,"delta_db":-0.8741732828369508},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9396326507323893e-05,"output_power":1.0436299051379507e-05,"reference_active":true,"delta_db":-2.6917297090181673}],"active_band_count":3,"active_band_mae_db":1.2316718999910574}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00012458719833187296,"log_rms_mae_db":3.3414769777162485,"rms_envelope_nrmse":0.10899279578984464,"envelope_correlation":0.9888942207809973,"rms_gain_error_db":-0.0013435790216913526,"reference_mean_power":0.0021726617564687806,"output_mean_power":0.0021719897029711947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.5403145620848265e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005607371759141369,"output_power":0.0005764369484241135,"reference_active":true,"delta_db":0.11992460385049875},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.84281540251884e-05,"output_power":5.145798012292677e-05,"reference_active":true,"delta_db":-1.2378208939199045},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9268319788025274e-06,"output_power":8.904938936520792e-07,"reference_active":true,"delta_db":-3.3521289976173034}],"active_band_count":3,"active_band_mae_db":1.5699581651292356}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00013740128966162354,"log_rms_mae_db":2.0859479057741948,"rms_envelope_nrmse":0.19282656363201925,"envelope_correlation":0.973300658041158,"rms_gain_error_db":-0.3693522908363357,"reference_mean_power":0.001344215530062727,"output_mean_power":0.0012346210562849977},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.2437342579426783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004237422220408671,"output_power":0.0003836369001099809,"reference_active":true,"delta_db":-0.4318136651769155},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010753440950326828,"output_power":9.564509516552235e-05,"reference_active":true,"delta_db":-0.5088475102106788},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.157936749643158e-05,"output_power":8.439830791599427e-06,"reference_active":true,"delta_db":-1.3735109779633454}],"active_band_count":3,"active_band_mae_db":0.7713907177836465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00011336523032200158,"log_rms_mae_db":1.7171288189889808,"rms_envelope_nrmse":0.11736510148601467,"envelope_correlation":0.9904984514286584,"rms_gain_error_db":-0.3693521139489292,"reference_mean_power":0.0013541602801659725,"output_mean_power":0.0012437550572673886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.1658618871138857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004110855381392752,"output_power":0.00036847700463013783,"reference_active":true,"delta_db":-0.475218084742301},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.54169136496196e-05,"output_power":4.00422562875276e-05,"reference_active":true,"delta_db":-0.546990770085824},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3416624074278085e-06,"output_power":8.395469056566654e-07,"reference_active":true,"delta_db":-2.0359828600290015}],"active_band_count":3,"active_band_mae_db":1.0193972382857088}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":2.2748625391889825,"rms_envelope_nrmse":0.24415436406452645,"envelope_correlation":0.9527554650088472,"rms_gain_error_db":-0.22289992894204236,"reference_mean_power":7.742789466080947e-05,"output_mean_power":7.355419683819876e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.015275941663668e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8994572701205544e-05,"output_power":3.63967121654033e-05,"reference_active":true,"delta_db":-0.2994201157262572},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.452200984753589e-06,"output_power":6.635441488660882e-06,"reference_active":true,"delta_db":-0.5041473532338604},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.182412963771586e-06,"output_power":1.123003205977286e-06,"reference_active":true,"delta_db":-0.22388186508729951}],"active_band_count":3,"active_band_mae_db":0.34248311134913906}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":1.3899085029216638,"rms_envelope_nrmse":0.12835593529023132,"envelope_correlation":0.9863869028178627,"rms_gain_error_db":-0.2228996335805766,"reference_mean_power":7.800066882867518e-05,"output_mean_power":7.409832029991578e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":6.50533884720925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.845568227980426e-05,"output_power":3.573510475056855e-05,"reference_active":true,"delta_db":-0.31865461512180593},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.058614755593347e-06,"output_power":2.7560711203181257e-06,"reference_active":true,"delta_db":-0.4523435901872211},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905753913869999e-08,"output_power":8.383850540438728e-08,"reference_active":true,"delta_db":-0.7244400720601456}],"active_band_count":3,"active_band_mae_db":0.4984794257897242}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":2.337717579994698,"rms_envelope_nrmse":0.27925997186349266,"envelope_correlation":0.9440948959982384,"rms_gain_error_db":-0.5676804035036943,"reference_mean_power":1.4177032937223138e-05,"output_mean_power":1.2439911995252262e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.931109844663439e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.831691262181725e-06,"output_power":3.420346650073816e-06,"reference_active":true,"delta_db":-0.49320384847225973},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0928531053800198e-06,"output_power":8.813190278526007e-07,"reference_active":true,"delta_db":-0.9342864394131439},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4875100105629565e-07,"output_power":1.2403087173082653e-07,"reference_active":true,"delta_db":-0.7893010102291725}],"active_band_count":3,"active_band_mae_db":0.7389304327048588}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":1.3059060072999098,"rms_envelope_nrmse":0.14963611127778842,"envelope_correlation":0.9852780409695001,"rms_gain_error_db":-0.5676809180044513,"reference_mean_power":1.428188481146513e-05,"output_mean_power":1.2531914817435832e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":7.397985811629504e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7144994114751925e-06,"output_power":3.307654279109119e-06,"reference_active":true,"delta_db":-0.5038018372064267},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.071953545728326e-07,"output_power":3.9265139789683017e-07,"reference_active":true,"delta_db":-1.1116811922396004},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3628863245634317e-08,"output_power":9.614606923110835e-09,"reference_active":true,"delta_db":-1.5152810022588226}],"active_band_count":3,"active_band_mae_db":1.043588010568283}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.317507266998291,"predicted_mos":3.317507266998291,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":98400,"input_sample_rate":24000,"input_seconds":4.1,"input_float32_sha256":"b48806e7b0cc475ce08a233a48192b824d9c51c30816a8ea13c62857ca72813a","internal_samples":65600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8694308680005634},"started_at":"2026-10-13T05:09:55.117563+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0034880330003943527,"native_input_samples":98160,"native_input_sha256":"adf9b8f686f3858487380ec7e58f35e84ea45deab9ed3afe90f6e57d93b241ec","output_sha256":"fa926ccb7a28df6ec3ff3871df147215b5c0e098e7884f969e709bf6a0d308d1","output_samples":98400,"length_delta_samples":240,"analysis_s":0.6892782180002541,"synthesis_s":0.09042991200112738,"total_s":0.7797081300013815,"rtf":0.1906376845969148,"output_peak":0.4517887532711029,"output_rms":0.04593927226817854,"finished_at":"2026-10-13T05:09:55.909019+00:00","analysis_stages":{"external_f0_s":0.06701270600024145},"evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"fa926ccb7a28df6ec3ff3871df147215b5c0e098e7884f969e709bf6a0d308d1","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":65440,"output_samples":65600,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.750645907352798,"output_peak":0.4522296987088557},"common_prefix":{"samples":65440,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975609756097561,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.187107379707174,"spectral_convergence":0.18207641718472647,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.458183236317979,"spectral_convergence":0.16469101068545952,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.949544455584893,"spectral_convergence":0.17340073119786675,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.00028078031438786275,"delta_db":-0.16660492333223242,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0009191140232469559,"delta_db":-0.11701352966187441,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0008229396192073874,"delta_db":0.07150120715705956,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":6.367238840990164e-05,"delta_db":-0.4979126872125594,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":2.444027011832327e-05,"delta_db":-0.676358366220448,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":7.394599996802667e-06,"delta_db":-0.6708065086437126,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3666995370379811,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07264652112631073,"pitch":null,"output_rms_dbfs":-26.740040770193218,"output_peak":0.4522296987088557,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.864945023870015},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65440,"output_samples":65440,"intelligibility":{"estoi":{"value":0.9455319165090025,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046412000424763004,"rms_dbfs":-26.667394249066906,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.473615807431017,"mean":-9.563744505343636e-05,"zero_energy":false},"output":{"rms":0.04602544132148102,"rms_dbfs":-26.740040770193218,"peak":0.4522296987088557,"peak_dbfs":-6.892818403641138,"crest_factor_db":19.847222366552074,"mean":-3.933961610503706e-06,"zero_energy":false},"rms_difference_db":-0.07264652112631254},"full_scale":{"reference":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":409,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":6.057582671993074e-07,"output_mean_power":5.690023298944251e-07,"output_rms_dbfs":-62.44885955295227,"output_zero_energy":false,"output_peak":0.006094777364089489,"output_energy_fraction":7.421184984764346e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":402,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":85,"both_estimated_voiced_frames":76,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.8735632183908046,"absolute_difference_cents_median":1.5505628265288753,"absolute_difference_cents_p95":4.618894195771725,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00016079151656388168,"log_rms_mae_db":1.145771005762262,"rms_envelope_nrmse":0.07964651794768766,"envelope_correlation":0.9944603321745314,"rms_gain_error_db":-0.0726466159520496,"reference_mean_power":0.0021567095381351967,"output_mean_power":0.002120933234407312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.6620838184249172e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000592954618765589,"output_power":0.0005976279716597511,"reference_active":true,"delta_db":0.03409459896739012},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014246995030046006,"output_power":0.00013881666032502618,"reference_active":true,"delta_db":-0.11281680940393764},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9396326507323893e-05,"output_power":2.403052690048341e-05,"reference_active":true,"delta_db":0.9304380719368983}],"active_band_count":3,"active_band_mae_db":0.3591164934360753}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00012458719833187296,"log_rms_mae_db":0.6642626443810066,"rms_envelope_nrmse":0.04643685623082774,"envelope_correlation":0.9980640174371617,"rms_gain_error_db":-0.07264690617899584,"reference_mean_power":0.0021726617564687806,"output_mean_power":0.0021366206885977926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.5403145620848265e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005607371759141369,"output_power":0.0005724129347914619,"reference_active":true,"delta_db":0.08950088899964918},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.84281540251884e-05,"output_power":7.140769487659454e-05,"reference_active":true,"delta_db":0.18510189505200406},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9268319788025274e-06,"output_power":1.92157343769569e-06,"reference_active":true,"delta_db":-0.01186858742641643}],"active_band_count":3,"active_band_mae_db":0.09549045715935656}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00013740128966162354,"log_rms_mae_db":1.0352506275481799,"rms_envelope_nrmse":0.0998502779909275,"envelope_correlation":0.9927836491253471,"rms_gain_error_db":0.004592529919915435,"reference_mean_power":0.001344215530062727,"output_mean_power":0.0013456377482838398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.2437342579426783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004237422220408671,"output_power":0.0004443771262631527,"reference_active":true,"delta_db":0.20649956464172406},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010753440950326828,"output_power":0.00010405935727167907,"reference_active":true,"delta_db":-0.14266315510488592},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.157936749643158e-05,"output_power":1.5524456764787264e-05,"reference_active":true,"delta_db":1.273315747561405}],"active_band_count":3,"active_band_mae_db":0.5408261557693383}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00011336523032200158,"log_rms_mae_db":0.6888306518550114,"rms_envelope_nrmse":0.06336767158321252,"envelope_correlation":0.9969567315719158,"rms_gain_error_db":0.004592548792126993,"reference_mean_power":0.0013541602801659725,"output_mean_power":0.0013555930261059457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.1658618871138857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004110855381392752,"output_power":0.0004373522216247876,"reference_active":true,"delta_db":0.2689913825435976},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.54169136496196e-05,"output_power":4.7025523840089386e-05,"reference_active":true,"delta_db":0.15116024081900573},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3416624074278085e-06,"output_power":1.4406456090303333e-06,"reference_active":true,"delta_db":0.30913908600081347}],"active_band_count":3,"active_band_mae_db":0.24309690312113896}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":1.2821940663781122,"rms_envelope_nrmse":0.14145454894116988,"envelope_correlation":0.9870587153847685,"rms_gain_error_db":-0.6060732828824857,"reference_mean_power":7.742789466080947e-05,"output_mean_power":6.734263758621701e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.015275941663668e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8994572701205544e-05,"output_power":3.3250342852271315e-05,"reference_active":true,"delta_db":-0.6920803796910541},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.452200984753589e-06,"output_power":5.740199779920232e-06,"reference_active":true,"delta_db":-1.133575515444439},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.182412963771586e-06,"output_power":1.0199155040058904e-06,"reference_active":true,"delta_db":-0.6420498894662128}],"active_band_count":3,"active_band_mae_db":0.8225685948672353}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":0.8522912295893227,"rms_envelope_nrmse":0.09719074648586531,"envelope_correlation":0.9955978073696135,"rms_gain_error_db":-0.6060733561976304,"reference_mean_power":7.800066882867518e-05,"output_mean_power":6.784080474806467e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":6.50533884720925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.845568227980426e-05,"output_power":3.264656872601982e-05,"reference_active":true,"delta_db":-0.7112297893322643},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.058614755593347e-06,"output_power":2.5990552690763563e-06,"reference_active":true,"delta_db":-0.7070926445359543},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905753913869999e-08,"output_power":6.965390096935209e-08,"reference_active":true,"delta_db":-1.5294209059292707}],"active_band_count":3,"active_band_mae_db":0.9825811132658298}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":1.7061504450391487,"rms_envelope_nrmse":0.1499393816443646,"envelope_correlation":0.9869040570343083,"rms_gain_error_db":-0.6517091590292858,"reference_mean_power":1.4177032937223138e-05,"output_mean_power":1.2201533944165319e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.931109844663439e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.831691262181725e-06,"output_power":3.480350763400242e-06,"reference_active":true,"delta_db":-0.4176749256464777},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0928531053800198e-06,"output_power":8.896374953475887e-07,"reference_active":true,"delta_db":-0.893487119349522},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4875100105629565e-07,"output_power":1.5137296962485058e-07,"reference_active":true,"delta_db":0.07588433869500413}],"active_band_count":3,"active_band_mae_db":0.46234879456366795}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":1.2057143887602404,"rms_envelope_nrmse":0.10004612788541897,"envelope_correlation":0.9961301649606734,"rms_gain_error_db":-0.6517054856111814,"reference_mean_power":1.428188481146513e-05,"output_mean_power":1.2291785628315786e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":7.397985811629504e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7144994114751925e-06,"output_power":3.3725710884398577e-06,"reference_active":true,"delta_db":-0.41939181197109787},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.071953545728326e-07,"output_power":4.163187096893916e-07,"reference_active":true,"delta_db":-0.8574933830734566},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3628863245634317e-08,"output_power":1.0348659791636357e-08,"reference_active":true,"delta_db":-1.1957552385998282}],"active_band_count":3,"active_band_mae_db":0.8242134778814609}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1089-134686-0032","method":"magphase","status":"measured","mos":{"status":"measured","value":3.922269105911255,"predicted_mos":3.922269105911255,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":197048,"input_sample_rate":48000,"input_seconds":4.105166666666666,"input_float32_sha256":"cfa4fc02013027f8cd02abed1e87fe70bf56e8439bde949fd29c9c208e4b6000","internal_samples":65683,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9240337510000245},"started_at":"2026-10-13T05:11:55.390477+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00657144300021173,"native_input_samples":196320,"native_input_sha256":"6f81bb8194635707ca0423bd552baf36787a24a460454f4bdf463a90df5c7f1f","output_sha256":"fa9f47a24b6b62da88264d2b0e0e107937cb417e47465d46ec3770d9e0b0ce65","output_samples":197048,"length_delta_samples":728,"analysis_s":1.884000363001178,"synthesis_s":0.7300062209997122,"total_s":2.61400658400089,"rtf":0.6391214141811468,"output_peak":0.4510455846253473,"output_rms":0.04491510782420744,"finished_at":"2026-10-13T05:11:58.019076+00:00","evaluation_input":{"reference":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"output_sha256":"fa9f47a24b6b62da88264d2b0e0e107937cb417e47465d46ec3770d9e0b0ce65","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":65440,"output_samples":65683,"length_delta_samples":243,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.946577582411784,"output_peak":0.4467756956075592},"common_prefix":{"samples":65440,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9963004125877319,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65440,"output_samples":65440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.678537403118345,"spectral_convergence":0.7272723192487297,"spectral_convergence_status":"ok","amplitude_floor":1.239800719019687e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.939062489365911,"spectral_convergence":0.3156518814600465,"spectral_convergence_status":"ok","amplitude_floor":7.959069981155994e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.657571580609852,"spectral_convergence":0.22084176102234396,"spectral_convergence_status":"ok","amplitude_floor":5.70263552352201e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002917609391844852,"output_power":0.0002735050696062516,"delta_db":-0.2806176761723886,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009442146735758444,"output_power":0.0008815528814751878,"delta_db":-0.29822375981789256,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008095018592590047,"output_power":0.0007651822396367139,"delta_db":-0.2445296935260019,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.140726670679603e-05,"output_power":6.91588564393818e-05,"delta_db":-0.13894606193301728,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.8558923829027427e-05,"output_power":2.984531586945249e-05,"delta_db":0.19134341509333366,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.629692823024492e-06,"output_power":8.213182278484247e-06,"delta_db":-0.21483875716494366,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2280832272845963,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26308805926040146,"pitch":null,"output_rms_dbfs":-26.930482308327306,"output_peak":0.4467756956075592,"reference_rms_dbfs":-26.667394249066906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.425057157698036},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65440,"output_samples":65440,"intelligibility":{"estoi":{"value":0.8594801155067072,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046412000424763004,"rms_dbfs":-26.667394249066906,"peak":0.43682861328125,"peak_dbfs":-7.19377844163589,"crest_factor_db":19.473615807431017,"mean":-9.563744505343636e-05,"zero_energy":false},"output":{"rms":0.04502729779268361,"rms_dbfs":-26.930482308327306,"peak":0.4467756956075592,"peak_dbfs":-6.998209206679985,"crest_factor_db":19.932273101647322,"mean":-0.00016171124009488666,"zero_energy":false},"rms_difference_db":-0.2630880592604008},"full_scale":{"reference":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":409,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.891882228665054e-06,"threshold_dbfs":-57.231059023067665,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":113,"quiet_samples":18080,"reference_mean_power":6.057582671993074e-07,"output_mean_power":4.754897318540149e-07,"output_rms_dbfs":-63.22858857164612,"output_zero_energy":false,"output_peak":0.0053769507205721455,"output_energy_fraction":6.479544981903004e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":402,"reference_estimated_voiced_frames":87,"output_estimated_voiced_frames":93,"both_estimated_voiced_frames":80,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9195402298850575,"absolute_difference_cents_median":4.891968832823381,"absolute_difference_cents_p95":20.360419260013323,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00016079151656388168,"log_rms_mae_db":2.487424979791474,"rms_envelope_nrmse":0.21322576768342882,"envelope_correlation":0.9598698116026919,"rms_gain_error_db":-0.2630866173478953,"reference_mean_power":0.0021567095381351967,"output_mean_power":0.0020299390457908903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.6620838184249172e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000592954618765589,"output_power":0.0005607102214151175,"reference_active":true,"delta_db":-0.24282983207097084},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014246995030046006,"output_power":0.0001365531869837095,"reference_active":true,"delta_db":-0.1842143230746934},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9396326507323893e-05,"output_power":1.6546162318579927e-05,"reference_active":true,"delta_db":-0.6902220572442962}],"active_band_count":3,"active_band_mae_db":0.37242207079665346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00012458719833187296,"log_rms_mae_db":1.58924364806594,"rms_envelope_nrmse":0.13211538526889077,"envelope_correlation":0.9841424214438812,"rms_gain_error_db":-0.26308204309948513,"reference_mean_power":0.0021726617564687806,"output_mean_power":0.0020449557532138036},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.5403145620848265e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005607371759141369,"output_power":0.0005305182028392353,"reference_active":true,"delta_db":-0.24059060274552516},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.84281540251884e-05,"output_power":6.528021447899445e-05,"reference_active":true,"delta_db":-0.20453251636760597},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9268319788025274e-06,"output_power":1.7471041651682687e-06,"reference_active":true,"delta_db":-0.425250464087829}],"active_band_count":3,"active_band_mae_db":0.2901245277336534}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":0.00013740128966162354,"log_rms_mae_db":2.106352577560653,"rms_envelope_nrmse":0.22485098231294373,"envelope_correlation":0.9630523055660986,"rms_gain_error_db":-0.2715446875175984,"reference_mean_power":0.001344215530062727,"output_mean_power":0.0012627414594848319},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":1.2437342579426783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004237422220408671,"output_power":0.0003990779455222532,"reference_active":true,"delta_db":-0.2604401185678581},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010753440950326828,"output_power":0.00010033242773387585,"reference_active":true,"delta_db":-0.3010613368077187},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.157936749643158e-05,"output_power":1.0418535269696318e-05,"reference_active":true,"delta_db":-0.4587817115398075}],"active_band_count":3,"active_band_mae_db":0.3400943889717947}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":0.00011336523032200158,"log_rms_mae_db":1.6476446391079063,"rms_envelope_nrmse":0.1404475233902196,"envelope_correlation":0.9853205526633539,"rms_gain_error_db":-0.27154486090004604,"reference_mean_power":0.0013541602801659725,"output_mean_power":0.0012720833988633053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":1.1658618871138857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004110855381392752,"output_power":0.0003872349372687136,"reference_active":true,"delta_db":-0.2595766520674137},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.54169136496196e-05,"output_power":4.245163378981197e-05,"reference_active":true,"delta_db":-0.2932320886355603},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3416624074278085e-06,"output_power":1.1860807422305092e-06,"reference_active":true,"delta_db":-0.5352899672365173}],"active_band_count":3,"active_band_mae_db":0.3626995693131638}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":4.166087007180048e-05,"log_rms_mae_db":2.389841192713579,"rms_envelope_nrmse":0.29488639818387197,"envelope_correlation":0.9322135954590091,"rms_gain_error_db":-0.0501240216170162,"reference_mean_power":7.742789466080947e-05,"output_mean_power":7.653939911038548e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.015275941663668e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8994572701205544e-05,"output_power":3.781127319159734e-05,"reference_active":true,"delta_db":-0.1338286445396059},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.452200984753589e-06,"output_power":7.575097423976969e-06,"reference_active":true,"delta_db":0.07103663468980227},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.182412963771586e-06,"output_power":1.2296316682649152e-06,"reference_active":true,"delta_db":0.1700585696611992}],"active_band_count":3,"active_band_mae_db":0.12497461629686912}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":3.3087684127894394e-05,"log_rms_mae_db":1.4639034801751443,"rms_envelope_nrmse":0.16225300345989865,"envelope_correlation":0.977970924586562,"rms_gain_error_db":-0.05012398794676111,"reference_mean_power":7.800066882867518e-05,"output_mean_power":7.710560121509204e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":6.50533884720925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.845568227980426e-05,"output_power":3.73762842284339e-05,"reference_active":true,"delta_db":-0.12364397294943612},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.058614755593347e-06,"output_power":3.1834873795699204e-06,"reference_active":true,"delta_db":0.1737835315906944},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905753913869999e-08,"output_power":9.444876312867884e-08,"reference_active":true,"delta_db":-0.20691259801842335}],"active_band_count":3,"active_band_mae_db":0.16811336751951797}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":817,"rms_floor":1.922446117673939e-05,"log_rms_mae_db":2.4538371955178384,"rms_envelope_nrmse":0.30761914257745,"envelope_correlation":0.933991867941457,"rms_gain_error_db":0.09482070221796861,"reference_mean_power":1.4177032937223138e-05,"output_mean_power":1.448996773992681e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":817,"frequency_resolution_hz":0.24479804161566707,"power_floor":7.931109844663439e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.831691262181725e-06,"output_power":3.35917381381842e-06,"reference_active":true,"delta_db":-0.5715803248447724},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0928531053800198e-06,"output_power":9.072981297043233e-07,"reference_active":true,"delta_db":-0.8081177504093671},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4875100105629565e-07,"output_power":2.1540633537064003e-07,"reference_active":true,"delta_db":1.607985751940086}],"active_band_count":3,"active_band_mae_db":0.9958946090647419}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65440,"uncovered_tail_samples":0,"metrics":{"frame_count":811,"rms_floor":1.4099069206239128e-05,"log_rms_mae_db":1.4611861536709776,"rms_envelope_nrmse":0.16697887142253354,"envelope_correlation":0.9793283675168052,"rms_gain_error_db":0.09481895267079024,"reference_mean_power":1.428188481146513e-05,"output_mean_power":1.459712816728553e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":811,"frequency_resolution_hz":0.24660912453760792,"power_floor":7.397985811629504e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7144994114751925e-06,"output_power":3.2600578574194285e-06,"reference_active":true,"delta_db":-0.5667498615843921},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.071953545728326e-07,"output_power":4.4900535756300905e-07,"reference_active":true,"delta_db":-0.529237441013368},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3628863245634317e-08,"output_power":1.5706384294905024e-08,"reference_active":true,"delta_db":0.6161658554705398}],"active_band_count":3,"active_band_mae_db":0.5707177193561}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"hnm","status":"measured","mos":{"status":"measured","value":2.604464292526245,"predicted_mos":2.604464292526245,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7c444d2fc6c66128529232fe6feeec192f1cf9df2d92d6899847e17cb774d18b","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9147132839989354},"started_at":"2026-10-13T04:37:59.139134+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4924999959475826e-05,"native_input_samples":96000,"native_input_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","output_sha256":"e533c64baefe7c3f28b4f34cbe692908f77dce2c8cbc3bd42bd8f5527702523c","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1155772369993429,"synthesis_s":0.513908549000007,"total_s":1.62948578599935,"rtf":0.271580964333225,"output_peak":0.24227797964759476,"output_rms":0.03014417682739575,"finished_at":"2026-10-13T04:38:00.773355+00:00","evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"e533c64baefe7c3f28b4f34cbe692908f77dce2c8cbc3bd42bd8f5527702523c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.0409526054611185,"spectral_convergence":0.32043179931760607,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.656071790185499,"spectral_convergence":0.2815251513247391,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.828295536132608,"spectral_convergence":0.30540239027720334,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.0002566172616976438,"delta_db":0.15252485134311525,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.0004405402444811658,"delta_db":-0.10713965954287989,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00013726987381618074,"delta_db":-0.4912680662082907,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.1407442127176516e-05,"delta_db":-0.2702210804865631,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":1.0398908620526029e-05,"delta_db":0.7935314037138751,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.2437665858610012e-05,"delta_db":-0.42848664518512813,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3738619510799753,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1031278972296274,"pitch":null,"output_rms_dbfs":-30.415931425689543,"output_peak":0.24227797964759476,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.841773310593076},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.0409526054611185,"spectral_convergence":0.32043179931760607,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.656071790185499,"spectral_convergence":0.2815251513247391,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.828295536132608,"spectral_convergence":0.30540239027720334,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.0002566172616976438,"delta_db":0.15252485134311525,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.0004405402444811658,"delta_db":-0.10713965954287989,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00013726987381618074,"delta_db":-0.4912680662082907,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.1407442127176516e-05,"delta_db":-0.2702210804865631,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":1.0398908620526029e-05,"delta_db":0.7935314037138751,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.2437665858610012e-05,"delta_db":-0.42848664518512813,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3738619510799753,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1031278972296274,"pitch":null,"output_rms_dbfs":-30.415931425689543,"output_peak":0.24227797964759476,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.841773310593076},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7204590496563736,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030504212914161243,"rms_dbfs":-30.312803528459913,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.593205152815347,"mean":-3.0800501505533854e-06,"zero_energy":false},"output":{"rms":0.03014417682739575,"rms_dbfs":-30.415931425689543,"peak":0.24227797964759476,"peak_dbfs":-12.313721131374942,"crest_factor_db":18.1022102943146,"mean":-0.00020784571183687063,"zero_energy":false},"rms_difference_db":-0.1031278972296299},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":2.268053437482614e-06,"output_rms_dbfs":-56.44346717270125,"output_zero_energy":false,"output_peak":0.006321289134525291,"output_energy_fraction":4.576019432625071e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":148,"both_estimated_voiced_frames":132,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9777777777777777,"absolute_difference_cents_median":1.2942915471765817,"absolute_difference_cents_p95":9.35070755327971,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":1.5242717907123151,"rms_envelope_nrmse":0.09835176618256401,"envelope_correlation":0.99154641281201,"rms_gain_error_db":-0.10313213756913799,"reference_mean_power":0.0009312818437003785,"output_mean_power":0.0009094271642185212},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.494393820063344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023666657525761142,"output_power":0.00023348250702447774,"reference_active":true,"delta_db":-0.05882578348030889},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5441929621447868e-05,"output_power":1.4865875921471657e-05,"reference_active":true,"delta_db":-0.16511065142068876},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.985747798516416e-06,"output_power":5.43996179551919e-06,"reference_active":true,"delta_db":-1.0861705287154055}],"active_band_count":3,"active_band_mae_db":0.4367023212054677}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":0.7663147809892163,"rms_envelope_nrmse":0.05044570774115426,"envelope_correlation":0.9977460018647478,"rms_gain_error_db":-0.10313334514699446,"reference_mean_power":0.0009359486205868301,"output_mean_power":0.0009139841702683143},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.189856446091556e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021037225251172397,"output_power":0.00020915012317904613,"reference_active":true,"delta_db":-0.025303323853230267},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.644252267790675e-06,"output_power":7.324542907386679e-06,"reference_active":true,"delta_db":-0.18554483304489208},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434290091658501e-07,"output_power":3.090861793147678e-07,"reference_active":true,"delta_db":-0.4575739070710258}],"active_band_count":3,"active_band_mae_db":0.2228073546563827}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":1.8594841885226179,"rms_envelope_nrmse":0.12891385635191369,"envelope_correlation":0.9893971860660897,"rms_gain_error_db":-0.2771439859334703,"reference_mean_power":0.00038327408961782034,"output_mean_power":0.0003595796142290196},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.028222657916465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015396875485141256,"output_power":0.00014233811262601202,"reference_active":true,"delta_db":-0.3411139476530455},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.75189127488202e-06,"output_power":6.958297970981137e-06,"reference_active":true,"delta_db":-0.9959889142459282},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1315883357233815e-06,"output_power":2.96952732965482e-06,"reference_active":true,"delta_db":-1.4342971605265946}],"active_band_count":3,"active_band_mae_db":0.9238000074751894}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":1.3013298880070951,"rms_envelope_nrmse":0.08900908421937202,"envelope_correlation":0.9951876762114152,"rms_gain_error_db":-0.2771498192832122,"reference_mean_power":0.00038520090394369153,"output_mean_power":0.0003613868251172741},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9052955697898775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058163319658512,"output_power":0.0001314468919741678,"reference_active":true,"delta_db":-0.29178262604038496},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.901478078483037e-06,"output_power":3.0116726057247154e-06,"reference_active":true,"delta_db":-1.1242141244403212},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3393802456709043e-07,"output_power":1.6913648464173857e-07,"reference_active":true,"delta_db":-1.4086351838580005}],"active_band_count":3,"active_band_mae_db":0.9415439781129024}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":1.7808563188168913,"rms_envelope_nrmse":0.20939755915668345,"envelope_correlation":0.9688454899362472,"rms_gain_error_db":-0.14979932306353994,"reference_mean_power":4.0846100044459596e-05,"output_mean_power":3.946123421695983e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0558567323423285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3562756220605465e-05,"output_power":1.224894396722578e-05,"reference_active":true,"delta_db":-0.4424930776998217},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5216123885214778e-06,"output_power":1.0758484177234834e-06,"reference_active":true,"delta_db":-1.5055295003402727},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602773162915517e-07,"output_power":5.404919407027831e-07,"reference_active":true,"delta_db":-0.8693715486547575}],"active_band_count":3,"active_band_mae_db":0.9391313755649507}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":1.1625338188105592,"rms_envelope_nrmse":0.12918098106434445,"envelope_correlation":0.9877819906657738,"rms_gain_error_db":-0.1498384449243977,"reference_mean_power":4.105094002568104e-05,"output_mean_power":3.9658771951473676e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8992373996258787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2457002317192888e-05,"output_power":1.1368087338983155e-05,"reference_active":true,"delta_db":-0.39726143694755844},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11943579398167e-07,"output_power":5.652119196160038e-07,"reference_active":true,"delta_db":-1.573145398889465},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742155092447495e-08,"output_power":3.0013285950953627e-08,"reference_active":true,"delta_db":-0.9580823675316976}],"active_band_count":3,"active_band_mae_db":0.9761630677895736}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":1.5970988664523877,"rms_envelope_nrmse":0.2546119796804402,"envelope_correlation":0.9611816460405975,"rms_gain_error_db":-0.22017478330515294,"reference_mean_power":2.620464364014994e-05,"output_mean_power":2.4909257358594625e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.347390510898902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9544955639169026e-05,"output_power":1.8894153463727287e-05,"reference_active":true,"delta_db":-0.1470725056037048},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7959648252038243e-06,"output_power":3.207288778693435e-06,"reference_active":true,"delta_db":0.596063613660731},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.147434020032559e-06,"output_power":5.414206036735118e-07,"reference_active":true,"delta_db":-3.2619294328882242}],"active_band_count":3,"active_band_mae_db":1.3350218507175533}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":1.6339956525330308,"rms_envelope_nrmse":0.16010363948070444,"envelope_correlation":0.9844857278246187,"rms_gain_error_db":-0.22017662912186176,"reference_mean_power":2.6336417520647037e-05,"output_mean_power":2.5034506559803938e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.2809655135480134e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9754827323518684e-05,"output_power":1.9148768420517148e-05,"reference_active":true,"delta_db":-0.13532390922649914},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4198245590922995e-06,"output_power":1.5864385557871993e-06,"reference_active":true,"delta_db":0.4818857212749802},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517276499876375e-08,"output_power":3.254761020771837e-08,"reference_active":true,"delta_db":-3.0154704040849185}],"active_band_count":3,"active_band_mae_db":1.2108933448621326}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.266627311706543,"predicted_mos":3.266627311706543,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ef0165e498631339ad7cc5ee78500ebdd26fb3140b68df7c39ef146f991698eb","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.875115258999358},"started_at":"2026-10-13T04:40:49.771359+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.195799970650114e-05,"native_input_samples":96000,"native_input_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","output_sha256":"3318f4180a2a97cdd8f7b6feb8b58f99a9cd8acb69de42d49dccced7a95af027","output_samples":96000,"length_delta_samples":0,"analysis_s":2.130751964999945,"synthesis_s":0.08015317399986088,"total_s":2.210905138999806,"rtf":0.36848418983330095,"output_peak":0.37958597542218114,"output_rms":0.03259321774458405,"finished_at":"2026-10-13T04:40:51.987935+00:00","evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"3318f4180a2a97cdd8f7b6feb8b58f99a9cd8acb69de42d49dccced7a95af027","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.859816553967868,"spectral_convergence":0.7674836640222197,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.570275255452129,"spectral_convergence":0.3222337254548995,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.721838171514785,"spectral_convergence":0.3210658051311424,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.0002516569285577983,"delta_db":0.0677551051183488,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.0005537261585857173,"delta_db":0.8859548607929867,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00017509910578440848,"delta_db":0.565818871031028,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.840341328653968e-05,"delta_db":0.4077547648276631,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":9.443696133311827e-06,"delta_db":0.3750738295365274,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.3988540598092863e-05,"delta_db":-0.13822513347835327,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.40676376079748455,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.575348288995471,"pitch":null,"output_rms_dbfs":-29.737455239464445,"output_peak":0.37958597542218114,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.0506433269782605},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.859816553967868,"spectral_convergence":0.7674836640222197,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.570275255452129,"spectral_convergence":0.3222337254548995,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.721838171514785,"spectral_convergence":0.3210658051311424,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.0002516569285577983,"delta_db":0.0677551051183488,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.0005537261585857173,"delta_db":0.8859548607929867,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00017509910578440848,"delta_db":0.565818871031028,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.840341328653968e-05,"delta_db":0.4077547648276631,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":9.443696133311827e-06,"delta_db":0.3750738295365274,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.3988540598092863e-05,"delta_db":-0.13822513347835327,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.40676376079748455,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.575348288995471,"pitch":null,"output_rms_dbfs":-29.737455239464445,"output_peak":0.37958597542218114,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.0506433269782605},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7460459931535802,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030504212914161243,"rms_dbfs":-30.312803528459913,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.593205152815347,"mean":-3.0800501505533854e-06,"zero_energy":false},"output":{"rms":0.03259321774458405,"rms_dbfs":-29.737455239464442,"peak":0.37958597542218114,"peak_dbfs":-8.413796836869336,"crest_factor_db":21.323658402595107,"mean":-9.209807028785997e-11,"zero_energy":false},"rms_difference_db":0.5753482889954711},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":8.594010294089362e-07,"output_rms_dbfs":-60.65804130444584,"output_zero_energy":false,"output_peak":0.003070416503466189,"output_energy_fraction":1.4831423235321365e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":138,"both_estimated_voiced_frames":115,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.8518518518518519,"absolute_difference_cents_median":6.3035861291053825,"absolute_difference_cents_p95":23.594955715459342,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":3.38317718455808,"rms_envelope_nrmse":0.15593132193000064,"envelope_correlation":0.9847025334214283,"rms_gain_error_db":0.5753531255963332,"reference_mean_power":0.0009312818437003785,"output_mean_power":0.0010632036247827262},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.494393820063344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023666657525761142,"output_power":0.0002888365963126042,"reference_active":true,"delta_db":0.8651529231162578},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5441929621447868e-05,"output_power":1.456389826073889e-05,"reference_active":true,"delta_db":-0.2542393235630122},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.985747798516416e-06,"output_power":3.829007270830175e-06,"reference_active":true,"delta_db":-2.6112671153776654}],"active_band_count":3,"active_band_mae_db":1.2435531206856452}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":2.9306442079013135,"rms_envelope_nrmse":0.11122772335122097,"envelope_correlation":0.9941744989457856,"rms_gain_error_db":0.5754142121415633,"reference_mean_power":0.0009359486205868301,"output_mean_power":0.0010685465089638546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.189856446091556e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021037225251172397,"output_power":0.0002613364268107877,"reference_active":true,"delta_db":0.9421149172246595},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.644252267790675e-06,"output_power":7.18331510668163e-06,"reference_active":true,"delta_db":-0.2701009294420429},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434290091658501e-07,"output_power":2.0653057952630638e-07,"reference_active":true,"delta_db":-2.2085261319760865}],"active_band_count":3,"active_band_mae_db":1.140247326214263}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":2.325898887378467,"rms_envelope_nrmse":0.20792630529151085,"envelope_correlation":0.9849647239591389,"rms_gain_error_db":1.079813578037387,"reference_mean_power":0.00038327408961782034,"output_mean_power":0.0004914629900354133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.028222657916465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015396875485141256,"output_power":0.00021211637731787685,"reference_active":true,"delta_db":1.3914160366469401},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.75189127488202e-06,"output_power":1.0288758119343817e-05,"reference_active":true,"delta_db":0.7026104365488279},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1315883357233815e-06,"output_power":1.7089004907627408e-06,"reference_active":true,"delta_db":-3.8340026815727275}],"active_band_count":3,"active_band_mae_db":1.9760097182561651}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":1.7158566789609813,"rms_envelope_nrmse":0.18003370634471394,"envelope_correlation":0.9909919544954179,"rms_gain_error_db":1.079802990795288,"reference_mean_power":0.00038520090394369153,"output_mean_power":0.0004939324928032331},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9052955697898775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058163319658512,"output_power":0.00019616595323289772,"reference_active":true,"delta_db":1.446950485884015},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.901478078483037e-06,"output_power":4.578243262525617e-06,"reference_active":true,"delta_db":0.6946969390177964},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3393802456709043e-07,"output_power":1.0401028734301042e-07,"reference_active":true,"delta_db":-3.520245221155436}],"active_band_count":3,"active_band_mae_db":1.8872975486857488}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":1.5710054080716858,"rms_envelope_nrmse":0.16278129714029949,"envelope_correlation":0.9833762145292287,"rms_gain_error_db":0.36214832045020073,"reference_mean_power":4.0846100044459596e-05,"output_mean_power":4.439820670132616e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0558567323423285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3562756220605465e-05,"output_power":1.4400864826126093e-05,"reference_active":true,"delta_db":0.26040618158600526},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5216123885214778e-06,"output_power":1.5819201433418468e-06,"reference_active":true,"delta_db":0.16880520658666226},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602773162915517e-07,"output_power":3.099551317096498e-07,"reference_active":true,"delta_db":-3.284275463643275}],"active_band_count":3,"active_band_mae_db":1.2378289506053142}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":0.8610278701878787,"rms_envelope_nrmse":0.08511024479709228,"envelope_correlation":0.996165962921663,"rms_gain_error_db":0.36215991867756747,"reference_mean_power":4.105094002568104e-05,"output_mean_power":4.462097938244286e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8992373996258787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2457002317192888e-05,"output_power":1.330174089610456e-05,"reference_active":true,"delta_db":0.2849493874780211},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11943579398167e-07,"output_power":8.251077392000821e-07,"reference_active":true,"delta_db":0.06984808760857386},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742155092447495e-08,"output_power":1.7437276653267018e-08,"reference_active":true,"delta_db":-3.3164312477778832}],"active_band_count":3,"active_band_mae_db":1.2237429076214927}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":1.4743219264060157,"rms_envelope_nrmse":0.3301049939235482,"envelope_correlation":0.9351434798845342,"rms_gain_error_db":-0.07176929856825319,"reference_mean_power":2.620464364014994e-05,"output_mean_power":2.577515752777221e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.347390510898902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9544955639169026e-05,"output_power":1.697828061574297e-05,"reference_active":true,"delta_db":-0.6114098176739287},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7959648252038243e-06,"output_power":1.3473890410664768e-06,"reference_active":true,"delta_db":-3.1703869272665997},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.147434020032559e-06,"output_power":5.344444715008945e-07,"reference_active":true,"delta_db":-3.3182513312323}],"active_band_count":3,"active_band_mae_db":2.3666826920576094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":1.1105033257111825,"rms_envelope_nrmse":0.22091359638260324,"envelope_correlation":0.9701693982756304,"rms_gain_error_db":-0.07176824465537181,"reference_mean_power":2.6336417520647037e-05,"output_mean_power":2.5904777960955565e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.2809655135480134e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9754827323518684e-05,"output_power":1.656620370556123e-05,"reference_active":true,"delta_db":-0.7645024053835339},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4198245590922995e-06,"output_power":6.673473530138585e-07,"reference_active":true,"delta_db":-3.2788274180435706},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517276499876375e-08,"output_power":3.3466460815662034e-08,"reference_active":true,"delta_db":-2.8945635985673808}],"active_band_count":3,"active_band_mae_db":2.3126311406648283}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"facodec","status":"measured","mos":{"status":"measured","value":3.2081079483032227,"predicted_mos":3.2081079483032227,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ff5d9fcb0dc475ef8d358b31558e219e50be39ae984517a690a153cb4e59177c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.372247904000687},"started_at":"2026-10-13T04:44:52.600734+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4770999516476877e-05,"native_input_samples":96000,"native_input_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","output_sha256":"65b6504c2524c746d8db147849d5e72668345d97ac5c9893d1929f6de5814baa","output_samples":96000,"length_delta_samples":0,"analysis_s":2.4405202899997676,"synthesis_s":7.588701540000329,"total_s":10.029221830000097,"rtf":1.6715369716666828,"output_peak":0.24584409594535828,"output_rms":0.030598307311399224,"finished_at":"2026-10-13T04:45:02.643851+00:00","evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"65b6504c2524c746d8db147849d5e72668345d97ac5c9893d1929f6de5814baa","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.433778228771606,"spectral_convergence":0.22656162528878812,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.76048985865528,"spectral_convergence":0.2327944746924367,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.788462596519022,"spectral_convergence":0.2467434753480722,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.00028845501825576687,"delta_db":0.6604471750998174,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.0004424101390167671,"delta_db":-0.08874482993980876,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00013601976364039273,"delta_db":-0.5310002523915844,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.006836247163542e-05,"delta_db":-0.4129891770122585,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":7.17463781344422e-06,"delta_db":-0.8183463941005997,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.212848912482094e-05,"delta_db":-0.4887457829391896,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5000456019138765,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.026751570737285804,"pitch":null,"output_rms_dbfs":-30.28605195772263,"output_peak":0.24584409594535828,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.66091022798197},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.433778228771606,"spectral_convergence":0.22656162528878812,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.76048985865528,"spectral_convergence":0.2327944746924367,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.788462596519022,"spectral_convergence":0.2467434753480722,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.00028845501825576687,"delta_db":0.6604471750998174,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.0004424101390167671,"delta_db":-0.08874482993980876,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00013601976364039273,"delta_db":-0.5310002523915844,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.006836247163542e-05,"delta_db":-0.4129891770122585,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":7.17463781344422e-06,"delta_db":-0.8183463941005997,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.212848912482094e-05,"delta_db":-0.4887457829391896,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5000456019138765,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.026751570737285804,"pitch":null,"output_rms_dbfs":-30.28605195772263,"output_peak":0.24584409594535828,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.66091022798197},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7409032315752091,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030504212914161243,"rms_dbfs":-30.312803528459913,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.593205152815347,"mean":-3.0800501505533854e-06,"zero_energy":false},"output":{"rms":0.030598307311399228,"rms_dbfs":-30.28605195772263,"peak":0.24584409594535828,"peak_dbfs":-12.186804340451012,"crest_factor_db":18.099247617271615,"mean":8.809932414992734e-05,"zero_energy":false},"rms_difference_db":0.026751570737282293},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":9.82840897099392e-07,"output_rms_dbfs":-60.0751678034107,"output_zero_energy":false,"output_peak":0.002897478174418211,"output_energy_fraction":1.9245528875943866e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":136,"both_estimated_voiced_frames":128,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9481481481481482,"absolute_difference_cents_median":2.6605803168085274,"absolute_difference_cents_p95":11.587345887185975,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":1.2821085905883094,"rms_envelope_nrmse":0.09718922868440692,"envelope_correlation":0.9917643016111495,"rms_gain_error_db":0.0267458688345893,"reference_mean_power":0.0009312818437003785,"output_mean_power":0.0009370348058263432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.494393820063344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023666657525761142,"output_power":0.0002597049814419678,"reference_active":true,"delta_db":0.40343353720287584},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5441929621447868e-05,"output_power":1.69353085365461e-05,"reference_active":true,"delta_db":0.4009154449206175},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.985747798516416e-06,"output_power":8.867251524955994e-06,"reference_active":true,"delta_db":1.0357612510538263}],"active_band_count":3,"active_band_mae_db":0.6133700777257732}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":0.9119379149418787,"rms_envelope_nrmse":0.08309381425574723,"envelope_correlation":0.9936257480370811,"rms_gain_error_db":0.026763277398396172,"reference_mean_power":0.0009359486205868301,"output_mean_power":0.0009417341864749658},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.189856446091556e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021037225251172397,"output_power":0.0002312718969770832,"reference_active":true,"delta_db":0.41134405568348736},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.644252267790675e-06,"output_power":8.473216041662153e-06,"reference_active":true,"delta_db":0.44713268996656735},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434290091658501e-07,"output_power":4.742292307870366e-07,"reference_active":true,"delta_db":1.4015134286359316}],"active_band_count":3,"active_band_mae_db":0.7533300580953287}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":1.5655478488077468,"rms_envelope_nrmse":0.10448296251600026,"envelope_correlation":0.9928654641252017,"rms_gain_error_db":-0.16150737405036267,"reference_mean_power":0.00038327408961782034,"output_mean_power":0.00036928249665617313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.028222657916465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015396875485141256,"output_power":0.00016005077870887878,"reference_active":true,"delta_db":0.16825193856529055},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.75189127488202e-06,"output_power":9.5651690621368e-06,"reference_active":true,"delta_db":0.3859073672100733},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1315883357233815e-06,"output_power":3.482832275027735e-06,"reference_active":true,"delta_db":-0.7418448229049628}],"active_band_count":3,"active_band_mae_db":0.4320013762267756}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":1.1037556851398624,"rms_envelope_nrmse":0.08804854923091868,"envelope_correlation":0.9948217176654449,"rms_gain_error_db":-0.1615073186359886,"reference_mean_power":0.00038520090394369153,"output_mean_power":0.0003711389764943785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9052955697898775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058163319658512,"output_power":0.00014736582669894682,"reference_active":true,"delta_db":0.20468200372769418},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.901478078483037e-06,"output_power":4.105289234525838e-06,"reference_active":true,"delta_db":0.2211458931078481},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3393802456709043e-07,"output_power":2.00362422660837e-07,"reference_active":true,"delta_db":-0.6728454401811947}],"active_band_count":3,"active_band_mae_db":0.36622444567224566}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":1.784588679009756,"rms_envelope_nrmse":0.1680554064037672,"envelope_correlation":0.9802542950985034,"rms_gain_error_db":-0.2617917994146208,"reference_mean_power":4.0846100044459596e-05,"output_mean_power":3.845664708643949e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0558567323423285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3562756220605465e-05,"output_power":1.082716823922304e-05,"reference_active":true,"delta_db":-0.97833070570463},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5216123885214778e-06,"output_power":1.5584925119732595e-06,"reference_active":true,"delta_db":0.10400684474575789},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602773162915517e-07,"output_power":7.738569737563496e-07,"reference_active":true,"delta_db":0.6893432325181449}],"active_band_count":3,"active_band_mae_db":0.5905602609895109}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":1.375689556799748,"rms_envelope_nrmse":0.13684260069366858,"envelope_correlation":0.98644277648136,"rms_gain_error_db":-0.26177568176649774,"reference_mean_power":4.105094002568104e-05,"output_mean_power":3.864964758618391e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8992373996258787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2457002317192888e-05,"output_power":9.885441779457529e-06,"reference_active":true,"delta_db":-1.0041746257118669},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11943579398167e-07,"output_power":8.38387461166412e-07,"reference_active":true,"delta_db":0.1391892253399434},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742155092447495e-08,"output_power":4.088670586366004e-08,"reference_active":true,"delta_db":0.3846033924256753}],"active_band_count":3,"active_band_mae_db":0.5093224144924952}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":1.5134213194589747,"rms_envelope_nrmse":0.22618409586363275,"envelope_correlation":0.97060124337761,"rms_gain_error_db":-0.5975996535459007,"reference_mean_power":2.620464364014994e-05,"output_mean_power":2.2835908422127212e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.347390510898902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9544955639169026e-05,"output_power":1.572276499176612e-05,"reference_active":true,"delta_db":-0.9450576598974638},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7959648252038243e-06,"output_power":3.2676378579662527e-06,"reference_active":true,"delta_db":0.6770221559353045},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.147434020032559e-06,"output_power":2.780585936262859e-06,"reference_active":true,"delta_db":3.844085997993678}],"active_band_count":3,"active_band_mae_db":1.822055271275482}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":1.4274745752043834,"rms_envelope_nrmse":0.16554767093767123,"envelope_correlation":0.9850644471554318,"rms_gain_error_db":-0.597598697779521,"reference_mean_power":2.6336417520647037e-05,"output_mean_power":2.2950747176068394e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.2809655135480134e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9754827323518684e-05,"output_power":1.731706165252611e-05,"reference_active":true,"delta_db":-0.5719903474176837},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4198245590922995e-06,"output_power":1.9195049943991216e-06,"reference_active":true,"delta_db":1.309545624414371},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517276499876375e-08,"output_power":1.8444111630954404e-07,"reference_active":true,"delta_db":4.517915954503954}],"active_band_count":3,"active_band_mae_db":2.133150642112003}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.7993268966674805,"predicted_mos":3.7993268966674805,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"df33169ae4d6b115fa9d7f98a955a7cce207e237aefd7bf64d26dc8783df831e","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.848605047000092},"started_at":"2026-10-13T04:57:50.517782+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00598721000005753,"native_input_samples":264600,"native_input_sha256":"44b28ca36f2ad744bb0c95e13ab6423dca8b4e790f06af2cec1c57790fb6657e","output_sha256":"4c6513fda47012f3319770465eb29ae27965906c73fdb8b0653d2c0687ac9ffc","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13572184700024081,"synthesis_s":6.400656975999482,"total_s":6.536378822999723,"rtf":1.0893964704999537,"output_peak":0.26132601499557495,"output_rms":0.027437286965100623,"finished_at":"2026-10-13T04:57:57.070565+00:00","evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"4c6513fda47012f3319770465eb29ae27965906c73fdb8b0653d2c0687ac9ffc","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.229396721366665,"output_peak":0.2605787513081038},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.047090917890799,"spectral_convergence":0.5297327173122175,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.92941388876902,"spectral_convergence":0.3908678002608113,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.314885100235037,"spectral_convergence":0.4513898855462866,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.00022140614886164762,"delta_db":-0.48843703398892435,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.00032871994405024514,"delta_db":-1.3787350009459856,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00014033375857984241,"delta_db":-0.3953988394335731,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":3.831002913131577e-05,"delta_db":-0.6078802936733942,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":1.1588173110308408e-05,"delta_db":1.26380352774811,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":1.3400369267368796e-05,"delta_db":-2.6670957457704825,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.133558406926745,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9148746888169007,"pitch":null,"output_rms_dbfs":-31.227678217276814,"output_peak":0.2605787513081038,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.76379663563162},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.5809321890604897,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030504212914161243,"rms_dbfs":-30.312803528459913,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.593205152815347,"mean":-3.0800501505533854e-06,"zero_energy":false},"output":{"rms":0.02745466122538627,"rms_dbfs":-31.227678217276814,"peak":0.2605787513081038,"peak_dbfs":-11.681220027716627,"crest_factor_db":19.546458189560187,"mean":-0.00012943944143962565,"zero_energy":false},"rms_difference_db":-0.9148746888169015},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":1.2023818843286664e-06,"output_rms_dbfs":-59.19957575664132,"output_zero_energy":false,"output_peak":0.0037220039084698655,"output_energy_fraction":2.9245003712996763e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":105,"both_estimated_voiced_frames":90,"reference_voiced_output_unvoiced_frames":45,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.6666666666666666,"absolute_difference_cents_median":10.522986245711214,"absolute_difference_cents_p95":53.70582658657783,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":2.638313659281909,"rms_envelope_nrmse":0.1724138001005565,"envelope_correlation":0.9822798581760706,"rms_gain_error_db":-0.9148717696061142,"reference_mean_power":0.0009312818437003785,"output_mean_power":0.0007543865887716069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.494393820063344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023666657525761142,"output_power":0.00021194884727222165,"reference_active":true,"delta_db":-0.4790586734943691},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5441929621447868e-05,"output_power":1.2938947964115477e-05,"reference_active":true,"delta_db":-0.7680260248085725},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.985747798516416e-06,"output_power":4.928296132194436e-06,"reference_active":true,"delta_db":-1.5151610666773763}],"active_band_count":3,"active_band_mae_db":0.9207485883267726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":2.348555006046477,"rms_envelope_nrmse":0.13795247103455843,"envelope_correlation":0.9920538244054885,"rms_gain_error_db":-0.9148330165365187,"reference_mean_power":0.0009359486205868301,"output_mean_power":0.0007581736853876299},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.189856446091556e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021037225251172397,"output_power":0.00018954081662812263,"reference_active":true,"delta_db":-0.4528570964879572},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.644252267790675e-06,"output_power":6.405098498183724e-06,"reference_active":true,"delta_db":-0.7680919800966817},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434290091658501e-07,"output_power":2.2343569756163274e-07,"reference_active":true,"delta_db":-1.8668441679636263}],"active_band_count":3,"active_band_mae_db":1.0292644148494217}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":3.2001218503135975,"rms_envelope_nrmse":0.1828396906422266,"envelope_correlation":0.9791777231580613,"rms_gain_error_db":-0.5737242470640875,"reference_mean_power":0.00038327408961782034,"output_mean_power":0.00033584356864539184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.028222657916465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015396875485141256,"output_power":0.00012872612026585932,"reference_active":true,"delta_db":-0.777659175853624},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.75189127488202e-06,"output_power":8.018551504865393e-06,"reference_active":true,"delta_db":-0.3800599064937116},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1315883357233815e-06,"output_power":3.082009587030142e-06,"reference_active":true,"delta_db":-1.272830573369318}],"active_band_count":3,"active_band_mae_db":0.8101832185722179}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":2.7091663054554314,"rms_envelope_nrmse":0.13134863189326115,"envelope_correlation":0.9903223288084182,"rms_gain_error_db":-0.5737294999693061,"reference_mean_power":0.00038520090394369153,"output_mean_power":0.0003375315296734732},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9052955697898775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058163319658512,"output_power":0.00011794330817570358,"reference_active":true,"delta_db":-0.7625527931440754},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.901478078483037e-06,"output_power":3.6141888083503257e-06,"reference_active":true,"delta_db":-0.3321833431956779},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3393802456709043e-07,"output_power":1.5131106683034094e-07,"reference_active":true,"delta_db":-1.8923012498476885}],"active_band_count":3,"active_band_mae_db":0.9956791287291473}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":3.4403838784415823,"rms_envelope_nrmse":0.22642918783490565,"envelope_correlation":0.965536699522326,"rms_gain_error_db":-0.5066768120610096,"reference_mean_power":4.0846100044459596e-05,"output_mean_power":3.634820054430909e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0558567323423285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3562756220605465e-05,"output_power":1.2527744592110637e-05,"reference_active":true,"delta_db":-0.34475065060290977},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5216123885214778e-06,"output_power":1.7245535250001436e-06,"reference_active":true,"delta_db":0.5437264261264203},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602773162915517e-07,"output_power":5.18716306964931e-07,"reference_active":true,"delta_db":-1.047964762293487}],"active_band_count":3,"active_band_mae_db":0.6454806130076056}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":2.74351259537231,"rms_envelope_nrmse":0.17064496172572666,"envelope_correlation":0.9807593210121621,"rms_gain_error_db":-0.5066551227248074,"reference_mean_power":4.105094002568104e-05,"output_mean_power":3.653066635239799e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8992373996258787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2457002317192888e-05,"output_power":1.1702205132351048e-05,"reference_active":true,"delta_db":-0.2714583848127637},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11943579398167e-07,"output_power":8.734340193116097e-07,"reference_active":true,"delta_db":0.31704251323506516},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742155092447495e-08,"output_power":3.311845649017946e-08,"reference_active":true,"delta_db":-0.5305169476257044}],"active_band_count":3,"active_band_mae_db":0.37300594855784447}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":2.3674212871120526,"rms_envelope_nrmse":0.25244459895276156,"envelope_correlation":0.9745455911396568,"rms_gain_error_db":-1.5807534271399628,"reference_mean_power":2.620464364014994e-05,"output_mean_power":1.8209705215140094e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.347390510898902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9544955639169026e-05,"output_power":1.4006876006627561e-05,"reference_active":true,"delta_db":-1.4469340490606588},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7959648252038243e-06,"output_power":2.2264927637892643e-06,"reference_active":true,"delta_db":-0.9891041545016322},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.147434020032559e-06,"output_power":4.170246203658921e-07,"reference_active":true,"delta_db":-4.395660263666767}],"active_band_count":3,"active_band_mae_db":2.2772328224096863}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":2.5172148750663794,"rms_envelope_nrmse":0.20798972353792583,"envelope_correlation":0.9886708349564342,"rms_gain_error_db":-1.5807511439516508,"reference_mean_power":2.6336417520647037e-05,"output_mean_power":1.830128500067662e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.2809655135480134e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9754827323518684e-05,"output_power":1.4032275074735648e-05,"reference_active":true,"delta_db":-1.4854514829459173},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4198245590922995e-06,"output_power":1.0371858092899755e-06,"reference_active":true,"delta_db":-1.3637811786452227},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517276499876375e-08,"output_power":2.8762503002154158e-08,"reference_active":true,"delta_db":-3.5523946961723802}],"active_band_count":3,"active_band_mae_db":2.133875785921173}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"golf","status":"measured","mos":{"status":"measured","value":2.944533109664917,"predicted_mos":2.944533109664917,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"a5193d09f723cd72545c13f77f3cb53bf758a1a3a3a54ace957dd1def67d2c24","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.012367835999612},"started_at":"2026-10-13T05:01:50.534761+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003461498000433494,"native_input_samples":144000,"native_input_sha256":"eee21f139648c85f2fe768624da8c14234ba6ddad4061a166c985d98d9bb42e2","output_sha256":"84721ad452ee8b4c9f2695397294fd421f6d0e028a56aa539e8fe41cc1c10c45","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.9206093729999338,"synthesis_s":0.1179860459997144,"total_s":1.0385954189996482,"rtf":0.17309923649994138,"output_peak":0.23662066459655762,"output_rms":0.026595320110743365,"finished_at":"2026-10-13T05:01:51.585590+00:00","analysis_stages":{"external_f0_s":0.17278193699985422},"evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"84721ad452ee8b4c9f2695397294fd421f6d0e028a56aa539e8fe41cc1c10c45","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.50151726477111,"output_peak":0.23684640027965712},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.865661479942618,"spectral_convergence":0.7242731685229842,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.60630218707523,"spectral_convergence":0.35663208221138964,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.702132071198827,"spectral_convergence":0.33782322159215583,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024817153938628444,"output_power":0.00021546636050731734,"delta_db":-0.6137249864759402,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004522964236989201,"output_power":0.0003368645059717842,"delta_db":-1.2796790023205284,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015396646098763943,"output_power":9.726944567259747e-05,"delta_db":-1.994496864785754,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.41391293608921e-05,"output_power":3.142078697243569e-05,"delta_db":-1.4760670422069007,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.67679204881589e-06,"output_power":6.413213555557762e-06,"delta_db":-1.3128348725283754,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.4805647390146256e-05,"output_power":2.0264184752287207e-05,"delta_db":-0.8782142961096291,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2591695107378547,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1959372967357709,"pitch":null,"output_rms_dbfs":-31.50151726477111,"output_peak":0.23684640027965712,"reference_rms_dbfs":-30.305579968035342,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.058031912738893},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.7681777440492947,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030529592084937825,"rms_dbfs":-30.305579968035342,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.585981592390773,"mean":-1.3972363607314275e-06,"zero_energy":false},"output":{"rms":0.02660260320780618,"rms_dbfs":-31.50151726477111,"peak":0.23684640027965712,"peak_dbfs":-12.510664230589732,"crest_factor_db":18.99085303418138,"mean":0.00029985264620987486,"zero_energy":false},"rms_difference_db":-1.1959372967357673},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":2.6074231896149997e-06,"output_rms_dbfs":-55.8378847631186,"output_zero_energy":false,"output_peak":0.004866153586872906,"output_energy_fraction":6.765955123724576e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":169,"both_estimated_voiced_frames":134,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9925925925925926,"absolute_difference_cents_median":4.214988001333023,"absolute_difference_cents_p95":20.36977055807385,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":2.6713221157373694,"rms_envelope_nrmse":0.1837639233986137,"envelope_correlation":0.9827262650967299,"rms_gain_error_db":-1.195948276957201,"reference_mean_power":0.0009328334953588469,"output_mean_power":0.0007082870546213162},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.484029159522767e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023670907478302118,"output_power":0.00017867681138807458,"reference_active":true,"delta_db":-1.2214671433714805},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.547857456639801e-05,"output_power":1.5471204486981785e-05,"reference_active":true,"delta_db":-0.002068373332575976},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969277237195987e-06,"output_power":4.880627796652901e-06,"reference_active":true,"delta_db":-1.54712052028844}],"active_band_count":3,"active_band_mae_db":0.9235520123308322}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":1.8966980416988568,"rms_envelope_nrmse":0.15476876077194235,"envelope_correlation":0.9921293885107051,"rms_gain_error_db":-1.1961176459063256,"reference_mean_power":0.0009375169286102652,"output_mean_power":0.0007118153577465333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.179397605913546e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021041358625737243,"output_power":0.0001588011133509855,"reference_active":true,"delta_db":-1.2222023554605277},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.654310811512285e-06,"output_power":8.235704684313101e-06,"reference_active":true,"delta_db":0.3179467206435992},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4219662660255285e-07,"output_power":3.106005826401286e-07,"reference_active":true,"delta_db":-0.420734578572181}],"active_band_count":3,"active_band_mae_db":0.6536278848921027}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":2.1037699600810473,"rms_envelope_nrmse":0.21873105028703016,"envelope_correlation":0.9834516295393341,"rms_gain_error_db":-1.5597810847660178,"reference_mean_power":0.00038391431857858903,"output_mean_power":0.0002680749301588},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.0202843451312154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015366385105766142,"output_power":0.00010654880593730562,"reference_active":true,"delta_db":-1.5902312603028157},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.743767030941421e-06,"output_power":7.145180160992839e-06,"reference_active":true,"delta_db":-0.8768539403395487},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.111716860577336e-06,"output_power":1.5214607151641864e-06,"reference_active":true,"delta_db":-4.317624576518958}],"active_band_count":3,"active_band_mae_db":2.261569925720441}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":1.4544421440890367,"rms_envelope_nrmse":0.19817495986801237,"envelope_correlation":0.989887772369254,"rms_gain_error_db":-1.559782352860887,"reference_mean_power":0.0003858475625119587,"output_mean_power":0.000269424773092724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.897458721767486e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014028177815589418,"output_power":9.784762588666169e-05,"reference_active":true,"delta_db":-1.5645096951475022},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.894859506264557e-06,"output_power":3.0423640403720176e-06,"reference_active":true,"delta_db":-1.0728061737081658},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.326623909487298e-07,"output_power":1.225736985139093e-07,"reference_active":true,"delta_db":-2.7832889618661936}],"active_band_count":3,"active_band_mae_db":1.8068682769072872}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":1.9109814999221477,"rms_envelope_nrmse":0.23160851421146147,"envelope_correlation":0.9730319103367158,"rms_gain_error_db":-1.3538426005577544,"reference_mean_power":4.091423123090152e-05,"output_mean_power":2.9956435472511288e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.0461406153644057e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3551495322644091e-05,"output_power":9.277529849939392e-06,"reference_active":true,"delta_db":-1.645548591893491},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5196075364423507e-06,"output_power":1.035063408161331e-06,"reference_active":true,"delta_db":-1.6676448319477946},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.593848152649842e-07,"output_power":4.582726221320794e-07,"reference_active":true,"delta_db":-1.5801502907080205}],"active_band_count":3,"active_band_mae_db":1.6311145715164355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":1.3943127907012818,"rms_envelope_nrmse":0.19106798235887024,"envelope_correlation":0.9864853009203456,"rms_gain_error_db":-1.3538704888403776,"reference_mean_power":4.1119757272939015e-05,"output_mean_power":3.010672346342552e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.8895484842339103e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2448866542459897e-05,"output_power":8.4630834596949e-06,"reference_active":true,"delta_db":-1.6760118738705634},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105964915700575e-07,"output_power":5.073405076731793e-07,"reference_active":true,"delta_db":-2.0350518054416105},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.710823462936071e-08,"output_power":3.400450693851318e-08,"reference_active":true,"delta_db":-0.37933811900612485}],"active_band_count":3,"active_band_mae_db":1.3634672661060996}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":1.7929925556297401,"rms_envelope_nrmse":0.21311245651115476,"envelope_correlation":0.9752686581147176,"rms_gain_error_db":-0.5961264247050928,"reference_mean_power":2.6248422096547663e-05,"output_mean_power":2.2881819677688073e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.34990873232847e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9565836306043246e-05,"output_power":1.7926075967716195e-05,"reference_active":true,"delta_db":-0.3801318320180763},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7936130852538458e-06,"output_power":4.220853998875318e-06,"reference_active":true,"delta_db":1.7923407372348699},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1457442320205715e-06,"output_power":1.5837541091401563e-06,"reference_active":true,"delta_db":1.4060007512563855}],"active_band_count":3,"active_band_mae_db":1.1928244401697772}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":1.8214012287595427,"rms_envelope_nrmse":0.14284528537940194,"envelope_correlation":0.9904620723035749,"rms_gain_error_db":-0.5961262938077154,"reference_mean_power":2.638063654213553e-05,"output_mean_power":2.2997077093366606e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.2834865217205475e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9772312362858262e-05,"output_power":1.9218332523815172e-05,"reference_active":true,"delta_db":-0.12341759362763235},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4168981378060552e-06,"output_power":2.1766154997057157e-06,"reference_active":true,"delta_db":1.8644308794922329},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.487613129635469e-08,"output_power":1.0902884396805888e-07,"reference_active":true,"delta_db":2.2545646323087287}],"active_band_count":3,"active_band_mae_db":1.4141377018095314}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.8070383071899414,"predicted_mos":2.8070383071899414,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"3451b0f600ab5b9f6741f9f0bdbdf3a46a603e4632b0ac15181920a94573c4a7","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7522654899985355},"started_at":"2026-10-13T05:03:58.789525+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030069750000620843,"native_input_samples":144000,"native_input_sha256":"eee21f139648c85f2fe768624da8c14234ba6ddad4061a166c985d98d9bb42e2","output_sha256":"3d2c8ca80d55dd9fd85717c755be1a29a73d494883f57c0ca460de2f5ead95c1","output_samples":144120,"length_delta_samples":120,"analysis_s":2.459540176999326,"synthesis_s":3.5347629430007146,"total_s":5.9943031200000405,"rtf":0.9990505200000067,"output_peak":0.26455286145210266,"output_rms":0.03099180368273374,"finished_at":"2026-10-13T05:04:04.794391+00:00","analysis_stages":{"external_f0_s":1.692593394999676},"evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"3d2c8ca80d55dd9fd85717c755be1a29a73d494883f57c0ca460de2f5ead95c1","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.173935967663784,"output_peak":0.2648390577654736},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.623743393838959,"spectral_convergence":0.6950616834540357,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.585665104943907,"spectral_convergence":0.37332662874019873,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.722633656426087,"spectral_convergence":0.36398921523072997,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.0003644061156584177,"delta_db":1.675522756477517,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.00039396499818473323,"delta_db":-0.5924191577758476,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00013639559281015463,"delta_db":-0.5190170338059839,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.1720764858509746e-05,"delta_db":-0.23748249930156645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":7.235462613540076e-06,"delta_db":-0.7816831771281703,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":1.7818818192059475e-05,"delta_db":-1.4294744405755542,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8725998441774397,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.14248499878494042,"pitch":null,"output_rms_dbfs":-30.170318529674972,"output_peak":0.2648390577654736,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.977347385069652},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7317227488560707,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030504212914161243,"rms_dbfs":-30.312803528459913,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.593205152815347,"mean":-3.0800501505533854e-06,"zero_energy":false},"output":{"rms":0.031008736709472944,"rms_dbfs":-30.17031852967498,"peak":0.2648390577654736,"peak_dbfs":-11.54035931809377,"crest_factor_db":18.629959211581205,"mean":0.00011559623169462552,"zero_energy":false},"rms_difference_db":0.14248499878493348},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":4.430081787307045e-07,"output_rms_dbfs":-63.535882558427865,"output_zero_energy":false,"output_peak":0.0023661958487413036,"output_energy_fraction":8.446660366529587e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":162,"both_estimated_voiced_frames":125,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9259259259259259,"absolute_difference_cents_median":9.234967732114871,"absolute_difference_cents_p95":28.306663282761367,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":4.759995449740294,"rms_envelope_nrmse":0.17152732447840893,"envelope_correlation":0.9756007961260929,"rms_gain_error_db":0.1424897750757037,"reference_mean_power":0.0009312818437003785,"output_mean_power":0.0009623434916729605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.494393820063344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023666657525761142,"output_power":0.0002453067826275979,"reference_active":true,"delta_db":0.15572630169573806},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5441929621447868e-05,"output_power":1.3370391836107912e-05,"reference_active":true,"delta_db":-0.6255743377760787},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.985747798516416e-06,"output_power":5.817972454829263e-06,"reference_active":true,"delta_db":-0.7944124180972784}],"active_band_count":3,"active_band_mae_db":0.5252376858563651}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":4.448937985784424,"rms_envelope_nrmse":0.11585188336167485,"envelope_correlation":0.9886377170248396,"rms_gain_error_db":0.14256146960077037,"reference_mean_power":0.0009359486205868301,"output_mean_power":0.000967181888971129},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.189856446091556e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021037225251172397,"output_power":0.0002184589773689428,"reference_active":true,"delta_db":0.163814392610519},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.644252267790675e-06,"output_power":6.489469912453865e-06,"reference_active":true,"delta_db":-0.7112578753564186},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434290091658501e-07,"output_power":3.1120001003915793e-07,"reference_active":true,"delta_db":-0.4279737454413677}],"active_band_count":3,"active_band_mae_db":0.43434867113610176}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":2.691839215524344,"rms_envelope_nrmse":0.20358625689882232,"envelope_correlation":0.975585544892615,"rms_gain_error_db":-0.8064116486635486,"reference_mean_power":0.00038327408961782034,"output_mean_power":0.0003183232028566454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.028222657916465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015396875485141256,"output_power":0.00012486994095676612,"reference_active":true,"delta_db":-0.9097469127258474},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.75189127488202e-06,"output_power":6.6891276557137465e-06,"reference_active":true,"delta_db":-1.1673242963034225},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1315883357233815e-06,"output_power":3.2026996749368406e-06,"reference_active":true,"delta_db":-1.1060082695000149}],"active_band_count":3,"active_band_mae_db":1.061026492843095}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":2.1845360875469684,"rms_envelope_nrmse":0.14071351002210367,"envelope_correlation":0.9909886419664392,"rms_gain_error_db":-0.8064074927641589,"reference_mean_power":0.00038520090394369153,"output_mean_power":0.00031992379903977847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9052955697898775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058163319658512,"output_power":0.00011391938451198866,"reference_active":true,"delta_db":-0.9133095451723222},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.901478078483037e-06,"output_power":2.9400769184379287e-06,"reference_active":true,"delta_db":-1.2287047846207555},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3393802456709043e-07,"output_power":1.7384879949669376e-07,"reference_active":true,"delta_db":-1.2892912225967779}],"active_band_count":3,"active_band_mae_db":1.1437685174632852}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":2.267270442121303,"rms_envelope_nrmse":0.22811698090877136,"envelope_correlation":0.9632009552483738,"rms_gain_error_db":-0.24360713574192583,"reference_mean_power":4.0846100044459596e-05,"output_mean_power":3.8618009332025335e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0558567323423285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3562756220605465e-05,"output_power":1.2776422321335883e-05,"reference_active":true,"delta_db":-0.25938696878303713},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5216123885214778e-06,"output_power":1.5645450662200276e-06,"reference_active":true,"delta_db":0.12084041870035273},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602773162915517e-07,"output_power":5.899094647880694e-07,"reference_active":true,"delta_db":-0.4894101316817051}],"active_band_count":3,"active_band_mae_db":0.28987917305503164}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":1.6588670159786822,"rms_envelope_nrmse":0.12053086710869558,"envelope_correlation":0.9896646171179683,"rms_gain_error_db":-0.24358186169298923,"reference_mean_power":4.105094002568104e-05,"output_mean_power":3.881190148117145e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8992373996258787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2457002317192888e-05,"output_power":1.1740675014984878e-05,"reference_active":true,"delta_db":-0.25720478402707964},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11943579398167e-07,"output_power":8.831903817345749e-07,"reference_active":true,"delta_db":0.3652847889334314},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742155092447495e-08,"output_power":3.162067228513482e-08,"reference_active":true,"delta_db":-0.731506835338171}],"active_band_count":3,"active_band_mae_db":0.45133213609956063}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":1.507095997887691,"rms_envelope_nrmse":0.28582684137054887,"envelope_correlation":0.9564270307663671,"rms_gain_error_db":-1.2318726159803945,"reference_mean_power":2.620464364014994e-05,"output_mean_power":1.973290370015987e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.347390510898902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9544955639169026e-05,"output_power":1.3645677665654632e-05,"reference_active":true,"delta_db":-1.5603958077567457},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7959648252038243e-06,"output_power":1.1794222883343724e-06,"reference_active":true,"delta_db":-3.7486237276774976},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.147434020032559e-06,"output_power":5.283861578545559e-07,"reference_active":true,"delta_db":-3.3677629017599204}],"active_band_count":3,"active_band_mae_db":2.892260812398055}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":1.1753656816483233,"rms_envelope_nrmse":0.20894031135453717,"envelope_correlation":0.98186852023838,"rms_gain_error_db":-1.2318736159459096,"reference_mean_power":2.6336417520647037e-05,"output_mean_power":1.9832128924128896e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.2809655135480134e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9754827323518684e-05,"output_power":1.3277667499931978e-05,"reference_active":true,"delta_db":-1.725514490550407},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4198245590922995e-06,"output_power":5.741731354380634e-07,"reference_active":true,"delta_db":-3.931918152604009},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517276499876375e-08,"output_power":2.979593762849235e-08,"reference_active":true,"delta_db":-3.3990909012516104}],"active_band_count":3,"active_band_mae_db":3.0188411814686753}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.6067087650299072,"predicted_mos":2.6067087650299072,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"c6a77c7d8cd9a0914e2afe8ad8aec766342dc34221cbb98df8e31e1cd27a9cd2","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.361086624001473},"started_at":"2026-10-13T05:09:55.910968+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026835669996216893,"native_input_samples":144000,"native_input_sha256":"eee21f139648c85f2fe768624da8c14234ba6ddad4061a166c985d98d9bb42e2","output_sha256":"20b7e2d7b142097908f0c30e7f490255f15d641db5ed7cb5c1fd226f4ea7c659","output_samples":144240,"length_delta_samples":240,"analysis_s":1.3611780169994745,"synthesis_s":0.11672563899992383,"total_s":1.4779036559993983,"rtf":0.24631727599989972,"output_peak":0.24263814091682434,"output_rms":0.029780286978414623,"finished_at":"2026-10-13T05:09:57.403879+00:00","analysis_stages":{"external_f0_s":0.23833266400106368},"evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"20b7e2d7b142097908f0c30e7f490255f15d641db5ed7cb5c1fd226f4ea7c659","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.519808683927675,"output_peak":0.24281964456250776},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.111845617494029,"spectral_convergence":0.2059571129919577,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.698845411513951,"spectral_convergence":0.19058525226035974,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.057318501963783,"spectral_convergence":0.20788758763334625,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.00024331019062039956,"delta_db":-0.07873082302674013,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.00043115026943836977,"delta_db":-0.2007089258934054,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00014316901259221808,"delta_db":-0.3085301135442091,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.0428937409079424e-05,"delta_db":-0.3740818463992723,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":7.380838971478174e-06,"delta_db":-0.6952889114463392,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.323438248759373e-05,"delta_db":-0.2769520625030419,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.322382113802168,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19977353061611594,"pitch":null,"output_rms_dbfs":-30.51257705907603,"output_peak":0.24281964456250776,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.9560031769905875},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8280652526352233,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030504212914161243,"rms_dbfs":-30.312803528459913,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.593205152815347,"mean":-3.0800501505533854e-06,"zero_energy":false},"output":{"rms":0.02981062950558305,"rms_dbfs":-30.51257705907603,"peak":0.24281964456250776,"peak_dbfs":-12.29432361882365,"crest_factor_db":18.21825344025238,"mean":4.3593735202926636e-05,"zero_energy":false},"rms_difference_db":-0.1997735306161168},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":2.06179790678889e-06,"output_rms_dbfs":-56.85753905622673,"output_zero_energy":false,"output_peak":0.0073380807967929245,"output_energy_fraction":4.25348822677604e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":138,"both_estimated_voiced_frames":129,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9555555555555556,"absolute_difference_cents_median":1.248227181560786,"absolute_difference_cents_p95":7.092066080959494,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":1.5552547808646704,"rms_envelope_nrmse":0.08826514024511191,"envelope_correlation":0.9935021557645466,"rms_gain_error_db":-0.19977858647781738,"reference_mean_power":0.0009312818437003785,"output_mean_power":0.000889412599416639},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.494393820063344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023666657525761142,"output_power":0.00022934942944933132,"reference_active":true,"delta_db":-0.13639262180884493},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5441929621447868e-05,"output_power":1.3258261499693687e-05,"reference_active":true,"delta_db":-0.6621498818535516},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.985747798516416e-06,"output_power":3.036112704543508e-06,"reference_active":true,"delta_db":-3.6189501344816573}],"active_band_count":3,"active_band_mae_db":1.472497546048018}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":0.8797263463271533,"rms_envelope_nrmse":0.0462512744488906,"envelope_correlation":0.9984970061995929,"rms_gain_error_db":-0.19977131230327483,"reference_mean_power":0.0009359486205868301,"output_mean_power":0.0008938710611442054},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.189856446091556e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021037225251172397,"output_power":0.00020603915570200694,"reference_active":true,"delta_db":-0.0903869550017458},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.644252267790675e-06,"output_power":7.083099423352863e-06,"reference_active":true,"delta_db":-0.3311167285831459},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434290091658501e-07,"output_power":1.6171435007834506e-07,"reference_active":true,"delta_db":-3.270884171799845}],"active_band_count":3,"active_band_mae_db":1.2307959517949123}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":1.2716659111338184,"rms_envelope_nrmse":0.11677152352614735,"envelope_correlation":0.9912390581809817,"rms_gain_error_db":-0.24497954181819162,"reference_mean_power":0.00038327408961782034,"output_mean_power":0.0003622525961150463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.028222657916465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015396875485141256,"output_power":0.00014550950501545086,"reference_active":true,"delta_db":-0.24541234276514307},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.75189127488202e-06,"output_power":6.854118656753406e-06,"reference_active":true,"delta_db":-1.061502952098356},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1315883357233815e-06,"output_power":1.8743920588786772e-06,"reference_active":true,"delta_db":-3.4325660701218874}],"active_band_count":3,"active_band_mae_db":1.5798271216617954}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":0.689280556869813,"rms_envelope_nrmse":0.0639734443318571,"envelope_correlation":0.9976594993810415,"rms_gain_error_db":-0.24498123939533234,"reference_mean_power":0.00038520090394369153,"output_mean_power":0.0003640735878467296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9052955697898775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058163319658512,"output_power":0.00013341098882125305,"reference_active":true,"delta_db":-0.22736981290623706},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.901478078483037e-06,"output_power":3.166884850669667e-06,"reference_active":true,"delta_db":-0.9059689853657593},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3393802456709043e-07,"output_power":1.1959023946434756e-07,"reference_active":true,"delta_db":-2.914050827853982}],"active_band_count":3,"active_band_mae_db":1.3491298753753262}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":1.2462513873531984,"rms_envelope_nrmse":0.14953922084804747,"envelope_correlation":0.9849612941255945,"rms_gain_error_db":-0.38277124275350877,"reference_mean_power":4.0846100044459596e-05,"output_mean_power":3.740016106049181e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0558567323423285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3562756220605465e-05,"output_power":1.220895865124397e-05,"reference_active":true,"delta_db":-0.45669332849830924},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5216123885214778e-06,"output_power":1.2651171820598368e-06,"reference_active":true,"delta_db":-0.8017328140519664},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602773162915517e-07,"output_power":3.513454653490798e-07,"reference_active":true,"delta_db":-2.7399202474598927}],"active_band_count":3,"active_band_mae_db":1.3327821300033895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":0.7086530877200783,"rms_envelope_nrmse":0.07700766784542407,"envelope_correlation":0.9968775125522471,"rms_gain_error_db":-0.38276284777507696,"reference_mean_power":4.105094002568104e-05,"output_mean_power":3.75877925862642e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8992373996258787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2457002317192888e-05,"output_power":1.1381339823803164e-05,"reference_active":true,"delta_db":-0.39220154542703556},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11943579398167e-07,"output_power":6.403587329501029e-07,"reference_active":true,"delta_db":-1.0310251526193794},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742155092447495e-08,"output_power":2.0625099985057736e-08,"reference_active":true,"delta_db":-2.5872572016352935}],"active_band_count":3,"active_band_mae_db":1.3368279665605696}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":1.001085392335978,"rms_envelope_nrmse":0.1699892325200527,"envelope_correlation":0.9829585737905844,"rms_gain_error_db":-0.13748942031300762,"reference_mean_power":2.620464364014994e-05,"output_mean_power":2.5388048308991117e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.347390510898902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9544955639169026e-05,"output_power":2.046476614041135e-05,"reference_active":true,"delta_db":0.19972097082761972},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7959648252038243e-06,"output_power":2.9005542734709455e-06,"reference_active":true,"delta_db":0.15949292702036827},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.147434020032559e-06,"output_power":5.068593757627492e-07,"reference_active":true,"delta_db":-3.548402376643714}],"active_band_count":3,"active_band_mae_db":1.3025387581639005}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":0.8798893646922213,"rms_envelope_nrmse":0.08476199608956936,"envelope_correlation":0.9958119956202307,"rms_gain_error_db":-0.13748836951383459,"reference_mean_power":2.6336417520647037e-05,"output_mean_power":2.5515721994182866e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.2809655135480134e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9754827323518684e-05,"output_power":2.0654144458181635e-05,"reference_active":true,"delta_db":0.1933397234001351},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4198245590922995e-06,"output_power":1.4303843709617233e-06,"reference_active":true,"delta_db":0.032180721631704795},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517276499876375e-08,"output_power":2.649366970000231e-08,"reference_active":true,"delta_db":-3.909240290187845}],"active_band_count":3,"active_band_mae_db":1.3782535784065617}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0003","method":"magphase","status":"measured","mos":{"status":"measured","value":3.3332009315490723,"predicted_mos":3.3332009315490723,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":290280,"input_sample_rate":48000,"input_seconds":6.0475,"input_float32_sha256":"50f1a5a83c4350b4e87dc8eeeca96175a0c77f3ddaa3ed4dfa27e2959dd37406","internal_samples":96760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8013630059995194},"started_at":"2026-10-13T05:11:58.020034+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008374368999284343,"native_input_samples":288000,"native_input_sha256":"102614371175b28d658b06a57ab2ad9f5322dad2544938b0bdab2abee6e708cc","output_sha256":"aeccb024ddab6aeaaa177b507399cda52d667e015d113290bde55eed18fd9527","output_samples":290280,"length_delta_samples":2280,"analysis_s":3.252018148999923,"synthesis_s":0.935256759999902,"total_s":4.187274908999825,"rtf":0.6978791514999708,"output_peak":0.2716731959383995,"output_rms":0.02952411207055878,"finished_at":"2026-10-13T05:12:02.228376+00:00","evaluation_input":{"reference":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"aeccb024ddab6aeaaa177b507399cda52d667e015d113290bde55eed18fd9527","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96760,"length_delta_samples":760,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.594895654115337,"output_peak":0.2690210803597354},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9921455146754857,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.889437490658235,"spectral_convergence":0.8910120558633844,"spectral_convergence_status":"ok","amplitude_floor":8.79928282883349e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.12627681234229,"spectral_convergence":0.5976687226080635,"spectral_convergence_status":"ok","amplitude_floor":4.444740709272264e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.506835901882868,"spectral_convergence":0.5956167680108925,"spectral_convergence_status":"ok","amplitude_floor":3.930299196146418e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002477612487613689,"output_power":0.00022918172934592834,"delta_db":-0.3385338910060565,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004515434578601184,"output_power":0.0004237874113796389,"delta_db":-0.27551502596248245,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015370997116088465,"output_power":0.00014680902871182146,"delta_db":-0.19949275699133795,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.4065683577917706e-05,"output_power":4.3671929983700104e-05,"delta_db":-0.038981265692435665,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.662327944441642e-06,"output_power":9.636128542198008e-06,"delta_db":0.4626796315612437,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.476431620775013e-05,"output_power":2.5797807632737665e-05,"delta_db":0.17756459269485286,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24879452731806817,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.24788071310216503,"pitch":null,"output_rms_dbfs":-30.560684241562083,"output_peak":0.2690210803597354,"reference_rms_dbfs":-30.312803528459913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.174183401627799},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.45606629692133704,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030504212914161243,"rms_dbfs":-30.312803528459913,"peak":0.259429931640625,"peak_dbfs":-11.71959837564457,"crest_factor_db":18.593205152815347,"mean":-3.0800501505533854e-06,"zero_energy":false},"output":{"rms":0.029645978405106223,"rms_dbfs":-30.560684241562083,"peak":0.2690210803597354,"peak_dbfs":-11.404273751237618,"crest_factor_db":19.15641049032446,"mean":-0.00019124447609318018,"zero_energy":false},"rms_difference_db":-0.24788071310216964},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":11,"quiet_samples":1760,"reference_mean_power":8.345015389336782e-07,"output_mean_power":2.5076236872019453e-06,"output_rms_dbfs":-56.00737636435363,"output_zero_energy":false,"output_peak":0.005347917771214972,"output_energy_fraction":5.230849471609587e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":148,"both_estimated_voiced_frames":117,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.8666666666666667,"absolute_difference_cents_median":12.577051508073888,"absolute_difference_cents_p95":40.921494899172735,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.467120737686064e-05,"log_rms_mae_db":4.446196550571609,"rms_envelope_nrmse":0.40934121747967783,"envelope_correlation":0.851882252722996,"rms_gain_error_db":-0.2478762693000417,"reference_mean_power":0.0009312818437003785,"output_mean_power":0.0008796167871031255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.494393820063344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023666657525761142,"output_power":0.00022801301589539592,"reference_active":true,"delta_db":-0.1617728728442476},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5441929621447868e-05,"output_power":1.5664554310220233e-05,"reference_active":true,"delta_db":0.06216474023996368},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.985747798516416e-06,"output_power":6.153519360059632e-06,"reference_active":true,"delta_db":-0.5508933125712885}],"active_band_count":3,"active_band_mae_db":0.2582769752184999}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.334380857742655e-05,"log_rms_mae_db":3.4016791099468224,"rms_envelope_nrmse":0.33207101326742716,"envelope_correlation":0.8971263696705647,"rms_gain_error_db":-0.24783385649420817,"reference_mean_power":0.0009359486205868301,"output_mean_power":0.000884033296882729},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.189856446091556e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021037225251172397,"output_power":0.0002039121448921506,"reference_active":true,"delta_db":-0.13545364197498624},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.644252267790675e-06,"output_power":7.50367351277857e-06,"reference_active":true,"delta_db":-0.08061081270495837},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434290091658501e-07,"output_power":3.2375909995317094e-07,"reference_active":true,"delta_db":-0.25614992820730975}],"active_band_count":3,"active_band_mae_db":0.15740479429575147}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.589555428061273e-05,"log_rms_mae_db":4.697418534806322,"rms_envelope_nrmse":0.5348296079804035,"envelope_correlation":0.8054219439907067,"rms_gain_error_db":-0.26814345349247726,"reference_mean_power":0.00038327408961782034,"output_mean_power":0.00036032559744581845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.028222657916465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015396875485141256,"output_power":0.00014258569429389655,"reference_active":true,"delta_db":-0.33356642957043314},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.75189127488202e-06,"output_power":8.584615719855001e-06,"reference_active":true,"delta_db":-0.08381054509578545},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1315883357233815e-06,"output_power":3.3276073504594e-06,"reference_active":true,"delta_db":-0.9398496770266519}],"active_band_count":3,"active_band_mae_db":0.4524088838976235}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.149058869949865e-05,"log_rms_mae_db":3.8602605139128294,"rms_envelope_nrmse":0.4381985641116239,"envelope_correlation":0.8645105283329012,"rms_gain_error_db":-0.26814653392213494,"reference_mean_power":0.00038520090394369153,"output_mean_power":0.0003621367871228903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9052955697898775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058163319658512,"output_power":0.00013073660500353274,"reference_active":true,"delta_db":-0.31531381470315634},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.901478078483037e-06,"output_power":3.7996129732911946e-06,"reference_active":true,"delta_db":-0.11489809209305991},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3393802456709043e-07,"output_power":2.1542545179751244e-07,"reference_active":true,"delta_db":-0.3580380585711447}],"active_band_count":3,"active_band_mae_db":0.26274998845578695}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.043858753489399e-05,"log_rms_mae_db":4.201122199887989,"rms_envelope_nrmse":0.5019024440458257,"envelope_correlation":0.8237092803814651,"rms_gain_error_db":-0.02176336240777047,"reference_mean_power":4.0846100044459596e-05,"output_mean_power":4.064192410347164e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0558567323423285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3562756220605465e-05,"output_power":1.3685986934169318e-05,"reference_active":true,"delta_db":0.03928165307507357},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5216123885214778e-06,"output_power":1.5548717475328507e-06,"reference_active":true,"delta_db":0.09390536869218316},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602773162915517e-07,"output_power":5.575777012445062e-07,"reference_active":true,"delta_db":-0.7342098037232508}],"active_band_count":3,"active_band_mae_db":0.28913227516350254}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.7804457921821966e-05,"log_rms_mae_db":3.285141832017806,"rms_envelope_nrmse":0.3986961165661899,"envelope_correlation":0.8832741556292966,"rms_gain_error_db":-0.021777823607161434,"reference_mean_power":4.105094002568104e-05,"output_mean_power":4.0845604149897515e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8992373996258787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2457002317192888e-05,"output_power":1.2774597888905302e-05,"reference_active":true,"delta_db":0.10933693328027122},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11943579398167e-07,"output_power":7.378868973727841e-07,"reference_active":true,"delta_db":-0.4153605322799683},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742155092447495e-08,"output_power":2.8013934518485676e-08,"reference_active":true,"delta_db":-1.2574767356728906}],"active_band_count":3,"active_band_mae_db":0.59405806707771}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.792437835062945e-05,"log_rms_mae_db":4.473503983844125,"rms_envelope_nrmse":0.762999555090055,"envelope_correlation":0.6787271544678946,"rms_gain_error_db":0.48726019627573014,"reference_mean_power":2.620464364014994e-05,"output_mean_power":2.9315970613881026e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.347390510898902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9544955639169026e-05,"output_power":2.3547086081275266e-05,"reference_active":true,"delta_db":0.8090248240067275},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7959648252038243e-06,"output_power":2.90729605902905e-06,"reference_active":true,"delta_db":0.16957556143252323},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.147434020032559e-06,"output_power":1.8127160334246456e-06,"reference_active":true,"delta_db":1.986020540449356}],"active_band_count":3,"active_band_mae_db":0.9882069752962023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.70505017465466e-05,"log_rms_mae_db":3.9343885425743332,"rms_envelope_nrmse":0.6231205665062605,"envelope_correlation":0.7796433960230073,"rms_gain_error_db":0.4872616583925231,"reference_mean_power":2.6336417520647037e-05,"output_mean_power":2.9463400176092095e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.2809655135480134e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9754827323518684e-05,"output_power":2.3874437239283414e-05,"reference_active":true,"delta_db":0.8225990544746808},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4198245590922995e-06,"output_power":1.3982715770250403e-06,"reference_active":true,"delta_db":-0.06643154270926334},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517276499876375e-08,"output_power":9.036077590481559e-08,"reference_active":true,"delta_db":1.4191380481525386}],"active_band_count":3,"active_band_mae_db":0.7693895484454942}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"hnm","status":"measured","mos":{"status":"measured","value":2.8923630714416504,"predicted_mos":2.8923630714416504,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68000,"input_sample_rate":16000,"input_seconds":4.25,"input_float32_sha256":"5cb7f3f761b25ff2c2d4b6c0054d8e4fefd21fd55b465c9ec5e6e7157c34a64d","internal_samples":68000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9633400259990594},"started_at":"2026-10-13T04:38:00.774624+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5458000234502833e-05,"native_input_samples":68000,"native_input_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","output_sha256":"f4de4f4fc0a47e1916d4d74150270b529efc3709f05319b7b81297cac63903f8","output_samples":68000,"length_delta_samples":0,"analysis_s":0.7975571500001024,"synthesis_s":0.3519657130000269,"total_s":1.1495228630001293,"rtf":0.2704759677647363,"output_peak":0.30222012397033254,"output_rms":0.04549087524860168,"finished_at":"2026-10-13T04:38:01.928135+00:00","evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"f4de4f4fc0a47e1916d4d74150270b529efc3709f05319b7b81297cac63903f8","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.077227732978913,"spectral_convergence":0.18325281924990622,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6261581769942905,"spectral_convergence":0.18294103080166468,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.703659196486805,"spectral_convergence":0.1787935621202677,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0006189619175620867,"delta_db":0.052870743890497805,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0010642520890023518,"delta_db":-0.18388618546403304,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.00029154039129246313,"delta_db":-0.29731753938307026,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":8.134591512908132e-05,"delta_db":-0.4423932126879244,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":8.165630994147907e-06,"delta_db":-0.8289112764555098,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":5.153786903710639e-06,"delta_db":-0.35729376135868285,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3604454532066197,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14463646261622287,"pitch":null,"output_rms_dbfs":-26.841514144481906,"output_peak":0.30222012397033254,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.802348368820002},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.077227732978913,"spectral_convergence":0.18325281924990622,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6261581769942905,"spectral_convergence":0.18294103080166468,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.703659196486805,"spectral_convergence":0.1787935621202677,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0006189619175620867,"delta_db":0.052870743890497805,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0010642520890023518,"delta_db":-0.18388618546403304,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.00029154039129246313,"delta_db":-0.29731753938307026,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":8.134591512908132e-05,"delta_db":-0.4423932126879244,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":8.165630994147907e-06,"delta_db":-0.8289112764555098,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":5.153786903710639e-06,"delta_db":-0.35729376135868285,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3604454532066197,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14463646261622287,"pitch":null,"output_rms_dbfs":-26.841514144481906,"output_peak":0.30222012397033254,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.802348368820002},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.820913089008821,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046254726345655035,"rms_dbfs":-26.69687768186568,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.21328676657332,"mean":-2.2544412051930147e-05,"zero_energy":false},"output":{"rms":0.04549087524860168,"rms_dbfs":-26.841514144481906,"peak":0.30222012397033254,"peak_dbfs":-10.393532412230655,"crest_factor_db":16.44798173225125,"mean":-0.0001495918062705153,"zero_energy":false},"rms_difference_db":-0.14463646261622642},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":51,"quiet_samples":8160,"reference_mean_power":1.0693246069565124e-06,"output_mean_power":3.4369551676709653e-06,"output_rms_dbfs":-54.63826132849711,"output_zero_energy":false,"output_peak":0.03871746903629192,"output_energy_fraction":0.00019929964615944133,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":132,"both_estimated_voiced_frames":121,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.983739837398374,"absolute_difference_cents_median":2.1111900311199587,"absolute_difference_cents_p95":8.683416000047101,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.0001318685794969069,"log_rms_mae_db":1.621352769511145,"rms_envelope_nrmse":0.06472329733932541,"envelope_correlation":0.9966133443014596,"rms_gain_error_db":-0.14466437864318313,"reference_mean_power":0.0021420152585098387,"output_mean_power":0.002071839564859794},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.3134511495070254e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003582591828573966,"output_power":0.00032909853871416567,"reference_active":true,"delta_db":-0.36871377082720697},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.463795601086153e-05,"output_power":4.443767882949859e-05,"reference_active":true,"delta_db":-0.019529340754414693},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.556588167851325e-06,"output_power":6.929566423798867e-06,"reference_active":true,"delta_db":-0.3761969184259445}],"active_band_count":3,"active_band_mae_db":0.2548133433358554}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00012588557284836115,"log_rms_mae_db":0.9464890719113325,"rms_envelope_nrmse":0.03887924294458006,"envelope_correlation":0.9988866378726771,"rms_gain_error_db":-0.14467514812574117,"reference_mean_power":0.0021571542863483557,"output_mean_power":0.002086477441028214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.2592546923162062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031562831458090785,"output_power":0.0002902389715817123,"reference_active":true,"delta_db":-0.36420229718835345},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.98767508506911e-05,"output_power":1.894299231270263e-05,"reference_active":true,"delta_db":-0.20896810975340466},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602361817361657e-07,"output_power":3.9557204607276754e-07,"reference_active":true,"delta_db":-2.2247372682573}],"active_band_count":3,"active_band_mae_db":0.9326358917330194}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00010444605382463925,"log_rms_mae_db":2.0477269620851084,"rms_envelope_nrmse":0.09900288471567917,"envelope_correlation":0.9939225631793062,"rms_gain_error_db":-0.28108365253472734,"reference_mean_power":0.0005604503758559807,"output_mean_power":0.0005253258833865062},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":5.643164593552106e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025020135526255573,"output_power":0.00023008801566223037,"reference_active":true,"delta_db":-0.3639565915568724},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8332303804350734e-05,"output_power":3.397288591924363e-05,"reference_active":true,"delta_db":-0.5243248074609291},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.791716397073577e-06,"output_power":3.60455552187174e-06,"reference_active":true,"delta_db":-2.751278232437056}],"active_band_count":3,"active_band_mae_db":1.2131865438182858}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":1.5828672320390604,"rms_envelope_nrmse":0.059294373518120315,"envelope_correlation":0.9982300870938127,"rms_gain_error_db":-0.28108733845844813,"reference_mean_power":0.0005644391167624198,"output_mean_power":0.0005290641932784528},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":5.412619397247975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022898404327685652,"output_power":0.00021144911142839402,"reference_active":true,"delta_db":-0.3459935518696639},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5949246143041614e-05,"output_power":1.4172143131896832e-05,"reference_active":true,"delta_db":-0.5130463066757192},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.408378605310093e-07,"output_power":3.0956688764361674e-07,"reference_active":true,"delta_db":-3.15993660997815}],"active_band_count":3,"active_band_mae_db":1.3396588228411777}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":2.0881805715156685,"rms_envelope_nrmse":0.16218261540734027,"envelope_correlation":0.9840624334054344,"rms_gain_error_db":-0.46012420074117594,"reference_mean_power":8.438545283737647e-05,"output_mean_power":7.590234003631334e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":6.948180578959807e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.55589320089413e-05,"output_power":2.8962865583994707e-05,"reference_active":true,"delta_db":-0.891071800033546},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.93670906958722e-06,"output_power":2.9997778847167456e-06,"reference_active":true,"delta_db":-1.1804422172544953},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579820903321111e-07,"output_power":8.897575011011262e-07,"reference_active":true,"delta_db":0.6961271380635558}],"active_band_count":3,"active_band_mae_db":0.9225470517838658}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":1.5322871045140312,"rms_envelope_nrmse":0.10112891960548302,"envelope_correlation":0.9947087849389037,"rms_gain_error_db":-0.46014036943988934,"reference_mean_power":8.498589632737071e-05,"output_mean_power":7.644213746465812e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":6.685624503092089e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.278937461914447e-05,"output_power":2.701171086785864e-05,"reference_active":true,"delta_db":-0.8418104118952157},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.678676767994248e-06,"output_power":1.2585154801551537e-06,"reference_active":true,"delta_db":-1.2510851818015913},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9241501206570436e-08,"output_power":4.783366012240609e-08,"reference_active":true,"delta_db":-0.9289244184309033}],"active_band_count":3,"active_band_mae_db":1.0072733373759035}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":1.9131713311958631,"rms_envelope_nrmse":0.1664897461722657,"envelope_correlation":0.9861235997707942,"rms_gain_error_db":-0.6808613248281791,"reference_mean_power":9.160796502339733e-06,"output_mean_power":7.831538790130267e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":3.736107367401369e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6569760918264079e-06,"output_power":1.481711802217868e-06,"reference_active":true,"delta_db":-0.48552501954219845},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.026752210402777e-07,"output_power":4.001775852266308e-07,"reference_active":true,"delta_db":-1.7783057659021286},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8917938491313026e-07,"output_power":6.866656584937902e-08,"reference_active":true,"delta_db":-4.401284810817269}],"active_band_count":3,"active_band_mae_db":2.2217051987538654}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":1.390984142398019,"rms_envelope_nrmse":0.10807853347166547,"envelope_correlation":0.9962633198126932,"rms_gain_error_db":-0.6808927388402973,"reference_mean_power":9.225978949789666e-06,"output_mean_power":7.887206027383875e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":3.5247411287487305e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.710468659980512e-06,"output_power":1.4639252826808988e-06,"reference_active":true,"delta_db":-0.6759620992303551},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.274879265893732e-07,"output_power":1.971737035063693e-07,"reference_active":true,"delta_db":-2.203462997506593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9472537885262174e-08,"output_power":6.088132444028726e-09,"reference_active":true,"delta_db":-5.049384657575419}],"active_band_count":3,"active_band_mae_db":2.642936584770789}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.245880603790283,"predicted_mos":3.245880603790283,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68000,"input_sample_rate":16000,"input_seconds":4.25,"input_float32_sha256":"594de0a626b6fb33f02f98640ea0de5998cf9135031f0ca5b155019ae13dcdff","internal_samples":68000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.746462535000319},"started_at":"2026-10-13T04:40:51.988745+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4837999262672383e-05,"native_input_samples":68000,"native_input_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","output_sha256":"fdb3e927ec14b13fbd4f15a5d687cf60090a1572162f7413e8843ddd2f16e631","output_samples":68000,"length_delta_samples":0,"analysis_s":1.1659911039996587,"synthesis_s":0.0711526800005231,"total_s":1.2371437840001818,"rtf":0.2910926550588663,"output_peak":0.5150196716309065,"output_rms":0.04993390373795378,"finished_at":"2026-10-13T04:40:53.230840+00:00","evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"fdb3e927ec14b13fbd4f15a5d687cf60090a1572162f7413e8843ddd2f16e631","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.1647525903893445,"spectral_convergence":0.8688399800098868,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.719691862044741,"spectral_convergence":0.3262136217359938,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.913383164699269,"spectral_convergence":0.2538200525300192,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0006363493892680928,"delta_db":0.17318776715521605,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0013346453436209155,"delta_db":0.7993274574410779,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.0004014904789210093,"delta_db":1.0924476445680575,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":0.00010565028207664835,"delta_db":0.6929559038287519,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":9.167528955349023e-06,"delta_db":-0.3262858751567278,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":6.091719669219588e-06,"delta_db":0.36884075892329216,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5755075678455206,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6647880759019495,"pitch":null,"output_rms_dbfs":-26.03208960596373,"output_peak":0.5150196716309065,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.265942539044452},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.1647525903893445,"spectral_convergence":0.8688399800098868,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.719691862044741,"spectral_convergence":0.3262136217359938,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.913383164699269,"spectral_convergence":0.2538200525300192,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0006363493892680928,"delta_db":0.17318776715521605,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0013346453436209155,"delta_db":0.7993274574410779,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.0004014904789210093,"delta_db":1.0924476445680575,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":0.00010565028207664835,"delta_db":0.6929559038287519,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":9.167528955349023e-06,"delta_db":-0.3262858751567278,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":6.091719669219588e-06,"delta_db":0.36884075892329216,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5755075678455206,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6647880759019495,"pitch":null,"output_rms_dbfs":-26.03208960596373,"output_peak":0.5150196716309065,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.265942539044452},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.8299198779902595,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046254726345655035,"rms_dbfs":-26.69687768186568,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.21328676657332,"mean":-2.2544412051930147e-05,"zero_energy":false},"output":{"rms":0.04993390373795378,"rms_dbfs":-26.03208960596373,"peak":0.5150196716309065,"peak_dbfs":-5.763523647619227,"crest_factor_db":20.268565958344507,"mean":1.661175262950418e-09,"zero_energy":false},"rms_difference_db":0.6647880759019493},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":51,"quiet_samples":8160,"reference_mean_power":1.0693246069565124e-06,"output_mean_power":5.314810134176594e-06,"output_rms_dbfs":-52.745122455625555,"output_zero_energy":false,"output_peak":0.07847045791325151,"output_energy_fraction":0.00025578670125005995,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":131,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.8617886178861789,"absolute_difference_cents_median":5.5744239917413,"absolute_difference_cents_p95":18.10683452512496,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.0001318685794969069,"log_rms_mae_db":5.609514033155309,"rms_envelope_nrmse":0.1377909710508401,"envelope_correlation":0.9904541392480063,"rms_gain_error_db":0.6647968160467643,"reference_mean_power":0.0021420152585098387,"output_mean_power":0.002496331413000455},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.3134511495070254e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003582591828573966,"output_power":0.00039725488105141443,"reference_active":true,"delta_db":0.4487191107006843},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.463795601086153e-05,"output_power":5.136392654199873e-05,"reference_active":true,"delta_db":0.6095391632292131},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.556588167851325e-06,"output_power":1.2918989178249386e-05,"reference_active":true,"delta_db":2.3290278051914584}],"active_band_count":3,"active_band_mae_db":1.1290953597071185}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00012588557284836115,"log_rms_mae_db":5.373528647577033,"rms_envelope_nrmse":0.10646715208234105,"envelope_correlation":0.9963352049065377,"rms_gain_error_db":0.665004834288455,"reference_mean_power":0.0021571542863483557,"output_mean_power":0.002514095042839513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.2592546923162062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031562831458090785,"output_power":0.00035541286938865606,"reference_active":true,"delta_db":0.5155719317696446},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.98767508506911e-05,"output_power":2.348325515045432e-05,"reference_active":true,"delta_db":0.7241290283220314},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602361817361657e-07,"output_power":9.490911503828899e-07,"reference_active":true,"delta_db":1.5760860341484741}],"active_band_count":3,"active_band_mae_db":0.9385956647467166}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00010444605382463925,"log_rms_mae_db":3.023235401830955,"rms_envelope_nrmse":0.18010093837674024,"envelope_correlation":0.9885805591006953,"rms_gain_error_db":0.9577895449156293,"reference_mean_power":0.0005604503758559807,"output_mean_power":0.0006987408264963103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":5.643164593552106e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025020135526255573,"output_power":0.0003207660112563892,"reference_active":true,"delta_db":1.0789868589495308},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8332303804350734e-05,"output_power":4.932001765698132e-05,"reference_active":true,"delta_db":1.0945830189503596},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.791716397073577e-06,"output_power":6.775731257314467e-06,"reference_active":true,"delta_db":-0.010233702578378114}],"active_band_count":3,"active_band_mae_db":0.7279345268260894}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":2.7102910580958457,"rms_envelope_nrmse":0.15068941177465114,"envelope_correlation":0.9942524078323819,"rms_gain_error_db":0.957782410769805,"reference_mean_power":0.0005644391167624198,"output_mean_power":0.0007037126283965424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":5.412619397247975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022898404327685652,"output_power":0.00029452825663969085,"reference_active":true,"delta_db":1.0932174710278908},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5949246143041614e-05,"output_power":2.0835689680937536e-05,"reference_active":true,"delta_db":1.1606772006898678},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.408378605310093e-07,"output_power":5.283726094827144e-07,"reference_active":true,"delta_db":-0.8380786587496809}],"active_band_count":3,"active_band_mae_db":1.0306577768224798}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":1.5605688885886104,"rms_envelope_nrmse":0.17285661326070512,"envelope_correlation":0.986810490548351,"rms_gain_error_db":0.7161981435395425,"reference_mean_power":8.438545283737647e-05,"output_mean_power":9.951473701432306e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":6.948180578959807e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.55589320089413e-05,"output_power":4.075558270632463e-05,"reference_active":true,"delta_db":0.5923839815790464},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.93670906958722e-06,"output_power":4.599935377693648e-06,"reference_active":true,"delta_db":0.6761840975352084},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579820903321111e-07,"output_power":4.2438633653983887e-07,"reference_active":true,"delta_db":-2.5189755121134216}],"active_band_count":3,"active_band_mae_db":1.2625145304092256}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":0.8789548638776472,"rms_envelope_nrmse":0.1276902205211852,"envelope_correlation":0.9946183755274226,"rms_gain_error_db":0.7161991353403354,"reference_mean_power":8.498589632737071e-05,"output_mean_power":0.00010022285559874039},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":6.685624503092089e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.278937461914447e-05,"output_power":3.773614572924052e-05,"reference_active":true,"delta_db":0.6102440680303686},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.678676767994248e-06,"output_power":1.968828142251476e-06,"reference_active":true,"delta_db":0.692407284425302},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9241501206570436e-08,"output_power":3.159149757127537e-08,"reference_active":true,"delta_db":-2.7305584141051447}],"active_band_count":3,"active_band_mae_db":1.3444032555202716}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":1.598113209459749,"rms_envelope_nrmse":0.29911190813049987,"envelope_correlation":0.9468692092064209,"rms_gain_error_db":0.08474219677380984,"reference_mean_power":9.160796502339733e-06,"output_mean_power":9.341302917874877e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":3.736107367401369e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6569760918264079e-06,"output_power":1.8447949476294544e-06,"reference_active":true,"delta_db":0.4663185842562495},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.026752210402777e-07,"output_power":6.135534777622832e-07,"reference_active":true,"delta_db":0.07769086291691504},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8917938491313026e-07,"output_power":8.06965366191637e-08,"reference_active":true,"delta_db":-3.7001891327479375}],"active_band_count":3,"active_band_mae_db":1.4147328599737008}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":0.8991468307084132,"rms_envelope_nrmse":0.18573225839873866,"envelope_correlation":0.9787939503752137,"rms_gain_error_db":0.08474204284350742,"reference_mean_power":9.225978949789666e-06,"output_mean_power":9.407769401631502e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":3.5247411287487305e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.710468659980512e-06,"output_power":1.8485515190781184e-06,"reference_active":true,"delta_db":0.33716437608485184},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.274879265893732e-07,"output_power":3.0396429673369866e-07,"reference_active":true,"delta_db":-0.32372718685845625},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9472537885262174e-08,"output_power":7.45327793527216e-09,"reference_active":true,"delta_db":-4.170752409430735}],"active_band_count":3,"active_band_mae_db":1.6105479907913478}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"facodec","status":"measured","mos":{"status":"measured","value":4.278822898864746,"predicted_mos":4.278822898864746,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68000,"input_sample_rate":16000,"input_seconds":4.25,"input_float32_sha256":"4d6c1da36d082d6a10cb64b18c1e5c4c683171b27921da858f356ac9df629904","internal_samples":68000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3774629999988974},"started_at":"2026-10-13T04:45:02.644669+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.8944000405317638e-05,"native_input_samples":68000,"native_input_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","output_sha256":"a8ab4617ffd6b19e2065dbebd5a620373fa614b607e765b32cfe1ce736766779","output_samples":68000,"length_delta_samples":0,"analysis_s":1.2726721369999723,"synthesis_s":9.17969644599998,"total_s":10.452368582999952,"rtf":2.4593808430588124,"output_peak":0.32441389560699463,"output_rms":0.04574989083357321,"finished_at":"2026-10-13T04:45:13.101775+00:00","evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"a8ab4617ffd6b19e2065dbebd5a620373fa614b607e765b32cfe1ce736766779","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.458561665488908,"spectral_convergence":0.1879188049531544,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.768034514772183,"spectral_convergence":0.18556461883779052,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.740713269736849,"spectral_convergence":0.19274674288437038,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0006524815782288677,"delta_db":0.2819139965552775,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0010393529363385418,"delta_db":-0.2867008252337154,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.00028445362920184476,"delta_db":-0.4041900719034583,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":0.00010743520046232902,"delta_db":0.7657152797939365,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":6.226127719724144e-06,"delta_db":-2.006628525143691,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":3.103039332558499e-06,"delta_db":-2.560685532072569,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0509723717837745,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09532107595747216,"pitch":null,"output_rms_dbfs":-26.792198757823154,"output_peak":0.32441389560699463,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.65576981666598},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.458561665488908,"spectral_convergence":0.1879188049531544,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.768034514772183,"spectral_convergence":0.18556461883779052,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.740713269736849,"spectral_convergence":0.19274674288437038,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0006524815782288677,"delta_db":0.2819139965552775,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0010393529363385418,"delta_db":-0.2867008252337154,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.00028445362920184476,"delta_db":-0.4041900719034583,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":0.00010743520046232902,"delta_db":0.7657152797939365,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":6.226127719724144e-06,"delta_db":-2.006628525143691,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":3.103039332558499e-06,"delta_db":-2.560685532072569,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0509723717837745,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09532107595747216,"pitch":null,"output_rms_dbfs":-26.792198757823154,"output_peak":0.32441389560699463,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.65576981666598},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.8292322815260024,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046254726345655035,"rms_dbfs":-26.69687768186568,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.21328676657332,"mean":-2.2544412051930147e-05,"zero_energy":false},"output":{"rms":0.045749890833573205,"rms_dbfs":-26.792198757823154,"peak":0.32441389560699463,"peak_dbfs":-9.778011038897363,"crest_factor_db":17.01418771892579,"mean":0.00010072122742169942,"zero_energy":false},"rms_difference_db":-0.09532107595747519},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":51,"quiet_samples":8160,"reference_mean_power":1.0693246069565124e-06,"output_mean_power":1.5583479066168116e-06,"output_rms_dbfs":-58.07335578077755,"output_zero_energy":false,"output_peak":0.0048521291464567184,"output_energy_fraction":8.934403116303646e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":127,"both_estimated_voiced_frames":119,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.967479674796748,"absolute_difference_cents_median":2.890716754546709,"absolute_difference_cents_p95":12.323135464613374,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.0001318685794969069,"log_rms_mae_db":1.4694996767553543,"rms_envelope_nrmse":0.08172757970225959,"envelope_correlation":0.9943593192142492,"rms_gain_error_db":-0.09531442368396728,"reference_mean_power":0.0021420152585098387,"output_mean_power":0.002095516659275113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.3134511495070254e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003582591828573966,"output_power":0.0003459028396713231,"reference_active":true,"delta_db":-0.15243203650424728},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.463795601086153e-05,"output_power":4.410889477142457e-05,"reference_active":true,"delta_db":-0.05178123997115965},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.556588167851325e-06,"output_power":1.0001626597027541e-05,"reference_active":true,"delta_db":1.217448825436032}],"active_band_count":3,"active_band_mae_db":0.47388736730381303}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00012588557284836115,"log_rms_mae_db":1.0976892164472576,"rms_envelope_nrmse":0.07095372162985476,"envelope_correlation":0.9956233479484476,"rms_gain_error_db":-0.0953143243214702,"reference_mean_power":0.0021571542863483557,"output_mean_power":0.0021103270992737233},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.2592546923162062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031562831458090785,"output_power":0.0003056641926152025,"reference_active":true,"delta_db":-0.13931390466919938},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.98767508506911e-05,"output_power":1.9881245157613802e-05,"reference_active":true,"delta_db":0.0009818667441967585},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602361817361657e-07,"output_power":7.588075440402189e-07,"reference_active":true,"delta_db":0.6043231947392428}],"active_band_count":3,"active_band_mae_db":0.24820632205087964}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00010444605382463925,"log_rms_mae_db":1.8304330765896302,"rms_envelope_nrmse":0.1263722102233539,"envelope_correlation":0.9907498187682272,"rms_gain_error_db":-0.480212908253967,"reference_mean_power":0.0005604503758559807,"output_mean_power":0.0005017829194031661},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":5.643164593552106e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025020135526255573,"output_power":0.00020471134631873579,"reference_active":true,"delta_db":-0.8714774329188991},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8332303804350734e-05,"output_power":3.206734482966529e-05,"reference_active":true,"delta_db":-0.7750191959700446},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.791716397073577e-06,"output_power":6.7599405436048515e-06,"reference_active":true,"delta_db":-0.020366665376721542}],"active_band_count":3,"active_band_mae_db":0.5556210980885551}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":1.4813206466249378,"rms_envelope_nrmse":0.11418216255098596,"envelope_correlation":0.9925737373993249,"rms_gain_error_db":-0.4802128256299417,"reference_mean_power":0.0005644391167624198,"output_mean_power":0.0005053541320052798},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":5.412619397247975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022898404327685652,"output_power":0.00018800711013011015,"reference_active":true,"delta_db":-0.8563094573871156},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5949246143041614e-05,"output_power":1.3365163512132062e-05,"reference_active":true,"delta_db":-0.7676588414612775},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.408378605310093e-07,"output_power":6.722450509786762e-07,"reference_active":true,"delta_db":0.2077945182426004}],"active_band_count":3,"active_band_mae_db":0.6105876056969979}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":1.8567745783423917,"rms_envelope_nrmse":0.15952470212462225,"envelope_correlation":0.9900259187461682,"rms_gain_error_db":0.7471318843464058,"reference_mean_power":8.438545283737647e-05,"output_mean_power":0.00010022608668649621},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":6.948180578959807e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.55589320089413e-05,"output_power":4.2889652249166734e-05,"reference_active":true,"delta_db":0.8140381625672872},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.93670906958722e-06,"output_power":5.488427007775914e-06,"reference_active":true,"delta_db":1.4431457199503992},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579820903321111e-07,"output_power":1.0961271283016745e-06,"reference_active":true,"delta_db":1.60201982131563}],"active_band_count":3,"active_band_mae_db":1.2864012346111056}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":1.524166271433197,"rms_envelope_nrmse":0.12993951442005597,"envelope_correlation":0.994751302053966,"rms_gain_error_db":0.7471359365526606,"reference_mean_power":8.498589632737071e-05,"output_mean_power":0.00010093933816335205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":6.685624503092089e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.278937461914447e-05,"output_power":3.988363698100519e-05,"reference_active":true,"delta_db":0.850616210798526},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.678676767994248e-06,"output_power":2.4172885968731002e-06,"reference_active":true,"delta_db":1.5836142328142744},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9241501206570436e-08,"output_power":8.38312555141761e-08,"reference_active":true,"delta_db":1.5077991529019403}],"active_band_count":3,"active_band_mae_db":1.3140098655049135}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":2.3183515527766336,"rms_envelope_nrmse":0.3006358235749126,"envelope_correlation":0.9793452225071492,"rms_gain_error_db":-2.5090837849012813,"reference_mean_power":9.160796502339733e-06,"output_mean_power":5.140730743153911e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":3.736107367401369e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6569760918264079e-06,"output_power":9.198019757087576e-07,"reference_active":true,"delta_db":-2.556219040257756},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.026752210402777e-07,"output_power":3.556169547491846e-07,"reference_active":true,"delta_db":-2.291008771910512},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8917938491313026e-07,"output_power":2.039057051558508e-07,"reference_active":true,"delta_db":0.32555568145867303}],"active_band_count":3,"active_band_mae_db":1.7242611645423136}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":1.796467120133476,"rms_envelope_nrmse":0.2791094015249645,"envelope_correlation":0.9895601558264326,"rms_gain_error_db":-2.509088517825371,"reference_mean_power":9.225978949789666e-06,"output_mean_power":5.177303297114414e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":3.5247411287487305e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.710468659980512e-06,"output_power":9.798425530304966e-07,"reference_active":true,"delta_db":-2.4195882498375703},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.274879265893732e-07,"output_power":2.0714907043278054e-07,"reference_active":true,"delta_db":-1.9891230483128688},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9472537885262174e-08,"output_power":1.957906228251036e-08,"reference_active":true,"delta_db":0.023693304478903162}],"active_band_count":3,"active_band_mae_db":1.4774682008764473}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.9759910106658936,"predicted_mos":3.9759910106658936,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":187904,"input_sample_rate":44100,"input_seconds":4.260861678004535,"input_float32_sha256":"a53a5970c8bc0d963a8e98b53579118e506fc68870bf2f558d713f02d1a756eb","internal_samples":68174,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.844333491000725},"started_at":"2026-10-13T04:57:57.071251+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00481092600057309,"native_input_samples":187425,"native_input_sha256":"87db71c6c7a3a90cd1818e36cdd24ceb092d779817624f11933f37242759dcd7","output_sha256":"8cee81afb13617b8df6e009d9011599ea5dfe07e90827d3502913d72bbb63171","output_samples":187904,"length_delta_samples":479,"analysis_s":0.09250728100050765,"synthesis_s":24.15756308500022,"total_s":24.250070366000728,"rtf":5.70589890964723,"output_peak":0.3700203001499176,"output_rms":0.04057993896412264,"finished_at":"2026-10-13T04:58:21.337463+00:00","evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"8cee81afb13617b8df6e009d9011599ea5dfe07e90827d3502913d72bbb63171","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":68000,"output_samples":68174,"length_delta_samples":174,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.830415957855145,"output_peak":0.3698475571738008},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.997447707337108,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.862478638495432,"spectral_convergence":0.4791391801350258,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.782120193141658,"spectral_convergence":0.4210846933653901,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.41356977694735,"spectral_convergence":0.4621623057794077,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0006344005382472589,"delta_db":0.15986687860535886,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.000701202483952391,"delta_db":-1.9958968398854895,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.00023057765058869713,"delta_db":-1.3161327653666366,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":7.17177409525861e-05,"delta_db":-0.9894846933240358,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":1.2467091993565534e-05,"delta_db":1.0088428625396166,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":1.8550318082668863e-06,"delta_db":-4.795044728527685,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.7108781280414702,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1224421555704909,"pitch":null,"output_rms_dbfs":-27.81931983743617,"output_peak":0.3698475571738008,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.68605620286148},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.6353927914071069,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046254726345655035,"rms_dbfs":-26.69687768186568,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.21328676657332,"mean":-2.2544412051930147e-05,"zero_energy":false},"output":{"rms":0.04064751576102487,"rms_dbfs":-27.81931983743617,"peak":0.3698475571738008,"peak_dbfs":-8.639544908955791,"crest_factor_db":19.17977492848038,"mean":-0.00015289505197387297,"zero_energy":false},"rms_difference_db":-1.1224421555704893},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":51,"quiet_samples":8160,"reference_mean_power":1.0693246069565124e-06,"output_mean_power":1.2087127610711385e-06,"output_rms_dbfs":-59.17676892772804,"output_zero_energy":false,"output_peak":0.004678398081433515,"output_energy_fraction":8.778823893827934e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":130,"both_estimated_voiced_frames":109,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.8861788617886179,"absolute_difference_cents_median":8.820995272433896,"absolute_difference_cents_p95":59.897484726567654,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.0001318685794969069,"log_rms_mae_db":2.4838817211568753,"rms_envelope_nrmse":0.16053670623718935,"envelope_correlation":0.989963665413149,"rms_gain_error_db":-1.12244189777971,"reference_mean_power":0.0021420152585098387,"output_mean_power":0.0016541632590609638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.3134511495070254e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003582591828573966,"output_power":0.0002931494049889075,"reference_active":true,"delta_db":-0.8710831397163181},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.463795601086153e-05,"output_power":3.241334855112978e-05,"reference_active":true,"delta_db":-1.3898040052448288},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.556588167851325e-06,"output_power":5.366796542458489e-06,"reference_active":true,"delta_db":-1.4861062261260496}],"active_band_count":3,"active_band_mae_db":1.2489977903623988}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00012588557284836115,"log_rms_mae_db":2.1500985541606115,"rms_envelope_nrmse":0.1512364987032583,"envelope_correlation":0.9924827193149401,"rms_gain_error_db":-1.1223699772524265,"reference_mean_power":0.0021571542863483557,"output_mean_power":0.0016658819038940875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.2592546923162062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031562831458090785,"output_power":0.00025792316103634525,"reference_active":true,"delta_db":-0.876856134911119},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.98767508506911e-05,"output_power":1.4878481843291809e-05,"reference_active":true,"delta_db":-1.2578677466419774},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602361817361657e-07,"output_power":3.630019387384438e-07,"reference_active":true,"delta_db":-2.597903759625084}],"active_band_count":3,"active_band_mae_db":1.5775425470593933}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00010444605382463925,"log_rms_mae_db":2.5770801352565367,"rms_envelope_nrmse":0.1770222922965659,"envelope_correlation":0.9785924354639255,"rms_gain_error_db":-0.10315802077014327,"reference_mean_power":0.0005604503758559807,"output_mean_power":0.0005472948514990611},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":5.643164593552106e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025020135526255573,"output_power":0.0001982279812895594,"reference_active":true,"delta_db":-1.011246995696057},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8332303804350734e-05,"output_power":2.6053112757297697e-05,"reference_active":true,"delta_db":-1.6770530241331012},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.791716397073577e-06,"output_power":3.6699291029554427e-06,"reference_active":true,"delta_db":-2.6732186822943667}],"active_band_count":3,"active_band_mae_db":1.7871729007078416}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":2.155586931814733,"rms_envelope_nrmse":0.1569351274072232,"envelope_correlation":0.9824783881154245,"rms_gain_error_db":-0.10315727280835604,"reference_mean_power":0.0005644391167624198,"output_mean_power":0.0005511900591013667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":5.412619397247975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022898404327685652,"output_power":0.00017933436394281766,"reference_active":true,"delta_db":-1.0614170285657412},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5949246143041614e-05,"output_power":1.0791283810202514e-05,"reference_active":true,"delta_db":-1.6966704590767325},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.408378605310093e-07,"output_power":3.322441988942191e-07,"reference_active":true,"delta_db":-2.852907550973158}],"active_band_count":3,"active_band_mae_db":1.8703316795385438}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":3.2906454826540954,"rms_envelope_nrmse":0.1956406770907032,"envelope_correlation":0.980311962766619,"rms_gain_error_db":-0.9175141540257126,"reference_mean_power":8.438545283737647e-05,"output_mean_power":6.83150152917695e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":6.948180578959807e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.55589320089413e-05,"output_power":2.9481605330924723e-05,"reference_active":true,"delta_db":-0.8139758070877022},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.93670906958722e-06,"output_power":2.5937393530932073e-06,"reference_active":true,"delta_db":-1.8120698942727387},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579820903321111e-07,"output_power":6.378707127411013e-07,"reference_active":true,"delta_db":-0.7492628183635761}],"active_band_count":3,"active_band_mae_db":1.1251028399080056}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":2.8252046320802755,"rms_envelope_nrmse":0.15334764695978806,"envelope_correlation":0.9903720961310718,"rms_gain_error_db":-0.9175073847811894,"reference_mean_power":8.498589632737071e-05,"output_mean_power":6.880121705060692e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":6.685624503092089e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.278937461914447e-05,"output_power":2.7012983547280412e-05,"reference_active":true,"delta_db":-0.8416057952264151},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.678676767994248e-06,"output_power":9.807040277982105e-07,"reference_active":true,"delta_db":-2.334291210908032},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9241501206570436e-08,"output_power":5.843096643412012e-08,"reference_active":true,"delta_db":-0.05982985712338824}],"active_band_count":3,"active_band_mae_db":1.078575621085945}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":4.533662083858606,"rms_envelope_nrmse":0.2894587764503488,"envelope_correlation":0.9656560907706213,"rms_gain_error_db":-1.8679603501528566,"reference_mean_power":9.160796502339733e-06,"output_mean_power":5.9585035280545784e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":3.736107367401369e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6569760918264079e-06,"output_power":1.3066196842336089e-06,"reference_active":true,"delta_db":-1.0316704556499336},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.026752210402777e-07,"output_power":3.129982869012683e-07,"reference_active":true,"delta_db":-2.845413752631647},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8917938491313026e-07,"output_power":2.0257698658595836e-07,"reference_active":true,"delta_db":0.2971629744649249}],"active_band_count":3,"active_band_mae_db":1.3914157275821684}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":4.446051151523076,"rms_envelope_nrmse":0.2643779634726055,"envelope_correlation":0.9752529827401409,"rms_gain_error_db":-1.8679524904309859,"reference_mean_power":9.225978949789666e-06,"output_mean_power":6.000911339592244e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":3.5247411287487305e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.710468659980512e-06,"output_power":1.2989526383789514e-06,"reference_active":true,"delta_db":-1.195218048636826},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.274879265893732e-07,"output_power":1.6401849388990874e-07,"reference_active":true,"delta_db":-3.00302473961073},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9472537885262174e-08,"output_power":1.8737157333861406e-08,"reference_active":true,"delta_db":-0.16718853936229014}],"active_band_count":3,"active_band_mae_db":1.4551437758699486}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"golf","status":"measured","mos":{"status":"measured","value":3.2908573150634766,"predicted_mos":3.2908573150634766,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":101760,"input_sample_rate":24000,"input_seconds":4.24,"input_float32_sha256":"32d866991fe45d003238ed0c6fc77ec9e474772d55c5462012354e9a55e8310e","internal_samples":67840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1697087850006938},"started_at":"2026-10-13T05:01:51.586190+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0025065119998544105,"native_input_samples":102000,"native_input_sha256":"f8deefe411feed52bf888f873f97f6d1bda632b1fcaa298b9c4bcf5cc5171a25","output_sha256":"c6518a5f990211223f55841571ea36aee33be81562eee518170c430be2927452","output_samples":101760,"length_delta_samples":-240,"analysis_s":0.48264715200002684,"synthesis_s":0.10493263199987268,"total_s":0.5875797839998995,"rtf":0.13825406682350577,"output_peak":0.341322660446167,"output_rms":0.04142376614963272,"finished_at":"2026-10-13T05:01:52.181866+00:00","analysis_stages":{"external_f0_s":0.04154660499989404},"evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"c6518a5f990211223f55841571ea36aee33be81562eee518170c430be2927452","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":68000,"output_samples":67840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.652247894198272,"output_peak":0.3394279983211738},"common_prefix":{"samples":67840,"sample_rate":16000,"reference_coverage":0.9976470588235294,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":67840,"output_samples":67840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.131919564894902,"spectral_convergence":0.7926097746388098,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.693371673286205,"spectral_convergence":0.3347480464726944,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.8412361635227015,"spectral_convergence":0.26759373156208344,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006174987785523294,"output_power":0.0005318835834627343,"delta_db":-0.6481951680468643,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011084825617552847,"output_power":0.0008880064285174311,"delta_db":-0.963127555767575,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031285874547808357,"output_power":0.00021728994903804472,"delta_db":-1.5830866144442346,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.018710719680499e-05,"output_power":6.599500706380605e-05,"delta_db":-1.3563337722122806,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.90753237169182e-06,"output_power":1.05907763962667e-05,"delta_db":0.28962298832846484,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.607631139209343e-06,"output_power":3.253688931479341e-06,"delta_db":-2.3640340880225903,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2007333644703349,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9655941775649389,"pitch":null,"output_rms_dbfs":-27.652247894198272,"output_peak":0.3394279983211738,"reference_rms_dbfs":-26.686653716633334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.222175800567936},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":67840,"output_samples":67840,"intelligibility":{"estoi":{"value":0.8187641421968338,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04630920379895776,"rms_dbfs":-26.686653716633334,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.20306280134097,"mean":-2.053953566641178e-05,"zero_energy":false},"output":{"rms":0.04143693320468785,"rms_dbfs":-27.652247894198272,"peak":0.3394279983211738,"peak_dbfs":-9.385046739784958,"crest_factor_db":18.267201154413314,"mean":0.0006703021046959485,"zero_energy":false},"rms_difference_db":-0.9655941775649381},"full_scale":{"reference":{"sample_count":67840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":67840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":424,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":50,"quiet_samples":8000,"reference_mean_power":1.0623650159686805e-06,"output_mean_power":1.7561343946791398e-06,"output_rms_dbfs":-57.5544225116675,"output_zero_energy":false,"output_peak":0.0056983223548288985,"output_energy_fraction":0.0001206109354953557,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":417,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":142,"both_estimated_voiced_frames":121,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.983739837398374,"absolute_difference_cents_median":6.18853116313973,"absolute_difference_cents_p95":24.034952094213146,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":67840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":847,"rms_floor":0.0001318685794969069,"log_rms_mae_db":2.845380869323528,"rms_envelope_nrmse":0.15172011138832345,"envelope_correlation":0.9889825830931339,"rms_gain_error_db":-0.965583299209088,"reference_mean_power":0.002147067110001966,"output_mean_power":0.0017190451733294625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":847,"frequency_resolution_hz":0.23612750885478156,"power_floor":1.314473412334891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003578910792416915,"output_power":0.0002972606198832675,"reference_active":true,"delta_db":-0.8061349424277969},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.442365135066064e-05,"output_power":3.6233532192253594e-05,"reference_active":true,"delta_db":-0.8850357911965232},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.52488746886903e-06,"output_power":1.305757340192994e-05,"reference_active":true,"delta_db":2.39362466215787}],"active_band_count":3,"active_band_mae_db":1.36159846526073}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":841,"rms_floor":0.00012588557284836115,"log_rms_mae_db":2.020642791677405,"rms_envelope_nrmse":0.1287723155738333,"envelope_correlation":0.9947701884526114,"rms_gain_error_db":-0.9655901794572794,"reference_mean_power":0.0021622754087721194,"output_mean_power":0.0017312189259785652},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":841,"frequency_resolution_hz":0.23781212841854935,"power_floor":1.2603020282928996e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003152017659580686,"output_power":0.00026252324286080674,"reference_active":true,"delta_db":-0.7942088170369094},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9781382301322475e-05,"output_power":1.678995289663723e-05,"reference_active":true,"delta_db":-0.7120715859555266},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.592897819816882e-07,"output_power":9.685263371907673e-07,"reference_active":true,"delta_db":1.6703509013726017}],"active_band_count":3,"active_band_mae_db":1.0588771014550125}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":847,"rms_floor":0.00010444605382463925,"log_rms_mae_db":1.9789028490013254,"rms_envelope_nrmse":0.18278253061635077,"envelope_correlation":0.9850576758451652,"rms_gain_error_db":-1.0898681984995915,"reference_mean_power":0.0005617736983086631,"output_mean_power":0.0004370937355986215},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":847,"frequency_resolution_hz":0.23612750885478156,"power_floor":5.639997020623192e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000250136863396395,"output_power":0.00020195507468452362,"reference_active":true,"delta_db":-0.9292292918362439},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.809378778279824e-05,"output_power":3.0104906908023202e-05,"reference_active":true,"delta_db":-1.0221686945068378},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.770831514241446e-06,"output_power":4.958887084999081e-06,"reference_active":true,"delta_db":-1.3525778750483506}],"active_band_count":3,"active_band_mae_db":1.1013252871304775}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":841,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":1.5263907353727346,"rms_envelope_nrmse":0.1618763069254956,"envelope_correlation":0.990177423099993,"rms_gain_error_db":-1.0898679193481275,"reference_mean_power":0.0005657813733834166,"output_mean_power":0.0004402119762183472},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":841,"frequency_resolution_hz":0.23781212841854935,"power_floor":5.409158196049644e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002287657778740115,"output_power":0.0001849014361736342,"reference_active":true,"delta_db":-0.9245077242754558},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5872835313513352e-05,"output_power":1.2413324965230013e-05,"reference_active":true,"delta_db":-1.0676638548859598},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.40169960997678e-07,"output_power":4.335638785160434e-07,"reference_active":true,"delta_db":-1.6924219899300836}],"active_band_count":3,"active_band_mae_db":1.2281978563638332}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":847,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":1.838595200455937,"rms_envelope_nrmse":0.20727603791366775,"envelope_correlation":0.9853120299794629,"rms_gain_error_db":-1.4461592552791709,"reference_mean_power":8.458463381516759e-05,"output_mean_power":6.0628321924915825e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":847,"frequency_resolution_hz":0.23612750885478156,"power_floor":6.970843665669566e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.572875984429195e-05,"output_power":2.47330878109801e-05,"reference_active":true,"delta_db":-1.597396025832321},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.825216052225613e-06,"output_power":3.0176310916625995e-06,"reference_active":true,"delta_db":-1.0298982382985729},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.5449517277093e-07,"output_power":5.798879013148456e-07,"reference_active":true,"delta_db":-1.1431241772769316}],"active_band_count":3,"active_band_mae_db":1.256806147135942}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":841,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":1.3694736929334939,"rms_envelope_nrmse":0.1811439106691973,"envelope_correlation":0.9927166837091478,"rms_gain_error_db":-1.4461612851384764,"reference_mean_power":8.518793272185747e-05,"output_mean_power":6.106072417924748e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":841,"frequency_resolution_hz":0.23781212841854935,"power_floor":6.707680334681063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.288434560210087e-05,"output_power":2.2455223565280317e-05,"reference_active":true,"delta_db":-1.656718205518133},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.625785952089863e-06,"output_power":1.3551240910750374e-06,"reference_active":true,"delta_db":-0.7908430049577351},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9279335751842487e-08,"output_power":6.19467411056629e-08,"reference_active":true,"delta_db":0.1911513538877006}],"active_band_count":3,"active_band_mae_db":0.8795708547878562}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":847,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":2.755690204298338,"rms_envelope_nrmse":0.2251352696670231,"envelope_correlation":0.972852142918354,"rms_gain_error_db":-0.36887933196430517,"reference_mean_power":9.18241959550987e-06,"output_mean_power":8.43469146440551e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":847,"frequency_resolution_hz":0.23612750885478156,"power_floor":3.754155562103874e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.667065413077745e-06,"output_power":1.4253815573407853e-06,"reference_active":true,"delta_db":-0.6802150594007204},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.053378742489232e-07,"output_power":6.863688421050813e-07,"reference_active":true,"delta_db":0.5455971305668952},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8848540463329543e-07,"output_power":2.1121832708977676e-07,"reference_active":true,"delta_db":0.4945387230978786}],"active_band_count":3,"active_band_mae_db":0.5734503043551648}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":67840,"uncovered_tail_samples":0,"metrics":{"frame_count":841,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":2.568318209061805,"rms_envelope_nrmse":0.1404447943621144,"envelope_correlation":0.9911231552745526,"rms_gain_error_db":-0.3688767252371519,"reference_mean_power":9.24790886192181e-06,"output_mean_power":8.494853012229037e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":841,"frequency_resolution_hz":0.23781212841854935,"power_floor":3.542740386452164e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7215289461794163e-06,"output_power":1.5707482869891832e-06,"reference_active":true,"delta_db":-0.39807734668580874},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.27942884593343e-07,"output_power":3.6582117126818857e-07,"reference_active":true,"delta_db":0.4747062349712766},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9492974779317407e-08,"output_power":1.8513740284421315e-08,"reference_active":true,"delta_db":-0.22383953845906646}],"active_band_count":3,"active_band_mae_db":0.3655410400387173}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3535821437835693,"predicted_mos":3.3535821437835693,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":102120,"input_sample_rate":24000,"input_seconds":4.255,"input_float32_sha256":"aecab6cfef1b22d6d26dffdfc9a763ac034bbc43e6b521e7744f62a7195dd4de","internal_samples":68080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.044375945999491},"started_at":"2026-10-13T05:04:04.795027+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0020896980004181387,"native_input_samples":102000,"native_input_sha256":"f8deefe411feed52bf888f873f97f6d1bda632b1fcaa298b9c4bcf5cc5171a25","output_sha256":"091c9ab14fe7fa69ac4e11a39929fb29b88b33e8fd9b8fc9e29a7f78c28802eb","output_samples":102120,"length_delta_samples":120,"analysis_s":1.5741693729996769,"synthesis_s":8.933276271000068,"total_s":10.507445643999745,"rtf":2.4723401515293517,"output_peak":0.35908839106559753,"output_rms":0.04886819248997594,"finished_at":"2026-10-13T05:04:15.312472+00:00","analysis_stages":{"external_f0_s":1.0596630119998736},"evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"091c9ab14fe7fa69ac4e11a39929fb29b88b33e8fd9b8fc9e29a7f78c28802eb","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":68000,"output_samples":68080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.21720960887866,"output_peak":0.35933857256050356},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988249118683902,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.12478391519176,"spectral_convergence":0.8225656283237538,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.707809856611929,"spectral_convergence":0.35998229534581233,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.730098771353029,"spectral_convergence":0.2687053184997666,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0009569715276097793,"delta_db":1.9452216165988758,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0010516917851269235,"delta_db":-0.23544648252339967,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.0002890731418683037,"delta_db":-0.33422743551159756,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":7.996323279132637e-05,"delta_db":-0.5168472687043454,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":9.755689635445811e-06,"delta_db":-0.05622902507401217,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":4.701244888564643e-06,"delta_db":-0.7564295948029522,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6407335705358639,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.48477379041804447,"pitch":null,"output_rms_dbfs":-26.212103891447637,"output_peak":0.35933857256050356,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.187564181052239},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.8015567418083541,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046254726345655035,"rms_dbfs":-26.69687768186568,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.21328676657332,"mean":-2.2544412051930147e-05,"zero_energy":false},"output":{"rms":0.04890967820299316,"rms_dbfs":-26.212103891447637,"peak":0.35933857256050356,"peak_dbfs":-8.889923234473745,"crest_factor_db":17.32218065697389,"mean":5.618443691219863e-05,"zero_energy":false},"rms_difference_db":0.4847737904180427},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":51,"quiet_samples":8160,"reference_mean_power":1.0693246069565124e-06,"output_mean_power":1.7472262620660717e-05,"output_rms_dbfs":-47.576508511682015,"output_zero_energy":false,"output_peak":0.10765770813705708,"output_energy_fraction":0.0008764775246179942,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":138,"both_estimated_voiced_frames":120,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.975609756097561,"absolute_difference_cents_median":5.3051167190605195,"absolute_difference_cents_p95":25.672245933862442,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.0001318685794969069,"log_rms_mae_db":6.697170486904744,"rms_envelope_nrmse":0.14151274775908734,"envelope_correlation":0.9867567360254602,"rms_gain_error_db":0.48478162409360037,"reference_mean_power":0.0021420152585098387,"output_mean_power":0.0023949735562903566},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.3134511495070254e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003582591828573966,"output_power":0.0003740055120413456,"reference_active":true,"delta_db":0.18680671822143413},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.463795601086153e-05,"output_power":4.439321295811653e-05,"reference_active":true,"delta_db":-0.02387721654369567},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.556588167851325e-06,"output_power":1.0503966990023974e-05,"reference_active":true,"delta_db":1.430275943380391}],"active_band_count":3,"active_band_mae_db":0.5469866260485069}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00012588557284836115,"log_rms_mae_db":6.580000882558777,"rms_envelope_nrmse":0.10915583638677956,"envelope_correlation":0.9929663184102709,"rms_gain_error_db":0.48499381878219733,"reference_mean_power":0.0021571542863483557,"output_mean_power":0.0024120182539065587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.2592546923162062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031562831458090785,"output_power":0.0003271175136165015,"reference_active":true,"delta_db":0.1552784029105331},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.98767508506911e-05,"output_power":2.012682985857612e-05,"reference_active":true,"delta_db":0.054299813205865216},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602361817361657e-07,"output_power":8.427451754658164e-07,"reference_active":true,"delta_db":1.0599695445203259}],"active_band_count":3,"active_band_mae_db":0.4231825868789081}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00010444605382463925,"log_rms_mae_db":2.2097610203421985,"rms_envelope_nrmse":0.17562197731419413,"envelope_correlation":0.9798416884485228,"rms_gain_error_db":-0.4552444931170404,"reference_mean_power":0.0005604503758559807,"output_mean_power":0.000504676073447384},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":5.643164593552106e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025020135526255573,"output_power":0.00021595114662203172,"reference_active":true,"delta_db":-0.6393414346591342},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8332303804350734e-05,"output_power":2.93881970615067e-05,"reference_active":true,"delta_db":-1.153919781815112},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.791716397073577e-06,"output_power":6.514045862786538e-06,"reference_active":true,"delta_db":-0.181287307684848}],"active_band_count":3,"active_band_mae_db":0.6581828413863647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":1.9350838644245825,"rms_envelope_nrmse":0.12547467834323958,"envelope_correlation":0.990225048575661,"rms_gain_error_db":-0.4552438816623967,"reference_mean_power":0.0005644391167624198,"output_mean_power":0.0005082679386533382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":5.412619397247975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022898404327685652,"output_power":0.00019701583621062525,"reference_active":true,"delta_db":-0.653040833136465},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5949246143041614e-05,"output_power":1.240105912661341e-05,"reference_active":true,"delta_db":-1.0928138235599814},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.408378605310093e-07,"output_power":5.365921950606352e-07,"reference_active":true,"delta_db":-0.7710381043508364}],"active_band_count":3,"active_band_mae_db":0.8389642536824277}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":2.1964731841513045,"rms_envelope_nrmse":0.19214697927890576,"envelope_correlation":0.9769094575074421,"rms_gain_error_db":-0.4635077767591015,"reference_mean_power":8.438545283737647e-05,"output_mean_power":7.584322776808602e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":6.948180578959807e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.55589320089413e-05,"output_power":3.2689564549718814e-05,"reference_active":true,"delta_db":-0.3653957324091198},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.93670906958722e-06,"output_power":3.69444656389554e-06,"reference_active":true,"delta_db":-0.275839314257024},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579820903321111e-07,"output_power":5.807348012046056e-07,"reference_active":true,"delta_db":-1.1568109179257189}],"active_band_count":3,"active_band_mae_db":0.5993486548639543}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":1.4523305435917266,"rms_envelope_nrmse":0.13656322048485803,"envelope_correlation":0.9888220546469756,"rms_gain_error_db":-0.46350700562088043,"reference_mean_power":8.498589632737071e-05,"output_mean_power":7.638290273967457e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":6.685624503092089e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.278937461914447e-05,"output_power":3.0131879896036874e-05,"reference_active":true,"delta_db":-0.3670690590305351},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.678676767994248e-06,"output_power":1.6022893429683577e-06,"reference_active":true,"delta_db":-0.20226135912463594},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9241501206570436e-08,"output_power":5.089401089612245e-08,"reference_active":true,"delta_db":-0.659593767575317}],"active_band_count":3,"active_band_mae_db":0.409641395243496}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":1.7540219493773697,"rms_envelope_nrmse":0.25379636595258626,"envelope_correlation":0.9610414651707009,"rms_gain_error_db":-0.12221754124888326,"reference_mean_power":9.160796502339733e-06,"output_mean_power":8.906590429943924e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":3.736107367401369e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6569760918264079e-06,"output_power":1.7611754068941315e-06,"reference_active":true,"delta_db":0.2648637021162104},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.026752210402777e-07,"output_power":6.863004236745928e-07,"reference_active":true,"delta_db":0.5643093115383755},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8917938491313026e-07,"output_power":2.1923308704041113e-07,"reference_active":true,"delta_db":0.640322901466483}],"active_band_count":3,"active_band_mae_db":0.489831971707023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":1.0438156678604178,"rms_envelope_nrmse":0.16261880531385883,"envelope_correlation":0.9834511558377592,"rms_gain_error_db":-0.12221606678645341,"reference_mean_power":9.225978949789666e-06,"output_mean_power":8.969967152757464e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":3.5247411287487305e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.710468659980512e-06,"output_power":1.8513657612121449e-06,"reference_active":true,"delta_db":0.3437710646817036},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.274879265893732e-07,"output_power":3.728085693871511e-07,"reference_active":true,"delta_db":0.5629059289723242},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9472537885262174e-08,"output_power":2.3107909674903446e-08,"reference_active":true,"delta_db":0.7433810382111787}],"active_band_count":3,"active_band_mae_db":0.5500193439550688}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.957899332046509,"predicted_mos":2.957899332046509,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":102240,"input_sample_rate":24000,"input_seconds":4.26,"input_float32_sha256":"20c22bb17167e26ec15e3569e5b1648c4e2d4f4740cc6e4a2b9aa17357a490dc","internal_samples":68160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.159259796000697},"started_at":"2026-10-13T05:09:57.408577+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0046513230008713435,"native_input_samples":102000,"native_input_sha256":"f8deefe411feed52bf888f873f97f6d1bda632b1fcaa298b9c4bcf5cc5171a25","output_sha256":"c7a484426455756c6584d12af5f49057de31f27e54831b4f4aa89d4dbd599d99","output_samples":102240,"length_delta_samples":240,"analysis_s":0.6017643169998337,"synthesis_s":0.0751736410002195,"total_s":0.6769379580000532,"rtf":0.1592795195294243,"output_peak":0.3135829567909241,"output_rms":0.04559674183889903,"finished_at":"2026-10-13T05:09:58.097498+00:00","analysis_stages":{"external_f0_s":0.06275827700119407},"evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"c7a484426455756c6584d12af5f49057de31f27e54831b4f4aa89d4dbd599d99","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":68000,"output_samples":68160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.818740536603446,"output_peak":0.3134505743226399},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9976525821596244,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.15454586788694,"spectral_convergence":0.16738740209823794,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.723727735770934,"spectral_convergence":0.1610219217450809,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.110296706390106,"spectral_convergence":0.17022948742961366,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0005881650757428893,"delta_db":-0.16877621661882325,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0011148854541742582,"delta_db":0.017971194238864906,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.00028699867303838615,"delta_db":-0.36550597533045437,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":8.35406005690922e-05,"delta_db":-0.32677477472711774,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":7.313414607626428e-06,"delta_db":-1.3076068297658279,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":4.291486317035497e-06,"delta_db":-1.152481015640142,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5565193343868716,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11165618489283552,"pitch":null,"output_rms_dbfs":-26.808533866758516,"output_peak":0.3134505743226399,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.996190103349327},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.8885513133960795,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046254726345655035,"rms_dbfs":-26.69687768186568,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.21328676657332,"mean":-2.2544412051930147e-05,"zero_energy":false},"output":{"rms":0.04566393220528964,"rms_dbfs":-26.808533866758516,"peak":0.3134505743226399,"peak_dbfs":-10.076618601686292,"crest_factor_db":16.731915265072224,"mean":4.297957776974373e-05,"zero_energy":false},"rms_difference_db":-0.11165618489283702},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":51,"quiet_samples":8160,"reference_mean_power":1.0693246069565124e-06,"output_mean_power":2.888367633118509e-06,"output_rms_dbfs":-55.393475303290515,"output_zero_energy":false,"output_peak":0.05837683732153551,"output_energy_fraction":0.000166221463748519,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":120,"both_estimated_voiced_frames":115,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9349593495934959,"absolute_difference_cents_median":2.3728298560018644,"absolute_difference_cents_p95":8.756269798188862,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.0001318685794969069,"log_rms_mae_db":1.7729009099595219,"rms_envelope_nrmse":0.06851193634139503,"envelope_correlation":0.9961409352161642,"rms_gain_error_db":-0.11166164230579798,"reference_mean_power":0.0021420152585098387,"output_mean_power":0.002087643780347709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.3134511495070254e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003582591828573966,"output_power":0.000337181290068024,"reference_active":true,"delta_db":-0.26333863044770817},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.463795601086153e-05,"output_power":3.444920851227414e-05,"reference_active":true,"delta_db":-1.1252505164758773},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.556588167851325e-06,"output_power":7.116389300596324e-06,"reference_active":true,"delta_db":-0.2606605587199452}],"active_band_count":3,"active_band_mae_db":0.5497499018811768}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00012588557284836115,"log_rms_mae_db":1.1448557842398497,"rms_envelope_nrmse":0.03637691375379991,"envelope_correlation":0.9990070709463359,"rms_gain_error_db":-0.11162191873327756,"reference_mean_power":0.0021571542863483557,"output_mean_power":0.0021024177593377204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.2592546923162062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031562831458090785,"output_power":0.0002979689452391577,"reference_active":true,"delta_db":-0.2500495261777634},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.98767508506911e-05,"output_power":1.5223449429459824e-05,"reference_active":true,"delta_db":-1.1583232510823465},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602361817361657e-07,"output_power":4.256867361883618e-07,"reference_active":true,"delta_db":-1.9060920213511965}],"active_band_count":3,"active_band_mae_db":1.104821599537102}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00010444605382463925,"log_rms_mae_db":1.330896418029636,"rms_envelope_nrmse":0.09567479363010398,"envelope_correlation":0.9941096122484968,"rms_gain_error_db":-0.22848236774542946,"reference_mean_power":0.0005604503758559807,"output_mean_power":0.000531727262654474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":5.643164593552106e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025020135526255573,"output_power":0.00022649941255629754,"reference_active":true,"delta_db":-0.4322257782720391},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8332303804350734e-05,"output_power":3.233352504975077e-05,"reference_active":true,"delta_db":-0.739118668765223},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.791716397073577e-06,"output_power":7.09794651864656e-06,"reference_active":true,"delta_db":0.1915318001586544}],"active_band_count":3,"active_band_mae_db":0.4542920823986389}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":0.8294996206758791,"rms_envelope_nrmse":0.05626263835205413,"envelope_correlation":0.9981957836929487,"rms_gain_error_db":-0.22848569196263072,"reference_mean_power":0.0005644391167624198,"output_mean_power":0.0005355111704696109},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":5.412619397247975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022898404327685652,"output_power":0.0002087696783418372,"reference_active":true,"delta_db":-0.4013779755635598},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5949246143041614e-05,"output_power":1.4227360718901773e-05,"reference_active":true,"delta_db":-0.4961581781472219},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.408378605310093e-07,"output_power":5.084066657049858e-07,"reference_active":true,"delta_db":-1.0053692589546839}],"active_band_count":3,"active_band_mae_db":0.6343018042218219}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":1.2902857753063315,"rms_envelope_nrmse":0.11674497338949855,"envelope_correlation":0.9918771519747221,"rms_gain_error_db":-0.3277435510734675,"reference_mean_power":8.438545283737647e-05,"output_mean_power":7.825160070436646e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":6.948180578959807e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.55589320089413e-05,"output_power":3.222234508356424e-05,"reference_active":true,"delta_db":-0.42791564326944664},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.93670906958722e-06,"output_power":3.876736931861774e-06,"reference_active":true,"delta_db":-0.06666989180628856},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579820903321111e-07,"output_power":5.797455227920732e-07,"reference_active":true,"delta_db":-1.1642154081323848}],"active_band_count":3,"active_band_mae_db":0.5529336477360399}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":0.7960684791232951,"rms_envelope_nrmse":0.06591710204186618,"envelope_correlation":0.9979639159268962,"rms_gain_error_db":-0.3277456806324581,"reference_mean_power":8.498589632737071e-05,"output_mean_power":7.88083602130942e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":6.685624503092089e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.278937461914447e-05,"output_power":3.008575279717213e-05,"reference_active":true,"delta_db":-0.3737225083693369},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.678676767994248e-06,"output_power":1.6225586154231576e-06,"reference_active":true,"delta_db":-0.1476668528668425},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9241501206570436e-08,"output_power":3.2731825908136976e-08,"reference_active":true,"delta_db":-2.576558225946652}],"active_band_count":3,"active_band_mae_db":1.0326491957276105}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":1.446399965163459,"rms_envelope_nrmse":0.16975875567763982,"envelope_correlation":0.9887250716119947,"rms_gain_error_db":-1.0007608780215937,"reference_mean_power":9.160796502339733e-06,"output_mean_power":7.275404561730221e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":3.736107367401369e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6569760918264079e-06,"output_power":1.4473749229845187e-06,"reference_active":true,"delta_db":-0.5873519828735426},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.026752210402777e-07,"output_power":4.6275688124210416e-07,"reference_active":true,"delta_db":-1.147304503641293},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8917938491313026e-07,"output_power":9.968064123367411e-08,"reference_active":true,"delta_db":-2.782629860053019}],"active_band_count":3,"active_band_mae_db":1.5057621155226182}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":1.1094096513591873,"rms_envelope_nrmse":0.1285654866844204,"envelope_correlation":0.996819732246099,"rms_gain_error_db":-1.0007741318907384,"reference_mean_power":9.225978949789666e-06,"output_mean_power":7.327149388665733e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":3.5247411287487305e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.710468659980512e-06,"output_power":1.4589800044678318e-06,"reference_active":true,"delta_db":-0.6906578137864089},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.274879265893732e-07,"output_power":2.3498911172938578e-07,"reference_active":true,"delta_db":-1.4414755399285248},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9472537885262174e-08,"output_power":7.248717664965486e-09,"reference_active":true,"delta_db":-4.291613730019253}],"active_band_count":3,"active_band_mae_db":2.1412490279113956}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0031","method":"magphase","status":"measured","mos":{"status":"measured","value":4.002053737640381,"predicted_mos":4.002053737640381,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":205627,"input_sample_rate":48000,"input_seconds":4.283895833333333,"input_float32_sha256":"887a3e39e53a9cf884beeff7ec27c2300b690d09e82b2b649698fb67307e2940","internal_samples":68543,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2149083860003884},"started_at":"2026-10-13T05:12:02.229197+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.0075149339991185116,"native_input_samples":204000,"native_input_sha256":"320e52707496d6770bc7f1b180f9cdb40b4dc58458b00611d36cec4ffb1fc89d","output_sha256":"611c684dd0c5f4d73bb79f93137466a11eceabb8642c16e656d64d04c3da6a19","output_samples":205627,"length_delta_samples":1627,"analysis_s":2.167552650000289,"synthesis_s":0.7645082529998035,"total_s":2.9320609030000924,"rtf":0.6898966830588452,"output_peak":0.3030546277730927,"output_rms":0.04467221981944794,"finished_at":"2026-10-13T05:12:05.178939+00:00","evaluation_input":{"reference":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"611c684dd0c5f4d73bb79f93137466a11eceabb8642c16e656d64d04c3da6a19","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":68000,"output_samples":68543,"length_delta_samples":543,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.996790788164923,"output_peak":0.3031968019464887},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9920779656565951,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.197978478960827,"spectral_convergence":0.7017975243816311,"spectral_convergence_status":"ok","amplitude_floor":1.0380731662877042e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.336407306092417,"spectral_convergence":0.41703483158379656,"spectral_convergence_status":"ok","amplitude_floor":5.867181533968612e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.665171729022004,"spectral_convergence":0.3816019762281574,"spectral_convergence_status":"ok","amplitude_floor":4.6154090777751923e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006114723941341076,"output_power":0.0005473963691064617,"delta_db":-0.4807494252184749,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011102815674838913,"output_power":0.0010687080091998005,"delta_db":-0.16574065737473287,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00031219826577364904,"output_power":0.0002938252243374736,"delta_db":-0.26341409864824594,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.006894391321554e-05,"output_power":8.876147747467303e-05,"delta_db":-0.0635054788687335,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.882819827254369e-06,"output_power":9.487158310827654e-06,"delta_db":-0.1774473020595227,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.595718179316582e-06,"output_power":4.495091131871418e-06,"delta_db":-0.951173328340547,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.35033838175170945,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2653894012483309,"pitch":null,"output_rms_dbfs":-26.962267083114014,"output_peak":0.3031968019464887,"reference_rms_dbfs":-26.69687768186568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.39985250469175},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.6140074049230878,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.046254726345655035,"rms_dbfs":-26.69687768186568,"peak":0.299102783203125,"peak_dbfs":-10.483590915292362,"crest_factor_db":16.21328676657332,"mean":-2.2544412051930147e-05,"zero_energy":false},"output":{"rms":0.04486282792648172,"rms_dbfs":-26.962267083114014,"peak":0.3031968019464887,"peak_dbfs":-10.365507677179476,"crest_factor_db":16.596759405934538,"mean":-0.000409234428569138,"zero_energy":false},"rms_difference_db":-0.2653894012483349},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.738932225853205e-06,"threshold_dbfs":-57.597173441108865,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":51,"quiet_samples":8160,"reference_mean_power":1.0693246069565124e-06,"output_mean_power":1.8188877471229026e-06,"output_rms_dbfs":-57.401941026459184,"output_zero_energy":false,"output_peak":0.007503482337333065,"output_energy_fraction":0.00010844607838190237,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":123,"output_estimated_voiced_frames":123,"both_estimated_voiced_frames":109,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.8861788617886179,"absolute_difference_cents_median":7.774513664219086,"absolute_difference_cents_p95":36.60782887652639,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.0001318685794969069,"log_rms_mae_db":3.9666094440619513,"rms_envelope_nrmse":0.272416670053466,"envelope_correlation":0.9367907773272521,"rms_gain_error_db":-0.26540157987523993,"reference_mean_power":0.0021420152585098387,"output_mean_power":0.00201503411008998},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.3134511495070254e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003582591828573966,"output_power":0.00033906248134177904,"reference_active":true,"delta_db":-0.23917595031407388},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.463795601086153e-05,"output_power":4.103725297305048e-05,"reference_active":true,"delta_db":-0.3652601839808855},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.556588167851325e-06,"output_power":1.0190858399075542e-05,"reference_active":true,"delta_db":1.2988501323162949}],"active_band_count":3,"active_band_mae_db":0.6344287555370848}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00012588557284836115,"log_rms_mae_db":2.8869693729098813,"rms_envelope_nrmse":0.21019794088461538,"envelope_correlation":0.9612347774754131,"rms_gain_error_db":-0.26531516592589366,"reference_mean_power":0.0021571542863483557,"output_mean_power":0.0020293160568131386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.2592546923162062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031562831458090785,"output_power":0.00029976374156986013,"reference_active":true,"delta_db":-0.22396855378460698},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.98767508506911e-05,"output_power":1.9298741285964375e-05,"reference_active":true,"delta_db":-0.1281640988058719},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.602361817361657e-07,"output_power":7.738490861772203e-07,"reference_active":true,"delta_db":0.6895695357374965}],"active_band_count":3,"active_band_mae_db":0.3472340627759918}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00010444605382463925,"log_rms_mae_db":3.703621228335867,"rms_envelope_nrmse":0.4032978184500166,"envelope_correlation":0.8877516589543756,"rms_gain_error_db":-0.20827887027042846,"reference_mean_power":0.0005604503758559807,"output_mean_power":0.0005342066349315307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":5.643164593552106e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025020135526255573,"output_power":0.0002305535726918312,"reference_active":true,"delta_db":-0.355178013029835},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8332303804350734e-05,"output_power":3.4447487817852794e-05,"reference_active":true,"delta_db":-0.46407366285749607},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.791716397073577e-06,"output_power":5.3770604733833915e-06,"reference_active":true,"delta_db":-1.0143462188200718}],"active_band_count":3,"active_band_mae_db":0.611199298235801}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":9.32306023645191e-05,"log_rms_mae_db":3.024564294667356,"rms_envelope_nrmse":0.30691817354166556,"envelope_correlation":0.9323887287321629,"rms_gain_error_db":-0.2082785564639711,"reference_mean_power":0.0005644391167624198,"output_mean_power":0.0005380086372788673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":5.412619397247975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022898404327685652,"output_power":0.00021071969261716645,"reference_active":true,"delta_db":-0.36100095539655574},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5949246143041614e-05,"output_power":1.4348444898481884e-05,"reference_active":true,"delta_db":-0.45935326250089403},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.408378605310093e-07,"output_power":5.698921591004576e-07,"reference_active":true,"delta_db":-0.5095548009192783}],"active_band_count":3,"active_band_mae_db":0.4433030062722427}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":4.754025004833883e-05,"log_rms_mae_db":3.6240954297938828,"rms_envelope_nrmse":0.37177171805208525,"envelope_correlation":0.9113700419971464,"rms_gain_error_db":-0.06394071623652747,"reference_mean_power":8.438545283737647e-05,"output_mean_power":8.315215588763637e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":6.948180578959807e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.55589320089413e-05,"output_power":3.559935820276994e-05,"reference_active":true,"delta_db":0.0049345971469404466},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.93670906958722e-06,"output_power":3.7842037998780136e-06,"reference_active":true,"delta_db":-0.1715880330358152},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.579820903321111e-07,"output_power":5.813137191062902e-07,"reference_active":true,"delta_db":-1.152483717011433}],"active_band_count":3,"active_band_mae_db":0.4430021157313962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.260586736236331e-05,"log_rms_mae_db":2.7611739608157615,"rms_envelope_nrmse":0.2904294985102393,"envelope_correlation":0.94377757160927,"rms_gain_error_db":-0.0639368818484346,"reference_mean_power":8.498589632737071e-05,"output_mean_power":8.374389780821342e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":6.685624503092089e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.278937461914447e-05,"output_power":3.2891225084601335e-05,"reference_active":true,"delta_db":0.013469161603820567},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.678676767994248e-06,"output_power":1.6535537320116419e-06,"reference_active":true,"delta_db":-0.06548768241402032},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9241501206570436e-08,"output_power":4.5489288153258426e-08,"reference_active":true,"delta_db":-1.1471691445498453}],"active_band_count":3,"active_band_mae_db":0.4087086628558954}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":2.78037949707682e-05,"log_rms_mae_db":3.4950590516810407,"rms_envelope_nrmse":0.40432120356515444,"envelope_correlation":0.897989756143923,"rms_gain_error_db":-0.43524243772826876,"reference_mean_power":9.160796502339733e-06,"output_mean_power":8.287222364304994e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":3.736107367401369e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6569760918264079e-06,"output_power":1.4663387314747391e-06,"reference_active":true,"delta_db":-0.5308193605652662},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.026752210402777e-07,"output_power":4.684666638806767e-07,"reference_active":true,"delta_db":-1.094046439416342},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8917938491313026e-07,"output_power":1.8587324253398523e-07,"reference_active":true,"delta_db":-0.07656933879505114}],"active_band_count":3,"active_band_mae_db":0.5671450462588865}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":2.0585812044473273e-05,"log_rms_mae_db":2.449168695633627,"rms_envelope_nrmse":0.2866440808577207,"envelope_correlation":0.9476269802934386,"rms_gain_error_db":-0.4352447923540186,"reference_mean_power":9.225978949789666e-06,"output_mean_power":8.34618448433988e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":3.5247411287487305e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.710468659980512e-06,"output_power":1.4845277144776557e-06,"reference_active":true,"delta_db":-0.6152681144251083},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.274879265893732e-07,"output_power":2.4825234708047073e-07,"reference_active":true,"delta_db":-1.2030193041665949},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9472537885262174e-08,"output_power":1.639294661473592e-08,"reference_active":true,"delta_db":-0.7476553290543707}],"active_band_count":3,"active_band_mae_db":0.855314249215358}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"hnm","status":"measured","mos":{"status":"measured","value":2.887166738510132,"predicted_mos":2.887166738510132,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bf9fcac5dfab5fb90021f0ab842f0f0e35de9d61e0f3c6b13ffb9df85f5bd888","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.989717622000171},"started_at":"2026-10-13T04:38:01.929143+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.362999955745181e-05,"native_input_samples":96000,"native_input_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","output_sha256":"d81fbe6c317cc0289135233f1d4cc74537218aea349579cbe4274dccc005ed01","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2019986100003734,"synthesis_s":0.5576380390002669,"total_s":1.7596366490006403,"rtf":0.29327277483344005,"output_peak":0.4379455572529207,"output_rms":0.05150541656857222,"finished_at":"2026-10-13T04:38:03.693620+00:00","evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"d81fbe6c317cc0289135233f1d4cc74537218aea349579cbe4274dccc005ed01","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.816741386029486,"spectral_convergence":0.1799962035530116,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.37511915428396,"spectral_convergence":0.16941598293718704,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.499066863211654,"spectral_convergence":0.1765319247932727,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007484888253859053,"delta_db":0.04011120304671298,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.001329410650706204,"delta_db":-0.07326458487243949,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.00035593387187984416,"delta_db":-0.1882742788749982,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00018341056673468518,"delta_db":-0.13110051073519224,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.4743593489625579e-05,"delta_db":-0.19689476856971175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":2.082042770588966e-05,"delta_db":-0.0402096683448883,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11164250240732382,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06177562974120265,"pitch":null,"output_rms_dbfs":-25.762941919314795,"output_peak":0.4379455572529207,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5636424678417},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.816741386029486,"spectral_convergence":0.1799962035530116,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.37511915428396,"spectral_convergence":0.16941598293718704,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.499066863211654,"spectral_convergence":0.1765319247932727,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007484888253859053,"delta_db":0.04011120304671298,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.001329410650706204,"delta_db":-0.07326458487243949,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.00035593387187984416,"delta_db":-0.1882742788749982,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00018341056673468518,"delta_db":-0.13110051073519224,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.4743593489625579e-05,"delta_db":-0.19689476856971175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":2.082042770588966e-05,"delta_db":-0.0402096683448883,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11164250240732382,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06177562974120265,"pitch":null,"output_rms_dbfs":-25.762941919314795,"output_peak":0.4379455572529207,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5636424678417},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8923662772085076,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051873038222359884,"rms_dbfs":-25.701166289573592,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.345694075201205,"mean":8.414904276529947e-06,"zero_energy":false},"output":{"rms":0.05150541656857222,"rms_dbfs":-25.762941919314795,"peak":0.4379455572529207,"peak_dbfs":-7.171597500156659,"crest_factor_db":18.591344419158137,"mean":3.5842739540973626e-05,"zero_energy":false},"rms_difference_db":-0.06177562974120221},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":1.54155072078312e-06,"output_rms_dbfs":-58.12042181353406,"output_zero_energy":false,"output_peak":0.006253495791714839,"output_energy_fraction":0.00013074784174235639,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":196,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9786096256684492,"absolute_difference_cents_median":1.058735344598144,"absolute_difference_cents_p95":6.490849583570328,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018114550052554645,"log_rms_mae_db":1.149398053472874,"rms_envelope_nrmse":0.05578397580630989,"envelope_correlation":0.9968796272861127,"rms_gain_error_db":-0.06150705499604715,"reference_mean_power":0.002692702060931978,"output_mean_power":0.002654835383386508},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0110233554726448e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033858489572855915,"output_power":0.00033289500895860853,"reference_active":true,"delta_db":-0.07360296431093585},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.201391758025471e-05,"output_power":3.923447102726454e-05,"reference_active":true,"delta_db":-0.29725377461651703},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.156677750109103e-06,"output_power":7.13544865017095e-06,"reference_active":true,"delta_db":-0.012901771900475105}],"active_band_count":3,"active_band_mae_db":0.12791950360930932}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015936923488846764,"log_rms_mae_db":0.6263679081600846,"rms_envelope_nrmse":0.03213359962575855,"envelope_correlation":0.9989474058472373,"rms_gain_error_db":-0.058962363420662564,"reference_mean_power":0.002702794865208283,"output_mean_power":0.0026663481096202125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.583684100944911e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031173336044584235,"output_power":0.00030648961743646005,"reference_active":true,"delta_db":-0.07367514150864121},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.830633609453067e-05,"output_power":1.7437421670017072e-05,"reference_active":true,"delta_db":-0.21119161769215744},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.642076702713704e-07,"output_power":5.014440288779484e-07,"reference_active":true,"delta_db":0.3351015197984262}],"active_band_count":3,"active_band_mae_db":0.2066560929997416}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001259260583850441,"log_rms_mae_db":1.7751692029014163,"rms_envelope_nrmse":0.09223808773385099,"envelope_correlation":0.99378023856796,"rms_gain_error_db":-0.11970654647541287,"reference_mean_power":0.0007105481356967998,"output_mean_power":0.0006912304313395397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8068878391559104e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001608363031836456,"output_power":0.00015405020260126652,"reference_active":true,"delta_db":-0.18721808481564645},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.800835490835068e-05,"output_power":1.6871327876544262e-05,"reference_active":true,"delta_db":-0.2832477558253654},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.325041103743567e-06,"output_power":2.090648101749758e-06,"reference_active":true,"delta_db":-0.4614969661627228}],"active_band_count":3,"active_band_mae_db":0.3106542689345782}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010786919997015104,"log_rms_mae_db":1.2951382315113356,"rms_envelope_nrmse":0.05729546332687004,"envelope_correlation":0.9976189354930853,"rms_gain_error_db":-0.12147725647120071,"reference_mean_power":0.0007140165693307253,"output_mean_power":0.0006943214213377095},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6527252040517484e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015126423656686204,"output_power":0.00014573048829131458,"reference_active":true,"delta_db":-0.1618583975862353},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.688501064647705e-06,"output_power":7.492389710065033e-06,"reference_active":true,"delta_db":-0.1122132010205798},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2694181944274497e-07,"output_power":2.027789126699378e-07,"reference_active":true,"delta_db":-0.48891742400943544}],"active_band_count":3,"active_band_mae_db":0.25432967420541686}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":1.6975973456527695,"rms_envelope_nrmse":0.11169205857629992,"envelope_correlation":0.9918638175899411,"rms_gain_error_db":-0.173851458277426,"reference_mean_power":0.0001538181527968681,"output_mean_power":0.00014778230829905905},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.418336470516265e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.474851415047919e-05,"output_power":2.2922814439100753e-05,"reference_active":true,"delta_db":-0.33281191121495424},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.277056498060546e-06,"output_power":2.707268112335332e-06,"reference_active":true,"delta_db":-0.8295266095706489},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.913659986768022e-07,"output_power":5.823926142354429e-07,"reference_active":true,"delta_db":-1.848402051010236}],"active_band_count":3,"active_band_mae_db":1.003580190598613}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":1.0711870926154066,"rms_envelope_nrmse":0.05828542215773472,"envelope_correlation":0.9979131410246762,"rms_gain_error_db":-0.17390654200809125,"reference_mean_power":0.00015458728634615255,"output_mean_power":0.00014851937718982696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.197451137326506e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3402826997507212e-05,"output_power":2.1424292778292712e-05,"reference_active":true,"delta_db":-0.3836182751148251},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.382612079388441e-06,"output_power":1.1586385888514253e-06,"reference_active":true,"delta_db":-0.7675235810990173},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.610921131466672e-08,"output_power":3.146569312343672e-08,"reference_active":true,"delta_db":-1.6595039100732754}],"active_band_count":3,"active_band_mae_db":0.936881922095706}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":1.890157633542356,"rms_envelope_nrmse":0.1719580979208887,"envelope_correlation":0.9812158815882759,"rms_gain_error_db":-0.14842353182321028,"reference_mean_power":2.0928545188569674e-05,"output_mean_power":2.0225379859109818e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.488290124260659e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1348414996445457e-05,"output_power":1.1094693379133468e-05,"reference_active":true,"delta_db":-0.09819904724348757},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8308296968161897e-06,"output_power":2.161218602217104e-06,"reference_active":true,"delta_db":-1.1721504594481116},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.670421987320987e-07,"output_power":3.013774574448189e-07,"reference_active":true,"delta_db":-3.4504254466809323}],"active_band_count":3,"active_band_mae_db":1.5735916511241772}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":1.3040290372538226,"rms_envelope_nrmse":0.09045607401929841,"envelope_correlation":0.9947185491734853,"rms_gain_error_db":-0.1484685585312575,"reference_mean_power":2.1033540736202732e-05,"output_mean_power":2.0326636982609222e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3907968246524526e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1287290435083048e-05,"output_power":1.0847758345554204e-05,"reference_active":true,"delta_db":-0.17249698194211105},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5147848321624166e-06,"output_power":1.107588089925302e-06,"reference_active":true,"delta_db":-1.3597267073523025},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4338124913221415e-08,"output_power":1.7735926987111445e-08,"reference_active":true,"delta_db":-3.9792343031929587}],"active_band_count":3,"active_band_mae_db":1.8371526641624574}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9190213680267334,"predicted_mos":2.9190213680267334,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"0b8f05ab0c0465dcbd223a89735ef8cd1c2a05f3c49eecbb80f0ed80fdb69652","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9238470820000657},"started_at":"2026-10-13T04:40:53.231394+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5323000045318622e-05,"native_input_samples":96000,"native_input_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","output_sha256":"34fad5044d28d529c8417da9bde3e459917960a179ab822229c7a95e41645721","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5542487970005823,"synthesis_s":0.0841276339997421,"total_s":1.6383764310003244,"rtf":0.2730627385000541,"output_peak":0.6797615360485552,"output_rms":0.05555320209993551,"finished_at":"2026-10-13T04:40:54.875757+00:00","evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"34fad5044d28d529c8417da9bde3e459917960a179ab822229c7a95e41645721","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.972413307127161,"spectral_convergence":0.8310824950106358,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.451875331066102,"spectral_convergence":0.3008664488769026,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.689418978852883,"spectral_convergence":0.242562590160484,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007557064841665593,"delta_db":0.08178948003965997,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.0016333119264510532,"delta_db":0.8208352096464049,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.00043699341043020165,"delta_db":0.7027814134762029,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00022154970110680667,"delta_db":0.6893676455684101,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.648501852928724e-05,"delta_db":0.2879661476459872,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":2.211172287237007e-05,"delta_db":0.22111969045789315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.46730993113909297,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5953482263094814,"pitch":null,"output_rms_dbfs":-25.10581806326411,"output_peak":0.6797615360485552,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.0379025390153815},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.972413307127161,"spectral_convergence":0.8310824950106358,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.451875331066102,"spectral_convergence":0.3008664488769026,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.689418978852883,"spectral_convergence":0.242562590160484,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007557064841665593,"delta_db":0.08178948003965997,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.0016333119264510532,"delta_db":0.8208352096464049,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.00043699341043020165,"delta_db":0.7027814134762029,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00022154970110680667,"delta_db":0.6893676455684101,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.648501852928724e-05,"delta_db":0.2879661476459872,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":2.211172287237007e-05,"delta_db":0.22111969045789315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.46730993113909297,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5953482263094814,"pitch":null,"output_rms_dbfs":-25.10581806326411,"output_peak":0.6797615360485552,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.0379025390153815},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9128532385096394,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051873038222359884,"rms_dbfs":-25.701166289573592,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.345694075201205,"mean":8.414904276529947e-06,"zero_energy":false},"output":{"rms":0.0555532020999355,"rms_dbfs":-25.10581806326411,"peak":0.6797615360485552,"peak_dbfs":-3.3528682676825605,"crest_factor_db":21.75294979558155,"mean":5.959173892465958e-07,"zero_energy":false},"rms_difference_db":0.5953482263094827},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":7.311335800903616e-07,"output_rms_dbfs":-61.36003269011717,"output_zero_energy":false,"output_peak":0.010237827015580861,"output_energy_fraction":5.33041540555237e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":192,"both_estimated_voiced_frames":172,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9197860962566845,"absolute_difference_cents_median":6.395604983152124,"absolute_difference_cents_p95":18.52002829410911,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018114550052554645,"log_rms_mae_db":3.0243551585415136,"rms_envelope_nrmse":0.13023031278370426,"envelope_correlation":0.9889117538853048,"rms_gain_error_db":0.5958602920749407,"reference_mean_power":0.002692702060931978,"output_mean_power":0.0030886900706730736},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0110233554726448e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033858489572855915,"output_power":0.0003916696137748255,"reference_active":true,"delta_db":0.6325229940912225},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.201391758025471e-05,"output_power":4.59792469824529e-05,"reference_active":true,"delta_db":0.3916867526747225},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.156677750109103e-06,"output_power":7.764481809116701e-06,"reference_active":true,"delta_db":0.3540101451008013}],"active_band_count":3,"active_band_mae_db":0.4594066306222488}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015936923488846764,"log_rms_mae_db":2.7034858418192838,"rms_envelope_nrmse":0.09906240960673428,"envelope_correlation":0.9954631342909908,"rms_gain_error_db":0.5985717608061911,"reference_mean_power":0.002702794865208283,"output_mean_power":0.003102203341173496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.583684100944911e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031173336044584235,"output_power":0.0003616558688507751,"reference_active":true,"delta_db":0.6451223589409917},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.830633609453067e-05,"output_power":2.1142674797456687e-05,"reference_active":true,"delta_db":0.6255849830240261},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.642076702713704e-07,"output_power":5.610859813603623e-07,"reference_active":true,"delta_db":0.8231710587971443}],"active_band_count":3,"active_band_mae_db":0.6979594669207208}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001259260583850441,"log_rms_mae_db":2.304475687447077,"rms_envelope_nrmse":0.1721183895642047,"envelope_correlation":0.9867542375446736,"rms_gain_error_db":0.8284924843243269,"reference_mean_power":0.0007105481356967998,"output_mean_power":0.000859889711596398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8068878391559104e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001608363031836456,"output_power":0.00020875851114062545,"reference_active":true,"delta_db":1.1326010851605752},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.800835490835068e-05,"output_power":2.1779618117845417e-05,"reference_active":true,"delta_db":0.8257621945532876},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.325041103743567e-06,"output_power":2.018984588310813e-06,"reference_active":true,"delta_db":-0.6129763124429074}],"active_band_count":3,"active_band_mae_db":0.8571131973855901}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010786919997015104,"log_rms_mae_db":1.8879130282709853,"rms_envelope_nrmse":0.1442381043093667,"envelope_correlation":0.9923340561961276,"rms_gain_error_db":0.825021823522314,"reference_mean_power":0.0007140165693307253,"output_mean_power":0.0008633968749450886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6527252040517484e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015126423656686204,"output_power":0.0001957037291100281,"reference_active":true,"delta_db":1.118628413337128},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.688501064647705e-06,"output_power":9.088770911414153e-06,"reference_active":true,"delta_db":0.7266347803788462},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2694181944274497e-07,"output_power":1.7223713716314064e-07,"reference_active":true,"delta_db":-1.197877341190579}],"active_band_count":3,"active_band_mae_db":1.0143801783021844}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":1.4887488249756182,"rms_envelope_nrmse":0.14133060261375777,"envelope_correlation":0.9914605658070683,"rms_gain_error_db":0.6457297755656223,"reference_mean_power":0.0001538181527968681,"output_mean_power":0.00017847630623446205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.418336470516265e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.474851415047919e-05,"output_power":2.6895679203093347e-05,"reference_active":true,"delta_db":0.3613338618869847},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.277056498060546e-06,"output_power":3.2714263326361433e-06,"reference_active":true,"delta_db":-0.007467837793029819},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.913659986768022e-07,"output_power":5.182051368438503e-07,"reference_active":true,"delta_db":-2.3555435009294836}],"active_band_count":3,"active_band_mae_db":0.9081150668698328}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":0.8995145682775247,"rms_envelope_nrmse":0.1094027543626854,"envelope_correlation":0.9962592782206023,"rms_gain_error_db":0.6458110242544233,"reference_mean_power":0.00015458728634615255,"output_mean_power":0.0001793720931076803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.197451137326506e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3402826997507212e-05,"output_power":2.511262228214139e-05,"reference_active":true,"delta_db":0.3062374232832814},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.382612079388441e-06,"output_power":1.4327553564425367e-06,"reference_active":true,"delta_db":0.1547169393007642},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.610921131466672e-08,"output_power":2.596917809476622e-08,"reference_active":true,"delta_db":-2.49329489047259}],"active_band_count":3,"active_band_mae_db":0.9847497510188785}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":1.6569992864958356,"rms_envelope_nrmse":0.2857943180999145,"envelope_correlation":0.9512027579618265,"rms_gain_error_db":0.3068610453817153,"reference_mean_power":2.0928545188569674e-05,"output_mean_power":2.2460796067793216e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.488290124260659e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1348414996445457e-05,"output_power":1.2058235407335639e-05,"reference_active":true,"delta_db":0.2634854915589232},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8308296968161897e-06,"output_power":2.9390717016995384e-06,"reference_active":true,"delta_db":0.1629643834278307},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.670421987320987e-07,"output_power":7.423724509936954e-07,"reference_active":true,"delta_db":0.4646853777227029}],"active_band_count":3,"active_band_mae_db":0.2970450842364856}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":1.0185578764956018,"rms_envelope_nrmse":0.1933113022281211,"envelope_correlation":0.9771510476110716,"rms_gain_error_db":0.30688750156027983,"reference_mean_power":2.1033540736202732e-05,"output_mean_power":2.2573616213752417e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3907968246524526e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1287290435083048e-05,"output_power":1.2158133862031089e-05,"reference_active":true,"delta_db":0.32277220602401346},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5147848321624166e-06,"output_power":1.5898938654085813e-06,"reference_active":true,"delta_db":0.21017185830192175},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4338124913221415e-08,"output_power":4.521263389350241e-08,"reference_active":true,"delta_db":0.08482485314967576}],"active_band_count":3,"active_band_mae_db":0.20592297249187033}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"facodec","status":"measured","mos":{"status":"measured","value":3.946070671081543,"predicted_mos":3.946070671081543,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b990d84313f8433e0042395456a1fbe451e9864aa00ff6f10a39ab4a94247212","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9711563410000963},"started_at":"2026-10-13T04:45:13.102439+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2045000150683336e-05,"native_input_samples":96000,"native_input_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","output_sha256":"c8a94584d920aaddb9ab24b5dec7b3c6cc6c44e5b8a37445b26fe2cede15655a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6697511069996835,"synthesis_s":6.83544647799954,"total_s":8.505197584999223,"rtf":1.4175329308332039,"output_peak":0.3959997892379761,"output_rms":0.05079217110051228,"finished_at":"2026-10-13T04:45:21.613866+00:00","evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"c8a94584d920aaddb9ab24b5dec7b3c6cc6c44e5b8a37445b26fe2cede15655a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.437109392763284,"spectral_convergence":0.1896496005551022,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.769136395260418,"spectral_convergence":0.18626223534316355,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.73872340224839,"spectral_convergence":0.1953067304345274,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007523898904458434,"delta_db":0.0626875088661945,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.0012649648714267814,"delta_db":-0.28907147182616755,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.0003770700456174488,"delta_db":0.062252866930910644,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00015169527811901036,"delta_db":-0.9556234136844644,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.5306731974806544e-05,"delta_db":-0.03410347217725733,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":1.841782751982487e-05,"delta_db":-0.5727221209147964,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3294101423999651,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.18289816942055387,"pitch":null,"output_rms_dbfs":-25.884064458994146,"output_peak":0.3959997892379761,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.64832306342403},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.437109392763284,"spectral_convergence":0.1896496005551022,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.769136395260418,"spectral_convergence":0.18626223534316355,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.73872340224839,"spectral_convergence":0.1953067304345274,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007523898904458434,"delta_db":0.0626875088661945,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.0012649648714267814,"delta_db":-0.28907147182616755,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.0003770700456174488,"delta_db":0.062252866930910644,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00015169527811901036,"delta_db":-0.9556234136844644,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.5306731974806544e-05,"delta_db":-0.03410347217725733,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":1.841782751982487e-05,"delta_db":-0.5727221209147964,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3294101423999651,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.18289816942055387,"pitch":null,"output_rms_dbfs":-25.884064458994146,"output_peak":0.3959997892379761,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.64832306342403},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.906551939817179,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051873038222359884,"rms_dbfs":-25.701166289573592,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.345694075201205,"mean":8.414904276529947e-06,"zero_energy":false},"output":{"rms":0.05079217110051229,"rms_dbfs":-25.884064458994146,"peak":0.3959997892379761,"peak_dbfs":-8.046100904358862,"crest_factor_db":17.837963554635284,"mean":7.838891724325473e-05,"zero_energy":false},"rms_difference_db":-0.1828981694205538},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":1.1746168918257605e-06,"output_rms_dbfs":-59.301037579848305,"output_zero_energy":false,"output_peak":0.005305091850459576,"output_energy_fraction":0.00010244368829045173,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":195,"both_estimated_voiced_frames":184,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.983957219251337,"absolute_difference_cents_median":1.9097559201264565,"absolute_difference_cents_p95":7.864308181615076,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018114550052554645,"log_rms_mae_db":1.1274215375266654,"rms_envelope_nrmse":0.07966252787315643,"envelope_correlation":0.9938872413798983,"rms_gain_error_db":-0.1827311956323391,"reference_mean_power":0.002692702060931978,"output_mean_power":0.0025817559297632918},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0110233554726448e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033858489572855915,"output_power":0.00034386358250475377,"reference_active":true,"delta_db":0.06718603184882013},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.201391758025471e-05,"output_power":4.491953369527834e-05,"reference_active":true,"delta_db":0.29042060112077567},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.156677750109103e-06,"output_power":8.85008739964267e-06,"reference_active":true,"delta_db":0.9223609732119011}],"active_band_count":3,"active_band_mae_db":0.4266558687271657}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015936923488846764,"log_rms_mae_db":0.8091192144527659,"rms_envelope_nrmse":0.07013850462603526,"envelope_correlation":0.9951446386588985,"rms_gain_error_db":-0.1824404177532987,"reference_mean_power":0.002702794865208283,"output_mean_power":0.0025916063978079773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.583684100944911e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031173336044584235,"output_power":0.00031648939483214586,"reference_active":true,"delta_db":0.0657588072050383},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.830633609453067e-05,"output_power":1.9425254212206167e-05,"reference_active":true,"delta_db":0.2576527947230281},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.642076702713704e-07,"output_power":5.79097447961629e-07,"reference_active":true,"delta_db":0.9603933885075246}],"active_band_count":3,"active_band_mae_db":0.4279349968118637}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001259260583850441,"log_rms_mae_db":1.5571699766649396,"rms_envelope_nrmse":0.0956596282214606,"envelope_correlation":0.9935481374648767,"rms_gain_error_db":-0.23401687216644704,"reference_mean_power":0.0007105481356967998,"output_mean_power":0.0006732739590105608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8068878391559104e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001608363031836456,"output_power":0.00014734079809647517,"reference_active":true,"delta_db":-0.38061064348216667},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.800835490835068e-05,"output_power":1.704272994019925e-05,"reference_active":true,"delta_db":-0.2393487894183733},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.325041103743567e-06,"output_power":2.780840723757988e-06,"reference_active":true,"delta_db":0.7774547973639572}],"active_band_count":3,"active_band_mae_db":0.4658047434214991}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010786919997015104,"log_rms_mae_db":1.198898925984522,"rms_envelope_nrmse":0.07934363437648631,"envelope_correlation":0.9955326373520149,"rms_gain_error_db":-0.23404091462275173,"reference_mean_power":0.0007140165693307253,"output_mean_power":0.0006765566989539592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6527252040517484e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015126423656686204,"output_power":0.00013850069916559364,"reference_active":true,"delta_db":-0.3828429406140317},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.688501064647705e-06,"output_power":7.529410033404389e-06,"reference_active":true,"delta_db":-0.09080730429928766},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2694181944274497e-07,"output_power":2.639568973984428e-07,"reference_active":true,"delta_db":0.6561848256128687}],"active_band_count":3,"active_band_mae_db":0.376611690175396}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":1.559065155472832,"rms_envelope_nrmse":0.15489311153577484,"envelope_correlation":0.9867991882589696,"rms_gain_error_db":-0.6655642356866166,"reference_mean_power":0.0001538181527968681,"output_mean_power":0.00013196269248419608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.418336470516265e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.474851415047919e-05,"output_power":2.2206020219483284e-05,"reference_active":true,"delta_db":-0.4707839898956209},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.277056498060546e-06,"output_power":3.1501603225529285e-06,"reference_active":true,"delta_db":-0.17151271796538362},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.913659986768022e-07,"output_power":8.335797511329573e-07,"reference_active":true,"delta_db":-0.2910890738398615}],"active_band_count":3,"active_band_mae_db":0.3111285939002887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":1.1742195315160495,"rms_envelope_nrmse":0.13722133851592394,"envelope_correlation":0.9901653378446742,"rms_gain_error_db":-0.665565171846073,"reference_mean_power":0.00015458728634615255,"output_mean_power":0.00013262251406477265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.197451137326506e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3402826997507212e-05,"output_power":2.0852478618498946e-05,"reference_active":true,"delta_db":-0.5011063761007288},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.382612079388441e-06,"output_power":1.4171298410873516e-06,"reference_active":true,"delta_db":0.10709296410944522},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.610921131466672e-08,"output_power":3.923149462975002e-08,"reference_active":true,"delta_db":-0.701528397296789}],"active_band_count":3,"active_band_mae_db":0.43657591250232103}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":2.004475670082827,"rms_envelope_nrmse":0.17414163780997252,"envelope_correlation":0.9812110200927443,"rms_gain_error_db":-0.32469664982991897,"reference_mean_power":2.0928545188569674e-05,"output_mean_power":1.9420901034815143e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.488290124260659e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1348414996445457e-05,"output_power":1.0531894129571538e-05,"reference_active":true,"delta_db":-0.3242872421592875},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8308296968161897e-06,"output_power":2.640392925699634e-06,"reference_active":true,"delta_db":-0.30245182429886125},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.670421987320987e-07,"output_power":7.291106087729712e-07,"reference_active":true,"delta_db":0.38640108069243295}],"active_band_count":3,"active_band_mae_db":0.3377133823835272}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":1.620519008629074,"rms_envelope_nrmse":0.12185505178728445,"envelope_correlation":0.9908601558019131,"rms_gain_error_db":-0.32472313865576746,"reference_mean_power":2.1033540736202732e-05,"output_mean_power":1.9518213897851188e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3907968246524526e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1287290435083048e-05,"output_power":1.0500194634215105e-05,"reference_active":true,"delta_db":-0.31392350754590426},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5147848321624166e-06,"output_power":1.4504428654060135e-06,"reference_active":true,"delta_db":-0.18850321672676423},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4338124913221415e-08,"output_power":4.799681555179028e-08,"reference_active":true,"delta_db":0.3443510157980165}],"active_band_count":3,"active_band_mae_db":0.28225924669022834}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.774413585662842,"predicted_mos":3.774413585662842,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"864f68109139682da3f812c8b3906f02305d2b3571511e5e83d910518b42f912","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.975263159998576},"started_at":"2026-10-13T04:58:21.339133+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.011737248999452277,"native_input_samples":264600,"native_input_sha256":"b9d9e193faf7f6ea8bbb58d46d39b620f746748e4c372a1371d05c63cd1f4a8b","output_sha256":"43cc04aa3efb0e385204677293aab06d3dea9333bd552caa6d19cc446aa0b6db","output_samples":264704,"length_delta_samples":104,"analysis_s":0.171309170999848,"synthesis_s":5.8110243189994435,"total_s":5.9823334899992915,"rtf":0.9970555816665486,"output_peak":0.5220205187797546,"output_rms":0.047365488039509324,"finished_at":"2026-10-13T04:58:27.353851+00:00","evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"43cc04aa3efb0e385204677293aab06d3dea9333bd552caa6d19cc446aa0b6db","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.487512143192923,"output_peak":0.5202047415542903},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.818635742793745,"spectral_convergence":0.4648223573480999,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.83657843072485,"spectral_convergence":0.40090108823082005,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.250707220898558,"spectral_convergence":0.446720304486731,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007914036047546629,"delta_db":0.2822382350295397,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.0009176791111918654,"delta_db":-1.6829476620922248,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.00037513833862287975,"delta_db":0.03994703917919693,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00013124170039567108,"delta_db":-1.5846255534044715,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":2.056324118240323e-05,"delta_db":1.2479874444111014,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":9.919840020698065e-06,"delta_db":-3.260059453878945,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3496342313325798,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7848409255305222,"pitch":null,"output_rms_dbfs":-26.486007215104117,"output_peak":0.5202047415542903,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.635307131472384},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7215508957956296,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051873038222359884,"rms_dbfs":-25.701166289573592,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.345694075201205,"mean":8.414904276529947e-06,"zero_energy":false},"output":{"rms":0.047391410995750914,"rms_dbfs":-26.486007215104117,"peak":0.5202047415542903,"peak_dbfs":-5.676513872426944,"crest_factor_db":20.80949334267717,"mean":-0.00035556148852894593,"zero_energy":false},"rms_difference_db":-0.7848409255305242},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":1.0093135415919345e-06,"output_rms_dbfs":-59.95973899936889,"output_zero_energy":false,"output_peak":0.006280335653559834,"output_energy_fraction":0.00010111354566129434,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":182,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.8502673796791443,"absolute_difference_cents_median":7.672611833795879,"absolute_difference_cents_p95":25.074881374905516,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018114550052554645,"log_rms_mae_db":2.0445645470448333,"rms_envelope_nrmse":0.1375196123280167,"envelope_correlation":0.9885648900330751,"rms_gain_error_db":-0.7846607826364806,"reference_mean_power":0.002692702060931978,"output_mean_power":0.002247616566455388},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0110233554726448e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033858489572855915,"output_power":0.00029462764196048954,"reference_active":true,"delta_db":-0.603940904099691},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.201391758025471e-05,"output_power":4.074588849790575e-05,"reference_active":true,"delta_db":-0.13309386742835339},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.156677750109103e-06,"output_power":4.988938305498702e-06,"reference_active":true,"delta_db":-1.5670332892812568}],"active_band_count":3,"active_band_mae_db":0.7680226869364337}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015936923488846764,"log_rms_mae_db":1.6692427212907046,"rms_envelope_nrmse":0.12106880574607552,"envelope_correlation":0.9927046835445158,"rms_gain_error_db":-0.7819129058313089,"reference_mean_power":0.002702794865208283,"output_mean_power":0.0022574689967956667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.583684100944911e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031173336044584235,"output_power":0.0002734903259836357,"reference_active":true,"delta_db":-0.5684131231454678},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.830633609453067e-05,"output_power":1.772317140711634e-05,"reference_active":true,"delta_db":-0.14059993794630113},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.642076702713704e-07,"output_power":3.655749729304163e-07,"reference_active":true,"delta_db":-1.0373585567554189}],"active_band_count":3,"active_band_mae_db":0.5821238726157293}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001259260583850441,"log_rms_mae_db":2.7771609335907237,"rms_envelope_nrmse":0.20675028298047876,"envelope_correlation":0.9703445334113682,"rms_gain_error_db":0.29373720793471664,"reference_mean_power":0.0007105481356967998,"output_mean_power":0.0007602688995596263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8068878391559104e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001608363031836456,"output_power":0.00018167539640933373,"reference_active":true,"delta_db":0.5291203417830891},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.800835490835068e-05,"output_power":2.2741026314233427e-05,"reference_active":true,"delta_db":1.0133601958268694},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.325041103743567e-06,"output_power":2.9218129588458644e-06,"reference_active":true,"delta_db":0.9922177587091747}],"active_band_count":3,"active_band_mae_db":0.8448994321063777}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010786919997015104,"log_rms_mae_db":2.327553515571243,"rms_envelope_nrmse":0.1896240217657461,"envelope_correlation":0.9739932321297449,"rms_gain_error_db":0.29403091929748987,"reference_mean_power":0.0007140165693307253,"output_mean_power":0.000764031706974774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6527252040517484e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015126423656686204,"output_power":0.00017227035485657257,"reference_active":true,"delta_db":0.5647428848323981},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.688501064647705e-06,"output_power":9.745763638127668e-06,"reference_active":true,"delta_db":1.0297419548549043},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2694181944274497e-07,"output_power":2.739044777593069e-07,"reference_active":true,"delta_db":0.8168459973328397}],"active_band_count":3,"active_band_mae_db":0.8037769456733806}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":2.999248368606141,"rms_envelope_nrmse":0.2007905372085555,"envelope_correlation":0.9821096371655439,"rms_gain_error_db":-1.2055063737744334,"reference_mean_power":0.0001538181527968681,"output_mean_power":0.00011653515394845499},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.418336470516265e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.474851415047919e-05,"output_power":2.0371605848903992e-05,"reference_active":true,"delta_db":-0.8452386508245981},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.277056498060546e-06,"output_power":4.587306561262971e-06,"reference_active":true,"delta_db":1.460738353068163},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.913659986768022e-07,"output_power":9.90906562098805e-07,"reference_active":true,"delta_db":0.4597664063531306}],"active_band_count":3,"active_band_mae_db":0.921914470081964}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":2.6076988258481086,"rms_envelope_nrmse":0.1728738170275083,"envelope_correlation":0.9897010174495819,"rms_gain_error_db":-1.2054802274234284,"reference_mean_power":0.00015458728634615255,"output_mean_power":0.00011711856723386128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.197451137326506e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3402826997507212e-05,"output_power":1.9579178551889596e-05,"reference_active":true,"delta_db":-0.7747385522200877},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.382612079388441e-06,"output_power":2.077713488099581e-06,"reference_active":true,"delta_db":1.7688531230316495},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.610921131466672e-08,"output_power":6.399883092524336e-08,"reference_active":true,"delta_db":1.4238434694830464}],"active_band_count":3,"active_band_mae_db":1.3224783815782613}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":3.9777466806235067,"rms_envelope_nrmse":0.2773150029218775,"envelope_correlation":0.9653427582944405,"rms_gain_error_db":-1.7048460026201262,"reference_mean_power":2.0928545188569674e-05,"output_mean_power":1.4133653497866904e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.488290124260659e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1348414996445457e-05,"output_power":7.27410721271549e-06,"reference_active":true,"delta_db":-1.9315551117459486},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8308296968161897e-06,"output_power":1.6655768183068529e-06,"reference_active":true,"delta_db":-2.303490752758761},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.670421987320987e-07,"output_power":3.694555505708038e-07,"reference_active":true,"delta_db":-2.565911136920266}],"active_band_count":3,"active_band_mae_db":2.266985667141659}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":4.140713854245612,"rms_envelope_nrmse":0.24287306073574857,"envelope_correlation":0.9783623072067522,"rms_gain_error_db":-1.704821477699302,"reference_mean_power":2.1033540736202732e-05,"output_mean_power":1.4204640250471155e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3907968246524526e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1287290435083048e-05,"output_power":7.188564884143705e-06,"reference_active":true,"delta_db":-1.9594750304553685},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5147848321624166e-06,"output_power":8.514795987249561e-07,"reference_active":true,"delta_db":-2.501767009774154},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4338124913221415e-08,"output_power":2.1398085680787575e-08,"reference_active":true,"delta_db":-3.1640240042448746}],"active_band_count":3,"active_band_mae_db":2.5417553481581323}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"golf","status":"measured","mos":{"status":"measured","value":2.8311707973480225,"predicted_mos":2.8311707973480225,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"2ce8db54444fdff073223a3b29d2dd127fa1127a3a279f6bda815b8049be8610","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.034676646000662},"started_at":"2026-10-13T05:01:52.182750+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035108189995298744,"native_input_samples":144000,"native_input_sha256":"daa2889ab84f21a64a7b93e4cf4e9aed59dfecf076fc02d8b5d01057bba43c75","output_sha256":"5f291f9d3f4f5d89551b58f8c77346d83f0fd5e95898b16711146550b4d1f319","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.8400868939997963,"synthesis_s":0.13146504100041057,"total_s":0.9715519350002069,"rtf":0.16192532250003447,"output_peak":0.4148748517036438,"output_rms":0.04608705718980903,"finished_at":"2026-10-13T05:01:53.167432+00:00","analysis_stages":{"external_f0_s":0.08245377100047335},"evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"5f291f9d3f4f5d89551b58f8c77346d83f0fd5e95898b16711146550b4d1f319","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.72632901779562,"output_peak":0.41522752739787167},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.20408604395256,"spectral_convergence":0.7309582820037869,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.564063615045611,"spectral_convergence":0.30353550557228987,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.610295013717456,"spectral_convergence":0.2631673638783056,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000741002290460984,"output_power":0.0006214149465210337,"delta_db":-0.76437855529771,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001353999958793735,"output_power":0.0010809034770958136,"delta_db":-0.9783173723099465,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00037231828979505927,"output_power":0.00027009435649089216,"delta_db":-1.3939886118127032,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001893479547510113,"output_power":0.00012516067422004614,"delta_db":-1.7979272448047359,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5456431282704244e-05,"output_power":1.1724496561586363e-05,"delta_db":-1.200150238056204,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.105141834913528e-05,"output_power":1.5741995071851774e-05,"delta_db":-1.2622158984972716,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2328296534630951,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0289769279684522,"pitch":null,"output_rms_dbfs":-26.72632901779562,"output_peak":0.41522752739787167,"reference_rms_dbfs":-25.69735208982717,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.126148224238541},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8730070559969177,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05189582202290113,"rms_dbfs":-25.69735208982717,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.341879875454786,"mean":6.2783293811626145e-06,"zero_energy":false},"output":{"rms":0.04609815555921109,"rms_dbfs":-26.72632901779562,"peak":0.41522752739787167,"peak_dbfs":-7.634277255373503,"crest_factor_db":19.09205176242212,"mean":0.0007740083703215167,"zero_energy":false},"rms_difference_db":-1.0289769279684506},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":1.5760639482798052e-06,"output_rms_dbfs":-58.02426165132243,"output_zero_energy":false,"output_peak":0.006057906586199989,"output_energy_fraction":0.00016715281032862256,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":199,"both_estimated_voiced_frames":181,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9679144385026738,"absolute_difference_cents_median":2.8529982834223206,"absolute_difference_cents_p95":13.987389971409904,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00018114550052554645,"log_rms_mae_db":2.323468249534014,"rms_envelope_nrmse":0.15459863386710238,"envelope_correlation":0.987187170121542,"rms_gain_error_db":-1.02899448143606,"reference_mean_power":0.002695161961216469,"output_mean_power":0.0021265980942991244},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0113978250433682e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033937478969393723,"output_power":0.0002741674308737297,"reference_active":true,"delta_db":-0.9266371536144984},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163500622021476e-05,"output_power":3.5630529027165746e-05,"reference_active":true,"delta_db":-0.6763636325595811},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.166092742149844e-06,"output_power":7.448117637050227e-06,"reference_active":true,"delta_db":0.16764102941373976}],"active_band_count":3,"active_band_mae_db":0.5902139385292731}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015936923488846764,"log_rms_mae_db":1.6476717101998404,"rms_envelope_nrmse":0.13057055445590995,"envelope_correlation":0.9945625038448919,"rms_gain_error_db":-1.0267968452856766,"reference_mean_power":0.0027048926875179484,"output_mean_power":0.0021353563227687507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.587125410617658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031233616795952804,"output_power":0.0002532453887995341,"reference_active":true,"delta_db":-0.9108073136796873},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8065725706526167e-05,"output_power":1.5428532966083046e-05,"reference_active":true,"delta_db":-0.6853077921416446},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.620178606697803e-07,"output_power":5.546613307511675e-07,"reference_active":true,"delta_db":0.7936912446362361}],"active_band_count":3,"active_band_mae_db":0.7966021168191894}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001259260583850441,"log_rms_mae_db":1.926312785888812,"rms_envelope_nrmse":0.18491137113606582,"envelope_correlation":0.9811292899518191,"rms_gain_error_db":-0.9836365615504749,"reference_mean_power":0.0007116958539710507,"output_mean_power":0.0005674541541221113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.8131403605168577e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001613761054681581,"output_power":0.00012133369297713601,"reference_active":true,"delta_db":-1.2385781409587953},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7825710396322598e-05,"output_power":1.3674087879256179e-05,"reference_active":true,"delta_db":-1.1514847986819128},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.329816126319897e-06,"output_power":3.2107833997348407e-06,"reference_active":true,"delta_db":1.3928936193630737}],"active_band_count":3,"active_band_mae_db":1.2609855196679272}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010786919997015104,"log_rms_mae_db":1.4232418877639585,"rms_envelope_nrmse":0.15667031046112664,"envelope_correlation":0.9886412465197033,"rms_gain_error_db":-0.9835168540608032,"reference_mean_power":0.0007151417813402843,"output_mean_power":0.0005702174013163104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.658573159837144e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015167796699146075,"output_power":0.00011532523260999749,"reference_active":true,"delta_db":-1.1899815973781351},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591564758874584e-06,"output_power":6.121119745751325e-06,"reference_active":true,"delta_db":-0.9350042551032429},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.264398182394501e-07,"output_power":2.8362404159595977e-07,"reference_active":true,"delta_db":0.9779024374593364}],"active_band_count":3,"active_band_mae_db":1.0342960966469048}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":1.880329565571304,"rms_envelope_nrmse":0.21475756874494784,"envelope_correlation":0.9891056073968972,"rms_gain_error_db":-1.7297509918207108,"reference_mean_power":0.00015407280774909997,"output_mean_power":0.00010345486011795868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.430230604583135e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4758797954411242e-05,"output_power":1.6964505578556612e-05,"reference_active":true,"delta_db":-1.6418834887333564},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.256598704030731e-06,"output_power":2.7582838833995014e-06,"reference_active":true,"delta_db":-0.7212528385410106},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.950740697440856e-07,"output_power":7.678299252142349e-07,"reference_active":true,"delta_db":-0.6659394162150052}],"active_band_count":3,"active_band_mae_db":1.0096919144964573}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":1.4827290520923275,"rms_envelope_nrmse":0.1958940934084106,"envelope_correlation":0.9952307109729037,"rms_gain_error_db":-1.72990679987958,"reference_mean_power":0.0001548432684725802,"output_mean_power":0.00010396846928581902},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.2091924178621476e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3405125294447035e-05,"output_power":1.6180943725730213e-05,"reference_active":true,"delta_db":-1.6030712288033142},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3649006496138356e-06,"output_power":1.2318271843385266e-06,"reference_active":true,"delta_db":-0.4455125646158753},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6140003393478904e-08,"output_power":4.077654338874311e-08,"reference_active":true,"delta_db":-0.5366721409324042}],"active_band_count":3,"active_band_mae_db":0.861751978117198}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":2.628982985957573,"rms_envelope_nrmse":0.20993620566640436,"envelope_correlation":0.9751933679441877,"rms_gain_error_db":-0.4536266851746379,"reference_mean_power":2.096339751326292e-05,"output_mean_power":1.8884217817328255e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.4856521598778733e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1323382158513378e-05,"output_power":1.1436647166549348e-05,"reference_active":true,"delta_db":0.04322558056475395},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.833589718603797e-06,"output_power":2.882805187960171e-06,"reference_active":true,"delta_db":0.07478326899219946},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.656075420136726e-07,"output_power":7.14500800318603e-07,"reference_active":true,"delta_db":0.3078448527345612}],"active_band_count":3,"active_band_mae_db":0.14195123409717153}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":2.3660044141856558,"rms_envelope_nrmse":0.15472602403020597,"envelope_correlation":0.9877032534790728,"rms_gain_error_db":-0.4536355412946842,"reference_mean_power":2.106867265066476e-05,"output_mean_power":1.897901291430995e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.3879439320628843e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1261183956343054e-05,"output_power":1.1472908309354818e-05,"reference_active":true,"delta_db":0.08089469828909103},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.512176418522691e-06,"output_power":1.4986007439958316e-06,"reference_active":true,"delta_db":-0.03916517346921994},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.400248676218554e-08,"output_power":5.403414537121379e-08,"reference_active":true,"delta_db":0.8919106590394674}],"active_band_count":3,"active_band_mae_db":0.33732351026592616}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.359966993331909,"predicted_mos":3.359966993331909,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"274bf79693c16972776d9795ed917416ab58dc09379be4d47f86430c90153297","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8132989879995876},"started_at":"2026-10-13T05:04:15.313313+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031349979999504285,"native_input_samples":144000,"native_input_sha256":"daa2889ab84f21a64a7b93e4cf4e9aed59dfecf076fc02d8b5d01057bba43c75","output_sha256":"c052b93c0a6a2119cdb1d2fe5238d1318bffbe2f1fa69e5973e5f6baa4f880d0","output_samples":144120,"length_delta_samples":120,"analysis_s":2.18894129399996,"synthesis_s":3.4928821789999347,"total_s":5.681823472999895,"rtf":0.9469705788333158,"output_peak":0.42442116141319275,"output_rms":0.05395127392167218,"finished_at":"2026-10-13T05:04:21.009394+00:00","analysis_stages":{"external_f0_s":1.4371566870004244},"evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"c052b93c0a6a2119cdb1d2fe5238d1318bffbe2f1fa69e5973e5f6baa4f880d0","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.358367592681077,"output_peak":0.4255611882589766},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.308885658873642,"spectral_convergence":0.8287360690516307,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.493049662200904,"spectral_convergence":0.34094657073810214,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.463173311694649,"spectral_convergence":0.27632141248765085,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0011165551779291068,"delta_db":1.777059893103369,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.0012691531616322407,"delta_db":-0.27471576339159487,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.0003360384661766652,"delta_db":-0.4380775306508993,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00016104071042482519,"delta_db":-0.6959872614551073,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.4900287824112015e-05,"delta_db":-0.1509816724375355,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":1.6457583645260507e-05,"delta_db":-1.0614454258139037,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7330445911420682,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.3462784402047123,"pitch":null,"output_rms_dbfs":-25.35488784936888,"output_peak":0.4255611882589766,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.088369544256398},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8717850225265152,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051873038222359884,"rms_dbfs":-25.701166289573592,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.345694075201205,"mean":8.414904276529947e-06,"zero_energy":false},"output":{"rms":0.05398282493193747,"rms_dbfs":-25.35488784936888,"peak":0.4255611882589766,"peak_dbfs":-7.420759742717323,"crest_factor_db":17.934128106651556,"mean":7.955437698107534e-06,"zero_energy":false},"rms_difference_db":0.34627844020471343},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":5.822868260212771e-07,"output_rms_dbfs":-62.34863035499025,"output_zero_energy":false,"output_peak":0.019784531821924523,"output_energy_fraction":4.495813297813489e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":218,"both_estimated_voiced_frames":185,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":33,"paired_voiced_fraction":0.9893048128342246,"absolute_difference_cents_median":5.871204304104725,"absolute_difference_cents_p95":24.415838402741045,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018114550052554645,"log_rms_mae_db":4.1795970576668,"rms_envelope_nrmse":0.13225086184163687,"envelope_correlation":0.9847552533371092,"rms_gain_error_db":0.34677230176468754,"reference_mean_power":0.002692702060931978,"output_mean_power":0.0029165238545436215},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0110233554726448e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033858489572855915,"output_power":0.0003878383062479039,"reference_active":true,"delta_db":0.5898312119147573},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.201391758025471e-05,"output_power":4.3218163467126816e-05,"reference_active":true,"delta_db":0.12273128688622166},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.156677750109103e-06,"output_power":6.064191722143856e-06,"reference_active":true,"delta_db":-0.7193853904656288}],"active_band_count":3,"active_band_mae_db":0.47731596308886926}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015936923488846764,"log_rms_mae_db":3.8003271974182335,"rms_envelope_nrmse":0.09642817688265494,"envelope_correlation":0.9923548401526815,"rms_gain_error_db":0.3492635460933324,"reference_mean_power":0.002702794865208283,"output_mean_power":0.002929135347882209},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.583684100944911e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031173336044584235,"output_power":0.0003565825505830613,"reference_active":true,"delta_db":0.583768058919644},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.830633609453067e-05,"output_power":1.9276257039813126e-05,"reference_active":true,"delta_db":0.22421277266809622},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.642076702713704e-07,"output_power":4.650328837068442e-07,"reference_active":true,"delta_db":0.007713519047275956}],"active_band_count":3,"active_band_mae_db":0.2718981168783387}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001259260583850441,"log_rms_mae_db":2.504175324459172,"rms_envelope_nrmse":0.1513087215840813,"envelope_correlation":0.9843587324042171,"rms_gain_error_db":-0.4932992478040759,"reference_mean_power":0.0007105481356967998,"output_mean_power":0.0006342545326016713},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8068878391559104e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001608363031836456,"output_power":0.0001441337476171045,"reference_active":true,"delta_db":-0.47618403446871227},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.800835490835068e-05,"output_power":1.5827987320828202e-05,"reference_active":true,"delta_db":-0.5604834744695626},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.325041103743567e-06,"output_power":2.173010548982908e-06,"reference_active":true,"delta_db":-0.2936880043422862}],"active_band_count":3,"active_band_mae_db":0.44345183776018704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010786919997015104,"log_rms_mae_db":2.143755657359112,"rms_envelope_nrmse":0.11062212492789279,"envelope_correlation":0.9924273179161169,"rms_gain_error_db":-0.49323143206587683,"reference_mean_power":0.0007140165693307253,"output_mean_power":0.0006373605029133544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6527252040517484e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015126423656686204,"output_power":0.0001359039591937615,"reference_active":true,"delta_db":-0.4650415091979762},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.688501064647705e-06,"output_power":6.980085500759765e-06,"reference_active":true,"delta_db":-0.4198093640434339},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2694181944274497e-07,"output_power":1.9221325737774092e-07,"reference_active":true,"delta_db":-0.7213119374944268}],"active_band_count":3,"active_band_mae_db":0.5353876035786124}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":2.1091419805836664,"rms_envelope_nrmse":0.1805469446502241,"envelope_correlation":0.9792759789727354,"rms_gain_error_db":-0.4912721547208818,"reference_mean_power":0.0001538181527968681,"output_mean_power":0.00013736635553315894},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.418336470516265e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.474851415047919e-05,"output_power":2.4564605609609513e-05,"reference_active":true,"delta_db":-0.032393340821992606},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.277056498060546e-06,"output_power":3.3717279714736216e-06,"reference_active":true,"delta_db":0.12368599910832717},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.913659986768022e-07,"output_power":6.299227850019e-07,"reference_active":true,"delta_db":-1.5076874627725079}],"active_band_count":3,"active_band_mae_db":0.5545889342342759}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":1.3990553706103117,"rms_envelope_nrmse":0.12825756352120646,"envelope_correlation":0.9902621393535401,"rms_gain_error_db":-0.49128960279216494,"reference_mean_power":0.00015458728634615255,"output_mean_power":0.000138052670883021},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.197451137326506e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3402826997507212e-05,"output_power":2.3165637934523196e-05,"reference_active":true,"delta_db":-0.04424057874833149},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.382612079388441e-06,"output_power":1.547770640762431e-06,"reference_active":true,"delta_db":0.4900625754974861},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.610921131466672e-08,"output_power":3.5781400784067565e-08,"reference_active":true,"delta_db":-1.1013035530378066}],"active_band_count":3,"active_band_mae_db":0.5452022357612081}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":1.963031274350084,"rms_envelope_nrmse":0.2279189288581748,"envelope_correlation":0.96720580563119,"rms_gain_error_db":-0.45647611185224474,"reference_mean_power":2.0928545188569674e-05,"output_mean_power":1.884045682690182e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.488290124260659e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1348414996445457e-05,"output_power":1.0005506442935484e-05,"reference_active":true,"delta_db":-0.5469613302457838},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8308296968161897e-06,"output_power":2.828663724900339e-06,"reference_active":true,"delta_db":-0.0033242187130339555},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.670421987320987e-07,"output_power":4.3622165286674153e-07,"reference_active":true,"delta_db":-1.844460903391338}],"active_band_count":3,"active_band_mae_db":0.7982488174500518}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":1.210823224735749,"rms_envelope_nrmse":0.1440903677688814,"envelope_correlation":0.9874155570764083,"rms_gain_error_db":-0.45650437025966434,"reference_mean_power":2.1033540736202732e-05,"output_mean_power":1.8934853526360953e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3907968246524526e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1287290435083048e-05,"output_power":9.973255483135686e-06,"reference_active":true,"delta_db":-0.5375275566117669},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5147848321624166e-06,"output_power":1.491482023281136e-06,"reference_active":true,"delta_db":-0.06732924583324829},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4338124913221415e-08,"output_power":2.7383620744696498e-08,"reference_active":true,"delta_db":-2.092864512667588}],"active_band_count":3,"active_band_mae_db":0.8992404383708678}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.001253366470337,"predicted_mos":3.001253366470337,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"09ead476956e5e8af15d6488c841d2b8845899cf732506dd602c4476d13a553f","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.713610138000149},"started_at":"2026-10-13T05:09:58.099595+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032453949988848763,"native_input_samples":144000,"native_input_sha256":"daa2889ab84f21a64a7b93e4cf4e9aed59dfecf076fc02d8b5d01057bba43c75","output_sha256":"e5dd1f3d2705c0ee5951880a2171b8e8a3dddb044a1aea66f0e31a0f69efecfd","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9653905630002555,"synthesis_s":0.11972831099956238,"total_s":1.0851188739998179,"rtf":0.18085314566663632,"output_peak":0.43206366896629333,"output_rms":0.05096340310140591,"finished_at":"2026-10-13T05:09:59.197494+00:00","analysis_stages":{"external_f0_s":0.14498923499922967},"evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"e5dd1f3d2705c0ee5951880a2171b8e8a3dddb044a1aea66f0e31a0f69efecfd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.85352743084721,"output_peak":0.43037207348239875},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.930500751217095,"spectral_convergence":0.17073238913150002,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.515411975061044,"spectral_convergence":0.1488472860556818,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.915226413651517,"spectral_convergence":0.16566722430624956,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0007329709765844086,"delta_db":-0.05087422349800156,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.0013023918949207189,"delta_db":-0.16243927492156157,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.00035554474376324013,"delta_db":-0.19302484219284177,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.00018101009252058476,"delta_db":-0.18831613543732945,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.3583555020740777e-05,"delta_db":-0.5527937864390124,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":1.6481755556817436e-05,"delta_db":-1.0550714487124224,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3670866185335282,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1457891319592988,"pitch":null,"output_rms_dbfs":-25.84695542153289,"output_peak":0.43037207348239875,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.787046379976552},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9488730845749981,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051873038222359884,"rms_dbfs":-25.701166289573592,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.345694075201205,"mean":8.414904276529947e-06,"zero_energy":false},"output":{"rms":0.05100963652454809,"rms_dbfs":-25.84695542153289,"peak":0.43037207348239875,"peak_dbfs":-7.323118349341878,"crest_factor_db":18.52383707219101,"mean":7.842926687621682e-06,"zero_energy":false},"rms_difference_db":-0.14578913195929744},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":1.2943320481654732e-06,"output_rms_dbfs":-58.879542953929715,"output_zero_energy":false,"output_peak":0.005611509757474778,"output_energy_fraction":0.00011192413969713698,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":185,"both_estimated_voiced_frames":177,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.946524064171123,"absolute_difference_cents_median":0.8769250789344126,"absolute_difference_cents_p95":5.0881524705673105,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018114550052554645,"log_rms_mae_db":1.0663140075194362,"rms_envelope_nrmse":0.06172605174496754,"envelope_correlation":0.9964106808522666,"rms_gain_error_db":-0.14629949599803638,"reference_mean_power":0.002692702060931978,"output_mean_power":0.0026035046227654737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0110233554726448e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033858489572855915,"output_power":0.00032868700894993257,"reference_active":true,"delta_db":-0.12885040996112201},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.201391758025471e-05,"output_power":3.2397292131353716e-05,"reference_active":true,"delta_db":-1.1288446712724518},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.156677750109103e-06,"output_power":6.674902074108326e-06,"reference_active":true,"delta_db":-0.3026656364974639}],"active_band_count":3,"active_band_mae_db":0.5201202392436793}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015936923488846764,"log_rms_mae_db":0.6210377456270965,"rms_envelope_nrmse":0.0382084279757934,"envelope_correlation":0.9987786269775847,"rms_gain_error_db":-0.1460897662689229,"reference_mean_power":0.002702794865208283,"output_mean_power":0.002613389299486043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.583684100944911e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031173336044584235,"output_power":0.00030332330184878705,"reference_active":true,"delta_db":-0.11877506320440648},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.830633609453067e-05,"output_power":1.4075624129088008e-05,"reference_active":true,"delta_db":-1.1413377045475468},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.642076702713704e-07,"output_power":4.7245498099274176e-07,"reference_active":true,"delta_db":0.07648119840192713}],"active_band_count":3,"active_band_mae_db":0.4455313220512935}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001259260583850441,"log_rms_mae_db":1.1196137696428408,"rms_envelope_nrmse":0.09351386685146572,"envelope_correlation":0.9939578844451392,"rms_gain_error_db":-0.2548122464463872,"reference_mean_power":0.0007105481356967998,"output_mean_power":0.000670057819397486},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8068878391559104e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001608363031836456,"output_power":0.00014610389670701235,"reference_active":true,"delta_db":-0.41722283235662705},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.800835490835068e-05,"output_power":1.3227847420242985e-05,"reference_active":true,"delta_db":-1.3398486434762709},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.325041103743567e-06,"output_power":1.920856397541486e-06,"reference_active":true,"delta_db":-0.8293573667012111}],"active_band_count":3,"active_band_mae_db":0.8621429475113697}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010786919997015104,"log_rms_mae_db":0.7282797285282853,"rms_envelope_nrmse":0.06370081527910947,"envelope_correlation":0.9974571050131618,"rms_gain_error_db":-0.2550612621565328,"reference_mean_power":0.0007140165693307253,"output_mean_power":0.0006732899994554122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6527252040517484e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015126423656686204,"output_power":0.0001366697552161858,"reference_active":true,"delta_db":-0.4406384325785318},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.688501064647705e-06,"output_power":5.831254483954032e-06,"reference_active":true,"delta_db":-1.2007968379624192},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2694181944274497e-07,"output_power":1.7737697000553174e-07,"reference_active":true,"delta_db":-1.0701730046944318}],"active_band_count":3,"active_band_mae_db":0.903869425078461}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":1.1019064533382241,"rms_envelope_nrmse":0.08098624339953603,"envelope_correlation":0.9960462572890353,"rms_gain_error_db":-0.23212749877132624,"reference_mean_power":0.0001538181527968681,"output_mean_power":0.00014581252873655583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.418336470516265e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.474851415047919e-05,"output_power":2.2892445190292075e-05,"reference_active":true,"delta_db":-0.3385694685072494},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.277056498060546e-06,"output_power":2.483682222081133e-06,"reference_active":true,"delta_db":-1.203879002042041},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.913659986768022e-07,"output_power":7.011944122268383e-07,"reference_active":true,"delta_db":-1.0421761725088559}],"active_band_count":3,"active_band_mae_db":0.8615415476860487}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":0.7024321864022219,"rms_envelope_nrmse":0.04954653042522163,"envelope_correlation":0.9988086161498458,"rms_gain_error_db":-0.2321269495138748,"reference_mean_power":0.00015458728634615255,"output_mean_power":0.00014654165047203125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.197451137326506e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3402826997507212e-05,"output_power":2.151881752458692e-05,"reference_active":true,"delta_db":-0.36449919327403874},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.382612079388441e-06,"output_power":1.0304485997993985e-06,"reference_active":true,"delta_db":-1.276740133925239},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.610921131466672e-08,"output_power":4.531153954357671e-08,"reference_active":true,"delta_db":-0.07578875405488965}],"active_band_count":3,"active_band_mae_db":0.5723426937513891}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":1.3173571802408488,"rms_envelope_nrmse":0.13559947416649412,"envelope_correlation":0.9894518155335671,"rms_gain_error_db":-0.44268886680700165,"reference_mean_power":2.0928545188569674e-05,"output_mean_power":1.8900363355851438e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.488290124260659e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1348414996445457e-05,"output_power":1.0301679432930236e-05,"reference_active":true,"delta_db":-0.42027177581844954},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8308296968161897e-06,"output_power":2.5388903536068165e-06,"reference_active":true,"delta_db":-0.47269797406367964},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.670421987320987e-07,"output_power":3.7544833158292005e-07,"reference_active":true,"delta_db":-2.496031305734009}],"active_band_count":3,"active_band_mae_db":1.1296670185387128}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":1.0437624005922697,"rms_envelope_nrmse":0.08482461608108056,"envelope_correlation":0.996740272606879,"rms_gain_error_db":-0.4427189660508287,"reference_mean_power":2.1033540736202732e-05,"output_mean_power":1.8995052154731692e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3907968246524526e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1287290435083048e-05,"output_power":1.0182730946515609e-05,"reference_active":true,"delta_db":-0.44725431328032056},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5147848321624166e-06,"output_power":1.3355805745157239e-06,"reference_active":true,"delta_db":-0.5468085400568939},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4338124913221415e-08,"output_power":1.9078250644306392e-08,"reference_active":true,"delta_db":-3.662387724651965}],"active_band_count":3,"active_band_mae_db":1.5521501926630599}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_1188-133604-0037","method":"magphase","status":"measured","mos":{"status":"measured","value":2.9491283893585205,"predicted_mos":2.9491283893585205,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288747,"input_sample_rate":48000,"input_seconds":6.0155625,"input_float32_sha256":"408822268b1b14356fa906238169986aab799a681c6805e5ff93bbf8e570efe0","internal_samples":96249,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0500317540008837},"started_at":"2026-10-13T05:12:05.179727+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00898135800161981,"native_input_samples":288000,"native_input_sha256":"08d668c03933291c9030b3bd678b7b84d262f53de5965a8dbd486f71d9b3ebea","output_sha256":"f7943d93c85c150803bb7a149f21bae1e14ed8d0915f131ed953b0f02f0267d4","output_samples":288747,"length_delta_samples":747,"analysis_s":3.0783126639998954,"synthesis_s":0.8746062420013914,"total_s":3.952918906001287,"rtf":0.6588198176668811,"output_peak":0.45277590149399044,"output_rms":0.0504158351284567,"finished_at":"2026-10-13T05:12:09.154497+00:00","evaluation_input":{"reference":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"output_sha256":"f7943d93c85c150803bb7a149f21bae1e14ed8d0915f131ed953b0f02f0267d4","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96249,"length_delta_samples":249,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.947258509739758,"output_peak":0.45197888200348435},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9974129601346508,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.902792971462455,"spectral_convergence":0.8093740215781701,"spectral_convergence_status":"ok","amplitude_floor":1.0507333742302167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.836257238781102,"spectral_convergence":0.4968673863063767,"spectral_convergence_status":"ok","amplitude_floor":5.595595188050952e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.235013830034873,"spectral_convergence":0.5166234119361893,"spectral_convergence_status":"ok","amplitude_floor":5.269142735102521e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007416076491485813,"output_power":0.0006890012235345995,"delta_db":-0.31954207461842143,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001352027771357195,"output_power":0.001274667124185982,"delta_db":-0.2558882756613729,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000371703582440024,"output_power":0.00035936257426065016,"delta_db":-0.14663901179460195,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018903159721465413,"output_power":0.0001871195429841096,"delta_db":-0.044152558106569956,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5427403014296353e-05,"output_power":1.6482596308450422e-05,"delta_db":0.2873279715750456,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1014091243658427e-05,"output_power":1.9737856368926317e-05,"delta_db":-0.27210629325577906,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22094269750196513,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2396170552214383,"pitch":null,"output_rms_dbfs":-25.94078334479503,"output_peak":0.45197888200348435,"reference_rms_dbfs":-25.701166289573592,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.991354680092811},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.5864083970808713,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051873038222359884,"rms_dbfs":-25.701166289573592,"peak":0.42877197265625,"peak_dbfs":-7.355472214372385,"crest_factor_db":18.345694075201205,"mean":8.414904276529947e-06,"zero_energy":false},"output":{"rms":0.050461578628127744,"rms_dbfs":-25.94078334479503,"peak":0.45197888200348435,"peak_dbfs":-6.897637128710964,"crest_factor_db":19.043146216084065,"mean":-0.0005548039719783353,"zero_energy":false},"rms_difference_db":-0.23961705522143717},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2813692360650753e-06,"threshold_dbfs":-54.83944897878765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.285817189556029e-06,"output_mean_power":5.719054811779231e-06,"output_rms_dbfs":-52.42675741127887,"output_zero_energy":false,"output_peak":0.07455273163757715,"output_energy_fraction":0.000505341670270708,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":194,"both_estimated_voiced_frames":168,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8983957219251337,"absolute_difference_cents_median":10.118668541215111,"absolute_difference_cents_p95":76.68247414785702,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018114550052554645,"log_rms_mae_db":3.9961173698375054,"rms_envelope_nrmse":0.33080484177860053,"envelope_correlation":0.8888593865366257,"rms_gain_error_db":-0.23929331963260325,"reference_mean_power":0.002692702060931978,"output_mean_power":0.002548349386096676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0110233554726448e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033858489572855915,"output_power":0.0003285310566738036,"reference_active":true,"delta_db":-0.1309114979858405},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.201391758025471e-05,"output_power":4.271293322518642e-05,"reference_active":true,"delta_db":0.0716621761744847},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.156677750109103e-06,"output_power":6.8740208302838316e-06,"reference_active":true,"delta_db":-0.17500618460752027}],"active_band_count":3,"active_band_mae_db":0.12585995292261518}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015936923488846764,"log_rms_mae_db":3.035485037458345,"rms_envelope_nrmse":0.28310477805543344,"envelope_correlation":0.9148644403123063,"rms_gain_error_db":-0.23958028757487776,"reference_mean_power":0.002702794865208283,"output_mean_power":0.0025577321143258093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.583684100944911e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031173336044584235,"output_power":0.00030369106568216606,"reference_active":true,"delta_db":-0.11351265652787622},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.830633609453067e-05,"output_power":1.8566923863593373e-05,"reference_active":true,"delta_db":0.06138525014183515},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.642076702713704e-07,"output_power":4.5994886236221023e-07,"reference_active":true,"delta_db":-0.04002763173114947}],"active_band_count":3,"active_band_mae_db":0.07164184613362028}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001259260583850441,"log_rms_mae_db":4.803961700167572,"rms_envelope_nrmse":0.43539247967627637,"envelope_correlation":0.8568778999332467,"rms_gain_error_db":-0.16378527452005368,"reference_mean_power":0.0007105481356967998,"output_mean_power":0.0006842502690545313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8068878391559104e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001608363031836456,"output_power":0.00015154406187392394,"reference_active":true,"delta_db":-0.2584515872877246},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.800835490835068e-05,"output_power":1.7373782153829793e-05,"reference_active":true,"delta_db":-0.15579669460336748},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.325041103743567e-06,"output_power":2.4340970088555324e-06,"reference_active":true,"delta_db":0.19907247611229154}],"active_band_count":3,"active_band_mae_db":0.20444025266779453}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010786919997015104,"log_rms_mae_db":4.223680707660878,"rms_envelope_nrmse":0.38619238715856324,"envelope_correlation":0.8823390700383166,"rms_gain_error_db":-0.16441324660037113,"reference_mean_power":0.0007140165693307253,"output_mean_power":0.0006874909180501492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6527252040517484e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015126423656686204,"output_power":0.00014299494865996167,"reference_active":true,"delta_db":-0.24415563678639784},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.688501064647705e-06,"output_power":7.698816713193475e-06,"reference_active":true,"delta_db":0.005823016040254217},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2694181944274497e-07,"output_power":2.0600566798383947e-07,"reference_active":true,"delta_db":-0.42035362781972463}],"active_band_count":3,"active_band_mae_db":0.22344409354879224}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.418520257385567e-05,"log_rms_mae_db":4.4039489800065965,"rms_envelope_nrmse":0.6240435201286949,"envelope_correlation":0.7425161639003702,"rms_gain_error_db":-0.0059409195560971035,"reference_mean_power":0.0001538181527968681,"output_mean_power":0.0001536078815258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.418336470516265e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.474851415047919e-05,"output_power":2.578522516072772e-05,"reference_active":true,"delta_db":0.17821798177201867},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.277056498060546e-06,"output_power":3.254437619291442e-06,"reference_active":true,"delta_db":-0.030079774487029302},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.913659986768022e-07,"output_power":8.03667867711207e-07,"reference_active":true,"delta_db":-0.4497945913030791}],"active_band_count":3,"active_band_mae_db":0.21936411585404236}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.421085502718719e-05,"log_rms_mae_db":3.4776316245417624,"rms_envelope_nrmse":0.5533070207766876,"envelope_correlation":0.7899398921159483,"rms_gain_error_db":-0.006080469489540184,"reference_mean_power":0.00015458728634615255,"output_mean_power":0.00015437100324507648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.197451137326506e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3402826997507212e-05,"output_power":2.43585333872678e-05,"reference_active":true,"delta_db":0.17382813972035938},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.382612079388441e-06,"output_power":1.3058494816948856e-06,"reference_active":true,"delta_db":-0.2480722582374324},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.610921131466672e-08,"output_power":4.373112504660849e-08,"reference_active":true,"delta_db":-0.22997043451774088}],"active_band_count":3,"active_band_mae_db":0.21729027749184424}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.007885661618193e-05,"log_rms_mae_db":4.558675214943635,"rms_envelope_nrmse":0.6520266931512367,"envelope_correlation":0.7426568124527398,"rms_gain_error_db":0.3376970748588773,"reference_mean_power":2.0928545188569674e-05,"output_mean_power":2.2620841026699462e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.488290124260659e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1348414996445457e-05,"output_power":1.1668421351559264e-05,"reference_active":true,"delta_db":0.12076894301750983},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8308296968161897e-06,"output_power":2.9605756134594184e-06,"reference_active":true,"delta_db":0.19462414570327627},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.670421987320987e-07,"output_power":4.987729671172319e-07,"reference_active":true,"delta_db":-1.262504021076628}],"active_band_count":3,"active_band_mae_db":0.5259657032658047}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3258654185194866e-05,"log_rms_mae_db":3.7905769825706033,"rms_envelope_nrmse":0.5033170269484363,"envelope_correlation":0.8400273575870204,"rms_gain_error_db":0.3376927772931052,"reference_mean_power":2.1033540736202732e-05,"output_mean_power":2.2734304086149083e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3907968246524526e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1287290435083048e-05,"output_power":1.1640479711701572e-05,"reference_active":true,"delta_db":0.13381178149801246},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5147848321624166e-06,"output_power":1.4993900963855366e-06,"reference_active":true,"delta_db":-0.0443630983161504},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4338124913221415e-08,"output_power":3.4165050696811235e-08,"reference_active":true,"delta_db":-1.131952530842844}],"active_band_count":3,"active_band_mae_db":0.436709136885669}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},{"case":"librispeech_121-121726-0012","method":"hnm","status":"measured","mos":{"status":"measured","value":2.757395029067993,"predicted_mos":2.757395029067993,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64720,"input_sample_rate":16000,"input_seconds":4.045,"input_float32_sha256":"24333d0434a474a5abc3874a8d296bfad4d1306d8af7d54c9f3c73d763a45e01","internal_samples":64720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1121072829992045},"started_at":"2026-10-13T04:38:03.694985+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2946000424562953e-05,"native_input_samples":64720,"native_input_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","output_sha256":"f1ce08aa4e6f134dd7865d956fb3a9b47322ae14cf9c25cc96c779317769a1e9","output_samples":64720,"length_delta_samples":0,"analysis_s":0.6596766419997948,"synthesis_s":0.36394565499995224,"total_s":1.023622296999747,"rtf":0.2530586642768225,"output_peak":0.502101428464692,"output_rms":0.05733364605578627,"finished_at":"2026-10-13T04:38:04.722212+00:00","evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"f1ce08aa4e6f134dd7865d956fb3a9b47322ae14cf9c25cc96c779317769a1e9","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.870150559122727,"spectral_convergence":0.23750375418101574,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.480930129224576,"spectral_convergence":0.22998495895303409,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.682073866607197,"spectral_convergence":0.23509682739140664,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.0003227699414994076,"delta_db":0.11922381828711778,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.000545913186813234,"delta_db":-0.06919057975980365,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.0010744613743228712,"delta_db":-0.06224110048235472,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.0007920505949892988,"delta_db":-0.15309713357736188,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.0004560252735999752,"delta_db":-0.16106419867396277,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":9.592659882538956e-05,"delta_db":-0.36076470720075116,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15426358966355866,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09085807192494626,"pitch":null,"output_rms_dbfs":-24.831808779521822,"output_peak":0.502101428464692,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.677718184984833},"common_prefix":{"samples":64720,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.870150559122727,"spectral_convergence":0.23750375418101574,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.480930129224576,"spectral_convergence":0.22998495895303409,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.682073866607197,"spectral_convergence":0.23509682739140664,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.0003227699414994076,"delta_db":0.11922381828711778,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.000545913186813234,"delta_db":-0.06919057975980365,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.0010744613743228712,"delta_db":-0.06224110048235472,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.0007920505949892988,"delta_db":-0.15309713357736188,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.0004560252735999752,"delta_db":-0.16106419867396277,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":9.592659882538956e-05,"delta_db":-0.36076470720075116,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15426358966355866,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09085807192494626,"pitch":null,"output_rms_dbfs":-24.831808779521822,"output_peak":0.502101428464692,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.677718184984833},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64720,"output_samples":64720,"intelligibility":{"estoi":{"value":0.9510658055147053,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057936527896013786,"rms_dbfs":-24.740950707596877,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.303807011740542,"mean":-1.1361573623638366e-05,"zero_energy":false},"output":{"rms":0.05733364605578627,"rms_dbfs":-24.831808779521822,"peak":0.502101428464692,"peak_dbfs":-5.9841708613618625,"crest_factor_db":18.847637918159958,"mean":-8.681346171788323e-05,"zero_energy":false},"rms_difference_db":-0.090858071924945},"full_scale":{"reference":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":9.444100419783097e-07,"output_rms_dbfs":-60.24839403685029,"output_zero_energy":false,"output_peak":0.04430386231134573,"output_energy_fraction":0.00011222252479235735,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":115,"both_estimated_voiced_frames":104,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9043478260869565,"absolute_difference_cents_median":0.8999647644162236,"absolute_difference_cents_p95":7.582770804708602,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00017261805365043103,"log_rms_mae_db":0.6070888848669685,"rms_envelope_nrmse":0.0673860736484842,"envelope_correlation":0.995959228092371,"rms_gain_error_db":-0.09085808089281737,"reference_mean_power":0.003360795516302892,"output_mean_power":0.0032912152072728513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.9828935159721717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046042689591210996,"output_power":0.0004510205182079093,"reference_active":true,"delta_db":-0.08964385461388184},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011507848708673394,"output_power":9.307278021832291e-05,"reference_active":true,"delta_db":-0.9217145656551485},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.734946402574847e-05,"output_power":1.6349398668820686e-05,"reference_active":true,"delta_db":-3.5878259002938635}],"active_band_count":3,"active_band_mae_db":1.5330614401876312}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00013379439656886598,"log_rms_mae_db":0.43388231711888775,"rms_envelope_nrmse":0.036317102197108625,"envelope_correlation":0.9988398980652817,"rms_gain_error_db":-0.09085809154180487,"reference_mean_power":0.0033859385941668236,"output_mean_power":0.003315837726702269},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.884361456130501e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004231625539622664,"output_power":0.00041137430756257545,"reference_active":true,"delta_db":-0.12270065329605725},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5839616945196026e-05,"output_power":4.284524569241247e-05,"reference_active":true,"delta_db":-1.1503979280514998},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3554104634271333e-06,"output_power":1.0458357313440814e-06,"reference_active":true,"delta_db":-5.062821789192149}],"active_band_count":3,"active_band_mae_db":2.111973456846569}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00011786303332041996,"log_rms_mae_db":0.811651397272933,"rms_envelope_nrmse":0.07002476228945176,"envelope_correlation":0.9961108982651272,"rms_gain_error_db":-0.08747433839898894,"reference_mean_power":0.0013229704572187174,"output_mean_power":0.0012965901331947892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":9.694412540428036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023578263015181652,"output_power":0.0002284995544485072,"reference_active":true,"delta_db":-0.13626450363078335},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.91091063998743e-05,"output_power":5.267660764897824e-05,"reference_active":true,"delta_db":-0.5003659500665227},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3510218569590892e-05,"output_power":9.582957220386664e-06,"reference_active":true,"delta_db":-1.4916282573253472}],"active_band_count":3,"active_band_mae_db":0.7094195703408844}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00010411406065660645,"log_rms_mae_db":0.5852060723827184,"rms_envelope_nrmse":0.04148584023458384,"envelope_correlation":0.998619181241972,"rms_gain_error_db":-0.08747434468792448,"reference_mean_power":0.0013328679887454193,"output_mean_power":0.0013062903038584149},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":9.215200160073525e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002182221456522984,"output_power":0.00021066546607346199,"reference_active":true,"delta_db":-0.1530547312267491},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7295024423763924e-05,"output_power":2.4582641716484807e-05,"reference_active":true,"delta_db":-0.45454935745776964},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19264005709586e-06,"output_power":7.141285890707073e-07,"reference_active":true,"delta_db":-2.2273297211167478}],"active_band_count":3,"active_band_mae_db":0.9449779366004222}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.0001403728741075033,"log_rms_mae_db":0.7076883450675157,"rms_envelope_nrmse":0.1106254398515481,"envelope_correlation":0.9909121189631804,"rms_gain_error_db":-0.13545583227081712,"reference_mean_power":0.0010435885844267173,"output_mean_power":0.0010115415734356834},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":7.131098310180302e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021758763787569643,"output_power":0.00021194805197060878,"reference_active":true,"delta_db":-0.11404788228681145},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802785604261901e-05,"output_power":3.96507423019486e-05,"reference_active":true,"delta_db":-1.653852028052534},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.302295499160149e-05,"output_power":1.0546568841219747e-05,"reference_active":true,"delta_db":-3.390498726845135}],"active_band_count":3,"active_band_mae_db":1.7194662123948268}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.0001001506434690652,"log_rms_mae_db":0.4920442999443725,"rms_envelope_nrmse":0.06109939707857351,"envelope_correlation":0.9972353715160954,"rms_gain_error_db":-0.13545583668110828,"reference_mean_power":0.0010513959774729575,"output_mean_power":0.0010191092123350107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":6.799409328431787e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000202136265751209,"output_power":0.0001937986910743061,"reference_active":true,"delta_db":-0.18293398839437985},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0065691390851225e-05,"output_power":1.818350022841004e-05,"reference_active":true,"delta_db":-2.183937089924957},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1255469357025125e-06,"output_power":6.215903369409268e-07,"reference_active":true,"delta_db":-5.339664449399991}],"active_band_count":3,"active_band_mae_db":2.568845175906443}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":0.9448368891377679,"rms_envelope_nrmse":0.1372124602788992,"envelope_correlation":0.9866250467137456,"rms_gain_error_db":-0.3795303835099847,"reference_mean_power":0.00018380542444204246,"output_mean_power":0.0001684245073644794},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.2570064040799187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4129018402509474e-05,"output_power":4.034564892307824e-05,"reference_active":true,"delta_db":-0.389275621001994},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0833298446779832e-05,"output_power":8.484711458510088e-06,"reference_active":true,"delta_db":-1.0612362994489248},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.606324451113386e-06,"output_power":2.174758834641053e-06,"reference_active":true,"delta_db":-3.2594342089865194}],"active_band_count":3,"active_band_mae_db":1.5699820431458127}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":0.6955923515001258,"rms_envelope_nrmse":0.08803675786785917,"envelope_correlation":0.9951160911540713,"rms_gain_error_db":-0.3795303869956241,"reference_mean_power":0.00018518052639831994,"output_mean_power":0.00016968454006064616},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.1869756093173805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2060123324004174e-05,"output_power":3.7959855263069005e-05,"reference_active":true,"delta_db":-0.4454599390274476},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.372206568913875e-06,"output_power":3.746586934402991e-06,"reference_active":true,"delta_db":-1.5651688929475902},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.05541130391232e-07,"output_power":1.5920228736235101e-07,"reference_active":true,"delta_db":-4.060856041296303}],"active_band_count":3,"active_band_mae_db":2.023828291090447}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-121726-0012","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.5604608058929443,"predicted_mos":2.5604608058929443,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64720,"input_sample_rate":16000,"input_seconds":4.045,"input_float32_sha256":"8eef5625d37224491a2d768cb8b80a76b66e8231a1c733928fd70c302595bd7f","internal_samples":64720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.58073015699847},"started_at":"2026-10-13T04:40:54.876697+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.8202000066812616e-05,"native_input_samples":64720,"native_input_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","output_sha256":"1cd5615b6d8278e67771b328d36fe1b183b5344794704783c756ac9dd32c91de","output_samples":64720,"length_delta_samples":0,"analysis_s":1.0849954870000147,"synthesis_s":0.07086816700029885,"total_s":1.1558636540003135,"rtf":0.28575121236101697,"output_peak":0.7857418710614739,"output_rms":0.06304676820616961,"finished_at":"2026-10-13T04:40:56.036643+00:00","evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"1cd5615b6d8278e67771b328d36fe1b183b5344794704783c756ac9dd32c91de","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.691627034658044,"spectral_convergence":0.5563833604767777,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.437535821116528,"spectral_convergence":0.23551420708288467,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.856796611404224,"spectral_convergence":0.2831733482573431,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.00033280775608264757,"delta_db":0.2522273727572706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.0006751061011562226,"delta_db":0.8532938985407735,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.0013584895032309027,"delta_db":0.9564136930422207,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.000982759787495847,"delta_db":0.7838473983150317,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.0005193318232925034,"delta_db":0.4034960339167959,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":0.00010640000998435599,"delta_db":0.08926151599531627,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5564233187612347,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7342072962151263,"pitch":null,"output_rms_dbfs":-24.00674341138175,"output_peak":0.7857418710614739,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.995319822392932},"common_prefix":{"samples":64720,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.691627034658044,"spectral_convergence":0.5563833604767777,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.437535821116528,"spectral_convergence":0.23551420708288467,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.856796611404224,"spectral_convergence":0.2831733482573431,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.00033280775608264757,"delta_db":0.2522273727572706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.0006751061011562226,"delta_db":0.8532938985407735,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.0013584895032309027,"delta_db":0.9564136930422207,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.000982759787495847,"delta_db":0.7838473983150317,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.0005193318232925034,"delta_db":0.4034960339167959,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":0.00010640000998435599,"delta_db":0.08926151599531627,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5564233187612347,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7342072962151263,"pitch":null,"output_rms_dbfs":-24.00674341138175,"output_peak":0.7857418710614739,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.995319822392932},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64720,"output_samples":64720,"intelligibility":{"estoi":{"value":0.9694277110724232,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057936527896013786,"rms_dbfs":-24.740950707596877,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.303807011740542,"mean":-1.1361573623638366e-05,"zero_energy":false},"output":{"rms":0.06304676820616961,"rms_dbfs":-24.00674341138175,"peak":0.7857418710614739,"peak_dbfs":-2.094402066122565,"crest_factor_db":21.912341345259186,"mean":1.743249966505404e-12,"zero_energy":false},"rms_difference_db":0.7342072962151285},"full_scale":{"reference":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":4.197030886200163e-07,"output_rms_dbfs":-63.77057834777802,"output_zero_energy":false,"output_peak":0.010915409423849783,"output_energy_fraction":4.1243457659599526e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":119,"both_estimated_voiced_frames":105,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9130434782608695,"absolute_difference_cents_median":4.391993964741687,"absolute_difference_cents_p95":14.589822780730575,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00017261805365043103,"log_rms_mae_db":1.0605551084555702,"rms_envelope_nrmse":0.14144590945775515,"envelope_correlation":0.9898811036933928,"rms_gain_error_db":0.7342073033408645,"reference_mean_power":0.003360795516302892,"output_mean_power":0.003979814403569793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.9828935159721717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046042689591210996,"output_power":0.0005717572024499752,"reference_active":true,"delta_db":0.9405095905778855},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011507848708673394,"output_power":0.00011427462950815248,"reference_active":true,"delta_db":-0.030443217457040653},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.734946402574847e-05,"output_power":3.085536600552603e-05,"reference_active":true,"delta_db":-0.8295367159003529}],"active_band_count":3,"active_band_mae_db":0.6001631746450931}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00013379439656886598,"log_rms_mae_db":0.9748404868698538,"rms_envelope_nrmse":0.10548854729298651,"envelope_correlation":0.9970959195406947,"rms_gain_error_db":0.7342073208144518,"reference_mean_power":0.0033859385941668236,"output_mean_power":0.004009588555899225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.884361456130501e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004231625539622664,"output_power":0.0005205642778944947,"reference_active":true,"delta_db":0.899671330714919},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5839616945196026e-05,"output_power":5.3070744079877315e-05,"reference_active":true,"delta_db":-0.22087253693606657},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3554104634271333e-06,"output_power":1.6269995090339155e-06,"reference_active":true,"delta_db":-3.1435823260596814}],"active_band_count":3,"active_band_mae_db":1.4213753979035557}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00011786303332041996,"log_rms_mae_db":1.1832586298528736,"rms_envelope_nrmse":0.14940335089822324,"envelope_correlation":0.9914820668893874,"rms_gain_error_db":0.8503127244543482,"reference_mean_power":0.0013229704572187174,"output_mean_power":0.0016090940119500422},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":9.694412540428036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023578263015181652,"output_power":0.00030522737132589184,"reference_active":true,"delta_db":1.1211166853558978},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.91091063998743e-05,"output_power":6.200998409099849e-05,"reference_active":true,"delta_db":0.20807226108441215},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3510218569590892e-05,"output_power":1.3313398885239884e-05,"reference_active":true,"delta_db":-0.06373430956316369}],"active_band_count":3,"active_band_mae_db":0.46430775200115787}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00010411406065660645,"log_rms_mae_db":1.015265648063883,"rms_envelope_nrmse":0.12487208624646019,"envelope_correlation":0.9961903633919305,"rms_gain_error_db":0.8503127089077458,"reference_mean_power":0.0013328679887454193,"output_mean_power":0.0016211321122328601},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":9.215200160073525e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002182221456522984,"output_power":0.0002796183706407203,"reference_active":true,"delta_db":1.0766687908775432},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7295024423763924e-05,"output_power":2.8547030841301138e-05,"reference_active":true,"delta_db":0.19477456983590266},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19264005709586e-06,"output_power":8.941047300095156e-07,"reference_active":true,"delta_db":-1.2512099946580855}],"active_band_count":3,"active_band_mae_db":0.8408844517905104}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.0001403728741075033,"log_rms_mae_db":0.9440755658449989,"rms_envelope_nrmse":0.15924183064730846,"envelope_correlation":0.9872603143506227,"rms_gain_error_db":0.6916036501158801,"reference_mean_power":0.0010435885844267173,"output_mean_power":0.0012237414903309288},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":7.131098310180302e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021758763787569643,"output_power":0.00026278411805309644,"reference_active":true,"delta_db":0.8196489660062783},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802785604261901e-05,"output_power":6.578219159235843e-05,"reference_active":true,"delta_db":0.5447181332912597},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.302295499160149e-05,"output_power":1.3479493341033172e-05,"reference_active":true,"delta_db":-2.324874959696979}],"active_band_count":3,"active_band_mae_db":1.2297473529981724}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.0001001506434690652,"log_rms_mae_db":0.7253878578685768,"rms_envelope_nrmse":0.10696050938143192,"envelope_correlation":0.9967156923427717,"rms_gain_error_db":0.691603650692484,"reference_mean_power":0.0010513959774729575,"output_mean_power":0.0012328966604002506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":6.799409328431787e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000202136265751209,"output_power":0.00024059590025873518,"reference_active":true,"delta_db":0.756439843939045},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0065691390851225e-05,"output_power":3.1955801358701566e-05,"reference_active":true,"delta_db":0.2647851753027093},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1255469357025125e-06,"output_power":8.180601263908718e-07,"reference_active":true,"delta_db":-4.146854744292153}],"active_band_count":3,"active_band_mae_db":1.7226932545113023}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":0.999919990293897,"rms_envelope_nrmse":0.16434967263210917,"envelope_correlation":0.9814103173269159,"rms_gain_error_db":0.32533794480832645,"reference_mean_power":0.00018380542444204246,"output_mean_power":0.00019810348633592643},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.2570064040799187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4129018402509474e-05,"output_power":4.764672204751323e-05,"reference_active":true,"delta_db":0.33308760645136687},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0833298446779832e-05,"output_power":1.040022492091729e-05,"reference_active":true,"delta_db":-0.17717976010761788},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.606324451113386e-06,"output_power":2.6373226374019315e-06,"reference_active":true,"delta_db":-2.4219126212849864}],"active_band_count":3,"active_band_mae_db":0.9773933292813237}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":0.6952031299131609,"rms_envelope_nrmse":0.07755587834262799,"envelope_correlation":0.9965085221001405,"rms_gain_error_db":0.32533794605034944,"reference_mean_power":0.00018518052639831994,"output_mean_power":0.00019958555631794265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.1869756093173805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2060123324004174e-05,"output_power":4.5210438924094765e-05,"reference_active":true,"delta_db":0.3136818164676767},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.372206568913875e-06,"output_power":4.870831275518806e-06,"reference_active":true,"delta_db":-0.42549617581828614},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.05541130391232e-07,"output_power":1.6400414011059214e-07,"reference_active":true,"delta_db":-3.931800958764384}],"active_band_count":3,"active_band_mae_db":1.5569929836834489}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-121726-0012","method":"facodec","status":"measured","mos":{"status":"measured","value":4.180410385131836,"predicted_mos":4.180410385131836,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64800,"input_sample_rate":16000,"input_seconds":4.05,"input_float32_sha256":"bcae8dd93f53d362911974cb28c01c99057d9b2e7f9462f3840720dbcc0b2974","internal_samples":64800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6121643070000573},"started_at":"2026-10-13T04:45:21.614495+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3513999622082338e-05,"native_input_samples":64720,"native_input_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","output_sha256":"59cf0982c320b4ec771548e050001796d55092e66a25d23b4e8f6337ad958ae2","output_samples":64800,"length_delta_samples":80,"analysis_s":1.0010071130000142,"synthesis_s":7.09183088699956,"total_s":8.092837999999574,"rtf":2.000701606922021,"output_peak":0.7351899743080139,"output_rms":0.05437065028898718,"finished_at":"2026-10-13T04:45:29.713969+00:00","evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"59cf0982c320b4ec771548e050001796d55092e66a25d23b4e8f6337ad958ae2","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":64720,"output_samples":64800,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.292709453546852,"output_peak":0.7351899743080139},"common_prefix":{"samples":64720,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987654320987654,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.81258594346161,"spectral_convergence":0.26036037856023697,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.035527989707111,"spectral_convergence":0.2723048217537597,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.090249214485435,"spectral_convergence":0.2886749715092706,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.00029265194363754907,"delta_db":-0.30619288879806184,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.00048507294877565864,"delta_db":-0.5823558715811029,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.0009794155420093224,"delta_db":-0.46447926351415153,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.0007010339542607245,"delta_db":-0.6832358457358333,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.00041043417057821664,"delta_db":-0.6185182177545672,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":9.121315423762543e-05,"delta_db":-0.579580428100334,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5390604192473418,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5463937736227858,"pitch":null,"output_rms_dbfs":-25.287344481219662,"output_peak":0.7351899743080139,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.979454382551385},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64720,"output_samples":64720,"intelligibility":{"estoi":{"value":0.9391349075892114,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057936527896013786,"rms_dbfs":-24.740950707596877,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.303807011740542,"mean":-1.1361573623638366e-05,"zero_energy":false},"output":{"rms":0.054404243524738924,"rms_dbfs":-25.287344481219662,"peak":0.7351899743080139,"peak_dbfs":-2.672008480001331,"crest_factor_db":22.61533600121833,"mean":4.9352798949926824e-05,"zero_energy":false},"rms_difference_db":-0.5463937736227855},"full_scale":{"reference":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":1.416440054598402e-07,"output_rms_dbfs":-68.4880180060561,"output_zero_energy":false,"output_peak":0.005715109407901764,"output_energy_fraction":1.8692664383346725e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":109,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9217391304347826,"absolute_difference_cents_median":2.307346064137976,"absolute_difference_cents_p95":10.0418843647082,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00017261805365043103,"log_rms_mae_db":1.028995878449064,"rms_envelope_nrmse":0.09920420768999368,"envelope_correlation":0.994118153710452,"rms_gain_error_db":-0.5463938053347244,"reference_mean_power":0.003360795516302892,"output_mean_power":0.0029634848310783027},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.9828935159721717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046042689591210996,"output_power":0.0003611738016155157,"reference_active":true,"delta_db":-1.0544444462086204},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011507848708673394,"output_power":0.00010550848507985829,"reference_active":true,"delta_db":-0.37706756314801215},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.734946402574847e-05,"output_power":3.835382546240673e-05,"reference_active":true,"delta_db":0.11524313603677205}],"active_band_count":3,"active_band_mae_db":0.5155850484644682}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00013379439656886598,"log_rms_mae_db":1.020449678131787,"rms_envelope_nrmse":0.08915887571932452,"envelope_correlation":0.9957459133765948,"rms_gain_error_db":-0.5463940449549234,"reference_mean_power":0.0033859385941668236,"output_mean_power":0.002985655348702275},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.884361456130501e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004231625539622664,"output_power":0.00033748901229842184,"reference_active":true,"delta_db":-0.9824759163594955},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5839616945196026e-05,"output_power":5.1668888260824094e-05,"reference_active":true,"delta_db":-0.337133133014258},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3554104634271333e-06,"output_power":3.2956770364032095e-06,"reference_active":true,"delta_db":-0.07801008557863998}],"active_band_count":3,"active_band_mae_db":0.4658730449841311}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00011786303332041996,"log_rms_mae_db":0.8290134392974561,"rms_envelope_nrmse":0.10020918311203017,"envelope_correlation":0.9931630111041547,"rms_gain_error_db":-0.3863581530762703,"reference_mean_power":0.0013229704572187174,"output_mean_power":0.0012103593540753275},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":9.694412540428036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023578263015181652,"output_power":0.00020613637290268157,"reference_active":true,"delta_db":-0.5835717784208723},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.91091063998743e-05,"output_power":5.849875670529285e-05,"reference_active":true,"delta_db":-0.04507757851414557},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3510218569590892e-05,"output_power":1.721853419992956e-05,"reference_active":true,"delta_db":1.053338024126975}],"active_band_count":3,"active_band_mae_db":0.5606624603539977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00010411406065660645,"log_rms_mae_db":0.6933378654016854,"rms_envelope_nrmse":0.09134242971426093,"envelope_correlation":0.9943114036551972,"rms_gain_error_db":-0.3863582143910326,"reference_mean_power":0.0013328679887454193,"output_mean_power":0.0012194143915661378},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":9.215200160073525e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002182221456522984,"output_power":0.00019222047349965683,"reference_active":true,"delta_db":-0.5509917890550575},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7295024423763924e-05,"output_power":2.7158229729288845e-05,"reference_active":true,"delta_db":-0.021820295908360734},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19264005709586e-06,"output_power":1.5017981503219958e-06,"reference_active":true,"delta_db":1.0010217323183228}],"active_band_count":3,"active_band_mae_db":0.524611272427247}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.0001403728741075033,"log_rms_mae_db":0.8735858432930609,"rms_envelope_nrmse":0.14361726152742063,"envelope_correlation":0.9875930173352298,"rms_gain_error_db":-0.6432756678917877,"reference_mean_power":0.0010435885844267173,"output_mean_power":0.0008999155433485703},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":7.131098310180302e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021758763787569643,"output_power":0.00015046937340332947,"reference_active":true,"delta_db":-1.6018610506869366},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802785604261901e-05,"output_power":4.904582122940876e-05,"reference_active":true,"delta_db":-0.7303451415396371},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.302295499160149e-05,"output_power":2.220946113399585e-05,"reference_active":true,"delta_db":-0.15623043048561036}],"active_band_count":3,"active_band_mae_db":0.8294788742373947}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.0001001506434690652,"log_rms_mae_db":0.7942063921039166,"rms_envelope_nrmse":0.12612297464062805,"envelope_correlation":0.9909480766188938,"rms_gain_error_db":-0.6432756994611587,"reference_mean_power":0.0010513959774729575,"output_mean_power":0.0009066480695398189},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":6.799409328431787e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000202136265751209,"output_power":0.00014234648975270403,"reference_active":true,"delta_db":-1.5229747635284638},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0065691390851225e-05,"output_power":2.6588128610586035e-05,"reference_active":true,"delta_db":-0.5338342448352641},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1255469357025125e-06,"output_power":1.8650688837925112e-06,"reference_active":true,"delta_db":-0.5677582285283976}],"active_band_count":3,"active_band_mae_db":0.8748557456307084}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":1.2802964006244004,"rms_envelope_nrmse":0.2015031862060183,"envelope_correlation":0.9717659363358695,"rms_gain_error_db":-0.6603818039914576,"reference_mean_power":0.00018380542444204246,"output_mean_power":0.00015787746473252065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.2570064040799187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4129018402509474e-05,"output_power":2.8990299056916057e-05,"reference_active":true,"delta_db":-1.8247157173581368},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0833298446779832e-05,"output_power":8.425506162099957e-06,"reference_active":true,"delta_db":-1.0916470714096178},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.606324451113386e-06,"output_power":3.012564771510031e-06,"reference_active":true,"delta_db":-1.8441813146408927}],"active_band_count":3,"active_band_mae_db":1.586848034469549}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":1.2093624816525674,"rms_envelope_nrmse":0.17830333186342476,"envelope_correlation":0.9777598654648029,"rms_gain_error_db":-0.6603818622726897,"reference_mean_power":0.00018518052639831994,"output_mean_power":0.0001590585899301436},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.1869756093173805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2060123324004174e-05,"output_power":2.7670641162200898e-05,"reference_active":true,"delta_db":-1.8185131935641452},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.372206568913875e-06,"output_power":4.372959803346226e-06,"reference_active":true,"delta_db":-0.8937721826476076},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.05541130391232e-07,"output_power":2.474848052089323e-07,"reference_active":true,"delta_db":-2.1448636759752784}],"active_band_count":3,"active_band_mae_db":1.6190496840623438}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-121726-0012","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5956509113311768,"predicted_mos":3.5956509113311768,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":178688,"input_sample_rate":44100,"input_seconds":4.051882086167801,"input_float32_sha256":"f1858e1282e174322df7b77dee207f6cba80bab9bfea410826fb870d73ca57ed","internal_samples":64831,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9042826580007386},"started_at":"2026-10-13T04:58:27.355026+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006858772000668978,"native_input_samples":178385,"native_input_sha256":"e62c379b1f957b98487fb753ff54d0e9f8a2b732a8d37aca7e31caf8711c9bed","output_sha256":"fd40ef97edb42ad957ab175f0ccd2ddbffaaa9de539edb120fd0f41a0a39bf93","output_samples":178688,"length_delta_samples":303,"analysis_s":0.1277128760002597,"synthesis_s":21.54219145800016,"total_s":21.66990433400042,"rtf":5.357192483277286,"output_peak":0.5313270688056946,"output_rms":0.054557333414572855,"finished_at":"2026-10-13T04:58:49.047459+00:00","evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"fd40ef97edb42ad957ab175f0ccd2ddbffaaa9de539edb120fd0f41a0a39bf93","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":64720,"output_samples":64831,"length_delta_samples":111,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.25969977042311,"output_peak":0.5244031314079157},"common_prefix":{"samples":64720,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998287856118215,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.04208499230677,"spectral_convergence":0.5142189975311767,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.700512446640823,"spectral_convergence":0.5153885187670466,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.352225760641573,"spectral_convergence":0.593760967567816,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.00020478065513110494,"delta_db":-1.8568177363835376,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.0006511795109721334,"delta_db":0.696580842291949,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.00118602381133022,"delta_db":0.366784905155616,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.0005404096966686084,"delta_db":-1.8133950486128056,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.00036291301681076326,"delta_db":-1.152927868750457,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":3.8524399632125514e-05,"delta_db":-4.3227963775573315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.701550463125283,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5113069439427835,"pitch":null,"output_rms_dbfs":-25.252257651539658,"output_peak":0.5244031314079157,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.698274399863055},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64720,"output_samples":64720,"intelligibility":{"estoi":{"value":0.8243705075679525,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057936527896013786,"rms_dbfs":-24.740950707596877,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.303807011740542,"mean":-1.1361573623638366e-05,"zero_energy":false},"output":{"rms":0.05462445505947822,"rms_dbfs":-25.252257651539658,"peak":0.5244031314079157,"peak_dbfs":-5.606694472777206,"crest_factor_db":19.64556317876245,"mean":-5.325383596587582e-05,"zero_energy":false},"rms_difference_db":-0.5113069439427811},"full_scale":{"reference":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":1.8847921836895193e-07,"output_rms_dbfs":-67.24736527926285,"output_zero_energy":false,"output_peak":0.014165747690177517,"output_energy_fraction":2.467333175362813e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":127,"both_estimated_voiced_frames":91,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.7913043478260869,"absolute_difference_cents_median":10.601985035582894,"absolute_difference_cents_p95":41.43409515311053,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00017261805365043103,"log_rms_mae_db":1.008649776143707,"rms_envelope_nrmse":0.11806221935180185,"envelope_correlation":0.9901690801318661,"rms_gain_error_db":-0.5113069404759667,"reference_mean_power":0.003360795516302892,"output_mean_power":0.0029875239466712906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.9828935159721717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046042689591210996,"output_power":0.0004579557807289463,"reference_active":true,"delta_db":-0.02337139637914948},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011507848708673394,"output_power":0.00010255082628461976,"reference_active":true,"delta_db":-0.5005497970810976},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.734946402574847e-05,"output_power":3.966772287854928e-05,"reference_active":true,"delta_db":0.2615289660021607}],"active_band_count":3,"active_band_mae_db":0.26181671982080257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00013379439656886598,"log_rms_mae_db":1.0165457907407514,"rms_envelope_nrmse":0.09641329071288479,"envelope_correlation":0.9942828604125542,"rms_gain_error_db":-0.5113069409096291,"reference_mean_power":0.0033859385941668236,"output_mean_power":0.003009874472264726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.884361456130501e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004231625539622664,"output_power":0.0004183885372428305,"reference_active":true,"delta_db":-0.049274521317623235},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5839616945196026e-05,"output_power":5.2313403214076676e-05,"reference_active":true,"delta_db":-0.2832945671516778},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3554104634271333e-06,"output_power":3.260974462756404e-06,"reference_active":true,"delta_db":-0.12398256703825593}],"active_band_count":3,"active_band_mae_db":0.15218388516918566}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00011786303332041996,"log_rms_mae_db":1.478756015030583,"rms_envelope_nrmse":0.19269717144644083,"envelope_correlation":0.9749308008367572,"rms_gain_error_db":0.5125670656179014,"reference_mean_power":0.0013229704572187174,"output_mean_power":0.001488698849656927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":9.694412540428036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023578263015181652,"output_power":0.0003177119648289863,"reference_active":true,"delta_db":1.2952176244842502},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.91091063998743e-05,"output_power":4.419388200246474e-05,"reference_active":true,"delta_db":-1.2629224203607825},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3510218569590892e-05,"output_power":1.4491199139606342e-05,"reference_active":true,"delta_db":0.30441949477113744}],"active_band_count":3,"active_band_mae_db":0.95418651320539}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00010411406065660645,"log_rms_mae_db":1.2529568808049818,"rms_envelope_nrmse":0.17219319371718234,"envelope_correlation":0.9796156393706719,"rms_gain_error_db":0.5125670604312579,"reference_mean_power":0.0013328679887454193,"output_mean_power":0.0014998362422934228},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":9.215200160073525e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002182221456522984,"output_power":0.0002858518683355843,"reference_active":true,"delta_db":1.1724221349399937},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7295024423763924e-05,"output_power":2.0061088160637902e-05,"reference_active":true,"delta_db":-1.3372900077347738},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19264005709586e-06,"output_power":1.003802560387874e-06,"reference_active":true,"delta_db":-0.7486109276745555}],"active_band_count":3,"active_band_mae_db":1.086107690116441}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.0001403728741075033,"log_rms_mae_db":1.8075120290231337,"rms_envelope_nrmse":0.240539327441779,"envelope_correlation":0.9734885964747962,"rms_gain_error_db":-1.578159092905256,"reference_mean_power":0.0010435885844267173,"output_mean_power":0.0007256269605925677},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":7.131098310180302e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021758763787569643,"output_power":0.0001613687898776673,"reference_active":true,"delta_db":-1.298146753287071},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802785604261901e-05,"output_power":5.244519592649129e-05,"reference_active":true,"delta_db":-0.4393081276417757},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.302295499160149e-05,"output_power":2.2286907135115725e-05,"reference_active":true,"delta_db":-0.14111261022133192}],"active_band_count":3,"active_band_mae_db":0.6261891637167262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.0001001506434690652,"log_rms_mae_db":1.672461456593735,"rms_envelope_nrmse":0.2205826755623893,"envelope_correlation":0.9807392663894864,"rms_gain_error_db":-1.5781591447835766,"reference_mean_power":0.0010513959774729575,"output_mean_power":0.000731055580507942},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":6.799409328431787e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000202136265751209,"output_power":0.00015550473810031667,"reference_active":true,"delta_db":-1.1390061216134488},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0065691390851225e-05,"output_power":2.8682791548667856e-05,"reference_active":true,"delta_db":-0.20449778632987325},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1255469357025125e-06,"output_power":2.175724355876062e-06,"reference_active":true,"delta_db":0.10133174050746674}],"active_band_count":3,"active_band_mae_db":0.48161188281692957}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":2.228244500019136,"rms_envelope_nrmse":0.2861871752175379,"envelope_correlation":0.9457848506667954,"rms_gain_error_db":-1.2557270835605148,"reference_mean_power":0.00018380542444204246,"output_mean_power":0.00013765297936907636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.2570064040799187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4129018402509474e-05,"output_power":3.0341190276746134e-05,"reference_active":true,"delta_db":-1.6269165314999958},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0833298446779832e-05,"output_power":5.715819388600969e-06,"reference_active":true,"delta_db":-2.776822104241647},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.606324451113386e-06,"output_power":2.7710129465928294e-06,"reference_active":true,"delta_db":-2.2071596979921844}],"active_band_count":3,"active_band_mae_db":2.2036327779112757}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":2.14054347240063,"rms_envelope_nrmse":0.25155376767204984,"envelope_correlation":0.9596942024351169,"rms_gain_error_db":-1.2557270776008969,"reference_mean_power":0.00018518052639831994,"output_mean_power":0.00013868280162145032},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.1869756093173805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2060123324004174e-05,"output_power":2.8598136442847932e-05,"reference_active":true,"delta_db":-1.6753280787720826},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.372206568913875e-06,"output_power":2.884932057067675e-06,"reference_active":true,"delta_db":-2.7001711399435786},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.05541130391232e-07,"output_power":2.3339603357265065e-07,"reference_active":true,"delta_db":-2.3994143617156825}],"active_band_count":3,"active_band_mae_db":2.258304526810448}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-121726-0012","method":"golf","status":"measured","mos":{"status":"measured","value":3.215839147567749,"predicted_mos":3.215839147567749,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96960,"input_sample_rate":24000,"input_seconds":4.04,"input_float32_sha256":"9e85f2cb65aa5f1a2677be9748d79ad846ab37f6c9343984712057ca57a85c92","internal_samples":64640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5728327020005963},"started_at":"2026-10-13T05:01:53.168076+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0019193320003978442,"native_input_samples":97080,"native_input_sha256":"860bfee4fe05677f8aa5194dc04ed6b2df2f07a650010902030a9c324650eaad","output_sha256":"fbae8a1f22c76c721011b020e8c9438d578cd1c5bc1562dd4cdd42d8b2811eb2","output_samples":96960,"length_delta_samples":-120,"analysis_s":0.427734650000275,"synthesis_s":0.08786477299963735,"total_s":0.5155994229999123,"rtf":0.12746586477130095,"output_peak":0.5933609008789062,"output_rms":0.05381772264402648,"finished_at":"2026-10-13T05:01:53.693632+00:00","analysis_stages":{"external_f0_s":0.04108920499947999},"evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"fbae8a1f22c76c721011b020e8c9438d578cd1c5bc1562dd4cdd42d8b2811eb2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":64720,"output_samples":64640,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.378722994452453,"output_peak":0.5936322963624834},"common_prefix":{"samples":64640,"sample_rate":16000,"reference_coverage":0.9987639060568603,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64640,"output_samples":64640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.0547612581468275,"spectral_convergence":0.5793761751388775,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.346175488566171,"spectral_convergence":0.2707705714935542,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.492997292217909,"spectral_convergence":0.2888062572251451,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003144183524659153,"output_power":0.0002916200258576163,"delta_db":-0.32690543557598767,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005552251097744475,"output_power":0.0005179843035237927,"delta_db":-0.30152499118229137,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001091461375067352,"output_power":0.0008607884952496544,"delta_db":-1.0311191752770215,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008214850357591343,"output_power":0.0007600513304629657,"delta_db":-0.33756733122561655,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00047384131656732305,"output_power":0.0003850025205106022,"delta_db":-0.9016935359110827,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00010436433261915653,"output_power":8.274897630831441e-05,"delta_db":-1.0078947046029925,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6511175289624987,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6431438934450748,"pitch":null,"output_rms_dbfs":-25.378722994452453,"output_peak":0.5936322963624834,"reference_rms_dbfs":-24.735579101007374,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.964644679643636},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64640,"output_samples":64640,"intelligibility":{"estoi":{"value":0.9303074524936535,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05797236860999668,"rms_dbfs":-24.735579101007374,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.29843540515104,"mean":-1.1353917641214805e-05,"zero_energy":false},"output":{"rms":0.05383489251324782,"rms_dbfs":-25.378722994452453,"peak":0.5936322963624834,"peak_dbfs":-4.529649588864222,"crest_factor_db":20.84907340558823,"mean":0.00040229591634001374,"zero_energy":false},"rms_difference_db":-0.6431438934450782},"full_scale":{"reference":{"sample_count":64640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":2.9904032933189386e-07,"output_rms_dbfs":-65.2427023767824,"output_zero_energy":false,"output_peak":0.005312845531015662,"output_energy_fraction":4.035318176315541e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":115,"both_estimated_voiced_frames":110,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9565217391304348,"absolute_difference_cents_median":2.922377191301482,"absolute_difference_cents_p95":13.163316746155084,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":0.00017261805365043103,"log_rms_mae_db":2.776296726510843,"rms_envelope_nrmse":0.13574382139829883,"envelope_correlation":0.987389446260142,"rms_gain_error_db":-0.6431439507661613,"reference_mean_power":0.0033649600693529597,"output_mean_power":0.002901786927698362},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":1.9830808326752542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004602051439035623,"output_power":0.00044936474068359465,"reference_active":true,"delta_db":-0.1035247606065631},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011504714634262295,"output_power":0.00010283978148978567,"reference_active":true,"delta_db":-0.48714705593648266},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.742851756199668e-05,"output_power":1.943715337856604e-05,"reference_active":true,"delta_db":-2.845699645796601}],"active_band_count":3,"active_band_mae_db":1.1454571541132157}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":0.00013379439656886598,"log_rms_mae_db":3.312247434143875,"rms_envelope_nrmse":0.09850984841038972,"envelope_correlation":0.9955268239929326,"rms_gain_error_db":-0.6431454723129726,"reference_mean_power":0.0033901657325970088,"output_mean_power":0.0029235221091049616},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":1.8843909992883445e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000422755116386511,"output_power":0.00041366717998716856,"reference_active":true,"delta_db":-0.09437806646701155},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5813018054290785e-05,"output_power":4.7471145821883804e-05,"reference_active":true,"delta_db":-0.7030579284611015},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3583249605043095e-06,"output_power":1.6453540166224648e-06,"reference_active":true,"delta_db":-3.098633615954604}],"active_band_count":3,"active_band_mae_db":1.298689870294239}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":0.00011786303332041996,"log_rms_mae_db":1.3516172087898124,"rms_envelope_nrmse":0.14635571955821292,"envelope_correlation":0.9873759081946466,"rms_gain_error_db":-0.7469683422996832,"reference_mean_power":0.0013246098256647055,"output_mean_power":0.0011152985510772973},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":9.694876337410067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023571903230875784,"output_power":0.0002051605101523043,"reference_active":true,"delta_db":-0.6030087921677783},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.915518113993878e-05,"output_power":4.618859111786613e-05,"reference_active":true,"delta_db":-1.074580733136752},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3521897588080491e-05,"output_power":1.0934645617697611e-05,"reference_active":true,"delta_db":-0.9223292986299185}],"active_band_count":3,"active_band_mae_db":0.8666396079781497}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":0.00010411406065660645,"log_rms_mae_db":1.1498114005884557,"rms_envelope_nrmse":0.1178186976149046,"envelope_correlation":0.9935957708887033,"rms_gain_error_db":-0.7469683471648846,"reference_mean_power":0.001334531993617383,"output_mean_power":0.0011236528435333614},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":9.214712542309975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002181094246432695,"output_power":0.0001897431315111321,"reference_active":true,"delta_db":-0.6050836823942084},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.732690253398315e-05,"output_power":2.2115408332390965e-05,"reference_active":true,"delta_db":-0.9189544534770158},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1925868097252376e-06,"output_power":8.800642546026535e-07,"reference_active":true,"delta_db":-1.3197561994658946}],"active_band_count":3,"active_band_mae_db":0.9479314451123729}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":0.0001403728741075033,"log_rms_mae_db":1.1452557987029097,"rms_envelope_nrmse":0.1637509303410055,"envelope_correlation":0.9812776281059838,"rms_gain_error_db":-0.46109756939431157,"reference_mean_power":0.0010448817546624529,"output_mean_power":0.00093963111330994},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":7.136771389584324e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021759481433422247,"output_power":0.0002305318360436873,"reference_active":true,"delta_db":0.25082368025451324},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.811881210755672e-05,"output_power":5.41042960160491e-05,"reference_active":true,"delta_db":-0.3108497861574717},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3032230634419253e-05,"output_power":6.990263588824154e-06,"reference_active":true,"delta_db":-5.178424482932197}],"active_band_count":3,"active_band_mae_db":1.9133659831147274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":0.0001001506434690652,"log_rms_mae_db":0.9311623758414067,"rms_envelope_nrmse":0.11039077027833426,"envelope_correlation":0.9923861648691019,"rms_gain_error_db":-0.4610975701273523,"reference_mean_power":0.001052708581437675,"output_mean_power":0.0009466695458949774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":6.804652052047638e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020210498399501184,"output_power":0.00021122515219820475,"reference_active":true,"delta_db":0.1916860818016802},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0052957500586122e-05,"output_power":2.6144582457139336e-05,"reference_active":true,"delta_db":-0.6050550689577071},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.124952972622094e-06,"output_power":7.475570641631119e-07,"reference_active":true,"delta_db":-4.537049733331719}],"active_band_count":3,"active_band_mae_db":1.7779302946970352}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":1.7245171636860885,"rms_envelope_nrmse":0.25464247381826516,"envelope_correlation":0.9601644406407182,"rms_gain_error_db":-1.2431771074540303,"reference_mean_power":0.00018403318820424335,"output_mean_power":0.00013822240308597605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":1.2575612496047337e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.407807949424467e-05,"output_power":3.316169551800886e-05,"reference_active":true,"delta_db":-1.2358593640486812},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0760228804590543e-05,"output_power":8.32284855277734e-06,"reference_active":true,"delta_db":-1.1154951419074863},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.617882024387646e-06,"output_power":3.053579929472334e-06,"reference_active":true,"delta_db":-1.7963354113339123}],"active_band_count":3,"active_band_mae_db":1.38256330576336}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":1.6075936287700134,"rms_envelope_nrmse":0.2088942306455786,"envelope_correlation":0.9770255848913332,"rms_gain_error_db":-1.2431771087517518,"reference_mean_power":0.00018541171298779382,"output_mean_power":0.0001392577761211185},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":1.1874631216114719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.197073694176745e-05,"output_power":3.241400450268215e-05,"reference_active":true,"delta_db":-1.1221390671054274},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.317287132961811e-06,"output_power":3.931714942709828e-06,"reference_active":true,"delta_db":-1.3110808973239951},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.04952091675133e-07,"output_power":2.452426356988833e-07,"reference_active":true,"delta_db":-2.1780767156531438}],"active_band_count":3,"active_band_mae_db":1.5370988933608556}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-121726-0012","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.5662779808044434,"predicted_mos":3.5662779808044434,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":97200,"input_sample_rate":24000,"input_seconds":4.05,"input_float32_sha256":"797fe2124f768432a8a59627a058f7b54b95f471649f86a1931ee66f00288f82","internal_samples":64800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7107435040015844},"started_at":"2026-10-13T05:04:21.010102+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00210355199942569,"native_input_samples":97080,"native_input_sha256":"860bfee4fe05677f8aa5194dc04ed6b2df2f07a650010902030a9c324650eaad","output_sha256":"f857d2b6a5a08cc7983c1b369e8f80a67c6e6c3c772e18e3b0f171a6c4623b0f","output_samples":97200,"length_delta_samples":120,"analysis_s":1.7239220569999816,"synthesis_s":8.188567793999937,"total_s":9.912489850999918,"rtf":2.450553733250907,"output_peak":0.7061163783073425,"output_rms":0.056863159267178685,"finished_at":"2026-10-13T05:04:30.931103+00:00","analysis_stages":{"external_f0_s":1.174545002000741},"evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"f857d2b6a5a08cc7983c1b369e8f80a67c6e6c3c772e18e3b0f171a6c4623b0f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":64720,"output_samples":64800,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.90162039964995,"output_peak":0.6656253938811241},"common_prefix":{"samples":64720,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987654320987654,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.427023558999042,"spectral_convergence":0.5411474768629255,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.3547290651216555,"spectral_convergence":0.267733342007854,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.367223079232134,"spectral_convergence":0.28379187619255825,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.000351329952120437,"delta_db":0.4874447481922507,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.0005586623732289574,"delta_db":0.031067789283129223,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.0010454372519434515,"delta_db":-0.18116946556495395,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.000776248554225169,"delta_db":-0.2406183360010653,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.00042756981040228227,"delta_db":-0.4408829923607259,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":7.94799059978602e-05,"delta_db":-1.177581723216519,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.42646084243644067,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1553047370102242,"pitch":null,"output_rms_dbfs":-24.8962554446071,"output_peak":0.6656253938811241,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.716325234450944},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64720,"output_samples":64720,"intelligibility":{"estoi":{"value":0.9523772774631517,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057936527896013786,"rms_dbfs":-24.740950707596877,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.303807011740542,"mean":-1.1361573623638366e-05,"zero_energy":false},"output":{"rms":0.05690982206893779,"rms_dbfs":-24.8962554446071,"peak":0.6656253938811241,"peak_dbfs":-3.53540235778192,"crest_factor_db":21.360853086825177,"mean":-2.981724019592687e-05,"zero_energy":false},"rms_difference_db":-0.15530473701022274},"full_scale":{"reference":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":4.120004697386313e-06,"output_rms_dbfs":-53.851022888096345,"output_zero_energy":false,"output_peak":0.11324418539126181,"output_energy_fraction":0.0004968917849373061,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":122,"both_estimated_voiced_frames":105,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9130434782608695,"absolute_difference_cents_median":4.07690215896431,"absolute_difference_cents_p95":21.25423480850251,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00017261805365043103,"log_rms_mae_db":1.3071339569834743,"rms_envelope_nrmse":0.13283303502821797,"envelope_correlation":0.984081897536026,"rms_gain_error_db":-0.15530473366395933,"reference_mean_power":0.003360795516302892,"output_mean_power":0.0032427361698851875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.9828935159721717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046042689591210996,"output_power":0.00042548382674314773,"reference_active":true,"delta_db":-0.3427762850992578},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011507848708673394,"output_power":0.00010177812456403438,"reference_active":true,"delta_db":-0.5333969968272482},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.734946402574847e-05,"output_power":3.213032617247992e-05,"reference_active":true,"delta_db":-0.6536923962627005}],"active_band_count":3,"active_band_mae_db":0.5099552260630689}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00013379439656886598,"log_rms_mae_db":1.2185019698799455,"rms_envelope_nrmse":0.07187974543668635,"envelope_correlation":0.9953157444625478,"rms_gain_error_db":-0.15530478719172908,"reference_mean_power":0.0033859385941668236,"output_mean_power":0.003266995971556777},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.884361456130501e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004231625539622664,"output_power":0.00038958238074060427,"reference_active":true,"delta_db":-0.35907922374239337},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5839616945196026e-05,"output_power":4.9492736413957815e-05,"reference_active":true,"delta_db":-0.5240096413300616},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3554104634271333e-06,"output_power":2.3590169644092485e-06,"reference_active":true,"delta_db":-1.5301459044874406}],"active_band_count":3,"active_band_mae_db":0.8044115898532985}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00011786303332041996,"log_rms_mae_db":1.2204439109097216,"rms_envelope_nrmse":0.13782472181498526,"envelope_correlation":0.9846768891125357,"rms_gain_error_db":-0.07869702153919071,"reference_mean_power":0.0013229704572187174,"output_mean_power":0.0012992132589853362},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":9.694412540428036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023578263015181652,"output_power":0.00021843605614628694,"reference_active":true,"delta_db":-0.33187481172688554},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.91091063998743e-05,"output_power":5.440345475297122e-05,"reference_active":true,"delta_db":-0.3602791447795125},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3510218569590892e-05,"output_power":1.3723415028937984e-05,"reference_active":true,"delta_db":0.0679982250823333}],"active_band_count":3,"active_band_mae_db":0.2533840605295771}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00010411406065660645,"log_rms_mae_db":0.954933606408827,"rms_envelope_nrmse":0.08147369978309826,"envelope_correlation":0.9944143427924633,"rms_gain_error_db":-0.0786970159295971,"reference_mean_power":0.0013328679887454193,"output_mean_power":0.0013089330576076619},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":9.215200160073525e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002182221456522984,"output_power":0.00020178969120872113,"reference_active":true,"delta_db":-0.33999845877472984},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7295024423763924e-05,"output_power":2.501479089511471e-05,"reference_active":true,"delta_db":-0.37886610393108655},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19264005709586e-06,"output_power":1.0737151029264407e-06,"reference_active":true,"delta_db":-0.4562032990598805}],"active_band_count":3,"active_band_mae_db":0.3916892872552323}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.0001403728741075033,"log_rms_mae_db":1.2118739416040198,"rms_envelope_nrmse":0.1655534915939136,"envelope_correlation":0.979659310177382,"rms_gain_error_db":-0.27382386556263455,"reference_mean_power":0.0010435885844267173,"output_mean_power":0.0009798214287101723},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":7.131098310180302e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021758763787569643,"output_power":0.00020314830301890603,"reference_active":true,"delta_db":-0.2982101868260725},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802785604261901e-05,"output_power":5.453325997754475e-05,"reference_active":true,"delta_db":-0.2697506454454595},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.302295499160149e-05,"output_power":1.7816354916065077e-05,"reference_active":true,"delta_db":-1.1134220887371078}],"active_band_count":3,"active_band_mae_db":0.5604609736695466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.0001001506434690652,"log_rms_mae_db":0.9323476342915386,"rms_envelope_nrmse":0.08895889504709205,"envelope_correlation":0.9944105074310929,"rms_gain_error_db":-0.2738238638640665,"reference_mean_power":0.0010513959774729575,"output_mean_power":0.0009871517613011375},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":6.799409328431787e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000202136265751209,"output_power":0.00018818735567612818,"reference_active":true,"delta_db":-0.3105379852861466},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0065691390851225e-05,"output_power":2.8413241203860505e-05,"reference_active":true,"delta_db":-0.2455041720333331},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1255469357025125e-06,"output_power":1.459775558682168e-06,"reference_active":true,"delta_db":-1.6318461147722338}],"active_band_count":3,"active_band_mae_db":0.7292960906972379}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":1.3487542790900342,"rms_envelope_nrmse":0.201168994873707,"envelope_correlation":0.9720525163779795,"rms_gain_error_db":-0.7301273095812632,"reference_mean_power":0.00018380542444204246,"output_mean_power":0.0001553622825320894},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.2570064040799187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4129018402509474e-05,"output_power":3.7226495227902874e-05,"reference_active":true,"delta_db":-0.7387211667830542},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0833298446779832e-05,"output_power":8.330913199598382e-06,"reference_active":true,"delta_db":-1.1406809815251417},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.606324451113386e-06,"output_power":4.109369628955316e-06,"reference_active":true,"delta_db":-0.4957931780295349}],"active_band_count":3,"active_band_mae_db":0.7917317754459102}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":1.0351648361301706,"rms_envelope_nrmse":0.12662314420749896,"envelope_correlation":0.9917551482611761,"rms_gain_error_db":-0.7301273171763645,"reference_mean_power":0.00018518052639831994,"output_mean_power":0.00015652459277933035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.1869756093173805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2060123324004174e-05,"output_power":3.520428588502899e-05,"reference_active":true,"delta_db":-0.7727500261301673},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.372206568913875e-06,"output_power":4.276952868865331e-06,"reference_active":true,"delta_db":-0.9901823910468979},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.05541130391232e-07,"output_power":3.199184914898162e-07,"reference_active":true,"delta_db":-1.029955640626166}],"active_band_count":3,"active_band_mae_db":0.9309626859344103}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-121726-0012","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.5473530292510986,"predicted_mos":3.5473530292510986,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":97320,"input_sample_rate":24000,"input_seconds":4.055,"input_float32_sha256":"4de693dd42485999c3e02809dd2cc1631906dbfca2fe9c97f6f644774c96cacf","internal_samples":64880,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1115138920013123},"started_at":"2026-10-13T05:09:59.202045+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0024603170004411368,"native_input_samples":97080,"native_input_sha256":"860bfee4fe05677f8aa5194dc04ed6b2df2f07a650010902030a9c324650eaad","output_sha256":"1bfe0f3462f86b9823064ff437dd2ae5d64e485608013e8979c8bbbbae587cb7","output_samples":97320,"length_delta_samples":240,"analysis_s":0.40750392399968405,"synthesis_s":0.07735845499882998,"total_s":0.48486237899851403,"rtf":0.1198670899872717,"output_peak":0.6554796099662781,"output_rms":0.05707464358903399,"finished_at":"2026-10-13T05:09:59.696692+00:00","analysis_stages":{"external_f0_s":0.062260668999442714},"evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"1bfe0f3462f86b9823064ff437dd2ae5d64e485608013e8979c8bbbbae587cb7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":64720,"output_samples":64880,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.86982427672094,"output_peak":0.6206508081105805},"common_prefix":{"samples":64720,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975339087546239,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.056050088503805,"spectral_convergence":0.16760091467086746,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.6839974778399127,"spectral_convergence":0.1681220990394856,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.024587419940184,"spectral_convergence":0.18846433230708284,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.00031234709482066655,"delta_db":-0.023332308174316543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.0005489759281643569,"delta_db":-0.04489341359922717,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.001061945647553057,"delta_db":-0.11312628481811708,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.0007953034328270172,"delta_db":-0.13529780871178326,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.0004604858133479841,"delta_db":-0.1187907745618374,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":8.749655662738901e-05,"delta_db":-0.7602455514361834,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19928102355024416,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11815024322342953,"pitch":null,"output_rms_dbfs":-24.859100950820302,"output_peak":0.6206508081105805,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":3.921544995427967},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64720,"output_samples":64720,"intelligibility":{"estoi":{"value":0.971866975352932,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057936527896013786,"rms_dbfs":-24.740950707596877,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.303807011740542,"mean":-1.1361573623638366e-05,"zero_energy":false},"output":{"rms":0.05715377916936439,"rms_dbfs":-24.859100950820302,"peak":0.6206508081105805,"peak_dbfs":-4.143053496117985,"crest_factor_db":20.71604745470232,"mean":-1.3389096924191895e-05,"zero_energy":false},"rms_difference_db":-0.11815024322342538},"full_scale":{"reference":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":7.916429279509052e-07,"output_rms_dbfs":-61.01470663598149,"output_zero_energy":false,"output_peak":0.037434493625393236,"output_energy_fraction":9.466250493858648e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":113,"both_estimated_voiced_frames":109,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9478260869565217,"absolute_difference_cents_median":0.49499498437823675,"absolute_difference_cents_p95":3.163862821749486,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00017261805365043103,"log_rms_mae_db":0.520739056120122,"rms_envelope_nrmse":0.04716547571470481,"envelope_correlation":0.9981455972159845,"rms_gain_error_db":-0.11815024527059993,"reference_mean_power":0.003360795516302892,"output_mean_power":0.003270597230099317},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.9828935159721717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046042689591210996,"output_power":0.00045329474481256576,"reference_active":true,"delta_db":-0.06780000905234818},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011507848708673394,"output_power":0.00010707597600506374,"reference_active":true,"delta_db":-0.3130210196390071},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.734946402574847e-05,"output_power":2.1246661272088827e-05,"reference_active":true,"delta_db":-2.4499367981085642}],"active_band_count":3,"active_band_mae_db":0.9435859422666398}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00013379439656886598,"log_rms_mae_db":0.46595619452609727,"rms_envelope_nrmse":0.02922417154265738,"envelope_correlation":0.9993756725353551,"rms_gain_error_db":-0.11815022832226489,"reference_mean_power":0.0033859385941668236,"output_mean_power":0.0032950655215010015},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.884361456130501e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004231625539622664,"output_power":0.0004149853929808616,"reference_active":true,"delta_db":-0.0847441931153977},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5839616945196026e-05,"output_power":5.0179895657895264e-05,"reference_active":true,"delta_db":-0.46412676479180914},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3554104634271333e-06,"output_power":1.704661575934683e-06,"reference_active":true,"delta_db":-2.9410748247620226}],"active_band_count":3,"active_band_mae_db":1.1633152608897432}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00011786303332041996,"log_rms_mae_db":0.53487812972069,"rms_envelope_nrmse":0.043764261856611496,"envelope_correlation":0.9985464372360551,"rms_gain_error_db":-0.09597430901562654,"reference_mean_power":0.0013229704572187174,"output_mean_power":0.0012940549409580994},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":9.694412540428036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023578263015181652,"output_power":0.00022906515424362084,"reference_active":true,"delta_db":-0.12552779293498273},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.91091063998743e-05,"output_power":5.2308406481290665e-05,"reference_active":true,"delta_db":-0.5308290389796064},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3510218569590892e-05,"output_power":8.295083175972468e-06,"reference_active":true,"delta_db":-2.118416300215553}],"active_band_count":3,"active_band_mae_db":0.9249243773767141}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00010411406065660645,"log_rms_mae_db":0.3566762793681753,"rms_envelope_nrmse":0.02653184685864955,"envelope_correlation":0.9995027383216492,"rms_gain_error_db":-0.09597430447013233,"reference_mean_power":0.0013328679887454193,"output_mean_power":0.0013037361483582935},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":9.215200160073525e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002182221456522984,"output_power":0.00021104177961912008,"reference_active":true,"delta_db":-0.14530381200996095},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7295024423763924e-05,"output_power":2.404694331883959e-05,"reference_active":true,"delta_db":-0.5502360754507644},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19264005709586e-06,"output_power":6.007427743224809e-07,"reference_active":true,"delta_db":-2.978208360107249}],"active_band_count":3,"active_band_mae_db":1.2245827491893246}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.0001403728741075033,"log_rms_mae_db":0.5514614968292469,"rms_envelope_nrmse":0.06979802802502566,"envelope_correlation":0.996448548984043,"rms_gain_error_db":-0.10563143585547914,"reference_mean_power":0.0010435885844267173,"output_mean_power":0.0010185120610795898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":7.131098310180302e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021758763787569643,"output_power":0.00021861867810678441,"reference_active":true,"delta_db":0.020530464620196823},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802785604261901e-05,"output_power":5.988965764555441e-05,"reference_active":true,"delta_db":0.13715305515197876},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.302295499160149e-05,"output_power":1.4976625877740368e-05,"reference_active":true,"delta_db":-1.8674708340845259}],"active_band_count":3,"active_band_mae_db":0.6750514512855671}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.0001001506434690652,"log_rms_mae_db":0.43279732553795125,"rms_envelope_nrmse":0.0394920324771633,"envelope_correlation":0.998903623190494,"rms_gain_error_db":-0.10563142991237293,"reference_mean_power":0.0010513959774729575,"output_mean_power":0.0010261318506856359},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":6.799409328431787e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000202136265751209,"output_power":0.00020329697763166114,"reference_active":true,"delta_db":0.02486683813900397},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0065691390851225e-05,"output_power":2.981255024073757e-05,"reference_active":true,"delta_db":-0.03672067068305135},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1255469357025125e-06,"output_power":1.25659980224147e-06,"reference_active":true,"delta_db":-2.282737123396501}],"active_band_count":3,"active_band_mae_db":0.7814415440728522}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":0.6582230960671239,"rms_envelope_nrmse":0.09344391380559715,"envelope_correlation":0.9944479239206014,"rms_gain_error_db":-0.3619883870555527,"reference_mean_power":0.00018380542444204246,"output_mean_power":0.00016910618239832526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.2570064040799187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4129018402509474e-05,"output_power":3.9806050759907065e-05,"reference_active":true,"delta_db":-0.4477517471742762},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0833298446779832e-05,"output_power":9.228709965640782e-06,"reference_active":true,"delta_db":-0.6961971024462655},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.606324451113386e-06,"output_power":3.1599932546515955e-06,"reference_active":true,"delta_db":-1.6366836910518059}],"active_band_count":3,"active_band_mae_db":0.9268775135574492}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":0.4768700161118682,"rms_envelope_nrmse":0.05878343104752509,"envelope_correlation":0.9985162895511698,"rms_gain_error_db":-0.3619883802389705,"reference_mean_power":0.00018518052639831994,"output_mean_power":0.00017037131530771565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.1869756093173805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2060123324004174e-05,"output_power":3.776488165938009e-05,"reference_active":true,"delta_db":-0.467824136186453},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.372206568913875e-06,"output_power":4.211046580672303e-06,"reference_active":true,"delta_db":-1.0576265813877403},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.05541130391232e-07,"output_power":2.375005065966584e-07,"reference_active":true,"delta_db":-2.3237036703937286}],"active_band_count":3,"active_band_mae_db":1.2830514626559741}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-121726-0012","method":"magphase","status":"measured","mos":{"status":"measured","value":3.724271059036255,"predicted_mos":3.724271059036255,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":194342,"input_sample_rate":48000,"input_seconds":4.048791666666666,"input_float32_sha256":"745349d709fdf9171414f293f850710890de8fd9d22991719ac7e43ac3d8cfb0","internal_samples":64781,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.990769440999429},"started_at":"2026-10-13T05:12:09.155181+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005977967999569955,"native_input_samples":194160,"native_input_sha256":"6122a70e8f4ec3e4d4ed8f731d31b12101133e6cd093a307019ba78b65d9f9b2","output_sha256":"693be4e67da55e2f0e2b1a8c3f8f48cd653698553b0403c6e5fcea097bb59737","output_samples":194342,"length_delta_samples":182,"analysis_s":2.1606700789998285,"synthesis_s":0.7526617530002113,"total_s":2.91333183200004,"rtf":0.7202303663782547,"output_peak":0.8558050373189906,"output_rms":0.057370117836010225,"finished_at":"2026-10-13T05:12:12.083141+00:00","evaluation_input":{"reference":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"693be4e67da55e2f0e2b1a8c3f8f48cd653698553b0403c6e5fcea097bb59737","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":64720,"output_samples":64781,"length_delta_samples":61,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.825651630773994,"output_peak":0.8408008367701911},"common_prefix":{"samples":64720,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990583658788842,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64720,"output_samples":64720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.012485704192495,"spectral_convergence":0.3070058292028344,"spectral_convergence_status":"ok","amplitude_floor":1.1807658352589429e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.319564964444623,"spectral_convergence":0.157568690524928,"spectral_convergence_status":"ok","amplitude_floor":7.24953185483845e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.673918386755324,"spectral_convergence":0.18364246139565035,"spectral_convergence_status":"ok","amplitude_floor":4.6531510695198173e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00031402968364977683,"output_power":0.000286893961025161,"delta_db":-0.39249295031149756,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005546801738700932,"output_power":0.000521258829937521,"delta_db":-0.26989218177973373,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010899709362407707,"output_power":0.0010516452822107456,"delta_db":-0.15545639876397221,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008204698714335104,"output_power":0.0008083275538423833,"delta_db":-0.06475254853758294,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004732551323003662,"output_power":0.0005117821368371825,"delta_db":0.33989790823788724,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001042354671510656,"output_power":0.00011500542013711339,"delta_db":0.4270279176402697,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2749199842118239,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0806095316635389,"pitch":null,"output_rms_dbfs":-24.821560239260414,"output_peak":0.8408008367701911,"reference_rms_dbfs":-24.740950707596877,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.66865635179748},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64720,"output_samples":64720,"intelligibility":{"estoi":{"value":0.9891018267887846,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057936527896013786,"rms_dbfs":-24.740950707596877,"peak":0.755340576171875,"peak_dbfs":-2.4371436958563346,"crest_factor_db":22.303807011740542,"mean":-1.1361573623638366e-05,"zero_energy":false},"output":{"rms":0.057401334339805266,"rms_dbfs":-24.821560239260414,"peak":0.8408008367701911,"peak_dbfs":-1.506137295340913,"crest_factor_db":23.3154229439195,"mean":0.0005368325843314808,"zero_energy":false},"rms_difference_db":-0.08060953166353713},"full_scale":{"reference":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.979699244606309e-06,"threshold_dbfs":-55.25827569145064,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":158,"quiet_samples":25280,"reference_mean_power":1.2681338364329143e-07,"output_mean_power":6.332004067921575e-07,"output_rms_dbfs":-61.98458814802208,"output_zero_energy":false,"output_peak":0.06165472834725884,"output_energy_fraction":7.506470291619751e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":115,"output_estimated_voiced_frames":116,"both_estimated_voiced_frames":107,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9304347826086956,"absolute_difference_cents_median":1.8562266355393227,"absolute_difference_cents_p95":7.209292091943128,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00017261805365043103,"log_rms_mae_db":0.614004636933203,"rms_envelope_nrmse":0.057536564721309275,"envelope_correlation":0.997112082507484,"rms_gain_error_db":-0.08060952490296605,"reference_mean_power":0.003360795516302892,"output_mean_power":0.003298991044778326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.9828935159721717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046042689591210996,"output_power":0.0004578721719890553,"reference_active":true,"delta_db":-0.024164357859887554},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011507848708673394,"output_power":0.00010834623070867988,"reference_active":true,"delta_db":-0.2618033656491932},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.734946402574847e-05,"output_power":3.783484292707563e-05,"reference_active":true,"delta_db":0.056075613552001795}],"active_band_count":3,"active_band_mae_db":0.11401444568702752}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00013379439656886598,"log_rms_mae_db":0.5206401146624896,"rms_envelope_nrmse":0.02968451929252954,"envelope_correlation":0.99927143153328,"rms_gain_error_db":-0.08060951070977232,"reference_mean_power":0.0033859385941668236,"output_mean_power":0.0033236717564771533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.884361456130501e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004231625539622664,"output_power":0.0004207051749550852,"reference_active":true,"delta_db":-0.025293754528060187},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5839616945196026e-05,"output_power":5.22123268727446e-05,"reference_active":true,"delta_db":-0.2916938217538753},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3554104634271333e-06,"output_power":3.2041718957411623e-06,"reference_active":true,"delta_db":-0.20029847648589488}],"active_band_count":3,"active_band_mae_db":0.1724286842559435}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.00011786303332041996,"log_rms_mae_db":0.5191165790167298,"rms_envelope_nrmse":0.05196584247998439,"envelope_correlation":0.9980948534948848,"rms_gain_error_db":-0.1686829133231606,"reference_mean_power":0.0013229704572187174,"output_mean_power":0.0012725705100717427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":9.694412540428036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023578263015181652,"output_power":0.0002283191645314196,"reference_active":true,"delta_db":-0.1396944134133379},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.91091063998743e-05,"output_power":5.569712938195866e-05,"reference_active":true,"delta_db":-0.25821581511338076},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3510218569590892e-05,"output_power":1.3754878067155909e-05,"reference_active":true,"delta_db":0.0779436971723976}],"active_band_count":3,"active_band_mae_db":0.15861797523303875}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.00010411406065660645,"log_rms_mae_db":0.3581537067695229,"rms_envelope_nrmse":0.03417459898716802,"envelope_correlation":0.9993138167882644,"rms_gain_error_db":-0.16868291690945808,"reference_mean_power":0.0013328679887454193,"output_mean_power":0.0012820909836953266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":9.215200160073525e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002182221456522984,"output_power":0.00021153353217117968,"reference_active":true,"delta_db":-0.13519600351918504},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7295024423763924e-05,"output_power":2.5770919015965493e-05,"reference_active":true,"delta_db":-0.24953581062183272},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19264005709586e-06,"output_power":1.1598564812775014e-06,"reference_active":true,"delta_db":-0.12105138194354069}],"active_band_count":3,"active_band_mae_db":0.16859439869485282}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":0.0001403728741075033,"log_rms_mae_db":0.4387662395982016,"rms_envelope_nrmse":0.06184969282786011,"envelope_correlation":0.9971911234808952,"rms_gain_error_db":0.04543128851385931,"reference_mean_power":0.0010435885844267173,"output_mean_power":0.001054562803044103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":7.131098310180302e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021758763787569643,"output_power":0.0002156612133781384,"reference_active":true,"delta_db":-0.03862173161915239},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802785604261901e-05,"output_power":5.249024604922465e-05,"reference_active":true,"delta_db":-0.43557916428186577},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.302295499160149e-05,"output_power":2.340677087505509e-05,"reference_active":true,"delta_db":0.07180439439889863}],"active_band_count":3,"active_band_mae_db":0.18200176343330562}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":0.0001001506434690652,"log_rms_mae_db":0.3107862307685259,"rms_envelope_nrmse":0.028447701890144655,"envelope_correlation":0.9993953903904474,"rms_gain_error_db":0.045431288944769985,"reference_mean_power":0.0010513959774729575,"output_mean_power":0.0010624522975519845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":6.799409328431787e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000202136265751209,"output_power":0.0001997507683121651,"reference_active":true,"delta_db":-0.05155779878685568},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0065691390851225e-05,"output_power":2.731790968465653e-05,"reference_active":true,"delta_db":-0.416237304557754},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1255469357025125e-06,"output_power":2.118199027464998e-06,"reference_active":true,"delta_db":-0.01503935069532233}],"active_band_count":3,"active_band_mae_db":0.16094481801331065}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":808,"rms_floor":5.7697319934351994e-05,"log_rms_mae_db":0.6555262275079179,"rms_envelope_nrmse":0.13770834801962262,"envelope_correlation":0.9904480465531669,"rms_gain_error_db":0.6077635909480004,"reference_mean_power":0.00018380542444204246,"output_mean_power":0.00021141445755909344},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":808,"frequency_resolution_hz":0.24752475247524752,"power_floor":1.2570064040799187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4129018402509474e-05,"output_power":4.877748829070292e-05,"reference_active":true,"delta_db":0.43495166147058006},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0833298446779832e-05,"output_power":1.1319485273849234e-05,"reference_active":true,"delta_db":0.19065971112914995},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.606324451113386e-06,"output_power":6.131514253571333e-06,"reference_active":true,"delta_db":1.2421321750294299}],"active_band_count":3,"active_band_mae_db":0.6225811825430533}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64720,"uncovered_tail_samples":0,"metrics":{"frame_count":802,"rms_floor":3.9028161750453004e-05,"log_rms_mae_db":0.4725547642713128,"rms_envelope_nrmse":0.09817027606998706,"envelope_correlation":0.9968016845403982,"rms_gain_error_db":0.6077635833899167,"reference_mean_power":0.00018518052639831994,"output_mean_power":0.00021299611036898396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":802,"frequency_resolution_hz":0.24937655860349128,"power_floor":1.1869756093173805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2060123324004174e-05,"output_power":4.644025420831607e-05,"reference_active":true,"delta_db":0.4302404660630764},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.372206568913875e-06,"output_power":5.801956506004978e-06,"reference_active":true,"delta_db":0.33421765265463266},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.05541130391232e-07,"output_power":4.963539153012376e-07,"reference_active":true,"delta_db":0.8775654463162932}],"active_band_count":3,"active_band_mae_db":0.5473411883446674}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"hnm","status":"measured","mos":{"status":"measured","value":2.4902844429016113,"predicted_mos":2.4902844429016113,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e25b76bd555ec5acdad68dda3a5d40e35f881187d0f167664846c1f37b5a3af1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0720181969991245},"started_at":"2026-10-13T04:38:04.723127+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2508000509114936e-05,"native_input_samples":96000,"native_input_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","output_sha256":"321cc43da312b30b3755787db149d7bf27a655a1f05074266fdb147ebd54bd8b","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9240275889997065,"synthesis_s":0.5018802890008374,"total_s":1.4259078780005439,"rtf":0.23765131300009065,"output_peak":0.5120446648119916,"output_rms":0.049897315100791755,"finished_at":"2026-10-13T04:38:06.154098+00:00","evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"321cc43da312b30b3755787db149d7bf27a655a1f05074266fdb147ebd54bd8b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.256790916301071,"spectral_convergence":0.30793755931485245,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.159696525120257,"spectral_convergence":0.2810407499398818,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.304293317790257,"spectral_convergence":0.28907386284514475,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.0002650230311925606,"delta_db":0.012113625292896108,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0008253634642818528,"delta_db":0.10572255555465648,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0006110914509005155,"delta_db":0.0011358667597495158,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.0002648241445958804,"delta_db":-0.17872925706047402,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.0003351458566097611,"delta_db":-0.34465169754756464,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.00018829410668713,"delta_db":-0.16141889981608118,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.13396198367190365,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04371753285534543,"pitch":null,"output_rms_dbfs":-26.038456449568834,"output_peak":0.5120446648119916,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.240260253070528},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.256790916301071,"spectral_convergence":0.30793755931485245,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.159696525120257,"spectral_convergence":0.2810407499398818,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.304293317790257,"spectral_convergence":0.28907386284514475,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.0002650230311925606,"delta_db":0.012113625292896108,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0008253634642818528,"delta_db":0.10572255555465648,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0006110914509005155,"delta_db":0.0011358667597495158,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.0002648241445958804,"delta_db":-0.17872925706047402,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.0003351458566097611,"delta_db":-0.34465169754756464,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.00018829410668713,"delta_db":-0.16141889981608118,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.13396198367190365,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04371753285534543,"pitch":null,"output_rms_dbfs":-26.038456449568834,"output_peak":0.5120446648119916,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.240260253070528},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8993249132349539,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05014908969947167,"rms_dbfs":-25.994738916713484,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.445737665615834,"mean":-2.8959274291992187e-05,"zero_energy":false},"output":{"rms":0.049897315100791755,"rms_dbfs":-26.038456449568827,"peak":0.5120446648119916,"peak_dbfs":-5.813843091602732,"crest_factor_db":20.2246133579661,"mean":2.559893886626187e-05,"zero_energy":false},"rms_difference_db":-0.043717532855342967},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":6.628926472190292e-06,"output_rms_dbfs":-51.78556798134245,"output_zero_energy":false,"output_peak":0.14059428511168623,"output_energy_fraction":0.0008031860766525245,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":230,"output_estimated_voiced_frames":201,"both_estimated_voiced_frames":200,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.8695652173913043,"absolute_difference_cents_median":1.3230308309724954,"absolute_difference_cents_p95":9.44282078208168,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017615675305240148,"log_rms_mae_db":0.6370608911367943,"rms_envelope_nrmse":0.07795064602643043,"envelope_correlation":0.9945817421033638,"rms_gain_error_db":-0.043652049551384284,"reference_mean_power":0.0025168568603362385,"output_mean_power":0.002491685999370576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0162973895242792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002469993456433397,"output_power":0.0002472734817780539,"reference_active":true,"delta_db":0.004817413226024619},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010173002443957726,"output_power":8.831409432581365e-05,"reference_active":true,"delta_db":-0.6141912929274748},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.108431545016983e-05,"output_power":1.576188679504699e-05,"reference_active":true,"delta_db":-1.2635130123739367}],"active_band_count":3,"active_band_mae_db":0.6275072395091453}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016525538510559444,"log_rms_mae_db":0.3322045070530203,"rms_envelope_nrmse":0.044889539302204025,"envelope_correlation":0.9981157180403558,"rms_gain_error_db":-0.04360457855206946,"reference_mean_power":0.002527970677484516,"output_mean_power":0.0025027160242669785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.625293642327984e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023064746753792458,"output_power":0.00022925319987494375,"reference_active":true,"delta_db":-0.02633284317117445},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.045652222183457e-05,"output_power":4.8775152016535624e-05,"reference_active":true,"delta_db":-0.9324452902458606},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2874374178901025e-06,"output_power":1.3930422845509305e-06,"reference_active":true,"delta_db":-2.153849217734036}],"active_band_count":3,"active_band_mae_db":1.0375424503836903}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001279910087199623,"log_rms_mae_db":1.0299969421425923,"rms_envelope_nrmse":0.0635134090658168,"envelope_correlation":0.9972167268150433,"rms_gain_error_db":0.07890348486836936,"reference_mean_power":0.0009939686093891967,"output_mean_power":0.001012192272986705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.147046828130101e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001388713090731631,"output_power":0.00014002695525676845,"reference_active":true,"delta_db":0.03599116216915903},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4048550736300918e-05,"output_power":2.4562211151072334e-05,"reference_active":true,"delta_db":0.09178551351126789},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.939409383024551e-06,"output_power":5.402428931188694e-06,"reference_active":true,"delta_db":0.3891404006054549}],"active_band_count":3,"active_band_mae_db":0.17230569209529392}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011859895721274409,"log_rms_mae_db":0.6690765369091936,"rms_envelope_nrmse":0.03281249806089142,"envelope_correlation":0.9992763466308733,"rms_gain_error_db":0.07882471724347964,"reference_mean_power":0.0009986896907851761,"output_mean_power":0.0010169814667691285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.009742136856967e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012678804378904427,"output_power":0.0001281275328293387,"reference_active":true,"delta_db":0.04564162462662377},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2988815650779826e-05,"output_power":1.3295034490981434e-05,"reference_active":true,"delta_db":0.10119915411041487},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.803999059830681e-07,"output_power":2.7784338858646037e-07,"reference_active":true,"delta_db":-0.039777966737148035}],"active_band_count":3,"active_band_mae_db":0.06220624849139556}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":1.0196223622961218,"rms_envelope_nrmse":0.12022709377067145,"envelope_correlation":0.9891577024571212,"rms_gain_error_db":-0.18044214966580246,"reference_mean_power":0.00041460293864459263,"output_mean_power":0.00039772982673694267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1875762004578592e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604229953230026e-05,"output_power":8.424423971922044e-05,"reference_active":true,"delta_db":-0.09171793395886335},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2897806527831643e-05,"output_power":2.2748037709360536e-05,"reference_active":true,"delta_db":-0.02849942023179465},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2516459896674549e-05,"output_power":8.628424035870328e-06,"reference_active":true,"delta_db":-1.6155003224859086}],"active_band_count":3,"active_band_mae_db":0.5785725588921888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":0.6514633866963968,"rms_envelope_nrmse":0.06749803519142056,"envelope_correlation":0.9966079025778981,"rms_gain_error_db":-0.18044714068410098,"reference_mean_power":0.0004166837249389971,"output_mean_power":0.0003997254718161108},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0350152285144472e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.63217391903156e-05,"output_power":8.334770424702508e-05,"reference_active":true,"delta_db":-0.1522653982967325},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4669499526204744e-05,"output_power":1.2981067332282819e-05,"reference_active":true,"delta_db":-0.5310489484207935},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2709360758910158e-06,"output_power":7.586773577915136e-07,"reference_active":true,"delta_db":-2.240665844146495}],"active_band_count":3,"active_band_mae_db":0.9746600636213403}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011544071276927884,"log_rms_mae_db":1.1163756746182332,"rms_envelope_nrmse":0.19134534674698941,"envelope_correlation":0.9729594814232801,"rms_gain_error_db":-0.37082149620481464,"reference_mean_power":0.0003386242858189928,"output_mean_power":0.00031091090325034655},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7742895295757829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.994910309174717e-05,"output_power":5.1283155300407066e-05,"reference_active":true,"delta_db":-0.6780795174940364},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.552315938457758e-05,"output_power":3.247928667000235e-05,"reference_active":true,"delta_db":-1.4662591277529649},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.055143212029712e-06,"output_power":3.163515648959223e-06,"reference_active":true,"delta_db":-3.4833584677919265}],"active_band_count":3,"active_band_mae_db":1.8758990376796423}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":0.6747004061696829,"rms_envelope_nrmse":0.12314568370753567,"envelope_correlation":0.9887944131896826,"rms_gain_error_db":-0.3708402108116984,"reference_mean_power":0.000340321515207851,"output_mean_power":0.00031246788303771725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6870888960608095e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.866841336444384e-05,"output_power":5.5330769889519826e-05,"reference_active":true,"delta_db":-0.9379029964358367},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5733889899301674e-05,"output_power":1.7333987234453663e-05,"reference_active":true,"delta_db":-1.7160696613377293},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.935502106463875e-07,"output_power":4.057728354061697e-07,"reference_active":true,"delta_db":-3.4283599184975873}],"active_band_count":3,"active_band_mae_db":2.027444192090384}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0239028930664062,"predicted_mos":3.0239028930664062,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e905f5a1ed10511acaa43d35b487c6002ac8cb214699426be8ea0b05abe7c5f8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0724806229991373},"started_at":"2026-10-13T04:40:56.037335+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6940999305224977e-05,"native_input_samples":96000,"native_input_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","output_sha256":"46c9030096f83a4d41020dfd698fbbef9076957bb547c5561318c0a3dbb5e750","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2061834969999836,"synthesis_s":0.10181118600030459,"total_s":1.3079946830002882,"rtf":0.21799911383338136,"output_peak":0.7115662675182428,"output_rms":0.053822988725628455,"finished_at":"2026-10-13T04:40:57.350285+00:00","evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"46c9030096f83a4d41020dfd698fbbef9076957bb547c5561318c0a3dbb5e750","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.872344578382498,"spectral_convergence":0.5719208752068335,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.151814055510519,"spectral_convergence":0.30660944301563786,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.395868365063651,"spectral_convergence":0.3224775831273349,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.00025389983418775265,"delta_db":-0.1740983711785178,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0009497294153446622,"delta_db":0.7152690449189232,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0007150497966080178,"delta_db":0.6834366609382845,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.0003549308602360451,"delta_db":1.0931325781248493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.0004167377026489464,"delta_db":0.6016376890356814,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.0002065665063337037,"delta_db":0.24081286870361046,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5847312021499778,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6140951156323273,"pitch":null,"output_rms_dbfs":-25.380643801081156,"output_peak":0.7115662675182428,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.473342332985556},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.872344578382498,"spectral_convergence":0.5719208752068335,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.151814055510519,"spectral_convergence":0.30660944301563786,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.395868365063651,"spectral_convergence":0.3224775831273349,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.00025389983418775265,"delta_db":-0.1740983711785178,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0009497294153446622,"delta_db":0.7152690449189232,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0007150497966080178,"delta_db":0.6834366609382845,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.0003549308602360451,"delta_db":1.0931325781248493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.0004167377026489464,"delta_db":0.6016376890356814,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.0002065665063337037,"delta_db":0.24081286870361046,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5847312021499778,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6140951156323273,"pitch":null,"output_rms_dbfs":-25.380643801081156,"output_peak":0.7115662675182428,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.473342332985556},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.935451491702408,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05014908969947167,"rms_dbfs":-25.994738916713484,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.445737665615834,"mean":-2.8959274291992187e-05,"zero_energy":false},"output":{"rms":0.05382298872562846,"rms_dbfs":-25.380643801081156,"peak":0.7115662675182428,"peak_dbfs":-2.955692964936147,"crest_factor_db":22.42495083614501,"mean":6.834528133431944e-06,"zero_energy":false},"rms_difference_db":0.6140951156323275},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":8.76511799370339e-07,"output_rms_dbfs":-60.572422331732255,"output_zero_energy":false,"output_peak":0.039557163201665274,"output_energy_fraction":9.12745018598085e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":230,"output_estimated_voiced_frames":231,"both_estimated_voiced_frames":208,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.9043478260869565,"absolute_difference_cents_median":3.4108199977481966,"absolute_difference_cents_p95":17.113554355496227,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017615675305240148,"log_rms_mae_db":1.1138944268203859,"rms_envelope_nrmse":0.15509894376343394,"envelope_correlation":0.9845773408245381,"rms_gain_error_db":0.614325742222846,"reference_mean_power":0.0025168568603362385,"output_mean_power":0.002899286218105518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0162973895242792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002469993456433397,"output_power":0.00028180190803783647,"reference_active":true,"delta_db":0.5724812660743194},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010173002443957726,"output_power":0.000131442695825639,"reference_active":true,"delta_db":1.1128730885405775},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.108431545016983e-05,"output_power":2.490196627615617e-05,"reference_active":true,"delta_db":0.7227413540626866}],"active_band_count":3,"active_band_mae_db":0.8026985695591945}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016525538510559444,"log_rms_mae_db":0.7530668173983113,"rms_envelope_nrmse":0.10395258675391787,"envelope_correlation":0.9952793118535667,"rms_gain_error_db":0.6142181040018103,"reference_mean_power":0.002527970677484516,"output_mean_power":0.0029120165745651293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.625293642327984e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023064746753792458,"output_power":0.0002635517939872303,"reference_active":true,"delta_db":0.5791728628160252},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.045652222183457e-05,"output_power":7.685823713367949e-05,"reference_active":true,"delta_db":1.0424725835543778},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2874374178901025e-06,"output_power":2.4509273848163076e-06,"reference_active":true,"delta_db":0.2998122330650171}],"active_band_count":3,"active_band_mae_db":0.6404858931451399}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001279910087199623,"log_rms_mae_db":1.1483269659465523,"rms_envelope_nrmse":0.1397061436774671,"envelope_correlation":0.9925619802297978,"rms_gain_error_db":0.7500137347712468,"reference_mean_power":0.0009939686093891967,"output_mean_power":0.0011813376422925704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.147046828130101e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001388713090731631,"output_power":0.00016343973236689424,"reference_active":true,"delta_db":0.7074511305895447},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4048550736300918e-05,"output_power":2.711616355689013e-05,"reference_active":true,"delta_db":0.5213933564195512},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.939409383024551e-06,"output_power":6.728572366099416e-06,"reference_active":true,"delta_db":1.34247905227458}],"active_band_count":3,"active_band_mae_db":0.857107846427892}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011859895721274409,"log_rms_mae_db":0.82671092386164,"rms_envelope_nrmse":0.11563945339990986,"envelope_correlation":0.9964486411956166,"rms_gain_error_db":0.7499322274994148,"reference_mean_power":0.0009986896907851761,"output_mean_power":0.0011869263996541815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.009742136856967e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012678804378904427,"output_power":0.0001483125162938254,"reference_active":true,"delta_db":0.680995021038157},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2988815650779826e-05,"output_power":1.4757395399693823e-05,"reference_active":true,"delta_db":0.5544016069064894},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.803999059830681e-07,"output_power":3.3157720809621785e-07,"reference_active":true,"delta_db":0.7280680675853702}],"active_band_count":3,"active_band_mae_db":0.6544882318433389}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":1.1019419289895094,"rms_envelope_nrmse":0.20362061447319416,"envelope_correlation":0.9807153826022461,"rms_gain_error_db":0.9350131545416003,"reference_mean_power":0.00041460293864459263,"output_mean_power":0.0005142019170377789},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1875762004578592e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604229953230026e-05,"output_power":0.00010649727253680707,"reference_active":true,"delta_db":0.9262647670577873},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2897806527831643e-05,"output_power":2.730812010497285e-05,"reference_active":true,"delta_db":0.7649792273967588},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2516459896674549e-05,"output_power":1.3316095329010777e-05,"reference_active":true,"delta_db":0.2689538337834018}],"active_band_count":3,"active_band_mae_db":0.653399276079316}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":0.7609504912721985,"rms_envelope_nrmse":0.14300044475118812,"envelope_correlation":0.9946376757891296,"rms_gain_error_db":0.9350265473549648,"reference_mean_power":0.0004166837249389971,"output_mean_power":0.0005167841588368598},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0350152285144472e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.63217391903156e-05,"output_power":0.0001056345198047318,"reference_active":true,"delta_db":0.8768568060264206},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4669499526204744e-05,"output_power":1.7539992214305552e-05,"reference_active":true,"delta_db":0.7761409881976399},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2709360758910158e-06,"output_power":1.267015693097771e-06,"reference_active":true,"delta_db":-0.013417134155639795}],"active_band_count":3,"active_band_mae_db":0.5554716427932334}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011544071276927884,"log_rms_mae_db":1.23811078073131,"rms_envelope_nrmse":0.2550580271884878,"envelope_correlation":0.9546396609416137,"rms_gain_error_db":0.3406901096444659,"reference_mean_power":0.0003386242858189928,"output_mean_power":0.000366257994515293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7742895295757829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.994910309174717e-05,"output_power":7.535149088811167e-05,"reference_active":true,"delta_db":0.9930915965795183},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.552315938457758e-05,"output_power":5.9081246898484444e-05,"reference_active":true,"delta_db":1.1321725723149347},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.055143212029712e-06,"output_power":8.583319887456273e-06,"reference_active":true,"delta_db":0.851494644266998}],"active_band_count":3,"active_band_mae_db":0.9922529377204837}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":0.8851931220991425,"rms_envelope_nrmse":0.15241184462817908,"envelope_correlation":0.9833367154615495,"rms_gain_error_db":0.3407020180279726,"reference_mean_power":0.000340321515207851,"output_mean_power":0.0003680947369978385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6870888960608095e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.866841336444384e-05,"output_power":8.825281618508264e-05,"reference_active":true,"delta_db":1.089715599954047},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5733889899301674e-05,"output_power":3.2835244877264374e-05,"reference_active":true,"delta_db":1.0583482097095351},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.935502106463875e-07,"output_power":9.38526621061044e-07,"reference_active":true,"delta_db":0.21327634245423208}],"active_band_count":3,"active_band_mae_db":0.7871133840392713}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"facodec","status":"measured","mos":{"status":"measured","value":4.438041687011719,"predicted_mos":4.438041687011719,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"135ac5cf7574465691a76a72633669ce02af4ea9947058af96e785d36affb898","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.141572736998569},"started_at":"2026-10-13T04:45:29.715803+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.588000047078822e-05,"native_input_samples":96000,"native_input_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","output_sha256":"24c54c214fc98535be02b3d62c3042b20818a37ff77ec1c25c5a2f51a0a123d9","output_samples":96000,"length_delta_samples":0,"analysis_s":1.9650402490005945,"synthesis_s":6.206990754999424,"total_s":8.172031004000019,"rtf":1.3620051673333364,"output_peak":0.41816747188568115,"output_rms":0.04624653904114145,"finished_at":"2026-10-13T04:45:37.895992+00:00","evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"24c54c214fc98535be02b3d62c3042b20818a37ff77ec1c25c5a2f51a0a123d9","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.087632480369803,"spectral_convergence":0.31231234008384984,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.442325821561038,"spectral_convergence":0.3317122686863998,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.398344324440004,"spectral_convergence":0.35088781758571996,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.0002940784199853564,"delta_db":0.46390901978909554,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0007171112056389452,"delta_db":-0.5048647596148103,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0005261272630406918,"delta_db":-0.6490181455744617,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.00022611468860932204,"delta_db":-0.8650172844184683,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.0002455242421116217,"delta_db":-1.696046452291382,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.00012978655389788313,"delta_db":-1.7774891630204457,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9927241374514439,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7036763446291154,"pitch":null,"output_rms_dbfs":-26.6984152613426,"output_peak":0.41816747188568115,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.3094342087902815},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.087632480369803,"spectral_convergence":0.31231234008384984,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.442325821561038,"spectral_convergence":0.3317122686863998,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.398344324440004,"spectral_convergence":0.35088781758571996,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.0002940784199853564,"delta_db":0.46390901978909554,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0007171112056389452,"delta_db":-0.5048647596148103,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0005261272630406918,"delta_db":-0.6490181455744617,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.00022611468860932204,"delta_db":-0.8650172844184683,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.0002455242421116217,"delta_db":-1.696046452291382,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.00012978655389788313,"delta_db":-1.7774891630204457,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9927241374514439,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7036763446291154,"pitch":null,"output_rms_dbfs":-26.6984152613426,"output_peak":0.41816747188568115,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.3094342087902815},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9196832860194792,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05014908969947167,"rms_dbfs":-25.994738916713484,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.445737665615834,"mean":-2.8959274291992187e-05,"zero_energy":false},"output":{"rms":0.04624653904114145,"rms_dbfs":-26.6984152613426,"peak":0.41816747188568115,"peak_dbfs":-7.572995055904153,"crest_factor_db":19.12542020543845,"mean":-0.00010411838021410939,"zero_energy":false},"rms_difference_db":-0.7036763446291161},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":4.2562716917656395e-07,"output_rms_dbfs":-63.709706573886336,"output_zero_energy":false,"output_peak":0.05598447471857071,"output_energy_fraction":6.003412611642522e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":230,"output_estimated_voiced_frames":237,"both_estimated_voiced_frames":224,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9739130434782609,"absolute_difference_cents_median":3.761065369386315,"absolute_difference_cents_p95":15.97265055667405,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017615675305240148,"log_rms_mae_db":1.061545652010687,"rms_envelope_nrmse":0.13076992911507537,"envelope_correlation":0.9893742506774303,"rms_gain_error_db":-0.7038078662191223,"reference_mean_power":0.0025168568603362385,"output_mean_power":0.0021403151729564493},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0162973895242792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002469993456433397,"output_power":0.00021737470326025057,"reference_active":true,"delta_db":-0.5548680056501398},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010173002443957726,"output_power":7.912603543587849e-05,"reference_active":true,"delta_db":-1.0912974263302648},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.108431545016983e-05,"output_power":1.6547362096648637e-05,"reference_active":true,"delta_db":-1.0523073480325442}],"active_band_count":3,"active_band_mae_db":0.8994909266709831}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016525538510559444,"log_rms_mae_db":0.920004697071866,"rms_envelope_nrmse":0.1223163479698875,"envelope_correlation":0.9909651683086699,"rms_gain_error_db":-0.7043919879758169,"reference_mean_power":0.002527970677484516,"output_mean_power":0.0021494771531999956},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.625293642327984e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023064746753792458,"output_power":0.0001997106132087248,"reference_active":true,"delta_db":-0.6254754533708563},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.045652222183457e-05,"output_power":4.7253007281770317e-05,"reference_active":true,"delta_db":-1.0701370750182848},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2874374178901025e-06,"output_power":1.6355626179346766e-06,"reference_active":true,"delta_db":-1.4568204516246825}],"active_band_count":3,"active_band_mae_db":1.0508109933379413}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001279910087199623,"log_rms_mae_db":1.2238257455758634,"rms_envelope_nrmse":0.11030159062419352,"envelope_correlation":0.9926163912476761,"rms_gain_error_db":-0.4250118815763966,"reference_mean_power":0.0009939686093891967,"output_mean_power":0.0009013043787224123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.147046828130101e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001388713090731631,"output_power":0.0001383007499691444,"reference_active":true,"delta_db":-0.01787994225178845},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4048550736300918e-05,"output_power":2.5327240831556e-05,"reference_active":true,"delta_db":0.2249897086117355},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.939409383024551e-06,"output_power":4.592126574162859e-06,"reference_active":true,"delta_db":-0.3166117225572843}],"active_band_count":3,"active_band_mae_db":0.18649379114026943}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011859895721274409,"log_rms_mae_db":1.1310291084294806,"rms_envelope_nrmse":0.10176117985909079,"envelope_correlation":0.9937589880170565,"rms_gain_error_db":-0.42494384143751696,"reference_mean_power":0.0009986896907851761,"output_mean_power":0.0009055995179015034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.009742136856967e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012678804378904427,"output_power":0.0001268246597765524,"reference_active":true,"delta_db":0.0012540477043252202},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2988815650779826e-05,"output_power":1.3449956366328193e-05,"reference_active":true,"delta_db":0.15151322486889085},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.803999059830681e-07,"output_power":2.7729778905627565e-07,"reference_active":true,"delta_db":-0.048314567916194914}],"active_band_count":3,"active_band_mae_db":0.06702728016313698}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":1.2202975372206573,"rms_envelope_nrmse":0.18106866513697908,"envelope_correlation":0.9841645035217096,"rms_gain_error_db":-1.109790441891395,"reference_mean_power":0.00041460293864459263,"output_mean_power":0.0003211096312176695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1875762004578592e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604229953230026e-05,"output_power":6.352295451859869e-05,"reference_active":true,"delta_db":-1.3178131927596457},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2897806527831643e-05,"output_power":2.1458190566338826e-05,"reference_active":true,"delta_db":-0.28200783669089113},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2516459896674549e-05,"output_power":1.0177742941463213e-05,"reference_active":true,"delta_db":-0.8983003447110336}],"active_band_count":3,"active_band_mae_db":0.8327071247205234}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":0.9773534929773222,"rms_envelope_nrmse":0.1640296465294195,"envelope_correlation":0.9885605647710407,"rms_gain_error_db":-1.1098704006456785,"reference_mean_power":0.0004166837249389971,"output_mean_power":0.0003227152568478808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0350152285144472e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.63217391903156e-05,"output_power":6.418233459333508e-05,"reference_active":true,"delta_db":-1.2870467158306935},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4669499526204744e-05,"output_power":1.4311011874980651e-05,"reference_active":true,"delta_db":-0.10744955346099726},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2709360758910158e-06,"output_power":1.0905631983657566e-06,"reference_active":true,"delta_db":-0.6647286934924317}],"active_band_count":3,"active_band_mae_db":0.6864083209280407}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011544071276927884,"log_rms_mae_db":1.6624417519564672,"rms_envelope_nrmse":0.2630884776252975,"envelope_correlation":0.9704307720178499,"rms_gain_error_db":-1.858200587535135,"reference_mean_power":0.0003386242858189928,"output_mean_power":0.0002207486434053922},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7742895295757829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.994910309174717e-05,"output_power":3.8904616926112876e-05,"reference_active":true,"delta_db":-1.8778154653680912},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.552315938457758e-05,"output_power":2.6445559566274798e-05,"reference_active":true,"delta_db":-2.358796344415999},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.055143212029712e-06,"output_power":5.318720087096047e-06,"reference_active":true,"delta_db":-1.2269869897479577}],"active_band_count":3,"active_band_mae_db":1.821199599844016}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":1.4490139970015932,"rms_envelope_nrmse":0.23655721689496995,"envelope_correlation":0.9813122845556396,"rms_gain_error_db":-1.8583982938227959,"reference_mean_power":0.000340321515207851,"output_mean_power":0.00022184496519360726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6870888960608095e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.866841336444384e-05,"output_power":4.288054531318849e-05,"reference_active":true,"delta_db":-2.044967131522445},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5733889899301674e-05,"output_power":1.5324123061983378e-05,"reference_active":true,"delta_db":-2.251298075546798},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.935502106463875e-07,"output_power":4.98051165061671e-07,"reference_active":true,"delta_db":-2.538450011803612}],"active_band_count":3,"active_band_mae_db":2.2782384062909515}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.686676025390625,"predicted_mos":3.686676025390625,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"c3db80ecc32adcb3b252cd6cc57f66bba9c9af95ebdaf79c32bf7ce793164779","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0684049949995824},"started_at":"2026-10-13T04:58:49.049525+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008029811999222147,"native_input_samples":264600,"native_input_sha256":"92deb31dcf4c7cecd4ad90535ad5cada14b25f296b03957bfa6eb4bbbe488da4","output_sha256":"c5bfecc9b512343cc47b269812e765edc16acdd0d92b9e24b0ff427b07683157","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1417986149999706,"synthesis_s":3.8865895560002173,"total_s":4.028388171000188,"rtf":0.6713980285000313,"output_peak":0.406786173582077,"output_rms":0.04558788172805714,"finished_at":"2026-10-13T04:58:53.100846+00:00","evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"c5bfecc9b512343cc47b269812e765edc16acdd0d92b9e24b0ff427b07683157","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.81839920451999,"output_peak":0.4045041730195339},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.96036545779527,"spectral_convergence":0.5418848326318275,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.105759768508408,"spectral_convergence":0.4702784541761658,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.905513669173853,"spectral_convergence":0.5338733353388507,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.0001926326696235038,"delta_db":-1.37342310969327,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0007352256593849697,"delta_db":-0.39652329312122286,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0006652459920268921,"delta_db":0.3698964533829533,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.000177445628411764,"delta_db":-1.9176519941144257,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.00021680712482649025,"delta_db":-2.2362547425033656,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":9.389746410430575e-05,"delta_db":-3.1832475406714162,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5794995222477757,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8220091045587031,"pitch":null,"output_rms_dbfs":-26.81674802127219,"output_peak":0.4045041730195339,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.990546298492511},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7864017276740397,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05014908969947167,"rms_dbfs":-25.994738916713484,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.445737665615834,"mean":-2.8959274291992187e-05,"zero_energy":false},"output":{"rms":0.0456207687175252,"rms_dbfs":-26.81674802127219,"peak":0.4045041730195339,"peak_dbfs":-7.861539873617213,"crest_factor_db":18.955208147654975,"mean":-6.984556926652992e-05,"zero_energy":false},"rms_difference_db":-0.8220091045587061},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":2.914234156991667e-07,"output_rms_dbfs":-65.35475555859172,"output_zero_energy":false,"output_peak":0.029325601079706994,"output_energy_fraction":4.224025883499036e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":230,"output_estimated_voiced_frames":232,"both_estimated_voiced_frames":208,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9043478260869565,"absolute_difference_cents_median":8.725030461526107,"absolute_difference_cents_p95":34.80358882339039,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017615675305240148,"log_rms_mae_db":1.2579810064188222,"rms_envelope_nrmse":0.13347510192814332,"envelope_correlation":0.9907477169201693,"rms_gain_error_db":-0.8218533376495667,"reference_mean_power":0.0025168568603362385,"output_mean_power":0.002082922844426188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0162973895242792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002469993456433397,"output_power":0.00021261270122225888,"reference_active":true,"delta_db":-0.6510659753525033},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010173002443957726,"output_power":8.721985212143596e-05,"reference_active":true,"delta_db":-0.6683380282732239},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.108431545016983e-05,"output_power":1.5669237901695006e-05,"reference_active":true,"delta_db":-1.289116308509892}],"active_band_count":3,"active_band_mae_db":0.869506770711873}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016525538510559444,"log_rms_mae_db":1.0295304777863705,"rms_envelope_nrmse":0.10859442276741749,"envelope_correlation":0.9963784746708862,"rms_gain_error_db":-0.8218420648156283,"reference_mean_power":0.002527970677484516,"output_mean_power":0.0020921259467816962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.625293642327984e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023064746753792458,"output_power":0.00019645086181285025,"reference_active":true,"delta_db":-0.6969475214163676},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.045652222183457e-05,"output_power":5.1341861026618536e-05,"reference_active":true,"delta_db":-0.7097155383432838},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2874374178901025e-06,"output_power":1.5283291525048258e-06,"reference_active":true,"delta_db":-1.7513232376593981}],"active_band_count":3,"active_band_mae_db":1.0526620991396831}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001279910087199623,"log_rms_mae_db":2.048671840308755,"rms_envelope_nrmse":0.1645467128831583,"envelope_correlation":0.9824325933501112,"rms_gain_error_db":0.30820865957870525,"reference_mean_power":0.0009939686093891967,"output_mean_power":0.0010670715354357165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.147046828130101e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001388713090731631,"output_power":0.00016613913450228408,"reference_active":true,"delta_db":0.7785941428137976},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4048550736300918e-05,"output_power":3.210691416561545e-05,"reference_active":true,"delta_db":1.2550965786128676},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.939409383024551e-06,"output_power":5.512823745809108e-06,"reference_active":true,"delta_db":0.4769908518087754}],"active_band_count":3,"active_band_mae_db":0.8368938577451468}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011859895721274409,"log_rms_mae_db":1.6780756076076138,"rms_envelope_nrmse":0.13582975374997788,"envelope_correlation":0.9878872703575446,"rms_gain_error_db":0.3092647111015523,"reference_mean_power":0.0009986896907851761,"output_mean_power":0.0010724005743313863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.009742136856967e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012678804378904427,"output_power":0.00015430138918663005,"reference_active":true,"delta_db":0.8529153489574387},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2988815650779826e-05,"output_power":1.7258402966852726e-05,"reference_active":true,"delta_db":1.2343105218599022},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.803999059830681e-07,"output_power":3.303230590567327e-07,"reference_active":true,"delta_db":0.7116102812335547}],"active_band_count":3,"active_band_mae_db":0.9329453840169654}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":2.6571026174797243,"rms_envelope_nrmse":0.2964484311092522,"envelope_correlation":0.9474401790198235,"rms_gain_error_db":-1.666222065035686,"reference_mean_power":0.00041460293864459263,"output_mean_power":0.00028249461233460034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1875762004578592e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604229953230026e-05,"output_power":4.7934501364605525e-05,"reference_active":true,"delta_db":-2.54063794640289},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2897806527831643e-05,"output_power":2.4549524145677976e-05,"reference_active":true,"delta_db":0.3024919688518893},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2516459896674549e-05,"output_power":8.060680602698036e-06,"reference_active":true,"delta_db":-1.9110979924069704}],"active_band_count":3,"active_band_mae_db":1.5847426358872498}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":2.363795269612291,"rms_envelope_nrmse":0.2740057917749139,"envelope_correlation":0.9564729278935723,"rms_gain_error_db":-1.6661789148903863,"reference_mean_power":0.0004166837249389971,"output_mean_power":0.000283915201547474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0350152285144472e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.63217391903156e-05,"output_power":4.884657811547774e-05,"reference_active":true,"delta_db":-2.4728603665094866},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4669499526204744e-05,"output_power":1.5775360268631253e-05,"reference_active":true,"delta_db":0.3156398876314441},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2709360758910158e-06,"output_power":8.790003590153618e-07,"reference_active":true,"delta_db":-1.6013465499411157}],"active_band_count":3,"active_band_mae_db":1.4632822680273485}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011544071276927884,"log_rms_mae_db":3.386859489621786,"rms_envelope_nrmse":0.4040656106431319,"envelope_correlation":0.9310498557564928,"rms_gain_error_db":-3.2753004326180992,"reference_mean_power":0.0003386242858189928,"output_mean_power":0.00015928983368738612},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7742895295757829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.994910309174717e-05,"output_power":2.7714860227881145e-05,"reference_active":true,"delta_db":-3.3506999724162605},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.552315938457758e-05,"output_power":1.8430199626030964e-05,"reference_active":true,"delta_db":-3.9270235591832674},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.055143212029712e-06,"output_power":2.7633519212157046e-06,"reference_active":true,"delta_db":-4.070696367506668}],"active_band_count":3,"active_band_mae_db":3.7828066330353987}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":3.1776011893579,"rms_envelope_nrmse":0.3861983570795518,"envelope_correlation":0.9416664784824537,"rms_gain_error_db":-3.2752273156631047,"reference_mean_power":0.000340321515207851,"output_mean_power":0.0001600909104878911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6870888960608095e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.866841336444384e-05,"output_power":3.028907134015459e-05,"reference_active":true,"delta_db":-3.5547105469575535},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5733889899301674e-05,"output_power":1.0409122430278709e-05,"reference_active":true,"delta_db":-3.930913218652366},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.935502106463875e-07,"output_power":3.13321979640394e-07,"reference_active":true,"delta_db":-4.551280996837886}],"active_band_count":3,"active_band_mae_db":4.012301587482601}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"golf","status":"measured","mos":{"status":"measured","value":3.5304722785949707,"predicted_mos":3.5304722785949707,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"730ea6176018ae33bb13d78c3ea32eff7fbd0babba53b7044c9b8561a40dd54c","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.058690301999377},"started_at":"2026-10-13T05:01:53.694422+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032540190004510805,"native_input_samples":144000,"native_input_sha256":"72515e36fb1c3a4e7f7251cd4bef09c3d49ff46c8058ef40440bb282a055c5fa","output_sha256":"784076db01cc02c3834aca0ca899dc35e89866a12883785e596b41c473e37814","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.7195505469999262,"synthesis_s":0.12394406099974731,"total_s":0.8434946079996735,"rtf":0.14058243466661224,"output_peak":0.5409121513366699,"output_rms":0.0492435011423073,"finished_at":"2026-10-13T05:01:54.551137+00:00","analysis_stages":{"external_f0_s":0.09070813299967995},"evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"784076db01cc02c3834aca0ca899dc35e89866a12883785e596b41c473e37814","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.14834487823037,"output_peak":0.536372784476902},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.939133159719438,"spectral_convergence":0.616627671551997,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.935000905125669,"spectral_convergence":0.3089199175153419,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.973434849896661,"spectral_convergence":0.32613958390241626,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026419310219024435,"output_power":0.0002693045782363947,"delta_db":0.08322262129935026,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008066865061240552,"output_power":0.0008903131434092164,"delta_db":0.42837992446486567,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006119414786285588,"output_power":0.0006664407778188049,"delta_db":0.37051671037308526,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027640849684133317,"output_power":0.0002426426686605177,"delta_db":-0.565842153526708,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00036342710967103277,"output_power":0.00024246550745708078,"delta_db":-1.7576735449472414,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019574659107593994,"output_power":0.00011636839110005747,"delta_db":-2.2585917772277444,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9107044553064991,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1595976491434437,"pitch":null,"output_rms_dbfs":-26.14834487823037,"output_peak":0.536372784476902,"reference_rms_dbfs":-25.988747229086925,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.2825229715805895},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9171089766574438,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05018369540529239,"rms_dbfs":-25.988747229086925,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.439745977989272,"mean":-2.7389717420472925e-05,"zero_energy":false},"output":{"rms":0.04927002198783832,"rms_dbfs":-26.14834487823037,"peak":0.536372784476902,"peak_dbfs":-5.410665326750317,"crest_factor_db":20.73767955148005,"mean":0.00023494536656357677,"zero_energy":false},"rms_difference_db":-0.15959764914344277},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":3.130620290446585e-07,"output_rms_dbfs":-65.04369604297486,"output_zero_energy":false,"output_peak":0.009086303690570141,"output_energy_fraction":3.896876443260196e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":229,"output_estimated_voiced_frames":236,"both_estimated_voiced_frames":219,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9563318777292577,"absolute_difference_cents_median":2.811031235840435,"absolute_difference_cents_p95":11.852190369043369,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017615675305240148,"log_rms_mae_db":1.5153701362289786,"rms_envelope_nrmse":0.1353607138061664,"envelope_correlation":0.9846224926941712,"rms_gain_error_db":-0.1596887527960357,"reference_mean_power":0.002520263859276242,"output_mean_power":0.00242927754832716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0163904057057423e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024822954213472994,"output_power":0.00023411666619202042,"reference_active":true,"delta_db":-0.25421135056809196},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010194623001864031,"output_power":7.39355431563045e-05,"reference_active":true,"delta_db":-1.3951790245856655},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1164004334544265e-05,"output_power":1.120475915592528e-05,"reference_active":true,"delta_db":-2.7619531590904907}],"active_band_count":3,"active_band_mae_db":1.4704478447480824}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016525538510559444,"log_rms_mae_db":1.2794518109982467,"rms_envelope_nrmse":0.10490255784532358,"envelope_correlation":0.9907981166061379,"rms_gain_error_db":-0.15943559487689934,"reference_mean_power":0.0025309564542671625,"output_mean_power":0.002439726332144441},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.620689298426675e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023179560004236752,"output_power":0.0002166977184631762,"reference_active":true,"delta_db":-0.29250849098069065},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.045111918378287e-05,"output_power":4.0130986752572166e-05,"reference_active":true,"delta_db":-1.779245073137774},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.287732353523494e-06,"output_power":8.337849368355036e-07,"reference_active":true,"delta_db":-4.383511692011919}],"active_band_count":3,"active_band_mae_db":2.1517550853767946}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001279910087199623,"log_rms_mae_db":1.1326509140481826,"rms_envelope_nrmse":0.15056123419541265,"envelope_correlation":0.9866667926474978,"rms_gain_error_db":0.44969142744146734,"reference_mean_power":0.0009954307347153904,"output_mean_power":0.0011040282556811461},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.1505987120302114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001396886889417044,"output_power":0.00014422731177359216,"reference_active":true,"delta_db":0.13886267586236126},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.384871053292915e-05,"output_power":2.7717720171356963e-05,"reference_active":true,"delta_db":0.6529260366048104},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9434173101932864e-06,"output_power":5.48363174815115e-06,"reference_active":true,"delta_db":0.4504100816696853}],"active_band_count":3,"active_band_mae_db":0.41406626471228564}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00011859895721274409,"log_rms_mae_db":0.793874794081901,"rms_envelope_nrmse":0.11810258874139536,"envelope_correlation":0.9922612729406407,"rms_gain_error_db":0.4502112081097188,"reference_mean_power":0.001000171888867752,"output_mean_power":0.001109419422554282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.012563021828067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001275292064263241,"output_power":0.0001337704516127232,"reference_active":true,"delta_db":0.2075053604304707},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2790427549200445e-05,"output_power":1.4890997863900772e-05,"reference_active":true,"delta_db":0.6603873932496817},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.802252036256774e-07,"output_power":2.7422838845029745e-07,"reference_active":true,"delta_db":-0.09394781952730727}],"active_band_count":3,"active_band_mae_db":0.3206135244024865}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":1.422561357017471,"rms_envelope_nrmse":0.19908912462891856,"envelope_correlation":0.9746330113014063,"rms_gain_error_db":-0.8392277002491878,"reference_mean_power":0.0004152913431734597,"output_mean_power":0.0003423182933111565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.189468129708466e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.640952710597733e-05,"output_power":5.914439118570656e-05,"reference_active":true,"delta_db":-1.6464806271812502},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.28858551549755e-05,"output_power":1.560469050175627e-05,"reference_active":true,"delta_db":-1.663119855198024},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.259606055853487e-05,"output_power":5.964139773362296e-06,"reference_active":true,"delta_db":-3.2468692707854547}],"active_band_count":3,"active_band_mae_db":2.185489917721576}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":1.1198789589544196,"rms_envelope_nrmse":0.15821336966741598,"envelope_correlation":0.9859097214008682,"rms_gain_error_db":-0.8392242624602009,"reference_mean_power":0.00041738143514653917,"output_mean_power":0.0003440413964476244},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.036217622078777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.666863103990894e-05,"output_power":5.7419664569354246e-05,"reference_active":true,"delta_db":-1.7880128565780677},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.465658021877047e-05,"output_power":8.704375769061495e-06,"reference_active":true,"delta_db":-2.2629501802896717},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.273252830538872e-06,"output_power":5.137948861554888e-07,"reference_active":true,"delta_db":-3.941248729376303}],"active_band_count":3,"active_band_mae_db":2.664070588748014}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00011544071276927884,"log_rms_mae_db":2.1720897286702914,"rms_envelope_nrmse":0.28543319763763636,"envelope_correlation":0.9687242776235315,"rms_gain_error_db":-2.1613398361442213,"reference_mean_power":0.0003391841790414643,"output_mean_power":0.00020620614495678788},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.777963622278667e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.021356734141063e-05,"output_power":3.344982626758651e-05,"reference_active":true,"delta_db":-2.5530049120195244},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.570354582431191e-05,"output_power":2.207959600391566e-05,"reference_active":true,"delta_db":-3.1595877257257863},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.047340241714703e-06,"output_power":2.6544268491371717e-06,"reference_active":true,"delta_db":-4.240544779750857}],"active_band_count":3,"active_band_mae_db":3.3177124724987226}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":1.9132472159382798,"rms_envelope_nrmse":0.2560142399765312,"envelope_correlation":0.9825468498360977,"rms_gain_error_db":-2.161340359301018,"reference_mean_power":0.0003408905215323047,"output_mean_power":0.00020724348649887802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.690276537928215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.89839333718906e-05,"output_power":3.649613255084217e-05,"reference_active":true,"delta_db":-2.7650110837362476},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5781430976790208e-05,"output_power":1.1789923873431117e-05,"reference_active":true,"delta_db":-3.3979601793842558},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.95764549918906e-07,"output_power":3.4645832845101044e-07,"reference_active":true,"delta_db":-4.12542865424324}],"active_band_count":3,"active_band_mae_db":3.429466639121248}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.7016642093658447,"predicted_mos":3.7016642093658447,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"c082802d85fc4b823bf7acc5c0a55e8ccb870f4e8738e1b8b182e0b42811a0f5","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1617340450011397},"started_at":"2026-10-13T05:04:30.931807+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00305284399928496,"native_input_samples":144000,"native_input_sha256":"72515e36fb1c3a4e7f7251cd4bef09c3d49ff46c8058ef40440bb282a055c5fa","output_sha256":"f7d43e1e5c6822cf201bc8e0866f20ebb4fac672fac3cec11452080a52e2f977","output_samples":144120,"length_delta_samples":120,"analysis_s":2.102739724999992,"synthesis_s":3.5652893750002477,"total_s":5.66802910000024,"rtf":0.9446715166667067,"output_peak":0.5523520112037659,"output_rms":0.04840085465223026,"finished_at":"2026-10-13T05:04:36.612690+00:00","analysis_stages":{"external_f0_s":1.3457387769994966},"evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"f7d43e1e5c6822cf201bc8e0866f20ebb4fac672fac3cec11452080a52e2f977","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.298805858224856,"output_peak":0.5070224783099873},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.771057296612495,"spectral_convergence":0.534762252396334,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.991250687983219,"spectral_convergence":0.3226781384505115,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.7770458398661715,"spectral_convergence":0.31693075112427044,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.00026642728547918095,"delta_db":0.035064455804995,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0007842625603141818,"delta_db":-0.11611501992650877,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0005898006686682366,"delta_db":-0.15287360772262262,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.0002627342703301865,"delta_db":-0.21313779205939382,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.0003119491786159817,"delta_db":-0.656151779181945,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.00013153814619292582,"delta_db":-1.7192690054649984,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4821019433600773,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.30066293924211945,"pitch":null,"output_rms_dbfs":-26.295401855955603,"output_peak":0.5070224783099873,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.1797846081539625},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9131293293262857,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05014908969947167,"rms_dbfs":-25.994738916713484,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.445737665615834,"mean":-2.8959274291992187e-05,"zero_energy":false},"output":{"rms":0.04844287470413676,"rms_dbfs":-26.295401855955607,"peak":0.5070224783099873,"peak_dbfs":-5.899455724986606,"crest_factor_db":20.395946130969,"mean":-0.00021327651343318768,"zero_energy":false},"rms_difference_db":-0.3006629392421232},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":1.0733905039854614e-05,"output_rms_dbfs":-49.69242251125522,"output_zero_energy":false,"output_peak":0.14907021539011295,"output_energy_fraction":0.0013798289702610562,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":230,"output_estimated_voiced_frames":252,"both_estimated_voiced_frames":222,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9652173913043478,"absolute_difference_cents_median":5.031268949110884,"absolute_difference_cents_p95":26.229624937035798,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017615675305240148,"log_rms_mae_db":1.5017966900957065,"rms_envelope_nrmse":0.14598375599027222,"envelope_correlation":0.981421942225835,"rms_gain_error_db":-0.30058985994594356,"reference_mean_power":0.0025168568603362385,"output_mean_power":0.0023485484872422307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0162973895242792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002469993456433397,"output_power":0.0002506058458699747,"reference_active":true,"delta_db":0.06295394824141291},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010173002443957726,"output_power":7.904347622564206e-05,"reference_active":true,"delta_db":-1.0958311714642734},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.108431545016983e-05,"output_power":9.438964507121635e-06,"reference_active":true,"delta_db":-3.490351521939672}],"active_band_count":3,"active_band_mae_db":1.5497122138817863}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016525538510559444,"log_rms_mae_db":1.1147837774828455,"rms_envelope_nrmse":0.09570125245127209,"envelope_correlation":0.9922141151542342,"rms_gain_error_db":-0.3000876651969682,"reference_mean_power":0.002527970677484516,"output_mean_power":0.0023591918847309353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.625293642327984e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023064746753792458,"output_power":0.00023114371470834075,"reference_active":true,"delta_db":0.009333980336438008},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.045652222183457e-05,"output_power":4.37453732993269e-05,"reference_active":true,"delta_db":-1.405110337335548},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2874374178901025e-06,"output_power":8.316740644137242e-07,"reference_active":true,"delta_db":-4.393960627078759}],"active_band_count":3,"active_band_mae_db":1.9361349815835815}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001279910087199623,"log_rms_mae_db":1.4119614014452775,"rms_envelope_nrmse":0.1286348504523721,"envelope_correlation":0.988153792061113,"rms_gain_error_db":-0.09903086382937315,"reference_mean_power":0.0009939686093891967,"output_mean_power":0.0009715599027579918},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.147046828130101e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001388713090731631,"output_power":0.0001405497395614408,"reference_active":true,"delta_db":0.05217515789505174},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4048550736300918e-05,"output_power":2.4192860182681873e-05,"reference_active":true,"delta_db":0.025983063927468294},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.939409383024551e-06,"output_power":4.3948719696611095e-06,"reference_active":true,"delta_db":-0.5072879456666781}],"active_band_count":3,"active_band_mae_db":0.1951487224963994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011859895721274409,"log_rms_mae_db":1.0462232912380665,"rms_envelope_nrmse":0.08361555981347654,"envelope_correlation":0.9948713212554516,"rms_gain_error_db":-0.09879727421466174,"reference_mean_power":0.0009986896907851761,"output_mean_power":0.0009762270548061042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.009742136856967e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012678804378904427,"output_power":0.0001285731051147968,"reference_active":true,"delta_db":0.06071831323591282},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2988815650779826e-05,"output_power":1.2996024840531003e-05,"reference_active":true,"delta_db":0.0024097984432939583},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.803999059830681e-07,"output_power":2.4375098119477243e-07,"reference_active":true,"delta_db":-0.6083149109515782}],"active_band_count":3,"active_band_mae_db":0.2238143408769283}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":1.4610038113021926,"rms_envelope_nrmse":0.19073480122889572,"envelope_correlation":0.9723775253163364,"rms_gain_error_db":-0.30783093028620734,"reference_mean_power":0.00041460293864459263,"output_mean_power":0.0003862329159778229},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1875762004578592e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604229953230026e-05,"output_power":8.073390407770574e-05,"reference_active":true,"delta_db":-0.27656054330494423},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2897806527831643e-05,"output_power":2.0712778375388084e-05,"reference_active":true,"delta_db":-0.43555523239517313},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2516459896674549e-05,"output_power":6.364803454145496e-06,"reference_active":true,"delta_db":-2.9369651506800407}],"active_band_count":3,"active_band_mae_db":1.216360308793386}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":1.0189902874625913,"rms_envelope_nrmse":0.11733883497320907,"envelope_correlation":0.9895304033470265,"rms_gain_error_db":-0.3078277883781545,"reference_mean_power":0.0004166837249389971,"output_mean_power":0.00038817160119483256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0350152285144472e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.63217391903156e-05,"output_power":7.785671364905526e-05,"reference_active":true,"delta_db":-0.4482411373455565},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4669499526204744e-05,"output_power":1.1453423855166216e-05,"reference_active":true,"delta_db":-1.074799645465368},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2709360758910158e-06,"output_power":5.104398174198744e-07,"reference_active":true,"delta_db":-3.961791627957484}],"active_band_count":3,"active_band_mae_db":1.8282774702561362}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011544071276927884,"log_rms_mae_db":1.7250694872107144,"rms_envelope_nrmse":0.2750425601682632,"envelope_correlation":0.951438578987582,"rms_gain_error_db":-1.312905662646744,"reference_mean_power":0.0003386242858189928,"output_mean_power":0.0002502808007669899},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7742895295757829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.994910309174717e-05,"output_power":4.097691853260026e-05,"reference_active":true,"delta_db":-1.6524339361838014},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.552315938457758e-05,"output_power":2.7030790200748436e-05,"reference_active":true,"delta_db":-2.2637365343755116},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.055143212029712e-06,"output_power":2.3494128141987084e-06,"reference_active":true,"delta_db":-4.775465007460188}],"active_band_count":3,"active_band_mae_db":2.8972118260065005}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":1.2727697192837912,"rms_envelope_nrmse":0.19982710173286417,"envelope_correlation":0.9812173209498096,"rms_gain_error_db":-1.3129433229880783,"reference_mean_power":0.000340321515207851,"output_mean_power":0.0002515330598003768},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6870888960608095e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.866841336444384e-05,"output_power":4.3755047013985096e-05,"reference_active":true,"delta_db":-1.9572885798908881},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5733889899301674e-05,"output_power":1.4736873963749492e-05,"reference_active":true,"delta_db":-2.421000692698613},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.935502106463875e-07,"output_power":2.830590435895206e-07,"reference_active":true,"delta_db":-4.992419266099873}],"active_band_count":3,"active_band_mae_db":3.123569512896458}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.424367904663086,"predicted_mos":3.424367904663086,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"ebde2c8b1e5059099d8bd525318e651b5c083637ac4a96e3e4d845d6d0884dd6","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.025877322999804},"started_at":"2026-10-13T05:09:59.705609+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030552019998140167,"native_input_samples":144000,"native_input_sha256":"72515e36fb1c3a4e7f7251cd4bef09c3d49ff46c8058ef40440bb282a055c5fa","output_sha256":"2b04c650d6c25d07bc94fb4ade0ffc2050146234d2bf1283c8c4b8a163a6f699","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7376505120009824,"synthesis_s":0.12447607399917615,"total_s":0.8621265860001586,"rtf":0.14368776433335975,"output_peak":0.4797019958496094,"output_rms":0.049294060906126855,"finished_at":"2026-10-13T05:10:00.579991+00:00","analysis_stages":{"external_f0_s":0.15106605300024967},"evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"2b04c650d6c25d07bc94fb4ade0ffc2050146234d2bf1283c8c4b8a163a6f699","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.140105042648614,"output_peak":0.4732870855362739},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":3.6589526848224803,"spectral_convergence":0.24027152292751983,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.4190628864864827,"spectral_convergence":0.20450261259162755,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":3.6939047967720557,"spectral_convergence":0.22298630842943198,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026428484162924415,"output_power":0.00025489976176400745,"delta_db":-0.15702824804129753,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008055137994793597,"output_power":0.0007947201659442811,"delta_db":-0.05858751502163277,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006109316451161528,"output_power":0.0006106577094344643,"delta_db":-0.0019477700342535476,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027595006700553216,"output_power":0.0002641526661398562,"delta_db":-0.18975505399084489,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003628265208503686,"output_power":0.00033491648397460544,"delta_db":-0.3476250111515441,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001954243236049976,"output_power":0.00017684494516998376,"delta_db":-0.4338596675396567,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19813387762987159,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13814643309370164,"pitch":null,"output_rms_dbfs":-26.132885349807186,"output_peak":0.4732870855362739,"reference_rms_dbfs":-25.994738916713484,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":3.590640122693673},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9598142950001851,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05014908969947167,"rms_dbfs":-25.994738916713484,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.445737665615834,"mean":-2.8959274291992187e-05,"zero_energy":false},"output":{"rms":0.04935779302630131,"rms_dbfs":-26.132885349807186,"peak":0.4732870855362739,"peak_dbfs":-6.497506917056921,"crest_factor_db":19.635378432750265,"mean":0.0001818148775974479,"zero_energy":false},"rms_difference_db":-0.13814643309370211},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":4.910479695337303e-06,"output_rms_dbfs":-53.08876080410225,"output_zero_energy":false,"output_peak":0.11370232901889445,"output_energy_fraction":0.0006080506807856595,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":230,"output_estimated_voiced_frames":227,"both_estimated_voiced_frames":220,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9565217391304348,"absolute_difference_cents_median":0.6225610205606984,"absolute_difference_cents_p95":6.5917657231079145,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017615675305240148,"log_rms_mae_db":0.608869651653114,"rms_envelope_nrmse":0.06110880956477874,"envelope_correlation":0.9968664480313779,"rms_gain_error_db":-0.1380703089191658,"reference_mean_power":0.0025168568603362385,"output_mean_power":0.0024380998401226975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0162973895242792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002469993456433397,"output_power":0.00023293755171139744,"reference_active":true,"delta_db":-0.254562962077868},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010173002443957726,"output_power":9.957928528988316e-05,"reference_active":true,"delta_db":-0.09280143942385083},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.108431545016983e-05,"output_power":1.5167086351765764e-05,"reference_active":true,"delta_db":-1.430573458599412}],"active_band_count":3,"active_band_mae_db":0.5926459533670436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016525538510559444,"log_rms_mae_db":0.4151022509808124,"rms_envelope_nrmse":0.037307035081978794,"envelope_correlation":0.9989417521488059,"rms_gain_error_db":-0.13797925947441358,"reference_mean_power":0.002527970677484516,"output_mean_power":0.00244891722654625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.625293642327984e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023064746753792458,"output_power":0.0002170885276200613,"reference_active":true,"delta_db":-0.2631181741040605},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.045652222183457e-05,"output_power":5.709492529277003e-05,"reference_active":true,"delta_db":-0.24845651621245501},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2874374178901025e-06,"output_power":1.5105255680593653e-06,"reference_active":true,"delta_db":-1.8022114021777424}],"active_band_count":3,"active_band_mae_db":0.7712620308314193}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001279910087199623,"log_rms_mae_db":0.5391392798355876,"rms_envelope_nrmse":0.04776846970007181,"envelope_correlation":0.9983758726953312,"rms_gain_error_db":-0.019082941603635172,"reference_mean_power":0.0009939686093891967,"output_mean_power":0.0009896106830881092},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.147046828130101e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001388713090731631,"output_power":0.00013710015780332038,"reference_active":true,"delta_db":-0.05574574741319809},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4048550736300918e-05,"output_power":2.3292450158599426e-05,"reference_active":true,"delta_db":-0.13873734287995002},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.939409383024551e-06,"output_power":3.647154918982838e-06,"reference_active":true,"delta_db":-1.317208113123439}],"active_band_count":3,"active_band_mae_db":0.503897067805529}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011859895721274409,"log_rms_mae_db":0.31115323732549127,"rms_envelope_nrmse":0.0274257593072726,"envelope_correlation":0.9994502486860755,"rms_gain_error_db":-0.01903961461205077,"reference_mean_power":0.0009986896907851761,"output_mean_power":0.000994320985218582},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.009742136856967e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012678804378904427,"output_power":0.0001249550020221611,"reference_active":true,"delta_db":-0.06324655296044471},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2988815650779826e-05,"output_power":1.2608185745524835e-05,"reference_active":true,"delta_db":-0.12916954658922183},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.803999059830681e-07,"output_power":1.8487393204924044e-07,"reference_active":true,"delta_db":-1.8090218542960799}],"active_band_count":3,"active_band_mae_db":0.6671459846152489}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":0.6953228526856869,"rms_envelope_nrmse":0.08979053552041967,"envelope_correlation":0.9942571604865923,"rms_gain_error_db":-0.2258856664184927,"reference_mean_power":0.00041460293864459263,"output_mean_power":0.0003935897767190832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1875762004578592e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604229953230026e-05,"output_power":8.062688761107071e-05,"reference_active":true,"delta_db":-0.2823211332549834},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2897806527831643e-05,"output_power":2.332900798616384e-05,"reference_active":true,"delta_db":0.0810239025313636},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2516459896674549e-05,"output_power":9.726026242344238e-06,"reference_active":true,"delta_db":-1.0954607517338348}],"active_band_count":3,"active_band_mae_db":0.4862685958400606}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":0.4526299202449079,"rms_envelope_nrmse":0.04881021043942227,"envelope_correlation":0.9985879514324396,"rms_gain_error_db":-0.225877532020065,"reference_mean_power":0.0004166837249389971,"output_mean_power":0.00039556584421862503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0350152285144472e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.63217391903156e-05,"output_power":8.032687741322484e-05,"reference_active":true,"delta_db":-0.3125929708166726},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4669499526204744e-05,"output_power":1.4076222740634761e-05,"reference_active":true,"delta_db":-0.17929166987907524},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2709360758910158e-06,"output_power":8.550789712412641e-07,"reference_active":true,"delta_db":-1.721174813872999}],"active_band_count":3,"active_band_mae_db":0.737686484856249}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011544071276927884,"log_rms_mae_db":0.7621689848303989,"rms_envelope_nrmse":0.13730802620402516,"envelope_correlation":0.9868653533302959,"rms_gain_error_db":-0.388125126346872,"reference_mean_power":0.0003386242858189928,"output_mean_power":0.00030967460296185863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7742895295757829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.994910309174717e-05,"output_power":5.375458226906843e-05,"reference_active":true,"delta_db":-0.47367198660110477},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.552315938457758e-05,"output_power":4.027212370790603e-05,"reference_active":true,"delta_db":-0.5322786293553373},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.055143212029712e-06,"output_power":4.5970800234094875e-06,"reference_active":true,"delta_db":-1.8602377013104332}],"active_band_count":3,"active_band_mae_db":0.9553961057556251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":0.5146226514300323,"rms_envelope_nrmse":0.08675333919505433,"envelope_correlation":0.995305032576963,"rms_gain_error_db":-0.3881170121298301,"reference_mean_power":0.000340321515207851,"output_mean_power":0.00031122731422501013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6870888960608095e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.866841336444384e-05,"output_power":6.09930164175219e-05,"reference_active":true,"delta_db":-0.5147690082946825},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5733889899301674e-05,"output_power":2.1987355404783228e-05,"reference_active":true,"delta_db":-0.6833244209865565},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.935502106463875e-07,"output_power":5.206199413239564e-07,"reference_active":true,"delta_db":-2.345981626468884}],"active_band_count":3,"active_band_mae_db":1.1813583519167077}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-123852-0000","method":"magphase","status":"measured","mos":{"status":"measured","value":3.753305435180664,"predicted_mos":3.753305435180664,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287181,"input_sample_rate":48000,"input_seconds":5.9829375,"input_float32_sha256":"3489236abc79e3d46afbaaceff32a5a2d232654ff7828699f4f6aab12da5afc5","internal_samples":95727,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9149464429992804},"started_at":"2026-10-13T05:12:12.083874+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009295515999838244,"native_input_samples":288000,"native_input_sha256":"eb24192699670f1fd8fe164ca89d5b4b19f2c1d893298245c2a6a09bcdc703c2","output_sha256":"862e3bb815f1417f11db0adcb661b69e446dbaf24a521268e9a40a271f48d637","output_samples":287181,"length_delta_samples":-819,"analysis_s":3.3577424059985788,"synthesis_s":1.1108997370010911,"total_s":4.46864214299967,"rtf":0.744773690499945,"output_peak":0.510589242642425,"output_rms":0.049873703473675056,"finished_at":"2026-10-13T05:12:16.574229+00:00","evaluation_input":{"reference":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"862e3bb815f1417f11db0adcb661b69e446dbaf24a521268e9a40a271f48d637","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95727,"length_delta_samples":-273,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.0382328610781,"output_peak":0.5030845884755845},"common_prefix":{"samples":95727,"sample_rate":16000,"reference_coverage":0.99715625,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95727,"output_samples":95727,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.167145189740851,"spectral_convergence":0.593619889699031,"spectral_convergence_status":"ok","amplitude_floor":1.0172154037614098e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.1054329586397555,"spectral_convergence":0.25549583778901686,"spectral_convergence_status":"ok","amplitude_floor":7.045944280558953e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.057971213419274,"spectral_convergence":0.2529033989180332,"spectral_convergence_status":"ok","amplitude_floor":6.386940858702166e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00026364618351556193,"output_power":0.00023492712660992535,"delta_db":-0.5008832209065932,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008080165229383154,"output_power":0.0007567717424936087,"delta_db":-0.2845533430862805,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000612506738317692,"output_power":0.0006037306107114656,"delta_db":-0.06267674558396095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002767413414168173,"output_power":0.00027688172376300253,"delta_db":0.002202483366251372,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00036390728325867385,"output_power":0.00038866858938672974,"delta_db":0.2858869647242164,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019592136469136844,"output_power":0.00022889044437261955,"delta_db":0.6754586509156439,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3019435680971578,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.053512415960851295,"pitch":null,"output_rms_dbfs":-26.0382328610781,"output_peak":0.5030845884755845,"reference_rms_dbfs":-25.984720445117247,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.443516453933293},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95727,"output_samples":95727,"intelligibility":{"estoi":{"value":0.9492501897346751,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05020696599216517,"rms_dbfs":-25.984720445117247,"peak":0.470489501953125,"peak_dbfs":-6.549001251097652,"crest_factor_db":19.435719194019597,"mean":-3.634297435676455e-05,"zero_energy":false},"output":{"rms":0.04989859955286673,"rms_dbfs":-26.038232861078097,"peak":0.5030845884755845,"peak_dbfs":-5.96717973349018,"crest_factor_db":20.07105312758792,"mean":-0.00039940191040407496,"zero_energy":false},"rms_difference_db":-0.053512415960849324},"full_scale":{"reference":{"sample_count":95727,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95727,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":47,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.103120164596476e-06,"threshold_dbfs":-55.0820140655869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":181,"quiet_samples":28960,"reference_mean_power":1.609320263506958e-07,"output_mean_power":2.4778220600057685e-07,"output_rms_dbfs":-66.05929884857589,"output_zero_energy":false,"output_peak":0.01504443570778296,"output_energy_fraction":3.0106308674198682e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":228,"output_estimated_voiced_frames":229,"both_estimated_voiced_frames":219,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9605263157894737,"absolute_difference_cents_median":3.5310334099659686,"absolute_difference_cents_p95":13.714396641977622,"difference_over_600_cents_frames":0,"uncovered_tail_samples":47,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95727,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1195,"rms_floor":0.00017615675305240148,"log_rms_mae_db":0.9861328684811247,"rms_envelope_nrmse":0.12254849641198035,"envelope_correlation":0.9867397914162896,"rms_gain_error_db":-0.052849072041043645,"reference_mean_power":0.0025232341125506173,"output_mean_power":0.0024927150753743478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.0162777505654903e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002494915574960452,"output_power":0.0002467816206123324,"reference_active":true,"delta_db":-0.04743042278679717},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010215121397549333,"output_power":0.00013096614449151022,"reference_active":true,"delta_db":1.0791551032067275},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.09778193904383e-05,"output_power":2.8535401662988233e-05,"reference_active":true,"delta_db":1.3362364802493707}],"active_band_count":3,"active_band_mae_db":0.8209406687476317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1189,"rms_floor":0.00016525538510559444,"log_rms_mae_db":0.634869742551772,"rms_envelope_nrmse":0.07119478713969901,"envelope_correlation":0.9952866766008702,"rms_gain_error_db":-0.05307083454331136,"reference_mean_power":0.002533574426532055,"output_mean_power":0.002502802517999722},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":9.614612685219072e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023294961874703707,"output_power":0.00023687653686922384,"reference_active":true,"delta_db":0.07260041045746092},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.041217873194756e-05,"output_power":7.944235698595019e-05,"reference_active":true,"delta_db":1.1892762221657711},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3209217605722515e-06,"output_power":2.9954723238187175e-06,"reference_active":true,"delta_db":1.1080481093187873}],"active_band_count":3,"active_band_mae_db":0.789974913980673}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1195,"rms_floor":0.0001279910087199623,"log_rms_mae_db":1.0516277270797052,"rms_envelope_nrmse":0.09148958034664684,"envelope_correlation":0.994085571961628,"rms_gain_error_db":-0.11762167817084135,"reference_mean_power":0.0009969243532022805,"output_mean_power":0.0009702866083553778},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":4.154436001329861e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000140530511745822,"output_power":0.00013219784144813672,"reference_active":true,"delta_db":-0.2654626379521623},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3631210089741425e-05,"output_power":2.2240107188092282e-05,"reference_active":true,"delta_db":-0.2634908520198686},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.943877418685852e-06,"output_power":4.790123542116464e-06,"reference_active":true,"delta_db":-0.13720979617738518}],"active_band_count":3,"active_band_mae_db":0.22205442871647205}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1189,"rms_floor":0.00011859895721274409,"log_rms_mae_db":0.7307519926233073,"rms_envelope_nrmse":0.054006200951804244,"envelope_correlation":0.9979503718892269,"rms_gain_error_db":-0.11775649160145256,"reference_mean_power":0.001001664370909136,"output_mean_power":0.0009748697108239884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":4.015503330473632e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012827877424671558,"output_power":0.00012105606678817662,"reference_active":true,"delta_db":-0.25168242206020314},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2582956952464403e-05,"output_power":1.1758565349595888e-05,"reference_active":true,"delta_db":-0.29428373690382265},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.839591868554536e-07,"output_power":2.9795782509499813e-07,"reference_active":true,"delta_db":0.2089887167705769}],"active_band_count":3,"active_band_mae_db":0.25165162524486756}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1195,"rms_floor":8.316292588094132e-05,"log_rms_mae_db":1.0236311442479682,"rms_envelope_nrmse":0.148331323723004,"envelope_correlation":0.9834454128248852,"rms_gain_error_db":0.03368335630020233,"reference_mean_power":0.0004159845560856241,"output_mean_power":0.0004192234260456071},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":2.191456779553608e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.677355985424721e-05,"output_power":8.91367567439776e-05,"reference_active":true,"delta_db":0.11669413566177139},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.28754459275524e-05,"output_power":2.4666853606316193e-05,"reference_active":true,"delta_db":0.3274418752165035},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2573049440909511e-05,"output_power":8.131367055590274e-06,"reference_active":true,"delta_db":-1.8927705746649215}],"active_band_count":3,"active_band_mae_db":0.7789688618477322}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1189,"rms_floor":7.786085823827924e-05,"log_rms_mae_db":0.6225683882711947,"rms_envelope_nrmse":0.07581225041242115,"envelope_correlation":0.9954210629670388,"rms_gain_error_db":0.03367721709725182,"reference_mean_power":0.0004180820480394539,"output_mean_power":0.0004213366535399393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":2.0374653429085834e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.700923622678812e-05,"output_power":8.786069776238501e-05,"reference_active":true,"delta_db":0.04229291569901393},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4644584105854576e-05,"output_power":1.368055469072406e-05,"reference_active":true,"delta_db":-0.2957333597467351},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2810724284639456e-06,"output_power":7.236328302638642e-07,"reference_active":true,"delta_db":-2.4805542230532023}],"active_band_count":3,"active_band_mae_db":0.9395268328329838}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1195,"rms_floor":0.00011544071276927884,"log_rms_mae_db":1.0991291351815782,"rms_envelope_nrmse":0.24487380410383877,"envelope_correlation":0.9637984828242656,"rms_gain_error_db":0.6408318446347354,"reference_mean_power":0.00033974948403059125,"output_mean_power":0.00039376942395934536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.781754120713965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0485172541194556e-05,"output_power":7.443193028268065e-05,"reference_active":true,"delta_db":0.9011035806390483},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.58805845915507e-05,"output_power":6.671788130873318e-05,"reference_active":true,"delta_db":1.626133035383601},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.010562172483794e-06,"output_power":2.0901814449303333e-05,"reference_active":true,"delta_db":4.744311429193545}],"active_band_count":3,"active_band_mae_db":2.423849348405398}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":47,"metrics":{"frame_count":1189,"rms_floor":8.669712818495073e-05,"log_rms_mae_db":0.7358946509878801,"rms_envelope_nrmse":0.16469481054253662,"envelope_correlation":0.9847473092172946,"rms_gain_error_db":0.6408258483206253,"reference_mean_power":0.00034146256320141796,"output_mean_power":0.0003957543351492367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":1.6935250130085004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.930468683293948e-05,"output_power":9.536011101077785e-05,"reference_active":true,"delta_db":1.38604142550292},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5822253489559612e-05,"output_power":4.153715465594317e-05,"reference_active":true,"delta_db":2.064426033815067},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.028669209681422e-07,"output_power":2.266802358694257e-06,"reference_active":true,"delta_db":3.99789914136878}],"active_band_count":3,"active_band_mae_db":2.482788866895589}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"hnm","status":"measured","mos":{"status":"measured","value":2.244995594024658,"predicted_mos":2.244995594024658,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e5ed77401447955a1ffe6b63fa05bc75cde6f7be2f4b2b3bdfa1693f3506b486","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3312942720003775},"started_at":"2026-10-13T04:38:06.155789+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4745000044058543e-05,"native_input_samples":96000,"native_input_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","output_sha256":"491fb6635c52232dde9d272614c93d9e1e5a5bcf4324e76536b05f3c73aba295","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9056811299997207,"synthesis_s":0.5357169310000245,"total_s":1.4413980609997452,"rtf":0.2402330101666242,"output_peak":0.43437158437384027,"output_rms":0.040573496626094246,"finished_at":"2026-10-13T04:38:07.604080+00:00","evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"491fb6635c52232dde9d272614c93d9e1e5a5bcf4324e76536b05f3c73aba295","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.395208882865694,"spectral_convergence":0.3725136994860082,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.353356744312127,"spectral_convergence":0.3657705079030013,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.440338545449306,"spectral_convergence":0.38787920250448316,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.0003952206636987841,"delta_db":-0.047922864209742885,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.00029743521455661664,"delta_db":-0.10995230642984472,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.00027943894914588143,"delta_db":-0.08589490468820546,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.0001670058401625163,"delta_db":-0.17231008531503134,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.0002828684821757717,"delta_db":0.4643676567178378,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.0002242394787281107,"delta_db":-0.13315121014049497,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1689331712501929,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.006984250750376292,"pitch":null,"output_rms_dbfs":-27.835151263226216,"output_peak":0.43437158437384027,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.396301390875709},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.395208882865694,"spectral_convergence":0.3725136994860082,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.353356744312127,"spectral_convergence":0.3657705079030013,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.440338545449306,"spectral_convergence":0.38787920250448316,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.0003952206636987841,"delta_db":-0.047922864209742885,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.00029743521455661664,"delta_db":-0.10995230642984472,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.00027943894914588143,"delta_db":-0.08589490468820546,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.0001670058401625163,"delta_db":-0.17231008531503134,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.0002828684821757717,"delta_db":0.4643676567178378,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.0002242394787281107,"delta_db":-0.13315121014049497,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1689331712501929,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.006984250750376292,"pitch":null,"output_rms_dbfs":-27.835151263226216,"output_peak":0.43437158437384027,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.396301390875709},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9054321861817344,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04060613455341415,"rms_dbfs":-27.828167012475838,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.2864855868614,"mean":-1.2833913167317708e-05,"zero_energy":false},"output":{"rms":0.040573496626094246,"rms_dbfs":-27.835151263226216,"peak":0.43437158437384027,"peak_dbfs":-7.242771861445167,"crest_factor_db":20.592379401781045,"mean":3.553941599843728e-05,"zero_energy":false},"rms_difference_db":-0.006984250750377896},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":220,"quiet_samples":35200,"reference_mean_power":1.6131530918011612e-07,"output_mean_power":4.2251278485812445e-07,"output_rms_dbfs":-63.741601451537456,"output_zero_energy":false,"output_peak":0.03102845713551746,"output_energy_fraction":9.410797135244176e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":171,"both_estimated_voiced_frames":157,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9457831325301205,"absolute_difference_cents_median":1.5438874367398119,"absolute_difference_cents_p95":6.595768287001355,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015167567220720444,"log_rms_mae_db":0.6396242287791585,"rms_envelope_nrmse":0.08834335345509481,"envelope_correlation":0.9926693839565393,"rms_gain_error_db":-0.006984230136486097,"reference_mean_power":0.0016502333338478435,"output_mean_power":0.0016475815970171382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103350035549282e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021122178949431673,"output_power":0.0002140380847018894,"reference_active":true,"delta_db":0.057523383870486056},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018786794884032712,"output_power":0.00019194071224814486,"reference_active":true,"delta_db":0.09314408308764054},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6179250514948536e-05,"output_power":2.2680792772207195e-05,"reference_active":true,"delta_db":-0.622989784063531}],"active_band_count":3,"active_band_mae_db":0.25788575034055256}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014379702931337134,"log_rms_mae_db":0.36777846384048646,"rms_envelope_nrmse":0.050730031905510856,"envelope_correlation":0.9973726208959894,"rms_gain_error_db":-0.006984013378362182,"reference_mean_power":0.0016585326982667058,"output_mean_power":0.0016558677079485799},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0955155560805602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019170493341265235,"output_power":0.0001974760902464542,"reference_active":true,"delta_db":0.128812308698101},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.87936190430084e-05,"output_power":8.109591912718378e-05,"reference_active":true,"delta_db":0.1250795201699297},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1061472825588798e-06,"output_power":1.756290283525974e-06,"reference_active":true,"delta_db":-0.78892439407185}],"active_band_count":3,"active_band_mae_db":0.3476054076466269}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010769717002798511,"log_rms_mae_db":1.3358652631090613,"rms_envelope_nrmse":0.1043777059413648,"envelope_correlation":0.9922474064585092,"rms_gain_error_db":-0.11624149769544909,"reference_mean_power":0.0005226470255969395,"output_mean_power":0.0005088436197443052},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.88003979644585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021484485335499559,"output_power":0.0002100230363141509,"reference_active":true,"delta_db":-0.09858021738970157},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090774498801903e-05,"output_power":3.266900002052509e-05,"reference_active":true,"delta_db":-0.9766969879415257},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.643789476493196e-06,"output_power":8.164344541504597e-06,"reference_active":true,"delta_db":-0.24782856910384599}],"active_band_count":3,"active_band_mae_db":0.4410352581450245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":0.8439096588527151,"rms_envelope_nrmse":0.06773888206026732,"envelope_correlation":0.996610609163137,"rms_gain_error_db":-0.11624149445140924,"reference_mean_power":0.0005252755682472565,"output_mean_power":0.0005114027414669704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.5461080823927677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019511319942898067,"output_power":0.00019111893511133737,"reference_active":true,"delta_db":-0.08982933537985417},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8168801206395614e-05,"output_power":1.4739789470411796e-05,"reference_active":true,"delta_db":-0.908349926139397},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.35857242269651e-07,"output_power":5.842194147064267e-07,"reference_active":true,"delta_db":-1.002175832412655}],"active_band_count":3,"active_band_mae_db":0.6667850313106354}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":1.1399544001167712,"rms_envelope_nrmse":0.15902475758229084,"envelope_correlation":0.981068551639651,"rms_gain_error_db":-0.04833042249798307,"reference_mean_power":0.0002851540241046427,"output_mean_power":0.00028199828199484506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.768616554938334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966944293503785,"output_power":0.0001107682315429225,"reference_active":true,"delta_db":0.04329584609577185},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.898846044442001e-05,"output_power":4.728371160880359e-05,"reference_active":true,"delta_db":-0.15382231305337823},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.52294853557438e-06,"output_power":5.549381043027831e-06,"reference_active":true,"delta_db":-1.8634531977460522}],"active_band_count":3,"active_band_mae_db":0.6868571189650674}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":0.6962561533414715,"rms_envelope_nrmse":0.08294138301459895,"envelope_correlation":0.9944297334977595,"rms_gain_error_db":-0.048330632356153896,"reference_mean_power":0.0002865881294452576,"output_mean_power":0.00028341650268556265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.511088390714625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010245851234810725,"output_power":0.00010382821367597558,"reference_active":true,"delta_db":0.057673363864943886},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.514042968169313e-05,"output_power":2.3891515062749675e-05,"reference_active":true,"delta_db":-0.2212900496817079},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.446669785409341e-07,"output_power":4.986006944388023e-07,"reference_active":true,"delta_db":-1.7420921690414248}],"active_band_count":3,"active_band_mae_db":0.6736851941960255}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001420862345348912,"log_rms_mae_db":1.1712853429681498,"rms_envelope_nrmse":0.1703049156650454,"envelope_correlation":0.9839100924920963,"rms_gain_error_db":0.30920538849864904,"reference_mean_power":0.000234926135231895,"output_mean_power":0.0002522620223103579},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.4459776587443103e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015814203638019883,"output_power":0.0001818871903509096,"reference_active":true,"delta_db":0.6075478750277883},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.639730090455678e-05,"output_power":0.00011177645829466579,"reference_active":true,"delta_db":0.6428547065389802},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2676469905844767e-05,"output_power":1.3897422722697835e-05,"reference_active":true,"delta_db":0.3993593806164598}],"active_band_count":3,"active_band_mae_db":0.5499206540610762}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010454414213188797,"log_rms_mae_db":0.8623826301994926,"rms_envelope_nrmse":0.10586224256065727,"envelope_correlation":0.9939457041927013,"rms_gain_error_db":0.30920536478467364,"reference_mean_power":0.00023610764979113878,"output_mean_power":0.0002535307229025087},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.234204737076497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017362277943143455,"output_power":0.0002030562286736438,"reference_active":true,"delta_db":0.6800961152888212},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.868815527339496e-05,"output_power":5.662127132300229e-05,"reference_active":true,"delta_db":0.6555629635282968},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.341220903531157e-06,"output_power":1.2164499734011896e-06,"reference_active":true,"delta_db":-0.4240606024432842}],"active_band_count":3,"active_band_mae_db":0.5865732270868007}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.621840715408325,"predicted_mos":2.621840715408325,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6b4f16a81452fb7eefe0413ed59b93c99c281bdb1f161c96a75521545df7c07a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.1698287219987833},"started_at":"2026-10-13T04:40:57.350780+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4017999749048613e-05,"native_input_samples":96000,"native_input_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","output_sha256":"d61948b38f2b49f7e27ab03ef03742c252ed806dff49be776dda7478887f62a4","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3165635729992573,"synthesis_s":0.1038113330005217,"total_s":1.420374905999779,"rtf":0.23672915099996317,"output_peak":0.6858167802718961,"output_rms":0.04489536816617872,"finished_at":"2026-10-13T04:40:58.777239+00:00","evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"d61948b38f2b49f7e27ab03ef03742c252ed806dff49be776dda7478887f62a4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.939450505967419,"spectral_convergence":0.5907747567799377,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.274716887834203,"spectral_convergence":0.3863547529455583,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.5368822860608455,"spectral_convergence":0.4140527836904754,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.00047980149621908223,"delta_db":0.7942966838069603,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.00039253451853756634,"delta_db":1.0949023781736447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.0003741495281302512,"delta_db":1.1816877193620223,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00021075482297432113,"delta_db":0.8381485479082594,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.0003170514113067654,"delta_db":0.9598189628302688,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00024130230560874734,"delta_db":0.18534275352573898,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8423661742678158,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8721977596545645,"pitch":null,"output_rms_dbfs":-26.955969252821276,"output_peak":0.6858167802718961,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.583683226620823},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.939450505967419,"spectral_convergence":0.5907747567799377,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.274716887834203,"spectral_convergence":0.3863547529455583,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.5368822860608455,"spectral_convergence":0.4140527836904754,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.00047980149621908223,"delta_db":0.7942966838069603,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.00039253451853756634,"delta_db":1.0949023781736447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.0003741495281302512,"delta_db":1.1816877193620223,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00021075482297432113,"delta_db":0.8381485479082594,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.0003170514113067654,"delta_db":0.9598189628302688,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00024130230560874734,"delta_db":0.18534275352573898,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8423661742678158,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8721977596545645,"pitch":null,"output_rms_dbfs":-26.955969252821276,"output_peak":0.6858167802718961,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.583683226620823},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9361802812777054,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04060613455341415,"rms_dbfs":-27.828167012475838,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.2864855868614,"mean":-1.2833913167317708e-05,"zero_energy":false},"output":{"rms":0.04489536816617872,"rms_dbfs":-26.955969252821276,"peak":0.6858167802718961,"peak_dbfs":-3.275837859185742,"crest_factor_db":23.680131393635534,"mean":5.837321815954644e-08,"zero_energy":false},"rms_difference_db":0.8721977596545614},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":220,"quiet_samples":35200,"reference_mean_power":1.6131530918011612e-07,"output_mean_power":2.260950238735055e-07,"output_rms_dbfs":-66.45708995938848,"output_zero_energy":false,"output_peak":0.008019604678335944,"output_energy_fraction":4.113006158433024e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":177,"both_estimated_voiced_frames":148,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.891566265060241,"absolute_difference_cents_median":3.6151293089196663,"absolute_difference_cents_p95":23.670631051977974,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015167567220720444,"log_rms_mae_db":1.1539041360843316,"rms_envelope_nrmse":0.17273765842196087,"envelope_correlation":0.9841883203736395,"rms_gain_error_db":0.8721978019774157,"reference_mean_power":0.0016502333338478435,"output_mean_power":0.0020172751357345063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103350035549282e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021122178949431673,"output_power":0.0002803094776020259,"reference_active":true,"delta_db":1.2289906433715927},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018786794884032712,"output_power":0.0002230636215173414,"reference_active":true,"delta_db":0.7457605511764549},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6179250514948536e-05,"output_power":3.156564445357176e-05,"reference_active":true,"delta_db":0.8125745145121603}],"active_band_count":3,"active_band_mae_db":0.9291085696867359}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014379702931337134,"log_rms_mae_db":0.8749158096693265,"rms_envelope_nrmse":0.13044141003865273,"envelope_correlation":0.9946094714869405,"rms_gain_error_db":0.8721979987462946,"reference_mean_power":0.0016585326982667058,"output_mean_power":0.002027420521073619},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0955155560805602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019170493341265235,"output_power":0.00025801234432872857,"reference_active":true,"delta_db":1.2900719548960722},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.87936190430084e-05,"output_power":9.169138775587233e-05,"reference_active":true,"delta_db":0.6583749748810274},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1061472825588798e-06,"output_power":2.381665034140035e-06,"reference_active":true,"delta_db":0.5339194265924223}],"active_band_count":3,"active_band_mae_db":0.827455452123174}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010769717002798511,"log_rms_mae_db":1.3870459688485868,"rms_envelope_nrmse":0.20772812982028097,"envelope_correlation":0.9890584485225747,"rms_gain_error_db":1.2801765534021157,"reference_mean_power":0.0005226470255969395,"output_mean_power":0.0007018206434952751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.88003979644585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021484485335499559,"output_power":0.0002958865025109649,"reference_active":true,"delta_db":1.3900019980419154},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090774498801903e-05,"output_power":5.884820096156832e-05,"reference_active":true,"delta_db":1.5792765067257042},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.643789476493196e-06,"output_power":1.151499319288257e-05,"reference_active":true,"delta_db":1.2455950466982}],"active_band_count":3,"active_band_mae_db":1.4049578504886064}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":0.9954891948271407,"rms_envelope_nrmse":0.17669570982227153,"envelope_correlation":0.9962083263713748,"rms_gain_error_db":1.2801766682609887,"reference_mean_power":0.0005252755682472565,"output_mean_power":0.0007053503206075587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.5461080823927677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019511319942898067,"output_power":0.0002702613767525872,"reference_active":true,"delta_db":1.414973343400734},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8168801206395614e-05,"output_power":2.6060054657056012e-05,"reference_active":true,"delta_db":1.566490491264988},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.35857242269651e-07,"output_power":9.234583433579397e-07,"reference_active":true,"delta_db":0.986237409207411}],"active_band_count":3,"active_band_mae_db":1.3225670812910444}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":1.228641819347102,"rms_envelope_nrmse":0.2114657368390284,"envelope_correlation":0.9791524780033339,"rms_gain_error_db":0.9449255344279168,"reference_mean_power":0.0002851540241046427,"output_mean_power":0.00035446393809736364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.768616554938334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966944293503785,"output_power":0.00014337724264140963,"reference_active":true,"delta_db":1.163945865180621},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.898846044442001e-05,"output_power":6.340991481104146e-05,"reference_active":true,"delta_db":1.1206337858784539},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.52294853557438e-06,"output_power":8.525293659287819e-06,"reference_active":true,"delta_db":0.001194814436479169}],"active_band_count":3,"active_band_mae_db":0.7619248218318514}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":0.8823370622670704,"rms_envelope_nrmse":0.15352134196276898,"envelope_correlation":0.9926330168513788,"rms_gain_error_db":0.9449256610021318,"reference_mean_power":0.0002865881294452576,"output_mean_power":0.0003562466293602875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.511088390714625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010245851234810725,"output_power":0.00013462432502214696,"reference_active":true,"delta_db":1.1857549302698436},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.514042968169313e-05,"output_power":3.256333395721507e-05,"reference_active":true,"delta_db":1.1235616714963959},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.446669785409341e-07,"output_power":7.419858586987634e-07,"reference_active":true,"delta_db":-0.015664676863331752}],"active_band_count":3,"active_band_mae_db":0.7749937595431904}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001420862345348912,"log_rms_mae_db":1.273334696194218,"rms_envelope_nrmse":0.22736621846301253,"envelope_correlation":0.9757511188777314,"rms_gain_error_db":0.7438371547219947,"reference_mean_power":0.000234926135231895,"output_mean_power":0.0002788143033910164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.4459776587443103e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015814203638019883,"output_power":0.00018647870351476132,"reference_active":true,"delta_db":0.7158191424226155},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.639730090455678e-05,"output_power":0.00010624857133039616,"reference_active":true,"delta_db":0.42258225070114663},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2676469905844767e-05,"output_power":1.3178955575731077e-05,"reference_active":true,"delta_db":0.16882664306488296}],"active_band_count":3,"active_band_mae_db":0.43574267872954836}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010454414213188797,"log_rms_mae_db":0.9754675090754796,"rms_envelope_nrmse":0.15832497026510167,"envelope_correlation":0.9899090379862895,"rms_gain_error_db":0.7438371601075594,"reference_mean_power":0.00023610764979113878,"output_mean_power":0.00028021654516412057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.234204737076497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017362277943143455,"output_power":0.00020252353288194045,"reference_active":true,"delta_db":0.6686879042374474},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.868815527339496e-05,"output_power":5.422825848831755e-05,"reference_active":true,"delta_db":0.4680233748041338},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.341220903531157e-06,"output_power":1.292818329821778e-06,"reference_active":true,"delta_db":-0.15962812501281304}],"active_band_count":3,"active_band_mae_db":0.4321131346847981}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9728617668151855,"predicted_mos":3.9728617668151855,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"60e59d268fbc0e5eefab24bb221a0e4876fcc5b1ece3ea92838804238f554d75","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4804086469994218},"started_at":"2026-10-13T04:45:37.896969+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.153599977347767e-05,"native_input_samples":96000,"native_input_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","output_sha256":"8c375d411d72144a156c483a31748e2972635938a7cd2cf15c886a710e3a2c4a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5863179150001088,"synthesis_s":5.884193686999424,"total_s":7.470511601999533,"rtf":1.2450852669999222,"output_peak":0.44169527292251587,"output_rms":0.03812490057694466,"finished_at":"2026-10-13T04:45:45.374649+00:00","evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"8c375d411d72144a156c483a31748e2972635938a7cd2cf15c886a710e3a2c4a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.264301444541294,"spectral_convergence":0.3670913901881537,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.894544527676516,"spectral_convergence":0.40098379258028805,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.870813849009318,"spectral_convergence":0.42153895424523163,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.0003814363328398428,"delta_db":-0.2020987116119677,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.0003029330989571897,"delta_db":-0.03040888112749969,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.0002473316699705682,"delta_db":-0.6159670006165177,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00014289846604554937,"delta_db":-0.8493510029134965,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.00019411385921605208,"delta_db":-1.170912504684556,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00018479461697271302,"delta_db":-0.9733788003334021,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6403528168812399,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5476585916748462,"pitch":null,"output_rms_dbfs":-28.375825604150688,"output_peak":0.44169527292251587,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676553273742376},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.264301444541294,"spectral_convergence":0.3670913901881537,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.894544527676516,"spectral_convergence":0.40098379258028805,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.870813849009318,"spectral_convergence":0.42153895424523163,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.0003814363328398428,"delta_db":-0.2020987116119677,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.0003029330989571897,"delta_db":-0.03040888112749969,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.0002473316699705682,"delta_db":-0.6159670006165177,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00014289846604554937,"delta_db":-0.8493510029134965,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.00019411385921605208,"delta_db":-1.170912504684556,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00018479461697271302,"delta_db":-0.9733788003334021,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6403528168812399,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5476585916748462,"pitch":null,"output_rms_dbfs":-28.375825604150688,"output_peak":0.44169527292251587,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676553273742376},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9216601854208145,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04060613455341415,"rms_dbfs":-27.828167012475838,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.2864855868614,"mean":-1.2833913167317708e-05,"zero_energy":false},"output":{"rms":0.03812490057694466,"rms_dbfs":-28.375825604150688,"peak":0.44169527292251587,"peak_dbfs":-7.0975449718008115,"crest_factor_db":21.278280632349876,"mean":3.798948322087152e-05,"zero_energy":false},"rms_difference_db":-0.5476585916748498},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":220,"quiet_samples":35200,"reference_mean_power":1.6131530918011612e-07,"output_mean_power":1.4314354219860313e-07,"output_rms_dbfs":-68.44228240033401,"output_zero_energy":false,"output_peak":0.007299977354705334,"output_energy_fraction":3.6109855524646815e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":177,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9216867469879518,"absolute_difference_cents_median":2.7833968544441197,"absolute_difference_cents_p95":11.308691796675662,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015167567220720444,"log_rms_mae_db":0.9043793374385156,"rms_envelope_nrmse":0.12658720984969205,"envelope_correlation":0.9877576787888657,"rms_gain_error_db":-0.5476585569351602,"reference_mean_power":0.0016502333338478435,"output_mean_power":0.0014547203014190434},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103350035549282e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021122178949431673,"output_power":0.00017940590921713276,"reference_active":true,"delta_db":-0.7090197408951773},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018786794884032712,"output_power":0.00014998832141367293,"reference_active":true,"delta_db":-0.9779524896094305},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6179250514948536e-05,"output_power":2.3419569619370312e-05,"reference_active":true,"delta_db":-0.4837829919295218}],"active_band_count":3,"active_band_mae_db":0.7235850741447098}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014379702931337134,"log_rms_mae_db":0.7839412952291482,"rms_envelope_nrmse":0.1111663632540622,"envelope_correlation":0.99069724697396,"rms_gain_error_db":-0.5476581469365707,"reference_mean_power":0.0016585326982667058,"output_mean_power":0.0014620365284240516},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0955155560805602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019170493341265235,"output_power":0.00016114910925091368,"reference_active":true,"delta_db":-0.7540538003325149},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.87936190430084e-05,"output_power":6.135409634437346e-05,"reference_active":true,"delta_db":-1.0864748439800398},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1061472825588798e-06,"output_power":1.7074676731609268e-06,"reference_active":true,"delta_db":-0.9113624796003404}],"active_band_count":3,"active_band_mae_db":0.9172970413042983}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010769717002798511,"log_rms_mae_db":1.1576060454666717,"rms_envelope_nrmse":0.1163603902466049,"envelope_correlation":0.9910037023870811,"rms_gain_error_db":-0.32742383610633746,"reference_mean_power":0.0005226470255969395,"output_mean_power":0.00048469227670467426},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.88003979644585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021484485335499559,"output_power":0.0001865105032345412,"reference_active":true,"delta_db":-0.6142166066243591},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090774498801903e-05,"output_power":3.774091783840879e-05,"reference_active":true,"delta_db":-0.3499308220562838},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.643789476493196e-06,"output_power":6.8950973340868684e-06,"reference_active":true,"delta_db":-0.9816377965796204}],"active_band_count":3,"active_band_mae_db":0.6485950750867544}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":0.9458370422818951,"rms_envelope_nrmse":0.10420553757349532,"envelope_correlation":0.9925735094356246,"rms_gain_error_db":-0.3274236482548312,"reference_mean_power":0.0005252755682472565,"output_mean_power":0.0004871299550452054},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.5461080823927677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019511319942898067,"output_power":0.00016930140937436788,"reference_active":true,"delta_db":-0.6162607698489813},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8168801206395614e-05,"output_power":1.685696183843719e-05,"reference_active":true,"delta_db":-0.3254696947863811},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.35857242269651e-07,"output_power":5.803582739777115e-07,"reference_active":true,"delta_db":-1.0309738812328535}],"active_band_count":3,"active_band_mae_db":0.6575681152894053}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":1.2007607028478888,"rms_envelope_nrmse":0.17452414041398526,"envelope_correlation":0.9839902464760424,"rms_gain_error_db":-0.9739752750907393,"reference_mean_power":0.0002851540241046427,"output_mean_power":0.00022786728469447208},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.768616554938334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966944293503785,"output_power":8.214296975939385e-05,"reference_active":true,"delta_db":-1.2551523739479762},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.898846044442001e-05,"output_power":3.859793490639882e-05,"reference_active":true,"delta_db":-1.0352972177294486},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.52294853557438e-06,"output_power":7.246876668389355e-06,"reference_active":true,"delta_db":-0.7043899573240775}],"active_band_count":3,"active_band_mae_db":0.9982798496671675}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":1.0164906135595253,"rms_envelope_nrmse":0.15370371306116998,"envelope_correlation":0.9887966513035564,"rms_gain_error_db":-0.973974956357468,"reference_mean_power":0.0002865881294452576,"output_mean_power":0.0002290132986204243},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.511088390714625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010245851234810725,"output_power":7.568019488594554e-05,"reference_active":true,"delta_db":-1.315658040663006},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.514042968169313e-05,"output_power":2.0173552284040618e-05,"reference_active":true,"delta_db":-0.9559031784565666},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.446669785409341e-07,"output_power":5.952198180807277e-07,"reference_active":true,"delta_db":-0.9728471316083876}],"active_band_count":3,"active_band_mae_db":1.0814694502426534}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001420862345348912,"log_rms_mae_db":1.30412220716418,"rms_envelope_nrmse":0.21564440049955141,"envelope_correlation":0.9766713572645582,"rms_gain_error_db":-1.0143606982839222,"reference_mean_power":0.000234926135231895,"output_mean_power":0.00018599242820232693},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.4459776587443103e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015814203638019883,"output_power":0.00011711108663883364,"reference_active":true,"delta_db":-1.3044931618260323},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.639730090455678e-05,"output_power":7.445033434716926e-05,"reference_active":true,"delta_db":-1.1219822150320007},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2676469905844767e-05,"output_power":1.1743600421420139e-05,"reference_active":true,"delta_db":-0.3319706388146054}],"active_band_count":3,"active_band_mae_db":0.9194820052242126}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010454414213188797,"log_rms_mae_db":1.131889850072214,"rms_envelope_nrmse":0.16963168269888543,"envelope_correlation":0.987944362880434,"rms_gain_error_db":-1.014360610126776,"reference_mean_power":0.00023610764979113878,"output_mean_power":0.00018692784414923877},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.234204737076497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017362277943143455,"output_power":0.00013167019474739575,"reference_active":true,"delta_db":-1.2011922641149384},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.868815527339496e-05,"output_power":3.811482407784148e-05,"reference_active":true,"delta_db":-1.0632940045449304},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.341220903531157e-06,"output_power":1.2146388635153188e-06,"reference_active":true,"delta_db":-0.43053140814253427}],"active_band_count":3,"active_band_mae_db":0.898339225600801}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.0038158893585205,"predicted_mos":3.0038158893585205,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"ebc15f34292686058986e5a95bae25390c50aec1edda2c4a4904466ada7c49a1","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.225989335000122},"started_at":"2026-10-13T04:58:53.101485+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006800050000492774,"native_input_samples":264600,"native_input_sha256":"1fdbc5c941e8dc15c4a1549bcadb7c88bea1df749fcf6a67b64c0283f40deb39","output_sha256":"a931b3be0caa1f2aae0ae0205422fa05ec8a0d1c026b1cb69ac16d381913ed82","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13976756800002477,"synthesis_s":3.790906490999987,"total_s":3.9306740590000118,"rtf":0.6551123431666687,"output_peak":0.3977620005607605,"output_rms":0.03553554903996149,"finished_at":"2026-10-13T04:58:57.050347+00:00","evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"a931b3be0caa1f2aae0ae0205422fa05ec8a0d1c026b1cb69ac16d381913ed82","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.009008672335543,"output_peak":0.386041087009826},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.363815277625681,"spectral_convergence":0.5806273552764328,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.332972249859241,"spectral_convergence":0.5346833783231859,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.024539611672608,"spectral_convergence":0.5684855705352995,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.0002431716459410273,"delta_db":-2.1571899518415054,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.00033380256143546476,"delta_db":0.39102049165584124,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.0002363574925228383,"delta_db":-0.8130705592541431,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00011490794096441248,"delta_db":-1.796126245454128,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.0001582326291202939,"delta_db":-2.0585175026204077,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00017034171850237268,"delta_db":-1.3270617186036993,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4238310782382875,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1791229294531644,"pitch":null,"output_rms_dbfs":-29.007289941929002,"output_peak":0.386041087009826,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.240442379719177},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7600787361028337,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04060613455341415,"rms_dbfs":-27.828167012475838,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.2864855868614,"mean":-1.2833913167317708e-05,"zero_energy":false},"output":{"rms":0.03545157244025164,"rms_dbfs":-29.007289941929002,"peak":0.386041087009826,"peak_dbfs":-8.267329403350633,"crest_factor_db":20.73996053857837,"mean":-8.537565433609131e-05,"zero_energy":false},"rms_difference_db":-1.1791229294531647},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":220,"quiet_samples":35200,"reference_mean_power":1.6131530918011612e-07,"output_mean_power":1.7583779090041733e-07,"output_rms_dbfs":-67.54887781049936,"output_zero_energy":false,"output_peak":0.009977642474677964,"output_energy_fraction":5.1299442283525746e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":168,"both_estimated_voiced_frames":141,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.8493975903614458,"absolute_difference_cents_median":6.563805303638462,"absolute_difference_cents_p95":41.49786987064468,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015167567220720444,"log_rms_mae_db":1.3333452689234582,"rms_envelope_nrmse":0.1713620080905694,"envelope_correlation":0.9858214472556629,"rms_gain_error_db":-1.1791228926158865,"reference_mean_power":0.0016502333338478435,"output_mean_power":0.0012578621993778692},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103350035549282e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021122178949431673,"output_power":0.0001762317490267558,"reference_active":true,"delta_db":-0.7865456625527509},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018786794884032712,"output_power":0.00013301266340591514,"reference_active":true,"delta_db":-1.4995970399931124},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6179250514948536e-05,"output_power":2.5005251371861482e-05,"reference_active":true,"delta_db":-0.19925984222737703}],"active_band_count":3,"active_band_mae_db":0.8284675149244135}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014379702931337134,"log_rms_mae_db":1.064631959114158,"rms_envelope_nrmse":0.14625225208744908,"envelope_correlation":0.9938423364495415,"rms_gain_error_db":-1.1791226324668953,"reference_mean_power":0.0016585326982667058,"output_mean_power":0.0012641883240137677},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0955155560805602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019170493341265235,"output_power":0.00016222897172636392,"reference_active":true,"delta_db":-0.7250487387589589},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.87936190430084e-05,"output_power":5.609007613952848e-05,"reference_active":true,"delta_db":-1.4760501882664632},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1061472825588798e-06,"output_power":1.903192796339741e-06,"reference_active":true,"delta_db":-0.44005952847947405}],"active_band_count":3,"active_band_mae_db":0.8803861518349655}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010769717002798511,"log_rms_mae_db":2.2249594709488383,"rms_envelope_nrmse":0.21211967700292816,"envelope_correlation":0.9673866450106186,"rms_gain_error_db":-0.29459201607129054,"reference_mean_power":0.0005226470255969395,"output_mean_power":0.0004883703414675722},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.88003979644585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021484485335499559,"output_power":0.00017985072748546665,"reference_active":true,"delta_db":-0.7721275569298676},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090774498801903e-05,"output_power":3.230605231536024e-05,"reference_active":true,"delta_db":-1.0252164794696321},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.643789476493196e-06,"output_power":7.958834377840133e-06,"reference_active":true,"delta_db":-0.35854713703781227}],"active_band_count":3,"active_band_mae_db":0.7186303911457707}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":1.808578349609585,"rms_envelope_nrmse":0.17301018222261458,"envelope_correlation":0.97730497059751,"rms_gain_error_db":-0.29459219776541223,"reference_mean_power":0.0005252755682472565,"output_mean_power":0.0004908264762517039},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.5461080823927677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019511319942898067,"output_power":0.00016297988295004476,"reference_active":true,"delta_db":-0.7815264889319075},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8168801206395614e-05,"output_power":1.4541933655888695e-05,"reference_active":true,"delta_db":-0.9670411417140502},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.35857242269651e-07,"output_power":6.133228507784103e-07,"reference_active":true,"delta_db":-0.7910442289204432}],"active_band_count":3,"active_band_mae_db":0.8465372865221337}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":2.2786257610169183,"rms_envelope_nrmse":0.2864589723054416,"envelope_correlation":0.9558869367773055,"rms_gain_error_db":-1.7421222182939737,"reference_mean_power":0.0002851540241046427,"output_mean_power":0.00019092697111313523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.768616554938334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966944293503785,"output_power":6.248467501036251e-05,"reference_active":true,"delta_db":-2.4431212212906708},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.898846044442001e-05,"output_power":3.280668671084598e-05,"reference_active":true,"delta_db":-1.7413141981730456},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.52294853557438e-06,"output_power":7.22648851728103e-06,"reference_active":true,"delta_db":-0.7166254910313181}],"active_band_count":3,"active_band_mae_db":1.6336869701650114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":1.9327192655372112,"rms_envelope_nrmse":0.2499755242896341,"envelope_correlation":0.9710870829710977,"rms_gain_error_db":-1.7421222135772214,"reference_mean_power":0.0002865881294452576,"output_mean_power":0.00019188718708493355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.511088390714625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010245851234810725,"output_power":5.8677431236445174e-05,"reference_active":true,"delta_db":-2.4207695265435225},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.514042968169313e-05,"output_power":1.7342101165453382e-05,"reference_active":true,"delta_db":-1.6127098069525183},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.446669785409341e-07,"output_power":6.666115862754336e-07,"reference_active":true,"delta_db":-0.4808923815001248}],"active_band_count":3,"active_band_mae_db":1.5047905716653887}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001420862345348912,"log_rms_mae_db":2.2846114923135326,"rms_envelope_nrmse":0.270505406004625,"envelope_correlation":0.964487468695598,"rms_gain_error_db":-1.4424269425615013,"reference_mean_power":0.000234926135231895,"output_mean_power":0.0001685344313048903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.4459776587443103e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015814203638019883,"output_power":9.890227582356876e-05,"reference_active":true,"delta_db":-2.0384104167772836},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.639730090455678e-05,"output_power":5.9240659981538335e-05,"reference_active":true,"delta_db":-2.114449857439574},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2676469905844767e-05,"output_power":1.100375105968398e-05,"reference_active":true,"delta_db":-0.6145757305205852}],"active_band_count":3,"active_band_mae_db":1.589145334912481}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010454414213188797,"log_rms_mae_db":1.9438936481885092,"rms_envelope_nrmse":0.23725944401770205,"envelope_correlation":0.9749527640781419,"rms_gain_error_db":-1.4424268309785826,"reference_mean_power":0.00023610764979113878,"output_mean_power":0.00016938204626481583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.234204737076497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017362277943143455,"output_power":0.00010711723313184065,"reference_active":true,"delta_db":-2.0974735812119727},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.868815527339496e-05,"output_power":3.0049463965103587e-05,"reference_active":true,"delta_db":-2.095865907612751},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.341220903531157e-06,"output_power":1.2459852811035452e-06,"reference_active":true,"delta_db":-0.3198740147687837}],"active_band_count":3,"active_band_mae_db":1.504404501197836}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"golf","status":"measured","mos":{"status":"measured","value":3.3677804470062256,"predicted_mos":3.3677804470062256,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"86c1496d2dc8136e6926106d4c17f067844a212bca6f14a41c6b13d7a0ed3e46","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.9663253420003457},"started_at":"2026-10-13T05:01:54.551764+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002882879999560828,"native_input_samples":144000,"native_input_sha256":"d69401da71d8da5b373ebdf9dde5d574f0b758885253ac0a04da1fed429896d1","output_sha256":"3579715213111c650dbaf04957e0b0984d7181e7826c1109dc269b50b5f1b175","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6876667240003371,"synthesis_s":0.11982612199972209,"total_s":0.8074928460000592,"rtf":0.13458214100000987,"output_peak":0.4946644902229309,"output_rms":0.03701420729700557,"finished_at":"2026-10-13T05:01:55.370271+00:00","analysis_stages":{"external_f0_s":0.08556255499934196},"evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"3579715213111c650dbaf04957e0b0984d7181e7826c1109dc269b50b5f1b175","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.653304467207295,"output_peak":0.49214941684493274},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.314902982194716,"spectral_convergence":0.6038827964690168,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.357075151040198,"spectral_convergence":0.4028675019575724,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.361984048440994,"spectral_convergence":0.42375729407165996,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004002723286661304,"output_power":0.0003309470532539737,"delta_db":-0.8259704932240092,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000305570322524036,"output_power":0.00027330671216666577,"delta_db":-0.4846087487571412,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002854978032949647,"output_power":0.0002449632713256581,"delta_db":-0.6650179788964645,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017405516872744928,"output_power":0.00014010496179052634,"delta_db":-0.9423340856924955,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00025460814429081316,"output_power":0.00016153089352235027,"delta_db":-1.9761669613892776,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023160697943511784,"output_power":0.00021269235310866162,"delta_db":-0.36999766570429254,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.87734932227728,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8323814686242742,"pitch":null,"output_rms_dbfs":-28.653304467207295,"output_peak":0.49214941684493274,"reference_rms_dbfs":-27.82092299858302,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.677987393891969},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8915914536542684,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.040640014110953644,"rms_dbfs":-27.82092299858302,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.279241572968584,"mean":-1.2809485943369158e-05,"zero_energy":false},"output":{"rms":0.03692621352329312,"rms_dbfs":-28.653304467207295,"peak":0.49214941684493274,"peak_dbfs":-6.158060503197438,"crest_factor_db":22.495243964009855,"mean":0.00032147584376237095,"zero_energy":false},"rms_difference_db":-0.832381468624277},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":219,"quiet_samples":35040,"reference_mean_power":1.6177516987708878e-07,"output_mean_power":1.8023627834964057e-07,"output_rms_dbfs":-67.44157788821408,"output_zero_energy":false,"output_peak":0.003283537899820687,"output_energy_fraction":4.832701269853222e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":186,"both_estimated_voiced_frames":155,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.9337349397590361,"absolute_difference_cents_median":3.5866282984549867,"absolute_difference_cents_p95":15.105281268614714,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015167567220720444,"log_rms_mae_db":1.6054115878482647,"rms_envelope_nrmse":0.16208054717860948,"envelope_correlation":0.9818861168323592,"rms_gain_error_db":-0.8323825603565678,"reference_mean_power":0.001652990523875522,"output_mean_power":0.0013646840254111117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.2105600954129169e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021125716718216816,"output_power":0.00019696246426652384,"reference_active":true,"delta_db":-0.30427982661349046},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018795562591071604,"output_power":0.00015947730699239244,"reference_active":true,"delta_db":-0.7135643624510369},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6193922537649617e-05,"output_power":2.825201600140733e-05,"reference_active":true,"delta_db":0.32848904698770265}],"active_band_count":3,"active_band_mae_db":0.44877774535074333}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014379702931337134,"log_rms_mae_db":1.3523209366907778,"rms_envelope_nrmse":0.12123093517052619,"envelope_correlation":0.9930340682665272,"rms_gain_error_db":-0.8323877587574945,"reference_mean_power":0.001661317630140058,"output_mean_power":0.0013715571169274304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0956377789212525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019172201568611695,"output_power":0.0001802465361032184,"reference_active":true,"delta_db":-0.26805058940796367},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.86738536317732e-05,"output_power":6.601037576197376e-05,"reference_active":true,"delta_db":-0.762182184967456},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.106882032158945e-06,"output_power":2.091124073066291e-06,"reference_active":true,"delta_db":-0.03260417755963469}],"active_band_count":3,"active_band_mae_db":0.3542789839783514}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00010769717002798511,"log_rms_mae_db":1.3248308458562776,"rms_envelope_nrmse":0.16875334181146595,"envelope_correlation":0.9813413826734144,"rms_gain_error_db":-0.56730428590236,"reference_mean_power":0.0005235202625125058,"output_mean_power":0.0004594127742381774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.876800379047144e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021483555620397402,"output_power":0.00019099058806103854,"reference_active":true,"delta_db":-0.5109419465301885},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.092333801705394e-05,"output_power":3.588058494175557e-05,"reference_active":true,"delta_db":-0.5711153633089767},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.626286895426639e-06,"output_power":9.93284711824141e-06,"reference_active":true,"delta_db":0.6124985312901194}],"active_band_count":3,"active_band_mae_db":0.5648519470430949}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":0.9379144348186299,"rms_envelope_nrmse":0.11472069816085088,"envelope_correlation":0.992604282990821,"rms_gain_error_db":-0.5673043787179622,"reference_mean_power":0.000526157626115977,"output_mean_power":0.0004617271705481047},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.542915501483868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019514218398875631,"output_power":0.0001737611975859159,"reference_active":true,"delta_db":-0.5039836007426189},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8152494260747315e-05,"output_power":1.6552819978899975e-05,"reference_active":true,"delta_db":-0.4006431623385891},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.315869881847544e-07,"output_power":7.837327106310137e-07,"reference_active":true,"delta_db":0.2990200065852568}],"active_band_count":3,"active_band_mae_db":0.4012155898888216}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":1.7161081196886827,"rms_envelope_nrmse":0.25688837099865935,"envelope_correlation":0.9608763136218859,"rms_gain_error_db":-1.3450524107562818,"reference_mean_power":0.0002856304517719728,"output_mean_power":0.00020955559741333071},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.7704597986657677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010975196286445492,"output_power":9.33354167234882e-05,"reference_active":true,"delta_db":-0.7036582529292679},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.906157293996632e-05,"output_power":3.74377029870302e-05,"reference_active":true,"delta_db":-1.1743227332208048},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51755302019265e-06,"output_power":1.0011752573158677e-05,"reference_active":true,"delta_db":0.7019526242047882}],"active_band_count":3,"active_band_mae_db":0.859977870118287}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":1.39498562702917,"rms_envelope_nrmse":0.19962453891318477,"envelope_correlation":0.9820068198516329,"rms_gain_error_db":-1.3450528201626641,"reference_mean_power":0.0002870693640506278,"output_mean_power":0.00021061124974084295},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.5127134383627273e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010253807092460142,"output_power":9.230124048264512e-05,"reference_active":true,"delta_db":-0.45677604919782866},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517194026268463e-05,"output_power":1.8855029495228608e-05,"reference_active":true,"delta_db":-1.254894763055817},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.44673146169822e-07,"output_power":8.152102458044326e-07,"reference_active":true,"delta_db":0.3930393644247037}],"active_band_count":3,"active_band_mae_db":0.701570058892783}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001420862345348912,"log_rms_mae_db":1.8424188734411044,"rms_envelope_nrmse":0.2238018081795242,"envelope_correlation":0.9697228228101881,"rms_gain_error_db":-0.28715973892816127,"reference_mean_power":0.00023531865571805875,"output_mean_power":0.000220262409053372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.4532676373909597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001585162255607495,"output_power":0.00013412455994625191,"reference_active":true,"delta_db":-0.7256541268851782},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.65729914102014e-05,"output_power":8.358887452092092e-05,"reference_active":true,"delta_db":-0.6270720643345905},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2680563811495701e-05,"output_power":9.166665365106654e-06,"reference_active":true,"delta_db":-1.4092718641358561}],"active_band_count":3,"active_band_mae_db":0.9206660184518749}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010454414213188797,"log_rms_mae_db":1.616724406787206,"rms_envelope_nrmse":0.1760881921488812,"envelope_correlation":0.980274635081654,"rms_gain_error_db":-0.28715972695871317,"reference_mean_power":0.0002365041285961905,"output_mean_power":0.00022137203317042166},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.2409765221414167e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017405629118635345,"output_power":0.00014824977279578213,"reference_active":true,"delta_db":-0.6969568883570132},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.867446063814007e-05,"output_power":4.193634245437872e-05,"reference_active":true,"delta_db":-0.6471059771154369},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3417072257111778e-06,"output_power":1.007480873861046e-06,"reference_active":true,"delta_db":-1.2442094846490284}],"active_band_count":3,"active_band_mae_db":0.8627574500404928}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4166808128356934,"predicted_mos":3.4166808128356934,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"e34542527a2d6e463d17d6889e2b116b8e572a6564ae0ba4a024c0300e485ce9","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4025571860001946},"started_at":"2026-10-13T05:04:36.613339+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026403299998492002,"native_input_samples":144000,"native_input_sha256":"d69401da71d8da5b373ebdf9dde5d574f0b758885253ac0a04da1fed429896d1","output_sha256":"be4efc815810a2ec916f21737263f93cd985c58b240e48243e372239f52da489","output_samples":144120,"length_delta_samples":120,"analysis_s":2.036478992999946,"synthesis_s":3.324296897999375,"total_s":5.360775890999321,"rtf":0.8934626484998868,"output_peak":0.5514600872993469,"output_rms":0.03944433382012971,"finished_at":"2026-10-13T05:04:41.985525+00:00","analysis_stages":{"external_f0_s":1.3554188600000998},"evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"be4efc815810a2ec916f21737263f93cd985c58b240e48243e372239f52da489","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.101003374761504,"output_peak":0.5122301793525651},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.059459884565055,"spectral_convergence":0.6027924415824286,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.3748671046795495,"spectral_convergence":0.41452138416109024,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.190293448462232,"spectral_convergence":0.41040274004232746,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.0004395946155327287,"delta_db":0.4142043467390568,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.0002938723783196621,"delta_db":-0.16228848008178465,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.0002778761369950459,"delta_db":-0.11025177171979163,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00016053013032554277,"delta_db":-0.34406108866446067,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.00022404490429970382,"delta_db":-0.5481272371782796,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00015383099335141077,"delta_db":-1.7698335172373334,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5581277402701178,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26921889455956466,"pitch":null,"output_rms_dbfs":-28.097385907035406,"output_peak":0.5122301793525651,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.541540145902279},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8972799642702795,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04060613455341415,"rms_dbfs":-27.828167012475838,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.2864855868614,"mean":-1.2833913167317708e-05,"zero_energy":false},"output":{"rms":0.03936685355504164,"rms_dbfs":-28.097385907035406,"peak":0.5122301793525651,"peak_dbfs":-5.810696750847656,"crest_factor_db":22.286689156187748,"mean":-3.761156959641348e-05,"zero_energy":false},"rms_difference_db":-0.26921889455956816},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":220,"quiet_samples":35200,"reference_mean_power":1.6131530918011612e-07,"output_mean_power":1.7342640220687842e-06,"output_rms_dbfs":-57.60884785407079,"output_zero_energy":false,"output_peak":0.11436383414749368,"output_energy_fraction":0.0004103224089338355,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":166,"both_estimated_voiced_frames":149,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.8975903614457831,"absolute_difference_cents_median":4.727656206409563,"absolute_difference_cents_p95":23.87062627028919,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015167567220720444,"log_rms_mae_db":1.6730411780257812,"rms_envelope_nrmse":0.18787722619528702,"envelope_correlation":0.9666627712969418,"rms_gain_error_db":-0.2692190022675415,"reference_mean_power":0.0016502333338478435,"output_mean_power":0.0015510416325566587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103350035549282e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021122178949431673,"output_power":0.00019612901069459041,"reference_active":true,"delta_db":-0.3219688000259283},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018786794884032712,"output_power":0.00015270614879426455,"reference_active":true,"delta_db":-0.899961692229158},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6179250514948536e-05,"output_power":2.379781472197864e-05,"reference_active":true,"delta_db":-0.4142012998838538}],"active_band_count":3,"active_band_mae_db":0.5453772640463134}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014379702931337134,"log_rms_mae_db":1.1624083163443897,"rms_envelope_nrmse":0.11619942693784466,"envelope_correlation":0.9867082075767125,"rms_gain_error_db":-0.26921884283580694,"reference_mean_power":0.0016585326982667058,"output_mean_power":0.0015588421986399298},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0955155560805602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019170493341265235,"output_power":0.00017456194731524618,"reference_active":true,"delta_db":-0.4068371131542084},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.87936190430084e-05,"output_power":6.0117858566392854e-05,"reference_active":true,"delta_db":-1.1748754600543534},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1061472825588798e-06,"output_power":1.7830309791930794e-06,"reference_active":true,"delta_db":-0.7232984918719954}],"active_band_count":3,"active_band_mae_db":0.7683370216935191}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010769717002798511,"log_rms_mae_db":1.6808102178403057,"rms_envelope_nrmse":0.18270476023701415,"envelope_correlation":0.9758933840669033,"rms_gain_error_db":-0.1466247865497849,"reference_mean_power":0.0005226470255969395,"output_mean_power":0.0005052961677605693},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.88003979644585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021484485335499559,"output_power":0.00020514283660272764,"reference_active":true,"delta_db":-0.20068598138486063},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090774498801903e-05,"output_power":3.901280750459222e-05,"reference_active":true,"delta_db":-0.20598335107611188},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.643789476493196e-06,"output_power":9.400482583779175e-06,"reference_active":true,"delta_db":0.3644596824525007}],"active_band_count":3,"active_band_mae_db":0.25704300497115773}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":1.2176246710536245,"rms_envelope_nrmse":0.11344972119790402,"envelope_correlation":0.9902720841710395,"rms_gain_error_db":-0.14662471091880927,"reference_mean_power":0.0005252755682472565,"output_mean_power":0.0005078374567857618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.5461080823927677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019511319942898067,"output_power":0.00018621701168468045,"reference_active":true,"delta_db":-0.20267297424288266},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8168801206395614e-05,"output_power":1.7400735745422218e-05,"reference_active":true,"delta_db":-0.18758661420925848},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.35857242269651e-07,"output_power":8.191152776767161e-07,"reference_active":true,"delta_db":0.46551457685738906}],"active_band_count":3,"active_band_mae_db":0.28525805510317676}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":1.7995920676373565,"rms_envelope_nrmse":0.24665505400999185,"envelope_correlation":0.9539168637120577,"rms_gain_error_db":-0.3040024190008586,"reference_mean_power":0.0002851540241046427,"output_mean_power":0.000265876077433485},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.768616554938334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966944293503785,"output_power":9.752557421339936e-05,"reference_active":true,"delta_db":-0.5096712143976243},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.898846044442001e-05,"output_power":4.0231018620850584e-05,"reference_active":true,"delta_db":-0.8553276240171226},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.52294853557438e-06,"output_power":8.99917308628996e-06,"reference_active":true,"delta_db":0.2361273889804228}],"active_band_count":3,"active_band_mae_db":0.5337087424650565}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":1.1572244339724356,"rms_envelope_nrmse":0.13493710537621637,"envelope_correlation":0.9857061177022785,"rms_gain_error_db":-0.3040024785172955,"reference_mean_power":0.0002865881294452576,"output_mean_power":0.000267213225873073},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.511088390714625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010245851234810725,"output_power":9.014043928860098e-05,"reference_active":true,"delta_db":-0.5562837560186561},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.514042968169313e-05,"output_power":2.0790601843675386e-05,"reference_active":true,"delta_db":-0.8250563464014409},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.446669785409341e-07,"output_power":7.402256871860295e-07,"reference_active":true,"delta_db":-0.025979441130929593}],"active_band_count":3,"active_band_mae_db":0.46910651451700885}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001420862345348912,"log_rms_mae_db":1.8722588087322944,"rms_envelope_nrmse":0.32477830080578973,"envelope_correlation":0.9365318359253666,"rms_gain_error_db":-1.0221049348434328,"reference_mean_power":0.000234926135231895,"output_mean_power":0.00018566106642617554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.4459776587443103e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015814203638019883,"output_power":0.00012627211241372972,"reference_active":true,"delta_db":-0.9773988100002879},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.639730090455678e-05,"output_power":7.3458285631865e-05,"reference_active":true,"delta_db":-1.1802408538732658},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2676469905844767e-05,"output_power":8.990596323423951e-06,"reference_active":true,"delta_db":-1.4920983140111475}],"active_band_count":3,"active_band_mae_db":1.2165793259615671}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010454414213188797,"log_rms_mae_db":1.498429467779736,"rms_envelope_nrmse":0.21387954697002806,"envelope_correlation":0.9757204907616038,"rms_gain_error_db":-1.0221049360537708,"reference_mean_power":0.00023610764979113878,"output_mean_power":0.0001865948120080966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.234204737076497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017362277943143455,"output_power":0.00013672489908647765,"reference_active":true,"delta_db":-1.0375909281785198},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.868815527339496e-05,"output_power":3.693994475579807e-05,"reference_active":true,"delta_db":-1.1992707843140218},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.341220903531157e-06,"output_power":8.888902677345368e-07,"reference_active":true,"delta_db":-1.7865216225412155}],"active_band_count":3,"active_band_mae_db":1.3411277783445856}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.4312591552734375,"predicted_mos":3.4312591552734375,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"d6e21b073d30d37933eafb5d081fa4c3c3b5ba144e6adf88c91a04f0c9eec1eb","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2709532420012692},"started_at":"2026-10-13T05:10:00.583433+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032477679997100495,"native_input_samples":144000,"native_input_sha256":"d69401da71d8da5b373ebdf9dde5d574f0b758885253ac0a04da1fed429896d1","output_sha256":"ce12cb52f8a6c0422dffa7817a170d98e67df5ae131238a22255a158661adf3b","output_samples":144240,"length_delta_samples":240,"analysis_s":0.6967498780013557,"synthesis_s":0.11082132499905128,"total_s":0.807571203000407,"rtf":0.13459520050006782,"output_peak":0.4001028537750244,"output_rms":0.039410399960817806,"finished_at":"2026-10-13T05:10:01.403541+00:00","analysis_stages":{"external_f0_s":0.15360888200120826},"evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"ce12cb52f8a6c0422dffa7817a170d98e67df5ae131238a22255a158661adf3b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.121663105237285,"output_peak":0.40634675589116237},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":3.7762423055331666,"spectral_convergence":0.2867201394051328,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.6115950300013404,"spectral_convergence":0.27962093844813923,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":3.9189697992845294,"spectral_convergence":0.3156550608543798,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.0003961984368918015,"delta_db":-0.037191717639771894,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.0002999476666132939,"delta_db":-0.07342127977084698,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.00027808150970614145,"delta_db":-0.10704317337606593,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00015368410137142185,"delta_db":-0.5333372510430668,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.00023430093935600021,"delta_db":-0.3537376467322585,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00018146603166644656,"delta_db":-1.052318541003475,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.35950826826091414,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2862639108864363,"pitch":null,"output_rms_dbfs":-28.114430923362278,"output_peak":0.40634675589116237,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":3.7689357116063458},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9614331870281286,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04060613455341415,"rms_dbfs":-27.828167012475838,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.2864855868614,"mean":-1.2833913167317708e-05,"zero_energy":false},"output":{"rms":0.03928967657801608,"rms_dbfs":-28.114430923362278,"peak":0.40634675589116237,"peak_dbfs":-7.822064062590718,"crest_factor_db":20.29236686077156,"mean":3.677143443652976e-06,"zero_energy":false},"rms_difference_db":-0.28626391088644},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":220,"quiet_samples":35200,"reference_mean_power":1.6131530918011612e-07,"output_mean_power":1.5295743640271322e-07,"output_rms_dbfs":-68.15429403869527,"output_zero_energy":false,"output_peak":0.006609943374137432,"output_energy_fraction":3.633164976018116e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":181,"both_estimated_voiced_frames":162,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9759036144578314,"absolute_difference_cents_median":0.8343861152442285,"absolute_difference_cents_p95":5.465689561176172,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015167567220720444,"log_rms_mae_db":0.5724222584629134,"rms_envelope_nrmse":0.07898788380187535,"envelope_correlation":0.9949682637497035,"rms_gain_error_db":-0.28626392122071026,"reference_mean_power":0.0016502333338478435,"output_mean_power":0.00154496613126973},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103350035549282e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021122178949431673,"output_power":0.0002034336817039206,"reference_active":true,"delta_db":-0.16315858746006812},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018786794884032712,"output_power":0.00016723056540666206,"reference_active":true,"delta_db":-0.5053703561303542},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6179250514948536e-05,"output_power":2.139029445334812e-05,"reference_active":true,"delta_db":-0.8774044598783625}],"active_band_count":3,"active_band_mae_db":0.515311134489595}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014379702931337134,"log_rms_mae_db":0.3893384775885756,"rms_envelope_nrmse":0.058124417967735376,"envelope_correlation":0.9975465813767571,"rms_gain_error_db":-0.28626371005462353,"reference_mean_power":0.0016585326982667058,"output_mean_power":0.001552736160674129},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0955155560805602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019170493341265235,"output_power":0.0001841748690172335,"reference_active":true,"delta_db":-0.17402919670081937},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.87936190430084e-05,"output_power":7.015206469110371e-05,"reference_active":true,"delta_db":-0.504505908138115},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1061472825588798e-06,"output_power":1.8119260820750104e-06,"reference_active":true,"delta_db":-0.653482614956251}],"active_band_count":3,"active_band_mae_db":0.4440059065983952}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010769717002798511,"log_rms_mae_db":0.7656777867719228,"rms_envelope_nrmse":0.06714166591559771,"envelope_correlation":0.9967992574276832,"rms_gain_error_db":-0.0690909049716029,"reference_mean_power":0.0005226470255969395,"output_mean_power":0.000514398143662632},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.88003979644585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021484485335499559,"output_power":0.00020970965648356757,"reference_active":true,"delta_db":-0.10506525697734431},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090774498801903e-05,"output_power":3.9046331780191135e-05,"reference_active":true,"delta_db":-0.2022529976720231},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.643789476493196e-06,"output_power":6.082738844970637e-06,"reference_active":true,"delta_db":-1.5260500994092179}],"active_band_count":3,"active_band_mae_db":0.6111227846861951}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":0.43163469519952674,"rms_envelope_nrmse":0.0387164811051464,"envelope_correlation":0.9989015421018915,"rms_gain_error_db":-0.06909085364837977,"reference_mean_power":0.0005252755682472565,"output_mean_power":0.0005169852064158236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.5461080823927677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019511319942898067,"output_power":0.00019152343412845565,"reference_active":true,"delta_db":-0.08064730167790342},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8168801206395614e-05,"output_power":1.695548472058222e-05,"reference_active":true,"delta_db":-0.3001606328843854},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.35857242269651e-07,"output_power":5.28011336609741e-07,"reference_active":true,"delta_db":-1.4415032143228497}],"active_band_count":3,"active_band_mae_db":0.6074370496283795}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":0.8042435403920016,"rms_envelope_nrmse":0.12578909262634824,"envelope_correlation":0.9895066262413862,"rms_gain_error_db":-0.4378524811614684,"reference_mean_power":0.0002851540241046427,"output_mean_power":0.00025780673372868845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.768616554938334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966944293503785,"output_power":9.919581862560193e-05,"reference_active":true,"delta_db":-0.4359227163783892},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.898846044442001e-05,"output_power":4.412164148450489e-05,"reference_active":true,"delta_db":-0.4544212974751842},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.52294853557438e-06,"output_power":7.985336969964502e-06,"reference_active":true,"delta_db":-0.2829661855840649}],"active_band_count":3,"active_band_mae_db":0.3911033998125461}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":0.592312818354902,"rms_envelope_nrmse":0.08124040776940754,"envelope_correlation":0.9964484559156304,"rms_gain_error_db":-0.4378523068072759,"reference_mean_power":0.0002865881294452576,"output_mean_power":0.00025910331364211765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.511088390714625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010245851234810725,"output_power":9.392085582767572e-05,"reference_active":true,"delta_db":-0.3778600451471898},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.514042968169313e-05,"output_power":2.2784561723094523e-05,"reference_active":true,"delta_db":-0.42732017012403156},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.446669785409341e-07,"output_power":6.808929438881857e-07,"reference_active":true,"delta_db":-0.3888326235069224}],"active_band_count":3,"active_band_mae_db":0.39800427959271456}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001420862345348912,"log_rms_mae_db":0.7433709390030929,"rms_envelope_nrmse":0.12079075569622331,"envelope_correlation":0.9920812077546535,"rms_gain_error_db":-0.40247838116969886,"reference_mean_power":0.000234926135231895,"output_mean_power":0.0002141329480084316},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.4459776587443103e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015814203638019883,"output_power":0.00014675103247541025,"reference_active":true,"delta_db":-0.32466161432432616},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.639730090455678e-05,"output_power":8.597953982014176e-05,"reference_active":true,"delta_db":-0.4966975760269794},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2676469905844767e-05,"output_power":8.910256566141172e-06,"reference_active":true,"delta_db":-1.5310812024224487}],"active_band_count":3,"active_band_mae_db":0.7841467975912515}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010454414213188797,"log_rms_mae_db":0.5217337328508045,"rms_envelope_nrmse":0.07334332279454674,"envelope_correlation":0.9977588703272898,"rms_gain_error_db":-0.4024783157282702,"reference_mean_power":0.00023610764979113878,"output_mean_power":0.00021520989058560776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.234204737076497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017362277943143455,"output_power":0.00015906768358156737,"reference_active":true,"delta_db":-0.38024747602784803},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.868815527339496e-05,"output_power":4.2882281375336226e-05,"reference_active":true,"delta_db":-0.5514543789237635},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.341220903531157e-06,"output_power":9.214007884861022e-07,"reference_active":true,"delta_db":-1.6305173392131946}],"active_band_count":3,"active_band_mae_db":0.854073064721602}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_121-127105-0035","method":"magphase","status":"measured","mos":{"status":"measured","value":3.3714587688446045,"predicted_mos":3.3714587688446045,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288535,"input_sample_rate":48000,"input_seconds":6.011145833333333,"input_float32_sha256":"b69247a41b4b4f99447ab1bc67ad74c17ba72722f28302518865e61839cc99c8","internal_samples":96179,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2216288889994757},"started_at":"2026-10-13T05:12:16.578846+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010109718999956385,"native_input_samples":288000,"native_input_sha256":"7e5a192b5e0d6893a3a7665bad6b262fd8bb80a9e88e2185e185c7aedd8eee4d","output_sha256":"6fd27f98a2410e965431716d3cc2d3939bd2984398ce7fbee3f9a34887a6522f","output_samples":288535,"length_delta_samples":535,"analysis_s":3.7079247960009525,"synthesis_s":1.0765039069992781,"total_s":4.784428703000231,"rtf":0.7974047838333718,"output_peak":0.5927711685788192,"output_rms":0.03937287082410142,"finished_at":"2026-10-13T05:12:21.385432+00:00","evaluation_input":{"reference":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"output_sha256":"6fd27f98a2410e965431716d3cc2d3939bd2984398ce7fbee3f9a34887a6522f","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96179,"length_delta_samples":179,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.13240812188944,"output_peak":0.45401278088030533},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981388868671955,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.029340742223712,"spectral_convergence":0.6351250867045076,"spectral_convergence_status":"ok","amplitude_floor":1.0971770400792783e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.7193240976588795,"spectral_convergence":0.5099782476768301,"spectral_convergence_status":"ok","amplitude_floor":6.752711905169626e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.146801985187395,"spectral_convergence":0.4633253809012304,"spectral_convergence_status":"ok","amplitude_floor":5.858543156261471e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003996059347809583,"output_power":0.00034760864344959825,"delta_db":-0.6054136271872557,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506165058519967,"output_power":0.0002839824498942383,"delta_db":-0.3109611425899745,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002850207180759584,"output_power":0.00026832869343101353,"delta_db":-0.2620931394629804,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017376514501848305,"output_power":0.00016769215642290034,"delta_db":-0.15449917604578575,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002541837569284722,"output_power":0.000267337454938386,"delta_db":0.21912014796364895,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023122095798090296,"output_power":0.00020521895558829888,"delta_db":-0.5180972315319888,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3450307441302723,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.29615103177428004,"pitch":null,"output_rms_dbfs":-28.12431804425012,"output_peak":0.45401278088030533,"reference_rms_dbfs":-27.828167012475838,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.631822275023329},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.795659171942857,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04060613455341415,"rms_dbfs":-27.828167012475838,"peak":0.419677734375,"peak_dbfs":-7.541681425614438,"crest_factor_db":20.2864855868614,"mean":-1.2833913167317708e-05,"zero_energy":false},"output":{"rms":0.039244978707147174,"rms_dbfs":-28.12431804425012,"peak":0.45401278088030533,"peak_dbfs":-6.858638423577678,"crest_factor_db":21.265679620672444,"mean":-3.5014597189826616e-05,"zero_energy":false},"rms_difference_db":-0.2961510317742828},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.273209779523313e-06,"threshold_dbfs":-56.43360484255024,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":220,"quiet_samples":35200,"reference_mean_power":1.6131530918011612e-07,"output_mean_power":3.1131936705702146e-07,"output_rms_dbfs":-65.06793861185885,"output_zero_energy":false,"output_peak":0.009621830380327195,"output_energy_fraction":7.41155564659772e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":166,"output_estimated_voiced_frames":152,"both_estimated_voiced_frames":132,"reference_voiced_output_unvoiced_frames":34,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.7951807228915663,"absolute_difference_cents_median":6.245525954680588,"absolute_difference_cents_p95":21.396013224464156,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015167567220720444,"log_rms_mae_db":2.0106931037427835,"rms_envelope_nrmse":0.281578286043182,"envelope_correlation":0.9246330308869052,"rms_gain_error_db":-0.29615099306489906,"reference_mean_power":0.0016502333338478435,"output_mean_power":0.0015414528891393496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103350035549282e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021122178949431673,"output_power":0.00020631106231129969,"reference_active":true,"delta_db":-0.10216202409743753},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018786794884032712,"output_power":0.00018144892554983652,"reference_active":true,"delta_db":-0.15098292861033336},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6179250514948536e-05,"output_power":3.5492932383558786e-05,"reference_active":true,"delta_db":1.3218467273198067}],"active_band_count":3,"active_band_mae_db":0.5249972266758592}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014379702931337134,"log_rms_mae_db":1.3310402504729675,"rms_envelope_nrmse":0.17145385440363553,"envelope_correlation":0.9702331837614643,"rms_gain_error_db":-0.29615092560620104,"reference_mean_power":0.0016585326982667058,"output_mean_power":0.0015492051982884942},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0955155560805602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019170493341265235,"output_power":0.00018870204674226952,"reference_active":true,"delta_db":-0.06856678600288914},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.87936190430084e-05,"output_power":7.637912842668496e-05,"reference_active":true,"delta_db":-0.13516350101793628},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1061472825588798e-06,"output_power":2.7314609164852763e-06,"reference_active":true,"delta_db":1.1290625266946008}],"active_band_count":3,"active_band_mae_db":0.4442642712384754}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010769717002798511,"log_rms_mae_db":2.352441642967671,"rms_envelope_nrmse":0.29210259192223087,"envelope_correlation":0.9377239902435855,"rms_gain_error_db":-0.29692436135742195,"reference_mean_power":0.0005226470255969395,"output_mean_power":0.0004881081363262621},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.88003979644585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021484485335499559,"output_power":0.00019436807588120415,"reference_active":true,"delta_db":-0.43500019095804066},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.090774498801903e-05,"output_power":3.61085516034695e-05,"reference_active":true,"delta_db":-0.5419547173445939},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.643789476493196e-06,"output_power":9.195410270857265e-06,"reference_active":true,"delta_db":0.26866929991028987}],"active_band_count":3,"active_band_mae_db":0.41520806940430816}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.974444405364767e-05,"log_rms_mae_db":1.659525320236814,"rms_envelope_nrmse":0.19726803791944653,"envelope_correlation":0.970376346887563,"rms_gain_error_db":-0.2969243005920841,"reference_mean_power":0.0005252755682472565,"output_mean_power":0.0004905629798034522},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.5461080823927677e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019511319942898067,"output_power":0.00017721955790687673,"reference_active":true,"delta_db":-0.41775001628615294},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8168801206395614e-05,"output_power":1.6223120418930097e-05,"reference_active":true,"delta_db":-0.49191881296209056},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.35857242269651e-07,"output_power":7.543705383487222e-07,"reference_active":true,"delta_db":0.10791150318044508}],"active_band_count":3,"active_band_mae_db":0.33919344414289626}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.224188331064747e-05,"log_rms_mae_db":2.545556219000559,"rms_envelope_nrmse":0.37221680061682455,"envelope_correlation":0.8976673082862228,"rms_gain_error_db":0.02646168504132019,"reference_mean_power":0.0002851540241046427,"output_mean_power":0.0002868967795382438},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.768616554938334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966944293503785,"output_power":0.00011067911888553401,"reference_active":true,"delta_db":0.03980055591635143},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.898846044442001e-05,"output_power":4.7435972038198104e-05,"reference_active":true,"delta_db":-0.13985986750693002},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.52294853557438e-06,"output_power":8.978041960710415e-06,"reference_active":true,"delta_db":0.22591764799644698}],"active_band_count":3,"active_band_mae_db":0.1351926904732428}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.641451530936751e-05,"log_rms_mae_db":1.801907954247368,"rms_envelope_nrmse":0.23830586959490313,"envelope_correlation":0.9544503034578733,"rms_gain_error_db":0.026461548469446386,"reference_mean_power":0.0002865881294452576,"output_mean_power":0.000288339640530425},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.511088390714625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010245851234810725,"output_power":0.00010220750136522866,"reference_active":true,"delta_db":-0.010652744678118933},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.514042968169313e-05,"output_power":2.4129344521388596e-05,"reference_active":true,"delta_db":-0.17827171684519635},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.446669785409341e-07,"output_power":8.111907676044045e-07,"reference_active":true,"delta_db":0.37160903239888177}],"active_band_count":3,"active_band_mae_db":0.1868444979740657}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001420862345348912,"log_rms_mae_db":2.628560490151746,"rms_envelope_nrmse":0.4959774814747012,"envelope_correlation":0.8545768048294159,"rms_gain_error_db":0.1856739486126065,"reference_mean_power":0.000234926135231895,"output_mean_power":0.0002451877278350011},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.4459776587443103e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015814203638019883,"output_power":0.00016399854113116318,"reference_active":true,"delta_db":0.1579265787896753},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.639730090455678e-05,"output_power":9.599380937204669e-05,"reference_active":true,"delta_db":-0.01821647608682348},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2676469905844767e-05,"output_power":1.9859084219732246e-05,"reference_active":true,"delta_db":1.9496088785924086}],"active_band_count":3,"active_band_mae_db":0.7085839778229691}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010454414213188797,"log_rms_mae_db":2.1463764557811364,"rms_envelope_nrmse":0.31552104873434894,"envelope_correlation":0.9373825234965257,"rms_gain_error_db":0.18567387300971713,"reference_mean_power":0.00023610764979113878,"output_mean_power":0.0002464208467560852},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.234204737076497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017362277943143455,"output_power":0.00018175631502407802,"reference_active":true,"delta_db":0.19882804796409603},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.868815527339496e-05,"output_power":4.869840771092701e-05,"reference_active":true,"delta_db":0.0009144130209002036},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.341220903531157e-06,"output_power":2.154357284343458e-06,"reference_active":true,"delta_db":2.0581741600285866}],"active_band_count":3,"active_band_mae_db":0.7526388736711942}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2227303981781006,"predicted_mos":2.2227303981781006,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"faefd8b4ed00dbd7d2b377faefb8de491a4c2e369a233186e0489bcdaf00ee3d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.504432398000063},"started_at":"2026-10-13T04:38:07.605475+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.9297000108053908e-05,"native_input_samples":96000,"native_input_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","output_sha256":"a1b0534287b19348621a70b7c850bc1fdb3de32aedc738504bd18a8598c54274","output_samples":96000,"length_delta_samples":0,"analysis_s":1.107031360000292,"synthesis_s":0.5155048979995627,"total_s":1.6225362579998546,"rtf":0.27042270966664245,"output_peak":0.359666227007926,"output_rms":0.03615587378659344,"finished_at":"2026-10-13T04:38:09.234649+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"a1b0534287b19348621a70b7c850bc1fdb3de32aedc738504bd18a8598c54274","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.3621274195414035,"spectral_convergence":0.2989721563572629,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1965732980125825,"spectral_convergence":0.2738272602296485,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.521151565545912,"spectral_convergence":0.2792046778773485,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":0.00010188746130600289,"delta_db":-0.01862375146611951,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.0003652275736687115,"delta_db":0.0718541589573288,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.00038669748523204007,"delta_db":-0.043576514527359785,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.00021870814055052892,"delta_db":-0.12442390823878957,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":0.00010686880372228749,"delta_db":-0.21811523537489177,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.00012785774479250384,"delta_db":-0.09493047680398041,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.09525400756141163,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.043076950790658104,"pitch":null,"output_rms_dbfs":-28.836422766346864,"output_peak":0.359666227007926,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.359950761033299},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.3621274195414035,"spectral_convergence":0.2989721563572629,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1965732980125825,"spectral_convergence":0.2738272602296485,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.521151565545912,"spectral_convergence":0.2792046778773485,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":0.00010188746130600289,"delta_db":-0.01862375146611951,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.0003652275736687115,"delta_db":0.0718541589573288,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.00038669748523204007,"delta_db":-0.043576514527359785,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.00021870814055052892,"delta_db":-0.12442390823878957,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":0.00010686880372228749,"delta_db":-0.21811523537489177,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.00012785774479250384,"delta_db":-0.09493047680398041,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.09525400756141163,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.043076950790658104,"pitch":null,"output_rms_dbfs":-28.836422766346864,"output_peak":0.359666227007926,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.359950761033299},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8560155388468386,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03633563122790996,"rms_dbfs":-28.793345815556208,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.25840632339133,"mean":-0.00027747472127278644,"zero_energy":false},"output":{"rms":0.03615587378659344,"rms_dbfs":-28.836422766346864,"peak":0.359666227007926,"peak_dbfs":-8.882006818426326,"crest_factor_db":19.954415947920538,"mean":3.2393146181003906e-05,"zero_energy":false},"rms_difference_db":-0.0430769507906561},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":268,"both_estimated_voiced_frames":265,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.916955017301038,"absolute_difference_cents_median":1.1393598667425977,"absolute_difference_cents_p95":5.190633949603857,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001036839322212551,"log_rms_mae_db":0.8562333918908532,"rms_envelope_nrmse":0.06339826943711127,"envelope_correlation":0.9949629785148003,"rms_gain_error_db":-0.04309021717748305,"reference_mean_power":0.0013213586135692834,"output_mean_power":0.0013083130651121142},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.530367535346178e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002613603148630271,"output_power":0.00025304540006894247,"reference_active":true,"delta_db":-0.1404119767772252},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.246118337274924e-05,"output_power":7.586194236780976e-05,"reference_active":true,"delta_db":-0.3622560474707638},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.028948119628933e-05,"output_power":1.0549207702182178e-05,"reference_active":true,"delta_db":0.10826365342907951}],"active_band_count":3,"active_band_mae_db":0.20364389255902282}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":0.40987758768208765,"rms_envelope_nrmse":0.03343808000753281,"envelope_correlation":0.9984717032831616,"rms_gain_error_db":-0.0430563235510887,"reference_mean_power":0.0013278763848253465,"output_mean_power":0.0013147767483898036},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.825313121587468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002333069672738863,"output_power":0.00022646761751539233,"reference_active":true,"delta_db":-0.1292159712508708},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6274669036315644e-05,"output_power":4.268278337345941e-05,"reference_active":true,"delta_db":-0.3509058897816136},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.136183723659666e-07,"output_power":6.236689047646884e-07,"reference_active":true,"delta_db":0.07055736932561966}],"active_band_count":3,"active_band_mae_db":0.18355974345270135}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":1.326230582418545,"rms_envelope_nrmse":0.06162588168108709,"envelope_correlation":0.9967337747868036,"rms_gain_error_db":-0.030570967512462972,"reference_mean_power":0.0005515675704872272,"output_mean_power":0.0005476985953013012},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3413004486315456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020691279435977171,"output_power":0.00020231469621083712,"reference_active":true,"delta_db":-0.09759914679507085},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313846735404118e-05,"output_power":2.87999930650638e-05,"reference_active":true,"delta_db":-0.6094003535222197},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.722807599466502e-06,"output_power":2.4209173281039624e-06,"reference_active":true,"delta_db":-0.5103699509332534}],"active_band_count":3,"active_band_mae_db":0.40578981708351464}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":0.7457495663792276,"rms_envelope_nrmse":0.03732239668442376,"envelope_correlation":0.9987048135371401,"rms_gain_error_db":-0.030573682963801717,"reference_mean_power":0.0005543343450028261,"output_mean_power":0.000550445618085264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0568367997569596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019480947374469874,"output_power":0.00019030791148723383,"reference_active":true,"delta_db":-0.10153229985594606},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.636770712067895e-05,"output_power":1.4368192471391359e-05,"reference_active":true,"delta_db":-0.5658570826718687},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1494053604425803e-07,"output_power":1.8782785745565319e-07,"reference_active":true,"delta_db":-0.5855832308310227}],"active_band_count":3,"active_band_mae_db":0.4176575377862792}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":1.2444146882160596,"rms_envelope_nrmse":0.0861027918075621,"envelope_correlation":0.9942508135338237,"rms_gain_error_db":-0.15646695971067748,"reference_mean_power":0.00027894413436310234,"output_mean_power":0.0002690732581394661},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0416852970454036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011680629747545272,"output_power":0.00011151229631922057,"reference_active":true,"delta_db":-0.20143498763251735},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1432581274087066e-05,"output_power":1.07409213788815e-05,"reference_active":true,"delta_db":-0.27102759800281423},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6649698962790796e-06,"output_power":1.739373698528985e-06,"reference_active":true,"delta_db":0.1898651312998305}],"active_band_count":3,"active_band_mae_db":0.22077590564505403}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":0.7188405642365678,"rms_envelope_nrmse":0.05046892734435441,"envelope_correlation":0.9980738878937068,"rms_gain_error_db":-0.156432175260771,"reference_mean_power":0.0002803113215319917,"output_mean_power":0.00027039423093456815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.911634143181416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010971300365261624,"output_power":0.00010504992743998812,"reference_active":true,"delta_db":-0.18862348291211822},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3756201199309736e-06,"output_power":5.0340641381437745e-06,"reference_active":true,"delta_db":-0.2850982623256492},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905227736918007e-08,"output_power":9.883551040647696e-08,"reference_active":true,"delta_db":-0.00951455716115815}],"active_band_count":3,"active_band_mae_db":0.16107876746630853}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":1.283465214707313,"rms_envelope_nrmse":0.1475651696958693,"envelope_correlation":0.9822511992658371,"rms_gain_error_db":-0.20440254578200318,"reference_mean_power":9.153664281853565e-05,"output_mean_power":8.732824365096384e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.93958925235022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.468609555304607e-05,"output_power":3.313638716147996e-05,"reference_active":true,"delta_db":-0.19850260329780822},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1944580647265697e-05,"output_power":2.0452063932561664e-05,"reference_active":true,"delta_db":-0.305901444064192},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8366553780699585e-06,"output_power":4.9600277851130055e-06,"reference_active":true,"delta_db":0.1093896524038305}],"active_band_count":3,"active_band_mae_db":0.20459789992194355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":0.735363015505234,"rms_envelope_nrmse":0.09163221445267096,"envelope_correlation":0.9927537711182342,"rms_gain_error_db":-0.20455344052946361,"reference_mean_power":9.19626023411585e-05,"output_mean_power":8.77315714063386e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.164079440499498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.035826325529751e-05,"output_power":3.86032607608175e-05,"reference_active":true,"delta_db":-0.19308478452110897},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2217507839367813e-05,"output_power":1.1526966146435903e-05,"reference_active":true,"delta_db":-0.2526760868383452},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.523678505678969e-07,"output_power":4.4661141379639016e-07,"reference_active":true,"delta_db":-0.055619144815781126}],"active_band_count":3,"active_band_mae_db":0.16712667205841178}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.164355516433716,"predicted_mos":3.164355516433716,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6aa8a65e0e86b141aca11831845c8e75cd023d32b94c2fa87ad61f987a97feb1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2031846909994783},"started_at":"2026-10-13T04:40:58.778361+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3966000298969448e-05,"native_input_samples":96000,"native_input_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","output_sha256":"509548080a615e66e4e985b69d0bdcc139a5293756d3b84ebcad55d7036b628b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3350190010005463,"synthesis_s":0.20063434900021093,"total_s":1.5356533500007572,"rtf":0.2559422250001262,"output_peak":0.4812580304974493,"output_rms":0.039619213565734036,"finished_at":"2026-10-13T04:41:00.319321+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"509548080a615e66e4e985b69d0bdcc139a5293756d3b84ebcad55d7036b628b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.103461928137469,"spectral_convergence":0.5642011379715971,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.353493336015238,"spectral_convergence":0.30700064822576,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.862500156795943,"spectral_convergence":0.333160019086707,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":0.00011420927972446645,"delta_db":0.47718276249173697,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.0004050770596205718,"delta_db":0.521595066225217,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.000505630010638414,"delta_db":1.1210384346824172,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.00026817510994417903,"delta_db":0.7611112827616695,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":0.00013460025887175053,"delta_db":0.7838342364127766,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.0001419903647678621,"delta_db":0.3603878729405876,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6708582759190674,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7514628279702952,"pitch":null,"output_rms_dbfs":-28.04188298758591,"output_peak":0.4812580304974493,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7731518069828835},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.103461928137469,"spectral_convergence":0.5642011379715971,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.353493336015238,"spectral_convergence":0.30700064822576,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.862500156795943,"spectral_convergence":0.333160019086707,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":0.00011420927972446645,"delta_db":0.47718276249173697,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.0004050770596205718,"delta_db":0.521595066225217,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.000505630010638414,"delta_db":1.1210384346824172,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.00026817510994417903,"delta_db":0.7611112827616695,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":0.00013460025887175053,"delta_db":0.7838342364127766,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.0001419903647678621,"delta_db":0.3603878729405876,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6708582759190674,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7514628279702952,"pitch":null,"output_rms_dbfs":-28.04188298758591,"output_peak":0.4812580304974493,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7731518069828835},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8858420285414901,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03633563122790996,"rms_dbfs":-28.793345815556208,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.25840632339133,"mean":-0.00027747472127278644,"zero_energy":false},"output":{"rms":0.039619213565734036,"rms_dbfs":-28.04188298758591,"peak":0.4812580304974493,"peak_dbfs":-6.352440211627045,"crest_factor_db":21.689442775958867,"mean":1.2535194699842264e-08,"zero_energy":false},"rms_difference_db":0.7514628279702968},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":330,"both_estimated_voiced_frames":278,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":52,"paired_voiced_fraction":0.9619377162629758,"absolute_difference_cents_median":4.580180833419717,"absolute_difference_cents_p95":17.20006664432122,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001036839322212551,"log_rms_mae_db":1.6911538344293127,"rms_envelope_nrmse":0.13743129240550436,"envelope_correlation":0.988464597192206,"rms_gain_error_db":0.751499210722954,"reference_mean_power":0.0013213586135692834,"output_mean_power":0.0015709798735957102},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.530367535346178e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002613603148630271,"output_power":0.0003116485028221359,"reference_active":true,"delta_db":0.7642540019061105},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.246118337274924e-05,"output_power":0.00010328566505540343,"reference_active":true,"delta_db":0.9779048735290992},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.028948119628933e-05,"output_power":1.3323895609693671e-05,"reference_active":true,"delta_db":1.1223774356782765}],"active_band_count":3,"active_band_mae_db":0.9548454370378288}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":1.3472680278738354,"rms_envelope_nrmse":0.11569521501041637,"envelope_correlation":0.9939325768272803,"rms_gain_error_db":0.7516891220848623,"reference_mean_power":0.0013278763848253465,"output_mean_power":0.0015787979710966554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.825313121587468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002333069672738863,"output_power":0.00027943824106103064,"reference_active":true,"delta_db":0.7835813057282579},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6274669036315644e-05,"output_power":5.835554917048869e-05,"reference_active":true,"delta_db":1.007388391652264},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.136183723659666e-07,"output_power":8.856215958754052e-07,"reference_active":true,"delta_db":1.593498439075821}],"active_band_count":3,"active_band_mae_db":1.1281560454854476}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":1.6822656941219287,"rms_envelope_nrmse":0.17645387846858904,"envelope_correlation":0.9897235518462965,"rms_gain_error_db":1.0848151188286255,"reference_mean_power":0.0005515675704872272,"output_mean_power":0.0007080765891423328},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3413004486315456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020691279435977171,"output_power":0.0002540909586477427,"reference_active":true,"delta_db":0.8920186586679706},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313846735404118e-05,"output_power":4.4515564345947296e-05,"reference_active":true,"delta_db":1.2817946497929367},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.722807599466502e-06,"output_power":2.5380153924762528e-06,"reference_active":true,"delta_db":-0.30522702405531854}],"active_band_count":3,"active_band_mae_db":0.8263467775054086}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":1.3615228775621113,"rms_envelope_nrmse":0.16434480274482144,"envelope_correlation":0.9922728591865138,"rms_gain_error_db":1.084830359473325,"reference_mean_power":0.0005543343450028261,"output_mean_power":0.000711630941870891},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0568367997569596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019480947374469874,"output_power":0.00023982881345719517,"reference_active":true,"delta_db":0.9029128564994575},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.636770712067895e-05,"output_power":2.189118115328345e-05,"reference_active":true,"delta_db":1.2628134955891892},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1494053604425803e-07,"output_power":2.2400879291318558e-07,"reference_active":true,"delta_db":0.1794673819095348}],"active_band_count":3,"active_band_mae_db":0.7817312446660605}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":1.2618541059191208,"rms_envelope_nrmse":0.13876009080077023,"envelope_correlation":0.9927995860164607,"rms_gain_error_db":0.7996271464491966,"reference_mean_power":0.00027894413436310234,"output_mean_power":0.00033533582126743614},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0416852970454036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011680629747545272,"output_power":0.00013673956534565596,"reference_active":true,"delta_db":0.6842793718479199},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1432581274087066e-05,"output_power":1.4919882402239748e-05,"reference_active":true,"delta_db":1.1562110260484275},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6649698962790796e-06,"output_power":2.353919366490718e-06,"reference_active":true,"delta_db":1.5038519641323453}],"active_band_count":3,"active_band_mae_db":1.1147807873428974}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":0.906595414384752,"rms_envelope_nrmse":0.11706893614555283,"envelope_correlation":0.9966023181816204,"rms_gain_error_db":0.7998321792400467,"reference_mean_power":0.0002803113215319917,"output_mean_power":0.00033699531002089257},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.911634143181416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010971300365261624,"output_power":0.00012952171841781572,"reference_active":true,"delta_db":0.720844926890315},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3756201199309736e-06,"output_power":6.913560205943174e-06,"reference_active":true,"delta_db":1.0927317844884967},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905227736918007e-08,"output_power":1.293819661502227e-07,"reference_active":true,"delta_db":1.1600928149194771}],"active_band_count":3,"active_band_mae_db":0.9912231754327628}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":1.3217612250949218,"rms_envelope_nrmse":0.1842417774272149,"envelope_correlation":0.9765042807021929,"rms_gain_error_db":0.49799921082600573,"reference_mean_power":9.153664281853565e-05,"output_mean_power":0.00010265849695184184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.93958925235022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.468609555304607e-05,"output_power":4.0018853818271824e-05,"reference_active":true,"delta_db":0.621092296790049},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1944580647265697e-05,"output_power":2.572748062601452e-05,"reference_active":true,"delta_db":0.6906997377399983},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8366553780699585e-06,"output_power":6.185355897533619e-06,"reference_active":true,"delta_db":1.0681954934773403}],"active_band_count":3,"active_band_mae_db":0.7933291760024627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":0.8326836735576746,"rms_envelope_nrmse":0.12110558992179243,"envelope_correlation":0.9904451341802601,"rms_gain_error_db":0.4981573832441669,"reference_mean_power":9.19626023411585e-05,"output_mean_power":0.00010313996760937819},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.164079440499498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.035826325529751e-05,"output_power":4.6674475363752174e-05,"reference_active":true,"delta_db":0.6314697614646075},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2217507839367813e-05,"output_power":1.4691357422946182e-05,"reference_active":true,"delta_db":0.8007929836006645},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.523678505678969e-07,"output_power":5.369935966972112e-07,"reference_active":true,"delta_db":0.744773746701196}],"active_band_count":3,"active_band_mae_db":0.7256788305888228}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"facodec","status":"measured","mos":{"status":"measured","value":4.754413604736328,"predicted_mos":4.754413604736328,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6fbb42a04cd54d9ce8cb15dbdb6e635667d02f3cb88c90238bb0c9e8f778a750","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.439043071000924},"started_at":"2026-10-13T04:45:45.375525+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.141499928460689e-05,"native_input_samples":96000,"native_input_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","output_sha256":"2619b6a050a6c2bbb499bd538207ea83d098b75deba6441b48fca2780aa816f3","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6564463700005945,"synthesis_s":5.607114046999413,"total_s":7.263560417000008,"rtf":1.2105934028333347,"output_peak":0.3171205520629883,"output_rms":0.03455733242570103,"finished_at":"2026-10-13T04:45:52.645531+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"2619b6a050a6c2bbb499bd538207ea83d098b75deba6441b48fca2780aa816f3","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.363910188786088,"spectral_convergence":0.3214466877967488,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.037017321086992,"spectral_convergence":0.3295576758437163,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.3064633935170775,"spectral_convergence":0.3449329064653977,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":9.465778639937062e-05,"delta_db":-0.3382677221495274,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.00034700870848023693,"delta_db":-0.15037768314005354,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.00035602625759123264,"delta_db":-0.40246970063839843,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.0001953461068523895,"delta_db":-0.6150257868570845,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":9.88379335765767e-05,"delta_db":-0.5573881456230773,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.00010233243148060169,"delta_db":-1.0620679423128696,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5209328301201684,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4358499698294426,"pitch":null,"output_rms_dbfs":-29.22919578538565,"output_peak":0.3171205520629883,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.902463634463385},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.363910188786088,"spectral_convergence":0.3214466877967488,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.037017321086992,"spectral_convergence":0.3295576758437163,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.3064633935170775,"spectral_convergence":0.3449329064653977,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":9.465778639937062e-05,"delta_db":-0.3382677221495274,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.00034700870848023693,"delta_db":-0.15037768314005354,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.00035602625759123264,"delta_db":-0.40246970063839843,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.0001953461068523895,"delta_db":-0.6150257868570845,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":9.88379335765767e-05,"delta_db":-0.5573881456230773,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.00010233243148060169,"delta_db":-1.0620679423128696,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5209328301201684,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4358499698294426,"pitch":null,"output_rms_dbfs":-29.22919578538565,"output_peak":0.3171205520629883,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.902463634463385},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8836519364489094,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03633563122790996,"rms_dbfs":-28.793345815556208,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.25840632339133,"mean":-0.00027747472127278644,"zero_energy":false},"output":{"rms":0.03455733242570103,"rms_dbfs":-29.22919578538565,"peak":0.3171205520629883,"peak_dbfs":-9.975512222951572,"crest_factor_db":19.253683562434077,"mean":3.692326619039932e-05,"zero_energy":false},"rms_difference_db":-0.4358499698294409},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":296,"both_estimated_voiced_frames":278,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9619377162629758,"absolute_difference_cents_median":3.2440312430571554,"absolute_difference_cents_p95":12.788875367189307,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001036839322212551,"log_rms_mae_db":1.1531038201327881,"rms_envelope_nrmse":0.11445987961549679,"envelope_correlation":0.9858370141700773,"rms_gain_error_db":-0.4358531029582349,"reference_mean_power":0.0013213586135692834,"output_mean_power":0.001195185704092471},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.530367535346178e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002613603148630271,"output_power":0.00023978467631333367,"reference_active":true,"delta_db":-0.3741821905248537},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.246118337274924e-05,"output_power":7.636645972006335e-05,"reference_active":true,"delta_db":-0.33346905354054857},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.028948119628933e-05,"output_power":1.288646275217892e-05,"reference_active":true,"delta_db":0.9774024493553726}],"active_band_count":3,"active_band_mae_db":0.5616845644735916}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":0.9597828409914301,"rms_envelope_nrmse":0.09937100003734205,"envelope_correlation":0.9890624625346383,"rms_gain_error_db":-0.43577905139840734,"reference_mean_power":0.0013278763848253465,"output_mean_power":0.00120110159106053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.825313121587468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002333069672738863,"output_power":0.00021647302185351895,"reference_active":true,"delta_db":-0.3252392865704221},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6274669036315644e-05,"output_power":4.2772958826256554e-05,"reference_active":true,"delta_db":-0.3417402760777544},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.136183723659666e-07,"output_power":8.771793080133585e-07,"reference_active":true,"delta_db":1.5519002409319445}],"active_band_count":3,"active_band_mae_db":0.7396266011933736}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":1.6448204626416314,"rms_envelope_nrmse":0.11813156285788398,"envelope_correlation":0.9879756313956505,"rms_gain_error_db":-0.1752561781309564,"reference_mean_power":0.0005515675704872272,"output_mean_power":0.0005297526111444261},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3413004486315456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020691279435977171,"output_power":0.00019628837078609576,"reference_active":true,"delta_db":-0.22892775556810208},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313846735404118e-05,"output_power":3.53577173110441e-05,"reference_active":true,"delta_db":0.28151800724191006},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.722807599466502e-06,"output_power":3.5203201618458015e-06,"reference_active":true,"delta_db":1.1156520889897212}],"active_band_count":3,"active_band_mae_db":0.5420326172665778}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":1.2719661410727372,"rms_envelope_nrmse":0.10717995186715015,"envelope_correlation":0.9894166762287476,"rms_gain_error_db":-0.1752534071567618,"reference_mean_power":0.0005543343450028261,"output_mean_power":0.0005324102971099624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0568367997569596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019480947374469874,"output_power":0.00018499303666372804,"reference_active":true,"delta_db":-0.22454691752814535},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.636770712067895e-05,"output_power":1.7775450756614873e-05,"reference_active":true,"delta_db":0.3583277725575479},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1494053604425803e-07,"output_power":3.2938480459203544e-07,"reference_active":true,"delta_db":1.8538523248446026}],"active_band_count":3,"active_band_mae_db":0.8122423383100986}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":1.7428253240023235,"rms_envelope_nrmse":0.13425168252584133,"envelope_correlation":0.9889819715727121,"rms_gain_error_db":-0.6451586071445286,"reference_mean_power":0.00027894413436310234,"output_mean_power":0.00024043704000950922},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0416852970454036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011680629747545272,"output_power":0.00010242045927422869,"reference_active":true,"delta_db":-0.5707953892086038},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1432581274087066e-05,"output_power":1.1770985094137777e-05,"reference_active":true,"delta_db":0.12668512288123093},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6649698962790796e-06,"output_power":2.1707026694346787e-06,"reference_active":true,"delta_db":1.151939547369214}],"active_band_count":3,"active_band_mae_db":0.6164733531530162}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":1.3506764420795034,"rms_envelope_nrmse":0.11033224392974536,"envelope_correlation":0.9935493738403519,"rms_gain_error_db":-0.6450839361018768,"reference_mean_power":0.0002803113215319917,"output_mean_power":0.0002416196468800784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.911634143181416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010971300365261624,"output_power":9.676578858088896e-05,"reference_active":true,"delta_db":-0.5453626486410681},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3756201199309736e-06,"output_power":5.548303592713926e-06,"reference_active":true,"delta_db":0.13731646075501214},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905227736918007e-08,"output_power":1.3112827006225479e-07,"reference_active":true,"delta_db":1.2183186661500285}],"active_band_count":3,"active_band_mae_db":0.6336659251820362}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":1.9407061191661943,"rms_envelope_nrmse":0.254322639620401,"envelope_correlation":0.950117326594779,"rms_gain_error_db":-0.7060199036202398,"reference_mean_power":9.153664281853565e-05,"output_mean_power":7.780239926062986e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.93958925235022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.468609555304607e-05,"output_power":3.337668614406462e-05,"reference_active":true,"delta_db":-0.16712201268253898},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1944580647265697e-05,"output_power":2.5535089403356394e-05,"reference_active":true,"delta_db":0.6581009668932501},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8366553780699585e-06,"output_power":7.87923755730803e-06,"reference_active":true,"delta_db":2.1193905047133126}],"active_band_count":3,"active_band_mae_db":0.9815378280963672}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":1.6213836233745949,"rms_envelope_nrmse":0.21069150641701223,"envelope_correlation":0.9638751693702461,"rms_gain_error_db":-0.7059062769367414,"reference_mean_power":9.19626023411585e-05,"output_mean_power":7.816649250093576e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.164079440499498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.035826325529751e-05,"output_power":4.220287873802317e-05,"reference_active":true,"delta_db":0.19409607127288617},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2217507839367813e-05,"output_power":1.5089381529021994e-05,"reference_active":true,"delta_db":0.9168881327507882},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.523678505678969e-07,"output_power":6.998806679024624e-07,"reference_active":true,"delta_db":1.895322652224786}],"active_band_count":3,"active_band_mae_db":1.0021022854161534}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.4077937602996826,"predicted_mos":3.4077937602996826,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"53dc4009b8cb7b64672fba74a7365e041b33fb8bd857bf2f6904cc1e03957f6b","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4338631829996302},"started_at":"2026-10-13T04:58:57.051051+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006674812000710517,"native_input_samples":264600,"native_input_sha256":"1f1e9b247655e1f6ef2f240a2f215ca99b84c840a0cd25054d9f45509975db02","output_sha256":"073a8942986be4eb1c6a3aea471eb679a7737299cc8120c7d75b3f7f70fa550b","output_samples":264704,"length_delta_samples":104,"analysis_s":0.14327805099947,"synthesis_s":3.8368363990002763,"total_s":3.9801144499997463,"rtf":0.663352408333291,"output_peak":0.33051741123199463,"output_rms":0.032365574100962496,"finished_at":"2026-10-13T04:59:01.048575+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"073a8942986be4eb1c6a3aea471eb679a7737299cc8120c7d75b3f7f70fa550b","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.810146105122257,"output_peak":0.3275608289888391},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.876568374209029,"spectral_convergence":0.47946841296824605,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.932978211605157,"spectral_convergence":0.4692841713458196,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.874270809811179,"spectral_convergence":0.5281772810580501,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":8.15973380451485e-05,"delta_db":-0.9830712538936698,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.0002984875900794054,"delta_db":-0.8045186262044136,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.0003626110135932838,"delta_db":-0.32288008437764637,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.0001436747326544414,"delta_db":-1.9492694138215252,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":7.435501656008017e-05,"delta_db":-1.793521968248242,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":8.436182078557045e-05,"delta_db":-1.9007414301206538,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.292333796111025,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0151276027967957,"pitch":null,"output_rms_dbfs":-29.808473418353,"output_peak":0.3275608289888391,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.894605798541788},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7461495288656678,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03633563122790996,"rms_dbfs":-28.793345815556208,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.25840632339133,"mean":-0.00027747472127278644,"zero_energy":false},"output":{"rms":0.03232781328388807,"rms_dbfs":-29.808473418353003,"peak":0.3275608289888391,"peak_dbfs":-9.694160768948098,"crest_factor_db":20.114312649404905,"mean":0.00013372299663150412,"zero_energy":false},"rms_difference_db":-1.0151276027967953},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":303,"both_estimated_voiced_frames":268,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9273356401384083,"absolute_difference_cents_median":7.375474881172208,"absolute_difference_cents_p95":43.920581815747774,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001036839322212551,"log_rms_mae_db":1.6603696359931615,"rms_envelope_nrmse":0.14411638398292403,"envelope_correlation":0.9885222370979059,"rms_gain_error_db":-1.015129208125999,"reference_mean_power":0.0013213586135692834,"output_mean_power":0.0010459424256994572},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.530367535346178e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002613603148630271,"output_power":0.00021876331414393522,"reference_active":true,"delta_db":-0.7726515059093897},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.246118337274924e-05,"output_power":6.566942163598113e-05,"reference_active":true,"delta_db":-0.9888637154015073},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.028948119628933e-05,"output_power":8.025243576789779e-06,"reference_active":true,"delta_db":-1.0793525515545586}],"active_band_count":3,"active_band_mae_db":0.9469559242884852}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":1.3856537273466811,"rms_envelope_nrmse":0.12858101920796705,"envelope_correlation":0.9937667792046895,"rms_gain_error_db":-1.014986041932329,"reference_mean_power":0.0013278763848253465,"output_mean_power":0.0010511363215810466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.825313121587468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002333069672738863,"output_power":0.00019426448070479746,"reference_active":true,"delta_db":-0.7953430687016138},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6274669036315644e-05,"output_power":3.5812279388303435e-05,"reference_active":true,"delta_db":-1.113113573258516},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.136183723659666e-07,"output_power":6.233829250812851e-07,"reference_active":true,"delta_db":0.06856548097513948}],"active_band_count":3,"active_band_mae_db":0.6590073743117564}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":2.780565424068677,"rms_envelope_nrmse":0.1839069642487325,"envelope_correlation":0.9734559007054152,"rms_gain_error_db":-0.6631398833751971,"reference_mean_power":0.0005515675704872272,"output_mean_power":0.00047346157175101676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3413004486315456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020691279435977171,"output_power":0.00017362013787257168,"reference_active":true,"delta_db":-0.7618724913992128},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313846735404118e-05,"output_power":2.8486160037242117e-05,"reference_active":true,"delta_db":-0.656985086788331},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.722807599466502e-06,"output_power":2.3280726795136127e-06,"reference_active":true,"delta_db":-0.680204197484271}],"active_band_count":3,"active_band_mae_db":0.6996872585572715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":2.119042517033187,"rms_envelope_nrmse":0.16528016478555121,"envelope_correlation":0.9778870659318535,"rms_gain_error_db":-0.6631651109224237,"reference_mean_power":0.0005543343450028261,"output_mean_power":0.00047583378671747366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0568367997569596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019480947374469874,"output_power":0.00016437719647910393,"reference_active":true,"delta_db":-0.7376850405041214},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.636770712067895e-05,"output_power":1.3617466025889786e-05,"reference_active":true,"delta_db":-0.7989154486110894},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1494053604425803e-07,"output_power":1.820426622669861e-07,"reference_active":true,"delta_db":-0.721451494550831}],"active_band_count":3,"active_band_mae_db":0.7526839945553473}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":3.157814777412011,"rms_envelope_nrmse":0.29907279461008524,"envelope_correlation":0.9664885394592068,"rms_gain_error_db":-2.381504099233931,"reference_mean_power":0.00027894413436310234,"output_mean_power":0.0001612006629623742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0416852970454036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011680629747545272,"output_power":6.263243217499407e-05,"reference_active":true,"delta_db":-2.7066698111197223},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1432581274087066e-05,"output_power":7.923686387250013e-06,"reference_active":true,"delta_db":-1.5921701925667444},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6649698962790796e-06,"output_power":1.646961297353575e-06,"reference_active":true,"delta_db":-0.04722991972958956}],"active_band_count":3,"active_band_mae_db":1.4486899744720185}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":2.6662771236843033,"rms_envelope_nrmse":0.2822324025903504,"envelope_correlation":0.9755377633809369,"rms_gain_error_db":-2.381656085762624,"reference_mean_power":0.0002803113215319917,"output_mean_power":0.00016198508581929895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.911634143181416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010971300365261624,"output_power":5.817288909614608e-05,"reference_active":true,"delta_db":-2.755374719391363},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3756201199309736e-06,"output_power":3.727355236119663e-06,"reference_active":true,"delta_db":-1.5902778552777028},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905227736918007e-08,"output_power":1.2586347072787633e-07,"reference_active":true,"delta_db":1.0403523828488679}],"active_band_count":3,"active_band_mae_db":1.795334985839311}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":4.331270784895108,"rms_envelope_nrmse":0.3204976710422087,"envelope_correlation":0.9302598747792777,"rms_gain_error_db":-1.3865468850106502,"reference_mean_power":9.153664281853565e-05,"output_mean_power":6.651816985620644e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.93958925235022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.468609555304607e-05,"output_power":2.8304373417562082e-05,"reference_active":true,"delta_db":-0.8830187093291988},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1944580647265697e-05,"output_power":2.0282536827395443e-05,"reference_active":true,"delta_db":-0.3420501284753927},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8366553780699585e-06,"output_power":3.5808903255642496e-06,"reference_active":true,"delta_db":-1.3055412431567288}],"active_band_count":3,"active_band_mae_db":0.8435366936537735}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":3.9013600185895663,"rms_envelope_nrmse":0.27767705053141684,"envelope_correlation":0.9495641594001972,"rms_gain_error_db":-1.3857129495331155,"reference_mean_power":9.19626023411585e-05,"output_mean_power":6.684054115966483e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.164079440499498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.035826325529751e-05,"output_power":3.337507995970161e-05,"reference_active":true,"delta_db":-0.8251015403869633},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2217507839367813e-05,"output_power":1.1396731614685657e-05,"reference_active":true,"delta_db":-0.3020230531173452},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.523678505678969e-07,"output_power":3.052574712945343e-07,"reference_active":true,"delta_db":-1.7082543010610305}],"active_band_count":3,"active_band_mae_db":0.9451262981884462}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"golf","status":"measured","mos":{"status":"measured","value":3.0296523571014404,"predicted_mos":3.0296523571014404,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"d0b597b09dda3704b44801c13b822bb964ef716ddf1aab04950908dba0c6dbb2","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.644784584001172},"started_at":"2026-10-13T05:01:55.371059+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003389708999748109,"native_input_samples":144000,"native_input_sha256":"9f9a5e5df266e123b8bdf48369c47089e4b37c456b4f9a57e50603fe3840d0e4","output_sha256":"9a72d78c57ae091dc1f2a0341b219e99b1d7c9c7de1fd4d693344ee2048feb88","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.7151487949995499,"synthesis_s":0.11834947500028647,"total_s":0.8334982699998363,"rtf":0.13891637833330606,"output_peak":0.40777096152305603,"output_rms":0.0339083004472697,"finished_at":"2026-10-13T05:01:56.215732+00:00","analysis_stages":{"external_f0_s":0.0863220100000035},"evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"9a72d78c57ae091dc1f2a0341b219e99b1d7c9c7de1fd4d693344ee2048feb88","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.398319867449313,"output_peak":0.4083480742185354},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.640681549607092,"spectral_convergence":0.5538317688193771,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.867397805513512,"spectral_convergence":0.357971131214849,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.035775758745212,"spectral_convergence":0.4014369477781935,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010247991392839571,"output_power":9.133180344497087e-05,"delta_db":-0.5001671863269376,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00035983798334207834,"output_power":0.0003419800889488745,"delta_db":-0.22106183225000425,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003912464812318744,"output_power":0.00032694443824539014,"delta_db":-0.7797649079985367,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022542700449699479,"output_power":0.00019899208857660812,"delta_db":-0.5417012981086684,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001125369423765841,"output_power":8.737470238911565e-05,"delta_db":-1.099094017312981,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013087562236247457,"output_power":0.00010197476606200511,"delta_db":-1.0836604223913957,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7042416107314207,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6119612379837454,"pitch":null,"output_rms_dbfs":-29.398319867449313,"output_peak":0.4083480742185354,"reference_rms_dbfs":-28.786358629465568,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.181285037955273},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.873145599428455,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03636487244221272,"rms_dbfs":-28.786358629465568,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.251419137300694,"mean":-0.00027771728464677455,"zero_energy":false},"output":{"rms":0.03389097059198754,"rms_dbfs":-29.398319867449313,"peak":0.4083480742185354,"peak_dbfs":-7.779389764322633,"crest_factor_db":21.61893010312668,"mean":7.336673727417613e-05,"zero_energy":false},"rms_difference_db":-0.6119612379837456},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":275,"both_estimated_voiced_frames":252,"reference_voiced_output_unvoiced_frames":37,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.8719723183391004,"absolute_difference_cents_median":4.377991230798278,"absolute_difference_cents_p95":11.315004557101757,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001036839322212551,"log_rms_mae_db":1.7216855259955215,"rms_envelope_nrmse":0.1836287142178093,"envelope_correlation":0.9620558542090752,"rms_gain_error_db":-0.6120554023819079,"reference_mean_power":0.0013234822406659546,"output_mean_power":0.0011495095343961267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.53147142139832e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026134828919373526,"output_power":0.00024357068249109345,"reference_active":true,"delta_db":-0.3059464848686396},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.260349353799019e-05,"output_power":7.215379877653484e-05,"reference_active":true,"delta_db":-0.5873921435675551},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0189793623473777e-05,"output_power":1.2346904048919347e-05,"reference_active":true,"delta_db":0.833926849229595}],"active_band_count":3,"active_band_mae_db":0.5757551592219299}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":1.2185138285600503,"rms_envelope_nrmse":0.15935753860460442,"envelope_correlation":0.9695718804751001,"rms_gain_error_db":-0.6123775736416825,"reference_mean_power":0.0013300569570584924,"output_mean_power":0.0011551343053912574},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.825418801008536e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023330242886879989,"output_power":0.0002175853604477648,"reference_active":true,"delta_db":-0.30289588317862565},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.626239064076535e-05,"output_power":4.0226167049489554e-05,"reference_active":true,"delta_db":-0.6071941868871709},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.133925981461212e-07,"output_power":7.877769767216356e-07,"reference_active":true,"delta_db":1.086647531143667}],"active_band_count":3,"active_band_mae_db":0.6655792004031545}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":1.6562142069691184,"rms_envelope_nrmse":0.22625018398570673,"envelope_correlation":0.9550418136172983,"rms_gain_error_db":-0.45909522004687053,"reference_mean_power":0.0005524863168910917,"output_mean_power":0.0004970636499879433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.3388638197983775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020671845013754738,"output_power":0.00017719823175373043,"reference_active":true,"delta_db":-0.6691985643899169},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3066717636237624e-05,"output_power":2.414934391094755e-05,"reference_active":true,"delta_db":-1.3648575052241163},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.709179177742964e-06,"output_power":3.862805515573948e-06,"reference_active":true,"delta_db":1.5406511380962686}],"active_band_count":3,"active_band_mae_db":1.1915690692367673}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":1.1449048911264286,"rms_envelope_nrmse":0.20313164907608164,"envelope_correlation":0.9609869495948049,"rms_gain_error_db":-0.459090186132464,"reference_mean_power":0.0005552626765472618,"output_mean_power":0.0004995620781579147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.054522365146115e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019463978238041968,"output_power":0.00016578951605369666,"reference_active":true,"delta_db":-0.6967454652843499},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.630682028775885e-05,"output_power":1.1910163846291277e-05,"reference_active":true,"delta_db":-1.3645154896047937},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1418180999223107e-07,"output_power":2.690941854809631e-07,"reference_active":true,"delta_db":0.9912172938902851}],"active_band_count":3,"active_band_mae_db":1.017492749593143}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":1.6392131944981572,"rms_envelope_nrmse":0.1890656259278678,"envelope_correlation":0.9751292052784707,"rms_gain_error_db":-0.7414780854040639,"reference_mean_power":0.00027938797303743874,"output_mean_power":0.00023553741170783925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0451066731721532e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011677346049165434,"output_power":0.00010274598792113839,"reference_active":true,"delta_db":-0.5557927805965833},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1414092263179511e-05,"output_power":9.84352260021788e-06,"reference_active":true,"delta_db":-0.642908359990234},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6509790752196263e-06,"output_power":1.5720699907202098e-06,"reference_active":true,"delta_db":-0.21269691455936865}],"active_band_count":3,"active_band_mae_db":0.470466018382062}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":1.1796735896694757,"rms_envelope_nrmse":0.16071827634421496,"envelope_correlation":0.9825768428286469,"rms_gain_error_db":-0.7415505581666538,"reference_mean_power":0.0002807698933190258,"output_mean_power":0.0002366984866077816},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.9148717585671515e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010966706761362735,"output_power":9.654161889901482e-05,"reference_active":true,"delta_db":-0.5536165383567347},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.361263474534157e-06,"output_power":4.6937408860424795e-06,"reference_active":true,"delta_db":-0.5774803969826128},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.871589068447308e-08,"output_power":8.63024537286184e-08,"reference_active":true,"delta_db":-0.5836392474785084}],"active_band_count":3,"active_band_mae_db":0.571578727605952}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":2.8909432553749106,"rms_envelope_nrmse":0.24422454471537774,"envelope_correlation":0.9602203185302726,"rms_gain_error_db":-0.7927917813119796,"reference_mean_power":9.166182930245578e-05,"output_mean_power":7.636763514105329e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.955808595384871e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.477171057797415e-05,"output_power":3.087410582586239e-05,"reference_active":true,"delta_db":-0.5163166768469553},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.201403706694246e-05,"output_power":2.3503856885180238e-05,"reference_active":true,"delta_db":0.2843944063870467},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.823230963381859e-06,"output_power":4.892401873702467e-06,"reference_active":true,"delta_db":0.061840649485756995}],"active_band_count":3,"active_band_mae_db":0.28751724423991964}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":2.6542617062099563,"rms_envelope_nrmse":0.1969795105036992,"envelope_correlation":0.9756596952828013,"rms_gain_error_db":-0.7926570252690884,"reference_mean_power":9.210383843228777e-05,"output_mean_power":7.673827408608858e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.178987663523889e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.047459033949941e-05,"output_power":3.7991600468596554e-05,"reference_active":true,"delta_db":-0.2749487234586905},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2222550340356595e-05,"output_power":1.3310405848633862e-05,"reference_active":true,"delta_db":0.3702946313910618},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.5267130124840183e-07,"output_power":4.486637721000865e-07,"reference_active":true,"delta_db":-0.03861958367422128}],"active_band_count":3,"active_band_mae_db":0.2279543128413245}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.294013738632202,"predicted_mos":3.294013738632202,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"b52e36f5f16b107e26056f2d1292ca119c762e76ce690d0b8a1b52c94115272f","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.224423752000803},"started_at":"2026-10-13T05:04:41.986112+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027530799998203292,"native_input_samples":144000,"native_input_sha256":"9f9a5e5df266e123b8bdf48369c47089e4b37c456b4f9a57e50603fe3840d0e4","output_sha256":"acbfe4f555555e3f6a5ff6eab8b947b8848b00bd45d59763fb5ce3ce4a4afe69","output_samples":144120,"length_delta_samples":120,"analysis_s":2.043272257999888,"synthesis_s":3.655552997000086,"total_s":5.698825254999974,"rtf":0.9498042091666624,"output_peak":0.43336954712867737,"output_rms":0.035049411156090414,"finished_at":"2026-10-13T05:04:47.696806+00:00","analysis_stages":{"external_f0_s":1.3095344819994352},"evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"acbfe4f555555e3f6a5ff6eab8b947b8848b00bd45d59763fb5ce3ce4a4afe69","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.11299592214156,"output_peak":0.41222881927846905},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.937608353991166,"spectral_convergence":0.5451624251089516,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.360090370961492,"spectral_convergence":0.33751606846075555,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.385931960569872,"spectral_convergence":0.34239437729543026,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":8.579395188977114e-05,"delta_db":-0.7652644335312573,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.00034989805381263064,"delta_db":-0.11436615356790712,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.00037777049075624617,"delta_db":-0.14500968894593927,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.0002175042962767233,"delta_db":-0.1483949924343848,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":0.00011040123602243902,"delta_db":-0.07688535845969113,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":8.622447091554983e-05,"delta_db":-1.8058954942031207,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.50930268685705,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.31611192122540743,"pitch":null,"output_rms_dbfs":-29.109457736781614,"output_peak":0.41222881927846905,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.56121022850751},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8653689517643307,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03633563122790996,"rms_dbfs":-28.793345815556208,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.25840632339133,"mean":-0.00027747472127278644,"zero_energy":false},"output":{"rms":0.03503701613541542,"rms_dbfs":-29.109457736781614,"peak":0.41222881927846905,"peak_dbfs":-7.697232991742756,"crest_factor_db":21.412224745038856,"mean":0.00018445987615716471,"zero_energy":false},"rms_difference_db":-0.31611192122540643},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":319,"both_estimated_voiced_frames":281,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":38,"paired_voiced_fraction":0.972318339100346,"absolute_difference_cents_median":4.919388381848577,"absolute_difference_cents_p95":26.44041205408258,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001036839322212551,"log_rms_mae_db":2.273225949310798,"rms_envelope_nrmse":0.14590771892411403,"envelope_correlation":0.9749081855627656,"rms_gain_error_db":-0.31608569278703563,"reference_mean_power":0.0013213586135692834,"output_mean_power":0.0012286045824471963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.530367535346178e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002613603148630271,"output_power":0.0002644795202935631,"reference_active":true,"delta_db":0.05152403849449877},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.246118337274924e-05,"output_power":6.502862950995671e-05,"reference_active":true,"delta_db":-1.0314496165415916},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.028948119628933e-05,"output_power":7.249247186474345e-06,"reference_active":true,"delta_db":-1.5210056916968129}],"active_band_count":3,"active_band_mae_db":0.8679931155776345}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":1.714666456383978,"rms_envelope_nrmse":0.10311424777830827,"envelope_correlation":0.9871737969837018,"rms_gain_error_db":-0.31590313400613756,"reference_mean_power":0.0013278763848253465,"output_mean_power":0.001234716733625917},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.825313121587468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002333069672738863,"output_power":0.00023522974296431056,"reference_active":true,"delta_db":0.03564525646803446},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6274669036315644e-05,"output_power":3.549321518565307e-05,"reference_active":true,"delta_db":-1.1519797919785133},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.136183723659666e-07,"output_power":4.811185553360778e-07,"reference_active":true,"delta_db":-1.0564624766559145}],"active_band_count":3,"active_band_mae_db":0.7480291750341541}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":1.994469061670732,"rms_envelope_nrmse":0.1485790166777438,"envelope_correlation":0.9806002712751637,"rms_gain_error_db":-0.14797993464869422,"reference_mean_power":0.0005515675704872272,"output_mean_power":0.0005330902387531068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3413004486315456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020691279435977171,"output_power":0.00019832621987463065,"reference_active":true,"delta_db":-0.18407211654191263},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313846735404118e-05,"output_power":3.1533199503812875e-05,"reference_active":true,"delta_db":-0.21564380030742333},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.722807599466502e-06,"output_power":2.831428596002684e-06,"reference_active":true,"delta_db":0.1698865983648854}],"active_band_count":3,"active_band_mae_db":0.18986750507140712}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":1.3360865677882787,"rms_envelope_nrmse":0.1014608504884813,"envelope_correlation":0.9902833235283963,"rms_gain_error_db":-0.14796643818229785,"reference_mean_power":0.0005543343450028261,"output_mean_power":0.0005357659922194194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0568367997569596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019480947374469874,"output_power":0.00018679445680808157,"reference_active":true,"delta_db":-0.18246088911346275},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.636770712067895e-05,"output_power":1.5591640612024507e-05,"reference_active":true,"delta_db":-0.2109602957094076},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1494053604425803e-07,"output_power":2.409676256328794e-07,"reference_active":true,"delta_db":0.4964037073217006}],"active_band_count":3,"active_band_mae_db":0.29660829738152367}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":1.8195943155149645,"rms_envelope_nrmse":0.14996164624666067,"envelope_correlation":0.9819961054487606,"rms_gain_error_db":-0.08420121736143718,"reference_mean_power":0.00027894413436310234,"output_mean_power":0.0002735880422927855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0416852970454036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011680629747545272,"output_power":0.00011276198408245073,"reference_active":true,"delta_db":-0.15303549074323225},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1432581274087066e-05,"output_power":1.1643250558964068e-05,"reference_active":true,"delta_db":0.07929945978092905},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6649698962790796e-06,"output_power":1.6096224978287963e-06,"reference_active":true,"delta_db":-0.14682352002552052}],"active_band_count":3,"active_band_mae_db":0.12638615684989393}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":1.1569461717727938,"rms_envelope_nrmse":0.09429995043421315,"envelope_correlation":0.9924486511111075,"rms_gain_error_db":-0.08399153896461728,"reference_mean_power":0.0002803113215319917,"output_mean_power":0.0002749422516366828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.911634143181416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010971300365261624,"output_power":0.00010597133899131385,"reference_active":true,"delta_db":-0.15069683198761225},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3756201199309736e-06,"output_power":5.445405841904605e-06,"reference_active":true,"delta_db":0.056016821800442915},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905227736918007e-08,"output_power":9.84356474240255e-08,"reference_active":true,"delta_db":-0.027120630602202222}],"active_band_count":3,"active_band_mae_db":0.07794476146341912}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":1.9065548222413544,"rms_envelope_nrmse":0.27247874139496214,"envelope_correlation":0.9393476339748517,"rms_gain_error_db":-0.7200701141842302,"reference_mean_power":9.153664281853565e-05,"output_mean_power":7.755110116912125e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.93958925235022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.468609555304607e-05,"output_power":2.9373087027473347e-05,"reference_active":true,"delta_db":-0.7220582412230305},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1944580647265697e-05,"output_power":1.6706628061463072e-05,"reference_active":true,"delta_db":-1.1843848200685414},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8366553780699585e-06,"output_power":3.0425069884899137e-06,"reference_active":true,"delta_db":-2.013135596335718}],"active_band_count":3,"active_band_mae_db":1.3065262192090967}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":1.2137203713907345,"rms_envelope_nrmse":0.16482877239158072,"envelope_correlation":0.9794727557745496,"rms_gain_error_db":-0.7200032458635017,"reference_mean_power":9.19626023411585e-05,"output_mean_power":7.791317954776653e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.164079440499498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.035826325529751e-05,"output_power":3.277798634799707e-05,"reference_active":true,"delta_db":-0.9035019887620765},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2217507839367813e-05,"output_power":9.013051399615926e-06,"reference_active":true,"delta_db":-1.321107785980355},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.523678505678969e-07,"output_power":2.835510523340893e-07,"reference_active":true,"delta_db":-2.0286046896988372}],"active_band_count":3,"active_band_mae_db":1.4177381548137564}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8041791915893555,"predicted_mos":2.8041791915893555,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"131f5cbfc576c4e7ab1895a2f80fc252c4eec614d209e06a33a82e3b7d72346c","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.404239633000543},"started_at":"2026-10-13T05:10:01.405460+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003386653999768896,"native_input_samples":144000,"native_input_sha256":"9f9a5e5df266e123b8bdf48369c47089e4b37c456b4f9a57e50603fe3840d0e4","output_sha256":"71a6192bbec9de97442a6ec40c13a9bed0f950784a1da44fb827f0742328bb05","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7078587489995698,"synthesis_s":0.1302352430011524,"total_s":0.8380939920007222,"rtf":0.13968233200012037,"output_peak":0.33784714341163635,"output_rms":0.03525500396658565,"finished_at":"2026-10-13T05:10:02.255023+00:00","analysis_stages":{"external_f0_s":0.15657044799991127},"evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"71a6192bbec9de97442a6ec40c13a9bed0f950784a1da44fb827f0742328bb05","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.062614610853075,"output_peak":0.33814085526559906},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.68708715617969,"spectral_convergence":0.2766175536549339,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.527973049109606,"spectral_convergence":0.2654873213789037,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.183209864936633,"spectral_convergence":0.32883049604330955,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":9.823231718698932e-05,"delta_db":-0.1772872768620627,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.00035072086547448245,"delta_db":-0.10416538054174726,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.0003632055307214848,"delta_db":-0.3157654610491897,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.0002092634031853671,"delta_db":-0.31614055230441906,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":0.0001057011690391412,"delta_db":-0.26582680944074144,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.00011584309557225055,"delta_db":-0.523499333068788,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.28378080221115803,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26206036114550546,"pitch":null,"output_rms_dbfs":-29.055406176701712,"output_peak":0.33814085526559906,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.799423356741976},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9028232937741132,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03633563122790996,"rms_dbfs":-28.793345815556208,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.25840632339133,"mean":-0.00027747472127278644,"zero_energy":false},"output":{"rms":0.03525572834561379,"rms_dbfs":-29.055406176701712,"peak":0.33814085526559906,"peak_dbfs":-9.418047064160103,"crest_factor_db":19.63735911254161,"mean":-2.27004988984771e-06,"zero_energy":false},"rms_difference_db":-0.2620603611455046},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":256,"both_estimated_voiced_frames":245,"reference_voiced_output_unvoiced_frames":44,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8477508650519031,"absolute_difference_cents_median":0.6868304216409654,"absolute_difference_cents_p95":3.7814269181264533,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001036839322212551,"log_rms_mae_db":0.7927184081430536,"rms_envelope_nrmse":0.07974690594629408,"envelope_correlation":0.9929326524262758,"rms_gain_error_db":-0.26204989840091025,"reference_mean_power":0.0013213586135692834,"output_mean_power":0.0012439866229667807},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.530367535346178e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002613603148630271,"output_power":0.00025194248770752265,"reference_active":true,"delta_db":-0.15938231352307725},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.246118337274924e-05,"output_power":7.308328020171135e-05,"reference_active":true,"delta_db":-0.5243153133833539},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.028948119628933e-05,"output_power":1.076271314108914e-05,"reference_active":true,"delta_db":0.19528287320104656}],"active_band_count":3,"active_band_mae_db":0.2929935000358259}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":0.4535343363201189,"rms_envelope_nrmse":0.06430001419917408,"envelope_correlation":0.9953212351870927,"rms_gain_error_db":-0.26202192950160774,"reference_mean_power":0.0013278763848253465,"output_mean_power":0.0012501307978437868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.825313121587468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002333069672738863,"output_power":0.00022439416957519288,"reference_active":true,"delta_db":-0.1691613990107575},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6274669036315644e-05,"output_power":4.1396473177303303e-05,"reference_active":true,"delta_db":-0.48379978841574567},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.136183723659666e-07,"output_power":5.96420534310422e-07,"reference_active":true,"delta_db":-0.12345766888875896}],"active_band_count":3,"active_band_mae_db":0.2588062854384207}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":0.8856519396520137,"rms_envelope_nrmse":0.10149983878365818,"envelope_correlation":0.9914235920470482,"rms_gain_error_db":-0.24746482539987757,"reference_mean_power":0.0005515675704872272,"output_mean_power":0.0005210174167567774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3413004486315456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020691279435977171,"output_power":0.0001969433891815799,"reference_active":true,"delta_db":-0.21445938516836321},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313846735404118e-05,"output_power":2.8013890984713886e-05,"reference_active":true,"delta_db":-0.7295898426218546},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.722807599466502e-06,"output_power":3.48418212241567e-06,"reference_active":true,"delta_db":1.070838939592773}],"active_band_count":3,"active_band_mae_db":0.6716293891276637}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":0.4868405723033801,"rms_envelope_nrmse":0.07854392861416853,"envelope_correlation":0.994717874674272,"rms_gain_error_db":-0.24746366476126155,"reference_mean_power":0.0005543343450028261,"output_mean_power":0.0005236310854597478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0568367997569596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019480947374469874,"output_power":0.00018517493453248715,"reference_active":true,"delta_db":-0.22027873372093126},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.636770712067895e-05,"output_power":1.3916126923930854e-05,"reference_active":true,"delta_db":-0.7046946415411671},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1494053604425803e-07,"output_power":2.1941110688597118e-07,"reference_active":true,"delta_db":0.08940280729675801}],"active_band_count":3,"active_band_mae_db":0.33812539418628546}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":0.827690771419012,"rms_envelope_nrmse":0.08997019334590216,"envelope_correlation":0.9942854652916063,"rms_gain_error_db":-0.29630702677312487,"reference_mean_power":0.00027894413436310234,"output_mean_power":0.0002605472722847941},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0416852970454036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011680629747545272,"output_power":0.00011206749319730385,"reference_active":true,"delta_db":-0.1798660044383275},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1432581274087066e-05,"output_power":9.895645642945558e-06,"reference_active":true,"delta_db":-0.6270016237996573},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6649698962790796e-06,"output_power":1.6443270996310427e-06,"reference_active":true,"delta_db":-0.05418171271412142}],"active_band_count":3,"active_band_mae_db":0.2870164469840354}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":0.5376657940233038,"rms_envelope_nrmse":0.06850767244242731,"envelope_correlation":0.9969030839524372,"rms_gain_error_db":-0.2962769398847599,"reference_mean_power":0.0002803113215319917,"output_mean_power":0.0002618261048894832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.911634143181416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010971300365261624,"output_power":0.00010540849013105325,"reference_active":true,"delta_db":-0.1738251250594745},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3756201199309736e-06,"output_power":4.678684793236808e-06,"reference_active":true,"delta_db":-0.6030478334265964},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905227736918007e-08,"output_power":8.851369657836096e-08,"reference_active":true,"delta_db":-0.48853986635669916}],"active_band_count":3,"active_band_mae_db":0.42180427494759004}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":0.9285222496737614,"rms_envelope_nrmse":0.10773982022991796,"envelope_correlation":0.9910341316439445,"rms_gain_error_db":-0.25119189597775704,"reference_mean_power":9.153664281853565e-05,"output_mean_power":8.639245007164251e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.93958925235022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.468609555304607e-05,"output_power":3.321814492550817e-05,"reference_active":true,"delta_db":-0.18780040628975847},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1944580647265697e-05,"output_power":2.0827636595050687e-05,"reference_active":true,"delta_db":-0.22687294646038475},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8366553780699585e-06,"output_power":3.7274216583280564e-06,"reference_active":true,"delta_db":-1.1313661972176878}],"active_band_count":3,"active_band_mae_db":0.5153465166559437}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":0.6292536837642745,"rms_envelope_nrmse":0.06568267948394294,"envelope_correlation":0.9968856383845485,"rms_gain_error_db":-0.25100221580754306,"reference_mean_power":9.19626023411585e-05,"output_mean_power":8.67982623166585e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.164079440499498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.035826325529751e-05,"output_power":3.8388487028532974e-05,"reference_active":true,"delta_db":-0.21731472900584164},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2217507839367813e-05,"output_power":1.1345304241845e-05,"reference_active":true,"delta_db":-0.321664797115016},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.523678505678969e-07,"output_power":3.393217233140979e-07,"reference_active":true,"delta_db":-1.2488006839060815}],"active_band_count":3,"active_band_mae_db":0.5959267366756463}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135766-0005","method":"magphase","status":"measured","mos":{"status":"measured","value":3.970635414123535,"predicted_mos":3.970635414123535,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288407,"input_sample_rate":48000,"input_seconds":6.008479166666667,"input_float32_sha256":"be3fa97c90fb14fd7d34779b223d7aa6e5fc01befc6cf3ba9d75f9d6f4f7b924","internal_samples":96136,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2802072529993893},"started_at":"2026-10-13T05:12:21.386242+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009051266000824398,"native_input_samples":288000,"native_input_sha256":"deece7841137ca06cf75ebfb4b2d7b23b2c210591b8c00caa7e3c3643d56f41a","output_sha256":"84a8de8ab791dfa80a44189a4eece39b82c207478e21ba3f91d2cc30843e2da0","output_samples":288407,"length_delta_samples":407,"analysis_s":3.786515107000014,"synthesis_s":1.1641068380013166,"total_s":4.9506219450013305,"rtf":0.8251036575002217,"output_peak":0.43241871571376944,"output_rms":0.03548710420775418,"finished_at":"2026-10-13T05:12:26.357760+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"84a8de8ab791dfa80a44189a4eece39b82c207478e21ba3f91d2cc30843e2da0","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96136,"length_delta_samples":136,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.009775763443475,"output_peak":0.41695635591657354},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9985853374386287,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.567978375039199,"spectral_convergence":0.34858075686906326,"spectral_convergence_status":"ok","amplitude_floor":6.643709912602353e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.068131222808017,"spectral_convergence":0.24546497303482334,"spectral_convergence_status":"ok","amplitude_floor":5.274693968404161e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.457641634815721,"spectral_convergence":0.2641001478052246,"spectral_convergence_status":"ok","amplitude_floor":3.8895875894681354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010232532114960487,"output_power":8.380941078637209e-05,"delta_db":-0.8669032895681301,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003592345857428397,"output_power":0.0003337345778133994,"delta_db":-0.31976936550375346,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003905970862424533,"output_power":0.00036978636066355467,"delta_db":-0.23778110793824112,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022506467673995083,"output_power":0.0002300144872531337,"delta_db":0.09447851455004923,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011237312839533447,"output_power":0.00012093034663991574,"delta_db":0.31872826244605806,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00013068329846048222,"output_power":0.0001195568557882076,"delta_db":-0.38645602523534583,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.37068609420692966,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2104276575289435,"pitch":null,"output_rms_dbfs":-29.00377347308515,"output_peak":0.41695635591657354,"reference_rms_dbfs":-28.793345815556208,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.36458374422098},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8872561363477174,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03633563122790996,"rms_dbfs":-28.793345815556208,"peak":0.37432861328125,"peak_dbfs":-8.534939492164874,"crest_factor_db":20.25840632339133,"mean":-0.00027747472127278644,"zero_energy":false},"output":{"rms":0.035465927859631464,"rms_dbfs":-29.00377347308515,"peak":0.41695635591657354,"peak_dbfs":-7.59818803129567,"crest_factor_db":21.40558544178948,"mean":0.0006532559713592422,"zero_energy":false},"rms_difference_db":-0.21042765752894255},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.04815257654991e-06,"threshold_dbfs":-59.795754937544,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":289,"output_estimated_voiced_frames":277,"both_estimated_voiced_frames":272,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9411764705882353,"absolute_difference_cents_median":2.56179427836174,"absolute_difference_cents_p95":8.531980789433355,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001036839322212551,"log_rms_mae_db":1.1539534021015925,"rms_envelope_nrmse":0.072673195229861,"envelope_correlation":0.9942966969385307,"rms_gain_error_db":-0.2104170148874893,"reference_mean_power":0.0013213586135692834,"output_mean_power":0.0012588645347883876},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.530367535346178e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002613603148630271,"output_power":0.00026776720758418517,"reference_active":true,"delta_db":0.10517744898497836},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.246118337274924e-05,"output_power":7.321392749772233e-05,"reference_active":true,"delta_db":-0.5165585809468902},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.028948119628933e-05,"output_power":8.533729317252084e-06,"reference_active":true,"delta_db":-0.8125461456164982}],"active_band_count":3,"active_band_mae_db":0.47809405851612224}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.673473491549852e-05,"log_rms_mae_db":0.7486620728884154,"rms_envelope_nrmse":0.05106677614070309,"envelope_correlation":0.9973701618587717,"rms_gain_error_db":-0.2102744316828009,"reference_mean_power":0.0013278763848253465,"output_mean_power":0.0012651155802885943},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.825313121587468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002333069672738863,"output_power":0.00023822104597014435,"reference_active":true,"delta_db":0.09052418817180387},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6274669036315644e-05,"output_power":4.0128805056404264e-05,"reference_active":true,"delta_db":-0.6188709365910906},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.136183723659666e-07,"output_power":5.965699455707995e-07,"reference_active":true,"delta_db":-0.12236983981803812}],"active_band_count":3,"active_band_mae_db":0.2772549881936442}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.601318795186366e-05,"log_rms_mae_db":1.313751517550461,"rms_envelope_nrmse":0.052236751675412404,"envelope_correlation":0.998211241299268,"rms_gain_error_db":-0.23951576287652904,"reference_mean_power":0.0005515675704872272,"output_mean_power":0.0005219719286777098},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3413004486315456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020691279435977171,"output_power":0.00019726415443250007,"reference_active":true,"delta_db":-0.20739170665927978},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313846735404118e-05,"output_power":3.057441730008026e-05,"reference_active":true,"delta_db":-0.3497422971197511},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.722807599466502e-06,"output_power":2.5474977361725055e-06,"reference_active":true,"delta_db":-0.28903147323845185}],"active_band_count":3,"active_band_mae_db":0.28205515900582756}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.977903967560393e-05,"log_rms_mae_db":0.8845219392286718,"rms_envelope_nrmse":0.03957886821229355,"envelope_correlation":0.9991899710717501,"rms_gain_error_db":-0.23951890979769974,"reference_mean_power":0.0005543343450028261,"output_mean_power":0.0005245898653460231},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0568367997569596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019480947374469874,"output_power":0.0001856322114686147,"reference_active":true,"delta_db":-0.20956734625848947},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.636770712067895e-05,"output_power":1.519726508400566e-05,"reference_active":true,"delta_db":-0.322224064102953},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1494053604425803e-07,"output_power":1.8148413261778712e-07,"reference_active":true,"delta_db":-0.7347966750935626}],"active_band_count":3,"active_band_mae_db":0.4221960284850017}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.251077329795778e-05,"log_rms_mae_db":0.9397638597977126,"rms_envelope_nrmse":0.05771513003688673,"envelope_correlation":0.997645645080142,"rms_gain_error_db":0.1311025021884192,"reference_mean_power":0.00027894413436310234,"output_mean_power":0.0002874931383181866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0416852970454036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011680629747545272,"output_power":0.00012330419814871057,"reference_active":true,"delta_db":0.23511605420645731},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1432581274087066e-05,"output_power":1.22935472687101e-05,"reference_active":true,"delta_db":0.3153291797051442},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6649698962790796e-06,"output_power":1.70839561343188e-06,"reference_active":true,"delta_db":0.11182062047190516}],"active_band_count":3,"active_band_mae_db":0.22075528479450224}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.285244854844391e-05,"log_rms_mae_db":0.6153222807007759,"rms_envelope_nrmse":0.037672399617653014,"envelope_correlation":0.9990635999821563,"rms_gain_error_db":0.1312883908820522,"reference_mean_power":0.0002803113215319917,"output_mean_power":0.00028891459265630904},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.911634143181416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010971300365261624,"output_power":0.00011576288830684542,"reference_active":true,"delta_db":0.23311248921244643},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3756201199309736e-06,"output_power":5.824000616186935e-06,"reference_active":true,"delta_db":0.34792841502591426},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.905227736918007e-08,"output_power":1.0067047660463743e-07,"reference_active":true,"delta_db":0.07037659589068591}],"active_band_count":3,"active_band_mae_db":0.21713916670968217}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.076518340019584e-05,"log_rms_mae_db":1.4480365754811506,"rms_envelope_nrmse":0.1608871667973314,"envelope_correlation":0.981445195797812,"rms_gain_error_db":0.3589070581376608,"reference_mean_power":9.153664281853565e-05,"output_mean_power":9.94227304681097e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.93958925235022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.468609555304607e-05,"output_power":3.4755836138345446e-05,"reference_active":true,"delta_db":0.008723248563373127},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1944580647265697e-05,"output_power":2.158258142383389e-05,"reference_active":true,"delta_db":-0.07223898021495964},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8366553780699585e-06,"output_power":4.57745847516629e-06,"reference_active":true,"delta_db":-0.23920730840584478}],"active_band_count":3,"active_band_mae_db":0.10672317906139252}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.411903094098525e-05,"log_rms_mae_db":1.108528694945951,"rms_envelope_nrmse":0.11292306195157246,"envelope_correlation":0.9911930478307386,"rms_gain_error_db":0.35939862782802423,"reference_mean_power":9.19626023411585e-05,"output_mean_power":9.989669383910407e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.164079440499498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.035826325529751e-05,"output_power":3.907098184144518e-05,"reference_active":true,"delta_db":-0.14078143845987742},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2217507839367813e-05,"output_power":1.1975088357553385e-05,"reference_active":true,"delta_db":-0.08703899848302303},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.523678505678969e-07,"output_power":4.2954173192600744e-07,"reference_active":true,"delta_db":-0.22486368484068225}],"active_band_count":3,"active_band_mae_db":0.15089470726119422}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"hnm","status":"measured","mos":{"status":"measured","value":2.0289130210876465,"predicted_mos":2.0289130210876465,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e72017c55e1a046a015a8b4a6bc703ff279b6dee1ed742f9af9b65c33e6077bb","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.520422920000783},"started_at":"2026-10-13T04:38:09.236024+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.645700012886664e-05,"native_input_samples":96000,"native_input_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","output_sha256":"36bc3f9b0bb5aea6c3bf8fd0d6b393e2dd2662868cda3ce5263f3026541bda4c","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1640304279999327,"synthesis_s":0.5350002100003621,"total_s":1.6990306380002949,"rtf":0.28317177300004914,"output_peak":0.2945119914007137,"output_rms":0.03647371551023896,"finished_at":"2026-10-13T04:38:10.941507+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"36bc3f9b0bb5aea6c3bf8fd0d6b393e2dd2662868cda3ce5263f3026541bda4c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.478735409582556,"spectral_convergence":0.37469992579414846,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.294769967896118,"spectral_convergence":0.3634413388798525,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.616074119561195,"spectral_convergence":0.3827448829668065,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":7.31165560560787e-05,"delta_db":-0.10256706405517524,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.00038466791824178274,"delta_db":-0.08183633128286107,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.0003738190849693119,"delta_db":-0.16897269132881415,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00014311935785805022,"delta_db":-0.17537852007241223,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":9.143842437855187e-05,"delta_db":-0.2847551756308283,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.0002641705816180704,"delta_db":0.18689275196134386,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16673375572190582,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08041599966046328,"pitch":null,"output_rms_dbfs":-28.760399873070188,"output_peak":0.2945119914007137,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.463193165679956},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.478735409582556,"spectral_convergence":0.37469992579414846,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.294769967896118,"spectral_convergence":0.3634413388798525,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.616074119561195,"spectral_convergence":0.3827448829668065,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":7.31165560560787e-05,"delta_db":-0.10256706405517524,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.00038466791824178274,"delta_db":-0.08183633128286107,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.0003738190849693119,"delta_db":-0.16897269132881415,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00014311935785805022,"delta_db":-0.17537852007241223,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":9.143842437855187e-05,"delta_db":-0.2847551756308283,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.0002641705816180704,"delta_db":0.18689275196134386,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16673375572190582,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08041599966046328,"pitch":null,"output_rms_dbfs":-28.760399873070188,"output_peak":0.2945119914007137,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.463193165679956},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8652336567227503,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03681296571269971,"rms_dbfs":-28.67998387340972,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.2140993950677,"mean":-0.0002700039545694987,"zero_energy":false},"output":{"rms":0.03647371551023896,"rms_dbfs":-28.760399873070188,"peak":0.2945119914007137,"peak_dbfs":-10.61794035416659,"crest_factor_db":18.142459518903596,"mean":-3.0877729546352714e-05,"zero_energy":false},"rms_difference_db":-0.08041599966046675},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":286,"output_estimated_voiced_frames":256,"both_estimated_voiced_frames":254,"reference_voiced_output_unvoiced_frames":32,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.8881118881118881,"absolute_difference_cents_median":1.6393811033513686,"absolute_difference_cents_p95":7.379522292512611,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":0.8678347411611979,"rms_envelope_nrmse":0.08643714572712001,"envelope_correlation":0.9899150961591262,"rms_gain_error_db":-0.07929869051241129,"reference_mean_power":0.0013552494150092414,"output_mean_power":0.0013307281974006225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.097356675930481e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024105831474109917,"output_power":0.00023320796602159318,"reference_active":true,"delta_db":-0.1437873488204616},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7795131277493678e-05,"output_power":2.3814336257323312e-05,"reference_active":true,"delta_db":-0.6713024801723058},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0822198872325408e-05,"output_power":9.67184015758367e-06,"reference_active":true,"delta_db":-0.4880639989912369}],"active_band_count":3,"active_band_mae_db":0.43438460932800144}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":0.4490627737261544,"rms_envelope_nrmse":0.04725815259519829,"envelope_correlation":0.9967653192757006,"rms_gain_error_db":-0.07909923338072186,"reference_mean_power":0.0013563897957452433,"output_mean_power":0.0013319091134354547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.582417446963071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021668841148527672,"output_power":0.00021050675146144858,"reference_active":true,"delta_db":-0.12569656563815323},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1881573874210085e-05,"output_power":1.0205281254135019e-05,"reference_active":true,"delta_db":-0.6604899437145577},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.720949484930855e-07,"output_power":5.83186876668799e-07,"reference_active":true,"delta_db":-0.6162288864623625}],"active_band_count":3,"active_band_mae_db":0.4674717986050245}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":1.3985817626473553,"rms_envelope_nrmse":0.10785383965275866,"envelope_correlation":0.9900155426994961,"rms_gain_error_db":-0.1822365623031419,"reference_mean_power":0.0005726237314993749,"output_mean_power":0.0005490927209073466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0939920051603367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018116410089920082,"output_power":0.00016889327757984259,"reference_active":true,"delta_db":-0.304597791846005},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.90482495868649e-05,"output_power":2.3813726524694112e-05,"reference_active":true,"delta_db":-0.8629260553390204},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.620984223411991e-06,"output_power":3.96922281670463e-06,"reference_active":true,"delta_db":-1.5110688715541158}],"active_band_count":3,"active_band_mae_db":0.8928642395797137}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":0.7881498841880855,"rms_envelope_nrmse":0.060004153762528586,"envelope_correlation":0.9969738036444346,"rms_gain_error_db":-0.1817559972712229,"reference_mean_power":0.0005714831913152958,"output_mean_power":0.0005480596910515544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7768006225472703e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016765591938032763,"output_power":0.00015570968768774838,"reference_active":true,"delta_db":-0.3210325792419336},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3133776830395815e-05,"output_power":1.0929603786250728e-05,"reference_active":true,"delta_db":-0.797852140272089},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7638611347445743e-07,"output_power":3.0098319388318e-07,"reference_active":true,"delta_db":-1.9941684703003828}],"active_band_count":3,"active_band_mae_db":1.0376843966048017}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":1.315234053640365,"rms_envelope_nrmse":0.12226936345847704,"envelope_correlation":0.9889148739717155,"rms_gain_error_db":-0.28973268487315795,"reference_mean_power":0.00020459070686192212,"output_mean_power":0.00019138708789211633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3670238086467602e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.952541720107037e-05,"output_power":6.380130721270906e-05,"reference_active":true,"delta_db":-0.37314026625877306},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.449912227458813e-06,"output_power":5.875859256471101e-06,"reference_active":true,"delta_db":-1.0307977140070814},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5578562674598195e-06,"output_power":1.1699351659429378e-06,"reference_active":true,"delta_db":-1.2436559065510062}],"active_band_count":3,"active_band_mae_db":0.8825312956056202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":0.7962208202430207,"rms_envelope_nrmse":0.07865923339900475,"envelope_correlation":0.9957540888083153,"rms_gain_error_db":-0.2925994515074545,"reference_mean_power":0.00020499017149025155,"output_mean_power":0.00019163423335136223},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2857171527440002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.293887842389662e-05,"output_power":5.7745502060792215e-05,"reference_active":true,"delta_db":-0.3740083786593846},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7135208044031714e-06,"output_power":2.9790531443505283e-06,"reference_active":true,"delta_db":-0.9570761051110434},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.333412649997229e-07,"output_power":9.539623569771224e-08,"reference_active":true,"delta_db":-1.4543333299902448}],"active_band_count":3,"active_band_mae_db":0.928472604586891}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":1.3105710659777292,"rms_envelope_nrmse":0.15340506297308765,"envelope_correlation":0.9864177264696332,"rms_gain_error_db":0.407433671872279,"reference_mean_power":0.00010545678906217639,"output_mean_power":0.00011582916112103033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.56955140861397e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.404725630082888e-05,"output_power":3.621291992161269e-05,"reference_active":true,"delta_db":0.2678142425396423},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.550080432854382e-06,"output_power":6.949011527638354e-06,"reference_active":true,"delta_db":-0.36028546172861503},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.516411722896476e-06,"output_power":2.3431726319702443e-06,"reference_active":true,"delta_db":-1.7629543303568376}],"active_band_count":3,"active_band_mae_db":0.7970180115416983}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":0.8092865519623624,"rms_envelope_nrmse":0.10359844130554527,"envelope_correlation":0.9945304427088273,"rms_gain_error_db":0.40746778126947714,"reference_mean_power":0.0001059674543662488,"output_mean_power":0.00011639096786083784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.064372154195949e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.39238699115062e-05,"output_power":3.5484093733557646e-05,"reference_active":true,"delta_db":0.1952832864935156},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.204493449095075e-06,"output_power":3.646712881649012e-06,"reference_active":true,"delta_db":-0.618121090050558},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.457795181461805e-07,"output_power":1.4228376271706237e-07,"reference_active":true,"delta_db":-2.3739034679661923}],"active_band_count":3,"active_band_mae_db":1.0624359481700887}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.108083963394165,"predicted_mos":3.108083963394165,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"83c5a1ab26757549d838606ec7a2cdf051bfcc4261bad6d5c729418129873739","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.356164600001648},"started_at":"2026-10-13T04:41:00.319961+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.055000004475005e-05,"native_input_samples":96000,"native_input_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","output_sha256":"71d8a938e1f280ac485e4ee87ee0feda323967dd290b1bb347c48efa589ca8d8","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2977572859999782,"synthesis_s":0.5081396680006947,"total_s":1.8058969540006728,"rtf":0.3009828256667788,"output_peak":0.3816178104403332,"output_rms":0.040047052265426965,"finished_at":"2026-10-13T04:41:02.133070+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"71d8a938e1f280ac485e4ee87ee0feda323967dd290b1bb347c48efa589ca8d8","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.277094450539818,"spectral_convergence":0.615898701634721,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.412007753594387,"spectral_convergence":0.37134683725884937,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.881358775516516,"spectral_convergence":0.3951471174512605,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":6.916324049439895e-05,"delta_db":-0.3439710041799785,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.000480954275751447,"delta_db":0.8883418975390568,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.000495025256353086,"delta_db":1.0506861812411472,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00017132630539018218,"delta_db":0.6058781853466181,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":0.00011217067220077199,"delta_db":0.6027507099284077,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.00027512664495995295,"delta_db":0.36337475007910797,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6425004547190528,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7313949649743178,"pitch":null,"output_rms_dbfs":-27.948588908435408,"output_peak":0.3816178104403332,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.856820326550241},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.277094450539818,"spectral_convergence":0.615898701634721,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.412007753594387,"spectral_convergence":0.37134683725884937,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.881358775516516,"spectral_convergence":0.3951471174512605,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":6.916324049439895e-05,"delta_db":-0.3439710041799785,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.000480954275751447,"delta_db":0.8883418975390568,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.000495025256353086,"delta_db":1.0506861812411472,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00017132630539018218,"delta_db":0.6058781853466181,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":0.00011217067220077199,"delta_db":0.6027507099284077,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.00027512664495995295,"delta_db":0.36337475007910797,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6425004547190528,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7313949649743178,"pitch":null,"output_rms_dbfs":-27.948588908435408,"output_peak":0.3816178104403332,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.856820326550241},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.904938059369171,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03681296571269971,"rms_dbfs":-28.67998387340972,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.2140993950677,"mean":-0.0002700039545694987,"zero_energy":false},"output":{"rms":0.040047052265426965,"rms_dbfs":-27.948588908435408,"peak":0.3816178104403332,"peak_dbfs":-8.367427291756382,"crest_factor_db":19.581161616679026,"mean":3.289144813627752e-06,"zero_energy":false},"rms_difference_db":0.7313949649743137},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":286,"output_estimated_voiced_frames":313,"both_estimated_voiced_frames":278,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.972027972027972,"absolute_difference_cents_median":5.787956929092985,"absolute_difference_cents_p95":20.345873775549705,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":1.7565922267653888,"rms_envelope_nrmse":0.15482796979253435,"envelope_correlation":0.9807812720376159,"rms_gain_error_db":0.7347374876782248,"reference_mean_power":0.0013552494150092414,"output_mean_power":0.0016050663032254008},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.097356675930481e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024105831474109917,"output_power":0.00031479523972110455,"reference_active":true,"delta_db":1.1590604036423915},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7795131277493678e-05,"output_power":3.628422773615306e-05,"reference_active":true,"delta_db":1.1574915453848036},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0822198872325408e-05,"output_power":1.5234781950105205e-05,"reference_active":true,"delta_db":1.4852073229444518}],"active_band_count":3,"active_band_mae_db":1.2672530906572155}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":1.4689918624265115,"rms_envelope_nrmse":0.11881910348528504,"envelope_correlation":0.9918677671561237,"rms_gain_error_db":0.7338336804187402,"reference_mean_power":0.0013563897957452433,"output_mean_power":0.00160608261797854},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.582417446963071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021668841148527672,"output_power":0.0002849683767614331,"reference_active":true,"delta_db":1.1896098273494629},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1881573874210085e-05,"output_power":1.6086719297884518e-05,"reference_active":true,"delta_db":1.3159351114572444},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.720949484930855e-07,"output_power":8.660774236858833e-07,"reference_active":true,"delta_db":1.1012608662727454}],"active_band_count":3,"active_band_mae_db":1.202268601693151}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":1.884768157467068,"rms_envelope_nrmse":0.19692179157949985,"envelope_correlation":0.9856212596104965,"rms_gain_error_db":1.1489695715328117,"reference_mean_power":0.0005726237314993749,"output_mean_power":0.0007460471916131646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0939920051603367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018116410089920082,"output_power":0.00025439540119028404,"reference_active":true,"delta_db":1.4743711315179728},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.90482495868649e-05,"output_power":4.266610876926217e-05,"reference_active":true,"delta_db":1.6696306873444429},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.620984223411991e-06,"output_power":7.663766310436298e-06,"reference_active":true,"delta_db":1.3462988707789023}],"active_band_count":3,"active_band_mae_db":1.496766896547106}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":1.4796252515666253,"rms_envelope_nrmse":0.17619886937709603,"envelope_correlation":0.9907827678436728,"rms_gain_error_db":1.1498464485565643,"reference_mean_power":0.0005714831913152958,"output_mean_power":0.000744711578485877},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7768006225472703e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016765591938032763,"output_power":0.0002378643913976443,"reference_active":true,"delta_db":1.5191054091756597},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3133776830395815e-05,"output_power":2.003266120209834e-05,"reference_active":true,"delta_db":1.8334901371213286},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7638611347445743e-07,"output_power":5.41288825630195e-07,"reference_active":true,"delta_db":0.5546996817637091}],"active_band_count":3,"active_band_mae_db":1.302431742686899}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":1.2482720343456846,"rms_envelope_nrmse":0.13422354971329323,"envelope_correlation":0.9918775845055839,"rms_gain_error_db":0.6823493895028593,"reference_mean_power":0.00020459070686192212,"output_mean_power":0.00023939817855672373},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3670238086467602e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.952541720107037e-05,"output_power":8.170779119856548e-05,"reference_active":true,"delta_db":0.7011986678857232},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.449912227458813e-06,"output_power":8.991652027141976e-06,"reference_active":true,"delta_db":0.8168833549421526},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5578562674598195e-06,"output_power":1.39921965201523e-06,"reference_active":true,"delta_db":-0.46641489688137916}],"active_band_count":3,"active_band_mae_db":0.6614989732364184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":0.8641431054624718,"rms_envelope_nrmse":0.10079348384615067,"envelope_correlation":0.9973474245273048,"rms_gain_error_db":0.681602875819043,"reference_mean_power":0.00020499017149025155,"output_mean_power":0.0002398243777641333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2857171527440002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.293887842389662e-05,"output_power":7.391071034732251e-05,"reference_active":true,"delta_db":0.6978837677084064},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7135208044031714e-06,"output_power":4.4999436143920246e-06,"reference_active":true,"delta_db":0.8342121064928449},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.333412649997229e-07,"output_power":1.1737093743062538e-07,"reference_active":true,"delta_db":-0.5540399766630768}],"active_band_count":3,"active_band_mae_db":0.695378616954776}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":1.3227490773425425,"rms_envelope_nrmse":0.22258639160804003,"envelope_correlation":0.9679180575264847,"rms_gain_error_db":0.2673091179088372,"reference_mean_power":0.00010545678906217639,"output_mean_power":0.00011215159531267737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.56955140861397e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.404725630082888e-05,"output_power":3.3963563926693685e-05,"reference_active":true,"delta_db":-0.010688638804738596},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.550080432854382e-06,"output_power":7.166078916550883e-06,"reference_active":true,"delta_db":-0.22669991768339895},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.516411722896476e-06,"output_power":4.241190385572334e-06,"reference_active":true,"delta_db":0.8138804900536772}],"active_band_count":3,"active_band_mae_db":0.3504230155139383}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":0.7710911383955362,"rms_envelope_nrmse":0.12186581966524603,"envelope_correlation":0.990326591161927,"rms_gain_error_db":0.26722240156592025,"reference_mean_power":0.0001059674543662488,"output_mean_power":0.00011269242946013814},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.064372154195949e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.39238699115062e-05,"output_power":3.418449930008531e-05,"reference_active":true,"delta_db":0.03323834033814247},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.204493449095075e-06,"output_power":3.887382044918651e-06,"reference_active":true,"delta_db":-0.340564557446256},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.457795181461805e-07,"output_power":2.54079247081188e-07,"reference_active":true,"delta_db":0.14423505344319967}],"active_band_count":3,"active_band_mae_db":0.17267931707586606}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"facodec","status":"measured","mos":{"status":"measured","value":4.438922882080078,"predicted_mos":4.438922882080078,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d85291701da9608ca390f9481a75e13b53d3447c81c972611bc03a9aee0899bc","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4361448400013614},"started_at":"2026-10-13T04:45:52.646151+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4152000150934327e-05,"native_input_samples":96000,"native_input_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","output_sha256":"b012c68866f88b162628b3d376767098e45ceb638892011960532f95aa626455","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5871254490002684,"synthesis_s":5.57892055100001,"total_s":7.166046000000279,"rtf":1.1943410000000465,"output_peak":0.2950056195259094,"output_rms":0.033880105172734855,"finished_at":"2026-10-13T04:45:59.817667+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"b012c68866f88b162628b3d376767098e45ceb638892011960532f95aa626455","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.301296058335135,"spectral_convergence":0.3717681301525394,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.970833530654455,"spectral_convergence":0.39105069483416877,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.236076549113238,"spectral_convergence":0.39701006820855655,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":7.108336918052084e-05,"delta_db":-0.2250442929373349,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.0003756729467607299,"delta_db":-0.18459678820643471,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.000339723242126572,"delta_db":-0.5843347912998822,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00013372722475715092,"delta_db":-0.47016399215742277,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":7.512788699077006e-05,"delta_db":-1.1380307720531018,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.00015252685669983124,"delta_db":-2.1984885683682034,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.80010986750373,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7211211300948718,"pitch":null,"output_rms_dbfs":-29.401105003504597,"output_peak":0.2950056195259094,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.836068712700943},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.301296058335135,"spectral_convergence":0.3717681301525394,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.970833530654455,"spectral_convergence":0.39105069483416877,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.236076549113238,"spectral_convergence":0.39701006820855655,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":7.108336918052084e-05,"delta_db":-0.2250442929373349,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.0003756729467607299,"delta_db":-0.18459678820643471,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.000339723242126572,"delta_db":-0.5843347912998822,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00013372722475715092,"delta_db":-0.47016399215742277,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":7.512788699077006e-05,"delta_db":-1.1380307720531018,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.00015252685669983124,"delta_db":-2.1984885683682034,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.80010986750373,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7211211300948718,"pitch":null,"output_rms_dbfs":-29.401105003504597,"output_peak":0.2950056195259094,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.836068712700943},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8954344110111301,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03681296571269971,"rms_dbfs":-28.67998387340972,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.2140993950677,"mean":-0.0002700039545694987,"zero_energy":false},"output":{"rms":0.033880105172734855,"rms_dbfs":-29.401105003504597,"peak":0.2950056195259094,"peak_dbfs":-10.60339422241311,"crest_factor_db":18.797710781091485,"mean":0.00017156545690356022,"zero_energy":false},"rms_difference_db":-0.7211211300948754},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":286,"output_estimated_voiced_frames":295,"both_estimated_voiced_frames":279,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9755244755244755,"absolute_difference_cents_median":3.398102919663024,"absolute_difference_cents_p95":13.912537687128442,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":1.2096065113988266,"rms_envelope_nrmse":0.1732849014670695,"envelope_correlation":0.9650144382686104,"rms_gain_error_db":-0.7218956547333286,"reference_mean_power":0.0013552494150092414,"output_mean_power":0.0011477033862677475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.097356675930481e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024105831474109917,"output_power":0.00017872453162165688,"reference_active":true,"delta_db":-1.2993794845044038},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7795131277493678e-05,"output_power":2.6119800400055664e-05,"reference_active":true,"delta_db":-0.2699887570611294},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0822198872325408e-05,"output_power":1.0825441867961822e-05,"reference_active":true,"delta_db":0.0013012181208698983}],"active_band_count":3,"active_band_mae_db":0.5235564865621344}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":1.0617248455404575,"rms_envelope_nrmse":0.15592309026364184,"envelope_correlation":0.9711710426413617,"rms_gain_error_db":-0.730253246903695,"reference_mean_power":0.0013563897957452433,"output_mean_power":0.0011464607454747386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.582417446963071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021668841148527672,"output_power":0.0001592226130104301,"reference_active":true,"delta_db":-1.3383093896443723},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1881573874210085e-05,"output_power":1.1148268783834664e-05,"reference_active":true,"delta_db":-0.2766654162762472},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.720949484930855e-07,"output_power":5.595640048521341e-07,"reference_active":true,"delta_db":-0.7958086142941102}],"active_band_count":3,"active_band_mae_db":0.8035944734049099}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":1.7697732480000237,"rms_envelope_nrmse":0.12166728094863118,"envelope_correlation":0.9878842281725395,"rms_gain_error_db":-0.34315694505411837,"reference_mean_power":0.0005726237314993749,"output_mean_power":0.0005291193550608567},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0939920051603367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018116410089920082,"output_power":0.00016283109854548728,"reference_active":true,"delta_db":-0.46334790108311424},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.90482495868649e-05,"output_power":3.068792411650753e-05,"reference_active":true,"delta_db":0.23847544147824223},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.620984223411991e-06,"output_power":5.562610929145375e-06,"reference_active":true,"delta_db":-0.045336816056571624}],"active_band_count":3,"active_band_mae_db":0.24905338620597603}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":1.4268672706040193,"rms_envelope_nrmse":0.10975655366261243,"envelope_correlation":0.9896746946827675,"rms_gain_error_db":-0.3556586855136931,"reference_mean_power":0.0005714831913152958,"output_mean_power":0.0005265475456025078},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7768006225472703e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016765591938032763,"output_power":0.0001508176760804427,"reference_active":true,"delta_db":-0.45966647001979993},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3133776830395815e-05,"output_power":1.4006015597539418e-05,"reference_active":true,"delta_db":0.27924973189284236},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7638611347445743e-07,"output_power":4.782593959022026e-07,"reference_active":true,"delta_db":0.01704417519674708}],"active_band_count":3,"active_band_mae_db":0.25198679236979643}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":1.7804459893581466,"rms_envelope_nrmse":0.16660623348985903,"envelope_correlation":0.9823437028381132,"rms_gain_error_db":-0.7558226011274667,"reference_mean_power":0.00020459070686192212,"output_mean_power":0.0001719109905474016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3670238086467602e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.952541720107037e-05,"output_power":4.998499346264836e-05,"reference_active":true,"delta_db":-1.4330396396975713},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.449912227458813e-06,"output_power":7.714975669416489e-06,"reference_active":true,"delta_db":0.15183404718066895},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5578562674598195e-06,"output_power":1.6275649580619472e-06,"reference_active":true,"delta_db":0.19010944937863286}],"active_band_count":3,"active_band_mae_db":0.5916610454189577}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":1.4873260700691984,"rms_envelope_nrmse":0.15081342522841604,"envelope_correlation":0.9858707312477577,"rms_gain_error_db":-0.7611493802921833,"reference_mean_power":0.00020499017149025155,"output_mean_power":0.0001720355105824413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2857171527440002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.293887842389662e-05,"output_power":4.492275863044712e-05,"reference_active":true,"delta_db":-1.4645258178582954},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7135208044031714e-06,"output_power":3.959432017574172e-06,"reference_active":true,"delta_db":0.2784702931796638},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.333412649997229e-07,"output_power":1.3895281009303607e-07,"reference_active":true,"delta_db":0.1790276316491329}],"active_band_count":3,"active_band_mae_db":0.6406745808956974}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":1.7460317845539408,"rms_envelope_nrmse":0.27285333035187065,"envelope_correlation":0.9556204294209654,"rms_gain_error_db":-1.2606112422091467,"reference_mean_power":0.00010545678906217639,"output_mean_power":7.888844936295905e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.56955140861397e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.404725630082888e-05,"output_power":2.6853694100896085e-05,"reference_active":true,"delta_db":-1.030780826046726},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.550080432854382e-06,"output_power":1.011282688721847e-05,"reference_active":true,"delta_db":1.269209946915115},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.516411722896476e-06,"output_power":3.4768466367648886e-06,"reference_active":true,"delta_db":-0.04914184786436054}],"active_band_count":3,"active_band_mae_db":0.7830442069420672}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":1.3855066255099673,"rms_envelope_nrmse":0.2325674362640685,"envelope_correlation":0.9704910759127441,"rms_gain_error_db":-1.2608897642300456,"reference_mean_power":0.0001059674543662488,"output_mean_power":7.926537618378752e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.064372154195949e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.39238699115062e-05,"output_power":2.754052803573062e-05,"reference_active":true,"delta_db":-0.905331263956808},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.204493449095075e-06,"output_power":5.6831193187199855e-06,"reference_active":true,"delta_db":1.3087309412337456},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.457795181461805e-07,"output_power":2.7172955192478276e-07,"reference_active":true,"delta_db":0.4359118419674184}],"active_band_count":3,"active_band_mae_db":0.8833246823859907}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.8319091796875,"predicted_mos":2.8319091796875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"525614d712f36be0b3a78aa6f59a74c3f6877ceec07837e5b9b4970e16be72d6","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3443503899998177},"started_at":"2026-10-13T04:59:01.049775+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006086358999709773,"native_input_samples":264600,"native_input_sha256":"c489532580bc70b378f86607c04dcaccad0e1e895a37fd4119f49a816b1b19e1","output_sha256":"743a083e28abbf467381c7941876bb88b41c441421aa74c0a2171b7e6488f3bb","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13609293499939668,"synthesis_s":3.87054064700078,"total_s":4.006633582000177,"rtf":0.6677722636666962,"output_peak":0.26591357588768005,"output_rms":0.03290421442576439,"finished_at":"2026-10-13T04:59:05.074034+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"743a083e28abbf467381c7941876bb88b41c441421aa74c0a2171b7e6488f3bb","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.6983345163086,"output_peak":0.26445024394677813},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.199333745952153,"spectral_convergence":0.4919374791200561,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.215957488576977,"spectral_convergence":0.4942716149566405,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.978840048099597,"spectral_convergence":0.5238007196353189,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":6.715661726120614e-05,"delta_db":-0.4718362121145234,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.0003038113376758188,"delta_db":-1.1066562337925079,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.0003431297010840949,"delta_db":-0.5410042729194798,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00010695622437709385,"delta_db":-1.440301673441712,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":0.00010037766061886215,"delta_db":0.12032818417717368,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.0001505561801685567,"delta_db":-2.2549658963881254,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9891820788055871,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0181180179095604,"pitch":null,"output_rms_dbfs":-29.698101891319286,"output_peak":0.26445024394677813,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.131377094209576},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7690149778330376,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03681296571269971,"rms_dbfs":-28.67998387340972,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.2140993950677,"mean":-0.0002700039545694987,"zero_energy":false},"output":{"rms":0.03274122357496177,"rms_dbfs":-29.698101891319283,"peak":0.26445024394677813,"peak_dbfs":-11.553120560444594,"crest_factor_db":18.144981330874693,"mean":0.0001820447068561663,"zero_energy":false},"rms_difference_db":-1.0181180179095612},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":286,"output_estimated_voiced_frames":301,"both_estimated_voiced_frames":262,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.916083916083916,"absolute_difference_cents_median":8.674336504697674,"absolute_difference_cents_p95":37.584988984382676,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":1.632475321480015,"rms_envelope_nrmse":0.14551616257551905,"envelope_correlation":0.9870618997936845,"rms_gain_error_db":-1.0170102126529215,"reference_mean_power":0.0013552494150092414,"output_mean_power":0.0010723046942808917},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.097356675930481e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024105831474109917,"output_power":0.000201511216701059,"reference_active":true,"delta_db":-0.7782289080143318},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7795131277493678e-05,"output_power":2.5253166661580173e-05,"reference_active":true,"delta_db":-0.41652884642510746},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0822198872325408e-05,"output_power":9.889586417956994e-06,"reference_active":true,"delta_db":-0.39137380607766087}],"active_band_count":3,"active_band_mae_db":0.5287105201723666}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":1.411043498587546,"rms_envelope_nrmse":0.12916632911427972,"envelope_correlation":0.9930093747489807,"rms_gain_error_db":-1.0165137766824979,"reference_mean_power":0.0013563897957452433,"output_mean_power":0.0010733296737121005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.582417446963071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021668841148527672,"output_power":0.00018123453489940044,"reference_active":true,"delta_db":-0.7759472821621867},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1881573874210085e-05,"output_power":1.1271258316675034e-05,"reference_active":true,"delta_db":-0.22901569464801247},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.720949484930855e-07,"output_power":6.104751084266601e-07,"reference_active":true,"delta_db":-0.4176267055393217}],"active_band_count":3,"active_band_mae_db":0.47419656078317357}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":2.7944114179443904,"rms_envelope_nrmse":0.19767214300562921,"envelope_correlation":0.9756898436783695,"rms_gain_error_db":-1.1206500140905737,"reference_mean_power":0.0005726237314993749,"output_mean_power":0.00044238902201008745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0939920051603367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018116410089920082,"output_power":0.00013191758854663708,"reference_active":true,"delta_db":-1.3776943917015638},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.90482495868649e-05,"output_power":2.019143301734876e-05,"reference_active":true,"delta_db":-1.579528248523273},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.620984223411991e-06,"output_power":5.392903042571435e-06,"reference_active":true,"delta_db":-0.179897540423527}],"active_band_count":3,"active_band_mae_db":1.0457067268827878}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":2.221096207179476,"rms_envelope_nrmse":0.18040088510142238,"envelope_correlation":0.9805000653588749,"rms_gain_error_db":-1.1160802224275446,"reference_mean_power":0.0005714831913152958,"output_mean_power":0.00044197269449369807},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7768006225472703e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016765591938032763,"output_power":0.00012116949054070333,"reference_active":true,"delta_db":-1.4102560965394353},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3133776830395815e-05,"output_power":9.919773520585705e-06,"reference_active":true,"delta_db":-1.2188787562352796},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7638611347445743e-07,"output_power":3.376034124964546e-07,"reference_active":true,"delta_db":-1.4955226560338821}],"active_band_count":3,"active_band_mae_db":1.3748858362695324}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":3.260722715325449,"rms_envelope_nrmse":0.22393512506865776,"envelope_correlation":0.9704676664690687,"rms_gain_error_db":-1.2152815075683923,"reference_mean_power":0.00020459070686192212,"output_mean_power":0.0001546527874422612},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3670238086467602e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.952541720107037e-05,"output_power":5.776718926675567e-05,"reference_active":true,"delta_db":-0.8046236668797759},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.449912227458813e-06,"output_power":5.8337091316063756e-06,"reference_active":true,"delta_db":-1.062063846281217},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5578562674598195e-06,"output_power":1.477452912899446e-06,"reference_active":true,"delta_db":-0.23013737249960917}],"active_band_count":3,"active_band_mae_db":0.698941628553534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":2.5620824734113254,"rms_envelope_nrmse":0.1978150118002789,"envelope_correlation":0.9793021864210012,"rms_gain_error_db":-1.2256467507064879,"reference_mean_power":0.00020499017149025155,"output_mean_power":0.00015458536071325847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2857171527440002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.293887842389662e-05,"output_power":5.26149208124471e-05,"reference_active":true,"delta_db":-0.7781007837759041},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7135208044031714e-06,"output_power":2.841155525412075e-06,"reference_active":true,"delta_db":-1.1629085358446476},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.333412649997229e-07,"output_power":1.2665695110058575e-07,"reference_active":true,"delta_db":-0.22335541497189543}],"active_band_count":3,"active_band_mae_db":0.7214549115308156}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":4.7814124895263435,"rms_envelope_nrmse":0.30471393279400727,"envelope_correlation":0.9401880680599257,"rms_gain_error_db":-0.09735547210969567,"reference_mean_power":0.00010545678906217639,"output_mean_power":0.00010311907214632278},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.56955140861397e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.404725630082888e-05,"output_power":3.757604085017389e-05,"reference_active":true,"delta_db":0.42828899286532857},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.550080432854382e-06,"output_power":6.022639511890331e-06,"reference_active":true,"delta_db":-0.9816470926481212},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.516411722896476e-06,"output_power":2.670493220967718e-06,"reference_active":true,"delta_db":-1.195082393688835}],"active_band_count":3,"active_band_mae_db":0.8683394930674283}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":4.150222175443645,"rms_envelope_nrmse":0.25066181564624435,"envelope_correlation":0.9585371027446307,"rms_gain_error_db":-0.09667369936655303,"reference_mean_power":0.0001059674543662488,"output_mean_power":0.00010363468496803371},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.064372154195949e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.39238699115062e-05,"output_power":3.7070496266652456e-05,"reference_active":true,"delta_db":0.3852301087139657},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.204493449095075e-06,"output_power":3.1359807845149375e-06,"reference_active":true,"delta_db":-1.2734028724031274},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.457795181461805e-07,"output_power":2.453479558036628e-07,"reference_active":true,"delta_db":-0.00763244583672391}],"active_band_count":3,"active_band_mae_db":0.5554218089846057}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"golf","status":"measured","mos":{"status":"measured","value":2.4080522060394287,"predicted_mos":2.4080522060394287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"baa6543861121ddcb683b884e4c9a9551f50a400a24f24e603902caeb82d7926","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4489405839995015},"started_at":"2026-10-13T05:01:56.216330+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031266099995264085,"native_input_samples":144000,"native_input_sha256":"271919eb76999201ad30c895eb15bf8903c21cb2f666da4c79d25fb0d8c90c16","output_sha256":"9c9bfc35b2180e9c1c925ae1878397e33fdf91a4204a8c870981af4c50cd361c","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6973659169998427,"synthesis_s":0.12353904799965676,"total_s":0.8209049649994995,"rtf":0.13681749416658326,"output_peak":0.29197460412979126,"output_rms":0.03274332203286858,"finished_at":"2026-10-13T05:01:57.048500+00:00","analysis_stages":{"external_f0_s":0.08325544299987087},"evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"9c9bfc35b2180e9c1c925ae1878397e33fdf91a4204a8c870981af4c50cd361c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.72649169026195,"output_peak":0.2921757825950145},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.597396752792617,"spectral_convergence":0.551159638704406,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.9962121751161765,"spectral_convergence":0.428155845223638,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.183343734403063,"spectral_convergence":0.46046291544194207,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.474639540565938e-05,"output_power":6.651057987380632e-05,"delta_db":-0.5069951968823674,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003927510430938073,"output_power":0.00031592514403965075,"delta_db":-0.9453315497097993,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038615206379246763,"output_power":0.0002848282511229615,"delta_db":-1.3217529701109545,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014882250290112057,"output_power":0.00012038696125348571,"delta_db":-0.9208915188882445,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.775910074349564e-05,"output_power":7.179809495100065e-05,"delta_db":-1.3404427689546725,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00025346126744946613,"output_power":0.00020555396835700762,"delta_db":-0.9098573671334427,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9908785619465802,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.041691511595505,"pitch":null,"output_rms_dbfs":-29.72649169026195,"output_peak":0.2921757825950145,"reference_rms_dbfs":-28.684800178666446,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.258984220770619},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8447596577380994,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.036792558668649514,"rms_dbfs":-28.684800178666443,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.21891570032442,"mean":-0.0002751886944141929,"zero_energy":false},"output":{"rms":0.03263438370182456,"rms_dbfs":-29.72649169026195,"peak":0.2921757825950145,"peak_dbfs":-10.687115680506551,"crest_factor_db":19.039376009755397,"mean":0.00011962758291376124,"zero_energy":false},"rms_difference_db":-1.0416915115955057},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":285,"output_estimated_voiced_frames":229,"both_estimated_voiced_frames":218,"reference_voiced_output_unvoiced_frames":67,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.7649122807017544,"absolute_difference_cents_median":3.94725671135384,"absolute_difference_cents_p95":19.4944202028355,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":2.285549403766303,"rms_envelope_nrmse":0.21656606180405466,"envelope_correlation":0.9537067798243346,"rms_gain_error_db":-1.041450258309085,"reference_mean_power":0.0013540006042442557,"output_mean_power":0.0010653046736233637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.105742351130961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024121484017359198,"output_power":0.0002046770490347279,"reference_active":true,"delta_db":-0.7133487640023621},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.787210966603033e-05,"output_power":2.8433576110184312e-05,"reference_active":true,"delta_db":0.08661642524995644},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0802536140133606e-05,"output_power":1.0161849803896455e-05,"reference_active":true,"delta_db":-0.2655295634801262}],"active_band_count":3,"active_band_mae_db":0.3551649175774816}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":1.8537389805218256,"rms_envelope_nrmse":0.19083544907298836,"envelope_correlation":0.9645919929598294,"rms_gain_error_db":-1.0294891287609207,"reference_mean_power":0.0013547589084803848,"output_mean_power":0.0010688409948787633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.5909353834355345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021686492823923438,"output_power":0.0001870557012779734,"reference_active":true,"delta_db":-0.6421837324481682},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1910385336822864e-05,"output_power":1.2318174544139089e-05,"reference_active":true,"delta_db":0.1462054115167543},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.685844459259575e-07,"output_power":6.358396721316338e-07,"reference_active":true,"delta_db":-0.21808647378287624}],"active_band_count":3,"active_band_mae_db":0.33549187258259955}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":2.353124575450152,"rms_envelope_nrmse":0.27270100156720756,"envelope_correlation":0.9422658967603709,"rms_gain_error_db":-1.0928637854609147,"reference_mean_power":0.0005711286012215311,"output_mean_power":0.00044406600853038957},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.1023770927764777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018127888675696264,"output_power":0.00014328895215235172,"reference_active":true,"delta_db":-1.0213451873563193},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.911192606295216e-05,"output_power":2.8611766336736413e-05,"reference_active":true,"delta_db":-0.07526269938729174},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.6207711745544935e-06,"output_power":5.492768750530831e-06,"reference_active":true,"delta_db":-0.10004589927100281}],"active_band_count":3,"active_band_mae_db":0.3988845953382046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":1.840403400670218,"rms_envelope_nrmse":0.24661535595811085,"envelope_correlation":0.9508255437579711,"rms_gain_error_db":-1.0714730008408655,"reference_mean_power":0.0005695862008353661,"output_mean_power":0.0004450534368717013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.784725321923225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000167755673364387,"output_power":0.00013451339791107557,"reference_active":true,"delta_db":-0.9591167314651778},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3147914953507592e-05,"output_power":1.3653167445120349e-05,"reference_active":true,"delta_db":0.16376530320234212},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769978202454618e-07,"output_power":4.1804110362248407e-07,"reference_active":true,"delta_db":-0.5729740883320467}],"active_band_count":3,"active_band_mae_db":0.5652853743331888}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":1.7489574778895505,"rms_envelope_nrmse":0.191565606492501,"envelope_correlation":0.9800113598114742,"rms_gain_error_db":-1.0827624637195705,"reference_mean_power":0.00020451896284517183,"output_mean_power":0.00015938862910537874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.371370632972142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.95451626460554e-05,"output_power":5.485151961066806e-05,"reference_active":true,"delta_db":-1.0307826334594583},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.468229211749217e-06,"output_power":5.894831064189851e-06,"reference_active":true,"delta_db":-1.0274627519524937},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5571349492124254e-06,"output_power":1.3845222314742185e-06,"reference_active":true,"delta_db":-0.510263186946502}],"active_band_count":3,"active_band_mae_db":0.8561695241194847}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":1.3274094566114998,"rms_envelope_nrmse":0.16081358485458686,"envelope_correlation":0.9886266050255569,"rms_gain_error_db":-1.075462762677018,"reference_mean_power":0.00020492196972728493,"output_mean_power":0.00015997136311132046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2898363654312675e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.2932702572922e-05,"output_power":4.9631397431651574e-05,"reference_active":true,"delta_db":-1.0311987897504733},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.715382170208954e-06,"output_power":2.884537654050263e-06,"reference_active":true,"delta_db":-1.0992728012061401},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3331500145359745e-07,"output_power":1.2030343730665715e-07,"reference_active":true,"delta_db":-0.4460098557887262}],"active_band_count":3,"active_band_mae_db":0.8588271489151132}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":2.848254138649887,"rms_envelope_nrmse":0.19954168941075096,"envelope_correlation":0.9768139265351302,"rms_gain_error_db":0.0769042014365017,"reference_mean_power":0.00010562659730039624,"output_mean_power":0.00010751367563815983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.532678642637495e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.375936849338044e-05,"output_power":3.58006667683393e-05,"reference_active":true,"delta_db":0.2549680115526992},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.5579365171316724e-06,"output_power":8.143270535084245e-06,"reference_active":true,"delta_db":0.3239562338821488},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4875543569951454e-06,"output_power":4.229775725493919e-06,"reference_active":true,"delta_db":0.837963552359769}],"active_band_count":3,"active_band_mae_db":0.4722959325982057}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":2.5738255677857764,"rms_envelope_nrmse":0.13724286311466305,"envelope_correlation":0.9901360860737263,"rms_gain_error_db":0.07731481802149463,"reference_mean_power":0.00010613504910757197,"output_mean_power":0.00010804142583554211},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.028648465216762e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.364360550011598e-05,"output_power":3.6122558443704986e-05,"reference_active":true,"delta_db":0.3087597053974651},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.191080260709659e-06,"output_power":4.698786613122632e-06,"reference_active":true,"delta_db":0.4965974496345378},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4483088580077784e-07,"output_power":3.274415261688407e-07,"reference_active":true,"delta_db":1.2626755198007054}],"active_band_count":3,"active_band_mae_db":0.6893442249442362}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.8073618412017822,"predicted_mos":2.8073618412017822,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"84f622d035d2f13d256429582e013a79d7adb09e8da9e12a56cc50df495c9a93","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0270945619995473},"started_at":"2026-10-13T05:04:47.697576+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030157489991324837,"native_input_samples":144000,"native_input_sha256":"271919eb76999201ad30c895eb15bf8903c21cb2f666da4c79d25fb0d8c90c16","output_sha256":"6fdf0ff8b91678b3ae506d84921f999ed1d4ab64c09d0a7933958bd349242ef4","output_samples":144120,"length_delta_samples":120,"analysis_s":2.2722238550004477,"synthesis_s":3.7058164549998764,"total_s":5.978040310000324,"rtf":0.9963400516667207,"output_peak":0.3385598361492157,"output_rms":0.03506708168370972,"finished_at":"2026-10-13T05:04:53.689579+00:00","analysis_stages":{"external_f0_s":1.409263753000232},"evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"6fdf0ff8b91678b3ae506d84921f999ed1d4ab64c09d0a7933958bd349242ef4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.130623562572573,"output_peak":0.33994052931973323},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.971727720000297,"spectral_convergence":0.5661108162245608,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.421970619227492,"spectral_convergence":0.4024281478809121,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.456907367596729,"spectral_convergence":0.4016792619348325,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":6.396012414773523e-05,"delta_db":-0.683631353791374,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.0003812441427847132,"delta_db":-0.12066420117290788,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.0003683161722157887,"delta_db":-0.233379507246187,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.00014514213891577847,"delta_db":-0.11442712137573532,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":9.757218894501849e-05,"delta_db":-0.002782033597617352,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.0001642320502501704,"delta_db":-1.877372625835718,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5053761405032566,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.454756365888423,"pitch":null,"output_rms_dbfs":-29.134740239298146,"output_peak":0.33994052931973323,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.616868568941506},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8735052004465188,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03681296571269971,"rms_dbfs":-28.67998387340972,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.2140993950677,"mean":-0.0002700039545694987,"zero_energy":false},"output":{"rms":0.03493518022365427,"rms_dbfs":-29.134740239298146,"peak":0.33994052931973323,"peak_dbfs":-9.371941073705262,"crest_factor_db":19.762799165592885,"mean":8.822946211233117e-05,"zero_energy":false},"rms_difference_db":-0.45475636588842505},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":286,"output_estimated_voiced_frames":314,"both_estimated_voiced_frames":274,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":40,"paired_voiced_fraction":0.958041958041958,"absolute_difference_cents_median":5.870210915178076,"absolute_difference_cents_p95":24.76301574927334,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":2.276335320575624,"rms_envelope_nrmse":0.16600398375260245,"envelope_correlation":0.9654665957440417,"rms_gain_error_db":-0.45196298591715245,"reference_mean_power":0.0013552494150092414,"output_mean_power":0.0012213016107976466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.097356675930481e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024105831474109917,"output_power":0.0002363524418377715,"reference_active":true,"delta_db":-0.085620225400999},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7795131277493678e-05,"output_power":2.6042134324180264e-05,"reference_active":true,"delta_db":-0.28292154925450036},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0822198872325408e-05,"output_power":9.532365642669579e-06,"reference_active":true,"delta_db":-0.5511481773947916}],"active_band_count":3,"active_band_mae_db":0.306563317350097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":1.7872186432028034,"rms_envelope_nrmse":0.12032683509183542,"envelope_correlation":0.9819451375250203,"rms_gain_error_db":-0.44914238233499726,"reference_mean_power":0.0013563897957452433,"output_mean_power":0.0012231234021237187},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.582417446963071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021668841148527672,"output_power":0.00021058939036925432,"reference_active":true,"delta_db":-0.12399198473388454},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1881573874210085e-05,"output_power":1.1506822535201986e-05,"reference_active":true,"delta_db":-0.13918557391410746},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.720949484930855e-07,"output_power":6.193601359819877e-07,"reference_active":true,"delta_db":-0.3548738190377024}],"active_band_count":3,"active_band_mae_db":0.20601712589523147}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":1.9911698256924792,"rms_envelope_nrmse":0.172044280200804,"envelope_correlation":0.973827375456393,"rms_gain_error_db":-0.1651723808175593,"reference_mean_power":0.0005726237314993749,"output_mean_power":0.0005512544453094846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0939920051603367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018116410089920082,"output_power":0.0001789337598033296,"reference_active":true,"delta_db":-0.05379855438352134},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.90482495868649e-05,"output_power":2.9243475170382677e-05,"reference_active":true,"delta_db":0.029090136208340965},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.620984223411991e-06,"output_power":5.930080690055823e-06,"reference_active":true,"delta_db":0.23248236461978605}],"active_band_count":3,"active_band_mae_db":0.10512368507054944}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":1.3013764993227779,"rms_envelope_nrmse":0.11192159420250192,"envelope_correlation":0.9880465793304398,"rms_gain_error_db":-0.15744960966165955,"reference_mean_power":0.0005714831913152958,"output_mean_power":0.0005511356451879851},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7768006225472703e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016765591938032763,"output_power":0.00016522846376267319,"reference_active":true,"delta_db":-0.06334026589333591},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3133776830395815e-05,"output_power":1.3409719818944027e-05,"reference_active":true,"delta_db":0.09030071383326702},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7638611347445743e-07,"output_power":4.965186879555406e-07,"reference_active":true,"delta_db":0.17976505672445903}],"active_band_count":3,"active_band_mae_db":0.11113534548368731}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":1.7806148473146413,"rms_envelope_nrmse":0.15648746533924954,"envelope_correlation":0.9810643355057536,"rms_gain_error_db":-0.004857789095675196,"reference_mean_power":0.00020459070686192212,"output_mean_power":0.00020436199042305744},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3670238086467602e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.952541720107037e-05,"output_power":7.317836137074234e-05,"reference_active":true,"delta_db":0.22239076808069863},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.449912227458813e-06,"output_power":7.339390233892313e-06,"reference_active":true,"delta_db":-0.06491176393188103},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5578562674598195e-06,"output_power":1.370339412450679e-06,"reference_active":true,"delta_db":-0.5569923717216623}],"active_band_count":3,"active_band_mae_db":0.2814316345780807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":1.1821388898539278,"rms_envelope_nrmse":0.09877466327375524,"envelope_correlation":0.9920131380948615,"rms_gain_error_db":-0.0023264053887040647,"reference_mean_power":0.00020499017149025155,"output_mean_power":0.00020488039286002398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2857171527440002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.293887842389662e-05,"output_power":6.625747057439655e-05,"reference_active":true,"delta_db":0.22315852882697496},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7135208044031714e-06,"output_power":3.565596650192933e-06,"reference_active":true,"delta_db":-0.17653648256139864},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.333412649997229e-07,"output_power":1.2044898251258067e-07,"reference_active":true,"delta_db":-0.4416143526543803}],"active_band_count":3,"active_band_mae_db":0.280436454680918}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":1.8200005586040444,"rms_envelope_nrmse":0.24123617887715562,"envelope_correlation":0.960426138761413,"rms_gain_error_db":-0.5374606963616028,"reference_mean_power":0.00010545678906217639,"output_mean_power":9.318123758075784e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.56955140861397e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.404725630082888e-05,"output_power":3.0320840955227466e-05,"reference_active":true,"delta_db":-0.5034087764447823},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.550080432854382e-06,"output_power":6.461077944904983e-06,"reference_active":true,"delta_db":-0.6764659800229083},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.516411722896476e-06,"output_power":3.2120476421300786e-06,"reference_active":true,"delta_db":-0.3931774097058352}],"active_band_count":3,"active_band_mae_db":0.524350722057842}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":1.1877637790401832,"rms_envelope_nrmse":0.14345320363819083,"envelope_correlation":0.9870103936576322,"rms_gain_error_db":-0.5374344533754453,"reference_mean_power":0.0001059674543662488,"output_mean_power":9.363302538918677e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.064372154195949e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.39238699115062e-05,"output_power":3.0147128273397846e-05,"reference_active":true,"delta_db":-0.5125944030937843},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.204493449095075e-06,"output_power":3.5673184664993e-06,"reference_active":true,"delta_db":-0.7137179815836414},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.457795181461805e-07,"output_power":2.157227775349774e-07,"reference_active":true,"delta_db":-0.5664968504648521}],"active_band_count":3,"active_band_mae_db":0.5976030783807592}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.3513519763946533,"predicted_mos":2.3513519763946533,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"61f323b12b7d8dd24390df2bc518ca7ba2ed2bf916a197c8757c66449020f2de","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.167146051999225},"started_at":"2026-10-13T05:10:02.257850+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002750686999206664,"native_input_samples":144000,"native_input_sha256":"271919eb76999201ad30c895eb15bf8903c21cb2f666da4c79d25fb0d8c90c16","output_sha256":"94bc40bbd1bd5395ccabf90c96d1c95faaf2a88729280403626b3d6ae57dc2ac","output_samples":144240,"length_delta_samples":240,"analysis_s":0.6945469450001838,"synthesis_s":0.1319660320004914,"total_s":0.8265129770006752,"rtf":0.13775216283344585,"output_peak":0.33567357063293457,"output_rms":0.03535875338903191,"finished_at":"2026-10-13T05:10:03.096099+00:00","analysis_stages":{"external_f0_s":0.16929813599927002},"evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"94bc40bbd1bd5395ccabf90c96d1c95faaf2a88729280403626b3d6ae57dc2ac","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.065093267607914,"output_peak":0.33097824575828455},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.865497845525453,"spectral_convergence":0.343642906474974,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.6465315294799,"spectral_convergence":0.32425992514643104,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.377144993304587,"spectral_convergence":0.4048787699648719,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486389753340235e-05,"output_power":8.022814847762487e-05,"delta_db":0.30054336149733674,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003919851371758182,"output_power":0.00037408663220466376,"delta_db":-0.20297411089969397,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038865006012116416,"output_power":0.0003546986755129882,"delta_db":-0.3969917245575776,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001490171424082218,"output_power":0.0001386885401303471,"delta_db":-0.31195654305239395,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.76347124356667e-05,"output_power":9.133501114589839e-05,"delta_db":-0.28966965359759045,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002530434948901305,"output_power":0.00020301269429719968,"delta_db":-0.9567198216179845,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.40980920253709624,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.37862637774975777,"pitch":null,"output_rms_dbfs":-29.058610251159482,"output_peak":0.33097824575828455,"reference_rms_dbfs":-28.67998387340972,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.96305812276998},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9048138500355459,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03681296571269971,"rms_dbfs":-28.67998387340972,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.2140993950677,"mean":-0.0002700039545694987,"zero_energy":false},"output":{"rms":0.035242725515611334,"rms_dbfs":-29.058610251159482,"peak":0.33097824575828455,"peak_dbfs":-9.604011004099808,"crest_factor_db":19.454599247059676,"mean":0.0001400435048283979,"zero_energy":false},"rms_difference_db":-0.37862637774976093},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":286,"output_estimated_voiced_frames":213,"both_estimated_voiced_frames":208,"reference_voiced_output_unvoiced_frames":78,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.7272727272727273,"absolute_difference_cents_median":0.8190073006312901,"absolute_difference_cents_p95":5.56361585373196,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":0.8951350671764996,"rms_envelope_nrmse":0.09904446883118267,"envelope_correlation":0.988813563756423,"rms_gain_error_db":-0.377894689986892,"reference_mean_power":0.0013552494150092414,"output_mean_power":0.0012423093660882218},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.097356675930481e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024105831474109917,"output_power":0.00023034498745400363,"reference_active":true,"delta_db":-0.197433500539229},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7795131277493678e-05,"output_power":2.472146886079425e-05,"reference_active":true,"delta_db":-0.508944581692462},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0822198872325408e-05,"output_power":8.959518944013817e-06,"reference_active":true,"delta_db":-0.8203081833340253}],"active_band_count":3,"active_band_mae_db":0.5088954218552387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":0.5747005532358673,"rms_envelope_nrmse":0.07397551254573223,"envelope_correlation":0.9943279960987046,"rms_gain_error_db":-0.3736575679026547,"reference_mean_power":0.0013563897957452433,"output_mean_power":0.001244568363257459},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.582417446963071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021668841148527672,"output_power":0.00020774288365314658,"reference_active":true,"delta_db":-0.18309530120566322},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1881573874210085e-05,"output_power":1.0809246080850215e-05,"reference_active":true,"delta_db":-0.4107856859434201},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.720949484930855e-07,"output_power":5.417320315491537e-07,"reference_active":true,"delta_db":-0.9364611590440914}],"active_band_count":3,"active_band_mae_db":0.5101140487310583}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":0.9995076644233747,"rms_envelope_nrmse":0.10457828943616092,"envelope_correlation":0.9910256868096264,"rms_gain_error_db":-0.2821844047128858,"reference_mean_power":0.0005726237314993749,"output_mean_power":0.0005366002890119044},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0939920051603367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018116410089920082,"output_power":0.00017302718930097604,"reference_active":true,"delta_db":-0.19957789931170686},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.90482495868649e-05,"output_power":2.6097329033018433e-05,"reference_active":true,"delta_db":-0.46523906295918893},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.620984223411991e-06,"output_power":4.45746729631243e-06,"reference_active":true,"delta_db":-1.0072420083352984}],"active_band_count":3,"active_band_mae_db":0.5573529902020647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":0.5252832976399968,"rms_envelope_nrmse":0.07404764193820265,"envelope_correlation":0.9955466774222558,"rms_gain_error_db":-0.27479438744182316,"reference_mean_power":0.0005714831913152958,"output_mean_power":0.000536443543424071},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7768006225472703e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016765591938032763,"output_power":0.0001600207438461798,"reference_active":true,"delta_db":-0.20242606680194716},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3133776830395815e-05,"output_power":1.1892133459931248e-05,"reference_active":true,"delta_db":-0.43129858019951406},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7638611347445743e-07,"output_power":3.3343110854035084e-07,"reference_active":true,"delta_db":-1.5495297713002578}],"active_band_count":3,"active_band_mae_db":0.7277514727672397}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":0.9071911764437487,"rms_envelope_nrmse":0.08506639589622492,"envelope_correlation":0.995244396248143,"rms_gain_error_db":-0.3130229348574101,"reference_mean_power":0.00020459070686192212,"output_mean_power":0.00019036346855622966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3670238086467602e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.952541720107037e-05,"output_power":6.688334785828484e-05,"reference_active":true,"delta_db":-0.16825599949030084},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.449912227458813e-06,"output_power":6.1551563907106e-06,"reference_active":true,"delta_db":-0.8291206405957969},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5578562674598195e-06,"output_power":1.2157368920724603e-06,"reference_active":true,"delta_db":-1.0768779010480773}],"active_band_count":3,"active_band_mae_db":0.6914181803780584}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":0.6042332053355877,"rms_envelope_nrmse":0.058724828017387015,"envelope_correlation":0.9981694224275058,"rms_gain_error_db":-0.30887627343813184,"reference_mean_power":0.00020499017149025155,"output_mean_power":0.00019091735608589382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2857171527440002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.293887842389662e-05,"output_power":6.055639066585964e-05,"reference_active":true,"delta_db":-0.16759017412259203},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7135208044031714e-06,"output_power":3.06004608747593e-06,"reference_active":true,"delta_db":-0.8405789386611564},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.333412649997229e-07,"output_power":1.0125376018830819e-07,"reference_active":true,"delta_db":-1.195534106629557}],"active_band_count":3,"active_band_mae_db":0.7345677398044351}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":0.8799792976922203,"rms_envelope_nrmse":0.10069186884967751,"envelope_correlation":0.9933242336948143,"rms_gain_error_db":-0.19955637225649223,"reference_mean_power":0.00010545678906217639,"output_mean_power":0.00010072073970371672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.56955140861397e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.404725630082888e-05,"output_power":3.244872113687744e-05,"reference_active":true,"delta_db":-0.2088453488682805},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.550080432854382e-06,"output_power":6.574736514532578e-06,"reference_active":true,"delta_db":-0.600732253317161},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.516411722896476e-06,"output_power":2.4972723286843393e-06,"reference_active":true,"delta_db":-1.4863381427647955}],"active_band_count":3,"active_band_mae_db":0.7653052483167456}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":0.5623271097409814,"rms_envelope_nrmse":0.052075925786765835,"envelope_correlation":0.9984140079880558,"rms_gain_error_db":-0.19950833890291078,"reference_mean_power":0.0001059674543662488,"output_mean_power":0.00010120959048263166},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.064372154195949e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.39238699115062e-05,"output_power":3.194778547679189e-05,"reference_active":true,"delta_db":-0.26064629576552883},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.204493449095075e-06,"output_power":3.475731751213689e-06,"reference_active":true,"delta_db":-0.82667428909039},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.457795181461805e-07,"output_power":1.5932061297947359e-07,"reference_active":true,"delta_db":-1.8827371977929852}],"active_band_count":3,"active_band_mae_db":0.990019260882968}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0006","method":"magphase","status":"measured","mos":{"status":"measured","value":3.6060221195220947,"predicted_mos":3.6060221195220947,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287098,"input_sample_rate":48000,"input_seconds":5.981208333333333,"input_float32_sha256":"8c28402c319d99fbaf6027a2b42624ebc09a4bbf767775c6bc3d1d961c7b26b9","internal_samples":95700,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.616344934000153},"started_at":"2026-10-13T05:12:26.358846+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010826949999682256,"native_input_samples":288000,"native_input_sha256":"0d4b1022cabacde27c054b0382a8f8afb71258c54ccd0777b16d89a664645af9","output_sha256":"216f6bc0710cded86481864d6930780b839a2767de1eed15b5dfb38d0fa8da42","output_samples":287098,"length_delta_samples":-902,"analysis_s":3.7521205830016697,"synthesis_s":1.1425416369984305,"total_s":4.8946622200001,"rtf":0.8157770366666833,"output_peak":0.3866835518401618,"output_rms":0.03584315507808574,"finished_at":"2026-10-13T05:12:31.276997+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"216f6bc0710cded86481864d6930780b839a2767de1eed15b5dfb38d0fa8da42","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95700,"length_delta_samples":-300,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.954671657558855,"output_peak":0.35929193636715107},"common_prefix":{"samples":95700,"sample_rate":16000,"reference_coverage":0.996875,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95700,"output_samples":95700,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.271797512008216,"spectral_convergence":0.540198826787416,"spectral_convergence_status":"ok","amplitude_floor":6.641091871383268e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.165307320477429,"spectral_convergence":0.3176215340131689,"spectral_convergence_status":"ok","amplitude_floor":5.299135910844449e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.528660916963287,"spectral_convergence":0.33338744104119455,"spectral_convergence_status":"ok","amplitude_floor":3.9942669914893874e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.486807841643471e-05,"output_power":6.0839912128400863e-05,"delta_db":-0.9010810842977662,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00039282961905705586,"output_power":0.00036190216859488127,"delta_db":-0.3561304032487433,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003845375394980986,"output_power":0.00036179696116402354,"delta_db":-0.26473828324165993,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014829789374165626,"output_power":0.00015383668702870356,"delta_db":0.1592493568572745,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.78583035074867e-05,"output_power":0.0001045419066479877,"delta_db":0.286927340824523,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00025383033420069135,"output_power":0.00022921628884265552,"delta_db":-0.4429804482869135,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4018511527928134,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26515088798972386,"pitch":null,"output_rms_dbfs":-28.954671657558855,"output_peak":0.35929193636715107,"reference_rms_dbfs":-28.689520769569135,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.65525524981631},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95700,"output_samples":95700,"intelligibility":{"estoi":{"value":0.9023396534238083,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03677256815102018,"rms_dbfs":-28.689520769569135,"peak":0.299713134765625,"peak_dbfs":-10.465884478342025,"crest_factor_db":18.22363629122711,"mean":-0.0002704217765273968,"zero_energy":false},"output":{"rms":0.03566698647778717,"rms_dbfs":-28.95467165755886,"peak":0.35929193636715107,"peak_dbfs":-8.891050592202468,"crest_factor_db":20.063621065356386,"mean":0.0007035643855498371,"zero_energy":false},"rms_difference_db":-0.26515088798972286},"full_scale":{"reference":{"sample_count":95700,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95700,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":20,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":284,"output_estimated_voiced_frames":272,"both_estimated_voiced_frames":264,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9295774647887324,"absolute_difference_cents_median":3.8610737119421215,"absolute_difference_cents_p95":13.267849677839287,"difference_over_600_cents_frames":0,"uncovered_tail_samples":20,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95700,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1195,"rms_floor":9.947854524209456e-05,"log_rms_mae_db":1.1548122469897744,"rms_envelope_nrmse":0.10554530760767009,"envelope_correlation":0.9864837954260419,"rms_gain_error_db":-0.2658277083528363,"reference_mean_power":0.001352403590652437,"output_mean_power":0.0012721067126791913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":4.1143911563999974e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002413691242433401,"output_power":0.0002323751531513612,"reference_active":true,"delta_db":-0.1649202568084249},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7960501424912344e-05,"output_power":3.0161764241159714e-05,"reference_active":true,"delta_db":0.3291178548071384},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0768585857738391e-05,"output_power":1.178950592306304e-05,"reference_active":true,"delta_db":0.3933692995318428}],"active_band_count":3,"active_band_mae_db":0.2958024703824687}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1189,"rms_floor":9.267346213866395e-05,"log_rms_mae_db":0.7895618110085724,"rms_envelope_nrmse":0.06654370136966892,"envelope_correlation":0.9949904817529132,"rms_gain_error_db":-0.26513121381340027,"reference_mean_power":0.0013529544392102489,"output_mean_power":0.001272828967496789},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":3.599628167914456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021704892277581365,"output_power":0.00020998811371242753,"reference_active":true,"delta_db":-0.14362922379959375},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1912309084844872e-05,"output_power":1.3345771374528395e-05,"reference_active":true,"delta_db":0.49347727320951024},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.830513680660678e-07,"output_power":7.551263117662855e-07,"reference_active":true,"delta_db":0.43566237662158175}],"active_band_count":3,"active_band_mae_db":0.3575896245435619}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1195,"rms_floor":8.299288573172057e-05,"log_rms_mae_db":1.4193898942213021,"rms_envelope_nrmse":0.07292800304948308,"envelope_correlation":0.9960431174744078,"rms_gain_error_db":-0.2690952512617844,"reference_mean_power":0.0005692812463557397,"output_mean_power":0.0005350783166340752},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":3.11101892035116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018139612694348758,"output_power":0.00016825655812373664,"reference_active":true,"delta_db":-0.32656009348618703},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9175599411708834e-05,"output_power":2.6936920512978292e-05,"reference_active":true,"delta_db":-0.346718426344403},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.614971921509107e-06,"output_power":5.791759715062825e-06,"reference_active":true,"delta_db":0.1346294697292395}],"active_band_count":3,"active_band_mae_db":0.26930266318660984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1189,"rms_floor":7.559252308295269e-05,"log_rms_mae_db":0.9073196105306106,"rms_envelope_nrmse":0.04984366657768977,"envelope_correlation":0.9984615849150582,"rms_gain_error_db":-0.26839211465658136,"reference_mean_power":0.0005675420554601409,"output_mean_power":0.0005335299911064642},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":2.792817878360976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016785734240671934,"output_power":0.00015576403319647708,"reference_active":true,"delta_db":-0.32473159045045596},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3154401140572944e-05,"output_power":1.220519670492276e-05,"reference_active":true,"delta_db":-0.3252629832798654},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.855829466501513e-07,"output_power":4.523723939723552e-07,"reference_active":true,"delta_db":-0.30767331962825856}],"active_band_count":3,"active_band_mae_db":0.3192226311195267}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1195,"rms_floor":5.686331455816738e-05,"log_rms_mae_db":1.0121180503790688,"rms_envelope_nrmse":0.08918232955550567,"envelope_correlation":0.9943608613243581,"rms_gain_error_db":0.17653090524715445,"reference_mean_power":0.00020442170627892135,"output_mean_power":0.00021290217563231048},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.375757245820818e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.95627140212973e-05,"output_power":7.197555577201332e-05,"reference_active":true,"delta_db":0.14808509184616936},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.487812831230777e-06,"output_power":8.106435372623719e-06,"reference_active":true,"delta_db":0.34474944658140405},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5543263671784844e-06,"output_power":1.5890984290032726e-06,"reference_active":true,"delta_db":0.09608583982930477}],"active_band_count":3,"active_band_mae_db":0.1963067927522927}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1189,"rms_floor":5.493590562585597e-05,"log_rms_mae_db":0.6118914670715323,"rms_envelope_nrmse":0.05380361894062111,"envelope_correlation":0.9980608524953004,"rms_gain_error_db":0.17706795533769465,"reference_mean_power":0.0002048172705805382,"output_mean_power":0.00021334053012065057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":1.2939961040270937e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.292651631493148e-05,"output_power":6.525275213415237e-05,"reference_active":true,"delta_db":0.15765144061697797},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.715048981795529e-06,"output_power":4.062055150111136e-06,"reference_active":true,"delta_db":0.3878127130544855},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3451008147553682e-07,"output_power":1.3563532976077838e-07,"reference_active":true,"delta_db":0.03617991844539266}],"active_band_count":3,"active_band_mae_db":0.1938813573722854}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1195,"rms_floor":6.417692344469313e-05,"log_rms_mae_db":1.2947376710691447,"rms_envelope_nrmse":0.2013021617282633,"envelope_correlation":0.9746047301628523,"rms_gain_error_db":0.35454369828918897,"reference_mean_power":0.00010578576034032043,"output_mean_power":0.00011478405992985076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":6.495939413418346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.347511768789387e-05,"output_power":3.600946723253109e-05,"reference_active":true,"delta_db":0.31694583414345756},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.566833168625048e-06,"output_power":8.877473915929547e-06,"reference_active":true,"delta_db":0.6937524604543566},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4560235701242222e-06,"output_power":4.210143736820371e-06,"reference_active":true,"delta_db":0.8572022745085033}],"active_band_count":3,"active_band_mae_db":0.6226335230354392}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":20,"metrics":{"frame_count":1189,"rms_floor":5.848412221971717e-05,"log_rms_mae_db":0.9571357939224292,"rms_envelope_nrmse":0.11113693351779391,"envelope_correlation":0.9927459205051198,"rms_gain_error_db":0.354397106722344,"reference_mean_power":0.00010630057050617279,"output_mean_power":0.00011533876743540862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":5.993000906219848e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.336645995319058e-05,"output_power":3.6105929964219036e-05,"reference_active":true,"delta_db":0.34268403196444225},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.172545558121855e-06,"output_power":4.8215399682317826e-06,"reference_active":true,"delta_db":0.6278468396280885},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.481380563081632e-07,"output_power":3.193090930250805e-07,"reference_active":true,"delta_db":1.0951791017091141}],"active_band_count":3,"active_band_mae_db":0.6885699911005482}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"hnm","status":"measured","mos":{"status":"measured","value":2.538820505142212,"predicted_mos":2.538820505142212,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60480,"input_sample_rate":16000,"input_seconds":3.78,"input_float32_sha256":"7a03a131119169ac12e9a4c00b5865b9bfddc1ae42df2b0aff3ce8ce31c0d6d3","internal_samples":60480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3914153200003057},"started_at":"2026-10-13T04:38:10.943469+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2739999874611385e-05,"native_input_samples":60480,"native_input_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","output_sha256":"1d55610e0395e1a51f98b08dcf18a3767faeae16e82982197096310023588cac","output_samples":60480,"length_delta_samples":0,"analysis_s":0.5998365329996886,"synthesis_s":0.32983473500007676,"total_s":0.9296712679997654,"rtf":0.24594477989411784,"output_peak":0.35237655929870776,"output_rms":0.04577188472888179,"finished_at":"2026-10-13T04:38:11.877220+00:00","evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"1d55610e0395e1a51f98b08dcf18a3767faeae16e82982197096310023588cac","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.003972272858582,"spectral_convergence":0.44741130528928885,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.8963798469816515,"spectral_convergence":0.4807609411413532,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.141328134030649,"spectral_convergence":0.4822377466362019,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":5.023392016209277e-05,"delta_db":0.36009968624633515,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0005414816689055985,"delta_db":-0.17749204619352704,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0007200792720005805,"delta_db":-0.36048552951089935,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.0002542121502934708,"delta_db":-0.4560888562176691,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00018400665922083933,"delta_db":-0.04946794598903597,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.00034505176105146,"delta_db":-0.08890632111437599,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2487567308786404,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23856593915862967,"pitch":null,"output_rms_dbfs":-26.78802408901095,"output_peak":0.35237655929870776,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.013893417956961},"common_prefix":{"samples":60480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.003972272858582,"spectral_convergence":0.44741130528928885,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.8963798469816515,"spectral_convergence":0.4807609411413532,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.141328134030649,"spectral_convergence":0.4822377466362019,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":5.023392016209277e-05,"delta_db":0.36009968624633515,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0005414816689055985,"delta_db":-0.17749204619352704,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0007200792720005805,"delta_db":-0.36048552951089935,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.0002542121502934708,"delta_db":-0.4560888562176691,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00018400665922083933,"delta_db":-0.04946794598903597,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.00034505176105146,"delta_db":-0.08890632111437599,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2487567308786404,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23856593915862967,"pitch":null,"output_rms_dbfs":-26.78802408901095,"output_peak":0.35237655929870776,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.013893417956961},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60480,"output_samples":60480,"intelligibility":{"estoi":{"value":0.67365890038556,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047046475372090966,"rms_dbfs":-26.54945814985232,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.546677348867952,"mean":-0.00027517540745003514,"zero_energy":false},"output":{"rms":0.04577188472888179,"rms_dbfs":-26.78802408901095,"peak":0.35237655929870776,"peak_dbfs":-9.059859786460683,"crest_factor_db":17.728164302550265,"mean":-8.66195048094955e-05,"zero_energy":false},"rms_difference_db":-0.23856593915862945},"full_scale":{"reference":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":378,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":371,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":40,"both_estimated_voiced_frames":38,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.7169811320754716,"absolute_difference_cents_median":3.1646753190150143,"absolute_difference_cents_p95":13.174730230870408,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00014780031156712295,"log_rms_mae_db":1.2769363978664596,"rms_envelope_nrmse":0.1282379263187393,"envelope_correlation":0.985679539276027,"rms_gain_error_db":-0.23862075002753014,"reference_mean_power":0.0022162793576100603,"output_mean_power":0.0020977920076248294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":1.2692207138635437e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044041649987677594,"output_power":0.00040929428726513663,"reference_active":true,"delta_db":-0.3182789811253621},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188839087946517,"output_power":9.001341221764681e-05,"reference_active":true,"delta_db":-1.2081585066192602},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.435323106110239e-05,"output_power":2.5087754050009965e-05,"reference_active":true,"delta_db":0.12905194055834335}],"active_band_count":3,"active_band_mae_db":0.5518298094343219}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00013486753957748745,"log_rms_mae_db":0.6461630488085514,"rms_envelope_nrmse":0.06698894931503398,"envelope_correlation":0.996370623288993,"rms_gain_error_db":-0.2387583556523265,"reference_mean_power":0.0022339157834068563,"output_mean_power":0.002114418553738299},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.176190391785113e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004109863651179866,"output_power":0.0003800697146207656,"reference_active":true,"delta_db":-0.33964149193164034},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3981406650238314e-05,"output_power":3.974965716264704e-05,"reference_active":true,"delta_db":-1.3291080997931104},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0096166790138942e-06,"output_power":1.2422088370898494e-06,"reference_active":true,"delta_db":0.9003809793000334}],"active_band_count":3,"active_band_mae_db":0.8563768570082614}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00012936155891805085,"log_rms_mae_db":1.7040462508090246,"rms_envelope_nrmse":0.1615268232591411,"envelope_correlation":0.9816680956262486,"rms_gain_error_db":-0.3066067344941122,"reference_mean_power":0.0012072233727869219,"output_mean_power":0.0011249337891255685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":8.078028680738071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004104434555454037,"output_power":0.00037524838680080987,"reference_active":true,"delta_db":-0.3893450142185057},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.258149083667088e-05,"output_power":7.556949371171042e-05,"reference_active":true,"delta_db":-0.8817765732122207},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4118213823838002e-05,"output_power":1.533465716685255e-05,"reference_active":true,"delta_db":0.3589431590569165}],"active_band_count":3,"active_band_mae_db":0.543354915495881}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00011150674016193665,"log_rms_mae_db":1.0483044114502167,"rms_envelope_nrmse":0.08970449378442193,"envelope_correlation":0.9946703525398152,"rms_gain_error_db":-0.3066095027670587,"reference_mean_power":0.0012168889554241998,"output_mean_power":0.0011339398009135577},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.50453906071102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003852443549284888,"output_power":0.0003530619009410364,"reference_active":true,"delta_db":-0.37885428432702906},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.764413089442773e-05,"output_power":3.713785691793503e-05,"reference_active":true,"delta_db":-1.0819257045419206},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0419619324101076e-06,"output_power":1.026512492627752e-06,"reference_active":true,"delta_db":-0.06487613418162333}],"active_band_count":3,"active_band_mae_db":0.5085520410168577}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":1.5808503877875804,"rms_envelope_nrmse":0.16661491799359635,"envelope_correlation":0.9800421516147847,"rms_gain_error_db":-0.44413461721189174,"reference_mean_power":0.0003662329710023158,"output_mean_power":0.0003306313104329377},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":2.347928764955057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015936395121522867,"output_power":0.00013512972851561212,"reference_active":true,"delta_db":-0.7163918472152121},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4803443060760012e-05,"output_power":1.8684930212846112e-05,"reference_active":true,"delta_db":-1.2302049096486174},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.158977700048839e-06,"output_power":6.818317499766698e-06,"reference_active":true,"delta_db":-0.21173789015854608}],"active_band_count":3,"active_band_mae_db":0.7194448823407918}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":1.0167401634185071,"rms_envelope_nrmse":0.09847397799834666,"envelope_correlation":0.993968335452833,"rms_gain_error_db":-0.4441676118728894,"reference_mean_power":0.0003691513253605656,"output_mean_power":0.0003332634384287549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":2.1457973458943643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015227225639161785,"output_power":0.00012916440303020608,"reference_active":true,"delta_db":-0.7147794214378433},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5037342052746154e-05,"output_power":1.1575705716346274e-05,"reference_active":true,"delta_db":-1.1362360129711966},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0212382710455385e-07,"output_power":3.349446613327762e-07,"reference_active":true,"delta_db":-0.7938674719967042}],"active_band_count":3,"active_band_mae_db":0.8816276354685814}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00011110909618382604,"log_rms_mae_db":1.4228515701835016,"rms_envelope_nrmse":0.20030667519786746,"envelope_correlation":0.9750749028717415,"rms_gain_error_db":0.19391212562814475,"reference_mean_power":0.00024441441718797425,"output_mean_power":0.0002555748016809196},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":3.4696433998856625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001104677897223138,"output_power":0.00011468809508780102,"reference_active":true,"delta_db":0.16282674801885239},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9627372395634915e-05,"output_power":5.589435039648028e-05,"reference_active":true,"delta_db":-0.28077758524008867},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.408768678752179e-06,"output_power":7.757384975296723e-06,"reference_active":true,"delta_db":2.4539803201931876}],"active_band_count":3,"active_band_mae_db":0.9658615511507095}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":0.9201943219041715,"rms_envelope_nrmse":0.11242347006680713,"envelope_correlation":0.992063344333731,"rms_gain_error_db":0.19392295209977117,"reference_mean_power":0.0002463642033912754,"output_mean_power":0.0002576142606928426},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.3341266023063217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010489106418058683,"output_power":0.00010983072431002755,"reference_active":true,"delta_db":0.1998535608654995},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.086235285110036e-05,"output_power":3.08294886656743e-05,"reference_active":true,"delta_db":-0.004627106399161575},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7476114035639134e-07,"output_power":4.426471706348171e-07,"reference_active":true,"delta_db":0.7230314079949554}],"active_band_count":3,"active_band_mae_db":0.3091706917532055}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.0743682384490967,"predicted_mos":2.0743682384490967,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60480,"input_sample_rate":16000,"input_seconds":3.78,"input_float32_sha256":"ee3e2c64352f7f76882aa8fb0c8430fc0c4060fc7e83bd371b756193bd99c86a","internal_samples":60480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0980813589994796},"started_at":"2026-10-13T04:41:02.134336+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.410400040593231e-05,"native_input_samples":60480,"native_input_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","output_sha256":"375b680c7b400d29bb4fb80887e78d8db818708b9ce2f46bb6f590e458fde666","output_samples":60480,"length_delta_samples":0,"analysis_s":1.6076126850002765,"synthesis_s":0.11161907499990775,"total_s":1.7192317600001843,"rtf":0.4548232169312657,"output_peak":0.49383870568827076,"output_rms":0.05579824890591493,"finished_at":"2026-10-13T04:41:03.859737+00:00","evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"375b680c7b400d29bb4fb80887e78d8db818708b9ce2f46bb6f590e458fde666","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.089504503135956,"spectral_convergence":0.5014396813171698,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.909923841553081,"spectral_convergence":0.45983822363832705,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.118469126420548,"spectral_convergence":0.5099684671746355,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":4.943829029366909e-05,"delta_db":0.2907634089862922,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0008385410936299273,"delta_db":1.721913871376353,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0011271481067912377,"delta_db":1.585521233363999,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.00038169490713478826,"delta_db":1.3091117547596787,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00022525165440067285,"delta_db":0.828876543585785,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.0004913705287161415,"delta_db":1.446342245510258,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1970881762637275,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4818695470872667,"pitch":null,"output_rms_dbfs":-25.067588602765056,"output_peak":0.49383870568827076,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.039299157036528},"common_prefix":{"samples":60480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.089504503135956,"spectral_convergence":0.5014396813171698,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.909923841553081,"spectral_convergence":0.45983822363832705,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.118469126420548,"spectral_convergence":0.5099684671746355,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":4.943829029366909e-05,"delta_db":0.2907634089862922,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0008385410936299273,"delta_db":1.721913871376353,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0011271481067912377,"delta_db":1.585521233363999,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.00038169490713478826,"delta_db":1.3091117547596787,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00022525165440067285,"delta_db":0.828876543585785,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.0004913705287161415,"delta_db":1.446342245510258,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1970881762637275,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4818695470872667,"pitch":null,"output_rms_dbfs":-25.067588602765056,"output_peak":0.49383870568827076,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.039299157036528},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60480,"output_samples":60480,"intelligibility":{"estoi":{"value":0.7439709020549296,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047046475372090966,"rms_dbfs":-26.54945814985232,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.546677348867952,"mean":-0.00027517540745003514,"zero_energy":false},"output":{"rms":0.055798248905914935,"rms_dbfs":-25.067588602765056,"peak":0.49383870568827076,"peak_dbfs":-6.128297485810776,"crest_factor_db":18.939291116954276,"mean":-7.555009288117041e-09,"zero_energy":false},"rms_difference_db":1.481869547087264},"full_scale":{"reference":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":378,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":371,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":59,"both_estimated_voiced_frames":42,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.7924528301886793,"absolute_difference_cents_median":12.28118034755009,"absolute_difference_cents_p95":1192.6848982116428,"difference_over_600_cents_frames":8,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00014780031156712295,"log_rms_mae_db":2.7039212562042563,"rms_envelope_nrmse":0.24423434826464954,"envelope_correlation":0.9833752297147279,"rms_gain_error_db":1.481911242232621,"reference_mean_power":0.0022162793576100603,"output_mean_power":0.0031175657793876955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":1.2692207138635437e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044041649987677594,"output_power":0.0006724111391308333,"reference_active":true,"delta_db":1.8377131820425534},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188839087946517,"output_power":0.00023129639658072966,"reference_active":true,"delta_db":2.8904579099232723},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.435323106110239e-05,"output_power":4.7809866653605254e-05,"reference_active":true,"delta_db":2.929609430822268}],"active_band_count":3,"active_band_mae_db":2.5525935075960313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00013486753957748745,"log_rms_mae_db":2.329276737034717,"rms_envelope_nrmse":0.2169658481082253,"envelope_correlation":0.9916292876029633,"rms_gain_error_db":1.4820740826207905,"reference_mean_power":0.0022339157834068563,"output_mean_power":0.003142492174306248},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.176190391785113e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004109863651179866,"output_power":0.0006410313511885728,"reference_active":true,"delta_db":1.930518562971579},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3981406650238314e-05,"output_power":9.664704251876386e-05,"reference_active":true,"delta_db":2.5294437155384255},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0096166790138942e-06,"output_power":2.4247187198361906e-06,"reference_active":true,"delta_db":3.805048488419879}],"active_band_count":3,"active_band_mae_db":2.755003588976628}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00012936155891805085,"log_rms_mae_db":2.1802808956843966,"rms_envelope_nrmse":0.2889409448034737,"envelope_correlation":0.980792564696038,"rms_gain_error_db":1.7464413273900579,"reference_mean_power":0.0012072233727869219,"output_mean_power":0.001804811160451499},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":8.078028680738071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004104434555454037,"output_power":0.0006396438601863813,"reference_active":true,"delta_db":1.9268490023492024},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.258149083667088e-05,"output_power":0.0002202218333998841,"reference_active":true,"delta_db":3.7633620399018355},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4118213823838002e-05,"output_power":2.818750855259741e-05,"reference_active":true,"delta_db":3.002769358880693}],"active_band_count":3,"active_band_mae_db":2.897660133710577}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00011150674016193665,"log_rms_mae_db":1.8992616504128588,"rms_envelope_nrmse":0.265783327371184,"envelope_correlation":0.9873269038310031,"rms_gain_error_db":1.7464378500183533,"reference_mean_power":0.0012168889554241998,"output_mean_power":0.0018192598476464079},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.50453906071102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003852443549284888,"output_power":0.0006190965646515891,"reference_active":true,"delta_db":2.0602211050932415},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.764413089442773e-05,"output_power":0.00010937620737502921,"reference_active":true,"delta_db":3.6091345099234284},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0419619324101076e-06,"output_power":1.6443457876720692e-06,"reference_active":true,"delta_db":1.981412976097861}],"active_band_count":3,"active_band_mae_db":2.550256197038177}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":1.4535059692676755,"rms_envelope_nrmse":0.2172886676500549,"envelope_correlation":0.9827859862866914,"rms_gain_error_db":1.1686261417806638,"reference_mean_power":0.0003662329710023158,"output_mean_power":0.00047931393386700335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":2.347928764955057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015936395121522867,"output_power":0.00021852221697920646,"reference_active":true,"delta_db":1.3710550896238527},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4803443060760012e-05,"output_power":2.617528214959519e-05,"reference_active":true,"delta_db":0.23379400651374332},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.158977700048839e-06,"output_power":9.418331271135711e-06,"reference_active":true,"delta_db":1.1912295221433074}],"active_band_count":3,"active_band_mae_db":0.9320262060936345}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":1.0532523235734441,"rms_envelope_nrmse":0.17932615978150931,"envelope_correlation":0.9920522688742904,"rms_gain_error_db":1.1686446644286117,"reference_mean_power":0.0003691513253605656,"output_mean_power":0.0004831354427595603},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":2.1457973458943643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015227225639161785,"output_power":0.00020749225160339416,"reference_active":true,"delta_db":1.3438110024344085},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5037342052746154e-05,"output_power":1.60322769434967e-05,"reference_active":true,"delta_db":0.27824127620857864},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0212382710455385e-07,"output_power":5.215528308599506e-07,"reference_active":true,"delta_db":1.1293849980591266}],"active_band_count":3,"active_band_mae_db":0.9171457589007046}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00011110909618382604,"log_rms_mae_db":1.409259875332215,"rms_envelope_nrmse":0.31449442743790534,"envelope_correlation":0.9672850410377994,"rms_gain_error_db":1.5087990000457274,"reference_mean_power":0.00024441441718797425,"output_mean_power":0.00034594473064270405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":3.4696433998856625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001104677897223138,"output_power":0.00016923512376866462,"reference_active":true,"delta_db":1.852548387779453},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9627372395634915e-05,"output_power":9.450785835096169e-05,"reference_active":true,"delta_db":2.000222500239914},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.408768678752179e-06,"output_power":2.4290374993029212e-05,"reference_active":true,"delta_db":7.411169067963864}],"active_band_count":3,"active_band_mae_db":3.75464665199441}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":0.9153681634684651,"rms_envelope_nrmse":0.23554215988506785,"envelope_correlation":0.9899116370410976,"rms_gain_error_db":1.5088183967390565,"reference_mean_power":0.0002463642033912754,"output_mean_power":0.00034870601994459475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.3341266023063217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010489106418058683,"output_power":0.00017126355903886463,"reference_active":true,"delta_db":2.1292647327778873},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.086235285110036e-05,"output_power":5.589045923302739e-05,"reference_active":true,"delta_db":2.5790864581342783},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7476114035639134e-07,"output_power":1.5321314600788091e-06,"reference_active":true,"delta_db":6.115414784792442}],"active_band_count":3,"active_band_mae_db":3.6079219919015357}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"facodec","status":"measured","mos":{"status":"measured","value":4.046996593475342,"predicted_mos":4.046996593475342,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60600,"input_sample_rate":16000,"input_seconds":3.7875,"input_float32_sha256":"cf84a6d2b7eddb87517669d60c391794835b737f4f14eda019d008144924bbfa","internal_samples":60600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.088960858998689},"started_at":"2026-10-13T04:45:59.818315+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5164999897242524e-05,"native_input_samples":60480,"native_input_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","output_sha256":"ec57abe5641d3806a2167bf85f2c25a8ae4c64a2e9c919db7c4edd15eb4c05f8","output_samples":60600,"length_delta_samples":120,"analysis_s":1.0153720879998218,"synthesis_s":7.503385478999917,"total_s":8.518757566999739,"rtf":2.253639567989349,"output_peak":0.3434785008430481,"output_rms":0.041764892973194495,"finished_at":"2026-10-13T04:46:08.342163+00:00","evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"ec57abe5641d3806a2167bf85f2c25a8ae4c64a2e9c919db7c4edd15eb4c05f8","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":60480,"output_samples":60600,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.583772543613613,"output_peak":0.3434785008430481},"common_prefix":{"samples":60480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998019801980198,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.51089848665645,"spectral_convergence":0.39560402637162695,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.1224680686171595,"spectral_convergence":0.45755223546289125,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.2683340042476985,"spectral_convergence":0.4897730182730606,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":3.556625249500977e-05,"delta_db":-1.139489950517303,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0005013718522700508,"delta_db":-0.5117301500967876,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0006176168351191711,"delta_db":-1.0270973700107016,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.00023691881386868238,"delta_db":-0.7620564013882631,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00012899648853929136,"delta_db":-1.5920244668614656,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.00022729674126353437,"delta_db":-1.9018167105950101,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1557025082449217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0257065402851053,"pitch":null,"output_rms_dbfs":-27.575164690137427,"output_peak":0.3434785008430481,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.967233519840437},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60480,"output_samples":60480,"intelligibility":{"estoi":{"value":0.7536360447277044,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047046475372090966,"rms_dbfs":-26.54945814985232,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.546677348867952,"mean":-0.00027517540745003514,"zero_energy":false},"output":{"rms":0.041806303155813,"rms_dbfs":-27.575164690137427,"peak":0.3434785008430481,"peak_dbfs":-9.282008826087765,"crest_factor_db":18.29315586404966,"mean":0.00011235330171977751,"zero_energy":false},"rms_difference_db":-1.0257065402851069},"full_scale":{"reference":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":378,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":371,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":56,"both_estimated_voiced_frames":49,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9245283018867925,"absolute_difference_cents_median":4.765203801010817,"absolute_difference_cents_p95":20.319605666142614,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00014780031156712295,"log_rms_mae_db":1.4848988325832968,"rms_envelope_nrmse":0.1814249334566296,"envelope_correlation":0.9795080485773187,"rms_gain_error_db":-1.0258484126993208,"reference_mean_power":0.0022162793576100603,"output_mean_power":0.0017500064936187118},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":1.2692207138635437e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044041649987677594,"output_power":0.00033418953319261526,"reference_active":true,"delta_db":-1.19870737445794},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188839087946517,"output_power":0.00011015226806733539,"reference_active":true,"delta_db":-0.3312963202936181},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.435323106110239e-05,"output_power":2.0984593824965587e-05,"reference_active":true,"delta_db":-0.6465602184339347}],"active_band_count":3,"active_band_mae_db":0.7255213043951643}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00013486753957748745,"log_rms_mae_db":1.1569171937016403,"rms_envelope_nrmse":0.16439923633819417,"envelope_correlation":0.9843887008439102,"rms_gain_error_db":-1.026078547125395,"reference_mean_power":0.0022339157834068563,"output_mean_power":0.0017638390035347154},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.176190391785113e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004109863651179866,"output_power":0.0003192142770855817,"reference_active":true,"delta_db":-1.0974510672038722},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3981406650238314e-05,"output_power":4.7424505714453914e-05,"reference_active":true,"delta_db":-0.5624138411887167},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0096166790138942e-06,"output_power":1.0285549941578166e-06,"reference_active":true,"delta_db":0.08071000679730386}],"active_band_count":3,"active_band_mae_db":0.5801916383966309}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00012936155891805085,"log_rms_mae_db":1.769776735044433,"rms_envelope_nrmse":0.16037071658357707,"envelope_correlation":0.9859067980588384,"rms_gain_error_db":-0.7908156832067131,"reference_mean_power":0.0012072233727869219,"output_mean_power":0.0010062504018391503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":8.078028680738071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004104434555454037,"output_power":0.00037193935295844356,"reference_active":true,"delta_db":-0.42781204083138674},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.258149083667088e-05,"output_power":7.772339540528976e-05,"reference_active":true,"delta_db":-0.759724051077588},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4118213823838002e-05,"output_power":1.3472081782954003e-05,"reference_active":true,"delta_db":-0.2034504453654912}],"active_band_count":3,"active_band_mae_db":0.4636621790914887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00011150674016193665,"log_rms_mae_db":1.4100135496701551,"rms_envelope_nrmse":0.14633704269033282,"envelope_correlation":0.9886324218564524,"rms_gain_error_db":-0.7908263853346365,"reference_mean_power":0.0012168889554241998,"output_mean_power":0.0010143044034153987},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.50453906071102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003852443549284888,"output_power":0.00034658948532372794,"reference_active":true,"delta_db":-0.459209004764639},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.764413089442773e-05,"output_power":4.150396971472284e-05,"reference_active":true,"delta_db":-0.5991977162523359},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0419619324101076e-06,"output_power":1.0146199433345054e-06,"reference_active":true,"delta_db":-0.11548457965795514}],"active_band_count":3,"active_band_mae_db":0.3912971002249766}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":2.390141630714412,"rms_envelope_nrmse":0.256775559788832,"envelope_correlation":0.9561733667532092,"rms_gain_error_db":-1.085357186789013,"reference_mean_power":0.0003662329710023158,"output_mean_power":0.0002852474176616634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":2.347928764955057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015936395121522867,"output_power":9.358796864032523e-05,"reference_active":true,"delta_db":-2.31170068097229},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4803443060760012e-05,"output_power":2.3767424276592234e-05,"reference_active":true,"delta_db":-0.18529852173520006},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.158977700048839e-06,"output_power":8.864559427942498e-06,"reference_active":true,"delta_db":0.9280614625966604}],"active_band_count":3,"active_band_mae_db":1.1416868884347169}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":2.037073032338672,"rms_envelope_nrmse":0.2196865207408703,"envelope_correlation":0.9686963505602328,"rms_gain_error_db":-1.0853495940570452,"reference_mean_power":0.0003691513253605656,"output_mean_power":0.0002875209353531809},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":2.1457973458943643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015227225639161785,"output_power":9.176822987904215e-05,"reference_active":true,"delta_db":-2.1992842855850894},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5037342052746154e-05,"output_power":1.4554890004688649e-05,"reference_active":true,"delta_db":-0.14162150982400687},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0212382710455385e-07,"output_power":5.574017096319262e-07,"reference_active":true,"delta_db":1.4180848924091507}],"active_band_count":3,"active_band_mae_db":1.2529968959394155}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00011110909618382604,"log_rms_mae_db":2.146103214515421,"rms_envelope_nrmse":0.37351238887953675,"envelope_correlation":0.9459506150284877,"rms_gain_error_db":-2.8694458296295617,"reference_mean_power":0.00024441441718797425,"output_mean_power":0.00012623571287970464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":3.4696433998856625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001104677897223138,"output_power":4.5970805904178646e-05,"reference_active":true,"delta_db":-3.8075354724248847},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9627372395634915e-05,"output_power":2.6605710496430522e-05,"reference_active":true,"delta_db":-3.504708105747362},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.408768678752179e-06,"output_power":2.550276004971231e-06,"reference_active":true,"delta_db":-2.37730127939051}],"active_band_count":3,"active_band_mae_db":3.2298482858542523}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":1.8453642795343919,"rms_envelope_nrmse":0.3593266957431599,"envelope_correlation":0.9537025151796116,"rms_gain_error_db":-2.869445704003206,"reference_mean_power":0.0002463642033912754,"output_mean_power":0.00012724274656365311},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.3341266023063217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010489106418058683,"output_power":4.550020890696728e-05,"reference_active":true,"delta_db":-3.627251009471988},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.086235285110036e-05,"output_power":1.309940891675929e-05,"reference_active":true,"delta_db":-3.72177332794235},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7476114035639134e-07,"output_power":1.950683355059643e-07,"reference_active":true,"delta_db":-2.8356777358465823}],"active_band_count":3,"active_band_mae_db":3.3949006910869737}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.8766567707061768,"predicted_mos":2.8766567707061768,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":166912,"input_sample_rate":44100,"input_seconds":3.7848526077097504,"input_float32_sha256":"26dfa25dc317a8c5eab6c5ab8bda236a842f7f5cc799dca5230db689ee0db9c5","internal_samples":60558,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6192624539999088},"started_at":"2026-10-13T04:59:05.074710+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0043146859998159925,"native_input_samples":166698,"native_input_sha256":"989662fa032f5d03c9833c0cef1897d0ed99da11596ca4c3e933596421c7f298","output_sha256":"c7bb844829bae510b48baf944235e4a83f4840f8863178215740bc5a88ee04b5","output_samples":166912,"length_delta_samples":214,"analysis_s":0.08807027000057133,"synthesis_s":19.910564203999456,"total_s":19.998634474000028,"rtf":5.290644040740748,"output_peak":0.3483743369579315,"output_rms":0.04263328633854329,"finished_at":"2026-10-13T04:59:25.086048+00:00","evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"c7bb844829bae510b48baf944235e4a83f4840f8863178215740bc5a88ee04b5","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":60480,"output_samples":60558,"length_delta_samples":78,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.43448410061425,"output_peak":0.3391592824403538},"common_prefix":{"samples":60480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998711978599029,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.617979259403658,"spectral_convergence":0.48834966519132134,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.88834639896964,"spectral_convergence":0.5336609088838675,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.802822501604318,"spectral_convergence":0.5380542894782719,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":2.7706477406706915e-05,"delta_db":-2.224057891736133,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0005877094121758928,"delta_db":0.17829682597919372,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0005903555450888892,"delta_db":-1.2231521573893942,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.0002702326502661142,"delta_db":-0.19067369290244665,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00014225602950695344,"delta_db":-1.1670965207350341,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.00018936826315712452,"delta_db":-2.6946768443989164,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.279658988856853,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8794503439859912,"pitch":null,"output_rms_dbfs":-27.428908493838314,"output_peak":0.3391592824403538,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.769716053325872},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60480,"output_samples":60480,"intelligibility":{"estoi":{"value":0.6151960119403798,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047046475372090966,"rms_dbfs":-26.54945814985232,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.546677348867952,"mean":-0.00027517540745003514,"zero_energy":false},"output":{"rms":0.042516213114548215,"rms_dbfs":-27.428908493838314,"peak":0.3391592824403538,"peak_dbfs":-9.391925844869801,"crest_factor_db":18.03698264896851,"mean":-9.529213392377692e-05,"zero_energy":false},"rms_difference_db":-0.8794503439859938},"full_scale":{"reference":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":378,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":371,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":75,"both_estimated_voiced_frames":50,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.9433962264150944,"absolute_difference_cents_median":8.50612820070959,"absolute_difference_cents_p95":25.46744542166657,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00014780031156712295,"log_rms_mae_db":2.0744863611457154,"rms_envelope_nrmse":0.14110247160154904,"envelope_correlation":0.9901935124138838,"rms_gain_error_db":-0.8794377942109386,"reference_mean_power":0.0022162793576100603,"output_mean_power":0.0018100089487504611},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":1.2692207138635437e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044041649987677594,"output_power":0.0003751807376618183,"reference_active":true,"delta_db":-0.6962304817093532},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188839087946517,"output_power":8.174890701637663e-05,"reference_active":true,"delta_db":-1.6264112100829102},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.435323106110239e-05,"output_power":1.8449764698931797e-05,"reference_active":true,"delta_db":-1.205657576166763}],"active_band_count":3,"active_band_mae_db":1.176099755986342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00013486753957748745,"log_rms_mae_db":1.7559445161067693,"rms_envelope_nrmse":0.11812523389207774,"envelope_correlation":0.995650568655236,"rms_gain_error_db":-0.879328377115969,"reference_mean_power":0.0022339157834068563,"output_mean_power":0.001824458372816826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.176190391785113e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004109863651179866,"output_power":0.00035177622585542414,"reference_active":true,"delta_db":-0.6756092883789906},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3981406650238314e-05,"output_power":3.688940736116523e-05,"reference_active":true,"delta_db":-1.6534251901731711},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0096166790138942e-06,"output_power":8.302695668256372e-07,"reference_active":true,"delta_db":-0.8493739722982694}],"active_band_count":3,"active_band_mae_db":1.0594694836168104}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00012936155891805085,"log_rms_mae_db":2.519083693880611,"rms_envelope_nrmse":0.1668851827588288,"envelope_correlation":0.9839574452031654,"rms_gain_error_db":-0.768531811914474,"reference_mean_power":0.0012072233727869219,"output_mean_power":0.0010114267927608443},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":8.078028680738071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004104434555454037,"output_power":0.0003396301512840919,"reference_active":true,"delta_db":-0.8224709673625931},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.258149083667088e-05,"output_power":6.778626861664064e-05,"reference_active":true,"delta_db":-1.3538244167017677},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4118213823838002e-05,"output_power":1.1594030144915477e-05,"reference_active":true,"delta_db":-0.8554532988850173}],"active_band_count":3,"active_band_mae_db":1.0105828943164594}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00011150674016193665,"log_rms_mae_db":1.9309569357386696,"rms_envelope_nrmse":0.14692030755735,"envelope_correlation":0.9881240446232784,"rms_gain_error_db":-0.7685509582565977,"reference_mean_power":0.0012168889554241998,"output_mean_power":0.0010195202437246575},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.50453906071102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003852443549284888,"output_power":0.00031844816849033793,"reference_active":true,"delta_db":-0.8269752817762115},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.764413089442773e-05,"output_power":3.4176129034626875e-05,"reference_active":true,"delta_db":-1.4428653839837375},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0419619324101076e-06,"output_power":8.046474124137068e-07,"reference_active":true,"delta_db":-1.1224623349140248}],"active_band_count":3,"active_band_mae_db":1.1307676668913247}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":2.8298389730114373,"rms_envelope_nrmse":0.2774133967848859,"envelope_correlation":0.9449657057905301,"rms_gain_error_db":-0.5871225284504915,"reference_mean_power":0.0003662329710023158,"output_mean_power":0.00031992279642585164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":2.347928764955057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015936395121522867,"output_power":0.0001293596476789026,"reference_active":true,"delta_db":-0.9059126489672619},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4803443060760012e-05,"output_power":2.454929210871421e-05,"reference_active":true,"delta_db":-0.044729975503635486},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.158977700048839e-06,"output_power":2.8682370619805863e-06,"reference_active":true,"delta_db":-3.9723596635932186}],"active_band_count":3,"active_band_mae_db":1.6410007626880387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":2.493514290171542,"rms_envelope_nrmse":0.23910034902803132,"envelope_correlation":0.9578510113097302,"rms_gain_error_db":-0.5871337326865517,"reference_mean_power":0.0003691513253605656,"output_mean_power":0.0003224712928175279},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":2.1457973458943643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015227225639161785,"output_power":0.000123377496665342,"reference_active":true,"delta_db":-0.9138482907664235},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5037342052746154e-05,"output_power":1.4696456735597739e-05,"reference_active":true,"delta_db":-0.09958438201401854},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0212382710455385e-07,"output_power":2.0485337071540766e-07,"reference_active":true,"delta_db":-2.9291669283019117}],"active_band_count":3,"active_band_mae_db":1.3141998670274513}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00011110909618382604,"log_rms_mae_db":4.31274005044611,"rms_envelope_nrmse":0.38659401474123317,"envelope_correlation":0.9276690786589723,"rms_gain_error_db":-2.590029011132286,"reference_mean_power":0.00024441441718797425,"output_mean_power":0.00013462444279720842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":3.4696433998856625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001104677897223138,"output_power":4.110063935549455e-05,"reference_active":true,"delta_db":-4.29387086831804},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9627372395634915e-05,"output_power":2.096600100849825e-05,"reference_active":true,"delta_db":-4.539300693256569},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.408768678752179e-06,"output_power":1.993771606593209e-06,"reference_active":true,"delta_db":-3.4464190577115454}],"active_band_count":3,"active_band_mae_db":4.093196873095384}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":3.976378661448897,"rms_envelope_nrmse":0.36740736127371754,"envelope_correlation":0.9378064334529793,"rms_gain_error_db":-2.589987126114364,"reference_mean_power":0.0002463642033912754,"output_mean_power":0.00013569970160367672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.3341266023063217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010489106418058683,"output_power":3.969697448834271e-05,"reference_active":true,"delta_db":-4.21981083406222},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.086235285110036e-05,"output_power":1.0144883004510213e-05,"reference_active":true,"delta_db":-4.831819893550995},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7476114035639134e-07,"output_power":1.6248907980632003e-07,"reference_active":true,"delta_db":-3.6293037256214715}],"active_band_count":3,"active_band_mae_db":4.226978151078229}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"golf","status":"measured","mos":{"status":"measured","value":2.903768301010132,"predicted_mos":2.903768301010132,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90480,"input_sample_rate":24000,"input_seconds":3.77,"input_float32_sha256":"8142fd3da26b0a9d76d994c04f73376f96bd6511c5fd5e79413483f001a1557b","internal_samples":60320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.858878476999962},"started_at":"2026-10-13T05:01:57.049244+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018385449993729708,"native_input_samples":90720,"native_input_sha256":"bdae00d9000f0fc06aec3a95a1c29a3b3e09b961f4b106ec43cb30b42cbb3877","output_sha256":"c2aef813402a04eb9685bf6d29bdf4c838149b09d08dbbd09e2adf3c1fd55f3d","output_samples":90480,"length_delta_samples":-240,"analysis_s":0.3771033829998487,"synthesis_s":0.10446775700074795,"total_s":0.48157114000059664,"rtf":0.12739977248693032,"output_peak":0.4067707061767578,"output_rms":0.04370559220392711,"finished_at":"2026-10-13T05:01:57.539867+00:00","analysis_stages":{"external_f0_s":0.04013980600029754},"evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"c2aef813402a04eb9685bf6d29bdf4c838149b09d08dbbd09e2adf3c1fd55f3d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60480,"output_samples":60320,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.2163920603915,"output_peak":0.40727241547302645},"common_prefix":{"samples":60320,"sample_rate":16000,"reference_coverage":0.9973544973544973,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60320,"output_samples":60320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.443567214217556,"spectral_convergence":0.4928430818768724,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.151777936464192,"spectral_convergence":0.46819356451763067,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.199526159728776,"spectral_convergence":0.5075096298113637,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6330010540814775e-05,"output_power":3.3151247663879735e-05,"delta_db":-1.4536252107243828,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000565566278051712,"output_power":0.0006417847397821713,"delta_db":0.5490587961556191,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007844727592037781,"output_power":0.000671140759276297,"delta_db":-0.6776425241796837,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002831108002004783,"output_power":0.0002401879538490528,"delta_db":-0.7140521485079645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018660738533793583,"output_power":0.00011735248798143953,"delta_db":-2.0143652629335,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003531146280665322,"output_power":0.00019466041490661922,"delta_db":-2.586380642329668,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3325207641384698,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6783737611478218,"pitch":null,"output_rms_dbfs":-27.2163920603915,"output_peak":0.40727241547302645,"reference_rms_dbfs":-26.538018299243674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.2649571034701745},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60320,"output_samples":60320,"intelligibility":{"estoi":{"value":0.7231706216088764,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047108479294811254,"rms_dbfs":-26.538018299243674,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.535237498259306,"mean":-0.00027479025033804084,"zero_energy":false},"output":{"rms":0.0435692813922281,"rms_dbfs":-27.2163920603915,"peak":0.40727241547302645,"peak_dbfs":-7.802300072944277,"crest_factor_db":19.414091987447218,"mean":-0.00011288471022411929,"zero_energy":false},"rms_difference_db":-0.6783737611478244},"full_scale":{"reference":{"sample_count":60320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":377,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":370,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":56,"both_estimated_voiced_frames":47,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.8867924528301887,"absolute_difference_cents_median":10.253514341804184,"absolute_difference_cents_p95":20.412152002441708,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":753,"rms_floor":0.00014780031156712295,"log_rms_mae_db":2.3408516659292946,"rms_envelope_nrmse":0.21969145628293515,"envelope_correlation":0.9613147260172198,"rms_gain_error_db":-0.6783550121340555,"reference_mean_power":0.002222133767712245,"output_mean_power":0.0019007924478326743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":753,"frequency_resolution_hz":0.2656042496679947,"power_floor":1.2684883722554531e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000439444844884082,"output_power":0.00043031411428807504,"reference_active":true,"delta_db":-0.09118782980801529},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011902020740442468,"output_power":7.754578133641162e-05,"reference_active":true,"delta_db":-1.860625265272097},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4014364330281287e-05,"output_power":2.285137671562636e-05,"reference_active":true,"delta_db":-0.21558725260886097}],"active_band_count":3,"active_band_mae_db":0.7224667825629911}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00013486753957748745,"log_rms_mae_db":1.7591573597842527,"rms_envelope_nrmse":0.1807641087529401,"envelope_correlation":0.9738765480620316,"rms_gain_error_db":-0.678317233017746,"reference_mean_power":0.0022398561271677548,"output_mean_power":0.0019159686552738745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.1751878382231853e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004103275553588298,"output_power":0.0003978166323266153,"reference_active":true,"delta_db":-0.13447746194985027},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3352333239170863e-05,"output_power":3.2350683933809916e-05,"reference_active":true,"delta_db":-2.1726995039762613},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0034016450858607e-06,"output_power":1.2155752624191364e-06,"reference_active":true,"delta_db":0.8330704466013439}],"active_band_count":3,"active_band_mae_db":1.0467491375091518}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":753,"rms_floor":0.00012936155891805085,"log_rms_mae_db":2.028071884266146,"rms_envelope_nrmse":0.2749544545168703,"envelope_correlation":0.947149844170314,"rms_gain_error_db":-0.07462828140189945,"reference_mean_power":0.0012104286938965506,"output_mean_power":0.0011898066233251307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":753,"frequency_resolution_hz":0.2656042496679947,"power_floor":8.067549758683906e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041036621828343533,"output_power":0.00045549684172734576,"reference_active":true,"delta_db":0.45313767994877174},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.270025081576261e-05,"output_power":6.748374085347578e-05,"reference_active":true,"delta_db":-1.378817601952956},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.370756535697394e-05,"output_power":1.2722022660917614e-05,"reference_active":true,"delta_db":-0.32404160432515683}],"active_band_count":3,"active_band_mae_db":0.7186656287422949}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00011150674016193665,"log_rms_mae_db":1.3679568573813594,"rms_envelope_nrmse":0.2270797486384023,"envelope_correlation":0.9618934141392044,"rms_gain_error_db":-0.07462321896992884,"reference_mean_power":0.001220145879602331,"output_mean_power":0.0011993596554172226},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":7.492952185928487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000385070846842501,"output_power":0.0004213801928301258,"reference_active":true,"delta_db":0.39133477480367046},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.752441243737112e-05,"output_power":3.270887956290723e-05,"reference_active":true,"delta_db":-1.6225108852425516},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0268641205693823e-06,"output_power":9.375122683418511e-07,"reference_active":true,"delta_db":-0.3953601989522123}],"active_band_count":3,"active_band_mae_db":0.8030686196661447}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":753,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":1.857218724460216,"rms_envelope_nrmse":0.24358027142447805,"envelope_correlation":0.9600459276876655,"rms_gain_error_db":-1.0023727706794447,"reference_mean_power":0.0003672041995756126,"output_mean_power":0.0002915213472403861},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":753,"frequency_resolution_hz":0.2656042496679947,"power_floor":2.349279001307805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001600993122383436,"output_power":0.00010953702542728135,"reference_active":true,"delta_db":-1.6482852315159633},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4427185152124117e-05,"output_power":1.7338652543837198e-05,"reference_active":true,"delta_db":-1.4885808057918506},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.186998396165689e-06,"output_power":6.508544293073953e-06,"reference_active":true,"delta_db":-0.4306368320953602}],"active_band_count":3,"active_band_mae_db":1.1891676231343915}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":1.4083856260875538,"rms_envelope_nrmse":0.19687847101503714,"envelope_correlation":0.9756006327373754,"rms_gain_error_db":-1.0023504011671251,"reference_mean_power":0.00037013798444358057,"output_mean_power":0.00029385197609451644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.1464885233849555e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015281279915839116,"output_power":0.00010424692874165006,"reference_active":true,"delta_db":-1.6609646209404856},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4840300860200075e-05,"output_power":9.97399570933169e-06,"reference_active":true,"delta_db":-1.7257352851046956},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9737029916096664e-07,"output_power":3.292371680353973e-07,"reference_active":true,"delta_db":-0.8168654589840312}],"active_band_count":3,"active_band_mae_db":1.4011884550097375}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":753,"rms_floor":0.00011110909618382604,"log_rms_mae_db":3.3010251476798755,"rms_envelope_nrmse":0.3777747445242345,"envelope_correlation":0.9335158468213502,"rms_gain_error_db":-2.5886881188527253,"reference_mean_power":0.0002450628596611517,"output_mean_power":0.00013502328977266543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":753,"frequency_resolution_hz":0.2656042496679947,"power_floor":3.4888510760876137e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011122057327605943,"output_power":3.834599748066813e-05,"reference_active":true,"delta_db":-4.624650898779495},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.968353237128163e-05,"output_power":2.959549126983286e-05,"reference_active":true,"delta_db":-3.0462896552324925},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.405077860323894e-06,"output_power":9.311335818804454e-06,"reference_active":true,"delta_db":3.250584009894294}],"active_band_count":3,"active_band_mae_db":3.6405081879687606}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60320,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":3.038888262086487,"rms_envelope_nrmse":0.34308072779686505,"envelope_correlation":0.953493749661624,"rms_gain_error_db":-2.588710514613429,"reference_mean_power":0.0002470228291631148,"output_mean_power":0.00013610248035780345},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":3.352345843045223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010560265861700354,"output_power":3.925736566591336e-05,"reference_active":true,"delta_db":-4.297536987288691},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0688172419943233e-05,"output_power":1.627987281415378e-05,"reference_active":true,"delta_db":-2.7532001790057943},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.746342191310223e-07,"output_power":5.837241283588649e-07,"reference_active":true,"delta_db":1.9260020197736425}],"active_band_count":3,"active_band_mae_db":2.9922463953560428}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.3823578357696533,"predicted_mos":2.3823578357696533,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90840,"input_sample_rate":24000,"input_seconds":3.785,"input_float32_sha256":"bac07c2327da3ae0d2f9b0566fc6884340ebb9bff69acbdd272c5e3b6665c95c","internal_samples":60560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2785990529991977},"started_at":"2026-10-13T05:04:53.690940+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0021032320000813343,"native_input_samples":90720,"native_input_sha256":"bdae00d9000f0fc06aec3a95a1c29a3b3e09b961f4b106ec43cb30b42cbb3877","output_sha256":"927452cf80d96b006c9e4ad6d28ad6f02d2f0b6358092bf492ad989ad18560e1","output_samples":90840,"length_delta_samples":120,"analysis_s":1.3072535960000096,"synthesis_s":8.916947834999519,"total_s":10.224201430999528,"rtf":2.7048151933861186,"output_peak":0.3308231830596924,"output_rms":0.04366619025660771,"finished_at":"2026-10-13T05:05:03.927179+00:00","analysis_stages":{"external_f0_s":0.8683156329998383},"evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"927452cf80d96b006c9e4ad6d28ad6f02d2f0b6358092bf492ad989ad18560e1","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60480,"output_samples":60560,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.21852236742293,"output_peak":0.31002393204851314},"common_prefix":{"samples":60480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986789960369881,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.076660471335854,"spectral_convergence":0.4613838719901879,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.812392397714279,"spectral_convergence":0.44768905418983956,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.727089604210398,"spectral_convergence":0.4355045604992747,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":2.7506335424875515e-05,"delta_db":-2.2555436772141766,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0005289147655194976,"delta_db":-0.2794727217747137,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0007014740536888638,"delta_db":-0.4741725071285036,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.00026794522594097,"delta_db":-0.22759166327011865,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.0001716576997267301,"delta_db":-0.3511704652717704,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.00020236206238225766,"delta_db":-2.406457834901823,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9990681449268509,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6633255327064081,"pitch":null,"output_rms_dbfs":-27.21278368255873,"output_peak":0.31002393204851314,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.872047491086843},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60480,"output_samples":60480,"intelligibility":{"estoi":{"value":0.7568873693428824,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047046475372090966,"rms_dbfs":-26.54945814985232,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.546677348867952,"mean":-0.00027517540745003514,"zero_energy":false},"output":{"rms":0.043587385132434764,"rms_dbfs":-27.21278368255873,"peak":0.31002393204851314,"peak_dbfs":-10.172095597157227,"crest_factor_db":17.040688085401502,"mean":0.00010050176113862474,"zero_energy":false},"rms_difference_db":-0.6633255327064091},"full_scale":{"reference":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":378,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":371,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":49,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9245283018867925,"absolute_difference_cents_median":10.071055685021703,"absolute_difference_cents_p95":32.327805525972565,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00014780031156712295,"log_rms_mae_db":3.50880077061728,"rms_envelope_nrmse":0.19080483810593907,"envelope_correlation":0.9715848781730303,"rms_gain_error_db":-0.6632951136835654,"reference_mean_power":0.0022162793576100603,"output_mean_power":0.0019023700065916337},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":1.2692207138635437e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044041649987677594,"output_power":0.00041651678322517327,"reference_active":true,"delta_db":-0.24231075465102303},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188839087946517,"output_power":8.465516765018077e-05,"reference_active":true,"delta_db":-1.4746960173635237},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.435323106110239e-05,"output_power":1.4788947591551824e-05,"reference_active":true,"delta_db":-2.166193194153699}],"active_band_count":3,"active_band_mae_db":1.2943999887227486}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00013486753957748745,"log_rms_mae_db":3.085275972009992,"rms_envelope_nrmse":0.13759645460988276,"envelope_correlation":0.9870812094795895,"rms_gain_error_db":-0.6631514045600313,"reference_mean_power":0.0022339157834068563,"output_mean_power":0.001917571896388312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.176190391785113e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004109863651179866,"output_power":0.00038873255425427115,"reference_active":true,"delta_db":-0.24176501988910953},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3981406650238314e-05,"output_power":3.845051942880133e-05,"reference_active":true,"delta_db":-1.4734198616610357},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0096166790138942e-06,"output_power":7.826469595760768e-07,"reference_active":true,"delta_db":-1.1059061413086153}],"active_band_count":3,"active_band_mae_db":0.9403636742862534}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00012936155891805085,"log_rms_mae_db":2.001424420220575,"rms_envelope_nrmse":0.1961030940702195,"envelope_correlation":0.9730093898705766,"rms_gain_error_db":-0.4208741005837924,"reference_mean_power":0.0012072233727869219,"output_mean_power":0.001095721605233513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":8.078028680738071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004104434555454037,"output_power":0.00037621870037846073,"reference_active":true,"delta_db":-0.3781295644968794},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.258149083667088e-05,"output_power":7.480855405300289e-05,"reference_active":true,"delta_db":-0.9257290942637041},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4118213823838002e-05,"output_power":1.2820034300809444e-05,"reference_active":true,"delta_db":-0.4189056794014828}],"active_band_count":3,"active_band_mae_db":0.5742547793873554}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00011150674016193665,"log_rms_mae_db":1.3551821164253943,"rms_envelope_nrmse":0.1325205251568653,"envelope_correlation":0.9878706529657937,"rms_gain_error_db":-0.4208725286792289,"reference_mean_power":0.0012168889554241998,"output_mean_power":0.0011044948534724276},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.50453906071102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003852443549284888,"output_power":0.0003524383950366963,"reference_active":true,"delta_db":-0.3865306877455129},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.764413089442773e-05,"output_power":3.789133361160532e-05,"reference_active":true,"delta_db":-0.9946951828878423},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0419619324101076e-06,"output_power":9.538746888504568e-07,"reference_active":true,"delta_db":-0.38360527610725115}],"active_band_count":3,"active_band_mae_db":0.5882770489135355}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":2.0674729062147033,"rms_envelope_nrmse":0.21119946933477982,"envelope_correlation":0.9664094465160199,"rms_gain_error_db":-0.16359811796585327,"reference_mean_power":0.0003662329710023158,"output_mean_power":0.0003526936400596643},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":2.347928764955057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015936395121522867,"output_power":0.000149167050167439,"reference_active":true,"delta_db":-0.28717187479403394},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4803443060760012e-05,"output_power":2.3197116149791213e-05,"reference_active":true,"delta_db":-0.2907797404569719},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.158977700048839e-06,"output_power":7.300998865720747e-06,"reference_active":true,"delta_db":0.08531271360368076}],"active_band_count":3,"active_band_mae_db":0.22108810961822886}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":1.37191507780624,"rms_envelope_nrmse":0.12761591532416305,"envelope_correlation":0.9870293782637514,"rms_gain_error_db":-0.16363092949031724,"reference_mean_power":0.0003691513253605656,"output_mean_power":0.0003555014193932129},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":2.1457973458943643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015227225639161785,"output_power":0.0001431142219283773,"reference_active":true,"delta_db":-0.269379894459421},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5037342052746154e-05,"output_power":1.4292802496297965e-05,"reference_active":true,"delta_db":-0.22053686190561317},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0212382710455385e-07,"output_power":3.8715839358737434e-07,"reference_active":true,"delta_db":-0.16471127984027006}],"active_band_count":3,"active_band_mae_db":0.2182093454017681}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00011110909618382604,"log_rms_mae_db":1.9266832547611386,"rms_envelope_nrmse":0.3091000726403734,"envelope_correlation":0.9510594589922939,"rms_gain_error_db":-1.750239039622248,"reference_mean_power":0.00024441441718797425,"output_mean_power":0.0001633438982499046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":3.4696433998856625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001104677897223138,"output_power":5.9937318802284145e-05,"reference_active":true,"delta_db":-2.655383529891124},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9627372395634915e-05,"output_power":3.061348133126945e-05,"reference_active":true,"delta_db":-2.895329517821661},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.408768678752179e-06,"output_power":4.3525629506577195e-06,"reference_active":true,"delta_db":-0.055722516769240615}],"active_band_count":3,"active_band_mae_db":1.868811854827342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":1.247405385655262,"rms_envelope_nrmse":0.252158353310596,"envelope_correlation":0.9757127236818832,"rms_gain_error_db":-1.7502981214979318,"reference_mean_power":0.0002463642033912754,"output_mean_power":0.00016464471441692468},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.3341266023063217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010489106418058683,"output_power":5.5872425663252696e-05,"reference_active":true,"delta_db":-2.7354096523469935},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.086235285110036e-05,"output_power":1.59973222737387e-05,"reference_active":true,"delta_db":-2.853817383036854},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7476114035639134e-07,"output_power":3.4359750073567153e-07,"reference_active":true,"delta_db":-0.3770455559616088}],"active_band_count":3,"active_band_mae_db":1.9887575304484857}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8474843502044678,"predicted_mos":2.8474843502044678,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90960,"input_sample_rate":24000,"input_seconds":3.79,"input_float32_sha256":"c16c2d6523e3d0a5051d4bdc832df3ce34701499c4a920d39d0e7e68b05b6399","internal_samples":60640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.240147630998763},"started_at":"2026-10-13T05:10:03.098701+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0023260059988388093,"native_input_samples":90720,"native_input_sha256":"bdae00d9000f0fc06aec3a95a1c29a3b3e09b961f4b106ec43cb30b42cbb3877","output_sha256":"cf1e42caeb47c1b0825ca19acfea930ebf85d5fe6a2caf86331a3e02d5726fb3","output_samples":90960,"length_delta_samples":240,"analysis_s":0.2484592930013605,"synthesis_s":0.06475298699842824,"total_s":0.31321227999978873,"rtf":0.08286039153433565,"output_peak":0.3934060335159302,"output_rms":0.04442236410783059,"finished_at":"2026-10-13T05:10:03.419459+00:00","analysis_stages":{"external_f0_s":0.05133224999917729},"evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"cf1e42caeb47c1b0825ca19acfea930ebf85d5fe6a2caf86331a3e02d5726fb3","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60480,"output_samples":60640,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.0768870758876,"output_peak":0.3918467567604899},"common_prefix":{"samples":60480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973614775725593,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.232920371183508,"spectral_convergence":0.33289561616347085,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.952432276933774,"spectral_convergence":0.3494359815229601,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.601211827154709,"spectral_convergence":0.38928496999246504,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":4.0763161532637576e-05,"delta_db":-0.54719242493498,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.000535190870328423,"delta_db":-0.22824267090749387,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0006698045393930615,"delta_db":-0.6748077606277446,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.00026715173584570965,"delta_db":-0.24047189421966103,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00017355971008161687,"delta_db":-0.3033141905532209,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.0002789590966580255,"delta_db":-1.0123435236951641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5010620774897108,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5159680006065185,"pitch":null,"output_rms_dbfs":-27.065426150458837,"output_peak":0.3918467567604899,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.262188158423997},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60480,"output_samples":60480,"intelligibility":{"estoi":{"value":0.7568018999829196,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047046475372090966,"rms_dbfs":-26.54945814985232,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.546677348867952,"mean":-0.00027517540745003514,"zero_energy":false},"output":{"rms":0.044333160431436355,"rms_dbfs":-27.065426150458837,"peak":0.3918467567604899,"peak_dbfs":-8.137674869045426,"crest_factor_db":18.927751281413414,"mean":2.0969112870947065e-06,"zero_energy":false},"rms_difference_db":-0.5159680006065166},"full_scale":{"reference":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":378,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":371,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":51,"both_estimated_voiced_frames":48,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9056603773584906,"absolute_difference_cents_median":0.976319755592711,"absolute_difference_cents_p95":9.982532298845053,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00014780031156712295,"log_rms_mae_db":1.219345215341339,"rms_envelope_nrmse":0.1186848293106374,"envelope_correlation":0.9899560060303554,"rms_gain_error_db":-0.5159891767569296,"reference_mean_power":0.0022162793576100603,"output_mean_power":0.001968002219036121},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":1.2692207138635437e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044041649987677594,"output_power":0.0004039452019517042,"reference_active":true,"delta_db":-0.37541127162679255},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188839087946517,"output_power":9.006266179604194e-05,"reference_active":true,"delta_db":-1.2057829749687066},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.435323106110239e-05,"output_power":1.6737202480588515e-05,"reference_active":true,"delta_db":-1.6287371922695313}],"active_band_count":3,"active_band_mae_db":1.0699771462883436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00013486753957748745,"log_rms_mae_db":0.7645679609030033,"rms_envelope_nrmse":0.09321546464619498,"envelope_correlation":0.9946689714831423,"rms_gain_error_db":-0.5160395954004461,"reference_mean_power":0.0022339157834068563,"output_mean_power":0.001983639907771354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.176190391785113e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004109863651179866,"output_power":0.0003749838043654984,"reference_active":true,"delta_db":-0.3981490310623972},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3981406650238314e-05,"output_power":4.075467842969817e-05,"reference_active":true,"delta_db":-1.2206672650132169},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0096166790138942e-06,"output_power":7.298278206921579e-07,"reference_active":true,"delta_db":-1.4093610212221654}],"active_band_count":3,"active_band_mae_db":1.00939243909926}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00012936155891805085,"log_rms_mae_db":1.129813173946647,"rms_envelope_nrmse":0.14645713699764332,"envelope_correlation":0.9862286498406614,"rms_gain_error_db":-0.4968372639669084,"reference_mean_power":0.0012072233727869219,"output_mean_power":0.0010767227981091068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":8.078028680738071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004104434555454037,"output_power":0.00038350663627316283,"reference_active":true,"delta_db":-0.2948045183590661},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.258149083667088e-05,"output_power":6.768716199590522e-05,"reference_active":true,"delta_db":-1.3601786466033978},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4118213823838002e-05,"output_power":1.1208336657285906e-05,"reference_active":true,"delta_db":-1.0023858802123617}],"active_band_count":3,"active_band_mae_db":0.8857896817249419}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00011150674016193665,"log_rms_mae_db":0.6789622115644076,"rms_envelope_nrmse":0.11072077023497001,"envelope_correlation":0.9926982181843865,"rms_gain_error_db":-0.4968366111191818,"reference_mean_power":0.0012168889554241998,"output_mean_power":0.0010853436965933307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.50453906071102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003852443549284888,"output_power":0.0003559062808687465,"reference_active":true,"delta_db":-0.34400631324327385},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.764413089442773e-05,"output_power":3.4205048587779974e-05,"reference_active":true,"delta_db":-1.4391919745634878},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0419619324101076e-06,"output_power":7.105254132321844e-07,"reference_active":true,"delta_db":-1.6627223663599724}],"active_band_count":3,"active_band_mae_db":1.1486402180555781}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":1.007200297413434,"rms_envelope_nrmse":0.10772209316486624,"envelope_correlation":0.9922010234813698,"rms_gain_error_db":-0.34726748638784655,"reference_mean_power":0.0003662329710023158,"output_mean_power":0.000338088728013645},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":2.347928764955057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015936395121522867,"output_power":0.0001475898567250462,"reference_active":true,"delta_db":-0.33333577814254267},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4803443060760012e-05,"output_power":2.5065825906190753e-05,"reference_active":true,"delta_db":0.045700478783963806},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.158977700048839e-06,"output_power":9.527306738562065e-06,"reference_active":true,"delta_db":1.241191382603767}],"active_band_count":3,"active_band_mae_db":0.5400758798434245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":0.6115246411306866,"rms_envelope_nrmse":0.06296572870709505,"envelope_correlation":0.9979963170899058,"rms_gain_error_db":-0.3472721786782429,"reference_mean_power":0.0003691513253605656,"output_mean_power":0.0003407824447071033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":2.1457973458943643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015227225639161785,"output_power":0.00014287915930245794,"reference_active":true,"delta_db":-0.27651897046790463},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5037342052746154e-05,"output_power":1.570867330699433e-05,"reference_active":true,"delta_db":0.18968429091601155},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0212382710455385e-07,"output_power":6.031573717471217e-07,"reference_active":true,"delta_db":1.760708329251387}],"active_band_count":3,"active_band_mae_db":0.742303863545101}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00011110909618382604,"log_rms_mae_db":0.9033617724574902,"rms_envelope_nrmse":0.11657059226669254,"envelope_correlation":0.9930281569356185,"rms_gain_error_db":-0.5119064492945403,"reference_mean_power":0.00024441441718797425,"output_mean_power":0.00021723818966113093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":3.4696433998856625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001104677897223138,"output_power":9.118120504268397e-05,"reference_active":true,"delta_db":-0.8333033708037971},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9627372395634915e-05,"output_power":4.5899533043075786e-05,"reference_active":true,"delta_db":-1.136374043929376},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.408768678752179e-06,"output_power":2.8380932131791863e-06,"reference_active":true,"delta_db":-1.9129065751985999}],"active_band_count":3,"active_band_mae_db":1.294194663310591}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":0.533078516549221,"rms_envelope_nrmse":0.08661104620442035,"envelope_correlation":0.9970588589590763,"rms_gain_error_db":-0.5119209275217398,"reference_mean_power":0.0002463642033912754,"output_mean_power":0.00021897045083318892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.3341266023063217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010489106418058683,"output_power":8.493447297411384e-05,"reference_active":true,"delta_db":-0.9165449530779484},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.086235285110036e-05,"output_power":2.4472973677069055e-05,"reference_active":true,"delta_db":-1.0074228919033157},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7476114035639134e-07,"output_power":2.7056063218413996e-07,"reference_active":true,"delta_db":-1.4148994677478512}],"active_band_count":3,"active_band_mae_db":1.112955770909705}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1221-135767-0019","method":"magphase","status":"measured","mos":{"status":"measured","value":3.6430115699768066,"predicted_mos":3.6430115699768066,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":181587,"input_sample_rate":48000,"input_seconds":3.7830625,"input_float32_sha256":"9c96688cc311dcc66d2356777b26cad9e7889f208d04d4ddb91c807ec010b98c","internal_samples":60529,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.264694569999847},"started_at":"2026-10-13T05:12:31.277599+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006661235000137822,"native_input_samples":181440,"native_input_sha256":"1f8a1aedda98bdc5e38596118fadca2876e122fcf77ec799792552035fb251c0","output_sha256":"337baee31ae3bc37323e0a4bc907c0a6cf539afad66958389bdf2ede396a210d","output_samples":181587,"length_delta_samples":147,"analysis_s":2.4401657180005714,"synthesis_s":1.0417333690002124,"total_s":3.4818990870007838,"rtf":0.921137324603382,"output_peak":0.38599615377351876,"output_rms":0.045904068243984866,"finished_at":"2026-10-13T05:12:34.783950+00:00","evaluation_input":{"reference":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"output_sha256":"337baee31ae3bc37323e0a4bc907c0a6cf539afad66958389bdf2ede396a210d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":60480,"output_samples":60529,"length_delta_samples":49,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.80637338944897,"output_peak":0.3857645593831585},"common_prefix":{"samples":60480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.999190470683474,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60480,"output_samples":60480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.671469085103202,"spectral_convergence":0.33790213804380437,"spectral_convergence_status":"ok","amplitude_floor":9.224568988143785e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.255114631814341,"spectral_convergence":0.26334583863480043,"spectral_convergence_status":"ok","amplitude_floor":8.440100575569802e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.539139864415905,"spectral_convergence":0.27954775326509335,"spectral_convergence_status":"ok","amplitude_floor":6.303303867538867e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.6236728988818796e-05,"output_power":3.658519500982853e-05,"delta_db":-1.016817281017737,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005640699493062965,"output_power":0.0005138571696263579,"delta_db":-0.40490542976759103,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007824000374604723,"output_power":0.0007206090291032468,"delta_db":-0.3572916311967373,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028236131111099825,"output_power":0.00028007567021429993,"delta_db":-0.035298059210962254,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018611455363091946,"output_power":0.00019756235379152478,"delta_db":0.25923856721763594,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00035218826443925665,"output_power":0.00033923183388369187,"delta_db":-0.16278280218590277,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3727222950994276,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2534007029562751,"pitch":null,"output_rms_dbfs":-26.802858852808598,"output_peak":0.3857645593831585,"reference_rms_dbfs":-26.54945814985232,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.48857452711115},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60480,"output_samples":60480,"intelligibility":{"estoi":{"value":0.7544147078308037,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047046475372090966,"rms_dbfs":-26.54945814985232,"peak":0.397979736328125,"peak_dbfs":-8.00278080098437,"crest_factor_db":18.546677348867952,"mean":-0.00027517540745003514,"zero_energy":false},"output":{"rms":0.04569377694641744,"rms_dbfs":-26.802858852808598,"peak":0.3857645593831585,"peak_dbfs":-8.273553479479412,"crest_factor_db":18.529305373329183,"mean":0.0005219221075801345,"zero_energy":false},"rms_difference_db":-0.25340070295627726},"full_scale":{"reference":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":378,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.184493209933862e-06,"threshold_dbfs":-56.606493008710245,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":371,"reference_estimated_voiced_frames":53,"output_estimated_voiced_frames":53,"both_estimated_voiced_frames":51,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9622641509433962,"absolute_difference_cents_median":3.5321077071318534,"absolute_difference_cents_p95":9.041996278516905,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00014780031156712295,"log_rms_mae_db":1.6567630461276053,"rms_envelope_nrmse":0.08922879048913714,"envelope_correlation":0.9936741726521903,"rms_gain_error_db":-0.2533630044143529,"reference_mean_power":0.0022162793576100603,"output_mean_power":0.0020906830633175087},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":1.2692207138635437e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044041649987677594,"output_power":0.0004288523686784897,"reference_active":true,"delta_db":-0.11555767967241261},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188839087946517,"output_power":0.00011558765375530333,"reference_active":true,"delta_db":-0.12211627520272385},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.435323106110239e-05,"output_power":2.559424809810061e-05,"reference_active":true,"delta_db":0.21585786145394856}],"active_band_count":3,"active_band_mae_db":0.151177272109695}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00013486753957748745,"log_rms_mae_db":1.2121536002366455,"rms_envelope_nrmse":0.055007572843489415,"envelope_correlation":0.997938965333071,"rms_gain_error_db":-0.2532555643425834,"reference_mean_power":0.0022339157834068563,"output_mean_power":0.0021073721685244975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.176190391785113e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004109863651179866,"output_power":0.000397185709018105,"reference_active":true,"delta_db":-0.14833800036581027},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3981406650238314e-05,"output_power":5.607992937540123e-05,"reference_active":true,"delta_db":0.1656326079192768},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0096166790138942e-06,"output_power":1.3251834590556135e-06,"reference_active":true,"delta_db":1.181194898114519}],"active_band_count":3,"active_band_mae_db":0.498388502133202}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00012936155891805085,"log_rms_mae_db":1.4895224081276806,"rms_envelope_nrmse":0.0659812954046674,"envelope_correlation":0.9981966223470657,"rms_gain_error_db":-0.38779422705187094,"reference_mean_power":0.0012072233727869219,"output_mean_power":0.001104099498010832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":8.078028680738071e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004104434555454037,"output_power":0.0003704761042422953,"reference_active":true,"delta_db":-0.4449313413461832},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.258149083667088e-05,"output_power":8.3988371674775e-05,"reference_active":true,"delta_db":-0.42305008478703465},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4118213823838002e-05,"output_power":1.2811964585719256e-05,"reference_active":true,"delta_db":-0.421640255714087}],"active_band_count":3,"active_band_mae_db":0.4298738939491016}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00011150674016193665,"log_rms_mae_db":1.1122368531429156,"rms_envelope_nrmse":0.05243887927668174,"envelope_correlation":0.999341797868731,"rms_gain_error_db":-0.38779723142512956,"reference_mean_power":0.0012168889554241998,"output_mean_power":0.0011129386538148618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.50453906071102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003852443549284888,"output_power":0.0003469740950549594,"reference_active":true,"delta_db":-0.45439231871573205},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.764413089442773e-05,"output_power":4.3559253142961583e-05,"reference_active":true,"delta_db":-0.38928984293175295},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0419619324101076e-06,"output_power":1.0117570509930026e-06,"reference_active":true,"delta_db":-0.12775612811049714}],"active_band_count":3,"active_band_mae_db":0.3238127632526607}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":7.756656945350893e-05,"log_rms_mae_db":1.0432844987040297,"rms_envelope_nrmse":0.09026584441494664,"envelope_correlation":0.9939625401616874,"rms_gain_error_db":0.042091942621698464,"reference_mean_power":0.0003662329710023158,"output_mean_power":0.000369799768025299},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":2.347928764955057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015936395121522867,"output_power":0.0001626111271665645,"reference_active":true,"delta_db":0.08760171248908419},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4803443060760012e-05,"output_power":2.5433301933553367e-05,"reference_active":true,"delta_db":0.10890775970257532},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.158977700048839e-06,"output_power":6.637602819764066e-06,"reference_active":true,"delta_db":-0.3283974802983082}],"active_band_count":3,"active_band_mae_db":0.17496898416332254}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":6.756499548428438e-05,"log_rms_mae_db":0.5672562036214712,"rms_envelope_nrmse":0.043910486392120716,"envelope_correlation":0.998476497525964,"rms_gain_error_db":0.04204985463752814,"reference_mean_power":0.0003691513253605656,"output_mean_power":0.0003727429323552066},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":2.1457973458943643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015227225639161785,"output_power":0.00015511369060963251,"reference_active":true,"delta_db":0.08029347886602853},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5037342052746154e-05,"output_power":1.54108145151788e-05,"reference_active":true,"delta_db":0.10654514613157903},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0212382710455385e-07,"output_power":3.846826269170374e-07,"reference_active":true,"delta_db":-0.1925723409242024}],"active_band_count":3,"active_band_mae_db":0.12647032197393665}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":755,"rms_floor":0.00011110909618382604,"log_rms_mae_db":1.3567860342309517,"rms_envelope_nrmse":0.15972281212644263,"envelope_correlation":0.9854156913002726,"rms_gain_error_db":0.34555488788288036,"reference_mean_power":0.00024441441718797425,"output_mean_power":0.00026465634393487786},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":755,"frequency_resolution_hz":0.26490066225165565,"power_floor":3.4696433998856625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001104677897223138,"output_power":0.00011554366386307575,"reference_active":true,"delta_db":0.19510470220698406},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9627372395634915e-05,"output_power":6.033266309803563e-05,"reference_active":true,"delta_db":0.05106823924317363},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.408768678752179e-06,"output_power":9.746221809234148e-06,"reference_active":true,"delta_db":3.4451897837492766}],"active_band_count":3,"active_band_mae_db":1.2304542417331448}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60480,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.166729119583492e-05,"log_rms_mae_db":1.0405836489600073,"rms_envelope_nrmse":0.09458280388684324,"envelope_correlation":0.9956867925780459,"rms_gain_error_db":0.34551224047168727,"reference_mean_power":0.0002463642033912754,"output_mean_power":0.00026676498801382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.3341266023063217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010489106418058683,"output_power":0.00011155921866355255,"reference_active":true,"delta_db":0.2676697229791034},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.086235285110036e-05,"output_power":3.301337354422079e-05,"reference_active":true,"delta_db":0.2926087366703511},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7476114035639134e-07,"output_power":7.110742426717338e-07,"reference_active":true,"delta_db":2.781603956341054}],"active_band_count":3,"active_band_mae_db":1.1139608053301695}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"hnm","status":"measured","mos":{"status":"measured","value":2.409708261489868,"predicted_mos":2.409708261489868,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":93920,"input_sample_rate":16000,"input_seconds":5.87,"input_float32_sha256":"1ce75f7082bb3a1b3a947b33a2ebd1915f02fa7d220da1ad6db01f7a42f95903","internal_samples":93920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3476339429998916},"started_at":"2026-10-13T04:38:11.878592+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.7792000107874628e-05,"native_input_samples":93920,"native_input_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","output_sha256":"9dd80790386e29099ebc2000d951698d7e17cf87f6eea57f844ded38be798500","output_samples":93920,"length_delta_samples":0,"analysis_s":1.2442487620000975,"synthesis_s":0.5918235159997494,"total_s":1.8360722779998468,"rtf":0.312789144463347,"output_peak":0.41950459170156323,"output_rms":0.06015946016925701,"finished_at":"2026-10-13T04:38:13.722457+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"9dd80790386e29099ebc2000d951698d7e17cf87f6eea57f844ded38be798500","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.659176132989015,"spectral_convergence":0.2769921064064206,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.683725806540736,"spectral_convergence":0.27872335467335957,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9332025181089705,"spectral_convergence":0.28987934359995415,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.00035827132734213215,"delta_db":-0.4241470267182387,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.001471698153024178,"delta_db":0.05282591736069978,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.0015718572667708032,"delta_db":-0.3157481611911884,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":6.870071924872539e-05,"delta_db":0.2537045750686506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":9.561676245130782e-05,"delta_db":-0.08636974436581034,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":5.30164190192738e-05,"delta_db":0.30930414271578627,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24034992790339568,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15543597272188547,"pitch":null,"output_rms_dbfs":-24.41392138934049,"output_peak":0.41950459170156323,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.758701485879574},"common_prefix":{"samples":93920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.659176132989015,"spectral_convergence":0.2769921064064206,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.683725806540736,"spectral_convergence":0.27872335467335957,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9332025181089705,"spectral_convergence":0.28987934359995415,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.00035827132734213215,"delta_db":-0.4241470267182387,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.001471698153024178,"delta_db":0.05282591736069978,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.0015718572667708032,"delta_db":-0.3157481611911884,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":6.870071924872539e-05,"delta_db":0.2537045750686506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":9.561676245130782e-05,"delta_db":-0.08636974436581034,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":5.30164190192738e-05,"delta_db":0.30930414271578627,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24034992790339568,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15543597272188547,"pitch":null,"output_rms_dbfs":-24.41392138934049,"output_peak":0.41950459170156323,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.758701485879574},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93920,"output_samples":93920,"intelligibility":{"estoi":{"value":0.8733607000562423,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.061245717832140306,"rms_dbfs":-24.258485416618605,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.058636746461097,"mean":-2.66944326371513e-05,"zero_energy":false},"output":{"rms":0.06015946016925701,"rms_dbfs":-24.41392138934049,"peak":0.41950459170156323,"peak_dbfs":-7.54526562448927,"crest_factor_db":16.86865576485122,"mean":0.00011725257454280848,"zero_energy":false},"rms_difference_db":-0.15543597272188592},"full_scale":{"reference":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":587,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":2.3024710544356792e-06,"output_mean_power":3.1387865656345926e-06,"output_rms_dbfs":-55.032382148888715,"output_zero_energy":false,"output_peak":0.01148047703411587,"output_energy_fraction":4.875619021084611e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":580,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":249,"both_estimated_voiced_frames":244,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9242424242424242,"absolute_difference_cents_median":2.094315977759755,"absolute_difference_cents_p95":11.578143638361365,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002156956510733609,"log_rms_mae_db":0.8855576389296083,"rms_envelope_nrmse":0.08315659193044433,"envelope_correlation":0.9922411464246534,"rms_gain_error_db":-0.15543864506708296,"reference_mean_power":0.003754233895192234,"output_mean_power":0.0036222419998930256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.680528929094005e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006742966926403748,"output_power":0.0006452238690503899,"reference_active":true,"delta_db":-0.19140604448820892},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002358850904334285,"output_power":0.0002238762678500127,"reference_active":true,"delta_db":-0.2269243299410682},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0118677140216256e-05,"output_power":2.724157087904812e-05,"reference_active":true,"delta_db":-0.4360374559291036}],"active_band_count":3,"active_band_mae_db":0.2847892767861269}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00016458512240675186,"log_rms_mae_db":0.46413762142361936,"rms_envelope_nrmse":0.04958671426599595,"envelope_correlation":0.9971623920322135,"rms_gain_error_db":-0.15544123727262196,"reference_mean_power":0.003773515049488898,"output_mean_power":0.0036408430914559647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.4929366354790985e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006129787155244603,"output_power":0.0005839561236799011,"reference_active":true,"delta_db":-0.21065177717209307},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073825851363778,"output_power":0.00010595169852984462,"reference_active":true,"delta_db":-0.1918976609793019},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6089511471242587e-06,"output_power":2.0897384120225587e-06,"reference_active":true,"delta_db":-0.9637402129966094}],"active_band_count":3,"active_band_mae_db":0.4554298837160015}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002116840322083979,"log_rms_mae_db":1.4129666607894311,"rms_envelope_nrmse":0.09794882694478362,"envelope_correlation":0.9917668819181602,"rms_gain_error_db":-0.1956149311310336,"reference_mean_power":0.002333704816993391,"output_mean_power":0.0022309222222451136},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.3033890876287954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006736793990758906,"output_power":0.0006408430431493152,"reference_active":true,"delta_db":-0.21701592881708895},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002058984330594163,"output_power":0.00018627920051052187,"reference_active":true,"delta_db":-0.43488676215349126},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6276513462408744e-05,"output_power":2.0448237326721095e-05,"reference_active":true,"delta_db":-1.0891186268795574}],"active_band_count":3,"active_band_mae_db":0.5803404392833792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00015464611695479593,"log_rms_mae_db":0.8332639234671938,"rms_envelope_nrmse":0.05553389007287161,"envelope_correlation":0.9973761585542485,"rms_gain_error_db":-0.19561522657885896,"reference_mean_power":0.002345702993864749,"output_mean_power":0.002242391814804813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.1648344391379404e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377113304270822,"output_power":0.0006054225799719708,"reference_active":true,"delta_db":-0.22565518844902224},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.690162436721034e-05,"output_power":8.865597906571224e-05,"reference_active":true,"delta_db":-0.3862302694773067},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0235059299775414e-06,"output_power":2.0883948181819006e-06,"reference_active":true,"delta_db":-1.6069821777179192}],"active_band_count":3,"active_band_mae_db":0.7396225452147495}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":1.6652738873729485,"rms_envelope_nrmse":0.19928754112359703,"envelope_correlation":0.9659455115301117,"rms_gain_error_db":0.14212974481291152,"reference_mean_power":0.00010718143407110891,"output_mean_power":0.00011074714470743533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":8.16723414342307e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.040560289520862e-05,"output_power":3.786212329684522e-05,"reference_active":true,"delta_db":-0.2823662587391225},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.346030743633643e-06,"output_power":7.819645655517506e-06,"reference_active":true,"delta_db":-0.28292906551336755},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1788672389317663e-06,"output_power":2.5905115675312795e-06,"reference_active":true,"delta_db":0.751547669487581}],"active_band_count":3,"active_band_mae_db":0.43894766458002366}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":1.15690608810246,"rms_envelope_nrmse":0.1329168671982429,"envelope_correlation":0.9833665969318686,"rms_gain_error_db":0.14212120287271612,"reference_mean_power":0.00010773182378082651,"output_mean_power":0.0001113156258334312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":7.38064524776292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.729428782271285e-05,"output_power":3.5036539571451654e-05,"reference_active":true,"delta_db":-0.2712111224955381},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.87881666371459e-06,"output_power":3.869301982746318e-06,"reference_active":true,"delta_db":-0.010666268050914626},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8761762452220107e-07,"output_power":1.7258187375564984e-07,"reference_active":true,"delta_db":-0.36278453040655956}],"active_band_count":3,"active_band_mae_db":0.21488730698433742}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":1.7373219642242765,"rms_envelope_nrmse":0.18130711295404048,"envelope_correlation":0.9716234781610581,"rms_gain_error_db":0.17460556245651443,"reference_mean_power":7.041944137047503e-05,"output_mean_power":7.330829718695291e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":5.931095881530664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9173785649337017e-05,"output_power":3.051155458399484e-05,"reference_active":true,"delta_db":0.1947154809140863},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1741151998348046e-05,"output_power":1.2086587457605784e-05,"reference_active":true,"delta_db":0.12592988546057576},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.929169683988833e-06,"output_power":2.5140031880889254e-06,"reference_active":true,"delta_db":1.1499539552706104}],"active_band_count":3,"active_band_mae_db":0.49019977388175745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":1.0567426948832028,"rms_envelope_nrmse":0.10301436794005404,"envelope_correlation":0.9899465868441275,"rms_gain_error_db":0.17454616878391602,"reference_mean_power":7.077979037597614e-05,"output_mean_power":7.368242130944443e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":5.2532775502110204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7931175609497946e-05,"output_power":2.9915369699080384e-05,"reference_active":true,"delta_db":0.2980515896649568},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173764837074772e-06,"output_power":5.8152173055258645e-06,"reference_active":true,"delta_db":0.5075926341003124},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7085501560216017e-07,"output_power":1.9217194540558102e-07,"reference_active":true,"delta_db":0.5106225421134409}],"active_band_count":3,"active_band_mae_db":0.4387555886262367}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.069908857345581,"predicted_mos":3.069908857345581,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":93920,"input_sample_rate":16000,"input_seconds":5.87,"input_float32_sha256":"ecbbf165adf021e1ca22a12451e0426c228ff848d344b4dc763f83d87608dea3","internal_samples":93920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2826015270002245},"started_at":"2026-10-13T04:41:03.860780+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.064500015170779e-05,"native_input_samples":93920,"native_input_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","output_sha256":"a6a8ff93d474f40fc4c0fb01d26afbb9a03c694d4bbe19a1682ca2df2c0c58d0","output_samples":93920,"length_delta_samples":0,"analysis_s":2.0119284079992212,"synthesis_s":0.12846617300056096,"total_s":2.140394580999782,"rtf":0.3646328076660617,"output_peak":0.5277819530168999,"output_rms":0.06967414225057021,"finished_at":"2026-10-13T04:41:06.082625+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"a6a8ff93d474f40fc4c0fb01d26afbb9a03c694d4bbe19a1682ca2df2c0c58d0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.830617254558779,"spectral_convergence":0.48837556502810103,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.398584253583114,"spectral_convergence":0.3127900436553969,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.830449943791972,"spectral_convergence":0.33460376804747144,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.0003685556636156285,"delta_db":-0.3012366597975658,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.0017404770604597926,"delta_db":0.7813215054023325,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.002501189685133266,"delta_db":1.7015870568797604,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":7.475619147158976e-05,"delta_db":0.6205634136921881,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":0.00011649419636188102,"delta_db":0.7713328074032245,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":5.3013301310534956e-05,"delta_db":0.30904874192860243,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7475150308506123,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1199180341102424,"pitch":null,"output_rms_dbfs":-23.138567382508363,"output_peak":0.5277819530168999,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.686550483977956},"common_prefix":{"samples":93920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.830617254558779,"spectral_convergence":0.48837556502810103,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.398584253583114,"spectral_convergence":0.3127900436553969,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.830449943791972,"spectral_convergence":0.33460376804747144,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.0003685556636156285,"delta_db":-0.3012366597975658,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.0017404770604597926,"delta_db":0.7813215054023325,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.002501189685133266,"delta_db":1.7015870568797604,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":7.475619147158976e-05,"delta_db":0.6205634136921881,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":0.00011649419636188102,"delta_db":0.7713328074032245,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":5.3013301310534956e-05,"delta_db":0.30904874192860243,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7475150308506123,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1199180341102424,"pitch":null,"output_rms_dbfs":-23.138567382508363,"output_peak":0.5277819530168999,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.686550483977956},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93920,"output_samples":93920,"intelligibility":{"estoi":{"value":0.9341290518082109,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.061245717832140306,"rms_dbfs":-24.258485416618605,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.058636746461097,"mean":-2.66944326371513e-05,"zero_energy":false},"output":{"rms":0.06967414225057021,"rms_dbfs":-23.138567382508363,"peak":0.5277819530168999,"peak_dbfs":-5.550909282667057,"crest_factor_db":17.587658099841306,"mean":-2.8239770058781397e-09,"zero_energy":false},"rms_difference_db":1.119918034110242},"full_scale":{"reference":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":587,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":2.3024710544356792e-06,"output_mean_power":3.6427685323364744e-06,"output_rms_dbfs":-54.385684236860854,"output_zero_energy":false,"output_peak":0.01479844521328246,"output_energy_fraction":4.21855925008201e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":580,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":293,"both_estimated_voiced_frames":252,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.9545454545454546,"absolute_difference_cents_median":5.57559682193569,"absolute_difference_cents_p95":26.366039587510688,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002156956510733609,"log_rms_mae_db":1.7952381379666897,"rms_envelope_nrmse":0.1911264967473382,"envelope_correlation":0.9829056537211508,"rms_gain_error_db":1.1199186569042903,"reference_mean_power":0.003754233895192234,"output_mean_power":0.004858622892421075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.680528929094005e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006742966926403748,"output_power":0.0008824009640231277,"reference_active":true,"delta_db":1.1681494438943643},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002358850904334285,"output_power":0.00033075604341790315,"reference_active":true,"delta_db":1.4680729681238902},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0118677140216256e-05,"output_power":3.708758924137918e-05,"reference_active":true,"delta_db":0.9039271129063062}],"active_band_count":3,"active_band_mae_db":1.1800498416415202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00016458512240675186,"log_rms_mae_db":1.3659210015699323,"rms_envelope_nrmse":0.16921533054168705,"envelope_correlation":0.9895802883874519,"rms_gain_error_db":1.1199381781160702,"reference_mean_power":0.003773515049488898,"output_mean_power":0.004883597966113173},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.4929366354790985e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006129787155244603,"output_power":0.0008191425280189615,"reference_active":true,"delta_db":1.2591407933886285},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073825851363778,"output_power":0.00016161013070240874,"reference_active":true,"delta_db":1.6417089205215074},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6089511471242587e-06,"output_power":3.095408825769411e-06,"reference_active":true,"delta_db":0.7425206957076723}],"active_band_count":3,"active_band_mae_db":1.214456803205936}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002116840322083979,"log_rms_mae_db":1.7872772545260207,"rms_envelope_nrmse":0.24129044573287717,"envelope_correlation":0.9868274625856089,"rms_gain_error_db":1.582181930660549,"reference_mean_power":0.002333704816993391,"output_mean_power":0.003359418547942054},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.3033890876287954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006736793990758906,"output_power":0.0009439809966975787,"reference_active":true,"delta_db":1.4650998464000689},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002058984330594163,"output_power":0.00031827747017384387,"reference_active":true,"delta_db":1.8915285587949353},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6276513462408744e-05,"output_power":3.3246654619963345e-05,"reference_active":true,"delta_db":1.0218021208661323}],"active_band_count":3,"active_band_mae_db":1.459476842020379}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00015464611695479593,"log_rms_mae_db":1.5740636486439563,"rms_envelope_nrmse":0.2210271906248783,"envelope_correlation":0.9928716737706409,"rms_gain_error_db":1.5821817060744594,"reference_mean_power":0.002345702993864749,"output_mean_power":0.003376690008377868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.1648344391379404e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377113304270822,"output_power":0.0009028857687413153,"reference_active":true,"delta_db":1.5100867443167432},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.690162436721034e-05,"output_power":0.00015146448171440483,"reference_active":true,"delta_db":1.9397974589562557},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0235059299775414e-06,"output_power":3.7032289019040066e-06,"reference_active":true,"delta_db":0.8806973252457861}],"active_band_count":3,"active_band_mae_db":1.4435271761729283}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":1.3923239026155916,"rms_envelope_nrmse":0.15324069284141426,"envelope_correlation":0.9903873743550478,"rms_gain_error_db":0.8765966345786012,"reference_mean_power":0.00010718143407110891,"output_mean_power":0.0001311533013525675},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":8.16723414342307e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.040560289520862e-05,"output_power":5.0067169434402245e-05,"reference_active":true,"delta_db":0.9311144780409699},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.346030743633643e-06,"output_power":1.0973581009459983e-05,"reference_active":true,"delta_db":1.1886839389683208},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1788672389317663e-06,"output_power":2.4580774938285158e-06,"reference_active":true,"delta_db":0.5236480146447882}],"active_band_count":3,"active_band_mae_db":0.8811488105513595}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":1.040030463230099,"rms_envelope_nrmse":0.13059160282836532,"envelope_correlation":0.9949768750859848,"rms_gain_error_db":0.876600436159656,"reference_mean_power":0.00010773182378082651,"output_mean_power":0.00013182690491270456},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":7.38064524776292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.729428782271285e-05,"output_power":4.654847707161063e-05,"reference_active":true,"delta_db":0.9626315850494703},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.87881666371459e-06,"output_power":5.070449429466212e-06,"reference_active":true,"delta_db":1.1634720289604212},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8761762452220107e-07,"output_power":2.013330033071451e-07,"reference_active":true,"delta_db":0.3064133906333173}],"active_band_count":3,"active_band_mae_db":0.8108390015477362}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":1.6501622057592955,"rms_envelope_nrmse":0.20372941227230928,"envelope_correlation":0.9659824780138528,"rms_gain_error_db":0.3784508751604416,"reference_mean_power":7.041944137047503e-05,"output_mean_power":7.683120826486064e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":5.931095881530664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9173785649337017e-05,"output_power":2.8083744025012033e-05,"reference_active":true,"delta_db":-0.16537781791935696},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1741151998348046e-05,"output_power":1.1732335485308154e-05,"reference_active":true,"delta_db":-0.0032623726198086453},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.929169683988833e-06,"output_power":2.4169414270712746e-06,"reference_active":true,"delta_db":0.9789569713817708}],"active_band_count":3,"active_band_mae_db":0.3825323873069788}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":1.0346309472601232,"rms_envelope_nrmse":0.1223433147156911,"envelope_correlation":0.9874070530113325,"rms_gain_error_db":0.37840391197698153,"reference_mean_power":7.077979037597614e-05,"output_mean_power":7.722353236558283e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":5.2532775502110204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7931175609497946e-05,"output_power":2.6630213479472456e-05,"reference_active":true,"delta_db":-0.20714567379716972},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173764837074772e-06,"output_power":5.213936913720773e-06,"reference_active":true,"delta_db":0.03359087237733265},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7085501560216017e-07,"output_power":1.8074240884933377e-07,"reference_active":true,"delta_db":0.2443233362887043}],"active_band_count":3,"active_band_mae_db":0.16168662748773557}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"facodec","status":"measured","mos":{"status":"measured","value":4.212867259979248,"predicted_mos":4.212867259979248,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":94000,"input_sample_rate":16000,"input_seconds":5.875,"input_float32_sha256":"32b6235209a75fcd726ae0209e39ec2fd8a1c599c2e37e89b10070564ce50af5","internal_samples":94000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2488709740009654},"started_at":"2026-10-13T04:46:08.342945+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.563799989729887e-05,"native_input_samples":93920,"native_input_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","output_sha256":"bb7a4607d0049b7fd66c4405f6adc34aa7e91fbd08ea314e6202647a214abab2","output_samples":94000,"length_delta_samples":80,"analysis_s":2.1992636830000265,"synthesis_s":11.678963106000083,"total_s":13.87822678900011,"rtf":2.36426350749576,"output_peak":0.38714709877967834,"output_rms":0.05832554914613259,"finished_at":"2026-10-13T04:46:22.229371+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"bb7a4607d0049b7fd66c4405f6adc34aa7e91fbd08ea314e6202647a214abab2","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":93920,"output_samples":94000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.682823270997126,"output_peak":0.38714709877967834},"common_prefix":{"samples":93920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991489361702127,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.370633544569803,"spectral_convergence":0.25265451648957393,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.0676008050867685,"spectral_convergence":0.24226254590513738,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.26628700145465,"spectral_convergence":0.2707667532920948,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.0003495298493884311,"delta_db":-0.5314248555279452,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.0014139703178761387,"delta_db":-0.1209586027471197,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.0014724575306615645,"delta_db":-0.5994514579430668,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":4.609715771273818e-05,"delta_db":-1.479166781220131,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":7.996664616405228e-05,"delta_db":-0.862621270929838,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":4.274582722230238e-05,"delta_db":-0.6258624978753085,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7032475777072348,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.42064019379498185,"pitch":null,"output_rms_dbfs":-24.67912561041359,"output_peak":0.38714709877967834,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.901507117037074},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93920,"output_samples":93920,"intelligibility":{"estoi":{"value":0.9036586386568253,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.061245717832140306,"rms_dbfs":-24.258485416618605,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.058636746461097,"mean":-2.66944326371513e-05,"zero_energy":false},"output":{"rms":0.058350384137769204,"rms_dbfs":-24.67912561041359,"peak":0.38714709877967834,"peak_dbfs":-8.242479818475099,"crest_factor_db":16.43664579193849,"mean":4.06916593315153e-05,"zero_energy":false},"rms_difference_db":-0.4206401937949842},"full_scale":{"reference":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":587,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":2.3024710544356792e-06,"output_mean_power":4.118753552578048e-06,"output_rms_dbfs":-53.852341934666185,"output_zero_energy":false,"output_peak":0.00942903757095337,"output_energy_fraction":6.800709223501208e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":580,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":280,"both_estimated_voiced_frames":261,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9886363636363636,"absolute_difference_cents_median":4.19885426260481,"absolute_difference_cents_p95":24.84941379228946,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002156956510733609,"log_rms_mae_db":1.3094019237882775,"rms_envelope_nrmse":0.12730019725046895,"envelope_correlation":0.983011865239849,"rms_gain_error_db":-0.4206404231979046,"reference_mean_power":0.003754233895192234,"output_mean_power":0.0034076680635103476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.680528929094005e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006742966926403748,"output_power":0.0005692422777940421,"reference_active":true,"delta_db":-0.7355388171868462},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002358850904334285,"output_power":0.00023139307687981752,"reference_active":true,"delta_db":-0.08350130322116903},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0118677140216256e-05,"output_power":3.2692700994122956e-05,"reference_active":true,"delta_db":0.3561490940284312}],"active_band_count":3,"active_band_mae_db":0.3917297381454821}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00016458512240675186,"log_rms_mae_db":1.089250240563779,"rms_envelope_nrmse":0.1118883397872987,"envelope_correlation":0.9858913366428969,"rms_gain_error_db":-0.4206383404775782,"reference_mean_power":0.003773515049488898,"output_mean_power":0.0034251709528402147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.4929366354790985e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006129787155244603,"output_power":0.000525015962358725,"reference_active":true,"delta_db":-0.6727288705319777},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073825851363778,"output_power":0.00010821982660562478,"reference_active":true,"delta_db":-0.09990855697684507},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6089511471242587e-06,"output_power":2.869738694379878e-06,"reference_active":true,"delta_db":0.41376406645654756}],"active_band_count":3,"active_band_mae_db":0.3954671646551235}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002116840322083979,"log_rms_mae_db":1.580146053398564,"rms_envelope_nrmse":0.13390535320834746,"envelope_correlation":0.9845402935705256,"rms_gain_error_db":-0.28887266946554285,"reference_mean_power":0.002333704816993391,"output_mean_power":0.002183527448527419},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.3033890876287954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006736793990758906,"output_power":0.0005872122979594786,"reference_active":true,"delta_db":-0.5965812454207331},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002058984330594163,"output_power":0.00020049552656070666,"reference_active":true,"delta_db":-0.11548354414320679},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6276513462408744e-05,"output_power":2.9367501692013123e-05,"reference_active":true,"delta_db":0.48299262734062764}],"active_band_count":3,"active_band_mae_db":0.39835247230152254}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00015464611695479593,"log_rms_mae_db":1.2895525206856988,"rms_envelope_nrmse":0.11306629883672399,"envelope_correlation":0.9879854809499345,"rms_gain_error_db":-0.2888744051719502,"reference_mean_power":0.002345702993864749,"output_mean_power":0.002194752647752813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.1648344391379404e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377113304270822,"output_power":0.0005633192439961329,"reference_active":true,"delta_db":-0.5386954554126776},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.690162436721034e-05,"output_power":9.467963060568257e-05,"reference_active":true,"delta_db":-0.10074502361398967},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0235059299775414e-06,"output_power":3.063613958537326e-06,"reference_active":true,"delta_db":0.05723214881969377}],"active_band_count":3,"active_band_mae_db":0.23222420928212031}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":1.9334420500693086,"rms_envelope_nrmse":0.21889422025182972,"envelope_correlation":0.9733329105057852,"rms_gain_error_db":-1.396523117758937,"reference_mean_power":0.00010718143407110891,"output_mean_power":7.770827201955278e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":8.16723414342307e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.040560289520862e-05,"output_power":2.6456196576743407e-05,"reference_active":true,"delta_db":-1.8391418244878968},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.346030743633643e-06,"output_power":5.50801954810875e-06,"reference_active":true,"delta_db":-1.8048450724556897},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1788672389317663e-06,"output_power":1.9785357892464564e-06,"reference_active":true,"delta_db":-0.418868584349874}],"active_band_count":3,"active_band_mae_db":1.3542851604311534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":1.6321628140543198,"rms_envelope_nrmse":0.19733534691771568,"envelope_correlation":0.9807706210491245,"rms_gain_error_db":-1.3965802951837851,"reference_mean_power":0.00010773182378082651,"output_mean_power":7.810628513277461e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":7.38064524776292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.729428782271285e-05,"output_power":2.454056421856048e-05,"reference_active":true,"delta_db":-1.817577747305559},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.87881666371459e-06,"output_power":2.453511698634896e-06,"reference_active":true,"delta_db":-1.9891111961735484},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8761762452220107e-07,"output_power":1.4093593238872875e-07,"reference_active":true,"delta_db":-1.2425189995450212}],"active_band_count":3,"active_band_mae_db":1.6830693143413764}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":1.9060538382994023,"rms_envelope_nrmse":0.20948421919680382,"envelope_correlation":0.9641817031788608,"rms_gain_error_db":-0.7654551454752744,"reference_mean_power":7.041944137047503e-05,"output_mean_power":5.9040096816569075e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":5.931095881530664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9173785649337017e-05,"output_power":2.1384674603640636e-05,"reference_active":true,"delta_db":-1.348901414343131},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1741151998348046e-05,"output_power":1.0811995101194276e-05,"reference_active":true,"delta_db":-0.35804870115931653},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.929169683988833e-06,"output_power":1.596296190388548e-06,"reference_active":true,"delta_db":-0.8225695137805986}],"active_band_count":3,"active_band_mae_db":0.8431732097610154}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":1.5276299038986085,"rms_envelope_nrmse":0.1645457107178007,"envelope_correlation":0.9777107039585635,"rms_gain_error_db":-0.7654684330532037,"reference_mean_power":7.077979037597614e-05,"output_mean_power":5.934203409801408e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":5.2532775502110204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7931175609497946e-05,"output_power":2.0349608781374817e-05,"reference_active":true,"delta_db":-1.3753315094203495},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173764837074772e-06,"output_power":4.85233441524043e-06,"reference_active":true,"delta_db":-0.27855960742292113},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7085501560216017e-07,"output_power":1.4484257609255433e-07,"reference_active":true,"delta_db":-0.7173149208286035}],"active_band_count":3,"active_band_mae_db":0.7904020125572914}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.309302806854248,"predicted_mos":3.309302806854248,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":259072,"input_sample_rate":44100,"input_seconds":5.874648526077097,"input_float32_sha256":"5dfae243c3980b640e3b4f43ef14b65d693e530130a976af0fae8bb3014c1bd4","internal_samples":93995,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.694081883999388},"started_at":"2026-10-13T04:59:25.087132+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007763762000649876,"native_input_samples":258867,"native_input_sha256":"4d9a7b4f73754ef5e42859053f1cba466ad630c028adfc5b8f7c8304a467761e","output_sha256":"2c258f543231cd639edaf12a06914934fcb2a86d096a8eb98920ddaa030d5994","output_samples":259072,"length_delta_samples":205,"analysis_s":0.12904372699995292,"synthesis_s":33.149259812999844,"total_s":33.2783035399998,"rtf":5.669216957410527,"output_peak":0.4409727156162262,"output_rms":0.057500417704653535,"finished_at":"2026-10-13T04:59:58.384268+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"2c258f543231cd639edaf12a06914934fcb2a86d096a8eb98920ddaa030d5994","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":93920,"output_samples":93995,"length_delta_samples":75,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.800043787012637,"output_peak":0.44297116102260786},"common_prefix":{"samples":93920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9992020852172988,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.610189483500669,"spectral_convergence":0.4496137825323324,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.688600652388093,"spectral_convergence":0.38628456978988635,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.475696347406187,"spectral_convergence":0.4191933741910096,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.0002280721773795506,"delta_db":-2.385544466524067,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.001383567093962996,"delta_db":-0.21535928671345994,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.0014930133178643328,"delta_db":-0.5392424159771579,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":0.00010411648701917647,"delta_db":2.0592867987336234,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":7.442796935710463e-05,"delta_db":-1.1743483883901649,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":3.072312433585627e-05,"delta_db":-2.0601459736060246,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4056545549907495,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5380941598724245,"pitch":null,"output_rms_dbfs":-24.79657957649103,"output_peak":0.44297116102260786,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.59149549443165},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93920,"output_samples":93920,"intelligibility":{"estoi":{"value":0.7703438032684903,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.061245717832140306,"rms_dbfs":-24.258485416618605,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.058636746461097,"mean":-2.66944326371513e-05,"zero_energy":false},"output":{"rms":0.05756665849186504,"rms_dbfs":-24.79657957649103,"peak":0.44297116102260786,"peak_dbfs":-7.072490939034513,"crest_factor_db":17.724088637456518,"mean":-7.852250543199929e-05,"zero_energy":false},"rms_difference_db":-0.5380941598724256},"full_scale":{"reference":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":587,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":2.3024710544356792e-06,"output_mean_power":2.1475909629398703e-06,"output_rms_dbfs":-56.680484322156,"output_zero_energy":false,"output_peak":0.007734175292482082,"output_energy_fraction":3.64321972033155e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":580,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":266,"both_estimated_voiced_frames":242,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9166666666666666,"absolute_difference_cents_median":9.919545688238648,"absolute_difference_cents_p95":33.30995909033725,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002156956510733609,"log_rms_mae_db":1.4727123923991998,"rms_envelope_nrmse":0.11095778306269206,"envelope_correlation":0.9901531688295542,"rms_gain_error_db":-0.5380934650661339,"reference_mean_power":0.003754233895192234,"output_mean_power":0.003316744211847114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.680528929094005e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006742966926403748,"output_power":0.0006451270227329057,"reference_active":true,"delta_db":-0.1920579574155039},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002358850904334285,"output_power":0.0002389218736802992,"reference_active":true,"delta_db":0.05555420582730102},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0118677140216256e-05,"output_power":2.4625859946761006e-05,"reference_active":true,"delta_db":-0.8744448784455852}],"active_band_count":3,"active_band_mae_db":0.37401901389613007}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00016458512240675186,"log_rms_mae_db":1.0604710150122594,"rms_envelope_nrmse":0.08225784984237632,"envelope_correlation":0.9962147491972139,"rms_gain_error_db":-0.5380839437412996,"reference_mean_power":0.003773515049488898,"output_mean_power":0.0033337857968702873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.4929366354790985e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006129787155244603,"output_power":0.0005847637753266712,"reference_active":true,"delta_db":-0.20464933450595},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073825851363778,"output_power":0.0001104488073863958,"reference_active":true,"delta_db":-0.011366586792791244},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6089511471242587e-06,"output_power":2.5986059291374398e-06,"reference_active":true,"delta_db":-0.017255219406489105}],"active_band_count":3,"active_band_mae_db":0.07775704690174344}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002116840322083979,"log_rms_mae_db":2.7389880792190313,"rms_envelope_nrmse":0.16414292768271396,"envelope_correlation":0.9762192107468063,"rms_gain_error_db":-0.19119081472423077,"reference_mean_power":0.002333704816993391,"output_mean_power":0.0022331959993458017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.3033890876287954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006736793990758906,"output_power":0.0006120540175258419,"reference_active":true,"delta_db":-0.41663513961991183},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002058984330594163,"output_power":0.00022673953364052784,"reference_active":true,"delta_db":0.4187420750488544},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6276513462408744e-05,"output_power":2.073868124949061e-05,"reference_active":true,"delta_db":-1.027866031019133}],"active_band_count":3,"active_band_mae_db":0.6210810818959663}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00015464611695479593,"log_rms_mae_db":2.1088536422676856,"rms_envelope_nrmse":0.13248101128694453,"envelope_correlation":0.9827467091936744,"rms_gain_error_db":-0.1911974724550408,"reference_mean_power":0.002345702993864749,"output_mean_power":0.0022446739934410755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.1648344391379404e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377113304270822,"output_power":0.0005826854067599944,"reference_active":true,"delta_db":-0.3918999190127731},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.690162436721034e-05,"output_power":9.932113072793155e-05,"reference_active":true,"delta_db":0.10710597950851114},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0235059299775414e-06,"output_power":2.578256255454559e-06,"reference_active":true,"delta_db":-0.6918474459772005}],"active_band_count":3,"active_band_mae_db":0.39695111483282824}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":3.6257174598248674,"rms_envelope_nrmse":0.3670967067295143,"envelope_correlation":0.9073900509580863,"rms_gain_error_db":0.8746651224243397,"reference_mean_power":0.00010718143407110891,"output_mean_power":0.00013109498427004955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":8.16723414342307e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.040560289520862e-05,"output_power":5.615858977612477e-05,"reference_active":true,"delta_db":1.429746021998649},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.346030743633643e-06,"output_power":1.2553296622193535e-05,"reference_active":true,"delta_db":1.772778107464736},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1788672389317663e-06,"output_power":3.026475620730975e-06,"reference_active":true,"delta_db":1.4270641092200158}],"active_band_count":3,"active_band_mae_db":1.5431960795611337}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":2.829286989705202,"rms_envelope_nrmse":0.31203245419792763,"envelope_correlation":0.9291939345446046,"rms_gain_error_db":0.8746655503410229,"reference_mean_power":0.00010773182378082651,"output_mean_power":0.00013176818595402612},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":7.38064524776292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.729428782271285e-05,"output_power":5.333767712497818e-05,"reference_active":true,"delta_db":1.5539177986993236},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.87881666371459e-06,"output_power":4.969351168945049e-06,"reference_active":true,"delta_db":1.076004353943294},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8761762452220107e-07,"output_power":2.2591409692072846e-07,"reference_active":true,"delta_db":0.8066969857219419}],"active_band_count":3,"active_band_mae_db":1.1455397127881863}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":4.930923847996697,"rms_envelope_nrmse":0.33003652830649677,"envelope_correlation":0.933941154803999,"rms_gain_error_db":-1.968764067368585,"reference_mean_power":7.041944137047503e-05,"output_mean_power":4.475238330683913e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":5.931095881530664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9173785649337017e-05,"output_power":1.8856645132051408e-05,"reference_active":true,"delta_db":-1.8952835965815276},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1741151998348046e-05,"output_power":8.50560284149965e-06,"reference_active":true,"delta_db":-1.4000561035514167},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.929169683988833e-06,"output_power":1.1666540013272257e-06,"reference_active":true,"delta_db":-2.184283536642833}],"active_band_count":3,"active_band_mae_db":1.8265410789252592}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":4.344974879078283,"rms_envelope_nrmse":0.29631870723243187,"envelope_correlation":0.9513750419144931,"rms_gain_error_db":-1.9687811198389438,"reference_mean_power":7.077979037597614e-05,"output_mean_power":4.498121272046008e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":5.2532775502110204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7931175609497946e-05,"output_power":1.8993559087261197e-05,"reference_active":true,"delta_db":-1.6748286316937808},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173764837074772e-06,"output_power":4.207683123073663e-06,"reference_active":true,"delta_db":-0.8976365881893901},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7085501560216017e-07,"output_power":1.0611135295954798e-07,"reference_active":true,"delta_db":-2.0686588043854597}],"active_band_count":3,"active_band_mae_db":1.547041341422877}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"golf","status":"measured","mos":{"status":"measured","value":3.1647708415985107,"predicted_mos":3.1647708415985107,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":140640,"input_sample_rate":24000,"input_seconds":5.86,"input_float32_sha256":"d53e77af4ebd7b6618c15411dfd0df87f6ed5cce68fbdbb1eff03be5436aa82b","internal_samples":93760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2906426660010766},"started_at":"2026-10-13T05:01:57.541268+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035740969997277716,"native_input_samples":140880,"native_input_sha256":"a7d7163049b3c80d7972dcfd9f76bd5c412343bd4b9c7e1e0ea5b49625eeb936","output_sha256":"942f6267cc3ad9f659048df166f42417b3dc9256755101b58be2c7fe54ad9d85","output_samples":140640,"length_delta_samples":-240,"analysis_s":0.6370434989994465,"synthesis_s":0.1202699350005787,"total_s":0.7573134340000252,"rtf":0.1290142136286244,"output_peak":0.3600919246673584,"output_rms":0.056477081405717014,"finished_at":"2026-10-13T05:01:58.313131+00:00","analysis_stages":{"external_f0_s":0.08082826000008936},"evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"942f6267cc3ad9f659048df166f42417b3dc9256755101b58be2c7fe54ad9d85","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":93920,"output_samples":93760,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.95702210129631,"output_peak":0.3600581606902645},"common_prefix":{"samples":93760,"sample_rate":16000,"reference_coverage":0.9982964224872232,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93760,"output_samples":93760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.305945963577005,"spectral_convergence":0.47977726513545466,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.646744676010698,"spectral_convergence":0.28136307355058177,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.798543254482344,"spectral_convergence":0.3211067909024698,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003982466300241428,"output_power":0.000366042942473896,"delta_db":-0.3662007166106662,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014541293910716815,"output_power":0.0013603110155135183,"delta_db":-0.28964837654587683,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016930045986002563,"output_power":0.0013103865340217268,"delta_db":-1.1125871627561912,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.489910143275e-05,"output_power":5.4937664705040344e-05,"delta_db":-0.7236848932279052,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.770699533974857e-05,"output_power":6.979861708941587e-05,"delta_db":-1.4607884008787175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.944650819910735e-05,"output_power":3.225022621865103e-05,"delta_db":-1.8560286278611824,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.96815636298009,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7059348138031685,"pitch":null,"output_rms_dbfs":-24.95702210129631,"output_peak":0.3600581606902645,"reference_rms_dbfs":-24.251087287493146,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.917077964690016},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93760,"output_samples":93760,"intelligibility":{"estoi":{"value":0.8979369485637302,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06129790554878435,"rms_dbfs":-24.251087287493146,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.051238617335635,"mean":-2.469886285046262e-05,"zero_energy":false},"output":{"rms":0.05651306928509766,"rms_dbfs":-24.95702210129631,"peak":0.3600581606902645,"peak_dbfs":-8.872546827616148,"crest_factor_db":16.084475273680166,"mean":-0.0004078192427498669,"zero_energy":false},"rms_difference_db":-0.7059348138031645},"full_scale":{"reference":{"sample_count":93760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":586,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.26790380111197e-06,"output_mean_power":2.034609882290506e-06,"output_rms_dbfs":-56.915188504242266,"output_zero_energy":false,"output_peak":0.005018678055335754,"output_energy_fraction":3.478850149263146e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":579,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":293,"both_estimated_voiced_frames":259,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9810606060606061,"absolute_difference_cents_median":6.651485597156624,"absolute_difference_cents_p95":20.406760182311583,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1171,"rms_floor":0.0002156956510733609,"log_rms_mae_db":1.8895549470397806,"rms_envelope_nrmse":0.1501167032345382,"envelope_correlation":0.980072585190348,"rms_gain_error_db":-0.705932353566552,"reference_mean_power":0.003760639015423336,"output_mean_power":0.0031964536550788676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1171,"frequency_resolution_hz":0.1707941929974381,"power_floor":1.6812503640734531e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006748860349589676,"output_power":0.0006245239160443535,"reference_active":true,"delta_db":-0.3368136735347033},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002360908288843899,"output_power":0.00023839266374857263,"reference_active":true,"delta_db":0.042137694469762135},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0089595689944443e-05,"output_power":2.807042821903884e-05,"reference_active":true,"delta_db":-0.30167314300444903}],"active_band_count":3,"active_band_mae_db":0.22687483700297148}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1165,"rms_floor":0.00016458512240675186,"log_rms_mae_db":1.454941636010159,"rms_envelope_nrmse":0.11743963639376387,"envelope_correlation":0.9895875978830277,"rms_gain_error_db":-0.7059299494900977,"reference_mean_power":0.0037799871060560067,"output_mean_power":0.003212900850137154},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1165,"frequency_resolution_hz":0.17167381974248927,"power_floor":1.4937574488856314e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006137907220238107,"output_power":0.0005781488031548828,"reference_active":true,"delta_db":-0.2598068847730384},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011063164950993407,"output_power":0.0001119282311200602,"reference_active":true,"delta_db":0.05060252616801941},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6075368040029037e-06,"output_power":2.1036846021622167e-06,"reference_active":true,"delta_db":-0.9324981882031986}],"active_band_count":3,"active_band_mae_db":0.41430253304808545}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1171,"rms_floor":0.0002116840322083979,"log_rms_mae_db":1.844180421371604,"rms_envelope_nrmse":0.18061179542620723,"envelope_correlation":0.9757082072150529,"rms_gain_error_db":-0.7930446453943235,"reference_mean_power":0.002337690588531355,"output_mean_power":0.0019475228587953048},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1171,"frequency_resolution_hz":0.1707941929974381,"power_floor":1.304152822630957e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006740239419837064,"output_power":0.0006119209614277407,"reference_active":true,"delta_db":-0.4197999310792806},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020611606890906466,"output_power":0.0002049921400552962,"reference_active":true,"delta_db":-0.023746415405948264},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6133921958752078e-05,"output_power":2.238135839735247e-05,"reference_active":true,"delta_db":-0.6731814813864023}],"active_band_count":3,"active_band_mae_db":0.37224260929054376}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1165,"rms_floor":0.00015464611695479593,"log_rms_mae_db":1.3922805618435206,"rms_envelope_nrmse":0.1530892085255732,"envelope_correlation":0.9827208264851387,"rms_gain_error_db":-0.7930445148935955,"reference_mean_power":0.002349729877081614,"output_mean_power":0.0019575528119801537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1165,"frequency_resolution_hz":0.17167381974248927,"power_floor":1.1657240548585992e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381284196972652,"output_power":0.0005833420527279171,"reference_active":true,"delta_db":-0.38984801249470685},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.688667677807707e-05,"output_power":9.641634425903089e-05,"reference_active":true,"delta_db":-0.02113399156162077},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0203625396804564e-06,"output_power":2.5427053298739804e-06,"reference_active":true,"delta_db":-0.7476304186622319}],"active_band_count":3,"active_band_mae_db":0.3862041409061865}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1171,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":1.71453131121285,"rms_envelope_nrmse":0.1958146897926559,"envelope_correlation":0.9732589935384213,"rms_gain_error_db":-0.9660978611871092,"reference_mean_power":0.00010736414311283751,"output_mean_power":8.595071150890568e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1171,"frequency_resolution_hz":0.1707941929974381,"power_floor":8.183041847010942e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0509559278891836e-05,"output_power":2.8622441926726528e-05,"reference_active":true,"delta_db":-1.5085083544574587},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.322055378226123e-06,"output_power":6.925653465392525e-06,"reference_active":true,"delta_db":-0.7976984419798083},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1783610402057645e-06,"output_power":1.3998806945799643e-06,"reference_active":true,"delta_db":-1.9203883675724922}],"active_band_count":3,"active_band_mae_db":1.4088650546699197}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1165,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":1.2502308540366833,"rms_envelope_nrmse":0.1508150754815452,"envelope_correlation":0.9873217203553917,"rms_gain_error_db":-0.9660983761540692,"reference_mean_power":0.00010791640453307677,"output_mean_power":8.639281592837059e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1165,"frequency_resolution_hz":0.17167381974248927,"power_floor":7.395294278444177e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.739624930784746e-05,"output_power":2.664681346890214e-05,"reference_active":true,"delta_db":-1.4718276458950872},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.850367830746337e-06,"output_power":3.1515847681503897e-06,"reference_active":true,"delta_db":-0.8697322738435632},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.879140124521217e-07,"output_power":1.135596867080265e-07,"reference_active":true,"delta_db":-2.1873498025350164}],"active_band_count":3,"active_band_mae_db":1.5096365740912223}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1171,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":2.833529498090409,"rms_envelope_nrmse":0.2526952581603455,"envelope_correlation":0.9645827951817619,"rms_gain_error_db":-1.5412998366417874,"reference_mean_power":7.053834323170161e-05,"output_mean_power":4.9464687681682316e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1171,"frequency_resolution_hz":0.1707941929974381,"power_floor":5.93207036625648e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9189742764699675e-05,"output_power":2.184600450950058e-05,"reference_active":true,"delta_db":-1.2585824880883298},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.171918904389644e-05,"output_power":8.78672225804688e-06,"reference_active":true,"delta_db":-1.2507066109728808},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9288004983096316e-06,"output_power":2.932305347311933e-06,"reference_active":true,"delta_db":1.8192188277937984}],"active_band_count":3,"active_band_mae_db":1.4428359756183362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93760,"uncovered_tail_samples":0,"metrics":{"frame_count":1165,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":2.5688098290831634,"rms_envelope_nrmse":0.21034623466642965,"envelope_correlation":0.9826183508740334,"rms_gain_error_db":-1.5413035464909903,"reference_mean_power":7.090046321407977e-05,"output_mean_power":4.97185801774293e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1165,"frequency_resolution_hz":0.17167381974248927,"power_floor":5.2537366939307095e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.793056896265401e-05,"output_power":2.2178160541190693e-05,"reference_active":true,"delta_db":-1.0015425975356202},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.1617802029781336e-06,"output_power":3.7994264032354796e-06,"reference_active":true,"delta_db":-1.3308147120477245},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7086265130352617e-07,"output_power":2.674146898305571e-07,"reference_active":true,"delta_db":1.9453811948433388}],"active_band_count":3,"active_band_mae_db":1.4259128348088945}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.199180841445923,"predicted_mos":3.199180841445923,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":141000,"input_sample_rate":24000,"input_seconds":5.875,"input_float32_sha256":"3fc6ba263146f2719371bb9c721a6412584b0b4a02f04865aacdb3af5c8182de","internal_samples":94000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.341899844999716},"started_at":"2026-10-13T05:05:03.931111+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004382551999697171,"native_input_samples":140880,"native_input_sha256":"a7d7163049b3c80d7972dcfd9f76bd5c412343bd4b9c7e1e0ea5b49625eeb936","output_sha256":"ac3b49e72989f6066149952a0d376608a0c52602cd0d757ce414c8a571fb98df","output_samples":141000,"length_delta_samples":120,"analysis_s":2.4045707889999903,"synthesis_s":14.519033321000279,"total_s":16.92360411000027,"rtf":2.883067139693402,"output_peak":0.39357009530067444,"output_rms":0.058543490450747186,"finished_at":"2026-10-13T05:05:20.872342+00:00","analysis_stages":{"external_f0_s":1.5213648250000915},"evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"ac3b49e72989f6066149952a0d376608a0c52602cd0d757ce414c8a571fb98df","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":93920,"output_samples":94000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.64489868822539,"output_peak":0.3957462176222885},"common_prefix":{"samples":93920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991489361702127,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.914684540576306,"spectral_convergence":0.45351453615936893,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.514027186400491,"spectral_convergence":0.32765388312821203,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.325943725648292,"spectral_convergence":0.3095706031477205,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.00033450789958281406,"delta_db":-0.7222037707331905,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.0013858739253523344,"delta_db":-0.20812429331382568,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.0015193971022584361,"delta_db":-0.46316629503116025,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":6.123841957958384e-05,"delta_db":-0.24566852291732252,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":9.541422405129833e-05,"delta_db":-0.09557886211862333,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":3.81962205574659e-05,"delta_db":-1.1145958337008017,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.47488959630248734,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.38271778633249987,"pitch":null,"output_rms_dbfs":-24.641203202951104,"output_peak":0.3957462176222885,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.584885150875029},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93920,"output_samples":93920,"intelligibility":{"estoi":{"value":0.9008095180959668,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.061245717832140306,"rms_dbfs":-24.258485416618605,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.058636746461097,"mean":-2.66944326371513e-05,"zero_energy":false},"output":{"rms":0.05860569760170024,"rms_dbfs":-24.641203202951104,"peak":0.3957462176222885,"peak_dbfs":-8.051664545035113,"crest_factor_db":16.58953865791599,"mean":-1.2260030055834789e-05,"zero_energy":false},"rms_difference_db":-0.3827177863324991},"full_scale":{"reference":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":587,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":2.3024710544356792e-06,"output_mean_power":6.433921128200123e-06,"output_rms_dbfs":-51.915242673068214,"output_zero_energy":false,"output_peak":0.0347948960491392,"output_energy_fraction":0.00010531055258912243,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":580,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":290,"both_estimated_voiced_frames":252,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":38,"paired_voiced_fraction":0.9545454545454546,"absolute_difference_cents_median":5.936889103293957,"absolute_difference_cents_p95":26.204835007837264,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002156956510733609,"log_rms_mae_db":2.3362606972191067,"rms_envelope_nrmse":0.17179575435550376,"envelope_correlation":0.9667492652294428,"rms_gain_error_db":-0.3827164910057089,"reference_mean_power":0.003754233895192234,"output_mean_power":0.0034375551727282194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.680528929094005e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006742966926403748,"output_power":0.0006455942503192297,"reference_active":true,"delta_db":-0.18891375617495698},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002358850904334285,"output_power":0.00022676012240493268,"reference_active":true,"delta_db":-0.17133808592089295},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0118677140216256e-05,"output_power":2.9501531903163622e-05,"reference_active":true,"delta_db":-0.08991325212840072}],"active_band_count":3,"active_band_mae_db":0.15005503140808354}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00016458512240675186,"log_rms_mae_db":1.7251959491803193,"rms_envelope_nrmse":0.11956442544295001,"envelope_correlation":0.9830630418462734,"rms_gain_error_db":-0.3826968921910825,"reference_mean_power":0.003773515049488898,"output_mean_power":0.003455225507642869},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.4929366354790985e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006129787155244603,"output_power":0.0005917549983936299,"reference_active":true,"delta_db":-0.15303459777814},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073825851363778,"output_power":0.00010885000288279971,"reference_active":true,"delta_db":-0.07469244570817019},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6089511471242587e-06,"output_power":2.7237864322849093e-06,"reference_active":true,"delta_db":0.1870710527558009}],"active_band_count":3,"active_band_mae_db":0.13826603208070368}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002116840322083979,"log_rms_mae_db":2.2405425178704372,"rms_envelope_nrmse":0.19311235288509182,"envelope_correlation":0.9672769587700784,"rms_gain_error_db":-0.39291499177134037,"reference_mean_power":0.002333704816993391,"output_mean_power":0.0021318391006369506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.3033890876287954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006736793990758906,"output_power":0.0006226848154249966,"reference_active":true,"delta_db":-0.341849916576229},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002058984330594163,"output_power":0.00019592478083531805,"reference_active":true,"delta_db":-0.21563671905455828},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6276513462408744e-05,"output_power":2.8590901857579545e-05,"reference_active":true,"delta_db":0.3666011503674283}],"active_band_count":3,"active_band_mae_db":0.30802926199940517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00015464611695479593,"log_rms_mae_db":1.6617212196728186,"rms_envelope_nrmse":0.1311830496696723,"envelope_correlation":0.984117412128212,"rms_gain_error_db":-0.39291501016383057,"reference_mean_power":0.002345702993864749,"output_mean_power":0.002142799424850419},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.1648344391379404e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377113304270822,"output_power":0.0005912533920753743,"reference_active":true,"delta_db":-0.3285048796818297},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.690162436721034e-05,"output_power":9.280686179193388e-05,"reference_active":true,"delta_db":-0.18750969699692796},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0235059299775414e-06,"output_power":3.3051014800666117e-06,"reference_active":true,"delta_db":0.3867397391344807}],"active_band_count":3,"active_band_mae_db":0.30091810527107943}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":2.153184441912019,"rms_envelope_nrmse":0.19818034916513516,"envelope_correlation":0.9654086799411643,"rms_gain_error_db":-0.0985594295589218,"reference_mean_power":0.00010718143407110891,"output_mean_power":0.00010477643572738037},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":8.16723414342307e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.040560289520862e-05,"output_power":3.959922559899749e-05,"reference_active":true,"delta_db":-0.08754898331057463},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.346030743633643e-06,"output_power":8.03620840779835e-06,"reference_active":true,"delta_db":-0.16428789115159267},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1788672389317663e-06,"output_power":2.022628087788824e-06,"reference_active":true,"delta_db":-0.3231473505608578}],"active_band_count":3,"active_band_mae_db":0.19166140834100834}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":1.2287138660279535,"rms_envelope_nrmse":0.11173306465764342,"envelope_correlation":0.9879145859581518,"rms_gain_error_db":-0.09856357526539279,"reference_mean_power":0.00010773182378082651,"output_mean_power":0.00010531437494638915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":7.38064524776292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.729428782271285e-05,"output_power":3.6839591679562046e-05,"reference_active":true,"delta_db":-0.053275102703018364},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.87881666371459e-06,"output_power":3.751155278899364e-06,"reference_active":true,"delta_db":-0.1453421058401998},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8761762452220107e-07,"output_power":1.6638544294189633e-07,"reference_active":true,"delta_db":-0.5215830573700957}],"active_band_count":3,"active_band_mae_db":0.24006675530443797}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":2.324340979369878,"rms_envelope_nrmse":0.24499311873103374,"envelope_correlation":0.946193109432186,"rms_gain_error_db":-0.5743628610696717,"reference_mean_power":7.041944137047503e-05,"output_mean_power":6.169589793612517e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":5.931095881530664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9173785649337017e-05,"output_power":2.4985155853253943e-05,"reference_active":true,"delta_db":-0.6731072486343679},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1741151998348046e-05,"output_power":9.399094672904083e-06,"reference_active":true,"delta_db":-0.966246862894899},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.929169683988833e-06,"output_power":1.570349314464549e-06,"reference_active":true,"delta_db":-0.8937415929690518}],"active_band_count":3,"active_band_mae_db":0.8443652348327729}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":1.3684108652927607,"rms_envelope_nrmse":0.14598355824739181,"envelope_correlation":0.9811975497530155,"rms_gain_error_db":-0.5743640221354347,"reference_mean_power":7.077979037597614e-05,"output_mean_power":6.20115904151179e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":5.2532775502110204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7931175609497946e-05,"output_power":2.3754044386320933e-05,"reference_active":true,"delta_db":-0.7035165167729849},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173764837074772e-06,"output_power":4.085548003498808e-06,"reference_active":true,"delta_db":-1.0255636721419814},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7085501560216017e-07,"output_power":1.329695336147235e-07,"reference_active":true,"delta_db":-1.088755869376591}],"active_band_count":3,"active_band_mae_db":0.9392786860971857}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.355379343032837,"predicted_mos":3.355379343032837,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":141120,"input_sample_rate":24000,"input_seconds":5.88,"input_float32_sha256":"c9f449724b1a5fcac33025dec3e170665b4dbee6c25a450b61131d7682dab100","internal_samples":94080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4097646350001014},"started_at":"2026-10-13T05:10:03.421055+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0037053240012028255,"native_input_samples":140880,"native_input_sha256":"a7d7163049b3c80d7972dcfd9f76bd5c412343bd4b9c7e1e0ea5b49625eeb936","output_sha256":"b6355df58b0e9fa8e008d42eab32dfda74015a870b176a632f3a6dd5edeaea77","output_samples":141120,"length_delta_samples":240,"analysis_s":0.6653381460000674,"synthesis_s":0.15877945899956103,"total_s":0.8241176049996284,"rtf":0.1403948219760866,"output_peak":0.43070468306541443,"output_rms":0.06001087174496175,"finished_at":"2026-10-13T05:10:04.259489+00:00","analysis_stages":{"external_f0_s":0.1601467260006757},"evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"b6355df58b0e9fa8e008d42eab32dfda74015a870b176a632f3a6dd5edeaea77","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":93920,"output_samples":94080,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.429759013001238,"output_peak":0.4278703376418579},"common_prefix":{"samples":93920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982993197278912,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.567090743460146,"spectral_convergence":0.1935733277931846,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.501253476491283,"spectral_convergence":0.17564781375848293,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.978540569849087,"spectral_convergence":0.2035985578897654,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.00037456182620044517,"delta_db":-0.23103241605959207,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.001415874617871757,"delta_db":-0.11511356795828587,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.0016238221292894646,"delta_db":-0.1744946766823774,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":6.3406282737501e-05,"delta_db":-0.0945853339203617,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":8.811792338312624e-05,"delta_db":-0.44106755124344804,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":4.634657021179044e-05,"delta_db":-0.27462375257041255,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2218195497390796,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16388164411914574,"pitch":null,"output_rms_dbfs":-24.42236706073775,"output_peak":0.4278703376418579,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.682294929933505},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93920,"output_samples":93920,"intelligibility":{"estoi":{"value":0.9507265568643126,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.061245717832140306,"rms_dbfs":-24.258485416618605,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.058636746461097,"mean":-2.66944326371513e-05,"zero_energy":false},"output":{"rms":0.06010099291770549,"rms_dbfs":-24.422367060737752,"peak":0.4278703376418579,"peak_dbfs":-7.373756403762988,"crest_factor_db":17.048610656974763,"mean":0.00014364095327672842,"zero_energy":false},"rms_difference_db":-0.1638816441191473},"full_scale":{"reference":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":587,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":2.3024710544356792e-06,"output_mean_power":4.077698676468396e-06,"output_rms_dbfs":-53.89584869774657,"output_zero_energy":false,"output_peak":0.02543747004139655,"output_energy_fraction":6.346403414376373e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":580,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":254,"both_estimated_voiced_frames":248,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9393939393939394,"absolute_difference_cents_median":1.102409649371416,"absolute_difference_cents_p95":7.117220804597835,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002156956510733609,"log_rms_mae_db":0.7423529103209182,"rms_envelope_nrmse":0.057255815750308935,"envelope_correlation":0.9965529071794278,"rms_gain_error_db":-0.16388098589321085,"reference_mean_power":0.003754233895192234,"output_mean_power":0.0036152074877599176},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.680528929094005e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006742966926403748,"output_power":0.0006682229832614656,"reference_active":true,"delta_db":-0.039296205662170826},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002358850904334285,"output_power":0.00021433248627697555,"reference_active":true,"delta_db":-0.41612489495861693},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0118677140216256e-05,"output_power":2.408163280345292e-05,"reference_active":true,"delta_db":-0.9714996305940122}],"active_band_count":3,"active_band_mae_db":0.4756402437382667}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00016458512240675186,"log_rms_mae_db":0.43153199387356345,"rms_envelope_nrmse":0.03517152685798451,"envelope_correlation":0.9988189276818609,"rms_gain_error_db":-0.16387769642566757,"reference_mean_power":0.003773515049488898,"output_mean_power":0.0036337773766367547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.4929366354790985e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006129787155244603,"output_power":0.000607800272194717,"reference_active":true,"delta_db":-0.0368450452529148},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073825851363778,"output_power":0.000101975100867408,"reference_active":true,"delta_db":-0.35803545883497745},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6089511471242587e-06,"output_power":2.176418995463924e-06,"reference_active":true,"delta_db":-0.7872343925940783}],"active_band_count":3,"active_band_mae_db":0.39403829889399017}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002116840322083979,"log_rms_mae_db":0.836600066501745,"rms_envelope_nrmse":0.05562580386078221,"envelope_correlation":0.9974947893519531,"rms_gain_error_db":-0.15407476196871742,"reference_mean_power":0.002333704816993391,"output_mean_power":0.002252363321289164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.3033890876287954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006736793990758906,"output_power":0.000655245518945968,"reference_active":true,"delta_db":-0.12049207352640576},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002058984330594163,"output_power":0.00019448479488522533,"reference_active":true,"delta_db":-0.24767388345819788},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6276513462408744e-05,"output_power":1.935454212256806e-05,"reference_active":true,"delta_db":-1.327848381981684}],"active_band_count":3,"active_band_mae_db":0.5653381129887626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00015464611695479593,"log_rms_mae_db":0.4663220987931981,"rms_envelope_nrmse":0.0323855114005491,"envelope_correlation":0.9992516335245201,"rms_gain_error_db":-0.15407474897581078,"reference_mean_power":0.002345702993864749,"output_mean_power":0.0022639433073769594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.1648344391379404e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377113304270822,"output_power":0.0006202816195406039,"reference_active":true,"delta_db":-0.12035221159400564},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.690162436721034e-05,"output_power":9.203397739510083e-05,"reference_active":true,"delta_db":-0.22382866053112765},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0235059299775414e-06,"output_power":2.2840659237398033e-06,"reference_active":true,"delta_db":-1.2180219016330445}],"active_band_count":3,"active_band_mae_db":0.5207342579193925}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":0.961334766841138,"rms_envelope_nrmse":0.08877442532509656,"envelope_correlation":0.9937372879867848,"rms_gain_error_db":-0.26672911082880496,"reference_mean_power":0.00010718143407110891,"output_mean_power":0.00010079677857611257},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":8.16723414342307e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.040560289520862e-05,"output_power":3.8946261863029636e-05,"reference_active":true,"delta_db":-0.15975811701728343},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.346030743633643e-06,"output_power":7.249359839980064e-06,"reference_active":true,"delta_db":-0.6118032258343233},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1788672389317663e-06,"output_power":1.6170716730606217e-06,"reference_active":true,"delta_db":-1.2950149949599934}],"active_band_count":3,"active_band_mae_db":0.6888587792705335}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":0.6072057523220054,"rms_envelope_nrmse":0.055015605316072094,"envelope_correlation":0.9979221923801869,"rms_gain_error_db":-0.26672947396957963,"reference_mean_power":0.00010773182378082651,"output_mean_power":0.00010131437383121594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":7.38064524776292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.729428782271285e-05,"output_power":3.589593982093634e-05,"reference_active":true,"delta_db":-0.16596989778149873},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.87881666371459e-06,"output_power":3.2943557576327883e-06,"reference_active":true,"delta_db":-0.7092875585911833},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8761762452220107e-07,"output_power":1.3749462843243838e-07,"reference_active":true,"delta_db":-1.349879012306325}],"active_band_count":3,"active_band_mae_db":0.7417121562263356}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":1.0010445348013464,"rms_envelope_nrmse":0.10437395293460407,"envelope_correlation":0.9906228639848811,"rms_gain_error_db":-0.22615740798608175,"reference_mean_power":7.041944137047503e-05,"output_mean_power":6.684621704500135e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":5.931095881530664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9173785649337017e-05,"output_power":2.7075881956579624e-05,"reference_active":true,"delta_db":-0.32410175682861564},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1741151998348046e-05,"output_power":1.0184813578833019e-05,"reference_active":true,"delta_db":-0.6175762621466931},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.929169683988833e-06,"output_power":1.372600903805584e-06,"reference_active":true,"delta_db":-1.4782614804496073}],"active_band_count":3,"active_band_mae_db":0.8066464998083053}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":0.582153446566763,"rms_envelope_nrmse":0.05962722653893594,"envelope_correlation":0.9969832734651478,"rms_gain_error_db":-0.22615352538492503,"reference_mean_power":7.077979037597614e-05,"output_mean_power":6.718834128246876e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":5.2532775502110204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7931175609497946e-05,"output_power":2.5582807598875663e-05,"reference_active":true,"delta_db":-0.38141010689778715},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173764837074772e-06,"output_power":4.481769703949819e-06,"reference_active":true,"delta_db":-0.6235714833948826},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7085501560216017e-07,"output_power":1.2361333444664177e-07,"reference_active":true,"delta_db":-1.405624108296585}],"active_band_count":3,"active_band_mae_db":0.8035352328630849}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0020","method":"magphase","status":"measured","mos":{"status":"measured","value":3.601630687713623,"predicted_mos":3.601630687713623,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":282206,"input_sample_rate":48000,"input_seconds":5.879291666666667,"input_float32_sha256":"6feb8abe1ac63c9469a2a19cf291de1411ca4fbb93a6dc78a4d5c6ae027328e5","internal_samples":94069,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.1810233790001803},"started_at":"2026-10-13T05:12:34.786726+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.019004437999683432,"native_input_samples":281760,"native_input_sha256":"8da5cd2c3d78583ce4417f6ea9d69642cbff4f88c9590ad276f6c15f6874c25e","output_sha256":"339e21e4b39ab6e89630ecf5c813e1d559a6003b8ab2bd9f16b62aac56cdfbf1","output_samples":282206,"length_delta_samples":446,"analysis_s":3.603051076999691,"synthesis_s":1.0845972120005172,"total_s":4.687648289000208,"rtf":0.7985772212947543,"output_peak":0.395757767576259,"output_rms":0.058551727929570405,"finished_at":"2026-10-13T05:12:39.508115+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"339e21e4b39ab6e89630ecf5c813e1d559a6003b8ab2bd9f16b62aac56cdfbf1","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":93920,"output_samples":94069,"length_delta_samples":149,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.64370236126451,"output_peak":0.39507494928350734},"common_prefix":{"samples":93920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9984160562990996,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":93920,"output_samples":93920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.099158367091698,"spectral_convergence":0.5149907543971466,"spectral_convergence_status":"ok","amplitude_floor":2.214924872151779e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.48018560024756,"spectral_convergence":0.316572001567319,"spectral_convergence_status":"ok","amplitude_floor":1.3113821592368984e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.803115868191139,"spectral_convergence":0.25637320459083973,"spectral_convergence_status":"ok","amplitude_floor":7.380112905156557e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00039502697710642294,"output_power":0.000338237812087537,"delta_db":-0.6740459901917456,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014539054117940992,"output_power":0.0013396562630738102,"delta_db":-0.35542774291287205,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0016903939251448888,"output_power":0.0015447275076282225,"delta_db":-0.3913604301607112,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.480236057553453e-05,"output_power":6.457358821629146e-05,"delta_db":-0.015359064449490478,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.753736279208405e-05,"output_power":9.971778839396842e-05,"delta_db":0.09601628880194407,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.937191536111839e-05,"output_power":5.118122871755768e-05,"delta_db":0.1563073206282221,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.28141947285749763,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.37833688133898513,"pitch":null,"output_rms_dbfs":-24.636822297957593,"output_peak":0.39507494928350734,"reference_rms_dbfs":-24.258485416618605,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.460819945176799},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":93920,"output_samples":93920,"intelligibility":{"estoi":{"value":0.8612013691632266,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.061245717832140306,"rms_dbfs":-24.258485416618605,"peak":0.4365234375,"peak_dbfs":-7.199848670157509,"crest_factor_db":17.058636746461097,"mean":-2.66944326371513e-05,"zero_energy":false},"output":{"rms":0.05863526403212821,"rms_dbfs":-24.636822297957586,"peak":0.39507494928350734,"peak_dbfs":-8.066410139494366,"crest_factor_db":16.570412158463224,"mean":0.0014735355855883911,"zero_energy":false},"rms_difference_db":-0.37833688133898136},"full_scale":{"reference":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":93920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":587,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.141902683768421e-06,"threshold_dbfs":-53.82800109305164,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":2.3024710544356792e-06,"output_mean_power":2.600794363701779e-06,"output_rms_dbfs":-55.84893984689978,"output_zero_energy":false,"output_peak":0.010073867429374196,"output_energy_fraction":4.252693503458417e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":580,"reference_estimated_voiced_frames":264,"output_estimated_voiced_frames":252,"both_estimated_voiced_frames":241,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9128787878787878,"absolute_difference_cents_median":4.997878385300921,"absolute_difference_cents_p95":22.42957600106794,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":93920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002156956510733609,"log_rms_mae_db":1.9826498943290227,"rms_envelope_nrmse":0.18176948938022486,"envelope_correlation":0.9626331757018061,"rms_gain_error_db":-0.3783374611732011,"reference_mean_power":0.003754233895192234,"output_mean_power":0.00344102303818518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.680528929094005e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006742966926403748,"output_power":0.0006377819085704111,"reference_active":true,"delta_db":-0.24178833698848312},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002358850904334285,"output_power":0.00023768347407075954,"reference_active":true,"delta_db":0.032984953100555436},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0118677140216256e-05,"output_power":2.742689630580336e-05,"reference_active":true,"delta_db":-0.40659228482493426}],"active_band_count":3,"active_band_mae_db":0.22712185830465761}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00016458512240675186,"log_rms_mae_db":1.2043414063635074,"rms_envelope_nrmse":0.11726157329304668,"envelope_correlation":0.983763000523501,"rms_gain_error_db":-0.378326790094589,"reference_mean_power":0.003773515049488898,"output_mean_power":0.0034587040892200944},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.4929366354790985e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006129787155244603,"output_power":0.0005785035906462328,"reference_active":true,"delta_db":-0.2513933589414608},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073825851363778,"output_power":0.00011178600047070143,"reference_active":true,"delta_db":0.040897285995055085},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6089511471242587e-06,"output_power":2.5346288372653345e-06,"reference_active":true,"delta_db":-0.12551575302313156}],"active_band_count":3,"active_band_mae_db":0.13926879931988248}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":0.0002116840322083979,"log_rms_mae_db":2.5174888384050647,"rms_envelope_nrmse":0.20625862772026962,"envelope_correlation":0.9624335514368543,"rms_gain_error_db":-0.38189265559434227,"reference_mean_power":0.002333704816993391,"output_mean_power":0.0021372565517161503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":1.3033890876287954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006736793990758906,"output_power":0.0006230950049273312,"reference_active":true,"delta_db":-0.33898997229191846},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002058984330594163,"output_power":0.00019802405018540248,"reference_active":true,"delta_db":-0.16935102646344108},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6276513462408744e-05,"output_power":2.5365924498400034e-05,"reference_active":true,"delta_db":-0.153170442698154}],"active_band_count":3,"active_band_mae_db":0.22050381381783782}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":0.00015464611695479593,"log_rms_mae_db":1.8195156976322524,"rms_envelope_nrmse":0.13573458458660836,"envelope_correlation":0.9827995069512078,"rms_gain_error_db":-0.381892884549763,"reference_mean_power":0.002345702993864749,"output_mean_power":0.0021482446242610444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":1.1648344391379404e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377113304270822,"output_power":0.0005915062636915817,"reference_active":true,"delta_db":-0.32664785406066205},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.690162436721034e-05,"output_power":9.388000006353748e-05,"reference_active":true,"delta_db":-0.13757975993370195},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0235059299775414e-06,"output_power":3.064171877242889e-06,"reference_active":true,"delta_db":0.05802297611796979}],"active_band_count":3,"active_band_mae_db":0.1740835300374446}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":4.7944659079237455e-05,"log_rms_mae_db":2.402557359932534,"rms_envelope_nrmse":0.24429255547060538,"envelope_correlation":0.9480395001366579,"rms_gain_error_db":0.0027595786811074525,"reference_mean_power":0.00010718143407110891,"output_mean_power":0.0001072495605620161},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":8.16723414342307e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.040560289520862e-05,"output_power":4.092777095040869e-05,"reference_active":true,"delta_db":0.05576501012054407},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.346030743633643e-06,"output_power":8.79934509864052e-06,"reference_active":true,"delta_db":0.22970370331685933},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1788672389317663e-06,"output_power":2.03380380555381e-06,"reference_active":true,"delta_db":-0.2992171332448154}],"active_band_count":3,"active_band_mae_db":0.19489528222740626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":3.945206980281183e-05,"log_rms_mae_db":1.435198810681642,"rms_envelope_nrmse":0.14430960038150695,"envelope_correlation":0.979887230822046,"rms_gain_error_db":0.002757537224533082,"reference_mean_power":0.00010773182378082651,"output_mean_power":0.00010780024943666328},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":7.38064524776292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.729428782271285e-05,"output_power":3.808827510193077e-05,"reference_active":true,"delta_db":0.09148987061567616},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.87881666371459e-06,"output_power":4.013179258502668e-06,"reference_active":true,"delta_db":0.14789306289343715},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8761762452220107e-07,"output_power":1.8843232119468653e-07,"reference_active":true,"delta_db":0.018817650520662046}],"active_band_count":3,"active_band_mae_db":0.08606686134325846}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1173,"rms_floor":5.718269674480752e-05,"log_rms_mae_db":2.639883766030076,"rms_envelope_nrmse":0.34135717225192547,"envelope_correlation":0.8967507694733557,"rms_gain_error_db":0.09034250624110157,"reference_mean_power":7.041944137047503e-05,"output_mean_power":7.189965825702423e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1173,"frequency_resolution_hz":0.17050298380221654,"power_floor":5.931095881530664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9173785649337017e-05,"output_power":2.6744853830574538e-05,"reference_active":true,"delta_db":-0.3775255903207237},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1741151998348046e-05,"output_power":1.3478846055354038e-05,"reference_active":true,"delta_db":0.5994200280706199},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.929169683988833e-06,"output_power":2.8221405179852174e-06,"reference_active":true,"delta_db":1.6520820546816672}],"active_band_count":3,"active_band_mae_db":0.8763425576910037}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":93920,"uncovered_tail_samples":0,"metrics":{"frame_count":1167,"rms_floor":4.6736281904897674e-05,"log_rms_mae_db":1.6658284523805382,"rms_envelope_nrmse":0.20699926625109952,"envelope_correlation":0.9566688387464668,"rms_gain_error_db":0.09034910289079709,"reference_mean_power":7.077979037597614e-05,"output_mean_power":7.226769156961538e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1167,"frequency_resolution_hz":0.1713796058269066,"power_floor":5.2532775502110204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7931175609497946e-05,"output_power":2.626623098335355e-05,"reference_active":true,"delta_db":-0.26691456228905},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173764837074772e-06,"output_power":6.529042964601505e-06,"reference_active":true,"delta_db":1.0104284164479063},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7085501560216017e-07,"output_power":2.169652912797605e-07,"reference_active":true,"delta_db":1.0376253132451132}],"active_band_count":3,"active_band_mae_db":0.7716560973273565}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"hnm","status":"measured","mos":{"status":"measured","value":2.7016665935516357,"predicted_mos":2.7016665935516357,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":52320,"input_sample_rate":16000,"input_seconds":3.27,"input_float32_sha256":"63909735762770252680fd40bc920a759aa6a2970104e8867808eb3d7fe1e9b6","internal_samples":52320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7772288839987596},"started_at":"2026-10-13T04:38:13.724033+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.288800012844149e-05,"native_input_samples":52320,"native_input_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","output_sha256":"2acee25ddc8e0a2431fd6932e0b5d8142b868048b64c5cdb8708a3bdb462bcfa","output_samples":52320,"length_delta_samples":0,"analysis_s":0.7091554270000415,"synthesis_s":0.2738418669996463,"total_s":0.9829972939996878,"rtf":0.3006107932720758,"output_peak":0.3770600684996338,"output_rms":0.06588679344368055,"finished_at":"2026-10-13T04:38:14.710157+00:00","evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"2acee25ddc8e0a2431fd6932e0b5d8142b868048b64c5cdb8708a3bdb462bcfa","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.608381828956337,"spectral_convergence":0.20792414118289393,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.498807151324647,"spectral_convergence":0.1980163810882508,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.773100881217547,"spectral_convergence":0.20469281065387782,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.0004791629210624787,"delta_db":0.05606091175194776,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.0018071753637050984,"delta_db":-0.02963019365803924,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0017217627698491455,"delta_db":-0.23476968141478077,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.0001344378938430915,"delta_db":-0.07306600892905969,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00011791479563568706,"delta_db":-0.1920229708929434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":8.061580619472488e-05,"delta_db":-0.3206390220421831,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15103146478149232,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11418665090378392,"pitch":null,"output_rms_dbfs":-23.624032560720828,"output_peak":0.3770600684996338,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.6267632871661775},"common_prefix":{"samples":52320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.608381828956337,"spectral_convergence":0.20792414118289393,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.498807151324647,"spectral_convergence":0.1980163810882508,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.773100881217547,"spectral_convergence":0.20469281065387782,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.0004791629210624787,"delta_db":0.05606091175194776,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.0018071753637050984,"delta_db":-0.02963019365803924,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0017217627698491455,"delta_db":-0.23476968141478077,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.0001344378938430915,"delta_db":-0.07306600892905969,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00011791479563568706,"delta_db":-0.1920229708929434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":8.061580619472488e-05,"delta_db":-0.3206390220421831,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15103146478149232,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11418665090378392,"pitch":null,"output_rms_dbfs":-23.624032560720828,"output_peak":0.3770600684996338,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.6267632871661775},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52320,"output_samples":52320,"intelligibility":{"estoi":{"value":0.8950688219202781,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0667586744040302,"rms_dbfs":-23.509845909817045,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.978446332783722,"mean":4.524732583888809e-05,"zero_energy":false},"output":{"rms":0.06588679344368055,"rms_dbfs":-23.624032560720828,"peak":0.3770600684996338,"peak_dbfs":-8.471789158225997,"crest_factor_db":15.152243402494832,"mean":-8.15169598931402e-05,"zero_energy":false},"rms_difference_db":-0.11418665090378255},"full_scale":{"reference":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":327,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":8.3193759936099e-07,"output_rms_dbfs":-60.79909247350514,"output_zero_energy":false,"output_peak":0.003359646326488524,"output_energy_fraction":1.1721312688173523e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":320,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":146,"both_estimated_voiced_frames":144,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9113924050632911,"absolute_difference_cents_median":1.6721238656509483,"absolute_difference_cents_p95":7.342761665328571,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00017439460044090568,"log_rms_mae_db":0.9113086308749214,"rms_envelope_nrmse":0.061068939482523725,"envelope_correlation":0.9955507838776808,"rms_gain_error_db":-0.11419096056188208,"reference_mean_power":0.004463526965610992,"output_mean_power":0.004347694969654415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":1.3056377024951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008270090313661458,"output_power":0.0007932977734929191,"reference_active":true,"delta_db":-0.18074016922256378},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010887958720618103,"output_power":9.688688300068831e-05,"reference_active":true,"delta_db":-0.5068148145059098},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8937894537156728e-05,"output_power":2.5261093754495975e-05,"reference_active":true,"delta_db":-0.5901477883886103}],"active_band_count":3,"active_band_mae_db":0.4259009240390279}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015806536675822195,"log_rms_mae_db":0.4968170378809166,"rms_envelope_nrmse":0.03728931533101147,"envelope_correlation":0.99829650798017,"rms_gain_error_db":-0.1141973716270274,"reference_mean_power":0.004504823508022685,"output_mean_power":0.004387913357222413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.092227651107542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000711888936919089,"output_power":0.0006777009220035847,"reference_active":true,"delta_db":-0.21374167361197666},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.550528203435593e-05,"output_power":6.22770231410936e-05,"reference_active":true,"delta_db":-0.2194847584894582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5248696894099064e-06,"output_power":1.2632807266322576e-06,"reference_active":true,"delta_db":-0.8173286148801819}],"active_band_count":3,"active_band_mae_db":0.41685168232720554}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00016816220894332027,"log_rms_mae_db":1.4603193485745534,"rms_envelope_nrmse":0.08419256306282105,"envelope_correlation":0.9936576575966237,"rms_gain_error_db":-0.15184749248207297,"reference_mean_power":0.002573247092376177,"output_mean_power":0.0024848303416049686},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":9.221294414644421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007324784888809201,"output_power":0.0006802296135852616,"reference_active":true,"delta_db":-0.32139340152549956},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.576574656453582e-05,"output_power":7.280307511960937e-05,"reference_active":true,"delta_db":-0.7116414855947506},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8048967652930613e-05,"output_power":1.6349276102917018e-05,"reference_active":true,"delta_db":-0.4295383844384092}],"active_band_count":3,"active_band_mae_db":0.48752442385288647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001473388036867024,"log_rms_mae_db":0.9238741729350545,"rms_envelope_nrmse":0.05200523992590268,"envelope_correlation":0.9975398088955205,"rms_gain_error_db":-0.15184988968098734,"reference_mean_power":0.002597106693874505,"output_mean_power":0.0025078687430926225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.911179205548874e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006603877113347142,"output_power":0.0006083501392438106,"reference_active":true,"delta_db":-0.3564537140849772},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5067104939947466e-05,"output_power":4.079775426838677e-05,"reference_active":true,"delta_db":-0.43223402575391207},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.161011125548218e-06,"output_power":9.570715575940085e-07,"reference_active":true,"delta_db":-0.8389197145655093}],"active_band_count":3,"active_band_mae_db":0.5425358181347996}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":1.4792945631079681,"rms_envelope_nrmse":0.13827788680863157,"envelope_correlation":0.9846709516284058,"rms_gain_error_db":-0.19438089920331206,"reference_mean_power":0.0001985957742888365,"output_mean_power":0.00018990303842076085},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":8.760825576371805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.45270565717188e-05,"output_power":4.601647935896855e-05,"reference_active":true,"delta_db":-0.7369866615731788},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.928893294666135e-06,"output_power":5.183799015397342e-06,"reference_active":true,"delta_db":-0.5832547933296716},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9788817769833534e-06,"output_power":1.9289554655564513e-06,"reference_active":true,"delta_db":-0.1109764815840728}],"active_band_count":3,"active_band_mae_db":0.47707264549564105}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":1.0309695485613872,"rms_envelope_nrmse":0.08561247095025556,"envelope_correlation":0.9939612050490897,"rms_gain_error_db":-0.19432499773921647,"reference_mean_power":0.000200416025269102,"output_mean_power":0.0001916460820050679},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.857113792607277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022793901144949e-05,"output_power":4.211741424152685e-05,"reference_active":true,"delta_db":-0.764836580083595},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6196674175617505e-06,"output_power":2.172711363015708e-06,"reference_active":true,"delta_db":-0.8124412286513503},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.185526690136049e-07,"output_power":1.1424319075707207e-07,"reference_active":true,"delta_db":-0.16081011596475844}],"active_band_count":3,"active_band_mae_db":0.5793626415665679}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":1.5168085041254435,"rms_envelope_nrmse":0.18075172005877524,"envelope_correlation":0.9683232782409497,"rms_gain_error_db":-0.26388103923689016,"reference_mean_power":8.71199669251843e-05,"output_mean_power":8.198409329593118e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":4.20340153760379e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.824817597815015e-05,"output_power":1.726406287735229e-05,"reference_active":true,"delta_db":-0.24076451368654858},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2434318429437933e-05,"output_power":1.1851019065061233e-05,"reference_active":true,"delta_db":-0.20866288200949062},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7676697520375684e-06,"output_power":5.077826040764027e-06,"reference_active":true,"delta_db":-1.2476133942685166}],"active_band_count":3,"active_band_mae_db":0.5656802633215187}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":0.9481315406549997,"rms_envelope_nrmse":0.10745264296481749,"envelope_correlation":0.9874022920230912,"rms_gain_error_db":-0.26389735440818607,"reference_mean_power":8.789733624317312e-05,"output_mean_power":8.271532460226435e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":3.223393789064046e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.051966991658495e-05,"output_power":1.8837849413798185e-05,"reference_active":true,"delta_db":-0.37139049442767935},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.88053445243811e-06,"output_power":5.589067110815071e-06,"reference_active":true,"delta_db":-0.22077474283312223},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.871896444237528e-07,"output_power":4.091791512690917e-07,"reference_active":true,"delta_db":-1.5686489065006068}],"active_band_count":3,"active_band_mae_db":0.7202713812538027}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.214627504348755,"predicted_mos":3.214627504348755,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":52320,"input_sample_rate":16000,"input_seconds":3.27,"input_float32_sha256":"470ea1f650c6232136b09d4162ad3754cd33dddebdc38cdadbe7463dedbd2323","internal_samples":52320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1049749670000892},"started_at":"2026-10-13T04:41:06.085425+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.800699960265774e-05,"native_input_samples":52320,"native_input_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","output_sha256":"4145840864abec0c4eb64a8b2b8a0be9e28439f18e78e160fdc8cc1873483cdf","output_samples":52320,"length_delta_samples":0,"analysis_s":1.0341611479998392,"synthesis_s":0.06363319999945816,"total_s":1.0977943479992973,"rtf":0.3357169259936689,"output_peak":0.4622224764757385,"output_rms":0.08050501669537424,"finished_at":"2026-10-13T04:41:07.200246+00:00","evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"4145840864abec0c4eb64a8b2b8a0be9e28439f18e78e160fdc8cc1873483cdf","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.167156766071802,"spectral_convergence":0.516843381859241,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.245449124678112,"spectral_convergence":0.3179655758566958,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.6286295603584735,"spectral_convergence":0.33984992912927964,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.00045337075345531407,"delta_db":-0.18423611868569806,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.002466185364097462,"delta_db":1.3206239924237761,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0031441406059144047,"delta_db":2.380516785879111,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.0001791276437146913,"delta_db":1.1733431221621955,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00014705765951407895,"delta_db":0.767170499843829,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":9.11756864265343e-05,"delta_db":0.2139493794503647,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0066399830741626,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6263047979776708,"pitch":null,"output_rms_dbfs":-21.883541111839374,"output_peak":0.4622224764757385,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.68041181703613},"common_prefix":{"samples":52320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.167156766071802,"spectral_convergence":0.516843381859241,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.245449124678112,"spectral_convergence":0.3179655758566958,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.6286295603584735,"spectral_convergence":0.33984992912927964,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.00045337075345531407,"delta_db":-0.18423611868569806,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.002466185364097462,"delta_db":1.3206239924237761,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0031441406059144047,"delta_db":2.380516785879111,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.0001791276437146913,"delta_db":1.1733431221621955,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00014705765951407895,"delta_db":0.767170499843829,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":9.11756864265343e-05,"delta_db":0.2139493794503647,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0066399830741626,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6263047979776708,"pitch":null,"output_rms_dbfs":-21.883541111839374,"output_peak":0.4622224764757385,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.68041181703613},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52320,"output_samples":52320,"intelligibility":{"estoi":{"value":0.9259565179114607,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0667586744040302,"rms_dbfs":-23.509845909817045,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.978446332783722,"mean":4.524732583888809e-05,"zero_energy":false},"output":{"rms":0.08050501669537424,"rms_dbfs":-21.883541111839374,"peak":0.4622224764757385,"peak_dbfs":-6.70297879842559,"crest_factor_db":15.180562313413784,"mean":-2.443584082215118e-08,"zero_energy":false},"rms_difference_db":1.6263047979776708},"full_scale":{"reference":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":327,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":5.265872192532801e-07,"output_rms_dbfs":-62.78529685820786,"output_zero_energy":false,"output_peak":0.0035788185473445066,"output_energy_fraction":4.969431040257212e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":320,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":155,"both_estimated_voiced_frames":141,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.8924050632911392,"absolute_difference_cents_median":3.765384148107697,"absolute_difference_cents_p95":15.67060292199558,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00017439460044090568,"log_rms_mae_db":2.1178300589609935,"rms_envelope_nrmse":0.24609399562920126,"envelope_correlation":0.9835074027117896,"rms_gain_error_db":1.6263176702563356,"reference_mean_power":0.004463526965610992,"output_mean_power":0.006490974901482813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":1.3056377024951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008270090313661458,"output_power":0.0012587546984932489,"reference_active":true,"delta_db":1.824308523361737},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010887958720618103,"output_power":0.00016803021439806986,"reference_active":true,"delta_db":1.8844091586343343},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8937894537156728e-05,"output_power":3.9621153146865705e-05,"reference_active":true,"delta_db":1.3646018169176013}],"active_band_count":3,"active_band_mae_db":1.6911064996378908}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015806536675822195,"log_rms_mae_db":1.772563845553689,"rms_envelope_nrmse":0.22975780441479232,"envelope_correlation":0.9902892460141866,"rms_gain_error_db":1.6263639744219394,"reference_mean_power":0.004504823508022685,"output_mean_power":0.006551099235052457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.092227651107542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000711888936919089,"output_power":0.0011080090978578137,"reference_active":true,"delta_db":1.921310825433179},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.550528203435593e-05,"output_power":9.176253432195748e-05,"reference_active":true,"delta_db":1.4638907867380024},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5248696894099064e-06,"output_power":2.004324326647211e-06,"reference_active":true,"delta_db":1.1873526575112838}],"active_band_count":3,"active_band_mae_db":1.5241847565608218}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00016816220894332027,"log_rms_mae_db":2.0014108527615138,"rms_envelope_nrmse":0.3074454381566695,"envelope_correlation":0.9866595951927567,"rms_gain_error_db":2.0934620554318197,"reference_mean_power":0.002573247092376177,"output_mean_power":0.004167040258414676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":9.221294414644421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007324784888809201,"output_power":0.0011904648335494838,"reference_active":true,"delta_db":2.109216957914563},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.576574656453582e-05,"output_power":0.00015347159454311317,"reference_active":true,"delta_db":2.5271413301683863},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8048967652930613e-05,"output_power":2.7942312803109947e-05,"reference_active":true,"delta_db":1.898099834939206}],"active_band_count":3,"active_band_mae_db":2.1781527076740517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001473388036867024,"log_rms_mae_db":1.6678790007811632,"rms_envelope_nrmse":0.29490388898518155,"envelope_correlation":0.9913117407544487,"rms_gain_error_db":2.0934641168332764,"reference_mean_power":0.002597106693874505,"output_mean_power":0.004205679787873666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.911179205548874e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006603877113347142,"output_power":0.0010972450004929455,"reference_active":true,"delta_db":2.2050462749231823},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5067104939947466e-05,"output_power":7.580839726982757e-05,"reference_active":true,"delta_db":2.258576545667545},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.161011125548218e-06,"output_power":1.4407073081731158e-06,"reference_active":true,"delta_db":0.9373937774755581}],"active_band_count":3,"active_band_mae_db":1.800338866022095}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":1.46234603896658,"rms_envelope_nrmse":0.17566126459657294,"envelope_correlation":0.9894763043288354,"rms_gain_error_db":1.0371768382454085,"reference_mean_power":0.0001985957742888365,"output_mean_power":0.0002521666722093793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":8.760825576371805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.45270565717188e-05,"output_power":7.158708657846295e-05,"reference_active":true,"delta_db":1.1822263339759593},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.928893294666135e-06,"output_power":8.39184627140001e-06,"reference_active":true,"delta_db":1.5088388530747234},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9788817769833534e-06,"output_power":2.6480250495936786e-06,"reference_active":true,"delta_db":1.2650223987899198}],"active_band_count":3,"active_band_mae_db":1.3186958619468674}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":1.049495855769079,"rms_envelope_nrmse":0.14236017248980373,"envelope_correlation":0.9968099429493402,"rms_gain_error_db":1.0372781523060395,"reference_mean_power":0.000200416025269102,"output_mean_power":0.0002544838696686367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.857113792607277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022793901144949e-05,"output_power":6.609894508184925e-05,"reference_active":true,"delta_db":1.1924917009428668},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6196674175617505e-06,"output_power":3.769032535357041e-06,"reference_active":true,"delta_db":1.5798372788000985},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.185526690136049e-07,"output_power":1.5462032046042134e-07,"reference_active":true,"delta_db":1.1535523320551235}],"active_band_count":3,"active_band_mae_db":1.3086271039326964}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":1.574466169697952,"rms_envelope_nrmse":0.32593403481827715,"envelope_correlation":0.900956232816372,"rms_gain_error_db":0.28801531419872706,"reference_mean_power":8.71199669251843e-05,"output_mean_power":9.309347313654861e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":4.20340153760379e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.824817597815015e-05,"output_power":2.055421038954753e-05,"reference_active":true,"delta_db":0.5168133712997882},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2434318429437933e-05,"output_power":1.3541658450348962e-05,"reference_active":true,"delta_db":0.37049870775636873},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7676697520375684e-06,"output_power":4.6855334835532666e-06,"reference_active":true,"delta_db":-1.5968011213400335}],"active_band_count":3,"active_band_mae_db":0.8280377334653969}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":1.0542378240786745,"rms_envelope_nrmse":0.18124532186442208,"envelope_correlation":0.9644910684369687,"rms_gain_error_db":0.2881867701925717,"reference_mean_power":8.789733624317312e-05,"output_mean_power":9.392785202666145e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":3.223393789064046e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.051966991658495e-05,"output_power":2.2062197522776584e-05,"reference_active":true,"delta_db":0.3147839815670375},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.88053445243811e-06,"output_power":6.230681944396409e-06,"reference_active":true,"delta_db":0.25118783802146694},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.871896444237528e-07,"output_power":3.9440290701121937e-07,"reference_active":true,"delta_db":-1.7283828051269183}],"active_band_count":3,"active_band_mae_db":0.7647848749051409}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"facodec","status":"measured","mos":{"status":"measured","value":4.072702407836914,"predicted_mos":4.072702407836914,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":52400,"input_sample_rate":16000,"input_seconds":3.275,"input_float32_sha256":"0159519d92a6830ce3e2dafd446d264423dcbde7b82c09d253a919703ecde20a","internal_samples":52400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.87879026399969},"started_at":"2026-10-13T04:46:22.230010+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.163000059837941e-05,"native_input_samples":52320,"native_input_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","output_sha256":"c4601d29dff38bdb24fcbc0106f27c48fc346ae39f3dff8f4cae0d28b2f4e0ff","output_samples":52400,"length_delta_samples":80,"analysis_s":0.9396574080001301,"synthesis_s":5.255341451999811,"total_s":6.1949988599999415,"rtf":1.8944950642201657,"output_peak":0.3154899477958679,"output_rms":0.06251252560380298,"finished_at":"2026-10-13T04:46:28.428817+00:00","evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"c4601d29dff38bdb24fcbc0106f27c48fc346ae39f3dff8f4cae0d28b2f4e0ff","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":52320,"output_samples":52400,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.080659091328847,"output_peak":0.3154899477958679},"common_prefix":{"samples":52320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9984732824427481,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.380900832412949,"spectral_convergence":0.23658575835525217,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.93346990745673,"spectral_convergence":0.23692726962301203,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.106490376822336,"spectral_convergence":0.2516257812473238,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.0004408056202624455,"delta_db":-0.3062998948672634,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.0017342011111654895,"delta_db":-0.20863856648860285,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0014686985254481104,"delta_db":-0.9251762200953209,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.0001085842255020606,"delta_db":-1.0006156290952566,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":9.054461703676986e-05,"delta_db":-1.3390796352351009,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":7.095674782417263e-05,"delta_db":-0.8749040152076997,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7757856601648742,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.564177957486873,"pitch":null,"output_rms_dbfs":-24.074023867303914,"output_peak":0.3154899477958679,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8069537055640055},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52320,"output_samples":52320,"intelligibility":{"estoi":{"value":0.8820989146045936,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0667586744040302,"rms_dbfs":-23.509845909817045,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.978446332783722,"mean":4.524732583888809e-05,"zero_energy":false},"output":{"rms":0.06256029769141966,"rms_dbfs":-24.074023867303914,"peak":0.3154899477958679,"peak_dbfs":-10.020289475549639,"crest_factor_db":14.053734391754277,"mean":2.0268234837256483e-05,"zero_energy":false},"rms_difference_db":-0.5641779574868693},"full_scale":{"reference":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":327,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":1.968021772466592e-06,"output_rms_dbfs":-57.059701012249036,"output_zero_energy":false,"output_peak":0.0044099618680775166,"output_energy_fraction":3.075491481904555e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":320,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":164,"both_estimated_voiced_frames":148,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9367088607594937,"absolute_difference_cents_median":5.590444890095033,"absolute_difference_cents_p95":16.743361166240224,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00017439460044090568,"log_rms_mae_db":1.2790889099750065,"rms_envelope_nrmse":0.12792356271681712,"envelope_correlation":0.983619803869256,"rms_gain_error_db":-0.5641684777380066,"reference_mean_power":0.004463526965610992,"output_mean_power":0.00391977659184307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":1.3056377024951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008270090313661458,"output_power":0.000686692277189819,"reference_active":true,"delta_db":-0.8074808912278906},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010887958720618103,"output_power":0.0001027099380232395,"reference_active":true,"delta_db":-0.2533399850363984},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8937894537156728e-05,"output_power":2.3553583753661745e-05,"reference_active":true,"delta_db":-0.8940993374457564}],"active_band_count":3,"active_band_mae_db":0.6516400712366818}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015806536675822195,"log_rms_mae_db":1.0378078059590747,"rms_envelope_nrmse":0.11812516270955176,"envelope_correlation":0.9852475907758494,"rms_gain_error_db":-0.5641319817287817,"reference_mean_power":0.004504823508022685,"output_mean_power":0.003956075601709699},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.092227651107542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000711888936919089,"output_power":0.0005852112121164544,"reference_active":true,"delta_db":-0.8509960563214785},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.550528203435593e-05,"output_power":6.208438501721193e-05,"reference_active":true,"delta_db":-0.2329393731258317},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5248696894099064e-06,"output_power":1.3383143524041143e-06,"reference_active":true,"delta_db":-0.5667459635719339}],"active_band_count":3,"active_band_mae_db":0.5502271310064147}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00016816220894332027,"log_rms_mae_db":1.5140328617995342,"rms_envelope_nrmse":0.14267147408880707,"envelope_correlation":0.9843110871387472,"rms_gain_error_db":-0.5958309058521555,"reference_mean_power":0.002573247092376177,"output_mean_power":0.0022433570462033836},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":9.221294414644421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007324784888809201,"output_power":0.0005667444512851912,"reference_active":true,"delta_db":-1.1140759817964838},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.576574656453582e-05,"output_power":8.35614140852935e-05,"reference_active":true,"delta_db":-0.11308091559454134},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8048967652930613e-05,"output_power":1.1346356858358708e-05,"reference_active":true,"delta_db":-2.015959280083969}],"active_band_count":3,"active_band_mae_db":1.081038725824998}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001473388036867024,"log_rms_mae_db":1.2118896838966036,"rms_envelope_nrmse":0.13063228371334337,"envelope_correlation":0.9863333268417109,"rms_gain_error_db":-0.5958279924193433,"reference_mean_power":0.002597106693874505,"output_mean_power":0.0022641593678264117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.911179205548874e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006603877113347142,"output_power":0.0005096454543506026,"reference_active":true,"delta_db":-1.1253082821273575},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5067104939947466e-05,"output_power":4.267925068265247e-05,"reference_active":true,"delta_db":-0.2364287444463698},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.161011125548218e-06,"output_power":8.393251472714364e-07,"reference_active":true,"delta_db":-1.4090614611017316}],"active_band_count":3,"active_band_mae_db":0.9235994958918196}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":1.7803327915750484,"rms_envelope_nrmse":0.19592150583227522,"envelope_correlation":0.9794486097593399,"rms_gain_error_db":-1.1662507205159847,"reference_mean_power":0.0001985957742888365,"output_mean_power":0.00015182557385819568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":8.760825576371805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.45270565717188e-05,"output_power":3.898533546481908e-05,"reference_active":true,"delta_db":-1.4571077901100726},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.928893294666135e-06,"output_power":3.8311698230224635e-06,"reference_active":true,"delta_db":-1.896422309543064},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9788817769833534e-06,"output_power":1.6936915310407263e-06,"reference_active":true,"delta_db":-0.6758553328294703}],"active_band_count":3,"active_band_mae_db":1.3431284774942023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":1.4895292737348111,"rms_envelope_nrmse":0.17230954495366876,"envelope_correlation":0.9864702854086252,"rms_gain_error_db":-1.1660131074353939,"reference_mean_power":0.000200416025269102,"output_mean_power":0.00015322553063638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.857113792607277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022793901144949e-05,"output_power":3.584925566454727e-05,"reference_active":true,"delta_db":-1.4646521537167727},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6196674175617505e-06,"output_power":1.6714384241304294e-06,"reference_active":true,"delta_db":-1.95155776744526},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.185526690136049e-07,"output_power":9.235555736296646e-08,"reference_active":true,"delta_db":-1.0844830222383008}],"active_band_count":3,"active_band_mae_db":1.5002309811334447}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":1.9171352680095106,"rms_envelope_nrmse":0.21716828751501885,"envelope_correlation":0.9617987142405396,"rms_gain_error_db":-0.8642753850435789,"reference_mean_power":8.71199669251843e-05,"output_mean_power":7.139867681661839e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":4.20340153760379e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.824817597815015e-05,"output_power":1.63267645978751e-05,"reference_active":true,"delta_db":-0.4831932938976435},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2434318429437933e-05,"output_power":1.1426284392249607e-05,"reference_active":true,"delta_db":-0.3671695584039518},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7676697520375684e-06,"output_power":7.126837722861585e-06,"reference_active":true,"delta_db":0.22457712186123885}],"active_band_count":3,"active_band_mae_db":0.35831332472094474}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":1.5843374002467665,"rms_envelope_nrmse":0.16931141505663772,"envelope_correlation":0.976911487040667,"rms_gain_error_db":-0.8632089668881205,"reference_mean_power":8.789733624317312e-05,"output_mean_power":7.205345613810343e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":3.223393789064046e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.051966991658495e-05,"output_power":1.9212795054188326e-05,"reference_active":true,"delta_db":-0.2857982025632467},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.88053445243811e-06,"output_power":5.869209784265836e-06,"reference_active":true,"delta_db":-0.008371658685439097},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.871896444237528e-07,"output_power":6.355560606045531e-07,"reference_active":true,"delta_db":0.34375476636667623}],"active_band_count":3,"active_band_mae_db":0.21264154253845402}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.7117233276367188,"predicted_mos":3.7117233276367188,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144384,"input_sample_rate":44100,"input_seconds":3.274013605442177,"input_float32_sha256":"004e9f48a9027375b0440ea07c211111a93ffc5a238874972866e8cc76275ced","internal_samples":52385,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9576635680004983},"started_at":"2026-10-13T04:59:58.384999+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004530675999376399,"native_input_samples":144207,"native_input_sha256":"dc46503355b563d4fea4f95205b0d31631c2aae838e36a3654682efd983d4e27","output_sha256":"6f921b424dc861f3ed1a8746e573d18f5b9ce9aa9b3ba1874922567f9685bcbc","output_samples":144384,"length_delta_samples":177,"analysis_s":0.07148904299992864,"synthesis_s":18.089501654999367,"total_s":18.160990697999296,"rtf":5.5538197853208855,"output_peak":0.38143959641456604,"output_rms":0.06230796839168864,"finished_at":"2026-10-13T05:00:16.559145+00:00","evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"6f921b424dc861f3ed1a8746e573d18f5b9ce9aa9b3ba1874922567f9685bcbc","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":52320,"output_samples":52385,"length_delta_samples":65,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.103425576246195,"output_peak":0.3698088236549347},"common_prefix":{"samples":52320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987591867901117,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.262684472367841,"spectral_convergence":0.4202255607473426,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.224481612499925,"spectral_convergence":0.36872960005045724,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.98030074400633,"spectral_convergence":0.49151629487800136,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.00027405408515170223,"delta_db":-2.3704083240044413,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.001681184948369522,"delta_db":-0.3434782368747463,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0016549008090377291,"delta_db":-0.4067831257982262,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.00012817914950252204,"delta_db":-0.2801091529944478,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00010655392077579592,"delta_db":-0.6320116500656731,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":4.733336094533383e-05,"delta_db":-2.633167606786073,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1109930160872679,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5881956067180362,"pitch":null,"output_rms_dbfs":-24.098041516535076,"output_peak":0.3698088236549347,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.155822276291365},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52320,"output_samples":52320,"intelligibility":{"estoi":{"value":0.7012581158227595,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0667586744040302,"rms_dbfs":-23.509845909817045,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.978446332783722,"mean":4.524732583888809e-05,"zero_energy":false},"output":{"rms":0.062387549028492896,"rms_dbfs":-24.098041516535076,"peak":0.3698088236549347,"peak_dbfs":-8.640454615352414,"crest_factor_db":15.457586901182665,"mean":-8.894938335741052e-05,"zero_energy":false},"rms_difference_db":-0.5881956067180312},"full_scale":{"reference":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":327,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":9.381297301818438e-07,"output_rms_dbfs":-60.27737100633367,"output_zero_energy":false,"output_peak":0.0035071604085878073,"output_energy_fraction":1.474175855869385e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":320,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":140,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.8860759493670886,"absolute_difference_cents_median":14.773387544674245,"absolute_difference_cents_p95":43.02770462518227,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00017439460044090568,"log_rms_mae_db":1.4695143957412196,"rms_envelope_nrmse":0.11334637726883735,"envelope_correlation":0.9895449179061964,"rms_gain_error_db":-0.5881906900342118,"reference_mean_power":0.004463526965610992,"output_mean_power":0.00389815491124593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":1.3056377024951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008270090313661458,"output_power":0.0007280383054820618,"reference_active":true,"delta_db":-0.5535602213149076},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010887958720618103,"output_power":0.00010709554192898414,"reference_active":true,"delta_db":-0.07175072851270876},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8937894537156728e-05,"output_power":2.212192240219511e-05,"reference_active":true,"delta_db":-1.166440649386862}],"active_band_count":3,"active_band_mae_db":0.5972505330714928}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015806536675822195,"log_rms_mae_db":1.0804265825135824,"rms_envelope_nrmse":0.093750425744136,"envelope_correlation":0.9941450274425829,"rms_gain_error_db":-0.5881689265802984,"reference_mean_power":0.004504823508022685,"output_mean_power":0.0039342403478563204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.092227651107542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000711888936919089,"output_power":0.0006289079886965852,"reference_active":true,"delta_db":-0.5382513246949763},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.550528203435593e-05,"output_power":6.289675570019575e-05,"reference_active":true,"delta_db":-0.1764807631927786},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5248696894099064e-06,"output_power":1.2112719092286837e-06,"reference_active":true,"delta_db":-0.9999108626554731}],"active_band_count":3,"active_band_mae_db":0.571547650181076}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00016816220894332027,"log_rms_mae_db":2.287353486337865,"rms_envelope_nrmse":0.13788584977551321,"envelope_correlation":0.9824654931686726,"rms_gain_error_db":-0.13489711511304045,"reference_mean_power":0.002573247092376177,"output_mean_power":0.0024945475070274623},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":9.221294414644421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007324784888809201,"output_power":0.0006416089472723665,"reference_active":true,"delta_db":-0.5752446345741569},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.576574656453582e-05,"output_power":9.752607735684519e-05,"reference_active":true,"delta_db":0.5580688423231603},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8048967652930613e-05,"output_power":1.652266622293316e-05,"reference_active":true,"delta_db":-0.38372236908135027}],"active_band_count":3,"active_band_mae_db":0.5056786153262225}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001473388036867024,"log_rms_mae_db":1.8175807678123554,"rms_envelope_nrmse":0.11534560768475932,"envelope_correlation":0.9865663323997409,"rms_gain_error_db":-0.13491368217262345,"reference_mean_power":0.002597106693874505,"output_mean_power":0.002517667787889555},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.911179205548874e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006603877113347142,"output_power":0.0005782163207408812,"reference_active":true,"delta_db":-0.5770863729297533},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5067104939947466e-05,"output_power":5.16537798269125e-05,"reference_active":true,"delta_db":0.5924244665613028},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.161011125548218e-06,"output_power":1.2321577178788349e-06,"reference_active":true,"delta_db":0.25829920231309367}],"active_band_count":3,"active_band_mae_db":0.4759366806013832}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":2.8201038049645843,"rms_envelope_nrmse":0.2578332683289101,"envelope_correlation":0.9456921013880099,"rms_gain_error_db":-0.5108155384127865,"reference_mean_power":0.0001985957742888365,"output_mean_power":0.00017655842642237113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":8.760825576371805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.45270565717188e-05,"output_power":4.065255269507826e-05,"reference_active":true,"delta_db":-1.2752423321067576},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.928893294666135e-06,"output_power":4.622895780524685e-06,"reference_active":true,"delta_db":-1.080595314668487},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9788817769833534e-06,"output_power":1.3599433215076673e-06,"reference_active":true,"delta_db":-1.6289904060197031}],"active_band_count":3,"active_band_mae_db":1.328276017598316}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":1.9666219878880347,"rms_envelope_nrmse":0.22241125452499994,"envelope_correlation":0.9573563028759619,"rms_gain_error_db":-0.5111005543766418,"reference_mean_power":0.000200416025269102,"output_mean_power":0.00017816499883428432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.857113792607277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022793901144949e-05,"output_power":3.6781991970342916e-05,"reference_active":true,"delta_db":-1.3531011299737459},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6196674175617505e-06,"output_power":2.042864267185108e-06,"reference_active":true,"delta_db":-1.0800664651764},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.185526690136049e-07,"output_power":9.754152686067872e-08,"reference_active":true,"delta_db":-0.8472178632038747}],"active_band_count":3,"active_band_mae_db":1.0934618194513404}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":3.564565801861323,"rms_envelope_nrmse":0.3260622327239774,"envelope_correlation":0.9287680662474909,"rms_gain_error_db":-2.120802414723117,"reference_mean_power":8.71199669251843e-05,"output_mean_power":5.346104706361971e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":4.20340153760379e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.824817597815015e-05,"output_power":1.404151798052202e-05,"reference_active":true,"delta_db":-1.1380540002799657},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2434318429437933e-05,"output_power":6.418205077975364e-06,"reference_active":true,"delta_db":-2.872083951918069},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7676697520375684e-06,"output_power":2.0617632229439993e-06,"reference_active":true,"delta_db":-5.162003693413308}],"active_band_count":3,"active_band_mae_db":3.057380548537114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":3.0821125979192634,"rms_envelope_nrmse":0.2919182046867029,"envelope_correlation":0.9458877983895356,"rms_gain_error_db":-2.119989510519842,"reference_mean_power":8.789733624317312e-05,"output_mean_power":5.394817564169724e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":3.223393789064046e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.051966991658495e-05,"output_power":1.448614090527789e-05,"reference_active":true,"delta_db":-1.5121766512491748},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.88053445243811e-06,"output_power":2.771982073315931e-06,"reference_active":true,"delta_db":-3.266263814054512},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.871896444237528e-07,"output_power":1.7405699282561604e-07,"reference_active":true,"delta_db":-5.280869118662452}],"active_band_count":3,"active_band_mae_db":3.35310319465538}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"golf","status":"measured","mos":{"status":"measured","value":3.541219472885132,"predicted_mos":3.541219472885132,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78240,"input_sample_rate":24000,"input_seconds":3.26,"input_float32_sha256":"649f853d105f3cbf7b50e0365c6988a2bc736acb0b3715fd6553fd7716156f83","internal_samples":52160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3869922949998},"started_at":"2026-10-13T05:01:58.313652+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0015962039997248212,"native_input_samples":78480,"native_input_sha256":"6ee6514f6cbc5f558e7996ec0e7fac4e87bb542c3bcdd1371a2075a14ab17381","output_sha256":"63cd3bffa7bda02652d84e1f038eedc57a3c3050234273452c7595662255a232","output_samples":78240,"length_delta_samples":-240,"analysis_s":0.35060077300022385,"synthesis_s":0.07089455099958286,"total_s":0.4214953239998067,"rtf":0.12889765259932928,"output_peak":0.38114476203918457,"output_rms":0.06517301039129561,"finished_at":"2026-10-13T05:01:58.742343+00:00","analysis_stages":{"external_f0_s":0.04047326799991424},"evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"63cd3bffa7bda02652d84e1f038eedc57a3c3050234273452c7595662255a232","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":52320,"output_samples":52160,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.71336256702151,"output_peak":0.3783732880166155},"common_prefix":{"samples":52160,"sample_rate":16000,"reference_coverage":0.9969418960244648,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52160,"output_samples":52160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.172525651533286,"spectral_convergence":0.4212971242733799,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.293418189529517,"spectral_convergence":0.21903741007796879,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4232918104121355,"spectral_convergence":0.24971480676911875,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047440408619083875,"output_power":0.0004374160399716769,"delta_db":-0.352537165882903,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018245958411121708,"output_power":0.0018663643258771227,"delta_db":0.0982974403815009,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018233144875127076,"output_power":0.0016620924807472818,"delta_db":-0.40206397991697407,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013733271447142958,"output_power":0.00012970474892089095,"delta_db":-0.2481812692220672,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012363585623747229,"output_power":9.285843392058539e-05,"delta_db":-1.2432308596751669,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.705148055006551e-05,"output_power":6.425414038703092e-05,"delta_db":-1.3187504406028008,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6105101926135689,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21676274540239548,"pitch":null,"output_rms_dbfs":-23.71336256702151,"output_peak":0.3783732880166155,"reference_rms_dbfs":-23.496599821619117,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.629745217158312},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52160,"output_samples":52160,"intelligibility":{"estoi":{"value":0.8940367292875369,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06686055986958743,"rms_dbfs":-23.496599821619117,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.965200244585795,"mean":4.016607085619967e-05,"zero_energy":false},"output":{"rms":0.06521265344873332,"rms_dbfs":-23.71336256702151,"peak":0.3783732880166155,"peak_dbfs":-8.441590620693882,"crest_factor_db":15.27177194632763,"mean":0.0005451358485535666,"zero_energy":false},"rms_difference_db":-0.2167627454023915},"full_scale":{"reference":{"sample_count":52160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":326,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":1.67298893697801e-06,"output_rms_dbfs":-57.76506930899703,"output_zero_energy":false,"output_peak":0.003457487886838317,"output_energy_fraction":2.4134689807662125e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":319,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":166,"both_estimated_voiced_frames":152,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9620253164556962,"absolute_difference_cents_median":6.379345174319788,"absolute_difference_cents_p95":18.074026749454816,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":651,"rms_floor":0.00017439460044090568,"log_rms_mae_db":1.5671435581211057,"rms_envelope_nrmse":0.1006940311844064,"envelope_correlation":0.9884233159808241,"rms_gain_error_db":-0.21675380699767044,"reference_mean_power":0.004477181642921785,"output_mean_power":0.004259212748643658},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":651,"frequency_resolution_hz":0.30721966205837176,"power_floor":1.302478123344161e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008272442696230258,"output_power":0.0008076695955756401,"reference_active":true,"delta_db":-0.10400033121799812},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001092499931016536,"output_power":0.00011474886587013469,"reference_active":true,"delta_db":0.21326983625328155},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8222595967027185e-05,"output_power":2.3609934237986634e-05,"reference_active":true,"delta_db":-0.7750218105553777}],"active_band_count":3,"active_band_mae_db":0.3640973260088858}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":0.00015806536675822195,"log_rms_mae_db":0.9676003452407738,"rms_envelope_nrmse":0.0705322860545569,"envelope_correlation":0.9941589779985074,"rms_gain_error_db":-0.2167140928226261,"reference_mean_power":0.0045187221905970115,"output_mean_power":0.00429877022979067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":1.0888821725908645e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007139037977762057,"output_power":0.0006866104854497352,"reference_active":true,"delta_db":-0.16929260903382337},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.31051685411487e-05,"output_power":6.787555553473419e-05,"reference_active":true,"delta_db":0.31648466154027877},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5135319473039672e-06,"output_power":1.5031657614579768e-06,"reference_active":true,"delta_db":-0.02984717373603582}],"active_band_count":3,"active_band_mae_db":0.17187481477004599}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":651,"rms_floor":0.00016816220894332027,"log_rms_mae_db":1.7946466552301035,"rms_envelope_nrmse":0.11895214675416177,"envelope_correlation":0.9870103576203616,"rms_gain_error_db":-0.14682312169912534,"reference_mean_power":0.0025811503858213384,"output_mean_power":0.002495347286243083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":651,"frequency_resolution_hz":0.30721966205837176,"power_floor":9.210892947152988e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007321350268305062,"output_power":0.0006938779956762082,"reference_active":true,"delta_db":-0.2330806956133054},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.605382431527204e-05,"output_power":8.150047515166918e-05,"reference_active":true,"delta_db":-0.23610034854764272},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.740290314150437e-05,"output_power":1.4698518413292517e-05,"reference_active":true,"delta_db":-0.7334814228406995}],"active_band_count":3,"active_band_mae_db":0.4008874890005492}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":0.0001473388036867024,"log_rms_mae_db":1.2043820684395992,"rms_envelope_nrmse":0.08538973674331617,"envelope_correlation":0.9927655238193519,"rms_gain_error_db":-0.14682461699879268,"reference_mean_power":0.0026051573218816623,"output_mean_power":0.002518555311946321},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":7.902499983752958e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006616241865456379,"output_power":0.0006221398468527388,"reference_active":true,"delta_db":-0.26723355088583056},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.319180651867406e-05,"output_power":4.0141659195355366e-05,"reference_active":true,"delta_db":-0.31806049574574236},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1562870660817352e-06,"output_power":1.2533152479098255e-06,"reference_active":true,"delta_db":0.34994655586506607}],"active_band_count":3,"active_band_mae_db":0.31174686749887964}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":651,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":1.6470423008262602,"rms_envelope_nrmse":0.15615921742539504,"envelope_correlation":0.9832064302147621,"rms_gain_error_db":-0.5934423882734833,"reference_mean_power":0.0001991899383463661,"output_mean_power":0.00017374933791673423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":651,"frequency_resolution_hz":0.30721966205837176,"power_floor":8.796146169207539e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.443067843725354e-05,"output_power":4.327907507010426e-05,"reference_active":true,"delta_db":-0.9956577674112245},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942904802704991e-06,"output_power":4.489778813415931e-06,"reference_active":true,"delta_db":-1.2177382726203074},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9452344587137615e-06,"output_power":1.1323333445637652e-06,"reference_active":true,"delta_db":-2.3499765778448642}],"active_band_count":3,"active_band_mae_db":1.5211242059587988}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":1.2744783786633465,"rms_envelope_nrmse":0.1176148028045767,"envelope_correlation":0.9917876322288638,"rms_gain_error_db":-0.5934576295781822,"reference_mean_power":0.00020102068440386152,"output_mean_power":0.0001753456451563672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":7.89303153146948e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.01472067957796e-05,"output_power":3.9551878495507094e-05,"reference_active":true,"delta_db":-1.030796328761012},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.587727965571062e-06,"output_power":2.039850421616554e-06,"reference_active":true,"delta_db":-1.0332029666994549},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1842915528172704e-07,"output_power":6.80106147362224e-08,"reference_active":true,"delta_db":-2.4088193122675348}],"active_band_count":3,"active_band_mae_db":1.4909395359093338}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":651,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":2.3571892959528062,"rms_envelope_nrmse":0.24021730846107023,"envelope_correlation":0.9566381457396417,"rms_gain_error_db":-0.8775678866805253,"reference_mean_power":8.736707940821119e-05,"output_mean_power":7.138238087942383e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":651,"frequency_resolution_hz":0.30721966205837176,"power_floor":4.207244503056726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8248088764824364e-05,"output_power":1.6318932553063717e-05,"reference_active":true,"delta_db":-0.485256373630876},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2520109305804558e-05,"output_power":1.4074046118041153e-05,"reference_active":true,"delta_db":0.5081084933233819},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.731342349875329e-06,"output_power":7.507916482802475e-06,"reference_active":true,"delta_db":0.47417753998417794}],"active_band_count":3,"active_band_mae_db":0.4891808023128119}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52160,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":1.970815121102944,"rms_envelope_nrmse":0.18865001031318562,"envelope_correlation":0.9744902431982467,"rms_gain_error_db":-0.8766710551403712,"reference_mean_power":8.814526895871144e-05,"output_mean_power":7.203306622753028e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":3.2252161856776474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.050163006823651e-05,"output_power":2.0487952783933427e-05,"reference_active":true,"delta_db":-0.002898282396347237},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.924869853784549e-06,"output_power":7.214220011554427e-06,"reference_active":true,"delta_db":0.8551056779632633},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.801895937786788e-07,"output_power":6.953744804970015e-07,"reference_active":true,"delta_db":0.7864881341962983}],"active_band_count":3,"active_band_mae_db":0.5481640315186364}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.7814202308654785,"predicted_mos":3.7814202308654785,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78600,"input_sample_rate":24000,"input_seconds":3.275,"input_float32_sha256":"15a0562e55dd7e7e69bc7daf25e3448a8e013eb792bbf63d7890d3cbd615e287","internal_samples":52400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.422224998999809},"started_at":"2026-10-13T05:05:20.874030+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0025452299996686634,"native_input_samples":78480,"native_input_sha256":"6ee6514f6cbc5f558e7996ec0e7fac4e87bb542c3bcdd1371a2075a14ab17381","output_sha256":"7170636182baa97f46ae2043fc185cb500cddc56dacf3558c6c0ff6f92e94a00","output_samples":78600,"length_delta_samples":120,"analysis_s":1.1995807769999374,"synthesis_s":7.670629657000063,"total_s":8.870210434,"rtf":2.712602579204893,"output_peak":0.4057076871395111,"output_rms":0.06564166459872711,"finished_at":"2026-10-13T05:05:29.756955+00:00","analysis_stages":{"external_f0_s":0.7775452779997067},"evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"7170636182baa97f46ae2043fc185cb500cddc56dacf3558c6c0ff6f92e94a00","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":52320,"output_samples":52400,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.651633067297446,"output_peak":0.40078379998495944},"common_prefix":{"samples":52320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9984732824427481,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.875901962879316,"spectral_convergence":0.4411415372902507,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.27151453311116,"spectral_convergence":0.24582456549199577,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.1851988601671115,"spectral_convergence":0.24180310877498404,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.0005131979450352805,"delta_db":0.35407796078269294,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.001770336392691643,"delta_db":-0.1190751959029973,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0017200564384974696,"delta_db":-0.23907583642284544,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.00013136210900807825,"delta_db":-0.17358188521254114,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00012175211297829142,"delta_db":-0.05294092695692642,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":6.344347064880115e-05,"delta_db":-1.360971708292633,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3832872522617727,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13516736717627847,"pitch":null,"output_rms_dbfs":-23.645013276993325,"output_peak":0.40078379998495944,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.444205118719196},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52320,"output_samples":52320,"intelligibility":{"estoi":{"value":0.9032309740149085,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0667586744040302,"rms_dbfs":-23.509845909817045,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.978446332783722,"mean":4.524732583888809e-05,"zero_energy":false},"output":{"rms":0.06572783633179753,"rms_dbfs":-23.645013276993325,"peak":0.40078379998495944,"peak_dbfs":-7.941796826615061,"crest_factor_db":15.70321645037826,"mean":6.564284294680251e-05,"zero_energy":false},"rms_difference_db":-0.13516736717627964},"full_scale":{"reference":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":327,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":2.051992955385808e-07,"output_rms_dbfs":-66.8782413451652,"output_zero_energy":false,"output_peak":0.002850929190698076,"output_energy_fraction":2.9050889615800787e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":320,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":172,"both_estimated_voiced_frames":155,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9810126582278481,"absolute_difference_cents_median":7.062447303039801,"absolute_difference_cents_p95":26.793805657365812,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00017439460044090568,"log_rms_mae_db":2.6736987942042663,"rms_envelope_nrmse":0.12944865620340731,"envelope_correlation":0.9797513491012085,"rms_gain_error_db":-0.13515244875188848,"reference_mean_power":0.004463526965610992,"output_mean_power":0.004326761114476334},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":1.3056377024951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008270090313661458,"output_power":0.0008369665168683351,"reference_active":true,"delta_db":0.051978319428512054},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010887958720618103,"output_power":0.00010648045194467652,"reference_active":true,"delta_db":-0.09676579877015269},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8937894537156728e-05,"output_power":2.4892770847763626e-05,"reference_active":true,"delta_db":-0.6539368834038285}],"active_band_count":3,"active_band_mae_db":0.2675603338674977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015806536675822195,"log_rms_mae_db":2.187837528523909,"rms_envelope_nrmse":0.08275517029196662,"envelope_correlation":0.9910582257818431,"rms_gain_error_db":-0.13511010915794738,"reference_mean_power":0.004504823508022685,"output_mean_power":0.0043668348718031505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.092227651107542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000711888936919089,"output_power":0.000718593623336891,"reference_active":true,"delta_db":0.04071115185309053},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.550528203435593e-05,"output_power":6.569753864791603e-05,"reference_active":true,"delta_db":0.012727782178129754},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5248696894099064e-06,"output_power":1.3156386784310977e-06,"reference_active":true,"delta_db":-0.6409609900990115}],"active_band_count":3,"active_band_mae_db":0.23146664137674391}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00016816220894332027,"log_rms_mae_db":2.107529534405053,"rms_envelope_nrmse":0.15679191247097865,"envelope_correlation":0.977436459774655,"rms_gain_error_db":-0.17946867401793676,"reference_mean_power":0.002573247092376177,"output_mean_power":0.002469076939617103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":9.221294414644421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007324784888809201,"output_power":0.0007034062786106906,"reference_active":true,"delta_db":-0.17588634521767915},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.576574656453582e-05,"output_power":8.493319868358978e-05,"reference_active":true,"delta_db":-0.04236391937154003},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8048967652930613e-05,"output_power":1.7736411561996398e-05,"reference_active":true,"delta_db":-0.07586608857315094}],"active_band_count":3,"active_band_mae_db":0.09803878438745671}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001473388036867024,"log_rms_mae_db":1.5300413737235248,"rms_envelope_nrmse":0.10295433227913442,"envelope_correlation":0.9895406828644343,"rms_gain_error_db":-0.1794665453853528,"reference_mean_power":0.002597106693874505,"output_mean_power":0.0024919718784659393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.911179205548874e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006603877113347142,"output_power":0.0006340115512109599,"reference_active":true,"delta_db":-0.17701812731116617},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5067104939947466e-05,"output_power":4.405536022716481e-05,"reference_active":true,"delta_db":-0.09860903693007432},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.161011125548218e-06,"output_power":1.1463545981001633e-06,"reference_active":true,"delta_db":-0.05517404147230815}],"active_band_count":3,"active_band_mae_db":0.11026706857118289}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":1.8766367499037497,"rms_envelope_nrmse":0.16515582253886502,"envelope_correlation":0.9778706465520335,"rms_gain_error_db":-0.13048381319272953,"reference_mean_power":0.0001985957742888365,"output_mean_power":0.00019271770804599433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":8.760825576371805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.45270565717188e-05,"output_power":5.1703466167291613e-05,"reference_active":true,"delta_db":-0.2309239574921128},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.928893294666135e-06,"output_power":5.494013819109266e-06,"reference_active":true,"delta_db":-0.33083886726619605},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9788817769833534e-06,"output_power":1.582083184832633e-06,"reference_active":true,"delta_db":-0.971905345517126}],"active_band_count":3,"active_band_mae_db":0.511222723425145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":1.1970299279662513,"rms_envelope_nrmse":0.0954288304223747,"envelope_correlation":0.9922045137013994,"rms_gain_error_db":-0.1303927893432153,"reference_mean_power":0.000200416025269102,"output_mean_power":0.00019448815921408876},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.857113792607277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022793901144949e-05,"output_power":4.7215616478077463e-05,"reference_active":true,"delta_db":-0.26859693751801067},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6196674175617505e-06,"output_power":2.4302599260228732e-06,"reference_active":true,"delta_db":-0.32593432923849974},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.185526690136049e-07,"output_power":9.416305494422404e-08,"reference_active":true,"delta_db":-1.0003079589659134}],"active_band_count":3,"active_band_mae_db":0.5316130752408079}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":2.092753564289766,"rms_envelope_nrmse":0.32726589146518104,"envelope_correlation":0.8937443980214498,"rms_gain_error_db":-0.8980646978927795,"reference_mean_power":8.71199669251843e-05,"output_mean_power":7.084533073148619e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":4.20340153760379e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.824817597815015e-05,"output_power":1.3954012694865233e-05,"reference_active":true,"delta_db":-1.1652034670300897},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2434318429437933e-05,"output_power":9.000342688659766e-06,"reference_active":true,"delta_db":-1.4036293944015654},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7676697520375684e-06,"output_power":3.893438607248764e-06,"reference_active":true,"delta_db":-2.401058268783843}],"active_band_count":3,"active_band_mae_db":1.6566303767384996}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":1.322222156310663,"rms_envelope_nrmse":0.1999548991896612,"envelope_correlation":0.961273778869411,"rms_gain_error_db":-0.8978401438160797,"reference_mean_power":8.789733624317312e-05,"output_mean_power":7.148117774532648e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":3.223393789064046e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.051966991658495e-05,"output_power":1.4746239268720721e-05,"reference_active":true,"delta_db":-1.434890939743267},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.88053445243811e-06,"output_power":3.841669087076107e-06,"reference_active":true,"delta_db":-1.848968457631606},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.871896444237528e-07,"output_power":3.006105702217378e-07,"reference_active":true,"delta_db":-2.907741404294689}],"active_band_count":3,"active_band_mae_db":2.063866933889854}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.831763505935669,"predicted_mos":3.831763505935669,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78720,"input_sample_rate":24000,"input_seconds":3.28,"input_float32_sha256":"caa81e1a8f3a3b64f5b292a92c036a8425088cd05e1a41bb2eb8c7cb3117de87","internal_samples":52480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4969280019995495},"started_at":"2026-10-13T05:10:04.262445+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0019277699993835995,"native_input_samples":78480,"native_input_sha256":"6ee6514f6cbc5f558e7996ec0e7fac4e87bb542c3bcdd1371a2075a14ab17381","output_sha256":"f376f38d26ea0b1ec36233e33033ae63f64522a14cab66b339907c38780d33b5","output_samples":78720,"length_delta_samples":240,"analysis_s":0.3542407210006786,"synthesis_s":0.0815302459996019,"total_s":0.4357709670002805,"rtf":0.13326329266063625,"output_peak":0.5038565397262573,"output_rms":0.06592252115906949,"finished_at":"2026-10-13T05:10:04.705220+00:00","analysis_stages":{"external_f0_s":0.055992404000789975},"evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"f376f38d26ea0b1ec36233e33033ae63f64522a14cab66b339907c38780d33b5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":52320,"output_samples":52480,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.614201090451317,"output_peak":0.5030228465183768},"common_prefix":{"samples":52320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9969512195121951,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.516372589504495,"spectral_convergence":0.1811088029773362,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.310140261300311,"spectral_convergence":0.16306373766545554,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.79805516634344,"spectral_convergence":0.2124156919214313,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.0004627214890836128,"delta_db":-0.09557443727969621,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.0017693346169357,"delta_db":-0.12153342263278646,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.0018048867850060909,"delta_db":-0.03000315715949537,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.00013320406611829318,"delta_db":-0.11310818830595266,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00011774581938680783,"delta_db":-0.198251034503033,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":7.631550959104286e-05,"delta_db":-0.5587129515292937,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18619719856837622,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0910993947719212,"pitch":null,"output_rms_dbfs":-23.600945304588965,"output_peak":0.5030228465183768,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.541522672382748},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52320,"output_samples":52320,"intelligibility":{"estoi":{"value":0.9500467766817547,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0667586744040302,"rms_dbfs":-23.509845909817045,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.978446332783722,"mean":4.524732583888809e-05,"zero_energy":false},"output":{"rms":0.0660621547190337,"rms_dbfs":-23.600945304588965,"peak":0.5030228465183768,"peak_dbfs":-5.96824579027172,"crest_factor_db":17.632699514317245,"mean":-9.525027287430843e-05,"zero_energy":false},"rms_difference_db":-0.09109939477191986},"full_scale":{"reference":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":327,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":9.197614927387515e-07,"output_rms_dbfs":-60.36324776802801,"output_zero_energy":false,"output_peak":0.004445283290755689,"output_energy_fraction":1.288997267354622e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":320,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":155,"both_estimated_voiced_frames":146,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9240506329113924,"absolute_difference_cents_median":1.002376542829331,"absolute_difference_cents_p95":7.3280223582604025,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00017439460044090568,"log_rms_mae_db":0.7625604132037443,"rms_envelope_nrmse":0.054130524573346044,"envelope_correlation":0.9966003201842509,"rms_gain_error_db":-0.09109669930693795,"reference_mean_power":0.004463526965610992,"output_mean_power":0.004370876070415737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":1.3056377024951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008270090313661458,"output_power":0.000848037054534761,"reference_active":true,"delta_db":0.10904576642444375},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010887958720618103,"output_power":9.955070225307402e-05,"reference_active":true,"delta_db":-0.3890213767783584},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8937894537156728e-05,"output_power":2.2444606502893136e-05,"reference_active":true,"delta_db":-1.1035493375476468}],"active_band_count":3,"active_band_mae_db":0.5338721602501496}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015806536675822195,"log_rms_mae_db":0.47669103492736775,"rms_envelope_nrmse":0.029692205764700157,"envelope_correlation":0.9990749473399982,"rms_gain_error_db":-0.0910792055936517,"reference_mean_power":0.004504823508022685,"output_mean_power":0.00441133317608446},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.092227651107542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000711888936919089,"output_power":0.0007297977130065027,"reference_active":true,"delta_db":0.1079025421262324},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.550528203435593e-05,"output_power":6.158437327853547e-05,"reference_active":true,"delta_db":-0.2680579471727705},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5248696894099064e-06,"output_power":1.28858188588851e-06,"reference_active":true,"delta_db":-0.7312070981981404}],"active_band_count":3,"active_band_mae_db":0.3690558624990478}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00016816220894332027,"log_rms_mae_db":0.8936242338795714,"rms_envelope_nrmse":0.06472742499252201,"envelope_correlation":0.9962375955139451,"rms_gain_error_db":-0.08599031772965089,"reference_mean_power":0.002573247092376177,"output_mean_power":0.0025227978891733675},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":9.221294414644421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007324784888809201,"output_power":0.0007252135871602251,"reference_active":true,"delta_db":-0.043289428597075165},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.576574656453582e-05,"output_power":7.439239065221335e-05,"reference_active":true,"delta_db":-0.6178535707430628},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8048967652930613e-05,"output_power":1.2701273499292806e-05,"reference_active":true,"delta_db":-1.5260509872004147}],"active_band_count":3,"active_band_mae_db":0.7290646621801842}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001473388036867024,"log_rms_mae_db":0.5194498713975709,"rms_envelope_nrmse":0.03558407565437937,"envelope_correlation":0.9988548580735187,"rms_gain_error_db":-0.08598913748696832,"reference_mean_power":0.002597106693874505,"output_mean_power":0.002546190408704787},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.911179205548874e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006603877113347142,"output_power":0.0006516374728538851,"reference_active":true,"delta_db":-0.05792932449249988},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5067104939947466e-05,"output_power":3.919213386887173e-05,"reference_active":true,"delta_db":-0.6066075050495355},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.161011125548218e-06,"output_power":8.795661869973272e-07,"reference_active":true,"delta_db":-1.2056785593888857}],"active_band_count":3,"active_band_mae_db":0.6234051296436404}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":0.8831654834632766,"rms_envelope_nrmse":0.08493141660322741,"envelope_correlation":0.9943482273575899,"rms_gain_error_db":-0.13620262980311224,"reference_mean_power":0.0001985957742888365,"output_mean_power":0.00019246410318643666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":8.760825576371805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.45270565717188e-05,"output_power":5.33175302110637e-05,"reference_active":true,"delta_db":-0.09742030941053315},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.928893294666135e-06,"output_power":5.646245176712523e-06,"reference_active":true,"delta_db":-0.21213901619844625},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9788817769833534e-06,"output_power":1.6924070401086178e-06,"reference_active":true,"delta_db":-0.6791502596539916}],"active_band_count":3,"active_band_mae_db":0.3295698617543236}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":0.537908231787154,"rms_envelope_nrmse":0.045687187600987604,"envelope_correlation":0.9984430113430008,"rms_gain_error_db":-0.13612363283493845,"reference_mean_power":0.000200416025269102,"output_mean_power":0.00019423168666281068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.857113792607277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022793901144949e-05,"output_power":4.9233695698216444e-05,"reference_active":true,"delta_db":-0.08682921190284859},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6196674175617505e-06,"output_power":2.4000878227260828e-06,"reference_active":true,"delta_db":-0.3801902506576521},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.185526690136049e-07,"output_power":9.803180499232621e-08,"reference_active":true,"delta_db":-0.8254433675511098}],"active_band_count":3,"active_band_mae_db":0.43082094337053684}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":0.967897034653758,"rms_envelope_nrmse":0.14155499923001574,"envelope_correlation":0.9815724923556566,"rms_gain_error_db":-0.3674962403616366,"reference_mean_power":8.71199669251843e-05,"output_mean_power":8.005124263031418e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":4.20340153760379e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.824817597815015e-05,"output_power":1.7893822136183675e-05,"reference_active":true,"delta_db":-0.08516344291597991},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2434318429437933e-05,"output_power":1.1106064185780783e-05,"reference_active":true,"delta_db":-0.49061805904604583},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7676697520375684e-06,"output_power":5.000961059134211e-06,"reference_active":true,"delta_db":-1.31385685046136}],"active_band_count":3,"active_band_mae_db":0.6298794508077953}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":0.6168552103246793,"rms_envelope_nrmse":0.08370046796084636,"envelope_correlation":0.9936590357777875,"rms_gain_error_db":-0.36751319702959184,"reference_mean_power":8.789733624317312e-05,"output_mean_power":8.076522255388289e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":3.223393789064046e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.051966991658495e-05,"output_power":1.9503042139645187e-05,"reference_active":true,"delta_db":-0.22068011222798803},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.88053445243811e-06,"output_power":5.0922194793877464e-06,"reference_active":true,"delta_db":-0.6250968484820664},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.871896444237528e-07,"output_power":4.131741747068724e-07,"reference_active":true,"delta_db":-1.526452195461819}],"active_band_count":3,"active_band_mae_db":0.7907430520572912}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0027","method":"magphase","status":"measured","mos":{"status":"measured","value":3.516442060470581,"predicted_mos":3.516442060470581,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":157282,"input_sample_rate":48000,"input_seconds":3.276708333333333,"input_float32_sha256":"090da1343b05ec6c3f20c6431f7e4ac06dc095b607cb3c37845780159e468a52","internal_samples":52428,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.321569661000467},"started_at":"2026-10-13T05:12:39.508715+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005727044999730424,"native_input_samples":156960,"native_input_sha256":"c746b125d39a98aa39228cacc38be2a8e9fd962d4019ad3f6d2d69b95e851ba7","output_sha256":"b86676fd8a7652f46dee74fdae4ab2babea25ceb59b75e98f53d0d7fc01a5c20","output_samples":157282,"length_delta_samples":322,"analysis_s":1.9448749050006882,"synthesis_s":0.7349045949995343,"total_s":2.6797795000002225,"rtf":0.8195044342508325,"output_peak":0.3908208450949639,"output_rms":0.0642531963936018,"finished_at":"2026-10-13T05:12:42.206093+00:00","evaluation_input":{"reference":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"b86676fd8a7652f46dee74fdae4ab2babea25ceb59b75e98f53d0d7fc01a5c20","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":52320,"output_samples":52428,"length_delta_samples":108,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.83749805727119,"output_peak":0.38634283241762785},"common_prefix":{"samples":52320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.997940032043946,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":52320,"output_samples":52320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.2812999653335835,"spectral_convergence":0.4780407803444322,"spectral_convergence_status":"ok","amplitude_floor":1.2746778473169441e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.030634587121246,"spectral_convergence":0.15945586045355525,"spectral_convergence_status":"ok","amplitude_floor":1.090624511245042e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.243421428752323,"spectral_convergence":0.16667698496773603,"spectral_convergence_status":"ok","amplitude_floor":7.935262595002381e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004730173956506593,"output_power":0.00043964426152065034,"delta_db":-0.3177570373128694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018195471619226037,"output_power":0.0016556878737255453,"delta_db":-0.4098484878505828,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018173989836957952,"output_power":0.001677507232160498,"delta_db":-0.347858796565039,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013671881984837986,"output_power":0.00013697196673988137,"delta_db":0.008033907959578665,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012324536624073353,"output_power":0.00012823145341395392,"delta_db":0.17223964997191354,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.679288082514536e-05,"output_power":0.00010333243921994146,"delta_db":0.7575257758550444,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.335543942585838,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.31870827746151636,"pitch":null,"output_rms_dbfs":-23.828554187278563,"output_peak":0.38634283241762785,"reference_rms_dbfs":-23.509845909817045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5184519937357175},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":52320,"output_samples":52320,"intelligibility":{"estoi":{"value":0.9570782006980971,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0667586744040302,"rms_dbfs":-23.509845909817045,"peak":0.374481201171875,"peak_dbfs":-8.531399577033323,"crest_factor_db":14.978446332783722,"mean":4.524732583888809e-05,"zero_energy":false},"output":{"rms":0.06435351759445998,"rms_dbfs":-23.828554187278563,"peak":0.38634283241762785,"peak_dbfs":-8.260542810884443,"crest_factor_db":15.568011376394118,"mean":0.0017418714821679322,"zero_energy":false},"rms_difference_db":-0.31870827746151775},"full_scale":{"reference":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":52320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":327,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.041347666294314e-06,"threshold_dbfs":-55.169339314103624,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":9.64982609730214e-07,"output_mean_power":5.96692297565677e-07,"output_rms_dbfs":-62.242495682331835,"output_zero_energy":false,"output_peak":0.0029848232532082696,"output_energy_fraction":8.812276055340101e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":320,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":155,"both_estimated_voiced_frames":150,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9493670886075949,"absolute_difference_cents_median":3.324178203987491,"absolute_difference_cents_p95":10.52884280240797,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":52320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00017439460044090568,"log_rms_mae_db":1.2335060102488655,"rms_envelope_nrmse":0.07910045709624712,"envelope_correlation":0.9936570433257139,"rms_gain_error_db":-0.3187002870456037,"reference_mean_power":0.004463526965610992,"output_mean_power":0.004147707616140477},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":1.3056377024951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008270090313661458,"output_power":0.000756518676178172,"reference_active":true,"delta_db":-0.38690598378346236},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010887958720618103,"output_power":9.915195904848418e-05,"reference_active":true,"delta_db":-0.4064516622883002},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8937894537156728e-05,"output_power":2.445738535402664e-05,"reference_active":true,"delta_db":-0.7305690312998796}],"active_band_count":3,"active_band_mae_db":0.5079755591238807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015806536675822195,"log_rms_mae_db":0.7577822699751169,"rms_envelope_nrmse":0.05158186450574954,"envelope_correlation":0.9980481287064619,"rms_gain_error_db":-0.31871185956896286,"reference_mean_power":0.004504823508022685,"output_mean_power":0.004186071044051846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.092227651107542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000711888936919089,"output_power":0.0006489900936121854,"reference_active":true,"delta_db":-0.40174176209677676},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.550528203435593e-05,"output_power":5.761891848838669e-05,"reference_active":true,"delta_db":-0.5571121856552856},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5248696894099064e-06,"output_power":1.1487600014960684e-06,"reference_active":true,"delta_db":-1.2300342629524366}],"active_band_count":3,"active_band_mae_db":0.7296294035681662}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":0.00016816220894332027,"log_rms_mae_db":1.2504203775934732,"rms_envelope_nrmse":0.07589303225059611,"envelope_correlation":0.995850049881609,"rms_gain_error_db":-0.3325600118946872,"reference_mean_power":0.002573247092376177,"output_mean_power":0.002383556711832339},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":9.221294414644421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007324784888809201,"output_power":0.0006726359016502634,"reference_active":true,"delta_db":-0.3701483115904546},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.576574656453582e-05,"output_power":7.545881272689302e-05,"reference_active":true,"delta_db":-0.5560390466252217},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8048967652930613e-05,"output_power":1.4367576726678262e-05,"reference_active":true,"delta_db":-0.9906884154426571}],"active_band_count":3,"active_band_mae_db":0.6389585912194445}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001473388036867024,"log_rms_mae_db":0.8267861392667056,"rms_envelope_nrmse":0.053806035275178785,"envelope_correlation":0.998446456171486,"rms_gain_error_db":-0.33255768639948624,"reference_mean_power":0.002597106693874505,"output_mean_power":0.002405658758767156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.911179205548874e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006603877113347142,"output_power":0.0006031711536629441,"reference_active":true,"delta_db":-0.39358419744342893},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5067104939947466e-05,"output_power":4.0046324267931316e-05,"reference_active":true,"delta_db":-0.5129700071789856},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.161011125548218e-06,"output_power":8.757464331535894e-07,"reference_active":true,"delta_db":-1.2245800429887135}],"active_band_count":3,"active_band_mae_db":0.7103780825370426}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":5.3802593839835606e-05,"log_rms_mae_db":1.0002625231782927,"rms_envelope_nrmse":0.09293212939073599,"envelope_correlation":0.9932066479406021,"rms_gain_error_db":0.07324458638716112,"reference_mean_power":0.0001985957742888365,"output_mean_power":0.00020197353281780367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":8.760825576371805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.45270565717188e-05,"output_power":5.568514403085485e-05,"reference_active":true,"delta_db":0.09127292667259689},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.928893294666135e-06,"output_power":5.662314738954986e-06,"reference_active":true,"delta_db":-0.19979628298044852},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9788817769833534e-06,"output_power":1.7288720351878482e-06,"reference_active":true,"delta_db":-0.5865699964780395}],"active_band_count":3,"active_band_mae_db":0.292546402043695}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":5.051409218298903e-05,"log_rms_mae_db":0.5255450651411877,"rms_envelope_nrmse":0.04170469736074767,"envelope_correlation":0.9985952010195618,"rms_gain_error_db":0.07334690736103441,"reference_mean_power":0.000200416025269102,"output_mean_power":0.00020382954523157974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":7.857113792607277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022793901144949e-05,"output_power":5.1183874058932465e-05,"reference_active":true,"delta_db":0.08187795888291804},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6196674175617505e-06,"output_power":2.468843901549261e-06,"reference_active":true,"delta_db":-0.2575252702687273},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.185526690136049e-07,"output_power":1.0009016134105733e-07,"reference_active":true,"delta_db":-0.7351994651823259}],"active_band_count":3,"active_band_mae_db":0.3582008981113238}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":653,"rms_floor":6.112392927022402e-05,"log_rms_mae_db":1.3024621191469752,"rms_envelope_nrmse":0.24098410446547222,"envelope_correlation":0.949660453236963,"rms_gain_error_db":0.4545067105182401,"reference_mean_power":8.71199669251843e-05,"output_mean_power":9.673160035440836e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":653,"frequency_resolution_hz":0.30627871362940273,"power_floor":4.20340153760379e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.824817597815015e-05,"output_power":2.153911732708917e-05,"reference_active":true,"delta_db":0.7200844148910537},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2434318429437933e-05,"output_power":1.2899562706936058e-05,"reference_active":true,"delta_db":0.15953003050737274},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7676697520375684e-06,"output_power":6.103681767691096e-06,"reference_active":true,"delta_db":-0.44847275501710643}],"active_band_count":3,"active_band_mae_db":0.44269573347184427}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":52320,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":3.6265226913688176e-05,"log_rms_mae_db":0.7816563782093567,"rms_envelope_nrmse":0.14427085917023375,"envelope_correlation":0.9804775444398108,"rms_gain_error_db":0.4541503631755638,"reference_mean_power":8.789733624317312e-05,"output_mean_power":9.758672652525872e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":3.223393789064046e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.051966991658495e-05,"output_power":2.3272283579430287e-05,"reference_active":true,"delta_db":0.5466862992352451},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.88053445243811e-06,"output_power":5.838577749238976e-06,"reference_active":true,"delta_db":-0.031097308769939512},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.871896444237528e-07,"output_power":4.853456292860363e-07,"reference_active":true,"delta_db":-0.8272726483432591}],"active_band_count":3,"active_band_mae_db":0.4683520854494813}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"hnm","status":"measured","mos":{"status":"measured","value":2.9910778999328613,"predicted_mos":2.9910778999328613,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"35e8e087a33d18052dd435e5b9cdb134e91071ed0cd326afd190ca5fa2b21bc8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6464486820004822},"started_at":"2026-10-13T04:38:14.711375+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.436300044588279e-05,"native_input_samples":96000,"native_input_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","output_sha256":"311bc44d4257bed8ecb6525943e8b3d7fe1d7f23d63365b5f96f4449410a922a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1022718660005921,"synthesis_s":0.5168354460001865,"total_s":1.6191073120007786,"rtf":0.26985121866679646,"output_peak":0.5508650945923664,"output_rms":0.05560375934869822,"finished_at":"2026-10-13T04:38:16.335665+00:00","evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"311bc44d4257bed8ecb6525943e8b3d7fe1d7f23d63365b5f96f4449410a922a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.60244755417561,"spectral_convergence":0.21229275486022428,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5205110825565225,"spectral_convergence":0.2072470422689199,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.730516662800133,"spectral_convergence":0.20817326455633045,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.00034145797713480675,"delta_db":0.06980484704914576,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.0011149154333276708,"delta_db":-0.03607232893764163,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0014846156695277313,"delta_db":-0.1639874369486731,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":7.990025038501287e-05,"delta_db":-0.07385636890541526,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":5.109663397590442e-05,"delta_db":-0.15546332081841874,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":1.9792089356817535e-05,"delta_db":0.06869135157284059,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.0946459423720225,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08885313354097046,"pitch":null,"output_rms_dbfs":-25.097916898861715,"output_peak":0.5508650945923664,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.6178250998440875},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.60244755417561,"spectral_convergence":0.21229275486022428,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5205110825565225,"spectral_convergence":0.2072470422689199,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.730516662800133,"spectral_convergence":0.20817326455633045,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.00034145797713480675,"delta_db":0.06980484704914576,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.0011149154333276708,"delta_db":-0.03607232893764163,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0014846156695277313,"delta_db":-0.1639874369486731,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":7.990025038501287e-05,"delta_db":-0.07385636890541526,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":5.109663397590442e-05,"delta_db":-0.15546332081841874,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":1.9792089356817535e-05,"delta_db":0.06869135157284059,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.0946459423720225,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08885313354097046,"pitch":null,"output_rms_dbfs":-25.097916898861715,"output_peak":0.5508650945923664,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.6178250998440875},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9016105143938604,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05617548255220346,"rms_dbfs":-25.009063765320747,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.132000780931644,"mean":-3.836313883463542e-06,"zero_energy":false},"output":{"rms":0.05560375934869822,"rms_dbfs":-25.097916898861715,"peak":0.5508650945923664,"peak_dbfs":-5.179094913985512,"crest_factor_db":19.918821984876203,"mean":-4.971237179709902e-05,"zero_energy":false},"rms_difference_db":-0.0888531335409688},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":2.9935060414616724e-06,"output_rms_dbfs":-55.23819861028963,"output_zero_energy":false,"output_peak":0.025018349670126,"output_energy_fraction":0.00015168702442915706,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":297,"both_estimated_voiced_frames":289,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.932258064516129,"absolute_difference_cents_median":1.9212100920184687,"absolute_difference_cents_p95":14.68150439103146,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025375764068356454,"log_rms_mae_db":0.8974960460830799,"rms_envelope_nrmse":0.07897341509807032,"envelope_correlation":0.9936293620317977,"rms_gain_error_db":-0.08885128752450502,"reference_mean_power":0.0031583120304531104,"output_mean_power":0.0030943533553993837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1807904078754827e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008327873716353349,"output_power":0.0008252260533213563,"reference_active":true,"delta_db":-0.03961200236563266},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013733244323789595,"output_power":0.00012416575015937527,"reference_active":true,"delta_db":-0.4377132981192746},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.747498790065721e-05,"output_power":1.6528247244723577e-05,"reference_active":true,"delta_db":-0.2419008292032007}],"active_band_count":3,"active_band_mae_db":0.23974204322936932}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002402786371599188,"log_rms_mae_db":0.5006386954719367,"rms_envelope_nrmse":0.045421998368486295,"envelope_correlation":0.9977530593466662,"rms_gain_error_db":-0.08884887135718471,"reference_mean_power":0.0031741738524038612,"output_mean_power":0.003109895691278383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0384864838590044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007821613563464859,"output_power":0.0007750030524036875,"reference_active":true,"delta_db":-0.03992942278880648},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.85949543312588e-05,"output_power":5.307043503133042e-05,"reference_active":true,"delta_db":-0.4300757244417913},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3154862476080321e-06,"output_power":1.1856563113632017e-06,"reference_active":true,"delta_db":-0.4512749479750964}],"active_band_count":3,"active_band_mae_db":0.3070933650685647}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002501327279402343,"log_rms_mae_db":1.2412645956034738,"rms_envelope_nrmse":0.09901202659649196,"envelope_correlation":0.992230008638375,"rms_gain_error_db":-0.12025367483191002,"reference_mean_power":0.0021283926153372414,"output_mean_power":0.0020702670782121986},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0698576130619868e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007463968445819582,"output_power":0.0007355760852306795,"reference_active":true,"delta_db":-0.06342193568738816},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014378870760237378,"output_power":0.00012806998678754065,"reference_active":true,"delta_db":-0.5027741549630201},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4173849116031066e-05,"output_power":1.4043076642529192e-05,"reference_active":true,"delta_db":-0.04025539191731867}],"active_band_count":3,"active_band_mae_db":0.20215049418924233}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00023719519351520602,"log_rms_mae_db":0.7321627422356534,"rms_envelope_nrmse":0.05864233932798885,"envelope_correlation":0.9971403412510679,"rms_gain_error_db":-0.12025369329406019,"reference_mean_power":0.0021390960955203144,"output_mean_power":0.0020806782418573747},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.624199297765067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007211274534636877,"output_power":0.0007079415981188207,"reference_active":true,"delta_db":-0.08014597654265118},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.136224424017926e-05,"output_power":5.430444234988541e-05,"reference_active":true,"delta_db":-0.5306587665617567},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3649293472151977e-06,"output_power":1.1682074493348674e-06,"reference_active":true,"delta_db":-0.6759020030222433}],"active_band_count":3,"active_band_mae_db":0.42890224870888377}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":1.6490360389045775,"rms_envelope_nrmse":0.17718163858455938,"envelope_correlation":0.9768318695369249,"rms_gain_error_db":-0.20121391378737838,"reference_mean_power":9.73047769043704e-05,"output_mean_power":9.289937022615367e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.517204404381221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.13601836765628e-05,"output_power":3.8435206966121406e-05,"reference_active":true,"delta_db":-0.3185323474684335},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.821587959830413e-06,"output_power":4.5436508743522654e-06,"reference_active":true,"delta_db":-0.25785140360010395},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.214811101263015e-06,"output_power":2.062863072855896e-06,"reference_active":true,"delta_db":-0.30866289989221884}],"active_band_count":3,"active_band_mae_db":0.2950155503202521}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":1.1205874433301017,"rms_envelope_nrmse":0.11878988716436521,"envelope_correlation":0.9891849813278655,"rms_gain_error_db":-0.20123447313412218,"reference_mean_power":9.778712425213711e-05,"output_mean_power":9.335943766903807e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.922472441667661e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.880871023429881e-05,"output_power":3.603859407666174e-05,"reference_active":true,"delta_db":-0.32161369556906017},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.146701619692129e-06,"output_power":2.0305263542257694e-06,"reference_active":true,"delta_db":-0.2416305339150633},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6121018734527105e-07,"output_power":1.561954136411925e-07,"reference_active":true,"delta_db":-0.1372420512081207}],"active_band_count":3,"active_band_mae_db":0.23349542689741473}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":1.9265558288867826,"rms_envelope_nrmse":0.22236239527355783,"envelope_correlation":0.9557586983296077,"rms_gain_error_db":-0.10067737812415162,"reference_mean_power":3.136903007980472e-05,"output_mean_power":3.065020283063104e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.994485115758079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0899878382466337e-05,"output_power":1.05628719304317e-05,"reference_active":true,"delta_db":-0.13639638024496642},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.551902798888814e-06,"output_power":2.08953678815781e-06,"reference_active":true,"delta_db":-0.8681410654634831},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4088478818809362e-06,"output_power":6.701603856509222e-07,"reference_active":true,"delta_db":-3.22685351027983}],"active_band_count":3,"active_band_mae_db":1.4104636519960934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":1.325536996227401,"rms_envelope_nrmse":0.14304928590590676,"envelope_correlation":0.9795902042130117,"rms_gain_error_db":-0.10055519321305734,"reference_mean_power":3.151908471872328e-05,"output_mean_power":3.0797685392605955e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7532611532165262e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0780259770731747e-05,"output_power":1.0183260251687992e-05,"reference_active":true,"delta_db":-0.24742383030377907},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1761748042676719e-06,"output_power":9.154433288382788e-07,"reference_active":true,"delta_db":-1.088404076490879},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.367404555815228e-08,"output_power":3.798877434577995e-08,"reference_active":true,"delta_db":-3.9196399489394422}],"active_band_count":3,"active_band_mae_db":1.7518226185780332}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.5149478912353516,"predicted_mos":3.5149478912353516,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"cc419f805bdf2904d7de10477b8c08064ac4fde3567d9c0bd46dc8ab7be0b0fc","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3962740380011383},"started_at":"2026-10-13T04:41:07.201446+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4525000299036037e-05,"native_input_samples":96000,"native_input_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","output_sha256":"a2f76a43cfd90e56e9cf04e1f653ff14392090bc981bb5f380906c0806a078e3","output_samples":96000,"length_delta_samples":0,"analysis_s":2.839946167000562,"synthesis_s":0.0987587439994968,"total_s":2.938704911000059,"rtf":0.4897841518333432,"output_peak":0.6436996514869824,"output_rms":0.06710335170175519,"finished_at":"2026-10-13T04:41:10.147421+00:00","evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"a2f76a43cfd90e56e9cf04e1f653ff14392090bc981bb5f380906c0806a078e3","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.032713594001548,"spectral_convergence":0.5129039588787736,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.284829886142564,"spectral_convergence":0.30489090541216707,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.684006402279887,"spectral_convergence":0.3332535065490408,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.0003571448430698404,"delta_db":0.2648760513952899,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.0013950122047450358,"delta_db":0.9372884704160823,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0025442564137720725,"delta_db":2.1754809425508896,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":0.000117250845053367,"delta_db":1.5918220429301535,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":6.648075650981755e-05,"delta_db":0.9875732907974573,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":2.2714746459317297e-05,"delta_db":0.6668538573409883,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.10398244257181,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5439480253991278,"pitch":null,"output_rms_dbfs":-23.465115739921618,"output_peak":0.6436996514869824,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.667183294141332},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.032713594001548,"spectral_convergence":0.5129039588787736,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.284829886142564,"spectral_convergence":0.30489090541216707,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.684006402279887,"spectral_convergence":0.3332535065490408,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.0003571448430698404,"delta_db":0.2648760513952899,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.0013950122047450358,"delta_db":0.9372884704160823,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0025442564137720725,"delta_db":2.1754809425508896,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":0.000117250845053367,"delta_db":1.5918220429301535,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":6.648075650981755e-05,"delta_db":0.9875732907974573,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":2.2714746459317297e-05,"delta_db":0.6668538573409883,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.10398244257181,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5439480253991278,"pitch":null,"output_rms_dbfs":-23.465115739921618,"output_peak":0.6436996514869824,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.667183294141332},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9433998204380387,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05617548255220346,"rms_dbfs":-25.009063765320747,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.132000780931644,"mean":-3.836313883463542e-06,"zero_energy":false},"output":{"rms":0.06710335170175519,"rms_dbfs":-23.465115739921615,"peak":0.6436996514869824,"peak_dbfs":-3.826334520149259,"crest_factor_db":19.63878121977236,"mean":4.8424792352767345e-09,"zero_energy":false},"rms_difference_db":1.543948025399132},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":2.830674558402327e-06,"output_rms_dbfs":-55.48110058439383,"output_zero_energy":false,"output_peak":0.017060824790653868,"output_energy_fraction":9.848682087251016e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":331,"both_estimated_voiced_frames":296,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9548387096774194,"absolute_difference_cents_median":7.000486802557205,"absolute_difference_cents_p95":25.779995360269673,"difference_over_600_cents_frames":8,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025375764068356454,"log_rms_mae_db":1.6994640221310646,"rms_envelope_nrmse":0.24202111842227617,"envelope_correlation":0.9846452948246368,"rms_gain_error_db":1.5439529641018352,"reference_mean_power":0.0031583120304531104,"output_mean_power":0.004506613683441746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1807904078754827e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008327873716353349,"output_power":0.0012673543100597544,"reference_active":true,"delta_db":1.8236391520969464},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013733244323789595,"output_power":0.00021907334854271766,"reference_active":true,"delta_db":2.028163999811562},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.747498790065721e-05,"output_power":2.066895386578183e-05,"reference_active":true,"delta_db":0.7290161217041508}],"active_band_count":3,"active_band_mae_db":1.5269397578708863}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002402786371599188,"log_rms_mae_db":1.442504840263694,"rms_envelope_nrmse":0.22417848123832396,"envelope_correlation":0.9908847557469079,"rms_gain_error_db":1.5439659224506663,"reference_mean_power":0.0031741738524038612,"output_mean_power":0.004529260523077046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0384864838590044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007821613563464859,"output_power":0.0012045974508853418,"reference_active":true,"delta_db":1.8754558451345904},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.85949543312588e-05,"output_power":9.626014336250755e-05,"reference_active":true,"delta_db":2.155862840406632},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3154862476080321e-06,"output_power":1.392667125664466e-06,"reference_active":true,"delta_db":0.24761011827911641}],"active_band_count":3,"active_band_mae_db":1.4263096012734462}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002501327279402343,"log_rms_mae_db":1.567798094287366,"rms_envelope_nrmse":0.29233899499352395,"envelope_correlation":0.9860953562671041,"rms_gain_error_db":1.921018248606439,"reference_mean_power":0.0021283926153372414,"output_mean_power":0.00331248231493668},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0698576130619868e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007463968445819582,"output_power":0.0012056377731433907,"reference_active":true,"delta_db":2.0824705140263213},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014378870760237378,"output_power":0.00024176630575528893,"reference_active":true,"delta_db":2.2567099420750987},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4173849116031066e-05,"output_power":1.4036752514893285e-05,"reference_active":true,"delta_db":-0.042211624461446384}],"active_band_count":3,"active_band_mae_db":1.4604640268542888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00023719519351520602,"log_rms_mae_db":1.2857927447491304,"rms_envelope_nrmse":0.2749213448361459,"envelope_correlation":0.991796906982706,"rms_gain_error_db":1.9210185439123773,"reference_mean_power":0.0021390960955203144,"output_mean_power":0.003329140693830077},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.624199297765067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007211274534636877,"output_power":0.001175030873419236,"reference_active":true,"delta_db":2.120372481022905},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.136224424017926e-05,"output_power":0.00010515055090275998,"reference_active":true,"delta_db":2.339103174174272},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3649293472151977e-06,"output_power":1.3843106519663907e-06,"reference_active":true,"delta_db":0.06123389119903313}],"active_band_count":3,"active_band_mae_db":1.5069031821320704}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":1.4722367840005919,"rms_envelope_nrmse":0.27241374293872633,"envelope_correlation":0.9761080711596365,"rms_gain_error_db":1.4776103608142903,"reference_mean_power":9.73047769043704e-05,"output_mean_power":0.0001367398809139394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.517204404381221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.13601836765628e-05,"output_power":5.333146429837878e-05,"reference_active":true,"delta_db":1.104010493278634},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.821587959830413e-06,"output_power":6.424213987108808e-06,"reference_active":true,"delta_db":1.2462990465886765},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.214811101263015e-06,"output_power":2.3307484891900606e-06,"reference_active":true,"delta_db":0.22158719727007017}],"active_band_count":3,"active_band_mae_db":0.8572989123791269}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":1.157989800266575,"rms_envelope_nrmse":0.23850277266249195,"envelope_correlation":0.985782868028417,"rms_gain_error_db":1.4776853278442572,"reference_mean_power":9.778712425213711e-05,"output_mean_power":0.00013742008323670756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.922472441667661e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.880871023429881e-05,"output_power":5.040438001373149e-05,"reference_active":true,"delta_db":1.1353906751216405},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.146701619692129e-06,"output_power":2.7764731752448673e-06,"reference_active":true,"delta_db":1.1172179794728445},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6121018734527105e-07,"output_power":1.5959063721749307e-07,"reference_active":true,"delta_db":-0.043850738864154}],"active_band_count":3,"active_band_mae_db":0.7654864644862128}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":1.6641417675660393,"rms_envelope_nrmse":0.22851776325395165,"envelope_correlation":0.9665504127297898,"rms_gain_error_db":0.8389914639047671,"reference_mean_power":3.136903007980472e-05,"output_mean_power":3.8053993271355136e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.994485115758079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0899878382466337e-05,"output_power":1.3038302450407536e-05,"reference_active":true,"delta_db":0.7779939893859429},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.551902798888814e-06,"output_power":3.1702765716532345e-06,"reference_active":true,"delta_db":0.9423302303765919},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4088478818809362e-06,"output_power":1.4325899505389368e-06,"reference_active":true,"delta_db":0.07257796923449557}],"active_band_count":3,"active_band_mae_db":0.5976340629990101}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":1.1343129482281937,"rms_envelope_nrmse":0.15463458419470236,"envelope_correlation":0.98779803108618,"rms_gain_error_db":0.8393545922748259,"reference_mean_power":3.151908471872328e-05,"output_mean_power":3.823922279764549e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7532611532165262e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0780259770731747e-05,"output_power":1.2871106544677583e-05,"reference_active":true,"delta_db":0.7698665921094049},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1761748042676719e-06,"output_power":1.52643543209969e-06,"reference_active":true,"delta_db":1.1320656664297006},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.367404555815228e-08,"output_power":8.611486947139319e-08,"reference_active":true,"delta_db":-0.36541129133850936}],"active_band_count":3,"active_band_mae_db":0.7557811832925383}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"facodec","status":"measured","mos":{"status":"measured","value":4.382434368133545,"predicted_mos":4.382434368133545,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e5a3d82dc3bc7e2f649768d38464a05db660f2ff71f2990faa581e90bce926b4","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.558778167998753},"started_at":"2026-10-13T04:46:28.429350+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4267000299005304e-05,"native_input_samples":96000,"native_input_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","output_sha256":"6580086efee1dee5677ea6a278de5f57a166f6e693b4055f71b1f93be9dd6cb4","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5796518449997166,"synthesis_s":6.091960810999808,"total_s":7.671612655999525,"rtf":1.278602109333254,"output_peak":0.389055460691452,"output_rms":0.05218870642655182,"finished_at":"2026-10-13T04:46:36.108056+00:00","evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"6580086efee1dee5677ea6a278de5f57a166f6e693b4055f71b1f93be9dd6cb4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.34431828592589,"spectral_convergence":0.25759077192418717,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.824162526833428,"spectral_convergence":0.24437998879729214,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.866324062265405,"spectral_convergence":0.27869253651923037,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.00030342070090456436,"delta_db":-0.443115712059729,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.0011579727340695315,"delta_db":0.12849173340170497,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0011406764908396434,"delta_db":-1.308502928416257,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":6.840329789374116e-05,"delta_db":-0.7485673651802395,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":3.562622969795087e-05,"delta_db":-1.7216876006246584,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":1.756162507137938e-05,"delta_db":-0.45057806845955733,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8001572346903577,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6394055873852754,"pitch":null,"output_rms_dbfs":-25.64846935270602,"output_peak":0.389055460691452,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.678268291674907},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.34431828592589,"spectral_convergence":0.25759077192418717,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.824162526833428,"spectral_convergence":0.24437998879729214,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.866324062265405,"spectral_convergence":0.27869253651923037,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.00030342070090456436,"delta_db":-0.443115712059729,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.0011579727340695315,"delta_db":0.12849173340170497,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0011406764908396434,"delta_db":-1.308502928416257,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":6.840329789374116e-05,"delta_db":-0.7485673651802395,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":3.562622969795087e-05,"delta_db":-1.7216876006246584,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":1.756162507137938e-05,"delta_db":-0.45057806845955733,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8001572346903577,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6394055873852754,"pitch":null,"output_rms_dbfs":-25.64846935270602,"output_peak":0.389055460691452,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.678268291674907},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9129897072319697,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05617548255220346,"rms_dbfs":-25.009063765320747,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.132000780931644,"mean":-3.836313883463542e-06,"zero_energy":false},"output":{"rms":0.05218870642655182,"rms_dbfs":-25.64846935270602,"peak":0.389055460691452,"peak_dbfs":-8.199769693002366,"crest_factor_db":17.448699659703653,"mean":4.7023547196545225e-06,"zero_energy":false},"rms_difference_db":-0.6394055873852729},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":2.973626215862401e-06,"output_rms_dbfs":-55.26713623100725,"output_zero_energy":false,"output_peak":0.011105012148618698,"output_energy_fraction":0.000171044815683275,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":303,"both_estimated_voiced_frames":293,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9451612903225807,"absolute_difference_cents_median":5.645354110350546,"absolute_difference_cents_p95":26.332135430310554,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025375764068356454,"log_rms_mae_db":1.1578932823262424,"rms_envelope_nrmse":0.14633324895367802,"envelope_correlation":0.982433713623088,"rms_gain_error_db":-0.6394053234204022,"reference_mean_power":0.0031583120304531104,"output_mean_power":0.002725928763515371},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1807904078754827e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008327873716353349,"output_power":0.000713564590343435,"reference_active":true,"delta_db":-0.671008402835454},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013733244323789595,"output_power":0.000127218823138893,"reference_active":true,"delta_db":-0.3322177278315009},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.747498790065721e-05,"output_power":2.1042356446788565e-05,"reference_active":true,"delta_db":0.806774893337855}],"active_band_count":3,"active_band_mae_db":0.6033336746682699}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002402786371599188,"log_rms_mae_db":0.896588536419592,"rms_envelope_nrmse":0.13187648316485887,"envelope_correlation":0.9859167553022743,"rms_gain_error_db":-0.6394054959558373,"reference_mean_power":0.0031741738524038612,"output_mean_power":0.0027396189412220644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0384864838590044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007821613563464859,"output_power":0.0006711252599207015,"reference_active":true,"delta_db":-0.6649277011810495},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.85949543312588e-05,"output_power":5.225470078532386e-05,"reference_active":true,"delta_db":-0.49734854829382086},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3154862476080321e-06,"output_power":1.604589128074542e-06,"reference_active":true,"delta_db":0.8627753307325474}],"active_band_count":3,"active_band_mae_db":0.6750171934024726}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002501327279402343,"log_rms_mae_db":1.6261856014006228,"rms_envelope_nrmse":0.16989339597426686,"envelope_correlation":0.9807455257595846,"rms_gain_error_db":-0.7016305124093785,"reference_mean_power":0.0021283926153372414,"output_mean_power":0.0018108759134127105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0698576130619868e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007463968445819582,"output_power":0.0006240395121300182,"reference_active":true,"delta_db":-0.7775770609602803},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014378870760237378,"output_power":0.00013644242124238875,"reference_active":true,"delta_db":-0.2277536262821931},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4173849116031066e-05,"output_power":1.6747006448372368e-05,"reference_active":true,"delta_db":0.7244938226063504}],"active_band_count":3,"active_band_mae_db":0.576608169949608}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00023719519351520602,"log_rms_mae_db":1.3233610130720364,"rms_envelope_nrmse":0.15217953671788917,"envelope_correlation":0.9847959356089192,"rms_gain_error_db":-0.7016310787087016,"reference_mean_power":0.0021390960955203144,"output_mean_power":0.0018199823955578212},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.624199297765067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007211274534636877,"output_power":0.0006063029919282748,"reference_active":true,"delta_db":-0.7532231819507196},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.136224424017926e-05,"output_power":5.6307717520896094e-05,"reference_active":true,"delta_db":-0.3733331174030591},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3649293472151977e-06,"output_power":1.6101469046593046e-06,"reference_active":true,"delta_db":0.717553298903473}],"active_band_count":3,"active_band_mae_db":0.6147031994190838}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":1.6472620781754959,"rms_envelope_nrmse":0.23810116724478628,"envelope_correlation":0.9653702145699555,"rms_gain_error_db":-1.1605917786686946,"reference_mean_power":9.73047769043704e-05,"output_mean_power":7.44860567181278e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.517204404381221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.13601836765628e-05,"output_power":3.297853818089895e-05,"reference_active":true,"delta_db":-0.9835105771541458},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.821587959830413e-06,"output_power":4.5648860652553185e-06,"reference_active":true,"delta_db":-0.2376015149496247},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.214811101263015e-06,"output_power":2.9003538831075155e-06,"reference_active":true,"delta_db":1.17114299385597}],"active_band_count":3,"active_band_mae_db":0.7974183619865802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":1.3611475805755058,"rms_envelope_nrmse":0.21353041028041628,"envelope_correlation":0.9726510748127793,"rms_gain_error_db":-1.1607619920971342,"reference_mean_power":9.778712425213711e-05,"output_mean_power":7.485235614237999e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.922472441667661e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.880871023429881e-05,"output_power":3.105924478969108e-05,"reference_active":true,"delta_db":-0.9673831808645538},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.146701619692129e-06,"output_power":2.156319063032409e-06,"reference_active":true,"delta_db":0.019413383460868537},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6121018734527105e-07,"output_power":2.1341199029398184e-07,"reference_active":true,"delta_db":1.2182633340270783}],"active_band_count":3,"active_band_mae_db":0.7350199661175002}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":1.9087328996923296,"rms_envelope_nrmse":0.25046729083090735,"envelope_correlation":0.948899338762815,"rms_gain_error_db":-0.9664957527768894,"reference_mean_power":3.136903007980472e-05,"output_mean_power":2.511027768192969e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.994485115758079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0899878382466337e-05,"output_power":7.330745769884117e-06,"reference_active":true,"delta_db":-1.7227349379834036},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.551902798888814e-06,"output_power":2.2408385003881966e-06,"reference_active":true,"delta_db":-0.5645357063665888},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4088478818809362e-06,"output_power":1.3459486279757222e-06,"reference_active":true,"delta_db":-0.19835619283147646}],"active_band_count":3,"active_band_mae_db":0.8285422790604896}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":1.493393794551283,"rms_envelope_nrmse":0.21436073601534783,"envelope_correlation":0.9610588434725138,"rms_gain_error_db":-0.9662207305667453,"reference_mean_power":3.151908471872328e-05,"output_mean_power":2.5231991198104222e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7532611532165262e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0780259770731747e-05,"output_power":7.148356358913868e-06,"reference_active":true,"delta_db":-1.7842303135884554},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1761748042676719e-06,"output_power":1.0586824748935934e-06,"reference_active":true,"delta_db":-0.45706147853360746},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.367404555815228e-08,"output_power":8.740351911715556e-08,"reference_active":true,"delta_db":-0.30090357844006854}],"active_band_count":3,"active_band_mae_db":0.8473984568540439}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.925689697265625,"predicted_mos":3.925689697265625,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"796f0f30b0cc775aa1b670d1b60685877f1cb57dd3d0eeb9ce217a6d6d57db3e","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.39444017400092},"started_at":"2026-10-13T05:00:16.560122+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006170147999910114,"native_input_samples":264600,"native_input_sha256":"f6fde145483701f299814c9dd45b5583a71daee9165c3d02912134675a468e90","output_sha256":"76823442fa2183a5c2c2c2763e0ec98168131b33e2fad5a202103d3556fe6099","output_samples":264704,"length_delta_samples":104,"analysis_s":0.14519822600050247,"synthesis_s":4.251648533999287,"total_s":4.396846759999789,"rtf":0.7328077933332983,"output_peak":0.4970381259918213,"output_rms":0.05219220393686633,"finished_at":"2026-10-13T05:00:20.977334+00:00","evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"76823442fa2183a5c2c2c2763e0ec98168131b33e2fad5a202103d3556fe6099","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.641268624950015,"output_peak":0.5002585656379603},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.709932699508638,"spectral_convergence":0.4335676364916896,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.532265461321563,"spectral_convergence":0.3307861671569409,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.225030198039097,"spectral_convergence":0.3907582157119032,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.00023135274235413134,"delta_db":-1.6208212664440302,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.00111979175229277,"delta_db":-0.017118957387144588,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.001230625977083153,"delta_db":-0.9788670544099916,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":9.703506633908496e-05,"delta_db":0.7699492994247569,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":3.848978468623811e-05,"delta_db":-1.3859314208999807,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":1.1963946499071687e-05,"delta_db":-2.1174804560827063,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1483614091081018,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6304882951363298,"pitch":null,"output_rms_dbfs":-25.63955206045707,"output_peak":0.5002585656379603,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.489076119623098},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7991363853595442,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05617548255220346,"rms_dbfs":-25.009063765320747,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.132000780931644,"mean":-3.836313883463542e-06,"zero_energy":false},"output":{"rms":0.05224231301593039,"rms_dbfs":-25.639552060457078,"peak":0.5002585656379603,"peak_dbfs":-6.016109329099299,"crest_factor_db":19.623442731357773,"mean":1.0491676114186305e-05,"zero_energy":false},"rms_difference_db":-0.6304882951363311},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":1.5690066614753761e-06,"output_rms_dbfs":-58.04375212540626,"output_zero_energy":false,"output_peak":0.0077756250910451415,"output_energy_fraction":9.006511268469205e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":309,"both_estimated_voiced_frames":291,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9387096774193548,"absolute_difference_cents_median":9.724691652586092,"absolute_difference_cents_p95":47.26123408654875,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025375764068356454,"log_rms_mae_db":1.4750806461712862,"rms_envelope_nrmse":0.1146213252053784,"envelope_correlation":0.9913020186371617,"rms_gain_error_db":-0.6304862206580013,"reference_mean_power":0.0031583120304531104,"output_mean_power":0.0027315327540122303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1807904078754827e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008327873716353349,"output_power":0.0007097161183647339,"reference_active":true,"delta_db":-0.6944946231882748},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013733244323789595,"output_power":0.00013536597616810384,"reference_active":true,"delta_db":-0.06263627070914461},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.747498790065721e-05,"output_power":1.658104316202299e-05,"reference_active":true,"delta_db":-0.22805033908239242}],"active_band_count":3,"active_band_mae_db":0.3283937443266039}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002402786371599188,"log_rms_mae_db":1.0735825463965383,"rms_envelope_nrmse":0.09148223490968056,"envelope_correlation":0.996033508223137,"rms_gain_error_db":-0.6304793501646326,"reference_mean_power":0.0031741738524038612,"output_mean_power":0.0027452555281423673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0384864838590044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007821613563464859,"output_power":0.0006724721708853354,"reference_active":true,"delta_db":-0.6562203877835691},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.85949543312588e-05,"output_power":5.710926276626057e-05,"reference_active":true,"delta_db":-0.11153666290779447},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3154862476080321e-06,"output_power":1.4729952004389846e-06,"reference_active":true,"delta_db":0.4911501952120444}],"active_band_count":3,"active_band_mae_db":0.41963574863446934}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002501327279402343,"log_rms_mae_db":2.7272736039321273,"rms_envelope_nrmse":0.14440884176360974,"envelope_correlation":0.9860193002798707,"rms_gain_error_db":-0.6005931467044898,"reference_mean_power":0.0021283926153372414,"output_mean_power":0.0018534993102862966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0698576130619868e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007463968445819582,"output_power":0.00062269456369082,"reference_active":true,"delta_db":-0.7869472052998325},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014378870760237378,"output_power":0.00013888975059485716,"reference_active":true,"delta_db":-0.1505458216928762},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4173849116031066e-05,"output_power":1.2389639471724514e-05,"reference_active":true,"delta_db":-0.5842913633363837}],"active_band_count":3,"active_band_mae_db":0.5072614634430308}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00023719519351520602,"log_rms_mae_db":2.1570226864440762,"rms_envelope_nrmse":0.11897534754341584,"envelope_correlation":0.9909721356488851,"rms_gain_error_db":-0.6005998690919291,"reference_mean_power":0.0021390960955203144,"output_mean_power":0.0018628174952473155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.624199297765067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007211274534636877,"output_power":0.0006060520884746781,"reference_active":true,"delta_db":-0.7550207739054954},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.136224424017926e-05,"output_power":5.9400999683616724e-05,"reference_active":true,"delta_db":-0.14107480996822733},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3649293472151977e-06,"output_power":1.4350092148501193e-06,"reference_active":true,"delta_db":0.21744518292175916}],"active_band_count":3,"active_band_mae_db":0.37118025559849394}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":3.1733451269106405,"rms_envelope_nrmse":0.2780983518510005,"envelope_correlation":0.9435072397500707,"rms_gain_error_db":-0.0930754997192034,"reference_mean_power":9.73047769043704e-05,"output_mean_power":9.524158434471168e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.517204404381221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.13601836765628e-05,"output_power":4.44452777765816e-05,"reference_active":true,"delta_db":0.31243166072318934},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.821587959830413e-06,"output_power":5.0587712786918175e-06,"reference_active":true,"delta_db":0.20854950206776315},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.214811101263015e-06,"output_power":1.6494399094114537e-06,"reference_active":true,"delta_db":-1.2800019322421587}],"active_band_count":3,"active_band_mae_db":0.6003276983443704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":2.6535121019262644,"rms_envelope_nrmse":0.23308856817865198,"envelope_correlation":0.9572454587901429,"rms_gain_error_db":-0.09302007551956712,"reference_mean_power":9.778712425213711e-05,"output_mean_power":9.57149257823822e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.922472441667661e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.880871023429881e-05,"output_power":4.1698469560362855e-05,"reference_active":true,"delta_db":0.3119090588648936},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.146701619692129e-06,"output_power":2.516705039469721e-06,"reference_active":true,"delta_db":0.6906063478433783},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6121018734527105e-07,"output_power":1.255998120441628e-07,"reference_active":true,"delta_db":-1.084034931854684}],"active_band_count":3,"active_band_mae_db":0.6955167795209852}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":5.102429453852313,"rms_envelope_nrmse":0.3095542983985588,"envelope_correlation":0.9440680086870261,"rms_gain_error_db":-1.824766164431331,"reference_mean_power":3.136903007980472e-05,"output_mean_power":2.060746041218617e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.994485115758079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0899878382466337e-05,"output_power":7.3247921959045925e-06,"reference_active":true,"delta_db":-1.7262634395985288},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.551902798888814e-06,"output_power":1.3592924934076645e-06,"reference_active":true,"delta_db":-2.7355120968850395},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4088478818809362e-06,"output_power":6.884238709654166e-07,"reference_active":true,"delta_db":-3.1100818238747046}],"active_band_count":3,"active_band_mae_db":2.5239524534527575}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":4.557526045054289,"rms_envelope_nrmse":0.27033926497967165,"envelope_correlation":0.9636512215356691,"rms_gain_error_db":-1.8246151814214189,"reference_mean_power":3.151908471872328e-05,"output_mean_power":2.070675663991627e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7532611532165262e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0780259770731747e-05,"output_power":6.953658734334192e-06,"reference_active":true,"delta_db":-1.9041585318284975},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1761748042676719e-06,"output_power":5.87271384094705e-07,"reference_active":true,"delta_db":-3.0163303227309317},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.367404555815228e-08,"output_power":5.1723833142266565e-08,"reference_active":true,"delta_db":-2.5792857474285174}],"active_band_count":3,"active_band_mae_db":2.4999248673293155}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"golf","status":"measured","mos":{"status":"measured","value":3.6879308223724365,"predicted_mos":3.6879308223724365,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"82c233084a7010ba9911918d90e6363a34cb3598ee4981e553d3ce487d87d574","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.286266180999519},"started_at":"2026-10-13T05:01:58.742923+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003280461999565887,"native_input_samples":144000,"native_input_sha256":"1dd97abec91ea7c29e58c5ecf3372892bffa19ee22c0694339207fc541cd2b8b","output_sha256":"5a3b5b6f26952ce760f199b306ba48810e8d711f8b331f9bd3d5c1f1484724d6","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6032718030000979,"synthesis_s":0.1205613150004865,"total_s":0.7238331180005844,"rtf":0.1206388530000974,"output_peak":0.38443523645401,"output_rms":0.050085834713811804,"finished_at":"2026-10-13T05:01:59.477798+00:00","analysis_stages":{"external_f0_s":0.08282578099988314},"evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"5a3b5b6f26952ce760f199b306ba48810e8d711f8b331f9bd3d5c1f1484724d6","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.999565449914563,"output_peak":0.38481645214884164},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.385273118072798,"spectral_convergence":0.44804535165188564,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.645563525134683,"spectral_convergence":0.27616635602989903,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.699969166407952,"spectral_convergence":0.298382777132762,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003365723233078328,"output_power":0.0002836760274525932,"delta_db":-0.7425576410254262,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011260883322941973,"output_power":0.0010191922179887182,"delta_db":-0.43316359691669,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015443214665304852,"output_power":0.0011067342758656145,"delta_db":-1.446943479363766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.140612595294537e-05,"output_power":5.705127433406933e-05,"delta_db":-1.5439173799177452,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.304100735296745e-05,"output_power":3.579582046285508e-05,"delta_db":-1.7077944240847944,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9511981053084483e-05,"output_power":9.688164460317787e-06,"delta_db":-3.040598630751026,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4858291920099083,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9977296715033566,"pitch":null,"output_rms_dbfs":-25.999565449914563,"output_peak":0.38481645214884164,"reference_rms_dbfs":-25.001835778411206,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.910268603205144},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9130202401781083,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05622224858978439,"rms_dbfs":-25.001835778411206,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.124772794022103,"mean":-2.022617448351419e-06,"zero_energy":false},"output":{"rms":0.05012123083648454,"rms_dbfs":-25.999565449914563,"peak":0.38481645214884164,"peak_dbfs":-8.294927374850937,"crest_factor_db":17.704638075063624,"mean":0.00019082089738711123,"zero_energy":false},"rms_difference_db":-0.9977296715033575},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":1.972998092076244e-06,"output_rms_dbfs":-57.04873334717986,"output_zero_energy":false,"output_peak":0.00842605990462553,"output_energy_fraction":0.0001232492376006338,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":317,"both_estimated_voiced_frames":290,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9354838709677419,"absolute_difference_cents_median":4.524644389501229,"absolute_difference_cents_p95":30.713077442400856,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00025375764068356454,"log_rms_mae_db":1.8193778978252586,"rms_envelope_nrmse":0.19117802748541732,"envelope_correlation":0.9713585657465593,"rms_gain_error_db":-0.9977336927484373,"reference_mean_power":0.003163577316472044,"output_mean_power":0.002514230460234659},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1772769580561777e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082953986151575,"output_power":0.0006694303551710449,"reference_active":true,"delta_db":-0.9313185836912226},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013714403929222555,"output_power":0.00012474065720768695,"reference_active":true,"delta_db":-0.41168908715976255},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744905078984129e-05,"output_power":1.5001413750562677e-05,"reference_active":true,"delta_db":-0.6563961737296421}],"active_band_count":3,"active_band_mae_db":0.6664679481935424}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002402786371599188,"log_rms_mae_db":1.4086936367387937,"rms_envelope_nrmse":0.16850267310696734,"envelope_correlation":0.9787498812687938,"rms_gain_error_db":-0.9977426497120621,"reference_mean_power":0.0031794956531267147,"output_mean_power":0.002526876233034716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0349881387369278e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000778720918505812,"output_power":0.0006282213560513797,"reference_active":true,"delta_db":-0.9326914522886725},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.852923390206432e-05,"output_power":5.5637679697255486e-05,"reference_active":true,"delta_db":-0.22003829638823874},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3081879232027254e-06,"output_power":1.0205942138049376e-06,"reference_active":true,"delta_db":-1.0781703374068454}],"active_band_count":3,"active_band_mae_db":0.7436333620279189}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002501327279402343,"log_rms_mae_db":1.7977506628177289,"rms_envelope_nrmse":0.21758518445108985,"envelope_correlation":0.9682014431080658,"rms_gain_error_db":-1.0098572436707043,"reference_mean_power":0.002131948250783903,"output_mean_power":0.001689627363852766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.066107286516685e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007428314997104514,"output_power":0.0005853635026004816,"reference_active":true,"delta_db":-1.0346467098907528},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014365955862227154,"output_power":0.00012962992022604505,"reference_active":true,"delta_db":-0.4462927401155704},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4149457363621171e-05,"output_power":1.2101981372963186e-05,"reference_active":true,"delta_db":-0.6788330469956555}],"active_band_count":3,"active_band_mae_db":0.7199241656673262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00023719519351520602,"log_rms_mae_db":1.3988710174595573,"rms_envelope_nrmse":0.1913393924837815,"envelope_correlation":0.975853546637661,"rms_gain_error_db":-1.0098568230482825,"reference_mean_power":0.0021426878497774523,"output_mean_power":0.0016981389545852423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.587316635973934e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007174671113357602,"output_power":0.0005672578883815241,"reference_active":true,"delta_db":-1.020214538508094},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.130984880414411e-05,"output_power":5.797431253699416e-05,"reference_active":true,"delta_db":-0.2429463766793327},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.356445356751355e-06,"output_power":1.0601985917386027e-06,"reference_active":true,"delta_db":-1.0701508024821378}],"active_band_count":3,"active_band_mae_db":0.7777705725565216}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":1.940533392078698,"rms_envelope_nrmse":0.2771559432470692,"envelope_correlation":0.9559554874390804,"rms_gain_error_db":-1.561973505209901,"reference_mean_power":9.746567824558402e-05,"output_mean_power":6.802277714889192e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.531468509554753e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.137689346278488e-05,"output_power":3.4000311260544655e-05,"reference_active":true,"delta_db":-0.85274988273997},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.8179476439466095e-06,"output_power":4.344163693901356e-06,"reference_active":true,"delta_db":-0.44955894576069605},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.219910088250087e-06,"output_power":1.57686405920723e-06,"reference_active":true,"delta_db":-1.4854113022904465}],"active_band_count":3,"active_band_mae_db":0.9292400435970375}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":1.6062215840574716,"rms_envelope_nrmse":0.244999767636661,"envelope_correlation":0.9685253507693581,"rms_gain_error_db":-1.562040665600739,"reference_mean_power":9.794927075573179e-05,"output_mean_power":6.835922657404581e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.935073873440217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8801093654727125e-05,"output_power":3.185927360842772e-05,"reference_active":true,"delta_db":-0.8560809720849827},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1461516536409356e-06,"output_power":1.8713998541194061e-06,"reference_active":true,"delta_db":-0.5949381600447885},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6117100095750884e-07,"output_power":1.2814138533857644e-07,"reference_active":true,"delta_db":-0.9959748827410073}],"active_band_count":3,"active_band_mae_db":0.8156646716235928}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":3.2418950218067906,"rms_envelope_nrmse":0.29064917592060807,"envelope_correlation":0.9621449613951972,"rms_gain_error_db":-2.131862394178859,"reference_mean_power":3.1416363901466996e-05,"output_mean_power":1.9229574714200084e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.9998851030195438e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0929134094971484e-05,"output_power":6.976162763132864e-06,"reference_active":true,"delta_db":-1.949691498596739},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.549657316671302e-06,"output_power":1.7247207285038163e-06,"reference_active":true,"delta_db":-1.6976303062713862},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4107214256087128e-06,"output_power":1.134089121014364e-06,"reference_active":true,"delta_db":-0.9479407792436677}],"active_band_count":3,"active_band_mae_db":1.5317541947039308}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":3.0499822354140558,"rms_envelope_nrmse":0.2548650547626259,"envelope_correlation":0.9813483378236189,"rms_gain_error_db":-2.131856558959076,"reference_mean_power":3.1568492623806487e-05,"output_mean_power":1.932271681932128e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.758022093946705e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0805501623318211e-05,"output_power":7.096970466605733e-06,"reference_active":true,"delta_db":-1.8257193496470625},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1684480107084793e-06,"output_power":7.748927590882368e-07,"reference_active":true,"delta_db":-1.7836779083509657},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395382334492549e-08,"output_power":8.165791105461066e-08,"reference_active":true,"delta_db":-0.6091619216770939}],"active_band_count":3,"active_band_mae_db":1.4061863932250407}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"sifigan","status":"measured","mos":{"status":"measured","value":4.019847393035889,"predicted_mos":4.019847393035889,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"d8a2d6b016b956af77115e497f5812946f9e3f571f95b4105187a3da9d86bf62","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.468883153000206},"started_at":"2026-10-13T05:05:29.758139+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031056449997777236,"native_input_samples":144000,"native_input_sha256":"1dd97abec91ea7c29e58c5ecf3372892bffa19ee22c0694339207fc541cd2b8b","output_sha256":"1b64dab25e5e4a4b152d8280290ed4f99f5e7d5e9357e9002ffcd7e1ec5dec27","output_samples":144120,"length_delta_samples":120,"analysis_s":2.418054183000095,"synthesis_s":3.6591441869995833,"total_s":6.077198369999678,"rtf":1.0128663949999464,"output_peak":0.41642579436302185,"output_rms":0.054103401068496815,"finished_at":"2026-10-13T05:05:35.851920+00:00","analysis_stages":{"external_f0_s":1.610155479000241},"evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"1b64dab25e5e4a4b152d8280290ed4f99f5e7d5e9357e9002ffcd7e1ec5dec27","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.32961744810735,"output_peak":0.413895430001247},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.960636100351944,"spectral_convergence":0.4472500369336442,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.438796642621849,"spectral_convergence":0.26055392345526945,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.281692373775326,"spectral_convergence":0.24596943029388885,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.0003435697976331104,"delta_db":0.09658200641292355,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.001082766027543541,"delta_db":-0.16314539106389686,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0013664740184753663,"delta_db":-0.5241140511467364,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":7.386790999040638e-05,"delta_db":-0.4147796584400447,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":5.17392155976338e-05,"delta_db":-0.10118783501440104,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":1.517363553604215e-05,"delta_db":-1.0853285958020333,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.39752292298000597,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.31694117400445254,"pitch":null,"output_rms_dbfs":-25.3260049393252,"output_peak":0.413895430001247,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.560375038916373},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9189052733343507,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05617548255220346,"rms_dbfs":-25.009063765320747,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.132000780931644,"mean":-3.836313883463542e-06,"zero_energy":false},"output":{"rms":0.054162631073241825,"rms_dbfs":-25.3260049393252,"peak":0.413895430001247,"peak_dbfs":-7.662187376125791,"crest_factor_db":17.663817563199405,"mean":8.276603263698202e-05,"zero_energy":false},"rms_difference_db":-0.31694117400445165},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":2.523471720985097e-06,"output_rms_dbfs":-55.98001557777508,"output_zero_energy":false,"output_peak":0.041906197630505716,"output_energy_fraction":0.00013476451087301785,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":324,"both_estimated_voiced_frames":296,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9548387096774194,"absolute_difference_cents_median":7.76184102199657,"absolute_difference_cents_p95":31.217408000751497,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025375764068356454,"log_rms_mae_db":2.2488212289609817,"rms_envelope_nrmse":0.16221847782131638,"envelope_correlation":0.9732906989266733,"rms_gain_error_db":-0.31693711729375584,"reference_mean_power":0.0031583120304531104,"output_mean_power":0.0029360356385011053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1807904078754827e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008327873716353349,"output_power":0.0007830131900761387,"reference_active":true,"delta_db":-0.26765052991351407},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013733244323789595,"output_power":0.0001270136408849137,"reference_active":true,"delta_db":-0.33922781132946445},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.747498790065721e-05,"output_power":1.3861030348589466e-05,"reference_active":true,"delta_db":-1.0062136931255463}],"active_band_count":3,"active_band_mae_db":0.5376973447895083}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002402786371599188,"log_rms_mae_db":1.770037990612012,"rms_envelope_nrmse":0.10665633424743579,"envelope_correlation":0.988097602154129,"rms_gain_error_db":-0.31692769080557287,"reference_mean_power":0.0031741738524038612,"output_mean_power":0.0029507875383548137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0384864838590044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007821613563464859,"output_power":0.0007315410373292855,"reference_active":true,"delta_db":-0.2905766148291101},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.85949543312588e-05,"output_power":5.435783633172557e-05,"reference_active":true,"delta_db":-0.325980584308301},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3154862476080321e-06,"output_power":1.1117083749273829e-06,"reference_active":true,"delta_db":-0.7309543485141595}],"active_band_count":3,"active_band_mae_db":0.44917051588385687}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002501327279402343,"log_rms_mae_db":2.1311596625269082,"rms_envelope_nrmse":0.19406780123598663,"envelope_correlation":0.9702441026356705,"rms_gain_error_db":-0.40823866219368043,"reference_mean_power":0.0021283926153372414,"output_mean_power":0.0019374382840701306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0698576130619868e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007463968445819582,"output_power":0.0006836865203293745,"reference_active":true,"delta_db":-0.3811277738128135},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014378870760237378,"output_power":0.00013164827278784296,"reference_active":true,"delta_db":-0.3830961463617131},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4173849116031066e-05,"output_power":1.0719741093741482e-05,"reference_active":true,"delta_db":-1.213035090042004}],"active_band_count":3,"active_band_mae_db":0.6590863367388434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00023719519351520602,"log_rms_mae_db":1.536498958879463,"rms_envelope_nrmse":0.12923862104705508,"envelope_correlation":0.9866686572444883,"rms_gain_error_db":-0.4082385606856139,"reference_mean_power":0.0021390960955203144,"output_mean_power":0.0019471815189398246},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.624199297765067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007211274534636877,"output_power":0.0006583722604394932,"reference_active":true,"delta_db":-0.39540505330363157},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.136224424017926e-05,"output_power":5.5935844995565834e-05,"reference_active":true,"delta_db":-0.4021103181648538},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3649293472151977e-06,"output_power":1.1458116529440549e-06,"reference_active":true,"delta_db":-0.7599693686425333}],"active_band_count":3,"active_band_mae_db":0.5191615800370063}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":2.1054982993645748,"rms_envelope_nrmse":0.23434015568184188,"envelope_correlation":0.9589113504646072,"rms_gain_error_db":-0.2534578122312869,"reference_mean_power":9.73047769043704e-05,"output_mean_power":9.178852268295904e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.517204404381221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.13601836765628e-05,"output_power":3.985658139758123e-05,"reference_active":true,"delta_db":-0.16082413352665495},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.821587959830413e-06,"output_power":4.327374780245308e-06,"reference_active":true,"delta_db":-0.4696558426874154},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.214811101263015e-06,"output_power":1.6708500734015735e-06,"reference_active":true,"delta_db":-1.2239920957243169}],"active_band_count":3,"active_band_mae_db":0.6181573573127958}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":1.4240350263038641,"rms_envelope_nrmse":0.14036231920351194,"envelope_correlation":0.9845055726578367,"rms_gain_error_db":-0.25340930933942474,"reference_mean_power":9.778712425213711e-05,"output_mean_power":9.224455572970423e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.922472441667661e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.880871023429881e-05,"output_power":3.7252470777022145e-05,"reference_active":true,"delta_db":-0.1777412696880541},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.146701619692129e-06,"output_power":1.902475075974481e-06,"reference_active":true,"delta_db":-0.5245270811094351},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6121018734527105e-07,"output_power":1.2778782552229672e-07,"reference_active":true,"delta_db":-1.0090300256946125}],"active_band_count":3,"active_band_mae_db":0.5704327921640339}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":2.2402614754590764,"rms_envelope_nrmse":0.25903100845022153,"envelope_correlation":0.940558883507792,"rms_gain_error_db":-0.5762315925321586,"reference_mean_power":3.136903007980472e-05,"output_mean_power":2.7471219019047783e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.994485115758079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0899878382466337e-05,"output_power":9.289183760999165e-06,"reference_active":true,"delta_db":-0.6944409795395523},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.551902798888814e-06,"output_power":2.0318438414755644e-06,"reference_active":true,"delta_db":-0.9897380129197025},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4088478818809362e-06,"output_power":1.0131708074276387e-06,"reference_active":true,"delta_db":-1.4318143544500341}],"active_band_count":3,"active_band_mae_db":1.0386644489697632}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":1.5018284470661045,"rms_envelope_nrmse":0.14667455205314134,"envelope_correlation":0.9815346193674821,"rms_gain_error_db":-0.5760446765143095,"reference_mean_power":3.151908471872328e-05,"output_mean_power":2.7603816384160496e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7532611532165262e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0780259770731747e-05,"output_power":9.027465648623223e-06,"reference_active":true,"delta_db":-0.7706338160153993},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1761748042676719e-06,"output_power":9.508635467104731e-07,"reference_active":true,"delta_db":-0.9235367365979231},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.367404555815228e-08,"output_power":6.964510931580992e-08,"reference_active":true,"delta_db":-1.287286523795359}],"active_band_count":3,"active_band_mae_db":0.9938190254695605}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.7571310997009277,"predicted_mos":3.7571310997009277,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"cc060960abad7464bf0dbf32be6e1ebd2470bcc2da787ea6976f3f11aa00e127","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.268564187001175},"started_at":"2026-10-13T05:10:04.706740+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029705199995078146,"native_input_samples":144000,"native_input_sha256":"1dd97abec91ea7c29e58c5ecf3372892bffa19ee22c0694339207fc541cd2b8b","output_sha256":"2bc4216a57128502d0a8f82881f481aa113e6053db1e47f71b6dfe7a42c6a49b","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7148365579996607,"synthesis_s":0.1513773320002656,"total_s":0.8662138899999263,"rtf":0.14436898166665438,"output_peak":0.4951316714286804,"output_rms":0.054687762099632525,"finished_at":"2026-10-13T05:10:05.585011+00:00","analysis_stages":{"external_f0_s":0.1500062960003561},"evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"2bc4216a57128502d0a8f82881f481aa113e6053db1e47f71b6dfe7a42c6a49b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.23613072936974,"output_peak":0.4951991772483022},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.564545248634895,"spectral_convergence":0.18594606979270742,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.398684151303968,"spectral_convergence":0.18939824835936234,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.809071806949826,"spectral_convergence":0.20312912708435074,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.0003319975833955173,"delta_db":-0.05221855953741635,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.001090711817346707,"delta_db":-0.13139141623228226,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0014395834587908616,"delta_db":-0.2977593589447202,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":7.317045258215424e-05,"delta_db":-0.4559803587494913,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":4.7775903033801716e-05,"delta_db":-0.44729719126846673,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":1.6683264371853557e-05,"delta_db":-0.6734147596204147,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3430102740587986,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21983591132027636,"pitch":null,"output_rms_dbfs":-25.22889967664102,"output_peak":0.4951991772483022,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.590767068962896},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9684511547038747,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05617548255220346,"rms_dbfs":-25.009063765320747,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.132000780931644,"mean":-3.836313883463542e-06,"zero_energy":false},"output":{"rms":0.05477154808402713,"rms_dbfs":-25.228899676641024,"peak":0.4951991772483022,"peak_dbfs":-6.104401710971175,"crest_factor_db":19.124497965669846,"mean":1.7922204677848985e-05,"zero_energy":false},"rms_difference_db":-0.2198359113202777},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":2.5405392301965556e-06,"output_rms_dbfs":-55.95074094465025,"output_zero_energy":false,"output_peak":0.021972409632818775,"output_energy_fraction":0.0001326760326134692,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":293,"both_estimated_voiced_frames":284,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9161290322580645,"absolute_difference_cents_median":0.8632434178897742,"absolute_difference_cents_p95":9.390587561366374,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025375764068356454,"log_rms_mae_db":0.6890125349802286,"rms_envelope_nrmse":0.10291839362732036,"envelope_correlation":0.989478446050394,"rms_gain_error_db":-0.2198365780263549,"reference_mean_power":0.0031583120304531104,"output_mean_power":0.0030024195328308394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1807904078754827e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008327873716353349,"output_power":0.000825743573804762,"reference_active":true,"delta_db":-0.03688928362043322},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013733244323789595,"output_power":0.00012772905731419122,"reference_active":true,"delta_db":-0.31483439581355743},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.747498790065721e-05,"output_power":1.5029444481185097e-05,"reference_active":true,"delta_db":-0.6547395524392018}],"active_band_count":3,"active_band_mae_db":0.3354877439577308}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002402786371599188,"log_rms_mae_db":0.4385884361959705,"rms_envelope_nrmse":0.06925725107945618,"envelope_correlation":0.9951393028991359,"rms_gain_error_db":-0.21983502893413553,"reference_mean_power":0.0031741738524038612,"output_mean_power":0.0030174995005275277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0384864838590044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007821613563464859,"output_power":0.00077507586879576,"reference_active":true,"delta_db":-0.03952139508312061},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.85949543312588e-05,"output_power":5.41370202307185e-05,"reference_active":true,"delta_db":-0.34365872125165503},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3154862476080321e-06,"output_power":1.1639925216077903e-06,"reference_active":true,"delta_db":-0.5313612215461718}],"active_band_count":3,"active_band_mae_db":0.3048471126269825}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002501327279402343,"log_rms_mae_db":0.854868074913828,"rms_envelope_nrmse":0.12184747456631287,"envelope_correlation":0.9884133515332257,"rms_gain_error_db":-0.23690157710990156,"reference_mean_power":0.0021283926153372414,"output_mean_power":0.0020154015383348663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0698576130619868e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007463968445819582,"output_power":0.000739884190073964,"reference_active":true,"delta_db":-0.038060473786399304},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014378870760237378,"output_power":0.0001323462699319709,"reference_active":true,"delta_db":-0.36013074629458414},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4173849116031066e-05,"output_power":1.2043712830265187e-05,"reference_active":true,"delta_db":-0.7072741361632675}],"active_band_count":3,"active_band_mae_db":0.36848845208141695}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00023719519351520602,"log_rms_mae_db":0.5089664469785331,"rms_envelope_nrmse":0.08289007375920383,"envelope_correlation":0.994501153845876,"rms_gain_error_db":-0.23690131768064382,"reference_mean_power":0.0021390960955203144,"output_mean_power":0.002025536918339611},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.624199297765067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007211274534636877,"output_power":0.0007134817371147399,"reference_active":true,"delta_db":-0.046291685297442445},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.136224424017926e-05,"output_power":5.5959368628683484e-05,"reference_active":true,"delta_db":-0.40028429113886765},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3649293472151977e-06,"output_power":1.1623988239433336e-06,"reference_active":true,"delta_db":-0.6975500969730231}],"active_band_count":3,"active_band_mae_db":0.3813753578031111}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":1.0038350433189223,"rms_envelope_nrmse":0.1364032354546683,"envelope_correlation":0.9877800309778771,"rms_gain_error_db":-0.473428730365805,"reference_mean_power":9.73047769043704e-05,"output_mean_power":8.725519395414859e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.517204404381221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.13601836765628e-05,"output_power":3.942610413273945e-05,"reference_active":true,"delta_db":-0.2079859410991329},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.821587959830413e-06,"output_power":4.491652037963371e-06,"reference_active":true,"delta_db":-0.3078398926502094},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.214811101263015e-06,"output_power":1.4953385189286914e-06,"reference_active":true,"delta_db":-1.7059717105892829}],"active_band_count":3,"active_band_mae_db":0.7405991814462084}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":0.6536522428711161,"rms_envelope_nrmse":0.0967489725345628,"envelope_correlation":0.9945464356758903,"rms_gain_error_db":-0.4734029399093345,"reference_mean_power":9.778712425213711e-05,"output_mean_power":8.768824547019455e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.922472441667661e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.880871023429881e-05,"output_power":3.684333243002917e-05,"reference_active":true,"delta_db":-0.22570305015224768},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.146701619692129e-06,"output_power":1.9949028262307432e-06,"reference_active":true,"delta_db":-0.3184993836047958},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6121018734527105e-07,"output_power":1.0735001686708618e-07,"reference_active":true,"delta_db":-1.765903656706049}],"active_band_count":3,"active_band_mae_db":0.7700353634876974}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":1.1769609125148424,"rms_envelope_nrmse":0.1258835705087525,"envelope_correlation":0.9884039148936083,"rms_gain_error_db":-0.5255457629420115,"reference_mean_power":3.136903007980472e-05,"output_mean_power":2.779370952637348e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.994485115758079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0899878382466337e-05,"output_power":9.445840986121991e-06,"reference_active":true,"delta_db":-0.6218102197189497},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.551902798888814e-06,"output_power":2.183547849538999e-06,"reference_active":true,"delta_db":-0.6770141488879143},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4088478818809362e-06,"output_power":9.039062997982587e-07,"reference_active":true,"delta_db":-1.9274069015162194}],"active_band_count":3,"active_band_mae_db":1.075410423374361}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":0.8411099733626705,"rms_envelope_nrmse":0.0781714587569539,"envelope_correlation":0.9971986591823543,"rms_gain_error_db":-0.5256399650179961,"reference_mean_power":3.151908471872328e-05,"output_mean_power":2.792605577283676e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7532611532165262e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0780259770731747e-05,"output_power":9.23312315281527e-06,"reference_active":true,"delta_db":-0.6728059783730055},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1761748042676719e-06,"output_power":9.922881936518023e-07,"reference_active":true,"delta_db":-0.7383404770899066},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.367404555815228e-08,"output_power":5.4817521074854715e-08,"reference_active":true,"delta_db":-2.326998845335541}],"active_band_count":3,"active_band_mae_db":1.2460484335994844}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1284-1180-0028","method":"magphase","status":"measured","mos":{"status":"measured","value":3.630284547805786,"predicted_mos":3.630284547805786,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288433,"input_sample_rate":48000,"input_seconds":6.0090208333333335,"input_float32_sha256":"3a5225a4b40184c986083bc6f4c44b81bc7f294185fe78519e9d5bdecd1554e9","internal_samples":96145,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2058014960002765},"started_at":"2026-10-13T05:12:42.206802+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00916818199948466,"native_input_samples":288000,"native_input_sha256":"24b5496ea462917a1c31f13ab735edfa3b6ad307d5596506297fd79bcb5c7f07","output_sha256":"9d55267545ce449ba8e1bbf3015346ecd431a23abadeeb704aabf18927985fac","output_samples":288433,"length_delta_samples":433,"analysis_s":3.600724595000429,"synthesis_s":1.1088114899994252,"total_s":4.709536084999854,"rtf":0.7849226808333091,"output_peak":0.4576420703120772,"output_rms":0.05327710983972712,"finished_at":"2026-10-13T05:12:46.937701+00:00","evaluation_input":{"reference":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"output_sha256":"9d55267545ce449ba8e1bbf3015346ecd431a23abadeeb704aabf18927985fac","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96145,"length_delta_samples":145,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.46332121083058,"output_peak":0.4535774263690948},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9984918612512351,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.436691305547371,"spectral_convergence":0.41267080690070657,"spectral_convergence_status":"ok","amplitude_floor":1.8586249352090494e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.633647766750471,"spectral_convergence":0.20137790279715984,"spectral_convergence_status":"ok","amplitude_floor":1.6686015821705495e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.403560590273273,"spectral_convergence":0.1864604859478446,"spectral_convergence_status":"ok","amplitude_floor":1.5267872572435527e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003360135402914431,"output_power":0.0003031611293249239,"delta_db":-0.4468326223250818,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011242144422279699,"output_power":0.0010018435763805428,"delta_db":-0.5004924223907714,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015417458345753522,"output_power":0.0013817917975669758,"delta_db":-0.47570173394458215,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.127065821964785e-05,"output_power":7.952095470953964e-05,"delta_db":-0.09452191707863324,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.295885949753132e-05,"output_power":5.726380691572035e-05,"delta_db":0.33941592951222244,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.948150516097124e-05,"output_power":2.2993029746572184e-05,"delta_db":0.7197369330083815,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.42945025970994544,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4477110023060792,"pitch":null,"output_rms_dbfs":-25.456774767626822,"output_peak":0.4535774263690948,"reference_rms_dbfs":-25.009063765320747,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.824633220857038},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9382208545484668,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05617548255220346,"rms_dbfs":-25.009063765320747,"peak":0.508331298828125,"peak_dbfs":-5.877062984389099,"crest_factor_db":19.132000780931644,"mean":-3.836313883463542e-06,"zero_energy":false},"output":{"rms":0.053353296942590854,"rms_dbfs":-25.456774767626822,"peak":0.4535774263690948,"peak_dbfs":-6.86697135070191,"crest_factor_db":18.589803416924916,"mean":0.001116535519692175,"zero_energy":false},"rms_difference_db":-0.4477110023060753},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.4392940205289046e-06,"threshold_dbfs":-51.91161744409661,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":94,"quiet_samples":15040,"reference_mean_power":2.224621873230059e-06,"output_mean_power":2.0523417371459136e-06,"output_rms_dbfs":-56.877503228001196,"output_zero_energy":false,"output_peak":0.009779755183433084,"output_energy_fraction":0.00011295455714065852,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":310,"output_estimated_voiced_frames":303,"both_estimated_voiced_frames":290,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9354838709677419,"absolute_difference_cents_median":4.64039903596362,"absolute_difference_cents_p95":21.78019567639418,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025375764068356454,"log_rms_mae_db":1.4992936379309638,"rms_envelope_nrmse":0.12401587664782937,"envelope_correlation":0.9860873809140591,"rms_gain_error_db":-0.447710312527995,"reference_mean_power":0.0031583120304531104,"output_mean_power":0.0028489445949700077},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1807904078754827e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008327873716353349,"output_power":0.0007612359575230409,"reference_active":true,"delta_db":-0.3901483661399706},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013733244323789595,"output_power":0.00012429222782948214,"reference_active":true,"delta_db":-0.4332917408482539},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.747498790065721e-05,"output_power":1.717394446433707e-05,"reference_active":true,"delta_db":-0.0754682954088576}],"active_band_count":3,"active_band_mae_db":0.2996361341323607}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002402786371599188,"log_rms_mae_db":0.9812513765219749,"rms_envelope_nrmse":0.08882015080659614,"envelope_correlation":0.9935825419536441,"rms_gain_error_db":-0.44771186717569866,"reference_mean_power":0.0031741738524038612,"output_mean_power":0.0028632516723996835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0384864838590044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007821613563464859,"output_power":0.0007157710896943622,"reference_active":true,"delta_db":-0.38522202222523116},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.85949543312588e-05,"output_power":5.3735873061400694e-05,"reference_active":true,"delta_db":-0.37595910734840776},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3154862476080321e-06,"output_power":1.1891138091751575e-06,"reference_active":true,"delta_db":-0.43862889630958124}],"active_band_count":3,"active_band_mae_db":0.39993667529440674}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002501327279402343,"log_rms_mae_db":1.7339568169230732,"rms_envelope_nrmse":0.14068992291412385,"envelope_correlation":0.9858366970266341,"rms_gain_error_db":-0.4922143180541043,"reference_mean_power":0.0021283926153372414,"output_mean_power":0.0019003356320495296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0698576130619868e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007463968445819582,"output_power":0.000667856369534309,"reference_active":true,"delta_db":-0.4828672243041031},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014378870760237378,"output_power":0.00012587070999330418,"reference_active":true,"delta_db":-0.5780009830615939},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4173849116031066e-05,"output_power":1.2638932426813854e-05,"reference_active":true,"delta_db":-0.4977741335582015}],"active_band_count":3,"active_band_mae_db":0.5195474469746328}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00023719519351520602,"log_rms_mae_db":1.2639940014418793,"rms_envelope_nrmse":0.1006165976595218,"envelope_correlation":0.9934692345998513,"rms_gain_error_db":-0.4922138303928719,"reference_mean_power":0.0021390960955203144,"output_mean_power":0.0019098924502247397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.624199297765067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007211274534636877,"output_power":0.0006446562569738353,"reference_active":true,"delta_db":-0.4868382730938043},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.136224424017926e-05,"output_power":5.407752959883766e-05,"reference_active":true,"delta_db":-0.548843912646442},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3649293472151977e-06,"output_power":1.1973848337552865e-06,"reference_active":true,"delta_db":-0.5687641858066338}],"active_band_count":3,"active_band_mae_db":0.5348154571822934}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.486004098842285e-05,"log_rms_mae_db":1.6887527889209002,"rms_envelope_nrmse":0.15807169116380054,"envelope_correlation":0.9816183537306888,"rms_gain_error_db":-0.02628729757385049,"reference_mean_power":9.73047769043704e-05,"output_mean_power":9.671758225512243e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.517204404381221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.13601836765628e-05,"output_power":4.293462505446766e-05,"reference_active":true,"delta_db":0.16225215833540385},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.821587959830413e-06,"output_power":5.045190447725177e-06,"reference_active":true,"delta_db":0.19687470823924078},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.214811101263015e-06,"output_power":2.1728388052136486e-06,"reference_active":true,"delta_db":-0.08309182820940395}],"active_band_count":3,"active_band_mae_db":0.14740623159468288}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.523096269973881e-05,"log_rms_mae_db":1.1196993674414402,"rms_envelope_nrmse":0.08828043153886644,"envelope_correlation":0.9938029759185582,"rms_gain_error_db":-0.026395112592966664,"reference_mean_power":9.778712425213711e-05,"output_mean_power":9.719460591565183e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.922472441667661e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.880871023429881e-05,"output_power":4.0419711500565916e-05,"reference_active":true,"delta_db":0.1766399970293843},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.146701619692129e-06,"output_power":2.2172054144179287e-06,"reference_active":true,"delta_db":0.1403424650189291},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6121018734527105e-07,"output_power":1.6232936068599412e-07,"reference_active":true,"delta_db":0.03004595560543475}],"active_band_count":3,"active_band_mae_db":0.11567613921791604}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.595322597915998e-05,"log_rms_mae_db":1.761119032965968,"rms_envelope_nrmse":0.2506558714638726,"envelope_correlation":0.9569486461390307,"rms_gain_error_db":0.7244980215783856,"reference_mean_power":3.136903007980472e-05,"output_mean_power":3.706388099995087e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.994485115758079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0899878382466337e-05,"output_power":1.3320555466048982e-05,"reference_active":true,"delta_db":0.8710068301377849},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.551902798888814e-06,"output_power":3.758948716967513e-06,"reference_active":true,"delta_db":1.682022724750401},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4088478818809362e-06,"output_power":1.6773833186562223e-06,"reference_active":true,"delta_db":0.7576821632292775}],"active_band_count":3,"active_band_mae_db":1.1035705727058212}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.9738091980327884e-05,"log_rms_mae_db":1.1539128631645965,"rms_envelope_nrmse":0.15790723005256205,"envelope_correlation":0.9853050044981435,"rms_gain_error_db":0.7246731225647031,"reference_mean_power":3.151908471872328e-05,"output_mean_power":3.7242678657020236e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7532611532165262e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0780259770731747e-05,"output_power":1.3384477883787129e-05,"reference_active":true,"delta_db":0.9397220830066738},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1761748042676719e-06,"output_power":1.798412133614807e-06,"reference_active":true,"delta_db":1.8441735220381095},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.367404555815228e-08,"output_power":1.1798657442723234e-07,"reference_active":true,"delta_db":1.0021331541260106}],"active_band_count":3,"active_band_mae_db":1.2620095863902645}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},{"case":"librispeech_1320-122617-0019","method":"hnm","status":"measured","mos":{"status":"measured","value":2.6974215507507324,"predicted_mos":2.6974215507507324,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c4a3c3ab992dbc93cfafc3962996584b1c1123daa429a6e2abdc9c4fe2c5bedb","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.284068949000357},"started_at":"2026-10-13T04:38:16.336787+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.637700006540399e-05,"native_input_samples":96000,"native_input_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","output_sha256":"91d57d5de949f7dc77ac205ec3e3b2ed57d55fc4be4f8c17d1347a6bf1d57e93","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3359167119997437,"synthesis_s":0.5139219370003048,"total_s":1.8498386490000485,"rtf":0.3083064415000081,"output_peak":0.6061180473494444,"output_rms":0.08907172238004034,"finished_at":"2026-10-13T04:38:18.193553+00:00","evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"91d57d5de949f7dc77ac205ec3e3b2ed57d55fc4be4f8c17d1347a6bf1d57e93","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.92182176767597,"spectral_convergence":0.18973594340589214,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.54600806720554,"spectral_convergence":0.19897298338958297,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.565548494022616,"spectral_convergence":0.19708179078248478,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.004049973856024657,"delta_db":-0.09289425131004392,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.0025754676321773098,"delta_db":-0.06403632960765708,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.0007867635295516601,"delta_db":-0.11125820877925105,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.00040949517367836434,"delta_db":-0.16396334702068727,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":6.80377500588299e-05,"delta_db":-0.13982826744575594,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":4.403378625615781e-05,"delta_db":-0.6095418400117855,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1969203740291968,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09253039130554973,"pitch":null,"output_rms_dbfs":-21.005202992999887,"output_peak":0.6061180473494444,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.677792776301376},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.92182176767597,"spectral_convergence":0.18973594340589214,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.54600806720554,"spectral_convergence":0.19897298338958297,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.565548494022616,"spectral_convergence":0.19708179078248478,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.004049973856024657,"delta_db":-0.09289425131004392,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.0025754676321773098,"delta_db":-0.06403632960765708,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.0007867635295516601,"delta_db":-0.11125820877925105,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.00040949517367836434,"delta_db":-0.16396334702068727,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":6.80377500588299e-05,"delta_db":-0.13982826744575594,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":4.403378625615781e-05,"delta_db":-0.6095418400117855,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1969203740291968,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09253039130554973,"pitch":null,"output_rms_dbfs":-21.005202992999887,"output_peak":0.6061180473494444,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.677792776301376},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8939476544883853,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900256715962862,"rms_dbfs":-20.91267260169434,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.018647707074955,"mean":-0.0005092315673828125,"zero_energy":false},"output":{"rms":0.08907172238004034,"rms_dbfs":-21.005202992999887,"peak":0.6061180473494444,"peak_dbfs":-4.348855690940266,"crest_factor_db":16.65634730205962,"mean":6.395361499908566e-06,"zero_energy":false},"rms_difference_db":-0.09253039130554797},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":4.368299068202485e-06,"output_rms_dbfs":-53.596876360487826,"output_zero_energy":false,"output_peak":0.011138099268205049,"output_energy_fraction":5.3224232359676866e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":292,"both_estimated_voiced_frames":268,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9745454545454545,"absolute_difference_cents_median":1.7501214498862123,"absolute_difference_cents_p95":8.662145792354119,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025953969251226393,"log_rms_mae_db":0.9469349678474418,"rms_envelope_nrmse":0.061343308282246316,"envelope_correlation":0.9956327931986654,"rms_gain_error_db":-0.09252756580198251,"reference_mean_power":0.008109529837298932,"output_mean_power":0.007938581714057783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5635786081597607e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013500334014131179,"output_power":0.0013458433528097029,"reference_active":true,"delta_db":-0.013499997355589492},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020457869490256438,"output_power":0.00018486838857943518,"reference_active":true,"delta_db":-0.43997748051959407},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.388028407734598e-05,"output_power":3.378969863178933e-05,"reference_active":true,"delta_db":-0.011627247730784555}],"active_band_count":3,"active_band_mae_db":0.1550349085353227}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024049824298809385,"log_rms_mae_db":0.4884109585512161,"rms_envelope_nrmse":0.03333084318737713,"envelope_correlation":0.998678187921266,"rms_gain_error_db":-0.0928650323852705,"reference_mean_power":0.008139876536629037,"output_mean_power":0.007967669559579591},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2818780696338306e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012006574574251973,"output_power":0.0012046754833130772,"reference_active":true,"delta_db":0.014509493175099909},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.159020687440259e-05,"output_power":8.352190595347815e-05,"reference_active":true,"delta_db":-0.40048643672506845},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.235316344629429e-06,"output_power":2.2662036376381194e-06,"reference_active":true,"delta_db":0.059599386380462305}],"active_band_count":3,"active_band_mae_db":0.15819843876021022}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002074995834794702,"log_rms_mae_db":1.7779557332129212,"rms_envelope_nrmse":0.11456439954335496,"envelope_correlation":0.9900727817834278,"rms_gain_error_db":-0.1893721449602778,"reference_mean_power":0.0020359851811281022,"output_mean_power":0.0019491147057491872},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.345568889858965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045532158312677236,"output_power":0.00043300390837755243,"reference_active":true,"delta_db":-0.21826420799908597},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.561332252218219e-05,"output_power":7.141840192626304e-05,"reference_active":true,"delta_db":-0.7873122358514935},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0669601144719856e-05,"output_power":1.0571083141468974e-05,"reference_active":true,"delta_db":-0.04028696196395716}],"active_band_count":3,"active_band_mae_db":0.34862113527151223}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001833599147850085,"log_rms_mae_db":1.085392086198682,"rms_envelope_nrmse":0.058734244009977565,"envelope_correlation":0.9974959930620765,"rms_gain_error_db":-0.18926314911527442,"reference_mean_power":0.0020457391187242045,"output_mean_power":0.0019585016191716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.75393473065512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00043556686984963635,"output_power":0.0004122970144714263,"reference_active":true,"delta_db":-0.23844648974234267},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6912146426205404e-05,"output_power":3.201552680267752e-05,"reference_active":true,"delta_db":-0.6180864761546502},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.183033529622032e-07,"output_power":6.875979095910942e-07,"reference_active":true,"delta_db":-0.18973345986080528}],"active_band_count":3,"active_band_mae_db":0.3487554752525994}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001109356585659573,"log_rms_mae_db":1.9744915661169506,"rms_envelope_nrmse":0.18030543654319753,"envelope_correlation":0.9757079866718533,"rms_gain_error_db":-0.1718661876332188,"reference_mean_power":0.000425572053658366,"output_mean_power":0.0004090594992132591},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9996440016989412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011758808293094378,"output_power":0.00010985684742975826,"reference_active":true,"delta_db":-0.29536178126346896},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.147530754834166e-05,"output_power":1.8504284183994887e-05,"reference_active":true,"delta_db":-0.6466710249269956},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.05615572182738e-06,"output_power":3.5611692933100443e-06,"reference_active":true,"delta_db":-1.5222782112888882}],"active_band_count":3,"active_band_mae_db":0.8214370058264509}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010031642073947205,"log_rms_mae_db":1.1751353095463595,"rms_envelope_nrmse":0.13080603993369316,"envelope_correlation":0.9863548822384771,"rms_gain_error_db":-0.1719435355827151,"reference_mean_power":0.0004276640413209328,"output_mean_power":0.0004110629948841305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.824716997106709e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011139646210262632,"output_power":0.00010408790154448506,"reference_active":true,"delta_db":-0.29471145008991495},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.037822413553872e-05,"output_power":9.20544209702188e-06,"reference_active":true,"delta_db":-0.5207839519326286},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.220864105203889e-07,"output_power":2.0982384166815282e-07,"reference_active":true,"delta_db":-1.8611756716981165}],"active_band_count":3,"active_band_mae_db":0.89222369124022}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":2.1893936748705376,"rms_envelope_nrmse":0.382654596602772,"envelope_correlation":0.9038623468605177,"rms_gain_error_db":-0.538580097289596,"reference_mean_power":6.36168552564465e-05,"output_mean_power":5.6197136556506606e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6976388849888463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.191278258055639e-06,"output_power":5.124778775934004e-06,"reference_active":true,"delta_db":-0.8210520047925095},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.016318108845026e-06,"output_power":1.6853837107966563e-06,"reference_active":true,"delta_db":-0.7786025844248683},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.069070181696061e-07,"output_power":6.444111155907801e-07,"reference_active":true,"delta_db":0.2604086342690419}],"active_band_count":3,"active_band_mae_db":0.6200210744954732}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":1.5152552602137441,"rms_envelope_nrmse":0.33837742364675544,"envelope_correlation":0.9216390463914916,"rms_gain_error_db":-0.5387632113008928,"reference_mean_power":6.392327613647515e-05,"output_mean_power":5.646543832645316e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.609918878818377e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.077093160223487e-06,"output_power":5.10734018314597e-06,"reference_active":true,"delta_db":-0.7550110827014672},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.511523340291827e-07,"output_power":8.821213061353205e-07,"reference_active":true,"delta_db":-0.3272176611070814},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7938163224025673e-08,"output_power":4.4785402746367465e-08,"reference_active":true,"delta_db":2.0493863377876855}],"active_band_count":3,"active_band_mae_db":1.0438716938654113}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0019","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.3203508853912354,"predicted_mos":3.3203508853912354,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"956ad9e0805fbcb9bea14564a1a7f5f109849efdddc4b25b1d2c5ee3e104d4e1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.450644189999366},"started_at":"2026-10-13T04:41:10.149115+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6326000806875527e-05,"native_input_samples":96000,"native_input_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","output_sha256":"64a48de82ec2efe67d9567c9fbd9ffe588e53b07765eaad8d2769652d6e00657","output_samples":96000,"length_delta_samples":0,"analysis_s":2.4378863590000037,"synthesis_s":0.07565092900040327,"total_s":2.513537288000407,"rtf":0.41892288133340116,"output_peak":0.8589965646002675,"output_rms":0.09701730878942981,"finished_at":"2026-10-13T04:41:12.670264+00:00","evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"64a48de82ec2efe67d9567c9fbd9ffe588e53b07765eaad8d2769652d6e00657","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.126695397239665,"spectral_convergence":0.6772961608885465,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.258319739538572,"spectral_convergence":0.24833245616997657,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4057926928957265,"spectral_convergence":0.24823199239398447,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.0046435201879128,"delta_db":0.5010569316859715,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.00315150735488688,"delta_db":0.8125859589638098,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.0009445387536237634,"delta_db":0.682497405002209,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.0005453036306304791,"delta_db":1.079932670065469,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":7.355096597191831e-05,"delta_db":0.19855610617195268,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":5.393731171773128e-05,"delta_db":0.27149083057411827,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5910199837439217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.649657068696885,"pitch":null,"output_rms_dbfs":-20.26301553299745,"output_peak":0.8589965646002675,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.930269276557987},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.126695397239665,"spectral_convergence":0.6772961608885465,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.258319739538572,"spectral_convergence":0.24833245616997657,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4057926928957265,"spectral_convergence":0.24823199239398447,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.0046435201879128,"delta_db":0.5010569316859715,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.00315150735488688,"delta_db":0.8125859589638098,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.0009445387536237634,"delta_db":0.682497405002209,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.0005453036306304791,"delta_db":1.079932670065469,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":7.355096597191831e-05,"delta_db":0.19855610617195268,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":5.393731171773128e-05,"delta_db":0.27149083057411827,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5910199837439217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.649657068696885,"pitch":null,"output_rms_dbfs":-20.26301553299745,"output_peak":0.8589965646002675,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.930269276557987},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9358777958420582,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900256715962862,"rms_dbfs":-20.91267260169434,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.018647707074955,"mean":-0.0005092315673828125,"zero_energy":false},"output":{"rms":0.09701730878942981,"rms_dbfs":-20.263015532997454,"peak":0.8589965646002675,"peak_dbfs":-1.3201714609334867,"crest_factor_db":18.942844072063963,"mean":4.7423596875046755e-06,"zero_energy":false},"rms_difference_db":0.6496570686968859},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":3.4421720494866547e-06,"output_rms_dbfs":-54.63167426200418,"output_zero_energy":false,"output_peak":0.016147321331439108,"output_energy_fraction":3.535174617019459e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":292,"both_estimated_voiced_frames":254,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":38,"paired_voiced_fraction":0.9236363636363636,"absolute_difference_cents_median":5.855943968362507,"absolute_difference_cents_p95":25.262943927938817,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025953969251226393,"log_rms_mae_db":2.3617468829831862,"rms_envelope_nrmse":0.16074281008283706,"envelope_correlation":0.978333271058627,"rms_gain_error_db":0.6506350794599346,"reference_mean_power":0.008109529837298932,"output_mean_power":0.009420179623655702},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5635786081597607e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013500334014131179,"output_power":0.0015512383838692978,"reference_active":true,"delta_db":0.6033402880988441},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020457869490256438,"output_power":0.00018813617472485905,"reference_active":true,"delta_db":-0.36388094231432794},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.388028407734598e-05,"output_power":2.9532160651483976e-05,"reference_active":true,"delta_db":-0.5965182085209588}],"active_band_count":3,"active_band_mae_db":0.5212464796447103}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024049824298809385,"log_rms_mae_db":1.7779521793844635,"rms_envelope_nrmse":0.1095748351724612,"envelope_correlation":0.9928347420956073,"rms_gain_error_db":0.6521896233198476,"reference_mean_power":0.008139876536629037,"output_mean_power":0.009458816058520435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2818780696338306e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012006574574251973,"output_power":0.0014071116742120215,"reference_active":true,"delta_db":0.6890944354583627},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.159020687440259e-05,"output_power":8.649881526670625e-05,"reference_active":true,"delta_db":-0.24838880778714015},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.235316344629429e-06,"output_power":1.831850066949047e-06,"reference_active":true,"delta_db":-0.8644906898713528}],"active_band_count":3,"active_band_mae_db":0.6006579777056186}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002074995834794702,"log_rms_mae_db":2.0915044092781425,"rms_envelope_nrmse":0.21114355374784394,"envelope_correlation":0.9803693023027742,"rms_gain_error_db":1.025611696877767,"reference_mean_power":0.0020359851811281022,"output_mean_power":0.0025783138549330922},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.345568889858965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045532158312677236,"output_power":0.0005746996220844233,"reference_active":true,"delta_db":1.011226743494646},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.561332252218219e-05,"output_power":0.00011078375995347768,"reference_active":true,"delta_db":1.119347491473596},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0669601144719856e-05,"output_power":8.667768416981961e-06,"reference_active":true,"delta_db":-0.902408853145962}],"active_band_count":3,"active_band_mae_db":1.0109943627047346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001833599147850085,"log_rms_mae_db":1.5845730069742183,"rms_envelope_nrmse":0.1698753447181276,"envelope_correlation":0.9904193765398986,"rms_gain_error_db":1.026217937362889,"reference_mean_power":0.0020457391187242045,"output_mean_power":0.002591027626265519},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.75393473065512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00043556686984963635,"output_power":0.0005494862384369714,"reference_active":true,"delta_db":1.0090198127301568},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6912146426205404e-05,"output_power":5.030149545748766e-05,"reference_active":true,"delta_db":1.3441159679494898},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.183033529622032e-07,"output_power":5.401320853566714e-07,"reference_active":true,"delta_db":-1.238079173231459}],"active_band_count":3,"active_band_mae_db":1.1970716513037019}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001109356585659573,"log_rms_mae_db":1.8462574248043393,"rms_envelope_nrmse":0.22064231401028078,"envelope_correlation":0.9783028166024332,"rms_gain_error_db":1.0052236374938377,"reference_mean_power":0.000425572053658366,"output_mean_power":0.0005364082699298264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9996440016989412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011758808293094378,"output_power":0.0001368324771944104,"reference_active":true,"delta_db":0.6582587942420663},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.147530754834166e-05,"output_power":2.3285466987167863e-05,"reference_active":true,"delta_db":0.3514555998990717},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.05615572182738e-06,"output_power":4.919274148026519e-06,"reference_active":true,"delta_db":-0.11919414893410907}],"active_band_count":3,"active_band_mae_db":0.3763028476917491}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010031642073947205,"log_rms_mae_db":1.0966499017195033,"rms_envelope_nrmse":0.1737311116679094,"envelope_correlation":0.9898170516008904,"rms_gain_error_db":1.0053148779503205,"reference_mean_power":0.0004276640413209328,"output_mean_power":0.0005390564208517711},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.824716997106709e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011139646210262632,"output_power":0.00013061014629698668,"reference_active":true,"delta_db":0.6910551783466723},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.037822413553872e-05,"output_power":1.1068785433109175e-05,"reference_active":true,"delta_db":0.2797692299470099},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.220864105203889e-07,"output_power":3.1540348660645195e-07,"reference_active":true,"delta_db":-0.09105911541957445}],"active_band_count":3,"active_band_mae_db":0.35396117457108556}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":2.0389142688854855,"rms_envelope_nrmse":0.2828780152756619,"envelope_correlation":0.9514353824129762,"rms_gain_error_db":0.22894460991696922,"reference_mean_power":6.36168552564465e-05,"output_mean_power":6.706048005364703e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6976388849888463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.191278258055639e-06,"output_power":6.362468623146569e-06,"reference_active":true,"delta_db":0.11845330414167098},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.016318108845026e-06,"output_power":2.0748880236974953e-06,"reference_active":true,"delta_db":0.12435613338863433},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.069070181696061e-07,"output_power":6.401754666466882e-07,"reference_active":true,"delta_db":0.231768670943074}],"active_band_count":3,"active_band_mae_db":0.15819270282445977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":1.2051643045524953,"rms_envelope_nrmse":0.14663466879585169,"envelope_correlation":0.9866662564232461,"rms_gain_error_db":0.2291956244923925,"reference_mean_power":6.392327613647515e-05,"output_mean_power":6.73873824695583e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.609918878818377e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.077093160223487e-06,"output_power":6.291881511178632e-06,"reference_active":true,"delta_db":0.15084641369704932},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.511523340291827e-07,"output_power":9.944385176699982e-07,"reference_active":true,"delta_db":0.1932785956617346},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7938163224025673e-08,"output_power":3.01161156030216e-08,"reference_active":true,"delta_db":0.32601105229926597}],"active_band_count":3,"active_band_mae_db":0.22337868721934995}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0019","method":"facodec","status":"measured","mos":{"status":"measured","value":3.534456729888916,"predicted_mos":3.534456729888916,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"181be0d5ab1948ffd06f4ce6b990abbbbe4bd8ceb186cdeef5af560634e13d11","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2632538299985754},"started_at":"2026-10-13T04:46:36.109097+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3111999325919896e-05,"native_input_samples":96000,"native_input_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","output_sha256":"a4ea5ccbf1664a58750ceb15420295ad572718f8746b3c2f1eadd2988f78d950","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6409719670000413,"synthesis_s":6.814104546000635,"total_s":8.455076513000677,"rtf":1.409179418833446,"output_peak":0.5573245882987976,"output_rms":0.08661733886837003,"finished_at":"2026-10-13T04:46:44.582297+00:00","evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"a4ea5ccbf1664a58750ceb15420295ad572718f8746b3c2f1eadd2988f78d950","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.310195691566711,"spectral_convergence":0.19059234027904912,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.4508533426123345,"spectral_convergence":0.18463327444466515,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.469992654646767,"spectral_convergence":0.20052558587848207,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.0039019032308659374,"delta_db":-0.25465150360694055,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.0024393391789667516,"delta_db":-0.29987537933331665,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.000703053986441221,"delta_db":-0.5998136574747095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.00035520924974268063,"delta_db":-0.7816085582486334,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":7.622326299833271e-05,"delta_db":0.3535476606652343,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":2.6834483623122863e-05,"delta_db":-2.760469714262041,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8416610789318125,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3352306616265519,"pitch":null,"output_rms_dbfs":-21.24790326332089,"output_peak":0.5573245882987976,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.410347229608604},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.310195691566711,"spectral_convergence":0.19059234027904912,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.4508533426123345,"spectral_convergence":0.18463327444466515,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.469992654646767,"spectral_convergence":0.20052558587848207,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.0039019032308659374,"delta_db":-0.25465150360694055,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.0024393391789667516,"delta_db":-0.29987537933331665,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.000703053986441221,"delta_db":-0.5998136574747095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.00035520924974268063,"delta_db":-0.7816085582486334,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":7.622326299833271e-05,"delta_db":0.3535476606652343,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":2.6834483623122863e-05,"delta_db":-2.760469714262041,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8416610789318125,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3352306616265519,"pitch":null,"output_rms_dbfs":-21.24790326332089,"output_peak":0.5573245882987976,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.410347229608604},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9140494503249889,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900256715962862,"rms_dbfs":-20.91267260169434,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.018647707074955,"mean":-0.0005092315673828125,"zero_energy":false},"output":{"rms":0.08661733886837003,"rms_dbfs":-21.24790326332089,"peak":0.5573245882987976,"peak_dbfs":-5.077835922406486,"crest_factor_db":16.170067340914404,"mean":-5.348878975555257e-05,"zero_energy":false},"rms_difference_db":-0.33523066162655013},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":3.2723996280712953e-06,"output_rms_dbfs":-54.8513366537103,"output_zero_energy":false,"output_peak":0.009819533675909042,"output_energy_fraction":4.216318443337584e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":275,"both_estimated_voiced_frames":264,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.96,"absolute_difference_cents_median":4.501743740835117,"absolute_difference_cents_p95":16.058290214227586,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025953969251226393,"log_rms_mae_db":1.210351052652024,"rms_envelope_nrmse":0.10136028395597942,"envelope_correlation":0.9891883718214836,"rms_gain_error_db":-0.33555240767322725,"reference_mean_power":0.008109529837298932,"output_mean_power":0.007506550925679021},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5635786081597607e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013500334014131179,"output_power":0.0013553799372134192,"reference_active":true,"delta_db":0.017165391957308724},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020457869490256438,"output_power":0.00021732461132805127,"reference_active":true,"delta_db":0.26248507862496323},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.388028407734598e-05,"output_power":3.712578373739202e-05,"reference_active":true,"delta_db":0.39728587400165444}],"active_band_count":3,"active_band_mae_db":0.22564544819464213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024049824298809385,"log_rms_mae_db":0.9466543886767707,"rms_envelope_nrmse":0.0950088636154526,"envelope_correlation":0.9896484241374469,"rms_gain_error_db":-0.3367333339957443,"reference_mean_power":0.008139876536629037,"output_mean_power":0.007532592687000009},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2818780696338306e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012006574574251973,"output_power":0.0012141617887518933,"reference_active":true,"delta_db":0.048574382746588096},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.159020687440259e-05,"output_power":9.506940145738931e-05,"reference_active":true,"delta_db":0.16191719707531235},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.235316344629429e-06,"output_power":2.472134976812702e-06,"reference_active":true,"delta_db":0.4373318554478542}],"active_band_count":3,"active_band_mae_db":0.2159411450899182}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002074995834794702,"log_rms_mae_db":1.5448682052882152,"rms_envelope_nrmse":0.13695883748053608,"envelope_correlation":0.9890939543164855,"rms_gain_error_db":-0.6412299909424144,"reference_mean_power":0.0020359851811281022,"output_mean_power":0.0017565139917748627},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.345568889858965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045532158312677236,"output_power":0.0004358994984236249,"reference_active":true,"delta_db":-0.18931867909045508},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.561332252218219e-05,"output_power":8.064827291923611e-05,"reference_active":true,"delta_db":-0.2594628026358754},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0669601144719856e-05,"output_power":1.1218533496186349e-05,"reference_active":true,"delta_db":0.21787904234628264}],"active_band_count":3,"active_band_mae_db":0.22222017469087105}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001833599147850085,"log_rms_mae_db":1.195625468155326,"rms_envelope_nrmse":0.12748133714571336,"envelope_correlation":0.9904813298911527,"rms_gain_error_db":-0.6413434613108443,"reference_mean_power":0.0020457391187242045,"output_mean_power":0.0017648829344489034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.75393473065512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00043556686984963635,"output_power":0.0004160244528433909,"reference_active":true,"delta_db":-0.19935980884655802},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6912146426205404e-05,"output_power":3.4154805762361136e-05,"reference_active":true,"delta_db":-0.33717480121737464},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.183033529622032e-07,"output_power":7.780022232451297e-07,"reference_active":true,"delta_db":0.3467294436827029}],"active_band_count":3,"active_band_mae_db":0.2944213512488785}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001109356585659573,"log_rms_mae_db":1.4851918807106435,"rms_envelope_nrmse":0.16201153246493477,"envelope_correlation":0.9826502910496747,"rms_gain_error_db":-0.5926082187462962,"reference_mean_power":0.000425572053658366,"output_mean_power":0.00037128916794265296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9996440016989412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011758808293094378,"output_power":0.000105444156354575,"reference_active":true,"delta_db":-0.47340793579609075},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.147530754834166e-05,"output_power":2.3221268318225277e-05,"reference_active":true,"delta_db":0.3394654461315113},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.05615572182738e-06,"output_power":6.968736258469915e-06,"reference_active":true,"delta_db":1.3933358726829734}],"active_band_count":3,"active_band_mae_db":0.7354030848701919}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010031642073947205,"log_rms_mae_db":1.2031510717226832,"rms_envelope_nrmse":0.14019001021198857,"envelope_correlation":0.9869759196521974,"rms_gain_error_db":-0.5927931484697385,"reference_mean_power":0.0004276640413209328,"output_mean_power":0.00037309842934397653},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.824716997106709e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011139646210262632,"output_power":0.00010025340025390064,"reference_active":true,"delta_db":-0.45772286743785745},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.037822413553872e-05,"output_power":1.1700435536979948e-05,"reference_active":true,"delta_db":0.5207898242972411},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.220864105203889e-07,"output_power":4.160736629751897e-07,"reference_active":true,"delta_db":1.1119782483922838}],"active_band_count":3,"active_band_mae_db":0.696830313375794}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":1.8072115231971062,"rms_envelope_nrmse":0.3224475987956746,"envelope_correlation":0.9463532479236059,"rms_gain_error_db":-1.653711599222584,"reference_mean_power":6.36168552564465e-05,"output_mean_power":4.3471140769538615e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6976388849888463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.191278258055639e-06,"output_power":6.1668363703236475e-06,"reference_active":true,"delta_db":-0.01717898098438398},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.016318108845026e-06,"output_power":2.2425274513172633e-06,"reference_active":true,"delta_db":0.4617871745214198},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.069070181696061e-07,"output_power":8.975131498494035e-07,"reference_active":true,"delta_db":1.6991866070427708}],"active_band_count":3,"active_band_mae_db":0.7260509208495248}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":1.4641110531044295,"rms_envelope_nrmse":0.29640191524389964,"envelope_correlation":0.9581917526793162,"rms_gain_error_db":-1.6548052073600703,"reference_mean_power":6.392327613647515e-05,"output_mean_power":4.3669528633762874e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.609918878818377e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.077093160223487e-06,"output_power":6.086356197909357e-06,"reference_active":true,"delta_db":0.006614713987594198},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.511523340291827e-07,"output_power":1.0950158420324458e-06,"reference_active":true,"delta_db":0.6117032437610768},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7938163224025673e-08,"output_power":4.454076027896208e-08,"reference_active":true,"delta_db":2.025597755432665}],"active_band_count":3,"active_band_mae_db":0.881305237727112}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0019","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.9777190685272217,"predicted_mos":2.9777190685272217,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"6c4722dcb9d80a853e1d46ea8baab3e88b25a416ea3355f4db5b61eb7327fb8a","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2860974300001544},"started_at":"2026-10-13T05:00:20.978228+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008428253000602126,"native_input_samples":264600,"native_input_sha256":"3f10cc0a1933975f87e175a6ccac73abfe1208f7726774103a5b2917992dde35","output_sha256":"cd909c8ac1c8862673b5d938867fa686b05f7dd436161d3be5423ae4e3a73fcc","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1478203400001803,"synthesis_s":4.335747805999745,"total_s":4.483568145999925,"rtf":0.7472613576666541,"output_peak":0.6016417741775513,"output_rms":0.08174486410392771,"finished_at":"2026-10-13T05:00:25.484659+00:00","evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"cd909c8ac1c8862673b5d938867fa686b05f7dd436161d3be5423ae4e3a73fcc","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.746892610420932,"output_peak":0.6034307622833821},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.48374838175169,"spectral_convergence":0.48944960620879663,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.201352690684852,"spectral_convergence":0.349917260602263,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.827876055695135,"spectral_convergence":0.39045990729764096,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.002410883689038993,"delta_db":-2.345653861064209,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.0028840722541255102,"delta_db":0.42746407373544565,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.0009809537229363376,"delta_db":0.8467847891053009,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.0003278698388102681,"delta_db":-1.129436550513781,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":6.473734731671606e-05,"delta_db":-0.35577870491207414,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":2.1981385088252314e-05,"delta_db":-3.626851593276803,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.455328262101269,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8327427921372321,"pitch":null,"output_rms_dbfs":-21.745415393831568,"output_peak":0.6034307622833821,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.170992376043893},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7672430168297943,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900256715962862,"rms_dbfs":-20.91267260169434,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.018647707074955,"mean":-0.0005092315673828125,"zero_energy":false},"output":{"rms":0.08179546587260249,"rms_dbfs":-21.745415393831568,"peak":0.6034307622833821,"peak_dbfs":-4.387451074159253,"crest_factor_db":17.357964319672313,"mean":-0.0002709317216357863,"zero_energy":false},"rms_difference_db":-0.8327427921372283},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":4.381947040263057e-06,"output_rms_dbfs":-53.58332875579628,"output_zero_energy":false,"output_peak":0.024401452788658816,"output_energy_fraction":6.33119087498661e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":232,"both_estimated_voiced_frames":214,"reference_voiced_output_unvoiced_frames":61,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.7781818181818182,"absolute_difference_cents_median":9.690976572926068,"absolute_difference_cents_p95":54.66093585552461,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025953969251226393,"log_rms_mae_db":1.9622076104316866,"rms_envelope_nrmse":0.13921729081567813,"envelope_correlation":0.9866016265753642,"rms_gain_error_db":-0.8330149297838886,"reference_mean_power":0.008109529837298932,"output_mean_power":0.006694130631307065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5635786081597607e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013500334014131179,"output_power":0.0012079778010655722,"reference_active":true,"delta_db":-0.4828556022864144},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020457869490256438,"output_power":0.0001610871879204517,"reference_active":true,"delta_db":-1.0379940357189954},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.388028407734598e-05,"output_power":3.121085380927414e-05,"reference_active":true,"delta_db":-0.35641393660451626}],"active_band_count":3,"active_band_mae_db":0.6257545248699753}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024049824298809385,"log_rms_mae_db":1.5131033539054137,"rms_envelope_nrmse":0.11568984777116462,"envelope_correlation":0.9935871706838464,"rms_gain_error_db":-0.8329384643293722,"reference_mean_power":0.008139876536629037,"output_mean_power":0.006719299064887495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2818780696338306e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012006574574251973,"output_power":0.0010896983212739604,"reference_active":true,"delta_db":-0.42112840817339076},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.159020687440259e-05,"output_power":7.719259143547838e-05,"reference_active":true,"delta_db":-0.7427341906686868},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.235316344629429e-06,"output_power":1.7169430739428342e-06,"reference_active":true,"delta_db":-1.1458309752809603}],"active_band_count":3,"active_band_mae_db":0.7698978580410127}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002074995834794702,"log_rms_mae_db":2.729824926778373,"rms_envelope_nrmse":0.22180553166428102,"envelope_correlation":0.9663557110229071,"rms_gain_error_db":0.47772432706083534,"reference_mean_power":0.0020359851811281022,"output_mean_power":0.00227272581721123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.345568889858965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045532158312677236,"output_power":0.0005465344638383399,"reference_active":true,"delta_db":0.7929931606516347},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.561332252218219e-05,"output_power":9.40931832384276e-05,"reference_active":true,"delta_db":0.41016809585559066},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0669601144719856e-05,"output_power":1.4552150339132277e-05,"reference_active":true,"delta_db":1.3477898804218156}],"active_band_count":3,"active_band_mae_db":0.8503170456430137}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001833599147850085,"log_rms_mae_db":2.0467039716154334,"rms_envelope_nrmse":0.1950976912048439,"envelope_correlation":0.9723837238398328,"rms_gain_error_db":0.47820418331953163,"reference_mean_power":0.0020457391187242045,"output_mean_power":0.002283866257541638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.75393473065512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00043556686984963635,"output_power":0.0005209939719696102,"reference_active":true,"delta_db":0.777778594627992},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6912146426205404e-05,"output_power":4.1942515413420615e-05,"reference_active":true,"delta_db":0.5548517282469332},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.183033529622032e-07,"output_power":8.018098773337644e-07,"reference_active":true,"delta_db":0.47763508255766307}],"active_band_count":3,"active_band_mae_db":0.6034218018108627}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001109356585659573,"log_rms_mae_db":3.4646234562993046,"rms_envelope_nrmse":0.2599566992078673,"envelope_correlation":0.9566883946075245,"rms_gain_error_db":-1.1132497303203994,"reference_mean_power":0.000425572053658366,"output_mean_power":0.000329342765579517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9996440016989412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011758808293094378,"output_power":8.673378838643738e-05,"reference_active":true,"delta_db":-1.3217499394965628},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.147530754834166e-05,"output_power":1.4908334978507147e-05,"reference_active":true,"delta_db":-1.5851024964315457},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.05615572182738e-06,"output_power":3.076478097996709e-06,"reference_active":true,"delta_db":-2.1576661346886126}],"active_band_count":3,"active_band_mae_db":1.6881728568722405}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010031642073947205,"log_rms_mae_db":2.849740169742184,"rms_envelope_nrmse":0.22388396546325695,"envelope_correlation":0.9687682233453222,"rms_gain_error_db":-1.1130381492625736,"reference_mean_power":0.0004276640413209328,"output_mean_power":0.00033097784252349643},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.824716997106709e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011139646210262632,"output_power":8.187956144047814e-05,"reference_active":true,"delta_db":-1.3369589022686976},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.037822413553872e-05,"output_power":7.270810904947968e-06,"reference_active":true,"delta_db":-1.545401958833152},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.220864105203889e-07,"output_power":1.7634501107108734e-07,"reference_active":true,"delta_db":-2.6160922377044993}],"active_band_count":3,"active_band_mae_db":1.8328176996021164}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":3.2375277083865903,"rms_envelope_nrmse":0.319956757417206,"envelope_correlation":0.9659375048689479,"rms_gain_error_db":-2.4772101921327825,"reference_mean_power":6.36168552564465e-05,"output_mean_power":3.596260788947939e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6976388849888463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.191278258055639e-06,"output_power":4.080566030470237e-06,"reference_active":true,"delta_db":-1.8105991332036633},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.016318108845026e-06,"output_power":1.1938131134341609e-06,"reference_active":true,"delta_db":-2.2762270551219626},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.069070181696061e-07,"output_power":3.2144704562969974e-07,"reference_active":true,"delta_db":-2.7601272104493764}],"active_band_count":3,"active_band_mae_db":2.282317799591667}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":2.787503281698363,"rms_envelope_nrmse":0.2806374490993779,"envelope_correlation":0.9851918694450343,"rms_gain_error_db":-2.47670033990742,"reference_mean_power":6.392327613647515e-05,"output_mean_power":3.614007012726781e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.609918878818377e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.077093160223487e-06,"output_power":3.9404622289724535e-06,"reference_active":true,"delta_db":-1.8814872505531233},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.511523340291827e-07,"output_power":5.716482609963982e-07,"reference_active":true,"delta_db":-2.2112119128919985},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7938163224025673e-08,"output_power":1.9443428803094723e-08,"reference_active":true,"delta_db":-1.5742499615669203}],"active_band_count":3,"active_band_mae_db":1.8889830416706808}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0019","method":"golf","status":"measured","mos":{"status":"measured","value":2.742090940475464,"predicted_mos":2.742090940475464,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"c6e5ef1fc935cbac26e50892d196acaecd0916d02e5816447b55899e6e7b400e","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.337738045998776},"started_at":"2026-10-13T05:01:59.478369+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032282020001730416,"native_input_samples":144000,"native_input_sha256":"d1b998dc37c6a02704cad5a3295ac96a901a4fde0683ee5caa90cf884f2c8d99","output_sha256":"f42b4803945c2d541f61ebf9fe3d6f518cb7a39e78652dd96e05942d4c35adbe","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6305563559999428,"synthesis_s":0.12015530200005742,"total_s":0.7507116580000002,"rtf":0.12511860966666669,"output_peak":0.5846099853515625,"output_rms":0.08120216607671031,"finished_at":"2026-10-13T05:02:00.240368+00:00","analysis_stages":{"external_f0_s":0.09430823799993959},"evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"f42b4803945c2d541f61ebf9fe3d6f518cb7a39e78652dd96e05942d4c35adbe","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.805917503711232,"output_peak":0.5849987106553859},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.93489967074019,"spectral_convergence":0.6383984218200952,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.075261163448779,"spectral_convergence":0.24698738683317328,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.245543313110633,"spectral_convergence":0.22982954034451777,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0041411696979745385,"output_power":0.0033454851935091033,"delta_db":-0.9266391539037626,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026155886576241837,"output_power":0.002273126050999131,"delta_db":-0.6094592619886934,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008085530056450193,"output_power":0.0005706234505850628,"delta_db":-1.5135888022282942,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004259699665993953,"output_power":0.00031169505578891176,"delta_db":-1.3564906644664423,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.035394124891427e-05,"output_power":4.9461861734845854e-05,"delta_db":-1.530179726668657,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.0755003572041864e-05,"output_power":4.754668228742889e-05,"delta_db":-0.2835864454742504,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0366573424550165,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8974058599788126,"pitch":null,"output_rms_dbfs":-21.805917503711232,"output_peak":0.5849987106553859,"reference_rms_dbfs":-20.908511643732417,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.751901382433201},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9056287760953351,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900688085447126,"rms_dbfs":-20.908511643732417,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.01448674911304,"mean":-0.000502340542852182,"zero_energy":false},"output":{"rms":0.08122769413755683,"rms_dbfs":-21.805917503711232,"peak":0.5849987106553859,"peak_dbfs":-4.656901822146758,"crest_factor_db":17.149015681564475,"mean":0.0005666211346521928,"zero_energy":false},"rms_difference_db":-0.8974058599788144},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":3.420692117022431e-06,"output_rms_dbfs":-54.65886013185732,"output_zero_energy":false,"output_peak":0.00973246909165837,"output_energy_fraction":5.0200368873052594e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":306,"both_estimated_voiced_frames":264,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":42,"paired_voiced_fraction":0.96,"absolute_difference_cents_median":4.16106622886184,"absolute_difference_cents_p95":15.909290184473209,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00025953969251226393,"log_rms_mae_db":1.9773594841352253,"rms_envelope_nrmse":0.15283237199287006,"envelope_correlation":0.9820395535676125,"rms_gain_error_db":-0.8977765469032207,"reference_mean_power":0.008117105271475649,"output_mean_power":0.006601209617247776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.5601406746492887e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001348726941862631,"output_power":0.0010841767920772084,"reference_active":true,"delta_db":-0.948239264362416},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020507524938250047,"output_power":0.00019409109836135557,"reference_active":true,"delta_db":-0.23907630637800242},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.398059683802693e-05,"output_power":3.0694421797696584e-05,"reference_active":true,"delta_db":-0.4417154575860724}],"active_band_count":3,"active_band_mae_db":0.543010342775497}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00024049824298809385,"log_rms_mae_db":1.3315540795592695,"rms_envelope_nrmse":0.11727280684292132,"envelope_correlation":0.9940121327869282,"rms_gain_error_db":-0.8994782080670193,"reference_mean_power":0.008145002833853759,"output_mean_power":0.006621302339639445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.278268398543304e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011996499044289954,"output_power":0.0009569834963043484,"reference_active":true,"delta_db":-0.9815007554990441},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.18060406487138e-05,"output_power":9.141904418316921e-05,"reference_active":true,"delta_db":-0.018345815430735525},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2337650913057897e-06,"output_power":2.507642844296858e-06,"reference_active":true,"delta_db":0.502281817476658}],"active_band_count":3,"active_band_mae_db":0.5007094628021459}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002074995834794702,"log_rms_mae_db":1.86957692897553,"rms_envelope_nrmse":0.18067606259344865,"envelope_correlation":0.9825028909819931,"rms_gain_error_db":-1.0292026320077927,"reference_mean_power":0.0020389976183601295,"output_mean_power":0.001608779247030691},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.328066119550333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004537079967301407,"output_power":0.0003384177444119162,"reference_active":true,"delta_db":-1.2732330737037811},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.570339733713959e-05,"output_power":7.333228969847708e-05,"reference_active":true,"delta_db":-0.6770279252686053},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0630701559258227e-05,"output_power":8.89252588538978e-06,"reference_active":true,"delta_db":-0.775367881117739}],"active_band_count":3,"active_band_mae_db":0.9085429600300419}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001833599147850085,"log_rms_mae_db":1.2747380373046815,"rms_envelope_nrmse":0.14221122113983076,"envelope_correlation":0.9928372472211682,"rms_gain_error_db":-1.029547291296762,"reference_mean_power":0.002048874504893202,"output_mean_power":0.0016164438722934943},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.738580038021443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004341414194829828,"output_power":0.0003235611375652671,"reference_active":true,"delta_db":-1.2767486827659582},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692334831572881e-05,"output_power":3.277887712934906e-05,"reference_active":true,"delta_db":-0.5170700488449865},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.149213351255052e-07,"output_power":5.692469873134755e-07,"reference_active":true,"delta_db":-0.989575169142911}],"active_band_count":3,"active_band_mae_db":0.9277979669179519}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001109356585659573,"log_rms_mae_db":2.101486016589523,"rms_envelope_nrmse":0.22283676228466662,"envelope_correlation":0.9762931940025887,"rms_gain_error_db":-1.3959286109558775,"reference_mean_power":0.00042626120454498706,"output_mean_power":0.00030908857105013465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.999121029351334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011753101808786183,"output_power":7.690272519473596e-05,"reference_active":true,"delta_db":-1.842107680263364},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1465232469331486e-05,"output_power":1.3416343414433141e-05,"reference_active":true,"delta_db":-2.041014300115555},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.060462938151611e-06,"output_power":3.1106944193716254e-06,"reference_active":true,"delta_db":-2.113328984497422}],"active_band_count":3,"active_band_mae_db":1.9988169882921138}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010031642073947205,"log_rms_mae_db":1.726175670495993,"rms_envelope_nrmse":0.17506570928626938,"envelope_correlation":0.992137485978769,"rms_gain_error_db":-1.3956280418689337,"reference_mean_power":0.00042827258025522245,"output_mean_power":0.000310568543854063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8240240252200601e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011134228673168037,"output_power":7.272750283568194e-05,"reference_active":true,"delta_db":-1.8496146060416263},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0362036098519702e-05,"output_power":6.876673840756815e-06,"reference_active":true,"delta_db":-1.7806667458631533},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2229436205745165e-07,"output_power":2.0214398585493575e-07,"reference_active":true,"delta_db":-2.0259188357727522}],"active_band_count":3,"active_band_mae_db":1.8854000625591774}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":2.4167143867284393,"rms_envelope_nrmse":0.23573590219938398,"envelope_correlation":0.9673081471248153,"rms_gain_error_db":-0.3931932172810716,"reference_mean_power":6.371661261706501e-05,"output_mean_power":5.820139008742058e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.702508431509748e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.197426570892186e-06,"output_power":4.90423708289656e-06,"reference_active":true,"delta_db":-1.0163993292683273},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0159509466559385e-06,"output_power":1.569354822382127e-06,"reference_active":true,"delta_db":-1.087588141329807},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.087936340119592e-07,"output_power":5.605938987363396e-07,"reference_active":true,"delta_db":-0.3582173571295312}],"active_band_count":3,"active_band_mae_db":0.8207349425758884}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":2.054907905701737,"rms_envelope_nrmse":0.1385633833889186,"envelope_correlation":0.990212039185305,"rms_gain_error_db":-0.39251764895290575,"reference_mean_power":6.401226694985459e-05,"output_mean_power":5.8480549242146605e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.6141973267031008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.085205548575498e-06,"output_power":4.868123245999611e-06,"reference_active":true,"delta_db":-0.9691368788552774},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.456520104986284e-07,"output_power":6.807613624698392e-07,"reference_active":true,"delta_db":-1.4273645141796767},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7810526588656804e-08,"output_power":2.445450685215975e-08,"reference_active":true,"delta_db":-0.5585030295673611}],"active_band_count":3,"active_band_mae_db":0.9850014742007717}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0019","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.299659013748169,"predicted_mos":3.299659013748169,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"a4d376fe5288edd27bf2095a56f871e9c8aaba95aaa7c3629a429a3a468383c7","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4492868630004523},"started_at":"2026-10-13T05:05:35.852687+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00328962200001115,"native_input_samples":144000,"native_input_sha256":"d1b998dc37c6a02704cad5a3295ac96a901a4fde0683ee5caa90cf884f2c8d99","output_sha256":"50eb8ce65c5a0dac3be27504fb6b11fefd73cfdf3f5d22051787349a3e694bad","output_samples":144120,"length_delta_samples":120,"analysis_s":2.4148409709996486,"synthesis_s":3.462260762000369,"total_s":5.877101733000018,"rtf":0.979516955500003,"output_peak":0.6229123473167419,"output_rms":0.09638524409949949,"finished_at":"2026-10-13T05:05:41.743028+00:00","analysis_stages":{"external_f0_s":1.5402874719993633},"evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"50eb8ce65c5a0dac3be27504fb6b11fefd73cfdf3f5d22051787349a3e694bad","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.31741407728546,"output_peak":0.6235415999778542},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.5403470987591055,"spectral_convergence":0.6164653272214381,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.1916752669965645,"spectral_convergence":0.27988456288553965,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.027411675591624,"spectral_convergence":0.27189693181491786,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.005599498213876254,"delta_db":1.3140746561673335,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.002507018551565154,"delta_db":-0.1810218142417217,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.000706005718299516,"delta_db":-0.581618222193252,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.00038244167055233606,"delta_db":-0.46079915644384395,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":7.020479787356142e-05,"delta_db":-0.0036597692371432553,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":3.6168971553586764e-05,"delta_db":-1.464040539930015,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6675356930355515,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5981268303487969,"pitch":null,"output_rms_dbfs":-20.31454577134554,"output_peak":0.6235415999778542,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.586478013782432},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8960792376371411,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900256715962862,"rms_dbfs":-20.91267260169434,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.018647707074955,"mean":-0.0005092315673828125,"zero_energy":false},"output":{"rms":0.09644344417180677,"rms_dbfs":-20.31454577134554,"peak":0.6235415999778542,"peak_dbfs":-4.102691339678228,"crest_factor_db":16.211854431667312,"mean":0.00019511791144626785,"zero_energy":false},"rms_difference_db":0.5981268303488001},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":7.202739398741308e-06,"output_rms_dbfs":-51.42502298084068,"output_zero_energy":false,"output_peak":0.03992302540967073,"output_energy_fraction":7.485641466367632e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":318,"both_estimated_voiced_frames":268,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":50,"paired_voiced_fraction":0.9745454545454545,"absolute_difference_cents_median":5.9089285415069615,"absolute_difference_cents_p95":29.896349651935353,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025953969251226393,"log_rms_mae_db":2.892271645776022,"rms_envelope_nrmse":0.1732342970891574,"envelope_correlation":0.972202869157808,"rms_gain_error_db":0.5988347526836212,"reference_mean_power":0.008109529837298932,"output_mean_power":0.009308488173377247},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5635786081597607e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013500334014131179,"output_power":0.0015940934470443635,"reference_active":true,"delta_db":0.7216926291548796},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020457869490256438,"output_power":0.00020247884636746274,"reference_active":true,"delta_db":-0.04480745981481188},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.388028407734598e-05,"output_power":2.852362088496496e-05,"reference_active":true,"delta_db":-0.7474238762925705}],"active_band_count":3,"active_band_mae_db":0.5046413217540874}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024049824298809385,"log_rms_mae_db":2.3470144719513537,"rms_envelope_nrmse":0.1308190634424601,"envelope_correlation":0.9850524749731617,"rms_gain_error_db":0.5990888564067605,"reference_mean_power":0.008139876536629037,"output_mean_power":0.009343868188274058},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2818780696338306e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012006574574251973,"output_power":0.0014181803461826299,"reference_active":true,"delta_db":0.7231233975479778},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.159020687440259e-05,"output_power":9.099143923978766e-05,"reference_active":true,"delta_db":-0.028485055034545983},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.235316344629429e-06,"output_power":1.8170107233896995e-06,"reference_active":true,"delta_db":-0.8998150331298798}],"active_band_count":3,"active_band_mae_db":0.5504744952374678}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002074995834794702,"log_rms_mae_db":2.360331480104465,"rms_envelope_nrmse":0.17348690726178226,"envelope_correlation":0.9770401352715568,"rms_gain_error_db":-0.3336175706916873,"reference_mean_power":0.0020359851811281022,"output_mean_power":0.0018854406039478227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.345568889858965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045532158312677236,"output_power":0.0004137985765468884,"reference_active":true,"delta_db":-0.41529244823533473},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.561332252218219e-05,"output_power":7.346039924559651e-05,"reference_active":true,"delta_db":-0.6648806734540859},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0669601144719856e-05,"output_power":8.578368041243657e-06,"reference_active":true,"delta_db":-0.9474350974148174}],"active_band_count":3,"active_band_mae_db":0.6758694063680792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001833599147850085,"log_rms_mae_db":1.6276262951493008,"rms_envelope_nrmse":0.11649049858859911,"envelope_correlation":0.989464607651102,"rms_gain_error_db":-0.3333293051061134,"reference_mean_power":0.0020457391187242045,"output_mean_power":0.0018945990680196484},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.75393473065512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00043556686984963635,"output_power":0.00039288317965788835,"reference_active":true,"delta_db":-0.44791403027107884},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6912146426205404e-05,"output_power":3.2443762837867595e-05,"reference_active":true,"delta_db":-0.5603808190474107},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.183033529622032e-07,"output_power":5.747048503916095e-07,"reference_active":true,"delta_db":-0.9686303117167415}],"active_band_count":3,"active_band_mae_db":0.6589750536784104}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001109356585659573,"log_rms_mae_db":2.5464394474854184,"rms_envelope_nrmse":0.20230635274513425,"envelope_correlation":0.9697683478997089,"rms_gain_error_db":-0.39356154746196537,"reference_mean_power":0.000425572053658366,"output_mean_power":0.0003887021531631509},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9996440016989412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011758808293094378,"output_power":0.00011368657740449211,"reference_active":true,"delta_db":-0.14654117997370528},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.147530754834166e-05,"output_power":1.947642744831638e-05,"reference_active":true,"delta_db":-0.42430094667653356},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.05615572182738e-06,"output_power":4.4484902618828734e-06,"reference_active":true,"delta_db":-0.5560779688465836}],"active_band_count":3,"active_band_mae_db":0.3756400318322741}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010031642073947205,"log_rms_mae_db":1.5687888144384858,"rms_envelope_nrmse":0.12921678972803347,"envelope_correlation":0.98767238632898,"rms_gain_error_db":-0.393498263558233,"reference_mean_power":0.0004276640413209328,"output_mean_power":0.00039061859110447577},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.824716997106709e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011139646210262632,"output_power":0.00010741849914473285,"reference_active":true,"delta_db":-0.15792317972329925},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.037822413553872e-05,"output_power":9.36142694677811e-06,"reference_active":true,"delta_db":-0.44780993222669196},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.220864105203889e-07,"output_power":2.904047030149877e-07,"reference_active":true,"delta_db":-0.44968756089494916}],"active_band_count":3,"active_band_mae_db":0.35180689094831347}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":2.3979069504733097,"rms_envelope_nrmse":0.28264515935787515,"envelope_correlation":0.9539859168169884,"rms_gain_error_db":-1.162378457075162,"reference_mean_power":6.36168552564465e-05,"output_mean_power":4.86781821330275e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6976388849888463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.191278258055639e-06,"output_power":5.236799721419053e-06,"reference_active":true,"delta_db":-0.7271435832866708},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.016318108845026e-06,"output_power":1.5214152815902217e-06,"reference_active":true,"delta_db":-1.2231127646359004},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.069070181696061e-07,"output_power":4.499131425280648e-07,"reference_active":true,"delta_db":-1.299934799888495}],"active_band_count":3,"active_band_mae_db":1.0833970492703553}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":1.5279829108425103,"rms_envelope_nrmse":0.19640262311696774,"envelope_correlation":0.9832940376993792,"rms_gain_error_db":-1.1624199161195525,"reference_mean_power":6.392327613647515e-05,"output_mean_power":4.891218154358524e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.609918878818377e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.077093160223487e-06,"output_power":5.041791560896394e-06,"reference_active":true,"delta_db":-0.8111100702515496},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.511523340291827e-07,"output_power":7.107172018563982e-07,"reference_active":true,"delta_db":-1.2655325093373708},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7938163224025673e-08,"output_power":2.0003734208750588e-08,"reference_active":true,"delta_db":-1.4508677490687496}],"active_band_count":3,"active_band_mae_db":1.1758367762192234}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0019","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0176258087158203,"predicted_mos":3.0176258087158203,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"3ec03975e8fc59a73844fbb1446531144c7b319753e8a07ecb6fd0525521d99c","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4048495100014406},"started_at":"2026-10-13T05:10:05.587397+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003360013999554212,"native_input_samples":144000,"native_input_sha256":"d1b998dc37c6a02704cad5a3295ac96a901a4fde0683ee5caa90cf884f2c8d99","output_sha256":"d7d124c94b393e80631e8a0fb9078b38a118ce2ae978632dd3f1245e63016cb2","output_samples":144240,"length_delta_samples":240,"analysis_s":1.1012372820005112,"synthesis_s":0.1431562859997939,"total_s":1.244393568000305,"rtf":0.20739892800005086,"output_peak":0.6202001571655273,"output_rms":0.08912086793894008,"finished_at":"2026-10-13T05:10:06.842769+00:00","analysis_stages":{"external_f0_s":0.1650603609996324},"evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"d7d124c94b393e80631e8a0fb9078b38a118ce2ae978632dd3f1245e63016cb2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.99761789892956,"output_peak":0.616669562047463},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.794070066116046,"spectral_convergence":0.1723037112072907,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.342415752302958,"spectral_convergence":0.16161330415394282,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.518447158226536,"spectral_convergence":0.17148064765708235,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.004113819132001636,"delta_db":-0.024964512287715228,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.0025557543620476457,"delta_db":-0.09740618407301851,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.0007932164950747426,"delta_db":-0.07578304001016946,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.00039016254335466584,"delta_db":-0.37399548591181986,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":6.150105566960325e-05,"delta_db":-0.578501997008476,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":4.598500923972258e-05,"delta_db":-0.4212393590300949,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.26198176305354903,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07795743050456017,"pitch":null,"output_rms_dbfs":-20.990630032198894,"output_peak":0.616669562047463,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.551644325548513},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.951022608399785,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900256715962862,"rms_dbfs":-20.91267260169434,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.018647707074955,"mean":-0.0005092315673828125,"zero_energy":false},"output":{"rms":0.08922129004552679,"rms_dbfs":-20.990630032198897,"peak":0.616669562047463,"peak_dbfs":-4.198949744334565,"crest_factor_db":16.79168028786433,"mean":0.000250187677316984,"zero_energy":false},"rms_difference_db":-0.07795743050455783},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":3.1915366297365e-06,"output_rms_dbfs":-54.96000166743765,"output_zero_energy":false,"output_peak":0.012273076987285504,"output_energy_fraction":3.8756056436692464e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":269,"both_estimated_voiced_frames":250,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9090909090909091,"absolute_difference_cents_median":1.2805211888717845,"absolute_difference_cents_p95":5.5695014303000185,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025953969251226393,"log_rms_mae_db":0.9569530924262137,"rms_envelope_nrmse":0.07381953242937311,"envelope_correlation":0.993567973938086,"rms_gain_error_db":-0.07750963517102666,"reference_mean_power":0.008109529837298932,"output_mean_power":0.007966080898570033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5635786081597607e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013500334014131179,"output_power":0.0013484107453339926,"reference_active":true,"delta_db":-0.005223088771216957},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020457869490256438,"output_power":0.00019496363429132265,"reference_active":true,"delta_db":-0.20906791847823486},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.388028407734598e-05,"output_power":4.723702273114334e-05,"reference_active":true,"delta_db":1.4433547384479946}],"active_band_count":3,"active_band_mae_db":0.5525485818991488}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024049824298809385,"log_rms_mae_db":0.537001046075635,"rms_envelope_nrmse":0.04605121844125328,"envelope_correlation":0.9972639463048841,"rms_gain_error_db":-0.07662998882436386,"reference_mean_power":0.008139876536629037,"output_mean_power":0.007997510497311592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2818780696338306e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012006574574251973,"output_power":0.001217420891993617,"reference_active":true,"delta_db":0.06021627683162021},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.159020687440259e-05,"output_power":9.121792531897405e-05,"reference_active":true,"delta_db":-0.017688496326976114},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.235316344629429e-06,"output_power":2.816671035377372e-06,"reference_active":true,"delta_db":1.0039713414021496}],"active_band_count":3,"active_band_mae_db":0.3606253715202486}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002074995834794702,"log_rms_mae_db":1.2172296735601267,"rms_envelope_nrmse":0.0830544566247529,"envelope_correlation":0.9948311703414211,"rms_gain_error_db":-0.15243166937244496,"reference_mean_power":0.0020359851811281022,"output_mean_power":0.001965764309509574},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.345568889858965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045532158312677236,"output_power":0.00044127557816360204,"reference_active":true,"delta_db":-0.1360834456170368},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.561332252218219e-05,"output_power":7.210075420286481e-05,"reference_active":true,"delta_db":-0.7460154404458976},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0669601144719856e-05,"output_power":7.384011100914567e-06,"reference_active":true,"delta_db":-1.5985584383413987}],"active_band_count":3,"active_band_mae_db":0.8268857748014443}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001833599147850085,"log_rms_mae_db":0.6986655716420553,"rms_envelope_nrmse":0.05211332449595692,"envelope_correlation":0.997959877677927,"rms_gain_error_db":-0.15233853469227096,"reference_mean_power":0.0020457391187242045,"output_mean_power":0.001975224193363124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.75393473065512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00043556686984963635,"output_power":0.0004195418911469605,"reference_active":true,"delta_db":-0.16279507451097291},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6912146426205404e-05,"output_power":3.187788403044268e-05,"reference_active":true,"delta_db":-0.636798136136001},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.183033529622032e-07,"output_power":4.3839575383521023e-07,"reference_active":true,"delta_db":-2.1444155462206687}],"active_band_count":3,"active_band_mae_db":0.9813362522892142}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001109356585659573,"log_rms_mae_db":1.3066197808855646,"rms_envelope_nrmse":0.10673653934247639,"envelope_correlation":0.9926951286741624,"rms_gain_error_db":-0.3904048087648536,"reference_mean_power":0.000425572053658366,"output_mean_power":0.0003889847902288579},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9996440016989412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011758808293094378,"output_power":0.00011345254102742085,"reference_active":true,"delta_db":-0.15549082667458772},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.147530754834166e-05,"output_power":1.7598719630761736e-05,"reference_active":true,"delta_db":-0.864583195880424},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.05615572182738e-06,"output_power":3.864142232640599e-06,"reference_active":true,"delta_db":-1.1676733738873932}],"active_band_count":3,"active_band_mae_db":0.7292491321474683}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010031642073947205,"log_rms_mae_db":0.7666177412496824,"rms_envelope_nrmse":0.07068435017344672,"envelope_correlation":0.9974766735138442,"rms_gain_error_db":-0.3903868269777644,"reference_mean_power":0.0004276640413209328,"output_mean_power":0.0003908985441087253},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.824716997106709e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011139646210262632,"output_power":0.0001074931770540566,"reference_active":true,"delta_db":-0.15490499050079207},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.037822413553872e-05,"output_power":8.719946599330739e-06,"reference_active":true,"delta_db":-0.7560922045848468},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.220864105203889e-07,"output_power":2.6875246460243314e-07,"reference_active":true,"delta_db":-0.7861994578794609}],"active_band_count":3,"active_band_mae_db":0.5657322176550332}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":1.3885917519915456,"rms_envelope_nrmse":0.13780910445591846,"envelope_correlation":0.9890571616552881,"rms_gain_error_db":-0.435242629205902,"reference_mean_power":6.36168552564465e-05,"output_mean_power":5.755034534923986e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6976388849888463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.191278258055639e-06,"output_power":5.477116062603545e-06,"reference_active":true,"delta_db":-0.5322837926380288},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.016318108845026e-06,"output_power":1.7171869268607319e-06,"reference_active":true,"delta_db":-0.6974147711288285},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.069070181696061e-07,"output_power":4.1215560253481844e-07,"reference_active":true,"delta_db":-1.6806095193305706}],"active_band_count":3,"active_band_mae_db":0.9701026943658094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":0.9038856312352126,"rms_envelope_nrmse":0.06822666090853416,"envelope_correlation":0.9984470828921331,"rms_gain_error_db":-0.4350305452762338,"reference_mean_power":6.392327613647515e-05,"output_mean_power":5.7830369929226087e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.609918878818377e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.077093160223487e-06,"output_power":5.378909395767801e-06,"reference_active":true,"delta_db":-0.530016650821412},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.511523340291827e-07,"output_power":8.118118193077874e-07,"reference_active":true,"delta_db":-0.6879470796227649},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7938163224025673e-08,"output_power":2.0562423412183082e-08,"reference_active":true,"delta_db":-1.3312355261220332}],"active_band_count":3,"active_band_mae_db":0.8497330855220699}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0019","method":"magphase","status":"measured","mos":{"status":"measured","value":3.5549590587615967,"predicted_mos":3.5549590587615967,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288561,"input_sample_rate":48000,"input_seconds":6.0116875,"input_float32_sha256":"6971c41eb9b7e5c77164edff09b08b63b737bb347ed090d558c640892f16d54f","internal_samples":96187,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.423696983998525},"started_at":"2026-10-13T05:12:46.938389+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010950961999697029,"native_input_samples":288000,"native_input_sha256":"fabfc1ddb093c10a7feb66f862bb9ef2e875e5d4e654c3a0a42a7a1778c6773b","output_sha256":"4412d19811762adbf7466ecbe341ff580413f964da1d552a676fa57fa4736c5e","output_samples":288561,"length_delta_samples":561,"analysis_s":2.760193222000453,"synthesis_s":1.3072228649998578,"total_s":4.067416087000311,"rtf":0.6779026811667185,"output_peak":0.6003515321361742,"output_rms":0.08661763931019126,"finished_at":"2026-10-13T05:12:51.032134+00:00","evaluation_input":{"reference":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"4412d19811762adbf7466ecbe341ff580413f964da1d552a676fa57fa4736c5e","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96187,"length_delta_samples":187,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.244973708851447,"output_peak":0.5947327944196829},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980558703359081,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.113609417970848,"spectral_convergence":0.6793488111960777,"spectral_convergence_status":"ok","amplitude_floor":1.884454455950008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.227555033352003,"spectral_convergence":0.46788154909426655,"spectral_convergence_status":"ok","amplitude_floor":1.301578596614512e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.092465560021933,"spectral_convergence":0.4811711712474742,"spectral_convergence_status":"ok","amplitude_floor":1.1393421242566376e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004137534659655858,"output_power":0.003696710740404641,"delta_db":-0.48926175532775723,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026137240204796635,"output_power":0.0024824059652499763,"delta_db":-0.22386922917002006,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008071793462917449,"output_power":0.0007822088545773033,"delta_db":-0.13647313014824156,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004252507832785635,"output_power":0.0004163188682418291,"delta_db":-0.09219027963267694,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.026398390024539e-05,"output_power":7.42983904060832e-05,"delta_db":0.24246635111164944,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.066875275629503e-05,"output_power":6.0314433491967236e-05,"delta_db":0.7568103904662622,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.32351185597610127,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3296228818532304,"pitch":null,"output_rms_dbfs":-21.24229548354757,"output_peak":0.5947327944196829,"reference_rms_dbfs":-20.91267260169434,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.144543337114927},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6499099860122476,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0900256715962862,"rms_dbfs":-20.91267260169434,"peak":0.569244384765625,"peak_dbfs":-4.89402489461938,"crest_factor_db":16.018647707074955,"mean":-0.0005092315673828125,"zero_energy":false},"output":{"rms":0.08667327876786364,"rms_dbfs":-21.24229548354757,"peak":0.5947327944196829,"peak_dbfs":-4.513562264390638,"crest_factor_db":16.72873321915693,"mean":0.002469844678395359,"zero_energy":false},"rms_difference_db":-0.3296228818532292},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.736085198936052e-06,"threshold_dbfs":-51.71592428457072,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":3.1619597675985302e-06,"output_mean_power":9.907022409739436e-06,"output_rms_dbfs":-50.0405685462982,"output_zero_energy":false,"output_peak":0.044802939664416674,"output_energy_fraction":0.00012748216691316268,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":289,"both_estimated_voiced_frames":235,"reference_voiced_output_unvoiced_frames":40,"reference_unvoiced_output_voiced_frames":54,"paired_voiced_fraction":0.8545454545454545,"absolute_difference_cents_median":13.438163180872037,"absolute_difference_cents_p95":107.3650822582187,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025953969251226393,"log_rms_mae_db":3.9814991628484093,"rms_envelope_nrmse":0.3508667311388368,"envelope_correlation":0.8525243841598034,"rms_gain_error_db":-0.3306569382002382,"reference_mean_power":0.008109529837298932,"output_mean_power":0.007515017257137635},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5635786081597607e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013500334014131179,"output_power":0.0013154628439094645,"reference_active":true,"delta_db":-0.112659279146918},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020457869490256438,"output_power":0.00019343819121375852,"reference_active":true,"delta_db":-0.24318181156089483},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.388028407734598e-05,"output_power":3.1195343654655e-05,"reference_active":true,"delta_db":-0.35857268845192514}],"active_band_count":3,"active_band_mae_db":0.23813792638657935}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024049824298809385,"log_rms_mae_db":2.9607158706149854,"rms_envelope_nrmse":0.27875492908395555,"envelope_correlation":0.8964554067870671,"rms_gain_error_db":-0.3368781518157712,"reference_mean_power":0.008139876536629037,"output_mean_power":0.007532341512852008},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2818780696338306e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012006574574251973,"output_power":0.0011865512276713445,"reference_active":true,"delta_db":-0.051326296578124264},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.159020687440259e-05,"output_power":8.584521854376471e-05,"reference_active":true,"delta_db":-0.2813292937114489},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.235316344629429e-06,"output_power":2.3024161348165966e-06,"reference_active":true,"delta_db":0.12844826352259375}],"active_band_count":3,"active_band_mae_db":0.15370128460405563}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002074995834794702,"log_rms_mae_db":5.269047937286352,"rms_envelope_nrmse":0.4908269178100067,"envelope_correlation":0.8101366165797422,"rms_gain_error_db":-0.229798948657043,"reference_mean_power":0.0020359851811281022,"output_mean_power":0.0019310553305752207},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.345568889858965e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045532158312677236,"output_power":0.0004259486704466723,"reference_active":true,"delta_db":-0.28960970218917287},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.561332252218219e-05,"output_power":8.229862101275575e-05,"reference_active":true,"delta_db":-0.1714879340310932},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0669601144719856e-05,"output_power":1.0393272555108401e-05,"reference_active":true,"delta_db":-0.11395868303901516}],"active_band_count":3,"active_band_mae_db":0.19168543975309374}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001833599147850085,"log_rms_mae_db":4.152704818997171,"rms_envelope_nrmse":0.3899136427409067,"envelope_correlation":0.8704101614679337,"rms_gain_error_db":-0.23075458870578547,"reference_mean_power":0.0020457391187242045,"output_mean_power":0.0019398796670066021},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.75393473065512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00043556686984963635,"output_power":0.00040896561480622715,"reference_active":true,"delta_db":-0.2736804423343384},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6912146426205404e-05,"output_power":3.520981156251436e-05,"reference_active":true,"delta_db":-0.20505599177032816},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.183033529622032e-07,"output_power":7.351599534287352e-07,"reference_active":true,"delta_db":0.10073947901470272}],"active_band_count":3,"active_band_mae_db":0.19315863770645644}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001109356585659573,"log_rms_mae_db":5.592237431819532,"rms_envelope_nrmse":0.4461478004166726,"envelope_correlation":0.8526380943232412,"rms_gain_error_db":-0.04500265894340231,"reference_mean_power":0.000425572053658366,"output_mean_power":0.00042118494112757313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9996440016989412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011758808293094378,"output_power":0.00011778146411852577,"reference_active":true,"delta_db":0.007136386386523743},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.147530754834166e-05,"output_power":2.1959015566101946e-05,"reference_active":true,"delta_db":0.0967347448228975},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.05615572182738e-06,"output_power":4.650799635563425e-06,"reference_active":true,"delta_db":-0.3629281132074269}],"active_band_count":3,"active_band_mae_db":0.1555997481389494}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010031642073947205,"log_rms_mae_db":4.187741370532633,"rms_envelope_nrmse":0.34106180369292904,"envelope_correlation":0.9070535626094248,"rms_gain_error_db":-0.051879516362696304,"reference_mean_power":0.0004276640413209328,"output_mean_power":0.0004225856875746664},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.824716997106709e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011139646210262632,"output_power":0.00011190294999139592,"reference_active":true,"delta_db":0.019701374974721478},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.037822413553872e-05,"output_power":1.0576186287525997e-05,"reference_active":true,"delta_db":0.08206046019740856},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.220864105203889e-07,"output_power":3.369746368937756e-07,"reference_active":true,"delta_db":0.19624812571710473}],"active_band_count":3,"active_band_mae_db":0.09933665362974492}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.985234496629771e-05,"log_rms_mae_db":5.349521475409125,"rms_envelope_nrmse":0.3769826391625236,"envelope_correlation":0.9223986195983472,"rms_gain_error_db":0.6798241475536746,"reference_mean_power":6.36168552564465e-05,"output_mean_power":7.439686097787946e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6976388849888463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.191278258055639e-06,"output_power":6.92450124109376e-06,"reference_active":true,"delta_db":0.4860817425215592},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.016318108845026e-06,"output_power":2.334513559687857e-06,"reference_active":true,"delta_db":0.6363735021187278},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.069070181696061e-07,"output_power":8.017205722631778e-07,"reference_active":true,"delta_db":1.2090086817106753}],"active_band_count":3,"active_band_mae_db":0.7771546421169875}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.357134521214993e-05,"log_rms_mae_db":3.9107686832914186,"rms_envelope_nrmse":0.25939953768298735,"envelope_correlation":0.9636133226177526,"rms_gain_error_db":0.6733016314017537,"reference_mean_power":6.392327613647515e-05,"output_mean_power":7.464301755171739e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.609918878818377e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.077093160223487e-06,"output_power":6.982123826121672e-06,"reference_active":true,"delta_db":0.6029165262700842},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.511523340291827e-07,"output_power":1.0983129299249578e-06,"reference_active":true,"delta_db":0.6247601841974039},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7938163224025673e-08,"output_power":3.3386746578768347e-08,"reference_active":true,"delta_db":0.7737625032918731}],"active_band_count":3,"active_band_mae_db":0.6671464045864538}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"hnm","status":"measured","mos":{"status":"measured","value":3.063091278076172,"predicted_mos":3.063091278076172,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ba87e46e5e9a46737a4da4e3ea15dc8754956f15d51df837aabfe3b8a2b39524","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8769588340001064},"started_at":"2026-10-13T04:38:18.209019+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5310999515349977e-05,"native_input_samples":96000,"native_input_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","output_sha256":"99bbb31f564a036a732a82e1eda3b22d7b7e54a8b39c274402ab3ce0c830f602","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2899420789999567,"synthesis_s":0.5122401630005697,"total_s":1.8021822420005265,"rtf":0.30036370700008774,"output_peak":0.56214407382108,"output_rms":0.07638193183240395,"finished_at":"2026-10-13T04:38:20.017656+00:00","evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"99bbb31f564a036a732a82e1eda3b22d7b7e54a8b39c274402ab3ce0c830f602","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.7486298419175625,"spectral_convergence":0.1666910002435213,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2947298999365096,"spectral_convergence":0.16440238747502464,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4106905897156246,"spectral_convergence":0.1652887112991687,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0020761524550753203,"delta_db":0.054927004468896964,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.0024635612924920596,"delta_db":-0.12001273199287955,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.00095480656950067,"delta_db":-0.08880770747278724,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.0002677911127048142,"delta_db":-0.4217983680923251,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.6389063655274e-05,"delta_db":-0.19707361959757672,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":3.549901702186451e-05,"delta_db":0.03743221199056612,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1533419406025053,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06747970968911084,"pitch":null,"output_rms_dbfs":-22.3401872351631,"output_peak":0.56214407382108,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.484683443856565},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.7486298419175625,"spectral_convergence":0.1666910002435213,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2947298999365096,"spectral_convergence":0.16440238747502464,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4106905897156246,"spectral_convergence":0.1652887112991687,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0020761524550753203,"delta_db":0.054927004468896964,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.0024635612924920596,"delta_db":-0.12001273199287955,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.00095480656950067,"delta_db":-0.08880770747278724,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.0002677911127048142,"delta_db":-0.4217983680923251,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.6389063655274e-05,"delta_db":-0.19707361959757672,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":3.549901702186451e-05,"delta_db":0.03743221199056612,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1533419406025053,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06747970968911084,"pitch":null,"output_rms_dbfs":-22.3401872351631,"output_peak":0.56214407382108,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.484683443856565},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9205112494248392,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07697764557868021,"rms_dbfs":-22.27270752547399,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.699994329705078,"mean":-0.0004927848180135092,"zero_energy":false},"output":{"rms":0.07638193183240395,"rms_dbfs":-22.340187235163103,"peak":0.56214407382108,"peak_dbfs":-5.00304726703167,"crest_factor_db":17.33713996813143,"mean":-0.00025252500692441474,"zero_energy":false},"rms_difference_db":-0.06747970968911332},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":8.091255928850592e-06,"output_rms_dbfs":-50.91984061745538,"output_zero_energy":false,"output_peak":0.04582615599618868,"output_energy_fraction":0.00010170354425708406,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":233,"output_estimated_voiced_frames":263,"both_estimated_voiced_frames":230,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":33,"paired_voiced_fraction":0.9871244635193133,"absolute_difference_cents_median":1.751994289079747,"absolute_difference_cents_p95":7.229990950393786,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019946925088635918,"log_rms_mae_db":0.9017698426010031,"rms_envelope_nrmse":0.053888809210109676,"envelope_correlation":0.9965471913307172,"rms_gain_error_db":-0.0658434739745349,"reference_mean_power":0.0059224241180442505,"output_mean_power":0.005833311353699319},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.077279189796661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245127053592495,"output_power":0.0010174338084276413,"reference_active":true,"delta_db":-0.03011183858863014},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329567656774176,"output_power":0.00012208994936035804,"reference_active":true,"delta_db":-0.3703052826678219},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3018966563501677e-05,"output_power":2.0570307192313692e-05,"reference_active":true,"delta_db":-0.48845044682084787}],"active_band_count":3,"active_band_mae_db":0.2962891893591}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001930447992601615,"log_rms_mae_db":0.5026722026948546,"rms_envelope_nrmse":0.029642445024589403,"envelope_correlation":0.9989150410114204,"rms_gain_error_db":-0.06652562424635826,"reference_mean_power":0.005919423744261807,"output_mean_power":0.005829440418158562},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8759871534324143e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000947359244704089,"output_power":0.0009418918052277464,"reference_active":true,"delta_db":-0.02513679083196489},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.123049684235516e-05,"output_power":5.5921302091774104e-05,"reference_active":true,"delta_db":-0.3939050831909852},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3299097007749553e-06,"output_power":1.0513682454781881e-06,"reference_active":true,"delta_db":-1.020672980446016}],"active_band_count":3,"active_band_mae_db":0.4799049514896554}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015110036587973006,"log_rms_mae_db":1.585500923522214,"rms_envelope_nrmse":0.07399731709624606,"envelope_correlation":0.9954671478632525,"rms_gain_error_db":-0.15900657744792837,"reference_mean_power":0.0025365732398871734,"output_mean_power":0.0024453822116788705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.31579034521966e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006679887319597139,"output_power":0.0006437358198206122,"reference_active":true,"delta_db":-0.16061461055622706},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.085773760663629e-05,"output_power":7.562459542994417e-05,"reference_active":true,"delta_db":-0.7969885419511049},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6589878271689648e-05,"output_power":1.1369812093455036e-05,"reference_active":true,"delta_db":-1.6408991214673017}],"active_band_count":3,"active_band_mae_db":0.8661674246582112}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013775289901134618,"log_rms_mae_db":1.000923013173523,"rms_envelope_nrmse":0.046740959448722454,"envelope_correlation":0.9982548153405069,"rms_gain_error_db":-0.15998493637949726,"reference_mean_power":0.0025400163932975937,"output_mean_power":0.0024481500118001026},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2092894134690622e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006291541070562533,"output_power":0.000605455670699637,"reference_active":true,"delta_db":-0.16674684827368608},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9750836404697825e-05,"output_power":3.36669072246638e-05,"reference_active":true,"delta_db":-0.7214304894848498},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.784451638039305e-07,"output_power":6.826824125424745e-07,"reference_active":true,"delta_db":-1.5631777640209557}],"active_band_count":3,"active_band_mae_db":0.8171183672598304}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010600794217574422,"log_rms_mae_db":1.6857637480066472,"rms_envelope_nrmse":0.1579451815631446,"envelope_correlation":0.9830774104332503,"rms_gain_error_db":-0.4821003713748039,"reference_mean_power":0.00026914786021620774,"output_mean_power":0.00024086899150634866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.028210835890849e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001073533616295001,"output_power":9.481231257463558e-05,"reference_active":true,"delta_db":-0.5395090872984746},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.315168535027105e-06,"output_power":7.118666697437328e-06,"reference_active":true,"delta_db":-0.6747239674851945},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4089826752830493e-06,"output_power":3.832487127051462e-06,"reference_active":true,"delta_db":2.016470280888599}],"active_band_count":3,"active_band_mae_db":1.0769011118907559}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":1.0908427448223923,"rms_envelope_nrmse":0.09922540072791544,"envelope_correlation":0.9945459265923015,"rms_gain_error_db":-0.4942423503081861,"reference_mean_power":0.00026921330106685767,"output_mean_power":0.00024025491352177048},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9187018796507895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.751052870210858e-05,"output_power":8.717218527366986e-05,"reference_active":true,"delta_db":-0.48673577967276926},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.818200149529569e-06,"output_power":3.342895006554458e-06,"reference_active":true,"delta_db":-0.5773595379991852},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4850622077020123e-07,"output_power":2.0504149811461052e-07,"reference_active":true,"delta_db":1.400971201400325}],"active_band_count":3,"active_band_mae_db":0.8216888396907599}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":2.092699066507608,"rms_envelope_nrmse":0.17927116213424796,"envelope_correlation":0.9794477440964126,"rms_gain_error_db":0.033441748924660744,"reference_mean_power":4.950856594995313e-05,"output_mean_power":4.9891265699418285e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1644362822449518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0281496224399232e-05,"output_power":9.932401859646463e-06,"reference_active":true,"delta_db":-0.15002037811413263},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.187880808785332e-06,"output_power":3.678141252746537e-06,"reference_active":true,"delta_db":-0.5636590917200336},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.689088055617458e-07,"output_power":5.450041965021376e-07,"reference_active":true,"delta_db":-0.18642809490125617}],"active_band_count":3,"active_band_mae_db":0.30003585491180745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":1.3270952795548474,"rms_envelope_nrmse":0.10748649488461216,"envelope_correlation":0.9923255808889148,"rms_gain_error_db":0.033500717552938694,"reference_mean_power":4.9751584691373285e-05,"output_mean_power":5.013684372338533e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0334524009764525e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.035252748280077e-05,"output_power":1.0098013331096627e-05,"reference_active":true,"delta_db":-0.10810452392750466},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2970706912154717e-06,"output_power":1.9902902541142125e-06,"reference_active":true,"delta_db":-0.6225794420512544},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.669301652148395e-08,"output_power":3.749758483495255e-08,"reference_active":true,"delta_db":-0.952486353292647}],"active_band_count":3,"active_band_mae_db":0.5610567730904686}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.698338508605957,"predicted_mos":3.698338508605957,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"53053db10ef49aa21802ebc9387a7ec8fe031278d422379ac2aa60a3855037e5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4802958060008677},"started_at":"2026-10-13T04:41:12.671104+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.378899989707861e-05,"native_input_samples":96000,"native_input_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","output_sha256":"5316d4fe732a09c59a55da1661c3c5129797c16875ae70f9830c3d6a716c62f2","output_samples":96000,"length_delta_samples":0,"analysis_s":1.550763471999744,"synthesis_s":0.08877084000050672,"total_s":1.6395343120002508,"rtf":0.27325571866670845,"output_peak":0.7840552585481667,"output_rms":0.08275556456490114,"finished_at":"2026-10-13T04:41:14.316789+00:00","evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"5316d4fe732a09c59a55da1661c3c5129797c16875ae70f9830c3d6a716c62f2","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.47311794459889,"spectral_convergence":0.721063822582441,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.225758521150411,"spectral_convergence":0.26368516672625947,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.505926652031966,"spectral_convergence":0.2217351294438677,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0021589003295964843,"delta_db":0.22466051818983437,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.0030583963045236365,"delta_db":0.8192911522056988,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.00121553237195598,"delta_db":0.9597036019294968,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.00034307393484782843,"delta_db":0.6541172703452293,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.8652205664672936e-05,"delta_db":0.06496039065818987,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":3.392831986691831e-05,"delta_db":-0.15910752401131978,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.48030674288996145,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6286516417595468,"pitch":null,"output_rms_dbfs":-21.644055883714444,"output_peak":0.7840552585481667,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.068267705927089},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.47311794459889,"spectral_convergence":0.721063822582441,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.225758521150411,"spectral_convergence":0.26368516672625947,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.505926652031966,"spectral_convergence":0.2217351294438677,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0021589003295964843,"delta_db":0.22466051818983437,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.0030583963045236365,"delta_db":0.8192911522056988,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.00121553237195598,"delta_db":0.9597036019294968,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.00034307393484782843,"delta_db":0.6541172703452293,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.8652205664672936e-05,"delta_db":0.06496039065818987,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":3.392831986691831e-05,"delta_db":-0.15910752401131978,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.48030674288996145,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6286516417595468,"pitch":null,"output_rms_dbfs":-21.644055883714444,"output_peak":0.7840552585481667,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.068267705927089},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9240907489821771,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07697764557868021,"rms_dbfs":-22.27270752547399,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.699994329705078,"mean":-0.0004927848180135092,"zero_energy":false},"output":{"rms":0.08275556456490114,"rms_dbfs":-21.644055883714444,"peak":0.7840552585481667,"peak_dbfs":-2.1130665616978153,"crest_factor_db":19.530989322016627,"mean":1.8776768483835443e-05,"zero_energy":false},"rms_difference_db":0.6286516417595465},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":6.568879282171456e-06,"output_rms_dbfs":-51.82508719190018,"output_zero_energy":false,"output_peak":0.08057415456962126,"output_energy_fraction":7.033934102132293e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":233,"output_estimated_voiced_frames":250,"both_estimated_voiced_frames":215,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9227467811158798,"absolute_difference_cents_median":6.037341883963324,"absolute_difference_cents_p95":19.335678452198962,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019946925088635918,"log_rms_mae_db":1.9229138105263202,"rms_envelope_nrmse":0.15678071869958043,"envelope_correlation":0.9790240060184415,"rms_gain_error_db":0.63426078915094,"reference_mean_power":0.0059224241180442505,"output_mean_power":0.006853707784883938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.077279189796661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245127053592495,"output_power":0.0011758991722086514,"reference_active":true,"delta_db":0.5985273602281249},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329567656774176,"output_power":0.00014672213504206454,"reference_active":true,"delta_db":0.4278519618451155},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3018966563501677e-05,"output_power":2.5835903065130284e-05,"reference_active":true,"delta_db":0.501378243446819}],"active_band_count":3,"active_band_mae_db":0.5092525218400198}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001930447992601615,"log_rms_mae_db":1.3736223645969923,"rms_envelope_nrmse":0.09877041767923929,"envelope_correlation":0.9952215394326875,"rms_gain_error_db":0.6360425085755732,"reference_mean_power":0.005919423744261807,"output_mean_power":0.006853046538426549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8759871534324143e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000947359244704089,"output_power":0.0010987793024202102,"reference_active":true,"delta_db":0.6439577257970928},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.123049684235516e-05,"output_power":6.942030130188653e-05,"reference_active":true,"delta_db":0.5451871084489747},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3299097007749553e-06,"output_power":1.2948699305038479e-06,"reference_active":true,"delta_db":-0.11596008118918377}],"active_band_count":3,"active_band_mae_db":0.4350349718117504}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015110036587973006,"log_rms_mae_db":2.1513014744789616,"rms_envelope_nrmse":0.21173643085683885,"envelope_correlation":0.9799732927743582,"rms_gain_error_db":1.1120381420907697,"reference_mean_power":0.0025365732398871734,"output_mean_power":0.003276809700834767},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.31579034521966e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006679887319597139,"output_power":0.000877324418227309,"reference_active":true,"delta_db":1.183910804813713},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.085773760663629e-05,"output_power":0.00011389636840466896,"reference_active":true,"delta_db":0.9814795827359374},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6589878271689648e-05,"output_power":1.444129360561407e-05,"reference_active":true,"delta_db":-0.6023710168976666}],"active_band_count":3,"active_band_mae_db":0.9225871348157724}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013775289901134618,"log_rms_mae_db":1.6793140404942002,"rms_envelope_nrmse":0.1601161574120903,"envelope_correlation":0.9943977241564731,"rms_gain_error_db":1.1152546528412168,"reference_mean_power":0.0025400163932975937,"output_mean_power":0.0032836887484496095},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2092894134690622e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006291541070562533,"output_power":0.0008290213731240037,"reference_active":true,"delta_db":1.198086913249974},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9750836404697825e-05,"output_power":5.107284463324378e-05,"reference_active":true,"delta_db":1.0884377651658625},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.784451638039305e-07,"output_power":7.693288477416064e-07,"reference_active":true,"delta_db":-1.0442447337431877}],"active_band_count":3,"active_band_mae_db":1.1102564707196747}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010600794217574422,"log_rms_mae_db":1.709429704006701,"rms_envelope_nrmse":0.20623542443048545,"envelope_correlation":0.9748362661572421,"rms_gain_error_db":0.6158795797219115,"reference_mean_power":0.00026914786021620774,"output_mean_power":0.0003101550732697277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.028210835890849e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001073533616295001,"output_power":0.000118883783466415,"reference_active":true,"delta_db":0.4430696967540834},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.315168535027105e-06,"output_power":8.536141617037883e-06,"reference_active":true,"delta_db":0.11390555454226614},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4089826752830493e-06,"output_power":3.2051254543279295e-06,"reference_active":true,"delta_db":1.2401135656087998}],"active_band_count":3,"active_band_mae_db":0.5990296056350498}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":1.1193481966707608,"rms_envelope_nrmse":0.10610599883472356,"envelope_correlation":0.9958398390788005,"rms_gain_error_db":0.6168920016871732,"reference_mean_power":0.00026921330106685767,"output_mean_power":0.0003103028136307819},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9187018796507895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.751052870210858e-05,"output_power":0.00010845930912219031,"reference_active":true,"delta_db":0.4621532249500615},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.818200149529569e-06,"output_power":3.968680241857859e-06,"reference_active":true,"delta_db":0.16787418726260522},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4850622077020123e-07,"output_power":1.6283419977638002e-07,"reference_active":true,"delta_db":0.4000097805257192}],"active_band_count":3,"active_band_mae_db":0.34334573091279535}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":2.1068978234394256,"rms_envelope_nrmse":0.21776406085911448,"envelope_correlation":0.9691908001800934,"rms_gain_error_db":-0.08537019404782935,"reference_mean_power":4.950856594995313e-05,"output_mean_power":4.8544868329397746e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1644362822449518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0281496224399232e-05,"output_power":9.056270261857333e-06,"reference_active":true,"delta_db":-0.5510694589493927},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.187880808785332e-06,"output_power":3.124790237154799e-06,"reference_active":true,"delta_db":-1.27173443662892},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.689088055617458e-07,"output_power":6.280713651180638e-07,"reference_active":true,"delta_db":0.4296633778045695}],"active_band_count":3,"active_band_mae_db":0.7508224244609608}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":1.2996789054240712,"rms_envelope_nrmse":0.12906556533673313,"envelope_correlation":0.9887606446131552,"rms_gain_error_db":-0.08545774028727435,"reference_mean_power":4.9751584691373285e-05,"output_mean_power":4.8782173271352386e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0334524009764525e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.035252748280077e-05,"output_power":9.178081046128517e-06,"reference_active":true,"delta_db":-0.5229450372560053},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2970706912154717e-06,"output_power":1.6187497802048852e-06,"reference_active":true,"delta_db":-1.5199463814228118},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.669301652148395e-08,"output_power":5.295675559985433e-08,"reference_active":true,"delta_db":0.5466943847804724}],"active_band_count":3,"active_band_mae_db":0.8631952678197631}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9553513526916504,"predicted_mos":3.9553513526916504,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7b2c074cca871a7f5f00f8f4c19086981a1eaaf1ddf2072b5f9a722a063d5dbe","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.504529685000307},"started_at":"2026-10-13T04:46:44.585376+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.4191999222675804e-05,"native_input_samples":96000,"native_input_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","output_sha256":"7c79a685a5a66de70f589046cad24d4bda2057fd1c50cade25eec9f574f3550c","output_samples":96000,"length_delta_samples":0,"analysis_s":2.5658033129993782,"synthesis_s":7.70623398600037,"total_s":10.272037298999749,"rtf":1.712006216499958,"output_peak":0.4170346260070801,"output_rms":0.07351090078158998,"finished_at":"2026-10-13T04:46:55.043287+00:00","evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"7c79a685a5a66de70f589046cad24d4bda2057fd1c50cade25eec9f574f3550c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.333303511522153,"spectral_convergence":0.16290778686538426,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.43356158763647,"spectral_convergence":0.1619485951167452,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.571719207737737,"spectral_convergence":0.17203426886752668,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0019111490794140689,"delta_db":-0.30471975244055977,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.002419927470212179,"delta_db":-0.19762295479178446,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.0008031848298950185,"delta_db":-0.8398067206696511,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.0002110483521161576,"delta_db":-1.4559403111786589,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.3442189603043765e-05,"delta_db":-0.5638353783358968,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":2.5100612480297298e-05,"delta_db":-1.4678878744730839,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8049688319816058,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4002575844701575,"pitch":null,"output_rms_dbfs":-22.672965109944148,"output_peak":0.4170346260070801,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.446194768965454},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.333303511522153,"spectral_convergence":0.16290778686538426,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.43356158763647,"spectral_convergence":0.1619485951167452,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.571719207737737,"spectral_convergence":0.17203426886752668,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0019111490794140689,"delta_db":-0.30471975244055977,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.002419927470212179,"delta_db":-0.19762295479178446,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.0008031848298950185,"delta_db":-0.8398067206696511,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.0002110483521161576,"delta_db":-1.4559403111786589,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.3442189603043765e-05,"delta_db":-0.5638353783358968,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":2.5100612480297298e-05,"delta_db":-1.4678878744730839,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8049688319816058,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4002575844701575,"pitch":null,"output_rms_dbfs":-22.672965109944148,"output_peak":0.4170346260070801,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.446194768965454},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.920584826480866,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07697764557868021,"rms_dbfs":-22.27270752547399,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.699994329705078,"mean":-0.0004927848180135092,"zero_energy":false},"output":{"rms":0.07351090078158998,"rms_dbfs":-22.672965109944148,"peak":0.4170346260070801,"peak_dbfs":-7.596557689038183,"crest_factor_db":15.076407420905964,"mean":-6.534188745642193e-05,"zero_energy":false},"rms_difference_db":-0.40025758447015747},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":3.017627676276557e-06,"output_rms_dbfs":-55.20334345777645,"output_zero_energy":false,"output_peak":0.030487772077322006,"output_energy_fraction":4.095091323817308e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":233,"output_estimated_voiced_frames":236,"both_estimated_voiced_frames":220,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.944206008583691,"absolute_difference_cents_median":2.9056313430875145,"absolute_difference_cents_p95":13.09285223858029,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019946925088635918,"log_rms_mae_db":1.2329758627604581,"rms_envelope_nrmse":0.08644277652882389,"envelope_correlation":0.9931331355921885,"rms_gain_error_db":-0.40054141879294536,"reference_mean_power":0.0059224241180442505,"output_mean_power":0.005400641671020686},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.077279189796661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245127053592495,"output_power":0.0009328933144548188,"reference_active":true,"delta_db":-0.4068536788464677},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329567656774176,"output_power":0.00012413871622489188,"reference_active":true,"delta_db":-0.29803192032523285},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3018966563501677e-05,"output_power":2.1104769376841608e-05,"reference_active":true,"delta_db":-0.3770521133364521}],"active_band_count":3,"active_band_mae_db":0.3606459041693842}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001930447992601615,"log_rms_mae_db":0.9734729258324509,"rms_envelope_nrmse":0.07968202833298225,"envelope_correlation":0.9940213306667308,"rms_gain_error_db":-0.3977898478688825,"reference_mean_power":0.005919423744261807,"output_mean_power":0.005401326687561972},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8759871534324143e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000947359244704089,"output_power":0.0008640004083804922,"reference_active":true,"delta_db":-0.4000074973687585},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.123049684235516e-05,"output_power":5.6858152219505586e-05,"reference_active":true,"delta_db":-0.3217504159474877},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3299097007749553e-06,"output_power":1.3395336586019384e-06,"reference_active":true,"delta_db":0.03131476734503442}],"active_band_count":3,"active_band_mae_db":0.2510242268870935}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015110036587973006,"log_rms_mae_db":1.6349434860868792,"rms_envelope_nrmse":0.10068623338439406,"envelope_correlation":0.9925540938117922,"rms_gain_error_db":-0.3904685888498629,"reference_mean_power":0.0025365732398871734,"output_mean_power":0.0023184650186207955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.31579034521966e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006679887319597139,"output_power":0.000577307644570385,"reference_active":true,"delta_db":-0.6336182823298033},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.085773760663629e-05,"output_power":8.9834599350921e-05,"reference_active":true,"delta_db":-0.04918283255213851},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6589878271689648e-05,"output_power":1.654087851163306e-05,"reference_active":true,"delta_db":-0.01284627517713639}],"active_band_count":3,"active_band_mae_db":0.2318824633530261}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013775289901134618,"log_rms_mae_db":1.3084906254761532,"rms_envelope_nrmse":0.08986021112051226,"envelope_correlation":0.9940138737957203,"rms_gain_error_db":-0.3851597824440042,"reference_mean_power":0.0025400163932975937,"output_mean_power":0.0023244517800671445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2092894134690622e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006291541070562533,"output_power":0.0005463336925121307,"reference_active":true,"delta_db":-0.6129905160934899},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9750836404697825e-05,"output_power":3.882150928060865e-05,"reference_active":true,"delta_db":-0.10273855524124383},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.784451638039305e-07,"output_power":9.967622521854075e-07,"reference_active":true,"delta_db":0.08055091727423716}],"active_band_count":3,"active_band_mae_db":0.26542666286965694}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010600794217574422,"log_rms_mae_db":1.8577519072200936,"rms_envelope_nrmse":0.19368593369243947,"envelope_correlation":0.9864836959526887,"rms_gain_error_db":-1.381878389428113,"reference_mean_power":0.00026914786021620774,"output_mean_power":0.00019579567418801763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.028210835890849e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001073533616295001,"output_power":7.824538344499118e-05,"reference_active":true,"delta_db":-1.3735692517598996},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.315168535027105e-06,"output_power":7.314657494146349e-06,"reference_active":true,"delta_db":-0.5567706081653498},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4089826752830493e-06,"output_power":2.4005716165204017e-06,"reference_active":true,"delta_db":-0.015190098333133575}],"active_band_count":3,"active_band_mae_db":0.6485099860861278}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":1.5452289941491872,"rms_envelope_nrmse":0.1791156903356344,"envelope_correlation":0.9905934842604188,"rms_gain_error_db":-1.3772450792779443,"reference_mean_power":0.00026921330106685767,"output_mean_power":0.00019605232879896432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9187018796507895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.751052870210858e-05,"output_power":7.13897398119835e-05,"reference_active":true,"delta_db":-1.3541571206347844},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.818200149529569e-06,"output_power":3.343156673619715e-06,"reference_active":true,"delta_db":-0.577019604752986},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4850622077020123e-07,"output_power":1.3158315830527235e-07,"reference_active":true,"delta_db":-0.5254433990043607}],"active_band_count":3,"active_band_mae_db":0.8188733747973771}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":2.2111153619785884,"rms_envelope_nrmse":0.23537307511730352,"envelope_correlation":0.9695730352724935,"rms_gain_error_db":-1.0763836529455932,"reference_mean_power":4.950856594995313e-05,"output_mean_power":3.8640432746910554e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1644362822449518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0281496224399232e-05,"output_power":8.429954547333654e-06,"reference_active":true,"delta_db":-0.8623108737083476},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.187880808785332e-06,"output_power":3.977343950179972e-06,"reference_active":true,"delta_db":-0.22401163486237133},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.689088055617458e-07,"output_power":6.946505475421222e-07,"reference_active":true,"delta_db":0.867237265491168}],"active_band_count":3,"active_band_mae_db":0.6511865913539623}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":1.8784618128173558,"rms_envelope_nrmse":0.19269200187921035,"envelope_correlation":0.9824807101557208,"rms_gain_error_db":-1.0761127694099772,"reference_mean_power":4.9751584691373285e-05,"output_mean_power":3.8832525986248314e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0334524009764525e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.035252748280077e-05,"output_power":8.528289834808699e-06,"reference_active":true,"delta_db":-0.8418444061922686},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2970706912154717e-06,"output_power":2.2606362905741386e-06,"reference_active":true,"delta_db":-0.0694366543725722},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.669301652148395e-08,"output_power":5.690062743212371e-08,"reference_active":true,"delta_db":0.8586512362188956}],"active_band_count":3,"active_band_mae_db":0.5899774322612455}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.437150716781616,"predicted_mos":3.437150716781616,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"da3dab763888a3f2b7853b0a28b688636a14ba633046ccdfbe6d6deb5131a84c","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.374712131999331},"started_at":"2026-10-13T05:00:25.485689+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0072287700004380895,"native_input_samples":264600,"native_input_sha256":"0651a8c9258714393d75e3a7ae9422cd762682ff90c347aaffe27025daeefef6","output_sha256":"502d83683bc17cbff41e796f8ddc6ff34aa599a401ad40c16a7f7bf21173f428","output_samples":264704,"length_delta_samples":104,"analysis_s":0.16328854200037313,"synthesis_s":4.458368431999588,"total_s":4.621656973999961,"rtf":0.7702761623333269,"output_peak":0.5715859532356262,"output_rms":0.07053908748183654,"finished_at":"2026-10-13T05:00:30.132038+00:00","evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"502d83683bc17cbff41e796f8ddc6ff34aa599a401ad40c16a7f7bf21173f428","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.02653347572734,"output_peak":0.5683542363282155},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.55741885382442,"spectral_convergence":0.4468626825758865,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.367687003817597,"spectral_convergence":0.295724807396142,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.300361347508298,"spectral_convergence":0.33208114277513634,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.001425127467248108,"delta_db":-1.579128302610485,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.0023702948486335465,"delta_db":-0.28762272366352076,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.0009111445305943974,"delta_db":-0.2920889669517291,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.00021996112215836813,"delta_db":-1.2763006999102422,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.559263724008089e-05,"delta_db":-0.29318074356935225,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":1.6587783573127294e-05,"delta_db":-3.266847459514872,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1658614827033669,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.756125577231505,"pitch":null,"output_rms_dbfs":-23.02883310270549,"output_peak":0.5683542363282155,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.408489068383439},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8156972703045088,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07697764557868021,"rms_dbfs":-22.27270752547399,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.699994329705078,"mean":-0.0004927848180135092,"zero_energy":false},"output":{"rms":0.07055996307714191,"rms_dbfs":-23.02883310270549,"peak":0.5683542363282155,"peak_dbfs":-4.907617971339857,"crest_factor_db":18.121215131365638,"mean":-8.28981074999195e-05,"zero_energy":false},"rms_difference_db":-0.7561255772315008},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":1.3283757710718758e-06,"output_rms_dbfs":-58.76679054321581,"output_zero_energy":false,"output_peak":0.006365035713519426,"output_energy_fraction":1.956616366975961e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":233,"output_estimated_voiced_frames":198,"both_estimated_voiced_frames":182,"reference_voiced_output_unvoiced_frames":51,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.7811158798283262,"absolute_difference_cents_median":9.700527857000825,"absolute_difference_cents_p95":23.448768443911003,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019946925088635918,"log_rms_mae_db":1.6236652604277333,"rms_envelope_nrmse":0.12446981701477111,"envelope_correlation":0.9895192182348574,"rms_gain_error_db":-0.7522612771789873,"reference_mean_power":0.0059224241180442505,"output_mean_power":0.004980504963526379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.077279189796661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245127053592495,"output_power":0.0009226229933640095,"reference_active":true,"delta_db":-0.45493074838080855},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329567656774176,"output_power":0.00011575250163813726,"reference_active":true,"delta_db":-0.6018005612245186},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3018966563501677e-05,"output_power":2.1895968719343677e-05,"reference_active":true,"delta_db":-0.2172165811285514}],"active_band_count":3,"active_band_mae_db":0.4246492969112929}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001930447992601615,"log_rms_mae_db":1.3701477513607232,"rms_envelope_nrmse":0.10708760462519408,"envelope_correlation":0.9941638410720115,"rms_gain_error_db":-0.7482088543710742,"reference_mean_power":0.005919423744261807,"output_mean_power":0.004982628924377589},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8759871534324143e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000947359244704089,"output_power":0.0008606651130761832,"reference_active":true,"delta_db":-0.4168049822596697},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.123049684235516e-05,"output_power":5.225456437430292e-05,"reference_active":true,"delta_db":-0.6884355189530582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3299097007749553e-06,"output_power":1.1461093554620142e-06,"reference_active":true,"delta_db":-0.6459609630619558}],"active_band_count":3,"active_band_mae_db":0.5837338214248946}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015110036587973006,"log_rms_mae_db":2.0309726263743753,"rms_envelope_nrmse":0.13464878578802422,"envelope_correlation":0.9842727535991262,"rms_gain_error_db":-0.13335806895149105,"reference_mean_power":0.0025365732398871734,"output_mean_power":0.002459866848459091},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.31579034521966e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006679887319597139,"output_power":0.0005777457931592075,"reference_active":true,"delta_db":-0.6303234473500429},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.085773760663629e-05,"output_power":8.684547164257916e-05,"reference_active":true,"delta_db":-0.19614740423636975},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6589878271689648e-05,"output_power":1.5606112956577874e-05,"reference_active":true,"delta_db":-0.26548453396897925}],"active_band_count":3,"active_band_mae_db":0.36398512851846393}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013775289901134618,"log_rms_mae_db":1.4928529303995979,"rms_envelope_nrmse":0.10989129938531307,"envelope_correlation":0.9889205682304065,"rms_gain_error_db":-0.15019378123489519,"reference_mean_power":0.0025400163932975937,"output_mean_power":0.002453675584490554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2092894134690622e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006291541070562533,"output_power":0.0005444112398127493,"reference_active":true,"delta_db":-0.6282995299400268},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9750836404697825e-05,"output_power":3.921370939092587e-05,"reference_active":true,"delta_db":-0.059083451605189644},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.784451638039305e-07,"output_power":8.838521018762759e-07,"reference_active":true,"delta_db":-0.4415689193210556}],"active_band_count":3,"active_band_mae_db":0.3763173002887574}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010600794217574422,"log_rms_mae_db":2.8581457954068257,"rms_envelope_nrmse":0.2269075496804692,"envelope_correlation":0.9696215389584274,"rms_gain_error_db":-1.0750598484680376,"reference_mean_power":0.00026914786021620774,"output_mean_power":0.00021012849335238777},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.028210835890849e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001073533616295001,"output_power":9.155684435616047e-05,"reference_active":true,"delta_db":-0.6912483269606278},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.315168535027105e-06,"output_power":8.376259094654432e-06,"reference_active":true,"delta_db":0.03179046394279357},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4089826752830493e-06,"output_power":2.8463579712965383e-06,"reference_active":true,"delta_db":0.7245584140180533}],"active_band_count":3,"active_band_mae_db":0.48253240164049155}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":2.4400700347895663,"rms_envelope_nrmse":0.1922485797871207,"envelope_correlation":0.9800068537003177,"rms_gain_error_db":-1.0561543035354102,"reference_mean_power":0.00026921330106685767,"output_mean_power":0.00021109652443238058},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9187018796507895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.751052870210858e-05,"output_power":8.40985799260372e-05,"reference_active":true,"delta_db":-0.6426284875048256},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.818200149529569e-06,"output_power":3.796367240792686e-06,"reference_active":true,"delta_db":-0.024904731382605506},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4850622077020123e-07,"output_power":2.0344164486524898e-07,"reference_active":true,"delta_db":1.3669521236146576}],"active_band_count":3,"active_band_mae_db":0.6781617808340296}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":3.8027703027840016,"rms_envelope_nrmse":0.28772707828087285,"envelope_correlation":0.9756843785946725,"rms_gain_error_db":-2.2919072036363146,"reference_mean_power":4.950856594995313e-05,"output_mean_power":2.9207179955498287e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1644362822449518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0281496224399232e-05,"output_power":5.87294730482255e-06,"reference_active":true,"delta_db":-2.432002162294407},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.187880808785332e-06,"output_power":2.6670337950534354e-06,"reference_active":true,"delta_db":-1.9596579388970694},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.689088055617458e-07,"output_power":3.074834063524878e-07,"reference_active":true,"delta_db":-2.6722097221248315}],"active_band_count":3,"active_band_mae_db":2.354623274438769}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":3.3884425049070557,"rms_envelope_nrmse":0.25469793872955215,"envelope_correlation":0.9903270078840907,"rms_gain_error_db":-2.2935263794086524,"reference_mean_power":4.9751584691373285e-05,"output_mean_power":2.933960621037251e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0334524009764525e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.035252748280077e-05,"output_power":5.925818005619849e-06,"reference_active":true,"delta_db":-2.422980828467157},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2970706912154717e-06,"output_power":1.4573380785146624e-06,"reference_active":true,"delta_db":-1.9761404794924184},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.669301652148395e-08,"output_power":2.9643446995520502e-08,"reference_active":true,"delta_db":-1.973232288609643}],"active_band_count":3,"active_band_mae_db":2.124117865523073}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"golf","status":"measured","mos":{"status":"measured","value":2.9423038959503174,"predicted_mos":2.9423038959503174,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"67727f239d14ee9b0fcf9b79d33ee91b3e572767a6e60910e5f79fd31fdfc089","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.915817356999469},"started_at":"2026-10-13T05:02:00.241938+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031796450002730126,"native_input_samples":144000,"native_input_sha256":"ca16dc2355b94f00f5386fb440dc3179497144bc00249f27c6ebbda44be8ea02","output_sha256":"b5ce18d5f9253dc9ec0f1c330a9f96bb807e7e627420d7037686982c8c5b09ef","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.7385173359998589,"synthesis_s":0.13410350899994228,"total_s":0.8726208449998012,"rtf":0.14543680749996687,"output_peak":0.4717522859573364,"output_rms":0.06984953808743949,"finished_at":"2026-10-13T05:02:01.131981+00:00","analysis_stages":{"external_f0_s":0.08692095000060363},"evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"b5ce18d5f9253dc9ec0f1c330a9f96bb807e7e627420d7037686982c8c5b09ef","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.112870571455353,"output_peak":0.4575936877100967},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.596960711724076,"spectral_convergence":0.5628904446690748,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.99367964675505,"spectral_convergence":0.24596978318095758,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.32825910619662,"spectral_convergence":0.22949950086861942,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020438864240225514,"output_power":0.0015536182485034263,"delta_db":-1.1911244524790305,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00252799640048965,"output_power":0.0023475443398287715,"delta_db":-0.3216264737581653,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009718591019328043,"output_power":0.0007122820000357749,"delta_db":-1.349513369019234,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029501071489106855,"output_power":0.00022042613319655976,"delta_db":-1.2657470784019855,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.824395085733052e-05,"output_power":2.5587607596606936e-05,"delta_db":-1.745330695676077,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.5387711423272274e-05,"output_power":2.3836468836217716e-05,"delta_db":-1.7161055815079294,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2649079418070703,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8304971249098518,"pitch":null,"output_rms_dbfs":-23.112870571455353,"output_peak":0.4575936877100967,"reference_rms_dbfs":-22.282373446545503,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.639633154891915},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9119452682263187,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07689203016969104,"rms_dbfs":-22.282373446545503,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.709660250776587,"mean":-0.0005091738820275002,"zero_energy":false},"output":{"rms":0.06988057525519775,"rms_dbfs":-23.11287057145535,"peak":0.4575936877100967,"peak_dbfs":-6.790399501240383,"crest_factor_db":16.322471070214974,"mean":0.0006904659412946977,"zero_energy":false},"rms_difference_db":-0.8304971249098472},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":4.492656440181213e-06,"output_rms_dbfs":-53.474967912530374,"output_zero_energy":false,"output_peak":0.01424301046012351,"output_energy_fraction":6.757967828763497e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":232,"output_estimated_voiced_frames":267,"both_estimated_voiced_frames":228,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.9827586206896551,"absolute_difference_cents_median":6.844035440351095,"absolute_difference_cents_p95":23.938716903827203,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019946925088635918,"log_rms_mae_db":1.7983369339810273,"rms_envelope_nrmse":0.1327427822679023,"envelope_correlation":0.9878525866828294,"rms_gain_error_db":-0.8319900848619542,"reference_mean_power":0.005910869457197018,"output_mean_power":0.004880365624844156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0774323857696722e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245416649083147,"output_power":0.0008798420264011082,"reference_active":true,"delta_db":-0.6612492191514995},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013165929287112422,"output_power":0.00010677558766188581,"reference_active":true,"delta_db":-0.9097954787376974},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.292542371735639e-05,"output_power":2.5249809161877847e-05,"reference_active":true,"delta_db":0.4194072883814308}],"active_band_count":3,"active_band_mae_db":0.6634839954235425}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001930447992601615,"log_rms_mae_db":1.2971984836009696,"rms_envelope_nrmse":0.10798281185683992,"envelope_correlation":0.9952556143248321,"rms_gain_error_db":-0.8309483402173772,"reference_mean_power":0.00590877835000785,"output_mean_power":0.004879809464587819},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8758194353968543e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009469683470568022,"output_power":0.0008190875514163566,"reference_active":true,"delta_db":-0.630051371845251},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.041007801126947e-05,"output_power":4.792870016115139e-05,"reference_active":true,"delta_db":-1.0051374558926816},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3035514683915637e-06,"output_power":1.4455165737245978e-06,"reference_active":true,"delta_db":0.4489489230346883}],"active_band_count":3,"active_band_mae_db":0.6947125835908735}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015110036587973006,"log_rms_mae_db":1.82398560050567,"rms_envelope_nrmse":0.14700626140293935,"envelope_correlation":0.9877954936055013,"rms_gain_error_db":-0.8545133972012459,"reference_mean_power":0.0025342768612421863,"output_mean_power":0.0020816260716179542},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3164961185804597e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006681999361189144,"output_power":0.0005647958476380468,"reference_active":true,"delta_db":-0.7301493464718384},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.036066629271582e-05,"output_power":7.606930561619769e-05,"reference_active":true,"delta_db":-0.7476997271791521},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.652420567572365e-05,"output_power":1.5776496100431365e-05,"reference_active":true,"delta_db":-0.20110037537484443}],"active_band_count":3,"active_band_mae_db":0.559649816341945}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013775289901134618,"log_rms_mae_db":1.2716552133411494,"rms_envelope_nrmse":0.11791668151244855,"envelope_correlation":0.9948082211194101,"rms_gain_error_db":-0.8545837838202461,"reference_mean_power":0.002537676371614413,"output_mean_power":0.002084384608473277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2098570341412897e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006292499297047673,"output_power":0.0005363852684253544,"reference_active":true,"delta_db":-0.6934633405898284},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.937620858638868e-05,"output_power":3.2550881956436666e-05,"reference_active":true,"delta_db":-0.8267113679666939},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.706907835938042e-07,"output_power":8.870530013395925e-07,"reference_active":true,"delta_db":-0.3913133648860134}],"active_band_count":3,"active_band_mae_db":0.6371626911475119}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00010600794217574422,"log_rms_mae_db":1.907057663369429,"rms_envelope_nrmse":0.23097578685485387,"envelope_correlation":0.9727249064605796,"rms_gain_error_db":-1.360275039413263,"reference_mean_power":0.0002686997092136819,"output_mean_power":0.00019644441787534743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.019776113896628e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010674756305236168,"output_power":7.476882496979204e-05,"reference_active":true,"delta_db":-1.546374137298163},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.298874033607721e-06,"output_power":4.912693939774241e-06,"reference_active":true,"delta_db":-2.2769946411244577},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4073181897221613e-06,"output_power":2.4366506717907317e-06,"reference_active":true,"delta_db":0.05259774031197438}],"active_band_count":3,"active_band_mae_db":1.2919888395781984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":1.5865240348510243,"rms_envelope_nrmse":0.1806302172768214,"envelope_correlation":0.9895379882007033,"rms_gain_error_db":-1.353994637902682,"reference_mean_power":0.00026909391972083,"output_mean_power":0.00019701732631064974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.910213100043573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.696462114403794e-05,"output_power":6.86514130960697e-05,"reference_active":true,"delta_db":-1.499638239136571},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.80005527016239e-06,"output_power":2.219107660602331e-06,"reference_active":true,"delta_db":-2.336115406321784},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4753515991858198e-07,"output_power":1.3250425846624395e-07,"reference_active":true,"delta_db":-0.46665695766169335}],"active_band_count":3,"active_band_mae_db":1.4341368677066828}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":2.527614738340033,"rms_envelope_nrmse":0.27226596978612716,"envelope_correlation":0.9699192936049482,"rms_gain_error_db":-1.8207864286704158,"reference_mean_power":4.9587875651275995e-05,"output_mean_power":3.260595017805848e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.1704196407361677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0316801581839808e-05,"output_power":4.980882661834165e-06,"reference_active":true,"delta_db":-3.1623876717037107},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.181494935992546e-06,"output_power":1.8922665064415726e-06,"reference_active":true,"delta_db":-3.4434927284530508},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.690766103512596e-07,"output_power":2.6770745818190313e-07,"reference_active":true,"delta_db":-3.275102654541635}],"active_band_count":3,"active_band_mae_db":3.293661018232799}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":2.235923480730861,"rms_envelope_nrmse":0.23423422559563736,"envelope_correlation":0.9843469143925413,"rms_gain_error_db":-1.8208025454591978,"reference_mean_power":4.9831318679736745e-05,"output_mean_power":3.276590180929087e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.0390170327569967e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0382920421488995e-05,"output_power":4.921660765664072e-06,"reference_active":true,"delta_db":-3.2420784956559685},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.287845504487032e-06,"output_power":1.0212704576548456e-06,"reference_active":true,"delta_db":-3.5028592448648475},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6521451922879915e-08,"output_power":2.0068371017306518e-08,"reference_active":true,"delta_db":-3.6514113896956886}],"active_band_count":3,"active_band_mae_db":3.465449710072168}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3468568325042725,"predicted_mos":3.3468568325042725,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"d57858b389399b0ccdf73eb3a2994b8b98c96a222411dca5ffcdd30cf2fc7cc4","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4664015599992126},"started_at":"2026-10-13T05:05:41.743803+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032062660002338816,"native_input_samples":144000,"native_input_sha256":"ca16dc2355b94f00f5386fb440dc3179497144bc00249f27c6ebbda44be8ea02","output_sha256":"291cee92cc2da5f2f01e2a7cf147e73c5a61fbd947d5016763b6f19cdebfc033","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3630712009999115,"synthesis_s":3.3357391819999975,"total_s":5.698810382999909,"rtf":0.9498017304999848,"output_peak":0.5929363965988159,"output_rms":0.079940916133864,"finished_at":"2026-10-13T05:05:47.456742+00:00","analysis_stages":{"external_f0_s":1.478025601999434},"evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"291cee92cc2da5f2f01e2a7cf147e73c5a61fbd947d5016763b6f19cdebfc033","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.94141762155443,"output_peak":0.6025514359553273},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.816949495209047,"spectral_convergence":0.7124581996578715,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.229852499448754,"spectral_convergence":0.2939730917377936,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.251731767746664,"spectral_convergence":0.25850907342832946,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0029876247656086324,"delta_db":1.6355951000732034,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.002279691511686032,"delta_db":-0.45688562870668126,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.0008142226081411164,"delta_db":-0.7805301217053038,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.000247740397642447,"delta_db":-0.7597916627749359,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.7529471559714595e-05,"delta_db":-0.06305793350103105,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":2.946907484888331e-05,"delta_db":-0.7710660444366502,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7444877485329675,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.3319807209822262,"pitch":null,"output_rms_dbfs":-21.940726804491764,"output_peak":0.6025514359553273,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.766177920801489},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9077646111484361,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07697764557868021,"rms_dbfs":-22.27270752547399,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.699994329705078,"mean":-0.0004927848180135092,"zero_energy":false},"output":{"rms":0.0799767330508494,"rms_dbfs":-21.940726804491764,"peak":0.6025514359553273,"peak_dbfs":-4.400117484668453,"crest_factor_db":17.54060931982331,"mean":2.088391661689379e-06,"zero_energy":false},"rms_difference_db":0.33198072098222653},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":5.408440782033167e-05,"output_rms_dbfs":-42.66927921096769,"output_zero_energy":false,"output_peak":0.07615384826276735,"output_energy_fraction":0.0006200778034594113,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":233,"output_estimated_voiced_frames":273,"both_estimated_voiced_frames":224,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.9613733905579399,"absolute_difference_cents_median":6.286419010056413,"absolute_difference_cents_p95":22.300735228424685,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019946925088635918,"log_rms_mae_db":2.501188348810441,"rms_envelope_nrmse":0.15239136480076154,"envelope_correlation":0.9747116721016393,"rms_gain_error_db":0.3326378625178806,"reference_mean_power":0.0059224241180442505,"output_mean_power":0.006393862480629516},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.077279189796661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245127053592495,"output_power":0.0010690817157393813,"reference_active":true,"delta_db":0.18493553354207526},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329567656774176,"output_power":0.00012854392754086067,"reference_active":true,"delta_db":-0.1465887662425053},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3018966563501677e-05,"output_power":2.3458905806634507e-05,"reference_active":true,"delta_db":0.08221929380620689}],"active_band_count":3,"active_band_mae_db":0.13791453119692915}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001930447992601615,"log_rms_mae_db":1.8834222362034296,"rms_envelope_nrmse":0.10357478652003609,"envelope_correlation":0.9885040352059591,"rms_gain_error_db":0.33321743892832817,"reference_mean_power":0.005919423744261807,"output_mean_power":0.006391476171228343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8759871534324143e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000947359244704089,"output_power":0.0009890454503064176,"reference_active":true,"delta_db":0.1870155200733908},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.123049684235516e-05,"output_power":6.032211863500492e-05,"reference_active":true,"delta_db":-0.06491197012292219},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3299097007749553e-06,"output_power":1.2827037196877391e-06,"reference_active":true,"delta_db":-0.1569579976696155}],"active_band_count":3,"active_band_mae_db":0.13629516262197616}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015110036587973006,"log_rms_mae_db":2.4351365459323797,"rms_envelope_nrmse":0.17802464959654973,"envelope_correlation":0.9760923042402807,"rms_gain_error_db":-0.7189498809047713,"reference_mean_power":0.0025365732398871734,"output_mean_power":0.0021495740885322913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.31579034521966e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006679887319597139,"output_power":0.000574112754762374,"reference_active":true,"delta_db":-0.6577194112598443},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.085773760663629e-05,"output_power":6.80005164666997e-05,"reference_active":true,"delta_db":-1.2584970730692338},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6589878271689648e-05,"output_power":1.194070984849316e-05,"reference_active":true,"delta_db":-1.428130540018687}],"active_band_count":3,"active_band_mae_db":1.114782341449255}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013775289901134618,"log_rms_mae_db":1.802181684164169,"rms_envelope_nrmse":0.12496544064768147,"envelope_correlation":0.9906314787502659,"rms_gain_error_db":-0.7181261063163751,"reference_mean_power":0.0025400163932975937,"output_mean_power":0.0021529002537176616},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2092894134690622e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006291541070562533,"output_power":0.0005427195614196003,"reference_active":true,"delta_db":-0.6418156070020531},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9750836404697825e-05,"output_power":2.9627289625909047e-05,"reference_active":true,"delta_db":-1.2765434814183354},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.784451638039305e-07,"output_power":7.298698822079005e-07,"reference_active":true,"delta_db":-1.2729104795038684}],"active_band_count":3,"active_band_mae_db":1.063756522641419}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010600794217574422,"log_rms_mae_db":2.2219760528922348,"rms_envelope_nrmse":0.2151669367908913,"envelope_correlation":0.9672667001052906,"rms_gain_error_db":-0.577526577384855,"reference_mean_power":0.00026914786021620774,"output_mean_power":0.00023563417254853454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.028210835890849e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001073533616295001,"output_power":9.631426850634308e-05,"reference_active":true,"delta_db":-0.47125017768719035},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.315168535027105e-06,"output_power":7.596099568979408e-06,"reference_active":true,"delta_db":-0.3928040729728472},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4089826752830493e-06,"output_power":1.8050158633678936e-06,"reference_active":true,"delta_db":-1.2535265363528525}],"active_band_count":3,"active_band_mae_db":0.7058602623376299}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":1.499367116575164,"rms_envelope_nrmse":0.1358384672880545,"envelope_correlation":0.9882346167075541,"rms_gain_error_db":-0.5724612179614136,"reference_mean_power":0.00026921330106685767,"output_mean_power":0.00023596652210594094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9187018796507895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.751052870210858e-05,"output_power":8.756729639249881e-05,"reference_active":true,"delta_db":-0.46709569954296015},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.818200149529569e-06,"output_power":3.4618630573496725e-06,"reference_active":true,"delta_db":-0.42548806036808223},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4850622077020123e-07,"output_power":1.2866402467535615e-07,"reference_active":true,"delta_db":-0.622875139492088}],"active_band_count":3,"active_band_mae_db":0.5051529664677101}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":2.3784248519598123,"rms_envelope_nrmse":0.2464467002712445,"envelope_correlation":0.9610594498940435,"rms_gain_error_db":-0.5951772346184661,"reference_mean_power":4.950856594995313e-05,"output_mean_power":4.3168069123237323e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1644362822449518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0281496224399232e-05,"output_power":9.734165373104303e-06,"reference_active":true,"delta_db":-0.23757600200021597},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.187880808785332e-06,"output_power":4.0583698105977625e-06,"reference_active":true,"delta_db":-0.13642694042299808},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.689088055617458e-07,"output_power":4.402628337868622e-07,"reference_active":true,"delta_db":-1.113306311915641}],"active_band_count":3,"active_band_mae_db":0.49576975144628505}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":1.4377223020301804,"rms_envelope_nrmse":0.15911376708643984,"envelope_correlation":0.984902660024399,"rms_gain_error_db":-0.5951145254695518,"reference_mean_power":4.9751584691373285e-05,"output_mean_power":4.338059115606174e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0334524009764525e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.035252748280077e-05,"output_power":9.833339130583698e-06,"reference_active":true,"delta_db":-0.22345374798123363},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2970706912154717e-06,"output_power":2.2215058238449538e-06,"reference_active":true,"delta_db":-0.14526904484667316},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.669301652148395e-08,"output_power":3.9604228044318605e-08,"reference_active":true,"delta_db":-0.715103791309427}],"active_band_count":3,"active_band_mae_db":0.36127552804577795}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.283848762512207,"predicted_mos":3.283848762512207,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"9ac7745ba8510a0f1a6170bdb3e659645a5dba327f49ad0b4525adb3bfc812e8","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.293549989000894},"started_at":"2026-10-13T05:10:06.845169+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029355110000324203,"native_input_samples":144000,"native_input_sha256":"ca16dc2355b94f00f5386fb440dc3179497144bc00249f27c6ebbda44be8ea02","output_sha256":"a7e54069620a6288df23d9250aefbaffe510365ffe5ad09cddbdc468a4e4009a","output_samples":144240,"length_delta_samples":240,"analysis_s":1.1490635319987632,"synthesis_s":0.14759075800066057,"total_s":1.2966542899994238,"rtf":0.2161090483332373,"output_peak":0.5330260396003723,"output_rms":0.07566550893494238,"finished_at":"2026-10-13T05:10:08.153533+00:00","analysis_stages":{"external_f0_s":0.16440162300023076},"evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"a7e54069620a6288df23d9250aefbaffe510365ffe5ad09cddbdc468a4e4009a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.4181415858725,"output_peak":0.5312268341843552},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.70542477355794,"spectral_convergence":0.13146648231916136,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.268207243352192,"spectral_convergence":0.12192116572704535,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.621725189772283,"spectral_convergence":0.12813236692805585,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.0020323747594030305,"delta_db":-0.037627481594988486,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.002433830400040633,"delta_db":-0.17274333623122046,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.0009379092512357306,"delta_db":-0.1663534961261024,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.0002694837585311806,"delta_db":-0.39443400718122346,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":3.264069356223721e-05,"delta_db":-0.6691886440779543,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":3.359797661818395e-05,"delta_db":-0.20159982761377174,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27365779880421015,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13829700134640702,"pitch":null,"output_rms_dbfs":-22.4110045268204,"output_peak":0.5312268341843552,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.531785735560805},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9570455156301879,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07697764557868021,"rms_dbfs":-22.27270752547399,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.699994329705078,"mean":-0.0004927848180135092,"zero_energy":false},"output":{"rms":0.07576171090591206,"rms_dbfs":-22.411004526820392,"peak":0.5312268341843552,"peak_dbfs":-5.494399906089722,"crest_factor_db":16.916604620730674,"mean":1.1320363011313414e-05,"zero_energy":false},"rms_difference_db":-0.13829700134640177},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":1.3423011488033098e-05,"output_rms_dbfs":-48.721500381229305,"output_zero_energy":false,"output_peak":0.053467848422041796,"output_energy_fraction":0.00017149514931046977,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":233,"output_estimated_voiced_frames":228,"both_estimated_voiced_frames":216,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.927038626609442,"absolute_difference_cents_median":1.4828996198574385,"absolute_difference_cents_p95":7.065379072261679,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019946925088635918,"log_rms_mae_db":0.8544783194227275,"rms_envelope_nrmse":0.054802190577857075,"envelope_correlation":0.9966264616635965,"rms_gain_error_db":-0.13629892553083234,"reference_mean_power":0.0059224241180442505,"output_mean_power":0.005739441219477009},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.077279189796661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245127053592495,"output_power":0.0009935639265338204,"reference_active":true,"delta_db":-0.13321533488884033},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329567656774176,"output_power":0.00011597245262922573,"reference_active":true,"delta_db":-0.5935559998471244},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3018966563501677e-05,"output_power":1.611241915655534e-05,"reference_active":true,"delta_db":-1.5492507076973991}],"active_band_count":3,"active_band_mae_db":0.7586740141444546}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001930447992601615,"log_rms_mae_db":0.4653593656522764,"rms_envelope_nrmse":0.02939199063125309,"envelope_correlation":0.9991615461636891,"rms_gain_error_db":-0.13410510325458624,"reference_mean_power":0.005919423744261807,"output_mean_power":0.005739432067527621},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8759871534324143e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000947359244704089,"output_power":0.0009221009446585221,"reference_active":true,"delta_db":-0.11736230549146165},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.123049684235516e-05,"output_power":5.5805962699480496e-05,"reference_active":true,"delta_db":-0.4028717896626699},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3299097007749553e-06,"output_power":9.973595954157331e-07,"reference_active":true,"delta_db":-1.249703836139712}],"active_band_count":3,"active_band_mae_db":0.5899793104312812}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015110036587973006,"log_rms_mae_db":1.1832226939126775,"rms_envelope_nrmse":0.06774738271168503,"envelope_correlation":0.9963917527416598,"rms_gain_error_db":-0.19885833289626487,"reference_mean_power":0.0025365732398871734,"output_mean_power":0.002423045521008025},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.31579034521966e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006679887319597139,"output_power":0.0006444076202894179,"reference_active":true,"delta_db":-0.15608469169922093},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.085773760663629e-05,"output_power":7.648777534780895e-05,"reference_active":true,"delta_db":-0.7476988888881708},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6589878271689648e-05,"output_power":1.1022536270182336e-05,"reference_active":true,"delta_db":-1.7756166283077683}],"active_band_count":3,"active_band_mae_db":0.8931334029650534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013775289901134618,"log_rms_mae_db":0.7143078968808658,"rms_envelope_nrmse":0.041292726374663094,"envelope_correlation":0.9988575310218563,"rms_gain_error_db":-0.1959715237277941,"reference_mean_power":0.0025400163932975937,"output_mean_power":0.0024279479222598715},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2092894134690622e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006291541070562533,"output_power":0.0006050621956807705,"reference_active":true,"delta_db":-0.16957016943605327},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9750836404697825e-05,"output_power":3.452907334657184e-05,"reference_active":true,"delta_db":-0.6116134777260631},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.784451638039305e-07,"output_power":6.781988199238909e-07,"reference_active":true,"delta_db":-1.59179461216714}],"active_band_count":3,"active_band_mae_db":0.7909927531097521}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010600794217574422,"log_rms_mae_db":1.202819693135268,"rms_envelope_nrmse":0.10310548239241431,"envelope_correlation":0.9936271941380177,"rms_gain_error_db":-0.4230186033419906,"reference_mean_power":0.00026914786021620774,"output_mean_power":0.0002441681828456287},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.028210835890849e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001073533616295001,"output_power":9.72849187485494e-05,"reference_active":true,"delta_db":-0.4277012783068959},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.315168535027105e-06,"output_power":6.255194330122459e-06,"reference_active":true,"delta_db":-1.2363024962627167},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4089826752830493e-06,"output_power":1.5753968570326134e-06,"reference_active":true,"delta_db":-1.8444370198470883}],"active_band_count":3,"active_band_mae_db":1.1694802648055669}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":0.7943768665489668,"rms_envelope_nrmse":0.07145827282013764,"envelope_correlation":0.9977352465189565,"rms_gain_error_db":-0.4228059165718312,"reference_mean_power":0.00026921330106685767,"output_mean_power":0.0002442395109489593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9187018796507895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.751052870210858e-05,"output_power":8.815873336714788e-05,"reference_active":true,"delta_db":-0.4378616942316576},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.818200149529569e-06,"output_power":2.8335534088908087e-06,"reference_active":true,"delta_db":-1.295272873564247},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4850622077020123e-07,"output_power":8.898772597223818e-08,"reference_active":true,"delta_db":-2.2241453739741575}],"active_band_count":3,"active_band_mae_db":1.3190933139233538}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":1.442835405936842,"rms_envelope_nrmse":0.11729123550668072,"envelope_correlation":0.991671581366944,"rms_gain_error_db":-0.30868396523160635,"reference_mean_power":4.950856594995313e-05,"output_mean_power":4.6111786714110265e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1644362822449518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0281496224399232e-05,"output_power":9.287250534262184e-06,"reference_active":true,"delta_db":-0.441691590718914},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.187880808785332e-06,"output_power":3.6517719451218212e-06,"reference_active":true,"delta_db":-0.5949066491963649},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.689088055617458e-07,"output_power":3.7443079103682185e-07,"reference_active":true,"delta_db":-1.816711004057177}],"active_band_count":3,"active_band_mae_db":0.951103081324152}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":0.9836050844816339,"rms_envelope_nrmse":0.06843371678373385,"envelope_correlation":0.997636606700162,"rms_gain_error_db":-0.30859113591364984,"reference_mean_power":4.9751584691373285e-05,"output_mean_power":4.6339122432859496e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0334524009764525e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.035252748280077e-05,"output_power":9.399941842168229e-06,"reference_active":true,"delta_db":-0.41921225485285374},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2970706912154717e-06,"output_power":1.9170088672081013e-06,"reference_active":true,"delta_db":-0.7855023886799593},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.669301652148395e-08,"output_power":3.334670615731001e-08,"reference_active":true,"delta_db":-1.46198989030326}],"active_band_count":3,"active_band_mae_db":0.888901511278691}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0023","method":"magphase","status":"measured","mos":{"status":"measured","value":3.5344130992889404,"predicted_mos":3.5344130992889404,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288667,"input_sample_rate":48000,"input_seconds":6.013895833333334,"input_float32_sha256":"b2c42d6a1f7dc5fe31a1a980e0fab551ac3fc39ea000c8db7186d5f1daa7bdfd","internal_samples":96223,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.228465629999846},"started_at":"2026-10-13T05:12:51.033592+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009246489998986362,"native_input_samples":288000,"native_input_sha256":"31b2fa26e320d39fc1cda3bf7ea38a9f409bb45b4f1a796caa0c46cca07e3ba7","output_sha256":"aaed75f909457a1b1ea575ac11e10d66d792cff23c9a9f6d52a16d164590fd53","output_samples":288667,"length_delta_samples":667,"analysis_s":2.9559302659999958,"synthesis_s":0.9030263049990026,"total_s":3.8589565709989984,"rtf":0.6431594284998331,"output_peak":0.5449213700096893,"output_rms":0.07398962610705814,"finished_at":"2026-10-13T05:12:54.918462+00:00","evaluation_input":{"reference":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"aaed75f909457a1b1ea575ac11e10d66d792cff23c9a9f6d52a16d164590fd53","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96223,"length_delta_samples":223,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.612634567959905,"output_peak":0.5432308034646103},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9976824667698991,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.548333831837114,"spectral_convergence":0.6676590263632273,"spectral_convergence_status":"ok","amplitude_floor":1.2959192499433599e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.40530802159897,"spectral_convergence":0.4286650378484482,"spectral_convergence_status":"ok","amplitude_floor":8.646137696849433e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.7458647289469456,"spectral_convergence":0.4217327770268648,"spectral_convergence_status":"ok","amplitude_floor":6.5781524348855e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0020500598531285057,"output_power":0.001823084693340924,"delta_db":-0.5095969607750133,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0025325885847670413,"output_power":0.0023690479598586917,"delta_db":-0.28990792205005855,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009745321457083826,"output_power":0.0009155306995515524,"delta_db":-0.2712325760144023,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029510461183058167,"output_power":0.00028581988616484385,"delta_db":-0.13883554693604863,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.807836063095196e-05,"output_power":4.1614242021412275e-05,"delta_db":0.38563745979298564,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.519436277144243e-05,"output_power":3.258024668320251e-05,"delta_db":-0.3351873801118987,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3217329742800679,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.34926338711297034,"pitch":null,"output_rms_dbfs":-22.62197091258696,"output_peak":0.5432308034646103,"reference_rms_dbfs":-22.27270752547399,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.566502194127676},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.722468216500512,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07697764557868021,"rms_dbfs":-22.27270752547399,"peak":0.526458740234375,"peak_dbfs":-5.572713195768912,"crest_factor_db":16.699994329705078,"mean":-0.0004927848180135092,"zero_energy":false},"output":{"rms":0.07394374704882507,"rms_dbfs":-22.621970912586956,"peak":0.5432308034646103,"peak_dbfs":-5.300312234219407,"crest_factor_db":17.32165867836755,"mean":0.001503039943231649,"zero_energy":false},"rms_difference_db":-0.34926338711296623},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.956296421238221e-06,"threshold_dbfs":-54.0271117681062,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":44,"quiet_samples":7040,"reference_mean_power":1.8675109128129076e-06,"output_mean_power":4.1139883464319485e-06,"output_rms_dbfs":-53.85736942852152,"output_zero_energy":false,"output_peak":0.0358605627638662,"output_energy_fraction":5.5177443471898e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":233,"output_estimated_voiced_frames":234,"both_estimated_voiced_frames":188,"reference_voiced_output_unvoiced_frames":45,"reference_unvoiced_output_voiced_frames":46,"paired_voiced_fraction":0.8068669527896996,"absolute_difference_cents_median":9.914145451110315,"absolute_difference_cents_p95":42.20794409106061,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019946925088635918,"log_rms_mae_db":3.374768606581745,"rms_envelope_nrmse":0.2789421244201902,"envelope_correlation":0.906127119911627,"rms_gain_error_db":-0.3489879074400506,"reference_mean_power":0.0059224241180442505,"output_mean_power":0.005465132732921264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.077279189796661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010245127053592495,"output_power":0.0009582864572886622,"reference_active":true,"delta_db":-0.29021997968243807},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329567656774176,"output_power":0.00011834735395549416,"reference_active":true,"delta_db":-0.5055188957620935},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3018966563501677e-05,"output_power":2.8068926333679925e-05,"reference_active":true,"delta_db":0.8613997864500587}],"active_band_count":3,"active_band_mae_db":0.5523795539648634}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001930447992601615,"log_rms_mae_db":2.5798326517769303,"rms_envelope_nrmse":0.21373481953650292,"envelope_correlation":0.9408637619936022,"rms_gain_error_db":-0.34251280999326106,"reference_mean_power":0.005919423744261807,"output_mean_power":0.005470514194226286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8759871534324143e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000947359244704089,"output_power":0.0008893735653376759,"reference_active":true,"delta_db":-0.27430480628737997},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.123049684235516e-05,"output_power":5.311529449777716e-05,"reference_active":true,"delta_db":-0.617481896774208},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3299097007749553e-06,"output_power":1.5143087133316782e-06,"reference_active":true,"delta_db":0.5639226737682487}],"active_band_count":3,"active_band_mae_db":0.4852364589432789}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015110036587973006,"log_rms_mae_db":4.488911223701332,"rms_envelope_nrmse":0.3377849465244981,"envelope_correlation":0.8992952020474878,"rms_gain_error_db":-0.3318486974637399,"reference_mean_power":0.0025365732398871734,"output_mean_power":0.0023499711831830705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.31579034521966e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006679887319597139,"output_power":0.000619965185041652,"reference_active":true,"delta_db":-0.3240183478722862},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.085773760663629e-05,"output_power":8.38403786299683e-05,"reference_active":true,"delta_db":-0.3490868755673249},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6589878271689648e-05,"output_power":1.6376259015125893e-05,"reference_active":true,"delta_db":-0.056285006701222405}],"active_band_count":3,"active_band_mae_db":0.24313007671361117}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013775289901134618,"log_rms_mae_db":3.6728527764093926,"rms_envelope_nrmse":0.2602476415409942,"envelope_correlation":0.9367143372344611,"rms_gain_error_db":-0.3302833885194872,"reference_mean_power":0.0025400163932975937,"output_mean_power":0.002354009334896963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2092894134690622e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006291541070562533,"output_power":0.0005850326218007253,"reference_active":true,"delta_db":-0.31576952679539017},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9750836404697825e-05,"output_power":3.658078332719435e-05,"reference_active":true,"delta_db":-0.3609327009975402},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.784451638039305e-07,"output_power":9.395886962573514e-07,"reference_active":true,"delta_db":-0.1759870763293264}],"active_band_count":3,"active_band_mae_db":0.2842297680407523}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010600794217574422,"log_rms_mae_db":4.05157307144561,"rms_envelope_nrmse":0.38260517074915684,"envelope_correlation":0.8938602767506245,"rms_gain_error_db":-0.08247374375258457,"reference_mean_power":0.00026914786021620774,"output_mean_power":0.0002640848923872249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.028210835890849e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001073533616295001,"output_power":0.00010628200033943303,"reference_active":true,"delta_db":-0.043559286506736405},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.315168535027105e-06,"output_power":8.139747740518603e-06,"reference_active":true,"delta_db":-0.09260110244347754},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4089826752830493e-06,"output_power":3.25768423463794e-06,"reference_active":true,"delta_db":1.3107530942371386}],"active_band_count":3,"active_band_mae_db":0.48230449439578416}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.461607013685598e-05,"log_rms_mae_db":3.28219972982119,"rms_envelope_nrmse":0.2928407948834905,"envelope_correlation":0.9346347149711627,"rms_gain_error_db":-0.06233124951324825,"reference_mean_power":0.00026921330106685767,"output_mean_power":0.0002653770660957933},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9187018796507895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.751052870210858e-05,"output_power":9.728356108764393e-05,"reference_active":true,"delta_db":-0.010120515180536},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.818200149529569e-06,"output_power":3.523930231034543e-06,"reference_active":true,"delta_db":-0.3483138891920028},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4850622077020123e-07,"output_power":1.8778543257664657e-07,"reference_active":true,"delta_db":1.0191725273968668}],"active_band_count":3,"active_band_mae_db":0.45920231058980193}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.633594187238401e-05,"log_rms_mae_db":4.315907557957546,"rms_envelope_nrmse":0.49238269380030014,"envelope_correlation":0.8425939498347854,"rms_gain_error_db":-0.07265473854720356,"reference_mean_power":4.950856594995313e-05,"output_mean_power":4.868720829823433e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1644362822449518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0281496224399232e-05,"output_power":1.0696681669010876e-05,"reference_active":true,"delta_db":0.17192751113380852},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.187880808785332e-06,"output_power":4.7414552084525295e-06,"reference_active":true,"delta_db":0.5391733949064536},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.689088055617458e-07,"output_power":8.924010673546393e-07,"reference_active":true,"delta_db":1.9551742522611186}],"active_band_count":3,"active_band_mae_db":0.8887583861004602}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.3561893402191574e-05,"log_rms_mae_db":3.064729436916236,"rms_envelope_nrmse":0.3622397120095522,"envelope_correlation":0.9105736563254171,"rms_gain_error_db":-0.07237102432998516,"reference_mean_power":4.9751584691373285e-05,"output_mean_power":4.8929391642927085e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0334524009764525e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.035252748280077e-05,"output_power":1.1123386721191655e-05,"reference_active":true,"delta_db":0.3119064428669098},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2970706912154717e-06,"output_power":2.6606915782126666e-06,"reference_active":true,"delta_db":0.6382017436472254},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.669301652148395e-08,"output_power":7.947744476286029e-08,"reference_active":true,"delta_db":2.3099196421071047}],"active_band_count":3,"active_band_mae_db":1.0866759428737467}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"hnm","status":"measured","mos":{"status":"measured","value":3.028024196624756,"predicted_mos":3.028024196624756,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"74e85b5aad7f40e77ec01dbb4f96ec1eecba478f7312e8e2ebb96dbd8421b683","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9228542339988053},"started_at":"2026-10-13T04:38:20.019589+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6197999684663955e-05,"native_input_samples":96000,"native_input_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","output_sha256":"1ddb3a3b62087bf523be518d71810f644c9cf48e66e6b12f5f0fc088a4641e81","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2797812159997193,"synthesis_s":0.520756395000717,"total_s":1.8005376110004363,"rtf":0.30008960183340605,"output_peak":0.5670881630709146,"output_rms":0.08569986134923661,"finished_at":"2026-10-13T04:38:21.826498+00:00","evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"1ddb3a3b62087bf523be518d71810f644c9cf48e66e6b12f5f0fc088a4641e81","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.68713046610186,"spectral_convergence":0.16885602172786163,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.269515828495426,"spectral_convergence":0.1619800098359814,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.461858759410179,"spectral_convergence":0.16122821957355976,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.003822361262780755,"delta_db":0.022266465445674753,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.0024449057037053794,"delta_db":-0.08566986851241676,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.000669963414278338,"delta_db":-0.10178474027514267,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.0002318553145821338,"delta_db":-0.5444626016717746,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00011881013549613292,"delta_db":-0.30305271388298954,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":5.657040443564098e-05,"delta_db":0.1557085484485426,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20215748970609015,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04857622217227438,"pitch":null,"output_rms_dbfs":-21.340397614117173,"output_peak":0.5670881630709146,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.472835018002488},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.68713046610186,"spectral_convergence":0.16885602172786163,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.269515828495426,"spectral_convergence":0.1619800098359814,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.461858759410179,"spectral_convergence":0.16122821957355976,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.003822361262780755,"delta_db":0.022266465445674753,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.0024449057037053794,"delta_db":-0.08566986851241676,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.000669963414278338,"delta_db":-0.10178474027514267,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.0002318553145821338,"delta_db":-0.5444626016717746,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00011881013549613292,"delta_db":-0.30305271388298954,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":5.657040443564098e-05,"delta_db":0.1557085484485426,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20215748970609015,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04857622217227438,"pitch":null,"output_rms_dbfs":-21.340397614117173,"output_peak":0.5670881630709146,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.472835018002488},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9198096638080558,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618048431617668,"rms_dbfs":-21.2918213919449,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.105627225441197,"mean":-0.0005158704121907552,"zero_energy":false},"output":{"rms":0.08569986134923661,"rms_dbfs":-21.340397614117173,"peak":0.5670881630709146,"peak_dbfs":-4.926988354461145,"crest_factor_db":16.413409259656028,"mean":4.337271790717149e-05,"zero_energy":false},"rms_difference_db":-0.04857622217227231},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":3.4022476838685028e-06,"output_rms_dbfs":-54.68234072903996,"output_zero_energy":false,"output_peak":0.010118732265161318,"output_energy_fraction":2.4706112601447948e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":313,"both_estimated_voiced_frames":294,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.98,"absolute_difference_cents_median":1.7096425427361295,"absolute_difference_cents_p95":8.630712885689057,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025738076998093343,"log_rms_mae_db":0.7672573504940388,"rms_envelope_nrmse":0.05798327988023417,"envelope_correlation":0.9955385310760524,"rms_gain_error_db":-0.04830348257891203,"reference_mean_power":0.007432639167823318,"output_mean_power":0.007350429243899579},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.369590439777949e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014473681837183883,"output_power":0.0014129114247999055,"reference_active":true,"delta_db":-0.10464084815804095},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002579988221497507,"output_power":0.00023245344230003764,"reference_active":true,"delta_db":-0.4528174142801786},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2677922934851034e-05,"output_power":3.0709221200640866e-05,"reference_active":true,"delta_db":-0.2698564154762615}],"active_band_count":3,"active_band_mae_db":0.275771559304827}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022253035095363898,"log_rms_mae_db":0.4249474842096047,"rms_envelope_nrmse":0.03089000012339408,"envelope_correlation":0.9986114226582088,"rms_gain_error_db":-0.048307321585462526,"reference_mean_power":0.0074694638279480755,"output_mean_power":0.007386840069131451},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0676309697735546e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013241614802002638,"output_power":0.001289037587080707,"reference_active":true,"delta_db":-0.11675368979314449},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001197578791096237,"output_power":0.00011102477407445384,"reference_active":true,"delta_db":-0.328841980304379},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429382109480039e-06,"output_power":1.9571906462823913e-06,"reference_active":true,"delta_db":-0.9386269738874399}],"active_band_count":3,"active_band_mae_db":0.4614075479949878}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017894181312046162,"log_rms_mae_db":1.727009378886572,"rms_envelope_nrmse":0.09868646154278131,"envelope_correlation":0.9923204560081462,"rms_gain_error_db":-0.1398175845605524,"reference_mean_power":0.0017811195275832714,"output_mean_power":0.0017246910380846413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1758328243827778e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007391946250824912,"output_power":0.0007169447752381928,"reference_active":true,"delta_db":-0.1327309617389832},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015375544244981539,"output_power":0.00013398836827650744,"reference_active":true,"delta_db":-0.5976339908124882},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532547306567537e-05,"output_power":9.794420072552326e-06,"reference_active":true,"delta_db":-0.31554691898689435}],"active_band_count":3,"active_band_mae_db":0.3486372905127886}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016320284141686103,"log_rms_mae_db":1.1100750787027531,"rms_envelope_nrmse":0.06615100814137492,"envelope_correlation":0.9963344426524848,"rms_gain_error_db":-0.1398275589571982,"reference_mean_power":0.0017900337371132,"output_mean_power":0.001733318851429527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0791360125207298e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007228891058818755,"output_power":0.0006985221196699837,"reference_active":true,"delta_db":-0.1489151666674749},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980895739386467e-05,"output_power":6.182499255436399e-05,"reference_active":true,"delta_db":-0.5274707907764452},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.398569514368019e-07,"output_power":8.327944809982349e-07,"reference_active":true,"delta_db":-0.5252391952577838}],"active_band_count":3,"active_band_mae_db":0.4005417175672347}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":1.7414004844089235,"rms_envelope_nrmse":0.15232450187846802,"envelope_correlation":0.980942785329796,"rms_gain_error_db":-0.4974126141873936,"reference_mean_power":0.0003180382981236394,"output_mean_power":0.0002836208533562608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6089659140343427e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.507437744605699e-05,"output_power":8.020037402288384e-05,"reference_active":true,"delta_db":-0.25622386049442614},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3978123456671115e-05,"output_power":1.983970603630357e-05,"reference_active":true,"delta_db":-0.8227995897650741},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.655957629268445e-06,"output_power":3.6188390385426753e-06,"reference_active":true,"delta_db":-1.0943975277887437}],"active_band_count":3,"active_band_mae_db":0.7244736593494147}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":1.1356341008072424,"rms_envelope_nrmse":0.09665733966635678,"envelope_correlation":0.9934811307719619,"rms_gain_error_db":-0.49771023617519583,"reference_mean_power":0.0003195237933424113,"output_mean_power":0.00028492606471358124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4169476048907228e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.318339905927386e-05,"output_power":7.825984659162258e-05,"reference_active":true,"delta_db":-0.2649767039053443},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1060276940136074e-05,"output_power":9.249165196612573e-06,"reference_active":true,"delta_db":-0.776634651563239},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5597258507742536e-07,"output_power":2.1723783717812973e-07,"reference_active":true,"delta_db":-0.7125798412220405}],"active_band_count":3,"active_band_mae_db":0.5847303988968746}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":1.85324881118505,"rms_envelope_nrmse":0.16769433521257132,"envelope_correlation":0.9806170234065092,"rms_gain_error_db":0.01487309141891542,"reference_mean_power":8.301357548516478e-05,"output_mean_power":8.329835577171626e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.362198065588498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.2275109921127754e-05,"output_power":3.1138656539055034e-05,"reference_active":true,"delta_db":-0.15567858812235852},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8777051155174135e-06,"output_power":2.255828673414873e-06,"reference_active":true,"delta_db":-1.0574017622212293},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.623133887825749e-07,"output_power":7.215891073299205e-07,"reference_active":true,"delta_db":0.3722643426255987}],"active_band_count":3,"active_band_mae_db":0.5284482309897288}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":1.2293411764072395,"rms_envelope_nrmse":0.09759217131243968,"envelope_correlation":0.9930842323369675,"rms_gain_error_db":0.013962644843349924,"reference_mean_power":8.336414632310506e-05,"output_mean_power":8.3632594838519e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1059391757915856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.094184470310103e-05,"output_power":2.9936785288946516e-05,"reference_active":true,"delta_db":-0.1434103954246688},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2229001330453797e-06,"output_power":9.999736210143512e-07,"reference_active":true,"delta_db":-0.8740244864525448},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.748361881351689e-08,"output_power":4.867551036886374e-08,"reference_active":true,"delta_db":0.10766703926965435}],"active_band_count":3,"active_band_mae_db":0.3750339737156227}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0912601947784424,"predicted_mos":3.0912601947784424,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"5134196d38d260f3d8fef86b510ee8414f46571ccf304b374a319c2293b7a55a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9872483990002365},"started_at":"2026-10-13T04:41:14.317598+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.531700026040198e-05,"native_input_samples":96000,"native_input_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","output_sha256":"b4ba69832d535d3ef1dc212c6170a6b86c06ad633afa2a4e605ebf8b1c2ae8c4","output_samples":96000,"length_delta_samples":0,"analysis_s":2.6720698620001713,"synthesis_s":0.20040355099990848,"total_s":2.8724734130000797,"rtf":0.4787455688333466,"output_peak":0.7602949050445239,"output_rms":0.09373111581473971,"finished_at":"2026-10-13T04:41:17.200282+00:00","evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"b4ba69832d535d3ef1dc212c6170a6b86c06ad633afa2a4e605ebf8b1c2ae8c4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.225958192172653,"spectral_convergence":0.7175664517919512,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.151061148304087,"spectral_convergence":0.26129928078168824,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.473296686019283,"spectral_convergence":0.25009995093247595,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.004307004890952845,"delta_db":0.5407028035640395,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.003146908982739942,"delta_db":1.0105508154815603,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0008252007961354001,"delta_db":0.8033007710924758,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.00029804616746442156,"delta_db":0.5462022626595496,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00014675585397210233,"delta_db":0.6143667136242816,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":6.160538061143641e-05,"delta_db":0.5260021629936175,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6735209215692541,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7294971335797555,"pitch":null,"output_rms_dbfs":-20.56232425836514,"output_peak":0.7602949050445239,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.95010534216534},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.225958192172653,"spectral_convergence":0.7175664517919512,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.151061148304087,"spectral_convergence":0.26129928078168824,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.473296686019283,"spectral_convergence":0.25009995093247595,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.004307004890952845,"delta_db":0.5407028035640395,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.003146908982739942,"delta_db":1.0105508154815603,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0008252007961354001,"delta_db":0.8033007710924758,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.00029804616746442156,"delta_db":0.5462022626595496,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00014675585397210233,"delta_db":0.6143667136242816,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":6.160538061143641e-05,"delta_db":0.5260021629936175,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6735209215692541,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7294971335797555,"pitch":null,"output_rms_dbfs":-20.56232425836514,"output_peak":0.7602949050445239,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.95010534216534},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9399910659415494,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618048431617668,"rms_dbfs":-21.2918213919449,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.105627225441197,"mean":-0.0005158704121907552,"zero_energy":false},"output":{"rms":0.09373111581473971,"rms_dbfs":-20.56232425836514,"peak":0.7602949050445239,"peak_dbfs":-2.38035839672193,"crest_factor_db":18.18196586164321,"mean":2.265665245995021e-06,"zero_energy":false},"rms_difference_db":0.7294971335797591},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":5.033762008064461e-06,"output_rms_dbfs":-52.981073213716165,"output_zero_energy":false,"output_peak":0.02974433612811839,"output_energy_fraction":3.055792301247273e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":306,"both_estimated_voiced_frames":272,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9066666666666666,"absolute_difference_cents_median":5.541008647677112,"absolute_difference_cents_p95":25.444744556105977,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025738076998093343,"log_rms_mae_db":1.753656913424675,"rms_envelope_nrmse":0.16425446164975988,"envelope_correlation":0.9764993001596286,"rms_gain_error_db":0.7298625580027516,"reference_mean_power":0.007432639167823318,"output_mean_power":0.008792842731082945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.369590439777949e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014473681837183883,"output_power":0.0018419497057113742,"reference_active":true,"delta_db":1.0469874598342297},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002579988221497507,"output_power":0.00023995201295179164,"reference_active":true,"delta_db":-0.3149332569939619},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2677922934851034e-05,"output_power":3.2303714438331546e-05,"reference_active":true,"delta_db":-0.05001981735344528}],"active_band_count":3,"active_band_mae_db":0.47064684472721224}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022253035095363898,"log_rms_mae_db":1.2468848396867198,"rms_envelope_nrmse":0.11619726816366835,"envelope_correlation":0.9921412105386078,"rms_gain_error_db":0.730039603719438,"reference_mean_power":0.0074694638279480755,"output_mean_power":0.008836766690381921},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0676309697735546e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013241614802002638,"output_power":0.0016927960382041754,"reference_active":true,"delta_db":1.066636838181351},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001197578791096237,"output_power":0.00011152638543941586,"reference_active":true,"delta_db":-0.3092646914475417},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429382109480039e-06,"output_power":2.1926186992412262e-06,"reference_active":true,"delta_db":-0.44532715352594787}],"active_band_count":3,"active_band_mae_db":0.6070762277182801}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017894181312046162,"log_rms_mae_db":2.0771198114135205,"rms_envelope_nrmse":0.24484948698325212,"envelope_correlation":0.9738535514368359,"rms_gain_error_db":1.2044232232983685,"reference_mean_power":0.0017811195275832714,"output_mean_power":0.002350365412248334},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1758328243827778e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007391946250824912,"output_power":0.001094010796834804,"reference_active":true,"delta_db":1.7026280775634461},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015375544244981539,"output_power":0.00018727900911112293,"reference_active":true,"delta_db":0.8565860550054951},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532547306567537e-05,"output_power":9.599123609134458e-06,"reference_active":true,"delta_db":-0.4030183424581799}],"active_band_count":3,"active_band_mae_db":0.9874108250090404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016320284141686103,"log_rms_mae_db":1.5698725991315376,"rms_envelope_nrmse":0.198945067438452,"envelope_correlation":0.9869259959279374,"rms_gain_error_db":1.2044966436175784,"reference_mean_power":0.0017900337371132,"output_mean_power":0.0023621685372273433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0791360125207298e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007228891058818755,"output_power":0.0010658608918926726,"reference_active":true,"delta_db":1.6862884768510948},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980895739386467e-05,"output_power":8.429737278496735e-05,"reference_active":true,"delta_db":0.8190288778926802},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.398569514368019e-07,"output_power":7.59293994194791e-07,"reference_active":true,"delta_db":-0.926517931879029}],"active_band_count":3,"active_band_mae_db":1.1439450955409347}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":1.704140920283877,"rms_envelope_nrmse":0.21545180403599806,"envelope_correlation":0.9650340174823795,"rms_gain_error_db":0.5864341680895266,"reference_mean_power":0.0003180382981236394,"output_mean_power":0.0003640179818946871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6089659140343427e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.507437744605699e-05,"output_power":0.00010307300199641036,"reference_active":true,"delta_db":0.8334614532053365},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3978123456671115e-05,"output_power":2.828880257403278e-05,"reference_active":true,"delta_db":0.7179937283218251},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.655957629268445e-06,"output_power":5.278626408166731e-06,"reference_active":true,"delta_db":0.5451190670090109}],"active_band_count":3,"active_band_mae_db":0.6988580828453909}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":1.0647956389573157,"rms_envelope_nrmse":0.12764672418557285,"envelope_correlation":0.989320492897365,"rms_gain_error_db":0.5874001531898645,"reference_mean_power":0.0003195237933424113,"output_mean_power":0.0003657995936985014},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4169476048907228e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.318339905927386e-05,"output_power":0.00010206863741428597,"reference_active":true,"delta_db":0.8885565447551222},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1060276940136074e-05,"output_power":1.3236528618576927e-05,"reference_active":true,"delta_db":0.7800810149106696},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5597258507742536e-07,"output_power":2.7454067927826414e-07,"reference_active":true,"delta_db":0.3041324946959629}],"active_band_count":3,"active_band_mae_db":0.6575900181205849}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":1.9622863929716452,"rms_envelope_nrmse":0.2195080949813343,"envelope_correlation":0.9725826735240258,"rms_gain_error_db":0.6210288150170227,"reference_mean_power":8.301357548516478e-05,"output_mean_power":9.577496485780464e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.362198065588498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.2275109921127754e-05,"output_power":3.8409459114726235e-05,"reference_active":true,"delta_db":0.7557046139463874},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8777051155174135e-06,"output_power":2.8355678768704254e-06,"reference_active":true,"delta_db":-0.06406241071852478},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.623133887825749e-07,"output_power":1.2955262638494445e-06,"reference_active":true,"delta_db":2.9138268726870735}],"active_band_count":3,"active_band_mae_db":1.2445312991173285}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":1.2626837160286792,"rms_envelope_nrmse":0.1338611983194932,"envelope_correlation":0.9916422908809381,"rms_gain_error_db":0.6235589007257651,"reference_mean_power":8.336414632310506e-05,"output_mean_power":9.623547567181673e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1059391757915856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.094184470310103e-05,"output_power":3.7282559776511575e-05,"reference_active":true,"delta_db":0.809595207768101},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2229001330453797e-06,"output_power":1.190492779767542e-06,"reference_active":true,"delta_db":-0.11664226454630294},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.748361881351689e-08,"output_power":7.918429818225974e-08,"reference_active":true,"delta_db":2.220952619602569}],"active_band_count":3,"active_band_mae_db":1.0490633639723244}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"facodec","status":"measured","mos":{"status":"measured","value":3.3346261978149414,"predicted_mos":3.3346261978149414,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2381944d1c39f2b0e241099186b1f5a4da56058eeceafe8e83e54fc0546530ab","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.049876311999469},"started_at":"2026-10-13T04:46:55.044033+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4090999431791715e-05,"native_input_samples":96000,"native_input_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","output_sha256":"938a8cf2d03da10e485f3f8709055ae132ff0799aec7a390ba7a478069dd5a91","output_samples":96000,"length_delta_samples":0,"analysis_s":2.025261763999879,"synthesis_s":6.335008218999974,"total_s":8.360269982999853,"rtf":1.3933783304999754,"output_peak":0.4866059124469757,"output_rms":0.08406433454313128,"finished_at":"2026-10-13T04:47:03.412653+00:00","evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"938a8cf2d03da10e485f3f8709055ae132ff0799aec7a390ba7a478069dd5a91","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.210226286533513,"spectral_convergence":0.18976303580606127,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.295118320067244,"spectral_convergence":0.1867876619129468,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.413925250778088,"spectral_convergence":0.19582960617144882,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.0036058938090492284,"delta_db":-0.2309215166503746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.0024395103025296816,"delta_db":-0.09526443993428056,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0006183657954729349,"delta_db":-0.4498410234964506,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.0002315572989686778,"delta_db":-0.5500484036332404,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00012811891250128353,"delta_db":0.024544810720786816,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":4.336622365769011e-05,"delta_db":-0.9986682346790174,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.391548071519025,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21594301942749386,"pitch":null,"output_rms_dbfs":-21.507764411372392,"output_peak":0.4866059124469757,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.306423285792949},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.210226286533513,"spectral_convergence":0.18976303580606127,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.295118320067244,"spectral_convergence":0.1867876619129468,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.413925250778088,"spectral_convergence":0.19582960617144882,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.0036058938090492284,"delta_db":-0.2309215166503746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.0024395103025296816,"delta_db":-0.09526443993428056,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0006183657954729349,"delta_db":-0.4498410234964506,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.0002315572989686778,"delta_db":-0.5500484036332404,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00012811891250128353,"delta_db":0.024544810720786816,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":4.336622365769011e-05,"delta_db":-0.9986682346790174,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.391548071519025,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21594301942749386,"pitch":null,"output_rms_dbfs":-21.507764411372392,"output_peak":0.4866059124469757,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.306423285792949},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9289859737613789,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618048431617668,"rms_dbfs":-21.2918213919449,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.105627225441197,"mean":-0.0005158704121907552,"zero_energy":false},"output":{"rms":0.0840643345431313,"rms_dbfs":-21.507764411372392,"peak":0.4866059124469757,"peak_dbfs":-6.256452370599012,"crest_factor_db":15.25131204077338,"mean":-1.5322424780284413e-05,"zero_energy":false},"rms_difference_db":-0.21594301942749183},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":4.033646967091786e-06,"output_rms_dbfs":-53.94302114787909,"output_zero_energy":false,"output_peak":0.03402245417237282,"output_energy_fraction":3.0441991074372718e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":299,"both_estimated_voiced_frames":288,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.96,"absolute_difference_cents_median":3.2253389126282204,"absolute_difference_cents_p95":13.829322627361545,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025738076998093343,"log_rms_mae_db":1.0772583328174894,"rms_envelope_nrmse":0.09545546393780814,"envelope_correlation":0.9885903055688601,"rms_gain_error_db":-0.2158784519013565,"reference_mean_power":0.007432639167823318,"output_mean_power":0.007072210919057605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.369590439777949e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014473681837183883,"output_power":0.001420526556189805,"reference_active":true,"delta_db":-0.08129664562625893},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002579988221497507,"output_power":0.0002396636783518837,"reference_active":true,"delta_db":-0.32015502696679826},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2677922934851034e-05,"output_power":3.8635829569529293e-05,"reference_active":true,"delta_db":0.7273579740742786}],"active_band_count":3,"active_band_mae_db":0.3762698822224453}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022253035095363898,"log_rms_mae_db":0.896945646786571,"rms_envelope_nrmse":0.08900834827335216,"envelope_correlation":0.9890185748747171,"rms_gain_error_db":-0.21581968194182707,"reference_mean_power":0.0074694638279480755,"output_mean_power":0.007107346032396006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0676309697735546e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013241614802002638,"output_power":0.0013002890402107874,"reference_active":true,"delta_db":-0.07901048110254243},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001197578791096237,"output_power":0.00011198499602405257,"reference_active":true,"delta_db":-0.2914425721034732},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429382109480039e-06,"output_power":2.931691872443727e-06,"reference_active":true,"delta_db":0.8162249409814247}],"active_band_count":3,"active_band_mae_db":0.39555933139581345}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017894181312046162,"log_rms_mae_db":1.5459502427291885,"rms_envelope_nrmse":0.09742125988061352,"envelope_correlation":0.9926637166760348,"rms_gain_error_db":-0.19288458270486572,"reference_mean_power":0.0017811195275832714,"output_mean_power":0.0017037450492001264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1758328243827778e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007391946250824912,"output_power":0.0007234838406847582,"reference_active":true,"delta_db":-0.093299649173832},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015375544244981539,"output_power":0.00014616204636725924,"reference_active":true,"delta_db":-0.21995882549308846},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532547306567537e-05,"output_power":1.153160139403974e-05,"reference_active":true,"delta_db":0.3935620355659028}],"active_band_count":3,"active_band_mae_db":0.23560683674427443}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016320284141686103,"log_rms_mae_db":1.174977370770272,"rms_envelope_nrmse":0.0821186836269046,"envelope_correlation":0.9944125660952622,"rms_gain_error_db":-0.19292828261494313,"reference_mean_power":0.0017900337371132,"output_mean_power":0.0017122547829576064},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0791360125207298e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007228891058818755,"output_power":0.000707603797752087,"reference_active":true,"delta_db":-0.09281524742212743},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980895739386467e-05,"output_power":6.71961729393131e-05,"reference_active":true,"delta_db":-0.1656661266345297},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.398569514368019e-07,"output_power":1.065726230597411e-06,"reference_active":true,"delta_db":0.5458389721900591}],"active_band_count":3,"active_band_mae_db":0.26810678208223876}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":1.46632413715072,"rms_envelope_nrmse":0.133270376358936,"envelope_correlation":0.9850373051446718,"rms_gain_error_db":-0.37678727521152156,"reference_mean_power":0.0003180382981236394,"output_mean_power":0.00029160884536290154},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6089659140343427e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.507437744605699e-05,"output_power":8.563800549108039e-05,"reference_active":true,"delta_db":0.028677643065122806},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3978123456671115e-05,"output_power":2.4351000221434725e-05,"reference_active":true,"delta_db":0.06701612708124918},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.655957629268445e-06,"output_power":4.984405498842095e-06,"reference_active":true,"delta_db":0.2960434702678803}],"active_band_count":3,"active_band_mae_db":0.13057908013808409}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":1.1657547564889317,"rms_envelope_nrmse":0.10382387760467203,"envelope_correlation":0.9907092209764256,"rms_gain_error_db":-0.3768817931654673,"reference_mean_power":0.0003195237933424113,"output_mean_power":0.0002929645177116705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4169476048907228e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.318339905927386e-05,"output_power":8.336408466125317e-05,"reference_active":true,"delta_db":0.009423233414560195},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1060276940136074e-05,"output_power":1.1432771491574518e-05,"reference_active":true,"delta_db":0.14385521795702233},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5597258507742536e-07,"output_power":3.0601326048376317e-07,"reference_active":true,"delta_db":0.7754679176341297}],"active_band_count":3,"active_band_mae_db":0.3095821230019041}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":1.7307424325111975,"rms_envelope_nrmse":0.1890101711322066,"envelope_correlation":0.9761336531175971,"rms_gain_error_db":-0.4985847049659683,"reference_mean_power":8.301357548516478e-05,"output_mean_power":7.401004176755627e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.362198065588498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.2275109921127754e-05,"output_power":3.593525060284247e-05,"reference_active":true,"delta_db":0.4665294781535634},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8777051155174135e-06,"output_power":3.360005482131625e-06,"reference_active":true,"delta_db":0.6729369716482336},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.623133887825749e-07,"output_power":7.575807590141564e-07,"reference_active":true,"delta_db":0.5836540147728091}],"active_band_count":3,"active_band_mae_db":0.5743734881915353}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":1.2937872425133699,"rms_envelope_nrmse":0.15463713225751047,"envelope_correlation":0.9841678936874614,"rms_gain_error_db":-0.49868493305677786,"reference_mean_power":8.336414632310506e-05,"output_mean_power":7.432087496910214e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1059391757915856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.094184470310103e-05,"output_power":3.456611911503312e-05,"reference_active":true,"delta_db":0.4810442011897258},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2229001330453797e-06,"output_power":1.4482414275872166e-06,"reference_active":true,"delta_db":0.7344997426192164},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.748361881351689e-08,"output_power":5.14296405851459e-08,"reference_active":true,"delta_db":0.3466967935337015}],"active_band_count":3,"active_band_mae_db":0.5207469124475479}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.175039052963257,"predicted_mos":3.175039052963257,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"0dceec58f3893749b7eac28b0ba6dd6026b4e32c0d34aa4d41e09d7a54a29f5e","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.072459125000023},"started_at":"2026-10-13T05:00:30.133648+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.010658211999725609,"native_input_samples":264600,"native_input_sha256":"0a5bdcb342da5a64fca00b23e3d296749f2d343d9f7cb2070e8acf61a242a4ea","output_sha256":"8fe3e6b9f9aa5745667c89c1aec5b49eae731b0ca4da68e8225bff28d265d9ff","output_samples":264704,"length_delta_samples":104,"analysis_s":0.14559509399987292,"synthesis_s":5.776225525999507,"total_s":5.92182061999938,"rtf":0.98697010333323,"output_peak":0.6352872252464294,"output_rms":0.07857297304294696,"finished_at":"2026-10-13T05:00:36.079535+00:00","evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"8fe3e6b9f9aa5745667c89c1aec5b49eae731b0ca4da68e8225bff28d265d9ff","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.090886721265935,"output_peak":0.6342759104250363},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.035879343274582,"spectral_convergence":0.45895812547809084,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.842525694284197,"spectral_convergence":0.32428537248926814,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.746810493220415,"spectral_convergence":0.3542995969180119,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.0023215250789494135,"delta_db":-2.143317049631285,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.002819618413219144,"delta_db":0.5336123746295176,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0007181395252353682,"delta_db":0.19979269059537252,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.00017995945654475823,"delta_db":-1.644886424210873,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00010404285393861573,"delta_db":-0.8794650599282526,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":3.785062700427696e-05,"delta_db":-1.5894535070108593,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1650878510010267,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7974956474485306,"pitch":null,"output_rms_dbfs":-22.089317039393425,"output_peak":0.6342759104250363,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.875071843593065},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7992527106648731,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618048431617668,"rms_dbfs":-21.2918213919449,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.105627225441197,"mean":-0.0005158704121907552,"zero_energy":false},"output":{"rms":0.07862020067954277,"rms_dbfs":-22.089317039393425,"peak":0.6342759104250363,"peak_dbfs":-3.9544356528057807,"crest_factor_db":18.134881386587647,"mean":-0.00023821812982453345,"zero_energy":false},"rms_difference_db":-0.7974956474485246},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":3.2411716063237865e-06,"output_rms_dbfs":-54.89297974303648,"output_zero_energy":false,"output_peak":0.021828480430549583,"output_energy_fraction":2.7966135501970866e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":249,"both_estimated_voiced_frames":231,"reference_voiced_output_unvoiced_frames":69,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.77,"absolute_difference_cents_median":8.416566903739387,"absolute_difference_cents_p95":40.49516998378688,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025738076998093343,"log_rms_mae_db":1.5142318418873055,"rms_envelope_nrmse":0.12055742951455294,"envelope_correlation":0.9907608491576567,"rms_gain_error_db":-0.7972690175725459,"reference_mean_power":0.007432639167823318,"output_mean_power":0.006186088770868435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.369590439777949e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014473681837183883,"output_power":0.001256572303161778,"reference_active":true,"delta_db":-0.6139153871428853},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002579988221497507,"output_power":0.00020453608678664415,"reference_active":true,"delta_db":-1.0084778056077552},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2677922934851034e-05,"output_power":2.3375465047714004e-05,"reference_active":true,"delta_db":-1.4549418444510065}],"active_band_count":3,"active_band_mae_db":1.0257783457338823}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022253035095363898,"log_rms_mae_db":1.2443358906341122,"rms_envelope_nrmse":0.10286066019156546,"envelope_correlation":0.9959537108186264,"rms_gain_error_db":-0.7971211991985351,"reference_mean_power":0.0074694638279480755,"output_mean_power":0.0062169490550598476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0676309697735546e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013241614802002638,"output_power":0.0011538832876423059,"reference_active":true,"delta_db":-0.5977906686183239},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001197578791096237,"output_power":9.276159843433762e-05,"reference_active":true,"delta_db":-1.1093587249041075},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429382109480039e-06,"output_power":1.8010529449726238e-06,"reference_active":true,"delta_db":-1.299693490989052}],"active_band_count":3,"active_band_mae_db":1.0022809615038277}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017894181312046162,"log_rms_mae_db":2.6758996047323342,"rms_envelope_nrmse":0.20626175493874374,"envelope_correlation":0.9722412786426821,"rms_gain_error_db":0.5972385193542379,"reference_mean_power":0.0017811195275832714,"output_mean_power":0.0020436989276910616},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1758328243827778e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007391946250824912,"output_power":0.0007390127930150695,"reference_active":true,"delta_db":-0.0010684382035405335},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015375544244981539,"output_power":0.00014222793389917713,"reference_active":true,"delta_db":-0.338455961281081},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532547306567537e-05,"output_power":1.0499128054455608e-05,"reference_active":true,"delta_db":-0.013801857196596433}],"active_band_count":3,"active_band_mae_db":0.11777541889373933}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016320284141686103,"log_rms_mae_db":2.012690245477434,"rms_envelope_nrmse":0.17602937853074888,"envelope_correlation":0.9790239371158116,"rms_gain_error_db":0.5972511202905779,"reference_mean_power":0.0017900337371132,"output_mean_power":0.0020539332632422184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0791360125207298e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007228891058818755,"output_power":0.0007149044113655432,"reference_active":true,"delta_db":-0.04823702953155574},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980895739386467e-05,"output_power":6.241667603548931e-05,"reference_active":true,"delta_db":-0.4861051496852972},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.398569514368019e-07,"output_power":1.0842133684924158e-06,"reference_active":true,"delta_db":0.6205299995493317}],"active_band_count":3,"active_band_mae_db":0.38495739292206155}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":2.8652449262788164,"rms_envelope_nrmse":0.2562971481816651,"envelope_correlation":0.9523509156222778,"rms_gain_error_db":-1.3186780436361314,"reference_mean_power":0.0003180382981236394,"output_mean_power":0.0002347532515958806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6089659140343427e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.507437744605699e-05,"output_power":6.743673743996109e-05,"reference_active":true,"delta_db":-1.0090222856964908},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3978123456671115e-05,"output_power":1.8759608591723296e-05,"reference_active":true,"delta_db":-1.0659141909102166},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.655957629268445e-06,"output_power":3.818599304499975e-06,"reference_active":true,"delta_db":-0.8610493031747445}],"active_band_count":3,"active_band_mae_db":0.9786619265938171}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":2.36545732553565,"rms_envelope_nrmse":0.22214711880187935,"envelope_correlation":0.9655573632990563,"rms_gain_error_db":-1.31803763472889,"reference_mean_power":0.0003195237933424113,"output_mean_power":0.00023588451938487656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4169476048907228e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.318339905927386e-05,"output_power":6.629704556554753e-05,"reference_active":true,"delta_db":-0.9854248751883532},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1060276940136074e-05,"output_power":8.426478093667256e-06,"reference_active":true,"delta_db":-1.181199053916935},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5597258507742536e-07,"output_power":2.428758516564955e-07,"reference_active":true,"delta_db":-0.22809117938388554}],"active_band_count":3,"active_band_mae_db":0.7982383694963913}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":3.1674083426103,"rms_envelope_nrmse":0.23746381692027618,"envelope_correlation":0.9758458163405065,"rms_gain_error_db":-1.5562310664627923,"reference_mean_power":8.301357548516478e-05,"output_mean_power":5.801309198739377e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.362198065588498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.2275109921127754e-05,"output_power":2.5164864063150736e-05,"reference_active":true,"delta_db":-1.08073141296366},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8777051155174135e-06,"output_power":2.0213043455877807e-06,"reference_active":true,"delta_db":-1.5341457912900855},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.623133887825749e-07,"output_power":3.5723842032440046e-07,"reference_active":true,"delta_db":-2.6810537418743055}],"active_band_count":3,"active_band_mae_db":1.765310315376017}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":2.7839137525925017,"rms_envelope_nrmse":0.19818281001337792,"envelope_correlation":0.9892838691671265,"rms_gain_error_db":-1.553869777630368,"reference_mean_power":8.336414632310506e-05,"output_mean_power":5.828976835766242e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1059391757915856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.094184470310103e-05,"output_power":2.404774620373022e-05,"reference_active":true,"delta_db":-1.094718222860454},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2229001330453797e-06,"output_power":8.413995194972271e-07,"reference_active":true,"delta_db":-1.6238873258019697},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.748361881351689e-08,"output_power":2.7021244173709893e-08,"reference_active":true,"delta_db":-2.4483846799862445}],"active_band_count":3,"active_band_mae_db":1.7223300762162228}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"golf","status":"measured","mos":{"status":"measured","value":3.223958730697632,"predicted_mos":3.223958730697632,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"a5a7b7b67b5d89e0b81e1f6e73944994c3bb8d29f160fcffa392cf2ef856d63c","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6538501899995026},"started_at":"2026-10-13T05:02:01.133133+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030778710006416077,"native_input_samples":144000,"native_input_sha256":"396336e666c918a83951c8524a6548f52db9930f0aac9aaa7bc6bce3e691e00c","output_sha256":"4dfb979ecd4aaacd7f9994b6443d41c62d75247024e278e4cf8a18aa5d9c423f","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6122165010001481,"synthesis_s":0.12311149400011345,"total_s":0.7353279950002616,"rtf":0.12255466583337693,"output_peak":0.5537991523742676,"output_rms":0.07779507502847617,"finished_at":"2026-10-13T05:02:01.880358+00:00","analysis_stages":{"external_f0_s":0.08138591300030384},"evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"4dfb979ecd4aaacd7f9994b6443d41c62d75247024e278e4cf8a18aa5d9c423f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.178396660920484,"output_peak":0.553892725507934},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.011743747075066,"spectral_convergence":0.6254194779177613,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.969530569167087,"spectral_convergence":0.24233854626309614,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.285916507183294,"spectral_convergence":0.22854123926810377,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038085829443051363,"output_power":0.003006660582354006,"delta_db":-1.0267901423013197,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024975974400892457,"output_power":0.0022204941656313914,"delta_db":-0.5107280419024992,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006869823799926772,"output_power":0.000503089713205025,"delta_db":-1.3530016094445696,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000263174775453501,"output_power":0.0001974426037785365,"delta_db":-1.248033913990443,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012721773490741758,"output_power":8.805880317897528e-05,"delta_db":-1.597748801625002,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.4647853530650685e-05,"output_power":3.989810433998979e-05,"delta_db":-1.366208465690693,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1837518291590878,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8930770033091512,"pitch":null,"output_rms_dbfs":-22.178396660920484,"output_peak":0.553892725507934,"reference_rms_dbfs":-21.285319657611335,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.755730274475149},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9163269510976958,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08624501799106213,"rms_dbfs":-21.285319657611335,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.099125491107635,"mean":-0.0005035814339410085,"zero_energy":false},"output":{"rms":0.07781801830224105,"rms_dbfs":-22.178396660920484,"peak":0.553892725507934,"peak_dbfs":-5.1314867715385715,"crest_factor_db":17.046909889381915,"mean":0.0009509826195693321,"zero_energy":false},"rms_difference_db":-0.8930770033091484},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":2.9353055614584026e-06,"output_rms_dbfs":-55.32346682577223,"output_zero_energy":false,"output_peak":0.007876949936815225,"output_energy_fraction":2.589501336155842e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":337,"both_estimated_voiced_frames":297,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":40,"paired_voiced_fraction":0.99,"absolute_difference_cents_median":6.7564752430887856,"absolute_difference_cents_p95":22.149278637807104,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00025738076998093343,"log_rms_mae_db":1.4942340764436213,"rms_envelope_nrmse":0.15031821313657978,"envelope_correlation":0.9807089580167063,"rms_gain_error_db":-0.893072152699217,"reference_mean_power":0.007444405365881566,"output_mean_power":0.0060607001523842515},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.3703791723524877e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014491302222405882,"output_power":0.0010875604322961522,"reference_active":true,"delta_db":-1.2465401527652553},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025632567516355295,"output_power":0.0002091388135456654,"reference_active":true,"delta_db":-0.8835747006524175},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.24954246099332e-05,"output_power":2.508471872887853e-05,"reference_active":true,"delta_db":-1.1241298044883186}],"active_band_count":3,"active_band_mae_db":1.084748219301997}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00022253035095363898,"log_rms_mae_db":1.1478718994250674,"rms_envelope_nrmse":0.11737289125125365,"envelope_correlation":0.9929523849656655,"rms_gain_error_db":-0.8930731841114576,"reference_mean_power":0.007481786267407753,"output_mean_power":0.006091131550921265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.0684095991885964e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013251979183713047,"output_power":0.0009947843665703618,"reference_active":true,"delta_db":-1.2455179347609495},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011842019261661955,"output_power":9.742159193152234e-05,"reference_active":true,"delta_db":-0.8477054120405867},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429550099490576e-06,"output_power":2.089888448173484e-06,"reference_active":true,"delta_db":-0.654027536068888}],"active_band_count":3,"active_band_mae_db":0.9157502942901413}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017894181312046162,"log_rms_mae_db":1.8359986183613823,"rms_envelope_nrmse":0.18033540303462126,"envelope_correlation":0.9800475780432684,"rms_gain_error_db":-0.9172967948809695,"reference_mean_power":0.0017840456057423013,"output_mean_power":0.0014443627271501654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1758314511200167e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007401665286049699,"output_power":0.0005369287233331544,"reference_active":true,"delta_db":-1.3941280434620185},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001523075912263928,"output_power":0.00012387116495391634,"reference_active":true,"delta_db":-0.8975132776602867},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0517872602325305e-05,"output_power":8.401256072509076e-06,"reference_active":true,"delta_db":-0.9758368379854878}],"active_band_count":3,"active_band_mae_db":1.0891593863692643}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016320284141686103,"log_rms_mae_db":1.2340436556745125,"rms_envelope_nrmse":0.13434909933720998,"envelope_correlation":0.9923917403377059,"rms_gain_error_db":-0.9172894841484992,"reference_mean_power":0.0017930270905362803,"output_mean_power":0.0014516365778651512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0792145937749595e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007232927091453493,"output_power":0.0005240437371205421,"reference_active":true,"delta_db":-1.3994655232187612},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.895227774578327e-05,"output_power":5.649315818759381e-05,"reference_active":true,"delta_db":-0.8655276291490905},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.370702620107776e-07,"output_power":7.711111280020938e-07,"reference_active":true,"delta_db":-0.8465518520391081}],"active_band_count":3,"active_band_mae_db":1.0371816681356532}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":2.017708597613746,"rms_envelope_nrmse":0.2333414668945787,"envelope_correlation":0.9652629460634549,"rms_gain_error_db":-1.3633927753235269,"reference_mean_power":0.0003184407642615612,"output_mean_power":0.00023264267348033042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.6098478477258045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.523795201067637e-05,"output_power":6.011034496685448e-05,"reference_active":true,"delta_db":-1.5168378614548002},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3787638487307002e-05,"output_power":1.7945291033103722e-05,"reference_active":true,"delta_db":-1.224008236594404},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.667219904763696e-06,"output_power":3.5793186768861046e-06,"reference_active":true,"delta_db":-1.1525789738113288}],"active_band_count":3,"active_band_mae_db":1.2978083572868444}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":1.6137744261884979,"rms_envelope_nrmse":0.17068992520131177,"envelope_correlation":0.9906834817494355,"rms_gain_error_db":-1.363327433043793,"reference_mean_power":0.0003200279538265462,"output_mean_power":0.00023380574131679174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.4177148174005372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.331135676530876e-05,"output_power":5.955515189657863e-05,"reference_active":true,"delta_db":-1.4578487056292733},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0898811813759945e-05,"output_power":8.244355027028142e-06,"reference_active":true,"delta_db":-1.2122246827298992},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.547775140080932e-07,"output_power":2.0942935778447293e-07,"reference_active":true,"delta_db":-0.8512353470504943}],"active_band_count":3,"active_band_mae_db":1.173769578469889}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":2.3730920749933513,"rms_envelope_nrmse":0.21142421542238948,"envelope_correlation":0.9788415503920479,"rms_gain_error_db":-1.2309124109157923,"reference_mean_power":8.306716977639144e-05,"output_mean_power":6.25659686258366e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.380731089636223e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.239020736466819e-05,"output_power":2.7451879027360306e-05,"reference_active":true,"delta_db":-0.718416519963569},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8786831404162764e-06,"output_power":2.2810928090570804e-06,"reference_active":true,"delta_db":-1.0105090883356795},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.640278712805448e-07,"output_power":7.924907629515216e-07,"reference_active":true,"delta_db":0.7680790048729659}],"active_band_count":3,"active_band_mae_db":0.8323348710574048}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":1.994853287492681,"rms_envelope_nrmse":0.17493610585145244,"envelope_correlation":0.989273779005869,"rms_gain_error_db":-1.2308501776274496,"reference_mean_power":8.348274921267069e-05,"output_mean_power":6.287988298452035e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.1236070332441925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.104968560288019e-05,"output_power":2.6470135950730866e-05,"reference_active":true,"delta_db":-0.6930103519135309},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2220524386988854e-06,"output_power":9.58953250304216e-07,"reference_active":true,"delta_db":-1.0529240653249514},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7430267514382606e-08,"output_power":4.7242746722973676e-08,"reference_active":true,"delta_db":-0.017204343203732247}],"active_band_count":3,"active_band_mae_db":0.5877129201474048}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.0609278678894043,"predicted_mos":3.0609278678894043,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"856d61a8b4b3623c963263062f5f8e570101df1039c884fd7fd49166470419b1","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3199267410000175},"started_at":"2026-10-13T05:05:47.457714+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0037691669995183474,"native_input_samples":144000,"native_input_sha256":"396336e666c918a83951c8524a6548f52db9930f0aac9aaa7bc6bce3e691e00c","output_sha256":"e8854b1fcb3e7c2b943d0300593c40530ef276c53d7e8ab7ba46f2e6b88b6008","output_samples":144120,"length_delta_samples":120,"analysis_s":2.417740505999973,"synthesis_s":3.692355111999859,"total_s":6.110095617999832,"rtf":1.0183492696666387,"output_peak":0.5086508989334106,"output_rms":0.09384690162996444,"finished_at":"2026-10-13T05:05:53.580976+00:00","analysis_stages":{"external_f0_s":1.4879557140002362},"evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"e8854b1fcb3e7c2b943d0300593c40530ef276c53d7e8ab7ba46f2e6b88b6008","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.54940647427334,"output_peak":0.5091097256262084},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.87969939007694,"spectral_convergence":0.7426049476708779,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.012114370161787,"spectral_convergence":0.30598740127220064,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.081692712761887,"spectral_convergence":0.27162080776444103,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.0054084902537335795,"delta_db":1.5297096684927214,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.0023960294379824017,"delta_db":-0.17336950998055325,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0006086730868437179,"delta_db":-0.5184546182487606,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.0002366869782590083,"delta_db":-0.4548895027229729,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.00012206398289450571,"delta_db":-0.1857122586649489,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":4.644863522077914e-05,"delta_db":-0.700454709450235,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5937650445933653,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7457155807552779,"pitch":null,"output_rms_dbfs":-20.546105811189616,"output_peak":0.5091097256262084,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.6578354910002036},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9065905327653151,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618048431617668,"rms_dbfs":-21.2918213919449,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.105627225441197,"mean":-0.0005158704121907552,"zero_energy":false},"output":{"rms":0.09390629571511164,"rms_dbfs":-20.546105811189616,"peak":0.5091097256262084,"peak_dbfs":-5.863772129359258,"crest_factor_db":14.68233368183036,"mean":4.769042784148117e-05,"zero_energy":false},"rms_difference_db":0.7457155807552844},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":3.51716270956646e-06,"output_rms_dbfs":-54.538075400616954,"output_zero_energy":false,"output_peak":0.02982137038439539,"output_energy_fraction":2.1271678918492433e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":346,"both_estimated_voiced_frames":296,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":50,"paired_voiced_fraction":0.9866666666666667,"absolute_difference_cents_median":7.3402890687505415,"absolute_difference_cents_p95":30.57643478169712,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025738076998093343,"log_rms_mae_db":2.2233358691444707,"rms_envelope_nrmse":0.18456401365442468,"envelope_correlation":0.9679032414126106,"rms_gain_error_db":0.7457397545572522,"reference_mean_power":0.007432639167823318,"output_mean_power":0.00882504696088816},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.369590439777949e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014473681837183883,"output_power":0.001669151989881004,"reference_active":true,"delta_db":0.6191686286560949},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002579988221497507,"output_power":0.00024977937541830684,"reference_active":true,"delta_db":-0.14061147965427273},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2677922934851034e-05,"output_power":2.8200165746932707e-05,"reference_active":true,"delta_db":-0.6400278333149563}],"active_band_count":3,"active_band_mae_db":0.4666026472084413}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022253035095363898,"log_rms_mae_db":1.7293328094741565,"rms_envelope_nrmse":0.1362932281910962,"envelope_correlation":0.9854359051156112,"rms_gain_error_db":0.7451363143017953,"reference_mean_power":0.0074694638279480755,"output_mean_power":0.008867538034076807},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0676309697735546e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013241614802002638,"output_power":0.001522074095887172,"reference_active":true,"delta_db":0.6049484465973636},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001197578791096237,"output_power":0.00011548640502780037,"reference_active":true,"delta_db":-0.15773233707005446},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429382109480039e-06,"output_power":2.2855626991177237e-06,"reference_active":true,"delta_db":-0.26502689269172297}],"active_band_count":3,"active_band_mae_db":0.342569225453047}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017894181312046162,"log_rms_mae_db":2.3623175651128574,"rms_envelope_nrmse":0.2065767310833591,"envelope_correlation":0.9656760242509311,"rms_gain_error_db":-0.31771348095287255,"reference_mean_power":0.0017811195275832714,"output_mean_power":0.001655471532541302},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1758328243827778e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007391946250824912,"output_power":0.0007001306271427262,"reference_active":true,"delta_db":-0.23579724096013877},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015375544244981539,"output_power":0.00013876372442679118,"reference_active":true,"delta_db":-0.4455454953537876},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532547306567537e-05,"output_power":9.213364399252658e-06,"reference_active":true,"delta_db":-0.5811517006861646}],"active_band_count":3,"active_band_mae_db":0.4208314790000303}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016320284141686103,"log_rms_mae_db":1.6864417409379588,"rms_envelope_nrmse":0.1352680030226085,"envelope_correlation":0.9845463517917955,"rms_gain_error_db":-0.31780339198880614,"reference_mean_power":0.0017900337371132,"output_mean_power":0.0016637224505370773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0791360125207298e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007228891058818755,"output_power":0.0006863168348191527,"reference_active":true,"delta_db":-0.2254702794291058},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980895739386467e-05,"output_power":6.370375244504602e-05,"reference_active":true,"delta_db":-0.39746136770218},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.398569514368019e-07,"output_power":8.547599019389807e-07,"reference_active":true,"delta_db":-0.41217617361901926}],"active_band_count":3,"active_band_mae_db":0.3450359402501017}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":2.278658580708639,"rms_envelope_nrmse":0.23060686551143542,"envelope_correlation":0.9514487418943117,"rms_gain_error_db":-0.32344256250485215,"reference_mean_power":0.0003180382981236394,"output_mean_power":0.0002952127868256502},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6089659140343427e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.507437744605699e-05,"output_power":7.8648808808958e-05,"reference_active":true,"delta_db":-0.34106630399940685},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3978123456671115e-05,"output_power":2.2236765551484817e-05,"reference_active":true,"delta_db":-0.32743574723693847},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.655957629268445e-06,"output_power":3.2596113426101114e-06,"reference_active":true,"delta_db":-1.5484319907065756}],"active_band_count":3,"active_band_mae_db":0.7389780139809736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":1.4935582242638674,"rms_envelope_nrmse":0.12887394505333147,"envelope_correlation":0.9840924323040148,"rms_gain_error_db":-0.32524165074818506,"reference_mean_power":0.0003195237933424113,"output_mean_power":0.0002964688292590241},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4169476048907228e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.318339905927386e-05,"output_power":7.643803102119762e-05,"reference_active":true,"delta_db":-0.3672717116316314},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1060276940136074e-05,"output_power":1.0240823778813315e-05,"reference_active":true,"delta_db":-0.3343110848967929},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5597258507742536e-07,"output_power":1.8706926429536573e-07,"reference_active":true,"delta_db":-1.3619101616789226}],"active_band_count":3,"active_band_mae_db":0.6878309860691156}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":2.153479609330608,"rms_envelope_nrmse":0.2124466349108295,"envelope_correlation":0.9696250763220673,"rms_gain_error_db":-0.5835041265687713,"reference_mean_power":8.301357548516478e-05,"output_mean_power":7.257694897117568e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.362198065588498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.2275109921127754e-05,"output_power":2.768387628163514e-05,"reference_active":true,"delta_db":-0.6664083031693168},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8777051155174135e-06,"output_power":2.523650358890688e-06,"reference_active":true,"delta_db":-0.5701710374181563},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.623133887825749e-07,"output_power":6.695930891226903e-07,"reference_active":true,"delta_db":0.04747428174848385}],"active_band_count":3,"active_band_mae_db":0.4280178741119857}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":1.4025214839262652,"rms_envelope_nrmse":0.14070500060201663,"envelope_correlation":0.9878825070886738,"rms_gain_error_db":-0.5895324051826463,"reference_mean_power":8.336414632310506e-05,"output_mean_power":7.278234877538959e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1059391757915856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.094184470310103e-05,"output_power":2.6576098621443844e-05,"reference_active":true,"delta_db":-0.6605497533794393},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2229001330453797e-06,"output_power":1.0121207217229707e-06,"reference_active":true,"delta_db":-0.8215867573924265},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.748361881351689e-08,"output_power":4.1473543982991636e-08,"reference_active":true,"delta_db":-0.5877266178922471}],"active_band_count":3,"active_band_mae_db":0.689954376221371}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.7541933059692383,"predicted_mos":2.7541933059692383,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"00a9aef7058a4cf26a38fff8069abd9d5d1b8d011278af12ffabe1cf999f9057","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4201328469989676},"started_at":"2026-10-13T05:10:08.155684+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002674699000635883,"native_input_samples":144000,"native_input_sha256":"396336e666c918a83951c8524a6548f52db9930f0aac9aaa7bc6bce3e691e00c","output_sha256":"1667ad3113356d42fb7b162b1cd631d89b9e6b95cf77f6686848fe7118508fe9","output_samples":144240,"length_delta_samples":240,"analysis_s":1.2277726040010748,"synthesis_s":0.1569130560001213,"total_s":1.384685660001196,"rtf":0.2307809433335327,"output_peak":0.7393411993980408,"output_rms":0.08552515509021813,"finished_at":"2026-10-13T05:10:09.555250+00:00","analysis_stages":{"external_f0_s":0.1660089839988359},"evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"1667ad3113356d42fb7b162b1cd631d89b9e6b95cf77f6686848fe7118508fe9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.355497101567288,"output_peak":0.739595507820407},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.598538408802267,"spectral_convergence":0.1875070320797893,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.20419458018718,"spectral_convergence":0.187313414195463,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.540583068680072,"spectral_convergence":0.233235582238665,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.003759492235361696,"delta_db":-0.049758923872173524,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.0024885742249338328,"delta_db":-0.008785018903057818,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0006761266845299301,"delta_db":-0.0620148471907166,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.00024154412140966268,"delta_db":-0.36666843110403824,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.0001160725196394127,"delta_db":-0.4042935058502111,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":4.934732775666013e-05,"delta_db":-0.43754788940631184,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22151143605441817,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05645333858161049,"pitch":null,"output_rms_dbfs":-21.34827473052651,"output_peak":0.739595507820407,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.447772019223173},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9567493802076645,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618048431617668,"rms_dbfs":-21.2918213919449,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.105627225441197,"mean":-0.0005158704121907552,"zero_energy":false},"output":{"rms":0.08562217652939684,"rms_dbfs":-21.34827473052651,"peak":0.739595507820407,"peak_dbfs":-2.6201147067400865,"crest_factor_db":18.72816002378642,"mean":0.00010445849694129513,"zero_energy":false},"rms_difference_db":-0.05645333858160839},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":4.072705687881583e-06,"output_rms_dbfs":-53.9011697286201,"output_zero_energy":false,"output_peak":0.016141419104433567,"output_energy_fraction":2.962847019285396e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":291,"both_estimated_voiced_frames":275,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9166666666666666,"absolute_difference_cents_median":1.5473508551701898,"absolute_difference_cents_p95":7.988077261449133,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025738076998093343,"log_rms_mae_db":0.7602678452876751,"rms_envelope_nrmse":0.08223324519281014,"envelope_correlation":0.9910745533123815,"rms_gain_error_db":-0.056430834443214964,"reference_mean_power":0.007432639167823318,"output_mean_power":0.0073366865729409605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.369590439777949e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014473681837183883,"output_power":0.0013946780394577503,"reference_active":true,"delta_db":-0.16105059089602194},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002579988221497507,"output_power":0.00022205179410973537,"reference_active":true,"delta_db":-0.6516343678051012},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2677922934851034e-05,"output_power":2.2823921096144644e-05,"reference_active":true,"delta_db":-1.5586418696533202}],"active_band_count":3,"active_band_mae_db":0.7904422761181479}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022253035095363898,"log_rms_mae_db":0.45651493069021265,"rms_envelope_nrmse":0.052049618337969096,"envelope_correlation":0.996115000729775,"rms_gain_error_db":-0.05648661332161546,"reference_mean_power":0.0074694638279480755,"output_mean_power":0.007372941145030565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0676309697735546e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013241614802002638,"output_power":0.0012765302101445087,"reference_active":true,"delta_db":-0.1590985292106441},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001197578791096237,"output_power":0.00010585049722820838,"reference_active":true,"delta_db":-0.5361119368630635},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429382109480039e-06,"output_power":1.4870000499556744e-06,"reference_active":true,"delta_db":-2.131848458088767}],"active_band_count":3,"active_band_mae_db":0.9423529747208249}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017894181312046162,"log_rms_mae_db":1.2275323470465502,"rms_envelope_nrmse":0.08300875165277152,"envelope_correlation":0.9947886328490744,"rms_gain_error_db":-0.19103304407667576,"reference_mean_power":0.0017811195275832714,"output_mean_power":0.0017044715659858126},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1758328243827778e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007391946250824912,"output_power":0.0007022237739745864,"reference_active":true,"delta_db":-0.22283271796604626},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015375544244981539,"output_power":0.0001332846386524386,"reference_active":true,"delta_db":-0.6205039844769418},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532547306567537e-05,"output_power":6.3163127828658235e-06,"reference_active":true,"delta_db":-2.220697903550702}],"active_band_count":3,"active_band_mae_db":1.0213448686645634}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016320284141686103,"log_rms_mae_db":0.7503167014558265,"rms_envelope_nrmse":0.050364423608168805,"envelope_correlation":0.9981567910192005,"rms_gain_error_db":-0.19105770461591617,"reference_mean_power":0.0017900337371132,"output_mean_power":0.0017129924381962807},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0791360125207298e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007228891058818755,"output_power":0.000685034074474555,"reference_active":true,"delta_db":-0.23359505475785067},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980895739386467e-05,"output_power":6.169271066027902e-05,"reference_active":true,"delta_db":-0.5367729907243688},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.398569514368019e-07,"output_power":4.951774464871784e-07,"reference_active":true,"delta_db":-2.783009019853581}],"active_band_count":3,"active_band_mae_db":1.1844590217786}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":1.245518512082289,"rms_envelope_nrmse":0.1031016073025689,"envelope_correlation":0.9917690125454933,"rms_gain_error_db":-0.3792690248685182,"reference_mean_power":0.0003180382981236394,"output_mean_power":0.0002914422548478388},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6089659140343427e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.507437744605699e-05,"output_power":8.089873340898029e-05,"reference_active":true,"delta_db":-0.21857057582402914},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3978123456671115e-05,"output_power":2.1295114628737714e-05,"reference_active":true,"delta_db":-0.5153520976304024},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.655957629268445e-06,"output_power":3.55322074761509e-06,"reference_active":true,"delta_db":-1.173868301048123}],"active_band_count":3,"active_band_mae_db":0.6359303248341849}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":0.8300325159645549,"rms_envelope_nrmse":0.06856379691574387,"envelope_correlation":0.9970533666198034,"rms_gain_error_db":-0.37885602265943374,"reference_mean_power":0.0003195237933424113,"output_mean_power":0.0002928313712465759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4169476048907228e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.318339905927386e-05,"output_power":7.908706322272677e-05,"reference_active":true,"delta_db":-0.21931213654573783},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1060276940136074e-05,"output_power":1.026377858894028e-05,"reference_active":true,"delta_db":-0.3245872664513917},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5597258507742536e-07,"output_power":2.06685284185939e-07,"reference_active":true,"delta_db":-0.928838980892348}],"active_band_count":3,"active_band_mae_db":0.4909127946298259}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":1.3463578504402092,"rms_envelope_nrmse":0.12551973044090564,"envelope_correlation":0.9901818116791489,"rms_gain_error_db":-0.41483292073256595,"reference_mean_power":8.301357548516478e-05,"output_mean_power":7.545114378044081e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.362198065588498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.2275109921127754e-05,"output_power":3.1381812115446e-05,"reference_active":true,"delta_db":-0.1218971210960046},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8777051155174135e-06,"output_power":2.298649120693703e-06,"reference_active":true,"delta_db":-0.9757360574631146},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.623133887825749e-07,"output_power":6.014757208151054e-07,"reference_active":true,"delta_db":-0.41845433188469094}],"active_band_count":3,"active_band_mae_db":0.5053625034812701}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":0.9389105544316237,"rms_envelope_nrmse":0.06992582855536796,"envelope_correlation":0.997972128536339,"rms_gain_error_db":-0.41459313238509665,"reference_mean_power":8.336414632310506e-05,"output_mean_power":7.577396167410307e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1059391757915856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.094184470310103e-05,"output_power":3.0081931908020074e-05,"reference_active":true,"delta_db":-0.12240478198831833},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2229001330453797e-06,"output_power":9.961620926247567e-07,"reference_active":true,"delta_db":-0.8906098092609293},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.748361881351689e-08,"output_power":3.9441928094897337e-08,"reference_active":true,"delta_db":-0.8058567290912256}],"active_band_count":3,"active_band_mae_db":0.6062904401134911}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1320-122617-0039","method":"magphase","status":"measured","mos":{"status":"measured","value":3.032550096511841,"predicted_mos":3.032550096511841,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288783,"input_sample_rate":48000,"input_seconds":6.0163125,"input_float32_sha256":"19411daa38d9e725acc8078db9e3823eb5b887ee0c675bbe1809b55d4791a8cf","internal_samples":96261,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7301449280003},"started_at":"2026-10-13T05:12:54.919064+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00930804299969168,"native_input_samples":288000,"native_input_sha256":"fee91c33b6a041c5c93d1c7e66090d25c37e179d6cfd171d4da137af1ebcbd5d","output_sha256":"6cc807b4628342d43e70f2824c1786dab03fa74c6e76652d7141f7217983f74b","output_samples":288783,"length_delta_samples":783,"analysis_s":2.6312799299994367,"synthesis_s":0.9186860829995567,"total_s":3.5499660129989934,"rtf":0.5916610021664989,"output_peak":0.5642328893196872,"output_rms":0.08308757921479959,"finished_at":"2026-10-13T05:12:58.488615+00:00","evaluation_input":{"reference":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"output_sha256":"6cc807b4628342d43e70f2824c1786dab03fa74c6e76652d7141f7217983f74b","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96261,"length_delta_samples":261,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.606572945183345,"output_peak":0.5645751303592371},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9972886215601334,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.2449039372899,"spectral_convergence":0.7171812584616369,"spectral_convergence_status":"ok","amplitude_floor":2.2281588625700377e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.091350342125086,"spectral_convergence":0.45107171065672197,"spectral_convergence_status":"ok","amplitude_floor":1.3568197926870596e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.2821199918929524,"spectral_convergence":0.4207742074948114,"spectral_convergence_status":"ok","amplitude_floor":1.1580179000740219e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0038028140043242973,"output_power":0.0034129816333259124,"delta_db":-0.4697113341739193,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002493613270814922,"output_power":0.002414916543889564,"delta_db":-0.13926973814245525,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006858506604602334,"output_power":0.0006580036121051913,"delta_db":-0.18001283513402005,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002628229680145485,"output_power":0.00026339376772298913,"delta_db":0.009421793114463669,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001273968713381539,"output_power":0.0001267626854131698,"delta_db":-0.021673315995205267,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.457810201861998e-05,"output_power":5.051355186557916e-05,"delta_db":-0.3361052173694896,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19269903898825888,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.30299520737763175,"pitch":null,"output_rms_dbfs":-21.594816599322527,"output_peak":0.5645751303592371,"reference_rms_dbfs":-21.2918213919449,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.206124757102646},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6751697682543103,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618048431617668,"rms_dbfs":-21.2918213919449,"peak":0.5504150390625,"peak_dbfs":-5.186194166503699,"crest_factor_db":16.105627225441197,"mean":-0.0005158704121907552,"zero_energy":false},"output":{"rms":0.08322602834643984,"rms_dbfs":-21.594816599322527,"peak":0.5645751303592371,"peak_dbfs":-4.965565130428614,"crest_factor_db":16.629251468893916,"mean":0.0023194683962917944,"zero_energy":false},"rms_difference_db":-0.3029952073776272},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.624486075597816e-06,"threshold_dbfs":-51.78847808438614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.9253438697196544e-06,"output_mean_power":2.5042358052271468e-05,"output_rms_dbfs":-46.01324779260625,"output_zero_energy":false,"output_peak":0.04625331322683927,"output_energy_fraction":0.0001928215673082665,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":300,"output_estimated_voiced_frames":290,"both_estimated_voiced_frames":254,"reference_voiced_output_unvoiced_frames":46,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.8466666666666667,"absolute_difference_cents_median":11.281564836632505,"absolute_difference_cents_p95":74.41023090971376,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00025738076998093343,"log_rms_mae_db":3.6824678170405596,"rms_envelope_nrmse":0.33107446299302823,"envelope_correlation":0.851581197839821,"rms_gain_error_db":-0.3026507108004613,"reference_mean_power":0.007432639167823318,"output_mean_power":0.006932310051599915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.369590439777949e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014473681837183883,"output_power":0.001368606750886704,"reference_active":true,"delta_db":-0.24300343766056234},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002579988221497507,"output_power":0.0002271485268722699,"reference_active":true,"delta_db":-0.5530779864219072},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2677922934851034e-05,"output_power":2.713943366873363e-05,"reference_active":true,"delta_db":-0.8065366343921082}],"active_band_count":3,"active_band_mae_db":0.5342060194915259}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022253035095363898,"log_rms_mae_db":2.916724446411002,"rms_envelope_nrmse":0.25493856978434776,"envelope_correlation":0.9017440322929164,"rms_gain_error_db":-0.3025581399207097,"reference_mean_power":0.0074694638279480755,"output_mean_power":0.006966804352195038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0676309697735546e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013241614802002638,"output_power":0.0012536749039229922,"reference_active":true,"delta_db":-0.2375601788256759},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001197578791096237,"output_power":0.00010414594051820802,"reference_active":true,"delta_db":-0.6066174978214189},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429382109480039e-06,"output_power":1.9518243508493027e-06,"reference_active":true,"delta_db":-0.9505509698843955}],"active_band_count":3,"active_band_mae_db":0.5982428821771634}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017894181312046162,"log_rms_mae_db":5.6236265027962435,"rms_envelope_nrmse":0.5022145592346953,"envelope_correlation":0.793525493515456,"rms_gain_error_db":-0.22621801654054097,"reference_mean_power":0.0017811195275832714,"output_mean_power":0.0016907183392127407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1758328243827778e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007391946250824912,"output_power":0.0006890953704019551,"reference_active":true,"delta_db":-0.30479468158000156},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015375544244981539,"output_power":0.00013603200746713383,"reference_active":true,"delta_db":-0.531893902255753},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532547306567537e-05,"output_power":1.0254788288960253e-05,"reference_active":true,"delta_db":-0.11606719673991378}],"active_band_count":3,"active_band_mae_db":0.31758526019188943}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016320284141686103,"log_rms_mae_db":4.523505248496715,"rms_envelope_nrmse":0.38617508422718355,"envelope_correlation":0.8661763567782272,"rms_gain_error_db":-0.22613206501271632,"reference_mean_power":0.0017900337371132,"output_mean_power":0.0016992137346283072},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0791360125207298e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007228891058818755,"output_power":0.0006704299636264621,"reference_active":true,"delta_db":-0.32718263826623684},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980895739386467e-05,"output_power":6.236117377962686e-05,"reference_active":true,"delta_db":-0.4899687080139076},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.398569514368019e-07,"output_power":8.030094931588496e-07,"reference_active":true,"delta_db":-0.6834107840241622}],"active_band_count":3,"active_band_mae_db":0.5001873767681022}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.442874453255206e-05,"log_rms_mae_db":5.209736241597885,"rms_envelope_nrmse":0.5045107113309165,"envelope_correlation":0.7743710851792913,"rms_gain_error_db":0.021287533505881757,"reference_mean_power":0.0003180382981236394,"output_mean_power":0.0003196010328591303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6089659140343427e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.507437744605699e-05,"output_power":8.65164948751831e-05,"reference_active":true,"delta_db":0.07300136422849447},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3978123456671115e-05,"output_power":2.5963665866849267e-05,"reference_active":true,"delta_db":0.3455081951883004},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.655957629268445e-06,"output_power":4.848239331115242e-06,"reference_active":true,"delta_db":0.17575030997183538}],"active_band_count":3,"active_band_mae_db":0.19808662312954342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.70740568335455e-05,"log_rms_mae_db":3.996397282176606,"rms_envelope_nrmse":0.3728297344372014,"envelope_correlation":0.8624219712181076,"rms_gain_error_db":0.022612820466515488,"reference_mean_power":0.0003195237933424113,"output_mean_power":0.00032119182680354487},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4169476048907228e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.318339905927386e-05,"output_power":8.520777805408619e-05,"reference_active":true,"delta_db":0.10442577811998693},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1060276940136074e-05,"output_power":1.212496195077094e-05,"reference_active":true,"delta_db":0.39914382943549404},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5597258507742536e-07,"output_power":2.930891118699864e-07,"reference_active":true,"delta_db":0.5880623046693835}],"active_band_count":3,"active_band_mae_db":0.36387730407495483}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.041296803251993e-05,"log_rms_mae_db":5.1941564722524385,"rms_envelope_nrmse":0.43862085645810595,"envelope_correlation":0.8632215885873935,"rms_gain_error_db":-0.20967296751204664,"reference_mean_power":8.301357548516478e-05,"output_mean_power":7.910097252236623e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.362198065588498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.2275109921127754e-05,"output_power":3.222224853839177e-05,"reference_active":true,"delta_db":-0.007118868901522386},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8777051155174135e-06,"output_power":2.981502556707543e-06,"reference_active":true,"delta_db":0.15388897285917716},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.623133887825749e-07,"output_power":1.1305219314895334e-06,"reference_active":true,"delta_db":2.3221545735617695}],"active_band_count":3,"active_band_mae_db":0.8277208051074897}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.404038681520289e-05,"log_rms_mae_db":3.9499631303542566,"rms_envelope_nrmse":0.33094353844921215,"envelope_correlation":0.9176835715957361,"rms_gain_error_db":-0.20635255316228454,"reference_mean_power":8.336414632310506e-05,"output_mean_power":7.94957757832953e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1059391757915856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.094184470310103e-05,"output_power":3.128448521057493e-05,"reference_active":true,"delta_db":0.04782810937017529},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2229001330453797e-06,"output_power":1.2548676907590253e-06,"reference_active":true,"delta_db":0.11206945361717197},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.748361881351689e-08,"output_power":6.311511689266195e-08,"reference_active":true,"delta_db":1.2358958096745174}],"active_band_count":3,"active_band_mae_db":0.4652644575539549}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},{"case":"librispeech_1580-141083-0010","method":"hnm","status":"measured","mos":{"status":"measured","value":2.1818768978118896,"predicted_mos":2.1818768978118896,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85120,"input_sample_rate":16000,"input_seconds":5.32,"input_float32_sha256":"59a9c2dc846ced58e35de2a21c0815d1adcb7811f41b3be9a9115087919511a9","internal_samples":85120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4847633820008923},"started_at":"2026-10-13T04:38:21.827666+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4141000065137632e-05,"native_input_samples":85120,"native_input_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","output_sha256":"ca915e324c7f9389862901ae28048399aacb6a610dc84fb7cfa85d2587025cb5","output_samples":85120,"length_delta_samples":0,"analysis_s":0.8394678439999552,"synthesis_s":0.48336304500026017,"total_s":1.3228308890002154,"rtf":0.24865242274440139,"output_peak":0.6173433862923519,"output_rms":0.05341790216685251,"finished_at":"2026-10-13T04:38:23.156513+00:00","evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"ca915e324c7f9389862901ae28048399aacb6a610dc84fb7cfa85d2587025cb5","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.497387885965726,"spectral_convergence":0.35077592385568035,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.722003109263566,"spectral_convergence":0.375116371772414,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.934338190457901,"spectral_convergence":0.37797530886746494,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":4.911030756341792e-06,"delta_db":3.5172745046448552,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0004882968396926196,"delta_db":-0.05987749417464248,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.0008367256963574641,"delta_db":0.05193840171321234,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0007578437033175815,"delta_db":-0.06956512616994054,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0006012976930829823,"delta_db":-0.17250714375830029,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.00016439730870043616,"delta_db":-0.07493917328815858,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6576836406248515,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.050793409802089955,"pitch":null,"output_rms_dbfs":-25.4462634327136,"output_peak":0.6173433862923519,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.717909728562398},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.497387885965726,"spectral_convergence":0.35077592385568035,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.722003109263566,"spectral_convergence":0.375116371772414,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.934338190457901,"spectral_convergence":0.37797530886746494,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":4.911030756341792e-06,"delta_db":3.5172745046448552,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0004882968396926196,"delta_db":-0.05987749417464248,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.0008367256963574641,"delta_db":0.05193840171321234,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0007578437033175815,"delta_db":-0.06956512616994054,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0006012976930829823,"delta_db":-0.17250714375830029,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.00016439730870043616,"delta_db":-0.07493917328815858,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6576836406248515,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.050793409802089955,"pitch":null,"output_rms_dbfs":-25.4462634327136,"output_peak":0.6173433862923519,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.717909728562398},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.8630141708895033,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053731194916086494,"rms_dbfs":-25.39547002291151,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.088525425009674,"mean":1.1236147772996945e-06,"zero_energy":false},"output":{"rms":0.05341790216685251,"rms_dbfs":-25.4462634327136,"peak":0.6173433862923519,"peak_dbfs":-4.1894640033362665,"crest_factor_db":21.25679942937733,"mean":-2.986436538839752e-06,"zero_energy":false},"rms_difference_db":-0.050793409802089684},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":3.91808107428413e-06,"output_rms_dbfs":-54.06926581677186,"output_zero_energy":false,"output_peak":0.02822411069427957,"output_energy_fraction":0.00012388803558802176,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":218,"both_estimated_voiced_frames":214,"reference_voiced_output_unvoiced_frames":34,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.8629032258064516,"absolute_difference_cents_median":1.3105573100837593,"absolute_difference_cents_p95":9.19170132935453,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0002036923290800657,"log_rms_mae_db":0.8224810849743632,"rms_envelope_nrmse":0.08090252224902485,"envelope_correlation":0.9939668706408566,"rms_gain_error_db":-0.05079208073000972,"reference_mean_power":0.0028897508304992817,"output_mean_power":0.002856151164416561},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.5709970652253345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007599193486798771,"output_power":0.0007550020064765105,"reference_active":true,"delta_db":-0.02819396634094052},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001405506492845202,"output_power":0.00013258935384740487,"reference_active":true,"delta_db":-0.25324202052672906},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.357724113057032e-05,"output_power":2.3602301115416955e-05,"reference_active":true,"delta_db":0.004613615834698786}],"active_band_count":3,"active_band_mae_db":0.09534986756745611}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.0001938989159680862,"log_rms_mae_db":0.3787510433710168,"rms_envelope_nrmse":0.04384117701461856,"envelope_correlation":0.9980980907985432,"rms_gain_error_db":-0.05079550698689289,"reference_mean_power":0.002906120097080002,"output_mean_power":0.0028723278364629536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.4139945205650041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006810740061577038,"output_power":0.0006819646944204854,"reference_active":true,"delta_db":0.005675863092021876},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.5174236911387e-05,"output_power":6.378440947350152e-05,"reference_active":true,"delta_db":-0.09361415847189726},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3046113483012312e-06,"output_power":1.2639142101873353e-06,"reference_active":true,"delta_db":-0.1376355531308715}],"active_band_count":3,"active_band_mae_db":0.07897519156493021}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013759847516455276,"log_rms_mae_db":1.5662142706710287,"rms_envelope_nrmse":0.11377632232238016,"envelope_correlation":0.9892602550069806,"rms_gain_error_db":-0.016763539877145305,"reference_mean_power":0.001146104301837711,"output_mean_power":0.0011416889262124148},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":6.878091269334419e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003194650474713193,"output_power":0.0003142476457995881,"reference_active":true,"delta_db":-0.0715131629409406},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.126588772977069e-05,"output_power":6.317406543520462e-05,"reference_active":true,"delta_db":-0.5234287342666815},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2614620971957042e-05,"output_power":1.6666153088362116e-05,"reference_active":true,"delta_db":1.2096116083382467}],"active_band_count":3,"active_band_mae_db":0.6015178351819562}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00012449215623942515,"log_rms_mae_db":0.8380273498002102,"rms_envelope_nrmse":0.06922578370594595,"envelope_correlation":0.9956713027834757,"rms_gain_error_db":-0.016763702253013015,"reference_mean_power":0.001152608697675057,"output_mean_power":0.0011481682208866388},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":6.232509709606391e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029975403822873686,"output_power":0.0002950943205707836,"reference_active":true,"delta_db":-0.06804191457855464},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2561293827419036e-05,"output_power":2.9134858448642845e-05,"reference_active":true,"delta_db":-0.48288740893253834},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.743814177835148e-07,"output_power":9.089567423537879e-07,"reference_active":true,"delta_db":0.6958829244544852}],"active_band_count":3,"active_band_mae_db":0.4156040826551927}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013997823549865974,"log_rms_mae_db":1.3671880735445392,"rms_envelope_nrmse":0.12514066774108726,"envelope_correlation":0.9883135509957537,"rms_gain_error_db":-0.11647074731457013,"reference_mean_power":0.0010964326399058029,"output_mean_power":0.001067418881942495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":8.778805813988018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004566223368026275,"output_power":0.00044847367071088436,"reference_active":true,"delta_db":-0.07820200974027375},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7374973721885583e-05,"output_power":4.4266195330304e-05,"reference_active":true,"delta_db":-0.29476786011971406},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4063527720029607e-05,"output_power":1.5214803251852845e-05,"reference_active":true,"delta_db":0.341720671697474}],"active_band_count":3,"active_band_mae_db":0.23823018051915393}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013340521153934581,"log_rms_mae_db":0.679640502058891,"rms_envelope_nrmse":0.06699714361749506,"envelope_correlation":0.9965439586401881,"rms_gain_error_db":-0.11647937304664228,"reference_mean_power":0.0011026403148293473,"output_mean_power":0.0010734601575600538},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.248821386475401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236383893109588,"output_power":0.0004178001463126017,"reference_active":true,"delta_db":-0.06026720431445517},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0735941322111324e-05,"output_power":2.0996473532312984e-05,"reference_active":true,"delta_db":0.054226035862070913},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.366449484579512e-07,"output_power":9.3126692233117e-07,"reference_active":true,"delta_db":1.0181596250013332}],"active_band_count":3,"active_band_mae_db":0.37755095505928643}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00010989278407704884,"log_rms_mae_db":1.5151924108871804,"rms_envelope_nrmse":0.1617158147580112,"envelope_correlation":0.9811700003657502,"rms_gain_error_db":-0.08042089740665893,"reference_mean_power":0.00030755753866752836,"output_mean_power":0.0003019147194009544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.696003640873152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014789902573029848,"output_power":0.00013611354179307803,"reference_active":true,"delta_db":-0.3606397828191803},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177704720777167e-05,"output_power":3.693836711803394e-05,"reference_active":true,"delta_db":0.6536415412889846},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203391401191374e-06,"output_power":4.955505331482614e-06,"reference_active":true,"delta_db":2.008861489953119}],"active_band_count":3,"active_band_mae_db":1.0077142713537612}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010174719078202196,"log_rms_mae_db":0.7819574996213936,"rms_envelope_nrmse":0.09023679818276734,"envelope_correlation":0.9937728618938332,"rms_gain_error_db":-0.08042536413322003,"reference_mean_power":0.00030929077108025465,"output_mean_power":0.0003036158395846529},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.5256223542383673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014158879062551859,"output_power":0.00013416672988037705,"reference_active":true,"delta_db":-0.23384037659368606},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5011511578101904e-05,"output_power":1.808587008825379e-05,"reference_active":true,"delta_db":0.8091498140371243},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.631467090177329e-07,"output_power":3.25598557129821e-07,"reference_active":true,"delta_db":0.9248452868615034}],"active_band_count":3,"active_band_mae_db":0.6559451591641046}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141083-0010","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.770113706588745,"predicted_mos":2.770113706588745,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85120,"input_sample_rate":16000,"input_seconds":5.32,"input_float32_sha256":"0db9f16f5c9b56bfc93ccfeb77797d7300640efb52c662e98a521abf81d69552","internal_samples":85120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3792349149989604},"started_at":"2026-10-13T04:41:17.201139+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1116999960213434e-05,"native_input_samples":85120,"native_input_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","output_sha256":"012a86b6a8a8b9e42fc59eac26382a88310e60bd571af68a818eb29d1f1ac8b5","output_samples":85120,"length_delta_samples":0,"analysis_s":2.179965973000435,"synthesis_s":0.1858255580000332,"total_s":2.3657915310004682,"rtf":0.44469765620309554,"output_peak":0.9082551527172927,"output_rms":0.06233840196167613,"finished_at":"2026-10-13T04:41:19.574337+00:00","evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"012a86b6a8a8b9e42fc59eac26382a88310e60bd571af68a818eb29d1f1ac8b5","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.810284939918121,"spectral_convergence":0.4685478913015548,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.396384580906157,"spectral_convergence":0.3303088245679989,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.798946902814793,"spectral_convergence":0.37884858321587483,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":3.1460804092660656e-06,"delta_db":1.5832461483379892,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0007589829168448775,"delta_db":1.8556033801169245,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.001274135949272974,"delta_db":1.8782650287271223,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0009600258570160029,"delta_db":0.9574677117490015,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0007022406902744503,"delta_db":0.50145738369178,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.0001875448653179354,"delta_db":0.4971655746790571,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.212200871216979,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.290583322722025,"pitch":null,"output_rms_dbfs":-24.104886700189482,"output_peak":0.9082551527172927,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.668538807879689},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.810284939918121,"spectral_convergence":0.4685478913015548,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.396384580906157,"spectral_convergence":0.3303088245679989,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.798946902814793,"spectral_convergence":0.37884858321587483,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":3.1460804092660656e-06,"delta_db":1.5832461483379892,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0007589829168448775,"delta_db":1.8556033801169245,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.001274135949272974,"delta_db":1.8782650287271223,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0009600258570160029,"delta_db":0.9574677117490015,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0007022406902744503,"delta_db":0.50145738369178,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.0001875448653179354,"delta_db":0.4971655746790571,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.212200871216979,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.290583322722025,"pitch":null,"output_rms_dbfs":-24.104886700189482,"output_peak":0.9082551527172927,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.668538807879689},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.9242058025370615,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053731194916086494,"rms_dbfs":-25.39547002291151,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.088525425009674,"mean":1.1236147772996945e-06,"zero_energy":false},"output":{"rms":0.06233840196167612,"rms_dbfs":-24.104886700189482,"peak":0.9082551527172927,"peak_dbfs":-0.8358425923362938,"crest_factor_db":23.26904410785319,"mean":3.158507245811964e-08,"zero_energy":false},"rms_difference_db":1.2905833227220285},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":3.5720246040010865e-06,"output_rms_dbfs":-54.47085558365488,"output_zero_energy":false,"output_peak":0.03495494249140266,"output_energy_fraction":8.293401992422061e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":261,"both_estimated_voiced_frames":231,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9314516129032258,"absolute_difference_cents_median":4.153275518633684,"absolute_difference_cents_p95":25.028575940791537,"difference_over_600_cents_frames":11,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0002036923290800657,"log_rms_mae_db":1.3954792208277502,"rms_envelope_nrmse":0.1893093237378297,"envelope_correlation":0.9921666922982117,"rms_gain_error_db":1.2905864446893587,"reference_mean_power":0.0028897508304992817,"output_mean_power":0.003889726285240178},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.5709970652253345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007599193486798771,"output_power":0.001004316605580901,"reference_active":true,"delta_db":1.2110314103156976},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001405506492845202,"output_power":0.00019597105632513997,"reference_active":true,"delta_db":1.443590773634393},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.357724113057032e-05,"output_power":2.737861685469883e-05,"reference_active":true,"delta_db":0.6491851899237062}],"active_band_count":3,"active_band_mae_db":1.101269124624599}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.0001938989159680862,"log_rms_mae_db":1.1012943361745073,"rms_envelope_nrmse":0.17546968742404373,"envelope_correlation":0.9958314178137506,"rms_gain_error_db":1.2905991948225353,"reference_mean_power":0.002906120097080002,"output_mean_power":0.003911771491520411},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.4139945205650041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006810740061577038,"output_power":0.0009110820659451995,"reference_active":true,"delta_db":1.2636319258542539},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.5174236911387e-05,"output_power":9.073736921218323e-05,"reference_active":true,"delta_db":1.4371022846520884},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3046113483012312e-06,"output_power":1.5767066066674365e-06,"reference_active":true,"delta_db":0.8226973543391867}],"active_band_count":3,"active_band_mae_db":1.174477188281843}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013759847516455276,"log_rms_mae_db":1.7497501035430405,"rms_envelope_nrmse":0.30133428002890755,"envelope_correlation":0.9873389282161,"rms_gain_error_db":2.0343104273846633,"reference_mean_power":0.001146104301837711,"output_mean_power":0.001830860206180617},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":6.878091269334419e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003194650474713193,"output_power":0.0005055554738854272,"reference_active":true,"delta_db":1.9934546793196828},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.126588772977069e-05,"output_power":0.00012535373657866753,"reference_active":true,"delta_db":2.4525558472075977},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2614620971957042e-05,"output_power":1.857295099639017e-05,"reference_active":true,"delta_db":1.68006706951395}],"active_band_count":3,"active_band_mae_db":2.042025865347077}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00012449215623942515,"log_rms_mae_db":1.3534910847862225,"rms_envelope_nrmse":0.28988935738194727,"envelope_correlation":0.9910579404073465,"rms_gain_error_db":2.0343092291213365,"reference_mean_power":0.001152608697675057,"output_mean_power":0.0018412502354657465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":6.232509709606391e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029975403822873686,"output_power":0.00048139141457743083,"reference_active":true,"delta_db":2.0573329799838365},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2561293827419036e-05,"output_power":5.5999506075591823e-05,"reference_active":true,"delta_db":2.3548254316772637},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.743814177835148e-07,"output_power":1.13342624048871e-06,"reference_active":true,"delta_db":1.6543833997423212}],"active_band_count":3,"active_band_mae_db":2.0221806038011403}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013997823549865974,"log_rms_mae_db":1.5501745295009448,"rms_envelope_nrmse":0.14128617955872252,"envelope_correlation":0.9936362131935008,"rms_gain_error_db":0.8472362940609063,"reference_mean_power":0.0010964326399058029,"output_mean_power":0.0013326177237283161},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":8.778805813988018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004566223368026275,"output_power":0.0005501378936390719,"reference_active":true,"delta_db":0.8091440804061766},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7374973721885583e-05,"output_power":6.13886239880206e-05,"reference_active":true,"delta_db":1.1253891679759065},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4063527720029607e-05,"output_power":1.2317607349816884e-05,"reference_active":true,"delta_db":-0.5756791763531911}],"active_band_count":3,"active_band_mae_db":0.8367374749117581}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013340521153934581,"log_rms_mae_db":1.195038068590915,"rms_envelope_nrmse":0.11891705503380588,"envelope_correlation":0.9974256580660038,"rms_gain_error_db":0.8472533388181914,"reference_mean_power":0.0011026403148293473,"output_mean_power":0.0013401678679718465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.248821386475401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236383893109588,"output_power":0.0005125778414416785,"reference_active":true,"delta_db":0.8276451969293017},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0735941322111324e-05,"output_power":2.721803610679692e-05,"reference_active":true,"delta_db":1.1813303062553422},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.366449484579512e-07,"output_power":6.580371915142564e-07,"reference_active":true,"delta_db":-0.49007774885282657}],"active_band_count":3,"active_band_mae_db":0.8330177506791568}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00010989278407704884,"log_rms_mae_db":1.5044146468886423,"rms_envelope_nrmse":0.14079266246977606,"envelope_correlation":0.988298979922821,"rms_gain_error_db":0.4330721844375828,"reference_mean_power":0.00030755753866752836,"output_mean_power":0.0003398079968555638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.696003640873152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014789902573029848,"output_power":0.00016068540073199407,"reference_active":true,"delta_db":0.36011107065224096},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177704720777167e-05,"output_power":3.1956373652086137e-05,"reference_active":true,"delta_db":0.024439513742997583},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203391401191374e-06,"output_power":3.723781304528862e-06,"reference_active":true,"delta_db":0.767823684775527}],"active_band_count":3,"active_band_mae_db":0.3841247563902552}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010174719078202196,"log_rms_mae_db":0.9662435395036808,"rms_envelope_nrmse":0.09234474143696793,"envelope_correlation":0.9957221479050135,"rms_gain_error_db":0.4330773412046224,"reference_mean_power":0.00030929077108025465,"output_mean_power":0.0003417233816349287},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.5256223542383673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014158879062551859,"output_power":0.00015254731903347236,"reference_active":true,"delta_db":0.3237570720483373},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5011511578101904e-05,"output_power":1.5042355593999688e-05,"reference_active":true,"delta_db":0.008914254215898159},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.631467090177329e-07,"output_power":2.8709884170912603e-07,"reference_active":true,"delta_db":0.3783349736745023}],"active_band_count":3,"active_band_mae_db":0.23700209997957924}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141083-0010","method":"facodec","status":"measured","mos":{"status":"measured","value":4.219021320343018,"predicted_mos":4.219021320343018,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85200,"input_sample_rate":16000,"input_seconds":5.325,"input_float32_sha256":"321f107c4e3ae15c75a9048c78f977ada805f0a8a148cb8303b082348e838787","internal_samples":85200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3335941580007784},"started_at":"2026-10-13T04:47:03.413367+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.620800013479311e-05,"native_input_samples":85120,"native_input_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","output_sha256":"665416ce4ef6f05ab6d3dab5195a4d3b61786ff5f134b99f7f6bd441d42403ad","output_samples":85200,"length_delta_samples":80,"analysis_s":1.285176123000383,"synthesis_s":11.65611182199973,"total_s":12.941287945000113,"rtf":2.4325729219925023,"output_peak":0.5737216472625732,"output_rms":0.04948864388929695,"finished_at":"2026-10-13T04:47:16.361078+00:00","evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"665416ce4ef6f05ab6d3dab5195a4d3b61786ff5f134b99f7f6bd441d42403ad","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":85120,"output_samples":85200,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.109888935279994,"output_peak":0.5737216472625732},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990610328638497,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.14868513029955,"spectral_convergence":0.2929865119522796,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.044565020164061,"spectral_convergence":0.32879520820945096,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.157767934765466,"spectral_convergence":0.3581906817140369,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":2.049687529779992e-06,"delta_db":-0.2775754464228244,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.00043288098201152317,"delta_db":-0.5830315676835923,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.0007062862859076499,"delta_db":-0.6840849291193571,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0006488805172581367,"delta_db":-0.7437142457195296,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0005322913230336143,"delta_db":-0.7019086529967346,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.00012903868770934517,"delta_db":-1.1266868314606322,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6861669455671117,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7103394695646892,"pitch":null,"output_rms_dbfs":-26.105809492476197,"output_peak":0.5737216472625732,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.783672695076359},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.9217143822390703,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053731194916086494,"rms_dbfs":-25.39547002291151,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.088525425009674,"mean":1.1236147772996945e-06,"zero_energy":false},"output":{"rms":0.04951189234365871,"rms_dbfs":-26.105809492476197,"peak":0.5737216472625732,"peak_dbfs":-4.8259752662758695,"crest_factor_db":21.27983422620033,"mean":7.889113651392896e-05,"zero_energy":false},"rms_difference_db":-0.710339469564687},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":1.8656784993216333e-06,"output_rms_dbfs":-57.291631933871,"output_zero_energy":false,"output_peak":0.008114248514175415,"output_energy_fraction":6.866688727774723e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":261,"both_estimated_voiced_frames":239,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9637096774193549,"absolute_difference_cents_median":2.88118845971813,"absolute_difference_cents_p95":20.820956707628426,"difference_over_600_cents_frames":6,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0002036923290800657,"log_rms_mae_db":1.1030575857581835,"rms_envelope_nrmse":0.11771774389610709,"envelope_correlation":0.9928221637402299,"rms_gain_error_db":-0.7103437154272361,"reference_mean_power":0.0028897508304992817,"output_mean_power":0.002453725779089602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.5709970652253345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007599193486798771,"output_power":0.0006578415725836296,"reference_active":true,"delta_db":-0.626461869940367},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001405506492845202,"output_power":0.000126643886846954,"reference_active":true,"delta_db":-0.45248625147416616},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.357724113057032e-05,"output_power":2.5142512337129448e-05,"reference_active":true,"delta_db":0.2791568674528744}],"active_band_count":3,"active_band_mae_db":0.4527016629558025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.0001938989159680862,"log_rms_mae_db":0.902951842006096,"rms_envelope_nrmse":0.10836345046063892,"envelope_correlation":0.9947416921127653,"rms_gain_error_db":-0.7103481192237857,"reference_mean_power":0.002906120097080002,"output_mean_power":0.002467622638375011},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.4139945205650041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006810740061577038,"output_power":0.0005910075138967087,"reference_active":true,"delta_db":-0.6160130292430539},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.5174236911387e-05,"output_power":5.933864879086962e-05,"reference_active":true,"delta_db":-0.40738302249781255},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3046113483012312e-06,"output_power":1.3384967451155174e-06,"reference_active":true,"delta_db":0.11136167470382652}],"active_band_count":3,"active_band_mae_db":0.37825257548156427}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013759847516455276,"log_rms_mae_db":1.3985217533814382,"rms_envelope_nrmse":0.14829989952924472,"envelope_correlation":0.9862719493537042,"rms_gain_error_db":-0.7180333088625497,"reference_mean_power":0.001146104301837711,"output_mean_power":0.0009714508033292902},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":6.878091269334419e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003194650474713193,"output_power":0.00026716256878490127,"reference_active":true,"delta_db":-0.7764773854604822},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.126588772977069e-05,"output_power":7.103772769215225e-05,"reference_active":true,"delta_db":-0.013926382677365075},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2614620971957042e-05,"output_power":1.1649553118829697e-05,"reference_active":true,"delta_db":-0.3456493993845069}],"active_band_count":3,"active_band_mae_db":0.37868438917411806}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00012449215623942515,"log_rms_mae_db":1.0485478124474612,"rms_envelope_nrmse":0.13743676542184835,"envelope_correlation":0.9883208277130698,"rms_gain_error_db":-0.7180352496461013,"reference_mean_power":0.001152608697675057,"output_mean_power":0.0009769635652798962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":6.232509709606391e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029975403822873686,"output_power":0.0002524613716479758,"reference_active":true,"delta_db":-0.7457010509125794},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2561293827419036e-05,"output_power":3.431872558247209e-05,"reference_active":true,"delta_db":0.22829498746078536},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.743814177835148e-07,"output_power":7.289478264483511e-07,"reference_active":true,"delta_db":-0.26258477672810604}],"active_band_count":3,"active_band_mae_db":0.4121936050338236}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013997823549865974,"log_rms_mae_db":1.550710574822969,"rms_envelope_nrmse":0.1309998211342757,"envelope_correlation":0.9917569662091547,"rms_gain_error_db":-0.7577095039966486,"reference_mean_power":0.0010964326399058029,"output_mean_power":0.000920896887784011},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":8.778805813988018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004566223368026275,"output_power":0.00039756951497214036,"reference_active":true,"delta_db":-0.6014407629535543},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7374973721885583e-05,"output_power":4.502160040034397e-05,"reference_active":true,"delta_db":-0.22128053106430648},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4063527720029607e-05,"output_power":1.5458897871523194e-05,"reference_active":true,"delta_db":0.41084254503223705}],"active_band_count":3,"active_band_mae_db":0.4111879463500326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013340521153934581,"log_rms_mae_db":1.235983697620388,"rms_envelope_nrmse":0.11544619044502874,"envelope_correlation":0.994510513611273,"rms_gain_error_db":-0.757719586204882,"reference_mean_power":0.0011026403148293473,"output_mean_power":0.0009261085816034048},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.248821386475401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236383893109588,"output_power":0.0003694930158461126,"reference_active":true,"delta_db":-0.5938907438686849},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0735941322111324e-05,"output_power":1.9112973029029487e-05,"reference_active":true,"delta_db":-0.3539550828558875},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.366449484579512e-07,"output_power":8.132570378700293e-07,"reference_active":true,"delta_db":0.42969615301003006}],"active_band_count":3,"active_band_mae_db":0.4591806599115342}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00010989278407704884,"log_rms_mae_db":1.7576052734034822,"rms_envelope_nrmse":0.1779420116606007,"envelope_correlation":0.982962977933121,"rms_gain_error_db":-0.9283288379384537,"reference_mean_power":0.00030755753866752836,"output_mean_power":0.0002483667720003097},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.696003640873152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014789902573029848,"output_power":0.00012142060520180204,"reference_active":true,"delta_db":-0.8567291992136585},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177704720777167e-05,"output_power":3.0312453550021884e-05,"reference_active":true,"delta_db":-0.20492448714093642},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203391401191374e-06,"output_power":3.7229996454818267e-06,"reference_active":true,"delta_db":0.7669119614030183}],"active_band_count":3,"active_band_mae_db":0.609521882585871}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010174719078202196,"log_rms_mae_db":1.4624809145064972,"rms_envelope_nrmse":0.14679053964017047,"envelope_correlation":0.9904140048100211,"rms_gain_error_db":-0.9283129573628311,"reference_mean_power":0.00030929077108025465,"output_mean_power":0.0002497673497127062},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.5256223542383673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014158879062551859,"output_power":0.00011868926684032505,"reference_active":true,"delta_db":-0.7661742511131213},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5011511578101904e-05,"output_power":1.4180318717029268e-05,"reference_active":true,"delta_db":-0.24738433387587694},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.631467090177329e-07,"output_power":3.06011038930759e-07,"reference_active":true,"delta_db":0.65539150359391}],"active_band_count":3,"active_band_mae_db":0.5563166961943028}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141083-0010","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.1031007766723633,"predicted_mos":3.1031007766723633,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":235008,"input_sample_rate":44100,"input_seconds":5.328979591836735,"input_float32_sha256":"33154b2b8ca004682c9e6e02123a1983d161ecb629a503ae1a754de6e3759151","internal_samples":85264,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5377213820011093},"started_at":"2026-10-13T05:00:36.080392+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007722453000496898,"native_input_samples":234612,"native_input_sha256":"4f6198db10209c08060b27c7b27f4cc0d8a48b5524f7f3347d092cfb68165ca6","output_sha256":"2e358ae4ff5d8d1ead44df803f6f2e2aa584635f926bb3b0fbb0254ddb00dd9f","output_samples":235008,"length_delta_samples":396,"analysis_s":0.12155743699986488,"synthesis_s":45.81768261700017,"total_s":45.93924005400004,"rtf":8.635195498872188,"output_peak":0.6338417530059814,"output_rms":0.050762309175587425,"finished_at":"2026-10-13T05:01:22.050706+00:00","evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"2e358ae4ff5d8d1ead44df803f6f2e2aa584635f926bb3b0fbb0254ddb00dd9f","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":85120,"output_samples":85264,"length_delta_samples":144,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.883837551371297,"output_peak":0.6371453096163985},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983111277913305,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.78698231226007,"spectral_convergence":0.46610382008675355,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.178020021610234,"spectral_convergence":0.4958743972702261,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.085559710684471,"spectral_convergence":0.4870834788940438,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":1.7491717328697102e-06,"delta_db":-0.9661275313148179,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0003834030405343897,"delta_db":-1.1101611023321625,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.0007809831767421543,"delta_db":-0.24747587682694225,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0007198477424526918,"delta_db":-0.2929551218529157,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0005868414028859689,"delta_db":-0.2781950539156892,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.00011151665961469715,"delta_db":-1.7604886885258988,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7759005624614045,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.48103147490554116,"pitch":null,"output_rms_dbfs":-25.876501497817053,"output_peak":0.6371453096163985,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.01685401485159},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.7663692111205928,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053731194916086494,"rms_dbfs":-25.39547002291151,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.088525425009674,"mean":1.1236147772996945e-06,"zero_energy":false},"output":{"rms":0.05083641602200898,"rms_dbfs":-25.876501497817053,"peak":0.6371453096163985,"peak_dbfs":-3.9152301926085187,"crest_factor_db":21.961271305208534,"mean":-2.6400496773255132e-05,"zero_energy":false},"rms_difference_db":-0.4810314749055422},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":2.6958851076906573e-06,"output_rms_dbfs":-55.69298620356004,"output_zero_energy":false,"output_peak":0.015855864754215184,"output_energy_fraction":9.411983009270427e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":242,"both_estimated_voiced_frames":222,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.8951612903225806,"absolute_difference_cents_median":6.155647966051047,"absolute_difference_cents_p95":37.34932473005941,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0002036923290800657,"log_rms_mae_db":1.5159134959811422,"rms_envelope_nrmse":0.10923743928017397,"envelope_correlation":0.9918683388622604,"rms_gain_error_db":-0.4810268384674193,"reference_mean_power":0.0028897508304992817,"output_mean_power":0.0025867693911929704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.5709970652253345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007599193486798771,"output_power":0.0007098229591211108,"reference_active":true,"delta_db":-0.29617460005095747},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001405506492845202,"output_power":0.0001253037722150148,"reference_active":true,"delta_db":-0.4986871076927415},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.357724113057032e-05,"output_power":2.0432412737466133e-05,"reference_active":true,"delta_db":-0.6217333234374827}],"active_band_count":3,"active_band_mae_db":0.47219834372706054}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.0001938989159680862,"log_rms_mae_db":1.1798312517304907,"rms_envelope_nrmse":0.08503738773100135,"envelope_correlation":0.9961105901222357,"rms_gain_error_db":-0.48101215320583196,"reference_mean_power":0.002906120097080002,"output_mean_power":0.002601431187268694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.4139945205650041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006810740061577038,"output_power":0.0006387949621811499,"reference_active":true,"delta_db":-0.278328228916111},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.5174236911387e-05,"output_power":5.876652899140449e-05,"reference_active":true,"delta_db":-0.44945914821818395},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3046113483012312e-06,"output_power":1.3365398737621051e-06,"reference_active":true,"delta_db":0.10500767882817315}],"active_band_count":3,"active_band_mae_db":0.27759835198748933}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013759847516455276,"log_rms_mae_db":2.52588900359391,"rms_envelope_nrmse":0.1791044272511067,"envelope_correlation":0.9745039846078273,"rms_gain_error_db":-0.48924386825055755,"reference_mean_power":0.001146104301837711,"output_mean_power":0.0010239995264681747},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":6.878091269334419e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003194650474713193,"output_power":0.0002564943075754348,"reference_active":true,"delta_db":-0.9534561794878704},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.126588772977069e-05,"output_power":6.381969425809727e-05,"reference_active":true,"delta_db":-0.47926980054001594},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2614620971957042e-05,"output_power":1.2529917831120286e-05,"reference_active":true,"delta_db":-0.02925982942176795}],"active_band_count":3,"active_band_mae_db":0.48732860314988474}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00012449215623942515,"log_rms_mae_db":1.9199608491354139,"rms_envelope_nrmse":0.1523524324874811,"envelope_correlation":0.9806894698668782,"rms_gain_error_db":-0.48925023242498245,"reference_mean_power":0.001152608697675057,"output_mean_power":0.001029809441563116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":6.232509709606391e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029975403822873686,"output_power":0.00024166833208687284,"reference_active":true,"delta_db":-0.9354529780662131},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2561293827419036e-05,"output_power":2.8637550993878914e-05,"reference_active":true,"delta_db":-0.557657777790266},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.743814177835148e-07,"output_power":8.36549277559911e-07,"reference_active":true,"delta_db":0.33536605487435933}],"active_band_count":3,"active_band_mae_db":0.6094922702436129}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013997823549865974,"log_rms_mae_db":2.8946055863953997,"rms_envelope_nrmse":0.20669553767323198,"envelope_correlation":0.9709851442234336,"rms_gain_error_db":-0.06556348104808113,"reference_mean_power":0.0010964326399058029,"output_mean_power":0.001080004605699625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":8.778805813988018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004566223368026275,"output_power":0.0005197141439460212,"reference_active":true,"delta_db":0.5620738399120193},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7374973721885583e-05,"output_power":5.620755041624265e-05,"reference_active":true,"delta_db":0.7424567661797327},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4063527720029607e-05,"output_power":1.2295247372612703e-05,"reference_active":true,"delta_db":-0.5835700264770474}],"active_band_count":3,"active_band_mae_db":0.6293668775229332}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013340521153934581,"log_rms_mae_db":2.324528937889816,"rms_envelope_nrmse":0.16753966864141,"envelope_correlation":0.9804910263104154,"rms_gain_error_db":-0.06554196634913045,"reference_mean_power":0.0011026403148293473,"output_mean_power":0.0010861246505739976},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.248821386475401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236383893109588,"output_power":0.00048136223291240653,"reference_active":true,"delta_db":0.5547670490115877},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0735941322111324e-05,"output_power":2.5230393752470662e-05,"reference_active":true,"delta_db":0.8520027304979071},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.366449484579512e-07,"output_power":8.411378835180089e-07,"reference_active":true,"delta_db":0.5760897832812668}],"active_band_count":3,"active_band_mae_db":0.6609531875969206}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00010989278407704884,"log_rms_mae_db":4.068645708857544,"rms_envelope_nrmse":0.2738916454549418,"envelope_correlation":0.9706537709875643,"rms_gain_error_db":-1.9266698517881795,"reference_mean_power":0.00030755753866752836,"output_mean_power":0.00019736011588080289},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.696003640873152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014789902573029848,"output_power":8.627646004187926e-05,"reference_active":true,"delta_db":-2.340729956183805},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177704720777167e-05,"output_power":1.7082093559401173e-05,"reference_active":true,"delta_db":-2.6957244297222105},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203391401191374e-06,"output_power":3.21217094144346e-06,"reference_active":true,"delta_db":0.12596850259253956}],"active_band_count":3,"active_band_mae_db":1.7208076294995183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010174719078202196,"log_rms_mae_db":3.734895131352999,"rms_envelope_nrmse":0.24974440680014545,"envelope_correlation":0.9807668009788809,"rms_gain_error_db":-1.9266150220814617,"reference_mean_power":0.00030929077108025465,"output_mean_power":0.0001984748393486816},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.5256223542383673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014158879062551859,"output_power":8.357212356858548e-05,"reference_active":true,"delta_db":-2.289674345299272},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5011511578101904e-05,"output_power":7.4731602662313945e-06,"reference_active":true,"delta_db":-3.029201295011542},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.631467090177329e-07,"output_power":3.3519104106681654e-07,"reference_active":true,"delta_db":1.050944593733433}],"active_band_count":3,"active_band_mae_db":2.1232734113480825}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141083-0010","method":"golf","status":"measured","mos":{"status":"measured","value":2.6787829399108887,"predicted_mos":2.6787829399108887,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":127440,"input_sample_rate":24000,"input_seconds":5.31,"input_float32_sha256":"0f30f4b262fd7bc0be8df21df018b01280d1e5d450bc54b743bd1fccea004427","internal_samples":84960,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4310201839998626},"started_at":"2026-10-13T05:02:01.881049+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028378249999150285,"native_input_samples":127680,"native_input_sha256":"0b7ae54e4ae3bef8e0583e093d7adf63b0f21140bf4513bc2de5b91e7c888316","output_sha256":"7319831e9cf6326c7384c7731540fe0e5608c98ff7c0aaa7077f6cc049daf9b2","output_samples":127440,"length_delta_samples":-240,"analysis_s":0.598173962999681,"synthesis_s":0.1287940600004731,"total_s":0.7269680230001541,"rtf":0.1366481246240891,"output_peak":0.5731508135795593,"output_rms":0.047714849111231146,"finished_at":"2026-10-13T05:02:02.618232+00:00","analysis_stages":{"external_f0_s":0.04748694799945952},"evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"7319831e9cf6326c7384c7731540fe0e5608c98ff7c0aaa7077f6cc049daf9b2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":85120,"output_samples":84960,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.422797176671388,"output_peak":0.5520690266943826},"common_prefix":{"samples":84960,"sample_rate":16000,"reference_coverage":0.9981203007518797,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84960,"output_samples":84960,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.527339251841338,"spectral_convergence":0.5411013044270954,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.79437715496123,"spectral_convergence":0.45428620248070095,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.812862296425921,"spectral_convergence":0.4913503886354457,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1887653142988217e-06,"output_power":4.39096950955513e-06,"delta_db":3.02361223655243,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004960076927814973,"output_power":0.0004020438227447309,"delta_db":-0.9121501846048541,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008283354070589984,"output_power":0.0006794523417095015,"delta_db":-0.8604722583998041,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007715247564499366,"output_power":0.0006374254348332926,"delta_db":-0.8292047723588718,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006268355117856337,"output_power":0.0003933586377281854,"delta_db":-2.0236490110781267,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001675707035485501,"output_power":0.00016220262992652125,"delta_db":-0.1414020154622448,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2984150797427219,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0354750540638233,"pitch":null,"output_rms_dbfs":-26.422797176671388,"output_peak":0.5520690266943826,"reference_rms_dbfs":-25.387322122607564,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.04485956774283},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84960,"output_samples":84960,"intelligibility":{"estoi":{"value":0.8845671243530827,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053781621739576756,"rms_dbfs":-25.387322122607564,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.08037752470573,"mean":5.431094412076271e-07,"zero_energy":false},"output":{"rms":0.04773755163863965,"rms_dbfs":-26.422797176671388,"peak":0.5520690266943826,"peak_dbfs":-5.1601323570723165,"crest_factor_db":21.262664819599074,"mean":0.00025441732090305895,"zero_energy":false},"rms_difference_db":-1.0354750540638236},"full_scale":{"reference":{"sample_count":84960,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84960,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":531,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":1.3659195530818151e-06,"output_rms_dbfs":-58.64574878020133,"output_zero_energy":false,"output_peak":0.004747042784441418,"output_energy_fraction":5.4181566340305295e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":524,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":243,"both_estimated_voiced_frames":228,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9193548387096774,"absolute_difference_cents_median":4.516283041106795,"absolute_difference_cents_p95":20.201260426283262,"difference_over_600_cents_frames":7,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84960,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1061,"rms_floor":0.0002036923290800657,"log_rms_mae_db":1.8191846542912118,"rms_envelope_nrmse":0.21925071185550618,"envelope_correlation":0.963851489436054,"rms_gain_error_db":-1.0354824678621695,"reference_mean_power":0.002895183261371113,"output_mean_power":0.002281013273137536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1061,"frequency_resolution_hz":0.18850141376060323,"power_floor":1.575254067723739e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007626135951607811,"output_power":0.0006891000125550102,"reference_active":true,"delta_db":-0.44022285374565345},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014011756378496084,"output_power":0.00012613169653083234,"reference_active":true,"delta_db":-0.45668340424647613},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.346033419495059e-05,"output_power":2.5432359186525205e-05,"reference_active":true,"delta_db":0.3505245418721965}],"active_band_count":3,"active_band_mae_db":0.41581026662144205}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":0.0001938989159680862,"log_rms_mae_db":1.5170676866343378,"rms_envelope_nrmse":0.17994098803093628,"envelope_correlation":0.9780097122344362,"rms_gain_error_db":-1.0354902780827566,"reference_mean_power":0.00291161618672308,"output_mean_power":0.0022939560728555777},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":1.4177236999702131e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006829878802922884,"output_power":0.000618456748972417,"reference_active":true,"delta_db":-0.4310366397273969},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.488422536277129e-05,"output_power":5.789451327699577e-05,"reference_active":true,"delta_db":-0.4950171692387683},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3058107158680727e-06,"output_power":1.851187610920901e-06,"reference_active":true,"delta_db":1.5157020686842482}],"active_band_count":3,"active_band_mae_db":0.8139186258834711}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1061,"rms_floor":0.00013759847516455276,"log_rms_mae_db":2.1018726649709905,"rms_envelope_nrmse":0.28073379543770305,"envelope_correlation":0.9384672299205367,"rms_gain_error_db":-0.9446385531096511,"reference_mean_power":0.0011482643591173983,"output_mean_power":0.0009238001526895161},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1061,"frequency_resolution_hz":0.18850141376060323,"power_floor":6.894466043598042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000320458649116509,"output_power":0.000261297013137379,"reference_active":true,"delta_db":-0.8863755222713272},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.142677450115982e-05,"output_power":5.927520244807022e-05,"reference_active":true,"delta_db":-0.8098799270631385},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2540429771748847e-05,"output_power":1.6427805126515117e-05,"reference_active":true,"delta_db":1.1726712201943719}],"active_band_count":3,"active_band_mae_db":0.9563088898429459}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":0.00012449215623942515,"log_rms_mae_db":1.5927655166026469,"rms_envelope_nrmse":0.22480115547147167,"envelope_correlation":0.9607077862185872,"rms_gain_error_db":-0.9446371823326063,"reference_mean_power":0.0011547932981056592,"output_mean_power":0.000929053099454975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":6.247315950268651e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003006883796278352,"output_power":0.00025051207402302923,"reference_active":true,"delta_db":-0.7928798216559575},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.257991872639369e-05,"output_power":2.625125523922049e-05,"reference_active":true,"delta_db":-0.9379992197065259},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.772291528597912e-07,"output_power":9.934901085994556e-07,"reference_active":true,"delta_db":1.0661446548107807}],"active_band_count":3,"active_band_mae_db":0.9323412320577548}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1061,"rms_floor":0.00013997823549865974,"log_rms_mae_db":2.0921307860739926,"rms_envelope_nrmse":0.24386120119936824,"envelope_correlation":0.9635659943268794,"rms_gain_error_db":-1.1799659529529198,"reference_mean_power":0.0010984933433972862,"output_mean_power":0.0008371452825328145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1061,"frequency_resolution_hz":0.18850141376060323,"power_floor":8.807732881851024e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004579188505357619,"output_power":0.00040034241094256474,"reference_active":true,"delta_db":-0.5835692166432419},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7484419529720305e-05,"output_power":3.862804739385471e-05,"reference_active":true,"delta_db":-0.8964837778505608},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.390859834636459e-05,"output_power":8.067104720078132e-06,"reference_active":true,"delta_db":-2.365656711041337}],"active_band_count":3,"active_band_mae_db":1.28190323517838}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":0.00013340521153934581,"log_rms_mae_db":1.731022154310398,"rms_envelope_nrmse":0.20544549032812787,"envelope_correlation":0.9766853135089322,"rms_gain_error_db":-1.1799941476460332,"reference_mean_power":0.001104724284832373,"output_mean_power":0.0008418883238255529},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":8.27582674609568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042474231853939417,"output_power":0.0003691695453950429,"reference_active":true,"delta_db":-0.6089966664927173},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0669330321239686e-05,"output_power":1.8032155851348116e-05,"reference_active":true,"delta_db":-0.5927875356087476},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.339781539033454e-07,"output_power":6.196506179360433e-07,"reference_active":true,"delta_db":-0.7353624665943446}],"active_band_count":3,"active_band_mae_db":0.6457155562319364}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1061,"rms_floor":0.00010989278407704884,"log_rms_mae_db":2.831861666058335,"rms_envelope_nrmse":0.30898295474423937,"envelope_correlation":0.9359232877921677,"rms_gain_error_db":-0.8543194861714973,"reference_mean_power":0.0003081303134437354,"output_mean_power":0.00025310602136822084},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1061,"frequency_resolution_hz":0.18850141376060323,"power_floor":2.702167318881018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014851314561838808,"output_power":0.00018341843831827038,"reference_active":true,"delta_db":0.9167809454259769},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.144441368751895e-05,"output_power":3.7163957917754826e-05,"reference_active":true,"delta_db":0.7257845834963684},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203126420174407e-06,"output_power":5.530179816445956e-06,"reference_active":true,"delta_db":2.4854114217418104}],"active_band_count":3,"active_band_mae_db":1.375992316888052}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84960,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":0.00010174719078202196,"log_rms_mae_db":2.43631922412068,"rms_envelope_nrmse":0.25246701176487296,"envelope_correlation":0.957210456184473,"rms_gain_error_db":-0.8542506785119168,"reference_mean_power":0.0003098720357582632,"output_mean_power":0.00025454074886190346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":2.531173855165388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001420046963120987,"output_power":0.00018136163792158725,"reference_active":true,"delta_db":1.0624272195442914},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4792734196938853e-05,"output_power":1.637393672267619e-05,"reference_active":true,"delta_db":0.44104654013358957},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.635371782262716e-07,"output_power":4.119473997900608e-07,"reference_active":true,"delta_db":1.9399987443642606}],"active_band_count":3,"active_band_mae_db":1.1478241680140473}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141083-0010","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.9289145469665527,"predicted_mos":2.9289145469665527,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":127800,"input_sample_rate":24000,"input_seconds":5.325,"input_float32_sha256":"92d7d08bce3b6598da4c76ea9552ea6f556fb353a8ca24564511b65020d0d71f","internal_samples":85200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.326543991001017},"started_at":"2026-10-13T05:05:53.581775+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002549683000324876,"native_input_samples":127680,"native_input_sha256":"0b7ae54e4ae3bef8e0583e093d7adf63b0f21140bf4513bc2de5b91e7c888316","output_sha256":"1e37c08deeb51f4d3e57a23a32c00866000ffa43cf93a809e3ccf025491a1a2b","output_samples":127800,"length_delta_samples":120,"analysis_s":2.2279544729999543,"synthesis_s":11.720281144999717,"total_s":13.948235617999671,"rtf":2.621848800375878,"output_peak":0.5792047381401062,"output_rms":0.052209281930943034,"finished_at":"2026-10-13T05:06:07.541507+00:00","analysis_stages":{"external_f0_s":1.496802433999619},"evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"1e37c08deeb51f4d3e57a23a32c00866000ffa43cf93a809e3ccf025491a1a2b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":85120,"output_samples":85200,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.640902790475366,"output_peak":0.5773336799210441},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990610328638497,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.756117959522352,"spectral_convergence":0.4557386661680387,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.672364327997712,"spectral_convergence":0.3743601867761485,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.388369218823957,"spectral_convergence":0.346337445255272,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":2.591621188335523e-06,"delta_db":0.7412631841795148,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0005097612192324949,"delta_db":0.12695129943444083,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.000815593215215134,"delta_db":-0.05915662170866267,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0007125100800308251,"delta_db":-0.33745146384900565,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0005652444674439045,"delta_db":-0.4410393181858175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.00012523337648278477,"delta_db":-1.2566853090401195,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4937578660662602,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.24141796523644865,"pitch":null,"output_rms_dbfs":-25.63688798814796,"output_peak":0.5773336799210441,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.6056171687813405},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.8797085664253849,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053731194916086494,"rms_dbfs":-25.39547002291151,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.088525425009674,"mean":1.1236147772996945e-06,"zero_energy":false},"output":{"rms":0.05225833885222031,"rms_dbfs":-25.63688798814796,"peak":0.5773336799210441,"peak_dbfs":-4.771462126410395,"crest_factor_db":20.865425861737563,"mean":7.654554139549608e-05,"zero_energy":false},"rms_difference_db":-0.24141796523645098},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":8.97122694053221e-06,"output_rms_dbfs":-50.47148157060309,"output_zero_energy":false,"output_peak":0.0765253331421533,"output_energy_fraction":0.0002963945725969435,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":251,"both_estimated_voiced_frames":234,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9435483870967742,"absolute_difference_cents_median":5.665799414725415,"absolute_difference_cents_p95":26.87635510869471,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0002036923290800657,"log_rms_mae_db":1.857972706747392,"rms_envelope_nrmse":0.1565189286125653,"envelope_correlation":0.9775780354498549,"rms_gain_error_db":-0.24143301917206195,"reference_mean_power":0.0028897508304992817,"output_mean_power":0.0027334875192609053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.5709970652253345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007599193486798771,"output_power":0.0007072730760406892,"reference_active":true,"delta_db":-0.3118037644690113},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001405506492845202,"output_power":0.00012697150341809966,"reference_active":true,"delta_db":-0.44126594279948805},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.357724113057032e-05,"output_power":2.1772978094877996e-05,"reference_active":true,"delta_db":-0.345751495156277}],"active_band_count":3,"active_band_mae_db":0.3662737341415921}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.0001938989159680862,"log_rms_mae_db":1.229824825782261,"rms_envelope_nrmse":0.10337164180157798,"envelope_correlation":0.9897971629206199,"rms_gain_error_db":-0.24143319306288724,"reference_mean_power":0.002906120097080002,"output_mean_power":0.0027489715074841764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.4139945205650041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006810740061577038,"output_power":0.0006336047051026198,"reference_active":true,"delta_db":-0.3137591170094756},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.5174236911387e-05,"output_power":6.05702156613185e-05,"reference_active":true,"delta_db":-0.3181683498024876},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3046113483012312e-06,"output_power":1.2720031108339424e-06,"reference_active":true,"delta_db":-0.10992978527684011}],"active_band_count":3,"active_band_mae_db":0.2472857506962678}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013759847516455276,"log_rms_mae_db":2.2396059485723026,"rms_envelope_nrmse":0.1742140181733686,"envelope_correlation":0.9751433889396114,"rms_gain_error_db":0.038648320044310305,"reference_mean_power":0.001146104301837711,"output_mean_power":0.0011563491211050907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":6.878091269334419e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003194650474713193,"output_power":0.0003214667075214608,"reference_active":true,"delta_db":0.027126530521119494},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.126588772977069e-05,"output_power":6.893075974315567e-05,"reference_active":true,"delta_db":-0.14468634095054794},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2614620971957042e-05,"output_power":1.3111002255964984e-05,"reference_active":true,"delta_db":0.16761686194141}],"active_band_count":3,"active_band_mae_db":0.11314324447102582}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00012449215623942515,"log_rms_mae_db":1.5292579609108314,"rms_envelope_nrmse":0.11458014505028427,"envelope_correlation":0.9883080621243042,"rms_gain_error_db":0.038641582152484574,"reference_mean_power":0.001152608697675057,"output_mean_power":0.001162909854355614},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":6.232509709606391e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029975403822873686,"output_power":0.0003021821614421948,"reference_active":true,"delta_db":0.03503780743979455},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2561293827419036e-05,"output_power":3.204823283856653e-05,"reference_active":true,"delta_db":-0.06897566089180204},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.743814177835148e-07,"output_power":8.227238512647536e-07,"reference_active":true,"delta_db":0.2629916494456775}],"active_band_count":3,"active_band_mae_db":0.12233503925909135}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013997823549865974,"log_rms_mae_db":2.2101877182057494,"rms_envelope_nrmse":0.1811656044518293,"envelope_correlation":0.9756880653263594,"rms_gain_error_db":-0.3490885718393989,"reference_mean_power":0.0010964326399058029,"output_mean_power":0.001011749733034925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":8.778805813988018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004566223368026275,"output_power":0.0004148902965900627,"reference_active":true,"delta_db":-0.41623874539839667},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7374973721885583e-05,"output_power":4.342102435300675e-05,"reference_active":true,"delta_db":-0.3784891732362419},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4063527720029607e-05,"output_power":1.076139405237551e-05,"reference_active":true,"delta_db":-1.162257391959853}],"active_band_count":3,"active_band_mae_db":0.6523284368648304}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013340521153934581,"log_rms_mae_db":1.3957188887495287,"rms_envelope_nrmse":0.11735304443105843,"envelope_correlation":0.989904615525546,"rms_gain_error_db":-0.3490776985898652,"reference_mean_power":0.0011026403148293473,"output_mean_power":0.0010174805059914068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.248821386475401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236383893109588,"output_power":0.00038345427516882954,"reference_active":true,"delta_db":-0.4328172404078201},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0735941322111324e-05,"output_power":1.9713198486451947e-05,"reference_active":true,"delta_db":-0.21966660523120046},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.366449484579512e-07,"output_power":6.330694161803831e-07,"reference_active":true,"delta_db":-0.6580688192364101}],"active_band_count":3,"active_band_mae_db":0.43685088829181024}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00010989278407704884,"log_rms_mae_db":2.3225592216384623,"rms_envelope_nrmse":0.20956717344338663,"envelope_correlation":0.9725151520235509,"rms_gain_error_db":-0.8799182548819907,"reference_mean_power":0.00030755753866752836,"output_mean_power":0.00025115079149789},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.696003640873152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014789902573029848,"output_power":0.00011265382388080751,"reference_active":true,"delta_db":-1.1821937529038617},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177704720777167e-05,"output_power":2.4680117517339424e-05,"reference_active":true,"delta_db":-1.0976631584852081},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203391401191374e-06,"output_power":2.206346170834486e-06,"reference_active":true,"delta_db":-1.5052814542900226}],"active_band_count":3,"active_band_mae_db":1.2617127885596975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010174719078202196,"log_rms_mae_db":1.4546999519844852,"rms_envelope_nrmse":0.15726518935445458,"envelope_correlation":0.9869782567104801,"rms_gain_error_db":-0.8798953026235723,"reference_mean_power":0.00030929077108025465,"output_mean_power":0.00025256747997553163},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.5256223542383673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014158879062551859,"output_power":0.00010653525797276588,"reference_active":true,"delta_db":-1.2353551041581388},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5011511578101904e-05,"output_power":1.183460022958716e-05,"reference_active":true,"delta_db":-1.0327083342007435},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.631467090177329e-07,"output_power":1.7320858142627242e-07,"reference_active":true,"delta_db":-1.8162853813492008}],"active_band_count":3,"active_band_mae_db":1.361449606569361}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141083-0010","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.669668197631836,"predicted_mos":2.669668197631836,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":127920,"input_sample_rate":24000,"input_seconds":5.33,"input_float32_sha256":"f8680dc4df6c7267df51b468299b5c659eb3ab23076022e1aaa172cadf085658","internal_samples":85280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.153095015000872},"started_at":"2026-10-13T05:10:09.560029+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002910265000537038,"native_input_samples":127680,"native_input_sha256":"0b7ae54e4ae3bef8e0583e093d7adf63b0f21140bf4513bc2de5b91e7c888316","output_sha256":"f5b9f6299c300c832445ce3498e48a2f7374d02e6c45de9b9dd6f3ec1d20f695","output_samples":127920,"length_delta_samples":240,"analysis_s":0.4865095669993025,"synthesis_s":0.17256206500132976,"total_s":0.6590716320006322,"rtf":0.12388564511290079,"output_peak":0.6204578876495361,"output_rms":0.05055186070026728,"finished_at":"2026-10-13T05:10:10.236513+00:00","analysis_stages":{"external_f0_s":0.08172744700095791},"evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"f5b9f6299c300c832445ce3498e48a2f7374d02e6c45de9b9dd6f3ec1d20f695","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":85120,"output_samples":85280,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.921123232815777,"output_peak":0.5848110255394134},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.99812382739212,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.485625156992499,"spectral_convergence":0.29833633046950137,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.589554031032715,"spectral_convergence":0.3099428659155242,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.029496656483606,"spectral_convergence":0.3734698825095972,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":3.230334882730923e-06,"delta_db":1.6980234362134095,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0003773653057747561,"delta_db":-1.1790969342977382,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.0008147916418437645,"delta_db":-0.06342701122519763,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0007227869119479949,"delta_db":-0.27525879305011797,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0004931708141364864,"delta_db":-1.0334288490223706,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.00015138569578422172,"delta_db":-0.4330377964072299,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7803788033693438,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5175002544998366,"pitch":null,"output_rms_dbfs":-25.912970277411347,"output_peak":0.5848110255394134,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.701558614836273},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.9253708975031165,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053731194916086494,"rms_dbfs":-25.39547002291151,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.088525425009674,"mean":1.1236147772996945e-06,"zero_energy":false},"output":{"rms":0.05062342051234739,"rms_dbfs":-25.912970277411347,"peak":0.5848110255394134,"peak_dbfs":-4.6596889629399865,"crest_factor_db":21.25328131447136,"mean":6.585688428643181e-06,"zero_energy":false},"rms_difference_db":-0.5175002544998364},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":3.0673681513156073e-06,"output_rms_dbfs":-55.132340960456936,"output_zero_energy":false,"output_peak":0.0353349910311905,"output_energy_fraction":0.00010799223683526787,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":227,"both_estimated_voiced_frames":222,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.8951612903225806,"absolute_difference_cents_median":0.7615665661229736,"absolute_difference_cents_p95":3.887868628806263,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0002036923290800657,"log_rms_mae_db":0.6228375231692482,"rms_envelope_nrmse":0.11993055113848221,"envelope_correlation":0.9893659131483625,"rms_gain_error_db":-0.5174991802367979,"reference_mean_power":0.0028897508304992817,"output_mean_power":0.0025651364928201913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.5709970652253345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007599193486798771,"output_power":0.0007202282085163889,"reference_active":true,"delta_db":-0.23297375454901992},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001405506492845202,"output_power":0.00013100423424279494,"reference_active":true,"delta_db":-0.3054752330811186},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.357724113057032e-05,"output_power":2.0538168326419466e-05,"reference_active":true,"delta_db":-0.5993127623776464}],"active_band_count":3,"active_band_mae_db":0.37925391666926167}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.0001938989159680862,"log_rms_mae_db":0.40229526348239775,"rms_envelope_nrmse":0.10421266734629112,"envelope_correlation":0.9921955442432412,"rms_gain_error_db":-0.5174952961573313,"reference_mean_power":0.002906120097080002,"output_mean_power":0.0025796692579245005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.4139945205650041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006810740061577038,"output_power":0.0006482746038812578,"reference_active":true,"delta_db":-0.2143529685785281},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.5174236911387e-05,"output_power":6.149529908251146e-05,"reference_active":true,"delta_db":-0.25234036916648783},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3046113483012312e-06,"output_power":1.0671231712723217e-06,"reference_active":true,"delta_db":-0.872666017794186}],"active_band_count":3,"active_band_mae_db":0.4464531185130673}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013759847516455276,"log_rms_mae_db":0.8250524968080283,"rms_envelope_nrmse":0.16306886891484954,"envelope_correlation":0.9793965555510509,"rms_gain_error_db":-0.49895924851468765,"reference_mean_power":0.001146104301837711,"output_mean_power":0.001021711349711128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":6.878091269334419e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003194650474713193,"output_power":0.00030440441065892037,"reference_active":true,"delta_db":-0.20972408286094937},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.126588772977069e-05,"output_power":6.823369666368451e-05,"reference_active":true,"delta_db":-0.18882798858345412},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2614620971957042e-05,"output_power":1.0347172415712728e-05,"reference_active":true,"delta_db":-0.8605252009392972}],"active_band_count":3,"active_band_mae_db":0.41969242412790025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00012449215623942515,"log_rms_mae_db":0.469766764910891,"rms_envelope_nrmse":0.12295417135229708,"envelope_correlation":0.9886170257910932,"rms_gain_error_db":-0.4989590930295334,"reference_mean_power":0.001152608697675057,"output_mean_power":0.001027509824772033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":6.232509709606391e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029975403822873686,"output_power":0.0002863953293218492,"reference_active":true,"delta_db":-0.19799111539877323},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2561293827419036e-05,"output_power":3.131115792858049e-05,"reference_active":true,"delta_db":-0.1700252460422523},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.743814177835148e-07,"output_power":5.990662374646862e-07,"reference_active":true,"delta_db":-1.114800789548126}],"active_band_count":3,"active_band_mae_db":0.49427238366305054}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013997823549865974,"log_rms_mae_db":0.8711293878001144,"rms_envelope_nrmse":0.12846581331099868,"envelope_correlation":0.9896690639668674,"rms_gain_error_db":-0.5254478219180191,"reference_mean_power":0.0010964326399058029,"output_mean_power":0.0009714874026892795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":8.778805813988018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004566223368026275,"output_power":0.00043353271459463925,"reference_active":true,"delta_db":-0.22535277127868453},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7374973721885583e-05,"output_power":4.509049297216221e-05,"reference_active":true,"delta_db":-0.21463998627297154},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4063527720029607e-05,"output_power":1.1823549830768295e-05,"reference_active":true,"delta_db":-0.7534638748143342}],"active_band_count":3,"active_band_mae_db":0.3978188774553301}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013340521153934581,"log_rms_mae_db":0.564898565147578,"rms_envelope_nrmse":0.10821653650397424,"envelope_correlation":0.993041972285936,"rms_gain_error_db":-0.5254427536382004,"reference_mean_power":0.0011026403148293473,"output_mean_power":0.0009769888150668107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.248821386475401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236383893109588,"output_power":0.00040331718603682753,"reference_active":true,"delta_db":-0.2134857973020507},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0735941322111324e-05,"output_power":1.9864150618434854e-05,"reference_active":true,"delta_db":-0.18653755675528688},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.366449484579512e-07,"output_power":5.881789361844503e-07,"reference_active":true,"delta_db":-0.9774874749212238}],"active_band_count":3,"active_band_mae_db":0.4591702763261871}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00010989278407704884,"log_rms_mae_db":0.9667338802037796,"rms_envelope_nrmse":0.1275909027616724,"envelope_correlation":0.9910278156813228,"rms_gain_error_db":-0.5962807454497556,"reference_mean_power":0.00030755753866752836,"output_mean_power":0.00026810091865885706},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.696003640873152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014789902573029848,"output_power":0.00013581709702290711,"reference_active":true,"delta_db":-0.3701086958799407},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177704720777167e-05,"output_power":2.8411265057655094e-05,"reference_active":true,"delta_db":-0.48622967340278317},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203391401191374e-06,"output_power":2.6243930552371897e-06,"reference_active":true,"delta_db":-0.7517291888173728}],"active_band_count":3,"active_band_mae_db":0.5360225193666989}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010174719078202196,"log_rms_mae_db":0.5852027677282058,"rms_envelope_nrmse":0.09726606381775736,"envelope_correlation":0.9960175434732738,"rms_gain_error_db":-0.596267507327698,"reference_mean_power":0.00030929077108025465,"output_mean_power":0.00026961261615751327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.5256223542383673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014158879062551859,"output_power":0.00012998286811266493,"reference_active":true,"delta_db":-0.3714275668446354},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5011511578101904e-05,"output_power":1.3814201425194251e-05,"reference_active":true,"delta_db":-0.3609864137849964},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.631467090177329e-07,"output_power":2.0173122943119467e-07,"reference_active":true,"delta_db":-1.154248076892728}],"active_band_count":3,"active_band_mae_db":0.6288873525074532}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141083-0010","method":"magphase","status":"measured","mos":{"status":"measured","value":2.896369457244873,"predicted_mos":2.896369457244873,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":255591,"input_sample_rate":48000,"input_seconds":5.3248125,"input_float32_sha256":"26bc0359c4cdf87130895cdab345b0019e9601d67b159c42d2bf0621c677e3be","internal_samples":85197,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6404782099998556},"started_at":"2026-10-13T05:12:58.489322+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010522504000618937,"native_input_samples":255360,"native_input_sha256":"4709fd8bf0a6147baa7d2b01d05612888e08b1e328e44d0587ff6b9f9b7b5940","output_sha256":"877309eb1e972764331b18b8e234e1f8aa49b5cf269fdffff10319835f88976d","output_samples":255591,"length_delta_samples":231,"analysis_s":3.537685921999582,"synthesis_s":1.1659967630002939,"total_s":4.703682684999876,"rtf":0.8841508806390744,"output_peak":0.6478873327496827,"output_rms":0.05335485092875394,"finished_at":"2026-10-13T05:13:03.215761+00:00","evaluation_input":{"reference":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"877309eb1e972764331b18b8e234e1f8aa49b5cf269fdffff10319835f88976d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":85120,"output_samples":85197,"length_delta_samples":77,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.45355781468324,"output_peak":0.6292405549502154},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990962123079451,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.14414546840121,"spectral_convergence":0.3205057617883775,"spectral_convergence_status":"ok","amplitude_floor":1.1374446115550809e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.047658919701646,"spectral_convergence":0.19320201338404513,"spectral_convergence_status":"ok","amplitude_floor":9.193083193108671e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.426424187021172,"spectral_convergence":0.21755491061938384,"spectral_convergence_status":"ok","amplitude_floor":7.316110787468302e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.1849686170678433e-06,"output_power":2.3955797076836333e-06,"delta_db":0.39965422100807785,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004950757600852583,"output_power":0.0004567206003661812,"delta_db":-0.3502106202640952,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008267786751374999,"output_power":0.0007701351955553843,"delta_db":-0.30822295153105295,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007700805582515175,"output_power":0.0007659562076624162,"delta_db":-0.023322189241638088,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006256626820185013,"output_power":0.0006748154908190033,"delta_db":0.32844791511710664,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00016725866300063496,"output_power":0.00018123488377983976,"delta_db":0.34853172759346457,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2930649374592392,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.054164873020528345,"pitch":null,"output_rms_dbfs":-25.44963489593204,"output_peak":0.6292405549502154,"reference_rms_dbfs":-25.39547002291151,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.20607619170801},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.9496484026029606,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053731194916086494,"rms_dbfs":-25.39547002291151,"peak":0.542816162109375,"peak_dbfs":-5.306944597901834,"crest_factor_db":20.088525425009674,"mean":1.1236147772996945e-06,"zero_energy":false},"output":{"rms":0.05339717181546705,"rms_dbfs":-25.44963489593204,"peak":0.6292405549502154,"peak_dbfs":-4.023665891874492,"crest_factor_db":21.425969004057546,"mean":0.0002538886212181252,"zero_energy":false},"rms_difference_db":-0.05416487302052886},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.8831495051272214e-06,"threshold_dbfs":-54.10815888361541,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":1.6478400842364256e-06,"output_mean_power":1.7667567119456085e-06,"output_rms_dbfs":-57.52823250116308,"output_zero_energy":false,"output_peak":0.018580571450245893,"output_energy_fraction":5.590747066063702e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":248,"output_estimated_voiced_frames":240,"both_estimated_voiced_frames":235,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9475806451612904,"absolute_difference_cents_median":2.4940429605196264,"absolute_difference_cents_p95":7.4887574164384265,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0002036923290800657,"log_rms_mae_db":0.7452052665513362,"rms_envelope_nrmse":0.0575480249955462,"envelope_correlation":0.9970401250806531,"rms_gain_error_db":-0.05416208289527849,"reference_mean_power":0.0028897508304992817,"output_mean_power":0.0028539357316814014},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.5709970652253345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007599193486798771,"output_power":0.0007518105366457108,"reference_active":true,"delta_db":-0.046590943663906086},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001405506492845202,"output_power":0.00013912148987565556,"reference_active":true,"delta_db":-0.04438636251908635},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.357724113057032e-05,"output_power":2.4751735666316443e-05,"reference_active":true,"delta_db":0.21112673217376504}],"active_band_count":3,"active_band_mae_db":0.10070134611891916}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.0001938989159680862,"log_rms_mae_db":0.4424509639194945,"rms_envelope_nrmse":0.03012057917762064,"envelope_correlation":0.9991732021604737,"rms_gain_error_db":-0.05416568712025535,"reference_mean_power":0.002906120097080002,"output_mean_power":0.002870099738344612},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.4139945205650041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006810740061577038,"output_power":0.0006731037572029825,"reference_active":true,"delta_db":-0.051122905785225894},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.5174236911387e-05,"output_power":6.40067106084934e-05,"reference_active":true,"delta_db":-0.07850446171521434},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3046113483012312e-06,"output_power":1.2186700730841172e-06,"reference_active":true,"delta_db":-0.2959500567935414}],"active_band_count":3,"active_band_mae_db":0.14185914143132722}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013759847516455276,"log_rms_mae_db":1.0710435120113444,"rms_envelope_nrmse":0.05364460975747026,"envelope_correlation":0.9987189192826426,"rms_gain_error_db":-0.32923996092901747,"reference_mean_power":0.001146104301837711,"output_mean_power":0.0010624295623560282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":6.878091269334419e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003194650474713193,"output_power":0.00029875998397528916,"reference_active":true,"delta_db":-0.291009216553175},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.126588772977069e-05,"output_power":6.671344099921788e-05,"reference_active":true,"delta_db":-0.2866835771738652},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2614620971957042e-05,"output_power":1.1834654290359868e-05,"reference_active":true,"delta_db":-0.2771862992582704}],"active_band_count":3,"active_band_mae_db":0.28495969766177015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00012449215623942515,"log_rms_mae_db":0.7806847311794114,"rms_envelope_nrmse":0.04413146058434997,"envelope_correlation":0.9994743775248854,"rms_gain_error_db":-0.3292396354670739,"reference_mean_power":0.001152608697675057,"output_mean_power":0.0010684591655790463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":6.232509709606391e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029975403822873686,"output_power":0.0002808079767115786,"reference_active":true,"delta_db":-0.28355602215940023},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2561293827419036e-05,"output_power":3.065625571486569e-05,"reference_active":true,"delta_db":-0.26182543289878174},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.743814177835148e-07,"output_power":7.085880090247965e-07,"reference_active":true,"delta_db":-0.3856112422970292}],"active_band_count":3,"active_band_mae_db":0.3103308991184037}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00013997823549865974,"log_rms_mae_db":0.8181456669600642,"rms_envelope_nrmse":0.06242596920322333,"envelope_correlation":0.9972158836913525,"rms_gain_error_db":0.06035912910490219,"reference_mean_power":0.0010964326399058029,"output_mean_power":0.0011117774691870518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":8.778805813988018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004566223368026275,"output_power":0.000466049786994677,"reference_active":true,"delta_db":0.08875161564315784},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7374973721885583e-05,"output_power":4.958305553164147e-05,"reference_active":true,"delta_db":0.19784304362111146},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4063527720029607e-05,"output_power":1.3873924716514757e-05,"reference_active":true,"delta_db":-0.05894939964754917}],"active_band_count":3,"active_band_mae_db":0.11518135297060617}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013340521153934581,"log_rms_mae_db":0.514373934768466,"rms_envelope_nrmse":0.033597943762721696,"envelope_correlation":0.9992001758104703,"rms_gain_error_db":0.06035888602049732,"reference_mean_power":0.0011026403148293473,"output_mean_power":0.001118071959380966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.248821386475401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236383893109588,"output_power":0.00043220352670879435,"reference_active":true,"delta_db":0.08692998167681305},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0735941322111324e-05,"output_power":2.156830179408625e-05,"reference_active":true,"delta_db":0.1709219651578127},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.366449484579512e-07,"output_power":7.083183810056973e-07,"reference_active":true,"delta_db":-0.1702970304894999}],"active_band_count":3,"active_band_mae_db":0.14271632577470855}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00010989278407704884,"log_rms_mae_db":0.9651395175139522,"rms_envelope_nrmse":0.12762771024390274,"envelope_correlation":0.9902300337532814,"rms_gain_error_db":0.3788927017468582,"reference_mean_power":0.00030755753866752836,"output_mean_power":0.0003355951276119491},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.696003640873152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014789902573029848,"output_power":0.0001483303879458305,"reference_active":true,"delta_db":0.012648194518983604},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177704720777167e-05,"output_power":2.8630800750795595e-05,"reference_active":true,"delta_db":-0.45280044551207643},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1203391401191374e-06,"output_power":4.712975138774639e-06,"reference_active":true,"delta_db":1.7909335011342438}],"active_band_count":3,"active_band_mae_db":0.7521273803884346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010174719078202196,"log_rms_mae_db":0.6214509355709074,"rms_envelope_nrmse":0.0806513601187493,"envelope_correlation":0.9966992934381251,"rms_gain_error_db":0.37887886377782143,"reference_mean_power":0.00030929077108025465,"output_mean_power":0.0003374852897845757},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.5256223542383673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014158879062551859,"output_power":0.0001409362727690439,"reference_active":true,"delta_db":-0.020060903969796623},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5011511578101904e-05,"output_power":1.3637624687801333e-05,"reference_active":true,"delta_db":-0.41685691215683446},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.631467090177329e-07,"output_power":2.9726618820098663e-07,"reference_active":true,"delta_db":0.5294757128878285}],"active_band_count":3,"active_band_mae_db":0.3221311763381532}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2352092266082764,"predicted_mos":2.2352092266082764,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":95360,"input_sample_rate":16000,"input_seconds":5.96,"input_float32_sha256":"9b923c8abb11d688a505a53498bb6019289871cbb429467174eb0063d7304681","internal_samples":95360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.090544114000295},"started_at":"2026-10-13T04:38:23.158326+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.7608999769436195e-05,"native_input_samples":95360,"native_input_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","output_sha256":"6cb62f5a676f4594648a6f559505de164066dc5d629ef1439d9766a58f9b2b25","output_samples":95360,"length_delta_samples":0,"analysis_s":0.9558676319993538,"synthesis_s":0.5104836610007624,"total_s":1.4663512930001161,"rtf":0.24603209614095908,"output_peak":0.49873233179501486,"output_rms":0.04452740875761405,"finished_at":"2026-10-13T04:38:24.631827+00:00","evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"6cb62f5a676f4594648a6f559505de164066dc5d629ef1439d9766a58f9b2b25","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.891201817846659,"spectral_convergence":0.40566977869505394,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.737513115992015,"spectral_convergence":0.4068711650959494,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.878115860676428,"spectral_convergence":0.42606685869634536,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":7.890930401129223e-06,"delta_db":1.9496805296276054,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.0001514061318146715,"delta_db":0.19223324059377875,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0004623875661286435,"delta_db":-0.20662184869313913,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0006784130915850873,"delta_db":-0.07565913627712365,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.00027839425494927,"delta_db":-0.24468993920208143,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.0004041981557888425,"delta_db":-0.062301427121008134,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.45519768691912277,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10178928490972834,"pitch":null,"output_rms_dbfs":-27.0274515521313,"output_peak":0.49873233179501486,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.835610264838368},"common_prefix":{"samples":95360,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.891201817846659,"spectral_convergence":0.40566977869505394,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.737513115992015,"spectral_convergence":0.4068711650959494,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.878115860676428,"spectral_convergence":0.42606685869634536,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":7.890930401129223e-06,"delta_db":1.9496805296276054,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.0001514061318146715,"delta_db":0.19223324059377875,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0004623875661286435,"delta_db":-0.20662184869313913,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0006784130915850873,"delta_db":-0.07565913627712365,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.00027839425494927,"delta_db":-0.24468993920208143,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.0004041981557888425,"delta_db":-0.062301427121008134,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.45519768691912277,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10178928490972834,"pitch":null,"output_rms_dbfs":-27.0274515521313,"output_peak":0.49873233179501486,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.835610264838368},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95360,"output_samples":95360,"intelligibility":{"estoi":{"value":0.822990142061189,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.045052291623524414,"rms_dbfs":-26.925662267221572,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.913020028217947,"mean":-3.6546847964293206e-06,"zero_energy":false},"output":{"rms":0.04452740875761405,"rms_dbfs":-27.027451552131303,"peak":0.49873233179501486,"peak_dbfs":-6.042649528954831,"crest_factor_db":20.98480202317647,"mean":3.0218345968826945e-05,"zero_energy":false},"rms_difference_db":-0.10178928490973149},"full_scale":{"reference":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":596,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":78,"quiet_samples":12480,"reference_mean_power":1.0344867731850498e-06,"output_mean_power":1.933447223696957e-06,"output_rms_dbfs":-57.13667678139665,"output_zero_energy":false,"output_peak":0.03527529650969771,"output_energy_fraction":0.00012762208043961168,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":589,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":142,"both_estimated_voiced_frames":138,"reference_voiced_output_unvoiced_frames":47,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.745945945945946,"absolute_difference_cents_median":1.6080852925333655,"absolute_difference_cents_p95":5.433734892970879,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95360,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001898595271537964,"log_rms_mae_db":0.9694074063881526,"rms_envelope_nrmse":0.08653623344103407,"envelope_correlation":0.9931159994425266,"rms_gain_error_db":-0.10178827093226143,"reference_mean_power":0.002031411160407025,"output_mean_power":0.0019843533423096076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.927152563814217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381525961065371,"output_power":0.0006098111397354533,"reference_active":true,"delta_db":-0.19729186772087037},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010674365906282644,"output_power":9.260133808259885e-05,"reference_active":true,"delta_db":-0.6172482363987253},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4567838102832594e-05,"output_power":3.568306901885376e-05,"reference_active":true,"delta_db":0.13789980675833824}],"active_band_count":3,"active_band_mae_db":0.31747997029264463}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00017019756464998765,"log_rms_mae_db":0.49193034555344317,"rms_envelope_nrmse":0.0475885248275162,"envelope_correlation":0.9977750123489627,"rms_gain_error_db":-0.10178784877914393,"reference_mean_power":0.0020416854120747243,"output_mean_power":0.0019943897838976383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":8.390142156345837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005509511464645055,"output_power":0.0005317672241235226,"reference_active":true,"delta_db":-0.15391525367206757},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5897699076585876e-05,"output_power":4.130557378391472e-05,"reference_active":true,"delta_db":-0.4578225481052366},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.848237969563024e-06,"output_power":1.9170268331761492e-06,"reference_active":true,"delta_db":0.1587030386961466}],"active_band_count":3,"active_band_mae_db":0.2568136134911503}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":1.8730578840695555,"rms_envelope_nrmse":0.11749702373492937,"envelope_correlation":0.9898929435378684,"rms_gain_error_db":-0.15855948541796946,"reference_mean_power":0.0005236182881812637,"output_mean_power":0.000504845928400934},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.785583338040323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002893453145707382,"output_power":0.00027026982643483827,"reference_active":true,"delta_db":-0.2961889139714056},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619778623468816e-05,"output_power":3.0029235032575677e-05,"reference_active":true,"delta_db":-0.8113774190679361},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0243815325354377e-05,"output_power":9.543411075310086e-06,"reference_active":true,"delta_db":-0.3075810930174485}],"active_band_count":3,"active_band_mae_db":0.47171580868559676}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":1.1368681569450476,"rms_envelope_nrmse":0.07467244545397074,"envelope_correlation":0.9957844687015139,"rms_gain_error_db":-0.15856096783041596,"reference_mean_power":0.000526268698480611,"output_mean_power":0.000507401145035184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.4283863314229963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002786455084699618,"output_power":0.00025968908806875884,"reference_active":true,"delta_db":-0.30598345688606804},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6309295594140896e-05,"output_power":1.3424550105021438e-05,"reference_active":true,"delta_db":-0.8453546390187896},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.788822138965803e-07,"output_power":6.886713929561223e-07,"reference_active":true,"delta_db":-0.5345974365743272}],"active_band_count":3,"active_band_mae_db":0.5619785108263949}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017019739791537767,"log_rms_mae_db":1.486142463416418,"rms_envelope_nrmse":0.10895195129923384,"envelope_correlation":0.9915871110572066,"rms_gain_error_db":-0.11201118672682843,"reference_mean_power":0.0008446325483556025,"output_mean_power":0.0008231267099497867},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0046345863274685e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028725854625924984,"output_power":0.00026866133850658074,"reference_active":true,"delta_db":-0.2906778416131251},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3436307686794475e-05,"output_power":3.947153941158164e-05,"reference_active":true,"delta_db":-0.41568836401001186},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.134050674510124e-05,"output_power":1.289347479923063e-05,"reference_active":true,"delta_db":0.5573751452146125}],"active_band_count":3,"active_band_mae_db":0.4212471169459165}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00015333717053704136,"log_rms_mae_db":0.8351731608495114,"rms_envelope_nrmse":0.06281967604189287,"envelope_correlation":0.9970977670218498,"rms_gain_error_db":-0.1120116754538762,"reference_mean_power":0.0008489024890862698,"output_mean_power":0.0008272878373404494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.579273675799128e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027065671114903563,"output_power":0.00025620453741064566,"reference_active":true,"delta_db":-0.23831983358672443},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9395570039165576e-05,"output_power":1.793204511093326e-05,"reference_active":true,"delta_db":-0.3407272538125524},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.660068008029238e-07,"output_power":8.778446018867682e-07,"reference_active":true,"delta_db":0.5918501745437322}],"active_band_count":3,"active_band_mae_db":0.3902990873143364}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010107776753147984,"log_rms_mae_db":1.5817693998791738,"rms_envelope_nrmse":0.1503738819676683,"envelope_correlation":0.9829004007820795,"rms_gain_error_db":-0.3341358148504886,"reference_mean_power":0.00032790731816785486,"output_mean_power":0.0003036249999817585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1725104205733306e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020494142716961813,"output_power":0.0001865257507502974,"reference_active":true,"delta_db":-0.40890959554246553},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.658028435672292e-05,"output_power":2.506863797660585e-05,"reference_active":true,"delta_db":-1.641163385276216},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.565019545039448e-06,"output_power":6.919122800857675e-06,"reference_active":true,"delta_db":-1.4063482324898102}],"active_band_count":3,"active_band_mae_db":1.1521404044361638}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":0.8634158388860467,"rms_envelope_nrmse":0.0900248856887965,"envelope_correlation":0.9940901956503022,"rms_gain_error_db":-0.33413141400678364,"reference_mean_power":0.0003295653902210023,"output_mean_power":0.00030516059707758205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":2.853826866715285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019529233091956366,"output_power":0.00017417196505417753,"reference_active":true,"delta_db":-0.497069372954023},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.517039000031073e-05,"output_power":1.040225334248475e-05,"reference_active":true,"delta_db":-1.638693191444144},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.363045655471933e-07,"output_power":4.612463821742506e-07,"reference_active":true,"delta_db":-2.031245206272244}],"active_band_count":3,"active_band_mae_db":1.3890025902234704}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.1402227878570557,"predicted_mos":2.1402227878570557,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":95360,"input_sample_rate":16000,"input_seconds":5.96,"input_float32_sha256":"2d492975e670451501f35b8d9f859b5b77db6a85d3eaa97634dc947d830658cf","internal_samples":95360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9698540570007026},"started_at":"2026-10-13T04:41:19.576421+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5411000024178065e-05,"native_input_samples":95360,"native_input_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","output_sha256":"ed8fddd44ac6c0bdf60155b0e277d2c54f38c6541d9341b740eb7cea58728fac","output_samples":95360,"length_delta_samples":0,"analysis_s":2.829055026999413,"synthesis_s":0.10936592500002007,"total_s":2.938420951999433,"rtf":0.4930236496643344,"output_peak":0.5960165909932796,"output_rms":0.050254229173220054,"finished_at":"2026-10-13T04:41:22.526212+00:00","evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"ed8fddd44ac6c0bdf60155b0e277d2c54f38c6541d9341b740eb7cea58728fac","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.170760380595011,"spectral_convergence":0.5588659807050211,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.582727149925583,"spectral_convergence":0.3975140498533497,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.86592110372342,"spectral_convergence":0.43111145626250164,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":7.061716751674511e-06,"delta_db":1.46750133937153,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.00017927678709412547,"delta_db":0.9260392039166403,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0006675660911655499,"delta_db":1.3882593669282837,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0008989873806691127,"delta_db":1.1469346176961075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.00033106342065174953,"delta_db":0.5078193532403811,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.00044153215346230896,"delta_db":0.32137865096371787,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9596554220194434,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9491145858107203,"pitch":null,"output_rms_dbfs":-25.976547681410853,"output_peak":0.5960165909932796,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.873136211414671},"common_prefix":{"samples":95360,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.170760380595011,"spectral_convergence":0.5588659807050211,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.582727149925583,"spectral_convergence":0.3975140498533497,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.86592110372342,"spectral_convergence":0.43111145626250164,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":7.061716751674511e-06,"delta_db":1.46750133937153,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.00017927678709412547,"delta_db":0.9260392039166403,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0006675660911655499,"delta_db":1.3882593669282837,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0008989873806691127,"delta_db":1.1469346176961075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.00033106342065174953,"delta_db":0.5078193532403811,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.00044153215346230896,"delta_db":0.32137865096371787,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9596554220194434,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9491145858107203,"pitch":null,"output_rms_dbfs":-25.976547681410853,"output_peak":0.5960165909932796,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.873136211414671},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95360,"output_samples":95360,"intelligibility":{"estoi":{"value":0.8679517139008773,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.045052291623524414,"rms_dbfs":-26.925662267221572,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.913020028217947,"mean":-3.6546847964293206e-06,"zero_energy":false},"output":{"rms":0.050254229173220054,"rms_dbfs":-25.976547681410857,"peak":0.5960165909932796,"peak_dbfs":-4.4948330173917945,"crest_factor_db":21.48171466401906,"mean":7.240779000964614e-09,"zero_energy":false},"rms_difference_db":0.9491145858107153},"full_scale":{"reference":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":596,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":78,"quiet_samples":12480,"reference_mean_power":1.0344867731850498e-06,"output_mean_power":1.4663158002698795e-06,"output_rms_dbfs":-58.337724856672594,"output_zero_energy":false,"output_peak":0.04816578286133729,"output_energy_fraction":7.598548247993504e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":589,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":160,"both_estimated_voiced_frames":134,"reference_voiced_output_unvoiced_frames":51,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.7243243243243244,"absolute_difference_cents_median":4.0096884414855,"absolute_difference_cents_p95":19.559505625341338,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95360,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001898595271537964,"log_rms_mae_db":1.6368931267440048,"rms_envelope_nrmse":0.1785481561038947,"envelope_correlation":0.984820470670437,"rms_gain_error_db":0.9491176306758785,"reference_mean_power":0.002031411160407025,"output_mean_power":0.002527607277801801},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.927152563814217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381525961065371,"output_power":0.0007839041476611374,"reference_active":true,"delta_db":0.893384219628205},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010674365906282644,"output_power":0.0001359391258836445,"reference_active":true,"delta_db":1.0500238710214675},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4567838102832594e-05,"output_power":4.553955082650627e-05,"reference_active":true,"delta_db":1.1971652338647463}],"active_band_count":3,"active_band_mae_db":1.0468577748381396}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00017019756464998765,"log_rms_mae_db":1.2203252095409882,"rms_envelope_nrmse":0.14503046203966702,"envelope_correlation":0.9930757014206819,"rms_gain_error_db":0.9491219899127133,"reference_mean_power":0.0020416854120747243,"output_mean_power":0.002540393686450267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":8.390142156345837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005509511464645055,"output_power":0.0006858953245984552,"reference_active":true,"delta_db":0.9514475138510357},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5897699076585876e-05,"output_power":5.984169238220408e-05,"reference_active":true,"delta_db":1.1521295308567145},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.848237969563024e-06,"output_power":2.4214438332327552e-06,"reference_active":true,"delta_db":1.173165117960052}],"active_band_count":3,"active_band_mae_db":1.092247387555934}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":2.1015005195805374,"rms_envelope_nrmse":0.22409190992034306,"envelope_correlation":0.9888674471921064,"rms_gain_error_db":1.4324557671998095,"reference_mean_power":0.0005236182881812637,"output_mean_power":0.0007282162791285813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.785583338040323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002893453145707382,"output_power":0.00042047728894030895,"reference_active":true,"delta_db":1.623260892705609},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619778623468816e-05,"output_power":5.373755568121344e-05,"reference_active":true,"delta_db":1.7159589713783663},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0243815325354377e-05,"output_power":1.4062779400387302e-05,"reference_active":true,"delta_db":1.3760942369508449}],"active_band_count":3,"active_band_mae_db":1.5717713670116067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":1.5825999945123121,"rms_envelope_nrmse":0.20362443638695144,"envelope_correlation":0.9938489966182228,"rms_gain_error_db":1.4324583866371923,"reference_mean_power":0.000526268698480611,"output_mean_power":0.0007319027490605689},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.4283863314229963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002786455084699618,"output_power":0.0004076271480922745,"reference_active":true,"delta_db":1.6521105345371216},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6309295594140896e-05,"output_power":2.3705423692150863e-05,"reference_active":true,"delta_db":1.6241251789321853},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.788822138965803e-07,"output_power":8.273871235780624e-07,"reference_active":true,"delta_db":0.26236971138741366}],"active_band_count":3,"active_band_mae_db":1.1795351416189068}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017019739791537767,"log_rms_mae_db":1.6385452564252285,"rms_envelope_nrmse":0.18545240862025478,"envelope_correlation":0.989559141648092,"rms_gain_error_db":1.0759999902590929,"reference_mean_power":0.0008446325483556025,"output_mean_power":0.0010821010344332729},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0046345863274685e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028725854625924984,"output_power":0.00037676459168804483,"reference_active":true,"delta_db":1.177971227779178},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3436307686794475e-05,"output_power":5.998027838729078e-05,"reference_active":true,"delta_db":1.4015557627647757},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.134050674510124e-05,"output_power":1.193733119034209e-05,"reference_active":true,"delta_db":0.22274781905643715}],"active_band_count":3,"active_band_mae_db":0.9340916032001303}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00015333717053704136,"log_rms_mae_db":1.1237627191221227,"rms_envelope_nrmse":0.15341553412897638,"envelope_correlation":0.996093102995317,"rms_gain_error_db":1.0760028829217871,"reference_mean_power":0.0008489024890862698,"output_mean_power":0.0010875721936167144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.579273675799128e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027065671114903563,"output_power":0.0003580975515167717,"reference_active":true,"delta_db":1.215825513055087},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9395570039165576e-05,"output_power":2.6758196323792892e-05,"reference_active":true,"delta_db":1.3975428770012148},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.660068008029238e-07,"output_power":8.640634641282874e-07,"reference_active":true,"delta_db":0.5231301648358743}],"active_band_count":3,"active_band_mae_db":1.045499518297392}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010107776753147984,"log_rms_mae_db":1.6558854180535971,"rms_envelope_nrmse":0.22143622853420214,"envelope_correlation":0.963908735885458,"rms_gain_error_db":0.31136836188822375,"reference_mean_power":0.00032790731816785486,"output_mean_power":0.00035227996576099043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1725104205733306e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020494142716961813,"output_power":0.00020416544271872425,"reference_active":true,"delta_db":-0.016475214376838025},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.658028435672292e-05,"output_power":4.2664244177698274e-05,"reference_active":true,"delta_db":0.6681697911309645},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.565019545039448e-06,"output_power":1.676315801743465e-05,"reference_active":true,"delta_db":2.4366997710130596}],"active_band_count":3,"active_band_mae_db":1.0404482588402875}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":0.9973335518166773,"rms_envelope_nrmse":0.1314681774972925,"envelope_correlation":0.9865303712911785,"rms_gain_error_db":0.3113694351716631,"reference_mean_power":0.0003295653902210023,"output_mean_power":0.00035406136625336913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":2.853826866715285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019529233091956366,"output_power":0.0002004397320771393,"reference_active":true,"delta_db":0.1129862457181219},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.517039000031073e-05,"output_power":1.9449008436521677e-05,"reference_active":true,"delta_db":1.079007189386941},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.363045655471933e-07,"output_power":1.1475241101262204e-06,"reference_active":true,"delta_db":1.9270432569470168}],"active_band_count":3,"active_band_mae_db":1.0396788973506934}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"facodec","status":"measured","mos":{"status":"measured","value":3.550316572189331,"predicted_mos":3.550316572189331,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":95400,"input_sample_rate":16000,"input_seconds":5.9625,"input_float32_sha256":"3c5d0cdf084a3a02b80859f16d1a42ffc91deb48b458a520ab4ade716fb9dd13","internal_samples":95400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5220789309987595},"started_at":"2026-10-13T04:47:16.361720+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3565999981656205e-05,"native_input_samples":95360,"native_input_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","output_sha256":"bebb80051ae24e286441f9c52f0a33cd284b45845e18144a54c56ce65d36a8ee","output_samples":95400,"length_delta_samples":40,"analysis_s":1.560128108999379,"synthesis_s":12.788915359000384,"total_s":14.349043467999763,"rtf":2.4075576288590206,"output_peak":0.5608283877372742,"output_rms":0.04201401646240844,"finished_at":"2026-10-13T04:47:30.719025+00:00","evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"bebb80051ae24e286441f9c52f0a33cd284b45845e18144a54c56ce65d36a8ee","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":95360,"output_samples":95400,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.53211597453373,"output_peak":0.5608283877372742},"common_prefix":{"samples":95360,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.99958071278826,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.167343866059595,"spectral_convergence":0.38507731785436133,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.857410736653584,"spectral_convergence":0.40663545366119236,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.900479501079164,"spectral_convergence":0.41885695702528825,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":7.396503662795931e-06,"delta_db":1.668663167409691,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.00014541223118301987,"delta_db":0.016807982376558075,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0004398498290484104,"delta_db":-0.42363904492488547,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0005923759398367358,"delta_db":-0.6646272295650772,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.00028044606166336794,"delta_db":-0.212799171430533,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.00030043726165500324,"delta_db":-1.3507067543069886,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7228738916689555,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6046328246740023,"pitch":null,"output_rms_dbfs":-27.530295091895578,"output_peak":0.5608283877372742,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.641744701264115},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95360,"output_samples":95360,"intelligibility":{"estoi":{"value":0.9050648030871322,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.045052291623524414,"rms_dbfs":-26.925662267221572,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.913020028217947,"mean":-3.6546847964293206e-06,"zero_energy":false},"output":{"rms":0.04202282507220729,"rms_dbfs":-27.530295091895578,"peak":0.5608283877372742,"peak_dbfs":-5.023400231583157,"crest_factor_db":22.506894860312418,"mean":6.11501127005988e-05,"zero_energy":false},"rms_difference_db":-0.604632824674006},"full_scale":{"reference":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":596,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":78,"quiet_samples":12480,"reference_mean_power":1.0344867731850498e-06,"output_mean_power":9.840029322429429e-07,"output_rms_dbfs":-60.0700360740704,"output_zero_energy":false,"output_peak":0.0049202051013708115,"output_energy_fraction":7.292463175085814e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":589,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":176,"both_estimated_voiced_frames":161,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.8702702702702703,"absolute_difference_cents_median":3.4234629005930306,"absolute_difference_cents_p95":16.084015475146117,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95360,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001898595271537964,"log_rms_mae_db":1.1802685284957073,"rms_envelope_nrmse":0.13438460298313906,"envelope_correlation":0.9866155478467404,"rms_gain_error_db":-0.6046362606854334,"reference_mean_power":0.002031411160407025,"output_mean_power":0.001767397384782155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.927152563814217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381525961065371,"output_power":0.0005555177178731822,"reference_active":true,"delta_db":-0.6022662532042912},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010674365906282644,"output_power":9.924464685084897e-05,"reference_active":true,"delta_db":-0.31634995199601956},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4567838102832594e-05,"output_power":3.348930987070319e-05,"reference_active":true,"delta_db":-0.13766021376068033}],"active_band_count":3,"active_band_mae_db":0.35209213965366365}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00017019756464998765,"log_rms_mae_db":0.9169634156170966,"rms_envelope_nrmse":0.119720668871523,"envelope_correlation":0.9892999017920165,"rms_gain_error_db":-0.6046584551047325,"reference_mean_power":0.0020416854120747243,"output_mean_power":0.001776327258267121},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":8.390142156345837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005509511464645055,"output_power":0.00048269176794654294,"reference_active":true,"delta_db":-0.5744319889905483},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5897699076585876e-05,"output_power":4.006770290239793e-05,"reference_active":true,"delta_db":-0.5899646935338518},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.848237969563024e-06,"output_power":1.8933219674742469e-06,"reference_active":true,"delta_db":0.10466585876866472}],"active_band_count":3,"active_band_mae_db":0.4230208470976882}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":1.6562525180598295,"rms_envelope_nrmse":0.10413913392053824,"envelope_correlation":0.9922569626686755,"rms_gain_error_db":-0.21708918588939055,"reference_mean_power":0.0005236182881812637,"output_mean_power":0.0004980877811543392},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.785583338040323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002893453145707382,"output_power":0.0002894444128060198,"reference_active":true,"delta_db":0.001487165895800028},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619778623468816e-05,"output_power":3.8739823587568156e-05,"reference_active":true,"delta_db":0.29475627718200054},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0243815325354377e-05,"output_power":1.0876909068217557e-05,"reference_active":true,"delta_db":0.2604375737081013}],"active_band_count":3,"active_band_mae_db":0.1855603389286339}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":1.211231427742759,"rms_envelope_nrmse":0.08937088555894565,"envelope_correlation":0.9939770891790097,"rms_gain_error_db":-0.21709600517863567,"reference_mean_power":0.000526268698480611,"output_mean_power":0.0005006081770649222},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.4283863314229963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002786455084699618,"output_power":0.0002800772723530377,"reference_active":true,"delta_db":0.022258212345676154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6309295594140896e-05,"output_power":1.7462868900254447e-05,"reference_active":true,"delta_db":0.2968038954909137},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.788822138965803e-07,"output_power":8.400267890331335e-07,"reference_active":true,"delta_db":0.3282134957768071}],"active_band_count":3,"active_band_mae_db":0.21575853453779895}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017019739791537767,"log_rms_mae_db":1.566279621165085,"rms_envelope_nrmse":0.15162617265472866,"envelope_correlation":0.9854608907072935,"rms_gain_error_db":-0.5461948600018987,"reference_mean_power":0.0008446325483556025,"output_mean_power":0.0007448148503799506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0046345863274685e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028725854625924984,"output_power":0.00028083764940802067,"reference_active":true,"delta_db":-0.09817628930735278},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3436307686794475e-05,"output_power":4.4810606400916277e-05,"reference_active":true,"delta_db":0.13527920207651928},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.134050674510124e-05,"output_power":1.300951613027335e-05,"reference_active":true,"delta_db":0.596286826895567}],"active_band_count":3,"active_band_mae_db":0.276580772759813}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00015333717053704136,"log_rms_mae_db":1.1943573989011886,"rms_envelope_nrmse":0.13296691853219622,"envelope_correlation":0.9888735976464097,"rms_gain_error_db":-0.5462226940440216,"reference_mean_power":0.0008489024890862698,"output_mean_power":0.0007485753767983853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.579273675799128e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027065671114903563,"output_power":0.0002688533228507939,"reference_active":true,"delta_db":-0.02903391797319673},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9395570039165576e-05,"output_power":1.8421014985315664e-05,"reference_active":true,"delta_db":-0.22388992264995228},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.660068008029238e-07,"output_power":7.867865803952209e-07,"reference_active":true,"delta_db":0.11624318461937669}],"active_band_count":3,"active_band_mae_db":0.12305567508084188}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010107776753147984,"log_rms_mae_db":1.7828192966762302,"rms_envelope_nrmse":0.2018599900562208,"envelope_correlation":0.972041909205159,"rms_gain_error_db":-0.8373634290453755,"reference_mean_power":0.00032790731816785486,"output_mean_power":0.0002704050302467847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1725104205733306e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020494142716961813,"output_power":0.00015720318953913408,"reference_active":true,"delta_db":-1.151684029037539},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.658028435672292e-05,"output_power":3.070503013495498e-05,"reference_active":true,"delta_db":-0.7603754919059157},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.565019545039448e-06,"output_power":9.211634508248372e-06,"reference_active":true,"delta_db":-0.16349163745700515}],"active_band_count":3,"active_band_mae_db":0.6918503861334866}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":1.4676768031989134,"rms_envelope_nrmse":0.16603561542629067,"envelope_correlation":0.9815453548686589,"rms_gain_error_db":-0.8374076787663604,"reference_mean_power":0.0003295653902210023,"output_mean_power":0.0002717695714153781},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":2.853826866715285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019529233091956366,"output_power":0.00015046873828937734,"reference_active":true,"delta_db":-1.1323890964771492},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.517039000031073e-05,"output_power":1.3174569099502452e-05,"reference_active":true,"delta_db":-0.6126032611895894},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.363045655471933e-07,"output_power":7.09508109332498e-07,"reference_active":true,"delta_db":-0.16100129785009606}],"active_band_count":3,"active_band_mae_db":0.6353312185056116}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.583444595336914,"predicted_mos":2.583444595336914,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":263168,"input_sample_rate":44100,"input_seconds":5.9675283446712015,"input_float32_sha256":"50cd5cda4d4c7cccfacb7e81b3dcd60094ce5a63edb04869f317d89efb0bad86","internal_samples":95481,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":21.22736077100126},"started_at":"2026-10-13T07:03:44.645287+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00886144099968078,"native_input_samples":262836,"native_input_sha256":"c109029ec5e03c948308d63bad9c2826db09ebe2a8ac965039cc45d01e1c60bf","output_sha256":"c056027bedcee938f86b25558e90518532cf7b42b0df102ecf9dc066adf4eaa0","output_samples":263168,"length_delta_samples":332,"analysis_s":0.250443565999376,"synthesis_s":38.51566584200009,"total_s":38.766109407999465,"rtf":6.504380773154272,"output_peak":0.4364529252052307,"output_rms":0.04128427595017111,"finished_at":"2026-10-13T07:04:23.440874+00:00","evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"c056027bedcee938f86b25558e90518532cf7b42b0df102ecf9dc066adf4eaa0","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":95360,"output_samples":95481,"length_delta_samples":121,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.70076980485876,"output_peak":0.4353200351443125},"common_prefix":{"samples":95360,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987327321666091,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.68502394405116,"spectral_convergence":0.5354483828024234,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.840855537377669,"spectral_convergence":0.51470797929905,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.681622179427292,"spectral_convergence":0.5430194525981457,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":5.332234799585024e-06,"delta_db":0.24749104971355881,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.00012578282177747388,"delta_db":-0.6129880665833083,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0004088479210556475,"delta_db":-0.7410653825029807,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0006570921896429495,"delta_db":-0.21433828604687177,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.0002623341369722632,"delta_db":-0.5027445461687403,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.000240707729239545,"delta_db":-2.313344344649349,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7719952792774681,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7696006371984193,"pitch":null,"output_rms_dbfs":-27.69526290441999,"output_peak":0.4353200351443125,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.735833886952042},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95360,"output_samples":95360,"intelligibility":{"estoi":{"value":0.7015469962350236,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.045052291623524414,"rms_dbfs":-26.925662267221572,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.913020028217947,"mean":-3.6546847964293206e-06,"zero_energy":false},"output":{"rms":0.04123223294326254,"rms_dbfs":-27.69526290441999,"peak":0.4353200351443125,"peak_dbfs":-7.2238268887663315,"crest_factor_db":20.471436015653662,"mean":-2.8078986773958346e-05,"zero_energy":false},"rms_difference_db":-0.769600637198419},"full_scale":{"reference":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":596,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":78,"quiet_samples":12480,"reference_mean_power":1.0344867731850498e-06,"output_mean_power":7.334109576678668e-07,"output_rms_dbfs":-61.346526056664736,"output_zero_energy":false,"output_peak":0.009001090421160103,"output_energy_fraction":5.645754998756819e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":589,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":156,"both_estimated_voiced_frames":140,"reference_voiced_output_unvoiced_frames":45,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.7567567567567568,"absolute_difference_cents_median":5.4686683117685675,"absolute_difference_cents_p95":22.05137013457729,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95360,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001898595271537964,"log_rms_mae_db":1.8729877982258127,"rms_envelope_nrmse":0.15220915271939853,"envelope_correlation":0.9847930189660964,"rms_gain_error_db":-0.7696005875037255,"reference_mean_power":0.002031411160407025,"output_mean_power":0.0017015228095487171},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.927152563814217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381525961065371,"output_power":0.0005143442669325709,"reference_active":true,"delta_db":-0.93670636961746},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010674365906282644,"output_power":8.230729563031913e-05,"reference_active":true,"delta_db":-1.1290375357573665},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4567838102832594e-05,"output_power":2.5124338986870707e-05,"reference_active":true,"delta_db":-1.385775748728073}],"active_band_count":3,"active_band_mae_db":1.1505065513676331}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00017019756464998765,"log_rms_mae_db":1.5239644785841855,"rms_envelope_nrmse":0.11982418666868962,"envelope_correlation":0.9928859316881824,"rms_gain_error_db":-0.7695972935428576,"reference_mean_power":0.0020416854120747243,"output_mean_power":0.0017101298846621711},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":8.390142156345837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005509511464645055,"output_power":0.00045291691002026844,"reference_active":true,"delta_db":-0.8509455539516181},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5897699076585876e-05,"output_power":3.574608786662443e-05,"reference_active":true,"delta_db":-1.085623956622888},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.848237969563024e-06,"output_power":1.2430719444005308e-06,"reference_active":true,"delta_db":-1.7226162327052728}],"active_band_count":3,"active_band_mae_db":1.2197285810932597}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":3.1080925250832343,"rms_envelope_nrmse":0.1850846007755999,"envelope_correlation":0.979233819797762,"rms_gain_error_db":-0.8583151537453048,"reference_mean_power":0.0005236182881812637,"output_mean_power":0.00042971774809634055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.785583338040323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002893453145707382,"output_power":0.00021630847721861664,"reference_active":true,"delta_db":-1.263429141821856},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619778623468816e-05,"output_power":2.698123729128519e-05,"reference_active":true,"delta_db":-1.276201495513295},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0243815325354377e-05,"output_power":8.352292378317231e-06,"reference_active":true,"delta_db":-0.886560517110036}],"active_band_count":3,"active_band_mae_db":1.1420637181483955}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":2.3455043048741113,"rms_envelope_nrmse":0.15358070841466143,"envelope_correlation":0.9869465095018856,"rms_gain_error_db":-0.8583217193729121,"reference_mean_power":0.000526268698480611,"output_mean_power":0.0004318922070183911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.4283863314229963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002786455084699618,"output_power":0.0002083963305636565,"reference_active":true,"delta_db":-1.2616197933814233},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6309295594140896e-05,"output_power":1.2211737291791434e-05,"reference_active":true,"delta_db":-1.2565775121716316},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.788822138965803e-07,"output_power":6.014695293151996e-07,"reference_active":true,"delta_db":-1.1225815595280806}],"active_band_count":3,"active_band_mae_db":1.213592955027045}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017019739791537767,"log_rms_mae_db":2.8079819200080296,"rms_envelope_nrmse":0.19825398817537815,"envelope_correlation":0.9727400038292111,"rms_gain_error_db":-0.33713928548965477,"reference_mean_power":0.0008446325483556025,"output_mean_power":0.0007815448121814115},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0046345863274685e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028725854625924984,"output_power":0.000254377205619544,"reference_active":true,"delta_db":-0.5279476608577732},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3436307686794475e-05,"output_power":4.580762856181708e-05,"reference_active":true,"delta_db":0.23084908299716375},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.134050674510124e-05,"output_power":1.6487008590284567e-05,"reference_active":true,"delta_db":1.6250940298936192}],"active_band_count":3,"active_band_mae_db":0.7946302579161854}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00015333717053704136,"log_rms_mae_db":2.2577134088580504,"rms_envelope_nrmse":0.13882084986963591,"envelope_correlation":0.9864476954772597,"rms_gain_error_db":-0.3371350591583086,"reference_mean_power":0.0008489024890862698,"output_mean_power":0.0007854965846698416},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.579273675799128e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027065671114903563,"output_power":0.00024665951161454965,"reference_active":true,"delta_db":-0.40320932976230217},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9395570039165576e-05,"output_power":2.0003375044899787e-05,"reference_active":true,"delta_db":0.13400729535960126},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.660068008029238e-07,"output_power":9.872814311145218e-07,"reference_active":true,"delta_db":1.102083434114327}],"active_band_count":3,"active_band_mae_db":0.5464333530787435}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010107776753147984,"log_rms_mae_db":3.399818977465804,"rms_envelope_nrmse":0.24843002989354096,"envelope_correlation":0.9610046479584972,"rms_gain_error_db":-1.2492143387728092,"reference_mean_power":0.00032790731816785486,"output_mean_power":0.0002459402869402599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1725104205733306e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020494142716961813,"output_power":0.0001601359598193513,"reference_active":true,"delta_db":-1.0714088902362997},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.658028435672292e-05,"output_power":2.7963350500247898e-05,"reference_active":true,"delta_db":-1.1665787084659607},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.565019545039448e-06,"output_power":7.63117004665818e-06,"reference_active":true,"delta_db":-0.980947306760763}],"active_band_count":3,"active_band_mae_db":1.072978301821008}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":3.090447165156463,"rms_envelope_nrmse":0.20949116786380478,"envelope_correlation":0.9743549815803454,"rms_gain_error_db":-1.2491973891155208,"reference_mean_power":0.0003295653902210023,"output_mean_power":0.00024718485523742467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":2.853826866715285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019529233091956366,"output_power":0.00015251358278174762,"reference_active":true,"delta_db":-1.073766655383981},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.517039000031073e-05,"output_power":1.236060282122409e-05,"reference_active":true,"delta_db":-0.8895709414832381},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.363045655471933e-07,"output_power":5.199210246882851e-07,"reference_active":true,"delta_db":-1.511201134907504}],"active_band_count":3,"active_band_mae_db":1.158179577258241}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"golf","status":"measured","mos":{"status":"measured","value":2.516199827194214,"predicted_mos":2.516199827194214,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":142800,"input_sample_rate":24000,"input_seconds":5.95,"input_float32_sha256":"c0b2509e8fccd37b864c8358c7dfae03dafbdcb80e09422409aced5257f3b08b","internal_samples":95200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0227894700001343},"started_at":"2026-10-13T05:02:02.618833+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028280400001676753,"native_input_samples":143040,"native_input_sha256":"a5f7c3acfbff85a9528bca3f869db64785de95a44484db36fde0aeccd737ae88","output_sha256":"2c463cf7641d2ad102da2d86d663e4e3afea17b2fe0fea453dac90f7b0bc4a63","output_samples":142800,"length_delta_samples":-240,"analysis_s":0.6269886230002157,"synthesis_s":0.12208756200016069,"total_s":0.7490761850003764,"rtf":0.12568392365778128,"output_peak":0.4990382492542267,"output_rms":0.04167334777204063,"finished_at":"2026-10-13T05:02:03.381187+00:00","analysis_stages":{"external_f0_s":0.0917984540001271},"evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"2c463cf7641d2ad102da2d86d663e4e3afea17b2fe0fea453dac90f7b0bc4a63","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":95360,"output_samples":95200,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.619469751840334,"output_peak":0.5170570894664893},"common_prefix":{"samples":95200,"sample_rate":16000,"reference_coverage":0.9983221476510067,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95200,"output_samples":95200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.255872566930333,"spectral_convergence":0.483491371039602,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.596762183819337,"spectral_convergence":0.3992803200729,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.680643535937484,"spectral_convergence":0.4356401924889685,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.0718951648848e-06,"output_power":8.672909236477336e-06,"delta_db":2.3299453326433204,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0001451955805689493,"output_power":0.00012015317021185027,"delta_db":-0.82218163561414,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00048560674996019786,"output_power":0.00041647292289700034,"delta_db":-0.6669794394201568,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006915025831364215,"output_power":0.000569216704520205,"delta_db":-0.8451617014936772,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029502909314189045,"output_power":0.00023338829268415053,"delta_db":-1.0178577734007015,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00041071365902962796,"output_power":0.0003821235730185368,"delta_db":-0.3133531641509227,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.999246507787153,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7010989398764933,"pitch":null,"output_rms_dbfs":-27.619469751840334,"output_peak":0.5170570894664893,"reference_rms_dbfs":-26.918370811963843,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.844426095562384},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95200,"output_samples":95200,"intelligibility":{"estoi":{"value":0.8640463453271366,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04509012709010667,"rms_dbfs":-26.918370811963843,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.905728572960214,"mean":-3.184470809808298e-06,"zero_energy":false},"output":{"rms":0.041593600139543344,"rms_dbfs":-27.619469751840334,"peak":0.5170570894664893,"peak_dbfs":-5.729230056053095,"crest_factor_db":21.89023969578724,"mean":0.00011559012011311917,"zero_energy":false},"rms_difference_db":-0.7010989398764913},"full_scale":{"reference":{"sample_count":95200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":595,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":77,"quiet_samples":12320,"reference_mean_power":1.0425502331564565e-06,"output_mean_power":1.0729175278005468e-06,"output_rms_dbfs":-59.69433659765347,"output_zero_energy":false,"output_peak":0.007363560395648982,"output_energy_fraction":8.025776748194929e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":588,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":187,"both_estimated_voiced_frames":154,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":33,"paired_voiced_fraction":0.8324324324324325,"absolute_difference_cents_median":4.598575642577757,"absolute_difference_cents_p95":14.191588558733699,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1189,"rms_floor":0.0001898595271537964,"log_rms_mae_db":1.7718499403659096,"rms_envelope_nrmse":0.1442611458843717,"envelope_correlation":0.985532327909011,"rms_gain_error_db":-0.7010981636012208,"reference_mean_power":0.002034827529876105,"output_mean_power":0.0017314812315344338},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":9.937198170833554e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000639909033815774,"output_power":0.0005211658710400933,"reference_active":true,"delta_db":-0.8914227344329719},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010503522741468661,"output_power":9.76480677569326e-05,"reference_active":true,"delta_db":-0.3167132611434297},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.486708243431572e-05,"output_power":3.19462666689136e-05,"reference_active":true,"delta_db":-0.3799549594814547}],"active_band_count":3,"active_band_mae_db":0.5293636516859521}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1183,"rms_floor":0.00017019756464998765,"log_rms_mae_db":1.3170215832368082,"rms_envelope_nrmse":0.11794400709839173,"envelope_correlation":0.9913807391548722,"rms_gain_error_db":-0.701103138395598,"reference_mean_power":0.002045136225535127,"output_mean_power":0.0017402511427373096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1183,"frequency_resolution_hz":0.16906170752324598,"power_floor":8.396696959578792e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005517849376647123,"output_power":0.0004506680369638568,"reference_active":true,"delta_db":-0.8791308376254083},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.504419350609339e-05,"output_power":4.072386339194981e-05,"reference_active":true,"delta_db":-0.43789843560193353},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8366556698998151e-06,"output_power":1.7240896504201485e-06,"reference_active":true,"delta_db":-0.27467899026872655}],"active_band_count":3,"active_band_mae_db":0.5305694211653561}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1189,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":2.0446351846825936,"rms_envelope_nrmse":0.20518438464737881,"envelope_correlation":0.9718528178336006,"rms_gain_error_db":-0.7692970634754809,"reference_mean_power":0.000524498978864778,"output_mean_power":0.00043935436011991705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":3.791467810186333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002903346019676198,"output_power":0.0001878192208538864,"reference_active":true,"delta_db":-1.8915876337285686},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.549190312135195e-05,"output_power":2.406509448493108e-05,"reference_active":true,"delta_db":-1.6874171624515057},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0328670631241012e-05,"output_power":7.740990680536067e-06,"reference_active":true,"delta_db":-1.252478839166217}],"active_band_count":3,"active_band_mae_db":1.6104945451154304}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1183,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":1.4462292311744145,"rms_envelope_nrmse":0.18236905608549642,"envelope_correlation":0.9772266724837466,"rms_gain_error_db":-0.7692943838502948,"reference_mean_power":0.0005271583048056999,"output_mean_power":0.0004415822564527962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1183,"frequency_resolution_hz":0.16906170752324598,"power_floor":3.433520374920014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002793382391683558,"output_power":0.0001799347889097889,"reference_active":true,"delta_db":-1.91015252482154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.591323127144436e-05,"output_power":1.0829697686434847e-05,"reference_active":true,"delta_db":-1.6714204118574782},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.797510540762293e-07,"output_power":5.545698690014451e-07,"reference_active":true,"delta_db":-1.479997008036054}],"active_band_count":3,"active_band_mae_db":1.6871899815716909}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1189,"rms_floor":0.00017019739791537767,"log_rms_mae_db":1.9847090333640045,"rms_envelope_nrmse":0.15868527357470277,"envelope_correlation":0.9879466157879385,"rms_gain_error_db":-0.9021507125549122,"reference_mean_power":0.0008460531391691786,"output_mean_power":0.0006873573325078692},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":4.013081739201161e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002882368338309994,"output_power":0.0002299450934059035,"reference_active":true,"delta_db":-0.9812533160077204},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2841437678241196e-05,"output_power":3.27715930009318e-05,"reference_active":true,"delta_db":-1.163664839090813},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1417222017347204e-05,"output_power":9.985291873552146e-06,"reference_active":true,"delta_db":-0.5819968238196783}],"active_band_count":3,"active_band_mae_db":0.9089716596394041}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1183,"rms_floor":0.00015333717053704136,"log_rms_mae_db":1.5580484062604605,"rms_envelope_nrmse":0.128646162634652,"envelope_correlation":0.9943436268125344,"rms_gain_error_db":-0.902145208104647,"reference_mean_power":0.0008503375006346524,"output_mean_power":0.000690838943693843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1183,"frequency_resolution_hz":0.16906170752324598,"power_floor":3.5866718630807327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027135837092556546,"output_power":0.0002184721830591256,"reference_active":true,"delta_db":-0.9414707499756012},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9034863476338416e-05,"output_power":1.415433492041232e-05,"reference_active":true,"delta_db":-1.2866029858988894},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.646971434371503e-07,"output_power":6.314943198111692e-07,"reference_active":true,"delta_db":-0.8312001936349923}],"active_band_count":3,"active_band_mae_db":1.019757976503161}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1189,"rms_floor":0.00010107776753147984,"log_rms_mae_db":2.5225884369660463,"rms_envelope_nrmse":0.19352279785881116,"envelope_correlation":0.9740823441638796,"rms_gain_error_db":0.05302107849372506,"reference_mean_power":0.00032845883849953733,"output_mean_power":0.0003324934242464473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":3.1736999031947997e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020485553603964853,"output_power":0.00024385947141971156,"reference_active":true,"delta_db":0.7569192323457542},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.617753249462584e-05,"output_power":4.1292232824855204e-05,"reference_active":true,"delta_db":0.5742942515271846},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.931586116400226e-06,"output_power":1.2315244167016826e-05,"reference_active":true,"delta_db":0.9342441405467796}],"active_band_count":3,"active_band_mae_db":0.7551525414732394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95200,"uncovered_tail_samples":0,"metrics":{"frame_count":1183,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":2.062903651265499,"rms_envelope_nrmse":0.15359845634071734,"envelope_correlation":0.9827532848452327,"rms_gain_error_db":0.05303810357822282,"reference_mean_power":0.0003301225084893171,"output_mean_power":0.0003341788397675561},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1183,"frequency_resolution_hz":0.16906170752324598,"power_floor":2.854553902363997e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019510956970136345,"output_power":0.00023580137726156623,"reference_active":true,"delta_db":0.8226776626094034},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5162661110276339e-05,"output_power":1.765846163041248e-05,"reference_active":true,"delta_db":0.6617743755096641},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.283989958442329e-07,"output_power":9.439588207710677e-07,"reference_active":true,"delta_db":1.1258371064901023}],"active_band_count":3,"active_band_mae_db":0.8700963815363899}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.0319275856018066,"predicted_mos":2.0319275856018066,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143160,"input_sample_rate":24000,"input_seconds":5.965,"input_float32_sha256":"8b78adfcda7d4a67e79d9278b59b449a7b90f1d830e60107aad1569f28d7af7d","internal_samples":95440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0527970049988653},"started_at":"2026-10-13T05:06:07.542356+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028431680002540816,"native_input_samples":143040,"native_input_sha256":"a5f7c3acfbff85a9528bca3f869db64785de95a44484db36fde0aeccd737ae88","output_sha256":"a2c5be076ab6a5998a22bd9fc6e7b6ce15ae8d18519067c4e2ac383abf896f0a","output_samples":143160,"length_delta_samples":120,"analysis_s":2.7644581900003686,"synthesis_s":14.934745014999862,"total_s":17.69920320500023,"rtf":2.969664967281918,"output_peak":0.4731246531009674,"output_rms":0.04368074828636741,"finished_at":"2026-10-13T05:06:25.253109+00:00","analysis_stages":{"external_f0_s":1.7615532110003187},"evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"a2c5be076ab6a5998a22bd9fc6e7b6ce15ae8d18519067c4e2ac383abf896f0a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":95360,"output_samples":95440,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.20517244247057,"output_peak":0.472612174906842},"common_prefix":{"samples":95360,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991617770326907,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.502507245182003,"spectral_convergence":0.5440559416775269,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.774742809541571,"spectral_convergence":0.4461609956706461,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.541242190362339,"spectral_convergence":0.446221163668845,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":6.515154540943759e-06,"delta_db":1.1176456182311036,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.00014531726466590487,"delta_db":0.013970744572731068,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0005102345604661744,"delta_db":0.22101539968613143,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.000658038094414654,"delta_db":-0.20809097658140913,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.0002757092800471625,"delta_db":-0.28677878494895664,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.00030897466141421805,"delta_db":-1.2290160537224541,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5127529296237977,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2758689534352752,"pitch":null,"output_rms_dbfs":-27.20153122065685,"output_peak":0.472612174906842,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9394974150286375},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95360,"output_samples":95360,"intelligibility":{"estoi":{"value":0.8499355354807513,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.045052291623524414,"rms_dbfs":-26.925662267221572,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.913020028217947,"mean":-3.6546847964293206e-06,"zero_energy":false},"output":{"rms":0.043643888639179,"rms_dbfs":-27.20153122065685,"peak":0.472612174906842,"peak_dbfs":-6.509901895005992,"crest_factor_db":20.691629325650855,"mean":-6.489632059292136e-07,"zero_energy":false},"rms_difference_db":-0.2758689534352783},"full_scale":{"reference":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":596,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":78,"quiet_samples":12480,"reference_mean_power":1.0344867731850498e-06,"output_mean_power":4.8013705603992035e-06,"output_rms_dbfs":-53.186347747377276,"output_zero_energy":false,"output_peak":0.08878469920281018,"output_energy_fraction":0.0003298881309039933,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":589,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":168,"both_estimated_voiced_frames":146,"reference_voiced_output_unvoiced_frames":39,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.7891891891891892,"absolute_difference_cents_median":5.574738508371392,"absolute_difference_cents_p95":22.520147483556595,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95360,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001898595271537964,"log_rms_mae_db":2.460346058915399,"rms_envelope_nrmse":0.176975462127853,"envelope_correlation":0.9713772265763623,"rms_gain_error_db":-0.27586829249627515,"reference_mean_power":0.002031411160407025,"output_mean_power":0.0019063867236118932},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.927152563814217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381525961065371,"output_power":0.0006029649899120111,"reference_active":true,"delta_db":-0.2463244401902798},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010674365906282644,"output_power":9.696694216944133e-05,"reference_active":true,"delta_db":-0.4171838544284884},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4567838102832594e-05,"output_power":2.971760358575919e-05,"reference_active":true,"delta_db":-0.6565843411805306}],"active_band_count":3,"active_band_mae_db":0.44003087859976625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00017019756464998765,"log_rms_mae_db":1.7381231956461172,"rms_envelope_nrmse":0.12090464040006144,"envelope_correlation":0.9857463804522839,"rms_gain_error_db":-0.2758640141304118,"reference_mean_power":0.0020416854120747243,"output_mean_power":0.0019160305277432076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":8.390142156345837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005509511464645055,"output_power":0.0005230488228570574,"reference_active":true,"delta_db":-0.22570862082509868},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5897699076585876e-05,"output_power":4.1636586454673593e-05,"reference_active":true,"delta_db":-0.4231579724158108},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.848237969563024e-06,"output_power":1.5220639594126226e-06,"reference_active":true,"delta_db":-0.8432498547167765}],"active_band_count":3,"active_band_mae_db":0.49737214931922863}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":2.644378175927383,"rms_envelope_nrmse":0.1818701875676518,"envelope_correlation":0.9766005077651579,"rms_gain_error_db":0.19581788084686746,"reference_mean_power":0.0005236182881812637,"output_mean_power":0.0005477679205149418},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.785583338040323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002893453145707382,"output_power":0.00030325493979501084,"reference_active":true,"delta_db":0.20391429791849908},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619778623468816e-05,"output_power":3.732382073742541e-05,"reference_active":true,"delta_db":0.13304083898033583},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0243815325354377e-05,"output_power":1.0589572901742563e-05,"reference_active":true,"delta_db":0.14416704069551067}],"active_band_count":3,"active_band_mae_db":0.1603740591981152}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":1.9368133086155959,"rms_envelope_nrmse":0.12248976668268448,"envelope_correlation":0.9887525371999678,"rms_gain_error_db":0.19581906663468765,"reference_mean_power":0.000526268698480611,"output_mean_power":0.0005505407198621075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.4283863314229963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002786455084699618,"output_power":0.00029211231759538117,"reference_active":true,"delta_db":0.20497823408018842},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6309295594140896e-05,"output_power":1.701040430950222e-05,"reference_active":true,"delta_db":0.18279432135027962},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.788822138965803e-07,"output_power":7.683296115510655e-07,"reference_active":true,"delta_db":-0.05924215363101649}],"active_band_count":3,"active_band_mae_db":0.14900490302049485}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017019739791537767,"log_rms_mae_db":2.388502953571051,"rms_envelope_nrmse":0.17893182606773778,"envelope_correlation":0.9771656374588081,"rms_gain_error_db":-0.19662713894800854,"reference_mean_power":0.0008446325483556025,"output_mean_power":0.0008072445084772268},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0046345863274685e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028725854625924984,"output_power":0.0002797687815857603,"reference_active":true,"delta_db":-0.11473706827541691},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3436307686794475e-05,"output_power":4.069531493812663e-05,"reference_active":true,"delta_db":-0.283084870444963},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.134050674510124e-05,"output_power":1.0621738966271806e-05,"reference_active":true,"delta_db":-0.2843683698731319}],"active_band_count":3,"active_band_mae_db":0.2273967695311706}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00015333717053704136,"log_rms_mae_db":1.5403749551803014,"rms_envelope_nrmse":0.11656113874679683,"envelope_correlation":0.9898588551356363,"rms_gain_error_db":-0.19662043888501385,"reference_mean_power":0.0008489024890862698,"output_mean_power":0.0008113266900302316},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.579273675799128e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027065671114903563,"output_power":0.0002643429065231387,"reference_active":true,"delta_db":-0.10251139362464809},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9395570039165576e-05,"output_power":1.790526448782258e-05,"reference_active":true,"delta_db":-0.34721807532460325},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.660068008029238e-07,"output_power":7.357463988081649e-07,"reference_active":true,"delta_db":-0.17504480351784588}],"active_band_count":3,"active_band_mae_db":0.20825809082236577}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010107776753147984,"log_rms_mae_db":2.5361109177646157,"rms_envelope_nrmse":0.2508392946941701,"envelope_correlation":0.9509872944985869,"rms_gain_error_db":-0.6311462618016515,"reference_mean_power":0.00032790731816785486,"output_mean_power":0.00028355446028817616},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1725104205733306e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020494142716961813,"output_power":0.00017438201421753424,"reference_active":true,"delta_db":-0.7012806649786738},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.658028435672292e-05,"output_power":3.161770156171156e-05,"reference_active":true,"delta_db":-0.6331678128939787},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.565019545039448e-06,"output_power":9.126953404808728e-06,"reference_active":true,"delta_db":-0.2036002843322387}],"active_band_count":3,"active_band_mae_db":0.5126829207349638}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":1.5988497616391824,"rms_envelope_nrmse":0.1555357465058329,"envelope_correlation":0.9818448852882653,"rms_gain_error_db":-0.6311475029492607,"reference_mean_power":0.0003295653902210023,"output_mean_power":0.000284988179493547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":2.853826866715285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019529233091956366,"output_power":0.00016638761722138323,"reference_active":true,"delta_db":-0.6956418658000572},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.517039000031073e-05,"output_power":1.3609022203853449e-05,"reference_active":true,"delta_db":-0.47169823118794},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.363045655471933e-07,"output_power":6.545034724248548e-07,"reference_active":true,"delta_db":-0.5114553836798666}],"active_band_count":3,"active_band_mae_db":0.5595984935559546}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0883829593658447,"predicted_mos":3.0883829593658447,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143280,"input_sample_rate":24000,"input_seconds":5.97,"input_float32_sha256":"5f8ffaea8108cb30a72da299d76aa37db846460b9211676e5137435edbc345f6","internal_samples":95520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.111037928998485},"started_at":"2026-10-13T05:10:10.239549+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003836750000118627,"native_input_samples":143040,"native_input_sha256":"a5f7c3acfbff85a9528bca3f869db64785de95a44484db36fde0aeccd737ae88","output_sha256":"bd14c82a8b05493b5786625533a17b9a08d3a3615078e0ab068db4e6590fb00b","output_samples":143280,"length_delta_samples":240,"analysis_s":0.6519183899999916,"synthesis_s":0.14608139899974049,"total_s":0.7979997889997321,"rtf":0.1338925820469349,"output_peak":0.5372427701950073,"output_rms":0.04368118422124363,"finished_at":"2026-10-13T05:10:11.052280+00:00","analysis_stages":{"external_f0_s":0.15298081999935675},"evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"bd14c82a8b05493b5786625533a17b9a08d3a3615078e0ab068db4e6590fb00b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":95360,"output_samples":95520,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.209442514467987,"output_peak":0.5379305317383541},"common_prefix":{"samples":95360,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998324958123953,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.826610035028939,"spectral_convergence":0.2942029984082377,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.597387398108219,"spectral_convergence":0.27805733477950206,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.031591594014801,"spectral_convergence":0.3155266045718395,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":5.0011842788932034e-06,"delta_db":-0.030873025977997265,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.00013993076495448807,"delta_db":-0.15006931723039127,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0004697508311702144,"delta_db":-0.13800774799353915,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0006591533635617352,"delta_db":-0.20073661136400953,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.00027411544191988004,"delta_db":-0.3119576106708104,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.0003565607490009017,"delta_db":-0.6069093699650479,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23975894720029922,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.27649983492415386,"pitch":null,"output_rms_dbfs":-27.20216210214573,"output_peak":0.5379305317383541,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.818529675717319},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95360,"output_samples":95360,"intelligibility":{"estoi":{"value":0.9189375805570342,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.045052291623524414,"rms_dbfs":-26.925662267221572,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.913020028217947,"mean":-3.6546847964293206e-06,"zero_energy":false},"output":{"rms":0.04364071877141935,"rms_dbfs":-27.20216210214573,"peak":0.5379305317383541,"peak_dbfs":-5.38547610863022,"crest_factor_db":21.816685993515506,"mean":-4.23643123665471e-06,"zero_energy":false},"rms_difference_db":-0.2764998349241594},"full_scale":{"reference":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":596,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":78,"quiet_samples":12480,"reference_mean_power":1.0344867731850498e-06,"output_mean_power":1.7037848766929209e-06,"output_rms_dbfs":-57.68585241003049,"output_zero_energy":false,"output_peak":0.03379455425013637,"output_energy_fraction":0.00011707908297760304,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":589,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":183,"both_estimated_voiced_frames":168,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9081081081081082,"absolute_difference_cents_median":1.4198370922950438,"absolute_difference_cents_p95":5.374832404385845,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95360,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001898595271537964,"log_rms_mae_db":0.7982661587863661,"rms_envelope_nrmse":0.06397179722964595,"envelope_correlation":0.9970402934905029,"rms_gain_error_db":-0.27649858201118194,"reference_mean_power":0.002031411160407025,"output_mean_power":0.0019061100706896277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.927152563814217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381525961065371,"output_power":0.0005901462097877556,"reference_active":true,"delta_db":-0.3396491814165093},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010674365906282644,"output_power":9.29291489305757e-05,"reference_active":true,"delta_db":-0.6019012610496219},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4567838102832594e-05,"output_power":2.62083212671201e-05,"reference_active":true,"delta_db":-1.2023301551994152}],"active_band_count":3,"active_band_mae_db":0.7146268658885155}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00017019756464998765,"log_rms_mae_db":0.5122621770031938,"rms_envelope_nrmse":0.04569874702914072,"envelope_correlation":0.9988539703468191,"rms_gain_error_db":-0.27649868464807337,"reference_mean_power":0.0020416854120747243,"output_mean_power":0.0019157505427822418},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":8.390142156345837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005509511464645055,"output_power":0.0005094858448839741,"reference_active":true,"delta_db":-0.3398096867150979},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5897699076585876e-05,"output_power":4.002442343997923e-05,"reference_active":true,"delta_db":-0.594658296861929},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.848237969563024e-06,"output_power":1.3862271705947794e-06,"reference_active":true,"delta_db":-1.2492348105958837}],"active_band_count":3,"active_band_mae_db":0.7279009313909702}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":1.0903997657270497,"rms_envelope_nrmse":0.058574518070828814,"envelope_correlation":0.9976355910056284,"rms_gain_error_db":-0.1475536610175782,"reference_mean_power":0.0005236182881812637,"output_mean_power":0.000506126923692456},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.785583338040323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002893453145707382,"output_power":0.0002794405564120275,"reference_active":true,"delta_db":-0.15127016752958752},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619778623468816e-05,"output_power":3.2800211628332715e-05,"reference_active":true,"delta_db":-0.42805365183314026},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0243815325354377e-05,"output_power":6.768743459899714e-06,"reference_active":true,"delta_db":-1.7995368609711695}],"active_band_count":3,"active_band_mae_db":0.792953560111299}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":0.5839078429800496,"rms_envelope_nrmse":0.037505957219620925,"envelope_correlation":0.9990937991999994,"rms_gain_error_db":-0.14755350287311245,"reference_mean_power":0.000526268698480611,"output_mean_power":0.0005086888160877883},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.4283863314229963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002786455084699618,"output_power":0.00026767649105307326,"reference_active":true,"delta_db":-0.17441816460204487},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6309295594140896e-05,"output_power":1.4596030484016084e-05,"reference_active":true,"delta_db":-0.48200442346352224},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.788822138965803e-07,"output_power":4.914580453496777e-07,"reference_active":true,"delta_db":-1.99985337632308}],"active_band_count":3,"active_band_mae_db":0.8854253214628823}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017019739791537767,"log_rms_mae_db":0.9832055496490846,"rms_envelope_nrmse":0.06283402956567935,"envelope_correlation":0.9976170677670492,"rms_gain_error_db":-0.2268346669742543,"reference_mean_power":0.0008446325483556025,"output_mean_power":0.0008016491685722948},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0046345863274685e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028725854625924984,"output_power":0.00027258989162328166,"reference_active":true,"delta_db":-0.22763211320130516},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3436307686794475e-05,"output_power":3.792541659868059e-05,"reference_active":true,"delta_db":-0.5892254077645795},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.134050674510124e-05,"output_power":7.793899474826659e-06,"reference_active":true,"delta_db":-1.6287766123819856}],"active_band_count":3,"active_band_mae_db":0.8152113777826234}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00015333717053704136,"log_rms_mae_db":0.5981523643738339,"rms_envelope_nrmse":0.04031965610793144,"envelope_correlation":0.9992718022103168,"rms_gain_error_db":-0.22683487234071703,"reference_mean_power":0.0008489024890862698,"output_mean_power":0.0008057017737710396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.579273675799128e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027065671114903563,"output_power":0.0002556684832278617,"reference_active":true,"delta_db":-0.2474160527389177},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9395570039165576e-05,"output_power":1.6957517576809496e-05,"reference_active":true,"delta_db":-0.5834027222145686},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.660068008029238e-07,"output_power":5.32627651139119e-07,"reference_active":true,"delta_db":-1.5780891652399973}],"active_band_count":3,"active_band_mae_db":0.8029693133978278}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010107776753147984,"log_rms_mae_db":1.1330035479529441,"rms_envelope_nrmse":0.10458086978897324,"envelope_correlation":0.991911138903717,"rms_gain_error_db":-0.2687575514271639,"reference_mean_power":0.00032790731816785486,"output_mean_power":0.00030823031784689214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1725104205733306e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020494142716961813,"output_power":0.00019267480932959034,"reference_active":true,"delta_db":-0.26804818336036756},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.658028435672292e-05,"output_power":3.378990329745785e-05,"reference_active":true,"delta_db":-0.34460128260930356},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.565019545039448e-06,"output_power":8.78408610629785e-06,"reference_active":true,"delta_db":-0.3698927748298271}],"active_band_count":3,"active_band_mae_db":0.32751408026649936}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":0.7198791511934851,"rms_envelope_nrmse":0.059363267044435314,"envelope_correlation":0.9976644128633281,"rms_gain_error_db":-0.2687568402682199,"reference_mean_power":0.0003295653902210023,"output_mean_power":0.00030978894335397753},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":2.853826866715285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019529233091956366,"output_power":0.0001843066237355894,"reference_active":true,"delta_db":-0.25144245521092035},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.517039000031073e-05,"output_power":1.4150745652061336e-05,"reference_active":true,"delta_db":-0.30217420818900287},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.363045655471933e-07,"output_power":6.196173651854382e-07,"reference_active":true,"delta_db":-0.7493391273440482}],"active_band_count":3,"active_band_mae_db":0.4343185969146572}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0016","method":"magphase","status":"measured","mos":{"status":"measured","value":2.313089609146118,"predicted_mos":2.313089609146118,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":286772,"input_sample_rate":48000,"input_seconds":5.9744166666666665,"input_float32_sha256":"8ce5de88e295fea894c5233a53a418054dc83ef9a0e3e7c69887515ee9fb5f65","internal_samples":95591,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8072553039983177},"started_at":"2026-10-13T05:13:03.216843+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.013668841000253451,"native_input_samples":286080,"native_input_sha256":"27efe9267682869aa9283dc237725bcd9626e272d631a33df0e3afc6c1c1521c","output_sha256":"8f80757a2cb9de1016846fd3bbd48c3592b7213b54cf545478d38f8ff5bbb316","output_samples":286772,"length_delta_samples":692,"analysis_s":3.643017905000306,"synthesis_s":1.1455579489993397,"total_s":4.788575853999646,"rtf":0.8034523244965849,"output_peak":0.5794947641945967,"output_rms":0.04462627213418575,"finished_at":"2026-10-13T05:13:08.030856+00:00","evaluation_input":{"reference":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"8f80757a2cb9de1016846fd3bbd48c3592b7213b54cf545478d38f8ff5bbb316","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":95360,"output_samples":95591,"length_delta_samples":231,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.024992006057616,"output_peak":0.5807495318456347},"common_prefix":{"samples":95360,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975834545093157,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95360,"output_samples":95360,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.728254491133367,"spectral_convergence":0.5725764920258309,"spectral_convergence_status":"ok","amplitude_floor":1.2569032294341943e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.732923204609833,"spectral_convergence":0.3583454226764753,"spectral_convergence_status":"ok","amplitude_floor":9.073914887165124e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.672217617167645,"spectral_convergence":0.37128394787362645,"spectral_convergence_status":"ok","amplitude_floor":7.969434026985316e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.036863249031963e-06,"output_power":4.853947692803529e-06,"delta_db":-0.1606506794200821,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00014485054714414723,"output_power":0.00013342865158151492,"delta_db":-0.3567104289968434,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004849180298355377,"output_power":0.0004560086918258615,"delta_db":-0.2669521132481508,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006903353830986323,"output_power":0.0006809760602161118,"delta_db":-0.05928289182713372,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00029452981369653373,"output_power":0.0003045661837465715,"delta_db":0.14552418891851038,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004100383435072048,"output_power":0.00040878493798512163,"delta_db":-0.013295848794169932,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1670693585341484,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08882303850579426,"pitch":null,"output_rms_dbfs":-27.01448530572737,"output_peak":0.5807495318456347,"reference_rms_dbfs":-26.925662267221572,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.044465104303614},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95360,"output_samples":95360,"intelligibility":{"estoi":{"value":0.8843405161283412,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.045052291623524414,"rms_dbfs":-26.925662267221572,"peak":0.5615234375,"peak_dbfs":-5.01264223900363,"crest_factor_db":21.913020028217947,"mean":-3.6546847964293206e-06,"zero_energy":false},"output":{"rms":0.04459392865680243,"rms_dbfs":-27.01448530572737,"peak":0.5807495318456347,"peak_dbfs":-4.720222632429338,"crest_factor_db":22.294262673298032,"mean":0.00016932601304573842,"zero_energy":false},"rms_difference_db":-0.08882303850579731},"full_scale":{"reference":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95360,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":596,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.21956716128625e-06,"threshold_dbfs":-54.9220251094403,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":78,"quiet_samples":12480,"reference_mean_power":1.0344867731850498e-06,"output_mean_power":9.388134172165177e-07,"output_rms_dbfs":-60.274207122333756,"output_zero_energy":false,"output_peak":0.008763403377330592,"output_energy_fraction":6.178401983988862e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":589,"reference_estimated_voiced_frames":185,"output_estimated_voiced_frames":169,"both_estimated_voiced_frames":158,"reference_voiced_output_unvoiced_frames":27,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8540540540540541,"absolute_difference_cents_median":3.019194179822306,"absolute_difference_cents_p95":12.037000696431347,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95360,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001898595271537964,"log_rms_mae_db":1.4174210807916505,"rms_envelope_nrmse":0.12096065774250125,"envelope_correlation":0.986559595721064,"rms_gain_error_db":-0.08882022101328253,"reference_mean_power":0.002031411160407025,"output_mean_power":0.0019902874842915507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.927152563814217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006381525961065371,"output_power":0.0006223273520467298,"reference_active":true,"delta_db":-0.10905651179284588},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010674365906282644,"output_power":0.00010810911764618346,"reference_active":true,"delta_db":0.055202368917564026},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4567838102832594e-05,"output_power":4.3142413470534166e-05,"reference_active":true,"delta_db":0.9623221741816885}],"active_band_count":3,"active_band_mae_db":0.3755270182973662}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00017019756464998765,"log_rms_mae_db":0.8503147363734456,"rms_envelope_nrmse":0.06693401449166367,"envelope_correlation":0.9954777645591262,"rms_gain_error_db":-0.08882399071967507,"reference_mean_power":0.0020416854120747243,"output_mean_power":0.002000352008757517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":8.390142156345837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005509511464645055,"output_power":0.0005396815231269578,"reference_active":true,"delta_db":-0.08975541574497066},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5897699076585876e-05,"output_power":4.483939483518481e-05,"reference_active":true,"delta_db":-0.10131171616973217},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.848237969563024e-06,"output_power":2.3941206208817523e-06,"reference_active":true,"delta_db":1.1238813930419342}],"active_band_count":3,"active_band_mae_db":0.4383161749855457}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.246932648014587e-05,"log_rms_mae_db":1.9678187292426066,"rms_envelope_nrmse":0.10989384948609876,"envelope_correlation":0.9916163666768872,"rms_gain_error_db":-0.2838297923142191,"reference_mean_power":0.0005236182881812637,"output_mean_power":0.0004904918858821029},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.785583338040323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002893453145707382,"output_power":0.0002723191362649197,"reference_active":true,"delta_db":-0.2633829321281245},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619778623468816e-05,"output_power":3.395393494942389e-05,"reference_active":true,"delta_db":-0.27791898712550733},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0243815325354377e-05,"output_power":9.631933898721436e-06,"reference_active":true,"delta_db":-0.26748246960425315}],"active_band_count":3,"active_band_mae_db":0.26959479628596167}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":8.334521268786246e-05,"log_rms_mae_db":1.276979389264734,"rms_envelope_nrmse":0.07036540320644062,"envelope_correlation":0.9967851466535773,"rms_gain_error_db":-0.2838285485803045,"reference_mean_power":0.000526268698480611,"output_mean_power":0.0004929747607146771},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.4283863314229963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002786455084699618,"output_power":0.0002623889510219004,"reference_active":true,"delta_db":-0.26106503675536274},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6309295594140896e-05,"output_power":1.5247836524203058e-05,"reference_active":true,"delta_db":-0.29226976928575477},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.788822138965803e-07,"output_power":6.429788418530364e-07,"reference_active":true,"delta_db":-0.8327510458283356}],"active_band_count":3,"active_band_mae_db":0.46202861728981776}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017019739791537767,"log_rms_mae_db":1.5904911487009328,"rms_envelope_nrmse":0.12156167558091893,"envelope_correlation":0.9895288328551481,"rms_gain_error_db":-0.052250369522396124,"reference_mean_power":0.0008446325483556025,"output_mean_power":0.0008345315809693128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0046345863274685e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028725854625924984,"output_power":0.00028874289281467477,"reference_active":true,"delta_db":0.022383449120831414},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3436307686794475e-05,"output_power":4.340204443813105e-05,"reference_active":true,"delta_db":-0.0034271356456400904},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.134050674510124e-05,"output_power":1.3472061513567905e-05,"reference_active":true,"delta_db":0.7480159592208663}],"active_band_count":3,"active_band_mae_db":0.25794218132911256}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":0.00015333717053704136,"log_rms_mae_db":0.9559660489362413,"rms_envelope_nrmse":0.0654448303896678,"envelope_correlation":0.9967749577402283,"rms_gain_error_db":-0.0522602418363097,"reference_mean_power":0.0008489024890862698,"output_mean_power":0.0008387485508125144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":3.579273675799128e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027065671114903563,"output_power":0.000273571089303246,"reference_active":true,"delta_db":0.04651399463336112},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9395570039165576e-05,"output_power":1.9129930757170263e-05,"reference_active":true,"delta_db":-0.05989150058923743},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.660068008029238e-07,"output_power":9.302232123569224e-07,"reference_active":true,"delta_db":0.8435454706657324}],"active_band_count":3,"active_band_mae_db":0.316650321962777}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010107776753147984,"log_rms_mae_db":1.6587628974060107,"rms_envelope_nrmse":0.19230184841118486,"envelope_correlation":0.975512599386982,"rms_gain_error_db":0.4976860127259194,"reference_mean_power":0.00032790731816785486,"output_mean_power":0.000367722082142127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1725104205733306e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020494142716961813,"output_power":0.00022825124348455032,"reference_active":true,"delta_db":0.4678339597762909},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.658028435672292e-05,"output_power":4.438251922568618e-05,"reference_active":true,"delta_db":0.8396487273323415},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.565019545039448e-06,"output_power":1.645569965665114e-05,"reference_active":true,"delta_db":2.3563049038056545}],"active_band_count":3,"active_band_mae_db":1.2212625303047624}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95360,"uncovered_tail_samples":0,"metrics":{"frame_count":1185,"rms_floor":7.865643818167714e-05,"log_rms_mae_db":1.033974661833643,"rms_envelope_nrmse":0.12040633453863953,"envelope_correlation":0.9910850089400256,"rms_gain_error_db":0.49768423760094216,"reference_mean_power":0.0003295653902210023,"output_mean_power":0.00036958132754370213},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1185,"frequency_resolution_hz":0.16877637130801687,"power_floor":2.853826866715285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019529233091956366,"output_power":0.00022234539566735244,"reference_active":true,"delta_db":0.5634295151148545},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.517039000031073e-05,"output_power":1.9132372941060072e-05,"reference_active":true,"delta_db":1.007720920714622},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.363045655471933e-07,"output_power":1.1818066945067722e-06,"reference_active":true,"delta_db":2.0548895239001306}],"active_band_count":3,"active_band_mae_db":1.2086799865765359}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"hnm","status":"measured","mos":{"status":"measured","value":3.1711089611053467,"predicted_mos":3.1711089611053467,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84000,"input_sample_rate":16000,"input_seconds":5.25,"input_float32_sha256":"ae41311b8054c04dcf2faffd11472e67367a71ed4093f69b77668da8f6878fcd","internal_samples":84000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.194958038999175},"started_at":"2026-10-13T04:38:24.633116+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6187000255740713e-05,"native_input_samples":84000,"native_input_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","output_sha256":"7a52ab8a0d31445f5f42e65293c16147aa19cb150749760ef26ca88ac065044f","output_samples":84000,"length_delta_samples":0,"analysis_s":0.9462754590003897,"synthesis_s":0.5071678590002193,"total_s":1.453443318000609,"rtf":0.27684634628583027,"output_peak":0.632326517553141,"output_rms":0.044041440278565644,"finished_at":"2026-10-13T04:38:26.091347+00:00","evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"7a52ab8a0d31445f5f42e65293c16147aa19cb150749760ef26ca88ac065044f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.511103943038622,"spectral_convergence":0.3421841801599182,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.463568548638778,"spectral_convergence":0.3436790293571638,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.92815304254252,"spectral_convergence":0.3546433690232922,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.7199900437332853e-05,"delta_db":0.6570809731392557,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.0002540663444014383,"delta_db":-0.10412818347334755,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0005845700579802947,"delta_db":-0.06311024464175984,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0005207498669277745,"delta_db":-0.06267601846292917,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.0003056255531033654,"delta_db":-0.032648880776575503,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.00025743673896025854,"delta_db":-0.06688823569411197,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16442208936466327,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0582229645732808,"pitch":null,"output_rms_dbfs":-27.12276973641922,"output_peak":0.632326517553141,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.634275178073306},"common_prefix":{"samples":84000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.511103943038622,"spectral_convergence":0.3421841801599182,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.463568548638778,"spectral_convergence":0.3436790293571638,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.92815304254252,"spectral_convergence":0.3546433690232922,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.7199900437332853e-05,"delta_db":0.6570809731392557,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.0002540663444014383,"delta_db":-0.10412818347334755,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0005845700579802947,"delta_db":-0.06311024464175984,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0005207498669277745,"delta_db":-0.06267601846292917,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.0003056255531033654,"delta_db":-0.032648880776575503,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.00025743673896025854,"delta_db":-0.06688823569411197,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16442208936466327,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0582229645732808,"pitch":null,"output_rms_dbfs":-27.12276973641922,"output_peak":0.632326517553141,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.634275178073306},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84000,"output_samples":84000,"intelligibility":{"estoi":{"value":0.9017836220170528,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04433764904553492,"rms_dbfs":-27.064546771845936,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.51650626449395,"mean":-1.6414097377232142e-06,"zero_energy":false},"output":{"rms":0.044041440278565644,"rms_dbfs":-27.12276973641922,"peak":0.632326517553141,"peak_dbfs":-3.9811721003948133,"crest_factor_db":23.141597636024404,"mean":-1.6537735270329845e-05,"zero_energy":false},"rms_difference_db":-0.058222964573282354},"full_scale":{"reference":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":525,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":1.5888158540131854e-06,"output_mean_power":4.216411123185783e-06,"output_rms_dbfs":-53.75057049639448,"output_zero_energy":false,"output_peak":0.041462008148028466,"output_energy_fraction":0.0002815590772999322,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":518,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":181,"both_estimated_voiced_frames":175,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.8706467661691543,"absolute_difference_cents_median":1.0128453539552127,"absolute_difference_cents_p95":5.176019189187911,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00019312594202620684,"log_rms_mae_db":0.7927887865706882,"rms_envelope_nrmse":0.07285090472646932,"envelope_correlation":0.994947961631494,"rms_gain_error_db":-0.058223611816536965,"reference_mean_power":0.0019676998400409485,"output_mean_power":0.0019414959508909835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.0378624939923077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018862042103601662,"output_power":0.00020085117275907968,"reference_active":true,"delta_db":0.2728566195190675},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.862312285918816e-05,"output_power":9.806056929847341e-05,"reference_active":true,"delta_db":-0.024843399138690086},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3789377405749525e-05,"output_power":1.131208451318233e-05,"reference_active":true,"delta_db":-0.8600201699904108}],"active_band_count":3,"active_band_mae_db":0.3859067295493894}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00018244538675777041,"log_rms_mae_db":0.4215070571558602,"rms_envelope_nrmse":0.03829168041427458,"envelope_correlation":0.9985044812353915,"rms_gain_error_db":-0.058225539141464853,"reference_mean_power":0.001979010526094432,"output_mean_power":0.0019526551458083803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":3.3102005537920435e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017836230285954771,"output_power":0.00019201976753010838,"reference_active":true,"delta_db":0.32042868627216975},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3320761072042273e-05,"output_power":4.347407490652717e-05,"reference_active":true,"delta_db":0.015342714993486228},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.132775238046075e-07,"output_power":6.825605151388418e-07,"reference_active":true,"delta_db":-0.7609760852112315}],"active_band_count":3,"active_band_mae_db":0.36558249549229577}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014606965842361196,"log_rms_mae_db":1.5382242979460012,"rms_envelope_nrmse":0.09432888085001737,"envelope_correlation":0.9929836866960503,"rms_gain_error_db":-0.09303880049691524,"reference_mean_power":0.0007714288246159527,"output_mean_power":0.0007550782877016688},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.155891354220712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.355627855804231e-05,"output_power":8.317795045559528e-05,"reference_active":true,"delta_db":-0.019708741293879208},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.117824128471733e-05,"output_power":1.9496798144556053e-05,"reference_active":true,"delta_db":-0.35926596589222604},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3214088476749303e-06,"output_power":3.5885461940060166e-06,"reference_active":true,"delta_db":0.3359620289572297}],"active_band_count":3,"active_band_mae_db":0.23831224538111165}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001346093108760191,"log_rms_mae_db":0.9072010282071523,"rms_envelope_nrmse":0.04307727073822354,"envelope_correlation":0.9985227113512994,"rms_gain_error_db":-0.09304078523318535,"reference_mean_power":0.0007758659430163304,"output_mean_power":0.0007594210137279316},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.945223621465206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.346976557991944e-05,"output_power":7.289463399834244e-05,"reference_active":true,"delta_db":-0.03413094421089826},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57892144723787e-06,"output_power":9.068125814815142e-06,"reference_active":true,"delta_db":-0.23799074732643277},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.32297115664878e-07,"output_power":2.3330236286486765e-07,"reference_active":true,"delta_db":0.01875319927931078}],"active_band_count":3,"active_band_mae_db":0.0969582969388806}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00012444218003818797,"log_rms_mae_db":1.2599915004268594,"rms_envelope_nrmse":0.0896799544650417,"envelope_correlation":0.9941227378632463,"rms_gain_error_db":-0.0896793507258316,"reference_mean_power":0.000687550351321365,"output_mean_power":0.0006734984076330823},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.3390858183888006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.254326674690696e-05,"output_power":3.9855941821644576e-05,"reference_active":true,"delta_db":-0.2833775853725571},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4859458498211297e-05,"output_power":1.5381707917654348e-05,"reference_active":true,"delta_db":0.15001576958671758},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141405389451218e-06,"output_power":3.569227440269827e-06,"reference_active":true,"delta_db":0.5545023870325667}],"active_band_count":3,"active_band_mae_db":0.3292985806639471}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001136976512596423,"log_rms_mae_db":0.7574261848042319,"rms_envelope_nrmse":0.0454779270041748,"envelope_correlation":0.9984744431000558,"rms_gain_error_db":-0.0896777830425389,"reference_mean_power":0.0006915015836222035,"output_mean_power":0.0006773691303687713},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.214722623247202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7827908938064065e-05,"output_power":3.624281646749158e-05,"reference_active":true,"delta_db":-0.1859039510363587},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.619804659862373e-06,"output_power":6.150552549173037e-06,"reference_active":true,"delta_db":0.39192913296114507},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.015023849518023e-07,"output_power":2.3937467218437754e-07,"reference_active":true,"delta_db":0.7479800580985108}],"active_band_count":3,"active_band_mae_db":0.4419377140320049}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00010848831624356853,"log_rms_mae_db":1.4478412344958749,"rms_envelope_nrmse":0.15141666697406103,"envelope_correlation":0.9822095411922329,"rms_gain_error_db":-0.015515467794379883,"reference_mean_power":0.0002381846371729614,"output_mean_power":0.00023733522444050479},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.715890316390027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.877460217795426e-05,"output_power":9.862734032732343e-05,"reference_active":true,"delta_db":0.45708586918169253},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.954888518804952e-05,"output_power":4.2428338963665114e-05,"reference_active":true,"delta_db":0.30521783783824646},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.296333400161522e-06,"output_power":5.783486393724584e-06,"reference_active":true,"delta_db":-1.009149523206812}],"active_band_count":3,"active_band_mae_db":0.5904844100755836}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":0.8827900235228942,"rms_envelope_nrmse":0.08359638925527321,"envelope_correlation":0.9940555407149374,"rms_gain_error_db":-0.015511883953335574,"reference_mean_power":0.00023955325002336745,"output_mean_power":0.00023869915352817456},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.5691901914560289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.470858674877873e-05,"output_power":0.0001050141359533558,"reference_active":true,"delta_db":0.44858407313767823},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8332956489415002e-05,"output_power":1.9709565454390556e-05,"reference_active":true,"delta_db":0.31444541593791075},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.840671065644935e-07,"output_power":3.9737728389079964e-07,"reference_active":true,"delta_db":-0.857025353693255}],"active_band_count":3,"active_band_mae_db":0.540018280922948}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.2041003704071045,"predicted_mos":3.2041003704071045,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84000,"input_sample_rate":16000,"input_seconds":5.25,"input_float32_sha256":"b83a3f95680c75b18bb69e35c1669889788f57338447ea2f5a1d4dfb12c98bfe","internal_samples":84000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.193427313999564},"started_at":"2026-10-13T04:41:22.527242+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5851000827969983e-05,"native_input_samples":84000,"native_input_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","output_sha256":"b18eefa0b58cce0334459c2fe02d631824d6e80a9a0836e06cc02b82ac78bab3","output_samples":84000,"length_delta_samples":0,"analysis_s":1.4055713959996865,"synthesis_s":0.09169144499992399,"total_s":1.4972628409996105,"rtf":0.2851929220951639,"output_peak":0.7865849114536864,"output_rms":0.050414208125235005,"finished_at":"2026-10-13T04:41:24.032265+00:00","evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"b18eefa0b58cce0334459c2fe02d631824d6e80a9a0836e06cc02b82ac78bab3","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5911975572390675,"spectral_convergence":0.606384016951396,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.339852717635891,"spectral_convergence":0.3793736639010918,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.958484953288709,"spectral_convergence":0.4213207207511064,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.868518008191147e-05,"delta_db":1.0167945213903267,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.0003779253338700017,"delta_db":1.620460482234824,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0008419447741081186,"delta_db":1.5213601441015576,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0006466033259956971,"delta_db":0.8774116471692377,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.0003660304949880647,"delta_db":0.7506271863972137,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.000290403271850718,"delta_db":0.45642156066092343,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0405125903256807,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.115605770913343,"pitch":null,"output_rms_dbfs":-25.948941000932596,"output_peak":0.7865849114536864,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.963178409387889},"common_prefix":{"samples":84000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5911975572390675,"spectral_convergence":0.606384016951396,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.339852717635891,"spectral_convergence":0.3793736639010918,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.958484953288709,"spectral_convergence":0.4213207207511064,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.868518008191147e-05,"delta_db":1.0167945213903267,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.0003779253338700017,"delta_db":1.620460482234824,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0008419447741081186,"delta_db":1.5213601441015576,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0006466033259956971,"delta_db":0.8774116471692377,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.0003660304949880647,"delta_db":0.7506271863972137,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.000290403271850718,"delta_db":0.45642156066092343,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0405125903256807,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.115605770913343,"pitch":null,"output_rms_dbfs":-25.948941000932596,"output_peak":0.7865849114536864,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.963178409387889},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84000,"output_samples":84000,"intelligibility":{"estoi":{"value":0.9193645066923738,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04433764904553492,"rms_dbfs":-27.064546771845936,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.51650626449395,"mean":-1.6414097377232142e-06,"zero_energy":false},"output":{"rms":0.05041420812523501,"rms_dbfs":-25.948941000932596,"peak":0.7865849114536864,"peak_dbfs":-2.085087772694099,"crest_factor_db":23.863853228238497,"mean":-7.514756975847216e-09,"zero_energy":false},"rms_difference_db":1.1156057709133407},"full_scale":{"reference":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":525,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":1.5888158540131854e-06,"output_mean_power":1.889066976931651e-06,"output_rms_dbfs":-57.23752643880087,"output_zero_energy":false,"output_peak":0.021741738185349142,"output_energy_fraction":9.62700207700887e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":518,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":221,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":38,"paired_voiced_fraction":0.9104477611940298,"absolute_difference_cents_median":3.5380308840089016,"absolute_difference_cents_p95":1200.6779500447826,"difference_over_600_cents_frames":19,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00019312594202620684,"log_rms_mae_db":1.5666432166802997,"rms_envelope_nrmse":0.18933161604393067,"envelope_correlation":0.9860227406044856,"rms_gain_error_db":1.1156075832077623,"reference_mean_power":0.0019676998400409485,"output_mean_power":0.002544014654214665},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.0378624939923077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018862042103601662,"output_power":0.00023326516176188543,"reference_active":true,"delta_db":0.9226117165042645},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.862312285918816e-05,"output_power":0.0001278499868538938,"reference_active":true,"delta_db":1.1272193754358293},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3789377405749525e-05,"output_power":1.5338131103230758e-05,"reference_active":true,"delta_db":0.4622778749622663}],"active_band_count":3,"active_band_mae_db":0.8373696556341201}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00018244538675777041,"log_rms_mae_db":1.189197671534683,"rms_envelope_nrmse":0.1608949054094776,"envelope_correlation":0.9937393406333174,"rms_gain_error_db":1.115616520865936,"reference_mean_power":0.001979010526094432,"output_mean_power":0.002558643365171826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":3.3102005537920435e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017836230285954771,"output_power":0.00021457425855241462,"reference_active":true,"delta_db":0.8027454957492398},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3320761072042273e-05,"output_power":5.766890192146111e-05,"reference_active":true,"delta_db":1.2424560421570738},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.132775238046075e-07,"output_power":1.029185639070103e-06,"reference_active":true,"delta_db":1.0225494751391067}],"active_band_count":3,"active_band_mae_db":1.0225836710151401}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014606965842361196,"log_rms_mae_db":1.9429098757724859,"rms_envelope_nrmse":0.2535863413087718,"envelope_correlation":0.98766545950116,"rms_gain_error_db":1.6712292336358314,"reference_mean_power":0.0007714288246159527,"output_mean_power":0.0011334928519333362},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.155891354220712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.355627855804231e-05,"output_power":0.00014939715355037826,"reference_active":true,"delta_db":2.523632339203477},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.117824128471733e-05,"output_power":3.657065354545056e-05,"reference_active":true,"delta_db":2.3724282964575703},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3214088476749303e-06,"output_power":4.599608336131303e-06,"reference_active":true,"delta_db":1.4139851418105034}],"active_band_count":3,"active_band_mae_db":2.103348592490517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001346093108760191,"log_rms_mae_db":1.5582535121809087,"rms_envelope_nrmse":0.2361766005299678,"envelope_correlation":0.9927128427151021,"rms_gain_error_db":1.6712300648797385,"reference_mean_power":0.0007758659430163304,"output_mean_power":0.001140012715017171},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.945223621465206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.346976557991944e-05,"output_power":0.00013420229001571854,"reference_active":true,"delta_db":2.6165127251530023},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57892144723787e-06,"output_power":1.5859382476847504e-05,"reference_active":true,"delta_db":2.1896966120714922},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.32297115664878e-07,"output_power":2.7344026319377625e-07,"reference_active":true,"delta_db":0.7081864601459621}],"active_band_count":3,"active_band_mae_db":1.8381319324568188}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00012444218003818797,"log_rms_mae_db":1.4966297109916304,"rms_envelope_nrmse":0.16485795493669644,"envelope_correlation":0.9894267806655065,"rms_gain_error_db":0.8743220722008238,"reference_mean_power":0.000687550351321365,"output_mean_power":0.0008408852131542945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.3390858183888006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.254326674690696e-05,"output_power":5.055804387029939e-05,"reference_active":true,"delta_db":0.7495942762746083},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4859458498211297e-05,"output_power":1.985711587760394e-05,"reference_active":true,"delta_db":1.2591318682517814},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141405389451218e-06,"output_power":3.9413125530154626e-06,"reference_active":true,"delta_db":0.9851689206179043}],"active_band_count":3,"active_band_mae_db":0.9979650217147648}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001136976512596423,"log_rms_mae_db":0.9961774129693012,"rms_envelope_nrmse":0.12375010576794923,"envelope_correlation":0.997145158671478,"rms_gain_error_db":0.8743332190631554,"reference_mean_power":0.0006915015836222035,"output_mean_power":0.0008457198049127684},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.214722623247202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7827908938064065e-05,"output_power":4.5073758260071485e-05,"reference_active":true,"delta_db":0.761114362499618},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.619804659862373e-06,"output_power":8.233446655278983e-06,"reference_active":true,"delta_db":1.6586045595546506},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.015023849518023e-07,"output_power":2.4775223895457717e-07,"reference_active":true,"delta_db":0.8973739717366157}],"active_band_count":3,"active_band_mae_db":1.105697631263628}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00010848831624356853,"log_rms_mae_db":1.4131155562939335,"rms_envelope_nrmse":0.16320323401087047,"envelope_correlation":0.9829564148139055,"rms_gain_error_db":0.48803431088413923,"reference_mean_power":0.0002381846371729614,"output_mean_power":0.0002665122509270624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.715890316390027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.877460217795426e-05,"output_power":9.387970321004277e-05,"reference_active":true,"delta_db":0.24282973164671998},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.954888518804952e-05,"output_power":4.163902059768749e-05,"reference_active":true,"delta_db":0.22366259778540315},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.296333400161522e-06,"output_power":7.805439804003336e-06,"reference_active":true,"delta_db":0.29292708256686195}],"active_band_count":3,"active_band_mae_db":0.2531398039996617}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":0.8691196904902725,"rms_envelope_nrmse":0.09439917274993279,"envelope_correlation":0.9955132638262509,"rms_gain_error_db":0.48803563496999486,"reference_mean_power":0.00023955325002336745,"output_mean_power":0.00026804371643463133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.5691901914560289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.470858674877873e-05,"output_power":9.959063458995745e-05,"reference_active":true,"delta_db":0.218291436238451},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8332956489415002e-05,"output_power":1.9386143247026935e-05,"reference_active":true,"delta_db":0.24258909771703693},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.840671065644935e-07,"output_power":4.947327132214501e-07,"reference_active":true,"delta_db":0.09465055808215453}],"active_band_count":3,"active_band_mae_db":0.18517703067921412}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"facodec","status":"measured","mos":{"status":"measured","value":4.133120536804199,"predicted_mos":4.133120536804199,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84000,"input_sample_rate":16000,"input_seconds":5.25,"input_float32_sha256":"e464c9d8192dfdeb20d43ccd8c3d189aa69be702420b5926b42e0a22a73b22d6","internal_samples":84000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1030248240003857},"started_at":"2026-10-13T04:47:30.719705+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.567000046838075e-05,"native_input_samples":84000,"native_input_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","output_sha256":"ff0d76d791e1f67f30aa2514c92b440e1a5a3bf86b3f5442e5ef9945907b7101","output_samples":84000,"length_delta_samples":0,"analysis_s":1.3022923139997147,"synthesis_s":9.226291660000243,"total_s":10.528583973999957,"rtf":2.005444566476182,"output_peak":0.580653727054596,"output_rms":0.04032675305129407,"finished_at":"2026-10-13T04:47:41.255383+00:00","evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"ff0d76d791e1f67f30aa2514c92b440e1a5a3bf86b3f5442e5ef9945907b7101","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.125792225157091,"spectral_convergence":0.33850843506972217,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.812376257476685,"spectral_convergence":0.36409694052976177,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.092359272680254,"spectral_convergence":0.38814370139823495,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.6243187687848326e-05,"delta_db":0.4085342689147253,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.0002581823840135742,"delta_db":-0.03433350503898885,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0005081162584743962,"delta_db":-0.671844997471408,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.00040940225053973534,"delta_db":-1.1074654377788342,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.0002419515026796602,"delta_db":-1.0472622679654682,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.00019235142826484167,"delta_db":-1.3326393358497726,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7670133021698661,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8235881129079975,"pitch":null,"output_rms_dbfs":-27.88813488475394,"output_peak":0.580653727054596,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.676842585104676},"common_prefix":{"samples":84000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.125792225157091,"spectral_convergence":0.33850843506972217,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.812376257476685,"spectral_convergence":0.36409694052976177,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.092359272680254,"spectral_convergence":0.38814370139823495,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.6243187687848326e-05,"delta_db":0.4085342689147253,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.0002581823840135742,"delta_db":-0.03433350503898885,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0005081162584743962,"delta_db":-0.671844997471408,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.00040940225053973534,"delta_db":-1.1074654377788342,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.0002419515026796602,"delta_db":-1.0472622679654682,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.00019235142826484167,"delta_db":-1.3326393358497726,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7670133021698661,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8235881129079975,"pitch":null,"output_rms_dbfs":-27.88813488475394,"output_peak":0.580653727054596,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.676842585104676},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84000,"output_samples":84000,"intelligibility":{"estoi":{"value":0.9371480258254727,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04433764904553492,"rms_dbfs":-27.064546771845936,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.51650626449395,"mean":-1.6414097377232142e-06,"zero_energy":false},"output":{"rms":0.04032675305129407,"rms_dbfs":-27.88813488475394,"peak":0.580653727054596,"peak_dbfs":-4.721655640021268,"crest_factor_db":23.16647924473267,"mean":7.067227610748108e-05,"zero_energy":false},"rms_difference_db":-0.823588112908002},"full_scale":{"reference":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":525,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":1.5888158540131854e-06,"output_mean_power":2.1648845575739305e-06,"output_rms_dbfs":-56.64565257437797,"output_zero_energy":false,"output_peak":0.008268672041594982,"output_energy_fraction":0.0001724240494007173,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":518,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":210,"both_estimated_voiced_frames":196,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9751243781094527,"absolute_difference_cents_median":2.9344924486304773,"absolute_difference_cents_p95":10.09428541059756,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00019312594202620684,"log_rms_mae_db":1.1102955802533043,"rms_envelope_nrmse":0.14087879521251453,"envelope_correlation":0.9885834101986504,"rms_gain_error_db":-0.8235886996770598,"reference_mean_power":0.0019676998400409485,"output_mean_power":0.001627796012739683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.0378624939923077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018862042103601662,"output_power":0.00014478147722858575,"reference_active":true,"delta_db":-1.148757064539116},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.862312285918816e-05,"output_power":8.061586640648986e-05,"reference_active":true,"delta_db":-0.8755822430490556},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3789377405749525e-05,"output_power":1.3990808282793654e-05,"reference_active":true,"delta_db":0.0629814735695634}],"active_band_count":3,"active_band_mae_db":0.695773593719245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00018244538675777041,"log_rms_mae_db":0.932664945676241,"rms_envelope_nrmse":0.13087776253238168,"envelope_correlation":0.9911044581474412,"rms_gain_error_db":-0.8236053340124888,"reference_mean_power":0.001979010526094432,"output_mean_power":0.0016371466009815813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":3.3102005537920435e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017836230285954771,"output_power":0.00013049541565223009,"reference_active":true,"delta_db":-1.3570781591660341},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3320761072042273e-05,"output_power":3.631803480990821e-05,"reference_active":true,"delta_db":-0.7657373727864748},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.132775238046075e-07,"output_power":7.445594090512013e-07,"reference_active":true,"delta_db":-0.38339413849180515}],"active_band_count":3,"active_band_mae_db":0.835403223481438}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014606965842361196,"log_rms_mae_db":1.2956583171686202,"rms_envelope_nrmse":0.13490756217522573,"envelope_correlation":0.9870760495845649,"rms_gain_error_db":-0.42957611875436125,"reference_mean_power":0.0007714288246159527,"output_mean_power":0.0006987764345608757},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.155891354220712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.355627855804231e-05,"output_power":9.601112588628101e-05,"reference_active":true,"delta_db":0.6034247346280927},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.117824128471733e-05,"output_power":2.2662373689218926e-05,"reference_active":true,"delta_db":0.2941550459587764},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3214088476749303e-06,"output_power":3.1365442474203994e-06,"reference_active":true,"delta_db":-0.24870919682954945}],"active_band_count":3,"active_band_mae_db":0.3820963258054728}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001346093108760191,"log_rms_mae_db":0.9517932168886086,"rms_envelope_nrmse":0.12515539783820406,"envelope_correlation":0.9885120823608958,"rms_gain_error_db":-0.42957632732064444,"reference_mean_power":0.0007758659430163304,"output_mean_power":0.0007027956358662678},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.945223621465206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.346976557991944e-05,"output_power":8.374507183513537e-05,"reference_active":true,"delta_db":0.5685060529428732},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57892144723787e-06,"output_power":1.0597224350432702e-05,"reference_active":true,"delta_db":0.43875516936969716},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.32297115664878e-07,"output_power":2.366875463772028e-07,"reference_active":true,"delta_db":0.08131590179100269}],"active_band_count":3,"active_band_mae_db":0.3628590413678577}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00012444218003818797,"log_rms_mae_db":1.5320272968387718,"rms_envelope_nrmse":0.16216154784010423,"envelope_correlation":0.9902409674203375,"rms_gain_error_db":-1.1111658386281553,"reference_mean_power":0.000687550351321365,"output_mean_power":0.0005323385587486817},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.3390858183888006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.254326674690696e-05,"output_power":3.723296514033468e-05,"reference_active":true,"delta_db":-0.5790321106025218},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4859458498211297e-05,"output_power":1.2367613223426441e-05,"reference_active":true,"delta_db":-0.797170883986248},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141405389451218e-06,"output_power":3.157978350712762e-06,"reference_active":true,"delta_db":0.022851639232418054}],"active_band_count":3,"active_band_mae_db":0.4663515446070626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001136976512596423,"log_rms_mae_db":1.2751928560375039,"rms_envelope_nrmse":0.1490974315057058,"envelope_correlation":0.9935266120458117,"rms_gain_error_db":-1.1112057962349484,"reference_mean_power":0.0006915015836222035,"output_mean_power":0.0005353928899224305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.214722623247202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7827908938064065e-05,"output_power":3.212834070778027e-05,"reference_active":true,"delta_db":-0.7092403823516519},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.619804659862373e-06,"output_power":4.958435119777962e-06,"reference_active":true,"delta_db":-0.5437658514311517},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.015023849518023e-07,"output_power":1.7113907918260943e-07,"reference_active":true,"delta_db":-0.7093099982243067}],"active_band_count":3,"active_band_mae_db":0.6541054106690368}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00010848831624356853,"log_rms_mae_db":2.0249493384617425,"rms_envelope_nrmse":0.20632716252380642,"envelope_correlation":0.9700902267703818,"rms_gain_error_db":-0.7419103970088925,"reference_mean_power":0.0002381846371729614,"output_mean_power":0.00020078104326228728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.715890316390027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.877460217795426e-05,"output_power":7.106719944337988e-05,"reference_active":true,"delta_db":-0.9661953332490323},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.954888518804952e-05,"output_power":3.542604011469171e-05,"reference_active":true,"delta_db":-0.4781163595768513},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.296333400161522e-06,"output_power":7.706217436758213e-06,"reference_active":true,"delta_db":0.2373658828211489}],"active_band_count":3,"active_band_mae_db":0.5605591918823442}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":1.8339613239512742,"rms_envelope_nrmse":0.17746563377548996,"envelope_correlation":0.9770651766296601,"rms_gain_error_db":-0.7419396640313546,"reference_mean_power":0.00023955325002336745,"output_mean_power":0.00020193337361593672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.5691901914560289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.470858674877873e-05,"output_power":7.85679089231669e-05,"reference_active":true,"delta_db":-0.8114416147789734},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8332956489415002e-05,"output_power":1.644220375250331e-05,"reference_active":true,"delta_db":-0.47272482007852495},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.840671065644935e-07,"output_power":4.80663119431255e-07,"reference_active":true,"delta_db":-0.030647717333857485}],"active_band_count":3,"active_band_mae_db":0.4382713840637853}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.985501766204834,"predicted_mos":2.985501766204834,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":231936,"input_sample_rate":44100,"input_seconds":5.259319727891157,"input_float32_sha256":"5da955e48b39988b9645cdfb5b7038fd1d8168aefe05b920fefcf11d153eb273","internal_samples":84150,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.904061831002764},"started_at":"2026-10-13T07:04:23.441696+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00680948300032469,"native_input_samples":231525,"native_input_sha256":"233cc32f2ed713b2b310d3a3e94e5cd04d3fe14f317e0e0747a7d3f3b2735e1f","output_sha256":"f28f6f4d5c4b245a8904a50fd4553963c9b988a5fb2ab9b22babc77134891be1","output_samples":231936,"length_delta_samples":411,"analysis_s":0.13076574099977734,"synthesis_s":31.91664388299978,"total_s":32.047409623999556,"rtf":6.104268499809439,"output_peak":0.573847770690918,"output_rms":0.0406820897101714,"finished_at":"2026-10-13T07:04:55.511316+00:00","evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"f28f6f4d5c4b245a8904a50fd4553963c9b988a5fb2ab9b22babc77134891be1","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":84000,"output_samples":84150,"length_delta_samples":150,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.82317873438886,"output_peak":0.5471825453693305},"common_prefix":{"samples":84000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982174688057041,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.099567491819327,"spectral_convergence":0.5512568775870261,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.0431371967597,"spectral_convergence":0.4924159326605717,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.074423774847585,"spectral_convergence":0.5154064204961776,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.1912311064828813e-05,"delta_db":-0.9382181008616491,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.00017481183329169894,"delta_db":-1.7278912981295926,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0005427569505976016,"delta_db":-0.38542197459432115,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0004636603665701953,"delta_db":-0.5669679525463498,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.00025611080168213754,"delta_db":-0.8002665492796632,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.00020444816252313626,"delta_db":-1.0677613731323485,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9144212080906543,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7508848460989247,"pitch":null,"output_rms_dbfs":-27.815431617944864,"output_peak":0.5471825453693305,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.072376154475537},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84000,"output_samples":84000,"intelligibility":{"estoi":{"value":0.7626897009416772,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04433764904553492,"rms_dbfs":-27.064546771845936,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.51650626449395,"mean":-1.6414097377232142e-06,"zero_energy":false},"output":{"rms":0.04066571560577286,"rms_dbfs":-27.815431617944864,"peak":0.5471825453693305,"peak_dbfs":-5.237355293400868,"crest_factor_db":22.578076324543996,"mean":-2.1619242136502757e-05,"zero_energy":false},"rms_difference_db":-0.750884846098927},"full_scale":{"reference":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":525,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":1.5888158540131854e-06,"output_mean_power":9.677122375804625e-07,"output_rms_dbfs":-60.14253766875871,"output_zero_energy":false,"output_peak":0.0042466905692222275,"output_energy_fraction":7.579472895094141e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":518,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":187,"both_estimated_voiced_frames":171,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.8507462686567164,"absolute_difference_cents_median":6.103365500285186,"absolute_difference_cents_p95":1198.407022089595,"difference_over_600_cents_frames":10,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00019312594202620684,"log_rms_mae_db":1.7630931618628187,"rms_envelope_nrmse":0.14106013925119726,"envelope_correlation":0.9876750255029554,"rms_gain_error_db":-0.7508839470688113,"reference_mean_power":0.0019676998400409485,"output_mean_power":0.0016552761424991557},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.0378624939923077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018862042103601662,"output_power":0.00014537138963500586,"reference_active":true,"delta_db":-1.1310976797684273},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.862312285918816e-05,"output_power":6.740642370940878e-05,"reference_active":true,"delta_db":-1.6527746418203342},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3789377405749525e-05,"output_power":9.816089281886221e-06,"reference_active":true,"delta_db":-1.4760615825433727}],"active_band_count":3,"active_band_mae_db":1.419977968044045}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00018244538675777041,"log_rms_mae_db":1.4996453010552242,"rms_envelope_nrmse":0.11285770558816262,"envelope_correlation":0.9946053108028978,"rms_gain_error_db":-0.7508773673890824,"reference_mean_power":0.001979010526094432,"output_mean_power":0.0016647934841902288},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":3.3102005537920435e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017836230285954771,"output_power":0.00013534347951910507,"reference_active":true,"delta_db":-1.1986573346484457},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3320761072042273e-05,"output_power":2.960561155873729e-05,"reference_active":true,"delta_db":-1.6532204116767528},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.132775238046075e-07,"output_power":5.496993370652781e-07,"reference_active":true,"delta_db":-1.7011355685024019}],"active_band_count":3,"active_band_mae_db":1.5176711049425335}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014606965842361196,"log_rms_mae_db":2.760153231395615,"rms_envelope_nrmse":0.16244858418121064,"envelope_correlation":0.9831983716013788,"rms_gain_error_db":-0.6969153038377105,"reference_mean_power":0.0007714288246159527,"output_mean_power":0.0006570589449054158},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.155891354220712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.355627855804231e-05,"output_power":6.217443674284459e-05,"reference_active":true,"delta_db":-1.2836722951296022},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.117824128471733e-05,"output_power":1.3429039668579068e-05,"reference_active":true,"delta_db":-1.9784493522677233},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3214088476749303e-06,"output_power":3.1481610929611847e-06,"reference_active":true,"delta_db":-0.23265390898979177}],"active_band_count":3,"active_band_mae_db":1.1649251854623726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001346093108760191,"log_rms_mae_db":2.328204487975055,"rms_envelope_nrmse":0.13191253557915827,"envelope_correlation":0.9902745316228223,"rms_gain_error_db":-0.6969245336356706,"reference_mean_power":0.0007758659430163304,"output_mean_power":0.0006608368241051556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.945223621465206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.346976557991944e-05,"output_power":5.446285011719213e-05,"reference_active":true,"delta_db":-1.3000828930572452},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57892144723787e-06,"output_power":6.060136703260936e-06,"reference_active":true,"delta_db":-1.9883419080643723},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.32297115664878e-07,"output_power":1.8683985044697138e-07,"reference_active":true,"delta_db":-0.9457430639097775}],"active_band_count":3,"active_band_mae_db":1.4113892883437984}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00012444218003818797,"log_rms_mae_db":2.9681272183662264,"rms_envelope_nrmse":0.22346515551168547,"envelope_correlation":0.9658388402515725,"rms_gain_error_db":-0.6096192207204796,"reference_mean_power":0.000687550351321365,"output_mean_power":0.0005975064341151202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.3390858183888006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.254326674690696e-05,"output_power":4.303034093699692e-05,"reference_active":true,"delta_db":0.04943952304578006},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4859458498211297e-05,"output_power":1.4197046667890596e-05,"reference_active":true,"delta_db":-0.1980497341648228},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141405389451218e-06,"output_power":2.5250495197775353e-06,"reference_active":true,"delta_db":-0.9485408482083055}],"active_band_count":3,"active_band_mae_db":0.3986767018063028}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001136976512596423,"log_rms_mae_db":2.4934023684927134,"rms_envelope_nrmse":0.1810780772548217,"envelope_correlation":0.9779681792060108,"rms_gain_error_db":-0.6096059220507499,"reference_mean_power":0.0006915015836222035,"output_mean_power":0.0006009420398437839},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.214722623247202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7827908938064065e-05,"output_power":3.8964234083045965e-05,"reference_active":true,"delta_db":0.12853808863804622},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.619804659862373e-06,"output_power":6.189272873691294e-06,"reference_active":true,"delta_db":0.4191841027683399},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.015023849518023e-07,"output_power":1.4828430728951836e-07,"reference_active":true,"delta_db":-1.3318499803851835}],"active_band_count":3,"active_band_mae_db":0.6265240572638565}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00010848831624356853,"log_rms_mae_db":3.732365478269658,"rms_envelope_nrmse":0.26372148749727126,"envelope_correlation":0.9495498986399529,"rms_gain_error_db":-0.5330814822873007,"reference_mean_power":0.0002381846371729614,"output_mean_power":0.00021067140807258536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.715890316390027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.877460217795426e-05,"output_power":6.741739326534544e-05,"reference_active":true,"delta_db":-1.1951677854326161},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.954888518804952e-05,"output_power":2.9777304926989383e-05,"reference_active":true,"delta_db":-1.2324885779106958},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.296333400161522e-06,"output_power":4.2334440755804944e-06,"reference_active":true,"delta_db":-2.364108434637725}],"active_band_count":3,"active_band_mae_db":1.5972549326603456}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":3.4960801236202834,"rms_envelope_nrmse":0.22366742045876487,"envelope_correlation":0.962554128608632,"rms_gain_error_db":-0.5330613320390957,"reference_mean_power":0.00023955325002336745,"output_mean_power":0.00021188291254004804},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.5691901914560289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.470858674877873e-05,"output_power":7.1953033933975e-05,"reference_active":true,"delta_db":-1.1934024518097945},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8332956489415002e-05,"output_power":1.3277078387584193e-05,"reference_active":true,"delta_db":-1.4012998837802826},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.840671065644935e-07,"output_power":3.488864442109508e-07,"reference_active":true,"delta_db":-1.4222147686346163}],"active_band_count":3,"active_band_mae_db":1.338972368074898}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"golf","status":"measured","mos":{"status":"measured","value":3.198162794113159,"predicted_mos":3.198162794113159,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":125760,"input_sample_rate":24000,"input_seconds":5.24,"input_float32_sha256":"cb174f2048b0275992e1d160727186697adc7d9a408007bbcc2d3f1aa145dd9d","internal_samples":83840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0407358159991418},"started_at":"2026-10-13T05:02:03.382076+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0025231089994122158,"native_input_samples":126000,"native_input_sha256":"5783c3f64b7cd9b74b60e40bc7ba400acdfc6da9e0b66a1abd912d61aa38ca1d","output_sha256":"0979afe9bb48bb5005e2f85b4cf985ac6cf5f76c3fd0965bcc2b2391d67e626e","output_samples":125760,"length_delta_samples":-240,"analysis_s":0.5458716939992883,"synthesis_s":0.1255435690000013,"total_s":0.6714152629992896,"rtf":0.1278886215236742,"output_peak":0.6490410566329956,"output_rms":0.04220384464609749,"finished_at":"2026-10-13T05:02:04.064315+00:00","analysis_stages":{"external_f0_s":0.04257950400005939},"evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"0979afe9bb48bb5005e2f85b4cf985ac6cf5f76c3fd0965bcc2b2391d67e626e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":84000,"output_samples":83840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.505299068472166,"output_peak":0.6498316834820552},"common_prefix":{"samples":83840,"sample_rate":16000,"reference_coverage":0.9980952380952381,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83840,"output_samples":83840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.535429124691487,"spectral_convergence":0.5623799468630579,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.540162396161551,"spectral_convergence":0.35951592556293316,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.694478799472261,"spectral_convergence":0.3841164605939793,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.5179476107238957e-05,"output_power":2.269479000772177e-05,"delta_db":1.746693857081742,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002605154082110824,"output_power":0.0001961794339095684,"delta_db":-1.2317993777718788,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005941191861890084,"output_power":0.0005996648279782495,"delta_db":0.04034999504217672,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005295265080890353,"output_power":0.0004938360790696161,"delta_db":-0.30304889975732807,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00030831624010802623,"output_power":0.000238525740666183,"delta_db":-1.1146114806997875,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002619182247522661,"output_power":0.00022521008400940175,"delta_db":-0.6557788589664076,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8487137448865535,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4490243951115946,"pitch":null,"output_rms_dbfs":-27.505299068472166,"output_peak":0.6498316834820552,"reference_rms_dbfs":-27.056274673360573,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.923356773441767},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83840,"output_samples":83840,"intelligibility":{"estoi":{"value":0.898377956863976,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044379894585911954,"rms_dbfs":-27.056274673360573,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.508234166008585,"mean":-1.6598301079437023e-06,"zero_energy":false},"output":{"rms":0.042143931421270375,"rms_dbfs":-27.505299068472166,"peak":0.6498316834820552,"peak_dbfs":-3.743982356405047,"crest_factor_db":23.76131671206712,"mean":0.0002259642269591155,"zero_energy":false},"rms_difference_db":-0.44902439511159287},"full_scale":{"reference":{"sample_count":83840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":524,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":67,"quiet_samples":10720,"reference_mean_power":1.5839214137852638e-06,"output_mean_power":1.4908811346564928e-06,"output_rms_dbfs":-58.26556980705696,"output_zero_energy":false,"output_peak":0.005034878624012956,"output_energy_fraction":0.0001073288415536495,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":517,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":215,"both_estimated_voiced_frames":190,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.945273631840796,"absolute_difference_cents_median":3.055326586015122,"absolute_difference_cents_p95":13.17015461733018,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1047,"rms_floor":0.00019312594202620684,"log_rms_mae_db":1.6734343946331407,"rms_envelope_nrmse":0.14922122090366446,"envelope_correlation":0.9829827707684408,"rms_gain_error_db":-0.44902586250410437,"reference_mean_power":0.001971454620858345,"output_mean_power":0.0017778053084065587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1047,"frequency_resolution_hz":0.19102196752626552,"power_floor":4.0494219123623997e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001891297518376484,"output_power":0.0001646128704856346,"reference_active":true,"delta_db":-0.6029606458811263},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.877882270622028e-05,"output_power":0.00010326620616893369,"reference_active":true,"delta_db":0.19294376322227907},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3746748998428123e-05,"output_power":1.5126794917908989e-05,"reference_active":true,"delta_db":0.41546916005977824}],"active_band_count":3,"active_band_mae_db":0.40379118972106115}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1041,"rms_floor":0.00018244538675777041,"log_rms_mae_db":1.2884193259744312,"rms_envelope_nrmse":0.11703243629938759,"envelope_correlation":0.9900564769377329,"rms_gain_error_db":-0.44902617076870177,"reference_mean_power":0.0019828092723699715,"output_mean_power":0.0017880445040284198},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1041,"frequency_resolution_hz":0.19212295869356388,"power_floor":3.320129933562179e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001789016561884765,"output_power":0.0001642083090329151,"reference_active":true,"delta_db":-0.3721923219787549},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.325137422111182e-05,"output_power":4.536997919036141e-05,"reference_active":true,"delta_db":0.20768669259806108},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.130793241033871e-07,"output_power":1.0406543491636937e-06,"reference_active":true,"delta_db":1.0717358611911159}],"active_band_count":3,"active_band_mae_db":0.5505382919226439}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1047,"rms_floor":0.00014606965842361196,"log_rms_mae_db":1.9550781035376248,"rms_envelope_nrmse":0.1641022450254657,"envelope_correlation":0.9814481147439857,"rms_gain_error_db":-0.3442291529739282,"reference_mean_power":0.0007729021770406794,"output_mean_power":0.0007140055955437093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1047,"frequency_resolution_hz":0.19102196752626552,"power_floor":2.161681458627025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.361921447239105e-05,"output_power":5.97335901328603e-05,"reference_active":true,"delta_db":-1.4608746574846725},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1194601648150128e-05,"output_power":2.0382690884766886e-05,"reference_active":true,"delta_db":-0.16963740228854918},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2942986005277316e-06,"output_power":2.5083528335137897e-06,"reference_active":true,"delta_db":-1.1837433596163869}],"active_band_count":3,"active_band_mae_db":0.9380851397965362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1041,"rms_floor":0.0001346093108760191,"log_rms_mae_db":1.5581354587756548,"rms_envelope_nrmse":0.13980465474399376,"envelope_correlation":0.9865940632599376,"rms_gain_error_db":-0.3442277963995019,"reference_mean_power":0.0007773563533707997,"output_mean_power":0.0007181205796783075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1041,"frequency_resolution_hz":0.19212295869356388,"power_floor":1.9509073250180956e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.357131969292107e-05,"output_power":5.3578358420106726e-05,"reference_active":true,"delta_db":-1.3771914312553877},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.5177300819668e-06,"output_power":9.396968670601984e-06,"reference_active":true,"delta_db":-0.05545605340152437},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3248886654323842e-07,"output_power":1.8860576005948346e-07,"reference_active":true,"delta_db":-0.9084720812544879}],"active_band_count":3,"active_band_mae_db":0.7803731886371333}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1047,"rms_floor":0.00012444218003818797,"log_rms_mae_db":2.168549911591472,"rms_envelope_nrmse":0.19367533978288892,"envelope_correlation":0.976285978262325,"rms_gain_error_db":-0.6290469704204732,"reference_mean_power":0.0006888617565889911,"output_mean_power":0.0005959740880156313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1047,"frequency_resolution_hz":0.19102196752626552,"power_floor":1.3444931563631107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.268388290672756e-05,"output_power":3.0781394539644474e-05,"reference_active":true,"delta_db":-1.4197562835813677},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.488530875174395e-05,"output_power":1.2798294236726125e-05,"reference_active":true,"delta_db":-0.6560575672011371},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1413455206978135e-06,"output_power":2.5991208180640807e-06,"reference_active":true,"delta_db":-0.8228923983938905}],"active_band_count":3,"active_band_mae_db":0.966235416392132}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1041,"rms_floor":0.0001136976512596423,"log_rms_mae_db":1.8264470665382877,"rms_envelope_nrmse":0.1459322655973865,"envelope_correlation":0.9879016009255217,"rms_gain_error_db":-0.6290462690544075,"reference_mean_power":0.0006928284791196402,"output_mean_power":0.0005994060255600863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1041,"frequency_resolution_hz":0.19212295869356388,"power_floor":1.2199050120783687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.796076178207305e-05,"output_power":2.830986900458371e-05,"reference_active":true,"delta_db":-1.2739705929606913},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.610178246946617e-06,"output_power":4.706880526511027e-06,"reference_active":true,"delta_db":-0.7624348504190325},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0022531153358464e-07,"output_power":2.047732374883273e-07,"reference_active":true,"delta_db":0.09754218526914357}],"active_band_count":3,"active_band_mae_db":0.7113158762162891}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1047,"rms_floor":0.00010848831624356853,"log_rms_mae_db":2.5285902990717797,"rms_envelope_nrmse":0.24054923891289123,"envelope_correlation":0.9596248658162173,"rms_gain_error_db":0.014765288162220875,"reference_mean_power":0.0002386387637288976,"output_mean_power":0.00023945147649165663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1047,"frequency_resolution_hz":0.19102196752626552,"power_floor":1.7202542212384264e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.909245434818499e-05,"output_power":0.00010649730474634198,"reference_active":true,"delta_db":0.7749769352096676},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9584889013642676e-05,"output_power":4.890367017464058e-05,"reference_active":true,"delta_db":0.9181202211899613},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.286394038898521e-06,"output_power":9.595198200797607e-06,"reference_active":true,"delta_db":1.1954129642607854}],"active_band_count":3,"active_band_mae_db":0.9628367068868048}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83840,"uncovered_tail_samples":0,"metrics":{"frame_count":1041,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":2.2297841675127135,"rms_envelope_nrmse":0.18871601754961245,"envelope_correlation":0.9731531824628117,"rms_gain_error_db":0.014779654839915007,"reference_mean_power":0.00024001269523029378,"output_mean_power":0.00024083088375983111},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1041,"frequency_resolution_hz":0.19212295869356388,"power_floor":1.572983839006426e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.502173464854692e-05,"output_power":0.00011651995173671574,"reference_active":true,"delta_db":0.8857734174448633},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.82997233705859e-05,"output_power":2.23332376324994e-05,"reference_active":true,"delta_db":0.86507162248239},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.852728256057268e-07,"output_power":5.977689202923888e-07,"reference_active":true,"delta_db":0.9054735878768543}],"active_band_count":3,"active_band_mae_db":0.8854395426013691}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3856303691864014,"predicted_mos":3.3856303691864014,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":126120,"input_sample_rate":24000,"input_seconds":5.255,"input_float32_sha256":"08c70fd73837bbf611bc581ad243876060787be0411a532217d579a97244af9a","internal_samples":84080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.047302999000749},"started_at":"2026-10-13T05:06:25.254021+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029849939992345753,"native_input_samples":126000,"native_input_sha256":"5783c3f64b7cd9b74b60e40bc7ba400acdfc6da9e0b66a1abd912d61aa38ca1d","output_sha256":"74db780f173cdc4fdd97ea592ec00cecb0d8a508e3d9fb58458163e941ca7452","output_samples":126120,"length_delta_samples":120,"analysis_s":2.248443552000026,"synthesis_s":12.574710748999678,"total_s":14.823154300999704,"rtf":2.8234579620951816,"output_peak":0.7001799941062927,"output_rms":0.04350129302227185,"finished_at":"2026-10-13T05:06:40.089774+00:00","analysis_stages":{"external_f0_s":1.4968671510005152},"evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"74db780f173cdc4fdd97ea592ec00cecb0d8a508e3d9fb58458163e941ca7452","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":84000,"output_samples":84080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.24303953600675,"output_peak":0.630065443148505},"common_prefix":{"samples":84000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990485252140818,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.097116145718483,"spectral_convergence":0.5332450704058092,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.387495385990251,"spectral_convergence":0.39246445403028857,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.47484270575121,"spectral_convergence":0.40049886653421224,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.5798228795432653e-05,"delta_db":0.28790563477734343,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.0002572071520114493,"delta_db":-0.05076916595130037,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0006077344030653966,"delta_db":0.10566230658704728,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0005265351436032001,"delta_db":-0.014694060001241738,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.00028223215877436436,"delta_db":-0.3784805284440665,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.0001989597146508136,"delta_db":-1.1859419960695097,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.33724228197175155,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.17435968447934957,"pitch":null,"output_rms_dbfs":-27.23890645632529,"output_peak":0.630065443148505,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.653151412486648},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84000,"output_samples":84000,"intelligibility":{"estoi":{"value":0.8994825876852357,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04433764904553492,"rms_dbfs":-27.064546771845936,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.51650626449395,"mean":-1.6414097377232142e-06,"zero_energy":false},"output":{"rms":0.04345649319607666,"rms_dbfs":-27.238906456325292,"peak":0.630065443148505,"peak_dbfs":-4.0122867848272,"crest_factor_db":23.22661967149809,"mean":-1.7017243031457538e-05,"zero_energy":false},"rms_difference_db":-0.1743596844793558},"full_scale":{"reference":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":525,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":1.5888158540131854e-06,"output_mean_power":9.79654244283405e-06,"output_rms_dbfs":-50.089271756290586,"output_zero_energy":false,"output_peak":0.08329816275378964,"output_energy_fraction":0.0006719130549461552,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":518,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":212,"both_estimated_voiced_frames":190,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.945273631840796,"absolute_difference_cents_median":3.6366783062747707,"absolute_difference_cents_p95":15.413720543992703,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00019312594202620684,"log_rms_mae_db":2.0606404111562426,"rms_envelope_nrmse":0.17477644665359215,"envelope_correlation":0.9710583313997297,"rms_gain_error_db":-0.17435833221247188,"reference_mean_power":0.0019676998400409485,"output_mean_power":0.0018902664104237553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.0378624939923077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018862042103601662,"output_power":0.00016195486753188047,"reference_active":true,"delta_db":-0.6619470474014908},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.862312285918816e-05,"output_power":8.313957393051952e-05,"reference_active":true,"delta_db":-0.7417095563715763},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3789377405749525e-05,"output_power":1.1455857463447844e-05,"reference_active":true,"delta_db":-0.8051705670074938}],"active_band_count":3,"active_band_mae_db":0.7362757235935202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00018244538675777041,"log_rms_mae_db":1.39959577374894,"rms_envelope_nrmse":0.10800474110394204,"envelope_correlation":0.988274269288252,"rms_gain_error_db":-0.1743492923074669,"reference_mean_power":0.001979010526094432,"output_mean_power":0.001901135952692871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":3.3102005537920435e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017836230285954771,"output_power":0.00014824345485260222,"reference_active":true,"delta_db":-0.8032754317230224},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3320761072042273e-05,"output_power":3.653792131502156e-05,"reference_active":true,"delta_db":-0.7395224163573102},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.132775238046075e-07,"output_power":7.470245475655458e-07,"reference_active":true,"delta_db":-0.3690389689566848}],"active_band_count":3,"active_band_mae_db":0.6372789390123391}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014606965842361196,"log_rms_mae_db":2.1270349539720748,"rms_envelope_nrmse":0.1702520406112028,"envelope_correlation":0.9774840068827051,"rms_gain_error_db":0.08887746075054347,"reference_mean_power":0.0007714288246159527,"output_mean_power":0.00078737860255083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.155891354220712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.355627855804231e-05,"output_power":8.80809042178469e-05,"reference_active":true,"delta_db":0.22902675284793955},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.117824128471733e-05,"output_power":2.274061041535595e-05,"reference_active":true,"delta_db":0.3091222613273691},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3214088476749303e-06,"output_power":3.2013809836253684e-06,"reference_active":true,"delta_db":-0.159849773385039}],"active_band_count":3,"active_band_mae_db":0.2326662625201159}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001346093108760191,"log_rms_mae_db":1.422489135985144,"rms_envelope_nrmse":0.10828497652876769,"envelope_correlation":0.990320368415048,"rms_gain_error_db":0.08887863234848589,"reference_mean_power":0.0007758659430163304,"output_mean_power":0.0007919076748084082},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.945223621465206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.346976557991944e-05,"output_power":7.82101945281153e-05,"reference_active":true,"delta_db":0.2715471194462842},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57892144723787e-06,"output_power":1.0336862953843458e-05,"reference_active":true,"delta_db":0.3307214662952837},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.32297115664878e-07,"output_power":2.1730990852632419e-07,"reference_active":true,"delta_db":-0.2896428839209014}],"active_band_count":3,"active_band_mae_db":0.2973038232208231}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00012444218003818797,"log_rms_mae_db":1.932378586557662,"rms_envelope_nrmse":0.18696559759137302,"envelope_correlation":0.9741823771796649,"rms_gain_error_db":-0.07816110418311456,"reference_mean_power":0.000687550351321365,"output_mean_power":0.0006752870135854568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.3390858183888006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.254326674690696e-05,"output_power":4.2657745760602837e-05,"reference_active":true,"delta_db":0.011670667377756499},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4859458498211297e-05,"output_power":1.3836437149453657e-05,"reference_active":true,"delta_db":-0.3097870865408148},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141405389451218e-06,"output_power":2.950296408818757e-06,"reference_active":true,"delta_db":-0.272583338471309}],"active_band_count":3,"active_band_mae_db":0.19801369746329345}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001136976512596423,"log_rms_mae_db":1.116526132409555,"rms_envelope_nrmse":0.10013749405811469,"envelope_correlation":0.9922400838607209,"rms_gain_error_db":-0.07815191093529557,"reference_mean_power":0.0006915015836222035,"output_mean_power":0.0006791692082968234},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.214722623247202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7827908938064065e-05,"output_power":3.767793436100207e-05,"reference_active":true,"delta_db":-0.017252497039585448},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.619804659862373e-06,"output_power":5.486143006041979e-06,"reference_active":true,"delta_db":-0.10454096013842258},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.015023849518023e-07,"output_power":1.8570601848976477e-07,"reference_active":true,"delta_db":-0.35454211866073615}],"active_band_count":3,"active_band_mae_db":0.1587785252795814}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00010848831624356853,"log_rms_mae_db":2.25850906254594,"rms_envelope_nrmse":0.2517409102224373,"envelope_correlation":0.9513223864265096,"rms_gain_error_db":-0.6598087602188554,"reference_mean_power":0.0002381846371729614,"output_mean_power":0.00020461283378967225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.715890316390027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.877460217795426e-05,"output_power":7.036597751951284e-05,"reference_active":true,"delta_db":-1.0092600955665205},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.954888518804952e-05,"output_power":3.103826165745031e-05,"reference_active":true,"delta_db":-1.0523685604653679},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.296333400161522e-06,"output_power":5.9119912338318e-06,"reference_active":true,"delta_db":-0.9137088909017939}],"active_band_count":3,"active_band_mae_db":0.9917791823112273}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":1.482567405437955,"rms_envelope_nrmse":0.1559562025519257,"envelope_correlation":0.9824831813093431,"rms_gain_error_db":-0.6598005518324996,"reference_mean_power":0.00023955325002336745,"output_mean_power":0.00020578893145575304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.5691901914560289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.470858674877873e-05,"output_power":7.487177146312467e-05,"reference_active":true,"delta_db":-1.0207124743383131},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8332956489415002e-05,"output_power":1.4437155913902539e-05,"reference_active":true,"delta_db":-1.0375086104849645},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.840671065644935e-07,"output_power":4.055566157595756e-07,"reference_active":true,"delta_db":-0.7685408201942269}],"active_band_count":3,"active_band_mae_db":0.9422539683391683}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"llsm2","status":"measured","mos":{"status":"measured","value":4.120872974395752,"predicted_mos":4.120872974395752,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":126240,"input_sample_rate":24000,"input_seconds":5.26,"input_float32_sha256":"e7078e93aa089c89140d8db0d8a6872ed31be95b0acc0383f23fee9b9e81be89","internal_samples":84160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0725718249996135},"started_at":"2026-10-13T05:10:11.054831+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003411556001083227,"native_input_samples":126000,"native_input_sha256":"5783c3f64b7cd9b74b60e40bc7ba400acdfc6da9e0b66a1abd912d61aa38ca1d","output_sha256":"6b7ed77809707ad715ff90a664b19eee50053757930a8f9a4234c1cd985090ff","output_samples":126240,"length_delta_samples":240,"analysis_s":0.5439779289990838,"synthesis_s":0.12224478000098316,"total_s":0.666222709000067,"rtf":0.12689956361906038,"output_peak":0.6249218583106995,"output_rms":0.04317219639576583,"finished_at":"2026-10-13T05:10:11.735028+00:00","analysis_stages":{"external_f0_s":0.08445013900018239},"evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"6b7ed77809707ad715ff90a664b19eee50053757930a8f9a4234c1cd985090ff","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":84000,"output_samples":84160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.312399700770417,"output_peak":0.62517169957402},"common_prefix":{"samples":84000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980988593155894,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.449596456875069,"spectral_convergence":0.24583754007571698,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.324881885003976,"spectral_convergence":0.23586447322353354,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.904562767162657,"spectral_convergence":0.27093975816651195,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.4707640498689092e-05,"delta_db":-0.022748298668722078,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.00025587883069496054,"delta_db":-0.07325599764089487,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0005744909841989509,"delta_db":-0.13864373781237868,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0005011063160379133,"delta_db":-0.22966892780632017,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.00029312074617920877,"delta_db":-0.21407992961429967,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.00022101025429057823,"delta_db":-0.7294692473503349,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2346443564821584,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23958888203226553,"pitch":null,"output_rms_dbfs":-27.304135653878205,"output_peak":0.62517169957402,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.559680369680568},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84000,"output_samples":84000,"intelligibility":{"estoi":{"value":0.9641528370667412,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04433764904553492,"rms_dbfs":-27.064546771845936,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.51650626449395,"mean":-1.6414097377232142e-06,"zero_energy":false},"output":{"rms":0.04313136645065052,"rms_dbfs":-27.304135653878205,"peak":0.62517169957402,"peak_dbfs":-4.080013799142238,"crest_factor_db":23.22412185473597,"mean":-3.067079280502379e-05,"zero_energy":false},"rms_difference_db":-0.2395888820322689},"full_scale":{"reference":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":525,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":1.5888158540131854e-06,"output_mean_power":3.352266727409028e-06,"output_rms_dbfs":-54.74661433468556,"output_zero_energy":false,"output_peak":0.04955341860163949,"output_energy_fraction":0.0002334004780440463,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":518,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":205,"both_estimated_voiced_frames":191,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9502487562189055,"absolute_difference_cents_median":0.6816837712463644,"absolute_difference_cents_p95":4.147120355897571,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00019312594202620684,"log_rms_mae_db":0.7106603026102792,"rms_envelope_nrmse":0.06238280523642306,"envelope_correlation":0.9969343348915392,"rms_gain_error_db":-0.2395891667501196,"reference_mean_power":0.0019676998400409485,"output_mean_power":0.0018620868521447263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.0378624939923077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018862042103601662,"output_power":0.00017801797549038805,"reference_active":true,"delta_db":-0.2512485223093755},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.862312285918816e-05,"output_power":9.494579582586343e-05,"reference_active":true,"delta_db":-0.16503011105985346},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3789377405749525e-05,"output_power":1.1489243748013367e-05,"reference_active":true,"delta_db":-0.792532148674032}],"active_band_count":3,"active_band_mae_db":0.40293692734775366}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00018244538675777041,"log_rms_mae_db":0.48736314889774823,"rms_envelope_nrmse":0.04588918922417166,"envelope_correlation":0.998590158663145,"rms_gain_error_db":-0.23958821367902228,"reference_mean_power":0.001979010526094432,"output_mean_power":0.0018727908670881468},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":3.3102005537920435e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017836230285954771,"output_power":0.0001694382405908298,"reference_active":true,"delta_db":-0.2229163777881708},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3320761072042273e-05,"output_power":4.274038771554571e-05,"reference_active":true,"delta_db":-0.05857620326332391},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.132775238046075e-07,"output_power":6.845104490727279e-07,"reference_active":true,"delta_db":-0.7485868802023439}],"active_band_count":3,"active_band_mae_db":0.3433598204179462}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014606965842361196,"log_rms_mae_db":0.7709131110950306,"rms_envelope_nrmse":0.04723302841382845,"envelope_correlation":0.9983806792949197,"rms_gain_error_db":-0.12524339864381717,"reference_mean_power":0.0007714288246159527,"output_mean_power":0.0007494998020487524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.155891354220712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.355627855804231e-05,"output_power":8.358188291916131e-05,"reference_active":true,"delta_db":0.00133061566973697},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.117824128471733e-05,"output_power":1.943869965004539e-05,"reference_active":true,"delta_db":-0.3722268246093349},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3214088476749303e-06,"output_power":2.8949421440905017e-06,"reference_active":true,"delta_db":-0.5968244951684216}],"active_band_count":3,"active_band_mae_db":0.3234606451491645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001346093108760191,"log_rms_mae_db":0.3819952117724897,"rms_envelope_nrmse":0.026162772441134255,"envelope_correlation":0.9995897686980477,"rms_gain_error_db":-0.1252439440979278,"reference_mean_power":0.0007758659430163304,"output_mean_power":0.0007538106940214858},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.945223621465206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.346976557991944e-05,"output_power":7.335624241783702e-05,"reference_active":true,"delta_db":-0.006715771048663834},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57892144723787e-06,"output_power":8.73199358802924e-06,"reference_active":true,"delta_db":-0.40203203673373716},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.32297115664878e-07,"output_power":2.0706798230169334e-07,"reference_active":true,"delta_db":-0.499308656870519}],"active_band_count":3,"active_band_mae_db":0.30268548821764}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00012444218003818797,"log_rms_mae_db":0.8371747243725282,"rms_envelope_nrmse":0.0700848948079551,"envelope_correlation":0.9969102118501034,"rms_gain_error_db":-0.25135911851320814,"reference_mean_power":0.000687550351321365,"output_mean_power":0.0006488862873238929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.3390858183888006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.254326674690696e-05,"output_power":4.050873020361305e-05,"reference_active":true,"delta_db":-0.21282204885746997},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4859458498211297e-05,"output_power":1.3202593001904334e-05,"reference_active":true,"delta_db":-0.5134374791121045},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141405389451218e-06,"output_power":2.7378143758332475e-06,"reference_active":true,"delta_db":-0.5971998491110886}],"active_band_count":3,"active_band_mae_db":0.44115312569355436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001136976512596423,"log_rms_mae_db":0.5157314975099134,"rms_envelope_nrmse":0.04357163929117755,"envelope_correlation":0.9991498988798609,"rms_gain_error_db":-0.25135748513406614,"reference_mean_power":0.0006915015836222035,"output_mean_power":0.0006526155694211067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.214722623247202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7827908938064065e-05,"output_power":3.605024962072595e-05,"reference_active":true,"delta_db":-0.20904058675220036},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.619804659862373e-06,"output_power":4.984533824736561e-06,"reference_active":true,"delta_db":-0.5209667264564803},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.015023849518023e-07,"output_power":1.7922466715315386e-07,"reference_active":true,"delta_db":-0.5088240824204139}],"active_band_count":3,"active_band_mae_db":0.41294379854303154}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00010848831624356853,"log_rms_mae_db":0.9181962202711608,"rms_envelope_nrmse":0.09682915569748894,"envelope_correlation":0.9931138654741809,"rms_gain_error_db":-0.22784801707951324,"reference_mean_power":0.0002381846371729614,"output_mean_power":0.00022601067091285282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.715890316390027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.877460217795426e-05,"output_power":8.800788946262896e-05,"reference_active":true,"delta_db":-0.037671285903233154},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.954888518804952e-05,"output_power":3.9739134720948515e-05,"reference_active":true,"delta_db":0.020841605323789657},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.296333400161522e-06,"output_power":7.017850830646618e-06,"reference_active":true,"delta_db":-0.1690053777234875}],"active_band_count":3,"active_band_mae_db":0.07583942298350344}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":0.6176333921377761,"rms_envelope_nrmse":0.0623065725865752,"envelope_correlation":0.9972552005695149,"rms_gain_error_db":-0.22784585022702425,"reference_mean_power":0.00023955325002336745,"output_mean_power":0.00022730944536539309},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.5691901914560289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.470858674877873e-05,"output_power":9.445692953870681e-05,"reference_active":true,"delta_db":-0.011555319837124187},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8332956489415002e-05,"output_power":1.8972598403177583e-05,"reference_active":true,"delta_db":0.14894306307335758},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.840671065644935e-07,"output_power":4.723183206448369e-07,"reference_active":true,"delta_db":-0.10670780717985312}],"active_band_count":3,"active_band_mae_db":0.08906873003011162}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1580-141084-0047","method":"magphase","status":"measured","mos":{"status":"measured","value":3.6200883388519287,"predicted_mos":3.6200883388519287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":252292,"input_sample_rate":48000,"input_seconds":5.256083333333334,"input_float32_sha256":"773e74400b310d148bd8ba7cb5191f8931388d36ee44e0d4db00b23886a4d757","internal_samples":84098,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.595480028001475},"started_at":"2026-10-13T05:13:08.033810+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008008845001313603,"native_input_samples":252000,"native_input_sha256":"7a8b570cfb2afcd9fa6d7a667e5650674d2b5f523307e134dcafb422bf50efa7","output_sha256":"2cc0834f616ca35a6b029444598748fd49e3ad8712ead19034f8f57586c9ea43","output_samples":252292,"length_delta_samples":292,"analysis_s":3.1495248850005737,"synthesis_s":0.995005246000801,"total_s":4.144530131001375,"rtf":0.7894343106669285,"output_peak":0.696664626962469,"output_rms":0.04385629207345212,"finished_at":"2026-10-13T05:13:12.199807+00:00","evaluation_input":{"reference":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"output_sha256":"2cc0834f616ca35a6b029444598748fd49e3ad8712ead19034f8f57586c9ea43","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":84000,"output_samples":84098,"length_delta_samples":98,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.177928957728625,"output_peak":0.6979913696153301},"common_prefix":{"samples":84000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988346928583319,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84000,"output_samples":84000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.5187253856483505,"spectral_convergence":0.3853375349887329,"spectral_convergence_status":"ok","amplitude_floor":1.02822042353332e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.038533608057308,"spectral_convergence":0.27625569300202746,"spectral_convergence_status":"ok","amplitude_floor":8.928413196720056e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.652951059876067,"spectral_convergence":0.33290053971775246,"spectral_convergence_status":"ok","amplitude_floor":7.736392334055093e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":1.4784881078912855e-05,"output_power":1.389697769518616e-05,"delta_db":-0.26897475192557957,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002602315552138397,"output_power":0.00023577803466105757,"delta_db":-0.4285661389644486,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005931268593866254,"output_power":0.0005516897869673885,"delta_db":-0.3145264698469674,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005283196568297616,"output_power":0.0005315405859010929,"delta_db":0.026396612916636808,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003079318069923152,"output_power":0.00033887567791480696,"delta_db":0.41585848880869836,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00026143236338356946,"output_power":0.00024562200960947736,"delta_db":-0.2709206868365032,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.287540524883139,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10831903797626913,"pitch":null,"output_rms_dbfs":-27.172865809822206,"output_peak":0.6979913696153301,"reference_rms_dbfs":-27.06454677184594,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.403403351193909},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84000,"output_samples":84000,"intelligibility":{"estoi":{"value":0.9477070983461888,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04433764904553492,"rms_dbfs":-27.064546771845936,"peak":0.592376708984375,"peak_dbfs":-4.548040507351988,"crest_factor_db":22.51650626449395,"mean":-1.6414097377232142e-06,"zero_energy":false},"output":{"rms":0.0437881613309923,"rms_dbfs":-27.172865809822206,"peak":0.6979913696153301,"peak_dbfs":-3.1229989444310697,"crest_factor_db":24.049866865391138,"mean":0.0001229025861893902,"zero_energy":false},"rms_difference_db":-0.10831903797626907},"full_scale":{"reference":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":525,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.673642724752426e-06,"threshold_dbfs":-54.34903082709951,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":1.5888158540131854e-06,"output_mean_power":1.6003484751539384e-06,"output_rms_dbfs":-57.957854396203,"output_zero_energy":false,"output_peak":0.019827614298376362,"output_energy_fraction":0.000108106237031514,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":518,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":183,"both_estimated_voiced_frames":175,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.8706467661691543,"absolute_difference_cents_median":2.089157297513686,"absolute_difference_cents_p95":7.445977699376367,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00019312594202620684,"log_rms_mae_db":0.9169456785161247,"rms_envelope_nrmse":0.0725159926791139,"envelope_correlation":0.9951863623400372,"rms_gain_error_db":-0.10831823092987347,"reference_mean_power":0.0019676998400409485,"output_mean_power":0.0019192300160737288},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.0378624939923077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018862042103601662,"output_power":0.00017968202354071673,"reference_active":true,"delta_db":-0.21084080068088842},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.862312285918816e-05,"output_power":9.583863059165553e-05,"reference_active":true,"delta_db":-0.1243815057170187},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3789377405749525e-05,"output_power":1.3735100152394803e-05,"reference_active":true,"delta_db":-0.01712827582636403}],"active_band_count":3,"active_band_mae_db":0.11745019407475706}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00018244538675777041,"log_rms_mae_db":0.6030141190389199,"rms_envelope_nrmse":0.04241355032099202,"envelope_correlation":0.9983530491825698,"rms_gain_error_db":-0.10831678300906138,"reference_mean_power":0.001979010526094432,"output_mean_power":0.0019302627325642405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":3.3102005537920435e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017836230285954771,"output_power":0.00017059203564919975,"reference_active":true,"delta_db":-0.1934431934703032},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3320761072042273e-05,"output_power":4.283404772032517e-05,"reference_active":true,"delta_db":-0.0490696165623996},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.132775238046075e-07,"output_power":8.616980399815557e-07,"reference_active":true,"delta_db":0.251163350411688}],"active_band_count":3,"active_band_mae_db":0.16455872014813028}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014606965842361196,"log_rms_mae_db":1.1783547207290603,"rms_envelope_nrmse":0.06256567691671747,"envelope_correlation":0.9981676923949394,"rms_gain_error_db":-0.36395166718326244,"reference_mean_power":0.0007714288246159527,"output_mean_power":0.0007094155367542127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.155891354220712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.355627855804231e-05,"output_power":7.640618711588958e-05,"reference_active":true,"delta_db":-0.3885056136160941},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.117824128471733e-05,"output_power":1.979826728851053e-05,"reference_active":true,"delta_db":-0.2926270874251467},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3214088476749303e-06,"output_power":3.029373889451083e-06,"reference_active":true,"delta_db":-0.39969460360874565}],"active_band_count":3,"active_band_mae_db":0.36027576821666213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001346093108760191,"log_rms_mae_db":0.7665895192894592,"rms_envelope_nrmse":0.048339361014619844,"envelope_correlation":0.9994285567625218,"rms_gain_error_db":-0.3639511647930174,"reference_mean_power":0.0007758659430163304,"output_mean_power":0.0007134960485300049},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.945223621465206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.346976557991944e-05,"output_power":6.75328354211093e-05,"reference_active":true,"delta_db":-0.3659366983200216},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57892144723787e-06,"output_power":8.91339935294272e-06,"reference_active":true,"delta_db":-0.31273246861841014},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.32297115664878e-07,"output_power":2.1586218927190157e-07,"reference_active":true,"delta_db":-0.3186724002594944}],"active_band_count":3,"active_band_mae_db":0.33244718906597537}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00012444218003818797,"log_rms_mae_db":0.8884296993440437,"rms_envelope_nrmse":0.06683298915882621,"envelope_correlation":0.9968520133510397,"rms_gain_error_db":0.07079532402773964,"reference_mean_power":0.000687550351321365,"output_mean_power":0.0006988501146041665},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.3390858183888006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.254326674690696e-05,"output_power":4.233451562290814e-05,"reference_active":true,"delta_db":-0.02136239696820842},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4859458498211297e-05,"output_power":1.4196307457003885e-05,"reference_active":true,"delta_db":-0.198275868215574},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141405389451218e-06,"output_power":3.5187912579337878e-06,"reference_active":true,"delta_db":0.4926951985568444}],"active_band_count":3,"active_band_mae_db":0.23744448791354225}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.0001136976512596423,"log_rms_mae_db":0.5324837624666676,"rms_envelope_nrmse":0.03303913968650503,"envelope_correlation":0.9992528175424128,"rms_gain_error_db":0.07079672124299345,"reference_mean_power":0.0006915015836222035,"output_mean_power":0.0007028665108061805},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.214722623247202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7827908938064065e-05,"output_power":3.7821095813279635e-05,"reference_active":true,"delta_db":-0.0007822713639450279},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.619804659862373e-06,"output_power":5.203797072043598e-06,"reference_active":true,"delta_db":-0.33400867670480616},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.015023849518023e-07,"output_power":2.0570291203593833e-07,"reference_active":true,"delta_db":0.08960249082240587}],"active_band_count":3,"active_band_mae_db":0.1414644796303857}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00010848831624356853,"log_rms_mae_db":1.1199034593684192,"rms_envelope_nrmse":0.14402618030333228,"envelope_correlation":0.9873488047709861,"rms_gain_error_db":0.4906455683993015,"reference_mean_power":0.0002381846371729614,"output_mean_power":0.00026667254340339316},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.715890316390027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.877460217795426e-05,"output_power":9.815854698392342e-05,"reference_active":true,"delta_db":0.43639386223795823},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.954888518804952e-05,"output_power":4.153295484680128e-05,"reference_active":true,"delta_db":0.21258583904459533},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.296333400161522e-06,"output_power":7.984926801604607e-06,"reference_active":true,"delta_db":0.3916626876886128}],"active_band_count":3,"active_band_mae_db":0.34688079632372215}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84000,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":7.397564116979366e-05,"log_rms_mae_db":0.8343818952073003,"rms_envelope_nrmse":0.0947831155998122,"envelope_correlation":0.9955165588156465,"rms_gain_error_db":0.490636116192737,"reference_mean_power":0.00023955325002336745,"output_mean_power":0.0002682042644983982},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.5691901914560289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.470858674877873e-05,"output_power":0.00010399773048760674,"reference_active":true,"delta_db":0.40634505845784163},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8332956489415002e-05,"output_power":1.9277489659512558e-05,"reference_active":true,"delta_db":0.2181797113405988},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.840671065644935e-07,"output_power":5.272357666274197e-07,"reference_active":true,"delta_db":0.3709929189207831}],"active_band_count":3,"active_band_mae_db":0.3318392295730745}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"hnm","status":"measured","mos":{"status":"measured","value":3.0572102069854736,"predicted_mos":3.0572102069854736,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bf27783b196e231e4745201d14d03a06ae978662ac4704045b5ae617192ef1c5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0208969060004165},"started_at":"2026-10-13T04:38:26.092722+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4475999705609865e-05,"native_input_samples":96000,"native_input_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","output_sha256":"2e2b3289c6fdd506bb5b0ca7b88fc75eddd2a5618c965765d0b16929b5d1e1eb","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2275157100002616,"synthesis_s":0.5202664709995588,"total_s":1.7477821809998204,"rtf":0.2912970301666367,"output_peak":0.29556243369469054,"output_rms":0.0665709155846394,"finished_at":"2026-10-13T04:38:27.846396+00:00","evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"2e2b3289c6fdd506bb5b0ca7b88fc75eddd2a5618c965765d0b16929b5d1e1eb","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.112039637831189,"spectral_convergence":0.1335200099551048,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.873841190855845,"spectral_convergence":0.1288709669473348,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.019832980598277,"spectral_convergence":0.13310506333278282,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.0013655320355939503,"delta_db":0.01898026094199549,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.002217925500114068,"delta_db":-0.0759620494608452,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0006697143715557329,"delta_db":-0.1772804495127908,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.00010201427621514311,"delta_db":-0.18645659383106153,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":2.41996180434155e-05,"delta_db":-0.5434004394286508,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":5.230100025487458e-05,"delta_db":-0.28980567339465363,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2153142444283329,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07037051023805374,"pitch":null,"output_rms_dbfs":-23.534309398559444,"output_peak":0.29556243369469054,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.001904603095103},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.112039637831189,"spectral_convergence":0.1335200099551048,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.873841190855845,"spectral_convergence":0.1288709669473348,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.019832980598277,"spectral_convergence":0.13310506333278282,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.0013655320355939503,"delta_db":0.01898026094199549,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.002217925500114068,"delta_db":-0.0759620494608452,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0006697143715557329,"delta_db":-0.1772804495127908,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.00010201427621514311,"delta_db":-0.18645659383106153,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":2.41996180434155e-05,"delta_db":-0.5434004394286508,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":5.230100025487458e-05,"delta_db":-0.28980567339465363,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2153142444283329,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07037051023805374,"pitch":null,"output_rms_dbfs":-23.534309398559444,"output_peak":0.29556243369469054,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.001904603095103},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9588399112880327,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06711244415402366,"rms_dbfs":-23.46393888832139,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.860774560984794,"mean":-7.2549184163411456e-06,"zero_energy":false},"output":{"rms":0.0665709155846394,"rms_dbfs":-23.534309398559444,"peak":0.29556243369469054,"peak_dbfs":-10.587015321367941,"crest_factor_db":12.947294077191504,"mean":-1.574624388741831e-05,"zero_energy":false},"rms_difference_db":-0.07037051023805319},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":1.5883571059780247e-06,"output_rms_dbfs":-57.99051849692128,"output_zero_energy":false,"output_peak":0.02401092976312277,"output_energy_fraction":4.898258402721035e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":393,"output_estimated_voiced_frames":390,"both_estimated_voiced_frames":380,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9669211195928753,"absolute_difference_cents_median":1.4983615910710606,"absolute_difference_cents_p95":12.076692376562372,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016450226550328756,"log_rms_mae_db":0.6260666122747014,"rms_envelope_nrmse":0.048797570082931174,"envelope_correlation":0.9968224420722049,"rms_gain_error_db":-0.07033765691320143,"reference_mean_power":0.004505860165227205,"output_mean_power":0.004433471734956138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3773993983732174e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011178733037365022,"output_power":0.001081935556490554,"reference_active":true,"delta_db":-0.14191191175743764},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011348254393479664,"output_power":0.00010381190369683088,"reference_active":true,"delta_db":-0.3868190761600205},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1976543085088728e-05,"output_power":1.907606949706999e-05,"reference_active":true,"delta_db":-0.6147048287864986}],"active_band_count":3,"active_band_mae_db":0.3811452722346522}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014948818281718847,"log_rms_mae_db":0.33596879391609,"rms_envelope_nrmse":0.027402175483250548,"envelope_correlation":0.9989509214841964,"rms_gain_error_db":-0.07131823633421959,"reference_mean_power":0.004512140106290299,"output_mean_power":0.004438648485085661},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1895509906107845e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130343468129068,"output_power":0.0009820702991672637,"reference_active":true,"delta_db":-0.1348159347678714},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.38575581348341e-05,"output_power":4.85944247747002e-05,"reference_active":true,"delta_db":-0.4466021327340982},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0969765535512897e-06,"output_power":1.6652998220367222e-06,"reference_active":true,"delta_db":-1.0010113897364048}],"active_band_count":3,"active_band_mae_db":0.5274764857461248}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012259006245866322,"log_rms_mae_db":1.4731529337542597,"rms_envelope_nrmse":0.07881583119718419,"envelope_correlation":0.994104876966685,"rms_gain_error_db":-0.11291365384975062,"reference_mean_power":0.001387125934818555,"output_mean_power":0.0013515263780427742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.931300849598276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005535817649712056,"output_power":0.0005225234344981537,"reference_active":true,"delta_db":-0.250760032244489},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.308179193682493e-05,"output_power":6.261475591419946e-05,"reference_active":true,"delta_db":-0.6713249559357124},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.782827787780852e-06,"output_power":6.646810651437357e-06,"reference_active":true,"delta_db":0.6047304791079813}],"active_band_count":3,"active_band_mae_db":0.5089384890960609}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001104267603656376,"log_rms_mae_db":0.9731649338378242,"rms_envelope_nrmse":0.046843669060026925,"envelope_correlation":0.9978943143608971,"rms_gain_error_db":-0.11517708228024648,"reference_mean_power":0.0013888741987160204,"output_mean_power":0.001352524690088284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.150356442217087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005149581742784205,"output_power":0.0004860327103913174,"reference_active":true,"delta_db":-0.2510645777245589},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177504958330739e-05,"output_power":2.74202934116669e-05,"reference_active":true,"delta_db":-0.6401413926105417},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.976808620230845e-07,"output_power":6.997229657667349e-07,"reference_active":true,"delta_db":-0.5690304434927753}],"active_band_count":3,"active_band_mae_db":0.4867454712759587}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":1.2673202566763906,"rms_envelope_nrmse":0.10394549510536139,"envelope_correlation":0.9908226220275443,"rms_gain_error_db":-0.22532509131467476,"reference_mean_power":0.00011418247196706925,"output_mean_power":0.00010840939775512908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.040353626229237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.425624857591464e-05,"output_power":4.0628802875875576e-05,"reference_active":true,"delta_db":-0.37140571689978275},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4228259000934756e-06,"output_power":2.9469134766806074e-06,"reference_active":true,"delta_db":-0.6501742471619743},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0991841496467986e-06,"output_power":1.0306379524231325e-06,"reference_active":true,"delta_db":-0.2796432621916415}],"active_band_count":3,"active_band_mae_db":0.4337410754177995}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":0.7990555268858748,"rms_envelope_nrmse":0.06219628458777281,"envelope_correlation":0.996890048135031,"rms_gain_error_db":-0.2253881547528688,"reference_mean_power":0.0001146281153210772,"output_mean_power":0.00010883092901448112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.50871602576763e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.069122651510938e-05,"output_power":3.727495685130507e-05,"reference_active":true,"delta_db":-0.38083631161259984},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.65433446148589e-06,"output_power":1.4015599288968569e-06,"reference_active":true,"delta_db":-0.7201164427356563},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.669640028113729e-08,"output_power":5.7544437782824686e-08,"reference_active":true,"delta_db":0.06447865384373477}],"active_band_count":3,"active_band_mae_db":0.388477136063997}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":1.4140646765588507,"rms_envelope_nrmse":0.1570205074209418,"envelope_correlation":0.9844833544622807,"rms_gain_error_db":-0.3565300565948853,"reference_mean_power":4.027652017807585e-05,"output_mean_power":3.7102137924637015e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067374336554811e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6348738836101644e-05,"output_power":1.5381981410320856e-05,"reference_active":true,"delta_db":-0.26471974064051546},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.842264617940261e-06,"output_power":5.0056397112314735e-06,"reference_active":true,"delta_db":-0.6712163652817299},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8971365027113356e-06,"output_power":9.08159086165362e-07,"reference_active":true,"delta_db":-3.199366479240661}],"active_band_count":3,"active_band_mae_db":1.378434195054302}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":0.8085188004095314,"rms_envelope_nrmse":0.08815351073435934,"envelope_correlation":0.9956265422799876,"rms_gain_error_db":-0.35689199140741157,"reference_mean_power":4.047626840683264e-05,"output_mean_power":3.7283035810312874e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8536334265621368e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7722714720556834e-05,"output_power":1.6067649284940924e-05,"reference_active":true,"delta_db":-0.42577903133072603},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.276595703569918e-06,"output_power":2.502658125854999e-06,"reference_active":true,"delta_db":-1.1702133033463744},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.751293843816523e-07,"output_power":7.65170178524668e-08,"reference_active":true,"delta_db":-3.596009853921118}],"active_band_count":3,"active_band_mae_db":1.7306673961994061}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.268627643585205,"predicted_mos":3.268627643585205,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"71e21fb5107c697e95a3aa773f6f1638b7969f41db4d218a710948c81fba9c40","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0371388450002996},"started_at":"2026-10-13T04:41:24.033541+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1611999727610964e-05,"native_input_samples":96000,"native_input_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","output_sha256":"9cf2c02668ad7b906d1854822fee6f69681589a46729be4c4a582601b3c065d1","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3957147350001833,"synthesis_s":0.10754152199933742,"total_s":1.5032562569995207,"rtf":0.25054270949992014,"output_peak":0.40955936492958606,"output_rms":0.076184988191965,"finished_at":"2026-10-13T04:41:25.544090+00:00","evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"9cf2c02668ad7b906d1854822fee6f69681589a46729be4c4a582601b3c065d1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.839942618998341,"spectral_convergence":0.47103589431551596,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.770735911972126,"spectral_convergence":0.22527020756015476,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.154222986597346,"spectral_convergence":0.25360468494289523,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.0016310306828820416,"delta_db":0.7905826392986905,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.0029838474983983803,"delta_db":1.2123346389287342,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0009490817712801219,"delta_db":1.3368596850403383,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.00014803965045887294,"delta_db":1.4307143899383614,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":3.300561543257039e-05,"delta_db":0.8043927989171602,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":5.914720735785979e-05,"delta_db":0.24443682030969427,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9698868287388298,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1013269797669383,"pitch":null,"output_rms_dbfs":-22.362611908554456,"output_peak":0.40955936492958606,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.254967172522605},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.839942618998341,"spectral_convergence":0.47103589431551596,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.770735911972126,"spectral_convergence":0.22527020756015476,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.154222986597346,"spectral_convergence":0.25360468494289523,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.0016310306828820416,"delta_db":0.7905826392986905,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.0029838474983983803,"delta_db":1.2123346389287342,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0009490817712801219,"delta_db":1.3368596850403383,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.00014803965045887294,"delta_db":1.4307143899383614,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":3.300561543257039e-05,"delta_db":0.8043927989171602,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":5.914720735785979e-05,"delta_db":0.24443682030969427,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9698868287388298,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1013269797669383,"pitch":null,"output_rms_dbfs":-22.362611908554456,"output_peak":0.40955936492958606,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.254967172522605},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9624280291062257,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06711244415402366,"rms_dbfs":-23.46393888832139,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.860774560984794,"mean":-7.2549184163411456e-06,"zero_energy":false},"output":{"rms":0.076184988191965,"rms_dbfs":-22.362611908554456,"peak":0.40955936492958606,"peak_dbfs":-7.753662781967266,"crest_factor_db":14.608949126587188,"mean":1.8218810226640866e-05,"zero_energy":false},"rms_difference_db":1.1013269797669345},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":3.958401547370975e-06,"output_rms_dbfs":-54.02480152281737,"output_zero_energy":false,"output_peak":0.02494863486386977,"output_energy_fraction":9.320595069173844e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":393,"output_estimated_voiced_frames":388,"both_estimated_voiced_frames":365,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.9287531806615776,"absolute_difference_cents_median":6.5480370057515325,"absolute_difference_cents_p95":25.619211785340227,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016450226550328756,"log_rms_mae_db":1.6056613731728144,"rms_envelope_nrmse":0.17427138436769032,"envelope_correlation":0.9855768713306392,"rms_gain_error_db":1.103111021208409,"reference_mean_power":0.004505860165227205,"output_mean_power":0.005808831941303563},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3773993983732174e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011178733037365022,"output_power":0.00141125154267685,"reference_active":true,"delta_db":1.0121184491079749},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011348254393479664,"output_power":0.00015803447387980013,"reference_active":true,"delta_db":1.4382277214658576},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1976543085088728e-05,"output_power":2.906877186572294e-05,"reference_active":true,"delta_db":1.2146730468961042}],"active_band_count":3,"active_band_mae_db":1.221673072489979}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014948818281718847,"log_rms_mae_db":1.1776904322337232,"rms_envelope_nrmse":0.15812434818861554,"envelope_correlation":0.9912437350360819,"rms_gain_error_db":1.1060885263797802,"reference_mean_power":0.004512140106290299,"output_mean_power":0.005820917299239222},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1895509906107845e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130343468129068,"output_power":0.0012944598075696856,"reference_active":true,"delta_db":1.0646440000567454},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.38575581348341e-05,"output_power":7.903972997708655e-05,"reference_active":true,"delta_db":1.6659878900016682},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0969765535512897e-06,"output_power":2.1345923179555987e-06,"reference_active":true,"delta_db":0.07721367525099786}],"active_band_count":3,"active_band_mae_db":0.9359485217698037}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012259006245866322,"log_rms_mae_db":1.7318810681541978,"rms_envelope_nrmse":0.22580590790910643,"envelope_correlation":0.9811701895201238,"rms_gain_error_db":1.3427166948973146,"reference_mean_power":0.001387125934818555,"output_mean_power":0.0018896769307365634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.931300849598276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005535817649712056,"output_power":0.0007651216585849612,"reference_active":true,"delta_db":1.4054871988588262},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.308179193682493e-05,"output_power":0.00012086925564189007,"reference_active":true,"delta_db":2.185066600272088},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.782827787780852e-06,"output_power":9.40691353460422e-06,"reference_active":true,"delta_db":2.113068929087928}],"active_band_count":3,"active_band_mae_db":1.9012075760729477}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001104267603656376,"log_rms_mae_db":1.3444940306977127,"rms_envelope_nrmse":0.2117825551962457,"envelope_correlation":0.9849956322595589,"rms_gain_error_db":1.3454404944615408,"reference_mean_power":0.0013888741987160204,"output_mean_power":0.0018932456144535315},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.150356442217087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005149581742784205,"output_power":0.0007273973315953953,"reference_active":true,"delta_db":1.4999974715828337},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177504958330739e-05,"output_power":5.2868759189183276e-05,"reference_active":true,"delta_db":2.2111288086853005},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.976808620230845e-07,"output_power":9.346648578075538e-07,"reference_active":true,"delta_db":0.6882674151124367}],"active_band_count":3,"active_band_mae_db":1.4664645651268569}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":1.361044849436141,"rms_envelope_nrmse":0.23099892469107786,"envelope_correlation":0.9847859008394793,"rms_gain_error_db":1.4272993066174087,"reference_mean_power":0.00011418247196706925,"output_mean_power":0.00015860956410700358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.040353626229237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.425624857591464e-05,"output_power":6.823961091367542e-05,"reference_active":true,"delta_db":1.8806194416484465},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4228259000934756e-06,"output_power":6.258767781460395e-06,"reference_active":true,"delta_db":2.6210402867877622},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0991841496467986e-06,"output_power":1.5086848568262116e-06,"reference_active":true,"delta_db":1.3752807401603304}],"active_band_count":3,"active_band_mae_db":1.9589801561988465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":1.0468742871844552,"rms_envelope_nrmse":0.20947712373057714,"envelope_correlation":0.9911499777139556,"rms_gain_error_db":1.4258215163443102,"reference_mean_power":0.0001146281153210772,"output_mean_power":0.0001591744300842128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.50871602576763e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.069122651510938e-05,"output_power":6.324550954368734e-05,"reference_active":true,"delta_db":1.9152891530508764},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.65433446148589e-06,"output_power":3.0139237969343358e-06,"reference_active":true,"delta_db":2.605089509180643},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.669640028113729e-08,"output_power":6.777965519363895e-08,"reference_active":true,"delta_db":0.7754386926064422}],"active_band_count":3,"active_band_mae_db":1.7652724516126537}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":1.4456440420858079,"rms_envelope_nrmse":0.3412045888285737,"envelope_correlation":0.9338376847251014,"rms_gain_error_db":0.6511573261694892,"reference_mean_power":4.027652017807585e-05,"output_mean_power":4.679157608551439e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067374336554811e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6348738836101644e-05,"output_power":1.962438112359566e-05,"reference_active":true,"delta_db":0.7931171345716259},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.842264617940261e-06,"output_power":9.782496982749986e-06,"reference_active":true,"delta_db":2.2386849872291497},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8971365027113356e-06,"output_power":4.110779835826603e-06,"reference_active":true,"delta_db":3.3582563720510255}],"active_band_count":3,"active_band_mae_db":2.1300194979506}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":0.9185091997751971,"rms_envelope_nrmse":0.2117390126774941,"envelope_correlation":0.9752261264894885,"rms_gain_error_db":0.6512389425660586,"reference_mean_power":4.047626840683264e-05,"output_mean_power":4.702451893813312e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8536334265621368e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7722714720556834e-05,"output_power":2.290254990236632e-05,"reference_active":true,"delta_db":1.113535913233021},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.276595703569918e-06,"output_power":5.906495420872308e-06,"reference_active":true,"delta_db":2.5590701415250843},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.751293843816523e-07,"output_power":3.9389302184422126e-07,"reference_active":true,"delta_db":3.5201926593896484}],"active_band_count":3,"active_band_mae_db":2.397599571382585}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"facodec","status":"measured","mos":{"status":"measured","value":4.536036968231201,"predicted_mos":4.536036968231201,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"641afa71e5203d1a197ee1fb70ff6d3a06f9950352f1efda096bbc97431b8713","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.021901340000113},"started_at":"2026-10-13T04:47:41.256651+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.718599989748327e-05,"native_input_samples":96000,"native_input_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","output_sha256":"dd511ec2b5ee50200c86ea6cd2731c8d2b5a4f4c19cc30d9a4028a02bef0b5fc","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4303213769999275,"synthesis_s":6.008123192999847,"total_s":7.438444569999774,"rtf":1.239740761666629,"output_peak":0.26562783122062683,"output_rms":0.0633934438690877,"finished_at":"2026-10-13T04:47:48.703371+00:00","evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"dd511ec2b5ee50200c86ea6cd2731c8d2b5a4f4c19cc30d9a4028a02bef0b5fc","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.210053123863224,"spectral_convergence":0.18636594496461417,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.663782549500025,"spectral_convergence":0.17185830562386692,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.600542307280938,"spectral_convergence":0.17596973382561623,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.001189690094524703,"delta_db":-0.5797002149454213,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.0019749313252302764,"delta_db":-0.5799116070106471,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0006732311425649931,"delta_db":-0.1545346609662755,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.00010169794960152634,"delta_db":-0.19994415086298262,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":2.3393716155559918e-05,"delta_db":-0.6904933910491922,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":5.578449750611416e-05,"delta_db":-0.009770367457983122,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3690590653820837,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.49517419969246856,"pitch":null,"output_rms_dbfs":-23.95911308801386,"output_peak":0.26562783122062683,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.4914593268813965},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.210053123863224,"spectral_convergence":0.18636594496461417,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.663782549500025,"spectral_convergence":0.17185830562386692,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.600542307280938,"spectral_convergence":0.17596973382561623,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.001189690094524703,"delta_db":-0.5797002149454213,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.0019749313252302764,"delta_db":-0.5799116070106471,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0006732311425649931,"delta_db":-0.1545346609662755,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.00010169794960152634,"delta_db":-0.19994415086298262,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":2.3393716155559918e-05,"delta_db":-0.6904933910491922,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":5.578449750611416e-05,"delta_db":-0.009770367457983122,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3690590653820837,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.49517419969246856,"pitch":null,"output_rms_dbfs":-23.95911308801386,"output_peak":0.26562783122062683,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.4914593268813965},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9340731268348518,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06711244415402366,"rms_dbfs":-23.46393888832139,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.860774560984794,"mean":-7.2549184163411456e-06,"zero_energy":false},"output":{"rms":0.0633934438690877,"rms_dbfs":-23.95911308801386,"peak":0.26562783122062683,"peak_dbfs":-11.514528472201304,"crest_factor_db":12.444584615812555,"mean":5.890203747588056e-05,"zero_energy":false},"rms_difference_db":-0.495174199692471},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":4.078374652703996e-07,"output_rms_dbfs":-63.8951288102766,"output_zero_energy":false,"output_peak":0.007299205753952265,"output_energy_fraction":1.3869507181577585e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":393,"output_estimated_voiced_frames":387,"both_estimated_voiced_frames":382,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9720101781170484,"absolute_difference_cents_median":2.7849308622599693,"absolute_difference_cents_p95":13.518613272112775,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016450226550328756,"log_rms_mae_db":0.8926035409680244,"rms_envelope_nrmse":0.10084808118219203,"envelope_correlation":0.9897659140382228,"rms_gain_error_db":-0.4957964799856931,"reference_mean_power":0.004505860165227205,"output_mean_power":0.004019740909233656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3773993983732174e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011178733037365022,"output_power":0.0008812139414787274,"reference_active":true,"delta_db":-1.0331122535737063},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011348254393479664,"output_power":9.590630232866518e-05,"reference_active":true,"delta_db":-0.7308191572640945},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1976543085088728e-05,"output_power":2.0893207591728968e-05,"reference_active":true,"delta_db":-0.21954259475290983}],"active_band_count":3,"active_band_mae_db":0.6611580018635702}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014948818281718847,"log_rms_mae_db":0.746081911522832,"rms_envelope_nrmse":0.09480510140682895,"envelope_correlation":0.9905063827984127,"rms_gain_error_db":-0.49790347157675113,"reference_mean_power":0.004512140106290299,"output_mean_power":0.00402339089983228},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1895509906107845e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130343468129068,"output_power":0.0007901586754878768,"reference_active":true,"delta_db":-1.0790985754403575},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.38575581348341e-05,"output_power":4.555092108371671e-05,"reference_active":true,"delta_db":-0.7274949565191664},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0969765535512897e-06,"output_power":1.9764045384445208e-06,"reference_active":true,"delta_db":-0.2571773211919475}],"active_band_count":3,"active_band_mae_db":0.6879236177171572}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012259006245866322,"log_rms_mae_db":1.3289886892761553,"rms_envelope_nrmse":0.11884972656610975,"envelope_correlation":0.9870626303689884,"rms_gain_error_db":-0.32637299572442846,"reference_mean_power":0.001387125934818555,"output_mean_power":0.0012867038254725002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.931300849598276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005535817649712056,"output_power":0.0004448431634884552,"reference_active":true,"delta_db":-0.9497485542841515},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.308179193682493e-05,"output_power":5.8238981207766635e-05,"reference_active":true,"delta_db":-0.9859541841531245},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.782827787780852e-06,"output_power":6.723169214632249e-06,"reference_active":true,"delta_db":0.6543378286157163}],"active_band_count":3,"active_band_mae_db":0.8633468556843308}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001104267603656376,"log_rms_mae_db":1.000240248827508,"rms_envelope_nrmse":0.10971808531859535,"envelope_correlation":0.9881630997041544,"rms_gain_error_db":-0.33060903845299927,"reference_mean_power":0.0013888741987160204,"output_mean_power":0.0012870695217856822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.150356442217087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005149581742784205,"output_power":0.0004103091055291144,"reference_active":true,"delta_db":-0.9866080150188772},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177504958330739e-05,"output_power":2.5840631499852668e-05,"reference_active":true,"delta_db":-0.8978311406993111},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.976808620230845e-07,"output_power":8.292348374311311e-07,"reference_active":true,"delta_db":0.16848366610962529}],"active_band_count":3,"active_band_mae_db":0.6843076072759379}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":1.3734464344347515,"rms_envelope_nrmse":0.1584575576036701,"envelope_correlation":0.9782828226033433,"rms_gain_error_db":-0.14585336887981987,"reference_mean_power":0.00011418247196706925,"output_mean_power":0.00011041144762746234},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.040353626229237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.425624857591464e-05,"output_power":4.511841444334178e-05,"reference_active":true,"delta_db":0.08379231459414348},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4228259000934756e-06,"output_power":5.150351206375506e-06,"reference_active":true,"delta_db":1.7745203534134377},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0991841496467986e-06,"output_power":1.2492738167707006e-06,"reference_active":true,"delta_db":0.555871806095596}],"active_band_count":3,"active_band_mae_db":0.8047281580343925}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":1.030443479329719,"rms_envelope_nrmse":0.1345783121751503,"envelope_correlation":0.9832447422372151,"rms_gain_error_db":-0.146835062714933,"reference_mean_power":0.0001146281153210772,"output_mean_power":0.00011081732068052144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.50871602576763e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.069122651510938e-05,"output_power":4.163550571729723e-05,"reference_active":true,"delta_db":0.09963063540385397},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.65433446148589e-06,"output_power":2.4543804346575285e-06,"reference_active":true,"delta_db":1.713185636207767},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.669640028113729e-08,"output_power":6.248480085921702e-08,"reference_active":true,"delta_db":0.4221890413426544}],"active_band_count":3,"active_band_mae_db":0.7450017709847585}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":1.7235143942771924,"rms_envelope_nrmse":0.26815058162681377,"envelope_correlation":0.9524175002761196,"rms_gain_error_db":-0.4029509392818532,"reference_mean_power":4.027652017807585e-05,"output_mean_power":3.670767241007228e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067374336554811e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6348738836101644e-05,"output_power":1.510147063336192e-05,"reference_active":true,"delta_db":-0.34465013773709},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.842264617940261e-06,"output_power":5.804196003209185e-06,"reference_active":true,"delta_db":-0.0283915409681125},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8971365027113356e-06,"output_power":1.3809311755651549e-06,"reference_active":true,"delta_db":-1.3792654610650694}],"active_band_count":3,"active_band_mae_db":0.584102379923424}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":1.3968326067742534,"rms_envelope_nrmse":0.20892989099271633,"envelope_correlation":0.9696583330587615,"rms_gain_error_db":-0.4032061747376787,"reference_mean_power":4.047626840683264e-05,"output_mean_power":3.688755326927322e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8536334265621368e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7722714720556834e-05,"output_power":1.6264449225385273e-05,"reference_active":true,"delta_db":-0.3729088573152609},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.276595703569918e-06,"output_power":2.946209417371241e-06,"reference_active":true,"delta_db":-0.46159244077377865},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.751293843816523e-07,"output_power":1.0781589649155466e-07,"reference_active":true,"delta_db":-2.10676222619229}],"active_band_count":3,"active_band_mae_db":0.9804211747604432}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.621885299682617,"predicted_mos":3.621885299682617,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"d99f58960ba501eb6a940ce2217be1d8b51edd836440ffd7c929022209db40b5","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.994359984997573},"started_at":"2026-10-13T07:04:55.512676+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006698523999148165,"native_input_samples":264600,"native_input_sha256":"be27c45f858a6150def3403cdfb72f68d14a2d6e2618d5b8b7073ad98940b7b3","output_sha256":"b0304abd9438f03cf9ce1458a9baa39b09322c35334c99f0edb3ab265afaf17e","output_samples":264704,"length_delta_samples":104,"analysis_s":0.16815932099962083,"synthesis_s":36.047020216999954,"total_s":36.215179537999575,"rtf":6.035863256333262,"output_peak":0.32446494698524475,"output_rms":0.062374868567797725,"finished_at":"2026-10-13T07:05:31.757856+00:00","evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"b0304abd9438f03cf9ce1458a9baa39b09322c35334c99f0edb3ab265afaf17e","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.097810421530824,"output_peak":0.32367044290703245},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.792327988533158,"spectral_convergence":0.4303548044271281,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.594869492291215,"spectral_convergence":0.2719379689729998,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.354025134885495,"spectral_convergence":0.3178624645346161,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.0010246360312024638,"delta_db":-1.2283424339523816,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.002083812344425849,"delta_db":-0.3468455265174846,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0006487144053588141,"delta_db":-0.31564121628204067,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":7.121522521636158e-05,"delta_db":-1.7473375993193905,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":2.2683469683366713e-05,"delta_db":-0.824390698902346,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":4.039294124566535e-05,"delta_db":-1.411850843176855,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9790680530250829,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6349416349697534,"pitch":null,"output_rms_dbfs":-24.098880523291143,"output_peak":0.32367044290703245,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.580407538569956},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8329029827985137,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06711244415402366,"rms_dbfs":-23.46393888832139,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.860774560984794,"mean":-7.2549184163411456e-06,"zero_energy":false},"output":{"rms":0.06238152304274496,"rms_dbfs":-24.098880523291143,"peak":0.32367044290703245,"peak_dbfs":-9.797939157976533,"crest_factor_db":14.300941365314612,"mean":-9.76114625205912e-05,"zero_energy":false},"rms_difference_db":-0.6349416349697528},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":1.091778182074585e-06,"output_rms_dbfs":-59.61865588816732,"output_zero_energy":false,"output_peak":0.027103776770080926,"output_energy_fraction":3.834290958840888e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":393,"output_estimated_voiced_frames":385,"both_estimated_voiced_frames":367,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9338422391857506,"absolute_difference_cents_median":9.202568348346684,"absolute_difference_cents_p95":45.83093885234314,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016450226550328756,"log_rms_mae_db":1.2586211097551974,"rms_envelope_nrmse":0.10557749139108721,"envelope_correlation":0.9912822421098805,"rms_gain_error_db":-0.6343136306821315,"reference_mean_power":0.004505860165227205,"output_mean_power":0.003893555293729849},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3773993983732174e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011178733037365022,"output_power":0.0009963626647080755,"reference_active":true,"delta_db":-0.4997513929623295},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011348254393479664,"output_power":8.941233747233182e-05,"reference_active":true,"delta_db":-1.035316141771403},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1976543085088728e-05,"output_power":1.912378157848556e-05,"reference_active":true,"delta_db":-0.6038560407045174}],"active_band_count":3,"active_band_mae_db":0.7129745251460834}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014948818281718847,"log_rms_mae_db":0.9259463596100918,"rms_envelope_nrmse":0.0883870123909141,"envelope_correlation":0.9955682639718397,"rms_gain_error_db":-0.6316004931975282,"reference_mean_power":0.004512140106290299,"output_mean_power":0.0039014183932895206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1895509906107845e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130343468129068,"output_power":0.0009044571960612143,"reference_active":true,"delta_db":-0.49236151859548044},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.38575581348341e-05,"output_power":4.2116253881847194e-05,"reference_active":true,"delta_db":-1.0679692390283966},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0969765535512897e-06,"output_power":1.7969313071207999e-06,"reference_active":true,"delta_db":-0.6706209934622837}],"active_band_count":3,"active_band_mae_db":0.7436505836953868}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012259006245866322,"log_rms_mae_db":2.2949182163218875,"rms_envelope_nrmse":0.19468856578838384,"envelope_correlation":0.96269043454185,"rms_gain_error_db":-0.006830890035267255,"reference_mean_power":0.001387125934818555,"output_mean_power":0.001384945880198534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.931300849598276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005535817649712056,"output_power":0.0005060418772732149,"reference_active":true,"delta_db":-0.38995317812407493},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.308179193682493e-05,"output_power":6.58376266043384e-05,"reference_active":true,"delta_db":-0.45335021035840334},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.782827787780852e-06,"output_power":9.559919380273651e-06,"reference_active":true,"delta_db":2.1831397060813735}],"active_band_count":3,"active_band_mae_db":1.0088143648546173}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001104267603656376,"log_rms_mae_db":1.8539007552420328,"rms_envelope_nrmse":0.17330873501394867,"envelope_correlation":0.967479766352354,"rms_gain_error_db":-0.004823994130839254,"reference_mean_power":0.0013888741987160204,"output_mean_power":0.0013873323413776822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.150356442217087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005149581742784205,"output_power":0.00046622297653366385,"reference_active":true,"delta_db":-0.43178283648916205},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177504958330739e-05,"output_power":3.0399435586677636e-05,"reference_active":true,"delta_db":-0.1922071659680219},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.976808620230845e-07,"output_power":1.0980330404132338e-06,"reference_active":true,"delta_db":1.3878623589519825}],"active_band_count":3,"active_band_mae_db":0.6706174538030555}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":3.297336810929905,"rms_envelope_nrmse":0.3098931744735495,"envelope_correlation":0.9344999656068019,"rms_gain_error_db":-1.8472730617051365,"reference_mean_power":0.00011418247196706925,"output_mean_power":7.462290208713995e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.040353626229237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.425624857591464e-05,"output_power":2.4381400621199e-05,"reference_active":true,"delta_db":-2.5891594714687156},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4228259000934756e-06,"output_power":2.9372989683583935e-06,"reference_active":true,"delta_db":-0.6643665682771218},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0991841496467986e-06,"output_power":8.760237863599849e-07,"reference_active":true,"delta_db":-0.9855455861703519}],"active_band_count":3,"active_band_mae_db":1.4130238753053963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":2.8267943150810075,"rms_envelope_nrmse":0.28332990157142923,"envelope_correlation":0.9485939601245396,"rms_gain_error_db":-1.843200681877343,"reference_mean_power":0.0001146281153210772,"output_mean_power":7.498442814396754e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.50871602576763e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.069122651510938e-05,"output_power":2.233047267308919e-05,"reference_active":true,"delta_db":-2.6060286459923363},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.65433446148589e-06,"output_power":1.4632846642068252e-06,"reference_active":true,"delta_db":-0.5329449559389314},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.669640028113729e-08,"output_power":5.620742262852977e-08,"reference_active":true,"delta_db":-0.03761814588431764}],"active_band_count":3,"active_band_mae_db":1.0588639159385285}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":3.366946444141574,"rms_envelope_nrmse":0.26329976313837533,"envelope_correlation":0.9543252335566258,"rms_gain_error_db":-0.3061971560708594,"reference_mean_power":4.027652017807585e-05,"output_mean_power":3.7534637472673935e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067374336554811e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6348738836101644e-05,"output_power":1.6561331397316513e-05,"reference_active":true,"delta_db":0.056109913990924916},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.842264617940261e-06,"output_power":4.368624096836123e-06,"reference_active":true,"delta_db":-1.2623654693778525},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8971365027113356e-06,"output_power":1.1920484857819872e-06,"reference_active":true,"delta_db":-2.018046599533891}],"active_band_count":3,"active_band_mae_db":1.1121739943008895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":2.8158653453534734,"rms_envelope_nrmse":0.22096704370601897,"envelope_correlation":0.965829500819447,"rms_gain_error_db":-0.30596808343657206,"reference_mean_power":4.047626840683264e-05,"output_mean_power":3.772277722040407e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8536334265621368e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7722714720556834e-05,"output_power":1.7108103972664984e-05,"reference_active":true,"delta_db":-0.1532836580452976},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.276595703569918e-06,"output_power":2.175431327315967e-06,"reference_active":true,"delta_db":-1.7787747914887033},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.751293843816523e-07,"output_power":1.1461182548993592e-07,"reference_active":true,"delta_db":-1.8412959118736416}],"active_band_count":3,"active_band_mae_db":1.257784787135881}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"golf","status":"measured","mos":{"status":"measured","value":2.8684403896331787,"predicted_mos":2.8684403896331787,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"757eb08228c52e481922793e7f6cb3ba0eeb539c2f857995824162270ff3b15e","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3368535049994534},"started_at":"2026-10-13T05:02:04.065179+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002798393000375654,"native_input_samples":144000,"native_input_sha256":"d086d02f1071621e604cce910000ce5ef03b3f290ce0aa589c79a8b5ad3c3d48","output_sha256":"5f1eca015dbe455afe18973619048925b4d05dee6463dd9499aa18ee4b26aadd","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.7531555659998048,"synthesis_s":0.6644946780006649,"total_s":1.4176502440004697,"rtf":0.23627504066674496,"output_peak":0.3180811405181885,"output_rms":0.06091048388334796,"finished_at":"2026-10-13T05:02:05.534566+00:00","analysis_stages":{"external_f0_s":0.08688887700009218},"evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"5f1eca015dbe455afe18973619048925b4d05dee6463dd9499aa18ee4b26aadd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.30369970337123,"output_peak":0.31850180331292505},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.09022517555015,"spectral_convergence":0.4030890410362223,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.22028941816266,"spectral_convergence":0.1716716719432803,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.328784473393203,"spectral_convergence":0.18085900799546162,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001356083258303267,"output_power":0.0011168548216992954,"delta_db":-0.8428963087336907,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257579784608889,"output_power":0.0018536928491437736,"delta_db":-0.8560533296714055,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006985464363416991,"output_power":0.0006001553068297729,"delta_db":-0.6593163075694252,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001065894799362309,"output_power":7.674798361416693e-05,"delta_db":-1.4264736907650524,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.746820967698779e-05,"output_power":2.3746136494040116e-05,"delta_db":-0.6323739407504725,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.600051207775998e-05,"output_power":4.099147623578492e-05,"delta_db":-1.3549843933487093,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9620163284731258,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8380128261453132,"pitch":null,"output_rms_dbfs":-24.30369970337123,"output_peak":0.31850180331292505,"reference_rms_dbfs":-23.465686877225913,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.546433022368671},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9482729232845651,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06709893949195347,"rms_dbfs":-23.465686877225913,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.86252254988932,"mean":-3.375784184578464e-05,"zero_energy":false},"output":{"rms":0.06092773238859981,"rms_dbfs":-24.30369970337123,"peak":0.31850180331292505,"peak_dbfs":-9.937762088133972,"crest_factor_db":14.365937615237257,"mean":0.00033326382263761063,"zero_energy":false},"rms_difference_db":-0.8380128261453166},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":5.302479140854382e-07,"output_rms_dbfs":-62.755210312564095,"output_zero_energy":false,"output_peak":0.006266022299767476,"output_energy_fraction":1.9554016129057264e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":392,"output_estimated_voiced_frames":401,"both_estimated_voiced_frames":384,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9795918367346939,"absolute_difference_cents_median":3.4947220839679147,"absolute_difference_cents_p95":16.067595443176756,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00016450226550328756,"log_rms_mae_db":1.7809828673733918,"rms_envelope_nrmse":0.12072019608003227,"envelope_correlation":0.9909126641989904,"rms_gain_error_db":-0.8379609050264605,"reference_mean_power":0.0045028235447229685,"output_mean_power":0.0037126912782012786},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3776144024949717e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011179835350197734,"output_power":0.0009836896571930303,"reference_active":true,"delta_db":-0.5557730246193461},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011324318282145996,"output_power":0.00010425843781659314,"reference_active":true,"delta_db":-0.35900853902674623},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1979621575658254e-05,"output_power":2.1404634023240583e-05,"reference_active":true,"delta_db":-0.11512404219966653}],"active_band_count":3,"active_band_mae_db":0.3433018686152529}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014948818281718847,"log_rms_mae_db":1.6943013664732267,"rms_envelope_nrmse":0.10776585304728556,"envelope_correlation":0.9947698254752724,"rms_gain_error_db":-0.8399888998556448,"reference_mean_power":0.004509123852410744,"output_mean_power":0.003716150330192453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1897864659756212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130545568112854,"output_power":0.0008934029707383666,"reference_active":true,"delta_db":-0.545854421353065},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.375845448001967e-05,"output_power":5.092078760472468e-05,"reference_active":true,"delta_db":-0.235516621700089},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0828194582324503e-06,"output_power":1.8906785368753603e-06,"reference_active":true,"delta_db":-0.4203393222666888}],"active_band_count":3,"active_band_mae_db":0.4005701217732809}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00012259006245866322,"log_rms_mae_db":1.550794305862977,"rms_envelope_nrmse":0.12801723657085187,"envelope_correlation":0.9874638441699519,"rms_gain_error_db":-0.6011170801075937,"reference_mean_power":0.0013859767082523829,"output_mean_power":0.0012068247933947434},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.928895256091993e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005534244023853744,"output_power":0.0005085307330542197,"reference_active":true,"delta_db":-0.36741100126194204},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.272029346555577e-05,"output_power":6.192121754708797e-05,"reference_active":true,"delta_db":-0.6981613560725232},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.744857902743943e-06,"output_power":6.720247914983284e-06,"reference_active":true,"delta_db":0.6810600377881841}],"active_band_count":3,"active_band_mae_db":0.5822107983742164}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001104267603656376,"log_rms_mae_db":1.0395947096564193,"rms_envelope_nrmse":0.10902444525691671,"envelope_correlation":0.9914166010138629,"rms_gain_error_db":-0.6011784163064539,"reference_mean_power":0.0013875784012634727,"output_mean_power":0.0012082023871312878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.148847442195771e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005147563053777727,"output_power":0.0004734322491034927,"reference_active":true,"delta_db":-0.36343837373061744},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1541213631419644e-05,"output_power":2.7489056432559195e-05,"reference_active":true,"delta_db":-0.5971856708743232},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.960922015945642e-07,"output_power":6.678875071008215e-07,"reference_active":true,"delta_db":-0.7626004963205619}],"active_band_count":3,"active_band_mae_db":0.5744081803085008}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":1.6432262324689515,"rms_envelope_nrmse":0.19460451718261434,"envelope_correlation":0.9823386615258483,"rms_gain_error_db":-1.3505368676650336,"reference_mean_power":0.0001143249984562233,"output_mean_power":8.376980749446241e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.04318641429584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4281205292676005e-05,"output_power":3.0999271663417116e-05,"reference_active":true,"delta_db":-1.5486794337059928},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.388027117897534e-06,"output_power":3.0990126157285072e-06,"reference_active":true,"delta_db":-0.3872353357567437},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0970907503396972e-06,"output_power":1.078349582875893e-06,"reference_active":true,"delta_db":-0.07482978826758521}],"active_band_count":3,"active_band_mae_db":0.6702481859101073}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":1.3358151249068007,"rms_envelope_nrmse":0.1680575206069837,"envelope_correlation":0.9916857956859717,"rms_gain_error_db":-1.3487936365094906,"reference_mean_power":0.00011470526861651363,"output_mean_power":8.408218764717937e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.512755753376485e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.070516308485891e-05,"output_power":2.852424268529573e-05,"reference_active":true,"delta_db":-1.5443537629423474},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6333740990622453e-06,"output_power":1.5083526046297504e-06,"reference_active":true,"delta_db":-0.3458278698866008},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.6312732379634804e-08,"output_power":6.338897579207482e-08,"reference_active":true,"delta_db":0.5140713422912134}],"active_band_count":3,"active_band_mae_db":0.8014176583733872}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":2.0995523927310975,"rms_envelope_nrmse":0.23466586220875682,"envelope_correlation":0.9653255878750174,"rms_gain_error_db":-0.14393118140463781,"reference_mean_power":4.034148789813246e-05,"output_mean_power":3.9026427174726875e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.076485388357221e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6403498075336392e-05,"output_power":1.7136771860431347e-05,"reference_active":true,"delta_db":0.1899254306432328},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.853624356742713e-06,"output_power":6.074037787617647e-06,"reference_active":true,"delta_db":0.1605264029569567},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8998112094127464e-06,"output_power":1.6541658442100002e-06,"reference_active":true,"delta_db":-0.6013139668443428}],"active_band_count":3,"active_band_mae_db":0.3172552668148441}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":1.7573317705485911,"rms_envelope_nrmse":0.14422908250705638,"envelope_correlation":0.9865388304745605,"rms_gain_error_db":-0.1438062915310295,"reference_mean_power":4.0542574585969874e-05,"output_mean_power":3.9222086685955084e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.8623021113923146e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.77784480947121e-05,"output_power":1.8552412485044332e-05,"reference_active":true,"delta_db":0.1850654349961433},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2787493934827795e-06,"output_power":3.0697929571105895e-06,"reference_active":true,"delta_db":-0.28599137886681636},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7487144271280286e-07,"output_power":1.3711436091496118e-07,"reference_active":true,"delta_db":-1.0563594926949522}],"active_band_count":3,"active_band_mae_db":0.5091387688526373}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.5379858016967773,"predicted_mos":3.5379858016967773,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"d2e63b3abedc29cd5e0f66f339354f6f558b7ea83c24721a489b91de58de14cb","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1825480720017367},"started_at":"2026-10-13T05:06:40.090431+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026681079998525092,"native_input_samples":144000,"native_input_sha256":"d086d02f1071621e604cce910000ce5ef03b3f290ce0aa589c79a8b5ad3c3d48","output_sha256":"14fc859a17e53a741dda7b09a556badd194120298301080fab19eb00962d1a8b","output_samples":144120,"length_delta_samples":120,"analysis_s":2.103963517000011,"synthesis_s":3.250283605999357,"total_s":5.354247122999368,"rtf":0.8923745204998946,"output_peak":0.2874285578727722,"output_rms":0.0660360377798348,"finished_at":"2026-10-13T05:06:45.456986+00:00","analysis_stages":{"external_f0_s":1.3283473399997092},"evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"14fc859a17e53a741dda7b09a556badd194120298301080fab19eb00962d1a8b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.602896266288166,"output_peak":0.2869437632612829},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.821953375642161,"spectral_convergence":0.4301274696209146,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.882772065944338,"spectral_convergence":0.18479915239690023,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.715014139858636,"spectral_convergence":0.19650260892440066,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.0014222693697312776,"delta_db":0.1957798997205849,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.0021112455484883184,"delta_db":-0.2900441221761991,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0006600729605367541,"delta_db":-0.24025721090753083,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.0001023653382384904,"delta_db":-0.17153686224227835,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":2.610274971924295e-05,"delta_db":-0.2146229600239558,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":4.137167897735663e-05,"delta_db":-1.3078741602707606,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4033525358902183,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1377833227166936,"pitch":null,"output_rms_dbfs":-23.601722211038087,"output_peak":0.2869437632612829,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.139913193815045},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.939797605088871,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06711244415402366,"rms_dbfs":-23.46393888832139,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.860774560984794,"mean":-7.2549184163411456e-06,"zero_energy":false},"output":{"rms":0.06605624607629047,"rms_dbfs":-23.601722211038087,"peak":0.2869437632612829,"peak_dbfs":-10.84406420458512,"crest_factor_db":12.757658006452965,"mean":4.0526058383167236e-05,"zero_energy":false},"rms_difference_db":-0.1377833227166967},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":3.862154726124528e-06,"output_rms_dbfs":-54.13170331448377,"output_zero_energy":false,"output_peak":0.03462617664521535,"output_energy_fraction":0.00012096632634473563,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":393,"output_estimated_voiced_frames":401,"both_estimated_voiced_frames":381,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9694656488549618,"absolute_difference_cents_median":6.387952486487254,"absolute_difference_cents_p95":30.357397156259328,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016450226550328756,"log_rms_mae_db":1.847000094696757,"rms_envelope_nrmse":0.12041153709443166,"envelope_correlation":0.9800632487713212,"rms_gain_error_db":-0.13708852108236314,"reference_mean_power":0.004505860165227205,"output_mean_power":0.004365850475258447},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3773993983732174e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011178733037365022,"output_power":0.0011327159086598055,"reference_active":true,"delta_db":0.05728415329511482},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011348254393479664,"output_power":0.0001126531508385455,"reference_active":true,"delta_db":-0.03185719623160916},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1976543085088728e-05,"output_power":2.1195487898911704e-05,"reference_active":true,"delta_db":-0.1571596080180844}],"active_band_count":3,"active_band_mae_db":0.0821003191816028}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014948818281718847,"log_rms_mae_db":1.390756944872149,"rms_envelope_nrmse":0.08144753398306123,"envelope_correlation":0.9899290344666184,"rms_gain_error_db":-0.13602086166371982,"reference_mean_power":0.004512140106290299,"output_mean_power":0.004373010199459808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1895509906107845e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130343468129068,"output_power":0.0010287930692938034,"reference_active":true,"delta_db":0.06703859548881257},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.38575581348341e-05,"output_power":5.298727942916532e-05,"reference_active":true,"delta_db":-0.07075037167763419},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0969765535512897e-06,"output_power":2.0581475154986715e-06,"reference_active":true,"delta_db":-0.08117075484443567}],"active_band_count":3,"active_band_mae_db":0.07298657400362747}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012259006245866322,"log_rms_mae_db":1.9145180161235509,"rms_envelope_nrmse":0.14453527575367406,"envelope_correlation":0.9794062255107747,"rms_gain_error_db":-0.10811097609290916,"reference_mean_power":0.001387125934818555,"output_mean_power":0.0013530218002269845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.931300849598276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005535817649712056,"output_power":0.0005475878776873446,"reference_active":true,"delta_db":-0.047279506823575956},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.308179193682493e-05,"output_power":7.292981907608152e-05,"reference_active":true,"delta_db":-0.009040513722914659},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.782827787780852e-06,"output_power":8.706140379654433e-06,"reference_active":true,"delta_db":1.7768540629823235}],"active_band_count":3,"active_band_mae_db":0.6110580278429381}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001104267603656376,"log_rms_mae_db":1.321558309774231,"rms_envelope_nrmse":0.09809840123618807,"envelope_correlation":0.9896242645542125,"rms_gain_error_db":-0.1047139596937638,"reference_mean_power":0.0013888741987160204,"output_mean_power":0.0013557871521136485},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.150356442217087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005149581742784205,"output_power":0.0005101737785265464,"reference_active":true,"delta_db":-0.04053823032564213},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177504958330739e-05,"output_power":3.1564865594684815e-05,"reference_active":true,"delta_db":-0.028822925081876078},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.976808620230845e-07,"output_power":9.292191619167048e-07,"reference_active":true,"delta_db":0.6628898446191367}],"active_band_count":3,"active_band_mae_db":0.24408366667555162}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":1.6802953119852844,"rms_envelope_nrmse":0.15076624638004882,"envelope_correlation":0.979888796890768,"rms_gain_error_db":-0.10591298076478432,"reference_mean_power":0.00011418247196706925,"output_mean_power":0.00011143154263804918},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.040353626229237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.425624857591464e-05,"output_power":4.440522880198943e-05,"reference_active":true,"delta_db":0.014595143890043802},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4228259000934756e-06,"output_power":3.302054017465551e-06,"reference_active":true,"delta_db":-0.1560063608311365},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0991841496467986e-06,"output_power":9.777738641013666e-07,"reference_active":true,"delta_db":-0.5083203279764896}],"active_band_count":3,"active_band_mae_db":0.22630727756588997}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":1.0787486290620318,"rms_envelope_nrmse":0.0882733388674652,"envelope_correlation":0.9926981862354544,"rms_gain_error_db":-0.10443616268325587,"reference_mean_power":0.0001146281153210772,"output_mean_power":0.00011190449602269313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.50871602576763e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.069122651510938e-05,"output_power":4.082991259820596e-05,"reference_active":true,"delta_db":0.014776697030210744},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.65433446148589e-06,"output_power":1.6040588915670715e-06,"reference_active":true,"delta_db":-0.1340300767063573},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.669640028113729e-08,"output_power":5.04665664439263e-08,"reference_active":true,"delta_db":-0.5055172794855958}],"active_band_count":3,"active_band_mae_db":0.21810801774072128}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":1.9327382925566554,"rms_envelope_nrmse":0.3002155630511885,"envelope_correlation":0.940601660413162,"rms_gain_error_db":-0.7301687913592875,"reference_mean_power":4.027652017807585e-05,"output_mean_power":3.4043567312224656e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067374336554811e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6348738836101644e-05,"output_power":1.3887482684096322e-05,"reference_active":true,"delta_db":-0.7086072578284918},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.842264617940261e-06,"output_power":4.3976291138676075e-06,"reference_active":true,"delta_db":-1.233626247478404},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8971365027113356e-06,"output_power":1.1989009534027204e-06,"reference_active":true,"delta_db":-1.9931527479567572}],"active_band_count":3,"active_band_mae_db":1.311795417754551}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":1.2068047236562152,"rms_envelope_nrmse":0.18813195557827714,"envelope_correlation":0.978151250159924,"rms_gain_error_db":-0.7302379648281432,"reference_mean_power":4.047626840683264e-05,"output_mean_power":3.4211858778994364e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8536334265621368e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7722714720556834e-05,"output_power":1.4560681799348171e-05,"reference_active":true,"delta_db":-0.8534853561967044},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.276595703569918e-06,"output_power":2.313414971809794e-06,"reference_active":true,"delta_db":-1.5116931555763522},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.751293843816523e-07,"output_power":1.0016862716519203e-07,"reference_active":true,"delta_db":-2.4262729932984266}],"active_band_count":3,"active_band_mae_db":1.5971505016904946}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.535681962966919,"predicted_mos":3.535681962966919,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"5ef98272f81ab999c21ebf817ac0138ddbb8dd3d7cbd51c97ec929176e511e4d","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9763352810005017},"started_at":"2026-10-13T05:10:11.737113+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028950440009793965,"native_input_samples":144000,"native_input_sha256":"d086d02f1071621e604cce910000ce5ef03b3f290ce0aa589c79a8b5ad3c3d48","output_sha256":"674547c48dbede45bfb373bdfbe84ed1ce2dd529aa319e7c6d774473d1940060","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9619697159996576,"synthesis_s":0.17166236699995352,"total_s":1.1336320829996112,"rtf":0.1889386804999352,"output_peak":0.29594042897224426,"output_rms":0.06643276876348081,"finished_at":"2026-10-13T05:10:12.882354+00:00","analysis_stages":{"external_f0_s":0.16185597900039284},"evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"674547c48dbede45bfb373bdfbe84ed1ce2dd529aa319e7c6d774473d1940060","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.55077970103281,"output_peak":0.2951941800153473},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.090782224802832,"spectral_convergence":0.0979118457864788,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.8674422591998114,"spectral_convergence":0.08982632297174077,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.210431900826558,"spectral_convergence":0.09695414774872392,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013595771819321111,"output_power":0.0013559945122606316,"delta_db":-0.011459351818036223,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002257060285350754,"output_power":0.0022193122683333774,"delta_db":-0.07324745161253467,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006976179772550826,"output_power":0.0006787724312783438,"delta_db":-0.11893468722545061,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001064894568705692,"output_power":0.00010276120948429352,"delta_db":-0.15477404797487032,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7425121182630722e-05,"output_power":2.35743331815935e-05,"delta_db":-0.6570913803125717,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.591013773579702e-05,"output_power":4.1743766204780304e-05,"delta_db":-1.268989327531539,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3807493744125004,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07971804528510035,"pitch":null,"output_rms_dbfs":-23.543656933606492,"output_peak":0.2951941800153473,"reference_rms_dbfs":-23.46393888832139,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.056218794943067},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.984203980739213,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06711244415402366,"rms_dbfs":-23.46393888832139,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.860774560984794,"mean":-7.2549184163411456e-06,"zero_energy":false},"output":{"rms":0.06649931218248066,"rms_dbfs":-23.543656933606492,"peak":0.2951941800153473,"peak_dbfs":-10.597844184414607,"crest_factor_db":12.945812749191886,"mean":3.535442304033889e-05,"zero_energy":false},"rms_difference_db":-0.07971804528510162},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":6.27267350210668e-07,"output_rms_dbfs":-62.0254731726478,"output_zero_energy":false,"output_peak":0.00775232093152616,"output_energy_fraction":1.9385677256933946e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":393,"output_estimated_voiced_frames":395,"both_estimated_voiced_frames":386,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9821882951653944,"absolute_difference_cents_median":0.6476262111728635,"absolute_difference_cents_p95":6.50544002742233,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016450226550328756,"log_rms_mae_db":0.4943369339744584,"rms_envelope_nrmse":0.03807918987505849,"envelope_correlation":0.9981118794376298,"rms_gain_error_db":-0.0791727872146418,"reference_mean_power":0.004505860165227205,"output_mean_power":0.004424461608008122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3773993983732174e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011178733037365022,"output_power":0.0011100836283103863,"reference_active":true,"delta_db":-0.03036887104812737},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011348254393479664,"output_power":0.00010429063217157478,"reference_active":true,"delta_db":-0.3668376279955527},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1976543085088728e-05,"output_power":1.7028993389103968e-05,"reference_active":true,"delta_db":-1.1077040196185186}],"active_band_count":3,"active_band_mae_db":0.5016368395540662}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014948818281718847,"log_rms_mae_db":0.31122395583766155,"rms_envelope_nrmse":0.026457653109865895,"envelope_correlation":0.9990582545337746,"rms_gain_error_db":-0.07818163663280353,"reference_mean_power":0.004512140106290299,"output_mean_power":0.004431639378794944},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1895509906107845e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130343468129068,"output_power":0.0010097866574716196,"reference_active":true,"delta_db":-0.013945423433757473},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.38575581348341e-05,"output_power":4.950613807779101e-05,"reference_active":true,"delta_db":-0.3658761011136693},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0969765535512897e-06,"output_power":1.5305509518687118e-06,"reference_active":true,"delta_db":-1.3674578282565484}],"active_band_count":3,"active_band_mae_db":0.5824264509346584}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012259006245866322,"log_rms_mae_db":0.7538101653807602,"rms_envelope_nrmse":0.03427594504974425,"envelope_correlation":0.9989875461266754,"rms_gain_error_db":-0.10206900756607969,"reference_mean_power":0.001387125934818555,"output_mean_power":0.0013549054539787266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.931300849598276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005535817649712056,"output_power":0.0005353787231465766,"reference_active":true,"delta_db":-0.14520668389737526},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.308179193682493e-05,"output_power":6.691362604662508e-05,"reference_active":true,"delta_db":-0.38294622624006985},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.782827787780852e-06,"output_power":4.138555509693643e-06,"reference_active":true,"delta_db":-1.4529147454899636}],"active_band_count":3,"active_band_mae_db":0.6603558852091362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001104267603656376,"log_rms_mae_db":0.42628156268880535,"rms_envelope_nrmse":0.020426851379298335,"envelope_correlation":0.9996946689844085,"rms_gain_error_db":-0.10047630258950661,"reference_mean_power":0.0013888741987160204,"output_mean_power":0.001357110716019123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.150356442217087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005149581742784205,"output_power":0.0004971726773050917,"reference_active":true,"delta_db":-0.1526470291150481},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.177504958330739e-05,"output_power":2.907068331612554e-05,"reference_active":true,"delta_db":-0.3863099679313574},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.976808620230845e-07,"output_power":5.46388578220796e-07,"reference_active":true,"delta_db":-1.6432756035439302}],"active_band_count":3,"active_band_mae_db":0.7274108668634452}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":0.7607874097841596,"rms_envelope_nrmse":0.06952846845606847,"envelope_correlation":0.9960644823530926,"rms_gain_error_db":-0.18259111863341115,"reference_mean_power":0.00011418247196706925,"output_mean_power":0.00010948139665395102},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.040353626229237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.425624857591464e-05,"output_power":4.282896555708924e-05,"reference_active":true,"delta_db":-0.14237012695692874},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4228259000934756e-06,"output_power":3.233017828555225e-06,"reference_active":true,"delta_db":-0.24776710035847505},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0991841496467986e-06,"output_power":7.403773322453177e-07,"reference_active":true,"delta_db":-1.7161734347562572}],"active_band_count":3,"active_band_mae_db":0.702103554023887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":0.4786392056720164,"rms_envelope_nrmse":0.04224311565611222,"envelope_correlation":0.9987369618204546,"rms_gain_error_db":-0.18232523826103045,"reference_mean_power":0.0001146281153210772,"output_mean_power":0.00010991542110706767},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.50871602576763e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.069122651510938e-05,"output_power":3.941900152225652e-05,"reference_active":true,"delta_db":-0.1379516109172477},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.65433446148589e-06,"output_power":1.5742899097042084e-06,"reference_active":true,"delta_db":-0.2153860475925715},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.669640028113729e-08,"output_power":3.1628890432253476e-08,"reference_active":true,"delta_db":-2.534715291463763}],"active_band_count":3,"active_band_mae_db":0.9626843166578608}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":0.9266677039219099,"rms_envelope_nrmse":0.18363233031025236,"envelope_correlation":0.9833683062413461,"rms_gain_error_db":-0.9342493447810076,"reference_mean_power":4.027652017807585e-05,"output_mean_power":3.248082164296469e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067374336554811e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6348738836101644e-05,"output_power":1.318048151058097e-05,"reference_active":true,"delta_db":-0.9355297998842614},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.842264617940261e-06,"output_power":3.855517031023068e-06,"reference_active":true,"delta_db":-1.8049859801931998},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8971365027113356e-06,"output_power":8.614605461658618e-07,"reference_active":true,"delta_db":-3.4286318827098263}],"active_band_count":3,"active_band_mae_db":2.056382554262429}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":0.6329338871813956,"rms_envelope_nrmse":0.13702732921457278,"envelope_correlation":0.9938286918477193,"rms_gain_error_db":-0.9342810616541868,"reference_mean_power":4.047626840683264e-05,"output_mean_power":3.26416693334067e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8536334265621368e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7722714720556834e-05,"output_power":1.3566046402122542e-05,"reference_active":true,"delta_db":-1.1607694855684836},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.276595703569918e-06,"output_power":1.990619727631091e-06,"reference_active":true,"delta_db":-2.1643455371006275},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.751293843816523e-07,"output_power":6.915368777598026e-08,"reference_active":true,"delta_db":-4.035436762425368}],"active_band_count":3,"active_band_mae_db":2.4535172616981598}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0000","method":"magphase","status":"measured","mos":{"status":"measured","value":3.3882195949554443,"predicted_mos":3.3882195949554443,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287425,"input_sample_rate":48000,"input_seconds":5.988020833333334,"input_float32_sha256":"942ae6a3c97c4795bbd923bd2c0ad6c1e78cdc28a8a718e5581d47992e08c33f","internal_samples":95809,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1619210129993007},"started_at":"2026-10-13T05:13:12.202792+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009922777999236132,"native_input_samples":288000,"native_input_sha256":"bfe64c08bf21ee523c32627158a797634c36c7ce312d7f867e2e80e0d87ecd27","output_sha256":"b7511b9ab502d56e04b1e4c7142bbf972e9aa0048e9f6ca9af200e5882feee40","output_samples":287425,"length_delta_samples":-575,"analysis_s":3.6296465859995806,"synthesis_s":1.1227579009992041,"total_s":4.752404486998785,"rtf":0.7920674144997975,"output_peak":0.30964336891159916,"output_rms":0.0641739338366283,"finished_at":"2026-10-13T05:13:16.976730+00:00","evaluation_input":{"reference":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"b7511b9ab502d56e04b1e4c7142bbf972e9aa0048e9f6ca9af200e5882feee40","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95809,"length_delta_samples":-191,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.85174864793227,"output_peak":0.3087350704064881},"common_prefix":{"samples":95809,"sample_rate":16000,"reference_coverage":0.9980104166666667,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95809,"output_samples":95809,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.812352636636375,"spectral_convergence":0.4111155263770454,"spectral_convergence_status":"ok","amplitude_floor":1.1304927744850054e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.910856899239796,"spectral_convergence":0.15448845278716283,"spectral_convergence_status":"ok","amplitude_floor":1.0019114511304393e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.071055038088429,"spectral_convergence":0.1627887483172932,"spectral_convergence_status":"ok","amplitude_floor":8.812050137930174e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013578518772931196,"output_power":0.001205096859724945,"delta_db":-0.5183044229970041,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002254894056128256,"output_power":0.0020762874669555216,"delta_db":-0.3583865936805781,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006978120996953585,"output_power":0.0006491407849128322,"delta_db":-0.31399599308198534,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010660437623914517,"output_power":0.00010328694510122962,"delta_db":-0.13729600805403372,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.7476252751763213e-05,"output_power":2.8214641382663785e-05,"delta_db":0.11517031854403534,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.601778975693188e-05,"output_power":5.7289553052604404e-05,"delta_db":0.09749464829540291,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2567746641088399,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.38450728041900817,"pitch":null,"output_rms_dbfs":-23.85174864793227,"output_peak":0.3087350704064881,"reference_rms_dbfs":-23.46724136751326,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.2647548579882},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95809,"output_samples":95809,"intelligibility":{"estoi":{"value":0.9685668320316467,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06708693204987523,"rms_dbfs":-23.46724136751326,"peak":0.295013427734375,"peak_dbfs":-10.603164327336597,"crest_factor_db":12.864077040176664,"mean":-6.617360431137732e-06,"zero_energy":false},"output":{"rms":0.06418189971580616,"rms_dbfs":-23.85174864793227,"peak":0.3087350704064881,"peak_dbfs":-10.208280690635661,"crest_factor_db":13.64346795729661,"mean":0.0017004918597273903,"zero_energy":false},"rms_difference_db":-0.38450728041900817},"full_scale":{"reference":{"sample_count":95809,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95809,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":129,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.432272993610241e-06,"threshold_dbfs":-56.13987682292056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":82,"quiet_samples":13120,"reference_mean_power":4.669256712787035e-07,"output_mean_power":4.850787461160826e-07,"output_rms_dbfs":-63.14187753711843,"output_zero_energy":false,"output_peak":0.012884588292294196,"output_energy_fraction":1.6125553811484013e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":391,"output_estimated_voiced_frames":385,"both_estimated_voiced_frames":378,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9667519181585678,"absolute_difference_cents_median":3.253677791650293,"absolute_difference_cents_p95":21.739391083904806,"difference_over_600_cents_frames":0,"uncovered_tail_samples":129,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95809,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1196,"rms_floor":0.00016450226550328756,"log_rms_mae_db":1.0832499514757596,"rms_envelope_nrmse":0.09197872771442382,"envelope_correlation":0.9905960818355978,"rms_gain_error_db":-0.38233939228595204,"reference_mean_power":0.004500246502353536,"output_mean_power":0.004120997620971699},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":1.3777774995976345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011180350054252385,"output_power":0.0010339637294608088,"reference_active":true,"delta_db":-0.33950097069820745},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011314173514391694,"output_power":0.0001023184593734288,"reference_active":true,"delta_db":-0.43668842530923785},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1953822963296324e-05,"output_power":2.4234165249772567e-05,"reference_active":true,"delta_db":0.4291790725163558}],"active_band_count":3,"active_band_mae_db":0.4017894895079337}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1190,"rms_floor":0.00014948818281718847,"log_rms_mae_db":0.7190742257378521,"rms_envelope_nrmse":0.06496901686636697,"envelope_correlation":0.9962286948930146,"rms_gain_error_db":-0.38436008563764196,"reference_mean_power":0.00450705141960834,"output_mean_power":0.00412530918997406},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":1.1899338134222094e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010130567947743742,"output_power":0.0009354382785233786,"reference_active":true,"delta_db":-0.34618656349384425},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3638982492590076e-05,"output_power":4.76059819882962e-05,"reference_active":true,"delta_db":-0.5181900292900271},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.15360265672152e-06,"output_power":2.286895392505376e-06,"reference_active":true,"delta_db":0.2608072120365601}],"active_band_count":3,"active_band_mae_db":0.37506126827347713}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1196,"rms_floor":0.00012259006245866322,"log_rms_mae_db":1.2297708744931115,"rms_envelope_nrmse":0.09462670886294093,"envelope_correlation":0.9916525064365029,"rms_gain_error_db":-0.22321902562260287,"reference_mean_power":0.0013854629981263748,"output_mean_power":0.001316051937571694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":7.927604353369866e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005533527789607117,"output_power":0.0005303544340972875,"reference_active":true,"delta_db":-0.18435891154521253},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.253319428385659e-05,"output_power":6.357124331284029e-05,"reference_active":true,"delta_db":-0.5727609839221812},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.719330070768544e-06,"output_power":6.669111988406081e-06,"reference_active":true,"delta_db":0.6672284916751088}],"active_band_count":3,"active_band_mae_db":0.47478279571416754}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1190,"rms_floor":0.0001104267603656376,"log_rms_mae_db":0.8544416130882685,"rms_envelope_nrmse":0.064872335550746,"envelope_correlation":0.9960665933237364,"rms_gain_error_db":-0.22474651053410666,"reference_mean_power":0.0013867257878572212,"output_mean_power":0.0013167882446413221},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":7.148265281011132e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000514657435599618,"output_power":0.0004917996049526766,"reference_active":true,"delta_db":-0.19730076151955944},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.142300161216597e-05,"output_power":2.7304592942611827e-05,"reference_active":true,"delta_db":-0.6101196124804155},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.969078142654669e-07,"output_power":9.80015252830534e-07,"reference_active":true,"delta_db":0.8982474962621261}],"active_band_count":3,"active_band_mae_db":0.5685559567540337}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1196,"rms_floor":4.1111372612584816e-05,"log_rms_mae_db":1.1262518886658857,"rms_envelope_nrmse":0.09985012545751541,"envelope_correlation":0.9913179881680724,"rms_gain_error_db":-0.08580987450628409,"reference_mean_power":0.00011439218853601968,"output_mean_power":0.00011215415858041175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":8.044671880846613e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.429374663327891e-05,"output_power":4.439569373650573e-05,"reference_active":true,"delta_db":0.009984296595685033},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3708786931410055e-06,"output_power":3.1399981136862464e-06,"reference_active":true,"delta_db":-0.3081373679295071},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0939407564525515e-06,"output_power":1.0658236047694535e-06,"reference_active":true,"delta_db":-0.11308468625101606}],"active_band_count":3,"active_band_mae_db":0.14373545025873605}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1190,"rms_floor":3.646632245602786e-05,"log_rms_mae_db":0.7333872369264753,"rms_envelope_nrmse":0.055246859357011915,"envelope_correlation":0.997242406694627,"rms_gain_error_db":-0.08617244170846,"reference_mean_power":0.00011473704382401828,"output_mean_power":0.00011248287600206906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":7.514853088799144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.071202600011086e-05,"output_power":4.0858584579445466e-05,"reference_active":true,"delta_db":0.01560602533654008},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6215770186649155e-06,"output_power":1.4134157146391987e-06,"reference_active":true,"delta_db":-0.596676652359762},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.783479181767893e-08,"output_power":6.336148156193883e-08,"reference_active":true,"delta_db":0.39636147040922964}],"active_band_count":3,"active_band_mae_db":0.3362147160351772}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1196,"rms_floor":4.742254313282892e-05,"log_rms_mae_db":1.3241018710618486,"rms_envelope_nrmse":0.22907456091546977,"envelope_correlation":0.967441199680672,"rms_gain_error_db":0.2632798661354455,"reference_mean_power":4.037440051318317e-05,"output_mean_power":4.289770650001702e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":3.0810487422157395e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6431143555594897e-05,"output_power":1.8235104675979694e-05,"reference_active":true,"delta_db":0.4524047065306387},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.858984807350286e-06,"output_power":5.3570084893936775e-06,"reference_active":true,"delta_db":-0.38900037249583197},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8923155054150606e-06,"output_power":1.6145988063450464e-06,"reference_active":true,"delta_db":-0.6892892086328629}],"active_band_count":3,"active_band_mae_db":0.5102314292197779}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":49,"metrics":{"frame_count":1190,"rms_floor":3.228791607619837e-05,"log_rms_mae_db":0.8893365666114201,"rms_envelope_nrmse":0.13488275768077104,"envelope_correlation":0.9883751647576673,"rms_gain_error_db":0.26321026386700963,"reference_mean_power":4.057573974367643e-05,"output_mean_power":4.3110938039817574e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":2.866652297103511e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7806262073815962e-05,"output_power":1.9145633074442502e-05,"reference_active":true,"delta_db":0.31496970350661374},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2784876397632586e-06,"output_power":2.7982453368543745e-06,"reference_active":true,"delta_db":-0.6878776196901486},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7640338839512634e-07,"output_power":1.2465361220235455e-07,"reference_active":true,"delta_db":-1.508020549272289}],"active_band_count":3,"active_band_mae_db":0.8369559574896837}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"hnm","status":"measured","mos":{"status":"measured","value":2.58378267288208,"predicted_mos":2.58378267288208,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7ea9253c7a0376695cbb476b1037f65d2a9ad594f61c937e626171a77387599e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9913717430008546},"started_at":"2026-10-13T04:38:27.847692+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.9409000742598437e-05,"native_input_samples":96000,"native_input_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","output_sha256":"887538e821592d38fe81ef53e549f69918c66067b76153441a07d2f7b4e44d34","output_samples":96000,"length_delta_samples":0,"analysis_s":1.13935986100023,"synthesis_s":0.41140797800017026,"total_s":1.5507678390004003,"rtf":0.2584613065000667,"output_peak":0.45279706291197,"output_rms":0.0743594473042181,"finished_at":"2026-10-13T04:38:29.405065+00:00","evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"887538e821592d38fe81ef53e549f69918c66067b76153441a07d2f7b4e44d34","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.936623458456343,"spectral_convergence":0.15039059110573466,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.774055781418571,"spectral_convergence":0.15791252767868297,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.032716935581264,"spectral_convergence":0.17097687937787726,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0014695066479431227,"delta_db":0.01478202233865017,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.0024594510645381167,"delta_db":-0.0311479773977141,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.001292792055741144,"delta_db":-0.16471862421358563,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.00023502000918905572,"delta_db":-0.1395350572468195,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":4.1766470171550517e-05,"delta_db":-0.26801638518865195,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":3.079115580579987e-05,"delta_db":-0.487229640219747,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18423828443419474,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05980750032040098,"pitch":null,"output_rms_dbfs":-22.57327693787914,"output_peak":0.45279706291197,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.914465391818726},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.936623458456343,"spectral_convergence":0.15039059110573466,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.774055781418571,"spectral_convergence":0.15791252767868297,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.032716935581264,"spectral_convergence":0.17097687937787726,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0014695066479431227,"delta_db":0.01478202233865017,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.0024594510645381167,"delta_db":-0.0311479773977141,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.001292792055741144,"delta_db":-0.16471862421358563,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.00023502000918905572,"delta_db":-0.1395350572468195,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":4.1766470171550517e-05,"delta_db":-0.26801638518865195,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":3.079115580579987e-05,"delta_db":-0.487229640219747,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18423828443419474,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05980750032040098,"pitch":null,"output_rms_dbfs":-22.57327693787914,"output_peak":0.45279706291197,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.914465391818726},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9473974299255743,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07487322298437474,"rms_dbfs":-22.513469437558737,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.708326506096444,"mean":-3.738085428873698e-05,"zero_energy":false},"output":{"rms":0.0743594473042181,"rms_dbfs":-22.57327693787914,"peak":0.45279706291197,"peak_dbfs":-6.881927977617487,"crest_factor_db":15.691348960261653,"mean":6.0072230817410036e-05,"zero_energy":false},"rms_difference_db":-0.05980750032040305},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":6.710066923371249e-07,"output_rms_dbfs":-61.732731483398155,"output_zero_energy":false,"output_peak":0.014772099364295068,"output_energy_fraction":1.5169265703227332e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":349,"both_estimated_voiced_frames":339,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9390581717451524,"absolute_difference_cents_median":1.215147376397311,"absolute_difference_cents_p95":9.49732754765685,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002110151907366841,"log_rms_mae_db":0.5358404758164732,"rms_envelope_nrmse":0.04820564337832947,"envelope_correlation":0.997005216301465,"rms_gain_error_db":-0.059133547719925766,"reference_mean_power":0.005604891233518836,"output_mean_power":0.005529092230972952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2250047113785757e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011643562088722453,"output_power":0.001140814384244564,"reference_active":true,"delta_db":-0.08870874920488632},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011188245530999946,"output_power":0.00010556296339076777,"reference_active":true,"delta_db":-0.25250415242705987},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8450792038001933e-05,"output_power":2.004982848901758e-05,"reference_active":true,"delta_db":-1.5198369925433357}],"active_band_count":3,"active_band_mae_db":0.620349964725094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020713167740975364,"log_rms_mae_db":0.29323441360082864,"rms_envelope_nrmse":0.027525941930530977,"envelope_correlation":0.9989732267448377,"rms_gain_error_db":-0.05714752773623752,"reference_mean_power":0.005603505754865554,"output_mean_power":0.005530253885037445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0483171696395765e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010458478077031374,"output_power":0.001027632066301911,"reference_active":true,"delta_db":-0.07630842817882427},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.131979407845486e-05,"output_power":5.77419920788164e-05,"reference_active":true,"delta_db":-0.26108924942345674},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.629293286100356e-06,"output_power":1.2422616721676588e-06,"reference_active":true,"delta_db":-1.1778618187514605}],"active_band_count":3,"active_band_mae_db":0.5050864987845806}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015831513319989338,"log_rms_mae_db":1.3491199451748896,"rms_envelope_nrmse":0.0752652068879019,"envelope_correlation":0.9950557879975521,"rms_gain_error_db":-0.09297997916727559,"reference_mean_power":0.0022309508329157127,"output_mean_power":0.002183695108216857},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0339679622838414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046453953265101344,"output_power":0.00043773627966564046,"reference_active":true,"delta_db":-0.2581013616402746},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.019812862224435e-05,"output_power":6.216477652847875e-05,"reference_active":true,"delta_db":-0.5278115786362504},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5452757374862184e-05,"output_power":1.0422627978277741e-05,"reference_active":true,"delta_db":-1.7102874927271114}],"active_band_count":3,"active_band_mae_db":0.8320668110012122}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014375880677218868,"log_rms_mae_db":0.9170701463296628,"rms_envelope_nrmse":0.04355906742981114,"envelope_correlation":0.9982969031359088,"rms_gain_error_db":-0.08892094921437488,"reference_mean_power":0.0022214967634666704,"output_mean_power":0.00217647453348286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.581275628581243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004040718027646314,"output_power":0.0003839654188488445,"reference_active":true,"delta_db":-0.22166433061885069},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.035103681693411e-05,"output_power":3.493952266988452e-05,"reference_active":true,"delta_db":-0.6253773084053413},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1841518228550596e-06,"output_power":7.908274480691425e-07,"reference_active":true,"delta_db":-1.753256534481871}],"active_band_count":3,"active_band_mae_db":0.8667660578353543}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":1.4167009163763504,"rms_envelope_nrmse":0.16917376779099091,"envelope_correlation":0.9789655013933485,"rms_gain_error_db":-0.17052384891514685,"reference_mean_power":0.0002507880581966276,"output_mean_power":0.00024113178880165283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4212041991971411e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.17102478326245e-05,"output_power":6.54951792461735e-05,"reference_active":true,"delta_db":-0.3937188839044889},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.275733290040069e-06,"output_power":7.220227438709586e-06,"reference_active":true,"delta_db":-0.03325892392943},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.07244494044588e-06,"output_power":1.2283073327560869e-06,"reference_active":true,"delta_db":0.5893203985236629}],"active_band_count":3,"active_band_mae_db":0.33876606878586063}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":0.9940548983635918,"rms_envelope_nrmse":0.1428784640433703,"envelope_correlation":0.9843980893068323,"rms_gain_error_db":-0.17186423114116586,"reference_mean_power":0.0002518169764916093,"output_mean_power":0.00024204637456013155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3611130645010087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.48897476225386e-05,"output_power":5.995694261059155e-05,"reference_active":true,"delta_db":-0.3433660632279999},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.71133538995873e-06,"output_power":3.6393163052584473e-06,"reference_active":true,"delta_db":-0.08510399689015664},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.48681833738951e-08,"output_power":9.439527436874613e-08,"reference_active":true,"delta_db":0.4620534703586735}],"active_band_count":3,"active_band_mae_db":0.29684117682561}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":1.510661059014658,"rms_envelope_nrmse":0.1824401238169322,"envelope_correlation":0.9745197695188459,"rms_gain_error_db":-0.3647470627500047,"reference_mean_power":2.817379488191862e-05,"output_mean_power":2.5904227296498058e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.938019558986415e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.123435668533647e-06,"output_power":5.652976310668156e-06,"reference_active":true,"delta_db":-0.3471799408869886},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7222454435376484e-06,"output_power":2.135113631038224e-06,"reference_active":true,"delta_db":-1.0550628637498285},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.909737577930933e-07,"output_power":6.550805279900577e-07,"reference_active":true,"delta_db":-0.2316686340481055}],"active_band_count":3,"active_band_mae_db":0.5446371462283076}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":0.9340813842613581,"rms_envelope_nrmse":0.10833450755966378,"envelope_correlation":0.9911549162435264,"rms_gain_error_db":-0.36453993974103405,"reference_mean_power":2.8310170166312666e-05,"output_mean_power":2.6030858166353385e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8087824895787787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.951898475340427e-06,"output_power":5.548966835139021e-06,"reference_active":true,"delta_db":-0.30443385465205786},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2501096733423978e-06,"output_power":9.72774841069248e-07,"reference_active":true,"delta_db":-1.08935785867321},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.567916579161295e-08,"output_power":4.4507607872873356e-08,"reference_active":true,"delta_db":-1.6899337494644073}],"active_band_count":3,"active_band_mae_db":1.0279084875965585}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0972702503204346,"predicted_mos":3.0972702503204346,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a90e27c6e61522b96df21c2a167967487b4a67fb2c9025e7f8f74224869a945e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9905984079996415},"started_at":"2026-10-13T04:41:25.545555+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.937499994004611e-05,"native_input_samples":96000,"native_input_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","output_sha256":"c2accb9cd92c153756a1486ecbeb4440747a89f5d8567483a996aafbaa053de7","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3795726620001005,"synthesis_s":0.092654594999658,"total_s":1.4722272569997585,"rtf":0.24537120949995975,"output_peak":0.5957602507576347,"output_rms":0.08637411863844047,"finished_at":"2026-10-13T04:41:27.024612+00:00","evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"c2accb9cd92c153756a1486ecbeb4440747a89f5d8567483a996aafbaa053de7","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.024826611328832,"spectral_convergence":0.5558402800540513,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.583002400205809,"spectral_convergence":0.24148659924415752,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.124837469099152,"spectral_convergence":0.26718408885702954,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0016640352762926972,"delta_db":0.5546917587129865,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.003584930142346771,"delta_db":1.605277136120677,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.0017798999317306765,"delta_db":1.223950492963248,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.00034591830011868475,"delta_db":1.539151934699905,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":5.207097124117446e-05,"delta_db":0.689662667135335,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":3.3633748837382654e-05,"delta_db":-0.10373678587138646,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9527451292505896,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.241142014832834,"pitch":null,"output_rms_dbfs":-21.2723274227259,"output_peak":0.5957602507576347,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.244222160211264},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.024826611328832,"spectral_convergence":0.5558402800540513,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.583002400205809,"spectral_convergence":0.24148659924415752,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.124837469099152,"spectral_convergence":0.26718408885702954,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0016640352762926972,"delta_db":0.5546917587129865,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.003584930142346771,"delta_db":1.605277136120677,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.0017798999317306765,"delta_db":1.223950492963248,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.00034591830011868475,"delta_db":1.539151934699905,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":5.207097124117446e-05,"delta_db":0.689662667135335,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":3.3633748837382654e-05,"delta_db":-0.10373678587138646,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9527451292505896,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.241142014832834,"pitch":null,"output_rms_dbfs":-21.2723274227259,"output_peak":0.5957602507576347,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.244222160211264},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9585595075154584,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07487322298437474,"rms_dbfs":-22.513469437558737,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.708326506096444,"mean":-3.738085428873698e-05,"zero_energy":false},"output":{"rms":0.08637411863844047,"rms_dbfs":-21.2723274227259,"peak":0.5957602507576347,"peak_dbfs":-4.498569527370748,"crest_factor_db":16.773757895355153,"mean":1.298649678817718e-05,"zero_energy":false},"rms_difference_db":1.2411420148328354},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":1.2674549656063519e-06,"output_rms_dbfs":-58.97067462792353,"output_zero_energy":false,"output_peak":0.014352256585700126,"output_energy_fraction":2.1236125951999154e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":362,"both_estimated_voiced_frames":342,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9473684210526315,"absolute_difference_cents_median":6.034197644845701,"absolute_difference_cents_p95":28.45173844032323,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002110151907366841,"log_rms_mae_db":1.5980982243258612,"rms_envelope_nrmse":0.2057057499888373,"envelope_correlation":0.9801557723903391,"rms_gain_error_db":1.2441804702265293,"reference_mean_power":0.005604891233518836,"output_mean_power":0.007464233835144127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2250047113785757e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011643562088722453,"output_power":0.001693605998106584,"reference_active":true,"delta_db":1.6272651974570904},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011188245530999946,"output_power":0.00012857452146924577,"reference_active":true,"delta_db":0.6039292804315068},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8450792038001933e-05,"output_power":3.546780197860119e-05,"reference_active":true,"delta_db":0.957399139077641}],"active_band_count":3,"active_band_mae_db":1.062864538988746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020713167740975364,"log_rms_mae_db":1.1435034823787023,"rms_envelope_nrmse":0.18315906320578793,"envelope_correlation":0.9890398590695396,"rms_gain_error_db":1.2442602389692128,"reference_mean_power":0.005603505754865554,"output_mean_power":0.007462525809913921},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0483171696395765e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010458478077031374,"output_power":0.0015504445052643037,"reference_active":true,"delta_db":1.7098773586141733},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.131979407845486e-05,"output_power":6.816957816419788e-05,"reference_active":true,"delta_db":0.4598991882151121},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.629293286100356e-06,"output_power":1.8447807254435394e-06,"reference_active":true,"delta_db":0.5394548449807892}],"active_band_count":3,"active_band_mae_db":0.9030771306033584}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015831513319989338,"log_rms_mae_db":1.799438068902146,"rms_envelope_nrmse":0.20308217709224372,"envelope_correlation":0.985601881775937,"rms_gain_error_db":1.21317989241188,"reference_mean_power":0.0022309508329157127,"output_mean_power":0.002949904717297871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0339679622838414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046453953265101344,"output_power":0.0005384212166775867,"reference_active":true,"delta_db":0.6409948640243145},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.019812862224435e-05,"output_power":8.30215239881898e-05,"reference_active":true,"delta_db":0.7286516662224185},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5452757374862184e-05,"output_power":1.7805470749444444e-05,"reference_active":true,"delta_db":0.6154747448774404}],"active_band_count":3,"active_band_mae_db":0.6617070917080579}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014375880677218868,"log_rms_mae_db":1.411578957169778,"rms_envelope_nrmse":0.17938621457462545,"envelope_correlation":0.9917566941210777,"rms_gain_error_db":1.2070796518692137,"reference_mean_power":0.0022214967634666704,"output_mean_power":0.0029332808708667354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.581275628581243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004040718027646314,"output_power":0.0004704452495366651,"reference_active":true,"delta_db":0.6605054222717588},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.035103681693411e-05,"output_power":4.6121712124219905e-05,"reference_active":true,"delta_db":0.580507223790294},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1841518228550596e-06,"output_power":1.1606726221877853e-06,"reference_active":true,"delta_db":-0.08697647387334495}],"active_band_count":3,"active_band_mae_db":0.4426630399784659}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":1.488086949137222,"rms_envelope_nrmse":0.22352009282466315,"envelope_correlation":0.9907548994228124,"rms_gain_error_db":1.4722262773320476,"reference_mean_power":0.0002507880581966276,"output_mean_power":0.00035198931525530296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4212041991971411e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.17102478326245e-05,"output_power":0.00011448016833448418,"reference_active":true,"delta_db":2.0314903591138536},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.275733290040069e-06,"output_power":1.0793994861523017e-05,"reference_active":true,"delta_db":1.7130543643634675},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.07244494044588e-06,"output_power":1.4339801383372198e-06,"reference_active":true,"delta_db":1.2616813142363135}],"active_band_count":3,"active_band_mae_db":1.6687420125712116}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":1.0698798588052745,"rms_envelope_nrmse":0.20197511533432294,"envelope_correlation":0.9962134211496904,"rms_gain_error_db":1.4716782167210183,"reference_mean_power":0.0002518169764916093,"output_mean_power":0.00035338883709766327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3611130645010087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.48897476225386e-05,"output_power":0.0001054169841644472,"reference_active":true,"delta_db":2.107345024375901},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.71133538995873e-06,"output_power":5.304066801627102e-06,"reference_active":true,"delta_db":1.5507878220249243},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.48681833738951e-08,"output_power":9.264190190169838e-08,"reference_active":true,"delta_db":0.380625562823581}],"active_band_count":3,"active_band_mae_db":1.346252803074802}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":1.5895465215332036,"rms_envelope_nrmse":0.2301889169021175,"envelope_correlation":0.959850810425382,"rms_gain_error_db":0.15471664990130807,"reference_mean_power":2.817379488191862e-05,"output_mean_power":2.9195573682800633e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.938019558986415e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.123435668533647e-06,"output_power":5.9199877825450056e-06,"reference_active":true,"delta_db":-0.14674349157808883},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7222454435376484e-06,"output_power":2.011279675814444e-06,"reference_active":true,"delta_db":-1.3145481452639043},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.909737577930933e-07,"output_power":5.954821958388002e-07,"reference_active":true,"delta_db":-0.6459277263708234}],"active_band_count":3,"active_band_mae_db":0.7024064544042722}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":0.9612671488020901,"rms_envelope_nrmse":0.13690550826266298,"envelope_correlation":0.984707964864071,"rms_gain_error_db":0.1552702172945627,"reference_mean_power":2.8310170166312666e-05,"output_mean_power":2.9340634513222546e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8087824895787787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.951898475340427e-06,"output_power":5.6550001729390405e-06,"reference_active":true,"delta_db":-0.22222892068934225},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2501096733423978e-06,"output_power":8.56664413918727e-07,"reference_active":true,"delta_db":-1.641373891981448},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.567916579161295e-08,"output_power":3.5061840957615456e-08,"reference_active":true,"delta_db":-2.7259291262791834}],"active_band_count":3,"active_band_mae_db":1.5298439796499912}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"facodec","status":"measured","mos":{"status":"measured","value":4.124079704284668,"predicted_mos":4.124079704284668,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c98350e089fafc470ff63f02d9c279811e7dc0cc8c8befb307ea27143c6ce720","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7076112929989904},"started_at":"2026-10-13T04:47:48.704001+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6699000045482535e-05,"native_input_samples":96000,"native_input_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","output_sha256":"0fe4ff555a431bc2c5b8e77b48b7ebaa5c3f1f61a979f12c567be8f726f8173d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.508632647000013,"synthesis_s":5.879399094000291,"total_s":7.3880317410003045,"rtf":1.2313386235000507,"output_peak":0.40503397583961487,"output_rms":0.06876303129812866,"finished_at":"2026-10-13T04:47:56.098502+00:00","evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"0fe4ff555a431bc2c5b8e77b48b7ebaa5c3f1f61a979f12c567be8f726f8173d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.165226805071272,"spectral_convergence":0.22463187756243108,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.4885308309406,"spectral_convergence":0.22143210047707387,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.58551670918938,"spectral_convergence":0.23052477849037523,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0011383202958740594,"delta_db":-1.0942887365041023,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.002117587680687277,"delta_db":-0.6811158186348322,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.001219605043851472,"delta_db":-0.41781325318982077,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.0001881392548990677,"delta_db":-1.105789249109059,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":3.476886468641763e-05,"delta_db":-1.0643890178623172,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":2.9933333309126772e-05,"delta_db":-0.6099387372106899,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8288891354184701,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7394302914275934,"pitch":null,"output_rms_dbfs":-23.25289972898633,"output_peak":0.40503397583961487,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.413091448400418},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.165226805071272,"spectral_convergence":0.22463187756243108,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.4885308309406,"spectral_convergence":0.22143210047707387,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.58551670918938,"spectral_convergence":0.23052477849037523,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0011383202958740594,"delta_db":-1.0942887365041023,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.002117587680687277,"delta_db":-0.6811158186348322,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.001219605043851472,"delta_db":-0.41781325318982077,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.0001881392548990677,"delta_db":-1.105789249109059,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":3.476886468641763e-05,"delta_db":-1.0643890178623172,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":2.9933333309126772e-05,"delta_db":-0.6099387372106899,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8288891354184701,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7394302914275934,"pitch":null,"output_rms_dbfs":-23.25289972898633,"output_peak":0.40503397583961487,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.413091448400418},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9073804795081657,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07487322298437474,"rms_dbfs":-22.513469437558737,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.708326506096444,"mean":-3.738085428873698e-05,"zero_energy":false},"output":{"rms":0.06876303129812865,"rms_dbfs":-23.25289972898633,"peak":0.40503397583961487,"peak_dbfs":-7.850170898631567,"crest_factor_db":15.402728830354764,"mean":4.8974529046611555e-05,"zero_energy":false},"rms_difference_db":-0.7394302914275919},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":3.9240174281721486e-07,"output_rms_dbfs":-64.0626907240774,"output_zero_energy":false,"output_peak":0.006959328427910805,"output_energy_fraction":1.037363381469196e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":359,"both_estimated_voiced_frames":345,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9556786703601108,"absolute_difference_cents_median":3.6614431219737322,"absolute_difference_cents_p95":15.667589307812232,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002110151907366841,"log_rms_mae_db":1.1533911203856744,"rms_envelope_nrmse":0.14084444210523714,"envelope_correlation":0.9812617081069189,"rms_gain_error_db":-0.7395950329250284,"reference_mean_power":0.005604891233518836,"output_mean_power":0.004727240371690185},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2250047113785757e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011643562088722453,"output_power":0.0008488036399247885,"reference_active":true,"delta_db":-1.3727863009143018},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011188245530999946,"output_power":9.536660668054179e-05,"reference_active":true,"delta_db":-0.6936565867664141},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8450792038001933e-05,"output_power":2.337212205343316e-05,"reference_active":true,"delta_db":-0.853962155603491}],"active_band_count":3,"active_band_mae_db":0.9734683477614023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020713167740975364,"log_rms_mae_db":1.0626654988178643,"rms_envelope_nrmse":0.13520593028107025,"envelope_correlation":0.9819585004431596,"rms_gain_error_db":-0.7413060673526677,"reference_mean_power":0.005603505754865554,"output_mean_power":0.0047242102282697195},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0483171696395765e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010458478077031374,"output_power":0.0007601827038967381,"reference_active":true,"delta_db":-1.3855050631076322},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.131979407845486e-05,"output_power":5.1521297334904836e-05,"reference_active":true,"delta_db":-0.7561389748948398},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.629293286100356e-06,"output_power":1.3180582288647535e-06,"reference_active":true,"delta_db":-0.9206467105910541}],"active_band_count":3,"active_band_mae_db":1.0207635828645087}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015831513319989338,"log_rms_mae_db":1.4275841463659906,"rms_envelope_nrmse":0.12854218581107454,"envelope_correlation":0.9863101935586819,"rms_gain_error_db":-0.34188966574912183,"reference_mean_power":0.0022309508329157127,"output_mean_power":0.0020620587404022255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0339679622838414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046453953265101344,"output_power":0.0004399515807547469,"reference_active":true,"delta_db":-0.23617796209911268},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.019812862224435e-05,"output_power":5.9099829907884484e-05,"reference_active":true,"delta_db":-0.7473930367591696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5452757374862184e-05,"output_power":1.199326425451805e-05,"reference_active":true,"delta_db":-1.1006858290963821}],"active_band_count":3,"active_band_mae_db":0.6947522759848881}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014375880677218868,"log_rms_mae_db":1.2229305668197408,"rms_envelope_nrmse":0.11875670563480173,"envelope_correlation":0.9876431501473566,"rms_gain_error_db":-0.33919698012785826,"reference_mean_power":0.0022214967634666704,"output_mean_power":0.002054593864308165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.581275628581243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004040718027646314,"output_power":0.0003863186230824082,"reference_active":true,"delta_db":-0.19512900761653104},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.035103681693411e-05,"output_power":3.4388045173649886e-05,"reference_active":true,"delta_db":-0.6944720979140477},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1841518228550596e-06,"output_power":8.528214273971334e-07,"reference_active":true,"delta_db":-1.4254928426059255}],"active_band_count":3,"active_band_mae_db":0.7716979827121682}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":1.6356841002136573,"rms_envelope_nrmse":0.16495517831818518,"envelope_correlation":0.9892352400692719,"rms_gain_error_db":-1.0973745363420508,"reference_mean_power":0.0002507880581966276,"output_mean_power":0.00019479122965654707},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4212041991971411e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.17102478326245e-05,"output_power":5.512091622977654e-05,"reference_active":true,"delta_db":-1.1426479555770008},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.275733290040069e-06,"output_power":5.467530173161559e-06,"reference_active":true,"delta_db":-1.2408558212943004},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.07244494044588e-06,"output_power":1.3876363981613851e-06,"reference_active":true,"delta_db":1.1190067835033193}],"active_band_count":3,"active_band_mae_db":1.1675035201248736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":1.4070253533160109,"rms_envelope_nrmse":0.15017627919667112,"envelope_correlation":0.9928855357708122,"rms_gain_error_db":-1.0978996271163157,"reference_mean_power":0.0002518169764916093,"output_mean_power":0.00019556676075621114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3611130645010087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.48897476225386e-05,"output_power":4.995358916480025e-05,"reference_active":true,"delta_db":-1.1360938731215926},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.71133538995873e-06,"output_power":2.7452281773312683e-06,"reference_active":true,"delta_db":-1.3095175502531362},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.48681833738951e-08,"output_power":1.1216043281854621e-07,"reference_active":true,"delta_db":1.210947704481558}],"active_band_count":3,"active_band_mae_db":1.2188530426187623}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":2.034388337462202,"rms_envelope_nrmse":0.26982252809731777,"envelope_correlation":0.9478362558784115,"rms_gain_error_db":-0.964210438395669,"reference_mean_power":2.817379488191862e-05,"output_mean_power":2.2564426741955804e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.938019558986415e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.123435668533647e-06,"output_power":4.783190814955115e-06,"reference_active":true,"delta_db":-1.0727745315275954},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7222454435376484e-06,"output_power":2.948720609028639e-06,"reference_active":true,"delta_db":0.3470634558151765},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.909737577930933e-07,"output_power":9.759678242949608e-07,"reference_active":true,"delta_db":1.499739462865215}],"active_band_count":3,"active_band_mae_db":0.9731924834026623}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":1.7069901104079328,"rms_envelope_nrmse":0.2262354583949209,"envelope_correlation":0.9626224501919547,"rms_gain_error_db":-0.9637627291688029,"reference_mean_power":2.8310170166312666e-05,"output_mean_power":2.2675987395025032e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8087824895787787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.951898475340427e-06,"output_power":5.1442146779788774e-06,"reference_active":true,"delta_db":-0.6333643037292868},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2501096733423978e-06,"output_power":1.3862129820786375e-06,"reference_active":true,"delta_db":0.4488184600027084},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.567916579161295e-08,"output_power":1.0959952184235207e-07,"reference_active":true,"delta_db":2.2238103136719314}],"active_band_count":3,"active_band_mae_db":1.1019976924679755}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.188446044921875,"predicted_mos":3.188446044921875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"f3b7a55ced181f6b6c47c705d44816ba32a1b2de81b944aa455e686b30434559","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.3139809720014455},"started_at":"2026-10-13T07:05:31.761542+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008981712000604603,"native_input_samples":264600,"native_input_sha256":"ac0872db5690f69fb216d6bbc1da2e18e7c661c812e59a92be57b69bce4b7660","output_sha256":"870bb7b50fd7e6cd22e4f345b87bf9a222dbbe28f39ec0447ce31d5f9c2def3f","output_samples":264704,"length_delta_samples":104,"analysis_s":0.153218133998962,"synthesis_s":4.054722586000935,"total_s":4.207940719999897,"rtf":0.7013234533333161,"output_peak":0.45350438356399536,"output_rms":0.06955265025523054,"finished_at":"2026-10-13T07:05:35.999205+00:00","evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"870bb7b50fd7e6cd22e4f345b87bf9a222dbbe28f39ec0447ce31d5f9c2def3f","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.14971728049517,"output_peak":0.45275607210913554},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.566225823604707,"spectral_convergence":0.3714368564820863,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.440963572607831,"spectral_convergence":0.2890960706691233,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.211862602709317,"spectral_convergence":0.3699893496878323,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.001134873074318444,"delta_db":-1.1074606071135602,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.0021719916356159573,"delta_db":-0.5709483516231254,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.0012934119905188965,"delta_db":-0.16263654365490807,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.00017733622221214785,"delta_db":-1.3626089217083828,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":3.947370498645829e-05,"delta_db":-0.5132152059205197,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":2.317840905559472e-05,"delta_db":-1.7206533223558609,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9062538253960595,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6378391343687135,"pitch":null,"output_rms_dbfs":-23.15130857192745,"output_peak":0.45275607210913554,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.406350666307285},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7808899739318238,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07487322298437474,"rms_dbfs":-22.513469437558737,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.708326506096444,"mean":-3.738085428873698e-05,"zero_energy":false},"output":{"rms":0.06957201331503567,"rms_dbfs":-23.15130857192745,"peak":0.45275607210913554,"peak_dbfs":-6.882714329242963,"crest_factor_db":16.26859424268449,"mean":-0.00014187290215274213,"zero_energy":false},"rms_difference_db":-0.6378391343687149},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":4.662215419256236e-07,"output_rms_dbfs":-63.314076636269256,"output_zero_energy":false,"output_peak":0.005715754298139224,"output_energy_fraction":1.2040186291192284e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":366,"both_estimated_voiced_frames":331,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9168975069252078,"absolute_difference_cents_median":10.284897198816022,"absolute_difference_cents_p95":46.87569088305189,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002110151907366841,"log_rms_mae_db":1.0626372056619773,"rms_envelope_nrmse":0.10346745108565815,"envelope_correlation":0.9920880942461329,"rms_gain_error_db":-0.6369531943329079,"reference_mean_power":0.005604891233518836,"output_mean_power":0.0048402954296782805},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2250047113785757e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011643562088722453,"output_power":0.0010933027349769142,"reference_active":true,"delta_db":-0.27345428848224185},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011188245530999946,"output_power":9.171757025057166e-05,"reference_active":true,"delta_db":-0.8630944755418808},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8450792038001933e-05,"output_power":2.289197170867323e-05,"reference_active":true,"delta_db":-0.9441116066007929}],"active_band_count":3,"active_band_mae_db":0.6935534568749718}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020713167740975364,"log_rms_mae_db":0.8267584092528724,"rms_envelope_nrmse":0.08798657592675112,"envelope_correlation":0.9957322382128191,"rms_gain_error_db":-0.6325842628723871,"reference_mean_power":0.005603505754865554,"output_mean_power":0.004843969456041134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0483171696395765e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010458478077031374,"output_power":0.0009876938936254464,"reference_active":true,"delta_db":-0.24846121608460264},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.131979407845486e-05,"output_power":4.876052261507514e-05,"reference_active":true,"delta_db":-0.9953233601767253},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.629293286100356e-06,"output_power":1.2871065902892316e-06,"reference_active":true,"delta_db":-1.0238475389553054}],"active_band_count":3,"active_band_mae_db":0.7558773717388778}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015831513319989338,"log_rms_mae_db":2.166389874618184,"rms_envelope_nrmse":0.13440023656454037,"envelope_correlation":0.9846127590701178,"rms_gain_error_db":-0.322990148509999,"reference_mean_power":0.0022309508329157127,"output_mean_power":0.002071051909318082},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0339679622838414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046453953265101344,"output_power":0.00043073632617330574,"reference_active":true,"delta_db":-0.3281117911992351},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.019812862224435e-05,"output_power":5.6428144489746694e-05,"reference_active":true,"delta_db":-0.9482976491306603},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5452757374862184e-05,"output_power":1.156684420042161e-05,"reference_active":true,"delta_db":-1.2579109986395016}],"active_band_count":3,"active_band_mae_db":0.8447734796564657}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014375880677218868,"log_rms_mae_db":1.590541693317666,"rms_envelope_nrmse":0.10788477310086389,"envelope_correlation":0.9896611088801002,"rms_gain_error_db":-0.30334415960177485,"reference_mean_power":0.0022214967634666704,"output_mean_power":0.0020716255987475937},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.581275628581243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004040718027646314,"output_power":0.0003817782896972245,"reference_active":true,"delta_db":-0.24647317185752332},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.035103681693411e-05,"output_power":3.200190287820454e-05,"reference_active":true,"delta_db":-1.006788955172609},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1841518228550596e-06,"output_power":7.773364151422214e-07,"reference_active":true,"delta_db":-1.8279837469682838}],"active_band_count":3,"active_band_mae_db":1.0270819579994719}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":2.8262181510120783,"rms_envelope_nrmse":0.23740378477234877,"envelope_correlation":0.9750571004340463,"rms_gain_error_db":-1.5838829506009335,"reference_mean_power":0.0002507880581966276,"output_mean_power":0.00017414802669263289},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4212041991971411e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.17102478326245e-05,"output_power":4.464739110146352e-05,"reference_active":true,"delta_db":-2.057851367722255},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.275733290040069e-06,"output_power":3.9844143631905545e-06,"reference_active":true,"delta_db":-2.6151227355417905},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.07244494044588e-06,"output_power":6.799201878427309e-07,"reference_active":true,"delta_db":-1.979170684415559}],"active_band_count":3,"active_band_mae_db":2.2173815958932015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":2.376058406805801,"rms_envelope_nrmse":0.21427114632721364,"envelope_correlation":0.9833330989972615,"rms_gain_error_db":-1.5808227462790425,"reference_mean_power":0.0002518169764916093,"output_mean_power":0.00017498576904103016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3611130645010087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.48897476225386e-05,"output_power":4.070635870791097e-05,"reference_active":true,"delta_db":-2.025138297287133},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.71133538995873e-06,"output_power":1.970777948893341e-06,"reference_active":true,"delta_db":-2.7489250862112007},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.48681833738951e-08,"output_power":4.3776916708574555e-08,"reference_active":true,"delta_db":-2.8749973596342664}],"active_band_count":3,"active_band_mae_db":2.549686914377533}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":3.9801341566060113,"rms_envelope_nrmse":0.276249807419924,"envelope_correlation":0.9458805001643897,"rms_gain_error_db":-0.5187171446140303,"reference_mean_power":2.817379488191862e-05,"output_mean_power":2.5001935701767065e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.938019558986415e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.123435668533647e-06,"output_power":7.62133771896608e-06,"reference_active":true,"delta_db":0.950360470377683},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7222454435376484e-06,"output_power":2.5901221222717756e-06,"reference_active":true,"delta_db":-0.21607038346794621},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.909737577930933e-07,"output_power":5.437497198397447e-07,"reference_active":true,"delta_db":-1.0406250759873514}],"active_band_count":3,"active_band_mae_db":0.7356853099443269}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":3.420119233972116,"rms_envelope_nrmse":0.2244870681439408,"envelope_correlation":0.963137175645061,"rms_gain_error_db":-0.5182679118743813,"reference_mean_power":2.8310170166312666e-05,"output_mean_power":2.5125556443017722e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8087824895787787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.951898475340427e-06,"output_power":7.1813561062526e-06,"reference_active":true,"delta_db":0.81550948272684},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2501096733423978e-06,"output_power":1.2043376112594725e-06,"reference_active":true,"delta_db":-0.16199866250479103},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.567916579161295e-08,"output_power":5.354191450167743e-08,"reference_active":true,"delta_db":-0.8873373183353577}],"active_band_count":3,"active_band_mae_db":0.6216151545223295}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"golf","status":"measured","mos":{"status":"measured","value":3.0835518836975098,"predicted_mos":3.0835518836975098,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"d85b854761cc96da72b16386bf46502a746ba52f360dc3ab890a9974aa194073","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1718550999994477},"started_at":"2026-10-13T05:02:05.551689+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009196871999847644,"native_input_samples":144000,"native_input_sha256":"36f80382f9af4ee921a50efb1dfe2cc01979e856582819bd245725a8544f3d76","output_sha256":"ec6aa58244820ea80066cf2894045ea3569458a79349a46733bb7e31a38d7c9c","output_samples":143760,"length_delta_samples":-240,"analysis_s":3.5604391040005794,"synthesis_s":0.1229980929992962,"total_s":3.6834371969998756,"rtf":0.6139061994999793,"output_peak":0.5426712036132812,"output_rms":0.06808925594298317,"finished_at":"2026-10-13T05:02:09.264224+00:00","analysis_stages":{"external_f0_s":2.7386917100002393},"evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"ec6aa58244820ea80066cf2894045ea3569458a79349a46733bb7e31a38d7c9c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.33473670842198,"output_peak":0.5443981314806946},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.581579457076234,"spectral_convergence":0.5530980642099067,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.074574860157291,"spectral_convergence":0.21589202559659032,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.286016127190622,"spectral_convergence":0.23255924589622162,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014643261603613927,"output_power":0.0011576580608300443,"delta_db":-1.0205752103769803,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002462312969852845,"output_power":0.002116112992687626,"delta_db":-0.658043989560998,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001344178132507633,"output_power":0.001099499652929244,"delta_db":-0.8726172928318071,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024329376779963144,"output_power":0.00021144892349001063,"delta_db":-0.6092550576164815,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.452905557794614e-05,"output_power":3.04102410193266e-05,"delta_db":-1.6562362202442371,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.45262247981124e-05,"output_power":2.4959326331542093e-05,"delta_db":-1.4091623459480516,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0376483527630926,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8113140396112108,"pitch":null,"output_rms_dbfs":-23.33473670842198,"output_peak":0.5443981314806946,"reference_rms_dbfs":-22.523422668810767,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.647390148141382},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9272874280277011,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07478747429147184,"rms_dbfs":-22.523422668810767,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.718279737348475,"mean":-4.173049544651242e-05,"zero_energy":false},"output":{"rms":0.06811820019119555,"rms_dbfs":-23.33473670842198,"peak":0.5443981314806946,"peak_dbfs":-5.2816674816083085,"crest_factor_db":18.053069226813673,"mean":0.0005662637972443275,"zero_energy":false},"rms_difference_db":-0.8113140396112115},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":5.647320938207807e-07,"output_rms_dbfs":-62.48157530532633,"output_zero_energy":false,"output_peak":0.00737808487888816,"output_energy_fraction":1.5238793405817512e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":368,"both_estimated_voiced_frames":347,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.961218836565097,"absolute_difference_cents_median":4.0823785030841675,"absolute_difference_cents_p95":16.869806425178368,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002110151907366841,"log_rms_mae_db":1.868484069415018,"rms_envelope_nrmse":0.15231598846600453,"envelope_correlation":0.9788514164196319,"rms_gain_error_db":-0.8110580922530781,"reference_mean_power":0.005592763738138928,"output_mean_power":0.004640028670697591},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.2253024253046896e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011628606318792015,"output_power":0.0010033568829015077,"reference_active":true,"delta_db":-0.6407223370089886},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011167283054267222,"output_power":0.00010274639078728873,"reference_active":true,"delta_db":-0.3618094904505639},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8329826117376648e-05,"output_power":2.9265679976397686e-05,"reference_active":true,"delta_db":0.14114710307679432}],"active_band_count":3,"active_band_mae_db":0.3812263101787823}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00020713167740975364,"log_rms_mae_db":1.5353957044695015,"rms_envelope_nrmse":0.11835955789660739,"envelope_correlation":0.9905177898264945,"rms_gain_error_db":-0.7988245463116926,"reference_mean_power":0.005595170324657781,"output_mean_power":0.004655119745103332},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.0488666704154796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010443710729859644,"output_power":0.0008973201690617374,"reference_active":true,"delta_db":-0.6590740472296548},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.10579473604591e-05,"output_power":5.5921539233741354e-05,"reference_active":true,"delta_db":-0.3816308333632959},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6278003177874578e-06,"output_power":1.561061270160631e-06,"reference_active":true,"delta_db":-0.1818117989077197}],"active_band_count":3,"active_band_mae_db":0.4075055598335568}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015831513319989338,"log_rms_mae_db":1.6277821834692583,"rms_envelope_nrmse":0.16431587666028097,"envelope_correlation":0.9781406364246047,"rms_gain_error_db":-0.5405491667294935,"reference_mean_power":0.0022188678341308337,"output_mean_power":0.0019591898301758847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0303736616980481e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004630694103534692,"output_power":0.00043120520222835064,"reference_active":true,"delta_db":-0.3096210141308478},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.008976697190523e-05,"output_power":6.098804656660619e-05,"reference_active":true,"delta_db":-0.6040989285703495},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5300250466715245e-05,"output_power":1.75228049681178e-05,"reference_active":true,"delta_db":0.5890508689045665}],"active_band_count":3,"active_band_mae_db":0.5009236038685879}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014375880677218868,"log_rms_mae_db":1.141236818801343,"rms_envelope_nrmse":0.11415233316219082,"envelope_correlation":0.9902652630837886,"rms_gain_error_db":-0.5185057177943363,"reference_mean_power":0.002214595659103923,"output_mean_power":0.0019653679633856664},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.550919233754655e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004027083290910328,"output_power":0.00038040127674218184,"reference_active":true,"delta_db":-0.247486461859595},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.024294850999427e-05,"output_power":3.492865502213102e-05,"reference_active":true,"delta_db":-0.6150793016112955},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1803004365123525e-06,"output_power":1.0414796425210034e-06,"reference_active":true,"delta_db":-0.5434178239258961}],"active_band_count":3,"active_band_mae_db":0.4686611957989289}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":1.7828112261803024,"rms_envelope_nrmse":0.16511872059912563,"envelope_correlation":0.9839749945854901,"rms_gain_error_db":-0.7481287855755968,"reference_mean_power":0.0002510646197622326,"output_mean_power":0.00021133558838818173},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4131726257405377e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.127243419484173e-05,"output_power":6.586211036465158e-05,"reference_active":true,"delta_db":-0.3428594934205518},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.224682966622807e-06,"output_power":6.753669601733249e-06,"reference_active":true,"delta_db":-0.29278983575565887},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0578191294822028e-06,"output_power":1.0751464956239561e-06,"reference_active":true,"delta_db":0.07056227212914871}],"active_band_count":3,"active_band_mae_db":0.23540386710178643}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":1.2712918296207798,"rms_envelope_nrmse":0.12422794558957524,"envelope_correlation":0.9929142010072602,"rms_gain_error_db":-0.7456094867411986,"reference_mean_power":0.00025214322949227265,"output_mean_power":0.00021236667254073504},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.3536870204248514e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.44779696940117e-05,"output_power":6.028738774946076e-05,"reference_active":true,"delta_db":-0.2918488748086439},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6759173050717935e-06,"output_power":3.331046510841826e-06,"reference_active":true,"delta_db":-0.42785036044695923},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429697290577143e-08,"output_power":8.397906292527383e-08,"reference_active":true,"delta_db":-0.016409551585912667}],"active_band_count":3,"active_band_mae_db":0.24536959561383856}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":2.6351872855449052,"rms_envelope_nrmse":0.24967542761346867,"envelope_correlation":0.9596777883480931,"rms_gain_error_db":-0.6945047160178902,"reference_mean_power":2.8216655406324318e-05,"output_mean_power":2.4046676608159992e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.9363800983558643e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.109994963869437e-06,"output_power":6.706237333892414e-06,"reference_active":true,"delta_db":0.404380667786027},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.721869074067349e-06,"output_power":3.3621432898382973e-06,"reference_active":true,"delta_db":0.9174898735836399},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.90430284621993e-07,"output_power":7.790295252035499e-07,"reference_active":true,"delta_db":0.5243408497778567}],"active_band_count":3,"active_band_mae_db":0.6154037970491745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":2.50606529111067,"rms_envelope_nrmse":0.20946122608027243,"envelope_correlation":0.9726591628190144,"rms_gain_error_db":-0.6944616745933568,"reference_mean_power":2.835552914734482e-05,"output_mean_power":2.4165266482346856e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8074847183074995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.936763899602731e-06,"output_power":6.701293223895675e-06,"reference_active":true,"delta_db":0.5260884375574963},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.250632401857866e-06,"output_power":1.6992640827829159e-06,"reference_active":true,"delta_db":1.331312015021155},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.578984947573361e-08,"output_power":4.705887143276916e-08,"reference_active":true,"delta_db":-1.4551738504982685}],"active_band_count":3,"active_band_mae_db":1.1041914343589732}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4769959449768066,"predicted_mos":3.4769959449768066,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"ca8777c4b853d6df09d812fb8b7e49191efd339134fa3dae615e7a3d13327b2a","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0332206370003405},"started_at":"2026-10-13T05:06:45.457693+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028239850007594214,"native_input_samples":144000,"native_input_sha256":"36f80382f9af4ee921a50efb1dfe2cc01979e856582819bd245725a8544f3d76","output_sha256":"f1110c802cb7bfafe81cdab66af9c4dc70f473f90041e93f974ac5d02bc52ee0","output_samples":144120,"length_delta_samples":120,"analysis_s":2.0892637910001213,"synthesis_s":3.2585972489996493,"total_s":5.347861039999771,"rtf":0.8913101733332951,"output_peak":0.37306228280067444,"output_rms":0.07407395720200323,"finished_at":"2026-10-13T05:06:50.818291+00:00","analysis_stages":{"external_f0_s":1.280101360000117},"evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"f1110c802cb7bfafe81cdab66af9c4dc70f473f90041e93f974ac5d02bc52ee0","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.603671682732525,"output_peak":0.3674363093517678},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.706536458931095,"spectral_convergence":0.45617615920621857,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.79178815482894,"spectral_convergence":0.20213852408072663,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.821371422518392,"spectral_convergence":0.20814158912254097,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0016463675733934197,"delta_db":0.5083345080777089,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.0023312622759494826,"delta_db":-0.2636184754090251,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.0012108171128794158,"delta_db":-0.4492198674740778,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.0002353701605826086,"delta_db":-0.13306940996460134,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":4.540660911663464e-05,"delta_db":0.09489659175929535,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":2.329084712622275e-05,"delta_db":-1.6996367049457441,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5247959262717421,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08881837881145749,"pitch":null,"output_rms_dbfs":-22.602287816370197,"output_peak":0.3674363093517678,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.106565345426142},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9324955718932747,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07487322298437474,"rms_dbfs":-22.513469437558737,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.708326506096444,"mean":-3.738085428873698e-05,"zero_energy":false},"output":{"rms":0.07411150099038465,"rms_dbfs":-22.602287816370197,"peak":0.3674363093517678,"peak_dbfs":-8.696358595228062,"crest_factor_db":13.90592922114213,"mean":5.078063853735891e-07,"zero_energy":false},"rms_difference_db":-0.08881837881146026},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":1.368470244283131e-06,"output_rms_dbfs":-58.637646413551614,"output_zero_energy":false,"output_peak":0.0364978749909195,"output_energy_fraction":3.1143982974196705e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":377,"both_estimated_voiced_frames":349,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9667590027700831,"absolute_difference_cents_median":6.113166741112613,"absolute_difference_cents_p95":27.510838271743566,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002110151907366841,"log_rms_mae_db":1.3995775797212349,"rms_envelope_nrmse":0.11989524720098547,"envelope_correlation":0.9810722299995247,"rms_gain_error_db":-0.08593921475650709,"reference_mean_power":0.005604891233518836,"output_mean_power":0.005495070488884845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2250047113785757e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011643562088722453,"output_power":0.0011583783065835785,"reference_active":true,"delta_db":-0.022354476520043534},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011188245530999946,"output_power":0.0001152192780413503,"reference_active":true,"delta_db":0.12763161041274304},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8450792038001933e-05,"output_power":2.692056152029911e-05,"reference_active":true,"delta_db":-0.24010246826711068}],"active_band_count":3,"active_band_mae_db":0.13002951839996577}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020713167740975364,"log_rms_mae_db":0.9685155166366386,"rms_envelope_nrmse":0.07560058095443999,"envelope_correlation":0.9919626919042374,"rms_gain_error_db":-0.08121002853801913,"reference_mean_power":0.005603505754865554,"output_mean_power":0.005499697712755265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0483171696395765e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010458478077031374,"output_power":0.0010368154237095474,"reference_active":true,"delta_db":-0.037670412211887334},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.131979407845486e-05,"output_power":6.368099263544616e-05,"reference_active":true,"delta_db":0.16409136646606798},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.629293286100356e-06,"output_power":1.4322288344265823e-06,"reference_active":true,"delta_db":-0.5598685496363436}],"active_band_count":3,"active_band_mae_db":0.2538767761047663}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015831513319989338,"log_rms_mae_db":1.928438494047064,"rms_envelope_nrmse":0.1555755610709902,"envelope_correlation":0.9787123151083147,"rms_gain_error_db":-0.29725673441416905,"reference_mean_power":0.0022309508329157127,"output_mean_power":0.0020833600198472716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0339679622838414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046453953265101344,"output_power":0.0004338323632758434,"reference_active":true,"delta_db":-0.2970073204174486},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.019812862224435e-05,"output_power":6.887200587502269e-05,"reference_active":true,"delta_db":-0.08282802782199582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5452757374862184e-05,"output_power":1.8419586584373852e-05,"reference_active":true,"delta_db":0.7627389275645318}],"active_band_count":3,"active_band_mae_db":0.38085809193465875}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014375880677218868,"log_rms_mae_db":1.355723476151882,"rms_envelope_nrmse":0.0998811098030437,"envelope_correlation":0.9911546033641473,"rms_gain_error_db":-0.2910052985205102,"reference_mean_power":0.0022214967634666704,"output_mean_power":0.002077519720754543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.581275628581243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004040718027646314,"output_power":0.00037778599851417385,"reference_active":true,"delta_db":-0.29212687159799533},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.035103681693411e-05,"output_power":3.976923821249917e-05,"reference_active":true,"delta_db":-0.06307426296046775},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1841518228550596e-06,"output_power":1.2094827747234333e-06,"reference_active":true,"delta_db":0.09192299722719077}],"active_band_count":3,"active_band_mae_db":0.14904137726188463}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":1.799194904422851,"rms_envelope_nrmse":0.15286528801194912,"envelope_correlation":0.9828928950450481,"rms_gain_error_db":-0.09354146813756134,"reference_mean_power":0.0002507880581966276,"output_mean_power":0.0002454441616151034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4212041991971411e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.17102478326245e-05,"output_power":6.878954046868572e-05,"reference_active":true,"delta_db":-0.18058815174538098},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.275733290040069e-06,"output_power":6.6425980374748536e-06,"reference_active":true,"delta_db":-0.3953879770646923},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.07244494044588e-06,"output_power":1.2664934792049337e-06,"reference_active":true,"delta_db":0.7222795342338705}],"active_band_count":3,"active_band_mae_db":0.4327518876813146}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":1.2075084014282347,"rms_envelope_nrmse":0.10010883760840084,"envelope_correlation":0.9923613257326479,"rms_gain_error_db":-0.0929233463341672,"reference_mean_power":0.0002518169764916093,"output_mean_power":0.00024648623463805516},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3611130645010087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.48897476225386e-05,"output_power":6.236442658305863e-05,"reference_active":true,"delta_db":-0.1723915152150044},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.71133538995873e-06,"output_power":3.35296147783123e-06,"reference_active":true,"delta_db":-0.44101639065399223},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.48681833738951e-08,"output_power":1.0078883388186174e-07,"reference_active":true,"delta_db":0.7466751439455711}],"active_band_count":3,"active_band_mae_db":0.4533610166048559}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":1.9277608804315463,"rms_envelope_nrmse":0.25930999965038626,"envelope_correlation":0.9475458668809108,"rms_gain_error_db":-0.5604947685390004,"reference_mean_power":2.817379488191862e-05,"output_mean_power":2.476257885654901e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.938019558986415e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.123435668533647e-06,"output_power":4.950897343562636e-06,"reference_active":true,"delta_db":-0.9231123823501259},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7222454435376484e-06,"output_power":1.997068390961817e-06,"reference_active":true,"delta_db":-1.3453434172804444},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.909737577930933e-07,"output_power":5.023879857166703e-07,"reference_active":true,"delta_db":-1.3842230881463637}],"active_band_count":3,"active_band_mae_db":1.2175596292589779}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":1.2000064989757986,"rms_envelope_nrmse":0.15909790273974356,"envelope_correlation":0.9806013695790963,"rms_gain_error_db":-0.5604687169182114,"reference_mean_power":2.8310170166312666e-05,"output_mean_power":2.4882591406555923e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8087824895787787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.951898475340427e-06,"output_power":4.752593711356582e-06,"reference_active":true,"delta_db":-0.9772482559748786},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2501096733423978e-06,"output_power":9.050148290011309e-07,"reference_active":true,"delta_db":-1.4029242042000671},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.567916579161295e-08,"output_power":4.432200298910148e-08,"reference_active":true,"delta_db":-1.7080824936408876}],"active_band_count":3,"active_band_mae_db":1.3627516512719444}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.308347225189209,"predicted_mos":3.308347225189209,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"2a26a697ab6e0fb492965919bea9a52eba8b19d86c8141971d5a17be41bbfb7a","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0467362030012737},"started_at":"2026-10-13T05:10:12.885091+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002690958999664872,"native_input_samples":144000,"native_input_sha256":"36f80382f9af4ee921a50efb1dfe2cc01979e856582819bd245725a8544f3d76","output_sha256":"b6ca02ce360711a7ec5084c5bf65cc79a82a3ddd8327f154ea93f572de30e73d","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9002709209999011,"synthesis_s":0.1721792230000574,"total_s":1.0724501439999585,"rtf":0.17874169066665976,"output_peak":0.46458542346954346,"output_rms":0.0739804008540383,"finished_at":"2026-10-13T05:10:13.969484+00:00","analysis_stages":{"external_f0_s":0.16271227299876045},"evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"b6ca02ce360711a7ec5084c5bf65cc79a82a3ddd8327f154ea93f572de30e73d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.614574715942926,"output_peak":0.45989857502528225},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":3.900114669727084,"spectral_convergence":0.13552913812015804,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.772109883384744,"spectral_convergence":0.13729318696194473,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.2069310807873315,"spectral_convergence":0.16769114159608714,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0014645134106370877,"output_power":0.0014820614942702291,"delta_db":0.05172871035787587,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024771538682985465,"output_power":0.0023906993712139446,"delta_db":-0.1542801628579832,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013427666217707404,"output_power":0.0013039723169749131,"delta_db":-0.12732165578885646,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024269361162520585,"output_power":0.00024178945859313519,"delta_db":-0.016209818366143297,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4425202041437504e-05,"output_power":3.9932942916392925e-05,"delta_db":-0.46298094272124307,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.4446805694883255e-05,"output_power":2.6259461850999916e-05,"delta_db":-1.1786313369869696,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.33185877117984525,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09498987185329505,"pitch":null,"output_rms_dbfs":-22.608459309412034,"output_peak":0.45989857502528225,"reference_rms_dbfs":-22.513469437558737,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":3.9597185446330534},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9726914498050532,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07487322298437474,"rms_dbfs":-22.513469437558737,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.708326506096444,"mean":-3.738085428873698e-05,"zero_energy":false},"output":{"rms":0.0740588620343279,"rms_dbfs":-22.608459309412034,"peak":0.45989857502528225,"peak_dbfs":-6.746758721309661,"crest_factor_db":15.861700588102373,"mean":-2.1631062013798275e-05,"zero_energy":false},"rms_difference_db":-0.0949898718532971},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":5.464963021841709e-07,"output_rms_dbfs":-62.624127723171014,"output_zero_energy":false,"output_peak":0.010879372149969683,"output_energy_fraction":1.2454983933046436e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":345,"both_estimated_voiced_frames":335,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9279778393351801,"absolute_difference_cents_median":0.9597606345717576,"absolute_difference_cents_p95":8.585750169635336,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002110151907366841,"log_rms_mae_db":0.46350375308335307,"rms_envelope_nrmse":0.05073355837270277,"envelope_correlation":0.9967693408284117,"rms_gain_error_db":-0.09217146365630771,"reference_mean_power":0.005604891233518836,"output_mean_power":0.005487190562307372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2250047113785757e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011643562088722453,"output_power":0.0011867611692476542,"reference_active":true,"delta_db":0.08277464414777497},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011188245530999946,"output_power":0.0001064683079711196,"reference_active":true,"delta_db":-0.21541636415347523},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8450792038001933e-05,"output_power":2.0179947384831085e-05,"reference_active":true,"delta_db":-1.4917433160201257}],"active_band_count":3,"active_band_mae_db":0.596644774773792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020713167740975364,"log_rms_mae_db":0.29014035544227695,"rms_envelope_nrmse":0.029962425841421735,"envelope_correlation":0.9988580941980808,"rms_gain_error_db":-0.08242994444447016,"reference_mean_power":0.005603505754865554,"output_mean_power":0.005498153086516889},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.0483171696395765e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010458478077031374,"output_power":0.0010697262227322847,"reference_active":true,"delta_db":0.09804151633233331},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.131979407845486e-05,"output_power":5.769303706562832e-05,"reference_active":true,"delta_db":-0.26477286149879903},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.629293286100356e-06,"output_power":1.2053728124951915e-06,"reference_active":true,"delta_db":-1.308778762851745}],"active_band_count":3,"active_band_mae_db":0.5571977135609592}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015831513319989338,"log_rms_mae_db":0.8181544797444321,"rms_envelope_nrmse":0.06580806564171239,"envelope_correlation":0.9963161257743757,"rms_gain_error_db":-0.1328621600696901,"reference_mean_power":0.0022309508329157127,"output_mean_power":0.0021637335699340884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0339679622838414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046453953265101344,"output_power":0.00044781539402720107,"reference_active":true,"delta_db":-0.15923659943422988},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.019812862224435e-05,"output_power":6.492630838664418e-05,"reference_active":true,"delta_db":-0.3390482440258446},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5452757374862184e-05,"output_power":1.0985939361511356e-05,"reference_active":true,"delta_db":-1.481687882164702}],"active_band_count":3,"active_band_mae_db":0.6599909085415921}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014375880677218868,"log_rms_mae_db":0.498667349740906,"rms_envelope_nrmse":0.03621376073461304,"envelope_correlation":0.998901334823638,"rms_gain_error_db":-0.11458123343120864,"reference_mean_power":0.0022214967634666704,"output_mean_power":0.002163652752371193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.581275628581243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004040718027646314,"output_power":0.0003920385592044512,"reference_active":true,"delta_db":-0.13129760809333121},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.035103681693411e-05,"output_power":3.6000917295660526e-05,"reference_active":true,"delta_db":-0.49541131727716736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1841518228550596e-06,"output_power":8.993171760421033e-07,"reference_active":true,"delta_db":-1.1949449978586486}],"active_band_count":3,"active_band_mae_db":0.6072179744097158}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":0.8251189777296443,"rms_envelope_nrmse":0.07382291256453882,"envelope_correlation":0.9961279755404563,"rms_gain_error_db":-0.10078955548879805,"reference_mean_power":0.0002507880581966276,"output_mean_power":0.00024503487319316717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4212041991971411e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.17102478326245e-05,"output_power":7.003722710743441e-05,"reference_active":true,"delta_db":-0.10252280141285303},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.275733290040069e-06,"output_power":6.868127570574763e-06,"reference_active":true,"delta_db":-0.25038417272901603},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.07244494044588e-06,"output_power":8.243621128639594e-07,"reference_active":true,"delta_db":-1.142569809774642}],"active_band_count":3,"active_band_mae_db":0.49849226130550367}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":0.513933707124839,"rms_envelope_nrmse":0.04450919869546912,"envelope_correlation":0.9986242527705014,"rms_gain_error_db":-0.09837473244460666,"reference_mean_power":0.0002518169764916093,"output_mean_power":0.0002461770323053272},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3611130645010087e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.48897476225386e-05,"output_power":6.363106443675055e-05,"reference_active":true,"delta_db":-0.08506896762324954},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.71133538995873e-06,"output_power":3.5259219692014316e-06,"reference_active":true,"delta_db":-0.22257506045794698},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.48681833738951e-08,"output_power":6.051534615217292e-08,"reference_active":true,"delta_db":-1.4687938425964033}],"active_band_count":3,"active_band_mae_db":0.5921459568925332}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":0.9859315042305276,"rms_envelope_nrmse":0.13486638567180148,"envelope_correlation":0.9889215554673703,"rms_gain_error_db":-0.6230404805835145,"reference_mean_power":2.817379488191862e-05,"output_mean_power":2.4408511760618628e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.938019558986415e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.123435668533647e-06,"output_power":5.280526188864698e-06,"reference_active":true,"delta_db":-0.6431795874583351},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7222454435376484e-06,"output_power":2.1272725241608713e-06,"reference_active":true,"delta_db":-1.0710414877705667},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.909737577930933e-07,"output_power":3.857879285194983e-07,"reference_active":true,"delta_db":-2.531129195489213}],"active_band_count":3,"active_band_mae_db":1.4151167569060383}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":0.6977294135182353,"rms_envelope_nrmse":0.09101138007358767,"envelope_correlation":0.9968334659816243,"rms_gain_error_db":-0.6227143173481837,"reference_mean_power":2.8310170166312666e-05,"output_mean_power":2.4528503255072838e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8087824895787787e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.951898475340427e-06,"output_power":5.046175514846909e-06,"reference_active":true,"delta_db":-0.7169316264275128},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2501096733423978e-06,"output_power":9.509965921772457e-07,"reference_active":true,"delta_db":-1.1876915507896337},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.567916579161295e-08,"output_power":3.8176356412795245e-08,"reference_active":true,"delta_db":-2.356331514943109}],"active_band_count":3,"active_band_mae_db":1.4203182307200855}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1836-0014","method":"magphase","status":"measured","mos":{"status":"measured","value":2.941920042037964,"predicted_mos":2.941920042037964,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287127,"input_sample_rate":48000,"input_seconds":5.9818125,"input_float32_sha256":"f663d085c4c139d68806e0a79e406e4daf01e837d77aa5046f754586f85fdf32","internal_samples":95709,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.452069978999134},"started_at":"2026-10-13T05:13:16.977344+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008515384999554954,"native_input_samples":288000,"native_input_sha256":"43fceb6a4bacded7d06d028a0150ebe7252120c2293e5388ca3423e93ceef3a0","output_sha256":"eb865e5a4f7ca6a9fe594a67ec7c78576c252d145c5c193fe0e30f8c08ca0414","output_samples":287127,"length_delta_samples":-873,"analysis_s":3.3257158160013205,"synthesis_s":1.056675275998714,"total_s":4.382391092000034,"rtf":0.7303985153333391,"output_peak":0.4851140964331441,"output_rms":0.07191280653854694,"finished_at":"2026-10-13T05:13:21.379345+00:00","evaluation_input":{"reference":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"eb865e5a4f7ca6a9fe594a67ec7c78576c252d145c5c193fe0e30f8c08ca0414","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95709,"length_delta_samples":-291,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.86141294562146,"output_peak":0.47381204624193424},"common_prefix":{"samples":95709,"sample_rate":16000,"reference_coverage":0.99696875,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95709,"output_samples":95709,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.156067056001531,"spectral_convergence":0.44521908561050433,"spectral_convergence_status":"ok","amplitude_floor":1.7246496227499867e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.951905765155866,"spectral_convergence":0.16667078424461582,"spectral_convergence_status":"ok","amplitude_floor":1.4526318235172953e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.940109963888985,"spectral_convergence":0.16973167108842355,"spectral_convergence_status":"ok","amplitude_floor":1.3962213853324247e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001468748041948761,"output_power":0.0013052533247521494,"delta_db":-0.5125249251684212,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024483019405024597,"output_power":0.00227109339155899,"delta_db":-0.3262998321474592,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013452849468903224,"output_power":0.0012762072043292403,"delta_db":-0.22893090736325494,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024331616785465762,"output_power":0.00023852006768237247,"delta_db":-0.08646043948156049,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.4555586523099467e-05,"output_power":4.7276281568918944e-05,"delta_db":0.2574114553403398,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.455637570376525e-05,"output_power":3.603432720800548e-05,"delta_db":0.18188230724233467,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.26558497779056167,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3314604691655927,"pitch":null,"output_rms_dbfs":-22.86141294562146,"output_peak":0.47381204624193424,"reference_rms_dbfs":-22.529952476455865,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.349360928348794},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95709,"output_samples":95709,"intelligibility":{"estoi":{"value":0.9607446293343511,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07473127229897178,"rms_dbfs":-22.529952476455865,"peak":0.456817626953125,"peak_dbfs":-6.805142931462294,"crest_factor_db":15.724809544993574,"mean":-6.638495538886102e-05,"zero_energy":false},"output":{"rms":0.07193319537668043,"rms_dbfs":-22.86141294562146,"peak":0.47381204624193424,"peak_dbfs":-6.487878038584425,"crest_factor_db":16.373534907037033,"mean":0.0013874517055290734,"zero_energy":false},"rms_difference_db":-0.33146046916559513},"full_scale":{"reference":{"sample_count":95709,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95709,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":29,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.4527410721639175e-06,"threshold_dbfs":-53.513725584506055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":3.921421399960915e-07,"output_mean_power":3.8405841582750416e-07,"output_rms_dbfs":-64.15602713804184,"output_zero_energy":false,"output_peak":0.009591530278390875,"output_energy_fraction":9.306085712050098e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":361,"output_estimated_voiced_frames":364,"both_estimated_voiced_frames":350,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9695290858725761,"absolute_difference_cents_median":4.978089276317138,"absolute_difference_cents_p95":18.9150880067002,"difference_over_600_cents_frames":0,"uncovered_tail_samples":29,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95709,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1195,"rms_floor":0.0002110151907366841,"log_rms_mae_db":1.1206115781340653,"rms_envelope_nrmse":0.10658001310444806,"envelope_correlation":0.9865472645572697,"rms_gain_error_db":-0.32905280033071926,"reference_mean_power":0.005582935511320634,"output_mean_power":0.005175559804459413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":2.2256189148998026e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001161337080331329,"output_power":0.0010664406646668594,"reference_active":true,"delta_db":-0.37021595955235803},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011137877729930416,"output_power":0.0001037522777488034,"reference_active":true,"delta_db":-0.3080480608590785},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.820972573797712e-05,"output_power":2.92896048927701e-05,"reference_active":true,"delta_db":0.1631464937390315}],"active_band_count":3,"active_band_mae_db":0.28047017138348934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1189,"rms_floor":0.00020713167740975364,"log_rms_mae_db":0.7532277724560913,"rms_envelope_nrmse":0.06830695878176558,"envelope_correlation":0.9951222124178842,"rms_gain_error_db":-0.3288311302253132,"reference_mean_power":0.005588474032804485,"output_mean_power":0.005180958627817944},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":2.049429005229749e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001042853703702277,"output_power":0.0009527116316408594,"reference_active":true,"delta_db":-0.39261920334678857},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0750063381167894e-05,"output_power":5.651693941511491e-05,"reference_active":true,"delta_db":-0.3136810008751463},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6375010916447458e-06,"output_power":1.5106450703282043e-06,"reference_active":true,"delta_db":-0.35019160390403936}],"active_band_count":3,"active_band_mae_db":0.3521639360419914}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1195,"rms_floor":0.00015831513319989338,"log_rms_mae_db":1.593726716071014,"rms_envelope_nrmse":0.11564615773939729,"envelope_correlation":0.9883739528522694,"rms_gain_error_db":-0.21105269992194903,"reference_mean_power":0.0022093694930249096,"output_mean_power":0.0021045685830995637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.0268756900730488e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004615547420959083,"output_power":0.0004456432608803988,"reference_active":true,"delta_db":-0.1523587393694503},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.997034329679248e-05,"output_power":6.642263168818754e-05,"reference_active":true,"delta_db":-0.22597926200033058},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5155731130608956e-05,"output_power":1.4355353746272328e-05,"reference_active":true,"delta_db":-0.2356299319581478}],"active_band_count":3,"active_band_mae_db":0.20465597777597622}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1189,"rms_floor":0.00014375880677218868,"log_rms_mae_db":1.1106677681767017,"rms_envelope_nrmse":0.07307141361293895,"envelope_correlation":0.9953085688330856,"rms_gain_error_db":-0.2085597939656058,"reference_mean_power":0.0022111660725955516,"output_mean_power":0.0021074893210840384},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":9.522354493434173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004013186359678699,"output_power":0.00038841927186596885,"reference_active":true,"delta_db":-0.14188557676460073},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.013036831510902e-05,"output_power":3.792327350114335e-05,"reference_active":true,"delta_db":-0.24567327518120113},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.18256964263377e-06,"output_power":1.0751138579475409e-06,"reference_active":true,"delta_db":-0.4137226632951573}],"active_band_count":3,"active_band_mae_db":0.26709383841365303}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1195,"rms_floor":7.952734804355277e-05,"log_rms_mae_db":1.300828253751822,"rms_envelope_nrmse":0.13508712890161056,"envelope_correlation":0.9866799008544447,"rms_gain_error_db":-0.051582077649583793,"reference_mean_power":0.00025130280621448906,"output_mean_power":0.00024833568487742514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.40508983074906e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.083460437799832e-05,"output_power":6.86973533295259e-05,"reference_active":true,"delta_db":-0.1330546712286861},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.171087583325197e-06,"output_power":7.156976790092325e-06,"reference_active":true,"delta_db":-0.008554179347419259},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.043016996065392e-06,"output_power":1.1586813631347617e-06,"reference_active":true,"delta_db":0.4567263623494188}],"active_band_count":3,"active_band_mae_db":0.19944507097517472}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1189,"rms_floor":7.039214481073796e-05,"log_rms_mae_db":0.8900684868351428,"rms_envelope_nrmse":0.08753239789250546,"envelope_correlation":0.9941495806992171,"rms_gain_error_db":-0.05158642621424381,"reference_mean_power":0.0002524840468222845,"output_mean_power":0.0002495027288028658},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":1.346307886480931e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.406600675037926e-05,"output_power":6.202181252409799e-05,"reference_active":true,"delta_db":-0.14083201256626982},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6387961501485645e-06,"output_power":3.6109897979384913e-06,"reference_active":true,"delta_db":-0.03331465130270955},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.438703539766785e-08,"output_power":9.784711836956432e-08,"reference_active":true,"delta_db":0.6427231011802723}],"active_band_count":3,"active_band_mae_db":0.2722899216830839}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1195,"rms_floor":2.8930089256612984e-05,"log_rms_mae_db":1.5177458932696926,"rms_envelope_nrmse":0.21335287416841348,"envelope_correlation":0.9703913210686581,"rms_gain_error_db":0.5525775275704278,"reference_mean_power":2.8260956195596447e-05,"output_mean_power":3.209553391585328e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.934711217096051e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.095355809564783e-06,"output_power":6.656309109217868e-06,"reference_active":true,"delta_db":0.38234419600730046},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.722565974572538e-06,"output_power":2.832067493890699e-06,"reference_active":true,"delta_db":0.17125186660908423},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.897697123798998e-07,"output_power":1.2071257164002215e-06,"reference_active":true,"delta_db":2.4304838147169794}],"active_band_count":3,"active_band_mae_db":0.9946932924444547}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":29,"metrics":{"frame_count":1189,"rms_floor":2.3848385775447814e-05,"log_rms_mae_db":0.9838267796582075,"rms_envelope_nrmse":0.13384815049452686,"envelope_correlation":0.9892651648051315,"rms_gain_error_db":0.552548138457325,"reference_mean_power":2.8401295235126688e-05,"output_mean_power":3.225469653642063e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":1.8061664931435912e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9208716979882304e-06,"output_power":6.524224459504112e-06,"reference_active":true,"delta_db":0.42143243835359},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2516682807118814e-06,"output_power":1.4098147519252506e-06,"reference_active":true,"delta_db":0.516728039469654},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.603284167552047e-08,"output_power":8.854994081269144e-08,"reference_active":true,"delta_db":1.2742828801748887}],"active_band_count":3,"active_band_mae_db":0.737481119332711}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"hnm","status":"measured","mos":{"status":"measured","value":3.1821227073669434,"predicted_mos":3.1821227073669434,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1935679a1231e084e52a6bb388df02d62da3636d01298dc6e33131770b2702e5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6254658950001613},"started_at":"2026-10-13T04:38:29.406599+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6510000023117755e-05,"native_input_samples":96000,"native_input_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","output_sha256":"a0b88d38f451978feeb7dbf9655fcbc589204f02c58ffdbc7b4318cc3074f2a7","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0795820160001313,"synthesis_s":0.47804502899998624,"total_s":1.5576270450001175,"rtf":0.25960450750001957,"output_peak":0.38081213498468236,"output_rms":0.07438324619639183,"finished_at":"2026-10-13T04:38:30.969919+00:00","evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"a0b88d38f451978feeb7dbf9655fcbc589204f02c58ffdbc7b4318cc3074f2a7","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.593900161712391,"spectral_convergence":0.12420178914577755,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.550369058357183,"spectral_convergence":0.11881206545017373,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.86283415353385,"spectral_convergence":0.1159936052015742,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0014870914119580784,"delta_db":-0.041142602849087474,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.002711646056101496,"delta_db":-0.0555963463813539,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0011076294972392265,"delta_db":-0.07776981180383982,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.0001583485290841634,"delta_db":-0.17983205444059086,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":3.9686573003234067e-05,"delta_db":0.03788907699294453,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":2.8465247326842094e-05,"delta_db":-0.02713352756584928,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.06989390333894431,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05896455503674526,"pitch":null,"output_rms_dbfs":-22.570497446022053,"output_peak":0.38081213498468236,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.6690344578678085},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.593900161712391,"spectral_convergence":0.12420178914577755,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.550369058357183,"spectral_convergence":0.11881206545017373,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.86283415353385,"spectral_convergence":0.1159936052015742,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0014870914119580784,"delta_db":-0.041142602849087474,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.002711646056101496,"delta_db":-0.0555963463813539,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0011076294972392265,"delta_db":-0.07776981180383982,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.0001583485290841634,"delta_db":-0.17983205444059086,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":3.9686573003234067e-05,"delta_db":0.03788907699294453,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":2.8465247326842094e-05,"delta_db":-0.02713352756584928,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.06989390333894431,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05896455503674526,"pitch":null,"output_rms_dbfs":-22.570497446022053,"output_peak":0.38081213498468236,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.6690344578678085},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9575745334038547,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0748899180673492,"rms_dbfs":-22.511532890985304,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.637965780126057,"mean":-3.1950950622558594e-05,"zero_energy":false},"output":{"rms":0.07438324619639183,"rms_dbfs":-22.570497446022053,"peak":0.38081213498468236,"peak_dbfs":-8.385784416207407,"crest_factor_db":14.184713029814642,"mean":-5.2017483830694374e-05,"zero_energy":false},"rms_difference_db":-0.05896455503674858},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":2.5114844002316823e-06,"output_rms_dbfs":-56.00069515061406,"output_zero_energy":false,"output_peak":0.06543491606173196,"output_energy_fraction":6.960121268011358e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":440,"output_estimated_voiced_frames":434,"both_estimated_voiced_frames":430,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9772727272727273,"absolute_difference_cents_median":1.1744625970314873,"absolute_difference_cents_p95":9.380573562690943,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018834661138467316,"log_rms_mae_db":0.5762207958016737,"rms_envelope_nrmse":0.05173989013410462,"envelope_correlation":0.9962206725805041,"rms_gain_error_db":-0.0564453410137717,"reference_mean_power":0.005606081823103195,"output_mean_power":0.005533690917149483},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7511238572638392e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009260227506744139,"output_power":0.0009220666626783199,"reference_active":true,"delta_db":-0.018593362404758933},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010953814679795108,"output_power":0.00010168563400255931,"reference_active":true,"delta_db":-0.3230578835121749},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.583845526426034e-05,"output_power":2.2532242496478434e-05,"reference_active":true,"delta_db":-0.5946212952552943}],"active_band_count":3,"active_band_mae_db":0.31209084705740936}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015988223985812614,"log_rms_mae_db":0.3355106510433134,"rms_envelope_nrmse":0.028604110669709717,"envelope_correlation":0.9987640421289028,"rms_gain_error_db":-0.053724221279609206,"reference_mean_power":0.00558793457857638,"output_mean_power":0.0055192350607563105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5615527006220852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008311543470725633,"output_power":0.0008320337859906332,"reference_active":true,"delta_db":0.00459281180873986},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4975058641843e-05,"output_power":5.0296963877775305e-05,"reference_active":true,"delta_db":-0.38623931167583403},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.054233222620555e-06,"output_power":1.5430233336021302e-06,"reference_active":true,"delta_db":-1.2427725516567258}],"active_band_count":3,"active_band_mae_db":0.5445348917137666}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016200114302549476,"log_rms_mae_db":1.2639289402213463,"rms_envelope_nrmse":0.07292644505562572,"envelope_correlation":0.9949075460325472,"rms_gain_error_db":-0.04713079340782041,"reference_mean_power":0.0020825301889685474,"output_mean_power":0.002060052206054166},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.598844168109501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038009245216565226,"output_power":0.0003750972251262812,"reference_active":true,"delta_db":-0.05745394153742933},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.129961338774397e-05,"output_power":5.899548549523561e-05,"reference_active":true,"delta_db":-0.16638956016113735},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.439999050166804e-05,"output_power":1.3071388741639648e-05,"reference_active":true,"delta_db":-0.4204047491493023}],"active_band_count":3,"active_band_mae_db":0.21474941694928965}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012179662863477574,"log_rms_mae_db":0.7988470437509892,"rms_envelope_nrmse":0.03885580918044455,"envelope_correlation":0.9984534940477735,"rms_gain_error_db":-0.04472988259095946,"reference_mean_power":0.0020746475669658963,"output_mean_power":0.002053389527773716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.889619852823584e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035505972043036534,"output_power":0.0003518380999527892,"reference_active":true,"delta_db":-0.03958540096203603},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2260222284985983e-05,"output_power":3.135000175322447e-05,"reference_active":true,"delta_db":-0.12429786060195194},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1886983564224354e-06,"output_power":9.488330827623326e-07,"reference_active":true,"delta_db":-0.9788184348259368}],"active_band_count":3,"active_band_mae_db":0.38090056546330825}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":1.3314242501789166,"rms_envelope_nrmse":0.10094780951686912,"envelope_correlation":0.9921976942325926,"rms_gain_error_db":-0.14827877744077742,"reference_mean_power":0.00015543877011727882,"output_mean_power":0.00015022128532674975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4080839081981733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.101806003601559e-05,"output_power":6.791561309656893e-05,"reference_active":true,"delta_db":-0.193991791625687},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0132703192354953e-06,"output_power":1.9401285355446927e-06,"reference_active":true,"delta_db":-0.16071587622359296},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.589135119212959e-07,"output_power":5.298548829099247e-07,"reference_active":true,"delta_db":-1.560353442459141}],"active_band_count":3,"active_band_mae_db":0.6383537034361403}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":0.8028794451456672,"rms_envelope_nrmse":0.057444749492949095,"envelope_correlation":0.9975057346376166,"rms_gain_error_db":-0.14892396806714442,"reference_mean_power":0.00015383740434286896,"output_mean_power":0.000148651585860431},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3572636139953195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.549215276795425e-05,"output_power":6.310562149879294e-05,"reference_active":true,"delta_db":-0.16121217930081033},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.982536907090337e-07,"output_power":9.783941167636778e-07,"reference_active":true,"delta_db":-0.08727091972218132},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.830181704119615e-08,"output_power":3.173900839501437e-08,"reference_active":true,"delta_db":-1.8237011436051938}],"active_band_count":3,"active_band_mae_db":0.6907280808760619}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":1.4028459742305355,"rms_envelope_nrmse":0.16288471300573024,"envelope_correlation":0.981222443609929,"rms_gain_error_db":0.045458175621184434,"reference_mean_power":2.2945322907755718e-05,"output_mean_power":2.318675597734702e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.468236349213916e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8725122830509825e-06,"output_power":4.007765958643509e-06,"reference_active":true,"delta_db":0.14909548168361275},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.737803238621211e-07,"output_power":5.448638073955926e-07,"reference_active":true,"delta_db":-0.22457690460762614},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4333843373978682e-07,"output_power":1.5371736717764963e-07,"reference_active":true,"delta_db":0.3036028277759121}],"active_band_count":3,"active_band_mae_db":0.2257584046890503}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":0.8011268401921154,"rms_envelope_nrmse":0.07513308255075814,"envelope_correlation":0.9958324602159857,"rms_gain_error_db":0.04649068662911468,"reference_mean_power":2.2958960928334712e-05,"output_mean_power":2.3206053953244956e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.967813463196928e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.583722018207827e-06,"output_power":3.7428307967484933e-06,"reference_active":true,"delta_db":0.1886587929447056},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.664756741318266e-07,"output_power":2.540082504745553e-07,"reference_active":true,"delta_db":-0.20809746339518537},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1431200548978786e-08,"output_power":9.899047285658432e-09,"reference_active":true,"delta_db":-0.624984452896787}],"active_band_count":3,"active_band_mae_db":0.34058023641222596}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.212294578552246,"predicted_mos":3.212294578552246,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2c399096c7b27282530873419877ab1d3a55e96e65ec70b6f994c3387dead98f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6465507539996906},"started_at":"2026-10-13T04:41:27.025883+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.05539997498272e-05,"native_input_samples":96000,"native_input_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","output_sha256":"fae16fd6380138add06a92dc912bc5228afea9a5eee2d5153d4431a8d0fd240b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3701247100007095,"synthesis_s":0.09861388799981796,"total_s":1.4687385980005274,"rtf":0.24478976633342123,"output_peak":0.4878265400222446,"output_rms":0.08434485447340992,"finished_at":"2026-10-13T04:41:28.501530+00:00","evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"fae16fd6380138add06a92dc912bc5228afea9a5eee2d5153d4431a8d0fd240b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.283250306091446,"spectral_convergence":0.4441232189337805,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.53299317363853,"spectral_convergence":0.21593292689035223,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.132542377363999,"spectral_convergence":0.24185850093761666,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0018005790819343205,"delta_db":0.7896027462035221,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.003443006223887433,"delta_db":0.9814517153843039,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0015394547065884618,"delta_db":1.3519542164694922,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.0002450964041396012,"delta_db":1.7173970070956723,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":5.3343653433116976e-05,"delta_db":1.3222806577259727,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":3.25744061577623e-05,"delta_db":0.5584819273627296,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1201947117069488,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0327047558741436,"pitch":null,"output_rms_dbfs":-21.478828135111165,"output_peak":0.4878265400222446,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.982928619031325},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.283250306091446,"spectral_convergence":0.4441232189337805,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.53299317363853,"spectral_convergence":0.21593292689035223,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.132542377363999,"spectral_convergence":0.24185850093761666,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0018005790819343205,"delta_db":0.7896027462035221,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.003443006223887433,"delta_db":0.9814517153843039,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0015394547065884618,"delta_db":1.3519542164694922,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.0002450964041396012,"delta_db":1.7173970070956723,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":5.3343653433116976e-05,"delta_db":1.3222806577259727,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":3.25744061577623e-05,"delta_db":0.5584819273627296,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1201947117069488,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0327047558741436,"pitch":null,"output_rms_dbfs":-21.478828135111165,"output_peak":0.4878265400222446,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.982928619031325},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.965459662607744,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0748899180673492,"rms_dbfs":-22.511532890985304,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.637965780126057,"mean":-3.1950950622558594e-05,"zero_energy":false},"output":{"rms":0.08434485447340992,"rms_dbfs":-21.478828135111165,"peak":0.4878265400222446,"peak_dbfs":-6.234691514981811,"crest_factor_db":15.244136620129352,"mean":3.1452482294882935e-05,"zero_energy":false},"rms_difference_db":1.0327047558741391},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":2.4667805834916215e-06,"output_rms_dbfs":-56.07869478648374,"output_zero_energy":false,"output_peak":0.04247122575486543,"output_energy_fraction":5.3167949547935785e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":440,"output_estimated_voiced_frames":442,"both_estimated_voiced_frames":432,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9818181818181818,"absolute_difference_cents_median":6.057115881485459,"absolute_difference_cents_p95":24.39177576924487,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018834661138467316,"log_rms_mae_db":1.4905871621428313,"rms_envelope_nrmse":0.16118561420993438,"envelope_correlation":0.9874757032999603,"rms_gain_error_db":1.0371960433605445,"reference_mean_power":0.005606081823103195,"output_mean_power":0.007118345060344812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7511238572638392e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009260227506744139,"output_power":0.0011676154067848266,"reference_active":true,"delta_db":1.00678160277494},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010953814679795108,"output_power":0.00013647306459252352,"reference_active":true,"delta_db":0.9548155490825521},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.583845526426034e-05,"output_power":3.650590265333703e-05,"reference_active":true,"delta_db":1.5009654529707017}],"active_band_count":3,"active_band_mae_db":1.1541875349427313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015988223985812614,"log_rms_mae_db":1.2278879449756253,"rms_envelope_nrmse":0.14576600629741793,"envelope_correlation":0.9929763972052879,"rms_gain_error_db":1.0415600368053246,"reference_mean_power":0.00558793457857638,"output_mean_power":0.007102435797355505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5615527006220852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008311543470725633,"output_power":0.001053900446756288,"reference_active":true,"delta_db":1.0311790796003164},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4975058641843e-05,"output_power":6.974560350275599e-05,"reference_active":true,"delta_db":1.0335113524357282},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.054233222620555e-06,"output_power":2.169728698876679e-06,"reference_active":true,"delta_db":0.23755684731646173}],"active_band_count":3,"active_band_mae_db":0.7674157597841688}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016200114302549476,"log_rms_mae_db":1.6084401440081735,"rms_envelope_nrmse":0.21901970470786913,"envelope_correlation":0.9823986874643209,"rms_gain_error_db":1.308457310679529,"reference_mean_power":0.0020825301889685474,"output_mean_power":0.002814731911335113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.598844168109501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038009245216565226,"output_power":0.00047552741175973417,"reference_active":true,"delta_db":0.9728631136766931},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.129961338774397e-05,"output_power":9.2504368361442e-05,"reference_active":true,"delta_db":1.7870450657848138},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.439999050166804e-05,"output_power":2.1288398048028187e-05,"reference_active":true,"delta_db":1.697807763759846}],"active_band_count":3,"active_band_mae_db":1.4859053144071177}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012179662863477574,"log_rms_mae_db":1.3333796665025357,"rms_envelope_nrmse":0.20048890154107213,"envelope_correlation":0.9878441141452303,"rms_gain_error_db":1.3125278751154592,"reference_mean_power":0.0020746475669658963,"output_mean_power":0.0028067072636071846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.889619852823584e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035505972043036534,"output_power":0.00044879687095240227,"reference_active":true,"delta_db":1.0174841354376607},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2260222284985983e-05,"output_power":4.892940832848191e-05,"reference_active":true,"delta_db":1.809026086364127},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1886983564224354e-06,"output_power":1.5710444088142147e-06,"reference_active":true,"delta_db":1.2111679922319547}],"active_band_count":3,"active_band_mae_db":1.3458927380112475}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":1.47314664125604,"rms_envelope_nrmse":0.2379885710617256,"envelope_correlation":0.9895907193163908,"rms_gain_error_db":1.5777673206742642,"reference_mean_power":0.00015543877011727882,"output_mean_power":0.0002235301365360334},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4080839081981733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.101806003601559e-05,"output_power":0.00011433352917680463,"reference_active":true,"delta_db":2.0680480450509804},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0132703192354953e-06,"output_power":3.22423321760719e-06,"reference_active":true,"delta_db":2.0452435700754474},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.589135119212959e-07,"output_power":9.606164116459415e-07,"reference_active":true,"delta_db":1.0235771686501935}],"active_band_count":3,"active_band_mae_db":1.7122895945922074}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":1.1701060240545609,"rms_envelope_nrmse":0.22632483120320931,"envelope_correlation":0.9933210965239992,"rms_gain_error_db":1.5917421792320374,"reference_mean_power":0.00015383740434286896,"output_mean_power":0.00022194029628975828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3572636139953195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.549215276795425e-05,"output_power":0.00010618587090687774,"reference_active":true,"delta_db":2.098774671670521},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.982536907090337e-07,"output_power":1.602776840525024e-06,"reference_active":true,"delta_db":2.0563213389907578},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.830181704119615e-08,"output_power":4.679633117780968e-08,"reference_active":true,"delta_db":-0.13751662743621187}],"active_band_count":3,"active_band_mae_db":1.4308708793658302}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":1.3726473978869924,"rms_envelope_nrmse":0.21980500485516466,"envelope_correlation":0.9726722159475252,"rms_gain_error_db":0.6763832876094032,"reference_mean_power":2.2945322907755718e-05,"output_mean_power":2.681220323285563e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.468236349213916e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8725122830509825e-06,"output_power":4.11306268230944e-06,"reference_active":true,"delta_db":0.26172524173699896},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.737803238621211e-07,"output_power":5.953011763352444e-07,"reference_active":true,"delta_db":0.1599108944191595},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4333843373978682e-07,"output_power":1.5834978750339518e-07,"reference_active":true,"delta_db":0.4325483028912222}],"active_band_count":3,"active_band_mae_db":0.2847281463491269}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":0.9842638374256499,"rms_envelope_nrmse":0.15565090801453357,"envelope_correlation":0.987849476332549,"rms_gain_error_db":0.6749476046468684,"reference_mean_power":2.2958960928334712e-05,"output_mean_power":2.681927227975857e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.967813463196928e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.583722018207827e-06,"output_power":3.8343895539089136e-06,"reference_active":true,"delta_db":0.2936191784420122},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.664756741318266e-07,"output_power":2.5883081042361306e-07,"reference_active":true,"delta_db":-0.12641597410065789},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1431200548978786e-08,"output_power":1.0520216952546603e-08,"reference_active":true,"delta_db":-0.36067147957894363}],"active_band_count":3,"active_band_mae_db":0.2602355440405379}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9946846961975098,"predicted_mos":3.9946846961975098,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6dd2a226d88984f1206bf51bcb4b58937847f0a60013d09c70f7aff028c69033","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.526035734999823},"started_at":"2026-10-13T04:47:56.099211+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.977699932671385e-05,"native_input_samples":96000,"native_input_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","output_sha256":"7f98b0506c625e55348933b2591fff0f7f4bb5ff9abb413befa6141ea64ca1bc","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5055367940003634,"synthesis_s":5.74872930499987,"total_s":7.254266099000233,"rtf":1.2090443498333723,"output_peak":0.3584807217121124,"output_rms":0.07068575324504119,"finished_at":"2026-10-13T04:48:03.360763+00:00","evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"7f98b0506c625e55348933b2591fff0f7f4bb5ff9abb413befa6141ea64ca1bc","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.773121692992785,"spectral_convergence":0.16746181252837664,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.301666190012755,"spectral_convergence":0.1564651316164044,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.502806795002777,"spectral_convergence":0.16248179899691584,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0013299066584439422,"delta_db":-0.526307654336216,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.0024115441274518653,"delta_db":-0.5649742300241588,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0010461852637466606,"delta_db":-0.3256289569577443,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.00015148623965591072,"delta_db":-0.37224053736135065,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":3.2780802600573046e-05,"delta_db":-0.7923510825902969,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":2.4572619919899164e-05,"delta_db":-0.6657685118843248,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5412118288590152,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5018293028213806,"pitch":null,"output_rms_dbfs":-23.013362193806685,"output_peak":0.3584807217121124,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.192531559336106},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.773121692992785,"spectral_convergence":0.16746181252837664,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.301666190012755,"spectral_convergence":0.1564651316164044,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.502806795002777,"spectral_convergence":0.16248179899691584,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0013299066584439422,"delta_db":-0.526307654336216,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.0024115441274518653,"delta_db":-0.5649742300241588,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0010461852637466606,"delta_db":-0.3256289569577443,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.00015148623965591072,"delta_db":-0.37224053736135065,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":3.2780802600573046e-05,"delta_db":-0.7923510825902969,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":2.4572619919899164e-05,"delta_db":-0.6657685118843248,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5412118288590152,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5018293028213806,"pitch":null,"output_rms_dbfs":-23.013362193806685,"output_peak":0.3584807217121124,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.192531559336106},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9381937406344208,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0748899180673492,"rms_dbfs":-22.511532890985304,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.637965780126057,"mean":-3.1950950622558594e-05,"zero_energy":false},"output":{"rms":0.07068575324504119,"rms_dbfs":-23.013362193806685,"peak":0.3584807217121124,"peak_dbfs":-8.910683894995977,"crest_factor_db":14.102678298810709,"mean":4.6804935597528183e-05,"zero_energy":false},"rms_difference_db":-0.5018293028213812},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":3.1455177563349356e-07,"output_rms_dbfs":-65.02307858895279,"output_zero_energy":false,"output_peak":0.005341320764273405,"output_energy_fraction":9.653058484746416e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":440,"output_estimated_voiced_frames":440,"both_estimated_voiced_frames":431,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9795454545454545,"absolute_difference_cents_median":2.4851060084748786,"absolute_difference_cents_p95":11.29338270629431,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018834661138467316,"log_rms_mae_db":0.9731769214617653,"rms_envelope_nrmse":0.0956466428897099,"envelope_correlation":0.9909360982688196,"rms_gain_error_db":-0.5025702273023622,"reference_mean_power":0.005606081823103195,"output_mean_power":0.0049934695907181215},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7511238572638392e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009260227506744139,"output_power":0.0008310132428600055,"reference_active":true,"delta_db":-0.470137119601741},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010953814679795108,"output_power":0.00011059723003740763,"reference_active":true,"delta_db":0.04178860864341865},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.583845526426034e-05,"output_power":2.6308895099240316e-05,"reference_active":true,"delta_db":0.0783606322511832}],"active_band_count":3,"active_band_mae_db":0.19676212016544758}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015988223985812614,"log_rms_mae_db":0.79782358387686,"rms_envelope_nrmse":0.09000487777307005,"envelope_correlation":0.9916666222905286,"rms_gain_error_db":-0.4992421836527305,"reference_mean_power":0.00558793457857638,"output_mean_power":0.00498112103361016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5615527006220852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008311543470725633,"output_power":0.0007482014240522023,"reference_active":true,"delta_db":-0.4566315019646605},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4975058641843e-05,"output_power":5.452855305848793e-05,"reference_active":true,"delta_db":-0.035417276688646},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.054233222620555e-06,"output_power":2.1684986301237595e-06,"reference_active":true,"delta_db":0.2350940345334633}],"active_band_count":3,"active_band_mae_db":0.2423809377289233}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016200114302549476,"log_rms_mae_db":1.2615495404730948,"rms_envelope_nrmse":0.09747340159602226,"envelope_correlation":0.9909253668379272,"rms_gain_error_db":-0.14232030022239067,"reference_mean_power":0.0020825301889685474,"output_mean_power":0.002015390818706745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.598844168109501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038009245216565226,"output_power":0.00038550750565018665,"reference_active":true,"delta_db":0.06143592462599144},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.129961338774397e-05,"output_power":6.684295277088822e-05,"reference_active":true,"delta_db":0.37597891033347275},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.439999050166804e-05,"output_power":1.5455778809962986e-05,"reference_active":true,"delta_db":0.30728688228761747}],"active_band_count":3,"active_band_mae_db":0.24823390574902723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012179662863477574,"log_rms_mae_db":1.0131692365646845,"rms_envelope_nrmse":0.0862792896371146,"envelope_correlation":0.9922898443512297,"rms_gain_error_db":-0.14264743037919853,"reference_mean_power":0.0020746475669658963,"output_mean_power":0.002007611099124548},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.889619852823584e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035505972043036534,"output_power":0.0003620456006804478,"reference_active":true,"delta_db":0.08461867833182592},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2260222284985983e-05,"output_power":3.432316662096959e-05,"reference_active":true,"delta_db":0.26919993084416943},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1886983564224354e-06,"output_power":1.3092017924423115e-06,"reference_active":true,"delta_db":0.4193492901305921}],"active_band_count":3,"active_band_mae_db":0.2577226331021958}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":1.507112465491674,"rms_envelope_nrmse":0.14959497409887973,"envelope_correlation":0.9831322258262979,"rms_gain_error_db":-0.3067260022066303,"reference_mean_power":0.00015543877011727882,"output_mean_power":0.00014483941134860534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4080839081981733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.101806003601559e-05,"output_power":6.0479990931877374e-05,"reference_active":true,"delta_db":-0.6975708735081009},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0132703192354953e-06,"output_power":2.168899052466213e-06,"reference_active":true,"delta_db":0.32337248053966905},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.589135119212959e-07,"output_power":9.648096879047188e-07,"reference_active":true,"delta_db":1.0424937047862979}],"active_band_count":3,"active_band_mae_db":0.6878123529446892}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":1.1994072488455358,"rms_envelope_nrmse":0.12596946810211382,"envelope_correlation":0.9877701744303441,"rms_gain_error_db":-0.328214539668776,"reference_mean_power":0.00015383740434286896,"output_mean_power":0.00014263972392090745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3572636139953195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.549215276795425e-05,"output_power":5.5666192584928174e-05,"reference_active":true,"delta_db":-0.7059774842425816},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.982536907090337e-07,"output_power":1.023147862705518e-06,"reference_active":true,"delta_db":0.10697476844780533},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.830181704119615e-08,"output_power":6.0707408163183e-08,"reference_active":true,"delta_db":0.9927822297840163}],"active_band_count":3,"active_band_mae_db":0.6019114941581344}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":1.5006842765276835,"rms_envelope_nrmse":0.21788961425539768,"envelope_correlation":0.9662412126963316,"rms_gain_error_db":-0.4290903309394836,"reference_mean_power":2.2945322907755718e-05,"output_mean_power":2.0786680497914444e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.468236349213916e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8725122830509825e-06,"output_power":4.1533723694914535e-06,"reference_active":true,"delta_db":0.3040806545834016},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.737803238621211e-07,"output_power":8.565885686536772e-07,"reference_active":true,"delta_db":1.7402662306096106},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4333843373978682e-07,"output_power":2.044675633077462e-07,"reference_active":true,"delta_db":1.5426176680004702}],"active_band_count":3,"active_band_mae_db":1.195654851064494}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":1.1223491248661597,"rms_envelope_nrmse":0.17046999726875822,"envelope_correlation":0.9788179182914928,"rms_gain_error_db":-0.4286088787329151,"reference_mean_power":2.2958960928334712e-05,"output_mean_power":2.080134136236498e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.967813463196928e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.583722018207827e-06,"output_power":3.96264574320631e-06,"reference_active":true,"delta_db":0.436509335668355},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.664756741318266e-07,"output_power":4.234654170694997e-07,"reference_active":true,"delta_db":2.011603792494588},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1431200548978786e-08,"output_power":1.6651822412037803e-08,"reference_active":true,"delta_db":1.6336992650589468}],"active_band_count":3,"active_band_mae_db":1.3606041310739634}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.259925127029419,"predicted_mos":3.259925127029419,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"e708ad47e0fd0a816bf1a0210d86301a28cc6045496baa57818b4a9002c6591e","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.932818158002192},"started_at":"2026-10-13T07:05:36.001207+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008254772999862325,"native_input_samples":264600,"native_input_sha256":"efbac690fe2e000cb64dc788ec5b3facacff5e1e2756d2bb2cbad36bc6d31640","output_sha256":"5c47cfe6f46b2e4cf3e21efc8a5d0f9f034b61730b731bfcf051ab1a4f672530","output_samples":264704,"length_delta_samples":104,"analysis_s":0.14837967099992966,"synthesis_s":4.029633475000082,"total_s":4.178013146000012,"rtf":0.6963355243333353,"output_peak":0.3923090100288391,"output_rms":0.07011622076504025,"finished_at":"2026-10-13T07:05:40.208809+00:00","evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"5c47cfe6f46b2e4cf3e21efc8a5d0f9f034b61730b731bfcf051ab1a4f672530","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.079896118368673,"output_peak":0.388452829106158},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.111625518330916,"spectral_convergence":0.42985281374666756,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.940754609644545,"spectral_convergence":0.29596717934628275,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.77453078866051,"spectral_convergence":0.35854854503017397,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0011006611785645067,"delta_db":-1.3479827710120464,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.0027497351950219583,"delta_db":0.0049823618336437584,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0008980963566691676,"delta_db":-0.9884855950879285,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.00010830239272891908,"delta_db":-1.8295918734847392,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":3.9773512959395955e-05,"delta_db":0.047392605167496236,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":1.904920351995199e-05,"delta_db":-1.7715149259788823,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9983250220941228,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5726859923047644,"pitch":null,"output_rms_dbfs":-23.08421888329007,"output_peak":0.388452829106158,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.942303638878657},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8378259661210923,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0748899180673492,"rms_dbfs":-22.511532890985304,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.637965780126057,"mean":-3.1950950622558594e-05,"zero_energy":false},"output":{"rms":0.0701114672465489,"rms_dbfs":-23.08421888329007,"peak":0.388452829106158,"peak_dbfs":-8.213234224658553,"crest_factor_db":14.870984658631517,"mean":-6.407085359782452e-05,"zero_energy":false},"rms_difference_db":-0.5726859923047662},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":4.2944974123089e-07,"output_rms_dbfs":-63.67087654610997,"output_zero_energy":false,"output_peak":0.02145510277780622,"output_energy_fraction":1.3395866495848532e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":440,"output_estimated_voiced_frames":423,"both_estimated_voiced_frames":416,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9454545454545454,"absolute_difference_cents_median":11.101153714578986,"absolute_difference_cents_p95":47.043717632258016,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018834661138467316,"log_rms_mae_db":1.1143420208165031,"rms_envelope_nrmse":0.1005887129901717,"envelope_correlation":0.991263814088169,"rms_gain_error_db":-0.5711696558328379,"reference_mean_power":0.005606081823103195,"output_mean_power":0.0049152144026681785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7511238572638392e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009260227506744139,"output_power":0.000874095771955502,"reference_active":true,"delta_db":-0.25062637091746953},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010953814679795108,"output_power":9.683078351419228e-05,"reference_active":true,"delta_db":-0.5355194312266133},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.583845526426034e-05,"output_power":2.6990596318307464e-05,"reference_active":true,"delta_db":0.1894593369522701}],"active_band_count":3,"active_band_mae_db":0.32520171303211765}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015988223985812614,"log_rms_mae_db":0.8701991463035398,"rms_envelope_nrmse":0.0814770415945082,"envelope_correlation":0.9960713375450561,"rms_gain_error_db":-0.5714604523801013,"reference_mean_power":0.00558793457857638,"output_mean_power":0.004898975501615284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5615527006220852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008311543470725633,"output_power":0.0007901531709927542,"reference_active":true,"delta_db":-0.2197039332528512},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4975058641843e-05,"output_power":4.8527570171550136e-05,"reference_active":true,"delta_db":-0.5417715503403826},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.054233222620555e-06,"output_power":2.062016446246616e-06,"reference_active":true,"delta_db":0.016423761300211722}],"active_band_count":3,"active_band_mae_db":0.25929974829781516}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016200114302549476,"log_rms_mae_db":2.2124750596643037,"rms_envelope_nrmse":0.1849555468909176,"envelope_correlation":0.9683123268470599,"rms_gain_error_db":0.20885108149013243,"reference_mean_power":0.0020825301889685474,"output_mean_power":0.0021851256446066536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.598844168109501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038009245216565226,"output_power":0.0004462266902869221,"reference_active":true,"delta_db":0.6966629779058607},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.129961338774397e-05,"output_power":5.5943141650267114e-05,"reference_active":true,"delta_db":-0.3971088366944391},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.439999050166804e-05,"output_power":1.662434237962748e-05,"reference_active":true,"delta_db":0.6238226900218253}],"active_band_count":3,"active_band_mae_db":0.5725315015407084}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012179662863477574,"log_rms_mae_db":1.7394156643517622,"rms_envelope_nrmse":0.16338869647405155,"envelope_correlation":0.9729736031868724,"rms_gain_error_db":0.1883971549111344,"reference_mean_power":0.0020746475669658963,"output_mean_power":0.0021666264793252193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.889619852823584e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035505972043036534,"output_power":0.0004182155220273962,"reference_active":true,"delta_db":0.7109874079806242},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2260222284985983e-05,"output_power":2.956539836481272e-05,"reference_active":true,"delta_db":-0.37883620446785415},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1886983564224354e-06,"output_power":1.396750194854252e-06,"reference_active":true,"delta_db":0.7004707839807002}],"active_band_count":3,"active_band_mae_db":0.5967647988097262}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":3.032928364284683,"rms_envelope_nrmse":0.2619628030818955,"envelope_correlation":0.9570607850723486,"rms_gain_error_db":-1.3692776330332794,"reference_mean_power":0.00015543877011727882,"output_mean_power":0.000113404839434467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4080839081981733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.101806003601559e-05,"output_power":5.644584709601774e-05,"reference_active":true,"delta_db":-0.997368097443977},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0132703192354953e-06,"output_power":2.995516765124356e-06,"reference_active":true,"delta_db":1.725696635964912},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.589135119212959e-07,"output_power":1.147845982560126e-06,"reference_active":true,"delta_db":1.7969133333658478}],"active_band_count":3,"active_band_mae_db":1.506659355591579}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":2.587980244427021,"rms_envelope_nrmse":0.22860085180710993,"envelope_correlation":0.9693486324419776,"rms_gain_error_db":-1.3261597106077971,"reference_mean_power":0.00015383740434286896,"output_mean_power":0.00011335638106833686},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3572636139953195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.549215276795425e-05,"output_power":5.2658095646065574e-05,"reference_active":true,"delta_db":-0.9472411718490079},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.982536907090337e-07,"output_power":1.4439012782675885e-06,"reference_active":true,"delta_db":1.602965763447304},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.830181704119615e-08,"output_power":7.04070292442754e-08,"reference_active":true,"delta_db":1.6365255151459999}],"active_band_count":3,"active_band_mae_db":1.3955774834807706}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":4.801448966384162,"rms_envelope_nrmse":0.2873030781663603,"envelope_correlation":0.9458337053498783,"rms_gain_error_db":-0.556265033444004,"reference_mean_power":2.2945322907755718e-05,"output_mean_power":2.018680884249136e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.468236349213916e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8725122830509825e-06,"output_power":4.1254335892310775e-06,"reference_active":true,"delta_db":0.27476796330457864},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.737803238621211e-07,"output_power":5.297962019300472e-07,"reference_active":true,"delta_db":-0.34636810744043184},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4333843373978682e-07,"output_power":1.3844316854455821e-07,"reference_active":true,"delta_db":-0.15091124199825331}],"active_band_count":3,"active_band_mae_db":0.2573491042477546}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":4.325802413484472,"rms_envelope_nrmse":0.25709800709973174,"envelope_correlation":0.9559609881522683,"rms_gain_error_db":-0.5403811467358441,"reference_mean_power":2.2958960928334712e-05,"output_mean_power":2.027281766354802e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.967813463196928e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.583722018207827e-06,"output_power":3.864406101599132e-06,"reference_active":true,"delta_db":0.3274844417004558},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.664756741318266e-07,"output_power":2.4270520998191783e-07,"reference_active":true,"delta_db":-0.4057847043734495},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1431200548978786e-08,"output_power":1.1562020975317519e-08,"reference_active":true,"delta_db":0.04941908838030204}],"active_band_count":3,"active_band_mae_db":0.2608960781514025}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"golf","status":"measured","mos":{"status":"measured","value":3.4767918586730957,"predicted_mos":3.4767918586730957,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"3c9073361bcc9ea357d60ed5e5c05e3f141556f72112ce2625a2ef5e9fa39137","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.557368495999981},"started_at":"2026-10-13T05:02:09.264866+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029532870003095013,"native_input_samples":144000,"native_input_sha256":"6185536cf32b5cef047fb22a876313cef6ea8ad6ad0a9b7f66fc39b8a83ec9ab","output_sha256":"3bc1be4fdea4e1569279bae1e54a1a6cedf2bff8993264c51f7a2cf3f3c7e3f6","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.7119274240003506,"synthesis_s":0.11994634199982102,"total_s":0.8318737660001716,"rtf":0.13864562766669528,"output_peak":0.34692978858947754,"output_rms":0.0682843815153847,"finished_at":"2026-10-13T05:02:10.107938+00:00","analysis_stages":{"external_f0_s":0.09619664799993188},"evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"3bc1be4fdea4e1569279bae1e54a1a6cedf2bff8993264c51f7a2cf3f3c7e3f6","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.310267346517037,"output_peak":0.34716970295968214},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.990247219400892,"spectral_convergence":0.4455083658120545,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.073123559686858,"spectral_convergence":0.17578745548969268,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.307588147221616,"spectral_convergence":0.19007376040626134,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015047614912801679,"output_power":0.001206811478813597,"delta_db":-0.9582823615211971,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002728665222494989,"output_power":0.0024101205723167786,"delta_db":-0.539114859478989,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011206207533448502,"output_power":0.0008817028968807818,"delta_db":-1.041363934343423,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016414151218902408,"output_power":0.00011559486037697583,"delta_db":-1.5227990539115752,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.929981970645483e-05,"output_power":3.1192153351716595e-05,"delta_db":-1.0034520066066142,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8672160546533183e-05,"output_power":2.088458028134236e-05,"delta_db":-1.3763466812517593,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.073559816185593,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7814017995197107,"pitch":null,"output_rms_dbfs":-23.310267346517037,"output_peak":0.34716970295968214,"reference_rms_dbfs":-22.528865546997324,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.456986308769789},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9407088390428673,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07474062455961965,"rms_dbfs":-22.528865546997324,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.655298436138079,"mean":-5.7228419537934315e-05,"zero_energy":false},"output":{"rms":0.06831036921303524,"rms_dbfs":-23.310267346517037,"peak":0.34716970295968214,"peak_dbfs":-9.189163642926172,"crest_factor_db":14.121103703590865,"mean":0.00044911199684320583,"zero_energy":false},"rms_difference_db":-0.7814017995197133},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":4.3966074240938754e-07,"output_rms_dbfs":-63.5688231108288,"output_zero_energy":false,"output_peak":0.005726710437711709,"output_energy_fraction":1.447122941498982e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":439,"output_estimated_voiced_frames":447,"both_estimated_voiced_frames":434,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9886104783599089,"absolute_difference_cents_median":3.8009710251625455,"absolute_difference_cents_p95":18.549407667799528,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00018834661138467316,"log_rms_mae_db":1.3475055814253574,"rms_envelope_nrmse":0.12347069774120152,"envelope_correlation":0.9880248637965313,"rms_gain_error_db":-0.7830353152708478,"reference_mean_power":0.005584502660834917,"output_mean_power":0.004663167018675229},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.749981211723326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009253800731659645,"output_power":0.0007512084483810471,"reference_active":true,"delta_db":-0.9055967977880082},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001094232472960387,"output_power":0.00010660235261531639,"reference_active":true,"delta_db":-0.11342809687325736},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.572182878073087e-05,"output_power":2.2685701484100185e-05,"reference_active":true,"delta_db":-0.545496299872106}],"active_band_count":3,"active_band_mae_db":0.5215070648444572}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015988223985812614,"log_rms_mae_db":1.1316328046912856,"rms_envelope_nrmse":0.1078366234741084,"envelope_correlation":0.9930813177670853,"rms_gain_error_db":-0.790409007398022,"reference_mean_power":0.005565801454097752,"output_mean_power":0.004639666977104336},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5604413461656273e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008303161980890314,"output_power":0.0006763489815039244,"reference_active":true,"delta_db":-0.8907266999928387},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.486336077809595e-05,"output_power":5.289649039182668e-05,"reference_active":true,"delta_db":-0.15855549605536962},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0492136887407997e-06,"output_power":2.0198034192095165e-06,"reference_active":true,"delta_db":-0.06278145242706312}],"active_band_count":3,"active_band_mae_db":0.3706878828250905}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00016200114302549588,"log_rms_mae_db":1.4124203841395664,"rms_envelope_nrmse":0.14233821332557411,"envelope_correlation":0.9826687818288353,"rms_gain_error_db":-0.5211297020977861,"reference_mean_power":0.0020741102035403834,"output_mean_power":0.001839580756234356},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.596752961053405e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003801147472946862,"output_power":0.0003124925517001217,"reference_active":true,"delta_db":-0.8507504889435002},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.139153155750043e-05,"output_power":5.899289097621689e-05,"reference_active":true,"delta_db":-0.17308788512719994},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4323085648354184e-05,"output_power":1.3041398981828115e-05,"reference_active":true,"delta_db":-0.4071240720532981}],"active_band_count":3,"active_band_mae_db":0.47698748204133273}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012179662863477574,"log_rms_mae_db":1.0041471514175668,"rms_envelope_nrmse":0.12346258622879581,"envelope_correlation":0.9868711906323882,"rms_gain_error_db":-0.536879937684362,"reference_mean_power":0.0020648348411433305,"output_mean_power":0.0018247245953960315},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.886482500908843e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003550471024710956,"output_power":0.0002939419852137486,"reference_active":true,"delta_db":-0.8202434976403554},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2251811422891965e-05,"output_power":3.078512982244414e-05,"reference_active":true,"delta_db":-0.2021312243149763},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1885181265761218e-06,"output_power":1.1028114907902576e-06,"reference_active":true,"delta_db":-0.32504527110751824}],"active_band_count":3,"active_band_mae_db":0.44913999768761664}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":1.7538211624609397,"rms_envelope_nrmse":0.20937218510475006,"envelope_correlation":0.9798102367052547,"rms_gain_error_db":-1.3802122229144675,"reference_mean_power":0.0001545150792776248,"output_mean_power":0.00011244745918802764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4080673348523324e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.084941738676609e-05,"output_power":5.073384654982717e-05,"reference_active":true,"delta_db":-1.4503849228205943},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9974811079649697e-06,"output_power":1.8353402313832366e-06,"reference_active":true,"delta_db":-0.36766095860504},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.541041873063484e-07,"output_power":8.372093818295013e-07,"reference_active":true,"delta_db":0.45402734132319217}],"active_band_count":3,"active_band_mae_db":0.7573577409162756}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":1.4064849768251362,"rms_envelope_nrmse":0.19226923802365745,"envelope_correlation":0.9859587185867692,"rms_gain_error_db":-1.4001336576064904,"reference_mean_power":0.00015249422275408518,"output_mean_power":0.0001104688988455566},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.3574979408829573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.531470616492437e-05,"output_power":4.668710009602976e-05,"reference_active":true,"delta_db":-1.4581407822992514},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.82985839114727e-07,"output_power":8.994168325104767e-07,"reference_active":true,"delta_db":-0.3858625034201789},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.807495159525782e-08,"output_power":5.093846357657572e-08,"reference_active":true,"delta_db":0.251269858488723}],"active_band_count":3,"active_band_mae_db":0.6984243814027177}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":2.5676031952235854,"rms_envelope_nrmse":0.20981647531266395,"envelope_correlation":0.9754339052659053,"rms_gain_error_db":-0.15070590194390335,"reference_mean_power":2.2918735577662457e-05,"output_mean_power":2.2137066200522506e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.41449179558794e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.854283636028979e-06,"output_power":3.4569487974842777e-06,"reference_active":true,"delta_db":-0.47250724099514374},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.692284864712767e-07,"output_power":6.197127373775642e-07,"reference_active":true,"delta_db":0.3690379680158182},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.426655016174459e-07,"output_power":2.0619918263968492e-07,"reference_active":true,"delta_db":1.5996797170631845}],"active_band_count":3,"active_band_mae_db":0.8137416420247154}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":2.399641821295892,"rms_envelope_nrmse":0.1637563127932231,"envelope_correlation":0.9868584483467217,"rms_gain_error_db":-0.14347915886436627,"reference_mean_power":2.2929214155928255e-05,"output_mean_power":2.2184071416966e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.9147344778456375e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.567016317439521e-06,"output_power":3.3218044168946858e-06,"reference_active":true,"delta_db":-0.3093103780944697},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.629770318450219e-07,"output_power":2.824965929477579e-07,"reference_active":true,"delta_db":0.31095395081508764},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1255428901154453e-08,"output_power":1.4302277403065696e-08,"reference_active":true,"delta_db":1.0404314829758052}],"active_band_count":3,"active_band_mae_db":0.5535652706284542}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4338293075561523,"predicted_mos":3.4338293075561523,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"4ab6f5fd5ac1d023fc36cb78c40cabb45772029fa185290edd756b5566e7db6c","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4808711590012535},"started_at":"2026-10-13T05:06:50.819097+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028554550008266233,"native_input_samples":144000,"native_input_sha256":"6185536cf32b5cef047fb22a876313cef6ea8ad6ad0a9b7f66fc39b8a83ec9ab","output_sha256":"38e9159c0d56f2a69c98bfbc9ede35bddcfbfb79f5efb75c80a15d86d1752ee5","output_samples":144120,"length_delta_samples":120,"analysis_s":2.1487031339993337,"synthesis_s":3.258343672000592,"total_s":5.407046805999926,"rtf":0.9011744676666543,"output_peak":0.34731391072273254,"output_rms":0.07381790262257949,"finished_at":"2026-10-13T05:06:56.237315+00:00","analysis_stages":{"external_f0_s":1.3498298019994763},"evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"38e9159c0d56f2a69c98bfbc9ede35bddcfbfb79f5efb75c80a15d86d1752ee5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.63392375869254,"output_peak":0.34625518600799954},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.565091956085919,"spectral_convergence":0.47434455055469665,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.76166518433524,"spectral_convergence":0.17466726077774405,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.789201124757274,"spectral_convergence":0.17113268294730705,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.001567710601394117,"delta_db":0.18813969359160548,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.002610429140867873,"delta_db":-0.22080727292799607,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0010646165490544334,"delta_db":-0.249782812580918,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.00014940686664796025,"delta_db":-0.4322668111833926,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":3.9379638624718934e-05,"delta_db":0.004170355996811333,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":2.0825827988323877e-05,"delta_db":-1.3842603738898336,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4132378866950928,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12261500898041358,"pitch":null,"output_rms_dbfs":-22.63414789996572,"output_peak":0.34625518600799954,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.038652755059477},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9411519437922078,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0748899180673492,"rms_dbfs":-22.511532890985304,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.637965780126057,"mean":-3.1950950622558594e-05,"zero_energy":false},"output":{"rms":0.07384015590840412,"rms_dbfs":-22.63414789996572,"peak":0.34625518600799954,"peak_dbfs":-9.212074265392312,"crest_factor_db":13.422073634573406,"mean":7.706751625450621e-05,"zero_energy":false},"rms_difference_db":-0.12261500898041433},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":7.604065347218527e-06,"output_rms_dbfs":-51.18954159584807,"output_zero_energy":false,"output_peak":0.07343294378790753,"output_energy_fraction":0.00021384406793733106,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":440,"output_estimated_voiced_frames":440,"both_estimated_voiced_frames":432,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9818181818181818,"absolute_difference_cents_median":6.4186524301386605,"absolute_difference_cents_p95":28.559526502782163,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018834661138467316,"log_rms_mae_db":1.34884017976184,"rms_envelope_nrmse":0.11606393322488372,"envelope_correlation":0.9807186234927325,"rms_gain_error_db":-0.12005285707776242,"reference_mean_power":0.005606081823103195,"output_mean_power":0.005453234159107945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7511238572638392e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009260227506744139,"output_power":0.0008763110336303671,"reference_active":true,"delta_db":-0.23963376700462347},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010953814679795108,"output_power":0.0001092808790740791,"reference_active":true,"delta_db":-0.010212092288173536},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.583845526426034e-05,"output_power":2.2937869441492184e-05,"reference_active":true,"delta_db":-0.5171346959933166}],"active_band_count":3,"active_band_mae_db":0.2556601850953712}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015988223985812614,"log_rms_mae_db":0.9724470671308417,"rms_envelope_nrmse":0.0704018943558899,"envelope_correlation":0.9923317338696164,"rms_gain_error_db":-0.11389575875281532,"reference_mean_power":0.00558793457857638,"output_mean_power":0.005443293313557224},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5615527006220852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008311543470725633,"output_power":0.0007857951791074355,"reference_active":true,"delta_db":-0.24372320586695367},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4975058641843e-05,"output_power":5.5101464085246245e-05,"reference_active":true,"delta_db":0.00997437225001158},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.054233222620555e-06,"output_power":1.9035533920430706e-06,"reference_active":true,"delta_db":-0.3308468599905023}],"active_band_count":3,"active_band_mae_db":0.19484814603582254}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016200114302549476,"log_rms_mae_db":1.8207722534639765,"rms_envelope_nrmse":0.15717476940296424,"envelope_correlation":0.9761265312233879,"rms_gain_error_db":-0.1949910792070224,"reference_mean_power":0.0020825301889685474,"output_mean_power":0.001991095990714222},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.598844168109501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038009245216565226,"output_power":0.00035875464278184335,"reference_active":true,"delta_db":-0.25091715366026923},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.129961338774397e-05,"output_power":6.14030790188545e-05,"reference_active":true,"delta_db":0.007324136484127179},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.439999050166804e-05,"output_power":1.3100238004724215e-05,"reference_active":true,"delta_db":-0.4108301965541005}],"active_band_count":3,"active_band_mae_db":0.22302382889949898}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012179662863477574,"log_rms_mae_db":1.303499020328895,"rms_envelope_nrmse":0.09870939884660525,"envelope_correlation":0.9899775528714067,"rms_gain_error_db":-0.1892952707304134,"reference_mean_power":0.0020746475669658963,"output_mean_power":0.001986162619409731},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.889619852823584e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035505972043036534,"output_power":0.00033447110214879125,"reference_active":true,"delta_db":-0.2594280551073152},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2260222284985983e-05,"output_power":3.205836382543505e-05,"reference_active":true,"delta_db":-0.02726002178787626},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1886983564224354e-06,"output_power":1.0023405865532571e-06,"reference_active":true,"delta_db":-0.7405634615067386}],"active_band_count":3,"active_band_mae_db":0.34241717946731004}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":1.844262602877882,"rms_envelope_nrmse":0.142392759141248,"envelope_correlation":0.9847823886281171,"rms_gain_error_db":-0.32276873759386066,"reference_mean_power":0.00015543877011727882,"output_mean_power":0.00014430536497526595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4080839081981733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.101806003601559e-05,"output_power":6.471624409108089e-05,"reference_active":true,"delta_db":-0.4035550035472865},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0132703192354953e-06,"output_power":2.0807501360282126e-06,"reference_active":true,"delta_db":0.14317840727423672},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.589135119212959e-07,"output_power":7.235963234048616e-07,"reference_active":true,"delta_db":-0.20695933661268873}],"active_band_count":3,"active_band_mae_db":0.251230915811404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":1.206606888641989,"rms_envelope_nrmse":0.09449641409508734,"envelope_correlation":0.9935948658360363,"rms_gain_error_db":-0.30971160022633326,"reference_mean_power":0.00015383740434286896,"output_mean_power":0.00014324873105647112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3572636139953195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.549215276795425e-05,"output_power":5.9732432669789424e-05,"reference_active":true,"delta_db":-0.3997906396092504},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.982536907090337e-07,"output_power":1.0125195695898415e-06,"reference_active":true,"delta_db":0.061625013170887664},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.830181704119615e-08,"output_power":4.51840874474834e-08,"reference_active":true,"delta_db":-0.2897795300666676}],"active_band_count":3,"active_band_mae_db":0.25039839428226857}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":1.8994814692488664,"rms_envelope_nrmse":0.22507563611334638,"envelope_correlation":0.964389420201119,"rms_gain_error_db":-0.5627463130790591,"reference_mean_power":2.2945322907755718e-05,"output_mean_power":2.015670512740106e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.468236349213916e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8725122830509825e-06,"output_power":3.210843073872585e-06,"reference_active":true,"delta_db":-0.8137372349283136},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.737803238621211e-07,"output_power":4.749873912286622e-07,"reference_active":true,"delta_db":-0.8206357007164676},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4333843373978682e-07,"output_power":1.3629978871482396e-07,"reference_active":true,"delta_db":-0.21867472006749808}],"active_band_count":3,"active_band_mae_db":0.617682551904093}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":1.259056754349401,"rms_envelope_nrmse":0.14125600784408504,"envelope_correlation":0.9871119716835713,"rms_gain_error_db":-0.5596477119325065,"reference_mean_power":2.2958960928334712e-05,"output_mean_power":2.018308075017203e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.967813463196928e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.583722018207827e-06,"output_power":2.9647515627293367e-06,"reference_active":true,"delta_db":-0.8234600833283883},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.664756741318266e-07,"output_power":2.22091872718584e-07,"reference_active":true,"delta_db":-0.7912490339688902},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1431200548978786e-08,"output_power":1.0876416995973207e-08,"reference_active":true,"delta_db":-0.2160599422868577}],"active_band_count":3,"active_band_mae_db":0.610256353194712}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.727771282196045,"predicted_mos":3.727771282196045,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"00bacf7dfa3c29e88b3517c94db95dec4bfc9bb234f01e2a111c98a8a6ac3425","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0392854580004496},"started_at":"2026-10-13T05:10:13.972826+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003156827000566409,"native_input_samples":144000,"native_input_sha256":"6185536cf32b5cef047fb22a876313cef6ea8ad6ad0a9b7f66fc39b8a83ec9ab","output_sha256":"813c1bf41792698061d1f72000220ad9f726b5db774f6f81df6f8848a10becb9","output_samples":144240,"length_delta_samples":240,"analysis_s":1.038777598998422,"synthesis_s":0.24501408300056937,"total_s":1.2837916819989914,"rtf":0.21396528033316523,"output_peak":0.3589308261871338,"output_rms":0.07406025188948684,"finished_at":"2026-10-13T05:10:15.271324+00:00","analysis_stages":{"external_f0_s":0.16580021899972053},"evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"813c1bf41792698061d1f72000220ad9f726b5db774f6f81df6f8848a10becb9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.605512122754913,"output_peak":0.35427954142599194},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":3.655178748634472,"spectral_convergence":0.08824310147241195,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.6483763588691835,"spectral_convergence":0.08006873909212416,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.1438886701532995,"spectral_convergence":0.0868778983977331,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001501246216341323,"output_power":0.0014862471716129656,"delta_db":-0.04360884692341905,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002746582421816528,"output_power":0.0026783557828048256,"delta_db":-0.10924369667537073,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011276426514964012,"output_power":0.0011120582955556995,"delta_db":-0.06043939958202211,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016504304929751857,"output_power":0.0001617254084523426,"delta_db":-0.0881898230001033,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.934184208485853e-05,"output_power":3.6164491958175716e-05,"delta_db":-0.365723223582795,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8643647097645308e-05,"output_power":2.2451974372031687e-05,"delta_db":-1.0577377656859772,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2874904592416146,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08720727062185933,"pitch":null,"output_rms_dbfs":-22.59874016160716,"output_peak":0.35427954142599194,"reference_rms_dbfs":-22.511532890985304,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":3.815814592552318},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9850828221736339,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0748899180673492,"rms_dbfs":-22.511532890985304,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.637965780126057,"mean":-3.1950950622558594e-05,"zero_energy":false},"output":{"rms":0.0741417771890858,"rms_dbfs":-22.598740161607168,"peak":0.35427954142599194,"peak_dbfs":-9.013078522759919,"crest_factor_db":13.585661638847244,"mean":-6.114211382536076e-05,"zero_energy":false},"rms_difference_db":-0.08720727062186384},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":1.2333546175456772e-06,"output_rms_dbfs":-59.08912035893576,"output_zero_energy":false,"output_peak":0.03824685760039956,"output_energy_fraction":3.44031775857377e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":440,"output_estimated_voiced_frames":436,"both_estimated_voiced_frames":433,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9840909090909091,"absolute_difference_cents_median":0.673310429361076,"absolute_difference_cents_p95":5.849542353513846,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018834661138467316,"log_rms_mae_db":0.46095261458415604,"rms_envelope_nrmse":0.034264997604115006,"envelope_correlation":0.9984490483475084,"rms_gain_error_db":-0.08530961752469596,"reference_mean_power":0.005606081823103195,"output_mean_power":0.005497034598378992},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7511238572638392e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009260227506744139,"output_power":0.0009158316518393582,"reference_active":true,"delta_db":-0.048060076299052575},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010953814679795108,"output_power":9.420926963881211e-05,"reference_active":true,"delta_db":-0.6547175219782788},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.583845526426034e-05,"output_power":1.905929689493221e-05,"reference_active":true,"delta_db":-1.3215967077656745}],"active_band_count":3,"active_band_mae_db":0.6747914353476686}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015988223985812614,"log_rms_mae_db":0.3066400914816693,"rms_envelope_nrmse":0.021410358456797337,"envelope_correlation":0.9994315965061714,"rms_gain_error_db":-0.08378048491660754,"reference_mean_power":0.00558793457857638,"output_mean_power":0.005481169903835408},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5615527006220852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008311543470725633,"output_power":0.0008232124167466982,"reference_active":true,"delta_db":-0.04169768261574946},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4975058641843e-05,"output_power":4.6999241049475866e-05,"reference_active":true,"delta_db":-0.6807485635218796},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.054233222620555e-06,"output_power":1.361871681210976e-06,"reference_active":true,"delta_db":-1.7851355943516558}],"active_band_count":3,"active_band_mae_db":0.8358606134964283}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016200114302549476,"log_rms_mae_db":0.7034213617468953,"rms_envelope_nrmse":0.040361379802143764,"envelope_correlation":0.9985450399002326,"rms_gain_error_db":-0.09655054431085888,"reference_mean_power":0.0020825301889685474,"output_mean_power":0.002036743089000454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.598844168109501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038009245216565226,"output_power":0.0003702396786090326,"reference_active":true,"delta_db":-0.11406285239088867},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.129961338774397e-05,"output_power":5.275231259890373e-05,"reference_active":true,"delta_db":-0.6521623212020089},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.439999050166804e-05,"output_power":1.0579137264623096e-05,"reference_active":true,"delta_db":-1.3391195347991505}],"active_band_count":3,"active_band_mae_db":0.7017815694640159}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012179662863477574,"log_rms_mae_db":0.39841564203254837,"rms_envelope_nrmse":0.025805424648318368,"envelope_correlation":0.999416983953067,"rms_gain_error_db":-0.09541328263274074,"reference_mean_power":0.0020746475669658963,"output_mean_power":0.002029565177400854},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.889619852823584e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035505972043036534,"output_power":0.00034409056166043183,"reference_active":true,"delta_db":-0.1362864661098451},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2260222284985983e-05,"output_power":2.7869606724560865e-05,"reference_active":true,"delta_db":-0.6353651520300795},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1886983564224354e-06,"output_power":8.736522188188585e-07,"reference_active":true,"delta_db":-1.3373307799187972}],"active_band_count":3,"active_band_mae_db":0.7029941326862406}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":0.825090821341019,"rms_envelope_nrmse":0.05831485238207527,"envelope_correlation":0.9974894308742562,"rms_gain_error_db":-0.11587651734390056,"reference_mean_power":0.00015543877011727882,"output_mean_power":0.00015134626217898892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4080839081981733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.101806003601559e-05,"output_power":7.046594991108116e-05,"reference_active":true,"delta_db":-0.03389493948616503},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0132703192354953e-06,"output_power":1.8157036126574911e-06,"reference_active":true,"delta_db":-0.44857133289047735},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.589135119212959e-07,"output_power":5.760584102289505e-07,"reference_active":true,"delta_db":-1.1972576366525514}],"active_band_count":3,"active_band_mae_db":0.5599079696763979}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":0.5188267440677449,"rms_envelope_nrmse":0.0329877459862484,"envelope_correlation":0.9992702675072759,"rms_gain_error_db":-0.1064226793687506,"reference_mean_power":0.00015383740434286896,"output_mean_power":0.0001501134739627871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3572636139953195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.549215276795425e-05,"output_power":6.522532243460026e-05,"reference_active":true,"delta_db":-0.017730315270353028},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.982536907090337e-07,"output_power":9.138608866532634e-07,"reference_active":true,"delta_db":-0.38360834655179893},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.830181704119615e-08,"output_power":3.3504760949327466e-08,"reference_active":true,"delta_db":-1.5885694489241682}],"active_band_count":3,"active_band_mae_db":0.6633027035821067}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":0.9167951672199877,"rms_envelope_nrmse":0.11344130607528233,"envelope_correlation":0.9919016680343264,"rms_gain_error_db":-0.3947793360678238,"reference_mean_power":2.2945322907755718e-05,"output_mean_power":2.0951553983556463e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.468236349213916e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8725122830509825e-06,"output_power":3.5488515017520634e-06,"reference_active":true,"delta_db":-0.3790497685639393},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.737803238621211e-07,"output_power":4.859151332886281e-07,"reference_active":true,"delta_db":-0.721852263970134},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4333843373978682e-07,"output_power":9.254017628354085e-08,"reference_active":true,"delta_db":-1.9003233215030129}],"active_band_count":3,"active_band_mae_db":1.0004084513456954}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":0.6223885710182726,"rms_envelope_nrmse":0.0706024064926251,"envelope_correlation":0.9976345780919031,"rms_gain_error_db":-0.39357322057381683,"reference_mean_power":2.2958960928334712e-05,"output_mean_power":2.0969829864986752e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.967813463196928e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.583722018207827e-06,"output_power":3.2780440504708645e-06,"reference_active":true,"delta_db":-0.3871952964731298},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.664756741318266e-07,"output_power":2.1889087476775534e-07,"reference_active":true,"delta_db":-0.8542991270353785},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1431200548978786e-08,"output_power":7.330394816741635e-09,"reference_active":true,"delta_db":-1.9296447760879687}],"active_band_count":3,"active_band_mae_db":1.0570463998654922}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_1995-1837-0023","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2175559997558594,"predicted_mos":3.2175559997558594,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":286820,"input_sample_rate":48000,"input_seconds":5.975416666666667,"input_float32_sha256":"ed6043449fe4d65a2d94c2d3ac7291ec6e586721bfe804a615990336950d8146","internal_samples":95607,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.10157124899888},"started_at":"2026-10-13T05:13:21.380027+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.01045876100033638,"native_input_samples":288000,"native_input_sha256":"f0028827efcb09d7a3ed896f80169fb05c9aa0a382499ccb3f90442eec1791bf","output_sha256":"a19b9b5001eab651bfa08999ec5fb2a9c53c225acbc5f9a5c78e99d7334566d1","output_samples":286820,"length_delta_samples":-1180,"analysis_s":3.5043138740002178,"synthesis_s":1.1208769040003972,"total_s":4.625190778000615,"rtf":0.7708651296667691,"output_peak":0.36780925753108046,"output_rms":0.07164998065737763,"finished_at":"2026-10-13T05:13:26.028432+00:00","evaluation_input":{"reference":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"output_sha256":"a19b9b5001eab651bfa08999ec5fb2a9c53c225acbc5f9a5c78e99d7334566d1","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95607,"length_delta_samples":-393,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.893098185260186,"output_peak":0.36791238100166374},"common_prefix":{"samples":95607,"sample_rate":16000,"reference_coverage":0.99590625,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95607,"output_samples":95607,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.64446161291767,"spectral_convergence":0.44616170536281147,"spectral_convergence_status":"ok","amplitude_floor":1.5362649016028353e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.576474110282069,"spectral_convergence":0.112559220804919,"spectral_convergence_status":"ok","amplitude_floor":1.0749003550909146e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.007838024831302,"spectral_convergence":0.12119639345331869,"spectral_convergence_status":"ok","amplitude_floor":8.625899017329306e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015143932150953722,"output_power":0.001349034633314103,"delta_db":-0.502155558877715,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002703313427111477,"output_power":0.0025253783670830623,"delta_db":-0.295699456252195,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011095298339820119,"output_power":0.0010355917165172576,"delta_db":-0.29950416458999235,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016200250510838403,"output_power":0.00016042814584435344,"delta_db":-0.04241166109598065,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.922935519409493e-05,"output_power":3.918847448054529e-05,"delta_db":-0.004528120707476253,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.870919988321863e-05,"output_power":2.7149383605611384e-05,"delta_db":-0.24261115110015094,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2311516854372517,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.341640900344827,"pitch":null,"output_rms_dbfs":-22.893098185260186,"output_peak":0.36791238100166374,"reference_rms_dbfs":-22.55145728491536,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.076257916010348},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95607,"output_samples":95607,"intelligibility":{"estoi":{"value":0.9800964579423335,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07454647903405338,"rms_dbfs":-22.55145728491536,"peak":0.360015869140625,"peak_dbfs":-8.873567110859247,"crest_factor_db":13.677890174056115,"mean":-2.8573022957099115e-05,"zero_energy":false},"output":{"rms":0.07167126844729996,"rms_dbfs":-22.893098185260186,"peak":0.36791238100166374,"peak_dbfs":-8.685111940565367,"crest_factor_db":14.20798624469482,"mean":0.0019577285242667994,"zero_energy":false},"rms_difference_db":-0.3416409003448244},"full_scale":{"reference":{"sample_count":95607,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95607,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":597,"uncovered_tail_samples":87,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.292106799199246e-06,"threshold_dbfs":-54.82526084288884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":92,"quiet_samples":14720,"reference_mean_power":4.016050851255979e-07,"output_mean_power":4.6369738360102333e-07,"output_rms_dbfs":-63.33765354591243,"output_zero_energy":false,"output_peak":0.012059930533034205,"output_energy_fraction":1.3898328839155697e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":590,"reference_estimated_voiced_frames":437,"output_estimated_voiced_frames":434,"both_estimated_voiced_frames":430,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9839816933638444,"absolute_difference_cents_median":2.932836648995882,"absolute_difference_cents_p95":11.508096524237061,"difference_over_600_cents_frames":0,"uncovered_tail_samples":87,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95607,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1194,"rms_floor":0.00018834661138467316,"log_rms_mae_db":0.8871462599575547,"rms_envelope_nrmse":0.07707237417508521,"envelope_correlation":0.9935302107799302,"rms_gain_error_db":-0.3432570421188659,"reference_mean_power":0.005554388856130748,"output_mean_power":0.005132282773615107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":1.7483235745868259e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242974994894068,"output_power":0.0008675178439019345,"reference_active":true,"delta_db":-0.27533361615666296},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010925661116474526,"output_power":0.00010114454736648449,"reference_active":true,"delta_db":-0.3350525056554138},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.553978219290416e-05,"output_power":2.242773771845868e-05,"reference_active":true,"delta_db":-0.5643172062234937}],"active_band_count":3,"active_band_mae_db":0.39156777601185677}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1188,"rms_floor":0.00015988223985812614,"log_rms_mae_db":0.6552637796747709,"rms_envelope_nrmse":0.05001530744212434,"envelope_correlation":0.9984271470715579,"rms_gain_error_db":-0.3434786865435332,"reference_mean_power":0.005534916383056086,"output_mean_power":0.005114029108419856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":1.558830950316775e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000828929550341382,"output_power":0.0007798299183416881,"reference_active":true,"delta_db":-0.2651772901570065},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.465988069764166e-05,"output_power":4.9677154883568955e-05,"reference_active":true,"delta_db":-0.41511964541913937},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0769354088583176e-06,"output_power":1.8584664434181084e-06,"reference_active":true,"delta_db":-0.48268266655487435}],"active_band_count":3,"active_band_mae_db":0.3876598673770067}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1194,"rms_floor":0.00016200114303312375,"log_rms_mae_db":0.9743492878243156,"rms_envelope_nrmse":0.07412603163521446,"envelope_correlation":0.9951036962595956,"rms_gain_error_db":-0.20165385302927777,"reference_mean_power":0.0020613880899198887,"output_mean_power":0.0019678608116576158},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":8.594097726660197e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003801268957939535,"output_power":0.0003686631272432787,"reference_active":true,"delta_db":-0.1329889636088346},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.149846938841415e-05,"output_power":5.594093687009436e-05,"reference_active":true,"delta_db":-0.4113457148531051},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4242692051180783e-05,"output_power":1.1971601175322068e-05,"reference_active":true,"delta_db":-0.7543984405518607}],"active_band_count":3,"active_band_mae_db":0.4329110396712668}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1188,"rms_floor":0.00012179662863477574,"log_rms_mae_db":0.6217773898431875,"rms_envelope_nrmse":0.043960319220727455,"envelope_correlation":0.9984815834425387,"rms_gain_error_db":-0.20172667624160703,"reference_mean_power":0.00204991858093371,"output_mean_power":0.001956878872440364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":7.882264205943119e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003549995099445186,"output_power":0.000343776464613709,"reference_active":true,"delta_db":-0.1395161253745567},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2229883605650744e-05,"output_power":2.8782612396742878e-05,"reference_active":true,"delta_db":-0.491285278458147},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2008070923225013e-06,"output_power":1.0230940547608395e-06,"reference_active":true,"delta_db":-0.6955768334349972}],"active_band_count":3,"active_band_mae_db":0.4421260790892336}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1194,"rms_floor":5.226796695404074e-05,"log_rms_mae_db":0.9486095380056568,"rms_envelope_nrmse":0.07422340696679018,"envelope_correlation":0.9957749984791723,"rms_gain_error_db":-0.03369001061207797,"reference_mean_power":0.0001519372076047013,"output_mean_power":0.0001507631279420048},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":1.4084197051270672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.05879757392047e-05,"output_power":7.04784501922418e-05,"reference_active":true,"delta_db":-0.0067438229780392815},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9710070147807163e-06,"output_power":1.9830867146518173e-06,"reference_active":true,"delta_db":0.026535351343658534},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.47561524468716e-07,"output_power":6.961175751609311e-07,"reference_active":true,"delta_db":-0.3096434228641347}],"active_band_count":3,"active_band_mae_db":0.11430753239527751}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1188,"rms_floor":5.184196106800431e-05,"log_rms_mae_db":0.5866958549998609,"rms_envelope_nrmse":0.040140016562166154,"envelope_correlation":0.9987209831370156,"rms_gain_error_db":-0.038521561106974335,"reference_mean_power":0.00015052785313104326,"output_mean_power":0.0001491985874851676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":1.3578222971869857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.503929371848133e-05,"output_power":6.488249391382239e-05,"reference_active":true,"delta_db":-0.010482818044673645},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.57700791318288e-07,"output_power":9.507288993378129e-07,"reference_active":true,"delta_db":-0.03173150889787526},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.874929104434867e-08,"output_power":5.157409310245827e-08,"reference_active":true,"delta_db":0.24463295737846932}],"active_band_count":3,"active_band_mae_db":0.09561576144033941}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1194,"rms_floor":2.781014610300033e-05,"log_rms_mae_db":1.1595326303364357,"rms_envelope_nrmse":0.18597206298918442,"envelope_correlation":0.9770512879016995,"rms_gain_error_db":0.29487242382550605,"reference_mean_power":2.2869244396953844e-05,"output_mean_power":2.4475921862534875e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":7.334653948335133e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.829128987128715e-06,"output_power":4.322312328915825e-06,"reference_active":true,"delta_db":0.5261614952763979},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.607352852931224e-07,"output_power":6.502199359503389e-07,"reference_active":true,"delta_db":0.6430239523862424},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4143757357901104e-07,"output_power":3.242526828086086e-07,"reference_active":true,"delta_db":3.6031878081709086}],"active_band_count":3,"active_band_mae_db":1.5907910852778497}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":7,"metrics":{"frame_count":1188,"rms_floor":2.4952988731037583e-05,"log_rms_mae_db":0.8005133634454265,"rms_envelope_nrmse":0.11219535143301765,"envelope_correlation":0.9917592638761044,"rms_gain_error_db":0.29117140092466864,"reference_mean_power":2.288255525125011e-05,"output_mean_power":2.446930643485e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":6.836423053480652e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5424529324690876e-06,"output_power":4.109365273586227e-06,"reference_active":true,"delta_db":0.6447065796531173},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5743867635038754e-07,"output_power":3.2529445962137845e-07,"reference_active":true,"delta_db":1.0160287270284134},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1361030247208987e-08,"output_power":2.43183462895496e-08,"reference_active":true,"delta_db":3.305163223642414}],"active_band_count":3,"active_band_mae_db":1.6552995101079817}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"hnm","status":"measured","mos":{"status":"measured","value":3.1396071910858154,"predicted_mos":3.1396071910858154,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":51840,"input_sample_rate":16000,"input_seconds":3.24,"input_float32_sha256":"58793b4ffacbda2615051bb4fd02661efc5d358cbc5d6af5487c0934ad46b076","internal_samples":51840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.330051419001393},"started_at":"2026-10-13T04:38:30.971125+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.1363999874447472e-05,"native_input_samples":51840,"native_input_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","output_sha256":"4f9a33a60342ed43f085899bf82c897cd93eab4af79b1e9227add9b28c00433d","output_samples":51840,"length_delta_samples":0,"analysis_s":0.5530479050003123,"synthesis_s":0.24414084799991542,"total_s":0.7971887530002277,"rtf":0.24604591141982335,"output_peak":0.5158652929434137,"output_rms":0.08191562461718822,"finished_at":"2026-10-13T04:38:31.771576+00:00","evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"4f9a33a60342ed43f085899bf82c897cd93eab4af79b1e9227add9b28c00433d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.270206591150415,"spectral_convergence":0.258154410090934,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.195291225430087,"spectral_convergence":0.23274549390105098,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.539856772596859,"spectral_convergence":0.23683494929336296,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0012039146083966715,"delta_db":-0.16009950119046598,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.003607383898617449,"delta_db":-0.026099912008284147,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006283717083490324,"delta_db":0.013244258594580772,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.0008165843585388904,"delta_db":0.008686004750443609,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.0002139912675633843,"delta_db":-0.07052279176082817,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.0002399237149586657,"delta_db":0.0017321311031763351,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.046730766567963176,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0430312696150109,"pitch":null,"output_rms_dbfs":-21.732665056895787,"output_peak":0.5158652929434137,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.335118196392454},"common_prefix":{"samples":51840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.270206591150415,"spectral_convergence":0.258154410090934,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.195291225430087,"spectral_convergence":0.23274549390105098,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.539856772596859,"spectral_convergence":0.23683494929336296,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0012039146083966715,"delta_db":-0.16009950119046598,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.003607383898617449,"delta_db":-0.026099912008284147,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006283717083490324,"delta_db":0.013244258594580772,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.0008165843585388904,"delta_db":0.008686004750443609,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.0002139912675633843,"delta_db":-0.07052279176082817,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.0002399237149586657,"delta_db":0.0017321311031763351,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.046730766567963176,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0430312696150109,"pitch":null,"output_rms_dbfs":-21.732665056895787,"output_peak":0.5158652929434137,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.335118196392454},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51840,"output_samples":51840,"intelligibility":{"estoi":{"value":0.889758424390996,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0823224544816197,"rms_dbfs":-21.689633787280776,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.55268657733706,"mean":0.0004703509954758632,"zero_energy":false},"output":{"rms":0.08191562461718822,"rms_dbfs":-21.732665056895787,"peak":0.5158652929434137,"peak_dbfs":-5.749273803464263,"crest_factor_db":15.983391253431522,"mean":-0.00015182769138942043,"zero_energy":false},"rms_difference_db":-0.04303126961501036},"full_scale":{"reference":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":324,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":317,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":179,"both_estimated_voiced_frames":174,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.93048128342246,"absolute_difference_cents_median":1.3124577461798634,"absolute_difference_cents_p95":9.238277311854697,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00019883564266349852,"log_rms_mae_db":1.042351709615178,"rms_envelope_nrmse":0.07043363060295722,"envelope_correlation":0.9925641671177401,"rms_gain_error_db":-0.04304342610617462,"reference_mean_power":0.006787407573323676,"output_mean_power":0.006720469061026634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.139266319409744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011382823319318319,"output_power":0.0011828402097821523,"reference_active":true,"delta_db":0.16676084709102965},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037355144462451696,"output_power":0.00034463541847153535,"reference_active":true,"delta_db":-0.34990512068223767},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.066072767849799e-05,"output_power":5.470953320433548e-05,"reference_active":true,"delta_db":-0.44844606465947906}],"active_band_count":3,"active_band_mae_db":0.3217040108109155}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00017748806718771712,"log_rms_mae_db":0.5397125321344781,"rms_envelope_nrmse":0.03895282373613246,"envelope_correlation":0.9974435244292735,"rms_gain_error_db":-0.04309654272381651,"reference_mean_power":0.006850619249749789,"output_mean_power":0.006782974373630301},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.7219585966772077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009980124299609617,"output_power":0.0010382000392058863,"reference_active":true,"delta_db":0.1714509060792892},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018464455345700424,"output_power":0.00017479346695212025,"reference_active":true,"delta_db":-0.23811305104801084},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884647670701787e-06,"output_power":4.733464425999141e-06,"reference_active":true,"delta_db":-0.1365412599590242}],"active_band_count":3,"active_band_mae_db":0.18203507236210806}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015076905043222532,"log_rms_mae_db":1.4267069014894354,"rms_envelope_nrmse":0.10489897034688933,"envelope_correlation":0.9890705352443622,"rms_gain_error_db":-0.12137844177353033,"reference_mean_power":0.0024151558510913072,"output_mean_power":0.002348590592990285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.3965349131952106e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007806812222828867,"output_power":0.0007410955899140708,"reference_active":true,"delta_db":-0.22599504519856028},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024651476965397817,"output_power":0.00021548154012650443,"reference_active":true,"delta_db":-0.5843287370731955},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.071673933883429e-05,"output_power":2.1544053409110585e-05,"reference_active":true,"delta_db":-1.5404769527688318}],"active_band_count":3,"active_band_mae_db":0.7836002450135292}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00011516776273837431,"log_rms_mae_db":0.9752372251627517,"rms_envelope_nrmse":0.06190530417124038,"envelope_correlation":0.9959430066874978,"rms_gain_error_db":-0.12138969749334869,"reference_mean_power":0.0024377564716401134,"output_mean_power":0.002370562163242449},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.1901399200800372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007614290342229631,"output_power":0.0007086444717185972,"reference_active":true,"delta_db":-0.31201029522431295},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011269364256369603,"output_power":0.00010174014815126847,"reference_active":true,"delta_db":-0.4440705098669927},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5333706288121514e-06,"output_power":2.63167824183101e-06,"reference_active":true,"delta_db":-1.2795640481179003}],"active_band_count":3,"active_band_mae_db":0.6785482844030687}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001448910076840584,"log_rms_mae_db":1.5129380836268114,"rms_envelope_nrmse":0.10216659287541695,"envelope_correlation":0.9924139112669099,"rms_gain_error_db":-0.03261913875870108,"reference_mean_power":0.0008223932804627595,"output_mean_power":0.0008162395595502648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":6.859948035043366e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004581143880627975,"output_power":0.0004614886429994084,"reference_active":true,"delta_db":0.031870857893592064},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.242911398972496e-05,"output_power":8.040350408314741e-05,"reference_active":true,"delta_db":-0.6053381373499543},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.706958400708282e-06,"output_power":5.256016835769044e-06,"reference_active":true,"delta_db":-2.192097218719879}],"active_band_count":3,"active_band_mae_db":0.9431020713211419}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00013523948700125144,"log_rms_mae_db":1.04404450372406,"rms_envelope_nrmse":0.05601154234621118,"envelope_correlation":0.9975824101542174,"rms_gain_error_db":-0.03262302759252025,"reference_mean_power":0.000830089539077811,"output_mean_power":0.0008238774916522397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":6.395142215880025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004454767193500785,"output_power":0.0004426961494907055,"reference_active":true,"delta_db":-0.027192681381400408},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.066351786105918e-05,"output_power":3.637498975437243e-05,"reference_active":true,"delta_db":-0.48402067717477376},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.874405735020645e-07,"output_power":5.571034532797981e-07,"reference_active":true,"delta_db":-2.02203430140816}],"active_band_count":3,"active_band_mae_db":0.8444158866547781}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":1.4973567300041408,"rms_envelope_nrmse":0.1359266500975396,"envelope_correlation":0.9878865399778948,"rms_gain_error_db":-0.12279852194109603,"reference_mean_power":0.0002494763103331032,"output_mean_power":0.00024252106134480518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.730062412281665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013367937231840974,"output_power":0.00013051806944755352,"reference_active":true,"delta_db":-0.10393756377476907},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8654341368924397e-05,"output_power":2.1409413990821494e-05,"reference_active":true,"delta_db":-1.265856503452689},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.31912639173653e-06,"output_power":7.4309129058660175e-06,"reference_active":true,"delta_db":-0.4903555136463616}],"active_band_count":3,"active_band_mae_db":0.6200498602912732}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":0.9711582574616119,"rms_envelope_nrmse":0.0732365694948045,"envelope_correlation":0.9964170104433581,"rms_gain_error_db":-0.12280165197328113,"reference_mean_power":0.0002518109985947062,"output_mean_power":0.0002447904834819021},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":2.5627299576593646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013078438211609959,"output_power":0.00012633671484534959,"reference_active":true,"delta_db":-0.15026305249347954},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1887209147638681e-05,"output_power":8.935816801540345e-06,"reference_active":true,"delta_db":-1.2394564742848888},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.396397638044023e-07,"output_power":6.728821265088533e-07,"reference_active":true,"delta_db":0.22003538233918432}],"active_band_count":3,"active_band_mae_db":0.5365849697058509}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.250094413757324,"predicted_mos":3.250094413757324,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":51840,"input_sample_rate":16000,"input_seconds":3.24,"input_float32_sha256":"3b3ff2b3696afe57702ad2132fdd9862065bd7fa0c882955f506fa1654337803","internal_samples":51840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3452187269995193},"started_at":"2026-10-13T04:41:28.502631+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5034999453055207e-05,"native_input_samples":51840,"native_input_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","output_sha256":"b3435d091c477083c5b99c6d8d63f5b84b9ad936d54a4b2f900fac9a889418ed","output_samples":51840,"length_delta_samples":0,"analysis_s":0.5996392569995805,"synthesis_s":0.05312854300063918,"total_s":0.6527678000002197,"rtf":0.20147154320994434,"output_peak":0.6269872038564273,"output_rms":0.09096410660016156,"finished_at":"2026-10-13T04:41:29.159460+00:00","evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"b3435d091c477083c5b99c6d8d63f5b84b9ad936d54a4b2f900fac9a889418ed","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.148085988312716,"spectral_convergence":0.5381338128109033,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.126212136228834,"spectral_convergence":0.2892744014205687,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.7880427458651145,"spectral_convergence":0.3151413257652671,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0012683832377067435,"delta_db":0.0664485961426671,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.0048094945971886415,"delta_db":1.2229708750390629,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006762203267342109,"delta_db":0.33196023847553113,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.0010425488770882702,"delta_db":1.0696396845317904,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.00024712496434112815,"delta_db":0.5546828921603146,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.00023069668650656266,"delta_db":-0.1685860747249551,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5690480601790535,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8670349566551766,"pitch":null,"output_rms_dbfs":-20.8225988306256,"output_peak":0.6269872038564273,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.687446956802222},"common_prefix":{"samples":51840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.148085988312716,"spectral_convergence":0.5381338128109033,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.126212136228834,"spectral_convergence":0.2892744014205687,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.7880427458651145,"spectral_convergence":0.3151413257652671,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0012683832377067435,"delta_db":0.0664485961426671,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.0048094945971886415,"delta_db":1.2229708750390629,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006762203267342109,"delta_db":0.33196023847553113,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.0010425488770882702,"delta_db":1.0696396845317904,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.00024712496434112815,"delta_db":0.5546828921603146,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.00023069668650656266,"delta_db":-0.1685860747249551,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5690480601790535,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8670349566551766,"pitch":null,"output_rms_dbfs":-20.8225988306256,"output_peak":0.6269872038564273,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.687446956802222},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51840,"output_samples":51840,"intelligibility":{"estoi":{"value":0.9180559343285432,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0823224544816197,"rms_dbfs":-21.689633787280776,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.55268657733706,"mean":0.0004703509954758632,"zero_energy":false},"output":{"rms":0.09096410660016156,"rms_dbfs":-20.8225988306256,"peak":0.6269872038564273,"peak_dbfs":-4.0548264513681955,"crest_factor_db":16.767772379257405,"mean":-3.663072963269457e-08,"zero_energy":false},"rms_difference_db":0.8670349566551749},"full_scale":{"reference":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":324,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":317,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":205,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9786096256684492,"absolute_difference_cents_median":4.982802251045859,"absolute_difference_cents_p95":29.309275319273883,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00019883564266349852,"log_rms_mae_db":2.7415602461611734,"rms_envelope_nrmse":0.16830581668669017,"envelope_correlation":0.9787522167333145,"rms_gain_error_db":0.8670677283497218,"reference_mean_power":0.006787407573323676,"output_mean_power":0.008287254984156051},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.139266319409744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011382823319318319,"output_power":0.0017008422312928902,"reference_active":true,"delta_db":1.7441403576972987},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037355144462451696,"output_power":0.0005005768691518368,"reference_active":true,"delta_db":1.2712035740699543},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.066072767849799e-05,"output_power":3.483643204696678e-05,"reference_active":true,"delta_db":-2.40873947646263}],"active_band_count":3,"active_band_mae_db":1.8080278027432943}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00017748806718771712,"log_rms_mae_db":2.448715629557636,"rms_envelope_nrmse":0.14768020170237456,"envelope_correlation":0.9859946774886993,"rms_gain_error_db":0.8672480306513182,"reference_mean_power":0.006850619249749789,"output_mean_power":0.008364782127238717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.7219585966772077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009980124299609617,"output_power":0.0015349398138756157,"reference_active":true,"delta_db":1.8695540080418245},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018464455345700424,"output_power":0.0002523322867132868,"reference_active":true,"delta_db":1.3563632182674032},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884647670701787e-06,"output_power":3.598575920250959e-06,"reference_active":true,"delta_db":-1.3270257399062002}],"active_band_count":3,"active_band_mae_db":1.5176476554051426}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015076905043222532,"log_rms_mae_db":2.327255653098251,"rms_envelope_nrmse":0.25922046158013146,"envelope_correlation":0.9768923792092898,"rms_gain_error_db":1.5632490505605123,"reference_mean_power":0.0024151558510913072,"output_mean_power":0.0034615456732204806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.3965349131952106e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007806812222828867,"output_power":0.0012549479789560939,"reference_active":true,"delta_db":2.061519900801399},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024651476965397817,"output_power":0.0003540392058384578,"reference_active":true,"delta_db":1.5720841323550105},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.071673933883429e-05,"output_power":3.301856293637599e-05,"reference_active":true,"delta_db":0.3138305525668075}],"active_band_count":3,"active_band_mae_db":1.3158115285744056}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00011516776273837431,"log_rms_mae_db":2.1259463820467097,"rms_envelope_nrmse":0.25032107565877226,"envelope_correlation":0.97853534048208,"rms_gain_error_db":1.5632314760886898,"reference_mean_power":0.0024377564716401134,"output_mean_power":0.0034939240942033424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.1901399200800372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007614290342229631,"output_power":0.0012223472587431416,"reference_active":true,"delta_db":2.0556516992923894},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011269364256369603,"output_power":0.000159818120583368,"reference_active":true,"delta_db":1.5172660248782375},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5333706288121514e-06,"output_power":3.7147865373272005e-06,"reference_active":true,"delta_db":0.21744668322196928}],"active_band_count":3,"active_band_mae_db":1.2634548024641987}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001448910076840584,"log_rms_mae_db":1.475256512873558,"rms_envelope_nrmse":0.17308092255002328,"envelope_correlation":0.9916667945968797,"rms_gain_error_db":1.0622094873379635,"reference_mean_power":0.0008223932804627595,"output_mean_power":0.0010502688911139236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":6.859948035043366e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004581143880627975,"output_power":0.0006418421294984427,"reference_active":true,"delta_db":1.4645428815747432},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.242911398972496e-05,"output_power":0.00012328946629773592,"reference_active":true,"delta_db":1.2511718260656393},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.706958400708282e-06,"output_power":7.392935204981434e-06,"reference_active":true,"delta_db":-0.7104956967141997}],"active_band_count":3,"active_band_mae_db":1.1420701347848607}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00013523948700125144,"log_rms_mae_db":1.046579529028197,"rms_envelope_nrmse":0.15412366631298266,"envelope_correlation":0.9953874515106051,"rms_gain_error_db":1.0622118317261082,"reference_mean_power":0.000830089539077811,"output_mean_power":0.0010600982657134163},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":6.395142215880025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004454767193500785,"output_power":0.0006258748233901609,"reference_active":true,"delta_db":1.4766246915677914},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.066351786105918e-05,"output_power":5.461879750609203e-05,"reference_active":true,"delta_db":1.281371870451563},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.874405735020645e-07,"output_power":8.203128840333565e-07,"reference_active":true,"delta_db":-0.34159747901536114}],"active_band_count":3,"active_band_mae_db":1.0331980136782386}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":1.321941333335575,"rms_envelope_nrmse":0.22014309752317474,"envelope_correlation":0.9682099286648227,"rms_gain_error_db":-0.023447850392327368,"reference_mean_power":0.0002494763103331032,"output_mean_power":0.0002481330005722053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.730062412281665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013367937231840974,"output_power":0.00012778099478977922,"reference_active":true,"delta_db":-0.19598132850650046},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8654341368924397e-05,"output_power":2.3444768784925834e-05,"reference_active":true,"delta_db":-0.8714447644991148},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.31912639173653e-06,"output_power":9.017413748927748e-06,"reference_active":true,"delta_db":0.35004274462640017}],"active_band_count":3,"active_band_mae_db":0.47248961254400507}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":0.8144272219411188,"rms_envelope_nrmse":0.12427712422017502,"envelope_correlation":0.9895445024354492,"rms_gain_error_db":-0.023449059027506637,"reference_mean_power":0.0002518109985947062,"output_mean_power":0.00025045504796083307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":2.5627299576593646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013078438211609959,"output_power":0.00012471588630203365,"reference_active":true,"delta_db":-0.20634107515769384},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1887209147638681e-05,"output_power":1.0683741500943596e-05,"reference_active":true,"delta_db":-0.4635653241916965},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.396397638044023e-07,"output_power":6.687716564843224e-07,"reference_active":true,"delta_db":0.19342404552826636}],"active_band_count":3,"active_band_mae_db":0.28777681495921886}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"facodec","status":"measured","mos":{"status":"measured","value":4.289660453796387,"predicted_mos":4.289660453796387,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":52000,"input_sample_rate":16000,"input_seconds":3.25,"input_float32_sha256":"684397fa56f9cb51900f1698cbf0c941d3d855830cdd56d992691859f4f106b7","internal_samples":52000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.013017118000789},"started_at":"2026-10-13T04:48:03.361311+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1273000129440334e-05,"native_input_samples":51840,"native_input_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","output_sha256":"1327a276aead045a3e410b565f5626b1246c1e7f78e4d9927f6d15ed80479cb1","output_samples":52000,"length_delta_samples":160,"analysis_s":0.8502005609998378,"synthesis_s":5.499090356000124,"total_s":6.349290916999962,"rtf":1.9596576904320868,"output_peak":0.4313991963863373,"output_rms":0.0775203412971751,"finished_at":"2026-10-13T04:48:09.715460+00:00","evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"1327a276aead045a3e410b565f5626b1246c1e7f78e4d9927f6d15ed80479cb1","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":51840,"output_samples":52000,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.211686477878136,"output_peak":0.4313991963863373},"common_prefix":{"samples":51840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9969230769230769,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.172082196249294,"spectral_convergence":0.229094517050848,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.705395788402004,"spectral_convergence":0.23040303953625352,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.993389570541776,"spectral_convergence":0.2411675557116358,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0009345936637214461,"delta_db":-1.2598280206316612,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.00348859716689413,"delta_db":-0.17151529933462478,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006642606989072507,"delta_db":0.2544636071539881,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.0005802318001008253,"delta_db":-1.4753092939369958,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.00017965860578944575,"delta_db":-0.8300030535080122,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.00018060801014489964,"delta_db":-1.231629564756295,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8704581398869294,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5086698387654305,"pitch":null,"output_rms_dbfs":-22.198303626046204,"output_peak":0.4313991963863373,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.623622518397691},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51840,"output_samples":51840,"intelligibility":{"estoi":{"value":0.9142610743158426,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0823224544816197,"rms_dbfs":-21.689633787280776,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.55268657733706,"mean":0.0004703509954758632,"zero_energy":false},"output":{"rms":0.07763987342569537,"rms_dbfs":-22.198303626046204,"peak":0.4313991963863373,"peak_dbfs":-7.302413364794852,"crest_factor_db":14.895890261251354,"mean":0.00031979875289430814,"zero_energy":false},"rms_difference_db":-0.508669838765428},"full_scale":{"reference":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":324,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":317,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":181,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9679144385026738,"absolute_difference_cents_median":2.3218189772292672,"absolute_difference_cents_p95":15.834997438845098,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00019883564266349852,"log_rms_mae_db":1.408972583480997,"rms_envelope_nrmse":0.10296940538498824,"envelope_correlation":0.9884529143301113,"rms_gain_error_db":-0.5086562519396294,"reference_mean_power":0.006787407573323676,"output_mean_power":0.006037238090969518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.139266319409744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011382823319318319,"output_power":0.0009466848391391682,"reference_active":true,"delta_db":-0.8004457283398587},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037355144462451696,"output_power":0.00035441412784614713,"reference_active":true,"delta_db":-0.2283939464292692},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.066072767849799e-05,"output_power":6.401242819980838e-05,"reference_active":true,"delta_db":0.23356685963750928}],"active_band_count":3,"active_band_mae_db":0.42080217813554577}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00017748806718771712,"log_rms_mae_db":1.1594068792619974,"rms_envelope_nrmse":0.091859032406892,"envelope_correlation":0.9905948973329328,"rms_gain_error_db":-0.5085484258409377,"reference_mean_power":0.006850619249749789,"output_mean_power":0.00609361466788787},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.7219585966772077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009980124299609617,"output_power":0.00081366922437308,"reference_active":true,"delta_db":-0.8868806046444612},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018464455345700424,"output_power":0.00017538927906324712,"reference_active":true,"delta_db":-0.22333458667485007},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884647670701787e-06,"output_power":5.016809580541375e-06,"reference_active":true,"delta_db":0.11594373581481893}],"active_band_count":3,"active_band_mae_db":0.4087196423780434}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015076905043222532,"log_rms_mae_db":1.6224681010311246,"rms_envelope_nrmse":0.0898394423057764,"envelope_correlation":0.9918909723254342,"rms_gain_error_db":-0.06314521780235056,"reference_mean_power":0.0024151558510913072,"output_mean_power":0.0023802942057742143},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.3965349131952106e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007806812222828867,"output_power":0.0007378746876228217,"reference_active":true,"delta_db":-0.24491121037834207},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024651476965397817,"output_power":0.0002606004537594091,"reference_active":true,"delta_db":0.2413222291783867},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.071673933883429e-05,"output_power":3.3466193680050846e-05,"reference_active":true,"delta_db":0.3723120780733844}],"active_band_count":3,"active_band_mae_db":0.2861818392100377}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00011516776273837431,"log_rms_mae_db":1.1595233875270938,"rms_envelope_nrmse":0.07909702073735682,"envelope_correlation":0.9929199748026064,"rms_gain_error_db":-0.06314206040587378,"reference_mean_power":0.0024377564716401134,"output_mean_power":0.002402570343643895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.1901399200800372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007614290342229631,"output_power":0.0007205076754951303,"reference_active":true,"delta_db":-0.23990821305658247},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011269364256369603,"output_power":0.00011962034976735821,"reference_active":true,"delta_db":0.25905651255327516},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5333706288121514e-06,"output_power":4.031481892943667e-06,"reference_active":true,"delta_db":0.572755189032413}],"active_band_count":3,"active_band_mae_db":0.35723997154742354}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001448910076840584,"log_rms_mae_db":1.9126024385234455,"rms_envelope_nrmse":0.18198922400383208,"envelope_correlation":0.9851066298328217,"rms_gain_error_db":-1.140317154503983,"reference_mean_power":0.0008223932804627595,"output_mean_power":0.0006324815156458887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":6.859948035043366e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004581143880627975,"output_power":0.00029504186963203864,"reference_active":true,"delta_db":-1.9109028050264374},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.242911398972496e-05,"output_power":6.7688911793613e-05,"reference_active":true,"delta_db":-1.352912577988923},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.706958400708282e-06,"output_power":6.817691521002542e-06,"reference_active":true,"delta_db":-1.062291226484173}],"active_band_count":3,"active_band_mae_db":1.4420355364998445}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00013523948700125144,"log_rms_mae_db":1.6429688548565011,"rms_envelope_nrmse":0.17191270777457007,"envelope_correlation":0.987279486282316,"rms_gain_error_db":-1.14031848124067,"reference_mean_power":0.000830089539077811,"output_mean_power":0.0006384003151312681},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":6.395142215880025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004454767193500785,"output_power":0.000288572038378444,"reference_active":true,"delta_db":-1.8857076552135803},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.066351786105918e-05,"output_power":3.1417372145569604e-05,"reference_active":true,"delta_db":-1.120350910098214},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.874405735020645e-07,"output_power":5.809500679247069e-07,"reference_active":true,"delta_db":-1.8400047382117286}],"active_band_count":3,"active_band_mae_db":1.615354434507841}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":1.7452993727740118,"rms_envelope_nrmse":0.2469613595769891,"envelope_correlation":0.9649644508773938,"rms_gain_error_db":-1.081630506341288,"reference_mean_power":0.0002494763103331032,"output_mean_power":0.00019447611123783924},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.730062412281665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013367937231840974,"output_power":0.00011077799045441862,"reference_active":true,"delta_db":-0.8161091504105913},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8654341368924397e-05,"output_power":2.0745345286331637e-05,"reference_active":true,"delta_db":-1.4026976284316053},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.31912639173653e-06,"output_power":6.422202575442855e-06,"reference_active":true,"delta_db":-1.123937221534233}],"active_band_count":3,"active_band_mae_db":1.1142480001254766}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":1.3935108326045267,"rms_envelope_nrmse":0.20429606408746087,"envelope_correlation":0.9784633521567874,"rms_gain_error_db":-1.0816364041384747,"reference_mean_power":0.0002518109985947062,"output_mean_power":0.00019629582145283078},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":2.5627299576593646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013078438211609959,"output_power":0.00010845575161327915,"reference_active":true,"delta_db":-0.8130329650544849},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1887209147638681e-05,"output_power":9.038848664418953e-06,"reference_active":true,"delta_db":-1.189667887486991},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.396397638044023e-07,"output_power":4.324188079159295e-07,"reference_active":true,"delta_db":-1.700308790463967}],"active_band_count":3,"active_band_mae_db":1.234336547668481}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5362284183502197,"predicted_mos":3.5362284183502197,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143360,"input_sample_rate":44100,"input_seconds":3.250793650793651,"input_float32_sha256":"69867243786dd3b4b5193a8d99025d8dfa8a2ed9b973797b3fb9161b53fba758","internal_samples":52013,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9421144610023475},"started_at":"2026-10-13T07:05:40.211402+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005470359999890206,"native_input_samples":142884,"native_input_sha256":"34f762c75e8d76dcbd0ccb0c9d45b3d3dbcba619b3969ff2f0586d7c50bb2d1d","output_sha256":"c102739b346718776ffa47db9004ca9e5346ebd7dc40a30662154cc02f2452d6","output_samples":143360,"length_delta_samples":476,"analysis_s":0.09864366500005417,"synthesis_s":19.52363767099996,"total_s":19.622281336000015,"rtf":6.056259671604942,"output_peak":0.47447633743286133,"output_rms":0.07422482359628417,"finished_at":"2026-10-13T07:05:59.857946+00:00","evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"c102739b346718776ffa47db9004ca9e5346ebd7dc40a30662154cc02f2452d6","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":51840,"output_samples":52013,"length_delta_samples":173,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.58651559656466,"output_peak":0.4741534956920637},"common_prefix":{"samples":51840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9966739084459655,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.09161760575,"spectral_convergence":0.4461697224153735,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.20074663938108,"spectral_convergence":0.35569218219882687,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.080231466589147,"spectral_convergence":0.4199815378981596,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0007068484574732189,"delta_db":-2.472793197763863,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.0032096047300804193,"delta_db":-0.5335080232425571,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006498818317777423,"delta_db":0.15942198533906257,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.0005720113745921665,"delta_db":-1.5372779145820379,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.00021324639055975093,"delta_db":-0.08566641470702366,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.00017928578808461398,"delta_db":-1.2635409970228173,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0087014221095603,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8824249739733119,"pitch":null,"output_rms_dbfs":-22.572058761254084,"output_peak":0.4741534956920637,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.12419857057341},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51840,"output_samples":51840,"intelligibility":{"estoi":{"value":0.7962090466671257,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0823224544816197,"rms_dbfs":-21.689633787280776,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.55268657733706,"mean":0.0004703509954758632,"zero_energy":false},"output":{"rms":0.07436987678198688,"rms_dbfs":-22.572058761254084,"peak":0.4741534956920637,"peak_dbfs":-6.481620865221328,"crest_factor_db":16.09043789603276,"mean":-0.0007646819937648653,"zero_energy":false},"rms_difference_db":-0.8824249739733077},"full_scale":{"reference":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":324,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":317,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":173,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9251336898395722,"absolute_difference_cents_median":11.013485334649175,"absolute_difference_cents_p95":43.84067477702686,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00019883564266349852,"log_rms_mae_db":1.9591802324125238,"rms_envelope_nrmse":0.14829040156868573,"envelope_correlation":0.980425828433456,"rms_gain_error_db":-0.8823974992817486,"reference_mean_power":0.006787407573323676,"output_mean_power":0.005539418521227824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.139266319409744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011382823319318319,"output_power":0.0010646167020148805,"reference_active":true,"delta_db":-0.2905671968943348},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037355144462451696,"output_power":0.0003274572967741889,"reference_active":true,"delta_db":-0.5719574796874803},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.066072767849799e-05,"output_power":3.9347528713273305e-05,"reference_active":true,"delta_db":-1.8799015478094425}],"active_band_count":3,"active_band_mae_db":0.9141420747970859}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00017748806718771712,"log_rms_mae_db":1.676627860642911,"rms_envelope_nrmse":0.1292179431047474,"envelope_correlation":0.9874047248532329,"rms_gain_error_db":-0.8822826422077941,"reference_mean_power":0.006850619249749789,"output_mean_power":0.005591155440858726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.7219585966772077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009980124299609617,"output_power":0.0009336398009942641,"reference_active":true,"delta_db":-0.2895659296236518},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018464455345700424,"output_power":0.000169337038560499,"reference_active":true,"delta_db":-0.37584541225338275},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884647670701787e-06,"output_power":3.2948433418763583e-06,"reference_active":true,"delta_db":-1.7099847329611142}],"active_band_count":3,"active_band_mae_db":0.7917986916127163}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015076905043222532,"log_rms_mae_db":2.7337263454849694,"rms_envelope_nrmse":0.16879936361494646,"envelope_correlation":0.9714382099246699,"rms_gain_error_db":-0.1402154371967292,"reference_mean_power":0.0024151558510913072,"output_mean_power":0.002338425926618381},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.3965349131952106e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007806812222828867,"output_power":0.0007747718606693118,"reference_active":true,"delta_db":-0.03299894436998213},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024651476965397817,"output_power":0.00024499206245047353,"reference_active":true,"delta_db":-0.026909308621917895},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.071673933883429e-05,"output_power":1.704311583352788e-05,"reference_active":true,"delta_db":-2.558261166367866}],"active_band_count":3,"active_band_mae_db":0.8727231397865887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00011516776273837431,"log_rms_mae_db":1.9846921271893008,"rms_envelope_nrmse":0.13911500447266814,"envelope_correlation":0.9783083250964636,"rms_gain_error_db":-0.14021027051914878,"reference_mean_power":0.0024377564716401134,"output_mean_power":0.002360311329486602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.1901399200800372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007614290342229631,"output_power":0.0007357314173940543,"reference_active":true,"delta_db":-0.14910131181114272},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011269364256369603,"output_power":0.00012064790905279803,"reference_active":true,"delta_db":0.29620382868071615},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5333706288121514e-06,"output_power":2.1664188242562014e-06,"reference_active":true,"delta_db":-2.1244677398504734}],"active_band_count":3,"active_band_mae_db":0.8565909601141107}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001448910076840584,"log_rms_mae_db":3.4354655931173297,"rms_envelope_nrmse":0.20627506124325104,"envelope_correlation":0.9830034159327701,"rms_gain_error_db":-1.393714963172725,"reference_mean_power":0.0008223932804627595,"output_mean_power":0.0005966340800390432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":6.859948035043366e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004581143880627975,"output_power":0.00034368111796925755,"reference_active":true,"delta_db":-1.24818259524381},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.242911398972496e-05,"output_power":7.578256884861592e-05,"reference_active":true,"delta_db":-0.8623946724392382},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.706958400708282e-06,"output_power":7.685196131929521e-06,"reference_active":true,"delta_db":-0.5421151403704485}],"active_band_count":3,"active_band_mae_db":0.8842308026844989}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00013523948700125144,"log_rms_mae_db":3.0104137309218424,"rms_envelope_nrmse":0.18425254366978772,"envelope_correlation":0.9895555074167242,"rms_gain_error_db":-1.3937143663250275,"reference_mean_power":0.000830089539077811,"output_mean_power":0.0006022176838339701},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":6.395142215880025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004454767193500785,"output_power":0.0003352779693528813,"reference_active":true,"delta_db":-1.2341999521818676},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.066351786105918e-05,"output_power":3.6244659664272046e-05,"reference_active":true,"delta_db":-0.4996092128341266},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.874405735020645e-07,"output_power":5.902636642702819e-07,"reference_active":true,"delta_db":-1.7709323110168933}],"active_band_count":3,"active_band_mae_db":1.168247158677629}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":3.7111719846760085,"rms_envelope_nrmse":0.2134831720247523,"envelope_correlation":0.9712995715909938,"rms_gain_error_db":-0.2806609893949227,"reference_mean_power":0.0002494763103331032,"output_mean_power":0.0002338639137664271},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.730062412281665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013367937231840974,"output_power":0.00012249279251799362,"reference_active":true,"delta_db":-0.3795386214915855},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8654341368924397e-05,"output_power":3.0702053165278026e-05,"reference_active":true,"delta_db":0.2997698892089807},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.31912639173653e-06,"output_power":7.859617363888846e-06,"reference_active":true,"delta_db":-0.24676319100104216}],"active_band_count":3,"active_band_mae_db":0.30869056723386945}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":3.274291109952874,"rms_envelope_nrmse":0.1685478112277653,"envelope_correlation":0.9816165737453765,"rms_gain_error_db":-0.28065512628519695,"reference_mean_power":0.0002518109985947062,"output_mean_power":0.00023605281433281832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":2.5627299576593646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013078438211609959,"output_power":0.00011810338829332709,"reference_active":true,"delta_db":-0.4429352756335446},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1887209147638681e-05,"output_power":1.1855759284723715e-05,"reference_active":true,"delta_db":-0.011505309412096577},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.396397638044023e-07,"output_power":6.414719760322004e-07,"reference_active":true,"delta_db":0.012422338785706194}],"active_band_count":3,"active_band_mae_db":0.15562097461044913}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"golf","status":"measured","mos":{"status":"measured","value":3.2163949012756348,"predicted_mos":3.2163949012756348,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":77520,"input_sample_rate":24000,"input_seconds":3.23,"input_float32_sha256":"cb5fed6b6ba1101ebc1faa3c89cdb17d6cdb55ff5e5b47368716a4e4c4a2f641","internal_samples":51680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1052299220009445},"started_at":"2026-10-13T05:02:10.108644+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0017106780005633482,"native_input_samples":77760,"native_input_sha256":"286afab90f0a65bccf5d9c4545f52cf4a7e8ff838d9f80bc7ddffb450a5f01c0","output_sha256":"ffc950be11d459c9c09b7911ebf010b40162e2c22966bb053df2841fad3595a8","output_samples":77520,"length_delta_samples":-240,"analysis_s":0.3933639309998398,"synthesis_s":0.10395921500003169,"total_s":0.4973231459998715,"rtf":0.15349479814810849,"output_peak":0.5935373306274414,"output_rms":0.0823842820439861,"finished_at":"2026-10-13T05:02:10.614349+00:00","analysis_stages":{"external_f0_s":0.037220985000203655},"evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"ffc950be11d459c9c09b7911ebf010b40162e2c22966bb053df2841fad3595a8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":51840,"output_samples":51680,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.682485331012384,"output_peak":0.5925842216061308},"common_prefix":{"samples":51680,"sample_rate":16000,"reference_coverage":0.9969135802469136,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51680,"output_samples":51680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.166935677499925,"spectral_convergence":0.4622144106879524,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.40823891071177,"spectral_convergence":0.2392498272917944,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.629678608476339,"spectral_convergence":0.252343334383639,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012529042815149912,"output_power":0.0011500197648636017,"delta_db":-0.37212588903990507,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003640368265345664,"output_power":0.003994462159937364,"delta_db":0.4031299160541046,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006283981538675021,"output_power":0.0005770911193795285,"delta_db":-0.3699050930364029,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000817475268014754,"output_power":0.0006566627641826596,"delta_db":-0.9513223185978348,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00021816776758968389,"output_power":0.00018419657056382437,"delta_db":-0.7350904762952972,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024056805781370486,"output_power":0.00022571815198293483,"delta_db":-0.2767147608230529,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5180480756410996,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.006222926734969942,"pitch":null,"output_rms_dbfs":-21.682485331012384,"output_peak":0.5925842216061308,"reference_rms_dbfs":-21.676262404277416,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.7349510655626785},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51680,"output_samples":51680,"intelligibility":{"estoi":{"value":0.9327825866364924,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08244928234024261,"rms_dbfs":-21.676262404277416,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.539315194333698,"mean":0.00046028751337860393,"zero_energy":false},"output":{"rms":0.08239023345586266,"rms_dbfs":-21.682485331012384,"peak":0.5925842216061308,"peak_dbfs":-4.544998328149331,"crest_factor_db":17.137487002863054,"mean":0.0005876652016236484,"zero_energy":false},"rms_difference_db":-0.00622292673496716},"full_scale":{"reference":{"sample_count":51680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":323,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":316,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":204,"both_estimated_voiced_frames":184,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.983957219251337,"absolute_difference_cents_median":3.5215443180468133,"absolute_difference_cents_p95":18.09431433096682,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":0.00019883564266349852,"log_rms_mae_db":1.6128765768146314,"rms_envelope_nrmse":0.10866896547878947,"envelope_correlation":0.9830103721459964,"rms_gain_error_db":-0.006217951343872726,"reference_mean_power":0.00680840368564024,"output_mean_power":0.0067986628224065955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":2.135397027389978e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011427397397678319,"output_power":0.0012028258168627287,"reference_active":true,"delta_db":0.2225541040066842},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000372988602566232,"output_power":0.0004118692023684011,"reference_active":true,"delta_db":0.4306375743206745},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0568414971559014e-05,"output_power":6.457225109370037e-05,"reference_active":true,"delta_db":0.2779971821950507}],"active_band_count":3,"active_band_mae_db":0.31039628684080317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":639,"rms_floor":0.00017748806718771712,"log_rms_mae_db":0.9982322228118884,"rms_envelope_nrmse":0.08414751663310957,"envelope_correlation":0.9884674465170378,"rms_gain_error_db":-0.0062073390084053555,"reference_mean_power":0.006871942059433067,"output_mean_power":0.006862127059240388},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":639,"frequency_resolution_hz":0.31298904538341155,"power_floor":1.71761673805953e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010020317606015327,"output_power":0.0010690278596950473,"reference_active":true,"delta_db":0.2810753617204575},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018408905856030577,"output_power":0.00019141089465343707,"reference_active":true,"delta_db":0.1693867642729387},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.978049238829859e-06,"output_power":5.6304220084934176e-06,"reference_active":true,"delta_db":0.5348175910486603}],"active_band_count":3,"active_band_mae_db":0.3284265723473522}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":0.00015076905043222532,"log_rms_mae_db":1.656112947706731,"rms_envelope_nrmse":0.15325727179585752,"envelope_correlation":0.9808084428045932,"rms_gain_error_db":0.4770164996092069,"reference_mean_power":0.0024226430487431613,"output_mean_power":0.00270390282814859},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":1.3943186158918981e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007797568277478662,"output_power":0.0009515934036261391,"reference_active":true,"delta_db":0.8649223664091605},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024641124114032915,"output_power":0.0003077316044093939,"reference_active":true,"delta_db":0.9651158489759393},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.036847491500593e-05,"output_power":4.322304071591172e-05,"reference_active":true,"delta_db":1.5329233350713012}],"active_band_count":3,"active_band_mae_db":1.120987183485467}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":639,"rms_floor":0.00011516776273837431,"log_rms_mae_db":1.1360350049917323,"rms_envelope_nrmse":0.1387329743392429,"envelope_correlation":0.9828938045022761,"rms_gain_error_db":0.4770214535194172,"reference_mean_power":0.002445384543425529,"output_mean_power":0.0027292876383846498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":639,"frequency_resolution_hz":0.31298904538341155,"power_floor":1.1878352853888112e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007608872366135183,"output_power":0.0009318825653305765,"reference_active":true,"delta_db":0.8804088740214239},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011226910313789305,"output_power":0.0001409713794469676,"reference_active":true,"delta_db":0.9887069616763545},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542129234087384e-06,"output_power":4.815048859827685e-06,"reference_active":true,"delta_db":1.3333629609047524}],"active_band_count":3,"active_band_mae_db":1.0674929322008435}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":0.0001448910076840584,"log_rms_mae_db":1.5769343362942705,"rms_envelope_nrmse":0.1554258733822356,"envelope_correlation":0.9885688238188867,"rms_gain_error_db":-0.9022623254333295,"reference_mean_power":0.000824942824596888,"output_mean_power":0.0006701894960776945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":6.902741188410601e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004611910548196111,"output_power":0.00037307688313091276,"reference_active":true,"delta_db":-0.9208253541040701},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.280840411329098e-05,"output_power":7.350190660561401e-05,"reference_active":true,"delta_db":-1.0128870019708647},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.683294982341308e-06,"output_power":7.931089644103544e-06,"reference_active":true,"delta_db":-0.3935169611651397}],"active_band_count":3,"active_band_mae_db":0.7757431057466914}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":639,"rms_floor":0.00013523948700125144,"log_rms_mae_db":0.9546239195240273,"rms_envelope_nrmse":0.1356657328891641,"envelope_correlation":0.9928997498129947,"rms_gain_error_db":-0.9022631982206433,"reference_mean_power":0.000832687102811657,"output_mean_power":0.0006764808675623558},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":639,"frequency_resolution_hz":0.31298904538341155,"power_floor":6.435426597846404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044842617039161876,"output_power":0.0003567043465215109,"reference_active":true,"delta_db":-0.9938254897616133},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067132144442288e-05,"output_power":3.170404376632987e-05,"reference_active":true,"delta_db":-1.0817362436096438},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.015879703888309e-07,"output_power":9.564871223718652e-07,"reference_active":true,"delta_db":0.2567101887644755}],"active_band_count":3,"active_band_mae_db":0.7774239740452442}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":645,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":2.5237839026656306,"rms_envelope_nrmse":0.19279250803420483,"envelope_correlation":0.9788156972027742,"rms_gain_error_db":-0.5886718890299726,"reference_mean_power":0.00025024969339703686,"output_mean_power":0.00021852763475145413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":645,"frequency_resolution_hz":0.31007751937984496,"power_floor":2.703313915851376e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013233269620133864,"output_power":0.000122883689824478,"reference_active":true,"delta_db":-0.3217291765521646},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8497480298232564e-05,"output_power":2.226316288002304e-05,"reference_active":true,"delta_db":-1.0721959840867799},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.182088958046791e-06,"output_power":1.1434320925706932e-05,"reference_active":true,"delta_db":1.4534618037645446}],"active_band_count":3,"active_band_mae_db":0.9491289881344963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51680,"uncovered_tail_samples":0,"metrics":{"frame_count":639,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":2.3059851829513507,"rms_envelope_nrmse":0.1414309962599252,"envelope_correlation":0.9901528190184387,"rms_gain_error_db":-0.5886698250777798,"reference_mean_power":0.0002525989843698935,"output_mean_power":0.000220579230603581},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":639,"frequency_resolution_hz":0.31298904538341155,"power_floor":2.5363388477366076e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001294356091439165,"output_power":0.0001220244158494427,"reference_active":true,"delta_db":-0.256070346105805},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1785984366235537e-05,"output_power":9.349893441258411e-06,"reference_active":true,"delta_db":-1.0055919934142101},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.360733607716069e-07,"output_power":1.0358142321561866e-06,"reference_active":true,"delta_db":2.117746665783653}],"active_band_count":3,"active_band_mae_db":1.126469668434556}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.5548624992370605,"predicted_mos":3.5548624992370605,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":77880,"input_sample_rate":24000,"input_seconds":3.245,"input_float32_sha256":"d6cc9e3d4a5fad1106cb8249f19a7eb198d2c63109bf4462dc100b5d96f1f91e","internal_samples":51920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0662942069993733},"started_at":"2026-10-13T05:06:56.237956+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0015836599995964207,"native_input_samples":77760,"native_input_sha256":"286afab90f0a65bccf5d9c4545f52cf4a7e8ff838d9f80bc7ddffb450a5f01c0","output_sha256":"91636d9491d09b03c72d80701e85fe9a46f38e4014bddf8f0d2fa7814c3295fd","output_samples":77880,"length_delta_samples":120,"analysis_s":0.9722848219998923,"synthesis_s":5.688345348000439,"total_s":6.660630170000331,"rtf":2.055750052469238,"output_peak":0.4674936830997467,"output_rms":0.08113853137124367,"finished_at":"2026-10-13T05:07:02.905551+00:00","analysis_stages":{"external_f0_s":0.5583865479993619},"evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"91636d9491d09b03c72d80701e85fe9a46f38e4014bddf8f0d2fa7814c3295fd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":51840,"output_samples":51920,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.813413208291827,"output_peak":0.4751034585951082},"common_prefix":{"samples":51840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9984591679506933,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.056408358994618,"spectral_convergence":0.5160856758028829,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.353478184367212,"spectral_convergence":0.2821172217298432,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.414040598381623,"spectral_convergence":0.26834491694445156,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0011812097877279852,"delta_db":-0.24278597414418018,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.003661796623520251,"delta_db":0.038918661884852566,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006246652394222645,"delta_db":-0.012448578315317295,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.000731677939233938,"delta_db":-0.4681249554290093,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.0002089749978931193,"delta_db":-0.17354000914672868,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.00018839290868924801,"delta_db":-1.048354128172046,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3306953845153557,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11708735394281002,"pitch":null,"output_rms_dbfs":-21.806721141223587,"output_peak":0.4751034585951082,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.607975713914485},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51840,"output_samples":51840,"intelligibility":{"estoi":{"value":0.9044506162374231,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0823224544816197,"rms_dbfs":-21.689633787280776,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.55268657733706,"mean":0.0004703509954758632,"zero_energy":false},"output":{"rms":0.08122017912124305,"rms_dbfs":-21.806721141223587,"peak":0.4751034585951082,"peak_dbfs":-6.464236160999034,"crest_factor_db":15.342484980224551,"mean":0.00032410828756371474,"zero_energy":false},"rms_difference_db":-0.11708735394281078},"full_scale":{"reference":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":324,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":317,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":217,"both_estimated_voiced_frames":186,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.9946524064171123,"absolute_difference_cents_median":4.54931439794623,"absolute_difference_cents_p95":26.588633743494103,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00019883564266349852,"log_rms_mae_db":3.2934191572115856,"rms_envelope_nrmse":0.12810662024143815,"envelope_correlation":0.9772350266845566,"rms_gain_error_db":-0.11705599192808422,"reference_mean_power":0.006787407573323676,"output_mean_power":0.006606909066659817},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.139266319409744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011382823319318319,"output_power":0.0011665241708771574,"reference_active":true,"delta_db":0.10643747072336876},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037355144462451696,"output_power":0.0003600092973569493,"reference_active":true,"delta_db":-0.16036703574620462},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.066072767849799e-05,"output_power":4.3199502284649404e-05,"reference_active":true,"delta_db":-1.4742887255072545}],"active_band_count":3,"active_band_mae_db":0.5803644106589426}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00017748806718771712,"log_rms_mae_db":2.8641264678553138,"rms_envelope_nrmse":0.08445618722105397,"envelope_correlation":0.9897480775597619,"rms_gain_error_db":-0.11686912138931084,"reference_mean_power":0.006850619249749789,"output_mean_power":0.006668726685219404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.7219585966772077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009980124299609617,"output_power":0.001031170522271947,"reference_active":true,"delta_db":0.14194539155876396},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018464455345700424,"output_power":0.0001771437863042272,"reference_active":true,"delta_db":-0.18010578465318725},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884647670701787e-06,"output_power":3.5083089007334116e-06,"reference_active":true,"delta_db":-1.4373541827298948}],"active_band_count":3,"active_band_mae_db":0.5864684529806153}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015076905043222532,"log_rms_mae_db":2.200844752846796,"rms_envelope_nrmse":0.16485005509194375,"envelope_correlation":0.9727141264546469,"rms_gain_error_db":0.006075032032699787,"reference_mean_power":0.0024151558510913072,"output_mean_power":0.002418536602278097},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.3965349131952106e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007806812222828867,"output_power":0.0007962128149611904,"reference_active":true,"delta_db":0.08555429818388885},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024651476965397817,"output_power":0.00023874045354815735,"reference_active":true,"delta_db":-0.13916930072560568},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.071673933883429e-05,"output_power":2.844378720446381e-05,"reference_active":true,"delta_db":-0.3338769141808664}],"active_band_count":3,"active_band_mae_db":0.18620017103012032}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00011516776273837431,"log_rms_mae_db":1.646328119199667,"rms_envelope_nrmse":0.10430913282536432,"envelope_correlation":0.987693435877737,"rms_gain_error_db":0.006058258108333318,"reference_mean_power":0.0024377564716401134,"output_mean_power":0.0024411594307224773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.1901399200800372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007614290342229631,"output_power":0.0007728728514530259,"reference_active":true,"delta_db":0.06478619216834816},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011269364256369603,"output_power":0.00010732286548787999,"reference_active":true,"delta_db":-0.21207157005109892},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5333706288121514e-06,"output_power":3.3409653056485135e-06,"reference_active":true,"delta_db":-0.2431722892211015}],"active_band_count":3,"active_band_mae_db":0.17334335048018287}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001448910076840584,"log_rms_mae_db":1.8175943934759966,"rms_envelope_nrmse":0.17673777843622732,"envelope_correlation":0.9777834304426573,"rms_gain_error_db":-0.40335751424521255,"reference_mean_power":0.0008223932804627595,"output_mean_power":0.0007494519634885519},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":6.859948035043366e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004581143880627975,"output_power":0.0003978322554123354,"reference_active":true,"delta_db":-0.6127394002814166},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.242911398972496e-05,"output_power":8.452300037111998e-05,"reference_active":true,"delta_db":-0.38833884877079033},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.706958400708282e-06,"output_power":7.514124958393311e-06,"reference_active":true,"delta_db":-0.63988056515216}],"active_band_count":3,"active_band_mae_db":0.5469862714014556}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00013523948700125144,"log_rms_mae_db":1.0542943592784773,"rms_envelope_nrmse":0.11878254872088322,"envelope_correlation":0.9902399576485251,"rms_gain_error_db":-0.4033590510408308,"reference_mean_power":0.000830089539077811,"output_mean_power":0.0007564653427644103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":6.395142215880025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004454767193500785,"output_power":0.00038779938701335336,"reference_active":true,"delta_db":-0.602178944379262},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.066351786105918e-05,"output_power":3.78526746504205e-05,"reference_active":true,"delta_db":-0.3110837544125113},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.874405735020645e-07,"output_power":7.468477401422141e-07,"reference_active":true,"delta_db":-0.7490720942247104}],"active_band_count":3,"active_band_mae_db":0.5541115976721612}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":1.789933811555316,"rms_envelope_nrmse":0.2221077038577786,"envelope_correlation":0.969027205151415,"rms_gain_error_db":-0.6728177118294734,"reference_mean_power":0.0002494763103331032,"output_mean_power":0.00021367196363802554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.730062412281665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013367937231840974,"output_power":0.0001131582961330889,"reference_active":true,"delta_db":-0.7237799819328807},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8654341368924397e-05,"output_power":2.2161265750538986e-05,"reference_active":true,"delta_db":-1.1159586879544043},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.31912639173653e-06,"output_power":6.434829612093327e-06,"reference_active":true,"delta_db":-1.115406708390549}],"active_band_count":3,"active_band_mae_db":0.9850484594259447}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":1.0781511018178993,"rms_envelope_nrmse":0.1270374431101606,"envelope_correlation":0.9922040501785644,"rms_gain_error_db":-0.6728194636815135,"reference_mean_power":0.0002518109985947062,"output_mean_power":0.00021567149506006398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":2.5627299576593646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013078438211609959,"output_power":0.00011006182899543834,"reference_active":true,"delta_db":-0.7491915932942067},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1887209147638681e-05,"output_power":8.545239731132878e-06,"reference_active":true,"delta_db":-1.4335565282316725},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.396397638044023e-07,"output_power":4.999189605864097e-07,"reference_active":true,"delta_db":-1.070358455454744}],"active_band_count":3,"active_band_mae_db":1.084368858993541}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.5666420459747314,"predicted_mos":3.5666420459747314,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78000,"input_sample_rate":24000,"input_seconds":3.25,"input_float32_sha256":"cd07fdda144b0e8ab84b54f2f476c94c6ff444530832d1d36bc8d9e1a06832fc","internal_samples":52000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.040821541999321},"started_at":"2026-10-13T05:10:15.274124+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0017877339996630326,"native_input_samples":77760,"native_input_sha256":"286afab90f0a65bccf5d9c4545f52cf4a7e8ff838d9f80bc7ddffb450a5f01c0","output_sha256":"f9b563f0a23e216dbe6b2a63047adaeeb46bf4bbf8a5bdbf0f21f6b97d3e90e1","output_samples":78000,"length_delta_samples":240,"analysis_s":0.48992913699839846,"synthesis_s":0.1437686429999303,"total_s":0.6336977799983288,"rtf":0.1955857345673854,"output_peak":0.5130819082260132,"output_rms":0.08101228929290244,"finished_at":"2026-10-13T05:10:15.918540+00:00","analysis_stages":{"external_f0_s":0.08958370400068816},"evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"f9b563f0a23e216dbe6b2a63047adaeeb46bf4bbf8a5bdbf0f21f6b97d3e90e1","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":51840,"output_samples":52000,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.826828056530644,"output_peak":0.4998721584344339},"common_prefix":{"samples":51840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9969230769230769,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.283508246810673,"spectral_convergence":0.22629791601746743,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.193597450762665,"spectral_convergence":0.15930767013157884,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.781371354630793,"spectral_convergence":0.17757859830321648,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0012059663498141654,"delta_db":-0.15270444552871706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.003570986829004306,"delta_db":-0.07014105324256967,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.000613713133789662,"delta_db":-0.08926779864461057,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.000782934407396794,"delta_db":-0.17407077092927592,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.0001979904091659478,"delta_db":-0.4080417728470337,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.00021491294601026572,"delta_db":-0.47637386791893255,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22843328485185657,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12381656406450692,"pitch":null,"output_rms_dbfs":-21.81345035134528,"output_peak":0.4998721584344339,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.419492350734711},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51840,"output_samples":51840,"intelligibility":{"estoi":{"value":0.9550759248249907,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0823224544816197,"rms_dbfs":-21.689633787280776,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.55268657733706,"mean":0.0004703509954758632,"zero_energy":false},"output":{"rms":0.08115727986558656,"rms_dbfs":-21.81345035134528,"peak":0.4998721584344339,"peak_dbfs":-6.0228210327024385,"crest_factor_db":15.790629318642841,"mean":0.0011088392672946089,"zero_energy":false},"rms_difference_db":-0.12381656406450503},"full_scale":{"reference":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":324,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":317,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":196,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9786096256684492,"absolute_difference_cents_median":0.7238458046298609,"absolute_difference_cents_p95":7.447503973627655,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00019883564266349852,"log_rms_mae_db":0.8718299854926634,"rms_envelope_nrmse":0.05083410718976953,"envelope_correlation":0.9964279364026168,"rms_gain_error_db":-0.12383316676708418,"reference_mean_power":0.006787407573323676,"output_mean_power":0.006596607010967474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.139266319409744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011382823319318319,"output_power":0.0011597883212308103,"reference_active":true,"delta_db":0.08128736283707631},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037355144462451696,"output_power":0.00035155370902598343,"reference_active":true,"delta_db":-0.2635873601438401},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.066072767849799e-05,"output_power":5.386621908986326e-05,"reference_active":true,"delta_db":-0.5159112258636339}],"active_band_count":3,"active_band_mae_db":0.28692864961485015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00017748806718771712,"log_rms_mae_db":0.4786932809275437,"rms_envelope_nrmse":0.030059443647074986,"envelope_correlation":0.9988615030919418,"rms_gain_error_db":-0.12380283459610682,"reference_mean_power":0.006850619249749789,"output_mean_power":0.0066580882478309115},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.7219585966772077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009980124299609617,"output_power":0.001010064458254107,"reference_active":true,"delta_db":0.05213139260034549},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018464455345700424,"output_power":0.00018274150100590597,"reference_active":true,"delta_db":-0.044993137899507527},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884647670701787e-06,"output_power":4.289672746213344e-06,"reference_active":true,"delta_db":-0.5640908188980919}],"active_band_count":3,"active_band_mae_db":0.22040511646598163}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015076905043222532,"log_rms_mae_db":0.8064323926861694,"rms_envelope_nrmse":0.039648117367583176,"envelope_correlation":0.9984859187518181,"rms_gain_error_db":-0.07498255064594499,"reference_mean_power":0.0024151558510913072,"output_mean_power":0.0023738151986833793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.3965349131952106e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007806812222828867,"output_power":0.0007712240624567253,"reference_active":true,"delta_db":-0.052931621659977654},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024651476965397817,"output_power":0.00023526580175953197,"reference_active":true,"delta_db":-0.2028414194470067},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.071673933883429e-05,"output_power":2.2364787730263104e-05,"reference_active":true,"delta_db":-1.3781033179416122}],"active_band_count":3,"active_band_mae_db":0.5446254530161988}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00011516776273837431,"log_rms_mae_db":0.37043181639658035,"rms_envelope_nrmse":0.020342568943710657,"envelope_correlation":0.9996128134521697,"rms_gain_error_db":-0.07498379281681702,"reference_mean_power":0.0024377564716401134,"output_mean_power":0.002396028275076051},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.1901399200800372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007614290342229631,"output_power":0.0007488198606541926,"reference_active":true,"delta_db":-0.07252078471069262},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011269364256369603,"output_power":0.00010819230891821192,"reference_active":true,"delta_db":-0.1770302757516205},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5333706288121514e-06,"output_power":2.5575980060951612e-06,"reference_active":true,"delta_db":-1.4035690995242454}],"active_band_count":3,"active_band_mae_db":0.5510400533288529}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001448910076840584,"log_rms_mae_db":0.8532756523926356,"rms_envelope_nrmse":0.055286116641187925,"envelope_correlation":0.9980714152049234,"rms_gain_error_db":-0.18451003863565127,"reference_mean_power":0.0008223932804627595,"output_mean_power":0.0007881856965194069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":6.859948035043366e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004581143880627975,"output_power":0.0004514686007318876,"reference_active":true,"delta_db":-0.06346381060453551},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.242911398972496e-05,"output_power":8.883169697120429e-05,"reference_active":true,"delta_db":-0.1724083137109032},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.706958400708282e-06,"output_power":5.9890598902379374e-06,"reference_active":true,"delta_db":-1.6250781350833152}],"active_band_count":3,"active_band_mae_db":0.620316753132918}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00013523948700125144,"log_rms_mae_db":0.4793040833356561,"rms_envelope_nrmse":0.03438930964523149,"envelope_correlation":0.9994418862994914,"rms_gain_error_db":-0.18451005840424553,"reference_mean_power":0.000830089539077811,"output_mean_power":0.000795561824368537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":6.395142215880025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004454767193500785,"output_power":0.0004383727759355992,"reference_active":true,"delta_db":-0.0698143717886531},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.066351786105918e-05,"output_power":4.010554551254472e-05,"reference_active":true,"delta_db":-0.06000519428709189},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.874405735020645e-07,"output_power":6.022214734154612e-07,"reference_active":true,"delta_db":-1.6838304354633387}],"active_band_count":3,"active_band_mae_db":0.6045500005130279}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":0.9438851182240646,"rms_envelope_nrmse":0.13120306655170497,"envelope_correlation":0.9893621693427016,"rms_gain_error_db":-0.35074917022671875,"reference_mean_power":0.0002494763103331032,"output_mean_power":0.0002301200158896614},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.730062412281665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013367937231840974,"output_power":0.00012570938828141494,"reference_active":true,"delta_db":-0.266966846322222},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8654341368924397e-05,"output_power":2.2438608154289315e-05,"reference_active":true,"delta_db":-1.0619451594022307},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.31912639173653e-06,"output_power":4.9588757418018345e-06,"reference_active":true,"delta_db":-2.2469449655345057}],"active_band_count":3,"active_band_mae_db":1.191952323752986}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":0.606521643350486,"rms_envelope_nrmse":0.06834240086570242,"envelope_correlation":0.9977975353054931,"rms_gain_error_db":-0.3507486538877192,"reference_mean_power":0.0002518109985947062,"output_mean_power":0.0002322735886618164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":2.5627299576593646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013078438211609959,"output_power":0.0001213707719671596,"reference_active":true,"delta_db":-0.32441770684853605},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1887209147638681e-05,"output_power":9.091171846574036e-06,"reference_active":true,"delta_db":-1.164600367722829},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.396397638044023e-07,"output_power":3.67191583115513e-07,"reference_active":true,"delta_db":-2.4104273663182325}],"active_band_count":3,"active_band_mae_db":1.2998151469631993}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0032","method":"magphase","status":"measured","mos":{"status":"measured","value":4.263149261474609,"predicted_mos":4.263149261474609,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":155743,"input_sample_rate":48000,"input_seconds":3.2446458333333332,"input_float32_sha256":"5163c99291af6c9a02025719e63c6f616292d5b14a8b175b2c265ea4230e5d03","internal_samples":51915,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.050514808999651},"started_at":"2026-10-13T05:13:26.029253+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005431408000731608,"native_input_samples":155520,"native_input_sha256":"143b94141e84864c50ae7144b4ae66a572e4dc1556606bf18276d5f44cf6a864","output_sha256":"f93985857e340c7c17650160e29875410a5215f5f724ac782dea9581bff5f132","output_samples":155743,"length_delta_samples":223,"analysis_s":1.9341763769989484,"synthesis_s":0.7173852949999855,"total_s":2.651561671998934,"rtf":0.8183832320984363,"output_peak":0.589350083923,"output_rms":0.08033216167234171,"finished_at":"2026-10-13T05:13:28.694820+00:00","evaluation_input":{"reference":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"f93985857e340c7c17650160e29875410a5215f5f724ac782dea9581bff5f132","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":51840,"output_samples":51915,"length_delta_samples":75,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.900355680316792,"output_peak":0.5481449000579307},"common_prefix":{"samples":51840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9985553308292401,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":51840,"output_samples":51840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.405444372209989,"spectral_convergence":0.5167601596048913,"spectral_convergence_status":"ok","amplitude_floor":1.3508970735580905e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.055350848947778,"spectral_convergence":0.207601932498401,"spectral_convergence_status":"ok","amplitude_floor":9.199259876037329e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.354032770950968,"spectral_convergence":0.2158973335596209,"spectral_convergence_status":"ok","amplitude_floor":7.633709769762664e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012491242347822075,"output_power":0.0010828443431819517,"delta_db":-0.6203960226383455,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0036291285647318994,"output_power":0.0033623772790699934,"delta_db":-0.3315591139092247,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006264583429363991,"output_power":0.0006304402892000854,"delta_db":0.027517622041706213,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0008149528003014777,"output_power":0.0008207416585837454,"delta_db":0.030740211795182236,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002174945257645064,"output_power":0.0002567197759853415,"delta_db":0.7201099468136368,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023982804336185733,"output_power":0.00031221027592906545,"delta_db":1.1454722883518915,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.47929920092499784,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20445681160899373,"pitch":null,"output_rms_dbfs":-21.89409059888977,"output_peak":0.5481449000579307,"reference_rms_dbfs":-21.689633787280776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.604942664036245},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":51840,"output_samples":51840,"intelligibility":{"estoi":{"value":0.9271774142625077,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0823224544816197,"rms_dbfs":-21.689633787280776,"peak":0.49334716796875,"peak_dbfs":-6.136947209943715,"crest_factor_db":15.55268657733706,"mean":0.0004703509954758632,"zero_energy":false},"output":{"rms":0.08040729831271651,"rms_dbfs":-21.89409059888977,"peak":0.5481449000579307,"peak_dbfs":-5.222092444280078,"crest_factor_db":16.671998154609692,"mean":0.0028708335323705206,"zero_energy":false},"rms_difference_db":-0.20445681160899198},"full_scale":{"reference":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":51840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":324,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.953561279340647e-06,"threshold_dbfs":-54.03011525344222,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":317,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":175,"both_estimated_voiced_frames":174,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.93048128342246,"absolute_difference_cents_median":2.793582246506395,"absolute_difference_cents_p95":10.048855606039957,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":51840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00019883564266349852,"log_rms_mae_db":1.4987859189804664,"rms_envelope_nrmse":0.1009519192678156,"envelope_correlation":0.9858604367726014,"rms_gain_error_db":-0.20442431201537054,"reference_mean_power":0.006787407573323676,"output_mean_power":0.006475323907302441},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.139266319409744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011382823319318319,"output_power":0.0011281319653204161,"reference_active":true,"delta_db":-0.03890089899001644},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037355144462451696,"output_power":0.00040077700274070274,"reference_active":true,"delta_db":0.30552372742774553},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.066072767849799e-05,"output_power":5.886161862653731e-05,"reference_active":true,"delta_db":-0.1307541522363203}],"active_band_count":3,"active_band_mae_db":0.15839292621802742}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00017748806718771712,"log_rms_mae_db":0.8674113189956875,"rms_envelope_nrmse":0.056940818126426,"envelope_correlation":0.9955550694272541,"rms_gain_error_db":-0.2043488891569462,"reference_mean_power":0.006850619249749789,"output_mean_power":0.006535742626927002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.7219585966772077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009980124299609617,"output_power":0.0009703144880469076,"reference_active":true,"delta_db":-0.12223434329364825},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018464455345700424,"output_power":0.00020136533857127974,"reference_active":true,"delta_db":0.3764821505360752},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.884647670701787e-06,"output_power":5.004172546075075e-06,"reference_active":true,"delta_db":0.10499032391122798}],"active_band_count":3,"active_band_mae_db":0.2012356059136505}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.00015076905043222532,"log_rms_mae_db":1.3129579821463753,"rms_envelope_nrmse":0.0756931874143314,"envelope_correlation":0.9946624605420193,"rms_gain_error_db":-0.2041687732483842,"reference_mean_power":0.0024151558510913072,"output_mean_power":0.0023042430374203924},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":1.3965349131952106e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007806812222828867,"output_power":0.000744820172889565,"reference_active":true,"delta_db":-0.204223027404905},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024651476965397817,"output_power":0.0002211913375936295,"reference_active":true,"delta_db":-0.47074829749260333},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.071673933883429e-05,"output_power":2.426549013895493e-05,"reference_active":true,"delta_db":-1.0238604426609625}],"active_band_count":3,"active_band_mae_db":0.5662772558528236}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00011516776273837431,"log_rms_mae_db":0.7660056320233104,"rms_envelope_nrmse":0.04873405306860527,"envelope_correlation":0.9978775497059219,"rms_gain_error_db":-0.20417040987516893,"reference_mean_power":0.0024377564716401134,"output_mean_power":0.002325804878118069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":1.1901399200800372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007614290342229631,"output_power":0.0007200332088204821,"reference_active":true,"delta_db":-0.2427690588450153},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011269364256369603,"output_power":0.00010027722565757471,"reference_active":true,"delta_db":-0.5069710674405011},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5333706288121514e-06,"output_power":2.9616002642017343e-06,"reference_active":true,"delta_db":-0.7666275448811021}],"active_band_count":3,"active_band_mae_db":0.5054558903888728}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":0.0001448910076840584,"log_rms_mae_db":1.0573699196587931,"rms_envelope_nrmse":0.0819013888358179,"envelope_correlation":0.9953251516447613,"rms_gain_error_db":0.10647474241608157,"reference_mean_power":0.0008223932804627595,"output_mean_power":0.0008428048524861441},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":6.859948035043366e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004581143880627975,"output_power":0.0004725577486020894,"reference_active":true,"delta_db":0.13480956815747658},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.242911398972496e-05,"output_power":9.589890380870291e-05,"reference_active":true,"delta_db":0.16004852929345034},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.706958400708282e-06,"output_power":7.838767150184424e-06,"reference_active":true,"delta_db":-0.4561870556964097}],"active_band_count":3,"active_band_mae_db":0.25034838438244555}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":0.00013523948700125144,"log_rms_mae_db":0.5507189467950866,"rms_envelope_nrmse":0.04292405745706156,"envelope_correlation":0.9987241778216436,"rms_gain_error_db":0.10647335160441906,"reference_mean_power":0.000830089539077811,"output_mean_power":0.0008506918576642094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":6.395142215880025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004454767193500785,"output_power":0.000461018884217676,"reference_active":true,"delta_db":0.14893702588255198},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.066351786105918e-05,"output_power":4.269146277412364e-05,"reference_active":true,"delta_db":0.2113608827726711},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.874405735020645e-07,"output_power":7.299366035959275e-07,"reference_active":true,"delta_db":-0.8485413819700929}],"active_band_count":3,"active_band_mae_db":0.4029464302084387}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":647,"rms_floor":8.84129869609077e-05,"log_rms_mae_db":1.2997800167315943,"rms_envelope_nrmse":0.2622860017773179,"envelope_correlation":0.9730392546533067,"rms_gain_error_db":1.172905248174683,"reference_mean_power":0.0002494763103331032,"output_mean_power":0.000326828437127172},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":647,"frequency_resolution_hz":0.3091190108191654,"power_floor":2.730062412281665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013367937231840974,"output_power":0.00016632403858587888,"reference_active":true,"delta_db":0.9489062405093915},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8654341368924397e-05,"output_power":5.319690334445002e-05,"reference_active":true,"delta_db":2.686959217456217},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.31912639173653e-06,"output_power":1.571672819766288e-05,"reference_active":true,"delta_db":2.762844200755375}],"active_band_count":3,"active_band_mae_db":2.1329032195736612}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":51840,"uncovered_tail_samples":0,"metrics":{"frame_count":641,"rms_floor":7.608900761326357e-05,"log_rms_mae_db":1.047206476926562,"rms_envelope_nrmse":0.19575489015122305,"envelope_correlation":0.9897502655791097,"rms_gain_error_db":1.1729067061778424,"reference_mean_power":0.0002518109985947062,"output_mean_power":0.0003298871249197267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":641,"frequency_resolution_hz":0.31201248049922,"power_floor":2.5627299576593646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013078438211609959,"output_power":0.00017134618383760647,"reference_active":true,"delta_db":1.1731855149576307},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1887209147638681e-05,"output_power":2.4335433405690334e-05,"reference_active":true,"delta_db":3.1115918138404823},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.396397638044023e-07,"output_power":8.458820774591528e-07,"reference_active":true,"delta_db":1.2137436899383733}],"active_band_count":3,"active_band_mae_db":1.8328403395788289}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"hnm","status":"measured","mos":{"status":"measured","value":2.6334526538848877,"predicted_mos":2.6334526538848877,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"3c760914e0e442a689b08bd5f9469f429846e29de88dd5844ba8debaa7833fd3","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1876699490003375},"started_at":"2026-10-13T04:38:31.772735+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3639999451697804e-05,"native_input_samples":96000,"native_input_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","output_sha256":"f4f172e881bb054e4db42182fdf61ebbf7facb2640bffd8981adf14fe025c43a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0670545600005426,"synthesis_s":0.42614749899985327,"total_s":1.4932020590003958,"rtf":0.2488670098333993,"output_peak":0.4995994196498588,"output_rms":0.08229968634385201,"finished_at":"2026-10-13T04:38:33.272100+00:00","evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"f4f172e881bb054e4db42182fdf61ebbf7facb2640bffd8981adf14fe025c43a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.286990252156443,"spectral_convergence":0.2544928726631055,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1193182858606345,"spectral_convergence":0.22616173158343264,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.416206042995276,"spectral_convergence":0.23013690664139397,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.0020895298022010978,"delta_db":-0.11266906495848039,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.0031035224938867068,"delta_db":-0.03930112082328249,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.0006753562891899857,"delta_db":-0.1481493178281857,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.00040896637574109923,"delta_db":-0.298687285881047,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.00014726152126203233,"delta_db":-0.032665290862244234,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.0003486018900154986,"delta_db":0.380500595293913,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1686621126078588,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06841966762487507,"pitch":null,"output_rms_dbfs":-21.692036398887986,"output_peak":0.4995994196498588,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.274171527004118},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.286990252156443,"spectral_convergence":0.2544928726631055,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1193182858606345,"spectral_convergence":0.22616173158343264,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.416206042995276,"spectral_convergence":0.23013690664139397,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.0020895298022010978,"delta_db":-0.11266906495848039,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.0031035224938867068,"delta_db":-0.03930112082328249,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.0006753562891899857,"delta_db":-0.1481493178281857,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.00040896637574109923,"delta_db":-0.298687285881047,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.00014726152126203233,"delta_db":-0.032665290862244234,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.0003486018900154986,"delta_db":0.380500595293913,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1686621126078588,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06841966762487507,"pitch":null,"output_rms_dbfs":-21.692036398887986,"output_peak":0.4995994196498588,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.274171527004118},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9352604860671652,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08295052965772477,"rms_dbfs":-21.62361673126311,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.770507654342886,"mean":0.00047166951497395836,"zero_energy":false},"output":{"rms":0.08229968634385201,"rms_dbfs":-21.692036398887986,"peak":0.4995994196498588,"peak_dbfs":-6.0275614957502865,"crest_factor_db":15.664474903137698,"mean":-8.8804094479503e-05,"zero_energy":false},"rms_difference_db":-0.06841966762487672},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":363,"output_estimated_voiced_frames":350,"both_estimated_voiced_frames":343,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9449035812672176,"absolute_difference_cents_median":1.3666189580155366,"absolute_difference_cents_p95":8.442190338045874,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020729159449367856,"log_rms_mae_db":0.9086799780485619,"rms_envelope_nrmse":0.0658802319462866,"envelope_correlation":0.9928388721010118,"rms_gain_error_db":-0.06841809267949113,"reference_mean_power":0.0068864212769201745,"output_mean_power":0.006778783721648814},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4588043207823824e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006447788388285564,"output_power":0.0006328434541222449,"reference_active":true,"delta_db":-0.08114483509001133},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001410929626621107,"output_power":0.0001283009186736387,"reference_active":true,"delta_db":-0.41275586785868945},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.775085636154685e-05,"output_power":2.5990312455091043e-05,"reference_active":true,"delta_db":-0.28464888912354513}],"active_band_count":3,"active_band_mae_db":0.25951653069074865}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001768834026970723,"log_rms_mae_db":0.4175257055981277,"rms_envelope_nrmse":0.030833657097124816,"envelope_correlation":0.9983308951915474,"rms_gain_error_db":-0.06845620838492367,"reference_mean_power":0.006918665749586704,"output_mean_power":0.0068104644280720565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2539809640994155e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005542590263527942,"output_power":0.0005434525876679804,"reference_active":true,"delta_db":-0.08551113518285976},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.178340930501843e-05,"output_power":6.701109283208153e-05,"reference_active":true,"delta_db":-0.2987738044657691},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4823995625387325e-06,"output_power":1.4633105796391253e-06,"reference_active":true,"delta_db":-0.056287656202098454}],"active_band_count":3,"active_band_mae_db":0.14685753195024245}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016054516244843413,"log_rms_mae_db":1.5066156512101936,"rms_envelope_nrmse":0.0859643414559666,"envelope_correlation":0.9927294615329793,"rms_gain_error_db":-0.08135691672826849,"reference_mean_power":0.0021921303983993327,"output_mean_power":0.0021514472027812587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.01765222997539e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005084502896759698,"output_power":0.0004800051293814065,"reference_active":true,"delta_db":-0.25002620784638485},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.688692893438752e-05,"output_power":8.609045968468948e-05,"reference_active":true,"delta_db":-0.513101633468174},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1104523408685175e-05,"output_power":1.4452625892018567e-05,"reference_active":true,"delta_db":-1.6442878812018715}],"active_band_count":3,"active_band_mae_db":0.8024719075054767}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013960343413021469,"log_rms_mae_db":0.9054595032626379,"rms_envelope_nrmse":0.04836996237475691,"envelope_correlation":0.9975879719546703,"rms_gain_error_db":-0.08137884058322259,"reference_mean_power":0.00220313483790695,"output_mean_power":0.0021622364983385936},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.111026246795902e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004662873703819768,"output_power":0.0004369873053654392,"reference_active":true,"delta_db":-0.2818483179198519},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.729339985490381e-05,"output_power":4.18663728640148e-05,"reference_active":true,"delta_db":-0.5293519897224582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4261322714398465e-06,"output_power":1.1434102200642557e-06,"reference_active":true,"delta_db":-0.9595773776823433}],"active_band_count":3,"active_band_mae_db":0.5902592284415511}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":1.4806173580162725,"rms_envelope_nrmse":0.11478563678078867,"envelope_correlation":0.9902154069526653,"rms_gain_error_db":-0.2891037227398625,"reference_mean_power":0.0004376816454019563,"output_mean_power":0.00040949439545609327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.844102522756596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.867945533757683e-05,"output_power":9.366622712020905e-05,"reference_active":true,"delta_db":-0.22643716549438336},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.286345958459762e-05,"output_power":1.0227689426317261e-05,"reference_active":true,"delta_db":-0.9958025456925006},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.692984829070904e-06,"output_power":2.3966236471793554e-06,"reference_active":true,"delta_db":-0.5063406669017901}],"active_band_count":3,"active_band_mae_db":0.5761934593628913}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":0.9687092943568719,"rms_envelope_nrmse":0.07299188369777394,"envelope_correlation":0.9963938044651038,"rms_gain_error_db":-0.2891156256668047,"reference_mean_power":0.00043987971266632326,"output_mean_power":0.00041154977645986},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7052935118245518e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.166386964932703e-05,"output_power":8.700783210681438e-05,"reference_active":true,"delta_db":-0.22639839437947407},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4533799725351494e-06,"output_power":4.665319528837919e-06,"reference_active":true,"delta_db":-0.6778436483598609},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.547364135494233e-07,"output_power":1.3011665875203116e-07,"reference_active":true,"delta_db":-0.7525962408086231}],"active_band_count":3,"active_band_mae_db":0.5522794278493194}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012710775449688842,"log_rms_mae_db":1.6578142326267344,"rms_envelope_nrmse":0.16849651644976096,"envelope_correlation":0.9842011811806215,"rms_gain_error_db":0.36606391713483477,"reference_mean_power":0.0003026125473377707,"output_mean_power":0.000329225391732275},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.325105138813022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.305521750251846e-05,"output_power":4.219193854463578e-05,"reference_active":true,"delta_db":-0.08796307631836411},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.253216912773389e-05,"output_power":1.107932822118024e-05,"reference_active":true,"delta_db":-0.5351281888338886},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7715638592807243e-06,"output_power":1.6582655988294174e-06,"reference_active":true,"delta_db":-0.28702720622086464}],"active_band_count":3,"active_band_mae_db":0.3033728237910391}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010399786917892937,"log_rms_mae_db":1.026011177526607,"rms_envelope_nrmse":0.08438355088753986,"envelope_correlation":0.9967944127542206,"rms_gain_error_db":0.36606031698734715,"reference_mean_power":0.0003041332985815956,"output_mean_power":0.0003308796090656666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.904406483303463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.33923405223367e-05,"output_power":4.231534502618954e-05,"reference_active":true,"delta_db":-0.10915189671829044},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.902501195013478e-06,"output_power":5.367312358914031e-06,"reference_active":true,"delta_db":-0.4127921322646668},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.630835506393022e-07,"output_power":1.337902401634204e-07,"reference_active":true,"delta_db":-0.8598572497163945}],"active_band_count":3,"active_band_mae_db":0.4606004262331173}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.3789920806884766,"predicted_mos":3.3789920806884766,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"88655b787f39bafc9a0f74337edb608364281b6ea76b29d8eb4b7f9ef153fb1f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.629446315999303},"started_at":"2026-10-13T04:41:29.160372+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.414899972791318e-05,"native_input_samples":96000,"native_input_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","output_sha256":"8798f1191986e59ea164c14e8d43db332900d82d498f1dedc1f95bedd96ef23d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2203951190003863,"synthesis_s":0.09690786599912826,"total_s":1.3173029849995146,"rtf":0.2195504974999191,"output_peak":0.6323670275119461,"output_rms":0.0907861182284021,"finished_at":"2026-10-13T04:41:30.483252+00:00","evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"8798f1191986e59ea164c14e8d43db332900d82d498f1dedc1f95bedd96ef23d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.244723560076275,"spectral_convergence":0.5263777842755997,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.99575013532353,"spectral_convergence":0.28498506541254137,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.57694617957467,"spectral_convergence":0.30409074666834673,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.002376491578798787,"delta_db":0.44620803488742145,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.003971384505077065,"delta_db":1.0315692749392638,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.0008733836921996,"delta_db":0.9685719819740425,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.0005382474744996166,"delta_db":0.8942566936903791,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.00016989820698662479,"delta_db":0.5883298392414714,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.00031271380541970966,"delta_db":-0.09132623341300977,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.670043676357598,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7840056757597956,"pitch":null,"output_rms_dbfs":-20.839611055503312,"output_peak":0.6323670275119461,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.605806624991492},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.244723560076275,"spectral_convergence":0.5263777842755997,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.99575013532353,"spectral_convergence":0.28498506541254137,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.57694617957467,"spectral_convergence":0.30409074666834673,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.002376491578798787,"delta_db":0.44620803488742145,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.003971384505077065,"delta_db":1.0315692749392638,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.0008733836921996,"delta_db":0.9685719819740425,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.0005382474744996166,"delta_db":0.8942566936903791,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.00016989820698662479,"delta_db":0.5883298392414714,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.00031271380541970966,"delta_db":-0.09132623341300977,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.670043676357598,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7840056757597956,"pitch":null,"output_rms_dbfs":-20.839611055503312,"output_peak":0.6323670275119461,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.605806624991492},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9491287938173277,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08295052965772477,"rms_dbfs":-21.62361673126311,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.770507654342886,"mean":0.00047166951497395836,"zero_energy":false},"output":{"rms":0.0907861182284021,"rms_dbfs":-20.839611055503315,"peak":0.6323670275119461,"peak_dbfs":-3.980615657242372,"crest_factor_db":16.85899539826094,"mean":6.669569177656591e-06,"zero_energy":false},"rms_difference_db":0.7840056757597935},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":363,"output_estimated_voiced_frames":393,"both_estimated_voiced_frames":359,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9889807162534435,"absolute_difference_cents_median":6.404346189373562,"absolute_difference_cents_p95":22.325953023165926,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020729159449367856,"log_rms_mae_db":2.088518337544595,"rms_envelope_nrmse":0.14819062909118302,"envelope_correlation":0.981649229463496,"rms_gain_error_db":0.7840306373130419,"reference_mean_power":0.0068864212769201745,"output_mean_power":0.00824891162659433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4588043207823824e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006447788388285564,"output_power":0.0007978069827473283,"reference_active":true,"delta_db":0.924870573185525},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001410929626621107,"output_power":0.0002154804167573534,"reference_active":true,"delta_db":1.8390245398435305},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.775085636154685e-05,"output_power":3.6882282572206496e-05,"reference_active":true,"delta_db":1.2354140132234637}],"active_band_count":3,"active_band_mae_db":1.3331030420841732}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001768834026970723,"log_rms_mae_db":1.792189332643047,"rms_envelope_nrmse":0.12408863372411148,"envelope_correlation":0.9903950269164901,"rms_gain_error_db":0.7835099486525333,"reference_mean_power":0.006918665749586704,"output_mean_power":0.008286542165982026},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2539809640994155e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005542590263527942,"output_power":0.0006935711233389636,"reference_active":true,"delta_db":0.973782285805592},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.178340930501843e-05,"output_power":0.00011023978799399861,"reference_active":true,"delta_db":1.8631428834437538},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4823995625387325e-06,"output_power":2.0451157487566885e-06,"reference_active":true,"delta_db":1.397526149394156}],"active_band_count":3,"active_band_mae_db":1.4114837728811673}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016054516244843413,"log_rms_mae_db":1.9910301944047657,"rms_envelope_nrmse":0.22007561718575674,"envelope_correlation":0.9836324309779668,"rms_gain_error_db":1.3571733849034764,"reference_mean_power":0.0021921303983993327,"output_mean_power":0.00299628915926442},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.01765222997539e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005084502896759698,"output_power":0.0007483977467968273,"reference_active":true,"delta_db":1.6788397220032607},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.688692893438752e-05,"output_power":0.00017563442187152736,"reference_active":true,"delta_db":2.5834444540012145},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1104523408685175e-05,"output_power":2.762282910794116e-05,"reference_active":true,"delta_db":1.1689260743348133}],"active_band_count":3,"active_band_mae_db":1.810403416779763}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013960343413021469,"log_rms_mae_db":1.6721253128238132,"rms_envelope_nrmse":0.20710464798328182,"envelope_correlation":0.9872727676880639,"rms_gain_error_db":1.3571631075850505,"reference_mean_power":0.00220313483790695,"output_mean_power":0.0030113233294170897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.111026246795902e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004662873703819768,"output_power":0.0006982087646494425,"reference_active":true,"delta_db":1.7533164369410312},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.729339985490381e-05,"output_power":8.772071385908874e-05,"reference_active":true,"delta_db":2.683016213014337},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4261322714398465e-06,"output_power":1.640095266524793e-06,"reference_active":true,"delta_db":0.6070926770863629}],"active_band_count":3,"active_band_mae_db":1.6811417756805769}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":1.381546908922301,"rms_envelope_nrmse":0.15004679434622012,"envelope_correlation":0.9911443228964469,"rms_gain_error_db":0.8336912475003186,"reference_mean_power":0.0004376816454019563,"output_mean_power":0.0005303071213631781},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.844102522756596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.867945533757683e-05,"output_power":0.00012178619688184744,"reference_active":true,"delta_db":0.913713248801571},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.286345958459762e-05,"output_power":1.5340303403003105e-05,"reference_active":true,"delta_db":0.7647616289271277},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.692984829070904e-06,"output_power":2.840929316994914e-06,"reference_active":true,"delta_db":0.2322652168553986}],"active_band_count":3,"active_band_mae_db":0.6369133648613657}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":1.0038618310604837,"rms_envelope_nrmse":0.12538358462913016,"envelope_correlation":0.9957513500363387,"rms_gain_error_db":0.8336976143286772,"reference_mean_power":0.00043987971266632326,"output_mean_power":0.0005329711414692229},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7052935118245518e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.166386964932703e-05,"output_power":0.00011362023927226213,"reference_active":true,"delta_db":0.9325751220346941},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4533799725351494e-06,"output_power":6.564671847866134e-06,"reference_active":true,"delta_db":0.8054726276402934},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.547364135494233e-07,"output_power":1.9535343267048232e-07,"reference_active":true,"delta_db":1.0122852022757736}],"active_band_count":3,"active_band_mae_db":0.9167776506502537}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012710775449688842,"log_rms_mae_db":1.2917477292568933,"rms_envelope_nrmse":0.1535728406726673,"envelope_correlation":0.9854755167096557,"rms_gain_error_db":0.07325243919969919,"reference_mean_power":0.0003026125473377707,"output_mean_power":0.00030776000141517556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.325105138813022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.305521750251846e-05,"output_power":4.3513644887195235e-05,"reference_active":true,"delta_db":0.04599675658774095},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.253216912773389e-05,"output_power":1.2433615299374297e-05,"reference_active":true,"delta_db":-0.03428821255869196},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7715638592807243e-06,"output_power":1.4792051389533146e-06,"reference_active":true,"delta_db":-0.7832840494556649}],"active_band_count":3,"active_band_mae_db":0.2878563395340326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010399786917892937,"log_rms_mae_db":0.7564930890509903,"rms_envelope_nrmse":0.08464736673039276,"envelope_correlation":0.995464808031993,"rms_gain_error_db":0.07325311300572972,"reference_mean_power":0.0003041332985815956,"output_mean_power":0.0003093066687000596},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.904406483303463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.33923405223367e-05,"output_power":4.4065990481240916e-05,"reference_active":true,"delta_db":0.06690460303483471},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.902501195013478e-06,"output_power":5.849977613156124e-06,"reference_active":true,"delta_db":-0.03881879561390542},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.630835506393022e-07,"output_power":1.4425912294947932e-07,"reference_active":true,"delta_db":-0.5326687081359966}],"active_band_count":3,"active_band_mae_db":0.2127973689282456}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"facodec","status":"measured","mos":{"status":"measured","value":4.314762592315674,"predicted_mos":4.314762592315674,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ce31bdbc766e226b1d7a0fd6c31842943a78b32ae6eed807329b64c056d21ab0","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.972569053999905},"started_at":"2026-10-13T04:48:09.716077+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5766000362636987e-05,"native_input_samples":96000,"native_input_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","output_sha256":"a2cae3d54f8b59329068cb76b906a1f65cad2483afc40aacaea9b03ffae9ec0d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.9177786919999562,"synthesis_s":5.835255445000257,"total_s":7.753034137000213,"rtf":1.2921723561667022,"output_peak":0.4562281370162964,"output_rms":0.07602374962292191,"finished_at":"2026-10-13T04:48:17.474863+00:00","evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"a2cae3d54f8b59329068cb76b906a1f65cad2483afc40aacaea9b03ffae9ec0d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.063213901093302,"spectral_convergence":0.25279836273872375,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.579094107240279,"spectral_convergence":0.2528534332818314,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.764829936220514,"spectral_convergence":0.2632025106539879,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.0015884468774453114,"delta_db":-1.303427809421038,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.0028940993401038355,"delta_db":-0.34271575300810825,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.0006272794587132817,"delta_db":-0.46886814236353747,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.0003162542184298563,"delta_db":-1.4152000422702118,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.0001384189349595472,"delta_db":-0.3016030834692707,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.0002151116770768856,"delta_db":-1.716156916887746,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9246619579033187,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7573975497185841,"pitch":null,"output_rms_dbfs":-22.38101428098169,"output_peak":0.4562281370162964,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.469045981518033},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.063213901093302,"spectral_convergence":0.25279836273872375,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.579094107240279,"spectral_convergence":0.2528534332818314,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.764829936220514,"spectral_convergence":0.2632025106539879,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.0015884468774453114,"delta_db":-1.303427809421038,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.0028940993401038355,"delta_db":-0.34271575300810825,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.0006272794587132817,"delta_db":-0.46886814236353747,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.0003162542184298563,"delta_db":-1.4152000422702118,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.0001384189349595472,"delta_db":-0.3016030834692707,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.0002151116770768856,"delta_db":-1.716156916887746,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9246619579033187,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7573975497185841,"pitch":null,"output_rms_dbfs":-22.38101428098169,"output_peak":0.4562281370162964,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.469045981518033},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.936440763221228,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08295052965772477,"rms_dbfs":-21.62361673126311,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.770507654342886,"mean":0.00047166951497395836,"zero_energy":false},"output":{"rms":0.0760237496229219,"rms_dbfs":-22.38101428098169,"peak":0.4562281370162964,"peak_dbfs":-6.816358678684473,"crest_factor_db":15.564655602297218,"mean":0.0005270258173730818,"zero_energy":false},"rms_difference_db":-0.7573975497185828},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":363,"output_estimated_voiced_frames":359,"both_estimated_voiced_frames":351,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9669421487603306,"absolute_difference_cents_median":3.595585037358692,"absolute_difference_cents_p95":13.796032637248015,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020729159449367856,"log_rms_mae_db":1.4054660683869789,"rms_envelope_nrmse":0.14239674905172012,"envelope_correlation":0.9758092512708245,"rms_gain_error_db":-0.7574765531954041,"reference_mean_power":0.0068864212769201745,"output_mean_power":0.0057842350374847544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4588043207823824e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006447788388285564,"output_power":0.000539499462501229,"reference_active":true,"delta_db":-0.774197594674158},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001410929626621107,"output_power":0.00013658245673129772,"reference_active":true,"delta_db":-0.14110432673136825},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.775085636154685e-05,"output_power":2.6558357432407496e-05,"reference_active":true,"delta_db":-0.19075178020135064}],"active_band_count":3,"active_band_mae_db":0.36868456720229226}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001768834026970723,"log_rms_mae_db":1.237528795914117,"rms_envelope_nrmse":0.13701443274732117,"envelope_correlation":0.9759570832866837,"rms_gain_error_db":-0.7584310439781418,"reference_mean_power":0.006918665749586704,"output_mean_power":0.005810041645236968},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2539809640994155e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005542590263527942,"output_power":0.00045986260183821046,"reference_active":true,"delta_db":-0.810846823548395},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.178340930501843e-05,"output_power":6.791467604108001e-05,"reference_active":true,"delta_db":-0.24060447602313492},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4823995625387325e-06,"output_power":1.4738946473621245e-06,"reference_active":true,"delta_db":-0.024988364879375977}],"active_band_count":3,"active_band_mae_db":0.3588132214836353}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016054516244843413,"log_rms_mae_db":1.529566706025387,"rms_envelope_nrmse":0.10339740999700395,"envelope_correlation":0.9902405018115485,"rms_gain_error_db":-0.3102685349351552,"reference_mean_power":0.0021921303983993327,"output_mean_power":0.0020409836750714476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.01765222997539e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005084502896759698,"output_power":0.00046330520019063096,"reference_active":true,"delta_db":-0.40381324296494947},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.688692893438752e-05,"output_power":0.00010332457122950698,"reference_active":true,"delta_db":0.279384216368144},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1104523408685175e-05,"output_power":2.0653483147688202e-05,"reference_active":true,"delta_db":-0.09382244553955676}],"active_band_count":3,"active_band_mae_db":0.25900663495755005}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013960343413021469,"log_rms_mae_db":1.1048928231058548,"rms_envelope_nrmse":0.08815525896366859,"envelope_correlation":0.9926178974303614,"rms_gain_error_db":-0.31029516789457556,"reference_mean_power":0.00220313483790695,"output_mean_power":0.002051216782771765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.111026246795902e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004662873703819768,"output_power":0.00042740739883703917,"reference_active":true,"delta_db":-0.3781161658630726},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.729339985490381e-05,"output_power":5.078932080499754e-05,"reference_active":true,"delta_db":0.3097186919071593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4261322714398465e-06,"output_power":1.345273509421192e-06,"reference_active":true,"delta_db":-0.2534921715967313}],"active_band_count":3,"active_band_mae_db":0.3137756764556544}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":1.9117142763482589,"rms_envelope_nrmse":0.18330235601879005,"envelope_correlation":0.9860931550042085,"rms_gain_error_db":-1.263946864249319,"reference_mean_power":0.0004376816454019563,"output_mean_power":0.00032716259781057603},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.844102522756596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.867945533757683e-05,"output_power":7.744412708348005e-05,"reference_active":true,"delta_db":-1.0523825477880715},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.286345958459762e-05,"output_power":1.084723957775116e-05,"reference_active":true,"delta_db":-0.7403855403615343},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.692984829070904e-06,"output_power":2.3581470203361635e-06,"reference_active":true,"delta_db":-0.5766302882220969}],"active_band_count":3,"active_band_mae_db":0.7897994587905676}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":1.564532402078573,"rms_envelope_nrmse":0.17199334680295006,"envelope_correlation":0.9891673986156548,"rms_gain_error_db":-1.2639669114250862,"reference_mean_power":0.00043987971266632326,"output_mean_power":0.0003288041130581661},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7052935118245518e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.166386964932703e-05,"output_power":7.21259593160573e-05,"reference_active":true,"delta_db":-1.0410658475271768},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4533799725351494e-06,"output_power":4.5561213680572e-06,"reference_active":true,"delta_db":-0.7807047437157439},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.547364135494233e-07,"output_power":1.3713066684674392e-07,"reference_active":true,"delta_db":-0.5245793871913972}],"active_band_count":3,"active_band_mae_db":0.782116659478106}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012710775449688842,"log_rms_mae_db":1.6721770679325738,"rms_envelope_nrmse":0.25094844387608134,"envelope_correlation":0.9696073258429172,"rms_gain_error_db":-1.3606274712725874,"reference_mean_power":0.0003026125473377707,"output_mean_power":0.00022121989621876013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.325105138813022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.305521750251846e-05,"output_power":4.405540939544624e-05,"reference_active":true,"delta_db":0.09973453879888071},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.253216912773389e-05,"output_power":1.5462242941278753e-05,"reference_active":true,"delta_db":0.9124624531637893},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7715638592807243e-06,"output_power":2.900673402850265e-06,"reference_active":true,"delta_db":2.1414202089727254}],"active_band_count":3,"active_band_mae_db":1.0512057336451317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010399786917892937,"log_rms_mae_db":1.3538637453666655,"rms_envelope_nrmse":0.2263156422032797,"envelope_correlation":0.9777237719812191,"rms_gain_error_db":-1.3606700247934524,"reference_mean_power":0.0003041332985815956,"output_mean_power":0.0002223294377923063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.904406483303463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.33923405223367e-05,"output_power":4.602796207696972e-05,"reference_active":true,"delta_db":0.25608670601407124},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.902501195013478e-06,"output_power":7.424858613310029e-06,"reference_active":true,"delta_db":0.9965210440682228},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.630835506393022e-07,"output_power":2.764010765113963e-07,"reference_active":true,"delta_db":2.2912957187068606}],"active_band_count":3,"active_band_mae_db":1.1813011562630515}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"hifiglot","status":"measured","mos":{"status":"measured","value":4.114757537841797,"predicted_mos":4.114757537841797,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"d2d4a47398a3be46feeaa50f208f5dd1ff5e0c4895f1b5065b6398a35f22974a","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6708519109997724},"started_at":"2026-10-13T07:05:59.861764+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009769373000381165,"native_input_samples":264600,"native_input_sha256":"5821220e883e8cd7be1d9480d243118510412b20fdd8339ca4325d0f867059a2","output_sha256":"048dfe170ef75647e9220313819ac69e866bae8c7271f1c8799823f6768cbcff","output_samples":264704,"length_delta_samples":104,"analysis_s":0.17219722899972112,"synthesis_s":4.95012237499941,"total_s":5.122319603999131,"rtf":0.8537199339998551,"output_peak":0.5014533400535583,"output_rms":0.0763191368940073,"finished_at":"2026-10-13T07:06:05.018038+00:00","evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"048dfe170ef75647e9220313819ac69e866bae8c7271f1c8799823f6768cbcff","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.343246772107875,"output_peak":0.49880456302838205},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.964243061843485,"spectral_convergence":0.43895716097282533,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.13056606189056,"spectral_convergence":0.38294228391238544,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.99988258189827,"spectral_convergence":0.4421123522118158,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.0014173367116732268,"delta_db":-1.798424398935504,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.0030464628941253014,"delta_db":-0.11989116727285926,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.000729423712314226,"delta_db":0.18631997049264093,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.00028336000838818215,"delta_db":-1.8921777468444614,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.00013997888949791653,"delta_db":-0.2529326788186606,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.00021582346010296213,"delta_db":-1.7018102695628308,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9919260386544928,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.717920921768213,"pitch":null,"output_rms_dbfs":-22.34153765303132,"output_peak":0.49880456302838205,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.031563901877439},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.844652574339341,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08295052965772477,"rms_dbfs":-21.62361673126311,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.770507654342886,"mean":0.00047166951497395836,"zero_energy":false},"output":{"rms":0.07637005745776164,"rms_dbfs":-22.34153765303132,"peak":0.49880456302838205,"peak_dbfs":-6.041391645610764,"crest_factor_db":16.300146007420558,"mean":-0.00023003712453788236,"zero_energy":false},"rms_difference_db":-0.7179209217682114},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":363,"output_estimated_voiced_frames":358,"both_estimated_voiced_frames":338,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.931129476584022,"absolute_difference_cents_median":10.15237542627695,"absolute_difference_cents_p95":60.89524412341775,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020729159449367856,"log_rms_mae_db":1.6043021035602725,"rms_envelope_nrmse":0.12482506055829236,"envelope_correlation":0.985448677286799,"rms_gain_error_db":-0.717865909541521,"reference_mean_power":0.0068864212769201745,"output_mean_power":0.005837232558723197},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4588043207823824e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006447788388285564,"output_power":0.0006423589498476153,"reference_active":true,"delta_db":-0.01632996371816527},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001410929626621107,"output_power":0.00013775218787056358,"reference_active":true,"delta_db":-0.10406847524338439},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.775085636154685e-05,"output_power":2.2034200614753704e-05,"reference_active":true,"delta_db":-1.001790902911078}],"active_band_count":3,"active_band_mae_db":0.37406311395754255}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001768834026970723,"log_rms_mae_db":1.3232764329745448,"rms_envelope_nrmse":0.10411551155118756,"envelope_correlation":0.9928240668585259,"rms_gain_error_db":-0.7175555142545313,"reference_mean_power":0.006918665749586704,"output_mean_power":0.005864983549452157},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2539809640994155e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005542590263527942,"output_power":0.0005524056919719077,"reference_active":true,"delta_db":-0.014546296342322532},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.178340930501843e-05,"output_power":6.913713502717034e-05,"reference_active":true,"delta_db":-0.16312702157099998},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4823995625387325e-06,"output_power":1.2725664209265806e-06,"reference_active":true,"delta_db":-0.662848187794839}],"active_band_count":3,"active_band_mae_db":0.2801738352360538}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016054516244843413,"log_rms_mae_db":2.662312268740309,"rms_envelope_nrmse":0.1808519030776896,"envelope_correlation":0.9737425284280824,"rms_gain_error_db":0.5447391557293669,"reference_mean_power":0.0021921303983993327,"output_mean_power":0.00248507957605328},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.01765222997539e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005084502896759698,"output_power":0.0005725649942388942,"reference_active":true,"delta_db":0.5157629328890811},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.688692893438752e-05,"output_power":8.544565244788429e-05,"reference_active":true,"delta_db":-0.5457521978718901},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1104523408685175e-05,"output_power":1.7759574841112472e-05,"reference_active":true,"delta_db":-0.7494298449106812}],"active_band_count":3,"active_band_mae_db":0.6036483252238841}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013960343413021469,"log_rms_mae_db":1.924711741714056,"rms_envelope_nrmse":0.15376563182652125,"envelope_correlation":0.9805405830774424,"rms_gain_error_db":0.5445711611743336,"reference_mean_power":0.00220313483790695,"output_mean_power":0.0024974580040922484},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.111026246795902e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004662873703819768,"output_power":0.0005194962540911912,"reference_active":true,"delta_db":0.46928767817787165},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.729339985490381e-05,"output_power":4.2046067377799436e-05,"reference_active":true,"delta_db":-0.5107515402936611},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4261322714398465e-06,"output_power":1.014365426588234e-06,"reference_active":true,"delta_db":-1.479653690973379}],"active_band_count":3,"active_band_mae_db":0.8198976364816373}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":3.754643998777509,"rms_envelope_nrmse":0.25426722292200676,"envelope_correlation":0.9712827828138426,"rms_gain_error_db":-1.7878994543377513,"reference_mean_power":0.0004376816454019563,"output_mean_power":0.00028998022934598926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.844102522756596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.867945533757683e-05,"output_power":6.81217318336387e-05,"reference_active":true,"delta_db":-1.6094106339158667},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.286345958459762e-05,"output_power":9.741342981711265e-06,"reference_active":true,"delta_db":-1.2073895171746647},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.692984829070904e-06,"output_power":2.945759362538091e-06,"reference_active":true,"delta_db":0.38963359857675406}],"active_band_count":3,"active_band_mae_db":1.0688112498890951}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":2.971122170338856,"rms_envelope_nrmse":0.2273088627124709,"envelope_correlation":0.9825119261079724,"rms_gain_error_db":-1.787876741609944,"reference_mean_power":0.00043987971266632326,"output_mean_power":0.0002914380541195539},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7052935118245518e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.166386964932703e-05,"output_power":6.368198108106126e-05,"reference_active":true,"delta_db":-1.5818162192190075},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4533799725351494e-06,"output_power":4.043852169072522e-06,"reference_active":true,"delta_db":-1.2987048814424018},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.547364135494233e-07,"output_power":1.756060796606811e-07,"reference_active":true,"delta_db":0.5494702090106848}],"active_band_count":3,"active_band_mae_db":1.1433304365573647}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012710775449688842,"log_rms_mae_db":3.6592095506583986,"rms_envelope_nrmse":0.22427197544477864,"envelope_correlation":0.979794780126163,"rms_gain_error_db":-1.4192970452445026,"reference_mean_power":0.0003026125473377707,"output_mean_power":0.00021825149479476826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.325105138813022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.305521750251846e-05,"output_power":3.32051772241031e-05,"reference_active":true,"delta_db":-1.1281998498658656},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.253216912773389e-05,"output_power":9.174786126779703e-06,"reference_active":true,"delta_db":-1.3543029805121338},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7715638592807243e-06,"output_power":1.1718579802618166e-06,"reference_active":true,"delta_db":-1.7948183001369236}],"active_band_count":3,"active_band_mae_db":1.425773710171641}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010399786917892937,"log_rms_mae_db":3.2213930910247637,"rms_envelope_nrmse":0.19670525367052408,"envelope_correlation":0.9879513349256052,"rms_gain_error_db":-1.4192816261007524,"reference_mean_power":0.0003041332985815956,"output_mean_power":0.00021934907617938897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.904406483303463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.33923405223367e-05,"output_power":3.326071983598483e-05,"reference_active":true,"delta_db":-1.1548143191558609},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.902501195013478e-06,"output_power":4.39057241228655e-06,"reference_active":true,"delta_db":-1.285149394509755},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.630835506393022e-07,"output_power":1.1081294737581095e-07,"reference_active":true,"delta_db":-1.6781965201134652}],"active_band_count":3,"active_band_mae_db":1.3727200779263604}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"golf","status":"measured","mos":{"status":"measured","value":3.078092336654663,"predicted_mos":3.078092336654663,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"c8d0547a137dd590e8005dff4824adfccba78681b841177a0a1546f56ab645b9","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3198485960001562},"started_at":"2026-10-13T05:02:10.615174+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032698840004741214,"native_input_samples":144000,"native_input_sha256":"da63ddd0065074804e30721613e39f5c4b0ab35ef21c2b02d41014deb49b80d6","output_sha256":"a255e92df78ccc5a9de65ecaf46c34d416ffa92efa1b408407a15dc5ba0c8b20","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.7220990510004412,"synthesis_s":0.11631276899970544,"total_s":0.8384118200001467,"rtf":0.1397353033333578,"output_peak":0.6277903318405151,"output_rms":0.08240274971181001,"finished_at":"2026-10-13T05:02:11.467261+00:00","analysis_stages":{"external_f0_s":0.09448925600008806},"evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"a255e92df78ccc5a9de65ecaf46c34d416ffa92efa1b408407a15dc5ba0c8b20","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.680914732194488,"output_peak":0.630755442094506},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.260117885986051,"spectral_convergence":0.48921626785006805,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.303624271431343,"spectral_convergence":0.2588713382089897,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4996168386816295,"spectral_convergence":0.2738737262418811,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0021474111101487904,"output_power":0.0020532488156015096,"delta_db":-0.19473614726126365,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00313696022356068,"output_power":0.0032273196340419814,"delta_db":0.12332968192043461,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006999763263605853,"output_power":0.0006960937139397992,"delta_db":-0.02415640259259954,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004388213268225449,"output_power":0.0003640023891696147,"delta_db":-0.8118349203448821,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000148624472642103,"output_power":0.00012426747134028346,"delta_db":-0.7773286546680662,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00031989462413578487,"output_power":0.0003256738809465824,"delta_db":0.07775987754091701,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3348576140546939,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06417086364824204,"pitch":null,"output_rms_dbfs":-21.680914732194488,"output_peak":0.630755442094506,"reference_rms_dbfs":-21.616743868546244,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.687786332033007},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9351442706937577,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08301619169577998,"rms_dbfs":-21.616743868546244,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.763634791626025,"mean":0.00047588475757528824,"zero_energy":false},"output":{"rms":0.0824051327590689,"rms_dbfs":-21.680914732194484,"peak":0.630755442094506,"peak_dbfs":-4.002779874902388,"crest_factor_db":17.678134857292097,"mean":0.000282458456354843,"zero_energy":false},"rms_difference_db":-0.06417086364823987},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":362,"output_estimated_voiced_frames":376,"both_estimated_voiced_frames":358,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.988950276243094,"absolute_difference_cents_median":3.957758252183078,"absolute_difference_cents_p95":12.658141435598438,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00020729159449367856,"log_rms_mae_db":1.4001241000431222,"rms_envelope_nrmse":0.11499255804388973,"envelope_correlation":0.9790388842692574,"rms_gain_error_db":-0.06427148626000671,"reference_mean_power":0.006897175076123719,"output_mean_power":0.006795854961917542},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4573249969183746e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006444039902745231,"output_power":0.0007173491492214115,"reference_active":true,"delta_db":0.465723665769887},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014107186456560066,"output_power":0.0001587728708595127,"reference_active":true,"delta_db":0.513358910622646},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7545803422812777e-05,"output_power":3.6098415254848627e-05,"reference_active":true,"delta_db":1.1743269262659695}],"active_band_count":3,"active_band_mae_db":0.7178031675528341}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001768834026970723,"log_rms_mae_db":0.8014203135568179,"rms_envelope_nrmse":0.07015996181785787,"envelope_correlation":0.9914194279998765,"rms_gain_error_db":-0.06459843283950922,"reference_mean_power":0.006928465544259041,"output_mean_power":0.0068261718615002355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2525102025921929e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000554004101948127,"output_power":0.0006199945781684229,"reference_active":true,"delta_db":0.48874911287885237},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.160809649440295e-05,"output_power":8.316149476054728e-05,"reference_active":true,"delta_db":0.6496015747369205},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.475079881485453e-06,"output_power":2.0572075150850203e-06,"reference_active":true,"delta_db":1.4446256241485527}],"active_band_count":3,"active_band_mae_db":0.8609921039214418}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00016054516244843413,"log_rms_mae_db":1.581450966733493,"rms_envelope_nrmse":0.1454621812373207,"envelope_correlation":0.9813647376675545,"rms_gain_error_db":0.35177457325338307,"reference_mean_power":0.002195785029212415,"output_mean_power":0.0023810432112987075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0178360622708142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005085363260312031,"output_power":0.0005354887899855917,"reference_active":true,"delta_db":0.22428402551641763},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.71057655935914e-05,"output_power":0.00011117199938305572,"reference_active":true,"delta_db":0.587503997057694},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0717756384717797e-05,"output_power":2.334958116384731e-05,"reference_active":true,"delta_db":0.5193637276293024}],"active_band_count":3,"active_band_mae_db":0.44371725006780466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013960343413021469,"log_rms_mae_db":1.0327115655674268,"rms_envelope_nrmse":0.10923191004958362,"envelope_correlation":0.9892270089145095,"rms_gain_error_db":0.35176179918345296,"reference_mean_power":0.002206805192038297,"output_mean_power":0.002392986105662261},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.111849573163542e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046643599119271196,"output_power":0.0004933856340045814,"reference_active":true,"delta_db":0.24394445780319346},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.707071909583374e-05,"output_power":5.6596006722729874e-05,"reference_active":true,"delta_db":0.8003495646022489},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4100083438354296e-06,"output_power":1.5245095677563158e-06,"reference_active":true,"delta_db":0.3390847168860917}],"active_band_count":3,"active_band_mae_db":0.4611262464305114}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":1.5757351334185292,"rms_envelope_nrmse":0.16669826173609062,"envelope_correlation":0.9808908819244984,"rms_gain_error_db":-0.661305821518023,"reference_mean_power":0.000438411134072676,"output_mean_power":0.00037648787406444046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.8442174070083068e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.863962654990803e-05,"output_power":8.860178806934153e-05,"reference_active":true,"delta_db":-0.46608932893787763},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2860571132382426e-05,"output_power":1.1026638617786929e-05,"reference_active":true,"delta_db":-0.6681711441505879},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.678380735942325e-06,"output_power":5.557900694641353e-06,"reference_active":true,"delta_db":3.1703847009469603}],"active_band_count":3,"active_band_mae_db":1.4348817246784753}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":1.0090597428359311,"rms_envelope_nrmse":0.1335998408236453,"envelope_correlation":0.9887629194494757,"rms_gain_error_db":-0.6613219600142517,"reference_mean_power":0.0004406166894224758,"output_mean_power":0.0003783805002812702},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.705408822505444e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.163901169244841e-05,"output_power":8.174521981994527e-05,"reference_active":true,"delta_db":-0.49618030744336117},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.433903181856091e-06,"output_power":4.745684403347725e-06,"reference_active":true,"delta_db":-0.588130426749895},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5369214536206926e-07,"output_power":3.039310228574483e-07,"reference_active":true,"delta_db":2.9612335878339646}],"active_band_count":3,"active_band_mae_db":1.3485147740090735}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00012710775449688842,"log_rms_mae_db":2.1682640353817666,"rms_envelope_nrmse":0.17629500308903223,"envelope_correlation":0.9823385899919256,"rms_gain_error_db":0.07976793391596598,"reference_mean_power":0.00030311790990912867,"output_mean_power":0.0003087367947481603},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.32395929030589e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3082031843258484e-05,"output_power":3.7809458223910554e-05,"reference_active":true,"delta_db":-0.5669572303934014},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.236736271487868e-05,"output_power":1.1460093696440631e-05,"reference_active":true,"delta_db":-0.3308892974567894},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7676742841801943e-06,"output_power":2.8444346324107154e-06,"reference_active":true,"delta_db":2.065937138258989}],"active_band_count":3,"active_band_mae_db":0.9879278887030599}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010399786917892937,"log_rms_mae_db":1.8572409518326316,"rms_envelope_nrmse":0.11585431872868177,"envelope_correlation":0.9927596513629818,"rms_gain_error_db":0.07977891686675237,"reference_mean_power":0.0003046431423979852,"output_mean_power":0.00031029108511168037},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.902987347825966e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3368915644751185e-05,"output_power":3.908659984214808e-05,"reference_active":true,"delta_db":-0.4515067096010551},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.76174306076353e-06,"output_power":5.469825186523406e-06,"reference_active":true,"delta_db":-0.2258044076986766},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6203765027292713e-07,"output_power":2.77329150509127e-07,"reference_active":true,"delta_db":2.3337958456079133}],"active_band_count":3,"active_band_mae_db":1.003702320969215}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.251533269882202,"predicted_mos":3.251533269882202,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"74203483fcd2020ecb9f494dda163786174a14742ff1f107c8ca46df77d31cd5","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1677414549994865},"started_at":"2026-10-13T05:07:02.906141+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003185759999723814,"native_input_samples":144000,"native_input_sha256":"da63ddd0065074804e30721613e39f5c4b0ab35ef21c2b02d41014deb49b80d6","output_sha256":"689d04d3bacc47018aeb009a23f751df909b5e0f47d53a15872ed59ea2ad5213","output_samples":144120,"length_delta_samples":120,"analysis_s":1.7832597249998798,"synthesis_s":3.099520334999397,"total_s":4.882780059999277,"rtf":0.8137966766665462,"output_peak":0.43459099531173706,"output_rms":0.0820015595845406,"finished_at":"2026-10-13T05:07:07.802404+00:00","analysis_stages":{"external_f0_s":0.9951931860005061},"evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"689d04d3bacc47018aeb009a23f751df909b5e0f47d53a15872ed59ea2ad5213","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.72212823002699,"output_peak":0.4349090660407768},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.317585515468803,"spectral_convergence":0.4972581082047002,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.2909212921289885,"spectral_convergence":0.28799911619307994,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.2837842242622814,"spectral_convergence":0.27753108698579537,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.002113362776982,"delta_db":-0.06341422351222226,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.0031644668424112956,"delta_db":0.04515539935569758,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.0006734730322473448,"delta_db":-0.16027670291190838,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.00040742864447647733,"delta_db":-0.3150477235641353,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.00013276599842285234,"delta_db":-0.48268945876479424,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.0002405702776418203,"delta_db":-1.2303770845134112,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38282676543702815,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09489860378221068,"pitch":null,"output_rms_dbfs":-21.718515335045318,"output_peak":0.4349090660407768,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.630763677286691},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9267004104728235,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08295052965772477,"rms_dbfs":-21.62361673126311,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.770507654342886,"mean":0.00047166951497395836,"zero_energy":false},"output":{"rms":0.0820491777666406,"rms_dbfs":-21.718515335045318,"peak":0.4349090660407768,"peak_dbfs":-7.232030780220836,"crest_factor_db":14.486484554824484,"mean":0.00023134446788606296,"zero_energy":false},"rms_difference_db":-0.09489860378220882},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":363,"output_estimated_voiced_frames":385,"both_estimated_voiced_frames":361,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9944903581267218,"absolute_difference_cents_median":5.992096305311196,"absolute_difference_cents_p95":25.631218655789965,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020729159449367856,"log_rms_mae_db":2.388756963886004,"rms_envelope_nrmse":0.1239843685999172,"envelope_correlation":0.9756561044242301,"rms_gain_error_db":-0.09484865276378303,"reference_mean_power":0.0068864212769201745,"output_mean_power":0.006737654264727241},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4588043207823824e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006447788388285564,"output_power":0.0006718905212718709,"reference_active":true,"delta_db":0.1788773849985033},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001410929626621107,"output_power":0.00014244280949774182,"reference_active":true,"delta_db":0.04135178123551065},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.775085636154685e-05,"output_power":2.6353054252108167e-05,"reference_active":true,"delta_db":-0.22445433431725562}],"active_band_count":3,"active_band_mae_db":0.14822783351708987}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001768834026970723,"log_rms_mae_db":1.8625970758629373,"rms_envelope_nrmse":0.07980433727675504,"envelope_correlation":0.9891107379317028,"rms_gain_error_db":-0.09443865042379011,"reference_mean_power":0.006918665749586704,"output_mean_power":0.006769841248711229},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2539809640994155e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005542590263527942,"output_power":0.0005747657455116219,"reference_active":true,"delta_db":0.15778102932158672},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.178340930501843e-05,"output_power":7.464458974826236e-05,"reference_active":true,"delta_db":0.16974254455585702},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4823995625387325e-06,"output_power":1.5452368539971302e-06,"reference_active":true,"delta_db":0.18029779398238066}],"active_band_count":3,"active_band_mae_db":0.16927378928660816}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016054516244843413,"log_rms_mae_db":1.9107906124871183,"rms_envelope_nrmse":0.15470873289406406,"envelope_correlation":0.9763825419653144,"rms_gain_error_db":0.037950671474641845,"reference_mean_power":0.0021921303983993327,"output_mean_power":0.0022113701939604556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.01765222997539e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005084502896759698,"output_power":0.0005214324401536997,"reference_active":true,"delta_db":0.10949547568368219},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.688692893438752e-05,"output_power":9.353264774035457e-05,"reference_active":true,"delta_db":-0.15301961497896557},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1104523408685175e-05,"output_power":1.994199568894347e-05,"reference_active":true,"delta_db":-0.24606931144525754}],"active_band_count":3,"active_band_mae_db":0.16952813403596845}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013960343413021469,"log_rms_mae_db":1.1486693750503978,"rms_envelope_nrmse":0.09691746715997616,"envelope_correlation":0.9898284521188245,"rms_gain_error_db":0.03796433447865659,"reference_mean_power":0.00220313483790695,"output_mean_power":0.002222478208714681},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.111026246795902e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004662873703819768,"output_power":0.00047778315446626456,"reference_active":true,"delta_db":0.10577180921933547},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.729339985490381e-05,"output_power":4.589708941402767e-05,"reference_active":true,"delta_db":-0.13015390529816848},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4261322714398465e-06,"output_power":1.3795208458265945e-06,"reference_active":true,"delta_db":-0.14431540034919255}],"active_band_count":3,"active_band_mae_db":0.12674703828889883}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":1.9445143957851247,"rms_envelope_nrmse":0.16806629485086946,"envelope_correlation":0.9777927281897679,"rms_gain_error_db":-0.25823702389890585,"reference_mean_power":0.0004376816454019563,"output_mean_power":0.0004124151703341938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.844102522756596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.867945533757683e-05,"output_power":9.477821257730155e-05,"reference_active":true,"delta_db":-0.17518229651571351},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.286345958459762e-05,"output_power":1.1420981260831229e-05,"reference_active":true,"delta_db":-0.516543673931599},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.692984829070904e-06,"output_power":2.571202287186892e-06,"reference_active":true,"delta_db":-0.20097661078900836}],"active_band_count":3,"active_band_mae_db":0.2975675270787736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":1.1450931926435872,"rms_envelope_nrmse":0.10218295495449363,"envelope_correlation":0.9917193197891854,"rms_gain_error_db":-0.258230381573996,"reference_mean_power":0.00043987971266632326,"output_mean_power":0.0004144869815639437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7052935118245518e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.166386964932703e-05,"output_power":8.765261213578739e-05,"reference_active":true,"delta_db":-0.194333242832168},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4533799725351494e-06,"output_power":4.778285680802925e-06,"reference_active":true,"delta_db":-0.5739364738407189},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.547364135494233e-07,"output_power":1.485604007033555e-07,"reference_active":true,"delta_db":-0.17689464514348152}],"active_band_count":3,"active_band_mae_db":0.3150547872721228}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012710775449688842,"log_rms_mae_db":2.0919691279576376,"rms_envelope_nrmse":0.22396713132524612,"envelope_correlation":0.9717908127795136,"rms_gain_error_db":-0.8797431314270715,"reference_mean_power":0.0003026125473377707,"output_mean_power":0.0002471226874373393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.325105138813022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.305521750251846e-05,"output_power":3.606783238712098e-05,"reference_active":true,"delta_db":-0.7690574472981184},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.253216912773389e-05,"output_power":9.550033809017965e-06,"reference_active":true,"delta_db":-1.1802133818978096},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7715638592807243e-06,"output_power":1.8848982194292021e-06,"reference_active":true,"delta_db":0.26931092283635794}],"active_band_count":3,"active_band_mae_db":0.7395272506774285}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010399786917892937,"log_rms_mae_db":1.26847329789722,"rms_envelope_nrmse":0.15302194151181917,"envelope_correlation":0.9900218163582799,"rms_gain_error_db":-0.879741887806839,"reference_mean_power":0.0003041332985815956,"output_mean_power":0.0002483646506654997},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.904406483303463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.33923405223367e-05,"output_power":3.6067112138917556e-05,"reference_active":true,"delta_db":-0.8030170585607215},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.902501195013478e-06,"output_power":4.6011209644399075e-06,"reference_active":true,"delta_db":-1.08172432569177},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.630835506393022e-07,"output_power":1.510766712210741e-07,"reference_active":true,"delta_db":-0.3321275115647094}],"active_band_count":3,"active_band_mae_db":0.7389562986057335}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8545145988464355,"predicted_mos":2.8545145988464355,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"04c212c8277f99f3697c6b085a33f403853b5d5baae226a33f17ea9b8243f276","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.201089038000646},"started_at":"2026-10-13T05:10:15.920341+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005681753000317258,"native_input_samples":144000,"native_input_sha256":"da63ddd0065074804e30721613e39f5c4b0ab35ef21c2b02d41014deb49b80d6","output_sha256":"0c5b6ec0bd684bd4dea0e9a580af516f356e4e4168d31724ef6586c77988b2f4","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0016353680002794,"synthesis_s":0.1555876000002172,"total_s":1.1572229680004966,"rtf":0.19287049466674944,"output_peak":0.49552252888679504,"output_rms":0.08164666416966498,"finished_at":"2026-10-13T05:10:17.096525+00:00","analysis_stages":{"external_f0_s":0.2378572850011551},"evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"0c5b6ec0bd684bd4dea0e9a580af516f356e4e4168d31724ef6586c77988b2f4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.76029956954141,"output_peak":0.4888771900087767},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.243030313373662,"spectral_convergence":0.22344433443704717,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.089004695124653,"spectral_convergence":0.18818855844198068,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.595702847222504,"spectral_convergence":0.21732631003011635,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00214444778551191,"output_power":0.002124438224926144,"delta_db":-0.04071369143942581,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003131735025595309,"output_power":0.0030096364060912444,"delta_db":-0.1727097783457067,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006987919290953902,"output_power":0.000689452192982029,"delta_db":-0.05843723004235907,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004380829302633605,"output_power":0.0004258938624528519,"delta_db":-0.12254949876121843,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014837331900785606,"output_power":0.00014005877774531336,"delta_db":-0.25045479601838566,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003193593810232492,"output_power":0.0002892275042125732,"delta_db":-0.4304008786129523,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.179210978870008,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12945937877332214,"pitch":null,"output_rms_dbfs":-21.75307611003643,"output_peak":0.4888771900087767,"reference_rms_dbfs":-21.62361673126311,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.309245951906941},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9701852978624621,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08295052965772477,"rms_dbfs":-21.62361673126311,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.770507654342886,"mean":0.00047166951497395836,"zero_energy":false},"output":{"rms":0.08172335632125101,"rms_dbfs":-21.75307611003643,"peak":0.4888771900087767,"peak_dbfs":-6.2160045107846305,"crest_factor_db":15.537071599251801,"mean":-0.00040201154956651217,"zero_energy":false},"rms_difference_db":-0.1294593787733227},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":363,"output_estimated_voiced_frames":362,"both_estimated_voiced_frames":356,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9807162534435262,"absolute_difference_cents_median":0.8932584232928347,"absolute_difference_cents_p95":5.627329367459598,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020729159449367856,"log_rms_mae_db":0.7191179374510612,"rms_envelope_nrmse":0.050950098123538984,"envelope_correlation":0.9960464770407231,"rms_gain_error_db":-0.12944352241263374,"reference_mean_power":0.0068864212769201745,"output_mean_power":0.006684196903912386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4588043207823824e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006447788388285564,"output_power":0.0006414565237211049,"reference_active":true,"delta_db":-0.022435494609280876},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001410929626621107,"output_power":0.00013231057470273488,"reference_active":true,"delta_db":-0.27910797013475425},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.775085636154685e-05,"output_power":2.161566102459296e-05,"reference_active":true,"delta_db":-1.0850786837237045}],"active_band_count":3,"active_band_mae_db":0.4622073828225799}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001768834026970723,"log_rms_mae_db":0.3825944076609508,"rms_envelope_nrmse":0.02862343823683411,"envelope_correlation":0.9989310399053568,"rms_gain_error_db":-0.12938435286725192,"reference_mean_power":0.006918665749586704,"output_mean_power":0.0067155859905197995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2539809640994155e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005542590263527942,"output_power":0.0005543671660759488,"reference_active":true,"delta_db":0.0008472555584613174},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.178340930501843e-05,"output_power":6.79968024812905e-05,"reference_active":true,"delta_db":-0.235355903271128},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4823995625387325e-06,"output_power":1.1457030759079782e-06,"reference_active":true,"delta_db":-1.1189319908934972}],"active_band_count":3,"active_band_mae_db":0.4517117165743622}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016054516244843413,"log_rms_mae_db":0.8887681005560952,"rms_envelope_nrmse":0.09542562936916556,"envelope_correlation":0.9912918346057454,"rms_gain_error_db":-0.21087753941090984,"reference_mean_power":0.0021921303983993327,"output_mean_power":0.002088231442042587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.01765222997539e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005084502896759698,"output_power":0.0004903128244186156,"reference_active":true,"delta_db":-0.15775246477166077},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.688692893438752e-05,"output_power":9.114822964528222e-05,"reference_active":true,"delta_db":-0.26516952256020143},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1104523408685175e-05,"output_power":1.829016876185645e-05,"reference_active":true,"delta_db":-0.6215783648490081}],"active_band_count":3,"active_band_mae_db":0.34816678406029006}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013960343413021469,"log_rms_mae_db":0.45082409093039366,"rms_envelope_nrmse":0.05795616956514661,"envelope_correlation":0.9968842469575523,"rms_gain_error_db":-0.21088066071260325,"reference_mean_power":0.00220313483790695,"output_mean_power":0.0020987128030418656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.111026246795902e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004662873703819768,"output_power":0.0004515751657475714,"reference_active":true,"delta_db":-0.13923602534872684},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.729339985490381e-05,"output_power":4.517409961600172e-05,"reference_active":true,"delta_db":-0.19911030734045526},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4261322714398465e-06,"output_power":1.0175201239284703e-06,"reference_active":true,"delta_db":-1.4661680027911208}],"active_band_count":3,"active_band_mae_db":0.6015047784934343}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":0.7675570461101134,"rms_envelope_nrmse":0.06872671051546944,"envelope_correlation":0.9964547195610176,"rms_gain_error_db":-0.15113920520333865,"reference_mean_power":0.0004376816454019563,"output_mean_power":0.0004227118414558183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.844102522756596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.867945533757683e-05,"output_power":9.388501887159876e-05,"reference_active":true,"delta_db":-0.21630445825312924},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.286345958459762e-05,"output_power":1.1314955636214788e-05,"reference_active":true,"delta_db":-0.5570493084356668},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.692984829070904e-06,"output_power":2.1409536019919553e-06,"reference_active":true,"delta_db":-0.9962665130439018}],"active_band_count":3,"active_band_mae_db":0.5898734265775659}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":0.42588161212287196,"rms_envelope_nrmse":0.038645346777113926,"envelope_correlation":0.9989872418608514,"rms_gain_error_db":-0.15113485660823212,"reference_mean_power":0.00043987971266632326,"output_mean_power":0.00042483515472494887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7052935118245518e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.166386964932703e-05,"output_power":8.741117199554713e-05,"reference_active":true,"delta_db":-0.20631244088479433},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4533799725351494e-06,"output_power":4.905508198416594e-06,"reference_active":true,"delta_db":-0.4598175299696704},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.547364135494233e-07,"output_power":1.1789429591988609e-07,"reference_active":true,"delta_db":-1.1809973351295313}],"active_band_count":3,"active_band_mae_db":0.6157091019946653}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012710775449688842,"log_rms_mae_db":0.9331997978544516,"rms_envelope_nrmse":0.09787789425753214,"envelope_correlation":0.9942946194861768,"rms_gain_error_db":-0.22387681982613844,"reference_mean_power":0.0003026125473377707,"output_mean_power":0.0002874082626230148},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.325105138813022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.305521750251846e-05,"output_power":3.9684217455515625e-05,"reference_active":true,"delta_db":-0.3540796675442353},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.253216912773389e-05,"output_power":1.0662454778432195e-05,"reference_active":true,"delta_db":-0.7016904501386971},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7715638592807243e-06,"output_power":1.6687345992440525e-06,"reference_active":true,"delta_db":-0.25969541273241753}],"active_band_count":3,"active_band_mae_db":0.43848851013844997}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010399786917892937,"log_rms_mae_db":0.5567009247928416,"rms_envelope_nrmse":0.0471967018426443,"envelope_correlation":0.9989554135209192,"rms_gain_error_db":-0.22387638242010638,"reference_mean_power":0.0003041332985815956,"output_mean_power":0.0002888526352387032},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.904406483303463e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.33923405223367e-05,"output_power":3.9911959314663756e-05,"reference_active":true,"delta_db":-0.36310027778209303},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.902501195013478e-06,"output_power":5.103520049746349e-06,"reference_active":true,"delta_db":-0.631662583930226},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.630835506393022e-07,"output_power":1.2915227500373257e-07,"reference_active":true,"delta_db":-1.013080976722275}],"active_band_count":3,"active_band_mae_db":0.669281279478198}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0039","method":"magphase","status":"measured","mos":{"status":"measured","value":3.9499869346618652,"predicted_mos":3.9499869346618652,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287011,"input_sample_rate":48000,"input_seconds":5.979395833333333,"input_float32_sha256":"9457d4d87f09bb43c607d6cc149ea183d3235c335ea3833f241a76fed6773297","internal_samples":95671,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0484451669999544},"started_at":"2026-10-13T05:13:28.695829+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008607050000136951,"native_input_samples":288000,"native_input_sha256":"a9ce7415fa74932412d6c3e67af9e262e44b50d02e88cda45af71b9392e53144","output_sha256":"73d0c512ca3e37c6515789b9a9754db6648e8a16bd81a73f04bd14a02208517e","output_samples":287011,"length_delta_samples":-989,"analysis_s":3.6082854630003567,"synthesis_s":1.1040532579991122,"total_s":4.712338720999469,"rtf":0.7853897868332448,"output_peak":0.5407441170725078,"output_rms":0.08008081882402543,"finished_at":"2026-10-13T05:13:33.428816+00:00","evaluation_input":{"reference":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"73d0c512ca3e37c6515789b9a9754db6648e8a16bd81a73f04bd14a02208517e","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95671,"length_delta_samples":-329,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.928899169464763,"output_peak":0.5225931714678389},"common_prefix":{"samples":95671,"sample_rate":16000,"reference_coverage":0.9965729166666667,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95671,"output_samples":95671,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.889716481245434,"spectral_convergence":0.40026370550148593,"spectral_convergence_status":"ok","amplitude_floor":1.4935395043887774e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.752768532033652,"spectral_convergence":0.1940813837188079,"spectral_convergence_status":"ok","amplitude_floor":1.0494802571211602e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.076567389827179,"spectral_convergence":0.1980867252229809,"spectral_convergence_status":"ok","amplitude_floor":9.412333460941768e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0021520167909949367,"output_power":0.0018827132069013592,"delta_db":-0.5806148645099405,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031403142363770407,"output_power":0.0029026580968603056,"delta_db":-0.3417722459972854,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007011052654563912,"output_power":0.0006746643716788671,"delta_db":-0.16695452649579498,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00043949018486453287,"output_power":0.0004479667264552899,"delta_db":0.08296576763983085,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014888480226902444,"output_power":0.00016649211645508252,"delta_db":0.4854330562474658,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003204488802597337,"output_power":0.0003392267633152157,"delta_db":0.24731350251469658,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3175089939008357,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.318812416526414,"pitch":null,"output_rms_dbfs":-21.928899169464763,"output_peak":0.5225931714678389,"reference_rms_dbfs":-21.610086752938344,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.239684134368755},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95671,"output_samples":95671,"intelligibility":{"estoi":{"value":0.9625349574049361,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08307984208110691,"rms_dbfs":-21.610086752938344,"peak":0.509735107421875,"peak_dbfs":-5.853109076920221,"crest_factor_db":15.756977676018124,"mean":0.00047460063737313294,"zero_energy":false},"output":{"rms":0.08008571209439372,"rms_dbfs":-21.928899169464763,"peak":0.5225931714678389,"peak_dbfs":-5.6367253867476155,"crest_factor_db":16.292173782717146,"mean":0.0032405426082140034,"zero_energy":false},"rms_difference_db":-0.3188124165264199},"full_scale":{"reference":{"sample_count":95671,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95671,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":597,"uncovered_tail_samples":151,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":3.941215087543242e-06,"threshold_dbfs":-54.04369863336833,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":590,"reference_estimated_voiced_frames":360,"output_estimated_voiced_frames":350,"both_estimated_voiced_frames":344,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9555555555555556,"absolute_difference_cents_median":1.8232362906469364,"absolute_difference_cents_p95":6.935677596860878,"difference_over_600_cents_frames":1,"uncovered_tail_samples":151,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95671,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1194,"rms_floor":0.00020729159449367856,"log_rms_mae_db":1.2851505392917988,"rms_envelope_nrmse":0.08089109781803935,"envelope_correlation":0.9914349539555871,"rms_gain_error_db":-0.31852417710067077,"reference_mean_power":0.006911473809425537,"output_mean_power":0.006422709051271471},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":1.45535805324733e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439389628815784,"output_power":0.0006149765432960368,"reference_active":true,"delta_db":-0.19986152746915384},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014087043896073392,"output_power":0.00013538859385477274,"reference_active":true,"delta_db":-0.17237790208755985},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7342965205519544e-05,"output_power":2.737156861218081e-05,"reference_active":true,"delta_db":0.004540768948031045}],"active_band_count":3,"active_band_mae_db":0.12559339950158158}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1188,"rms_floor":0.0001768834026970723,"log_rms_mae_db":0.8741497859481474,"rms_envelope_nrmse":0.05228820762603669,"envelope_correlation":0.9974992118337902,"rms_gain_error_db":-0.318322819593094,"reference_mean_power":0.006942757842426064,"output_mean_power":0.006452079882959277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":1.250382299113575e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005537028252392527,"output_power":0.0005272532637878098,"reference_active":true,"delta_db":-0.21257462705202557},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.128821181126967e-05,"output_power":6.944532398200062e-05,"reference_active":true,"delta_db":-0.11374712678378794},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4864507069209313e-06,"output_power":1.2877947940557165e-06,"reference_active":true,"delta_db":-0.6230384676842395}],"active_band_count":3,"active_band_mae_db":0.316453407173351}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1194,"rms_floor":0.00016054516244843413,"log_rms_mae_db":1.2084207832541702,"rms_envelope_nrmse":0.07238813485883107,"envelope_correlation":0.9953295505669061,"rms_gain_error_db":-0.22942318986950022,"reference_mean_power":0.0022012774193300648,"output_mean_power":0.0020880094488199526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":1.0181406995914003e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000508664391190348,"output_power":0.0004637635771270469,"reference_active":true,"delta_db":-0.4013469891605328},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.738980660113488e-05,"output_power":9.11267116630736e-05,"reference_active":true,"delta_db":-0.2886780450368314},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0157215306055344e-05,"output_power":1.847268863209336e-05,"reference_active":true,"delta_db":-0.37900424670859834}],"active_band_count":3,"active_band_mae_db":0.35634309363532085}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1188,"rms_floor":0.00013960343413021469,"log_rms_mae_db":0.7166020787337409,"rms_envelope_nrmse":0.047859302147622615,"envelope_correlation":0.9982009004066081,"rms_gain_error_db":-0.22946911773394238,"reference_mean_power":0.002212293810151605,"output_mean_power":0.002098436793364284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":9.113491788243329e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666799495038692,"output_power":0.00042459824123542567,"reference_active":true,"delta_db":-0.41040951297652056},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.662326569935427e-05,"output_power":4.4561652491402925e-05,"reference_active":true,"delta_db":-0.19641402632700045},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4541701796762719e-06,"output_power":1.2384624475832477e-06,"reference_active":true,"delta_db":-0.6973239189606626}],"active_band_count":3,"active_band_mae_db":0.43471581942139453}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1194,"rms_floor":9.530788197178379e-05,"log_rms_mae_db":0.9524245641312776,"rms_envelope_nrmse":0.07149601926404953,"envelope_correlation":0.9962139285680081,"rms_gain_error_db":0.12467406513270818,"reference_mean_power":0.00043950923576440885,"output_mean_power":0.00045230919040415454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":1.8443947527735381e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.859000273356408e-05,"output_power":0.00010270732448863235,"reference_active":true,"delta_db":0.17768537401072707},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2850031697394328e-05,"output_power":1.316621652917861e-05,"reference_active":true,"delta_db":0.1055679419176059},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642602882527956e-06,"output_power":3.064969220112576e-06,"reference_active":true,"delta_db":0.6439421310486947}],"active_band_count":3,"active_band_mae_db":0.30906514899234255}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1188,"rms_floor":9.347481502172348e-05,"log_rms_mae_db":0.5379077558442233,"rms_envelope_nrmse":0.038593845239205525,"envelope_correlation":0.998950741312281,"rms_gain_error_db":0.12467684356551453,"reference_mean_power":0.0004417274006417925,"output_mean_power":0.0004545922463679016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":1.7055961310619746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.16095224605036e-05,"output_power":9.514601913004117e-05,"reference_active":true,"delta_db":0.16450003032802915},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.401021683473013e-06,"output_power":5.598223879597048e-06,"reference_active":true,"delta_db":0.15574341581598664},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.549806581967859e-07,"output_power":1.6171988806303763e-07,"reference_active":true,"delta_db":0.1848593106545503}],"active_band_count":3,"active_band_mae_db":0.16836758559952203}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1194,"rms_floor":0.00012710775449688842,"log_rms_mae_db":1.4278459982385248,"rms_envelope_nrmse":0.17686363213013842,"envelope_correlation":0.9829862899162793,"rms_gain_error_db":0.4315968030770393,"reference_mean_power":0.00030387758381410625,"output_mean_power":0.00033562812358131953},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":8.321935382222033e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.314816336427805e-05,"output_power":4.5435328548720186e-05,"reference_active":true,"delta_db":0.22431358482974884},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2082127810151176e-05,"output_power":1.0977866772222054e-05,"reference_active":true,"delta_db":-0.4162546971455241},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7589959297992567e-06,"output_power":1.9060595706430788e-06,"reference_active":true,"delta_db":0.34871635118415945}],"active_band_count":3,"active_band_mae_db":0.3297615443864775}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":71,"metrics":{"frame_count":1188,"rms_floor":0.00010399786917892937,"log_rms_mae_db":1.1016136025154004,"rms_envelope_nrmse":0.11962265165033589,"envelope_correlation":0.9928638909745923,"rms_gain_error_db":0.4315943882052648,"reference_mean_power":0.0003054112159869582,"output_mean_power":0.00033732180919264585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":7.90101686406905e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3330141093920334e-05,"output_power":4.4923690785096113e-05,"reference_active":true,"delta_db":0.1568532607520277},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5456662207745294e-06,"output_power":5.212353252538227e-06,"reference_active":true,"delta_db":-0.2691988585179337},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6446202761958655e-07,"output_power":1.335871351182655e-07,"reference_active":true,"delta_db":-0.9030100412411645}],"active_band_count":3,"active_band_mae_db":0.4430207201703753}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"hnm","status":"measured","mos":{"status":"measured","value":2.133727550506592,"predicted_mos":2.133727550506592,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":38160,"input_sample_rate":16000,"input_seconds":2.385,"input_float32_sha256":"554992762eb0282cd280385e79f8afa70c0916c2ac1119002727fa0b47b74b1c","internal_samples":38160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0229759579997335},"started_at":"2026-10-13T04:38:33.273427+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3030000445724e-05,"native_input_samples":38160,"native_input_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","output_sha256":"1f31d6ead1daedde5c8b89fbcba413bbae5f82be0ee50c5c73654216994efb8e","output_samples":38160,"length_delta_samples":0,"analysis_s":0.3729941769997822,"synthesis_s":0.19538046500019846,"total_s":0.5683746419999807,"rtf":0.23831221886791645,"output_peak":1.2953775984198879,"output_rms":0.08772715241439301,"finished_at":"2026-10-13T04:38:33.844419+00:00","evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"1f31d6ead1daedde5c8b89fbcba413bbae5f82be0ee50c5c73654216994efb8e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.169126287752202,"spectral_convergence":0.6002153965337679,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.771619788415547,"spectral_convergence":0.5844024008372948,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.893781789093306,"spectral_convergence":0.5275514445400181,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.005333438371125816,"delta_db":1.7032106708176098,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.001579252119549835,"delta_db":0.2519854825128432,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.0004374033702874314,"delta_db":-0.009386955287402118,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.0002300331653108863,"delta_db":-0.1491436725831852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":8.777606505197393e-05,"delta_db":0.34347491131582697,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.8150179412198243e-05,"delta_db":0.19490702468816296,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4420181195341717,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1704979462046194,"pitch":null,"output_rms_dbfs":-21.137319348492,"output_peak":1.2953775984198879,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.944842621753685},"common_prefix":{"samples":38160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.169126287752202,"spectral_convergence":0.6002153965337679,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.771619788415547,"spectral_convergence":0.5844024008372948,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.893781789093306,"spectral_convergence":0.5275514445400181,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.005333438371125816,"delta_db":1.7032106708176098,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.001579252119549835,"delta_db":0.2519854825128432,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.0004374033702874314,"delta_db":-0.009386955287402118,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.0002300331653108863,"delta_db":-0.1491436725831852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":8.777606505197393e-05,"delta_db":0.34347491131582697,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.8150179412198243e-05,"delta_db":0.19490702468816296,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4420181195341717,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1704979462046194,"pitch":null,"output_rms_dbfs":-21.137319348492,"output_peak":1.2953775984198879,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.944842621753685},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38160,"output_samples":38160,"intelligibility":{"estoi":{"value":0.8012420591482878,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07666711752372073,"rms_dbfs":-22.307817294696616,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.052362335702647,"mean":0.0005484319083095846,"zero_energy":false},"output":{"rms":0.08772715241439301,"rms_dbfs":-21.137319348492,"peak":1.2953775984198879,"peak_dbfs":2.2479276464409956,"crest_factor_db":23.38524699493299,"mean":0.0008880124274455687,"zero_energy":false},"rms_difference_db":1.1704979462046161},"full_scale":{"reference":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":36,"at_or_above_full_scale_fraction":0.0009433962264150943,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0009433962264150943,"longest_full_scale_run_samples":36,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":1.5121603116675437e-05,"output_rms_dbfs":-48.20402164666852,"output_zero_energy":false,"output_peak":0.014618802081623266,"output_energy_fraction":0.00010709883313202356,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":67,"both_estimated_voiced_frames":67,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":0,"paired_voiced_fraction":0.9178082191780822,"absolute_difference_cents_median":1.4750982379814654,"absolute_difference_cents_p95":7.066717853700859,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.0004033177834583336,"log_rms_mae_db":1.7090784246292297,"rms_envelope_nrmse":0.3567066219884458,"envelope_correlation":0.9306641149322759,"rms_gain_error_db":1.1704823701344338,"reference_mean_power":0.005890133913079112,"output_mean_power":0.007712113378587559},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":4.7027834380481505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019529842144495992,"output_power":0.0027758285091389304,"reference_active":true,"delta_db":1.5269389887762812},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0010897406087412391,"output_power":0.002562198295282399,"reference_active":true,"delta_db":3.712896028983803},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00042203241052064084,"output_power":0.0014913074666327083,"reference_active":true,"delta_db":5.482213878319091}],"active_band_count":3,"active_band_mae_db":3.5740162986930586}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.0002641139400842597,"log_rms_mae_db":0.9887263326612736,"rms_envelope_nrmse":0.3002398554525687,"envelope_correlation":0.9502605712575077,"rms_gain_error_db":1.1705179328572501,"reference_mean_power":0.0059649962331930016,"output_mean_power":0.007810196615641526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.959398727002631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020291985866854774,"output_power":0.0037263943233063323,"reference_active":true,"delta_db":2.639642584338828},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005646977153887027,"output_power":0.0015961668297044163,"reference_active":true,"delta_db":4.512622506919585},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7144804121490974e-05,"output_power":0.00013576199526939265,"reference_active":true,"delta_db":5.628801396108967}],"active_band_count":3,"active_band_mae_db":4.260355495789127}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.00011878678262042868,"log_rms_mae_db":2.356555494913065,"rms_envelope_nrmse":0.12130208450080374,"envelope_correlation":0.9899102774156419,"rms_gain_error_db":-0.04847763837268709,"reference_mean_power":0.0009978964081743123,"output_mean_power":0.0009868194382746944},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":1.0827926032603135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006103304193321489,"output_power":0.0005730502711754231,"reference_active":true,"delta_db":-0.27372293683214516},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011556410212235238,"output_power":0.00010474207703789343,"reference_active":true,"delta_db":-0.4270176766354073},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8931134571049757e-05,"output_power":1.1121456481796404e-05,"reference_active":true,"delta_db":-2.310149758628242}],"active_band_count":3,"active_band_mae_db":1.0036301240319314}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":1.6429676962187676,"rms_envelope_nrmse":0.06674590601999814,"envelope_correlation":0.9968682454573166,"rms_gain_error_db":-0.04848777618547139,"reference_mean_power":0.0010106286473201321,"output_mean_power":0.0009994080125475438},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.735673127654544e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005396175923660179,"output_power":0.0005086504761620719,"reference_active":true,"delta_db":-0.25666644511512027},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.461345336739452e-05,"output_power":5.7864441681727434e-05,"reference_active":true,"delta_db":-0.4791118612064293},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3971113807678808e-06,"output_power":1.1337810295603034e-06,"reference_active":true,"delta_db":-3.2515902771762524}],"active_band_count":3,"active_band_mae_db":1.329122861165934}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":2.2784761068797685,"rms_envelope_nrmse":0.20559883348893632,"envelope_correlation":0.9726094811455677,"rms_gain_error_db":-0.026237405780520764,"reference_mean_power":0.0002825436588167163,"output_mean_power":0.0002808418493480835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":3.405184045067586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020464089970485737,"output_power":0.00019015882752241504,"reference_active":true,"delta_db":-0.31875945552944646},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.521356834551285e-05,"output_power":3.5163196878115314e-05,"reference_active":true,"delta_db":-1.0918043150060532},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018106896586624e-06,"output_power":9.476566205654306e-06,"reference_active":true,"delta_db":0.7257915911939044}],"active_band_count":3,"active_band_mae_db":0.7121184539098012}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":1.6625700130235201,"rms_envelope_nrmse":0.1271585866916343,"envelope_correlation":0.9891575264543183,"rms_gain_error_db":-0.02625175186957283,"reference_mean_power":0.0002861487173387815,"output_mean_power":0.00028442425443454307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.083140770966501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019461067978585592,"output_power":0.00017839715557064277,"reference_active":true,"delta_db":-0.37778744121571395},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0239988981192974e-05,"output_power":1.5676778081668093e-05,"reference_active":true,"delta_db":-1.1095346116240528},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.640717194960497e-07,"output_power":7.056848841698071e-07,"reference_active":true,"delta_db":-0.3452331057699238}],"active_band_count":3,"active_band_mae_db":0.6108517195365635}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":2.258361635192694,"rms_envelope_nrmse":0.1915712951693245,"envelope_correlation":0.977238423148945,"rms_gain_error_db":0.03335170857786272,"reference_mean_power":4.55291382465523e-05,"output_mean_power":4.588012591550931e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":6.179039674292896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.30631712224655e-05,"output_power":3.1775489988401436e-05,"reference_active":true,"delta_db":-0.17252249933335154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2207009907392539e-05,"output_power":8.231432677331265e-06,"reference_active":true,"delta_db":-1.7113386649335405},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.251387780144131e-06,"output_power":9.62111069141493e-07,"reference_active":true,"delta_db":-3.692250932804847}],"active_band_count":3,"active_band_mae_db":1.8587040323572463}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":1.5944399750538585,"rms_envelope_nrmse":0.11929898191502063,"envelope_correlation":0.9907442569445309,"rms_gain_error_db":0.033321018712401965,"reference_mean_power":4.61098055796954e-05,"output_mean_power":4.6464941307861594e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":5.621286439624597e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.22612462378357e-05,"output_power":2.9334826855122715e-05,"reference_active":true,"delta_db":-0.41297610787884576},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2476871573483135e-06,"output_power":4.6749732770852315e-06,"reference_active":true,"delta_db":-1.904203062605171},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.30977962238882e-07,"output_power":1.0109437248609784e-07,"reference_active":true,"delta_db":-3.5884356469258845}],"active_band_count":3,"active_band_mae_db":1.9685382724699672}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.049264430999756,"predicted_mos":2.049264430999756,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":38160,"input_sample_rate":16000,"input_seconds":2.385,"input_float32_sha256":"101cc30ec1b4b1af1590375a05ee946a933b78e38e968b57c9907d13cf34e033","internal_samples":38160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.055186424000567},"started_at":"2026-10-13T04:41:30.484228+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.893799981189659e-05,"native_input_samples":38160,"native_input_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","output_sha256":"6f99e4bc86868bb8eb0b19517d73ca5def287bcfd535d8cc425ccb7f43b53577","output_samples":38160,"length_delta_samples":0,"analysis_s":0.49461431599956995,"synthesis_s":0.04379664000043704,"total_s":0.538410956000007,"rtf":0.22574882851153336,"output_peak":0.551041753074889,"output_rms":0.06525686250916393,"finished_at":"2026-10-13T04:41:31.025747+00:00","evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"6f99e4bc86868bb8eb0b19517d73ca5def287bcfd535d8cc425ccb7f43b53577","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.373189127694517,"spectral_convergence":0.7493038986251717,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.554057611837967,"spectral_convergence":0.7111450796867919,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.983808067300253,"spectral_convergence":0.6727584375204921,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.001185108566852945,"delta_db":-4.829280764797861,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.0021338174932648947,"delta_db":1.559043510082868,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.0005193424123322921,"delta_db":0.7363297032202049,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.00030966715254733264,"delta_db":1.1419031797632075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":8.636248017544789e-05,"delta_db":0.27296489345223934,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.4159999367012132e-05,"delta_db":-0.468935460867726,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5014095853640181,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.3996589159330601,"pitch":null,"output_rms_dbfs":-23.70747621062968,"output_peak":0.551041753074889,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.970351602277579},"common_prefix":{"samples":38160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.373189127694517,"spectral_convergence":0.7493038986251717,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.554057611837967,"spectral_convergence":0.7111450796867919,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.983808067300253,"spectral_convergence":0.6727584375204921,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.001185108566852945,"delta_db":-4.829280764797861,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.0021338174932648947,"delta_db":1.559043510082868,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.0005193424123322921,"delta_db":0.7363297032202049,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.00030966715254733264,"delta_db":1.1419031797632075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":8.636248017544789e-05,"delta_db":0.27296489345223934,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.4159999367012132e-05,"delta_db":-0.468935460867726,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5014095853640181,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.3996589159330601,"pitch":null,"output_rms_dbfs":-23.70747621062968,"output_peak":0.551041753074889,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.970351602277579},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38160,"output_samples":38160,"intelligibility":{"estoi":{"value":0.8587149880995615,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07666711752372073,"rms_dbfs":-22.307817294696616,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.052362335702647,"mean":0.0005484319083095846,"zero_energy":false},"output":{"rms":0.06525686250916393,"rms_dbfs":-23.70747621062968,"peak":0.551041753074889,"peak_dbfs":-5.176309858063201,"crest_factor_db":18.531166352566476,"mean":2.1413473341682645e-08,"zero_energy":false},"rms_difference_db":-1.3996589159330632},"full_scale":{"reference":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":6.193066281118933e-06,"output_rms_dbfs":-52.080942719409904,"output_zero_energy":false,"output_peak":0.008557650555371606,"output_energy_fraction":7.926990141037956e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":92,"both_estimated_voiced_frames":73,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":6.332703818256954,"absolute_difference_cents_p95":26.624562656468107,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.0004033177834583336,"log_rms_mae_db":4.4694898417667694,"rms_envelope_nrmse":0.5138648160620384,"envelope_correlation":0.7560042721881165,"rms_gain_error_db":-1.399615673692104,"reference_mean_power":0.005890133913079112,"output_mean_power":0.0042674024408129035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":4.7027834380481505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019529842144495992,"output_power":0.0014859131903433524,"reference_active":true,"delta_db":-1.1870529507359564},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0010897406087412391,"output_power":0.0003010919375115718,"reference_active":true,"delta_db":-5.586240086568319},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00042203241052064084,"output_power":6.399434766827734e-05,"reference_active":true,"delta_db":-8.192041880369706}],"active_band_count":3,"active_band_mae_db":4.988444972557994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.0002641139400842597,"log_rms_mae_db":4.327279080471463,"rms_envelope_nrmse":0.47763659412950893,"envelope_correlation":0.7690815480663367,"rms_gain_error_db":-1.3994346090616363,"reference_mean_power":0.0059649962331930016,"output_mean_power":0.004321820377244668},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.959398727002631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020291985866854774,"output_power":0.0012718756051162373,"reference_active":true,"delta_db":-2.0287991363586},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005646977153887027,"output_power":0.00014977630495371388,"reference_active":true,"delta_db":-5.763729185549181},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7144804121490974e-05,"output_power":3.237229041543257e-06,"reference_active":true,"delta_db":-10.59724644650478}],"active_band_count":3,"active_band_mae_db":6.129924922804187}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.00011878678262042868,"log_rms_mae_db":3.6137347650648786,"rms_envelope_nrmse":0.3016264133316079,"envelope_correlation":0.9612655016751526,"rms_gain_error_db":1.3189766963740193,"reference_mean_power":0.0009978964081743123,"output_mean_power":0.0013520200405798335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":1.0827926032603135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006103304193321489,"output_power":0.0006831371924043086,"reference_active":true,"delta_db":0.48942914442512697},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011556410212235238,"output_power":0.0001537782333477217,"reference_active":true,"delta_db":1.2407191790180288},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8931134571049757e-05,"output_power":3.803837575871277e-05,"reference_active":true,"delta_db":3.0304532159877406}],"active_band_count":3,"active_band_mae_db":1.5868671798102987}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":3.4949711711426157,"rms_envelope_nrmse":0.265554202184464,"envelope_correlation":0.9716082386587727,"rms_gain_error_db":1.3189240148584975,"reference_mean_power":0.0010106286473201321,"output_mean_power":0.0013692539614873886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.735673127654544e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005396175923660179,"output_power":0.0006051508630390156,"reference_active":true,"delta_db":0.49777557095152536},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.461345336739452e-05,"output_power":8.758005559894716e-05,"reference_active":true,"delta_db":1.320822633513634},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3971113807678808e-06,"output_power":2.6011609944198034e-06,"reference_active":true,"delta_db":0.3547901913125031}],"active_band_count":3,"active_band_mae_db":0.7244627985925541}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":1.5725685760805637,"rms_envelope_nrmse":0.20968466110023457,"envelope_correlation":0.9848066250182888,"rms_gain_error_db":1.0599125171456134,"reference_mean_power":0.0002825436588167163,"output_mean_power":0.00036064242656141894},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":3.405184045067586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020464089970485737,"output_power":0.0002538602477952915,"reference_active":true,"delta_db":0.9360226311704105},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.521356834551285e-05,"output_power":6.617951302047238e-05,"reference_active":true,"delta_db":1.6545478331690182},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018106896586624e-06,"output_power":9.062957396371977e-06,"reference_active":true,"delta_db":0.5319809652032658}],"active_band_count":3,"active_band_mae_db":1.0408504765142315}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":0.9670075429159605,"rms_envelope_nrmse":0.17278887654502464,"envelope_correlation":0.9925524903486753,"rms_gain_error_db":1.0599206047776633,"reference_mean_power":0.0002861487173387815,"output_mean_power":0.0003652446506488885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.083140770966501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019461067978585592,"output_power":0.0002486002159545349,"reference_active":true,"delta_db":1.063348319024357},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0239988981192974e-05,"output_power":2.9776835818011713e-05,"reference_active":true,"delta_db":1.6766827461017921},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.640717194960497e-07,"output_power":6.272652210207446e-07,"reference_active":true,"delta_db":-0.856829169181057}],"active_band_count":3,"active_band_mae_db":1.1989534114357354}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":1.459518950513047,"rms_envelope_nrmse":0.18822132898123298,"envelope_correlation":0.9776125962424526,"rms_gain_error_db":-0.367072387109792,"reference_mean_power":4.55291382465523e-05,"output_mean_power":4.183908629103895e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":6.179039674292896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.30631712224655e-05,"output_power":2.8866098471405923e-05,"reference_active":true,"delta_db":-0.5895641736854362},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2207009907392539e-05,"output_power":1.064703240601396e-05,"reference_active":true,"delta_db":-0.5938072113625839},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.251387780144131e-06,"output_power":1.9496357810565988e-06,"reference_active":true,"delta_db":-0.624968178187588}],"active_band_count":3,"active_band_mae_db":0.6027798544118693}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":0.8074910255049677,"rms_envelope_nrmse":0.11012812503354577,"envelope_correlation":0.9927714231544551,"rms_gain_error_db":-0.3670744773778726,"reference_mean_power":4.61098055796954e-05,"output_mean_power":4.2372671223008714e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":5.621286439624597e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.22612462378357e-05,"output_power":2.8266955241055537e-05,"reference_active":true,"delta_db":-0.5740210875631059},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2476871573483135e-06,"output_power":6.27002704153228e-06,"reference_active":true,"delta_db":-0.6293002510653409},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.30977962238882e-07,"output_power":1.6238763368869181e-07,"reference_active":true,"delta_db":-1.5301759222207147}],"active_band_count":3,"active_band_mae_db":0.9111657536163872}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"facodec","status":"measured","mos":{"status":"measured","value":4.102581977844238,"predicted_mos":4.102581977844238,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":38200,"input_sample_rate":16000,"input_seconds":2.3875,"input_float32_sha256":"aef7d92fd034d0d56829a93929f17ab6a627c77b39a0006843a82ac63a93c58f","internal_samples":38200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.039488349999374},"started_at":"2026-10-13T04:48:17.475442+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.884399989648955e-05,"native_input_samples":38160,"native_input_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","output_sha256":"bf63aaa4e441b05cca93496f62087baf9793428ac6ca68cfc37350bbf705f7d8","output_samples":38200,"length_delta_samples":40,"analysis_s":0.600426607999907,"synthesis_s":3.928632041000128,"total_s":4.529058649000035,"rtf":1.898976372746346,"output_peak":0.3295063078403473,"output_rms":0.05676733030971122,"finished_at":"2026-10-13T04:48:22.008085+00:00","evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"bf63aaa4e441b05cca93496f62087baf9793428ac6ca68cfc37350bbf705f7d8","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":38160,"output_samples":38200,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.91803059145568,"output_peak":0.3295063078403473},"common_prefix":{"samples":38160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9989528795811519,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1492567260552295,"spectral_convergence":0.468593729394167,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.291802181631183,"spectral_convergence":0.4498519282451202,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.517422534154988,"spectral_convergence":0.4331865309512923,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.0014416875134210204,"delta_db":-3.978150770097413,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.0011625246432355298,"delta_db":-1.07850752343671,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.00037406132998666605,"delta_db":-0.6887800690189688,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.00015755582794989195,"delta_db":-1.7927035047507578,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":7.127447810589797e-05,"delta_db":-0.5609457070159706,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":1.8786350561995e-05,"delta_db":-1.5614804244524714,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6100946664620486,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-2.6056869699627834,"pitch":null,"output_rms_dbfs":-24.9135042646594,"output_peak":0.3295063078403473,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.319493813947133},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38160,"output_samples":38160,"intelligibility":{"estoi":{"value":0.8446492606568473,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07666711752372073,"rms_dbfs":-22.307817294696616,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.052362335702647,"mean":0.0005484319083095846,"zero_energy":false},"output":{"rms":0.05679692019168822,"rms_dbfs":-24.9135042646594,"peak":0.3295063078403473,"peak_dbfs":-9.642725343157592,"crest_factor_db":15.270778921501808,"mean":0.0008976446294476452,"zero_energy":false},"rms_difference_db":-2.605686969962786},"full_scale":{"reference":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":1.5404228456407546e-05,"output_rms_dbfs":-48.123600490845305,"output_zero_energy":false,"output_peak":0.010802613571286201,"output_energy_fraction":0.0002602827258573048,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":73,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.9863013698630136,"absolute_difference_cents_median":2.650959641912846,"absolute_difference_cents_p95":15.831201474257254,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.0004033177834583336,"log_rms_mae_db":2.0713610161827325,"rms_envelope_nrmse":0.3968017899516534,"envelope_correlation":0.8958476876526197,"rms_gain_error_db":-2.605705055050938,"reference_mean_power":0.005890133913079112,"output_mean_power":0.0032326200564981203},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":4.7027834380481505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019529842144495992,"output_power":0.0009960787030816813,"reference_active":true,"delta_db":-2.92405078341047},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0010897406087412391,"output_power":0.0002712080852198331,"reference_active":true,"delta_db":-6.0402050248804064},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00042203241052064084,"output_power":6.209960016801686e-05,"reference_active":true,"delta_db":-8.322570004899209}],"active_band_count":3,"active_band_mae_db":5.762275271063362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.0002641139400842597,"log_rms_mae_db":1.7774210794467857,"rms_envelope_nrmse":0.38416426731487524,"envelope_correlation":0.895266627670207,"rms_gain_error_db":-2.605908533926243,"reference_mean_power":0.0059649962331930016,"output_mean_power":0.0032735525753638897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.959398727002631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020291985866854774,"output_power":0.0008279601057617204,"reference_active":true,"delta_db":-3.893151398200174},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005646977153887027,"output_power":0.0001510205672857906,"reference_active":true,"delta_db":-5.727799334084382},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7144804121490974e-05,"output_power":3.7850436188720583e-06,"reference_active":true,"delta_db":-9.918271838148552}],"active_band_count":3,"active_band_mae_db":6.513074190144369}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.00011878678262042868,"log_rms_mae_db":1.9844987512730372,"rms_envelope_nrmse":0.0994154474753103,"envelope_correlation":0.9935180947433218,"rms_gain_error_db":-0.26579947240513724,"reference_mean_power":0.0009978964081743123,"output_mean_power":0.0009386539502609386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":1.0827926032603135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006103304193321489,"output_power":0.0005761382706523005,"reference_active":true,"delta_db":-0.2503829136423905},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011556410212235238,"output_power":0.00010855782902036609,"reference_active":true,"delta_db":-0.27161799827588473},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8931134571049757e-05,"output_power":1.5841077918757915e-05,"reference_active":true,"delta_db":-0.7739191248524815}],"active_band_count":3,"active_band_mae_db":0.4319733455902523}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":1.2261796153708315,"rms_envelope_nrmse":0.07866004907259959,"envelope_correlation":0.99587002569387,"rms_gain_error_db":-0.26578824716008803,"reference_mean_power":0.0010106286473201321,"output_mean_power":0.0009506327673088318},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.735673127654544e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005396175923660179,"output_power":0.0005145670439845495,"reference_active":true,"delta_db":-0.20644132006218852},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.461345336739452e-05,"output_power":5.862496053755624e-05,"reference_active":true,"delta_db":-0.42240389825544955},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3971113807678808e-06,"output_power":2.1366924987352427e-06,"reference_active":true,"delta_db":-0.4994618847149278}],"active_band_count":3,"active_band_mae_db":0.376102367677522}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":1.8334751228798476,"rms_envelope_nrmse":0.20860966237824127,"envelope_correlation":0.9870156096307815,"rms_gain_error_db":-1.5498375581388006,"reference_mean_power":0.0002825436588167163,"output_mean_power":0.00019774331432169945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":3.405184045067586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020464089970485737,"output_power":0.00014276591321344987,"reference_active":true,"delta_db":-1.5636790884164828},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.521356834551285e-05,"output_power":2.7382046700095088e-05,"reference_active":true,"delta_db":-2.178028769661558},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018106896586624e-06,"output_power":6.685247397143766e-06,"reference_active":true,"delta_db":-0.7895435833902595}],"active_band_count":3,"active_band_mae_db":1.5104171471561}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":1.3934349909157777,"rms_envelope_nrmse":0.1910674638693535,"envelope_correlation":0.9920791265741178,"rms_gain_error_db":-1.5498447064371512,"reference_mean_power":0.0002861487173387815,"output_mean_power":0.00020026605041403205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.083140770966501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019461067978585592,"output_power":0.00013217136316488544,"reference_active":true,"delta_db":-1.6802930049847666},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0239988981192974e-05,"output_power":1.2859558425903768e-05,"reference_active":true,"delta_db":-1.9698421578132028},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.640717194960497e-07,"output_power":6.042526653720768e-07,"reference_active":true,"delta_db":-1.019155507194189}],"active_band_count":3,"active_band_mae_db":1.5564302233307192}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":1.6305988456836191,"rms_envelope_nrmse":0.21691435490564473,"envelope_correlation":0.9753427732951674,"rms_gain_error_db":-1.0432769887808329,"reference_mean_power":4.55291382465523e-05,"output_mean_power":3.580648842075563e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":6.179039674292896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.30631712224655e-05,"output_power":2.470814016187834e-05,"reference_active":true,"delta_db":-1.2650444987817338},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2207009907392539e-05,"output_power":8.072408290161835e-06,"reference_active":true,"delta_db":-1.7960617731276414},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.251387780144131e-06,"output_power":2.225304349460204e-06,"reference_active":true,"delta_db":-0.05060887761062044}],"active_band_count":3,"active_band_mae_db":1.0372383831733318}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":1.2372531353923653,"rms_envelope_nrmse":0.1844222725115397,"envelope_correlation":0.9837713516400449,"rms_gain_error_db":-1.0432914134859834,"reference_mean_power":4.61098055796954e-05,"output_mean_power":3.626303504562118e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":5.621286439624597e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.22612462378357e-05,"output_power":2.3754123913755135e-05,"reference_active":true,"delta_db":-1.3294212233727951},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2476871573483135e-06,"output_power":5.042897016025193e-06,"reference_active":true,"delta_db":-1.5751933971170662},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.30977962238882e-07,"output_power":2.1131747080853744e-07,"reference_active":true,"delta_db":-0.3863514142940325}],"active_band_count":3,"active_band_mae_db":1.096988678261298}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.0045413970947266,"predicted_mos":3.0045413970947266,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":105472,"input_sample_rate":44100,"input_seconds":2.391655328798186,"input_float32_sha256":"3bc84543a17e5b14f69a33351cd49dd6974ffd2b57dc2cfb6e4338d3bff03162","internal_samples":38267,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7280698089998623},"started_at":"2026-10-13T07:06:05.019873+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005323188999682316,"native_input_samples":105179,"native_input_sha256":"39212b2156441b2b6f91e7b595a498f0ea762286adcbaa9f0bb79866f721cdc2","output_sha256":"c5dedb77d75b73fb8c165177942b6c64d41f70e22cc42d29bbf828142cf28187","output_samples":105472,"length_delta_samples":293,"analysis_s":0.074596972001018,"synthesis_s":15.982985577998988,"total_s":16.057582550000006,"rtf":6.732707008575859,"output_peak":0.3536963164806366,"output_rms":0.05799394341440562,"finished_at":"2026-10-13T07:06:21.094598+00:00","evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"c5dedb77d75b73fb8c165177942b6c64d41f70e22cc42d29bbf828142cf28187","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":38160,"output_samples":38267,"length_delta_samples":107,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.72912711430557,"output_peak":0.34064917355652197},"common_prefix":{"samples":38160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9972038571092586,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.972923662476672,"spectral_convergence":0.5991740143991288,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.82306345107887,"spectral_convergence":0.5288876596409797,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.79147790269185,"spectral_convergence":0.5219038034026595,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.0013954594086572356,"delta_db":-4.119690058381574,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.0013116701579679955,"delta_db":-0.5542828224377659,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.00038037925030169037,"delta_db":-0.6160400244014665,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.00015471415528465788,"delta_db":-1.8717477223906784,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":0.00010913896956779576,"delta_db":1.289512324343568,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.38655707114969e-05,"delta_db":-0.5221864029037054,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.495576559143126,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-2.4091521807762315,"pitch":null,"output_rms_dbfs":-24.716969475472848,"output_peak":0.34064917355652197,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.862488338749131},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38160,"output_samples":38160,"intelligibility":{"estoi":{"value":0.7039876466163082,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07666711752372073,"rms_dbfs":-22.307817294696616,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.052362335702647,"mean":0.0005484319083095846,"zero_energy":false},"output":{"rms":0.058096708275864245,"rms_dbfs":-24.716969475472848,"peak":0.34064917355652197,"peak_dbfs":-9.353853207163182,"crest_factor_db":15.363116268309668,"mean":-0.0017092887340221044,"zero_energy":false},"rms_difference_db":-2.409152180776232},"full_scale":{"reference":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":1.0847956424796364e-05,"output_rms_dbfs":-49.646520679929125,"output_zero_energy":false,"output_peak":0.010520266344646733,"output_energy_fraction":0.0001751861823025903,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":63,"both_estimated_voiced_frames":58,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.7945205479452054,"absolute_difference_cents_median":14.782827067419488,"absolute_difference_cents_p95":36.11998149041343,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.0004033177834583336,"log_rms_mae_db":2.8187026195624263,"rms_envelope_nrmse":0.3094801180434452,"envelope_correlation":0.9583985103575932,"rms_gain_error_db":-2.4091290142217145,"reference_mean_power":0.005890133913079112,"output_mean_power":0.003382301102957423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":4.7027834380481505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019529842144495992,"output_power":0.001027884219844875,"reference_active":true,"delta_db":-2.78754534204031},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0010897406087412391,"output_power":0.0004094185577591602,"reference_active":true,"delta_db":-4.2515561092064225},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00042203241052064084,"output_power":0.00014958947517611714,"reference_active":true,"delta_db":-4.504447659544153}],"active_band_count":3,"active_band_mae_db":3.847849703596962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.0002641139400842597,"log_rms_mae_db":2.3435259227270224,"rms_envelope_nrmse":0.29096922713208057,"envelope_correlation":0.9674518271335398,"rms_gain_error_db":-2.409120401430091,"reference_mean_power":0.0059649962331930016,"output_mean_power":0.003425296206789259},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.959398727002631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020291985866854774,"output_power":0.0009298423799569341,"reference_active":true,"delta_db":-3.3891521477846083},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005646977153887027,"output_power":0.0002081361078897928,"reference_active":true,"delta_db":-4.3346860163679715},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7144804121490974e-05,"output_power":1.16424588601479e-05,"reference_active":true,"delta_db":-5.038533605104865}],"active_band_count":3,"active_band_mae_db":4.254123923085815}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.00011878678262042868,"log_rms_mae_db":3.7335301022909504,"rms_envelope_nrmse":0.20452932123764783,"envelope_correlation":0.9705299080201093,"rms_gain_error_db":-0.20090256985812271,"reference_mean_power":0.0009978964081743123,"output_mean_power":0.0009527856387523511},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":1.0827926032603135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006103304193321489,"output_power":0.0006132982524653519,"reference_active":true,"delta_db":0.02106711145503043},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011556410212235238,"output_power":7.651529246730192e-05,"reference_active":true,"delta_db":-1.7907470677406725},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8931134571049757e-05,"output_power":2.140649760086725e-05,"reference_active":true,"delta_db":0.5336897386373449}],"active_band_count":3,"active_band_mae_db":0.7818346392776826}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":2.833086718582314,"rms_envelope_nrmse":0.15171924046490384,"envelope_correlation":0.9826531320530738,"rms_gain_error_db":-0.20087985727837437,"reference_mean_power":0.0010106286473201321,"output_mean_power":0.0009649473524600108},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.735673127654544e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005396175923660179,"output_power":0.0005431291727030168,"reference_active":true,"delta_db":0.02817030462594081},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.461345336739452e-05,"output_power":3.9772876595812646e-05,"reference_active":true,"delta_db":-2.1073595049431924},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3971113807678808e-06,"output_power":2.170234343090741e-06,"reference_active":true,"delta_db":-0.4318158207598513}],"active_band_count":3,"active_band_mae_db":0.8557818767763282}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":4.109808825313513,"rms_envelope_nrmse":0.2774376510152624,"envelope_correlation":0.9517327556607945,"rms_gain_error_db":-0.828639293031957,"reference_mean_power":0.0002825436588167163,"output_mean_power":0.0002334649210489122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":3.405184045067586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020464089970485737,"output_power":0.00019449942834389334,"reference_active":true,"delta_db":-0.22074107294532597},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.521356834551285e-05,"output_power":3.2041317139689036e-05,"reference_active":true,"delta_db":-1.4955842322218116},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018106896586624e-06,"output_power":5.168062622838176e-06,"reference_active":true,"delta_db":-1.907440744382528}],"active_band_count":3,"active_band_mae_db":1.2079220165165552}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":3.4319648174337907,"rms_envelope_nrmse":0.21002825146379983,"envelope_correlation":0.9733514999379276,"rms_gain_error_db":-0.8286286322201601,"reference_mean_power":0.0002861487173387815,"output_mean_power":0.0002364443497776514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.083140770966501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019461067978585592,"output_power":0.00017980810793339865,"reference_active":true,"delta_db":-0.3435739855835872},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0239988981192974e-05,"output_power":1.621110671842135e-05,"reference_active":true,"delta_db":-0.9639760696026747},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.640717194960497e-07,"output_power":5.293192013815008e-07,"reference_active":true,"delta_db":-1.5941647690800798}],"active_band_count":3,"active_band_mae_db":0.9672382747554472}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":5.398260330862139,"rms_envelope_nrmse":0.2112533059639327,"envelope_correlation":0.9745368273423584,"rms_gain_error_db":-0.6439684510154025,"reference_mean_power":4.55291382465523e-05,"output_mean_power":3.9254783383418336e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":6.179039674292896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.30631712224655e-05,"output_power":3.0323664523129565e-05,"reference_active":true,"delta_db":-0.3756282286742965},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2207009907392539e-05,"output_power":1.1402725672440473e-05,"reference_active":true,"delta_db":-0.2960062089112174},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.251387780144131e-06,"output_power":1.7331112401114669e-06,"reference_active":true,"delta_db":-1.136238656478474}],"active_band_count":3,"active_band_mae_db":0.602624364687996}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":4.6939790397671395,"rms_envelope_nrmse":0.14693111839501136,"envelope_correlation":0.9895127189713875,"rms_gain_error_db":-0.6439290033693332,"reference_mean_power":4.61098055796954e-05,"output_mean_power":3.97557902577542e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":5.621286439624597e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.22612462378357e-05,"output_power":2.9640952746219636e-05,"reference_active":true,"delta_db":-0.3678898096659774},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2476871573483135e-06,"output_power":6.762115995398018e-06,"reference_active":true,"delta_db":-0.301168227209815},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.30977962238882e-07,"output_power":2.0858047197638017e-07,"reference_active":true,"delta_db":-0.44296899697591263}],"active_band_count":3,"active_band_mae_db":0.37067567795056827}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"golf","status":"measured","mos":{"status":"measured","value":2.682457447052002,"predicted_mos":2.682457447052002,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":57120,"input_sample_rate":24000,"input_seconds":2.38,"input_float32_sha256":"952a09e16c0e2b9c91e5a86842f9b6823a3c1319f47536cd307c3465714b3a1a","internal_samples":38080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4059405160005554},"started_at":"2026-10-13T05:02:11.468258+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0016722210002626525,"native_input_samples":57240,"native_input_sha256":"4f8c6dd430ce539306e95760f94d929084c20a22e6dc488bf29e8faf76f81111","output_sha256":"c368dfb3a059771b8c6e7b3e1c6616d82925c953fa4810660df8f7d18f80a07a","output_samples":57120,"length_delta_samples":-120,"analysis_s":0.24117237600057706,"synthesis_s":0.05841020699972432,"total_s":0.2995825830003014,"rtf":0.12561114591207606,"output_peak":0.4947223663330078,"output_rms":0.06695255371966542,"finished_at":"2026-10-13T05:02:11.774301+00:00","analysis_stages":{"external_f0_s":0.020699505000266072},"evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"c368dfb3a059771b8c6e7b3e1c6616d82925c953fa4810660df8f7d18f80a07a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":38160,"output_samples":38080,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.482295932849443,"output_peak":0.49331989349885685},"common_prefix":{"samples":38080,"sample_rate":16000,"reference_coverage":0.9979035639412998,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38080,"output_samples":38080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.33113834840701,"spectral_convergence":0.6489797084269885,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.415719656078894,"spectral_convergence":0.4841787537241671,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.690469890686198,"spectral_convergence":0.4350975322261519,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0036107005318460296,"output_power":0.0021931343495793284,"delta_db":-2.1652623292902637,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014932732864718157,"output_power":0.0015837184408221663,"delta_db":0.2553867770913463,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004393605533047506,"output_power":0.00041491152339900595,"delta_db":-0.24865565760597977,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023857015938232742,"output_power":0.0002047062060867235,"delta_db":-0.6648511131407321,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.127208986534482e-05,"output_power":6.624981353779054e-05,"delta_db":-0.887567672559064,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6971385873256304e-05,"output_power":2.2361864129485034e-05,"delta_db":-0.8139525804486769,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8392793550226773,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1835580109486437,"pitch":null,"output_rms_dbfs":-23.482295932849443,"output_peak":0.49331989349885685,"reference_rms_dbfs":-22.2987379219008,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.812442631724035},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38080,"output_samples":38080,"intelligibility":{"estoi":{"value":0.8703036486785426,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07674729967069542,"rms_dbfs":-22.298737921900802,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.04328296290683,"mean":0.0005415058937393316,"zero_energy":false},"output":{"rms":0.06697075628626648,"rms_dbfs":-23.482295932849443,"peak":0.49331989349885685,"peak_dbfs":-6.137427418607287,"crest_factor_db":17.344868514242155,"mean":-0.0022949045332128714,"zero_energy":false},"rms_difference_db":-1.1835580109486408},"full_scale":{"reference":{"sample_count":38080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":1.7741064835929945e-05,"output_rms_dbfs":-47.51020316945678,"output_zero_energy":false,"output_peak":0.014428988715256115,"output_energy_fraction":0.00021606064666420817,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":81,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9863013698630136,"absolute_difference_cents_median":5.1895663397400345,"absolute_difference_cents_p95":12.635183015741605,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":475,"rms_floor":0.0004033177834583336,"log_rms_mae_db":2.9204916202660804,"rms_envelope_nrmse":0.3542046908325351,"envelope_correlation":0.8923719839464404,"rms_gain_error_db":-1.1835456357282008,"reference_mean_power":0.005902508732740229,"output_mean_power":0.004494507140296849},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":475,"frequency_resolution_hz":0.42105263157894735,"power_floor":4.717908668477957e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019361135365921196,"output_power":0.0014741250101287426,"reference_active":true,"delta_db":-1.1839650686493532},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0011220077050803647,"output_power":0.0006433251753219073,"reference_active":true,"delta_db":-2.415652922972823},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.000422759607892854,"output_power":0.0003194573398899016,"reference_active":true,"delta_db":-1.2168061543882647}],"active_band_count":3,"active_band_mae_db":1.6054747153368139}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":0.0002641139400842597,"log_rms_mae_db":1.7165341515155697,"rms_envelope_nrmse":0.239925853006726,"envelope_correlation":0.954505068031497,"rms_gain_error_db":-1.183635393338581,"reference_mean_power":0.005977638337910964,"output_mean_power":0.004551621039650358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":3.970761246024513e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002038678675183512,"output_power":0.0014358974372220403,"reference_active":true,"delta_db":-1.5222535979131986},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000567214753167372,"output_power":0.0003124719170250263,"reference_active":true,"delta_db":-2.5893652642680154},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7336080754885705e-05,"output_power":2.377653198525826e-05,"reference_active":true,"delta_db":-1.9598021796079792}],"active_band_count":3,"active_band_mae_db":2.023807013929731}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":475,"rms_floor":0.00011878678262042868,"log_rms_mae_db":1.9991231143020545,"rms_envelope_nrmse":0.13450838829738648,"envelope_correlation":0.9879655658771995,"rms_gain_error_db":0.19578553731199982,"reference_mean_power":0.000999996378265583,"output_mean_power":0.001046109102004317},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":475,"frequency_resolution_hz":0.42105263157894735,"power_floor":1.0829179763617924e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005992319497299148,"output_power":0.0006363097456276833,"reference_active":true,"delta_db":0.26073613942850205},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012763511937137354,"output_power":0.0001207797527754092,"reference_active":true,"delta_db":-0.23976052574724332},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8969742399993153e-05,"output_power":1.927804441542196e-05,"reference_active":true,"delta_db":0.07001543106078095}],"active_band_count":3,"active_band_mae_db":0.19017069874550882}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":1.0257160724745944,"rms_envelope_nrmse":0.09538083936851693,"envelope_correlation":0.9936464681943606,"rms_gain_error_db":0.19580704303362567,"reference_mean_power":0.0010127821454062977,"output_mean_power":0.0010594897042763291},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":9.73231626637206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000540417746873412,"output_power":0.0005776641328232939,"reference_active":true,"delta_db":0.2894580056906647},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.456778602612762e-05,"output_power":6.153693206629847e-05,"reference_active":true,"delta_db":-0.20880054447223828},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.395181403687281e-06,"output_power":2.5504669443645993e-06,"reference_active":true,"delta_db":0.2728128805033435}],"active_band_count":3,"active_band_mae_db":0.25702381022208215}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":475,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":1.6304867399009968,"rms_envelope_nrmse":0.1738218448429791,"envelope_correlation":0.9836361724884585,"rms_gain_error_db":-0.7767036985985458,"reference_mean_power":0.00028313822704949574,"output_mean_power":0.00023677079829510493},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":475,"frequency_resolution_hz":0.42105263157894735,"power_floor":3.4130865998200165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002025656248527448,"output_power":0.00016933715942711326,"reference_active":true,"delta_db":-0.7781347766086871},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.8272947883247473e-05,"output_power":4.1550840023553355e-05,"reference_active":true,"delta_db":-0.6512401244367836},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.934533483604734e-06,"output_power":8.522668687091373e-06,"reference_active":true,"delta_db":0.31054208673733497}],"active_band_count":3,"active_band_mae_db":0.5799723292609352}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":0.9675726434988474,"rms_envelope_nrmse":0.1151625932239864,"envelope_correlation":0.9954650409681244,"rms_gain_error_db":-0.7767074342527596,"reference_mean_power":0.00028675855705298406,"output_mean_power":0.00023979804767816152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":3.0896961056478475e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019538408050750507,"output_power":0.00016583957964013592,"reference_active":true,"delta_db":-0.7120098713973984},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.024390521060236e-05,"output_power":1.7560601187196774e-05,"reference_active":true,"delta_db":-0.6175491521872026},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.654743383219002e-07,"output_power":6.802608855103028e-07,"reference_active":true,"delta_db":-0.5125513596541589}],"active_band_count":3,"active_band_mae_db":0.61403679441292}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":475,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":2.6101808329421905,"rms_envelope_nrmse":0.1516692249852344,"envelope_correlation":0.9892619533336056,"rms_gain_error_db":-0.7251538358013108,"reference_mean_power":4.56248832445191e-05,"output_mean_power":3.860880702173354e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":475,"frequency_resolution_hz":0.42105263157894735,"power_floor":6.196742184587273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.073308488610773e-05,"output_power":2.7375766420838454e-05,"reference_active":true,"delta_db":-0.5023986883030217},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4683056260359484e-05,"output_power":1.4309696897690629e-05,"reference_active":true,"delta_db":-0.11186028055165273},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2512264429291995e-06,"output_power":2.2312276812082103e-06,"reference_active":true,"delta_db":-0.03875292083051962}],"active_band_count":3,"active_band_mae_db":0.2176706298950647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38080,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":2.301279891338821,"rms_envelope_nrmse":0.11424187815432192,"envelope_correlation":0.9958168241388867,"rms_gain_error_db":-0.7251470046406915,"reference_mean_power":4.620801886818527e-05,"output_mean_power":3.9102331066776376e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":5.636749189187334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.244224972502318e-05,"output_power":2.9962125194079386e-05,"reference_active":true,"delta_db":-0.3453834861923161},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.182187465953632e-06,"output_power":6.9239205335918225e-06,"reference_active":true,"delta_db":-0.15904661882407395},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2996975516005069e-07,"output_power":2.1150160132411082e-07,"reference_active":true,"delta_db":-0.36357062979748267}],"active_band_count":3,"active_band_mae_db":0.2893335782712909}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.4546120166778564,"predicted_mos":2.4546120166778564,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":57360,"input_sample_rate":24000,"input_seconds":2.39,"input_float32_sha256":"50b43c304592788dd40dcfccca8a304aa3620f119ed44d6e0c2a7d3b633d0c63","internal_samples":38240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.340289419000328},"started_at":"2026-10-13T05:07:07.803031+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0012834330000259797,"native_input_samples":57240,"native_input_sha256":"4f8c6dd430ce539306e95760f94d929084c20a22e6dc488bf29e8faf76f81111","output_sha256":"fa08a614a6f0a6d0ed4325dd705b24dfdf8aa1d3cd47f4f7a60de5038f1a3742","output_samples":57360,"length_delta_samples":120,"analysis_s":0.7264217190004274,"synthesis_s":4.660130665999532,"total_s":5.386552384999959,"rtf":2.258512530398306,"output_peak":0.38335055112838745,"output_rms":0.058551832386745044,"finished_at":"2026-10-13T05:07:13.195246+00:00","analysis_stages":{"external_f0_s":0.4485144769996623},"evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"fa08a614a6f0a6d0ed4325dd705b24dfdf8aa1d3cd47f4f7a60de5038f1a3742","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":38160,"output_samples":38240,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.64611702856078,"output_peak":0.3884936677550504},"common_prefix":{"samples":38160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.997907949790795,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.368096109618378,"spectral_convergence":0.7422352936030262,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.697886975701862,"spectral_convergence":0.7147751237096003,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.625677990645391,"spectral_convergence":0.6648367436701561,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.0012328689472823685,"delta_db":-4.657693000337125,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.0014526323610205886,"delta_db":-0.11097205262504482,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.00042644999858187016,"delta_db":-0.11952703274118429,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.00022288925342726606,"delta_db":-0.2861569318419998,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":8.08505693011716e-05,"delta_db":-0.013455343600081056,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.2243257634030636e-05,"delta_db":-0.8279207253525063,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0026208477496572,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-2.329206865819677,"pitch":null,"output_rms_dbfs":-24.63702416051629,"output_peak":0.3884936677550504,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.897220358655211},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38160,"output_samples":38160,"intelligibility":{"estoi":{"value":0.8089348772954988,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07666711752372073,"rms_dbfs":-22.307817294696616,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.052362335702647,"mean":0.0005484319083095846,"zero_energy":false},"output":{"rms":0.05863390134766146,"rms_dbfs":-24.63702416051629,"peak":0.3884936677550504,"peak_dbfs":-8.212321111596053,"crest_factor_db":16.42470304892024,"mean":0.0002928738067346252,"zero_energy":false},"rms_difference_db":-2.3292068658196747},"full_scale":{"reference":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":3.0155819437652115e-06,"output_rms_dbfs":-55.206288657532895,"output_zero_energy":false,"output_peak":0.011929433589825617,"output_energy_fraction":4.781107616726532e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":89,"both_estimated_voiced_frames":73,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":6.736589910457443,"absolute_difference_cents_p95":21.526230262408344,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.0004033177834583336,"log_rms_mae_db":5.874407209517071,"rms_envelope_nrmse":0.5428855443973417,"envelope_correlation":0.7271364394469544,"rms_gain_error_db":-2.3291669404386357,"reference_mean_power":0.005890133913079112,"output_mean_power":0.0034451526891998733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":4.7027834380481505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019529842144495992,"output_power":0.0011822292719268566,"reference_active":true,"delta_db":-2.1799702474168243},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0010897406087412391,"output_power":0.00021810407202132523,"reference_active":true,"delta_db":-6.986593610373303},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00042203241052064084,"output_power":6.75554908721701e-05,"reference_active":true,"delta_db":-7.956851504415004}],"active_band_count":3,"active_band_mae_db":5.707805120735043}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.0002641139400842597,"log_rms_mae_db":5.585786409013689,"rms_envelope_nrmse":0.4860174076884808,"envelope_correlation":0.7700459406884387,"rms_gain_error_db":-2.328949582966849,"reference_mean_power":0.0059649962331930016,"output_mean_power":0.003489114450332741},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.959398727002631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020291985866854774,"output_power":0.0009840288771279407,"reference_active":true,"delta_db":-3.143167078209042},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005646977153887027,"output_power":0.00012111605503388133,"reference_active":true,"delta_db":-6.68614314064846},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7144804121490974e-05,"output_power":4.4749033714621e-06,"reference_active":true,"delta_db":-9.191144106495262}],"active_band_count":3,"active_band_mae_db":6.340151441784255}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.00011878678262042868,"log_rms_mae_db":3.078793302807804,"rms_envelope_nrmse":0.22248774651299233,"envelope_correlation":0.9648021927412811,"rms_gain_error_db":-0.13377209724645117,"reference_mean_power":0.0009978964081743123,"output_mean_power":0.0009676276051177427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":1.0827926032603135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006103304193321489,"output_power":0.0005899571195889437,"reference_active":true,"delta_db":-0.14744569510343034},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011556410212235238,"output_power":9.524307597090412e-05,"reference_active":true,"delta_db":-0.839895364480546},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8931134571049757e-05,"output_power":3.2278064622310065e-05,"reference_active":true,"delta_db":2.317308440782674}],"active_band_count":3,"active_band_mae_db":1.10154983345555}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":2.330188864567221,"rms_envelope_nrmse":0.14226284765663724,"envelope_correlation":0.9845314653004172,"rms_gain_error_db":-0.1337666888139718,"reference_mean_power":0.0010106286473201321,"output_mean_power":0.0009799748626116956},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.735673127654544e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005396175923660179,"output_power":0.0005179834929161331,"reference_active":true,"delta_db":-0.17770180019779838},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.461345336739452e-05,"output_power":5.488463218058507e-05,"reference_active":true,"delta_db":-0.7087219516245068},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3971113807678808e-06,"output_power":3.3617570197470058e-06,"reference_active":true,"delta_db":1.4687810659511844}],"active_band_count":3,"active_band_mae_db":0.7850682725911632}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":2.280500457178971,"rms_envelope_nrmse":0.2444251020883617,"envelope_correlation":0.9604139661480726,"rms_gain_error_db":-0.2043729174164144,"reference_mean_power":0.0002825436588167163,"output_mean_power":0.00026955554683716636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":3.405184045067586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020464089970485737,"output_power":0.0001980860189502827,"reference_active":true,"delta_db":-0.1413861270483154},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.521356834551285e-05,"output_power":3.8061151767879934e-05,"reference_active":true,"delta_db":-0.7478685745067993},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018106896586624e-06,"output_power":8.283648280390778e-06,"reference_active":true,"delta_db":0.14149808760948565}],"active_band_count":3,"active_band_mae_db":0.34358426305486683}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":1.534488642048018,"rms_envelope_nrmse":0.1452486028902789,"envelope_correlation":0.9853133673507679,"rms_gain_error_db":-0.20437269859090768,"reference_mean_power":0.0002861487173387815,"output_mean_power":0.00027299489993332386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.083140770966501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019461067978585592,"output_power":0.0001862401338682032,"reference_active":true,"delta_db":-0.19093394529186622},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0239988981192974e-05,"output_power":1.7098822297002298e-05,"reference_active":true,"delta_db":-0.7324407289366669},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.640717194960497e-07,"output_power":6.454678584587881e-07,"reference_active":true,"delta_db":-0.7325950435460332}],"active_band_count":3,"active_band_mae_db":0.5519899059248554}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":2.0875333293711176,"rms_envelope_nrmse":0.26965322903686223,"envelope_correlation":0.9525373350064523,"rms_gain_error_db":-0.4332809626988029,"reference_mean_power":4.55291382465523e-05,"output_mean_power":4.1206083096585336e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":6.179039674292896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.30631712224655e-05,"output_power":2.9012298966213445e-05,"reference_active":true,"delta_db":-0.5676236203866744},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2207009907392539e-05,"output_power":1.1047771462756663e-05,"reference_active":true,"delta_db":-0.43334615324787096},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.251387780144131e-06,"output_power":2.1613697400687614e-06,"reference_active":true,"delta_db":-0.17721237639591075}],"active_band_count":3,"active_band_mae_db":0.3927273833434854}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":1.2463084870559387,"rms_envelope_nrmse":0.1623726784777164,"envelope_correlation":0.9826465179842404,"rms_gain_error_db":-0.4332964123930266,"reference_mean_power":4.61098055796954e-05,"output_mean_power":4.1731466800154484e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":5.621286439624597e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.22612462378357e-05,"output_power":2.8318641262262304e-05,"reference_active":true,"delta_db":-0.5660872805763779},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2476871573483135e-06,"output_power":6.606387408200932e-06,"reference_active":true,"delta_db":-0.4023540121307798},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.30977962238882e-07,"output_power":2.0570755250796017e-07,"reference_active":true,"delta_db":-0.5032030852846707}],"active_band_count":3,"active_band_mae_db":0.49054812599727615}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.0699753761291504,"predicted_mos":2.0699753761291504,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":57480,"input_sample_rate":24000,"input_seconds":2.395,"input_float32_sha256":"9fa1d722338f82c245a92f182b76e9ee9575de7b9ec164c096b6cde75a983813","internal_samples":38320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2928441280000698},"started_at":"2026-10-13T05:10:17.100519+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001903712000057567,"native_input_samples":57240,"native_input_sha256":"4f8c6dd430ce539306e95760f94d929084c20a22e6dc488bf29e8faf76f81111","output_sha256":"1c1954331ceb97fe60e021ceb6f2f89fd7be02b43e551a57742ec785c51fbdbd","output_samples":57480,"length_delta_samples":240,"analysis_s":0.1725525390011171,"synthesis_s":0.05040784899938444,"total_s":0.22296038800050155,"rtf":0.0934844394132082,"output_peak":0.5398839116096497,"output_rms":0.07300938231183561,"finished_at":"2026-10-13T05:10:17.330085+00:00","analysis_stages":{"external_f0_s":0.028748631999405916},"evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"1c1954331ceb97fe60e021ceb6f2f89fd7be02b43e551a57742ec785c51fbdbd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":38160,"output_samples":38320,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.730471479792346,"output_peak":0.539840515030139},"common_prefix":{"samples":38160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9958246346555324,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.257945750726827,"spectral_convergence":0.4828126998905431,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.835020447666829,"spectral_convergence":0.3898715948201716,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.325931547493629,"spectral_convergence":0.336481019335016,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.003136112877819463,"delta_db":-0.602945282056784,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.0014803650874697875,"delta_db":-0.028840859317793986,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.0004139749258519071,"delta_db":-0.2484678354675678,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.00022753097314309094,"delta_db":-0.19664298436540473,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":7.3970518573065e-05,"delta_db":-0.3996995330405462,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.3172838822839084e-05,"delta_db":-0.6501122372144963,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3544514552437655,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4044847771216382,"pitch":null,"output_rms_dbfs":-22.712302071818257,"output_peak":0.539840515030139,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.139632581962428},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38160,"output_samples":38160,"intelligibility":{"estoi":{"value":0.8908374758396406,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07666711752372073,"rms_dbfs":-22.307817294696616,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.052362335702647,"mean":0.0005484319083095846,"zero_energy":false},"output":{"rms":0.07317873476413864,"rms_dbfs":-22.712302071818257,"peak":0.539840515030139,"peak_dbfs":-5.354690495118106,"crest_factor_db":17.357611576700148,"mean":6.091002030484497e-05,"zero_energy":false},"rms_difference_db":-0.40448477712164177},"full_scale":{"reference":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":1.2776515932495607e-05,"output_rms_dbfs":-48.93587559133837,"output_zero_energy":false,"output_peak":0.010305420901860091,"output_energy_fraction":0.00013004618518156925,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":69,"both_estimated_voiced_frames":67,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9178082191780822,"absolute_difference_cents_median":1.3913746278602808,"absolute_difference_cents_p95":6.999753720318998,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.0004033177834583336,"log_rms_mae_db":1.656170619481411,"rms_envelope_nrmse":0.22614732180726915,"envelope_correlation":0.9563769025576274,"rms_gain_error_db":-0.40450843057130736,"reference_mean_power":0.005890133913079112,"output_mean_power":0.005366292309254645},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":4.7027834380481505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019529842144495992,"output_power":0.0017641591779511121,"reference_active":true,"delta_db":-0.4416096456323706},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0010897406087412391,"output_power":0.00096315595691672,"reference_active":true,"delta_db":-0.5362652000576529},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00042203241052064084,"output_power":0.0003273550878670693,"reference_active":true,"delta_db":-1.103267091776134}],"active_band_count":3,"active_band_mae_db":0.6937139791553859}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.0002641139400842597,"log_rms_mae_db":0.968907615527013,"rms_envelope_nrmse":0.10273784243376605,"envelope_correlation":0.991549741965675,"rms_gain_error_db":-0.4045284043194522,"reference_mean_power":0.0059649962331930016,"output_mean_power":0.0054344717227173475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.959398727002631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020291985866854774,"output_power":0.0018144335439216916,"reference_active":true,"delta_db":-0.48583484958031115},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005646977153887027,"output_power":0.0004585772307183198,"reference_active":true,"delta_db":-0.9040354336980299},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7144804121490974e-05,"output_power":2.126832125772952e-05,"reference_active":true,"delta_db":-2.4216486078830393}],"active_band_count":3,"active_band_mae_db":1.2705062970537935}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.00011878678262042868,"log_rms_mae_db":1.3329143362903495,"rms_envelope_nrmse":0.09239262731766434,"envelope_correlation":0.9943003763429843,"rms_gain_error_db":-0.21348175054368204,"reference_mean_power":0.0009978964081743123,"output_mean_power":0.0009500299228701342},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":1.0827926032603135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006103304193321489,"output_power":0.0005879920299400058,"reference_active":true,"delta_db":-0.1619357668185799},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011556410212235238,"output_power":8.932015807901158e-05,"reference_active":true,"delta_db":-1.1187346636895692},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8931134571049757e-05,"output_power":1.1563907611808125e-05,"reference_active":true,"delta_db":-2.1407202937935965}],"active_band_count":3,"active_band_mae_db":1.1404635747672485}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":0.7045368114368371,"rms_envelope_nrmse":0.06826300578547313,"envelope_correlation":0.9968298847176664,"rms_gain_error_db":-0.21348925494341195,"reference_mean_power":0.0010106286473201321,"output_mean_power":0.0009621497671966319},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.735673127654544e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005396175923660179,"output_power":0.0005201044291822292,"reference_active":true,"delta_db":-0.15995547681028582},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.461345336739452e-05,"output_power":5.0962118511823365e-05,"reference_active":true,"delta_db":-1.0307547974521518},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3971113807678808e-06,"output_power":1.4072169088388106e-06,"reference_active":true,"delta_db":-2.313271689351843}],"active_band_count":3,"active_band_mae_db":1.1679939878714267}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":1.3060166444210688,"rms_envelope_nrmse":0.0902375112939181,"envelope_correlation":0.9951174590146336,"rms_gain_error_db":-0.25785345908959667,"reference_mean_power":0.0002825436588167163,"output_mean_power":0.0002662565000401606},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":3.405184045067586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020464089970485737,"output_power":0.00019580400970814526,"reference_active":true,"delta_db":-0.1917085540170492},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.521356834551285e-05,"output_power":3.972347467273495e-05,"reference_active":true,"delta_db":-0.5622155389498709},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018106896586624e-06,"output_power":6.45287136704998e-06,"reference_active":true,"delta_db":-0.9431883405084752}],"active_band_count":3,"active_band_mae_db":0.5657041444917984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":0.6382648974823685,"rms_envelope_nrmse":0.05226335735457685,"envelope_correlation":0.9986546019283012,"rms_gain_error_db":-0.2578563866630382,"reference_mean_power":0.0002861487173387815,"output_mean_power":0.00026965356408925704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.083140770966501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019461067978585592,"output_power":0.00018605099797899626,"reference_active":true,"delta_db":-0.195346657443866},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0239988981192974e-05,"output_power":1.7872108783024537e-05,"reference_active":true,"delta_db":-0.5403447251384045},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.640717194960497e-07,"output_power":5.236388794919916e-07,"reference_active":true,"delta_db":-1.641022406211429}],"active_band_count":3,"active_band_mae_db":0.7922379295978997}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":1.4573744671256057,"rms_envelope_nrmse":0.11585752889186804,"envelope_correlation":0.9921577945586031,"rms_gain_error_db":-0.35040548412435335,"reference_mean_power":4.55291382465523e-05,"output_mean_power":4.199996049072873e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":6.179039674292896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.30631712224655e-05,"output_power":3.0200942559259486e-05,"reference_active":true,"delta_db":-0.39324008848715364},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2207009907392539e-05,"output_power":1.0171287842485568e-05,"reference_active":true,"delta_db":-0.7923335224478745},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.251387780144131e-06,"output_power":1.726375488724525e-06,"reference_active":true,"delta_db":-1.1531504331938285}],"active_band_count":3,"active_band_mae_db":0.7795746813762855}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":0.8867374348613809,"rms_envelope_nrmse":0.06485798878319418,"envelope_correlation":0.9981208516447246,"rms_gain_error_db":-0.35041025358481365,"reference_mean_power":4.61098055796954e-05,"output_mean_power":4.253557085341807e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":5.621286439624597e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.22612462378357e-05,"output_power":2.9190287051693745e-05,"reference_active":true,"delta_db":-0.4344277420513172},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2476871573483135e-06,"output_power":6.05793384282194e-06,"reference_active":true,"delta_db":-0.7787491277610679},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.30977962238882e-07,"output_power":1.582940714515825e-07,"reference_active":true,"delta_db":-1.6410589589924318}],"active_band_count":3,"active_band_mae_db":0.9514119429349389}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2094-142345-0040","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1901767253875732,"predicted_mos":3.1901767253875732,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":114856,"input_sample_rate":48000,"input_seconds":2.3928333333333334,"input_float32_sha256":"683ddd956280da019fb07249cf232e3b19be29f9c4a9e0a538cafb36712d13ea","internal_samples":38286,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4744806320013595},"started_at":"2026-10-13T05:13:33.430017+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.004513089999818476,"native_input_samples":114480,"native_input_sha256":"29ea2baa3ea58d05177ba55070e3abc20a63cefd97a69b2904b730b893ea8943","output_sha256":"2cf3370b4430ea94718c2094450053e276b37fe93adb156379cc6f0829bda05d","output_samples":114856,"length_delta_samples":376,"analysis_s":1.4417695390002336,"synthesis_s":0.6069945210001606,"total_s":2.0487640600003942,"rtf":0.8590205702307734,"output_peak":0.8049853499245729,"output_rms":0.07193522958470683,"finished_at":"2026-10-13T05:13:35.491127+00:00","evaluation_input":{"reference":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"output_sha256":"2cf3370b4430ea94718c2094450053e276b37fe93adb156379cc6f0829bda05d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":38160,"output_samples":38286,"length_delta_samples":126,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.859143286182046,"output_peak":0.804945433504017},"common_prefix":{"samples":38160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.996708979783733,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":38160,"output_samples":38160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.700733142092804,"spectral_convergence":0.6019498240330217,"spectral_convergence_status":"ok","amplitude_floor":5.674431450769592e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.7000992547176805,"spectral_convergence":0.5853614844963339,"spectral_convergence_status":"ok","amplitude_floor":3.1139992269810527e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.658955157685955,"spectral_convergence":0.5945196181437017,"spectral_convergence_status":"ok","amplitude_floor":1.28722240515293e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.003603182122246331,"output_power":0.0030522125682321917,"delta_db":-0.7207143768784792,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001490228689925631,"output_power":0.0013700506632746395,"delta_db":-0.3651629290866046,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00043834980790852295,"output_power":0.000426193151814049,"delta_db":-0.12214353502075204,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023807008265260595,"output_power":0.00023507167986694207,"delta_db":-0.05504511839964622,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.11014495688624e-05,"output_power":8.524527651654579e-05,"delta_db":0.21641707202314886,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.691475709405268e-05,"output_power":2.538494447887951e-05,"delta_db":-0.25414246790452516,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2889375832188594,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5370309056244092,"pitch":null,"output_rms_dbfs":-22.844848200321024,"output_peak":0.804945433504017,"reference_rms_dbfs":-22.307817294696616,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.019929184832146},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":38160,"output_samples":38160,"intelligibility":{"estoi":{"value":0.81509097036686,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07666711752372073,"rms_dbfs":-22.307817294696616,"peak":0.6126708984375,"peak_dbfs":-4.255454958993972,"crest_factor_db":18.052362335702647,"mean":0.0005484319083095846,"zero_energy":false},"output":{"rms":0.07207050911561017,"rms_dbfs":-22.844848200321024,"peak":0.804945433504017,"peak_dbfs":-1.8846711809982621,"crest_factor_db":20.960177019322764,"mean":0.004011114273999206,"zero_energy":false},"rms_difference_db":-0.5370309056244089},"full_scale":{"reference":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":38160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":238,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6266523445374332e-05,"threshold_dbfs":-47.88705256517615,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":1.066119965309134e-05,"output_mean_power":1.3349012264764276e-05,"output_rms_dbfs":-48.74550867919811,"output_zero_energy":false,"output_peak":0.015814006944935247,"output_energy_fraction":0.00014008412477005918,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":231,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":71,"both_estimated_voiced_frames":66,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9041095890410958,"absolute_difference_cents_median":3.3894304359430585,"absolute_difference_cents_p95":13.172049996736995,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":38160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.0004033177834583336,"log_rms_mae_db":2.6546726057580567,"rms_envelope_nrmse":0.33691441308016323,"envelope_correlation":0.903414884275299,"rms_gain_error_db":-0.5369875699498399,"reference_mean_power":0.005890133913079112,"output_mean_power":0.005205068048721418},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":4.7027834380481505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019529842144495992,"output_power":0.0023530375575904878,"reference_active":true,"delta_db":0.8093012615211228},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0010897406087412391,"output_power":0.0013113909593870994,"reference_active":true,"delta_db":0.8040905034705721},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00042203241052064084,"output_power":0.0003422004297661441,"reference_active":true,"delta_db":-0.9106525381297326}],"active_band_count":3,"active_band_mae_db":0.8413481010404759}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.0002641139400842597,"log_rms_mae_db":1.8153586471057614,"rms_envelope_nrmse":0.28128956525122634,"envelope_correlation":0.9260631280299131,"rms_gain_error_db":-0.5368326548177857,"reference_mean_power":0.0059649962331930016,"output_mean_power":0.005271411361465871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.959398727002631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020291985866854774,"output_power":0.0025233305848366178,"reference_active":true,"delta_db":0.9464960056834494},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005646977153887027,"output_power":0.0006121633379367123,"reference_active":true,"delta_db":0.3505128572202244},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7144804121490974e-05,"output_power":2.424376121075502e-05,"reference_active":true,"delta_db":-1.8529807472294133}],"active_band_count":3,"active_band_mae_db":1.0499965367110289}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":0.00011878678262042868,"log_rms_mae_db":2.175867822932963,"rms_envelope_nrmse":0.14048781658214157,"envelope_correlation":0.9862851042858518,"rms_gain_error_db":-0.22945247280237407,"reference_mean_power":0.0009978964081743123,"output_mean_power":0.0009465427037376441},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":1.0827926032603135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006103304193321489,"output_power":0.0006002261833755558,"reference_active":true,"delta_db":-0.07250079543693475},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011556410212235238,"output_power":0.00010140897474556243,"reference_active":true,"delta_db":-0.5674655740484011},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8931134571049757e-05,"output_power":1.9619418063043456e-05,"reference_active":true,"delta_db":0.15509478848897246}],"active_band_count":3,"active_band_mae_db":0.2650203859914361}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":9.505044043492436e-05,"log_rms_mae_db":1.338955107131857,"rms_envelope_nrmse":0.09464342438011236,"envelope_correlation":0.9935403806255771,"rms_gain_error_db":-0.2294530731076893,"reference_mean_power":0.0010106286473201321,"output_mean_power":0.0009586195844039585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.735673127654544e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005396175923660179,"output_power":0.0005327131216705326,"reference_active":true,"delta_db":-0.05592705539617979},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.461345336739452e-05,"output_power":5.766929558393907e-05,"reference_active":true,"delta_db":-0.4937830656595205},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3971113807678808e-06,"output_power":2.291558248664963e-06,"reference_active":true,"delta_db":-0.19557312845682603}],"active_band_count":3,"active_band_mae_db":0.2484277498375088}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":9.972124238081891e-05,"log_rms_mae_db":1.6610995744390986,"rms_envelope_nrmse":0.16405995721475872,"envelope_correlation":0.9823935994629718,"rms_gain_error_db":-0.01973501014017545,"reference_mean_power":0.0002825436588167163,"output_mean_power":0.0002812626496706008},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":3.405184045067586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020464089970485737,"output_power":0.00020972408556064331,"reference_active":true,"delta_db":0.10655872944213862},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.521356834551285e-05,"output_power":4.007217060948659e-05,"reference_active":true,"delta_db":-0.5242591606040599},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018106896586624e-06,"output_power":8.598254024380366e-06,"reference_active":true,"delta_db":0.3033842972061616}],"active_band_count":3,"active_band_mae_db":0.31140072908412003}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":8.262892288945791e-05,"log_rms_mae_db":0.9752097478380988,"rms_envelope_nrmse":0.10330947612544804,"envelope_correlation":0.9925389144427748,"rms_gain_error_db":-0.01973404626451719,"reference_mean_power":0.0002861487173387815,"output_mean_power":0.0002848514266363035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":3.083140770966501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019461067978585592,"output_power":0.00019767731292445807,"reference_active":true,"delta_db":0.0679015929702079},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0239988981192974e-05,"output_power":1.7554451087029768e-05,"reference_active":true,"delta_db":-0.6182301774525285},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.640717194960497e-07,"output_power":6.941382604473936e-07,"reference_active":true,"delta_db":-0.4168814239584501}],"active_band_count":3,"active_band_mae_db":0.36767106479372885}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":476,"rms_floor":4.3426949519701344e-05,"log_rms_mae_db":1.611315978997912,"rms_envelope_nrmse":0.18447234354918893,"envelope_correlation":0.9782286007373658,"rms_gain_error_db":-0.16502708863388307,"reference_mean_power":4.55291382465523e-05,"output_mean_power":4.3831539168195816e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":476,"frequency_resolution_hz":0.42016806722689076,"power_floor":6.179039674292896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.30631712224655e-05,"output_power":3.120871859294715e-05,"reference_active":true,"delta_db":-0.25068568954227954},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2207009907392539e-05,"output_power":9.866435345285854e-06,"reference_active":true,"delta_db":-0.9244902275193662},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.251387780144131e-06,"output_power":2.223991297924749e-06,"reference_active":true,"delta_db":-0.05317220941884816}],"active_band_count":3,"active_band_mae_db":0.40944937549349797}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":38160,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":3.683581645514391e-05,"log_rms_mae_db":1.0695011747560026,"rms_envelope_nrmse":0.1178978616390985,"envelope_correlation":0.9909464063616735,"rms_gain_error_db":-0.1650336055248169,"reference_mean_power":4.61098055796954e-05,"output_mean_power":4.439048913280999e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":5.621286439624597e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.22612462378357e-05,"output_power":2.9563966671117947e-05,"reference_active":true,"delta_db":-0.3791843597721479},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2476871573483135e-06,"output_power":5.975923381673506e-06,"reference_active":true,"delta_db":-0.8379441829771045},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.30977962238882e-07,"output_power":2.0939494115722954e-07,"reference_active":true,"delta_db":-0.42604360339973607}],"active_band_count":3,"active_band_mae_db":0.5477240487163295}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},{"case":"librispeech_2300-131720-0022","method":"hnm","status":"measured","mos":{"status":"measured","value":2.764047861099243,"predicted_mos":2.764047861099243,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"945bfccdf42a6b444111727c5dbec3bb0ec76b2958e5cfd92a8f996fecf5c10e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.336347038000895},"started_at":"2026-10-13T04:38:33.845312+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5263999557646457e-05,"native_input_samples":96000,"native_input_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","output_sha256":"ee2e1be0dcc5ce806f1cac0e1c88ebf56e567615ac7f13e7013b5a9759dc4c03","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2240268359992115,"synthesis_s":0.5009544820004521,"total_s":1.7249813179996636,"rtf":0.28749688633327725,"output_peak":0.6668041400988419,"output_rms":0.0768517235694638,"finished_at":"2026-10-13T04:38:35.574990+00:00","evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"ee2e1be0dcc5ce806f1cac0e1c88ebf56e567615ac7f13e7013b5a9759dc4c03","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.9590975312507,"spectral_convergence":0.2791917713506903,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.67690674700723,"spectral_convergence":0.2796646366293311,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.176772468856673,"spectral_convergence":0.2938517014138143,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.002458297902376768,"delta_db":-0.12620665664857675,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.0026333928851813484,"delta_db":-0.06174251443388173,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.0004914025870643458,"delta_db":0.23679521806326667,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00011096941281725473,"delta_db":-0.4976805086101255,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.0001660689840100565,"delta_db":0.09246422042874605,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":4.605564414750491e-05,"delta_db":1.0823257905830765,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.34953581812794554,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.061282966418962394,"pitch":null,"output_rms_dbfs":-22.28692776086092,"output_peak":0.6668041400988419,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.937592249038201},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.9590975312507,"spectral_convergence":0.2791917713506903,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.67690674700723,"spectral_convergence":0.2796646366293311,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.176772468856673,"spectral_convergence":0.2938517014138143,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.002458297902376768,"delta_db":-0.12620665664857675,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.0026333928851813484,"delta_db":-0.06174251443388173,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.0004914025870643458,"delta_db":0.23679521806326667,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00011096941281725473,"delta_db":-0.4976805086101255,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.0001660689840100565,"delta_db":0.09246422042874605,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":4.605564414750491e-05,"delta_db":1.0823257905830765,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.34953581812794554,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.061282966418962394,"pitch":null,"output_rms_dbfs":-22.28692776086092,"output_peak":0.6668041400988419,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.937592249038201},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8911755319375176,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0773958653322626,"rms_dbfs":-22.225644794441955,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.47852725154813,"mean":1.5013694763183594e-05,"zero_energy":false},"output":{"rms":0.0768517235694638,"rms_dbfs":-22.28692776086092,"peak":0.6668041400988419,"peak_dbfs":-3.5200342471715955,"crest_factor_db":18.766893513689325,"mean":-3.2747222868415546e-05,"zero_energy":false},"rms_difference_db":-0.06128296641896469},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":3.574484684524214e-06,"output_rms_dbfs":-54.467865594010156,"output_zero_energy":false,"output_peak":0.01654890495258599,"output_energy_fraction":0.00011297256489848631,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":173,"both_estimated_voiced_frames":152,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.8735632183908046,"absolute_difference_cents_median":1.9313119055572887,"absolute_difference_cents_p95":11.417335521616543,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031508938028130336,"log_rms_mae_db":1.0214411435965658,"rms_envelope_nrmse":0.0841293005532006,"envelope_correlation":0.991079299329291,"rms_gain_error_db":-0.061259266744270194,"reference_mean_power":0.0059947180836205155,"output_mean_power":0.0059107533558987605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.905621652404067e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011843610245185533,"output_power":0.0012109996544612666,"reference_active":true,"delta_db":0.09659912228634704},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002763334576078544,"output_power":0.000256734164749351,"reference_active":true,"delta_db":-0.3194980514218199},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0009701982183084e-05,"output_power":5.238281972918833e-05,"reference_active":true,"delta_db":0.2013460596767511}],"active_band_count":3,"active_band_mae_db":0.205814411128306}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002456604014165908,"log_rms_mae_db":0.6009881177606597,"rms_envelope_nrmse":0.041542477207777734,"envelope_correlation":0.9976576045656775,"rms_gain_error_db":-0.06114823509772667,"reference_mean_power":0.006022035579915752,"output_mean_power":0.0059378400357762255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5974133279358877e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010595070071700143,"output_power":0.0010918895257039916,"reference_active":true,"delta_db":0.13074866591081968},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014348331799351297,"output_power":0.00013269045187404584,"reference_active":true,"delta_db":-0.33961737932855873},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4072600706208143e-06,"output_power":3.123794644027489e-06,"reference_active":true,"delta_db":-0.3772280788223861}],"active_band_count":3,"active_band_mae_db":0.28253137468725487}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026889287764829997,"log_rms_mae_db":1.513786862292706,"rms_envelope_nrmse":0.12315968185152523,"envelope_correlation":0.9870570070302478,"rms_gain_error_db":-0.03665670404042633,"reference_mean_power":0.0016839373335741825,"output_mean_power":0.0016697838455643878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1138357532326787e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006324020549599114,"output_power":0.0006344723216753275,"reference_active":true,"delta_db":0.014194086288224484},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002190350723697022,"output_power":0.00019330388279652938,"reference_active":true,"delta_db":-0.5427308301229596},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1194141233781235e-05,"output_power":2.645383777500502e-05,"reference_active":true,"delta_db":-0.7158434816287225}],"active_band_count":3,"active_band_mae_db":0.4242561326799688}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001970017370857221,"log_rms_mae_db":1.0628006620488657,"rms_envelope_nrmse":0.07112198401135071,"envelope_correlation":0.9952325532222253,"rms_gain_error_db":-0.03667474865681308,"reference_mean_power":0.001692399182790771,"output_mean_power":0.0016781676002873684},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.64481232375879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970882254118313,"output_power":0.0006025404380901353,"reference_active":true,"delta_db":0.039476919089267495},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011301601543605377,"output_power":0.00010144622593447116,"reference_active":true,"delta_db":-0.4690409669478651},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.966963236644965e-06,"output_power":1.947675447606734e-06,"reference_active":true,"delta_db":-1.8279557547198468}],"active_band_count":3,"active_band_mae_db":0.7788245469189933}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":2.0520904918363763,"rms_envelope_nrmse":0.21275400900763428,"envelope_correlation":0.9645559751049341,"rms_gain_error_db":-0.34130684054766336,"reference_mean_power":0.0002025729815791723,"output_mean_power":0.0001872625082986312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0818748871802835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.592086384233799e-05,"output_power":5.630890488737032e-05,"reference_active":true,"delta_db":-0.6844580868971444},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.925318155530913e-05,"output_power":1.469312924838389e-05,"reference_active":true,"delta_db":-1.1738820730192976},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.819375654274542e-06,"output_power":3.688134625786643e-06,"reference_active":true,"delta_db":-0.15185609995981345}],"active_band_count":3,"active_band_mae_db":0.6700654199587518}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":1.4525493483197707,"rms_envelope_nrmse":0.12511966762150967,"envelope_correlation":0.9876795747546868,"rms_gain_error_db":-0.3414012049402144,"reference_mean_power":0.0002035903915876016,"output_mean_power":0.00018819893315248905},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.438814668171718e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.334752463763687e-05,"output_power":5.436951583383884e-05,"reference_active":true,"delta_db":-0.6637418351596069},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.661338887813939e-06,"output_power":6.811986211966479e-06,"reference_active":true,"delta_db":-1.5176355574787532},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.128964187832131e-07,"output_power":2.6107655877200437e-07,"reference_active":true,"delta_db":-0.7863271284861474}],"active_band_count":3,"active_band_mae_db":0.989234840374836}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":2.168260841234534,"rms_envelope_nrmse":0.3348101854205156,"envelope_correlation":0.9413652818469501,"rms_gain_error_db":0.651865640013715,"reference_mean_power":8.701244577979692e-05,"output_mean_power":0.00010110390737516547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.568258861947873e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4501869786005614e-05,"output_power":2.4613066644005428e-05,"reference_active":true,"delta_db":0.01966500126277046},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.528335672589259e-06,"output_power":7.269964064708496e-06,"reference_active":true,"delta_db":2.0559365177219155},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7809461115210602e-06,"output_power":1.7713384875834806e-06,"reference_active":true,"delta_db":-0.023492195616747372}],"active_band_count":3,"active_band_mae_db":0.6996979048671444}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":1.7100237137952399,"rms_envelope_nrmse":0.266057198634472,"envelope_correlation":0.9628429467924554,"rms_gain_error_db":0.6517916138991227,"reference_mean_power":8.744947435640977e-05,"output_mean_power":0.00010160997973378257},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.34827454492152e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1350322554557276e-05,"output_power":2.2089405564906208e-05,"reference_active":true,"delta_db":0.14779588376259634},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413574366249565e-06,"output_power":3.907179360264497e-06,"reference_active":true,"delta_db":2.0920266340571265},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6206121669861414e-07,"output_power":1.8019268011697296e-07,"reference_active":true,"delta_db":0.4605804974605587}],"active_band_count":3,"active_band_mae_db":0.9001343384267605}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0022","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.6796395778656006,"predicted_mos":3.6796395778656006,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"573a2db65a080560b5ebfe33d12164b3669075c65a8b4f6b083f42ee36f6b8f0","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.338607128000149},"started_at":"2026-10-13T04:41:31.026388+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.374500036239624e-05,"native_input_samples":96000,"native_input_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","output_sha256":"067cd1a6712611bb9c9ffcc54f71efb79f7a0fc0bc53d2d8bda219f5b5c2b557","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6143238940003357,"synthesis_s":0.08593108100012614,"total_s":1.7002549750004619,"rtf":0.28337582916674364,"output_peak":0.8550572803147956,"output_rms":0.08505741352098616,"finished_at":"2026-10-13T04:41:32.732298+00:00","evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"067cd1a6712611bb9c9ffcc54f71efb79f7a0fc0bc53d2d8bda219f5b5c2b557","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.782508360769657,"spectral_convergence":0.6464434184127339,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.144500840305401,"spectral_convergence":0.2692825252967261,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.616997470943169,"spectral_convergence":0.27430876994126824,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.002736209644497654,"delta_db":0.3389419310851333,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.003522047171456071,"delta_db":1.2010525894952309,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.0005967569725439033,"delta_db":1.080395852772377,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00014960835762961415,"delta_db":0.7998451648745407,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.00018963177387464595,"delta_db":0.6686900092287342,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":4.050967487814724e-05,"delta_db":0.5250847585253915,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.769001717663568,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8198882400547143,"pitch":null,"output_rms_dbfs":-21.405756554387242,"output_peak":0.8550572803147956,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.848002224006076},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.782508360769657,"spectral_convergence":0.6464434184127339,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.144500840305401,"spectral_convergence":0.2692825252967261,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.616997470943169,"spectral_convergence":0.27430876994126824,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.002736209644497654,"delta_db":0.3389419310851333,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.003522047171456071,"delta_db":1.2010525894952309,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.0005967569725439033,"delta_db":1.080395852772377,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00014960835762961415,"delta_db":0.7998451648745407,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.00018963177387464595,"delta_db":0.6686900092287342,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":4.050967487814724e-05,"delta_db":0.5250847585253915,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.769001717663568,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8198882400547143,"pitch":null,"output_rms_dbfs":-21.405756554387242,"output_peak":0.8550572803147956,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.848002224006076},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.957404892697998,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0773958653322626,"rms_dbfs":-22.225644794441955,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.47852725154813,"mean":1.5013694763183594e-05,"zero_energy":false},"output":{"rms":0.08505741352098614,"rms_dbfs":-21.405756554387242,"peak":0.8550572803147956,"peak_dbfs":-1.3600958179189053,"crest_factor_db":20.045660736468342,"mean":6.8438389451329305e-06,"zero_energy":false},"rms_difference_db":0.8198882400547127},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":4.163979618741105e-06,"output_rms_dbfs":-53.80491404879545,"output_zero_energy":false,"output_peak":0.020196525858501603,"output_energy_fraction":0.00010743629495349504,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":189,"both_estimated_voiced_frames":162,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9310344827586207,"absolute_difference_cents_median":5.942746550479201,"absolute_difference_cents_p95":24.86370736616085,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031508938028130336,"log_rms_mae_db":2.225233806667508,"rms_envelope_nrmse":0.1766763413514735,"envelope_correlation":0.97711956324197,"rms_gain_error_db":0.8201605010165346,"reference_mean_power":0.0059947180836205155,"output_mean_power":0.007240771029199329},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.905621652404067e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011843610245185533,"output_power":0.0015757181085833893,"reference_active":true,"delta_db":1.2399441907871769},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002763334576078544,"output_power":0.0003597621139619385,"reference_active":true,"delta_db":1.1458195533700626},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0009701982183084e-05,"output_power":5.259967976943434e-05,"reference_active":true,"delta_db":0.21928833638611528}],"active_band_count":3,"active_band_mae_db":0.8683506935144516}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002456604014165908,"log_rms_mae_db":1.8410991846562967,"rms_envelope_nrmse":0.1405650201872214,"envelope_correlation":0.9889530936220684,"rms_gain_error_db":0.8199494293531083,"reference_mean_power":0.006022035579915752,"output_mean_power":0.007273413194470838},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5974133279358877e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010595070071700143,"output_power":0.0014507971151370759,"reference_active":true,"delta_db":1.3650285001656026},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014348331799351297,"output_power":0.00019233405053663022,"reference_active":true,"delta_db":1.2725476689604187},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4072600706208143e-06,"output_power":2.7830599485986122e-06,"reference_active":true,"delta_db":-0.8788272245463024}],"active_band_count":3,"active_band_mae_db":1.1721344645574412}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026889287764829997,"log_rms_mae_db":1.9058233810943557,"rms_envelope_nrmse":0.27009541443956264,"envelope_correlation":0.9757530559227082,"rms_gain_error_db":1.530666539966258,"reference_mean_power":0.0016839373335741825,"output_mean_power":0.0023954801669515283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1138357532326787e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006324020549599114,"output_power":0.0009214038350968286,"reference_active":true,"delta_db":1.6345674307356246},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002190350723697022,"output_power":0.00031462624510490537,"reference_active":true,"delta_db":1.5728128666067176},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1194141233781235e-05,"output_power":3.732285649944822e-05,"reference_active":true,"delta_db":0.7790184089006849}],"active_band_count":3,"active_band_mae_db":1.3287995687476757}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001970017370857221,"log_rms_mae_db":1.661856357217609,"rms_envelope_nrmse":0.24016503689695615,"envelope_correlation":0.9852864392899908,"rms_gain_error_db":1.5306653094677074,"reference_mean_power":0.001692399182790771,"output_mean_power":0.0024075168638691058},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.64481232375879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970882254118313,"output_power":0.0008862836012087876,"reference_active":true,"delta_db":1.7153420671089805},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011301601543605377,"output_power":0.00016475743985383965,"reference_active":true,"delta_db":1.637050434152517},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.966963236644965e-06,"output_power":2.859753255297952e-06,"reference_active":true,"delta_db":-0.15983602019922497}],"active_band_count":3,"active_band_mae_db":1.1707428404869074}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":1.6611367650150428,"rms_envelope_nrmse":0.19643914857438946,"envelope_correlation":0.9789890137554759,"rms_gain_error_db":0.7855891659507841,"reference_mean_power":0.0002025729815791723,"output_mean_power":0.00024273949099236881},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0818748871802835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.592086384233799e-05,"output_power":8.318227461044952e-05,"reference_active":true,"delta_db":1.010079021054024},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.925318155530913e-05,"output_power":2.155959831002194e-05,"reference_active":true,"delta_db":0.491381587908103},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.819375654274542e-06,"output_power":3.1198445800720653e-06,"reference_active":true,"delta_db":-0.8785941596056133}],"active_band_count":3,"active_band_mae_db":0.7933515895225801}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":1.068569988598886,"rms_envelope_nrmse":0.13422290912149984,"envelope_correlation":0.9929897903402706,"rms_gain_error_db":0.7856009687365074,"reference_mean_power":0.0002035903915876016,"output_mean_power":0.000243959297763289},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.438814668171718e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.334752463763687e-05,"output_power":7.941402912929354e-05,"reference_active":true,"delta_db":0.9816758173297528},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.661338887813939e-06,"output_power":1.0786441548445901e-05,"reference_active":true,"delta_db":0.47840878389789726},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.128964187832131e-07,"output_power":2.1188401280349393e-07,"reference_active":true,"delta_db":-1.6930240320589907}],"active_band_count":3,"active_band_mae_db":1.051036211095547}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":1.8211178208132077,"rms_envelope_nrmse":0.2567034283501001,"envelope_correlation":0.9662873780042467,"rms_gain_error_db":0.6034969340339464,"reference_mean_power":8.701244577979692e-05,"output_mean_power":9.998412943677937e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.568258861947873e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4501869786005614e-05,"output_power":2.7896622131058755e-05,"reference_active":true,"delta_db":0.563523923117301},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.528335672589259e-06,"output_power":7.102410376667691e-06,"reference_active":true,"delta_db":1.9546714980587077},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7809461115210602e-06,"output_power":2.6365258673196897e-06,"reference_active":true,"delta_db":1.703812579810713}],"active_band_count":3,"active_band_mae_db":1.4073360003289073}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":1.0523691136867712,"rms_envelope_nrmse":0.18267286002740396,"envelope_correlation":0.9838164370880239,"rms_gain_error_db":0.6035104006782809,"reference_mean_power":8.744947435640977e-05,"output_mean_power":0.00010048662116119812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.34827454492152e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1350322554557276e-05,"output_power":2.5365650959664348e-05,"reference_active":true,"delta_db":0.7484157150499405},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413574366249565e-06,"output_power":3.718422412642903e-06,"reference_active":true,"delta_db":1.8769803937698253},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6206121669861414e-07,"output_power":2.0170374243814223e-07,"reference_active":true,"delta_db":0.950348611317155}],"active_band_count":3,"active_band_mae_db":1.191914906712307}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0022","method":"facodec","status":"measured","mos":{"status":"measured","value":3.89001202583313,"predicted_mos":3.89001202583313,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a534540b7ebbeb6ed7b4b3a65a59044891712446a668c427c120f23ced57e25f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3026531619998423},"started_at":"2026-10-13T04:48:22.008612+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.0402999982470647e-05,"native_input_samples":96000,"native_input_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","output_sha256":"b9dd67697f4ef8b702ad5e4f32ccb6e1df734efe47a6e36cd417ec0b446c46ba","output_samples":96000,"length_delta_samples":0,"analysis_s":1.98269967700071,"synthesis_s":5.901696318000177,"total_s":7.884395995000887,"rtf":1.3140659991668144,"output_peak":0.5403710603713989,"output_rms":0.07411039801715677,"finished_at":"2026-10-13T04:48:29.900312+00:00","evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"b9dd67697f4ef8b702ad5e4f32ccb6e1df734efe47a6e36cd417ec0b446c46ba","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.379143686408794,"spectral_convergence":0.20992564391563787,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.442449423024011,"spectral_convergence":0.20361469390833747,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.698075474475281,"spectral_convergence":0.2233117470521211,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.0023006282720753938,"delta_db":-0.41408723773097705,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.002457226212876417,"delta_db":-0.3624476972821238,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.00045357310441294533,"delta_db":-0.11110622525283594,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00011200794308792514,"delta_db":-0.4572251695094193,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.00013810581393510017,"delta_db":-0.7083014491341165,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":3.0809747873611824e-05,"delta_db":-0.6636213743618666,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4527981922118898,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3767722916757624,"pitch":null,"output_rms_dbfs":-22.602417086117722,"output_peak":0.5403710603713989,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.506556194636029},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.379143686408794,"spectral_convergence":0.20992564391563787,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.442449423024011,"spectral_convergence":0.20361469390833747,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.698075474475281,"spectral_convergence":0.2233117470521211,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.0023006282720753938,"delta_db":-0.41408723773097705,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.002457226212876417,"delta_db":-0.3624476972821238,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.00045357310441294533,"delta_db":-0.11110622525283594,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00011200794308792514,"delta_db":-0.4572251695094193,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.00013810581393510017,"delta_db":-0.7083014491341165,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":3.0809747873611824e-05,"delta_db":-0.6636213743618666,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4527981922118898,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3767722916757624,"pitch":null,"output_rms_dbfs":-22.602417086117722,"output_peak":0.5403710603713989,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.506556194636029},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9278163928577449,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0773958653322626,"rms_dbfs":-22.225644794441955,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.47852725154813,"mean":1.5013694763183594e-05,"zero_energy":false},"output":{"rms":0.07411039801715677,"rms_dbfs":-22.602417086117722,"peak":0.5403710603713989,"peak_dbfs":-5.346158354271519,"crest_factor_db":17.256258731846202,"mean":-0.00023765996769354084,"zero_energy":false},"rms_difference_db":-0.376772291675767},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":3.843023673566305e-06,"output_rms_dbfs":-54.15326940212324,"output_zero_energy":false,"output_peak":0.011378530412912369,"output_energy_fraction":0.00013061153716397307,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":188,"both_estimated_voiced_frames":164,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9425287356321839,"absolute_difference_cents_median":4.250575945891596,"absolute_difference_cents_p95":23.449014726537953,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031508938028130336,"log_rms_mae_db":1.2294967072707106,"rms_envelope_nrmse":0.10475195816732463,"envelope_correlation":0.9880162017655637,"rms_gain_error_db":-0.37666666541942395,"reference_mean_power":0.0059947180836205155,"output_mean_power":0.005496700797429151},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.905621652404067e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011843610245185533,"output_power":0.0010643392318017745,"reference_active":true,"delta_db":-0.4640403634984678},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002763334576078544,"output_power":0.00027394287302053976,"reference_active":true,"delta_db":-0.037734649614828315},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0009701982183084e-05,"output_power":4.767877773628046e-05,"reference_active":true,"delta_db":-0.20729152938113282}],"active_band_count":3,"active_band_mae_db":0.23635551416480963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002456604014165908,"log_rms_mae_db":0.9063280377182913,"rms_envelope_nrmse":0.09253077873148795,"envelope_correlation":0.9903742642882601,"rms_gain_error_db":-0.3758113523709284,"reference_mean_power":0.006022035579915752,"output_mean_power":0.0055228364424795935},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5974133279358877e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010595070071700143,"output_power":0.0009473663021990618,"reference_active":true,"delta_db":-0.48585900491829825},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014348331799351297,"output_power":0.00014028769306192248,"reference_active":true,"delta_db":-0.09781837407529435},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4072600706208143e-06,"output_power":3.2395342479097073e-06,"reference_active":true,"delta_db":-0.2192270821762703}],"active_band_count":3,"active_band_mae_db":0.2676348203899543}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026889287764829997,"log_rms_mae_db":1.3983898046924417,"rms_envelope_nrmse":0.1305597435623044,"envelope_correlation":0.9863919414603806,"rms_gain_error_db":-0.34504805847792686,"reference_mean_power":0.0016839373335741825,"output_mean_power":0.0015553248813217114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1138357532326787e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006324020549599114,"output_power":0.0005227604242686336,"reference_active":true,"delta_db":-0.8269057071964624},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002190350723697022,"output_power":0.00019588517065787322,"reference_active":true,"delta_db":-0.48512101284002657},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1194141233781235e-05,"output_power":2.8883856538238507e-05,"reference_active":true,"delta_db":-0.33417854869271785}],"active_band_count":3,"active_band_mae_db":0.5487350895764023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001970017370857221,"log_rms_mae_db":1.0979271687713832,"rms_envelope_nrmse":0.11250990288345222,"envelope_correlation":0.9892366664521206,"rms_gain_error_db":-0.3450405281495916,"reference_mean_power":0.001692399182790771,"output_mean_power":0.0015631431584350916},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.64481232375879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970882254118313,"output_power":0.0004893332761923733,"reference_active":true,"delta_db":-0.8643375682647387},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011301601543605377,"output_power":0.00010195349803233899,"reference_active":true,"delta_db":-0.4473786042474917},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.966963236644965e-06,"output_power":2.7472157645189052e-06,"reference_active":true,"delta_db":-0.3341939508618283}],"active_band_count":3,"active_band_mae_db":0.5486367077913529}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":1.9162876395467963,"rms_envelope_nrmse":0.20675619758898572,"envelope_correlation":0.9700288472242954,"rms_gain_error_db":-0.7882016982479745,"reference_mean_power":0.0002025729815791723,"output_mean_power":0.0001689512271307225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0818748871802835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.592086384233799e-05,"output_power":5.6448108071927654e-05,"reference_active":true,"delta_db":-0.6737349923342288},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.925318155530913e-05,"output_power":1.6444469038950793e-05,"reference_active":true,"delta_db":-0.6848265070372282},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.819375654274542e-06,"output_power":4.899782318851376e-06,"reference_active":true,"delta_db":1.0818441072705016}],"active_band_count":3,"active_band_mae_db":0.8134685355473196}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":1.5287410559747026,"rms_envelope_nrmse":0.17746696740100917,"envelope_correlation":0.9777299114978679,"rms_gain_error_db":-0.7882103835529839,"reference_mean_power":0.0002035903915876016,"output_mean_power":0.0001697994344241429},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.438814668171718e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.334752463763687e-05,"output_power":5.4021342740736207e-05,"reference_active":true,"delta_db":-0.6916427439899785},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.661338887813939e-06,"output_power":8.084209931572494e-06,"reference_active":true,"delta_db":-0.7739973319847182},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.128964187832131e-07,"output_power":3.9392669022175465e-07,"reference_active":true,"delta_db":1.0001481460718304}],"active_band_count":3,"active_band_mae_db":0.8219294073488425}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":2.1947370670465576,"rms_envelope_nrmse":0.20567992395779966,"envelope_correlation":0.9747920591903326,"rms_gain_error_db":-0.11004368371321364,"reference_mean_power":8.701244577979692e-05,"output_mean_power":8.483537966282948e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.568258861947873e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4501869786005614e-05,"output_power":2.2517962955464707e-05,"reference_active":true,"delta_db":-0.366701271626409},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.528335672589259e-06,"output_power":8.060413839772464e-06,"reference_active":true,"delta_db":2.5041872764989774},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7809461115210602e-06,"output_power":2.0978042681521655e-06,"reference_active":true,"delta_db":0.7111418605308346}],"active_band_count":3,"active_band_mae_db":1.1940101362187403}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":1.7342604506821784,"rms_envelope_nrmse":0.1508693517142542,"envelope_correlation":0.9859774572837801,"rms_gain_error_db":-0.11007307125091313,"reference_mean_power":8.744947435640977e-05,"output_mean_power":8.526089677103187e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.34827454492152e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1350322554557276e-05,"output_power":2.0010593409661307e-05,"reference_active":true,"delta_db":-0.2814447287456256},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413574366249565e-06,"output_power":4.084711214426925e-06,"reference_active":true,"delta_db":2.285006730092554},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6206121669861414e-07,"output_power":2.0967049119986614e-07,"reference_active":true,"delta_db":1.1185821751276472}],"active_band_count":3,"active_band_mae_db":1.2283445446552756}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0022","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.690927505493164,"predicted_mos":3.690927505493164,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"415fcef25cf5d0edb64f7f3b043abf9fdb58c35ccfe83fe645bc9b8e951b6a1a","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5256525139993755},"started_at":"2026-10-13T07:06:21.097049+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007874437000282342,"native_input_samples":264600,"native_input_sha256":"29de6064c83c21af9e43103a586b866ac94a9a3834d6a149ddf174e6039e334c","output_sha256":"98490419974fb2e01baab01132a15e0f8150cbee8448f7a7873a605d6c93d7de","output_samples":264704,"length_delta_samples":104,"analysis_s":0.2630580699988059,"synthesis_s":7.308848512000623,"total_s":7.5719065819994285,"rtf":1.2619844303332381,"output_peak":0.6202402114868164,"output_rms":0.07080805108515008,"finished_at":"2026-10-13T07:06:28.698555+00:00","evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"98490419974fb2e01baab01132a15e0f8150cbee8448f7a7873a605d6c93d7de","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.994991949621998,"output_peak":0.615293791046903},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.97499212333815,"spectral_convergence":0.48718800114043315,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.568357147545266,"spectral_convergence":0.34773058173016286,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.649350998649599,"spectral_convergence":0.3865795372403362,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.0016914018102241407,"delta_db":-1.750083851813843,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.0026865433514744607,"delta_db":0.025039436547794496,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.0004061979723842487,"delta_db":-0.5902016554442925,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":8.301419548063639e-05,"delta_db":-1.7581897551825971,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.00011840274511069636,"delta_db":-1.3768033525703278,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":3.3624357239087686e-05,"delta_db":-0.28396290645631006,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9640468263358608,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7680236625836752,"pitch":null,"output_rms_dbfs":-22.993668457025635,"output_peak":0.615293791046903,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.730900089844338},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7934626428214366,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0773958653322626,"rms_dbfs":-22.225644794441955,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.47852725154813,"mean":1.5013694763183594e-05,"zero_energy":false},"output":{"rms":0.07084620266403324,"rms_dbfs":-22.993668457025635,"peak":0.615293791046903,"peak_dbfs":-4.218349347440586,"crest_factor_db":18.77531910958505,"mean":-0.00023959114813253368,"zero_energy":false},"rms_difference_db":-0.7680236625836798},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":1.956266668162771e-06,"output_rms_dbfs":-57.08571944733617,"output_zero_energy":false,"output_peak":0.011694456874270108,"output_energy_fraction":7.275480369583688e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":161,"both_estimated_voiced_frames":134,"reference_voiced_output_unvoiced_frames":40,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.7701149425287356,"absolute_difference_cents_median":11.79000311526103,"absolute_difference_cents_p95":54.50560507133278,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031508938028130336,"log_rms_mae_db":1.7951640339494777,"rms_envelope_nrmse":0.13814207134186413,"envelope_correlation":0.9845607936661243,"rms_gain_error_db":-0.768042415533634,"reference_mean_power":0.0059947180836205155,"output_mean_power":0.005023015549601618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.905621652404067e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011843610245185533,"output_power":0.0010824062622564567,"reference_active":true,"delta_db":-0.39093810790119315},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002763334576078544,"output_power":0.00023202299642550016,"reference_active":true,"delta_db":-0.7590244002187106},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0009701982183084e-05,"output_power":3.351780373546535e-05,"reference_active":true,"delta_db":-1.7377871289842315}],"active_band_count":3,"active_band_mae_db":0.962583212368045}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002456604014165908,"log_rms_mae_db":1.572178072276532,"rms_envelope_nrmse":0.11111541042993034,"envelope_correlation":0.9929584966642967,"rms_gain_error_db":-0.7681218283808464,"reference_mean_power":0.006022035579915752,"output_mean_power":0.00504581280182776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5974133279358877e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010595070071700143,"output_power":0.0009806283833468708,"reference_active":true,"delta_db":-0.335993739834181},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014348331799351297,"output_power":0.00011669229173830823,"reference_active":true,"delta_db":-0.897592418999455},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4072600706208143e-06,"output_power":2.2091662955466465e-06,"reference_active":true,"delta_db":-1.8817687509674785}],"active_band_count":3,"active_band_mae_db":1.0384516366003715}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026889287764829997,"log_rms_mae_db":2.5030376640093235,"rms_envelope_nrmse":0.19051627192459283,"envelope_correlation":0.9691893914514088,"rms_gain_error_db":-0.03787477355415984,"reference_mean_power":0.0016839373335741825,"output_mean_power":0.0016693155855054729},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1138357532326787e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006324020549599114,"output_power":0.0006482699722479342,"reference_active":true,"delta_db":0.10762633142725042},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002190350723697022,"output_power":0.0001893040060922004,"reference_active":true,"delta_db":-0.6335385588574904},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1194141233781235e-05,"output_power":2.4499102431810626e-05,"reference_active":true,"delta_db":-1.0492286061680471}],"active_band_count":3,"active_band_mae_db":0.5967978321509294}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001970017370857221,"log_rms_mae_db":1.9509205906019895,"rms_envelope_nrmse":0.15064299791831015,"envelope_correlation":0.9787325915395902,"rms_gain_error_db":-0.03788054358027052,"reference_mean_power":0.001692399182790771,"output_mean_power":0.0016777017308893057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.64481232375879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970882254118313,"output_power":0.0006096588635370049,"reference_active":true,"delta_db":0.09048385208881429},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011301601543605377,"output_power":9.775424665120722e-05,"reference_active":true,"delta_db":-0.6300435832305801},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.966963236644965e-06,"output_power":2.260602720676128e-06,"reference_active":true,"delta_db":-1.1808791910277887}],"active_band_count":3,"active_band_mae_db":0.6338022087823943}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":3.301452941980805,"rms_envelope_nrmse":0.31013693966511524,"envelope_correlation":0.941005849419071,"rms_gain_error_db":-1.766193469172678,"reference_mean_power":0.0002025729815791723,"output_mean_power":0.00013488453944627795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0818748871802835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.592086384233799e-05,"output_power":3.877940768829605e-05,"reference_active":true,"delta_db":-2.304217183872119},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.925318155530913e-05,"output_power":1.3466747409334125e-05,"reference_active":true,"delta_db":-1.5523979189377612},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.819375654274542e-06,"output_power":3.420717502060422e-06,"reference_active":true,"delta_db":-0.47875165748270765}],"active_band_count":3,"active_band_mae_db":1.4451222534308625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":2.840177560866773,"rms_envelope_nrmse":0.27885351523936397,"envelope_correlation":0.9544363206015427,"rms_gain_error_db":-1.7661984456432647,"reference_mean_power":0.0002035903915876016,"output_mean_power":0.0001355618331913448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.438814668171718e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.334752463763687e-05,"output_power":3.7226596546036784e-05,"reference_active":true,"delta_db":-2.308763165816564},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.661338887813939e-06,"output_power":6.9083276520577544e-06,"reference_active":true,"delta_db":-1.4566438864335851},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.128964187832131e-07,"output_power":2.7664645398923133e-07,"reference_active":true,"delta_db":-0.5347548464408567}],"active_band_count":3,"active_band_mae_db":1.4333872995636685}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":4.703704406047507,"rms_envelope_nrmse":0.2279428718032872,"envelope_correlation":0.9715998334899072,"rms_gain_error_db":-0.6790142609633688,"reference_mean_power":8.701244577979692e-05,"output_mean_power":7.441833516303145e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.568258861947873e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4501869786005614e-05,"output_power":2.049509825673215e-05,"reference_active":true,"delta_db":-0.7754922276385559},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.528335672589259e-06,"output_power":4.800665188181281e-06,"reference_active":true,"delta_db":0.2536280573040716},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7809461115210602e-06,"output_power":6.643796565584082e-07,"reference_active":true,"delta_db":-4.282344528338213}],"active_band_count":3,"active_band_mae_db":1.7704882710936136}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":4.577142605486546,"rms_envelope_nrmse":0.1609263036191978,"envelope_correlation":0.9880109592595332,"rms_gain_error_db":-0.6789936268856916,"reference_mean_power":8.744947435640977e-05,"output_mean_power":7.479246393000578e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.34827454492152e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1350322554557276e-05,"output_power":1.8030592981802558e-05,"reference_active":true,"delta_db":-0.7339443077390851},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413574366249565e-06,"output_power":2.0053143402903748e-06,"reference_active":true,"delta_db":-0.8047822517138982},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6206121669861414e-07,"output_power":7.428121779654634e-08,"reference_active":true,"delta_db":-3.3880007998677915}],"active_band_count":3,"active_band_mae_db":1.6422424531069248}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0022","method":"golf","status":"measured","mos":{"status":"measured","value":3.0113601684570312,"predicted_mos":3.0113601684570312,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"9e08bb6dd127b9c2845169fb6a7ba31b984690faa125b845abce205de1c16893","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3845403630002693},"started_at":"2026-10-13T05:02:11.775034+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002569385000242619,"native_input_samples":144000,"native_input_sha256":"8da681ac7b73bffdeb882b0e2c10b437655cd5606b620fd0bb6e53198ca5bc63","output_sha256":"ccea6a621dbecb9144ad92e41e1c3aa2f5ebbb7f09eef88aada6d76e07ada825","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6050695320000159,"synthesis_s":0.1221981290000258,"total_s":0.7272676610000417,"rtf":0.12121127683334028,"output_peak":0.6131124496459961,"output_rms":0.07398285714951498,"finished_at":"2026-10-13T05:02:12.512855+00:00","analysis_stages":{"external_f0_s":0.08990711400019791},"evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"ccea6a621dbecb9144ad92e41e1c3aa2f5ebbb7f09eef88aada6d76e07ada825","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.61489121950992,"output_peak":0.5989050650373251},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.305518657565385,"spectral_convergence":0.5580728786202747,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.053936226940614,"spectral_convergence":0.22361443412550192,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4402298300964596,"spectral_convergence":0.22780953650140431,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002533486114263211,"output_power":0.0022490937727754697,"delta_db":-0.5171096521100567,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002675453287078469,"output_power":0.0025997481902660636,"delta_db":-0.12466088117082677,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00046606799060122845,"output_power":0.0003759776866767282,"delta_db":-0.9328720533721341,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012463977967559957,"output_power":8.383466415900858e-05,"delta_db":-1.7223304373661061,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016283409645220413,"output_power":0.0001289903287240032,"delta_db":-1.0118819912281058,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.595170372488576e-05,"output_power":3.8953541123779314e-05,"delta_db":0.34827467820473346,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7761882822419938,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3952693223664469,"pitch":null,"output_rms_dbfs":-22.61489121950992,"output_peak":0.5989050650373251,"reference_rms_dbfs":-22.219621897143472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.599894904867487},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9233943712676663,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07744955114005243,"rms_dbfs":-22.219621897143472,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.472504354249647,"mean":4.704687153556709e-06,"zero_energy":false},"output":{"rms":0.07400404167155367,"rms_dbfs":-22.614891219509914,"peak":0.5989050650373251,"peak_dbfs":-4.4528402800269875,"crest_factor_db":18.16205093948293,"mean":0.0006635298098846728,"zero_energy":false},"rms_difference_db":-0.39526932236644186},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":3.853636270551692e-06,"output_rms_dbfs":-54.14129279080397,"output_zero_energy":false,"output_peak":0.01760375433061194,"output_energy_fraction":0.00013156823334063423,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":200,"both_estimated_voiced_frames":166,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9540229885057471,"absolute_difference_cents_median":6.64548161631282,"absolute_difference_cents_p95":18.83118288481903,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00031508938028130336,"log_rms_mae_db":1.8057792760763582,"rms_envelope_nrmse":0.1284421868781763,"envelope_correlation":0.9818214249630633,"rms_gain_error_db":-0.39513478828446646,"reference_mean_power":0.006003054743647262,"output_mean_power":0.005480987668985848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.9067552506897725e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011859671725170844,"output_power":0.001115463780264985,"reference_active":true,"delta_db":-0.26617194892517243},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002771961770571282,"output_power":0.00028687186559308697,"reference_active":true,"delta_db":0.14900721316184115},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.989479679490919e-05,"output_power":4.8049386394462033e-05,"reference_active":true,"delta_db":-0.16367412261970557}],"active_band_count":3,"active_band_mae_db":0.19295109490223972}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002456604014165908,"log_rms_mae_db":1.1864956238867645,"rms_envelope_nrmse":0.08640988812063519,"envelope_correlation":0.9925751216542593,"rms_gain_error_db":-0.39388916981851924,"reference_mean_power":0.006030156666147423,"output_mean_power":0.005507311970300833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5982106056129403e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010610290913416146,"output_power":0.0010101315021219994,"reference_active":true,"delta_db":-0.21349376273642928},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014388786854457582,"output_power":0.00014441467480865228,"reference_active":true,"delta_db":0.01587147345253511},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.40082558392092e-06,"output_power":3.529391841529937e-06,"reference_active":true,"delta_db":0.1611551834341875}],"active_band_count":3,"active_band_mae_db":0.1301734732077173}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00026889287764829997,"log_rms_mae_db":1.7423238237168566,"rms_envelope_nrmse":0.1405031381575945,"envelope_correlation":0.9845901636890387,"rms_gain_error_db":-0.35741099985413555,"reference_mean_power":0.001686747994220838,"output_mean_power":0.0015534922907780551},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1149632050429066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006333130837589486,"output_power":0.0005888516512936088,"reference_active":true,"delta_db":-0.31612563041541647},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021957501448394197,"output_power":0.0002180199635628075,"reference_active":true,"delta_db":-0.03086657346947496},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1160302773896666e-05,"output_power":3.220913498187543e-05,"reference_active":true,"delta_db":0.14377392523853308}],"active_band_count":3,"active_band_mae_db":0.16358870970780817}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001970017370857221,"log_rms_mae_db":1.3222052174706251,"rms_envelope_nrmse":0.10727150991512069,"envelope_correlation":0.990847483182053,"rms_gain_error_db":-0.3574233305816814,"reference_mean_power":0.0016952360804995293,"output_mean_power":0.0015613053721467898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.65278669186941e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005981048548448826,"output_power":0.0005599641442761718,"reference_active":true,"delta_db":-0.2861710854002285},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011317826126691098,"output_power":0.00011294786538117267,"reference_active":true,"delta_db":-0.008849901418934998},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9675357344093136e-06,"output_power":2.7754994110467177e-06,"reference_active":true,"delta_db":-0.29054817824458945}],"active_band_count":3,"active_band_mae_db":0.19518972168791762}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":1.9864289338406627,"rms_envelope_nrmse":0.22429409355894328,"envelope_correlation":0.9767043999637067,"rms_gain_error_db":-1.4940512351464146,"reference_mean_power":0.00020291098851362707,"output_mean_power":0.00014384687883891683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0821256684172297e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.595439447675222e-05,"output_power":4.4276799112506955e-05,"reference_active":true,"delta_db":-1.73067520423642},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9229002774340723e-05,"output_power":1.0290454829955452e-05,"reference_active":true,"delta_db":-2.7152219143550256},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.820633954233586e-06,"output_power":1.7875151190493597e-06,"reference_active":true,"delta_db":-3.2988570715859207}],"active_band_count":3,"active_band_mae_db":2.581584730059122}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":1.597047817961983,"rms_envelope_nrmse":0.18413796066312382,"envelope_correlation":0.9910675819812386,"rms_gain_error_db":-1.494054088015615,"reference_mean_power":0.00020393211867284934,"output_mean_power":0.00014457067954231828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.441079235311567e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.337797856020578e-05,"output_power":4.1814895708610475e-05,"reference_active":true,"delta_db":-1.8060736531899981},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.642663119496166e-06,"output_power":4.914680538703894e-06,"reference_active":true,"delta_db":-2.927017009948345},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1234176496572926e-07,"output_power":1.5846244636021774e-07,"reference_active":true,"delta_db":-2.947037037322043}],"active_band_count":3,"active_band_mae_db":2.5600425668201288}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":2.4577697285609164,"rms_envelope_nrmse":0.292776728931224,"envelope_correlation":0.9499729193501204,"rms_gain_error_db":-0.1883330477446031,"reference_mean_power":8.71574937348812e-05,"output_mean_power":8.3458664059076e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.519260056872149e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4285652559112058e-05,"output_power":2.2709864711782124e-05,"reference_active":true,"delta_db":-0.29135230494438563},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.488517406836407e-06,"output_power":4.429896010898554e-06,"reference_active":true,"delta_db":-0.05709382206943929},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7669726577963266e-06,"output_power":2.981946046909585e-06,"reference_active":true,"delta_db":2.2726995212109626}],"active_band_count":3,"active_band_mae_db":0.8737152160749292}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":2.296957483784085,"rms_envelope_nrmse":0.206112646166313,"envelope_correlation":0.975119020723654,"rms_gain_error_db":-0.18832503740724638,"reference_mean_power":8.759621614753331e-05,"output_mean_power":8.387892247904759e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.300643654765101e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1161628268232665e-05,"output_power":2.024817389721966e-05,"reference_active":true,"delta_db":-0.1916321915189398},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3807664094473305e-06,"output_power":2.2382386054248414e-06,"reference_active":true,"delta_db":-0.2681040415479236},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6039105236844088e-07,"output_power":2.3421595613331773e-07,"reference_active":true,"delta_db":1.6443634145395591}],"active_band_count":3,"active_band_mae_db":0.7013665492021408}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0022","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.7329471111297607,"predicted_mos":3.7329471111297607,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"90f5775f9a63066078e21baf4593fd2f64dc2227bb2e0004c8d834c9e8335504","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6618768610005645},"started_at":"2026-10-13T05:07:13.195928+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004841441000280611,"native_input_samples":144000,"native_input_sha256":"8da681ac7b73bffdeb882b0e2c10b437655cd5606b620fd0bb6e53198ca5bc63","output_sha256":"64680665623a27c28cbf0191cbf698f11ab8764c397a8e81e229eca8a760d7c0","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3165140589999282,"synthesis_s":3.2613471340000615,"total_s":5.57786119299999,"rtf":0.929643532166665,"output_peak":0.5433101654052734,"output_rms":0.0807054891945447,"finished_at":"2026-10-13T05:07:18.786818+00:00","analysis_stages":{"external_f0_s":1.4748489579997113},"evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"64680665623a27c28cbf0191cbf698f11ab8764c397a8e81e229eca8a760d7c0","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.859570491615603,"output_peak":0.5441780822654313},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.320455390541429,"spectral_convergence":0.5544940634497532,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.213713904035461,"spectral_convergence":0.29949480942243,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.319858378730856,"spectral_convergence":0.2860401675534376,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.0033078317188416097,"delta_db":1.1628823109744997,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.0024873595641010846,"delta_db":-0.309513393931367,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.00042616363470263194,"delta_db":-0.38181528911414625,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00011235621935940314,"delta_db":-0.4437422154834757,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.00015765343859326317,"delta_db":-0.13338659046281837,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":2.9006808873571984e-05,"delta_db":-0.9255032840554756,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5594738473369637,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.36836812236427297,"pitch":null,"output_rms_dbfs":-21.857276672077685,"output_peak":0.5441780822654313,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.618009224435916},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9171925182515323,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0773958653322626,"rms_dbfs":-22.225644794441955,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.47852725154813,"mean":1.5013694763183594e-05,"zero_energy":false},"output":{"rms":0.08074881661344373,"rms_dbfs":-21.857276672077685,"peak":0.5441780822654313,"peak_dbfs":-5.285179083645204,"crest_factor_db":16.57209758843248,"mean":-9.682947282818603e-05,"zero_energy":false},"rms_difference_db":0.3683681223642701},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":3.250337799059725e-06,"output_rms_dbfs":-54.8807150159218,"output_zero_energy":false,"output_peak":0.03718951312013711,"output_energy_fraction":9.305140562025215e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":220,"both_estimated_voiced_frames":169,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":51,"paired_voiced_fraction":0.9712643678160919,"absolute_difference_cents_median":6.147672740562241,"absolute_difference_cents_p95":33.791820777341734,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031508938028130336,"log_rms_mae_db":2.799650125970185,"rms_envelope_nrmse":0.16883140228390917,"envelope_correlation":0.9678958077147589,"rms_gain_error_db":0.3685296119800595,"reference_mean_power":0.0059947180836205155,"output_mean_power":0.00652561917358382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.905621652404067e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011843610245185533,"output_power":0.0012771840808448686,"reference_active":true,"delta_db":0.3276938974588856},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002763334576078544,"output_power":0.0002619302326307435,"reference_active":true,"delta_db":-0.23247842490377774},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0009701982183084e-05,"output_power":3.711090379603874e-05,"reference_active":true,"delta_db":-1.2955273525638629}],"active_band_count":3,"active_band_mae_db":0.6185665583088421}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002456604014165908,"log_rms_mae_db":2.4952718108101473,"rms_envelope_nrmse":0.1293211978110961,"envelope_correlation":0.9802937106375663,"rms_gain_error_db":0.3682297348022877,"reference_mean_power":0.006022035579915752,"output_mean_power":0.006554903320780841},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5974133279358877e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010595070071700143,"output_power":0.001148050221482415,"reference_active":true,"delta_db":0.34857053387025017},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014348331799351297,"output_power":0.00013566788776922444,"reference_active":true,"delta_db":-0.24324347537127214},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4072600706208143e-06,"output_power":2.740720899404503e-06,"reference_active":true,"delta_db":-0.9454047229833684}],"active_band_count":3,"active_band_mae_db":0.5124062440749636}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026889287764829997,"log_rms_mae_db":2.1136964488354772,"rms_envelope_nrmse":0.20177585580089544,"envelope_correlation":0.9656354679456616,"rms_gain_error_db":-0.42674206337077014,"reference_mean_power":0.0016839373335741825,"output_mean_power":0.001526341523525399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1138357532326787e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006324020549599114,"output_power":0.0005759563020907701,"reference_active":true,"delta_db":-0.4060373780948945},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002190350723697022,"output_power":0.00019435718801191448,"reference_active":true,"delta_db":-0.5191305357963026},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1194141233781235e-05,"output_power":2.6258873751459824e-05,"reference_active":true,"delta_db":-0.747969389495336}],"active_band_count":3,"active_band_mae_db":0.5577124344621777}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001970017370857221,"log_rms_mae_db":1.6032245152333193,"rms_envelope_nrmse":0.13628434042180065,"envelope_correlation":0.983850396582852,"rms_gain_error_db":-0.42675805093542635,"reference_mean_power":0.001692399182790771,"output_mean_power":0.001534005800691851},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.64481232375879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970882254118313,"output_power":0.0005417118522434952,"reference_active":true,"delta_db":-0.42270169248708167},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011301601543605377,"output_power":0.00010073502106242287,"reference_active":true,"delta_db":-0.49959509894764553},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.966963236644965e-06,"output_power":2.245222902973498e-06,"reference_active":true,"delta_db":-1.2105270134642692}],"active_band_count":3,"active_band_mae_db":0.7109412682996655}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":2.3161525298450583,"rms_envelope_nrmse":0.22080093609333384,"envelope_correlation":0.9622653841059136,"rms_gain_error_db":-0.09726956163125366,"reference_mean_power":0.0002025729815791723,"output_mean_power":0.00019808635663172097},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0818748871802835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.592086384233799e-05,"output_power":6.292630606934373e-05,"reference_active":true,"delta_db":-0.20190651492672199},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.925318155530913e-05,"output_power":1.6662420786620538e-05,"reference_active":true,"delta_db":-0.627644084100186},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.819375654274542e-06,"output_power":3.2564430802110885e-06,"reference_active":true,"delta_db":-0.6924888397317526}],"active_band_count":3,"active_band_mae_db":0.5073464795862201}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":1.517841930707741,"rms_envelope_nrmse":0.12946081640923154,"envelope_correlation":0.9860162390795593,"rms_gain_error_db":-0.09730454542173715,"reference_mean_power":0.0002035903915876016,"output_mean_power":0.00019907962919399035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.438814668171718e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.334752463763687e-05,"output_power":6.022946001736355e-05,"reference_active":true,"delta_db":-0.2192067954368247},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.661338887813939e-06,"output_power":8.234210245907185e-06,"reference_active":true,"delta_db":-0.6941536425086192},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.128964187832131e-07,"output_power":2.7005174030770354e-07,"reference_active":true,"delta_db":-0.6395361214001285}],"active_band_count":3,"active_band_mae_db":0.5176321864485242}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":2.363684226445204,"rms_envelope_nrmse":0.2646905468494713,"envelope_correlation":0.9584376615962693,"rms_gain_error_db":-0.698709803421316,"reference_mean_power":8.701244577979692e-05,"output_mean_power":7.4081607202058e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.568258861947873e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4501869786005614e-05,"output_power":1.9959076117285062e-05,"reference_active":true,"delta_db":-0.8905879307384814},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.528335672589259e-06,"output_power":4.23004586270402e-06,"reference_active":true,"delta_db":-0.2959353630546469},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7809461115210602e-06,"output_power":1.258697724361785e-06,"reference_active":true,"delta_db":-1.5072933160718183}],"active_band_count":3,"active_band_mae_db":0.8979388699549822}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":1.4470654757684704,"rms_envelope_nrmse":0.17032631745256563,"envelope_correlation":0.9845505345074234,"rms_gain_error_db":-0.6987561993447131,"reference_mean_power":8.744947435640977e-05,"output_mean_power":7.445289397998246e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.34827454492152e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1350322554557276e-05,"output_power":1.7354278026984644e-05,"reference_active":true,"delta_db":-0.8999788975155633},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413574366249565e-06,"output_power":2.0859507388971373e-06,"reference_active":true,"delta_db":-0.6335663663690038},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6206121669861414e-07,"output_power":1.30843187857507e-07,"reference_active":true,"delta_db":-0.929279784006062}],"active_band_count":3,"active_band_mae_db":0.8209416826302096}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0022","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.1925971508026123,"predicted_mos":3.1925971508026123,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"68eec2a70e50a49fe31c3c4986ad91c2dd0f58e7f62702d81129dee74529069a","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.048766403000627},"started_at":"2026-10-13T05:10:17.332057+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004026855000120122,"native_input_samples":144000,"native_input_sha256":"8da681ac7b73bffdeb882b0e2c10b437655cd5606b620fd0bb6e53198ca5bc63","output_sha256":"bdf33537ecf6cc68890d467b4cf9b81c468de36c435da04beb7c134af5937ab2","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0110927959995024,"synthesis_s":0.14856976400005806,"total_s":1.1596625599995605,"rtf":0.19327709333326007,"output_peak":0.597366213798523,"output_rms":0.07585154320656491,"finished_at":"2026-10-13T05:10:18.506190+00:00","analysis_stages":{"external_f0_s":0.1644725419992028},"evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"bdf33537ecf6cc68890d467b4cf9b81c468de36c435da04beb7c134af5937ab2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.398023957956724,"output_peak":0.594614146393603},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.78623352413066,"spectral_convergence":0.177177627844189,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.364316045037695,"spectral_convergence":0.15123162120692718,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.81922217251128,"spectral_convergence":0.1592015587537183,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0025307845601282375,"output_power":0.0024785850056894473,"delta_db":-0.09051358156559722,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026710985360254404,"output_power":0.0025550266477172333,"delta_db":-0.1929447533031557,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004653266401104453,"output_power":0.00045180186857770536,"delta_db":-0.1280989399560248,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012444324831517497,"output_power":0.00011122531419366045,"delta_db":-0.4876769764546062,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016257063671523992,"output_power":0.00013828712970622795,"delta_db":-0.7026034418593147,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.589634923518972e-05,"output_power":3.141551042212073e-05,"delta_db":-0.5790616150211219,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.36348321802663675,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16535163307575618,"pitch":null,"output_rms_dbfs":-22.390996427517713,"output_peak":0.594614146393603,"reference_rms_dbfs":-22.225644794441955,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.656590580559878},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9648184425274666,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0773958653322626,"rms_dbfs":-22.225644794441955,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.47852725154813,"mean":1.5013694763183594e-05,"zero_energy":false},"output":{"rms":0.07593643049489748,"rms_dbfs":-22.390996427517713,"peak":0.594614146393603,"peak_dbfs":-4.515295255208301,"crest_factor_db":17.875701172309416,"mean":-0.00030055073989193696,"zero_energy":false},"rms_difference_db":-0.1653516330757583},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":2.759036945443142e-06,"output_rms_dbfs":-55.59242483973638,"output_zero_energy":false,"output_peak":0.013293601273652625,"output_energy_fraction":8.93149030337945e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":181,"both_estimated_voiced_frames":165,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9482758620689655,"absolute_difference_cents_median":1.5695546452912053,"absolute_difference_cents_p95":7.215517965750874,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031508938028130336,"log_rms_mae_db":0.9299814431247623,"rms_envelope_nrmse":0.060387807630195106,"envelope_correlation":0.9957772817079265,"rms_gain_error_db":-0.1654394806612801,"reference_mean_power":0.0059947180836205155,"output_mean_power":0.005770651085486663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.905621652404067e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011843610245185533,"output_power":0.0011573237390456697,"reference_active":true,"delta_db":-0.1002924552494261},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002763334576078544,"output_power":0.000248416145308147,"reference_active":true,"delta_db":-0.4625365262553457},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0009701982183084e-05,"output_power":3.67063368159469e-05,"reference_active":true,"delta_db":-1.3431322114839381}],"active_band_count":3,"active_band_mae_db":0.6353203976629033}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002456604014165908,"log_rms_mae_db":0.5364003325232075,"rms_envelope_nrmse":0.03371849640144496,"envelope_correlation":0.9988905610292769,"rms_gain_error_db":-0.16464013489291468,"reference_mean_power":0.006022035579915752,"output_mean_power":0.005798014586820633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5974133279358877e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010595070071700143,"output_power":0.0010330322166879613,"reference_active":true,"delta_db":-0.10989967444466099},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014348331799351297,"output_power":0.00012903838880104715,"reference_active":true,"delta_db":-0.4608247927614054},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4072600706208143e-06,"output_power":2.2948308809984213e-06,"reference_active":true,"delta_db":-1.716545983764317}],"active_band_count":3,"active_band_mae_db":0.7624234836567946}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026889287764829997,"log_rms_mae_db":0.9899203614368598,"rms_envelope_nrmse":0.07731049575514243,"envelope_correlation":0.9956300464888782,"rms_gain_error_db":-0.2692311483325321,"reference_mean_power":0.0016839373335741825,"output_mean_power":0.0015827153365269965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1138357532326787e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006324020549599114,"output_power":0.0006121511446781417,"reference_active":true,"delta_db":-0.14134606552569662},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002190350723697022,"output_power":0.00018991615832911587,"reference_active":true,"delta_db":-0.6195174389678577},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1194141233781235e-05,"output_power":2.2861514249427597e-05,"reference_active":true,"delta_db":-1.349680412915415}],"active_band_count":3,"active_band_mae_db":0.7035146391363231}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001970017370857221,"log_rms_mae_db":0.6001981516484685,"rms_envelope_nrmse":0.05115795851626129,"envelope_correlation":0.9983891677171739,"rms_gain_error_db":-0.26929365671872774,"reference_mean_power":0.001692399182790771,"output_mean_power":0.0015906456468871744},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.64481232375879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970882254118313,"output_power":0.0005740527007565402,"reference_active":true,"delta_db":-0.17086742454550527},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011301601543605377,"output_power":9.753320538612959e-05,"reference_active":true,"delta_db":-0.6398749412302883},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.966963236644965e-06,"output_power":1.987286936767109e-06,"reference_active":true,"delta_db":-1.7405158729123873}],"active_band_count":3,"active_band_mae_db":0.8504194128960604}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":1.3578428157778908,"rms_envelope_nrmse":0.1523607857334183,"envelope_correlation":0.9851713336216942,"rms_gain_error_db":-0.664360359786414,"reference_mean_power":0.0002025729815791723,"output_mean_power":0.0001738383072320253},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0818748871802835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.592086384233799e-05,"output_power":6.135237799757898e-05,"reference_active":true,"delta_db":-0.31191489324682187},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.925318155530913e-05,"output_power":1.915092951705539e-05,"reference_active":true,"delta_db":-0.023126482906737432},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.819375654274542e-06,"output_power":3.1036958638534276e-06,"reference_active":true,"delta_db":-0.9011321800354062}],"active_band_count":3,"active_band_mae_db":0.4120578520629885}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":0.8651871510538957,"rms_envelope_nrmse":0.11323245743154065,"envelope_correlation":0.993321504433987,"rms_gain_error_db":-0.6644175713616812,"reference_mean_power":0.0002035903915876016,"output_mean_power":0.00017470909761593137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.438814668171718e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.334752463763687e-05,"output_power":5.9808238779172806e-05,"reference_active":true,"delta_db":-0.24968635512607418},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.661338887813939e-06,"output_power":9.886085097388171e-06,"reference_active":true,"delta_db":0.09987028467728488},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.128964187832131e-07,"output_power":2.3325381790826586e-07,"reference_active":true,"delta_db":-1.2757183148452595}],"active_band_count":3,"active_band_mae_db":0.5417583182162061}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":1.270389130832885,"rms_envelope_nrmse":0.12970071870136252,"envelope_correlation":0.9909478672819276,"rms_gain_error_db":-0.435310069535939,"reference_mean_power":8.701244577979692e-05,"output_mean_power":7.871370760495974e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.568258861947873e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4501869786005614e-05,"output_power":2.1270723729974372e-05,"reference_active":true,"delta_db":-0.614169606043954},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.528335672589259e-06,"output_power":5.1253330577404235e-06,"reference_active":true,"delta_db":0.5378347986953312},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7809461115210602e-06,"output_power":1.1336099641244441e-06,"reference_active":true,"delta_db":-1.9618712404349288}],"active_band_count":3,"active_band_mae_db":1.0379585483914047}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":0.9796990701469652,"rms_envelope_nrmse":0.07965782421858902,"envelope_correlation":0.9974455433086534,"rms_gain_error_db":-0.4353588297170048,"reference_mean_power":8.744947435640977e-05,"output_mean_power":7.910816676424947e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.34827454492152e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1350322554557276e-05,"output_power":1.8576421382764886e-05,"reference_active":true,"delta_db":-0.6044238667666978},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413574366249565e-06,"output_power":2.6481775403936676e-06,"reference_active":true,"delta_db":0.402864132256635},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6206121669861414e-07,"output_power":9.216548090637353e-08,"reference_active":true,"delta_db":-2.451108015971795}],"active_band_count":3,"active_band_mae_db":1.1527986716650427}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0022","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2837111949920654,"predicted_mos":3.2837111949920654,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287431,"input_sample_rate":48000,"input_seconds":5.988145833333333,"input_float32_sha256":"3857926a2c065b6dfe0eb005a13a063e9d19b4959388b09fb97a09b0da76081b","internal_samples":95811,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0628965229989262},"started_at":"2026-10-13T05:13:35.491917+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010094880999531597,"native_input_samples":288000,"native_input_sha256":"f1ef87f1778b00da2b1093079485cd57e0edc7a336bcac63fb149953ec08f5f4","output_sha256":"990aee25221797b2d1e1231f3fbceae553372df6b87a81f4962dbaa2b8566ab9","output_samples":287431,"length_delta_samples":-569,"analysis_s":3.2457433700001275,"synthesis_s":0.9763278850005008,"total_s":4.222071255000628,"rtf":0.7036785425001048,"output_peak":0.6788515368539957,"output_rms":0.07426108642407026,"finished_at":"2026-10-13T05:13:39.736269+00:00","evaluation_input":{"reference":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"990aee25221797b2d1e1231f3fbceae553372df6b87a81f4962dbaa2b8566ab9","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95811,"length_delta_samples":-189,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.582048531553728,"output_peak":0.6665478154604385},"common_prefix":{"samples":95811,"sample_rate":16000,"reference_coverage":0.99803125,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95811,"output_samples":95811,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.0563775059515965,"spectral_convergence":0.6102953846260799,"spectral_convergence_status":"ok","amplitude_floor":2.1903561449018732e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.570111939218122,"spectral_convergence":0.2454449509387918,"spectral_convergence_status":"ok","amplitude_floor":2.0456135368211886e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.481197656868367,"spectral_convergence":0.22407582238347024,"spectral_convergence_status":"ok","amplitude_floor":1.0394418151882858e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002533942257991319,"output_power":0.0022447938495980097,"delta_db":-0.5262025042555003,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026763659979622684,"output_power":0.0024994333852374545,"delta_db":-0.2970393727456199,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004662643148783451,"output_power":0.00044261825722170177,"delta_db":-0.2260285454586122,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012468137007693773,"output_power":0.00011930982315361904,"delta_db":-0.1912536395530913,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00016288443140098386,"output_power":0.0001755801439327978,"delta_db":0.3259582455901056,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.5964965929369615e-05,"output_power":3.643544360358189e-05,"delta_db":0.056444082352818566,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2704877316592913,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3636358334578127,"pitch":null,"output_rms_dbfs":-22.582048531553728,"output_peak":0.6665478154604385,"reference_rms_dbfs":-22.218412698095914,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.035895700679362},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95811,"output_samples":95811,"intelligibility":{"estoi":{"value":0.9142160016643316,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07746033396674212,"rms_dbfs":-22.21841269809591,"peak":0.64959716796875,"peak_dbfs":-3.747117542893826,"crest_factor_db":18.471295155202085,"mean":1.0211385301368057e-05,"zero_energy":false},"output":{"rms":0.07428439205342643,"rms_dbfs":-22.582048531553728,"peak":0.6665478154604385,"peak_dbfs":-3.5233738118511733,"crest_factor_db":19.05867471970255,"mean":0.0023729427325154834,"zero_energy":false},"rms_difference_db":-0.3636358334578169},"full_scale":{"reference":{"sample_count":95811,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95811,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":131,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.92813175660558e-06,"threshold_dbfs":-50.03132467932869,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":112,"quiet_samples":17920,"reference_mean_power":2.835093515126833e-06,"output_mean_power":3.169030917420274e-06,"output_rms_dbfs":-54.99073523763464,"output_zero_energy":false,"output_peak":0.016506852876748978,"output_energy_fraction":0.00010741228653354112,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":174,"output_estimated_voiced_frames":165,"both_estimated_voiced_frames":151,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.867816091954023,"absolute_difference_cents_median":4.897666138325675,"absolute_difference_cents_p95":30.088285852236467,"difference_over_600_cents_frames":1,"uncovered_tail_samples":131,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95811,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1196,"rms_floor":0.00031508938028130336,"log_rms_mae_db":1.8588595923976585,"rms_envelope_nrmse":0.13293431055310334,"envelope_correlation":0.9794376247037042,"rms_gain_error_db":-0.3634139290551168,"reference_mean_power":0.006007159427746688,"output_mean_power":0.005524942534091784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":1.907340962853286e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011867669229289847,"output_power":0.0011439160178085492,"reference_active":true,"delta_db":-0.1597129204777474},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027764798350746504,"output_power":0.00026650544658487845,"reference_active":true,"delta_db":-0.17788434489620059},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.972046580013219e-05,"output_power":4.7863138320903965e-05,"reference_active":true,"delta_db":-0.16534017313928326}],"active_band_count":3,"active_band_mae_db":0.16764581283774374}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1190,"rms_floor":0.0002456604014165908,"log_rms_mae_db":1.2660180700735364,"rms_envelope_nrmse":0.08109229822771075,"envelope_correlation":0.993109016862579,"rms_gain_error_db":-0.3630714988228735,"reference_mean_power":0.006034229806464894,"output_mean_power":0.0055502774815520284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":1.5986197031440293e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010617875289451854,"output_power":0.001031858410626178,"reference_active":true,"delta_db":-0.1241751166677839},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014402732550033556,"output_power":0.00013781408751332424,"reference_active":true,"delta_db":-0.19151282261429403},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4614744821871644e-06,"output_power":3.5389873462348856e-06,"reference_active":true,"delta_db":0.09617875443604557}],"active_band_count":3,"active_band_mae_db":0.13728889790604118}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1196,"rms_floor":0.00026889287764829997,"log_rms_mae_db":1.7992484024629263,"rms_envelope_nrmse":0.15456950406017572,"envelope_correlation":0.9797353213332496,"rms_gain_error_db":-0.2509147983932129,"reference_mean_power":0.0016881546697690068,"output_mean_power":0.0015933851106752772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":1.1155330089425426e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006337820190195963,"output_power":0.0005983298952834277,"reference_active":true,"delta_db":-0.249992109819441},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021982703243859082,"output_power":0.0002019150742136328,"reference_active":true,"delta_db":-0.3691235424033764},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1127429938089466e-05,"output_power":2.9166960481040884e-05,"reference_active":true,"delta_db":-0.2825209107586625}],"active_band_count":3,"active_band_mae_db":0.30054552099382664}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1190,"rms_floor":0.0001970017370857221,"log_rms_mae_db":1.3983967278330593,"rms_envelope_nrmse":0.10367055695178168,"envelope_correlation":0.9903158307096449,"rms_gain_error_db":-0.25092559919273816,"reference_mean_power":0.0016966575699121965,"output_mean_power":0.001601406692686551},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":9.656789750421897e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005986259456952069,"output_power":0.0005652775540440363,"reference_active":true,"delta_db":-0.24893795236804292},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011320330180143014,"output_power":0.00010409159818926406,"reference_active":true,"delta_db":-0.36443417500296277},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0159004797289938e-06,"output_power":2.6093491986762077e-06,"reference_active":true,"delta_db":-0.6288480348838701}],"active_band_count":3,"active_band_mae_db":0.41407338741829197}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1196,"rms_floor":7.658803471872854e-05,"log_rms_mae_db":2.0475928863516923,"rms_envelope_nrmse":0.2303883416099168,"envelope_correlation":0.9590175579899836,"rms_gain_error_db":-0.029903607814979083,"reference_mean_power":0.00020308060924087022,"output_mean_power":0.00020168708857983236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":1.082253099755478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.59722410300857e-05,"output_power":6.405816959940306e-05,"reference_active":true,"delta_db":-0.12786712047659254},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9216614934514734e-05,"output_power":1.85626840457312e-05,"reference_active":true,"delta_db":-0.15036115133072586},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.817879439185974e-06,"output_power":3.5808559416696753e-06,"reference_active":true,"delta_db":-0.27835360507631}],"active_band_count":3,"active_band_mae_db":0.1855272922945428}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1190,"rms_floor":6.479997453028038e-05,"log_rms_mae_db":1.2386314621141863,"rms_envelope_nrmse":0.1375828919956461,"envelope_correlation":0.9840530689824442,"rms_gain_error_db":-0.02991295452364762,"reference_mean_power":0.00020410346548050467,"output_mean_power":0.00020270248982452775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":9.442237848364784e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.339388861276472e-05,"output_power":6.153440999091509e-05,"reference_active":true,"delta_db":-0.12929351612564863},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628775963550532e-06,"output_power":9.340899067563806e-06,"reference_active":true,"delta_db":-0.1318240255980008},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.160888614396936e-07,"output_power":3.162886984064084e-07,"reference_active":true,"delta_db":0.002744818928570919}],"active_band_count":3,"active_band_mae_db":0.08795412021740678}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1196,"rms_floor":6.558280881155458e-05,"log_rms_mae_db":2.0777865627144485,"rms_envelope_nrmse":0.30168536304762433,"envelope_correlation":0.9491804506917326,"rms_gain_error_db":0.36575309289829305,"reference_mean_power":8.723035246781682e-05,"output_mean_power":9.48949140121179e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":6.494764517237398e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4177733140740963e-05,"output_power":2.3854704972920896e-05,"reference_active":true,"delta_db":-0.05841530073131379},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469024725428864e-06,"output_power":7.751422543233778e-06,"reference_active":true,"delta_db":2.3916865421084683},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7445613721850562e-06,"output_power":2.9634431291963083e-06,"reference_active":true,"delta_db":2.301103450391207}],"active_band_count":3,"active_band_mae_db":1.5837350977436628}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":51,"metrics":{"frame_count":1190,"rms_floor":5.713713309096249e-05,"log_rms_mae_db":1.2988385583830933,"rms_envelope_nrmse":0.18670890462993928,"envelope_correlation":0.9805018958616736,"rms_gain_error_db":0.36574694403191105,"reference_mean_power":8.7669809320809e-05,"output_mean_power":9.537284904542348e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":6.276822027006862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.106799719354269e-05,"output_power":2.161656321584606e-05,"reference_active":true,"delta_db":0.11163395460290257},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.361051315969183e-06,"output_power":4.165556121087692e-06,"reference_active":true,"delta_db":2.4656756364432035},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6298174891853744e-07,"output_power":2.4635575066182626e-07,"reference_active":true,"delta_db":1.7942373062504855}],"active_band_count":3,"active_band_mae_db":1.4571822990988637}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"hnm","status":"measured","mos":{"status":"measured","value":3.2497663497924805,"predicted_mos":3.2497663497924805,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"803f6ab95ff7dfe5f80df00f784e179cfbd0d98555e95504919a6546a27ca834","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1228206070009037},"started_at":"2026-10-13T04:38:35.576105+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6696000784577336e-05,"native_input_samples":96000,"native_input_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","output_sha256":"8b726f8b82a1df2b799f6a4bea5fc424a33c23949258da6fc8bc216016f7fe37","output_samples":96000,"length_delta_samples":0,"analysis_s":1.294695153000248,"synthesis_s":0.5005759080004282,"total_s":1.7952710610006761,"rtf":0.2992118435001127,"output_peak":0.5176862870743544,"output_rms":0.06286098219254158,"finished_at":"2026-10-13T04:38:37.377986+00:00","evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"8b726f8b82a1df2b799f6a4bea5fc424a33c23949258da6fc8bc216016f7fe37","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.999773902430519,"spectral_convergence":0.21088459093840176,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.45380138373778,"spectral_convergence":0.2121162169269198,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.727634703144092,"spectral_convergence":0.21401164816829615,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.0015603894112878698,"delta_db":0.009978434510136816,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.001847040708979701,"delta_db":-0.2546445962314444,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.0002108834184916705,"delta_db":0.023681040853658786,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00016061608862849366,"delta_db":-0.7029529876735663,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":9.60660397782127e-05,"delta_db":-0.5235834404196443,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":7.650741504508136e-05,"delta_db":-0.16758754618543853,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.28040467431231486,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.162175821727561,"pitch":null,"output_rms_dbfs":-24.03237674970683,"output_peak":0.5176862870743544,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.727069996437464},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.999773902430519,"spectral_convergence":0.21088459093840176,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.45380138373778,"spectral_convergence":0.2121162169269198,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.727634703144092,"spectral_convergence":0.21401164816829615,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.0015603894112878698,"delta_db":0.009978434510136816,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.001847040708979701,"delta_db":-0.2546445962314444,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.0002108834184916705,"delta_db":0.023681040853658786,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00016061608862849366,"delta_db":-0.7029529876735663,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":9.60660397782127e-05,"delta_db":-0.5235834404196443,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":7.650741504508136e-05,"delta_db":-0.16758754618543853,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.28040467431231486,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.162175821727561,"pitch":null,"output_rms_dbfs":-24.03237674970683,"output_peak":0.5176862870743544,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.727069996437464},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8963480036579523,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06404569659064947,"rms_dbfs":-23.870200927979273,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.867078192515194,"mean":-2.6813824971516927e-05,"zero_energy":false},"output":{"rms":0.06286098219254158,"rms_dbfs":-24.03237674970683,"peak":0.5176862870743544,"peak_dbfs":-5.718666776731994,"crest_factor_db":18.313709972974838,"mean":6.0475446021842865e-05,"zero_energy":false},"rms_difference_db":-0.1621758217275584},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":2.3777390981586174e-06,"output_rms_dbfs":-56.238358008721974,"output_zero_energy":false,"output_peak":0.023224266173485344,"output_energy_fraction":0.00010429991704579837,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":244,"both_estimated_voiced_frames":227,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9458333333333333,"absolute_difference_cents_median":2.334215749792135,"absolute_difference_cents_p95":8.415790378158457,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023872023988588299,"log_rms_mae_db":0.8518566701071917,"rms_envelope_nrmse":0.06592494178921925,"envelope_correlation":0.9949907477594534,"rms_gain_error_db":-0.16137183882374287,"reference_mean_power":0.004104010412713017,"output_mean_power":0.003954315072322327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.972155761177528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003715407512890731,"output_power":0.0003598643727890607,"reference_active":true,"delta_db":-0.138676020880779},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.499306084417706e-05,"output_power":5.0077810489603764e-05,"reference_active":true,"delta_db":-0.4066256031432478},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6530449727349195e-05,"output_power":1.5472182965125398e-05,"reference_active":true,"delta_db":-0.28733076660386014}],"active_band_count":3,"active_band_mae_db":0.27754413020929564}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00021497588341920653,"log_rms_mae_db":0.49101132800409053,"rms_envelope_nrmse":0.03682815230299764,"envelope_correlation":0.9986460767718954,"rms_gain_error_db":-0.16080057189488456,"reference_mean_power":0.00411739985369924,"output_mean_power":0.003967738006605828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.973189563381596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033993681960372955,"output_power":0.00032883702706858176,"reference_active":true,"delta_db":-0.14417493697875708},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413396717591137e-05,"output_power":2.2030360584538954e-05,"reference_active":true,"delta_db":-0.3960711214738644},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.827774931478278e-07,"output_power":7.640154044568864e-07,"reference_active":true,"delta_db":-0.6274913695214078}],"active_band_count":3,"active_band_mae_db":0.38924580932467645}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001937632671887944,"log_rms_mae_db":1.715850145545658,"rms_envelope_nrmse":0.10020813071637935,"envelope_correlation":0.9928576345041884,"rms_gain_error_db":-0.29414915808054304,"reference_mean_power":0.0014622039835396937,"output_mean_power":0.0013664478036796218},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.89031570773124e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020584065582039213,"output_power":0.00018992623904282038,"reference_active":true,"delta_db":-0.3494618855163501},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2632804837837524e-05,"output_power":3.470598535388437e-05,"reference_active":true,"delta_db":-0.893395269104445},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.198246611681044e-06,"output_power":6.794288294876934e-06,"reference_active":true,"delta_db":-0.815770073160142}],"active_band_count":3,"active_band_mae_db":0.6862090759269791}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001823230477151681,"log_rms_mae_db":1.1178642166821915,"rms_envelope_nrmse":0.06008920046119009,"envelope_correlation":0.9978834370093631,"rms_gain_error_db":-0.29317536997695,"reference_mean_power":0.0014670926652510829,"output_mean_power":0.001371323785384276},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5556250379109994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020148633268320188,"output_power":0.00018389650201112192,"reference_active":true,"delta_db":-0.39672123825809724},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.931250874616967e-05,"output_power":1.671556433342288e-05,"reference_active":true,"delta_db":-0.627176500777003},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.933225179066709e-07,"output_power":5.128958745271445e-07,"reference_active":true,"delta_db":-1.3090609971900715}],"active_band_count":3,"active_band_mae_db":0.7776529120750574}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":1.7771209729249644,"rms_envelope_nrmse":0.17822623396396198,"envelope_correlation":0.9786737331352348,"rms_gain_error_db":-0.6717661919907916,"reference_mean_power":0.00020535819062638303,"output_mean_power":0.0001759281800430023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.949818894937679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.42703704863136e-05,"output_power":4.850799206850647e-05,"reference_active":true,"delta_db":-0.4874948849101551},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.92519989001781e-06,"output_power":6.788442182914734e-06,"reference_active":true,"delta_db":-1.1884782794017594},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.574401681634857e-06,"output_power":1.911252675138511e-06,"reference_active":true,"delta_db":-1.293582041489447}],"active_band_count":3,"active_band_mae_db":0.9898517352671204}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":1.0712086962193281,"rms_envelope_nrmse":0.12510398623094343,"envelope_correlation":0.9914187630027286,"rms_gain_error_db":-0.671964525475172,"reference_mean_power":0.0002063770763803802,"output_mean_power":0.00017679297468205105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.280870922430584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2679956690924704e-05,"output_power":4.697344070923756e-05,"reference_active":true,"delta_db":-0.49793036711010363},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1805452510023395e-06,"output_power":3.101282664461391e-06,"reference_active":true,"delta_db":-1.2969157713975088},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0193067246169139e-07,"output_power":1.2508188648645083e-07,"reference_active":true,"delta_db":-2.080078688947896}],"active_band_count":3,"active_band_mae_db":1.2916416091518361}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":2.016766953905585,"rms_envelope_nrmse":0.17030006419239996,"envelope_correlation":0.9830884832355702,"rms_gain_error_db":-0.4358707473733875,"reference_mean_power":0.00014718641611289373,"output_mean_power":0.0001331314480913156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4339203545791116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.0687614509069e-05,"output_power":6.401534702095553e-05,"reference_active":true,"delta_db":-0.43059221696140043},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.343067234178676e-06,"output_power":5.267953314109041e-06,"reference_active":true,"delta_db":-2.4884755662730957},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.359519184803567e-06,"output_power":2.140450378038515e-06,"reference_active":true,"delta_db":-0.42318348801222344}],"active_band_count":3,"active_band_mae_db":1.1140837570822397}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":1.4544414582310343,"rms_envelope_nrmse":0.10419812028618866,"envelope_correlation":0.9945294366195107,"rms_gain_error_db":-0.43587448378745497,"reference_mean_power":0.00014792010871620984,"output_mean_power":0.00013379496459493202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3826641859038568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.726675521598139e-05,"output_power":6.038695312749764e-05,"reference_active":true,"delta_db":-0.4685736140052826},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.270863847079585e-06,"output_power":2.2458537388545276e-06,"reference_active":true,"delta_db":-2.791342571314993},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6847652723937036e-07,"output_power":1.0790419016764677e-07,"reference_active":true,"delta_db":-1.935010922771414}],"active_band_count":3,"active_band_mae_db":1.7316423693638967}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.492929458618164,"predicted_mos":3.492929458618164,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e739cd78978ee05da0ace0fb814c27af223de6d9e85ba8115e6dded71329c874","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0847089039998536},"started_at":"2026-10-13T04:41:32.733321+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6235999939672183e-05,"native_input_samples":96000,"native_input_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","output_sha256":"72d2f6315f8bba1a6ff5e772863970fbd84d5ab19754f8f53d9b65fc9934991a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6334653750000143,"synthesis_s":0.08009114000014961,"total_s":1.713556515000164,"rtf":0.2855927525000273,"output_peak":0.6943533334235288,"output_rms":0.07271835425613583,"finished_at":"2026-10-13T04:41:34.453231+00:00","evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"72d2f6315f8bba1a6ff5e772863970fbd84d5ab19754f8f53d9b65fc9934991a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.077914674672632,"spectral_convergence":0.5635348164425449,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.423590707604511,"spectral_convergence":0.38356039643071954,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.629485640398524,"spectral_convergence":0.3841155464488178,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.001481524453615022,"delta_db":-0.21526326886951297,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.00310570459432469,"delta_db":2.002192174499723,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.00025549672631542016,"delta_db":0.8571101093632041,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00023402632803820283,"delta_db":0.931803740331405,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":0.0001239829577703614,"delta_db":0.5843376100573838,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":8.722398565717078e-05,"delta_db":0.4017364433475541,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8320738910781303,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1030817575691925,"pitch":null,"output_rms_dbfs":-22.767119170410083,"output_peak":0.6943533334235288,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.043663674225222},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.077914674672632,"spectral_convergence":0.5635348164425449,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.423590707604511,"spectral_convergence":0.38356039643071954,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.629485640398524,"spectral_convergence":0.3841155464488178,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.001481524453615022,"delta_db":-0.21526326886951297,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.00310570459432469,"delta_db":2.002192174499723,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.00025549672631542016,"delta_db":0.8571101093632041,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00023402632803820283,"delta_db":0.931803740331405,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":0.0001239829577703614,"delta_db":0.5843376100573838,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":8.722398565717078e-05,"delta_db":0.4017364433475541,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8320738910781303,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1030817575691925,"pitch":null,"output_rms_dbfs":-22.767119170410083,"output_peak":0.6943533334235288,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.043663674225222},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8742953988941995,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06404569659064947,"rms_dbfs":-23.870200927979273,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.867078192515194,"mean":-2.6813824971516927e-05,"zero_energy":false},"output":{"rms":0.07271835425613582,"rms_dbfs":-22.767119170410076,"peak":0.6943533334235288,"peak_dbfs":-3.168389504260783,"crest_factor_db":19.598729666149296,"mean":2.2774670822642285e-07,"zero_energy":false},"rms_difference_db":1.1030817575691962},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":1.9309726734370085e-06,"output_rms_dbfs":-57.14223872186162,"output_zero_energy":false,"output_peak":0.01589656320315907,"output_energy_fraction":6.329510632902177e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":220,"both_estimated_voiced_frames":194,"reference_voiced_output_unvoiced_frames":46,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8083333333333333,"absolute_difference_cents_median":5.5299873149036465,"absolute_difference_cents_p95":1172.711714556395,"difference_over_600_cents_frames":12,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023872023988588299,"log_rms_mae_db":2.180625364238255,"rms_envelope_nrmse":0.24272721918565732,"envelope_correlation":0.9649087852121384,"rms_gain_error_db":1.1034570117734828,"reference_mean_power":0.004104010412713017,"output_mean_power":0.005291199727172878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.972155761177528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003715407512890731,"output_power":0.00046816765410383297,"reference_active":true,"delta_db":1.0039494986688722},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.499306084417706e-05,"output_power":6.997967008478012e-05,"reference_active":true,"delta_db":1.0466399805889846},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6530449727349195e-05,"output_power":3.253595697109297e-05,"reference_active":true,"delta_db":2.9407891589495745}],"active_band_count":3,"active_band_mae_db":1.663792879402477}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00021497588341920653,"log_rms_mae_db":1.4790639109254953,"rms_envelope_nrmse":0.20626399962888015,"envelope_correlation":0.979652961729041,"rms_gain_error_db":1.1034006136414949,"reference_mean_power":0.00411739985369924,"output_mean_power":0.005308393468213887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.973189563381596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033993681960372955,"output_power":0.0004314734312411194,"reference_active":true,"delta_db":1.035558515009101},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413396717591137e-05,"output_power":3.214505694529058e-05,"reference_active":true,"delta_db":1.2448548173246523},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.827774931478278e-07,"output_power":1.358140451804635e-06,"reference_active":true,"delta_db":1.8709343266150913}],"active_band_count":3,"active_band_mae_db":1.3837825529829482}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001937632671887944,"log_rms_mae_db":2.6695938349605006,"rms_envelope_nrmse":0.32666948945062213,"envelope_correlation":0.9719358124967021,"rms_gain_error_db":1.9016631412495313,"reference_mean_power":0.0014622039835396937,"output_mean_power":0.0022655532629586466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.89031570773124e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020584065582039213,"output_power":0.00025997397616357596,"reference_active":true,"delta_db":1.0139871963937221},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2632804837837524e-05,"output_power":5.7785647268402386e-05,"reference_active":true,"delta_db":1.3207607617634245},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.198246611681044e-06,"output_power":1.3571623878045856e-05,"reference_active":true,"delta_db":2.18910836966357}],"active_band_count":3,"active_band_mae_db":1.5079521092735721}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001823230477151681,"log_rms_mae_db":2.0342682693450933,"rms_envelope_nrmse":0.29034763673328967,"envelope_correlation":0.9845520481878265,"rms_gain_error_db":1.892523703200365,"reference_mean_power":0.0014670926652510829,"output_mean_power":0.0022683492183918673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5556250379109994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020148633268320188,"output_power":0.0002546774254515778,"reference_active":true,"delta_db":1.017448587003734},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.931250874616967e-05,"output_power":2.7381394342432468e-05,"reference_active":true,"delta_db":1.5161686648358874},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.933225179066709e-07,"output_power":7.693761494074304e-07,"reference_active":true,"delta_db":0.4520341353347778}],"active_band_count":3,"active_band_mae_db":0.9952171290581332}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":1.988333718453956,"rms_envelope_nrmse":0.2223390750365749,"envelope_correlation":0.9737648659492721,"rms_gain_error_db":0.8704620112164475,"reference_mean_power":0.00020535819062638303,"output_mean_power":0.0002509332608851828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.949818894937679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.42703704863136e-05,"output_power":6.376449267322041e-05,"reference_active":true,"delta_db":0.7001612229950017},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.92519989001781e-06,"output_power":9.835435787671479e-06,"reference_active":true,"delta_db":0.42175655952723407},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.574401681634857e-06,"output_power":1.6300140281189136e-06,"reference_active":true,"delta_db":-1.984849684257361}],"active_band_count":3,"active_band_mae_db":1.0355891555931989}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":1.072448397900041,"rms_envelope_nrmse":0.14920203877186805,"envelope_correlation":0.9918276218008285,"rms_gain_error_db":0.8690052624260186,"reference_mean_power":0.0002063770763803802,"output_mean_power":0.0002520936939225328},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.280870922430584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2679956690924704e-05,"output_power":6.148427640081135e-05,"reference_active":true,"delta_db":0.6711865700679085},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1805452510023395e-06,"output_power":4.64365356181724e-06,"reference_active":true,"delta_db":0.4562688320509565},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0193067246169139e-07,"output_power":1.199956111806504e-07,"reference_active":true,"delta_db":-2.2603692945917424}],"active_band_count":3,"active_band_mae_db":1.129274898903536}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":2.0231619792944406,"rms_envelope_nrmse":0.20378092797542915,"envelope_correlation":0.9787862149787966,"rms_gain_error_db":0.5412103096374471,"reference_mean_power":0.00014718641611289373,"output_mean_power":0.00016672040692139892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4339203545791116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.0687614509069e-05,"output_power":7.61470811345152e-05,"reference_active":true,"delta_db":0.3230993484303521},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.343067234178676e-06,"output_power":1.4813818915696312e-05,"reference_active":true,"delta_db":2.001775574167913},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.359519184803567e-06,"output_power":3.3126313746538835e-06,"reference_active":true,"delta_db":1.4734959798147782}],"active_band_count":3,"active_band_mae_db":1.2661236341376811}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":1.1040487690275205,"rms_envelope_nrmse":0.13938711725335867,"envelope_correlation":0.9910728617979225,"rms_gain_error_db":0.5411784195070803,"reference_mean_power":0.00014792010871620984,"output_mean_power":0.00016755024194275208},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3826641859038568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.726675521598139e-05,"output_power":7.398738642452764e-05,"reference_active":true,"delta_db":0.41357207427298376},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.270863847079585e-06,"output_power":7.4141801428728305e-06,"reference_active":true,"delta_db":2.395474073412027},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6847652723937036e-07,"output_power":1.7948337367903576e-07,"reference_active":true,"delta_db":0.2748482226174071}],"active_band_count":3,"active_band_mae_db":1.0279647901008058}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9167726039886475,"predicted_mos":3.9167726039886475,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f02574eb3361872e2ecece7cfa39059c22cc52f53c4e6c9836612f39bd11c804","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.303827296000236},"started_at":"2026-10-13T04:48:29.901422+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4977999601105694e-05,"native_input_samples":96000,"native_input_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","output_sha256":"3c60b644a222e0518713eec44437b2e54f91e8854b56a68aba2bee498a4079bb","output_samples":96000,"length_delta_samples":0,"analysis_s":1.964438353999867,"synthesis_s":6.270462522000344,"total_s":8.23490087600021,"rtf":1.3724834793333685,"output_peak":0.4348481595516205,"output_rms":0.062323044757636545,"finished_at":"2026-10-13T04:48:38.145078+00:00","evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"3c60b644a222e0518713eec44437b2e54f91e8854b56a68aba2bee498a4079bb","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.507231324859679,"spectral_convergence":0.24139815338657575,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.47971803871422,"spectral_convergence":0.24905427506847325,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.627560953998772,"spectral_convergence":0.2672273441870684,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.0016238075582174208,"delta_db":0.18299407451083258,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.0017113012570459435,"delta_db":-0.5861445770343552,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.00021740414840206362,"delta_db":0.15593497882776336,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00016162889073762727,"delta_db":-0.6756535202652287,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":0.0001073630109767101,"delta_db":-0.04073549070009641,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":6.26570424826027e-05,"delta_db":-1.0349239172684768,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.44606442643445887,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23682582406461072,"pitch":null,"output_rms_dbfs":-24.10702675204388,"output_peak":0.4348481595516205,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.538170105857557},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.507231324859679,"spectral_convergence":0.24139815338657575,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.47971803871422,"spectral_convergence":0.24905427506847325,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.627560953998772,"spectral_convergence":0.2672273441870684,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.0016238075582174208,"delta_db":0.18299407451083258,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.0017113012570459435,"delta_db":-0.5861445770343552,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.00021740414840206362,"delta_db":0.15593497882776336,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00016162889073762727,"delta_db":-0.6756535202652287,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":0.0001073630109767101,"delta_db":-0.04073549070009641,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":6.26570424826027e-05,"delta_db":-1.0349239172684768,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.44606442643445887,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23682582406461072,"pitch":null,"output_rms_dbfs":-24.10702675204388,"output_peak":0.4348481595516205,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.538170105857557},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8963374839297706,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06404569659064947,"rms_dbfs":-23.870200927979273,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.867078192515194,"mean":-2.6813824971516927e-05,"zero_energy":false},"output":{"rms":0.062323044757636545,"rms_dbfs":-24.10702675204388,"peak":0.4348481595516205,"peak_dbfs":-7.2332472738085904,"crest_factor_db":16.873779478235292,"mean":-0.0003031270075772833,"zero_energy":false},"rms_difference_db":-0.2368258240646064},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":1.7362974965416551e-06,"output_rms_dbfs":-57.60375860935356,"output_zero_energy":false,"output_peak":0.02077590487897396,"output_energy_fraction":7.748344169811345e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":238,"both_estimated_voiced_frames":226,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9416666666666667,"absolute_difference_cents_median":3.751543315953189,"absolute_difference_cents_p95":14.657848489516624,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023872023988588299,"log_rms_mae_db":1.1188989211053961,"rms_envelope_nrmse":0.12136152577168735,"envelope_correlation":0.9823375383135547,"rms_gain_error_db":-0.23634273555576388,"reference_mean_power":0.004104010412713017,"output_mean_power":0.00388663878711967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.972155761177528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003715407512890731,"output_power":0.00039768368653406876,"reference_active":true,"delta_db":0.2953132112084428},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.499306084417706e-05,"output_power":6.272745306375681e-05,"reference_active":true,"delta_db":0.5714976150733475},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6530449727349195e-05,"output_power":1.9971572377960833e-05,"reference_active":true,"delta_db":0.8212758942461367}],"active_band_count":3,"active_band_mae_db":0.562695573509309}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00021497588341920653,"log_rms_mae_db":0.8734131343063286,"rms_envelope_nrmse":0.1110047755603996,"envelope_correlation":0.9843790731228835,"rms_gain_error_db":-0.23843225329302836,"reference_mean_power":0.00411739985369924,"output_mean_power":0.0038974434223472613},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.973189563381596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033993681960372955,"output_power":0.0003685348406964059,"reference_active":true,"delta_db":0.35080344740562713},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413396717591137e-05,"output_power":2.845555424865524e-05,"reference_active":true,"delta_db":0.7153833136017173},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.827774931478278e-07,"output_power":1.04534399982408e-06,"reference_active":true,"delta_db":0.7340797869689469}],"active_band_count":3,"active_band_mae_db":0.6000888493254304}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001937632671887944,"log_rms_mae_db":1.3429809140348392,"rms_envelope_nrmse":0.15273284124634584,"envelope_correlation":0.9839793421571278,"rms_gain_error_db":-0.501560777305198,"reference_mean_power":0.0014622039835396937,"output_mean_power":0.0013027224126164988},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.89031570773124e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020584065582039213,"output_power":0.00021854913338165228,"reference_active":true,"delta_db":0.2601793181814427},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2632804837837524e-05,"output_power":4.589170407256299e-05,"reference_active":true,"delta_db":0.3199027834125309},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.198246611681044e-06,"output_power":1.0301366323870347e-05,"reference_active":true,"delta_db":0.9917385306355851}],"active_band_count":3,"active_band_mae_db":0.5239402107431862}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001823230477151681,"log_rms_mae_db":1.0397715876465614,"rms_envelope_nrmse":0.14166310891457765,"envelope_correlation":0.9860976116013117,"rms_gain_error_db":-0.5034629661068204,"reference_mean_power":0.0014670926652510829,"output_mean_power":0.0013065055207857246},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5556250379109994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020148633268320188,"output_power":0.00021564733295671959,"reference_active":true,"delta_db":0.2949849912178302},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.931250874616967e-05,"output_power":2.125873349860471e-05,"reference_active":true,"delta_db":0.4169869416150248},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.933225179066709e-07,"output_power":8.273559835504277e-07,"reference_active":true,"delta_db":0.7675710656885759}],"active_band_count":3,"active_band_mae_db":0.4931809995071436}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":1.9931618408716394,"rms_envelope_nrmse":0.1806787468724338,"envelope_correlation":0.9784128036338477,"rms_gain_error_db":-0.7336800163376712,"reference_mean_power":0.00020535819062638303,"output_mean_power":0.00017343790844204636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.949818894937679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.42703704863136e-05,"output_power":4.719783425766035e-05,"reference_active":true,"delta_db":-0.6064071553043124},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.92519989001781e-06,"output_power":7.378101662013589e-06,"reference_active":true,"delta_db":-0.8267331638706582},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.574401681634857e-06,"output_power":3.0722410031940804e-06,"reference_active":true,"delta_db":0.7677897066071172}],"active_band_count":3,"active_band_mae_db":0.7336433419273627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":1.5998103838583961,"rms_envelope_nrmse":0.1550511233488132,"envelope_correlation":0.9845209368105127,"rms_gain_error_db":-0.7344661372778628,"reference_mean_power":0.0002063770763803802,"output_mean_power":0.00017426687446381802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.280870922430584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2679956690924704e-05,"output_power":4.596374292974462e-05,"reference_active":true,"delta_db":-0.5923002222793343},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1805452510023395e-06,"output_power":3.448555830738878e-06,"reference_active":true,"delta_db":-0.835956672432412},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0193067246169139e-07,"output_power":2.159905684763356e-07,"reference_active":true,"delta_db":0.29232495952812426}],"active_band_count":3,"active_band_mae_db":0.5735272847466235}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":2.250747381847973,"rms_envelope_nrmse":0.22383749236834335,"envelope_correlation":0.9699839079199536,"rms_gain_error_db":-0.5062468056106938,"reference_mean_power":0.00014718641611289373,"output_mean_power":0.00013099148032474716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4339203545791116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.0687614509069e-05,"output_power":7.243025255750325e-05,"reference_active":true,"delta_db":0.1057667375025485},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.343067234178676e-06,"output_power":1.0864302181502798e-05,"reference_active":true,"delta_db":0.6551236253300013},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.359519184803567e-06,"output_power":3.956210781190005e-06,"reference_active":true,"delta_db":2.244559092640988}],"active_band_count":3,"active_band_mae_db":1.001816485157846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":1.8621979996820863,"rms_envelope_nrmse":0.19079767154565774,"envelope_correlation":0.9780847014684878,"rms_gain_error_db":-0.5062531957907566,"reference_mean_power":0.00014792010871620984,"output_mean_power":0.0001316442509584076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3826641859038568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.726675521598139e-05,"output_power":7.035346255839748e-05,"reference_active":true,"delta_db":0.19484997912914853},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.270863847079585e-06,"output_power":5.466744598935512e-06,"reference_active":true,"delta_db":1.0721305797877034},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6847652723937036e-07,"output_power":3.243851448077882e-07,"reference_active":true,"delta_db":2.845215555726717}],"active_band_count":3,"active_band_mae_db":1.3707320382145232}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8421528339385986,"predicted_mos":3.8421528339385986,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"203c574ef14058c775c3c35a6318dcd5242dc9e647d37cd0d6fd59438b6e474b","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7319747309993545},"started_at":"2026-10-13T07:06:28.699473+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006191671998749371,"native_input_samples":264600,"native_input_sha256":"b84948c9ebf047f8f90fff6fbe9cd73fa6ca266d9c732023a87b49db431681f3","output_sha256":"39264df6809651e972e210b038b097ce132661029b7fda90050bd4fa0ddd7e09","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15872559000126785,"synthesis_s":5.561009400998955,"total_s":5.719734991000223,"rtf":0.9532891651667038,"output_peak":0.42374715209007263,"output_rms":0.055140119874006445,"finished_at":"2026-10-13T07:06:34.436382+00:00","evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"39264df6809651e972e210b038b097ce132661029b7fda90050bd4fa0ddd7e09","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.166997799503626,"output_peak":0.42275765543627264},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.254440002301646,"spectral_convergence":0.5632080046697171,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.744933797941968,"spectral_convergence":0.3986947376217509,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.409505307037584,"spectral_convergence":0.46945243357946437,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.0008190490837211798,"delta_db":-2.7892522229532686,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.0018025581480753863,"delta_db":-0.3605164369983139,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.00019883363429673437,"delta_db":-0.23184478291750366,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":9.437990031566359e-05,"delta_db":-3.0120483009227303,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":7.847278521793652e-05,"delta_db":-1.4020916457307901,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":5.031486747827394e-05,"delta_db":-1.9876595016745489,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.630568815199526,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.295911304442571,"pitch":null,"output_rms_dbfs":-25.166112232421845,"output_peak":0.42275765543627264,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.802959702427065},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7354906810110091,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06404569659064947,"rms_dbfs":-23.870200927979273,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.867078192515194,"mean":-2.6813824971516927e-05,"zero_energy":false},"output":{"rms":0.05516890808331424,"rms_dbfs":-25.166112232421845,"peak":0.42275765543627264,"peak_dbfs":-7.478170386201003,"crest_factor_db":17.68794184622084,"mean":-0.00014624152498646088,"zero_energy":false},"rms_difference_db":-1.2959113044425727},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":1.0018833415105467e-06,"output_rms_dbfs":-59.991828444246295,"output_zero_energy":false,"output_peak":0.013047231032384211,"output_energy_fraction":5.705720161144042e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":155,"both_estimated_voiced_frames":137,"reference_voiced_output_unvoiced_frames":103,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.5708333333333333,"absolute_difference_cents_median":14.607953206993212,"absolute_difference_cents_p95":119.13637976821026,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023872023988588299,"log_rms_mae_db":2.302428358697503,"rms_envelope_nrmse":0.19512582156795472,"envelope_correlation":0.9767928080614952,"rms_gain_error_db":-1.2964766225007485,"reference_mean_power":0.004104010412713017,"output_mean_power":0.0030448141677240238},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.972155761177528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003715407512890731,"output_power":0.000315309005400372,"reference_active":true,"delta_db":-0.7127008032875761},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.499306084417706e-05,"output_power":5.314694080863405e-05,"reference_active":true,"delta_db":-0.14829621484318165},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6530449727349195e-05,"output_power":1.1226422694470606e-05,"reference_active":true,"delta_db":-1.680432790069701}],"active_band_count":3,"active_band_mae_db":0.8471432694001528}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00021497588341920653,"log_rms_mae_db":2.075693993685015,"rms_envelope_nrmse":0.16768760464933685,"envelope_correlation":0.9891761109939936,"rms_gain_error_db":-1.293536340276991,"reference_mean_power":0.00411739985369924,"output_mean_power":0.003056816794329744},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.973189563381596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033993681960372955,"output_power":0.00029144024550215107,"reference_active":true,"delta_db":-0.6684868283036481},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413396717591137e-05,"output_power":2.3219576037315946e-05,"reference_active":true,"delta_db":-0.16774431964262626},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.827774931478278e-07,"output_power":6.284739151535215e-07,"reference_active":true,"delta_db":-1.4756399509823739}],"active_band_count":3,"active_band_mae_db":0.7706236996428828}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001937632671887944,"log_rms_mae_db":2.707426081853073,"rms_envelope_nrmse":0.19192118631317096,"envelope_correlation":0.972038261457684,"rms_gain_error_db":-0.39311929579235017,"reference_mean_power":0.0014622039835396937,"output_mean_power":0.0013356603479206407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.89031570773124e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020584065582039213,"output_power":0.00020107155653449388,"reference_active":true,"delta_db":-0.10180516995507455},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2632804837837524e-05,"output_power":3.6251951996406886e-05,"reference_active":true,"delta_db":-0.7041250983848732},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.198246611681044e-06,"output_power":5.614542786402269e-06,"reference_active":true,"delta_db":-1.6440658241712345}],"active_band_count":3,"active_band_mae_db":0.8166653641703941}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001823230477151681,"log_rms_mae_db":2.1336916871445317,"rms_envelope_nrmse":0.14724596302798085,"envelope_correlation":0.9832522904149619,"rms_gain_error_db":-0.38706746795381447,"reference_mean_power":0.0014670926652510829,"output_mean_power":0.0013419946950094463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5556250379109994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020148633268320188,"output_power":0.0001938253775258478,"reference_active":true,"delta_db":-0.16834953661292762},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.931250874616967e-05,"output_power":1.6496901653983975e-05,"reference_active":true,"delta_db":-0.6843630808082285},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.933225179066709e-07,"output_power":4.033116616585986e-07,"reference_active":true,"delta_db":-2.3529452584699144}],"active_band_count":3,"active_band_mae_db":1.0685526252970234}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":3.7349306386526804,"rms_envelope_nrmse":0.39369663911153474,"envelope_correlation":0.908747888376035,"rms_gain_error_db":-2.557235548366004,"reference_mean_power":0.00020535819062638303,"output_mean_power":0.00011396945575146638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.949818894937679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.42703704863136e-05,"output_power":4.0096781589548226e-05,"reference_active":true,"delta_db":-1.314532714633311},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.92519989001781e-06,"output_power":6.040800618489931e-06,"reference_active":true,"delta_db":-1.695234496873617},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.574401681634857e-06,"output_power":1.7258622369808664e-06,"reference_active":true,"delta_db":-1.7367018424660077}],"active_band_count":3,"active_band_mae_db":1.582156351324312}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":3.2589245855513003,"rms_envelope_nrmse":0.3606386587652826,"envelope_correlation":0.9288206316313313,"rms_gain_error_db":-2.557360145504359,"reference_mean_power":0.0002063770763803802,"output_mean_power":0.00011453162990356104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.280870922430584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2679956690924704e-05,"output_power":3.908136080032641e-05,"reference_active":true,"delta_db":-1.2967573194589586},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1805452510023395e-06,"output_power":2.8396812019151047e-06,"reference_active":true,"delta_db":-1.6796334216030109},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0193067246169139e-07,"output_power":1.2905233942760422e-07,"reference_active":true,"delta_db":-1.9443640981958021}],"active_band_count":3,"active_band_mae_db":1.6402516130859237}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":3.433831873640305,"rms_envelope_nrmse":0.25866677460748744,"envelope_correlation":0.9833868134575443,"rms_gain_error_db":-2.0791417898080327,"reference_mean_power":0.00014718641611289373,"output_mean_power":9.119133039027365e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4339203545791116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.0687614509069e-05,"output_power":4.264867451142898e-05,"reference_active":true,"delta_db":-2.194377876282465},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.343067234178676e-06,"output_power":5.811889509082187e-06,"reference_active":true,"delta_db":-2.061721248592793},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.359519184803567e-06,"output_power":9.652191767278867e-07,"reference_active":true,"delta_db":-3.881975710852738}],"active_band_count":3,"active_band_mae_db":2.7126916119093316}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":3.042622447635167,"rms_envelope_nrmse":0.2340695809146194,"envelope_correlation":0.9925147509976874,"rms_gain_error_db":-2.079016473628036,"reference_mean_power":0.00014792010871620984,"output_mean_power":9.164854403516262e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3826641859038568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.726675521598139e-05,"output_power":4.036612129065505e-05,"reference_active":true,"delta_db":-2.2178345805781916},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.270863847079585e-06,"output_power":2.5082438180365888e-06,"reference_active":true,"delta_db":-2.3114597603272364},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6847652723937036e-07,"output_power":7.041139557285807e-08,"reference_active":true,"delta_db":-3.7889644970279264}],"active_band_count":3,"active_band_mae_db":2.772752945977785}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"golf","status":"measured","mos":{"status":"measured","value":3.3147966861724854,"predicted_mos":3.3147966861724854,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"e57c230c876b8d8420ebd53a518511aaa7ccf63e49c96c18e5126dbeae629d1b","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.31601508499989},"started_at":"2026-10-13T05:02:12.513482+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003125857000668475,"native_input_samples":144000,"native_input_sha256":"1be42d1b2de3fa9211e421859f28acc21779929beaa6d50a3bf6c890c00ec6a9","output_sha256":"03c22f41a15e33dcd90f0826927d37aa9aec62ec21b0de5c2f7dd52c1ad6ce62","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6359780359998695,"synthesis_s":0.11388811399956467,"total_s":0.7498661499994341,"rtf":0.12497769166657235,"output_peak":0.49658966064453125,"output_rms":0.059866133709824305,"finished_at":"2026-10-13T05:02:13.277367+00:00","analysis_stages":{"external_f0_s":0.0883832990002702},"evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"03c22f41a15e33dcd90f0826927d37aa9aec62ec21b0de5c2f7dd52c1ad6ce62","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.454040507345468,"output_peak":0.496953879222766},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.075492875933872,"spectral_convergence":0.624435663137454,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.246821578952848,"spectral_convergence":0.29372055402336766,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.365542710669591,"spectral_convergence":0.2780182791854667,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001556863619677442,"output_power":0.0014347465987383189,"delta_db":-0.3547536643113537,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019621881259362125,"output_power":0.0017363043746792968,"delta_db":-0.5311478381042898,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020798738704851468,"output_power":0.0001556437145070703,"delta_db":-1.259054120106976,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001891986342603237,"output_power":0.0001281071324290609,"delta_db":-1.6934468711931263,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001085676305561318,"output_power":7.140448014212622e-05,"delta_db":-1.8197489799731308,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.964844968928286e-05,"output_power":5.967532842467536e-05,"delta_db":-1.2538250909408517,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1519960941049547,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5865942730784012,"pitch":null,"output_rms_dbfs":-24.454040507345468,"output_peak":0.496953879222766,"reference_rms_dbfs":-23.867446234267064,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8959523885187695},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8749263302947695,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06406601163774679,"rms_dbfs":-23.867446234267064,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.864323498802985,"mean":1.6102607739787667e-06,"zero_energy":false},"output":{"rms":0.05988223132883868,"rms_dbfs":-24.454040507345468,"peak":0.496953879222766,"peak_dbfs":-6.073678298914311,"crest_factor_db":18.380362208431155,"mean":0.0010158553299636761,"zero_energy":false},"rms_difference_db":-0.5865942730784042},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":1.8652678763441653e-06,"output_rms_dbfs":-57.292587891293174,"output_zero_energy":false,"output_peak":0.009577159082715217,"output_energy_fraction":9.031331394487421e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":273,"both_estimated_voiced_frames":233,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":40,"paired_voiced_fraction":0.9708333333333333,"absolute_difference_cents_median":5.787205576259459,"absolute_difference_cents_p95":28.404287923378476,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00023872023988588299,"log_rms_mae_db":1.7089088704865425,"rms_envelope_nrmse":0.15015038641572148,"envelope_correlation":0.9761977598201889,"rms_gain_error_db":-0.5851962773465591,"reference_mean_power":0.0041058943957472,"output_mean_power":0.003588295060919},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.961172180411551e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003728954652448579,"output_power":0.0002860550336645328,"reference_active":true,"delta_db":-1.1513750758997134},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.466148509579836e-05,"output_power":5.5046202314966966e-05,"reference_active":true,"delta_db":0.03045935142840015},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6492454155270608e-05,"output_power":1.7760725977545247e-05,"reference_active":true,"delta_db":0.32175428256253535}],"active_band_count":3,"active_band_mae_db":0.5011962366302164}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00021497588341920653,"log_rms_mae_db":1.040323300156469,"rms_envelope_nrmse":0.09872127300773444,"envelope_correlation":0.9922277619733105,"rms_gain_error_db":-0.5819085469134004,"reference_mean_power":0.004119936662020282,"output_mean_power":0.0036032938857093228},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.96076271517958e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003412935249288939,"output_power":0.0002628423808250445,"reference_active":true,"delta_db":-1.1343265648415533},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.363218799701937e-05,"output_power":2.4190392027047425e-05,"reference_active":true,"delta_db":0.10138973702687072},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.821485834407272e-07,"output_power":9.139645204463982e-07,"reference_active":true,"delta_db":0.1538759594447543}],"active_band_count":3,"active_band_mae_db":0.4631974204377261}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001937632671887944,"log_rms_mae_db":1.9668054363002074,"rms_envelope_nrmse":0.17679558909014692,"envelope_correlation":0.9804334781426708,"rms_gain_error_db":-0.7981955581148295,"reference_mean_power":0.001462594834053801,"output_mean_power":0.0012170389549347898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.8939494240458495e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020693497568187466,"output_power":0.00015073289236530992,"reference_active":true,"delta_db":-1.3762586757535098},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.240382754957755e-05,"output_power":3.697785804332321e-05,"reference_active":true,"delta_db":-0.5946330875009367},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.151327094521421e-06,"output_power":6.1614461086646455e-06,"reference_active":true,"delta_db":-1.2154566649360767}],"active_band_count":3,"active_band_mae_db":1.0621161427301744}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001823230477151681,"log_rms_mae_db":1.445388771737368,"rms_envelope_nrmse":0.13339637360823364,"envelope_correlation":0.9911889911904243,"rms_gain_error_db":-0.7972210856379071,"reference_mean_power":0.0014682377207830331,"output_mean_power":0.0012220086184421455},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.560112502136973e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020267550654064917,"output_power":0.00014985780649761791,"reference_active":true,"delta_db":-1.3112189566678756},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8872485114212806e-05,"output_power":1.619813641983367e-05,"reference_active":true,"delta_db":-0.6636403924109113},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.939600476703035e-07,"output_power":4.5399602620826343e-07,"reference_active":true,"delta_db":-1.8428241665619631}],"active_band_count":3,"active_band_mae_db":1.27256117188025}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":2.4155797962721057,"rms_envelope_nrmse":0.2619373832469292,"envelope_correlation":0.9651113004868197,"rms_gain_error_db":-1.6695465705239947,"reference_mean_power":0.00020569271982969603,"output_mean_power":0.0001400439216281897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.951299890224689e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.433279959345761e-05,"output_power":4.070627252744591e-05,"reference_active":true,"delta_db":-1.2540074753733175},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.876008819411584e-06,"output_power":6.33067144655506e-06,"reference_active":true,"delta_db":-1.4676795028623013},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.567497277063412e-06,"output_power":2.525483913548389e-06,"reference_active":true,"delta_db":-0.07165385129562636}],"active_band_count":3,"active_band_mae_db":0.9311136098437484}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":1.9446350550778049,"rms_envelope_nrmse":0.21728287887716577,"envelope_correlation":0.9835618396478488,"rms_gain_error_db":-1.6697470476959415,"reference_mean_power":0.00020671479905234769,"output_mean_power":0.0001407332978740379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.282630047664373e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2726759434451004e-05,"output_power":3.971970815347985e-05,"reference_active":true,"delta_db":-1.2302503179533157},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130328768312416e-06,"output_power":2.907088422909914e-06,"reference_active":true,"delta_db":-1.5252638064915593},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.01929818000009e-07,"output_power":1.448765507075695e-07,"reference_active":true,"delta_db":-1.4420235631388356}],"active_band_count":3,"active_band_mae_db":1.3991792291945702}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":2.771806623663376,"rms_envelope_nrmse":0.2483407823400951,"envelope_correlation":0.9765237326244823,"rms_gain_error_db":-1.631861696918985,"reference_mean_power":0.00014742606889726076,"output_mean_power":0.0001012483875667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4384414535548702e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.085192163953421e-05,"output_power":5.2683988137121434e-05,"reference_active":true,"delta_db":-1.2867299034937796},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.350694121813545e-06,"output_power":6.7419464417316594e-06,"reference_active":true,"delta_db":-1.4205855252017783},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.368103149482884e-06,"output_power":2.373547290520664e-06,"reference_active":true,"delta_db":0.009972736203446745}],"active_band_count":3,"active_band_mae_db":0.9057627216330015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":2.381201631360344,"rms_envelope_nrmse":0.2049067451312168,"envelope_correlation":0.9905596236653968,"rms_gain_error_db":-1.6319978377005717,"reference_mean_power":0.00014816554583489536,"output_mean_power":0.0001017530513223983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.3869850901456495e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.741531996659909e-05,"output_power":5.0270655849188195e-05,"reference_active":true,"delta_db":-1.2744404888158614},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.264721623194733e-06,"output_power":3.0646218244331953e-06,"reference_active":true,"delta_db":-1.4351379810467497},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6808910928498486e-07,"output_power":1.6792050626183027e-07,"reference_active":true,"delta_db":-0.004358409125557969}],"active_band_count":3,"active_band_mae_db":0.9046456263293897}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.0871951580047607,"predicted_mos":3.0871951580047607,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"283cdb8ad1d4a5635ba593c963f7fa19d5a986922fce7409d2d728cc815841e9","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.2047426980007},"started_at":"2026-10-13T05:07:18.787494+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002667082999323611,"native_input_samples":144000,"native_input_sha256":"1be42d1b2de3fa9211e421859f28acc21779929beaa6d50a3bf6c890c00ec6a9","output_sha256":"1c50dcd0f18af93e2a132fa98ec2786f424553cad27f9776e7f693b66309551e","output_samples":144120,"length_delta_samples":120,"analysis_s":2.4820407590004834,"synthesis_s":3.1560143059996335,"total_s":5.638055065000117,"rtf":0.9396758441666861,"output_peak":0.4446036219596863,"output_rms":0.06819932745361769,"finished_at":"2026-10-13T05:07:24.436044+00:00","analysis_stages":{"external_f0_s":1.649537883999983},"evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"1c50dcd0f18af93e2a132fa98ec2786f424553cad27f9776e7f693b66309551e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.322225943644494,"output_peak":0.4404569780124172},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.28023406275578,"spectral_convergence":0.7221774017915383,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.375330977765995,"spectral_convergence":0.3967832725475173,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.293709967207653,"spectral_convergence":0.3628660064207137,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.0023348692553647925,"delta_db":1.760274154865047,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.0018157907613805852,"delta_db":-0.32875124959591,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.0001763273728201754,"delta_db":-0.7535459213207827,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00016802476567578562,"delta_db":-0.5071104598207039,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":9.33432774799755e-05,"delta_db":-0.648451859628484,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":6.717728906600822e-05,"delta_db":-0.732398093879862,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7884219565184649,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5498947626678873,"pitch":null,"output_rms_dbfs":-23.320306165311386,"output_peak":0.4404569780124172,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9830916692431435},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8233060186611616,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06404569659064947,"rms_dbfs":-23.870200927979273,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.867078192515194,"mean":-2.6813824971516927e-05,"zero_energy":false},"output":{"rms":0.068231464309271,"rms_dbfs":-23.320306165311386,"peak":0.4404569780124172,"peak_dbfs":-7.121930104824168,"crest_factor_db":16.19837606048722,"mean":0.00018665906590580343,"zero_energy":false},"rms_difference_db":0.5498947626678863},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":9.969415208579005e-06,"output_rms_dbfs":-50.01330316024941,"output_zero_energy":false,"output_peak":0.09335307835334118,"output_energy_fraction":0.00037117813851893883,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":235,"both_estimated_voiced_frames":193,"reference_voiced_output_unvoiced_frames":47,"reference_unvoiced_output_voiced_frames":42,"paired_voiced_fraction":0.8041666666666667,"absolute_difference_cents_median":6.347960922727386,"absolute_difference_cents_p95":29.692183473740716,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023872023988588299,"log_rms_mae_db":2.9358892448528846,"rms_envelope_nrmse":0.213612276177426,"envelope_correlation":0.9522086833323373,"rms_gain_error_db":0.5496894744665881,"reference_mean_power":0.004104010412713017,"output_mean_power":0.004657763158748969},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.972155761177528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003715407512890731,"output_power":0.0004675066579024422,"reference_active":true,"delta_db":0.9978134522498652},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.499306084417706e-05,"output_power":5.934754919483518e-05,"reference_active":true,"delta_db":0.3309489648224146},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6530449727349195e-05,"output_power":1.8780194898170285e-05,"reference_active":true,"delta_db":0.5541542585323703}],"active_band_count":3,"active_band_mae_db":0.6276388918682168}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00021497588341920653,"log_rms_mae_db":2.4340212219356263,"rms_envelope_nrmse":0.16607479295593325,"envelope_correlation":0.9709866538204562,"rms_gain_error_db":0.5473100176955003,"reference_mean_power":0.00411739985369924,"output_mean_power":0.004670399665329596},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.973189563381596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033993681960372955,"output_power":0.0004261203220438938,"reference_active":true,"delta_db":0.9813403966749356},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413396717591137e-05,"output_power":2.525293548242466e-05,"reference_active":true,"delta_db":0.19683151453460795},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.827774931478278e-07,"output_power":1.002716285088259e-06,"reference_active":true,"delta_db":0.5532681628720333}],"active_band_count":3,"active_band_mae_db":0.5771466913605255}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001937632671887944,"log_rms_mae_db":2.5505388768607484,"rms_envelope_nrmse":0.18281362951838717,"envelope_correlation":0.9755304774574658,"rms_gain_error_db":-0.5068629528230869,"reference_mean_power":0.0014622039835396937,"output_mean_power":0.00130113292703765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.89031570773124e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020584065582039213,"output_power":0.00017941092989815548,"reference_active":true,"delta_db":-0.5968225971216264},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2632804837837524e-05,"output_power":3.391604074784794e-05,"reference_active":true,"delta_db":-0.9933875771306975},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.198246611681044e-06,"output_power":6.000137775634959e-06,"reference_active":true,"delta_db":-1.3555975539904268}],"active_band_count":3,"active_band_mae_db":0.9819359094142502}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001823230477151681,"log_rms_mae_db":1.8101923970368416,"rms_envelope_nrmse":0.1171554661545314,"envelope_correlation":0.9910711648225176,"rms_gain_error_db":-0.5048403747069675,"reference_mean_power":0.0014670926652510829,"output_mean_power":0.0013060912151322387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5556250379109994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020148633268320188,"output_power":0.0001748317531945926,"reference_active":true,"delta_db":-0.6162527955538498},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.931250874616967e-05,"output_power":1.5173197346663147e-05,"reference_active":true,"delta_db":-1.0476158704091076},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.933225179066709e-07,"output_power":4.3392099650121303e-07,"reference_active":true,"delta_db":-2.0352464034445283}],"active_band_count":3,"active_band_mae_db":1.233038356469162}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":2.558942940585693,"rms_envelope_nrmse":0.22948259661788334,"envelope_correlation":0.9598001318270191,"rms_gain_error_db":-0.454921750888151,"reference_mean_power":0.00020535819062638303,"output_mean_power":0.00018493531616289326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.949818894937679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.42703704863136e-05,"output_power":4.836030566284831e-05,"reference_active":true,"delta_db":-0.5007374920845672},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.92519989001781e-06,"output_power":7.809826448750732e-06,"reference_active":true,"delta_db":-0.5797656839171125},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.574401681634857e-06,"output_power":2.06040457918501e-06,"reference_active":true,"delta_db":-0.9672380402664673}],"active_band_count":3,"active_band_mae_db":0.6825804054227156}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":1.5500190950441466,"rms_envelope_nrmse":0.13751136141623285,"envelope_correlation":0.9859056107520735,"rms_gain_error_db":-0.4548362669871372,"reference_mean_power":0.0002063770763803802,"output_mean_power":0.00018585653196836267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.280870922430584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2679956690924704e-05,"output_power":4.675787400620248e-05,"reference_active":true,"delta_db":-0.5179065287490938},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1805452510023395e-06,"output_power":3.676221150383822e-06,"reference_active":true,"delta_db":-0.558312993320405},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0193067246169139e-07,"output_power":1.4578850393939885e-07,"reference_active":true,"delta_db":-1.4147901223501733}],"active_band_count":3,"active_band_mae_db":0.8303365481398908}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":2.546686057931807,"rms_envelope_nrmse":0.2197423156160606,"envelope_correlation":0.9730556679109287,"rms_gain_error_db":-0.8006048514025766,"reference_mean_power":0.00014718641611289373,"output_mean_power":0.00012240727940500843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4339203545791116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.0687614509069e-05,"output_power":5.870331874433523e-05,"reference_active":true,"delta_db":-0.806806713706158},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.343067234178676e-06,"output_power":6.790976789627121e-06,"reference_active":true,"delta_db":-1.3855722798359504},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.359519184803567e-06,"output_power":2.2735361514476153e-06,"reference_active":true,"delta_db":-0.161216485565972}],"active_band_count":3,"active_band_mae_db":0.78453182636936}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":1.4889800888330476,"rms_envelope_nrmse":0.1441541202135871,"envelope_correlation":0.9911176853223782,"rms_gain_error_db":-0.800564364356746,"reference_mean_power":0.00014792010871620984,"output_mean_power":0.00012301860017933152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3826641859038568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.726675521598139e-05,"output_power":5.5723310794404835e-05,"reference_active":true,"delta_db":-0.8176356680515011},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.270863847079585e-06,"output_power":3.1595272646752646e-06,"reference_active":true,"delta_db":-1.3089361916861626},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6847652723937036e-07,"output_power":1.502600287776109e-07,"reference_active":true,"delta_db":-0.4969593426088528}],"active_band_count":3,"active_band_mae_db":0.8745104007821721}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.9302444458007812,"predicted_mos":2.9302444458007812,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"2b89602a4f247f9527572797f1856dfce788b30c0a998d01f37e992a9e25814d","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2159947479995026},"started_at":"2026-10-13T05:10:18.509079+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030194909995771013,"native_input_samples":144000,"native_input_sha256":"1be42d1b2de3fa9211e421859f28acc21779929beaa6d50a3bf6c890c00ec6a9","output_sha256":"50adeab5fd26bd381cd262e0290d62b65f1ee54c05edc6a6694271d0438762bd","output_samples":144240,"length_delta_samples":240,"analysis_s":1.302151171999867,"synthesis_s":0.14565308099918184,"total_s":1.447804252999049,"rtf":0.2413007088331748,"output_peak":0.508889377117157,"output_rms":0.06286171616752755,"finished_at":"2026-10-13T05:10:19.967784+00:00","analysis_stages":{"external_f0_s":0.17780898199998774},"evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"50adeab5fd26bd381cd262e0290d62b65f1ee54c05edc6a6694271d0438762bd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.02957265382984,"output_peak":0.5128172746120724},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.852630013764005,"spectral_convergence":0.19920588649730858,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.295941209607068,"spectral_convergence":0.18966121544548104,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.630142579640817,"spectral_convergence":0.21847844441421088,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.0015216582134516621,"delta_db":-0.09918036565239333,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.0019123712344691185,"delta_db":-0.10368724570448534,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.00019591531165269127,"delta_db":-0.29605949580418933,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.00016218515803135267,"delta_db":-0.6607323607486162,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":9.62681500012145e-05,"delta_db":-0.5144560578159632,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":7.156528072183485e-05,"delta_db":-0.4575990414838307,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3552857612015797,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1528874089590872,"pitch":null,"output_rms_dbfs":-24.023088336938358,"output_peak":0.5128172746120724,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.592904601003963},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9326408401288979,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06404569659064947,"rms_dbfs":-23.870200927979273,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.867078192515194,"mean":-2.6813824971516927e-05,"zero_energy":false},"output":{"rms":0.06292823967288355,"rms_dbfs":-24.023088336938358,"peak":0.5128172746120724,"peak_dbfs":-5.800747074529747,"crest_factor_db":18.222341262408612,"mean":0.0001592478062441792,"zero_energy":false},"rms_difference_db":-0.15288740895908504},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":4.374747826415143e-06,"output_rms_dbfs":-53.590469759565906,"output_zero_energy":false,"output_peak":0.044331373003568965,"output_energy_fraction":0.00019148905091899022,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":231,"both_estimated_voiced_frames":225,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9375,"absolute_difference_cents_median":1.666609560796793,"absolute_difference_cents_p95":6.224574179963032,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023872023988588299,"log_rms_mae_db":0.8214291308342715,"rms_envelope_nrmse":0.06419006302100894,"envelope_correlation":0.9952566865002073,"rms_gain_error_db":-0.1532803699886926,"reference_mean_power":0.004104010412713017,"output_mean_power":0.0039616893410673445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.972155761177528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003715407512890731,"output_power":0.00034582078071357175,"reference_active":true,"delta_db":-0.3115536803744991},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.499306084417706e-05,"output_power":4.6349907483210586e-05,"reference_active":true,"delta_db":-0.7425902102854927},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6530449727349195e-05,"output_power":1.219277659060316e-05,"reference_active":true,"delta_db":-1.3218205289069365}],"active_band_count":3,"active_band_mae_db":0.7919881398556429}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00021497588341920653,"log_rms_mae_db":0.5113245774951224,"rms_envelope_nrmse":0.037457260347790684,"envelope_correlation":0.9985491640857924,"rms_gain_error_db":-0.15183014427257713,"reference_mean_power":0.00411739985369924,"output_mean_power":0.003975941907802116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.973189563381596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033993681960372955,"output_power":0.00031546685706107167,"reference_active":true,"delta_db":-0.32444467892459317},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413396717591137e-05,"output_power":2.114204237809663e-05,"reference_active":true,"delta_db":-0.5748177871099805},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.827774931478278e-07,"output_power":5.55261245905694e-07,"reference_active":true,"delta_db":-2.0135388884458534}],"active_band_count":3,"active_band_mae_db":0.970933784826809}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001937632671887944,"log_rms_mae_db":1.1615910782205976,"rms_envelope_nrmse":0.09143237621818531,"envelope_correlation":0.9938059862425928,"rms_gain_error_db":-0.19077140400082418,"reference_mean_power":0.0014622039835396937,"output_mean_power":0.00139936440361853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.89031570773124e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020584065582039213,"output_power":0.00018704621184338044,"reference_active":true,"delta_db":-0.41582239825333245},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2632804837837524e-05,"output_power":3.3095118638193895e-05,"reference_active":true,"delta_db":-1.0997996381160613},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.198246611681044e-06,"output_power":5.707457314041836e-06,"reference_active":true,"delta_db":-1.572783061064448}],"active_band_count":3,"active_band_mae_db":1.0294683658112807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001823230477151681,"log_rms_mae_db":0.7619970529938892,"rms_envelope_nrmse":0.05437099701961352,"envelope_correlation":0.9979781235055132,"rms_gain_error_db":-0.18978904175329217,"reference_mean_power":0.0014670926652510829,"output_mean_power":0.0014043606162823167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5556250379109994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020148633268320188,"output_power":0.00018198421461057218,"reference_active":true,"delta_db":-0.4421187347855505},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.931250874616967e-05,"output_power":1.4974205095001803e-05,"reference_active":true,"delta_db":-1.104949162800764},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.933225179066709e-07,"output_power":4.094052365429517e-07,"reference_active":true,"delta_db":-2.28781912301863}],"active_band_count":3,"active_band_mae_db":1.2782956735349815}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":1.355597493072003,"rms_envelope_nrmse":0.14441213905271238,"envelope_correlation":0.9872507108535309,"rms_gain_error_db":-0.6541500803586836,"reference_mean_power":0.00020535819062638303,"output_mean_power":0.00017664323966623833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.949818894937679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.42703704863136e-05,"output_power":4.517833130809333e-05,"reference_active":true,"delta_db":-0.7963260045473836},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.92519989001781e-06,"output_power":7.146639003235148e-06,"reference_active":true,"delta_db":-0.9651610618896032},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.574401681634857e-06,"output_power":1.7066779417189538e-06,"reference_active":true,"delta_db":-1.7852473503948094}],"active_band_count":3,"active_band_mae_db":1.1822448056105987}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":0.8334840358738278,"rms_envelope_nrmse":0.10049498524973116,"envelope_correlation":0.9957614931320088,"rms_gain_error_db":-0.6541875897170907,"reference_mean_power":0.0002063770763803802,"output_mean_power":0.00017751812283463105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.280870922430584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2679956690924704e-05,"output_power":4.376613817352998e-05,"reference_active":true,"delta_db":-0.8050718212238299},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1805452510023395e-06,"output_power":3.344966305627028e-06,"reference_active":true,"delta_db":-0.9684118126989855},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0193067246169139e-07,"output_power":1.1321062250165742e-07,"reference_active":true,"delta_db":-2.5131511312992476}],"active_band_count":3,"active_band_mae_db":1.4288782550740209}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":1.353981182879429,"rms_envelope_nrmse":0.1163592949529248,"envelope_correlation":0.9929636665095466,"rms_gain_error_db":-0.4446857659762856,"reference_mean_power":0.00014718641611289373,"output_mean_power":0.00013286150084516933},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4339203545791116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.0687614509069e-05,"output_power":6.445658174966989e-05,"reference_active":true,"delta_db":-0.4007605544035414},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.343067234178676e-06,"output_power":5.8817443563589416e-06,"reference_active":true,"delta_db":-2.00983329667681},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.359519184803567e-06,"output_power":1.5813575719959843e-06,"reference_active":true,"delta_db":-1.7379343041218263}],"active_band_count":3,"active_band_mae_db":1.382842718400726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":0.940290247046917,"rms_envelope_nrmse":0.07492933005141766,"envelope_correlation":0.9979532429433219,"rms_gain_error_db":-0.4447224834773622,"reference_mean_power":0.00014792010871620984,"output_mean_power":0.0001335226579548672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3826641859038568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.726675521598139e-05,"output_power":6.071434865900577e-05,"reference_active":true,"delta_db":-0.4450913859792665},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.270863847079585e-06,"output_power":2.56539286459603e-06,"reference_active":true,"delta_db":-2.213618441109121},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6847652723937036e-07,"output_power":8.148608596216541e-08,"reference_active":true,"delta_db":-3.1545594419082503}],"active_band_count":3,"active_band_mae_db":1.9377564229988795}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0028","method":"magphase","status":"measured","mos":{"status":"measured","value":3.4276952743530273,"predicted_mos":3.4276952743530273,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288645,"input_sample_rate":48000,"input_seconds":6.0134375,"input_float32_sha256":"78ef74fa4a68f9bc7c73dab19447ecda20c7a299f7a1ee6b5e35afa65738db74","internal_samples":96215,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.666438861999268},"started_at":"2026-10-13T05:13:39.738356+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009086217998628854,"native_input_samples":288000,"native_input_sha256":"10e817c55fffeb2ba8e37330bac674e041a4af825ce0c46f2f82bc43be2110c6","output_sha256":"75de6f1bf428e7d39dbbb0316d9d94964d4f471b1ce67a38f4903dc9ed416971","output_samples":288645,"length_delta_samples":645,"analysis_s":2.6847237950005365,"synthesis_s":0.8135822849999386,"total_s":3.498306080000475,"rtf":0.5830510133334125,"output_peak":0.5236200296496077,"output_rms":0.06091876980420346,"finished_at":"2026-10-13T05:13:43.261144+00:00","evaluation_input":{"reference":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"75de6f1bf428e7d39dbbb0316d9d94964d4f471b1ce67a38f4903dc9ed416971","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96215,"length_delta_samples":215,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.30234499510528,"output_peak":0.5224190438040951},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9977654211921219,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.01781398004593,"spectral_convergence":0.58160500088302,"spectral_convergence_status":"ok","amplitude_floor":1.711947899928738e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.754661423884595,"spectral_convergence":0.32891619196249344,"spectral_convergence_status":"ok","amplitude_floor":1.4743066880059505e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.30226290429535,"spectral_convergence":0.32117363031985113,"spectral_convergence_status":"ok","amplitude_floor":1.2829193968481007e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015568083457567275,"output_power":0.001375809652347319,"delta_db":-0.5367679907237419,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001958578248681761,"output_power":0.001776399512561441,"delta_db":-0.4240028173427386,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00020973665094298736,"output_power":0.00019404863392288705,"delta_db":-0.3376373923612296,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018883580517861778,"output_power":0.0001786062509794694,"delta_db":-0.24187689852825964,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010837478072549288,"output_power":0.00011296137076862516,"delta_db":0.18001722057663383,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.95174204959434e-05,"output_power":8.30727210726209e-05,"delta_db":0.1899615281898535,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.31837730795374286,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4233212568464756,"pitch":null,"output_rms_dbfs":-24.293522184825747,"output_peak":0.5224190438040951,"reference_rms_dbfs":-23.870200927979273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.024912769408625},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7374601622151533,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06404569659064947,"rms_dbfs":-23.870200927979273,"peak":0.501007080078125,"peak_dbfs":-6.00312273546408,"crest_factor_db":17.867078192515194,"mean":-2.6813824971516927e-05,"zero_energy":false},"output":{"rms":0.06099916508979744,"rms_dbfs":-24.293522184825747,"peak":0.5224190438040951,"peak_dbfs":-5.639620001111174,"crest_factor_db":18.653902183714575,"mean":0.0015893946100977753,"zero_energy":false},"rms_difference_db":-0.42332125684647437},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.9451430427143354e-06,"threshold_dbfs":-53.058211415262505,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":1.4717707997008872e-06,"output_mean_power":2.2671030369819995e-06,"output_rms_dbfs":-56.445287412868836,"output_zero_energy":false,"output_peak":0.021619289044165654,"output_energy_fraction":0.00010561012729999293,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":240,"output_estimated_voiced_frames":243,"both_estimated_voiced_frames":219,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9125,"absolute_difference_cents_median":9.535994409213442,"absolute_difference_cents_p95":41.487322864436344,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023872023988588299,"log_rms_mae_db":2.14966562060184,"rms_envelope_nrmse":0.19063127178497177,"envelope_correlation":0.9564519292376072,"rms_gain_error_db":-0.4243262688274907,"reference_mean_power":0.004104010412713017,"output_mean_power":0.00372199535769808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.972155761177528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003715407512890731,"output_power":0.00032598181207411616,"reference_active":true,"delta_db":-0.5681308535668754},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.499306084417706e-05,"output_power":5.1866016346315693e-05,"reference_active":true,"delta_db":-0.25425000054922325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6530449727349195e-05,"output_power":2.4268746285539034e-05,"reference_active":true,"delta_db":1.6676267227292896}],"active_band_count":3,"active_band_mae_db":0.8300025256151294}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00021497588341920653,"log_rms_mae_db":1.4423604724911898,"rms_envelope_nrmse":0.12469324909033029,"envelope_correlation":0.9816550754281802,"rms_gain_error_db":-0.4223153758441408,"reference_mean_power":0.00411739985369924,"output_mean_power":0.003735867865239906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.973189563381596e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033993681960372955,"output_power":0.0002999779662427235,"reference_active":true,"delta_db":-0.5430885048587251},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.413396717591137e-05,"output_power":2.1660655148924185e-05,"reference_active":true,"delta_db":-0.4695712954888917},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.827774931478278e-07,"output_power":1.0795473088967467e-06,"reference_active":true,"delta_db":0.8739042705564006}],"active_band_count":3,"active_band_mae_db":0.6288546903013391}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001937632671887944,"log_rms_mae_db":3.3187608923281102,"rms_envelope_nrmse":0.2295110527345878,"envelope_correlation":0.9592050943364266,"rms_gain_error_db":-0.44869575899329667,"reference_mean_power":0.0014622039835396937,"output_mean_power":0.0013186768646581708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.89031570773124e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020584065582039213,"output_power":0.0001806812893076662,"reference_active":true,"delta_db":-0.5661797594808617},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2632804837837524e-05,"output_power":3.5298019579879405e-05,"reference_active":true,"delta_db":-0.8199356640675406},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.198246611681044e-06,"output_power":7.856138966775293e-06,"reference_active":true,"delta_db":-0.18511821145071505}],"active_band_count":3,"active_band_mae_db":0.5237445449997058}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001823230477151681,"log_rms_mae_db":2.4621498972781724,"rms_envelope_nrmse":0.1612506378702352,"envelope_correlation":0.9797501811887375,"rms_gain_error_db":-0.4444918408019204,"reference_mean_power":0.0014670926652510829,"output_mean_power":0.0013243670338602448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5556250379109994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020148633268320188,"output_power":0.00017680516108825904,"reference_active":true,"delta_db":-0.5675065393059283},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.931250874616967e-05,"output_power":1.5055541645092861e-05,"reference_active":true,"delta_db":-1.081423089425287},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.933225179066709e-07,"output_power":6.519604404399065e-07,"reference_active":true,"delta_db":-0.2671406108104057}],"active_band_count":3,"active_band_mae_db":0.638690079847207}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.475794683802334e-05,"log_rms_mae_db":3.527112225267658,"rms_envelope_nrmse":0.33047276735995784,"envelope_correlation":0.9157450356304393,"rms_gain_error_db":-0.2075333181937172,"reference_mean_power":0.00020535819062638303,"output_mean_power":0.00019577566028550013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.949818894937679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.42703704863136e-05,"output_power":4.98694393174395e-05,"reference_active":true,"delta_db":-0.36728300908940475},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.92519989001781e-06,"output_power":8.368456595696285e-06,"reference_active":true,"delta_db":-0.2797258347684707},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.574401681634857e-06,"output_power":2.7638833368269447e-06,"reference_active":true,"delta_db":0.30843397130565037}],"active_band_count":3,"active_band_mae_db":0.3184809383878419}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.971385946015733e-05,"log_rms_mae_db":2.410468683105789,"rms_envelope_nrmse":0.22934809936638997,"envelope_correlation":0.9564383883816753,"rms_gain_error_db":-0.20755543534027304,"reference_mean_power":0.0002063770763803802,"output_mean_power":0.00019674600030102017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.280870922430584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2679956690924704e-05,"output_power":4.878237560654268e-05,"reference_active":true,"delta_db":-0.333824634622443},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1805452510023395e-06,"output_power":3.875867078894759e-06,"reference_active":true,"delta_db":-0.32864053977343277},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0193067246169139e-07,"output_power":2.1304824296864567e-07,"reference_active":true,"delta_db":0.23275665329342846}],"active_band_count":3,"active_band_mae_db":0.29840727589643473}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.517166954195252e-05,"log_rms_mae_db":3.8373508400306218,"rms_envelope_nrmse":0.34291139986777475,"envelope_correlation":0.9312262830060934,"rms_gain_error_db":0.22669466836725688,"reference_mean_power":0.00014718641611289373,"output_mean_power":0.00015507336020529667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4339203545791116e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.0687614509069e-05,"output_power":7.359071390106778e-05,"reference_active":true,"delta_db":0.1747969022554527},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.343067234178676e-06,"output_power":1.3382149502104845e-05,"reference_active":true,"delta_db":1.5603640329219233},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.359519184803567e-06,"output_power":6.419898213312979e-06,"reference_active":true,"delta_db":4.347046295789941}],"active_band_count":3,"active_band_mae_db":2.027402410322439}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.54521457203503e-05,"log_rms_mae_db":2.7024265882353498,"rms_envelope_nrmse":0.22247566379947983,"envelope_correlation":0.9703928639378556,"rms_gain_error_db":0.22681799345553633,"reference_mean_power":0.00014792010871620984,"output_mean_power":0.00015585079310503696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3826641859038568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.726675521598139e-05,"output_power":7.275788030366839e-05,"reference_active":true,"delta_db":0.34079559147131666},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.270863847079585e-06,"output_power":6.933499479686788e-06,"reference_active":true,"delta_db":2.1043676073282813},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6847652723937036e-07,"output_power":4.3263322291833064e-07,"reference_active":true,"delta_db":4.095804649338052}],"active_band_count":3,"active_band_mae_db":2.1803226160458835}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2284395694732666,"predicted_mos":2.2284395694732666,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"def4033d3025046d5d3f1ecc6c5edc2718b650d94fa859e9b4a0bd956686b060","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8973654210003588},"started_at":"2026-10-13T04:38:37.379213+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3126999874657486e-05,"native_input_samples":60000,"native_input_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","output_sha256":"63e1d0767e155130de2ab211dff8d9dbf8520d9a00b9083d1fd6fe49f59ff8a1","output_samples":60000,"length_delta_samples":0,"analysis_s":0.7195247019999442,"synthesis_s":0.3190044110006056,"total_s":1.0385291130005498,"rtf":0.2769410968001466,"output_peak":0.48775796123234794,"output_rms":0.06725940379433173,"finished_at":"2026-10-13T04:38:38.421262+00:00","evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"63e1d0767e155130de2ab211dff8d9dbf8520d9a00b9083d1fd6fe49f59ff8a1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.856568383687963,"spectral_convergence":0.21890327662785505,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.508883164107519,"spectral_convergence":0.21363419548125004,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.902165437477444,"spectral_convergence":0.2286271894843504,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0016028749477429852,"delta_db":0.08033556174480082,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.0021759498952223164,"delta_db":-0.18832240943807432,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.000362710609248266,"delta_db":0.11520392192234766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.00022164793426500314,"delta_db":-0.36711265689527506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00011753965867398084,"delta_db":-0.19977448435225997,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":4.310435361641441e-05,"delta_db":0.45196132671442557,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2337850601778639,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07460370376666854,"pitch":null,"output_rms_dbfs":-23.44493973417712,"output_peak":0.48775796123234794,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.755872328424309},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.856568383687963,"spectral_convergence":0.21890327662785505,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.508883164107519,"spectral_convergence":0.21363419548125004,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.902165437477444,"spectral_convergence":0.2286271894843504,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0016028749477429852,"delta_db":0.08033556174480082,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.0021759498952223164,"delta_db":-0.18832240943807432,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.000362710609248266,"delta_db":0.11520392192234766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.00022164793426500314,"delta_db":-0.36711265689527506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00011753965867398084,"delta_db":-0.19977448435225997,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":4.310435361641441e-05,"delta_db":0.45196132671442557,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2337850601778639,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07460370376666854,"pitch":null,"output_rms_dbfs":-23.44493973417712,"output_peak":0.48775796123234794,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.755872328424309},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9273817637764903,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06783958749363862,"rms_dbfs":-23.37033603041045,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.17949258376381,"mean":2.234954833984375e-05,"zero_energy":false},"output":{"rms":0.06725940379433173,"rms_dbfs":-23.44493973417712,"peak":0.48775796123234794,"peak_dbfs":-6.235912665585925,"crest_factor_db":17.209027068591194,"mean":-8.29245128263516e-06,"zero_energy":false},"rms_difference_db":-0.07460370376666958},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":3.4953892199237914e-07,"output_mean_power":8.999694653512771e-07,"output_rms_dbfs":-60.45772225287796,"output_zero_energy":false,"output_peak":0.028067714594704175,"output_energy_fraction":4.668454737977771e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":117,"both_estimated_voiced_frames":97,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9238095238095239,"absolute_difference_cents_median":2.3265448129411745,"absolute_difference_cents_p95":10.194341548407847,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00019859470734653442,"log_rms_mae_db":0.6144160200716244,"rms_envelope_nrmse":0.061563019829726555,"envelope_correlation":0.9957367159623541,"rms_gain_error_db":-0.07460370376661167,"reference_mean_power":0.004608354103444978,"output_mean_power":0.004529867221731335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.3747785986813366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009662478923264104,"output_power":0.0009389927643466653,"reference_active":true,"delta_db":-0.1242631389243227},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.594410683545364e-05,"output_power":9.161582275751026e-05,"reference_active":true,"delta_db":-0.20047818123219496},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744648631903194e-05,"output_power":1.8275966421141467e-05,"reference_active":true,"delta_db":0.20172377221982732}],"active_band_count":3,"active_band_mae_db":0.17548836412544833}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00016219647203140816,"log_rms_mae_db":0.3377831345333544,"rms_envelope_nrmse":0.03442450861863409,"envelope_correlation":0.9986259489370307,"rms_gain_error_db":-0.07460370376661069,"reference_mean_power":0.004645568268479526,"output_mean_power":0.004566447576146393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1771130170213754e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008502808342452142,"output_power":0.0008276908286640574,"reference_active":true,"delta_db":-0.11694247003305604},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1331740410000496e-05,"output_power":3.9372201236710376e-05,"reference_active":true,"delta_db":-0.21093996898283288},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2770439378675007e-06,"output_power":1.2873421089022368e-06,"reference_active":true,"delta_db":0.034881354172691385}],"active_band_count":3,"active_band_mae_db":0.12092126439619344}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015290889340024391,"log_rms_mae_db":1.3898626662781695,"rms_envelope_nrmse":0.10802436801549062,"envelope_correlation":0.9905342004858878,"rms_gain_error_db":-0.19111817790342728,"reference_mean_power":0.0016143224036409894,"output_mean_power":0.0015448220374207535},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":8.858385684123232e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006875363814202012,"output_power":0.0006366893216142749,"reference_active":true,"delta_db":-0.3336811802392471},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.173936122300265e-05,"output_power":5.078215500889433e-05,"reference_active":true,"delta_db":-0.8485100500900986},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0160658568889886e-05,"output_power":7.849966004995363e-06,"reference_active":true,"delta_db":-1.1205408190828496}],"active_band_count":3,"active_band_mae_db":0.7675773498040651}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012027897252950328,"log_rms_mae_db":0.9858211563907362,"rms_envelope_nrmse":0.07043667838943433,"envelope_correlation":0.9959026720645339,"rms_gain_error_db":-0.19111817790342625,"reference_mean_power":0.0016273586545452231,"output_mean_power":0.001557297047144205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.855745033342305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006306037982733972,"output_power":0.0005820022882332412,"reference_active":true,"delta_db":-0.3483189012638558},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.290965975606541e-05,"output_power":1.9648665770009932e-05,"reference_active":true,"delta_db":-0.6668557406583598},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0400487875049354e-06,"output_power":6.14467655858752e-07,"reference_active":true,"delta_db":-2.285546844588594}],"active_band_count":3,"active_band_mae_db":1.1002404955036031}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":1.4520378155715297,"rms_envelope_nrmse":0.1379464212428296,"envelope_correlation":0.9874299743655645,"rms_gain_error_db":-0.2893536656078955,"reference_mean_power":0.0003068909357402281,"output_mean_power":0.0002871102483502495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.0486708619890829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.911544017889025e-05,"output_power":5.759473625268392e-05,"reference_active":true,"delta_db":-0.7919228445298396},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0167211042917926e-05,"output_power":7.27693996907601e-06,"reference_active":true,"delta_db":-1.4525304615627064},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.330972389293411e-06,"output_power":2.477685099979541e-06,"reference_active":true,"delta_db":0.26508979860004256}],"active_band_count":3,"active_band_mae_db":0.8365143682308629}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":0.9463749199392638,"rms_envelope_nrmse":0.084746006029781,"envelope_correlation":0.9955108597262882,"rms_gain_error_db":-0.28935366560789705,"reference_mean_power":0.0003093691936331506,"output_mean_power":0.0002894287698712475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.094392572662044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.275291186820285e-05,"output_power":5.147684533871404e-05,"reference_active":true,"delta_db":-0.8602195867529623},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.365294966430716e-06,"output_power":3.2882272056335836e-06,"reference_active":true,"delta_db":-1.2305177643612442},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9190263566646987e-07,"output_power":1.681811378583688e-07,"reference_active":true,"delta_db":-0.573036531003346}],"active_band_count":3,"active_band_mae_db":0.8879246273725174}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":1.5122177847335097,"rms_envelope_nrmse":0.2181357473232493,"envelope_correlation":0.9681546942961492,"rms_gain_error_db":-0.09450007320424816,"reference_mean_power":6.929937891252565e-05,"output_mean_power":6.780775018113458e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.858105660801018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988342411251764e-05,"output_power":3.5401807174787955e-05,"reference_active":true,"delta_db":-0.5176700448234359},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1366606898498144e-05,"output_power":1.0721584981914723e-05,"reference_active":true,"delta_db":-0.2537184839968997},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7366453289643982e-06,"output_power":2.3374487605445555e-06,"reference_active":true,"delta_db":1.2903096688365348}],"active_band_count":3,"active_band_mae_db":0.6872327325522901}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":1.2798978655636217,"rms_envelope_nrmse":0.17394321386072292,"envelope_correlation":0.9788375056237572,"rms_gain_error_db":-0.09450007320424865,"reference_mean_power":6.985899704640877e-05,"output_mean_power":6.835532286092839e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.43114016941622e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0602300645347656e-05,"output_power":3.41663895971195e-05,"reference_active":true,"delta_db":-0.7495155386179854},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067895142462184e-06,"output_power":5.808429123643207e-06,"reference_active":true,"delta_db":-0.18979372940603134},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2311303789031625e-07,"output_power":1.7440023054591354e-07,"reference_active":true,"delta_db":1.5124300677921985}],"active_band_count":3,"active_band_mae_db":0.8172464452720717}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.4924468994140625,"predicted_mos":3.4924468994140625,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"831659cb6fe922add805ca0ff166d482a14590b9ba8dbb85d479808c188ff470","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9372306400000525},"started_at":"2026-10-13T04:41:34.454225+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1733000721724238e-05,"native_input_samples":60000,"native_input_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","output_sha256":"7884cb89da20f660b4c8180186a259d2d0ca78232df0cbdeef410f517c466e54","output_samples":60000,"length_delta_samples":0,"analysis_s":0.9012727919998724,"synthesis_s":0.05107524600043689,"total_s":0.9523480380003093,"rtf":0.25395947680008246,"output_peak":0.6646857006443889,"output_rms":0.0757057328929434,"finished_at":"2026-10-13T04:41:35.410694+00:00","evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"7884cb89da20f660b4c8180186a259d2d0ca78232df0cbdeef410f517c466e54","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.170868816270065,"spectral_convergence":0.6703320055690403,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.223076922653773,"spectral_convergence":0.31003915955631317,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.550635804087893,"spectral_convergence":0.2963188537715729,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0016176072493422592,"delta_db":0.12006999609307434,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.0031674630401086062,"delta_db":1.4423042422221997,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.0004784058781030586,"delta_db":1.317566404827082,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.0002925131796342052,"delta_db":0.8377048490363241,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00013563284543794925,"delta_db":0.4220299924414015,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":3.973580023161273e-05,"delta_db":0.09856959393660325,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7063741797594475,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9529113930666857,"pitch":null,"output_rms_dbfs":-22.417424637343764,"output_peak":0.6646857006443889,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.98152718100391},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.170868816270065,"spectral_convergence":0.6703320055690403,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.223076922653773,"spectral_convergence":0.31003915955631317,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.550635804087893,"spectral_convergence":0.2963188537715729,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0016176072493422592,"delta_db":0.12006999609307434,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.0031674630401086062,"delta_db":1.4423042422221997,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.0004784058781030586,"delta_db":1.317566404827082,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.0002925131796342052,"delta_db":0.8377048490363241,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00013563284543794925,"delta_db":0.4220299924414015,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":3.973580023161273e-05,"delta_db":0.09856959393660325,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7063741797594475,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9529113930666857,"pitch":null,"output_rms_dbfs":-22.417424637343764,"output_peak":0.6646857006443889,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.98152718100391},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9532691306411952,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06783958749363862,"rms_dbfs":-23.37033603041045,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.17949258376381,"mean":2.234954833984375e-05,"zero_energy":false},"output":{"rms":0.0757057328929434,"rms_dbfs":-22.417424637343764,"peak":0.6646857006443889,"peak_dbfs":-3.547673281686896,"crest_factor_db":18.86975135565687,"mean":2.5098990478570707e-14,"zero_energy":false},"rms_difference_db":0.9529113930666853},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":3.4953892199237914e-07,"output_mean_power":8.524661113419307e-07,"output_rms_dbfs":-60.69322876792632,"output_zero_energy":false,"output_peak":0.014086979230761604,"output_energy_fraction":3.49036617576845e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":104,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8857142857142857,"absolute_difference_cents_median":5.4651139395012915,"absolute_difference_cents_p95":16.03246421526079,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00019859470734653442,"log_rms_mae_db":1.4526757930549874,"rms_envelope_nrmse":0.19909113377194676,"envelope_correlation":0.974833608312625,"rms_gain_error_db":0.9529113930666535,"reference_mean_power":0.004608354103444978,"output_mean_power":0.00573901000619925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.3747785986813366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009662478923264104,"output_power":0.0011693686785214544,"reference_active":true,"delta_db":0.8286289750001536},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.594410683545364e-05,"output_power":0.00011984852171758425,"reference_active":true,"delta_db":0.9661437729070652},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744648631903194e-05,"output_power":2.495036084375008e-05,"reference_active":true,"delta_db":1.553688567510318}],"active_band_count":3,"active_band_mae_db":1.1161537718058456}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00016219647203140816,"log_rms_mae_db":1.0564195784535262,"rms_envelope_nrmse":0.15897808331554225,"envelope_correlation":0.9881563860058046,"rms_gain_error_db":0.9529113930666527,"reference_mean_power":0.004645568268479526,"output_mean_power":0.00578535463612818},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1771130170213754e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008502808342452142,"output_power":0.0010342540751880982,"reference_active":true,"delta_db":0.8506485081422825},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1331740410000496e-05,"output_power":5.394556587876403e-05,"reference_active":true,"delta_db":1.1567206004792425},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2770439378675007e-06,"output_power":1.754757623767467e-06,"reference_active":true,"delta_db":1.380112981156181}],"active_band_count":3,"active_band_mae_db":1.1291606965925685}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015290889340024391,"log_rms_mae_db":1.8011949595435346,"rms_envelope_nrmse":0.3000866939620286,"envelope_correlation":0.9721712750017031,"rms_gain_error_db":1.7131582501717513,"reference_mean_power":0.0016143224036409894,"output_mean_power":0.0023950032053253636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":8.858385684123232e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006875363814202012,"output_power":0.0009528543592093841,"reference_active":true,"delta_db":1.4173084108969847},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.173936122300265e-05,"output_power":8.89595059231705e-05,"reference_active":true,"delta_db":1.5863023038704773},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0160658568889886e-05,"output_power":1.0331517164929808e-05,"reference_active":true,"delta_db":0.07242243670227974}],"active_band_count":3,"active_band_mae_db":1.0253443838232472}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012027897252950328,"log_rms_mae_db":1.5451737655752016,"rms_envelope_nrmse":0.27083052250501505,"envelope_correlation":0.9819783919647909,"rms_gain_error_db":1.7131582501717517,"reference_mean_power":0.0016273586545452231,"output_mean_power":0.00241434374264966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.855745033342305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006306037982733972,"output_power":0.0008756483894759316,"reference_active":true,"delta_db":1.4257317096398676},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.290965975606541e-05,"output_power":3.4361223492587404e-05,"reference_active":true,"delta_db":1.7604997997186382},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0400487875049354e-06,"output_power":9.655896226733416e-07,"reference_active":true,"delta_db":-0.3226112234562309}],"active_band_count":3,"active_band_mae_db":1.169614244271579}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":1.432680358152062,"rms_envelope_nrmse":0.20252926165297155,"envelope_correlation":0.9795818276027425,"rms_gain_error_db":0.7477762651815889,"reference_mean_power":0.0003068909357402281,"output_mean_power":0.00036455384899414185},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.0486708619890829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.911544017889025e-05,"output_power":8.343875264646437e-05,"reference_active":true,"delta_db":0.817927247773153},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0167211042917926e-05,"output_power":1.2376245768548807e-05,"reference_active":true,"delta_db":0.8538708664232546},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.330972389293411e-06,"output_power":2.4780647663051876e-06,"reference_active":true,"delta_db":0.26575523569821824}],"active_band_count":3,"active_band_mae_db":0.6458511166315419}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":0.9286440929137041,"rms_envelope_nrmse":0.11684380874128238,"envelope_correlation":0.9963754419496419,"rms_gain_error_db":0.7477762651815874,"reference_mean_power":0.0003093691936331506,"output_mean_power":0.00036749775625385223},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.094392572662044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.275291186820285e-05,"output_power":7.569206927128654e-05,"reference_active":true,"delta_db":0.8141649534496874},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.365294966430716e-06,"output_power":5.3593516807515655e-06,"reference_active":true,"delta_db":0.8909866173578196},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9190263566646987e-07,"output_power":1.805140128734938e-07,"reference_active":true,"delta_db":-0.2657001875009017}],"active_band_count":3,"active_band_mae_db":0.6569505861028029}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":1.562124864893014,"rms_envelope_nrmse":0.23869870565168988,"envelope_correlation":0.9639511742197377,"rms_gain_error_db":0.26671255990062154,"reference_mean_power":6.929937891252565e-05,"output_mean_power":7.368864968215675e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.858105660801018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988342411251764e-05,"output_power":4.258771762693646e-05,"reference_active":true,"delta_db":0.2849192912985128},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1366606898498144e-05,"output_power":8.658166438627955e-06,"reference_active":true,"delta_db":-1.182049105256793},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7366453289643982e-06,"output_power":2.6059057006794977e-06,"reference_active":true,"delta_db":1.7624756339356775}],"active_band_count":3,"active_band_mae_db":1.0764813434969944}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":1.0996427715208947,"rms_envelope_nrmse":0.14519007129836933,"envelope_correlation":0.9864395869411523,"rms_gain_error_db":0.2667125599006134,"reference_mean_power":6.985899704640877e-05,"output_mean_power":7.428371280206635e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.43114016941622e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0602300645347656e-05,"output_power":4.356060674230241e-05,"reference_active":true,"delta_db":0.3054327760767557},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067895142462184e-06,"output_power":4.545248151610858e-06,"reference_active":true,"delta_db":-1.2548046838798679},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2311303789031625e-07,"output_power":1.993208894890046e-07,"reference_active":true,"delta_db":2.0924876865788953}],"active_band_count":3,"active_band_mae_db":1.217575048845173}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"facodec","status":"measured","mos":{"status":"measured","value":3.7411444187164307,"predicted_mos":3.7411444187164307,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"d9e2c639a365f5ce645a260bc6eb8822341c79fcaec38d43ed4cc5d66ed17e26","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.154082068000207},"started_at":"2026-10-13T04:48:38.146280+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4158000087481923e-05,"native_input_samples":60000,"native_input_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","output_sha256":"70260a4b8bb3c283455cfb9d0aad15363daa1fb75afbd350bd67e08e5779d1ec","output_samples":60000,"length_delta_samples":0,"analysis_s":1.094189113999164,"synthesis_s":6.972148630999982,"total_s":8.066337744999146,"rtf":2.151023398666439,"output_peak":0.38984453678131104,"output_rms":0.06483242545631664,"finished_at":"2026-10-13T04:48:46.222038+00:00","evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"70260a4b8bb3c283455cfb9d0aad15363daa1fb75afbd350bd67e08e5779d1ec","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.4818290136944725,"spectral_convergence":0.20705667023241106,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.439002598536141,"spectral_convergence":0.20382704732311727,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.596274922825819,"spectral_convergence":0.2214168933579758,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0014982562178464218,"delta_db":-0.212799962237413,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.00202080317175855,"delta_db":-0.5095711688963895,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.0003212777963275739,"delta_db":-0.4115915436782007,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.00020746411373171584,"delta_db":-0.6543196865082969,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.0001272709964296094,"delta_db":0.14567570641950442,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":2.8171094454983753e-05,"delta_db":-1.3952128444315166,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5548618186952202,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.39381858223595956,"pitch":null,"output_rms_dbfs":-23.764154612646408,"output_peak":0.38984453678131104,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.505702178352145},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.4818290136944725,"spectral_convergence":0.20705667023241106,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.439002598536141,"spectral_convergence":0.20382704732311727,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.596274922825819,"spectral_convergence":0.2214168933579758,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0014982562178464218,"delta_db":-0.212799962237413,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.00202080317175855,"delta_db":-0.5095711688963895,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.0003212777963275739,"delta_db":-0.4115915436782007,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.00020746411373171584,"delta_db":-0.6543196865082969,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.0001272709964296094,"delta_db":0.14567570641950442,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":2.8171094454983753e-05,"delta_db":-1.3952128444315166,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5548618186952202,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.39381858223595956,"pitch":null,"output_rms_dbfs":-23.764154612646408,"output_peak":0.38984453678131104,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.505702178352145},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9333813389048269,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06783958749363862,"rms_dbfs":-23.37033603041045,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.17949258376381,"mean":2.234954833984375e-05,"zero_energy":false},"output":{"rms":0.06483242545631664,"rms_dbfs":-23.764154612646408,"peak":0.38984453678131104,"peak_dbfs":-8.182170950674763,"crest_factor_db":15.581983661971645,"mean":-0.00014448276617183402,"zero_energy":false},"rms_difference_db":-0.39381858223595856},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":3.4953892199237914e-07,"output_mean_power":2.8629069807145284e-07,"output_rms_dbfs":-65.4319276249837,"output_zero_energy":false,"output_peak":0.0048803240060806274,"output_energy_fraction":1.5983581622982885e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":109,"both_estimated_voiced_frames":98,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9333333333333333,"absolute_difference_cents_median":5.201075330080384,"absolute_difference_cents_p95":18.357735501022272,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00019859470734653442,"log_rms_mae_db":0.813692779537325,"rms_envelope_nrmse":0.09927777264116287,"envelope_correlation":0.9907277288066121,"rms_gain_error_db":-0.39381858272352743,"reference_mean_power":0.004608354103444978,"output_mean_power":0.004208855197006312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.3747785986813366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009662478923264104,"output_power":0.0008583413749581979,"reference_active":true,"delta_db":-0.5142851210935944},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.594410683545364e-05,"output_power":0.00010261594455546132,"reference_active":true,"delta_db":0.2919654280896593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744648631903194e-05,"output_power":1.8191379360605208e-05,"reference_active":true,"delta_db":0.18157656450509463}],"active_band_count":3,"active_band_mae_db":0.32927570456278277}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00016219647203140816,"log_rms_mae_db":0.7084646938513671,"rms_envelope_nrmse":0.09092588659177084,"envelope_correlation":0.9921137909441751,"rms_gain_error_db":-0.39381858415262144,"reference_mean_power":0.004645568268479526,"output_mean_power":0.004242843259112227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1771130170213754e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008502808342452142,"output_power":0.0007448545406724571,"reference_active":true,"delta_db":-0.5749092037996791},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1331740410000496e-05,"output_power":4.831899702182961e-05,"reference_active":true,"delta_db":0.6783421779340215},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2770439378675007e-06,"output_power":1.3122329922529746e-06,"reference_active":true,"delta_db":0.11805112800263406}],"active_band_count":3,"active_band_mae_db":0.4571008365787783}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015290889340024391,"log_rms_mae_db":1.1819651603390888,"rms_envelope_nrmse":0.11980358297635466,"envelope_correlation":0.9900769557609546,"rms_gain_error_db":-0.48942679419018964,"reference_mean_power":0.0016143224036409894,"output_mean_power":0.0014422734023224269},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":8.858385684123232e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006875363814202012,"output_power":0.0006086606022223446,"reference_active":true,"delta_db":-0.5292049270782369},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.173936122300265e-05,"output_power":6.244809563062767e-05,"reference_active":true,"delta_db":0.049570671927505326},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0160658568889886e-05,"output_power":1.0196363982065717e-05,"reference_active":true,"delta_db":0.015234722837294882}],"active_band_count":3,"active_band_mae_db":0.1980034406143457}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012027897252950328,"log_rms_mae_db":0.9139696060360749,"rms_envelope_nrmse":0.10621594124071689,"envelope_correlation":0.9922103094854452,"rms_gain_error_db":-0.489426795009581,"reference_mean_power":0.0016273586545452231,"output_mean_power":0.0014539202935877286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.855745033342305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006306037982733972,"output_power":0.0005587755142367885,"reference_active":true,"delta_db":-0.5251921537269456},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.290965975606541e-05,"output_power":2.451948346530819e-05,"reference_active":true,"delta_db":0.29492677675207163},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0400487875049354e-06,"output_power":8.883869860182996e-07,"reference_active":true,"delta_db":-0.6845152409368173}],"active_band_count":3,"active_band_mae_db":0.5015447238052783}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":1.4105773684871747,"rms_envelope_nrmse":0.1440897056943055,"envelope_correlation":0.9883396582456592,"rms_gain_error_db":-0.6173191771224613,"reference_mean_power":0.0003068909357402281,"output_mean_power":0.00026622702781833874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.0486708619890829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.911544017889025e-05,"output_power":5.98783614998962e-05,"reference_active":true,"delta_db":-0.6230517039575991},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0167211042917926e-05,"output_power":1.121344546630917e-05,"reference_active":true,"delta_db":0.42537236923823496},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.330972389293411e-06,"output_power":3.1647126396404282e-06,"reference_active":true,"delta_db":1.3279715232391807}],"active_band_count":3,"active_band_mae_db":0.7921318654783382}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":1.1388560689517033,"rms_envelope_nrmse":0.12186900682508375,"envelope_correlation":0.9923676241286008,"rms_gain_error_db":-0.6173191802963706,"reference_mean_power":0.0003093691936331506,"output_mean_power":0.00026837690940808526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.094392572662044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.275291186820285e-05,"output_power":5.373663418969334e-05,"reference_active":true,"delta_db":-0.6736342202266876},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.365294966430716e-06,"output_power":5.249257907876945e-06,"reference_active":true,"delta_db":0.8008431650503577},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9190263566646987e-07,"output_power":2.631863158271303e-07,"reference_active":true,"delta_db":1.371823651686511}],"active_band_count":3,"active_band_mae_db":0.9487670123211854}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":1.74492743163157,"rms_envelope_nrmse":0.21000993548654034,"envelope_correlation":0.9710944208890939,"rms_gain_error_db":-0.05270881138839501,"reference_mean_power":6.929937891252565e-05,"output_mean_power":6.846339969951602e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.858105660801018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988342411251764e-05,"output_power":4.524818707358646e-05,"reference_active":true,"delta_db":0.5480874659235031},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1366606898498144e-05,"output_power":1.1796350457871374e-05,"reference_active":true,"delta_db":0.1611682586866298},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7366453289643982e-06,"output_power":2.2145868918681536e-06,"reference_active":true,"delta_db":1.0558159244833105}],"active_band_count":3,"active_band_mae_db":0.5883572163644811}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":1.4939384372729894,"rms_envelope_nrmse":0.17626710622902175,"envelope_correlation":0.9785667092863535,"rms_gain_error_db":-0.052708821192094094,"reference_mean_power":6.985899704640877e-05,"output_mean_power":6.901626683604422e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.43114016941622e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0602300645347656e-05,"output_power":4.643356687558054e-05,"reference_active":true,"delta_db":0.5828140336458316},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067895142462184e-06,"output_power":6.190196422020955e-06,"reference_active":true,"delta_db":0.08666362637558389},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2311303789031625e-07,"output_power":1.3346670391274954e-07,"reference_active":true,"delta_db":0.35068887357522166}],"active_band_count":3,"active_band_mae_db":0.34005551119887906}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.864448308944702,"predicted_mos":2.864448308944702,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":165376,"input_sample_rate":44100,"input_seconds":3.7500226757369615,"input_float32_sha256":"5fa818c5bc9545f728b45aa22b8ab0be4c1f7bbb996eff4cdd1050ca84a3a757","internal_samples":60001,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5394541309979104},"started_at":"2026-10-13T07:06:43.783244+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004695624000305543,"native_input_samples":165375,"native_input_sha256":"684c9df11c50c8ac3105326341a68a6ddf736d659c7229145a6eb263ba70a22b","output_sha256":"6e2411c8f99d6ce4e374e67891fd128ec46928afc7ee9941732dd58cc61cff25","output_samples":165376,"length_delta_samples":1,"analysis_s":0.11154868400080886,"synthesis_s":23.59862795799927,"total_s":23.71017664200008,"rtf":6.3227137712000205,"output_peak":0.40722960233688354,"output_rms":0.060254365840831,"finished_at":"2026-10-13T07:07:07.511116+00:00","evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"6e2411c8f99d6ce4e374e67891fd128ec46928afc7ee9941732dd58cc61cff25","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60001,"length_delta_samples":1,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.396945932418813,"output_peak":0.4036826074985804},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9999833336111065,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.384375606231607,"spectral_convergence":0.5159765342558005,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.952864934239548,"spectral_convergence":0.3546813883156618,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.937609872481312,"spectral_convergence":0.4581956578812882,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0013747678236251887,"delta_db":-0.5863672599952692,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.0016391607797144104,"delta_db":-1.4185957761912067,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.00032546508539755664,"delta_db":-0.3553546062832616,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.0001602625390960827,"delta_db":-1.775429346251215,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00010940640441100519,"delta_db":-0.5111912826989538,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":2.4332624829165892e-05,"delta_db":-2.0313604418627627,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1130497855471113,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0265375201979836,"pitch":null,"output_rms_dbfs":-24.396873550608433,"output_peak":0.4036826074985804,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.091616804317489},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.7925866106519079,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06783958749363862,"rms_dbfs":-23.37033603041045,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.17949258376381,"mean":2.234954833984375e-05,"zero_energy":false},"output":{"rms":0.060277651389826145,"rms_dbfs":-24.396873550608433,"peak":0.4036826074985804,"peak_dbfs":-7.879199231764469,"crest_factor_db":16.517674318843966,"mean":-0.00017951610389299347,"zero_energy":false},"rms_difference_db":-1.0265375201979836},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":3.4953892199237914e-07,"output_mean_power":5.189604879218077e-07,"output_rms_dbfs":-62.84865706758019,"output_zero_energy":false,"output_peak":0.009423607913954496,"output_energy_fraction":3.3517610723808186e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":82,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.780952380952381,"absolute_difference_cents_median":20.712252869814584,"absolute_difference_cents_p95":54.46975898021625,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00019859470734653442,"log_rms_mae_db":1.5389587220698053,"rms_envelope_nrmse":0.1648517645921952,"envelope_correlation":0.9825101468112806,"rms_gain_error_db":-1.0265375202099258,"reference_mean_power":0.004608354103444978,"output_mean_power":0.0036382462520661743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.3747785986813366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009662478923264104,"output_power":0.0007936954256262849,"reference_active":true,"delta_db":-0.8543468229640425},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.594410683545364e-05,"output_power":7.759240720905761e-05,"reference_active":true,"delta_db":-0.9219907873793327},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744648631903194e-05,"output_power":1.660648905125977e-05,"reference_active":true,"delta_db":-0.21430150732413722}],"active_band_count":3,"active_band_mae_db":0.6635463725558375}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00016219647203140816,"log_rms_mae_db":1.354024584680676,"rms_envelope_nrmse":0.1436470270477656,"envelope_correlation":0.9898476376112384,"rms_gain_error_db":-1.026537520223558,"reference_mean_power":0.004645568268479526,"output_mean_power":0.003667626437131912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1771130170213754e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008502808342452142,"output_power":0.0007062983476045443,"reference_active":true,"delta_db":-0.8057419975831231},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1331740410000496e-05,"output_power":3.29785815683398e-05,"reference_active":true,"delta_db":-0.9805172067278596},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2770439378675007e-06,"output_power":8.411717028853997e-07,"reference_active":true,"delta_db":-1.8132118527751246}],"active_band_count":3,"active_band_mae_db":1.199823685695369}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015290889340024391,"log_rms_mae_db":2.407884952408734,"rms_envelope_nrmse":0.1972227798854127,"envelope_correlation":0.9778176092918152,"rms_gain_error_db":-1.1373759120309754,"reference_mean_power":0.0016143224036409894,"output_mean_power":0.0012423749403667067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":8.858385684123232e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006875363814202012,"output_power":0.0004782279531009632,"reference_active":true,"delta_db":-1.5766072642705233},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.173936122300265e-05,"output_power":4.723324047250075e-05,"reference_active":true,"delta_db":-1.1631439010381381},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0160658568889886e-05,"output_power":8.602210453974008e-06,"reference_active":true,"delta_db":-0.7231179449739886}],"active_band_count":3,"active_band_mae_db":1.15428970342755}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012027897252950328,"log_rms_mae_db":1.9593932897465198,"rms_envelope_nrmse":0.16512983243635454,"envelope_correlation":0.9877228068738686,"rms_gain_error_db":-1.137375912031902,"reference_mean_power":0.0016273586545452231,"output_mean_power":0.0012524075778391178},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.855745033342305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006306037982733972,"output_power":0.0004377485982001343,"reference_active":true,"delta_db":-1.5853181825753206},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.290965975606541e-05,"output_power":1.8917399536484082e-05,"reference_active":true,"delta_db":-0.8315720302782592},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0400487875049354e-06,"output_power":5.36434648057575e-07,"reference_active":true,"delta_db":-2.8753689107793887}],"active_band_count":3,"active_band_mae_db":1.7640863745443227}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":2.9137945705542734,"rms_envelope_nrmse":0.27551170523834656,"envelope_correlation":0.9604697153597105,"rms_gain_error_db":-1.508236759236896,"reference_mean_power":0.0003068909357402281,"output_mean_power":0.00021685047883982552},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.0486708619890829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.911544017889025e-05,"output_power":4.553176737946328e-05,"reference_active":true,"delta_db":-1.8126056989239816},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0167211042917926e-05,"output_power":7.715968221262026e-06,"reference_active":true,"delta_db":-1.1981140808736965},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.330972389293411e-06,"output_power":2.299408678668772e-06,"reference_active":true,"delta_db":-0.05920963082208733}],"active_band_count":3,"active_band_mae_db":1.0233098035399217}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":2.564108719555636,"rms_envelope_nrmse":0.2364708343651131,"envelope_correlation":0.9750260222843384,"rms_gain_error_db":-1.508236759264274,"reference_mean_power":0.0003093691936331506,"output_mean_power":0.00021860162671602347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.094392572662044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.275291186820285e-05,"output_power":4.141997215031735e-05,"reference_active":true,"delta_db":-1.8042408025616894},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.365294966430716e-06,"output_power":3.3766408363849658e-06,"reference_active":true,"delta_db":-1.1152872622262824},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9190263566646987e-07,"output_power":1.4380215148176214e-07,"reference_active":true,"delta_db":-1.2531555579840312}],"active_band_count":3,"active_band_mae_db":1.390894540924001}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":2.9126606611823607,"rms_envelope_nrmse":0.24697121558112892,"envelope_correlation":0.9687879879725786,"rms_gain_error_db":-1.3405931941435076,"reference_mean_power":6.929937891252565e-05,"output_mean_power":5.089440282650003e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.858105660801018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988342411251764e-05,"output_power":3.0519855844691744e-05,"reference_active":true,"delta_db":-1.1620995876214428},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1366606898498144e-05,"output_power":6.815463398256818e-06,"reference_active":true,"delta_db":-2.2213545085514315},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7366453289643982e-06,"output_power":9.397723844691843e-07,"reference_active":true,"delta_db":-2.6668845356757274}],"active_band_count":3,"active_band_mae_db":2.016779543949534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":2.54868489531042,"rms_envelope_nrmse":0.2064853145965144,"envelope_correlation":0.9820485140911211,"rms_gain_error_db":-1.3405931942847986,"reference_mean_power":6.985899704640877e-05,"output_mean_power":5.130539396474879e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.43114016941622e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0602300645347656e-05,"output_power":3.0350138961958577e-05,"reference_active":true,"delta_db":-1.2638995878437573},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067895142462184e-06,"output_power":3.5368433432060784e-06,"reference_active":true,"delta_db":-2.3442224315247517},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2311303789031625e-07,"output_power":7.467262748716304e-08,"reference_active":true,"delta_db":-2.171426149181591}],"active_band_count":3,"active_band_mae_db":1.9265160561833667}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"golf","status":"measured","mos":{"status":"measured","value":2.90887713432312,"predicted_mos":2.90887713432312,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89760,"input_sample_rate":24000,"input_seconds":3.74,"input_float32_sha256":"5fef52b76e6dfea4ba3facdb523efcb571e28424ed35133d9a1e75696723153d","internal_samples":59840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9791780299983657},"started_at":"2026-10-13T05:02:13.278951+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030150330003380077,"native_input_samples":90000,"native_input_sha256":"307f80f5cfa16704dfae7650b9864bb91ff0611b1d0001c4602d485bf4deb58e","output_sha256":"ee02ba939ea36853a32ee483979c60c35c1a95d26e85054c43bae8d1f79cfd1d","output_samples":89760,"length_delta_samples":-240,"analysis_s":0.4314869310001086,"synthesis_s":0.08783919799952855,"total_s":0.5193261289996371,"rtf":0.13848696773323657,"output_peak":0.5140223503112793,"output_rms":0.06461390929429837,"finished_at":"2026-10-13T05:02:13.807330+00:00","analysis_stages":{"external_f0_s":0.04312801899959595},"evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"ee02ba939ea36853a32ee483979c60c35c1a95d26e85054c43bae8d1f79cfd1d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":59840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.79042044110352,"output_peak":0.5116953499136772},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":0.9973333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.681338679510267,"spectral_convergence":0.6292386802137425,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.061507461331948,"spectral_convergence":0.26206038351179645,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.296845386480324,"spectral_convergence":0.23602814386233886,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015780590644672969,"output_power":0.0014631041886391132,"delta_db":-0.3284800056125424,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0022793999973787487,"output_power":0.002104378686097565,"delta_db":-0.3469664898188751,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003529050089255872,"output_power":0.0002939374417940451,"delta_db":-0.7940291243299917,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024181849661513886,"output_power":0.00019997027202045396,"delta_db":-0.8252407945500573,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012340626522142252,"output_power":9.529492594070799e-05,"delta_db":-1.1226743210897263,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.892617204459772e-05,"output_power":2.1213669383454182e-05,"delta_db":-2.6362590173546847,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0089416254593129,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4316810659654569,"pitch":null,"output_rms_dbfs":-23.79042044110352,"output_peak":0.5116953499136772,"reference_rms_dbfs":-23.35873937513806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.67989717577418},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.9241738670619956,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06793022158548279,"rms_dbfs":-23.35873937513806,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.167895928491422,"mean":2.2409306490485042e-05,"zero_energy":false},"output":{"rms":0.06463667058160824,"rms_dbfs":-23.79042044110352,"peak":0.5116953499136772,"peak_dbfs":-5.819770594151815,"crest_factor_db":17.9706498469517,"mean":0.0010130018797889085,"zero_energy":false},"rms_difference_db":-0.43168106596545996},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":87,"quiet_samples":13920,"reference_mean_power":3.5355661075091227e-07,"output_mean_power":4.3583166555203325e-07,"output_rms_dbfs":-63.60681219076186,"output_zero_energy":false,"output_peak":0.008821669232117699,"output_energy_fraction":2.4266574519328115e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":119,"both_estimated_voiced_frames":102,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9714285714285714,"absolute_difference_cents_median":6.8514171381298254,"absolute_difference_cents_p95":26.848708732457244,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00019859470734653442,"log_rms_mae_db":1.42590399740456,"rms_envelope_nrmse":0.12795733437101534,"envelope_correlation":0.9841733897107776,"rms_gain_error_db":-0.4316810659656197,"reference_mean_power":0.0046206924009106936,"output_mean_power":0.004183492087735793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.3746011573420917e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009649833083916325,"output_power":0.000840558989151627,"reference_active":true,"delta_db":-0.599516043143496},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.476156013523496e-05,"output_power":8.856059413211031e-05,"reference_active":true,"delta_db":-0.2939168078689274},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7397685942152586e-05,"output_power":1.7527603949671256e-05,"reference_active":true,"delta_db":0.03231064606326636}],"active_band_count":3,"active_band_mae_db":0.3085811656918966}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00016219647203140816,"log_rms_mae_db":1.243936801596938,"rms_envelope_nrmse":0.08481091269749684,"envelope_correlation":0.9944652557010307,"rms_gain_error_db":-0.43168106596626354,"reference_mean_power":0.00465810691427839,"output_mean_power":0.004217366517595376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.1766829233849286e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008484745917512814,"output_power":0.000744707148650596,"reference_active":true,"delta_db":-0.5665331877695847},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.056750595188828e-05,"output_power":3.707178214626994e-05,"reference_active":true,"delta_db":-0.3913484424612057},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2803005518232767e-06,"output_power":1.4030882798084283e-06,"reference_active":true,"delta_db":0.3977306424480377}],"active_band_count":3,"active_band_mae_db":0.4518707575596094}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00015290889340024391,"log_rms_mae_db":1.7003961946101998,"rms_envelope_nrmse":0.16692864136375735,"envelope_correlation":0.9793001420972614,"rms_gain_error_db":-0.5705305376858818,"reference_mean_power":0.0016186445519773775,"output_mean_power":0.001419379198110496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":8.864319553713053e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006871717731957776,"output_power":0.0004909655597185535,"reference_active":true,"delta_db":-1.4601428349422643},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.122547801813984e-05,"output_power":5.511817987580219e-05,"reference_active":true,"delta_db":-0.45637316793631805},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023392505393483e-05,"output_power":8.658063146665708e-06,"reference_active":true,"delta_db":-0.635939882314911}],"active_band_count":3,"active_band_mae_db":0.8508186283978311}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00012027897252950328,"log_rms_mae_db":1.382092836507416,"rms_envelope_nrmse":0.13685874622794156,"envelope_correlation":0.9862625399016134,"rms_gain_error_db":-0.5705305376864432,"reference_mean_power":0.0016317509855966273,"output_mean_power":0.0014308721470828656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":7.860028972773865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006296807053174762,"output_power":0.00045113506009180856,"reference_active":true,"delta_db":-1.4481380611937604},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2594213016438003e-05,"output_power":2.153467172531126e-05,"reference_active":true,"delta_db":-0.2085896317160592},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.035916073559823e-06,"output_power":8.069494811386131e-07,"reference_active":true,"delta_db":-1.0847822507392415}],"active_band_count":3,"active_band_mae_db":0.9138366478830203}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":1.761026302914818,"rms_envelope_nrmse":0.23331338338100271,"envelope_correlation":0.9676453848561516,"rms_gain_error_db":-0.8756203955700803,"reference_mean_power":0.00030771259821878297,"output_mean_power":0.000251526204682646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.053032330470538e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.925969099434094e-05,"output_power":4.7825256148123596e-05,"reference_active":true,"delta_db":-1.608232448397342},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0198944748844678e-05,"output_power":5.418480140954406e-06,"reference_active":true,"delta_db":-2.7467775307749194},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.310708162202101e-06,"output_power":1.8593760884010765e-06,"reference_active":true,"delta_db":-0.943778567955891}],"active_band_count":3,"active_band_mae_db":1.7662628490427175}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":1.4284839148026724,"rms_envelope_nrmse":0.16114077649922715,"envelope_correlation":0.9881238839680749,"rms_gain_error_db":-0.8756203955714287,"reference_mean_power":0.00031020419820436016,"output_mean_power":0.000253562854113196},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":9.132483962305987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.281780104005181e-05,"output_power":4.3283216153262334e-05,"reference_active":true,"delta_db":-1.6176320620482076},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.364498006941983e-06,"output_power":2.316959379303894e-06,"reference_active":true,"delta_db":-2.7501587965724505},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9272935717464496e-07,"output_power":1.2057957738227495e-07,"reference_active":true,"delta_db":-2.036741154904471}],"active_band_count":3,"active_band_mae_db":2.1348440045083765}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":2.1317751608584907,"rms_envelope_nrmse":0.2761981144920085,"envelope_correlation":0.9713979217720172,"rms_gain_error_db":-1.9739841241250478,"reference_mean_power":6.948491941831556e-05,"output_mean_power":4.410543865844542e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":7.889302863050732e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988966611609443e-05,"output_power":2.4968091474810293e-05,"reference_active":true,"delta_db":-2.034750544284657},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1396571746535427e-05,"output_power":6.673476852441171e-06,"reference_active":true,"delta_db":-2.3242207050962835},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7261232510360692e-06,"output_power":1.0730610007771973e-06,"reference_active":true,"delta_db":-2.0644739138090866}],"active_band_count":3,"active_band_mae_db":2.1411483877300093}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":2.0303035580808904,"rms_envelope_nrmse":0.24496066702636757,"envelope_correlation":0.9840573809915963,"rms_gain_error_db":-1.973984124136152,"reference_mean_power":7.004755034477965e-05,"output_mean_power":4.446256771629486e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":7.46074251190508e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0579981382752805e-05,"output_power":2.518126249166221e-05,"reference_active":true,"delta_db":-2.0723434327928785},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.072598753931927e-06,"output_power":3.4618729304463403e-06,"reference_active":true,"delta_db":-2.44063463262827},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2102140824794873e-07,"output_power":7.978535348180742e-08,"reference_active":true,"delta_db":-1.8093902874711878}],"active_band_count":3,"active_band_mae_db":2.1074561176307784}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.092407464981079,"predicted_mos":3.092407464981079,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90120,"input_sample_rate":24000,"input_seconds":3.755,"input_float32_sha256":"3010ff67e4819381793d6e916c3ae0189a10ee3769542f2dde01f3e62634919d","internal_samples":60080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.937618981999549},"started_at":"2026-10-13T05:07:24.436741+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00207000499995047,"native_input_samples":90000,"native_input_sha256":"307f80f5cfa16704dfae7650b9864bb91ff0611b1d0001c4602d485bf4deb58e","output_sha256":"bef6d709b95fa2f96bc5aa16db4861b34e90f086d6e86ad37c684ea81e7be777","output_samples":90120,"length_delta_samples":120,"analysis_s":1.2577673020005022,"synthesis_s":7.83819843099991,"total_s":9.095965733000412,"rtf":2.425590862133443,"output_peak":0.576437771320343,"output_rms":0.0695455084015161,"finished_at":"2026-10-13T05:07:33.541303+00:00","analysis_stages":{"external_f0_s":0.7705591660005666},"evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"bef6d709b95fa2f96bc5aa16db4861b34e90f086d6e86ad37c684ea81e7be777","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.151808893572934,"output_peak":0.5602465039210754},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986684420772304,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.3055479286943195,"spectral_convergence":0.5702313283936165,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.136279751446575,"spectral_convergence":0.2913822291067614,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.237244733267895,"spectral_convergence":0.2833252640019343,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0020822567280468487,"delta_db":1.216681890798192,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.0020567629216274213,"delta_db":-0.43296897304904314,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.00032719620206411666,"delta_db":-0.33231612695790735,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.00023099285894063434,"delta_db":-0.18776399720381542,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00011785886103984485,"delta_db":-0.18799633797805715,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":3.1092825542658945e-05,"delta_db":-0.9666481412199732,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5540625778678313,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.22431387273566133,"pitch":null,"output_rms_dbfs":-23.14602215767479,"output_peak":0.5602465039210754,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5596908044695965},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9236102468155423,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06783958749363862,"rms_dbfs":-23.37033603041045,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.17949258376381,"mean":2.234954833984375e-05,"zero_energy":false},"output":{"rms":0.06961436918669539,"rms_dbfs":-23.14602215767479,"peak":0.5602465039210754,"peak_dbfs":-5.032416897822239,"crest_factor_db":18.113605259852548,"mean":0.0001533518183402944,"zero_energy":false},"rms_difference_db":0.22431387273566017},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":3.4953892199237914e-07,"output_mean_power":9.850743557479173e-07,"output_rms_dbfs":-60.06530986687933,"output_zero_energy":false,"output_peak":0.0307643775782635,"output_energy_fraction":4.770046728391501e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":133,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9809523809523809,"absolute_difference_cents_median":6.370208142407366,"absolute_difference_cents_p95":33.282934920076244,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00019859470734653442,"log_rms_mae_db":1.6235125126273173,"rms_envelope_nrmse":0.1364918930505533,"envelope_correlation":0.9800233712415378,"rms_gain_error_db":0.2243138669838972,"reference_mean_power":0.004608354103444978,"output_mean_power":0.004852630564929881},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.3747785986813366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009662478923264104,"output_power":0.0009818941898988218,"reference_active":true,"delta_db":0.06976130609468742},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.594410683545364e-05,"output_power":9.023799625861616e-05,"reference_active":true,"delta_db":-0.26628861062262815},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744648631903194e-05,"output_power":1.908312124756312e-05,"reference_active":true,"delta_db":0.3894143543908299}],"active_band_count":3,"active_band_mae_db":0.24182142370271512}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00016219647203140816,"log_rms_mae_db":1.259188383781509,"rms_envelope_nrmse":0.09524758948698786,"envelope_correlation":0.9899602440456652,"rms_gain_error_db":0.22431386084590577,"reference_mean_power":0.004645568268479526,"output_mean_power":0.004891817345889073},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1771130170213754e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008502808342452142,"output_power":0.000863700681268862,"reference_active":true,"delta_db":0.06800872187908606},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1331740410000496e-05,"output_power":4.136683704007288e-05,"reference_active":true,"delta_db":0.003686223913132607},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2770439378675007e-06,"output_power":1.456420590047444e-06,"reference_active":true,"delta_db":0.570809702938367}],"active_band_count":3,"active_band_mae_db":0.21416821624352855}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015290889340024391,"log_rms_mae_db":1.9654500178308931,"rms_envelope_nrmse":0.16787625412753188,"envelope_correlation":0.9786260292364307,"rms_gain_error_db":-0.535315994323272,"reference_mean_power":0.0016143224036409894,"output_mean_power":0.001427114026096779},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":8.858385684123232e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006875363814202012,"output_power":0.0005960076302543327,"reference_active":true,"delta_db":-0.6204386434234326},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.173936122300265e-05,"output_power":5.521137849630995e-05,"reference_active":true,"delta_db":-0.48533541235019606},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0160658568889886e-05,"output_power":6.516892277467173e-06,"reference_active":true,"delta_db":-1.9288131562494046}],"active_band_count":3,"active_band_mae_db":1.0115290706743443}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012027897252950328,"log_rms_mae_db":1.6266985747159695,"rms_envelope_nrmse":0.12581819992947885,"envelope_correlation":0.9885957020480401,"rms_gain_error_db":-0.5353159944904163,"reference_mean_power":0.0016273586545452231,"output_mean_power":0.001438638500007199},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.855745033342305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006306037982733972,"output_power":0.000549583383636242,"reference_active":true,"delta_db":-0.5972298869110128},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.290965975606541e-05,"output_power":2.11645203050775e-05,"reference_active":true,"delta_db":-0.3441020967631894},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0400487875049354e-06,"output_power":6.195332345942161e-07,"reference_active":true,"delta_db":-2.2498910315858263}],"active_band_count":3,"active_band_mae_db":1.0637410050866762}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":1.8132872724762994,"rms_envelope_nrmse":0.1744683475863796,"envelope_correlation":0.9793246142802282,"rms_gain_error_db":-0.12828486489352944,"reference_mean_power":0.0003068909357402281,"output_mean_power":0.00029795835980288006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.0486708619890829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.911544017889025e-05,"output_power":6.186251730769593e-05,"reference_active":true,"delta_db":-0.4814748995296294},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0167211042917926e-05,"output_power":9.309709657553596e-06,"reference_active":true,"delta_db":-0.38265701565897264},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.330972389293411e-06,"output_power":1.9480267402638404e-06,"reference_active":true,"delta_db":-0.7794221519265507}],"active_band_count":3,"active_band_mae_db":0.547851355705051}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":1.3127523341199518,"rms_envelope_nrmse":0.10967798030450694,"envelope_correlation":0.9915331289113576,"rms_gain_error_db":-0.12828486539269376,"reference_mean_power":0.0003093691936331506,"output_mean_power":0.0003003644838044504},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.094392572662044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.275291186820285e-05,"output_power":5.592406638156824e-05,"reference_active":true,"delta_db":-0.5003514025996133},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.365294966430716e-06,"output_power":4.057668885873056e-06,"reference_active":true,"delta_db":-0.3173698983422146},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9190263566646987e-07,"output_power":1.4288369812038362e-07,"reference_active":true,"delta_db":-1.2809825743467842}],"active_band_count":3,"active_band_mae_db":0.6995679584295372}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":1.787179829498458,"rms_envelope_nrmse":0.2125216775712809,"envelope_correlation":0.9717469223922023,"rms_gain_error_db":-0.6860678277998888,"reference_mean_power":6.929937891252565e-05,"output_mean_power":5.917285986517186e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.858105660801018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988342411251764e-05,"output_power":3.407660575662068e-05,"reference_active":true,"delta_db":-0.683361069360467},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1366606898498144e-05,"output_power":8.786855673369842e-06,"reference_active":true,"delta_db":-1.117973476128998},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7366453289643982e-06,"output_power":1.6097826062390533e-06,"reference_active":true,"delta_db":-0.3294390203820055}],"active_band_count":3,"active_band_mae_db":0.7102578552904902}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":1.3185331357981855,"rms_envelope_nrmse":0.14263833782885077,"envelope_correlation":0.989017611620384,"rms_gain_error_db":-0.6860678308766621,"reference_mean_power":6.985899704640877e-05,"output_mean_power":5.9650702567449184e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.43114016941622e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0602300645347656e-05,"output_power":3.454935258009545e-05,"reference_active":true,"delta_db":-0.7011072912517496},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067895142462184e-06,"output_power":4.682944586532088e-06,"reference_active":true,"delta_db":-1.1251904847367113},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2311303789031625e-07,"output_power":1.0736657434858414e-07,"reference_active":true,"delta_db":-0.5943495125075153}],"active_band_count":3,"active_band_mae_db":0.8068824294986587}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.825834274291992,"predicted_mos":2.825834274291992,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90240,"input_sample_rate":24000,"input_seconds":3.76,"input_float32_sha256":"2b7b2862e3535569b8bc62f13e586b6d0efeef697f92eb18dded991327713325","internal_samples":60160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9015319170011935},"started_at":"2026-10-13T05:10:19.970079+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002106620000631665,"native_input_samples":90000,"native_input_sha256":"307f80f5cfa16704dfae7650b9864bb91ff0611b1d0001c4602d485bf4deb58e","output_sha256":"cc3195591725531460dce82a416e28ec8f97a4642ad6fa83498aa2da3fbf0fca","output_samples":90240,"length_delta_samples":240,"analysis_s":0.6959226729995862,"synthesis_s":0.08693417399990722,"total_s":0.7828568469994934,"rtf":0.20876182586653158,"output_peak":0.4979838728904724,"output_rms":0.06687217655925119,"finished_at":"2026-10-13T05:10:20.763122+00:00","analysis_stages":{"external_f0_s":0.08197526599906269},"evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"cc3195591725531460dce82a416e28ec8f97a4642ad6fa83498aa2da3fbf0fca","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.491940566679645,"output_peak":0.4935453734239598},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973404255319149,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":3.8109690037885344,"spectral_convergence":0.1585857922589787,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.419430801925781,"spectral_convergence":0.14420646536826695,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":3.7470137801352044,"spectral_convergence":0.16115196546845936,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.001564249180937589,"delta_db":-0.025601486951492336,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.002219862718794942,"delta_db":-0.10155014197075746,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.0003402272037620526,"delta_db":-0.1627083081931906,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.00022209271111154778,"delta_db":-0.358406481482478,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00010757113205834744,"delta_db":-0.5846613457548523,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":3.3063703595102e-05,"delta_db":-0.6997350523183615,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3221104694451887,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11003876428561156,"pitch":null,"output_rms_dbfs":-23.48037479469606,"output_peak":0.4935453734239598,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":3.6591378619498394},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9602364424279941,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06783958749363862,"rms_dbfs":-23.37033603041045,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.17949258376381,"mean":2.234954833984375e-05,"zero_energy":false},"output":{"rms":0.0669855704630451,"rms_dbfs":-23.48037479469606,"peak":0.4935453734239598,"peak_dbfs":-6.1334582977054435,"crest_factor_db":17.34691649699062,"mean":0.0002721532597451992,"zero_energy":false},"rms_difference_db":-0.11003876428561199},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":3.4953892199237914e-07,"output_mean_power":5.789589164155477e-07,"output_rms_dbfs":-62.373522532105035,"output_zero_energy":false,"output_peak":0.019377159190458515,"output_energy_fraction":3.0278658562899245e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":109,"both_estimated_voiced_frames":99,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9428571428571428,"absolute_difference_cents_median":1.5905545248341202,"absolute_difference_cents_p95":8.57378673720777,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00019859470734653442,"log_rms_mae_db":0.5984118471678452,"rms_envelope_nrmse":0.05663981036698317,"envelope_correlation":0.9964943233598886,"rms_gain_error_db":-0.11003876435610274,"reference_mean_power":0.004608354103444978,"output_mean_power":0.004493057393377921},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.3747785986813366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009662478923264104,"output_power":0.0009327396990824689,"reference_active":true,"delta_db":-0.15328098090167216},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.594410683545364e-05,"output_power":7.643858987396725e-05,"reference_active":true,"delta_db":-0.9870563760094155},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744648631903194e-05,"output_power":1.4594884622832802e-05,"reference_active":true,"delta_db":-0.7750730812033966}],"active_band_count":3,"active_band_mae_db":0.6384701460381614}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00016219647203140816,"log_rms_mae_db":0.3573029817286841,"rms_envelope_nrmse":0.03303677899810971,"envelope_correlation":0.998852613761921,"rms_gain_error_db":-0.11003876469739479,"reference_mean_power":0.004645568268479526,"output_mean_power":0.004529340494448989},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1771130170213754e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008502808342452142,"output_power":0.0008228380972218254,"reference_active":true,"delta_db":-0.14247998745683949},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1331740410000496e-05,"output_power":3.303968006685153e-05,"reference_active":true,"delta_db":-0.9724785981141953},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2770439378675007e-06,"output_power":7.84636251997675e-07,"reference_active":true,"delta_db":-2.1153747016029913}],"active_band_count":3,"active_band_mae_db":1.0767777623913422}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015290889340024391,"log_rms_mae_db":0.855049575156548,"rms_envelope_nrmse":0.0735053153801184,"envelope_correlation":0.995569472949943,"rms_gain_error_db":-0.08733285361784653,"reference_mean_power":0.0016143224036409894,"output_mean_power":0.001582184000924862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":8.858385684123232e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006875363814202012,"output_power":0.0006507359555507096,"reference_active":true,"delta_db":-0.23890880314650195},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.173936122300265e-05,"output_power":5.0673986519046996e-05,"reference_active":true,"delta_db":-0.8577706025512061},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0160658568889886e-05,"output_power":9.167263904138247e-06,"reference_active":true,"delta_db":-0.44682124050283795}],"active_band_count":3,"active_band_mae_db":0.514500215400182}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012027897252950328,"log_rms_mae_db":0.5727965968435472,"rms_envelope_nrmse":0.046065112670104415,"envelope_correlation":0.9981726484431206,"rms_gain_error_db":-0.08733285370126684,"reference_mean_power":0.0016273586545452231,"output_mean_power":0.0015949607223014247},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.855745033342305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006306037982733972,"output_power":0.0005966505288312073,"reference_active":true,"delta_db":-0.24036552386641236},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.290965975606541e-05,"output_power":1.8800156324071837e-05,"reference_active":true,"delta_db":-0.858571788363674},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0400487875049354e-06,"output_power":6.518257620229401e-07,"reference_active":true,"delta_db":-2.0292219100508295}],"active_band_count":3,"active_band_mae_db":1.0427197407603053}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":1.0179719498589217,"rms_envelope_nrmse":0.09956407894969194,"envelope_correlation":0.9946940667124204,"rms_gain_error_db":-0.43336688909297527,"reference_mean_power":0.0003068909357402281,"output_mean_power":0.0002777457182142571},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.0486708619890829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.911544017889025e-05,"output_power":6.0094879742468365e-05,"reference_active":true,"delta_db":-0.6073760784438335},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0167211042917926e-05,"output_power":8.348202153447868e-06,"reference_active":true,"delta_db":-0.8560888133987893},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.330972389293411e-06,"output_power":1.5168246335224475e-06,"reference_active":true,"delta_db":-1.866017561930169}],"active_band_count":3,"active_band_mae_db":1.1098274845909306}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":0.6869257028732997,"rms_envelope_nrmse":0.06763176109166115,"envelope_correlation":0.9984321302719831,"rms_gain_error_db":-0.4333668904741176,"reference_mean_power":0.0003093691936331506,"output_mean_power":0.0002799886175993546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.094392572662044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.275291186820285e-05,"output_power":5.436495153590898e-05,"reference_active":true,"delta_db":-0.623148775931153},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.365294966430716e-06,"output_power":3.579595447480804e-06,"reference_active":true,"delta_db":-0.8617964757888037},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9190263566646987e-07,"output_power":1.0774267691834772e-07,"reference_active":true,"delta_db":-2.5069317794996544}],"active_band_count":3,"active_band_mae_db":1.3306256770732037}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":0.9578481664641065,"rms_envelope_nrmse":0.10718132716850208,"envelope_correlation":0.9945835433392421,"rms_gain_error_db":-0.5526934049363473,"reference_mean_power":6.929937891252565e-05,"output_mean_power":6.1018285669970715e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.858105660801018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988342411251764e-05,"output_power":3.547103476152825e-05,"reference_active":true,"delta_db":-0.5091857883276579},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1366606898498144e-05,"output_power":9.821548346806338e-06,"reference_active":true,"delta_db":-0.6345088184894052},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7366453289643982e-06,"output_power":1.511111393534407e-06,"reference_active":true,"delta_db":-0.6041465245154638}],"active_band_count":3,"active_band_mae_db":0.5826137104441756}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":0.692453151571715,"rms_envelope_nrmse":0.07606193360664085,"envelope_correlation":0.9985309500362761,"rms_gain_error_db":-0.5526934141043487,"reference_mean_power":6.985899704640877e-05,"output_mean_power":6.151103078106219e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.43114016941622e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0602300645347656e-05,"output_power":3.613198288544793e-05,"reference_active":true,"delta_db":-0.506588468233647},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067895142462184e-06,"output_power":5.21785895267188e-06,"reference_active":true,"delta_db":-0.6554573203525542},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2311303789031625e-07,"output_power":1.0550895615937519e-07,"reference_active":true,"delta_db":-0.6701472151138904}],"active_band_count":3,"active_band_mae_db":0.6107310012333639}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_2300-131720-0041","method":"magphase","status":"measured","mos":{"status":"measured","value":2.967974901199341,"predicted_mos":2.967974901199341,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":181112,"input_sample_rate":48000,"input_seconds":3.7731666666666666,"input_float32_sha256":"eeb66a9bbf6b6997476b2b176758798ac1a1d31872b9b2240d454e86a22db73f","internal_samples":60371,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9122137800004566},"started_at":"2026-10-13T05:13:43.262679+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00705847800054471,"native_input_samples":180000,"native_input_sha256":"d1b8a95baeae1e5027ce951e1aa311b7e090bb7530d5f4e1e3c017fe80b531fa","output_sha256":"b575bbc00600a623f2dcd5b65cc2195226dd099b9f6e0f4f541b4f6a8d87a995","output_samples":181112,"length_delta_samples":1112,"analysis_s":2.190962786000455,"synthesis_s":0.8289153959995019,"total_s":3.0198781819999567,"rtf":0.8053008485333218,"output_peak":0.4776827095571222,"output_rms":0.06266284167195184,"finished_at":"2026-10-13T05:13:46.299107+00:00","evaluation_input":{"reference":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"output_sha256":"b575bbc00600a623f2dcd5b65cc2195226dd099b9f6e0f4f541b4f6a8d87a995","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60371,"length_delta_samples":371,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.05667743181929,"output_peak":0.4776251159091096},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9938546653194414,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.612768389445401,"spectral_convergence":0.5463175256181678,"spectral_convergence_status":"ok","amplitude_floor":1.419137961520426e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.66468504278346,"spectral_convergence":0.36141078117218317,"spectral_convergence_status":"ok","amplitude_floor":1.2001614819603722e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.257342401617858,"spectral_convergence":0.3511556593201676,"spectral_convergence_status":"ok","amplitude_floor":7.977266337453151e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0015734976003854988,"output_power":0.0013019249152998317,"delta_db":-0.8228014664800425,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002272380907829781,"output_power":0.001955281005836452,"delta_db":-0.6527195031757248,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00035321559220182435,"output_power":0.00031393772877974654,"delta_db":-0.5119635465065279,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024119869501752642,"output_power":0.00022936000107829914,"delta_db":-0.21857271785513607,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012307274117421543,"output_power":0.00011611356229724912,"delta_db":-0.25278924489174925,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.88440946982043e-05,"output_power":3.713434503273348e-05,"delta_db":-0.19549235889786423,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.442389806301174,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6595702086311985,"pitch":null,"output_rms_dbfs":-24.029906239041647,"output_peak":0.4776251159091096,"reference_rms_dbfs":-23.37033603041045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.844931944615573},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.8280245464432107,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06783958749363862,"rms_dbfs":-23.37033603041045,"peak":0.49029541015625,"peak_dbfs":-6.190843446646642,"crest_factor_db":17.17949258376381,"mean":2.234954833984375e-05,"zero_energy":false},"output":{"rms":0.0628788641621675,"rms_dbfs":-24.029906239041647,"peak":0.4776251159091096,"peak_dbfs":-6.4182568781960825,"crest_factor_db":17.61164936084557,"mean":-0.00047719045753039763,"zero_energy":false},"rms_difference_db":-0.6595702086311981},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.943985778605566e-06,"threshold_dbfs":-54.040646597258224,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":3.4953892199237914e-07,"output_mean_power":2.1895275155639367e-06,"output_rms_dbfs":-56.59649592704102,"output_zero_energy":false,"output_peak":0.020406976078083307,"output_energy_fraction":0.00012995483304219067,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":105,"output_estimated_voiced_frames":89,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":33,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.6857142857142857,"absolute_difference_cents_median":10.218464592065356,"absolute_difference_cents_p95":32.34394618333862,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00019859470734653442,"log_rms_mae_db":2.122234239712335,"rms_envelope_nrmse":0.19940089558792332,"envelope_correlation":0.9579304746021222,"rms_gain_error_db":-0.6595702086538401,"reference_mean_power":0.004608354103444978,"output_mean_power":0.003959030265324132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.3747785986813366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009662478923264104,"output_power":0.0008312954544168627,"reference_active":true,"delta_db":-0.6533315384715492},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.594410683545364e-05,"output_power":8.881143501333117e-05,"reference_active":true,"delta_db":-0.335494168282524},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.744648631903194e-05,"output_power":1.4836113217412259e-05,"reference_active":true,"delta_db":-0.703878353638832}],"active_band_count":3,"active_band_mae_db":0.5642346867976351}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00016219647203140816,"log_rms_mae_db":1.6930874427771745,"rms_envelope_nrmse":0.15116168292945542,"envelope_correlation":0.9768781939624483,"rms_gain_error_db":-0.6595702088385325,"reference_mean_power":0.004645568268479526,"output_mean_power":0.003991000899867657},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1771130170213754e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008502808342452142,"output_power":0.0007253007214026785,"reference_active":true,"delta_db":-0.6904428054991023},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1331740410000496e-05,"output_power":3.762769680537241e-05,"reference_active":true,"delta_db":-0.4077605714616325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2770439378675007e-06,"output_power":1.0447019035201334e-06,"reference_active":true,"delta_db":-0.8721345379895686}],"active_band_count":3,"active_band_mae_db":0.6567793049834345}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015290889340024391,"log_rms_mae_db":3.0881787621100822,"rms_envelope_nrmse":0.27050805619006346,"envelope_correlation":0.9403302461038644,"rms_gain_error_db":-0.7084365480538297,"reference_mean_power":0.0016143224036409894,"output_mean_power":0.001371344658225351},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":8.858385684123232e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006875363814202012,"output_power":0.0005912499596891067,"reference_active":true,"delta_db":-0.6552455994004378},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.173936122300265e-05,"output_power":5.389643375484774e-05,"reference_active":true,"delta_db":-0.5900210227565219},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0160658568889886e-05,"output_power":9.806419678821875e-06,"reference_active":true,"delta_db":-0.15411382385118844}],"active_band_count":3,"active_band_mae_db":0.4664601486693827}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012027897252950328,"log_rms_mae_db":2.553514185775806,"rms_envelope_nrmse":0.20543254003663566,"envelope_correlation":0.965242229847174,"rms_gain_error_db":-0.708436548088319,"reference_mean_power":0.0016273586545452231,"output_mean_power":0.0013824187738931772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.855745033342305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006306037982733972,"output_power":0.000542690632107632,"reference_active":true,"delta_db":-0.6520425739697335},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.290965975606541e-05,"output_power":1.9394525607488056e-05,"reference_active":true,"delta_db":-0.7233947812263842},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0400487875049354e-06,"output_power":9.283394442044639e-07,"reference_active":true,"delta_db":-0.493469084576662}],"active_band_count":3,"active_band_mae_db":0.6229688132575932}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":9.957022642296084e-05,"log_rms_mae_db":3.0119679689886407,"rms_envelope_nrmse":0.24176630952054926,"envelope_correlation":0.9606092810224817,"rms_gain_error_db":-0.2479609471302186,"reference_mean_power":0.0003068909357402281,"output_mean_power":0.0002898597856334979},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.0486708619890829e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.911544017889025e-05,"output_power":6.099402557737473e-05,"reference_active":true,"delta_db":-0.5428778073987535},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0167211042917926e-05,"output_power":9.103106841719077e-06,"reference_active":true,"delta_db":-0.4801219838519136},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.330972389293411e-06,"output_power":2.709291949398213e-06,"reference_active":true,"delta_db":0.6531867723287632}],"active_band_count":3,"active_band_mae_db":0.5587288545264768}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":8.985216395929295e-05,"log_rms_mae_db":2.365451033623622,"rms_envelope_nrmse":0.1644217090639357,"envelope_correlation":0.981325603582291,"rms_gain_error_db":-0.247960947671768,"reference_mean_power":0.0003093691936331506,"output_mean_power":0.00029220051064928366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.094392572662044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.275291186820285e-05,"output_power":5.5200880165164425e-05,"reference_active":true,"delta_db":-0.5568788029702799},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.365294966430716e-06,"output_power":3.906001607641141e-06,"reference_active":true,"delta_db":-0.4828117695465014},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9190263566646987e-07,"output_power":2.1964545991492134e-07,"reference_active":true,"delta_db":0.5864129125253439}],"active_band_count":3,"active_band_mae_db":0.5420344950140418}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.809631559896483e-05,"log_rms_mae_db":3.020924265308229,"rms_envelope_nrmse":0.3022069643709271,"envelope_correlation":0.9394799664139778,"rms_gain_error_db":-0.19502010198876557,"reference_mean_power":6.929937891252565e-05,"output_mean_power":6.625632326937259e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":7.858105660801018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.988342411251764e-05,"output_power":4.35890875363832e-05,"reference_active":true,"delta_db":0.38585341160178255},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1366606898498144e-05,"output_power":1.0113285857053807e-05,"reference_active":true,"delta_db":-0.5073855769742561},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7366453289643982e-06,"output_power":1.4376926580719898e-06,"reference_active":true,"delta_db":-0.8204507763730445}],"active_band_count":3,"active_band_mae_db":0.5712299216496944}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.651257368418167e-05,"log_rms_mae_db":2.429486914285988,"rms_envelope_nrmse":0.2063738089678404,"envelope_correlation":0.9708198944002607,"rms_gain_error_db":-0.1950201061021553,"reference_mean_power":6.985899704640877e-05,"output_mean_power":6.679136753937707e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":7.43114016941622e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0602300645347656e-05,"output_power":4.4237521042056685e-05,"reference_active":true,"delta_db":0.37240139549957946},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067895142462184e-06,"output_power":5.200687667510114e-06,"reference_active":true,"delta_db":-0.6697729469312019},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2311303789031625e-07,"output_power":6.953101649700808e-08,"reference_active":true,"delta_db":-2.481254693999417}],"active_band_count":3,"active_band_mae_db":1.1744763454767329}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},{"case":"librispeech_237-126133-0003","method":"hnm","status":"measured","mos":{"status":"measured","value":2.3816981315612793,"predicted_mos":2.3816981315612793,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"0fac2ada9abcc6d5ab12f12fb843772ca0afb06c9cc16e36a3205a1d9a9b5bb8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.902367167000193},"started_at":"2026-10-13T04:38:38.422550+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4125999769021291e-05,"native_input_samples":96000,"native_input_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","output_sha256":"e9afa9adbcc982caa3341df335c67c714b068c8a48fa09e51e4bfb0a3db16e76","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0707055120001314,"synthesis_s":0.5156246730002749,"total_s":1.5863301850004063,"rtf":0.2643883641667344,"output_peak":0.2361700122934427,"output_rms":0.030060041197432438,"finished_at":"2026-10-13T04:38:40.014428+00:00","evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"e9afa9adbcc982caa3341df335c67c714b068c8a48fa09e51e4bfb0a3db16e76","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.65776985481447,"spectral_convergence":0.23932140969413185,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.470228474210614,"spectral_convergence":0.2372497675711963,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.805164748306881,"spectral_convergence":0.24060773161430046,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":3.3442962164812446e-05,"delta_db":0.24300952976677934,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.0002974879861715342,"delta_db":-0.1030952494969839,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.00039924528088504526,"delta_db":-0.20070042742854535,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":0.00011542011598942025,"delta_db":0.0043569167886361995,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":5.462974851624227e-05,"delta_db":-0.18306803105302663,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":3.3799830642809102e-06,"delta_db":0.5187469129201748,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2088295112423577,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12350431627678615,"pitch":null,"output_rms_dbfs":-30.440208570920127,"output_peak":0.2361700122934427,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.644387692443988},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.65776985481447,"spectral_convergence":0.23932140969413185,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.470228474210614,"spectral_convergence":0.2372497675711963,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.805164748306881,"spectral_convergence":0.24060773161430046,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":3.3442962164812446e-05,"delta_db":0.24300952976677934,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.0002974879861715342,"delta_db":-0.1030952494969839,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.00039924528088504526,"delta_db":-0.20070042742854535,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":0.00011542011598942025,"delta_db":0.0043569167886361995,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":5.462974851624227e-05,"delta_db":-0.18306803105302663,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":3.3799830642809102e-06,"delta_db":0.5187469129201748,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2088295112423577,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12350431627678615,"pitch":null,"output_rms_dbfs":-30.440208570920127,"output_peak":0.2361700122934427,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.644387692443988},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8968095875087768,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03049051692298232,"rms_dbfs":-30.31670425464334,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.682069594728176,"mean":-0.00010436312357584635,"zero_energy":false},"output":{"rms":0.030060041197432438,"rms_dbfs":-30.440208570920127,"peak":0.2361700122934427,"peak_dbfs":-12.535504955980771,"crest_factor_db":17.904703614939354,"mean":-2.0696728180224726e-05,"zero_energy":false},"rms_difference_db":-0.1235043162767866},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":8.385859118948277e-07,"output_rms_dbfs":-60.76452437946916,"output_zero_energy":false,"output_peak":0.017179061767659088,"output_energy_fraction":0.00010672502359695275,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":252,"both_estimated_voiced_frames":229,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.9015748031496063,"absolute_difference_cents_median":1.729760843033959,"absolute_difference_cents_p95":6.603975605365772,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":0.9676773971977828,"rms_envelope_nrmse":0.07830117035848569,"envelope_correlation":0.9938451734642316,"rms_gain_error_db":-0.12301934617045981,"reference_mean_power":0.0009300339703354519,"output_mean_power":0.0009040592149031845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3130295372770493e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016320463803082337,"output_power":0.0001546893033618874,"reference_active":true,"delta_db":-0.23272212959096},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1951400558986024e-05,"output_power":2.7678017042427194e-05,"reference_active":true,"delta_db":-0.6235492738964263},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7431760484027967e-06,"output_power":3.4861180808102305e-06,"reference_active":true,"delta_db":-0.30898159933520764}],"active_band_count":3,"active_band_mae_db":0.3884176676075313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":0.5017316372106704,"rms_envelope_nrmse":0.044574976098200304,"envelope_correlation":0.9979566530741624,"rms_gain_error_db":-0.12321659918146771,"reference_mean_power":0.0009316560347730994,"output_mean_power":0.0009055948446747924},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0476212509594803e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014768349458157853,"output_power":0.0001394842552083031,"reference_active":true,"delta_db":-0.24806772548147438},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4986865819787958e-05,"output_power":1.3177367288055512e-05,"reference_active":true,"delta_db":-0.5588216796538258},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.979624065594547e-07,"output_power":3.258475740684849e-07,"reference_active":true,"delta_db":-0.868275564422635}],"active_band_count":3,"active_band_mae_db":0.5583883231859784}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":1.533345237371479,"rms_envelope_nrmse":0.10307757547548796,"envelope_correlation":0.9910003536117244,"rms_gain_error_db":-0.20225550105702472,"reference_mean_power":0.0005508019269041907,"output_mean_power":0.000525738644168796},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0420364913089255e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.387221376866869e-05,"output_power":8.447632939706869e-05,"reference_active":true,"delta_db":-0.4580202490466365},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.257079739763329e-05,"output_power":1.8214598563566796e-05,"reference_active":true,"delta_db":-0.9312729827119103},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1515320396484854e-06,"output_power":2.647363292197069e-06,"reference_active":true,"delta_db":0.9006572494071642}],"active_band_count":3,"active_band_mae_db":0.763316827055237}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":0.8928348728743176,"rms_envelope_nrmse":0.061814479855006065,"envelope_correlation":0.9968028239298015,"rms_gain_error_db":-0.202762730319963,"reference_mean_power":0.0005517580103658142,"output_mean_power":0.0005265897166828843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.88495020416125e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.537091697949908e-05,"output_power":7.64205952945986e-05,"reference_active":true,"delta_db":-0.4809952986159648},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0155340206288246e-05,"output_power":8.32344527084685e-06,"reference_active":true,"delta_db":-0.8639134868938951},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.398901151759037e-07,"output_power":2.3717775236010519e-07,"reference_active":true,"delta_db":-0.04938403848571618}],"active_band_count":3,"active_band_mae_db":0.4647642746651921}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":1.5817649766136173,"rms_envelope_nrmse":0.1340072465829181,"envelope_correlation":0.9853438091366306,"rms_gain_error_db":-0.06072501252302957,"reference_mean_power":0.00014468505517057894,"output_mean_power":0.00014267608141569634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.678016723697028e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.684581121209276e-05,"output_power":3.4977658984493446e-05,"reference_active":true,"delta_db":-0.22597383292320042},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.802139012131185e-06,"output_power":6.859126093411221e-06,"reference_active":true,"delta_db":-0.559448972484969},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9028025525875444e-06,"output_power":1.2424462645776145e-06,"reference_active":true,"delta_db":-1.8511611064174143}],"active_band_count":3,"active_band_mae_db":0.8788613039418612}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":0.869118843914532,"rms_envelope_nrmse":0.07702663827805446,"envelope_correlation":0.9947340243202798,"rms_gain_error_db":-0.059929556116661224,"reference_mean_power":0.00014482485587598114,"output_mean_power":0.0001428401012509903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.245552781666509e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6110400828216855e-05,"output_power":3.398263375919841e-05,"reference_active":true,"delta_db":-0.263752741528931},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660726810319972e-06,"output_power":3.2416567520436947e-06,"reference_active":true,"delta_db":-0.5280029293458799},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4925257777124552e-07,"output_power":8.959981972593275e-08,"reference_active":true,"delta_db":-2.2161470478715977}],"active_band_count":3,"active_band_mae_db":1.0026342395821362}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":1.6805424708299896,"rms_envelope_nrmse":0.21167518710893024,"envelope_correlation":0.9608902273995145,"rms_gain_error_db":-0.08337633562995614,"reference_mean_power":8.01766615945731e-06,"output_mean_power":7.865210231933926e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.89214453053621e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.737362730986326e-06,"output_power":2.624108504196119e-06,"reference_active":true,"delta_db":-0.18350561362780715},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.579169756744006e-07,"output_power":7.195063152555901e-07,"reference_active":true,"delta_db":-0.22587024234876604},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8094260096505427e-07,"output_power":1.6731908854114324e-07,"reference_active":true,"delta_db":-0.33995338678429887}],"active_band_count":3,"active_band_mae_db":0.24977641425362404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":1.0109432420748379,"rms_envelope_nrmse":0.11553932094263852,"envelope_correlation":0.9870512901041556,"rms_gain_error_db":-0.08344805913054426,"reference_mean_power":8.04858299788563e-06,"output_mean_power":7.895408793329148e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.450576990854381e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.750016160596198e-06,"output_power":2.656525578211585e-06,"reference_active":true,"delta_db":-0.1502124414260895},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5474696245295724e-07,"output_power":3.489340024438818e-07,"reference_active":true,"delta_db":-0.07175393421396352},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2784081824019621e-08,"output_power":1.0188882077558069e-08,"reference_active":true,"delta_db":-0.9854300576154993}],"active_band_count":3,"active_band_mae_db":0.40246547775185076}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-126133-0003","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.7633655071258545,"predicted_mos":2.7633655071258545,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"88a90f6d4e2d6c6636f20a4db5d6d8d3249606969c11a104f19602eaf74d9677","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.399116644999594},"started_at":"2026-10-13T04:41:35.411276+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1226000171736814e-05,"native_input_samples":96000,"native_input_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","output_sha256":"d26e59fb650a67a70ded8faca58cf2f2cc1f131cb2bd4286d65daab8be740162","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3217364890006138,"synthesis_s":0.10433024699977977,"total_s":1.4260667360003936,"rtf":0.23767778933339892,"output_peak":0.2928557232132781,"output_rms":0.03504394781412128,"finished_at":"2026-10-13T04:41:36.842617+00:00","evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"d26e59fb650a67a70ded8faca58cf2f2cc1f131cb2bd4286d65daab8be740162","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.425420077685156,"spectral_convergence":0.5307917440693961,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.182108739387664,"spectral_convergence":0.2658209754585544,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.655093500702689,"spectral_convergence":0.2920262192658459,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":3.394466510040195e-05,"delta_db":0.30767743726531005,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.0003779988573112147,"delta_db":0.9371153025646437,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.000576411786757699,"delta_db":1.3942301872036134,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":0.00015735775290724228,"delta_db":1.3504233043337557,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":7.913487124131864e-05,"delta_db":1.4263189568087034,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":3.2303450809791543e-06,"delta_db":0.3220908531379184,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9563093402189908,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.2089647532582992,"pitch":null,"output_rms_dbfs":-29.10773950138504,"output_peak":0.2928557232132781,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.754207439258503},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.425420077685156,"spectral_convergence":0.5307917440693961,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.182108739387664,"spectral_convergence":0.2658209754585544,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.655093500702689,"spectral_convergence":0.2920262192658459,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":3.394466510040195e-05,"delta_db":0.30767743726531005,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.0003779988573112147,"delta_db":0.9371153025646437,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.000576411786757699,"delta_db":1.3942301872036134,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":0.00015735775290724228,"delta_db":1.3504233043337557,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":7.913487124131864e-05,"delta_db":1.4263189568087034,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":3.2303450809791543e-06,"delta_db":0.3220908531379184,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9563093402189908,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.2089647532582992,"pitch":null,"output_rms_dbfs":-29.10773950138504,"output_peak":0.2928557232132781,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.754207439258503},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9388946487743035,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03049051692298232,"rms_dbfs":-30.31670425464334,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.682069594728176,"mean":-0.00010436312357584635,"zero_energy":false},"output":{"rms":0.03504394781412128,"rms_dbfs":-29.10773950138504,"peak":0.2928557232132781,"peak_dbfs":-10.666925684681638,"crest_factor_db":18.440813816703404,"mean":1.6817844001839816e-06,"zero_energy":false},"rms_difference_db":1.2089647532583},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":1.0361846600458867e-06,"output_rms_dbfs":-59.845628414126665,"output_zero_energy":false,"output_peak":0.018984353514594446,"output_energy_fraction":9.703065187394816e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":300,"both_estimated_voiced_frames":251,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.9881889763779528,"absolute_difference_cents_median":4.936099177419228,"absolute_difference_cents_p95":25.04393219257222,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":1.8161910473712752,"rms_envelope_nrmse":0.1869755608145197,"envelope_correlation":0.9891146532799882,"rms_gain_error_db":1.2108615527313211,"reference_mean_power":0.0009300339703354519,"output_mean_power":0.0012290936278723961},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3130295372770493e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016320463803082337,"output_power":0.00020995546809025002,"reference_active":true,"delta_db":1.0939469331703846},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1951400558986024e-05,"output_power":3.983136138109575e-05,"reference_active":true,"delta_db":0.9573525045692401},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7431760484027967e-06,"output_power":4.6109098232455234e-06,"reference_active":true,"delta_db":0.9054637543603339}],"active_band_count":3,"active_band_mae_db":0.9855877306999862}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":1.4035686497472175,"rms_envelope_nrmse":0.16548306403813035,"envelope_correlation":0.9955205126450892,"rms_gain_error_db":1.2125760383225463,"reference_mean_power":0.0009316560347730994,"output_mean_power":0.001231723437272081},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0476212509594803e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014768349458157853,"output_power":0.00019076068627414436,"reference_active":true,"delta_db":1.1115691580964095},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4986865819787958e-05,"output_power":1.8723758318472977e-05,"reference_active":true,"delta_db":0.9668220772208103},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.979624065594547e-07,"output_power":3.5750902958339627e-07,"reference_active":true,"delta_db":-0.4655503324396858}],"active_band_count":3,"active_band_mae_db":0.8479805225856353}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":1.9362418551426146,"rms_envelope_nrmse":0.21292450809936433,"envelope_correlation":0.9888345401834548,"rms_gain_error_db":1.3908414241965925,"reference_mean_power":0.0005508019269041907,"output_mean_power":0.0007587166125900812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0420364913089255e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.387221376866869e-05,"output_power":0.00012557376926136473,"reference_active":true,"delta_db":1.263618705043642},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.257079739763329e-05,"output_power":3.0115839402391672e-05,"reference_active":true,"delta_db":1.2524807010710268},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1515320396484854e-06,"output_power":2.92520147810792e-06,"reference_active":true,"delta_db":1.3340796641677153}],"active_band_count":3,"active_band_mae_db":1.2833930234274613}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":1.5397805315356157,"rms_envelope_nrmse":0.19431174953466177,"envelope_correlation":0.9939783838800033,"rms_gain_error_db":1.392173877808608,"reference_mean_power":0.0005517580103658142,"output_mean_power":0.0007602668156777238},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.88495020416125e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.537091697949908e-05,"output_power":0.00011441859391291348,"reference_active":true,"delta_db":1.2718666017192617},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0155340206288246e-05,"output_power":1.350589160897645e-05,"reference_active":true,"delta_db":1.2382878303223288},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.398901151759037e-07,"output_power":2.5119553378017825e-07,"reference_active":true,"delta_db":0.1999956048409487}],"active_band_count":3,"active_band_mae_db":0.9033833456275131}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":1.673075439371183,"rms_envelope_nrmse":0.231194574834715,"envelope_correlation":0.9859360531484171,"rms_gain_error_db":1.4326655775476607,"reference_mean_power":0.00014468505517057894,"output_mean_power":0.00020122884384863863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.678016723697028e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.684581121209276e-05,"output_power":5.496082132504971e-05,"reference_active":true,"delta_db":1.7366509155713823},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.802139012131185e-06,"output_power":1.2364991202619774e-05,"reference_active":true,"delta_db":1.9998012775926983},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9028025525875444e-06,"output_power":3.2190866334662962e-06,"reference_active":true,"delta_db":2.283389394165226}],"active_band_count":3,"active_band_mae_db":2.006613862443102}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":1.2092286973786923,"rms_envelope_nrmse":0.2037636353643246,"envelope_correlation":0.9937459014729173,"rms_gain_error_db":1.4370602387465885,"reference_mean_power":0.00014482485587598114,"output_mean_power":0.00020162720452420368},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.245552781666509e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6110400828216855e-05,"output_power":5.454545610594621e-05,"reference_active":true,"delta_db":1.7912626849453797},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660726810319972e-06,"output_power":5.728435402901518e-06,"reference_active":true,"delta_db":1.9446870018190088},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4925257777124552e-07,"output_power":2.4044108734364087e-07,"reference_active":true,"delta_db":2.070868426519783}],"active_band_count":3,"active_band_mae_db":1.9356060377613904}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":1.6689568505783565,"rms_envelope_nrmse":0.22132625586590524,"envelope_correlation":0.9742557762308596,"rms_gain_error_db":0.9901927436839177,"reference_mean_power":8.01766615945731e-06,"output_mean_power":1.0070875879529617e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.89214453053621e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.737362730986326e-06,"output_power":3.3461036762883595e-06,"reference_active":true,"delta_db":0.8720704302529413},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.579169756744006e-07,"output_power":9.471716096903057e-07,"reference_active":true,"delta_db":0.9680703771176925},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8094260096505427e-07,"output_power":1.5363303806463736e-07,"reference_active":true,"delta_db":-0.7105621007385645}],"active_band_count":3,"active_band_mae_db":0.8502343027030661}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":1.068446242455617,"rms_envelope_nrmse":0.16479773019482746,"envelope_correlation":0.9899270297053611,"rms_gain_error_db":0.9914504920207725,"reference_mean_power":8.04858299788563e-06,"output_mean_power":1.0112638347845362e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.450576990854381e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.750016160596198e-06,"output_power":3.3402080475536634e-06,"reference_active":true,"delta_db":0.8443827204707217},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5474696245295724e-07,"output_power":4.4691307090253117e-07,"reference_active":true,"delta_db":1.0030437122486808},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2784081824019621e-08,"output_power":9.328768738510356e-09,"reference_active":true,"delta_db":-1.3684521467586124}],"active_band_count":3,"active_band_mae_db":1.0719595264926716}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-126133-0003","method":"facodec","status":"measured","mos":{"status":"measured","value":4.608487606048584,"predicted_mos":4.608487606048584,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"eb39630035b70fa5483258e88ab8a6ab909218f6da2d0c2699f1a9e5fa49c774","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.864629767000224},"started_at":"2026-10-13T04:48:46.223654+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00011141099912492791,"native_input_samples":96000,"native_input_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","output_sha256":"1785de465436764729122abeb85f5c84b8cbfe2159840e754baf461fbf902250","output_samples":96000,"length_delta_samples":0,"analysis_s":2.0883317729994815,"synthesis_s":6.053890990000582,"total_s":8.142222763000063,"rtf":1.3570371271666772,"output_peak":0.21478137373924255,"output_rms":0.027718134656702656,"finished_at":"2026-10-13T04:48:54.374046+00:00","evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"1785de465436764729122abeb85f5c84b8cbfe2159840e754baf461fbf902250","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.083555104003563,"spectral_convergence":0.22726474577795855,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.505537342052572,"spectral_convergence":0.2392568666070331,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6740938047528,"spectral_convergence":0.25838209765728787,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":2.9135867430689037e-05,"delta_db":-0.35575832137213936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.00024649000513928774,"delta_db":-0.9197964283909242,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.00035543352587820974,"delta_db":-0.7055144443115049,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":8.818150191833835e-05,"delta_db":-1.164683232195416,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":4.670057453194499e-05,"delta_db":-0.8641378089451879,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":2.353513948630869e-06,"delta_db":-1.0532305632387138,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.843853466408981,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.828015739951717,"pitch":null,"output_rms_dbfs":-31.144719994595057,"output_peak":0.21478137373924255,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.421062083602979},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.083555104003563,"spectral_convergence":0.22726474577795855,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.505537342052572,"spectral_convergence":0.2392568666070331,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6740938047528,"spectral_convergence":0.25838209765728787,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":2.9135867430689037e-05,"delta_db":-0.35575832137213936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.00024649000513928774,"delta_db":-0.9197964283909242,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.00035543352587820974,"delta_db":-0.7055144443115049,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":8.818150191833835e-05,"delta_db":-1.164683232195416,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":4.670057453194499e-05,"delta_db":-0.8641378089451879,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":2.353513948630869e-06,"delta_db":-1.0532305632387138,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.843853466408981,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.828015739951717,"pitch":null,"output_rms_dbfs":-31.144719994595057,"output_peak":0.21478137373924255,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.421062083602979},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.921397720876399,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03049051692298232,"rms_dbfs":-30.31670425464334,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.682069594728176,"mean":-0.00010436312357584635,"zero_energy":false},"output":{"rms":0.027718134656702656,"rms_dbfs":-31.144719994595057,"peak":0.21478137373924255,"peak_dbfs":-13.360067684125106,"crest_factor_db":17.78465231046995,"mean":-1.5943758237777198e-05,"zero_energy":false},"rms_difference_db":-0.8280157399517165},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":4.150651284296848e-07,"output_rms_dbfs":-63.81883752207428,"output_zero_energy":false,"output_peak":0.0041346061043441296,"output_energy_fraction":6.212781608928734e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":279,"both_estimated_voiced_frames":245,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9645669291338582,"absolute_difference_cents_median":3.0085696058302935,"absolute_difference_cents_p95":14.557456512482252,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":1.3224894521461839,"rms_envelope_nrmse":0.14396832894421102,"envelope_correlation":0.9862817846739818,"rms_gain_error_db":-0.8285135116234789,"reference_mean_power":0.0009300339703354519,"output_mean_power":0.0007685063506206164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3130295372770493e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016320463803082337,"output_power":0.00013308201256782422,"reference_active":true,"delta_db":-0.8861313661120683},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1951400558986024e-05,"output_power":2.8582953997018574e-05,"reference_active":true,"delta_db":-0.4838278946354874},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7431760484027967e-06,"output_power":3.9745315054479644e-06,"reference_active":true,"delta_db":0.2604569065620988}],"active_band_count":3,"active_band_mae_db":0.5434720557698848}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":1.1165592164548341,"rms_envelope_nrmse":0.13654833066466487,"envelope_correlation":0.9875119676638312,"rms_gain_error_db":-0.825798346671536,"reference_mean_power":0.0009316560347730994,"output_mean_power":0.0007703281469049466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0476212509594803e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014768349458157853,"output_power":0.00012046260245130812,"reference_active":true,"delta_db":-0.8847971900616718},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4986865819787958e-05,"output_power":1.3279773811417569e-05,"reference_active":true,"delta_db":-0.525201410594732},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.979624065594547e-07,"output_power":3.776029952750592e-07,"reference_active":true,"delta_db":-0.2280661785753824}],"active_band_count":3,"active_band_mae_db":0.5460215930772621}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":1.5128079116997994,"rms_envelope_nrmse":0.14373132871176744,"envelope_correlation":0.9875551522656686,"rms_gain_error_db":-0.7867597237104375,"reference_mean_power":0.0005508019269041907,"output_mean_power":0.000459535935347187},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0420364913089255e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.387221376866869e-05,"output_power":8.412492174064847e-05,"reference_active":true,"delta_db":-0.47612386643308735},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.257079739763329e-05,"output_power":2.095391016135921e-05,"reference_active":true,"delta_db":-0.32281824854711977},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1515320396484854e-06,"output_power":2.053414702508463e-06,"reference_active":true,"delta_db":-0.2027115052683395}],"active_band_count":3,"active_band_mae_db":0.3338845400828489}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":1.224499534357857,"rms_envelope_nrmse":0.13502069005471742,"envelope_correlation":0.9889561937978882,"rms_gain_error_db":-0.7838728252386387,"reference_mean_power":0.0005517580103658142,"output_mean_power":0.0004606396996883068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.88495020416125e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.537091697949908e-05,"output_power":7.680762825929285e-05,"reference_active":true,"delta_db":-0.45905591506823046},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0155340206288246e-05,"output_power":9.254927472120866e-06,"reference_active":true,"delta_db":-0.40321457275344424},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.398901151759037e-07,"output_power":2.533741678739598e-07,"reference_active":true,"delta_db":0.2374998244533892}],"active_band_count":3,"active_band_mae_db":0.36659010409168796}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":1.5624798995435962,"rms_envelope_nrmse":0.20581533080849243,"envelope_correlation":0.9743526723707863,"rms_gain_error_db":-1.1221693691122627,"reference_mean_power":0.00014468505517057894,"output_mean_power":0.00011173950350501486},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.678016723697028e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.684581121209276e-05,"output_power":3.117296867476105e-05,"reference_active":true,"delta_db":-0.7260995943859374},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.802139012131185e-06,"output_power":7.527689377511089e-06,"reference_active":true,"delta_db":-0.15551993882190365},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9028025525875444e-06,"output_power":2.9158751711500215e-06,"reference_active":true,"delta_db":1.853752025130213}],"active_band_count":3,"active_band_mae_db":0.9117905194460181}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":1.2244668849338476,"rms_envelope_nrmse":0.18760847010540438,"envelope_correlation":0.9792231693134354,"rms_gain_error_db":-1.1196842251382026,"reference_mean_power":0.00014482485587598114,"output_mean_power":0.00011191149114645577},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.245552781666509e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6110400828216855e-05,"output_power":3.148581571435191e-05,"reference_active":true,"delta_db":-0.5951735994844464},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660726810319972e-06,"output_power":3.6862344818039317e-06,"reference_active":true,"delta_db":0.030156373612934467},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4925257777124552e-07,"output_power":2.0856123686233718e-07,"reference_active":true,"delta_db":1.4531175307694988}],"active_band_count":3,"active_band_mae_db":0.6928158346222931}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":1.8308653265112496,"rms_envelope_nrmse":0.23381441553283228,"envelope_correlation":0.9537184342278867,"rms_gain_error_db":-0.4531886700341906,"reference_mean_power":8.01766615945731e-06,"output_mean_power":7.223191053082359e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.89214453053621e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.737362730986326e-06,"output_power":2.3487577610995893e-06,"reference_active":true,"delta_db":-0.6649412190064381},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.579169756744006e-07,"output_power":7.59848831554862e-07,"reference_active":true,"delta_db":0.011055655593291972},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8094260096505427e-07,"output_power":2.0907958822865681e-07,"reference_active":true,"delta_db":0.6277080730111955}],"active_band_count":3,"active_band_mae_db":0.4345683158703085}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":1.4407838715568142,"rms_envelope_nrmse":0.19673050007955387,"envelope_correlation":0.9638526264049115,"rms_gain_error_db":-0.4518808587185351,"reference_mean_power":8.04858299788563e-06,"output_mean_power":7.253228194125629e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.450576990854381e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.750016160596198e-06,"output_power":2.3997822826643387e-06,"reference_active":true,"delta_db":-0.5916340333031562},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5474696245295724e-07,"output_power":3.544663134724667e-07,"reference_active":true,"delta_db":-0.0034371694405144074},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2784081824019621e-08,"output_power":1.4621123260932442e-08,"reference_active":true,"delta_db":0.5831119670946825}],"active_band_count":3,"active_band_mae_db":0.39272772327945105}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-126133-0003","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.680866003036499,"predicted_mos":3.680866003036499,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"f3b860b121e5fbcd3fd11b2163aa5f257f80086e1a1febdcc74baec1a9cfa5e6","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6706821020015923},"started_at":"2026-10-13T07:07:07.512918+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009118422998653841,"native_input_samples":264600,"native_input_sha256":"df82911448eaae21da033bae27c3de55a49c66fb8b99fb390712310e92e3efbf","output_sha256":"1974eff1f83310178135d615d8eadcace88176b54559498a27449bbb098bf2f3","output_samples":264704,"length_delta_samples":104,"analysis_s":0.2078702959988732,"synthesis_s":36.26262556100119,"total_s":36.47049585700006,"rtf":6.078415976166677,"output_peak":0.22631853818893433,"output_rms":0.02943475285298625,"finished_at":"2026-10-13T07:07:44.048788+00:00","evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"1974eff1f83310178135d615d8eadcace88176b54559498a27449bbb098bf2f3","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.615139853297116,"output_peak":0.22559892670838877},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.021502183100587,"spectral_convergence":0.39061792478705576,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.176248332050495,"spectral_convergence":0.34294643983511053,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.158005591638897,"spectral_convergence":0.37273407732706343,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":1.6831357995565902e-05,"delta_db":-2.738846271943343,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.00026803936729988796,"delta_db":-0.5558037258854533,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.0004595395702824397,"delta_db":0.4101307961291841,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":8.49461432145671e-05,"delta_db":-1.327021492260267,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":3.7272384209721545e-05,"delta_db":-1.8434883014344456,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":1.3797634195951544e-06,"delta_db":-3.372352062071962,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.7079404416207755,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2980555325646299,"pitch":null,"output_rms_dbfs":-30.61475978720797,"output_peak":0.22559892670838877,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.118585368929994},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7926065486167736,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03049051692298232,"rms_dbfs":-30.31670425464334,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.682069594728176,"mean":-0.00010436312357584635,"zero_energy":false},"output":{"rms":0.029461985446024803,"rms_dbfs":-30.61475978720797,"peak":0.22559892670838877,"peak_dbfs":-12.93325941697719,"crest_factor_db":17.681500370230783,"mean":-1.0758168037521383e-05,"zero_energy":false},"rms_difference_db":-0.2980555325646286},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":3.449518549992384e-07,"output_rms_dbfs":-64.62241515266547,"output_zero_energy":false,"output_peak":0.003856215574750655,"output_energy_fraction":4.5701694597795685e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":273,"both_estimated_voiced_frames":240,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":33,"paired_voiced_fraction":0.9448818897637795,"absolute_difference_cents_median":6.679957781371641,"absolute_difference_cents_p95":32.09179853193763,"difference_over_600_cents_frames":4,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":1.5215379284434036,"rms_envelope_nrmse":0.09715019899473229,"envelope_correlation":0.9921398863827393,"rms_gain_error_db":-0.29677662626915025,"reference_mean_power":0.0009300339703354519,"output_mean_power":0.000868602648419514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3130295372770493e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016320463803082337,"output_power":0.00015600315977705296,"reference_active":true,"delta_db":-0.1959910169856288},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1951400558986024e-05,"output_power":2.558384457620564e-05,"reference_active":true,"delta_db":-0.9652409175528175},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7431760484027967e-06,"output_power":3.4926532306765955e-06,"reference_active":true,"delta_db":-0.30084784561753586}],"active_band_count":3,"active_band_mae_db":0.48735992671866074}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":1.1438251296233464,"rms_envelope_nrmse":0.07363308452024608,"envelope_correlation":0.9961142479335465,"rms_gain_error_db":-0.29387931219423,"reference_mean_power":0.0009316560347730994,"output_mean_power":0.000870698247273147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0476212509594803e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014768349458157853,"output_power":0.00014224281421455767,"reference_active":true,"delta_db":-0.1630162431229632},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4986865819787958e-05,"output_power":1.1517708871465135e-05,"reference_active":true,"delta_db":-1.1434472216223488},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.979624065594547e-07,"output_power":3.2135624000657015e-07,"reference_active":true,"delta_db":-0.9285531112999303}],"active_band_count":3,"active_band_mae_db":0.7450055253484141}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":2.523879634236946,"rms_envelope_nrmse":0.12878750922434212,"envelope_correlation":0.9872011831441847,"rms_gain_error_db":0.2037865837248547,"reference_mean_power":0.0005508019269041907,"output_mean_power":0.0005772635147185203},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0420364913089255e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.387221376866869e-05,"output_power":0.00010406300205536918,"reference_active":true,"delta_db":0.4475929059742697},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.257079739763329e-05,"output_power":1.9959751590423866e-05,"reference_active":true,"delta_db":-0.533917704593891},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1515320396484854e-06,"output_power":2.2439180025073966e-06,"reference_active":true,"delta_db":0.18259165053454074}],"active_band_count":3,"active_band_mae_db":0.3880340870342338}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":1.8190076079711723,"rms_envelope_nrmse":0.10633175139878483,"envelope_correlation":0.99104565447466,"rms_gain_error_db":0.20720271262619205,"reference_mean_power":0.0005517580103658142,"output_mean_power":0.0005787205686894836},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.88495020416125e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.537091697949908e-05,"output_power":9.532786057007301e-05,"reference_active":true,"delta_db":0.4790990006375525},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0155340206288246e-05,"output_power":8.81958970576786e-06,"reference_active":true,"delta_db":-0.6124609507019092},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.398901151759037e-07,"output_power":2.389765700150233e-07,"reference_active":true,"delta_db":-0.01657029465273971}],"active_band_count":3,"active_band_mae_db":0.3693767486640671}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":3.1085325185491004,"rms_envelope_nrmse":0.2917294236453244,"envelope_correlation":0.9389802259946771,"rms_gain_error_db":-1.3693241446092677,"reference_mean_power":0.00014468505517057894,"output_mean_power":0.0001055580259311959},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.678016723697028e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.684581121209276e-05,"output_power":2.4443919761111087e-05,"reference_active":true,"delta_db":-1.7821727311691},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.802139012131185e-06,"output_power":6.188815374637224e-06,"reference_active":true,"delta_db":-1.0060615697916373},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9028025525875444e-06,"output_power":1.1869222880209206e-06,"reference_active":true,"delta_db":-2.0497144025189966}],"active_band_count":3,"active_band_mae_db":1.612649567826578}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":2.493381920098946,"rms_envelope_nrmse":0.2585165364109403,"envelope_correlation":0.9520523381222926,"rms_gain_error_db":-1.362960385961867,"reference_mean_power":0.00014482485587598114,"output_mean_power":0.00010581495860562974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.245552781666509e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6110400828216855e-05,"output_power":2.3862717760488797e-05,"reference_active":true,"delta_db":-1.7991240455862334},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660726810319972e-06,"output_power":2.8321049576486367e-06,"reference_active":true,"delta_db":-1.1145797567813955},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4925257777124552e-07,"output_power":1.0809524054888466e-07,"reference_active":true,"delta_db":-1.4011526833617418}],"active_band_count":3,"active_band_mae_db":1.4382854952431237}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":4.530918043778193,"rms_envelope_nrmse":0.3353911939802363,"envelope_correlation":0.9378321401534377,"rms_gain_error_db":-2.2220037253700444,"reference_mean_power":8.01766615945731e-06,"output_mean_power":4.806706409876634e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.89214453053621e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.737362730986326e-06,"output_power":1.866377613304048e-06,"reference_active":true,"delta_db":-1.663328334974247},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.579169756744006e-07,"output_power":6.578179151309805e-07,"reference_active":true,"delta_db":-0.6151593745518721},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8094260096505427e-07,"output_power":6.016711056527939e-08,"reference_active":true,"delta_db":-4.781816731500143}],"active_band_count":3,"active_band_mae_db":2.353434813675421}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":4.048082967727781,"rms_envelope_nrmse":0.2980370132856226,"envelope_correlation":0.9588656865200259,"rms_gain_error_db":-2.2201363235243483,"reference_mean_power":8.04858299788563e-06,"output_mean_power":4.827316728465607e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.450576990854381e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.750016160596198e-06,"output_power":1.8867874790987315e-06,"reference_active":true,"delta_db":-1.6361226042550425},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5474696245295724e-07,"output_power":3.2078797913638853e-07,"reference_active":true,"delta_db":-0.43700599834024145},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2784081824019621e-08,"output_power":5.263605475679938e-09,"reference_active":true,"delta_db":-3.8538621161948954}],"active_band_count":3,"active_band_mae_db":1.9756635729300598}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-126133-0003","method":"golf","status":"measured","mos":{"status":"measured","value":3.3967747688293457,"predicted_mos":3.3967747688293457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"666380a1aef207546dad923a721df773f4ed846d6886060fbf7cf5509c43f4f0","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.285641623000629},"started_at":"2026-10-13T05:02:13.807955+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029934779995528515,"native_input_samples":144000,"native_input_sha256":"7f10d3ec84a9e138d9a285b3cf1a74a268bc738430bc31ed9ad28b5e50293476","output_sha256":"a5d9d03f20e2725ea72dd29ded120b73f6d3599768c42bc269cdc57207c7887f","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6650166110002829,"synthesis_s":0.13405453999985184,"total_s":0.7990711510001347,"rtf":0.13317852516668913,"output_peak":0.19545690715312958,"output_rms":0.02514241242561139,"finished_at":"2026-10-13T05:02:14.617531+00:00","analysis_stages":{"external_f0_s":0.09271793900006742},"evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"a5d9d03f20e2725ea72dd29ded120b73f6d3599768c42bc269cdc57207c7887f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.98536933067746,"output_peak":0.19297148173638198},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.788703987634797,"spectral_convergence":0.4472524868149964,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.815594298273312,"spectral_convergence":0.2950541741198589,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.0262433304109315,"spectral_convergence":0.3104220919744034,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.152980972993541e-05,"output_power":2.8568154587904792e-05,"delta_db":-0.42839162714581047,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030472316086000925,"output_power":0.00020772098973750113,"delta_db":-1.6642508133508764,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00041810653004353685,"output_power":0.00027979061382198145,"delta_db":-1.7445380941496513,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011501197186389244,"output_power":8.634408039511858e-05,"delta_db":-1.2451048100516533,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.706960082249132e-05,"output_power":2.9459882044061024e-05,"delta_db":-2.871738312389252,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.0062461417271605e-06,"output_power":1.202800187795567e-06,"delta_db":-3.9783104910081866,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.988722358015905,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.667617123122934,"pitch":null,"output_rms_dbfs":-31.98536933067746,"output_peak":0.19297148173638198,"reference_rms_dbfs":-30.317752207554527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.210180538773013},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8991609060128091,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030486838462877593,"rms_dbfs":-30.317752207554527,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.68311754763936,"mean":-0.00010344229874905442,"zero_energy":false},"output":{"rms":0.025161210638090584,"rms_dbfs":-31.98536933067746,"peak":0.19297148173638198,"peak_dbfs":-14.290137367995758,"crest_factor_db":17.6952319626817,"mean":0.0001541616968807394,"zero_energy":false},"rms_difference_db":-1.667617123122934},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":5.291443791974678e-07,"output_rms_dbfs":-62.76425812778449,"output_zero_energy":false,"output_peak":0.0041141672700244734,"output_energy_fraction":9.627940299048344e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":271,"both_estimated_voiced_frames":235,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.9288537549407114,"absolute_difference_cents_median":3.6138454788351924,"absolute_difference_cents_p95":14.100848604129757,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":2.365994909779046,"rms_envelope_nrmse":0.20408856878184656,"envelope_correlation":0.9866764466675229,"rms_gain_error_db":-1.6687155061362895,"reference_mean_power":0.0009297483403830102,"output_mean_power":0.0006331314122583071},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.309604459173917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016315954592213363,"output_power":0.0001056026420351542,"reference_active":true,"delta_db":-1.8893770419851819},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.170228798122289e-05,"output_power":2.212666273247671e-05,"reference_active":true,"delta_db":-1.5617469061883218},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732466055197301e-06,"output_power":2.754500298042356e-06,"reference_active":true,"delta_db":-1.3195304303577713}],"active_band_count":3,"active_band_mae_db":1.5902181261770918}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":2.0210319826452037,"rms_envelope_nrmse":0.18747235950861926,"envelope_correlation":0.9940558221624348,"rms_gain_error_db":-1.668889171722192,"reference_mean_power":0.0009309232436910003,"output_mean_power":0.0006339061378973419},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.0432318592179023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014758384292890757,"output_power":9.574286243499163e-05,"reference_active":true,"delta_db":-1.879324072398469},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4782360824005484e-05,"output_power":1.0412876488383267e-05,"reference_active":true,"delta_db":-1.5217308173830235},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.956204013316136e-07,"output_power":2.931234275873836e-07,"reference_active":true,"delta_db":-1.3022814881334277}],"active_band_count":3,"active_band_mae_db":1.5677787926383067}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":2.4166375169232164,"rms_envelope_nrmse":0.22678498208893344,"envelope_correlation":0.9817172854232405,"rms_gain_error_db":-1.7610145808069095,"reference_mean_power":0.0005506683294299569,"output_mean_power":0.0003671035984127002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0425988985450017e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.395924604659824e-05,"output_power":5.9447955968431824e-05,"reference_active":true,"delta_db":-1.9880259660095623},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.244438681388383e-05,"output_power":1.4642540929852007e-05,"reference_active":true,"delta_db":-1.8549129819675758},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.142648258367857e-06,"output_power":1.7557646168951401e-06,"reference_active":true,"delta_db":-0.8648458955687821}],"active_band_count":3,"active_band_mae_db":1.5692616145153064}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":2.1620047290849085,"rms_envelope_nrmse":0.2070469099187983,"envelope_correlation":0.9897656925801759,"rms_gain_error_db":-1.7613445022743521,"reference_mean_power":0.0005512759658678031,"output_mean_power":0.0003674807623039184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8845492484962513e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.541529186488066e-05,"output_power":5.426853343200718e-05,"reference_active":true,"delta_db":-1.9698754397101834},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.004122554046304e-05,"output_power":6.765239768787441e-06,"reference_active":true,"delta_db":-1.7150352890911948},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.396798065135247e-07,"output_power":1.9065675563169652e-07,"reference_active":true,"delta_db":-0.9937924698061198}],"active_band_count":3,"active_band_mae_db":1.559567732869166}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":2.3665107905981664,"rms_envelope_nrmse":0.24829110569731597,"envelope_correlation":0.9731691675747125,"rms_gain_error_db":-1.778451610774057,"reference_mean_power":0.00014452012770466692,"output_mean_power":9.595843924215974e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.663061405119821e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6804894238014375e-05,"output_power":2.0574379887321863e-05,"reference_active":true,"delta_db":-2.5257881965986675},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.770092944869329e-06,"output_power":4.977878278757205e-06,"reference_active":true,"delta_db":-1.933819409626478},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.89042950847152e-06,"output_power":9.509669753448941e-07,"reference_active":true,"delta_db":-2.983950524930292}],"active_band_count":3,"active_band_mae_db":2.481186043718479}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":2.0163251341623836,"rms_envelope_nrmse":0.21778685212038867,"envelope_correlation":0.985973071625535,"rms_gain_error_db":-1.776752803444213,"reference_mean_power":0.00014466631800174327,"output_mean_power":9.609308752694517e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.230802636041004e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.606306943338458e-05,"output_power":2.0087356965660884e-05,"reference_active":true,"delta_db":-2.541398906954023},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.629583324194086e-06,"output_power":2.406419199586715e-06,"reference_active":true,"delta_db":-1.7848548699551339},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4766974994325807e-07,"output_power":7.795090338304844e-08,"reference_active":true,"delta_db":-2.7747038681884306}],"active_band_count":3,"active_band_mae_db":2.3669858816991955}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":3.116689522387293,"rms_envelope_nrmse":0.3139966215610285,"envelope_correlation":0.9590491660253468,"rms_gain_error_db":-2.491533805635978,"reference_mean_power":8.024837436989596e-06,"output_mean_power":4.5215034118562085e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.878459766112128e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.730801358967967e-06,"output_power":1.526633291630003e-06,"reference_active":true,"delta_db":-2.5255538139919205},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.533362975091236e-07,"output_power":4.89245117117032e-07,"reference_active":true,"delta_db":-1.874623933627442},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8094939044779397e-07,"output_power":9.133516068632465e-08,"reference_active":true,"delta_db":-2.969191273007275}],"active_band_count":3,"active_band_mae_db":2.456456340208879}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":2.9189723080094203,"rms_envelope_nrmse":0.28657606079157344,"envelope_correlation":0.9747657299045598,"rms_gain_error_db":-2.490648676757055,"reference_mean_power":8.053065418987068e-06,"output_mean_power":4.5383330037551195e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.437206232252696e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.742725116574554e-06,"output_power":1.5500632671269472e-06,"reference_active":true,"delta_db":-2.4783285897864458},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5134081867774285e-07,"output_power":2.2612342298217884e-07,"reference_active":true,"delta_db":-1.9138305788910805},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2706619589138297e-08,"output_power":7.190890101661251e-09,"reference_active":true,"delta_db":-2.472473767219834}],"active_band_count":3,"active_band_mae_db":2.288210978632453}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-126133-0003","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.489304780960083,"predicted_mos":3.489304780960083,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"151dbc9a2340e973f0b8e747bcb1397d2276206e11aeba58693864312809de59","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2994595599993772},"started_at":"2026-10-13T05:07:33.541995+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004961708000337239,"native_input_samples":144000,"native_input_sha256":"7f10d3ec84a9e138d9a285b3cf1a74a268bc738430bc31ed9ad28b5e50293476","output_sha256":"297777b4c631be317a11579e8af27cfa4fbcc025459676973e11e962c54b0a5d","output_samples":144120,"length_delta_samples":120,"analysis_s":2.413856031999785,"synthesis_s":3.6297185600005832,"total_s":6.043574592000368,"rtf":1.0072624320000614,"output_peak":0.22379957139492035,"output_rms":0.02953540227744242,"finished_at":"2026-10-13T05:07:39.601256+00:00","analysis_stages":{"external_f0_s":1.61956112200005},"evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"297777b4c631be317a11579e8af27cfa4fbcc025459676973e11e962c54b0a5d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.587199474243988,"output_peak":0.21698449037419726},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.183922374161455,"spectral_convergence":0.46915939444858185,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.449202286701223,"spectral_convergence":0.27471506817490254,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.448538143504652,"spectral_convergence":0.2700716294216848,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":2.702458451689549e-05,"delta_db":-0.6824475881752169,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.00029226883107586084,"delta_db":-0.17996454093848355,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.00038076469010736894,"delta_db":-0.40653166537118346,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":0.00010973114647568953,"delta_db":-0.21515897979955872,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":6.155813468569022e-05,"delta_db":0.3354944712661744,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":2.3526752264955813e-06,"delta_db":-1.0547785350074101,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.47906263009300454,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2696720796150901,"pitch":null,"output_rms_dbfs":-30.58637633425843,"output_peak":0.21698449037419726,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.693887601455777},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.908259881055313,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03049051692298232,"rms_dbfs":-30.31670425464334,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.682069594728176,"mean":-0.00010436312357584635,"zero_energy":false},"output":{"rms":0.02955841778729032,"rms_dbfs":-30.58637633425843,"peak":0.21698449037419726,"peak_dbfs":-13.271426151197677,"crest_factor_db":17.314950183060752,"mean":3.299105960782248e-05,"zero_energy":false},"rms_difference_db":-0.26967207961508777},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":2.8837986827693565e-06,"output_rms_dbfs":-55.40035060878432,"output_zero_energy":false,"output_peak":0.03949425550807309,"output_energy_fraction":0.0003795774578817334,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":288,"both_estimated_voiced_frames":241,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":47,"paired_voiced_fraction":0.9488188976377953,"absolute_difference_cents_median":6.222093025560296,"absolute_difference_cents_p95":23.70245783895373,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":2.3110884993508205,"rms_envelope_nrmse":0.15206414145983294,"envelope_correlation":0.976748898882179,"rms_gain_error_db":-0.2686421794089644,"reference_mean_power":0.0009300339703354519,"output_mean_power":0.0008742478921975784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3130295372770493e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016320463803082337,"output_power":0.00016171477343841174,"reference_active":true,"delta_db":-0.03982799938828695},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1951400558986024e-05,"output_power":3.025902635686718e-05,"reference_active":true,"delta_db":-0.2363495015276139},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7431760484027967e-06,"output_power":3.244557751935541e-06,"reference_active":true,"delta_db":-0.6208474441132481}],"active_band_count":3,"active_band_mae_db":0.2990083150097163}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":1.7125499723444113,"rms_envelope_nrmse":0.100797840828659,"envelope_correlation":0.9892936326358479,"rms_gain_error_db":-0.26514505379938835,"reference_mean_power":0.0009316560347730994,"output_mean_power":0.0008764781543565903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0476212509594803e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014768349458157853,"output_power":0.00014650913333503373,"reference_active":true,"delta_db":-0.034672610374080294},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4986865819787958e-05,"output_power":1.3972732501992503e-05,"reference_active":true,"delta_db":-0.3042947413106659},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.979624065594547e-07,"output_power":3.411267832203876e-07,"reference_active":true,"delta_db":-0.669262295354046}],"active_band_count":3,"active_band_mae_db":0.33607654901293077}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":2.3958952645925424,"rms_envelope_nrmse":0.1644993643076181,"envelope_correlation":0.9767750406665023,"rms_gain_error_db":-0.3411391070413445,"reference_mean_power":0.0005508019269041907,"output_mean_power":0.000509191962997114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0420364913089255e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.387221376866869e-05,"output_power":8.841702372894542e-05,"reference_active":true,"delta_db":-0.26001168154990156},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.257079739763329e-05,"output_power":2.0916700294007625e-05,"reference_active":true,"delta_db":-0.33053728781672037},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1515320396484854e-06,"output_power":1.8803855503712873e-06,"reference_active":true,"delta_db":-0.5850091254712302}],"active_band_count":3,"active_band_mae_db":0.39185269827928404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":1.7692640755942242,"rms_envelope_nrmse":0.11142187015790263,"envelope_correlation":0.9891118807581479,"rms_gain_error_db":-0.337596108313879,"reference_mean_power":0.0005517580103658142,"output_mean_power":0.0005104921122685709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.88495020416125e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.537091697949908e-05,"output_power":8.056761558834419e-05,"reference_active":true,"delta_db":-0.25149435438488293},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0155340206288246e-05,"output_power":9.312485817713557e-06,"reference_active":true,"delta_db":-0.3762885259622528},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.398901151759037e-07,"output_power":2.0235530146008966e-07,"reference_active":true,"delta_db":-0.738977660993238}],"active_band_count":3,"active_band_mae_db":0.45558684711345787}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":2.283278159180266,"rms_envelope_nrmse":0.21193607214898605,"envelope_correlation":0.9635028458045102,"rms_gain_error_db":-0.04095279013509431,"reference_mean_power":0.00014468505517057894,"output_mean_power":0.00014332712690517962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.678016723697028e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.684581121209276e-05,"output_power":4.0868584564810595e-05,"reference_active":true,"delta_db":0.45001474130469044},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.802139012131185e-06,"output_power":8.505992544740113e-06,"reference_active":true,"delta_db":0.37511313782887346},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9028025525875444e-06,"output_power":2.1099027842195436e-06,"reference_active":true,"delta_db":0.44868719877500474}],"active_band_count":3,"active_band_mae_db":0.4246050259695229}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":1.521586107998493,"rms_envelope_nrmse":0.13931446455179386,"envelope_correlation":0.9827458042045016,"rms_gain_error_db":-0.03697274845997152,"reference_mean_power":0.00014482485587598114,"output_mean_power":0.00014359715319600962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.245552781666509e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6110400828216855e-05,"output_power":4.0256411085194845e-05,"reference_active":true,"delta_db":0.47202745168216764},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660726810319972e-06,"output_power":3.914790519960441e-06,"reference_active":true,"delta_db":0.2914120803956389},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4925257777124552e-07,"output_power":1.7603969653333733e-07,"reference_active":true,"delta_db":0.7168877057805272}],"active_band_count":3,"active_band_mae_db":0.49344241261944455}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":2.170282569756381,"rms_envelope_nrmse":0.24183823380606723,"envelope_correlation":0.9493772337282742,"rms_gain_error_db":-0.21238583344625836,"reference_mean_power":8.01766615945731e-06,"output_mean_power":7.63500605372737e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.89214453053621e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.737362730986326e-06,"output_power":2.5631170228442274e-06,"reference_active":true,"delta_db":-0.2856391508462189},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.579169756744006e-07,"output_power":7.051168743872413e-07,"reference_active":true,"delta_db":-0.31360526369038055},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8094260096505427e-07,"output_power":1.4139176156492093e-07,"reference_active":true,"delta_db":-1.071167235247694}],"active_band_count":3,"active_band_mae_db":0.5568038832614312}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":1.2946610972519146,"rms_envelope_nrmse":0.13848973056579825,"envelope_correlation":0.9818487592088194,"rms_gain_error_db":-0.2107629795374863,"reference_mean_power":8.04858299788563e-06,"output_mean_power":7.667311875125544e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.450576990854381e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.750016160596198e-06,"output_power":2.5666057624870487e-06,"reference_active":true,"delta_db":-0.2997608098949989},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5474696245295724e-07,"output_power":3.2541138592637065e-07,"reference_active":true,"delta_db":-0.37485940927246897},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2784081824019621e-08,"output_power":9.095659620598973e-09,"reference_active":true,"delta_db":-1.4783534192579506}],"active_band_count":3,"active_band_mae_db":0.7176578794751395}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-126133-0003","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.435497760772705,"predicted_mos":3.435497760772705,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"d2184ea774feedbfbb98996feeb7873064e304015e53281af11d55176fb37301","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5128838010004984},"started_at":"2026-10-13T05:10:20.764988+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027478029987832997,"native_input_samples":144000,"native_input_sha256":"7f10d3ec84a9e138d9a285b3cf1a74a268bc738430bc31ed9ad28b5e50293476","output_sha256":"9116eb76c6091d24d8c7235cb76a147fa3629260d0177343220693e4926b6fff","output_samples":144240,"length_delta_samples":240,"analysis_s":0.8433537559994875,"synthesis_s":0.16304177900019567,"total_s":1.0063955349996832,"rtf":0.16773258916661385,"output_peak":0.22856631875038147,"output_rms":0.0299043148862736,"finished_at":"2026-10-13T05:10:21.783320+00:00","analysis_stages":{"external_f0_s":0.16305969699897105},"evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"9116eb76c6091d24d8c7235cb76a147fa3629260d0177343220693e4926b6fff","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.479200093343287,"output_peak":0.227521798382585},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.5890532694274695,"spectral_convergence":0.16439699566782592,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.3758076572085916,"spectral_convergence":0.15713915277479099,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.825168629898313,"spectral_convergence":0.17566791410601948,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.162305217059986e-05,"output_power":3.201875190059529e-05,"delta_db":0.054006144635703156,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000304634409424632,"output_power":0.00029006299140476774,"delta_db":-0.2128663519698284,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004181285613649292,"output_power":0.0004073830785270879,"delta_db":-0.11306848437286403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001153043826104485,"output_power":0.00011113346312819096,"delta_db":-0.16000966737320824,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.6981779049220075e-05,"output_power":5.3856898660937265e-05,"delta_db":-0.2449466335191405,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.999437610841641e-06,"output_power":2.499703975699056e-06,"delta_db":-0.791512519792063,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2627349669438012,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15558973810694537,"pitch":null,"output_rms_dbfs":-30.472293992750288,"output_peak":0.227521798382585,"reference_rms_dbfs":-30.31670425464334,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.596676518844791},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9544196682209222,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03049051692298232,"rms_dbfs":-30.31670425464334,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.682069594728176,"mean":-0.00010436312357584635,"zero_energy":false},"output":{"rms":0.029949205124631912,"rms_dbfs":-30.472293992750288,"peak":0.227521798382585,"peak_dbfs":-12.859539763543623,"crest_factor_db":17.61275422920666,"mean":-3.990984155818324e-06,"zero_energy":false},"rms_difference_db":-0.1555897381069471},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":7.647806987458158e-07,"output_rms_dbfs":-61.16463081158162,"output_zero_energy":false,"output_peak":0.011628121791177245,"output_energy_fraction":9.805373890248223e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":262,"both_estimated_voiced_frames":237,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.9330708661417323,"absolute_difference_cents_median":0.8877519797868466,"absolute_difference_cents_p95":5.121789147036988,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":0.7508139421274495,"rms_envelope_nrmse":0.06572497648234994,"envelope_correlation":0.9958231951920722,"rms_gain_error_db":-0.15527513749055333,"reference_mean_power":0.0009300339703354519,"output_mean_power":0.0008973694866671031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3130295372770493e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016320463803082337,"output_power":0.00016076325782714442,"reference_active":true,"delta_db":-0.06545698112418809},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1951400558986024e-05,"output_power":2.8841713456650703e-05,"reference_active":true,"delta_db":-0.44468841989469093},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7431760484027967e-06,"output_power":3.2279815849715205e-06,"reference_active":true,"delta_db":-0.6430920476838753}],"active_band_count":3,"active_band_mae_db":0.3844124829009181}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":0.42574178207246527,"rms_envelope_nrmse":0.049679707778372766,"envelope_correlation":0.997538396086282,"rms_gain_error_db":-0.15501684928187073,"reference_mean_power":0.0009316560347730994,"output_mean_power":0.0008989880452313038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0476212509594803e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014768349458157853,"output_power":0.0001456576503885642,"reference_active":true,"delta_db":-0.05998660311047346},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4986865819787958e-05,"output_power":1.3485974648918612e-05,"reference_active":true,"delta_db":-0.4582848005648532},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.979624065594547e-07,"output_power":3.1939200896870397e-07,"reference_active":true,"delta_db":-0.9551800236477658}],"active_band_count":3,"active_band_mae_db":0.4911504757743641}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":0.9649904959545685,"rms_envelope_nrmse":0.08021017815782903,"envelope_correlation":0.9945746868071921,"rms_gain_error_db":-0.15613872565315917,"reference_mean_power":0.0005508019269041907,"output_mean_power":0.0005313510930892927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0420364913089255e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.387221376866869e-05,"output_power":9.175430690910734e-05,"reference_active":true,"delta_db":-0.09910600856598116},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.257079739763329e-05,"output_power":2.013984662403869e-05,"reference_active":true,"delta_db":-0.49490743581910046},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1515320396484854e-06,"output_power":1.885746763250194e-06,"reference_active":true,"delta_db":-0.5726444683190982}],"active_band_count":3,"active_band_mae_db":0.38888597090139326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":0.5949348076082374,"rms_envelope_nrmse":0.060891535977575316,"envelope_correlation":0.9967168996092629,"rms_gain_error_db":-0.1562774868484716,"reference_mean_power":0.0005517580103658142,"output_mean_power":0.0005322564073740218},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.88495020416125e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.537091697949908e-05,"output_power":8.31913726364648e-05,"reference_active":true,"delta_db":-0.1123165615206381},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0155340206288246e-05,"output_power":9.09604001087156e-06,"reference_active":true,"delta_db":-0.47842114931173163},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.398901151759037e-07,"output_power":2.0129158852558872e-07,"reference_active":true,"delta_db":-0.7618672577487801}],"active_band_count":3,"active_band_mae_db":0.4508683228603833}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":1.0087877303078614,"rms_envelope_nrmse":0.08004911714253589,"envelope_correlation":0.9950922287807731,"rms_gain_error_db":-0.1949777135871017,"reference_mean_power":0.00014468505517057894,"output_mean_power":0.00013833303420011537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.678016723697028e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.684581121209276e-05,"output_power":3.558750494011486e-05,"reference_active":true,"delta_db":-0.15090582122766155},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.802139012131185e-06,"output_power":7.3325120305604465e-06,"reference_active":true,"delta_db":-0.2696089974203112},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9028025525875444e-06,"output_power":1.497372011227679e-06,"reference_active":true,"delta_db":-1.04064014283286}],"active_band_count":3,"active_band_mae_db":0.48705165382694426}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":0.5913630877000853,"rms_envelope_nrmse":0.05085975587794761,"envelope_correlation":0.9981041219589902,"rms_gain_error_db":-0.19402713195379023,"reference_mean_power":0.00014482485587598114,"output_mean_power":0.00013849700815914796},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.245552781666509e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6110400828216855e-05,"output_power":3.4826361927406795e-05,"reference_active":true,"delta_db":-0.15724200111528672},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660726810319972e-06,"output_power":3.371810059429056e-06,"reference_active":true,"delta_db":-0.3570421792786592},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4925257777124552e-07,"output_power":1.0093852945409889e-07,"reference_active":true,"delta_db":-1.6986486732748107}],"active_band_count":3,"active_band_mae_db":0.7376442845562522}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":1.1885408929324668,"rms_envelope_nrmse":0.11267093811651646,"envelope_correlation":0.9900954892692201,"rms_gain_error_db":-0.3611506124699665,"reference_mean_power":8.01766615945731e-06,"output_mean_power":7.377902434955429e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.89214453053621e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.737362730986326e-06,"output_power":2.5888543944571685e-06,"reference_active":true,"delta_db":-0.24224725073527978},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.579169756744006e-07,"output_power":6.654067872122889e-07,"reference_active":true,"delta_db":-0.5653440801636027},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8094260096505427e-07,"output_power":1.376849828958738e-07,"reference_active":true,"delta_db":-1.1865425384308095}],"active_band_count":3,"active_band_mae_db":0.664711289776564}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":0.7677095670076135,"rms_envelope_nrmse":0.06231950481118317,"envelope_correlation":0.9977526352332893,"rms_gain_error_db":-0.36027731487857306,"reference_mean_power":8.04858299788563e-06,"output_mean_power":7.407841737375978e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.450576990854381e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.750016160596198e-06,"output_power":2.5817307970882244e-06,"reference_active":true,"delta_db":-0.27424290566290227},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5474696245295724e-07,"output_power":3.074360013154619e-07,"reference_active":true,"delta_db":-0.6216396265072787},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2784081824019621e-08,"output_power":8.783400138080206e-09,"reference_active":true,"delta_db":-1.6300687366722901}],"active_band_count":3,"active_band_mae_db":0.8419837562808237}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-126133-0003","method":"magphase","status":"measured","mos":{"status":"measured","value":3.657620429992676,"predicted_mos":3.657620429992676,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":286868,"input_sample_rate":48000,"input_seconds":5.976416666666666,"input_float32_sha256":"cfa1f15d60212fb689bfd45efbe303a5d26dc36e6c35d104f0072b70942f3713","internal_samples":95623,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.494250327001282},"started_at":"2026-10-13T05:13:46.300132+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008542535999367828,"native_input_samples":288000,"native_input_sha256":"07f41371feab4ea8626a2150a243253ba59697c326a1c4d125665ad65267db01","output_sha256":"0fe16332c8c681685702e709ac2998462b3c582075aac6c8673abc443f322f0b","output_samples":286868,"length_delta_samples":-1132,"analysis_s":3.727142014000492,"synthesis_s":1.120248070999878,"total_s":4.84739008500037,"rtf":0.8078983475000617,"output_peak":0.2445357398394844,"output_rms":0.029534729114591096,"finished_at":"2026-10-13T05:13:51.168630+00:00","evaluation_input":{"reference":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"0fe16332c8c681685702e709ac2998462b3c582075aac6c8673abc443f322f0b","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95623,"length_delta_samples":-377,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.58731577397808,"output_peak":0.2447887596301447},"common_prefix":{"samples":95623,"sample_rate":16000,"reference_coverage":0.9960729166666666,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95623,"output_samples":95623,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.863663265677506,"spectral_convergence":0.40952100962802185,"spectral_convergence_status":"ok","amplitude_floor":6.60007588392934e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.86240786602015,"spectral_convergence":0.13828942080990927,"spectral_convergence_status":"ok","amplitude_floor":5.570424644238683e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.188011997600083,"spectral_convergence":0.15853493492669007,"spectral_convergence_status":"ok","amplitude_floor":3.8555660806566813e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.143490545312991e-05,"output_power":2.9239476140877173e-05,"delta_db":-0.3144257099295946,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030506438759804423,"output_power":0.0002791593142426058,"delta_db":-0.38539389416467174,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004173403899162805,"output_power":0.000389544339355868,"delta_db":-0.2993351966510893,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011460799254850687,"output_power":0.00011519698234269067,"delta_db":0.022261970415320374,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.7185261885615396e-05,"output_power":5.756531603471961e-05,"delta_db":0.02876778946006976,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.0109417520392834e-06,"output_power":2.805660864750298e-06,"delta_db":-0.30667179615207274,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22614272612880307,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2658077796860764,"pitch":null,"output_rms_dbfs":-30.58731577397808,"output_peak":0.2447887596301447,"reference_rms_dbfs":-30.321507994292002,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.304694376432579},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95623,"output_samples":95623,"intelligibility":{"estoi":{"value":0.9636988105758808,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030473658775303246,"rms_dbfs":-30.321507994292002,"peak":0.233489990234375,"peak_dbfs":-12.634634659915168,"crest_factor_db":17.686873334376834,"mean":-0.0001047286252473908,"zero_energy":false},"output":{"rms":0.02955522101053402,"rms_dbfs":-30.58731577397808,"peak":0.2447887596301447,"peak_dbfs":-12.224170565715797,"crest_factor_db":18.36314520826228,"mean":0.00034918165066184017,"zero_energy":false},"rms_difference_db":-0.26580777968607805},"full_scale":{"reference":{"sample_count":95623,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95623,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":597,"uncovered_tail_samples":103,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":69,"quiet_samples":11040,"reference_mean_power":3.419871327291796e-07,"output_mean_power":4.7047334407160675e-07,"output_rms_dbfs":-63.27464977657669,"output_zero_energy":false,"output_peak":0.0116879946040979,"output_energy_fraction":6.218323449354909e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":590,"reference_estimated_voiced_frames":251,"output_estimated_voiced_frames":267,"both_estimated_voiced_frames":240,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9561752988047809,"absolute_difference_cents_median":2.7103098464289523,"absolute_difference_cents_p95":8.872408127007413,"difference_over_600_cents_frames":0,"uncovered_tail_samples":103,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95623,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1194,"rms_floor":9.667347842324586e-05,"log_rms_mae_db":1.058457475425795,"rms_envelope_nrmse":0.06868716014146523,"envelope_correlation":0.9960744252938728,"rms_gain_error_db":-0.2656938445377047,"reference_mean_power":0.0009287815184421741,"output_mean_power":0.0008736634766001482},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":3.303801689322474e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016313811774608855,"output_power":0.00015622238506446656,"reference_active":true,"delta_db":-0.18812183257760096},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1306865316668425e-05,"output_power":3.066054903336129e-05,"reference_active":true,"delta_db":-0.09059657474098302},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6994318254072617e-06,"output_power":3.4616792734053234e-06,"reference_active":true,"delta_db":-0.28848200613614916}],"active_band_count":3,"active_band_mae_db":0.18906680448491106}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1188,"rms_floor":9.141277463219576e-05,"log_rms_mae_db":0.7241844039407199,"rms_envelope_nrmse":0.040510171475565196,"envelope_correlation":0.9991911275845721,"rms_gain_error_db":-0.26508617900290066,"reference_mean_power":0.0009292557470041667,"output_mean_power":0.0008742318764289772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":3.0359618108271217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014744797737141344,"output_power":0.0001416668559832829,"reference_active":true,"delta_db":-0.17370563761034308},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4468711815048213e-05,"output_power":1.4065114717865288e-05,"reference_active":true,"delta_db":-0.12286587828153289},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.993632043609725e-07,"output_power":3.532641260941115e-07,"reference_active":true,"delta_db":-0.5326851116787802}],"active_band_count":3,"active_band_mae_db":0.27641887585688535}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1194,"rms_floor":8.73108999878072e-05,"log_rms_mae_db":1.1631711666474736,"rms_envelope_nrmse":0.06591231956213044,"envelope_correlation":0.9972278430758696,"rms_gain_error_db":-0.3132271297074054,"reference_mean_power":0.0005499988047324193,"output_mean_power":0.0005117278255906333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":2.0428415500771273e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.410553560839843e-05,"output_power":8.793540912267841e-05,"reference_active":true,"delta_db":-0.2945138236207189},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.225281434454023e-05,"output_power":2.095391499984389e-05,"reference_active":true,"delta_db":-0.2611976677043678},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1268010257586878e-06,"output_power":1.933480125045665e-06,"reference_active":true,"delta_db":-0.4138714894384905}],"active_band_count":3,"active_band_mae_db":0.3231943269211924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1188,"rms_floor":8.091336805501339e-05,"log_rms_mae_db":0.8207594647085946,"rms_envelope_nrmse":0.04471613832174455,"envelope_correlation":0.9992643718519496,"rms_gain_error_db":-0.312611742127306,"reference_mean_power":0.0005501962788669412,"output_mean_power":0.0005119841008340055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":1.8835368981507142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.548400993128531e-05,"output_power":7.982374983142887e-05,"reference_active":true,"delta_db":-0.29752760582738885},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.876463387203636e-06,"output_power":9.362346342224918e-06,"reference_active":true,"delta_db":-0.23216755123011992},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.415654266780246e-07,"output_power":2.1221035110542034e-07,"reference_active":true,"delta_db":-0.562682133990665}],"active_band_count":3,"active_band_mae_db":0.36412576368272465}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1194,"rms_floor":6.045309710707618e-05,"log_rms_mae_db":0.9023096826820677,"rms_envelope_nrmse":0.07458797356881178,"envelope_correlation":0.9955283305779625,"rms_gain_error_db":-0.005000781833752268,"reference_mean_power":0.00014433109148906265,"output_mean_power":0.00014416499384295524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":6.641294976728031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.675187141619839e-05,"output_power":3.648353817388599e-05,"reference_active":true,"delta_db":-0.03182508805286808},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.717320293411943e-06,"output_power":7.780382627037234e-06,"reference_active":true,"delta_db":0.03534430207115978},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8652849798296607e-06,"output_power":1.8712370179019022e-06,"reference_active":true,"delta_db":0.013836072576595162}],"active_band_count":3,"active_band_mae_db":0.02700182090020768}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1188,"rms_floor":5.539248843957364e-05,"log_rms_mae_db":0.5106851801687832,"rms_envelope_nrmse":0.03559918840285494,"envelope_correlation":0.9988934651283511,"rms_gain_error_db":-0.0033560155361466782,"reference_mean_power":0.00014434914495206898,"output_mean_power":0.00014423764207407392},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":6.207510272433012e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.599604677151923e-05,"output_power":3.576117219075855e-05,"reference_active":true,"delta_db":-0.028430616494074024},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5722495748815897e-06,"output_power":3.646161398942964e-06,"reference_active":true,"delta_db":0.08894095761568993},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.513429823574929e-07,"output_power":1.4819056317024868e-07,"reference_active":true,"delta_db":-0.09141739410970935}],"active_band_count":3,"active_band_mae_db":0.06959632273982443}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1194,"rms_floor":1.667503736683172e-05,"log_rms_mae_db":1.0334639481462233,"rms_envelope_nrmse":0.13566123137051744,"envelope_correlation":0.9859157606466121,"rms_gain_error_db":0.31541821593506536,"reference_mean_power":8.037281811027694e-06,"output_mean_power":8.6427314338287e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":4.858060653588983e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7213896383139267e-06,"output_power":2.7695143023908997e-06,"reference_active":true,"delta_db":0.07612885403242775},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.461692454365688e-07,"output_power":7.643219587322872e-07,"reference_active":true,"delta_db":0.10438991980318688},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8083440958036947e-07,"output_power":1.8154979064075312e-07,"reference_active":true,"delta_db":0.017146798915669823}],"active_band_count":3,"active_band_mae_db":0.06588852425042814}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":23,"metrics":{"frame_count":1188,"rms_floor":1.4241724067408683e-05,"log_rms_mae_db":0.651727511558477,"rms_envelope_nrmse":0.07700021617017669,"envelope_correlation":0.9958006343066588,"rms_gain_error_db":0.3197430689489844,"reference_mean_power":8.0462044852627e-06,"output_mean_power":8.660946828509385e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":4.415866318556709e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7318958393515643e-06,"output_power":2.7748771041728297e-06,"reference_active":true,"delta_db":0.06779616810404368},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.458309857474409e-07,"output_power":3.6067308845526834e-07,"reference_active":true,"delta_db":0.18249836161525107},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2892547666459682e-08,"output_power":1.2802141966320425e-08,"reference_active":true,"delta_db":-0.030561068674962347}],"active_band_count":3,"active_band_mae_db":0.0936185327980857}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"hnm","status":"measured","mos":{"status":"measured","value":1.4635086059570312,"predicted_mos":1.4635086059570312,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":33040,"input_sample_rate":16000,"input_seconds":2.065,"input_float32_sha256":"3cbf2ed618eb1f85220002b6e407cb35b0412474ab07099eb6b5251f0050647c","internal_samples":33040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.04033923300085},"started_at":"2026-10-13T04:38:40.016072+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2719999176624697e-05,"native_input_samples":33040,"native_input_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","output_sha256":"51c13f5a10d3c23db49271dd561ed5733e2d2efa77596844d3e77c558e7bc2ba","output_samples":33040,"length_delta_samples":0,"analysis_s":0.3616325009998036,"synthesis_s":0.17182836899974063,"total_s":0.5334608699995442,"rtf":0.2583345617431207,"output_peak":0.9612942841299921,"output_rms":0.08465975658756261,"finished_at":"2026-10-13T04:38:40.551966+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"51c13f5a10d3c23db49271dd561ed5733e2d2efa77596844d3e77c558e7bc2ba","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.877425183997666,"spectral_convergence":0.31041024380467613,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.691130563720353,"spectral_convergence":0.33715893627457966,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.507516129925867,"spectral_convergence":0.35629069341785313,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00021761763361808226,"delta_db":0.015222628774537053,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.001114682979947605,"delta_db":0.13900933603370044,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.002970498500844659,"delta_db":-0.028314582025116886,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.0018894136605933707,"delta_db":-0.37125825933593504,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0009223712581915172,"delta_db":-0.24159454727438331,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":5.2690352270117345e-05,"delta_db":4.341057326905707,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8560761133915632,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10277972981978822,"pitch":null,"output_rms_dbfs":-21.446459690741317,"output_peak":0.9612942841299921,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.025357292547962},"common_prefix":{"samples":33040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.877425183997666,"spectral_convergence":0.31041024380467613,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.691130563720353,"spectral_convergence":0.33715893627457966,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.507516129925867,"spectral_convergence":0.35629069341785313,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00021761763361808226,"delta_db":0.015222628774537053,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.001114682979947605,"delta_db":0.13900933603370044,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.002970498500844659,"delta_db":-0.028314582025116886,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.0018894136605933707,"delta_db":-0.37125825933593504,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0009223712581915172,"delta_db":-0.24159454727438331,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":5.2690352270117345e-05,"delta_db":4.341057326905707,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8560761133915632,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10277972981978822,"pitch":null,"output_rms_dbfs":-21.446459690741317,"output_peak":0.9612942841299921,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.025357292547962},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":33040,"output_samples":33040,"intelligibility":{"estoi":{"value":0.8404729931574609,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08566748199211072,"rms_dbfs":-21.343679960921527,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.764991009696175,"mean":-9.639823119230478e-05,"zero_energy":false},"output":{"rms":0.08465975658756261,"rms_dbfs":-21.446459690741317,"peak":0.9612942841299921,"peak_dbfs":-0.3428728000329949,"crest_factor_db":21.10358689070832,"mean":-2.4828703659318503e-05,"zero_energy":false},"rms_difference_db":-0.10277972981979033},"full_scale":{"reference":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":2.9347763462860665e-06,"output_rms_dbfs":-55.324249899103464,"output_zero_energy":false,"output_peak":0.022215766771571357,"output_energy_fraction":9.319632369098072e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":47,"both_estimated_voiced_frames":41,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.82,"absolute_difference_cents_median":1.6661678006530638,"absolute_difference_cents_p95":5.098748975533875,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":33040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00032988134947798775,"log_rms_mae_db":0.9393330528089996,"rms_envelope_nrmse":0.08343826727297633,"envelope_correlation":0.9935752514795789,"rms_gain_error_db":-0.10277893167416995,"reference_mean_power":0.007356724541990913,"output_mean_power":0.007184666304031415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":2.980285612038863e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020727352824950612,"output_power":0.001973499704524279,"reference_active":true,"delta_db":-0.2130677440169475},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028987718896809884,"output_power":0.00023547314967687678,"reference_active":true,"delta_db":-0.902726482313152},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.012977878291207e-05,"output_power":6.039911666834059e-05,"reference_active":true,"delta_db":-0.6487188181443798}],"active_band_count":3,"active_band_mae_db":0.5881710148248264}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0002960474071300256,"log_rms_mae_db":0.47687377787259966,"rms_envelope_nrmse":0.035138663983703085,"envelope_correlation":0.9988663281248789,"rms_gain_error_db":-0.10278074643083529,"reference_mean_power":0.007465415825536721,"output_mean_power":0.0072908124816209625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":2.5597102021075326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001877761726334087,"output_power":0.0017934319625517899,"reference_active":true,"delta_db":-0.1995557727323496},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012702101262043313,"output_power":0.0001042599211855317,"reference_active":true,"delta_db":-0.857581783533431},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.962314408759722e-06,"output_power":3.2239887676848655e-06,"reference_active":true,"delta_db":-0.8955541367755555}],"active_band_count":3,"active_band_mae_db":0.6508972310137787}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00021933697126022545,"log_rms_mae_db":1.4275678433607195,"rms_envelope_nrmse":0.12879107814327576,"envelope_correlation":0.9852078181684387,"rms_gain_error_db":-0.09880884977786435,"reference_mean_power":0.003363603458339666,"output_mean_power":0.0032879401638529733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.3609823101859838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008546335077009212,"output_power":0.0007547545086238757,"reference_active":true,"delta_db":-0.5397420022466505},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016660612815788933,"output_power":0.0001331673043223704,"reference_active":true,"delta_db":-0.9729336321538598},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.188726605138031e-05,"output_power":4.8232099366156526e-05,"reference_active":true,"delta_db":0.612541503471723}],"active_band_count":3,"active_band_mae_db":0.7084057126240776}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0001893291578274055,"log_rms_mae_db":0.8947400891992137,"rms_envelope_nrmse":0.06931174468958665,"envelope_correlation":0.9954548832223277,"rms_gain_error_db":-0.09881084250659597,"reference_mean_power":0.0034133051942474998,"output_mean_power":0.003336522342535857},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.15224102535942e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007711766819603955,"output_power":0.0006894807933663282,"reference_active":true,"delta_db":-0.4863171666217615},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.304889773698623e-05,"output_power":6.17288621015702e-05,"reference_active":true,"delta_db":-0.7312539574596664},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5271182055851987e-06,"output_power":2.2122835397619456e-06,"reference_active":true,"delta_db":-0.5778476842439996}],"active_band_count":3,"active_band_mae_db":0.5984729361084757}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00022567359951457066,"log_rms_mae_db":1.7044475323024255,"rms_envelope_nrmse":0.11415862623035475,"envelope_correlation":0.990967406164127,"rms_gain_error_db":-0.3453931117148834,"reference_mean_power":0.002711357875845493,"output_mean_power":0.002504076158656396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.6613986227009822e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012279896352023918,"output_power":0.0011344925620054944,"reference_active":true,"delta_db":-0.3439304827969203},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188832719696088,"output_power":9.301103951678058e-05,"reference_active":true,"delta_db":-1.065862512725359},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.234146057131371e-05,"output_power":2.9472851279522197e-05,"reference_active":true,"delta_db":-0.4033747651893468}],"active_band_count":3,"active_band_mae_db":0.6043892535705421}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.00020374269470838248,"log_rms_mae_db":0.9135264772085409,"rms_envelope_nrmse":0.05523187438719043,"envelope_correlation":0.9987051984467255,"rms_gain_error_db":-0.3453964556900426,"reference_mean_power":0.0027514125536590357,"output_mean_power":0.0025410667233076835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.4886745667367608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372871645153378,"output_power":0.001052881899119585,"reference_active":true,"delta_db":-0.3349047823339121},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.907501205097419e-05,"output_power":3.968671722732822e-05,"reference_active":true,"delta_db":-0.9221523813319215},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.526885866477726e-06,"output_power":1.75531750282881e-06,"reference_active":true,"delta_db":-1.58229942957}],"active_band_count":3,"active_band_mae_db":0.9464521977452778}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":2.007086972409457,"rms_envelope_nrmse":0.4106760456570009,"envelope_correlation":0.8961359607517007,"rms_gain_error_db":0.7538038636658568,"reference_mean_power":9.509723574878586e-05,"output_mean_power":0.00011312231379687155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":5.720070068605407e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5023814010404594e-05,"output_power":4.7504834883406174e-05,"reference_active":true,"delta_db":1.3237437495265074},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.975419310826886e-06,"output_power":8.637349649172128e-06,"reference_active":true,"delta_db":2.3955081322532568},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4719997044746735e-06,"output_power":1.6018169665672457e-05,"reference_active":true,"delta_db":10.367051666210482}],"active_band_count":3,"active_band_mae_db":4.695434515996749}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":1.6328792338935805,"rms_envelope_nrmse":0.3330332214895439,"envelope_correlation":0.9274429860995922,"rms_gain_error_db":0.753834897040037,"reference_mean_power":9.649976412079661e-05,"output_mean_power":0.00011479150279124675},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":5.0877964817728856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.168565282093968e-05,"output_power":5.532221610654496e-05,"reference_active":true,"delta_db":2.4203690948409373},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3362275317320777e-06,"output_power":4.893767107994415e-06,"reference_active":true,"delta_db":3.2112815998934914},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1955736150290604e-07,"output_power":1.114395269061965e-06,"reference_active":true,"delta_db":9.694629376310079}],"active_band_count":3,"active_band_mae_db":5.108760023681502}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.2438230514526367,"predicted_mos":2.2438230514526367,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":33040,"input_sample_rate":16000,"input_seconds":2.065,"input_float32_sha256":"f958babdd151658664312d3c680adcea7dad472944a463cda5d030dbef4e0f85","internal_samples":33040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.771994534999976},"started_at":"2026-10-13T04:41:36.843491+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.108499979338376e-05,"native_input_samples":33040,"native_input_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","output_sha256":"c7468503ced248cd444700ee08cdb22ea3b85a564ade4c245d77f83c8e12ce60","output_samples":33040,"length_delta_samples":0,"analysis_s":0.5168510980001884,"synthesis_s":0.03347306399973604,"total_s":0.5503241619999244,"rtf":0.26650080484257843,"output_peak":1.3745894747842422,"output_rms":0.09404013611760229,"finished_at":"2026-10-13T04:41:37.397176+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"c7468503ced248cd444700ee08cdb22ea3b85a564ade4c245d77f83c8e12ce60","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.81301275889998,"spectral_convergence":0.6230580707334445,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.0511888500612505,"spectral_convergence":0.24442077464337147,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.582548866605491,"spectral_convergence":0.29933651814390744,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00019315057571541377,"delta_db":-0.5027581361339621,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.001297502486406212,"delta_db":0.7985776241105,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.00375254576042323,"delta_db":0.9866520121719075,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.0024421341646757994,"delta_db":0.7431664236022278,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0011375089099733492,"delta_db":0.668895908501874,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":2.0705303823163577e-05,"delta_db":0.2845623814683699,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6641020809981403,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8099449423064016,"pitch":null,"output_rms_dbfs":-20.53373501861513,"output_peak":1.3745894747842422,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.815583491855574},"common_prefix":{"samples":33040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.81301275889998,"spectral_convergence":0.6230580707334445,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.0511888500612505,"spectral_convergence":0.24442077464337147,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.582548866605491,"spectral_convergence":0.29933651814390744,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00019315057571541377,"delta_db":-0.5027581361339621,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.001297502486406212,"delta_db":0.7985776241105,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.00375254576042323,"delta_db":0.9866520121719075,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.0024421341646757994,"delta_db":0.7431664236022278,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0011375089099733492,"delta_db":0.668895908501874,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":2.0705303823163577e-05,"delta_db":0.2845623814683699,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6641020809981403,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8099449423064016,"pitch":null,"output_rms_dbfs":-20.53373501861513,"output_peak":1.3745894747842422,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.815583491855574},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":33040,"output_samples":33040,"intelligibility":{"estoi":{"value":0.9434547249918532,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08566748199211072,"rms_dbfs":-21.343679960921527,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.764991009696175,"mean":-9.639823119230478e-05,"zero_energy":false},"output":{"rms":0.0940401361176023,"rms_dbfs":-20.53373501861513,"peak":1.3745894747842422,"peak_dbfs":2.7634602839586457,"crest_factor_db":23.297195302573776,"mean":-4.999360130963135e-10,"zero_energy":false},"rms_difference_db":0.809944942306398},"full_scale":{"reference":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":18,"at_or_above_full_scale_fraction":0.0005447941888619854,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0005447941888619854,"longest_full_scale_run_samples":2,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":2.2914104783374193e-06,"output_rms_dbfs":-56.398971052634124,"output_zero_energy":false,"output_peak":0.01067105432941786,"output_energy_fraction":5.8973132710919415e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":63,"both_estimated_voiced_frames":46,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.92,"absolute_difference_cents_median":4.690717016401241,"absolute_difference_cents_p95":14.176170166487207,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":33040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00032988134947798775,"log_rms_mae_db":1.808572187649905,"rms_envelope_nrmse":0.16616952118314274,"envelope_correlation":0.9846592226415872,"rms_gain_error_db":0.8099475607336565,"reference_mean_power":0.007356724541990913,"output_mean_power":0.008865010434525679},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":2.980285612038863e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020727352824950612,"output_power":0.0025188116562142827,"reference_active":true,"delta_db":0.8465185429373168},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028987718896809884,"output_power":0.00029914005721052526,"reference_active":true,"delta_db":0.13660531204211968},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.012977878291207e-05,"output_power":6.588731259820913e-05,"reference_active":true,"delta_db":-0.2710067499338972}],"active_band_count":3,"active_band_mae_db":0.4180435349711112}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0002960474071300256,"log_rms_mae_db":1.330262208005628,"rms_envelope_nrmse":0.11220467358295601,"envelope_correlation":0.997156036411137,"rms_gain_error_db":0.809950073139909,"reference_mean_power":0.007465415825536721,"output_mean_power":0.008995990960326536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":2.5597102021075326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001877761726334087,"output_power":0.00229664654249623,"reference_active":true,"delta_db":0.8745367897292378},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012702101262043313,"output_power":0.00013626000252902866,"reference_active":true,"delta_db":0.30492822021055677},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.962314408759722e-06,"output_power":3.766769413350539e-06,"reference_active":true,"delta_db":-0.21979898504345696}],"active_band_count":3,"active_band_mae_db":0.46642133166108385}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00021933697126022545,"log_rms_mae_db":1.8645183914715355,"rms_envelope_nrmse":0.1924339948363063,"envelope_correlation":0.9818397675983447,"rms_gain_error_db":0.9822677014951475,"reference_mean_power":0.003363603458339666,"output_mean_power":0.004217271495222708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.3609823101859838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008546335077009212,"output_power":0.0011065273959099831,"reference_active":true,"delta_db":1.121822546153391},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016660612815788933,"output_power":0.00018171816507021207,"reference_active":true,"delta_db":0.3770737108515038},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.188726605138031e-05,"output_power":3.693222105839809e-05,"reference_active":true,"delta_db":-0.546765888570878}],"active_band_count":3,"active_band_mae_db":0.681887381858591}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0001893291578274055,"log_rms_mae_db":1.412758803906956,"rms_envelope_nrmse":0.13667543496692527,"envelope_correlation":0.9961653917675819,"rms_gain_error_db":0.9822708018025345,"reference_mean_power":0.0034133051942474998,"output_mean_power":0.0042795903722184035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.15224102535942e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007711766819603955,"output_power":0.0010003833303151722,"reference_active":true,"delta_db":1.1301255699290456},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.304889773698623e-05,"output_power":8.64349439072435e-05,"reference_active":true,"delta_db":0.7307568741025361},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5271182055851987e-06,"output_power":2.133024011323215e-06,"reference_active":true,"delta_db":-0.7362981215217245}],"active_band_count":3,"active_band_mae_db":0.8657268551844354}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00022567359951457066,"log_rms_mae_db":1.763723902763238,"rms_envelope_nrmse":0.1635077653908462,"envelope_correlation":0.9868276064655331,"rms_gain_error_db":0.7335040476101172,"reference_mean_power":0.002711357875845493,"output_mean_power":0.003210238133135385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.6613986227009822e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012279896352023918,"output_power":0.0014239668663509452,"reference_active":true,"delta_db":0.6430518283872482},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188832719696088,"output_power":0.00013736334149306274,"reference_active":true,"delta_db":0.6275009743671076},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.234146057131371e-05,"output_power":2.9836979345520096e-05,"reference_active":true,"delta_db":-0.35004775513732067}],"active_band_count":3,"active_band_mae_db":0.5402001859638922}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.00020374269470838248,"log_rms_mae_db":1.1770624047189129,"rms_envelope_nrmse":0.10484513518113354,"envelope_correlation":0.9975747044082318,"rms_gain_error_db":0.7335021793937044,"reference_mean_power":0.0027514125536590357,"output_mean_power":0.0032576613286093813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.4886745667367608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372871645153378,"output_power":0.0013217838634632258,"reference_active":true,"delta_db":0.6529030793157031},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.907501205097419e-05,"output_power":5.9113983338333674e-05,"reference_active":true,"delta_db":0.8082980992820528},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.526885866477726e-06,"output_power":2.2010117623479568e-06,"reference_active":true,"delta_db":-0.5996326284020418}],"active_band_count":3,"active_band_mae_db":0.6869446023332659}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":1.577639362835105,"rms_envelope_nrmse":0.14460787208909337,"envelope_correlation":0.988122483816725,"rms_gain_error_db":0.4752704746972017,"reference_mean_power":9.509723574878586e-05,"output_mean_power":0.00010609500575099066},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":5.720070068605407e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5023814010404594e-05,"output_power":3.7805851514762606e-05,"reference_active":true,"delta_db":0.3319558631557973},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.975419310826886e-06,"output_power":5.007699743417403e-06,"reference_active":true,"delta_db":0.028085936932326437},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4719997044746735e-06,"output_power":1.2114251471432962e-06,"reference_active":true,"delta_db":-0.8461113816858954}],"active_band_count":3,"active_band_mae_db":0.40205106059133966}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":0.9529906972828192,"rms_envelope_nrmse":0.0758148555186796,"envelope_correlation":0.9981386799196095,"rms_gain_error_db":0.47530435320896175,"reference_mean_power":9.649976412079661e-05,"output_mean_power":0.00010766057304324322},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":5.0877964817728856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.168565282093968e-05,"output_power":3.4091966360241925e-05,"reference_active":true,"delta_db":0.31789391894621344},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3362275317320777e-06,"output_power":2.3212473046571815e-06,"reference_active":true,"delta_db":-0.027937251920872694},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1955736150290604e-07,"output_power":8.813267470154614e-08,"reference_active":true,"delta_db":-1.3243937153704874}],"active_band_count":3,"active_band_mae_db":0.5567416287458579}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"facodec","status":"measured","mos":{"status":"measured","value":2.946568727493286,"predicted_mos":2.946568727493286,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":33200,"input_sample_rate":16000,"input_seconds":2.075,"input_float32_sha256":"9d56aa9e94237cbcdd3a9deffb479d7ad71bf76491941b81cbc53fd163d6a4ad","internal_samples":33200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.221174226000585},"started_at":"2026-10-13T04:48:54.375016+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.831600002333289e-05,"native_input_samples":33040,"native_input_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","output_sha256":"c901636fac6fd2d2fba5839d807d4abc282113845192eed4084cca2231cb50a1","output_samples":33200,"length_delta_samples":160,"analysis_s":0.5281122650003454,"synthesis_s":3.5701890839991393,"total_s":4.098301348999485,"rtf":1.9846495636801378,"output_peak":0.7808858752250671,"output_rms":0.08069822157074644,"finished_at":"2026-10-13T04:48:58.477425+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"c901636fac6fd2d2fba5839d807d4abc282113845192eed4084cca2231cb50a1","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":33040,"output_samples":33200,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.862720723283964,"output_peak":0.7808858752250671},"common_prefix":{"samples":33040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9951807228915662,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.095171657200627,"spectral_convergence":0.2289627028036622,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.25137453090622,"spectral_convergence":0.2507945075700143,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.443297647987299,"spectral_convergence":0.2740055338320103,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00018727426574748437,"delta_db":-0.6369371758520209,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.000879119487388996,"delta_db":-0.8920252920861846,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.0027605827064086912,"delta_db":-0.34660032703309307,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.001965186973240259,"delta_db":-0.20049000307814369,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0007259249930262729,"delta_db":-1.281734671918584,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":2.5650491838164746e-05,"delta_db":1.214703275148312,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7620817908527231,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4980604039707235,"pitch":null,"output_rms_dbfs":-21.841740364892253,"output_peak":0.7808858752250671,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.263281278698048},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":33040,"output_samples":33040,"intelligibility":{"estoi":{"value":0.9208955394284487,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08566748199211072,"rms_dbfs":-21.343679960921527,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.764991009696175,"mean":-9.639823119230478e-05,"zero_energy":false},"output":{"rms":0.08089337993711146,"rms_dbfs":-21.841740364892253,"peak":0.7808858752250671,"peak_dbfs":-2.1482486536100325,"crest_factor_db":19.69349171128222,"mean":0.00014732093004832274,"zero_energy":false},"rms_difference_db":-0.49806040397072593},"full_scale":{"reference":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":2.212203381288637e-06,"output_rms_dbfs":-56.551749482032264,"output_zero_energy":false,"output_peak":0.01923457160592079,"output_energy_fraction":7.694437746176245e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":52,"both_estimated_voiced_frames":47,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.94,"absolute_difference_cents_median":2.836148682653121,"absolute_difference_cents_p95":18.51002303971833,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":33040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00032988134947798775,"log_rms_mae_db":0.9719573260662787,"rms_envelope_nrmse":0.09576828648233131,"envelope_correlation":0.9940571757578462,"rms_gain_error_db":-0.49805981091246065,"reference_mean_power":0.007356724541990913,"output_mean_power":0.006559617471766553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":2.980285612038863e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020727352824950612,"output_power":0.001843959483113589,"reference_active":true,"delta_db":-0.5079246593760463},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028987718896809884,"output_power":0.0002646853134983682,"reference_active":true,"delta_db":-0.39484196750543654},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.012977878291207e-05,"output_power":8.114315873615259e-05,"reference_active":true,"delta_db":0.6334944092344725}],"active_band_count":3,"active_band_mae_db":0.5120870120386517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0002960474071300256,"log_rms_mae_db":0.6405227309468274,"rms_envelope_nrmse":0.08445489501671902,"envelope_correlation":0.9956999628548351,"rms_gain_error_db":-0.4980625076218681,"reference_mean_power":0.007465415825536721,"output_mean_power":0.006656527833286961},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":2.5597102021075326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001877761726334087,"output_power":0.0016734563579971747,"reference_active":true,"delta_db":-0.500260918704765},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012702101262043313,"output_power":0.00011746523573698233,"reference_active":true,"delta_db":-0.339662160481375},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.962314408759722e-06,"output_power":4.535443621271848e-06,"reference_active":true,"delta_db":0.5867083899010344}],"active_band_count":3,"active_band_mae_db":0.4755438230290581}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00021933697126022545,"log_rms_mae_db":1.3513838169271462,"rms_envelope_nrmse":0.1108183049652238,"envelope_correlation":0.990738518558222,"rms_gain_error_db":-0.4370552085798965,"reference_mean_power":0.003363603458339666,"output_mean_power":0.0030415801797247107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.3609823101859838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008546335077009212,"output_power":0.0006857053808976502,"reference_active":true,"delta_db":-0.9564235874827297},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016660612815788933,"output_power":0.00014578331913101425,"reference_active":true,"delta_db":-0.5798313788543595},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.188726605138031e-05,"output_power":4.6092208784063096e-05,"reference_active":true,"delta_db":0.41545505212714984}],"active_band_count":3,"active_band_mae_db":0.6505700061547464}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0001893291578274055,"log_rms_mae_db":0.9433774785401617,"rms_envelope_nrmse":0.09604814706403557,"envelope_correlation":0.9928828264398183,"rms_gain_error_db":-0.4370528098689263,"reference_mean_power":0.0034133051942474998,"output_mean_power":0.0030865252961289925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.15224102535942e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007711766819603955,"output_power":0.000624034121918747,"reference_active":true,"delta_db":-0.919455519796829},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.304889773698623e-05,"output_power":6.057926436557762e-05,"reference_active":true,"delta_db":-0.8128967183808191},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5271182055851987e-06,"output_power":2.6952687280019138e-06,"reference_active":true,"delta_db":0.2797651594989301}],"active_band_count":3,"active_band_mae_db":0.6707057992255261}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00022567359951457066,"log_rms_mae_db":1.348420707288445,"rms_envelope_nrmse":0.12403728041102806,"envelope_correlation":0.9899888199095369,"rms_gain_error_db":-0.4745888733023167,"reference_mean_power":0.002711357875845493,"output_mean_power":0.0024306809491965358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.6613986227009822e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012279896352023918,"output_power":0.001170485574353034,"reference_active":true,"delta_db":-0.20828635571131354},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188832719696088,"output_power":0.00012780837603960194,"reference_active":true,"delta_db":0.31438567175561205},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.234146057131371e-05,"output_power":3.780203834392721e-05,"reference_active":true,"delta_db":0.6775558912590303}],"active_band_count":3,"active_band_mae_db":0.400075972908652}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.00020374269470838248,"log_rms_mae_db":0.9958499220441743,"rms_envelope_nrmse":0.10483827201508956,"envelope_correlation":0.992993106525158,"rms_gain_error_db":-0.4745969906830151,"reference_mean_power":0.0027514125536590357,"output_mean_power":0.0024665845984399837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.4886745667367608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372871645153378,"output_power":0.0010811482771179864,"reference_active":true,"delta_db":-0.21984877127658367},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.907501205097419e-05,"output_power":5.72153224386576e-05,"reference_active":true,"delta_db":0.6665193484379262},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.526885866477726e-06,"output_power":2.793906279845235e-06,"reference_active":true,"delta_db":0.4362620753929797}],"active_band_count":3,"active_band_mae_db":0.44087673170249647}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":1.7303904771461251,"rms_envelope_nrmse":0.21143718011212612,"envelope_correlation":0.969233196172731,"rms_gain_error_db":-0.4916693224511554,"reference_mean_power":9.509723574878586e-05,"output_mean_power":8.491823540225314e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":5.720070068605407e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5023814010404594e-05,"output_power":3.7085982891496994e-05,"reference_active":true,"delta_db":0.24846355663507488},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.975419310826886e-06,"output_power":4.409348258956018e-06,"reference_active":true,"delta_db":-0.5245528582518981},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4719997044746735e-06,"output_power":1.9986120309696457e-06,"reference_active":true,"delta_db":1.3282077457827968}],"active_band_count":3,"active_band_mae_db":0.7004080535565899}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":1.298406090241359,"rms_envelope_nrmse":0.1888833157490813,"envelope_correlation":0.9745956206086884,"rms_gain_error_db":-0.4916551014074879,"reference_mean_power":9.649976412079661e-05,"output_mean_power":8.617092236919084e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":5.0877964817728856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.168565282093968e-05,"output_power":3.306371113095537e-05,"reference_active":true,"delta_db":0.1848893862998422},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3362275317320777e-06,"output_power":2.1046405938276256e-06,"reference_active":true,"delta_db":-0.45337194938174336},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1955736150290604e-07,"output_power":1.4563005075701678e-07,"reference_active":true,"delta_db":0.8567467879894239}],"active_band_count":3,"active_band_mae_db":0.49833604122366987}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.635977268218994,"predicted_mos":3.635977268218994,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":91136,"input_sample_rate":44100,"input_seconds":2.066575963718821,"input_float32_sha256":"461ceb5e6e3db1f67537996ddc2e23bcc31f0deffe28a2d57181fac20b714196","internal_samples":33066,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.035011055999348},"started_at":"2026-10-13T07:07:44.051267+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004655175000152667,"native_input_samples":91067,"native_input_sha256":"c305625a84cd295b067627439b18d27f8f440b8bf47dc521eeef9ef5b9935236","output_sha256":"18717d29fcced50b4ddef3c29a3a2420cc3f35bc7426757a51be09fe34132ace","output_samples":91136,"length_delta_samples":69,"analysis_s":0.05432062100044277,"synthesis_s":11.776642625000022,"total_s":11.830963246000465,"rtf":5.729248565875898,"output_peak":0.8045237064361572,"output_rms":0.0798342999605012,"finished_at":"2026-10-13T07:07:55.897660+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"18717d29fcced50b4ddef3c29a3a2420cc3f35bc7426757a51be09fe34132ace","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":33040,"output_samples":33066,"length_delta_samples":26,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.948192195132933,"output_peak":0.7899971625507599},"common_prefix":{"samples":33040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9992136938244722,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.532303923675943,"spectral_convergence":0.4961950823936346,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.174978757527404,"spectral_convergence":0.4457743053568921,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.986214605539374,"spectral_convergence":0.4754680355868752,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00011273980242180959,"delta_db":-2.840945512013391,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.0009117581555498146,"delta_db":-0.733707796425998,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.0034440110923807657,"delta_db":0.614037458132418,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.0013586309249338751,"delta_db":-1.8035138123786472,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0005481325741194877,"delta_db":-2.501796044184133,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":1.5044257905803493e-05,"delta_db":-1.102545994551233,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5994244362809698,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6010961456687332,"pitch":null,"output_rms_dbfs":-21.944776106590265,"output_peak":0.7899971625507599,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.231165762247572},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":33040,"output_samples":33040,"intelligibility":{"estoi":{"value":0.7121259916774068,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08566748199211072,"rms_dbfs":-21.343679960921527,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.764991009696175,"mean":-9.639823119230478e-05,"zero_energy":false},"output":{"rms":0.07993945713670787,"rms_dbfs":-21.944776106590265,"peak":0.7899971625507599,"peak_dbfs":-2.047489371425619,"crest_factor_db":19.89728673516464,"mean":-3.08311951913377e-05,"zero_energy":false},"rms_difference_db":-0.6010961456687376},"full_scale":{"reference":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":1.9142743053164963e-06,"output_rms_dbfs":-57.17995830012881,"output_zero_energy":false,"output_peak":0.009478374787338161,"output_energy_fraction":6.818040593036368e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":59,"both_estimated_voiced_frames":47,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.94,"absolute_difference_cents_median":6.867382014816325,"absolute_difference_cents_p95":25.314552189099096,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":33040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00032988134947798775,"log_rms_mae_db":1.5386027647208833,"rms_envelope_nrmse":0.10812308522229377,"envelope_correlation":0.9929281525783609,"rms_gain_error_db":-0.6010945849990424,"reference_mean_power":0.007356724541990913,"output_mean_power":0.006405824505875902},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":2.980285612038863e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020727352824950612,"output_power":0.0018742480701428434,"reference_active":true,"delta_db":-0.4371676777158051},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028987718896809884,"output_power":0.00021503874895272077,"reference_active":true,"delta_db":-1.296973164209829},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.012977878291207e-05,"output_power":5.383542803530241e-05,"reference_active":true,"delta_db":-1.148342985682988}],"active_band_count":3,"active_band_mae_db":0.9608279425362074}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0002960474071300256,"log_rms_mae_db":1.2598960900737064,"rms_envelope_nrmse":0.08209951123000211,"envelope_correlation":0.9976529257721436,"rms_gain_error_db":-0.6010906162095109,"reference_mean_power":0.007465415825536721,"output_mean_power":0.006500472740413947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":2.5597102021075326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001877761726334087,"output_power":0.0016839533313714118,"reference_active":true,"delta_db":-0.47310431352935084},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012702101262043313,"output_power":9.609464669500517e-05,"reference_active":true,"delta_db":-1.2117637626866453},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.962314408759722e-06,"output_power":3.4438513758477373e-06,"reference_active":true,"delta_db":-0.6090453313043748}],"active_band_count":3,"active_band_mae_db":0.7646378025067904}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00021933697126022545,"log_rms_mae_db":1.8788800859756676,"rms_envelope_nrmse":0.14804701047467891,"envelope_correlation":0.984525718547181,"rms_gain_error_db":0.47594095261497543,"reference_mean_power":0.003363603458339666,"output_mean_power":0.00375317561168105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.3609823101859838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008546335077009212,"output_power":0.0009698925786221959,"reference_active":true,"delta_db":0.5494371998819318},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016660612815788933,"output_power":0.00017273235294844007,"reference_active":true,"delta_db":0.15682717274263996},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.188726605138031e-05,"output_power":3.485834737673664e-05,"reference_active":true,"delta_db":-0.7977522173248002}],"active_band_count":3,"active_band_mae_db":0.5013388633164574}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0001893291578274055,"log_rms_mae_db":1.3188212234597156,"rms_envelope_nrmse":0.11476130974206546,"envelope_correlation":0.9910726927340355,"rms_gain_error_db":0.4759417466959681,"reference_mean_power":0.0034133051942474998,"output_mean_power":0.0038086344931285063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.15224102535942e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007711766819603955,"output_power":0.0008664448879500366,"reference_active":true,"delta_db":0.5058705436644249},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.304889773698623e-05,"output_power":7.661041344400855e-05,"reference_active":true,"delta_db":0.20674139023688062},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5271182055851987e-06,"output_power":2.195633154788938e-06,"reference_active":true,"delta_db":-0.6106577630619722}],"active_band_count":3,"active_band_mae_db":0.4410898989877592}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00022567359951457066,"log_rms_mae_db":2.773077579882128,"rms_envelope_nrmse":0.25681912262342166,"envelope_correlation":0.9772190644946881,"rms_gain_error_db":-1.9915946207638786,"reference_mean_power":0.002711357875845493,"output_mean_power":0.0017140653764952164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.6613986227009822e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012279896352023918,"output_power":0.0009483600255813184,"reference_active":true,"delta_db":-1.1222146147371517},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188832719696088,"output_power":6.174887607368494e-05,"reference_active":true,"delta_db":-2.844916923047804},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.234146057131371e-05,"output_power":2.1093789784645296e-05,"reference_active":true,"delta_db":-1.8560501554026567}],"active_band_count":3,"active_band_mae_db":1.9410605643958707}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.00020374269470838248,"log_rms_mae_db":2.408530001354025,"rms_envelope_nrmse":0.23725686820416397,"envelope_correlation":0.9853082945292264,"rms_gain_error_db":-1.9915901852829665,"reference_mean_power":0.0027514125536590357,"output_mean_power":0.0017393889066750613},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.4886745667367608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372871645153378,"output_power":0.0008774332671744461,"reference_active":true,"delta_db":-1.1265604139667302},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.907501205097419e-05,"output_power":2.894257707644618e-05,"reference_active":true,"delta_db":-2.2932321639211466},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.526885866477726e-06,"output_power":1.671656377942829e-06,"reference_active":true,"delta_db":-1.7943861665140255}],"active_band_count":3,"active_band_mae_db":1.7380595814673008}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":3.3446829664283935,"rms_envelope_nrmse":0.2330616717058804,"envelope_correlation":0.9674787387050582,"rms_gain_error_db":-1.0378316978812738,"reference_mean_power":9.509723574878586e-05,"output_mean_power":7.488325664074112e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":5.720070068605407e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5023814010404594e-05,"output_power":2.7684474286698563e-05,"reference_active":true,"delta_db":-1.0212715705606916},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.975419310826886e-06,"output_power":5.6591271539982625e-06,"reference_active":true,"delta_db":0.5591976470190136},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4719997044746735e-06,"output_power":1.497759496917294e-06,"reference_active":true,"delta_db":0.07534359212948884}],"active_band_count":3,"active_band_mae_db":0.5519376032363981}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":2.8490857949653448,"rms_envelope_nrmse":0.19600530038889483,"envelope_correlation":0.9784916538630892,"rms_gain_error_db":-1.0377806651592345,"reference_mean_power":9.649976412079661e-05,"output_mean_power":7.598855486456013e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":5.0877964817728856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.168565282093968e-05,"output_power":2.5105235828143273e-05,"reference_active":true,"delta_db":-1.010983540310571},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3362275317320777e-06,"output_power":2.9427332269904884e-06,"reference_active":true,"delta_db":1.0023575533314606},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1955736150290604e-07,"output_power":1.1473673976921739e-07,"reference_active":true,"delta_db":-0.17873816784271424}],"active_band_count":3,"active_band_mae_db":0.7306930871615819}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"golf","status":"measured","mos":{"status":"measured","value":2.2496113777160645,"predicted_mos":2.2496113777160645,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49440,"input_sample_rate":24000,"input_seconds":2.06,"input_float32_sha256":"378f68056dd8187597effb7f418aed742aaa2c1536cae373ecf458d4c21a59f4","internal_samples":32960,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.418472852001287},"started_at":"2026-10-13T05:02:14.618097+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0014151780005704495,"native_input_samples":49560,"native_input_sha256":"1cfc8bdca50ae9e884f68fd64431b39b05df907db22838a1084a2d49fbfcf73c","output_sha256":"1ee3e637a3957e2d12df476f1e791d4ee0101a5505f14f47420cf1c1ba1db4ca","output_samples":49440,"length_delta_samples":-120,"analysis_s":0.20810612999957812,"synthesis_s":0.05551337399992917,"total_s":0.2636195039995073,"rtf":0.12766077675520934,"output_peak":1.1158907413482666,"output_rms":0.07695036375187962,"finished_at":"2026-10-13T05:02:14.887331+00:00","analysis_stages":{"external_f0_s":0.020263022999643},"evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"1ee3e637a3957e2d12df476f1e791d4ee0101a5505f14f47420cf1c1ba1db4ca","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":33040,"output_samples":32960,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.27006561114477,"output_peak":1.1166599889802213},"common_prefix":{"samples":32960,"sample_rate":16000,"reference_coverage":0.9975786924939467,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":32960,"output_samples":32960,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.145696238613125,"spectral_convergence":0.5745519366723197,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.234418908368604,"spectral_convergence":0.2730942351896806,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.484369471612575,"spectral_convergence":0.3089924991294206,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021738217175554296,"output_power":0.00021818131018965867,"delta_db":0.015936221287018926,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010819735923207302,"output_power":0.0007891635549056154,"delta_db":-1.370496406078131,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002997399842472293,"output_power":0.0023674074177746384,"delta_db":-1.0247167609963528,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020630300721186405,"output_power":0.001743521211060238,"delta_db":-0.730783233720989,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009775038639760088,"output_power":0.000796953420847622,"delta_db":-0.8868554365674602,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.943883643343838e-05,"output_power":1.3936755360846483e-05,"delta_db":-1.4450858873263104,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.912312324329377,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.936912826801728,"pitch":null,"output_rms_dbfs":-22.27006561114477,"output_peak":1.1166599889802213,"reference_rms_dbfs":-21.333152784343042,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.954828206198101},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":32960,"output_samples":32960,"intelligibility":{"estoi":{"value":0.8966566041549634,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08577137272468394,"rms_dbfs":-21.333152784343042,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.754463833117693,"mean":-9.578408546818113e-05,"zero_energy":false},"output":{"rms":0.07700106278577341,"rms_dbfs":-22.27006561114477,"peak":1.1166599889802213,"peak_dbfs":0.9584191044144669,"crest_factor_db":23.228484715559237,"mean":0.00045397559646967933,"zero_energy":false},"rms_difference_db":-0.936912826801727},"full_scale":{"reference":{"sample_count":32960,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":32960,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":4,"at_or_above_full_scale_fraction":0.00012135922330097087,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.00012135922330097087,"longest_full_scale_run_samples":1,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":2.9543561064441217e-06,"output_rms_dbfs":-55.295371577232835,"output_zero_energy":false,"output_peak":0.008799239781313589,"output_energy_fraction":0.00011368418193074009,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":67,"both_estimated_voiced_frames":50,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":4.705042819825248,"absolute_difference_cents_p95":10.457044758550449,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":32960,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":411,"rms_floor":0.00032988134947798775,"log_rms_mae_db":1.741466672218665,"rms_envelope_nrmse":0.15592735957006396,"envelope_correlation":0.9856678222506998,"rms_gain_error_db":-0.9369116762643939,"reference_mean_power":0.007374621729336458,"output_mean_power":0.005943586410872379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":411,"frequency_resolution_hz":0.48661800486618,"power_floor":2.9819548804955484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020723728128806056,"output_power":0.0018228410389908999,"reference_active":true,"delta_db":-0.5571908859560799},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002883193684023648,"output_power":0.0002219278128715057,"reference_active":true,"delta_db":-1.1366208478117825},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.003411334202304e-05,"output_power":6.907431878101383e-05,"reference_active":true,"delta_db":-0.05993024144439828}],"active_band_count":3,"active_band_mae_db":0.5845806584040869}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":405,"rms_floor":0.0002960474071300256,"log_rms_mae_db":1.3133870656483746,"rms_envelope_nrmse":0.1157003350246787,"envelope_correlation":0.9966685635698718,"rms_gain_error_db":-0.9369130603655105,"reference_mean_power":0.007483846506198729,"output_mean_power":0.006031614346027714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":405,"frequency_resolution_hz":0.4938271604938272,"power_floor":2.5610498591405636e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001875541722572381,"output_power":0.001659386148605864,"reference_active":true,"delta_db":-0.5317926915641047},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001255848862457006,"output_power":0.00010151364042974256,"reference_active":true,"delta_db":-0.9241297404357846},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.909720445112925e-06,"output_power":3.289212526213647e-06,"reference_active":true,"delta_db":-0.7505376985211611}],"active_band_count":3,"active_band_mae_db":0.7354867101736836}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":411,"rms_floor":0.00021933697126022545,"log_rms_mae_db":1.8180801237941129,"rms_envelope_nrmse":0.17120974104971506,"envelope_correlation":0.9832718791366172,"rms_gain_error_db":-1.0418345722841587,"reference_mean_power":0.0033717868958789135,"output_mean_power":0.0026526299039250677},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":411,"frequency_resolution_hz":0.48661800486618,"power_floor":1.3601824635936952e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000853724459789269,"output_power":0.0007474429159463898,"reference_active":true,"delta_db":-0.5773969430829433},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016570531712712153,"output_power":0.0001232700587879923,"reference_active":true,"delta_db":-1.28478841120101},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1603623176265326e-05,"output_power":3.515400497387269e-05,"reference_active":true,"delta_db":-0.7315634427220579}],"active_band_count":3,"active_band_mae_db":0.864582932335337}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":405,"rms_floor":0.0001893291578274055,"log_rms_mae_db":1.371381719514885,"rms_envelope_nrmse":0.13079526996250593,"envelope_correlation":0.9951395047317485,"rms_gain_error_db":-1.041832869084437,"reference_mean_power":0.0034217324351337322,"output_mean_power":0.0026919237842056927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":405,"frequency_resolution_hz":0.4938271604938272,"power_floor":1.1516542528399025e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007693710858026342,"output_power":0.0006765021755161611,"reference_active":true,"delta_db":-0.5586666324221512},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.223650307404179e-05,"output_power":5.715797967754566e-05,"reference_active":true,"delta_db":-1.0167984418583953},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.500445400506234e-06,"output_power":1.5911740092353817e-06,"reference_active":true,"delta_db":-1.9629969950917059}],"active_band_count":3,"active_band_mae_db":1.1794873564574175}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":411,"rms_floor":0.00022567359951457066,"log_rms_mae_db":1.8866140617246345,"rms_envelope_nrmse":0.1692796876221609,"envelope_correlation":0.9838314940702817,"rms_gain_error_db":-0.8798134445782007,"reference_mean_power":0.0027179539130431055,"output_mean_power":0.0022195285915193514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":411,"frequency_resolution_hz":0.48661800486618,"power_floor":1.6647011014944542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001228841491632942,"output_power":0.0010332152047854638,"reference_active":true,"delta_db":-0.7530507819558158},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001186051194734395,"output_power":0.00010024767071526556,"reference_active":true,"delta_db":-0.7302914488065738},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2380473081131186e-05,"output_power":3.845426040147121e-05,"reference_active":true,"delta_db":0.7466127335961221}],"active_band_count":3,"active_band_mae_db":0.7433183214528372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":405,"rms_floor":0.00020374269470838248,"log_rms_mae_db":1.4372064069157844,"rms_envelope_nrmse":0.11782931163770974,"envelope_correlation":0.9958299550561374,"rms_gain_error_db":-0.8798150022847633,"reference_mean_power":0.002758205118117753,"output_mean_power":0.0022523976200554473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":405,"frequency_resolution_hz":0.4938271604938272,"power_floor":1.4916349005092608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372967390103534,"output_power":0.0009637625043763981,"reference_active":true,"delta_db":-0.7190376802091933},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.872281514305466e-05,"output_power":4.497916922249029e-05,"reference_active":true,"delta_db":-0.3472094370309061},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.488015434158336e-06,"output_power":2.4579744051217388e-06,"reference_active":true,"delta_db":-0.0527571386294912}],"active_band_count":3,"active_band_mae_db":0.3730014186231969}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":411,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":2.244306817783748,"rms_envelope_nrmse":0.13441632464437642,"envelope_correlation":0.9877269922629754,"rms_gain_error_db":-0.013512228009976543,"reference_mean_power":9.532836067639418e-05,"output_mean_power":9.503222594721816e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":411,"frequency_resolution_hz":0.48661800486618,"power_floor":5.7403836552837704e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.516144572155743e-05,"output_power":3.840869224030185e-05,"reference_active":true,"delta_db":0.3836279681437946},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.921813537188541e-06,"output_power":4.3096896303479386e-06,"reference_active":true,"delta_db":-0.5767916161155404},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47560744395903e-06,"output_power":1.763328729388074e-06,"reference_active":true,"delta_db":0.7736244596570421}],"active_band_count":3,"active_band_mae_db":0.5780146813054591}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":32960,"uncovered_tail_samples":0,"metrics":{"frame_count":405,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":1.6018462040968828,"rms_envelope_nrmse":0.06020237841309605,"envelope_correlation":0.9977229464173967,"rms_gain_error_db":-0.01344447378811249,"reference_mean_power":9.67378123493293e-05,"output_mean_power":9.643880372897675e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":405,"frequency_resolution_hz":0.4938271604938272,"power_floor":5.104720484723869e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.177976736530917e-05,"output_power":3.514707067557834e-05,"reference_active":true,"delta_db":0.43738420937120703},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.285461893325413e-06,"output_power":1.991778448488031e-06,"reference_active":true,"delta_db":-0.597329555800161},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.199021436818828e-07,"output_power":1.2360964026771633e-07,"reference_active":true,"delta_db":0.13225394787715083}],"active_band_count":3,"active_band_mae_db":0.38898923768283966}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.552696704864502,"predicted_mos":2.552696704864502,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49680,"input_sample_rate":24000,"input_seconds":2.07,"input_float32_sha256":"65720eb1f37b6babaa7a0fac2ba17ff6ec61ee6c5b2efba613e855975a90e34b","internal_samples":33120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.464281775000927},"started_at":"2026-10-13T05:07:39.602363+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002144692999536346,"native_input_samples":49560,"native_input_sha256":"1cfc8bdca50ae9e884f68fd64431b39b05df907db22838a1084a2d49fbfcf73c","output_sha256":"cbd6be6a670edfcf88fbd58fc468140a0c10bf18da2839f82cf994e327201447","output_samples":49680,"length_delta_samples":120,"analysis_s":0.8853682040007698,"synthesis_s":4.186870485999862,"total_s":5.072238690000631,"rtf":2.4562899225184656,"output_peak":0.904175341129303,"output_rms":0.08109882398081775,"finished_at":"2026-10-13T05:07:44.683285+00:00","analysis_stages":{"external_f0_s":0.5746396470003674},"evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"cbd6be6a670edfcf88fbd58fc468140a0c10bf18da2839f82cf994e327201447","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":33040,"output_samples":33120,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.81433731244563,"output_peak":0.8943753857730132},"common_prefix":{"samples":33040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975845410628019,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.734482713152143,"spectral_convergence":0.5823945736975509,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.271139454104429,"spectral_convergence":0.3094651291921608,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.230448206959294,"spectral_convergence":0.3066166937869029,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.0002370602939928177,"delta_db":0.3868699808169641,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.0010459701848142102,"delta_db":-0.13731131125324578,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.0027183235664088893,"delta_db":-0.41359645493625885,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.001799358924259299,"delta_db":-0.5833507475683015,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0007873984008489723,"delta_db":-0.9287068783640589,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":1.2989417863641952e-05,"delta_db":-1.7403568135945922,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6983653644222368,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.460156401276826,"pitch":null,"output_rms_dbfs":-21.803836362198354,"output_peak":0.8943753857730132,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.745356791405289},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":33040,"output_samples":33040,"intelligibility":{"estoi":{"value":0.8890513167381008,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08566748199211072,"rms_dbfs":-21.343679960921527,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.764991009696175,"mean":-9.639823119230478e-05,"zero_energy":false},"output":{"rms":0.08124715864685872,"rms_dbfs":-21.803836362198354,"peak":0.8943753857730132,"peak_dbfs":-0.9696032314172012,"crest_factor_db":20.83423313078115,"mean":-7.887041052591096e-05,"zero_energy":false},"rms_difference_db":-0.46015640127682644},"full_scale":{"reference":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":4.46462840342475e-06,"output_rms_dbfs":-53.50214682141161,"output_zero_energy":false,"output_peak":0.055885291275712914,"output_energy_fraction":0.00015393832475997589,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":55,"both_estimated_voiced_frames":43,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.86,"absolute_difference_cents_median":6.042807577323712,"absolute_difference_cents_p95":24.10501606532148,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":33040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00032988134947798775,"log_rms_mae_db":2.504759737447901,"rms_envelope_nrmse":0.18220153526115612,"envelope_correlation":0.9699488154061334,"rms_gain_error_db":-0.46015719608941125,"reference_mean_power":0.007356724541990913,"output_mean_power":0.006617116417761802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":2.980285612038863e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020727352824950612,"output_power":0.0018810556225908855,"reference_active":true,"delta_db":-0.42142202333534223},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028987718896809884,"output_power":0.0002812181601905718,"reference_active":true,"delta_db":-0.13170678539282396},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.012977878291207e-05,"output_power":5.558691942953136e-05,"reference_active":true,"delta_db":-1.0092986241667987}],"active_band_count":3,"active_band_mae_db":0.5208091442983216}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0002960474071300256,"log_rms_mae_db":1.7919244173443178,"rms_envelope_nrmse":0.10011683466380396,"envelope_correlation":0.9920151658902525,"rms_gain_error_db":-0.4601529002690906,"reference_mean_power":0.007465415825536721,"output_mean_power":0.006714887067762041},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":2.5597102021075326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001877761726334087,"output_power":0.0016985523215786254,"reference_active":true,"delta_db":-0.4356155350419091},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012702101262043313,"output_power":0.000130345401676688,"reference_active":true,"delta_db":0.11220144096782382},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.962314408759722e-06,"output_power":2.8223151116513886e-06,"reference_active":true,"delta_db":-1.4734343260921783}],"active_band_count":3,"active_band_mae_db":0.6737504340339703}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00021933697126022545,"log_rms_mae_db":2.37978810291912,"rms_envelope_nrmse":0.19606542575327307,"envelope_correlation":0.9655864883099304,"rms_gain_error_db":-0.360473110881107,"reference_mean_power":0.003363603458339666,"output_mean_power":0.0030956901067176437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.3609823101859838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008546335077009212,"output_power":0.0008207354055025798,"reference_active":true,"delta_db":-0.17576747491839517},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016660612815788933,"output_power":0.00016458510201715403,"reference_active":true,"delta_db":-0.053004507114002675},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.188726605138031e-05,"output_power":2.9674841842637217e-05,"reference_active":true,"delta_db":-1.4969360224710961}],"active_band_count":3,"active_band_mae_db":0.5752360015011647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0001893291578274055,"log_rms_mae_db":1.6187197656222507,"rms_envelope_nrmse":0.09450661204572354,"envelope_correlation":0.9924522541053169,"rms_gain_error_db":-0.3604681338009324,"reference_mean_power":0.0034133051942474998,"output_mean_power":0.0031414366649721256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.15224102535942e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007711766819603955,"output_power":0.0007389669381403311,"reference_active":true,"delta_db":-0.18528881134859337},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.304889773698623e-05,"output_power":7.564503441998025e-05,"reference_active":true,"delta_db":0.15166757744712533},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5271182055851987e-06,"output_power":1.654351090260599e-06,"reference_active":true,"delta_db":-1.8399787446669167}],"active_band_count":3,"active_band_mae_db":0.7256450444875452}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00022567359951457066,"log_rms_mae_db":2.692579058456476,"rms_envelope_nrmse":0.20861848847053344,"envelope_correlation":0.9692009271863573,"rms_gain_error_db":-0.702834219203199,"reference_mean_power":0.002711357875845493,"output_mean_power":0.0023062342764906423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.6613986227009822e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012279896352023918,"output_power":0.0010344824243041545,"reference_active":true,"delta_db":-0.744715847067562},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188832719696088,"output_power":0.0001011342558843669,"reference_active":true,"delta_db":-0.702224660661741},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.234146057131371e-05,"output_power":2.9556515014744704e-05,"reference_active":true,"delta_db":-0.39106403744368584}],"active_band_count":3,"active_band_mae_db":0.6126681817243297}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.00020374269470838248,"log_rms_mae_db":1.8286618703350355,"rms_envelope_nrmse":0.13400088894959458,"envelope_correlation":0.9894930651223431,"rms_gain_error_db":-0.7028319335014016,"reference_mean_power":0.0027514125536590357,"output_mean_power":0.0023403053267679933},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.4886745667367608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372871645153378,"output_power":0.0009585963100059218,"reference_active":true,"delta_db":-0.7423438480482297},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.907501205097419e-05,"output_power":4.4804553554584074e-05,"reference_active":true,"delta_db":-0.3953826067177022},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.526885866477726e-06,"output_power":1.6622125255168925e-06,"reference_active":true,"delta_db":-1.8189907570692538}],"active_band_count":3,"active_band_mae_db":0.9855724039450618}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":2.4293270428055047,"rms_envelope_nrmse":0.22580529247483602,"envelope_correlation":0.9654716534650217,"rms_gain_error_db":-0.7048013344469841,"reference_mean_power":9.509723574878586e-05,"output_mean_power":8.085144008870565e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":5.720070068605407e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5023814010404594e-05,"output_power":3.0475584920728235e-05,"reference_active":true,"delta_db":-0.6041138813862053},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.975419310826886e-06,"output_power":4.1926088995511445e-06,"reference_active":true,"delta_db":-0.7434533556955009},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4719997044746735e-06,"output_power":1.5767548987452539e-06,"reference_active":true,"delta_db":0.2985646614588603}],"active_band_count":3,"active_band_mae_db":0.5487106328468555}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":1.6048081449837641,"rms_envelope_nrmse":0.15341492104078555,"envelope_correlation":0.9855424425277226,"rms_gain_error_db":-0.7047946878148024,"reference_mean_power":9.649976412079661e-05,"output_mean_power":8.204399188583596e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":5.0877964817728856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.168565282093968e-05,"output_power":2.8002199850914386e-05,"reference_active":true,"delta_db":-0.5367050851492134},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3362275317320777e-06,"output_power":2.0340231966781217e-06,"reference_active":true,"delta_db":-0.6015923620030467},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1955736150290604e-07,"output_power":1.1217124397774292e-07,"reference_active":true,"delta_db":-0.27694785858930393}],"active_band_count":3,"active_band_mae_db":0.47174843524718807}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.0491178035736084,"predicted_mos":2.0491178035736084,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49800,"input_sample_rate":24000,"input_seconds":2.075,"input_float32_sha256":"8b22e755336f8ed1d51d9bd13e68d41ce7b548279fe768647b6a2e86945adbf7","internal_samples":33200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4298603150000417},"started_at":"2026-10-13T05:10:21.785676+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0012271030009287642,"native_input_samples":49560,"native_input_sha256":"1cfc8bdca50ae9e884f68fd64431b39b05df907db22838a1084a2d49fbfcf73c","output_sha256":"45211bf02d73e1929679c611c60e22d5753589d304dea447b8aba569c0f3ef8c","output_samples":49800,"length_delta_samples":240,"analysis_s":0.23949214800086338,"synthesis_s":0.05367890900015482,"total_s":0.2931710570010182,"rtf":0.14197145617482723,"output_peak":0.9383136630058289,"output_rms":0.08411380995406327,"finished_at":"2026-10-13T05:10:22.084825+00:00","analysis_stages":{"external_f0_s":0.028604856999663753},"evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"45211bf02d73e1929679c611c60e22d5753589d304dea447b8aba569c0f3ef8c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":33040,"output_samples":33200,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.497135127573177,"output_peak":0.9114120278059853},"common_prefix":{"samples":33040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9951807228915662,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.544397462086404,"spectral_convergence":0.15178678997453196,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.185532243551556,"spectral_convergence":0.14780945519614655,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.666640053505145,"spectral_convergence":0.15257737280294031,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00020668773220184277,"delta_db":-0.20857120415541297,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.0010571505439674733,"delta_db":-0.09113598749850371,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.0029578811066258716,"delta_db":-0.04680082387116967,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.00197269340570797,"delta_db":-0.18393284118209952,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.0009095003050281186,"delta_db":-0.30262367216658365,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":1.452174993580816e-05,"delta_db":-1.2560641549512774,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.34818811397084115,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13247489704758694,"pitch":null,"output_rms_dbfs":-21.476154857969117,"output_peak":0.9114120278059853,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.465523253047702},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":33040,"output_samples":33040,"intelligibility":{"estoi":{"value":0.9569775295160606,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08566748199211072,"rms_dbfs":-21.343679960921527,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.764991009696175,"mean":-9.639823119230478e-05,"zero_energy":false},"output":{"rms":0.08437081748725138,"rms_dbfs":-21.476154857969117,"peak":0.9114120278059853,"peak_dbfs":-0.8057048875686366,"crest_factor_db":20.67044997040048,"mean":8.045092157603603e-05,"zero_energy":false},"rms_difference_db":-0.1324748970475902},"full_scale":{"reference":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":2.7288584554280485e-06,"output_rms_dbfs":-55.64018990391216,"output_zero_energy":false,"output_peak":0.02812771470717323,"output_energy_fraction":8.725178034213319e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":49,"both_estimated_voiced_frames":45,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9,"absolute_difference_cents_median":0.6718396327247915,"absolute_difference_cents_p95":3.6684654045978324,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":33040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00032988134947798775,"log_rms_mae_db":0.7521365668016438,"rms_envelope_nrmse":0.05787027141124873,"envelope_correlation":0.9970601399109661,"rms_gain_error_db":-0.13247400437491233,"reference_mean_power":0.007356724541990913,"output_mean_power":0.007135708404604939},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":2.980285612038863e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020727352824950612,"output_power":0.0020101047181895035,"reference_active":true,"delta_db":-0.13325157143561508},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028987718896809884,"output_power":0.0002736451433737786,"reference_active":true,"delta_db":-0.2502629644148594},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.012977878291207e-05,"output_power":5.179029096851875e-05,"reference_active":true,"delta_db":-1.3165411798362956}],"active_band_count":3,"active_band_mae_db":0.5666852385622567}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0002960474071300256,"log_rms_mae_db":0.4387215741085517,"rms_envelope_nrmse":0.03643287203874844,"envelope_correlation":0.9988714297974098,"rms_gain_error_db":-0.13247397744225672,"reference_mean_power":0.007465415825536721,"output_mean_power":0.00724113434945812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":2.5597102021075326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001877761726334087,"output_power":0.001831088684932654,"reference_active":true,"delta_db":-0.10931103811467036},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012702101262043313,"output_power":0.00011776279370665238,"reference_active":true,"delta_db":-0.3286747068595946},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.962314408759722e-06,"output_power":3.0880270727108037e-06,"reference_active":true,"delta_db":-1.0826783460264409}],"active_band_count":3,"active_band_mae_db":0.5068880303335687}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00021933697126022545,"log_rms_mae_db":0.8052376268118743,"rms_envelope_nrmse":0.08341392080209573,"envelope_correlation":0.9937820263748369,"rms_gain_error_db":-0.051945187882602155,"reference_mean_power":0.003363603458339666,"output_mean_power":0.0033236116419477863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.3609823101859838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008546335077009212,"output_power":0.0008456996985926775,"reference_active":true,"delta_db":-0.04563740433157952},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016660612815788933,"output_power":0.0001605422299006918,"reference_active":true,"delta_db":-0.16101680768415688},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.188726605138031e-05,"output_power":3.2404160353378975e-05,"reference_active":true,"delta_db":-1.1148124259986694}],"active_band_count":3,"active_band_mae_db":0.44048887933813524}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0001893291578274055,"log_rms_mae_db":0.42864373972549463,"rms_envelope_nrmse":0.051577902221268616,"envelope_correlation":0.9973954066326931,"rms_gain_error_db":-0.05194505519584189,"reference_mean_power":0.0034133051942474998,"output_mean_power":0.0033727225483713183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.15224102535942e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007711766819603955,"output_power":0.0007675429591065643,"reference_active":true,"delta_db":-0.020511972105464252},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.304889773698623e-05,"output_power":7.007018357750872e-05,"reference_active":true,"delta_db":-0.18080411775575084},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5271182055851987e-06,"output_power":2.127936529245166e-06,"reference_active":true,"delta_db":-0.7466688650297402}],"active_band_count":3,"active_band_mae_db":0.31599498496365175}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00022567359951457066,"log_rms_mae_db":1.1632293222806926,"rms_envelope_nrmse":0.06894616677276034,"envelope_correlation":0.9968706677193799,"rms_gain_error_db":-0.2355340190675989,"reference_mean_power":0.002711357875845493,"output_mean_power":0.002568227207822172},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.6613986227009822e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012279896352023918,"output_power":0.0011581100121561255,"reference_active":true,"delta_db":-0.2544488495921948},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188832719696088,"output_power":0.00011289590659904292,"reference_active":true,"delta_db":-0.22442553988591996},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.234146057131371e-05,"output_power":2.2653528326724432e-05,"reference_active":true,"delta_db":-1.5462377542423966}],"active_band_count":3,"active_band_mae_db":0.6750373812401705}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.00020374269470838248,"log_rms_mae_db":0.6475171056553305,"rms_envelope_nrmse":0.04230008280546799,"envelope_correlation":0.9991046302305393,"rms_gain_error_db":-0.23553669256534557,"reference_mean_power":0.0027514125536590357,"output_mean_power":0.002606165823128171},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.4886745667367608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372871645153378,"output_power":0.0010744595878516795,"reference_active":true,"delta_db":-0.24680052080012055},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.907501205097419e-05,"output_power":4.7440883828517886e-05,"reference_active":true,"delta_db":-0.14707643536229417},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.526885866477726e-06,"output_power":1.7375669875279604e-06,"reference_active":true,"delta_db":-1.6264406958021822}],"active_band_count":3,"active_band_mae_db":0.6734392173215323}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":1.1410626895278162,"rms_envelope_nrmse":0.06609900682724035,"envelope_correlation":0.9972202861151557,"rms_gain_error_db":-0.18656165291361218,"reference_mean_power":9.509723574878586e-05,"output_mean_power":9.109860526840837e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":5.720070068605407e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5023814010404594e-05,"output_power":3.3847132952151834e-05,"reference_active":true,"delta_db":-0.14841550637676137},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.975419310826886e-06,"output_power":4.728466510206071e-06,"reference_active":true,"delta_db":-0.22109369981368077},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4719997044746735e-06,"output_power":1.0576930845605755e-06,"reference_active":true,"delta_db":-1.4354805796470171}],"active_band_count":3,"active_band_mae_db":0.6016632619458198}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":0.8211759432064748,"rms_envelope_nrmse":0.0396370937081794,"envelope_correlation":0.9992393320725905,"rms_gain_error_db":-0.18653928140120926,"reference_mean_power":9.649976412079661e-05,"output_mean_power":9.244263658636463e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":5.0877964817728856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.168565282093968e-05,"output_power":3.104890958367681e-05,"reference_active":true,"delta_db":-0.08816306685028458},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3362275317320777e-06,"output_power":2.1820689035693673e-06,"reference_active":true,"delta_db":-0.296466773929639},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1955736150290604e-07,"output_power":8.511060193887851e-08,"reference_active":true,"delta_db":-1.475926600358559}],"active_band_count":3,"active_band_mae_db":0.6201854803794942}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0029","method":"magphase","status":"measured","mos":{"status":"measured","value":2.5098514556884766,"predicted_mos":2.5098514556884766,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":99305,"input_sample_rate":48000,"input_seconds":2.0688541666666667,"input_float32_sha256":"a7c5c0590afe54485b01c5014dfd38360e24c793b588b20c8a150907933c04b4","internal_samples":33102,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4522193419998075},"started_at":"2026-10-13T05:13:51.169956+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.004105407000679406,"native_input_samples":99120,"native_input_sha256":"7ce5e6d7763cb10288ad34ccd61f5ec04366c2e342550fff2186cf4b6c454fe7","output_sha256":"7426886a0d151e57d01d66188369a845614c619b28b2cfb1c45c2c4dac04b4aa","output_samples":99305,"length_delta_samples":185,"analysis_s":1.0819803069989575,"synthesis_s":0.5903973580007005,"total_s":1.672377664999658,"rtf":0.8098681186439022,"output_peak":1.0415231745669273,"output_rms":0.08504795825133149,"finished_at":"2026-10-13T05:13:52.852084+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"7426886a0d151e57d01d66188369a845614c619b28b2cfb1c45c2c4dac04b4aa","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":33040,"output_samples":33102,"length_delta_samples":62,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.401619273913287,"output_peak":1.027010708118166},"common_prefix":{"samples":33040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981270013896442,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":33040,"output_samples":33040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.367763082969325,"spectral_convergence":0.22964514485997423,"spectral_convergence_status":"ok","amplitude_floor":1.5756171103077008e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.574549479723547,"spectral_convergence":0.11382592290231766,"spectral_convergence_status":"ok","amplitude_floor":1.2505050051947005e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.962552071552397,"spectral_convergence":0.17514973769059802,"spectral_convergence_status":"ok","amplitude_floor":9.121352204658788e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021685618865241799,"output_power":0.00019952911242597708,"delta_db":-0.36165549911508077,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010795690773652512,"output_power":0.0010300735516666866,"delta_db":-0.2038220004964887,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002989928450689294,"output_power":0.002848152093202924,"delta_db":-0.21097618538684884,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020580355715967367,"output_power":0.002091211740823077,"delta_db":0.06945131585728131,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0009751360590152492,"output_power":0.001071451767997927,"delta_db":0.4090740953843675,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9392123549665832e-05,"output_power":1.482968814713562e-05,"delta_db":-1.1649335107523133,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4033187678320634,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04979744497515548,"pitch":null,"output_rms_dbfs":-21.393477405896686,"output_peak":1.027010708118166,"reference_rms_dbfs":-21.343679960921527,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.96828821141509},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":33040,"output_samples":33040,"intelligibility":{"estoi":{"value":0.9691453434955134,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08566748199211072,"rms_dbfs":-21.343679960921527,"peak":0.935546875,"peak_dbfs":-0.5786889512253508,"crest_factor_db":20.764991009696175,"mean":-9.639823119230478e-05,"zero_energy":false},"output":{"rms":0.08517774330342245,"rms_dbfs":-21.393477405896686,"peak":1.027010708118166,"peak_dbfs":0.23149943577004176,"crest_factor_db":21.62497684166673,"mean":0.0007905581600604536,"zero_energy":false},"rms_difference_db":-0.04979744497515881},"full_scale":{"reference":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":33040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":1,"at_or_above_full_scale_fraction":3.026634382566586e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":3.026634382566586e-05,"longest_full_scale_run_samples":1,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":206,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0839493764797227e-05,"threshold_dbfs":-49.649910001121114,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":47,"quiet_samples":7520,"reference_mean_power":2.162107093101169e-06,"output_mean_power":2.034198196392772e-06,"output_rms_dbfs":-56.916067350876546,"output_zero_energy":false,"output_peak":0.01794436287937264,"output_energy_fraction":6.381441722196987e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":199,"reference_estimated_voiced_frames":50,"output_estimated_voiced_frames":50,"both_estimated_voiced_frames":48,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.96,"absolute_difference_cents_median":1.9707167971848263,"absolute_difference_cents_p95":5.845272372261881,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":33040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00032988134947798775,"log_rms_mae_db":0.9571650829476412,"rms_envelope_nrmse":0.04457791660796925,"envelope_correlation":0.9982826666472211,"rms_gain_error_db":-0.04979597575316561,"reference_mean_power":0.007356724541990913,"output_mean_power":0.0072728544709753555},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":2.980285612038863e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0020727352824950612,"output_power":0.0021406964665669067,"reference_active":true,"delta_db":0.1401125210728798},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028987718896809884,"output_power":0.00028854474517461435,"reference_active":true,"delta_db":-0.02000871600396064},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.012977878291207e-05,"output_power":7.507989075935718e-05,"reference_active":true,"delta_db":0.2962116309055552}],"active_band_count":3,"active_band_mae_db":0.15211095599413188}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0002960474071300256,"log_rms_mae_db":0.6524907508278851,"rms_envelope_nrmse":0.02605233886359563,"envelope_correlation":0.9994496064754321,"rms_gain_error_db":-0.049795696055034075,"reference_mean_power":0.007465415825536721,"output_mean_power":0.0073803070988026995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":2.5597102021075326e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001877761726334087,"output_power":0.0019506371126143814,"reference_active":true,"delta_db":0.16536000035860637},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012702101262043313,"output_power":0.0001287771003496842,"reference_active":true,"delta_db":0.05963071294151804},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.962314408759722e-06,"output_power":4.0294416884550536e-06,"reference_active":true,"delta_db":0.07295941577560622}],"active_band_count":3,"active_band_mae_db":0.0993167096919102}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00021933697126022545,"log_rms_mae_db":1.0215096398214245,"rms_envelope_nrmse":0.045286205536962186,"envelope_correlation":0.9985804123401328,"rms_gain_error_db":-0.1942843148284472,"reference_mean_power":0.003363603458339666,"output_mean_power":0.0032164466984069932},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.3609823101859838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008546335077009212,"output_power":0.0008178057492540894,"reference_active":true,"delta_db":-0.1912975688035597},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016660612815788933,"output_power":0.00016443198083611706,"reference_active":true,"delta_db":-0.05704683149492412},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.188726605138031e-05,"output_power":4.2334758503373105e-05,"reference_active":true,"delta_db":0.04615071472416243}],"active_band_count":3,"active_band_mae_db":0.09816503834088208}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.0001893291578274055,"log_rms_mae_db":0.7930806824597072,"rms_envelope_nrmse":0.02777658253713623,"envelope_correlation":0.9997178238173514,"rms_gain_error_db":-0.1942821249222772,"reference_mean_power":0.0034133051942474998,"output_mean_power":0.003263975642381487},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.15224102535942e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007711766819603955,"output_power":0.0007386467303707258,"reference_active":true,"delta_db":-0.18717109590703443},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.304889773698623e-05,"output_power":7.404504847622686e-05,"reference_active":true,"delta_db":0.05882354702948514},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5271182055851987e-06,"output_power":2.4905429324557224e-06,"reference_active":true,"delta_db":-0.06331523871697098}],"active_band_count":3,"active_band_mae_db":0.10310329388449685}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":0.00022567359951457066,"log_rms_mae_db":0.959810964266001,"rms_envelope_nrmse":0.04254370346716958,"envelope_correlation":0.9989031606034355,"rms_gain_error_db":0.1482683533103654,"reference_mean_power":0.002711357875845493,"output_mean_power":0.0028055220075521897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":1.6613986227009822e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012279896352023918,"output_power":0.0013002624773041653,"reference_active":true,"delta_db":0.24836328776507352},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001188832719696088,"output_power":0.0001248662221775393,"reference_active":true,"delta_db":0.21324222886133287},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.234146057131371e-05,"output_power":3.5283369535078536e-05,"reference_active":true,"delta_db":0.37810424066803283}],"active_band_count":3,"active_band_mae_db":0.27990325243147973}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":0.00020374269470838248,"log_rms_mae_db":0.5699246446630426,"rms_envelope_nrmse":0.02576230396112878,"envelope_correlation":0.9997290359828519,"rms_gain_error_db":0.14826891923526364,"reference_mean_power":0.0027514125536590357,"output_mean_power":0.0028469681357033943},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":1.4886745667367608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011372871645153378,"output_power":0.001206239682328479,"reference_active":true,"delta_db":0.2556347389107496},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.907501205097419e-05,"output_power":5.2906732076277926e-05,"reference_active":true,"delta_db":0.3265052209897974},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.526885866477726e-06,"output_power":2.7165230450521177e-06,"reference_active":true,"delta_db":0.3142776733399005}],"active_band_count":3,"active_band_mae_db":0.29880587774681583}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":412,"rms_floor":4.7181572312282426e-05,"log_rms_mae_db":0.9634401095159641,"rms_envelope_nrmse":0.09377297761989228,"envelope_correlation":0.9973535028991578,"rms_gain_error_db":0.5755545014124385,"reference_mean_power":9.509723574878586e-05,"output_mean_power":0.00010857337632070361},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":412,"frequency_resolution_hz":0.4854368932038835,"power_floor":5.720070068605407e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5023814010404594e-05,"output_power":4.115278468610438e-05,"reference_active":true,"delta_db":0.7003578999058652},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.975419310826886e-06,"output_power":5.640957027423735e-06,"reference_active":true,"delta_db":0.5452310386625956},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4719997044746735e-06,"output_power":1.799945914859705e-06,"reference_active":true,"delta_db":0.8735173272003369}],"active_band_count":3,"active_band_mae_db":0.7063687552562659}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":33040,"uncovered_tail_samples":0,"metrics":{"frame_count":406,"rms_floor":4.01547990231072e-05,"log_rms_mae_db":0.6947271034525345,"rms_envelope_nrmse":0.0824182984337278,"envelope_correlation":0.9985829184187291,"rms_gain_error_db":0.5755034271158753,"reference_mean_power":9.649976412079661e-05,"output_mean_power":0.00011017336000160743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":406,"frequency_resolution_hz":0.4926108374384236,"power_floor":5.0877964817728856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.168565282093968e-05,"output_power":3.764899935914739e-05,"reference_active":true,"delta_db":0.748907786027586},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3362275317320777e-06,"output_power":2.7024697376643247e-06,"reference_active":true,"delta_db":0.6324570175596522},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1955736150290604e-07,"output_power":1.4157592283386845e-07,"reference_active":true,"delta_db":0.7341307908798882}],"active_band_count":3,"active_band_mae_db":0.7051651981557088}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"hnm","status":"measured","mos":{"status":"measured","value":2.0558533668518066,"predicted_mos":2.0558533668518066,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7dc3a66061bb62cd0b501018654279f240b68694dfd1c0bf67a176990433ec4e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4424107800005004},"started_at":"2026-10-13T04:38:40.552571+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.583199991728179e-05,"native_input_samples":96000,"native_input_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","output_sha256":"c89bdf7bc7868e9fd68db00718e1d02e891b718b559c8abbdd7be43553448e5b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0279799379995893,"synthesis_s":0.509239157000593,"total_s":1.5372190950001823,"rtf":0.2562031825000304,"output_peak":0.8111427281255558,"output_rms":0.08610914394700823,"finished_at":"2026-10-13T04:38:42.094782+00:00","evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"c89bdf7bc7868e9fd68db00718e1d02e891b718b559c8abbdd7be43553448e5b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.937058720009311,"spectral_convergence":0.3708652426229753,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.653736079730179,"spectral_convergence":0.4037261742302693,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.944608024951128,"spectral_convergence":0.39212761224799975,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00015236226974333348,"delta_db":0.6743515242084237,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.001455037181684023,"delta_db":0.07501392869914414,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.004280087582771394,"delta_db":-0.02854226798919223,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0012685486189892626,"delta_db":-0.12300778136420895,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.0002450423328286729,"delta_db":0.013574786749857303,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":1.3706685269899121e-05,"delta_db":1.3601601047871408,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3791083989663278,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.007833154437560252,"pitch":null,"output_rms_dbfs":-21.299014565614108,"output_peak":0.8111427281255558,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.845134274896874},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.937058720009311,"spectral_convergence":0.3708652426229753,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.653736079730179,"spectral_convergence":0.4037261742302693,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.944608024951128,"spectral_convergence":0.39212761224799975,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00015236226974333348,"delta_db":0.6743515242084237,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.001455037181684023,"delta_db":0.07501392869914414,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.004280087582771394,"delta_db":-0.02854226798919223,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0012685486189892626,"delta_db":-0.12300778136420895,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.0002450423328286729,"delta_db":0.013574786749857303,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":1.3706685269899121e-05,"delta_db":1.3601601047871408,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3791083989663278,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.007833154437560252,"pitch":null,"output_rms_dbfs":-21.299014565614108,"output_peak":0.8111427281255558,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.845134274896874},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8530044173442837,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618683437204885,"rms_dbfs":-21.291181411176545,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.435250156246948,"mean":-4.1497230529785154e-05,"zero_energy":false},"output":{"rms":0.08610914394700823,"rms_dbfs":-21.299014565614108,"peak":0.8111427281255558,"peak_dbfs":-1.818054418033087,"crest_factor_db":19.480960147581023,"mean":-0.000246545042005883,"zero_energy":false},"rms_difference_db":-0.00783315443756294},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":1.8720049411058425e-06,"output_mean_power":3.998921559840425e-06,"output_rms_dbfs":-53.980571146118066,"output_zero_energy":false,"output_peak":0.0358261632181178,"output_energy_fraction":9.258280110245701e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":164,"both_estimated_voiced_frames":161,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.8385416666666666,"absolute_difference_cents_median":1.1443301661488634,"absolute_difference_cents_p95":6.8795460410599585,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002631236493882886,"log_rms_mae_db":1.1313897823318912,"rms_envelope_nrmse":0.11776880135201663,"envelope_correlation":0.9864986333879797,"rms_gain_error_db":-0.007833012831769048,"reference_mean_power":0.007434364903539191,"output_mean_power":0.007420968235103399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.450307530079474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015621941098031631,"output_power":0.001582805178506085,"reference_active":true,"delta_db":0.05692466508227019},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031400948871223737,"output_power":0.0003429558095973421,"reference_active":true,"delta_db":0.38295392366027314},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9119711962951405e-05,"output_power":7.450761324092711e-05,"reference_active":true,"delta_db":1.8094484008871765}],"active_band_count":3,"active_band_mae_db":0.7497756632099066}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024033810696394756,"log_rms_mae_db":0.5924276107462026,"rms_envelope_nrmse":0.06374526648373702,"envelope_correlation":0.9955897757671035,"rms_gain_error_db":-0.00783356584601792,"reference_mean_power":0.007471745584669175,"output_mean_power":0.0074582806068237495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.136699874330792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014162642987993615,"output_power":0.0014642828128290733,"reference_active":true,"delta_db":0.14480657239237432},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014955063876354673,"output_power":0.00016247942223198518,"reference_active":true,"delta_db":0.3601009373367811},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0546566248780085e-06,"output_power":4.513172418493799e-06,"reference_active":true,"delta_db":1.6951952724620043}],"active_band_count":3,"active_band_mae_db":0.7333675940637199}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002303175773961387,"log_rms_mae_db":1.7814965345848477,"rms_envelope_nrmse":0.1412205660154931,"envelope_correlation":0.9828644597330674,"rms_gain_error_db":-0.03907643940691108,"reference_mean_power":0.004944923373977616,"output_mean_power":0.004900630089832457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6955719701225893e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012875788279854436,"output_power":0.0012749476280054072,"reference_active":true,"delta_db":-0.0428148155685809},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026650258378785487,"output_power":0.0002968818326453784,"reference_active":true,"delta_db":0.4688219829797189},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0302791253002337e-05,"output_power":6.874046260877565e-05,"reference_active":true,"delta_db":3.5572981641002306}],"active_band_count":3,"active_band_mae_db":1.3563116542161768}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020459712012168156,"log_rms_mae_db":0.9603054955979233,"rms_envelope_nrmse":0.07580272761727996,"envelope_correlation":0.9945135844152835,"rms_gain_error_db":-0.03907636585500317,"reference_mean_power":0.004969790940376444,"output_mean_power":0.004925274992779441},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.46585734788702e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011840763589577269,"output_power":0.0012044250729040624,"reference_active":true,"delta_db":0.07400077621597755},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012134699925967192,"output_power":0.00013008046994050012,"reference_active":true,"delta_db":0.301830559143801},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.482336680977432e-06,"output_power":4.323875414871528e-06,"reference_active":true,"delta_db":2.410124871959582}],"active_band_count":3,"active_band_mae_db":0.9286520691064535}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014083569934759886,"log_rms_mae_db":1.7620143299242486,"rms_envelope_nrmse":0.15247541225809244,"envelope_correlation":0.9816877151208534,"rms_gain_error_db":-0.19176058928692744,"reference_mean_power":0.0013181968247743341,"output_mean_power":0.0012612587792861492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.706116677050863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035629780608750625,"output_power":0.00033403072124995383,"reference_active":true,"delta_db":-0.2802673689039169},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927249258823345e-05,"output_power":7.0508353815922e-05,"reference_active":true,"delta_db":0.0767976030350178},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3444445693793866e-05,"output_power":1.8759493346484116e-05,"reference_active":true,"delta_db":1.4467820360802452}],"active_band_count":3,"active_band_mae_db":0.6012823360063934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010879087849669347,"log_rms_mae_db":1.0367554480336498,"rms_envelope_nrmse":0.08955719396103885,"envelope_correlation":0.9934238453043458,"rms_gain_error_db":-0.19176340611954557,"reference_mean_power":0.0013248216670349025,"output_mean_power":0.0012675966466916802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.107952744901023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000335569442501778,"output_power":0.00031848191857043604,"reference_active":true,"delta_db":-0.22697625574501404},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4091376965670345e-05,"output_power":3.646826182158947e-05,"reference_active":true,"delta_db":0.29270521200861366},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.097157965658072e-07,"output_power":1.1294737580223818e-06,"reference_active":true,"delta_db":1.4454353314502875}],"active_band_count":3,"active_band_mae_db":0.6550389330679717}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":1.977390461338661,"rms_envelope_nrmse":0.5261753175118735,"envelope_correlation":0.804196014817104,"rms_gain_error_db":1.0901617977097788,"reference_mean_power":2.393470871827269e-05,"output_mean_power":3.076410792993627e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3699291941082084e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.018836822609007e-06,"output_power":5.878562271910089e-06,"reference_active":true,"delta_db":1.6517075023104844},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.745214583455975e-06,"output_power":7.010408069350598e-06,"reference_active":true,"delta_db":1.694874421334057},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3999346848964708e-06,"output_power":4.8475433932756244e-06,"reference_active":true,"delta_db":5.394139316143413}],"active_band_count":3,"active_band_mae_db":2.9135737465959846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":1.338329167862101,"rms_envelope_nrmse":0.445271630231449,"envelope_correlation":0.8390019791882372,"rms_gain_error_db":1.0901616377341474,"reference_mean_power":2.405439007374668e-05,"output_mean_power":3.091793737158311e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.116507982063627e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.688364032412793e-06,"output_power":7.587900529847687e-06,"reference_active":true,"delta_db":3.1328785077788406},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3132434905641704e-06,"output_power":5.574156351305112e-06,"reference_active":true,"delta_db":2.2592579281820977},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.200263616160695e-07,"output_power":5.063998423988623e-07,"reference_active":true,"delta_db":6.252169204090098}],"active_band_count":3,"active_band_mae_db":3.8814352133503456}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.094667434692383,"predicted_mos":3.094667434692383,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6e51739168efb749cd4898fc913f025ea3de6fe0363de866777a28d26a05a14d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1716528090000793},"started_at":"2026-10-13T04:41:37.397824+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.163199951610295e-05,"native_input_samples":96000,"native_input_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","output_sha256":"048c43a45d285f99ab7a4c17db8abd9cb304f76b3b56ffa0651af9ed7752585c","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4012806889995773,"synthesis_s":0.10369048600023234,"total_s":1.5049711749998096,"rtf":0.25082852916663495,"output_peak":0.9103883664787014,"output_rms":0.09705710973637274,"finished_at":"2026-10-13T04:41:38.908307+00:00","evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"048c43a45d285f99ab7a4c17db8abd9cb304f76b3b56ffa0651af9ed7752585c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.310177525541947,"spectral_convergence":0.5411751546439693,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.165129682991532,"spectral_convergence":0.257675970224484,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.645469214343054,"spectral_convergence":0.3065152999592203,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00011748603476190836,"delta_db":-0.4545603500969543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.0016928792865640203,"delta_db":0.7325329270315126,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.005716047240094651,"delta_db":1.2278892654305398,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0016118584764257405,"delta_db":0.9171901245368818,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.00027197425863216735,"delta_db":0.46644161924419547,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":9.837253899810489e-06,"delta_db":-0.0804254954321701,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6465066302953757,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0317284975094494,"pitch":null,"output_rms_dbfs":-20.2594529136671,"output_peak":0.9103883664787014,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.706925474292177},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.310177525541947,"spectral_convergence":0.5411751546439693,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.165129682991532,"spectral_convergence":0.257675970224484,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.645469214343054,"spectral_convergence":0.3065152999592203,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00011748603476190836,"delta_db":-0.4545603500969543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.0016928792865640203,"delta_db":0.7325329270315126,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.005716047240094651,"delta_db":1.2278892654305398,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0016118584764257405,"delta_db":0.9171901245368818,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.00027197425863216735,"delta_db":0.46644161924419547,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":9.837253899810489e-06,"delta_db":-0.0804254954321701,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6465066302953757,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0317284975094494,"pitch":null,"output_rms_dbfs":-20.2594529136671,"output_peak":0.9103883664787014,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.706925474292177},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9335193324845861,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618683437204885,"rms_dbfs":-21.291181411176545,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.435250156246948,"mean":-4.1497230529785154e-05,"zero_energy":false},"output":{"rms":0.09705710973637274,"rms_dbfs":-20.2594529136671,"peak":0.9103883664787014,"peak_dbfs":-0.8154660120904982,"crest_factor_db":19.4439869015766,"mean":7.79338775494289e-10,"zero_energy":false},"rms_difference_db":1.0317284975094445},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":1.8720049411058425e-06,"output_mean_power":4.103509512518998e-06,"output_rms_dbfs":-53.868445554831574,"output_zero_energy":false,"output_peak":0.07799410189583919,"output_energy_fraction":7.478021512887965e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":235,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":52,"paired_voiced_fraction":0.953125,"absolute_difference_cents_median":5.015902024705415,"absolute_difference_cents_p95":21.914234543396507,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002631236493882886,"log_rms_mae_db":1.753989966588193,"rms_envelope_nrmse":0.17381500473608058,"envelope_correlation":0.9877700734911328,"rms_gain_error_db":1.0317287876053802,"reference_mean_power":0.007434364903539191,"output_mean_power":0.009427938755832846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.450307530079474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015621941098031631,"output_power":0.0020419025500367496,"reference_active":true,"delta_db":1.163000155254546},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031400948871223737,"output_power":0.00039050168470540465,"reference_active":true,"delta_db":0.9468014010847735},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9119711962951405e-05,"output_power":5.466280021682055e-05,"reference_active":true,"delta_db":0.46436064127984}],"active_band_count":3,"active_band_mae_db":0.8580540658730532}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024033810696394756,"log_rms_mae_db":1.2524545906502427,"rms_envelope_nrmse":0.1416700666322412,"envelope_correlation":0.9960968908529102,"rms_gain_error_db":1.0317299508026352,"reference_mean_power":0.007471745584669175,"output_mean_power":0.009475345850337763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.136699874330792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014162642987993615,"output_power":0.0018741311217675158,"reference_active":true,"delta_db":1.2165566501676903},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014955063876354673,"output_power":0.00019001739683944909,"reference_active":true,"delta_db":1.040050918015},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0546566248780085e-06,"output_power":3.402863892778298e-06,"reference_active":true,"delta_db":0.46882180486840386}],"active_band_count":3,"active_band_mae_db":0.908476457683698}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002303175773961387,"log_rms_mae_db":1.965473773985591,"rms_envelope_nrmse":0.19917189158895343,"envelope_correlation":0.9879351339302572,"rms_gain_error_db":1.2357564053357755,"reference_mean_power":0.004944923373977616,"output_mean_power":0.006572569796546821},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6955719701225893e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012875788279854436,"output_power":0.0017109273377179373,"reference_active":true,"delta_db":1.2345773880158895},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026650258378785487,"output_power":0.0003558790018953013,"reference_active":true,"delta_db":1.256009399259266},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0302791253002337e-05,"output_power":3.376896871217998e-05,"reference_active":true,"delta_db":0.4703516358542381}],"active_band_count":3,"active_band_mae_db":0.9869794743764646}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020459712012168156,"log_rms_mae_db":1.4800508877622491,"rms_envelope_nrmse":0.17043432678200848,"envelope_correlation":0.9953292176550663,"rms_gain_error_db":1.2357565514336473,"reference_mean_power":0.004969790940376444,"output_mean_power":0.0066056228698305785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.46585734788702e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011840763589577269,"output_power":0.001579786088918762,"reference_active":true,"delta_db":1.2521857508981673},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012134699925967192,"output_power":0.00016012773097739865,"reference_active":true,"delta_db":1.2043750839536835},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.482336680977432e-06,"output_power":2.7009721386329843e-06,"reference_active":true,"delta_db":0.36659419520515335}],"active_band_count":3,"active_band_mae_db":0.9410516766856681}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014083569934759886,"log_rms_mae_db":1.6289836908231476,"rms_envelope_nrmse":0.15327219147856655,"envelope_correlation":0.9891586963935385,"rms_gain_error_db":0.7626439217706626,"reference_mean_power":0.0013181968247743341,"output_mean_power":0.001571247694201376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.706116677050863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035629780608750625,"output_power":0.00042077841428634933,"reference_active":true,"delta_db":0.7224030445592343},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927249258823345e-05,"output_power":8.081162285560356e-05,"reference_active":true,"delta_db":0.6691301347559273},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3444445693793866e-05,"output_power":1.4921236928888433e-05,"reference_active":true,"delta_db":0.4526192519443675}],"active_band_count":3,"active_band_mae_db":0.6147174770865097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010879087849669347,"log_rms_mae_db":1.163858015284448,"rms_envelope_nrmse":0.10829657827441133,"envelope_correlation":0.9974144595035648,"rms_gain_error_db":0.7626488234885971,"reference_mean_power":0.0013248216670349025,"output_mean_power":0.0015791460727904699},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.107952744901023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000335569442501778,"output_power":0.00039602376397411386,"reference_active":true,"delta_db":0.7193884077429115},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4091376965670345e-05,"output_power":4.037741560447549e-05,"reference_active":true,"delta_db":0.7349397513118738},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.097157965658072e-07,"output_power":8.49112585486042e-07,"reference_active":true,"delta_db":0.2063266609679333}],"active_band_count":3,"active_band_mae_db":0.5535516066742395}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":1.513154438253303,"rms_envelope_nrmse":0.2367007372375385,"envelope_correlation":0.955410309910671,"rms_gain_error_db":0.48228673096074437,"reference_mean_power":2.393470871827269e-05,"output_mean_power":2.674587556044526e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3699291941082084e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.018836822609007e-06,"output_power":4.658992690845177e-06,"reference_active":true,"delta_db":0.6419165628501732},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.745214583455975e-06,"output_power":5.866356794295255e-06,"reference_active":true,"delta_db":0.921126169734751},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3999346848964708e-06,"output_power":1.5216771022052303e-06,"reference_active":true,"delta_db":0.3621473173247872}],"active_band_count":3,"active_band_mae_db":0.6417300166365704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":0.9446366606277894,"rms_envelope_nrmse":0.13532037625653603,"envelope_correlation":0.9847978408441469,"rms_gain_error_db":0.48232582924783834,"reference_mean_power":2.405439007374668e-05,"output_mean_power":2.6879855658192068e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.116507982063627e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.688364032412793e-06,"output_power":4.518756559995607e-06,"reference_active":true,"delta_db":0.881851668267353},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3132434905641704e-06,"output_power":3.962093609893194e-06,"reference_active":true,"delta_db":0.776713787306653},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.200263616160695e-07,"output_power":1.2195603935172892e-07,"reference_active":true,"delta_db":0.0692667032103302}],"active_band_count":3,"active_band_mae_db":0.575944052928112}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9256105422973633,"predicted_mos":3.9256105422973633,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"916120725e3fb1a47cbfd3ad59f1092cc1cb840b959fc4bebaa9d805adc8c6b4","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2130057779995695},"started_at":"2026-10-13T04:48:58.478152+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.3739000173227396e-05,"native_input_samples":96000,"native_input_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","output_sha256":"632130e421d1f8c726072049d51d7ca125bae05c63c387950b1ddec9bd7178c1","output_samples":96000,"length_delta_samples":0,"analysis_s":1.945229781999842,"synthesis_s":5.9311181189996205,"total_s":7.876347900999463,"rtf":1.312724650166577,"output_peak":0.6007810831069946,"output_rms":0.07939448312235405,"finished_at":"2026-10-13T04:49:06.362575+00:00","evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"632130e421d1f8c726072049d51d7ca125bae05c63c387950b1ddec9bd7178c1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.064533716153112,"spectral_convergence":0.19720364500166895,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.355339038457778,"spectral_convergence":0.21219404259548563,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.497444823385067,"spectral_convergence":0.23017440642717377,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00012507470357809447,"delta_db":-0.18272799011968852,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.0012812480287851826,"delta_db":-0.47739488221385634,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.003536855639447678,"delta_db":-0.856895488220344,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0011288223123840679,"delta_db":-0.6298230985962611,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.00022327570335902003,"delta_db":-0.3904217352508417,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":8.207562711717467e-06,"delta_db":-0.8670222049204641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5673808998869093,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7130120749819358,"pitch":null,"output_rms_dbfs":-22.004193486158485,"output_peak":0.6007810831069946,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.305772525998652},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.064533716153112,"spectral_convergence":0.19720364500166895,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.355339038457778,"spectral_convergence":0.21219404259548563,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.497444823385067,"spectral_convergence":0.23017440642717377,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00012507470357809447,"delta_db":-0.18272799011968852,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.0012812480287851826,"delta_db":-0.47739488221385634,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.003536855639447678,"delta_db":-0.856895488220344,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0011288223123840679,"delta_db":-0.6298230985962611,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.00022327570335902003,"delta_db":-0.3904217352508417,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":8.207562711717467e-06,"delta_db":-0.8670222049204641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5673808998869093,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7130120749819358,"pitch":null,"output_rms_dbfs":-22.004193486158485,"output_peak":0.6007810831069946,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.305772525998652},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9361334019195366,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618683437204885,"rms_dbfs":-21.291181411176545,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.435250156246948,"mean":-4.1497230529785154e-05,"zero_energy":false},"output":{"rms":0.07939448312235403,"rms_dbfs":-22.004193486158485,"peak":0.6007810831069946,"peak_dbfs":-4.425675009811018,"crest_factor_db":17.578518476347465,"mean":8.377459004774816e-05,"zero_energy":false},"rms_difference_db":-0.7130120749819397},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":1.8720049411058425e-06,"output_mean_power":2.219511849998319e-06,"output_rms_dbfs":-56.53742531929382,"output_zero_energy":false,"output_peak":0.009918044321238995,"output_energy_fraction":6.0445335297524495e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":204,"both_estimated_voiced_frames":187,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9739583333333334,"absolute_difference_cents_median":3.047891267907655,"absolute_difference_cents_p95":11.47355482275424,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002631236493882886,"log_rms_mae_db":1.099449272122492,"rms_envelope_nrmse":0.11015626228259497,"envelope_correlation":0.9939411314555158,"rms_gain_error_db":-0.7130120533105686,"reference_mean_power":0.007434364903539191,"output_mean_power":0.006308740570004073},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.450307530079474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015621941098031631,"output_power":0.001243996088694047,"reference_active":true,"delta_db":-0.9891598110950337},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031400948871223737,"output_power":0.0002755417634946681,"reference_active":true,"delta_db":-0.5675533813789229},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9119711962951405e-05,"output_power":4.8424534599753974e-05,"reference_active":true,"delta_db":-0.06190355967774572}],"active_band_count":3,"active_band_mae_db":0.5395389173839008}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024033810696394756,"log_rms_mae_db":0.8480593077964805,"rms_envelope_nrmse":0.10268271312290508,"envelope_correlation":0.9952094471615901,"rms_gain_error_db":-0.7130149683773267,"reference_mean_power":0.007471745584669175,"output_mean_power":0.006340457251011562},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.136699874330792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014162642987993615,"output_power":0.001126825788742065,"reference_active":true,"delta_db":-0.9928752981968147},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014955063876354673,"output_power":0.00012874388607268136,"reference_active":true,"delta_db":-0.6506165837598163},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0546566248780085e-06,"output_power":2.9240086277225317e-06,"reference_active":true,"delta_db":-0.18983748374890502}],"active_band_count":3,"active_band_mae_db":0.6111097885685121}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002303175773961387,"log_rms_mae_db":1.4154393132487884,"rms_envelope_nrmse":0.12141485962280431,"envelope_correlation":0.9927588587394646,"rms_gain_error_db":-0.7527135434438306,"reference_mean_power":0.004944923373977616,"output_mean_power":0.004158035681694323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6955719701225893e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012875788279854436,"output_power":0.0010219461940372896,"reference_active":true,"delta_db":-1.0034579626254556},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026650258378785487,"output_power":0.00022006352702784768,"reference_active":true,"delta_db":-0.8315335468753471},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0302791253002337e-05,"output_power":2.928468604412939e-05,"reference_active":true,"delta_db":-0.14842061715402283}],"active_band_count":3,"active_band_mae_db":0.6611373755516085}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020459712012168156,"log_rms_mae_db":1.0896249786015868,"rms_envelope_nrmse":0.11278483711167182,"envelope_correlation":0.99407469639769,"rms_gain_error_db":-0.7527133342659774,"reference_mean_power":0.004969790940376444,"output_mean_power":0.004178946263293205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.46585734788702e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011840763589577269,"output_power":0.0009345988601369568,"reference_active":true,"delta_db":-1.0275446315026031},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012134699925967192,"output_power":9.816913421287211e-05,"reference_active":true,"delta_db":-0.9205408048998152},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.482336680977432e-06,"output_power":2.227234712582724e-06,"reference_active":true,"delta_db":-0.470946981244859}],"active_band_count":3,"active_band_mae_db":0.8063441392157591}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014083569934759886,"log_rms_mae_db":1.58533439793885,"rms_envelope_nrmse":0.12832246484491105,"envelope_correlation":0.990332236932521,"rms_gain_error_db":-0.6385460032229345,"reference_mean_power":0.0013181968247743341,"output_mean_power":0.0011379565000412303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.706116677050863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035629780608750625,"output_power":0.000305156969706483,"reference_active":true,"delta_db":-0.6728985460046348},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927249258823345e-05,"output_power":7.373269472258276e-05,"reference_active":true,"delta_db":0.2709929157348117},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3444445693793866e-05,"output_power":1.3222599047863133e-05,"reference_active":true,"delta_db":-0.07226072317817886}],"active_band_count":3,"active_band_mae_db":0.3387173949725418}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010879087849669347,"log_rms_mae_db":1.2242696951597576,"rms_envelope_nrmse":0.1099433400118005,"envelope_correlation":0.993551548792383,"rms_gain_error_db":-0.6385537720807835,"reference_mean_power":0.0013248216670349025,"output_mean_power":0.0011436734653139306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.107952744901023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000335569442501778,"output_power":0.0002912839076554411,"reference_active":true,"delta_db":-0.6146591424916549},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4091376965670345e-05,"output_power":3.628262563826615e-05,"reference_active":true,"delta_db":0.2705416486903225},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.097157965658072e-07,"output_power":8.34531574285103e-07,"reference_active":true,"delta_db":0.1311016083139573}],"active_band_count":3,"active_band_mae_db":0.33876746649864486}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":1.8049622446400868,"rms_envelope_nrmse":0.22830027934786387,"envelope_correlation":0.9533247591170495,"rms_gain_error_db":-0.057092194885200766,"reference_mean_power":2.393470871827269e-05,"output_mean_power":2.3622123028317567e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3699291941082084e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.018836822609007e-06,"output_power":4.847697478115244e-06,"reference_active":true,"delta_db":0.8143513690742403},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.745214583455975e-06,"output_power":4.548749716359873e-06,"reference_active":true,"delta_db":-0.18363814934375097},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3999346848964708e-06,"output_power":1.1543656209760667e-06,"reference_active":true,"delta_db":-0.8376438955483856}],"active_band_count":3,"active_band_mae_db":0.6118778046554589}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":1.4952148858997139,"rms_envelope_nrmse":0.18571892532023784,"envelope_correlation":0.9624044698026804,"rms_gain_error_db":-0.05709391208264656,"reference_mean_power":2.405439007374668e-05,"output_mean_power":2.374023196644559e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.116507982063627e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.688364032412793e-06,"output_power":4.616102272200335e-06,"reference_active":true,"delta_db":0.974416439905941},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3132434905641704e-06,"output_power":3.016996327862786e-06,"reference_active":true,"delta_db":-0.40678571821827575},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.200263616160695e-07,"output_power":9.962734236582303e-08,"reference_active":true,"delta_db":-0.808980960986047}],"active_band_count":3,"active_band_mae_db":0.7300610397034212}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.515439510345459,"predicted_mos":3.515439510345459,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"4f122559d53102fc944908eb2af5839d3f8e136c0efe65d9ab4169d7c295167e","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.905829602001177},"started_at":"2026-10-13T07:07:55.899771+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.01584090100004687,"native_input_samples":264600,"native_input_sha256":"a63fee09f04a055a65305731933b01cce5f5b02cb7b33bbabe6340127069f75f","output_sha256":"c837ac87d5322300823c46992d41134a254a2c0b62fc65ce845355dd0b869f8c","output_samples":264704,"length_delta_samples":104,"analysis_s":0.18183059400143975,"synthesis_s":5.204698853998707,"total_s":5.386529448000147,"rtf":0.8977549080000244,"output_peak":0.5947355031967163,"output_rms":0.08152350647964511,"finished_at":"2026-10-13T07:08:01.329862+00:00","evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"c837ac87d5322300823c46992d41134a254a2c0b62fc65ce845355dd0b869f8c","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.76550682889821,"output_peak":0.5942636589080424},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.65121598983533,"spectral_convergence":0.43743469992419826,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.469022756638779,"spectral_convergence":0.3820485259763881,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.223454864086507,"spectral_convergence":0.43533050415250035,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":6.056689115780559e-05,"delta_db":-3.332069990050246,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.001053786862647511,"delta_db":-1.3261991844760739,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.004340171062657875,"delta_db":0.031999642441138315,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0010491477582592452,"delta_db":-0.9477123798741766,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.00015152708238247485,"delta_db":-2.073933785954373,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":7.054366734014266e-06,"delta_db":-1.524583967276325,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.539416491678722,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.47260671078330496,"pitch":null,"output_rms_dbfs":-21.763788121959852,"output_peak":0.5942636589080424,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.447897870186871},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8108928385520163,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618683437204885,"rms_dbfs":-21.291181411176545,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.435250156246948,"mean":-4.1497230529785154e-05,"zero_energy":false},"output":{"rms":0.08162263181151981,"rms_dbfs":-21.763788121959852,"peak":0.5942636589080424,"peak_dbfs":-4.520416548067364,"crest_factor_db":17.24337157389249,"mean":-0.0001304453017138416,"zero_energy":false},"rms_difference_db":-0.47260671078330674},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":1.8720049411058425e-06,"output_mean_power":1.7309480765093562e-06,"output_rms_dbfs":-57.61715959521304,"output_zero_energy":false,"output_peak":0.012800791146270949,"output_energy_fraction":4.4601434500124006e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":179,"both_estimated_voiced_frames":162,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.84375,"absolute_difference_cents_median":5.2675927541691046,"absolute_difference_cents_p95":23.69586131695232,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002631236493882886,"log_rms_mae_db":1.7291141974109063,"rms_envelope_nrmse":0.11021233810929865,"envelope_correlation":0.9906835979841649,"rms_gain_error_db":-0.47260683748364674,"reference_mean_power":0.007434364903539191,"output_mean_power":0.006667809602366354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.450307530079474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015621941098031631,"output_power":0.0014962309244247807,"reference_active":true,"delta_db":-0.18736369388471463},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031400948871223737,"output_power":0.0003014229034654591,"reference_active":true,"delta_db":-0.17766522868584522},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9119711962951405e-05,"output_power":3.720303684350591e-05,"reference_active":true,"delta_db":-1.206774191301382}],"active_band_count":3,"active_band_mae_db":0.5239343712906473}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024033810696394756,"log_rms_mae_db":1.2624244688613417,"rms_envelope_nrmse":0.081036256364864,"envelope_correlation":0.9959654335472545,"rms_gain_error_db":-0.47260537391542906,"reference_mean_power":0.007471745584669175,"output_mean_power":0.00670133823018731},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.136699874330792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014162642987993615,"output_power":0.0013672113636070931,"reference_active":true,"delta_db":-0.15308648214726187},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014955063876354673,"output_power":0.00014564491494956102,"reference_active":true,"delta_db":-0.1149294607891648},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0546566248780085e-06,"output_power":2.453357149471821e-06,"reference_active":true,"delta_db":-0.9520162253031599}],"active_band_count":3,"active_band_mae_db":0.4066773894131955}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002303175773961387,"log_rms_mae_db":2.2614062710008285,"rms_envelope_nrmse":0.138260801152097,"envelope_correlation":0.9835748685989918,"rms_gain_error_db":-0.0969622112237613,"reference_mean_power":0.004944923373977616,"output_mean_power":0.004835744483781342},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6955719701225893e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012875788279854436,"output_power":0.0012872221586032013,"reference_active":true,"delta_db":-0.0012031962576470414},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026650258378785487,"output_power":0.00026200701613401337,"reference_active":true,"delta_db":-0.07388502746179496},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0302791253002337e-05,"output_power":2.785259483796379e-05,"reference_active":true,"delta_db":-0.36616972435629247}],"active_band_count":3,"active_band_mae_db":0.1470859826919115}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020459712012168156,"log_rms_mae_db":1.4685327622156703,"rms_envelope_nrmse":0.10822629154666878,"envelope_correlation":0.9888547678900008,"rms_gain_error_db":-0.09696239744778011,"reference_mean_power":0.004969790940376444,"output_mean_power":0.00486006279115078},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.46585734788702e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011840763589577269,"output_power":0.0011801799804100584,"reference_active":true,"delta_db":-0.014314668054214772},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012134699925967192,"output_power":0.00011822023997417605,"reference_active":true,"delta_db":-0.11337204725938477},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.482336680977432e-06,"output_power":2.1158606966389835e-06,"reference_active":true,"delta_db":-0.6937361344458005}],"active_band_count":3,"active_band_mae_db":0.2738076165864667}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014083569934759886,"log_rms_mae_db":3.0341462280074416,"rms_envelope_nrmse":0.26402254195730074,"envelope_correlation":0.9563232964891928,"rms_gain_error_db":-1.3984773337309686,"reference_mean_power":0.0013181968247743341,"output_mean_power":0.0009552840527630096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.706116677050863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035629780608750625,"output_power":0.00029257933195946036,"reference_active":true,"delta_db":-0.8556950422077343},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927249258823345e-05,"output_power":6.175858396917571e-05,"reference_active":true,"delta_db":-0.49863485134401675},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3444445693793866e-05,"output_power":1.4468587561218445e-05,"reference_active":true,"delta_db":0.3188323566807278}],"active_band_count":3,"active_band_mae_db":0.557720750077493}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010879087849669347,"log_rms_mae_db":2.348335697373011,"rms_envelope_nrmse":0.23577894864708906,"envelope_correlation":0.9663538746291769,"rms_gain_error_db":-1.3984719724642547,"reference_mean_power":0.0013248216670349025,"output_mean_power":0.0009600861948836643},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.107952744901023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000335569442501778,"output_power":0.0002781049462878663,"reference_active":true,"delta_db":-0.8157369323749839},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4091376965670345e-05,"output_power":3.213819421916244e-05,"reference_active":true,"delta_db":-0.2562307195302168},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.097157965658072e-07,"output_power":8.761283203105719e-07,"reference_active":true,"delta_db":0.34235106950346705}],"active_band_count":3,"active_band_mae_db":0.4714395738028893}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":3.7292503923907754,"rms_envelope_nrmse":0.30370106380180695,"envelope_correlation":0.9254830877228127,"rms_gain_error_db":-0.9697079570403313,"reference_mean_power":2.393470871827269e-05,"output_mean_power":1.9145087291808277e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3699291941082084e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.018836822609007e-06,"output_power":3.110817569590865e-06,"reference_active":true,"delta_db":-1.1122582968828794},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.745214583455975e-06,"output_power":2.7737474845211642e-06,"reference_active":true,"delta_db":-2.3318893506425757},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3999346848964708e-06,"output_power":7.329194437762524e-07,"reference_active":true,"delta_db":-2.8105153044796705}],"active_band_count":3,"active_band_mae_db":2.084887650668375}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":3.009867636632571,"rms_envelope_nrmse":0.25765043736313853,"envelope_correlation":0.9413348061428999,"rms_gain_error_db":-0.9697112633364836,"reference_mean_power":2.405439007374668e-05,"output_mean_power":1.9240804328814583e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.116507982063627e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.688364032412793e-06,"output_power":2.811969523840179e-06,"reference_active":true,"delta_db":-1.178231688044165},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3132434905641704e-06,"output_power":1.8483338918695172e-06,"reference_active":true,"delta_db":-2.5347292639338486},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.200263616160695e-07,"output_power":6.585548210192476e-08,"reference_active":true,"delta_db":-2.606847080414361}],"active_band_count":3,"active_band_mae_db":2.106602677464125}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"golf","status":"measured","mos":{"status":"measured","value":2.656071424484253,"predicted_mos":2.656071424484253,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"91828cf7f21717b4a28967ffdd34bc766444db9f6ab8b0f7c66a6c62d7c25741","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.451909583000088},"started_at":"2026-10-13T05:02:14.888118+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0038300050000543706,"native_input_samples":144000,"native_input_sha256":"2875306e8b3926a52bcb5e9289ff0efcf17bd7b241f16d1b78541e6fc9673201","output_sha256":"6d260bc3b57374b3d3fa30591f65a61953b21e2969a89fc5edfec7420a57f6a4","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5809692689999792,"synthesis_s":0.14279187499960244,"total_s":0.7237611439995817,"rtf":0.12062685733326362,"output_peak":0.7294853925704956,"output_rms":0.0758922361599198,"finished_at":"2026-10-13T05:02:15.631482+00:00","analysis_stages":{"external_f0_s":0.08235723899997538},"evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"6d260bc3b57374b3d3fa30591f65a61953b21e2969a89fc5edfec7420a57f6a4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.388782502802812,"output_peak":0.7025992175911436},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.7534845880812515,"spectral_convergence":0.5666400249622194,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.487500256981352,"spectral_convergence":0.25248988017784674,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.659274913096498,"spectral_convergence":0.2791139741534537,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013066711990891274,"output_power":0.00012205134175167214,"delta_db":-0.2962376060584568,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001432506513453256,"output_power":0.0011337926587827608,"delta_db":-1.0156296469243455,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0043155030050477956,"output_power":0.003261972976072971,"delta_db":-1.2155106457014357,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0013071711372435672,"output_power":0.001057634972300389,"delta_db":-0.9199664695020328,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002446852822882272,"output_power":0.0001869949031112762,"delta_db":-1.167780783099771,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0037708894854436e-05,"output_power":6.834910418189149e-06,"delta_db":-1.66901769630477,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0473571412651355,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1048450187184113,"pitch":null,"output_rms_dbfs":-22.388782502802812,"output_peak":0.7025992175911436,"reference_rms_dbfs":-21.2839374840844,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.966753252719701},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9031800770615095,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08625874313272025,"rms_dbfs":-21.2839374840844,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.4280062291548,"mean":-4.116083823380765e-05,"zero_energy":false},"output":{"rms":0.07595578820891308,"rms_dbfs":-22.388782502802812,"peak":0.7025992175911436,"peak_dbfs":-3.06584676341155,"crest_factor_db":19.322935739391262,"mean":0.0003431124686582993,"zero_energy":false},"rms_difference_db":-1.104845018718411},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":102,"quiet_samples":16320,"reference_mean_power":1.886807834072148e-06,"output_mean_power":1.7165302936634393e-06,"output_rms_dbfs":-57.65348527656364,"output_zero_energy":false,"output_peak":0.010610144069476853,"output_energy_fraction":5.066441961852662e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":210,"both_estimated_voiced_frames":178,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9270833333333334,"absolute_difference_cents_median":3.6978506067063046,"absolute_difference_cents_p95":12.605103266729428,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002631236493882886,"log_rms_mae_db":1.9519201027730257,"rms_envelope_nrmse":0.16633933110880395,"envelope_correlation":0.9853365354949211,"rms_gain_error_db":-1.1048452023060222,"reference_mean_power":0.007446785869889712,"output_mean_power":0.005774100594980898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.452356766164656e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001560571077839441,"output_power":0.0012320493086236915,"reference_active":true,"delta_db":-1.026554644974951},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031093357664944906,"output_power":0.00026935515383069936,"reference_active":true,"delta_db":-0.6234233281098089},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.913904842359545e-05,"output_power":3.446816680032934e-05,"reference_active":true,"delta_db":-1.5400855618267586}],"active_band_count":3,"active_band_mae_db":1.0633545116371728}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00024033810696394756,"log_rms_mae_db":1.5546325196207267,"rms_envelope_nrmse":0.13376460778887161,"envelope_correlation":0.9956260522875621,"rms_gain_error_db":-1.1048456410640053,"reference_mean_power":0.007484291890408534,"output_mean_power":0.00580318148615421},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.1395490160178902e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014149017909731003,"output_power":0.0011296134110174415,"reference_active":true,"delta_db":-0.9779645653615796},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014655489338018406,"output_power":0.0001294571036997545,"reference_active":true,"delta_db":-0.5387443738318689},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.047957997550667e-06,"output_power":2.5981572450331975e-06,"reference_active":true,"delta_db":-0.6934354610840142}],"active_band_count":3,"active_band_mae_db":0.7367148000924876}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002303175773961387,"log_rms_mae_db":1.9927723059509752,"rms_envelope_nrmse":0.17465336361646294,"envelope_correlation":0.9859276677252609,"rms_gain_error_db":-1.1790010150194705,"reference_mean_power":0.004953185511712761,"output_mean_power":0.003775587089996799},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.6963652259754234e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012863678273339343,"output_power":0.0009873038845490108,"reference_active":true,"delta_db":-1.1491432388051868},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002639251854033185,"output_power":0.00022331563229620313,"reference_active":true,"delta_db":-0.7256171017008728},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0232309074402076e-05,"output_power":2.378337813524362e-05,"reference_active":true,"delta_db":-1.0419777802721617}],"active_band_count":3,"active_band_mae_db":0.9722460402594071}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00020459712012168156,"log_rms_mae_db":1.5732479267996835,"rms_envelope_nrmse":0.14581399603567932,"envelope_correlation":0.9943920479940224,"rms_gain_error_db":-1.179000102457881,"reference_mean_power":0.004978136446447089,"output_mean_power":0.0037946068453090323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.4673250129767045e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001183255919839933,"output_power":0.000913072898145924,"reference_active":true,"delta_db":-1.1257323369299834},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011871408695983572,"output_power":0.00010209446825238537,"reference_active":true,"delta_db":-0.6550004515879235},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4743159794369816e-06,"output_power":2.275717101788901e-06,"reference_active":true,"delta_db":-0.3633688666727658}],"active_band_count":3,"active_band_mae_db":0.7147005517302243}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014083569934759886,"log_rms_mae_db":1.9703545257365553,"rms_envelope_nrmse":0.1852954755295499,"envelope_correlation":0.981291128912362,"rms_gain_error_db":-1.061936177532878,"reference_mean_power":0.0013203991981887142,"output_mean_power":0.00103397879954858},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.710954831363522e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003562060074631555,"output_power":0.00026297608369976364,"reference_active":true,"delta_db":-1.317849861523812},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.918342856672123e-05,"output_power":6.182994173444068e-05,"reference_active":true,"delta_db":-0.48803243575833577},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.34457570709314e-05,"output_power":8.5175287387823e-06,"reference_active":true,"delta_db":-1.9827165285877317}],"active_band_count":3,"active_band_mae_db":1.2628662752899598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010879087849669347,"log_rms_mae_db":1.5615333485022254,"rms_envelope_nrmse":0.1373025652597433,"envelope_correlation":0.9945682241940148,"rms_gain_error_db":-1.0619335577570181,"reference_mean_power":0.0013270462430956217,"output_mean_power":0.0010391845973344012},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.113056019615109e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003355688349679864,"output_power":0.00024994977584978623,"reference_active":true,"delta_db":-1.2792886850299428},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.387024808678908e-05,"output_power":3.0271460344593468e-05,"reference_active":true,"delta_db":-0.4878500516380483},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.096931080079527e-07,"output_power":6.429034319756133e-07,"reference_active":true,"delta_db":-1.001746985015899}],"active_band_count":3,"active_band_mae_db":0.9229619072279635}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":2.4129590276615684,"rms_envelope_nrmse":0.2129733684675126,"envelope_correlation":0.9619728766792114,"rms_gain_error_db":-0.5304177039106359,"reference_mean_power":2.397461153299699e-05,"output_mean_power":2.1218262039450023e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3708587578007682e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.016462056546588e-06,"output_power":3.606897432973075e-06,"reference_active":true,"delta_db":-0.4670987580515744},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.755527075540814e-06,"output_power":4.158397803199795e-06,"reference_active":true,"delta_db":-0.5827262594069395},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3941732690430353e-06,"output_power":8.686638467454311e-07,"reference_active":true,"delta_db":-2.054650047847889}],"active_band_count":3,"active_band_mae_db":1.0348250217688009}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":1.8644915552685837,"rms_envelope_nrmse":0.13782726275679255,"envelope_correlation":0.9839281627472902,"rms_gain_error_db":-0.5303631393223932,"reference_mean_power":2.409465592734126e-05,"output_mean_power":2.1324772909758146e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1172574287196194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6853946669620687e-06,"output_power":3.405716395759844e-06,"reference_active":true,"delta_db":-0.342755229259665},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3158920571215833e-06,"output_power":2.6553685441501593e-06,"reference_active":true,"delta_db":-0.9647557819600462},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1993424923294007e-07,"output_power":6.870407513569011e-08,"reference_active":true,"delta_db":-2.4196072316433366}],"active_band_count":3,"active_band_mae_db":1.242372747621016}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.113330602645874,"predicted_mos":3.113330602645874,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"341493d80669728649335c7e5dbefdfff29c803a29d18648b128044212e78c65","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.215742428999874},"started_at":"2026-10-13T05:07:44.684098+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030072500003370806,"native_input_samples":144000,"native_input_sha256":"2875306e8b3926a52bcb5e9289ff0efcf17bd7b241f16d1b78541e6fc9673201","output_sha256":"46f5c1427b806a00c795d1026209c1023f98521e74df1b7874fdb95b19f59843","output_samples":144120,"length_delta_samples":120,"analysis_s":2.252226688000519,"synthesis_s":3.4058919489998516,"total_s":5.658118637000371,"rtf":0.9430197728333951,"output_peak":0.745739221572876,"output_rms":0.08271608075863734,"finished_at":"2026-10-13T05:07:50.355646+00:00","analysis_stages":{"external_f0_s":1.4681149059997551},"evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"46f5c1427b806a00c795d1026209c1023f98521e74df1b7874fdb95b19f59843","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.64099114878379,"output_peak":0.7466777817020169},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.293733756895378,"spectral_convergence":0.5414144426028447,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.428573458521253,"spectral_convergence":0.30243954162636105,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.287485165096534,"spectral_convergence":0.27156407686444906,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00012044725496723559,"delta_db":-0.34645374576255106,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.0013770094980614651,"delta_db":-0.16435762597567263,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.003974126392852637,"delta_db":-0.3506520753756224,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0011617731887126935,"delta_db":-0.5048654533657899,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.000217902305095185,"delta_db":-0.4962181518623912,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":7.770428806691505e-06,"delta_db":-1.1047144447411834,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4945435828472018,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.34619215484330756,"pitch":null,"output_rms_dbfs":-21.637373566019853,"output_peak":0.7466777817020169,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.669930793504388},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8966435020520639,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618683437204885,"rms_dbfs":-21.291181411176545,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.435250156246948,"mean":-4.1497230529785154e-05,"zero_energy":false},"output":{"rms":0.0828192554210427,"rms_dbfs":-21.637373566019853,"peak":0.7466777817020169,"peak_dbfs":-2.5373354287120513,"crest_factor_db":19.100038137307802,"mean":3.1556116794721986e-05,"zero_energy":false},"rms_difference_db":-0.34619215484330823},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":1.8720049411058425e-06,"output_mean_power":4.863364227104759e-06,"output_rms_dbfs":-53.13063204028711,"output_zero_energy":false,"output_peak":0.07419711123133105,"output_energy_fraction":0.0001217194908077654,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":205,"both_estimated_voiced_frames":174,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.90625,"absolute_difference_cents_median":4.959777608960064,"absolute_difference_cents_p95":22.306769905855024,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002631236493882886,"log_rms_mae_db":2.5266853245236867,"rms_envelope_nrmse":0.1708415905064888,"envelope_correlation":0.9721131993676307,"rms_gain_error_db":-0.3461922462142475,"reference_mean_power":0.007434364903539191,"output_mean_power":0.00686474879135886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.450307530079474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015621941098031631,"output_power":0.0014701649927579999,"reference_active":true,"delta_db":-0.26368918767931787},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031400948871223737,"output_power":0.00028825401716524665,"reference_active":true,"delta_db":-0.3716740332242837},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9119711962951405e-05,"output_power":4.095261278306833e-05,"reference_active":true,"delta_db":-0.7897419641017668}],"active_band_count":3,"active_band_mae_db":0.47503506166845605}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024033810696394756,"log_rms_mae_db":1.7724155738109195,"rms_envelope_nrmse":0.1045666637143247,"envelope_correlation":0.9894124400046412,"rms_gain_error_db":-0.34619183296456274,"reference_mean_power":0.007471745584669175,"output_mean_power":0.006899266045967216},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.136699874330792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014162642987993615,"output_power":0.00133741486992623,"reference_active":true,"delta_db":-0.24878160035844157},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014955063876354673,"output_power":0.00013554236620735825,"reference_active":true,"delta_db":-0.4271320934738414},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0546566248780085e-06,"output_power":2.8904493668880217e-06,"reference_active":true,"delta_db":-0.2399703203860792}],"active_band_count":3,"active_band_mae_db":0.3052946714061207}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002303175773961387,"log_rms_mae_db":2.8100008023473064,"rms_envelope_nrmse":0.1802229954002272,"envelope_correlation":0.9723050116918626,"rms_gain_error_db":-0.31744017142922715,"reference_mean_power":0.004944923373977616,"output_mean_power":0.004596376007918835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6955719701225893e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012875788279854436,"output_power":0.0012061139956038808,"reference_active":true,"delta_db":-0.28385469854853906},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026650258378785487,"output_power":0.00024630982130893113,"reference_active":true,"delta_db":-0.3421969478549517},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0302791253002337e-05,"output_power":2.76688905420976e-05,"reference_active":true,"delta_db":-0.3949088881558028}],"active_band_count":3,"active_band_mae_db":0.3403201781864312}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020459712012168156,"log_rms_mae_db":2.0088268307159667,"rms_envelope_nrmse":0.11430738187274823,"envelope_correlation":0.988341848749686,"rms_gain_error_db":-0.31743957965510555,"reference_mean_power":0.004969790940376444,"output_mean_power":0.004619491391018009},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.46585734788702e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011840763589577269,"output_power":0.0011090354228352986,"reference_active":true,"delta_db":-0.284342923291655},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012134699925967192,"output_power":0.00011072686804178096,"reference_active":true,"delta_db":-0.39776025431675327},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.482336680977432e-06,"output_power":2.4571197826001797e-06,"reference_active":true,"delta_db":-0.04434356306947159}],"active_band_count":3,"active_band_mae_db":0.24214891355929333}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014083569934759886,"log_rms_mae_db":2.6991646284425364,"rms_envelope_nrmse":0.21050238133705282,"envelope_correlation":0.9655406928023195,"rms_gain_error_db":-0.46591748779713993,"reference_mean_power":0.0013181968247743341,"output_mean_power":0.0011841003579812562},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.706116677050863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035629780608750625,"output_power":0.0003130497971265502,"reference_active":true,"delta_db":-0.5619972150574102},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927249258823345e-05,"output_power":6.127772753290904e-05,"reference_active":true,"delta_db":-0.5325816362543399},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3444445693793866e-05,"output_power":1.4751135237628792e-05,"reference_active":true,"delta_db":0.40282543382221864}],"active_band_count":3,"active_band_mae_db":0.49913476171132287}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010879087849669347,"log_rms_mae_db":1.8191030118961315,"rms_envelope_nrmse":0.12043886791313163,"envelope_correlation":0.9893236767789012,"rms_gain_error_db":-0.46591729695864176,"reference_mean_power":0.0013248216670349025,"output_mean_power":0.0011900513259080084},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.107952744901023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000335569442501778,"output_power":0.00029599585805728117,"reference_active":true,"delta_db":-0.5449677248445988},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4091376965670345e-05,"output_power":2.9923465474004577e-05,"reference_active":true,"delta_db":-0.5663265463496328},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.097157965658072e-07,"output_power":8.127166937916779e-07,"reference_active":true,"delta_db":0.016065686553041317}],"active_band_count":3,"active_band_mae_db":0.3757866525824243}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":2.2849805066873827,"rms_envelope_nrmse":0.3239854016595208,"envelope_correlation":0.9032922004577335,"rms_gain_error_db":-0.4101796748957673,"reference_mean_power":2.393470871827269e-05,"output_mean_power":2.1777608144608275e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3699291941082084e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.018836822609007e-06,"output_power":3.7253980742639274e-06,"reference_active":true,"delta_db":-0.32927687096911495},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.745214583455975e-06,"output_power":4.2114022578621144e-06,"reference_active":true,"delta_db":-0.518291308134329},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3999346848964708e-06,"output_power":1.0165853766303548e-06,"reference_active":true,"delta_db":-1.3896391560920556}],"active_band_count":3,"active_band_mae_db":0.7457357783984998}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":1.4059831413483583,"rms_envelope_nrmse":0.16375259493492636,"envelope_correlation":0.9719312892565195,"rms_gain_error_db":-0.4101875547838752,"reference_mean_power":2.405439007374668e-05,"output_mean_power":2.1886463582090305e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.116507982063627e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.688364032412793e-06,"output_power":3.425288906705159e-06,"reference_active":true,"delta_db":-0.32136570257111097},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3132434905641704e-06,"output_power":2.8112409355248727e-06,"reference_active":true,"delta_db":-0.71355285274845},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.200263616160695e-07,"output_power":8.166711392221151e-08,"reference_active":true,"delta_db":-1.672294333448452}],"active_band_count":3,"active_band_mae_db":0.9024042962560044}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8495943546295166,"predicted_mos":2.8495943546295166,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"f46acaa073db21d946092b4028d93bc82e1169947c56199535f278900dac5122","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.752361734999795},"started_at":"2026-10-13T05:10:22.086117+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035158849987055873,"native_input_samples":144000,"native_input_sha256":"2875306e8b3926a52bcb5e9289ff0efcf17bd7b241f16d1b78541e6fc9673201","output_sha256":"8db636cdb3c2eeb84f39d8b58c33a596def7fae42ae7f4ac16584fe1607bcbb4","output_samples":144240,"length_delta_samples":240,"analysis_s":0.8383712409995496,"synthesis_s":0.1956019400004152,"total_s":1.0339731809999648,"rtf":0.17232886349999413,"output_peak":0.6922863125801086,"output_rms":0.0843916047547565,"finished_at":"2026-10-13T05:10:23.133230+00:00","analysis_stages":{"external_f0_s":0.15457044599861547},"evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"8db636cdb3c2eeb84f39d8b58c33a596def7fae42ae7f4ac16584fe1607bcbb4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.466697902816925,"output_peak":0.687905647023278},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.637166394293451,"spectral_convergence":0.14986439615431435,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.364737671810731,"spectral_convergence":0.13859281829217923,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.770062950399211,"spectral_convergence":0.14667606690335663,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.00012926324109237129,"delta_db":-0.03967240267438826,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.0013701355543458483,"delta_db":-0.18609162252922173,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.004194109224088436,"delta_db":-0.11667146516963531,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0012312337724418427,"delta_db":-0.2526737540869233,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.00021326918125771048,"delta_db":-0.5895553798145482,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":7.831365341785194e-06,"delta_db":-1.070789454760471,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3759090131725313,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16828428914227994,"pitch":null,"output_rms_dbfs":-21.45946570031883,"output_peak":0.687905647023278,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.590655672167798},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9450381705633725,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618683437204885,"rms_dbfs":-21.291181411176545,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.435250156246948,"mean":-4.1497230529785154e-05,"zero_energy":false},"output":{"rms":0.0845330842840127,"rms_dbfs":-21.45946570031883,"peak":0.687905647023278,"peak_dbfs":-3.249422508173765,"crest_factor_db":18.21004319214506,"mean":-0.0005173968414315016,"zero_energy":false},"rms_difference_db":-0.1682842891422851},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":1.8720049411058425e-06,"output_mean_power":3.0050233606168586e-06,"output_rms_dbfs":-55.22152147505647,"output_zero_energy":false,"output_peak":0.028637913409679566,"output_energy_fraction":7.219055768811406e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":178,"both_estimated_voiced_frames":171,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.890625,"absolute_difference_cents_median":0.6394942921122665,"absolute_difference_cents_p95":3.874979611545834,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002631236493882886,"log_rms_mae_db":0.8830382133864119,"rms_envelope_nrmse":0.05779575712958532,"envelope_correlation":0.9970619946800848,"rms_gain_error_db":-0.16828419398809635,"reference_mean_power":0.007434364903539191,"output_mean_power":0.007151801541281473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.450307530079474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015621941098031631,"output_power":0.001540931385169261,"reference_active":true,"delta_db":-0.05951695163789887},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031400948871223737,"output_power":0.0002966998859167746,"reference_active":true,"delta_db":-0.24625392417970696},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9119711962951405e-05,"output_power":4.317118211040087e-05,"reference_active":true,"delta_db":-0.5606187080473426}],"active_band_count":3,"active_band_mae_db":0.28879652795498284}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024033810696394756,"log_rms_mae_db":0.5174812163122113,"rms_envelope_nrmse":0.03821804475493716,"envelope_correlation":0.9988134650258872,"rms_gain_error_db":-0.16828342390753737,"reference_mean_power":0.007471745584669175,"output_mean_power":0.007187762741950382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.136699874330792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014162642987993615,"output_power":0.0014010809180195368,"reference_active":true,"delta_db":-0.04681089339282378},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014955063876354673,"output_power":0.00014423065231029198,"reference_active":true,"delta_db":-0.15730704641664375},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0546566248780085e-06,"output_power":2.611617852012801e-06,"reference_active":true,"delta_db":-0.6805276950087289}],"active_band_count":3,"active_band_mae_db":0.2948818782727321}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002303175773961387,"log_rms_mae_db":1.0387145871096262,"rms_envelope_nrmse":0.06053606943783517,"envelope_correlation":0.9970191893204875,"rms_gain_error_db":-0.13359521647329164,"reference_mean_power":0.004944923373977616,"output_mean_power":0.0047951262327593745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6955719701225893e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012875788279854436,"output_power":0.0012796509362589205,"reference_active":true,"delta_db":-0.026823080837902774},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026650258378785487,"output_power":0.00024330439532812295,"reference_active":true,"delta_db":-0.3955146934924839},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0302791253002337e-05,"output_power":2.5097498039994103e-05,"reference_active":true,"delta_db":-0.8185220512336001}],"active_band_count":3,"active_band_mae_db":0.4136199418546623}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020459712012168156,"log_rms_mae_db":0.5763751668308673,"rms_envelope_nrmse":0.039354451284589906,"envelope_correlation":0.9987295902255492,"rms_gain_error_db":-0.13359469303609708,"reference_mean_power":0.004969790940376444,"output_mean_power":0.004819241063908517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.46585734788702e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011840763589577269,"output_power":0.0011767693083190708,"reference_active":true,"delta_db":-0.026883772614005687},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012134699925967192,"output_power":0.00011283198867846201,"reference_active":true,"delta_db":-0.31596798603875315},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.482336680977432e-06,"output_power":2.0123970419836845e-06,"reference_active":true,"delta_db":-0.9114701452145415}],"active_band_count":3,"active_band_mae_db":0.4181073012891001}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014083569934759886,"log_rms_mae_db":1.1572406205526862,"rms_envelope_nrmse":0.08810255720282369,"envelope_correlation":0.9947763837733964,"rms_gain_error_db":-0.32552466423908566,"reference_mean_power":0.0013181968247743341,"output_mean_power":0.0012230037671417738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.706116677050863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035629780608750625,"output_power":0.0003267233092487003,"reference_active":true,"delta_db":-0.3763302895746722},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927249258823345e-05,"output_power":6.141072378751822e-05,"reference_active":true,"delta_db":-0.5231659884907265},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3444445693793866e-05,"output_power":1.1655251225410959e-05,"reference_active":true,"delta_db":-0.6202126219118909}],"active_band_count":3,"active_band_mae_db":0.5065696333257632}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010879087849669347,"log_rms_mae_db":0.7101595136386389,"rms_envelope_nrmse":0.05967016529277124,"envelope_correlation":0.9980315238589864,"rms_gain_error_db":-0.32552319607644464,"reference_mean_power":0.0013248216670349025,"output_mean_power":0.0012291506145851766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.107952744901023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000335569442501778,"output_power":0.0003080785261726642,"reference_active":true,"delta_db":-0.37120978413091976},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4091376965670345e-05,"output_power":3.0085435913287784e-05,"reference_active":true,"delta_db":-0.5428823448372581},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.097157965658072e-07,"output_power":6.182792115012254e-07,"reference_active":true,"delta_db":-1.171479675205403}],"active_band_count":3,"active_band_mae_db":0.6951906013911936}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":1.1245635619730392,"rms_envelope_nrmse":0.15485919582192836,"envelope_correlation":0.9804039383954737,"rms_gain_error_db":-0.5110777008164056,"reference_mean_power":2.393470871827269e-05,"output_mean_power":2.127748908841252e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3699291941082084e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.018836822609007e-06,"output_power":3.29358104395364e-06,"reference_active":true,"delta_db":-0.8643201839496549},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.745214583455975e-06,"output_power":3.7271895584860543e-06,"reference_active":true,"delta_db":-1.0487437565789668},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3999346848964708e-06,"output_power":9.380220418719304e-07,"reference_active":true,"delta_db":-1.7389473012765346}],"active_band_count":3,"active_band_mae_db":1.2173370806017187}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":0.75681649042685,"rms_envelope_nrmse":0.09616905207584851,"envelope_correlation":0.9931770015372204,"rms_gain_error_db":-0.5110691964190635,"reference_mean_power":2.405439007374668e-05,"output_mean_power":2.1383925352638185e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.116507982063627e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.688364032412793e-06,"output_power":3.029157586914807e-06,"reference_active":true,"delta_db":-0.8551191091595133},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3132434905641704e-06,"output_power":2.4777904312825845e-06,"reference_active":true,"delta_db":-1.2618878193924086},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.200263616160695e-07,"output_power":6.321140837826502e-08,"reference_active":true,"delta_db":-2.7848117473012968}],"active_band_count":3,"active_band_mae_db":1.6339395586177397}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_237-134500-0031","method":"magphase","status":"measured","mos":{"status":"measured","value":3.318911075592041,"predicted_mos":3.318911075592041,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289020,"input_sample_rate":48000,"input_seconds":6.02125,"input_float32_sha256":"b7d61a44d06179f19fed05c6924541957261c9f71354b6e1e315a80d87fc7ce4","internal_samples":96340,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1055702379999275},"started_at":"2026-10-13T05:13:52.852935+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00939076800023031,"native_input_samples":288000,"native_input_sha256":"dd5e282425744beb64e3c388059697033eeb108210759b7ffdc2c922afbc582b","output_sha256":"bc469de4bf022d5f0764296da244233ec3a4d0b477ba064d807b5bd9b2a876b9","output_samples":289020,"length_delta_samples":1020,"analysis_s":3.4786310979998234,"synthesis_s":1.0700121150002815,"total_s":4.548643213000105,"rtf":0.7581072021666841,"output_peak":0.6665189488788784,"output_rms":0.0840920011374195,"finished_at":"2026-10-13T05:13:57.423142+00:00","evaluation_input":{"reference":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"output_sha256":"bc469de4bf022d5f0764296da244233ec3a4d0b477ba064d807b5bd9b2a876b9","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96340,"length_delta_samples":340,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.497832341568866,"output_peak":0.6591111955448187},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9964708324683413,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.793647690353717,"spectral_convergence":0.5682216042643876,"spectral_convergence_status":"ok","amplitude_floor":2.051312383514402e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.979874094931039,"spectral_convergence":0.41813412808810035,"spectral_convergence_status":"ok","amplitude_floor":1.425885904641783e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.137440903943865,"spectral_convergence":0.33961979954394483,"spectral_convergence_status":"ok","amplitude_floor":8.483800566678623e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001304494586920238,"output_power":0.0001147613494106548,"delta_db":-0.5564663509330545,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014301207186140645,"output_power":0.0013492836136429892,"delta_db":-0.2526945234530941,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.004308309383381418,"output_power":0.004041092175188891,"delta_db":-0.27808126219417123,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.001304992137209625,"output_power":0.0013258552934040338,"delta_db":0.0688823186780288,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002442775973124686,"output_power":0.00026706930347832877,"delta_db":0.38740334099537727,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0021123865382114e-05,"output_power":1.0015208328421134e-05,"delta_db":-0.002564426572134321,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25768203713764337,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19129738576271396,"pitch":null,"output_rms_dbfs":-21.48247879693926,"output_peak":0.6591111955448187,"reference_rms_dbfs":-21.29118141117655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9703208964095404},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7828974536360649,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08618683437204885,"rms_dbfs":-21.291181411176545,"peak":0.641510009765625,"peak_dbfs":-3.8559312549296005,"crest_factor_db":17.435250156246948,"mean":-4.1497230529785154e-05,"zero_energy":false},"output":{"rms":0.08430941195058425,"rms_dbfs":-21.48247879693926,"peak":0.6591111955448187,"peak_dbfs":-3.6208262289764725,"crest_factor_db":17.86165256796279,"mean":0.0010889501044430098,"zero_energy":false},"rms_difference_db":-0.1912973857627165},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.923405486741104e-06,"threshold_dbfs":-51.59680232089683,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":1.8720049411058425e-06,"output_mean_power":1.9424369405901206e-05,"output_rms_dbfs":-47.11653071264898,"output_zero_energy":false,"output_peak":0.11927290380412862,"output_energy_fraction":0.00046911658027058874,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":192,"output_estimated_voiced_frames":201,"both_estimated_voiced_frames":169,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.8802083333333334,"absolute_difference_cents_median":5.779939569610029,"absolute_difference_cents_p95":33.5047193413698,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002631236493882886,"log_rms_mae_db":3.3670595494627613,"rms_envelope_nrmse":0.2515867606679517,"envelope_correlation":0.9380024814253072,"rms_gain_error_db":-0.19129714255870267,"reference_mean_power":0.007434364903539191,"output_mean_power":0.007114004894613246},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.450307530079474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015621941098031631,"output_power":0.001519903698216441,"reference_active":true,"delta_db":-0.11918924258662128},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031400948871223737,"output_power":0.00028927631949927007,"reference_active":true,"delta_db":-0.35629888501012624},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9119711962951405e-05,"output_power":4.731774108603239e-05,"reference_active":true,"delta_db":-0.16231807892706437}],"active_band_count":3,"active_band_mae_db":0.21260206884127064}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00024033810696394756,"log_rms_mae_db":2.414270632357389,"rms_envelope_nrmse":0.1739400417280179,"envelope_correlation":0.9672030462935198,"rms_gain_error_db":-0.19129811027228039,"reference_mean_power":0.007471745584669175,"output_mean_power":0.007149773182552353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.136699874330792e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014162642987993615,"output_power":0.0013832623791961065,"reference_active":true,"delta_db":-0.10239742065930203},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014955063876354673,"output_power":0.0001356288916821794,"reference_active":true,"delta_db":-0.42436059491410266},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0546566248780085e-06,"output_power":2.8058786296750903e-06,"reference_active":true,"delta_db":-0.36893516738775467}],"active_band_count":3,"active_band_mae_db":0.29856439432038645}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002303175773961387,"log_rms_mae_db":4.282063455762507,"rms_envelope_nrmse":0.2716042789330956,"envelope_correlation":0.9356884905241689,"rms_gain_error_db":-0.2707284425043008,"reference_mean_power":0.004944923373977616,"output_mean_power":0.004646080410661176},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6955719701225893e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012875788279854436,"output_power":0.0012053786703208455,"reference_active":true,"delta_db":-0.2865032467049259},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026650258378785487,"output_power":0.0002410192204505814,"reference_active":true,"delta_db":-0.4364974650064687},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0302791253002337e-05,"output_power":2.7330855807375583e-05,"reference_active":true,"delta_db":-0.4482940320133455}],"active_band_count":3,"active_band_mae_db":0.39043158124158}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020459712012168156,"log_rms_mae_db":3.262750823312168,"rms_envelope_nrmse":0.19012287069556263,"envelope_correlation":0.9653619432640935,"rms_gain_error_db":-0.2707284991595481,"reference_mean_power":0.004969790940376444,"output_mean_power":0.004669445062274707},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.46585734788702e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011840763589577269,"output_power":0.0011086572476052285,"reference_active":true,"delta_db":-0.28582409712059065},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012134699925967192,"output_power":0.00010767335969661417,"reference_active":true,"delta_db":-0.5192077689712027},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.482336680977432e-06,"output_power":2.1862929921848295e-06,"reference_active":true,"delta_db":-0.551523220874562}],"active_band_count":3,"active_band_mae_db":0.45218502898878504}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014083569934759886,"log_rms_mae_db":3.9790476238949344,"rms_envelope_nrmse":0.2858753595096235,"envelope_correlation":0.9369938945258472,"rms_gain_error_db":0.11457797292411755,"reference_mean_power":0.0013181968247743341,"output_mean_power":0.00135343704212157},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.706116677050863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035629780608750625,"output_power":0.00035707346307234177,"reference_active":true,"delta_db":0.009444273849507683},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927249258823345e-05,"output_power":7.319665226220914e-05,"reference_active":true,"delta_db":0.23930403665698052},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3444445693793866e-05,"output_power":1.3881455052528356e-05,"reference_active":true,"delta_db":0.13892089967958496}],"active_band_count":3,"active_band_mae_db":0.12922307006202438}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010879087849669347,"log_rms_mae_db":2.9160615489983037,"rms_envelope_nrmse":0.19256435760872717,"envelope_correlation":0.9688402421349706,"rms_gain_error_db":0.11457461686629215,"reference_mean_power":0.0013248216670349025,"output_mean_power":0.0013602379394815117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.107952744901023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000335569442501778,"output_power":0.0003386573190294226,"reference_active":true,"delta_db":0.03978059577799526},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4091376965670345e-05,"output_power":3.635444199983551e-05,"reference_active":true,"delta_db":0.2791294016701698},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.097157965658072e-07,"output_power":8.881581892194343e-07,"reference_active":true,"delta_db":0.4015771263380734}],"active_band_count":3,"active_band_mae_db":0.24016237459541281}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.557718084358035e-05,"log_rms_mae_db":3.4901177655731477,"rms_envelope_nrmse":0.5081079369084555,"envelope_correlation":0.7969149048035492,"rms_gain_error_db":0.7243930221909672,"reference_mean_power":2.393470871827269e-05,"output_mean_power":2.8279221447622573e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3699291941082084e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.018836822609007e-06,"output_power":4.617680972757594e-06,"reference_active":true,"delta_db":0.603235521651334},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.745214583455975e-06,"output_power":6.591578408779102e-06,"reference_active":true,"delta_db":1.4273356605844723},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3999346848964708e-06,"output_power":1.9917401700880323e-06,"reference_active":true,"delta_db":1.531249086677714}],"active_band_count":3,"active_band_mae_db":1.1872734229711732}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.6234588308267947e-05,"log_rms_mae_db":2.4307471960109557,"rms_envelope_nrmse":0.29393642124794017,"envelope_correlation":0.9231208737108808,"rms_gain_error_db":0.7243698123697329,"reference_mean_power":2.405439007374668e-05,"output_mean_power":2.8420474897938488e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.116507982063627e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.688364032412793e-06,"output_power":4.498345064525842e-06,"reference_active":true,"delta_db":0.8621898847806975},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3132434905641704e-06,"output_power":4.633575587521519e-06,"reference_active":true,"delta_db":1.456628986654893},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.200263616160695e-07,"output_power":1.7515670204178476e-07,"reference_active":true,"delta_db":1.6415011799349144}],"active_band_count":3,"active_band_mae_db":1.3201066837901683}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},{"case":"librispeech_260-123286-0008","method":"hnm","status":"measured","mos":{"status":"measured","value":2.171614170074463,"predicted_mos":2.171614170074463,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59760,"input_sample_rate":16000,"input_seconds":3.735,"input_float32_sha256":"bffaf9307b1cf37427ab859f75647b67c96988548127ac05891d5eb86887cafd","internal_samples":59760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8380886770009965},"started_at":"2026-10-13T04:38:42.096202+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.576000067871064e-05,"native_input_samples":59760,"native_input_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","output_sha256":"dbac973d4fa173f5d94e3c4a07273092c50e05ea29dd4a8bf5fa937b20fd11b4","output_samples":59760,"length_delta_samples":0,"analysis_s":0.7827421709998816,"synthesis_s":0.3221723229999043,"total_s":1.104914493999786,"rtf":0.29582717376165624,"output_peak":0.5019204180729975,"output_rms":0.06514913864874479,"finished_at":"2026-10-13T04:38:43.204757+00:00","evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"dbac973d4fa173f5d94e3c4a07273092c50e05ea29dd4a8bf5fa937b20fd11b4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.747039244140858,"spectral_convergence":0.2952039111890726,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.433452210199871,"spectral_convergence":0.2911227330014557,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.58203750805042,"spectral_convergence":0.30431335436833074,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.0010659248000099823,"delta_db":-0.04316097368899355,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.0018705675891270961,"delta_db":-0.07124462960176077,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.0009206754450823678,"delta_db":-0.08604996198682761,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.0002669346231580366,"delta_db":-0.319583625760633,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":5.907214114334772e-05,"delta_db":-0.0034494314191841147,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":6.123566815254079e-05,"delta_db":0.4163517838991659,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15664006772609415,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07591203900179223,"pitch":null,"output_rms_dbfs":-23.72182643634796,"output_peak":0.5019204180729975,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.587509654130383},"common_prefix":{"samples":59760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.747039244140858,"spectral_convergence":0.2952039111890726,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.433452210199871,"spectral_convergence":0.2911227330014557,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.58203750805042,"spectral_convergence":0.30431335436833074,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.0010659248000099823,"delta_db":-0.04316097368899355,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.0018705675891270961,"delta_db":-0.07124462960176077,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.0009206754450823678,"delta_db":-0.08604996198682761,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.0002669346231580366,"delta_db":-0.319583625760633,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":5.907214114334772e-05,"delta_db":-0.0034494314191841147,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":6.123566815254079e-05,"delta_db":0.4163517838991659,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15664006772609415,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07591203900179223,"pitch":null,"output_rms_dbfs":-23.72182643634796,"output_peak":0.5019204180729975,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.587509654130383},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59760,"output_samples":59760,"intelligibility":{"estoi":{"value":0.8527406166189484,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06572101773044617,"rms_dbfs":-23.64591439734617,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.027337415210035,"mean":-4.259541012356719e-05,"zero_energy":false},"output":{"rms":0.06514913864874479,"rms_dbfs":-23.72182643634796,"peak":0.5019204180729975,"peak_dbfs":-5.9873027380397605,"crest_factor_db":17.7345236983082,"mean":-0.0005712067936625463,"zero_energy":false},"rms_difference_db":-0.07591203900178911},"full_scale":{"reference":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":2.7684499959750065e-07,"output_rms_dbfs":-65.57763316352246,"output_zero_energy":false,"output_peak":0.004863893743058792,"output_energy_fraction":1.1001939960120706e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":56,"both_estimated_voiced_frames":53,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.8833333333333333,"absolute_difference_cents_median":1.5416926438800356,"absolute_difference_cents_p95":12.536156365485974,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00019933639001390812,"log_rms_mae_db":1.002488706201167,"rms_envelope_nrmse":0.09347200124155022,"envelope_correlation":0.9923087515036225,"rms_gain_error_db":-0.07591205007516111,"reference_mean_power":0.004325042032530898,"output_mean_power":0.004250099792941194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":2.9832889988039093e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014406782878671795,"output_power":0.0013756889057598626,"reference_active":true,"delta_db":-0.20046776142422307},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010200208961765018,"output_power":0.00010393258464884834,"reference_active":true,"delta_db":0.0814265885668213},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7770617096993514e-05,"output_power":3.609338869317707e-05,"reference_active":true,"delta_db":1.138421280871043}],"active_band_count":3,"active_band_mae_db":0.4734385436206958}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00018454723874230657,"log_rms_mae_db":0.5159877757631325,"rms_envelope_nrmse":0.05280500980488157,"envelope_correlation":0.9974131047059355,"rms_gain_error_db":-0.07591209092921605,"reference_mean_power":0.004360109839005934,"output_mean_power":0.004284559921096563},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":2.740771226026708e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013018843976497203,"output_power":0.0012510111149961112,"reference_active":true,"delta_db":-0.17311253922384653},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.649956473480408e-05,"output_power":4.905221884820657e-05,"reference_active":true,"delta_db":0.23209769575942724},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9034855452992526e-06,"output_power":2.3758540381748284e-06,"reference_active":true,"delta_db":0.9627017280726076}],"active_band_count":3,"active_band_mae_db":0.45597065435196044}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00017949843182759416,"log_rms_mae_db":1.566003177095711,"rms_envelope_nrmse":0.1410036676306569,"envelope_correlation":0.9859831340261656,"rms_gain_error_db":-0.0711605645959828,"reference_mean_power":0.001722884360599687,"output_mean_power":0.0016948843605157004},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.0185329105758934e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006952905258544651,"output_power":0.0006576291523110499,"reference_active":true,"delta_db":-0.241852546803171},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.45148729518806e-05,"output_power":5.845493625394817e-05,"reference_active":true,"delta_db":-0.4283865511787272},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5243024554692384e-05,"output_power":2.7311807931830826e-05,"reference_active":true,"delta_db":2.5327930041209323}],"active_band_count":3,"active_band_mae_db":1.06767736736761}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00016126256915979383,"log_rms_mae_db":0.9099463931013806,"rms_envelope_nrmse":0.08002701989544335,"envelope_correlation":0.9952466260671909,"rms_gain_error_db":-0.07116057039137709,"reference_mean_power":0.0017368536839123,"output_mean_power":0.001708626654672354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":9.447738359038237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006450702777726167,"output_power":0.00061951193752766,"reference_active":true,"delta_db":-0.17557352503448226},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.981889248830953e-05,"output_power":2.6562191703507558e-05,"reference_active":true,"delta_db":-0.5022760242041179},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0738004145337803e-06,"output_power":1.7269026883104466e-06,"reference_active":true,"delta_db":2.063442982972004}],"active_band_count":3,"active_band_mae_db":0.9137641774035347}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":1.8134303085390424,"rms_envelope_nrmse":0.2131072466792601,"envelope_correlation":0.9643944184996955,"rms_gain_error_db":-0.20811403765565195,"reference_mean_power":0.00025607969857314143,"output_mean_power":0.00024409773307977451},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.9689769875456022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013521942671011414,"output_power":0.00011203617306895925,"reference_active":true,"delta_db":-0.8168082454928334},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4637782286244174e-05,"output_power":1.2679103135338916e-05,"reference_active":true,"delta_db":-0.62386748902102},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.589617396201186e-06,"output_power":5.033537339767984e-06,"reference_active":true,"delta_db":-0.45508787233676823}],"active_band_count":3,"active_band_mae_db":0.6319212022835405}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":1.233195701967199,"rms_envelope_nrmse":0.14669840527389705,"envelope_correlation":0.9819752436192397,"rms_gain_error_db":-0.20811403766088477,"reference_mean_power":0.00025815601744306934,"output_mean_power":0.0002460769010186944},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":1.774101433771278e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012602347588101524,"output_power":0.00010409363045861296,"reference_active":true,"delta_db":-0.8302729818858536},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.970334241514429e-06,"output_power":6.017602869142237e-06,"reference_active":true,"delta_db":-0.6383008074293461},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8125650262452097e-07,"output_power":3.141295267979424e-07,"reference_active":true,"delta_db":-0.8410849955237621}],"active_band_count":3,"active_band_mae_db":0.7698862616129873}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":2.213003652470718,"rms_envelope_nrmse":0.22829651881878393,"envelope_correlation":0.9677848958323163,"rms_gain_error_db":0.3089385002576408,"reference_mean_power":6.312530293815551e-05,"output_mean_power":6.777933839993144e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":4.348772018527747e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.235236211757751e-05,"output_power":2.3645246707745e-05,"reference_active":true,"delta_db":0.24420425100945775},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.837592049886237e-06,"output_power":9.720392833091302e-06,"reference_active":true,"delta_db":0.4134986623309266},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1202466049559018e-06,"output_power":2.37228624904965e-06,"reference_active":true,"delta_db":0.48780715004331926}],"active_band_count":3,"active_band_mae_db":0.38183668779456786}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":1.9153950661206003,"rms_envelope_nrmse":0.15878213676697436,"envelope_correlation":0.9839795205201757,"rms_gain_error_db":0.3089385288872821,"reference_mean_power":6.3637127906493e-05,"output_mean_power":6.832889910775994e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":3.943049206462403e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.36389645521954e-05,"output_power":2.5115884555658164e-05,"reference_active":true,"delta_db":0.263200287665659},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.125341563464701e-06,"output_power":4.414266416008423e-06,"reference_active":true,"delta_db":0.2939862836945978},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9054004218332713e-07,"output_power":2.120460949921802e-07,"reference_active":true,"delta_db":0.46444021940634733}],"active_band_count":3,"active_band_mae_db":0.340542263588868}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123286-0008","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0943894386291504,"predicted_mos":3.0943894386291504,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59760,"input_sample_rate":16000,"input_seconds":3.735,"input_float32_sha256":"55e3e28379c7d8832fa911a262d208275c40bd8d56857d67f43561e97aab18d2","internal_samples":59760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.526434565999807},"started_at":"2026-10-13T04:41:38.909148+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.876500027719885e-05,"native_input_samples":59760,"native_input_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","output_sha256":"66534accd572036134044d60d02a1cefcd64ac114cc1b9dd6b0024b45cf2e25b","output_samples":59760,"length_delta_samples":0,"analysis_s":0.8515449799997441,"synthesis_s":0.05677073199967708,"total_s":0.9083157119994212,"rtf":0.24319028433719445,"output_peak":0.6388974486427053,"output_rms":0.0764333375103198,"finished_at":"2026-10-13T04:41:39.821206+00:00","evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"66534accd572036134044d60d02a1cefcd64ac114cc1b9dd6b0024b45cf2e25b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.272910085871882,"spectral_convergence":0.5783702522657183,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.003545131675484,"spectral_convergence":0.3303078002700979,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.0886941508688945,"spectral_convergence":0.3510799717582633,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.0011740676382487407,"delta_db":0.3765045334208009,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.002876078352318627,"delta_db":1.797028451341447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.001286035814049941,"delta_db":1.3654150653321766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.0003813743435156712,"delta_db":1.2298820223377565,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":6.524423447585208e-05,"delta_db":0.4281448399758734,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":5.9254700357629054e-05,"delta_db":0.27353523827101506,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9117516917798448,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.311570872224738,"pitch":null,"output_rms_dbfs":-22.334343525121433,"output_peak":0.6388974486427053,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4550497894720875},"common_prefix":{"samples":59760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.272910085871882,"spectral_convergence":0.5783702522657183,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.003545131675484,"spectral_convergence":0.3303078002700979,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.0886941508688945,"spectral_convergence":0.3510799717582633,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.0011740676382487407,"delta_db":0.3765045334208009,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.002876078352318627,"delta_db":1.797028451341447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.001286035814049941,"delta_db":1.3654150653321766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.0003813743435156712,"delta_db":1.2298820223377565,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":6.524423447585208e-05,"delta_db":0.4281448399758734,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":5.9254700357629054e-05,"delta_db":0.27353523827101506,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9117516917798448,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.311570872224738,"pitch":null,"output_rms_dbfs":-22.334343525121433,"output_peak":0.6388974486427053,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4550497894720875},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59760,"output_samples":59760,"intelligibility":{"estoi":{"value":0.9020920292569362,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06572101773044617,"rms_dbfs":-23.64591439734617,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.027337415210035,"mean":-4.259541012356719e-05,"zero_energy":false},"output":{"rms":0.07643333751031982,"rms_dbfs":-22.334343525121433,"peak":0.6388974486427053,"peak_dbfs":-3.891376923305139,"crest_factor_db":18.442966601816295,"mean":2.4419464084321225e-10,"zero_energy":false},"rms_difference_db":1.3115708722247383},"full_scale":{"reference":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":2.6691769082236486e-07,"output_rms_dbfs":-65.73622641004852,"output_zero_energy":false,"output_peak":0.006840939218222141,"output_energy_fraction":7.706579356609949e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":55,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9166666666666666,"absolute_difference_cents_median":6.704237228188993,"absolute_difference_cents_p95":1209.0542441918894,"difference_over_600_cents_frames":7,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00019933639001390812,"log_rms_mae_db":1.810124329465292,"rms_envelope_nrmse":0.22698438716335018,"envelope_correlation":0.9818143536217464,"rms_gain_error_db":1.3115708903616592,"reference_mean_power":0.004325042032530898,"output_mean_power":0.005849886251671697},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":2.9832889988039093e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014406782878671795,"output_power":0.002072544287435935,"reference_active":true,"delta_db":1.579368085315636},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010200208961765018,"output_power":0.00012389656636511154,"reference_active":true,"delta_db":0.8445020180477487},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7770617096993514e-05,"output_power":3.126942941032755e-05,"reference_active":true,"delta_db":0.5153442610539948}],"active_band_count":3,"active_band_mae_db":0.9797381214724598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00018454723874230657,"log_rms_mae_db":1.4533263787645851,"rms_envelope_nrmse":0.20292267381956286,"envelope_correlation":0.9889674996532385,"rms_gain_error_db":1.3115709373123514,"reference_mean_power":0.004360109839005934,"output_mean_power":0.005897317687753327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":2.740771226026708e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013018843976497203,"output_power":0.0019042117219171344,"reference_active":true,"delta_db":1.6514281199550638},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.649956473480408e-05,"output_power":6.1025825196977826e-05,"reference_active":true,"delta_db":1.1806477306071939},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9034855452992526e-06,"output_power":1.8021804554410005e-06,"reference_active":true,"delta_db":-0.23751307737887356}],"active_band_count":3,"active_band_mae_db":1.0231963093137104}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00017949843182759416,"log_rms_mae_db":2.1231516475773224,"rms_envelope_nrmse":0.2204303352460543,"envelope_correlation":0.9853741270173783,"rms_gain_error_db":1.2629731713865342,"reference_mean_power":0.001722884360599687,"output_mean_power":0.0023043765436267225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.0185329105758934e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006952905258544651,"output_power":0.0009005802360622389,"reference_active":true,"delta_db":1.123561001430125},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.45148729518806e-05,"output_power":9.001015160099749e-05,"reference_active":true,"delta_db":1.4463164680761977},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5243024554692384e-05,"output_power":1.847585420671753e-05,"reference_active":true,"delta_db":0.835333773357514}],"active_band_count":3,"active_band_mae_db":1.1350704142879455}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00016126256915979383,"log_rms_mae_db":1.693324801764709,"rms_envelope_nrmse":0.18148442518136262,"envelope_correlation":0.9945790616821426,"rms_gain_error_db":1.2629731218944562,"reference_mean_power":0.0017368536839123,"output_mean_power":0.0023230606387971498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":9.447738359038237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006450702777726167,"output_power":0.000840980002634514,"reference_active":true,"delta_db":1.1517863716291294},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.981889248830953e-05,"output_power":4.39657753449133e-05,"reference_active":true,"delta_db":1.6862322726278653},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0738004145337803e-06,"output_power":1.1014649931017887e-06,"reference_active":true,"delta_db":0.11047131620708364}],"active_band_count":3,"active_band_mae_db":0.9828299868213595}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":1.6157638952090958,"rms_envelope_nrmse":0.2270153594612748,"envelope_correlation":0.9728947490425918,"rms_gain_error_db":0.9191225070054732,"reference_mean_power":0.00025607969857314143,"output_mean_power":0.00031643710355993426},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.9689769875456022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013521942671011414,"output_power":0.00015977879185513793,"reference_active":true,"delta_db":0.7248004265293506},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4637782286244174e-05,"output_power":1.8379233391605607e-05,"reference_active":true,"delta_db":0.9885210932550714},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.589617396201186e-06,"output_power":4.195248414692475e-06,"reference_active":true,"delta_db":-1.2462439991956422}],"active_band_count":3,"active_band_mae_db":0.9865218396600214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":1.14663262020995,"rms_envelope_nrmse":0.17749711035378674,"envelope_correlation":0.9855661432701158,"rms_gain_error_db":0.9191225296001964,"reference_mean_power":0.00025815601744306934,"output_mean_power":0.0003190028077447294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":1.774101433771278e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012602347588101524,"output_power":0.00014953634209392946,"reference_active":true,"delta_db":0.7429529906452823},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.970334241514429e-06,"output_power":8.896567746119472e-06,"reference_active":true,"delta_db":1.0596888630411452},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8125650262452097e-07,"output_power":2.4009595534515786e-07,"reference_active":true,"delta_db":-2.0083241564954744}],"active_band_count":3,"active_band_mae_db":1.2703220033939673}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":1.5647925127696087,"rms_envelope_nrmse":0.21256451362120038,"envelope_correlation":0.9732675226736258,"rms_gain_error_db":0.41808716998421175,"reference_mean_power":6.312530293815551e-05,"output_mean_power":6.950438298551632e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":4.348772018527747e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.235236211757751e-05,"output_power":2.462818391212413e-05,"reference_active":true,"delta_db":0.42108963467448846},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.837592049886237e-06,"output_power":9.253923813985792e-06,"reference_active":true,"delta_db":0.19991969386094594},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1202466049559018e-06,"output_power":2.0418992595800133e-06,"reference_active":true,"delta_db":-0.1635206481286738}],"active_band_count":3,"active_band_mae_db":0.2615099922213694}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":1.0326428959391056,"rms_envelope_nrmse":0.12362527352448718,"envelope_correlation":0.9914239003302892,"rms_gain_error_db":0.4180872057985466,"reference_mean_power":6.3637127906493e-05,"output_mean_power":7.00679306191924e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":3.943049206462403e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.36389645521954e-05,"output_power":2.5769271427765382e-05,"reference_active":true,"delta_db":0.3747369055269117},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.125341563464701e-06,"output_power":4.295461737925405e-06,"reference_active":true,"delta_db":0.17549942487589534},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9054004218332713e-07,"output_power":1.58514169626078e-07,"reference_active":true,"delta_db":-0.7991816703473016}],"active_band_count":3,"active_band_mae_db":0.4498060002500362}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123286-0008","method":"facodec","status":"measured","mos":{"status":"measured","value":3.697977304458618,"predicted_mos":3.697977304458618,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59800,"input_sample_rate":16000,"input_seconds":3.7375,"input_float32_sha256":"ba85e1be521578351f9922c1e48237429efb776ca73e05a154bf72170222b982","internal_samples":59800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8452422139998816},"started_at":"2026-10-13T04:49:06.363320+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.042699998128228e-05,"native_input_samples":59760,"native_input_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","output_sha256":"f058dfb51c612d87151980efe63591996c86e194ff7b4da9465d11283a39b532","output_samples":59800,"length_delta_samples":40,"analysis_s":1.4015885749995505,"synthesis_s":8.111959955000202,"total_s":9.513548529999753,"rtf":2.54713481392229,"output_peak":0.4501762390136719,"output_rms":0.06270106822651207,"finished_at":"2026-10-13T04:49:15.884270+00:00","evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"f058dfb51c612d87151980efe63591996c86e194ff7b4da9465d11283a39b532","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":59760,"output_samples":59800,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.054501202228515,"output_peak":0.4501762390136719},"common_prefix":{"samples":59760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9993311036789297,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.189878367007046,"spectral_convergence":0.2408936826961555,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.373868217461056,"spectral_convergence":0.24956404046553382,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.23700382041964,"spectral_convergence":0.2747395806349392,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.000916688395137694,"delta_db":-0.6982093066149679,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.00171431199894075,"delta_db":-0.4500800322313321,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.00095271179919897,"delta_db":0.06249987292478837,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.00024652826431682947,"delta_db":-0.6649655276995013,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":6.0911946479400056e-05,"delta_db":0.12974822045812762,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":4.290302696628316e-05,"delta_db":-1.128813481358024,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5223860735477902,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.40568085407746735,"pitch":null,"output_rms_dbfs":-24.051595251423635,"output_peak":0.4501762390136719,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.266916801629248},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59760,"output_samples":59760,"intelligibility":{"estoi":{"value":0.8813965719458546,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06572101773044617,"rms_dbfs":-23.64591439734617,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.027337415210035,"mean":-4.259541012356719e-05,"zero_energy":false},"output":{"rms":0.06272204900224423,"rms_dbfs":-24.051595251423635,"peak":0.4501762390136719,"peak_dbfs":-6.9323486290722744,"crest_factor_db":17.119246622351362,"mean":4.7947021541168704e-05,"zero_energy":false},"rms_difference_db":-0.40568085407746324},"full_scale":{"reference":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":2.077071610039924e-07,"output_rms_dbfs":-66.82548530278056,"output_zero_energy":false,"output_peak":0.003962693270295858,"output_energy_fraction":8.905553933867524e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":63,"both_estimated_voiced_frames":58,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9666666666666667,"absolute_difference_cents_median":2.8194235893952557,"absolute_difference_cents_p95":19.275554409836946,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00019933639001390812,"log_rms_mae_db":1.4468111510831767,"rms_envelope_nrmse":0.10954567411969031,"envelope_correlation":0.9909150249303104,"rms_gain_error_db":-0.40568090318614436,"reference_mean_power":0.004325042032530898,"output_mean_power":0.003939328899878605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":2.9832889988039093e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014406782878671795,"output_power":0.0013786548967830343,"reference_active":true,"delta_db":-0.19111443355395208},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010200208961765018,"output_power":9.459100626053425e-05,"reference_active":true,"delta_db":-0.3275922329859657},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7770617096993514e-05,"output_power":3.170397535664761e-05,"reference_active":true,"delta_db":0.5752819134045348}],"active_band_count":3,"active_band_mae_db":0.36466285998148423}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00018454723874230657,"log_rms_mae_db":1.1916350507281201,"rms_envelope_nrmse":0.09620418892475498,"envelope_correlation":0.9930302342817193,"rms_gain_error_db":-0.4056825279773264,"reference_mean_power":0.004360109839005934,"output_mean_power":0.003971267825923756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":2.740771226026708e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013018843976497203,"output_power":0.001246005473305579,"reference_active":true,"delta_db":-0.19052472212106056},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.649956473480408e-05,"output_power":4.2912450657614204e-05,"reference_active":true,"delta_db":-0.3486557057365606},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9034855452992526e-06,"output_power":2.066223731416983e-06,"reference_active":true,"delta_db":0.3562776207108901}],"active_band_count":3,"active_band_mae_db":0.2984860161895037}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00017949843182759416,"log_rms_mae_db":1.4292723067590367,"rms_envelope_nrmse":0.1167148800244345,"envelope_correlation":0.9905435026902292,"rms_gain_error_db":-0.18164414104010307,"reference_mean_power":0.001722884360599687,"output_mean_power":0.001652310697572762},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.0185329105758934e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006952905258544651,"output_power":0.0006834586733048271,"reference_active":true,"delta_db":-0.07454052442759994},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.45148729518806e-05,"output_power":6.471706473134749e-05,"reference_active":true,"delta_db":0.013589650876183296},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5243024554692384e-05,"output_power":1.9166533548256756e-05,"reference_active":true,"delta_db":0.9947242442333215}],"active_band_count":3,"active_band_mae_db":0.3609514731790349}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00016126256915979383,"log_rms_mae_db":0.9341415344425528,"rms_envelope_nrmse":0.08926318047028818,"envelope_correlation":0.9942983122702792,"rms_gain_error_db":-0.1816444572851889,"reference_mean_power":0.0017368536839123,"output_mean_power":0.0016657076810848847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":9.447738359038237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006450702777726167,"output_power":0.0006375865437434123,"reference_active":true,"delta_db":-0.0506788911839359},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.981889248830953e-05,"output_power":3.023333106871628e-05,"reference_active":true,"delta_db":0.0599449064408788},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0738004145337803e-06,"output_power":1.3088648016926432e-06,"reference_active":true,"delta_db":0.8597122142789749}],"active_band_count":3,"active_band_mae_db":0.3234453373012632}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":1.3771333317763486,"rms_envelope_nrmse":0.15472253317690704,"envelope_correlation":0.983072681496045,"rms_gain_error_db":-0.5127073086726942,"reference_mean_power":0.00025607969857314143,"output_mean_power":0.0002275644506837934},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.9689769875456022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013521942671011414,"output_power":0.00012169997369288018,"reference_active":true,"delta_db":-0.4574860597406103},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4637782286244174e-05,"output_power":1.4120440338212374e-05,"reference_active":true,"delta_db":-0.15627043266814833},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.589617396201186e-06,"output_power":7.969711152215605e-06,"reference_active":true,"delta_db":1.5406049958760313}],"active_band_count":3,"active_band_mae_db":0.71812049609493}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":1.066759810543813,"rms_envelope_nrmse":0.11651479535484952,"envelope_correlation":0.9908945915296218,"rms_gain_error_db":-0.5127074954125102,"reference_mean_power":0.00025815601744306934,"output_mean_power":0.00022940955531207025},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":1.774101433771278e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012602347588101524,"output_power":0.00011498132343091217,"reference_active":true,"delta_db":-0.39824150762620275},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.970334241514429e-06,"output_power":6.883368764652302e-06,"reference_active":true,"delta_db":-0.05452567131740104},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8125650262452097e-07,"output_power":5.151209183381896e-07,"reference_active":true,"delta_db":1.306919266823293}],"active_band_count":3,"active_band_mae_db":0.5865621485889656}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":1.6503536102782035,"rms_envelope_nrmse":0.21357828324935543,"envelope_correlation":0.9744840042358105,"rms_gain_error_db":-0.9263502441679912,"reference_mean_power":6.312530293815551e-05,"output_mean_power":5.099979740640645e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":4.348772018527747e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.235236211757751e-05,"output_power":1.5455702665977795e-05,"reference_active":true,"delta_db":-1.6023467030580214},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.837592049886237e-06,"output_power":6.309718692056415e-06,"reference_active":true,"delta_db":-1.4632395291428513},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1202466049559018e-06,"output_power":1.9102906405077807e-06,"reference_active":true,"delta_db":-0.4528692860962503}],"active_band_count":3,"active_band_mae_db":1.172818506099041}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":1.3700849979833376,"rms_envelope_nrmse":0.18252550160221467,"envelope_correlation":0.982613572838727,"rms_gain_error_db":-0.9263504683842666,"reference_mean_power":6.3637127906493e-05,"output_mean_power":5.1413305158782904e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":3.943049206462403e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.36389645521954e-05,"output_power":1.6353542595906957e-05,"reference_active":true,"delta_db":-1.6001660291532374},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.125341563464701e-06,"output_power":2.9145618927641616e-06,"reference_active":true,"delta_db":-1.5088663009685224},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9054004218332713e-07,"output_power":1.6505049668772816e-07,"reference_active":true,"delta_db":-0.623694214352612}],"active_band_count":3,"active_band_mae_db":1.2442421814914573}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123286-0008","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.987429141998291,"predicted_mos":2.987429141998291,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":164864,"input_sample_rate":44100,"input_seconds":3.7384126984126986,"input_float32_sha256":"45f1055e29bec892b28b3c0283542802b4c650de2f0efa9e9c1bb2bd3291e774","internal_samples":59815,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":11.936964782002178},"started_at":"2026-10-13T07:08:01.331993+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0053667390002374304,"native_input_samples":164714,"native_input_sha256":"9c244273ae0903f46e9f94fc2310b3e8cf8c0068d90903e9dc8e1a3661aafbee","output_sha256":"225025af8def857af936153c94aa316ee5d0fed253ec61950a0686bcd84efa4d","output_samples":164864,"length_delta_samples":150,"analysis_s":0.1289326070000243,"synthesis_s":22.425774792000084,"total_s":22.554707399000108,"rtf":6.038725283193322,"output_peak":0.4813450276851654,"output_rms":0.061884633328871126,"finished_at":"2026-10-13T07:08:23.908429+00:00","evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"225025af8def857af936153c94aa316ee5d0fed253ec61950a0686bcd84efa4d","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":59760,"output_samples":59815,"length_delta_samples":55,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.163539180455608,"output_peak":0.4778982871800101},"common_prefix":{"samples":59760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.999080498202792,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.392553816504817,"spectral_convergence":0.4735017570037568,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.051915136856705,"spectral_convergence":0.36556202868196264,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.628451531585392,"spectral_convergence":0.42448224898786535,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.0007253273976583486,"delta_db":-1.7150858177665866,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.0018596201995954684,"delta_db":-0.09673613182209771,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.0009092437122252049,"delta_db":-0.14031250211825347,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.00026302949640305984,"delta_db":-0.38358817379612636,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":5.0795474924689735e-05,"delta_db":-0.6590263060800787,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":2.9459081371736464e-05,"delta_db":-2.7615008244538615,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9593749593395007,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.513629600809267,"pitch":null,"output_rms_dbfs":-24.15954399815544,"output_peak":0.4778982871800101,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.024306828315638},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59760,"output_samples":59760,"intelligibility":{"estoi":{"value":0.7488979039152488,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06572101773044617,"rms_dbfs":-23.64591439734617,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.027337415210035,"mean":-4.259541012356719e-05,"zero_energy":false},"output":{"rms":0.06194735960618909,"rms_dbfs":-24.15954399815544,"peak":0.4778982871800101,"peak_dbfs":-6.413290520348011,"crest_factor_db":17.74625347780743,"mean":-0.00018137167254600697,"zero_energy":false},"rms_difference_db":-0.5136296008092671},"full_scale":{"reference":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":1.324537167555902e-07,"output_rms_dbfs":-68.77935850550625,"output_zero_energy":false,"output_peak":0.004054310256333001,"output_energy_fraction":5.821950284879013e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":61,"both_estimated_voiced_frames":54,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9,"absolute_difference_cents_median":11.039521607582516,"absolute_difference_cents_p95":57.732669678582,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00019933639001390812,"log_rms_mae_db":1.7302419128076207,"rms_envelope_nrmse":0.11724689381047466,"envelope_correlation":0.9912208704564542,"rms_gain_error_db":-0.5136295904231702,"reference_mean_power":0.004325042032530898,"output_mean_power":0.003842619421327627},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":2.9832889988039093e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014406782878671795,"output_power":0.0013678693124212005,"reference_active":true,"delta_db":-0.22522404624442008},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010200208961765018,"output_power":9.880524903741909e-05,"reference_active":true,"delta_db":-0.13829051689415633},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7770617096993514e-05,"output_power":2.3952261333606102e-05,"reference_active":true,"delta_db":-0.6423900897137863}],"active_band_count":3,"active_band_mae_db":0.33530155095078756}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00018454723874230657,"log_rms_mae_db":1.443814953979403,"rms_envelope_nrmse":0.08661615809925031,"envelope_correlation":0.9966594048533779,"rms_gain_error_db":-0.5136295508342967,"reference_mean_power":0.004360109839005934,"output_mean_power":0.0038737757397950593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":2.740771226026708e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013018843976497203,"output_power":0.0012424736733726075,"reference_active":true,"delta_db":-0.20285226968356898},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.649956473480408e-05,"output_power":4.583314609699475e-05,"reference_active":true,"delta_db":-0.06269218357176382},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9034855452992526e-06,"output_power":1.5772471237374767e-06,"reference_active":true,"delta_db":-0.8164983911389052}],"active_band_count":3,"active_band_mae_db":0.36068094813141266}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00017949843182759416,"log_rms_mae_db":2.2863052839282334,"rms_envelope_nrmse":0.15382701149169312,"envelope_correlation":0.9835429205722468,"rms_gain_error_db":-0.05020660669496218,"reference_mean_power":0.001722884360599687,"output_mean_power":0.0017030816438624399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.0185329105758934e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006952905258544651,"output_power":0.0006481355460327442,"reference_active":true,"delta_db":-0.30500471313296973},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.45148729518806e-05,"output_power":7.409779155212065e-05,"reference_active":true,"delta_db":0.6014541787950318},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5243024554692384e-05,"output_power":1.0877672464970486e-05,"reference_active":true,"delta_db":-1.4653517148973918}],"active_band_count":3,"active_band_mae_db":0.7906035356084645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00016126256915979383,"log_rms_mae_db":1.6630261290275161,"rms_envelope_nrmse":0.11093012302258237,"envelope_correlation":0.9909881243798478,"rms_gain_error_db":-0.05020664554703509,"reference_mean_power":0.0017368536839123,"output_mean_power":0.0017168903893549402},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":9.447738359038237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006450702777726167,"output_power":0.0006005060487105081,"reference_active":true,"delta_db":-0.31089645549093053},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.981889248830953e-05,"output_power":3.4957785633831276e-05,"reference_active":true,"delta_db":0.690524057158626},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0738004145337803e-06,"output_power":8.879365427487851e-07,"reference_active":true,"delta_db":-0.8254163767506084}],"active_band_count":3,"active_band_mae_db":0.608945629800055}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":2.9777411579842084,"rms_envelope_nrmse":0.27114132725456463,"envelope_correlation":0.9456293018382602,"rms_gain_error_db":-0.6240707076302655,"reference_mean_power":0.00025607969857314143,"output_mean_power":0.00022180333874225714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.9689769875456022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013521942671011414,"output_power":0.0001236224487991056,"reference_active":true,"delta_db":-0.3894174817845347},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4637782286244174e-05,"output_power":1.3831485193463665e-05,"reference_active":true,"delta_db":-0.24606467261548787},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.589617396201186e-06,"output_power":5.446637752993605e-06,"reference_active":true,"delta_db":-0.11253589859819782}],"active_band_count":3,"active_band_mae_db":0.24933935099940682}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":2.549656307318477,"rms_envelope_nrmse":0.22008164746522504,"envelope_correlation":0.9627543856743618,"rms_gain_error_db":-0.624070677469003,"reference_mean_power":0.00025815601744306934,"output_mean_power":0.0002236017431369552},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":1.774101433771278e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012602347588101524,"output_power":0.00011358483781244766,"reference_active":true,"delta_db":-0.45131091576418547},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.970334241514429e-06,"output_power":6.74600862237775e-06,"reference_active":true,"delta_db":-0.14206711944355824},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8125650262452097e-07,"output_power":4.0056352392422026e-07,"reference_active":true,"delta_db":0.21454139430882133}],"active_band_count":3,"active_band_mae_db":0.2693064765055217}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":2.843571403654731,"rms_envelope_nrmse":0.2735425094163356,"envelope_correlation":0.972942642113457,"rms_gain_error_db":-1.9893629203023269,"reference_mean_power":6.312530293815551e-05,"output_mean_power":3.9927046289292114e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":4.348772018527747e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.235236211757751e-05,"output_power":1.2381356569818816e-05,"reference_active":true,"delta_db":-2.5655519358513628},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.837592049886237e-06,"output_power":4.222039833391819e-06,"reference_active":true,"delta_db":-3.2081162391877944},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1202466049559018e-06,"output_power":1.1376749642454381e-06,"reference_active":true,"delta_db":-2.7036817539101903}],"active_band_count":3,"active_band_mae_db":2.8257833096497826}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":2.4189441493490094,"rms_envelope_nrmse":0.24410335026445049,"envelope_correlation":0.9846939977760096,"rms_gain_error_db":-1.9893628244417225,"reference_mean_power":6.3637127906493e-05,"output_mean_power":4.025077883923574e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":3.943049206462403e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.36389645521954e-05,"output_power":1.2554219883734278e-05,"reference_active":true,"delta_db":-2.7483871847334744},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.125341563464701e-06,"output_power":2.0763596060642643e-06,"reference_active":true,"delta_db":-2.981573409649379},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9054004218332713e-07,"output_power":9.455665388384778e-08,"reference_active":true,"delta_db":-3.0429416178199964}],"active_band_count":3,"active_band_mae_db":2.9243007374009498}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123286-0008","method":"golf","status":"measured","mos":{"status":"measured","value":2.5970189571380615,"predicted_mos":2.5970189571380615,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89520,"input_sample_rate":24000,"input_seconds":3.73,"input_float32_sha256":"7cedc01c43a57ad42931ea5c57ecefbd59e4f0cef7743225a8535cdbb444d10b","internal_samples":59680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.365076333999241},"started_at":"2026-10-13T05:02:15.632163+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001755225000124483,"native_input_samples":89640,"native_input_sha256":"400adfaa56d4be3af628ae4172823020acf50847ad353ccf2077720a21eedc36","output_sha256":"7b4633caa6bc5267b2c222dfcc3ce6a901b314eb8dce99ff2c00df2d13268d5a","output_samples":89520,"length_delta_samples":-120,"analysis_s":0.5183613790004529,"synthesis_s":0.07613518000016484,"total_s":0.5944965590006177,"rtf":0.1591690921019057,"output_peak":0.42880749702453613,"output_rms":0.061027850107328675,"finished_at":"2026-10-13T05:02:16.236976+00:00","analysis_stages":{"external_f0_s":0.06257509199986089},"evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"7b4633caa6bc5267b2c222dfcc3ce6a901b314eb8dce99ff2c00df2d13268d5a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59760,"output_samples":59680,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.286147229112434,"output_peak":0.42558377803591874},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":0.998661311914324,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.184685782913934,"spectral_convergence":0.55219988327383,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.990027523420327,"spectral_convergence":0.26625498880687964,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.951398444421822,"spectral_convergence":0.3058455007024072,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010780139316047734,"output_power":0.0009083336546684027,"delta_db":-0.7437896791764975,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019037210220096655,"output_power":0.0017487676387439469,"delta_db":-0.3687119768331914,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009406929630887111,"output_power":0.0007660649123915238,"delta_db":-0.891823242032649,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028770350524054205,"output_power":0.00021058617577341204,"delta_db":-1.3551529524547758,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.9198327166825896e-05,"output_power":4.8627322071695716e-05,"delta_db":-0.8542908111664382,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.571230373064044e-05,"output_power":4.484243546442879e-05,"delta_db":-0.9426192532534186,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8593979858194952,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6460505742377014,"pitch":null,"output_rms_dbfs":-24.286147229112434,"output_peak":0.42558377803591874,"reference_rms_dbfs":-23.640096654874732,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.375370583585362},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.8729130891193249,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06576505191088318,"rms_dbfs":-23.640096654874732,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.021519672738595,"mean":-4.2641770104620475e-05,"zero_energy":false},"output":{"rms":0.061050979837455596,"rms_dbfs":-24.286147229112434,"peak":0.42558377803591874,"peak_dbfs":-7.420298687744823,"crest_factor_db":16.86584854136761,"mean":0.0005470684588023271,"zero_energy":false},"rms_difference_db":-0.6460505742377016},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":2.535462773899705e-07,"output_rms_dbfs":-65.95942761455538,"output_zero_energy":false,"output_peak":0.0037974534397322945,"output_energy_fraction":1.1489567451961004e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":66,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9333333333333333,"absolute_difference_cents_median":2.75723928520385,"absolute_difference_cents_p95":13.978271021007677,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00019933639001390812,"log_rms_mae_db":2.140645560239122,"rms_envelope_nrmse":0.14177116852638774,"envelope_correlation":0.9860645515287847,"rms_gain_error_db":-0.6460505850304272,"reference_mean_power":0.004330847456382475,"output_mean_power":0.0037322250935879517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":2.985347337343345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014430132709376598,"output_power":0.0013203202332696148,"reference_active":true,"delta_db":-0.3859104649446043},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001015002214457508,"output_power":8.560156790556945e-05,"reference_active":true,"delta_db":-0.7398527033617006},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7761779799021895e-05,"output_power":2.4112875946754488e-05,"reference_active":true,"delta_db":-0.6119829331644812}],"active_band_count":3,"active_band_mae_db":0.579248700490262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00018454723874230657,"log_rms_mae_db":1.719684865529357,"rms_envelope_nrmse":0.10820543430491829,"envelope_correlation":0.9934570862844061,"rms_gain_error_db":-0.6460521222115057,"reference_mean_power":0.004366009850069157,"output_mean_power":0.003762525906954301},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":2.7423296848155848e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001303710992051316,"output_power":0.0011881240473843977,"reference_active":true,"delta_db":-0.40319541232533446},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.615676509788632e-05,"output_power":3.4711898454617494e-05,"reference_active":true,"delta_db":-1.2375699698823392},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.92139359737678e-06,"output_power":1.8555855487720061e-06,"reference_active":true,"delta_db":-0.15135357585460613}],"active_band_count":3,"active_band_mae_db":0.5973729860207599}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00017949843182759416,"log_rms_mae_db":1.9393669176617034,"rms_envelope_nrmse":0.17458721556395823,"envelope_correlation":0.9839501193370943,"rms_gain_error_db":-0.9112576961551606,"reference_mean_power":0.001725196956724554,"output_mean_power":0.0013986624443873801},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":1.0210653468453997e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006973203453753609,"output_power":0.0005910424555200758,"reference_active":true,"delta_db":-0.7181365852244928},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.445739110288533e-05,"output_power":5.5829874180018795e-05,"reference_active":true,"delta_db":-0.6240607445760595},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5195072481627374e-05,"output_power":1.122044616953716e-05,"reference_active":true,"delta_db":-1.3169264949965218}],"active_band_count":3,"active_band_mae_db":0.8863746082656915}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00016126256915979383,"log_rms_mae_db":1.3825018242450053,"rms_envelope_nrmse":0.14165523137146058,"envelope_correlation":0.9916099095767754,"rms_gain_error_db":-0.9112576883442426,"reference_mean_power":0.00173920395907541,"output_mean_power":0.0014100182912924286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":9.470985622908502e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000646764801447424,"output_power":0.0005486815732529731,"reference_active":true,"delta_db":-0.7142600125572431},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9747454901296846e-05,"output_power":2.4813532210749336e-05,"reference_active":true,"delta_db":-0.7876122421371867},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0733587208437206e-06,"output_power":7.774838764590701e-07,"reference_active":true,"delta_db":-1.4005349785624648}],"active_band_count":3,"active_band_mae_db":0.9674690777522983}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":1.784120445799658,"rms_envelope_nrmse":0.22341581303805758,"envelope_correlation":0.9716233784746805,"rms_gain_error_db":-1.2595733670936584,"reference_mean_power":0.00025642342956140086,"output_mean_power":0.00019186703651132304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":1.971132962432638e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013535240536864215,"output_power":9.78277415133835e-05,"reference_active":true,"delta_db":-1.4100395072797536},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4668259724286499e-05,"output_power":1.0389300909174102e-05,"reference_active":true,"delta_db":-1.4979226613058627},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.545661768109656e-06,"output_power":4.631075061445109e-06,"reference_active":true,"delta_db":-0.7827155808626605}],"active_band_count":3,"active_band_mae_db":1.230225916482759}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":1.2672218174462575,"rms_envelope_nrmse":0.17826369850228835,"envelope_correlation":0.9868848594440809,"rms_gain_error_db":-1.2595733851978483,"reference_mean_power":0.0002585053488756031,"output_mean_power":0.00019342481726090507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.7759898912323287e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261233782042813,"output_power":9.044428311072063e-05,"reference_active":true,"delta_db":-1.444144740136748},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.9322915229978e-06,"output_power":4.810814365564752e-06,"reference_active":true,"delta_db":-1.5865821862444196},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.035147967337228e-07,"output_power":3.657903135474442e-07,"reference_active":true,"delta_db":-0.42627263974268165}],"active_band_count":3,"active_band_mae_db":1.1523331887079498}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":1.9445591437731913,"rms_envelope_nrmse":0.1797255357679482,"envelope_correlation":0.9806447557291528,"rms_gain_error_db":-0.5377586439667438,"reference_mean_power":6.321003471390329e-05,"output_mean_power":5.5848326559682315e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":4.337577345911101e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.230713931732826e-05,"output_power":1.9903035525444683e-05,"reference_active":true,"delta_db":-0.495245614465433},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.815272107664393e-06,"output_power":9.372884566346055e-06,"reference_active":true,"delta_db":0.2663754587960327},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1140530816844676e-06,"output_power":2.2916275520416683e-06,"reference_active":true,"delta_db":0.350281473011188}],"active_band_count":3,"active_band_mae_db":0.37063418209088456}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":1.5712051262972484,"rms_envelope_nrmse":0.12064177646576996,"envelope_correlation":0.9922613436301487,"rms_gain_error_db":-0.5377585999094824,"reference_mean_power":6.372324023766896e-05,"output_mean_power":5.6301762567367665e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.931853932037475e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.360737856692625e-05,"output_power":2.1848582899486916e-05,"reference_active":true,"delta_db":-0.33624490562043097},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.079395537963689e-06,"output_power":4.403377989492615e-06,"reference_active":true,"delta_db":0.3319015071837967},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9503966783183838e-07,"output_power":2.3307577939634994e-07,"reference_active":true,"delta_db":0.7737419652250083}],"active_band_count":3,"active_band_mae_db":0.48062945934307866}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123286-0008","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.5963706970214844,"predicted_mos":3.5963706970214844,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89760,"input_sample_rate":24000,"input_seconds":3.74,"input_float32_sha256":"b858294f158c330fa7980e14b81537a5f0f36f3a7282dbe6d64de51ba85b4b0e","internal_samples":59840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7499199240010057},"started_at":"2026-10-13T05:07:50.357063+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002381168000283651,"native_input_samples":89640,"native_input_sha256":"400adfaa56d4be3af628ae4172823020acf50847ad353ccf2077720a21eedc36","output_sha256":"bd6e535898c5b99a92639807dc9f473a35af4aa7d6b4955e12a4875a819966a8","output_samples":89760,"length_delta_samples":120,"analysis_s":1.3685377540004993,"synthesis_s":7.442639244999555,"total_s":8.811176999000054,"rtf":2.359083533868823,"output_peak":0.4237719476222992,"output_rms":0.06496793430475956,"finished_at":"2026-10-13T05:07:59.180653+00:00","analysis_stages":{"external_f0_s":0.9066423049998775},"evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"bd6e535898c5b99a92639807dc9f473a35af4aa7d6b4955e12a4875a819966a8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59760,"output_samples":59840,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.7428410690934,"output_peak":0.4236243755346119},"common_prefix":{"samples":59760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986631016042781,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.182754571451367,"spectral_convergence":0.6006361766230092,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.990229643019051,"spectral_convergence":0.30847105701270666,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.710539238911104,"spectral_convergence":0.30352362916911757,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.001297052186295872,"delta_db":0.8091478599977853,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.0017078793843911706,"delta_db":-0.4664067187332437,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.0008639418143039647,"delta_db":-0.3622706250813861,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.0002625367393514022,"delta_db":-0.39173183739553186,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":5.616431182406168e-05,"delta_db":-0.22267213945785416,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":4.200209840839315e-05,"delta_db":-1.220982943516519,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.57886868736372,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09111672300170359,"pitch":null,"output_rms_dbfs":-23.737031120347876,"output_peak":0.4236243755346119,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.29450781779384},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59760,"output_samples":59760,"intelligibility":{"estoi":{"value":0.8720883037482035,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06572101773044617,"rms_dbfs":-23.64591439734617,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.027337415210035,"mean":-4.259541012356719e-05,"zero_energy":false},"output":{"rms":0.06503519458397018,"rms_dbfs":-23.737031120347876,"peak":0.4236243755346119,"peak_dbfs":-7.460381166966058,"crest_factor_db":16.276649953381817,"mean":9.25116289312536e-06,"zero_energy":false},"rms_difference_db":-0.09111672300170426},"full_scale":{"reference":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":8.969641502368129e-08,"output_rms_dbfs":-70.4722491444249,"output_zero_energy":false,"output_peak":0.005589601120551293,"output_energy_fraction":3.5770757813341483e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":74,"both_estimated_voiced_frames":59,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9833333333333333,"absolute_difference_cents_median":6.927815039716922,"absolute_difference_cents_p95":28.76436906910963,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00019933639001390812,"log_rms_mae_db":2.4770860889228326,"rms_envelope_nrmse":0.16985204315440164,"envelope_correlation":0.9747994815789565,"rms_gain_error_db":-0.0911167067913963,"reference_mean_power":0.004325042032530898,"output_mean_power":0.004235246203201281},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":2.9832889988039093e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014406782878671795,"output_power":0.0014185707216776225,"reference_active":true,"delta_db":-0.0671601900270384},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010200208961765018,"output_power":0.00010027953308774062,"reference_active":true,"delta_db":-0.07396765652700547},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7770617096993514e-05,"output_power":2.4491734600351485e-05,"reference_active":true,"delta_db":-0.5456598576443707}],"active_band_count":3,"active_band_mae_db":0.22892923473280488}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00018454723874230657,"log_rms_mae_db":1.917291312376528,"rms_envelope_nrmse":0.108878176519239,"envelope_correlation":0.9890133238075901,"rms_gain_error_db":-0.09111664148015301,"reference_mean_power":0.004360109839005934,"output_mean_power":0.00426958600170806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":2.740771226026708e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013018843976497203,"output_power":0.0012845209142026655,"reference_active":true,"delta_db":-0.058312425382683775},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.649956473480408e-05,"output_power":4.510350138913243e-05,"reference_active":true,"delta_db":-0.13238630130381265},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9034855452992526e-06,"output_power":1.7498777506598238e-06,"reference_active":true,"delta_db":-0.3654187399602814}],"active_band_count":3,"active_band_mae_db":0.1853724888822593}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00017949843182759416,"log_rms_mae_db":2.241963278801536,"rms_envelope_nrmse":0.2016231133742497,"envelope_correlation":0.9715568990243324,"rms_gain_error_db":-0.31685390262908175,"reference_mean_power":0.001722884360599687,"output_mean_power":0.001601661507876536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.0185329105758934e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006952905258544651,"output_power":0.0006273212435138435,"reference_active":true,"delta_db":-0.44676316985866377},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.45148729518806e-05,"output_power":5.706621551733504e-05,"reference_active":true,"delta_db":-0.5328077413544162},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5243024554692384e-05,"output_power":1.5803566206799522e-05,"reference_active":true,"delta_db":0.15683950966159557}],"active_band_count":3,"active_band_mae_db":0.3788034736248918}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00016126256915979383,"log_rms_mae_db":1.4510461499356289,"rms_envelope_nrmse":0.12433538238880759,"envelope_correlation":0.9891950238057357,"rms_gain_error_db":-0.3168538885268734,"reference_mean_power":0.0017368536839123,"output_mean_power":0.0016146479490943355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":9.447738359038237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006450702777726167,"output_power":0.0005826811571547553,"reference_active":true,"delta_db":-0.44176057796922513},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.981889248830953e-05,"output_power":2.657567936245906e-05,"reference_active":true,"delta_db":-0.5000713382185041},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0738004145337803e-06,"output_power":1.1507016575944177e-06,"reference_active":true,"delta_db":0.30039171414902555}],"active_band_count":3,"active_band_mae_db":0.414074543445585}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":2.2112034590377125,"rms_envelope_nrmse":0.235817130263199,"envelope_correlation":0.9566372184861102,"rms_gain_error_db":-0.3655084067795824,"reference_mean_power":0.00025607969857314143,"output_mean_power":0.0002354096758150236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.9689769875456022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013521942671011414,"output_power":0.00012621855522753129,"reference_active":true,"delta_db":-0.299158856505215},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4637782286244174e-05,"output_power":1.313125007876134e-05,"reference_active":true,"delta_db":-0.47169211074835415},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.589617396201186e-06,"output_power":4.9277591540667076e-06,"reference_active":true,"delta_db":-0.5473260849270716}],"active_band_count":3,"active_band_mae_db":0.4393923507268802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":1.4493603974566833,"rms_envelope_nrmse":0.12914147428399186,"envelope_correlation":0.9868246706269387,"rms_gain_error_db":-0.3655083889538974,"reference_mean_power":0.00025815601744306934,"output_mean_power":0.00023731840112289923},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":1.774101433771278e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012602347588101524,"output_power":0.00011745144454201112,"reference_active":true,"delta_db":-0.3059309125620753},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.970334241514429e-06,"output_power":6.245814077877423e-06,"reference_active":true,"delta_db":-0.47664551653359116},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8125650262452097e-07,"output_power":3.6697397896523273e-07,"reference_active":true,"delta_db":-0.1658198883413451}],"active_band_count":3,"active_band_mae_db":0.3161321058123372}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":2.149956966865593,"rms_envelope_nrmse":0.2655621306504839,"envelope_correlation":0.955527914764321,"rms_gain_error_db":-0.8917501990066744,"reference_mean_power":6.312530293815551e-05,"output_mean_power":5.1407733334033e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":4.348772018527747e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.235236211757751e-05,"output_power":1.8359971578367775e-05,"reference_active":true,"delta_db":-0.8545142001010333},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.837592049886237e-06,"output_power":6.660159389608718e-06,"reference_active":true,"delta_db":-1.2284932757457285},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1202466049559018e-06,"output_power":1.1889161224054558e-06,"reference_active":true,"delta_db":-2.512351600842022}],"active_band_count":3,"active_band_mae_db":1.5317863588962612}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":1.4373999923960656,"rms_envelope_nrmse":0.16742336461731885,"envelope_correlation":0.9857839291316595,"rms_gain_error_db":-0.8917501895469877,"reference_mean_power":6.3637127906493e-05,"output_mean_power":5.182455143053443e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":3.943049206462403e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.36389645521954e-05,"output_power":1.8960509387388872e-05,"reference_active":true,"delta_db":-0.9577844861918148},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.125341563464701e-06,"output_power":2.988068945520478e-06,"reference_active":true,"delta_db":-1.4006929839989801},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9054004218332713e-07,"output_power":1.1320399201795046e-07,"reference_active":true,"delta_db":-2.2612451497562787}],"active_band_count":3,"active_band_mae_db":1.5399075399823579}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123286-0008","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.207517147064209,"predicted_mos":3.207517147064209,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89880,"input_sample_rate":24000,"input_seconds":3.745,"input_float32_sha256":"396bddfc30cf488db5bd0aba71916d0892562ce8c44e8dee4c4983d6fd77205f","internal_samples":59920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.130818249001095},"started_at":"2026-10-13T05:10:23.136160+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027097000001958804,"native_input_samples":89640,"native_input_sha256":"400adfaa56d4be3af628ae4172823020acf50847ad353ccf2077720a21eedc36","output_sha256":"62141c4914ab8444d7b8a9517d3ccc4fa9a1539372afdd8807f1f87ff55a64f6","output_samples":89880,"length_delta_samples":240,"analysis_s":0.4796894600003725,"synthesis_s":0.07076697299999068,"total_s":0.5504564330003632,"rtf":0.14737789370826324,"output_peak":0.477708101272583,"output_rms":0.06399233267320249,"finished_at":"2026-10-13T05:10:23.697100+00:00","analysis_stages":{"external_f0_s":0.07520313500026532},"evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"62141c4914ab8444d7b8a9517d3ccc4fa9a1539372afdd8807f1f87ff55a64f6","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59760,"output_samples":59920,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.874256907932864,"output_peak":0.47843301768230434},"common_prefix":{"samples":59760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973297730307076,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.721063670193204,"spectral_convergence":0.18812845478930112,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.204886896306198,"spectral_convergence":0.17737851036491467,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.502816735620159,"spectral_convergence":0.18842778360503765,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.0010407771066561248,"delta_db":-0.14684933327981461,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.0018168598099524815,"delta_db":-0.19776449954168102,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.0008908125793073906,"delta_db":-0.22925215485903833,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.0002641720167094297,"delta_db":-0.3647645919601072,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":5.169985900338247e-05,"delta_db":-0.5823829714828681,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":4.467279842262726e-05,"delta_db":-0.9532612379117804,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4123791315058816,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21673035191461748,"pitch":null,"output_rms_dbfs":-23.862644749260788,"output_peak":0.47843301768230434,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.476255767373186},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59760,"output_samples":59760,"intelligibility":{"estoi":{"value":0.9284729940826806,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06572101773044617,"rms_dbfs":-23.64591439734617,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.027337415210035,"mean":-4.259541012356719e-05,"zero_energy":false},"output":{"rms":0.06410143656776685,"rms_dbfs":-23.862644749260788,"peak":0.47843301768230434,"peak_dbfs":-6.403577127950502,"crest_factor_db":17.459067621310286,"mean":-0.00039685067452992933,"zero_energy":false},"rms_difference_db":-0.21673035191461665},"full_scale":{"reference":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":2.1261996358530837e-07,"output_rms_dbfs":-66.72395960565804,"output_zero_energy":false,"output_peak":0.003962465663172718,"output_energy_fraction":8.728074762672451e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":55,"both_estimated_voiced_frames":53,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.8833333333333333,"absolute_difference_cents_median":0.6704844707519609,"absolute_difference_cents_p95":4.709414794270779,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00019933639001390812,"log_rms_mae_db":0.9801663585709192,"rms_envelope_nrmse":0.06916046381077394,"envelope_correlation":0.9963570318271459,"rms_gain_error_db":-0.21673035066850463,"reference_mean_power":0.004325042032530898,"output_mean_power":0.004114502186085346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":2.9832889988039093e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014406782878671795,"output_power":0.001398563925402031,"reference_active":true,"delta_db":-0.12884689224432544},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010200208961765018,"output_power":8.694916035984008e-05,"reference_active":true,"delta_db":-0.6934367630547604},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7770617096993514e-05,"output_power":1.9365050979691953e-05,"reference_active":true,"delta_db":-1.5656688578009723}],"active_band_count":3,"active_band_mae_db":0.7959841710333527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00018454723874230657,"log_rms_mae_db":0.6244774102012742,"rms_envelope_nrmse":0.04688396339068935,"envelope_correlation":0.9986196678297291,"rms_gain_error_db":-0.21673034156493992,"reference_mean_power":0.004360109839005934,"output_mean_power":0.0041478629263792385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":2.740771226026708e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013018843976497203,"output_power":0.001272296317844923,"reference_active":true,"delta_db":-0.09984151968173745},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.649956473480408e-05,"output_power":3.8868430577206994e-05,"reference_active":true,"delta_db":-0.7785188252173183},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9034855452992526e-06,"output_power":1.1826060042080723e-06,"reference_active":true,"delta_db":-2.067095035898064}],"active_band_count":3,"active_band_mae_db":0.9818184602657066}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00017949843182759416,"log_rms_mae_db":1.2651506685078737,"rms_envelope_nrmse":0.09785516011446448,"envelope_correlation":0.9939017653639054,"rms_gain_error_db":-0.2910585499772277,"reference_mean_power":0.001722884360599687,"output_mean_power":0.0016112030440857768},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.0185329105758934e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006952905258544651,"output_power":0.0006518445914854072,"reference_active":true,"delta_db":-0.2802224522705991},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.45148729518806e-05,"output_power":5.651988242830621e-05,"reference_active":true,"delta_db":-0.5745859659585834},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5243024554692384e-05,"output_power":9.381107722073429e-06,"reference_active":true,"delta_db":-2.1081702628723944}],"active_band_count":3,"active_band_mae_db":0.9876595603671922}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00016126256915979383,"log_rms_mae_db":0.7467211520735446,"rms_envelope_nrmse":0.06596444257148473,"envelope_correlation":0.9975760808074811,"rms_gain_error_db":-0.2910585490888992,"reference_mean_power":0.0017368536839123,"output_mean_power":0.0016242668441481735},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":9.447738359038237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006450702777726167,"output_power":0.0006080663081629765,"reference_active":true,"delta_db":-0.25656091203255826},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.981889248830953e-05,"output_power":2.5543234830450515e-05,"reference_active":true,"delta_db":-0.6721561310383688},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0738004145337803e-06,"output_power":5.64115887175116e-07,"reference_active":true,"delta_db":-2.795552363666065}],"active_band_count":3,"active_band_mae_db":1.2414231355789973}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":1.1766697915582853,"rms_envelope_nrmse":0.12350841073894962,"envelope_correlation":0.9897177844700188,"rms_gain_error_db":-0.4495088216283494,"reference_mean_power":0.00025607969857314143,"output_mean_power":0.00023090017803194878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.9689769875456022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013521942671011414,"output_power":0.0001237404174524449,"reference_active":true,"delta_db":-0.3852751349688366},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4637782286244174e-05,"output_power":1.300740382691642e-05,"reference_active":true,"delta_db":-0.5128465990100762},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.589617396201186e-06,"output_power":4.047343615845005e-06,"reference_active":true,"delta_db":-1.4021200470570394}],"active_band_count":3,"active_band_mae_db":0.7667472603453174}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":0.7558004513545072,"rms_envelope_nrmse":0.07633941294416899,"envelope_correlation":0.9971510968768571,"rms_gain_error_db":-0.4495088060715643,"reference_mean_power":0.00025815601744306934,"output_mean_power":0.00023277233975703076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":1.774101433771278e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012602347588101524,"output_power":0.00011637165669790617,"reference_active":true,"delta_db":-0.34604236719597925},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.970334241514429e-06,"output_power":6.106572980087473e-06,"reference_active":true,"delta_db":-0.5745605213895021},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8125650262452097e-07,"output_power":2.630971164890213e-07,"reference_active":true,"delta_db":-1.6110117134686113}],"active_band_count":3,"active_band_mae_db":0.8438715340180308}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":1.134235388824269,"rms_envelope_nrmse":0.13357151259494285,"envelope_correlation":0.9910786184899921,"rms_gain_error_db":-0.6489265883651212,"reference_mean_power":6.312530293815551e-05,"output_mean_power":5.436392647977941e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":4.348772018527747e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.235236211757751e-05,"output_power":1.9206525620587906e-05,"reference_active":true,"delta_db":-0.6587461465058525},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.837592049886237e-06,"output_power":8.024783022842081e-06,"reference_active":true,"delta_db":-0.41900651703267633},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1202466049559018e-06,"output_power":1.3758294722217712e-06,"reference_active":true,"delta_db":-1.8782176804991622}],"active_band_count":3,"active_band_mae_db":0.9853234480125637}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":0.8088622553078489,"rms_envelope_nrmse":0.09324510397177314,"envelope_correlation":0.997373407366516,"rms_gain_error_db":-0.6489265650705994,"reference_mean_power":6.3637127906493e-05,"output_mean_power":5.480471380610327e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":3.943049206462403e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.36389645521954e-05,"output_power":2.0320539725079467e-05,"reference_active":true,"delta_db":-0.6569321053071088},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.125341563464701e-06,"output_power":3.7275391614950225e-06,"reference_active":true,"delta_db":-0.44037697554726907},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9054004218332713e-07,"output_power":1.2262234941270473e-07,"reference_active":true,"delta_db":-1.9141662418767635}],"active_band_count":3,"active_band_mae_db":1.003825107577047}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123286-0008","method":"magphase","status":"measured","mos":{"status":"measured","value":3.4348747730255127,"predicted_mos":3.4348747730255127,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":180501,"input_sample_rate":48000,"input_seconds":3.7604375,"input_float32_sha256":"fcb143183586638869e68511d3dcccd6ad81659d6fd6e31d8350eab0db1a7f73","internal_samples":60167,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.56276196699946},"started_at":"2026-10-13T05:13:57.424197+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00655737699889869,"native_input_samples":179280,"native_input_sha256":"f972dab5115135ac34e189eb1f5e5ab344b0aadc0c527a9e802c138751cb1553","output_sha256":"90de909f094166f4bc45823ea574d7a31135db8670e00ad2dc379b045581d900","output_samples":180501,"length_delta_samples":1221,"analysis_s":1.9239448240005004,"synthesis_s":0.7066248939991056,"total_s":2.630569717999606,"rtf":0.7043024680052493,"output_peak":0.4913914797564982,"output_rms":0.06303097013646397,"finished_at":"2026-10-13T05:14:00.069406+00:00","evaluation_input":{"reference":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"90de909f094166f4bc45823ea574d7a31135db8670e00ad2dc379b045581d900","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":59760,"output_samples":60167,"length_delta_samples":407,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.005974562469206,"output_peak":0.48439498754875043},"common_prefix":{"samples":59760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9932354945401964,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59760,"output_samples":59760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.767563813724583,"spectral_convergence":0.5580362181873086,"spectral_convergence_status":"ok","amplitude_floor":1.6701466153856754e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.170555140258089,"spectral_convergence":0.45246647402721424,"spectral_convergence_status":"ok","amplitude_floor":1.1197689318182202e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.212443496050082,"spectral_convergence":0.40089225854717564,"spectral_convergence_status":"ok","amplitude_floor":1.0416079453540003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0010765709682483164,"output_power":0.0009598626529780964,"delta_db":-0.4983356991039353,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019015067355649686,"output_power":0.0017638848650934956,"delta_db":-0.3262763431114226,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009390993860770502,"output_power":0.0008824335280104084,"delta_db":-0.2702955589416151,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028731828855173817,"output_power":0.00027741182798026256,"delta_db":-0.15238296735884604,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.911907847527107e-05,"output_power":6.124235396656177e-05,"delta_db":0.15324219453399668,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.5637714608274427e-05,"output_power":5.783961377496845e-05,"delta_db":0.1685610181330956,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2615156301971519,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3305824873784336,"pitch":null,"output_rms_dbfs":-23.976496884724604,"output_peak":0.48439498754875043,"reference_rms_dbfs":-23.64591439734617,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.050187483344252},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59760,"output_samples":59760,"intelligibility":{"estoi":{"value":0.6534908329427368,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06572101773044617,"rms_dbfs":-23.64591439734617,"peak":0.46673583984375,"peak_dbfs":-6.618576982136135,"crest_factor_db":17.027337415210035,"mean":-4.259541012356719e-05,"zero_energy":false},"output":{"rms":0.06326669615053242,"rms_dbfs":-23.976496884724604,"peak":0.48439498754875043,"peak_dbfs":-6.296007190479851,"crest_factor_db":17.680489694244756,"mean":-0.0006630875553356693,"zero_energy":false},"rms_difference_db":-0.33058248737843243},"full_scale":{"reference":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9734996383776885e-06,"threshold_dbfs":-54.008268221715824,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.3247566931541004e-07,"output_mean_power":4.647183158577752e-07,"output_rms_dbfs":-63.32810210415027,"output_zero_energy":false,"output_peak":0.00953816042466664,"output_energy_fraction":1.958345982372862e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":60,"output_estimated_voiced_frames":60,"both_estimated_voiced_frames":50,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.8333333333333334,"absolute_difference_cents_median":14.304435071003162,"absolute_difference_cents_p95":43.998362589160394,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00019933639001390812,"log_rms_mae_db":3.2860211298879203,"rms_envelope_nrmse":0.2859897031614727,"envelope_correlation":0.9271693116947726,"rms_gain_error_db":-0.33058246901990307,"reference_mean_power":0.004325042032530898,"output_mean_power":0.004008040354916564},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":2.9832889988039093e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014406782878671795,"output_power":0.001404678674392768,"reference_active":true,"delta_db":-0.10990021986568413},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010200208961765018,"output_power":0.00010302084803823242,"reference_active":true,"delta_db":0.04316051725650312},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7770617096993514e-05,"output_power":2.4924755232055334e-05,"reference_active":true,"delta_db":-0.46954628297236745}],"active_band_count":3,"active_band_mae_db":0.20753567336485157}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00018454723874230657,"log_rms_mae_db":2.535606828694397,"rms_envelope_nrmse":0.21150986987249334,"envelope_correlation":0.9577350952443355,"rms_gain_error_db":-0.330582413412916,"reference_mean_power":0.004360109839005934,"output_mean_power":0.0040405379367225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":2.740771226026708e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013018843976497203,"output_power":0.0012776154354576887,"reference_active":true,"delta_db":-0.08172272184095385},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.649956473480408e-05,"output_power":4.696285160925711e-05,"reference_active":true,"delta_db":0.04305571992228496},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9034855452992526e-06,"output_power":2.008602118519426e-06,"reference_active":true,"delta_db":0.23344333202564965}],"active_band_count":3,"active_band_mae_db":0.11940725792962949}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":0.00017949843182759416,"log_rms_mae_db":3.99725623648125,"rms_envelope_nrmse":0.34215111401597104,"envelope_correlation":0.9160940043794589,"rms_gain_error_db":-0.3104089668763847,"reference_mean_power":0.001722884360599687,"output_mean_power":0.0016040401401448549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.0185329105758934e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006952905258544651,"output_power":0.0006682175110529381,"reference_active":true,"delta_db":-0.17248459249752257},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.45148729518806e-05,"output_power":6.464096467339961e-05,"reference_active":true,"delta_db":0.00847982673862413},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5243024554692384e-05,"output_power":1.5874883530497412e-05,"reference_active":true,"delta_db":0.17639398453612234}],"active_band_count":3,"active_band_mae_db":0.11911946792408967}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":0.00016126256915979383,"log_rms_mae_db":2.9189042622344816,"rms_envelope_nrmse":0.25022535135391477,"envelope_correlation":0.9528594658228431,"rms_gain_error_db":-0.31040897033986564,"reference_mean_power":0.0017368536839123,"output_mean_power":0.001617045861024742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":9.447738359038237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006450702777726167,"output_power":0.0006218895246643591,"reference_active":true,"delta_db":-0.15893790363864468},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.981889248830953e-05,"output_power":2.998920947724723e-05,"reference_active":true,"delta_db":0.024735086380613995},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0738004145337803e-06,"output_power":1.2329990631739827e-06,"reference_active":true,"delta_db":0.6003917933740164}],"active_band_count":3,"active_band_mae_db":0.2613549277977583}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":6.626476450042254e-05,"log_rms_mae_db":3.9969543771659466,"rms_envelope_nrmse":0.45997571292160344,"envelope_correlation":0.8360305336021606,"rms_gain_error_db":-0.094988400977527,"reference_mean_power":0.00025607969857314143,"output_mean_power":0.00025053955968992145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":1.9689769875456022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013521942671011414,"output_power":0.00013639339787363685,"reference_active":true,"delta_db":0.03754258466868193},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4637782286244174e-05,"output_power":1.5297631267027708e-05,"reference_active":true,"delta_db":0.19148905080607787},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.589617396201186e-06,"output_power":5.0736226582614255e-06,"reference_active":true,"delta_db":-0.4206391767424123}],"active_band_count":3,"active_band_mae_db":0.21655693740572404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":5.5643778901005376e-05,"log_rms_mae_db":2.9905289917506406,"rms_envelope_nrmse":0.30703528837613947,"envelope_correlation":0.9200370594247704,"rms_gain_error_db":-0.0949884043714163,"reference_mean_power":0.00025815601744306934,"output_mean_power":0.00025257095838260134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":1.774101433771278e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012602347588101524,"output_power":0.00012819336546043822,"reference_active":true,"delta_db":0.07414095404340083},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.970334241514429e-06,"output_power":7.187019763468164e-06,"reference_active":true,"delta_db":0.1329523522843659},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8125650262452097e-07,"output_power":3.5933174054051453e-07,"reference_active":true,"delta_db":-0.2572167858758333}],"active_band_count":3,"active_band_mae_db":0.15477003073453335}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":746,"rms_floor":3.541043489510121e-05,"log_rms_mae_db":4.27790592439808,"rms_envelope_nrmse":0.5932101060122129,"envelope_correlation":0.7818961555183002,"rms_gain_error_db":0.38803219530070376,"reference_mean_power":6.312530293815551e-05,"output_mean_power":6.902504442714044e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":746,"frequency_resolution_hz":0.2680965147453083,"power_floor":4.348772018527747e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.235236211757751e-05,"output_power":2.2577327311832264e-05,"reference_active":true,"delta_db":0.043491045770368525},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.837592049886237e-06,"output_power":9.840270466913064e-06,"reference_active":true,"delta_db":0.46673085173334933},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1202466049559018e-06,"output_power":2.116610067291154e-06,"reference_active":true,"delta_db":-0.007455190255583568}],"active_band_count":3,"active_band_mae_db":0.1725590292531005}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59760,"uncovered_tail_samples":0,"metrics":{"frame_count":740,"rms_floor":3.370430375904166e-05,"log_rms_mae_db":3.431310674059834,"rms_envelope_nrmse":0.4300090923395805,"envelope_correlation":0.8783014259470839,"rms_gain_error_db":0.38803219292624513,"reference_mean_power":6.3637127906493e-05,"output_mean_power":6.958470492986222e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":740,"frequency_resolution_hz":0.27027027027027023,"power_floor":3.943049206462403e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.36389645521954e-05,"output_power":2.4198762785607137e-05,"reference_active":true,"delta_db":0.1016471289386605},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.125341563464701e-06,"output_power":4.60840152330691e-06,"reference_active":true,"delta_db":0.4809039911092782},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9054004218332713e-07,"output_power":1.769128884415213e-07,"reference_active":true,"delta_db":-0.32226783785253943}],"active_band_count":3,"active_band_mae_db":0.30160631930015935}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"hnm","status":"measured","mos":{"status":"measured","value":2.5726871490478516,"predicted_mos":2.5726871490478516,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f6478415a93c8074efa6ea3074561b6c3d059c24a5bab9110bf6bbb3c3ee5724","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2049843770000734},"started_at":"2026-10-13T04:38:43.205859+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2971999240107834e-05,"native_input_samples":96000,"native_input_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","output_sha256":"3ac1c188c3362718a055ea360773d9dbca3bc67a85d3edcbfe6c67a8711ecc2b","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9720239810003477,"synthesis_s":0.5240535749999253,"total_s":1.496077556000273,"rtf":0.24934625933337884,"output_peak":0.3750992250337336,"output_rms":0.052200877117155264,"finished_at":"2026-10-13T04:38:44.708012+00:00","evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"3ac1c188c3362718a055ea360773d9dbca3bc67a85d3edcbfe6c67a8711ecc2b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.057136432499027,"spectral_convergence":0.35607512099031824,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.468742205098093,"spectral_convergence":0.3511607170516739,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.504281830367697,"spectral_convergence":0.3667022356359908,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0008417355524001815,"delta_db":-0.010338920483299187,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.001125018170392831,"delta_db":-0.18552974213222967,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0004709042398680211,"delta_db":-0.04735320781067068,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00018263090053028834,"delta_db":-0.016411966086892084,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":6.415574258994999e-05,"delta_db":-0.07470796476051958,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":4.0486966019071923e-05,"delta_db":0.3963238760433363,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12177761288615792,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08607294105295392,"pitch":null,"output_rms_dbfs":-25.64644399208495,"output_peak":0.3750992250337336,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676720155988273},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.057136432499027,"spectral_convergence":0.35607512099031824,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.468742205098093,"spectral_convergence":0.3511607170516739,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.504281830367697,"spectral_convergence":0.3667022356359908,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0008417355524001815,"delta_db":-0.010338920483299187,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.001125018170392831,"delta_db":-0.18552974213222967,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0004709042398680211,"delta_db":-0.04735320781067068,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00018263090053028834,"delta_db":-0.016411966086892084,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":6.415574258994999e-05,"delta_db":-0.07470796476051958,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":4.0486966019071923e-05,"delta_db":0.3963238760433363,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12177761288615792,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08607294105295392,"pitch":null,"output_rms_dbfs":-25.64644399208495,"output_peak":0.3750992250337336,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676720155988273},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.820405893601631,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05272073392617238,"rms_dbfs":-25.560371051032,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.410487802791902,"mean":-3.097852071126302e-05,"zero_energy":false},"output":{"rms":0.052200877117155264,"rms_dbfs":-25.646443992084954,"peak":0.3750992250337336,"peak_dbfs":-8.517076662275912,"crest_factor_db":17.129367329809043,"mean":-0.00015366908485688502,"zero_energy":false},"rms_difference_db":-0.08607294105295438},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":2.0393548702511727e-06,"output_rms_dbfs":-56.90507195613736,"output_zero_energy":false,"output_peak":0.022156809148904496,"output_energy_fraction":8.856136504262668e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":96,"both_estimated_voiced_frames":75,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.8928571428571429,"absolute_difference_cents_median":1.8135663829874848,"absolute_difference_cents_p95":17.064144785495532,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001544797211376954,"log_rms_mae_db":1.262711561210058,"rms_envelope_nrmse":0.12006464300075743,"envelope_correlation":0.9860853791754499,"rms_gain_error_db":-0.08607101741040155,"reference_mean_power":0.002781784143506645,"output_mean_power":0.0027271958385316117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.487428460245723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008814675226478596,"output_power":0.0008479712651763876,"reference_active":true,"delta_db":-0.1682517974409018},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001997406840389717,"output_power":0.0001899072016198388,"reference_active":true,"delta_db":-0.21925098572277352},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0345220429658976e-05,"output_power":3.782330511065242e-05,"reference_active":true,"delta_db":-0.28032616530164906}],"active_band_count":3,"active_band_mae_db":0.22260964948844145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013989081102856241,"log_rms_mae_db":0.6231858673293273,"rms_envelope_nrmse":0.06521256324725028,"envelope_correlation":0.9954838834892262,"rms_gain_error_db":-0.08620091515818756,"reference_mean_power":0.002795409730019158,"output_mean_power":0.0027404720746094756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3091173503801554e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008259246223406256,"output_power":0.0007964500035684012,"reference_active":true,"delta_db":-0.15781894879887065},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.58341852709606e-05,"output_power":8.906159450020306e-05,"reference_active":true,"delta_db":-0.31829988885403804},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329771424834818e-06,"output_power":2.956264289888991e-06,"reference_active":true,"delta_db":-0.5167116462804671}],"active_band_count":3,"active_band_mae_db":0.33094349464445855}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000132208474335363,"log_rms_mae_db":1.9339362704582896,"rms_envelope_nrmse":0.18099737649567332,"envelope_correlation":0.974887666113824,"rms_gain_error_db":-0.22800232878046317,"reference_mean_power":0.001270855716483403,"output_mean_power":0.0012058575694808037},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.478174051796597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005667765898001897,"output_power":0.0005330378131324805,"reference_active":true,"delta_db":-0.2665388529543348},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011084740591226797,"output_power":9.994953552393513e-05,"reference_active":true,"delta_db":-0.44944753819598054},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2190579526022388e-05,"output_power":2.2921457968035175e-05,"reference_active":true,"delta_db":0.14073594044098528}],"active_band_count":3,"active_band_mae_db":0.2855741105304335}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012422751968092996,"log_rms_mae_db":1.104299810973103,"rms_envelope_nrmse":0.0957166044630471,"envelope_correlation":0.9927242775094457,"rms_gain_error_db":-0.22800347080908626,"reference_mean_power":0.0012772403358267607,"output_mean_power":0.001211915327617918},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.710312666254066e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574547781532275,"output_power":0.0005243067695344116,"reference_active":true,"delta_db":-0.26624177212269945},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253668134122962e-05,"output_power":4.81186375924474e-05,"reference_active":true,"delta_db":-0.3814931318291697},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7043530583570791e-06,"output_power":1.622944119377097e-06,"reference_active":true,"delta_db":-0.21255997651397382}],"active_band_count":3,"active_band_mae_db":0.286764960155281}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":1.8931300737701249,"rms_envelope_nrmse":0.24428345239587543,"envelope_correlation":0.9503227591866948,"rms_gain_error_db":-0.07358299345723829,"reference_mean_power":0.00020286667816607566,"output_mean_power":0.00019945844025330288},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.368709420467863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.826983537989379e-05,"output_power":7.195614430599184e-05,"reference_active":true,"delta_db":-0.36526536310339575},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2413787132715606e-05,"output_power":2.1657866919215718e-05,"reference_active":true,"delta_db":-0.14899562270952832},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388555256663481e-06,"output_power":5.873190662844752e-06,"reference_active":true,"delta_db":-0.3652855581668959}],"active_band_count":3,"active_band_mae_db":0.2931821813266067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":1.1149488007880346,"rms_envelope_nrmse":0.13889924267925272,"envelope_correlation":0.9818949670343441,"rms_gain_error_db":-0.07385351689094409,"reference_mean_power":0.00020386380860129344,"output_mean_power":0.00020042633350539135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.180981609421061e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.558718162537126e-05,"output_power":7.11755080343852e-05,"reference_active":true,"delta_db":-0.2611757663728228},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0576597086854286e-05,"output_power":1.0272490820045749e-05,"reference_active":true,"delta_db":-0.12670198486153414},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.384274936793116e-07,"output_power":3.859483906047043e-07,"reference_active":true,"delta_db":-1.4459799385343959}],"active_band_count":3,"active_band_mae_db":0.6112858965895843}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":2.1293902804532023,"rms_envelope_nrmse":0.17382663283669347,"envelope_correlation":0.9801933253920923,"rms_gain_error_db":0.1756215853117607,"reference_mean_power":5.251733918815235e-05,"output_mean_power":5.468457898007661e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5789636326786885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2240876569039163e-05,"output_power":2.2278665344962823e-05,"reference_active":true,"delta_db":0.007372699761656762},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.963351102568177e-06,"output_power":8.401131653100246e-06,"reference_active":true,"delta_db":0.23241926431568688},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4165297947437375e-06,"output_power":1.3164032097928956e-06,"reference_active":true,"delta_db":-0.3183678116972916}],"active_band_count":3,"active_band_mae_db":0.1860532585915451}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":1.2426403940045443,"rms_envelope_nrmse":0.09258766579410098,"envelope_correlation":0.994149525506406,"rms_gain_error_db":0.17084206922246192,"reference_mean_power":5.256947927681557e-05,"output_mean_power":5.467866242225799e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.1875752001118554e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1809094457380056e-05,"output_power":2.2163856095493067e-05,"reference_active":true,"delta_db":0.07007688260652797},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.006377197072698e-06,"output_power":4.265705045728833e-06,"reference_active":true,"delta_db":0.27238987300386397},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0500724932970186e-07,"output_power":8.073583381122563e-08,"reference_active":true,"delta_db":-1.141529473872474}],"active_band_count":3,"active_band_mae_db":0.49466540982762197}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.2427408695220947,"predicted_mos":3.2427408695220947,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"0f5e778aedf3f77b964ee07e84f1c6e6a83a19d9239d47c1aba9871811f3d7ab","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.29253436199906},"started_at":"2026-10-13T04:41:39.821707+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0816999494854826e-05,"native_input_samples":96000,"native_input_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","output_sha256":"13f41f4cc25776ef55ed555a7615f9a60e505d6e094d7043559f0e974bc71792","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3787294359999578,"synthesis_s":0.08648760699998093,"total_s":1.4652170429999387,"rtf":0.24420284049998978,"output_peak":0.47081481637281086,"output_rms":0.0578930002429295,"finished_at":"2026-10-13T04:41:41.291677+00:00","evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"13f41f4cc25776ef55ed555a7615f9a60e505d6e094d7043559f0e974bc71792","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.353321134438185,"spectral_convergence":0.5902627526751845,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.253363768723771,"spectral_convergence":0.32015597378298144,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.437633734447878,"spectral_convergence":0.37503727601641446,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0009018135510040955,"delta_db":0.2890719398305973,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.0014953039026351837,"delta_db":1.0501695569656424,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0006215273547084605,"delta_db":1.1579232611244408,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.000220399949263529,"delta_db":0.7999603302377702,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":7.42380543317491e-05,"delta_db":0.5592024846201955,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":3.831666518481689e-05,"delta_db":0.1570485853020337,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.66889602634678,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8128921909896125,"pitch":null,"output_rms_dbfs":-24.747478860042385,"output_peak":0.47081481637281086,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.681439545869945},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.353321134438185,"spectral_convergence":0.5902627526751845,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.253363768723771,"spectral_convergence":0.32015597378298144,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.437633734447878,"spectral_convergence":0.37503727601641446,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0009018135510040955,"delta_db":0.2890719398305973,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.0014953039026351837,"delta_db":1.0501695569656424,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0006215273547084605,"delta_db":1.1579232611244408,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.000220399949263529,"delta_db":0.7999603302377702,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":7.42380543317491e-05,"delta_db":0.5592024846201955,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":3.831666518481689e-05,"delta_db":0.1570485853020337,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.66889602634678,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8128921909896125,"pitch":null,"output_rms_dbfs":-24.747478860042385,"output_peak":0.47081481637281086,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.681439545869945},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8830257309814142,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05272073392617238,"rms_dbfs":-25.560371051032,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.410487802791902,"mean":-3.097852071126302e-05,"zero_energy":false},"output":{"rms":0.057893000242929496,"rms_dbfs":-24.747478860042385,"peak":0.47081481637281086,"peak_dbfs":-6.542997570450386,"crest_factor_db":18.204481289592,"mean":1.2053963969174454e-08,"zero_energy":false},"rms_difference_db":0.812892190989615},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":1.2309409848141833e-06,"output_rms_dbfs":-59.097627680142,"output_zero_energy":false,"output_peak":0.0193366709458649,"output_energy_fraction":4.346024961028558e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":107,"both_estimated_voiced_frames":79,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9404761904761905,"absolute_difference_cents_median":10.406083558948044,"absolute_difference_cents_p95":26.504834273700723,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001544797211376954,"log_rms_mae_db":1.9676243176707053,"rms_envelope_nrmse":0.1823682516370603,"envelope_correlation":0.979463454560841,"rms_gain_error_db":0.8129009290515316,"reference_mean_power":0.002781784143506645,"output_mean_power":0.003354389733364145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.487428460245723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008814675226478596,"output_power":0.0010767340905910973,"reference_active":true,"delta_db":0.8690214800680394},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001997406840389717,"output_power":0.00021739369248520432,"reference_active":true,"delta_db":0.3678040636167852},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0345220429658976e-05,"output_power":4.246486178909288e-05,"reference_active":true,"delta_db":0.22237622304369892}],"active_band_count":3,"active_band_mae_db":0.4864005889095078}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013989081102856241,"log_rms_mae_db":1.1905866792152504,"rms_envelope_nrmse":0.12594757093815281,"envelope_correlation":0.9940140214785573,"rms_gain_error_db":0.8128446284386326,"reference_mean_power":0.002795409730019158,"output_mean_power":0.0033707763279930427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3091173503801554e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008259246223406256,"output_power":0.001007221027543231,"reference_active":true,"delta_db":0.8618437044196656},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.58341852709606e-05,"output_power":0.00010567793241422805,"reference_active":true,"delta_db":0.42463852657346524},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329771424834818e-06,"output_power":3.2792215712338547e-06,"reference_active":true,"delta_db":-0.06643659817722693}],"active_band_count":3,"active_band_mae_db":0.4509729430567859}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000132208474335363,"log_rms_mae_db":2.1998030209868413,"rms_envelope_nrmse":0.23192253618755865,"envelope_correlation":0.9786001791912532,"rms_gain_error_db":1.1903041292435876,"reference_mean_power":0.001270855716483403,"output_mean_power":0.0016715780502003473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.478174051796597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005667765898001897,"output_power":0.0007174016012325265,"reference_active":true,"delta_db":1.0235043788059774},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011084740591226797,"output_power":0.00013290616903478372,"reference_active":true,"delta_db":0.7881960575996889},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2190579526022388e-05,"output_power":1.994959365587043e-05,"reference_active":true,"delta_db":-0.46234590185498775}],"active_band_count":3,"active_band_mae_db":0.7580154460868846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012422751968092996,"log_rms_mae_db":1.4894352181573123,"rms_envelope_nrmse":0.1828915693198215,"envelope_correlation":0.9913576696486658,"rms_gain_error_db":1.190295385852572,"reference_mean_power":0.0012772403358267607,"output_mean_power":0.0016799724659766625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.710312666254066e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574547781532275,"output_power":0.0007045629267518502,"reference_active":true,"delta_db":1.017101450846583},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253668134122962e-05,"output_power":6.198242785467049e-05,"reference_active":true,"delta_db":0.7180594794870889},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7043530583570791e-06,"output_power":1.5551201939835455e-06,"reference_active":true,"delta_db":-0.3979560333532192}],"active_band_count":3,"active_band_mae_db":0.7110389878956304}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":1.8655679672109058,"rms_envelope_nrmse":0.2181578877158958,"envelope_correlation":0.9703632603758756,"rms_gain_error_db":0.7638423754773754,"reference_mean_power":0.00020286667816607566,"output_mean_power":0.00024187721278237477},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.368709420467863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.826983537989379e-05,"output_power":9.712790253375282e-05,"reference_active":true,"delta_db":0.9374958989993886},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2413787132715606e-05,"output_power":2.240858948177872e-05,"reference_active":true,"delta_db":-0.0010072250966435183},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388555256663481e-06,"output_power":6.819236599274062e-06,"reference_active":true,"delta_db":0.2833310342836636}],"active_band_count":3,"active_band_mae_db":0.40727805279323187}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":1.1889882224058583,"rms_envelope_nrmse":0.13109379222699974,"envelope_correlation":0.9925127929670151,"rms_gain_error_db":0.7635743222662017,"reference_mean_power":0.00020386380860129344,"output_mean_power":0.0002430510858754692},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.180981609421061e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.558718162537126e-05,"output_power":9.328133251610943e-05,"reference_active":true,"delta_db":0.9134658895862027},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0576597086854286e-05,"output_power":1.074884845376168e-05,"reference_active":true,"delta_db":0.07015979567405567},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.384274936793116e-07,"output_power":4.7575677886235e-07,"reference_active":true,"delta_db":-0.5374224303963051}],"active_band_count":3,"active_band_mae_db":0.5070160385521878}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":1.9841896065526923,"rms_envelope_nrmse":0.22621672942617596,"envelope_correlation":0.967794571820544,"rms_gain_error_db":0.3544412145266507,"reference_mean_power":5.251733918815235e-05,"output_mean_power":5.6983200197336065e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5789636326786885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2240876569039163e-05,"output_power":2.4876802952075727e-05,"reference_active":true,"delta_db":0.48642666263734735},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.963351102568177e-06,"output_power":9.191160767037194e-06,"reference_active":true,"delta_db":0.6227449851913597},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4165297947437375e-06,"output_power":1.7315354971349622e-06,"reference_active":true,"delta_db":0.8720568544258862}],"active_band_count":3,"active_band_mae_db":0.660409500751531}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":1.158584021034408,"rms_envelope_nrmse":0.14121183273925722,"envelope_correlation":0.9872006901440886,"rms_gain_error_db":0.3496901362025762,"reference_mean_power":5.256947927681557e-05,"output_mean_power":5.697740803086366e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.1875752001118554e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1809094457380056e-05,"output_power":2.470615021724862e-05,"reference_active":true,"delta_db":0.541674442015941},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.006377197072698e-06,"output_power":4.759796557104124e-06,"reference_active":true,"delta_db":0.7483655518656185},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0500724932970186e-07,"output_power":1.3448351528346335e-07,"reference_active":true,"delta_db":1.0744977040173647}],"active_band_count":3,"active_band_mae_db":0.7881792326329746}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"facodec","status":"measured","mos":{"status":"measured","value":3.8841779232025146,"predicted_mos":3.8841779232025146,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e94064aa30039a2b38893535c6e328591ae47a309a93f3a7dc163d78b952a59b","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2031274910004868},"started_at":"2026-10-13T04:49:15.885469+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7985999622615054e-05,"native_input_samples":96000,"native_input_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","output_sha256":"097b40540a563ec4574355af974d4dd196b8d71406127758c5effaf0a09848ae","output_samples":96000,"length_delta_samples":0,"analysis_s":1.791659362000246,"synthesis_s":6.579447731000073,"total_s":8.371107093000319,"rtf":1.395184515500053,"output_peak":0.33372172713279724,"output_rms":0.05113802386720367,"finished_at":"2026-10-13T04:49:24.264019+00:00","evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"097b40540a563ec4574355af974d4dd196b8d71406127758c5effaf0a09848ae","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.795888550733184,"spectral_convergence":0.24884943895635167,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.044924843615379,"spectral_convergence":0.2533319118449658,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.062286023584275,"spectral_convergence":0.2827576064689743,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0006802950267538148,"delta_db":-0.9351226682106193,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.0011508440438333816,"delta_db":-0.0869603671849205,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0005134956715923687,"delta_db":0.3286886566435501,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00016801715965983102,"delta_db":-0.3786181852016759,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":7.071088579974836e-05,"delta_db":0.3477994502274823,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":3.173469740354763e-05,"delta_db":-0.6614848397144795,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4564456945304545,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26475011855273484,"pitch":null,"output_rms_dbfs":-25.825121169584733,"output_peak":0.33372172713279724,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.967699805977613},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.795888550733184,"spectral_convergence":0.24884943895635167,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.044924843615379,"spectral_convergence":0.2533319118449658,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.062286023584275,"spectral_convergence":0.2827576064689743,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0006802950267538148,"delta_db":-0.9351226682106193,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.0011508440438333816,"delta_db":-0.0869603671849205,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0005134956715923687,"delta_db":0.3286886566435501,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00016801715965983102,"delta_db":-0.3786181852016759,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":7.071088579974836e-05,"delta_db":0.3477994502274823,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":3.173469740354763e-05,"delta_db":-0.6614848397144795,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4564456945304545,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26475011855273484,"pitch":null,"output_rms_dbfs":-25.825121169584733,"output_peak":0.33372172713279724,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.967699805977613},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8882532394651589,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05272073392617238,"rms_dbfs":-25.560371051032,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.410487802791902,"mean":-3.097852071126302e-05,"zero_energy":false},"output":{"rms":0.051138023867203675,"rms_dbfs":-25.825121169584733,"peak":0.33372172713279724,"peak_dbfs":-9.53231034897141,"crest_factor_db":16.292810820613322,"mean":-6.2495555235623595e-06,"zero_energy":false},"rms_difference_db":-0.2647501185527332},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":5.793528078531469e-07,"output_rms_dbfs":-62.37056883871097,"output_zero_energy":false,"output_peak":0.005988695193082094,"output_energy_fraction":2.621575269044696e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":87,"both_estimated_voiced_frames":76,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9047619047619048,"absolute_difference_cents_median":4.808711873669546,"absolute_difference_cents_p95":23.94611079897983,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001544797211376954,"log_rms_mae_db":1.5620906244008135,"rms_envelope_nrmse":0.11448764281250366,"envelope_correlation":0.9878839916177673,"rms_gain_error_db":-0.26474079029131986,"reference_mean_power":0.002781784143506645,"output_mean_power":0.0026172749481390366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.487428460245723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008814675226478596,"output_power":0.0007716192731749889,"reference_active":true,"delta_db":-0.5780324877849712},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001997406840389717,"output_power":0.00018323985528120447,"reference_active":true,"delta_db":-0.374465927350352},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0345220429658976e-05,"output_power":4.205204460509548e-05,"reference_active":true,"delta_db":0.17995023788088477}],"active_band_count":3,"active_band_mae_db":0.37748288433873595}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013989081102856241,"log_rms_mae_db":1.3475343062758571,"rms_envelope_nrmse":0.09814019321442372,"envelope_correlation":0.9902552022773499,"rms_gain_error_db":-0.26459947743031925,"reference_mean_power":0.002795409730019158,"output_mean_power":0.0026301803251979785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3091173503801554e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008259246223406256,"output_power":0.0007244264664609775,"reference_active":true,"delta_db":-0.5694610482275545},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.58341852709606e-05,"output_power":8.932490997813478e-05,"reference_active":true,"delta_db":-0.30547867881290197},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329771424834818e-06,"output_power":3.341602822925162e-06,"reference_active":true,"delta_db":0.015404071484954518}],"active_band_count":3,"active_band_mae_db":0.29678126617513695}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000132208474335363,"log_rms_mae_db":1.8016518396456982,"rms_envelope_nrmse":0.11464712736178453,"envelope_correlation":0.9900896599537979,"rms_gain_error_db":0.05268174347801191,"reference_mean_power":0.001270855716483403,"output_mean_power":0.0012863656103036898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.478174051796597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005667765898001897,"output_power":0.0005461966936466238,"reference_active":true,"delta_db":-0.16062836952527904},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011084740591226797,"output_power":0.00010524810294876466,"reference_active":true,"delta_db":-0.22511257456734457},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2190579526022388e-05,"output_power":2.8442034834117958e-05,"reference_active":true,"delta_db":1.0779201963206013}],"active_band_count":3,"active_band_mae_db":0.4878870468044083}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012422751968092996,"log_rms_mae_db":1.3590765244795355,"rms_envelope_nrmse":0.08809223420149162,"envelope_correlation":0.9936241143553872,"rms_gain_error_db":0.05268854897964936,"reference_mean_power":0.0012772403358267607,"output_mean_power":0.0012928301752988258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.710312666254066e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574547781532275,"output_power":0.0005375112735584031,"reference_active":true,"delta_db":-0.15822064900384566},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253668134122962e-05,"output_power":5.034192823656607e-05,"reference_active":true,"delta_db":-0.18532789399890973},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7043530583570791e-06,"output_power":2.1089484397832267e-06,"reference_active":true,"delta_db":0.925063978398168}],"active_band_count":3,"active_band_mae_db":0.4228708404669745}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":1.6426909961641072,"rms_envelope_nrmse":0.1498362769906451,"envelope_correlation":0.9818616978842968,"rms_gain_error_db":-0.28773009483720124,"reference_mean_power":0.00020286667816607566,"output_mean_power":0.00018986185006330842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.368709420467863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.826983537989379e-05,"output_power":7.193199042415664e-05,"reference_active":true,"delta_db":-0.36672342602370134},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2413787132715606e-05,"output_power":1.9809640939399495e-05,"reference_active":true,"delta_db":-0.5363863926933446},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388555256663481e-06,"output_power":6.807082595236021e-06,"reference_active":true,"delta_db":0.27558364746379216}],"active_band_count":3,"active_band_mae_db":0.3928978220602794}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":1.2710693159386945,"rms_envelope_nrmse":0.1091684281411849,"envelope_correlation":0.9894961380377896,"rms_gain_error_db":-0.2875370771260706,"reference_mean_power":0.00020386380860129344,"output_mean_power":0.00019080353903695073},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.180981609421061e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.558718162537126e-05,"output_power":6.829030050312515e-05,"reference_active":true,"delta_db":-0.4408912857891072},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0576597086854286e-05,"output_power":9.220829153950257e-06,"reference_active":true,"delta_db":-0.595759849456569},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.384274936793116e-07,"output_power":5.479354391824561e-07,"reference_active":true,"delta_db":0.07602162368355433}],"active_band_count":3,"active_band_mae_db":0.3708909196430768}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":1.8625753956365412,"rms_envelope_nrmse":0.17781812352548845,"envelope_correlation":0.9786217040171283,"rms_gain_error_db":-0.05815674204147206,"reference_mean_power":5.251733918815235e-05,"output_mean_power":5.1818762836040434e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5789636326786885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2240876569039163e-05,"output_power":2.0248850200407872e-05,"reference_active":true,"delta_db":-0.40751532376785743},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.963351102568177e-06,"output_power":7.0649705133888646e-06,"reference_active":true,"delta_db":-0.5198551050428567},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4165297947437375e-06,"output_power":2.1424157391225594e-06,"reference_active":true,"delta_db":1.7967803646088174}],"active_band_count":3,"active_band_mae_db":0.9080502644731773}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":1.5765954148940131,"rms_envelope_nrmse":0.11708584486633863,"envelope_correlation":0.9900003461407305,"rms_gain_error_db":-0.05371748679980595,"reference_mean_power":5.256947927681557e-05,"output_mean_power":5.192325697225209e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.1875752001118554e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1809094457380056e-05,"output_power":1.9938648937481802e-05,"reference_active":true,"delta_db":-0.38941906718136643},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.006377197072698e-06,"output_power":3.5999655220066004e-06,"reference_active":true,"delta_db":-0.46453493932451173},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0500724932970186e-07,"output_power":1.5437341276827669e-07,"reference_active":true,"delta_db":1.673532230488794}],"active_band_count":3,"active_band_mae_db":0.8424954123315574}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.2459163665771484,"predicted_mos":3.2459163665771484,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"073540395ce109c515fa94c1560527abcb0edfbff0751e51126f85724626c4a5","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.07643775500037},"started_at":"2026-10-13T07:08:23.912194+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.010027440999692772,"native_input_samples":264600,"native_input_sha256":"d885a2774e937833b38a4f7620a8fc74a305a529b2026c5a09f2ee6999cb662f","output_sha256":"c5f44974963b7571a499edace7f9bec270a46b0bcc4e22d77dec76e62d0379b7","output_samples":264704,"length_delta_samples":104,"analysis_s":0.16445382299934863,"synthesis_s":4.189894850000201,"total_s":4.354348672999549,"rtf":0.7257247788332583,"output_peak":0.34495359659194946,"output_rms":0.04739347264235801,"finished_at":"2026-10-13T07:08:28.297169+00:00","evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"c5f44974963b7571a499edace7f9bec270a46b0bcc4e22d77dec76e62d0379b7","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.481202833691803,"output_peak":0.34305296012188896},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.906172811300838,"spectral_convergence":0.4967801069024534,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.676481469188424,"spectral_convergence":0.38502901527020233,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.481843910198712,"spectral_convergence":0.4378835149315967,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0005463063097193699,"delta_db":-1.887733463651255,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.001099777530867891,"delta_db":-0.2840766856120803,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.00036956795747303663,"delta_db":-1.0997361179591398,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00014774814174792835,"delta_db":-0.9369342978429894,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":5.8602014883821076e-05,"delta_db":-0.4679378428466566,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":2.7316227472211366e-05,"delta_db":-1.3126212496929999,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9981732762675201,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9191200105116127,"pitch":null,"output_rms_dbfs":-26.47949106154361,"output_peak":0.34305296012188896,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.688166063562658},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7392861157544195,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05272073392617238,"rms_dbfs":-25.560371051032,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.410487802791902,"mean":-3.097852071126302e-05,"zero_energy":false},"output":{"rms":0.04742697736694021,"rms_dbfs":-26.479491061543612,"peak":0.34305296012188896,"peak_dbfs":-9.292776577965181,"crest_factor_db":17.186714483578427,"mean":-0.00020644109855310377,"zero_energy":false},"rms_difference_db":-0.9191200105116124},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":9.895596384012168e-07,"output_rms_dbfs":-60.045580267753635,"output_zero_energy":false,"output_peak":0.014641616316572794,"output_energy_fraction":5.205928244063463e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":86,"both_estimated_voiced_frames":69,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.8214285714285714,"absolute_difference_cents_median":11.933344559972893,"absolute_difference_cents_p95":38.77981928749263,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001544797211376954,"log_rms_mae_db":1.8580073097531642,"rms_envelope_nrmse":0.15233749963648616,"envelope_correlation":0.9864008400707802,"rms_gain_error_db":-0.9191241708454896,"reference_mean_power":0.002781784143506645,"output_mean_power":0.002251184086937659},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.487428460245723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008814675226478596,"output_power":0.0007681903797036398,"reference_active":true,"delta_db":-0.5973745145389149},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001997406840389717,"output_power":0.00016068736422480956,"reference_active":true,"delta_db":-0.9448480581953886},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0345220429658976e-05,"output_power":2.5582712305606514e-05,"reference_active":true,"delta_db":-1.97845505681536}],"active_band_count":3,"active_band_mae_db":1.173559209849888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013989081102856241,"log_rms_mae_db":1.3816569305568576,"rms_envelope_nrmse":0.12041049512602521,"envelope_correlation":0.9951133486346736,"rms_gain_error_db":-0.9191626341641723,"reference_mean_power":0.002795409730019158,"output_mean_power":0.002262190681398201},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3091173503801554e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008259246223406256,"output_power":0.0007224401965887793,"reference_active":true,"delta_db":-0.5813851129530595},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.58341852709606e-05,"output_power":7.707490242259894e-05,"reference_active":true,"delta_db":-0.946074714940931},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329771424834818e-06,"output_power":2.3284332227306875e-06,"reference_active":true,"delta_db":-1.5535063470167756}],"active_band_count":3,"active_band_mae_db":1.0269887249702554}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000132208474335363,"log_rms_mae_db":2.8764034814770274,"rms_envelope_nrmse":0.18445560180206458,"envelope_correlation":0.9757412152082751,"rms_gain_error_db":-0.5652100269808761,"reference_mean_power":0.001270855716483403,"output_mean_power":0.0011157714486030475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.478174051796597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005667765898001897,"output_power":0.0005316774755946475,"reference_active":true,"delta_db":-0.27763641892204305},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011084740591226797,"output_power":0.0001051752498731553,"reference_active":true,"delta_db":-0.228119815880319},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2190579526022388e-05,"output_power":1.693709018576746e-05,"reference_active":true,"delta_db":-1.1732984430956137}],"active_band_count":3,"active_band_mae_db":0.5596848926326586}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012422751968092996,"log_rms_mae_db":2.1064129077982243,"rms_envelope_nrmse":0.13730983863452148,"envelope_correlation":0.9868533068315263,"rms_gain_error_db":-0.5652154242032346,"reference_mean_power":0.0012772403358267607,"output_mean_power":0.0011213755504848105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.710312666254066e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574547781532275,"output_power":0.0005225998242325775,"reference_active":true,"delta_db":-0.2804038298807093},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253668134122962e-05,"output_power":4.778091450158394e-05,"reference_active":true,"delta_db":-0.4120817796193802},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7043530583570791e-06,"output_power":1.6673044898193756e-06,"reference_active":true,"delta_db":-0.09544644597434018}],"active_band_count":3,"active_band_mae_db":0.26264401849147656}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":2.9303343740699606,"rms_envelope_nrmse":0.2782996697266319,"envelope_correlation":0.9407749279030003,"rms_gain_error_db":-0.9718663540376226,"reference_mean_power":0.00020286667816607566,"output_mean_power":0.00016219000325365748},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.368709420467863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.826983537989379e-05,"output_power":6.296292611420517e-05,"reference_active":true,"delta_db":-0.945095173455593},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2413787132715606e-05,"output_power":1.4754683171443834e-05,"reference_active":true,"delta_db":-1.8158535475961888},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388555256663481e-06,"output_power":4.2697338858643e-06,"reference_active":true,"delta_db":-1.7500184733525483}],"active_band_count":3,"active_band_mae_db":1.50365573146811}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":2.346124314454871,"rms_envelope_nrmse":0.21835325550172588,"envelope_correlation":0.9637647330887275,"rms_gain_error_db":-0.9722760704828581,"reference_mean_power":0.00020386380860129344,"output_mean_power":0.00016297182406765177},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.180981609421061e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.558718162537126e-05,"output_power":6.04728824113844e-05,"reference_active":true,"delta_db":-0.9688748282063687},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0576597086854286e-05,"output_power":7.168581559756075e-06,"reference_active":true,"delta_db":-1.6891272957262955},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.384274936793116e-07,"output_power":3.7124818114423426e-07,"reference_active":true,"delta_db":-1.614628935338861}],"active_band_count":3,"active_band_mae_db":1.4242103530905084}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":3.236180817927405,"rms_envelope_nrmse":0.2976316456839412,"envelope_correlation":0.9584493456765117,"rms_gain_error_db":-1.9291946944900604,"reference_mean_power":5.251733918815235e-05,"output_mean_power":3.368086563579691e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5789636326786885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2240876569039163e-05,"output_power":1.6164937249355847e-05,"reference_active":true,"delta_db":-1.385778768244072},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.963351102568177e-06,"output_power":4.919554297530947e-06,"reference_active":true,"delta_db":-2.0917010582849214},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4165297947437375e-06,"output_power":7.708357647438038e-07,"reference_active":true,"delta_db":-2.642638571679548}],"active_band_count":3,"active_band_mae_db":2.040039466069514}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":2.7199700407795895,"rms_envelope_nrmse":0.269421108614281,"envelope_correlation":0.970136907345606,"rms_gain_error_db":-1.9352204300772011,"reference_mean_power":5.256947927681557e-05,"output_mean_power":3.3667559177927494e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.1875752001118554e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1809094457380056e-05,"output_power":1.590382453091273e-05,"reference_active":true,"delta_db":-1.371360580195019},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.006377197072698e-06,"output_power":2.4385106919798896e-06,"reference_active":true,"delta_db":-2.156271711912086},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0500724932970186e-07,"output_power":5.4953164809647595e-08,"reference_active":true,"delta_db":-2.812265733138677}],"active_band_count":3,"active_band_mae_db":2.113299341748594}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"golf","status":"measured","mos":{"status":"measured","value":3.162174940109253,"predicted_mos":3.162174940109253,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"a446127cad163b4413da8a63279fae2f99c44a6d5eb5b940fe2ec8bcc69b5004","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2002278300005855},"started_at":"2026-10-13T05:02:16.237540+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002788323999993736,"native_input_samples":144000,"native_input_sha256":"9284a2a6222858e45935a7cb320bddcd7e1104acbdabc82b3c5c31c222e7b875","output_sha256":"44ecdab7350e3d9491e57f2b331dbbc8c7f6d888c15f158f832cb43aee47cdc3","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6085442490002606,"synthesis_s":0.11027281899987429,"total_s":0.7188170680001349,"rtf":0.11980284466668915,"output_peak":0.3650522232055664,"output_rms":0.048496403278163416,"finished_at":"2026-10-13T05:02:16.966310+00:00","analysis_stages":{"external_f0_s":0.09496335699986957},"evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"44ecdab7350e3d9491e57f2b331dbbc8c7f6d888c15f158f832cb43aee47cdc3","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.283076506153055,"output_peak":0.3654075423945052},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.348842520170328,"spectral_convergence":0.6024947744353234,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.249631698362483,"spectral_convergence":0.29481251011746973,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.326719864982118,"spectral_convergence":0.32439809484132004,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008451374229069515,"output_power":0.0007281108043282237,"delta_db":-0.6472985727407841,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011760758400662726,"output_power":0.0010162756052802587,"delta_db":-0.6342382754212418,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047686136960666583,"output_power":0.0003851515138559473,"delta_db":-0.927605325807745,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001836235582825525,"output_power":0.0001420982301896937,"delta_db":-1.113397300613441,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.536876019563004e-05,"output_power":5.0926301258403e-05,"delta_db":-1.0842811326143615,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.700093065337359e-05,"output_power":3.0819122333573946e-05,"delta_db":-0.7939238092135004,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8667907360685123,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7298747073964134,"pitch":null,"output_rms_dbfs":-26.283076506153055,"output_peak":0.3654075423945052,"reference_rms_dbfs":-25.553201798756643,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.641731361171643},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8577722171920247,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.052764267091578616,"rms_dbfs":-25.553201798756643,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.403318550516545,"mean":-2.9777246643983462e-05,"zero_energy":false},"output":{"rms":0.04851166434215693,"rms_dbfs":-26.283076506153055,"peak":0.3654075423945052,"peak_dbfs":-8.744449851957109,"crest_factor_db":17.538626654195948,"mean":0.0006814703240406604,"zero_energy":false},"rms_difference_db":-0.7298747073964122},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":1.1520330399731634e-06,"output_rms_dbfs":-59.38535065294524,"output_zero_energy":false,"output_peak":0.015843039431960974,"output_energy_fraction":5.8023525328975955e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":124,"both_estimated_voiced_frames":83,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.9880952380952381,"absolute_difference_cents_median":4.505585676705819,"absolute_difference_cents_p95":20.619504446879407,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001544797211376954,"log_rms_mae_db":2.1242169976105822,"rms_envelope_nrmse":0.16785780738298953,"envelope_correlation":0.9781455162792505,"rms_gain_error_db":-0.7301150106522525,"reference_mean_power":0.002786360447137067,"output_mean_power":0.0023551891695837855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4879331024564772e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008816216819900816,"output_power":0.0007666723206960743,"reference_active":true,"delta_db":-0.6067247824511853},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001996103967590323,"output_power":0.000171880586371834,"reference_active":true,"delta_db":-0.6495633126755791},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.031332858322273e-05,"output_power":4.5180458211250516e-05,"reference_active":true,"delta_db":0.49501972757778745}],"active_band_count":3,"active_band_mae_db":0.5837692742348506}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013989081102856241,"log_rms_mae_db":1.4927459630543467,"rms_envelope_nrmse":0.11904916594880653,"envelope_correlation":0.9915755601760086,"rms_gain_error_db":-0.731207401640453,"reference_mean_power":0.002799520676291984,"output_mean_power":0.002365717808966171},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.3094923207670522e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008255518837178394,"output_power":0.0007163659970498116,"reference_active":true,"delta_db":-0.6160940920611905},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.597859573078861e-05,"output_power":8.313760419463855e-05,"reference_active":true,"delta_db":-0.6237688631663203},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3228071837753745e-06,"output_power":3.3356611944750573e-06,"reference_active":true,"delta_db":0.016767918791025876}],"active_band_count":3,"active_band_mae_db":0.4188769580061789}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.000132208474335363,"log_rms_mae_db":2.4093014549895795,"rms_envelope_nrmse":0.20594905314343212,"envelope_correlation":0.9709269988331853,"rms_gain_error_db":-0.7600424405079824,"reference_mean_power":0.0012729769265026464,"output_mean_power":0.0010686027508164434},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.480082020259753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005666764326281202,"output_power":0.00047174724299291687,"reference_active":true,"delta_db":-0.7962578052630607},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011066907794650637,"output_power":8.792072876071167e-05,"reference_active":true,"delta_db":-0.9993501235471942},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.22093201677972e-05,"output_power":2.3018979021720185e-05,"reference_active":true,"delta_db":0.15550792225776866}],"active_band_count":3,"active_band_mae_db":0.6503719503560078}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012422751968092996,"log_rms_mae_db":1.5943502175253248,"rms_envelope_nrmse":0.1500739801140631,"envelope_correlation":0.9859910718346097,"rms_gain_error_db":-0.7600751535888518,"reference_mean_power":0.0012793765178768753,"output_mean_power":0.0010739668095586202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.711614729178152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005569733451067102,"output_power":0.0004603443302208151,"reference_active":true,"delta_db":-0.8275161319915125},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253494212944629e-05,"output_power":3.977079159644934e-05,"reference_active":true,"delta_db":-1.2088402293679925},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7005900650304245e-06,"output_power":1.838796272099971e-06,"reference_active":true,"delta_db":0.339339770400549}],"active_band_count":3,"active_band_mae_db":0.7918987105866847}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":2.1000622644436278,"rms_envelope_nrmse":0.25054623334631865,"envelope_correlation":0.956234960346455,"rms_gain_error_db":-1.1717839798872896,"reference_mean_power":0.00020319686943111155,"output_mean_power":0.0001551452968415467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.369576788373715e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.830593542253845e-05,"output_power":5.67010550381122e-05,"reference_active":true,"delta_db":-1.4020354200145466},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.247585435599763e-05,"output_power":1.4510731475970788e-05,"reference_active":true,"delta_db":-1.9002690371972997},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.352596004531884e-06,"output_power":6.86739079222249e-06,"reference_active":true,"delta_db":0.33840524726137156}],"active_band_count":3,"active_band_mae_db":1.2135699014910728}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":1.555327680526434,"rms_envelope_nrmse":0.17909249985530118,"envelope_correlation":0.9829926867463052,"rms_gain_error_db":-1.173234649181075,"reference_mean_power":0.00020417172614144464,"output_mean_power":0.0001558375585161728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1817225789199817e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.561792018550131e-05,"output_power":5.5265173949398465e-05,"reference_active":true,"delta_db":-1.3617318692456524},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0594565952365007e-05,"output_power":6.975004537095131e-06,"reference_active":true,"delta_db":-1.8153867440203495},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.352174294608598e-07,"output_power":4.6309133167175294e-07,"reference_active":true,"delta_db":-0.6286359610262872}],"active_band_count":3,"active_band_mae_db":1.2685848580974295}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":2.331837506317906,"rms_envelope_nrmse":0.21454647163674148,"envelope_correlation":0.9712646644961078,"rms_gain_error_db":-0.5961620997902507,"reference_mean_power":5.258823518374834e-05,"output_mean_power":4.584293202995535e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.578044623136896e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2280425197387916e-05,"output_power":2.0629417660026795e-05,"reference_active":true,"delta_db":-0.3343650600019873},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.921124723407294e-06,"output_power":7.807005898961842e-06,"reference_active":true,"delta_db":-0.06302344083719416},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4134644776050184e-06,"output_power":1.327408134016789e-06,"reference_active":true,"delta_db":-0.2728042395836868}],"active_band_count":3,"active_band_mae_db":0.2233975801409561}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":1.9171261377786881,"rms_envelope_nrmse":0.14864999971296372,"envelope_correlation":0.9877317770330014,"rms_gain_error_db":-0.6264910505516339,"reference_mean_power":5.227974676452078e-05,"output_mean_power":4.5256854961035e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.1891399226756614e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1832048098118347e-05,"output_power":2.0609765802031823e-05,"reference_active":true,"delta_db":-0.25021422730534976},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.971698293982887e-06,"output_power":3.84736514184171e-06,"reference_active":true,"delta_db":-0.13812844493661378},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0452457422452964e-07,"output_power":1.0308137644589055e-07,"reference_active":true,"delta_db":-0.060381980969049474}],"active_band_count":3,"active_band_mae_db":0.14957488440367098}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3090267181396484,"predicted_mos":3.3090267181396484,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"354b183e7db3b3a15875702844a701ba8bc59bf1a4ccd432facc614457925af6","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3805929610007297},"started_at":"2026-10-13T05:07:59.182117+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0038503320001836983,"native_input_samples":144000,"native_input_sha256":"9284a2a6222858e45935a7cb320bddcd7e1104acbdabc82b3c5c31c222e7b875","output_sha256":"81eb159a2d7ab226e950936490d2532efac3ee360b140621d4e73030844bbbf9","output_samples":144120,"length_delta_samples":120,"analysis_s":2.138920839000093,"synthesis_s":3.258758162999584,"total_s":5.397679001999677,"rtf":0.8996131669999462,"output_peak":0.36476701498031616,"output_rms":0.05292029774213537,"finished_at":"2026-10-13T05:08:04.593802+00:00","analysis_stages":{"external_f0_s":1.414388279999912},"evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"81eb159a2d7ab226e950936490d2532efac3ee360b140621d4e73030844bbbf9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.524919870108068,"output_peak":0.36494181350803384},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.607162590335421,"spectral_convergence":0.621694980020933,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.353953961041086,"spectral_convergence":0.34845055736445474,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.0122657254032505,"spectral_convergence":0.3495667879115765,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0010096486728666376,"delta_db":0.77960715838403,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.0010898703930401084,"delta_db":-0.3233765621232592,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.000444759001907671,"delta_db":-0.2954317453605926,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00016882000277840576,"delta_db":-0.3579155409940374,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":6.193067112066631e-05,"delta_db":-0.2280054635983833,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":2.9521542401613462e-05,"delta_db":-0.9754380138523661,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.49329574738544474,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.03900336190736632,"pitch":null,"output_rms_dbfs":-25.52136768912463,"output_peak":0.36494181350803384,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.657794092259919},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.83748070014103,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05272073392617238,"rms_dbfs":-25.560371051032,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.410487802791902,"mean":-3.097852071126302e-05,"zero_energy":false},"output":{"rms":0.05295800491063748,"rms_dbfs":-25.52136768912463,"peak":0.36494181350803384,"peak_dbfs":-8.755527482750571,"crest_factor_db":16.76584020637406,"mean":0.00019596693409659158,"zero_energy":false},"rms_difference_db":0.03900336190736908},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":6.997618670833443e-06,"output_rms_dbfs":-51.55049727709801,"output_zero_energy":false,"output_peak":0.06583474898097974,"output_energy_fraction":0.0002952528779410204,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":111,"both_estimated_voiced_frames":76,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9047619047619048,"absolute_difference_cents_median":7.246781824244562,"absolute_difference_cents_p95":28.14755186536101,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001544797211376954,"log_rms_mae_db":2.953177467095176,"rms_envelope_nrmse":0.20337191299123394,"envelope_correlation":0.9614100025249681,"rms_gain_error_db":0.03899837361872277,"reference_mean_power":0.002781784143506645,"output_mean_power":0.00280687624233843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.487428460245723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008814675226478596,"output_power":0.0009562309357268064,"reference_active":true,"delta_db":0.35356474258697357},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001997406840389717,"output_power":0.00020402487586217873,"reference_active":true,"delta_db":0.09216589440929317},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0345220429658976e-05,"output_power":3.70613931351411e-05,"reference_active":true,"delta_db":-0.36870352226622555}],"active_band_count":3,"active_band_mae_db":0.27147805308749745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013989081102856241,"log_rms_mae_db":1.9765721212696024,"rms_envelope_nrmse":0.12159708189183378,"envelope_correlation":0.9848161073397079,"rms_gain_error_db":0.039286364505129255,"reference_mean_power":0.002795409730019158,"output_mean_power":0.0028208117820917263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3091173503801554e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008259246223406256,"output_power":0.0008912117470654308,"reference_active":true,"delta_db":0.3304048887683645},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.58341852709606e-05,"output_power":9.602923694066647e-05,"reference_active":true,"delta_db":0.008830228507933649},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329771424834818e-06,"output_power":2.96270223487418e-06,"reference_active":true,"delta_db":-0.5072641690130927}],"active_band_count":3,"active_band_mae_db":0.28216642876313025}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000132208474335363,"log_rms_mae_db":2.8602310967455273,"rms_envelope_nrmse":0.22243632838062513,"envelope_correlation":0.9623293556568552,"rms_gain_error_db":-0.3135792721828615,"reference_mean_power":0.001270855716483403,"output_mean_power":0.0011823289388419723},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.478174051796597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005667765898001897,"output_power":0.0005547387731920475,"reference_active":true,"delta_db":-0.09323382063355037},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011084740591226797,"output_power":0.00010227938904394818,"reference_active":true,"delta_db":-0.3493740892011888},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2190579526022388e-05,"output_power":1.9474081564889683e-05,"reference_active":true,"delta_db":-0.567116596878874}],"active_band_count":3,"active_band_mae_db":0.3365748355712044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012422751968092996,"log_rms_mae_db":1.8441946031979564,"rms_envelope_nrmse":0.13126022918163932,"envelope_correlation":0.9863873140768742,"rms_gain_error_db":-0.3135781142877175,"reference_mean_power":0.0012772403358267607,"output_mean_power":0.0011882691275949243},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.710312666254066e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574547781532275,"output_power":0.0005444512243195302,"reference_active":true,"delta_db":-0.10250663560310182},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253668134122962e-05,"output_power":4.686066899690324e-05,"reference_active":true,"delta_db":-0.49654151200501434},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7043530583570791e-06,"output_power":1.5416273604202363e-06,"reference_active":true,"delta_db":-0.435801547984754}],"active_band_count":3,"active_band_mae_db":0.34494989853095676}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":2.7964812563178447,"rms_envelope_nrmse":0.28220654202234885,"envelope_correlation":0.9333776960925749,"rms_gain_error_db":-0.31285504405477704,"reference_mean_power":0.00020286667816607566,"output_mean_power":0.0001887666260736874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.368709420467863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.826983537989379e-05,"output_power":7.520412167201107e-05,"reference_active":true,"delta_db":-0.17352777007159356},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2413787132715606e-05,"output_power":1.8333228380496825e-05,"reference_active":true,"delta_db":-0.8727629449191288},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388555256663481e-06,"output_power":4.659987608640115e-06,"reference_active":true,"delta_db":-1.3701789363763914}],"active_band_count":3,"active_band_mae_db":0.805489883789038}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":1.7033908403292055,"rms_envelope_nrmse":0.14267486522406994,"envelope_correlation":0.9816559234603309,"rms_gain_error_db":-0.3126838620186395,"reference_mean_power":0.00020386380860129344,"output_mean_power":0.00018970192908994782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.180981609421061e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.558718162537126e-05,"output_power":7.155563613433331e-05,"reference_active":true,"delta_db":-0.238043053817031},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0576597086854286e-05,"output_power":8.54659982981355e-06,"reference_active":true,"delta_db":-0.9255259048195285},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.384274936793116e-07,"output_power":3.3970483191434206e-07,"reference_active":true,"delta_db":-2.0002550386759044}],"active_band_count":3,"active_band_mae_db":1.0546079991041546}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":2.7499649018964707,"rms_envelope_nrmse":0.25295923083733535,"envelope_correlation":0.9562936863716207,"rms_gain_error_db":-0.5146413196745578,"reference_mean_power":5.251733918815235e-05,"output_mean_power":4.664859655284747e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5789636326786885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2240876569039163e-05,"output_power":1.880633816872979e-05,"reference_active":true,"delta_db":-0.7284765872012948},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.963351102568177e-06,"output_power":6.957320899367907e-06,"reference_active":true,"delta_db":-0.5865382889382522},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4165297947437375e-06,"output_power":1.3751384798489035e-06,"reference_active":true,"delta_db":-0.1287927883799784}],"active_band_count":3,"active_band_mae_db":0.48126922150650847}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":1.5500984296394664,"rms_envelope_nrmse":0.14295007138545962,"envelope_correlation":0.9866520219485911,"rms_gain_error_db":-0.5057929963403469,"reference_mean_power":5.256947927681557e-05,"output_mean_power":4.6790143328424925e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.1875752001118554e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1809094457380056e-05,"output_power":1.8395743189620714e-05,"reference_active":true,"delta_db":-0.7392029541199707},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.006377197072698e-06,"output_power":3.5825694791462938e-06,"reference_active":true,"delta_db":-0.4855721363383565},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0500724932970186e-07,"output_power":9.762737089786082e-08,"reference_active":true,"delta_db":-0.3164768833115856}],"active_band_count":3,"active_band_mae_db":0.5137506579233043}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.3792190551757812,"predicted_mos":3.3792190551757812,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"ff2cc954f5e89b0422ae083078155b533320eeb6523f8db39bcb7f689acd2a1e","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5112208479986293},"started_at":"2026-10-13T05:10:23.699137+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031589249992975965,"native_input_samples":144000,"native_input_sha256":"9284a2a6222858e45935a7cb320bddcd7e1104acbdabc82b3c5c31c222e7b875","output_sha256":"308d700983f1602ff0592471849bcb0b4a00d1f557f674452116522aeef270e5","output_samples":144240,"length_delta_samples":240,"analysis_s":0.6980489420002414,"synthesis_s":0.11066887400011183,"total_s":0.8087178160003532,"rtf":0.13478630266672553,"output_peak":0.3717769384384155,"output_rms":0.051287174776372464,"finished_at":"2026-10-13T05:10:24.522677+00:00","analysis_stages":{"external_f0_s":0.15062496699829353},"evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"308d700983f1602ff0592471849bcb0b4a00d1f557f674452116522aeef270e5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.797006036814196,"output_peak":0.37209714833386204},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.062547017859251,"spectral_convergence":0.22984472923155966,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.4935999914557145,"spectral_convergence":0.2158879341732847,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.9066871218478285,"spectral_convergence":0.25173899641831365,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.000816215890019172,"delta_db":-0.14404517589396448,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.0011103581360423125,"delta_db":-0.24249431928283244,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0004599886847852506,"delta_db":-0.14920772839810698,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00016080821267582053,"delta_db":-0.5690723224135937,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":5.7750313616002614e-05,"delta_db":-0.5315198566105395,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":3.134391038647632e-05,"delta_db":-0.7152966857290572,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3919393480546824,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.22940860301801053,"pitch":null,"output_rms_dbfs":-25.789779654050008,"output_peak":0.37209714833386204,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.8209447103875975},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9232039791203897,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05272073392617238,"rms_dbfs":-25.560371051032,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.410487802791902,"mean":-3.097852071126302e-05,"zero_energy":false},"output":{"rms":0.05134652030590812,"rms_dbfs":-25.789779654050008,"peak":0.37209714833386204,"peak_dbfs":-8.586873165955428,"crest_factor_db":17.20290648809458,"mean":-3.4583914239598676e-05,"zero_energy":false},"rms_difference_db":-0.22940860301800825},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":2.091961780828255e-06,"output_rms_dbfs":-56.794462540910985,"output_zero_energy":false,"output_peak":0.030121456803809094,"output_energy_fraction":9.389420943558782e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":91,"both_estimated_voiced_frames":81,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9642857142857143,"absolute_difference_cents_median":1.7273463022777202,"absolute_difference_cents_p95":8.285010409020215,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001544797211376954,"log_rms_mae_db":1.110575780769693,"rms_envelope_nrmse":0.08033154564055493,"envelope_correlation":0.9943206363298263,"rms_gain_error_db":-0.2294099547462397,"reference_mean_power":0.002781784143506645,"output_mean_power":0.0026386539135471545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.487428460245723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008814675226478596,"output_power":0.0008724969096729569,"reference_active":true,"delta_db":-0.044424181157039985},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001997406840389717,"output_power":0.00017835684575983,"reference_active":true,"delta_db":-0.49176749593867924},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0345220429658976e-05,"output_power":3.2956463972169665e-05,"reference_active":true,"delta_db":-0.878514842428456}],"active_band_count":3,"active_band_mae_db":0.47156883984139175}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013989081102856241,"log_rms_mae_db":0.628455292064168,"rms_envelope_nrmse":0.04781237311681345,"envelope_correlation":0.998280918274173,"rms_gain_error_db":-0.2293577684739706,"reference_mean_power":0.002795409730019158,"output_mean_power":0.0026516102896535546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3091173503801554e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008259246223406256,"output_power":0.0008174103138494626,"reference_active":true,"delta_db":-0.045003002141052705},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.58341852709606e-05,"output_power":8.524983545469003e-05,"reference_active":true,"delta_db":-0.5082690566281759},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329771424834818e-06,"output_power":2.600345196569945e-06,"reference_active":true,"delta_db":-1.073834174780083}],"active_band_count":3,"active_band_mae_db":0.5423687445164372}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000132208474335363,"log_rms_mae_db":1.367338094543765,"rms_envelope_nrmse":0.10165580557183553,"envelope_correlation":0.9925308157815638,"rms_gain_error_db":-0.24693806014419142,"reference_mean_power":0.001270855716483403,"output_mean_power":0.0012006113392729236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.478174051796597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005667765898001897,"output_power":0.0005456556450482008,"reference_active":true,"delta_db":-0.1649325127073212},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011084740591226797,"output_power":9.489679062362692e-05,"reference_active":true,"delta_db":-0.6747400907833111},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2190579526022388e-05,"output_power":1.914565765887709e-05,"reference_active":true,"delta_db":-0.6409835527212933}],"active_band_count":3,"active_band_mae_db":0.49355205207064184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012422751968092996,"log_rms_mae_db":0.7502391884373883,"rms_envelope_nrmse":0.061755218199730316,"envelope_correlation":0.997401978269104,"rms_gain_error_db":-0.2469366833088835,"reference_mean_power":0.0012772403358267607,"output_mean_power":0.0012066434422362867},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.710312666254066e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574547781532275,"output_power":0.0005338982717452144,"reference_active":true,"delta_db":-0.18751127284224287},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253668134122962e-05,"output_power":4.4425609840396466e-05,"reference_active":true,"delta_db":-0.7282923744654817},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7043530583570791e-06,"output_power":1.46604634472598e-06,"reference_active":true,"delta_db":-0.6541186480538199}],"active_band_count":3,"active_band_mae_db":0.5233074317871815}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":1.3399887800464207,"rms_envelope_nrmse":0.15403190861309027,"envelope_correlation":0.9827080177677964,"rms_gain_error_db":-0.5765937722915245,"reference_mean_power":0.00020286667816607566,"output_mean_power":0.00017764432604432128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.368709420467863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.826983537989379e-05,"output_power":7.23573050101577e-05,"reference_active":true,"delta_db":-0.3411203745727415},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2413787132715606e-05,"output_power":1.7640359874470034e-05,"reference_active":true,"delta_db":-1.0400780228609654},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388555256663481e-06,"output_power":3.173939982691664e-06,"reference_active":true,"delta_db":-3.038039452505918}],"active_band_count":3,"active_band_mae_db":1.4730792833132085}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":0.9137878698347965,"rms_envelope_nrmse":0.09911799513858999,"envelope_correlation":0.9943321534895142,"rms_gain_error_db":-0.5765748965108316,"reference_mean_power":0.00020386380860129344,"output_mean_power":0.00017851825945099995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.180981609421061e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.558718162537126e-05,"output_power":6.945156707770501e-05,"reference_active":true,"delta_db":-0.36766102928135264},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0576597086854286e-05,"output_power":8.066063149416405e-06,"reference_active":true,"delta_db":-1.1768434252778095},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.384274936793116e-07,"output_power":2.661222187069796e-07,"reference_active":true,"delta_db":-3.0604609254859714}],"active_band_count":3,"active_band_mae_db":1.5349884600150443}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":1.4240714485637755,"rms_envelope_nrmse":0.12910525177770157,"envelope_correlation":0.9901608298885324,"rms_gain_error_db":-0.4823047967673535,"reference_mean_power":5.251733918815235e-05,"output_mean_power":4.699722708552518e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5789636326786885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2240876569039163e-05,"output_power":2.0181013909701728e-05,"reference_active":true,"delta_db":-0.4220891814949412},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.963351102568177e-06,"output_power":6.281276036351245e-06,"reference_active":true,"delta_db":-1.03047984818926},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4165297947437375e-06,"output_power":9.595803105156311e-07,"reference_active":true,"delta_db":-1.691443855719594}],"active_band_count":3,"active_band_mae_db":1.0480042951345985}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":0.8990588191522398,"rms_envelope_nrmse":0.0772219953857273,"envelope_correlation":0.9976659817480749,"rms_gain_error_db":-0.4839434859151651,"reference_mean_power":5.256947927681557e-05,"output_mean_power":4.702613936435151e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.1875752001118554e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1809094457380056e-05,"output_power":1.953550062783167e-05,"reference_active":true,"delta_db":-0.4781308828791383},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.006377197072698e-06,"output_power":3.0917078607622328e-06,"reference_active":true,"delta_db":-1.1255338517491666},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0500724932970186e-07,"output_power":8.179180432287062e-08,"reference_active":true,"delta_db":-1.0850949345260734}],"active_band_count":3,"active_band_mae_db":0.8962532230514594}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123288-0003","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1415107250213623,"predicted_mos":3.1415107250213623,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":290055,"input_sample_rate":48000,"input_seconds":6.0428125,"input_float32_sha256":"a26d242d551107aa883e12035f221c0143f1b6b702bd106f3fad1292544ae4dd","internal_samples":96685,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6183784129989363},"started_at":"2026-10-13T05:14:00.070121+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009521402000245871,"native_input_samples":288000,"native_input_sha256":"7ba2e10d9cc0800a9a2938a77799cd1b7b0d27c510be6641dfce87e845ba2ae5","output_sha256":"9f7c285dd4b52dfbb15676f1053cf943bec1bde6616a350b46099f44e5f5bfb0","output_samples":290055,"length_delta_samples":2055,"analysis_s":3.020072285000424,"synthesis_s":0.8984387839991541,"total_s":3.918511068999578,"rtf":0.6530851781665964,"output_peak":0.4067431353526975,"output_rms":0.05042121881060466,"finished_at":"2026-10-13T05:14:04.009225+00:00","evaluation_input":{"reference":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"9f7c285dd4b52dfbb15676f1053cf943bec1bde6616a350b46099f44e5f5bfb0","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96685,"length_delta_samples":685,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.945100031713935,"output_peak":0.4040469188946166},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9929151367844029,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.08487559964894,"spectral_convergence":0.8434296562398377,"spectral_convergence_status":"ok","amplitude_floor":1.3116896377207775e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.542895446787275,"spectral_convergence":0.7761106739862246,"spectral_convergence_status":"ok","amplitude_floor":8.457684224428572e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.161908695867357,"spectral_convergence":0.7358251745764006,"spectral_convergence_status":"ok","amplitude_floor":6.717993311155654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008437417957183396,"output_power":0.0007551781914740057,"delta_db":-0.4816012328257302,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011741200681074208,"output_power":0.0010826671475120129,"delta_db":-0.35217552361157,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004760668283187788,"output_power":0.0004430190056907182,"delta_db":-0.31245563459184444,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001833223674641814,"output_power":0.00017925881294993758,"delta_db":-0.0973494097776796,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.526890589901711e-05,"output_power":6.442163403488891e-05,"delta_db":-0.05674596320776971,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6955820206525705e-05,"output_power":3.5874453716013716e-05,"delta_db":-0.12897549055967686,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23821720909571184,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3565181198862622,"pitch":null,"output_rms_dbfs":-25.91688917091826,"output_peak":0.4040469188946166,"reference_rms_dbfs":-25.560371051032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.263226580767858},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.36145219357392677,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05272073392617238,"rms_dbfs":-25.560371051032,"peak":0.39129638671875,"peak_dbfs":-8.149883248240098,"crest_factor_db":17.410487802791902,"mean":-3.097852071126302e-05,"zero_energy":false},"output":{"rms":0.0506005854252456,"rms_dbfs":-25.91688917091826,"peak":0.4040469188946166,"peak_dbfs":-7.871364012941741,"crest_factor_db":18.04552515797652,"mean":-0.0006639653058921838,"zero_energy":false},"rms_difference_db":-0.35651811988626037},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3863984242780135e-06,"threshold_dbfs":-56.22257046407471,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":71,"quiet_samples":11360,"reference_mean_power":4.726274325971452e-07,"output_mean_power":1.9523167467201693e-06,"output_rms_dbfs":-57.09449720381904,"output_zero_energy":false,"output_peak":0.035476655160448636,"output_energy_fraction":9.022903135061317e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":84,"output_estimated_voiced_frames":80,"both_estimated_voiced_frames":41,"reference_voiced_output_unvoiced_frames":43,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.4880952380952381,"absolute_difference_cents_median":31.875576632854212,"absolute_difference_cents_p95":73.37318705752705,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001544797211376954,"log_rms_mae_db":6.39672952096197,"rms_envelope_nrmse":0.5765938577337404,"envelope_correlation":0.6733941613109327,"rms_gain_error_db":-0.357227134709575,"reference_mean_power":0.002781784143506645,"output_mean_power":0.0025621273578711264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.487428460245723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008814675226478596,"output_power":0.0008259976180758489,"reference_active":true,"delta_db":-0.2822752055705444},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001997406840389717,"output_power":0.00019135842410594568,"reference_active":true,"delta_db":-0.18618947065373356},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0345220429658976e-05,"output_power":4.041393908079819e-05,"reference_active":true,"delta_db":0.007390898661486733}],"active_band_count":3,"active_band_mae_db":0.15861852496192158}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013989081102856241,"log_rms_mae_db":4.853844518915202,"rms_envelope_nrmse":0.45470795691578275,"envelope_correlation":0.7726885092021342,"rms_gain_error_db":-0.3598816069622627,"reference_mean_power":0.002795409730019158,"output_mean_power":0.002573103833309546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3091173503801554e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008259246223406256,"output_power":0.0007734687887724332,"reference_active":true,"delta_db":-0.28497619810459884},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.58341852709606e-05,"output_power":9.313136305554312e-05,"reference_active":true,"delta_db":-0.12424495807813865},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329771424834818e-06,"output_power":3.2223870623156423e-06,"reference_active":true,"delta_db":-0.14236716781193753}],"active_band_count":3,"active_band_mae_db":0.18386277466489168}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000132208474335363,"log_rms_mae_db":7.708292644906003,"rms_envelope_nrmse":0.6400525473730483,"envelope_correlation":0.67870204745471,"rms_gain_error_db":-0.38452081631914264,"reference_mean_power":0.001270855716483403,"output_mean_power":0.001163172605012503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.478174051796597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005667765898001897,"output_power":0.0005239611527792048,"reference_active":true,"delta_db":-0.34112814830669497},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011084740591226797,"output_power":9.932196138186577e-05,"reference_active":true,"delta_db":-0.4768024677481339},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2190579526022388e-05,"output_power":2.2685508742654754e-05,"reference_active":true,"delta_db":0.09579878797099822}],"active_band_count":3,"active_band_mae_db":0.304576468008609}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012422751968092996,"log_rms_mae_db":6.2663123408122825,"rms_envelope_nrmse":0.5078323041048574,"envelope_correlation":0.7788253629507705,"rms_gain_error_db":-0.38458838436277665,"reference_mean_power":0.0012772403358267607,"output_mean_power":0.0011689980503921875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.710312666254066e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574547781532275,"output_power":0.0005139784591268329,"reference_active":true,"delta_db":-0.3526472424739088},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.253668134122962e-05,"output_power":4.7545194061271995e-05,"reference_active":true,"delta_db":-0.4335601126174453},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7043530583570791e-06,"output_power":1.7503070953954697e-06,"reference_active":true,"delta_db":0.11554689069768206}],"active_band_count":3,"active_band_mae_db":0.3005847485963454}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.199705053096066e-05,"log_rms_mae_db":7.839694216607943,"rms_envelope_nrmse":0.6968236830341709,"envelope_correlation":0.6003402356819093,"rms_gain_error_db":-0.10235717737106077,"reference_mean_power":0.00020286667816607566,"output_mean_power":0.00019814129640590453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.368709420467863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.826983537989379e-05,"output_power":7.865742845581292e-05,"reference_active":true,"delta_db":0.02145323537446256},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2413787132715606e-05,"output_power":2.2390892366131578e-05,"reference_active":true,"delta_db":-0.004438407670153463},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388555256663481e-06,"output_power":6.061804831429383e-06,"reference_active":true,"delta_db":-0.22800705979990488}],"active_band_count":3,"active_band_mae_db":0.08463290094817362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3135372863327127e-05,"log_rms_mae_db":6.07789131389662,"rms_envelope_nrmse":0.5343074677670825,"envelope_correlation":0.7337912257878098,"rms_gain_error_db":-0.10545077812970621,"reference_mean_power":0.00020386380860129344,"output_mean_power":0.000198973415833554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.180981609421061e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.558718162537126e-05,"output_power":7.652513835111251e-05,"reference_active":true,"delta_db":0.053559710772026},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0576597086854286e-05,"output_power":1.0506987979160001e-05,"reference_active":true,"delta_db":-0.028677249254229792},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.384274936793116e-07,"output_power":4.992364707026081e-07,"reference_active":true,"delta_db":-0.32820923794645634}],"active_band_count":3,"active_band_mae_db":0.1368153993242374}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055375191564044e-05,"log_rms_mae_db":7.474853120071472,"rms_envelope_nrmse":0.777848493658192,"envelope_correlation":0.5939833769699442,"rms_gain_error_db":0.07494630518546741,"reference_mean_power":5.251733918815235e-05,"output_mean_power":5.343149733585816e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5789636326786885e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2240876569039163e-05,"output_power":2.3867362549662784e-05,"reference_active":true,"delta_db":0.3065253028684449},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.963351102568177e-06,"output_power":8.183709682740393e-06,"reference_active":true,"delta_db":0.11854350266393598},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4165297947437375e-06,"output_power":2.0825443164187263e-06,"reference_active":true,"delta_db":1.6736853820192237}],"active_band_count":3,"active_band_mae_db":0.6995847291838682}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.394748258818145e-05,"log_rms_mae_db":5.692870019812783,"rms_envelope_nrmse":0.5961334830731806,"envelope_correlation":0.7403262811053749,"rms_gain_error_db":-0.012306124528377822,"reference_mean_power":5.256947927681557e-05,"output_mean_power":5.2420729778923825e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.1875752001118554e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1809094457380056e-05,"output_power":2.3760543356363266e-05,"reference_active":true,"delta_db":0.37218734405320564},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.006377197072698e-06,"output_power":4.01126221505131e-06,"reference_active":true,"delta_db":0.005292172684705712},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0500724932970186e-07,"output_power":1.470141331971512e-07,"reference_active":true,"delta_db":1.4613980537344426}],"active_band_count":3,"active_band_mae_db":0.6129591901574513}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"hnm","status":"measured","mos":{"status":"measured","value":1.7928898334503174,"predicted_mos":1.7928898334503174,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":79920,"input_sample_rate":16000,"input_seconds":4.995,"input_float32_sha256":"8fdde44507a3b55bc3ce5d618ebf2a267e8222801a715f4dc0ca0b5ae9582f7d","internal_samples":79920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.560324418000164},"started_at":"2026-10-13T04:38:44.709411+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4684000234410632e-05,"native_input_samples":79920,"native_input_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","output_sha256":"4a86961b98cc1ecb66510c641115d46a3e0c151def510adce6920ca5a8fd77d0","output_samples":79920,"length_delta_samples":0,"analysis_s":0.8740831669992986,"synthesis_s":0.4284241010000187,"total_s":1.3025072679993173,"rtf":0.2607622158156791,"output_peak":0.45781398891155095,"output_rms":0.07302432409615485,"finished_at":"2026-10-13T04:38:46.016590+00:00","evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"4a86961b98cc1ecb66510c641115d46a3e0c151def510adce6920ca5a8fd77d0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.276484627930516,"spectral_convergence":0.26140565214256795,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.975607657013558,"spectral_convergence":0.252482082617658,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.116749602976656,"spectral_convergence":0.2589343961885255,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.001051946943011967,"delta_db":0.07052229598981329,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.003184726420505586,"delta_db":-0.10084840134927434,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0008066277195346742,"delta_db":0.05673151934393872,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.00022136082620209457,"delta_db":-0.2811312331169056,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":5.40313177954315e-05,"delta_db":0.012705087965738384,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.3858682650508416e-05,"delta_db":0.0965093731419413,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.10307465181793528,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04999919395102753,"pitch":null,"output_rms_dbfs":-22.73064908220465,"output_peak":0.45781398891155095,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.122947295973577},"common_prefix":{"samples":79920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.276484627930516,"spectral_convergence":0.26140565214256795,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.975607657013558,"spectral_convergence":0.252482082617658,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.116749602976656,"spectral_convergence":0.2589343961885255,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.001051946943011967,"delta_db":0.07052229598981329,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.003184726420505586,"delta_db":-0.10084840134927434,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0008066277195346742,"delta_db":0.05673151934393872,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.00022136082620209457,"delta_db":-0.2811312331169056,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":5.40313177954315e-05,"delta_db":0.012705087965738384,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.3858682650508416e-05,"delta_db":0.0965093731419413,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.10307465181793528,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04999919395102753,"pitch":null,"output_rms_dbfs":-22.73064908220465,"output_peak":0.45781398891155095,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.122947295973577},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79920,"output_samples":79920,"intelligibility":{"estoi":{"value":0.8571012707021751,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07344589130405568,"rms_dbfs":-22.68064988825362,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.093818785323364,"mean":-6.284627828512106e-05,"zero_energy":false},"output":{"rms":0.07302432409615485,"rms_dbfs":-22.73064908220465,"peak":0.45781398891155095,"peak_dbfs":-6.786218824130554,"crest_factor_db":15.944430258074092,"mean":0.0001183780517562727,"zero_energy":false},"rms_difference_db":-0.04999919395103092},"full_scale":{"reference":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":2.3147989534605955e-06,"output_rms_dbfs":-56.354867226565084,"output_zero_energy":false,"output_peak":0.02331174134086693,"output_energy_fraction":8.429744803255212e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":134,"both_estimated_voiced_frames":122,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9242424242424242,"absolute_difference_cents_median":1.6328627085625032,"absolute_difference_cents_p95":15.919090952808755,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001944445221026879,"log_rms_mae_db":1.276266721539359,"rms_envelope_nrmse":0.08477368463722834,"envelope_correlation":0.9929227142728585,"rms_gain_error_db":-0.04999852666440443,"reference_mean_power":0.005399702914122853,"output_mean_power":0.005337894836854387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":2.8891675057826967e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000788720956533661,"output_power":0.0007869602738346026,"reference_active":true,"delta_db":-0.009705708034054565},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016463444647070574,"output_power":0.00015155603252589705,"reference_active":true,"delta_db":-0.3594748015717485},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1066196116820634e-05,"output_power":5.44046842193018e-05,"reference_active":true,"delta_db":0.2750278508563764}],"active_band_count":3,"active_band_mae_db":0.21473612015405982}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00017595749087279735,"log_rms_mae_db":0.6150083139100644,"rms_envelope_nrmse":0.040676300205084504,"envelope_correlation":0.9982525273433882,"rms_gain_error_db":-0.049999124264065054,"reference_mean_power":0.005432355653378944,"output_mean_power":0.005370173075296808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":2.667750009815418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007158752148089716,"output_power":0.0007075551650124646,"reference_active":true,"delta_db":-0.050770208613097764},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.294187442575086e-05,"output_power":6.815126938538653e-05,"reference_active":true,"delta_db":-0.295029700010706},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.759052873918178e-06,"output_power":3.88874744687809e-06,"reference_active":true,"delta_db":0.1473130454802646}],"active_band_count":3,"active_band_mae_db":0.16437098470135614}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001659932762873133,"log_rms_mae_db":1.943246282243205,"rms_envelope_nrmse":0.13359282715356605,"envelope_correlation":0.9858198065573391,"rms_gain_error_db":-0.1621180108831134,"reference_mean_power":0.0017617711849040457,"output_mean_power":0.0016972182826947005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.325463642910379e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157322295664234,"output_power":0.000589735985707058,"reference_active":true,"delta_db":-0.18734257421783332},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015692982702212585,"output_power":0.00013177157577359523,"reference_active":true,"delta_db":-0.7588375661732949},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0206834209715196e-05,"output_power":3.5845302539651066e-05,"reference_active":true,"delta_db":-0.4986762888965406}],"active_band_count":3,"active_band_mae_db":0.4816188097625562}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00014319111528307625,"log_rms_mae_db":1.1798379993690005,"rms_envelope_nrmse":0.07887870351296114,"envelope_correlation":0.9947825311623155,"rms_gain_error_db":-0.1621183523626267,"reference_mean_power":0.0017724260962696937,"output_mean_power":0.0017074826541381987},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.2006657592514347e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005739716253322745,"output_power":0.00054992247461393,"reference_active":true,"delta_db":-0.18588954199743563},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.210701544963181e-05,"output_power":6.37308993677294e-05,"reference_active":true,"delta_db":-0.5362747296910119},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.009713658025179e-06,"output_power":2.8141193464749805e-06,"reference_active":true,"delta_db":-1.5377084798925638}],"active_band_count":3,"active_band_mae_db":0.7532909171936705}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001241617701093293,"log_rms_mae_db":2.0375520090127353,"rms_envelope_nrmse":0.17290620428956419,"envelope_correlation":0.9791271524934374,"rms_gain_error_db":-0.2171105022606046,"reference_mean_power":0.00024591411290482445,"output_mean_power":0.00023392271893165608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.707803249714263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.578711086190903e-05,"output_power":8.691556127790202e-05,"reference_active":true,"delta_db":-0.42209535160089356},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5502167493014163e-05,"output_power":1.5654612485837624e-05,"reference_active":true,"delta_db":0.04249896654916133},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901592249182956e-06,"output_power":5.445346437134261e-06,"reference_active":true,"delta_db":-0.34943684952736953}],"active_band_count":3,"active_band_mae_db":0.27134372255914146}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":1.1892710024382964,"rms_envelope_nrmse":0.0994358085651175,"envelope_correlation":0.9930018613583267,"rms_gain_error_db":-0.21711847560703773,"reference_mean_power":0.0002474004325310524,"output_mean_power":0.00023533612979328327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.5717362127850396e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.0470755798443e-05,"output_power":8.142145108783899e-05,"reference_active":true,"delta_db":-0.45769380375011937},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.549692811018895e-06,"output_power":8.103287658735679e-06,"reference_active":true,"delta_db":0.3073197519985902},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.070423026829598e-07,"output_power":3.403583548397231e-07,"reference_active":true,"delta_db":-0.7770313071204002}],"active_band_count":3,"active_band_mae_db":0.5140149542897032}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":2.3331248310309234,"rms_envelope_nrmse":0.30889941674178145,"envelope_correlation":0.9359851885936536,"rms_gain_error_db":-0.08395929413084173,"reference_mean_power":2.0515613806993188e-05,"output_mean_power":2.012280810729987e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":4.904304308942734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6213437163626398e-06,"output_power":2.0686277678653916e-06,"reference_active":true,"delta_db":1.0580725730283107},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.453801656353397e-07,"output_power":9.153574855931786e-07,"reference_active":true,"delta_db":0.8921290534149375},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4092114970939407e-07,"output_power":1.2397498528161253e-06,"reference_active":true,"delta_db":7.114591383859929}],"active_band_count":3,"active_band_mae_db":3.021597670101059}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":1.6123375849295833,"rms_envelope_nrmse":0.23174306836322814,"envelope_correlation":0.9615982234328386,"rms_gain_error_db":-0.0839602396493242,"reference_mean_power":2.063955290066711e-05,"output_mean_power":2.0244369772496264e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":4.623298998188315e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6281370009575067e-06,"output_power":2.607063091236952e-06,"reference_active":true,"delta_db":2.0446059510834385},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7397133003103594e-07,"output_power":6.740082126738231e-07,"reference_active":true,"delta_db":2.5582687945006914},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7174687173869285e-08,"output_power":1.1637457406259912e-07,"reference_active":true,"delta_db":8.30969269047755}],"active_band_count":3,"active_band_mae_db":4.304189145353893}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.3918466567993164,"predicted_mos":2.3918466567993164,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":79920,"input_sample_rate":16000,"input_seconds":4.995,"input_float32_sha256":"8e55c542a7cf570192c41fc3ba2159f76cce5af8683b273bfae2c9eb94af55e2","internal_samples":79920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1944870839997748},"started_at":"2026-10-13T04:41:41.292641+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.701799985516118e-05,"native_input_samples":79920,"native_input_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","output_sha256":"f920adf3484e0cfb0c46c251422bf50f83c6f27d999d6dce5058c50eb52828b1","output_samples":79920,"length_delta_samples":0,"analysis_s":1.3133945680001489,"synthesis_s":0.0801190249994761,"total_s":1.393513592999625,"rtf":0.27898170030022523,"output_peak":0.6142908173569714,"output_rms":0.08383700015634818,"finished_at":"2026-10-13T04:41:42.692257+00:00","evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"f920adf3484e0cfb0c46c251422bf50f83c6f27d999d6dce5058c50eb52828b1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.458554644365382,"spectral_convergence":0.5084695658153979,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.482210914315445,"spectral_convergence":0.27485228250812344,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.782889933097548,"spectral_convergence":0.3329065418805601,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.0012040338914252144,"delta_db":0.6569710543058072,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.0043817624112082,"delta_db":1.28491854516328,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0010534509625114664,"delta_db":1.2161433387466891,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.0003110790591236667,"delta_db":1.1965688633825402,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":6.468776550040589e-05,"delta_db":0.7944709854583544,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.3628505446569202e-05,"delta_db":0.023772193985742863,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8621408301737358,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1493644893355628,"pitch":null,"output_rms_dbfs":-21.531285398918058,"output_peak":0.6142908173569714,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9078851639261245},"common_prefix":{"samples":79920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.458554644365382,"spectral_convergence":0.5084695658153979,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.482210914315445,"spectral_convergence":0.27485228250812344,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.782889933097548,"spectral_convergence":0.3329065418805601,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.0012040338914252144,"delta_db":0.6569710543058072,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.0043817624112082,"delta_db":1.28491854516328,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0010534509625114664,"delta_db":1.2161433387466891,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.0003110790591236667,"delta_db":1.1965688633825402,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":6.468776550040589e-05,"delta_db":0.7944709854583544,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.3628505446569202e-05,"delta_db":0.023772193985742863,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8621408301737358,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1493644893355628,"pitch":null,"output_rms_dbfs":-21.531285398918058,"output_peak":0.6142908173569714,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9078851639261245},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79920,"output_samples":79920,"intelligibility":{"estoi":{"value":0.9130540932550363,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07344589130405568,"rms_dbfs":-22.68064988825362,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.093818785323364,"mean":-6.284627828512106e-05,"zero_energy":false},"output":{"rms":0.08383700015634817,"rms_dbfs":-21.531285398918058,"peak":0.6142908173569714,"peak_dbfs":-4.232519532482354,"crest_factor_db":17.2987658664357,"mean":1.1739183500472008e-07,"zero_energy":false},"rms_difference_db":1.1493644893355608},"full_scale":{"reference":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":2.084234389916302e-06,"output_rms_dbfs":-56.81053442511076,"output_zero_energy":false,"output_peak":0.01416510578530954,"output_energy_fraction":5.75852609346134e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":153,"both_estimated_voiced_frames":123,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9318181818181818,"absolute_difference_cents_median":7.027747737266533,"absolute_difference_cents_p95":36.44608614959856,"difference_over_600_cents_frames":5,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001944445221026879,"log_rms_mae_db":1.968802534821173,"rms_envelope_nrmse":0.18786330255257339,"envelope_correlation":0.9870282461720531,"rms_gain_error_db":1.1493652721122556,"reference_mean_power":0.005399702914122853,"output_mean_power":0.007035685100139764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":2.8891675057826967e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000788720956533661,"output_power":0.0009925732582296443,"reference_active":true,"delta_db":0.9983919001386515},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016463444647070574,"output_power":0.00018588996139831462,"reference_active":true,"delta_db":0.527352294001573},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1066196116820634e-05,"output_power":5.145929717510775e-05,"reference_active":true,"delta_db":0.033303415874201554}],"active_band_count":3,"active_band_mae_db":0.5196825366714753}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00017595749087279735,"log_rms_mae_db":1.4026687620765887,"rms_envelope_nrmse":0.16184140507019354,"envelope_correlation":0.9944656510986082,"rms_gain_error_db":1.149368086971562,"reference_mean_power":0.005432355653378944,"output_mean_power":0.007078235434308404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":2.667750009815418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007158752148089716,"output_power":0.0008969606559132991,"reference_active":true,"delta_db":0.9793606722548344},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.294187442575086e-05,"output_power":8.430058238990399e-05,"reference_active":true,"delta_db":0.6285365553996813},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.759052873918178e-06,"output_power":3.2700934507509942e-06,"reference_active":true,"delta_db":-0.6051827059430076}],"active_band_count":3,"active_band_mae_db":0.7376933111991745}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001659932762873133,"log_rms_mae_db":2.3324529912442804,"rms_envelope_nrmse":0.21248770601146166,"envelope_correlation":0.983096493699493,"rms_gain_error_db":1.1738970677370122,"reference_mean_power":0.0017617711849040457,"output_mean_power":0.0023085495963388147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.325463642910379e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157322295664234,"output_power":0.000790885420720426,"reference_active":true,"delta_db":1.0872168310663453},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015692982702212585,"output_power":0.00020030115158457655,"reference_active":true,"delta_db":1.0597795013438924},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0206834209715196e-05,"output_power":3.750322915685656e-05,"reference_active":true,"delta_db":-0.3023121555236919}],"active_band_count":3,"active_band_mae_db":0.8164361626446431}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00014319111528307625,"log_rms_mae_db":1.8007221512868277,"rms_envelope_nrmse":0.17432379150752164,"envelope_correlation":0.992732261611279,"rms_gain_error_db":1.1738970114284784,"reference_mean_power":0.0017724260962696937,"output_mean_power":0.0023225113062881595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.2006657592514347e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005739716253322745,"output_power":0.0007384916595088856,"reference_active":true,"delta_db":1.0945517148052228},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.210701544963181e-05,"output_power":9.379288285211642e-05,"reference_active":true,"delta_db":1.1419236446020182},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.009713658025179e-06,"output_power":3.794719737826046e-06,"reference_active":true,"delta_db":-0.23933653620441742}],"active_band_count":3,"active_band_mae_db":0.8252706318705529}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001241617701093293,"log_rms_mae_db":1.6080284307086086,"rms_envelope_nrmse":0.23188038717484186,"envelope_correlation":0.98131713450439,"rms_gain_error_db":1.199330555052219,"reference_mean_power":0.00024591411290482445,"output_mean_power":0.0003241279697577709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.707803249714263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.578711086190903e-05,"output_power":0.00011295014638729143,"reference_active":true,"delta_db":0.7157972388379262},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5502167493014163e-05,"output_power":1.7512236294169735e-05,"reference_active":true,"delta_db":0.5294918366501307},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901592249182956e-06,"output_power":5.623637106155943e-06,"reference_active":true,"delta_db":-0.20951912221704114}],"active_band_count":3,"active_band_mae_db":0.4849360659016993}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":0.9694987984554346,"rms_envelope_nrmse":0.19006005895083933,"envelope_correlation":0.991514116701459,"rms_gain_error_db":1.199338317037774,"reference_mean_power":0.0002474004325310524,"output_mean_power":0.0003260876014229138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.5717362127850396e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.0470755798443e-05,"output_power":0.00010587645381063554,"reference_active":true,"delta_db":0.6829116830802285},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.549692811018895e-06,"output_power":8.740319625025294e-06,"reference_active":true,"delta_db":0.6359803364593919},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.070423026829598e-07,"output_power":3.3664707317791134e-07,"reference_active":true,"delta_db":-0.8246470344266708}],"active_band_count":3,"active_band_mae_db":0.7145130179887639}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":1.626442645976585,"rms_envelope_nrmse":0.22067127972245526,"envelope_correlation":0.9741481292572118,"rms_gain_error_db":0.6531053322558112,"reference_mean_power":2.0515613806993188e-05,"output_mean_power":2.3844874905380942e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":4.904304308942734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6213437163626398e-06,"output_power":1.7914822176227732e-06,"reference_active":true,"delta_db":0.43337408925435583},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.453801656353397e-07,"output_power":7.990005404366461e-07,"reference_active":true,"delta_db":0.3016924095405683},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4092114970939407e-07,"output_power":1.7098265181578436e-07,"reference_active":true,"delta_db":-1.4892287863044016}],"active_band_count":3,"active_band_mae_db":0.7414317616997752}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":1.045115122083619,"rms_envelope_nrmse":0.16612541880956685,"envelope_correlation":0.9862646972963452,"rms_gain_error_db":0.6531118061995212,"reference_mean_power":2.063955290066711e-05,"output_mean_power":2.3988962518246297e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":4.623298998188315e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6281370009575067e-06,"output_power":1.787634049737465e-06,"reference_active":true,"delta_db":0.4058767207200549},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7397133003103594e-07,"output_power":3.819887392574964e-07,"reference_active":true,"delta_db":0.09212251502863786},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7174687173869285e-08,"output_power":1.1830170349029783e-08,"reference_active":true,"delta_db":-1.6189783708823837}],"active_band_count":3,"active_band_mae_db":0.7056592022103588}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"facodec","status":"measured","mos":{"status":"measured","value":4.234045505523682,"predicted_mos":4.234045505523682,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":80000,"input_sample_rate":16000,"input_seconds":5.0,"input_float32_sha256":"84c4f7d61ef8bd259ede0451be1106b0034f61836103e9b541e6b43af563710e","internal_samples":80000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.005207608999626},"started_at":"2026-10-13T04:49:24.264648+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.8634000045713037e-05,"native_input_samples":79920,"native_input_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","output_sha256":"ce6ae861460a63474431c5115684230301de4e9ec3c45c09499f5bec81d38f19","output_samples":80000,"length_delta_samples":80,"analysis_s":1.3334607990000222,"synthesis_s":9.922347689999697,"total_s":11.255808488999719,"rtf":2.253415112912857,"output_peak":0.42336124181747437,"output_rms":0.06914735984144833,"finished_at":"2026-10-13T04:49:35.527132+00:00","evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"ce6ae861460a63474431c5115684230301de4e9ec3c45c09499f5bec81d38f19","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":79920,"output_samples":80000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.204487946141775,"output_peak":0.42336124181747437},"common_prefix":{"samples":79920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.999,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.3701404142444575,"spectral_convergence":0.20417205585644047,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.687700706332974,"spectral_convergence":0.20268400650043528,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6554477769992735,"spectral_convergence":0.22430200112899054,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.0008481131911451476,"delta_db":-0.8648778821562425,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.002972595857740587,"delta_db":-0.40021102669828584,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0007149687484584531,"delta_db":-0.467129313536502,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.0001879158264589443,"delta_db":-0.9925053204169834,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":5.09109397875693e-05,"delta_db":-0.24563937140475695,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.163892547921532e-05,"delta_db":-0.6615812514119679,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6053240276041231,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5194929650918336,"pitch":null,"output_rms_dbfs":-23.200142853345454,"output_peak":0.42336124181747437,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.571096299192235},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79920,"output_samples":79920,"intelligibility":{"estoi":{"value":0.8989971414884474,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07344589130405568,"rms_dbfs":-22.68064988825362,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.093818785323364,"mean":-6.284627828512106e-05,"zero_energy":false},"output":{"rms":0.06918195927458196,"rms_dbfs":-23.200142853345454,"peak":0.42336124181747437,"peak_dbfs":-7.465778072070036,"crest_factor_db":15.734364781275417,"mean":3.33457854920979e-05,"zero_energy":false},"rms_difference_db":-0.5194929650918354},"full_scale":{"reference":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":1.2997866166672002e-06,"output_rms_dbfs":-58.86127939085756,"output_zero_energy":false,"output_peak":0.010560448281466961,"output_energy_fraction":5.273787031761882e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":130,"both_estimated_voiced_frames":127,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9621212121212122,"absolute_difference_cents_median":3.174693736153236,"absolute_difference_cents_p95":12.164010535157372,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001944445221026879,"log_rms_mae_db":1.3547397124951848,"rms_envelope_nrmse":0.10752299669967162,"envelope_correlation":0.9914477868839491,"rms_gain_error_db":-0.519492344531138,"reference_mean_power":0.005399702914122853,"output_mean_power":0.00479093889314156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":2.8891675057826967e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000788720956533661,"output_power":0.0006942209996995762,"reference_active":true,"delta_db":-0.554256336653222},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016463444647070574,"output_power":0.000150816712597723,"reference_active":true,"delta_db":-0.380712377391187},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1066196116820634e-05,"output_power":5.4257824567436594e-05,"reference_active":true,"delta_db":0.26328868231234515}],"active_band_count":3,"active_band_mae_db":0.39941913211891805}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00017595749087279735,"log_rms_mae_db":1.0767320075309326,"rms_envelope_nrmse":0.09626241098762711,"envelope_correlation":0.9932856407141858,"rms_gain_error_db":-0.5194914108106746,"reference_mean_power":0.005432355653378944,"output_mean_power":0.004819911389682648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":2.667750009815418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007158752148089716,"output_power":0.0006304763616893384,"reference_active":true,"delta_db":-0.5516851813043595},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.294187442575086e-05,"output_power":6.845077714919421e-05,"reference_active":true,"delta_db":-0.27598536197253337},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.759052873918178e-06,"output_power":3.6349734153166965e-06,"reference_active":true,"delta_db":-0.14577195479208852}],"active_band_count":3,"active_band_mae_db":0.3244808326896605}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001659932762873133,"log_rms_mae_db":1.7316312867090151,"rms_envelope_nrmse":0.11723113920018223,"envelope_correlation":0.9889950799023248,"rms_gain_error_db":-0.07332697016580055,"reference_mean_power":0.0017617711849040457,"output_mean_power":0.0017322748713151548},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.325463642910379e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157322295664234,"output_power":0.0005865566565982199,"reference_active":true,"delta_db":-0.21081918880838688},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015692982702212585,"output_power":0.00014783786875407633,"reference_active":true,"delta_db":-0.2591980295357777},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0206834209715196e-05,"output_power":4.008904352518701e-05,"reference_active":true,"delta_db":-0.012741844898775218}],"active_band_count":3,"active_band_mae_db":0.1609196877476466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00014319111528307625,"log_rms_mae_db":1.327285410783062,"rms_envelope_nrmse":0.09399980274998652,"envelope_correlation":0.9922154578119416,"rms_gain_error_db":-0.07332626757494935,"reference_mean_power":0.0017724260962696937,"output_mean_power":0.001742751675614855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.2006657592514347e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005739716253322745,"output_power":0.0005463899307970122,"reference_active":true,"delta_db":-0.21387735986343887},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.210701544963181e-05,"output_power":6.685906611461494e-05,"reference_active":true,"delta_db":-0.3281721415047665},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.009713658025179e-06,"output_power":3.49789525655314e-06,"reference_active":true,"delta_db":-0.5930655934057707}],"active_band_count":3,"active_band_mae_db":0.37837169825799205}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001241617701093293,"log_rms_mae_db":1.9654086292473718,"rms_envelope_nrmse":0.20128094997131424,"envelope_correlation":0.9746955204182853,"rms_gain_error_db":-0.774900320179934,"reference_mean_power":0.00024591411290482445,"output_mean_power":0.00020572800820812336},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.707803249714263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.578711086190903e-05,"output_power":8.570287890246958e-05,"reference_active":true,"delta_db":-0.4831166339025997},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5502167493014163e-05,"output_power":1.5764366755319802e-05,"reference_active":true,"delta_db":0.07284105189925072},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901592249182956e-06,"output_power":6.923453616231434e-06,"reference_active":true,"delta_db":0.6935358691514114}],"active_band_count":3,"active_band_mae_db":0.41649785165108727}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":1.6588430447381517,"rms_envelope_nrmse":0.1796425461168292,"envelope_correlation":0.9797010200960833,"rms_gain_error_db":-0.7749073131260018,"reference_mean_power":0.0002474004325310524,"output_mean_power":0.00020697110734870235},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.5717362127850396e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.0470755798443e-05,"output_power":8.040880113538947e-05,"reference_active":true,"delta_db":-0.5120463145657016},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.549692811018895e-06,"output_power":7.555824169268952e-06,"reference_active":true,"delta_db":0.0035256194200902887},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.070423026829598e-07,"output_power":5.001263088447606e-07,"reference_active":true,"delta_db":0.8944015445876202}],"active_band_count":3,"active_band_mae_db":0.46999115952447074}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":2.0792057505434753,"rms_envelope_nrmse":0.20572713052873884,"envelope_correlation":0.973070350955005,"rms_gain_error_db":-0.579703264543227,"reference_mean_power":2.0515613806993188e-05,"output_mean_power":1.7952055767767767e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":4.904304308942734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6213437163626398e-06,"output_power":1.3887220283205037e-06,"reference_active":true,"delta_db":-0.6725976829589233},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.453801656353397e-07,"output_power":6.925120504906339e-07,"reference_active":true,"delta_db":-0.31950497096859015},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4092114970939407e-07,"output_power":3.1757446636983084e-07,"reference_active":true,"delta_db":1.199706499855118}],"active_band_count":3,"active_band_mae_db":0.7306030512608771}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":1.6483231398779272,"rms_envelope_nrmse":0.13528923288024683,"envelope_correlation":0.9892665762939961,"rms_gain_error_db":-0.5797310998239054,"reference_mean_power":2.063955290066711e-05,"output_mean_power":1.8060392118652523e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":4.623298998188315e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6281370009575067e-06,"output_power":1.435981831540364e-06,"reference_active":true,"delta_db":-0.5454200104429686},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7397133003103594e-07,"output_power":3.4852832433709536e-07,"reference_active":true,"delta_db":-0.30600230637623654},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7174687173869285e-08,"output_power":2.4029303185530312e-08,"reference_active":true,"delta_db":1.4585234168228758}],"active_band_count":3,"active_band_mae_db":0.769981911214027}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8239240646362305,"predicted_mos":3.8239240646362305,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":220672,"input_sample_rate":44100,"input_seconds":5.00390022675737,"input_float32_sha256":"b433db8a384cdc222b06a8af606a6b6c093a2e8747b8dd9dfef39d4f75d5b212","internal_samples":80063,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.373933356000634},"started_at":"2026-10-13T07:08:28.298774+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007303054999283631,"native_input_samples":220280,"native_input_sha256":"e5cac63f933d0d98dbc25d7aaab2db2e7c3365a18ce958a63d7d224c0905c698","output_sha256":"d8fb5b5151bab5332b981820624ce16b75480bb7e2a41538531ffda30237f161","output_samples":220672,"length_delta_samples":392,"analysis_s":0.12028768100026355,"synthesis_s":27.125367652000932,"total_s":27.245655333001196,"rtf":5.45457327122459,"output_peak":0.4369792342185974,"output_rms":0.06796224688772255,"finished_at":"2026-10-13T07:08:55.565366+00:00","evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"d8fb5b5151bab5332b981820624ce16b75480bb7e2a41538531ffda30237f161","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":79920,"output_samples":80063,"length_delta_samples":143,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.34990995222112,"output_peak":0.4367084024476542},"common_prefix":{"samples":79920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998213906548593,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.395097173447585,"spectral_convergence":0.4295005461578715,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.278756478519243,"spectral_convergence":0.3296329360470246,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.700822811812865,"spectral_convergence":0.3951494903672891,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.0006195175492258868,"delta_db":-2.2288799303221216,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.002848577029656757,"delta_db":-0.5852900304056756,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0009063387097597836,"delta_db":0.5629053872108202,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.00020173428520633795,"delta_db":-0.6843417656558934,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":4.8736596308073234e-05,"delta_db":-0.4351985566346373,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":7.275130014040535e-06,"delta_db":-2.7023025365713838,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1998197011334222,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6614964966775001,"pitch":null,"output_rms_dbfs":-23.34214638493112,"output_peak":0.4367084024476542,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.124892154593232},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79920,"output_samples":79920,"intelligibility":{"estoi":{"value":0.7724942570790116,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07344589130405568,"rms_dbfs":-22.68064988825362,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.093818785323364,"mean":-6.284627828512106e-05,"zero_energy":false},"output":{"rms":0.06806011534056403,"rms_dbfs":-23.34214638493112,"peak":0.4367084024476542,"peak_dbfs":-7.196169039908668,"crest_factor_db":16.14597734502245,"mean":-4.616673528400954e-05,"zero_energy":false},"rms_difference_db":-0.6614964966775005},"full_scale":{"reference":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":3.1115009036135117e-06,"output_rms_dbfs":-55.07030068555678,"output_zero_energy":false,"output_peak":0.041732369186847904,"output_energy_fraction":0.0001304430099865614,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":128,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.803030303030303,"absolute_difference_cents_median":8.233010196169877,"absolute_difference_cents_p95":53.98937793025689,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001944445221026879,"log_rms_mae_db":1.9772446778512542,"rms_envelope_nrmse":0.137509386677073,"envelope_correlation":0.9863846277566969,"rms_gain_error_db":-0.6614959361474042,"reference_mean_power":0.005399702914122853,"output_mean_power":0.004636820378059671},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":2.8891675057826967e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000788720956533661,"output_power":0.000799773288103288,"reference_active":true,"delta_db":0.060435145887713926},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016463444647070574,"output_power":0.00015642936651371676,"reference_active":true,"delta_db":-0.22202421272049908},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1066196116820634e-05,"output_power":4.458193964407186e-05,"reference_active":true,"delta_db":-0.5897454920863512}],"active_band_count":3,"active_band_mae_db":0.2907349502315214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00017595749087279735,"log_rms_mae_db":1.4085484820374174,"rms_envelope_nrmse":0.10168483430592633,"envelope_correlation":0.9948360852924832,"rms_gain_error_db":-0.6614942404588546,"reference_mean_power":0.005432355653378944,"output_mean_power":0.004664861684053506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":2.667750009815418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007158752148089716,"output_power":0.0007263048499308173,"reference_active":true,"delta_db":0.06281617752624816},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.294187442575086e-05,"output_power":7.021967331155873e-05,"reference_active":true,"delta_db":-0.1651811478304206},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.759052873918178e-06,"output_power":3.307807309455055e-06,"reference_active":true,"delta_db":-0.5553823188685765}],"active_band_count":3,"active_band_mae_db":0.26112654807508173}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001659932762873133,"log_rms_mae_db":2.592720383014169,"rms_envelope_nrmse":0.1862033504782814,"envelope_correlation":0.9738719725804741,"rms_gain_error_db":0.1012189749983225,"reference_mean_power":0.0017617711849040457,"output_mean_power":0.0018033141895711568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.325463642910379e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157322295664234,"output_power":0.0007384939934126239,"reference_active":true,"delta_db":0.789550806307125},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015692982702212585,"output_power":0.0001543103000878673,"reference_active":true,"delta_db":-0.07310580204319303},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0206834209715196e-05,"output_power":3.654527903392142e-05,"reference_active":true,"delta_db":-0.41468596938423685}],"active_band_count":3,"active_band_mae_db":0.4257808592448516}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00014319111528307625,"log_rms_mae_db":1.897728410019931,"rms_envelope_nrmse":0.14058106866274017,"envelope_correlation":0.9840457351797725,"rms_gain_error_db":0.1012185215300056,"reference_mean_power":0.0017724260962696937,"output_mean_power":0.001814220156976929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.2006657592514347e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005739716253322745,"output_power":0.0006918909632471949,"reference_active":true,"delta_db":0.8114723504924242},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.210701544963181e-05,"output_power":7.217856731245996e-05,"reference_active":true,"delta_db":0.0043073716928049635},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.009713658025179e-06,"output_power":3.6664349847564292e-06,"reference_active":true,"delta_db":-0.38869371606073155}],"active_band_count":3,"active_band_mae_db":0.4014911460819868}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001241617701093293,"log_rms_mae_db":3.1339824773966796,"rms_envelope_nrmse":0.24678222010826215,"envelope_correlation":0.9590377426138844,"rms_gain_error_db":-0.7595321842269588,"reference_mean_power":0.00024591411290482445,"output_mean_power":0.00020645729598941194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.707803249714263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.578711086190903e-05,"output_power":8.73386128009081e-05,"reference_active":true,"delta_db":-0.4010078449228682},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5502167493014163e-05,"output_power":1.330680011749955e-05,"reference_active":true,"delta_db":-0.663187915518416},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901592249182956e-06,"output_power":4.8887775508848815e-06,"reference_active":true,"delta_db":-0.8176892362338274}],"active_band_count":3,"active_band_mae_db":0.6272949988917039}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":2.549910957733257,"rms_envelope_nrmse":0.19321778392495026,"envelope_correlation":0.975171585010836,"rms_gain_error_db":-0.7595224887572054,"reference_mean_power":0.0002474004325310524,"output_mean_power":0.00020770559994597132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.5717362127850396e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.0470755798443e-05,"output_power":8.250965647278205e-05,"reference_active":true,"delta_db":-0.4000343951256594},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.549692811018895e-06,"output_power":6.220202986591812e-06,"reference_active":true,"delta_db":-0.8412472356457362},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.070423026829598e-07,"output_power":3.816502877936076e-07,"reference_active":true,"delta_db":-0.2797395228166807}],"active_band_count":3,"active_band_mae_db":0.5070070511960254}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":4.064719916501026,"rms_envelope_nrmse":0.33656554909471376,"envelope_correlation":0.9280458674014906,"rms_gain_error_db":-1.2878412593016875,"reference_mean_power":2.0515613806993188e-05,"output_mean_power":1.525107262701133e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":4.904304308942734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6213437163626398e-06,"output_power":1.55301404335799e-06,"reference_active":true,"delta_db":-0.1869970984275551},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.453801656353397e-07,"output_power":6.360270297046472e-07,"reference_active":true,"delta_db":-0.6890225954283323},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4092114970939407e-07,"output_power":2.821445468696534e-07,"reference_active":true,"delta_db":0.685967338714253}],"active_band_count":3,"active_band_mae_db":0.5206623441900468}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":3.4816556546902215,"rms_envelope_nrmse":0.2677212205946025,"envelope_correlation":0.9565427803602595,"rms_gain_error_db":-1.2878233261146392,"reference_mean_power":2.063955290066711e-05,"output_mean_power":1.5343270887876016e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":4.623298998188315e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6281370009575067e-06,"output_power":1.5679774668002408e-06,"reference_active":true,"delta_db":-0.16351128961401362},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7397133003103594e-07,"output_power":3.229215999164615e-07,"reference_active":true,"delta_db":-0.6374121339025267},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7174687173869285e-08,"output_power":2.3288477105140687e-08,"reference_active":true,"delta_db":1.3225225440046737}],"active_band_count":3,"active_band_mae_db":0.7078153225070714}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"golf","status":"measured","mos":{"status":"measured","value":3.0752227306365967,"predicted_mos":3.0752227306365967,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":119760,"input_sample_rate":24000,"input_seconds":4.99,"input_float32_sha256":"31da13b4f9c3bb11a4f0a41bc729bf999b521579f84aacb3116ec914f99f2680","internal_samples":79840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7052988360010204},"started_at":"2026-10-13T05:02:16.967038+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003769348999412614,"native_input_samples":119880,"native_input_sha256":"f6a05ee4bd5792fa8c852d32da7f1168d28d75d498982647806928ee65fa7620","output_sha256":"4c27f0c09fc6926d336bf3bd1ee2c2ad2741d8ae9ec068b930debc8b5c435ab7","output_samples":119760,"length_delta_samples":-120,"analysis_s":0.5553248570004143,"synthesis_s":0.11795482999968954,"total_s":0.6732796870001039,"rtf":0.13479072812814893,"output_peak":0.3789236545562744,"output_rms":0.06637600203210103,"finished_at":"2026-10-13T05:02:17.651360+00:00","analysis_stages":{"external_f0_s":0.04693336399941472},"evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"4c27f0c09fc6926d336bf3bd1ee2c2ad2741d8ae9ec068b930debc8b5c435ab7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":79920,"output_samples":79840,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.55609263501243,"output_peak":0.3770513370704287},"common_prefix":{"samples":79840,"sample_rate":16000,"reference_coverage":0.998998998998999,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79840,"output_samples":79840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.353564204547112,"spectral_convergence":0.42942491264668947,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.576561341589743,"spectral_convergence":0.25671726900431097,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.571386301431891,"spectral_convergence":0.2841725744389763,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001036046046476064,"output_power":0.0008322708646055874,"delta_db":-0.9511436630281678,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032643144549746096,"output_power":0.0027421220897111535,"delta_db":-0.7570520083139998,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007954457891713772,"output_power":0.0006139246494466582,"delta_db":-1.1249551625109664,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023640044248147204,"output_power":0.0001679527667434319,"delta_db":-1.48461122580425,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.3927411030950785e-05,"output_power":4.4575008617269346e-05,"delta_db":-0.8271813519728153,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.356768331900497e-05,"output_power":8.668729443806097e-06,"delta_db":-1.9455024966281529,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1817409847097256,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8797904216367372,"pitch":null,"output_rms_dbfs":-23.55609263501243,"output_peak":0.3770513370704287,"reference_rms_dbfs":-22.676302213375692,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.833837282522915},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79840,"output_samples":79840,"intelligibility":{"estoi":{"value":0.8889412444066196,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.073482663448282,"rms_dbfs":-22.676302213375692,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.089471110445437,"mean":-6.328460448729491e-05,"zero_energy":false},"output":{"rms":0.06640417237318681,"rms_dbfs":-23.55609263501243,"peak":0.3770513370704287,"peak_dbfs":-8.471990296227363,"crest_factor_db":15.084102338785069,"mean":0.0005693788725775036,"zero_energy":false},"rms_difference_db":-0.8797904216367378},"full_scale":{"reference":{"sample_count":79840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":1.0356906419035757e-06,"output_rms_dbfs":-59.84769947851003,"output_zero_energy":false,"output_peak":0.009184408805195483,"output_energy_fraction":4.565732932200302e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":146,"both_estimated_voiced_frames":125,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.946969696969697,"absolute_difference_cents_median":4.71577327168978,"absolute_difference_cents_p95":23.044817923628283,"difference_over_600_cents_frames":4,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":997,"rms_floor":0.0001944445221026879,"log_rms_mae_db":2.202911970686342,"rms_envelope_nrmse":0.16874288445033678,"envelope_correlation":0.9799601228161656,"rms_gain_error_db":-0.8797903607202856,"reference_mean_power":0.005405117134183436,"output_mean_power":0.004413936427891924},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":997,"frequency_resolution_hz":0.20060180541624872,"power_floor":2.890355167344666e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00078920511829847,"output_power":0.0006339612158727629,"reference_active":true,"delta_db":-0.9512720348813951},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016427757872593748,"output_power":0.00011494994456310259,"reference_active":true,"delta_db":-1.550695270390404},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1105819469362816e-05,"output_power":3.9074618527273486e-05,"reference_active":true,"delta_db":-1.1657560964871623}],"active_band_count":3,"active_band_mae_db":1.222574467252987}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":991,"rms_floor":0.00017595749087279735,"log_rms_mae_db":1.7101202383641314,"rms_envelope_nrmse":0.13626380449313696,"envelope_correlation":0.9897199670802909,"rms_gain_error_db":-0.8797929997209191,"reference_mean_power":0.005437835488584355,"output_mean_power":0.004440652250646672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":991,"frequency_resolution_hz":0.20181634712411706,"power_floor":2.668561412377551e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007162531988624354,"output_power":0.0005718538259971806,"reference_active":true,"delta_db":-0.9778154347649177},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.251487280867327e-05,"output_power":5.204756856245066e-05,"reference_active":true,"delta_db":-1.4402664359601989},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7564653406099925e-06,"output_power":2.787553798931047e-06,"reference_active":true,"delta_db":-1.2955612824726994}],"active_band_count":3,"active_band_mae_db":1.2378810510659388}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":997,"rms_floor":0.0001659932762873133,"log_rms_mae_db":2.2551814811481394,"rms_envelope_nrmse":0.21738702630921483,"envelope_correlation":0.9654652324686147,"rms_gain_error_db":-0.7729632182692401,"reference_mean_power":0.0017635381449666874,"output_mean_power":0.0014760074038731124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":997,"frequency_resolution_hz":0.20060180541624872,"power_floor":1.3265284449820896e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006168256689153099,"output_power":0.00045497094688595605,"reference_active":true,"delta_db":-1.3217877361615025},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015693431111996167,"output_power":0.0001121066745879652,"reference_active":true,"delta_db":-1.460864350373017},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0232901404147895e-05,"output_power":2.8334189529311967e-05,"reference_active":true,"delta_db":-1.5227055833551923}],"active_band_count":3,"active_band_mae_db":1.4351192232965708}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":991,"rms_floor":0.00014319111528307625,"log_rms_mae_db":1.6367340161728503,"rms_envelope_nrmse":0.16720554768995297,"envelope_correlation":0.980340617890069,"rms_gain_error_db":-0.7729635010450436,"reference_mean_power":0.001774214203818478,"output_mean_power":0.001484942720323942},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":991,"frequency_resolution_hz":0.20181634712411706,"power_floor":1.2014630928165769e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005750737617163215,"output_power":0.00042222504808437034,"reference_active":true,"delta_db":-1.3417955912621127},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188659742457197e-05,"output_power":5.486378067639951e-05,"reference_active":true,"delta_db":-1.173621961984619},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.007981202908779e-06,"output_power":2.2824540449383303e-06,"reference_active":true,"delta_db":-2.44523633875593}],"active_band_count":3,"active_band_mae_db":1.6535512973342206}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":997,"rms_floor":0.0001241617701093293,"log_rms_mae_db":2.0763873463803617,"rms_envelope_nrmse":0.21784749498230008,"envelope_correlation":0.9774405486868301,"rms_gain_error_db":-1.3115347969948923,"reference_mean_power":0.0002461605517005965,"output_mean_power":0.00018199731331977468},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":997,"frequency_resolution_hz":0.20060180541624872,"power_floor":1.7078821776794767e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.58558407897561e-05,"output_power":7.137306059668062e-05,"reference_active":true,"delta_db":-1.280842604794875},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5484970401129606e-05,"output_power":1.1032967911231486e-05,"reference_active":true,"delta_db":-1.4721802449860983},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.905758137516535e-06,"output_power":4.272209108087108e-06,"reference_active":true,"delta_db":-1.4062315546599733}],"active_band_count":3,"active_band_mae_db":1.3864181348136488}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":991,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":1.6439184320919897,"rms_envelope_nrmse":0.17189947169624573,"envelope_correlation":0.9914416529538846,"rms_gain_error_db":-1.311538371175548,"reference_mean_power":0.0002476497110123701,"output_mean_power":0.00018309816355240602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":991,"frequency_resolution_hz":0.20181634712411706,"power_floor":1.571651512915129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.053071245336861e-05,"output_power":6.66231870483295e-05,"reference_active":true,"delta_db":-1.331705340053841},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.526538774164114e-06,"output_power":4.995870079973918e-06,"reference_active":true,"delta_db":-1.7798416750021049},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066575530832664e-07,"output_power":3.0389515840433104e-07,"reference_active":true,"delta_db":-1.2650506206685763}],"active_band_count":3,"active_band_mae_db":1.4588658785748407}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":997,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":2.350353818280426,"rms_envelope_nrmse":0.2243317058839306,"envelope_correlation":0.9729219599246128,"rms_gain_error_db":-1.0676868474829786,"reference_mean_power":2.0536171089055278e-05,"output_mean_power":1.6060194074913755e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":997,"frequency_resolution_hz":0.20060180541624872,"power_floor":4.8942975204686146e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.618172701581962e-06,"output_power":1.1829985086057379e-06,"reference_active":true,"delta_db":-1.3604067327452847},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.443343382563998e-07,"output_power":5.617277432615947e-07,"reference_active":true,"delta_db":-1.2224218084590426},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4055728492175065e-07,"output_power":2.1743140865824384e-07,"reference_active":true,"delta_db":-0.43896233903591103}],"active_band_count":3,"active_band_mae_db":1.007263626746746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79840,"uncovered_tail_samples":0,"metrics":{"frame_count":991,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":1.9955337283394983,"rms_envelope_nrmse":0.16881614920346205,"envelope_correlation":0.9882370591312616,"rms_gain_error_db":-1.0676744793252333,"reference_mean_power":2.066034453184498e-05,"output_mean_power":1.6157349213197945e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":991,"frequency_resolution_hz":0.20181634712411706,"power_floor":4.614549257849188e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6254110304119603e-06,"output_power":1.2107539503155112e-06,"reference_active":true,"delta_db":-1.279073079593458},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.731386096450129e-07,"output_power":2.860381461982762e-07,"reference_active":true,"delta_db":-1.1544623423889362},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7111019016380207e-08,"output_power":1.4774039446631498e-08,"reference_active":true,"delta_db":-0.6377661969743332}],"active_band_count":3,"active_band_mae_db":1.0237672063189092}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.8714189529418945,"predicted_mos":2.8714189529418945,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":120000,"input_sample_rate":24000,"input_seconds":5.0,"input_float32_sha256":"a1a11e1092192e026acf9190fb4c8f2ca2d5771e5f1cd4dd7e221a7cb7c7d66f","internal_samples":80000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.921384343000682},"started_at":"2026-10-13T05:08:04.594691+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0024010930001168163,"native_input_samples":119880,"native_input_sha256":"f6a05ee4bd5792fa8c852d32da7f1168d28d75d498982647806928ee65fa7620","output_sha256":"2c6a4163ac284bf5590f8bc9d05a1c42f874ce9f3c2d1048fd8a36b0d55b573a","output_samples":120000,"length_delta_samples":120,"analysis_s":1.8254013440000563,"synthesis_s":9.849232692999976,"total_s":11.674634037000033,"rtf":2.337264071471478,"output_peak":0.4140229821205139,"output_rms":0.07208598244768515,"finished_at":"2026-10-13T05:08:16.280416+00:00","analysis_stages":{"external_f0_s":1.2079396909994102},"evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"2c6a4163ac284bf5590f8bc9d05a1c42f874ce9f3c2d1048fd8a36b0d55b573a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":79920,"output_samples":80000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.83936708591707,"output_peak":0.41432909755992253},"common_prefix":{"samples":79920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.999,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.4387887975047295,"spectral_convergence":0.472415749662965,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.707149914901074,"spectral_convergence":0.2701183112625962,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.3688795574481425,"spectral_convergence":0.2802738804508849,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.0011653711128716791,"delta_db":0.5152264234470629,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.0030326981495361026,"delta_db":-0.31327784674835596,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0007265304326946314,"delta_db":-0.39746179785844943,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.0002152290823400481,"delta_db":-0.4031293638527328,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":5.480021520462218e-05,"delta_db":0.07407212909395937,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.12944067251139e-05,"delta_db":-0.792075895256571,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.41587390937618857,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1543723761938363,"pitch":null,"output_rms_dbfs":-22.835022264447456,"output_peak":0.41432909755992253,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.838272756617982},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79920,"output_samples":79920,"intelligibility":{"estoi":{"value":0.8881767496206016,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07344589130405568,"rms_dbfs":-22.68064988825362,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.093818785323364,"mean":-6.284627828512106e-05,"zero_energy":false},"output":{"rms":0.07215208520460235,"rms_dbfs":-22.835022264447456,"peak":0.41432909755992253,"peak_dbfs":-7.653091318781048,"crest_factor_db":15.181930945666409,"mean":4.422710356910023e-05,"zero_energy":false},"rms_difference_db":-0.15437237619383737},"full_scale":{"reference":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":3.7894890198160645e-06,"output_rms_dbfs":-54.21419346980556,"output_zero_energy":false,"output_peak":0.036817441575577876,"output_energy_fraction":0.00014135758637932937,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":151,"both_estimated_voiced_frames":125,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.946969696969697,"absolute_difference_cents_median":7.560272362506229,"absolute_difference_cents_p95":41.88763897494095,"difference_over_600_cents_frames":3,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001944445221026879,"log_rms_mae_db":2.933666909835781,"rms_envelope_nrmse":0.15371823673422377,"envelope_correlation":0.9767275134852164,"rms_gain_error_db":-0.15437174212147997,"reference_mean_power":0.005399702914122853,"output_mean_power":0.005211139411787966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":2.8891675057826967e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000788720956533661,"output_power":0.0007503121192285943,"reference_active":true,"delta_db":-0.21681418963995858},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016463444647070574,"output_power":0.0001454859104294914,"reference_active":true,"delta_db":-0.5369977168597658},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1066196116820634e-05,"output_power":4.339332352417631e-05,"reference_active":true,"delta_db":-0.7071059447262633}],"active_band_count":3,"active_band_mae_db":0.48697261707532924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00017595749087279735,"log_rms_mae_db":2.1871012926259006,"rms_envelope_nrmse":0.09577419715671409,"envelope_correlation":0.990392074169678,"rms_gain_error_db":-0.15437004892790251,"reference_mean_power":0.005432355653378944,"output_mean_power":0.005242653925817566},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":2.667750009815418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007158752148089716,"output_power":0.000679627901912868,"reference_active":true,"delta_db":-0.2256612604953773},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.294187442575086e-05,"output_power":6.430957752495535e-05,"reference_active":true,"delta_db":-0.547012628751627},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.759052873918178e-06,"output_power":3.113642417488546e-06,"reference_active":true,"delta_db":-0.8180969936730398}],"active_band_count":3,"active_band_mae_db":0.5302569609733481}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001659932762873133,"log_rms_mae_db":2.7461233713744146,"rms_envelope_nrmse":0.21220192954359407,"envelope_correlation":0.9637977269001644,"rms_gain_error_db":-0.2884700287546401,"reference_mean_power":0.0017617711849040457,"output_mean_power":0.0016485514269880255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.325463642910379e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157322295664234,"output_power":0.0005823750291565376,"reference_active":true,"delta_db":-0.24189141470937728},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015692982702212585,"output_power":0.00013941852252422605,"reference_active":true,"delta_db":-0.5138502001427028},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0206834209715196e-05,"output_power":3.437177428703584e-05,"reference_active":true,"delta_db":-0.6809792793700374}],"active_band_count":3,"active_band_mae_db":0.4789069647407058}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00014319111528307625,"log_rms_mae_db":1.8915829292873314,"rms_envelope_nrmse":0.12948799713677764,"envelope_correlation":0.9857487430437856,"rms_gain_error_db":-0.28846909949667043,"reference_mean_power":0.0017724260962696937,"output_mean_power":0.0016585219581739734},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.2006657592514347e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005739716253322745,"output_power":0.0005413498837906244,"reference_active":true,"delta_db":-0.2541237540610638},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.210701544963181e-05,"output_power":6.463474114021015e-05,"reference_active":true,"delta_db":-0.47511506423476535},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.009713658025179e-06,"output_power":3.812608216106614e-06,"reference_active":true,"delta_db":-0.21891180422704776}],"active_band_count":3,"active_band_mae_db":0.31605020750762564}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001241617701093293,"log_rms_mae_db":2.4845484898235144,"rms_envelope_nrmse":0.23355967606173667,"envelope_correlation":0.961491870618139,"rms_gain_error_db":-0.27481034893978085,"reference_mean_power":0.00024591411290482445,"output_mean_power":0.00023083540386156638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.707803249714263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.578711086190903e-05,"output_power":8.817170098513297e-05,"reference_active":true,"delta_db":-0.359778550662437},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5502167493014163e-05,"output_power":1.3935893703146794e-05,"reference_active":true,"delta_db":-0.46257599839280683},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901592249182956e-06,"output_power":4.628155543788273e-06,"reference_active":true,"delta_db":-1.0556125370984966}],"active_band_count":3,"active_band_mae_db":0.6259890287179135}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":1.5036526998053557,"rms_envelope_nrmse":0.13340765257453957,"envelope_correlation":0.9869885157442937,"rms_gain_error_db":-0.27481105529449534,"reference_mean_power":0.0002474004325310524,"output_mean_power":0.00023223054909631973},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.5717362127850396e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.0470755798443e-05,"output_power":8.284070269706128e-05,"reference_active":true,"delta_db":-0.3826444431137229},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.549692811018895e-06,"output_power":7.079444331474836e-06,"reference_active":true,"delta_db":-0.2793010993425758},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.070423026829598e-07,"output_power":3.072206843783005e-07,"reference_active":true,"delta_db":-1.2218909421676614}],"active_band_count":3,"active_band_mae_db":0.6279454948746533}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":2.383396952950163,"rms_envelope_nrmse":0.33975702739942476,"envelope_correlation":0.922731892328809,"rms_gain_error_db":-0.15203363420250465,"reference_mean_power":2.0515613806993188e-05,"output_mean_power":1.9809848388175978e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":4.904304308942734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6213437163626398e-06,"output_power":1.5537762789854625e-06,"reference_active":true,"delta_db":-0.1848660583558599},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.453801656353397e-07,"output_power":6.760318292906443e-07,"reference_active":true,"delta_db":-0.42410688005723296},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4092114970939407e-07,"output_power":1.9850479414841374e-07,"reference_active":true,"delta_db":-0.8410392701625241}],"active_band_count":3,"active_band_mae_db":0.48333740285853893}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":1.4828727124673768,"rms_envelope_nrmse":0.19358049362580443,"envelope_correlation":0.973265233388378,"rms_gain_error_db":-0.1520387322284727,"reference_mean_power":2.063955290066711e-05,"output_mean_power":1.992950041150402e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":4.623298998188315e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6281370009575067e-06,"output_power":1.5412456346984995e-06,"reference_active":true,"delta_db":-0.23819086620889715},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7397133003103594e-07,"output_power":3.4549310267374995e-07,"reference_active":true,"delta_db":-0.3439892725587157},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7174687173869285e-08,"output_power":1.3443223105065212e-08,"reference_active":true,"delta_db":-1.06385429115791}],"active_band_count":3,"active_band_mae_db":0.5486781433085076}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.7466797828674316,"predicted_mos":2.7466797828674316,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":120120,"input_sample_rate":24000,"input_seconds":5.005,"input_float32_sha256":"a91072b5427bfa9dd8ce7ef63e1d9862d580c9f9c899381db5d567fe8458d0d0","internal_samples":80080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.142973860998609},"started_at":"2026-10-13T05:10:24.524992+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004264970000804169,"native_input_samples":119880,"native_input_sha256":"f6a05ee4bd5792fa8c852d32da7f1168d28d75d498982647806928ee65fa7620","output_sha256":"db5095ce12d9a4440bbc2400d1176202ee26726f5214c84fd0ca5b79a6e0ae21","output_samples":120120,"length_delta_samples":240,"analysis_s":0.5979862160002085,"synthesis_s":0.10998699300034787,"total_s":0.7079732090005564,"rtf":0.14173637817828957,"output_peak":0.46026182174682617,"output_rms":0.07202370828707116,"finished_at":"2026-10-13T05:10:25.247475+00:00","analysis_stages":{"external_f0_s":0.07246668600055273},"evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"db5095ce12d9a4440bbc2400d1176202ee26726f5214c84fd0ca5b79a6e0ae21","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":79920,"output_samples":80080,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.84678920918053,"output_peak":0.46052194226286686},"common_prefix":{"samples":79920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998001998001998,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.074523862528967,"spectral_convergence":0.18471781757061784,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.70640888728381,"spectral_convergence":0.18202397718902924,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.924287574406616,"spectral_convergence":0.21369617474259595,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.0009723180546481543,"delta_db":-0.27133256189620303,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.003177864784963391,"delta_db":-0.1102155648580695,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0007794774420474831,"delta_db":-0.09196439107482143,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.00021265484701420346,"delta_db":-0.4553860441806085,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":4.787566063981477e-05,"delta_db":-0.5126027120585551,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.2040368511496475e-05,"delta_db":-0.5143123338697174,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3259689346563292,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1574536528763708,"pitch":null,"output_rms_dbfs":-22.83810354112999,"output_peak":0.46052194226286686,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.901740108073131},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79920,"output_samples":79920,"intelligibility":{"estoi":{"value":0.9296826007328326,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07344589130405568,"rms_dbfs":-22.68064988825362,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.093818785323364,"mean":-6.284627828512106e-05,"zero_energy":false},"output":{"rms":0.07212649414621886,"rms_dbfs":-22.83810354112999,"peak":0.46052194226286686,"peak_dbfs":-6.734993447164208,"crest_factor_db":16.103110093965782,"mean":-2.8451286487070157e-06,"zero_energy":false},"rms_difference_db":-0.15745365287637014},"full_scale":{"reference":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":1.759091639290797e-06,"output_rms_dbfs":-57.54711535527384,"output_zero_energy":false,"output_peak":0.019452064805891747,"output_energy_fraction":6.566516808696986e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":130,"both_estimated_voiced_frames":122,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9242424242424242,"absolute_difference_cents_median":0.9449436000280226,"absolute_difference_cents_p95":7.4592091220381365,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001944445221026879,"log_rms_mae_db":1.0997049411120263,"rms_envelope_nrmse":0.07445618632953302,"envelope_correlation":0.9948204302597021,"rms_gain_error_db":-0.15745344812379874,"reference_mean_power":0.005399702914122853,"output_mean_power":0.0052074429560638735},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":2.8891675057826967e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000788720956533661,"output_power":0.0007798793077566279,"reference_active":true,"delta_db":-0.04895982781110125},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016463444647070574,"output_power":0.00014213614552464454,"reference_active":true,"delta_db":-0.6381617384782334},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1066196116820634e-05,"output_power":3.6657540806023104e-05,"reference_active":true,"delta_db":-1.4397018209824033}],"active_band_count":3,"active_band_mae_db":0.7089411290905794}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00017595749087279735,"log_rms_mae_db":0.5698732726806044,"rms_envelope_nrmse":0.03978624564670112,"envelope_correlation":0.9987101224955451,"rms_gain_error_db":-0.15745279271149024,"reference_mean_power":0.005432355653378944,"output_mean_power":0.00523893386378705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":2.667750009815418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007158752148089716,"output_power":0.0007082368138774036,"reference_active":true,"delta_db":-0.046588289688186735},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.294187442575086e-05,"output_power":6.369001729776989e-05,"reference_active":true,"delta_db":-0.5890555257805116},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.759052873918178e-06,"output_power":2.7020001783824453e-06,"reference_active":true,"delta_db":-1.4339306107916425}],"active_band_count":3,"active_band_mae_db":0.6898581420867803}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001659932762873133,"log_rms_mae_db":1.312466355250184,"rms_envelope_nrmse":0.11253159415939963,"envelope_correlation":0.9901661876425727,"rms_gain_error_db":-0.22492741687750603,"reference_mean_power":0.0017617711849040457,"output_mean_power":0.0016728490760387707},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.325463642910379e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157322295664234,"output_power":0.0005664756052737189,"reference_active":true,"delta_db":-0.36210674541538695},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015692982702212585,"output_power":0.0001254693157264774,"reference_active":true,"delta_db":-0.9716796655855255},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0206834209715196e-05,"output_power":3.198942842751626e-05,"reference_active":true,"delta_db":-0.9929339875863243}],"active_band_count":3,"active_band_mae_db":0.7755734661957456}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00014319111528307625,"log_rms_mae_db":0.7192747677202925,"rms_envelope_nrmse":0.06221978245986802,"envelope_correlation":0.997101243044715,"rms_gain_error_db":-0.22492689897196083,"reference_mean_power":0.0017724260962696937,"output_mean_power":0.0016829664013582354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.2006657592514347e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005739716253322745,"output_power":0.0005222972250006359,"reference_active":true,"delta_db":-0.4097270488497947},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.210701544963181e-05,"output_power":5.9111870493418515e-05,"reference_active":true,"delta_db":-0.8630281816954726},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.009713658025179e-06,"output_power":2.662681049263903e-06,"reference_active":true,"delta_db":-1.7779421250968452}],"active_band_count":3,"active_band_mae_db":1.0168991185473708}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001241617701093293,"log_rms_mae_db":1.2395683009712715,"rms_envelope_nrmse":0.11077220599531305,"envelope_correlation":0.9929045924007047,"rms_gain_error_db":-0.4533224945006553,"reference_mean_power":0.00024591411290482445,"output_mean_power":0.00022153951671834631},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.707803249714263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.578711086190903e-05,"output_power":8.558780639194598e-05,"reference_active":true,"delta_db":-0.48895178644126747},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5502167493014163e-05,"output_power":1.3242675533557921e-05,"reference_active":true,"delta_db":-0.6841668662741236},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901592249182956e-06,"output_power":4.180704235936019e-06,"reference_active":true,"delta_db":-1.4971975561372246}],"active_band_count":3,"active_band_mae_db":0.8901054029508719}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":0.7676652458310624,"rms_envelope_nrmse":0.07782250657208889,"envelope_correlation":0.9972876836476878,"rms_gain_error_db":-0.45332246272576787,"reference_mean_power":0.0002474004325310524,"output_mean_power":0.00022287851645610206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.5717362127850396e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.0470755798443e-05,"output_power":8.07586496225506e-05,"reference_active":true,"delta_db":-0.4931917002617271},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.549692811018895e-06,"output_power":6.2403534336625445e-06,"reference_active":true,"delta_db":-0.827200935778992},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.070423026829598e-07,"output_power":2.714785083329502e-07,"reference_active":true,"delta_db":-1.7590409224123338}],"active_band_count":3,"active_band_mae_db":1.0264778528176843}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":1.420004645170567,"rms_envelope_nrmse":0.12527579503948882,"envelope_correlation":0.9900723051330335,"rms_gain_error_db":-0.3009486186172728,"reference_mean_power":2.0515613806993188e-05,"output_mean_power":1.9142103200279435e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":4.904304308942734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6213437163626398e-06,"output_power":1.4517489971166743e-06,"reference_active":true,"delta_db":-0.4798355807050816},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.453801656353397e-07,"output_power":6.211130007940586e-07,"reference_active":true,"delta_db":-0.7920721236008197},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4092114970939407e-07,"output_power":1.6372064989653333e-07,"reference_active":true,"delta_db":-1.6777146708918123}],"active_band_count":3,"active_band_mae_db":0.9832074583992378}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":0.9520022663884985,"rms_envelope_nrmse":0.06922463325310592,"envelope_correlation":0.9973735418341185,"rms_gain_error_db":-0.3009436972099598,"reference_mean_power":2.063955290066711e-05,"output_mean_power":1.9257766453268995e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":4.623298998188315e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6281370009575067e-06,"output_power":1.4407608522258418e-06,"reference_active":true,"delta_db":-0.5309904666840797},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7397133003103594e-07,"output_power":3.071671143310297e-07,"reference_active":true,"delta_db":-0.8546359114557822},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7174687173869285e-08,"output_power":1.1669172121463855e-08,"reference_active":true,"delta_db":-1.6784878962579497}],"active_band_count":3,"active_band_mae_db":1.0213714247992705}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_260-123440-0020","method":"magphase","status":"measured","mos":{"status":"measured","value":2.9061601161956787,"predicted_mos":2.9061601161956787,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":240710,"input_sample_rate":48000,"input_seconds":5.0147916666666665,"input_float32_sha256":"b4266da80e2ee5963d790683fb5a5a9e3f7ef81b1d990792ec1b94a5b14c0b68","internal_samples":80237,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1921854419997544},"started_at":"2026-10-13T05:14:04.010420+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.0075622340009431355,"native_input_samples":239760,"native_input_sha256":"cd3d1cd2410e4575237f389f5e8fa5d116c1efaa825e938a65e91b5369fea5bd","output_sha256":"4dcc7b99bf5301308ac636a0d8dfda80d3de94c865e05acbaa4b1a7313334eac","output_samples":240710,"length_delta_samples":950,"analysis_s":2.717890277999686,"synthesis_s":0.8865630279997276,"total_s":3.6044533059994137,"rtf":0.7216122734733561,"output_peak":0.49091860710835844,"output_rms":0.07011261198172893,"finished_at":"2026-10-13T05:14:07.633082+00:00","evaluation_input":{"reference":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"output_sha256":"4dcc7b99bf5301308ac636a0d8dfda80d3de94c865e05acbaa4b1a7313334eac","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":79920,"output_samples":80237,"length_delta_samples":317,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.08041388742135,"output_peak":0.4863671348192251},"common_prefix":{"samples":79920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9960492042324613,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79920,"output_samples":79920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.820215351960022,"spectral_convergence":0.5332275827996994,"spectral_convergence_status":"ok","amplitude_floor":1.5443736044991425e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.01647477530442,"spectral_convergence":0.3470442644700347,"spectral_convergence_status":"ok","amplitude_floor":1.1257392826698756e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.10314882012738,"spectral_convergence":0.33123727985569595,"spectral_convergence_status":"ok","amplitude_floor":1.1201855379193364e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001035002994445706,"output_power":0.0009097653575150297,"delta_db":-0.5601221075146642,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0032595449236567883,"output_power":0.002956261616709645,"delta_db":-0.4241410628799433,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007961593328208376,"output_power":0.0007686983323839943,"delta_db":-0.1524405149055139,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023616410570225164,"output_power":0.00022991736340854002,"delta_db":-0.11642119878310965,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.38734826098075e-05,"output_power":5.6393845420933516e-05,"delta_db":0.19856658472379324,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.3554110211771778e-05,"output_power":1.8401348054976783e-05,"delta_db":1.3277862726045584,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4632462902352638,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.38257484654705054,"pitch":null,"output_rms_dbfs":-23.06322473480067,"output_peak":0.4863671348192251,"reference_rms_dbfs":-22.68064988825362,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9799463157972745},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79920,"output_samples":79920,"intelligibility":{"estoi":{"value":0.7812302322240285,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07344589130405568,"rms_dbfs":-22.68064988825362,"peak":0.46844482421875,"peak_dbfs":-6.586831102930253,"crest_factor_db":16.093818785323364,"mean":-6.284627828512106e-05,"zero_energy":false},"output":{"rms":0.07028113447784633,"rms_dbfs":-23.06322473480067,"peak":0.4863671348192251,"peak_dbfs":-6.260715584611249,"crest_factor_db":16.80250915018942,"mean":-0.0012076730014895349,"zero_energy":false},"rms_difference_db":-0.3825748465470511},"full_scale":{"reference":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":499,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.780867217574269e-06,"threshold_dbfs":-54.22408574604508,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":9.962755433024511e-07,"output_mean_power":4.099015594980854e-06,"output_rms_dbfs":-53.87320429377753,"output_zero_energy":false,"output_peak":0.07728052854866803,"output_energy_fraction":0.00016115295959889163,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":492,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":127,"both_estimated_voiced_frames":117,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.8863636363636364,"absolute_difference_cents_median":9.026467151362661,"absolute_difference_cents_p95":75.18677577968445,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001944445221026879,"log_rms_mae_db":2.9613599363904823,"rms_envelope_nrmse":0.21638871343736016,"envelope_correlation":0.9541998186016282,"rms_gain_error_db":-0.3825742536518999,"reference_mean_power":0.005399702914122853,"output_mean_power":0.004944386827047553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":2.8891675057826967e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000788720956533661,"output_power":0.0007524561980287566,"reference_active":true,"delta_db":-0.20442156447192358},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016463444647070574,"output_power":0.00014936087993105604,"reference_active":true,"delta_db":-0.4228384432112506},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.1066196116820634e-05,"output_power":4.572241166274368e-05,"reference_active":true,"delta_db":-0.48004379271482817}],"active_band_count":3,"active_band_mae_db":0.3691012667993341}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00017595749087279735,"log_rms_mae_db":1.8430880979827737,"rms_envelope_nrmse":0.14836222687217882,"envelope_correlation":0.9776786637790645,"rms_gain_error_db":-0.38257394923330906,"reference_mean_power":0.005432355653378944,"output_mean_power":0.004974286556558645},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":2.667750009815418e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007158752148089716,"output_power":0.0006852333850063692,"reference_active":true,"delta_db":-0.1899881252557413},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.294187442575086e-05,"output_power":6.494451912257612e-05,"reference_active":true,"delta_db":-0.5043441397157621},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.759052873918178e-06,"output_power":3.1083139537941955e-06,"reference_active":true,"delta_db":-0.8255355638176229}],"active_band_count":3,"active_band_mae_db":0.5066226095963754}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001659932762873133,"log_rms_mae_db":3.591771963072535,"rms_envelope_nrmse":0.28725373661791265,"envelope_correlation":0.9329719894376317,"rms_gain_error_db":-0.24309046935523407,"reference_mean_power":0.0017617711849040457,"output_mean_power":0.0016658675003426108},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.325463642910379e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157322295664234,"output_power":0.0005780397151093999,"reference_active":true,"delta_db":-0.2743420835284163},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015692982702212585,"output_power":0.00013967228186319546,"reference_active":true,"delta_db":-0.5059526765207926},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0206834209715196e-05,"output_power":3.404510115976059e-05,"reference_active":true,"delta_db":-0.7224525018152519}],"active_band_count":3,"active_band_mae_db":0.5009157539548202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":0.00014319111528307625,"log_rms_mae_db":2.295079627506884,"rms_envelope_nrmse":0.1850736114552363,"envelope_correlation":0.9697310266213576,"rms_gain_error_db":-0.24309133759509022,"reference_mean_power":0.0017724260962696937,"output_mean_power":0.0016759420664537773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.2006657592514347e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005739716253322745,"output_power":0.0005393557998177456,"reference_active":true,"delta_db":-0.2701507013019383},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.210701544963181e-05,"output_power":6.186965502092158e-05,"reference_active":true,"delta_db":-0.6649982579372188},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.009713658025179e-06,"output_power":3.3051435347171117e-06,"reference_active":true,"delta_db":-0.8392303522398068}],"active_band_count":3,"active_band_mae_db":0.5914597704929879}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":0.0001241617701093293,"log_rms_mae_db":3.4023331868744098,"rms_envelope_nrmse":0.31020861760697843,"envelope_correlation":0.9325883007847553,"rms_gain_error_db":-0.07894871924197733,"reference_mean_power":0.00024591411290482445,"output_mean_power":0.0002414841226072263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":1.707803249714263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.578711086190903e-05,"output_power":9.600939076304082e-05,"reference_active":true,"delta_db":0.010066396091842655},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5502167493014163e-05,"output_power":1.5269823536469534e-05,"reference_active":true,"delta_db":-0.06558406685271242},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901592249182956e-06,"output_power":5.803620142715061e-06,"reference_active":true,"delta_db":-0.07270220725111036}],"active_band_count":3,"active_band_mae_db":0.04945089006522182}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":9.745648796239739e-05,"log_rms_mae_db":1.967995255921015,"rms_envelope_nrmse":0.1859248001581389,"envelope_correlation":0.9739742407299826,"rms_gain_error_db":-0.07896237484537687,"reference_mean_power":0.0002474004325310524,"output_mean_power":0.00024294290321562396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":1.5717362127850396e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.0470755798443e-05,"output_power":9.107369388752056e-05,"reference_active":true,"delta_db":0.028847331684208284},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.549692811018895e-06,"output_power":7.569562872990273e-06,"reference_active":true,"delta_db":0.011415195850662006},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.070423026829598e-07,"output_power":4.125249729017376e-07,"reference_active":true,"delta_db":0.05810697947013215}],"active_band_count":3,"active_band_mae_db":0.03278983566833415}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":998,"rms_floor":3.626382194129561e-05,"log_rms_mae_db":3.4288961741256494,"rms_envelope_nrmse":0.548169779776762,"envelope_correlation":0.8353090300895687,"rms_gain_error_db":1.0565420904066931,"reference_mean_power":2.0515613806993188e-05,"output_mean_power":2.6166083574832594e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":998,"frequency_resolution_hz":0.2004008016032064,"power_floor":4.904304308942734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6213437163626398e-06,"output_power":1.9852554298345323e-06,"reference_active":true,"delta_db":0.8794129976446335},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.453801656353397e-07,"output_power":9.202752649414346e-07,"reference_active":true,"delta_db":0.9153991715069436},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4092114970939407e-07,"output_power":2.4158038399055996e-07,"reference_active":true,"delta_db":0.011867402225490094}],"active_band_count":3,"active_band_mae_db":0.6022265237923558}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79920,"uncovered_tail_samples":0,"metrics":{"frame_count":992,"rms_floor":2.511407055347502e-05,"log_rms_mae_db":2.1794178493944703,"rms_envelope_nrmse":0.3691346385227822,"envelope_correlation":0.9257090186295905,"rms_gain_error_db":1.0565317168421222,"reference_mean_power":2.063955290066711e-05,"output_mean_power":2.632409545476384e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":992,"frequency_resolution_hz":0.20161290322580647,"power_floor":4.623298998188315e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6281370009575067e-06,"output_power":1.987638807349394e-06,"reference_active":true,"delta_db":0.8664652130706667},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7397133003103594e-07,"output_power":4.6140136698210937e-07,"reference_active":true,"delta_db":0.912405679393615},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7174687173869285e-08,"output_power":1.672077820364316e-08,"reference_active":true,"delta_db":-0.11632349280316699}],"active_band_count":3,"active_band_mae_db":0.6317314617558162}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"hnm","status":"measured","mos":{"status":"measured","value":2.6614232063293457,"predicted_mos":2.6614232063293457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59680,"input_sample_rate":16000,"input_seconds":3.73,"input_float32_sha256":"875e728789bb4dafb59e40de8dd85a29e958fee7755ff817b02a10641cb53e54","internal_samples":59680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.484903092999957},"started_at":"2026-10-13T04:38:46.017661+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.0842999472515658e-05,"native_input_samples":59680,"native_input_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","output_sha256":"14fa20b9231f7e5ffaf0b26e8e6839fe6509230fc54cb1864d339035b4fc5a1f","output_samples":59680,"length_delta_samples":0,"analysis_s":0.7614011959994968,"synthesis_s":0.3136047480002162,"total_s":1.075005943999713,"rtf":0.28820534691681315,"output_peak":0.9692718217192645,"output_rms":0.14179559391606325,"finished_at":"2026-10-13T04:38:47.096143+00:00","evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"14fa20b9231f7e5ffaf0b26e8e6839fe6509230fc54cb1864d339035b4fc5a1f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.789870235105239,"spectral_convergence":0.19810514813399543,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.479149724188452,"spectral_convergence":0.1883938719278663,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.659319079350597,"spectral_convergence":0.184590828670329,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.004350248626474489,"delta_db":0.06578904179695264,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.01201999406636754,"delta_db":-0.23022270169591535,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.0029962422455383224,"delta_db":-0.16008019478151034,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.0004893780638435625,"delta_db":-0.540723394602804,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00014008589871055332,"delta_db":-0.559484203942837,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":0.00011004155307464569,"delta_db":0.7953657179082813,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.39194420912138345,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16263577389065972,"pitch":null,"output_rms_dbfs":-16.96674527976855,"output_peak":0.9692718217192645,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.642779679548096},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.789870235105239,"spectral_convergence":0.19810514813399543,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.479149724188452,"spectral_convergence":0.1883938719278663,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.659319079350597,"spectral_convergence":0.184590828670329,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.004350248626474489,"delta_db":0.06578904179695264,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.01201999406636754,"delta_db":-0.23022270169591535,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.0029962422455383224,"delta_db":-0.16008019478151034,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.0004893780638435625,"delta_db":-0.540723394602804,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00014008589871055332,"delta_db":-0.559484203942837,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":0.00011004155307464569,"delta_db":0.7953657179082813,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.39194420912138345,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16263577389065972,"pitch":null,"output_rms_dbfs":-16.96674527976855,"output_peak":0.9692718217192645,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.642779679548096},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.8893619745056927,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14447560597487225,"rms_dbfs":-16.80410950587789,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.859693585918194,"mean":-0.0023839802269002386,"zero_energy":false},"output":{"rms":0.14179559391606325,"rms_dbfs":-16.96674527976855,"peak":0.9692718217192645,"peak_dbfs":-0.2710882542751347,"crest_factor_db":16.695657025493414,"mean":-0.0002855562870992768,"zero_energy":false},"rms_difference_db":-0.16263577389065986},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":59,"quiet_samples":9440,"reference_mean_power":6.490114235745396e-06,"output_mean_power":7.172103362159427e-06,"output_rms_dbfs":-51.44353460145548,"output_zero_energy":false,"output_peak":0.022227184794326692,"output_energy_fraction":5.642404895794324e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":116,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9906542056074766,"absolute_difference_cents_median":1.2225646274365944,"absolute_difference_cents_p95":6.335684045521351,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00043785897255075944,"log_rms_mae_db":1.0830353657937162,"rms_envelope_nrmse":0.05999301911278998,"envelope_correlation":0.9966228823348916,"rms_gain_error_db":-0.16263693626811299,"reference_mean_power":0.0209012125553907,"output_mean_power":0.020132967302944428},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.005551294872497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00444581511703579,"output_power":0.004264351635945967,"reference_active":true,"delta_db":-0.1809838907236782},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039454721092167534,"output_power":0.0003926005020793033,"reference_active":true,"delta_db":-0.021481270105337173},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010840168168493419,"output_power":0.0001246749991141161,"reference_active":true,"delta_db":0.6074335414067493}],"active_band_count":3,"active_band_mae_db":0.2699662340785882}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0004107294113879301,"log_rms_mae_db":0.5982089490057759,"rms_envelope_nrmse":0.03296440296824044,"envelope_correlation":0.9991819606172591,"rms_gain_error_db":-0.16263658862096383,"reference_mean_power":0.021070817398047652,"output_mean_power":0.02029633977239792},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.012296407388438e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00391054354874771,"output_power":0.0037698160755839628,"reference_active":true,"delta_db":-0.15916964597533262},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019349823972573515,"output_power":0.00019043689215309716,"reference_active":true,"delta_db":-0.06925933185316825},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.980555526095884e-06,"output_power":5.339688094248451e-06,"reference_active":true,"delta_db":-0.4922563747317204}],"active_band_count":3,"active_band_mae_db":0.24022845085340708}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.0003730035430310483,"log_rms_mae_db":1.5810854897487892,"rms_envelope_nrmse":0.08502188435995599,"envelope_correlation":0.9944008451792309,"rms_gain_error_db":-0.07978490041624792,"reference_mean_power":0.007423998211444209,"output_mean_power":0.007288855970301705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.725628531347034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002177971361103236,"output_power":0.002059948572601822,"reference_active":true,"delta_db":-0.24195786591469043},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032785256002961345,"output_power":0.00029321337348639807,"reference_active":true,"delta_db":-0.484948045664887},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.051887548388713e-05,"output_power":6.905269142062643e-05,"reference_active":true,"delta_db":-0.6671709050842912}],"active_band_count":3,"active_band_mae_db":0.46469227222128956}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00035212990311193784,"log_rms_mae_db":0.9814672788096163,"rms_envelope_nrmse":0.05116186448855444,"envelope_correlation":0.9979121261542527,"rms_gain_error_db":-0.07978586232329095,"reference_mean_power":0.007484242812710004,"output_mean_power":0.007348002285620127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.215922155621447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018646593945503165,"output_power":0.0017784984122262897,"reference_active":true,"delta_db":-0.20546031782674393},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001569482176698105,"output_power":0.0001494704054508928,"reference_active":true,"delta_db":-0.21201175561056387},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505413793795638e-06,"output_power":3.609214164700945e-06,"reference_active":true,"delta_db":-2.558622730798077}],"active_band_count":3,"active_band_mae_db":0.9920316014117949}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00013449969207959865,"log_rms_mae_db":1.7252752612853652,"rms_envelope_nrmse":0.1763711934729519,"envelope_correlation":0.9799948357993421,"rms_gain_error_db":-0.6223369102263737,"reference_mean_power":0.0006066736643199692,"output_mean_power":0.0005256799969033399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.452566559449028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001977891479666072,"output_power":0.000181760458625274,"reference_active":true,"delta_db":-0.3670304975800121},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.140914614473541e-05,"output_power":2.5897549393812358e-05,"reference_active":true,"delta_db":-0.8379746023428114},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0263986880205326e-05,"output_power":1.0524770680055088e-05,"reference_active":true,"delta_db":0.10896553708572693}],"active_band_count":3,"active_band_mae_db":0.43799021233618346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00012636239720129058,"log_rms_mae_db":1.1528797250901082,"rms_envelope_nrmse":0.11836933269113105,"envelope_correlation":0.9929376842996873,"rms_gain_error_db":-0.6223542388593982,"reference_mean_power":0.0006115927691385327,"output_mean_power":0.0005299402645662889},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.048532457224497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017739140051082302,"output_power":0.00016368425522136814,"reference_active":true,"delta_db":-0.34925655840269326},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4831454807134891e-05,"output_power":1.2876749604312182e-05,"reference_active":true,"delta_db":-0.6137750208851313},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.857972973983328e-07,"output_power":6.329241580672151e-07,"reference_active":true,"delta_db":1.1489657817058734}],"active_band_count":3,"active_band_mae_db":0.7039991203312326}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":2.0269882249774196,"rms_envelope_nrmse":0.3095447625593153,"envelope_correlation":0.9372830925386754,"rms_gain_error_db":0.33044938752807274,"reference_mean_power":9.493685130389179e-05,"output_mean_power":0.00010244240424348976},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.14467403295884e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.902158206464311e-05,"output_power":2.585836775359306e-05,"reference_active":true,"delta_db":-0.5011997597939487},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.98285507088674e-06,"output_power":1.6225509850180712e-05,"reference_active":true,"delta_db":2.1094358625727887},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.211971373682006e-06,"output_power":6.456934806288103e-06,"reference_active":true,"delta_db":3.0325473546306543}],"active_band_count":3,"active_band_mae_db":1.881060992332464}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":1.6359765516125797,"rms_envelope_nrmse":0.2617291188603766,"envelope_correlation":0.9527306300212934,"rms_gain_error_db":0.3304489728325767,"reference_mean_power":9.570694886911733e-05,"output_mean_power":0.00010327337460748609},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.518176746918327e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.649434521118165e-05,"output_power":2.385835556192181e-05,"reference_active":true,"delta_db":-0.45512684103115886},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515847029872691e-06,"output_power":8.782695488103945e-06,"reference_active":true,"delta_db":2.8888860306047963},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.652745912515814e-07,"output_power":6.429663202734729e-07,"reference_active":true,"delta_db":3.8449257037362417}],"active_band_count":3,"active_band_mae_db":2.396312858457399}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.8135781288146973,"predicted_mos":2.8135781288146973,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59680,"input_sample_rate":16000,"input_seconds":3.73,"input_float32_sha256":"8305051b1d4ba2d3ff8c11cadbb6be105212db41faac064997a5aca4bbd16aed","internal_samples":59680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.136215479999009},"started_at":"2026-10-13T04:41:42.692764+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.961099941487191e-05,"native_input_samples":59680,"native_input_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","output_sha256":"fa504dbb42c9456ef268a17267d471029839bf05feaa34863905dd786f5f49a4","output_samples":59680,"length_delta_samples":0,"analysis_s":0.84675239399985,"synthesis_s":0.06098396600009437,"total_s":0.9077363599999444,"rtf":0.24336095442357758,"output_peak":1.2619740184992332,"output_rms":0.17129676781235476,"finished_at":"2026-10-13T04:41:43.604252+00:00","evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"fa504dbb42c9456ef268a17267d471029839bf05feaa34863905dd786f5f49a4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.801380433431443,"spectral_convergence":0.6832860413541502,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.254164136531701,"spectral_convergence":0.3297981722313159,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.566677547222358,"spectral_convergence":0.34967145690706325,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.004172562362161585,"delta_db":-0.11532338123136858,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.019577328966320886,"delta_db":1.8882691507516935,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.004698115156127159,"delta_db":1.7933871508188775,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.0006354969660538367,"delta_db":0.5939664366272459,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00017013137381252273,"delta_db":0.2844156790736434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":8.894783848378674e-05,"delta_db":-0.12884740614410736,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8007015341078229,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.479092873243896,"pitch":null,"output_rms_dbfs":-15.325016632633993,"output_peak":1.2619740184992332,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.874074039061834},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.801380433431443,"spectral_convergence":0.6832860413541502,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.254164136531701,"spectral_convergence":0.3297981722313159,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.566677547222358,"spectral_convergence":0.34967145690706325,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.004172562362161585,"delta_db":-0.11532338123136858,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.019577328966320886,"delta_db":1.8882691507516935,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.004698115156127159,"delta_db":1.7933871508188775,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.0006354969660538367,"delta_db":0.5939664366272459,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00017013137381252273,"delta_db":0.2844156790736434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":8.894783848378674e-05,"delta_db":-0.12884740614410736,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8007015341078229,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.479092873243896,"pitch":null,"output_rms_dbfs":-15.325016632633993,"output_peak":1.2619740184992332,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.874074039061834},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.9314442356349313,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14447560597487225,"rms_dbfs":-16.80410950587789,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.859693585918194,"mean":-0.0023839802269002386,"zero_energy":false},"output":{"rms":0.17129676781235476,"rms_dbfs":-15.325016632633993,"peak":1.2619740184992332,"peak_dbfs":2.021008275047139,"crest_factor_db":17.34602490768113,"mean":2.9579323251134837e-08,"zero_energy":false},"rms_difference_db":1.4790928732438964},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":57,"at_or_above_full_scale_fraction":0.0009550938337801609,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0009718498659517426,"longest_full_scale_run_samples":4,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":59,"quiet_samples":9440,"reference_mean_power":6.490114235745396e-06,"output_mean_power":2.571862164250032e-06,"output_rms_dbfs":-55.89752310596061,"output_zero_energy":false,"output_peak":0.013956022038522272,"output_energy_fraction":1.3864127143503912e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":126,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9906542056074766,"absolute_difference_cents_median":6.34329326472529,"absolute_difference_cents_p95":1201.641700743341,"difference_over_600_cents_frames":8,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00043785897255075944,"log_rms_mae_db":2.8952334743583785,"rms_envelope_nrmse":0.25826277715765217,"envelope_correlation":0.9746230263119156,"rms_gain_error_db":1.4790934275966074,"reference_mean_power":0.0209012125553907,"output_mean_power":0.02938196415682869},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.005551294872497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00444581511703579,"output_power":0.007031250082920875,"reference_active":true,"delta_db":1.9908114667630934},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039454721092167534,"output_power":0.0005224307175701866,"reference_active":true,"delta_db":1.219297267312934},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010840168168493419,"output_power":0.0001555864154967346,"reference_active":true,"delta_db":1.5693565568678982}],"active_band_count":3,"active_band_mae_db":1.5931550969813086}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0004107294113879301,"log_rms_mae_db":2.451453320352856,"rms_envelope_nrmse":0.2239873182130403,"envelope_correlation":0.9876913833214257,"rms_gain_error_db":1.4791045417976996,"reference_mean_power":0.021070817398047652,"output_mean_power":0.02962046265374112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.012296407388438e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00391054354874771,"output_power":0.006271574668453142,"reference_active":true,"delta_db":2.051394706255244},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019349823972573515,"output_power":0.00024600921000430474,"reference_active":true,"delta_db":1.0427434781895186},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.980555526095884e-06,"output_power":8.536697938802585e-06,"reference_active":true,"delta_db":1.545483877679644}],"active_band_count":3,"active_band_mae_db":1.546540687374802}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.0003730035430310483,"log_rms_mae_db":2.0183686803095484,"rms_envelope_nrmse":0.29409682201521414,"envelope_correlation":0.9802342837456485,"rms_gain_error_db":1.822755834891305,"reference_mean_power":0.007423998211444209,"output_mean_power":0.011295707608817627},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.725628531347034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002177971361103236,"output_power":0.0029422702386374283,"reference_active":true,"delta_db":1.30630394090948},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032785256002961345,"output_power":0.00039430201405368554,"reference_active":true,"delta_db":0.8015041621511233},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.051887548388713e-05,"output_power":9.456585852696363e-05,"reference_active":true,"delta_db":0.6983666874494389}],"active_band_count":3,"active_band_mae_db":0.9353915968366807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00035212990311193784,"log_rms_mae_db":1.5894065976042553,"rms_envelope_nrmse":0.2664450232481233,"envelope_correlation":0.9895269742387537,"rms_gain_error_db":1.8227616644809073,"reference_mean_power":0.007484242812710004,"output_mean_power":0.01138738582053914},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.215922155621447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018646593945503165,"output_power":0.0024759836834146883,"reference_active":true,"delta_db":1.2314826479914085},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001569482176698105,"output_power":0.00018756059347130972,"reference_active":true,"delta_db":0.7738521010189863},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505413793795638e-06,"output_power":7.5651686626438e-06,"reference_active":true,"delta_db":0.6554368869398504}],"active_band_count":3,"active_band_mae_db":0.8869238786500816}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00013449969207959865,"log_rms_mae_db":1.4504131677708045,"rms_envelope_nrmse":0.1744139878723145,"envelope_correlation":0.9831552080531478,"rms_gain_error_db":0.606738112011336,"reference_mean_power":0.0006066736643199692,"output_mean_power":0.0006976361132741835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.452566559449028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001977891479666072,"output_power":0.00022711055519561603,"reference_active":true,"delta_db":0.6003485935000885},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.140914614473541e-05,"output_power":3.739617678659032e-05,"reference_active":true,"delta_db":0.7577107389990011},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0263986880205326e-05,"output_power":1.0350801351868003e-05,"reference_active":true,"delta_db":0.03657885635864957}],"active_band_count":3,"active_band_mae_db":0.464879396285913}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00012636239720129058,"log_rms_mae_db":0.8224267932755006,"rms_envelope_nrmse":0.0972685652259974,"envelope_correlation":0.9969984103053612,"rms_gain_error_db":0.6067367509334727,"reference_mean_power":0.0006115927691385327,"output_mean_power":0.0007032925504169642},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.048532457224497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017739140051082302,"output_power":0.00020485653843915755,"reference_active":true,"delta_db":0.625172674561244},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4831454807134891e-05,"output_power":1.785203275721856e-05,"reference_active":true,"delta_db":0.8050392230601583},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.857972973983328e-07,"output_power":5.489035764508004e-07,"reference_active":true,"delta_db":0.530409660954295}],"active_band_count":3,"active_band_mae_db":0.6535405195252325}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":1.5869177721102095,"rms_envelope_nrmse":0.20284682842690677,"envelope_correlation":0.9714022736431895,"rms_gain_error_db":-0.25023523225753497,"reference_mean_power":9.493685130389179e-05,"output_mean_power":8.962131263301517e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.14467403295884e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.902158206464311e-05,"output_power":2.4325579140440796e-05,"reference_active":true,"delta_db":-0.7665789491495798},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.98285507088674e-06,"output_power":8.533461559235947e-06,"reference_active":true,"delta_db":-0.681295297001248},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.211971373682006e-06,"output_power":3.3087015707124256e-06,"reference_active":true,"delta_db":0.12885931585589455}],"active_band_count":3,"active_band_mae_db":0.5255778540022408}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":0.840150537359709,"rms_envelope_nrmse":0.0978116893314641,"envelope_correlation":0.9937106927912119,"rms_gain_error_db":-0.25022745885554737,"reference_mean_power":9.570694886911733e-05,"output_mean_power":9.034845395173751e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.518176746918327e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.649434521118165e-05,"output_power":2.277074744251906e-05,"reference_active":true,"delta_db":-0.6577590427814466},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515847029872691e-06,"output_power":3.984546748755821e-06,"reference_active":true,"delta_db":-0.5436029562172415},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.652745912515814e-07,"output_power":2.1653304526048172e-07,"reference_active":true,"delta_db":-0.8817147029030782}],"active_band_count":3,"active_band_mae_db":0.694358900633922}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"facodec","status":"measured","mos":{"status":"measured","value":4.041409015655518,"predicted_mos":4.041409015655518,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59800,"input_sample_rate":16000,"input_seconds":3.7375,"input_float32_sha256":"5a534ade2cd80de177d444b0d62b1626aa5e43e5ed3cbc653219b66d1474bc13","internal_samples":59800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.738314820999221},"started_at":"2026-10-13T04:49:35.527879+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.362300074310042e-05,"native_input_samples":59680,"native_input_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","output_sha256":"a7a071fa96de259ba2d02ec241b1d51413ae3bf1f6efb2a37062d5fff2aae1c0","output_samples":59800,"length_delta_samples":120,"analysis_s":0.8087900959999388,"synthesis_s":2.6315507099998285,"total_s":3.4403408059997673,"rtf":0.9223433796246026,"output_peak":0.7509177327156067,"output_rms":0.1356589261015111,"finished_at":"2026-10-13T04:49:38.972713+00:00","evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"a7a071fa96de259ba2d02ec241b1d51413ae3bf1f6efb2a37062d5fff2aae1c0","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":59680,"output_samples":59800,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-17.351032504086284,"output_peak":0.7509177327156067},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9979933110367893,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.629368281673297,"spectral_convergence":0.2265875980200473,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.832772732457654,"spectral_convergence":0.220377422821676,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.760579206610901,"spectral_convergence":0.23173408420704628,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.003995974883067074,"delta_db":-0.3031242460954759,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.011359204991550357,"delta_db":-0.4757858642754632,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.0023888394570703536,"delta_db":-1.1439797874725337,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.00046681161987533295,"delta_db":-0.7457517988664892,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.0001488883909723778,"delta_db":-0.29482004495912506,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":8.062891809623428e-05,"delta_db":-0.5552930705707567,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5864591353733072,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5381993062071775,"pitch":null,"output_rms_dbfs":-17.342308812085065,"output_peak":0.7509177327156067,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.740906740247283},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.9068037394668138,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14447560597487225,"rms_dbfs":-16.80410950587789,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.859693585918194,"mean":-0.0023839802269002386,"zero_energy":false},"output":{"rms":0.13579524388074765,"rms_dbfs":-17.342308812085065,"peak":0.7509177327156067,"peak_dbfs":-2.488152796126923,"crest_factor_db":14.854156015958143,"mean":0.0002609573151508463,"zero_energy":false},"rms_difference_db":-0.5381993062071757},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":59,"quiet_samples":9440,"reference_mean_power":6.490114235745396e-06,"output_mean_power":3.7240821753000744e-06,"output_rms_dbfs":-54.28980744477585,"output_zero_energy":false,"output_peak":0.019980065524578094,"output_energy_fraction":3.194429564181315e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":120,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9626168224299065,"absolute_difference_cents_median":3.5420402110225857,"absolute_difference_cents_p95":7.728818369254741,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00043785897255075944,"log_rms_mae_db":2.147153491584484,"rms_envelope_nrmse":0.11518426259883921,"envelope_correlation":0.9896969287065649,"rms_gain_error_db":-0.5381982573677403,"reference_mean_power":0.0209012125553907,"output_mean_power":0.01846509966581967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.005551294872497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00444581511703579,"output_power":0.0037504511125384574,"reference_active":true,"delta_db":-0.7386788946831642},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039454721092167534,"output_power":0.0003952839199216954,"reference_active":true,"delta_db":0.00810170006648437},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010840168168493419,"output_power":0.00010492950866862419,"reference_active":true,"delta_db":-0.14138380376100845}],"active_band_count":3,"active_band_mae_db":0.29605479950355235}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0004107294113879301,"log_rms_mae_db":1.7950696926368224,"rms_envelope_nrmse":0.1044866652932231,"envelope_correlation":0.9918373832140853,"rms_gain_error_db":-0.5381833922190726,"reference_mean_power":0.021070817398047652,"output_mean_power":0.018615000158576638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.012296407388438e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00391054354874771,"output_power":0.003306905717396481,"reference_active":true,"delta_db":-0.7281531366098477},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019349823972573515,"output_power":0.0001956301539649096,"reference_active":true,"delta_db":0.047587781764820744},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.980555526095884e-06,"output_power":6.3266456733773935e-06,"reference_active":true,"delta_db":0.24431985366485037}],"active_band_count":3,"active_band_mae_db":0.34002025734650626}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.0003730035430310483,"log_rms_mae_db":2.044054302199916,"rms_envelope_nrmse":0.13901550875688043,"envelope_correlation":0.9867708755283581,"rms_gain_error_db":-0.5243140854886161,"reference_mean_power":0.007423998211444209,"output_mean_power":0.006579705416717358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.725628531347034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002177971361103236,"output_power":0.0020720670928434247,"reference_active":true,"delta_db":-0.2164835115756891},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032785256002961345,"output_power":0.0003609618039580857,"reference_active":true,"delta_db":0.41782669226424507},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.051887548388713e-05,"output_power":7.541810192405936e-05,"reference_active":true,"delta_db":-0.28422102805485105}],"active_band_count":3,"active_band_mae_db":0.30617707729826177}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00035212990311193784,"log_rms_mae_db":1.7271414507696272,"rms_envelope_nrmse":0.12618131412702385,"envelope_correlation":0.9890249027325729,"rms_gain_error_db":-0.5242975217726491,"reference_mean_power":0.007484242812710004,"output_mean_power":0.006633124010345997},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.215922155621447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018646593945503165,"output_power":0.0017927676617367015,"reference_active":true,"delta_db":-0.17075503867324393},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001569482176698105,"output_power":0.00016842503789543016,"reference_active":true,"delta_db":0.3065026559711318},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505413793795638e-06,"output_power":6.306659570190659e-06,"reference_active":true,"delta_db":-0.13475537545472893}],"active_band_count":3,"active_band_mae_db":0.20400435669970154}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00013449969207959865,"log_rms_mae_db":1.8767090663942836,"rms_envelope_nrmse":0.17065444319734444,"envelope_correlation":0.9829713427317646,"rms_gain_error_db":-0.7810686689455498,"reference_mean_power":0.0006066736643199692,"output_mean_power":0.0005068136179550639},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.452566559449028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001977891479666072,"output_power":0.00017968544787420543,"reference_active":true,"delta_db":-0.41689553149641534},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.140914614473541e-05,"output_power":3.205948271225324e-05,"reference_active":true,"delta_db":0.08900380328609428},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0263986880205326e-05,"output_power":1.2712175283619001e-05,"reference_active":true,"delta_db":0.9290378433472264}],"active_band_count":3,"active_band_mae_db":0.478312392709912}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00012636239720129058,"log_rms_mae_db":1.5499758491349769,"rms_envelope_nrmse":0.14631790832573863,"envelope_correlation":0.988506169407757,"rms_gain_error_db":-0.7810302910793849,"reference_mean_power":0.0006115927691385327,"output_mean_power":0.0005109275404216339},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.048532457224497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017739140051082302,"output_power":0.00016155283715837772,"reference_active":true,"delta_db":-0.4061797308840369},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4831454807134891e-05,"output_power":1.5539428350179767e-05,"reference_active":true,"delta_db":0.20251285580452932},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.857972973983328e-07,"output_power":7.183733316000519e-07,"reference_active":true,"delta_db":1.6989510711123286}],"active_band_count":3,"active_band_mae_db":0.7692145526002983}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":2.2035659678153263,"rms_envelope_nrmse":0.18742120000398046,"envelope_correlation":0.9768415639172287,"rms_gain_error_db":-0.5361392846953759,"reference_mean_power":9.493685130389179e-05,"output_mean_power":8.391138612950434e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.14467403295884e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.902158206464311e-05,"output_power":2.5116565428812927e-05,"reference_active":true,"delta_db":-0.6276083219396666},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.98285507088674e-06,"output_power":8.036216745561797e-06,"reference_active":true,"delta_db":-0.9420312452585493},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.211971373682006e-06,"output_power":2.2023628773254633e-06,"reference_active":true,"delta_db":-1.6388278797626579}],"active_band_count":3,"active_band_mae_db":1.069489148986958}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":1.7279641199921356,"rms_envelope_nrmse":0.14208766517910154,"envelope_correlation":0.9871360916374344,"rms_gain_error_db":-0.5361177078884883,"reference_mean_power":9.570694886911733e-05,"output_mean_power":8.459246890024664e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.518176746918327e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.649434521118165e-05,"output_power":2.259169033276649e-05,"reference_active":true,"delta_db":-0.6920446422341247},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515847029872691e-06,"output_power":3.409413481762383e-06,"reference_active":true,"delta_db":-1.2205954803826953},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.652745912515814e-07,"output_power":2.1523305932993003e-07,"reference_active":true,"delta_db":-0.907866752182013}],"active_band_count":3,"active_band_mae_db":0.9401689582662777}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8627660274505615,"predicted_mos":3.8627660274505615,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":164864,"input_sample_rate":44100,"input_seconds":3.7384126984126986,"input_float32_sha256":"5d6155a1fd95d165731386ada77d0a4fdeb460709d0c0414b9331d0127fba049","internal_samples":59815,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3628238210003474},"started_at":"2026-10-13T07:08:55.566643+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004375505001007696,"native_input_samples":164493,"native_input_sha256":"af6514e62003faee485ca9b6090c0b8468f5bde63e1decaf04f5f015a2c3329e","output_sha256":"0ff742e8662d3e8b0c9f1250e3b1826203417b1da7d3891677fce5af8707caeb","output_samples":164864,"length_delta_samples":371,"analysis_s":0.08906596199994965,"synthesis_s":2.5160279920000903,"total_s":2.60509395400004,"rtf":0.6984166096514852,"output_peak":1.1357290744781494,"output_rms":0.12926961122973893,"finished_at":"2026-10-13T07:08:58.187414+00:00","evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"0ff742e8662d3e8b0c9f1250e3b1826203417b1da7d3891677fce5af8707caeb","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":59680,"output_samples":59815,"length_delta_samples":135,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-17.765849181453266,"output_peak":1.1175062295814426},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9977430410432165,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.98960372599307,"spectral_convergence":0.3996534245160028,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.870217889818965,"spectral_convergence":0.32463394853701905,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.50752831869144,"spectral_convergence":0.3619413391775709,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.0029966872289539687,"delta_db":-1.552937573783725,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.010608681426445197,"delta_db":-0.772651154946222,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.002411622786561692,"delta_db":-1.1027556383410315,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.0005058125042317812,"delta_db":-0.3972727648316393,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00018066835847671935,"delta_db":0.545392576024363,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":6.124749744085091e-05,"delta_db":-1.7493179117484714,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0200546032792424,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9519275018684699,"pitch":null,"output_rms_dbfs":-17.756037007746357,"output_peak":1.1175062295814426,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.789116644834493},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.7207283094359185,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14447560597487225,"rms_dbfs":-16.80410950587789,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.859693585918194,"mean":-0.0023839802269002386,"zero_energy":false},"output":{"rms":0.1294786461240239,"rms_dbfs":-17.756037007746357,"peak":1.1175062295814426,"peak_dbfs":0.9649990560401834,"crest_factor_db":18.721036063786542,"mean":-0.00016722354386359146,"zero_energy":false},"rms_difference_db":-0.9519275018684681},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":4,"at_or_above_full_scale_fraction":6.702412868632708e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":6.702412868632708e-05,"longest_full_scale_run_samples":2,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":59,"quiet_samples":9440,"reference_mean_power":6.490114235745396e-06,"output_mean_power":7.565212848492565e-06,"output_rms_dbfs":-51.21178848523095,"output_zero_energy":false,"output_peak":0.03274109918321978,"output_energy_fraction":7.137860107043799e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":114,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.8691588785046729,"absolute_difference_cents_median":10.009408901614625,"absolute_difference_cents_p95":40.851611489307636,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00043785897255075944,"log_rms_mae_db":2.344073364246791,"rms_envelope_nrmse":0.14486002993883487,"envelope_correlation":0.9892306859331451,"rms_gain_error_db":-0.9519274613028332,"reference_mean_power":0.0209012125553907,"output_mean_power":0.016787218211453788},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.005551294872497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00444581511703579,"output_power":0.0038490816266874154,"reference_active":true,"delta_db":-0.6259427705029501},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039454721092167534,"output_power":0.0003441088051888462,"reference_active":true,"delta_db":-0.5940319201691028},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010840168168493419,"output_power":0.0001014277073168374,"reference_active":true,"delta_db":-0.28879410908155895}],"active_band_count":3,"active_band_mae_db":0.502922933251204}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0004107294113879301,"log_rms_mae_db":2.0640565471123797,"rms_envelope_nrmse":0.12675531182685876,"envelope_correlation":0.9942930299383879,"rms_gain_error_db":-0.9519200078873647,"reference_mean_power":0.021070817398047652,"output_mean_power":0.01692346870676383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.012296407388438e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00391054354874771,"output_power":0.003437429448201696,"reference_active":true,"delta_db":-0.5600333344047447},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019349823972573515,"output_power":0.00015892956327253588,"reference_active":true,"delta_db":-0.8547232856717637},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.980555526095884e-06,"output_power":4.775028576831893e-06,"reference_active":true,"delta_db":-0.9776555189279124}],"active_band_count":3,"active_band_mae_db":0.7974707130014735}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.0003730035430310483,"log_rms_mae_db":2.6663898888628235,"rms_envelope_nrmse":0.19250499294500356,"envelope_correlation":0.9710579892788501,"rms_gain_error_db":-0.44670225250475126,"reference_mean_power":0.007423998211444209,"output_mean_power":0.006698346867818512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.725628531347034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002177971361103236,"output_power":0.0018382100319800567,"reference_active":true,"delta_db":-0.7365703284511516},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032785256002961345,"output_power":0.0002647326240396038,"reference_active":true,"delta_db":-0.9287111478307195},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.051887548388713e-05,"output_power":7.149139040371544e-05,"reference_active":true,"delta_db":-0.5164395741368522}],"active_band_count":3,"active_band_mae_db":0.7272403501395743}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00035212990311193784,"log_rms_mae_db":2.0161919583043004,"rms_envelope_nrmse":0.16581842577535982,"envelope_correlation":0.9775403397730069,"rms_gain_error_db":-0.44669484351703714,"reference_mean_power":0.007484242812710004,"output_mean_power":0.0067527144402740515},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.215922155621447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018646593945503165,"output_power":0.0016118704993157904,"reference_active":true,"delta_db":-0.632693667421206},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001569482176698105,"output_power":0.00011827957560863535,"reference_active":true,"delta_db":-1.2284663055066867},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505413793795638e-06,"output_power":4.280779199932573e-06,"reference_active":true,"delta_db":-1.8175209844972064}],"active_band_count":3,"active_band_mae_db":1.2262269858083663}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00013449969207959865,"log_rms_mae_db":3.5215976203880377,"rms_envelope_nrmse":0.32995736813293985,"envelope_correlation":0.9279551885325605,"rms_gain_error_db":-0.27526090032589595,"reference_mean_power":0.0006066736643199692,"output_mean_power":0.0005694151955005109},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.452566559449028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001977891479666072,"output_power":0.0002025478826042777,"reference_active":true,"delta_db":0.10325247894700337},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.140914614473541e-05,"output_power":2.3375469998419224e-05,"reference_active":true,"delta_db":-1.282957785306613},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0263986880205326e-05,"output_power":7.874015980829213e-06,"reference_active":true,"delta_db":-1.1511979635241256}],"active_band_count":3,"active_band_mae_db":0.8458027425925806}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00012636239720129058,"log_rms_mae_db":3.1764291142765546,"rms_envelope_nrmse":0.2874204943487684,"envelope_correlation":0.9440033106414396,"rms_gain_error_db":-0.27523900268021717,"reference_mean_power":0.0006115927691385327,"output_mean_power":0.0005740350910396352},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.048532457224497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017739140051082302,"output_power":0.00018767174375269996,"reference_active":true,"delta_db":0.24466326763433352},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4831454807134891e-05,"output_power":1.1033803437912153e-05,"reference_active":true,"delta_db":-1.2845850982691178},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.857972973983328e-07,"output_power":5.017482312921291e-07,"reference_active":true,"delta_db":0.14030755838787007}],"active_band_count":3,"active_band_mae_db":0.5565186414304405}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":3.9969113701064267,"rms_envelope_nrmse":0.29611211506745727,"envelope_correlation":0.9416408988355585,"rms_gain_error_db":-1.0082790717885883,"reference_mean_power":9.493685130389179e-05,"output_mean_power":7.526740039845918e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.14467403295884e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.902158206464311e-05,"output_power":2.6593118327202848e-05,"reference_active":true,"delta_db":-0.3795181762392516},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.98285507088674e-06,"output_power":8.412744174048181e-06,"reference_active":true,"delta_db":-0.7431708364734346},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.211971373682006e-06,"output_power":3.127261566070445e-06,"reference_active":true,"delta_db":-0.11607458607012804}],"active_band_count":3,"active_band_mae_db":0.4129211995942714}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":3.774812715508831,"rms_envelope_nrmse":0.2545582562073141,"envelope_correlation":0.9568674313152424,"rms_gain_error_db":-1.0082571741847703,"reference_mean_power":9.570694886911733e-05,"output_mean_power":7.587832821511059e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.518176746918327e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.649434521118165e-05,"output_power":2.3438900437266465e-05,"reference_active":true,"delta_db":-0.5321595644895336},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515847029872691e-06,"output_power":4.187698085006332e-06,"reference_active":true,"delta_db":-0.3276385880556488},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.652745912515814e-07,"output_power":2.410434539218862e-07,"reference_active":true,"delta_db":-0.41600312290250985}],"active_band_count":3,"active_band_mae_db":0.4252670918158974}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"golf","status":"measured","mos":{"status":"measured","value":2.9400429725646973,"predicted_mos":2.9400429725646973,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89280,"input_sample_rate":24000,"input_seconds":3.72,"input_float32_sha256":"53a5c7fba80731d240cdd0eeea9b5bbaa08563dc4d57724dc1ac97ca4b99cf6b","internal_samples":59520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5344803190000675},"started_at":"2026-10-13T05:02:17.652045+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018344439995416906,"native_input_samples":89520,"native_input_sha256":"742f19a2c09138bb13a28e286ef7cc75438d1b7f570a24981e81691740d064c3","output_sha256":"d92650335c75c44d89bdacc8faa2269308dc3bbcd930201280b798c7a6ab67d2","output_samples":89280,"length_delta_samples":-240,"analysis_s":0.4058423900005437,"synthesis_s":0.114755277999393,"total_s":0.5205976679999367,"rtf":0.13957042037531814,"output_peak":1.0612602233886719,"output_rms":0.13338739243358924,"finished_at":"2026-10-13T05:02:18.180245+00:00","analysis_stages":{"external_f0_s":0.03791707399977895},"evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"d92650335c75c44d89bdacc8faa2269308dc3bbcd930201280b798c7a6ab67d2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59680,"output_samples":59520,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-17.49504998919855,"output_peak":1.0548247647151183},"common_prefix":{"samples":59520,"sample_rate":16000,"reference_coverage":0.9973190348525469,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59520,"output_samples":59520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.698698684577994,"spectral_convergence":0.633021973083192,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.216369897801338,"spectral_convergence":0.30269708052803196,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.3526253909926185,"spectral_convergence":0.30900551435348594,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004296358181966894,"output_power":0.004180076348575885,"delta_db":-0.11916266602136272,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012709991254280337,"output_power":0.011131933787985982,"delta_db":-0.5757463720242622,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.003115830997008452,"output_power":0.0019240236523620465,"delta_db":-2.0936348686902746,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005556446396243708,"output_power":0.00037705747553522825,"delta_db":-1.68389573484247,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015961043062438372,"output_power":0.00010989877438142323,"delta_db":-1.6206842022167167,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.186909125166661e-05,"output_power":8.008412732991009e-05,"delta_db":-0.5962297298497667,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1148922622741422,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7025979746028936,"pitch":null,"output_rms_dbfs":-17.49504998919855,"output_peak":1.0548247647151183,"reference_rms_dbfs":-16.792452014595657,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.755897991123984},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59520,"output_samples":59520,"intelligibility":{"estoi":{"value":0.8901591949534864,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14466963950586212,"rms_dbfs":-16.792452014595657,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.84803609463596,"mean":-0.0023883973398516256,"zero_energy":false},"output":{"rms":0.13342816106868324,"rms_dbfs":-17.49504998919855,"peak":1.0548247647151183,"peak_dbfs":0.4636063483450392,"crest_factor_db":17.95865633754359,"mean":0.0014312989562109805,"zero_energy":false},"rms_difference_db":-0.7025979746028916},"full_scale":{"reference":{"sample_count":59520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":4,"at_or_above_full_scale_fraction":6.720430107526882e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":6.720430107526882e-05,"longest_full_scale_run_samples":2,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":372,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":6.557875092077101e-06,"output_mean_power":7.992231707311386e-06,"output_rms_dbfs":-50.97331933722509,"output_zero_energy":false,"output_peak":0.01161923869970105,"output_energy_fraction":6.999356576886286e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":365,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":150,"both_estimated_voiced_frames":107,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":43,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":5.245870365112419,"absolute_difference_cents_p95":21.949784134495914,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00043785897255075944,"log_rms_mae_db":2.147794552846627,"rms_envelope_nrmse":0.16665469891971502,"envelope_correlation":0.976171533480209,"rms_gain_error_db":-0.7025974242210432,"reference_mean_power":0.020957465473323382,"output_mean_power":0.01782703088634622},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.006615434226001e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.004459299233401496,"output_power":0.004049442644811891,"reference_active":true,"delta_db":-0.41871363773431114},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039350311445259575,"output_power":0.0004146499184577232,"reference_active":true,"delta_db":0.22733410300859844},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010750126271492884,"output_power":0.00013691350749473047,"reference_active":true,"delta_db":1.0503273097291708}],"active_band_count":3,"active_band_mae_db":0.5654583501573601}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":0.0004107294113879301,"log_rms_mae_db":1.616216592115822,"rms_envelope_nrmse":0.11386498728976913,"envelope_correlation":0.9919001564679948,"rms_gain_error_db":-0.7025979033987123,"reference_mean_power":0.021127972469740217,"output_mean_power":0.017972067123569777},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":8.008920711213039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0039212748700043305,"output_power":0.003636463026745599,"reference_active":true,"delta_db":-0.32748109853315693},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019193519395315032,"output_power":0.00020091377137952348,"reference_active":true,"delta_db":0.19855090039079692},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9412903601338735e-06,"output_power":6.747485350413151e-06,"reference_active":true,"delta_db":0.5526117278043285}],"active_band_count":3,"active_band_mae_db":0.3595479089094275}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.0003730035430310483,"log_rms_mae_db":1.816191049531399,"rms_envelope_nrmse":0.18784447212828834,"envelope_correlation":0.9731249131916037,"rms_gain_error_db":-0.5287940086612621,"reference_mean_power":0.007443979503954865,"output_mean_power":0.006590612349742039},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":3.745651405174024e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0021931461808536888,"output_power":0.0018941521285462501,"reference_active":true,"delta_db":-0.6365272352180793},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032867398116542774,"output_power":0.0003345937924134627,"reference_active":true,"delta_db":0.07752554040157944},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.066577156866226e-05,"output_power":7.957585293566931e-05,"reference_active":true,"delta_db":-0.05907989877466194}],"active_band_count":3,"active_band_mae_db":0.25771089146477355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":0.00035212990311193784,"log_rms_mae_db":1.3193124419679558,"rms_envelope_nrmse":0.1325269949831372,"envelope_correlation":0.9874282291256774,"rms_gain_error_db":-0.5287894278140168,"reference_mean_power":0.007504545557157891,"output_mean_power":0.006644242205510534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":3.232389127378406e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018768390664311296,"output_power":0.0016470293735330868,"reference_active":true,"delta_db":-0.5672569015613751},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001575381547000851,"output_power":0.00016670548276722112,"reference_active":true,"delta_db":0.24564129430818069},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489797008550677e-06,"output_power":6.197719444631542e-06,"reference_active":true,"delta_db":-0.19999199995028744}],"active_band_count":3,"active_band_mae_db":0.3376300652732811}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00013449969207959865,"log_rms_mae_db":1.812708892331271,"rms_envelope_nrmse":0.24179192192354984,"envelope_correlation":0.9773243348899078,"rms_gain_error_db":-1.7211315418417361,"reference_mean_power":0.0006083059385170187,"output_mean_power":0.0004092690487291038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":3.470471650164028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019907940093928571,"output_power":0.0001250437654143331,"reference_active":true,"delta_db":-2.0196428219614373},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.169121410085626e-05,"output_power":1.582606498209352e-05,"reference_active":true,"delta_db":-3.0156593285801536},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.016385066470989e-05,"output_power":8.33203774203008e-06,"reference_active":true,"delta_db":-0.8630704690434949}],"active_band_count":3,"active_band_mae_db":1.9661242065283622}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":0.00012636239720129058,"log_rms_mae_db":1.3567696959678086,"rms_envelope_nrmse":0.2003948379350252,"envelope_correlation":0.9932033867608578,"rms_gain_error_db":-1.7211301866353805,"reference_mean_power":0.0006132512452294604,"output_mean_power":0.00041259638644679947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":3.0638277619783516e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001785270651712335,"output_power":0.00011108280866635663,"reference_active":true,"delta_db":-2.0605721351297883},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.492472581590345e-05,"output_power":7.466712811465751e-06,"reference_active":true,"delta_db":-3.0077691380118954},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.863579093174827e-07,"output_power":4.110916221047999e-07,"reference_active":true,"delta_db":-0.730173568036391}],"active_band_count":3,"active_band_mae_db":1.9328382803926916}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":2.373805096017096,"rms_envelope_nrmse":0.20477323154751964,"envelope_correlation":0.9732762948922621,"rms_gain_error_db":-0.3061399941219934,"reference_mean_power":9.519234452783782e-05,"output_mean_power":8.871315482537686e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":9.187189860194911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9068490159862067e-05,"output_power":3.1937484103690724e-05,"reference_active":true,"delta_db":0.4087822655574286},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0059714207120673e-05,"output_power":1.0999251498033817e-05,"reference_active":true,"delta_db":0.38777489565277745},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1944519043892312e-06,"output_power":2.5108796103992955e-06,"reference_active":true,"delta_db":-1.0457046377171249}],"active_band_count":3,"active_band_mae_db":0.6140872663091103}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59520,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":2.1114857676870895,"rms_envelope_nrmse":0.15218953471218005,"envelope_correlation":0.9858820793568699,"rms_gain_error_db":-0.3061302640912965,"reference_mean_power":9.596656736821234e-05,"output_mean_power":8.943488118791772e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":8.557251049668845e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.65288483079175e-05,"output_power":3.0208149335089023e-05,"reference_active":true,"delta_db":0.5640572312995352},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.539148854654746e-06,"output_power":4.883875149850105e-06,"reference_active":true,"delta_db":0.31790128200568896},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.644351361080623e-07,"output_power":2.2342671020967812e-07,"reference_active":true,"delta_db":-0.7318406945284528}],"active_band_count":3,"active_band_mae_db":0.5379330692778923}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.893915891647339,"predicted_mos":2.893915891647339,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89640,"input_sample_rate":24000,"input_seconds":3.735,"input_float32_sha256":"c4ce4ddd142159cd8f8d1be003520f22592a7b23981cba00d4bc28dce0cb456c","internal_samples":59760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8813847429992165},"started_at":"2026-10-13T05:08:16.281740+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002665435000380967,"native_input_samples":89520,"native_input_sha256":"742f19a2c09138bb13a28e286ef7cc75438d1b7f570a24981e81691740d064c3","output_sha256":"35edc172c81a7ead461f1c37e06cdb564faaf2afface8bc2950565f0b82de826","output_samples":89640,"length_delta_samples":120,"analysis_s":1.2883399020001889,"synthesis_s":9.289346784000372,"total_s":10.577686686000561,"rtf":2.835840934584601,"output_peak":0.8902040123939514,"output_rms":0.14047521830365164,"finished_at":"2026-10-13T05:08:26.870314+00:00","analysis_stages":{"external_f0_s":0.8131186119999256},"evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"35edc172c81a7ead461f1c37e06cdb564faaf2afface8bc2950565f0b82de826","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59680,"output_samples":59760,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-17.045043107921266,"output_peak":0.8912093144425799},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998661311914324,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.549051523958453,"spectral_convergence":0.6283791543978342,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.1656444780581205,"spectral_convergence":0.26495343461894943,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.050472403721361,"spectral_convergence":0.2729544040118422,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.005043269334478314,"delta_db":0.7077698777276321,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.011390924402130685,"delta_db":-0.46367553850792753,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.0026354325192350053,"delta_db":-0.717330420419195,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.0004888687240030074,"delta_db":-0.5452458424125721,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00014790352009733001,"delta_db":-0.32364330735054203,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":6.68228217719013e-05,"delta_db":-1.3709532846549013,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.688103045178795,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2351162196593706,"pitch":null,"output_rms_dbfs":-17.03922572553726,"output_peak":0.8912093144425799,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.588389468579312},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.8993865650182852,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14447560597487225,"rms_dbfs":-16.80410950587789,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.859693585918194,"mean":-0.0023839802269002386,"zero_energy":false},"output":{"rms":0.14061728671012053,"rms_dbfs":-17.03922572553726,"peak":0.8912093144425799,"peak_dbfs":-1.0004056626646647,"crest_factor_db":16.038820062872595,"mean":-1.239608108180785e-05,"zero_energy":false},"rms_difference_db":-0.23511621965937124},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":59,"quiet_samples":9440,"reference_mean_power":6.490114235745396e-06,"output_mean_power":1.4806541776548856e-06,"output_rms_dbfs":-58.29546363675899,"output_zero_energy":false,"output_peak":0.021393361956939992,"output_energy_fraction":1.1844572449126945e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":113,"both_estimated_voiced_frames":99,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9252336448598131,"absolute_difference_cents_median":5.852598604446105,"absolute_difference_cents_p95":18.25291213030702,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00043785897255075944,"log_rms_mae_db":3.4539797967494925,"rms_envelope_nrmse":0.15125758132268063,"envelope_correlation":0.976815747599249,"rms_gain_error_db":-0.2351160232388801,"reference_mean_power":0.0209012125553907,"output_mean_power":0.019799757878532834},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.005551294872497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00444581511703579,"output_power":0.004269767135823754,"reference_active":true,"delta_db":-0.17547208079400917},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039454721092167534,"output_power":0.00033375536005461484,"reference_active":true,"delta_db":-0.7267072855021406},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010840168168493419,"output_power":9.158959293228733e-05,"reference_active":true,"delta_db":-0.7318989085244886}],"active_band_count":3,"active_band_mae_db":0.5446927582735461}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0004107294113879301,"log_rms_mae_db":2.8244653761744147,"rms_envelope_nrmse":0.0847712264127667,"envelope_correlation":0.9927737962903818,"rms_gain_error_db":-0.23510304635433854,"reference_mean_power":0.021070817398047652,"output_mean_power":0.019960484506874054},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.012296407388438e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00391054354874771,"output_power":0.0037960848079152433,"reference_active":true,"delta_db":-0.12901220268973068},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019349823972573515,"output_power":0.00016615552807833664,"reference_active":true,"delta_db":-0.6616222349750749},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.980555526095884e-06,"output_power":4.657502014965434e-06,"reference_active":true,"delta_db":-1.0858847546066328}],"active_band_count":3,"active_band_mae_db":0.6255063974238128}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.0003730035430310483,"log_rms_mae_db":2.1990342343667657,"rms_envelope_nrmse":0.17461473634082916,"envelope_correlation":0.9767168056303185,"rms_gain_error_db":-0.4602874910407041,"reference_mean_power":0.007423998211444209,"output_mean_power":0.006677426394525602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.725628531347034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002177971361103236,"output_power":0.001957798661442838,"reference_active":true,"delta_db":-0.4628413752712387},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032785256002961345,"output_power":0.0002780365096775506,"reference_active":true,"delta_db":-0.7157675122492473},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.051887548388713e-05,"output_power":6.685113809040115e-05,"reference_active":true,"delta_db":-0.8078889576067095}],"active_band_count":3,"active_band_mae_db":0.662165948375732}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00035212990311193784,"log_rms_mae_db":1.5245650816274416,"rms_envelope_nrmse":0.11370899717515068,"envelope_correlation":0.9907785738330782,"rms_gain_error_db":-0.4602799583832983,"reference_mean_power":0.007484242812710004,"output_mean_power":0.0067316243562748415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.215922155621447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018646593945503165,"output_power":0.0016860086993176946,"reference_active":true,"delta_db":-0.43739702455010304},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001569482176698105,"output_power":0.00013570742374265841,"reference_active":true,"delta_db":-0.6315278217642841},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505413793795638e-06,"output_power":5.874650783617679e-06,"reference_active":true,"delta_db":-0.44292870967667064}],"active_band_count":3,"active_band_mae_db":0.5039511853303527}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00013449969207959865,"log_rms_mae_db":2.07605455301858,"rms_envelope_nrmse":0.2065184356252662,"envelope_correlation":0.9698794392937787,"rms_gain_error_db":-0.49265751884332737,"reference_mean_power":0.0006066736643199692,"output_mean_power":0.0005416133876766218},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.452566559449028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001977891479666072,"output_power":0.0001816704445769716,"reference_active":true,"delta_db":-0.36918180639422826},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.140914614473541e-05,"output_power":2.7805695206520134e-05,"reference_active":true,"delta_db":-0.5292237238379678},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0263986880205326e-05,"output_power":9.498555514832986e-06,"reference_active":true,"delta_db":-0.3365852290926999}],"active_band_count":3,"active_band_mae_db":0.411663586441632}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00012636239720129058,"log_rms_mae_db":1.2932424540616512,"rms_envelope_nrmse":0.13072191239722433,"envelope_correlation":0.988658050500886,"rms_gain_error_db":-0.4926468738250028,"reference_mean_power":0.0006115927691385327,"output_mean_power":0.0005460063012048762},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.048532457224497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017739140051082302,"output_power":0.0001641289923687776,"reference_active":true,"delta_db":-0.3374725924441568},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4831454807134891e-05,"output_power":1.3345764939033538e-05,"reference_active":true,"delta_db":-0.45840281474848166},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.857972973983328e-07,"output_power":4.7484970583140565e-07,"reference_active":true,"delta_db":-0.09898921100875653}],"active_band_count":3,"active_band_mae_db":0.29828820606713163}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":2.151388184068941,"rms_envelope_nrmse":0.2533832940540603,"envelope_correlation":0.9565193481300291,"rms_gain_error_db":-0.726114480134492,"reference_mean_power":9.493685130389179e-05,"output_mean_power":8.031994006020623e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.14467403295884e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.902158206464311e-05,"output_power":2.317631038370702e-05,"reference_active":true,"delta_db":-0.9767678517958733},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.98285507088674e-06,"output_power":8.11545409947705e-06,"reference_active":true,"delta_db":-0.8994194040819845},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.211971373682006e-06,"output_power":2.7587340824073137e-06,"reference_active":true,"delta_db":-0.6606182567980394}],"active_band_count":3,"active_band_mae_db":0.8456018375586325}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":1.2103312871102312,"rms_envelope_nrmse":0.15648168881469662,"envelope_correlation":0.9857461465591799,"rms_gain_error_db":-0.7261168075701525,"reference_mean_power":9.570694886911733e-05,"output_mean_power":8.097142649336698e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.518176746918327e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.649434521118165e-05,"output_power":2.1356962414175518e-05,"reference_active":true,"delta_db":-0.9361370735660325},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515847029872691e-06,"output_power":3.843588141558756e-06,"reference_active":true,"delta_db":-0.7000237737432506},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.652745912515814e-07,"output_power":2.176642935338784e-07,"reference_active":true,"delta_db":-0.8590846243400317}],"active_band_count":3,"active_band_mae_db":0.8317484905497716}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.4845845699310303,"predicted_mos":3.4845845699310303,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89760,"input_sample_rate":24000,"input_seconds":3.74,"input_float32_sha256":"691c74d089d767e0585130c891696bd9dcc15d37b66e1a8ac9f11a44fff37ab1","internal_samples":59840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7450684400009777},"started_at":"2026-10-13T05:10:25.249409+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0021058929996797815,"native_input_samples":89520,"native_input_sha256":"742f19a2c09138bb13a28e286ef7cc75438d1b7f570a24981e81691740d064c3","output_sha256":"f526505e3341f12910d0a9e26db0fe5fd042faba95aaef5647d53a66bea5f917","output_samples":89760,"length_delta_samples":240,"analysis_s":0.5576515749999089,"synthesis_s":0.09603436600082205,"total_s":0.653685941000731,"rtf":0.17525092252030322,"output_peak":0.9105510115623474,"output_rms":0.14146820515333222,"finished_at":"2026-10-13T05:10:25.911454+00:00","analysis_stages":{"external_f0_s":0.08340417600084038},"evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"f526505e3341f12910d0a9e26db0fe5fd042faba95aaef5647d53a66bea5f917","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59680,"output_samples":59840,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-16.98381633285919,"output_peak":0.9100302984160593},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973262032085561,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.6943415770868535,"spectral_convergence":0.1662323131866825,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.337792341581774,"spectral_convergence":0.14430778348174247,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.682580398288213,"spectral_convergence":0.15817073678794497,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.004135934570310864,"delta_db":-0.15361514703718093,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.01230451769530113,"delta_db":-0.1286192827610767,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.0029483025818718717,"delta_db":-0.23012889779907067,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.00047444111338194384,"delta_db":-0.6753452144886051,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00014023787789894606,"delta_db":-0.5547750971986072,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":7.737138008121494e-05,"delta_db":-0.7343979591606443,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4128135997408641,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16807925845101132,"pitch":null,"output_rms_dbfs":-16.972188764328898,"output_peak":0.9100302984160593,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.571571438985614},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.9589124277683554,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14447560597487225,"rms_dbfs":-16.80410950587789,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.859693585918194,"mean":-0.0023839802269002386,"zero_energy":false},"output":{"rms":0.14170675784466305,"rms_dbfs":-16.9721887643289,"peak":0.9100302984160593,"peak_dbfs":-0.8188829620208746,"crest_factor_db":16.153305802308026,"mean":-4.095668228517858e-05,"zero_energy":false},"rms_difference_db":-0.16807925845101224},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":59,"quiet_samples":9440,"reference_mean_power":6.490114235745396e-06,"output_mean_power":6.257591550905021e-06,"output_rms_dbfs":-52.035927877825415,"output_zero_energy":false,"output_peak":0.011477266951771628,"output_energy_fraction":4.9291186068299765e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":110,"both_estimated_voiced_frames":101,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9439252336448598,"absolute_difference_cents_median":1.0998922337551544,"absolute_difference_cents_p95":4.842400534144429,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00043785897255075944,"log_rms_mae_db":1.0292047222390788,"rms_envelope_nrmse":0.058079743301935285,"envelope_correlation":0.9968853711607344,"rms_gain_error_db":-0.16807942414915092,"reference_mean_power":0.0209012125553907,"output_mean_power":0.020107752890553617},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.005551294872497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00444581511703579,"output_power":0.004286639257742143,"reference_active":true,"delta_db":-0.15834461253875806},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039454721092167534,"output_power":0.0003565620796128115,"reference_active":true,"delta_db":-0.43963823609942676},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010840168168493419,"output_power":7.287278494722175e-05,"reference_active":true,"delta_db":-1.7247065258037722}],"active_band_count":3,"active_band_mae_db":0.7742297914806523}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0004107294113879301,"log_rms_mae_db":0.5573320548669984,"rms_envelope_nrmse":0.03461341605597088,"envelope_correlation":0.999086102887848,"rms_gain_error_db":-0.16807577684545802,"reference_mean_power":0.021070817398047652,"output_mean_power":0.02027093615459086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.012296407388438e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00391054354874771,"output_power":0.0037967719948196707,"reference_active":true,"delta_db":-0.1282260915086628},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019349823972573515,"output_power":0.00017488780359219117,"reference_active":true,"delta_db":-0.43917495047260213},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.980555526095884e-06,"output_power":3.732824281049066e-06,"reference_active":true,"delta_db":-2.047039804856245}],"active_band_count":3,"active_band_mae_db":0.87148028227917}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.0003730035430310483,"log_rms_mae_db":0.9640767810659834,"rms_envelope_nrmse":0.06873580206891922,"envelope_correlation":0.9964661025090206,"rms_gain_error_db":-0.16307503674663174,"reference_mean_power":0.007423998211444209,"output_mean_power":0.007150400289635639},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.725628531347034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002177971361103236,"output_power":0.002102514359694715,"reference_active":true,"delta_db":-0.1531319415676014},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032785256002961345,"output_power":0.00027011873542186416,"reference_active":true,"delta_db":-0.841238713193181},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.051887548388713e-05,"output_power":4.6439111974544275e-05,"reference_active":true,"delta_db":-2.3901379454873166}],"active_band_count":3,"active_band_mae_db":1.128169533416033}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00035212990311193784,"log_rms_mae_db":0.6190635489601822,"rms_envelope_nrmse":0.03992141949230771,"envelope_correlation":0.998919536536174,"rms_gain_error_db":-0.16307272902233702,"reference_mean_power":0.007484242812710004,"output_mean_power":0.007208428516298495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.215922155621447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018646593945503165,"output_power":0.0018118461705215843,"reference_active":true,"delta_db":-0.12478191175196109},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001569482176698105,"output_power":0.00012732431533044202,"reference_active":true,"delta_db":-0.9084503864962197},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505413793795638e-06,"output_power":3.4776370461189576e-06,"reference_active":true,"delta_db":-2.7199067260526357}],"active_band_count":3,"active_band_mae_db":1.2510463414336055}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00013449969207959865,"log_rms_mae_db":1.1135668137479622,"rms_envelope_nrmse":0.12892402604970418,"envelope_correlation":0.9914635190688652,"rms_gain_error_db":-0.6728286728163095,"reference_mean_power":0.0006066736643199692,"output_mean_power":0.0005196037483645541},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.452566559449028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001977891479666072,"output_power":0.00017364887668554102,"reference_active":true,"delta_db":-0.5653048136435628},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.140914614473541e-05,"output_power":2.4945061727389628e-05,"reference_active":true,"delta_db":-1.0007154734355161},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0263986880205326e-05,"output_power":8.71621250882809e-06,"reference_active":true,"delta_db":-0.7098827805669955}],"active_band_count":3,"active_band_mae_db":0.7586343558820249}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00012636239720129058,"log_rms_mae_db":0.7121500737303654,"rms_envelope_nrmse":0.0935260624147149,"envelope_correlation":0.9974195665905752,"rms_gain_error_db":-0.6728244487676105,"reference_mean_power":0.0006115927691385327,"output_mean_power":0.0005238173718309033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.048532457224497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017739140051082302,"output_power":0.00015722315556803278,"reference_active":true,"delta_db":-0.5241605390951375},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4831454807134891e-05,"output_power":1.1911078962425979e-05,"reference_active":true,"delta_db":-0.9523264902959315},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.857972973983328e-07,"output_power":4.764617159907284e-07,"reference_active":true,"delta_db":-0.08427083864448384}],"active_band_count":3,"active_band_mae_db":0.5202526226785177}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":1.228996870223752,"rms_envelope_nrmse":0.10748522000238414,"envelope_correlation":0.9931223501362954,"rms_gain_error_db":-0.3978946431515794,"reference_mean_power":9.493685130389179e-05,"output_mean_power":8.66254212200138e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.14467403295884e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.902158206464311e-05,"output_power":2.6828891653332855e-05,"reference_active":true,"delta_db":-0.34118352002275665},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.98285507088674e-06,"output_power":8.657006117370398e-06,"reference_active":true,"delta_db":-0.6188704175114406},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.211971373682006e-06,"output_power":2.457929303505461e-06,"reference_active":true,"delta_db":-1.1620227874376756}],"active_band_count":3,"active_band_mae_db":0.7073589083239575}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":0.8793925482576925,"rms_envelope_nrmse":0.0677068847733994,"envelope_correlation":0.9980630438612202,"rms_gain_error_db":-0.3978935226420347,"reference_mean_power":9.570694886911733e-05,"output_mean_power":8.732812163716459e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.518176746918327e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.649434521118165e-05,"output_power":2.4631818751452817e-05,"reference_active":true,"delta_db":-0.3165671047138797},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515847029872691e-06,"output_power":4.107554352564736e-06,"reference_active":true,"delta_db":-0.411559033012387},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.652745912515814e-07,"output_power":1.909565501215754e-07,"reference_active":true,"delta_db":-1.4276109404926964}],"active_band_count":3,"active_band_mae_db":0.7185790260729877}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0000","method":"magphase","status":"measured","mos":{"status":"measured","value":3.7893569469451904,"predicted_mos":3.7893569469451904,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":179659,"input_sample_rate":48000,"input_seconds":3.7428958333333333,"input_float32_sha256":"cf521df37fe94b00d9e77f14f5dc44023e83bcf1a8db2b31f723319915288501","internal_samples":59887,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.253896342999724},"started_at":"2026-10-13T05:14:07.633691+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006165085000247927,"native_input_samples":179040,"native_input_sha256":"2fd3128c4cb73fd3c53b4126e1bb8f51c8f1fb7dabe7837a0622028db03473e3","output_sha256":"a3075f1a740852e2fb01f82d79f4011652c84999732c4da25cfbe611ef82375a","output_samples":179659,"length_delta_samples":619,"analysis_s":1.802700737998748,"synthesis_s":0.6509244510016288,"total_s":2.453625189000377,"rtf":0.6578083616622994,"output_peak":0.9361388456982593,"output_rms":0.13884172409256806,"finished_at":"2026-10-13T05:14:10.100932+00:00","evaluation_input":{"reference":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"a3075f1a740852e2fb01f82d79f4011652c84999732c4da25cfbe611ef82375a","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":59680,"output_samples":59887,"length_delta_samples":207,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-17.146665690604703,"output_peak":0.930417636831583},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9965434902399519,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.996840278232828,"spectral_convergence":0.668983251627724,"spectral_convergence_status":"ok","amplitude_floor":3.189740694195829e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.325517897295155,"spectral_convergence":0.5064524479815479,"spectral_convergence_status":"ok","amplitude_floor":2.205927306089595e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.084980086587329,"spectral_convergence":0.5113897925382997,"spectral_convergence_status":"ok","amplitude_floor":1.9576049396551346e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004284845572396228,"output_power":0.003943012096759665,"delta_db":-0.36107064942283246,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.012674374150346739,"output_power":0.011721344153458359,"delta_db":-0.3394910588247301,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0031087438744962816,"output_power":0.002894366484726044,"delta_db":-0.31031421901812106,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005542642121172655,"output_power":0.000546145683625722,"delta_db":-0.06408332226397255,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001593466145796079,"output_power":0.00016452380081383449,"delta_db":0.13885892939840289,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.162629787041768e-05,"output_power":8.754061086756678e-05,"delta_db":-0.1981056655769099,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23532064075082815,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3275250637917705,"pitch":null,"output_rms_dbfs":-17.131634569669657,"output_peak":0.930417636831583,"reference_rms_dbfs":-16.80410950587789,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.135779420705104},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.5535468586988535,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.14447560597487225,"rms_dbfs":-16.80410950587789,"peak":0.89697265625,"peak_dbfs":-0.9444159199596966,"crest_factor_db":15.859693585918194,"mean":-0.0023839802269002386,"zero_energy":false},"output":{"rms":0.13912919474449348,"rms_dbfs":-17.131634569669657,"peak":0.930417636831583,"peak_dbfs":-0.6264413158493082,"crest_factor_db":16.505193253820348,"mean":-0.0023653905852952487,"zero_energy":false},"rms_difference_db":-0.32752506379176793},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.7623893892159688e-05,"threshold_dbfs":-47.53898130597884,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":59,"quiet_samples":9440,"reference_mean_power":6.490114235745396e-06,"output_mean_power":7.300099253360973e-05,"output_rms_dbfs":-41.36671235099551,"output_zero_energy":false,"output_peak":0.16670514143266557,"output_energy_fraction":0.0005965342746846476,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":114,"both_estimated_voiced_frames":87,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.8130841121495327,"absolute_difference_cents_median":10.852731068955144,"absolute_difference_cents_p95":44.24931653474261,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00043785897255075944,"log_rms_mae_db":4.078144881404845,"rms_envelope_nrmse":0.32202120372580495,"envelope_correlation":0.892602557927914,"rms_gain_error_db":-0.3275241685098724,"reference_mean_power":0.0209012125553907,"output_mean_power":0.01938291382816959},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.005551294872497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00444581511703579,"output_power":0.004305527740700445,"reference_active":true,"delta_db":-0.13925006482737606},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00039454721092167534,"output_power":0.00039945023610576263,"reference_active":true,"delta_db":0.05363704521331684},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010840168168493419,"output_power":0.0001189330246498295,"reference_active":true,"delta_db":0.402664441368829}],"active_band_count":3,"active_band_mae_db":0.19851718380317396}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0004107294113879301,"log_rms_mae_db":2.912127672601596,"rms_envelope_nrmse":0.2437967378025531,"envelope_correlation":0.9347821378850222,"rms_gain_error_db":-0.32751190628157323,"reference_mean_power":0.021070817398047652,"output_mean_power":0.019540253465392055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.012296407388438e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00391054354874771,"output_power":0.0038018943816919702,"reference_active":true,"delta_db":-0.12237078853952316},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019349823972573515,"output_power":0.00018730761380091865,"reference_active":true,"delta_db":-0.14121587332592359},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.980555526095884e-06,"output_power":5.431128465928673e-06,"reference_active":true,"delta_db":-0.4185145135976107}],"active_band_count":3,"active_band_mae_db":0.22736705848768582}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.0003730035430310483,"log_rms_mae_db":5.416262367650078,"rms_envelope_nrmse":0.4284620987289332,"envelope_correlation":0.850232436780026,"rms_gain_error_db":-0.2767972491544007,"reference_mean_power":0.007423998211444209,"output_mean_power":0.006965593619985256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.725628531347034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002177971361103236,"output_power":0.0021594400496848668,"reference_active":true,"delta_db":-0.037110130898997826},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00032785256002961345,"output_power":0.00030588564868147295,"reference_active":true,"delta_db":-0.30119477642162834},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.051887548388713e-05,"output_power":7.24367576199726e-05,"reference_active":true,"delta_db":-0.4593869856684782}],"active_band_count":3,"active_band_mae_db":0.26589729766303477}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00035212990311193784,"log_rms_mae_db":4.133783803803327,"rms_envelope_nrmse":0.3233429449706752,"envelope_correlation":0.9087035803644811,"rms_gain_error_db":-0.2767901020629988,"reference_mean_power":0.007484242812710004,"output_mean_power":0.0070221298948022134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.215922155621447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018646593945503165,"output_power":0.001857328292706186,"reference_active":true,"delta_db":-0.017108391870651697},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001569482176698105,"output_power":0.00014921749997056836,"reference_active":true,"delta_db":-0.2193662871334528},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505413793795638e-06,"output_power":5.669115708911645e-06,"reference_active":true,"delta_db":-0.5975960500343432}],"active_band_count":3,"active_band_mae_db":0.2780235763461492}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00013449969207959865,"log_rms_mae_db":5.177096606803545,"rms_envelope_nrmse":0.5013248439639426,"envelope_correlation":0.8191068108695168,"rms_gain_error_db":-0.07074987092981518,"reference_mean_power":0.0006066736643199692,"output_mean_power":0.0005968705562112257},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":3.452566559449028e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001977891479666072,"output_power":0.0002063971036730564,"reference_active":true,"delta_db":0.18501139015685328},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.140914614473541e-05,"output_power":3.241470124332278e-05,"reference_active":true,"delta_db":0.13685892897769955},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0263986880205326e-05,"output_power":1.1269137725890378e-05,"reference_active":true,"delta_db":0.40574598420576036}],"active_band_count":3,"active_band_mae_db":0.2425387677801044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00012636239720129058,"log_rms_mae_db":3.739007410508046,"rms_envelope_nrmse":0.36854012920119433,"envelope_correlation":0.8959746835680708,"rms_gain_error_db":-0.07072333091220159,"reference_mean_power":0.0006115927691385327,"output_mean_power":0.0006017138513835057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":3.048532457224497e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017739140051082302,"output_power":0.00018674069984638006,"reference_active":true,"delta_db":0.22306419574035197},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4831454807134891e-05,"output_power":1.5712340689886954e-05,"reference_active":true,"delta_db":0.25057134567319894},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.857972973983328e-07,"output_power":5.265715271429818e-07,"reference_active":true,"delta_db":0.3500227777043063}],"active_band_count":3,"active_band_mae_db":0.27455277303928577}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":5.531999738763852e-05,"log_rms_mae_db":5.194333717128609,"rms_envelope_nrmse":0.5590234314250715,"envelope_correlation":0.79581938723321,"rms_gain_error_db":0.34721583668055145,"reference_mean_power":9.493685130389179e-05,"output_mean_power":0.0001028386595941056},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":9.14467403295884e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.902158206464311e-05,"output_power":3.18896895454884e-05,"reference_active":true,"delta_db":0.4092920768374476},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.98285507088674e-06,"output_power":1.334617027874301e-05,"reference_active":true,"delta_db":1.2610189539170156},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.211971373682006e-06,"output_power":3.6591250354690484e-06,"reference_active":true,"delta_db":0.5660558393316848}],"active_band_count":3,"active_band_mae_db":0.7454556233620494}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":4.1914797815135596e-05,"log_rms_mae_db":3.6732291034830125,"rms_envelope_nrmse":0.3870512506401001,"envelope_correlation":0.8970194805678208,"rms_gain_error_db":0.34723766213117346,"reference_mean_power":9.570694886911733e-05,"output_mean_power":0.00010367337512564617},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":8.518176746918327e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.649434521118165e-05,"output_power":2.9380316292634753e-05,"reference_active":true,"delta_db":0.44903276154445104},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515847029872691e-06,"output_power":6.001566614266467e-06,"reference_active":true,"delta_db":1.2352540872463884},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.652745912515814e-07,"output_power":3.5352636020195795e-07,"reference_active":true,"delta_db":1.247261478808058}],"active_band_count":3,"active_band_mae_db":0.9771827758662992}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"hnm","status":"measured","mos":{"status":"measured","value":3.2113099098205566,"predicted_mos":3.2113099098205566,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88640,"input_sample_rate":16000,"input_seconds":5.54,"input_float32_sha256":"99285af7225f8a0e1662a21a94ae3839cbeee22116ed69c386c4df24561d607f","internal_samples":88640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5718108199998824},"started_at":"2026-10-13T04:38:47.097472+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3173999832361005e-05,"native_input_samples":88640,"native_input_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","output_sha256":"fd4cb1cdaf5c5b78dba57dc62e8a5c2b64cbfb5b8c64b8eca7ddd76ab8731e5f","output_samples":88640,"length_delta_samples":0,"analysis_s":0.9330833310004891,"synthesis_s":0.45892581499992957,"total_s":1.3920091460004187,"rtf":0.2512651888087398,"output_peak":0.5688475769844377,"output_rms":0.06200843319094478,"finished_at":"2026-10-13T04:38:48.493874+00:00","evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"fd4cb1cdaf5c5b78dba57dc62e8a5c2b64cbfb5b8c64b8eca7ddd76ab8731e5f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.135142052281211,"spectral_convergence":0.34057822514004094,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.751448653973341,"spectral_convergence":0.29243375475178013,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8895283296273115,"spectral_convergence":0.3069148462374022,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.0007109551986426736,"delta_db":-0.08036843985556355,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.0017120545246405963,"delta_db":-0.09121380862698966,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0009659361045159967,"delta_db":-0.07202045615712979,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.00023858988757476514,"delta_db":-0.481618361414676,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":6.39514658777557e-05,"delta_db":0.26507450279786154,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.0001535572483806163,"delta_db":-0.036642382162907626,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17115632516918802,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10187892605144636,"pitch":null,"output_rms_dbfs":-24.150984842541256,"output_peak":0.5688475769844377,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9253730119606205},"common_prefix":{"samples":88640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.135142052281211,"spectral_convergence":0.34057822514004094,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.751448653973341,"spectral_convergence":0.29243375475178013,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8895283296273115,"spectral_convergence":0.3069148462374022,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.0007109551986426736,"delta_db":-0.08036843985556355,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.0017120545246405963,"delta_db":-0.09121380862698966,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0009659361045159967,"delta_db":-0.07202045615712979,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.00023858988757476514,"delta_db":-0.481618361414676,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":6.39514658777557e-05,"delta_db":0.26507450279786154,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.0001535572483806163,"delta_db":-0.036642382162907626,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17115632516918802,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10187892605144636,"pitch":null,"output_rms_dbfs":-24.150984842541256,"output_peak":0.5688475769844377,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9253730119606205},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88640,"output_samples":88640,"intelligibility":{"estoi":{"value":0.7309631895435401,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06274002742130019,"rms_dbfs":-24.04910591648981,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.113070510727493,"mean":-0.006068708578171713,"zero_energy":false},"output":{"rms":0.06200843319094478,"rms_dbfs":-24.150984842541256,"peak":0.5688475769844377,"peak_dbfs":-4.90008174916263,"crest_factor_db":19.25090309337863,"mean":-0.00046954161388109957,"zero_energy":false},"rms_difference_db":-0.10187892605144455},"full_scale":{"reference":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":554,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":547,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":99,"both_estimated_voiced_frames":78,"reference_voiced_output_unvoiced_frames":123,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.3880597014925373,"absolute_difference_cents_median":0.7460812229426694,"absolute_difference_cents_p95":4.073146145268583,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00022148173042756648,"log_rms_mae_db":1.6027137705904815,"rms_envelope_nrmse":0.09573032778398903,"envelope_correlation":0.9898410096261309,"rms_gain_error_db":-0.10199220399640412,"reference_mean_power":0.0039398267431541775,"output_mean_power":0.0038483795964795236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.3094240146918987e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011389510990882513,"output_power":0.001117776363525192,"reference_active":true,"delta_db":-0.08150156232881889},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017715594495014008,"output_power":0.00017008465406593583,"reference_active":true,"delta_db":-0.17690599804988716},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5198627564616767e-05,"output_power":3.557776701495247e-05,"reference_active":true,"delta_db":1.49801799050874}],"active_band_count":3,"active_band_mae_db":0.5854751836291486}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.00020322184022683357,"log_rms_mae_db":0.7156073305175715,"rms_envelope_nrmse":0.045746987418141295,"envelope_correlation":0.9976097387418478,"rms_gain_error_db":-0.10235837198660071,"reference_mean_power":0.0039603691212238695,"output_mean_power":0.0038681190186966845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.1092095592585075e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010485738336262066,"output_power":0.0010318438056382215,"reference_active":true,"delta_db":-0.06985054662044517},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.703453851742181e-05,"output_power":8.062915936692576e-05,"reference_active":true,"delta_db":-0.3319949900068644},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8525148080015621e-06,"output_power":2.275970650725766e-06,"reference_active":true,"delta_db":0.894049692306007}],"active_band_count":3,"active_band_mae_db":0.43196507631110553}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00017913726163896074,"log_rms_mae_db":3.1311753588410185,"rms_envelope_nrmse":0.09926146515392972,"envelope_correlation":0.9939020780747121,"rms_gain_error_db":0.020819221215717415,"reference_mean_power":0.001348381907516537,"output_mean_power":0.0013548613025997141},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":1.724352265721046e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753679492081999,"output_power":0.0008439577438341453,"reference_active":true,"delta_db":-0.15869938819312726},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012951638097036193,"output_power":0.00011664351777869389,"reference_active":true,"delta_db":-0.4546409179410167},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6972503859652748e-05,"output_power":1.2634079252429002e-05,"reference_active":true,"delta_db":-1.2820231925105865}],"active_band_count":3,"active_band_mae_db":0.6317878328815768}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.0001658599445391014,"log_rms_mae_db":2.377995718330495,"rms_envelope_nrmse":0.0660485715635256,"envelope_correlation":0.9973561252959736,"rms_gain_error_db":0.020805393008963217,"reference_mean_power":0.001355729112927364,"output_mean_power":0.0013622394761589475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":1.6447728018068126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082538722946322,"output_power":0.0007991719079343029,"reference_active":true,"delta_db":-0.14017535714885793},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.001034181994635e-05,"output_power":5.446269807860912e-05,"reference_active":true,"delta_db":-0.4212694806405197},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.081003459949387e-06,"output_power":1.2909599067652844e-06,"reference_active":true,"delta_db":-2.073600476573051}],"active_band_count":3,"active_band_mae_db":0.8783484381208096}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00010266849739942128,"log_rms_mae_db":1.9431924733766273,"rms_envelope_nrmse":0.14998223068252586,"envelope_correlation":0.9860944115639545,"rms_gain_error_db":-0.3841278273803327,"reference_mean_power":0.00024063110594780816,"output_mean_power":0.0002202616987481009},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.4743465614681173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013528556332001237,"output_power":0.00011823626151932834,"reference_active":true,"delta_db":-0.5850076493135212},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4222986289891938e-05,"output_power":1.4202741391294963e-05,"reference_active":true,"delta_db":-0.0061861207794168555},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7979674926629104e-06,"output_power":2.4412431233504407e-06,"reference_active":true,"delta_db":-0.592316315171437}],"active_band_count":3,"active_band_mae_db":0.3945033617547917}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":1.2277830945337747,"rms_envelope_nrmse":0.10433577724579644,"envelope_correlation":0.9939226374130807,"rms_gain_error_db":-0.38414789665913135,"reference_mean_power":0.00024194203098044546,"output_mean_power":0.00022146063066444315},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.355449255956096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012689841334701806,"output_power":0.00011232919923818198,"reference_active":true,"delta_db":-0.5296352906331143},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.520296796235236e-06,"output_power":6.741447396921676e-06,"reference_active":true,"delta_db":0.14485785342336238},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9276932572800109e-07,"output_power":1.6036716023725633e-07,"reference_active":true,"delta_db":-0.7992248944983873}],"active_band_count":3,"active_band_mae_db":0.4912393461849547}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":1.9179339279159917,"rms_envelope_nrmse":0.1425475791742824,"envelope_correlation":0.9859628236072164,"rms_gain_error_db":-0.05173122994814156,"reference_mean_power":0.00013131711251995076,"output_mean_power":0.00012976220049022583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":6.872487837717031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.115138542486779e-05,"output_power":2.9462034778363854e-05,"reference_active":true,"delta_db":-0.24214628372509692},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2237072790355106e-06,"output_power":2.930727070957857e-06,"reference_active":true,"delta_db":-0.4138022396576624},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5460080528529544e-06,"output_power":1.2585103940412487e-06,"reference_active":true,"delta_db":-0.893549450043723}],"active_band_count":3,"active_band_mae_db":0.5164993244754941}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":1.2403050005370113,"rms_envelope_nrmse":0.07624092020680025,"envelope_correlation":0.9958362875346748,"rms_gain_error_db":-0.05174290990637896,"reference_mean_power":0.00013203257062264153,"output_mean_power":0.0001304688360446178},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":6.49214546612456e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9138042963833366e-05,"output_power":2.7664841432788342e-05,"reference_active":true,"delta_db":-0.22532194028418398},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5333654526802435e-06,"output_power":1.3465557427682272e-06,"reference_active":true,"delta_db":-0.5642133779206986},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.807699107498535e-08,"output_power":6.092666739610854e-08,"reference_active":true,"delta_db":-1.6005504618693247}],"active_band_count":3,"active_band_mae_db":0.7966952600247357}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.4062483310699463,"predicted_mos":3.4062483310699463,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88640,"input_sample_rate":16000,"input_seconds":5.54,"input_float32_sha256":"1a317790856f648054ebc35e658f6b83960df3cbcf14420d03026d69921215e6","internal_samples":88640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.710740011998496},"started_at":"2026-10-13T04:41:43.605098+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.138700074283406e-05,"native_input_samples":88640,"native_input_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","output_sha256":"72c535e50ef164235fba41a31dc7164188ad955e407f5cd04027d797f46ac046","output_samples":88640,"length_delta_samples":0,"analysis_s":1.0875276300002952,"synthesis_s":0.08269918099995266,"total_s":1.1702268110002478,"rtf":0.21123227635383535,"output_peak":0.7432229697270127,"output_rms":0.07068633060281487,"finished_at":"2026-10-13T04:41:44.779966+00:00","evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"72c535e50ef164235fba41a31dc7164188ad955e407f5cd04027d797f46ac046","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.698008107202144,"spectral_convergence":0.636548290926109,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.816094280954592,"spectral_convergence":0.37001369049108285,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.074724119399833,"spectral_convergence":0.38896290551072465,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.0005903695519527181,"delta_db":-0.8875512734824925,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.002718771937681365,"delta_db":1.9173380600717715,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0011434281292238622,"delta_db":0.6605842705660642,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.0003104818610439974,"delta_db":0.6622236425156041,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":6.800014771186901e-05,"delta_db":0.531668026675912,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.00016550498215873424,"delta_db":0.28876512244204555,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.824688399292315,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.035814668314929,"pitch":null,"output_rms_dbfs":-23.013291248174884,"output_peak":0.7432229697270127,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.196275502518856},"common_prefix":{"samples":88640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.698008107202144,"spectral_convergence":0.636548290926109,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.816094280954592,"spectral_convergence":0.37001369049108285,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.074724119399833,"spectral_convergence":0.38896290551072465,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.0005903695519527181,"delta_db":-0.8875512734824925,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.002718771937681365,"delta_db":1.9173380600717715,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0011434281292238622,"delta_db":0.6605842705660642,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.0003104818610439974,"delta_db":0.6622236425156041,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":6.800014771186901e-05,"delta_db":0.531668026675912,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.00016550498215873424,"delta_db":0.28876512244204555,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.824688399292315,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.035814668314929,"pitch":null,"output_rms_dbfs":-23.013291248174884,"output_peak":0.7432229697270127,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.196275502518856},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88640,"output_samples":88640,"intelligibility":{"estoi":{"value":0.8036279281368607,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06274002742130019,"rms_dbfs":-24.04910591648981,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.113070510727493,"mean":-0.006068708578171713,"zero_energy":false},"output":{"rms":0.07068633060281487,"rms_dbfs":-23.013291248174884,"peak":0.7432229697270127,"peak_dbfs":-2.577617533799066,"crest_factor_db":20.435673714375817,"mean":1.1615724523640409e-07,"zero_energy":false},"rms_difference_db":1.0358146683149272},"full_scale":{"reference":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":554,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":547,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":147,"both_estimated_voiced_frames":79,"reference_voiced_output_unvoiced_frames":122,"reference_unvoiced_output_voiced_frames":68,"paired_voiced_fraction":0.39303482587064675,"absolute_difference_cents_median":5.705870170221448,"absolute_difference_cents_p95":1203.6604065927963,"difference_over_600_cents_frames":12,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00022148173042756648,"log_rms_mae_db":3.757855535778798,"rms_envelope_nrmse":0.23236053335156667,"envelope_correlation":0.9718954209519406,"rms_gain_error_db":1.0358516126842479,"reference_mean_power":0.0039398267431541775,"output_mean_power":0.00500106253429286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.3094240146918987e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011389510990882513,"output_power":0.0014506401105915975,"reference_active":true,"delta_db":1.050546036235482},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017715594495014008,"output_power":0.000212883120345842,"reference_active":true,"delta_db":0.797854964371379},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5198627564616767e-05,"output_power":2.7410615651719868e-05,"reference_active":true,"delta_db":0.36541902338037385}],"active_band_count":3,"active_band_mae_db":0.737940007995745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.00020322184022683357,"log_rms_mae_db":3.157683743215171,"rms_envelope_nrmse":0.19648178239481995,"envelope_correlation":0.9850265032427781,"rms_gain_error_db":1.0367301624414167,"reference_mean_power":0.0039603691212238695,"output_mean_power":0.005028155289031165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.1092095592585075e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010485738336262066,"output_power":0.0013365899835069133,"reference_active":true,"delta_db":1.0539918614874701},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.703453851742181e-05,"output_power":0.00010703125653102917,"reference_active":true,"delta_db":0.8981899310450657},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8525148080015621e-06,"output_power":2.1425745719306705e-06,"reference_active":true,"delta_db":0.6317425822078375}],"active_band_count":3,"active_band_mae_db":0.8613081249134579}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00017913726163896074,"log_rms_mae_db":6.481436872081371,"rms_envelope_nrmse":0.18863679586871035,"envelope_correlation":0.98637763462783,"rms_gain_error_db":0.8829750563861765,"reference_mean_power":0.001348381907516537,"output_mean_power":0.0016523818736633783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":1.724352265721046e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753679492081999,"output_power":0.000979460532810664,"reference_active":true,"delta_db":0.48796299588358805},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012951638097036193,"output_power":0.00014941918571175587,"reference_active":true,"delta_db":0.6208166472887389},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6972503859652748e-05,"output_power":1.299917574672249e-05,"reference_active":true,"delta_db":-1.158301007238593}],"active_band_count":3,"active_band_mae_db":0.7556935501369733}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.0001658599445391014,"log_rms_mae_db":5.93484430993186,"rms_envelope_nrmse":0.14449083271727703,"envelope_correlation":0.9950156456637728,"rms_gain_error_db":0.8828284206491012,"reference_mean_power":0.001355729112927364,"output_mean_power":0.0016613294518366018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":1.6447728018068126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082538722946322,"output_power":0.0009253197234926076,"reference_active":true,"delta_db":0.4963407428251455},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.001034181994635e-05,"output_power":7.0755749084691e-05,"reference_active":true,"delta_db":0.7153563263993563},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.081003459949387e-06,"output_power":1.4631992126938268e-06,"reference_active":true,"delta_db":-1.5296934349864697}],"active_band_count":3,"active_band_mae_db":0.9137968347369906}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00010266849739942128,"log_rms_mae_db":1.5716444890302022,"rms_envelope_nrmse":0.2054873882725959,"envelope_correlation":0.9802060720407465,"rms_gain_error_db":0.8071386710414913,"reference_mean_power":0.00024063110594780816,"output_mean_power":0.0002897781491482923},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.4743465614681173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013528556332001237,"output_power":0.00016699144523466618,"reference_active":true,"delta_db":0.9144276898959752},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4222986289891938e-05,"output_power":1.929378387606116e-05,"reference_active":true,"delta_db":1.3242661795697208},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7979674926629104e-06,"output_power":2.638056158010418e-06,"reference_active":true,"delta_db":-0.25558628049609566}],"active_band_count":3,"active_band_mae_db":0.8314267166539305}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":0.9739657920305737,"rms_envelope_nrmse":0.12546920440093984,"envelope_correlation":0.996057755881148,"rms_gain_error_db":0.8071401743752253,"reference_mean_power":0.00024194203098044546,"output_mean_power":0.0002913569213395408},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.355449255956096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012689841334701806,"output_power":0.00015901783825677542,"reference_active":true,"delta_db":0.9798965321259012},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.520296796235236e-06,"output_power":8.949105305062364e-06,"reference_active":true,"delta_db":1.3751225374511968},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9276932572800109e-07,"output_power":1.6020440183310023e-07,"reference_active":true,"delta_db":-0.8036348354442047}],"active_band_count":3,"active_band_mae_db":1.052884635007101}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":1.5489356633672353,"rms_envelope_nrmse":0.21440942184287157,"envelope_correlation":0.9717463452162441,"rms_gain_error_db":0.4342508031401513,"reference_mean_power":0.00013131711251995076,"output_mean_power":0.00014512639420018175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":6.872487837717031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.115138542486779e-05,"output_power":3.492261035249768e-05,"reference_active":true,"delta_db":0.496293321469913},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2237072790355106e-06,"output_power":3.7498846541166075e-06,"reference_active":true,"delta_db":0.6566230924040212},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5460080528529544e-06,"output_power":2.0728465128463655e-06,"reference_active":true,"delta_db":1.273553935289565}],"active_band_count":3,"active_band_mae_db":0.8088234497211664}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":0.8556861643506123,"rms_envelope_nrmse":0.12544521498061162,"envelope_correlation":0.9910884766136142,"rms_gain_error_db":0.4342518687642724,"reference_mean_power":0.00013203257062264153,"output_mean_power":0.00014591712553952842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":6.49214546612456e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9138042963833366e-05,"output_power":3.337191370898732e-05,"reference_active":true,"delta_db":0.5892073261166254},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5333654526802435e-06,"output_power":1.60301349044498e-06,"reference_active":true,"delta_db":0.19291503045007213},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.807699107498535e-08,"output_power":1.2532191320361448e-07,"reference_active":true,"delta_db":1.5316454667850161}],"active_band_count":3,"active_band_mae_db":0.7712559411172378}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"facodec","status":"measured","mos":{"status":"measured","value":4.342762470245361,"predicted_mos":4.342762470245361,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88800,"input_sample_rate":16000,"input_seconds":5.55,"input_float32_sha256":"7a6cc4c0e536330cee288d9739b8df771a0083fc16539e86320c66ea249345b4","internal_samples":88800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2879008349991636},"started_at":"2026-10-13T04:49:38.973500+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4316000235558022e-05,"native_input_samples":88640,"native_input_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","output_sha256":"90ce1c659e6f71d7d9cd7d759a8e85898806ac08ef0f73e44934c6f86081c0f9","output_samples":88800,"length_delta_samples":160,"analysis_s":1.977012239999567,"synthesis_s":12.731140305000736,"total_s":14.708152545000303,"rtf":2.654901181407997,"output_peak":0.529787540435791,"output_rms":0.06264823271898329,"finished_at":"2026-10-13T04:49:53.688099+00:00","evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"90ce1c659e6f71d7d9cd7d759a8e85898806ac08ef0f73e44934c6f86081c0f9","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":88640,"output_samples":88800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.0618235153341,"output_peak":0.529787540435791},"common_prefix":{"samples":88640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981981981981982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6032269354254565,"spectral_convergence":0.2936855098380836,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.997260517591795,"spectral_convergence":0.3013266012861745,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.074420237346215,"spectral_convergence":0.3179371513764029,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.0006514552362138457,"delta_db":-0.45994499027149616,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.0020013306758642035,"delta_db":0.5867987955998248,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0009078243206690161,"delta_db":-0.3414863159768913,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.00018916189256471165,"delta_db":-1.4898021818995366,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":4.27150268230045e-05,"delta_db":-1.487623700112478,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.00013939666682316696,"delta_db":-0.45682170257962146,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8037462810733081,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.004887317266496524,"pitch":null,"output_rms_dbfs":-24.05399323375631,"output_peak":0.529787540435791,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.891635896787822},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88640,"output_samples":88640,"intelligibility":{"estoi":{"value":0.7809380715133185,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06274002742130019,"rms_dbfs":-24.04910591648981,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.113070510727493,"mean":-0.006068708578171713,"zero_energy":false},"output":{"rms":0.06270473521959041,"rms_dbfs":-24.05399323375631,"peak":0.529787540435791,"peak_dbfs":-5.517965193467784,"crest_factor_db":18.536028040288524,"mean":-0.00016474821223641894,"zero_energy":false},"rms_difference_db":-0.004887317266497604},"full_scale":{"reference":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":554,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":547,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":185,"both_estimated_voiced_frames":157,"reference_voiced_output_unvoiced_frames":44,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.7810945273631841,"absolute_difference_cents_median":0.0,"absolute_difference_cents_p95":12.7949455868502,"difference_over_600_cents_frames":7,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00022148173042756648,"log_rms_mae_db":1.7705818279513943,"rms_envelope_nrmse":0.13371331304132525,"envelope_correlation":0.9802441002055994,"rms_gain_error_db":-0.0048979544704064485,"reference_mean_power":0.0039398267431541775,"output_mean_power":0.003935385928151031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.3094240146918987e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011389510990882513,"output_power":0.001083221360186962,"reference_active":true,"delta_db":-0.21787862691629117},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017715594495014008,"output_power":0.00016684808683936937,"reference_active":true,"delta_db":-0.26034499658305865},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5198627564616767e-05,"output_power":2.7236450769844352e-05,"reference_active":true,"delta_db":0.3377362553214866}],"active_band_count":3,"active_band_mae_db":0.27198662627361214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.00020322184022683357,"log_rms_mae_db":1.5382272045506225,"rms_envelope_nrmse":0.11751997230378407,"envelope_correlation":0.9835709702394668,"rms_gain_error_db":-0.004388706427488888,"reference_mean_power":0.0039603691212238695,"output_mean_power":0.0039563690431480706},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.1092095592585075e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010485738336262066,"output_power":0.0009972734619990969,"reference_active":true,"delta_db":-0.21784753643167726},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.703453851742181e-05,"output_power":7.68024899768279e-05,"reference_active":true,"delta_db":-0.5431633059097752},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8525148080015621e-06,"output_power":2.0317934897569685e-06,"reference_active":true,"delta_db":0.40117876249976453}],"active_band_count":3,"active_band_mae_db":0.3873965349470723}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00017913726163896074,"log_rms_mae_db":2.4216275989080565,"rms_envelope_nrmse":0.12992528855717908,"envelope_correlation":0.9889120842950185,"rms_gain_error_db":-0.09568582369320801,"reference_mean_power":0.001348381907516537,"output_mean_power":0.0013189985988082845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":1.724352265721046e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753679492081999,"output_power":0.0008370674946552379,"reference_active":true,"delta_db":-0.19430163718835716},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012951638097036193,"output_power":0.0001095753041033629,"reference_active":true,"delta_db":-0.7261201595913026},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6972503859652748e-05,"output_power":1.7686109594973206e-05,"reference_active":true,"delta_db":0.17886395759054707}],"active_band_count":3,"active_band_mae_db":0.36642858479006896}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.0001658599445391014,"log_rms_mae_db":1.9334390621778177,"rms_envelope_nrmse":0.1137469227558294,"envelope_correlation":0.9911762697843898,"rms_gain_error_db":-0.09568453342518304,"reference_mean_power":0.001355729112927364,"output_mean_power":0.0013261860913540832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":1.6447728018068126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082538722946322,"output_power":0.0007907651695727072,"reference_active":true,"delta_db":-0.1861021319564836},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.001034181994635e-05,"output_power":4.910045410661887e-05,"reference_active":true,"delta_db":-0.8714059179851827},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.081003459949387e-06,"output_power":1.8562815548639195e-06,"reference_active":true,"delta_db":-0.49628953005036225}],"active_band_count":3,"active_band_mae_db":0.5179325266640095}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00010266849739942128,"log_rms_mae_db":1.9368914870561507,"rms_envelope_nrmse":0.2561248870919909,"envelope_correlation":0.9655819560627746,"rms_gain_error_db":-1.3500778157262445,"reference_mean_power":0.00024063110594780816,"output_mean_power":0.00017633721828337718},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.4743465614681173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013528556332001237,"output_power":0.00010403769302768776,"reference_active":true,"delta_db":-1.1406074094266385},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4222986289891938e-05,"output_power":1.0598488472871966e-05,"reference_active":true,"delta_db":-1.277468596023276},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7979674926629104e-06,"output_power":3.4100757454415458e-06,"reference_active":true,"delta_db":0.859213612891902}],"active_band_count":3,"active_band_mae_db":1.0924298727806054}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":1.5562430638413038,"rms_envelope_nrmse":0.23962384890889657,"envelope_correlation":0.9710429969907105,"rms_gain_error_db":-1.3500821338659943,"reference_mean_power":0.00024194203098044546,"output_mean_power":0.0001772977028102407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.355449255956096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012689841334701806,"output_power":9.786627174813517e-05,"reference_active":true,"delta_db":-1.1282314797573088},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.520296796235236e-06,"output_power":4.3621566671836215e-06,"reference_active":true,"delta_db":-1.7456610551382032},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9276932572800109e-07,"output_power":1.8636760442454116e-07,"reference_active":true,"delta_db":-0.14667505474146675}],"active_band_count":3,"active_band_mae_db":1.0068558632123261}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":1.9458894763404377,"rms_envelope_nrmse":0.21464482780315824,"envelope_correlation":0.9680974711764057,"rms_gain_error_db":-0.34703751984040876,"reference_mean_power":0.00013131711251995076,"output_mean_power":0.00012123208450991355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":6.872487837717031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.115138542486779e-05,"output_power":2.6794907988677424e-05,"reference_active":true,"delta_db":-0.6542509615803227},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2237072790355106e-06,"output_power":3.5429674392010826e-06,"reference_active":true,"delta_db":0.410115612111008},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5460080528529544e-06,"output_power":1.7616619188953089e-06,"reference_active":true,"delta_db":0.5671081473589177}],"active_band_count":3,"active_band_mae_db":0.5438249070167495}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":1.5683230604582936,"rms_envelope_nrmse":0.14181330730888533,"envelope_correlation":0.9860045469616107,"rms_gain_error_db":-0.34704085244388866,"reference_mean_power":0.00013203257062264153,"output_mean_power":0.00012189250258256331},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":6.49214546612456e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9138042963833366e-05,"output_power":2.5320748915424066e-05,"reference_active":true,"delta_db":-0.6098383260450084},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5333654526802435e-06,"output_power":1.6520128109098459e-06,"reference_active":true,"delta_db":0.3236773662603073},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.807699107498535e-08,"output_power":9.218307729623653e-08,"reference_active":true,"delta_db":0.19788732183387397}],"active_band_count":3,"active_band_mae_db":0.37713433804639657}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.750957489013672,"predicted_mos":3.750957489013672,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":244736,"input_sample_rate":44100,"input_seconds":5.5495691609977325,"input_float32_sha256":"bd4c2bf1bf9a1689fe40fae57dc796614b188fb466257b72c21f0d30b50ce64a","internal_samples":88794,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.633994907002489},"started_at":"2026-10-13T07:08:58.188364+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009490784999798052,"native_input_samples":244314,"native_input_sha256":"ee68ea6152e7334ade96d79f29746515035e9d86e7f7ad684b82d4e79b50c3bd","output_sha256":"6e094f2efd7cd5ad44cf845dd14bc3f22261360d7817fe90b380358073826492","output_samples":244736,"length_delta_samples":422,"analysis_s":0.1254345170000306,"synthesis_s":34.90524070699939,"total_s":35.030675223999424,"rtf":6.323226574729138,"output_peak":0.528749942779541,"output_rms":0.0549368856674832,"finished_at":"2026-10-13T07:09:33.249359+00:00","evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"6e094f2efd7cd5ad44cf845dd14bc3f22261360d7817fe90b380358073826492","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":88640,"output_samples":88794,"length_delta_samples":154,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.20002522779599,"output_peak":0.5264876818535985},"common_prefix":{"samples":88640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998265648579859,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.771791681260583,"spectral_convergence":0.5405881611227994,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.70651857560171,"spectral_convergence":0.4571590979465966,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.499648679528512,"spectral_convergence":0.5196815392534722,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.0007595260787393743,"delta_db":0.20663611995212136,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.0010755319173992373,"delta_db":-2.110156696105078,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0008241627214897204,"delta_db":-0.7613747830180275,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.00020028820985544857,"delta_db":-1.2415848375829377,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":5.000459924303919e-05,"delta_db":-0.8033310241032645,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.00011565818211857834,"delta_db":-1.267581976242667,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0651109061673492,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1433937794449405,"pitch":null,"output_rms_dbfs":-25.19249969593475,"output_peak":0.5264876818535985,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.659319645463603},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88640,"output_samples":88640,"intelligibility":{"estoi":{"value":0.5639362489535981,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06274002742130019,"rms_dbfs":-24.04910591648981,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.113070510727493,"mean":-0.006068708578171713,"zero_energy":false},"output":{"rms":0.055001560976812504,"rms_dbfs":-25.19249969593475,"peak":0.5264876818535985,"peak_dbfs":-5.572235709535402,"crest_factor_db":19.620263986399348,"mean":0.0007807935405429785,"zero_energy":false},"rms_difference_db":-1.1433937794449385},"full_scale":{"reference":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":554,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":547,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":84,"both_estimated_voiced_frames":60,"reference_voiced_output_unvoiced_frames":141,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.29850746268656714,"absolute_difference_cents_median":10.661110041322898,"absolute_difference_cents_p95":53.145388161223174,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00022148173042756648,"log_rms_mae_db":2.9382822735634853,"rms_envelope_nrmse":0.18908880985761908,"envelope_correlation":0.9805000822047432,"rms_gain_error_db":-1.1433559847450026,"reference_mean_power":0.0039398267431541775,"output_mean_power":0.0030278999816736522},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.3094240146918987e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011389510990882513,"output_power":0.0010189930382483925,"reference_active":true,"delta_db":-0.4833386110771841},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017715594495014008,"output_power":0.0001482458872269963,"reference_active":true,"delta_db":-0.7737307723902486},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5198627564616767e-05,"output_power":2.0001045598505888e-05,"reference_active":true,"delta_db":-1.0032418775873095}],"active_band_count":3,"active_band_mae_db":0.7534370870182473}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.00020322184022683357,"log_rms_mae_db":2.6044488406418065,"rms_envelope_nrmse":0.1629824787247796,"envelope_correlation":0.9906346540763362,"rms_gain_error_db":-1.142880539003158,"reference_mean_power":0.0039603691212238695,"output_mean_power":0.003044020772737952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.1092095592585075e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010485738336262066,"output_power":0.0009409691032639405,"reference_active":true,"delta_db":-0.4702365242488826},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.703453851742181e-05,"output_power":7.035936411448843e-05,"reference_active":true,"delta_db":-0.9236972511657988},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8525148080015621e-06,"output_power":1.508241456229512e-06,"reference_active":true,"delta_db":-0.8929081434312045}],"active_band_count":3,"active_band_mae_db":0.7622806396152954}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00017913726163896074,"log_rms_mae_db":3.595654677950823,"rms_envelope_nrmse":0.221124190979372,"envelope_correlation":0.9681374595247827,"rms_gain_error_db":-0.49921448493776827,"reference_mean_power":0.001348381907516537,"output_mean_power":0.001201964021445309},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":1.724352265721046e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753679492081999,"output_power":0.0007390469598222719,"reference_active":true,"delta_db":-0.7351860647503555},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012951638097036193,"output_power":0.00010359913823071513,"reference_active":true,"delta_db":-0.9696855775034945},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6972503859652748e-05,"output_power":1.2976452031497107e-05,"reference_active":true,"delta_db":-1.1658995043624996}],"active_band_count":3,"active_band_mae_db":0.9569237155387832}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.0001658599445391014,"log_rms_mae_db":2.9104560136326394,"rms_envelope_nrmse":0.19172145935843843,"envelope_correlation":0.975587867719743,"rms_gain_error_db":-0.4992131298147587,"reference_mean_power":0.001355729112927364,"output_mean_power":0.001208513786741905},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":1.6447728018068126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082538722946322,"output_power":0.0007028108660657161,"reference_active":true,"delta_db":-0.6981927766793394},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.001034181994635e-05,"output_power":4.650037673801283e-05,"reference_active":true,"delta_db":-1.107696290390182},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.081003459949387e-06,"output_power":1.2693908485783775e-06,"reference_active":true,"delta_db":-2.1467743925620466}],"active_band_count":3,"active_band_mae_db":1.317554486543856}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00010266849739942128,"log_rms_mae_db":4.344343111489095,"rms_envelope_nrmse":0.24956510617009234,"envelope_correlation":0.9609765798879579,"rms_gain_error_db":-0.7566668572423754,"reference_mean_power":0.00024063110594780816,"output_mean_power":0.00020215527647009351},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.4743465614681173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013528556332001237,"output_power":0.00013293221981446355,"reference_active":true,"delta_db":-0.07621197330501558},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4222986289891938e-05,"output_power":1.1275621270263094e-05,"reference_active":true,"delta_db":-1.008503112040353},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7979674926629104e-06,"output_power":2.229436247116195e-06,"reference_active":true,"delta_db":-0.9864760667352247}],"active_band_count":3,"active_band_mae_db":0.6903970506935311}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":3.818417559511835,"rms_envelope_nrmse":0.19543939576129962,"envelope_correlation":0.9772963310263502,"rms_gain_error_db":-0.756662861159191,"reference_mean_power":0.00024194203098044546,"output_mean_power":0.00020325677752048492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.355449255956096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012689841334701806,"output_power":0.0001253913312104193,"reference_active":true,"delta_db":-0.051886789318502},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.520296796235236e-06,"output_power":5.0088154743981015e-06,"reference_active":true,"delta_db":-1.1453233225278272},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9276932572800109e-07,"output_power":1.3084105999154164e-07,"reference_active":true,"delta_db":-1.6829387443966324}],"active_band_count":3,"active_band_mae_db":0.9600496187476538}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":5.3406079559384505,"rms_envelope_nrmse":0.286508166525739,"envelope_correlation":0.9603000477945305,"rms_gain_error_db":-1.7955839436630523,"reference_mean_power":0.00013131711251995076,"output_mean_power":8.684862164878389e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":6.872487837717031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.115138542486779e-05,"output_power":2.7110817732484495e-05,"reference_active":true,"delta_db":-0.6033474898721578},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2237072790355106e-06,"output_power":2.5528951236949545e-06,"reference_active":true,"delta_db":-1.013226260808297},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5460080528529544e-06,"output_power":1.0652292291108197e-06,"reference_active":true,"delta_db":-1.617686770897473}],"active_band_count":3,"active_band_mae_db":1.0780868405259758}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":4.9448500276247715,"rms_envelope_nrmse":0.25666888232614377,"envelope_correlation":0.9726428262903858,"rms_gain_error_db":-1.795579469712076,"reference_mean_power":0.00013203257062264153,"output_mean_power":8.732189098609309e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":6.49214546612456e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9138042963833366e-05,"output_power":2.57098711972031e-05,"reference_active":true,"delta_db":-0.5436047843524995},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5333654526802435e-06,"output_power":1.3232388952518145e-06,"reference_active":true,"delta_db":-0.6400741616942869},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.807699107498535e-08,"output_power":6.025865729946415e-08,"reference_active":true,"delta_db":-1.6484301928546614}],"active_band_count":3,"active_band_mae_db":0.9440363796338159}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"golf","status":"measured","mos":{"status":"measured","value":1.8253366947174072,"predicted_mos":1.8253366947174072,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":132720,"input_sample_rate":24000,"input_seconds":5.53,"input_float32_sha256":"52a6006a80e3cf0ed1e2638e308a14073b72e0eec47890f4a28035da4a1b2109","internal_samples":88480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3781334270006482},"started_at":"2026-10-13T05:02:18.180795+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00296516199978214,"native_input_samples":132960,"native_input_sha256":"be9600739a4f1b9dc7f5b5cedc822386716d2aeb6659f60e15fa2a0639b7f69c","output_sha256":"90c05a2d9f86b95b1dab4421319ef1d37beec1284400e30d107dd4cac6905159","output_samples":132720,"length_delta_samples":-240,"analysis_s":0.5467498640000485,"synthesis_s":0.11102510100045038,"total_s":0.6577749650004989,"rtf":0.11873194314088428,"output_peak":0.46794557571411133,"output_rms":0.04702911120371228,"finished_at":"2026-10-13T05:02:18.848703+00:00","analysis_stages":{"external_f0_s":0.07998272199984058},"evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"90c05a2d9f86b95b1dab4421319ef1d37beec1284400e30d107dd4cac6905159","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":88640,"output_samples":88480,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.55953702976376,"output_peak":0.4680266613016777},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":0.9981949458483754,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.399020307277661,"spectral_convergence":0.6517053380652498,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.926300780852791,"spectral_convergence":0.5769652300784995,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.153671502524361,"spectral_convergence":0.7209890502666885,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007251691699967882,"output_power":0.0004490286785070988,"delta_db":-2.081652527694365,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017515643759528341,"output_power":0.00104006051343971,"delta_db":-2.2636749532528,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009838614848842117,"output_power":0.00036133687772780415,"delta_db":-4.35021672078083,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002670554198356043,"output_power":0.00016349241158181735,"delta_db":-2.1310379640925565,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.027362489584717e-05,"output_power":5.212897214262944e-05,"delta_db":-0.6304814923837985,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015513958728401693,"output_power":0.00014219267173377827,"delta_db":-0.3784541727749202,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.9725863051632118,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-2.5178749202042177,"pitch":null,"output_rms_dbfs":-26.55953702976376,"output_peak":0.4680266613016777,"reference_rms_dbfs":-24.04166210955954,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.159664196884937},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.6920845921139616,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06279381866751936,"rms_dbfs":-24.04166210955954,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.105626703797224,"mean":-0.006066736217864383,"zero_energy":false},"output":{"rms":0.046991915529512505,"rms_dbfs":-26.55953702976376,"peak":0.4680266613016777,"peak_dbfs":-6.594588129697181,"crest_factor_db":19.96494890006658,"mean":-0.00028133764742967963,"zero_energy":false},"rms_difference_db":-2.517874920204221},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":200,"output_estimated_voiced_frames":22,"both_estimated_voiced_frames":19,"reference_voiced_output_unvoiced_frames":181,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.095,"absolute_difference_cents_median":23.47085038942522,"absolute_difference_cents_p95":1211.284398841394,"difference_over_600_cents_frames":8,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00022148173042756648,"log_rms_mae_db":3.9017448935359385,"rms_envelope_nrmse":0.384643209486951,"envelope_correlation":0.8756917477432469,"rms_gain_error_db":-2.5179075040043375,"reference_mean_power":0.003946556996852286,"output_mean_power":0.002210179920078461},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.3095248451460773e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011404131300046696,"output_power":0.0004333008485434925,"reference_active":true,"delta_db":-4.202726693371287},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001770806594011296,"output_power":0.00014867978589495282,"reference_active":true,"delta_db":-0.7591920348449552},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5165131238088292e-05,"output_power":5.060759137077349e-05,"reference_active":true,"delta_db":3.034164682721249}],"active_band_count":3,"active_band_mae_db":2.6653611369791634}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020322184022683357,"log_rms_mae_db":2.866066752415652,"rms_envelope_nrmse":0.33772706076229064,"envelope_correlation":0.9175855376211836,"rms_gain_error_db":-2.5182291457620294,"reference_mean_power":0.003967271393289564,"output_mean_power":0.0022216160083110216},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.108908566456897e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010690956893780732,"output_power":0.00040149145433206245,"reference_active":true,"delta_db":-4.253402726242012},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.14436495155423e-05,"output_power":7.881945029091593e-05,"reference_active":true,"delta_db":-0.6452014877555973},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8936718106860088e-06,"output_power":2.560061634548737e-06,"reference_active":true,"delta_db":1.3094570696999615}],"active_band_count":3,"active_band_mae_db":2.0693537612325232}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00017913726163896074,"log_rms_mae_db":2.711754521484206,"rms_envelope_nrmse":0.4496662675338758,"envelope_correlation":0.9011370835659694,"rms_gain_error_db":-3.3137218919000895,"reference_mean_power":0.0013508222095710003,"output_mean_power":0.0006298338579745704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.7253255020370045e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753201390921549,"output_power":0.000261654033707778,"reference_active":true,"delta_db":-5.244394864014513},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012941613986364918,"output_power":8.533923102744273e-05,"reference_active":true,"delta_db":-1.8083971662845122},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.69535723142722e-05,"output_power":3.307894291275999e-05,"reference_active":true,"delta_db":2.902903995057698}],"active_band_count":3,"active_band_mae_db":3.318565341785574}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.0001658599445391014,"log_rms_mae_db":1.9719269983991898,"rms_envelope_nrmse":0.40245987995328614,"envelope_correlation":0.9384729961522402,"rms_gain_error_db":-3.3137242436295997,"reference_mean_power":0.0013581961580897942,"output_mean_power":0.0006332716895313658},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.645622193911694e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008769987963817979,"output_power":0.00026710001590653443,"reference_active":true,"delta_db":-5.163250834435011},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.240493320386223e-05,"output_power":4.291201485270416e-05,"reference_active":true,"delta_db":-1.6264001618218664},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1070877936020416e-06,"output_power":2.447622198544841e-06,"reference_active":true,"delta_db":0.6506175211362063}],"active_band_count":3,"active_band_mae_db":2.480089505797695}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00010266849739942128,"log_rms_mae_db":1.9377166137849635,"rms_envelope_nrmse":0.3201335321399637,"envelope_correlation":0.9463883752398105,"rms_gain_error_db":-1.859254381681402,"reference_mean_power":0.00024106654582674693,"output_mean_power":0.00015711277770485938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.4786099002897244e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013562441490102612,"output_power":5.962811514461463e-05,"reference_active":true,"delta_db":-3.568867961649704},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4081836252427383e-05,"output_power":1.2948164902869963e-05,"reference_active":true,"delta_db":-0.36451068138622056},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7966121911785616e-06,"output_power":9.747355007173796e-06,"reference_active":true,"delta_db":5.422545372107864}],"active_band_count":3,"active_band_mae_db":3.118641338381263}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":1.417173828758252,"rms_envelope_nrmse":0.25968707750927106,"envelope_correlation":0.9758676334875063,"rms_gain_error_db":-1.859256993270118,"reference_mean_power":0.00024238224219500748,"output_mean_power":0.00015797017502838372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.359467964038251e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012914177979451215,"output_power":5.8340130452609836e-05,"reference_active":true,"delta_db":-3.4509937153189547},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.023680384153715e-06,"output_power":7.871348524723e-06,"reference_active":true,"delta_db":-0.08324478284158648},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9543080796525525e-07,"output_power":4.624740489071964e-07,"reference_active":true,"delta_db":3.740943404002029}],"active_band_count":3,"active_band_mae_db":2.4250606340541903}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":2.459012554551271,"rms_envelope_nrmse":0.18624956472888846,"envelope_correlation":0.9775825304987484,"rms_gain_error_db":0.045420294126623476,"reference_mean_power":0.0001315547544802074,"output_mean_power":0.00013293782770345104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":6.889741710795763e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.121215150053368e-05,"output_power":3.167818444111966e-05,"reference_active":true,"delta_db":0.06436576819162222},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.228703300123207e-06,"output_power":3.281409192287701e-06,"reference_active":true,"delta_db":0.07032252755976981},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5461163588235966e-06,"output_power":2.4407858058948387e-06,"reference_active":true,"delta_db":1.982874937697099}],"active_band_count":3,"active_band_mae_db":0.7058544111494971}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":2.088557863920183,"rms_envelope_nrmse":0.11149463850337824,"envelope_correlation":0.992131218015195,"rms_gain_error_db":0.045422224214608994,"reference_mean_power":0.0001322728034847645,"output_mean_power":0.00013366348516769526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":6.508421705054224e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9875677553163228e-05,"output_power":3.1232451106702206e-05,"reference_active":true,"delta_db":0.1928830530401687},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7544967613069286e-06,"output_power":1.6279689015595653e-06,"reference_active":true,"delta_db":-0.3250646639307432},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.857575656746704e-08,"output_power":1.0528329190231431e-07,"reference_active":true,"delta_db":0.7504458507163467}],"active_band_count":3,"active_band_mae_db":0.42279785589575286}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.618046760559082,"predicted_mos":2.618046760559082,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":133080,"input_sample_rate":24000,"input_seconds":5.545,"input_float32_sha256":"e2b5d0617cd82de27bb2c50ae60366da7e353818d66651f01a84e9459ecad49e","internal_samples":88720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5487186960017425},"started_at":"2026-10-13T05:08:26.871513+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030904169998393627,"native_input_samples":132960,"native_input_sha256":"be9600739a4f1b9dc7f5b5cedc822386716d2aeb6659f60e15fa2a0639b7f69c","output_sha256":"fbd98022cc823b78f9ffc7a762976e5e61790c0afe0a4d176e4d03cebbba7e96","output_samples":133080,"length_delta_samples":120,"analysis_s":1.966069627998877,"synthesis_s":13.499650134001058,"total_s":15.465719761999935,"rtf":2.791646166425981,"output_peak":0.583911657333374,"output_rms":0.06052420972937943,"finished_at":"2026-10-13T05:08:42.349176+00:00","analysis_stages":{"external_f0_s":1.1922436040003959},"evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"fbd98022cc823b78f9ffc7a762976e5e61790c0afe0a4d176e4d03cebbba7e96","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":88640,"output_samples":88720,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.362153559622474,"output_peak":0.5841545731755647},"common_prefix":{"samples":88640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990982867448152,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.463133129852539,"spectral_convergence":0.5882109833829569,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.676907599745209,"spectral_convergence":0.37703445265709223,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.642245796610742,"spectral_convergence":0.39984475901314676,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.000878091867126584,"delta_db":0.8366087654733838,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.0015361852005655312,"delta_db":-0.5619539576241316,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0008401764456387227,"delta_db":-0.6777994301679455,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.00023776819649538777,"delta_db":-0.496601050592018,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":5.485493566362913e-05,"delta_db":-0.4012734339294084,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.00011878004845921743,"delta_db":-1.1519106140999407,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6876912086478048,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3091339256409549,"pitch":null,"output_rms_dbfs":-24.358239842130764,"output_peak":0.5841545731755647,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.927428842069497},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88640,"output_samples":88640,"intelligibility":{"estoi":{"value":0.7703401198093827,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06274002742130019,"rms_dbfs":-24.04910591648981,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.113070510727493,"mean":-0.006068708578171713,"zero_energy":false},"output":{"rms":0.06054635568872337,"rms_dbfs":-24.358239842130764,"peak":0.5841545731755647,"peak_dbfs":-4.669444379849913,"crest_factor_db":19.688795462280854,"mean":7.989975266343741e-05,"zero_energy":false},"rms_difference_db":-0.3091339256409533},"full_scale":{"reference":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":554,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":547,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":129,"reference_unvoiced_output_voiced_frames":86,"paired_voiced_fraction":0.3582089552238806,"absolute_difference_cents_median":5.380303480556076,"absolute_difference_cents_p95":1201.0438725346094,"difference_over_600_cents_frames":14,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00022148173042756648,"log_rms_mae_db":4.193111329626602,"rms_envelope_nrmse":0.18513181225326383,"envelope_correlation":0.9663040179578786,"rms_gain_error_db":-0.30909972467153685,"reference_mean_power":0.0039398267431541775,"output_mean_power":0.0036691642329710217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.3094240146918987e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011389510990882513,"output_power":0.0010587243112191314,"reference_active":true,"delta_db":-0.3172219223944076},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017715594495014008,"output_power":0.0001725238482454407,"reference_active":true,"delta_db":-0.11506594073052914},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5198627564616767e-05,"output_power":2.319538206471779e-05,"reference_active":true,"delta_db":-0.35975357279933423}],"active_band_count":3,"active_band_mae_db":0.264013811974757}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.00020322184022683357,"log_rms_mae_db":3.6032616618077173,"rms_envelope_nrmse":0.1246501849443193,"envelope_correlation":0.9860282961846059,"rms_gain_error_db":-0.30816218126899525,"reference_mean_power":0.0039603691212238695,"output_mean_power":0.003689091673735873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.1092095592585075e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010485738336262066,"output_power":0.000967351281808346,"reference_active":true,"delta_db":-0.350148045556872},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.703453851742181e-05,"output_power":8.441200807829654e-05,"reference_active":true,"delta_db":-0.1328739907527487},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8525148080015621e-06,"output_power":1.986918171184363e-06,"reference_active":true,"delta_db":0.30418293406008495}],"active_band_count":3,"active_band_mae_db":0.2624016567899019}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00017913726163896074,"log_rms_mae_db":3.1911210951934166,"rms_envelope_nrmse":0.20524258030317677,"envelope_correlation":0.972740824790784,"rms_gain_error_db":-0.4832197516221277,"reference_mean_power":0.001348381907516537,"output_mean_power":0.0012063989246173822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":1.724352265721046e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753679492081999,"output_power":0.0007486782215993732,"reference_active":true,"delta_db":-0.6789544121376752},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012951638097036193,"output_power":0.00011531692410438769,"reference_active":true,"delta_db":-0.5043165082457112},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6972503859652748e-05,"output_power":1.6751638669307916e-05,"reference_active":true,"delta_db":-0.05688619348036621}],"active_band_count":3,"active_band_mae_db":0.4133857046212508}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.0001658599445391014,"log_rms_mae_db":2.428181730619506,"rms_envelope_nrmse":0.12276296575162873,"envelope_correlation":0.9913048113850257,"rms_gain_error_db":-0.4832241577344307,"reference_mean_power":0.001355729112927364,"output_mean_power":0.0012129712475765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":1.6447728018068126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082538722946322,"output_power":0.0007070064875267237,"reference_active":true,"delta_db":-0.672343461895947},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.001034181994635e-05,"output_power":5.3125006037661186e-05,"reference_active":true,"delta_db":-0.5292710810605905},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.081003459949387e-06,"output_power":2.0995675922953036e-06,"reference_active":true,"delta_db":0.03857058345233381}],"active_band_count":3,"active_band_mae_db":0.4133950421362904}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00010266849739942128,"log_rms_mae_db":2.2085557368335182,"rms_envelope_nrmse":0.2315675215515163,"envelope_correlation":0.9641661723881578,"rms_gain_error_db":-0.45249775573694395,"reference_mean_power":0.00024063110594780816,"output_mean_power":0.00021682132372421477},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.4743465614681173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013528556332001237,"output_power":0.00011710920418651178,"reference_active":true,"delta_db":-0.6266042461198772},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4222986289891938e-05,"output_power":1.4014089711756812e-05,"reference_active":true,"delta_db":-0.06425898084934076},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7979674926629104e-06,"output_power":2.6907720022332746e-06,"reference_active":true,"delta_db":-0.16965764311996057}],"active_band_count":3,"active_band_mae_db":0.28684029002972616}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":1.204422368691669,"rms_envelope_nrmse":0.12224735606987003,"envelope_correlation":0.9909317491981916,"rms_gain_error_db":-0.4525018842665535,"reference_mean_power":0.00024194203098044546,"output_mean_power":0.00021800232911211778},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.355449255956096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012689841334701806,"output_power":0.00011039619245099445,"reference_active":true,"delta_db":-0.6050209710414018},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.520296796235236e-06,"output_power":6.505368300582776e-06,"reference_active":true,"delta_db":-0.009954756360522479},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9276932572800109e-07,"output_power":1.7978275532457702e-07,"reference_active":true,"delta_db":-0.3028989620903181}],"active_band_count":3,"active_band_mae_db":0.30595822983074744}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":1.9722005056749612,"rms_envelope_nrmse":0.22459881819739608,"envelope_correlation":0.9653640641473755,"rms_gain_error_db":-0.4995723088334616,"reference_mean_power":0.00013131711251995076,"output_mean_power":0.0001170480259935004},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":6.872487837717031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.115138542486779e-05,"output_power":2.9306412334329915e-05,"reference_active":true,"delta_db":-0.26514710474436765},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2237072790355106e-06,"output_power":2.8629444218997017e-06,"reference_active":true,"delta_db":-0.5154268267933015},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5460080528529544e-06,"output_power":1.3264120422649786e-06,"reference_active":true,"delta_db":-0.6653329563463249}],"active_band_count":3,"active_band_mae_db":0.4819689626279981}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":1.1515797630298454,"rms_envelope_nrmse":0.13129707294089707,"envelope_correlation":0.9891781808932436,"rms_gain_error_db":-0.4995756511539271,"reference_mean_power":0.00013203257062264153,"output_mean_power":0.00011768565092690566},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":6.49214546612456e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9138042963833366e-05,"output_power":2.7446074757294557e-05,"reference_active":true,"delta_db":-0.25980137397881686},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5333654526802435e-06,"output_power":1.2750849323933135e-06,"reference_active":true,"delta_db":-0.8010656046341729},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.807699107498535e-08,"output_power":7.34794618592612e-08,"reference_active":true,"delta_db":-0.786965026498584}],"active_band_count":3,"active_band_mae_db":0.6159440017038579}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.331871271133423,"predicted_mos":3.331871271133423,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":133200,"input_sample_rate":24000,"input_seconds":5.55,"input_float32_sha256":"8ce7f2c5d62b05fff1a7e42f95a59e197694da5290b44c87d899bb3e943fc062","internal_samples":88800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3917506570014666},"started_at":"2026-10-13T05:10:25.913720+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0023989040000742534,"native_input_samples":132960,"native_input_sha256":"be9600739a4f1b9dc7f5b5cedc822386716d2aeb6659f60e15fa2a0639b7f69c","output_sha256":"de8f37aae497e735eab6aa09860c815e0dfed130a415b9a2ce6424346708ecc2","output_samples":133200,"length_delta_samples":240,"analysis_s":0.3290588199997728,"synthesis_s":0.03602361400044174,"total_s":0.36508243400021456,"rtf":0.06589935631772827,"output_peak":0.53885817527771,"output_rms":0.05965204590132433,"finished_at":"2026-10-13T05:10:26.288794+00:00","analysis_stages":{"external_f0_s":0.132386214001599},"evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"de8f37aae497e735eab6aa09860c815e0dfed130a415b9a2ce6424346708ecc2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":88640,"output_samples":88800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.48957512659667,"output_peak":0.5393527832680082},"common_prefix":{"samples":88640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981981981981982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.862245322184609,"spectral_convergence":0.5595364147933787,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.217551735617916,"spectral_convergence":0.4870148279543733,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.088362049317171,"spectral_convergence":0.7740256382875572,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.0007248699990295535,"delta_db":0.003810473177547182,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.0014435542294423882,"delta_db":-0.8320586911785023,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0009750302688870219,"delta_db":-0.031323467692729146,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.0002268053085703197,"delta_db":-0.7016065330134903,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":5.811259296418453e-05,"delta_db":-0.15072799483021074,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.00013470613161282723,"delta_db":-0.6054719495228686,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38749985156922473,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4326401351118645,"pitch":null,"output_rms_dbfs":-24.481746051601675,"output_peak":0.5393527832680082,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.722719702373233},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88640,"output_samples":88640,"intelligibility":{"estoi":{"value":0.7641509996863234,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06274002742130019,"rms_dbfs":-24.04910591648981,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.113070510727493,"mean":-0.006068708578171713,"zero_energy":false},"output":{"rms":0.05969152816458382,"rms_dbfs":-24.481746051601675,"peak":0.5393527832680082,"peak_dbfs":-5.362541515975919,"crest_factor_db":19.119204535625755,"mean":-0.0005258285151560907,"zero_energy":false},"rms_difference_db":-0.43264013511186405},"full_scale":{"reference":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":554,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":547,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":7,"both_estimated_voiced_frames":7,"reference_voiced_output_unvoiced_frames":194,"reference_unvoiced_output_voiced_frames":0,"paired_voiced_fraction":0.03482587064676617,"absolute_difference_cents_median":1202.1811787746904,"absolute_difference_cents_p95":1212.8529003375268,"difference_over_600_cents_frames":4,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00022148173042756648,"log_rms_mae_db":1.7705611194770294,"rms_envelope_nrmse":0.16203915357770018,"envelope_correlation":0.9724197117813785,"rms_gain_error_db":-0.4326959399964774,"reference_mean_power":0.0039398267431541775,"output_mean_power":0.0035662150611033756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.3094240146918987e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011389510990882513,"output_power":0.0011151671616052336,"reference_active":true,"delta_db":-0.09165105781758912},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017715594495014008,"output_power":0.00015871911419375085,"reference_active":true,"delta_db":-0.47726499814295587},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5198627564616767e-05,"output_power":4.329143698916698e-05,"reference_active":true,"delta_db":2.3502511404237096}],"active_band_count":3,"active_band_mae_db":0.9730557321280848}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.00020322184022683357,"log_rms_mae_db":0.9134062934010659,"rms_envelope_nrmse":0.09066023343491778,"envelope_correlation":0.992758161351769,"rms_gain_error_db":-0.4327447920028157,"reference_mean_power":0.0039603691212238695,"output_mean_power":0.0035847690925225107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.1092095592585075e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010485738336262066,"output_power":0.0010334202271967293,"reference_active":true,"delta_db":-0.06322058254717393},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.703453851742181e-05,"output_power":7.623181050334975e-05,"reference_active":true,"delta_db":-0.5755539653039209},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8525148080015621e-06,"output_power":2.531806339802201e-06,"reference_active":true,"delta_db":1.3566879485053516}],"active_band_count":3,"active_band_mae_db":0.6651541654521488}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00017913726163896074,"log_rms_mae_db":2.1055625264523736,"rms_envelope_nrmse":0.20861991786117365,"envelope_correlation":0.9716861370186561,"rms_gain_error_db":0.051502184195272784,"reference_mean_power":0.001348381907516537,"output_mean_power":0.001364467309070009},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":1.724352265721046e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753679492081999,"output_power":0.0009055659277665846,"reference_active":true,"delta_db":0.14729432395013078},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012951638097036193,"output_power":9.646653365491807e-05,"reference_active":true,"delta_db":-1.2794802733848067},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6972503859652748e-05,"output_power":3.543020962856233e-05,"reference_active":true,"delta_db":3.196278058065807}],"active_band_count":3,"active_band_mae_db":1.5410175518002482}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.0001658599445391014,"log_rms_mae_db":1.4745766610309001,"rms_envelope_nrmse":0.11637874396413084,"envelope_correlation":0.9911663205775739,"rms_gain_error_db":0.051500005444307446,"reference_mean_power":0.001355729112927364,"output_mean_power":0.0013719014740615225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":1.6447728018068126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082538722946322,"output_power":0.0008533491827355723,"reference_active":true,"delta_db":0.14469031781199784},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.001034181994635e-05,"output_power":4.323912360144656e-05,"reference_active":true,"delta_db":-1.4234921771330724},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.081003459949387e-06,"output_power":2.583785581435359e-06,"reference_active":true,"delta_db":0.9398366808122295}],"active_band_count":3,"active_band_mae_db":0.8360063919191}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00010266849739942128,"log_rms_mae_db":1.3571216247015843,"rms_envelope_nrmse":0.16602748000754539,"envelope_correlation":0.9826497072036305,"rms_gain_error_db":-0.4830383214917395,"reference_mean_power":0.00024063110594780816,"output_mean_power":0.00021530193595963303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.4743465614681173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013528556332001237,"output_power":0.00011397795007225411,"reference_active":true,"delta_db":-0.7443061256728787},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4222986289891938e-05,"output_power":1.2607214533401272e-05,"reference_active":true,"delta_db":-0.5237164824197439},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7979674926629104e-06,"output_power":3.5726585555608394e-06,"reference_active":true,"delta_db":1.0614884753189873}],"active_band_count":3,"active_band_mae_db":0.7765036944705367}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":0.7759411560260047,"rms_envelope_nrmse":0.09218905658858911,"envelope_correlation":0.9959192578813384,"rms_gain_error_db":-0.4830454651304025,"reference_mean_power":0.00024194203098044546,"output_mean_power":0.00021647451509661133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.355449255956096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012689841334701806,"output_power":0.0001081147165057631,"reference_active":true,"delta_db":-0.6957137813760383},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.520296796235236e-06,"output_power":5.744127400944442e-06,"reference_active":true,"delta_db":-0.5504330105020138},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9276932572800109e-07,"output_power":2.1319843333027316e-07,"reference_active":true,"delta_db":0.43746080705721746}],"active_band_count":3,"active_band_mae_db":0.5612025329784232}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":1.041683529879423,"rms_envelope_nrmse":0.11817417589880959,"envelope_correlation":0.990397720902951,"rms_gain_error_db":-0.14532884203207,"reference_mean_power":0.00013131711251995076,"output_mean_power":0.0001269955316201285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":6.872487837717031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.115138542486779e-05,"output_power":3.170876663495054e-05,"reference_active":true,"delta_db":0.07701983539972208},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2237072790355106e-06,"output_power":2.844517056154956e-06,"reference_active":true,"delta_db":-0.5434705765328849},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5460080528529544e-06,"output_power":1.1223690035313306e-06,"reference_active":true,"delta_db":-1.3907608746933608}],"active_band_count":3,"active_band_mae_db":0.6704170955419894}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":0.5191207010663863,"rms_envelope_nrmse":0.05303935366693503,"envelope_correlation":0.998134807264382,"rms_gain_error_db":-0.14533017413389424,"reference_mean_power":0.00013203257062264153,"output_mean_power":0.00012768740518700892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":6.49214546612456e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9138042963833366e-05,"output_power":2.951514500082516e-05,"reference_active":true,"delta_db":0.05584541817039782},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5333654526802435e-06,"output_power":1.252302684852482e-06,"reference_active":true,"delta_db":-0.8793636252622765},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.807699107498535e-08,"output_power":6.081739212660706e-08,"reference_active":true,"delta_db":-1.6083467616615024}],"active_band_count":3,"active_band_mae_db":0.8478519350313922}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0038","method":"magphase","status":"measured","mos":{"status":"measured","value":3.306969165802002,"predicted_mos":3.306969165802002,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":267325,"input_sample_rate":48000,"input_seconds":5.569270833333333,"input_float32_sha256":"7f5f0083b5bb8dbe9c18589dec55daa7f157178f7284e6b7388772e83a6c8490","internal_samples":89109,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.583489513999666},"started_at":"2026-10-13T05:14:10.101522+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.007831711000108044,"native_input_samples":265920,"native_input_sha256":"25ddc4903cf974020be826d0a806a84419eb450eb4b9d128d2618bed7ab65bb5","output_sha256":"9cff007aa63ca35752d778a993480ce6ad768988d9fdf7e3c439c009d32fd415","output_samples":267325,"length_delta_samples":1405,"analysis_s":2.8726839070004644,"synthesis_s":0.8567658290012332,"total_s":3.7294497360016976,"rtf":0.6731858729244942,"output_peak":0.6246057797484355,"output_rms":0.06035936359092956,"finished_at":"2026-10-13T05:14:13.848966+00:00","evaluation_input":{"reference":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"9cff007aa63ca35752d778a993480ce6ad768988d9fdf7e3c439c009d32fd415","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":88640,"output_samples":89109,"length_delta_samples":469,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.38742819387919,"output_peak":0.6075090385539219},"common_prefix":{"samples":88640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9947367830409948,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88640,"output_samples":88640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.2646762555813,"spectral_convergence":0.7488447469132964,"spectral_convergence_status":"ok","amplitude_floor":1.5935254482600421e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.6800872538255005,"spectral_convergence":0.5157258977362696,"spectral_convergence_status":"ok","amplitude_floor":9.146897175620752e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.899461712218303,"spectral_convergence":0.4871056835451139,"spectral_convergence_status":"ok","amplitude_floor":7.905477970735852e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007242342814613898,"output_power":0.000629641810813254,"delta_db":-0.6078551894245902,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001748392652036385,"output_power":0.0016063884736696211,"delta_db":-0.3678839333367637,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009820880914630636,"output_power":0.0009446572683988978,"delta_db":-0.16876174460034932,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002665715938534463,"output_power":0.00026217825958512755,"delta_db":-0.07217192505915751,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.016487980563797e-05,"output_power":6.371672905743463e-05,"delta_db":0.2491041946571212,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00015485832573279259,"output_power":0.0001537230133996873,"delta_db":-0.03195670273517565,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24962228163552627,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3157202089825677,"pitch":null,"output_rms_dbfs":-24.36482612547238,"output_peak":0.6075090385539219,"reference_rms_dbfs":-24.04910591648981,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.6147417405417},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88640,"output_samples":88640,"intelligibility":{"estoi":{"value":0.5409426617505563,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06274002742130019,"rms_dbfs":-24.04910591648981,"peak":0.566497802734375,"peak_dbfs":-4.936035405762318,"crest_factor_db":19.113070510727493,"mean":-0.006068708578171713,"zero_energy":false},"output":{"rms":0.060500462369998634,"rms_dbfs":-24.36482612547238,"peak":0.6075090385539219,"peak_dbfs":-4.3289451244731225,"crest_factor_db":20.035881000999257,"mean":-0.0015554209169441888,"zero_energy":false},"rms_difference_db":-0.3157202089825688},"full_scale":{"reference":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":554,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":4.905415691318922e-06,"threshold_dbfs":-53.09324184044178,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":547,"reference_estimated_voiced_frames":201,"output_estimated_voiced_frames":93,"both_estimated_voiced_frames":73,"reference_voiced_output_unvoiced_frames":128,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.36318407960199006,"absolute_difference_cents_median":8.910728339516623,"absolute_difference_cents_p95":24.73835302968478,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00022148173042756648,"log_rms_mae_db":3.606784127047849,"rms_envelope_nrmse":0.3373327222746711,"envelope_correlation":0.8766320037472309,"rms_gain_error_db":-0.315720099377299,"reference_mean_power":0.0039398267431541775,"output_mean_power":0.0036635752287954336},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.3094240146918987e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011389510990882513,"output_power":0.0011837294203027712,"reference_active":true,"delta_db":0.16747363633334134},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017715594495014008,"output_power":0.00017211452688370129,"reference_active":true,"delta_db":-0.12538203465054837},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5198627564616767e-05,"output_power":3.071734993504875e-05,"reference_active":true,"delta_db":0.8600685755897644}],"active_band_count":3,"active_band_mae_db":0.384308082191218}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.00020322184022683357,"log_rms_mae_db":2.434105339958739,"rms_envelope_nrmse":0.24528925384176015,"envelope_correlation":0.9304973172206399,"rms_gain_error_db":-0.31515912864644413,"reference_mean_power":0.0039603691212238695,"output_mean_power":0.0036831529387761695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.1092095592585075e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010485738336262066,"output_power":0.0010832019711826829,"reference_active":true,"delta_db":0.1411042563979143},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.703453851742181e-05,"output_power":8.62300146659421e-05,"reference_active":true,"delta_db":-0.04033170939458597},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8525148080015621e-06,"output_power":2.2130799594459394e-06,"reference_active":true,"delta_db":0.7723541726003318}],"active_band_count":3,"active_band_mae_db":0.317930046130944}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00017913726163896074,"log_rms_mae_db":4.90839232623132,"rms_envelope_nrmse":0.4536885622976648,"envelope_correlation":0.8624368123597982,"rms_gain_error_db":-0.16469549789080573,"reference_mean_power":0.001348381907516537,"output_mean_power":0.0012982052669323833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":1.724352265721046e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008753679492081999,"output_power":0.0008456349478190217,"reference_active":true,"delta_db":-0.15007718381796492},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012951638097036193,"output_power":0.00012560330782343103,"reference_active":true,"delta_db":-0.1332362367486688},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6972503859652748e-05,"output_power":1.704781361152309e-05,"reference_active":true,"delta_db":0.019227723623155675}],"active_band_count":3,"active_band_mae_db":0.10084704806326313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":0.0001658599445391014,"log_rms_mae_db":3.6628345137937983,"rms_envelope_nrmse":0.3397836618898321,"envelope_correlation":0.9196550723049413,"rms_gain_error_db":-0.16469575913360512,"reference_mean_power":0.001355729112927364,"output_mean_power":0.0013052789860601584},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":1.6447728018068126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00082538722946322,"output_power":0.0008000464412441044,"reference_active":true,"delta_db":-0.13542547382314227},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.001034181994635e-05,"output_power":5.896087556136853e-05,"reference_active":true,"delta_db":-0.07662176406322911},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.081003459949387e-06,"output_power":2.131390177608912e-06,"reference_active":true,"delta_db":0.10390157739104727}],"active_band_count":3,"active_band_mae_db":0.10531627175913955}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":0.00010266849739942128,"log_rms_mae_db":4.229693598224073,"rms_envelope_nrmse":0.4383350653202409,"envelope_correlation":0.8725913859230456,"rms_gain_error_db":-0.005717554500971432,"reference_mean_power":0.00024063110594780816,"output_mean_power":0.0002403145197897321},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":2.4743465614681173e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013528556332001237,"output_power":0.00013748735691066714,"reference_active":true,"delta_db":0.07011308706950778},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4222986289891938e-05,"output_power":1.6179361942629065e-05,"reference_active":true,"delta_db":0.5597059915568253},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7979674926629104e-06,"output_power":3.4862743162750972e-06,"reference_active":true,"delta_db":0.9551889196151669}],"active_band_count":3,"active_band_mae_db":0.5283359994138334}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":8.271272308600347e-05,"log_rms_mae_db":2.93221927650182,"rms_envelope_nrmse":0.31823217657405034,"envelope_correlation":0.9296356840617821,"rms_gain_error_db":-0.005718317211906419,"reference_mean_power":0.00024194203098044546,"output_mean_power":0.0002416236776705836},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":2.355449255956096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012689841334701806,"output_power":0.00012906798046113794,"reference_active":true,"delta_db":0.07362322660819642},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.520296796235236e-06,"output_power":7.793063202456162e-06,"reference_active":true,"delta_db":0.7744083366601006},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9276932572800109e-07,"output_power":2.4687502247517404e-07,"reference_active":true,"delta_db":1.0743922422196819}],"active_band_count":3,"active_band_mae_db":0.6408079351626595}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1107,"rms_floor":5.887039168463897e-05,"log_rms_mae_db":3.838804207633538,"rms_envelope_nrmse":0.4021648014709487,"envelope_correlation":0.8954023670693839,"rms_gain_error_db":0.41175416814454746,"reference_mean_power":0.00013131711251995076,"output_mean_power":0.0001443765771538054},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1107,"frequency_resolution_hz":0.18066847335140018,"power_floor":6.872487837717031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.115138542486779e-05,"output_power":3.70362338432134e-05,"reference_active":true,"delta_db":0.7514945131572621},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2237072790355106e-06,"output_power":5.282331945974619e-06,"reference_active":true,"delta_db":2.1447008928153273},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5460080528529544e-06,"output_power":2.474210662403293e-06,"reference_active":true,"delta_db":2.0422492237962024}],"active_band_count":3,"active_band_mae_db":1.6461482099229308}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88640,"uncovered_tail_samples":0,"metrics":{"frame_count":1101,"rms_floor":5.469229736232728e-05,"log_rms_mae_db":2.7415415073256555,"rms_envelope_nrmse":0.2926911557336274,"envelope_correlation":0.9425331931572524,"rms_gain_error_db":0.41175333052862945,"reference_mean_power":0.00013203257062264153,"output_mean_power":0.00014516315944646302},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1101,"frequency_resolution_hz":0.18165304268846502,"power_floor":6.49214546612456e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9138042963833366e-05,"output_power":3.540086731087564e-05,"reference_active":true,"delta_db":0.8455352291368099},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5333654526802435e-06,"output_power":2.7385914750213965e-06,"reference_active":true,"delta_db":2.518815776978347},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.807699107498535e-08,"output_power":1.4861313062602318e-07,"reference_active":true,"delta_db":2.271947132966506}],"active_band_count":3,"active_band_mae_db":1.8787660463605544}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"hnm","status":"measured","mos":{"status":"measured","value":2.0631396770477295,"predicted_mos":2.0631396770477295,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48320,"input_sample_rate":16000,"input_seconds":3.02,"input_float32_sha256":"f997ce118b9f136718768b50e6e59cef9a2b3092dbf8d15f8714341473b5e969","internal_samples":48320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.834069364998868},"started_at":"2026-10-13T04:38:48.495150+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.15550001282827e-05,"native_input_samples":48320,"native_input_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","output_sha256":"4fbb63f1a51cfd6681d26abccfa16a9745b4488eca04064edee4cb10e9a76eca","output_samples":48320,"length_delta_samples":0,"analysis_s":0.4525151050002023,"synthesis_s":0.25326648600002954,"total_s":0.7057815910002319,"rtf":0.23370251357623573,"output_peak":0.5557188932077991,"output_rms":0.07156715449910914,"finished_at":"2026-10-13T04:38:49.203776+00:00","evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"4fbb63f1a51cfd6681d26abccfa16a9745b4488eca04064edee4cb10e9a76eca","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.074726364623214,"spectral_convergence":0.31732712077344527,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.702560867017592,"spectral_convergence":0.2776393817115369,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.87452872174537,"spectral_convergence":0.2980876302045022,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.0009266881697493866,"delta_db":-0.07440447671474917,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.0023435829206521043,"delta_db":-0.1368584255850635,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0014144074902782662,"delta_db":-0.011313980687322393,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.00021240655754968874,"delta_db":-0.5247101811427203,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":5.7369493213794356e-05,"delta_db":0.1798392227448366,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.00016740244478840997,"delta_db":-0.5754159059714369,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25042369880768817,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11992058033552652,"pitch":null,"output_rms_dbfs":-22.905724998714632,"output_peak":0.5557188932077991,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.883938651128726},"common_prefix":{"samples":48320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.074726364623214,"spectral_convergence":0.31732712077344527,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.702560867017592,"spectral_convergence":0.2776393817115369,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.87452872174537,"spectral_convergence":0.2980876302045022,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.0009266881697493866,"delta_db":-0.07440447671474917,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.0023435829206521043,"delta_db":-0.1368584255850635,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0014144074902782662,"delta_db":-0.011313980687322393,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.00021240655754968874,"delta_db":-0.5247101811427203,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":5.7369493213794356e-05,"delta_db":0.1798392227448366,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.00016740244478840997,"delta_db":-0.5754159059714369,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25042369880768817,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11992058033552652,"pitch":null,"output_rms_dbfs":-22.905724998714632,"output_peak":0.5557188932077991,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.883938651128726},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48320,"output_samples":48320,"intelligibility":{"estoi":{"value":0.7850917950006866,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.072562089314983,"rms_dbfs":-22.785804418379104,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.662948143418326,"mean":-0.006056492533904828,"zero_energy":false},"output":{"rms":0.07156715449910914,"rms_dbfs":-22.905724998714632,"peak":0.5557188932077991,"peak_dbfs":-5.102896758321156,"crest_factor_db":17.802828240393474,"mean":-7.428385206027003e-05,"zero_energy":false},"rms_difference_db":-0.11992058033552766},"full_scale":{"reference":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":302,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":295,"reference_estimated_voiced_frames":101,"output_estimated_voiced_frames":42,"both_estimated_voiced_frames":40,"reference_voiced_output_unvoiced_frames":61,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.39603960396039606,"absolute_difference_cents_median":1.7870685306379284,"absolute_difference_cents_p95":5.722583421568211,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00023057786100875665,"log_rms_mae_db":1.747058146257836,"rms_envelope_nrmse":0.08613500469300538,"envelope_correlation":0.99277244906796,"rms_gain_error_db":-0.12015453972319746,"reference_mean_power":0.005273848963429562,"output_mean_power":0.005129939390341325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.9211277816293244e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014051118338624634,"output_power":0.001387470781225113,"reference_active":true,"delta_db":-0.054870453523672924},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013648411662722183,"output_power":0.00015050916138919355,"reference_active":true,"delta_db":0.424808227511887},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.767297964659028e-05,"output_power":3.3437253794209266e-05,"reference_active":true,"delta_db":-0.5179936938497719}],"active_band_count":3,"active_band_mae_db":0.3325574582951106}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00020448142520679275,"log_rms_mae_db":0.7895163185823791,"rms_envelope_nrmse":0.04326129730529655,"envelope_correlation":0.9982146301112264,"rms_gain_error_db":-0.120279416073159,"reference_mean_power":0.005325056781288949,"output_mean_power":0.0051796009449380075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.674469134143454e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013611842323655793,"output_power":0.001337326472552054,"reference_active":true,"delta_db":-0.07679468081456778},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.483051409804231e-05,"output_power":7.277391656503085e-05,"reference_active":true,"delta_db":0.5019628300194126},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.671629890449392e-06,"output_power":1.7308159841980689e-06,"reference_active":true,"delta_db":0.15110769113138733}],"active_band_count":3,"active_band_mae_db":0.24328840065512256}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00020294603022145288,"log_rms_mae_db":2.9620398403418724,"rms_envelope_nrmse":0.09652610464974504,"envelope_correlation":0.9939579216821233,"rms_gain_error_db":0.005252007984865426,"reference_mean_power":0.002254450133685994,"output_mean_power":0.002257178133439997},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.5930281169536517e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007344814515063544,"output_power":0.0007383514589500462,"reference_active":true,"delta_db":0.022823047120258844},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010026570767730039,"output_power":0.00010017123827946269,"reference_active":true,"delta_db":-0.004093810253568468},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.614259763381322e-05,"output_power":1.8336203947213296e-05,"reference_active":true,"delta_db":0.5533600903064064}],"active_band_count":3,"active_band_mae_db":0.19342564922674457}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00017314576936579135,"log_rms_mae_db":2.2128773445040335,"rms_envelope_nrmse":0.05076479481631268,"envelope_correlation":0.998382874591162,"rms_gain_error_db":0.00523512699740563,"reference_mean_power":0.002277104155750591,"output_mean_power":0.0022798507062513885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.4781762283606311e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007359452368517945,"output_power":0.0007422783406336475,"reference_active":true,"delta_db":0.03721289385260631},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.796294327084698e-05,"output_power":4.734360560988896e-05,"reference_active":true,"delta_db":-0.05644495955694165},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4655779178142778e-06,"output_power":1.3822239293521361e-06,"reference_active":true,"delta_db":-0.2543050546664537}],"active_band_count":3,"active_band_mae_db":0.11598763602533389}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":1.9069891701901291,"rms_envelope_nrmse":0.15503511729351582,"envelope_correlation":0.9842401467892334,"rms_gain_error_db":-0.3796699118250312,"reference_mean_power":0.0002815001552852329,"output_mean_power":0.00025793581406901976},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.4039983283954285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014419558923291264,"output_power":0.00012770526145847245,"reference_active":true,"delta_db":-0.5274318547797261},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9143053212884863e-05,"output_power":1.3714908727463201e-05,"reference_active":true,"delta_db":-1.448182848473122},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.190405195331928e-06,"output_power":3.127830587508791e-06,"reference_active":true,"delta_db":-1.2701279716666003}],"active_band_count":3,"active_band_mae_db":1.0819142249731495}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":1.3924047921270168,"rms_envelope_nrmse":0.10391559526448546,"envelope_correlation":0.9933753376544742,"rms_gain_error_db":-0.37970500187310685,"reference_mean_power":0.00028432562712779965,"output_mean_power":0.0002605226610579872},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.2325770901894177e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013845430163824743,"output_power":0.00012133754464366291,"reference_active":true,"delta_db":-0.5731125096936376},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.792343358100621e-06,"output_power":6.7958254144781995e-06,"reference_active":true,"delta_db":-1.5864441987631213},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.075134841414049e-07,"output_power":1.9754812399271805e-07,"reference_active":true,"delta_db":-1.9219125410504365}],"active_band_count":3,"active_band_mae_db":1.3604897498357318}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":1.814766166978385,"rms_envelope_nrmse":0.1590806973174039,"envelope_correlation":0.9838000232438276,"rms_gain_error_db":-0.42724661191805713,"reference_mean_power":0.00010138440655269987,"output_mean_power":9.188539829376848e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.3079247184162875e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.16171570932163e-05,"output_power":4.674438729563601e-05,"reference_active":true,"delta_db":-0.4306460970734599},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0705364401401468e-05,"output_power":1.3902774496180246e-05,"reference_active":true,"delta_db":-1.7298139967800146},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9746890612071323e-06,"output_power":2.62550384379909e-06,"reference_active":true,"delta_db":-0.5422891799032846}],"active_band_count":3,"active_band_mae_db":0.9009164245855863}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":1.2722816454805497,"rms_envelope_nrmse":0.09255739422427342,"envelope_correlation":0.9954095893551599,"rms_gain_error_db":-0.427272648484259,"reference_mean_power":0.00010240265187942673,"output_mean_power":9.280768477376481e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.2076196938321018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.154744016518147e-05,"output_power":5.532956116432962e-05,"reference_active":true,"delta_db":-0.46252769088195667},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.945675894844706e-06,"output_power":6.675640707970613e-06,"reference_active":true,"delta_db":-1.7314134821741174},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53780713465196e-07,"output_power":2.2418701443529666e-07,"reference_active":true,"delta_db":-0.5384816061375914}],"active_band_count":3,"active_band_mae_db":0.9108075930645553}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.4884068965911865,"predicted_mos":3.4884068965911865,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48320,"input_sample_rate":16000,"input_seconds":3.02,"input_float32_sha256":"737ae9e9230fe23eb86a5107568aeb16322017bb3b71041d9a0d33d2bab1145d","internal_samples":48320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.416579332000765},"started_at":"2026-10-13T04:41:44.780875+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0587000108207576e-05,"native_input_samples":48320,"native_input_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","output_sha256":"ad4c4b49b866d5c1c5268b01127c74f941553cf58db0d47aa9f70d6d3a00e6bb","output_samples":48320,"length_delta_samples":0,"analysis_s":0.5437165120001737,"synthesis_s":0.04770747800012032,"total_s":0.591423990000294,"rtf":0.19583575827824304,"output_peak":0.7078521145037081,"output_rms":0.08030978259555756,"finished_at":"2026-10-13T04:41:45.376362+00:00","evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"ad4c4b49b866d5c1c5268b01127c74f941553cf58db0d47aa9f70d6d3a00e6bb","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.619620509586981,"spectral_convergence":0.5789444649121818,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.916954011896045,"spectral_convergence":0.3335288911934364,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.067036487368575,"spectral_convergence":0.36011341165799077,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.0007930239888832558,"delta_db":-0.7508774158080607,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.003293074919015659,"delta_db":1.3403544433549694,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0018172933967998304,"delta_db":1.077191029282105,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.0002979339813199412,"delta_db":0.9448110162252911,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":6.304412753515435e-05,"delta_db":0.589475477331073,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.00018527913509018697,"delta_db":-0.1347687214655544,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8062463505778421,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8811734232056914,"pitch":null,"output_rms_dbfs":-21.904630995173413,"output_peak":0.7078521145037081,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.2012036696172},"common_prefix":{"samples":48320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.619620509586981,"spectral_convergence":0.5789444649121818,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.916954011896045,"spectral_convergence":0.3335288911934364,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.067036487368575,"spectral_convergence":0.36011341165799077,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.0007930239888832558,"delta_db":-0.7508774158080607,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.003293074919015659,"delta_db":1.3403544433549694,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0018172933967998304,"delta_db":1.077191029282105,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.0002979339813199412,"delta_db":0.9448110162252911,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":6.304412753515435e-05,"delta_db":0.589475477331073,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.00018527913509018697,"delta_db":-0.1347687214655544,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8062463505778421,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8811734232056914,"pitch":null,"output_rms_dbfs":-21.904630995173413,"output_peak":0.7078521145037081,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.2012036696172},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48320,"output_samples":48320,"intelligibility":{"estoi":{"value":0.8305856795817906,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.072562089314983,"rms_dbfs":-22.785804418379104,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.662948143418326,"mean":-0.006056492533904828,"zero_energy":false},"output":{"rms":0.08030978259555754,"rms_dbfs":-21.904630995173413,"peak":0.7078521145037081,"peak_dbfs":-3.001149325402517,"crest_factor_db":18.903481669770898,"mean":-1.0306439534681859e-07,"zero_energy":false},"rms_difference_db":0.8811734232056914},"full_scale":{"reference":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":302,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":295,"reference_estimated_voiced_frames":101,"output_estimated_voiced_frames":69,"both_estimated_voiced_frames":40,"reference_voiced_output_unvoiced_frames":61,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.39603960396039606,"absolute_difference_cents_median":5.712434905644632,"absolute_difference_cents_p95":15.013505293519415,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00023057786100875665,"log_rms_mae_db":4.2346674560916036,"rms_envelope_nrmse":0.19118935077048146,"envelope_correlation":0.9846386833492449,"rms_gain_error_db":0.881274838816545,"reference_mean_power":0.005273848963429562,"output_mean_power":0.006460336978159417},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.9211277816293244e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014051118338624634,"output_power":0.0020320119007378313,"reference_active":true,"delta_db":1.6021535569334644},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013648411662722183,"output_power":0.00015737126174544702,"reference_active":true,"delta_db":0.6184331365853052},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.767297964659028e-05,"output_power":3.8009759832349365e-05,"reference_active":true,"delta_db":0.03865154530919432}],"active_band_count":3,"active_band_mae_db":0.7530794129426547}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00020448142520679275,"log_rms_mae_db":3.6558732831470784,"rms_envelope_nrmse":0.15898785317391664,"envelope_correlation":0.9939452517544457,"rms_gain_error_db":0.8826140879238998,"reference_mean_power":0.005325056781288949,"output_mean_power":0.006525077162853467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.674469134143454e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013611842323655793,"output_power":0.0019834153064627064,"reference_active":true,"delta_db":1.6349675076975232},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.483051409804231e-05,"output_power":7.357482739970394e-05,"reference_active":true,"delta_db":0.5494978643622777},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.671629890449392e-06,"output_power":1.7403954797881304e-06,"reference_active":true,"delta_db":0.17507818488534319}],"active_band_count":3,"active_band_mae_db":0.7865145189817148}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00020294603022145288,"log_rms_mae_db":6.5715396124042265,"rms_envelope_nrmse":0.1661320435965711,"envelope_correlation":0.9907641038881666,"rms_gain_error_db":0.8852151378596536,"reference_mean_power":0.002254450133685994,"output_mean_power":0.0027641534400147695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.5930281169536517e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007344814515063544,"output_power":0.0008541135377811333,"reference_active":true,"delta_db":0.6553477287858607},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010026570767730039,"output_power":0.000116209313718859,"reference_active":true,"delta_db":0.6408851302279394},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.614259763381322e-05,"output_power":1.781727364263189e-05,"reference_active":true,"delta_db":0.42867828273194847}],"active_band_count":3,"active_band_mae_db":0.5749703805819161}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00017314576936579135,"log_rms_mae_db":6.106051259015937,"rms_envelope_nrmse":0.13601984941119258,"envelope_correlation":0.9960951703234779,"rms_gain_error_db":0.885086330850948,"reference_mean_power":0.002277104155750591,"output_mean_power":0.0027918464520752624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.4781762283606311e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007359452368517945,"output_power":0.0008594701343999598,"reference_active":true,"delta_db":0.6738529122682597},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.796294327084698e-05,"output_power":5.55067990319132e-05,"reference_active":true,"delta_db":0.6344035719384287},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4655779178142778e-06,"output_power":1.3722587429195733e-06,"reference_active":true,"delta_db":-0.28572906327710185}],"active_band_count":3,"active_band_mae_db":0.5313285158279301}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":1.6020884707768708,"rms_envelope_nrmse":0.23844087174657763,"envelope_correlation":0.977277410254469,"rms_gain_error_db":1.1114349416033855,"reference_mean_power":0.0002815001552852329,"output_mean_power":0.0003635985418753205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.4039983283954285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014419558923291264,"output_power":0.00018069123416248974,"reference_active":true,"delta_db":0.9798510818266555},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9143053212884863e-05,"output_power":2.4436416101865608e-05,"reference_active":true,"delta_db":1.0602630509057223},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.190405195331928e-06,"output_power":7.345843626255217e-06,"reference_active":true,"delta_db":2.4378565959039125}],"active_band_count":3,"active_band_mae_db":1.49265690954543}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":1.0581616921443957,"rms_envelope_nrmse":0.1860131422724434,"envelope_correlation":0.9900668450905452,"rms_gain_error_db":1.1114355616506608,"reference_mean_power":0.00028432562712779965,"output_mean_power":0.0003672481036340503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.2325770901894177e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013845430163824743,"output_power":0.00017533673558449756,"reference_active":true,"delta_db":1.025664632171102},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.792343358100621e-06,"output_power":1.2258357834074258e-05,"reference_active":true,"delta_db":0.9754566156139527},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.075134841414049e-07,"output_power":5.008279830756424e-07,"reference_active":true,"delta_db":2.1182442260477234}],"active_band_count":3,"active_band_mae_db":1.3731218246109258}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":1.6473426572182095,"rms_envelope_nrmse":0.22958670023366104,"envelope_correlation":0.9640753652664003,"rms_gain_error_db":0.0008340972166448214,"reference_mean_power":0.00010138440655269987,"output_mean_power":0.00010140388010716573},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.3079247184162875e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.16171570932163e-05,"output_power":5.457788089491473e-05,"reference_active":true,"delta_db":0.24222587907085052},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0705364401401468e-05,"output_power":1.5657796289787573e-05,"reference_active":true,"delta_db":-1.2135223968468591},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9746890612071323e-06,"output_power":3.962249129336117e-06,"reference_active":true,"delta_db":1.245002022157036}],"active_band_count":3,"active_band_mae_db":0.9002500993582485}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":1.0195966757136556,"rms_envelope_nrmse":0.15186609906186535,"envelope_correlation":0.983305680406652,"rms_gain_error_db":0.0008298212953230016,"reference_mean_power":0.00010240265187942673,"output_mean_power":0.00010242222017320046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.2076196938321018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.154744016518147e-05,"output_power":6.479731161287496e-05,"reference_active":true,"delta_db":0.22346992930120702},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.945675894844706e-06,"output_power":7.177605821450323e-06,"reference_active":true,"delta_db":-1.4165469839529714},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53780713465196e-07,"output_power":2.549285825103653e-07,"reference_active":true,"delta_db":0.019599171561052214}],"active_band_count":3,"active_band_mae_db":0.5532053616050768}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"facodec","status":"measured","mos":{"status":"measured","value":4.7093305587768555,"predicted_mos":4.7093305587768555,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48400,"input_sample_rate":16000,"input_seconds":3.025,"input_float32_sha256":"ffdf667bfd62b3cfa05ca8471053fbde72513a3f43f48a33a34edb497c78e310","internal_samples":48400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.310059519999413},"started_at":"2026-10-13T04:49:53.689333+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3535000764240976e-05,"native_input_samples":48320,"native_input_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","output_sha256":"9e30a68e14ceeca104ce478756fb2ebee90a630f1653e052bef1e7f19a5d8c2e","output_samples":48400,"length_delta_samples":80,"analysis_s":0.7375323080004819,"synthesis_s":5.28944378199958,"total_s":6.026976090000062,"rtf":1.995687447019888,"output_peak":0.5532426834106445,"output_rms":0.07202405563550875,"finished_at":"2026-10-13T04:49:59.723470+00:00","evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"9e30a68e14ceeca104ce478756fb2ebee90a630f1653e052bef1e7f19a5d8c2e","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":48320,"output_samples":48400,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.850448547789235,"output_peak":0.5532426834106445},"common_prefix":{"samples":48320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983471074380166,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.667767550110912,"spectral_convergence":0.2870821033207282,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.0336037605183135,"spectral_convergence":0.28565946854850416,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.0738050157057435,"spectral_convergence":0.30695488679594957,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.0009369585971351576,"delta_db":-0.026536659969133344,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.0025568237745688793,"delta_db":0.24134629036065042,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0013092189006585428,"delta_db":-0.3469367922398772,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.00021401230635402892,"delta_db":-0.49200191233098456,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":4.409887603708051e-05,"delta_db":-0.9626957058876161,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.00013491896302759418,"delta_db":-1.5123039225908106,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5969702138965121,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05747789791195889,"pitch":null,"output_rms_dbfs":-22.843282316291067,"output_peak":0.5532426834106445,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.925058775444989},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48320,"output_samples":48320,"intelligibility":{"estoi":{"value":0.8099621527160595,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.072562089314983,"rms_dbfs":-22.785804418379104,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.662948143418326,"mean":-0.006056492533904828,"zero_energy":false},"output":{"rms":0.07208350309057203,"rms_dbfs":-22.843282316291067,"peak":0.5532426834106445,"peak_dbfs":-5.141686417436404,"crest_factor_db":17.70159589885466,"mean":-5.66267151175332e-05,"zero_energy":false},"rms_difference_db":-0.05747789791196212},"full_scale":{"reference":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":302,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":295,"reference_estimated_voiced_frames":101,"output_estimated_voiced_frames":78,"both_estimated_voiced_frames":73,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.7227722772277227,"absolute_difference_cents_median":0.20738963289009885,"absolute_difference_cents_p95":11.58723136783156,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00023057786100875665,"log_rms_mae_db":2.278679614428859,"rms_envelope_nrmse":0.13216034326069487,"envelope_correlation":0.983891043143037,"rms_gain_error_db":-0.05743287209327256,"reference_mean_power":0.005273848963429562,"output_mean_power":0.0052045645678462545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.9211277816293244e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014051118338624634,"output_power":0.0013178686254137453,"reference_active":true,"delta_db":-0.2783877259685348},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013648411662722183,"output_power":0.00015323880370413616,"reference_active":true,"delta_db":0.5028663970930352},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.767297964659028e-05,"output_power":3.6329915246267023e-05,"reference_active":true,"delta_db":-0.1576558625488469}],"active_band_count":3,"active_band_mae_db":0.3129699952034723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00020448142520679275,"log_rms_mae_db":2.059719019676124,"rms_envelope_nrmse":0.11793136884881746,"envelope_correlation":0.9865527879178265,"rms_gain_error_db":-0.05651090050079654,"reference_mean_power":0.005325056781288949,"output_mean_power":0.005256215383733437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.674469134143454e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013611842323655793,"output_power":0.00129465652036166,"reference_active":true,"delta_db":-0.21762346752589745},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.483051409804231e-05,"output_power":7.046388099371916e-05,"reference_active":true,"delta_db":0.3618709395869416},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.671629890449392e-06,"output_power":1.8925454412410076e-06,"reference_active":true,"delta_db":0.5390618781908559}],"active_band_count":3,"active_band_mae_db":0.3728520951012317}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00020294603022145288,"log_rms_mae_db":2.2600536722584765,"rms_envelope_nrmse":0.09198115377520542,"envelope_correlation":0.995050694489257,"rms_gain_error_db":0.23726149065584323,"reference_mean_power":0.002254450133685994,"output_mean_power":0.0023810405004714343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.5930281169536517e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007344814515063544,"output_power":0.0007799281920356993,"reference_active":true,"delta_db":0.2607378640010899},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010026570767730039,"output_power":0.00012692129474839277,"reference_active":true,"delta_db":1.0238207035313638},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.614259763381322e-05,"output_power":2.0496238780690805e-05,"reference_active":true,"delta_db":1.0370075016053861}],"active_band_count":3,"active_band_mae_db":0.7738553563792799}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00017314576936579135,"log_rms_mae_db":1.8857054194034755,"rms_envelope_nrmse":0.07590333309644191,"envelope_correlation":0.9966545930383663,"rms_gain_error_db":0.23726332641139786,"reference_mean_power":0.002277104155750591,"output_mean_power":0.002404967592513676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.4781762283606311e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007359452368517945,"output_power":0.0007805747664484572,"reference_active":true,"delta_db":0.2556900895991703},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.796294327084698e-05,"output_power":6.172324291457796e-05,"reference_active":true,"delta_db":1.0954290969920024},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4655779178142778e-06,"output_power":1.9691314266852787e-06,"reference_active":true,"delta_db":1.2826579066850685}],"active_band_count":3,"active_band_mae_db":0.8779256977587471}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":2.0018200298530378,"rms_envelope_nrmse":0.19853978941145997,"envelope_correlation":0.9763456093530506,"rms_gain_error_db":-0.8012611489066412,"reference_mean_power":0.0002815001552852329,"output_mean_power":0.00023407364814652992},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.4039983283954285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014419558923291264,"output_power":0.00011560079625798257,"reference_active":true,"delta_db":-0.9599115053436027},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9143053212884863e-05,"output_power":1.861535931864068e-05,"reference_active":true,"delta_db":-0.12139783069950397},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.190405195331928e-06,"output_power":5.006779212346719e-06,"reference_active":true,"delta_db":0.7730242088876218}],"active_band_count":3,"active_band_mae_db":0.6181111816435761}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":1.6190321995210784,"rms_envelope_nrmse":0.15320083535348183,"envelope_correlation":0.9876158580564736,"rms_gain_error_db":-0.8012715993401949,"reference_mean_power":0.00028432562712779965,"output_mean_power":0.00023642252200257167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.2325770901894177e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013845430163824743,"output_power":0.00011150008509921085,"reference_active":true,"delta_db":-0.9403125454433403},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.792343358100621e-06,"output_power":8.942788683893286e-06,"reference_active":true,"delta_db":-0.3941366455081329},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.075134841414049e-07,"output_power":3.6268017897006716e-07,"reference_active":true,"delta_db":0.7165965748400261}],"active_band_count":3,"active_band_mae_db":0.6836819219304998}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":1.9417648381386061,"rms_envelope_nrmse":0.25686683713124575,"envelope_correlation":0.9557774780009934,"rms_gain_error_db":-0.7180463352248119,"reference_mean_power":0.00010138440655269987,"output_mean_power":8.59342972706581e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.3079247184162875e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.16171570932163e-05,"output_power":4.110763391705414e-05,"reference_active":true,"delta_db":-0.9887160105018579},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0705364401401468e-05,"output_power":1.4053819444334534e-05,"reference_active":true,"delta_db":-1.6828850855282758},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9746890612071323e-06,"output_power":6.9074852284907045e-06,"reference_active":true,"delta_db":3.6587838842893117}],"active_band_count":3,"active_band_mae_db":2.1101283267731485}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":1.4902555428796327,"rms_envelope_nrmse":0.1893280086600426,"envelope_correlation":0.9767914739505468,"rms_gain_error_db":-0.7180666278100162,"reference_mean_power":0.00010240265187942673,"output_mean_power":8.679696522703567e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.2076196938321018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.154744016518147e-05,"output_power":4.755383986368084e-05,"reference_active":true,"delta_db":-1.1202440381933392},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.945675894844706e-06,"output_power":5.911595948611389e-06,"reference_active":true,"delta_db":-2.259295597076857},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53780713465196e-07,"output_power":6.17357867826051e-07,"reference_active":true,"delta_db":3.8607837331787986}],"active_band_count":3,"active_band_mae_db":2.4134411228163315}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.4094760417938232,"predicted_mos":3.4094760417938232,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":133632,"input_sample_rate":44100,"input_seconds":3.030204081632653,"input_float32_sha256":"fbb8d1693cd33699d20c21a4fda63f660cb19f731cd929b69217b2d88d9ae2a0","internal_samples":48484,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8501256859999557},"started_at":"2026-10-13T07:09:33.251000+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005714716999136726,"native_input_samples":133182,"native_input_sha256":"23e276cb32c7a19349fda04eab2e9f51547ccc3b064aa42527ead97e3b60288a","output_sha256":"0de851193c9a2659fcc7999a6dad8d858be54a3aa74d5071241422fdac82e52a","output_samples":133632,"length_delta_samples":450,"analysis_s":0.07247499500044796,"synthesis_s":18.01059036799961,"total_s":18.08306536300006,"rtf":5.987769987748364,"output_peak":0.5525949597358704,"output_rms":0.06394349560394884,"finished_at":"2026-10-13T07:09:51.350552+00:00","evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"0de851193c9a2659fcc7999a6dad8d858be54a3aa74d5071241422fdac82e52a","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":48320,"output_samples":48484,"length_delta_samples":164,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.888204687860615,"output_peak":0.5493287899957838},"common_prefix":{"samples":48320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9966174408052141,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.707280349178497,"spectral_convergence":0.49852606947879213,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.789078523893396,"spectral_convergence":0.5030682656584755,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.5038069860735,"spectral_convergence":0.5701206404355413,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.001221195514305399,"delta_db":1.1241113392474753,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.001199948738040434,"delta_db":-3.0440347356462496,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0014013516931475124,"delta_db":-0.05158803052756399,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.0001336954832016855,"delta_db":-2.5352220333757023,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":4.351452404515021e-05,"delta_db":-1.0206285354650881,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":9.902176261932579e-05,"delta_db":-2.8557273420296396,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.77188533604862,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.08770475405559,"pitch":null,"output_rms_dbfs":-23.873509172434698,"output_peak":0.5493287899957838,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.666721953048464},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48320,"output_samples":48320,"intelligibility":{"estoi":{"value":0.6162497952290834,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.072562089314983,"rms_dbfs":-22.785804418379104,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.662948143418326,"mean":-0.006056492533904828,"zero_energy":false},"output":{"rms":0.06402130778479996,"rms_dbfs":-23.873509172434698,"peak":0.5493287899957838,"peak_dbfs":-5.203352785323742,"crest_factor_db":18.670156387110953,"mean":0.001914096212773041,"zero_energy":false},"rms_difference_db":-1.0877047540555935},"full_scale":{"reference":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":302,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":295,"reference_estimated_voiced_frames":101,"output_estimated_voiced_frames":52,"both_estimated_voiced_frames":40,"reference_voiced_output_unvoiced_frames":61,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.39603960396039606,"absolute_difference_cents_median":11.129235362925954,"absolute_difference_cents_p95":26.81880097584485,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00023057786100875665,"log_rms_mae_db":2.937913095664483,"rms_envelope_nrmse":0.16932191341260747,"envelope_correlation":0.9869900194770239,"rms_gain_error_db":-1.0877164415518619,"reference_mean_power":0.005273848963429562,"output_mean_power":0.0041054053489032695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.9211277816293244e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014051118338624634,"output_power":0.0011857791021703422,"reference_active":true,"delta_db":-0.7370709923179479},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013648411662722183,"output_power":0.00011357474518305253,"reference_active":true,"delta_db":-0.7980034208765657},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.767297964659028e-05,"output_power":2.7362250226925924e-05,"reference_active":true,"delta_db":-1.3887816076895978}],"active_band_count":3,"active_band_mae_db":0.974618673628037}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00020448142520679275,"log_rms_mae_db":2.5601982019319256,"rms_envelope_nrmse":0.14764611750396656,"envelope_correlation":0.9941064669789365,"rms_gain_error_db":-1.0873144711358653,"reference_mean_power":0.005325056781288949,"output_mean_power":0.004145651549208077},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.674469134143454e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013611842323655793,"output_power":0.0011571519052923193,"reference_active":true,"delta_db":-0.7052653492055752},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.483051409804231e-05,"output_power":5.500368851031094e-05,"reference_active":true,"delta_db":-0.7138765155236517},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.671629890449392e-06,"output_power":1.3751516190834737e-06,"reference_active":true,"delta_db":-0.8478954367544923}],"active_band_count":3,"active_band_mae_db":0.755679100494573}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00020294603022145288,"log_rms_mae_db":4.208662876186938,"rms_envelope_nrmse":0.164188492335591,"envelope_correlation":0.9860781521139613,"rms_gain_error_db":-0.7464564182098404,"reference_mean_power":0.002254450133685994,"output_mean_power":0.0018984317639374603},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.5930281169536517e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007344814515063544,"output_power":0.000609090366724564,"reference_active":true,"delta_db":-0.8129910182764527},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010026570767730039,"output_power":7.130334100258848e-05,"reference_active":true,"delta_db":-1.4804254370734582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.614259763381322e-05,"output_power":1.5769837129886482e-05,"reference_active":true,"delta_db":-0.10146213810148826}],"active_band_count":3,"active_band_mae_db":0.7982928644837997}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00017314576936579135,"log_rms_mae_db":3.8712249390096383,"rms_envelope_nrmse":0.13827497557926302,"envelope_correlation":0.9913778950962329,"rms_gain_error_db":-0.7464535068840397,"reference_mean_power":0.002277104155750591,"output_mean_power":0.0019175095924213541},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.4781762283606311e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007359452368517945,"output_power":0.000607986449391188,"reference_active":true,"delta_db":-0.8295159886082377},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.796294327084698e-05,"output_power":3.7704001566434194e-05,"reference_active":true,"delta_db":-1.0451838115571181},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4655779178142778e-06,"output_power":1.4110092317017622e-06,"reference_active":true,"delta_db":-0.16479057573524372}],"active_band_count":3,"active_band_mae_db":0.6798301253001999}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":4.876032052678084,"rms_envelope_nrmse":0.3982317269283613,"envelope_correlation":0.898928055887043,"rms_gain_error_db":-1.9487163637171805,"reference_mean_power":0.0002815001552852329,"output_mean_power":0.00017972438343426448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.4039983283954285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014419558923291264,"output_power":0.0001005038536439806,"reference_active":true,"delta_db":-1.5676926171252248},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9143053212884863e-05,"output_power":1.95246104935007e-05,"reference_active":true,"delta_db":0.0857117209828734},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.190405195331928e-06,"output_power":4.3173426317572994e-06,"reference_active":true,"delta_db":0.12960496856306566}],"active_band_count":3,"active_band_mae_db":0.5943364355570546}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":4.265371862814009,"rms_envelope_nrmse":0.3628099412492711,"envelope_correlation":0.9177234509062966,"rms_gain_error_db":-1.9487018957208766,"reference_mean_power":0.00028432562712779965,"output_mean_power":0.00018152891678839913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.2325770901894177e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013845430163824743,"output_power":9.968708755167592e-05,"reference_active":true,"delta_db":-1.4266754551325354},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.792343358100621e-06,"output_power":1.0816189502562416e-05,"reference_active":true,"delta_db":0.4318765447849713},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.075134841414049e-07,"output_power":2.7400764373930087e-07,"reference_active":true,"delta_db":-0.5010148577995887}],"active_band_count":3,"active_band_mae_db":0.7865222859056984}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":5.746122791036667,"rms_envelope_nrmse":0.33333840986798635,"envelope_correlation":0.928880646454435,"rms_gain_error_db":-1.3243351673090915,"reference_mean_power":0.00010138440655269987,"output_mean_power":7.473734171776724e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.3079247184162875e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.16171570932163e-05,"output_power":5.65285099157928e-05,"reference_active":true,"delta_db":0.39473456339569757},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0705364401401468e-05,"output_power":7.4894305927715494e-06,"reference_active":true,"delta_db":-4.416340778482606},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9746890612071323e-06,"output_power":2.832830980374284e-06,"reference_active":true,"delta_db":-0.21220913128707342}],"active_band_count":3,"active_band_mae_db":1.6744281577217925}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":5.611444512227115,"rms_envelope_nrmse":0.2712779283674377,"envelope_correlation":0.9557623759314321,"rms_gain_error_db":-1.3243119204414553,"reference_mean_power":0.00010240265187942673,"output_mean_power":7.548836367613445e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.2076196938321018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.154744016518147e-05,"output_power":6.020190580224e-05,"reference_active":true,"delta_db":-0.0959975492627345},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.945675894844706e-06,"output_power":4.132698979049493e-06,"reference_active":true,"delta_db":-3.8140052960345194},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53780713465196e-07,"output_power":2.251441237408323e-07,"reference_active":true,"delta_db":-0.5199799763453804}],"active_band_count":3,"active_band_mae_db":1.4766609405475448}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"golf","status":"measured","mos":{"status":"measured","value":2.052809476852417,"predicted_mos":2.052809476852417,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":72240,"input_sample_rate":24000,"input_seconds":3.01,"input_float32_sha256":"34463fa1ec6a26d877df8286631a6f39b31687725a9037e49a0b67cb8498b2d2","internal_samples":48160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.308287769999879},"started_at":"2026-10-13T05:02:18.849297+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0014911009993738844,"native_input_samples":72480,"native_input_sha256":"509a44b8d544e5c2f8b97001c669b620becc00ff21ae0fc249ce76be179b998f","output_sha256":"b42a6a03cca11e8cc38106b76dd5703fc2f250d95271b873411cd3ff00e053f8","output_samples":72240,"length_delta_samples":-240,"analysis_s":0.30163190299936105,"synthesis_s":0.06663630600087345,"total_s":0.3682682090002345,"rtf":0.12194311556299155,"output_peak":0.4003700911998749,"output_rms":0.05083100465859443,"finished_at":"2026-10-13T05:02:19.224475+00:00","analysis_stages":{"external_f0_s":0.039221081000505364},"evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"b42a6a03cca11e8cc38106b76dd5703fc2f250d95271b873411cd3ff00e053f8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":48320,"output_samples":48160,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.88834305385531,"output_peak":0.40063831209428347},"common_prefix":{"samples":48160,"sample_rate":16000,"reference_coverage":0.9966887417218543,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48160,"output_samples":48160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.441434356331762,"spectral_convergence":0.6749467549464857,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.981838743738206,"spectral_convergence":0.6400134009759045,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.07497802214807,"spectral_convergence":0.7745460555601255,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009459220878466249,"output_power":0.0005296036804513766,"delta_db":-2.519043720744036,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002425918679219363,"output_power":0.0010764352242822824,"delta_db":-3.528883377299788,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014228282099542566,"output_power":0.0005254904088555214,"delta_db":-4.325876734334852,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00024035144042611935,"output_power":0.00022086461423005633,"delta_db":-0.36720588138232124,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.522060775126978e-05,"output_power":5.1539649943576995e-05,"delta_db":-0.2995971761742162,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019175395579434218,"output_power":0.00017337070014594342,"delta_db":-0.4376862894656696,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.9130488632334808,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-3.1163228958537896,"pitch":null,"output_rms_dbfs":-25.88834305385531,"output_peak":0.40063831209428347,"reference_rms_dbfs":-22.772020158001524,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.166083707406012},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48160,"output_samples":48160,"intelligibility":{"estoi":{"value":0.7174441315405855,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07267733471304501,"rms_dbfs":-22.772020158001524,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.649163883040735,"mean":-0.0060354771408131745,"zero_energy":false},"output":{"rms":0.050767157473200694,"rms_dbfs":-25.88834305385531,"peak":0.40063831209428347,"peak_dbfs":-7.944950450047511,"crest_factor_db":17.9433926038078,"mean":-0.00020234282450410398,"zero_energy":false},"rms_difference_db":-3.1163228958537843},"full_scale":{"reference":{"sample_count":48160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":301,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":294,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":18,"both_estimated_voiced_frames":15,"reference_voiced_output_unvoiced_frames":85,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.15,"absolute_difference_cents_median":11.247476429263756,"absolute_difference_cents_p95":22.254410236455733,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00023057786100875665,"log_rms_mae_db":3.8590379375851738,"rms_envelope_nrmse":0.4378192238664951,"envelope_correlation":0.8513813413345934,"rms_gain_error_db":-3.116343748682292,"reference_mean_power":0.005290692347507094,"output_mean_power":0.0025815356881378113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.9394209377456444e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001413844026269174,"output_power":0.000665842620470696,"reference_active":true,"delta_db":-3.270299103850029},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013735372585208148,"output_power":0.0001255164948273398,"reference_active":true,"delta_db":-0.39139641928507896},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.772285033635853e-05,"output_power":8.734911857570521e-05,"reference_active":true,"delta_db":3.646540263826188}],"active_band_count":3,"active_band_mae_db":2.436078595653765}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":595,"rms_floor":0.00020448142520679275,"log_rms_mae_db":2.959566357706519,"rms_envelope_nrmse":0.39571816169936713,"envelope_correlation":0.8973971536559084,"rms_gain_error_db":-3.1167173220299653,"reference_mean_power":0.005342387707462288,"output_mean_power":0.0026065356560199013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":595,"frequency_resolution_hz":0.3361344537815126,"power_floor":2.6914560842690746e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013693757788350162,"output_power":0.0006481898695178919,"reference_active":true,"delta_db":-3.2482040269803596},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.413543257692763e-05,"output_power":5.201491120649653e-05,"reference_active":true,"delta_db":-0.9097016683863589},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.674728419756812e-06,"output_power":4.101840631554674e-06,"reference_active":true,"delta_db":3.8903439249549727}],"active_band_count":3,"active_band_mae_db":2.6827498734405637}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00020294603022145288,"log_rms_mae_db":2.5963066387408307,"rms_envelope_nrmse":0.48165742204637696,"envelope_correlation":0.8680368375833716,"rms_gain_error_db":-3.2424445927101693,"reference_mean_power":0.0022619521433475667,"output_mean_power":0.0010721090275484183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.6059592173229041e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007396875668559024,"output_power":0.0004142897084923862,"reference_active":true,"delta_db":-2.517441738829525},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010058798371492593,"output_power":7.806676871178884e-05,"reference_active":true,"delta_db":-1.1007989910152203},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.617571441833657e-05,"output_power":6.317383254833468e-05,"reference_active":true,"delta_db":5.916737540483505}],"active_band_count":3,"active_band_mae_db":3.178326090109417}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":595,"rms_floor":0.00017314576936579135,"log_rms_mae_db":1.8445322776486757,"rms_envelope_nrmse":0.43008217381712827,"envelope_correlation":0.9115711518608367,"rms_gain_error_db":-3.24244791920765,"reference_mean_power":0.002284757287918625,"output_mean_power":0.0010829172691063959},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":595,"frequency_resolution_hz":0.3361344537815126,"power_floor":1.4904009051407098e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007408819474493079,"output_power":0.00043693888994286164,"reference_active":true,"delta_db":-2.293283117563084},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.694179006482714e-05,"output_power":3.90287637529507e-05,"reference_active":true,"delta_db":-0.8017485174072727},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4604386300560187e-06,"output_power":4.0184722232348774e-06,"reference_active":true,"delta_db":4.395776587624107}],"active_band_count":3,"active_band_mae_db":2.4969360741981546}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":1.9149047421647718,"rms_envelope_nrmse":0.31767819709877354,"envelope_correlation":0.9398841492123233,"rms_gain_error_db":-1.5998261973789236,"reference_mean_power":0.0002824355674454734,"output_mean_power":0.0001954054927278231},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.4220001316558737e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014547450354044224,"output_power":9.979882166272527e-05,"reference_active":true,"delta_db":-1.6366147020710295},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9185039831881957e-05,"output_power":2.072825670840245e-05,"reference_active":true,"delta_db":0.3360007330130829},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167522399868022e-06,"output_power":6.584194106866633e-06,"reference_active":true,"delta_db":1.9862468293219404}],"active_band_count":3,"active_band_mae_db":1.3196207548020176}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":595,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":1.2764856720385418,"rms_envelope_nrmse":0.2615267337532244,"envelope_correlation":0.9655915644563177,"rms_gain_error_db":-1.599898407875345,"reference_mean_power":0.0002852790423116174,"output_mean_power":0.00019736949372751647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":595,"frequency_resolution_hz":0.3361344537815126,"power_floor":2.250092475926771e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013969151997361956,"output_power":9.95352078208359e-05,"reference_active":true,"delta_db":-1.4719331534750877},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.773292733664618e-06,"output_power":1.0526184854696926e-05,"reference_active":true,"delta_db":0.32230085442495476},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.068748671764992e-07,"output_power":4.858531980509773e-07,"reference_active":true,"delta_db":1.9954374425621189}],"active_band_count":3,"active_band_mae_db":1.2632238168207204}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":2.5926157894971333,"rms_envelope_nrmse":0.2572825760156713,"envelope_correlation":0.9631845827702179,"rms_gain_error_db":0.5827161168829946,"reference_mean_power":0.00010172149909191666,"output_mean_power":0.00011632802734727764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.3035067339221815e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.1466666685017e-05,"output_power":7.696406037729786e-05,"reference_active":true,"delta_db":1.7476193018501127},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.059623222129099e-05,"output_power":1.8213648406978483e-05,"reference_active":true,"delta_db":-0.5339083104971374},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.965497159878585e-06,"output_power":2.663748798713926e-06,"reference_active":true,"delta_db":-0.4660424557021057}],"active_band_count":3,"active_band_mae_db":0.9158566893497851}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48160,"uncovered_tail_samples":0,"metrics":{"frame_count":595,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":2.230229960769152,"rms_envelope_nrmse":0.18338684280041653,"envelope_correlation":0.9826053632944837,"rms_gain_error_db":0.5827261420782461,"reference_mean_power":0.00010274640874642475,"output_mean_power":0.00011750037842264296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":595,"frequency_resolution_hz":0.3361344537815126,"power_floor":1.2031763277777752e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.134239657246395e-05,"output_power":8.628183710983807e-05,"reference_active":true,"delta_db":1.4815864417921734},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.841949128033184e-06,"output_power":8.900582448928724e-06,"reference_active":true,"delta_db":-0.43662688330162835},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5137680893427457e-07,"output_power":2.3551682867663558e-07,"reference_active":true,"delta_db":-0.28303264088392405}],"active_band_count":3,"active_band_mae_db":0.7337486553259086}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.6145291328430176,"predicted_mos":2.6145291328430176,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":72600,"input_sample_rate":24000,"input_seconds":3.025,"input_float32_sha256":"2fbf538c6ea77e9e1fea1f33d062fc717d79013d62cb4329efb0eecf3d8b2b10","internal_samples":48400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2976969249994},"started_at":"2026-10-13T05:08:42.349861+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001668204999077716,"native_input_samples":72480,"native_input_sha256":"509a44b8d544e5c2f8b97001c669b620becc00ff21ae0fc249ce76be179b998f","output_sha256":"24f8373e623e59f67514fe649bb7b65dc68e46a5f6342cbb0162d94405b6ab0a","output_samples":72600,"length_delta_samples":120,"analysis_s":0.8946791519992985,"synthesis_s":5.474339610000243,"total_s":6.369018761999541,"rtf":2.108946609933623,"output_peak":0.48319360613822937,"output_rms":0.06993320066568708,"finished_at":"2026-10-13T05:08:48.725931+00:00","analysis_stages":{"external_f0_s":0.5676099079992127},"evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"24f8373e623e59f67514fe649bb7b65dc68e46a5f6342cbb0162d94405b6ab0a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":48320,"output_samples":48400,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.109523746058784,"output_peak":0.480318241920234},"common_prefix":{"samples":48320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983471074380166,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.595955155359547,"spectral_convergence":0.6186200517114935,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7690405468175845,"spectral_convergence":0.34865661477270005,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.663475280228004,"spectral_convergence":0.35339138501422773,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.00105186967135601,"delta_db":0.47587866787791266,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.002208346433924501,"delta_db":-0.3949896262872283,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0012455322385417225,"delta_db":-0.5635097262497772,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.00020967892379887635,"delta_db":-0.5808415966678114,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":5.081883096826342e-05,"delta_db":-0.3467242082905034,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.00012890000817027946,"delta_db":-1.7105044196814492,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6787413741757803,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.31653887010753895,"pitch":null,"output_rms_dbfs":-23.102343288486647,"output_peak":0.480318241920234,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.009490327468378},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48320,"output_samples":48320,"intelligibility":{"estoi":{"value":0.7986426018596379,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.072562089314983,"rms_dbfs":-22.785804418379104,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.662948143418326,"mean":-0.006056492533904828,"zero_energy":false},"output":{"rms":0.06996532173545278,"rms_dbfs":-23.102343288486647,"peak":0.480318241920234,"peak_dbfs":-6.369418381113991,"crest_factor_db":16.732924907372656,"mean":0.0004845295733799501,"zero_energy":false},"rms_difference_db":-0.3165388701075429},"full_scale":{"reference":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":302,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":295,"reference_estimated_voiced_frames":101,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":40,"reference_voiced_output_unvoiced_frames":61,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.39603960396039606,"absolute_difference_cents_median":5.980981500481979,"absolute_difference_cents_p95":13.224006105253592,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00023057786100875665,"log_rms_mae_db":4.770274146410616,"rms_envelope_nrmse":0.183504423168776,"envelope_correlation":0.9711142627175001,"rms_gain_error_db":-0.31643528580097663,"reference_mean_power":0.005273848963429562,"output_mean_power":0.004903251381780547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.9211277816293244e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014051118338624634,"output_power":0.001370273362534556,"reference_active":true,"delta_db":-0.10903676097429116},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013648411662722183,"output_power":0.00012141463852686272,"reference_active":true,"delta_db":-0.5081106196807252},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.767297964659028e-05,"output_power":3.27987010875644e-05,"reference_active":true,"delta_db":-0.6017332607690649}],"active_band_count":3,"active_band_mae_db":0.40629354714136046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00020448142520679275,"log_rms_mae_db":4.181563176334851,"rms_envelope_nrmse":0.12296074991955248,"envelope_correlation":0.9887627438276267,"rms_gain_error_db":-0.31510847553538546,"reference_mean_power":0.005325056781288949,"output_mean_power":0.004952373550702307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.674469134143454e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013611842323655793,"output_power":0.0013131080149434394,"reference_active":true,"delta_db":-0.15616457503133124},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.483051409804231e-05,"output_power":5.8423138668435e-05,"reference_active":true,"delta_db":-0.45194580635891124},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.671629890449392e-06,"output_power":1.6976460789182239e-06,"reference_active":true,"delta_db":0.06707026528536773}],"active_band_count":3,"active_band_mae_db":0.22506021555853672}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00020294603022145288,"log_rms_mae_db":3.0214030444415534,"rms_envelope_nrmse":0.17756823893856036,"envelope_correlation":0.9797403754372359,"rms_gain_error_db":-0.44256305808026086,"reference_mean_power":0.002254450133685994,"output_mean_power":0.0020360307262254846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.5930281169536517e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007344814515063544,"output_power":0.000655472290755812,"reference_active":true,"delta_db":-0.494264954755371},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010026570767730039,"output_power":9.065432262638062e-05,"reference_active":true,"delta_db":-0.43763906239121125},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.614259763381322e-05,"output_power":1.816536904515684e-05,"reference_active":true,"delta_db":0.5127080356589055}],"active_band_count":3,"active_band_mae_db":0.4815373509351626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00017314576936579135,"log_rms_mae_db":2.350365081609038,"rms_envelope_nrmse":0.11378543191410942,"envelope_correlation":0.9924294334287991,"rms_gain_error_db":-0.44256902811161924,"reference_mean_power":0.002277104155750591,"output_mean_power":0.002056487116492176},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.4781762283606311e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007359452368517945,"output_power":0.000658957461833094,"reference_active":true,"delta_db":-0.47988118657043394},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.796294327084698e-05,"output_power":4.198763402614404e-05,"reference_active":true,"delta_db":-0.5778442281366609},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4655779178142778e-06,"output_power":1.6494104583303628e-06,"reference_active":true,"delta_db":0.513198311817512}],"active_band_count":3,"active_band_mae_db":0.523641242174869}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":2.3707745187297795,"rms_envelope_nrmse":0.1972235832884951,"envelope_correlation":0.9737783546675853,"rms_gain_error_db":-0.42818369388671307,"reference_mean_power":0.0002815001552852329,"output_mean_power":0.0002550705210734525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.4039983283954285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014419558923291264,"output_power":0.00012766579183819888,"reference_active":true,"delta_db":-0.5287743278754694},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9143053212884863e-05,"output_power":1.8612887879790605e-05,"reference_active":true,"delta_db":-0.12197445320381069},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.190405195331928e-06,"output_power":4.514550105860648e-06,"reference_active":true,"delta_db":0.323584579848836}],"active_band_count":3,"active_band_mae_db":0.3247777869760387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":1.297327513602234,"rms_envelope_nrmse":0.11549418520106253,"envelope_correlation":0.9916820922246993,"rms_gain_error_db":-0.42816806571204447,"reference_mean_power":0.00028432562712779965,"output_mean_power":0.00025763164062484503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.2325770901894177e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013845430163824743,"output_power":0.00012356518854367269,"reference_active":true,"delta_db":-0.49410317208206905},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.792343358100621e-06,"output_power":9.449941530086494e-06,"reference_active":true,"delta_db":-0.1545751174839331},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.075134841414049e-07,"output_power":3.2686008837515587e-07,"reference_active":true,"delta_db":0.2649772991025665}],"active_band_count":3,"active_band_mae_db":0.3045518628895229}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":2.1284072036341652,"rms_envelope_nrmse":0.27441271660584665,"envelope_correlation":0.9487578054325513,"rms_gain_error_db":-0.7045981439830344,"reference_mean_power":0.00010138440655269987,"output_mean_power":8.620081044255773e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.3079247184162875e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.16171570932163e-05,"output_power":4.2650866528786124e-05,"reference_active":true,"delta_db":-0.8286622226122582},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0705364401401468e-05,"output_power":1.4333828924811085e-05,"reference_active":true,"delta_db":-1.597206613843019},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9746890612071323e-06,"output_power":1.4621089186269138e-06,"reference_active":true,"delta_db":-3.0846185018639503}],"active_band_count":3,"active_band_mae_db":1.8368291127730758}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":1.2233262386605603,"rms_envelope_nrmse":0.1638397298980172,"envelope_correlation":0.9836955808741021,"rms_gain_error_db":-0.7046008320989433,"reference_mean_power":0.00010240265187942673,"output_mean_power":8.706650677286056e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.2076196938321018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.154744016518147e-05,"output_power":5.050519977105498e-05,"reference_active":true,"delta_db":-0.8587390151412334},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.945675894844706e-06,"output_power":6.972247027663037e-06,"reference_active":true,"delta_db":-1.5426153671280058},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53780713465196e-07,"output_power":1.371618875409743e-07,"reference_active":true,"delta_db":-2.6722516101171445}],"active_band_count":3,"active_band_mae_db":1.6912019974621277}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.7609074115753174,"predicted_mos":2.7609074115753174,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":72720,"input_sample_rate":24000,"input_seconds":3.03,"input_float32_sha256":"72189f19202dc57872d8653b39260ac141398c93199f95c627046df371ababb7","internal_samples":48480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6874471959999937},"started_at":"2026-10-13T05:10:26.291215+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0015818990013940493,"native_input_samples":72480,"native_input_sha256":"509a44b8d544e5c2f8b97001c669b620becc00ff21ae0fc249ce76be179b998f","output_sha256":"250ea8f4a54f2dc2ab3e17f6f34648b6c7b75d7a4f0e7c4897ea6e4ed840a748","output_samples":72720,"length_delta_samples":240,"analysis_s":0.15779991899944434,"synthesis_s":0.023255075999259134,"total_s":0.18105499499870348,"rtf":0.059951985098908436,"output_peak":0.6465683579444885,"output_rms":0.06870961340030589,"finished_at":"2026-10-13T05:10:26.479167+00:00","analysis_stages":{"external_f0_s":0.042531938999673},"evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"250ea8f4a54f2dc2ab3e17f6f34648b6c7b75d7a4f0e7c4897ea6e4ed840a748","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":48320,"output_samples":48480,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.266615220704896,"output_peak":0.6482629162808712},"common_prefix":{"samples":48320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9966996699669967,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.703738051678215,"spectral_convergence":0.5193868329889741,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.13610798246696,"spectral_convergence":0.5061216645777452,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.897226783432161,"spectral_convergence":0.7206358681575614,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.0008880989598601807,"delta_db":-0.2591270503787967,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.0022890787632379858,"delta_db":-0.23905430641146988,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0011179428122819514,"delta_db":-1.0328635760172329,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.0002315661730600519,"delta_db":-0.14963820131548417,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":4.7822183881622335e-05,"delta_db":-0.6106768615665552,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.0001545111778542923,"delta_db":-0.9234348363948983,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5357991386807396,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4664831433485444,"pitch":null,"output_rms_dbfs":-23.25228756172765,"output_peak":0.6482629162808712,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.579024272525778},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48320,"output_samples":48320,"intelligibility":{"estoi":{"value":0.7998106224319934,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.072562089314983,"rms_dbfs":-22.785804418379104,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.662948143418326,"mean":-0.006056492533904828,"zero_energy":false},"output":{"rms":0.06876787777460165,"rms_dbfs":-23.25228756172765,"peak":0.6482629162808712,"peak_dbfs":-3.7649764278902778,"crest_factor_db":19.487311133837373,"mean":0.0001341117814907414,"zero_energy":false},"rms_difference_db":-0.4664831433485439},"full_scale":{"reference":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":302,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":295,"reference_estimated_voiced_frames":101,"output_estimated_voiced_frames":17,"both_estimated_voiced_frames":15,"reference_voiced_output_unvoiced_frames":86,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.1485148514851485,"absolute_difference_cents_median":1.9731868056822577,"absolute_difference_cents_p95":4.349001991025226,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00023057786100875665,"log_rms_mae_db":1.6568541888002004,"rms_envelope_nrmse":0.15616192800088927,"envelope_correlation":0.977247420550707,"rms_gain_error_db":-0.46654476112520926,"reference_mean_power":0.005273848963429562,"output_mean_power":0.0047366709055484445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.9211277816293244e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014051118338624634,"output_power":0.0012641748565142483,"reference_active":true,"delta_db":-0.45903743139717545},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013648411662722183,"output_power":0.0001490379287419346,"reference_active":true,"delta_db":0.3821469312267698},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.767297964659028e-05,"output_power":6.861946150268623e-05,"reference_active":true,"delta_db":2.6041733461195427}],"active_band_count":3,"active_band_mae_db":1.1484525695811627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00020448142520679275,"log_rms_mae_db":0.7860617211783553,"rms_envelope_nrmse":0.09543154895455751,"envelope_correlation":0.9928565205058576,"rms_gain_error_db":-0.46640019637472435,"reference_mean_power":0.005325056781288949,"output_mean_power":0.004782822056765317},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.674469134143454e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013611842323655793,"output_power":0.0012417578310069803,"reference_active":true,"delta_db":-0.39880002183408286},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.483051409804231e-05,"output_power":6.532710503054578e-05,"reference_active":true,"delta_db":0.03313947345636451},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.671629890449392e-06,"output_power":3.476109888835712e-06,"reference_active":true,"delta_db":3.1795336899708646}],"active_band_count":3,"active_band_mae_db":1.2038243950871041}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00020294603022145288,"log_rms_mae_db":1.7920499042536022,"rms_envelope_nrmse":0.2036783224556218,"envelope_correlation":0.972408915824769,"rms_gain_error_db":-0.3063211616530484,"reference_mean_power":0.002254450133685994,"output_mean_power":0.002100915172058678},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.5930281169536517e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007344814515063544,"output_power":0.0007190643414882069,"reference_active":true,"delta_db":-0.09213080094748526},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010026570767730039,"output_power":8.744549323100034e-05,"reference_active":true,"delta_db":-0.5941499171216903},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.614259763381322e-05,"output_power":3.129154346650388e-05,"reference_active":true,"delta_db":2.8745356361436025}],"active_band_count":3,"active_band_mae_db":1.1869387847375927}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00017314576936579135,"log_rms_mae_db":1.1072712193625822,"rms_envelope_nrmse":0.1417815186999577,"envelope_correlation":0.9867200651016271,"rms_gain_error_db":-0.3063284434922348,"reference_mean_power":0.002277104155750591,"output_mean_power":0.0021220228277996837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.4781762283606311e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007359452368517945,"output_power":0.0007304163772501303,"reference_active":true,"delta_db":-0.032749965244491246},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.796294327084698e-05,"output_power":3.839605031989094e-05,"reference_active":true,"delta_db":-0.9661927378325673},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4655779178142778e-06,"output_power":2.0983886965579048e-06,"reference_active":true,"delta_db":1.5587702540148272}],"active_band_count":3,"active_band_mae_db":0.8525709856972953}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":1.12250937200751,"rms_envelope_nrmse":0.2145887961491321,"envelope_correlation":0.9772534783316591,"rms_gain_error_db":-1.205767403174383,"reference_mean_power":0.0002815001552852329,"output_mean_power":0.00021325631400501396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.4039983283954285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014419558923291264,"output_power":9.991502353134844e-05,"reference_active":true,"delta_db":-1.593211810536137},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9143053212884863e-05,"output_power":1.5731466776786405e-05,"reference_active":true,"delta_db":-0.8524198913980414},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.190405195331928e-06,"output_power":4.3069282664934586e-06,"reference_active":true,"delta_db":0.11911618958470396}],"active_band_count":3,"active_band_mae_db":0.8549159638396274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":0.682247669849409,"rms_envelope_nrmse":0.16758482218509388,"envelope_correlation":0.9912046355805384,"rms_gain_error_db":-1.205781457265167,"reference_mean_power":0.00028432562712779965,"output_mean_power":0.00021539611209253604},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.2325770901894177e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013845430163824743,"output_power":9.61019372730592e-05,"reference_active":true,"delta_db":-1.5857431089947849},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.792343358100621e-06,"output_power":8.250139934489093e-06,"reference_active":true,"delta_db":-0.7442531825778441},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.075134841414049e-07,"output_power":3.028319548106597e-07,"reference_active":true,"delta_db":-0.06662463908112523}],"active_band_count":3,"active_band_mae_db":0.7988736435512515}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":1.0681640134857489,"rms_envelope_nrmse":0.12535938154690096,"envelope_correlation":0.9893576091085102,"rms_gain_error_db":-0.15308034827538752,"reference_mean_power":0.00010138440655269987,"output_mean_power":9.787305139588112e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.3079247184162875e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.16171570932163e-05,"output_power":5.3868497352102475e-05,"reference_active":true,"delta_db":0.18540779856813927},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0705364401401468e-05,"output_power":1.9039147333871955e-05,"reference_active":true,"delta_db":-0.36435383562444806},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9746890612071323e-06,"output_power":2.8575703009710142e-06,"reference_active":true,"delta_db":-0.1744465286752363}],"active_band_count":3,"active_band_mae_db":0.2414027209559412}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":0.6256397679573226,"rms_envelope_nrmse":0.07211130454390362,"envelope_correlation":0.9964239641670628,"rms_gain_error_db":-0.1530799706307513,"reference_mean_power":0.00010240265187942673,"output_mean_power":9.88560393335432e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.2076196938321018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.154744016518147e-05,"output_power":6.413204082239463e-05,"reference_active":true,"delta_db":0.17865065547753534},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.945675894844706e-06,"output_power":8.729286719924466e-06,"reference_active":true,"delta_db":-0.5665554408324576},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53780713465196e-07,"output_power":2.1056179180505933e-07,"reference_active":true,"delta_db":-0.8107904636087265}],"active_band_count":3,"active_band_mae_db":0.5186655199729064}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2830-3980-0042","method":"magphase","status":"measured","mos":{"status":"measured","value":3.9063572883605957,"predicted_mos":3.9063572883605957,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":145635,"input_sample_rate":48000,"input_seconds":3.0340625,"input_float32_sha256":"d3cc7fa95e99104d247aca268abfc04cc70df7d32164479f1e5a74b250470cc9","internal_samples":48545,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.55230076800035},"started_at":"2026-10-13T05:14:13.849581+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005127219999849331,"native_input_samples":144960,"native_input_sha256":"e10a6325c97a7a5c6a38f11c69712a0623600f125302c378d347a5fe858dc4a1","output_sha256":"7e1bb879894a68d40367e0305b3576f9858d28100c64d44e20acf26e90fd1a6e","output_samples":145635,"length_delta_samples":675,"analysis_s":1.655981891999545,"synthesis_s":0.6387561909996293,"total_s":2.2947380829991744,"rtf":0.75984704735072,"output_peak":0.5692310233154315,"output_rms":0.0697790561372038,"finished_at":"2026-10-13T05:14:16.156517+00:00","evaluation_input":{"reference":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"output_sha256":"7e1bb879894a68d40367e0305b3576f9858d28100c64d44e20acf26e90fd1a6e","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":48320,"output_samples":48545,"length_delta_samples":225,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.135724039772235,"output_peak":0.5664031198981956},"common_prefix":{"samples":48320,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9953651251416211,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48320,"output_samples":48320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.453187215249725,"spectral_convergence":0.7133449204960282,"spectral_convergence_status":"ok","amplitude_floor":1.6806044167073843e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.567370881759164,"spectral_convergence":0.33184456587099237,"spectral_convergence_status":"ok","amplitude_floor":1.0917469705551035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.028553164957511,"spectral_convergence":0.2838892353150738,"spectral_convergence_status":"ok","amplitude_floor":1.0075473322757857e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009427012142024427,"output_power":0.0008400812250489027,"delta_db":-0.5005278767231465,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024186117972299723,"output_power":0.0022308734031008973,"delta_db":-0.35091241117133937,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0014180970240603914,"output_power":0.0013601543755716223,"delta_db":-0.18117742668500372,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00023968394187503082,"output_power":0.0002304272539470313,"delta_db":-0.17105094267557125,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.504236566794763e-05,"output_power":6.179584319022604e-05,"delta_db":0.5026217127539834,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00019111878985838765,"output_power":0.00015588763893041704,"delta_db":-0.884917075624541,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4318679076055975,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.33069157427763296,"pitch":null,"output_rms_dbfs":-23.11649599265674,"output_peak":0.5664031198981956,"reference_rms_dbfs":-22.785804418379104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.683037087322134},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48320,"output_samples":48320,"intelligibility":{"estoi":{"value":0.7251613721039898,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.072562089314983,"rms_dbfs":-22.785804418379104,"peak":0.554443359375,"peak_dbfs":-5.1228562749607836,"crest_factor_db":17.662948143418326,"mean":-0.006056492533904828,"zero_energy":false},"output":{"rms":0.06985141374560058,"rms_dbfs":-23.11649599265674,"peak":0.5664031198981956,"peak_dbfs":-4.937487262056472,"crest_factor_db":18.179008730600266,"mean":-0.0012580899297302964,"zero_energy":false},"rms_difference_db":-0.33069157427763685},"full_scale":{"reference":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":302,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.31661499873735e-06,"threshold_dbfs":-52.743647878667055,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":295,"reference_estimated_voiced_frames":101,"output_estimated_voiced_frames":41,"both_estimated_voiced_frames":39,"reference_voiced_output_unvoiced_frames":62,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.38613861386138615,"absolute_difference_cents_median":3.985881258216297,"absolute_difference_cents_p95":18.81339063868632,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00023057786100875665,"log_rms_mae_db":3.256501815787086,"rms_envelope_nrmse":0.19950138770240913,"envelope_correlation":0.9635507750389756,"rms_gain_error_db":-0.3305978410278962,"reference_mean_power":0.005273848963429562,"output_mean_power":0.004887287682827226},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.9211277816293244e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014051118338624634,"output_power":0.0013816671064326136,"reference_active":true,"delta_db":-0.07307473030504512},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013648411662722183,"output_power":0.00013687859333686128,"reference_active":true,"delta_db":0.012534202428466368},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.767297964659028e-05,"output_power":2.9745217251422704e-05,"reference_active":true,"delta_db":-1.026128257092291}],"active_band_count":3,"active_band_mae_db":0.37057906327526746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00020448142520679275,"log_rms_mae_db":2.0910063761187745,"rms_envelope_nrmse":0.10810315316135481,"envelope_correlation":0.9907302930770432,"rms_gain_error_db":-0.32967692446707325,"reference_mean_power":0.005325056781288949,"output_mean_power":0.0049357886001398345},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.674469134143454e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013611842323655793,"output_power":0.0013330430843460149,"reference_active":true,"delta_db":-0.09072723522357959},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.483051409804231e-05,"output_power":6.071906642585946e-05,"reference_active":true,"delta_db":-0.28454379989830375},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.671629890449392e-06,"output_power":1.3573646511699209e-06,"reference_active":true,"delta_db":-0.904435931727397}],"active_band_count":3,"active_band_mae_db":0.4265689889497601}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":0.00020294603022145288,"log_rms_mae_db":3.5469224649677797,"rms_envelope_nrmse":0.19501324502279224,"envelope_correlation":0.9745812900070991,"rms_gain_error_db":-0.1963047877836391,"reference_mean_power":0.002254450133685994,"output_mean_power":0.002154815817233117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.5930281169536517e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007344814515063544,"output_power":0.000706281390383754,"reference_active":true,"delta_db":-0.1700306935684882},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010026570767730039,"output_power":9.283065978903079e-05,"reference_active":true,"delta_db":-0.33460986229410067},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.614259763381322e-05,"output_power":1.364400569835454e-05,"reference_active":true,"delta_db":-0.730315296734182}],"active_band_count":3,"active_band_mae_db":0.41165195086559025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":0.00017314576936579135,"log_rms_mae_db":2.3835994308829314,"rms_envelope_nrmse":0.12051238048154726,"envelope_correlation":0.9900977987259996,"rms_gain_error_db":-0.19630649438744385,"reference_mean_power":0.002277104155750591,"output_mean_power":0.0021764678006533514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.4781762283606311e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007359452368517945,"output_power":0.0007041056220373488,"reference_active":true,"delta_db":-0.19207686802219165},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.796294327084698e-05,"output_power":4.178786551692389e-05,"reference_active":true,"delta_db":-0.5985563757519198},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4655779178142778e-06,"output_power":1.3208530341861853e-06,"reference_active":true,"delta_db":-0.4515441459722447}],"active_band_count":3,"active_band_mae_db":0.4140591299154521}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":8.794013668187318e-05,"log_rms_mae_db":2.728847107292712,"rms_envelope_nrmse":0.2345361887646044,"envelope_correlation":0.9621724843204581,"rms_gain_error_db":-0.10601340962761177,"reference_mean_power":0.0002815001552852329,"output_mean_power":0.0002747117894008847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":2.4039983283954285e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014419558923291264,"output_power":0.00014162918171084943,"reference_active":true,"delta_db":-0.0779923010805611},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9143053212884863e-05,"output_power":1.90735192437763e-05,"reference_active":true,"delta_db":-0.01580374715523978},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.190405195331928e-06,"output_power":4.052361888406825e-06,"reference_active":true,"delta_db":-0.14547797270871551}],"active_band_count":3,"active_band_mae_db":0.07975800698150547}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":7.034799916708663e-05,"log_rms_mae_db":1.5697997473742624,"rms_envelope_nrmse":0.1175015590152469,"envelope_correlation":0.9901025893703049,"rms_gain_error_db":-0.10602164801256238,"reference_mean_power":0.00028432562712779965,"output_mean_power":0.0002774685987471135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":2.2325770901894177e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013845430163824743,"output_power":0.00013628223069378617,"reference_active":true,"delta_db":-0.06867219821626946},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.792343358100621e-06,"output_power":9.690039837798246e-06,"reference_active":true,"delta_db":-0.04561070611699894},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.075134841414049e-07,"output_power":2.7971854660652576e-07,"reference_active":true,"delta_db":-0.4114290086583685}],"active_band_count":3,"active_band_mae_db":0.17523730433054563}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":603,"rms_floor":5.0491059904452196e-05,"log_rms_mae_db":2.7337796492781874,"rms_envelope_nrmse":0.3729208652574646,"envelope_correlation":0.9051922655880964,"rms_gain_error_db":0.017108896421164883,"reference_mean_power":0.00010138440655269987,"output_mean_power":0.00010178459502542825},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":603,"frequency_resolution_hz":0.33167495854063017,"power_floor":1.3079247184162875e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.16171570932163e-05,"output_power":4.5427999317361314e-05,"reference_active":true,"delta_db":-0.5547047073900111},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0705364401401468e-05,"output_power":2.0152935898815416e-05,"reference_active":true,"delta_db":-0.11744554697510942},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9746890612071323e-06,"output_power":3.500151687930829e-06,"reference_active":true,"delta_db":0.7064528957735521}],"active_band_count":3,"active_band_mae_db":0.45953438337955754}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48320,"uncovered_tail_samples":0,"metrics":{"frame_count":597,"rms_floor":4.0590764349045424e-05,"log_rms_mae_db":1.6120489874026636,"rms_envelope_nrmse":0.2232342858781883,"envelope_correlation":0.9638887831951454,"rms_gain_error_db":0.017088167427522744,"reference_mean_power":0.00010240265187942673,"output_mean_power":0.00010280636891094811},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":597,"frequency_resolution_hz":0.33500837520938026,"power_floor":1.2076196938321018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.154744016518147e-05,"output_power":5.509795764787415e-05,"reference_active":true,"delta_db":-0.48074493921859535},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.945675894844706e-06,"output_power":8.925900441269837e-06,"reference_active":true,"delta_db":-0.46982264088051257},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53780713465196e-07,"output_power":2.960498544028701e-07,"reference_active":true,"delta_db":0.6690623783214071}],"active_band_count":3,"active_band_mae_db":0.5398766528068384}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},{"case":"librispeech_2961-960-0010","method":"hnm","status":"measured","mos":{"status":"measured","value":2.546583414077759,"predicted_mos":2.546583414077759,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b52af4cd48fdf5be5a751c427ff42449482b74a8de4a91a5ca2706d75fc49f71","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.321782674998758},"started_at":"2026-10-13T04:38:49.204881+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.1180000001331791e-05,"native_input_samples":96000,"native_input_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","output_sha256":"8875469d745ce9fad3f1f70dbf84bfc97477b0e3d17e2ea1acb5c17a4058454d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1383486950007864,"synthesis_s":0.48563509699943097,"total_s":1.6239837920002174,"rtf":0.27066396533336956,"output_peak":0.3924473128201879,"output_rms":0.03818942093133069,"finished_at":"2026-10-13T04:38:50.833622+00:00","evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"8875469d745ce9fad3f1f70dbf84bfc97477b0e3d17e2ea1acb5c17a4058454d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.336906213482559,"spectral_convergence":0.20756144570653048,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2591159719835066,"spectral_convergence":0.2104700098498939,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8151343577875725,"spectral_convergence":0.22163003823016786,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":2.9664711429756725e-05,"delta_db":0.24611118450334607,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.00039820167758853855,"delta_db":-0.01736878136925936,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0008172176418608782,"delta_db":-0.2892055833701807,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":0.0001378644768189421,"delta_db":-0.2628891415662801,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":5.935940541993597e-05,"delta_db":-0.29289277308757194,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.6123957952306643e-05,"delta_db":-0.29846883677843655,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23448938344584577,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20395715463416192,"pitch":null,"output_rms_dbfs":-28.361138536243708,"output_peak":0.3924473128201879,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.47038551441788},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.336906213482559,"spectral_convergence":0.20756144570653048,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2591159719835066,"spectral_convergence":0.2104700098498939,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8151343577875725,"spectral_convergence":0.22163003823016786,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":2.9664711429756725e-05,"delta_db":0.24611118450334607,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.00039820167758853855,"delta_db":-0.01736878136925936,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0008172176418608782,"delta_db":-0.2892055833701807,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":0.0001378644768189421,"delta_db":-0.2628891415662801,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":5.935940541993597e-05,"delta_db":-0.29289277308757194,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.6123957952306643e-05,"delta_db":-0.29846883677843655,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23448938344584577,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20395715463416192,"pitch":null,"output_rms_dbfs":-28.361138536243708,"output_peak":0.3924473128201879,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.47038551441788},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8270297016337382,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03909677463678235,"rms_dbfs":-28.157181381609547,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.76620826693273,"mean":4.716555277506511e-06,"zero_energy":false},"output":{"rms":0.03818942093133069,"rms_dbfs":-28.361138536243708,"peak":0.3924473128201879,"peak_dbfs":-8.124372805756034,"crest_factor_db":20.236765730487676,"mean":-1.188631939849504e-05,"zero_energy":false},"rms_difference_db":-0.20395715463416053},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":3.0482798458999485e-06,"output_rms_dbfs":-55.15945165301203,"output_zero_energy":false,"output_peak":0.004412159642347132,"output_energy_fraction":3.483513018749816e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":191,"output_estimated_voiced_frames":195,"both_estimated_voiced_frames":172,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.900523560209424,"absolute_difference_cents_median":1.6530876239918606,"absolute_difference_cents_p95":9.012717330926309,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012908667094909063,"log_rms_mae_db":0.7206722010046344,"rms_envelope_nrmse":0.07453836877814467,"envelope_correlation":0.9951211912063922,"rms_gain_error_db":-0.20175579032197813,"reference_mean_power":0.001527852496322364,"output_mean_power":0.00145849803746778},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.179575522794606e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002748180655831064,"output_power":0.00025671991352280946,"reference_active":true,"delta_db":-0.29585720540019667},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.543131824921543e-05,"output_power":4.643581269458126e-05,"reference_active":true,"delta_db":-0.7690215622386453},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2340545383743233e-05,"output_power":8.679597738693285e-06,"reference_active":true,"delta_db":-1.5283475553414416}],"active_band_count":3,"active_band_mae_db":0.8644087743267613}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001146348873577052,"log_rms_mae_db":0.4018799562062855,"rms_envelope_nrmse":0.04417247593347045,"envelope_correlation":0.998518522577893,"rms_gain_error_db":-0.20031053180370934,"reference_mean_power":0.0015234711554082644,"output_mean_power":0.0014547996327714126},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.718780555754887e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002667517105928089,"output_power":0.0002471616591035688,"reference_active":true,"delta_db":-0.33126111287692417},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8193242907874454e-05,"output_power":2.3368949031616894e-05,"reference_active":true,"delta_db":-0.8150585179303744},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.654413380386998e-07,"output_power":5.455761932615503e-07,"reference_active":true,"delta_db":-1.4705650160327413}],"active_band_count":3,"active_band_mae_db":0.8722948822800133}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001269978635547725,"log_rms_mae_db":1.08381510495659,"rms_envelope_nrmse":0.08950655771241314,"envelope_correlation":0.9935244344350774,"rms_gain_error_db":-0.22521826704443612,"reference_mean_power":0.0011723709196259178,"output_mean_power":0.001113123138754417},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.280903308109198e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025355561575728737,"output_power":0.00023464663972882596,"reference_active":true,"delta_db":-0.3365889450439842},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.608107242392719e-05,"output_power":4.765937122934731e-05,"reference_active":true,"delta_db":-0.7066800154375532},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1840508111446508e-05,"output_power":7.759748975167984e-06,"reference_active":true,"delta_db":-1.8352266742620682}],"active_band_count":3,"active_band_mae_db":0.9594985449145351}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":0.5909200377667757,"rms_envelope_nrmse":0.053335858428465055,"envelope_correlation":0.9979365588816432,"rms_gain_error_db":-0.22336452045743083,"reference_mean_power":0.0011689931508753605,"output_mean_power":0.0011103899300846948},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.891072559770158e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025025709822561247,"output_power":0.0002294102089195531,"reference_active":true,"delta_db":-0.37773664126380146},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.836247458716607e-05,"output_power":2.3625393805114462e-05,"reference_active":true,"delta_db":-0.7936506338895193},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.847292373572038e-07,"output_power":5.35437647626495e-07,"reference_active":true,"delta_db":-1.6601092982252381}],"active_band_count":3,"active_band_mae_db":0.9438321911261863}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":1.1093298371328504,"rms_envelope_nrmse":0.10446798570054364,"envelope_correlation":0.9924082747559351,"rms_gain_error_db":-0.272021889779202,"reference_mean_power":0.00015673633956673046,"output_mean_power":0.00014722023648672177},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.19793910027727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3248792094830594e-05,"output_power":3.0433716556064804e-05,"reference_active":true,"delta_db":-0.3842088080868184},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.64846750156482e-06,"output_power":2.6312084065356413e-06,"reference_active":true,"delta_db":-1.4195523414346536},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1792733873632037e-06,"output_power":1.1938701042065815e-06,"reference_active":true,"delta_db":0.053425791961997995}],"active_band_count":3,"active_band_mae_db":0.6190623138278234}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":0.5962473405380619,"rms_envelope_nrmse":0.058629368156125954,"envelope_correlation":0.998069788227905,"rms_gain_error_db":-0.27431213984406183,"reference_mean_power":0.00015617827045434713,"output_mean_power":0.00014661871032459737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.897719153255603e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.156452210778441e-05,"output_power":2.8836620468857492e-05,"reference_active":true,"delta_db":-0.3925485688574737},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6348481818884698e-06,"output_power":1.1786474364536539e-06,"reference_active":true,"delta_db":-1.420935127093059},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.181225496757285e-08,"output_power":7.170777474562464e-08,"reference_active":true,"delta_db":-0.0063231859645637815}],"active_band_count":3,"active_band_mae_db":0.6066022939716988}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":1.2523366848330837,"rms_envelope_nrmse":0.152147197804262,"envelope_correlation":0.9798553439545821,"rms_gain_error_db":-0.29779467159922335,"reference_mean_power":2.7841403311343517e-05,"output_mean_power":2.599630684045686e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.105832315496701e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.079425705940831e-06,"output_power":4.715725309764178e-06,"reference_active":true,"delta_db":-0.32266112935710534},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8241793691525187e-06,"output_power":1.5957751638407882e-06,"reference_active":true,"delta_db":-0.5809583813343556},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.319091479762647e-07,"output_power":4.534048155490053e-07,"reference_active":true,"delta_db":-0.6935133115101805}],"active_band_count":3,"active_band_mae_db":0.5323776074005472}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":0.8158528910083879,"rms_envelope_nrmse":0.09141134636551991,"envelope_correlation":0.9929432269606607,"rms_gain_error_db":-0.2978924182792854,"reference_mean_power":2.786874257727335e-05,"output_mean_power":2.6021248620442444e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.80064272849021e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.98417395020918e-06,"output_power":4.502665834753453e-06,"reference_active":true,"delta_db":-0.44123474010803154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.007140223829004e-06,"output_power":8.426539074602946e-07,"reference_active":true,"delta_db":-0.7744070246589945},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932914966264323e-08,"output_power":2.8381938882593868e-08,"reference_active":true,"delta_db":-1.4167249610017358}],"active_band_count":3,"active_band_mae_db":0.877455575256254}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0010","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0297837257385254,"predicted_mos":3.0297837257385254,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7c96602168a2b62d93616de5a9e0d56f45c68f7e7f502d1de71278d611277dda","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.089400615999693},"started_at":"2026-10-13T04:41:45.376861+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6824000087799504e-05,"native_input_samples":96000,"native_input_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","output_sha256":"bea137415716d015c3d16bfa10f49670e8154f71a86cb21432e6b66e40b5504b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5501974030003112,"synthesis_s":0.09060005899937096,"total_s":1.6407974619996821,"rtf":0.2734662436666137,"output_peak":0.4023529676163389,"output_rms":0.04726903945441302,"finished_at":"2026-10-13T04:41:47.023796+00:00","evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"bea137415716d015c3d16bfa10f49670e8154f71a86cb21432e6b66e40b5504b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.051954518787667,"spectral_convergence":0.5354500390951791,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.329324495881013,"spectral_convergence":0.3343829931006208,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.079380268965506,"spectral_convergence":0.37228813512410514,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":2.913330245969149e-05,"delta_db":0.167607081318991,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.0005309704239529538,"delta_db":1.2323036726266292,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0014119979035973633,"delta_db":2.085757600925974,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":0.0001708686260940507,"delta_db":0.6692103618956179,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":7.202059125399482e-05,"delta_db":0.5467786295348722,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.937124358479995e-05,"delta_db":0.49838961603412024,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8666744937227007,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6487169250714493,"pitch":null,"output_rms_dbfs":-26.508464456538093,"output_peak":0.4023529676163389,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.820219761211395},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.051954518787667,"spectral_convergence":0.5354500390951791,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.329324495881013,"spectral_convergence":0.3343829931006208,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.079380268965506,"spectral_convergence":0.37228813512410514,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":2.913330245969149e-05,"delta_db":0.167607081318991,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.0005309704239529538,"delta_db":1.2323036726266292,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0014119979035973633,"delta_db":2.085757600925974,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":0.0001708686260940507,"delta_db":0.6692103618956179,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":7.202059125399482e-05,"delta_db":0.5467786295348722,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.937124358479995e-05,"delta_db":0.49838961603412024,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8666744937227007,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6487169250714493,"pitch":null,"output_rms_dbfs":-26.508464456538093,"output_peak":0.4023529676163389,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.820219761211395},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8144026734194899,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03909677463678235,"rms_dbfs":-28.157181381609547,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.76620826693273,"mean":4.716555277506511e-06,"zero_energy":false},"output":{"rms":0.04726903945441301,"rms_dbfs":-26.508464456538093,"peak":0.4023529676163389,"peak_dbfs":-7.90785582238624,"crest_factor_db":18.600608634151858,"mean":2.298630836978585e-06,"zero_energy":false},"rms_difference_db":1.6487169250714544},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":1.5487019604561791e-06,"output_rms_dbfs":-58.10032151887792,"output_zero_energy":false,"output_peak":0.003995469944840932,"output_energy_fraction":1.1552155957875351e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":191,"output_estimated_voiced_frames":224,"both_estimated_voiced_frames":164,"reference_voiced_output_unvoiced_frames":27,"reference_unvoiced_output_voiced_frames":60,"paired_voiced_fraction":0.8586387434554974,"absolute_difference_cents_median":5.963991579302824,"absolute_difference_cents_p95":18.829714629144505,"difference_over_600_cents_frames":7,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012908667094909063,"log_rms_mae_db":1.4408895401047517,"rms_envelope_nrmse":0.2607023005201432,"envelope_correlation":0.9821077760233222,"rms_gain_error_db":1.6540019387442981,"reference_mean_power":0.001527852496322364,"output_mean_power":0.002236050573894319},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.179575522794606e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002748180655831064,"output_power":0.00045718899233217317,"reference_active":true,"delta_db":2.210504870752836},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.543131824921543e-05,"output_power":9.532331175616233e-05,"reference_active":true,"delta_db":2.354439153862775},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2340545383743233e-05,"output_power":1.643637126867922e-05,"reference_active":true,"delta_db":1.2447158911649456}],"active_band_count":3,"active_band_mae_db":1.9365533052601858}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001146348873577052,"log_rms_mae_db":1.1264976797096808,"rms_envelope_nrmse":0.23508976827704553,"envelope_correlation":0.9914550771784524,"rms_gain_error_db":1.6573050992127363,"reference_mean_power":0.0015234711554082644,"output_mean_power":0.002231334836434232},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.718780555754887e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002667517105928089,"output_power":0.00044815497879313804,"reference_active":true,"delta_db":2.2532101249924548},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8193242907874454e-05,"output_power":4.795202065792296e-05,"reference_active":true,"delta_db":2.306618795362318},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.654413380386998e-07,"output_power":8.385933392752009e-07,"reference_active":true,"delta_db":0.3963949578469764}],"active_band_count":3,"active_band_mae_db":1.65207462606725}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001269978635547725,"log_rms_mae_db":1.8221532901922977,"rms_envelope_nrmse":0.30105804295980404,"envelope_correlation":0.9816704817640977,"rms_gain_error_db":1.9440698867666533,"reference_mean_power":0.0011723709196259178,"output_mean_power":0.0018343070110184742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.280903308109198e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025355561575728737,"output_power":0.00044153818912255426,"reference_active":true,"delta_db":2.40895038387004},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.608107242392719e-05,"output_power":9.91696897339117e-05,"reference_active":true,"delta_db":2.4756264454547954},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1840508111446508e-05,"output_power":1.6416008767188805e-05,"reference_active":true,"delta_db":1.41897235716858}],"active_band_count":3,"active_band_mae_db":2.101183062164472}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":1.3566661996999239,"rms_envelope_nrmse":0.2760504390022781,"envelope_correlation":0.9910397809124631,"rms_gain_error_db":1.948022225507482,"reference_mean_power":0.0011689931508753605,"output_mean_power":0.0018306873862734118},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.891072559770158e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025025709822561247,"output_power":0.0004376133450231163,"reference_active":true,"delta_db":2.4270415280271616},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.836247458716607e-05,"output_power":4.979077047160508e-05,"reference_active":true,"delta_db":2.4440472691627115},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.847292373572038e-07,"output_power":8.782222187192396e-07,"reference_active":true,"delta_db":0.48884586600378116}],"active_band_count":3,"active_band_mae_db":1.7866448877312182}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":1.1905778004995913,"rms_envelope_nrmse":0.14804123730264945,"envelope_correlation":0.9893589648643453,"rms_gain_error_db":0.6643077906468786,"reference_mean_power":0.00015673633956673046,"output_mean_power":0.00018264192634794975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.19793910027727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3248792094830594e-05,"output_power":4.109845834469507e-05,"reference_active":true,"delta_db":0.9204965893124718},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.64846750156482e-06,"output_power":4.38191304136316e-06,"reference_active":true,"delta_db":0.7955327261634608},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1792733873632037e-06,"output_power":1.2580097626935547e-06,"reference_active":true,"delta_db":0.280695134969261}],"active_band_count":3,"active_band_mae_db":0.6655748168150646}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":0.6919877496085961,"rms_envelope_nrmse":0.09913124236871236,"envelope_correlation":0.9975592765495943,"rms_gain_error_db":0.6664882924142752,"reference_mean_power":0.00015617827045434713,"output_mean_power":0.00018208301597237147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.897719153255603e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.156452210778441e-05,"output_power":3.913208213032785e-05,"reference_active":true,"delta_db":0.9333373792977185},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6348481818884698e-06,"output_power":1.898953665715705e-06,"reference_active":true,"delta_db":0.6503693946301474},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.181225496757285e-08,"output_power":6.351471154291162e-08,"reference_active":true,"delta_db":-0.5332423412114713}],"active_band_count":3,"active_band_mae_db":0.7056497050464458}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":1.3339566112785173,"rms_envelope_nrmse":0.18730335284020466,"envelope_correlation":0.9730590002813894,"rms_gain_error_db":0.4527368317255622,"reference_mean_power":2.7841403311343517e-05,"output_mean_power":3.090045004844359e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.105832315496701e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.079425705940831e-06,"output_power":5.9445598009517504e-06,"reference_active":true,"delta_db":0.683050878056449},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8241793691525187e-06,"output_power":1.8573790206832059e-06,"reference_active":true,"delta_db":0.07832996136030843},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.319091479762647e-07,"output_power":5.12059084336042e-07,"reference_active":true,"delta_db":-0.16517384272859442}],"active_band_count":3,"active_band_mae_db":0.30885156071511727}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":0.8008526010703384,"rms_envelope_nrmse":0.11096792226688897,"envelope_correlation":0.9914265466329216,"rms_gain_error_db":0.4546423178207176,"reference_mean_power":2.786874257727335e-05,"output_mean_power":3.094436718809759e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.80064272849021e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.98417395020918e-06,"output_power":5.77884085195038e-06,"reference_active":true,"delta_db":0.6424754323907843},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.007140223829004e-06,"output_power":9.882866170649486e-07,"reference_active":true,"delta_db":-0.08207027069631032},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932914966264323e-08,"output_power":3.240559628622097e-08,"reference_active":true,"delta_db":-0.8409453941750935}],"active_band_count":3,"active_band_mae_db":0.5218303657540627}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0010","method":"facodec","status":"measured","mos":{"status":"measured","value":3.974656343460083,"predicted_mos":3.974656343460083,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"95a5bf504216b5f71e7e8d3e0b7b14ecb038f99ae31396b524fc3e8696392c95","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.469526554999902},"started_at":"2026-10-13T04:49:59.724423+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":8.66229993334855e-05,"native_input_samples":96000,"native_input_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","output_sha256":"d1f3e75cdb81915ad0e7d8fb8a0298f7dbbc02a9d7fd9dad6ea7dc56ab02d9d9","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7701440230002845,"synthesis_s":6.169415473999834,"total_s":7.939559497000118,"rtf":1.3232599161666865,"output_peak":0.30289188027381897,"output_rms":0.036263812490879184,"finished_at":"2026-10-13T04:50:07.671712+00:00","evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"d1f3e75cdb81915ad0e7d8fb8a0298f7dbbc02a9d7fd9dad6ea7dc56ab02d9d9","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.398898497084152,"spectral_convergence":0.25514879576458327,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.850708750827609,"spectral_convergence":0.2580260272098194,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.18681021805863,"spectral_convergence":0.2894588786088658,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":3.080976148939904e-05,"delta_db":0.4105932673876819,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.0003372679514180893,"delta_db":-0.7386488876686124,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0007890818833456109,"delta_db":-0.44136219362160445,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":9.707867021362418e-05,"delta_db":-1.7861747249518545,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":4.732467189323408e-05,"delta_db":-1.276912078042617,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.350115801368774e-05,"delta_db":-1.0694752046377267,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9538610593850163,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6533494093494842,"pitch":null,"output_rms_dbfs":-28.810530790959028,"output_peak":0.30289188027381897,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8121391553234645},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.398898497084152,"spectral_convergence":0.25514879576458327,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.850708750827609,"spectral_convergence":0.2580260272098194,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.18681021805863,"spectral_convergence":0.2894588786088658,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":3.080976148939904e-05,"delta_db":0.4105932673876819,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.0003372679514180893,"delta_db":-0.7386488876686124,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0007890818833456109,"delta_db":-0.44136219362160445,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":9.707867021362418e-05,"delta_db":-1.7861747249518545,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":4.732467189323408e-05,"delta_db":-1.276912078042617,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.350115801368774e-05,"delta_db":-1.0694752046377267,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9538610593850163,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6533494093494842,"pitch":null,"output_rms_dbfs":-28.810530790959028,"output_peak":0.30289188027381897,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8121391553234645},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8201631738517076,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03909677463678235,"rms_dbfs":-28.157181381609547,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.76620826693273,"mean":4.716555277506511e-06,"zero_energy":false},"output":{"rms":0.03626381249087919,"rms_dbfs":-28.810530790959028,"peak":0.30289188027381897,"peak_dbfs":-10.374247375834837,"crest_factor_db":18.43628341512419,"mean":1.104752225150681e-06,"zero_energy":false},"rms_difference_db":-0.6533494093494809},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":1.3892720215225984e-06,"output_rms_dbfs":-58.572127104324885,"output_zero_energy":false,"output_peak":0.003302698489278555,"output_energy_fraction":1.7607152195769836e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":191,"output_estimated_voiced_frames":197,"both_estimated_voiced_frames":167,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.8743455497382199,"absolute_difference_cents_median":3.7401721835709325,"absolute_difference_cents_p95":26.62625684790159,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012908667094909063,"log_rms_mae_db":1.421548850054239,"rms_envelope_nrmse":0.14201571323614803,"envelope_correlation":0.9846874427612732,"rms_gain_error_db":-0.6527036793370812,"reference_mean_power":0.001527852496322364,"output_mean_power":0.0013146527680493294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.179575522794606e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002748180655831064,"output_power":0.0002533593370261006,"reference_active":true,"delta_db":-0.35308364401192444},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.543131824921543e-05,"output_power":5.49950317705908e-05,"reference_active":true,"delta_db":-0.034317496878410195},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2340545383743233e-05,"output_power":1.598988086506412e-05,"reference_active":true,"delta_db":1.1251087445240593}],"active_band_count":3,"active_band_mae_db":0.504169961804798}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001146348873577052,"log_rms_mae_db":1.230806966915575,"rms_envelope_nrmse":0.13168601070764482,"envelope_correlation":0.9867383908848718,"rms_gain_error_db":-0.6517854934872973,"reference_mean_power":0.0015234711554082644,"output_mean_power":0.001311159984877224},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.718780555754887e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002667517105928089,"output_power":0.0002482399599868739,"reference_active":true,"delta_db":-0.3123552052271411},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8193242907874454e-05,"output_power":2.8992625306250806e-05,"reference_active":true,"delta_db":0.12142509699692668},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.654413380386998e-07,"output_power":9.038705864478713e-07,"reference_active":true,"delta_db":0.7219434109969226}],"active_band_count":3,"active_band_mae_db":0.3852412377403301}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001269978635547725,"log_rms_mae_db":1.8912510366977657,"rms_envelope_nrmse":0.1444303385783905,"envelope_correlation":0.984236238251391,"rms_gain_error_db":-0.540027991270194,"reference_mean_power":0.0011723709196259178,"output_mean_power":0.0010352905222517552},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.280903308109198e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025355561575728737,"output_power":0.0002439474587607764,"reference_active":true,"delta_db":-0.16776935399850618},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.608107242392719e-05,"output_power":5.5404309306373916e-05,"reference_active":true,"delta_db":-0.05272764792601893},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1840508111446508e-05,"output_power":1.5756772098548514e-05,"reference_active":true,"delta_db":1.2409691389315542}],"active_band_count":3,"active_band_mae_db":0.4871553802853598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":1.586880955632925,"rms_envelope_nrmse":0.1306779372030653,"envelope_correlation":0.9867766049278061,"rms_gain_error_db":-0.5375768268620195,"reference_mean_power":0.0011689931508753605,"output_mean_power":0.0010328905022129497},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.891072559770158e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025025709822561247,"output_power":0.00024292091173488465,"reference_active":true,"delta_db":-0.1292150208395362},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.836247458716607e-05,"output_power":2.8809796705306604e-05,"reference_active":true,"delta_db":0.06796073952449007},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.847292373572038e-07,"output_power":9.332912226959734e-07,"reference_active":true,"delta_db":0.752973476399451}],"active_band_count":3,"active_band_mae_db":0.31671641225449243}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":1.6269147255780225,"rms_envelope_nrmse":0.21135200513561872,"envelope_correlation":0.987240126350819,"rms_gain_error_db":-1.6549945961794283,"reference_mean_power":0.00015673633956673046,"output_mean_power":0.00010707060098889029},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.19793910027727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3248792094830594e-05,"output_power":2.1016323719823727e-05,"reference_active":true,"delta_db":-1.992191229346418},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.64846750156482e-06,"output_power":3.099766928303572e-06,"reference_active":true,"delta_db":-0.7078144173343608},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1792733873632037e-06,"output_power":1.0600799314345181e-06,"reference_active":true,"delta_db":-0.4627588505621013}],"active_band_count":3,"active_band_mae_db":1.0542548324142933}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":1.3164377224718635,"rms_envelope_nrmse":0.19915844654632034,"envelope_correlation":0.9920403870560152,"rms_gain_error_db":-1.6665438297533428,"reference_mean_power":0.00015617827045434713,"output_mean_power":0.00010640602659125167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.897719153255603e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.156452210778441e-05,"output_power":2.0004701859298345e-05,"reference_active":true,"delta_db":-1.9806713517925576},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6348481818884698e-06,"output_power":1.355157464366362e-06,"reference_active":true,"delta_db":-0.8148766707924591},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.181225496757285e-08,"output_power":6.993977921146247e-08,"reference_active":true,"delta_db":-0.11474307189447018}],"active_band_count":3,"active_band_mae_db":0.9700970314931623}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":1.9205833204992522,"rms_envelope_nrmse":0.24133453783585207,"envelope_correlation":0.9571139026497807,"rms_gain_error_db":-1.181537545473181,"reference_mean_power":2.7841403311343517e-05,"output_mean_power":2.1209838762523727e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.105832315496701e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.079425705940831e-06,"output_power":3.913060181715747e-06,"reference_active":true,"delta_db":-1.1329808523870604},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8241793691525187e-06,"output_power":2.390298329205244e-06,"reference_active":true,"delta_db":1.1738456821543628},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.319091479762647e-07,"output_power":6.585559906019029e-07,"reference_active":true,"delta_db":0.9275524515069068}],"active_band_count":3,"active_band_mae_db":1.0781263286827767}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":1.6265151719621858,"rms_envelope_nrmse":0.21282551152023765,"envelope_correlation":0.9679842941692934,"rms_gain_error_db":-1.184098123845937,"reference_mean_power":2.786874257727335e-05,"output_mean_power":2.1218152264090048e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.80064272849021e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.98417395020918e-06,"output_power":4.004782536906807e-06,"reference_active":true,"delta_db":-0.9501425251293166},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.007140223829004e-06,"output_power":1.354559635066005e-06,"reference_active":true,"delta_db":1.287081883371286},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932914966264323e-08,"output_power":5.363521554330022e-08,"reference_active":true,"delta_db":1.3473547364653837}],"active_band_count":3,"active_band_mae_db":1.1948597149886622}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0010","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.257254123687744,"predicted_mos":3.257254123687744,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"1c26a726190e952cbaeb59520ee059e55f8b385ba8cc09b6f06ba146f6a91fe3","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.499379503002274},"started_at":"2026-10-13T07:10:03.477966+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.01237710300119943,"native_input_samples":264600,"native_input_sha256":"678c54e15d85be96b80a39e91abd650f871fe293f8438568e1d194d2f2bccce4","output_sha256":"204c0f800d32b45c1c68558bd8cad13feea31650ee04bdd2b95429d2ca3747f9","output_samples":264704,"length_delta_samples":104,"analysis_s":0.20232728000155475,"synthesis_s":40.49065710499963,"total_s":40.69298438500118,"rtf":6.782164064166864,"output_peak":0.31939932703971863,"output_rms":0.03533083276982741,"finished_at":"2026-10-13T07:10:44.221131+00:00","evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"204c0f800d32b45c1c68558bd8cad13feea31650ee04bdd2b95429d2ca3747f9","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.02976035066548,"output_peak":0.3116424079796634},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.577649685426168,"spectral_convergence":0.4594790241988971,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.627234340723206,"spectral_convergence":0.45847624788452507,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.681970253698523,"spectral_convergence":0.5081777690310575,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":1.8438675445364626e-05,"delta_db":-1.8189928954280055,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.0004184354197566166,"delta_db":0.19788476134564054,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0006705480495745992,"delta_db":-1.1482838820822203,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":8.842938998220018e-05,"delta_db":-2.1914466211208454,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":4.7261223067525e-05,"delta_db":-1.282738629460939,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":6.912347566196372e-06,"delta_db":-3.9769297343114114,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.7693794206248439,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8736312641183686,"pitch":null,"output_rms_dbfs":-29.030812645727913,"output_peak":0.3116424079796634,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.628951426615965},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6094948022794632,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03909677463678235,"rms_dbfs":-28.157181381609547,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.76620826693273,"mean":4.716555277506511e-06,"zero_energy":false},"output":{"rms":0.035355694101410345,"rms_dbfs":-29.030812645727913,"peak":0.3116424079796634,"peak_dbfs":-10.126868972596725,"crest_factor_db":18.903943673131188,"mean":-7.532607792045972e-06,"zero_energy":false},"rms_difference_db":-0.8736312641183659},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":1.591111582087011e-06,"output_rms_dbfs":-57.98299362914427,"output_zero_energy":false,"output_peak":0.0032867267005714166,"output_energy_fraction":2.1214395017922585e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":191,"output_estimated_voiced_frames":184,"both_estimated_voiced_frames":149,"reference_voiced_output_unvoiced_frames":42,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.7801047120418848,"absolute_difference_cents_median":8.761760692066138,"absolute_difference_cents_p95":42.987824728292715,"difference_over_600_cents_frames":6,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012908667094909063,"log_rms_mae_db":1.7895020424357693,"rms_envelope_nrmse":0.14279735557703688,"envelope_correlation":0.9883823378792136,"rms_gain_error_db":-0.8721683598035637,"reference_mean_power":0.001527852496322364,"output_mean_power":0.0012498692750404728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.179575522794606e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002748180655831064,"output_power":0.00023772070169767877,"reference_active":true,"delta_db":-0.6297827480652864},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.543131824921543e-05,"output_power":4.11934200156309e-05,"reference_active":true,"delta_db":-1.289273569274969},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2340545383743233e-05,"output_power":1.008167448645986e-05,"reference_active":true,"delta_db":-0.8780168254973184}],"active_band_count":3,"active_band_mae_db":0.9323577142791913}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001146348873577052,"log_rms_mae_db":1.5864030492760754,"rms_envelope_nrmse":0.11770087944998713,"envelope_correlation":0.9946765726146608,"rms_gain_error_db":-0.8633402993348792,"reference_mean_power":0.0015234711554082644,"output_mean_power":0.0012488210369540308},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.718780555754887e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002667517105928089,"output_power":0.00022844687437837804,"reference_active":true,"delta_db":-0.6732199270264441},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8193242907874454e-05,"output_power":2.067348617531574e-05,"reference_active":true,"delta_db":-1.3473131516766244},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.654413380386998e-07,"output_power":5.99031796099787e-07,"reference_active":true,"delta_db":-1.0646203782653898}],"active_band_count":3,"active_band_mae_db":1.0283844856561528}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001269978635547725,"log_rms_mae_db":2.350367399151263,"rms_envelope_nrmse":0.18054277718118053,"envelope_correlation":0.9780234108460326,"rms_gain_error_db":-0.8775223595449269,"reference_mean_power":0.0011723709196259178,"output_mean_power":0.0009578837403738961},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.280903308109198e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025355561575728737,"output_power":0.00020417321160603116,"reference_active":true,"delta_db":-0.9407447355242953},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.608107242392719e-05,"output_power":3.613379461923659e-05,"reference_active":true,"delta_db":-1.9090273819565122},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1840508111446508e-05,"output_power":9.16662633052127e-06,"reference_active":true,"delta_db":-1.1116081157825004}],"active_band_count":3,"active_band_mae_db":1.320460077754436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":2.0528218168071963,"rms_envelope_nrmse":0.15593866289482108,"envelope_correlation":0.9847095562034117,"rms_gain_error_db":-0.8665060258366214,"reference_mean_power":0.0011689931508753605,"output_mean_power":0.0009575497871116691},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.891072559770158e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025025709822561247,"output_power":0.00019791367294224563,"reference_active":true,"delta_db":-1.019106058947251},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.836247458716607e-05,"output_power":1.787954840576279e-05,"reference_active":true,"delta_db":-2.0038757439638277},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.847292373572038e-07,"output_power":5.746073961428982e-07,"reference_active":true,"delta_db":-1.3534862203092402}],"active_band_count":3,"active_band_mae_db":1.458822674406773}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":2.6241786303379713,"rms_envelope_nrmse":0.32041365461563853,"envelope_correlation":0.9434239946480221,"rms_gain_error_db":-1.9905452203083067,"reference_mean_power":0.00015673633956673046,"output_mean_power":9.910947553861432e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.19793910027727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3248792094830594e-05,"output_power":1.9213606852698986e-05,"reference_active":true,"delta_db":-2.3816697235616986},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.64846750156482e-06,"output_power":2.200702397123443e-06,"reference_active":true,"delta_db":-2.195491656542263},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1792733873632037e-06,"output_power":7.105097596193618e-07,"reference_active":true,"delta_db":-2.2004445012775093}],"active_band_count":3,"active_band_mae_db":2.2592019604604903}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":2.2976869615588575,"rms_envelope_nrmse":0.29367653365652097,"envelope_correlation":0.9579737676132363,"rms_gain_error_db":-1.9911605806098474,"reference_mean_power":0.00015617827045434713,"output_mean_power":9.874259830048402e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.897719153255603e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.156452210778441e-05,"output_power":1.8316266793151914e-05,"reference_active":true,"delta_db":-2.3636225762361933},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6348481818884698e-06,"output_power":1.007343117532668e-06,"reference_active":true,"delta_db":-2.103000051077509},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.181225496757285e-08,"output_power":4.1886016987790706e-08,"reference_active":true,"delta_db":-2.3412949963990592}],"active_band_count":3,"active_band_mae_db":2.2693058745709203}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":3.3677473932861086,"rms_envelope_nrmse":0.3537440676986589,"envelope_correlation":0.9245932193854994,"rms_gain_error_db":-2.4676508740204124,"reference_mean_power":2.7841403311343517e-05,"output_mean_power":1.577342605444441e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.105832315496701e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.079425705940831e-06,"output_power":2.3421809896231217e-06,"reference_active":true,"delta_db":-3.3619416082051874},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8241793691525187e-06,"output_power":1.1478288289691798e-06,"reference_active":true,"delta_db":-2.0119041136981286},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.319091479762647e-07,"output_power":3.400353252068225e-07,"reference_active":true,"delta_db":-1.9431342266495366}],"active_band_count":3,"active_band_mae_db":2.438993316184284}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":3.0030389133878934,"rms_envelope_nrmse":0.3244667067741902,"envelope_correlation":0.9432525193712444,"rms_gain_error_db":-2.4680960288548284,"reference_mean_power":2.786874257727335e-05,"output_mean_power":1.5787296705027846e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.80064272849021e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.98417395020918e-06,"output_power":2.3611904012737383e-06,"reference_active":true,"delta_db":-3.2446218195815746},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.007140223829004e-06,"output_power":6.40561886948052e-07,"reference_active":true,"delta_db":-1.965288466201628},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932914966264323e-08,"output_power":2.673714960753197e-08,"reference_active":true,"delta_db":-1.6759945045703604}],"active_band_count":3,"active_band_mae_db":2.2953015967845207}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0010","method":"golf","status":"measured","mos":{"status":"measured","value":2.994140386581421,"predicted_mos":2.994140386581421,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"3d6ffbdbd21aa1ef64f109ed25b79b589320c20d896bb71399f1b764bbc0ecbd","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3179548410007556},"started_at":"2026-10-13T05:02:19.225187+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002775805999590375,"native_input_samples":144000,"native_input_sha256":"f3314c33989824d3ea8b552a437e8dacfb265f731b45d6fb4cb2b00137526aa1","output_sha256":"bc8ed00f238bfaf3087ab5579edf17a3c25abe53b155c59aa8b9ce0a8d9f2fab","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6459116599999106,"synthesis_s":0.11718259999997827,"total_s":0.7630942599998889,"rtf":0.12718237666664814,"output_peak":0.3634170591831207,"output_rms":0.03349354572432691,"finished_at":"2026-10-13T05:02:20.000310+00:00","analysis_stages":{"external_f0_s":0.09686697699999058},"evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"bc8ed00f238bfaf3087ab5579edf17a3c25abe53b155c59aa8b9ce0a8d9f2fab","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.49398989079029,"output_peak":0.36195685146779},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.855624045875725,"spectral_convergence":0.47244014866976314,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.057772316998954,"spectral_convergence":0.3240422213842608,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.363976969665949,"spectral_convergence":0.36571077043355155,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8075108947131676e-05,"output_power":3.895379508165503e-05,"delta_db":1.4222832505842755,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003971031560544685,"output_power":0.00035772998015174903,"delta_db":-0.4534800027689548,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008719994402493281,"output_power":0.0005661419495273474,"delta_db":-1.8759087007451756,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014666835430508275,"output_power":0.00011032177124010115,"delta_db":-1.2367519308535697,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.296352927223779e-05,"output_power":4.212248501348658e-05,"delta_db":-1.7457507894065494,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.72788347682752e-05,"output_power":8.302285721609281e-06,"delta_db":-3.183167761769072,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6528904060212664,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.3240915343564899,"pitch":null,"output_rms_dbfs":-29.49398989079029,"output_peak":0.36195685146779,"reference_rms_dbfs":-28.169898356433798,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.425791110846876},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.7966716608326377,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03903957509497925,"rms_dbfs":-28.169898356433798,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.778925241756983,"mean":-2.939673218384808e-06,"zero_energy":false},"output":{"rms":0.033519729514659695,"rms_dbfs":-29.493989890790285,"peak":0.36195685146779,"peak_dbfs":-8.826863964272373,"crest_factor_db":20.667125926517915,"mean":0.0002934020583668418,"zero_energy":false},"rms_difference_db":-1.3240915343564872},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":2.4432890158374666e-06,"output_rms_dbfs":-56.120251574394324,"output_zero_energy":false,"output_peak":0.004373524909464329,"output_energy_fraction":3.630337562459454e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":222,"both_estimated_voiced_frames":164,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":58,"paired_voiced_fraction":0.8631578947368421,"absolute_difference_cents_median":5.404821565608497,"absolute_difference_cents_p95":22.89925802024772,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00012908667094909063,"log_rms_mae_db":2.42316338212773,"rms_envelope_nrmse":0.20629563432428358,"envelope_correlation":0.9761449589109639,"rms_gain_error_db":-1.3278929801960917,"reference_mean_power":0.0015231962775564133,"output_mean_power":0.0011219320941171613},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.182148371615088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027490782959959206,"output_power":0.00021028063912918913,"reference_active":true,"delta_db":-1.163878207336683},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.546441782537001e-05,"output_power":5.817278907960923e-05,"reference_active":true,"delta_db":0.20705427599626908},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2353351090007925e-05,"output_power":1.655642489898928e-05,"reference_active":true,"delta_db":1.2718177895414662}],"active_band_count":3,"active_band_mae_db":0.880916757624806}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001146348873577052,"log_rms_mae_db":2.20543652411257,"rms_envelope_nrmse":0.18410889788486565,"envelope_correlation":0.985182917636393,"rms_gain_error_db":-1.357410246794416,"reference_mean_power":0.0015176848514001338,"output_mean_power":0.0011103006002660496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.720322431072923e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026694178084974313,"output_power":0.0002125846186312867,"reference_active":true,"delta_db":-0.9888471511409658},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8101409219599857e-05,"output_power":3.196900794365797e-05,"reference_active":true,"delta_db":0.5600106023642435},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.671997931244289e-07,"output_power":1.3022218310471519e-06,"reference_active":true,"delta_db":2.297764947297961}],"active_band_count":3,"active_band_mae_db":1.28220756693439}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001269978635547725,"log_rms_mae_db":2.857308924067517,"rms_envelope_nrmse":0.22899176878990887,"envelope_correlation":0.9697377672441021,"rms_gain_error_db":-1.3833950204861605,"reference_mean_power":0.0011681971507162055,"output_mean_power":0.0008495259324572678},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.281334124587761e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025354774014983857,"output_power":0.00019016558446537896,"reference_active":true,"delta_db":-1.2492782152750634},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.611496842954089e-05,"output_power":5.9664665258389416e-05,"reference_active":true,"delta_db":0.26638485490067965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1846783993513377e-05,"output_power":1.6326038948284243e-05,"reference_active":true,"delta_db":1.3928035835335835}],"active_band_count":3,"active_band_mae_db":0.9694888845697754}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":2.421939991543143,"rms_envelope_nrmse":0.2035166571019441,"envelope_correlation":0.9801959608812238,"rms_gain_error_db":-1.421723762040806,"reference_mean_power":0.0011645730248692487,"output_mean_power":0.0008394490656908346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.890809922723839e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002503577629221681,"output_power":0.00019713579464555054,"reference_active":true,"delta_db":-1.0379557442065699},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.823850921598352e-05,"output_power":3.295557990677948e-05,"reference_active":true,"delta_db":0.6708719270313388},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.870512080554987e-07,"output_power":1.356618936753122e-06,"reference_active":true,"delta_db":2.3645488513597144}],"active_band_count":3,"active_band_mae_db":1.3577921741992078}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":2.1466101117230703,"rms_envelope_nrmse":0.2003319226798809,"envelope_correlation":0.982856567369668,"rms_gain_error_db":-1.3168676392706957,"reference_mean_power":0.00015634458091279266,"output_mean_power":0.00011545056662581537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.21617091079238e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3278415866860326e-05,"output_power":2.603176070682968e-05,"reference_active":true,"delta_db":-1.066591012199176},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660714428925675e-06,"output_power":2.3609195058925064e-06,"reference_active":true,"delta_db":-1.904846707134259},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1791918388088946e-06,"output_power":9.651131275974947e-07,"reference_active":true,"delta_db":-0.8700624177837126}],"active_band_count":3,"active_band_mae_db":1.2805000457057159}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":1.762603693194428,"rms_envelope_nrmse":0.1629681451527701,"envelope_correlation":0.9939464748784931,"rms_gain_error_db":-1.3073699056921142,"reference_mean_power":0.00015535499998529917,"output_mean_power":0.00011497098302772196},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.91397231167222e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.159108692502282e-05,"output_power":2.4695051131600906e-05,"reference_active":true,"delta_db":-1.069546388128587},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6377102184107528e-06,"output_power":1.0198796822070207e-06,"reference_active":true,"delta_db":-2.0568811883047466},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.186669027707379e-08,"output_power":5.953641600860331e-08,"reference_active":true,"delta_db":-0.8174495716868148}],"active_band_count":3,"active_band_mae_db":1.3146257160400494}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":3.2666847179811325,"rms_envelope_nrmse":0.28174123011960417,"envelope_correlation":0.9717733965693817,"rms_gain_error_db":-2.2853524139453336,"reference_mean_power":2.7818246262694912e-05,"output_mean_power":1.6435938442432502e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1050567165659845e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.076935019002076e-06,"output_power":3.254247419411046e-06,"reference_active":true,"delta_db":-1.9315103566390492},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.823254355450826e-06,"output_power":9.927929405176933e-07,"reference_active":true,"delta_db":-2.6398857934246465},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.273366704999127e-07,"output_power":2.4811252508918195e-07,"reference_active":true,"delta_db":-3.2743928415772494}],"active_band_count":3,"active_band_mae_db":2.6152629972136485}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":3.0983988120795027,"rms_envelope_nrmse":0.2549977283669699,"envelope_correlation":0.9869437759398579,"rms_gain_error_db":-2.2895497981792277,"reference_mean_power":2.7835318337777966e-05,"output_mean_power":1.6430138055725275e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.79427114511403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.983339816882189e-06,"output_power":3.1791920559809167e-06,"reference_active":true,"delta_db":-1.9520373840578222},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0025899045430063e-06,"output_power":5.302441327575743e-07,"reference_active":true,"delta_db":-2.7664745555150345},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.901092518548671e-08,"output_power":1.9076039218430104e-08,"reference_active":true,"delta_db":-3.106980435596153}],"active_band_count":3,"active_band_mae_db":2.6084974583896696}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0010","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.26055645942688,"predicted_mos":3.26055645942688,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"de14b616a46e4e5833651e0161c0180781e514aa845ed499859a244c10d4c08b","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9951936899997236},"started_at":"2026-10-13T05:08:48.726511+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029421299987006932,"native_input_samples":144000,"native_input_sha256":"f3314c33989824d3ea8b552a437e8dacfb265f731b45d6fb4cb2b00137526aa1","output_sha256":"3e86b2fdf147951d88887d0ab5b183ccf363891fb5ac7ee35b0500c87d887117","output_samples":144120,"length_delta_samples":120,"analysis_s":3.472540452999965,"synthesis_s":3.356973730000391,"total_s":6.829514183000356,"rtf":1.1382523638333926,"output_peak":0.37430453300476074,"output_rms":0.03754509476028277,"finished_at":"2026-10-13T05:08:55.567462+00:00","analysis_stages":{"external_f0_s":2.060333957999319},"evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"3e86b2fdf147951d88887d0ab5b183ccf363891fb5ac7ee35b0500c87d887117","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.502787797857426,"output_peak":0.3703235448988162},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.862829211384923,"spectral_convergence":0.4899425234158451,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.272535632689259,"spectral_convergence":0.27420081477606395,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.7226644105663445,"spectral_convergence":0.2899730782002863,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":3.150726318932972e-05,"delta_db":0.5078167121085543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.00036222718564753187,"delta_db":-0.428589218782353,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0007974191396446719,"delta_db":-0.39571636441327207,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":0.0001430174363539062,"delta_db":-0.10352302349457496,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":6.119049144569308e-05,"delta_db":-0.16094878236642562,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.3845581011306697e-05,"delta_db":-0.9600735556537463,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.42611127613648775,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.35307040055372135,"pitch":null,"output_rms_dbfs":-28.510251782163266,"output_peak":0.3703235448988162,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.6193430848801755},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8048749286477254,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03909677463678235,"rms_dbfs":-28.157181381609547,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.76620826693273,"mean":4.716555277506511e-06,"zero_energy":false},"output":{"rms":0.03753940725813927,"rms_dbfs":-28.510251782163266,"peak":0.3703235448988162,"peak_dbfs":-8.628373498975305,"crest_factor_db":19.881878283187962,"mean":1.8722585273445307e-06,"zero_energy":false},"rms_difference_db":-0.35307040055371886},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":1.3800371776114464e-06,"output_rms_dbfs":-58.60109213733534,"output_zero_energy":false,"output_peak":0.0028520685725841695,"output_energy_fraction":1.6321674558018734e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":191,"output_estimated_voiced_frames":201,"both_estimated_voiced_frames":157,"reference_voiced_output_unvoiced_frames":34,"reference_unvoiced_output_voiced_frames":44,"paired_voiced_fraction":0.8219895287958116,"absolute_difference_cents_median":5.010114778829327,"absolute_difference_cents_p95":17.776530981179175,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012908667094909063,"log_rms_mae_db":1.9893779208284845,"rms_envelope_nrmse":0.15915527061429677,"envelope_correlation":0.9769598575895552,"rms_gain_error_db":-0.35086437061647774,"reference_mean_power":0.001527852496322364,"output_mean_power":0.0014092725444320383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.179575522794606e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002748180655831064,"output_power":0.00026220858790649475,"reference_active":true,"delta_db":-0.20398366649499805},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.543131824921543e-05,"output_power":4.9315732686587614e-05,"reference_active":true,"delta_db":-0.507697171176469},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2340545383743233e-05,"output_power":8.949973057684938e-06,"reference_active":true,"delta_db":-1.395126255806482}],"active_band_count":3,"active_band_mae_db":0.7022690311593163}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001146348873577052,"log_rms_mae_db":1.5919530544662965,"rms_envelope_nrmse":0.10212085514953434,"envelope_correlation":0.9909605167797007,"rms_gain_error_db":-0.35200050598886284,"reference_mean_power":0.0015234711554082644,"output_mean_power":0.0014048636816945343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.718780555754887e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002667517105928089,"output_power":0.00025363181717870744,"reference_active":true,"delta_db":-0.2190347982330855},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8193242907874454e-05,"output_power":2.459562439542383e-05,"reference_active":true,"delta_db":-0.592871809336514},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.654413380386998e-07,"output_power":5.441003393412652e-07,"reference_active":true,"delta_db":-1.4823291604631077}],"active_band_count":3,"active_band_mae_db":0.7647452560109024}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001269978635547725,"log_rms_mae_db":2.3738865290255764,"rms_envelope_nrmse":0.17444680661786946,"envelope_correlation":0.9747602570266758,"rms_gain_error_db":-0.3989089527626486,"reference_mean_power":0.0011723709196259178,"output_mean_power":0.001069483631488764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.280903308109198e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025355561575728737,"output_power":0.00023855678002576122,"reference_active":true,"delta_db":-0.2648146959723858},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.608107242392719e-05,"output_power":4.929734348793852e-05,"reference_active":true,"delta_db":-0.5599279308873133},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1840508111446508e-05,"output_power":8.555875137237254e-06,"reference_active":true,"delta_db":-1.4110590171840156}],"active_band_count":3,"active_band_mae_db":0.7452672146812382}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":1.881765485859992,"rms_envelope_nrmse":0.11295935036157001,"envelope_correlation":0.9898647362115528,"rms_gain_error_db":-0.3982669217817924,"reference_mean_power":0.0011689931508753605,"output_mean_power":0.0010665599569775835},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.891072559770158e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025025709822561247,"output_power":0.0002346448638028567,"reference_active":true,"delta_db":-0.2797535226143904},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.836247458716607e-05,"output_power":2.439714903072035e-05,"reference_active":true,"delta_db":-0.6540504066206747},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.847292373572038e-07,"output_power":5.538483138005952e-07,"reference_active":true,"delta_db":-1.5132899597370562}],"active_band_count":3,"active_band_mae_db":0.8156979629907072}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":1.5528192943821777,"rms_envelope_nrmse":0.16284812182542038,"envelope_correlation":0.9803739297972722,"rms_gain_error_db":-0.041682802509880504,"reference_mean_power":0.00015673633956673046,"output_mean_power":0.0001552392084869294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.19793910027727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3248792094830594e-05,"output_power":3.291466767839321e-05,"reference_active":true,"delta_db":-0.04386397699835823},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.64846750156482e-06,"output_power":3.511393965037865e-06,"reference_active":true,"delta_db":-0.16630923775362455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1792733873632037e-06,"output_power":1.304067472134424e-06,"reference_active":true,"delta_db":0.43685564332045446}],"active_band_count":3,"active_band_mae_db":0.21567628602414576}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":0.9575612775759009,"rms_envelope_nrmse":0.089067658121114,"envelope_correlation":0.9939262144029675,"rms_gain_error_db":-0.038044011015202626,"reference_mean_power":0.00015617827045434713,"output_mean_power":0.00015481613033471657},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.897719153255603e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.156452210778441e-05,"output_power":3.128538939823394e-05,"reference_active":true,"delta_db":-0.038576535632807274},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6348481818884698e-06,"output_power":1.607531365015362e-06,"reference_active":true,"delta_db":-0.07317973334483521},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.181225496757285e-08,"output_power":7.112839140940548e-08,"reference_active":true,"delta_db":-0.0415557706287302}],"active_band_count":3,"active_band_mae_db":0.05110401320212423}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":1.5524380912632336,"rms_envelope_nrmse":0.2175894192461611,"envelope_correlation":0.9590919485173277,"rms_gain_error_db":-0.585288780254404,"reference_mean_power":2.7841403311343517e-05,"output_mean_power":2.43311280050351e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.105832315496701e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.079425705940831e-06,"output_power":4.4892517791666555e-06,"reference_active":true,"delta_db":-0.5364064907177405},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8241793691525187e-06,"output_power":1.582397825544761e-06,"reference_active":true,"delta_db":-0.6175186222904122},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.319091479762647e-07,"output_power":3.6843494607172645e-07,"reference_active":true,"delta_db":-1.594766432165793}],"active_band_count":3,"active_band_mae_db":0.916230515057982}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":0.998093895317478,"rms_envelope_nrmse":0.1326736473768787,"envelope_correlation":0.9862262832445948,"rms_gain_error_db":-0.5831085985863256,"reference_mean_power":2.786874257727335e-05,"output_mean_power":2.4367249726307866e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.80064272849021e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.98417395020918e-06,"output_power":4.36675324545897e-06,"reference_active":true,"delta_db":-0.5743453931803184},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.007140223829004e-06,"output_power":8.73463350285658e-07,"reference_active":true,"delta_db":-0.6184525434641507},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932914966264323e-08,"output_power":2.670051234176367e-08,"reference_active":true,"delta_db":-1.681949617226068}],"active_band_count":3,"active_band_mae_db":0.9582491846235124}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0010","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.029985189437866,"predicted_mos":3.029985189437866,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"1bb25484651e97f511e5623cd27a12ecb0e2cf91a54da565247dbfbaff04be20","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.957203549998667},"started_at":"2026-10-13T05:10:26.480488+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029723740008194,"native_input_samples":144000,"native_input_sha256":"f3314c33989824d3ea8b552a437e8dacfb265f731b45d6fb4cb2b00137526aa1","output_sha256":"137ff158c1adff4f4c13e845ec08710f2b4d79a0e2d13fd26d060497bf9a92f8","output_samples":144240,"length_delta_samples":240,"analysis_s":0.8646541329999309,"synthesis_s":0.14668533700023545,"total_s":1.0113394700001663,"rtf":0.16855657833336105,"output_peak":0.3908519744873047,"output_rms":0.03853546154354388,"finished_at":"2026-10-13T05:10:27.506226+00:00","analysis_stages":{"external_f0_s":0.20175239999844052},"evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"137ff158c1adff4f4c13e845ec08710f2b4d79a0e2d13fd26d060497bf9a92f8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.276187005184347,"output_peak":0.3875523258967009},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.690027404842618,"spectral_convergence":0.1653670013056037,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.548556154852091,"spectral_convergence":0.19407922481453901,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.316525412340134,"spectral_convergence":0.23112466274482588,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8030381557821282e-05,"output_power":2.7576202757309015e-05,"delta_db":-0.07094545931667526,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003997973981863815,"output_power":0.0003901861343256717,"delta_db":-0.10568131254613927,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008734907128440196,"output_power":0.0008613395706222284,"delta_db":-0.060838923167421816,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001464675075572333,"output_power":0.00013772696522648876,"delta_db":-0.26722313178841317,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350074484130352e-05,"output_power":5.762806465022475e-05,"delta_db":-0.42144784727581985,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.7271042012588545e-05,"output_power":1.5215818026538383e-05,"delta_db":-0.5502323493419691,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2460615039060731,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1119098674389523,"pitch":null,"output_rms_dbfs":-28.269091249048497,"output_peak":0.3875523258967009,"reference_rms_dbfs":-28.157181381609547,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.851702990678281},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8841585208148056,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03909677463678235,"rms_dbfs":-28.157181381609547,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.76620826693273,"mean":4.716555277506511e-06,"zero_energy":false},"output":{"rms":0.03859627903837961,"rms_dbfs":-28.269091249048497,"peak":0.3875523258967009,"peak_dbfs":-8.233393046946148,"crest_factor_db":20.03569820210235,"mean":1.334275860140241e-05,"zero_energy":false},"rms_difference_db":-0.11190986743894982},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":2.404276182964201e-06,"output_rms_dbfs":-56.19015645717566,"output_zero_energy":false,"output_peak":0.004428665064805716,"output_energy_fraction":2.6899377440586134e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":191,"output_estimated_voiced_frames":152,"both_estimated_voiced_frames":135,"reference_voiced_output_unvoiced_frames":56,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.7068062827225131,"absolute_difference_cents_median":0.9068526927661124,"absolute_difference_cents_p95":7.472649878848493,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012908667094909063,"log_rms_mae_db":0.7875287171057974,"rms_envelope_nrmse":0.06958811430000789,"envelope_correlation":0.9955967301242452,"rms_gain_error_db":-0.10912875263195487,"reference_mean_power":0.001527852496322364,"output_mean_power":0.0014899392217617262},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.179575522794606e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002748180655831064,"output_power":0.0002689878468620623,"reference_active":true,"delta_db":-0.09312619743697655},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.543131824921543e-05,"output_power":5.1486925356641486e-05,"reference_active":true,"delta_db":-0.32058249113022874},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2340545383743233e-05,"output_power":9.830669786943513e-06,"reference_active":true,"delta_db":-0.9875124517413023}],"active_band_count":3,"active_band_mae_db":0.46707371343616916}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001146348873577052,"log_rms_mae_db":0.5389477695999851,"rms_envelope_nrmse":0.04501817104415203,"envelope_correlation":0.9981937859003579,"rms_gain_error_db":-0.10680049751229634,"reference_mean_power":0.0015234711554082644,"output_mean_power":0.0014864632829367204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.718780555754887e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002667517105928089,"output_power":0.0002606208083449878,"reference_active":true,"delta_db":-0.10098125631122637},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8193242907874454e-05,"output_power":2.612554165356899e-05,"reference_active":true,"delta_db":-0.33079729842358824},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.654413380386998e-07,"output_power":5.947348372413014e-07,"reference_active":true,"delta_db":-1.0958853429350621}],"active_band_count":3,"active_band_mae_db":0.5092212992232922}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001269978635547725,"log_rms_mae_db":1.0541150960389807,"rms_envelope_nrmse":0.08339123777638467,"envelope_correlation":0.9940888287116114,"rms_gain_error_db":-0.06404921775671002,"reference_mean_power":0.0011723709196259178,"output_mean_power":0.001155207807847364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.280903308109198e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025355561575728737,"output_power":0.0002485733402656575,"reference_active":true,"delta_db":-0.08618685463004691},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.608107242392719e-05,"output_power":5.238425595212536e-05,"reference_active":true,"delta_db":-0.2961553021786527},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1840508111446508e-05,"output_power":9.082549541456256e-06,"reference_active":true,"delta_db":-1.1516256422224913}],"active_band_count":3,"active_band_mae_db":0.5113225996770637}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":0.6288950242184145,"rms_envelope_nrmse":0.053627081482808404,"envelope_correlation":0.9974582146193879,"rms_gain_error_db":-0.06114742683826874,"reference_mean_power":0.0011689931508753605,"output_mean_power":0.0011526493878250778},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.891072559770158e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025025709822561247,"output_power":0.0002444537170543844,"reference_active":true,"delta_db":-0.10189759200357362},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.836247458716607e-05,"output_power":2.6490825744304435e-05,"reference_active":true,"delta_db":-0.2964862388300569},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.847292373572038e-07,"output_power":5.821018218216227e-07,"reference_active":true,"delta_db":-1.2972087529604752}],"active_band_count":3,"active_band_mae_db":0.5651975279313686}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":0.8686317183428443,"rms_envelope_nrmse":0.07778917171312805,"envelope_correlation":0.9964261780433071,"rms_gain_error_db":-0.3264450985915791,"reference_mean_power":0.00015673633956673046,"output_mean_power":0.00014538687111040363},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.19793910027727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3248792094830594e-05,"output_power":3.287131474304372e-05,"reference_active":true,"delta_db":-0.04958797552482615},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.64846750156482e-06,"output_power":2.994125544874944e-06,"reference_active":true,"delta_db":-0.8584047563304862},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1792733873632037e-06,"output_power":9.934996208026387e-07,"reference_active":true,"delta_db":-0.7444679225120531}],"active_band_count":3,"active_band_mae_db":0.5508202181224552}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":0.585844811419098,"rms_envelope_nrmse":0.05518040191525797,"envelope_correlation":0.9986753812034346,"rms_gain_error_db":-0.32380938677833043,"reference_mean_power":0.00015617827045434713,"output_mean_power":0.00014495715955525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.897719153255603e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.156452210778441e-05,"output_power":3.1273460213991356e-05,"reference_active":true,"delta_db":-0.04023282534563634},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6348481818884698e-06,"output_power":1.3311091183376952e-06,"reference_active":true,"delta_db":-0.8926377021827763},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.181225496757285e-08,"output_power":5.792237169043383e-08,"reference_active":true,"delta_db":-0.9335222798507004}],"active_band_count":3,"active_band_mae_db":0.6221309357930377}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":0.9288909168066155,"rms_envelope_nrmse":0.1421674835368009,"envelope_correlation":0.9830464051560305,"rms_gain_error_db":-0.26896974485497327,"reference_mean_power":2.7841403311343517e-05,"output_mean_power":2.6169422995941584e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.105832315496701e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.079425705940831e-06,"output_power":4.4659012115569235e-06,"reference_active":true,"delta_db":-0.5590550052790305},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8241793691525187e-06,"output_power":1.6453100999770653e-06,"reference_active":true,"delta_db":-0.44819776001595757},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.319091479762647e-07,"output_power":4.850052159370402e-07,"reference_active":true,"delta_db":-0.4009105034562147}],"active_band_count":3,"active_band_mae_db":0.4693877562504009}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":0.683559285615163,"rms_envelope_nrmse":0.09779972130934857,"envelope_correlation":0.9920421754657957,"rms_gain_error_db":-0.26789682563055445,"reference_mean_power":2.786874257727335e-05,"output_mean_power":2.6201592708810507e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.80064272849021e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.98417395020918e-06,"output_power":4.378902719829128e-06,"reference_active":true,"delta_db":-0.562278938491876},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.007140223829004e-06,"output_power":9.269727922368971e-07,"reference_active":true,"delta_db":-0.36022954185156275},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932914966264323e-08,"output_power":3.5871407196258233e-08,"reference_active":true,"delta_db":-0.39966142700313845}],"active_band_count":3,"active_band_mae_db":0.44072330244885904}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0010","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2999958992004395,"predicted_mos":3.2999958992004395,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287640,"input_sample_rate":48000,"input_seconds":5.9925,"input_float32_sha256":"a2caa46dfe0aea7e229e35d995cbb12758b1c2be887a05509893b892d8253e88","internal_samples":95880,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.015508893000515},"started_at":"2026-10-13T05:14:16.157161+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010434658999656676,"native_input_samples":288000,"native_input_sha256":"6bb8103baaee388a0d10a7960c4f4399cb0c8f9a5cc34bb30a899d091ae7e697","output_sha256":"52ef173f3483ae74f561e8b90e344cd99494c56c08c09e7d84360e7826083f15","output_samples":287640,"length_delta_samples":-360,"analysis_s":3.4778337309999188,"synthesis_s":1.0001966500003618,"total_s":4.478030381000281,"rtf":0.7463383968333801,"output_peak":0.398964126958756,"output_rms":0.0374541560999871,"finished_at":"2026-10-13T05:14:20.657057+00:00","evaluation_input":{"reference":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"52ef173f3483ae74f561e8b90e344cd99494c56c08c09e7d84360e7826083f15","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95880,"length_delta_samples":-120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.52363015462388,"output_peak":0.39129495803772624},"common_prefix":{"samples":95880,"sample_rate":16000,"reference_coverage":0.99875,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95880,"output_samples":95880,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.765681136149929,"spectral_convergence":0.5431666703084507,"spectral_convergence_status":"ok","amplitude_floor":9.509362929412646e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.87513745694077,"spectral_convergence":0.28154342957257633,"spectral_convergence_status":"ok","amplitude_floor":8.503880220761315e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.891897515301877,"spectral_convergence":0.26125922653634714,"spectral_convergence_status":"ok","amplitude_floor":5.089992363719204e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":2.8076024000475212e-05,"output_power":2.5796527858925324e-05,"delta_db":-0.36774349927579353,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003987300543060781,"output_power":0.0003582422107593863,"delta_db":-0.46502215542392644,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008716059737711422,"output_power":0.0007933317589346404,"delta_db":-0.4086535766818286,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014653417799657167,"output_power":0.00014290086539982011,"delta_db":-0.10904073569102779,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.300444410189779e-05,"output_power":6.617586260861223e-05,"delta_db":0.21328427009413464,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.728726004345478e-05,"output_power":1.842551078347236e-05,"delta_db":0.2769337109210641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3067796580146292,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3570061355286871,"pitch":null,"output_rms_dbfs":-28.52363015462388,"output_peak":0.39129495803772624,"reference_rms_dbfs":-28.166624019095195,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.177572036130859},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95880,"output_samples":95880,"intelligibility":{"estoi":{"value":0.7951744475080671,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03905429469662061,"rms_dbfs":-28.166624019095195,"peak":0.380584716796875,"peak_dbfs":-8.390973114676815,"crest_factor_db":19.77565090441838,"mean":5.258139243064247e-06,"zero_energy":false},"output":{"rms":0.03748163198992576,"rms_dbfs":-28.52363015462388,"peak":0.39129495803772624,"peak_dbfs":-8.149914961766662,"crest_factor_db":20.373715192857222,"mean":-5.2520445064805463e-05,"zero_energy":false},"rms_difference_db":-0.3570061355286853},"full_scale":{"reference":{"sample_count":95880,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95880,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":40,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6035728371934964e-06,"threshold_dbfs":-57.94911308842623,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":1.5343423001468181e-06,"output_mean_power":1.945993755591219e-06,"output_rms_dbfs":-57.10858557652779,"output_zero_energy":false,"output_peak":0.004765098919736783,"output_energy_fraction":2.3115133983391e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":165,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":37,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.8052631578947368,"absolute_difference_cents_median":5.1205122232932405,"absolute_difference_cents_p95":14.575778695135202,"difference_over_600_cents_frames":0,"uncovered_tail_samples":40,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95880,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1197,"rms_floor":0.00012908667094909063,"log_rms_mae_db":1.6751026638725486,"rms_envelope_nrmse":0.16626188120735974,"envelope_correlation":0.9746385269938909,"rms_gain_error_db":-0.35478017275847507,"reference_mean_power":0.0015231962775564133,"output_mean_power":0.0014037114822488782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.182148371615088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027490782959959206,"output_power":0.00025672552613136647,"reference_active":true,"delta_db":-0.29718056510943114},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.546441782537001e-05,"output_power":5.038317806522038e-05,"reference_active":true,"delta_db":-0.4172890042529469},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2353351090007925e-05,"output_power":1.1740536965023142e-05,"reference_active":true,"delta_db":-0.22096824265652387}],"active_band_count":3,"active_band_mae_db":0.3118126040063007}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1191,"rms_floor":0.0001146348873577052,"log_rms_mae_db":1.1222347548284324,"rms_envelope_nrmse":0.09848585699296063,"envelope_correlation":0.9915956755358057,"rms_gain_error_db":-0.35405027576809245,"reference_mean_power":0.0015176848514001338,"output_mean_power":0.001398867472359419},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.720322431072923e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026694178084974313,"output_power":0.0002488962394273774,"reference_active":true,"delta_db":-0.30398218547836425},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8101409219599857e-05,"output_power":2.5298540562028225e-05,"reference_active":true,"delta_db":-0.4563263125055916},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.671997931244289e-07,"output_power":7.328569974484575e-07,"reference_active":true,"delta_db":-0.198892381273313}],"active_band_count":3,"active_band_mae_db":0.3197336264190896}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1197,"rms_floor":0.0001269978635547725,"log_rms_mae_db":2.211405993123827,"rms_envelope_nrmse":0.17796016668357054,"envelope_correlation":0.9739754794485778,"rms_gain_error_db":-0.43415186063479816,"reference_mean_power":0.0011687249011511486,"output_mean_power":0.0010575407507668437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.281589069016777e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002535477401567451,"output_power":0.00023282753369763528,"reference_active":true,"delta_db":-0.3702540630912316},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6114968572255974e-05,"output_power":5.108099874234756e-05,"reference_active":true,"delta_db":-0.4081934396630136},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1846783994755571e-05,"output_power":1.0578834011836844e-05,"reference_active":true,"delta_db":-0.4916266708579671}],"active_band_count":3,"active_band_mae_db":0.42335805787073744}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1191,"rms_floor":9.809451727166842e-05,"log_rms_mae_db":1.4807878123246105,"rms_envelope_nrmse":0.11039090441805455,"envelope_correlation":0.9907042292350535,"rms_gain_error_db":-0.43207629441923734,"reference_mean_power":0.001164705508685638,"output_mean_power":0.001054407533289007},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.890873483485896e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025035776291711413,"output_power":0.00022974735430309552,"reference_active":true,"delta_db":-0.3731054335013584},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8238509254379514e-05,"output_power":2.5520570694846675e-05,"reference_active":true,"delta_db":-0.4395138412893902},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.870512070898288e-07,"output_power":7.206509814264078e-07,"reference_active":true,"delta_db":-0.38278006908732676}],"active_band_count":3,"active_band_mae_db":0.3984664479593585}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1197,"rms_floor":5.9423530315000806e-05,"log_rms_mae_db":1.6937405102800678,"rms_envelope_nrmse":0.19927977715860823,"envelope_correlation":0.970859391868271,"rms_gain_error_db":-0.012200314391535998,"reference_mean_power":0.00015632923759107938,"output_mean_power":0.0001558906896788182},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.21612385138477e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.327841586702687e-05,"output_power":3.3387899692332425e-05,"reference_active":true,"delta_db":0.01426455430615223},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.660714428595829e-06,"output_power":3.666411981956306e-06,"reference_active":true,"delta_db":0.006754125273356085},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1791918390671276e-06,"output_power":1.426385356690989e-06,"reference_active":true,"delta_db":0.8265240684502656}],"active_band_count":3,"active_band_mae_db":0.282514249343258}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1191,"rms_floor":5.3506304987563625e-05,"log_rms_mae_db":1.023993132056519,"rms_envelope_nrmse":0.11709854366650212,"envelope_correlation":0.9895654729705444,"rms_gain_error_db":-0.021090306104235564,"reference_mean_power":0.00015535114483136839,"output_mean_power":0.0001545985539684436},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.913959893350914e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1591086925541034e-05,"output_power":3.1779292864257634e-05,"reference_active":true,"delta_db":0.0257966075755608},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6377102181459912e-06,"output_power":1.6042976830860441e-06,"reference_active":true,"delta_db":-0.089521024261598},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.186669028456215e-08,"output_power":8.48064760202163e-08,"reference_active":true,"delta_db":0.7190137274276448}],"active_band_count":3,"active_band_mae_db":0.2781104530882679}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1197,"rms_floor":2.766512572532951e-05,"log_rms_mae_db":1.7043014800020795,"rms_envelope_nrmse":0.2999240836011698,"envelope_correlation":0.9295593564356122,"rms_gain_error_db":0.500619072839537,"reference_mean_power":2.78179191041041e-05,"output_mean_power":3.12166681137526e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1050555509022496e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.076935019051711e-06,"output_power":5.464067471018145e-06,"reference_active":true,"delta_db":0.31914448633941783},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8232543554120107e-06,"output_power":2.1659382439008896e-06,"reference_active":true,"delta_db":0.7479881000336284},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.273366705130611e-07,"output_power":1.1193313700048388e-06,"reference_active":true,"delta_db":3.2687070249691255}],"active_band_count":3,"active_band_mae_db":1.4452798704473906}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":40,"metrics":{"frame_count":1191,"rms_floor":2.524493687269012e-05,"log_rms_mae_db":1.0747695641165174,"rms_envelope_nrmse":0.1824735462706724,"envelope_correlation":0.9735468150943671,"rms_gain_error_db":0.5017377043190847,"reference_mean_power":2.7835236136091736e-05,"output_mean_power":3.124414756984422e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.794268022397042e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.983339816891276e-06,"output_power":5.557475104905316e-06,"reference_active":true,"delta_db":0.47357022944605026},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.002589904543984e-06,"output_power":1.2429465326801192e-06,"reference_active":true,"delta_db":0.9332911995224914},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.901092518561111e-08,"output_power":8.199265581143643e-08,"reference_active":true,"delta_db":3.2258870361576593}],"active_band_count":3,"active_band_mae_db":1.5442494883754003}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"hnm","status":"measured","mos":{"status":"measured","value":2.1381170749664307,"predicted_mos":2.1381170749664307,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ab99c6609ffe6d84b044e131bc19f0b2b59e82ec8ee319ef2cd9596343da3f40","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4270125219991314},"started_at":"2026-10-13T04:38:50.834990+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.605199940968305e-05,"native_input_samples":96000,"native_input_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","output_sha256":"c5b6695d2aac0fa24eb5a03200febb0783a1456deb1c5b003143a942763f9cd1","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0648281900002985,"synthesis_s":0.5056889749994298,"total_s":1.5705171649997283,"rtf":0.261752860833288,"output_peak":0.6079654891432077,"output_rms":0.05752540340710192,"finished_at":"2026-10-13T04:38:52.410533+00:00","evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"c5b6695d2aac0fa24eb5a03200febb0783a1456deb1c5b003143a942763f9cd1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.534941768685333,"spectral_convergence":0.2506144768956535,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5116626656972425,"spectral_convergence":0.2539872443313816,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.052996283471753,"spectral_convergence":0.26179695893295124,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.9865785488846626e-05,"delta_db":0.31047840874009835,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.0004388088219136273,"delta_db":0.49322175098312093,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0018953140481136543,"delta_db":-0.30831755903856245,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0007043143276929662,"delta_db":-0.13013672504949994,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.00018432376946510428,"delta_db":-0.18302412512065702,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":3.654528447561411e-05,"delta_db":-0.11502595666648373,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25670075426640376,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15372922005968473,"pitch":null,"output_rms_dbfs":-24.802806541630005,"output_peak":0.6079654891432077,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.699866905951443},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.534941768685333,"spectral_convergence":0.2506144768956535,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5116626656972425,"spectral_convergence":0.2539872443313816,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.052996283471753,"spectral_convergence":0.26179695893295124,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.9865785488846626e-05,"delta_db":0.31047840874009835,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.0004388088219136273,"delta_db":0.49322175098312093,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0018953140481136543,"delta_db":-0.30831755903856245,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0007043143276929662,"delta_db":-0.13013672504949994,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.00018432376946510428,"delta_db":-0.18302412512065702,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":3.654528447561411e-05,"delta_db":-0.11502595666648373,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25670075426640376,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15372922005968473,"pitch":null,"output_rms_dbfs":-24.802806541630005,"output_peak":0.6079654891432077,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.699866905951443},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8094141120349648,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05855259318441331,"rms_dbfs":-24.64907732157032,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.569645971137312,"mean":-2.2611618041992187e-06,"zero_energy":false},"output":{"rms":0.05752540340710192,"rms_dbfs":-24.802806541630005,"peak":0.6079654891432077,"peak_dbfs":-4.3224214507312295,"crest_factor_db":20.48038509089878,"mean":-2.046596426085468e-05,"zero_energy":false},"rms_difference_db":-0.15372922005968448},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":2.242113389111763e-06,"output_rms_dbfs":-56.493424278604365,"output_zero_energy":false,"output_peak":0.005866685270017118,"output_energy_fraction":9.033935030438987e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":106,"both_estimated_voiced_frames":94,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.831858407079646,"absolute_difference_cents_median":2.3164168174316666,"absolute_difference_cents_p95":10.560799446878596,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019822260308892205,"log_rms_mae_db":0.9026774404065947,"rms_envelope_nrmse":0.08885186999538719,"envelope_correlation":0.9932522803078285,"rms_gain_error_db":-0.1537078122427422,"reference_mean_power":0.003431222806852645,"output_mean_power":0.0033119070429483465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.922662834898126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006393448264747715,"output_power":0.0006266220884238075,"reference_active":true,"delta_db":-0.08729455559504906},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011197882141156391,"output_power":0.00010183294819123314,"reference_active":true,"delta_db":-0.4124757487150738},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.279682095777351e-05,"output_power":3.9502754609139886e-05,"reference_active":true,"delta_db":-0.3478412892076974}],"active_band_count":3,"active_band_mae_db":0.28253719783927345}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001679501427730559,"log_rms_mae_db":0.48394816325501244,"rms_envelope_nrmse":0.042548321500316955,"envelope_correlation":0.9985984953623243,"rms_gain_error_db":-0.15399924389810243,"reference_mean_power":0.003447933525092644,"output_mean_power":0.0033278133512190935},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8195786138583623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005865504590263928,"output_power":0.0005732660395911327,"reference_active":true,"delta_db":-0.09949164164923852},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16977320654735e-05,"output_power":5.56571048715615e-05,"reference_active":true,"delta_db":-0.44748588393612976},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.232527996839891e-06,"output_power":2.1468662962317553e-06,"reference_active":true,"delta_db":-0.16991915792089027}],"active_band_count":3,"active_band_mae_db":0.23896556116875287}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001900450506992256,"log_rms_mae_db":1.2492121179239626,"rms_envelope_nrmse":0.11520649187624209,"envelope_correlation":0.9891534684842849,"rms_gain_error_db":-0.21964193041273566,"reference_mean_power":0.002194955532035717,"output_mean_power":0.0020867072895540037},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4050180992423441e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048796679990628384,"output_power":0.00046347724921177693,"reference_active":true,"delta_db":-0.2236185393519682},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011744031611151056,"output_power":0.00010989222684915887,"reference_active":true,"delta_db":-0.288502375650691},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.011051692782721e-05,"output_power":3.821857311411997e-05,"reference_active":true,"delta_db":-0.2098379033732461}],"active_band_count":3,"active_band_mae_db":0.2406529394586351}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014767968623269652,"log_rms_mae_db":0.6509140214740572,"rms_envelope_nrmse":0.0571519796781222,"envelope_correlation":0.9975958928486752,"rms_gain_error_db":-0.2197138925833509,"reference_mean_power":0.002205773735941533,"output_mean_power":0.00209695722730428},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.315971715150661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045788234917548806,"output_power":0.00043258996319078894,"reference_active":true,"delta_db":-0.24677463447248416},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0620858906342295e-05,"output_power":5.746533160959962e-05,"reference_active":true,"delta_db":-0.2321616818751958},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2498045919301085e-06,"output_power":2.2376852910874294e-06,"reference_active":true,"delta_db":-0.0234579170725472}],"active_band_count":3,"active_band_mae_db":0.16746474447340906}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001241557965588199,"log_rms_mae_db":1.168034431376902,"rms_envelope_nrmse":0.08586169640557471,"envelope_correlation":0.9951195211546411,"rms_gain_error_db":-0.13714717174393157,"reference_mean_power":0.0006561031712122295,"output_mean_power":0.0006357076236808759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1091236930101454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020492597233413894,"output_power":0.00019780731684638248,"reference_active":true,"delta_db":-0.15354652397359028},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.133021650590788e-06,"output_power":7.251316945265117e-06,"reference_active":true,"delta_db":0.07143345746827288},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2691438355846274e-06,"output_power":5.108177847066121e-06,"reference_active":true,"delta_db":0.7792522205625105}],"active_band_count":3,"active_band_mae_db":0.33474406733479123}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011294582885045651,"log_rms_mae_db":0.6595357933529931,"rms_envelope_nrmse":0.053827121285609035,"envelope_correlation":0.9981614940531931,"rms_gain_error_db":-0.13819966043183027,"reference_mean_power":0.0006591401463258689,"output_mean_power":0.0006384954371427406},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.002660056119491e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019159735065259618,"output_power":0.00018467218128682855,"reference_active":true,"delta_db":-0.15988020535767203},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.047849758491007e-06,"output_power":3.2457215068225194e-06,"reference_active":true,"delta_db":0.273176982791936},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0860941276748168e-07,"output_power":2.431318665507208e-07,"reference_active":true,"delta_db":0.6650798367780043}],"active_band_count":3,"active_band_mae_db":0.36604567497587076}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":1.28262057115011,"rms_envelope_nrmse":0.14493438678601897,"envelope_correlation":0.9858493239550956,"rms_gain_error_db":-0.1919209035436746,"reference_mean_power":8.990510589252756e-05,"output_mean_power":8.601857952913416e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067721726798636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1343959051909258e-05,"output_power":1.0559232751894813e-05,"reference_active":true,"delta_db":-0.31132287255704805},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.893195477577151e-06,"output_power":3.120999861392615e-06,"reference_active":true,"delta_db":-0.9601246083967154},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973785070466086e-07,"output_power":4.1999191759419357e-07,"reference_active":true,"delta_db":-2.784235924544787}],"active_band_count":3,"active_band_mae_db":1.3518944684995169}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":0.758139304844051,"rms_envelope_nrmse":0.07711667787857451,"envelope_correlation":0.9960912621331058,"rms_gain_error_db":-0.19192889004751829,"reference_mean_power":9.03558904947943e-05,"output_mean_power":8.644971809310567e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9012156385948518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1053455137857865e-05,"output_power":1.0098130179729098e-05,"reference_active":true,"delta_db":-0.3925708784204508},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0905012137138105e-06,"output_power":1.5913201252182748e-06,"reference_active":true,"delta_db":-1.1849286875269511},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495315305956298e-08,"output_power":3.3970916175748166e-08,"reference_active":true,"delta_db":-2.8149297844001344}],"active_band_count":3,"active_band_mae_db":1.4641431167825119}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.6095192432403564,"predicted_mos":2.6095192432403564,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6f5a42e0310ab5ce46c1fe0512e1bb3296ae87d995bf7f83d0ec256bfea999ea","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0643970880009874},"started_at":"2026-10-13T04:41:47.025066+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4273999770230148e-05,"native_input_samples":96000,"native_input_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","output_sha256":"8f963244f8eab8f4d454d732667da855f163e5b1f1a3880c151676970fdc3f95","output_samples":96000,"length_delta_samples":0,"analysis_s":1.499326793000364,"synthesis_s":0.1034076100004313,"total_s":1.6027344030007953,"rtf":0.26712240050013253,"output_peak":0.6329470232829834,"output_rms":0.07030724341276473,"finished_at":"2026-10-13T04:41:48.634317+00:00","evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"8f963244f8eab8f4d454d732667da855f163e5b1f1a3880c151676970fdc3f95","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.205496327445334,"spectral_convergence":0.6123361094464099,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.365194274828981,"spectral_convergence":0.3455109296116877,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9346438664859775,"spectral_convergence":0.4007084912371656,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.742412610812541e-05,"delta_db":0.09244513232380931,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.00046841807227485804,"delta_db":0.7768046741047686,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0032244837753288994,"delta_db":1.9994725851118944,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0009408425287809877,"delta_db":1.1273674479920768,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.00021817849067993293,"delta_db":0.549281772840132,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":4.3761483128944914e-05,"delta_db":0.6675808980098547,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8688254183970895,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5890787329600897,"pitch":null,"output_rms_dbfs":-23.05999858861023,"output_peak":0.6329470232829834,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.835111489586764},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.205496327445334,"spectral_convergence":0.6123361094464099,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.365194274828981,"spectral_convergence":0.3455109296116877,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9346438664859775,"spectral_convergence":0.4007084912371656,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.742412610812541e-05,"delta_db":0.09244513232380931,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.00046841807227485804,"delta_db":0.7768046741047686,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0032244837753288994,"delta_db":1.9994725851118944,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0009408425287809877,"delta_db":1.1273674479920768,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.00021817849067993293,"delta_db":0.549281772840132,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":4.3761483128944914e-05,"delta_db":0.6675808980098547,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8688254183970895,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5890787329600897,"pitch":null,"output_rms_dbfs":-23.05999858861023,"output_peak":0.6329470232829834,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.835111489586764},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8398020107404753,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05855259318441331,"rms_dbfs":-24.64907732157032,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.569645971137312,"mean":-2.2611618041992187e-06,"zero_energy":false},"output":{"rms":0.07030724341276473,"rms_dbfs":-23.05999858861023,"peak":0.6329470232829834,"peak_dbfs":-3.9726527651723065,"crest_factor_db":19.08734582343793,"mean":4.921668196411263e-07,"zero_energy":false},"rms_difference_db":1.5890787329600897},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":2.80908910577673e-06,"output_rms_dbfs":-55.51434484531647,"output_zero_energy":false,"output_peak":0.008342106857145173,"output_energy_fraction":7.577119051690019e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":156,"both_estimated_voiced_frames":110,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":46,"paired_voiced_fraction":0.9734513274336283,"absolute_difference_cents_median":6.95331695304942,"absolute_difference_cents_p95":1907.6068840386097,"difference_over_600_cents_frames":15,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019822260308892205,"log_rms_mae_db":1.6102848099571099,"rms_envelope_nrmse":0.26315990475904705,"envelope_correlation":0.9792792432206411,"rms_gain_error_db":1.5891295494142998,"reference_mean_power":0.003431222806852645,"output_mean_power":0.004947227418754708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.922662834898126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006393448264747715,"output_power":0.0010363249652087967,"reference_active":true,"delta_db":2.0976080524120504},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011197882141156391,"output_power":0.00021877785851038658,"reference_active":true,"delta_db":2.90867474770977},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.279682095777351e-05,"output_power":5.315343041144362e-05,"reference_active":true,"delta_db":0.9411978836352228}],"active_band_count":3,"active_band_mae_db":1.9824935612523475}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001679501427730559,"log_rms_mae_db":1.216240911798314,"rms_envelope_nrmse":0.23030665493141414,"envelope_correlation":0.9904525283758917,"rms_gain_error_db":1.58924676744318,"reference_mean_power":0.003447933525092644,"output_mean_power":0.004971455552465644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8195786138583623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005865504590263928,"output_power":0.0009618121130492668,"reference_active":true,"delta_db":2.1478486327226602},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16977320654735e-05,"output_power":0.00011754043917090693,"reference_active":true,"delta_db":2.7991810886963195},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.232527996839891e-06,"output_power":2.664175691599071e-06,"reference_active":true,"delta_db":0.7676594757502092}],"active_band_count":3,"active_band_mae_db":1.9048963990563965}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001900450506992256,"log_rms_mae_db":1.9216525717893826,"rms_envelope_nrmse":0.32035624015580977,"envelope_correlation":0.9747571534191363,"rms_gain_error_db":1.9571339218158037,"reference_mean_power":0.002194955532035717,"output_mean_power":0.0034446025472322094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4050180992423441e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048796679990628384,"output_power":0.0008879217096700829,"reference_active":true,"delta_db":2.5998439996031},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011744031611151056,"output_power":0.00022568162593441352,"reference_active":true,"delta_db":2.836789905001585},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.011051692782721e-05,"output_power":5.248777098592713e-05,"reference_active":true,"delta_db":1.1679987081822458}],"active_band_count":3,"active_band_mae_db":2.2015442042623103}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014767968623269652,"log_rms_mae_db":1.4709118665966991,"rms_envelope_nrmse":0.2855609935914058,"envelope_correlation":0.98791309272165,"rms_gain_error_db":1.9571800577155007,"reference_mean_power":0.002205773735941533,"output_mean_power":0.0034616166177722204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.315971715150661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045788234917548806,"output_power":0.0008395706549116404,"reference_active":true,"delta_db":2.633033481900023},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0620858906342295e-05,"output_power":0.0001146501346732839,"reference_active":true,"delta_db":2.7675248383395497},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2498045919301085e-06,"output_power":2.6836200473072726e-06,"reference_active":true,"delta_db":0.765762286506691}],"active_band_count":3,"active_band_mae_db":2.0554402022487546}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001241557965588199,"log_rms_mae_db":1.2421385868798118,"rms_envelope_nrmse":0.17629432080516963,"envelope_correlation":0.9895256321808511,"rms_gain_error_db":0.9562797187819666,"reference_mean_power":0.0006561031712122295,"output_mean_power":0.0008177115060352761},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1091236930101454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020492597233413894,"output_power":0.00026798046821696186,"reference_active":true,"delta_db":1.1650613717888878},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.133021650590788e-06,"output_power":1.3013895900945602e-05,"reference_active":true,"delta_db":2.6113378623980337},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2691438355846274e-06,"output_power":4.415630317758461e-06,"reference_active":true,"delta_db":0.1465191929280044}],"active_band_count":3,"active_band_mae_db":1.3076394757049752}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011294582885045651,"log_rms_mae_db":0.7530342310263269,"rms_envelope_nrmse":0.13544690034732984,"envelope_correlation":0.9971114934226345,"rms_gain_error_db":0.9565158723281637,"reference_mean_power":0.0006591401463258689,"output_mean_power":0.0008215412061890372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.002660056119491e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019159735065259618,"output_power":0.0002521413217816527,"reference_active":true,"delta_db":1.1925452567564605},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.047849758491007e-06,"output_power":5.546249682802046e-06,"reference_active":true,"delta_db":2.6000586191810413},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0860941276748168e-07,"output_power":2.585797939840043e-07,"reference_active":true,"delta_db":0.9326068456036664}],"active_band_count":3,"active_band_mae_db":1.5750702405137227}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":1.2679993197081618,"rms_envelope_nrmse":0.16126443172248656,"envelope_correlation":0.9875328682868305,"rms_gain_error_db":0.6420799023967684,"reference_mean_power":8.990510589252756e-05,"output_mean_power":0.00010422990624886409},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067721726798636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1343959051909258e-05,"output_power":1.3601822558312126e-05,"reference_active":true,"delta_db":0.7883245481019568},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.893195477577151e-06,"output_power":4.25610398803942e-06,"reference_active":true,"delta_db":0.3870602017855196},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973785070466086e-07,"output_power":7.093293780102863e-07,"reference_active":true,"delta_db":-0.5081657806432032}],"active_band_count":3,"active_band_mae_db":0.5611835101768932}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":0.7229291482389861,"rms_envelope_nrmse":0.11036848545056346,"envelope_correlation":0.9960004759996535,"rms_gain_error_db":0.6420904905461833,"reference_mean_power":9.03558904947943e-05,"output_mean_power":0.00010475277085220977},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9012156385948518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1053455137857865e-05,"output_power":1.3186522258045557e-05,"reference_active":true,"delta_db":0.766322192420048},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0905012137138105e-06,"output_power":2.247601104279495e-06,"reference_active":true,"delta_db":0.3146881276678164},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495315305956298e-08,"output_power":4.8288980453850705e-08,"reference_active":true,"delta_db":-1.2875220208651708}],"active_band_count":3,"active_band_mae_db":0.7895107803176783}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"facodec","status":"measured","mos":{"status":"measured","value":3.8894670009613037,"predicted_mos":3.8894670009613037,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1520ca7e0d62f9c717e0738d4dcd32618654b43a1f1a5852a17137139b3f6ffb","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6486715049995837},"started_at":"2026-10-13T04:50:07.672268+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5162999918393325e-05,"native_input_samples":96000,"native_input_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","output_sha256":"b0350b3efa7b95fd6cc1bb64f81bfa566405d597633d123f8cddbcf20fab9012","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7060168179996253,"synthesis_s":6.70652535499994,"total_s":8.412542172999565,"rtf":1.402090362166594,"output_peak":0.5268740057945251,"output_rms":0.05239103240925583,"finished_at":"2026-10-13T04:50:16.092160+00:00","evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"b0350b3efa7b95fd6cc1bb64f81bfa566405d597633d123f8cddbcf20fab9012","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.391832881521843,"spectral_convergence":0.27055719984694926,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7790482538578996,"spectral_convergence":0.28175838951546833,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.000848136726546,"spectral_convergence":0.3058037171662693,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":5.09302804223537e-05,"delta_db":0.40221243821210517,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.00032143480407298636,"delta_db":-0.8586027586783482,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0016887742320090462,"delta_db":-0.8094134381157123,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0005122960191593685,"delta_db":-1.5125921456633704,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.0001456651678212686,"delta_db":-1.205280422417123,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":2.571977342267082e-05,"delta_db":-1.640668041637209,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0714615407873114,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9657835450404845,"pitch":null,"output_rms_dbfs":-25.614860866610805,"output_peak":0.5268740057945251,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.723909757368763},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.391832881521843,"spectral_convergence":0.27055719984694926,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7790482538578996,"spectral_convergence":0.28175838951546833,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.000848136726546,"spectral_convergence":0.3058037171662693,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":5.09302804223537e-05,"delta_db":0.40221243821210517,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.00032143480407298636,"delta_db":-0.8586027586783482,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0016887742320090462,"delta_db":-0.8094134381157123,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0005122960191593685,"delta_db":-1.5125921456633704,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.0001456651678212686,"delta_db":-1.205280422417123,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":2.571977342267082e-05,"delta_db":-1.640668041637209,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0714615407873114,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9657835450404845,"pitch":null,"output_rms_dbfs":-25.614860866610805,"output_peak":0.5268740057945251,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.723909757368763},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8435620242709583,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05855259318441331,"rms_dbfs":-24.64907732157032,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.569645971137312,"mean":-2.2611618041992187e-06,"zero_energy":false},"output":{"rms":0.05239103240925583,"rms_dbfs":-25.614860866610805,"peak":0.5268740057945251,"peak_dbfs":-5.565864550786742,"crest_factor_db":20.048996315824063,"mean":-8.808278283709115e-06,"zero_energy":false},"rms_difference_db":-0.9657835450404839},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":1.923964126166117e-06,"output_rms_dbfs":-57.15803029987057,"output_zero_energy":false,"output_peak":0.0054254732094705105,"output_energy_fraction":9.345914277654872e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":151,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":48,"paired_voiced_fraction":0.911504424778761,"absolute_difference_cents_median":3.5593849210173816,"absolute_difference_cents_p95":28.553320891433213,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019822260308892205,"log_rms_mae_db":1.359981260134239,"rms_envelope_nrmse":0.1551043006102988,"envelope_correlation":0.9874945076422913,"rms_gain_error_db":-0.9657529702964269,"reference_mean_power":0.003431222806852645,"output_mean_power":0.0027470946486807457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.922662834898126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006393448264747715,"output_power":0.0005105871894183951,"reference_active":true,"delta_db":-0.9766524105243675},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011197882141156391,"output_power":0.00010972494348601206,"reference_active":true,"delta_db":-0.08830526356535431},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.279682095777351e-05,"output_power":4.216785597915424e-05,"reference_active":true,"delta_db":-0.06429989947243696}],"active_band_count":3,"active_band_mae_db":0.3764191911873862}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001679501427730559,"log_rms_mae_db":1.1525885703585033,"rms_envelope_nrmse":0.1454625207177347,"envelope_correlation":0.9899355205716693,"rms_gain_error_db":-0.9655852828679141,"reference_mean_power":0.003447933525092644,"output_mean_power":0.0027605801185290177},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8195786138583623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005865504590263928,"output_power":0.0004717671071514996,"reference_active":true,"delta_db":-0.945777215520125},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16977320654735e-05,"output_power":6.059669715089067e-05,"reference_active":true,"delta_db":-0.07820246779759779},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.232527996839891e-06,"output_power":2.1109253980778872e-06,"reference_active":true,"delta_db":-0.24324028592433328}],"active_band_count":3,"active_band_mae_db":0.4224066564140187}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001900450506992256,"log_rms_mae_db":1.7192646612765565,"rms_envelope_nrmse":0.1617756215983462,"envelope_correlation":0.9826882776400544,"rms_gain_error_db":-0.7717934979058633,"reference_mean_power":0.002194955532035717,"output_mean_power":0.0018375805120994227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4050180992423441e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048796679990628384,"output_power":0.0004080899147790485,"reference_active":true,"delta_db":-0.7763441256488856},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011744031611151056,"output_power":0.0001148727543888542,"reference_active":true,"delta_db":-0.09600176962780693},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.011051692782721e-05,"output_power":4.4382138394050664e-05,"reference_active":true,"delta_db":0.43949964294844834}],"active_band_count":3,"active_band_mae_db":0.43728184607504694}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014767968623269652,"log_rms_mae_db":1.3561545308027894,"rms_envelope_nrmse":0.14624533047065633,"envelope_correlation":0.9863752822132319,"rms_gain_error_db":-0.7719202083610276,"reference_mean_power":0.002205773735941533,"output_mean_power":0.0018465834566979678},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.315971715150661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045788234917548806,"output_power":0.00038509702178021147,"reference_active":true,"delta_db":-0.751837423721454},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0620858906342295e-05,"output_power":6.166333413688204e-05,"reference_active":true,"delta_db":0.07404917956413738},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2498045919301085e-06,"output_power":2.2866708424031483e-06,"reference_active":true,"delta_db":0.07058855209103629}],"active_band_count":3,"active_band_mae_db":0.29882505179220925}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001241557965588199,"log_rms_mae_db":1.7190207741341965,"rms_envelope_nrmse":0.18412424141695394,"envelope_correlation":0.9893910254195307,"rms_gain_error_db":-1.3388569941039754,"reference_mean_power":0.0006561031712122295,"output_mean_power":0.0004820437290665498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1091236930101454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020492597233413894,"output_power":0.0001496292269766077,"reference_active":true,"delta_db":-1.3658057214244184},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.133021650590788e-06,"output_power":8.027363099719006e-06,"reference_active":true,"delta_db":0.5129936553508188},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2691438355846274e-06,"output_power":3.9543745495428995e-06,"reference_active":true,"delta_db":-0.33262984927301664}],"active_band_count":3,"active_band_mae_db":0.7371430753494179}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011294582885045651,"log_rms_mae_db":1.3312659624755843,"rms_envelope_nrmse":0.17244357028325302,"envelope_correlation":0.9925233610221912,"rms_gain_error_db":-1.3377394764697061,"reference_mean_power":0.0006591401463258689,"output_mean_power":0.0004843996452587773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.002660056119491e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019159735065259618,"output_power":0.00013988087789461713,"reference_active":true,"delta_db":-1.366311502128319},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.047849758491007e-06,"output_power":3.6657751121957587e-06,"reference_active":true,"delta_db":0.8017226339677813},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0860941276748168e-07,"output_power":1.9511399911566678e-07,"reference_active":true,"delta_db":-0.29045470103562177}],"active_band_count":3,"active_band_mae_db":0.8194962790439074}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":1.7847894204696186,"rms_envelope_nrmse":0.17585717350518545,"envelope_correlation":0.9883396795761676,"rms_gain_error_db":-1.1650533589373526,"reference_mean_power":8.990510589252756e-05,"output_mean_power":6.875100024277269e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067721726798636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1343959051909258e-05,"output_power":9.127905552932005e-06,"reference_active":true,"delta_db":-0.9439351241815558},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.893195477577151e-06,"output_power":2.459107471320697e-06,"reference_active":true,"delta_db":-1.9952870092501571},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973785070466086e-07,"output_power":5.306712516235787e-07,"reference_active":true,"delta_db":-1.7683896431809765}],"active_band_count":3,"active_band_mae_db":1.569203925537563}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":1.5162841057921703,"rms_envelope_nrmse":0.1568710267886192,"envelope_correlation":0.9929785180280252,"rms_gain_error_db":-1.1650264464361313,"reference_mean_power":9.03558904947943e-05,"output_mean_power":6.909614624079506e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9012156385948518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1053455137857865e-05,"output_power":8.686054383482294e-06,"reference_active":true,"delta_db":-1.0467550877405518},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0905012137138105e-06,"output_power":1.316245512779981e-06,"reference_active":true,"delta_db":-2.009135203911093},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495315305956298e-08,"output_power":4.019799848580679e-08,"reference_active":true,"delta_db":-2.083958086390887}],"active_band_count":3,"active_band_mae_db":1.7132827926808438}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.47721529006958,"predicted_mos":3.47721529006958,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"722ff5fd763712b6cbf5afeeabc36cbb97377e5877b0ce2f908613d5b8418cf0","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.869958586001303},"started_at":"2026-10-13T07:10:44.226944+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009999342999435612,"native_input_samples":264600,"native_input_sha256":"c8276f9125b7e208ffbee841bbbf05ab6d65211e0715d0db778542dd038eb466","output_sha256":"9565d1ee647249cd8bf544d1afeedd644b32b94c62446acae0df50fdb21b56f7","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1687173279988201,"synthesis_s":5.8197762119998515,"total_s":5.988493539998672,"rtf":0.9980822566664452,"output_peak":0.5799359679222107,"output_rms":0.05364450664848333,"finished_at":"2026-10-13T07:10:50.256841+00:00","evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"9565d1ee647249cd8bf544d1afeedd644b32b94c62446acae0df50fdb21b56f7","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.4017402473899,"output_peak":0.5748564558674594},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.258564809306035,"spectral_convergence":0.5013608744498939,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.221752099638877,"spectral_convergence":0.4347416879809582,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.131782246217707,"spectral_convergence":0.49224688824679264,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":3.2972188228526836e-05,"delta_db":-1.4860705315069587,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.00047819029723846515,"delta_db":0.8664758473102909,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0016193718784951782,"delta_db":-0.9916634433956263,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0005942428479874088,"delta_db":-0.8681623225447305,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.00014209870528570807,"delta_db":-1.3129363484534309,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":1.7105221517304354e-05,"delta_db":-3.412052397426462,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4895601484395833,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7509985207382079,"pitch":null,"output_rms_dbfs":-25.400075842308528,"output_peak":0.5748564558674594,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.204033051720874},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6305978102144627,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05855259318441331,"rms_dbfs":-24.64907732157032,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.569645971137312,"mean":-2.2611618041992187e-06,"zero_energy":false},"output":{"rms":0.05370271072071309,"rms_dbfs":-25.400075842308528,"peak":0.5748564558674594,"peak_dbfs":-4.808811739506885,"crest_factor_db":20.591264102801645,"mean":-2.7753452986566968e-05,"zero_energy":false},"rms_difference_db":-0.7509985207382073},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":1.2267947243233122e-06,"output_rms_dbfs":-59.112281003144304,"output_zero_energy":false,"output_peak":0.011065636397237321,"output_energy_fraction":5.6717648985917096e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":135,"both_estimated_voiced_frames":99,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.8761061946902655,"absolute_difference_cents_median":5.5073632769659895,"absolute_difference_cents_p95":50.96463000754434,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019822260308892205,"log_rms_mae_db":1.9471776999675998,"rms_envelope_nrmse":0.13585483875183685,"envelope_correlation":0.9891678436092555,"rms_gain_error_db":-0.7510047332546685,"reference_mean_power":0.003431222806852645,"output_mean_power":0.002886346370779204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.922662834898126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006393448264747715,"output_power":0.0005190008565884633,"reference_active":true,"delta_db":-0.9056708061886491},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011197882141156391,"output_power":0.00010527645243441608,"reference_active":true,"delta_db":-0.2680465036232299},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.279682095777351e-05,"output_power":3.365988399086392e-05,"reference_active":true,"delta_db":-1.042988950175858}],"active_band_count":3,"active_band_mae_db":0.7389020866625792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001679501427730559,"log_rms_mae_db":1.69696411793899,"rms_envelope_nrmse":0.10288961993877992,"envelope_correlation":0.9964205139789802,"rms_gain_error_db":-0.7508354935078704,"reference_mean_power":0.003447933525092644,"output_mean_power":0.002900516463174323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8195786138583623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005865504590263928,"output_power":0.0004714257134872982,"reference_active":true,"delta_db":-0.9489211196667622},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16977320654735e-05,"output_power":5.974233507556003e-05,"reference_active":true,"delta_db":-0.139870068461806},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.232527996839891e-06,"output_power":2.263588587358233e-06,"reference_active":true,"delta_db":0.06000581797220292}],"active_band_count":3,"active_band_mae_db":0.38293233536692367}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001900450506992256,"log_rms_mae_db":2.141045713409139,"rms_envelope_nrmse":0.16454665085928166,"envelope_correlation":0.982234341571037,"rms_gain_error_db":-0.7934620134771747,"reference_mean_power":0.002194955532035717,"output_mean_power":0.0018284349953328557},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4050180992423441e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048796679990628384,"output_power":0.00039358224860704214,"reference_active":true,"delta_db":-0.9335477205015664},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011744031611151056,"output_power":0.000104734812478683,"reference_active":true,"delta_db":-0.4972615208708913},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.011051692782721e-05,"output_power":3.188623824937519e-05,"reference_active":true,"delta_db":-0.9965497226683615}],"active_band_count":3,"active_band_mae_db":0.8091196546802731}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014767968623269652,"log_rms_mae_db":1.7249383539077539,"rms_envelope_nrmse":0.1284774501336479,"envelope_correlation":0.9913950800262186,"rms_gain_error_db":-0.7934833070853371,"reference_mean_power":0.002205773735941533,"output_mean_power":0.001837437732687517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.315971715150661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045788234917548806,"output_power":0.00036932466585561417,"reference_active":true,"delta_db":-0.9334558873662382},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0620858906342295e-05,"output_power":5.541526301467926e-05,"reference_active":true,"delta_db":-0.38992686528572773},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2498045919301085e-06,"output_power":2.161135170858175e-06,"reference_active":true,"delta_db":-0.1746286763730181}],"active_band_count":3,"active_band_mae_db":0.499337143008328}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001241557965588199,"log_rms_mae_db":2.520774794420022,"rms_envelope_nrmse":0.1825904464760484,"envelope_correlation":0.9781384419328096,"rms_gain_error_db":-0.4519464935887464,"reference_mean_power":0.0006561031712122295,"output_mean_power":0.000591258622890926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1091236930101454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020492597233413894,"output_power":0.00016124975408812864,"reference_active":true,"delta_db":-1.040979434143646},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.133021650590788e-06,"output_power":7.549036504322601e-06,"reference_active":true,"delta_db":0.24617983242264746},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2691438355846274e-06,"output_power":3.7231775055405233e-06,"reference_active":true,"delta_db":-0.5942704520095463}],"active_band_count":3,"active_band_mae_db":0.62714323952528}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011294582885045651,"log_rms_mae_db":2.0028762386937684,"rms_envelope_nrmse":0.1434526893556019,"envelope_correlation":0.9868629920558227,"rms_gain_error_db":-0.4511692877072414,"reference_mean_power":0.0006591401463258689,"output_mean_power":0.0005941017549974781},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.002660056119491e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019159735065259618,"output_power":0.0001490603697553035,"reference_active":true,"delta_db":-1.0902730530243276},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.047849758491007e-06,"output_power":3.6090915261516056e-06,"reference_active":true,"delta_db":0.7340434088936177},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0860941276748168e-07,"output_power":2.399308443294305e-07,"reference_active":true,"delta_db":0.6075218177300701}],"active_band_count":3,"active_band_mae_db":0.8106127598826718}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":4.171331827602278,"rms_envelope_nrmse":0.34749802292333887,"envelope_correlation":0.9548868724043619,"rms_gain_error_db":-2.712769115995707,"reference_mean_power":8.990510589252756e-05,"output_mean_power":4.8140150679988225e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067721726798636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1343959051909258e-05,"output_power":5.2000882127797095e-06,"reference_active":true,"delta_db":-3.387539392344129},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.893195477577151e-06,"output_power":1.8266802090532675e-06,"reference_active":true,"delta_db":-3.28643686729047},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973785070466086e-07,"output_power":3.539123158946061e-07,"reference_active":true,"delta_db":-3.527688492488915}],"active_band_count":3,"active_band_mae_db":3.4005549173745044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":3.750323817716324,"rms_envelope_nrmse":0.32666570580975374,"envelope_correlation":0.9666162978077096,"rms_gain_error_db":-2.7127343172706158,"reference_mean_power":9.03558904947943e-05,"output_mean_power":4.838191327871489e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9012156385948518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1053455137857865e-05,"output_power":4.9682047789698225e-06,"reference_active":true,"delta_db":-3.4729856468716407},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0905012137138105e-06,"output_power":8.857107014082928e-07,"reference_active":true,"delta_db":-3.729585320526066},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495315305956298e-08,"output_power":3.441898885919613e-08,"reference_active":true,"delta_db":-2.758021304050768}],"active_band_count":3,"active_band_mae_db":3.320197423816158}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"golf","status":"measured","mos":{"status":"measured","value":2.5249738693237305,"predicted_mos":2.5249738693237305,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"6574ee81e23e57180c8439eefde64e296e1c283270ea18763e7478da4f11cc18","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2125456379999378},"started_at":"2026-10-13T05:02:20.001622+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029638439991686027,"native_input_samples":144000,"native_input_sha256":"056e606e477add8239007daff6716be0de07f16a5b5f447c49b009a74ec30c4b","output_sha256":"cfdf6cd8e4ce4a0aaac42dc64524b2e5f802ca959055d5d7e1eddf933d36d173","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6597739700000602,"synthesis_s":0.1690045800005464,"total_s":0.8287785500006066,"rtf":0.13812975833343444,"output_peak":0.5979775786399841,"output_rms":0.05051895846076115,"finished_at":"2026-10-13T05:02:20.850203+00:00","analysis_stages":{"external_f0_s":0.09310946200002945},"evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"cfdf6cd8e4ce4a0aaac42dc64524b2e5f802ca959055d5d7e1eddf933d36d173","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.92379360041007,"output_peak":0.6107290823237265},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.5381427768190985,"spectral_convergence":0.5025454108630178,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.915415907005145,"spectral_convergence":0.3385266367172847,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.143157354114435,"spectral_convergence":0.4097283050777163,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.650015258246881e-05,"output_power":5.7372848451662727e-05,"delta_db":0.9125203434712839,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00039235712020654266,"output_power":0.00034921397538800573,"delta_db":-0.5058992261813694,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002038076296786429,"output_power":0.0014348762195318626,"delta_db":-1.524060000782942,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007268491123858984,"output_power":0.0005543551606936981,"delta_db":-1.1765616953375164,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019257205169468836,"output_power":0.0001393742558575262,"delta_db":-1.4041069584676291,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.758885001792126e-05,"output_power":2.1159455597168473e-05,"delta_db":-2.495545495162473,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3364489532338688,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.2817251589200889,"pitch":null,"output_rms_dbfs":-25.92379360041007,"output_peak":0.6107290823237265,"reference_rms_dbfs":-24.642068441489982,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.19890534597956},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.7936602527792754,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.058599859928791194,"rms_dbfs":-24.642068441489982,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.56263709105697,"mean":-2.6967171237544344e-06,"zero_energy":false},"output":{"rms":0.05056037890997183,"rms_dbfs":-25.92379360041007,"peak":0.6107290823237265,"peak_dbfs":-4.283027976634532,"crest_factor_db":21.64076562377554,"mean":0.00030796395110622407,"zero_energy":false},"rms_difference_db":-1.2817251589200893},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":1.5115545115314703e-06,"output_rms_dbfs":-58.20576186143835,"output_zero_energy":false,"output_peak":0.005187627790303513,"output_energy_fraction":7.897076379980861e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":142,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.8230088495575221,"absolute_difference_cents_median":9.094478397607503,"absolute_difference_cents_p95":39.725007762867804,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019822260308892205,"log_rms_mae_db":2.2775716319858743,"rms_envelope_nrmse":0.19829245412380792,"envelope_correlation":0.9799031304700128,"rms_gain_error_db":-1.281604220975126,"reference_mean_power":0.003436665458226562,"output_mean_power":0.00255844942159704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.9251706196653543e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006380818265983564,"output_power":0.0005613080141126863,"reference_active":true,"delta_db":-0.5567513259351005},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011201822248274835,"output_power":9.072284999475171e-05,"reference_active":true,"delta_db":-0.9157199213574457},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2917525554906997e-05,"output_power":2.727531177663885e-05,"reference_active":true,"delta_db":-1.9686495082125564}],"active_band_count":3,"active_band_mae_db":1.1470402518350342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001679501427730559,"log_rms_mae_db":1.9771048512082172,"rms_envelope_nrmse":0.16744380854906618,"envelope_correlation":0.9906466172535876,"rms_gain_error_db":-1.2813099907395131,"reference_mean_power":0.0034533076239279628,"output_mean_power":0.0025710129749059825},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8219070609611747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005851743441037445,"output_power":0.0005128702873275515,"reference_active":true,"delta_db":-0.5727773768560438},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16940543587153e-05,"output_power":4.9799114683117565e-05,"reference_active":true,"delta_db":-0.9302168974275786},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.234118702927627e-06,"output_power":1.5631351089759975e-06,"reference_active":true,"delta_db":-1.5510972655863886}],"active_band_count":3,"active_band_mae_db":1.0180305132900036}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001900450506992256,"log_rms_mae_db":2.7872785666046567,"rms_envelope_nrmse":0.22792740405717424,"envelope_correlation":0.9709888978486371,"rms_gain_error_db":-1.3928967172352817,"reference_mean_power":0.0021985431333935977,"output_mean_power":0.001595310849461183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4072638549255544e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004871295552152088,"output_power":0.0004005894685987239,"reference_active":true,"delta_db":-0.849449523573092},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011753835053341866,"output_power":9.419945730159238e-05,"reference_active":true,"delta_db":-0.9613119101969889},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.020244888714106e-05,"output_power":2.4754131884318943e-05,"reference_active":true,"delta_db":-2.106048080193452}],"active_band_count":3,"active_band_mae_db":1.3056031713211775}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014767968623269652,"log_rms_mae_db":2.299497322823061,"rms_envelope_nrmse":0.19400918207725754,"envelope_correlation":0.9839129352493542,"rms_gain_error_db":-1.3927248209312728,"reference_mean_power":0.0022093216638216975,"output_mean_power":0.0016031954403879908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.318048563317358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004570018667401013,"output_power":0.0003744900304005592,"reference_active":true,"delta_db":-0.8647771357182703},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064421339754299e-05,"output_power":4.898347801103301e-05,"reference_active":true,"delta_db":-0.9273974886588533},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2515614156024925e-06,"output_power":1.5609886735648306e-06,"reference_active":true,"delta_db":-1.5908404580461364}],"active_band_count":3,"active_band_mae_db":1.1276716941410865}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001241557965588199,"log_rms_mae_db":2.0153833308516504,"rms_envelope_nrmse":0.1698323789505832,"envelope_correlation":0.9886570869281334,"rms_gain_error_db":-1.0874492549845085,"reference_mean_power":0.0006570242425001301,"output_mean_power":0.000511489200713865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.117918428694023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020465900553461264,"output_power":0.0001720340831019048,"reference_active":true,"delta_db":-0.7541636238045536},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.114380757391017e-06,"output_power":7.1654373417976745e-06,"reference_active":true,"delta_db":0.031055980134401962},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.273968019174362e-06,"output_power":3.843422885835905e-06,"reference_active":true,"delta_db":-0.461130963053388}],"active_band_count":3,"active_band_mae_db":0.41545018899744784}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00011294582885045651,"log_rms_mae_db":1.6616807234600646,"rms_envelope_nrmse":0.1336832093546839,"envelope_correlation":0.9968985933271441,"rms_gain_error_db":-1.0865079784330387,"reference_mean_power":0.0006600292454819431,"output_mean_power":0.00051393995415405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.011218401305624e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019124746358779017,"output_power":0.00016061537209570784,"reference_active":true,"delta_db":-0.7580857587888366},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.028633485934512e-06,"output_power":3.1789860586036368e-06,"reference_active":true,"delta_db":0.21041903532254647},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0814581352130082e-07,"output_power":2.2489660166471174e-07,"reference_active":true,"delta_db":0.33615212752298645}],"active_band_count":3,"active_band_mae_db":0.43488564054478984}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":3.084913836605541,"rms_envelope_nrmse":0.27458095139287125,"envelope_correlation":0.9748527524198545,"rms_gain_error_db":-2.0926423673684718,"reference_mean_power":9.005455809846362e-05,"output_mean_power":5.562134195691309e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.0496554977461487e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1216080536300525e-05,"output_power":7.107875644088207e-06,"reference_active":true,"delta_db":-1.9810129805952192},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.880479549512068e-06,"output_power":1.986715926205032e-06,"reference_active":true,"delta_db":-2.907496257569655},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.936071058521691e-07,"output_power":4.517771026730889e-07,"reference_active":true,"delta_db":-2.4468133189507397}],"active_band_count":3,"active_band_mae_db":2.445107519038538}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":2.871843462403973,"rms_envelope_nrmse":0.250676656568198,"envelope_correlation":0.9849785252260411,"rms_gain_error_db":-2.0926987002015114,"reference_mean_power":9.050658474578348e-05,"output_mean_power":5.5899806832220495e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.8839313099922835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0931187478790267e-05,"output_power":6.867484107827719e-06,"reference_active":true,"delta_db":-2.0186967978647354},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.081492654427627e-06,"output_power":1.026983963687964e-06,"reference_active":true,"delta_db":-3.0681122044356046},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.468075707253601e-08,"output_power":2.830791243854464e-08,"reference_active":true,"delta_db":-3.5886725099094052}],"active_band_count":3,"active_band_mae_db":2.891827170736582}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.570922374725342,"predicted_mos":2.570922374725342,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"5664e9911fda68c6d2fb4d2205cfed3897d80ee0b19e3c198411e6517980b382","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.996718353000688},"started_at":"2026-10-13T05:08:55.568176+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027801560008811066,"native_input_samples":144000,"native_input_sha256":"056e606e477add8239007daff6716be0de07f16a5b5f447c49b009a74ec30c4b","output_sha256":"bb0fbbe8c396e64906d8af40590b0a9c41931a7f38cf64ef15c78bc468d97bde","output_samples":144120,"length_delta_samples":120,"analysis_s":2.232599418000973,"synthesis_s":3.2871423220003635,"total_s":5.519741740001336,"rtf":0.9199569566668894,"output_peak":0.5283111333847046,"output_rms":0.055938487661393775,"finished_at":"2026-10-13T05:09:01.099742+00:00","analysis_stages":{"external_f0_s":1.4787622389994795},"evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"bb0fbbe8c396e64906d8af40590b0a9c41931a7f38cf64ef15c78bc468d97bde","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.03869269041231,"output_peak":0.5236461619268297},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.12243788057849,"spectral_convergence":0.5417724201181641,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.322095005430537,"spectral_convergence":0.2950937725710391,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.620698063252371,"spectral_convergence":0.3159379193417977,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.891025349166477e-05,"delta_db":0.22645090096103868,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.0003612161747153136,"delta_db":-0.3518598591988759,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0018399625853605648,"delta_db":-0.4370394556625714,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0006940846684318833,"delta_db":-0.1936774431274282,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.00016770605921773038,"delta_db":-0.5933500182835485,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":2.491833204899562e-05,"delta_db":-1.778149738136048,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5967545692282518,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3860570530997817,"pitch":null,"output_rms_dbfs":-25.035134374670104,"output_peak":0.5236461619268297,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.688410316420466},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8296462075411659,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05855259318441331,"rms_dbfs":-24.64907732157032,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.569645971137312,"mean":-2.2611618041992187e-06,"zero_energy":false},"output":{"rms":0.05600712520087201,"rms_dbfs":-25.0351343746701,"peak":0.5236461619268297,"peak_dbfs":-5.619241505740682,"crest_factor_db":19.415892868929422,"mean":-0.00012445959825462287,"zero_energy":false},"rms_difference_db":-0.3860570530997798},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":1.596555965335811e-06,"output_rms_dbfs":-57.96815853190608,"output_zero_energy":false,"output_peak":0.020024181187429775,"output_energy_fraction":6.786351041391354e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":126,"both_estimated_voiced_frames":95,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.8407079646017699,"absolute_difference_cents_median":5.497841861927966,"absolute_difference_cents_p95":32.1308602761767,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019822260308892205,"log_rms_mae_db":2.1636112435888095,"rms_envelope_nrmse":0.16288652344629628,"envelope_correlation":0.9776502895832034,"rms_gain_error_db":-0.386021798726402,"reference_mean_power":0.003431222806852645,"output_mean_power":0.00313940062257569},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.922662834898126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006393448264747715,"output_power":0.0006006332279843482,"reference_active":true,"delta_db":-0.27125800863942695},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011197882141156391,"output_power":0.00010253481702331706,"reference_active":true,"delta_db":-0.3826453144590289},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.279682095777351e-05,"output_power":3.208105724345692e-05,"reference_active":true,"delta_db":-1.2516283767159295}],"active_band_count":3,"active_band_mae_db":0.6351772332714618}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001679501427730559,"log_rms_mae_db":1.694250952257374,"rms_envelope_nrmse":0.09805558139406755,"envelope_correlation":0.9926603230805348,"rms_gain_error_db":-0.3857510339875874,"reference_mean_power":0.003447933525092644,"output_mean_power":0.003154886798454765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8195786138583623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005865504590263928,"output_power":0.0005502251394385393,"reference_active":true,"delta_db":-0.27764949851918264},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16977320654735e-05,"output_power":5.59375370172553e-05,"reference_active":true,"delta_db":-0.42565860124077914},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.232527996839891e-06,"output_power":1.6270915022840766e-06,"reference_active":true,"delta_db":-1.3738493691073483}],"active_band_count":3,"active_band_mae_db":0.69238582295577}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001900450506992256,"log_rms_mae_db":2.402051004162276,"rms_envelope_nrmse":0.18268671343376428,"envelope_correlation":0.9728547595691204,"rms_gain_error_db":-0.432385106567103,"reference_mean_power":0.002194955532035717,"output_mean_power":0.0019869512672710476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4050180992423441e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048796679990628384,"output_power":0.00045050681843537317,"reference_active":true,"delta_db":-0.34688906224311145},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011744031611151056,"output_power":0.00011010012859689818,"reference_active":true,"delta_db":-0.28029385326656253},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.011051692782721e-05,"output_power":2.919636828275297e-05,"reference_active":true,"delta_db":-1.3792942582611052}],"active_band_count":3,"active_band_mae_db":0.6688257245902598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014767968623269652,"log_rms_mae_db":1.8330560719302174,"rms_envelope_nrmse":0.1117783648873397,"envelope_correlation":0.9905907003476302,"rms_gain_error_db":-0.4321795966289359,"reference_mean_power":0.002205773735941533,"output_mean_power":0.0019968387765131805},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.315971715150661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045788234917548806,"output_power":0.00042206806393365687,"reference_active":true,"delta_db":-0.3537141009929754},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0620858906342295e-05,"output_power":5.6578047062601455e-05,"reference_active":true,"delta_db":-0.29974132790286734},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2498045919301085e-06,"output_power":1.6874258011132011e-06,"reference_active":true,"delta_db":-1.2492011357225654}],"active_band_count":3,"active_band_mae_db":0.6342188548728027}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001241557965588199,"log_rms_mae_db":1.6581137599690847,"rms_envelope_nrmse":0.1631044897023184,"envelope_correlation":0.9820134674861187,"rms_gain_error_db":-0.2267486491766536,"reference_mean_power":0.0006561031712122295,"output_mean_power":0.0006227263917957027},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1091236930101454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020492597233413894,"output_power":0.000190064226778641,"reference_active":true,"delta_db":-0.32696621232294854},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.133021650590788e-06,"output_power":5.875286095791457e-06,"reference_active":true,"delta_db":-0.8424452288531941},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2691438355846274e-06,"output_power":4.852026292991789e-06,"reference_active":true,"delta_db":0.5558235837182125}],"active_band_count":3,"active_band_mae_db":0.5750783416314517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011294582885045651,"log_rms_mae_db":1.0441505915067377,"rms_envelope_nrmse":0.094994703478174,"envelope_correlation":0.9940632993739775,"rms_gain_error_db":-0.22614180428703082,"reference_mean_power":0.0006591401463258689,"output_mean_power":0.0006256962954168659},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.002660056119491e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019159735065259618,"output_power":0.00017748123965478696,"reference_active":true,"delta_db":-0.33237046022049904},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.047849758491007e-06,"output_power":2.6293149714222605e-06,"reference_active":true,"delta_db":-0.6415094065487769},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0860941276748168e-07,"output_power":1.9922514069429387e-07,"reference_active":true,"delta_db":-0.19989758347213696}],"active_band_count":3,"active_band_mae_db":0.39125915008047096}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":1.6543247495798237,"rms_envelope_nrmse":0.21900605675115165,"envelope_correlation":0.9736840469610673,"rms_gain_error_db":-1.069027460431957,"reference_mean_power":8.990510589252756e-05,"output_mean_power":7.028806877242715e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067721726798636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1343959051909258e-05,"output_power":8.612263779116227e-06,"reference_active":true,"delta_db":-1.196473271138546},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.893195477577151e-06,"output_power":3.028009320229942e-06,"reference_active":true,"delta_db":-1.091490026264781},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973785070466086e-07,"output_power":3.9177310439817055e-07,"reference_active":true,"delta_db":-3.08629907339041}],"active_band_count":3,"active_band_mae_db":1.791420790264579}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":1.00177300875849,"rms_envelope_nrmse":0.16721278908110299,"envelope_correlation":0.9886758990735176,"rms_gain_error_db":-1.0690246193739914,"reference_mean_power":9.03558904947943e-05,"output_mean_power":7.064053967446222e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9012156385948518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1053455137857865e-05,"output_power":8.27087498171405e-06,"reference_active":true,"delta_db":-1.25946596627956},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0905012137138105e-06,"output_power":1.5420382143945024e-06,"reference_active":true,"delta_db":-1.321552876089303},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495315305956298e-08,"output_power":3.4883604629120734e-08,"reference_active":true,"delta_db":-2.6997888258526337}],"active_band_count":3,"active_band_mae_db":1.760269222740499}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.495739221572876,"predicted_mos":2.495739221572876,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"fe315a17dad174f2b881f01afa447fb67b560977e2b5e1602460b8ebc81edf4a","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0522527339999215},"started_at":"2026-10-13T05:10:27.508509+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028529579994938103,"native_input_samples":144000,"native_input_sha256":"056e606e477add8239007daff6716be0de07f16a5b5f447c49b009a74ec30c4b","output_sha256":"f3298496bc55856f03689d512869439877db3c3cd7ee29869c0aca0e35fcb0b4","output_samples":144240,"length_delta_samples":240,"analysis_s":0.769365159998415,"synthesis_s":0.1345465360009257,"total_s":0.9039116959993407,"rtf":0.15065194933322346,"output_peak":0.6243997812271118,"output_rms":0.056852007303932105,"finished_at":"2026-10-13T05:10:28.423097+00:00","analysis_stages":{"external_f0_s":0.15310146999945573},"evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"f3298496bc55856f03689d512869439877db3c3cd7ee29869c0aca0e35fcb0b4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.89793776085658,"output_peak":0.6336237270696193},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.747834632285216,"spectral_convergence":0.17686101519543498,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.619152448750374,"spectral_convergence":0.17344842791210238,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.2774964426272275,"spectral_convergence":0.18977409172596285,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.630973149083977e-05,"delta_db":-0.010825606771905003,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.0003872701947899476,"delta_db":-0.04939101084170931,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0018953776214868188,"delta_db":-0.3081718887093963,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.000700198609562447,"delta_db":-0.15558951714529662,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.00017857075901558348,"delta_db":-0.32073411173811095,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":3.512341619484619e-05,"delta_db":-0.2873719263294625,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18868067692264678,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.24165362725879802,"pitch":null,"output_rms_dbfs":-24.89073094882912,"output_peak":0.6336237270696193,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.881494507887606},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8851757231747756,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05855259318441331,"rms_dbfs":-24.64907732157032,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.569645971137312,"mean":-2.2611618041992187e-06,"zero_energy":false},"output":{"rms":0.05694602999806469,"rms_dbfs":-24.89073094882912,"peak":0.6336237270696193,"peak_dbfs":-3.963371365035716,"crest_factor_db":20.9273595837934,"mean":-6.168491343135738e-06,"zero_energy":false},"rms_difference_db":-0.24165362725879902},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":2.1574800335731236e-06,"output_rms_dbfs":-56.66053214778056,"output_zero_energy":false,"output_peak":0.005995534756400304,"output_energy_fraction":8.87071480264886e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":111,"both_estimated_voiced_frames":100,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8849557522123894,"absolute_difference_cents_median":0.8431749744402512,"absolute_difference_cents_p95":5.705468581309847,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019822260308892205,"log_rms_mae_db":0.8440526501577098,"rms_envelope_nrmse":0.07445422054060988,"envelope_correlation":0.9956973916136922,"rms_gain_error_db":-0.241638076159255,"reference_mean_power":0.003431222806852645,"output_mean_power":0.003245526147162496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.922662834898126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006393448264747715,"output_power":0.0006060241873249478,"reference_active":true,"delta_db":-0.23245197402500467},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011197882141156391,"output_power":0.00010546219009772513,"reference_active":true,"delta_db":-0.26039106253521355},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.279682095777351e-05,"output_power":3.143121533489261e-05,"reference_active":true,"delta_db":-1.3405033587244228}],"active_band_count":3,"active_band_mae_db":0.6111154650948804}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001679501427730559,"log_rms_mae_db":0.5304139327191828,"rms_envelope_nrmse":0.04660381424699237,"envelope_correlation":0.9986610169882129,"rms_gain_error_db":-0.24148652597141196,"reference_mean_power":0.003447933525092644,"output_mean_power":0.003261446295311314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8195786138583623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005865504590263928,"output_power":0.0005537962876000706,"reference_active":true,"delta_db":-0.24955338893571363},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16977320654735e-05,"output_power":5.981340061701251e-05,"reference_active":true,"delta_db":-0.1347070579274388},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.232527996839891e-06,"output_power":1.7520676026819703e-06,"reference_active":true,"delta_db":-1.052460546320493}],"active_band_count":3,"active_band_mae_db":0.47890699772788176}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001900450506992256,"log_rms_mae_db":1.0691448383261846,"rms_envelope_nrmse":0.09324779028403674,"envelope_correlation":0.9933436672355452,"rms_gain_error_db":-0.27906585704154274,"reference_mean_power":0.002194955532035717,"output_mean_power":0.0020583496055180343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4050180992423441e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048796679990628384,"output_power":0.0004498266995388844,"reference_active":true,"delta_db":-0.35345045164093253},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011744031611151056,"output_power":0.00010997996967660473,"reference_active":true,"delta_db":-0.28503615951138206},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.011051692782721e-05,"output_power":2.8723456770479333e-05,"reference_active":true,"delta_db":-1.4502155444459501}],"active_band_count":3,"active_band_mae_db":0.6962340518660882}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014767968623269652,"log_rms_mae_db":0.6320509384421013,"rms_envelope_nrmse":0.05640476676355847,"envelope_correlation":0.9979939881870236,"rms_gain_error_db":-0.2790348178617585,"reference_mean_power":0.002205773735941533,"output_mean_power":0.0020685093080196637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.315971715150661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045788234917548806,"output_power":0.0004186654373628338,"reference_active":true,"delta_db":-0.38886792821455485},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0620858906342295e-05,"output_power":5.937652940559505e-05,"reference_active":true,"delta_db":-0.09007276150249309},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2498045919301085e-06,"output_power":1.7279653728358112e-06,"reference_active":true,"delta_db":-1.146097635669534}],"active_band_count":3,"active_band_mae_db":0.5416794417955274}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001241557965588199,"log_rms_mae_db":0.8604649025223443,"rms_envelope_nrmse":0.0635790718142834,"envelope_correlation":0.9975484084840007,"rms_gain_error_db":-0.17437503486169456,"reference_mean_power":0.0006561031712122295,"output_mean_power":0.0006302816066407291},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1091236930101454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020492597233413894,"output_power":0.0002020765721622503,"reference_active":true,"delta_db":-0.06081038098128422},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.133021650590788e-06,"output_power":6.444348521948506e-06,"reference_active":true,"delta_db":-0.4409452241196698},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2691438355846274e-06,"output_power":4.5337514059163265e-06,"reference_active":true,"delta_db":0.26116916004757895}],"active_band_count":3,"active_band_mae_db":0.254308255049511}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011294582885045651,"log_rms_mae_db":0.606102071873272,"rms_envelope_nrmse":0.03959984536541254,"envelope_correlation":0.9992336736448786,"rms_gain_error_db":-0.1738299921226816,"reference_mean_power":0.0006591401463258689,"output_mean_power":0.0006332785306391514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.002660056119491e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019159735065259618,"output_power":0.00018959639027977675,"reference_active":true,"delta_db":-0.04559434937071848},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.047849758491007e-06,"output_power":2.7859738987455737e-06,"reference_active":true,"delta_db":-0.39016511683177646},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0860941276748168e-07,"output_power":2.113212385623979e-07,"reference_active":true,"delta_db":0.056092468933889014}],"active_band_count":3,"active_band_mae_db":0.16395064504546134}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":0.9245816431722336,"rms_envelope_nrmse":0.08570292105828844,"envelope_correlation":0.9958227590077655,"rms_gain_error_db":-0.32300175187429847,"reference_mean_power":8.990510589252756e-05,"output_mean_power":8.346111487102784e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067721726798636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1343959051909258e-05,"output_power":1.0599310690110967e-05,"reference_active":true,"delta_db":-0.29487027668003557},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.893195477577151e-06,"output_power":3.4934519043444202e-06,"reference_active":true,"delta_db":-0.4705144168008428},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973785070466086e-07,"output_power":8.340011308043989e-07,"reference_active":true,"delta_db":0.1950211420311886}],"active_band_count":3,"active_band_mae_db":0.3201352785040223}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":0.7052137813818874,"rms_envelope_nrmse":0.053077000468020996,"envelope_correlation":0.9989429385766231,"rms_gain_error_db":-0.32298624811466975,"reference_mean_power":9.03558904947943e-05,"output_mean_power":8.387988871352817e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9012156385948518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1053455137857865e-05,"output_power":1.0363304777913445e-05,"reference_active":true,"delta_db":-0.2799978227665771},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0905012137138105e-06,"output_power":1.9025539698441672e-06,"reference_active":true,"delta_db":-0.40913438760638027},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495315305956298e-08,"output_power":6.98606715708206e-08,"reference_active":true,"delta_db":0.31632518059445136}],"active_band_count":3,"active_band_mae_db":0.3351524636558029}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-960-0019","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2419686317443848,"predicted_mos":3.2419686317443848,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288000,"input_sample_rate":48000,"input_seconds":6.0,"input_float32_sha256":"1fef78554942f43b89f887ced2c2c2c3bda9c7ef24eaa32c5475e859a7f57414","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1262003859992546},"started_at":"2026-10-13T05:14:20.660047+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009778587000255357,"native_input_samples":288000,"native_input_sha256":"5c805f79551fbe27d4274fd7647ed9a357a21c575fb76a9f166b732ae5cb60c3","output_sha256":"66eb96490b936251487c72fc87d6d8032df4ee1cd0c99e57db602cd76dc9571b","output_samples":288000,"length_delta_samples":0,"analysis_s":3.4846771110005648,"synthesis_s":1.0504306980001274,"total_s":4.535107809000692,"rtf":0.7558513015001154,"output_peak":0.6619444685874062,"output_rms":0.05690134919586093,"finished_at":"2026-10-13T05:14:25.226414+00:00","evaluation_input":{"reference":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"66eb96490b936251487c72fc87d6d8032df4ee1cd0c99e57db602cd76dc9571b","native_sample_rate":48000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.292873605357874,"spectral_convergence":0.5108900382883881,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.275181154937414,"spectral_convergence":0.18763717445493897,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.462316356721857,"spectral_convergence":0.16411439838225467,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.2267023319559154e-05,"delta_db":-0.4075316013204738,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.0003623177121098092,"delta_db":-0.3386360975932199,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0018745296444098612,"delta_db":-0.35620624639889237,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0007158719171877015,"delta_db":-0.05944869841971547,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.0002086373302407808,"delta_db":0.35508261008508524,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":3.946202670226802e-05,"delta_db":0.21845443578871815,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2892266149343508,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2413384875517101,"pitch":null,"output_rms_dbfs":-24.890415809122032,"output_peak":0.6623627807633694,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676790372339049},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.292873605357874,"spectral_convergence":0.5108900382883881,"spectral_convergence_status":"ok","amplitude_floor":1.5171791511760096e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.275181154937414,"spectral_convergence":0.18763717445493897,"spectral_convergence_status":"ok","amplitude_floor":1.2250623067014244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.462316356721857,"spectral_convergence":0.16411439838225467,"spectral_convergence_status":"ok","amplitude_floor":7.883290232655995e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.642531119863568e-05,"output_power":4.2267023319559154e-05,"delta_db":-0.4075316013204738,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003916996423883643,"output_power":0.0003623177121098092,"delta_db":-0.3386360975932199,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0020347588099002725,"output_power":0.0018745296444098612,"delta_db":-0.35620624639889237,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007257385551015356,"output_power":0.0007158719171877015,"delta_db":-0.05944869841971547,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001922577058777525,"output_power":0.0002086373302407808,"delta_db":0.35508261008508524,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.752614415284284e-05,"output_power":3.946202670226802e-05,"delta_db":0.21845443578871815,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2892266149343508,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2413384875517101,"pitch":null,"output_rms_dbfs":-24.890415809122032,"output_peak":0.6623627807633694,"reference_rms_dbfs":-24.64907732157032,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676790372339049},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8654881570993457,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05855259318441331,"rms_dbfs":-24.64907732157032,"peak":0.625213623046875,"peak_dbfs":-4.079431350433008,"crest_factor_db":20.569645971137312,"mean":-2.2611618041992187e-06,"zero_energy":false},"output":{"rms":0.05694809613999383,"rms_dbfs":-24.890415809122032,"peak":0.6623627807633694,"peak_dbfs":-3.578081584947358,"crest_factor_db":21.31233422417467,"mean":-0.00013893919683146698,"zero_energy":false},"rms_difference_db":-0.24133848755171172},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.2396896200953057e-06,"threshold_dbfs":-54.894965955763524,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":80,"quiet_samples":12800,"reference_mean_power":2.2128017735667525e-06,"output_mean_power":1.7438540757785526e-06,"output_rms_dbfs":-57.58489859274795,"output_zero_energy":false,"output_peak":0.007220528783241349,"output_energy_fraction":7.169526234555195e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":109,"both_estimated_voiced_frames":98,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8672566371681416,"absolute_difference_cents_median":2.1125691289516713,"absolute_difference_cents_p95":9.756842472340988,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019822260308892205,"log_rms_mae_db":1.3597104265811946,"rms_envelope_nrmse":0.11416415326758154,"envelope_correlation":0.9891262927361639,"rms_gain_error_db":-0.24129532665651604,"reference_mean_power":0.003431222806852645,"output_mean_power":0.0032457822974053677},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.922662834898126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006393448264747715,"output_power":0.0006219061420247657,"reference_active":true,"delta_db":-0.12010309269501875},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011197882141156391,"output_power":0.00010771671591778356,"reference_active":true,"delta_db":-0.16852788078868644},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.279682095777351e-05,"output_power":4.259446015569156e-05,"reference_active":true,"delta_db":-0.02058391514237818}],"active_band_count":3,"active_band_mae_db":0.1030716295420278}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001679501427730559,"log_rms_mae_db":0.9011625435954479,"rms_envelope_nrmse":0.05532733088451634,"envelope_correlation":0.9979091971219316,"rms_gain_error_db":-0.24138049062820816,"reference_mean_power":0.003447933525092644,"output_mean_power":0.0032615259262561004},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8195786138583623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005865504590263928,"output_power":0.0005711218521876315,"reference_active":true,"delta_db":-0.11576601599195252},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.16977320654735e-05,"output_power":5.908731878800622e-05,"reference_active":true,"delta_db":-0.1877491678650348},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.232527996839891e-06,"output_power":2.2472318864473275e-06,"reference_active":true,"delta_db":0.028509747031029382}],"active_band_count":3,"active_band_mae_db":0.11067497696267221}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001900450506992256,"log_rms_mae_db":1.7128425615217633,"rms_envelope_nrmse":0.1267641083025491,"envelope_correlation":0.987505500051166,"rms_gain_error_db":-0.34840065955904176,"reference_mean_power":0.002194955532035717,"output_mean_power":0.0020257491262953716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4050180992423441e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048796679990628384,"output_power":0.0004636739187583262,"reference_active":true,"delta_db":-0.22177606742168254},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011744031611151056,"output_power":0.00011244686703153907,"reference_active":true,"delta_db":-0.18869851826103712},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.011051692782721e-05,"output_power":4.11271704580628e-05,"reference_active":true,"delta_db":0.10870572128480274}],"active_band_count":3,"active_band_mae_db":0.17306010232250746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014767968623269652,"log_rms_mae_db":1.119662868172429,"rms_envelope_nrmse":0.06741181825610022,"envelope_correlation":0.9972954969987093,"rms_gain_error_db":-0.3484129520553167,"reference_mean_power":0.002205773735941533,"output_mean_power":0.002035727606212292},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.315971715150661e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045788234917548806,"output_power":0.00043545267609341066,"reference_active":true,"delta_db":-0.21812938439089163},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0620858906342295e-05,"output_power":5.8047403390349913e-05,"reference_active":true,"delta_db":-0.1883928797451148},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2498045919301085e-06,"output_power":2.2647150313338514e-06,"reference_active":true,"delta_db":0.02868763741003995}],"active_band_count":3,"active_band_mae_db":0.14506996718201545}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001241557965588199,"log_rms_mae_db":1.1517113548555642,"rms_envelope_nrmse":0.11144866786439293,"envelope_correlation":0.9916535023844024,"rms_gain_error_db":0.020447966225429777,"reference_mean_power":0.0006561031712122295,"output_mean_power":0.0006591995974882138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1091236930101454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020492597233413894,"output_power":0.00020299562119446073,"reference_active":true,"delta_db":-0.041103345311336165},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.133021650590788e-06,"output_power":6.89232989959922e-06,"reference_active":true,"delta_db":-0.149074855530412},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2691438355846274e-06,"output_power":4.601303402304121e-06,"reference_active":true,"delta_db":0.32540083607314724}],"active_band_count":3,"active_band_mae_db":0.1718596789716318}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011294582885045651,"log_rms_mae_db":0.6276875222555204,"rms_envelope_nrmse":0.048751495412320184,"envelope_correlation":0.9983669925822748,"rms_gain_error_db":0.020104256282342418,"reference_mean_power":0.0006591401463258689,"output_mean_power":0.0006621984954828925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.002660056119491e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019159735065259618,"output_power":0.00018941088959794052,"reference_active":true,"delta_db":-0.049845556839325594},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.047849758491007e-06,"output_power":3.128331217299363e-06,"reference_active":true,"delta_db":0.11319173508162547},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0860941276748168e-07,"output_power":2.285281889881341e-07,"reference_active":true,"delta_db":0.3960587745624915}],"active_band_count":3,"active_band_mae_db":0.18636535549448086}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.245162939338534e-05,"log_rms_mae_db":1.2440236121818968,"rms_envelope_nrmse":0.13280963998893794,"envelope_correlation":0.9898918449318567,"rms_gain_error_db":0.3718540869444613,"reference_mean_power":8.990510589252756e-05,"output_mean_power":9.794217983295467e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.067721726798636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1343959051909258e-05,"output_power":1.236269314056376e-05,"reference_active":true,"delta_db":0.3734843940138489},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.893195477577151e-06,"output_power":4.259383834930868e-06,"reference_active":true,"delta_db":0.3904056812377672},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973785070466086e-07,"output_power":7.224312933884501e-07,"reference_active":true,"delta_db":-0.4286797533962139}],"active_band_count":3,"active_band_mae_db":0.39752327621594336}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.7197745144437054e-05,"log_rms_mae_db":0.7533734369330978,"rms_envelope_nrmse":0.07683993525775311,"envelope_correlation":0.9973341296312497,"rms_gain_error_db":0.3718565633996604,"reference_mean_power":9.03558904947943e-05,"output_mean_power":9.843331848894635e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9012156385948518e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1053455137857865e-05,"output_power":1.2050225178839188e-05,"reference_active":true,"delta_db":0.3749710958235447},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0905012137138105e-06,"output_power":2.1807995570024546e-06,"reference_active":true,"delta_db":0.18365326238630553},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495315305956298e-08,"output_power":5.6975805225359516e-08,"reference_active":true,"delta_db":-0.5690976646094945}],"active_band_count":3,"active_band_mae_db":0.37590734093978156}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"hnm","status":"measured","mos":{"status":"measured","value":2.8297815322875977,"predicted_mos":2.8297815322875977,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b3ffcbd6f5265ab075a5d8369f722a62af2e9a806446c66b4d9ce92605d2cf52","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.994254485000056},"started_at":"2026-10-13T04:38:52.412156+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3466000382322818e-05,"native_input_samples":96000,"native_input_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","output_sha256":"715786ac519b6d0a47798b3f563713c22504dc0cd08647cc2cd900697e5e5b2c","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0824571690000084,"synthesis_s":0.5909346959997492,"total_s":1.6733918649997577,"rtf":0.2788986441666263,"output_peak":0.6081810744641096,"output_rms":0.04299894424288198,"finished_at":"2026-10-13T04:38:54.090865+00:00","evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"715786ac519b6d0a47798b3f563713c22504dc0cd08647cc2cd900697e5e5b2c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.730810234189973,"spectral_convergence":0.31231315015903616,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.619927221948608,"spectral_convergence":0.2974340158664754,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.2008626629761245,"spectral_convergence":0.3136983138155875,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.647892457267107e-05,"delta_db":0.15910954449543643,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.000265797850097292,"delta_db":0.23957414786050724,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0010527610928647412,"delta_db":-0.1664485291449887,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.00030926085233727585,"delta_db":-0.01545959082873024,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.00012879380998954558,"delta_db":-0.13576243669696345,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":3.581667614094748e-05,"delta_db":-0.14081731768677158,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.14286192778556628,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07269182753774478,"pitch":null,"output_rms_dbfs":-27.330844151254524,"output_peak":0.6081810744641096,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.850533373038235},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.730810234189973,"spectral_convergence":0.31231315015903616,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.619927221948608,"spectral_convergence":0.2974340158664754,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.2008626629761245,"spectral_convergence":0.3136983138155875,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.647892457267107e-05,"delta_db":0.15910954449543643,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.000265797850097292,"delta_db":0.23957414786050724,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0010527610928647412,"delta_db":-0.1664485291449887,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.00030926085233727585,"delta_db":-0.01545959082873024,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.00012879380998954558,"delta_db":-0.13576243669696345,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":3.581667614094748e-05,"delta_db":-0.14081731768677158,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.14286192778556628,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07269182753774478,"pitch":null,"output_rms_dbfs":-27.330844151254524,"output_peak":0.6081810744641096,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.850533373038235},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8444920144216925,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043360310532128375,"rms_dbfs":-27.258152323716782,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.07422203496477,"mean":1.2998580932617187e-06,"zero_energy":false},"output":{"rms":0.04299894424288198,"rms_dbfs":-27.330844151254524,"peak":0.6081810744641096,"peak_dbfs":-4.319341969391412,"crest_factor_db":23.011502181863115,"mean":1.1965274146950488e-05,"zero_energy":false},"rms_difference_db":-0.07269182753774217},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":8.178126158326046e-07,"output_rms_dbfs":-60.87346194167811,"output_zero_energy":false,"output_peak":0.010548485141539143,"output_energy_fraction":0.00010615720185859822,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":133,"both_estimated_voiced_frames":123,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9044117647058824,"absolute_difference_cents_median":1.9194111688169464,"absolute_difference_cents_p95":8.11152766929499,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019413100530357063,"log_rms_mae_db":0.859220578335755,"rms_envelope_nrmse":0.08680853472409339,"envelope_correlation":0.9934920731018162,"rms_gain_error_db":-0.07267025278861264,"reference_mean_power":0.0018816420370153617,"output_mean_power":0.0018504185846983324},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1507655155627414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005726194534917747,"output_power":0.0005650128317124638,"reference_active":true,"delta_db":-0.058077871725653096},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.421513884012437e-05,"output_power":8.587501923040656e-05,"reference_active":true,"delta_db":-0.4025384513694779},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6796574139632317e-05,"output_power":1.6744449499057446e-05,"reference_active":true,"delta_db":-0.013498372763197975}],"active_band_count":3,"active_band_mae_db":0.15803823195277633}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016416501002274435,"log_rms_mae_db":0.4583074000477412,"rms_envelope_nrmse":0.04091644126694627,"envelope_correlation":0.9985248473414565,"rms_gain_error_db":-0.07268444952242045,"reference_mean_power":0.001890759403997248,"output_mean_power":0.0018593785824235739},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0781635397104837e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005383866955860211,"output_power":0.0005303796812757389,"reference_active":true,"delta_db":-0.06507441229889488},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.887402542633377e-05,"output_power":3.6412575872986405e-05,"reference_active":true,"delta_db":-0.2840811181075881},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650199092025118e-06,"output_power":1.3732210769195572e-06,"reference_active":true,"delta_db":-0.5677940615477861}],"active_band_count":3,"active_band_mae_db":0.3056498639847564}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017990629858894173,"log_rms_mae_db":1.4368942358280667,"rms_envelope_nrmse":0.11079474076061202,"envelope_correlation":0.9904137713064451,"rms_gain_error_db":-0.10015315082602652,"reference_mean_power":0.0012781810161111924,"output_mean_power":0.0012490420167177996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.355790711503392e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004472139003696596,"output_power":0.00043711981538992027,"reference_active":true,"delta_db":-0.09914799645900435},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.621773891011176e-05,"output_power":7.185821291843012e-05,"reference_active":true,"delta_db":-0.2557964847115004},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7010237353267383e-05,"output_power":1.9102517355602394e-05,"reference_active":true,"delta_db":0.5038022932757645}],"active_band_count":3,"active_band_mae_db":0.28624892481542313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000149706333097475,"log_rms_mae_db":0.7878975273596475,"rms_envelope_nrmse":0.053988452794389546,"envelope_correlation":0.9976862404591822,"rms_gain_error_db":-0.10016708466125529,"reference_mean_power":0.0012844845607787134,"output_mean_power":0.0012551978308013975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.840745806614025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236215690036061,"output_power":0.00041463078586197096,"reference_active":true,"delta_db":-0.09316519597274865},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.160014209802635e-05,"output_power":3.081294554919872e-05,"reference_active":true,"delta_db":-0.10955819057272083},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5445822829049618e-06,"output_power":1.5362446214139045e-06,"reference_active":true,"delta_db":-0.02350673588762004}],"active_band_count":3,"active_band_mae_db":0.07541004081102984}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010005798255669509,"log_rms_mae_db":1.3803178149452822,"rms_envelope_nrmse":0.13652521540730433,"envelope_correlation":0.9872407416473188,"rms_gain_error_db":-0.049811797826044515,"reference_mean_power":0.0003605696849433612,"output_mean_power":0.00035645772470522917},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5037729366163554e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014066740308353966,"output_power":0.0001353319587407817,"reference_active":true,"delta_db":-0.16793102182703687},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0544824834133837e-05,"output_power":2.470744687366889e-05,"reference_active":true,"delta_db":0.8012542740695356},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.44585128264573e-06,"output_power":5.62334489168879e-06,"reference_active":true,"delta_db":2.126981904586902}],"active_band_count":3,"active_band_mae_db":1.0320557334944915}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":0.8930529305824519,"rms_envelope_nrmse":0.07364278335154642,"envelope_correlation":0.9961795851484101,"rms_gain_error_db":-0.04981200833131575,"reference_mean_power":0.0003623811119711957,"output_mean_power":0.000358248476736393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.381538826489277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013551929688065346,"output_power":0.00013261270818771872,"reference_active":true,"delta_db":-0.09415995513083461},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.971075792205562e-06,"output_power":1.0419978901260003e-05,"reference_active":true,"delta_db":1.1634990099286777},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4163220581925866e-07,"output_power":3.3209806265022016e-07,"reference_active":true,"delta_db":1.3811152356139451}],"active_band_count":3,"active_band_mae_db":0.8795914002244857}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":1.3436965467405644,"rms_envelope_nrmse":0.1311861985656453,"envelope_correlation":0.9884335404796161,"rms_gain_error_db":-0.30369713008888954,"reference_mean_power":6.230386432542119e-05,"output_mean_power":5.809587151737431e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.022113691992469e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1604215620852607e-05,"output_power":1.1013928055739548e-05,"reference_active":true,"delta_db":-0.22673554648325134},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3772446197384044e-06,"output_power":2.241847775162531e-06,"reference_active":true,"delta_db":-0.2546775315002265},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.450999961693561e-07,"output_power":4.2739634798785997e-07,"reference_active":true,"delta_db":-0.1762678365621581}],"active_band_count":3,"active_band_mae_db":0.21922697151521198}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":0.8700449138370174,"rms_envelope_nrmse":0.08018148248693291,"envelope_correlation":0.996199442317345,"rms_gain_error_db":-0.30369839262995924,"reference_mean_power":6.261700596722032e-05,"output_mean_power":5.838784664991257e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8467488015072894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1232117533806206e-05,"output_power":1.0635084482503845e-05,"reference_active":true,"delta_db":-0.23720695121516477},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0675811682412763e-06,"output_power":1.0709518607802603e-06,"reference_active":true,"delta_db":0.013690453804070548},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5798002699765385e-08,"output_power":2.3922176225943515e-08,"reference_active":true,"delta_db":-0.3278539852155193}],"active_band_count":3,"active_band_mae_db":0.1929171300782515}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.090877056121826,"predicted_mos":3.090877056121826,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"dd9fca548692b014e22b613c7dabef70e43cc636bec731c678e643bcb585665a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.257658892998734},"started_at":"2026-10-13T04:41:48.635392+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.588400002423441e-05,"native_input_samples":96000,"native_input_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","output_sha256":"8c7779e9f69cfe930a4020ca55f28d9b4212c18a58312c82a25d92864f0aa820","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6249924730000203,"synthesis_s":0.1123011959998621,"total_s":1.7372936689998824,"rtf":0.2895489448333137,"output_peak":0.6117655844088385,"output_rms":0.052085107815628875,"finished_at":"2026-10-13T04:41:50.377876+00:00","evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"8c7779e9f69cfe930a4020ca55f28d9b4212c18a58312c82a25d92864f0aa820","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.257181494427537,"spectral_convergence":0.6478433166210844,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.219797003828948,"spectral_convergence":0.3704259628920197,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.73662886766741,"spectral_convergence":0.4093652589770809,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.009099660846148e-05,"delta_db":-0.3621579169737585,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.0003265126512957468,"delta_db":1.133059643765653,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0017818288601680523,"delta_db":2.1189130973688006,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.0003683296750484874,"delta_db":0.7436580229943629,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.0001436070597520672,"delta_db":0.33704556040065525,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":4.248921329286861e-05,"delta_db":0.6011167824715106,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8826585039957903,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5924236733865156,"pitch":null,"output_rms_dbfs":-25.665728650330266,"output_peak":0.6117655844088385,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.737869121974632},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.257181494427537,"spectral_convergence":0.6478433166210844,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.219797003828948,"spectral_convergence":0.3704259628920197,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.73662886766741,"spectral_convergence":0.4093652589770809,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.009099660846148e-05,"delta_db":-0.3621579169737585,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.0003265126512957468,"delta_db":1.133059643765653,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0017818288601680523,"delta_db":2.1189130973688006,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.0003683296750484874,"delta_db":0.7436580229943629,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.0001436070597520672,"delta_db":0.33704556040065525,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":4.248921329286861e-05,"delta_db":0.6011167824715106,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8826585039957903,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5924236733865156,"pitch":null,"output_rms_dbfs":-25.665728650330266,"output_peak":0.6117655844088385,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.737869121974632},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9245942695463354,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043360310532128375,"rms_dbfs":-27.258152323716782,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.07422203496477,"mean":1.2998580932617187e-06,"zero_energy":false},"output":{"rms":0.05208510781562887,"rms_dbfs":-25.665728650330266,"peak":0.6117655844088385,"peak_dbfs":-4.268299168201744,"crest_factor_db":21.397429482128523,"mean":1.781200283283435e-06,"zero_energy":false},"rms_difference_db":1.5924236733865165},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":8.406598820211277e-07,"output_rms_dbfs":-60.753796774937136,"output_zero_energy":false,"output_peak":0.015171463399603565,"output_energy_fraction":7.437113839335102e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":167,"both_estimated_voiced_frames":128,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.9411764705882353,"absolute_difference_cents_median":4.9722836995662565,"absolute_difference_cents_p95":1906.2333478562932,"difference_over_600_cents_frames":14,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019413100530357063,"log_rms_mae_db":1.424775248654291,"rms_envelope_nrmse":0.27633540901621795,"envelope_correlation":0.9757032582179824,"rms_gain_error_db":1.5925185033608318,"reference_mean_power":0.0018816420370153617,"output_mean_power":0.002715118927215202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1507655155627414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005726194534917747,"output_power":0.0009544854075404207,"reference_active":true,"delta_db":2.219031950159323},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.421513884012437e-05,"output_power":0.00015680891409079913,"reference_active":true,"delta_db":2.212500548596283},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6796574139632317e-05,"output_power":2.5156314625711146e-05,"reference_active":true,"delta_db":1.7542630648581254}],"active_band_count":3,"active_band_mae_db":2.0619318545379106}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016416501002274435,"log_rms_mae_db":1.03075483552293,"rms_envelope_nrmse":0.24545468823683445,"envelope_correlation":0.9862935998185501,"rms_gain_error_db":1.5927531981364298,"reference_mean_power":0.001890759403997248,"output_mean_power":0.0027284222903392595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0781635397104837e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005383866955860211,"output_power":0.0009103759271154364,"reference_active":true,"delta_db":2.28126446019572},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.887402542633377e-05,"output_power":6.678109835529895e-05,"reference_active":true,"delta_db":2.3499404327313616},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650199092025118e-06,"output_power":1.7102394781358848e-06,"reference_active":true,"delta_db":0.38537060465745565}],"active_band_count":3,"active_band_mae_db":1.6721918325281793}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017990629858894173,"log_rms_mae_db":1.7579497218753417,"rms_envelope_nrmse":0.3311614062410895,"envelope_correlation":0.975832173068918,"rms_gain_error_db":2.0134013762581358,"reference_mean_power":0.0012781810161111924,"output_mean_power":0.0020320411603594597},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.355790711503392e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004472139003696596,"output_power":0.0008125887944790941,"reference_active":true,"delta_db":2.593555352582619},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.621773891011176e-05,"output_power":0.0001431230430773925,"reference_active":true,"delta_db":2.736535008475074},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7010237353267383e-05,"output_power":2.6614882297317833e-05,"reference_active":true,"delta_db":1.9441417626594797}],"active_band_count":3,"active_band_mae_db":2.4247440412390575}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000149706333097475,"log_rms_mae_db":1.2973247732689381,"rms_envelope_nrmse":0.3021276992198508,"envelope_correlation":0.9863189969486392,"rms_gain_error_db":2.01352500700659,"reference_mean_power":0.0012844845607787134,"output_mean_power":0.0020421206136148414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.840745806614025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236215690036061,"output_power":0.0007825250374980321,"reference_active":true,"delta_db":2.665201776097688},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.160014209802635e-05,"output_power":6.253484709127895e-05,"reference_active":true,"delta_db":2.964330567190302},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5445822829049618e-06,"output_power":1.7587651840448984e-06,"reference_active":true,"delta_db":0.5639681091139642}],"active_band_count":3,"active_band_mae_db":2.0645001508006513}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010005798255669509,"log_rms_mae_db":1.3549728955818385,"rms_envelope_nrmse":0.16964564687789177,"envelope_correlation":0.987058800645755,"rms_gain_error_db":0.7470641834399687,"reference_mean_power":0.0003605696849433612,"output_mean_power":0.0004282482813866095},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5037729366163554e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014066740308353966,"output_power":0.00017153078859158177,"reference_active":true,"delta_db":0.8614861438429076},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0544824834133837e-05,"output_power":2.4281862463061892e-05,"reference_active":true,"delta_db":0.725795520539544},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.44585128264573e-06,"output_power":4.342857127424823e-06,"reference_active":true,"delta_db":1.00479011980443}],"active_band_count":3,"active_band_mae_db":0.8640239280622938}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":0.8497180027915944,"rms_envelope_nrmse":0.11379338700523808,"envelope_correlation":0.9968322554881117,"rms_gain_error_db":0.7470673153108104,"reference_mean_power":0.0003623811119711957,"output_mean_power":0.0004304000219710546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.381538826489277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013551929688065346,"output_power":0.00016533337770843447,"reference_active":true,"delta_db":0.8635939860645432},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.971075792205562e-06,"output_power":9.32471699709843e-06,"reference_active":true,"delta_db":0.6811872135166009},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4163220581925866e-07,"output_power":2.2183039774591006e-07,"reference_active":true,"delta_db":-0.37133760611532696}],"active_band_count":3,"active_band_mae_db":0.6387062685654903}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":1.4462955804343294,"rms_envelope_nrmse":0.1820914422568664,"envelope_correlation":0.9778844415116598,"rms_gain_error_db":0.25089518387302406,"reference_mean_power":6.230386432542119e-05,"output_mean_power":6.600920503932399e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.022113691992469e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1604215620852607e-05,"output_power":1.214076479726441e-05,"reference_active":true,"delta_db":0.19630255675757374},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3772446197384044e-06,"output_power":2.441981887478929e-06,"reference_active":true,"delta_db":0.11668565236653451},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.450999961693561e-07,"output_power":6.293553520818485e-07,"reference_active":true,"delta_db":1.504383392743627}],"active_band_count":3,"active_band_mae_db":0.6057905339559118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":0.8329495687514892,"rms_envelope_nrmse":0.09124426601885473,"envelope_correlation":0.9946940315727328,"rms_gain_error_db":0.2508925841427526,"reference_mean_power":6.261700596722032e-05,"output_mean_power":6.634093015509857e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8467488015072894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1232117533806206e-05,"output_power":1.1867133818490848e-05,"reference_active":true,"delta_db":0.23884200366040828},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0675811682412763e-06,"output_power":1.126779525445801e-06,"reference_active":true,"delta_db":0.2343804250059711},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5798002699765385e-08,"output_power":4.324730053846903e-08,"reference_active":true,"delta_db":2.243729204680795}],"active_band_count":3,"active_band_mae_db":0.9056505444490582}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"facodec","status":"measured","mos":{"status":"measured","value":3.874553918838501,"predicted_mos":3.874553918838501,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1336d9ce2fe6764f9098140f755fdd30d5002621b4c12a38a125dc339828aeaa","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.974910623001051},"started_at":"2026-10-13T04:50:16.092966+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.618300044967327e-05,"native_input_samples":96000,"native_input_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","output_sha256":"8b3d644eaaa781fa73942829a0bdb40bf23f88ff2bb4ea7cb50b0f21ca090ca1","output_samples":96000,"length_delta_samples":0,"analysis_s":2.0679512239994438,"synthesis_s":6.794678444000056,"total_s":8.8626296679995,"rtf":1.4771049446665832,"output_peak":0.5439934134483337,"output_rms":0.04092553508062052,"finished_at":"2026-10-13T04:50:24.965286+00:00","evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"8b3d644eaaa781fa73942829a0bdb40bf23f88ff2bb4ea7cb50b0f21ca090ca1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.267106001964578,"spectral_convergence":0.26374249927714183,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.575488536906271,"spectral_convergence":0.27925112394924007,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.771103102408011,"spectral_convergence":0.31274358309283595,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.0836516739260904e-05,"delta_db":-0.29799678644702415,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.00023666726170212693,"delta_db":-0.2645586313304487,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0009763012524789067,"delta_db":-0.49390832699560927,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.00023888945659859451,"delta_db":-1.136739250117774,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.000136753285473861,"delta_db":0.12466534584599646,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":3.545164864235135e-05,"delta_db":-0.18530575351110787,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4171956823746601,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5019603502582632,"pitch":null,"output_rms_dbfs":-27.760112673975044,"output_peak":0.5439934134483337,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.53789921375962},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.267106001964578,"spectral_convergence":0.26374249927714183,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.575488536906271,"spectral_convergence":0.27925112394924007,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.771103102408011,"spectral_convergence":0.31274358309283595,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.0836516739260904e-05,"delta_db":-0.29799678644702415,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.00023666726170212693,"delta_db":-0.2645586313304487,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0009763012524789067,"delta_db":-0.49390832699560927,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.00023888945659859451,"delta_db":-1.136739250117774,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.000136753285473861,"delta_db":0.12466534584599646,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":3.545164864235135e-05,"delta_db":-0.18530575351110787,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4171956823746601,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5019603502582632,"pitch":null,"output_rms_dbfs":-27.760112673975044,"output_peak":0.5439934134483337,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.53789921375962},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9146383766311099,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043360310532128375,"rms_dbfs":-27.258152323716782,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.07422203496477,"mean":1.2998580932617187e-06,"zero_energy":false},"output":{"rms":0.04092553508062053,"rms_dbfs":-27.760112673975044,"peak":0.5439934134483337,"peak_dbfs":-5.288127172226127,"crest_factor_db":22.471985501748918,"mean":-6.42560392979874e-06,"zero_energy":false},"rms_difference_db":-0.5019603502582619},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":7.2531402288325e-07,"output_rms_dbfs":-61.394739260099975,"output_zero_energy":false,"output_peak":0.00920589454472065,"output_energy_fraction":0.00010393183211087439,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":133,"both_estimated_voiced_frames":119,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":3.7322380882096002,"absolute_difference_cents_p95":21.13009735830107,"difference_over_600_cents_frames":4,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019413100530357063,"log_rms_mae_db":1.1839760487026427,"rms_envelope_nrmse":0.11575393794714675,"envelope_correlation":0.9907529623279234,"rms_gain_error_db":-0.5019852508911162,"reference_mean_power":0.0018816420370153617,"output_mean_power":0.0016762488071796438},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1507655155627414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005726194534917747,"output_power":0.0004657380641460548,"reference_active":true,"delta_db":-0.8972436445243444},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.421513884012437e-05,"output_power":8.726424893098213e-05,"reference_active":true,"delta_db":-0.33284337321572877},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6796574139632317e-05,"output_power":1.9916797193005955e-05,"reference_active":true,"delta_db":0.7399878984256807}],"active_band_count":3,"active_band_mae_db":0.6566916387219179}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016416501002274435,"log_rms_mae_db":0.9391658521114362,"rms_envelope_nrmse":0.10028143377511518,"envelope_correlation":0.9935781922127287,"rms_gain_error_db":-0.5021206705939687,"reference_mean_power":0.001890759403997248,"output_mean_power":0.0016843184349433931},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0781635397104837e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005383866955860211,"output_power":0.0004402640435571565,"reference_active":true,"delta_db":-0.8738110122289554},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.887402542633377e-05,"output_power":3.7506549616240537e-05,"reference_active":true,"delta_db":-0.1555240095849681},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650199092025118e-06,"output_power":1.7676279075806995e-06,"reference_active":true,"delta_db":0.5287098291247669}],"active_band_count":3,"active_band_mae_db":0.5193482836462301}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017990629858894173,"log_rms_mae_db":1.5516983047010353,"rms_envelope_nrmse":0.12611957473171032,"envelope_correlation":0.9891171343204642,"rms_gain_error_db":-0.46408427982839,"reference_mean_power":0.0012781810161111924,"output_mean_power":0.001148639997434304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.355790711503392e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004472139003696596,"output_power":0.00037848497787623214,"reference_active":true,"delta_db":-0.7246664692744635},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.621773891011176e-05,"output_power":7.47520655141945e-05,"reference_active":true,"delta_db":-0.08432863380097169},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7010237353267383e-05,"output_power":1.9998337485337982e-05,"reference_active":true,"delta_db":0.7028351950984262}],"active_band_count":3,"active_band_mae_db":0.5039434327246205}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000149706333097475,"log_rms_mae_db":1.2138920991097755,"rms_envelope_nrmse":0.10781977546847941,"envelope_correlation":0.9922824111762737,"rms_gain_error_db":-0.4641805232345475,"reference_mean_power":0.0012844845607787134,"output_mean_power":0.0011542791107017909},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.840745806614025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236215690036061,"output_power":0.0003621940320487874,"reference_active":true,"delta_db":-0.6803677435049924},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.160014209802635e-05,"output_power":3.2479748814967124e-05,"reference_active":true,"delta_db":0.11923626389326773},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5445822829049618e-06,"output_power":1.6387755952444297e-06,"reference_active":true,"delta_db":0.2570843882064227}],"active_band_count":3,"active_band_mae_db":0.35222946520156095}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010005798255669509,"log_rms_mae_db":1.4411589683271646,"rms_envelope_nrmse":0.15372248116726114,"envelope_correlation":0.988055613026886,"rms_gain_error_db":-0.7900095797415779,"reference_mean_power":0.0003605696849433612,"output_mean_power":0.00030059949901328867},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5037729366163554e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014066740308353966,"output_power":0.00010832995077647575,"reference_active":true,"delta_db":-1.1344492404108326},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0544824834133837e-05,"output_power":1.660128228516141e-05,"reference_active":true,"delta_db":-0.9256080850753214},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.44585128264573e-06,"output_power":3.4214122432225274e-06,"reference_active":true,"delta_db":-0.030911249385130124}],"active_band_count":3,"active_band_mae_db":0.6969895249570947}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":1.1731392137173646,"rms_envelope_nrmse":0.13467468066532248,"envelope_correlation":0.9919165152722451,"rms_gain_error_db":-0.7900146559687725,"reference_mean_power":0.0003623811119711957,"output_mean_power":0.0003021092951935926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.381538826489277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013551929688065346,"output_power":0.00010430122189118489,"reference_active":true,"delta_db":-1.137117435016855},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.971075792205562e-06,"output_power":7.0441089335543e-06,"reference_active":true,"delta_db":-0.5369087510680183},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4163220581925866e-07,"output_power":2.8867062112991366e-07,"reference_active":true,"delta_db":0.7724776817035633}],"active_band_count":3,"active_band_mae_db":0.8155012892628122}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":1.807726723531455,"rms_envelope_nrmse":0.18389206703966818,"envelope_correlation":0.9764012641257876,"rms_gain_error_db":-0.23419114479207948,"reference_mean_power":6.230386432542119e-05,"output_mean_power":5.903313802503197e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.022113691992469e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1604215620852607e-05,"output_power":1.264212739393319e-05,"reference_active":true,"delta_db":0.37204372431604377},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3772446197384044e-06,"output_power":2.2656978361761693e-06,"reference_active":true,"delta_db":-0.20871883270039057},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.450999961693561e-07,"output_power":5.837768199068339e-07,"reference_active":true,"delta_db":1.1778925586865856}],"active_band_count":3,"active_band_mae_db":0.5862183719010067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":1.4733688868899573,"rms_envelope_nrmse":0.14371709539863597,"envelope_correlation":0.9850918484376531,"rms_gain_error_db":-0.23421045826834605,"reference_mean_power":6.261700596722032e-05,"output_mean_power":5.93295770247308e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8467488015072894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1232117533806206e-05,"output_power":1.215900262235415e-05,"reference_active":true,"delta_db":0.34436312823583975},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0675811682412763e-06,"output_power":1.056211711609178e-06,"reference_active":true,"delta_db":-0.046499256065376623},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5798002699765385e-08,"output_power":3.935386487799759e-08,"reference_active":true,"delta_db":1.8340130626417963}],"active_band_count":3,"active_band_mae_db":0.7416251489810043}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8522422313690186,"predicted_mos":3.8522422313690186,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"0f5cfb425b9b5325abc9185983f12acb7fec3d5046cb6d5e500f642468e3dbc4","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7664885470003355},"started_at":"2026-10-13T07:10:50.258494+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007422357000905322,"native_input_samples":264600,"native_input_sha256":"ef2e92b74fe046415b2c704be44c87ce0f4826e922c0db3e3be04b91dacf0c40","output_sha256":"fe68237f15eb7583a09d2f37e18cf787a2851661c11fe3ce20dfe68d20c56336","output_samples":264704,"length_delta_samples":104,"analysis_s":0.18686042100125633,"synthesis_s":4.71288791599909,"total_s":4.8997483370003465,"rtf":0.816624722833391,"output_peak":0.45308539271354675,"output_rms":0.04001978092639336,"finished_at":"2026-10-13T07:10:55.193874+00:00","evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"fe68237f15eb7583a09d2f37e18cf787a2851661c11fe3ce20dfe68d20c56336","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.94890347732749,"output_peak":0.453003330272306},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.768794769212823,"spectral_convergence":0.5245814441485441,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.47526807937476,"spectral_convergence":0.4696809210238835,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.363274141034026,"spectral_convergence":0.5324920931292744,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":2.018333835088251e-05,"delta_db":-4.309824634540477,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.00029235581341398526,"delta_db":0.6531768503488532,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0008653431679221512,"delta_db":-1.0178630998402785,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.0002878271685944602,"delta_db":-0.32739122091114725,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.00010457333535095689,"delta_db":-1.0405027439787113,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":3.399622747715365e-05,"delta_db":-0.3673628520730412,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.286020233615418,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6890485516619881,"pitch":null,"output_rms_dbfs":-27.94720087537877,"output_peak":0.453003330272306,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.535778996540536},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7170442775110792,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043360310532128375,"rms_dbfs":-27.258152323716782,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.07422203496477,"mean":1.2998580932617187e-06,"zero_energy":false},"output":{"rms":0.04005345242434902,"rms_dbfs":-27.94720087537877,"peak":0.453003330272306,"peak_dbfs":-6.877972104839637,"crest_factor_db":21.069228770539134,"mean":-1.8277273978760474e-05,"zero_energy":false},"rms_difference_db":-0.6890485516619869},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":5.654840369155435e-07,"output_rms_dbfs":-62.47579650291743,"output_zero_energy":false,"output_peak":0.013085485983045785,"output_energy_fraction":8.459636044356695e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":137,"both_estimated_voiced_frames":108,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.7941176470588235,"absolute_difference_cents_median":8.963502956616875,"absolute_difference_cents_p95":1901.4475862985548,"difference_over_600_cents_frames":10,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019413100530357063,"log_rms_mae_db":1.3844770548197292,"rms_envelope_nrmse":0.13087817248633646,"envelope_correlation":0.9896363292419456,"rms_gain_error_db":-0.6890320239434279,"reference_mean_power":0.0018816420370153617,"output_mean_power":0.0016055868572234535},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1507655155627414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005726194534917747,"output_power":0.00047476305542384206,"reference_active":true,"delta_db":-0.8138918198095075},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.421513884012437e-05,"output_power":7.629272415120228e-05,"reference_active":true,"delta_db":-0.9163757012133821},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6796574139632317e-05,"output_power":1.371950231255255e-05,"reference_active":true,"delta_db":-0.8788235402281709}],"active_band_count":3,"active_band_mae_db":0.8696970204170201}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016416501002274435,"log_rms_mae_db":1.1381833366186238,"rms_envelope_nrmse":0.10383059271045217,"envelope_correlation":0.9952801698447569,"rms_gain_error_db":-0.688948119542435,"reference_mean_power":0.001890759403997248,"output_mean_power":0.0016133977878956586},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0781635397104837e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005383866955860211,"output_power":0.00044503311543793005,"reference_active":true,"delta_db":-0.8270199067612858},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.887402542633377e-05,"output_power":3.0019231951063616e-05,"reference_active":true,"delta_db":-1.1225993786261386},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650199092025118e-06,"output_power":1.3308925154110506e-06,"reference_active":true,"delta_db":-0.7037688402909424}],"active_band_count":3,"active_band_mae_db":0.8844627085594556}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017990629858894173,"log_rms_mae_db":2.1602838159203173,"rms_envelope_nrmse":0.1828606952736256,"envelope_correlation":0.9770830686662139,"rms_gain_error_db":-0.696975913638125,"reference_mean_power":0.0012781810161111924,"output_mean_power":0.00108866628051978},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.355790711503392e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004472139003696596,"output_power":0.0003357835525531594,"reference_active":true,"delta_db":-1.2445587430910505},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.621773891011176e-05,"output_power":6.596643271670554e-05,"reference_active":true,"delta_db":-0.6273306150080358},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7010237353267383e-05,"output_power":1.1363290145420842e-05,"reference_active":true,"delta_db":-1.7520627769626937}],"active_band_count":3,"active_band_mae_db":1.2079840450205934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000149706333097475,"log_rms_mae_db":1.7472142492827958,"rms_envelope_nrmse":0.15562276422958488,"envelope_correlation":0.9840938365379235,"rms_gain_error_db":-0.6970734573243625,"reference_mean_power":0.0012844845607787134,"output_mean_power":0.0010940106323149197},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.840745806614025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236215690036061,"output_power":0.00031763277445619794,"reference_active":true,"delta_db":-1.2505275633648247},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.160014209802635e-05,"output_power":2.6928140615045497e-05,"reference_active":true,"delta_db":-0.6948266907400329},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5445822829049618e-06,"output_power":1.0934275348611838e-06,"reference_active":true,"delta_db":-1.5002104280219042}],"active_band_count":3,"active_band_mae_db":1.1485215607089205}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010005798255669509,"log_rms_mae_db":2.861830462724105,"rms_envelope_nrmse":0.2639894708345662,"envelope_correlation":0.9522091843215099,"rms_gain_error_db":-0.5450408796546653,"reference_mean_power":0.0003605696849433612,"output_mean_power":0.0003180424735239819},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5037729366163554e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014066740308353966,"output_power":0.00013169618552698258,"reference_active":true,"delta_db":-0.28620273693516624},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0544824834133837e-05,"output_power":1.656875478959871e-05,"reference_active":true,"delta_db":-0.9341257218384946},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.44585128264573e-06,"output_power":3.2599069240875125e-06,"reference_active":true,"delta_db":-0.2409132967532092}],"active_band_count":3,"active_band_mae_db":0.4870805851756234}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":2.5558277438085306,"rms_envelope_nrmse":0.23004678412204319,"envelope_correlation":0.9629960457772045,"rms_gain_error_db":-0.5450346157098267,"reference_mean_power":0.0003623811119711957,"output_mean_power":0.0003196407137216307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.381538826489277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013551929688065346,"output_power":0.00012686232744885231,"reference_active":true,"delta_db":-0.28668464915067404},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.971075792205562e-06,"output_power":6.5660437725454344e-06,"reference_active":true,"delta_db":-0.8421316496867405},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4163220581925866e-07,"output_power":2.8433182966704016e-07,"reference_active":true,"delta_db":0.706706612074287}],"active_band_count":3,"active_band_mae_db":0.6118409703039004}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":3.8180931281608785,"rms_envelope_nrmse":0.2765705921088069,"envelope_correlation":0.9547783585347761,"rms_gain_error_db":-1.3334235220381199,"reference_mean_power":6.230386432542119e-05,"output_mean_power":4.583240341547932e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.022113691992469e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1604215620852607e-05,"output_power":9.094039725719033e-06,"reference_active":true,"delta_db":-1.0585894292815416},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3772446197384044e-06,"output_power":2.4075580291532966e-06,"reference_active":true,"delta_db":0.055028905220545266},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.450999961693561e-07,"output_power":4.721794745201638e-07,"reference_active":true,"delta_db":0.2564951397594772}],"active_band_count":3,"active_band_mae_db":0.4567044914205214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":3.5447609738483674,"rms_envelope_nrmse":0.22926249657572428,"envelope_correlation":0.972707083209004,"rms_gain_error_db":-1.3334105978748692,"reference_mean_power":6.261700596722032e-05,"output_mean_power":4.606289593334369e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8467488015072894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1232117533806206e-05,"output_power":9.165841211901716e-06,"reference_active":true,"delta_db":-0.8828931000223205},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0675811682412763e-06,"output_power":1.1275809198623007e-06,"reference_active":true,"delta_db":0.23746814053344087},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5798002699765385e-08,"output_power":3.6318001074469395e-08,"reference_active":true,"delta_db":1.4853585327645753}],"active_band_count":3,"active_band_mae_db":0.8685732577734454}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"golf","status":"measured","mos":{"status":"measured","value":3.3547816276550293,"predicted_mos":3.3547816276550293,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"c9b14e6c32c1de4557740e62d54ac26309caf76243073ac6b414a7d6d655b5d6","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.927241845000026},"started_at":"2026-10-13T05:02:20.852319+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003425972000513866,"native_input_samples":144000,"native_input_sha256":"16bfd826568e50b29fd62127d3f4234760fe847e0d7ffea3faf50c6ed181a2e0","output_sha256":"ab32c5a89e665f66242591175e31087e934333da7d8b7447dcb76757856b93ef","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6467808959996546,"synthesis_s":0.22295787799976097,"total_s":0.8697387739994156,"rtf":0.14495646233323592,"output_peak":0.5529648065567017,"output_rms":0.03769992931505303,"finished_at":"2026-10-13T05:02:21.734624+00:00","analysis_stages":{"external_f0_s":0.09719019499971182},"evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"ab32c5a89e665f66242591175e31087e934333da7d8b7447dcb76757856b93ef","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.46695520028244,"output_peak":0.5611751231454606},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.393063449247026,"spectral_convergence":0.4350573711261162,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.599033636660637,"spectral_convergence":0.3224084142738913,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.778715631882527,"spectral_convergence":0.38011326789243205,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.440391682443362e-05,"output_power":5.5519503231755666e-05,"delta_db":0.08815403585435416,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025189803216352365,"output_power":0.00022071115027823745,"delta_db":-0.5740050061072224,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010957199913972687,"output_power":0.0007604127773853165,"delta_db":-1.586501796169377,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003108810664713635,"output_power":0.00026563201836391496,"delta_db":-0.6831385089234876,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013310514675148087,"output_power":9.935197854054755e-05,"delta_db":-1.2701832831170972,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.7058720482667074e-05,"output_power":2.1698890936818534e-05,"delta_db":-2.3245288349588855,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0877519108550706,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.2156126358865544,"pitch":null,"output_rms_dbfs":-28.46695520028244,"output_peak":0.5611751231454606,"reference_rms_dbfs":-27.251342564395884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.923604239263397},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.877035430110022,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04339431845404117,"rms_dbfs":-27.251342564395884,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.06741227564387,"mean":9.259715900198247e-07,"zero_energy":false},"output":{"rms":0.037726997213356255,"rms_dbfs":-28.46695520028244,"peak":0.5611751231454606,"peak_dbfs":-5.018031789228425,"crest_factor_db":23.448923411054015,"mean":0.000261362718625546,"zero_energy":false},"rms_difference_db":-1.2156126358865578},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":5.580329963705882e-07,"output_rms_dbfs":-62.53340120564228,"output_zero_energy":false,"output_peak":0.004751175376623876,"output_energy_fraction":9.425210733799805e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":145,"both_estimated_voiced_frames":113,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.8308823529411765,"absolute_difference_cents_median":5.543749281859325,"absolute_difference_cents_p95":40.034977471683355,"difference_over_600_cents_frames":4,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019413100530357063,"log_rms_mae_db":2.5038855015812325,"rms_envelope_nrmse":0.17265812801617259,"envelope_correlation":0.9882418817660703,"rms_gain_error_db":-1.215663899951365,"reference_mean_power":0.0018845854894181833,"output_mean_power":0.001424457357999844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.148960637371092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005707265401250985,"output_power":0.0004437015435925855,"reference_active":true,"delta_db":-1.0933712914681961},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.377940073139796e-05,"output_power":7.529258405641994e-05,"reference_active":true,"delta_db":-0.9535525071631944},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.674824080275267e-05,"output_power":1.4191949675875711e-05,"reference_active":true,"delta_db":-0.7192713396224382}],"active_band_count":3,"active_band_mae_db":0.9220650460846095}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016416501002274435,"log_rms_mae_db":2.2426275649933665,"rms_envelope_nrmse":0.151841331009842,"envelope_correlation":0.9945352619383442,"rms_gain_error_db":-1.2159625293846223,"reference_mean_power":0.0018936178584986407,"output_mean_power":0.0014311860280613103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.076492258678556e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005365047384510255,"output_power":0.0004225868531943577,"reference_active":true,"delta_db":-1.0365758014818292},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8646995819141884e-05,"output_power":3.031801315095961e-05,"reference_active":true,"delta_db":-1.0541500316751173},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5590455459720787e-06,"output_power":1.2606868457417915e-06,"reference_active":true,"delta_db":-0.9225158151537037}],"active_band_count":3,"active_band_mae_db":1.0044138827702167}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017990629858894173,"log_rms_mae_db":2.583997800617633,"rms_envelope_nrmse":0.19934920465285264,"envelope_correlation":0.9838087616527093,"rms_gain_error_db":-1.3774071779400816,"reference_mean_power":0.0012802494233928605,"output_mean_power":0.0009322961078073712},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.335490990013797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044534639296707484,"output_power":0.00031727522818834073,"reference_active":true,"delta_db":-1.4726177401477631},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.578548187823571e-05,"output_power":5.5924880538333405e-05,"reference_active":true,"delta_db":-1.3198095132726455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6944026611308588e-05,"output_power":1.2912824407814841e-05,"reference_active":true,"delta_db":-1.1799537943282703}],"active_band_count":3,"active_band_mae_db":1.3241270159162264}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.000149706333097475,"log_rms_mae_db":2.235004951277181,"rms_envelope_nrmse":0.1761113743028286,"envelope_correlation":0.9913051676526766,"rms_gain_error_db":-1.3774431781062133,"reference_mean_power":0.0012865249153156504,"output_mean_power":0.0009368582456259844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.821855482139166e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004218307413926247,"output_power":0.0003046707197258637,"reference_active":true,"delta_db":-1.413075077514209},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.129493091103924e-05,"output_power":2.25775597656909e-05,"reference_active":true,"delta_db":-1.4179699652604196},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5399437889680152e-06,"output_power":1.2203942172768835e-06,"reference_active":true,"delta_db":-1.0100472739580728}],"active_band_count":3,"active_band_mae_db":1.2803641055775672}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00010005798255669509,"log_rms_mae_db":2.008840055318135,"rms_envelope_nrmse":0.15496058457047573,"envelope_correlation":0.9882936831623823,"rms_gain_error_db":-0.7965208569903791,"reference_mean_power":0.00036117138362833497,"output_mean_power":0.0003006500270365709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.505273002538631e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014058363055342773,"output_power":0.00012455189467268837,"reference_active":true,"delta_db":-0.525844163492434},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.04506354842166e-05,"output_power":1.940255000180418e-05,"reference_active":true,"delta_db":-0.22847996525972536},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4471771597581703e-06,"output_power":4.267547519600244e-06,"reference_active":true,"delta_db":0.927147621798191}],"active_band_count":3,"active_band_mae_db":0.5604905835167835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":1.7185748137425927,"rms_envelope_nrmse":0.11996082171136574,"envelope_correlation":0.9952069929844963,"rms_gain_error_db":-0.7965631285206438,"reference_mean_power":0.00036298769020143284,"output_mean_power":0.0003021590347389717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.3830845778063543e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013538104239125022,"output_power":0.00012178108226515142,"reference_active":true,"delta_db":-0.4597802440554957},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.925164315408223e-06,"output_power":7.382592015895498e-06,"reference_active":true,"delta_db":-0.3079940669949124},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.390777362535073e-07,"output_power":2.664072761162004e-07,"reference_active":true,"delta_db":0.4700694711917505}],"active_band_count":3,"active_band_mae_db":0.41261459408071954}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":2.9206191891692392,"rms_envelope_nrmse":0.22442528886676127,"envelope_correlation":0.9802503776953168,"rms_gain_error_db":-1.5401762258727563,"reference_mean_power":6.240784430599746e-05,"output_mean_power":4.377453675381085e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.024628061036854e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1608242887940351e-05,"output_power":9.521873920516983e-06,"reference_active":true,"delta_db":-0.8604405992538091},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.386525700501394e-06,"output_power":2.3096030541539243e-06,"reference_active":true,"delta_db":-0.1422877047402098},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.454361746376869e-07,"output_power":4.571941202641715e-07,"reference_active":true,"delta_db":0.11315152591751415}],"active_band_count":3,"active_band_mae_db":0.37195994330384435}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":2.6423294229029604,"rms_envelope_nrmse":0.19470813065517692,"envelope_correlation":0.9900409659633311,"rms_gain_error_db":-1.5401773734146886,"reference_mean_power":6.272201994903646e-05,"output_mean_power":4.3994896356128704e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8488505624863064e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1236162174661177e-05,"output_power":9.551032093543184e-06,"reference_active":true,"delta_db":-0.7056769447687451},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0719281858609613e-06,"output_power":1.041001283315775e-06,"reference_active":true,"delta_db":-0.1271442575050573},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.58265498638088e-08,"output_power":3.000492288550132e-08,"reference_active":true,"delta_db":0.6512612174338429}],"active_band_count":3,"active_band_mae_db":0.49469413990254846}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.880856990814209,"predicted_mos":2.880856990814209,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"14db915999e19b9f8bcf99436518706aca1c96d25dc0fc5f99421d3794247d3a","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9553770250004163},"started_at":"2026-10-13T05:09:01.100395+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003392114998860052,"native_input_samples":144000,"native_input_sha256":"16bfd826568e50b29fd62127d3f4234760fe847e0d7ffea3faf50c6ed181a2e0","output_sha256":"d6ce454c06c67d2a1d8e21766569f348af6ed40603c8acba754c75f449ae67e0","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3432220550002967,"synthesis_s":3.107236694000676,"total_s":5.450458749000973,"rtf":0.9084097915001621,"output_peak":0.5343999862670898,"output_rms":0.04141995074207615,"finished_at":"2026-10-13T05:09:06.562589+00:00","analysis_stages":{"external_f0_s":1.589822030000505},"evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"d6ce454c06c67d2a1d8e21766569f348af6ed40603c8acba754c75f449ae67e0","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.65102724309514,"output_peak":0.5043581802555863},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.076020314847664,"spectral_convergence":0.5500771121972267,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.114981960868035,"spectral_convergence":0.31227672126403355,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.324129004266726,"spectral_convergence":0.3225352000213293,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.508777556723661e-05,"delta_db":0.050797718292987445,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.0002387871158166852,"delta_db":-0.22583159090484392,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0009889782268796608,"delta_db":-0.4378794878570666,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.0002846589506309505,"delta_db":-0.3754606353440316,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.0001247964717511017,"delta_db":-0.27268927259713255,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":2.657714406061243e-05,"delta_db":-1.436587012657191,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4665409529422089,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3893777293784826,"pitch":null,"output_rms_dbfs":-27.647530053095263,"output_peak":0.5043581802555863,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.505043759994142},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9053480921856224,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043360310532128375,"rms_dbfs":-27.258152323716782,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.07422203496477,"mean":1.2998580932617187e-06,"zero_energy":false},"output":{"rms":0.04145944626627625,"rms_dbfs":-27.647530053095263,"peak":0.5043581802555863,"peak_dbfs":-5.945218617918185,"crest_factor_db":21.70231143517708,"mean":-2.7459113450344343e-05,"zero_energy":false},"rms_difference_db":-0.3893777293784808},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":1.1303105159248862e-06,"output_rms_dbfs":-59.468022318905,"output_zero_energy":false,"output_peak":0.03661303701316515,"output_energy_fraction":0.0001578199913092724,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":140,"both_estimated_voiced_frames":118,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.8676470588235294,"absolute_difference_cents_median":5.47446027183558,"absolute_difference_cents_p95":16.705208254339606,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019413100530357063,"log_rms_mae_db":1.948890887266256,"rms_envelope_nrmse":0.15850180222274848,"envelope_correlation":0.9792737523780494,"rms_gain_error_db":-0.3893579506820092,"reference_mean_power":0.0018816420370153617,"output_mean_power":0.0017202882057405216},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1507655155627414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005726194534917747,"output_power":0.000520390585720504,"reference_active":true,"delta_db":-0.4153666691193395},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.421513884012437e-05,"output_power":8.771225054618263e-05,"reference_active":true,"delta_db":-0.31060438053973866},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6796574139632317e-05,"output_power":1.372802999348468e-05,"reference_active":true,"delta_db":-0.876124918836545}],"active_band_count":3,"active_band_mae_db":0.534031989498541}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016416501002274435,"log_rms_mae_db":1.4905069840782939,"rms_envelope_nrmse":0.09776129337678197,"envelope_correlation":0.992842443300578,"rms_gain_error_db":-0.38923535673729637,"reference_mean_power":0.001890759403997248,"output_mean_power":0.0017286725406350757},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0781635397104837e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005383866955860211,"output_power":0.0004885587023980877,"reference_active":true,"delta_db":-0.4217756571095091},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.887402542633377e-05,"output_power":3.64856448731883e-05,"reference_active":true,"delta_db":-0.27537487702557045},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650199092025118e-06,"output_power":1.0761784652152062e-06,"reference_active":true,"delta_db":-1.6263556935809584}],"active_band_count":3,"active_band_mae_db":0.774502075905346}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017990629858894173,"log_rms_mae_db":2.317929262615968,"rms_envelope_nrmse":0.17912706091786218,"envelope_correlation":0.975554504391009,"rms_gain_error_db":-0.39356286615094294,"reference_mean_power":0.0012781810161111924,"output_mean_power":0.0011674440508160895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.355790711503392e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004472139003696596,"output_power":0.00040279574640513855,"reference_active":true,"delta_db":-0.4543041821583601},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.621773891011176e-05,"output_power":7.042381849384834e-05,"reference_active":true,"delta_db":-0.34336491239883254},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7010237353267383e-05,"output_power":1.3089501032921828e-05,"reference_active":true,"delta_db":-1.1378728188978622}],"active_band_count":3,"active_band_mae_db":0.6451806378183517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000149706333097475,"log_rms_mae_db":1.935725519506483,"rms_envelope_nrmse":0.10795385370760524,"envelope_correlation":0.9917299344762718,"rms_gain_error_db":-0.39342351782110196,"reference_mean_power":0.0012844845607787134,"output_mean_power":0.0011732391233556419},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.840745806614025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236215690036061,"output_power":0.00038046434631738385,"reference_active":true,"delta_db":-0.4666409956460049},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.160014209802635e-05,"output_power":2.939860753624198e-05,"reference_active":true,"delta_db":-0.3136227497751764},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5445822829049618e-06,"output_power":1.0214910518456209e-06,"reference_active":true,"delta_db":-1.795764823551416}],"active_band_count":3,"active_band_mae_db":0.8586761896575325}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010005798255669509,"log_rms_mae_db":1.83671523536754,"rms_envelope_nrmse":0.1767615751556846,"envelope_correlation":0.9790216482085269,"rms_gain_error_db":-0.3792531461354046,"reference_mean_power":0.0003605696849433612,"output_mean_power":0.00033041815040649403},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5037729366163554e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014066740308353966,"output_power":0.00012543774795146877,"reference_active":true,"delta_db":-0.497652215529254},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0544824834133837e-05,"output_power":1.8929156400668463e-05,"reference_active":true,"delta_db":-0.35571183235601206},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.44585128264573e-06,"output_power":3.7323825114370743e-06,"reference_active":true,"delta_db":0.3468961574686492}],"active_band_count":3,"active_band_mae_db":0.40008673511797177}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":1.2400966786709922,"rms_envelope_nrmse":0.10708906306776272,"envelope_correlation":0.9928858854703523,"rms_gain_error_db":-0.3792524958595609,"reference_mean_power":0.0003623811119711957,"output_mean_power":0.0003320781521248287},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.381538826489277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013551929688065346,"output_power":0.00012116348846591049,"reference_active":true,"delta_db":-0.48629370942749134},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.971075792205562e-06,"output_power":7.553974691620882e-06,"reference_active":true,"delta_db":-0.23341413136751785},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4163220581925866e-07,"output_power":2.3057329601821846e-07,"reference_active":true,"delta_db":-0.20345810714180756}],"active_band_count":3,"active_band_mae_db":0.3077219826456056}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":1.8467152903963593,"rms_envelope_nrmse":0.20504101042129744,"envelope_correlation":0.9718253391659889,"rms_gain_error_db":-0.6242141488398218,"reference_mean_power":6.230386432542119e-05,"output_mean_power":5.396268731970834e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.022113691992469e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1604215620852607e-05,"output_power":1.0142004914691108e-05,"reference_active":true,"delta_db":-0.5849197327505078},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3772446197384044e-06,"output_power":1.9602224257868545e-06,"reference_active":true,"delta_db":-0.8376851975827508},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.450999961693561e-07,"output_power":3.2410792466492494e-07,"reference_active":true,"delta_db":-1.3776794057970392}],"active_band_count":3,"active_band_mae_db":0.9334281120434326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":1.2641730496225145,"rms_envelope_nrmse":0.12833115512291834,"envelope_correlation":0.9907579946121089,"rms_gain_error_db":-0.6242152163726429,"reference_mean_power":6.261700596722032e-05,"output_mean_power":5.423389255077559e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8467488015072894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1232117533806206e-05,"output_power":9.751799339264728e-06,"reference_active":true,"delta_db":-0.6137688298730994},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0675811682412763e-06,"output_power":8.856531283803021e-07,"reference_active":true,"delta_db":-0.8113724335870847},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5798002699765385e-08,"output_power":2.057767492566996e-08,"reference_active":true,"delta_db":-0.9818978166333482}],"active_band_count":3,"active_band_mae_db":0.8023463600311774}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8443403244018555,"predicted_mos":2.8443403244018555,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"64cd73c58389238dea41e1399412323803ff9aa8feb65c3149419e146b183a40","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9845636539994302},"started_at":"2026-10-13T05:10:28.426016+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028372479991958244,"native_input_samples":144000,"native_input_sha256":"16bfd826568e50b29fd62127d3f4234760fe847e0d7ffea3faf50c6ed181a2e0","output_sha256":"6400ba5d83ca57dcf315e561132e2f8f738bbf52b10e7dac62edb2cca8bca4ae","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7709335920008016,"synthesis_s":0.13539698999920802,"total_s":0.9063305820000096,"rtf":0.1510550970000016,"output_peak":0.6273297071456909,"output_rms":0.042152422374851575,"finished_at":"2026-10-13T05:10:29.343679+00:00","analysis_stages":{"external_f0_s":0.1523408320008457},"evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"6400ba5d83ca57dcf315e561132e2f8f738bbf52b10e7dac62edb2cca8bca4ae","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.498215706502535,"output_peak":0.6381723087383281},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.799734072824952,"spectral_convergence":0.19086693453658934,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.508322572552269,"spectral_convergence":0.21228591719082376,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.082358523611179,"spectral_convergence":0.2451866327583639,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.444718915891811e-05,"output_power":5.38131565898613e-05,"delta_db":-0.05086996396032514,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025153246625132694,"output_power":0.0002490409206619816,"delta_db":-0.04323335858827841,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010938925867664089,"output_power":0.0010396540202593848,"delta_db":-0.22085841871088033,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031036369064007244,"output_power":0.0002894364141278681,"delta_db":-0.30317738561848057,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001328835538454751,"output_power":0.00011937282141318417,"delta_db":-0.46565775725068814,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6997042780401114e-05,"output_power":3.0656693665673444e-05,"delta_db":-0.8164169751308308,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.31670230987658055,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23283397969708125,"pitch":null,"output_rms_dbfs":-27.490986303413862,"output_peak":0.6381723087383281,"reference_rms_dbfs":-27.258152323716782,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.796805056329466},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9377994774285282,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043360310532128375,"rms_dbfs":-27.258152323716782,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.07422203496477,"mean":1.2998580932617187e-06,"zero_energy":false},"output":{"rms":0.042213434197159956,"rms_dbfs":-27.490986303413862,"peak":0.6381723087383281,"peak_dbfs":-3.9012408885616674,"crest_factor_db":23.589745414852196,"mean":-4.986029405852787e-06,"zero_energy":false},"rms_difference_db":-0.2328339796970802},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":7.151654804874123e-07,"output_rms_dbfs":-61.45593456174717,"output_zero_energy":false,"output_peak":0.01061101092611835,"output_energy_fraction":9.631998712860346e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":122,"both_estimated_voiced_frames":108,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.7941176470588235,"absolute_difference_cents_median":0.9870465913088692,"absolute_difference_cents_p95":4.501873121545199,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019413100530357063,"log_rms_mae_db":0.8029770228238081,"rms_envelope_nrmse":0.06876227251846462,"envelope_correlation":0.9964602378611579,"rms_gain_error_db":-0.2328042292793987,"reference_mean_power":0.0018816420370153617,"output_mean_power":0.0017834321194344308},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1507655155627414e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005726194534917747,"output_power":0.000553877566900088,"reference_active":true,"delta_db":-0.14452322393596945},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.421513884012437e-05,"output_power":8.202038172777656e-05,"reference_active":true,"delta_db":-0.6019890626948068},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6796574139632317e-05,"output_power":1.4161441319815609e-05,"reference_active":true,"delta_db":-0.7411325412720577}],"active_band_count":3,"active_band_mae_db":0.49588160930094466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016416501002274435,"log_rms_mae_db":0.5551830424401141,"rms_envelope_nrmse":0.04199857877763238,"envelope_correlation":0.9990412742295105,"rms_gain_error_db":-0.23279247538659728,"reference_mean_power":0.001890759403997248,"output_mean_power":0.0017920784671554442},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0781635397104837e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005383866955860211,"output_power":0.0005188140202813742,"reference_active":true,"delta_db":-0.16082615402819272},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.887402542633377e-05,"output_power":3.550505146884507e-05,"reference_active":true,"delta_db":-0.39369367840716185},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650199092025118e-06,"output_power":1.128824709878976e-06,"reference_active":true,"delta_db":-1.4189335921942585}],"active_band_count":3,"active_band_mae_db":0.657817808209871}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017990629858894173,"log_rms_mae_db":1.028936805221524,"rms_envelope_nrmse":0.0789549143858414,"envelope_correlation":0.9954284282718029,"rms_gain_error_db":-0.1960793292250054,"reference_mean_power":0.0012781810161111924,"output_mean_power":0.0012217558581281687},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.355790711503392e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004472139003696596,"output_power":0.0004329796740010157,"reference_active":true,"delta_db":-0.14047784784551984},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.621773891011176e-05,"output_power":6.602656018547878e-05,"reference_active":true,"delta_db":-0.6233738853163859},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7010237353267383e-05,"output_power":1.42222340336031e-05,"reference_active":true,"delta_db":-0.777425527217756}],"active_band_count":3,"active_band_mae_db":0.5137590867932206}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000149706333097475,"log_rms_mae_db":0.5995755428430719,"rms_envelope_nrmse":0.043720133194066264,"envelope_correlation":0.9988199792810959,"rms_gain_error_db":-0.19605521063115608,"reference_mean_power":0.0012844845607787134,"output_mean_power":0.0012277879520369724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.840745806614025e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004236215690036061,"output_power":0.00040734614725107,"reference_active":true,"delta_db":-0.17014451385975748},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.160014209802635e-05,"output_power":2.8418662913553544e-05,"reference_active":true,"delta_db":-0.4608539484301555},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5445822829049618e-06,"output_power":1.086013784594609e-06,"reference_active":true,"delta_db":-1.529757112498551}],"active_band_count":3,"active_band_mae_db":0.7202518582628213}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010005798255669509,"log_rms_mae_db":0.98967611275695,"rms_envelope_nrmse":0.08805115615894295,"envelope_correlation":0.9956887932968094,"rms_gain_error_db":-0.3625405696316294,"reference_mean_power":0.0003605696849433612,"output_mean_power":0.00033169211949257975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5037729366163554e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014066740308353966,"output_power":0.00012992538109364408,"reference_active":true,"delta_db":-0.3449947049294577},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0544824834133837e-05,"output_power":1.81296128100623e-05,"reference_active":true,"delta_db":-0.543139137363591},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.44585128264573e-06,"output_power":3.0049634046195117e-06,"reference_active":true,"delta_db":-0.5945734267332522}],"active_band_count":3,"active_band_mae_db":0.49423575634210026}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":0.7211817957765287,"rms_envelope_nrmse":0.06000656175670834,"envelope_correlation":0.9985827670566108,"rms_gain_error_db":-0.3625426205498814,"reference_mean_power":0.0003623811119711957,"output_mean_power":0.00033335831421816826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.381538826489277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013551929688065346,"output_power":0.00012499998140094373,"reference_active":true,"delta_db":-0.35091191339432326},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.971075792205562e-06,"output_power":7.315527179794802e-06,"reference_active":true,"delta_db":-0.37271310387507006},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4163220581925866e-07,"output_power":2.175270165869755e-07,"reference_active":true,"delta_db":-0.45641615035516064}],"active_band_count":3,"active_band_mae_db":0.3933470558748513}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":1.1396590793099928,"rms_envelope_nrmse":0.09979107499799422,"envelope_correlation":0.9953376012366805,"rms_gain_error_db":-0.5474343268458877,"reference_mean_power":6.230386432542119e-05,"output_mean_power":5.492518791278542e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.022113691992469e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1604215620852607e-05,"output_power":1.0445638948434549e-05,"reference_active":true,"delta_db":-0.45680779488240236},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3772446197384044e-06,"output_power":2.1175200328740485e-06,"reference_active":true,"delta_db":-0.5024634548059832},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.450999961693561e-07,"output_power":3.272623738954238e-07,"reference_active":true,"delta_db":-1.3356151419784432}],"active_band_count":3,"active_band_mae_db":0.7649621305556096}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":0.9175511783168225,"rms_envelope_nrmse":0.07337641240221848,"envelope_correlation":0.9987033549843238,"rms_gain_error_db":-0.547433262722083,"reference_mean_power":6.261700596722032e-05,"output_mean_power":5.5201257565975515e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8467488015072894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1232117533806206e-05,"output_power":1.0129326092614883e-05,"reference_active":true,"delta_db":-0.4488108673570036},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0675811682412763e-06,"output_power":9.562885551007306e-07,"reference_active":true,"delta_db":-0.4781194622636932},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5798002699765385e-08,"output_power":1.9253804552842273e-08,"reference_active":true,"delta_db":-1.2706952493044308}],"active_band_count":3,"active_band_mae_db":0.7325418596417093}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_2961-961-0011","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2466020584106445,"predicted_mos":3.2466020584106445,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287401,"input_sample_rate":48000,"input_seconds":5.987520833333333,"input_float32_sha256":"714b628ce8ea2283139acae06cec56c5f4602274186112f274edd9c7de320af5","internal_samples":95801,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2489792450014647},"started_at":"2026-10-13T05:14:25.230017+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009816151001359685,"native_input_samples":288000,"native_input_sha256":"b953445d2ffd07798ca9b32a588e60e22bb0c4a4515b1dca4bdbe67dcd6743bf","output_sha256":"7c056aa8feeb534bb00b4be27b337a1cd2e0b14a503201454a48ce7c23884e29","output_samples":287401,"length_delta_samples":-599,"analysis_s":3.3864487640003063,"synthesis_s":1.0365518630005681,"total_s":4.423000627000874,"rtf":0.7371667711668124,"output_peak":0.6558340622522263,"output_rms":0.042470527899111266,"finished_at":"2026-10-13T05:14:29.673804+00:00","evaluation_input":{"reference":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"output_sha256":"7c056aa8feeb534bb00b4be27b337a1cd2e0b14a503201454a48ce7c23884e29","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95801,"length_delta_samples":-199,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.433485070336534,"output_peak":0.6301229458366028},"common_prefix":{"samples":95801,"sample_rate":16000,"reference_coverage":0.9979270833333334,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95801,"output_samples":95801,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6292619899002005,"spectral_convergence":0.5411697891268435,"spectral_convergence_status":"ok","amplitude_floor":1.4064891899964367e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.685144744915618,"spectral_convergence":0.3124547814782969,"spectral_convergence_status":"ok","amplitude_floor":1.2304331075199438e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.52988693361501,"spectral_convergence":0.24760838065505558,"spectral_convergence_status":"ok","amplitude_floor":7.896502575224385e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.4599715614677986e-05,"output_power":5.055753193965628e-05,"delta_db":-0.3340451577075451,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025183684832606057,"output_power":0.00023766734571224553,"delta_db":-0.25149759660202603,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010961384990206293,"output_power":0.001023509404429922,"delta_db":-0.29773593989242625,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003110038615813153,"output_power":0.0003113535764769517,"delta_db":0.004880772884263935,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013315875763597455,"output_power":0.00014466205625775328,"delta_db":0.35984899401216514,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.70736947007274e-05,"output_power":3.797459512580176e-05,"delta_db":0.10427283008414277,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22538021519709486,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1838592243240923,"pitch":null,"output_rms_dbfs":-27.433485070336534,"output_peak":0.6301229458366028,"reference_rms_dbfs":-27.249625846012442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.948097889476943},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95801,"output_samples":95801,"intelligibility":{"estoi":{"value":0.9035273984697065,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04340289595037854,"rms_dbfs":-27.249625846012442,"peak":0.61773681640625,"peak_dbfs":-4.183930288752015,"crest_factor_db":23.065695557260426,"mean":1.902709722660776e-06,"zero_energy":false},"output":{"rms":0.04249381731431446,"rms_dbfs":-27.433485070336534,"peak":0.6301229458366028,"peak_dbfs":-4.011494106514303,"crest_factor_db":23.421990963822232,"mean":-0.00011939159280331317,"zero_energy":false},"rms_difference_db":-0.18385922432409174},"full_scale":{"reference":{"sample_count":95801,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95801,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":121,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.7686847220174974e-06,"threshold_dbfs":-54.23810192933847,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":144,"quiet_samples":23040,"reference_mean_power":6.866943294880912e-07,"output_mean_power":6.547085344841868e-07,"output_rms_dbfs":-61.839519977747656,"output_zero_energy":false,"output_peak":0.007079924938113306,"output_energy_fraction":8.719848542393392e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":136,"output_estimated_voiced_frames":124,"both_estimated_voiced_frames":110,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.8088235294117647,"absolute_difference_cents_median":2.8353368015379363,"absolute_difference_cents_p95":14.452088353193014,"difference_over_600_cents_frames":3,"uncovered_tail_samples":121,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95801,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1196,"rms_floor":0.00019413100530357063,"log_rms_mae_db":1.5947449864532355,"rms_envelope_nrmse":0.17516043287710759,"envelope_correlation":0.9735816740885181,"rms_gain_error_db":-0.1837530051915094,"reference_mean_power":0.0018860362971144695,"output_mean_power":0.0018079014206833783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":1.148053787247503e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00056979470732408,"output_power":0.0005509421861424999,"reference_active":true,"delta_db":-0.14612382904268345},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.35510184513397e-05,"output_power":9.297994405578895e-05,"reference_active":true,"delta_db":-0.026592393059397323},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6697227469408008e-05,"output_power":1.539019942009687e-05,"reference_active":true,"delta_db":-0.3540011639703053}],"active_band_count":3,"active_band_mae_db":0.17557246202412868}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1190,"rms_floor":0.00016416501002274435,"log_rms_mae_db":1.0714789788965133,"rms_envelope_nrmse":0.10015931394433816,"envelope_correlation":0.9911874400148775,"rms_gain_error_db":-0.18418254489646113,"reference_mean_power":0.001895023613293585,"output_mean_power":0.0018163367555592378},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":1.07565251427859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005355741553058871,"output_power":0.0005191254826808951,"reference_active":true,"delta_db":-0.13547263585178226},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.852940284577273e-05,"output_power":3.806633400116783e-05,"reference_active":true,"delta_db":-0.05251224217526247},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5601955728833312e-06,"output_power":1.4708400087031663e-06,"reference_active":true,"delta_db":-0.2561360652420886}],"active_band_count":3,"active_band_mae_db":0.14804031442304444}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1196,"rms_floor":0.00017990629858894173,"log_rms_mae_db":2.1516673117345895,"rms_envelope_nrmse":0.19505005747193999,"envelope_correlation":0.9703889446782644,"rms_gain_error_db":-0.302711519539177,"reference_mean_power":0.0012812830259082187,"output_mean_power":0.0011950165523830802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":9.325291908470656e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004444331031979379,"output_power":0.0004194092211540788,"reference_active":true,"delta_db":-0.25168424848964044},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.555090356492566e-05,"output_power":7.095486254570124e-05,"reference_active":true,"delta_db":-0.27257499661732126},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6894032470249594e-05,"output_power":1.452978636783692e-05,"reference_active":true,"delta_db":-0.6547409566749341}],"active_band_count":3,"active_band_mae_db":0.39300006726063197}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1190,"rms_floor":0.000149706333097475,"log_rms_mae_db":1.4564333388972979,"rms_envelope_nrmse":0.1128505469504858,"envelope_correlation":0.990234815121855,"rms_gain_error_db":-0.30296358178935673,"reference_mean_power":0.0012875378970155734,"output_mean_power":0.00120078059960663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":8.812369353425445e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004209505695773079,"output_power":0.0003972428631255381,"reference_active":true,"delta_db":-0.2517499808365622},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1137231921545076e-05,"output_power":2.8634749660609242e-05,"reference_active":true,"delta_db":-0.36386610823007814},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5381917102501228e-06,"output_power":1.2769411895734881e-06,"reference_active":true,"delta_db":-0.808395705040198}],"active_band_count":3,"active_band_mae_db":0.47467059803561273}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1196,"rms_floor":0.00010005798255669509,"log_rms_mae_db":1.7874431959912214,"rms_envelope_nrmse":0.18830063844211523,"envelope_correlation":0.9761645569988379,"rms_gain_error_db":0.06287590836618619,"reference_mean_power":0.0003614729343359283,"output_mean_power":0.00036674430231308414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":2.5060231124791377e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014054084138237122,"output_power":0.00014111650312934946,"reference_active":true,"delta_db":0.017752568725302297},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0403656165167317e-05,"output_power":2.2239258097777324e-05,"reference_active":true,"delta_db":0.37412298725011583},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.436995805957306e-06,"output_power":3.902911240137248e-06,"reference_active":true,"delta_db":0.5520967250097472}],"active_band_count":3,"active_band_mae_db":0.3146574269950551}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1190,"rms_floor":9.182113809426457e-05,"log_rms_mae_db":1.1799652351486092,"rms_envelope_nrmse":0.10392425811861294,"envelope_correlation":0.9924989842382997,"rms_gain_error_db":0.06285911340962021,"reference_mean_power":0.00036329139519325666,"output_mean_power":0.00036858785642726903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":2.3838585201063594e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013531026562413223,"output_power":0.00013675829148796393,"reference_active":true,"delta_db":0.04622919950946784},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.894931764055532e-06,"output_power":8.631479000257304e-06,"reference_active":true,"delta_db":0.38736837515757083},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.46441478033426e-07,"output_power":3.304103204488627e-07,"reference_active":true,"delta_db":1.2733979934528863}],"active_band_count":3,"active_band_mae_db":0.5689985227066416}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1196,"rms_floor":4.283283642007834e-05,"log_rms_mae_db":1.7533492140121028,"rms_envelope_nrmse":0.18148145090810755,"envelope_correlation":0.9803865541140588,"rms_gain_error_db":0.4813477801553293,"reference_mean_power":6.24599445989733e-05,"output_mean_power":6.978086886584285e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1196,"frequency_resolution_hz":0.16722408026755853,"power_floor":2.0258972035813642e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1610047706661525e-05,"output_power":1.334107758047631e-05,"reference_active":true,"delta_db":0.6035690537228882},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3913832910183155e-06,"output_power":3.0790752603518993e-06,"reference_active":true,"delta_db":1.0977111385020024},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.449301158292651e-07,"output_power":7.007070115459455e-07,"reference_active":true,"delta_db":1.9724466055245882}],"active_band_count":3,"active_band_mae_db":1.2245755992498262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95760,"uncovered_tail_samples":41,"metrics":{"frame_count":1190,"rms_floor":3.7589606193577616e-05,"log_rms_mae_db":1.1031410511116295,"rms_envelope_nrmse":0.10302740634434365,"envelope_correlation":0.9949095259854384,"rms_gain_error_db":0.4813451917015338,"reference_mean_power":6.277463753286006e-05,"output_mean_power":7.013240512921528e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1190,"frequency_resolution_hz":0.1680672268907563,"power_floor":1.8499120883755288e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1238281952789991e-05,"output_power":1.3199357358525098e-05,"reference_active":true,"delta_db":0.6985286334433162},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0739864285086567e-06,"output_power":1.4299255170448705e-06,"reference_active":true,"delta_db":1.243146228101165},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5921830125729988e-08,"output_power":4.58291961655288e-08,"reference_active":true,"delta_db":2.474765796602683}],"active_band_count":3,"active_band_mae_db":1.4721468860490547}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"hnm","status":"measured","mos":{"status":"measured","value":3.1206791400909424,"predicted_mos":3.1206791400909424,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90800,"input_sample_rate":16000,"input_seconds":5.675,"input_float32_sha256":"58577f334782c85133cb8d90d9822a1f453bbc8e2776ec1f48a9cea7fd72ad27","internal_samples":90800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1384994039999583},"started_at":"2026-10-13T04:38:54.092552+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4813000234426e-05,"native_input_samples":90800,"native_input_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","output_sha256":"43830d5d24ca05b6a3789a61e13033ac9445220a4970100e79960d58f0d025f6","output_samples":90800,"length_delta_samples":0,"analysis_s":0.9634345170006782,"synthesis_s":0.4837173179994352,"total_s":1.4471518350001134,"rtf":0.25500472863438123,"output_peak":0.43161962546331406,"output_rms":0.07180773241370532,"finished_at":"2026-10-13T04:38:55.544724+00:00","evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"43830d5d24ca05b6a3789a61e13033ac9445220a4970100e79960d58f0d025f6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.900908734439058,"spectral_convergence":0.17576354827342652,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.441196937040845,"spectral_convergence":0.1721098905067382,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.410412825756551,"spectral_convergence":0.16587704147735788,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.001882201820723886,"delta_db":0.024169299554440436,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0019920712318958456,"delta_db":-0.04844631359364476,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0010110354966208419,"delta_db":-0.19286720740571978,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00016878548173827982,"delta_db":-0.31132909511232676,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":7.227228058799529e-05,"delta_db":-0.3719272117417727,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":2.9984122831457438e-05,"delta_db":-0.16303335673380218,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1852954140236178,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.065174510280683,"pitch":null,"output_rms_dbfs":-22.876575749195403,"output_peak":0.43161962546331406,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.584172832412151},"common_prefix":{"samples":90800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.900908734439058,"spectral_convergence":0.17576354827342652,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.441196937040845,"spectral_convergence":0.1721098905067382,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.410412825756551,"spectral_convergence":0.16587704147735788,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.001882201820723886,"delta_db":0.024169299554440436,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0019920712318958456,"delta_db":-0.04844631359364476,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0010110354966208419,"delta_db":-0.19286720740571978,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00016878548173827982,"delta_db":-0.31132909511232676,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":7.227228058799529e-05,"delta_db":-0.3719272117417727,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":2.9984122831457438e-05,"delta_db":-0.16303335673380218,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1852954140236178,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.065174510280683,"pitch":null,"output_rms_dbfs":-22.876575749195403,"output_peak":0.43161962546331406,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.584172832412151},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90800,"output_samples":90800,"intelligibility":{"estoi":{"value":0.9144050055909293,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07234856775571318,"rms_dbfs":-22.81140123891472,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.478772831760725,"mean":-3.846626449786619e-06,"zero_energy":false},"output":{"rms":0.07180773241370532,"rms_dbfs":-22.876575749195403,"peak":0.43161962546331406,"peak_dbfs":-7.297976329057503,"crest_factor_db":15.5785994201379,"mean":-9.286890382756491e-06,"zero_energy":false},"rms_difference_db":-0.06517451028068422},"full_scale":{"reference":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":1.343864364695181e-06,"output_rms_dbfs":-58.716445621132856,"output_zero_energy":false,"output_peak":0.02811297001595522,"output_energy_fraction":5.5109743920883964e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":244,"both_estimated_voiced_frames":231,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9705882352941176,"absolute_difference_cents_median":1.693773132786996,"absolute_difference_cents_p95":8.286961445523207,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.0001853404348361204,"log_rms_mae_db":1.0145786844510303,"rms_envelope_nrmse":0.06712911060172538,"envelope_correlation":0.995663900708884,"rms_gain_error_db":-0.0651746009110099,"reference_mean_power":0.005238930533969212,"output_mean_power":0.0051608968600718825},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":2.4623434734036186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012515931525511833,"output_power":0.0012220984326749506,"reference_active":true,"delta_db":-0.10356991339024664},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003238300007472361,"output_power":0.00029214213723152195,"reference_active":true,"delta_db":-0.44722878725339527},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.578860153055394e-05,"output_power":5.638825929675463e-05,"reference_active":true,"delta_db":-1.2841520529139594}],"active_band_count":3,"active_band_mae_db":0.6116502511858671}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00016148298564882675,"log_rms_mae_db":0.5403206998986346,"rms_envelope_nrmse":0.03345353286833444,"envelope_correlation":0.9988444546440659,"rms_gain_error_db":-0.06517467831281702,"reference_mean_power":0.005266794777733929,"output_mean_power":0.005188345974501441},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.1218089587296202e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001164760758537026,"output_power":0.0011338300339603612,"reference_active":true,"delta_db":-0.1168877377383039},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015381285066774078,"output_power":0.00013456660860687496,"reference_active":true,"delta_db":-0.5805531385624031},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.591871326154896e-06,"output_power":3.977564995469173e-06,"reference_active":true,"delta_db":-1.4793988438406325}],"active_band_count":3,"active_band_mae_db":0.7256132400471133}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.00014972999954250175,"log_rms_mae_db":1.6391478640982406,"rms_envelope_nrmse":0.08962577416057937,"envelope_correlation":0.9938567145651187,"rms_gain_error_db":-0.12484068507209242,"reference_mean_power":0.0025001891623953804,"output_mean_power":0.0024293427898613865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.4970716182960256e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007649864504896549,"output_power":0.0007313884546111663,"reference_active":true,"delta_db":-0.19505642490836206},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022996394368157333,"output_power":0.00020145292299271456,"reference_active":true,"delta_db":-0.5748617461959191},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80930068112851e-05,"output_power":3.0315714968916143e-05,"reference_active":true,"delta_db":-2.004141152060895}],"active_band_count":3,"active_band_mae_db":0.9246864410550587}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00013636056906360855,"log_rms_mae_db":1.0096871371267282,"rms_envelope_nrmse":0.04981755771920607,"envelope_correlation":0.9980458906010422,"rms_gain_error_db":-0.12484075140533082,"reference_mean_power":0.0025134871346063777,"output_mean_power":0.00244226390804429},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3161092414230216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215738253307795,"output_power":0.0006883689683661233,"reference_active":true,"delta_db":-0.20459486914546912},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011225926953226224,"output_power":9.374105975182737e-05,"reference_active":true,"delta_db":-0.7829235299809434},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.841934016741924e-06,"output_power":2.2316193097076804e-06,"reference_active":true,"delta_db":-2.3592979119484188}],"active_band_count":3,"active_band_mae_db":1.1156054370249437}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":1.868174556153501,"rms_envelope_nrmse":0.15763913043575348,"envelope_correlation":0.9809755595311668,"rms_gain_error_db":-0.3570253694702588,"reference_mean_power":0.00020292459461300388,"output_mean_power":0.00018690983327495568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.5147206216944186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.93765985961165e-05,"output_power":6.169924610542348e-05,"reference_active":true,"delta_db":-0.5093314590670761},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.392588995706175e-05,"output_power":2.0582889060518318e-05,"reference_active":true,"delta_db":-0.6536176768875136},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8985142545041605e-06,"output_power":4.470627437355137e-06,"reference_active":true,"delta_db":-1.203741541185555}],"active_band_count":3,"active_band_mae_db":0.7888968923800482}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":1.07633293603309,"rms_envelope_nrmse":0.09454815068544072,"envelope_correlation":0.9934028739311823,"rms_gain_error_db":-0.35702355733255114,"reference_mean_power":0.0002040033414802843,"output_mean_power":0.00018790352409750967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3444198629453335e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.55829624744988e-05,"output_power":5.813054001557589e-05,"reference_active":true,"delta_db":-0.523866730251666},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.319750377898722e-05,"output_power":1.1609491769380502e-05,"reference_active":true,"delta_db":-0.5567858705519207},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.991435827054759e-07,"output_power":3.91742553347641e-07,"reference_active":true,"delta_db":-1.0522474223495482}],"active_band_count":3,"active_band_mae_db":0.7109666743843782}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":1.7484474433950083,"rms_envelope_nrmse":0.1779853455385242,"envelope_correlation":0.979174451136099,"rms_gain_error_db":-0.304852865696719,"reference_mean_power":5.558522396873643e-05,"output_mean_power":5.1817215705065486e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":3.2309689753050144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093182032279317e-05,"output_power":1.1988046330565479e-05,"reference_active":true,"delta_db":-0.38296793133970375},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.882644346015486e-06,"output_power":6.191833956733188e-06,"reference_active":true,"delta_db":-0.4593602681014849},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3224614880429852e-06,"output_power":1.5674432022285223e-06,"reference_active":true,"delta_db":-3.2626814284506227}],"active_band_count":3,"active_band_mae_db":1.3683365426306038}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":0.9924312406670871,"rms_envelope_nrmse":0.10029759489060627,"envelope_correlation":0.9934673112435408,"rms_gain_error_db":-0.30485375400095516,"reference_mean_power":5.5880857927829866e-05,"output_mean_power":5.209279858971519e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.8916608161728035e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3816996130233868e-05,"output_power":1.2353312791983419e-05,"reference_active":true,"delta_db":-0.4863019784721523},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.814539997745628e-06,"output_power":3.1649171362308113e-06,"reference_active":true,"delta_db":-0.8107982927743295},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.118413559303931e-07,"output_power":1.4247960800156328e-07,"reference_active":true,"delta_db":-3.4018099855176036}],"active_band_count":3,"active_band_mae_db":1.5663034189213618}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.571845054626465,"predicted_mos":3.571845054626465,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90800,"input_sample_rate":16000,"input_seconds":5.675,"input_float32_sha256":"852ca4ff744ac72fd7e50c731cefbb2aaf4fc382d6b84c34f3926db13bbd3d44","internal_samples":90800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0888402149994363},"started_at":"2026-10-13T04:41:50.379005+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7533999855222646e-05,"native_input_samples":90800,"native_input_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","output_sha256":"deed3ad51ede1a5ee38e2f8b44d780a99bac126b86a6888408ea7e430512c52b","output_samples":90800,"length_delta_samples":0,"analysis_s":2.3641153179996763,"synthesis_s":0.10137394100001984,"total_s":2.465489258999696,"rtf":0.43444744651977024,"output_peak":0.5072865396849185,"output_rms":0.0797272945159709,"finished_at":"2026-10-13T04:41:52.850507+00:00","evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"deed3ad51ede1a5ee38e2f8b44d780a99bac126b86a6888408ea7e430512c52b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.380797400690229,"spectral_convergence":0.5390612256974197,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.284459313432032,"spectral_convergence":0.22853009154674545,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4925645949613475,"spectral_convergence":0.26420363791272783,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.0021374565195211815,"delta_db":0.5764802976761118,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0025176911599885396,"delta_db":0.9685295992686266,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0013719912666860523,"delta_db":1.1329822258890296,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.000212058197092671,"delta_db":0.6798706763048413,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":8.669273940304559e-05,"delta_db":0.41818245947262606,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":3.055160814487078e-05,"delta_db":-0.08160608856684222,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6429418911963461,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8435417768409454,"pitch":null,"output_rms_dbfs":-21.967859462073772,"output_peak":0.5072865396849185,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.719273769694536},"common_prefix":{"samples":90800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.380797400690229,"spectral_convergence":0.5390612256974197,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.284459313432032,"spectral_convergence":0.22853009154674545,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4925645949613475,"spectral_convergence":0.26420363791272783,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.0021374565195211815,"delta_db":0.5764802976761118,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0025176911599885396,"delta_db":0.9685295992686266,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0013719912666860523,"delta_db":1.1329822258890296,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.000212058197092671,"delta_db":0.6798706763048413,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":8.669273940304559e-05,"delta_db":0.41818245947262606,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":3.055160814487078e-05,"delta_db":-0.08160608856684222,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6429418911963461,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8435417768409454,"pitch":null,"output_rms_dbfs":-21.967859462073772,"output_peak":0.5072865396849185,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.719273769694536},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90800,"output_samples":90800,"intelligibility":{"estoi":{"value":0.9325445154410559,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07234856775571318,"rms_dbfs":-22.81140123891472,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.478772831760725,"mean":-3.846626449786619e-06,"zero_energy":false},"output":{"rms":0.07972729451597088,"rms_dbfs":-21.967859462073772,"peak":0.5072865396849185,"peak_dbfs":-5.894933221545438,"crest_factor_db":16.072926240528336,"mean":2.1574805841030445e-08,"zero_energy":false},"rms_difference_db":0.8435417768409472},"full_scale":{"reference":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":2.0585481241455477e-06,"output_rms_dbfs":-56.86438975736421,"output_zero_energy":false,"output_peak":0.06285607990382319,"output_energy_fraction":6.847977908676548e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":272,"both_estimated_voiced_frames":232,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":40,"paired_voiced_fraction":0.9747899159663865,"absolute_difference_cents_median":7.102114775610138,"absolute_difference_cents_p95":25.477955477976987,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.0001853404348361204,"log_rms_mae_db":1.8834028131119998,"rms_envelope_nrmse":0.1618898899804297,"envelope_correlation":0.9859773857944342,"rms_gain_error_db":0.8435421114829562,"reference_mean_power":0.005238930533969212,"output_mean_power":0.006362046676422263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":2.4623434734036186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012515931525511833,"output_power":0.0014274960015806448,"reference_active":true,"delta_db":0.5711172199826116},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003238300007472361,"output_power":0.0003528855362646621,"reference_active":true,"delta_db":0.37316777431825254},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.578860153055394e-05,"output_power":7.241678769175016e-05,"reference_active":true,"delta_db":-0.19764637226413861}],"active_band_count":3,"active_band_mae_db":0.38064378885500094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00016148298564882675,"log_rms_mae_db":1.2343878424376824,"rms_envelope_nrmse":0.128597771554086,"envelope_correlation":0.993925288279726,"rms_gain_error_db":0.8435423347436072,"reference_mean_power":0.005266794777733929,"output_mean_power":0.0063958847547919044},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.1218089587296202e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001164760758537026,"output_power":0.001332801476170662,"reference_active":true,"delta_db":0.5852873449103801},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015381285066774078,"output_power":0.00017309177138864575,"reference_active":true,"delta_db":0.5128380126938696},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.591871326154896e-06,"output_power":4.862877980292691e-06,"reference_active":true,"delta_db":-0.6066379700267062}],"active_band_count":3,"active_band_mae_db":0.568254442543652}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.00014972999954250175,"log_rms_mae_db":1.90238819242893,"rms_envelope_nrmse":0.2029566501916055,"envelope_correlation":0.9864992915974564,"rms_gain_error_db":1.1855676005115643,"reference_mean_power":0.0025001891623953804,"output_mean_power":0.0032849565410539355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.4970716182960256e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007649864504896549,"output_power":0.0009308559781475036,"reference_active":true,"delta_db":0.8522874926479},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022996394368157333,"output_power":0.00027631236674945033,"reference_active":true,"delta_db":0.7974057500285515},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80930068112851e-05,"output_power":4.80186606455253e-05,"reference_active":true,"delta_db":-0.006718879577168605}],"active_band_count":3,"active_band_mae_db":0.55213737408454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00013636056906360855,"log_rms_mae_db":1.3750956233672582,"rms_envelope_nrmse":0.18000300253101875,"envelope_correlation":0.9918341402350331,"rms_gain_error_db":1.1855678580134736,"reference_mean_power":0.0025134871346063777,"output_mean_power":0.00330242871916365},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3161092414230216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215738253307795,"output_power":0.0008818110933503467,"reference_active":true,"delta_db":0.8709478699055685},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011225926953226224,"output_power":0.0001342818629173371,"reference_active":true,"delta_db":0.7779514571460889},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.841934016741924e-06,"output_power":3.119813002072778e-06,"reference_active":true,"delta_db":-0.9042133810603775}],"active_band_count":3,"active_band_mae_db":0.8510375693706783}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":1.834487220177945,"rms_envelope_nrmse":0.16451678062868247,"envelope_correlation":0.9854210033912759,"rms_gain_error_db":0.7023098972777807,"reference_mean_power":0.00020292459461300388,"output_mean_power":0.00023854245072240103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.5147206216944186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.93765985961165e-05,"output_power":8.100453380534083e-05,"reference_active":true,"delta_db":0.6729632353432241},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.392588995706175e-05,"output_power":2.849253309531151e-05,"reference_active":true,"delta_db":0.7586296045636653},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8985142545041605e-06,"output_power":5.363637332994566e-06,"reference_active":true,"delta_db":-0.4128322830847287}],"active_band_count":3,"active_band_mae_db":0.6148083743305394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":1.1160131821337775,"rms_envelope_nrmse":0.10662510718304226,"envelope_correlation":0.9965252108157946,"rms_gain_error_db":0.7023016285360792,"reference_mean_power":0.0002040033414802843,"output_mean_power":0.00023981008547761915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3444198629453335e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.55829624744988e-05,"output_power":7.708016768162577e-05,"reference_active":true,"delta_db":0.701516202887388},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.319750377898722e-05,"output_power":1.529877854317559e-05,"reference_active":true,"delta_db":0.6416496303434459},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.991435827054759e-07,"output_power":4.019955844119565e-07,"reference_active":true,"delta_db":-0.9400420918407635}],"active_band_count":3,"active_band_mae_db":0.7610693083571992}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":1.6033550422855212,"rms_envelope_nrmse":0.2228572984793023,"envelope_correlation":0.9676436239057401,"rms_gain_error_db":0.09733771724989167,"reference_mean_power":5.558522396873643e-05,"output_mean_power":5.684511267091967e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":3.2309689753050144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093182032279317e-05,"output_power":1.304874664527098e-05,"reference_active":true,"delta_db":-0.014764069019192437},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.882644346015486e-06,"output_power":6.94513042883981e-06,"reference_active":true,"delta_db":0.03925077769649565},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3224614880429852e-06,"output_power":1.4944992527058392e-06,"reference_active":true,"delta_db":-3.4696425335073307}],"active_band_count":3,"active_band_mae_db":1.1745524600743396}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":0.9136875930100241,"rms_envelope_nrmse":0.1182119480764509,"envelope_correlation":0.9903902444281539,"rms_gain_error_db":0.09733729998688144,"reference_mean_power":5.5880857927829866e-05,"output_mean_power":5.7147441946869346e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.8916608161728035e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3816996130233868e-05,"output_power":1.3344587429205949e-05,"reference_active":true,"delta_db":-0.15108484584219725},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.814539997745628e-06,"output_power":3.4428902802307834e-06,"reference_active":true,"delta_db":-0.4451899048511298},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.118413559303931e-07,"output_power":1.2067323872042315e-07,"reference_active":true,"delta_db":-4.123227413356704}],"active_band_count":3,"active_band_mae_db":1.5731673880166772}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"facodec","status":"measured","mos":{"status":"measured","value":4.156701564788818,"predicted_mos":4.156701564788818,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90800,"input_sample_rate":16000,"input_seconds":5.675,"input_float32_sha256":"97971b41510aeac6d19af3ff62e3aebb277f774f29a6530bd070886bb05db512","internal_samples":90800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0966489920010645},"started_at":"2026-10-13T04:50:24.965801+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.0263000553532038e-05,"native_input_samples":90800,"native_input_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","output_sha256":"de42954074c91e9ec47f434edfab25376dbea7691f1503fcaabec15a1174fea3","output_samples":90800,"length_delta_samples":0,"analysis_s":2.0742680220000693,"synthesis_s":12.791312852999909,"total_s":14.865580874999978,"rtf":2.6194856167400844,"output_peak":0.4164831340312958,"output_rms":0.0681354684765723,"finished_at":"2026-10-13T04:50:39.839966+00:00","evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"de42954074c91e9ec47f434edfab25376dbea7691f1503fcaabec15a1174fea3","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.206106996527609,"spectral_convergence":0.17138990091654574,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.452072521921469,"spectral_convergence":0.17361357349276327,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.476350637637725,"spectral_convergence":0.18700499273485838,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.0015800985761322195,"delta_db":-0.73565077512061,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0019072778153625145,"delta_db":-0.2373553782031124,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0009143407762605401,"delta_db":-0.6294503617149095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00015311608313686915,"delta_db":-0.7344718622222648,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":5.958389012720164e-05,"delta_db":-1.2103562636202871,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":2.802492350263345e-05,"delta_db":-0.4565024818364008,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.667297853786264,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5211338345231795,"pitch":null,"output_rms_dbfs":-23.332535073437896,"output_peak":0.4164831340312958,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.3781767186956015},"common_prefix":{"samples":90800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.206106996527609,"spectral_convergence":0.17138990091654574,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.452072521921469,"spectral_convergence":0.17361357349276327,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.476350637637725,"spectral_convergence":0.18700499273485838,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.0015800985761322195,"delta_db":-0.73565077512061,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0019072778153625145,"delta_db":-0.2373553782031124,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0009143407762605401,"delta_db":-0.6294503617149095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00015311608313686915,"delta_db":-0.7344718622222648,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":5.958389012720164e-05,"delta_db":-1.2103562636202871,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":2.802492350263345e-05,"delta_db":-0.4565024818364008,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.667297853786264,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5211338345231795,"pitch":null,"output_rms_dbfs":-23.332535073437896,"output_peak":0.4164831340312958,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.3781767186956015},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90800,"output_samples":90800,"intelligibility":{"estoi":{"value":0.9211498597241762,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07234856775571318,"rms_dbfs":-22.81140123891472,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.478772831760725,"mean":-3.846626449786619e-06,"zero_energy":false},"output":{"rms":0.06813546847657231,"rms_dbfs":-23.332535073437896,"peak":0.4164831340312958,"peak_dbfs":-7.608051623220764,"crest_factor_db":15.724483450217132,"mean":2.0214981402813762e-05,"zero_energy":false},"rms_difference_db":-0.5211338345231766},"full_scale":{"reference":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":8.442744045168822e-07,"output_rms_dbfs":-60.735163768354816,"output_zero_energy":false,"output_peak":0.06476248055696487,"output_energy_fraction":3.845497299348761e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":237,"both_estimated_voiced_frames":229,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9621848739495799,"absolute_difference_cents_median":3.3735828312171208,"absolute_difference_cents_p95":16.583501657904574,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.0001853404348361204,"log_rms_mae_db":1.2338061479382074,"rms_envelope_nrmse":0.09168929915936111,"envelope_correlation":0.9948247715861399,"rms_gain_error_db":-0.5211337304263839,"reference_mean_power":0.005238930533969212,"output_mean_power":0.004646535578399612},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":2.4623434734036186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012515931525511833,"output_power":0.0012136852810459419,"reference_active":true,"delta_db":-0.1335709339019357},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003238300007472361,"output_power":0.0003028247440178273,"reference_active":true,"delta_db":-0.2912572206713906},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.578860153055394e-05,"output_power":7.375625899381972e-05,"reference_active":true,"delta_db":-0.1180501276823655}],"active_band_count":3,"active_band_mae_db":0.18095942741856394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00016148298564882675,"log_rms_mae_db":1.0153408507076978,"rms_envelope_nrmse":0.08403445679347041,"envelope_correlation":0.9957481855794212,"rms_gain_error_db":-0.521133823738059,"reference_mean_power":0.005266794777733929,"output_mean_power":0.004671248957088314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.1218089587296202e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001164760758537026,"output_power":0.0011367821205864111,"reference_active":true,"delta_db":-0.10559496233758017},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015381285066774078,"output_power":0.00014014486835879376,"reference_active":true,"delta_db":-0.4041542120624703},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.591871326154896e-06,"output_power":5.40396511087852e-06,"reference_active":true,"delta_db":-0.1484463289346976}],"active_band_count":3,"active_band_mae_db":0.21939850111158274}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.00014972999954250175,"log_rms_mae_db":1.579931272841281,"rms_envelope_nrmse":0.10037724781311581,"envelope_correlation":0.9925786311418119,"rms_gain_error_db":-0.2658296577894129,"reference_mean_power":0.0025001891623953804,"output_mean_power":0.0023517432301911096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.4970716182960256e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007649864504896549,"output_power":0.0007924877105499865,"reference_active":true,"delta_db":0.1533879319151474},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022996394368157333,"output_power":0.00022286606278595426,"reference_active":true,"delta_db":-0.13615807041701403},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80930068112851e-05,"output_power":4.6571947037494465e-05,"reference_active":true,"delta_db":-0.13957535443284752}],"active_band_count":3,"active_band_mae_db":0.14304045225500298}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00013636056906360855,"log_rms_mae_db":1.2484375115136326,"rms_envelope_nrmse":0.09194788178671733,"envelope_correlation":0.993263263524938,"rms_gain_error_db":-0.2658291690592349,"reference_mean_power":0.0025134871346063777,"output_mean_power":0.002364251916250404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3161092414230216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215738253307795,"output_power":0.0007570204828471358,"reference_active":true,"delta_db":0.20826859406574735},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011225926953226224,"output_power":0.00010742138888997498,"reference_active":true,"delta_db":-0.191314487253767},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.841934016741924e-06,"output_power":3.821820981351596e-06,"reference_active":true,"delta_db":-0.022795614789719306}],"active_band_count":3,"active_band_mae_db":0.1407928987030779}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":1.6302689773239836,"rms_envelope_nrmse":0.15632400324451168,"envelope_correlation":0.9840530721658126,"rms_gain_error_db":-0.6884463820250837,"reference_mean_power":0.00020292459461300388,"output_mean_power":0.00017317693495752517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.5147206216944186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.93765985961165e-05,"output_power":6.334694128194502e-05,"reference_active":true,"delta_db":-0.394873536467335},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.392588995706175e-05,"output_power":2.1165087135267803e-05,"reference_active":true,"delta_db":-0.5324804016675367},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8985142545041605e-06,"output_power":4.802097846906067e-06,"reference_active":true,"delta_db":-0.8931162834379404}],"active_band_count":3,"active_band_mae_db":0.6068234071909373}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":1.332971849464021,"rms_envelope_nrmse":0.13757377166404242,"envelope_correlation":0.9874656358541851,"rms_gain_error_db":-0.6884532066038385,"reference_mean_power":0.0002040033414802843,"output_mean_power":0.000174097269727102},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3444198629453335e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.55829624744988e-05,"output_power":6.015144600444155e-05,"reference_active":true,"delta_db":-0.3754495848950718},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.319750377898722e-05,"output_power":1.105943001719649e-05,"reference_active":true,"delta_db":-0.767590502275052},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.991435827054759e-07,"output_power":3.9397408272595225e-07,"reference_active":true,"delta_db":-1.0275783888453414}],"active_band_count":3,"active_band_mae_db":0.7235394920051551}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":1.6543638514661123,"rms_envelope_nrmse":0.31504331543881414,"envelope_correlation":0.9392686185397097,"rms_gain_error_db":-1.3562707043877547,"reference_mean_power":5.558522396873643e-05,"output_mean_power":4.067544280165835e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":3.2309689753050144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093182032279317e-05,"output_power":8.658951643922541e-06,"reference_active":true,"delta_db":-1.7957989151735918},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.882644346015486e-06,"output_power":7.83445630209031e-06,"reference_active":true,"delta_db":0.5625353416722769},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3224614880429852e-06,"output_power":2.822206580032683e-06,"reference_active":true,"delta_db":-0.7087115547264756}],"active_band_count":3,"active_band_mae_db":1.022348603857448}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":1.366725914799277,"rms_envelope_nrmse":0.2739979564950971,"envelope_correlation":0.9553625804534114,"rms_gain_error_db":-1.3562749152093267,"reference_mean_power":5.5880857927829866e-05,"output_mean_power":4.089173838359609e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.8916608161728035e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3816996130233868e-05,"output_power":9.793340623770484e-06,"reference_active":true,"delta_db":-1.4948277592130146},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.814539997745628e-06,"output_power":4.315176466187563e-06,"reference_active":true,"delta_db":0.5355638749586076},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.118413559303931e-07,"output_power":2.7281038474809133e-07,"reference_active":true,"delta_db":-0.5807281208154234}],"active_band_count":3,"active_band_mae_db":0.8703732516623486}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8228254318237305,"predicted_mos":3.8228254318237305,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":250368,"input_sample_rate":44100,"input_seconds":5.677278911564626,"input_float32_sha256":"8e8f8e093546c5fdd057ab6ce6c8dfba30151b180edc09996ff1554392f36aa8","internal_samples":90837,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.102223745001538},"started_at":"2026-10-13T07:10:55.196581+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008123713998429594,"native_input_samples":250268,"native_input_sha256":"abdc0081ef9d9354eb11eb989be15ceefa8e8ece2c6eabeadc29208f542f3d4d","output_sha256":"f01d5385b8b74692a2186c3c9e5409923dd407bd267da0e01b579af3588222b1","output_samples":250368,"length_delta_samples":100,"analysis_s":0.1543589040011284,"synthesis_s":35.91863526299858,"total_s":36.072994166999706,"rtf":6.356462043747851,"output_peak":0.4458174705505371,"output_rms":0.0668589903682926,"finished_at":"2026-10-13T07:11:31.302161+00:00","evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"f01d5385b8b74692a2186c3c9e5409923dd407bd267da0e01b579af3588222b1","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":90800,"output_samples":90837,"length_delta_samples":37,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.492057745722924,"output_peak":0.4414002686969108},"common_prefix":{"samples":90800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9995926769928554,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.100672816361639,"spectral_convergence":0.4680799848039071,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.719464893771585,"spectral_convergence":0.2988751376594845,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.354489700268436,"spectral_convergence":0.3202591064159719,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.0010734827167371195,"delta_db":-2.414542026252997,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0024010420267862598,"delta_db":0.7625026672488368,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0007993347126369532,"delta_db":-1.2132445087267443,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00013714611739966794,"delta_db":-1.2128447985063233,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":4.97994274753333e-05,"delta_db":-1.9894013061469007,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":1.6030592622881402e-05,"delta_db":-2.8824510623787116,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.745831061543419,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6788873036883198,"pitch":null,"output_rms_dbfs":-23.49028854260304,"output_peak":0.4414002686969108,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.724875803467221},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90800,"output_samples":90800,"intelligibility":{"estoi":{"value":0.8246866205984308,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07234856775571318,"rms_dbfs":-22.81140123891472,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.478772831760725,"mean":-3.846626449786619e-06,"zero_energy":false},"output":{"rms":0.06690915926581514,"rms_dbfs":-23.49028854260304,"peak":0.4414002686969108,"peak_dbfs":-7.103348136058632,"crest_factor_db":16.386940406544404,"mean":-0.00019325697418854686,"zero_energy":false},"rms_difference_db":-0.6788873036883203},"full_scale":{"reference":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":8.50139881286047e-07,"output_rms_dbfs":-60.70509609965086,"output_zero_energy":false,"output_peak":0.03493742120222312,"output_energy_fraction":4.0154537168118006e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":251,"both_estimated_voiced_frames":228,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.957983193277311,"absolute_difference_cents_median":6.540536297134624,"absolute_difference_cents_p95":35.95050258868193,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.0001853404348361204,"log_rms_mae_db":1.9216894865250718,"rms_envelope_nrmse":0.12592974373967208,"envelope_correlation":0.9896534789041461,"rms_gain_error_db":-0.6788870892393439,"reference_mean_power":0.005238930533969212,"output_mean_power":0.004480783196436913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":2.4623434734036186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012515931525511833,"output_power":0.0010613782073073585,"reference_active":true,"delta_db":-0.7159301220542537},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003238300007472361,"output_power":0.00028931632351317303,"reference_active":true,"delta_db":-0.48944143181929656},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.578860153055394e-05,"output_power":5.5713963305273975e-05,"reference_active":true,"delta_db":-1.336398396165595}],"active_band_count":3,"active_band_mae_db":0.8472566500130484}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00016148298564882675,"log_rms_mae_db":1.5363691418500973,"rms_envelope_nrmse":0.09710514273223955,"envelope_correlation":0.9959191956306734,"rms_gain_error_db":-0.6788869245933917,"reference_mean_power":0.005266794777733929,"output_mean_power":0.0045046152608461424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.1218089587296202e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001164760758537026,"output_power":0.0009948699439245316,"reference_active":true,"delta_db":-0.6847042001620725},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015381285066774078,"output_power":0.00013567809298641202,"reference_active":true,"delta_db":-0.5448289040629043},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.591871326154896e-06,"output_power":3.5900500931479956e-06,"reference_active":true,"delta_db":-1.9245666088297575}],"active_band_count":3,"active_band_mae_db":1.0513665710182447}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.00014972999954250175,"log_rms_mae_db":2.464354159863601,"rms_envelope_nrmse":0.1403530383612368,"envelope_correlation":0.9852516410267355,"rms_gain_error_db":0.191996788371116,"reference_mean_power":0.0025001891623953804,"output_mean_power":0.002613199383980002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.4970716182960256e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007649864504896549,"output_power":0.0007193080462809335,"reference_active":true,"delta_db":-0.26738824550856427},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022996394368157333,"output_power":0.00022360374097065176,"reference_active":true,"delta_db":-0.12180682632986514},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80930068112851e-05,"output_power":4.0950933742373324e-05,"reference_active":true,"delta_db":-0.6981812157191216}],"active_band_count":3,"active_band_mae_db":0.3624587625191837}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00013636056906360855,"log_rms_mae_db":1.5646385622848358,"rms_envelope_nrmse":0.10352282294801339,"envelope_correlation":0.9913582763814481,"rms_gain_error_db":0.19199612242059824,"reference_mean_power":0.0025134871346063777,"output_mean_power":0.002627098030583717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3161092414230216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215738253307795,"output_power":0.0006864353658129827,"reference_active":true,"delta_db":-0.2168112035264015},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011225926953226224,"output_power":0.00010780218423892515,"reference_active":true,"delta_db":-0.17594651534002376},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.841934016741924e-06,"output_power":2.941658260692048e-06,"reference_active":true,"delta_db":-1.15957683485398}],"active_band_count":3,"active_band_mae_db":0.5174448512401352}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":3.5615863367518386,"rms_envelope_nrmse":0.2982444443727247,"envelope_correlation":0.9432693795941092,"rms_gain_error_db":-1.5355858365931339,"reference_mean_power":0.00020292459461300388,"output_mean_power":0.00014248728238310868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.5147206216944186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.93765985961165e-05,"output_power":5.255658348767594e-05,"reference_active":true,"delta_db":-1.2058587782326893},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.392588995706175e-05,"output_power":1.880839068551767e-05,"reference_active":true,"delta_db":-1.0451646362145524},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8985142545041605e-06,"output_power":3.8869497686065325e-06,"reference_active":true,"delta_db":-1.8113370503969828}],"active_band_count":3,"active_band_mae_db":1.3541201549480748}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":3.186546340724405,"rms_envelope_nrmse":0.26644517339630047,"envelope_correlation":0.9555632913625346,"rms_gain_error_db":-1.5355774205636297,"reference_mean_power":0.0002040033414802843,"output_mean_power":0.0001432450221701521},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3444198629453335e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.55829624744988e-05,"output_power":5.166064713763164e-05,"reference_active":true,"delta_db":-1.0363118831820435},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.319750377898722e-05,"output_power":9.826039800422239e-06,"reference_active":true,"delta_db":-1.2811327611506877},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.991435827054759e-07,"output_power":3.2803098462921385e-07,"reference_active":true,"delta_db":-1.823106243832206}],"active_band_count":3,"active_band_mae_db":1.3801836293883125}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":2.951279313224607,"rms_envelope_nrmse":0.2901709535133587,"envelope_correlation":0.9596792250439793,"rms_gain_error_db":-1.7657691030705889,"reference_mean_power":5.558522396873643e-05,"output_mean_power":3.7015400232186786e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":3.2309689753050144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093182032279317e-05,"output_power":8.889657883648104e-06,"reference_active":true,"delta_db":-1.6816015826902349},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.882644346015486e-06,"output_power":5.521437795340845e-06,"reference_active":true,"delta_db":-0.9570314462937586},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3224614880429852e-06,"output_power":1.646209942510304e-06,"reference_active":true,"delta_db":-3.049747352245478}],"active_band_count":3,"active_band_mae_db":1.8961267937431572}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":2.4451145928992735,"rms_envelope_nrmse":0.23378627534704863,"envelope_correlation":0.9820787120136261,"rms_gain_error_db":-1.7657679148288195,"reference_mean_power":5.5880857927829866e-05,"output_mean_power":3.72122794470408e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.8916608161728035e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3816996130233868e-05,"output_power":9.36784247056896e-06,"reference_active":true,"delta_db":-1.687740569176346},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.814539997745628e-06,"output_power":2.8788300025506875e-06,"reference_active":true,"delta_db":-1.2222615293079393},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.118413559303931e-07,"output_power":1.708041235096516e-07,"reference_active":true,"delta_db":-2.6143535904525557}],"active_band_count":3,"active_band_mae_db":1.8414518963122803}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"golf","status":"measured","mos":{"status":"measured","value":3.3419225215911865,"predicted_mos":3.3419225215911865,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":136080,"input_sample_rate":24000,"input_seconds":5.67,"input_float32_sha256":"d54329639f66a1b02bb06100bbf0ba796b26ed53bceb5c6cc7fa338f6d4ed72a","internal_samples":90720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0413841169993248},"started_at":"2026-10-13T05:02:21.735420+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002712458999667433,"native_input_samples":136200,"native_input_sha256":"f66c8d856c15de3167e353f6999f8bfcee9d4e0d6e6d6e6bb78d9a57e187ffcd","output_sha256":"196a055ccf5f482a60a6a001c10daf88f6774648414d8a2dcab439ed9aed9e86","output_samples":136080,"length_delta_samples":-120,"analysis_s":0.6279934969998067,"synthesis_s":0.14969689999998081,"total_s":0.7776903969997875,"rtf":0.13703795541846475,"output_peak":0.4085502624511719,"output_rms":0.06612262140481118,"finished_at":"2026-10-13T05:02:22.526555+00:00","analysis_stages":{"external_f0_s":0.08549769399996876},"evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"196a055ccf5f482a60a6a001c10daf88f6774648414d8a2dcab439ed9aed9e86","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":90800,"output_samples":90720,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.588797639653734,"output_peak":0.40803887947855516},"common_prefix":{"samples":90720,"sample_rate":16000,"reference_coverage":0.9991189427312775,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90720,"output_samples":90720,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.3620239725470675,"spectral_convergence":0.5328855694499105,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.305138983958524,"spectral_convergence":0.20999362591342874,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.26148253782221,"spectral_convergence":0.21674701362901289,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018734064501982372,"output_power":0.001452712731022479,"delta_db":-1.1045226864769746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0020153703455059882,"output_power":0.0018184311726815558,"delta_db":-0.44657996263516075,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010587019389177344,"output_power":0.0008717912627478853,"delta_db":-0.8436119632735501,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018148907743805736,"output_power":0.00015617187828417784,"delta_db":-0.6524765939919039,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.880377974504834e-05,"output_power":5.6562941554542224e-05,"delta_db":-1.4401506164152245,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1158568726583855e-05,"output_power":2.0762529123961923e-05,"delta_db":-1.7629724542576701,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0417190461750807,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7812237293099673,"pitch":null,"output_rms_dbfs":-23.588797639653734,"output_peak":0.40803887947855516,"reference_rms_dbfs":-22.80757391034377,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.642881831442601},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90720,"output_samples":90720,"intelligibility":{"estoi":{"value":0.9041106071824511,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0723804542713822,"rms_dbfs":-22.80757391034377,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.474945503189776,"mean":-3.928061725811563e-06,"zero_energy":false},"output":{"rms":0.06615461068901096,"rms_dbfs":-23.588797639653734,"peak":0.40803887947855516,"peak_dbfs":-7.785969074548956,"crest_factor_db":15.80282856510478,"mean":0.0005229331753315324,"zero_energy":false},"rms_difference_db":-0.7812237293099642},"full_scale":{"reference":{"sample_count":90720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90720,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":4.785187090367579e-07,"output_rms_dbfs":-63.20101077589156,"output_zero_energy":false,"output_peak":0.007335752452593528,"output_energy_fraction":2.3140720323604965e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":255,"both_estimated_voiced_frames":234,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.9831932773109243,"absolute_difference_cents_median":3.6090479111699736,"absolute_difference_cents_p95":16.61480454630448,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90720,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1133,"rms_floor":0.0001853404348361204,"log_rms_mae_db":2.396458319909924,"rms_envelope_nrmse":0.14587805209330618,"envelope_correlation":0.9858479909203843,"rms_gain_error_db":-0.7812236165774488,"reference_mean_power":0.005243553852072456,"output_mean_power":0.004380295111222797},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1133,"frequency_resolution_hz":0.176522506619594,"power_floor":2.4624363698758623e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001251516393899419,"output_power":0.0010927054445780365,"reference_active":true,"delta_db":-0.5893343588108477},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003233995905153785,"output_power":0.0002686659815350601,"reference_active":true,"delta_db":-0.8052678609035268},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.588060995646397e-05,"output_power":5.6212339185460734e-05,"reference_active":true,"delta_db":-1.302991551201382}],"active_band_count":3,"active_band_mae_db":0.8991979236385855}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1127,"rms_floor":0.00016148298564882675,"log_rms_mae_db":1.8581234116198968,"rms_envelope_nrmse":0.11206477837850548,"envelope_correlation":0.9944152276502513,"rms_gain_error_db":-0.7812233927982601,"reference_mean_power":0.005271467417088359,"output_mean_power":0.004403613426551545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1127,"frequency_resolution_hz":0.1774622892635315,"power_floor":2.121817138129975e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011647338328131692,"output_power":0.0010186307121480811,"reference_active":true,"delta_db":-0.5820992470025583},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015323577776394463,"output_power":0.00012488317912117848,"reference_active":true,"delta_db":-0.8885623106401003},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.585081688657495e-06,"output_power":3.877394540434571e-06,"reference_active":true,"delta_db":-1.5848953512677832}],"active_band_count":3,"active_band_mae_db":1.018518969636814}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1133,"rms_floor":0.00014972999954250175,"log_rms_mae_db":2.3117606275237947,"rms_envelope_nrmse":0.1493637726599649,"envelope_correlation":0.983828451339981,"rms_gain_error_db":-0.4696202447624745,"reference_mean_power":0.00250239575023584,"output_mean_power":0.0022459183802259787},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1133,"frequency_resolution_hz":0.176522506619594,"power_floor":1.49702721958374e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007646791411364983,"output_power":0.0006666533019269755,"reference_active":true,"delta_db":-0.5957920918835822},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022984253084633319,"output_power":0.0001987823428244509,"reference_active":true,"delta_db":-0.6305259028409521},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8145101089943905e-05,"output_power":3.682207624047401e-05,"reference_active":true,"delta_db":-1.1644383004842216}],"active_band_count":3,"active_band_mae_db":0.7969187650695853}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1127,"rms_floor":0.00013636056906360855,"log_rms_mae_db":1.5818806703925556,"rms_envelope_nrmse":0.10527866368225171,"envelope_correlation":0.9924340974707437,"rms_gain_error_db":-0.4696190505861741,"reference_mean_power":0.0025157171675570145,"output_mean_power":0.0022578750699750755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1127,"frequency_resolution_hz":0.1774622892635315,"power_floor":1.3160775774528626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007213514844321301,"output_power":0.0006310530779031009,"reference_active":true,"delta_db":-0.5808104053362663},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011201731569321164,"output_power":9.514677525561947e-05,"reference_active":true,"delta_db":-0.7089108769221342},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.83586447464848e-06,"output_power":3.159314369447064e-06,"reference_active":true,"delta_db":-0.8427041187843621}],"active_band_count":3,"active_band_mae_db":0.7108084670142542}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1133,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":2.159322574946979,"rms_envelope_nrmse":0.2000151359000322,"envelope_correlation":0.9725362603811871,"rms_gain_error_db":-0.8230160232025118,"reference_mean_power":0.00020310353617388895,"output_mean_power":0.00016804124190427877},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1133,"frequency_resolution_hz":0.176522506619594,"power_floor":1.5143267966042753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.933162411577618e-05,"output_power":6.0236994457131304e-05,"reference_active":true,"delta_db":-0.6106807932693923},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3948392508428363e-05,"output_power":1.7583921212251532e-05,"reference_active":true,"delta_db":-1.3416063836120597},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.908897737659843e-06,"output_power":4.285283609894786e-06,"reference_active":true,"delta_db":-1.3952690411218005}],"active_band_count":3,"active_band_mae_db":1.115852072667751}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1127,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":1.6246613040657196,"rms_envelope_nrmse":0.13780817650971622,"envelope_correlation":0.9897064329625066,"rms_gain_error_db":-0.8230060853785177,"reference_mean_power":0.00020418416146590615,"output_mean_power":0.00016893570259768491},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1127,"frequency_resolution_hz":0.1774622892635315,"power_floor":1.3439840528808033e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.555076112703115e-05,"output_power":5.643280890748556e-05,"reference_active":true,"delta_db":-0.6504607122011463},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3202172876588523e-05,"output_power":9.442142801972365e-06,"reference_active":true,"delta_db":-1.45574850977231},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.98101248920639e-07,"output_power":2.997896995771391e-07,"reference_active":true,"delta_db":-2.2050092374866113}],"active_band_count":3,"active_band_mae_db":1.4370728198200224}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1133,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":2.3081220765055175,"rms_envelope_nrmse":0.2889892994490586,"envelope_correlation":0.9559484376264707,"rms_gain_error_db":-1.5627143085007769,"reference_mean_power":5.5634269537956214e-05,"output_mean_power":3.882137910906562e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1133,"frequency_resolution_hz":0.176522506619594,"power_floor":3.230715383503439e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3074231537935687e-05,"output_power":8.806293874513955e-06,"reference_active":true,"delta_db":-1.7162299763979942},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.869637350760251e-06,"output_power":5.832979111371691e-06,"reference_active":true,"delta_db":-0.7104339002054543},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3257906758317962e-06,"output_power":2.2616288197160716e-06,"reference_active":true,"delta_db":-1.6747358170315736}],"active_band_count":3,"active_band_mae_db":1.367133231211674}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90720,"uncovered_tail_samples":0,"metrics":{"frame_count":1127,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":1.9655937031607866,"rms_envelope_nrmse":0.2464069598195395,"envelope_correlation":0.97162395289127,"rms_gain_error_db":-1.5627146666622826,"reference_mean_power":5.5930433017320136e-05,"output_mean_power":3.902803762619706e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1127,"frequency_resolution_hz":0.1774622892635315,"power_floor":2.891541023890363e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.379174882512345e-05,"output_power":9.675550896696384e-06,"reference_active":true,"delta_db":-1.5394363738248638},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.808567459397196e-06,"output_power":3.2641107501989816e-06,"reference_active":true,"delta_db":-0.6699676665514532},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119646570137969e-07,"output_power":1.9793514460245136e-07,"reference_active":true,"delta_db":-1.9758248216348375}],"active_band_count":3,"active_band_mae_db":1.3950762873370515}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.284867763519287,"predicted_mos":3.284867763519287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":136320,"input_sample_rate":24000,"input_seconds":5.68,"input_float32_sha256":"5bd2cd101d065e95bb30054d950972e3da3f09f88cd703e02281e23eca78f919","internal_samples":90880,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0034382400008326},"started_at":"2026-10-13T05:09:06.564317+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028122359999542823,"native_input_samples":136200,"native_input_sha256":"f66c8d856c15de3167e353f6999f8bfcee9d4e0d6e6d6e6bb78d9a57e187ffcd","output_sha256":"8cfef342581888890d8ea5120f687335d12733ac6adad93eaa0e8fb385cd7216","output_samples":136320,"length_delta_samples":120,"analysis_s":2.0933494290002272,"synthesis_s":11.16837925300024,"total_s":13.261728682000467,"rtf":2.3368684902203465,"output_peak":0.435933917760849,"output_rms":0.0724469542997882,"finished_at":"2026-10-13T05:09:19.836394+00:00","analysis_stages":{"external_f0_s":1.346872796999378},"evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"8cfef342581888890d8ea5120f687335d12733ac6adad93eaa0e8fb385cd7216","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":90800,"output_samples":90880,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.79589963962478,"output_peak":0.4325154352600298},"common_prefix":{"samples":90800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991197183098591,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.364791075547191,"spectral_convergence":0.5108733854701967,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.3157155106386105,"spectral_convergence":0.24053267238973014,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.871560126753337,"spectral_convergence":0.22498561272038214,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.0020141860044288512,"delta_db":0.31850314711631306,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0019871451760557503,"delta_db":-0.05919898317698359,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0009875372696957277,"delta_db":-0.29499629420874207,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00017041029392224463,"delta_db":-0.2697217159187192,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":7.45363670340274e-05,"delta_db":-0.23796259437240547,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":2.3844572501821602e-05,"delta_db":-1.1580514341653805,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38973902815975725,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.019325956768158636,"pitch":null,"output_rms_dbfs":-22.79207528214656,"output_peak":0.4325154352600298,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.517355570979713},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90800,"output_samples":90800,"intelligibility":{"estoi":{"value":0.8835682076374908,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07234856775571318,"rms_dbfs":-22.81140123891472,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.478772831760725,"mean":-3.846626449786619e-06,"zero_energy":false},"output":{"rms":0.072509721304377,"rms_dbfs":-22.79207528214656,"peak":0.4325154352600298,"peak_dbfs":-7.279967783468074,"crest_factor_db":15.512107498678487,"mean":0.00017981867714876018,"zero_energy":false},"rms_difference_db":0.01932595676816007},"full_scale":{"reference":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":2.787290266079266e-06,"output_rms_dbfs":-55.548178018671265,"output_zero_energy":false,"output_peak":0.07399320465125554,"output_energy_fraction":0.00011209986937469051,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":258,"both_estimated_voiced_frames":231,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9705882352941176,"absolute_difference_cents_median":5.843281571222308,"absolute_difference_cents_p95":29.25370562074449,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.0001853404348361204,"log_rms_mae_db":2.3783678959998595,"rms_envelope_nrmse":0.1650282661878122,"envelope_correlation":0.9735805302104674,"rms_gain_error_db":0.019326151220353655,"reference_mean_power":0.005238930533969212,"output_mean_power":0.0052622957805129305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":2.4623434734036186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012515931525511833,"output_power":0.0012885139627652668,"reference_active":true,"delta_db":0.126259506101197},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003238300007472361,"output_power":0.00032331096736939216,"reference_active":true,"delta_db":-0.006966438268035741},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.578860153055394e-05,"output_power":6.221797950782433e-05,"reference_active":true,"delta_db":-0.8568798988926551}],"active_band_count":3,"active_band_mae_db":0.33003528108729596}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00016148298564882675,"log_rms_mae_db":1.6548606444223009,"rms_envelope_nrmse":0.09018594170775865,"envelope_correlation":0.9911893011066499,"rms_gain_error_db":0.019327123180717512,"reference_mean_power":0.005266794777733929,"output_mean_power":0.005290285480741492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.1218089587296202e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001164760758537026,"output_power":0.0011943837277671133,"reference_active":true,"delta_db":0.10907147364411714},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015381285066774078,"output_power":0.00015236068568931014,"reference_active":true,"delta_db":-0.04119702660015246},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.591871326154896e-06,"output_power":5.176996637879663e-06,"reference_active":true,"delta_db":-0.3347928642099173}],"active_band_count":3,"active_band_mae_db":0.16168712148472897}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.00014972999954250175,"log_rms_mae_db":2.856561247056333,"rms_envelope_nrmse":0.20055872180161516,"envelope_correlation":0.968163929291111,"rms_gain_error_db":-0.14171681494653285,"reference_mean_power":0.0025001891623953804,"output_mean_power":0.002419920991312392},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.4970716182960256e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007649864504896549,"output_power":0.0007305150740116074,"reference_active":true,"delta_db":-0.20024561027785218},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022996394368157333,"output_power":0.0002198867661356749,"reference_active":true,"delta_db":-0.1946065563554753},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80930068112851e-05,"output_power":4.885172102044304e-05,"reference_active":true,"delta_db":0.06797937947282638}],"active_band_count":3,"active_band_mae_db":0.15427718203538462}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00013636056906360855,"log_rms_mae_db":2.093979140211118,"rms_envelope_nrmse":0.11708688043296954,"envelope_correlation":0.988155689348665,"rms_gain_error_db":-0.14171574119295888,"reference_mean_power":0.0025134871346063777,"output_mean_power":0.0024327926357493378},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3161092414230216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215738253307795,"output_power":0.0006888685710964492,"reference_active":true,"delta_db":-0.20144400065365475},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011225926953226224,"output_power":0.0001064128105143338,"reference_active":true,"delta_db":-0.23228298270805056},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.841934016741924e-06,"output_power":4.40845116000568e-06,"reference_active":true,"delta_db":0.5973613192474619}],"active_band_count":3,"active_band_mae_db":0.3436961008697224}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":2.6790602858432333,"rms_envelope_nrmse":0.25209600940043525,"envelope_correlation":0.9490573104783103,"rms_gain_error_db":-0.17972378922138324,"reference_mean_power":0.00020292459461300388,"output_mean_power":0.00019469836662123623},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.5147206216944186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.93765985961165e-05,"output_power":6.511331059487732e-05,"reference_active":true,"delta_db":-0.2754322638643964},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.392588995706175e-05,"output_power":2.1530654673503223e-05,"reference_active":true,"delta_db":-0.4581086550946037},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8985142545041605e-06,"output_power":5.94454565328587e-06,"reference_active":true,"delta_db":0.03376033442573953}],"active_band_count":3,"active_band_mae_db":0.25576708446157986}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":1.6189988059439588,"rms_envelope_nrmse":0.12299686064261137,"envelope_correlation":0.9867565654781026,"rms_gain_error_db":-0.17972007605115028,"reference_mean_power":0.0002040033414802843,"output_mean_power":0.00019573355022173836},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3444198629453335e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.55829624744988e-05,"output_power":6.121592127518476e-05,"reference_active":true,"delta_db":-0.2992664063120542},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.319750377898722e-05,"output_power":1.1839514432613944e-05,"reference_active":true,"delta_db":-0.47157903753995134},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.991435827054759e-07,"output_power":5.236669035606836e-07,"reference_active":true,"delta_db":0.20829634854383966}],"active_band_count":3,"active_band_mae_db":0.32638059746528175}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":2.328009095481495,"rms_envelope_nrmse":0.3189108258636644,"envelope_correlation":0.9313854404476266,"rms_gain_error_db":-0.7685624767200652,"reference_mean_power":5.558522396873643e-05,"output_mean_power":4.6569664827224015e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":3.2309689753050144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093182032279317e-05,"output_power":1.1587968916638207e-05,"reference_active":true,"delta_db":-0.5303788425235931},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.882644346015486e-06,"output_power":5.406512696422554e-06,"reference_active":true,"delta_db":-1.0483810118045147},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3224614880429852e-06,"output_power":2.6992101157175613e-06,"reference_active":true,"delta_db":-0.9022326273042571}],"active_band_count":3,"active_band_mae_db":0.8269974938774549}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":1.3538962029879702,"rms_envelope_nrmse":0.19878428978634327,"envelope_correlation":0.9747945671546123,"rms_gain_error_db":-0.7685636065367772,"reference_mean_power":5.5880857927829866e-05,"output_mean_power":4.681733671557513e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.8916608161728035e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3816996130233868e-05,"output_power":1.1950273366054945e-05,"reference_active":true,"delta_db":-0.6303579594384203},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.814539997745628e-06,"output_power":2.9203016146138574e-06,"reference_active":true,"delta_db":-1.1601446452310646},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.118413559303931e-07,"output_power":2.469626834986188e-07,"reference_active":true,"delta_db":-1.0130237460437321}],"active_band_count":3,"active_band_mae_db":0.9345087835710725}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.272460460662842,"predicted_mos":3.272460460662842,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":136440,"input_sample_rate":24000,"input_seconds":5.685,"input_float32_sha256":"dc0eab9cdca689040886e801de0e1c6d874f1fcaa7818d2ae4348f76033a5c48","internal_samples":90960,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0096411690010427},"started_at":"2026-10-13T05:10:29.346806+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027505769994604634,"native_input_samples":136200,"native_input_sha256":"f66c8d856c15de3167e353f6999f8bfcee9d4e0d6e6d6e6bb78d9a57e187ffcd","output_sha256":"f5ec3d26a1d83a220ed83195da319227601da63630ceb87ad1cc216691eb5935","output_samples":136440,"length_delta_samples":240,"analysis_s":0.8143319199989492,"synthesis_s":0.12642957900061447,"total_s":0.9407614989995636,"rtf":0.1657729513655619,"output_peak":0.4313487112522125,"output_rms":0.07123555172046335,"finished_at":"2026-10-13T05:10:30.299430+00:00","analysis_stages":{"external_f0_s":0.15916622300028394},"evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"f5ec3d26a1d83a220ed83195da319227601da63630ceb87ad1cc216691eb5935","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":90800,"output_samples":90960,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.942153891605987,"output_peak":0.4316338155306825},"common_prefix":{"samples":90800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982409850483729,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.82032003318543,"spectral_convergence":0.13656825811796214,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.34246659266533,"spectral_convergence":0.13220966762385458,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.596554712134324,"spectral_convergence":0.14845655473872757,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.0018344667539036074,"delta_db":-0.08739413804093944,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0019651096447631283,"delta_db":-0.10762707925663936,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.001015847529312455,"delta_db":-0.1722459565061696,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00017341471530592809,"delta_db":-0.19382049844572366,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":7.241755387143364e-05,"delta_db":-0.36320629409038657,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":2.676839138731826e-05,"delta_db":-0.6557241040121391,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.26333634505866627,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1231067467824973,"pitch":null,"output_rms_dbfs":-22.934507985697216,"output_peak":0.4316338155306825,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.586447112661695},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90800,"output_samples":90800,"intelligibility":{"estoi":{"value":0.9547636434508618,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07234856775571318,"rms_dbfs":-22.81140123891472,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.478772831760725,"mean":-3.846626449786619e-06,"zero_energy":false},"output":{"rms":0.0713303903574337,"rms_dbfs":-22.934507985697216,"peak":0.4316338155306825,"peak_dbfs":-7.297690773615086,"crest_factor_db":15.636817212082128,"mean":-8.774632311081037e-06,"zero_energy":false},"rms_difference_db":-0.12310674678249711},"full_scale":{"reference":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":9.651120795227804e-07,"output_rms_dbfs":-60.15422248633619,"output_zero_energy":false,"output_peak":0.023854046754009064,"output_energy_fraction":4.010919356215228e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":247,"both_estimated_voiced_frames":227,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9537815126050421,"absolute_difference_cents_median":0.9428181352008934,"absolute_difference_cents_p95":6.544261724371681,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.0001853404348361204,"log_rms_mae_db":0.9323063194097561,"rms_envelope_nrmse":0.05023065991898907,"envelope_correlation":0.9977016368265593,"rms_gain_error_db":-0.12310653127693855,"reference_mean_power":0.005238930533969212,"output_mean_power":0.0050925111293448795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":2.4623434734036186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012515931525511833,"output_power":0.0012311055734287018,"reference_active":true,"delta_db":-0.07167881020463245},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003238300007472361,"output_power":0.00028862745742764466,"reference_active":true,"delta_db":-0.49979437240823643},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.578860153055394e-05,"output_power":5.463982332558204e-05,"reference_active":true,"delta_db":-1.4209460702191448}],"active_band_count":3,"active_band_mae_db":0.6641397509440045}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00016148298564882675,"log_rms_mae_db":0.5541708089833665,"rms_envelope_nrmse":0.028716012311459295,"envelope_correlation":0.9993166692723054,"rms_gain_error_db":-0.12310629766393602,"reference_mean_power":0.005266794777733929,"output_mean_power":0.005119596889179364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.1218089587296202e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001164760758537026,"output_power":0.0011449942502515303,"reference_active":true,"delta_db":-0.07433424772378537},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015381285066774078,"output_power":0.0001355067896750929,"reference_active":true,"delta_db":-0.5503156474243198},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.591871326154896e-06,"output_power":3.473507435590542e-06,"reference_active":true,"delta_db":-2.0678893646990115}],"active_band_count":3,"active_band_mae_db":0.8975130866157056}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.00014972999954250175,"log_rms_mae_db":1.2016881285565983,"rms_envelope_nrmse":0.06961954180100063,"envelope_correlation":0.9963518639336696,"rms_gain_error_db":-0.14471793025886726,"reference_mean_power":0.0025001891623953804,"output_mean_power":0.0024182493253060164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.4970716182960256e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007649864504896549,"output_power":0.0007493204028252904,"reference_active":true,"delta_db":-0.08986185037805808},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022996394368157333,"output_power":0.00020585622546421,"reference_active":true,"delta_db":-0.4809574242839901},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80930068112851e-05,"output_power":3.1410918140784394e-05,"reference_active":true,"delta_db":-1.8500129933638099}],"active_band_count":3,"active_band_mae_db":0.8069440893419527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00013636056906360855,"log_rms_mae_db":0.6971715144796131,"rms_envelope_nrmse":0.04269964099678802,"envelope_correlation":0.9986328745361608,"rms_gain_error_db":-0.1447178648886666,"reference_mean_power":0.0025134871346063777,"output_mean_power":0.0024311115136159173},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3161092414230216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215738253307795,"output_power":0.0007038622064100103,"reference_active":true,"delta_db":-0.10793124474097297},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011225926953226224,"output_power":9.754006765204997e-05,"reference_active":true,"delta_db":-0.6103915947698001},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.841934016741924e-06,"output_power":2.320775197221237e-06,"reference_active":true,"delta_db":-2.189168274115529}],"active_band_count":3,"active_band_mae_db":0.9691637045421007}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":1.2780584544017768,"rms_envelope_nrmse":0.09735327010508829,"envelope_correlation":0.992864136578797,"rms_gain_error_db":-0.22154507335849316,"reference_mean_power":0.00020292459461300388,"output_mean_power":0.00019283247689049797},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.5147206216944186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.93765985961165e-05,"output_power":6.765711117774668e-05,"reference_active":true,"delta_db":-0.10899553039029294},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.392588995706175e-05,"output_power":2.3012532493805864e-05,"reference_active":true,"delta_db":-0.16903686203813464},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8985142545041605e-06,"output_power":4.851475637342566e-06,"reference_active":true,"delta_db":-0.848687784560545}],"active_band_count":3,"active_band_mae_db":0.37557339232965753}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":0.82518562135219,"rms_envelope_nrmse":0.06306008301056583,"envelope_correlation":0.9970910540646339,"rms_gain_error_db":-0.22154227792088788,"reference_mean_power":0.0002040033414802843,"output_mean_power":0.0001938576988547594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3444198629453335e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.55829624744988e-05,"output_power":6.45681663725447e-05,"reference_active":true,"delta_db":-0.06772577120717681},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.319750377898722e-05,"output_power":1.2639224580562684e-05,"reference_active":true,"delta_db":-0.18771364335580146},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.991435827054759e-07,"output_power":4.047736801623782e-07,"reference_active":true,"delta_db":-0.9101322656640474}],"active_band_count":3,"active_band_mae_db":0.38852389340900856}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":1.1109546712695009,"rms_envelope_nrmse":0.13861055224361513,"envelope_correlation":0.9887465679870465,"rms_gain_error_db":-0.4959936426715268,"reference_mean_power":5.558522396873643e-05,"output_mean_power":4.958610499209923e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":3.2309689753050144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093182032279317e-05,"output_power":1.2207612604501757e-05,"reference_active":true,"delta_db":-0.30414466882149016},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.882644346015486e-06,"output_power":5.511313421223592e-06,"reference_active":true,"delta_db":-0.9650021889666971},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3224614880429852e-06,"output_power":2.3256009297401286e-06,"reference_active":true,"delta_db":-1.5492476317912356}],"active_band_count":3,"active_band_mae_db":0.9394648298598076}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":0.709498832778531,"rms_envelope_nrmse":0.08331754870847576,"envelope_correlation":0.9970957966076065,"rms_gain_error_db":-0.4959931734386278,"reference_mean_power":5.5880857927829866e-05,"output_mean_power":4.984983759711932e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.8916608161728035e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3816996130233868e-05,"output_power":1.2634596158377426e-05,"reference_active":true,"delta_db":-0.38852270903352226},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.814539997745628e-06,"output_power":2.9373706642590014e-06,"reference_active":true,"delta_db":-1.1348341986880737},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.118413559303931e-07,"output_power":2.2070038179620635e-07,"reference_active":true,"delta_db":-1.5013062569023619}],"active_band_count":3,"active_band_mae_db":1.0082210548746524}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0001","method":"magphase","status":"measured","mos":{"status":"measured","value":3.7447071075439453,"predicted_mos":3.7447071075439453,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":273005,"input_sample_rate":48000,"input_seconds":5.687604166666667,"input_float32_sha256":"5895ff7e4f1e79720a8b68e7ec161dc4a63f5df62c05596760bc52bec6f3d1fa","internal_samples":91002,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9374261039993144},"started_at":"2026-10-13T05:14:29.679156+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.014103839001109009,"native_input_samples":272400,"native_input_sha256":"9da0d7fb53d76a356f4a6d5ae143e196612ac61b21e4499bf2c286f4a4208a18","output_sha256":"4fbd1a986684b35fffdafc4b95ea72d5222908ae45a5007381925b6f03d0f0a9","output_samples":273005,"length_delta_samples":605,"analysis_s":3.261637883999356,"synthesis_s":0.9850156950014934,"total_s":4.246653579000849,"rtf":0.7483090007049955,"output_peak":0.44413320777858467,"output_rms":0.06995129544345689,"finished_at":"2026-10-13T05:14:33.958012+00:00","evaluation_input":{"reference":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"4fbd1a986684b35fffdafc4b95ea72d5222908ae45a5007381925b6f03d0f0a9","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":90800,"output_samples":91002,"length_delta_samples":202,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.10020066504221,"output_peak":0.4420756636822757},"common_prefix":{"samples":90800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.997780268565526,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":90800,"output_samples":90800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.35912217654554,"spectral_convergence":0.6089288680723473,"spectral_convergence_status":"ok","amplitude_floor":1.4038593411999917e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.0724385066416025,"spectral_convergence":0.25379803938622403,"spectral_convergence_status":"ok","amplitude_floor":9.854110842773577e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.701302851259027,"spectral_convergence":0.2246660309982104,"spectral_convergence_status":"ok","amplitude_floor":8.983114295610284e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717561089627995,"output_power":0.001688040168921184,"delta_db":-0.44866482269635377,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00201441754378434,"output_power":0.0019048709141331855,"delta_db":-0.24283944577474,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010569467904903444,"output_power":0.0010172983306283678,"delta_db":-0.16604792492175943,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00018132930505040446,"output_power":0.00018566902368498243,"delta_db":0.10271456632471251,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.873439242518033e-05,"output_power":8.186370443382637e-05,"delta_db":0.16926912627889223,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.113111558995142e-05,"output_power":3.0714340244816924e-05,"delta_db":-0.058534930411762755,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1980118027347035,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2791493117843702,"pitch":null,"output_rms_dbfs":-23.09055055069909,"output_peak":0.4420756636822757,"reference_rms_dbfs":-22.81140123891472,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.377621178148723},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":90800,"output_samples":90800,"intelligibility":{"estoi":{"value":0.8831100469272231,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07234856775571318,"rms_dbfs":-22.81140123891472,"peak":0.429901123046875,"peak_dbfs":-7.332628407153992,"crest_factor_db":15.478772831760725,"mean":-3.846626449786619e-06,"zero_energy":false},"output":{"rms":0.07006037740439572,"rms_dbfs":-23.09055055069909,"peak":0.4420756636822757,"peak_dbfs":-7.090067847976529,"crest_factor_db":16.00048270272256,"mean":-0.0007604725202292979,"zero_energy":false},"rms_difference_db":-0.2791493117843693},"full_scale":{"reference":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":90800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":567,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.4351076785242186e-06,"threshold_dbfs":-54.64059644789021,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":4.230425111018121e-07,"output_mean_power":1.2769775688391775e-06,"output_rms_dbfs":-58.93816731409199,"output_zero_energy":false,"output_peak":0.021720155585917802,"output_energy_fraction":5.5011527463423476e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":560,"reference_estimated_voiced_frames":238,"output_estimated_voiced_frames":243,"both_estimated_voiced_frames":229,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9621848739495799,"absolute_difference_cents_median":5.245690784761436,"absolute_difference_cents_p95":21.698057783440074,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":90800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.0001853404348361204,"log_rms_mae_db":2.135385664938376,"rms_envelope_nrmse":0.1906902231760941,"envelope_correlation":0.9647068857146962,"rms_gain_error_db":-0.2791490512100549,"reference_mean_power":0.005238930533969212,"output_mean_power":0.004912784733471326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":2.4623434734036186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012515931525511833,"output_power":0.0012168138295671857,"reference_active":true,"delta_db":-0.12239041461168504},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003238300007472361,"output_power":0.00031254907644479593,"reference_active":true,"delta_db":-0.1539886090699415},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.578860153055394e-05,"output_power":7.669107585231538e-05,"reference_active":true,"delta_db":0.05140936830816328}],"active_band_count":3,"active_band_mae_db":0.10926279732992994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00016148298564882675,"log_rms_mae_db":1.43699163634798,"rms_envelope_nrmse":0.11061312926682526,"envelope_correlation":0.9874936705527652,"rms_gain_error_db":-0.2791489167205193,"reference_mean_power":0.005266794777733929,"output_mean_power":0.004938914462002442},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.1218089587296202e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001164760758537026,"output_power":0.0011367937556661608,"reference_active":true,"delta_db":-0.10555051208646159},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015381285066774078,"output_power":0.00015060554934012477,"reference_active":true,"delta_db":-0.09151646624008486},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.591871326154896e-06,"output_power":5.5855736737939215e-06,"reference_active":true,"delta_db":-0.004893848207241034}],"active_band_count":3,"active_band_mae_db":0.0673202755112625}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":0.00014972999954250175,"log_rms_mae_db":2.7600131329459954,"rms_envelope_nrmse":0.20900324358526176,"envelope_correlation":0.9653569130871812,"rms_gain_error_db":-0.20108753912060728,"reference_mean_power":0.0025001891623953804,"output_mean_power":0.002387064281860289},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.4970716182960256e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007649864504896549,"output_power":0.0007447699925945086,"reference_active":true,"delta_db":-0.11631572719829257},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022996394368157333,"output_power":0.0002220868302868432,"reference_active":true,"delta_db":-0.15136942090921343},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80930068112851e-05,"output_power":4.9748274399473804e-05,"reference_active":true,"delta_db":0.146960907830463}],"active_band_count":3,"active_band_mae_db":0.138215351979323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":0.00013636056906360855,"log_rms_mae_db":2.0519775717622792,"rms_envelope_nrmse":0.129383409738184,"envelope_correlation":0.9856219171100745,"rms_gain_error_db":-0.20108737677809624,"reference_mean_power":0.0025134871346063777,"output_mean_power":0.0023997606566956613},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3161092414230216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215738253307795,"output_power":0.0007054103471865925,"reference_active":true,"delta_db":-0.09838945291051093},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011225926953226224,"output_power":0.00011011208841133365,"reference_active":true,"delta_db":-0.0838721230629666},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.841934016741924e-06,"output_power":4.176787833012393e-06,"reference_active":true,"delta_db":0.3629251330078643}],"active_band_count":3,"active_band_mae_db":0.1817289029937806}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.2746042839695326e-05,"log_rms_mae_db":2.5660836944630483,"rms_envelope_nrmse":0.25901082769786926,"envelope_correlation":0.9479522014338394,"rms_gain_error_db":0.1098984516282179,"reference_mean_power":0.00020292459461300388,"output_mean_power":0.0002081251348314132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":1.5147206216944186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.93765985961165e-05,"output_power":7.248836546501608e-05,"reference_active":true,"delta_db":0.1905530359933113},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.392588995706175e-05,"output_power":2.443173381505572e-05,"reference_active":true,"delta_db":0.09086187136131099},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8985142545041605e-06,"output_power":5.832032858659448e-06,"reference_active":true,"delta_db":-0.0492267107610046}],"active_band_count":3,"active_band_mae_db":0.11021387270520895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.512933910060941e-05,"log_rms_mae_db":1.6326633656832195,"rms_envelope_nrmse":0.1493919315699091,"envelope_correlation":0.9808417070141437,"rms_gain_error_db":0.10989681225325448,"reference_mean_power":0.0002040033414802843,"output_mean_power":0.0002092314487826687},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":1.3444198629453335e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.55829624744988e-05,"output_power":6.841700715031392e-05,"reference_active":true,"delta_db":0.183730419164081},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.319750377898722e-05,"output_power":1.312612115276445e-05,"reference_active":true,"delta_db":-0.023553865960148152},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.991435827054759e-07,"output_power":4.7335079791702806e-07,"reference_active":true,"delta_db":-0.23042378395436458}],"active_band_count":3,"active_band_mae_db":0.14590268969286457}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1134,"rms_floor":5.6247119079700065e-05,"log_rms_mae_db":2.3286477600378612,"rms_envelope_nrmse":0.3579574227127758,"envelope_correlation":0.9145918980686912,"rms_gain_error_db":-0.024892099258334386,"reference_mean_power":5.558522396873643e-05,"output_mean_power":5.526754200240419e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1134,"frequency_resolution_hz":0.1763668430335097,"power_floor":3.2309689753050144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093182032279317e-05,"output_power":1.2319187292412807e-05,"reference_active":true,"delta_db":-0.26463147856187613},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.882644346015486e-06,"output_power":5.438699673003862e-06,"reference_active":true,"delta_db":-1.0226025081028491},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3224614880429852e-06,"output_power":2.4289444008935294e-06,"reference_active":true,"delta_db":-1.3604238174945031}],"active_band_count":3,"active_band_mae_db":0.8825526013864096}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":90800,"uncovered_tail_samples":0,"metrics":{"frame_count":1128,"rms_floor":4.362122194789186e-05,"log_rms_mae_db":1.5706106768420556,"rms_envelope_nrmse":0.23121950057022783,"envelope_correlation":0.9616300434581913,"rms_gain_error_db":-0.02489249655785774,"reference_mean_power":5.5880857927829866e-05,"output_mean_power":5.556148126455754e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1128,"frequency_resolution_hz":0.1773049645390071,"power_floor":2.8916608161728035e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3816996130233868e-05,"output_power":1.2546704444613622e-05,"reference_active":true,"delta_db":-0.4188396826500193},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.814539997745628e-06,"output_power":2.8391206707734303e-06,"reference_active":true,"delta_db":-1.2825832141452052},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.118413559303931e-07,"output_power":1.9504839407079026e-07,"reference_active":true,"delta_db":-2.0379133124878965}],"active_band_count":3,"active_band_mae_db":1.2464454030943737}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"hnm","status":"measured","mos":{"status":"measured","value":3.347336769104004,"predicted_mos":3.347336769104004,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85280,"input_sample_rate":16000,"input_seconds":5.33,"input_float32_sha256":"23c566cc3b64a1dd86eb5d3879790af6a1ab19fbc004fad9e1f2255e67cb5eeb","internal_samples":85280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1435189849999006},"started_at":"2026-10-13T04:38:55.546157+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4238999938243069e-05,"native_input_samples":85280,"native_input_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","output_sha256":"e33c61f73e6394e11a3c2b21a0f1f8b96389d39e01bbe8a58e8c75eb0666913a","output_samples":85280,"length_delta_samples":0,"analysis_s":0.926096502999826,"synthesis_s":0.45266020400049456,"total_s":1.3787567070003206,"rtf":0.2586785566604729,"output_peak":0.5818677398100551,"output_rms":0.06286372522963973,"finished_at":"2026-10-13T04:38:56.930274+00:00","evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"e33c61f73e6394e11a3c2b21a0f1f8b96389d39e01bbe8a58e8c75eb0666913a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.61797870408133,"spectral_convergence":0.18394847423431984,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.3006964366369065,"spectral_convergence":0.18577484418173382,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.605906574261126,"spectral_convergence":0.18198587987211273,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.001245724698654336,"delta_db":0.03842573861099924,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.0018219605439716926,"delta_db":-0.05415784159140105,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.0006048235497377465,"delta_db":-0.10329902463217464,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.00016224344840988898,"delta_db":-0.31294776847535405,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":0.00010183257330376488,"delta_db":-0.14787506158946884,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.5263135670214666e-05,"delta_db":0.05610382435336597,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11880154320879395,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04580590509833145,"pitch":null,"output_rms_dbfs":-24.031997735621218,"output_peak":0.5818677398100551,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.508193904993121},"common_prefix":{"samples":85280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.61797870408133,"spectral_convergence":0.18394847423431984,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.3006964366369065,"spectral_convergence":0.18577484418173382,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.605906574261126,"spectral_convergence":0.18198587987211273,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.001245724698654336,"delta_db":0.03842573861099924,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.0018219605439716926,"delta_db":-0.05415784159140105,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.0006048235497377465,"delta_db":-0.10329902463217464,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.00016224344840988898,"delta_db":-0.31294776847535405,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":0.00010183257330376488,"delta_db":-0.14787506158946884,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.5263135670214666e-05,"delta_db":0.05610382435336597,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11880154320879395,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04580590509833145,"pitch":null,"output_rms_dbfs":-24.031997735621218,"output_peak":0.5818677398100551,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.508193904993121},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85280,"output_samples":85280,"intelligibility":{"estoi":{"value":0.9371270386072423,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06319611903854065,"rms_dbfs":-23.986191830522884,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.497017771838582,"mean":-1.0716221793283889e-05,"zero_energy":false},"output":{"rms":0.06286372522963973,"rms_dbfs":-24.031997735621218,"peak":0.5818677398100551,"peak_dbfs":-4.703514410036486,"crest_factor_db":19.32848332558473,"mean":-2.192218423556073e-05,"zero_energy":false},"rms_difference_db":-0.04580590509833371},"full_scale":{"reference":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":533,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":116,"quiet_samples":18560,"reference_mean_power":8.461537774941273e-07,"output_mean_power":1.352591852563609e-06,"output_rms_dbfs":-58.68833232907514,"output_zero_energy":false,"output_peak":0.01988249146924237,"output_energy_fraction":7.448988792589578e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":526,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":221,"both_estimated_voiced_frames":211,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9504504504504504,"absolute_difference_cents_median":1.2351361925822586,"absolute_difference_cents_p95":7.360691224460018,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.0001748275799785616,"log_rms_mae_db":0.904079591047336,"rms_envelope_nrmse":0.06351861451516215,"envelope_correlation":0.9962523866943708,"rms_gain_error_db":-0.04580587882909613,"reference_mean_power":0.003997498793754849,"output_mean_power":0.003955557968753488},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.9383161701154944e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012108392795166332,"output_power":0.0011956149358283567,"reference_active":true,"delta_db":-0.05495169303903742},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231683599850243,"output_power":0.0002056322569753599,"reference_active":true,"delta_db":-0.3554137963891836},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.764859371015312e-05,"output_power":4.955478918830802e-05,"reference_active":true,"delta_db":-1.3517313990466877}],"active_band_count":3,"active_band_mae_db":0.5873656294916362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00017019095451879833,"log_rms_mae_db":0.45939132773295,"rms_envelope_nrmse":0.03366398896153657,"envelope_correlation":0.9988740725505354,"rms_gain_error_db":-0.04580636986016415,"reference_mean_power":0.00402014109400192,"output_mean_power":0.003977962261502589},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.6741447857886242e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011176954288226882,"output_power":0.0011081596898744376,"reference_active":true,"delta_db":-0.03721126417666045},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.194819962775495e-05,"output_power":8.485988336556893e-05,"reference_active":true,"delta_db":-0.3484079952496197},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3053605179616585e-06,"output_power":2.5542952593966178e-06,"reference_active":true,"delta_db":-1.1194773784737846}],"active_band_count":3,"active_band_mae_db":0.5016988793000215}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.00014662911909855806,"log_rms_mae_db":1.75108703575241,"rms_envelope_nrmse":0.10495488857648781,"envelope_correlation":0.9912517084092691,"rms_gain_error_db":-0.0917345928816815,"reference_mean_power":0.0015423797799867744,"output_mean_power":0.0015101422696981088},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.0119951767590758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006526269487739279,"output_power":0.0006338326967372958,"reference_active":true,"delta_db":-0.1269036420363261},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016341376588393177,"output_power":0.00014981553704065012,"reference_active":true,"delta_db":-0.3773178303388719},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.788084498799748e-05,"output_power":3.724921097308029e-05,"reference_active":true,"delta_db":-1.0904472832045087}],"active_band_count":3,"active_band_mae_db":0.5315562518599023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00013732417437068112,"log_rms_mae_db":0.9738853152057556,"rms_envelope_nrmse":0.05769272647789234,"envelope_correlation":0.9971610409734994,"rms_gain_error_db":-0.09173609150134074,"reference_mean_power":0.0015511172002359399,"output_mean_power":0.001518696543764249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":8.798155816797883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006408046711327789,"output_power":0.0006194872505120074,"reference_active":true,"delta_db":-0.1469329615373121},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.486060187757687e-05,"output_power":6.787017003098507e-05,"reference_active":true,"delta_db":-0.42574377306203676},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5748897665955194e-06,"output_power":2.6189446842237835e-06,"reference_active":true,"delta_db":-1.3513632905432873}],"active_band_count":3,"active_band_mae_db":0.6413466750475454}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":1.5215906988676045,"rms_envelope_nrmse":0.1368585143783423,"envelope_correlation":0.9861263974100278,"rms_gain_error_db":-0.23787591820125645,"reference_mean_power":0.00020984803844209808,"output_mean_power":0.0001986631522023256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":2.0976407038326692e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011684508155422022,"output_power":0.00011725967273150024,"reference_active":true,"delta_db":0.015382417099921164},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.122844755814484e-05,"output_power":3.363169913011665e-05,"reference_active":true,"delta_db":-0.8844817275407151},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.871758765973679e-06,"output_power":5.599177763734965e-06,"reference_active":true,"delta_db":-1.9988546852515174}],"active_band_count":3,"active_band_mae_db":0.9662396099640512}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":0.8001930146549988,"rms_envelope_nrmse":0.07984268416688839,"envelope_correlation":0.9952352196188545,"rms_gain_error_db":-0.2378768797935233,"reference_mean_power":0.00021103616868638406,"output_mean_power":0.00019978791094903597},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.885738469163612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011739951182146861,"output_power":0.00011701247479225687,"reference_active":true,"delta_db":-0.0143412631111734},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.974678911762062e-05,"output_power":1.575808490094448e-05,"reference_active":true,"delta_db":-0.9799305219390876},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.379135569148538e-07,"output_power":3.693009255051772e-07,"reference_active":true,"delta_db":-2.3738143582059235}],"active_band_count":3,"active_band_mae_db":1.1226953810853948}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":1.6434582744357968,"rms_envelope_nrmse":0.17163937932378617,"envelope_correlation":0.9800852799449338,"rms_gain_error_db":-0.09609714979678045,"reference_mean_power":4.50355353361435e-05,"output_mean_power":4.404996975686399e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":3.018630059598247e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.579538283809341e-05,"output_power":1.5521922740314387e-05,"reference_active":true,"delta_db":-0.07584639088607509},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.352602360298045e-06,"output_power":3.566545256751562e-06,"reference_active":true,"delta_db":-0.8650125396867336},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8485554777348664e-06,"output_power":1.219887842177425e-06,"reference_active":true,"delta_db":-1.8051258588891925}],"active_band_count":3,"active_band_mae_db":0.9153282631540005}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":0.887176121481282,"rms_envelope_nrmse":0.0870345646505493,"envelope_correlation":0.9945496677204801,"rms_gain_error_db":-0.09609743382746187,"reference_mean_power":4.529062388817584e-05,"output_mean_power":4.429947300893397e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":2.7414916263445047e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5534436484200313e-05,"output_power":1.4835794102413652e-05,"reference_active":true,"delta_db":-0.19984706358141963},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0764100435071083e-06,"output_power":1.6777443332257498e-06,"reference_active":true,"delta_db":-0.9258734026454234},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47015628096149e-07,"output_power":9.559955543456267e-08,"reference_active":true,"delta_db":-1.8690763100833814}],"active_band_count":3,"active_band_mae_db":0.998265592103408}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.330228090286255,"predicted_mos":3.330228090286255,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85280,"input_sample_rate":16000,"input_seconds":5.33,"input_float32_sha256":"a6557d5d7bab604ca351e506e9a38036f42b6e4701bcb6c6e305c028049f829f","internal_samples":85280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0126126279992604},"started_at":"2026-10-13T04:41:52.851164+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4109000150929205e-05,"native_input_samples":85280,"native_input_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","output_sha256":"b2f0d8af78d89a2d1c6b878acd14322f9dcafd8c3dda0fc4877c05b999a95a4a","output_samples":85280,"length_delta_samples":0,"analysis_s":1.6113395230004244,"synthesis_s":0.09520325099947513,"total_s":1.7065427739998995,"rtf":0.3201768806754033,"output_peak":0.6475635112427698,"output_rms":0.06848937857266912,"finished_at":"2026-10-13T04:41:54.564266+00:00","evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"b2f0d8af78d89a2d1c6b878acd14322f9dcafd8c3dda0fc4877c05b999a95a4a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.53407775635151,"spectral_convergence":0.59941473655827,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.269105926573236,"spectral_convergence":0.22056588776648847,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.835404163700802,"spectral_convergence":0.2631210751805184,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.0014506824567678627,"delta_db":0.6999285780417169,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.002133994996192124,"delta_db":0.6323864485270461,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.0007579328767333743,"delta_db":0.8767215058577206,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.0002057415455635503,"delta_db":0.7186005267239506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":0.00012656008314379486,"delta_db":0.796225268000776,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.588301886968147e-05,"delta_db":0.2289966972050954,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6588098373927176,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6986563450595564,"pitch":null,"output_rms_dbfs":-23.28753548546333,"output_peak":0.6475635112427698,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.879529282208516},"common_prefix":{"samples":85280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.53407775635151,"spectral_convergence":0.59941473655827,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.269105926573236,"spectral_convergence":0.22056588776648847,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.835404163700802,"spectral_convergence":0.2631210751805184,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.0014506824567678627,"delta_db":0.6999285780417169,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.002133994996192124,"delta_db":0.6323864485270461,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.0007579328767333743,"delta_db":0.8767215058577206,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.0002057415455635503,"delta_db":0.7186005267239506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":0.00012656008314379486,"delta_db":0.796225268000776,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.588301886968147e-05,"delta_db":0.2289966972050954,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6588098373927176,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6986563450595564,"pitch":null,"output_rms_dbfs":-23.28753548546333,"output_peak":0.6475635112427698,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.879529282208516},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85280,"output_samples":85280,"intelligibility":{"estoi":{"value":0.9501864125464903,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06319611903854065,"rms_dbfs":-23.986191830522884,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.497017771838582,"mean":-1.0716221793283889e-05,"zero_energy":false},"output":{"rms":0.06848937857266912,"rms_dbfs":-23.28753548546333,"peak":0.6475635112427698,"peak_dbfs":-3.774352615061849,"crest_factor_db":19.513182870401483,"mean":-4.911075331595226e-09,"zero_energy":false},"rms_difference_db":0.698656345059554},"full_scale":{"reference":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":533,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":116,"quiet_samples":18560,"reference_mean_power":8.461537774941273e-07,"output_mean_power":1.8986767896215535e-06,"output_rms_dbfs":-57.21548958607158,"output_zero_energy":false,"output_peak":0.02752655221980954,"output_energy_fraction":8.80917768811711e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":526,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":235,"both_estimated_voiced_frames":217,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9774774774774775,"absolute_difference_cents_median":7.029475395316614,"absolute_difference_cents_p95":22.636777293604794,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.0001748275799785616,"log_rms_mae_db":1.5635504896639263,"rms_envelope_nrmse":0.14822922949819284,"envelope_correlation":0.9871252581397189,"rms_gain_error_db":0.6986566831767256,"reference_mean_power":0.003997498793754849,"output_mean_power":0.004695199061403306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.9383161701154944e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012108392795166332,"output_power":0.0013793044387507625,"reference_active":true,"delta_db":0.5657363283026945},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231683599850243,"output_power":0.00023987549260026864,"reference_active":true,"delta_db":0.31353257691462877},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.764859371015312e-05,"output_power":7.0447791535812e-05,"reference_active":true,"delta_db":0.17608610139494668}],"active_band_count":3,"active_band_mae_db":0.35178500220409}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00017019095451879833,"log_rms_mae_db":1.0815808634119883,"rms_envelope_nrmse":0.10690368946007067,"envelope_correlation":0.9959042742702697,"rms_gain_error_db":0.6986566487280966,"reference_mean_power":0.00402014109400192,"output_mean_power":0.004721793180033952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.6741447857886242e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011176954288226882,"output_power":0.001276916393712271,"reference_active":true,"delta_db":0.5783898795256397},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.194819962775495e-05,"output_power":9.798452908853266e-05,"reference_active":true,"delta_db":0.2761427970867082},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3053605179616585e-06,"output_power":3.578104646158805e-06,"reference_active":true,"delta_db":0.34434202698859384}],"active_band_count":3,"active_band_mae_db":0.3996249012003139}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.00014662911909855806,"log_rms_mae_db":2.0196904753885265,"rms_envelope_nrmse":0.1996306071924157,"envelope_correlation":0.981894650841226,"rms_gain_error_db":0.9839982568694718,"reference_mean_power":0.0015423797799867744,"output_mean_power":0.0019345998459158358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.0119951767590758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006526269487739279,"output_power":0.0008267810586271385,"reference_active":true,"delta_db":1.02725515433916},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016341376588393177,"output_power":0.0001871972705873034,"reference_active":true,"delta_db":0.5901087379638897},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.788084498799748e-05,"output_power":4.599055217710688e-05,"reference_active":true,"delta_db":-0.1749318222953189}],"active_band_count":3,"active_band_mae_db":0.5974319048661229}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00013732417437068112,"log_rms_mae_db":1.4472809276646192,"rms_envelope_nrmse":0.1584875242334009,"envelope_correlation":0.9916308957056967,"rms_gain_error_db":0.9839933623276966,"reference_mean_power":0.0015511172002359399,"output_mean_power":0.001945556959181763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":8.798155816797883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006408046711327789,"output_power":0.0008092350872559489,"reference_active":true,"delta_db":1.013490360152237},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.486060187757687e-05,"output_power":8.250606722940194e-05,"reference_active":true,"delta_db":0.422325718257237},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5748897665955194e-06,"output_power":3.2305279345452865e-06,"reference_active":true,"delta_db":-0.4398915390960463}],"active_band_count":3,"active_band_mae_db":0.6252358725018401}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":1.6544840621960162,"rms_envelope_nrmse":0.18230360198301793,"envelope_correlation":0.9828038679957154,"rms_gain_error_db":0.7553905089302693,"reference_mean_power":0.00020984803844209808,"output_mean_power":0.00024971461725324553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":2.0976407038326692e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011684508155422022,"output_power":0.00013727723553009716,"reference_active":true,"delta_db":0.6998808879412741},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.122844755814484e-05,"output_power":4.996377818528081e-05,"reference_active":true,"delta_db":0.8345828972312624},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.871758765973679e-06,"output_power":7.4197294062068374e-06,"reference_active":true,"delta_db":-0.7762165723395185}],"active_band_count":3,"active_band_mae_db":0.7702267858373517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":0.9838397830696607,"rms_envelope_nrmse":0.11665587062805159,"envelope_correlation":0.9960196959700852,"rms_gain_error_db":0.7553971740302692,"reference_mean_power":0.00021103616868638406,"output_mean_power":0.0002511288518971618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.885738469163612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011739951182146861,"output_power":0.00013886711803105656,"reference_active":true,"delta_db":0.7293313149320693},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.974678911762062e-05,"output_power":2.2291763090923567e-05,"reference_active":true,"delta_db":0.5264793065608506},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.379135569148538e-07,"output_power":5.869036914084616e-07,"reference_active":true,"delta_db":-0.36194990783686604}],"active_band_count":3,"active_band_mae_db":0.5392535097765953}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":1.7266301890260425,"rms_envelope_nrmse":0.23158586771783635,"envelope_correlation":0.973725049689496,"rms_gain_error_db":0.8371883608341729,"reference_mean_power":4.50355353361435e-05,"output_mean_power":5.461025011888866e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":3.018630059598247e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.579538283809341e-05,"output_power":1.9006692897787975e-05,"reference_active":true,"delta_db":0.8037640111109956},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.352602360298045e-06,"output_power":5.729730072442908e-06,"reference_active":true,"delta_db":1.1938516961068026},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8485554777348664e-06,"output_power":2.733319816115298e-06,"reference_active":true,"delta_db":1.698579610425415}],"active_band_count":3,"active_band_mae_db":1.2320651058810712}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":1.0514862543142443,"rms_envelope_nrmse":0.15654053591638648,"envelope_correlation":0.9907597277372304,"rms_gain_error_db":0.8371885801768174,"reference_mean_power":4.529062388817584e-05,"output_mean_power":5.491957417692166e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":2.7414916263445047e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5534436484200313e-05,"output_power":1.917031916333936e-05,"reference_active":true,"delta_db":0.9133383970290125},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0764100435071083e-06,"output_power":2.6474232525388154e-06,"reference_active":true,"delta_db":1.0551025810877879},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47015628096149e-07,"output_power":1.5573801083402535e-07,"reference_active":true,"delta_db":0.2503111966821744}],"active_band_count":3,"active_band_mae_db":0.7395840582663249}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"facodec","status":"measured","mos":{"status":"measured","value":4.265214920043945,"predicted_mos":4.265214920043945,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85400,"input_sample_rate":16000,"input_seconds":5.3375,"input_float32_sha256":"6a859847fa85646cf795deb86fe1194e4738357ead4224e8c2e3394132be5da8","internal_samples":85400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2056542740010627},"started_at":"2026-10-13T04:50:39.841885+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6704000447352882e-05,"native_input_samples":85280,"native_input_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","output_sha256":"1af563c2f3c044dda32adbc142ea28f300298ee1dd46415b57e452ab480cb0d3","output_samples":85400,"length_delta_samples":120,"analysis_s":1.4045938950002892,"synthesis_s":10.964712191999752,"total_s":12.36930608700004,"rtf":2.3206953258911898,"output_peak":0.5480087399482727,"output_rms":0.059034077810037125,"finished_at":"2026-10-13T04:50:52.217041+00:00","evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"1af563c2f3c044dda32adbc142ea28f300298ee1dd46415b57e452ab480cb0d3","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":85280,"output_samples":85400,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.577944332448677,"output_peak":0.5480087399482727},"common_prefix":{"samples":85280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9985948477751756,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.299989152227973,"spectral_convergence":0.19146347578627385,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.594880927646686,"spectral_convergence":0.19318290471415683,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.7632081539326565,"spectral_convergence":0.20885082899141574,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.0010436862726358877,"delta_db":-0.7300952998465043,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.001651848298405237,"delta_db":-0.4798459159754675,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.000527602782898328,"delta_db":-0.6965151759715806,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.00015014658716300827,"delta_db":-0.6494648020941782,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":9.96659503537346e-05,"delta_db":-0.24127412382245597,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.6976292651751873e-05,"delta_db":0.5180947122020751,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5525483383187103,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5856457571163264,"pitch":null,"output_rms_dbfs":-24.571837587639212,"output_peak":0.5480087399482727,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.552692744602439},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85280,"output_samples":85280,"intelligibility":{"estoi":{"value":0.9397568756991372,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06319611903854065,"rms_dbfs":-23.986191830522884,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.497017771838582,"mean":-1.0716221793283889e-05,"zero_energy":false},"output":{"rms":0.059075597196371586,"rms_dbfs":-24.571837587639212,"peak":0.5480087399482727,"peak_dbfs":-5.224250301807554,"crest_factor_db":19.34758728583166,"mean":2.082528784832838e-05,"zero_energy":false},"rms_difference_db":-0.5856457571163283},"full_scale":{"reference":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":533,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":116,"quiet_samples":18560,"reference_mean_power":8.461537774941273e-07,"output_mean_power":8.387180532981363e-07,"output_rms_dbfs":-60.763840087491104,"output_zero_energy":false,"output_peak":0.006277470383793116,"output_energy_fraction":5.230347334042789e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":526,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":223,"both_estimated_voiced_frames":214,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.963963963963964,"absolute_difference_cents_median":3.3919785610748727,"absolute_difference_cents_p95":16.903070850870353,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.0001748275799785616,"log_rms_mae_db":1.1603387790386366,"rms_envelope_nrmse":0.09823750892258426,"envelope_correlation":0.9946248722854709,"rms_gain_error_db":-0.5856469370286187,"reference_mean_power":0.003997498793754849,"output_mean_power":0.003493201577957856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.9383161701154944e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012108392795166332,"output_power":0.0010178920976495747,"reference_active":true,"delta_db":-0.753847582306687},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231683599850243,"output_power":0.00020656598006710408,"reference_active":true,"delta_db":-0.33573824013860587},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.764859371015312e-05,"output_power":5.971816529397282e-05,"reference_active":true,"delta_db":-0.5415231630507128}],"active_band_count":3,"active_band_mae_db":0.5437029951653352}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00017019095451879833,"log_rms_mae_db":0.9488617516812485,"rms_envelope_nrmse":0.08965270903042503,"envelope_correlation":0.9959295268625397,"rms_gain_error_db":-0.5856512096244457,"reference_mean_power":0.00402014109400192,"output_mean_power":0.0035129840237703037},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.6741447857886242e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011176954288226882,"output_power":0.0009432282369407268,"reference_active":true,"delta_db":-0.7370668121381065},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.194819962775495e-05,"output_power":8.645853443083875e-05,"reference_active":true,"delta_db":-0.26735360564942257},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3053605179616585e-06,"output_power":3.3672723865479964e-06,"reference_active":true,"delta_db":0.0805941370150591}],"active_band_count":3,"active_band_mae_db":0.3616715182675294}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.00014662911909855806,"log_rms_mae_db":1.5610662623369198,"rms_envelope_nrmse":0.09001356638221417,"envelope_correlation":0.9940754114263252,"rms_gain_error_db":-0.25984210546100467,"reference_mean_power":0.0015423797799867744,"output_mean_power":0.0014528042866645097},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.0119951767590758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006526269487739279,"output_power":0.0006025400594747971,"reference_active":true,"delta_db":-0.346790770575671},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016341376588393177,"output_power":0.00015321649789750455,"reference_active":true,"delta_db":-0.27983107106271604},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.788084498799748e-05,"output_power":4.56370616476126e-05,"reference_active":true,"delta_db":-0.20844131612415945}],"active_band_count":3,"active_band_mae_db":0.2783543859208488}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00013732417437068112,"log_rms_mae_db":1.1888612215912984,"rms_envelope_nrmse":0.0746902369215675,"envelope_correlation":0.9957629164452586,"rms_gain_error_db":-0.2598465746937935,"reference_mean_power":0.0015511172002359399,"output_mean_power":0.001461032767587403},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":8.798155816797883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006408046711327789,"output_power":0.0005900685814257234,"reference_active":true,"delta_db":-0.3582317789797959},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.486060187757687e-05,"output_power":6.979884665215882e-05,"reference_active":true,"delta_db":-0.3040506805835115},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5748897665955194e-06,"output_power":3.466852968919791e-06,"reference_active":true,"delta_db":-0.1332723128254277}],"active_band_count":3,"active_band_mae_db":0.2651849241295784}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":1.5331924328164717,"rms_envelope_nrmse":0.12976297028289335,"envelope_correlation":0.9886417757338717,"rms_gain_error_db":-0.4440127651746873,"reference_mean_power":0.00020984803844209808,"output_mean_power":0.0001894539382183381},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":2.0976407038326692e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011684508155422022,"output_power":0.00010233370989034986,"reference_active":true,"delta_db":-0.5759171720415978},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.122844755814484e-05,"output_power":3.76471133545972e-05,"reference_active":true,"delta_db":-0.3946530023941085},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.871758765973679e-06,"output_power":7.096677129213413e-06,"reference_active":true,"delta_db":-0.969546773065134}],"active_band_count":3,"active_band_mae_db":0.6467056491669467}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":1.1564470311308181,"rms_envelope_nrmse":0.09608543319998854,"envelope_correlation":0.9941548613638866,"rms_gain_error_db":-0.4440288472809251,"reference_mean_power":0.00021103616868638406,"output_mean_power":0.00019052589439281972},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.885738469163612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011739951182146861,"output_power":0.00010334104604325508,"reference_active":true,"delta_db":-0.5539343773639329},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.974678911762062e-05,"output_power":1.7879082917088383e-05,"reference_active":true,"delta_db":-0.4315124972802067},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.379135569148538e-07,"output_power":4.835006172084969e-07,"reference_active":true,"delta_db":-1.2036479904427808}],"active_band_count":3,"active_band_mae_db":0.7296982883623068}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":1.798805854788846,"rms_envelope_nrmse":0.208877118871315,"envelope_correlation":0.9708180765224215,"rms_gain_error_db":-0.132977286109762,"reference_mean_power":4.50355353361435e-05,"output_mean_power":4.3677482757938594e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":3.018630059598247e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.579538283809341e-05,"output_power":1.732813677050441e-05,"reference_active":true,"delta_db":0.4022171065655906},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.352602360298045e-06,"output_power":4.534150171830108e-06,"reference_active":true,"delta_db":0.17746906708714733},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8485554777348664e-06,"output_power":1.2319163437603765e-06,"reference_active":true,"delta_db":-1.7625127187010816}],"active_band_count":3,"active_band_mae_db":0.7807329641179398}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":1.4833781473787873,"rms_envelope_nrmse":0.15530924386113423,"envelope_correlation":0.9827624477553797,"rms_gain_error_db":-0.13299662577558488,"reference_mean_power":4.529062388817584e-05,"output_mean_power":4.392468347820475e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":2.7414916263445047e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5534436484200313e-05,"output_power":1.680469705795874e-05,"reference_active":true,"delta_db":0.3413518402275416},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0764100435071083e-06,"output_power":2.1275525935878083e-06,"reference_active":true,"delta_db":0.10567183872630925},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47015628096149e-07,"output_power":1.0970010823173531e-07,"reference_active":true,"delta_db":-1.2715644763861766}],"active_band_count":3,"active_band_mae_db":0.5728627184466758}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.9408116340637207,"predicted_mos":3.9408116340637207,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":235520,"input_sample_rate":44100,"input_seconds":5.340589569160998,"input_float32_sha256":"604281e47fc70e906ae33951d40a55cbc92717790f959b54aba1e93e2df8bc73","internal_samples":85450,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.2468688460030535},"started_at":"2026-10-13T07:11:31.304386+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007698729999901843,"native_input_samples":235053,"native_input_sha256":"68d301652f465d7ec294e74410442a04acf68ca8518719cf18b20db1f6003696","output_sha256":"12f48b784f002d6df3e6dfde201336ac98c374e08a66906637eff90f114cd464","output_samples":235520,"length_delta_samples":467,"analysis_s":0.1275853110000753,"synthesis_s":30.879962688999512,"total_s":31.007547999999588,"rtf":5.817551219512118,"output_peak":0.42718663811683655,"output_rms":0.05812521900574588,"finished_at":"2026-10-13T07:12:02.333571+00:00","evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"12f48b784f002d6df3e6dfde201336ac98c374e08a66906637eff90f114cd464","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":85280,"output_samples":85450,"length_delta_samples":170,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.708293128462614,"output_peak":0.4255714866206874},"common_prefix":{"samples":85280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980105324751316,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.31036208624725,"spectral_convergence":0.4528767200805609,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.023713025050089,"spectral_convergence":0.30472070971467186,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.617130129840624,"spectral_convergence":0.361224735287473,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.000745570701759063,"delta_db":-2.190906678409625,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.0019161095757640016,"delta_db":0.16465589371294015,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.000542932253034185,"delta_db":-0.5721295341040318,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.00010658796993779066,"delta_db":-2.1375375456334584,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":6.574469479029757e-05,"delta_db":-2.0481351109535533,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.177362352981397e-05,"delta_db":-1.0712323732289466,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3640995226737591,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7134528248718748,"pitch":null,"output_rms_dbfs":-24.69964465539476,"output_peak":0.4255714866206874,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.983735080379321},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85280,"output_samples":85280,"intelligibility":{"estoi":{"value":0.8364181441697501,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06319611903854065,"rms_dbfs":-23.986191830522884,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.497017771838582,"mean":-1.0716221793283889e-05,"zero_energy":false},"output":{"rms":0.05821270324263555,"rms_dbfs":-24.69964465539476,"peak":0.4255714866206874,"peak_dbfs":-7.42054955120214,"crest_factor_db":17.27909510419262,"mean":-8.868927834256153e-05,"zero_energy":false},"rms_difference_db":-0.7134528248718759},"full_scale":{"reference":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":533,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":116,"quiet_samples":18560,"reference_mean_power":8.461537774941273e-07,"output_mean_power":7.914412922195925e-07,"output_rms_dbfs":-61.01581294839143,"output_zero_energy":false,"output_peak":0.01366296144110968,"output_energy_fraction":5.082927917645699e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":526,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":212,"both_estimated_voiced_frames":197,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.8873873873873874,"absolute_difference_cents_median":9.672690298487629,"absolute_difference_cents_p95":37.974311865277976,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.0001748275799785616,"log_rms_mae_db":1.5085423106096385,"rms_envelope_nrmse":0.11629345813106497,"envelope_correlation":0.9928041295895877,"rms_gain_error_db":-0.7134526877431677,"reference_mean_power":0.003997498793754849,"output_mean_power":0.0033919002553342566},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.9383161701154944e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012108392795166332,"output_power":0.00103760838344305,"reference_active":true,"delta_db":-0.6705302897822941},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231683599850243,"output_power":0.00018171465319360077,"reference_active":true,"delta_db":-0.8924267233813514},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.764859371015312e-05,"output_power":6.220331358007554e-05,"reference_active":true,"delta_db":-0.3644525260309022}],"active_band_count":3,"active_band_mae_db":0.6424698463981826}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00017019095451879833,"log_rms_mae_db":1.2078120672905752,"rms_envelope_nrmse":0.09417608552326684,"envelope_correlation":0.997237399133254,"rms_gain_error_db":-0.7134520136813093,"reference_mean_power":0.00402014109400192,"output_mean_power":0.0034111129041334435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.6741447857886242e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011176954288226882,"output_power":0.0009624792072298344,"reference_active":true,"delta_db":-0.6493211867851894},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.194819962775495e-05,"output_power":7.628946179654396e-05,"reference_active":true,"delta_db":-0.8107867898119709},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3053605179616585e-06,"output_power":3.5488235134495054e-06,"reference_active":true,"delta_db":0.3086556677957393}],"active_band_count":3,"active_band_mae_db":0.5895878814642999}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.00014662911909855806,"log_rms_mae_db":2.5213182931915803,"rms_envelope_nrmse":0.14460997279610016,"envelope_correlation":0.9834892566009655,"rms_gain_error_db":0.04935195852095871,"reference_mean_power":0.0015423797799867744,"output_mean_power":0.0015600068990815627},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.0119951767590758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006526269487739279,"output_power":0.0006987807553073013,"reference_active":true,"delta_db":0.2967593279341867},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016341376588393177,"output_power":0.00015293550989991243,"reference_active":true,"delta_db":-0.28780303061172374},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.788084498799748e-05,"output_power":4.6828963806495055e-05,"reference_active":true,"delta_db":-0.09647257940983103}],"active_band_count":3,"active_band_mae_db":0.22701164598524715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00013732417437068112,"log_rms_mae_db":1.839105551061687,"rms_envelope_nrmse":0.10412580202760427,"envelope_correlation":0.9906290746689266,"rms_gain_error_db":0.04934733233905837,"reference_mean_power":0.0015511172002359399,"output_mean_power":0.0015688425039583026},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":8.798155816797883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006408046711327789,"output_power":0.0006798091024564319,"reference_active":true,"delta_db":0.2566130650071081},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.486060187757687e-05,"output_power":7.165085495111435e-05,"reference_active":true,"delta_db":-0.19031937664911036},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5748897665955194e-06,"output_power":3.5691036314603154e-06,"reference_active":true,"delta_db":-0.007034964816521817}],"active_band_count":3,"active_band_mae_db":0.15132246882424677}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":3.284852159526905,"rms_envelope_nrmse":0.2854401220922851,"envelope_correlation":0.9607901106967026,"rms_gain_error_db":-1.9361038912095043,"reference_mean_power":0.00020984803844209808,"output_mean_power":0.0001343675890988238},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":2.0976407038326692e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011684508155422022,"output_power":7.112560146200456e-05,"reference_active":true,"delta_db":-2.1558448407776196},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.122844755814484e-05,"output_power":2.617982461352368e-05,"reference_active":true,"delta_db":-1.9723024915624918},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.871758765973679e-06,"output_power":6.4804142812960146e-06,"reference_active":true,"delta_db":-1.3640695387319148}],"active_band_count":3,"active_band_mae_db":1.8307389570240087}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":2.802603192274576,"rms_envelope_nrmse":0.257956023611079,"envelope_correlation":0.9725067506035975,"rms_gain_error_db":-1.9360987101717058,"reference_mean_power":0.00021103616868638406,"output_mean_power":0.00013512852080089507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.885738469163612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011739951182146861,"output_power":7.024935687219129e-05,"reference_active":true,"delta_db":-2.230239383423038},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.974678911762062e-05,"output_power":1.3093597238155236e-05,"reference_active":true,"delta_db":-1.7843751042194547},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.379135569148538e-07,"output_power":4.538600439876825e-07,"reference_active":true,"delta_db":-1.4783988097171958}],"active_band_count":3,"active_band_mae_db":1.8310044324532295}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":3.6415924595958766,"rms_envelope_nrmse":0.27618039018934226,"envelope_correlation":0.9719989706475451,"rms_gain_error_db":-2.0219461357777906,"reference_mean_power":4.50355353361435e-05,"output_mean_power":2.8272272362254362e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":3.018630059598247e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.579538283809341e-05,"output_power":1.1457246401738067e-05,"reference_active":true,"delta_db":-1.3944990329873117},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.352602360298045e-06,"output_power":2.6160607029237794e-06,"reference_active":true,"delta_db":-2.21101176113987},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8485554777348664e-06,"output_power":7.468513520067459e-07,"reference_active":true,"delta_db":-3.935983173648837}],"active_band_count":3,"active_band_mae_db":2.513831322592006}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":3.1369029870955205,"rms_envelope_nrmse":0.24457775680938,"envelope_correlation":0.9847956064015306,"rms_gain_error_db":-2.0219439398134376,"reference_mean_power":4.529062388817584e-05,"output_mean_power":2.843242545961368e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":2.7414916263445047e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5534436484200313e-05,"output_power":1.104371352835327e-05,"reference_active":true,"delta_db":-1.4818037112509905},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0764100435071083e-06,"output_power":1.1996348127508866e-06,"reference_active":true,"delta_db":-2.382640606154283},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47015628096149e-07,"output_power":5.579761414946792e-08,"reference_active":true,"delta_db":-4.207478743607248}],"active_band_count":3,"active_band_mae_db":2.690641020337507}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"golf","status":"measured","mos":{"status":"measured","value":3.3430261611938477,"predicted_mos":3.3430261611938477,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":127680,"input_sample_rate":24000,"input_seconds":5.32,"input_float32_sha256":"e02b548f1f6b1cdbf5b02cb4b49f231bddaf53f261d0bbf68e1e77419abee0bc","internal_samples":85120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0140169659989624},"started_at":"2026-10-13T05:02:22.527810+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004365422999399016,"native_input_samples":127920,"native_input_sha256":"472122e94fe4ec6b3d742d5271ed9b4796453f060acce3f14f934acb60ae53e4","output_sha256":"5eec6378976372c58370edec62adf9762ccad288b8235bbc9b19d9d2d1f7e95c","output_samples":127680,"length_delta_samples":-240,"analysis_s":0.6247619919995486,"synthesis_s":0.11980331799986743,"total_s":0.7445653099994161,"rtf":0.1396933039398529,"output_peak":0.4763035774230957,"output_rms":0.05816960603560275,"finished_at":"2026-10-13T05:02:23.286141+00:00","analysis_stages":{"external_f0_s":0.05538528200031578},"evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"5eec6378976372c58370edec62adf9762ccad288b8235bbc9b19d9d2d1f7e95c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":85280,"output_samples":85120,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.702233585002237,"output_peak":0.4700869246905239},"common_prefix":{"samples":85120,"sample_rate":16000,"reference_coverage":0.99812382739212,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85120,"output_samples":85120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.47704183696066,"spectral_convergence":0.4711208820680081,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.598335086339854,"spectral_convergence":0.21750944228165664,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.653985420592249,"spectral_convergence":0.23087683711239207,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012370501569505716,"output_power":0.0009875521181717767,"delta_db":-0.9782728387744619,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018483067835241731,"output_power":0.0016455592589663362,"delta_db":-0.5046053094899243,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000620551172610952,"output_power":0.0005115824903021675,"delta_db":-0.8386192894288226,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001746937629312122,"output_power":0.0001557202851505574,"delta_db":-0.49932209385536597,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010555767463071633,"output_power":7.698044608382815e-05,"delta_db":-1.3710939101703938,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.509556563515496e-05,"output_power":9.304728660273235e-06,"delta_db":-2.1014567782383646,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0488950366595555,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7241960097186472,"pitch":null,"output_rms_dbfs":-24.702233585002237,"output_peak":0.4700869246905239,"reference_rms_dbfs":-23.978037575283587,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.909787447964255},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85120,"output_samples":85120,"intelligibility":{"estoi":{"value":0.9174559152393951,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06325547499057121,"rms_dbfs":-23.978037575283587,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.488863516599285,"mean":-1.0658923844645793e-05,"zero_energy":false},"output":{"rms":0.05819535486046063,"rms_dbfs":-24.702233585002237,"peak":0.4700869246905239,"peak_dbfs":-6.556436567969779,"crest_factor_db":18.145797017032457,"mean":0.0005259402495629757,"zero_energy":false},"rms_difference_db":-0.72419600971865},"full_scale":{"reference":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":532,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":115,"quiet_samples":18400,"reference_mean_power":8.470669854432344e-07,"output_mean_power":6.070654231830356e-07,"output_rms_dbfs":-62.167645026705394,"output_zero_energy":false,"output_peak":0.004977884932915385,"output_energy_fraction":3.874762285074712e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":525,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":242,"both_estimated_voiced_frames":220,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.990990990990991,"absolute_difference_cents_median":3.0956558820861897,"absolute_difference_cents_p95":18.694372564059076,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.0001748275799785616,"log_rms_mae_db":2.093884087769935,"rms_envelope_nrmse":0.13372944226221795,"envelope_correlation":0.9889466509865436,"rms_gain_error_db":-0.7241956965438343,"reference_mean_power":0.004005018288349805,"output_mean_power":0.00338988475541652},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.9399313060354302e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012123978866571611,"output_power":0.0010994279474214027,"reference_active":true,"delta_db":-0.4247839815206964},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022283900889466174,"output_power":0.0001817940795968681,"reference_active":true,"delta_db":-0.8841148256067692},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.74555247809626e-05,"output_power":5.253300390346912e-05,"reference_active":true,"delta_db":-1.0858529027985266}],"active_band_count":3,"active_band_mae_db":0.7982505699753307}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00017019095451879833,"log_rms_mae_db":1.7169079595653323,"rms_envelope_nrmse":0.1078699515557917,"envelope_correlation":0.994835902314832,"rms_gain_error_db":-0.7241935782642805,"reference_mean_power":0.004027746628211665,"output_mean_power":0.003409123896622724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.6755640842688293e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001119175106124645,"output_power":0.0010229189963501053,"reference_active":true,"delta_db":-0.3905679770641186},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.144893636353236e-05,"output_power":7.55162894451143e-05,"reference_active":true,"delta_db":-0.8313801619340635},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3058169698668944e-06,"output_power":2.9502972117483474e-06,"reference_active":true,"delta_db":-0.49413035884643286}],"active_band_count":3,"active_band_mae_db":0.572026165948205}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":0.00014662911909855806,"log_rms_mae_db":2.2990374547474284,"rms_envelope_nrmse":0.14421019491631368,"envelope_correlation":0.9849741448338171,"rms_gain_error_db":-0.46240617686991253,"reference_mean_power":0.0015452815052398637,"output_mean_power":0.0013892070834469346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":1.0132628042407608e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006540898169463482,"output_power":0.0005854491318372943,"reference_active":true,"delta_db":-0.48148221644360717},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016316704352439525,"output_power":0.00012010943723116325,"reference_active":true,"delta_db":-1.3305531238383352},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7544092760259446e-05,"output_power":3.5381649208195475e-05,"reference_active":true,"delta_db":-1.2831849221064637}],"active_band_count":3,"active_band_mae_db":1.0317400874628022}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013732417437068112,"log_rms_mae_db":1.6535231532296157,"rms_envelope_nrmse":0.10353497126403759,"envelope_correlation":0.992861401734,"rms_gain_error_db":-0.46240387027604996,"reference_mean_power":0.001554051956353719,"output_mean_power":0.0013970924553943333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":8.810304627829802e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006421197819925865,"output_power":0.0005686289358110884,"reference_active":true,"delta_db":-0.5278709372095327},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.443609419059771e-05,"output_power":5.4837646377395825e-05,"reference_active":true,"delta_db":-1.327047700933004},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5652236768482943e-06,"output_power":2.4150119032674806e-06,"reference_active":true,"delta_db":-1.691675063629736}],"active_band_count":3,"active_band_mae_db":1.1821979005907577}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":2.157278861203866,"rms_envelope_nrmse":0.18046693347659606,"envelope_correlation":0.978155524398174,"rms_gain_error_db":-0.5810665170038501,"reference_mean_power":0.00021024277817020606,"output_mean_power":0.00018391384963868022},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":2.1015276125960312e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011722939144352818,"output_power":0.00011287755184549387,"reference_active":true,"delta_db":-0.16428928884831343},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.117362060522162e-05,"output_power":3.79153194052829e-05,"reference_active":true,"delta_db":-0.3580433948180575},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.875268791396813e-06,"output_power":6.8645691844420965e-06,"reference_active":true,"delta_db":-1.1156822831434587}],"active_band_count":3,"active_band_mae_db":0.5460049889366099}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":1.7341477863847323,"rms_envelope_nrmse":0.13776701385457693,"envelope_correlation":0.9880867002577807,"rms_gain_error_db":-0.5810612153044588,"reference_mean_power":0.0002114354162877258,"output_mean_power":0.00018495735819466772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":1.8892079488915511e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011775089347720298,"output_power":0.00011422072816090848,"reference_active":true,"delta_db":-0.13219286699604807},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9670457264457344e-05,"output_power":1.8392191803110886e-05,"reference_active":true,"delta_db":-0.2918096842865567},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.40144136964917e-07,"output_power":5.508502500888752e-07,"reference_active":true,"delta_db":-0.6524422123215289}],"active_band_count":3,"active_band_mae_db":0.35881492120137787}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1063,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":3.150270066497733,"rms_envelope_nrmse":0.22886962122999008,"envelope_correlation":0.9768816532053517,"rms_gain_error_db":-1.3793957931125413,"reference_mean_power":4.51202630297885e-05,"output_mean_power":3.284218501947897e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1063,"frequency_resolution_hz":0.18814675446848542,"power_floor":3.026637778029664e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5852832875853344e-05,"output_power":1.2671658236550553e-05,"reference_active":true,"delta_db":-0.9727343003868791},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.358783740305791e-06,"output_power":3.145897919777887e-06,"reference_active":true,"delta_db":-1.4162069593209325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8405763201390477e-06,"output_power":1.4577659613016908e-06,"reference_active":true,"delta_db":-1.012660250526247}],"active_band_count":3,"active_band_mae_db":1.1338671700780196}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85120,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":2.9592773086037685,"rms_envelope_nrmse":0.176998863750041,"envelope_correlation":0.9926614409167869,"rms_gain_error_db":-1.3793931050465933,"reference_mean_power":4.5376316724610616e-05,"output_mean_power":3.302858209810193e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.748859856190735e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5588632633716037e-05,"output_power":1.2455287139464726e-05,"reference_active":true,"delta_db":-0.9745427840971785},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.066785560572957e-06,"output_power":1.438723912026182e-06,"reference_active":true,"delta_db":-1.5731795695138062},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4612404496152138e-07,"output_power":1.1889784686105545e-07,"reference_active":true,"delta_db":-0.8954769572850575}],"active_band_count":3,"active_band_mae_db":1.147733103632014}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.979693651199341,"predicted_mos":3.979693651199341,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":128040,"input_sample_rate":24000,"input_seconds":5.335,"input_float32_sha256":"9d2b95879f979364b1e4817658301019b1d6dfdffebdd346506a63e9ffc80ba9","internal_samples":85360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0833322849994147},"started_at":"2026-10-13T05:09:19.837596+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002439721998598543,"native_input_samples":127920,"native_input_sha256":"472122e94fe4ec6b3d742d5271ed9b4796453f060acce3f14f934acb60ae53e4","output_sha256":"23d9ea5eaae2a67d5933e6792502ac67123891c5e21d3383816b64bbe33f967b","output_samples":128040,"length_delta_samples":120,"analysis_s":1.872719662000236,"synthesis_s":10.472323775000405,"total_s":12.34504343700064,"rtf":2.316143233958844,"output_peak":0.5731388926506042,"output_rms":0.0632219193405037,"finished_at":"2026-10-13T05:09:32.191459+00:00","analysis_stages":{"external_f0_s":1.2165052289983578},"evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"23d9ea5eaae2a67d5933e6792502ac67123891c5e21d3383816b64bbe33f967b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":85280,"output_samples":85360,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.97915180739539,"output_peak":0.5738139323358622},"common_prefix":{"samples":85280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990627928772259,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.253550383983031,"spectral_convergence":0.49356635232894075,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.354079138348568,"spectral_convergence":0.2460436787904569,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.225403691253531,"spectral_convergence":0.21573040039226143,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.0013634051449722597,"delta_db":0.43045427213051146,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.00177755479662346,"delta_db":-0.16131754309975654,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.0005946521652958464,"delta_db":-0.17695590600458416,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.000163501434656834,"delta_db":-0.2794037756257809,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":9.324367452682254e-05,"delta_db":-0.5305484489387303,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.1623928550407468e-05,"delta_db":-1.1268045034801808,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4509140748799241,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.011112067036444954,"pitch":null,"output_rms_dbfs":-23.97507976348644,"output_peak":0.5738139323358622,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.611011071195044},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85280,"output_samples":85280,"intelligibility":{"estoi":{"value":0.9164265889266849,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06319611903854065,"rms_dbfs":-23.986191830522884,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.497017771838582,"mean":-1.0716221793283889e-05,"zero_energy":false},"output":{"rms":0.06327701908770379,"rms_dbfs":-23.97507976348644,"peak":0.5738139323358622,"peak_dbfs":-4.824578223817566,"crest_factor_db":19.150501539668877,"mean":0.0001383513777417575,"zero_energy":false},"rms_difference_db":0.011112067036442852},"full_scale":{"reference":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":533,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":116,"quiet_samples":18560,"reference_mean_power":8.461537774941273e-07,"output_mean_power":2.744168686578059e-06,"output_rms_dbfs":-55.61589195662697,"output_zero_energy":false,"output_peak":0.078993375735525,"output_energy_fraction":0.00014915903359239034,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":526,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":233,"both_estimated_voiced_frames":215,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9684684684684685,"absolute_difference_cents_median":4.691212872694847,"absolute_difference_cents_p95":19.130963680053814,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.0001748275799785616,"log_rms_mae_db":2.2232859398401996,"rms_envelope_nrmse":0.1542760062747009,"envelope_correlation":0.977866286972907,"rms_gain_error_db":0.01111211389642774,"reference_mean_power":0.003997498793754849,"output_mean_power":0.004007740125594901},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.9383161701154944e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012108392795166332,"output_power":0.0012474728317874127,"reference_active":true,"delta_db":0.12944595085608762},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231683599850243,"output_power":0.00020672845758597995,"reference_active":true,"delta_db":-0.3323235755756379},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.764859371015312e-05,"output_power":6.738132124058619e-05,"reference_active":true,"delta_db":-0.017192502909980614}],"active_band_count":3,"active_band_mae_db":0.1596540097805687}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00017019095451879833,"log_rms_mae_db":1.5991426678114113,"rms_envelope_nrmse":0.0830603074183656,"envelope_correlation":0.9930293688213302,"rms_gain_error_db":0.011116718848025323,"reference_mean_power":0.00402014109400192,"output_mean_power":0.004030444707538451},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.6741447857886242e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011176954288226882,"output_power":0.0011490611909270518,"reference_active":true,"delta_db":0.12019682030398034},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.194819962775495e-05,"output_power":8.582306004958249e-05,"reference_active":true,"delta_db":-0.2993923466017598},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3053605179616585e-06,"output_power":3.5423879109195916e-06,"reference_active":true,"delta_db":0.3007728188252428}],"active_band_count":3,"active_band_mae_db":0.24012066191032763}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.00014662911909855806,"log_rms_mae_db":2.7049841390148477,"rms_envelope_nrmse":0.21452708898970865,"envelope_correlation":0.9631674063332152,"rms_gain_error_db":-0.16496832786287632,"reference_mean_power":0.0015423797799867744,"output_mean_power":0.0014848907107390791},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.0119951767590758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006526269487739279,"output_power":0.0006262914709948552,"reference_active":true,"delta_db":-0.1788850564099985},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016341376588393177,"output_power":0.00015198762500483404,"reference_active":true,"delta_db":-0.3148040979833641},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.788084498799748e-05,"output_power":5.140849282928611e-05,"reference_active":true,"delta_db":0.3087306549469886}],"active_band_count":3,"active_band_mae_db":0.26747326978011704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00013732417437068112,"log_rms_mae_db":2.029003031078175,"rms_envelope_nrmse":0.12091809058741233,"envelope_correlation":0.9873705792726266,"rms_gain_error_db":-0.16496637067749412,"reference_mean_power":0.0015511172002359399,"output_mean_power":0.0014933031343892559},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":8.798155816797883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006408046711327789,"output_power":0.00061853277782663,"reference_active":true,"delta_db":-0.1536294978295783},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.486060187757687e-05,"output_power":7.049060021160229e-05,"reference_active":true,"delta_db":-0.26122106007004064},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5748897665955194e-06,"output_power":4.077200034070171e-06,"reference_active":true,"delta_db":0.5709936449974484}],"active_band_count":3,"active_band_mae_db":0.32861473429902244}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":2.3933022357481355,"rms_envelope_nrmse":0.2657721337437694,"envelope_correlation":0.9459019433605863,"rms_gain_error_db":-0.21216681108458563,"reference_mean_power":0.00020984803844209808,"output_mean_power":0.00019984267433409864},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":2.0976407038326692e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011684508155422022,"output_power":0.00010874080517214722,"reference_active":true,"delta_db":-0.31217891622099403},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.122844755814484e-05,"output_power":3.5547106136230735e-05,"reference_active":true,"delta_db":-0.6439273100163476},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.871758765973679e-06,"output_power":9.531434101883377e-06,"reference_active":true,"delta_db":0.31148525306750546}],"active_band_count":3,"active_band_mae_db":0.42253049310161567}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":1.4257735396772886,"rms_envelope_nrmse":0.14200165387465938,"envelope_correlation":0.9834624775829786,"rms_gain_error_db":-0.21216083087470916,"reference_mean_power":0.00021103616868638406,"output_mean_power":0.00020097443234659608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.885738469163612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011739951182146861,"output_power":0.00010851252818184316,"reference_active":true,"delta_db":-0.3418640898291031},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.974678911762062e-05,"output_power":1.7832213706083955e-05,"reference_active":true,"delta_db":-0.44291228016263534},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.379135569148538e-07,"output_power":7.426424979017325e-07,"reference_active":true,"delta_db":0.6601796635257079}],"active_band_count":3,"active_band_mae_db":0.4816520111724822}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":2.436521639534982,"rms_envelope_nrmse":0.29604412120218815,"envelope_correlation":0.9416285163290233,"rms_gain_error_db":-0.8869922560795604,"reference_mean_power":4.50355353361435e-05,"output_mean_power":3.6716062794563e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":3.018630059598247e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.579538283809341e-05,"output_power":1.2303225687252763e-05,"reference_active":true,"delta_db":-1.0851116584446605},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.352602360298045e-06,"output_power":3.2424919790826527e-06,"reference_active":true,"delta_db":-1.2787008277971335},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8485554777348664e-06,"output_power":1.555552310400251e-06,"reference_active":true,"delta_db":-0.7494786866412197}],"active_band_count":3,"active_band_mae_db":1.0377637242943378}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":1.5007593460552102,"rms_envelope_nrmse":0.19515350737428133,"envelope_correlation":0.9769878258939698,"rms_gain_error_db":-0.8869959009255115,"reference_mean_power":4.529062388817584e-05,"output_mean_power":3.692399752056781e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":2.7414916263445047e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5534436484200313e-05,"output_power":1.1944420526995404e-05,"reference_active":true,"delta_db":-1.141304185452906},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0764100435071083e-06,"output_power":1.574677328414091e-06,"reference_active":true,"delta_db":-1.2012154614445647},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47015628096149e-07,"output_power":1.1126116659390427e-07,"reference_active":true,"delta_db":-1.210198944082166}],"active_band_count":3,"active_band_mae_db":1.1842395303265458}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.269160032272339,"predicted_mos":3.269160032272339,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":128160,"input_sample_rate":24000,"input_seconds":5.34,"input_float32_sha256":"d2822ad5fed1c5b3d3e847e0a0ad18f05a0498cb869f4c54f04d4302bdfeac01","internal_samples":85440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0216986479990737},"started_at":"2026-10-13T05:10:30.302006+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002731375001530978,"native_input_samples":127920,"native_input_sha256":"472122e94fe4ec6b3d742d5271ed9b4796453f060acce3f14f934acb60ae53e4","output_sha256":"c971f83a57900b150a17c5f4356ce745a3a7c275f9f2216e9804cf1e97f53184","output_samples":128160,"length_delta_samples":240,"analysis_s":0.6700420329998451,"synthesis_s":0.11973488200055726,"total_s":0.7897769150004024,"rtf":0.14817578142596669,"output_peak":0.5907711386680603,"output_rms":0.062142948042863785,"finished_at":"2026-10-13T05:10:31.105175+00:00","analysis_stages":{"external_f0_s":0.06768731199917966},"evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"c971f83a57900b150a17c5f4356ce745a3a7c275f9f2216e9804cf1e97f53184","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":85280,"output_samples":85440,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.128513490121524,"output_peak":0.5701091478225462},"common_prefix":{"samples":85280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.99812734082397,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.734195867132786,"spectral_convergence":0.13655469279866353,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.422840986559184,"spectral_convergence":0.13572157156454243,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.9271528822672686,"spectral_convergence":0.14702170402161818,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.0012188458229016373,"delta_db":-0.056307280159958305,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.0018162063077264423,"delta_db":-0.06789572242300672,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.0005747149755767448,"delta_db":-0.3250608134584021,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.00016033812894601713,"delta_db":-0.36425134004901294,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":8.954480731753395e-05,"delta_db":-0.7063381838422851,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.2593761050180744e-05,"delta_db":-0.7787793325309217,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38310544541059777,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13418123479374588,"pitch":null,"output_rms_dbfs":-24.120373065316635,"output_peak":0.5701091478225462,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.694729911986413},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85280,"output_samples":85280,"intelligibility":{"estoi":{"value":0.9651969594164933,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06319611903854065,"rms_dbfs":-23.986191830522884,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.497017771838582,"mean":-1.0716221793283889e-05,"zero_energy":false},"output":{"rms":0.06222735574904783,"rms_dbfs":-24.120373065316635,"peak":0.5701091478225462,"peak_dbfs":-4.880839807281704,"crest_factor_db":19.23953325803493,"mean":9.823800019632932e-06,"zero_energy":false},"rms_difference_db":-0.13418123479375055},"full_scale":{"reference":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":533,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":116,"quiet_samples":18560,"reference_mean_power":8.461537774941273e-07,"output_mean_power":1.0069036331976306e-06,"output_rms_dbfs":-59.97012092081108,"output_zero_energy":false,"output_peak":0.014145870563681943,"output_energy_fraction":5.6592124076737406e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":526,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":229,"both_estimated_voiced_frames":219,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9864864864864865,"absolute_difference_cents_median":1.036947639784418,"absolute_difference_cents_p95":7.332330280738256,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.0001748275799785616,"log_rms_mae_db":0.775267774231734,"rms_envelope_nrmse":0.05626334600109577,"envelope_correlation":0.9972282002849177,"rms_gain_error_db":-0.13418148017428477,"reference_mean_power":0.003997498793754849,"output_mean_power":0.0038758788472304847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.9383161701154944e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012108392795166332,"output_power":0.001185383015373975,"reference_active":true,"delta_db":-0.09227800778361926},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231683599850243,"output_power":0.00019824247010893345,"reference_active":true,"delta_db":-0.5143592151363034},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.764859371015312e-05,"output_power":4.518087622265334e-05,"reference_active":true,"delta_db":-1.7530412361742314}],"active_band_count":3,"active_band_mae_db":0.786559486364718}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00017019095451879833,"log_rms_mae_db":0.4768603306402708,"rms_envelope_nrmse":0.03374018732868115,"envelope_correlation":0.9990859208149522,"rms_gain_error_db":-0.13418068864699492,"reference_mean_power":0.00402014109400192,"output_mean_power":0.003897832988292988},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.6741447857886242e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011176954288226882,"output_power":0.00109920418930199,"reference_active":true,"delta_db":-0.07245099839288194},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.194819962775495e-05,"output_power":7.95668942675253e-05,"reference_active":true,"delta_db":-0.6281082345933103},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3053605179616585e-06,"output_power":2.3619546634571877e-06,"reference_active":true,"delta_db":-1.4594727791381723}],"active_band_count":3,"active_band_mae_db":0.7200106707081216}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.00014662911909855806,"log_rms_mae_db":1.096488768356836,"rms_envelope_nrmse":0.07607355895838322,"envelope_correlation":0.9955982079820302,"rms_gain_error_db":-0.16029483271822045,"reference_mean_power":0.0015423797799867744,"output_mean_power":0.001486489479561947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.0119951767590758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006526269487739279,"output_power":0.0006231423654855491,"reference_active":true,"delta_db":-0.2007772447228072},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016341376588393177,"output_power":0.0001525174356074729,"reference_active":true,"delta_db":-0.29969143923781755},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.788084498799748e-05,"output_power":3.264489584286583e-05,"reference_active":true,"delta_db":-1.6634651883336526}],"active_band_count":3,"active_band_mae_db":0.7213112907647591}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00013732417437068112,"log_rms_mae_db":0.6256931567835498,"rms_envelope_nrmse":0.04539940734343066,"envelope_correlation":0.9984647381794403,"rms_gain_error_db":-0.16029567784097823,"reference_mean_power":0.0015511172002359399,"output_mean_power":0.0014949099961960863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":8.798155816797883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006408046711327789,"output_power":0.0006094125685794569,"reference_active":true,"delta_db":-0.21814261961266523},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.486060187757687e-05,"output_power":6.896320941504135e-05,"reference_active":true,"delta_db":-0.35635849994108565},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5748897665955194e-06,"output_power":2.5530275690534007e-06,"reference_active":true,"delta_db":-1.4620715007926117}],"active_band_count":3,"active_band_mae_db":0.6788575401154541}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":1.1238461844566627,"rms_envelope_nrmse":0.12035315122461715,"envelope_correlation":0.9906079989319401,"rms_gain_error_db":-0.4540519869692538,"reference_mean_power":0.00020984803844209808,"output_mean_power":0.00018901649920743653},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":2.0976407038326692e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011684508155422022,"output_power":0.00010632626729070838,"reference_active":true,"delta_db":-0.40969868313071733},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.122844755814484e-05,"output_power":3.3725245968621705e-05,"reference_active":true,"delta_db":-0.8724185611297491},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.871758765973679e-06,"output_power":4.485849208622229e-06,"reference_active":true,"delta_db":-2.96165053574611}],"active_band_count":3,"active_band_mae_db":1.4145892600021923}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":0.7296741195782722,"rms_envelope_nrmse":0.07794964164936923,"envelope_correlation":0.9969880299958518,"rms_gain_error_db":-0.4540517336313603,"reference_mean_power":0.00021103616868638406,"output_mean_power":0.00019008669527844355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.885738469163612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011739951182146861,"output_power":0.00010764895269219947,"reference_active":true,"delta_db":-0.37656482038090305},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.974678911762062e-05,"output_power":1.4896687925646792e-05,"reference_active":true,"delta_db":-1.2240676850208705},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.379135569148538e-07,"output_power":3.0425125648947217e-07,"reference_active":true,"delta_db":-3.215294514138445}],"active_band_count":3,"active_band_mae_db":1.6053090065134061}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":1.1752967468159883,"rms_envelope_nrmse":0.1420237526337972,"envelope_correlation":0.9882333444872974,"rms_gain_error_db":-0.5445728934087337,"reference_mean_power":4.50355353361435e-05,"output_mean_power":3.972812239704346e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":3.018630059598247e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.579538283809341e-05,"output_power":1.3800071525312384e-05,"reference_active":true,"delta_db":-0.5864881917160613},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.352602360298045e-06,"output_power":3.5646266205101605e-06,"reference_active":true,"delta_db":-0.867349471587598},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8485554777348664e-06,"output_power":1.6103862604098835e-06,"reference_active":true,"delta_db":-0.5990243233273317}],"active_band_count":3,"active_band_mae_db":0.684287328876997}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":0.858726257682976,"rms_envelope_nrmse":0.09182254341595543,"envelope_correlation":0.9963518191605049,"rms_gain_error_db":-0.5445723007422977,"reference_mean_power":4.529062388817584e-05,"output_mean_power":3.9953154356524275e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":2.7414916263445047e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5534436484200313e-05,"output_power":1.3622120221606133e-05,"reference_active":true,"delta_db":-0.5705079491312702},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0764100435071083e-06,"output_power":1.6945924716093396e-06,"reference_active":true,"delta_db":-0.8824784819585343},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47015628096149e-07,"output_power":1.166675277039462e-07,"reference_active":true,"delta_db":-1.0041350884400821}],"active_band_count":3,"active_band_mae_db":0.8190405065099622}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5694-0004","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2198827266693115,"predicted_mos":3.2198827266693115,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":256271,"input_sample_rate":48000,"input_seconds":5.338979166666666,"input_float32_sha256":"b5e9f0f9e26d91ee761807e31440b5ba4f03d3a93cb4075c8b566513bcfd8df0","internal_samples":85424,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9871787390002282},"started_at":"2026-10-13T05:14:33.958682+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00896202399962931,"native_input_samples":255840,"native_input_sha256":"c836fb649da8d320090189881fbd43bdd2c71b39d9dd9ee8c2fe09009e831e9e","output_sha256":"5c4647ab9c24c4e76cd5e64341a83c4031b078646eeb7969bacb8e8fab00d906","output_samples":256271,"length_delta_samples":431,"analysis_s":3.01195166800062,"synthesis_s":0.9495558969992999,"total_s":3.96150756499992,"rtf":0.7432471979361951,"output_peak":0.6374665882723517,"output_rms":0.0610986612279038,"finished_at":"2026-10-13T05:14:37.940048+00:00","evaluation_input":{"reference":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"5c4647ab9c24c4e76cd5e64341a83c4031b078646eeb7969bacb8e8fab00d906","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":85280,"output_samples":85424,"length_delta_samples":144,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.275720777213216,"output_peak":0.6317714688028836},"common_prefix":{"samples":85280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983142910657427,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":85280,"output_samples":85280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.651454352954782,"spectral_convergence":0.48375942955983303,"spectral_convergence_status":"ok","amplitude_floor":1.2882929606468862e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.704873550682902,"spectral_convergence":0.21001228072730438,"spectral_convergence_status":"ok","amplitude_floor":9.933866418820512e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.726477745402757,"spectral_convergence":0.19562995686657988,"spectral_convergence_status":"ok","amplitude_floor":8.99033450142914e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001234751326260816,"output_power":0.0010885430520261642,"delta_db":-0.5473389119183452,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018448232014630782,"output_power":0.0017448655591848074,"delta_db":-0.2419278143762979,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006193820215877917,"output_power":0.0005984145597537352,"delta_db":-0.14956443603685973,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00017436605640499536,"output_power":0.0001771935203096071,"delta_db":0.06985891041867223,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010535962705358282,"output_power":0.00011659315900990272,"delta_db":0.439988447007157,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.506722876313705e-05,"output_power":1.6879625622563973e-05,"delta_db":0.4932942793183475,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3236621331792799,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2822022919168316,"pitch":null,"output_rms_dbfs":-24.26839412243972,"output_peak":0.6317714688028836,"reference_rms_dbfs":-23.986191830522884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.027601883013481},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":85280,"output_samples":85280,"intelligibility":{"estoi":{"value":0.9153835582034789,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06319611903854065,"rms_dbfs":-23.986191830522884,"peak":0.596405029296875,"peak_dbfs":-4.489174058684304,"crest_factor_db":19.497017771838582,"mean":-1.0716221793283889e-05,"zero_energy":false},"output":{"rms":0.06117588966175139,"rms_dbfs":-24.268394122439716,"peak":0.6317714688028836,"peak_dbfs":-3.988799820002547,"crest_factor_db":20.27959430243717,"mean":-0.0006800141527128136,"zero_energy":false},"rms_difference_db":-0.2822022919168319},"full_scale":{"reference":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":85280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":533,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.0564682721160354e-06,"threshold_dbfs":-55.1478010807523,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":116,"quiet_samples":18560,"reference_mean_power":8.461537774941273e-07,"output_mean_power":2.4593576733827088e-06,"output_rms_dbfs":-56.091783056328254,"output_zero_energy":false,"output_peak":0.047113836708786196,"output_energy_fraction":0.0001430183906782495,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":526,"reference_estimated_voiced_frames":222,"output_estimated_voiced_frames":222,"both_estimated_voiced_frames":215,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9684684684684685,"absolute_difference_cents_median":4.1897911644776915,"absolute_difference_cents_p95":17.339667889312896,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":85280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.0001748275799785616,"log_rms_mae_db":1.6391233945747656,"rms_envelope_nrmse":0.13124988648135325,"envelope_correlation":0.9844981507578339,"rms_gain_error_db":-0.282202550649263,"reference_mean_power":0.003997498793754849,"output_mean_power":0.003746002702070038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.9383161701154944e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012108392795166332,"output_power":0.0011591233049174762,"reference_active":true,"delta_db":-0.18956863300233637},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231683599850243,"output_power":0.00020376115991446207,"reference_active":true,"delta_db":-0.39511217712981667},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.764859371015312e-05,"output_power":7.586366330342987e-05,"reference_active":true,"delta_db":0.49775037234574443}],"active_band_count":3,"active_band_mae_db":0.3608103941592991}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00017019095451879833,"log_rms_mae_db":1.092057727587756,"rms_envelope_nrmse":0.07128830418201959,"envelope_correlation":0.9957796325100942,"rms_gain_error_db":-0.2822003119029873,"reference_mean_power":0.00402014109400192,"output_mean_power":0.0037672224410347076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.6741447857886242e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011176954288226882,"output_power":0.0010805123776560998,"reference_active":true,"delta_db":-0.14693728500447753},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.194819962775495e-05,"output_power":8.273944609942733e-05,"reference_active":true,"delta_db":-0.45830620868516275},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3053605179616585e-06,"output_power":4.270115894708456e-06,"reference_active":true,"delta_db":1.1122082711480505}],"active_band_count":3,"active_band_mae_db":0.5724839216125636}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":0.00014662911909855806,"log_rms_mae_db":2.179726158308662,"rms_envelope_nrmse":0.16080627248207988,"envelope_correlation":0.9793967544339739,"rms_gain_error_db":-0.1824044925285087,"reference_mean_power":0.0015423797799867744,"output_mean_power":0.001478941085314948},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":1.0119951767590758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006526269487739279,"output_power":0.0006262307633812946,"reference_active":true,"delta_db":-0.17930604664675331},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016341376588393177,"output_power":0.00014328489978010189,"reference_active":true,"delta_db":-0.5708821420253862},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.788084498799748e-05,"output_power":4.972220649321957e-05,"reference_active":true,"delta_db":0.1638858671726022}],"active_band_count":3,"active_band_mae_db":0.30469135194824726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":0.00013732417437068112,"log_rms_mae_db":1.5664445557565012,"rms_envelope_nrmse":0.09220745313054253,"envelope_correlation":0.992826928048381,"rms_gain_error_db":-0.18240598111139913,"reference_mean_power":0.0015511172002359399,"output_mean_power":0.0014873186221974527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":8.798155816797883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006408046711327789,"output_power":0.0006151375461597722,"reference_active":true,"delta_db":-0.1775343299695048},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.486060187757687e-05,"output_power":6.546807233544207e-05,"reference_active":true,"delta_db":-0.5822376091706395},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5748897665955194e-06,"output_power":3.8227780773161526e-06,"reference_active":true,"delta_db":0.29116432103454776}],"active_band_count":3,"active_band_mae_db":0.3503120867248974}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":8.659636175169931e-05,"log_rms_mae_db":1.8571580327308819,"rms_envelope_nrmse":0.19202791251658832,"envelope_correlation":0.9730788524032229,"rms_gain_error_db":0.14858333640473628,"reference_mean_power":0.00020984803844209808,"output_mean_power":0.0002171517069954374},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":2.0976407038326692e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011684508155422022,"output_power":0.0001252840253198607,"reference_active":true,"delta_db":0.3028526258945329},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.122844755814484e-05,"output_power":3.9571108177660214e-05,"reference_active":true,"delta_db":-0.17818769181975702},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.871758765973679e-06,"output_power":8.128310796208142e-06,"reference_active":true,"delta_db":-0.38009423246618207}],"active_band_count":3,"active_band_mae_db":0.2870448500601573}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":7.401841529112699e-05,"log_rms_mae_db":1.0653275033802063,"rms_envelope_nrmse":0.10094357227664409,"envelope_correlation":0.9920433779319324,"rms_gain_error_db":0.14857852757892567,"reference_mean_power":0.00021103616868638406,"output_mean_power":0.00021838094778210563},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":1.885738469163612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011739951182146861,"output_power":0.00012609628286462072,"reference_active":true,"delta_db":0.3103599338798429},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.974678911762062e-05,"output_power":1.8119461968050874e-05,"reference_active":true,"delta_db":-0.3735119045427262},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.379135569148538e-07,"output_power":6.283277349550185e-07,"reference_active":true,"delta_db":-0.06575601567861697}],"active_band_count":3,"active_band_mae_db":0.249875951367062}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1065,"rms_floor":4.122543101057287e-05,"log_rms_mae_db":1.818224497992024,"rms_envelope_nrmse":0.27156479751087376,"envelope_correlation":0.9537981043976722,"rms_gain_error_db":0.3508936978267214,"reference_mean_power":4.50355353361435e-05,"output_mean_power":4.882527514428121e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1065,"frequency_resolution_hz":0.18779342723004694,"power_floor":3.018630059598247e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.579538283809341e-05,"output_power":1.6945813427053817e-05,"reference_active":true,"delta_db":0.3053226401416901},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.352602360298045e-06,"output_power":4.556707889710589e-06,"reference_active":true,"delta_db":0.19902195558278873},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8485554777348664e-06,"output_power":1.8574833221181747e-06,"reference_active":true,"delta_db":0.020924341383131238}],"active_band_count":3,"active_band_mae_db":0.17508964570253668}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":85280,"uncovered_tail_samples":0,"metrics":{"frame_count":1059,"rms_floor":3.168822925930831e-05,"log_rms_mae_db":1.1545156278222866,"rms_envelope_nrmse":0.1688396155292686,"envelope_correlation":0.9814726503221586,"rms_gain_error_db":0.3508901578247472,"reference_mean_power":4.529062388817584e-05,"output_mean_power":4.9101789371539647e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1059,"frequency_resolution_hz":0.18885741265344666,"power_floor":2.7414916263445047e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5534436484200313e-05,"output_power":1.6606864204863938e-05,"reference_active":true,"delta_db":0.2899213070325843},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0764100435071083e-06,"output_power":2.1680254979403386e-06,"reference_active":true,"delta_db":0.1875126471637389},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.47015628096149e-07,"output_power":1.463552341853772e-07,"reference_active":true,"delta_db":-0.019552447270311267}],"active_band_count":3,"active_band_mae_db":0.16566213382221148}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"hnm","status":"measured","mos":{"status":"measured","value":3.0901541709899902,"predicted_mos":3.0901541709899902,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ddc438d929100fd7b5cf15542783d9a9027a03078f00d12654f25421557ec287","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0831755449999037},"started_at":"2026-10-13T04:38:56.931868+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2874999811174348e-05,"native_input_samples":96000,"native_input_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","output_sha256":"be0fff9381c6695efbe0f1d70bd5b3466cf0e4e74b14e5c40fd531021bfffcea","output_samples":96000,"length_delta_samples":0,"analysis_s":1.129615694999302,"synthesis_s":0.46660585700010415,"total_s":1.596221551999406,"rtf":0.26603692533323436,"output_peak":0.7006643602595451,"output_rms":0.08379653313070698,"finished_at":"2026-10-13T04:38:58.534304+00:00","evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"be0fff9381c6695efbe0f1d70bd5b3466cf0e4e74b14e5c40fd531021bfffcea","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.571279791085914,"spectral_convergence":0.14507158638387546,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.308229939874918,"spectral_convergence":0.1417074564964475,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.488757949290679,"spectral_convergence":0.13692676817944635,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.004046411624191875,"delta_db":-0.024098828546604296,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0018073059743346324,"delta_db":-0.024787717683271288,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0007531922246988052,"delta_db":-0.1547693458668095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.0003000036052270722,"delta_db":-0.15548087833558838,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":7.758909656607793e-05,"delta_db":-0.17958297302576892,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":3.7356439707210075e-05,"delta_db":-0.6646778685990106,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20056626867617547,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04955083561231998,"pitch":null,"output_rms_dbfs":-21.53547897664246,"output_peak":0.7006643602595451,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.456089226750503},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.571279791085914,"spectral_convergence":0.14507158638387546,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.308229939874918,"spectral_convergence":0.1417074564964475,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.488757949290679,"spectral_convergence":0.13692676817944635,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.004046411624191875,"delta_db":-0.024098828546604296,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0018073059743346324,"delta_db":-0.024787717683271288,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0007531922246988052,"delta_db":-0.1547693458668095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.0003000036052270722,"delta_db":-0.15548087833558838,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":7.758909656607793e-05,"delta_db":-0.17958297302576892,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":3.7356439707210075e-05,"delta_db":-0.6646778685990106,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20056626867617547,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04955083561231998,"pitch":null,"output_rms_dbfs":-21.53547897664246,"output_peak":0.7006643602595451,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.456089226750503},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9241571485810357,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08427593760910047,"rms_dbfs":-21.485928141030136,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.931131229807892,"mean":0.001402181625366211,"zero_energy":false},"output":{"rms":0.08379653313070698,"rms_dbfs":-21.53547897664246,"peak":0.7006643602595451,"peak_dbfs":-3.0897994522196033,"crest_factor_db":18.445679524422857,"mean":1.9788721867527004e-05,"zero_energy":false},"rms_difference_db":-0.04955083561232243},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":6.4401924939058445e-06,"output_rms_dbfs":-51.911011516166276,"output_zero_energy":false,"output_peak":0.020201266955995865,"output_energy_fraction":1.9871875059808753e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":376,"output_estimated_voiced_frames":370,"both_estimated_voiced_frames":363,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9654255319148937,"absolute_difference_cents_median":1.3623733014792965,"absolute_difference_cents_p95":9.413011636106987,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019068326661124743,"log_rms_mae_db":0.7461338154697332,"rms_envelope_nrmse":0.050515457393448794,"envelope_correlation":0.9965124231927328,"rms_gain_error_db":-0.048513254681385114,"reference_mean_power":0.007104543911662494,"output_mean_power":0.007025623580560728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.7734542191089184e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001358301262786015,"output_power":0.001324171608722762,"reference_active":true,"delta_db":-0.11051832416496621},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002660490997406596,"output_power":0.0002541164341640368,"reference_active":true,"delta_db":-0.19929041134576791},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.883599774152297e-05,"output_power":5.8229752809574546e-05,"reference_active":true,"delta_db":-1.315796218929198}],"active_band_count":3,"active_band_mae_db":0.541868318146644}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001870958079240233,"log_rms_mae_db":0.361296096729037,"rms_envelope_nrmse":0.024477555326449622,"envelope_correlation":0.9990999040953192,"rms_gain_error_db":-0.04861158636136098,"reference_mean_power":0.007117638239873263,"output_mean_power":0.007038413087854484},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4206220628648626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103399895685935,"output_power":0.001190322399994755,"reference_active":true,"delta_db":-0.07242776178031543},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012555497766250794,"output_power":0.00012099742073662423,"reference_active":true,"delta_db":-0.1605782241915135},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0918852450489685e-06,"output_power":4.202777077241004e-06,"reference_active":true,"delta_db":-1.6121535891972112}],"active_band_count":3,"active_band_mae_db":0.6150531917230134}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013521834879092463,"log_rms_mae_db":1.4404843861793424,"rms_envelope_nrmse":0.07995083033713446,"envelope_correlation":0.9935201550173745,"rms_gain_error_db":-0.06478522529763943,"reference_mean_power":0.0018937311393695897,"output_mean_power":0.0018656913499252759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.102346197011796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006150154421416123,"output_power":0.0005993996386288974,"reference_active":true,"delta_db":-0.11169543635356816},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012427322589552618,"output_power":0.00011066023173774812,"reference_active":true,"delta_db":-0.5038599658603264},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6589172324725154e-05,"output_power":1.6357302128822892e-05,"reference_active":true,"delta_db":-2.109931432762182}],"active_band_count":3,"active_band_mae_db":0.9084956116586923}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012428224153258587,"log_rms_mae_db":0.7164706802373818,"rms_envelope_nrmse":0.0390026535187249,"envelope_correlation":0.9982865097042126,"rms_gain_error_db":-0.06626568891049628,"reference_mean_power":0.0018973976433641784,"output_mean_power":0.0018686664482585704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.796182335233462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005643605996811222,"output_power":0.0005491277818726544,"reference_active":true,"delta_db":-0.11883269860839193},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.691592416494353e-05,"output_power":4.9932758530879156e-05,"reference_active":true,"delta_db":-0.5684823276523668},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.24640973682173e-06,"output_power":1.2540568042424407e-06,"reference_active":true,"delta_db":-2.5317176394536745}],"active_band_count":3,"active_band_mae_db":1.0730108885714778}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001274818876901684,"log_rms_mae_db":1.7904695123572707,"rms_envelope_nrmse":0.13928325209251308,"envelope_correlation":0.9870592141189853,"rms_gain_error_db":-0.20549176407555225,"reference_mean_power":0.0003203305006386765,"output_mean_power":0.00030552666258306033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0949985705721986e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.663152856065708e-05,"output_power":6.390633612400875e-05,"reference_active":true,"delta_db":-0.18135857010351805},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2062707324852754e-05,"output_power":1.1236030133752864e-05,"reference_active":true,"delta_db":-0.30831895487505967},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2084052775211498e-06,"output_power":2.26779188007114e-06,"reference_active":true,"delta_db":-1.506860260906553}],"active_band_count":3,"active_band_mae_db":0.6655125952950436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012438051646139164,"log_rms_mae_db":0.99444691483626,"rms_envelope_nrmse":0.08434745960464887,"envelope_correlation":0.9952660472389171,"rms_gain_error_db":-0.20675764981929298,"reference_mean_power":0.00032140841726426223,"output_mean_power":0.0003064654222291558},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.99354395167411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.454820101360057e-05,"output_power":6.197610673767007e-05,"reference_active":true,"delta_db":-0.17659851877036895},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4052169228627875e-06,"output_power":5.3089631958828895e-06,"reference_active":true,"delta_db":-0.07803413174038092},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434599324165249e-07,"output_power":1.9462925610878083e-07,"reference_active":true,"delta_db":-2.4666795739410405}],"active_band_count":3,"active_band_mae_db":0.9071040748172634}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":1.9206325068827226,"rms_envelope_nrmse":0.20514053954941286,"envelope_correlation":0.970009130697954,"rms_gain_error_db":-0.5330375418918345,"reference_mean_power":5.5432329924189186e-05,"output_mean_power":4.902971615242141e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.209399643069474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1581941196726705e-05,"output_power":1.9808656830831194e-05,"reference_active":true,"delta_db":-0.3723547664996943},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7792321883200752e-05,"output_power":1.3321808580720365e-05,"reference_active":true,"delta_db":-1.256694378042358},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1497984756130903e-05,"output_power":5.737202505974764e-06,"reference_active":true,"delta_db":-3.0192154908810886}],"active_band_count":3,"active_band_mae_db":1.549421545141047}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":1.055198699307435,"rms_envelope_nrmse":0.12818348124419976,"envelope_correlation":0.9888060653187053,"rms_gain_error_db":-0.5334808943841738,"reference_mean_power":5.570117934209357e-05,"output_mean_power":4.9262483340162324e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.4557792596837174e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8496933872861596e-05,"output_power":2.364414134337894e-05,"reference_active":true,"delta_db":-0.8107458774124848},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0110098860775746e-05,"output_power":6.510853380791849e-06,"reference_active":true,"delta_db":-1.9111748683207161},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.149890921474854e-06,"output_power":5.311049212067136e-07,"reference_active":true,"delta_db":-3.3547631958007122}],"active_band_count":3,"active_band_mae_db":2.0255613138446376}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.1608128547668457,"predicted_mos":3.1608128547668457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2b36819c4a4b8ee462f76f6ae223b170ea432c51e308bb8145334d6294e8e07d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3388056580006378},"started_at":"2026-10-13T04:41:54.564852+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7516000045579858e-05,"native_input_samples":96000,"native_input_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","output_sha256":"3443a9733e6968088f2246bd9fc73ad7160a548dcbc66f611c1749f9326d05c3","output_samples":96000,"length_delta_samples":0,"analysis_s":1.9657351320001908,"synthesis_s":0.09450327000013203,"total_s":2.060238402000323,"rtf":0.3433730670000538,"output_peak":0.7844574008979924,"output_rms":0.09309118549704347,"finished_at":"2026-10-13T04:41:56.632347+00:00","evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"3443a9733e6968088f2246bd9fc73ad7160a548dcbc66f611c1749f9326d05c3","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.666005723604068,"spectral_convergence":0.6093848605761067,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.10646395665999,"spectral_convergence":0.20652643132355444,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.708076977325338,"spectral_convergence":0.24245843115943488,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.0048343764077529655,"delta_db":0.7486051938292775,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0021481013837262256,"delta_db":0.7254431887165941,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0011708595090540311,"delta_db":1.7612202470798166,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.0003713762050303963,"delta_db":0.7713951237713857,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":9.311695986552321e-05,"delta_db":0.6126979609781247,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":4.813835181581524e-05,"delta_db":0.43657951087972385,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8426568708758203,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8640993608237508,"pitch":null,"output_rms_dbfs":-20.62182878020639,"output_peak":0.7844574008979924,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.826848885863132},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.666005723604068,"spectral_convergence":0.6093848605761067,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.10646395665999,"spectral_convergence":0.20652643132355444,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.708076977325338,"spectral_convergence":0.24245843115943488,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.0048343764077529655,"delta_db":0.7486051938292775,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0021481013837262256,"delta_db":0.7254431887165941,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0011708595090540311,"delta_db":1.7612202470798166,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.0003713762050303963,"delta_db":0.7713951237713857,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":9.311695986552321e-05,"delta_db":0.6126979609781247,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":4.813835181581524e-05,"delta_db":0.43657951087972385,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8426568708758203,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8640993608237508,"pitch":null,"output_rms_dbfs":-20.62182878020639,"output_peak":0.7844574008979924,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.826848885863132},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9358839079723469,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08427593760910047,"rms_dbfs":-21.485928141030136,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.931131229807892,"mean":0.001402181625366211,"zero_energy":false},"output":{"rms":0.09309118549704348,"rms_dbfs":-20.62182878020639,"peak":0.7844574008979924,"peak_dbfs":-2.10861270648042,"crest_factor_db":18.513216073725967,"mean":2.386726140168666e-05,"zero_energy":false},"rms_difference_db":0.8640993608237473},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":2.989035223287154e-06,"output_rms_dbfs":-55.24468967132311,"output_zero_energy":false,"output_peak":0.010989922196006379,"output_energy_fraction":7.473189807585447e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":376,"output_estimated_voiced_frames":398,"both_estimated_voiced_frames":365,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":33,"paired_voiced_fraction":0.9707446808510638,"absolute_difference_cents_median":7.652207994864537,"absolute_difference_cents_p95":24.837477421563342,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019068326661124743,"log_rms_mae_db":1.6961907937151013,"rms_envelope_nrmse":0.1552978336398335,"envelope_correlation":0.9835761149876104,"rms_gain_error_db":0.8659842077582813,"reference_mean_power":0.007104543911662494,"output_mean_power":0.008672306603621116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.7734542191089184e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001358301262786015,"output_power":0.0016144568601045416,"reference_active":true,"delta_db":0.7503034027956135},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002660490997406596,"output_power":0.0002652397676855607,"reference_active":true,"delta_db":-0.013231548371900766},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.883599774152297e-05,"output_power":7.380176233843257e-05,"reference_active":true,"delta_db":-0.2865783576788038}],"active_band_count":3,"active_band_mae_db":0.3500377696154393}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001870958079240233,"log_rms_mae_db":1.185508666773089,"rms_envelope_nrmse":0.11861321015037886,"envelope_correlation":0.9956669624486745,"rms_gain_error_db":0.8669785688417426,"reference_mean_power":0.007117638239873263,"output_mean_power":0.008690279962352727},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4206220628648626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103399895685935,"output_power":0.0014432459875805966,"reference_active":true,"delta_db":0.764329761831301},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012555497766250794,"output_power":0.00012274163000604822,"reference_active":true,"delta_db":-0.09842048710471082},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0918852450489685e-06,"output_power":4.830520152998822e-06,"reference_active":true,"delta_db":-1.0075781545348106}],"active_band_count":3,"active_band_mae_db":0.6234428011569407}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013521834879092463,"log_rms_mae_db":1.9321351790403833,"rms_envelope_nrmse":0.22068684186752957,"envelope_correlation":0.9804265750097632,"rms_gain_error_db":1.3046010569314634,"reference_mean_power":0.0018937311393695897,"output_mean_power":0.002557280848450432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.102346197011796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006150154421416123,"output_power":0.000864473157855435,"reference_active":true,"delta_db":1.4786549243457439},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012427322589552618,"output_power":0.00017218190886958713,"reference_active":true,"delta_db":1.4160994631337052},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6589172324725154e-05,"output_power":2.3686026069601108e-05,"reference_active":true,"delta_db":-0.5021261575799606}],"active_band_count":3,"active_band_mae_db":1.1322935150198032}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012428224153258587,"log_rms_mae_db":1.3931384329802805,"rms_envelope_nrmse":0.19004577665165392,"envelope_correlation":0.990323031063057,"rms_gain_error_db":1.3042459915302627,"reference_mean_power":0.0018973976433641784,"output_mean_power":0.0025620225975456702},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.796182335233462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005643605996811222,"output_power":0.0007963668929088566,"reference_active":true,"delta_db":1.4955651068447233},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.691592416494353e-05,"output_power":7.537095071590775e-05,"reference_active":true,"delta_db":1.219702013767186},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.24640973682173e-06,"output_power":1.853228853892208e-06,"reference_active":true,"delta_db":-0.8355991951420617}],"active_band_count":3,"active_band_mae_db":1.1836221052513236}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001274818876901684,"log_rms_mae_db":1.693185240511615,"rms_envelope_nrmse":0.15998739474264687,"envelope_correlation":0.9893039395583939,"rms_gain_error_db":0.7534823479489599,"reference_mean_power":0.0003203305006386765,"output_mean_power":0.0003810189073652013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0949985705721986e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.663152856065708e-05,"output_power":8.238406357633082e-05,"reference_active":true,"delta_db":0.9216343331576056},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2062707324852754e-05,"output_power":1.1931489412669411e-05,"reference_active":true,"delta_db":-0.04750130536020722},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2084052775211498e-06,"output_power":2.6593326282482754e-06,"reference_active":true,"delta_db":-0.8151655993466157}],"active_band_count":3,"active_band_mae_db":0.5947670792881429}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012438051646139164,"log_rms_mae_db":1.025164354773932,"rms_envelope_nrmse":0.111909401141495,"envelope_correlation":0.997068862810418,"rms_gain_error_db":0.7439137242786721,"reference_mean_power":0.00032140841726426223,"output_mean_power":0.0003814596611050926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.99354395167411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.454820101360057e-05,"output_power":7.867576337879807e-05,"reference_active":true,"delta_db":0.8595682270282243},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4052169228627875e-06,"output_power":5.333595471384136e-06,"reference_active":true,"delta_db":-0.05793054431618991},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434599324165249e-07,"output_power":2.2816929106289144e-07,"reference_active":true,"delta_db":-1.7761888710718017}],"active_band_count":3,"active_band_mae_db":0.8978958808054053}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":1.8327520529588375,"rms_envelope_nrmse":0.24679812016529096,"envelope_correlation":0.9577610232907501,"rms_gain_error_db":0.3694021410112873,"reference_mean_power":5.5432329924189186e-05,"output_mean_power":6.035362321844176e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.209399643069474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1581941196726705e-05,"output_power":2.4499761805460166e-05,"reference_active":true,"delta_db":0.5507135713229555},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7792321883200752e-05,"output_power":1.9534835185707502e-05,"reference_active":true,"delta_db":0.40577124607654275},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1497984756130903e-05,"output_power":7.700905876801418e-06,"reference_active":true,"delta_db":-1.740799132061236}],"active_band_count":3,"active_band_mae_db":0.8990946498202447}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":1.161300393699177,"rms_envelope_nrmse":0.14732662997565818,"envelope_correlation":0.9838621670622689,"rms_gain_error_db":0.3694244070067209,"reference_mean_power":5.570117934209357e-05,"output_mean_power":6.064665207187062e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.4557792596837174e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8496933872861596e-05,"output_power":3.022038566721969e-05,"reference_active":true,"delta_db":0.25501867813348883},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0110098860775746e-05,"output_power":9.453673224511461e-06,"reference_active":true,"delta_db":-0.2915481591332284},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.149890921474854e-06,"output_power":7.905372966254988e-07,"reference_active":true,"delta_db":-1.6273428092359459}],"active_band_count":3,"active_band_mae_db":0.724636548834221}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"facodec","status":"measured","mos":{"status":"measured","value":4.326303482055664,"predicted_mos":4.326303482055664,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"dafc3fabef0adea5f4681798f79bdd5efc1b0a58fa571008d6881b1adf2aa2e6","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.999891920000664},"started_at":"2026-10-13T04:50:52.218175+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.382200000283774e-05,"native_input_samples":96000,"native_input_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","output_sha256":"a87bbe0828067936bc8d47174173e8126bfb4a5b364d34d5c838693d8d464c53","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5673224629999822,"synthesis_s":6.78247272399949,"total_s":8.349795186999472,"rtf":1.3916325311665787,"output_peak":0.6013345122337341,"output_rms":0.0764539198165007,"finished_at":"2026-10-13T04:51:00.574383+00:00","evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"a87bbe0828067936bc8d47174173e8126bfb4a5b364d34d5c838693d8d464c53","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.261742024553551,"spectral_convergence":0.1991252470201458,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.694216632089262,"spectral_convergence":0.18179407930544006,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.863472535198708,"spectral_convergence":0.19733283647941172,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.0033259087326962325,"delta_db":-0.8756961502549526,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0015758636876550664,"delta_db":-0.6199180772477383,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0006309679836415356,"delta_db":-0.9237543969276797,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.00020577919262088415,"delta_db":-1.7927310269659165,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":6.591075071200421e-05,"delta_db":-0.8880273414375237,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":4.077150798219438e-05,"delta_db":-0.2847649087933262,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8974819836045227,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8460767242946583,"pitch":null,"output_rms_dbfs":-22.332004865324798,"output_peak":0.6013345122337341,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.606477063947175},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.261742024553551,"spectral_convergence":0.1991252470201458,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.694216632089262,"spectral_convergence":0.18179407930544006,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.863472535198708,"spectral_convergence":0.19733283647941172,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.0033259087326962325,"delta_db":-0.8756961502549526,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0015758636876550664,"delta_db":-0.6199180772477383,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0006309679836415356,"delta_db":-0.9237543969276797,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.00020577919262088415,"delta_db":-1.7927310269659165,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":6.591075071200421e-05,"delta_db":-0.8880273414375237,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":4.077150798219438e-05,"delta_db":-0.2847649087933262,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8974819836045227,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8460767242946583,"pitch":null,"output_rms_dbfs":-22.332004865324798,"output_peak":0.6013345122337341,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.606477063947175},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9135524878394086,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08427593760910047,"rms_dbfs":-21.485928141030136,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.931131229807892,"mean":0.001402181625366211,"zero_energy":false},"output":{"rms":0.0764539198165007,"rms_dbfs":-22.332004865324798,"peak":0.6013345122337341,"peak_dbfs":-4.417677401801279,"crest_factor_db":17.91432746352352,"mean":3.923160581152748e-05,"zero_energy":false},"rms_difference_db":-0.8460767242946616},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":1.4125041834069215e-06,"output_rms_dbfs":-58.500102572587174,"output_zero_energy":false,"output_peak":0.01143968477845192,"output_energy_fraction":5.235791348994871e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":376,"output_estimated_voiced_frames":372,"both_estimated_voiced_frames":362,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9627659574468085,"absolute_difference_cents_median":5.0631459284321725,"absolute_difference_cents_p95":18.762283396009444,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019068326661124743,"log_rms_mae_db":1.2249474299801903,"rms_envelope_nrmse":0.11349194990062964,"envelope_correlation":0.9935267261140324,"rms_gain_error_db":-0.8465211644170973,"reference_mean_power":0.007104543911662494,"output_mean_power":0.0058463402401002975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.7734542191089184e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001358301262786015,"output_power":0.0010626425605463727,"reference_active":true,"delta_db":-1.0660889833183538},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002660490997406596,"output_power":0.00021164179536385292,"reference_active":true,"delta_db":-0.9936035668456782},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.883599774152297e-05,"output_power":6.261716044162217e-05,"reference_active":true,"delta_db":-1.000311990132932}],"active_band_count":3,"active_band_mae_db":1.0200015134323213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001870958079240233,"log_rms_mae_db":1.0208007743679126,"rms_envelope_nrmse":0.10851283349416142,"envelope_correlation":0.9945845552194833,"rms_gain_error_db":-0.8485037048333117,"reference_mean_power":0.007117638239873263,"output_mean_power":0.005854442438822018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4206220628648626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103399895685935,"output_power":0.0009456604970761755,"reference_active":true,"delta_db":-1.0717213489005009},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012555497766250794,"output_power":9.948286949794429e-05,"reference_active":true,"delta_db":-1.010856314875338},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0918852450489685e-06,"output_power":4.638831716327946e-06,"reference_active":true,"delta_db":-1.1834309595441992}],"active_band_count":3,"active_band_mae_db":1.0886695411066794}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013521834879092463,"log_rms_mae_db":1.4392870490463578,"rms_envelope_nrmse":0.1277823505749738,"envelope_correlation":0.9870551168220421,"rms_gain_error_db":-0.6071263078248212,"reference_mean_power":0.0018937311393695897,"output_mean_power":0.0016466666496882708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.102346197011796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006150154421416123,"output_power":0.0004903926234403235,"reference_active":true,"delta_db":-0.9834209157744055},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012427322589552618,"output_power":9.201497041254081e-05,"reference_active":true,"delta_db":-1.305190811284513},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6589172324725154e-05,"output_power":1.9196857079427755e-05,"reference_active":true,"delta_db":-1.4147468712151066}],"active_band_count":3,"active_band_mae_db":1.2344528660913416}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012428224153258587,"log_rms_mae_db":1.0122532192200415,"rms_envelope_nrmse":0.11529416826234465,"envelope_correlation":0.989073194002629,"rms_gain_error_db":-0.6108967935584506,"reference_mean_power":0.0018973976433641784,"output_mean_power":0.0016484230455159478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.796182335233462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005643605996811222,"output_power":0.0004485877773418987,"reference_active":true,"delta_db":-0.997092500697039},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.691592416494353e-05,"output_power":4.166559097709525e-05,"reference_active":true,"delta_db":-1.3545624588520289},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.24640973682173e-06,"output_power":1.7440193826209081e-06,"reference_active":true,"delta_db":-1.0993766558656088}],"active_band_count":3,"active_band_mae_db":1.1503438718048922}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001274818876901684,"log_rms_mae_db":1.702206723436873,"rms_envelope_nrmse":0.2091622342607819,"envelope_correlation":0.986284993436433,"rms_gain_error_db":-1.5231433309436493,"reference_mean_power":0.0003203305006386765,"output_mean_power":0.00022557136073555822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0949985705721986e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.663152856065708e-05,"output_power":4.226414575791148e-05,"reference_active":true,"delta_db":-1.977076808666805},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2062707324852754e-05,"output_power":1.0767146255930446e-05,"reference_active":true,"delta_db":-0.49344178398767125},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2084052775211498e-06,"output_power":3.401547479364292e-06,"reference_active":true,"delta_db":0.2538731525166043}],"active_band_count":3,"active_band_mae_db":0.9081305817236934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012438051646139164,"log_rms_mae_db":1.3524018574317938,"rms_envelope_nrmse":0.19758586289532923,"envelope_correlation":0.9901035835206584,"rms_gain_error_db":-1.524418200165454,"reference_mean_power":0.00032140841726426223,"output_mean_power":0.0002262639821014168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.99354395167411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.454820101360057e-05,"output_power":4.105327445394412e-05,"reference_active":true,"delta_db":-1.96536340156741},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4052169228627875e-06,"output_power":5.234549948114851e-06,"reference_active":true,"delta_db":-0.13933779599635604},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434599324165249e-07,"output_power":3.274582154524475e-07,"reference_active":true,"delta_db":-0.20720189406965014}],"active_band_count":3,"active_band_mae_db":0.7706343638778055}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":1.7231779746095222,"rms_envelope_nrmse":0.2520805027972264,"envelope_correlation":0.954018360700943,"rms_gain_error_db":-0.7597109919950006,"reference_mean_power":5.5432329924189186e-05,"output_mean_power":4.653631965949424e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.209399643069474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1581941196726705e-05,"output_power":1.3640836292714542e-05,"reference_active":true,"delta_db":-1.9924950802949513},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7792321883200752e-05,"output_power":1.2810591384392004e-05,"reference_active":true,"delta_db":-1.4266344799994428},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1497984756130903e-05,"output_power":1.0022270988233236e-05,"reference_active":true,"delta_db":-0.5965558724626486}],"active_band_count":3,"active_band_mae_db":1.338561810919014}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":1.3051886472589163,"rms_envelope_nrmse":0.223655877575648,"envelope_correlation":0.9598978712841983,"rms_gain_error_db":-0.7598046576746498,"reference_mean_power":5.570117934209357e-05,"output_mean_power":4.6761014486876286e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.4557792596837174e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8496933872861596e-05,"output_power":1.8879233607073264e-05,"reference_active":true,"delta_db":-1.7881377425908993},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0110098860775746e-05,"output_power":7.940327545518376e-06,"reference_active":true,"delta_db":-1.049169844976316},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.149890921474854e-06,"output_power":1.0050158561699686e-06,"reference_active":true,"delta_db":-0.5848373149148822}],"active_band_count":3,"active_band_mae_db":1.1407149674940325}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.489633798599243,"predicted_mos":3.489633798599243,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"8cfaa0380c3b605ea228acd781a11ce2f86dbf6260dec98fcb76acafee54b75b","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3075512720024562},"started_at":"2026-10-13T07:12:02.334346+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0064502929999434855,"native_input_samples":264600,"native_input_sha256":"18b5f88e05e714897deedb60039cf561fce5b96b9aa3a368f4aac55110409439","output_sha256":"897d74d204e15d6350e0f9235a5cdc9f92f125bf8e76029af0514d647e63fa7b","output_samples":264704,"length_delta_samples":104,"analysis_s":0.21901609700034896,"synthesis_s":4.1490770549989975,"total_s":4.3680931519993464,"rtf":0.7280155253332244,"output_peak":0.5278319120407104,"output_rms":0.07705500867098182,"finished_at":"2026-10-13T07:12:06.721193+00:00","evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"897d74d204e15d6350e0f9235a5cdc9f92f125bf8e76029af0514d647e63fa7b","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.26075003554535,"output_peak":0.5269667063738701},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.083788179606135,"spectral_convergence":0.4536190350563446,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.958385254625855,"spectral_convergence":0.3170204464378914,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.816694042882462,"spectral_convergence":0.3578575841943211,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.0028243710167268686,"delta_db":-1.5855819628097623,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.002230249921850325,"delta_db":0.8884307683058768,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0005655131203767846,"delta_db":-1.3994007730296352,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.00023951313224644023,"delta_db":-1.1334523132167245,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":5.98272217373156e-05,"delta_db":-1.3086015685238994,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":2.0748744422081268e-05,"delta_db":-3.218414443311577,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.588980304866246,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7760442539567286,"pitch":null,"output_rms_dbfs":-22.261972394986866,"output_peak":0.5269667063738701,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.952955825704817},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7933353023231474,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08427593760910047,"rms_dbfs":-21.485928141030136,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.931131229807892,"mean":0.001402181625366211,"zero_energy":false},"output":{"rms":0.07707284318980204,"rms_dbfs":-22.261972394986866,"peak":0.5269667063738701,"peak_dbfs":-5.564336450772471,"crest_factor_db":16.697635944214397,"mean":-0.00014637839557784065,"zero_energy":false},"rms_difference_db":-0.77604425395673},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":2.192024119104686e-06,"output_rms_dbfs":-56.59154671567027,"output_zero_energy":false,"output_peak":0.004705850091006167,"output_energy_fraction":7.995298266714265e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":376,"output_estimated_voiced_frames":350,"both_estimated_voiced_frames":341,"reference_voiced_output_unvoiced_frames":35,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9069148936170213,"absolute_difference_cents_median":8.85222503735834,"absolute_difference_cents_p95":32.30659869314291,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019068326661124743,"log_rms_mae_db":1.4566471853359213,"rms_envelope_nrmse":0.12468752316895129,"envelope_correlation":0.9879245376436789,"rms_gain_error_db":-0.7761648223890517,"reference_mean_power":0.007104543911662494,"output_mean_power":0.005941823138326612},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.7734542191089184e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001358301262786015,"output_power":0.001167844472491733,"reference_active":true,"delta_db":-0.6561109492117225},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002660490997406596,"output_power":0.0002372249353104987,"reference_active":true,"delta_db":-0.4980145681513076},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.883599774152297e-05,"output_power":5.944223587319918e-05,"reference_active":true,"delta_db":-1.2262943167963052}],"active_band_count":3,"active_band_mae_db":0.7934732780531117}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001870958079240233,"log_rms_mae_db":1.184899141858077,"rms_envelope_nrmse":0.09871409186514171,"envelope_correlation":0.9960286925498859,"rms_gain_error_db":-0.7754644489439324,"reference_mean_power":0.007117638239873263,"output_mean_power":0.005953734528320999},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4206220628648626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103399895685935,"output_power":0.0010556430023004953,"reference_active":true,"delta_db":-0.5939030941673171},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012555497766250794,"output_power":0.00011602990967995929,"reference_active":true,"delta_db":-0.34263980936995814},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0918852450489685e-06,"output_power":4.592076259874477e-06,"reference_active":true,"delta_db":-1.2274262207691276}],"active_band_count":3,"active_band_mae_db":0.7213230414354675}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013521834879092463,"log_rms_mae_db":2.3623807383413817,"rms_envelope_nrmse":0.20391320408371097,"envelope_correlation":0.9600134855640287,"rms_gain_error_db":0.3146744624810309,"reference_mean_power":0.0018937311393695897,"output_mean_power":0.002036037472465405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.102346197011796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006150154421416123,"output_power":0.0006223424424657319,"reference_active":true,"delta_db":0.05143399564773189},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012427322589552618,"output_power":0.00015749167987203654,"reference_active":true,"delta_db":1.0288004345092285},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6589172324725154e-05,"output_power":2.7575340009972047e-05,"reference_active":true,"delta_db":0.15816057390636587}],"active_band_count":3,"active_band_mae_db":0.41279833468777544}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012428224153258587,"log_rms_mae_db":1.6827119019320518,"rms_envelope_nrmse":0.17186088105463804,"envelope_correlation":0.9680420486744894,"rms_gain_error_db":0.3216373608509523,"reference_mean_power":0.0018973976433641784,"output_mean_power":0.0020432527538370607},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.796182335233462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005643605996811222,"output_power":0.0005718815414924322,"reference_active":true,"delta_db":0.05749392866908046},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.691592416494353e-05,"output_power":8.010850456571957e-05,"reference_active":true,"delta_db":1.4844483252859466},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.24640973682173e-06,"output_power":2.5519684196068553e-06,"reference_active":true,"delta_db":0.553863228683541}],"active_band_count":3,"active_band_mae_db":0.6986018275461894}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001274818876901684,"log_rms_mae_db":3.7873363567492415,"rms_envelope_nrmse":0.24894610057577854,"envelope_correlation":0.9688701736712672,"rms_gain_error_db":-1.2182096901707018,"reference_mean_power":0.0003203305006386765,"output_mean_power":0.00024197880266597582},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0949985705721986e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.663152856065708e-05,"output_power":4.4473040269223725e-05,"reference_active":true,"delta_db":-1.7558295658062404},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2062707324852754e-05,"output_power":1.3909168787571825e-05,"reference_active":true,"delta_db":0.6185638656667423},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2084052775211498e-06,"output_power":2.6067664107097087e-06,"reference_active":true,"delta_db":-0.9018710575058877}],"active_band_count":3,"active_band_mae_db":1.0920881629929566}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012438051646139164,"log_rms_mae_db":3.1475514024644062,"rms_envelope_nrmse":0.21331808718410455,"envelope_correlation":0.9801001983853024,"rms_gain_error_db":-1.214960949597614,"reference_mean_power":0.00032140841726426223,"output_mean_power":0.00024297475406858578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.99354395167411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.454820101360057e-05,"output_power":4.5245800633722275e-05,"reference_active":true,"delta_db":-1.5430586524813459},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4052169228627875e-06,"output_power":6.0345505651634485e-06,"reference_active":true,"delta_db":0.4783180291418437},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434599324165249e-07,"output_power":2.552378023525563e-07,"reference_active":true,"delta_db":-1.2893108352219043}],"active_band_count":3,"active_band_mae_db":1.103562505615031}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":4.41794040829654,"rms_envelope_nrmse":0.3510048718832281,"envelope_correlation":0.9566145857865702,"rms_gain_error_db":-2.9172158760720333,"reference_mean_power":5.5432329924189186e-05,"output_mean_power":2.8316628660090573e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.209399643069474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1581941196726705e-05,"output_power":1.2127024905478285e-05,"reference_active":true,"delta_db":-2.50336235413632},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7792321883200752e-05,"output_power":7.885119584592606e-06,"reference_active":true,"delta_db":-3.534243426880974},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1497984756130903e-05,"output_power":4.194400689909476e-06,"reference_active":true,"delta_db":-4.379518123887975}],"active_band_count":3,"active_band_mae_db":3.472374634968423}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":3.8651989745943793,"rms_envelope_nrmse":0.32489514443597695,"envelope_correlation":0.9713479577268458,"rms_gain_error_db":-2.9172222110918415,"reference_mean_power":5.570117934209357e-05,"output_mean_power":2.845392414078676e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.4557792596837174e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8496933872861596e-05,"output_power":1.471127274273242e-05,"reference_active":true,"delta_db":-2.871478873638151},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0110098860775746e-05,"output_power":3.995905910999461e-06,"reference_active":true,"delta_db":-4.031401487005643},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.149890921474854e-06,"output_power":4.2389529874232916e-07,"reference_active":true,"delta_db":-4.333980451771601}],"active_band_count":3,"active_band_mae_db":3.7456202708051314}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"golf","status":"measured","mos":{"status":"measured","value":3.2647178173065186,"predicted_mos":3.2647178173065186,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"14a26e20eb708898c97bf1e126e1efad2d762e7414ed62f52ddcf15b5aeaeefc","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.906503693000559},"started_at":"2026-10-13T05:02:23.286684+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027416879993324983,"native_input_samples":144000,"native_input_sha256":"09e3a6ada6d7398a056800d1c98272bbab4f3ad4d8e21a555e5cea199994b686","output_sha256":"3e6a4809b3a272c20ca74f46f3671f3976473185e4ff6f4910ae4e36567a0ff0","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6737122699996689,"synthesis_s":0.12815089800005808,"total_s":0.801863167999727,"rtf":0.13364386133328784,"output_peak":0.5159177780151367,"output_rms":0.07729864162834342,"finished_at":"2026-10-13T05:02:24.099346+00:00","analysis_stages":{"external_f0_s":0.09291361899977346},"evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"3e6a4809b3a272c20ca74f46f3671f3976473185e4ff6f4910ae4e36567a0ff0","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.233773636957533,"output_peak":0.5162629378870561},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.602036046884486,"spectral_convergence":0.5260643685958416,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.280396490165777,"spectral_convergence":0.18308971781502698,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.5037686203106215,"spectral_convergence":0.1921948093132157,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00406465321878383,"output_power":0.003363105708548889,"delta_db":-0.8228298091053741,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018182795262749823,"output_power":0.001662898923367522,"delta_db":-0.3879479649456949,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007796068737936506,"output_power":0.0006050021434825077,"delta_db":-1.1011874630505685,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031141308104497474,"output_power":0.0002495147790184906,"delta_db":-0.9624057691541874,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.097519376807928e-05,"output_power":7.149339371402758e-05,"delta_db":-0.5408608287585372,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.359594601842185e-05,"output_power":2.690359056419103e-05,"delta_db":-2.096358611669665,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9852650744473378,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7454540890310895,"pitch":null,"output_rms_dbfs":-22.233773636957533,"output_peak":0.5162629378870561,"reference_rms_dbfs":-21.488319547926444,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.795400385786962},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9166173226414612,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08425273787648647,"rms_dbfs":-21.488319547926444,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.933522636704197,"mean":0.0014006173670391407,"zero_energy":false},"output":{"rms":0.07732346693401446,"rms_dbfs":-22.233773636957533,"peak":0.5162629378870561,"peak_dbfs":-5.742581029905067,"crest_factor_db":16.49119260705247,"mean":0.0005570038190234294,"zero_energy":false},"rms_difference_db":-0.7454540890310888},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":2.7942191314694853e-06,"output_rms_dbfs":-55.53739538144442,"output_zero_energy":false,"output_peak":0.007803168526477294,"output_energy_fraction":1.0142718107224032e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":375,"output_estimated_voiced_frames":388,"both_estimated_voiced_frames":366,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.976,"absolute_difference_cents_median":3.830569113411582,"absolute_difference_cents_p95":14.640619226403706,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019068326661124743,"log_rms_mae_db":1.6737567357019463,"rms_envelope_nrmse":0.126667636133389,"envelope_correlation":0.9864991038122198,"rms_gain_error_db":-0.7466544314019745,"reference_mean_power":0.007100685367549582,"output_mean_power":0.005979086360068563},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.775128970261957e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013585900214317641,"output_power":0.0012070172237515435,"reference_active":true,"delta_db":-0.5137495307594125},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026455768120026276,"output_power":0.00022722411656460174,"reference_active":true,"delta_db":-0.6606595182625803},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.853121179279783e-05,"output_power":4.764967897034634e-05,"reference_active":true,"delta_db":-2.169823199618752}],"active_band_count":3,"active_band_mae_db":1.1147440828802482}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001870958079240233,"log_rms_mae_db":1.3192743704684062,"rms_envelope_nrmse":0.09804124034447789,"envelope_correlation":0.9956514861005269,"rms_gain_error_db":-0.7496439297531826,"reference_mean_power":0.007115032028168213,"output_mean_power":0.005987044228033541},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.4229544477081542e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103508455887709,"output_power":0.0010802868204175266,"reference_active":true,"delta_db":-0.4937220023685921},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001243381457136309,"output_power":0.00010395279956488546,"reference_active":true,"delta_db":-0.7776819647646233},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.059131831582797e-06,"output_power":3.5977597388142574e-06,"reference_active":true,"delta_db":-2.26378244307532}],"active_band_count":3,"active_band_mae_db":1.1783954700695116}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00013521834879092463,"log_rms_mae_db":2.0563974606954467,"rms_envelope_nrmse":0.14583077494254187,"envelope_correlation":0.9806708371215891,"rms_gain_error_db":-0.5339708065850808,"reference_mean_power":0.001893869920585969,"output_mean_power":0.0016747618745765313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.102586783082135e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006146611055601692,"output_power":0.000533001090246288,"reference_active":true,"delta_db":-0.6190763539377806},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012383406885837456,"output_power":0.0001040130914859609,"reference_active":true,"delta_db":-0.7575213821536037},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6558892349653455e-05,"output_power":1.7019014226970507e-05,"reference_active":true,"delta_db":-1.9327555729921424}],"active_band_count":3,"active_band_mae_db":1.103117769694509}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012428224153258587,"log_rms_mae_db":1.3109612136426732,"rms_envelope_nrmse":0.10563366677967151,"envelope_correlation":0.9905847646458718,"rms_gain_error_db":-0.5378400773781724,"reference_mean_power":0.001896018402592948,"output_mean_power":0.001675168664340197},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.800083939051905e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005640402218027871,"output_power":0.00048730185069770194,"reference_active":true,"delta_db":-0.6351201393166933},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6600677174574485e-05,"output_power":4.78776374997101e-05,"reference_active":true,"delta_db":-0.7268891495077694},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.240679302115356e-06,"output_power":1.3766399249223855e-06,"reference_active":true,"delta_db":-2.11559341771017}],"active_band_count":3,"active_band_mae_db":1.1592009021782108}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001274818876901684,"log_rms_mae_db":2.0525753636589137,"rms_envelope_nrmse":0.15661885646080245,"envelope_correlation":0.9870699714833512,"rms_gain_error_db":-0.719757029822246,"reference_mean_power":0.0003207966947011442,"output_mean_power":0.0002718029599662059},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0970868199411442e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.668853069391917e-05,"output_power":6.0176706677160936e-05,"reference_active":true,"delta_db":-0.4462273280148748},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2037423481781382e-05,"output_power":9.989135285505097e-06,"reference_active":true,"delta_db":-0.8100564453026424},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2115661332164785e-06,"output_power":2.883532833638935e-06,"reference_active":true,"delta_db":-0.46791968608401857}],"active_band_count":3,"active_band_mae_db":0.5747344864671785}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012438051646139164,"log_rms_mae_db":1.5362480735540547,"rms_envelope_nrmse":0.11211872728934197,"envelope_correlation":0.9955515139155791,"rms_gain_error_db":-0.7171974989178932,"reference_mean_power":0.000321401756636716,"output_mean_power":0.00027247615124114294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0010957368798705e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.456699310182142e-05,"output_power":5.855147453577037e-05,"reference_active":true,"delta_db":-0.4247272505709004},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.389152266358016e-06,"output_power":4.637238985477274e-06,"reference_active":true,"delta_db":-0.6526097608514609},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.430202064245703e-07,"output_power":2.4785269886104216e-07,"reference_active":true,"delta_db":-1.4112605166282408}],"active_band_count":3,"active_band_mae_db":0.8295325093502006}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":2.745540117321345,"rms_envelope_nrmse":0.25732466250131536,"envelope_correlation":0.963031319879359,"rms_gain_error_db":-1.4116024275482393,"reference_mean_power":5.5521773618596246e-05,"output_mean_power":4.011465747439635e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.213750320954792e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.165220363753765e-05,"output_power":1.6576856825742093e-05,"reference_active":true,"delta_db":-1.1599991645995618},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7767922813822933e-05,"output_power":1.3047802273522227e-05,"reference_active":true,"delta_db":-1.3409929213179697},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1475606310272497e-05,"output_power":7.5903293497711655e-06,"reference_active":true,"delta_db":-1.7951501998649988}],"active_band_count":3,"active_band_mae_db":1.4320474285941767}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":2.3500723962891574,"rms_envelope_nrmse":0.19879944856484857,"envelope_correlation":0.9840842198478374,"rms_gain_error_db":-1.4114518639851072,"reference_mean_power":5.5789200397705704e-05,"output_mean_power":4.030927160963084e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.4596634127215415e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8578658869512727e-05,"output_power":2.1389552390191417e-05,"reference_active":true,"delta_db":-1.2584014811974042},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.004970467450939e-05,"output_power":7.1432544779795316e-06,"reference_active":true,"delta_db":-1.4825717744295237},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1495841451622925e-06,"output_power":7.323292749911621e-07,"reference_active":true,"delta_db":-1.9583436998174806}],"active_band_count":3,"active_band_mae_db":1.5664389851481362}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4621644020080566,"predicted_mos":3.4621644020080566,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"fe2d17f6730cf9e775078d2c27500d9c37d8a8893932ef80bc719d4f7cefe495","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5447427639992384},"started_at":"2026-10-13T05:09:32.192176+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032024449992604787,"native_input_samples":144000,"native_input_sha256":"09e3a6ada6d7398a056800d1c98272bbab4f3ad4d8e21a555e5cea199994b686","output_sha256":"bee7fcceb5866e38173bd4b09ab3ccf3a51f6adb2d092ddc0e55996382f01275","output_samples":144120,"length_delta_samples":120,"analysis_s":1.9936541790011688,"synthesis_s":3.2207159390000015,"total_s":5.21437011800117,"rtf":0.8690616863335284,"output_peak":0.5748097896575928,"output_rms":0.08479065334521214,"finished_at":"2026-10-13T05:09:37.416627+00:00","analysis_stages":{"external_f0_s":1.1649147500011168},"evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"bee7fcceb5866e38173bd4b09ab3ccf3a51f6adb2d092ddc0e55996382f01275","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.430554824139357,"output_peak":0.5752844090652348},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.2156950797589055,"spectral_convergence":0.4639079591167971,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.355576099844277,"spectral_convergence":0.19978881059998313,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.108661190622227,"spectral_convergence":0.19158571794608537,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.004270403588850927,"delta_db":0.20988978963511706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0017641902593647067,"delta_db":-0.1296503611735971,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0007804830505707116,"delta_db":-0.00019286569820256425,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.0002681609807694133,"delta_db":-0.6427897628502344,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":7.833586380810975e-05,"delta_db":-0.1379835560536361,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":3.3312324750658855e-05,"delta_db":-1.1622832387892061,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38046492903333223,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.05616735324778798,"pitch":null,"output_rms_dbfs":-21.42976078778235,"output_peak":0.5752844090652348,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5599774567418025},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9113506902074865,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08427593760910047,"rms_dbfs":-21.485928141030136,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.931131229807892,"mean":0.001402181625366211,"zero_energy":false},"output":{"rms":0.08482267425703183,"rms_dbfs":-21.42976078778235,"peak":0.5752844090652348,"peak_dbfs":-4.80234791489024,"crest_factor_db":16.627412872892112,"mean":0.00020388833247065468,"zero_energy":false},"rms_difference_db":0.05616735324778688},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":1.1410489618609844e-05,"output_rms_dbfs":-49.42695719815062,"output_zero_energy":false,"output_peak":0.04791554191194511,"output_energy_fraction":3.4361527441763596e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":376,"output_estimated_voiced_frames":379,"both_estimated_voiced_frames":364,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9680851063829787,"absolute_difference_cents_median":5.3806533363608455,"absolute_difference_cents_p95":30.21375269212049,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019068326661124743,"log_rms_mae_db":1.9559411119053707,"rms_envelope_nrmse":0.13418285845248376,"envelope_correlation":0.9754320342617199,"rms_gain_error_db":0.0569947449168494,"reference_mean_power":0.007104543911662494,"output_mean_power":0.007198395053249323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.7734542191089184e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001358301262786015,"output_power":0.0013869534717915865,"reference_active":true,"delta_db":0.09065787568024997},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002660490997406596,"output_power":0.0002566334707267268,"reference_active":true,"delta_db":-0.15648496276864138},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.883599774152297e-05,"output_power":6.66519453163642e-05,"reference_active":true,"delta_db":-0.7291173905028213}],"active_band_count":3,"active_band_mae_db":0.3254200763172375}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001870958079240233,"log_rms_mae_db":1.238702892630677,"rms_envelope_nrmse":0.0775134916082836,"envelope_correlation":0.990685338813346,"rms_gain_error_db":0.05815024908486638,"reference_mean_power":0.007117638239873263,"output_mean_power":0.007213581381576694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4206220628648626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103399895685935,"output_power":0.0012346355663327042,"reference_active":true,"delta_db":0.08631401006131433},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012555497766250794,"output_power":0.00011989628840913532,"reference_active":true,"delta_db":-0.2002819612524928},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0918852450489685e-06,"output_power":5.387229191383072e-06,"reference_active":true,"delta_db":-0.5338626145557875}],"active_band_count":3,"active_band_mae_db":0.2734861952898649}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013521834879092463,"log_rms_mae_db":2.3652303072674643,"rms_envelope_nrmse":0.17466984597755347,"envelope_correlation":0.9687172681004395,"rms_gain_error_db":-0.027026187944654112,"reference_mean_power":0.0018937311393695897,"output_mean_power":0.0018819830242592394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.102346197011796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006150154421416123,"output_power":0.0006187890377402268,"reference_active":true,"delta_db":0.026565908656469624},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012427322589552618,"output_power":0.00011243179893287105,"reference_active":true,"delta_db":-0.4348841241312915},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6589172324725154e-05,"output_power":2.4941805249013005e-05,"reference_active":true,"delta_db":-0.2777693486228149}],"active_band_count":3,"active_band_mae_db":0.24640646047019202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012428224153258587,"log_rms_mae_db":1.509726157211071,"rms_envelope_nrmse":0.10670003478822876,"envelope_correlation":0.9866658545671904,"rms_gain_error_db":-0.025524812092152772,"reference_mean_power":0.0018973976433641784,"output_mean_power":0.0018862787648889763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.796182335233462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005643605996811222,"output_power":0.0005650913834620817,"reference_active":true,"delta_db":0.005619989385306914},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.691592416494353e-05,"output_power":5.1336467989313674e-05,"reference_active":true,"delta_db":-0.448078066627442},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.24640973682173e-06,"output_power":2.2345068177445072e-06,"reference_active":true,"delta_db":-0.023072887743630865}],"active_band_count":3,"active_band_mae_db":0.15892364791879324}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001274818876901684,"log_rms_mae_db":2.318438303833753,"rms_envelope_nrmse":0.2179588806474393,"envelope_correlation":0.9688377140430986,"rms_gain_error_db":-0.5758411573307524,"reference_mean_power":0.0003203305006386765,"output_mean_power":0.0002805525217704962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0949985705721986e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.663152856065708e-05,"output_power":6.147393268408707e-05,"reference_active":true,"delta_db":-0.34988779098868417},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2062707324852754e-05,"output_power":1.0629967268761207e-05,"reference_active":true,"delta_db":-0.5491286347841619},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2084052775211498e-06,"output_power":3.108088369985156e-06,"reference_active":true,"delta_db":-0.1379586376245725}],"active_band_count":3,"active_band_mae_db":0.3456583544658061}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012438051646139164,"log_rms_mae_db":1.3328117398419068,"rms_envelope_nrmse":0.14510068784983826,"envelope_correlation":0.9873829340868556,"rms_gain_error_db":-0.5738430889139843,"reference_mean_power":0.00032140841726426223,"output_mean_power":0.0002816261235044229},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.99354395167411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.454820101360057e-05,"output_power":5.922885275574676e-05,"reference_active":true,"delta_db":-0.3735082222039342},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4052169228627875e-06,"output_power":5.023720792582031e-06,"reference_active":true,"delta_db":-0.3178763353916102},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434599324165249e-07,"output_power":3.4141519062809575e-07,"reference_active":true,"delta_db":-0.02593239801701691}],"active_band_count":3,"active_band_mae_db":0.23910565187085378}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":2.461276592428251,"rms_envelope_nrmse":0.37401016604437337,"envelope_correlation":0.8916168094699661,"rms_gain_error_db":-0.6574853862597368,"reference_mean_power":5.5432329924189186e-05,"output_mean_power":4.7644699763752294e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.209399643069474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1581941196726705e-05,"output_power":1.939194305958843e-05,"reference_active":true,"delta_db":-0.4646917763000261},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7792321883200752e-05,"output_power":1.5012090026823879e-05,"reference_active":true,"delta_db":-0.7379146665759339},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1497984756130903e-05,"output_power":8.064515953946193e-06,"reference_active":true,"delta_db":-1.5404342311978305}],"active_band_count":3,"active_band_mae_db":0.9143468913579301}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":1.5362532627890753,"rms_envelope_nrmse":0.20899019552735526,"envelope_correlation":0.9647693430092992,"rms_gain_error_db":-0.6575053732064824,"reference_mean_power":5.570117934209357e-05,"output_mean_power":4.7875558475904836e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.4557792596837174e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8496933872861596e-05,"output_power":2.4604929963618062e-05,"reference_active":true,"delta_db":-0.637760014344332},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0110098860775746e-05,"output_power":7.977349459793236e-06,"reference_active":true,"delta_db":-1.0289678493384637},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.149890921474854e-06,"output_power":8.105458925271022e-07,"reference_active":true,"delta_db":-1.5187903585836646}],"active_band_count":3,"active_band_mae_db":1.0618394074221535}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.70059871673584,"predicted_mos":3.70059871673584,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"2404d601b9ee20c73a14472156fcdb3b9498edbdd7ce67895d5f7899593faa42","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.204254162001234},"started_at":"2026-10-13T05:10:31.107372+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00328662100037036,"native_input_samples":144000,"native_input_sha256":"09e3a6ada6d7398a056800d1c98272bbab4f3ad4d8e21a555e5cea199994b686","output_sha256":"8fadf7d96c45b68d59f3523fbf35d7acfac1a2feab2737d007a33bfc4f0dad8c","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0102604100011376,"synthesis_s":0.16385801599972183,"total_s":1.1741184260008595,"rtf":0.1956864043334766,"output_peak":0.6650905609130859,"output_rms":0.08305089828873113,"finished_at":"2026-10-13T05:10:32.299809+00:00","analysis_stages":{"external_f0_s":0.15691510499891592},"evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"8fadf7d96c45b68d59f3523fbf35d7acfac1a2feab2737d007a33bfc4f0dad8c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.610519725191217,"output_peak":0.6654312781501214},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.561496319867138,"spectral_convergence":0.11037599869378302,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.308276673507719,"spectral_convergence":0.09820943369926717,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.762161890339916,"spectral_convergence":0.10867132934618565,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.004068927413639468,"output_power":0.003993882469234807,"delta_db":-0.08084662025454406,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018176508169902323,"output_power":0.0017416950338592482,"delta_db":-0.18538342466300894,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007805177117872849,"output_power":0.0007629332612914719,"delta_db":-0.09896213607871517,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000310938544728363,"output_power":0.0003039767104766153,"delta_db":-0.09834250707766046,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.08647023559392e-05,"output_power":7.398009347553022e-05,"delta_db":-0.3864411663014792,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.353447039170603e-05,"output_power":3.61795822805488e-05,"delta_db":-0.8036971721219077,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2756121710828859,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1176279659019149,"pitch":null,"output_rms_dbfs":-21.603556106932054,"output_peak":0.6654312781501214,"reference_rms_dbfs":-21.485928141030136,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.543978294571591},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9650826310885681,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08427593760910047,"rms_dbfs":-21.485928141030136,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.931131229807892,"mean":0.001402181625366211,"zero_energy":false},"output":{"rms":0.08314233067829059,"rms_dbfs":-21.603556106932054,"peak":0.6654312781501214,"peak_dbfs":-3.537935785341968,"crest_factor_db":18.065620321590085,"mean":-8.748077870445896e-05,"zero_energy":false},"rms_difference_db":-0.11762796590191726},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":4.887196034585465e-06,"output_rms_dbfs":-53.10940240257441,"output_zero_energy":false,"output_peak":0.020036924851550477,"output_energy_fraction":1.531819071750948e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":376,"output_estimated_voiced_frames":379,"both_estimated_voiced_frames":375,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9973404255319149,"absolute_difference_cents_median":0.9167771849483017,"absolute_difference_cents_p95":7.361860183298548,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019068326661124743,"log_rms_mae_db":0.6719200566061391,"rms_envelope_nrmse":0.039564511661488624,"envelope_correlation":0.9980648275014806,"rms_gain_error_db":-0.11689721946814205,"reference_mean_power":0.007104543911662494,"output_mean_power":0.006915864583166223},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.7734542191089184e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001358301262786015,"output_power":0.0013366966666584423,"reference_active":true,"delta_db":-0.06963239401341086},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002660490997406596,"output_power":0.0002433629897966377,"reference_active":true,"delta_db":-0.38707261497356626},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.883599774152297e-05,"output_power":5.337448640156881e-05,"reference_active":true,"delta_db":-1.6939085935184537}],"active_band_count":3,"active_band_mae_db":0.7168712008351436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001870958079240233,"log_rms_mae_db":0.38969964174642113,"rms_envelope_nrmse":0.025767726677232938,"envelope_correlation":0.999259278654848,"rms_gain_error_db":-0.11622593960779615,"reference_mean_power":0.007117638239873263,"output_mean_power":0.006929682181658911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4206220628648626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103399895685935,"output_power":0.001199441854426157,"reference_active":true,"delta_db":-0.039281831548373336},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012555497766250794,"output_power":0.00011514774939950349,"reference_active":true,"delta_db":-0.3757848116226463},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0918852450489685e-06,"output_power":4.037947213477727e-06,"reference_active":true,"delta_db":-1.7859107650464592}],"active_band_count":3,"active_band_mae_db":0.733659136072493}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013521834879092463,"log_rms_mae_db":0.9837226341482836,"rms_envelope_nrmse":0.05530544991455966,"envelope_correlation":0.9970886253498668,"rms_gain_error_db":-0.14070763435617756,"reference_mean_power":0.0018937311393695897,"output_mean_power":0.0018333591782225492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.102346197011796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006150154421416123,"output_power":0.0006061131096867508,"reference_active":true,"delta_db":-0.06332342899493928},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012427322589552618,"output_power":0.00011009864308611473,"reference_active":true,"delta_db":-0.5259560540757352},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6589172324725154e-05,"output_power":1.8264107381317237e-05,"reference_active":true,"delta_db":-1.6310636683593733}],"active_band_count":3,"active_band_mae_db":0.7401143838100159}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012428224153258587,"log_rms_mae_db":0.5475073618271699,"rms_envelope_nrmse":0.035640837874937915,"envelope_correlation":0.9988003401237444,"rms_gain_error_db":-0.14038598512232078,"reference_mean_power":0.0018973976433641784,"output_mean_power":0.0018370448454792221},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.796182335233462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005643605996811222,"output_power":0.0005552880787146806,"reference_active":true,"delta_db":-0.07038336412270879},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.691592416494353e-05,"output_power":4.982003676324783e-05,"reference_active":true,"delta_db":-0.5782974836819073},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.24640973682173e-06,"output_power":1.4851530134484314e-06,"reference_active":true,"delta_db":-1.797177720835389}],"active_band_count":3,"active_band_mae_db":0.8152861895466682}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001274818876901684,"log_rms_mae_db":1.1154515061775763,"rms_envelope_nrmse":0.07988461535467914,"envelope_correlation":0.9958798414080701,"rms_gain_error_db":-0.13768987217959477,"reference_mean_power":0.0003203305006386765,"output_mean_power":0.0003103339614427519},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0949985705721986e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.663152856065708e-05,"output_power":6.284512884116832e-05,"reference_active":true,"delta_db":-0.25408155381586117},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2062707324852754e-05,"output_power":1.0862690194491968e-05,"reference_active":true,"delta_db":-0.4550739717557822},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2084052775211498e-06,"output_power":2.5089657438986553e-06,"reference_active":true,"delta_db":-1.0679449031550423}],"active_band_count":3,"active_band_mae_db":0.5923668095755619}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012438051646139164,"log_rms_mae_db":0.7211549989876176,"rms_envelope_nrmse":0.04750678445117487,"envelope_correlation":0.9986646696770632,"rms_gain_error_db":-0.13807027928911542,"reference_mean_power":0.00032140841726426223,"output_mean_power":0.000311350966533608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.99354395167411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.454820101360057e-05,"output_power":6.091853316725848e-05,"reference_active":true,"delta_db":-0.2513470520667738},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4052169228627875e-06,"output_power":4.814199920325141e-06,"reference_active":true,"delta_db":-0.5028900644652294},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.434599324165249e-07,"output_power":2.5278782630596945e-07,"reference_active":true,"delta_db":-1.331199246611587}],"active_band_count":3,"active_band_mae_db":0.6951454543811968}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":1.1859389843020702,"rms_envelope_nrmse":0.19232736659642516,"envelope_correlation":0.9741706357652339,"rms_gain_error_db":-0.6107228866641661,"reference_mean_power":5.5432329924189186e-05,"output_mean_power":4.816048418866696e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.209399643069474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1581941196726705e-05,"output_power":1.943430220809773e-05,"reference_active":true,"delta_db":-0.45521553072040655},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7792321883200752e-05,"output_power":1.5689358031801687e-05,"reference_active":true,"delta_db":-0.5462745311406706},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1497984756130903e-05,"output_power":6.062913819019891e-06,"reference_active":true,"delta_db":-2.7794033350670597}],"active_band_count":3,"active_band_mae_db":1.2602977989760455}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":0.7987933898494035,"rms_envelope_nrmse":0.1199469901712392,"envelope_correlation":0.9911453925665281,"rms_gain_error_db":-0.6106837091967212,"reference_mean_power":5.570117934209357e-05,"output_mean_power":4.839450137352201e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.4557792596837174e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8496933872861596e-05,"output_power":2.432249912796858e-05,"reference_active":true,"delta_db":-0.6878993812046018},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0110098860775746e-05,"output_power":7.677004384580074e-06,"reference_active":true,"delta_db":-1.1956361367863833},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.149890921474854e-06,"output_power":6.007662987771466e-07,"reference_active":true,"delta_db":-2.81951083145131}],"active_band_count":3,"active_band_mae_db":1.567682116480765}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3570-5696-0010","method":"magphase","status":"measured","mos":{"status":"measured","value":3.5216610431671143,"predicted_mos":3.5216610431671143,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287219,"input_sample_rate":48000,"input_seconds":5.9837291666666665,"input_float32_sha256":"111ba7663263eb40102ad016b2bed3e310a93e86fb7ba2d2b0d2465d7e240cbb","internal_samples":95740,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.624076086000059},"started_at":"2026-10-13T05:14:37.940658+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009561580000081449,"native_input_samples":288000,"native_input_sha256":"6859b5d928374b58fbf6a7f936688126eb928bd9497ca98e36055086d506785e","output_sha256":"cb9d0243e80c322564777df6f9bb92a83adcfb5ef8c1a640a6dc1cc16f469317","output_samples":287219,"length_delta_samples":-781,"analysis_s":3.2155196019994037,"synthesis_s":1.016294511000524,"total_s":4.231814112999928,"rtf":0.7053023521666546,"output_peak":0.7829122371572953,"output_rms":0.08034456967239936,"finished_at":"2026-10-13T05:14:42.193539+00:00","evaluation_input":{"reference":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"output_sha256":"cb9d0243e80c322564777df6f9bb92a83adcfb5ef8c1a640a6dc1cc16f469317","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95740,"length_delta_samples":-260,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.898259650959027,"output_peak":0.7492096104737228},"common_prefix":{"samples":95740,"sample_rate":16000,"reference_coverage":0.9972916666666667,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95740,"output_samples":95740,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.7046369160631105,"spectral_convergence":0.37975080896771785,"spectral_convergence_status":"ok","amplitude_floor":1.1745223847485646e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.0895705830461315,"spectral_convergence":0.20811005450797598,"spectral_convergence_status":"ok","amplitude_floor":9.481908348017771e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.632309000771408,"spectral_convergence":0.18158146959160715,"spectral_convergence_status":"ok","amplitude_floor":8.395402116778918e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00406813894672741,"output_power":0.0035774600557516864,"delta_db":-0.5582098444293795,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0018135479177784969,"output_power":0.0016986274473542794,"delta_db":-0.2843089767447199,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007791191118680645,"output_power":0.0007275209216193848,"delta_db":-0.2975837074260225,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003118473830540254,"output_power":0.0003196873404114073,"delta_db":0.10783334680837736,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.10706186124318e-05,"output_power":9.02239010133688e-05,"delta_db":0.46458113840042414,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3672299096325545e-05,"output_power":4.561047400384956e-05,"delta_db":0.18858530001879234,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3168503856379526,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.40925019467970086,"pitch":null,"output_rms_dbfs":-21.898259650959027,"output_peak":0.7492096104737228,"reference_rms_dbfs":-21.489009456279327,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.1421721666268825},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95740,"output_samples":95740,"intelligibility":{"estoi":{"value":0.8875084457637622,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08424604606233314,"rms_dbfs":-21.489009456279327,"peak":0.74517822265625,"peak_dbfs":-2.5547969112222466,"crest_factor_db":18.93421254505708,"mean":0.0014099570537923346,"zero_energy":false},"output":{"rms":0.08036871368980579,"rms_dbfs":-21.898259650959027,"peak":0.7492096104737228,"peak_dbfs":-2.507933207117621,"crest_factor_db":19.390326443841406,"mean":-0.0005423035634917226,"zero_energy":false},"rms_difference_db":-0.40925019467970003},"full_scale":{"reference":{"sample_count":95740,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95740,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":60,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.6163352150470022e-06,"threshold_dbfs":-54.417313195413605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":13,"quiet_samples":2080,"reference_mean_power":3.003840370533558e-06,"output_mean_power":2.4851379639999765e-06,"output_rms_dbfs":-56.04649496129564,"output_zero_energy":false,"output_peak":0.012756488572990471,"output_energy_fraction":8.358847092156855e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":374,"output_estimated_voiced_frames":368,"both_estimated_voiced_frames":362,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9679144385026738,"absolute_difference_cents_median":5.4712663855496295,"absolute_difference_cents_p95":25.481776828510153,"difference_over_600_cents_frames":0,"uncovered_tail_samples":60,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95740,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1195,"rms_floor":0.00019068326661124743,"log_rms_mae_db":1.7366558898200273,"rms_envelope_nrmse":0.14542624481029684,"envelope_correlation":0.9728351110922252,"rms_gain_error_db":-0.40707799406666645,"reference_mean_power":0.0070978229948091545,"output_mean_power":0.006462750133726981},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":2.776890244327359e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013589070596448866,"output_power":0.0012538190033047558,"reference_active":true,"delta_db":-0.3495490697924419},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026303187758845074,"output_power":0.0002492865645620558,"reference_active":true,"delta_db":-0.23309512482568154},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.815569748652546e-05,"output_power":6.993111452401391e-05,"reference_active":true,"delta_db":-0.48290193378033697}],"active_band_count":3,"active_band_mae_db":0.3551820427994868}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1189,"rms_floor":0.0001870958079240233,"log_rms_mae_db":1.1571013141389321,"rms_envelope_nrmse":0.09226844832402041,"envelope_correlation":0.9895699847075506,"rms_gain_error_db":-0.4086731913420485,"reference_mean_power":0.007113407016619202,"output_mean_power":0.006474561192543863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":2.425377919784419e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0012103132940805273,"output_power":0.0011174356454539164,"reference_active":true,"delta_db":-0.3467528265446935},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012302412890784672,"output_power":0.00011537737293104388,"reference_active":true,"delta_db":-0.278696524977631},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.168404979028706e-06,"output_power":5.480889669950983e-06,"reference_active":true,"delta_db":-0.5132181921517011}],"active_band_count":3,"active_band_mae_db":0.3795558478913419}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1195,"rms_floor":0.00013521834879092463,"log_rms_mae_db":2.3156733145783734,"rms_envelope_nrmse":0.18088203906668904,"envelope_correlation":0.9666147826854875,"rms_gain_error_db":-0.25440189900479104,"reference_mean_power":0.001892801223278294,"output_mean_power":0.001785109322570511},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.1027373593712692e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006143061369332651,"output_power":0.000576091593302631,"reference_active":true,"delta_db":-0.2789331618405121},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001233780815818339,"output_power":0.00011464560791027447,"reference_active":true,"delta_db":-0.3188059157140468},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6508966559059777e-05,"output_power":2.3170613172645193e-05,"reference_active":true,"delta_db":-0.5845527043476826}],"active_band_count":3,"active_band_mae_db":0.39409726063408046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1189,"rms_floor":0.00012428224153258587,"log_rms_mae_db":1.6520743574915764,"rms_envelope_nrmse":0.11568661749228357,"envelope_correlation":0.9850203814480858,"rms_gain_error_db":-0.25611872888826287,"reference_mean_power":0.0018943958520854863,"output_mean_power":0.0017859070892798578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":9.803772858520211e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005637097406109357,"output_power":0.0005305250028441529,"reference_active":true,"delta_db":-0.26349682903293303},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.625917009551748e-05,"output_power":5.215078495644617e-05,"reference_active":true,"delta_db":-0.3293247173264858},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2597357603671865e-06,"output_power":2.0559497951962766e-06,"reference_active":true,"delta_db":-0.41045153088423153}],"active_band_count":3,"active_band_mae_db":0.3344243590812168}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1195,"rms_floor":0.0001274818876901684,"log_rms_mae_db":2.232492402820909,"rms_envelope_nrmse":0.19923910891683708,"envelope_correlation":0.9742739014335126,"rms_gain_error_db":0.15569808179090708,"reference_mean_power":0.00032090653399024263,"output_mean_power":0.0003326200067632951},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.0988707692269039e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.674412509390585e-05,"output_power":7.046835090688796e-05,"reference_active":true,"delta_db":0.23581063756622433},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2011090449611565e-05,"output_power":1.2564295017411522e-05,"reference_active":true,"delta_db":0.1955568790853865},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.208508241073215e-06,"output_power":3.0459392385608643e-06,"reference_active":true,"delta_db":-0.22581923485331645}],"active_band_count":3,"active_band_mae_db":0.2190622505016424}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1189,"rms_floor":0.00012438051646139164,"log_rms_mae_db":1.3866493497185848,"rms_envelope_nrmse":0.1247386208883857,"envelope_correlation":0.989598683894661,"rms_gain_error_db":0.15782347145894,"reference_mean_power":0.0003212861913636893,"output_mean_power":0.00033317653514408093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":1.0027770234865758e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.458336099374133e-05,"output_power":6.814735509888318e-05,"reference_active":true,"delta_db":0.23328362414501483},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.369808185663129e-06,"output_power":5.5607052961058085e-06,"reference_active":true,"delta_db":0.15171106542099588},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.459266537741778e-07,"output_power":3.31164963675061e-07,"reference_active":true,"delta_db":-0.18939642446670768}],"active_band_count":3,"active_band_mae_db":0.1914637046775728}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1195,"rms_floor":6.885927766129989e-05,"log_rms_mae_db":2.3424160974655766,"rms_envelope_nrmse":0.3844999926140304,"envelope_correlation":0.8967051971833965,"rms_gain_error_db":0.38896471526383175,"reference_mean_power":5.56093576642216e-05,"output_mean_power":6.081971072045163e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":6.218054974928163e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1722695933400107e-05,"output_power":2.4063479817332445e-05,"reference_active":true,"delta_db":0.4444470768293418},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.774156145316298e-05,"output_power":1.878838300339119e-05,"reference_active":true,"delta_db":0.24897564861300572},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1451864888216117e-05,"output_power":1.068092276224586e-05,"reference_active":true,"delta_db":-0.30267440940305207}],"active_band_count":3,"active_band_mae_db":0.3320323782817998}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":60,"metrics":{"frame_count":1189,"rms_floor":3.785435651948458e-05,"log_rms_mae_db":1.469230200542122,"rms_envelope_nrmse":0.21472156765301298,"envelope_correlation":0.9644763220495076,"rms_gain_error_db":0.3887703012258404,"reference_mean_power":5.5876566954598054e-05,"output_mean_power":6.110922070482684e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":5.463479592394092e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8658449025996726e-05,"output_power":3.085556665219641e-05,"reference_active":true,"delta_db":0.3208084349494673},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.986609065003426e-06,"output_power":1.0107939688120473e-05,"reference_active":true,"delta_db":0.05244592562650884},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1534443683083378e-06,"output_power":1.0338599875671823e-06,"reference_active":true,"delta_db":-0.47534925343263973}],"active_band_count":3,"active_band_mae_db":0.28286787133620533}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"hnm","status":"measured","mos":{"status":"measured","value":2.3577494621276855,"predicted_mos":2.3577494621276855,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":55920,"input_sample_rate":16000,"input_seconds":3.495,"input_float32_sha256":"181dca288f51fb8cd51e46cc092b8a28d47fd473fd18bd0f2bacb9bf84dbe882","internal_samples":55920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2953336240007047},"started_at":"2026-10-13T04:38:58.535642+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3001000297663268e-05,"native_input_samples":55920,"native_input_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","output_sha256":"d184f7fa65b9ce65bcf592f9e0aba32eacab5f96b9c4ccb2bacff813403e1ae8","output_samples":55920,"length_delta_samples":0,"analysis_s":0.6193126450007185,"synthesis_s":0.2878324959992824,"total_s":0.9071451410000009,"rtf":0.2595551190271819,"output_peak":0.8706827940478363,"output_rms":0.10194094006714141,"finished_at":"2026-10-13T04:38:59.446401+00:00","evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"d184f7fa65b9ce65bcf592f9e0aba32eacab5f96b9c4ccb2bacff813403e1ae8","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.367333589961612,"spectral_convergence":0.39069125979072017,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.30783973295878,"spectral_convergence":0.38034616584761105,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.674857319987056,"spectral_convergence":0.34071812496242276,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.0016068729083638534,"delta_db":-0.8969086141544829,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.005225438482398725,"delta_db":0.18000886261160193,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0014387024809128465,"delta_db":-0.2301650772051858,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.0018205404030462746,"delta_db":-0.0943506624234047,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.00023958935221933325,"delta_db":-0.2513892650943416,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":6.081163483148687e-05,"delta_db":-0.19230451163699963,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3075211655210028,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1202526087006569,"pitch":null,"output_rms_dbfs":-19.833027316046923,"output_peak":0.8706827940478363,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.450010214302483},"common_prefix":{"samples":55920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.367333589961612,"spectral_convergence":0.39069125979072017,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.30783973295878,"spectral_convergence":0.38034616584761105,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.674857319987056,"spectral_convergence":0.34071812496242276,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.0016068729083638534,"delta_db":-0.8969086141544829,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.005225438482398725,"delta_db":0.18000886261160193,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0014387024809128465,"delta_db":-0.2301650772051858,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.0018205404030462746,"delta_db":-0.0943506624234047,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.00023958935221933325,"delta_db":-0.2513892650943416,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":6.081163483148687e-05,"delta_db":-0.19230451163699963,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3075211655210028,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1202526087006569,"pitch":null,"output_rms_dbfs":-19.833027316046923,"output_peak":0.8706827940478363,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.450010214302483},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55920,"output_samples":55920,"intelligibility":{"estoi":{"value":0.9105567129805101,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10336208580657574,"rms_dbfs":-19.712774707346266,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.2877957854351,"mean":-4.563447572983727e-06,"zero_energy":false},"output":{"rms":0.10194094006714141,"rms_dbfs":-19.833027316046923,"peak":0.8706827940478363,"peak_dbfs":-1.2028007548643205,"crest_factor_db":18.630226561182607,"mean":0.0005131775567802661,"zero_energy":false},"rms_difference_db":-0.1202526087006568},"full_scale":{"reference":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":349,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.2266614618871747e-05,"output_mean_power":1.3386954679541426e-05,"output_rms_dbfs":-48.73318206878267,"output_zero_energy":false,"output_peak":0.017198551059947623,"output_energy_fraction":0.00021377914618661713,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":342,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9481865284974094,"absolute_difference_cents_median":1.847472499899886,"absolute_difference_cents_p95":10.61491744613175,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.0006292775937517623,"log_rms_mae_db":0.7535801702215891,"rms_envelope_nrmse":0.1598382516763493,"envelope_correlation":0.970603385997405,"rms_gain_error_db":-0.1202533431681509,"reference_mean_power":0.010699017072596153,"output_mean_power":0.01040683206028842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":4.048774453403744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014613887448305465,"output_power":0.0014585541620709011,"reference_active":true,"delta_db":-0.008431973343105106},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0008261048603288557,"output_power":0.000802503772772304,"reference_active":true,"delta_db":-0.12588094501007205},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.0004730285880207083,"output_power":0.00025570128341936035,"reference_active":true,"delta_db":-2.6715448076335333}],"active_band_count":3,"active_band_mae_db":0.9352859086622368}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00037792225909995443,"log_rms_mae_db":0.4373356814770976,"rms_envelope_nrmse":0.07548096905332058,"envelope_correlation":0.9929866329268198,"rms_gain_error_db":-0.12025748158870363,"reference_mean_power":0.01079168973829664,"output_mean_power":0.010496963877414037},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":3.4667222612961444e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016650159461876711,"output_power":0.001584649072868138,"reference_active":true,"delta_db":-0.21485296305428794},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005334276556675961,"output_power":0.0004552625013451411,"reference_active":true,"delta_db":-0.6881364782362859},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9068036163681986e-05,"output_power":1.7855224203570668e-05,"reference_active":true,"delta_db":-3.4005627316566334}],"active_band_count":3,"active_band_mae_db":1.4345173909824025}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00018013905998130842,"log_rms_mae_db":1.4197807672218135,"rms_envelope_nrmse":0.11283938079581145,"envelope_correlation":0.9881031865504838,"rms_gain_error_db":-0.170294070586863,"reference_mean_power":0.002043548171012202,"output_mean_power":0.001964967955797862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.1637162379955263e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006991910792272858,"output_power":0.0006385776149557982,"reference_active":true,"delta_db":-0.3938218814931777},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001302332633123258,"output_power":0.00010780559740893675,"reference_active":true,"delta_db":-0.8208061242414609},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.350140195021887e-05,"output_power":2.100119068914354e-05,"reference_active":true,"delta_db":-2.0281906325089865}],"active_band_count":3,"active_band_mae_db":1.0809395460812083}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00014142899773968785,"log_rms_mae_db":0.9110228099050116,"rms_envelope_nrmse":0.06853530281388753,"envelope_correlation":0.9956545738694321,"rms_gain_error_db":-0.1702993371827385,"reference_mean_power":0.002061256155357224,"output_mean_power":0.0019819926144683314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.0272195933213497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006464643086637218,"output_power":0.0005856949197886011,"reference_active":true,"delta_db":-0.4287309539905211},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.880640785962183e-05,"output_power":5.530558616092748e-05,"reference_active":true,"delta_db":-0.9485988585893702},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.140839173745677e-06,"output_power":1.1292797696038707e-06,"reference_active":true,"delta_db":-2.777824949419378}],"active_band_count":3,"active_band_mae_db":1.3850515873330898}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00022296985761207554,"log_rms_mae_db":1.2832124550151174,"rms_envelope_nrmse":0.10180036270421697,"envelope_correlation":0.9932103819215881,"rms_gain_error_db":-0.10468619485761749,"reference_mean_power":0.001891605546476507,"output_mean_power":0.0018465537735626789},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.5501990909593355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010291296692043158,"output_power":0.0009462226973348882,"reference_active":true,"delta_db":-0.36476737443445584},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.517530440244301e-05,"output_power":5.2405264369691005e-05,"reference_active":true,"delta_db":-0.22369821627358538},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2165950116970767e-05,"output_power":1.960423933535796e-05,"reference_active":true,"delta_db":2.0720396446156397}],"active_band_count":3,"active_band_mae_db":0.886835078441227}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00021618630891631995,"log_rms_mae_db":0.8340222637215726,"rms_envelope_nrmse":0.05760997284999655,"envelope_correlation":0.9978259160658799,"rms_gain_error_db":-0.10469362041590319,"reference_mean_power":0.0019079864948334635,"output_mean_power":0.001862541397464969},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.466045308126094e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009690551708671024,"output_power":0.0008929867746141454,"reference_active":true,"delta_db":-0.35503476372620124},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1519400367413303e-05,"output_power":2.153905815127447e-05,"reference_active":true,"delta_db":0.00396543115763245},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.641667021707966e-07,"output_power":7.887654367549116e-07,"reference_active":true,"delta_db":0.13759762016910276}],"active_band_count":3,"active_band_mae_db":0.16553260501764547}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":1.5327435358556132,"rms_envelope_nrmse":0.1766752010773801,"envelope_correlation":0.9739132147048357,"rms_gain_error_db":-0.08997029041814858,"reference_mean_power":6.674409144257211e-05,"output_mean_power":6.537561637868208e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":5.3173969334771903e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5059703710589825e-05,"output_power":3.2970238642406615e-05,"reference_active":true,"delta_db":-0.2668615092412626},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3101289173881222e-06,"output_power":2.3662111146546745e-06,"reference_active":true,"delta_db":0.10417273519644983},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.893247149634713e-07,"output_power":1.5469502207743287e-06,"reference_active":true,"delta_db":1.9413748018053647}],"active_band_count":3,"active_band_mae_db":0.770803015414359}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":1.027874625632889,"rms_envelope_nrmse":0.1103159040709041,"envelope_correlation":0.988999985773568,"rms_gain_error_db":-0.09003643966432348,"reference_mean_power":6.730512481108481e-05,"output_mean_power":6.592414257748058e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":4.920971050071521e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3235368479868e-05,"output_power":3.1592076675424804e-05,"reference_active":true,"delta_db":-0.22022323368645358},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2494288246616962e-06,"output_power":1.1286232706615166e-06,"reference_active":true,"delta_db":-0.44162520435019503},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.10110394125166e-08,"output_power":1.029087784398827e-07,"reference_active":true,"delta_db":1.0390821859188377}],"active_band_count":3,"active_band_mae_db":0.5669768746518288}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.931370735168457,"predicted_mos":2.931370735168457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":55920,"input_sample_rate":16000,"input_seconds":3.495,"input_float32_sha256":"17f01357e7ce7f55c5fd412f88779ee82901b568b653731e7780d8dd3077efd1","internal_samples":55920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.252330179000637},"started_at":"2026-10-13T04:41:56.633293+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3780999981681816e-05,"native_input_samples":55920,"native_input_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","output_sha256":"a1caff47a12d7a8448950e94159cd358a8ea48142592dd8da8812fce2b2faea8","output_samples":55920,"length_delta_samples":0,"analysis_s":0.7280248249999204,"synthesis_s":0.05982640199999878,"total_s":0.7878512269999192,"rtf":0.22542238254647184,"output_peak":0.9567216349585681,"output_rms":0.11245432605594011,"finished_at":"2026-10-13T04:41:57.425088+00:00","evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"a1caff47a12d7a8448950e94159cd358a8ea48142592dd8da8812fce2b2faea8","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.005665980411654,"spectral_convergence":0.5563612401655409,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.382698133840426,"spectral_convergence":0.4834688177665305,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.948255643836438,"spectral_convergence":0.4833958509381798,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.00039399860715330165,"delta_db":-7.0017770356044915,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.007309863242506487,"delta_db":1.6378739896226264,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0021046838981186255,"delta_db":1.4219937827596647,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.002435856384528621,"delta_db":1.1701629131296845,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.00032789423764174994,"delta_db":1.111273446885384,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":7.367907874690312e-05,"delta_db":0.6412705687938491,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.1640586227992835,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7322980505934034,"pitch":null,"output_rms_dbfs":-18.980476656752863,"output_peak":0.9567216349585681,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.778873252696172},"common_prefix":{"samples":55920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.005665980411654,"spectral_convergence":0.5563612401655409,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.382698133840426,"spectral_convergence":0.4834688177665305,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.948255643836438,"spectral_convergence":0.4833958509381798,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.00039399860715330165,"delta_db":-7.0017770356044915,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.007309863242506487,"delta_db":1.6378739896226264,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0021046838981186255,"delta_db":1.4219937827596647,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.002435856384528621,"delta_db":1.1701629131296845,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.00032789423764174994,"delta_db":1.111273446885384,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":7.367907874690312e-05,"delta_db":0.6412705687938491,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.1640586227992835,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7322980505934034,"pitch":null,"output_rms_dbfs":-18.980476656752863,"output_peak":0.9567216349585681,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.778873252696172},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55920,"output_samples":55920,"intelligibility":{"estoi":{"value":0.9059353080695073,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10336208580657574,"rms_dbfs":-19.712774707346266,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.2877957854351,"mean":-4.563447572983727e-06,"zero_energy":false},"output":{"rms":0.1124543260559401,"rms_dbfs":-18.980476656752863,"peak":0.9567216349585681,"peak_dbfs":-0.38428809894588883,"crest_factor_db":18.596188557806975,"mean":-5.580638264642222e-08,"zero_energy":false},"rms_difference_db":0.7322980505934034},"full_scale":{"reference":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":349,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.2266614618871747e-05,"output_mean_power":8.49387407571364e-06,"output_rms_dbfs":-50.70894181846409,"output_zero_energy":false,"output_peak":0.02281039112644032,"output_energy_fraction":0.00011146391612691829,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":342,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":193,"both_estimated_voiced_frames":175,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9067357512953368,"absolute_difference_cents_median":9.237340484375137,"absolute_difference_cents_p95":22.81295407081221,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.0006292775937517623,"log_rms_mae_db":1.8122634033410128,"rms_envelope_nrmse":0.3082825075321792,"envelope_correlation":0.9072659979201773,"rms_gain_error_db":0.7322984474901933,"reference_mean_power":0.010699017072596153,"output_mean_power":0.012664082331169507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":4.048774453403744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014613887448305465,"output_power":0.0014427408522059188,"reference_active":true,"delta_db":-0.0557742890225911},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0008261048603288557,"output_power":0.00034182544323372286,"reference_active":true,"delta_db":-3.8323079170289676},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.0004730285880207083,"output_power":0.00010681791212512896,"reference_active":true,"delta_db":-6.4624330366442475}],"active_band_count":3,"active_band_mae_db":3.450171747565269}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00037792225909995443,"log_rms_mae_db":1.6231904502476528,"rms_envelope_nrmse":0.29616389508502994,"envelope_correlation":0.9067048972155094,"rms_gain_error_db":0.7323112191722996,"reference_mean_power":0.01079168973829664,"output_mean_power":0.012773813549392214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":3.4667222612961444e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016650159461876711,"output_power":0.0013144225099781812,"reference_active":true,"delta_db":-1.0268340924008894},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005334276556675961,"output_power":0.0001811723359176355,"reference_active":true,"delta_db":-4.689836441520125},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9068036163681986e-05,"output_power":8.213632114523221e-06,"reference_active":true,"delta_db":-6.772863339879453}],"active_band_count":3,"active_band_mae_db":4.163177957933489}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00018013905998130842,"log_rms_mae_db":2.175434233802694,"rms_envelope_nrmse":0.28431120054609504,"envelope_correlation":0.9608831906659929,"rms_gain_error_db":1.4400587197195438,"reference_mean_power":0.002043548171012202,"output_mean_power":0.0028470215300988103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.1637162379955263e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006991910792272858,"output_power":0.000929996066192841,"reference_active":true,"delta_db":1.2388523290122717},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001302332633123258,"output_power":0.00024762234332833177,"reference_active":true,"delta_db":2.7906790614304926},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.350140195021887e-05,"output_power":6.621481042575057e-05,"reference_active":true,"delta_db":2.958921584359979}],"active_band_count":3,"active_band_mae_db":2.329484324934248}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00014142899773968785,"log_rms_mae_db":1.9555984745245854,"rms_envelope_nrmse":0.267994655868645,"envelope_correlation":0.9649760637976112,"rms_gain_error_db":1.4400423942389078,"reference_mean_power":0.002061256155357224,"output_mean_power":0.0028716810676122527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.0272195933213497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006464643086637218,"output_power":0.0008790001988276511,"reference_active":true,"delta_db":1.334444208582498},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.880640785962183e-05,"output_power":0.00014414416382657486,"reference_active":true,"delta_db":3.21168177718927},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.140839173745677e-06,"output_power":4.695542426829336e-06,"reference_active":true,"delta_db":3.4110172597537747}],"active_band_count":3,"active_band_mae_db":2.6523810818418476}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00022296985761207554,"log_rms_mae_db":1.308752954685586,"rms_envelope_nrmse":0.1830656309087782,"envelope_correlation":0.9940736428675884,"rms_gain_error_db":1.2273223955797556,"reference_mean_power":0.001891605546476507,"output_mean_power":0.002509359118285306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.5501990909593355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010291296692043158,"output_power":0.001333539517880124,"reference_active":true,"delta_db":1.1253579118811867},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.517530440244301e-05,"output_power":8.685200931713227e-05,"reference_active":true,"delta_db":1.9703513243643238},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2165950116970767e-05,"output_power":1.4036825194445247e-05,"reference_active":true,"delta_db":0.6212286009338612}],"active_band_count":3,"active_band_mae_db":1.2389792790597907}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00021618630891631995,"log_rms_mae_db":1.039926184333026,"rms_envelope_nrmse":0.1662866277597306,"envelope_correlation":0.9973241773457759,"rms_gain_error_db":1.227335730946356,"reference_mean_power":0.0019079864948334635,"output_mean_power":0.002531097468348391},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.466045308126094e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009690551708671024,"output_power":0.0012627269779987393,"reference_active":true,"delta_db":1.1496095592720272},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1519400367413303e-05,"output_power":3.0071967306468405e-05,"reference_active":true,"delta_db":1.4533167489302092},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.641667021707966e-07,"output_power":7.727616636984226e-07,"reference_active":true,"delta_db":0.048574590197009604}],"active_band_count":3,"active_band_mae_db":0.8838336327997487}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":1.3791169088652437,"rms_envelope_nrmse":0.2386046572268264,"envelope_correlation":0.9630393616875162,"rms_gain_error_db":0.7392850792086626,"reference_mean_power":6.674409144257211e-05,"output_mean_power":7.913003056894855e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":5.3173969334771903e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5059703710589825e-05,"output_power":4.015253743348222e-05,"reference_active":true,"delta_db":0.5890475410308769},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3101289173881222e-06,"output_power":3.3430383780876947e-06,"reference_active":true,"delta_db":1.6050514582973672},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.893247149634713e-07,"output_power":1.3949060556063223e-06,"reference_active":true,"delta_db":1.492061010430423}],"active_band_count":3,"active_band_mae_db":1.2287200032528889}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":0.9562223537922583,"rms_envelope_nrmse":0.15600136210067078,"envelope_correlation":0.9862985805026546,"rms_gain_error_db":0.7395731154999409,"reference_mean_power":6.730512481108481e-05,"output_mean_power":7.980046931649572e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":4.920971050071521e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3235368479868e-05,"output_power":3.843179034495307e-05,"reference_active":true,"delta_db":0.6309011839292817},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2494288246616962e-06,"output_power":1.6255519460572098e-06,"reference_active":true,"delta_db":1.142893313297693},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.10110394125166e-08,"output_power":9.228521856414175e-08,"reference_active":true,"delta_db":0.5658794069499584}],"active_band_count":3,"active_band_mae_db":0.7798913013923111}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"facodec","status":"measured","mos":{"status":"measured","value":3.8002212047576904,"predicted_mos":3.8002212047576904,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":56000,"input_sample_rate":16000,"input_seconds":3.5,"input_float32_sha256":"6c64385bf351e9c0e1d445001c4bd9686276696d3721a62d920830298e44a6bc","internal_samples":56000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.444297767000535},"started_at":"2026-10-13T04:51:00.575053+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.79930000033346e-05,"native_input_samples":55920,"native_input_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","output_sha256":"6cbdf62c6c087e26dceb43471873b02bf2ab6fb9ec000b6ed77fa9481e54ee6c","output_samples":56000,"length_delta_samples":80,"analysis_s":0.8649693679999473,"synthesis_s":6.698585732000538,"total_s":7.563555100000485,"rtf":2.1641073247497813,"output_peak":0.577387809753418,"output_rms":0.08858741090952431,"finished_at":"2026-10-13T04:51:08.148400+00:00","evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"6cbdf62c6c087e26dceb43471873b02bf2ab6fb9ec000b6ed77fa9481e54ee6c","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":55920,"output_samples":56000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.05255981981869,"output_peak":0.577387809753418},"common_prefix":{"samples":55920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9985714285714286,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.008103887721189,"spectral_convergence":0.39533155871397513,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.630367462620952,"spectral_convergence":0.36864994333664475,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.972267791164235,"spectral_convergence":0.3502495528005086,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.0006049108702783137,"delta_db":-5.139810012189856,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.004065797397646618,"delta_db":-0.9097611946354811,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0013697082311141383,"delta_db":-0.44359434612893256,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.001541898867658255,"delta_db":-0.8157949836661165,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.00021688097388029655,"delta_db":-0.6838498510093387,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":5.975995736154866e-05,"delta_db":-0.2680685116223611,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3768131498753478,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.3335765550876122,"pitch":null,"output_rms_dbfs":-21.04635126243388,"output_peak":0.577387809753418,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.536913047168792},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55920,"output_samples":55920,"intelligibility":{"estoi":{"value":0.9332128466245921,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10336208580657574,"rms_dbfs":-19.712774707346266,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.2877957854351,"mean":-4.563447572983727e-06,"zero_energy":false},"output":{"rms":0.08865075463829493,"rms_dbfs":-21.04635126243388,"peak":0.577387809753418,"peak_dbfs":-4.770647790218206,"crest_factor_db":16.275703472215675,"mean":2.43682883504078e-05,"zero_energy":false},"rms_difference_db":-1.333576555087614},"full_scale":{"reference":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":349,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.2266614618871747e-05,"output_mean_power":1.2330859063622908e-05,"output_rms_dbfs":-49.09006666016217,"output_zero_energy":false,"output_peak":0.027462216094136238,"output_energy_fraction":0.000260380989880691,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":342,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":204,"both_estimated_voiced_frames":193,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":3.101050705508139,"absolute_difference_cents_p95":10.713839526098837,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.0006292775937517623,"log_rms_mae_db":1.2730720650116387,"rms_envelope_nrmse":0.2772869687439026,"envelope_correlation":0.9249118200951347,"rms_gain_error_db":-1.3335769350097062,"reference_mean_power":0.010699017072596153,"output_mean_power":0.007870207576768285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":4.048774453403744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014613887448305465,"output_power":0.0009016560116135175,"reference_active":true,"delta_db":-2.0972487561201496},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0008261048603288557,"output_power":0.00019940108298622695,"reference_active":true,"delta_db":-6.173076645902337},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.0004730285880207083,"output_power":7.958950291558987e-05,"reference_active":true,"delta_db":-7.7403159633223515}],"active_band_count":3,"active_band_mae_db":5.3368804551149465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00037792225909995443,"log_rms_mae_db":1.1129027667942735,"rms_envelope_nrmse":0.26572084603300544,"envelope_correlation":0.9260640102406733,"rms_gain_error_db":-1.3335725977640935,"reference_mean_power":0.01079168973829664,"output_mean_power":0.007938385609594098},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":3.4667222612961444e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016650159461876711,"output_power":0.0008108890207207953,"reference_active":true,"delta_db":-3.1245697699096757},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005334276556675961,"output_power":9.74893742522005e-05,"reference_active":true,"delta_db":-7.381182451995894},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9068036163681986e-05,"output_power":4.286203123308729e-06,"reference_active":true,"delta_db":-9.597488326320967}],"active_band_count":3,"active_band_mae_db":6.701080182742179}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00018013905998130842,"log_rms_mae_db":1.7472045000140977,"rms_envelope_nrmse":0.153886099633008,"envelope_correlation":0.9789832935267031,"rms_gain_error_db":-0.46173969023981587,"reference_mean_power":0.002043548171012202,"output_mean_power":0.0018374304550249046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.1637162379955263e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006991910792272858,"output_power":0.0006321207319133222,"reference_active":true,"delta_db":-0.43795844337510775},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001302332633123258,"output_power":0.00012796941333725642,"reference_active":true,"delta_db":-0.076157440288179},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.350140195021887e-05,"output_power":4.062718950843002e-05,"reference_active":true,"delta_db":0.8375379835509911}],"active_band_count":3,"active_band_mae_db":0.450551289071426}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00014142899773968785,"log_rms_mae_db":1.5215474441608021,"rms_envelope_nrmse":0.14045875727418888,"envelope_correlation":0.9815716214493979,"rms_gain_error_db":-0.46173091499736296,"reference_mean_power":0.002061256155357224,"output_mean_power":0.0018533561097047917},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.0272195933213497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006464643086637218,"output_power":0.0005876023786962865,"reference_active":true,"delta_db":-0.4146100724552041},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.880640785962183e-05,"output_power":6.866868718224297e-05,"reference_active":true,"delta_db":-0.008701408821695223},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.140839173745677e-06,"output_power":2.853902251847281e-06,"reference_active":true,"delta_db":1.2485505112603241}],"active_band_count":3,"active_band_mae_db":0.5572873308457411}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00022296985761207554,"log_rms_mae_db":1.5400225024420244,"rms_envelope_nrmse":0.14221535705506977,"envelope_correlation":0.9905571822209506,"rms_gain_error_db":-0.7653913249581281,"reference_mean_power":0.001891605546476507,"output_mean_power":0.0015859571392724836},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.5501990909593355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010291296692043158,"output_power":0.0009698355637122714,"reference_active":true,"delta_db":-0.2577199325678859},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.517530440244301e-05,"output_power":5.615724370799792e-05,"reference_active":true,"delta_db":0.07661045895767554},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2165950116970767e-05,"output_power":1.663154738895885e-05,"reference_active":true,"delta_db":1.3578662616215262}],"active_band_count":3,"active_band_mae_db":0.5640655510490292}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00021618630891631995,"log_rms_mae_db":1.2413992534710916,"rms_envelope_nrmse":0.128873732213316,"envelope_correlation":0.9929319401263944,"rms_gain_error_db":-0.7653835901393369,"reference_mean_power":0.0019079864948334635,"output_mean_power":0.001599694078958795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.466045308126094e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009690551708671024,"output_power":0.0009201885194803518,"reference_active":true,"delta_db":-0.22471692697370432},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1519400367413303e-05,"output_power":2.1980990581450308e-05,"reference_active":true,"delta_db":0.09217094511241034},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.641667021707966e-07,"output_power":1.1394203934273305e-06,"reference_active":true,"delta_db":1.7349587842352743}],"active_band_count":3,"active_band_mae_db":0.6839488854404631}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":1.7688916316011676,"rms_envelope_nrmse":0.17235854930270367,"envelope_correlation":0.9768333553025994,"rms_gain_error_db":-0.5097054231954111,"reference_mean_power":6.674409144257211e-05,"output_mean_power":5.935294642469345e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":5.3173969334771903e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5059703710589825e-05,"output_power":3.217738779164364e-05,"reference_active":true,"delta_db":-0.3725745703249673},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3101289173881222e-06,"output_power":3.216328011334591e-06,"reference_active":true,"delta_db":1.4372411657108528},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.893247149634713e-07,"output_power":1.1078701667035623e-06,"reference_active":true,"delta_db":0.49150009017856905}],"active_band_count":3,"active_band_mae_db":0.7671052754047963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":1.3600789683317784,"rms_envelope_nrmse":0.12351329971755155,"envelope_correlation":0.9883097374010289,"rms_gain_error_db":-0.5094599217671935,"reference_mean_power":6.730512481108481e-05,"output_mean_power":5.985523521070391e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":4.920971050071521e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3235368479868e-05,"output_power":3.0987758717668756e-05,"reference_active":true,"delta_db":-0.3041033242326359},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2494288246616962e-06,"output_power":1.675260784767168e-06,"reference_active":true,"delta_db":1.273709014421843},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.10110394125166e-08,"output_power":7.730547380056885e-08,"reference_active":true,"delta_db":-0.20333958149474943}],"active_band_count":3,"active_band_mae_db":0.5937173067164094}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.028127670288086,"predicted_mos":3.028127670288086,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":154624,"input_sample_rate":44100,"input_seconds":3.5062131519274375,"input_float32_sha256":"246f76b33ef9b52646ae63c2606f38a2429f76f7247769f2a6443ca7b8fe683a","internal_samples":56100,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":9.519888301001629},"started_at":"2026-10-13T07:12:06.722020+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005112161999932141,"native_input_samples":154130,"native_input_sha256":"ca6bd9d1ccf5ccf5e557f9decb3a07c532062b9eb739b878618e497992dc8c5b","output_sha256":"6d95cb0d90c2727ee38123b867e59208eb4a022af12e2fc7fc86b4012b59b4a9","output_samples":154624,"length_delta_samples":494,"analysis_s":0.08850204100053816,"synthesis_s":19.755260731999442,"total_s":19.84376277299998,"rtf":5.677739170111589,"output_peak":0.7148619294166565,"output_rms":0.09518516867859872,"finished_at":"2026-10-13T07:12:26.580636+00:00","evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"6d95cb0d90c2727ee38123b867e59208eb4a022af12e2fc7fc86b4012b59b4a9","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":55920,"output_samples":56100,"length_delta_samples":180,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.425266589077037,"output_peak":0.7148670323188798},"common_prefix":{"samples":55920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9967914438502674,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.204306901692075,"spectral_convergence":0.43400567922849037,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.356790477773144,"spectral_convergence":0.37319757291526606,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.459151855866715,"spectral_convergence":0.3934428772208278,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.0012485317310740318,"delta_db":-1.9927280579312932,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.0044910716219003985,"delta_db":-0.47771871937807364,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0015438018760620307,"delta_db":0.07604064130932305,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.0016379749064337876,"delta_db":-0.5532814372792868,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.0001424552309365374,"delta_db":-2.5092803872753078,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":3.255224590142778e-05,"delta_db":-2.9063617256095093,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.419235161463799,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6985357117629012,"pitch":null,"output_rms_dbfs":-20.41131041910917,"output_peak":0.7148670323188798,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.340083078443977},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55920,"output_samples":55920,"intelligibility":{"estoi":{"value":0.8321912174177692,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10336208580657574,"rms_dbfs":-19.712774707346266,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.2877957854351,"mean":-4.563447572983727e-06,"zero_energy":false},"output":{"rms":0.09537498420607059,"rms_dbfs":-20.41131041910917,"peak":0.7148670323188798,"peak_dbfs":-2.915494618535859,"crest_factor_db":17.49581580057331,"mean":-0.0004994005254829275,"zero_energy":false},"rms_difference_db":-0.6985357117629043},"full_scale":{"reference":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":349,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.2266614618871747e-05,"output_mean_power":7.553464786022373e-06,"output_rms_dbfs":-51.21853791114519,"output_zero_energy":false,"output_peak":0.017322869033260697,"output_energy_fraction":0.00013780280705502374,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":342,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":197,"both_estimated_voiced_frames":184,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9533678756476683,"absolute_difference_cents_median":11.894083567761111,"absolute_difference_cents_p95":41.6546384711943,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.0006292775937517623,"log_rms_mae_db":1.4938615812025182,"rms_envelope_nrmse":0.13615569480393747,"envelope_correlation":0.9846135045932262,"rms_gain_error_db":-0.6985353796999144,"reference_mean_power":0.010699017072596153,"output_mean_power":0.009109411953718031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":4.048774453403744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014613887448305465,"output_power":0.001384375044478086,"reference_active":true,"delta_db":-0.23511996362262438},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0008261048603288557,"output_power":0.000553047644152351,"reference_active":true,"delta_db":-1.742726306329251},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.0004730285880207083,"output_power":0.00024877924624006336,"reference_active":true,"delta_db":-2.7907324096679127}],"active_band_count":3,"active_band_mae_db":1.5895262265399293}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00037792225909995443,"log_rms_mae_db":1.224884132850098,"rms_envelope_nrmse":0.11188228021973362,"envelope_correlation":0.9912141364734609,"rms_gain_error_db":-0.6985251524565992,"reference_mean_power":0.01079168973829664,"output_mean_power":0.009188337427422484},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":3.4667222612961444e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016650159461876711,"output_power":0.0014194842102039078,"reference_active":true,"delta_db":-0.6928783122251504},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005334276556675961,"output_power":0.00032938322009885406,"reference_active":true,"delta_db":-2.09374057065427},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9068036163681986e-05,"output_power":2.0018269943175815e-05,"reference_active":true,"delta_db":-2.9039503971038756}],"active_band_count":3,"active_band_mae_db":1.8968564266610988}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00018013905998130842,"log_rms_mae_db":2.4557363579965257,"rms_envelope_nrmse":0.20047316494583156,"envelope_correlation":0.963234655954105,"rms_gain_error_db":0.1695245926217392,"reference_mean_power":0.002043548171012202,"output_mean_power":0.00212489433520101},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.1637162379955263e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006991910792272858,"output_power":0.0007622846759884449,"reference_active":true,"delta_db":0.3752131072169443},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001302332633123258,"output_power":0.00012196933539337393,"reference_active":true,"delta_db":-0.2847126563488482},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.350140195021887e-05,"output_power":2.6494577987656126e-05,"reference_active":true,"delta_db":-1.0190597520634355}],"active_band_count":3,"active_band_mae_db":0.5596618385430759}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00014142899773968785,"log_rms_mae_db":1.8565445642574443,"rms_envelope_nrmse":0.1682572903428077,"envelope_correlation":0.9721441598142646,"rms_gain_error_db":0.16952409447102196,"reference_mean_power":0.002061256155357224,"output_mean_power":0.00214330696366792},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.0272195933213497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006464643086637218,"output_power":0.0006944737147535781,"reference_active":true,"delta_db":0.31111260251711353},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.880640785962183e-05,"output_power":5.675522307666763e-05,"reference_active":true,"delta_db":-0.8362305044142191},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.140839173745677e-06,"output_power":1.9836231948777265e-06,"reference_active":true,"delta_db":-0.3312486508135875}],"active_band_count":3,"active_band_mae_db":0.4928639192483067}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00022296985761207554,"log_rms_mae_db":3.8218364811383414,"rms_envelope_nrmse":0.1645502882460228,"envelope_correlation":0.9873354404809919,"rms_gain_error_db":-0.8363147663758244,"reference_mean_power":0.001891605546476507,"output_mean_power":0.0015602676429416304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.5501990909593355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010291296692043158,"output_power":0.0009168939705926041,"reference_active":true,"delta_db":-0.5015098197642232},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.517530440244301e-05,"output_power":4.873486653552437e-05,"reference_active":true,"delta_db":-0.5390495679441081},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2165950116970767e-05,"output_power":8.969623020366656e-06,"reference_active":true,"delta_db":-1.323718407390217}],"active_band_count":3,"active_band_mae_db":0.7880925983661827}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00021618630891631995,"log_rms_mae_db":3.328010595385155,"rms_envelope_nrmse":0.13482494522112046,"envelope_correlation":0.9934783626737731,"rms_gain_error_db":-0.8363009177962488,"reference_mean_power":0.0019079864948334635,"output_mean_power":0.0015737842857567208},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.466045308126094e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009690551708671024,"output_power":0.0008710596427375345,"reference_active":true,"delta_db":-0.46300610476072857},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1519400367413303e-05,"output_power":1.987235281414995e-05,"reference_active":true,"delta_db":-0.3458087662043843},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.641667021707966e-07,"output_power":4.867990063346273e-07,"reference_active":true,"delta_db":-1.9583842673324827}],"active_band_count":3,"active_band_mae_db":0.9223997127658651}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":3.8146858627188824,"rms_envelope_nrmse":0.36215775782413856,"envelope_correlation":0.9121729046037393,"rms_gain_error_db":-2.1161273515507837,"reference_mean_power":6.674409144257211e-05,"output_mean_power":4.10015325876544e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":5.3173969334771903e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5059703710589825e-05,"output_power":1.606235755886041e-05,"reference_active":true,"delta_db":-3.3899895218757443},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3101289173881222e-06,"output_power":2.1185794407037517e-06,"reference_active":true,"delta_db":-0.37591463042762646},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.893247149634713e-07,"output_power":8.258481594728158e-07,"reference_active":true,"delta_db":-0.7843865337342304}],"active_band_count":3,"active_band_mae_db":1.5167635620125337}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":3.2427222742217467,"rms_envelope_nrmse":0.3303785409613991,"envelope_correlation":0.9272264440329256,"rms_gain_error_db":-2.1158804253513797,"reference_mean_power":6.730512481108481e-05,"output_mean_power":4.1348531622972536e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":4.920971050071521e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3235368479868e-05,"output_power":1.5266412850161683e-05,"reference_active":true,"delta_db":-3.378634953051554},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2494288246616962e-06,"output_power":1.0737004850524438e-06,"reference_active":true,"delta_db":-0.6582837171143013},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.10110394125166e-08,"output_power":6.05551499372269e-08,"reference_active":true,"delta_db":-1.26393120625416}],"active_band_count":3,"active_band_mae_db":1.7669499588066717}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"golf","status":"measured","mos":{"status":"measured","value":3.1235008239746094,"predicted_mos":3.1235008239746094,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83760,"input_sample_rate":24000,"input_seconds":3.49,"input_float32_sha256":"4640275ab9158817ee9c97607c16914140015b38e7533a7e20e6c5a4a73ec5ea","internal_samples":55840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6181552500002},"started_at":"2026-10-13T05:02:24.100046+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002038757000264013,"native_input_samples":83880,"native_input_sha256":"b65e71db39cec6b80ae62b10ce5e456b296fb4eef0a83df9f9526c3193128012","output_sha256":"3e7766b65f315f8c6670e94d18cb9b351f629f428e90249797e8069840b64dc7","output_samples":83760,"length_delta_samples":-120,"analysis_s":0.4386333780003042,"synthesis_s":0.07561353499932011,"total_s":0.5142469129996243,"rtf":0.14713788640904846,"output_peak":0.757988691329956,"output_rms":0.09345488311125862,"finished_at":"2026-10-13T05:02:24.622056+00:00","analysis_stages":{"external_f0_s":0.12423572900024737},"evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"3e7766b65f315f8c6670e94d18cb9b351f629f428e90249797e8069840b64dc7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":55920,"output_samples":55840,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.584662881982755,"output_peak":0.7534493984625054},"common_prefix":{"samples":55840,"sample_rate":16000,"reference_coverage":0.9985693848354793,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55840,"output_samples":55840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.800806440335958,"spectral_convergence":0.4862320753727499,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.333752181656087,"spectral_convergence":0.3613997778403715,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.605499843212341,"spectral_convergence":0.3699308833141327,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975277228223782,"output_power":0.0006572085567282598,"delta_db":-4.779248478152985,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0050246164587129235,"output_power":0.0047797919872804425,"delta_db":-0.21693919720358995,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015180528190541867,"output_power":0.0014650405518339198,"delta_db":-0.15437236659991266,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018631892240544088,"output_power":0.0015605388781965938,"delta_db":-0.7698237089594974,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002542307200651109,"output_power":0.00023982575453920167,"delta_db":-0.2533220800625962,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.365593856305903e-05,"output_power":3.804264211272309e-05,"delta_db":-2.235682542484957,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.40156472891059,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8781038451887507,"pitch":null,"output_rms_dbfs":-20.584662881982755,"output_peak":0.7534493984625054,"reference_rms_dbfs":-19.706559036794008,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.580019488401462},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55840,"output_samples":55840,"intelligibility":{"estoi":{"value":0.9276553039640346,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10343607875723765,"rms_dbfs":-19.706559036794005,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.28158011488284,"mean":-3.4086342185821095e-06,"zero_energy":false},"output":{"rms":0.09349036512224743,"rms_dbfs":-20.584662881982755,"peak":0.7534493984625054,"peak_dbfs":-2.458918190316634,"crest_factor_db":18.12574469166612,"mean":0.0011928688086766152,"zero_energy":false},"rms_difference_db":-0.8781038451887504},"full_scale":{"reference":{"sample_count":55840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":349,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.2266614618871747e-05,"output_mean_power":8.67663601820327e-06,"output_rms_dbfs":-50.61648620665214,"output_zero_energy":false,"output_peak":0.015445734493804725,"output_energy_fraction":0.00016497579662616218,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":342,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":204,"both_estimated_voiced_frames":189,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9792746113989638,"absolute_difference_cents_median":4.426949322992127,"absolute_difference_cents_p95":15.575519112128712,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":697,"rms_floor":0.0006292775937517623,"log_rms_mae_db":1.3173146546602694,"rms_envelope_nrmse":0.2542883753791843,"envelope_correlation":0.9287812372785264,"rms_gain_error_db":-0.8781003885384986,"reference_mean_power":0.010714356235657396,"output_mean_power":0.00875298215460856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":697,"frequency_resolution_hz":0.28694404591104733,"power_floor":4.038426135015158e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014600159049399475,"output_power":0.0010885018861074016,"reference_active":true,"delta_db":-1.2752840096967288},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0008211653567022091,"output_power":0.00021202964164349813,"reference_active":true,"delta_db":-5.880340397473538},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00047025903407089096,"output_power":4.7837386672343615e-05,"reference_active":true,"delta_db":-9.92569701082113}],"active_band_count":3,"active_band_mae_db":5.693773805997132}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":691,"rms_floor":0.00037792225909995443,"log_rms_mae_db":1.0200579258297864,"rms_envelope_nrmse":0.22165345137542264,"envelope_correlation":0.9435842903773956,"rms_gain_error_db":-0.878105856881467,"reference_mean_power":0.010807295924500906,"output_mean_power":0.008828897152309131},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":691,"frequency_resolution_hz":0.2894356005788712,"power_floor":3.4570823735369687e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016604496815630235,"output_power":0.0009684118476350204,"reference_active":true,"delta_db":-2.3416562523064415},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005301441023959951,"output_power":0.00010769666619849231,"reference_active":true,"delta_db":-6.92191674734233},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.88210630340621e-05,"output_power":2.8585687537122722e-06,"reference_active":true,"delta_db":-11.329187812280532}],"active_band_count":3,"active_band_mae_db":6.864253603976434}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":697,"rms_floor":0.00018013905998130842,"log_rms_mae_db":1.4738210053933445,"rms_envelope_nrmse":0.13749806802525308,"envelope_correlation":0.9824783097327714,"rms_gain_error_db":0.15112869446878652,"reference_mean_power":0.0020464774674373285,"output_mean_power":0.0021189457441493594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":697,"frequency_resolution_hz":0.28694404591104733,"power_floor":1.1642485793993803e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007003678032154894,"output_power":0.0007338162299529457,"reference_active":true,"delta_db":0.20261140090305124},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001304073271942749,"output_power":8.387982589926109e-05,"reference_active":true,"delta_db":-1.9164447340432933},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3241611755679726e-05,"output_power":3.195699032796509e-05,"reference_active":true,"delta_db":-0.1711620164196992}],"active_band_count":3,"active_band_mae_db":0.7634060504553478}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":691,"rms_floor":0.00014142899773968785,"log_rms_mae_db":1.0120935286137882,"rms_envelope_nrmse":0.10242961728073016,"envelope_correlation":0.9895057674740515,"rms_gain_error_db":0.15112671063032918,"reference_mean_power":0.002064237303033032,"output_mean_power":0.002137333500977402},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":691,"frequency_resolution_hz":0.2894356005788712,"power_floor":1.0275127449152043e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006476188296208531,"output_power":0.0006699904075192849,"reference_active":true,"delta_db":0.1474911735711833},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.862322703180955e-05,"output_power":4.308613551883087e-05,"reference_active":true,"delta_db":-2.021335939543147},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1210474608183942e-06,"output_power":1.8465220339489691e-06,"reference_active":true,"delta_db":-0.601958921387207}],"active_band_count":3,"active_band_mae_db":0.9235953448338456}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":697,"rms_floor":0.00022296985761207554,"log_rms_mae_db":1.4551790509994107,"rms_envelope_nrmse":0.14714497946950256,"envelope_correlation":0.9889629703209992,"rms_gain_error_db":-0.7150503233515446,"reference_mean_power":0.0018943187051213954,"output_mean_power":0.001606748915008861},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":697,"frequency_resolution_hz":0.28694404591104733,"power_floor":1.5532239110156713e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010314733784749532,"output_power":0.0008355668626722893,"reference_active":true,"delta_db":-0.9147681546881989},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.550482278463391e-05,"output_power":5.8900089039240325e-05,"reference_active":true,"delta_db":0.2578523093084783},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2019869930519208e-05,"output_power":1.4208512502180304e-05,"reference_active":true,"delta_db":0.7264884565130937}],"active_band_count":3,"active_band_mae_db":0.6330363068365904}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":691,"rms_floor":0.00021618630891631995,"log_rms_mae_db":0.9935603728199549,"rms_envelope_nrmse":0.11661287689046454,"envelope_correlation":0.9945541210659985,"rms_gain_error_db":-0.715056583071324,"reference_mean_power":0.0019107462646677856,"output_mean_power":0.0016206803291496106},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":691,"frequency_resolution_hz":0.2894356005788712,"power_floor":1.4688426857591842e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009713616559634793,"output_power":0.0007858024207176438,"reference_active":true,"delta_db":-0.9206759355708561},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1552680659984433e-05,"output_power":2.3230089814966272e-05,"reference_active":true,"delta_db":0.32549594779250746},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.617385851592998e-07,"output_power":9.061100763099914e-07,"reference_active":true,"delta_db":0.7537500502825801}],"active_band_count":3,"active_band_mae_db":0.6666406445486479}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":697,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":2.228003350500896,"rms_envelope_nrmse":0.22936287774281652,"envelope_correlation":0.9588518624234694,"rms_gain_error_db":-0.6366679610380449,"reference_mean_power":6.683974057715238e-05,"output_mean_power":5.772553405109944e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":697,"frequency_resolution_hz":0.28694404591104733,"power_floor":5.320548787034083e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.51131555365398e-05,"output_power":3.327051440358893e-05,"reference_active":true,"delta_db":-0.2341034458649443},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.29926959566329e-06,"output_power":2.1473681445285744e-06,"reference_active":true,"delta_db":-0.2968339026202653},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.79313814776281e-07,"output_power":1.2836371675133159e-06,"reference_active":true,"delta_db":1.1752040257176308}],"active_band_count":3,"active_band_mae_db":0.5687137914009468}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55840,"uncovered_tail_samples":0,"metrics":{"frame_count":691,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":1.8771284434417177,"rms_envelope_nrmse":0.17599945673748146,"envelope_correlation":0.975685590605982,"rms_gain_error_db":-0.6361984004668308,"reference_mean_power":6.740228235035123e-05,"output_mean_power":5.821766230912438e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":691,"frequency_resolution_hz":0.2894356005788712,"power_floor":4.9240214230086055e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3283858201314734e-05,"output_power":3.161666705318547e-05,"reference_active":true,"delta_db":-0.22317577309798386},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2318744238315927e-06,"output_power":9.622674150000025e-07,"reference_active":true,"delta_db":-1.0727065884964586},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.012363938741427e-08,"output_power":6.772683037757241e-08,"reference_active":true,"delta_db":-0.7299991659885654}],"active_band_count":3,"active_band_mae_db":0.6752938425276693}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.1218833923339844,"predicted_mos":3.1218833923339844,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84000,"input_sample_rate":24000,"input_seconds":3.5,"input_float32_sha256":"c7c90c5ed45e7df127fc116122f828c6d0c8d32bd6d7f5b671e2fadae1854a00","internal_samples":56000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1996278500009794},"started_at":"2026-10-13T05:09:37.417256+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0020388729990372667,"native_input_samples":83880,"native_input_sha256":"b65e71db39cec6b80ae62b10ce5e456b296fb4eef0a83df9f9526c3193128012","output_sha256":"2df8d3c9630c1c71479b34edfac36ab5671327c92413aa42e91ca29c625c8dca","output_samples":84000,"length_delta_samples":120,"analysis_s":1.1491342640001676,"synthesis_s":6.729277155000091,"total_s":7.878411419000258,"rtf":2.2541949696710324,"output_peak":0.7404671907424927,"output_rms":0.09302454239167487,"finished_at":"2026-10-13T05:09:45.302518+00:00","analysis_stages":{"external_f0_s":0.7082913989997905},"evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"2df8d3c9630c1c71479b34edfac36ab5671327c92413aa42e91ca29c625c8dca","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":55920,"output_samples":56000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.62555951460719,"output_peak":0.7333331602681362},"common_prefix":{"samples":55920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9985714285714286,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.696217431297964,"spectral_convergence":0.5474069343693414,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.221724120779947,"spectral_convergence":0.48250270584967375,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.407011706156311,"spectral_convergence":0.45034720179739546,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.0002837345757795081,"delta_db":-8.427601281209508,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.004966578400277833,"delta_db":-0.04064557570400883,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0014438465996227977,"delta_db":-0.2146644582795042,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.0016842182824960297,"delta_db":-0.43237010478012716,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.0002500190373631157,"delta_db":-0.06633361129695128,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":4.251559117590746e-05,"delta_db":-1.7466890781788016,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.8213840182414833,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.90657726242135,"pitch":null,"output_rms_dbfs":-20.619351969767617,"output_peak":0.7333331602681362,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.441651086078075},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55920,"output_samples":55920,"intelligibility":{"estoi":{"value":0.9047393988144494,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10336208580657574,"rms_dbfs":-19.712774707346266,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.2877957854351,"mean":-4.563447572983727e-06,"zero_energy":false},"output":{"rms":0.09311773454458175,"rms_dbfs":-20.619351969767617,"peak":0.7333331602681362,"peak_dbfs":-2.6939735278019143,"crest_factor_db":17.9253784419657,"mean":-5.578079210719598e-05,"zero_energy":false},"rms_difference_db":-0.9065772624213508},"full_scale":{"reference":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":349,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.2266614618871747e-05,"output_mean_power":7.961883142568252e-06,"output_rms_dbfs":-50.98984200892866,"output_zero_energy":false,"output_peak":0.03301555355024408,"output_energy_fraction":0.00015238134745389044,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":342,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":202,"both_estimated_voiced_frames":190,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9844559585492227,"absolute_difference_cents_median":6.382805093863663,"absolute_difference_cents_p95":23.53609888791815,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.0006292775937517623,"log_rms_mae_db":1.7659213493881643,"rms_envelope_nrmse":0.3304354648956019,"envelope_correlation":0.8706161680809423,"rms_gain_error_db":-0.9065766229801817,"reference_mean_power":0.010699017072596153,"output_mean_power":0.008683328241164983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":4.048774453403744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014613887448305465,"output_power":0.0008878439110763159,"reference_active":true,"delta_db":-2.164291376166192},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0008261048603288557,"output_power":0.00016546823900728466,"reference_active":true,"delta_db":-6.983205323564439},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.0004730285880207083,"output_power":6.560399823756021e-05,"reference_active":true,"delta_db":-8.579570803399934}],"active_band_count":3,"active_band_mae_db":5.909022501043521}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00037792225909995443,"log_rms_mae_db":1.4422143585014362,"rms_envelope_nrmse":0.2999671269497872,"envelope_correlation":0.8853848395818823,"rms_gain_error_db":-0.9065731367048007,"reference_mean_power":0.01079168973829664,"output_mean_power":0.008758548459363137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":3.4667222612961444e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016650159461876711,"output_power":0.0007837727002500611,"reference_active":true,"delta_db":-3.2722826478247735},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005334276556675961,"output_power":9.13999269602339e-05,"reference_active":true,"delta_db":-7.661296793458213},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9068036163681986e-05,"output_power":5.1448537995750645e-06,"reference_active":true,"delta_db":-8.804485430532994}],"active_band_count":3,"active_band_mae_db":6.579354957271993}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00018013905998130842,"log_rms_mae_db":1.8164280029451083,"rms_envelope_nrmse":0.18039542802050054,"envelope_correlation":0.9685224944750401,"rms_gain_error_db":-0.06331618537678686,"reference_mean_power":0.002043548171012202,"output_mean_power":0.002013971223743303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.1637162379955263e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006991910792272858,"output_power":0.000654112206578909,"reference_active":true,"delta_db":-0.2894362492277809},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001302332633123258,"output_power":0.00011694282840771686,"reference_active":true,"delta_db":-0.4674832942275175},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.350140195021887e-05,"output_power":3.716556681063907e-05,"reference_active":true,"delta_db":0.4507777903866403}],"active_band_count":3,"active_band_mae_db":0.4025657779473129}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00014142899773968785,"log_rms_mae_db":1.3140660610316002,"rms_envelope_nrmse":0.11581675834382847,"envelope_correlation":0.9858545534316545,"rms_gain_error_db":-0.06331953277801951,"reference_mean_power":0.002061256155357224,"output_mean_power":0.0020314213488309705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.0272195933213497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006464643086637218,"output_power":0.0005995878732323929,"reference_active":true,"delta_db":-0.32691711867202855},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.880640785962183e-05,"output_power":6.41089751840031e-05,"reference_active":true,"delta_db":-0.30710050940478567},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.140839173745677e-06,"output_power":2.4709355102551308e-06,"reference_active":true,"delta_db":0.6227736774099718}],"active_band_count":3,"active_band_mae_db":0.41893043516226197}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00022296985761207554,"log_rms_mae_db":1.792737653731809,"rms_envelope_nrmse":0.14279440498103627,"envelope_correlation":0.9873198980167659,"rms_gain_error_db":-0.40097213355284983,"reference_mean_power":0.001891605546476507,"output_mean_power":0.0017247786410191273},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.5501990909593355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010291296692043158,"output_power":0.0009297857693857861,"reference_active":true,"delta_db":-0.44087203928880225},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.517530440244301e-05,"output_power":5.112804164548243e-05,"reference_active":true,"delta_db":-0.3308557955735497},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2165950116970767e-05,"output_power":1.1302993534116316e-05,"reference_active":true,"delta_db":-0.31952552266617945}],"active_band_count":3,"active_band_mae_db":0.36375111917617714}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00021618630891631995,"log_rms_mae_db":1.2945218046213567,"rms_envelope_nrmse":0.09906621381507894,"envelope_correlation":0.9944597286121063,"rms_gain_error_db":-0.40096888307000783,"reference_mean_power":0.0019079864948334635,"output_mean_power":0.0017397162018354402},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.466045308126094e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009690551708671024,"output_power":0.0008754425062227921,"reference_active":true,"delta_db":-0.4412087376806603},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1519400367413303e-05,"output_power":2.0101866996588437e-05,"reference_active":true,"delta_db":-0.2959377049346293},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.641667021707966e-07,"output_power":7.467538388845534e-07,"reference_active":true,"delta_db":-0.10010645913319055}],"active_band_count":3,"active_band_mae_db":0.27908430058282674}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":1.9811835684524655,"rms_envelope_nrmse":0.24394175999935225,"envelope_correlation":0.9505626820515778,"rms_gain_error_db":-0.536083183474325,"reference_mean_power":6.674409144257211e-05,"output_mean_power":5.899354675919827e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":5.3173969334771903e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5059703710589825e-05,"output_power":3.2321893402324455e-05,"reference_active":true,"delta_db":-0.3531144792462133},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3101289173881222e-06,"output_power":2.2084032718279163e-06,"reference_active":true,"delta_db":-0.1955783460174284},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.893247149634713e-07,"output_power":6.753769029649582e-07,"reference_active":true,"delta_db":-1.6579265437364834}],"active_band_count":3,"active_band_mae_db":0.7355397896667083}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":1.4293591598363926,"rms_envelope_nrmse":0.15153209890834513,"envelope_correlation":0.9811000291139855,"rms_gain_error_db":-0.5362606409139945,"reference_mean_power":6.730512481108481e-05,"output_mean_power":5.948700032402884e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":4.920971050071521e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3235368479868e-05,"output_power":3.055161945432658e-05,"reference_active":true,"delta_db":-0.3656626224891447},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2494288246616962e-06,"output_power":1.1042442373164894e-06,"reference_active":true,"delta_db":-0.5364637951916467},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.10110394125166e-08,"output_power":5.225765028842556e-08,"reference_active":true,"delta_db":-1.903943262000968}],"active_band_count":3,"active_band_mae_db":0.9353565598939197}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.3758034706115723,"predicted_mos":3.3758034706115723,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84120,"input_sample_rate":24000,"input_seconds":3.505,"input_float32_sha256":"4fba74af7f26caac58a5e6d0ed57abee48a7a6fb46299a6703e0224861763836","internal_samples":56080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.296193136000511},"started_at":"2026-10-13T05:10:32.308604+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0040936400000646245,"native_input_samples":83880,"native_input_sha256":"b65e71db39cec6b80ae62b10ce5e456b296fb4eef0a83df9f9526c3193128012","output_sha256":"eea76ddc48c2451104e639f0b8653b5d6611aa6a74c8b9f08cc54dafdc172770","output_samples":84120,"length_delta_samples":240,"analysis_s":0.42911046399967745,"synthesis_s":0.09667281400106731,"total_s":0.5257832780007448,"rtf":0.1504387061518583,"output_peak":0.6280577182769775,"output_rms":0.09901055463524706,"finished_at":"2026-10-13T05:10:32.846364+00:00","analysis_stages":{"external_f0_s":0.07496048999928462},"evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"eea76ddc48c2451104e639f0b8653b5d6611aa6a74c8b9f08cc54dafdc172770","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":55920,"output_samples":56080,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.083982064253302,"output_peak":0.6285189308656324},"common_prefix":{"samples":55920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9971469329529244,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55920,"output_samples":55920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.364648313357284,"spectral_convergence":0.2845552387973517,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.3376619269683,"spectral_convergence":0.20546598711543243,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.932781697926856,"spectral_convergence":0.226601213077297,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001975478876015244,"output_power":0.0013985035290914319,"delta_db":-1.5000882344877202,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005013278749394044,"output_power":0.004965282183735997,"delta_db":-0.041779179402743026,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015170067046631896,"output_power":0.0014203275721851732,"delta_db":-0.28598982333236883,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018605244766058998,"output_power":0.001764044431880535,"delta_db":-0.23125868138702602,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002538671104412511,"output_power":0.0002384366880780202,"delta_db":-0.2723335904690721,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.356486516629775e-05,"output_power":4.995113278687464e-05,"delta_db":-1.0467178854138912,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5630278990821369,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3587992242887642,"pitch":null,"output_rms_dbfs":-20.071573931635033,"output_peak":0.6285189308656324,"reference_rms_dbfs":-19.712774707346266,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.545030646084147},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55920,"output_samples":55920,"intelligibility":{"estoi":{"value":0.9648549183252018,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10336208580657574,"rms_dbfs":-19.712774707346266,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.2877957854351,"mean":-4.563447572983727e-06,"zero_energy":false},"output":{"rms":0.09917936044237244,"rms_dbfs":-20.071573931635033,"peak":0.6285189308656324,"peak_dbfs":-4.0336327383460056,"crest_factor_db":16.037941193289026,"mean":0.00011964442900755082,"zero_energy":false},"rms_difference_db":-0.35879922428876654},"full_scale":{"reference":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":349,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.2266614618871747e-05,"output_mean_power":1.1303795245876539e-05,"output_rms_dbfs":-49.46775717790012,"output_zero_energy":false,"output_peak":0.01949498246766494,"output_energy_fraction":0.00019070517964507055,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":342,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":192,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9481865284974094,"absolute_difference_cents_median":0.9029867112072006,"absolute_difference_cents_p95":5.6573319018895445,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.0006292775937517623,"log_rms_mae_db":0.6091687357236502,"rms_envelope_nrmse":0.11180030028695735,"envelope_correlation":0.9873358952515109,"rms_gain_error_db":-0.3587996642316069,"reference_mean_power":0.010699017072596153,"output_mean_power":0.009850627897101723},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":4.048774453403744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014613887448305465,"output_power":0.0013368914999120306,"reference_active":true,"delta_db":-0.3866959620847027},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0008261048603288557,"output_power":0.000543636291469422,"reference_active":true,"delta_db":-1.8172673613221269},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.0004730285880207083,"output_power":0.00016909899011154168,"reference_active":true,"delta_db":-4.467463746883139}],"active_band_count":3,"active_band_mae_db":2.2238090234299897}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00037792225909995443,"log_rms_mae_db":0.40667627270713447,"rms_envelope_nrmse":0.0850087977090648,"envelope_correlation":0.9928667173980714,"rms_gain_error_db":-0.35880092020230986,"reference_mean_power":0.01079168973829664,"output_mean_power":0.009935949118359188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":3.4667222612961444e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016650159461876711,"output_power":0.0013545814249008347,"reference_active":true,"delta_db":-0.8961328126716744},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005334276556675961,"output_power":0.0003056253392589637,"reference_active":true,"delta_db":-2.4188616946924486},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9068036163681986e-05,"output_power":1.1100616183132615e-05,"reference_active":true,"delta_db":-5.464744943853842}],"active_band_count":3,"active_band_mae_db":2.926579817072655}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00018013905998130842,"log_rms_mae_db":0.843133389146303,"rms_envelope_nrmse":0.07344669153989118,"envelope_correlation":0.9951181418031482,"rms_gain_error_db":-0.17041003906140276,"reference_mean_power":0.002043548171012202,"output_mean_power":0.0019649154864934057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.1637162379955263e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006991910792272858,"output_power":0.0006744711731419276,"reference_active":true,"delta_db":-0.15632485942429902},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001302332633123258,"output_power":0.00010393177743999809,"reference_active":true,"delta_db":-0.9797356835437309},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.350140195021887e-05,"output_power":1.9428737549405763e-05,"reference_active":true,"delta_db":-2.366183998550677}],"active_band_count":3,"active_band_mae_db":1.1674148471729024}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00014142899773968785,"log_rms_mae_db":0.526714789646546,"rms_envelope_nrmse":0.04592920155482626,"envelope_correlation":0.9981635309267559,"rms_gain_error_db":-0.17041260881172166,"reference_mean_power":0.002061256155357224,"output_mean_power":0.001981940921293831},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.0272195933213497e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006464643086637218,"output_power":0.0006179103139619742,"reference_active":true,"delta_db":-0.19619108071741037},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.880640785962183e-05,"output_power":5.319359939514988e-05,"reference_active":true,"delta_db":-1.1176950719065895},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.140839173745677e-06,"output_power":9.974561120780014e-07,"reference_active":true,"delta_db":-3.3169024713094193}],"active_band_count":3,"active_band_mae_db":1.5435962079778065}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":0.00022296985761207554,"log_rms_mae_db":0.838255586831952,"rms_envelope_nrmse":0.07383975954346882,"envelope_correlation":0.9967834967596559,"rms_gain_error_db":-0.23171163126099834,"reference_mean_power":0.001891605546476507,"output_mean_power":0.0017933267045357024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":1.5501990909593355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010291296692043158,"output_power":0.0009672859898691571,"reference_active":true,"delta_db":-0.2691520132093164},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.517530440244301e-05,"output_power":5.468250054989138e-05,"reference_active":true,"delta_db":-0.03896371850887779},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2165950116970767e-05,"output_power":1.2769024001358009e-05,"reference_active":true,"delta_db":0.21011671862178993}],"active_band_count":3,"active_band_mae_db":0.17274415011332808}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":0.00021618630891631995,"log_rms_mae_db":0.6156344622455433,"rms_envelope_nrmse":0.04348465691812759,"envelope_correlation":0.9991648738227457,"rms_gain_error_db":-0.23170900037465167,"reference_mean_power":0.0019079864948334635,"output_mean_power":0.0018088576723773102},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.466045308126094e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009690551708671024,"output_power":0.0009132699934947822,"reference_active":true,"delta_db":-0.25749314632829706},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1519400367413303e-05,"output_power":2.1698918639029068e-05,"reference_active":true,"delta_db":0.036079257568459075},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.641667021707966e-07,"output_power":6.208263796327058e-07,"reference_active":true,"delta_db":-0.9021794748196179}],"active_band_count":3,"active_band_mae_db":0.3985839595721247}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":698,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":0.9614326737868214,"rms_envelope_nrmse":0.1387835459088268,"envelope_correlation":0.9858864345430154,"rms_gain_error_db":-0.501960050793831,"reference_mean_power":6.674409144257211e-05,"output_mean_power":5.94588931656464e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":698,"frequency_resolution_hz":0.28653295128939826,"power_floor":5.3173969334771903e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5059703710589825e-05,"output_power":3.088342100086629e-05,"reference_active":true,"delta_db":-0.5508283980103192},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3101289173881222e-06,"output_power":2.24398412350753e-06,"reference_active":true,"delta_db":-0.12616436590481744},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.893247149634713e-07,"output_power":6.150556805792423e-07,"reference_active":true,"delta_db":-2.064244246615074}],"active_band_count":3,"active_band_mae_db":0.9137456701767369}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55920,"uncovered_tail_samples":0,"metrics":{"frame_count":692,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":0.6511310052315634,"rms_envelope_nrmse":0.10064336569856135,"envelope_correlation":0.993248613281801,"rms_gain_error_db":-0.5019156970860337,"reference_mean_power":6.730512481108481e-05,"output_mean_power":5.995930141402091e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":4.920971050071521e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3235368479868e-05,"output_power":2.918096695615363e-05,"reference_active":true,"delta_db":-0.5650081934063665},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2494288246616962e-06,"output_power":1.0570341187124791e-06,"reference_active":true,"delta_db":-0.7262251542252849},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.10110394125166e-08,"output_power":3.936458115507233e-08,"reference_active":true,"delta_db":-3.1343856957954963}],"active_band_count":3,"active_band_mae_db":1.4752063478090491}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0034","method":"magphase","status":"measured","mos":{"status":"measured","value":2.941586494445801,"predicted_mos":2.941586494445801,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":166481,"input_sample_rate":48000,"input_seconds":3.468354166666667,"input_float32_sha256":"5d50018ffb9e5999852674ae4f27de0072597a27497c2d86fdff826b5dc6f35b","internal_samples":55494,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.773962944998857},"started_at":"2026-10-13T05:14:42.197370+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006828362000305788,"native_input_samples":167760,"native_input_sha256":"c628ab69005505687ed805b1b22c95c3d2125ad1b9e15de07b8e05c1c35fcdbf","output_sha256":"58c04da0b10ce7ec389add72c864e9bea0736b817dd4e53bc8ec2de37cb09ab9","output_samples":166481,"length_delta_samples":-1279,"analysis_s":2.0650011819998326,"synthesis_s":0.8740352080003504,"total_s":2.939036390000183,"rtf":0.840926005722513,"output_peak":0.9599363613951819,"output_rms":0.09582405348274114,"finished_at":"2026-10-13T05:14:45.151446+00:00","evaluation_input":{"reference":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"58c04da0b10ce7ec389add72c864e9bea0736b817dd4e53bc8ec2de37cb09ab9","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":55920,"output_samples":55494,"length_delta_samples":-426,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.36937969937361,"output_peak":0.9594725411686237},"common_prefix":{"samples":55494,"sample_rate":16000,"reference_coverage":0.992381974248927,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":55494,"output_samples":55494,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.123576308317371,"spectral_convergence":0.5644248410703865,"spectral_convergence_status":"ok","amplitude_floor":7.545858591823444e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.010098195157945,"spectral_convergence":0.3501229998120007,"spectral_convergence_status":"ok","amplitude_floor":3.9963548992246115e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.341518042577867,"spectral_convergence":0.27488427255339437,"spectral_convergence_status":"ok","amplitude_floor":1.7986227023732554e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0019898876245471524,"output_power":0.0010017882546090855,"delta_db":-2.9805261554406375,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005053248509717736,"output_power":0.004549042799205734,"delta_db":-0.4565063372420061,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015277476729861656,"output_power":0.0014415613360582347,"delta_db":-0.2521850525806769,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0018749237637889265,"output_power":0.0018428505456246067,"delta_db":-0.07493498035493856,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00025582223183289577,"output_power":0.0002652389853263655,"delta_db":0.15699074377818575,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.405323504988844e-05,"output_power":8.415578752077435e-05,"delta_db":1.1854291947004978,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8510954106828237,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6897955915402594,"pitch":null,"output_rms_dbfs":-20.36937969937361,"output_peak":0.9594725411686237,"reference_rms_dbfs":-19.679584107833353,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.491730848684394},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":55494,"output_samples":55494,"intelligibility":{"estoi":{"value":0.9688124200611113,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.10375780952739301,"rms_dbfs":-19.679584107833353,"peak":0.84869384765625,"peak_dbfs":-1.4249789219111684,"crest_factor_db":18.254605185922188,"mean":-5.973293213568134e-06,"zero_energy":false},"output":{"rms":0.09583651552693682,"rms_dbfs":-20.36937969937361,"peak":0.9594725411686237,"peak_dbfs":-0.3593489936527694,"crest_factor_db":20.01003070572084,"mean":-0.0006043733326266018,"zero_energy":false},"rms_difference_db":-0.6897955915402569},"full_scale":{"reference":{"sample_count":55494,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":55494,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":346,"uncovered_tail_samples":134,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9599028999800796e-05,"threshold_dbfs":-44.02315463195664,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.250903524289077e-05,"output_mean_power":8.655163025155933e-06,"output_rms_dbfs":-50.627247475160466,"output_zero_energy":false,"output_peak":0.017821922764914167,"output_energy_fraction":0.00014943415245795177,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":339,"reference_estimated_voiced_frames":193,"output_estimated_voiced_frames":199,"both_estimated_voiced_frames":191,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9896373056994818,"absolute_difference_cents_median":2.6898014650159774,"absolute_difference_cents_p95":11.929892766705523,"difference_over_600_cents_frames":0,"uncovered_tail_samples":134,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":55494,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":692,"rms_floor":0.0006292775937517623,"log_rms_mae_db":1.0771207829507539,"rms_envelope_nrmse":0.2824215261073934,"envelope_correlation":0.9061980013914932,"rms_gain_error_db":-0.689793708657195,"reference_mean_power":0.010791720391775688,"output_mean_power":0.009206855215009437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":3.985685304640933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014508689031471995,"output_power":0.0010907275027097658,"reference_active":true,"delta_db":-1.2391190844183817},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000800028175772533,"output_power":0.0003396698841789491,"reference_active":true,"delta_db":-3.7204823930370927},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.0004542495688395224,"output_power":0.0002494097801790842,"reference_active":true,"delta_db":-2.6038104416685774}],"active_band_count":3,"active_band_mae_db":2.521137306374684}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":686,"rms_floor":0.00037792225909995443,"log_rms_mae_db":0.8021157169874448,"rms_envelope_nrmse":0.179371529990291,"envelope_correlation":0.963411472703798,"rms_gain_error_db":-0.6897921644215194,"reference_mean_power":0.010885986808968982,"output_mean_power":0.009287281028580282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":686,"frequency_resolution_hz":0.29154518950437314,"power_floor":3.4082471746450887e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016361637910054351,"output_power":0.0011148292361316632,"reference_active":true,"delta_db":-1.6661842785147858},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0005147680006135021,"output_power":0.00025128780040414736,"reference_active":true,"delta_db":-3.114401372938956},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.76947750811316e-05,"output_power":2.867289418716968e-05,"reference_active":true,"delta_db":-1.1880962427379396}],"active_band_count":3,"active_band_mae_db":1.9895606313972272}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":692,"rms_floor":0.00018013905998130842,"log_rms_mae_db":1.0461868129086394,"rms_envelope_nrmse":0.06658601974634533,"envelope_correlation":0.9959737522597975,"rms_gain_error_db":-0.1535183553572484,"reference_mean_power":0.0020612561167491043,"output_mean_power":0.001989665739898507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.1667335793500796e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007065863895584184,"output_power":0.0006663657991878944,"reference_active":true,"delta_db":-0.2545256861650078},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013101708443451665,"output_power":0.00011760902169874076,"reference_active":true,"delta_db":-0.4688729329027568},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.193650080010762e-05,"output_power":2.9542512643785494e-05,"reference_active":true,"delta_db":-0.3383989981175013}],"active_band_count":3,"active_band_mae_db":0.3539325390617553}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":686,"rms_floor":0.00014142899773968785,"log_rms_mae_db":0.6151477586386399,"rms_envelope_nrmse":0.042159762591873685,"envelope_correlation":0.9984586036112805,"rms_gain_error_db":-0.15351380301504483,"reference_mean_power":0.0020792749482629108,"output_mean_power":0.0020070608554204035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":686,"frequency_resolution_hz":0.29154518950437314,"power_floor":1.0288430198430323e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006535596410565631,"output_power":0.0006103579320253258,"reference_active":true,"delta_db":-0.29700632677137706},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.741978168092372e-05,"output_power":6.055151897502502e-05,"reference_active":true,"delta_db":-0.4666229967173019},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2576035777428068e-06,"output_power":2.1156864774671534e-06,"reference_active":true,"delta_db":-0.28196374413310615}],"active_band_count":3,"active_band_mae_db":0.34853102254059504}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":692,"rms_floor":0.00022296985761207554,"log_rms_mae_db":0.9236049355703483,"rms_envelope_nrmse":0.07681767566363976,"envelope_correlation":0.9961199869542144,"rms_gain_error_db":-0.07814086481631961,"reference_mean_power":0.0019079982110547736,"output_mean_power":0.0018739753618577796},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":1.5683546234435866e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010433128530985625,"output_power":0.0010187858301281836,"reference_active":true,"delta_db":-0.10331661797536767},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7158374820521426e-05,"output_power":5.161333970112873e-05,"reference_active":true,"delta_db":-0.44317910526032817},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1205904211979195e-05,"output_power":1.2207130518325931e-05,"reference_active":true,"delta_db":0.37166682263155865}],"active_band_count":3,"active_band_mae_db":0.30605418195575146}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":686,"rms_floor":0.00021618630891631995,"log_rms_mae_db":0.5465034174885119,"rms_envelope_nrmse":0.03956952367123117,"envelope_correlation":0.9989786689302594,"rms_gain_error_db":-0.07815379991798929,"reference_mean_power":0.0019246652969419718,"output_mean_power":0.0018903396150629671},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":686,"frequency_resolution_hz":0.29154518950437314,"power_floor":1.4828015424640235e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000983071481213826,"output_power":0.000957883079955502,"reference_active":true,"delta_db":-0.11272595503625399},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1584824787359018e-05,"output_power":2.018788978552198e-05,"reference_active":true,"delta_db":-0.29057602599386456},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.998939993785077e-07,"output_power":8.155989511039346e-07,"reference_active":true,"delta_db":0.08444219761656947}],"active_band_count":3,"active_band_mae_db":0.16258139288222934}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":692,"rms_floor":3.608035305023674e-05,"log_rms_mae_db":1.4594497838364766,"rms_envelope_nrmse":0.17975363179167902,"envelope_correlation":0.9798135976299979,"rms_gain_error_db":0.6579301743254596,"reference_mean_power":6.73214703498144e-05,"output_mean_power":7.833333380135468e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":692,"frequency_resolution_hz":0.28901734104046245,"power_floor":5.336112643344646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.538614884786389e-05,"output_power":4.19835235963504e-05,"reference_active":true,"delta_db":0.742455852872347},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.223921405190483e-06,"output_power":2.673704187515702e-06,"reference_active":true,"delta_db":0.7999392130524462},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.380335543329365e-07,"output_power":9.920044336622073e-07,"reference_active":true,"delta_db":0.24295239418632528}],"active_band_count":3,"active_band_mae_db":0.5951158200370394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":55440,"uncovered_tail_samples":54,"metrics":{"frame_count":686,"rms_floor":2.6238288952821805e-05,"log_rms_mae_db":1.177852411772565,"rms_envelope_nrmse":0.13382796162279653,"envelope_correlation":0.9900892177679682,"rms_gain_error_db":0.6577426655936724,"reference_mean_power":6.789154044505658e-05,"output_mean_power":7.899324038402999e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":686,"frequency_resolution_hz":0.29154518950437314,"power_floor":4.938952261802857e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3524778599870035e-05,"output_power":3.9578241814103104e-05,"reference_active":true,"delta_db":0.7209057919682611},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1431530233482289e-06,"output_power":1.2842315759617957e-06,"reference_active":true,"delta_db":0.5053897461755658},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.5384486426126e-08,"output_power":6.692747177578952e-08,"reference_active":true,"delta_db":-0.5167756086234727}],"active_band_count":3,"active_band_mae_db":0.5810237155890998}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2282419204711914,"predicted_mos":2.2282419204711914,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b4d67f57c4aa254608e758b2ca498ea13c0e4df09f600514fa6f12707ec2c4ec","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5300162530002126},"started_at":"2026-10-13T04:38:59.447555+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3037999451626092e-05,"native_input_samples":96000,"native_input_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","output_sha256":"55d1c448fdce9d8414ccd227e4e10315a949b04ced7c566da3cbba7c215bb8e5","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0995311350006887,"synthesis_s":0.5002459809993525,"total_s":1.5997771160000411,"rtf":0.26662951933334017,"output_peak":0.6557666248446634,"output_rms":0.09355582275360819,"finished_at":"2026-10-13T04:39:01.052357+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"55d1c448fdce9d8414ccd227e4e10315a949b04ced7c566da3cbba7c215bb8e5","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.69883796767837,"spectral_convergence":0.23456845365969137,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.536794704636039,"spectral_convergence":0.23201783449866747,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.77159492589292,"spectral_convergence":0.23836608716943133,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0006459646016386351,"delta_db":0.255712598852455,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.004939059183095345,"delta_db":-0.10005194221275195,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0021548346337730098,"delta_db":-0.14408528683714095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0005620146509231323,"delta_db":-0.3389230967814391,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.0001333261251605981,"delta_db":0.11107866850090248,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.0003174927765138304,"delta_db":-0.007836048493236785,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15961460694632104,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09505419606591264,"pitch":null,"output_rms_dbfs":-20.578583551613768,"output_peak":0.6557666248446634,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.66907586606911},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.69883796767837,"spectral_convergence":0.23456845365969137,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.536794704636039,"spectral_convergence":0.23201783449866747,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.77159492589292,"spectral_convergence":0.23836608716943133,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0006459646016386351,"delta_db":0.255712598852455,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.004939059183095345,"delta_db":-0.10005194221275195,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0021548346337730098,"delta_db":-0.14408528683714095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0005620146509231323,"delta_db":-0.3389230967814391,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.0001333261251605981,"delta_db":0.11107866850090248,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.0003174927765138304,"delta_db":-0.007836048493236785,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15961460694632104,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09505419606591264,"pitch":null,"output_rms_dbfs":-20.578583551613768,"output_peak":0.6557666248446634,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.66907586606911},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8830212611533288,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09458527529619017,"rms_dbfs":-20.483529355547855,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.58908518923696,"mean":-7.225354512532552e-06,"zero_energy":false},"output":{"rms":0.09355582275360819,"rms_dbfs":-20.578583551613768,"peak":0.6557666248446634,"peak_dbfs":-3.6650138092842948,"crest_factor_db":16.91356974232947,"mean":-0.00013801345247179165,"zero_energy":false},"rms_difference_db":-0.09505419606591303},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":4.4089105859863335e-06,"output_rms_dbfs":-53.55668708705203,"output_zero_energy":false,"output_peak":0.008704123558365966,"output_energy_fraction":1.2593013099688543e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":332,"output_estimated_voiced_frames":310,"both_estimated_voiced_frames":306,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9216867469879518,"absolute_difference_cents_median":1.772356212403369,"absolute_difference_cents_p95":12.786406183524313,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002515751859318019,"log_rms_mae_db":0.8740676111321568,"rms_envelope_nrmse":0.07376187201923594,"envelope_correlation":0.9928706067032245,"rms_gain_error_db":-0.09505316656949168,"reference_mean_power":0.008953808913712473,"output_mean_power":0.00875996770473321},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.40642035802602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010131590544257714,"output_power":0.0009703041944138528,"reference_active":true,"delta_db":-0.1876972125174194},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021577628078402173,"output_power":0.00019910274738026233,"reference_active":true,"delta_db":-0.34926450318719054},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.554717092269403e-05,"output_power":4.655314643171184e-05,"reference_active":true,"delta_db":0.09487632625350892}],"active_band_count":3,"active_band_mae_db":0.2106126806527063}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002371340103862925,"log_rms_mae_db":0.4853076951025203,"rms_envelope_nrmse":0.041518299438663496,"envelope_correlation":0.9976276893352104,"rms_gain_error_db":-0.09500778495450744,"reference_mean_power":0.008997361160684636,"output_mean_power":0.008802669071175212},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.083478820328128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000878683378715502,"output_power":0.0008367479293645388,"reference_active":true,"delta_db":-0.21237764994177272},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.731034174913729e-05,"output_power":9.24639821927033e-05,"reference_active":true,"delta_db":-0.22186404297159193},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4907380289326388e-06,"output_power":2.183737248747655e-06,"reference_active":true,"delta_db":-0.5712766960335893}],"active_band_count":3,"active_band_mae_db":0.3351727963156514}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019737503490519328,"log_rms_mae_db":1.6512966946431287,"rms_envelope_nrmse":0.12403163732352881,"envelope_correlation":0.9871742571509821,"rms_gain_error_db":-0.08353111542012698,"reference_mean_power":0.0030829546180243575,"output_mean_power":0.0030242244504913424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6914511369831456e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008052884179451132,"output_power":0.0007800977588148521,"reference_active":true,"delta_db":-0.1380242270508637},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019776686451848029,"output_power":0.0001699476299271759,"reference_active":true,"delta_db":-0.6583841576558196},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.986877726391445e-05,"output_power":3.0106186215482446e-05,"reference_active":true,"delta_db":1.80484602343144}],"active_band_count":3,"active_band_mae_db":0.8670848027127077}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017438355068165795,"log_rms_mae_db":1.03501835448002,"rms_envelope_nrmse":0.0736746198896781,"envelope_correlation":0.9952398776578475,"rms_gain_error_db":-0.08365463226004573,"reference_mean_power":0.0030981947388459344,"output_mean_power":0.003039087812266065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5303812378070619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007441051144552584,"output_power":0.0007176657025676457,"reference_active":true,"delta_db":-0.15712097901732092},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.887152997212749e-05,"output_power":8.859922242667285e-05,"reference_active":true,"delta_db":-0.47641344225361953},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5498600424852703e-06,"output_power":1.8641774181746492e-06,"reference_active":true,"delta_db":0.8019476111281894}],"active_band_count":3,"active_band_mae_db":0.47849401079970993}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011259893106980884,"log_rms_mae_db":1.5225101740464222,"rms_envelope_nrmse":0.13712859311422185,"envelope_correlation":0.986477475091154,"rms_gain_error_db":-0.31797693405063343,"reference_mean_power":0.0006356182176644625,"output_mean_power":0.0005907430797065842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.9581273226653026e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020276258597112104,"output_power":0.00018642618056951494,"reference_active":true,"delta_db":-0.364809194426456},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6931765404372444e-05,"output_power":2.221623016241838e-05,"reference_active":true,"delta_db":-0.8359445668493225},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.128022173574558e-06,"output_power":5.8267296805388186e-06,"reference_active":true,"delta_db":-0.8754417168306062}],"active_band_count":3,"active_band_mae_db":0.692065159368795}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010212318931682781,"log_rms_mae_db":0.9156634040724977,"rms_envelope_nrmse":0.07875355135042351,"envelope_correlation":0.9960828449525717,"rms_gain_error_db":-0.318038209633583,"reference_mean_power":0.0006388062286465316,"output_mean_power":0.0005936976379744042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.690696537781622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018976599367877945,"output_power":0.000174329067035801,"reference_active":true,"delta_db":-0.36848582976008015},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3827689740852884e-05,"output_power":1.1231603488835742e-05,"reference_active":true,"delta_db":-0.9030786339989129},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.220792176423388e-07,"output_power":3.6138870344913566e-07,"reference_active":true,"delta_db":-0.6741939592299812}],"active_band_count":3,"active_band_mae_db":0.6485861409963247}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":1.7164352961337455,"rms_envelope_nrmse":0.16139648370546444,"envelope_correlation":0.9830998513500663,"rms_gain_error_db":0.20742472649267785,"reference_mean_power":0.0001805297653048896,"output_mean_power":0.0001893613278011766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6198199988921145e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.95993126484054e-05,"output_power":0.00010558826269886416,"reference_active":true,"delta_db":0.2535930295954927},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361167220740983e-06,"output_power":9.323807130775642e-06,"reference_active":true,"delta_db":-0.017367219491405333},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4041286689847184e-06,"output_power":2.319185224301142e-06,"reference_active":true,"delta_db":-0.15622271962890807}],"active_band_count":3,"active_band_mae_db":0.14239432290526868}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":1.088554727131284,"rms_envelope_nrmse":0.08977153256211638,"envelope_correlation":0.9947322294163736,"rms_gain_error_db":0.20740764766884795,"reference_mean_power":0.00018143628465930523,"output_mean_power":0.00019031144590558798},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5580116647442146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.473019444861232e-05,"output_power":0.0001007043408926163,"reference_active":true,"delta_db":0.265597625949733},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1793170529251645e-06,"output_power":4.392056923140421e-06,"reference_active":true,"delta_db":0.2156264113590215},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8116784694632847e-07,"output_power":1.6015769546083823e-07,"reference_active":true,"delta_db":-0.5353331203590027}],"active_band_count":3,"active_band_mae_db":0.3388523858892524}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9930434226989746,"predicted_mos":2.9930434226989746,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b7b848f97ed712ab7ab929c03de1a1eebd8d15ce93016a921e0f24224f2ddd1d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7188860189999104},"started_at":"2026-10-13T04:41:57.425728+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.061200004594866e-05,"native_input_samples":96000,"native_input_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","output_sha256":"3351bf3d4c36dabafc2a1b6c8dff07973332214cd755940f25a940498f829ce3","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2879404919995068,"synthesis_s":0.09990929200012033,"total_s":1.387849783999627,"rtf":0.23130829733327118,"output_peak":0.7841227173929007,"output_rms":0.10819979065022092,"finished_at":"2026-10-13T04:41:58.820574+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"3351bf3d4c36dabafc2a1b6c8dff07973332214cd755940f25a940498f829ce3","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.046725255087793,"spectral_convergence":0.49124969320176576,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5472385741333845,"spectral_convergence":0.3276779565486089,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.0108616756192985,"spectral_convergence":0.3711568210106987,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0008591993520404135,"delta_db":1.4945648107560727,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.006338442908481485,"delta_db":0.9833315849223085,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0032770185603551044,"delta_db":1.6765642529267277,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0007274666401615201,"delta_db":0.7817313559862825,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.00015804015874451832,"delta_db":0.8496006663846953,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.000347027076968591,"delta_db":0.3784590849485937,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0273752926541133,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1680577651285873,"pitch":null,"output_rms_dbfs":-19.315471590419268,"output_peak":0.7841227173929007,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.868275168280159},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.046725255087793,"spectral_convergence":0.49124969320176576,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5472385741333845,"spectral_convergence":0.3276779565486089,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.0108616756192985,"spectral_convergence":0.3711568210106987,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0008591993520404135,"delta_db":1.4945648107560727,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.006338442908481485,"delta_db":0.9833315849223085,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0032770185603551044,"delta_db":1.6765642529267277,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0007274666401615201,"delta_db":0.7817313559862825,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.00015804015874451832,"delta_db":0.8496006663846953,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.000347027076968591,"delta_db":0.3784590849485937,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0273752926541133,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1680577651285873,"pitch":null,"output_rms_dbfs":-19.315471590419268,"output_peak":0.7841227173929007,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.868275168280159},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8813609392285491,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09458527529619017,"rms_dbfs":-20.483529355547855,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.58908518923696,"mean":-7.225354512532552e-06,"zero_energy":false},"output":{"rms":0.10819979065022091,"rms_dbfs":-19.315471590419264,"peak":0.7841227173929007,"peak_dbfs":-2.1123192739666754,"crest_factor_db":17.20315231645259,"mean":2.3390301404921457e-06,"zero_energy":false},"rms_difference_db":1.1680577651285908},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":4.59968551947715e-06,"output_rms_dbfs":-53.3727186001941,"output_zero_energy":false,"output_peak":0.00844705372480615,"output_energy_fraction":9.822347792578808e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":332,"output_estimated_voiced_frames":342,"both_estimated_voiced_frames":313,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.9427710843373494,"absolute_difference_cents_median":7.001254729409081,"absolute_difference_cents_p95":1178.8877188399626,"difference_over_600_cents_frames":17,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002515751859318019,"log_rms_mae_db":1.97908934807855,"rms_envelope_nrmse":0.18485132654877712,"envelope_correlation":0.984706279157714,"rms_gain_error_db":1.1680680958831955,"reference_mean_power":0.008953808913712473,"output_mean_power":0.011716951474589435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.40642035802602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010131590544257714,"output_power":0.0013160184526190344,"reference_active":true,"delta_db":1.135843488055418},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021577628078402173,"output_power":0.00028788640017815125,"reference_active":true,"delta_db":1.2521744605534846},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.554717092269403e-05,"output_power":5.9415312505717294e-05,"reference_active":true,"delta_db":1.1543697895980498}],"active_band_count":3,"active_band_mae_db":1.1807959127356507}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002371340103862925,"log_rms_mae_db":1.7286106273027515,"rms_envelope_nrmse":0.16577391957846294,"envelope_correlation":0.9916391313749364,"rms_gain_error_db":1.1677383314308198,"reference_mean_power":0.008997361160684636,"output_mean_power":0.011773049956783174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.083478820328128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000878683378715502,"output_power":0.0011552088906888521,"reference_active":true,"delta_db":1.1882811133409257},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.731034174913729e-05,"output_power":0.00013880143085609744,"reference_active":true,"delta_db":1.5423494535369309},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4907380289326388e-06,"output_power":2.749863993488209e-06,"reference_active":true,"delta_db":0.4298316275991205}],"active_band_count":3,"active_band_mae_db":1.0534873981589923}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019737503490519328,"log_rms_mae_db":2.133210668904947,"rms_envelope_nrmse":0.23123801127329321,"envelope_correlation":0.9864291529695397,"rms_gain_error_db":1.4708054348022426,"reference_mean_power":0.0030829546180243575,"output_mean_power":0.004325613134821545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6914511369831456e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008052884179451132,"output_power":0.0011661279298275041,"reference_active":true,"delta_db":1.6079474432436465},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019776686451848029,"output_power":0.0003326599139545893,"reference_active":true,"delta_db":2.2584694275116095},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.986877726391445e-05,"output_power":2.8075053762688533e-05,"reference_active":true,"delta_db":1.5014945539447515}],"active_band_count":3,"active_band_mae_db":1.789303808233336}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017438355068165795,"log_rms_mae_db":1.7814051019107109,"rms_envelope_nrmse":0.21397138714214797,"envelope_correlation":0.9911659047795773,"rms_gain_error_db":1.4704231946279904,"reference_mean_power":0.0030981947388459344,"output_mean_power":0.004346613570545578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5303812378070619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007441051144552584,"output_power":0.0010940406519148918,"reference_active":true,"delta_db":1.6739917005122917},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.887152997212749e-05,"output_power":0.00017366050253505146,"reference_active":true,"delta_db":2.446297988099536},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5498600424852703e-06,"output_power":2.0583358782260195e-06,"reference_active":true,"delta_db":1.2322376246162197}],"active_band_count":3,"active_band_mae_db":1.7841757710760158}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011259893106980884,"log_rms_mae_db":1.3242010439158136,"rms_envelope_nrmse":0.16972928856055802,"envelope_correlation":0.9893824816993072,"rms_gain_error_db":0.9332782302047657,"reference_mean_power":0.0006356182176644625,"output_mean_power":0.0007879962653580967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.9581273226653026e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020276258597112104,"output_power":0.00023281695345950143,"reference_active":true,"delta_db":0.6002678050412564},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6931765404372444e-05,"output_power":3.630016914561547e-05,"reference_active":true,"delta_db":1.2964382590979027},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.128022173574558e-06,"output_power":6.5232872032604916e-06,"reference_active":true,"delta_db":-0.3850254222465931}],"active_band_count":3,"active_band_mae_db":0.7605771621285841}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010212318931682781,"log_rms_mae_db":0.9139962663822303,"rms_envelope_nrmse":0.13883453532075607,"envelope_correlation":0.9955451059448289,"rms_gain_error_db":0.9332590652010203,"reference_mean_power":0.0006388062286465316,"output_mean_power":0.0007919450499087971},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.690696537781622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018976599367877945,"output_power":0.0002187066252307421,"reference_active":true,"delta_db":0.6164355031361344},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3827689740852884e-05,"output_power":1.819953426503952e-05,"reference_active":true,"delta_db":1.1931064783362422},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.220792176423388e-07,"output_power":3.974087984446043e-07,"reference_active":true,"delta_db":-0.2615649091079569}],"active_band_count":3,"active_band_mae_db":0.6903689635267778}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":1.5438490045816273,"rms_envelope_nrmse":0.24031566644658625,"envelope_correlation":0.9680664773585456,"rms_gain_error_db":0.6780891778407284,"reference_mean_power":0.0001805297653048896,"output_mean_power":0.00021103657763565714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6198199988921145e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.95993126484054e-05,"output_power":0.00010785936549015633,"reference_active":true,"delta_db":0.3460151981830908},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361167220740983e-06,"output_power":1.1883631915934098e-05,"reference_active":true,"delta_db":1.036191882691827},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4041286689847184e-06,"output_power":5.721660509728863e-06,"reference_active":true,"delta_db":3.7656437831206797}],"active_band_count":3,"active_band_mae_db":1.7159502879985327}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":1.0202642765042087,"rms_envelope_nrmse":0.15630297500693544,"envelope_correlation":0.9881961018599977,"rms_gain_error_db":0.6780896478858085,"reference_mean_power":0.00018143628465930523,"output_mean_power":0.00021209630806824742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5580116647442146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.473019444861232e-05,"output_power":0.00010521235673094857,"reference_active":true,"delta_db":0.4557832003605883},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1793170529251645e-06,"output_power":5.154474733714763e-06,"reference_active":true,"delta_db":0.9107909612912696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8116784694632847e-07,"output_power":3.62140427161897e-07,"reference_active":true,"delta_db":3.0079588645525144}],"active_band_count":3,"active_band_mae_db":1.4581776754014575}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"facodec","status":"measured","mos":{"status":"measured","value":4.1155314445495605,"predicted_mos":4.1155314445495605,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1632a21dd17556936ccf321078d59d508d6a27f6837e19e68b95128209c55781","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4408148389993585},"started_at":"2026-10-13T04:51:08.149401+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.861400025722105e-05,"native_input_samples":96000,"native_input_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","output_sha256":"02a4229e48d66a9ad922dcfe35b72515bfdbb1df23596c24476b86e60aa8bcdc","output_samples":96000,"length_delta_samples":0,"analysis_s":2.2247034480005823,"synthesis_s":6.657277427999361,"total_s":8.881980875999943,"rtf":1.4803301459999905,"output_peak":0.552330732345581,"output_rms":0.08557267309473668,"finished_at":"2026-10-13T04:51:17.041009+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"02a4229e48d66a9ad922dcfe35b72515bfdbb1df23596c24476b86e60aa8bcdc","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.240926693466249,"spectral_convergence":0.23541905470508076,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.822607466347508,"spectral_convergence":0.26469422432365197,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.994388820832247,"spectral_convergence":0.2775299286547585,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0005683706860333146,"delta_db":-0.300057888544414,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.00407398292952961,"delta_db":-0.9363021949691391,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0018518617754930595,"delta_db":-0.80213908397349,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0004792420901661577,"delta_db":-1.0308499320428253,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.00012262699581364614,"delta_db":-0.2522130180302353,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.00022659790354288205,"delta_db":-1.4726156616686215,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7990296298714542,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8697686730871166,"pitch":null,"output_rms_dbfs":-21.35329802863497,"output_peak":0.552330732345581,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.685974326882001},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.240926693466249,"spectral_convergence":0.23541905470508076,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.822607466347508,"spectral_convergence":0.26469422432365197,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.994388820832247,"spectral_convergence":0.2775299286547585,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0005683706860333146,"delta_db":-0.300057888544414,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.00407398292952961,"delta_db":-0.9363021949691391,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0018518617754930595,"delta_db":-0.80213908397349,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0004792420901661577,"delta_db":-1.0308499320428253,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.00012262699581364614,"delta_db":-0.2522130180302353,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.00022659790354288205,"delta_db":-1.4726156616686215,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7990296298714542,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8697686730871166,"pitch":null,"output_rms_dbfs":-21.35329802863497,"output_peak":0.552330732345581,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.685974326882001},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8963843724258301,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09458527529619017,"rms_dbfs":-20.483529355547855,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.58908518923696,"mean":-7.225354512532552e-06,"zero_energy":false},"output":{"rms":0.08557267309473668,"rms_dbfs":-21.35329802863497,"peak":0.552330732345581,"peak_dbfs":-5.156015828744764,"crest_factor_db":16.197282199890207,"mean":0.00017109181075772995,"zero_energy":false},"rms_difference_db":-0.8697686730871155},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":3.7502013204030664e-06,"output_rms_dbfs":-54.25945417607399,"output_zero_energy":false,"output_peak":0.006362020969390869,"output_energy_fraction":1.280337288132764e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":332,"output_estimated_voiced_frames":342,"both_estimated_voiced_frames":325,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9789156626506024,"absolute_difference_cents_median":3.436917359470074,"absolute_difference_cents_p95":21.82379736475879,"difference_over_600_cents_frames":9,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002515751859318019,"log_rms_mae_db":1.4242976536696148,"rms_envelope_nrmse":0.14210880521483593,"envelope_correlation":0.9837518963497701,"rms_gain_error_db":-0.8697622609133725,"reference_mean_power":0.008953808913712473,"output_mean_power":0.00732877849252494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.40642035802602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010131590544257714,"output_power":0.000792018920506621,"reference_active":true,"delta_db":-1.0694207344104787},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021577628078402173,"output_power":0.00017935725325140346,"reference_active":true,"delta_db":-0.8028475879023784},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.554717092269403e-05,"output_power":4.088941405347114e-05,"reference_active":true,"delta_db":-0.46850519611554553}],"active_band_count":3,"active_band_mae_db":0.7802578394761343}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002371340103862925,"log_rms_mae_db":1.2693709867270615,"rms_envelope_nrmse":0.1356200923664838,"envelope_correlation":0.984983426856492,"rms_gain_error_db":-0.8703331264798618,"reference_mean_power":0.008997361160684636,"output_mean_power":0.007363458458886405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.083478820328128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000878683378715502,"output_power":0.0006879320372098769,"reference_active":true,"delta_db":-1.062868761600614},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.731034174913729e-05,"output_power":8.145861541478787e-05,"reference_active":true,"delta_db":-0.7722197384750973},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4907380289326388e-06,"output_power":2.1780475333993185e-06,"reference_active":true,"delta_db":-0.582606981704907}],"active_band_count":3,"active_band_mae_db":0.8058984939268727}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019737503490519328,"log_rms_mae_db":1.7147644225701328,"rms_envelope_nrmse":0.12191014551166351,"envelope_correlation":0.9897475891049149,"rms_gain_error_db":-0.4998835734617674,"reference_mean_power":0.0030829546180243575,"output_mean_power":0.002747759857049523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6914511369831456e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008052884179451132,"output_power":0.0006827150438238021,"reference_active":true,"delta_db":-0.7171198013822055},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019776686451848029,"output_power":0.00016302630240276367,"reference_active":true,"delta_db":-0.8389584966533702},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.986877726391445e-05,"output_power":1.88053914971534e-05,"reference_active":true,"delta_db":-0.23888762052197463}],"active_band_count":3,"active_band_mae_db":0.5983219728525168}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017438355068165795,"log_rms_mae_db":1.3114043844700094,"rms_envelope_nrmse":0.10986183680512863,"envelope_correlation":0.9916332510241304,"rms_gain_error_db":-0.5002072667689159,"reference_mean_power":0.0030981947388459344,"output_mean_power":0.002761137189273881},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5303812378070619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007441051144552584,"output_power":0.0006277189357899684,"reference_active":true,"delta_db":-0.738690597927832},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.887152997212749e-05,"output_power":7.843138637015726e-05,"reference_active":true,"delta_db":-1.0058136286943897},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5498600424852703e-06,"output_power":1.479809239865933e-06,"reference_active":true,"delta_db":-0.20086746994920546}],"active_band_count":3,"active_band_mae_db":0.6484572321904757}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011259893106980884,"log_rms_mae_db":1.8083850714348293,"rms_envelope_nrmse":0.14799840638002706,"envelope_correlation":0.9883967092738536,"rms_gain_error_db":-0.7921001971516959,"reference_mean_power":0.0006356182176644625,"output_mean_power":0.0005296467557714932},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.9581273226653026e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020276258597112104,"output_power":0.0001564827331611992,"reference_active":true,"delta_db":-1.1252139845283236},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6931765404372444e-05,"output_power":2.537611925895198e-05,"reference_active":true,"delta_db":-0.2583961611010533},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.128022173574558e-06,"output_power":6.160093948625281e-06,"reference_active":true,"delta_db":-0.6338170630445996}],"active_band_count":3,"active_band_mae_db":0.6724757362246588}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010212318931682781,"log_rms_mae_db":1.5013129058829426,"rms_envelope_nrmse":0.12450404660684404,"envelope_correlation":0.993254502358802,"rms_gain_error_db":-0.7921081259012563,"reference_mean_power":0.0006388062286465316,"output_mean_power":0.0005323022837758306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.690696537781622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018976599367877945,"output_power":0.00014698998257527718,"reference_active":true,"delta_db":-1.1093065049052113},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3827689740852884e-05,"output_power":1.3020273397871718e-05,"reference_active":true,"delta_db":-0.26129522903551505},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.220792176423388e-07,"output_power":3.655273197043506e-07,"reference_active":true,"delta_db":-0.6247412696216468}],"active_band_count":3,"active_band_mae_db":0.665114334520791}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":1.8396013804414164,"rms_envelope_nrmse":0.2660377836623474,"envelope_correlation":0.9603458414518282,"rms_gain_error_db":-1.3399628871083515,"reference_mean_power":0.0001805297653048896,"output_mean_power":0.00013260274939138147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6198199988921145e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.95993126484054e-05,"output_power":6.966071585935933e-05,"reference_active":true,"delta_db":-1.5526840817847376},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361167220740983e-06,"output_power":8.372953211860315e-06,"reference_active":true,"delta_db":-0.4845133884349801},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4041286689847184e-06,"output_power":3.309706307831495e-06,"reference_active":true,"delta_db":1.3883175026205774}],"active_band_count":3,"active_band_mae_db":1.1418383242800985}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":1.464320056469488,"rms_envelope_nrmse":0.2294236268461793,"envelope_correlation":0.9739463472052882,"rms_gain_error_db":-1.3399704230958844,"reference_mean_power":0.00018143628465930523,"output_mean_power":0.00013326837486819584},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5580116647442146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.473019444861232e-05,"output_power":6.716524397229615e-05,"reference_active":true,"delta_db":-1.4934383213572346},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1793170529251645e-06,"output_power":3.847298478878406e-06,"reference_active":true,"delta_db":-0.3594943820864971},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8116784694632847e-07,"output_power":2.50453503336231e-07,"reference_active":true,"delta_db":1.406459878952166}],"active_band_count":3,"active_band_mae_db":1.086464194131966}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.1787846088409424,"predicted_mos":3.1787846088409424,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"6902a3677883ef558f61f14001837b96ecc7bea655a56d5a6595e7ec6bdb6f1c","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.850455088999297},"started_at":"2026-10-13T07:12:26.582024+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.011814892999609583,"native_input_samples":264600,"native_input_sha256":"8e1443b059aea7dbdaeeb56f0aab2cde8ad0b3d1b8b5545f4e09537974da9603","output_sha256":"1f105000b000752fefe9a8619ec28fcea2f5bf14c5a589694023dc52825bff47","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15310783800123318,"synthesis_s":4.272672787999909,"total_s":4.425780626001142,"rtf":0.7376301043335237,"output_peak":0.7317626476287842,"output_rms":0.08702730972553373,"finished_at":"2026-10-13T07:12:31.034839+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"1f105000b000752fefe9a8619ec28fcea2f5bf14c5a589694023dc52825bff47","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.206540845794358,"output_peak":0.732686793614452},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.438456491683816,"spectral_convergence":0.44200529490220847,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.471822523182006,"spectral_convergence":0.36584994490575007,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.333294935388931,"spectral_convergence":0.4371154902235263,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0003601967830974249,"delta_db":-2.2809762992487066,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.004218085739294428,"delta_db":-0.785340212010716,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0023685853866526893,"delta_db":0.26666569655552186,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.00029441815798694006,"delta_db":-3.146753556041362,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.00011844527320272379,"delta_db":-0.4028965659296107,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.00021762395088759613,"delta_db":-1.6481076316959364,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.421789993580309,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7212941328445742,"pitch":null,"output_rms_dbfs":-21.204823488392428,"output_peak":0.732686793614452,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.414524650084918},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7721659598956616,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09458527529619017,"rms_dbfs":-20.483529355547855,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.58908518923696,"mean":-7.225354512532552e-06,"zero_energy":false},"output":{"rms":0.08704800566998536,"rms_dbfs":-21.204823488392428,"peak":0.732686793614452,"peak_dbfs":-2.7016327274854484,"crest_factor_db":18.50319076090698,"mean":-0.00021220531128441063,"zero_energy":false},"rms_difference_db":-0.721294132844573},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":3.056711961278367e-06,"output_rms_dbfs":-55.1474548359026,"output_zero_energy":false,"output_peak":0.006326211286960881,"output_energy_fraction":1.008502255681425e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":332,"output_estimated_voiced_frames":295,"both_estimated_voiced_frames":285,"reference_voiced_output_unvoiced_frames":47,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.858433734939759,"absolute_difference_cents_median":12.13680169580801,"absolute_difference_cents_p95":38.89966081784329,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002515751859318019,"log_rms_mae_db":1.592740018343879,"rms_envelope_nrmse":0.11772470205622096,"envelope_correlation":0.9898021446685443,"rms_gain_error_db":-0.7212860472130802,"reference_mean_power":0.008953808913712473,"output_mean_power":0.007583666339728595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.40642035802602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010131590544257714,"output_power":0.0008951704007710351,"reference_active":true,"delta_db":-0.5377191638998918},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021577628078402173,"output_power":0.00019834816415565366,"reference_active":true,"delta_db":-0.3657551798881681},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.554717092269403e-05,"output_power":4.414323545747307e-05,"reference_active":true,"delta_db":-0.13597245357568283}],"active_band_count":3,"active_band_mae_db":0.34648226578791425}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002371340103862925,"log_rms_mae_db":1.298875538472401,"rms_envelope_nrmse":0.09660081738880341,"envelope_correlation":0.9957853355216567,"rms_gain_error_db":-0.7214235653872052,"reference_mean_power":0.008997361160684636,"output_mean_power":0.007620312773410617},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.083478820328128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000878683378715502,"output_power":0.0007784833114402401,"reference_active":true,"delta_db":-0.5258310439603033},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.731034174913729e-05,"output_power":9.057199981229483e-05,"reference_active":true,"delta_db":-0.3116504079324605},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4907380289326388e-06,"output_power":2.32272204580518e-06,"reference_active":true,"delta_db":-0.3033080967271161}],"active_band_count":3,"active_band_mae_db":0.38026318287329336}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019737503490519328,"log_rms_mae_db":2.7613134984205514,"rms_envelope_nrmse":0.17701206213280188,"envelope_correlation":0.979259420361955,"rms_gain_error_db":0.5742827573978694,"reference_mean_power":0.0030829546180243575,"output_mean_power":0.0035188066785625145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6914511369831456e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008052884179451132,"output_power":0.0009062911190407542,"reference_active":true,"delta_db":0.5131627136022405},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019776686451848029,"output_power":0.00018715232393942522,"reference_active":true,"delta_db":-0.23958303820372195},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.986877726391445e-05,"output_power":2.0913960547510968e-05,"reference_active":true,"delta_db":0.22265143224677497}],"active_band_count":3,"active_band_mae_db":0.3251323946842458}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017438355068165795,"log_rms_mae_db":2.158124254313712,"rms_envelope_nrmse":0.13762118942955873,"envelope_correlation":0.9880855657135738,"rms_gain_error_db":0.5742656335977324,"reference_mean_power":0.0030981947388459344,"output_mean_power":0.003536187425395722},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5303812378070619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007441051144552584,"output_power":0.000841352634562796,"reference_active":true,"delta_db":0.5334376940919592},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.887152997212749e-05,"output_power":9.081343666419326e-05,"reference_active":true,"delta_db":-0.369211435832325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5498600424852703e-06,"output_power":1.5478296008688007e-06,"reference_active":true,"delta_db":-0.005693338053012694}],"active_band_count":3,"active_band_mae_db":0.30278082265909895}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011259893106980884,"log_rms_mae_db":3.9311586793959292,"rms_envelope_nrmse":0.3736930622019275,"envelope_correlation":0.9236345023840052,"rms_gain_error_db":-2.536697457429205,"reference_mean_power":0.0006356182176644625,"output_mean_power":0.0003544268300108358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.9581273226653026e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020276258597112104,"output_power":0.00014548633071703173,"reference_active":true,"delta_db":-1.441656306857039},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6931765404372444e-05,"output_power":1.8166413460662136e-05,"reference_active":true,"delta_db":-1.709956284436199},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.128022173574558e-06,"output_power":5.600769397394874e-06,"reference_active":true,"delta_db":-1.04721350352093}],"active_band_count":3,"active_band_mae_db":1.3996086982713891}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010212318931682781,"log_rms_mae_db":3.418834269163959,"rms_envelope_nrmse":0.3464581954343308,"envelope_correlation":0.940562466204728,"rms_gain_error_db":-2.5366972892254114,"reference_mean_power":0.0006388062286465316,"output_mean_power":0.00035620450938495554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.690696537781622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018976599367877945,"output_power":0.00013796870039280758,"reference_active":true,"delta_db":-1.3843781546721186},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3827689740852884e-05,"output_power":9.378251315867818e-06,"reference_active":true,"delta_db":-1.6862775981997142},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.220792176423388e-07,"output_power":3.177721112651857e-07,"reference_active":true,"delta_db":-1.2327818944823565}],"active_band_count":3,"active_band_mae_db":1.43447921578473}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":3.333663852797399,"rms_envelope_nrmse":0.24214462937559442,"envelope_correlation":0.9616629105723079,"rms_gain_error_db":-0.7386284407429982,"reference_mean_power":0.0001805297653048896,"output_mean_power":0.0001522951151047496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6198199988921145e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.95993126484054e-05,"output_power":7.227431347533017e-05,"reference_active":true,"delta_db":-1.3927236624154724},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361167220740983e-06,"output_power":1.0812063165220178e-05,"reference_active":true,"delta_db":0.6257857101153508},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4041286689847184e-06,"output_power":3.0254488952437235e-06,"reference_active":true,"delta_db":0.9983211399301017}],"active_band_count":3,"active_band_mae_db":1.0056101708203082}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":2.801284148715809,"rms_envelope_nrmse":0.20108945735007175,"envelope_correlation":0.9734508992087999,"rms_gain_error_db":-0.7386137530269667,"reference_mean_power":0.00018143628465930523,"output_mean_power":0.0001530603735069602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5580116647442146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.473019444861232e-05,"output_power":6.935368898422141e-05,"reference_active":true,"delta_db":-1.3541886223145674},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1793170529251645e-06,"output_power":5.6332437603034276e-06,"reference_active":true,"delta_db":1.2965322531822598},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8116784694632847e-07,"output_power":2.6853986394524957e-07,"reference_active":true,"delta_db":1.7092764147598687}],"active_band_count":3,"active_band_mae_db":1.4533324300855652}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"golf","status":"measured","mos":{"status":"measured","value":2.671227216720581,"predicted_mos":2.671227216720581,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"a62a4ed2280d245fb90607ae7fdc3b0536cc711c64c8a9010e5a51e0cba6f115","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3935327540002618},"started_at":"2026-10-13T05:02:24.622652+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031483950006077066,"native_input_samples":144000,"native_input_sha256":"1ba73157a41449ffb9ff66f3743e4957374e58fbb8ccfaa5034fc9fd18c2743e","output_sha256":"4257364e5ebcf6cd64b74485ed8bcd89f97cd52723b0fb53038b72c15b6228bf","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.8057724860000235,"synthesis_s":0.12779602600039652,"total_s":0.93356851200042,"rtf":0.15559475200007,"output_peak":0.5660176873207092,"output_rms":0.0889292624535795,"finished_at":"2026-10-13T05:02:25.567218+00:00","analysis_stages":{"external_f0_s":0.08531879700058198},"evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"4257364e5ebcf6cd64b74485ed8bcd89f97cd52723b0fb53038b72c15b6228bf","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.019948851175574,"output_peak":0.5662765928447664},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.950697483926984,"spectral_convergence":0.39183467809006844,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.400880313006876,"spectral_convergence":0.26651956106078517,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.482321394037102,"spectral_convergence":0.28431464022915054,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000610037946757047,"output_power":0.0005568099747658755,"delta_db":-0.39649843994386746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005062582286812263,"output_power":0.004618603296698607,"delta_db":-0.3986143405085721,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002231243947515385,"output_power":0.00188473278962844,"delta_db":-0.7329726913075576,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000608645141712815,"output_power":0.0004656928525687589,"delta_db":-1.1626458731669227,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013017616213461652,"output_power":0.00010681355069156175,"delta_db":-0.8590511138031719,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003185972634221577,"output_power":0.00027422693798977616,"delta_db":-0.6513192675769598,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7001836210511753,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5436506561206226,"pitch":null,"output_rms_dbfs":-21.019948851175574,"output_peak":0.5662765928447664,"reference_rms_dbfs":-20.47629819505495,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.611299730323654},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8849321168913777,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09466405203853406,"rms_dbfs":-20.47629819505495,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.581854028744058,"mean":-1.0883031982014295e-05,"zero_energy":false},"output":{"rms":0.08892063541351367,"rms_dbfs":-21.019948851175574,"peak":0.5662765928447664,"peak_dbfs":-4.939427793019501,"crest_factor_db":16.080521058156073,"mean":-0.00031201422907881055,"zero_energy":false},"rms_difference_db":-0.5436506561206222},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":6.071365115429382e-06,"output_rms_dbfs":-52.167136490495956,"output_zero_energy":false,"output_peak":0.013307308875615076,"output_energy_fraction":1.9228511783499934e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":331,"output_estimated_voiced_frames":339,"both_estimated_voiced_frames":320,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9667673716012085,"absolute_difference_cents_median":3.802884195214575,"absolute_difference_cents_p95":28.1597174473809,"difference_over_600_cents_frames":13,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002515751859318019,"log_rms_mae_db":1.5006505248425839,"rms_envelope_nrmse":0.12752486234399205,"envelope_correlation":0.9827374650580947,"rms_gain_error_db":-0.5436469418063729,"reference_mean_power":0.008968739455321504,"output_mean_power":0.007913465505518066},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.4064705193250025e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010133702782121783,"output_power":0.0009690014006136889,"reference_active":true,"delta_db":-0.19443757650611151},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021632531407264513,"output_power":0.00022009941213451566,"reference_active":true,"delta_db":0.07511539674436143},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.5472176438455244e-05,"output_power":5.33127404273241e-05,"reference_active":true,"delta_db":0.6908526570266462}],"active_band_count":3,"active_band_mae_db":0.32013521009237306}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002371340103862925,"log_rms_mae_db":1.0307637224296955,"rms_envelope_nrmse":0.0959032735215086,"envelope_correlation":0.9916031316767073,"rms_gain_error_db":-0.5429355322863645,"reference_mean_power":0.009007019987999644,"output_mean_power":0.00794854382841413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.084766750562158e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008787858059927794,"output_power":0.0008375598271684758,"reference_active":true,"delta_db":-0.20867195038651692},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.747229460240877e-05,"output_power":0.00010243738006329513,"reference_active":true,"delta_db":0.21577272486972504},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.493712627714086e-06,"output_power":2.9351491425332935e-06,"reference_active":true,"delta_db":0.7078376929247077}],"active_band_count":3,"active_band_mae_db":0.3774274560603166}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019737503490519328,"log_rms_mae_db":1.7288658209712877,"rms_envelope_nrmse":0.16391581568815775,"envelope_correlation":0.9792490837125406,"rms_gain_error_db":-0.5177023252640828,"reference_mean_power":0.0030881013820315143,"output_mean_power":0.0027410775154674996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.6938101069166783e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008061124961221262,"output_power":0.000696337918914661,"reference_active":true,"delta_db":-0.6357560829056661},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019803756611354168,"output_power":0.0002143213837783954,"reference_active":true,"delta_db":0.3431792445401547},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9920287623430582e-05,"output_power":3.174904548483615e-05,"reference_active":true,"delta_db":2.024350682438877}],"active_band_count":3,"active_band_mae_db":1.0010953366282325}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017438355068165795,"log_rms_mae_db":1.0843975217850748,"rms_envelope_nrmse":0.11503570531020935,"envelope_correlation":0.9905836737174932,"rms_gain_error_db":-0.5169262572686133,"reference_mean_power":0.0031016416313949147,"output_mean_power":0.0027535881979509403},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5328405171624645e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007446832681989692,"output_power":0.0006478781683123714,"reference_active":true,"delta_db":-0.6047825039557289},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.900332102350934e-05,"output_power":0.00010733632277005277,"reference_active":true,"delta_db":0.35096949669860933},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5499064653408548e-06,"output_power":1.973618151837185e-06,"reference_active":true,"delta_db":1.0495764090558939}],"active_band_count":3,"active_band_mae_db":0.668442803236744}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00011259893106980884,"log_rms_mae_db":1.60376122152134,"rms_envelope_nrmse":0.20389801785018402,"envelope_correlation":0.9742680278317819,"rms_gain_error_db":-0.9500437364325082,"reference_mean_power":0.0006366797664078877,"output_mean_power":0.0005115836717514387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.969188193763372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002031429893342895,"output_power":0.00017356434980176914,"reference_active":true,"delta_db":-0.6834131327664293},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6956503064779867e-05,"output_power":2.7209143210152297e-05,"reference_active":true,"delta_db":0.04051313858468504},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.149319708723847e-06,"output_power":6.283945418266675e-06,"reference_active":true,"delta_db":-0.5603231446672625}],"active_band_count":3,"active_band_mae_db":0.42808313867279235}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010212318931682781,"log_rms_mae_db":1.0904372062529926,"rms_envelope_nrmse":0.1533505558293158,"envelope_correlation":0.989072157785454,"rms_gain_error_db":-0.9501467258901231,"reference_mean_power":0.0006398474491567561,"output_mean_power":0.0005141167699420597},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.700909823047014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019007358252738413,"output_power":0.00016283339372034965,"reference_active":true,"delta_db":-0.6717828602840032},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3835342096842554e-05,"output_power":1.3686938441287831e-05,"reference_active":true,"delta_db":-0.04683588139488653},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.230900204086696e-07,"output_power":3.3947622011926745e-07,"reference_active":true,"delta_db":-0.9562342367503573}],"active_band_count":3,"active_band_mae_db":0.5582843261430823}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":2.172976771185848,"rms_envelope_nrmse":0.20272473258360516,"envelope_correlation":0.9734084894160873,"rms_gain_error_db":0.17130987166406267,"reference_mean_power":0.0001808313146299297,"output_mean_power":0.00018810685688124131},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.6154889353823893e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.912109405551609e-05,"output_power":8.44872774631821e-05,"reference_active":true,"delta_db":-0.6937477141290394},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.309738295917586e-06,"output_power":1.0662749007185768e-05,"reference_active":true,"delta_db":0.5893171358068551},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3973045926800337e-06,"output_power":3.645555405611783e-06,"reference_active":true,"delta_db":1.8204048568937155}],"active_band_count":3,"active_band_mae_db":1.0344899022765366}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":1.7862405939354018,"rms_envelope_nrmse":0.14553625616537316,"envelope_correlation":0.9854715602853186,"rms_gain_error_db":0.1713291999624178,"reference_mean_power":0.0001817379528391398,"output_mean_power":0.0001890508140172643},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5538923550737496e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.421462295326959e-05,"output_power":8.162416987688211e-05,"reference_active":true,"delta_db":-0.6229953691513959},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.145844143751639e-06,"output_power":5.116579363005051e-06,"reference_active":true,"delta_db":0.913667433631348},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8065855730520946e-07,"output_power":2.999703424185285e-07,"reference_active":true,"delta_db":2.20219781109369}],"active_band_count":3,"active_band_mae_db":1.2462868712921447}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.93621563911438,"predicted_mos":2.93621563911438,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"3a67f0401b46bf25253e795d76c5bc4c2de117f6ca5a571d6aab5addd9b4ba9a","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":12.369418933001725},"started_at":"2026-10-13T07:36:06.781339+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0034473300001991447,"native_input_samples":144000,"native_input_sha256":"1ba73157a41449ffb9ff66f3743e4957374e58fbb8ccfaa5034fc9fd18c2743e","output_sha256":"15c1719f6cce44297fc5248416c1233ab450e216a86bd496b2124173396094ab","output_samples":144120,"length_delta_samples":120,"analysis_s":2.2157431409978017,"synthesis_s":15.277865968000697,"total_s":17.4936091089985,"rtf":2.9156015181664166,"output_peak":0.5213633179664612,"output_rms":0.09146797944627968,"finished_at":"2026-10-13T07:36:24.297042+00:00","analysis_stages":{"external_f0_s":1.3190821339994727},"evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"15c1719f6cce44297fc5248416c1233ab450e216a86bd496b2124173396094ab","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.77606068909336,"output_peak":0.5220647730355499},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.212720757348732,"spectral_convergence":0.45803147235114094,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.457129225396249,"spectral_convergence":0.2797826749106103,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.396764339642906,"spectral_convergence":0.28654971227625864,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0005809471639684522,"delta_db":-0.20500823824359615,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.00485318411592566,"delta_db":-0.17622656901204126,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0020522179692049995,"delta_db":-0.35598989880941934,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0005376994061208946,"delta_db":-0.5310039002751925,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.00012598299168193843,"delta_db":-0.13495473398522354,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.00022055011354460004,"delta_db":-1.5901016797150693,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.49888083667342364,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.28891421676366547,"pitch":null,"output_rms_dbfs":-20.77244357231152,"output_peak":0.5220647730355499,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.688871440795963},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8704756938277968,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09458527529619017,"rms_dbfs":-20.483529355547855,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.58908518923696,"mean":-7.225354512532552e-06,"zero_energy":false},"output":{"rms":0.09149088348270852,"rms_dbfs":-20.77244357231152,"peak":0.5220647730355499,"peak_dbfs":-5.6455122071290536,"crest_factor_db":15.126931365182465,"mean":0.0001410089862561242,"zero_energy":false},"rms_difference_db":-0.2889142167636649},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":2.3559511134623406e-06,"output_rms_dbfs":-56.27833725503705,"output_zero_energy":false,"output_peak":0.006239304054948845,"output_energy_fraction":7.036401951758302e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":332,"output_estimated_voiced_frames":337,"both_estimated_voiced_frames":321,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9668674698795181,"absolute_difference_cents_median":7.758447284938406,"absolute_difference_cents_p95":58.84135629317936,"difference_over_600_cents_frames":13,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002515751859318019,"log_rms_mae_db":2.2375955740771807,"rms_envelope_nrmse":0.12916607744710948,"envelope_correlation":0.9792924782778827,"rms_gain_error_db":-0.28890269120799283,"reference_mean_power":0.008953808913712473,"output_mean_power":0.008377560109332946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.40642035802602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010131590544257714,"output_power":0.000951915638422731,"reference_active":true,"delta_db":-0.2707916836773694},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021577628078402173,"output_power":0.0001986695062619182,"reference_active":true,"delta_db":-0.35872490670751755},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.554717092269403e-05,"output_power":4.0949599337613125e-05,"reference_active":true,"delta_db":-0.4621174989956334}],"active_band_count":3,"active_band_mae_db":0.36387802979350675}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002371340103862925,"log_rms_mae_db":1.8469310900059386,"rms_envelope_nrmse":0.08213743211791583,"envelope_correlation":0.9918833575878007,"rms_gain_error_db":-0.2883734777113131,"reference_mean_power":0.008997361160684636,"output_mean_power":0.008419335306150742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.083478820328128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000878683378715502,"output_power":0.0008272477714096768,"reference_active":true,"delta_db":-0.2619680545821907},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.731034174913729e-05,"output_power":9.226295526995078e-05,"reference_active":true,"delta_db":-0.2313163642116441},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4907380289326388e-06,"output_power":2.2207578890071117e-06,"reference_active":true,"delta_db":-0.49826838105748583}],"active_band_count":3,"active_band_mae_db":0.3305175999504402}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019737503490519328,"log_rms_mae_db":2.0486899127834515,"rms_envelope_nrmse":0.1649153881803756,"envelope_correlation":0.9774653971565631,"rms_gain_error_db":-0.2719687562173322,"reference_mean_power":0.0030829546180243575,"output_mean_power":0.0028958112848906665},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6914511369831456e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008052884179451132,"output_power":0.000732761738604123,"reference_active":true,"delta_db":-0.4098866847788723},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019776686451848029,"output_power":0.00019099476010916492,"reference_active":true,"delta_db":-0.15132075442363901},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.986877726391445e-05,"output_power":2.1489456587644792e-05,"reference_active":true,"delta_db":0.34054292256957414}],"active_band_count":3,"active_band_mae_db":0.30058345392402847}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017438355068165795,"log_rms_mae_db":1.3224650490653658,"rms_envelope_nrmse":0.10672795471054519,"envelope_correlation":0.99032953492398,"rms_gain_error_db":-0.27164406993060985,"reference_mean_power":0.0030981947388459344,"output_mean_power":0.0029103438652474255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5303812378070619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007441051144552584,"output_power":0.000676918085591684,"reference_active":true,"delta_db":-0.4109817204657206},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.887152997212749e-05,"output_power":9.718204663808809e-05,"reference_active":true,"delta_db":-0.07485213636202014},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5498600424852703e-06,"output_power":1.6257659536293457e-06,"reference_active":true,"delta_db":0.2076554272408514}],"active_band_count":3,"active_band_mae_db":0.23116309468953075}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011259893106980884,"log_rms_mae_db":1.8913217867771872,"rms_envelope_nrmse":0.173548192179028,"envelope_correlation":0.9777720101980314,"rms_gain_error_db":-0.35250787535211436,"reference_mean_power":0.0006356182176644625,"output_mean_power":0.0005860646797639573},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.9581273226653026e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020276258597112104,"output_power":0.0001826873406160837,"reference_active":true,"delta_db":-0.4527936763839321},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6931765404372444e-05,"output_power":2.509992531499672e-05,"reference_active":true,"delta_db":-0.305923935490311},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.128022173574558e-06,"output_power":6.037677695333897e-06,"reference_active":true,"delta_db":-0.7209911631700059}],"active_band_count":3,"active_band_mae_db":0.493236258348083}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010212318931682781,"log_rms_mae_db":1.1649907584694699,"rms_envelope_nrmse":0.09771119435694008,"envelope_correlation":0.9934777131392654,"rms_gain_error_db":-0.3524716970415998,"reference_mean_power":0.0006388062286465316,"output_mean_power":0.0005890090563293447},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.690696537781622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018976599367877945,"output_power":0.00017127695984678063,"reference_active":true,"delta_db":-0.4451944324837799},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3827689740852884e-05,"output_power":1.2955983070824625e-05,"reference_active":true,"delta_db":-0.28279254626412526},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.220792176423388e-07,"output_power":3.4859371949099775e-07,"reference_active":true,"delta_db":-0.8307441057698512}],"active_band_count":3,"active_band_mae_db":0.5195770281725854}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":2.223748173849751,"rms_envelope_nrmse":0.2276783549226078,"envelope_correlation":0.9665049987170482,"rms_gain_error_db":-0.7381710904788885,"reference_mean_power":0.0001805297653048896,"output_mean_power":0.00015231115396355208},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6198199988921145e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.95993126484054e-05,"output_power":8.152741590952345e-05,"reference_active":true,"delta_db":-0.8695266413724767},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361167220740983e-06,"output_power":7.83434272835032e-06,"reference_active":true,"delta_db":-0.77327436544407},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4041286689847184e-06,"output_power":2.473798968951897e-06,"reference_active":true,"delta_db":0.12406696779890139}],"active_band_count":3,"active_band_mae_db":0.5889559915384828}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":1.4918816273210393,"rms_envelope_nrmse":0.1496248829076163,"envelope_correlation":0.987590211417916,"rms_gain_error_db":-0.7381593783203504,"reference_mean_power":0.00018143628465930523,"output_mean_power":0.00015307638807851237},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5580116647442146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.473019444861232e-05,"output_power":7.7590288880209e-05,"reference_active":true,"delta_db":-0.8668105994960883},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1793170529251645e-06,"output_power":3.5405024169093514e-06,"reference_active":true,"delta_db":-0.720404238128799},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8116784694632847e-07,"output_power":1.9162450192808155e-07,"reference_active":true,"delta_db":0.2436991597102069}],"active_band_count":3,"active_band_mae_db":0.6103046657783647}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.82051420211792,"predicted_mos":2.82051420211792,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"5f4abb9776ecb3180adf99ecd51eaa274018926d4410cdc131694ffe9fb9f20e","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3737209319988324},"started_at":"2026-10-13T05:10:32.848780+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003114686000117217,"native_input_samples":144000,"native_input_sha256":"1ba73157a41449ffb9ff66f3743e4957374e58fbb8ccfaa5034fc9fd18c2743e","output_sha256":"b62c5b329136445c52d51b8c4664461a73c2a2cf5211561fa19596791f578b2a","output_samples":144240,"length_delta_samples":240,"analysis_s":0.6438498000006803,"synthesis_s":0.13935971399951086,"total_s":0.7832095140001911,"rtf":0.13053491900003186,"output_peak":0.5318006277084351,"output_rms":0.09245978093158294,"finished_at":"2026-10-13T05:10:33.644688+00:00","analysis_stages":{"external_f0_s":0.16094604799945955},"evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"b62c5b329136445c52d51b8c4664461a73c2a2cf5211561fa19596791f578b2a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.684355101964368,"output_peak":0.5319120651677838},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.797934131895189,"spectral_convergence":0.20542172404987977,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.64059377840756,"spectral_convergence":0.20180308278077494,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.187936659058441,"spectral_convergence":0.24667274435548964,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006090282805006308,"output_power":0.0006098814684608897,"delta_db":0.006079775537052107,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005054165098664251,"output_power":0.004951908265766801,"delta_db":-0.08876833649943386,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022275244303280044,"output_power":0.0020823248460839533,"delta_db":-0.29273994741134995,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006076310340369887,"output_power":0.0005264957957796692,"delta_db":-0.6224503926560906,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001299593081283461,"output_power":0.00012372114300072783,"delta_db":-0.21363467011625625,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003180661511978585,"output_power":0.00026200025066903265,"delta_db":-0.842157468397512,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3443050984362825,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19359448552384642,"pitch":null,"output_rms_dbfs":-20.677123841071698,"output_peak":0.5319120651677838,"reference_rms_dbfs":-20.48352935554785,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.875488189787063},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9321493279084775,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09458527529619017,"rms_dbfs":-20.483529355547855,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.58908518923696,"mean":-7.225354512532552e-06,"zero_energy":false},"output":{"rms":0.09250044199765249,"rms_dbfs":-20.677123841071698,"peak":0.5319120651677838,"peak_dbfs":-5.483203172480494,"crest_factor_db":15.193920668591206,"mean":5.73761928381852e-05,"zero_energy":false},"rms_difference_db":-0.1935944855238425},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":4.640507926856479e-06,"output_rms_dbfs":-53.334344811351606,"output_zero_energy":false,"output_peak":0.0084245243160502,"output_energy_fraction":1.3558695629523434e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":332,"output_estimated_voiced_frames":315,"both_estimated_voiced_frames":311,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9367469879518072,"absolute_difference_cents_median":1.0464812979586453,"absolute_difference_cents_p95":9.8348621448632,"difference_over_600_cents_frames":6,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002515751859318019,"log_rms_mae_db":0.8029944874971735,"rms_envelope_nrmse":0.06403826925604007,"envelope_correlation":0.9951874492255728,"rms_gain_error_db":-0.19359592425421474,"reference_mean_power":0.008953808913712473,"output_mean_power":0.008563439410798785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.40642035802602e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010131590544257714,"output_power":0.0010057227935801635,"reference_active":true,"delta_db":-0.03199336959385045},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021577628078402173,"output_power":0.00021393551499942275,"reference_active":true,"delta_db":-0.037208162334498196},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.554717092269403e-05,"output_power":4.037119393875762e-05,"reference_active":true,"delta_db":-0.5238981332080175}],"active_band_count":3,"active_band_mae_db":0.1976998883787887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002371340103862925,"log_rms_mae_db":0.5268414434928322,"rms_envelope_nrmse":0.04402281921952424,"envelope_correlation":0.9979165004731446,"rms_gain_error_db":-0.19344887347636508,"reference_mean_power":0.008997361160684636,"output_mean_power":0.008605384231114802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.083478820328128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000878683378715502,"output_power":0.0008754566462589008,"reference_active":true,"delta_db":-0.01597767180654682},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.731034174913729e-05,"output_power":9.818295927993965e-05,"reference_active":true,"delta_db":0.038771200179693106},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4907380289326388e-06,"output_power":2.1967362019441775e-06,"reference_active":true,"delta_db":-0.5455014446672604}],"active_band_count":3,"active_band_mae_db":0.2000834388845001}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019737503490519328,"log_rms_mae_db":0.9704408525245841,"rms_envelope_nrmse":0.0833505809063547,"envelope_correlation":0.9946944146360132,"rms_gain_error_db":-0.23878831371932402,"reference_mean_power":0.0030829546180243575,"output_mean_power":0.0029180202436076864},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6914511369831456e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008052884179451132,"output_power":0.000769696649486876,"reference_active":true,"delta_db":-0.19631856833684816},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019776686451848029,"output_power":0.0001927541044366313,"reference_active":true,"delta_db":-0.11149893566954908},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.986877726391445e-05,"output_power":1.8023727062724984e-05,"reference_active":true,"delta_db":-0.42326539049247514}],"active_band_count":3,"active_band_mae_db":0.24369429816629076}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017438355068165795,"log_rms_mae_db":0.5172815733843805,"rms_envelope_nrmse":0.05682663911856825,"envelope_correlation":0.9977045568494668,"rms_gain_error_db":-0.23866210825912723,"reference_mean_power":0.0030981947388459344,"output_mean_power":0.002932530253986449},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5303812378070619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007441051144552584,"output_power":0.0007142278351591853,"reference_active":true,"delta_db":-0.1779751790613685},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.887152997212749e-05,"output_power":9.711058483909182e-05,"reference_active":true,"delta_db":-0.07804685004050593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5498600424852703e-06,"output_power":1.3447322420340038e-06,"reference_active":true,"delta_db":-0.6165666382305717}],"active_band_count":3,"active_band_mae_db":0.29086288911081537}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011259893106980884,"log_rms_mae_db":0.9569696395217911,"rms_envelope_nrmse":0.1333135722550524,"envelope_correlation":0.9878868687293864,"rms_gain_error_db":-0.4320122359123451,"reference_mean_power":0.0006356182176644625,"output_mean_power":0.0005754334615457414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.9581273226653026e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020276258597112104,"output_power":0.00018637771097884413,"reference_active":true,"delta_db":-0.3659384785548824},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6931765404372444e-05,"output_power":2.4206860078886446e-05,"reference_active":true,"delta_db":-0.4632636291236554},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.128022173574558e-06,"output_power":7.199360843402819e-06,"reference_active":true,"delta_db":0.04324899621623079}],"active_band_count":3,"active_band_mae_db":0.29081703463158953}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010212318931682781,"log_rms_mae_db":0.5799361954985188,"rms_envelope_nrmse":0.06886514608989197,"envelope_correlation":0.9980439210851137,"rms_gain_error_db":-0.43200707108490505,"reference_mean_power":0.0006388062286465316,"output_mean_power":0.0005783202972270274},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.690696537781622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018976599367877945,"output_power":0.0001758690507557475,"reference_active":true,"delta_db":-0.33028969397734137},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3827689740852884e-05,"output_power":1.2359163635531001e-05,"reference_active":true,"delta_db":-0.4876054414448685},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.220792176423388e-07,"output_power":3.7208218904087424e-07,"reference_active":true,"delta_db":-0.5475508730701274}],"active_band_count":3,"active_band_mae_db":0.4551486694974458}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":1.0425990601694213,"rms_envelope_nrmse":0.10833770054631735,"envelope_correlation":0.9921489860137056,"rms_gain_error_db":-0.1732547089016562,"reference_mean_power":0.0001805297653048896,"output_mean_power":0.00017346958786001367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6198199988921145e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.95993126484054e-05,"output_power":0.00010037685510821799,"reference_active":true,"delta_db":0.03377243450854593},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361167220740983e-06,"output_power":9.010142020123552e-06,"reference_active":true,"delta_db":-0.1659836672519311},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4041286689847184e-06,"output_power":2.8154672413731985e-06,"reference_active":true,"delta_db":0.6859277119077423}],"active_band_count":3,"active_band_mae_db":0.29522793788940643}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":0.6471628677124789,"rms_envelope_nrmse":0.05895510682025073,"envelope_correlation":0.997741329072504,"rms_gain_error_db":-0.17325295904885993,"reference_mean_power":0.00018143628465930523,"output_mean_power":0.0001743407252032733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5580116647442146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.473019444861232e-05,"output_power":9.623881651457904e-05,"reference_active":true,"delta_db":0.06861844916008308},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1793170529251645e-06,"output_power":3.99581990360581e-06,"reference_active":true,"delta_db":-0.19499413183022196},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8116784694632847e-07,"output_power":1.8641877879205128e-07,"reference_active":true,"delta_db":0.12408535563575061}],"active_band_count":3,"active_band_mae_db":0.12923264554201855}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0041","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1408987045288086,"predicted_mos":3.1408987045288086,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287716,"input_sample_rate":48000,"input_seconds":5.994083333333333,"input_float32_sha256":"239e73cc3f5a1d0bf21ec27d1334c2086890f3b1b615ab157882154544244e3d","internal_samples":95906,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3131487289992947},"started_at":"2026-10-13T05:14:45.152307+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010719314999732887,"native_input_samples":288000,"native_input_sha256":"42721f95ed753dc1bab61a9398d1a1e553c3b93bea343b7680382264fa961e64","output_sha256":"b71ef84aaa86a7d30215160749b98faeb75c19ff52980acf81d6fac2187c69a6","output_samples":287716,"length_delta_samples":-284,"analysis_s":3.4945329969996237,"synthesis_s":1.096556361999319,"total_s":4.591089358998943,"rtf":0.7651815598331572,"output_peak":0.6559998488934621,"output_rms":0.09044247866774625,"finished_at":"2026-10-13T05:14:49.764732+00:00","evaluation_input":{"reference":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"b71ef84aaa86a7d30215160749b98faeb75c19ff52980acf81d6fac2187c69a6","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95906,"length_delta_samples":-94,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.876615990474313,"output_peak":0.647182080404616},"common_prefix":{"samples":95906,"sample_rate":16000,"reference_coverage":0.9990208333333334,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95906,"output_samples":95906,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.281467170935487,"spectral_convergence":0.42969218838150997,"spectral_convergence_status":"ok","amplitude_floor":1.6781522163747868e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.767248292984069,"spectral_convergence":0.32863882190417876,"spectral_convergence_status":"ok","amplitude_floor":1.6368745556803986e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.304082753147244,"spectral_convergence":0.2955429634026947,"spectral_convergence_status":"ok","amplitude_floor":1.5514964797831996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006102447716979331,"output_power":0.0005442121341424707,"delta_db":-0.4973584637417532,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005058774986237961,"output_power":0.004599136925302713,"delta_db":-0.4136902290400745,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0022294158062184082,"output_power":0.0020566070319909735,"delta_db":-0.35039759465699255,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000608218780120645,"output_power":0.0005162593570315831,"delta_db":-0.7119188965163515,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013012518010875303,"output_power":0.00015208541505168046,"delta_db":0.6772622369753557,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003183408472273083,"output_power":0.00030388811489106356,"delta_db":-0.20178652676339842,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4754023246156543,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.39733026906975616,"pitch":null,"output_rms_dbfs":-20.876615990474313,"output_peak":0.647182080404616,"reference_rms_dbfs":-20.47928572140456,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.7842660723556},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95906,"output_samples":95906,"intelligibility":{"estoi":{"value":0.7963390996975207,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09463149777748954,"rms_dbfs":-20.47928572140456,"peak":0.638671875,"peak_dbfs":-3.8944441663108935,"crest_factor_db":16.584841555093664,"mean":-9.880521743784278e-06,"zero_energy":false},"output":{"rms":0.09040016027867696,"rms_dbfs":-20.876615990474313,"peak":0.647182080404616,"peak_dbfs":-3.77947032510771,"crest_factor_db":17.097145665366604,"mean":-0.001985512906050271,"zero_energy":false},"rms_difference_db":-0.3973302690697551},"full_scale":{"reference":{"sample_count":95906,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95906,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":66,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.32900741766207e-06,"threshold_dbfs":-51.98664395327536,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":15,"quiet_samples":2400,"reference_mean_power":4.280436939249436e-06,"output_mean_power":6.227086968029093e-06,"output_rms_dbfs":-52.057150688423725,"output_zero_energy":false,"output_peak":0.009962738458688504,"output_energy_fraction":1.9068300784113814e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":331,"output_estimated_voiced_frames":318,"both_estimated_voiced_frames":307,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9274924471299094,"absolute_difference_cents_median":5.5252240475897825,"absolute_difference_cents_p95":41.36310867780327,"difference_over_600_cents_frames":3,"uncovered_tail_samples":66,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95906,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1197,"rms_floor":0.0002515751859318019,"log_rms_mae_db":2.417930143656932,"rms_envelope_nrmse":0.17984477426170653,"envelope_correlation":0.9582768436895642,"rms_gain_error_db":-0.3979359859969103,"reference_mean_power":0.008968739455321504,"output_mean_power":0.008183475927041673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.4064705193250025e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010133702782121783,"output_power":0.0009399215045365938,"reference_active":true,"delta_db":-0.32676576467636725},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021632531407264513,"output_power":0.0002121659318789346,"reference_active":true,"delta_db":-0.08431693725529114},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.5472176438455244e-05,"output_power":4.556891560733117e-05,"reference_active":true,"delta_db":0.00922952528247544}],"active_band_count":3,"active_band_mae_db":0.14010407573804462}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1191,"rms_floor":0.0002371340103862925,"log_rms_mae_db":1.788447954834783,"rms_envelope_nrmse":0.12292435711457392,"envelope_correlation":0.9802755810496084,"rms_gain_error_db":-0.4068986903848631,"reference_mean_power":0.009007019987999644,"output_mean_power":0.00820146163107525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.084766750562158e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008787858059927794,"output_power":0.0008156274741679318,"reference_active":true,"delta_db":-0.3239118721794604},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.747229460240877e-05,"output_power":9.736135957034753e-05,"reference_active":true,"delta_db":-0.004945601178088524},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.493712627714086e-06,"output_power":2.7366527478537455e-06,"reference_active":true,"delta_db":0.40373289072611845}],"active_band_count":3,"active_band_mae_db":0.2441967880278891}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1197,"rms_floor":0.00019737503490519328,"log_rms_mae_db":2.9821471721773993,"rms_envelope_nrmse":0.20498311649602458,"envelope_correlation":0.9646243844703829,"rms_gain_error_db":-0.24424298940795286,"reference_mean_power":0.0030881016785350643,"output_mean_power":0.002919223137143057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.6938087926046955e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008061124961264632,"output_power":0.000764713209922883,"reference_active":true,"delta_db":-0.22897061106270614},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019803756611673555,"output_power":0.0001858080151773997,"reference_active":true,"delta_db":-0.2768313604048254},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9920287623541665e-05,"output_power":1.9249595084983543e-05,"reference_active":true,"delta_db":-0.14874006230205478}],"active_band_count":3,"active_band_mae_db":0.21818067792319548}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1191,"rms_floor":0.00017438355068165795,"log_rms_mae_db":2.2059461251740298,"rms_envelope_nrmse":0.14735567128634905,"envelope_correlation":0.9807190543397385,"rms_gain_error_db":-0.254160996725903,"reference_mean_power":0.003101641708234998,"output_mean_power":0.00292533446859688},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5328405057741684e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007446832681990293,"output_power":0.0007084877579450961,"reference_active":true,"delta_db":-0.21639245658748524},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.90033210235262e-05,"output_power":9.21629050651311e-05,"reference_active":true,"delta_db":-0.31093607379426336},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5499064653441906e-06,"output_power":1.4524352675176792e-06,"reference_active":true,"delta_db":-0.2820870415203657}],"active_band_count":3,"active_band_mae_db":0.2698051906340381}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1197,"rms_floor":0.00011259893106980884,"log_rms_mae_db":2.6136940043743926,"rms_envelope_nrmse":0.22331967910551392,"envelope_correlation":0.9625002855963475,"rms_gain_error_db":-0.4366079943299731,"reference_mean_power":0.0006366797668038263,"output_mean_power":0.0005757848705884349},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.9691881737499126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002031429893343061,"output_power":0.0002020007924521861,"reference_active":true,"delta_db":-0.024487657946486843},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6956503064786538e-05,"output_power":2.9170008876448496e-05,"reference_active":true,"delta_db":0.34273008658168513},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.149319708723445e-06,"output_power":7.0352663675152585e-06,"reference_active":true,"delta_db":-0.06984173379442561}],"active_band_count":3,"active_band_mae_db":0.14568649277419918}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1191,"rms_floor":0.00010212318931682781,"log_rms_mae_db":1.804307669237091,"rms_envelope_nrmse":0.15772189585583044,"envelope_correlation":0.9814228031827418,"rms_gain_error_db":-0.4374459837802249,"reference_mean_power":0.0006398474492562399,"output_mean_power":0.0005785379396478821},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.70090982257624e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001900735825273846,"output_power":0.00019016530710761872,"reference_active":true,"delta_db":0.0020952869845329075},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3835342096842654e-05,"output_power":1.4981464816034721e-05,"reference_active":true,"delta_db":0.3456437657238194},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2309002040866685e-07,"output_power":3.788502055553554e-07,"reference_active":true,"delta_db":-0.4796525435306829}],"active_band_count":3,"active_band_mae_db":0.27579719874634506}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1197,"rms_floor":8.113996831641667e-05,"log_rms_mae_db":3.2696327067122284,"rms_envelope_nrmse":0.4065162733875907,"envelope_correlation":0.9030972169204554,"rms_gain_error_db":0.7756200882224991,"reference_mean_power":0.0001808313149683438,"output_mean_power":0.00021619002374831018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.6154889420353469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.91210940555437e-05,"output_power":0.00011840543409362707,"reference_active":true,"delta_db":0.7720554738741068},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.309738295932303e-06,"output_power":1.2588771715652179e-05,"reference_active":true,"delta_db":1.3104588534170583},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.397304592678925e-06,"output_power":6.8933926920081935e-06,"reference_active":true,"delta_db":4.587098020129902}],"active_band_count":3,"active_band_mae_db":2.223204115807022}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":66,"metrics":{"frame_count":1191,"rms_floor":7.456026972891914e-05,"log_rms_mae_db":2.3951208737587835,"rms_envelope_nrmse":0.2658379969876314,"envelope_correlation":0.9590912102785342,"rms_gain_error_db":0.7754261087916343,"reference_mean_power":0.00018173795292416954,"output_mean_power":0.0002172642360290114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5538923553293036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.421462295327057e-05,"output_power":0.00011524364347193998,"reference_active":true,"delta_db":0.8749866579425697},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.145844143751972e-06,"output_power":5.901274183242561e-06,"reference_active":true,"delta_db":1.5333282169648952},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8065855730518095e-07,"output_power":4.696151983626743e-07,"reference_active":true,"delta_db":4.148836059885335}],"active_band_count":3,"active_band_mae_db":2.1857169782642667}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"hnm","status":"measured","mos":{"status":"measured","value":2.491795539855957,"predicted_mos":2.491795539855957,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"92f16c9b9bb50ba909be1a1bb02902e2e3fe2a01b9b1d4e99ea73b43276e60c0","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.75074895599937},"started_at":"2026-10-13T04:39:01.054351+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.7026000023179222e-05,"native_input_samples":96000,"native_input_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","output_sha256":"4c819a12d2935c7f9a133ef77b741749d202d3111df85c6e9ffcdc376d79e59a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0662271020000844,"synthesis_s":0.5081000720001612,"total_s":1.5743271740002456,"rtf":0.26238786233337424,"output_peak":0.6702436918505661,"output_rms":0.06651382816749937,"finished_at":"2026-10-13T04:39:02.634835+00:00","evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"4c819a12d2935c7f9a133ef77b741749d202d3111df85c6e9ffcdc376d79e59a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.806190018121504,"spectral_convergence":0.24241643917133163,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5699951146762805,"spectral_convergence":0.24354450583983484,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.709532359317035,"spectral_convergence":0.2503437993447695,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.00047543696712107136,"delta_db":0.020430830527163255,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.00162050685333725,"delta_db":-0.06634216289284642,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0017540894955832694,"delta_db":-0.1397283096386007,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.0004018559761291129,"delta_db":-0.35089198377087044,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":0.00010936763056532685,"delta_db":-0.06674152106323747,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":6.283241475960256e-05,"delta_db":0.07654031313132822,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12011252017067443,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11094938347458229,"pitch":null,"output_rms_dbfs":-23.54176111723784,"output_peak":0.6702436918505661,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.695239164038274},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.806190018121504,"spectral_convergence":0.24241643917133163,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5699951146762805,"spectral_convergence":0.24354450583983484,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.709532359317035,"spectral_convergence":0.2503437993447695,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.00047543696712107136,"delta_db":0.020430830527163255,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.00162050685333725,"delta_db":-0.06634216289284642,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0017540894955832694,"delta_db":-0.1397283096386007,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.0004018559761291129,"delta_db":-0.35089198377087044,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":0.00010936763056532685,"delta_db":-0.06674152106323747,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":6.283241475960256e-05,"delta_db":0.07654031313132822,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12011252017067443,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11094938347458229,"pitch":null,"output_rms_dbfs":-23.54176111723784,"output_peak":0.6702436918505661,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.695239164038274},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9199766004739945,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06736889333987882,"rms_dbfs":-23.430811733763257,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.418110778783316,"mean":-4.31219736735026e-06,"zero_energy":false},"output":{"rms":0.06651382816749937,"rms_dbfs":-23.541761117237833,"peak":0.6702436918505661,"peak_dbfs":-3.475345295723209,"crest_factor_db":20.066415821514628,"mean":1.15642523221738e-05,"zero_energy":false},"rms_difference_db":-0.11094938347457628},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":3.173703279246656e-06,"output_rms_dbfs":-54.98433679411555,"output_zero_energy":false,"output_peak":0.03674012609089723,"output_energy_fraction":0.00017336410591536013,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":244,"both_estimated_voiced_frames":238,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9333333333333333,"absolute_difference_cents_median":1.9400870926560634,"absolute_difference_cents_p95":12.319635119028897,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027737662806153396,"log_rms_mae_db":1.1440308714940992,"rms_envelope_nrmse":0.07323972570461458,"envelope_correlation":0.9951832900138401,"rms_gain_error_db":-0.11034522756921863,"reference_mean_power":0.004541375985905238,"output_mean_power":0.0044274425688181915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.847489317868867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100074703458764,"output_power":0.0006767975584592795,"reference_active":true,"delta_db":-0.20804134853295947},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018446902300009587,"output_power":0.00014804527018487257,"reference_active":true,"delta_db":-0.9552891070784906},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9341715429184454e-05,"output_power":2.8331152057094814e-05,"reference_active":true,"delta_db":-0.15221265753116633}],"active_band_count":3,"active_band_mae_db":0.4385143710475388}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002314838901045675,"log_rms_mae_db":0.6224895606717109,"rms_envelope_nrmse":0.03867431757475594,"envelope_correlation":0.9986892480120753,"rms_gain_error_db":-0.11018587554240579,"reference_mean_power":0.0045632674272876395,"output_mean_power":0.004448948039244587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7204098415154257e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818437378247786,"output_power":0.0006434833293480999,"reference_active":true,"delta_db":-0.25147556049561004},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591872232322897e-05,"output_power":6.199078489309896e-05,"reference_active":true,"delta_db":-0.8802175532330827},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0593810422650423e-06,"output_power":1.4298820185201266e-06,"reference_active":true,"delta_db":-1.584365057804936}],"active_band_count":3,"active_band_mae_db":0.9053527238445428}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002609159645197948,"log_rms_mae_db":1.4175379344300845,"rms_envelope_nrmse":0.10799193643439702,"envelope_correlation":0.9917448751872026,"rms_gain_error_db":-0.12715394079334785,"reference_mean_power":0.0022069594688974205,"output_mean_power":0.002143280256207209},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0859718062152696e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048225083122535774,"output_power":0.0004408070660839357,"reference_active":true,"delta_db":-0.3902443729506271},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016124782840735576,"output_power":0.00013632778308408518,"reference_active":true,"delta_db":-0.7290950209983239},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2038625150435638e-05,"output_power":1.844734540301562e-05,"reference_active":true,"delta_db":-0.7725061870528056}],"active_band_count":3,"active_band_mae_db":0.6306151936672522}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001981315811908706,"log_rms_mae_db":0.9833239715124513,"rms_envelope_nrmse":0.06354510669584863,"envelope_correlation":0.9970802952631009,"rms_gain_error_db":-0.1272381878824504,"reference_mean_power":0.002217812362061406,"output_mean_power":0.0021537782212311916},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0257217759681142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004916402889453351,"output_power":0.0004440472532579406,"reference_active":true,"delta_db":-0.4421827725182737},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.925750439584526e-05,"output_power":6.130883250785336e-05,"reference_active":true,"delta_db":-0.5294379223738179},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6222165494821603e-06,"output_power":1.4054671982848042e-06,"reference_active":true,"delta_db":-0.6228811344814418}],"active_band_count":3,"active_band_mae_db":0.5315006097911777}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014069890621343104,"log_rms_mae_db":1.4628885508074658,"rms_envelope_nrmse":0.1431808164810319,"envelope_correlation":0.9873210665608301,"rms_gain_error_db":-0.3252017725146582,"reference_mean_power":0.000503617873892436,"output_mean_power":0.0004672840507222227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6929874202041835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020569637799863847,"output_power":0.00019458375080747478,"reference_active":true,"delta_db":-0.24120073872159964},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.720458225580812e-05,"output_power":5.6660564701385414e-05,"reference_active":true,"delta_db":-0.7411798730493419},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.876518957296052e-06,"output_power":7.1866638273629396e-06,"reference_active":true,"delta_db":0.19158685616113932}],"active_band_count":3,"active_band_mae_db":0.39132248931069363}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012225910578362226,"log_rms_mae_db":0.8769994950370524,"rms_envelope_nrmse":0.08364176105359276,"envelope_correlation":0.9961109239938556,"rms_gain_error_db":-0.325363330399873,"reference_mean_power":0.0005061264034494053,"output_mean_power":0.00046959413159876524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5254708519662443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021591129559099358,"output_power":0.00020399826647143454,"reference_active":true,"delta_db":-0.24648886518959964},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.951038057605059e-05,"output_power":2.5192126140681127e-05,"reference_active":true,"delta_db":-0.6870998813728507},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.54700119986915e-07,"output_power":5.497322347677e-07,"reference_active":true,"delta_db":-1.9166256139584021}],"active_band_count":3,"active_band_mae_db":0.9500714535069509}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":1.9400085865565515,"rms_envelope_nrmse":0.19536106228783096,"envelope_correlation":0.9780554369665073,"rms_gain_error_db":0.37155953612153136,"reference_mean_power":4.741724973734132e-05,"output_mean_power":5.165261512355902e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.272212516285401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.356579821454211e-05,"output_power":3.7059783752022265e-05,"reference_active":true,"delta_db":0.4300590140997088},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.145230698082003e-06,"output_power":6.104556258414658e-06,"reference_active":true,"delta_db":-0.6836215495703792},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6230912610162062e-06,"output_power":1.2921903665525665e-06,"reference_active":true,"delta_db":-0.9901644045082787}],"active_band_count":3,"active_band_mae_db":0.7012816560594555}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":1.249990190109471,"rms_envelope_nrmse":0.12286790290961161,"envelope_correlation":0.9915712716541395,"rms_gain_error_db":0.3710806126546054,"reference_mean_power":4.7645706194829206e-05,"output_mean_power":5.189575440238374e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.96519383630983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1852605209893146e-05,"output_power":3.503314988858452e-05,"reference_active":true,"delta_db":0.4133422834059481},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5945787260643978e-06,"output_power":2.8103038244189733e-06,"reference_active":true,"delta_db":-1.0689472527782402},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8710953646003796e-07,"output_power":1.3521479037971634e-07,"reference_active":true,"delta_db":-1.4107172364667873}],"active_band_count":3,"active_band_mae_db":0.9643355908836586}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.4201157093048096,"predicted_mos":2.4201157093048096,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d895d460d2fcf55866eec129519ed0848a6052280fee5a25845280e19c3cf181","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6313081919997785},"started_at":"2026-10-13T04:41:58.821528+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.588299958006246e-05,"native_input_samples":96000,"native_input_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","output_sha256":"8b4321bad1eafbc0df0cedf9359368b20d59bf858ff574e12b8b59df3f5bb08a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4355721229994742,"synthesis_s":0.1039653489997363,"total_s":1.5395374719992105,"rtf":0.2565895786665351,"output_peak":1.0404653337314105,"output_rms":0.08078936712566974,"finished_at":"2026-10-13T04:42:00.367291+00:00","evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"8b4321bad1eafbc0df0cedf9359368b20d59bf858ff574e12b8b59df3f5bb08a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.548487128178644,"spectral_convergence":0.5692491565247899,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.590656251211849,"spectral_convergence":0.3355208665137117,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.92582240752167,"spectral_convergence":0.3763280153019956,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.0005543625990109466,"delta_db":0.6874405841602099,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.0022661861485937526,"delta_db":1.3901049237003864,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.002966786411118398,"delta_db":2.1426170238595623,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.000538204747624249,"delta_db":0.9178789483168919,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":0.00013106357872099886,"delta_db":0.7191906689963795,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":7.03183554978999e-05,"delta_db":0.5653898612484355,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0704370017136442,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5778958545731403,"pitch":null,"output_rms_dbfs":-21.852915879190114,"output_peak":1.0404653337314105,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.021655262304054},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.548487128178644,"spectral_convergence":0.5692491565247899,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.590656251211849,"spectral_convergence":0.3355208665137117,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.92582240752167,"spectral_convergence":0.3763280153019956,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.0005543625990109466,"delta_db":0.6874405841602099,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.0022661861485937526,"delta_db":1.3901049237003864,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.002966786411118398,"delta_db":2.1426170238595623,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.000538204747624249,"delta_db":0.9178789483168919,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":0.00013106357872099886,"delta_db":0.7191906689963795,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":7.03183554978999e-05,"delta_db":0.5653898612484355,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0704370017136442,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5778958545731403,"pitch":null,"output_rms_dbfs":-21.852915879190114,"output_peak":1.0404653337314105,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.021655262304054},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9353423608126441,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06736889333987882,"rms_dbfs":-23.430811733763257,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.418110778783316,"mean":-4.31219736735026e-06,"zero_energy":false},"output":{"rms":0.08078936712566974,"rms_dbfs":-21.852915879190114,"peak":1.0404653337314105,"peak_dbfs":0.34455229893058825,"crest_factor_db":22.1974681781207,"mean":1.0477581469701149e-08,"zero_energy":false},"rms_difference_db":1.5778958545731427},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":1,"at_or_above_full_scale_fraction":1.0416666666666666e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":1.0416666666666666e-05,"longest_full_scale_run_samples":1,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":1.807353502863256e-06,"output_rms_dbfs":-57.42956894850332,"output_zero_energy":false,"output_peak":0.02156703445527421,"output_energy_fraction":6.691930854918803e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":292,"both_estimated_voiced_frames":246,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":46,"paired_voiced_fraction":0.9647058823529412,"absolute_difference_cents_median":7.620006806185379,"absolute_difference_cents_p95":32.84565472504505,"difference_over_600_cents_frames":5,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027737662806153396,"log_rms_mae_db":2.970558749368376,"rms_envelope_nrmse":0.24378325448687174,"envelope_correlation":0.9859645744184427,"rms_gain_error_db":1.5786512664151364,"reference_mean_power":0.004541375985905238,"output_mean_power":0.00653209640603974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.847489317868867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100074703458764,"output_power":0.0009171094737177564,"reference_active":true,"delta_db":1.111582615460803},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018446902300009587,"output_power":0.00022392944361042762,"reference_active":true,"delta_db":0.8418775339905754},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9341715429184454e-05,"output_power":4.017873280483756e-05,"reference_active":true,"delta_db":1.365107346649363}],"active_band_count":3,"active_band_mae_db":1.1061891653669138}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002314838901045675,"log_rms_mae_db":2.62509149567675,"rms_envelope_nrmse":0.2227755505693893,"envelope_correlation":0.992994648839792,"rms_gain_error_db":1.5792411472605004,"reference_mean_power":0.0045632674272876395,"output_mean_power":0.006564475560448845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7204098415154257e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818437378247786,"output_power":0.0008762742034870923,"reference_active":true,"delta_db":1.0895517053100037},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591872232322897e-05,"output_power":9.403278936984134e-05,"reference_active":true,"delta_db":0.9293042861749813},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0593810422650423e-06,"output_power":2.3651108880231805e-06,"reference_active":true,"delta_db":0.6011479690644557}],"active_band_count":3,"active_band_mae_db":0.8733346535164802}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002609159645197948,"log_rms_mae_db":1.8816258808898698,"rms_envelope_nrmse":0.3078196151807226,"envelope_correlation":0.9844333116723402,"rms_gain_error_db":1.9861918889957497,"reference_mean_power":0.0022069594688974205,"output_mean_power":0.003486691686675542},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0859718062152696e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048225083122535774,"output_power":0.0006654315836302269,"reference_active":true,"delta_db":1.3983042522220606},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016124782840735576,"output_power":0.00021755269419622563,"reference_active":true,"delta_db":1.3007059149275604},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2038625150435638e-05,"output_power":2.7064669039905928e-05,"reference_active":true,"delta_db":0.8921822253685956}],"active_band_count":3,"active_band_mae_db":1.1970641308394054}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001981315811908706,"log_rms_mae_db":1.8118296101346378,"rms_envelope_nrmse":0.2845532755534928,"envelope_correlation":0.9917819643622273,"rms_gain_error_db":1.9862404358256982,"reference_mean_power":0.002217812362061406,"output_mean_power":0.003503876928823987},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0257217759681142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004916402889453351,"output_power":0.000674606113686796,"reference_active":true,"delta_db":1.374028074660356},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.925750439584526e-05,"output_power":9.427314142218649e-05,"reference_active":true,"delta_db":1.3392114086724582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6222165494821603e-06,"output_power":1.6055942032069697e-06,"reference_active":true,"delta_db":-0.04473036143006592}],"active_band_count":3,"active_band_mae_db":0.9193232815876268}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014069890621343104,"log_rms_mae_db":1.285281028914901,"rms_envelope_nrmse":0.17894832924764367,"envelope_correlation":0.9885259510172231,"rms_gain_error_db":0.9042488073259741,"reference_mean_power":0.000503617873892436,"output_mean_power":0.0006201917945139288},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6929874202041835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020569637799863847,"output_power":0.00025977216703366817,"reference_active":true,"delta_db":1.0136597279188762},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.720458225580812e-05,"output_power":8.100928634574873e-05,"reference_active":true,"delta_db":0.8113592039592505},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.876518957296052e-06,"output_power":8.359327460113862e-06,"reference_active":true,"delta_db":0.8480269359323745}],"active_band_count":3,"active_band_mae_db":0.8910152892701672}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012225910578362226,"log_rms_mae_db":0.9272586311087183,"rms_envelope_nrmse":0.13476814750297564,"envelope_correlation":0.9963413711201631,"rms_gain_error_db":0.9042860123593104,"reference_mean_power":0.0005061264034494053,"output_mean_power":0.000623286320367015},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5254708519662443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021591129559099358,"output_power":0.00027105202711863994,"reference_active":true,"delta_db":0.9877729616013944},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.951038057605059e-05,"output_power":3.661112898685444e-05,"reference_active":true,"delta_db":0.9363831119173348},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.54700119986915e-07,"output_power":8.251708777563483e-07,"reference_active":true,"delta_db":-0.1526987264152442}],"active_band_count":3,"active_band_mae_db":0.6922849333113246}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":1.6784649233800761,"rms_envelope_nrmse":0.22862482801550604,"envelope_correlation":0.971833445849819,"rms_gain_error_db":0.5743888752142552,"reference_mean_power":4.741724973734132e-05,"output_mean_power":5.412217587720891e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.272212516285401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.356579821454211e-05,"output_power":3.561609872496351e-05,"reference_active":true,"delta_db":0.2574936706219691},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.145230698082003e-06,"output_power":9.612245199167967e-06,"reference_active":true,"delta_db":1.2880858528899253},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6230912610162062e-06,"output_power":1.625432847044001e-06,"reference_active":true,"delta_db":0.006260923626054916}],"active_band_count":3,"active_band_mae_db":0.5172801490459831}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":1.02624527198906,"rms_envelope_nrmse":0.13924109777755586,"envelope_correlation":0.99080551380245,"rms_gain_error_db":0.5744624054573387,"reference_mean_power":4.7645706194829206e-05,"output_mean_power":5.438385745314884e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.96519383630983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1852605209893146e-05,"output_power":3.432007860860148e-05,"reference_active":true,"delta_db":0.324033150403619},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5945787260643978e-06,"output_power":4.846502784498215e-06,"reference_active":true,"delta_db":1.2978046691229044},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8710953646003796e-07,"output_power":1.983863985165829e-07,"reference_active":true,"delta_db":0.2541597050105001}],"active_band_count":3,"active_band_mae_db":0.6253325081790079}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"facodec","status":"measured","mos":{"status":"measured","value":3.4174838066101074,"predicted_mos":3.4174838066101074,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"0787e257f065a42e31b0c8889da186a2052ad7cc7c1010961f1ef1a734d959be","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7347511510015465},"started_at":"2026-10-13T04:51:17.041955+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.684500032046344e-05,"native_input_samples":96000,"native_input_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","output_sha256":"862d651b95d8ed3b34eaa3a5c8d5834684e252e64189106a6baf0a721953d995","output_samples":96000,"length_delta_samples":0,"analysis_s":2.1604686840000795,"synthesis_s":5.607322287999523,"total_s":7.767790971999602,"rtf":1.2946318286666003,"output_peak":0.5666621923446655,"output_rms":0.06252127477537557,"finished_at":"2026-10-13T04:51:24.817010+00:00","evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"862d651b95d8ed3b34eaa3a5c8d5834684e252e64189106a6baf0a721953d995","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.162799172839707,"spectral_convergence":0.20807430560573797,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.563543720057273,"spectral_convergence":0.21832178397642543,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.746766826936629,"spectral_convergence":0.23805762788962992,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.00045739049023165324,"delta_db":-0.14762732987440647,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.0013712731513649868,"delta_db":-0.7916111546085747,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0015626084127217674,"delta_db":-0.6417442062251508,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.00037464194019341064,"delta_db":-0.6554323572005795,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":8.901933276473685e-05,"delta_db":-0.9607862064352907,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":5.397647226145832e-05,"delta_db":-0.5832522244244351,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6300755797947395,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.648631776841057,"pitch":null,"output_rms_dbfs":-24.07944351060431,"output_peak":0.5666621923446655,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.49103657327787},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.162799172839707,"spectral_convergence":0.20807430560573797,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.563543720057273,"spectral_convergence":0.21832178397642543,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.746766826936629,"spectral_convergence":0.23805762788962992,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.00045739049023165324,"delta_db":-0.14762732987440647,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.0013712731513649868,"delta_db":-0.7916111546085747,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0015626084127217674,"delta_db":-0.6417442062251508,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.00037464194019341064,"delta_db":-0.6554323572005795,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":8.901933276473685e-05,"delta_db":-0.9607862064352907,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":5.397647226145832e-05,"delta_db":-0.5832522244244351,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6300755797947395,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.648631776841057,"pitch":null,"output_rms_dbfs":-24.07944351060431,"output_peak":0.5666621923446655,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.49103657327787},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9452190872500739,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06736889333987882,"rms_dbfs":-23.430811733763257,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.418110778783316,"mean":-4.31219736735026e-06,"zero_energy":false},"output":{"rms":0.06252127477537557,"rms_dbfs":-24.07944351060431,"peak":0.5666621923446655,"peak_dbfs":-4.933515249687543,"crest_factor_db":19.14592826091677,"mean":-7.464010584726566e-06,"zero_energy":false},"rms_difference_db":-0.6486317768410537},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":2.5350105330558185e-06,"output_rms_dbfs":-55.9602023181841,"output_zero_energy":false,"output_peak":0.02080056630074978,"output_energy_fraction":0.00015672593559485425,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":269,"both_estimated_voiced_frames":250,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9803921568627451,"absolute_difference_cents_median":3.2249673170648743,"absolute_difference_cents_p95":14.953621223300676,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027737662806153396,"log_rms_mae_db":1.3498642012824549,"rms_envelope_nrmse":0.11395221052025134,"envelope_correlation":0.9923725950132952,"rms_gain_error_db":-0.6482465421604017,"reference_mean_power":0.004541375985905238,"output_mean_power":0.003911675364807972},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.847489317868867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100074703458764,"output_power":0.0005746513300439424,"reference_active":true,"delta_db":-0.9185850194924118},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018446902300009587,"output_power":0.0001604689914371358,"reference_active":true,"delta_db":-0.6053232458334284},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9341715429184454e-05,"output_power":3.449648941307921e-05,"reference_active":true,"delta_db":0.70289399923399}],"active_band_count":3,"active_band_mae_db":0.7422674215199434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002314838901045675,"log_rms_mae_db":1.0751069301656637,"rms_envelope_nrmse":0.10459466594950835,"envelope_correlation":0.9941225514316614,"rms_gain_error_db":-0.6479245226224546,"reference_mean_power":0.0045632674272876395,"output_mean_power":0.003930822821431963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7204098415154257e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818437378247786,"output_power":0.0005587079326066133,"reference_active":true,"delta_db":-0.8650001864380584},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591872232322897e-05,"output_power":6.472361420863237e-05,"reference_active":true,"delta_db":-0.6928612993955547},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0593810422650423e-06,"output_power":2.1007226411152803e-06,"reference_active":true,"delta_db":0.08632005687759095}],"active_band_count":3,"active_band_mae_db":0.5480605142370681}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002609159645197948,"log_rms_mae_db":1.2944491702146836,"rms_envelope_nrmse":0.10025010695527228,"envelope_correlation":0.995182246530519,"rms_gain_error_db":-0.5312968852488729,"reference_mean_power":0.0022069594688974205,"output_mean_power":0.0019528310369201897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0859718062152696e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048225083122535774,"output_power":0.0004138832892061582,"reference_active":true,"delta_db":-0.6639509303622083},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016124782840735576,"output_power":0.00014532495643433034,"reference_active":true,"delta_db":-0.4515367308419557},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2038625150435638e-05,"output_power":2.540216817468365e-05,"reference_active":true,"delta_db":0.6168628878337756}],"active_band_count":3,"active_band_mae_db":0.5774501830126465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001981315811908706,"log_rms_mae_db":1.1343509884255338,"rms_envelope_nrmse":0.08789574424006716,"envelope_correlation":0.9968071958587499,"rms_gain_error_db":-0.5312683730271331,"reference_mean_power":0.002217812362061406,"output_mean_power":0.001962447117727017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0257217759681142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004916402889453351,"output_power":0.00042418635330178156,"reference_active":true,"delta_db":-0.6409077272935444},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.925750439584526e-05,"output_power":6.008241049099267e-05,"reference_active":true,"delta_db":-0.617194901286524},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6222165494821603e-06,"output_power":1.7435076268668102e-06,"reference_active":true,"delta_db":0.3131502385018696}],"active_band_count":3,"active_band_mae_db":0.5237509556939793}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014069890621343104,"log_rms_mae_db":1.3770276146518645,"rms_envelope_nrmse":0.1415268298795781,"envelope_correlation":0.9893399928388035,"rms_gain_error_db":-0.6002846389244381,"reference_mean_power":0.000503617873892436,"output_mean_power":0.0004386040841232822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6929874202041835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020569637799863847,"output_power":0.00015841325321731288,"reference_active":true,"delta_db":-1.13435131645424},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.720458225580812e-05,"output_power":5.2482297303661057e-05,"reference_active":true,"delta_db":-1.0738604873642212},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.876518957296052e-06,"output_power":7.844026362162702e-06,"reference_active":true,"delta_db":0.5717039997471733}],"active_band_count":3,"active_band_mae_db":0.9266386011885448}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012225910578362226,"log_rms_mae_db":1.116245533586195,"rms_envelope_nrmse":0.11714742143227488,"envelope_correlation":0.9933667206355072,"rms_gain_error_db":-0.6002461150385225,"reference_mean_power":0.0005061264034494053,"output_mean_power":0.00044079268885426553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5254708519662443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021591129559099358,"output_power":0.00016676387213366826,"reference_active":true,"delta_db":-1.1217339286198897},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.951038057605059e-05,"output_power":2.3009237682976033e-05,"reference_active":true,"delta_db":-1.0807258001274682},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.54700119986915e-07,"output_power":7.585474437300936e-07,"reference_active":true,"delta_db":-0.5183101578917804}],"active_band_count":3,"active_band_mae_db":0.9069232955463793}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":1.9885240956470986,"rms_envelope_nrmse":0.2882485648643587,"envelope_correlation":0.9658710839348681,"rms_gain_error_db":-1.908728003595892,"reference_mean_power":4.741724973734132e-05,"output_mean_power":3.0553682432495087e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.272212516285401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.356579821454211e-05,"output_power":1.618724698687022e-05,"reference_active":true,"delta_db":-3.1672398614116517},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.145230698082003e-06,"output_power":5.779266239737334e-06,"reference_active":true,"delta_db":-0.9214355337739509},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6230912610162062e-06,"output_power":1.6089364111053893e-06,"reference_active":true,"delta_db":-0.0380405933104332}],"active_band_count":3,"active_band_mae_db":1.3755719961653454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":1.6770581375424456,"rms_envelope_nrmse":0.2658240569306676,"envelope_correlation":0.9750387596606618,"rms_gain_error_db":-1.9095036019826535,"reference_mean_power":4.7645706194829206e-05,"output_mean_power":3.069540784502258e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.96519383630983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1852605209893146e-05,"output_power":1.578647547737585e-05,"reference_active":true,"delta_db":-3.0485977956327197},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5945787260643978e-06,"output_power":3.113933046204774e-06,"reference_active":true,"delta_db":-0.6233872926911019},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8710953646003796e-07,"output_power":1.5390052815644318e-07,"reference_active":true,"delta_db":-0.8485581261549286}],"active_band_count":3,"active_band_mae_db":1.5068477381595835}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.9741625785827637,"predicted_mos":2.9741625785827637,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"64d2558d1cbcd24e16081787505e9fe1ec879aa76c4d2a55b48ef9ab0a7ec2ac","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.170633674999408},"started_at":"2026-10-13T07:12:31.035642+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007171744999141083,"native_input_samples":264600,"native_input_sha256":"ad88025d1575be8173663d9724eedb303d310e7055430c5c206c955e48296bcd","output_sha256":"2cd30361d4e5ab7a4b9c11335e52d91bf15a8fe8327507fa7767f23d038d8b29","output_samples":264704,"length_delta_samples":104,"analysis_s":0.17837119300020277,"synthesis_s":4.063470694998614,"total_s":4.241841887998817,"rtf":0.7069736479998028,"output_peak":0.7069530487060547,"output_rms":0.06377639726416816,"finished_at":"2026-10-13T07:12:35.296276+00:00","evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"2cd30361d4e5ab7a4b9c11335e52d91bf15a8fe8327507fa7767f23d038d8b29","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.90182247966054,"output_peak":0.6981084097310496},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.0203929500729,"spectral_convergence":0.4438008039007581,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.46293630773674,"spectral_convergence":0.3432948949618394,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.243420843454384,"spectral_convergence":0.41537991823149845,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.0002897670582130404,"delta_db":-2.130008520367274,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.0015746349128621844,"delta_db":-0.19105211944353723,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0018020303779065598,"delta_db":-0.022624708760336062,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.00028780606989736485,"delta_db":-1.8005968055210888,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":7.520304324090289e-05,"delta_db":-1.6932753985425673,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":4.4079261452956164e-05,"delta_db":-1.462954101098103,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2167519422888178,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.46948896547650076,"pitch":null,"output_rms_dbfs":-23.900300699239754,"output_peak":0.6981084097310496,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.575583367088008},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8428074387473296,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06736889333987882,"rms_dbfs":-23.430811733763257,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.418110778783316,"mean":-4.31219736735026e-06,"zero_energy":false},"output":{"rms":0.06382413903510967,"rms_dbfs":-23.900300699239754,"peak":0.6981084097310496,"peak_dbfs":-3.121542605069391,"crest_factor_db":20.778758094170364,"mean":-0.00012265502182881124,"zero_energy":false},"rms_difference_db":-0.4694889654764971},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":1.3481435038534075e-06,"output_rms_dbfs":-58.70263876632635,"output_zero_energy":false,"output_peak":0.008090476890243854,"output_energy_fraction":7.998028459244052e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":223,"both_estimated_voiced_frames":205,"reference_voiced_output_unvoiced_frames":50,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.803921568627451,"absolute_difference_cents_median":9.448387349803431,"absolute_difference_cents_p95":41.763111867080355,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027737662806153396,"log_rms_mae_db":1.9513570569756116,"rms_envelope_nrmse":0.111262162966672,"envelope_correlation":0.9915360920218711,"rms_gain_error_db":-0.4690688090505293,"reference_mean_power":0.004541375985905238,"output_mean_power":0.004076435530164266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.847489317868867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100074703458764,"output_power":0.0007213838177850434,"reference_active":true,"delta_db":0.06903477743731469},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018446902300009587,"output_power":0.00020632596097417674,"reference_active":true,"delta_db":0.4863042891132534},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9341715429184454e-05,"output_power":3.148805588754348e-05,"reference_active":true,"delta_db":0.30660346802769156}],"active_band_count":3,"active_band_mae_db":0.28731417819275323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002314838901045675,"log_rms_mae_db":1.5792380658267637,"rms_envelope_nrmse":0.08408007640051132,"envelope_correlation":0.9963116261969346,"rms_gain_error_db":-0.4687204310591827,"reference_mean_power":0.0045632674272876395,"output_mean_power":0.004096414341735133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7204098415154257e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818437378247786,"output_power":0.0007074738528627695,"reference_active":true,"delta_db":0.16025537425613676},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591872232322897e-05,"output_power":8.680099678956319e-05,"reference_active":true,"delta_db":0.5817582197073857},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0593810422650423e-06,"output_power":2.2326421962792818e-06,"reference_active":true,"delta_db":0.35082418015769645}],"active_band_count":3,"active_band_mae_db":0.36427925804040634}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002609159645197948,"log_rms_mae_db":2.079735165803409,"rms_envelope_nrmse":0.14279319657587144,"envelope_correlation":0.9864652154693402,"rms_gain_error_db":0.2328856601114243,"reference_mean_power":0.0022069594688974205,"output_mean_power":0.0023285358214657735},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0859718062152696e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048225083122535774,"output_power":0.0006037404434038417,"reference_active":true,"delta_db":0.9757728449217605},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016124782840735576,"output_power":0.00019270004442284715,"reference_active":true,"delta_db":0.7738794025332847},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2038625150435638e-05,"output_power":2.5310409151763952e-05,"reference_active":true,"delta_db":0.6011466758371817}],"active_band_count":3,"active_band_mae_db":0.783599641097409}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001981315811908706,"log_rms_mae_db":1.6979267652329235,"rms_envelope_nrmse":0.10800243291087104,"envelope_correlation":0.9920698419428732,"rms_gain_error_db":0.23303919906642928,"reference_mean_power":0.002217812362061406,"output_mean_power":0.0023400693042379605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0257217759681142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004916402889453351,"output_power":0.0006182575178567336,"reference_active":true,"delta_db":0.9952194073461108},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.925750439584526e-05,"output_power":8.230237968289642e-05,"reference_active":true,"delta_db":0.7494555428347155},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6222165494821603e-06,"output_power":1.619932877681024e-06,"reference_active":true,"delta_db":-0.006118078441559603}],"active_band_count":3,"active_band_mae_db":0.583597676207462}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014069890621343104,"log_rms_mae_db":2.889166908029534,"rms_envelope_nrmse":0.25586767293011903,"envelope_correlation":0.975029320696408,"rms_gain_error_db":-1.7516542157910764,"reference_mean_power":0.000503617873892436,"output_mean_power":0.0003364617610207751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6929874202041835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020569637799863847,"output_power":0.0001249889720219052,"reference_active":true,"delta_db":-2.1635494829711774},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.720458225580812e-05,"output_power":4.3284604813354944e-05,"reference_active":true,"delta_db":-1.9106542907905146},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.876518957296052e-06,"output_power":5.61888219308622e-06,"reference_active":true,"delta_db":-0.8771871793636178}],"active_band_count":3,"active_band_mae_db":1.6504636510417698}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012225910578362226,"log_rms_mae_db":2.5796487208712926,"rms_envelope_nrmse":0.22776469393924847,"envelope_correlation":0.9849000934899567,"rms_gain_error_db":-1.7515658854699123,"reference_mean_power":0.0005061264034494053,"output_mean_power":0.000338144560403905},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5254708519662443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021591129559099358,"output_power":0.00013293239991459633,"reference_active":true,"delta_db":-2.106445180084821},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.951038057605059e-05,"output_power":1.9055767349942837e-05,"reference_active":true,"delta_db":-1.8994836843665928},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.54700119986915e-07,"output_power":6.826877051647498e-07,"reference_active":true,"delta_db":-0.9759168341342489}],"active_band_count":3,"active_band_mae_db":1.6606152328618877}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":3.381365819152046,"rms_envelope_nrmse":0.27298925988913014,"envelope_correlation":0.9543368802226633,"rms_gain_error_db":-0.8737750724743032,"reference_mean_power":4.741724973734132e-05,"output_mean_power":3.8775629179675666e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.272212516285401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.356579821454211e-05,"output_power":2.5800039977796363e-05,"reference_active":true,"delta_db":-1.1427660050975874},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.145230698082003e-06,"output_power":6.7119215136731986e-06,"reference_active":true,"delta_db":-0.2716938602318427},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6230912610162062e-06,"output_power":1.4917938901592883e-06,"reference_active":true,"delta_db":-0.36634115336518935}],"active_band_count":3,"active_band_mae_db":0.5936003395648731}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":3.059975136028155,"rms_envelope_nrmse":0.2379785744522481,"envelope_correlation":0.9648506166281684,"rms_gain_error_db":-0.8732495026283705,"reference_mean_power":4.7645706194829206e-05,"output_mean_power":3.896716568370353e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.96519383630983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1852605209893146e-05,"output_power":2.4665146887580663e-05,"reference_active":true,"delta_db":-1.1106125272323026},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5945787260643978e-06,"output_power":3.3718690489916717e-06,"reference_active":true,"delta_db":-0.27777299801845784},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8710953646003796e-07,"output_power":1.7903876001679078e-07,"reference_active":true,"delta_db":-0.19148861496755343}],"active_band_count":3,"active_band_mae_db":0.5266247134061046}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"golf","status":"measured","mos":{"status":"measured","value":2.5084753036499023,"predicted_mos":2.5084753036499023,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"e5dcb7d8b0a2fe0bdc99240d2e4be60e515866b271c1aaa6d51b2677b39a71c5","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2330911720000586},"started_at":"2026-10-13T05:02:25.568183+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028737779994116863,"native_input_samples":144000,"native_input_sha256":"bb121b95e88c63da74de3c75ed8426ef74dcd4719c604859c64577785ab2b82f","output_sha256":"09b446f8d04396e550151676f89d386696b0fc2a8c14ea17c68423766120825e","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5603361949997634,"synthesis_s":0.11425786000017979,"total_s":0.6745940549999432,"rtf":0.11243234249999053,"output_peak":0.6900485157966614,"output_rms":0.059406508817207285,"finished_at":"2026-10-13T05:02:26.255100+00:00","analysis_stages":{"external_f0_s":0.08695206299944402},"evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"09b446f8d04396e550151676f89d386696b0fc2a8c14ea17c68423766120825e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.519223330248273,"output_peak":0.6901259829533756},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.203207695720592,"spectral_convergence":0.4444812437728274,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.418202413491925,"spectral_convergence":0.26388618535967395,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.569522154585835,"spectral_convergence":0.2805383743379213,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047330699114450815,"output_power":0.00038662472066677215,"delta_db":-0.8785330017502252,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016473967181183415,"output_power":0.001435191668208889,"delta_db":-0.5988829198791337,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018140005685480464,"output_power":0.0012491244191283704,"delta_db":-1.6203172037962497,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00043638160470564315,"output_power":0.0003357205793940471,"delta_db":-1.1388847103188646,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011122580793845085,"output_power":8.789444237505007e-05,"delta_db":-1.0224415391541046,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.182955593571425e-05,"output_power":3.7907539110460275e-05,"delta_db":-2.124705356544255,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2306274552404723,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0937415519405331,"pitch":null,"output_rms_dbfs":-24.519223330248273,"output_peak":0.6901259829533756,"reference_rms_dbfs":-23.42548177830774,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.730310754599451},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9143675183479222,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06741024585618054,"rms_dbfs":-23.42548177830774,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.4127808233278,"mean":-1.4176790622717551e-05,"zero_energy":false},"output":{"rms":0.0594345301056851,"rms_dbfs":-24.519223330248273,"peak":0.6901259829533756,"peak_dbfs":-3.2214324256273623,"crest_factor_db":21.29779090462091,"mean":0.00011446916380976263,"zero_energy":false},"rms_difference_db":-1.093741551940532},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":2.099256173811472e-06,"output_rms_dbfs":-56.77934560901069,"output_zero_energy":false,"output_peak":0.009136770522458298,"output_energy_fraction":0.00014385632184976807,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":266,"both_estimated_voiced_frames":249,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9764705882352941,"absolute_difference_cents_median":4.449789418916165,"absolute_difference_cents_p95":37.26713496247157,"difference_over_600_cents_frames":7,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00027737662806153396,"log_rms_mae_db":2.187472580225539,"rms_envelope_nrmse":0.1584683254908,"envelope_correlation":0.988568736024284,"rms_gain_error_db":-1.0937267378310298,"reference_mean_power":0.004547892376945262,"output_mean_power":0.0035353914317368244},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.847839858424573e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106811148990416,"output_power":0.0005196130977844094,"reference_active":true,"delta_db":-1.359946852696459},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001827754588332032,"output_power":0.0001527332450297489,"reference_active":true,"delta_db":-0.7798430377167793},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9268201808042884e-05,"output_power":2.328246080275015e-05,"reference_active":true,"delta_db":-0.9936716041846567}],"active_band_count":3,"active_band_mae_db":1.0444871648659653}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002314838901045675,"log_rms_mae_db":1.6714367338706273,"rms_envelope_nrmse":0.14063971538814732,"envelope_correlation":0.9937039057070024,"rms_gain_error_db":-1.093758078104665,"reference_mean_power":0.004570563264020534,"output_mean_power":0.0035529894411910966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.721146827957421e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818080641138064,"output_power":0.0005041087559414767,"reference_active":true,"delta_db":-1.311378927292416},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.490730075701568e-05,"output_power":6.343055356618015e-05,"reference_active":true,"delta_db":-0.7222564623755318},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0518232169937487e-06,"output_power":1.4043569612337545e-06,"reference_active":true,"delta_db":-1.6466242833187092}],"active_band_count":3,"active_band_mae_db":1.2267532243288857}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002609159645197948,"log_rms_mae_db":1.631139326719882,"rms_envelope_nrmse":0.197084282843549,"envelope_correlation":0.9841449942671261,"rms_gain_error_db":-1.3095823059196339,"reference_mean_power":0.0022103503922765158,"output_mean_power":0.0016349440473145086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0863498364294629e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004828812536856572,"output_power":0.00033736598519599706,"reference_active":true,"delta_db":-1.5573905276308366},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001604104550635534,"output_power":0.0001299899301509685,"reference_active":true,"delta_db":-0.9132296046909759},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2018115158985312e-05,"output_power":1.4588748257116051e-05,"reference_active":true,"delta_db":-1.7876210865354933}],"active_band_count":3,"active_band_mae_db":1.4194137396191018}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001981315811908706,"log_rms_mae_db":1.2956351558168215,"rms_envelope_nrmse":0.17364570525272266,"envelope_correlation":0.9908485471304145,"rms_gain_error_db":-1.3096395380885875,"reference_mean_power":0.002221446667632762,"output_mean_power":0.0016431300468269924},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0262886406950672e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004918125144155888,"output_power":0.00034762136118407944,"reference_active":true,"delta_db":-1.5069311957683384},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.87961976579859e-05,"output_power":5.981406330919615e-05,"reference_active":true,"delta_db":-0.6076112953923531},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6154241946577913e-06,"output_power":1.0343853300088427e-06,"reference_active":true,"delta_db":-1.9360423053376605}],"active_band_count":3,"active_band_mae_db":1.3501949321661175}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014069890621343104,"log_rms_mae_db":1.5559142434923892,"rms_envelope_nrmse":0.20516484674198632,"envelope_correlation":0.9809118860211776,"rms_gain_error_db":-1.1743954951178257,"reference_mean_power":0.000504436783870768,"output_mean_power":0.00038491709421240695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.686855433576991e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002058598732896101,"output_power":0.00014092013708077436,"reference_active":true,"delta_db":-1.6459864415241208},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.647795697069528e-05,"output_power":4.735899579662608e-05,"reference_active":true,"delta_db":-1.4727517905271967},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.854565497212275e-06,"output_power":7.3671225007514885e-06,"reference_active":true,"delta_db":0.3131796079240027}],"active_band_count":3,"active_band_mae_db":1.1439726133251067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012225910578362226,"log_rms_mae_db":1.2337929693206804,"rms_envelope_nrmse":0.16652115408117504,"envelope_correlation":0.9910206754155252,"rms_gain_error_db":-1.1745042487626924,"reference_mean_power":0.000506964577074292,"output_mean_power":0.00038683627285791177},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.5196352298078345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021587658113612985,"output_power":0.0001501714046264372,"reference_active":true,"delta_db":-1.5761828851997481},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9003195542094344e-05,"output_power":2.0906690766326127e-05,"reference_active":true,"delta_db":-1.421605550731285},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.477599488643678e-07,"output_power":8.369012144954114e-07,"reference_active":true,"delta_db":-0.055986970323375594}],"active_band_count":3,"active_band_mae_db":1.0179251354181365}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":2.2513945834399824,"rms_envelope_nrmse":0.21373678883782646,"envelope_correlation":0.9756646767628887,"rms_gain_error_db":-0.8747157760864627,"reference_mean_power":4.749085483762607e-05,"output_mean_power":3.882740889703954e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.26634324803031e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.353777789508225e-05,"output_power":3.039749587700389e-05,"reference_active":true,"delta_db":-0.4269647604522685},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.127573832130495e-06,"output_power":6.6666538109977735e-06,"reference_active":true,"delta_db":-0.29033821243081903},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.62141974758968e-06,"output_power":2.520857061226675e-06,"reference_active":true,"delta_db":1.91652762741827}],"active_band_count":3,"active_band_mae_db":0.8779435334337858}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":2.074473993418771,"rms_envelope_nrmse":0.14474027783507545,"envelope_correlation":0.992272030312562,"rms_gain_error_db":-0.8748200150824513,"reference_mean_power":4.7717923718620126e-05,"output_mean_power":3.901211871083803e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.959457305469273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.182281088277951e-05,"output_power":2.9778162517758476e-05,"reference_active":true,"delta_db":-0.2884064208047949},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5784320176095223e-06,"output_power":3.5618223608107356e-06,"reference_active":true,"delta_db":-0.020205149348943595},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.869028351048552e-07,"output_power":2.3452676227189684e-07,"reference_active":true,"delta_db":0.985765188201947}],"active_band_count":3,"active_band_mae_db":0.4314589194518952}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.596487045288086,"predicted_mos":2.596487045288086,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"7d81c1ebd7d2aff7869972c19db88590eb16794b114a4de37b970d57db2e63c2","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6078976079988934},"started_at":"2026-10-13T07:36:24.301005+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005849108001712011,"native_input_samples":144000,"native_input_sha256":"bb121b95e88c63da74de3c75ed8426ef74dcd4719c604859c64577785ab2b82f","output_sha256":"42cd2433972a473594ae41c2764141986cfc43b48ceae842e54d72f790448a38","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3823962019996543,"synthesis_s":4.084832792999805,"total_s":6.467228994999459,"rtf":1.0778714991665765,"output_peak":0.7368617653846741,"output_rms":0.06503292802916016,"finished_at":"2026-10-13T07:36:30.787594+00:00","analysis_stages":{"external_f0_s":1.6313314649996755},"evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"42cd2433972a473594ae41c2764141986cfc43b48ceae842e54d72f790448a38","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.73479977127083,"output_peak":0.7427202842119206},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.271732106199701,"spectral_convergence":0.4526442598366645,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.527314889069032,"spectral_convergence":0.2846100214895064,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.376292698538077,"spectral_convergence":0.2814754666351329,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.00045285679875236326,"delta_db":-0.19088971577533648,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.0016129599252370551,"delta_db":-0.08661511081744519,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0016284498684124428,"delta_db":-0.462501853277689,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.0003875069550171717,"delta_db":-0.508801279444404,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":0.00010539299975383935,"delta_db":-0.2275118957525468,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":4.618641593447276e-05,"delta_db":-1.2601545711455095,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4560790710354885,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.30234346588300687,"pitch":null,"output_rms_dbfs":-23.733155199646262,"output_peak":0.7427202842119206,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.725113231268936},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9159497537934901,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06736889333987882,"rms_dbfs":-23.430811733763257,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.418110778783316,"mean":-4.31219736735026e-06,"zero_energy":false},"output":{"rms":0.06506422183587034,"rms_dbfs":-23.733155199646262,"peak":0.7427202842119206,"peak_dbfs":-2.5834943007102282,"crest_factor_db":21.14966089893603,"mean":6.683984480741562e-05,"zero_energy":false},"rms_difference_db":-0.30234346588300554},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":2.9443167250286402e-06,"output_rms_dbfs":-55.3101547404459,"output_zero_energy":false,"output_peak":0.0668427954731853,"output_energy_fraction":0.0001680802935048214,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":268,"both_estimated_voiced_frames":246,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9647058823529412,"absolute_difference_cents_median":6.019552574802974,"absolute_difference_cents_p95":38.59975279359164,"difference_over_600_cents_frames":5,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027737662806153396,"log_rms_mae_db":3.6668252924508735,"rms_envelope_nrmse":0.14238856463930175,"envelope_correlation":0.9820805868914755,"rms_gain_error_db":-0.3015779752149599,"reference_mean_power":0.004541375985905238,"output_mean_power":0.004236719011793431},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.847489317868867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100074703458764,"output_power":0.0006544585680992208,"reference_active":true,"delta_db":-0.35380760358833896},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018446902300009587,"output_power":0.0001583828517521384,"reference_active":true,"delta_db":-0.6621528926353647},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9341715429184454e-05,"output_power":2.8617925616513135e-05,"reference_active":true,"delta_db":-0.10847350240500775}],"active_band_count":3,"active_band_mae_db":0.3748113328762372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002314838901045675,"log_rms_mae_db":3.199711619910863,"rms_envelope_nrmse":0.09265454203244473,"envelope_correlation":0.9926461408174321,"rms_gain_error_db":-0.30111079767368604,"reference_mean_power":0.0045632674272876395,"output_mean_power":0.0042575998439912015},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7204098415154257e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818437378247786,"output_power":0.0006263436390386293,"reference_active":true,"delta_db":-0.368721850012783},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591872232322897e-05,"output_power":6.261840672213957e-05,"reference_active":true,"delta_db":-0.8364687737887431},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0593810422650423e-06,"output_power":2.330795360837839e-06,"reference_active":true,"delta_db":0.5376743450445444}],"active_band_count":3,"active_band_mae_db":0.5809549896153569}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002609159645197948,"log_rms_mae_db":1.9801727174275523,"rms_envelope_nrmse":0.19605977365667254,"envelope_correlation":0.9727770505618063,"rms_gain_error_db":-0.3693481465240867,"reference_mean_power":0.0022069594688974205,"output_mean_power":0.0020270270428129817},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0859718062152696e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048225083122535774,"output_power":0.00043589270855891007,"reference_active":true,"delta_db":-0.4389338051073047},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016124782840735576,"output_power":0.00014525340694070731,"reference_active":true,"delta_db":-0.4536754689455452},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2038625150435638e-05,"output_power":2.0225427042230964e-05,"reference_active":true,"delta_db":-0.3728679802613782}],"active_band_count":3,"active_band_mae_db":0.42182575143807605}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001981315811908706,"log_rms_mae_db":1.5017034094398198,"rms_envelope_nrmse":0.12299132881977215,"envelope_correlation":0.9894774346946925,"rms_gain_error_db":-0.3694614882360045,"reference_mean_power":0.002217812362061406,"output_mean_power":0.0020369419438257927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0257217759681142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004916402889453351,"output_power":0.00044448285857130683,"reference_active":true,"delta_db":-0.4379244812905356},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.925750439584526e-05,"output_power":6.157591097667225e-05,"reference_active":true,"delta_db":-0.510559925872449},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6222165494821603e-06,"output_power":1.4808062910164415e-06,"reference_active":true,"delta_db":-0.3961057686702852}],"active_band_count":3,"active_band_mae_db":0.4481967252777566}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014069890621343104,"log_rms_mae_db":1.8996866794077376,"rms_envelope_nrmse":0.22786849656002042,"envelope_correlation":0.9667161629358556,"rms_gain_error_db":-0.45359147665675625,"reference_mean_power":0.000503617873892436,"output_mean_power":0.00045367201204917553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6929874202041835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020569637799863847,"output_power":0.00017450509731726312,"reference_active":true,"delta_db":-0.7141852721419827},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.720458225580812e-05,"output_power":5.454820054254181e-05,"reference_active":true,"delta_db":-0.9061845737610457},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.876518957296052e-06,"output_power":8.185715917245876e-06,"reference_active":true,"delta_db":0.7568802391621297}],"active_band_count":3,"active_band_mae_db":0.7924166950217192}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012225910578362226,"log_rms_mae_db":1.3041775822146784,"rms_envelope_nrmse":0.1489154191883957,"envelope_correlation":0.9861960974091168,"rms_gain_error_db":-0.45362591170933936,"reference_mean_power":0.0005061264034494053,"output_mean_power":0.0004559281453308392},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5254708519662443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021591129559099358,"output_power":0.00018391570516444633,"reference_active":true,"delta_db":-0.6965654677984663},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.951038057605059e-05,"output_power":2.383601510324048e-05,"reference_active":true,"delta_db":-0.9274115841224174},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.54700119986915e-07,"output_power":1.0424894642954602e-06,"reference_active":true,"delta_db":0.8625790963296387}],"active_band_count":3,"active_band_mae_db":0.8288520494168408}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":2.2617669909641855,"rms_envelope_nrmse":0.2751288556458867,"envelope_correlation":0.9521325236833158,"rms_gain_error_db":-0.6489845476258559,"reference_mean_power":4.741724973734132e-05,"output_mean_power":4.083550266982176e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.272212516285401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.356579821454211e-05,"output_power":2.9235600941696182e-05,"reference_active":true,"delta_db":-0.5998495417714798},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.145230698082003e-06,"output_power":6.209303737486304e-06,"reference_active":true,"delta_db":-0.6097335088320097},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6230912610162062e-06,"output_power":1.5121184939719332e-06,"reference_active":true,"delta_db":-0.3075711437666194}],"active_band_count":3,"active_band_mae_db":0.5057180647900363}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":1.483259612517046,"rms_envelope_nrmse":0.156416307668017,"envelope_correlation":0.9861704025882083,"rms_gain_error_db":-0.649055996490872,"reference_mean_power":4.7645706194829206e-05,"output_mean_power":4.103157320429892e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.96519383630983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1852605209893146e-05,"output_power":2.8018549933756277e-05,"reference_active":true,"delta_db":-0.5569930372500854},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5945787260643978e-06,"output_power":3.089480523147571e-06,"reference_active":true,"delta_db":-0.6576253802066307},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8710953646003796e-07,"output_power":1.6674165931662534e-07,"reference_active":true,"delta_db":-0.5005180383285145}],"active_band_count":3,"active_band_mae_db":0.5717121519284102}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.86484432220459,"predicted_mos":2.86484432220459,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"c0ab6064f3acf22bd97a89de1fc99ee801775161ab7a582c5055a536710cff1e","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2254694949988334},"started_at":"2026-10-13T05:10:33.647009+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0033392580007785,"native_input_samples":144000,"native_input_sha256":"bb121b95e88c63da74de3c75ed8426ef74dcd4719c604859c64577785ab2b82f","output_sha256":"b0f6c5fa438eed59533e3b1ede0ee93cf94d128dd3a77288791bcc222ff77c6d","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7775995450010669,"synthesis_s":0.12718755499918188,"total_s":0.9047871000002488,"rtf":0.15079785000004145,"output_peak":0.643333375453949,"output_rms":0.06618861573999438,"finished_at":"2026-10-13T05:10:34.563008+00:00","analysis_stages":{"external_f0_s":0.16028782300054445},"evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"b0f6c5fa438eed59533e3b1ede0ee93cf94d128dd3a77288791bcc222ff77c6d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.581170674926604,"output_peak":0.6431557882820615},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.945599867457083,"spectral_convergence":0.18993064149090824,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.603906752888125,"spectral_convergence":0.17142520679476073,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.078627342498957,"spectral_convergence":0.20767946510331137,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.00046190063650959314,"delta_db":-0.1050130306637105,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.001625574923215526,"delta_db":-0.052780972990814363,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0017248141010820944,"delta_db":-0.2128228468171227,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.0004232785331106559,"delta_db":-0.1253338616050177,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":0.00010510946457828045,"delta_db":-0.2392113157122555,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":5.00181720082497e-05,"delta_db":-0.9140190370170985,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2748635108010032,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1438547462724426,"pitch":null,"output_rms_dbfs":-23.574666480035695,"output_peak":0.6431557882820615,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.876044654281388},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9548306645419404,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06736889333987882,"rms_dbfs":-23.430811733763257,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.418110778783316,"mean":-4.31219736735026e-06,"zero_energy":false},"output":{"rms":0.06626232587605418,"rms_dbfs":-23.574666480035695,"peak":0.6431557882820615,"peak_dbfs":-3.8336763487827086,"crest_factor_db":19.74099013125299,"mean":6.1122179473387e-05,"zero_energy":false},"rms_difference_db":-0.14385474627243866},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":2.4086088586859877e-06,"output_rms_dbfs":-56.182337206768196,"output_zero_energy":false,"output_peak":0.03345160028583195,"output_energy_fraction":0.0001325713501123074,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":238,"both_estimated_voiced_frames":230,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9019607843137255,"absolute_difference_cents_median":0.9071875586555842,"absolute_difference_cents_p95":13.722346589196896,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027737662806153396,"log_rms_mae_db":1.0305351988696754,"rms_envelope_nrmse":0.060331144212767855,"envelope_correlation":0.9968881507563523,"rms_gain_error_db":-0.14344869944523023,"reference_mean_power":0.004541375985905238,"output_mean_power":0.004393823316734088},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.847489317868867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100074703458764,"output_power":0.0006982661502978283,"reference_active":true,"delta_db":-0.07241928825397624},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018446902300009587,"output_power":0.00016224765700547766,"reference_active":true,"delta_db":-0.5574501368946986},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9341715429184454e-05,"output_power":3.169072358419261e-05,"reference_active":true,"delta_db":0.33446654640947515}],"active_band_count":3,"active_band_mae_db":0.3214453238527167}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002314838901045675,"log_rms_mae_db":0.6416898626774239,"rms_envelope_nrmse":0.0336034680013133,"envelope_correlation":0.999174555706629,"rms_gain_error_db":-0.14319043943943938,"reference_mean_power":0.0045632674272876395,"output_mean_power":0.004415266041866728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7204098415154257e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818437378247786,"output_power":0.0006695141997892773,"reference_active":true,"delta_db":-0.07925063778298524},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591872232322897e-05,"output_power":6.783014805949823e-05,"reference_active":true,"delta_db":-0.48926125463591386},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0593810422650423e-06,"output_power":2.1441387037970896e-06,"reference_active":true,"delta_db":0.17516165826375357}],"active_band_count":3,"active_band_mae_db":0.24789118356088422}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002609159645197948,"log_rms_mae_db":0.8798322868828671,"rms_envelope_nrmse":0.07670631721462776,"envelope_correlation":0.9960249355650204,"rms_gain_error_db":-0.1731844820393744,"reference_mean_power":0.0022069594688974205,"output_mean_power":0.0021206837529127724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0859718062152696e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048225083122535774,"output_power":0.00046386816556688213,"reference_active":true,"delta_db":-0.16878416185853085},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016124782840735576,"output_power":0.00014339794118952358,"reference_active":true,"delta_db":-0.5095095844874123},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2038625150435638e-05,"output_power":1.991336451038267e-05,"reference_active":true,"delta_db":-0.44039854660631816}],"active_band_count":3,"active_band_mae_db":0.3728974309840871}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001981315811908706,"log_rms_mae_db":0.4962259798210213,"rms_envelope_nrmse":0.042018454570700414,"envelope_correlation":0.9989461702365786,"rms_gain_error_db":-0.1733090010877208,"reference_mean_power":0.002217812362061406,"output_mean_power":0.0021310512771358556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0257217759681142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004916402889453351,"output_power":0.0004695151961416909,"reference_active":true,"delta_db":-0.19997812211377908},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.925750439584526e-05,"output_power":6.330309047757868e-05,"reference_active":true,"delta_db":-0.39041925324639326},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6222165494821603e-06,"output_power":1.2714262156077463e-06,"reference_active":true,"delta_db":-1.0581766573164733}],"active_band_count":3,"active_band_mae_db":0.5495246775588819}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014069890621343104,"log_rms_mae_db":0.9122000591625025,"rms_envelope_nrmse":0.11948523587347044,"envelope_correlation":0.9909725112128471,"rms_gain_error_db":-0.16424981902307897,"reference_mean_power":0.000503617873892436,"output_mean_power":0.0004849267629460003},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6929874202041835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020569637799863847,"output_power":0.00020846765229475981,"reference_active":true,"delta_db":0.058120310272460264},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.720458225580812e-05,"output_power":6.199939387583538e-05,"reference_active":true,"delta_db":-0.3501144213043469},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.876518957296052e-06,"output_power":7.933327187630164e-06,"reference_active":true,"delta_db":0.6208672130437762}],"active_band_count":3,"active_band_mae_db":0.34303398154019443}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012225910578362226,"log_rms_mae_db":0.6447065951269068,"rms_envelope_nrmse":0.06708822359622432,"envelope_correlation":0.9972224715471029,"rms_gain_error_db":-0.1642380566310905,"reference_mean_power":0.0005061264034494053,"output_mean_power":0.00048734351166294546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5254708519662443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021591129559099358,"output_power":0.00021936240815102672,"reference_active":true,"delta_db":0.06886841707924816},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.951038057605059e-05,"output_power":2.749725269064346e-05,"reference_active":true,"delta_db":-0.30685505640807437},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.54700119986915e-07,"output_power":8.034848043956404e-07,"reference_active":true,"delta_db":-0.26836097181615354}],"active_band_count":3,"active_band_mae_db":0.2146948151011587}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":1.2226183489126412,"rms_envelope_nrmse":0.12174883887064579,"envelope_correlation":0.9910740477055546,"rms_gain_error_db":-0.3006169508571474,"reference_mean_power":4.741724973734132e-05,"output_mean_power":4.4246066287007285e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.272212516285401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.356579821454211e-05,"output_power":3.100526441830162e-05,"reference_active":true,"delta_db":-0.34461539985653167},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.145230698082003e-06,"output_power":6.431350679634597e-06,"reference_active":true,"delta_db":-0.4571406459398761},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6230912610162062e-06,"output_power":1.2155980571056037e-06,"reference_active":true,"delta_db":-1.255529421709836}],"active_band_count":3,"active_band_mae_db":0.6857618225020813}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":0.9456314768078604,"rms_envelope_nrmse":0.0637230950358096,"envelope_correlation":0.9979991288677554,"rms_gain_error_db":-0.3006824565091625,"reference_mean_power":4.7645706194829206e-05,"output_mean_power":4.445857338860315e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.96519383630983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1852605209893146e-05,"output_power":2.941363529738112e-05,"reference_active":true,"delta_db":-0.34596255612821764},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5945787260643978e-06,"output_power":3.1767861104920622e-06,"reference_active":true,"delta_db":-0.5366002443702247},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8710953646003796e-07,"output_power":1.404962409287815e-07,"reference_active":true,"delta_db":-1.2443121830684605}],"active_band_count":3,"active_band_mae_db":0.7089583278556342}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3575-170457-0054","method":"magphase","status":"measured","mos":{"status":"measured","value":2.893986463546753,"predicted_mos":2.893986463546753,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288731,"input_sample_rate":48000,"input_seconds":6.015229166666667,"input_float32_sha256":"9ea47aafd1ba6fc6310973dc947e5a2f5020d8f84cf01fa29ed76a14024cf382","internal_samples":96244,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.33871515699866},"started_at":"2026-10-13T05:14:49.765516+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010165876001337892,"native_input_samples":288000,"native_input_sha256":"8b01d71753837bc63b48e0f366d0c9d266d8e5ecd003cc2e8783fae714da77bd","output_sha256":"8042ec4f4996e105871a362de2cdc0eb4c4a1a5b7fada7cbf3a8ef930fb26bd8","output_samples":288731,"length_delta_samples":731,"analysis_s":3.451208452001083,"synthesis_s":1.0125871210002515,"total_s":4.463795573001335,"rtf":0.7439659288335557,"output_peak":0.6549965387742904,"output_rms":0.06498671901347501,"finished_at":"2026-10-13T05:14:54.253668+00:00","evaluation_input":{"reference":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"output_sha256":"8042ec4f4996e105871a362de2cdc0eb4c4a1a5b7fada7cbf3a8ef930fb26bd8","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96244,"length_delta_samples":244,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.740723411193546,"output_peak":0.6442716820285486},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9974647770250613,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.405693515056144,"spectral_convergence":0.3796187486149102,"spectral_convergence_status":"ok","amplitude_floor":1.8779397470586322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.732380764078267,"spectral_convergence":0.22767335078389445,"spectral_convergence_status":"ok","amplitude_floor":1.7710051963816244e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.738674830779968,"spectral_convergence":0.20915108463163706,"spectral_convergence_status":"ok","amplitude_floor":1.1208016909489659e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004732055872254276,"output_power":0.0004107580917852706,"delta_db":-0.6146373614603335,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016454515092618504,"output_power":0.0015250272606117873,"delta_db":-0.3300748137055305,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001811442607675158,"output_power":0.0016790415814177868,"delta_db":-0.3296312718134548,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004356719801952226,"output_power":0.00044516435443413906,"delta_db":0.09360752248833414,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011106135183657726,"output_power":0.00011995050561816867,"delta_db":0.3343912756337476,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.173475364573245e-05,"output_power":5.616088116509209e-05,"delta_db":-0.4109580615779833,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3522167177798973,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.299523485924938,"pitch":null,"output_rms_dbfs":-23.730335219688193,"output_peak":0.6442716820285486,"reference_rms_dbfs":-23.430811733763257,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.958916369971459},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9271361534021205,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06736889333987882,"rms_dbfs":-23.430811733763257,"peak":0.630035400390625,"peak_dbfs":-4.012700954979939,"crest_factor_db":19.418110778783316,"mean":-4.31219736735026e-06,"zero_energy":false},"output":{"rms":0.06508534915810352,"rms_dbfs":-23.730335219688193,"peak":0.6442716820285486,"peak_dbfs":-3.818619139610817,"crest_factor_db":19.911716080077376,"mean":-0.0009403804458223005,"zero_energy":false},"rms_difference_db":-0.2995234859249365},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.532477545901202e-06,"threshold_dbfs":-51.23062154272028,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":145,"quiet_samples":23200,"reference_mean_power":2.127250833115701e-06,"output_mean_power":1.6903277775409068e-06,"output_rms_dbfs":-57.72029071614479,"output_zero_energy":false,"output_peak":0.018625319634932025,"output_energy_fraction":9.643200623537207e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":248,"both_estimated_voiced_frames":238,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9333333333333333,"absolute_difference_cents_median":4.471597906020805,"absolute_difference_cents_p95":20.181982558816063,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027737662806153396,"log_rms_mae_db":1.7506657222990707,"rms_envelope_nrmse":0.11748139577711812,"envelope_correlation":0.9881884263446317,"rms_gain_error_db":-0.29859899034108306,"reference_mean_power":0.004541375985905238,"output_mean_power":0.00423962612943176},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.847489317868867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100074703458764,"output_power":0.0006822072698219164,"reference_active":true,"delta_db":-0.17346575095014247},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018446902300009587,"output_power":0.0001804637135378486,"reference_active":true,"delta_db":-0.09533557696027489},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9341715429184454e-05,"output_power":2.8199142733335834e-05,"reference_active":true,"delta_db":-0.1724959499825423}],"active_band_count":3,"active_band_mae_db":0.14709909263098656}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002314838901045675,"log_rms_mae_db":1.1784112714174106,"rms_envelope_nrmse":0.07823381338298972,"envelope_correlation":0.9950953246589974,"rms_gain_error_db":-0.29787296572067334,"reference_mean_power":0.0045632674272876395,"output_mean_power":0.004260775231531112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7204098415154257e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006818437378247786,"output_power":0.0006579503832070151,"reference_active":true,"delta_db":-0.15491712012475467},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.591872232322897e-05,"output_power":7.402257273609145e-05,"reference_active":true,"delta_db":-0.10984715101950124},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0593810422650423e-06,"output_power":2.024588867719384e-06,"reference_active":true,"delta_db":-0.07399865970676504}],"active_band_count":3,"active_band_mae_db":0.11292097695034033}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002609159645197948,"log_rms_mae_db":1.6429731003026682,"rms_envelope_nrmse":0.13437330782889742,"envelope_correlation":0.9874554754215524,"rms_gain_error_db":-0.2692699195008446,"reference_mean_power":0.0022069594688974205,"output_mean_power":0.002074279930575752},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0859718062152696e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048225083122535774,"output_power":0.0004508701881699141,"reference_active":true,"delta_db":-0.2922146446031195},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016124782840735576,"output_power":0.0001496994642994753,"reference_active":true,"delta_db":-0.3227362829922764},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2038625150435638e-05,"output_power":1.9485295220363872e-05,"reference_active":true,"delta_db":-0.5347750801171459}],"active_band_count":3,"active_band_mae_db":0.38324200257084734}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001981315811908706,"log_rms_mae_db":1.2197448369367585,"rms_envelope_nrmse":0.08889666025262702,"envelope_correlation":0.9945864501841308,"rms_gain_error_db":-0.26884550622933134,"reference_mean_power":0.002217812362061406,"output_mean_power":0.002084684077203329},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0257217759681142e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004916402889453351,"output_power":0.0004592381347781448,"reference_active":true,"delta_db":-0.2960952087915532},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.925750439584526e-05,"output_power":6.48058073952732e-05,"reference_active":true,"delta_db":-0.2885291253094265},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6222165494821603e-06,"output_power":1.4413846222213869e-06,"reference_active":true,"delta_db":-0.5132894329349638}],"active_band_count":3,"active_band_mae_db":0.3659712556786478}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014069890621343104,"log_rms_mae_db":1.5031592276681718,"rms_envelope_nrmse":0.19271889145141205,"envelope_correlation":0.9768012079276611,"rms_gain_error_db":0.11875590084031232,"reference_mean_power":0.000503617873892436,"output_mean_power":0.0005175790931567929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6929874202041835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020569637799863847,"output_power":0.00020923405910723648,"reference_active":true,"delta_db":0.07405735884000039},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.720458225580812e-05,"output_power":6.778071271049973e-05,"reference_active":true,"delta_db":0.03707245248304446},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.876518957296052e-06,"output_power":6.240143550985455e-06,"reference_active":true,"delta_db":-0.4217406410097182}],"active_band_count":3,"active_band_mae_db":0.17762348411092102}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012225910578362226,"log_rms_mae_db":0.9932995042120341,"rms_envelope_nrmse":0.12493984346985795,"envelope_correlation":0.9898817620818556,"rms_gain_error_db":0.1188662493025095,"reference_mean_power":0.0005061264034494053,"output_mean_power":0.0005201703804652783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5254708519662443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021591129559099358,"output_power":0.00021968041905698422,"reference_active":true,"delta_db":0.0751598491326507},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.951038057605059e-05,"output_power":2.8978840290499572e-05,"reference_active":true,"delta_db":-0.07893808974816217},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.54700119986915e-07,"output_power":8.324735208876532e-07,"reference_active":true,"delta_db":-0.11443336461166541}],"active_band_count":3,"active_band_mae_db":0.08951043449749276}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.38440402041993e-05,"log_rms_mae_db":1.9743184166677004,"rms_envelope_nrmse":0.24176307892227877,"envelope_correlation":0.9666560800505506,"rms_gain_error_db":0.4545584264189201,"reference_mean_power":4.741724973734132e-05,"output_mean_power":5.264925174798076e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.272212516285401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.356579821454211e-05,"output_power":3.5887265268454044e-05,"reference_active":true,"delta_db":0.29043385455548154},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.145230698082003e-06,"output_power":8.499931713268774e-06,"reference_active":true,"delta_db":0.7539918127391168},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6230912610162062e-06,"output_power":2.9882465310632094e-06,"reference_active":true,"delta_db":2.650734845763098}],"active_band_count":3,"active_band_mae_db":1.231720171019232}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.68596247913149e-05,"log_rms_mae_db":1.423785775307691,"rms_envelope_nrmse":0.1407991493924858,"envelope_correlation":0.9893080705892834,"rms_gain_error_db":0.4541406200218069,"reference_mean_power":4.7645706194829206e-05,"output_mean_power":5.2897826807791894e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.96519383630983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.1852605209893146e-05,"output_power":3.481723033479426e-05,"reference_active":true,"delta_db":0.38649261726964124},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5945787260643978e-06,"output_power":4.5078871285226e-06,"reference_active":true,"delta_db":0.9832503367636529},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8710953646003796e-07,"output_power":3.1713031626557936e-07,"reference_active":true,"delta_db":2.2914183779386055}],"active_band_count":3,"active_band_mae_db":1.2203871106572999}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"hnm","status":"measured","mos":{"status":"measured","value":3.206979751586914,"predicted_mos":3.206979751586914,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"5430892f0b5c7024553716f792d20b94ec0de08cb996349237f6a67cdb361c91","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.088874766999652},"started_at":"2026-10-13T04:39:02.636024+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4891999853716698e-05,"native_input_samples":96000,"native_input_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","output_sha256":"85cfc6493da07fdec9bd5e9658825fbd158527b72011934cc67268f53054991b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1966997429999537,"synthesis_s":0.4906138550004471,"total_s":1.6873135980004008,"rtf":0.2812189330000668,"output_peak":0.3743638405003378,"output_rms":0.04737817653743918,"finished_at":"2026-10-13T04:39:04.329017+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"85cfc6493da07fdec9bd5e9658825fbd158527b72011934cc67268f53054991b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.332949753891908,"spectral_convergence":0.23573699228656467,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.32261813127996,"spectral_convergence":0.23761654506468333,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9364234361135315,"spectral_convergence":0.24534750569136324,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":9.532766533807288e-05,"delta_db":0.11741476846961364,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0009023069583168293,"delta_db":-0.051094307760623126,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0007050219927189853,"delta_db":-0.2390381695624325,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.000380466826806481,"delta_db":-0.2761142550054493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":0.00011183909985212125,"delta_db":-0.015834594421053096,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":4.972906898026212e-05,"delta_db":-0.030011647228479207,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12158462374127514,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14040633531804167,"pitch":null,"output_rms_dbfs":-26.488433163091326,"output_peak":0.3743638405003378,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.530663773761799},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.332949753891908,"spectral_convergence":0.23573699228656467,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.32261813127996,"spectral_convergence":0.23761654506468333,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9364234361135315,"spectral_convergence":0.24534750569136324,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":9.532766533807288e-05,"delta_db":0.11741476846961364,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0009023069583168293,"delta_db":-0.051094307760623126,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0007050219927189853,"delta_db":-0.2390381695624325,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.000380466826806481,"delta_db":-0.2761142550054493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":0.00011183909985212125,"delta_db":-0.015834594421053096,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":4.972906898026212e-05,"delta_db":-0.030011647228479207,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12158462374127514,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14040633531804167,"pitch":null,"output_rms_dbfs":-26.488433163091326,"output_peak":0.3743638405003378,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.530663773761799},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8831956814106386,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04815026244547317,"rms_dbfs":-26.348026827773285,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.44194825234116,"mean":-2.5399525960286456e-07,"zero_energy":false},"output":{"rms":0.04737817653743918,"rms_dbfs":-26.488433163091326,"peak":0.3743638405003378,"peak_dbfs":-8.534122121197845,"crest_factor_db":17.954311041893483,"mean":-1.0250129974552019e-05,"zero_energy":false},"rms_difference_db":-0.14040633531804048},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":1.1367630493457863e-06,"output_rms_dbfs":-59.44330051658734,"output_zero_energy":false,"output_peak":0.01399161337642477,"output_energy_fraction":9.284418310801061e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":290,"both_estimated_voiced_frames":224,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":66,"paired_voiced_fraction":0.9180327868852459,"absolute_difference_cents_median":1.5884842733329134,"absolute_difference_cents_p95":10.40381841393093,"difference_over_600_cents_frames":4,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001795198121335632,"log_rms_mae_db":0.6531896422686229,"rms_envelope_nrmse":0.06782100705162905,"envelope_correlation":0.9956620940934883,"rms_gain_error_db":-0.139781218299291,"reference_mean_power":0.002319729266015563,"output_mean_power":0.0022462556361100526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2030894540936354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042099620393065603,"output_power":0.00041881802541022395,"reference_active":true,"delta_db":-0.022528149682635234},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010212932714007924,"output_power":0.00010027499231244942,"reference_active":true,"delta_db":-0.07957833290741352},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7753962605616007e-05,"output_power":1.491504069177564e-05,"reference_active":true,"delta_db":-0.7567085825177843}],"active_band_count":3,"active_band_mae_db":0.2862716883692777}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017377061993537507,"log_rms_mae_db":0.3474597777998764,"rms_envelope_nrmse":0.04132474916026361,"envelope_correlation":0.9984279264796041,"rms_gain_error_db":-0.13957052154553,"reference_mean_power":0.002329667623369792,"output_mean_power":0.002255988658256344},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1345793651623817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003814703455166119,"output_power":0.00037925843487281917,"reference_active":true,"delta_db":-0.0252553424808102},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.449258478788483e-05,"output_power":5.3594066154137956e-05,"reference_active":true,"delta_db":-0.0722070061869419},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.451644818904226e-07,"output_power":8.771455672659499e-07,"reference_active":true,"delta_db":-0.3243572004687468}],"active_band_count":3,"active_band_mae_db":0.14060651637883295}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001705289920117241,"log_rms_mae_db":1.213761667813964,"rms_envelope_nrmse":0.08491279322068762,"envelope_correlation":0.9942961460015716,"rms_gain_error_db":-0.15719315226755295,"reference_mean_power":0.0014327712363456494,"output_mean_power":0.0013818392967971688},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.476686730813627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004044339102676846,"output_power":0.00040141464917471556,"reference_active":true,"delta_db":-0.03254344882559642},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628475293024379e-05,"output_power":9.246479570210506e-05,"reference_active":true,"delta_db":-0.17581105885654452},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2048085119397813e-05,"output_power":1.0903919899901594e-05,"reference_active":true,"delta_db":-0.4333537463137398}],"active_band_count":3,"active_band_mae_db":0.21390275133196024}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015393777861692204,"log_rms_mae_db":0.6694512482868974,"rms_envelope_nrmse":0.05114357418313031,"envelope_correlation":0.9979581504557065,"rms_gain_error_db":-0.1564342310917225,"reference_mean_power":0.0014385306131629134,"output_mean_power":0.0013876364057231326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.958921932585308e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037465347734378165,"output_power":0.00036946226639504104,"reference_active":true,"delta_db":-0.06059677851972513},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.015607336171047e-05,"output_power":4.797692680644028e-05,"reference_active":true,"delta_db":-0.19291103432774795},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018660953269709e-07,"output_power":7.434232926435038e-07,"reference_active":true,"delta_db":-0.3286568681742632}],"active_band_count":3,"active_band_mae_db":0.19405489367391207}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":1.2492292819618427,"rms_envelope_nrmse":0.11692761346850603,"envelope_correlation":0.9898595408788325,"rms_gain_error_db":-0.2542202351190152,"reference_mean_power":0.0004617837976357088,"output_mean_power":0.0004355285882030939},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1028770536544704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.713589283902283e-05,"output_power":8.088306610322712e-05,"reference_active":true,"delta_db":-0.3233947949303241},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3909625300628061e-05,"output_power":1.2040478013932375e-05,"reference_active":true,"delta_db":-0.6267170206012123},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.457858491278475e-06,"output_power":3.395700630104035e-06,"reference_active":true,"delta_db":-1.1819688304113385}],"active_band_count":3,"active_band_mae_db":0.710693548647625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":0.7300852372364666,"rms_envelope_nrmse":0.06929764467213337,"envelope_correlation":0.9966707488478089,"rms_gain_error_db":-0.2551723916129291,"reference_mean_power":0.00046390064262223966,"output_mean_power":0.00043742916429670995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.977269818051972e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.049771362210367e-05,"output_power":7.515709955815371e-05,"reference_active":true,"delta_db":-0.29813533689428806},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.625262232859731e-06,"output_power":5.856653894791944e-06,"reference_active":true,"delta_db":-0.5355351296042212},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.83834356187152e-07,"output_power":2.1367974965092773e-07,"reference_active":true,"delta_db":-1.233015964583408}],"active_band_count":3,"active_band_mae_db":0.6888954770273058}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":1.4787549445464827,"rms_envelope_nrmse":0.2078445062316258,"envelope_correlation":0.9690889433258995,"rms_gain_error_db":0.2735957704830994,"reference_mean_power":4.298191756596504e-05,"output_mean_power":4.577679300658404e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5620745474492655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.382029340071455e-06,"output_power":5.820051220761995e-06,"reference_active":true,"delta_db":-0.40031989662067724},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1967382600492444e-06,"output_power":1.3259237951148368e-06,"reference_active":true,"delta_db":0.4451938879259815},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059624559400327e-07,"output_power":5.419680025629998e-07,"reference_active":true,"delta_db":0.2985535486041181}],"active_band_count":3,"active_band_mae_db":0.38135577771692564}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":0.9207573995250128,"rms_envelope_nrmse":0.13895309668989086,"envelope_correlation":0.9857348589261967,"rms_gain_error_db":0.2733342451141797,"reference_mean_power":4.31749976072588e-05,"output_mean_power":4.597965906409649e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4490976093596581e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.797984044318815e-06,"output_power":5.3825042841556554e-06,"reference_active":true,"delta_db":-0.32292631701559854},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942574114894925e-07,"output_power":6.586415861551389e-07,"reference_active":true,"delta_db":0.4467454139246523},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1049309542740296e-08,"output_power":3.2469597247166205e-08,"reference_active":true,"delta_db":1.882390469483792}],"active_band_count":3,"active_band_mae_db":0.8840207334746809}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.6527342796325684,"predicted_mos":3.6527342796325684,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"8da2b695139e6ef60a1e9b89ee484735b2162bb89d00ce4ac753ddbb96639f1d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.586780418001581},"started_at":"2026-10-13T04:42:00.369108+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.780800059554167e-05,"native_input_samples":96000,"native_input_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","output_sha256":"5242b930a664ac78de48c7da511dfeaf0806f1f6bc7d5a2a1c70cb539a31d08b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2902539690003323,"synthesis_s":0.09211632300048223,"total_s":1.3823702920008145,"rtf":0.2303950486668024,"output_peak":0.5491744251651681,"output_rms":0.05953880318850538,"finished_at":"2026-10-13T04:42:01.757977+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"5242b930a664ac78de48c7da511dfeaf0806f1f6bc7d5a2a1c70cb539a31d08b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1952500009592475,"spectral_convergence":0.5469080241964667,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.3100456155771125,"spectral_convergence":0.36936440158334544,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.071120569722514,"spectral_convergence":0.40242777105240857,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":0.00012079646164172847,"delta_db":1.1457673312407932,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0014458400355726816,"delta_db":1.9965650896679021,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0012419455276689278,"delta_db":2.2199606618942243,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.0005538442916601802,"delta_db":1.3545946211230389,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":0.0001288109799627085,"delta_db":0.5977576185391328,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":5.3631788613351225e-05,"delta_db":0.2981078763169042,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.268792199796999,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.8440288374633895,"pitch":null,"output_rms_dbfs":-24.503997990309895,"output_peak":0.5491744251651681,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.858805395419625},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1952500009592475,"spectral_convergence":0.5469080241964667,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.3100456155771125,"spectral_convergence":0.36936440158334544,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.071120569722514,"spectral_convergence":0.40242777105240857,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":0.00012079646164172847,"delta_db":1.1457673312407932,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0014458400355726816,"delta_db":1.9965650896679021,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0012419455276689278,"delta_db":2.2199606618942243,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.0005538442916601802,"delta_db":1.3545946211230389,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":0.0001288109799627085,"delta_db":0.5977576185391328,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":5.3631788613351225e-05,"delta_db":0.2981078763169042,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.268792199796999,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.8440288374633895,"pitch":null,"output_rms_dbfs":-24.503997990309895,"output_peak":0.5491744251651681,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.858805395419625},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9215473373163324,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04815026244547317,"rms_dbfs":-26.348026827773285,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.44194825234116,"mean":-2.5399525960286456e-07,"zero_energy":false},"output":{"rms":0.05953880318850538,"rms_dbfs":-24.503997990309895,"peak":0.5491744251651681,"peak_dbfs":-5.205793917717433,"crest_factor_db":19.298204072592462,"mean":2.478962445931654e-06,"zero_energy":false},"rms_difference_db":1.8440288374633909},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":1.408855730673518e-06,"output_rms_dbfs":-58.5113347713926,"output_zero_energy":false,"output_peak":0.038362016149208915,"output_energy_fraction":7.286311880299864e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":314,"both_estimated_voiced_frames":234,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":80,"paired_voiced_fraction":0.9590163934426229,"absolute_difference_cents_median":6.491276931726231,"absolute_difference_cents_p95":1211.8731269311288,"difference_over_600_cents_frames":16,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001795198121335632,"log_rms_mae_db":1.685635297392041,"rms_envelope_nrmse":0.29848401962895893,"envelope_correlation":0.9758199056573658,"rms_gain_error_db":1.8448481014957896,"reference_mean_power":0.002319729266015563,"output_mean_power":0.003547497612065023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2030894540936354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042099620393065603,"output_power":0.0007297444624143546,"reference_active":true,"delta_db":2.3889262820151993},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010212932714007924,"output_power":0.00020320422124004288,"reference_active":true,"delta_db":2.987822548238017},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7753962605616007e-05,"output_power":2.9899209645455844e-05,"reference_active":true,"delta_db":2.2636440754269165}],"active_band_count":3,"active_band_mae_db":2.546797635226711}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017377061993537507,"log_rms_mae_db":1.435364800684612,"rms_envelope_nrmse":0.27926284085659026,"envelope_correlation":0.9831503094249653,"rms_gain_error_db":1.8455982475893007,"reference_mean_power":0.002329667623369792,"output_mean_power":0.003563311495064397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1345793651623817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003814703455166119,"output_power":0.0006758264826425653,"reference_active":true,"delta_db":2.4837442313422615},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.449258478788483e-05,"output_power":0.00011148278794971526,"reference_active":true,"delta_db":3.108704123568284},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.451644818904226e-07,"output_power":1.4691972772650754e-06,"reference_active":true,"delta_db":1.9157272181528588}],"active_band_count":3,"active_band_mae_db":2.5027251910211348}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001705289920117241,"log_rms_mae_db":1.9574674465676314,"rms_envelope_nrmse":0.3677407605830234,"envelope_correlation":0.9788688306390988,"rms_gain_error_db":2.376056369720077,"reference_mean_power":0.0014327712363456494,"output_mean_power":0.0024761815947854527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.476686730813627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004044339102676846,"output_power":0.0007390290714248407,"reference_active":true,"delta_db":2.6181396030255772},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628475293024379e-05,"output_power":0.00020716936105635733,"reference_active":true,"delta_db":3.3276800628556997},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2048085119397813e-05,"output_power":2.1762253656538187e-05,"reference_active":true,"delta_db":2.567858408981235}],"active_band_count":3,"active_band_mae_db":2.837892691620837}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015393777861692204,"log_rms_mae_db":1.6605071420028878,"rms_envelope_nrmse":0.3510777232901171,"envelope_correlation":0.9852033962744853,"rms_gain_error_db":2.3774234817452826,"reference_mean_power":0.0014385306131629134,"output_mean_power":0.002486917948137523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.958921932585308e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037465347734378165,"output_power":0.0006951142132495348,"reference_active":true,"delta_db":2.6842640092120575},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.015607336171047e-05,"output_power":0.0001106716208137798,"reference_active":true,"delta_db":3.437127414079539},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018660953269709e-07,"output_power":1.3022376790898712e-06,"reference_active":true,"delta_db":2.1058840618490597}],"active_band_count":3,"active_band_mae_db":2.7424251617135518}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":1.3577305533161066,"rms_envelope_nrmse":0.23296486645050385,"envelope_correlation":0.9818587123222142,"rms_gain_error_db":1.3017956230104923,"reference_mean_power":0.0004617837976357088,"output_mean_power":0.0006231868107810385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1028770536544704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.713589283902283e-05,"output_power":0.00011209197361823005,"reference_active":true,"delta_db":1.093774303538665},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3909625300628061e-05,"output_power":1.983427664701366e-05,"reference_active":true,"delta_db":1.5410093538073646},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.457858491278475e-06,"output_power":5.969561849027781e-06,"reference_active":true,"delta_db":1.2681617789691946}],"active_band_count":3,"active_band_mae_db":1.3009818121050747}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":0.9767944258681361,"rms_envelope_nrmse":0.19950993205904463,"envelope_correlation":0.9907790961936772,"rms_gain_error_db":1.3024272902108018,"reference_mean_power":0.00046390064262223966,"output_mean_power":0.0006261345996054688},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.977269818051972e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.049771362210367e-05,"output_power":0.00010392707835014026,"reference_active":true,"delta_db":1.1094517308512744},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.625262232859731e-06,"output_power":9.558368987063237e-06,"reference_active":true,"delta_db":1.5918071917001426},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.83834356187152e-07,"output_power":3.6498070049874515e-07,"reference_active":true,"delta_db":1.0920493772967366}],"active_band_count":3,"active_band_mae_db":1.2644360999493844}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":1.5057948497062228,"rms_envelope_nrmse":0.2561465614115089,"envelope_correlation":0.9524521213120034,"rms_gain_error_db":0.32969598910949144,"reference_mean_power":4.298191756596504e-05,"output_mean_power":4.6371952871879945e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5620745474492655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.382029340071455e-06,"output_power":7.451780718742981e-06,"reference_active":true,"delta_db":0.6730127012030057},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1967382600492444e-06,"output_power":1.7726473202894776e-06,"reference_active":true,"delta_db":1.7062316271947964},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059624559400327e-07,"output_power":6.146706981072188e-07,"reference_active":true,"delta_db":0.8452421844219872}],"active_band_count":3,"active_band_mae_db":1.0748288376065964}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":0.9339506156339727,"rms_envelope_nrmse":0.19146604695090383,"envelope_correlation":0.9720627219910358,"rms_gain_error_db":0.3296937872098601,"reference_mean_power":4.31749976072588e-05,"output_mean_power":4.65802377507569e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4490976093596581e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.797984044318815e-06,"output_power":6.9450633055967245e-06,"reference_active":true,"delta_db":0.7839919289997526},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942574114894925e-07,"output_power":8.81873265571836e-07,"reference_active":true,"delta_db":1.7143157005888943},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1049309542740296e-08,"output_power":3.3012025105504746e-08,"reference_active":true,"delta_db":1.9543431199358763}],"active_band_count":3,"active_band_mae_db":1.4842169165081742}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"facodec","status":"measured","mos":{"status":"measured","value":4.347302436828613,"predicted_mos":4.347302436828613,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"940bf23dab3121e2bf441b6e4593410ad3f3a59a7c949151f0c676e6a7876659","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7696173380008986},"started_at":"2026-10-13T04:51:24.817944+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7611999939836096e-05,"native_input_samples":96000,"native_input_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","output_sha256":"93f8615d31c2bd197d02376d6640b9816ddc88899188693b8f7adb994c982d73","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7781986290001441,"synthesis_s":5.622533085000214,"total_s":7.400731714000358,"rtf":1.2334552856667262,"output_peak":0.38460057973861694,"output_rms":0.044518108785710206,"finished_at":"2026-10-13T04:51:32.228271+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"93f8615d31c2bd197d02376d6640b9816ddc88899188693b8f7adb994c982d73","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.442189531188306,"spectral_convergence":0.3097988416047718,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.063752855092595,"spectral_convergence":0.3030266250085401,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.4450222001038115,"spectral_convergence":0.3616472012212013,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":7.787820491431981e-05,"delta_db":-0.7606154752111176,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0008512639377140641,"delta_db":-0.3039950163648883,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0006086697120488785,"delta_db":-0.8772479029549006,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.00031047783305272454,"delta_db":-1.158976231598978,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":9.293153565022424e-05,"delta_db":-0.8201400867362963,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":4.064078647611625e-05,"delta_db":-0.9064938946165801,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8045781012471268,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6812390443174331,"pitch":null,"output_rms_dbfs":-27.02926587209072,"output_peak":0.38460057973861694,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.983654862128237},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.442189531188306,"spectral_convergence":0.3097988416047718,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.063752855092595,"spectral_convergence":0.3030266250085401,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.4450222001038115,"spectral_convergence":0.3616472012212013,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":7.787820491431981e-05,"delta_db":-0.7606154752111176,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0008512639377140641,"delta_db":-0.3039950163648883,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0006086697120488785,"delta_db":-0.8772479029549006,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.00031047783305272454,"delta_db":-1.158976231598978,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":9.293153565022424e-05,"delta_db":-0.8201400867362963,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":4.064078647611625e-05,"delta_db":-0.9064938946165801,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8045781012471268,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6812390443174331,"pitch":null,"output_rms_dbfs":-27.02926587209072,"output_peak":0.38460057973861694,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.983654862128237},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8809296113102295,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04815026244547317,"rms_dbfs":-26.348026827773285,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.44194825234116,"mean":-2.5399525960286456e-07,"zero_energy":false},"output":{"rms":0.044518108785710206,"rms_dbfs":-27.02926587209072,"peak":0.38460057973861694,"peak_dbfs":-8.299801309018251,"crest_factor_db":18.729464563072465,"mean":7.66260994850872e-06,"zero_energy":false},"rms_difference_db":-0.681239044317433},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":8.854359085713031e-07,"output_rms_dbfs":-60.52842869325612,"output_zero_energy":false,"output_peak":0.007758779916912317,"output_energy_fraction":8.190777953464753e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":266,"both_estimated_voiced_frames":230,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.9426229508196722,"absolute_difference_cents_median":7.350009446426334,"absolute_difference_cents_p95":36.763229097117545,"difference_over_600_cents_frames":9,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001795198121335632,"log_rms_mae_db":1.1260939311962648,"rms_envelope_nrmse":0.13093422013957542,"envelope_correlation":0.987797471361707,"rms_gain_error_db":-0.6806851835248289,"reference_mean_power":0.002319729266015563,"output_mean_power":0.0019832103638065452},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2030894540936354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042099620393065603,"output_power":0.00037457503431087444,"reference_active":true,"delta_db":-0.5073935190353441},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010212932714007924,"output_power":8.787726909850468e-05,"reference_active":true,"delta_db":-0.6527391846233377},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7753962605616007e-05,"output_power":1.6964059002515122e-05,"reference_active":true,"delta_db":-0.19765526504668313}],"active_band_count":3,"active_band_mae_db":0.45259598956845504}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017377061993537507,"log_rms_mae_db":0.8840494767579375,"rms_envelope_nrmse":0.11597255804801172,"envelope_correlation":0.9910562692994934,"rms_gain_error_db":-0.6801742997216291,"reference_mean_power":0.002329667623369792,"output_mean_power":0.0019919412906644575},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1345793651623817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003814703455166119,"output_power":0.00034115976215335156,"reference_active":true,"delta_db":-0.48502979752031466},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.449258478788483e-05,"output_power":4.591886793585526e-05,"reference_active":true,"delta_db":-0.7434623599914613},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.451644818904226e-07,"output_power":9.58565948135839e-07,"reference_active":true,"delta_db":0.06114604139283598}],"active_band_count":3,"active_band_mae_db":0.42987939963487065}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001705289920117241,"log_rms_mae_db":1.5027568092713108,"rms_envelope_nrmse":0.13879699072335716,"envelope_correlation":0.9855267996872467,"rms_gain_error_db":-0.43101968469450863,"reference_mean_power":0.0014327712363456494,"output_mean_power":0.0012974029639836648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.476686730813627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004044339102676846,"output_power":0.0003757419884805662,"reference_active":true,"delta_db":-0.3195783302039301},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628475293024379e-05,"output_power":8.692989865002909e-05,"reference_active":true,"delta_db":-0.4438834703833592},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2048085119397813e-05,"output_power":1.2914173564139657e-05,"reference_active":true,"delta_db":0.3014859180143402}],"active_band_count":3,"active_band_mae_db":0.3549825728672098}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015393777861692204,"log_rms_mae_db":1.1510542171072118,"rms_envelope_nrmse":0.11750469110013491,"envelope_correlation":0.9895147473868492,"rms_gain_error_db":-0.430036063070111,"reference_mean_power":0.0014385306131629134,"output_mean_power":0.0013029132545182471},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.958921932585308e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037465347734378165,"output_power":0.0003481033098502078,"reference_active":true,"delta_db":-0.31921615173623985},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.015607336171047e-05,"output_power":4.380786949076031e-05,"reference_active":true,"delta_db":-0.58771396300151},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018660953269709e-07,"output_power":1.0281922580423203e-06,"reference_active":true,"delta_db":1.0797247851265634}],"active_band_count":3,"active_band_mae_db":0.6622182999547711}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":1.6377122233117325,"rms_envelope_nrmse":0.1809318602775591,"envelope_correlation":0.9841381915654417,"rms_gain_error_db":-1.116649781177386,"reference_mean_power":0.0004617837976357088,"output_mean_power":0.00035708673128187257},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1028770536544704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.713589283902283e-05,"output_power":7.283127398281366e-05,"reference_active":true,"delta_db":-0.7787917875045015},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3909625300628061e-05,"output_power":9.593526378636856e-06,"reference_active":true,"delta_db":-1.613371570259424},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.457858491278475e-06,"output_power":4.050289747743305e-06,"reference_active":true,"delta_db":-0.4164018558630773}],"active_band_count":3,"active_band_mae_db":0.9361884045423343}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":1.3371229599151024,"rms_envelope_nrmse":0.16161902019462146,"envelope_correlation":0.9892496545786521,"rms_gain_error_db":-1.1163813219306238,"reference_mean_power":0.00046390064262223966,"output_mean_power":0.00035874581375120704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.977269818051972e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.049771362210367e-05,"output_power":6.753787296522551e-05,"reference_active":true,"delta_db":-0.7623616640603894},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.625262232859731e-06,"output_power":4.614143482670345e-06,"reference_active":true,"delta_db":-1.5711197731582764},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.83834356187152e-07,"output_power":2.8463830779677026e-07,"reference_active":true,"delta_db":0.01228386174636412}],"active_band_count":3,"active_band_mae_db":0.7819217663216765}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":1.7085228390919833,"rms_envelope_nrmse":0.22189433740777925,"envelope_correlation":0.9686044328023266,"rms_gain_error_db":-1.0112191365139287,"reference_mean_power":4.298191756596504e-05,"output_mean_power":3.405366608788427e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5620745474492655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.382029340071455e-06,"output_power":5.744214846532479e-06,"reference_active":true,"delta_db":-0.4572812131517389},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1967382600492444e-06,"output_power":1.1097651807878792e-06,"reference_active":true,"delta_db":-0.3276808119117286},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059624559400327e-07,"output_power":6.15792473576391e-07,"reference_active":true,"delta_db":0.8531608453356878}],"active_band_count":3,"active_band_mae_db":0.5460409567997184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":1.4106188382212863,"rms_envelope_nrmse":0.18841857021056602,"envelope_correlation":0.9799610590478939,"rms_gain_error_db":-1.0109051208584305,"reference_mean_power":4.31749976072588e-05,"output_mean_power":3.420911272218683e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4490976093596581e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.797984044318815e-06,"output_power":5.383458178357223e-06,"reference_active":true,"delta_db":-0.32215672304648124},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942574114894925e-07,"output_power":5.214527571229677e-07,"reference_active":true,"delta_db":-0.5675963880640925},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1049309542740296e-08,"output_power":2.732619632213873e-08,"reference_active":true,"delta_db":1.1334132942963262}],"active_band_count":3,"active_band_mae_db":0.6743888018023}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"hifiglot","status":"measured","mos":{"status":"measured","value":4.44533109664917,"predicted_mos":4.44533109664917,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"4b8c4e61e60abdc25b2f3189dd5f40fba90bb828df6947c5040aa6ae3ccd1450","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.6605305559969565},"started_at":"2026-10-13T07:12:35.296873+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0074277509993407875,"native_input_samples":264600,"native_input_sha256":"d9769990624f0ea64fe8b04e2fc6d04d253cc39fa5b370af1cdecb1dd535cded","output_sha256":"f8435ca58f5040595ff299bad6a39a4dad7cb58cdd510ccb80f0071dc9000835","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15952917899994645,"synthesis_s":4.088725792000332,"total_s":4.248254971000279,"rtf":0.7080424951667131,"output_peak":0.38386788964271545,"output_rms":0.04512254062301742,"finished_at":"2026-10-13T07:12:39.564860+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"f8435ca58f5040595ff299bad6a39a4dad7cb58cdd510ccb80f0071dc9000835","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.916097684580876,"output_peak":0.36846598341092374},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.809464114799525,"spectral_convergence":0.47196547402646594,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.664500891730935,"spectral_convergence":0.3915577168750468,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.687475811334442,"spectral_convergence":0.43962357763537985,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":2.3052770016173273e-05,"delta_db":-6.047543624372816,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0009319544172183062,"delta_db":0.08930933811462025,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0006705507878848807,"delta_db":-0.45674804567083416,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.0002915735626091211,"delta_db":-1.4318007845909513,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":9.342699504322354e-05,"delta_db":-0.7970474183657406,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":2.4209905797414283e-05,"delta_db":-3.156183936666684,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.9964388579636079,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5668232458219491,"pitch":null,"output_rms_dbfs":-26.914850073595233,"output_peak":0.36846598341092374,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.720480272621634},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6852103985812877,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04815026244547317,"rms_dbfs":-26.348026827773285,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.44194825234116,"mean":-2.5399525960286456e-07,"zero_energy":false},"output":{"rms":0.04510840762617452,"rms_dbfs":-26.914850073595233,"peak":0.36846598341092374,"peak_dbfs":-8.672051995926864,"crest_factor_db":18.24279807766837,"mean":-2.8482865375896893e-05,"zero_energy":false},"rms_difference_db":-0.5668232458219471},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":6.503759591420215e-07,"output_rms_dbfs":-61.868355206195965,"output_zero_energy":false,"output_peak":0.008972617330022378,"output_energy_fraction":5.859909670764254e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":233,"both_estimated_voiced_frames":199,"reference_voiced_output_unvoiced_frames":45,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.8155737704918032,"absolute_difference_cents_median":8.621799016338125,"absolute_difference_cents_p95":171.79087207166017,"difference_over_600_cents_frames":10,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001795198121335632,"log_rms_mae_db":1.558932919962054,"rms_envelope_nrmse":0.11290495822166946,"envelope_correlation":0.9913894587187396,"rms_gain_error_db":-0.5665451677465781,"reference_mean_power":0.002319729266015563,"output_mean_power":0.0020360234929723423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2030894540936354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042099620393065603,"output_power":0.00037215092508029873,"reference_active":true,"delta_db":-0.5355907699119234},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010212932714007924,"output_power":8.792068884898946e-05,"reference_active":true,"delta_db":-0.6505938857080652},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7753962605616007e-05,"output_power":1.5687701335150048e-05,"reference_active":true,"delta_db":-0.5373598823616733}],"active_band_count":3,"active_band_mae_db":0.5745148459938872}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017377061993537507,"log_rms_mae_db":1.2821298441961255,"rms_envelope_nrmse":0.08626757992014547,"envelope_correlation":0.9966170546238887,"rms_gain_error_db":-0.5665275019383494,"reference_mean_power":0.002329667623369792,"output_mean_power":0.002044754694304327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1345793651623817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003814703455166119,"output_power":0.00033629226650657747,"reference_active":true,"delta_db":-0.5474390246958079},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.449258478788483e-05,"output_power":4.587279097107685e-05,"reference_active":true,"delta_db":-0.7478224453118882},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.451644818904226e-07,"output_power":8.319310948330607e-07,"reference_active":true,"delta_db":-0.5542003592632834}],"active_band_count":3,"active_band_mae_db":0.6164872764236599}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001705289920117241,"log_rms_mae_db":2.2995948427874193,"rms_envelope_nrmse":0.1323346830931713,"envelope_correlation":0.9857506313003817,"rms_gain_error_db":-0.14129944805571715,"reference_mean_power":0.0014327712363456494,"output_mean_power":0.0013869056245468011},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.476686730813627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004044339102676846,"output_power":0.00036930667718389535,"reference_active":true,"delta_db":-0.3946040250400396},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628475293024379e-05,"output_power":8.559997324052088e-05,"reference_active":true,"delta_db":-0.5108389141584136},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2048085119397813e-05,"output_power":9.662356076807097e-06,"reference_active":true,"delta_db":-0.9583498909589883}],"active_band_count":3,"active_band_mae_db":0.6212642767191472}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015393777861692204,"log_rms_mae_db":1.7893081968273559,"rms_envelope_nrmse":0.10297014782302565,"envelope_correlation":0.9908386660259882,"rms_gain_error_db":-0.14081404880101667,"reference_mean_power":0.0014385306131629134,"output_mean_power":0.0013926362757532684},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.958921932585308e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037465347734378165,"output_power":0.00033594293908806744,"reference_active":true,"delta_db":-0.4736425041569204},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.015607336171047e-05,"output_power":4.402072180774421e-05,"reference_active":true,"delta_db":-0.5666636953520764},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018660953269709e-07,"output_power":7.893467269691376e-07,"reference_active":true,"delta_db":-0.06834038429812204}],"active_band_count":3,"active_band_mae_db":0.36954886126903963}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":3.247759636686511,"rms_envelope_nrmse":0.25756230693606735,"envelope_correlation":0.9616066980578782,"rms_gain_error_db":-1.4658043622170438,"reference_mean_power":0.0004617837976357088,"output_mean_power":0.0003295021515605314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1028770536544704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.713589283902283e-05,"output_power":6.537103285077379e-05,"reference_active":true,"delta_db":-1.2481173876234548},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3909625300628061e-05,"output_power":1.0703934749802607e-05,"reference_active":true,"delta_db":-1.1377197805321682},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.457858491278475e-06,"output_power":3.5489074766482034e-06,"reference_active":true,"delta_db":-0.9903160123880784}],"active_band_count":3,"active_band_mae_db":1.1253843935145673}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":2.7639796399297247,"rms_envelope_nrmse":0.2208350845318301,"envelope_correlation":0.9760841456065402,"rms_gain_error_db":-1.4648280405938174,"reference_mean_power":0.00046390064262223966,"output_mean_power":0.00033108703154929555},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.977269818051972e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.049771362210367e-05,"output_power":6.017314000405486e-05,"reference_active":true,"delta_db":-1.2638087047647975},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.625262232859731e-06,"output_power":5.0755651260463e-06,"reference_active":true,"delta_db":-1.1571866794576666},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.83834356187152e-07,"output_power":1.9353759770086372e-07,"reference_active":true,"delta_db":-1.662996166122224}],"active_band_count":3,"active_band_mae_db":1.3613305167815628}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":3.519709843756082,"rms_envelope_nrmse":0.40759402192751976,"envelope_correlation":0.8655433811596231,"rms_gain_error_db":-0.9519405869100983,"reference_mean_power":4.298191756596504e-05,"output_mean_power":3.452166455410878e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5620745474492655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.382029340071455e-06,"output_power":8.857953519540283e-06,"reference_active":true,"delta_db":1.4237460063346694},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1967382600492444e-06,"output_power":1.6936858515326812e-06,"reference_active":true,"delta_db":1.5083368383207514},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059624559400327e-07,"output_power":5.754329441821577e-07,"reference_active":true,"delta_db":0.5587643014059731}],"active_band_count":3,"active_band_mae_db":1.1636157153537978}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":3.016659457761796,"rms_envelope_nrmse":0.37146238828213923,"envelope_correlation":0.8813105766465844,"rms_gain_error_db":-0.950522287738852,"reference_mean_power":4.31749976072588e-05,"output_mean_power":3.4688066523508345e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4490976093596581e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.797984044318815e-06,"output_power":8.156250513894967e-06,"reference_active":true,"delta_db":1.4821354051166227},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942574114894925e-07,"output_power":8.054312183373146e-07,"reference_active":true,"delta_db":1.3205385186671075},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1049309542740296e-08,"output_power":2.8015425918277493e-08,"reference_active":true,"delta_db":1.2415937470606906}],"active_band_count":3,"active_band_mae_db":1.3480892236148068}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"golf","status":"measured","mos":{"status":"measured","value":3.4768569469451904,"predicted_mos":3.4768569469451904,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"5f9ddbfe8f7c37ab2015ffa09c4a2e225bb5b9b6dcba53890fcbe03b5f99417e","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.573387480999372},"started_at":"2026-10-13T05:02:26.256162+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002766223000435275,"native_input_samples":144000,"native_input_sha256":"66415967409011bce98a6f3f353f92acf24c415a4ed73f31afd28e690dd591dc","output_sha256":"6c85fe1aa6b9a182e45d5196fbd5f1e9bf8c6bb95990bfdea7306bd9fd779728","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5290426270003081,"synthesis_s":0.1228523089994269,"total_s":0.651894935999735,"rtf":0.10864915599995584,"output_peak":0.38629743456840515,"output_rms":0.04244655359661236,"finished_at":"2026-10-13T05:02:26.918103+00:00","analysis_stages":{"external_f0_s":0.09155704700060596},"evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"6c85fe1aa6b9a182e45d5196fbd5f1e9bf8c6bb95990bfdea7306bd9fd779728","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.439938617481733,"output_peak":0.3865890886143634},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.868947640719399,"spectral_convergence":0.4713032220742858,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.340697614671833,"spectral_convergence":0.3614682441620164,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.758471759193977,"spectral_convergence":0.412988333420504,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.291789885041953e-05,"output_power":7.68800470340621e-05,"delta_db":-0.8228574019607688,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009144414632616836,"output_power":0.0006947474519716858,"delta_db":-1.193289472750017,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449721249770868,"output_power":0.0006638623410506427,"delta_db":-0.5006198971921483,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040589346940147434,"output_power":0.0002747592413121309,"delta_db":-1.6945975532226845,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011241646466940715,"output_power":7.42336568913674e-05,"delta_db":-1.8022906800298666,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.015489425280975e-05,"output_power":1.8560486356935016e-05,"delta_db":-4.317239668402103,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.7218157789262645,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.096308880404009,"pitch":null,"output_rms_dbfs":-27.439938617481733,"output_peak":0.3865890886143634,"reference_rms_dbfs":-26.343629737077727,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.656039004861737},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8252699925321221,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.048174643905408175,"rms_dbfs":-26.343629737077727,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.4375511616456,"mean":-2.0277122026293824e-06,"zero_energy":false},"output":{"rms":0.04246225647109234,"rms_dbfs":-27.439938617481733,"peak":0.3865890886143634,"peak_dbfs":-8.255008159871322,"crest_factor_db":19.184930457610413,"mean":0.0003092508590030839,"zero_energy":false},"rms_difference_db":-1.0963088804040062},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":1.080411976620032e-06,"output_rms_dbfs":-59.664106102017634,"output_zero_energy":false,"output_peak":0.003967350240593532,"output_energy_fraction":0.00011003962819541703,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":221,"both_estimated_voiced_frames":204,"reference_voiced_output_unvoiced_frames":40,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.8360655737704918,"absolute_difference_cents_median":6.901191429164974,"absolute_difference_cents_p95":131.32949818670983,"difference_over_600_cents_frames":9,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001795198121335632,"log_rms_mae_db":2.4464308468899634,"rms_envelope_nrmse":0.19528868060455742,"envelope_correlation":0.9770619928956172,"rms_gain_error_db":-1.0966613517301227,"reference_mean_power":0.0023223931703046617,"output_mean_power":0.0018041374037100167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.2033273144039954e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042062346487872543,"output_power":0.0003348039422508291,"reference_active":true,"delta_db":-0.9910293347170334},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010173343920563461,"output_power":7.962667891368962e-05,"reference_active":true,"delta_db":-1.0640512406829055},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7642537472955714e-05,"output_power":1.834469232274529e-05,"reference_active":true,"delta_db":0.16949383610517127}],"active_band_count":3,"active_band_mae_db":0.7415248038350367}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017377061993537507,"log_rms_mae_db":2.114799202628384,"rms_envelope_nrmse":0.16462696079760689,"envelope_correlation":0.9877995645997834,"rms_gain_error_db":-1.0972751833958625,"reference_mean_power":0.0023327549437440107,"output_mean_power":0.001811930769815003},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1350032776478166e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000381129408448059,"output_power":0.000305021470977958,"reference_active":true,"delta_db":-0.9674204965343242},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.400550087958404e-05,"output_power":4.239566466753096e-05,"reference_active":true,"delta_db":-1.0511654993774437},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.387488869082391e-07,"output_power":8.221371108737574e-07,"reference_active":true,"delta_db":-0.5760518256995885}],"active_band_count":3,"active_band_mae_db":0.8648792738704522}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001705289920117241,"log_rms_mae_db":2.62634888372676,"rms_envelope_nrmse":0.2002083886714292,"envelope_correlation":0.9727177710124606,"rms_gain_error_db":-0.7293044374474826,"reference_mean_power":0.0014340897533679329,"output_mean_power":0.0012123998920267696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.473695679033684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00040407026825217045,"output_power":0.00030940095164516045,"reference_active":true,"delta_db":-1.1593525079958777},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.597475743967453e-05,"output_power":7.874467701365854e-05,"reference_active":true,"delta_db":-0.8593581718811008},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1975939874173226e-05,"output_power":1.5624996414393308e-05,"reference_active":true,"delta_db":1.1551031941219627}],"active_band_count":3,"active_band_mae_db":1.0579379579996468}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015393777861692204,"log_rms_mae_db":2.0565987728413453,"rms_envelope_nrmse":0.15887903330624564,"envelope_correlation":0.9844867794737968,"rms_gain_error_db":-0.7298498809855902,"reference_mean_power":0.0014402674269585922,"output_mean_power":0.0012174696699305545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.956828162799189e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037426993455050777,"output_power":0.0002861050707827908,"reference_active":true,"delta_db":-1.1665938614772513},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.978700598278765e-05,"output_power":4.194592386172732e-05,"reference_active":true,"delta_db":-0.7442624590011899},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.009765324767222e-07,"output_power":8.64702599730226e-07,"reference_active":true,"delta_db":0.3324697265203396}],"active_band_count":3,"active_band_mae_db":0.7477753489995936}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":2.2743646290743826,"rms_envelope_nrmse":0.25168650777239504,"envelope_correlation":0.972492035639583,"rms_gain_error_db":-1.7644484853512288,"reference_mean_power":0.00046239047384780444,"output_mean_power":0.00030800944177987243},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.107470537917513e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.717909274714403e-05,"output_power":5.617838426517675e-05,"reference_active":true,"delta_db":-1.908431007100265},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3846289100984193e-05,"output_power":8.153486243550383e-06,"reference_active":true,"delta_db":-2.299900523382508},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.466943962577706e-06,"output_power":2.7476182778098273e-06,"reference_active":true,"delta_db":-2.1105410757550565}],"active_band_count":3,"active_band_mae_db":2.1062908687459436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":1.9591827697064554,"rms_envelope_nrmse":0.21875935274187808,"envelope_correlation":0.986182041626173,"rms_gain_error_db":-1.7658662290942286,"reference_mean_power":0.00046457759919978797,"output_mean_power":0.0003093653306596738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.9820655050701825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.053415109341896e-05,"output_power":5.15736909351584e-05,"reference_active":true,"delta_db":-1.935518718237868},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.572833133373123e-06,"output_power":3.887441119543066e-06,"reference_active":true,"delta_db":-2.2808878272856363},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.826668608781266e-07,"output_power":1.586910691033667e-07,"reference_active":true,"delta_db":-2.5072240993012977}],"active_band_count":3,"active_band_mae_db":2.2412102149416007}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":3.000432831889533,"rms_envelope_nrmse":0.3104063622058948,"envelope_correlation":0.9313269993091355,"rms_gain_error_db":-0.9652066809038022,"reference_mean_power":4.305075858104394e-05,"output_mean_power":3.4471496706420585e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.5590971317494033e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.360126903341135e-06,"output_power":4.617978739803554e-06,"reference_active":true,"delta_db":-1.3901385200825085},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1852362910609442e-06,"output_power":1.3069983746592146e-06,"reference_active":true,"delta_db":0.4247010671806828},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.033472380938384e-07,"output_power":4.844616219189795e-07,"reference_active":true,"delta_db":-0.16608311136803247}],"active_band_count":3,"active_band_mae_db":0.660307566210408}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":2.62618221830127,"rms_envelope_nrmse":0.24327199852238968,"envelope_correlation":0.9586878717366414,"rms_gain_error_db":-0.9640105107128376,"reference_mean_power":4.322592773932139e-05,"output_mean_power":3.462129209346461e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.4465441771969583e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.773754203828433e-06,"output_power":4.4187225388294426e-06,"reference_active":true,"delta_db":-1.1616155933525332},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.872398395512772e-07,"output_power":5.890682214316261e-07,"reference_active":true,"delta_db":0.013500830730634239},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.088555872107466e-08,"output_power":2.2877803970795245e-08,"reference_active":true,"delta_db":0.3956823659969362}],"active_band_count":3,"active_band_mae_db":0.5235995966933679}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.762068271636963,"predicted_mos":3.762068271636963,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"35eccec12e011ab81abd2ccf0cef359d40d32dff3a99d6fd0669be0a2079cc4f","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4952868340005807},"started_at":"2026-10-13T07:36:30.789303+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031571750005241483,"native_input_samples":144000,"native_input_sha256":"66415967409011bce98a6f3f353f92acf24c415a4ed73f31afd28e690dd591dc","output_sha256":"e85a231fce5631d81a11e19db8b601db72cff655265cd68dea45b714f9a8977d","output_samples":144120,"length_delta_samples":120,"analysis_s":2.1063285159980296,"synthesis_s":3.8989355730009265,"total_s":6.005264088998956,"rtf":1.0008773481664928,"output_peak":0.41068530082702637,"output_rms":0.047256406997043005,"finished_at":"2026-10-13T07:36:36.808387+00:00","analysis_stages":{"external_f0_s":1.2845698349992745},"evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"e85a231fce5631d81a11e19db8b601db72cff655265cd68dea45b714f9a8977d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.516050028199448,"output_peak":0.41183931549110014},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.0395047942521956,"spectral_convergence":0.45549424760506,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.430126033626764,"spectral_convergence":0.28762305953069267,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.746520071392775,"spectral_convergence":0.29970039332098886,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":0.00010019466189885954,"delta_db":0.3336710407443854,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0008820024463407273,"delta_db":-0.14993947601697621,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0007136122077661185,"delta_db":-0.1864421079440917,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.0003970769803119292,"delta_db":-0.0905351128933688,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":0.00010679435574882845,"delta_db":-0.21628822249849258,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":3.091971532347863e-05,"delta_db":-2.0937600650518657,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5117726708581968,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16775547827980877,"pitch":null,"output_rms_dbfs":-26.515782306053094,"output_peak":0.41183931549110014,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.738716966423911},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.888583570102168,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04815026244547317,"rms_dbfs":-26.348026827773285,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.44194825234116,"mean":-2.5399525960286456e-07,"zero_energy":false},"output":{"rms":0.04722923212788815,"rms_dbfs":-26.515782306053094,"peak":0.41183931549110014,"peak_dbfs":-7.705443932124727,"crest_factor_db":18.810338373928367,"mean":1.788625198508272e-05,"zero_energy":false},"rms_difference_db":-0.1677554782798083},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":1.3472874548286824e-06,"output_rms_dbfs":-58.70539734100977,"output_zero_energy":false,"output_peak":0.04025642576388257,"output_energy_fraction":0.00011073373055207849,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":243,"both_estimated_voiced_frames":218,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.8934426229508197,"absolute_difference_cents_median":4.395315001947461,"absolute_difference_cents_p95":25.03791145715056,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001795198121335632,"log_rms_mae_db":1.9940238397141186,"rms_envelope_nrmse":0.1561215634863759,"envelope_correlation":0.9763979289625013,"rms_gain_error_db":-0.16677411868275382,"reference_mean_power":0.002319729266015563,"output_mean_power":0.0022323376796199875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2030894540936354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042099620393065603,"output_power":0.0004115346420194864,"reference_active":true,"delta_db":-0.09871780898694857},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010212932714007924,"output_power":0.00010128496104312318,"reference_active":true,"delta_db":-0.03605505258030563},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7753962605616007e-05,"output_power":1.512843891068258e-05,"reference_active":true,"delta_db":-0.6950118491208163}],"active_band_count":3,"active_band_mae_db":0.2765949035626902}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017377061993537507,"log_rms_mae_db":1.444865072760791,"rms_envelope_nrmse":0.0995877574793617,"envelope_correlation":0.9899082951619059,"rms_gain_error_db":-0.16678654561156456,"reference_mean_power":0.002329667623369792,"output_mean_power":0.0022418952124317296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1345793651623817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003814703455166119,"output_power":0.00037340087076058215,"reference_active":true,"delta_db":-0.09285456335689496},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.449258478788483e-05,"output_power":5.424852277656943e-05,"reference_active":true,"delta_db":-0.019494920571083905},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.451644818904226e-07,"output_power":8.409802028520509e-07,"reference_active":true,"delta_db":-0.5072162063733148}],"active_band_count":3,"active_band_mae_db":0.20652189676709787}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001705289920117241,"log_rms_mae_db":2.0982605781177983,"rms_envelope_nrmse":0.1763667605757248,"envelope_correlation":0.9745416685809846,"rms_gain_error_db":-0.15005177590666635,"reference_mean_power":0.0014327712363456494,"output_mean_power":0.0013841134109817738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.476686730813627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004044339102676846,"output_power":0.0003867262776380512,"reference_active":true,"delta_db":-0.1944387948786116},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628475293024379e-05,"output_power":9.568725588793857e-05,"reference_active":true,"delta_db":-0.027034202322880736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2048085119397813e-05,"output_power":1.2732968853314864e-05,"reference_active":true,"delta_db":0.24011649585549633}],"active_band_count":3,"active_band_mae_db":0.15386316435232955}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015393777861692204,"log_rms_mae_db":1.3117591259027728,"rms_envelope_nrmse":0.11103644791319117,"envelope_correlation":0.9893026570186908,"rms_gain_error_db":-0.14975824231330323,"reference_mean_power":0.0014385306131629134,"output_mean_power":0.00138977112520853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.958921932585308e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037465347734378165,"output_power":0.00035911871195978004,"reference_active":true,"delta_db":-0.18391733128688792},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.015607336171047e-05,"output_power":5.001518202776402e-05,"reference_active":true,"delta_db":-0.01221675199971981},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018660953269709e-07,"output_power":8.996018893145213e-07,"reference_active":true,"delta_db":0.4994850792302731}],"active_band_count":3,"active_band_mae_db":0.23187305417229362}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":1.9198684481675108,"rms_envelope_nrmse":0.18115999525812979,"envelope_correlation":0.9747482196581565,"rms_gain_error_db":-0.07585128222654482,"reference_mean_power":0.0004617837976357088,"output_mean_power":0.00045378858051365256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1028770536544704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.713589283902283e-05,"output_power":8.559260484669246e-05,"reference_active":true,"delta_db":-0.07760842061498045},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3909625300628061e-05,"output_power":1.3713287707859309e-05,"reference_active":true,"delta_db":-0.06173843365542595},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.457858491278475e-06,"output_power":4.0331636566775375e-06,"reference_active":true,"delta_db":-0.4348043321390819}],"active_band_count":3,"active_band_mae_db":0.19138372880316276}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":1.1687215730777651,"rms_envelope_nrmse":0.10271220937043898,"envelope_correlation":0.9914088961959635,"rms_gain_error_db":-0.07594777995118195,"reference_mean_power":0.00046390064262223966,"output_mean_power":0.0004558586459108444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.977269818051972e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.049771362210367e-05,"output_power":7.874023651089604e-05,"reference_active":true,"delta_db":-0.0958683032951635},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.625262232859731e-06,"output_power":6.66233573130988e-06,"reference_active":true,"delta_db":0.024234414112507617},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.83834356187152e-07,"output_power":2.604847206255211e-07,"reference_active":true,"delta_db":-0.37282708881270765}],"active_band_count":3,"active_band_mae_db":0.16430993540679292}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":2.025602408655219,"rms_envelope_nrmse":0.2566524237193627,"envelope_correlation":0.9522684731252768,"rms_gain_error_db":-0.9198845942124457,"reference_mean_power":4.298191756596504e-05,"output_mean_power":3.4777417375203466e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5620745474492655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.382029340071455e-06,"output_power":5.146943715994521e-06,"reference_active":true,"delta_db":-0.9340937735955039},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1967382600492444e-06,"output_power":1.0280083498205734e-06,"reference_active":true,"delta_db":-0.6600253360910104},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059624559400327e-07,"output_power":3.6847578486988455e-07,"reference_active":true,"delta_db":-1.3770933937804197}],"active_band_count":3,"active_band_mae_db":0.9904041678223114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":1.1799021796300837,"rms_envelope_nrmse":0.15708483706578769,"envelope_correlation":0.9870688161886035,"rms_gain_error_db":-0.9188463055241604,"reference_mean_power":4.31749976072588e-05,"output_mean_power":3.49419945481969e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4490976093596581e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.797984044318815e-06,"output_power":4.617377525739769e-06,"reference_active":true,"delta_db":-0.9888163110818524},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942574114894925e-07,"output_power":5.104629876163579e-07,"reference_active":true,"delta_db":-0.6601034892270184},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1049309542740296e-08,"output_power":1.4189238087017937e-08,"reference_active":true,"delta_db":-1.7127877875015804}],"active_band_count":3,"active_band_mae_db":1.1205691959368171}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.6695828437805176,"predicted_mos":3.6695828437805176,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"c0647cfe95ff81eedd5a8b6470765068bd7d42476699fa7ff9d36f96aa671528","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4642968339994695},"started_at":"2026-10-13T05:10:34.566761+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003152171999317943,"native_input_samples":144000,"native_input_sha256":"66415967409011bce98a6f3f353f92acf24c415a4ed73f31afd28e690dd591dc","output_sha256":"2e339cd211af916af524631f74f3eedde9c532684332869758c3ee57a1b9bfb2","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0216388899989397,"synthesis_s":0.19641578499977186,"total_s":1.2180546749987116,"rtf":0.20300911249978526,"output_peak":0.38156214356422424,"output_rms":0.04731008786177797,"finished_at":"2026-10-13T05:10:35.796258+00:00","analysis_stages":{"external_f0_s":0.1688812569991569},"evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"2e339cd211af916af524631f74f3eedde9c532684332869758c3ee57a1b9bfb2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.506884309148433,"output_peak":0.38032596388075024},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.334533021811261,"spectral_convergence":0.18400139934143842,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.31327861296824,"spectral_convergence":0.20342457408572817,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.970258752943072,"spectral_convergence":0.24082399692424006,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":9.153554828787936e-05,"delta_db":-0.05887692776760113,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0008810461360471815,"delta_db":-0.15465086432873,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0007497623451419432,"delta_db":0.028171436227169413,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.00038541450804416137,"delta_db":-0.22000163144250262,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":0.00010126534828017284,"delta_db":-0.44716261687755837,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":2.9722223031071487e-05,"delta_db":-2.2653020450216483,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5290275869442017,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15192473212744462,"pitch":null,"output_rms_dbfs":-26.49995155990073,"output_peak":0.38032596388075024,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.539356795907524},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.934871287352573,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04815026244547317,"rms_dbfs":-26.348026827773285,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.44194825234116,"mean":-2.5399525960286456e-07,"zero_energy":false},"output":{"rms":0.047315389767309424,"rms_dbfs":-26.49995155990073,"peak":0.38032596388075024,"peak_dbfs":-8.396880508058699,"crest_factor_db":18.103071051842033,"mean":9.01783118247491e-06,"zero_energy":false},"rms_difference_db":-0.15192473212744417},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":1.0960277216346975e-06,"output_rms_dbfs":-59.60178461179457,"output_zero_energy":false,"output_peak":0.013655745092994597,"output_energy_fraction":8.975489218731678e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":265,"both_estimated_voiced_frames":203,"reference_voiced_output_unvoiced_frames":41,"reference_unvoiced_output_voiced_frames":62,"paired_voiced_fraction":0.8319672131147541,"absolute_difference_cents_median":0.9243562057073075,"absolute_difference_cents_p95":4.873039244941293,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001795198121335632,"log_rms_mae_db":0.7746406116247228,"rms_envelope_nrmse":0.07953876035857844,"envelope_correlation":0.9942729107352606,"rms_gain_error_db":-0.15137551646430075,"reference_mean_power":0.002319729266015563,"output_mean_power":0.0022402668369715492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2030894540936354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042099620393065603,"output_power":0.0004255170688373196,"reference_active":true,"delta_db":0.04638805822083279},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010212932714007924,"output_power":9.584180613583916e-05,"reference_active":true,"delta_db":-0.27595481261095867},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7753962605616007e-05,"output_power":1.9978821022237237e-05,"reference_active":true,"delta_db":0.5127455552017047}],"active_band_count":3,"active_band_mae_db":0.2783628086778321}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017377061993537507,"log_rms_mae_db":0.5913246978257182,"rms_envelope_nrmse":0.049774812234940034,"envelope_correlation":0.9979985351359583,"rms_gain_error_db":-0.15069312908131355,"reference_mean_power":0.002329667623369792,"output_mean_power":0.0022502182943556015},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1345793651623817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003814703455166119,"output_power":0.0003893645326571107,"reference_active":true,"delta_db":0.08895606246238047},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.449258478788483e-05,"output_power":5.230570101521638e-05,"reference_active":true,"delta_db":-0.17788381607194154},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.451644818904226e-07,"output_power":1.1307560788259382e-06,"reference_active":true,"delta_db":0.7786153812705634}],"active_band_count":3,"active_band_mae_db":0.34848508660162847}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001705289920117241,"log_rms_mae_db":0.9159225193223629,"rms_envelope_nrmse":0.07915539740777988,"envelope_correlation":0.9949490211208738,"rms_gain_error_db":-0.06415195664131461,"reference_mean_power":0.0014327712363456494,"output_mean_power":0.0014117625546801532},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.476686730813627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004044339102676846,"output_power":0.00040647181975047973,"reference_active":true,"delta_db":0.021828793989620357},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628475293024379e-05,"output_power":9.363662813758357e-05,"reference_active":true,"delta_db":-0.12111754000482396},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2048085119397813e-05,"output_power":1.78244256111783e-05,"reference_active":true,"delta_db":1.7009751654364704}],"active_band_count":3,"active_band_mae_db":0.6146404998103049}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015393777861692204,"log_rms_mae_db":0.5559891910070169,"rms_envelope_nrmse":0.038882092531464685,"envelope_correlation":0.9987615341140447,"rms_gain_error_db":-0.06306432601342235,"reference_mean_power":0.0014385306131629134,"output_mean_power":0.0014177925040479273},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.958921932585308e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037465347734378165,"output_power":0.0003796684921721703,"reference_active":true,"delta_db":0.05774789666066747},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.015607336171047e-05,"output_power":5.023516804875241e-05,"reference_active":true,"delta_db":0.006843304805283111},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018660953269709e-07,"output_power":1.092372365416222e-06,"reference_active":true,"delta_db":1.342688539925564}],"active_band_count":3,"active_band_mae_db":0.4690932471305049}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":1.0382946185179156,"rms_envelope_nrmse":0.09618271581052618,"envelope_correlation":0.9933693908470057,"rms_gain_error_db":-0.2450758043315714,"reference_mean_power":0.0004617837976357088,"output_mean_power":0.0004364465959266502},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1028770536544704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.713589283902283e-05,"output_power":8.065672662545303e-05,"reference_active":true,"delta_db":-0.3355649294711169},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3909625300628061e-05,"output_power":1.1435054562787643e-05,"reference_active":true,"delta_db":-0.8507718986602774},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.457858491278475e-06,"output_power":3.7531222855447052e-06,"reference_active":true,"delta_db":-0.7473356329034925}],"active_band_count":3,"active_band_mae_db":0.644557487011629}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":0.6829931867928549,"rms_envelope_nrmse":0.06058957028684636,"envelope_correlation":0.9976279763543169,"rms_gain_error_db":-0.24468764202486562,"reference_mean_power":0.00046390064262223966,"output_mean_power":0.0004384864827916644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.977269818051972e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.049771362210367e-05,"output_power":7.460863444169895e-05,"reference_active":true,"delta_db":-0.3299445407864847},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.625262232859731e-06,"output_power":5.358299727896469e-06,"reference_active":true,"delta_db":-0.9217606951353896},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.83834356187152e-07,"output_power":2.2892699303276654e-07,"reference_active":true,"delta_db":-0.9336795881637576}],"active_band_count":3,"active_band_mae_db":0.7284616080285439}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":1.1710569444717407,"rms_envelope_nrmse":0.1697265701125578,"envelope_correlation":0.979151742204219,"rms_gain_error_db":-0.4969692413390698,"reference_mean_power":4.298191756596504e-05,"output_mean_power":3.8334416999562704e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5620745474492655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.382029340071455e-06,"output_power":5.77281663642527e-06,"reference_active":true,"delta_db":-0.4357103334020685},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1967382600492444e-06,"output_power":1.074108758915928e-06,"reference_active":true,"delta_db":-0.46950917673397596},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059624559400327e-07,"output_power":3.2509161731463864e-07,"reference_active":true,"delta_db":-1.9211252087943278}],"active_band_count":3,"active_band_mae_db":0.9421149063101241}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":0.7821932002021501,"rms_envelope_nrmse":0.08697344556378778,"envelope_correlation":0.9961093280375385,"rms_gain_error_db":-0.49696700196480115,"reference_mean_power":4.31749976072588e-05,"output_mean_power":3.8506639754109735e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4490976093596581e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.797984044318815e-06,"output_power":5.276338321913188e-06,"reference_active":true,"delta_db":-0.4094438088549649},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942574114894925e-07,"output_power":5.299750363902624e-07,"reference_active":true,"delta_db":-0.4971919354047363},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1049309542740296e-08,"output_power":1.3739424202310846e-08,"reference_active":true,"delta_db":-1.8526932221768608}],"active_band_count":3,"active_band_mae_db":0.9197763221455206}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0000","method":"magphase","status":"measured","mos":{"status":"measured","value":3.916459560394287,"predicted_mos":3.916459560394287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288660,"input_sample_rate":48000,"input_seconds":6.01375,"input_float32_sha256":"b348df72eb9cebb74adcddaefe9124b72eced342a74177b201d46b579d08a2a6","internal_samples":96220,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.463295122001},"started_at":"2026-10-13T05:14:54.257013+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.011761137000576127,"native_input_samples":288000,"native_input_sha256":"0c800329dad125e209915d280c7918e27a6cff80a03d58ab4114343a3a29e237","output_sha256":"dd121fb60e0024d51559803ab38a2c24c512c5056a1b81015020d162d1c63d84","output_samples":288660,"length_delta_samples":660,"analysis_s":3.517259752999962,"synthesis_s":1.029195858998719,"total_s":4.546455611998681,"rtf":0.7577426019997802,"output_peak":0.43240110168690615,"output_rms":0.04642620831355037,"finished_at":"2026-10-13T05:14:58.826108+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"dd121fb60e0024d51559803ab38a2c24c512c5056a1b81015020d162d1c63d84","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96220,"length_delta_samples":220,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.676139633793557,"output_peak":0.4327161511824729},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9977135730617335,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.718368088146753,"spectral_convergence":0.5025974863161158,"spectral_convergence_status":"ok","amplitude_floor":1.174323222508143e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.652229642246057,"spectral_convergence":0.21604182144621084,"spectral_convergence_status":"ok","amplitude_floor":1.1752576893073956e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9252989671407725,"spectral_convergence":0.21003725773694307,"spectral_convergence_status":"ok","amplitude_floor":9.275961087692859e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.278493760532488e-05,"output_power":8.48283554909498e-05,"delta_db":-0.3893643232086895,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009129851997232178,"output_power":0.0008231814767772401,"delta_db":-0.44968147930094465,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007449145920476126,"output_power":0.0006796980956207484,"delta_db":-0.3979042858602591,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00040544152038629675,"output_power":0.00040960437406811935,"delta_db":0.04436363930549168,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011224761509504284,"output_power":0.00012061122879040038,"delta_db":0.31210619692156455,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.007390871044891e-05,"output_power":3.544547779519133e-05,"delta_db":-1.5005065834236802,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5156544180034383,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.32078858861396586,"pitch":null,"output_rms_dbfs":-26.66881541638725,"output_peak":0.4327161511824729,"reference_rms_dbfs":-26.348026827773285,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.098632232511194},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8924859833605043,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04815026244547317,"rms_dbfs":-26.348026827773285,"peak":0.402435302734375,"peak_dbfs":-7.906078575432125,"crest_factor_db":18.44194825234116,"mean":-2.5399525960286456e-07,"zero_energy":false},"output":{"rms":0.04640440721033563,"rms_dbfs":-26.66881541638725,"peak":0.4327161511824729,"peak_dbfs":-7.275937888097782,"crest_factor_db":19.392877528289468,"mean":-0.00035252397370162024,"zero_energy":false},"rms_difference_db":-0.320788588613965},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.222736294846982e-06,"threshold_dbfs":-54.917752298206764,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.0135754498399116e-06,"output_mean_power":7.897815543722536e-07,"output_rms_dbfs":-61.02493013582372,"output_zero_energy":false,"output_peak":0.011161815816850496,"output_energy_fraction":6.72403496074457e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":206,"both_estimated_voiced_frames":199,"reference_voiced_output_unvoiced_frames":45,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.8155737704918032,"absolute_difference_cents_median":3.6216598013545442,"absolute_difference_cents_p95":17.50398430758682,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001795198121335632,"log_rms_mae_db":1.5466576397732035,"rms_envelope_nrmse":0.12060448940981551,"envelope_correlation":0.9868786066206291,"rms_gain_error_db":-0.32005957706037547,"reference_mean_power":0.002319729266015563,"output_mean_power":0.0021549209528801514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2030894540936354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042099620393065603,"output_power":0.00038839832673251445,"reference_active":true,"delta_db":-0.3500082962883114},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010212932714007924,"output_power":9.611632369513303e-05,"reference_active":true,"delta_db":-0.2635331931080386},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7753962605616007e-05,"output_power":1.6434027908778027e-05,"reference_active":true,"delta_db":-0.335512806393887}],"active_band_count":3,"active_band_mae_db":0.31635143193007903}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017377061993537507,"log_rms_mae_db":1.0398422333797175,"rms_envelope_nrmse":0.07431795431204653,"envelope_correlation":0.9955820386793132,"rms_gain_error_db":-0.31865970152368256,"reference_mean_power":0.002329667623369792,"output_mean_power":0.002164850917142306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1345793651623817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003814703455166119,"output_power":0.0003516303816981857,"reference_active":true,"delta_db":-0.35374390617509366},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.449258478788483e-05,"output_power":5.1479091932982534e-05,"reference_active":true,"delta_db":-0.2470653106138908},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.451644818904226e-07,"output_power":9.197526606292431e-07,"reference_active":true,"delta_db":-0.11836340201403138}],"active_band_count":3,"active_band_mae_db":0.23972420626767196}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001705289920117241,"log_rms_mae_db":1.7413885122490276,"rms_envelope_nrmse":0.11722794612131499,"envelope_correlation":0.9904957397116809,"rms_gain_error_db":-0.4503215146110768,"reference_mean_power":0.0014327712363456494,"output_mean_power":0.00129164956721888},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.476686730813627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004044339102676846,"output_power":0.00035701592839893116,"reference_active":true,"delta_db":-0.5415996965545334},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.628475293024379e-05,"output_power":8.777307387659263e-05,"reference_active":true,"delta_db":-0.4019621239776334},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2048085119397813e-05,"output_power":1.1990570914411795e-05,"reference_active":true,"delta_db":-0.02078165225464426}],"active_band_count":3,"active_band_mae_db":0.32144782426227037}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015393777861692204,"log_rms_mae_db":1.191412753788645,"rms_envelope_nrmse":0.07761743630196521,"envelope_correlation":0.9968187537741268,"rms_gain_error_db":-0.4481231922445251,"reference_mean_power":0.0014385306131629134,"output_mean_power":0.001297498275371078},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.958921932585308e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037465347734378165,"output_power":0.00033026633020411656,"reference_active":true,"delta_db":-0.5476546710223994},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.015607336171047e-05,"output_power":4.616375883393952e-05,"reference_active":true,"delta_db":-0.36022365393181544},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.018660953269709e-07,"output_power":8.610148717062641e-07,"reference_active":true,"delta_db":0.30908801849942263}],"active_band_count":3,"active_band_mae_db":0.40565544781787916}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.828766409891504e-05,"log_rms_mae_db":1.5531303664862017,"rms_envelope_nrmse":0.16240558302096975,"envelope_correlation":0.9801893537461006,"rms_gain_error_db":0.08675836351046345,"reference_mean_power":0.0004617837976357088,"output_mean_power":0.00047110154370728625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1028770536544704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.713589283902283e-05,"output_power":9.030937527543866e-05,"reference_active":true,"delta_db":0.1553575245191155},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3909625300628061e-05,"output_power":1.554869230469961e-05,"reference_active":true,"delta_db":0.4837843826249312},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.457858491278475e-06,"output_power":4.358842909841541e-06,"reference_active":true,"delta_db":-0.09755060689204201}],"active_band_count":3,"active_band_mae_db":0.2455641713453629}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.324411735838956e-05,"log_rms_mae_db":0.9126745569364402,"rms_envelope_nrmse":0.08612106250587025,"envelope_correlation":0.9941416537565537,"rms_gain_error_db":0.08721058090358509,"reference_mean_power":0.00046390064262223966,"output_mean_power":0.0004733103835864504},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.977269818051972e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.049771362210367e-05,"output_power":8.350466078545064e-05,"reference_active":true,"delta_db":0.15927170895526516},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.625262232859731e-06,"output_power":7.43349534278867e-06,"reference_active":true,"delta_db":0.49990001019853136},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.83834356187152e-07,"output_power":2.9875606563150147e-07,"reference_active":true,"delta_db":0.22251768859096394}],"active_band_count":3,"active_band_mae_db":0.29389646924825347}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.433462630283972e-05,"log_rms_mae_db":1.7540392573777948,"rms_envelope_nrmse":0.30425229080333854,"envelope_correlation":0.9318748295675393,"rms_gain_error_db":0.2749592114709283,"reference_mean_power":4.298191756596504e-05,"output_mean_power":4.5791166607157624e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5620745474492655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.382029340071455e-06,"output_power":7.446104019278546e-06,"reference_active":true,"delta_db":0.6697030241374392},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1967382600492444e-06,"output_power":1.5017849537210424e-06,"reference_active":true,"delta_db":0.9860857308273523},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059624559400327e-07,"output_power":6.252409968828751e-07,"reference_active":true,"delta_db":0.9192915488018037}],"active_band_count":3,"active_band_mae_db":0.8583601012555317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.178045699602676e-05,"log_rms_mae_db":1.051553333626987,"rms_envelope_nrmse":0.1620914074385231,"envelope_correlation":0.979837308823902,"rms_gain_error_db":0.2749315724708298,"reference_mean_power":4.31749976072588e-05,"output_mean_power":4.599657341101544e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4490976093596581e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.797984044318815e-06,"output_power":6.812520913638311e-06,"reference_active":true,"delta_db":0.7003083264975684},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942574114894925e-07,"output_power":7.138321860192643e-07,"reference_active":true,"delta_db":0.7962151909947518},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1049309542740296e-08,"output_power":2.8598836035891196e-08,"reference_active":true,"delta_db":1.3311050310971324}],"active_band_count":3,"active_band_mae_db":0.9425428495298176}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"hnm","status":"measured","mos":{"status":"measured","value":2.7441649436950684,"predicted_mos":2.7441649436950684,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"41750e5427621f7b83c67d160b6a47c8ff8af3113dbfcdc744766abc7f7fb565","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3508442459988146},"started_at":"2026-10-13T04:39:04.330652+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3949999811302405e-05,"native_input_samples":96000,"native_input_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","output_sha256":"d5bec7ac00f157c7fdf18e4929e0bb9254f5181d8890f201770559b3aed73da1","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1464869609999369,"synthesis_s":0.4961432770005558,"total_s":1.6426302380004927,"rtf":0.27377170633341547,"output_peak":0.324004507016577,"output_rms":0.044166905440962405,"finished_at":"2026-10-13T04:39:05.978280+00:00","evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"d5bec7ac00f157c7fdf18e4929e0bb9254f5181d8890f201770559b3aed73da1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.158712445959529,"spectral_convergence":0.20982438053505484,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.267313981677057,"spectral_convergence":0.20327886023876757,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9959248142727155,"spectral_convergence":0.21429830227954474,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":8.317493445625315e-05,"delta_db":0.08243896714037716,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.0005089884823197461,"delta_db":0.01005542746483369,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0009136533469938374,"delta_db":-0.08341260546222744,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.0003392489844386275,"delta_db":-0.19341201340068515,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":9.005873137783903e-05,"delta_db":-0.30771116217670114,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":1.5591056644610874e-05,"delta_db":0.4404938332467905,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18625400148193583,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07840093198598651,"pitch":null,"output_rms_dbfs":-27.09806057131615,"output_peak":0.324004507016577,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.473983747303101},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.158712445959529,"spectral_convergence":0.20982438053505484,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.267313981677057,"spectral_convergence":0.20327886023876757,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9959248142727155,"spectral_convergence":0.21429830227954474,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":8.317493445625315e-05,"delta_db":0.08243896714037716,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.0005089884823197461,"delta_db":0.01005542746483369,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0009136533469938374,"delta_db":-0.08341260546222744,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.0003392489844386275,"delta_db":-0.19341201340068515,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":9.005873137783903e-05,"delta_db":-0.30771116217670114,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":1.5591056644610874e-05,"delta_db":0.4404938332467905,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18625400148193583,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07840093198598651,"pitch":null,"output_rms_dbfs":-27.09806057131615,"output_peak":0.324004507016577,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.473983747303101},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8931926134520461,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04456737119893152,"rms_dbfs":-27.01965963933016,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.1790027369579,"mean":2.3174285888671874e-07,"zero_energy":false},"output":{"rms":0.044166905440962405,"rms_dbfs":-27.09806057131615,"peak":0.324004507016577,"peak_dbfs":-9.788978971249527,"crest_factor_db":17.30908160006662,"mean":-3.356423472968565e-05,"zero_energy":false},"rms_difference_db":-0.07840093198598908},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":8.237005059161351e-07,"output_rms_dbfs":-60.84230667272941,"output_zero_energy":false,"output_peak":0.0026746693198219495,"output_energy_fraction":3.8706760830371075e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":305,"output_estimated_voiced_frames":308,"both_estimated_voiced_frames":283,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.9278688524590164,"absolute_difference_cents_median":1.046019651058323,"absolute_difference_cents_p95":11.578976050327912,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001511941088699898,"log_rms_mae_db":0.5998489795795416,"rms_envelope_nrmse":0.057457615376175816,"envelope_correlation":0.996357064718397,"rms_gain_error_db":-0.07840316094556554,"reference_mean_power":0.0019872702200913737,"output_mean_power":0.0019517159370839251},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.92641346613152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000630498462430925,"output_power":0.0006259514124866816,"reference_active":true,"delta_db":-0.031434080424498106},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.948331074534803e-05,"output_power":7.478183280245725e-05,"reference_active":true,"delta_db":-0.2647984370604787},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.312350608813158e-05,"output_power":1.3013780601119729e-05,"reference_active":true,"delta_db":-0.03646396356491222}],"active_band_count":3,"active_band_mae_db":0.11089882701662968}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013307215895407113,"log_rms_mae_db":0.3269225080754127,"rms_envelope_nrmse":0.032582503554443865,"envelope_correlation":0.9988045893300053,"rms_gain_error_db":-0.07900295199130679,"reference_mean_power":0.0019932283004473585,"output_mean_power":0.001957297086249472},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.959638607230672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005794424432114469,"output_power":0.0005768105867814571,"reference_active":true,"delta_db":-0.019770805680906323},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.108494970400102e-05,"output_power":3.8831962310232945e-05,"reference_active":true,"delta_db":-0.24493422180910268},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.934091733283845e-07,"output_power":6.757419378836889e-07,"reference_active":true,"delta_db":-0.6971634450495512}],"active_band_count":3,"active_band_mae_db":0.3206228241798534}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014956359127228178,"log_rms_mae_db":1.0688761836341576,"rms_envelope_nrmse":0.0728809473089485,"envelope_correlation":0.9951118942795886,"rms_gain_error_db":-0.07075187196705239,"reference_mean_power":0.001265981682968443,"output_mean_power":0.0012455243857569186},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.65319874591582e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005127917421071942,"output_power":0.0005061472551298815,"reference_active":true,"delta_db":-0.05664136794854785},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.49871434628785e-05,"output_power":6.938837850431033e-05,"reference_active":true,"delta_db":-0.3370007119528108},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1228212970044594e-05,"output_power":9.684185265794678e-06,"reference_active":true,"delta_db":-0.642475522477946}],"active_band_count":3,"active_band_mae_db":0.34537253412643487}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012969159209010741,"log_rms_mae_db":0.599050616493911,"rms_envelope_nrmse":0.03938034002762386,"envelope_correlation":0.998503390764297,"rms_gain_error_db":-0.07082345908155176,"reference_mean_power":0.0012704534730935044,"output_mean_power":0.0012499033121646224},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.90499086389905e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004798268169104881,"output_power":0.00047265423398834506,"reference_active":true,"delta_db":-0.0654096386359588},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.776254953569435e-05,"output_power":3.523710891748928e-05,"reference_active":true,"delta_db":-0.3006103912871559},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.708330714477494e-07,"output_power":5.680196837314942e-07,"reference_active":true,"delta_db":-1.3259695345304086}],"active_band_count":3,"active_band_mae_db":0.5639965214845077}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":1.090252971679023,"rms_envelope_nrmse":0.10273700845489636,"envelope_correlation":0.9900121927556496,"rms_gain_error_db":-0.21507962546854967,"reference_mean_power":0.0004068165018391945,"output_mean_power":0.0003871601045807655},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.83132236545664e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016437493457690725,"output_power":0.00015979911253239892,"reference_active":true,"delta_db":-0.12261229912593777},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7703924674415496e-05,"output_power":2.728250929575117e-05,"reference_active":true,"delta_db":-0.06656985869205144},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.687130809571587e-06,"output_power":4.734337345149775e-06,"reference_active":true,"delta_db":0.043521262596640034}],"active_band_count":3,"active_band_mae_db":0.0775678068048764}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":0.6042571851136084,"rms_envelope_nrmse":0.057332852935908146,"envelope_correlation":0.9971267835930097,"rms_gain_error_db":-0.21811456829763676,"reference_mean_power":0.0004069857359569261,"output_mean_power":0.00038705058794338577},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.5932421541601994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001538886928367015,"output_power":0.0001509333498246206,"reference_active":true,"delta_db":-0.08421499763529107},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4161267354242535e-05,"output_power":1.383642681694152e-05,"reference_active":true,"delta_db":-0.10078171642878656},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.792056845121366e-07,"output_power":2.742726861154126e-07,"reference_active":true,"delta_db":-0.07741696250365784}],"active_band_count":3,"active_band_mae_db":0.08747122552257851}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":1.6235433982932532,"rms_envelope_nrmse":0.26056330266294986,"envelope_correlation":0.9397077287721113,"rms_gain_error_db":-0.051571306529667724,"reference_mean_power":1.611693883749463e-05,"output_mean_power":1.5926686339828632e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.125384681951734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2265606388859267e-06,"output_power":2.582806589392251e-06,"reference_active":true,"delta_db":0.6445735838288638},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4345900244965858e-06,"output_power":1.2769757070734332e-06,"reference_active":true,"delta_db":-0.5054517120655235},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80760798762104e-07,"output_power":5.448325834432852e-07,"reference_active":true,"delta_db":0.5433402438400056}],"active_band_count":3,"active_band_mae_db":0.5644551799114642}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":1.0372546115394232,"rms_envelope_nrmse":0.19198159168546228,"envelope_correlation":0.9641512421179013,"rms_gain_error_db":-0.051555775815855485,"reference_mean_power":1.6196973858403443e-05,"output_mean_power":1.6005833824894557e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.254945521069739e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.488949340159725e-06,"output_power":2.885710322829121e-06,"reference_active":true,"delta_db":0.642366759157164},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.648300260916528e-07,"output_power":6.877504516353721e-07,"reference_active":true,"delta_db":-0.46134045103034377},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.810620519312689e-08,"output_power":4.5328137436449914e-08,"reference_active":true,"delta_db":-0.25833225421104006}],"active_band_count":3,"active_band_mae_db":0.45401315479951593}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.4080660343170166,"predicted_mos":3.4080660343170166,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b02b8f05e57ca370d2a1148565219ef231b715d604e5ae7ad88c50f46e15851a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.250961173000178},"started_at":"2026-10-13T04:42:01.758943+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5841999558906537e-05,"native_input_samples":96000,"native_input_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","output_sha256":"06897f8a9e76371878ac8a7e7b23e550f69da8fab415abc6ef65eba428ddbaa0","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3215998309997303,"synthesis_s":0.1784391590008454,"total_s":1.5000389900005757,"rtf":0.2500064983334293,"output_peak":0.40070628071943154,"output_rms":0.05366914744084067,"finished_at":"2026-10-13T04:42:03.265035+00:00","evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"06897f8a9e76371878ac8a7e7b23e550f69da8fab415abc6ef65eba428ddbaa0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.060904070308429,"spectral_convergence":0.5353391164703495,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.051330086940604,"spectral_convergence":0.3237196179843437,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.011560010033144,"spectral_convergence":0.3602335586138617,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":9.441532607928538e-05,"delta_db":0.632939267920416,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.0007286591464208317,"delta_db":1.5682200847051175,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0014553478967513904,"delta_db":1.9384411252517852,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.00046444500748867544,"delta_db":1.1707454124069119,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":0.00012083333813158871,"delta_db":0.9688983228205845,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":1.6676672154922667e-05,"delta_db":0.7328322501243838,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1686794105381997,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6141535668039564,"pitch":null,"output_rms_dbfs":-25.405506072526208,"output_peak":0.40070628071943154,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7079313890940595},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.060904070308429,"spectral_convergence":0.5353391164703495,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.051330086940604,"spectral_convergence":0.3237196179843437,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.011560010033144,"spectral_convergence":0.3602335586138617,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":9.441532607928538e-05,"delta_db":0.632939267920416,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.0007286591464208317,"delta_db":1.5682200847051175,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0014553478967513904,"delta_db":1.9384411252517852,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.00046444500748867544,"delta_db":1.1707454124069119,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":0.00012083333813158871,"delta_db":0.9688983228205845,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":1.6676672154922667e-05,"delta_db":0.7328322501243838,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1686794105381997,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6141535668039564,"pitch":null,"output_rms_dbfs":-25.405506072526208,"output_peak":0.40070628071943154,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7079313890940595},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9378153469436993,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04456737119893152,"rms_dbfs":-27.01965963933016,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.1790027369579,"mean":2.3174285888671874e-07,"zero_energy":false},"output":{"rms":0.05366914744084067,"rms_dbfs":-25.405506072526208,"peak":0.40070628071943154,"peak_dbfs":-7.943477006578629,"crest_factor_db":17.462029065947576,"mean":1.8700414224189872e-06,"zero_energy":false},"rms_difference_db":1.614153566803953},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":1.1635481020372904e-06,"output_rms_dbfs":-59.34215657914386,"output_zero_energy":false,"output_peak":0.00334498685199411,"output_energy_fraction":3.702937556046591e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":305,"output_estimated_voiced_frames":361,"both_estimated_voiced_frames":280,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":81,"paired_voiced_fraction":0.9180327868852459,"absolute_difference_cents_median":4.952968891756244,"absolute_difference_cents_p95":1906.8686012069495,"difference_over_600_cents_frames":24,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001511941088699898,"log_rms_mae_db":1.4790496650219258,"rms_envelope_nrmse":0.25365536319884435,"envelope_correlation":0.9799950487715399,"rms_gain_error_db":1.6155121572590345,"reference_mean_power":0.0019872702200913737,"output_mean_power":0.002882757696222013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.92641346613152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000630498462430925,"output_power":0.0009282094679809645,"reference_active":true,"delta_db":1.6796196220826807},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.948331074534803e-05,"output_power":0.00013292355926769434,"reference_active":true,"delta_db":2.2332601314913534},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.312350608813158e-05,"output_power":1.5924955895359563e-05,"reference_active":true,"delta_db":0.8402836115724083}],"active_band_count":3,"active_band_mae_db":1.5843877883821476}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013307215895407113,"log_rms_mae_db":1.3213980569105652,"rms_envelope_nrmse":0.23816703335877318,"envelope_correlation":0.985934317711214,"rms_gain_error_db":1.6170442273200625,"reference_mean_power":0.0019932283004473585,"output_mean_power":0.00289242074366767},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.959638607230672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005794424432114469,"output_power":0.000860992705523091,"reference_active":true,"delta_db":1.719891686588937},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.108494970400102e-05,"output_power":6.802863406963242e-05,"reference_active":true,"delta_db":2.1900899130628897},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.934091733283845e-07,"output_power":9.27229205698683e-07,"reference_active":true,"delta_db":0.6768988505794976}],"active_band_count":3,"active_band_mae_db":1.5289601500771084}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014956359127228178,"log_rms_mae_db":1.6926081984333847,"rms_envelope_nrmse":0.31406990889965924,"envelope_correlation":0.978595700932702,"rms_gain_error_db":2.0148476299497693,"reference_mean_power":0.001265981682968443,"output_mean_power":0.0020133171132746793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.65319874591582e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005127917421071942,"output_power":0.000807950826441556,"reference_active":true,"delta_db":1.974439067404448},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.49871434628785e-05,"output_power":0.000128230955962119,"reference_active":true,"delta_db":2.330060699110437},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1228212970044594e-05,"output_power":1.3685075538965e-05,"reference_active":true,"delta_db":0.8593655749499238}],"active_band_count":3,"active_band_mae_db":1.7212884471549363}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012969159209010741,"log_rms_mae_db":1.470558891609891,"rms_envelope_nrmse":0.29768701571360057,"envelope_correlation":0.9847559892691543,"rms_gain_error_db":2.0160914442645463,"reference_mean_power":0.0012704534730935044,"output_mean_power":0.002021007425798035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.90499086389905e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004798268169104881,"output_power":0.0007618310570683618,"reference_active":true,"delta_db":2.007741568085628},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.776254953569435e-05,"output_power":6.375879035473695e-05,"reference_active":true,"delta_db":2.274787609641987},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.708330714477494e-07,"output_power":9.296951776955898e-07,"reference_active":true,"delta_db":0.813802391480712}],"active_band_count":3,"active_band_mae_db":1.698777189736109}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":1.2950446188045124,"rms_envelope_nrmse":0.1948461098970514,"envelope_correlation":0.9838251639472049,"rms_gain_error_db":1.1215682972739514,"reference_mean_power":0.0004068165018391945,"output_mean_power":0.0005266903857392362},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.83132236545664e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016437493457690725,"output_power":0.00021544418250805974,"reference_active":true,"delta_db":1.1749917864605934},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7703924674415496e-05,"output_power":3.9092858876126345e-05,"reference_active":true,"delta_db":1.495561340016317},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.687130809571587e-06,"output_power":5.976687312880648e-06,"reference_active":true,"delta_db":1.0555346101308483}],"active_band_count":3,"active_band_mae_db":1.2420292455359196}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":0.9519948893751903,"rms_envelope_nrmse":0.16651290507530664,"envelope_correlation":0.9920833475682392,"rms_gain_error_db":1.1237361859018131,"reference_mean_power":0.0004069857359569261,"output_mean_power":0.0005271725725339549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.5932421541601994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001538886928367015,"output_power":0.0002035454905360287,"reference_active":true,"delta_db":1.2145477453604816},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4161267354242535e-05,"output_power":1.959766100587262e-05,"reference_active":true,"delta_db":1.4110211908783208},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.792056845121366e-07,"output_power":3.5522868498453993e-07,"reference_active":true,"delta_db":1.045837720329559}],"active_band_count":3,"active_band_mae_db":1.2238022188561206}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":1.5893599376433574,"rms_envelope_nrmse":0.21893805317827142,"envelope_correlation":0.9673526045820063,"rms_gain_error_db":0.7330697629825421,"reference_mean_power":1.611693883749463e-05,"output_mean_power":1.908049046617026e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.125384681951734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2265606388859267e-06,"output_power":2.9152546473169735e-06,"reference_active":true,"delta_db":1.1704196900137767},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4345900244965858e-06,"output_power":1.7124488943877503e-06,"reference_active":true,"delta_db":0.7688981283554115},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80760798762104e-07,"output_power":5.718034164658792e-07,"reference_active":true,"delta_db":0.7531769782985777}],"active_band_count":3,"active_band_mae_db":0.8974982655559219}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":1.0378271004944533,"rms_envelope_nrmse":0.15259367743805052,"envelope_correlation":0.9861906818266342,"rms_gain_error_db":0.7332109407361433,"reference_mean_power":1.6196973858403443e-05,"output_mean_power":1.917586551935821e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.254945521069739e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.488949340159725e-06,"output_power":3.227283242037616e-06,"reference_active":true,"delta_db":1.1282102575814736},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.648300260916528e-07,"output_power":9.170429307002234e-07,"reference_active":true,"delta_db":0.788247379678696},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.810620519312689e-08,"output_power":5.510026671320484e-08,"reference_active":true,"delta_db":0.5895260166712056}],"active_band_count":3,"active_band_mae_db":0.8353278846437918}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"facodec","status":"measured","mos":{"status":"measured","value":4.126815319061279,"predicted_mos":4.126815319061279,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ee4683d658aa1af8be71488f236a1d6aa5911e14557bd5d9a7dafe4a950504e9","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.387355151999145},"started_at":"2026-10-13T04:51:32.229998+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.1944999793486204e-05,"native_input_samples":96000,"native_input_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","output_sha256":"79c38f19367d6adfe57c6612b053fdca81d91b6b3a99dc82950948d6498ad54e","output_samples":96000,"length_delta_samples":0,"analysis_s":1.8279702539994105,"synthesis_s":5.38097961800031,"total_s":7.208949871999721,"rtf":1.201491645333287,"output_peak":0.3469199240207672,"output_rms":0.041426943833279745,"finished_at":"2026-10-13T04:51:39.448116+00:00","evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"79c38f19367d6adfe57c6612b053fdca81d91b6b3a99dc82950948d6498ad54e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.344509912612676,"spectral_convergence":0.28453475708121856,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.037291196894736,"spectral_convergence":0.27341689024878846,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.498770557158271,"spectral_convergence":0.3144101111346818,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":6.873289488366938e-05,"delta_db":-0.7458393473679135,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.000491928772140715,"delta_db":-0.13800187735582847,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0007644887153885661,"delta_db":-0.857516308823011,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.00030400425353510553,"delta_db":-0.6698009742060591,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":7.761862069671956e-05,"delta_db":-0.9533102020546219,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":9.418418720939317e-06,"delta_db":-1.7484817182058148,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8521584046688747,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6346824503396798,"pitch":null,"output_rms_dbfs":-27.65434208966984,"output_peak":0.3469199240207672,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.960190555555227},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.344509912612676,"spectral_convergence":0.28453475708121856,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.037291196894736,"spectral_convergence":0.27341689024878846,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.498770557158271,"spectral_convergence":0.3144101111346818,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":6.873289488366938e-05,"delta_db":-0.7458393473679135,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.000491928772140715,"delta_db":-0.13800187735582847,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0007644887153885661,"delta_db":-0.857516308823011,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.00030400425353510553,"delta_db":-0.6698009742060591,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":7.761862069671956e-05,"delta_db":-0.9533102020546219,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":9.418418720939317e-06,"delta_db":-1.7484817182058148,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8521584046688747,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6346824503396798,"pitch":null,"output_rms_dbfs":-27.65434208966984,"output_peak":0.3469199240207672,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.960190555555227},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8784915235102968,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04456737119893152,"rms_dbfs":-27.01965963933016,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.1790027369579,"mean":2.3174285888671874e-07,"zero_energy":false},"output":{"rms":0.041426943833279745,"rms_dbfs":-27.65434208966984,"peak":0.3469199240207672,"peak_dbfs":-9.195415147938345,"crest_factor_db":18.458926941731498,"mean":1.3600091996342673e-07,"zero_energy":false},"rms_difference_db":-0.6346824503396782},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":7.4562429113247e-07,"output_rms_dbfs":-61.27479951870616,"output_zero_energy":false,"output_peak":0.00514734722673893,"output_energy_fraction":3.9825908920834734e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":305,"output_estimated_voiced_frames":292,"both_estimated_voiced_frames":249,"reference_voiced_output_unvoiced_frames":56,"reference_unvoiced_output_voiced_frames":43,"paired_voiced_fraction":0.8163934426229508,"absolute_difference_cents_median":4.052432211306331,"absolute_difference_cents_p95":21.414471406866085,"difference_over_600_cents_frames":6,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001511941088699898,"log_rms_mae_db":1.0953127472847828,"rms_envelope_nrmse":0.12171865880489312,"envelope_correlation":0.9882145517915898,"rms_gain_error_db":-0.6355632143191303,"reference_mean_power":0.0019872702200913737,"output_mean_power":0.001716724491888394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.92641346613152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000630498462430925,"output_power":0.0005510869020864281,"reference_active":true,"delta_db":-0.5846394272639468},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.948331074534803e-05,"output_power":7.25716999557781e-05,"reference_active":true,"delta_db":-0.39508652298410263},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.312350608813158e-05,"output_power":1.3882205224591382e-05,"reference_active":true,"delta_db":0.24408583352455793}],"active_band_count":3,"active_band_mae_db":0.40793726125753577}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013307215895407113,"log_rms_mae_db":0.9256490581456995,"rms_envelope_nrmse":0.11061904456154191,"envelope_correlation":0.9906715270575551,"rms_gain_error_db":-0.6406185158217597,"reference_mean_power":0.0019932283004473585,"output_mean_power":0.00171986830552743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.959638607230672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005794424432114469,"output_power":0.0005092427555414136,"reference_active":true,"delta_db":-0.5608544392604069},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.108494970400102e-05,"output_power":3.60882684018503e-05,"reference_active":true,"delta_db":-0.5631671553777549},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.934091733283845e-07,"output_power":7.968557270724059e-07,"reference_active":true,"delta_db":0.018824808568913557}],"active_band_count":3,"active_band_mae_db":0.3809488010690251}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014956359127228178,"log_rms_mae_db":1.3687663754022314,"rms_envelope_nrmse":0.12887184400591548,"envelope_correlation":0.9867255306009854,"rms_gain_error_db":-0.48831305007127374,"reference_mean_power":0.001265981682968443,"output_mean_power":0.001131347748945991},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.65319874591582e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005127917421071942,"output_power":0.0004442282955287371,"reference_active":true,"delta_db":-0.6233480489914781},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.49871434628785e-05,"output_power":7.177284399976621e-05,"reference_active":true,"delta_db":-0.19026654537958176},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1228212970044594e-05,"output_power":1.1605406873912336e-05,"reference_active":true,"delta_db":0.1434972950817209}],"active_band_count":3,"active_band_mae_db":0.31903729648426027}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012969159209010741,"log_rms_mae_db":1.097955580274644,"rms_envelope_nrmse":0.11520597401056835,"envelope_correlation":0.9891945389389739,"rms_gain_error_db":-0.4926279753661378,"reference_mean_power":0.0012704534730935044,"output_mean_power":0.0011342165167047224},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.90499086389905e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004798268169104881,"output_power":0.0004185170919645022,"reference_active":true,"delta_db":-0.5937131748205241},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.776254953569435e-05,"output_power":3.503085582517829e-05,"reference_active":true,"delta_db":-0.3261056044876959},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.708330714477494e-07,"output_power":7.70500662422354e-07,"reference_active":true,"delta_db":-0.0018732271218091985}],"active_band_count":3,"active_band_mae_db":0.3072306688100097}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":1.6297029858018808,"rms_envelope_nrmse":0.17609429375812008,"envelope_correlation":0.9745848061562166,"rms_gain_error_db":-0.7122816406365398,"reference_mean_power":0.0004068165018391945,"output_mean_power":0.00034527918427720443},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.83132236545664e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016437493457690725,"output_power":0.00014549571078033654,"reference_active":true,"delta_db":-0.5298540248865038},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7703924674415496e-05,"output_power":2.103431848911294e-05,"reference_active":true,"delta_db":-1.1961285225699574},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.687130809571587e-06,"output_power":4.430383121881271e-06,"reference_active":true,"delta_db":-0.24465790180445615}],"active_band_count":3,"active_band_mae_db":0.6568801497536391}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":1.4019028953754293,"rms_envelope_nrmse":0.15558223891910405,"envelope_correlation":0.9804691895189132,"rms_gain_error_db":-0.7216851772295587,"reference_mean_power":0.0004069857359569261,"output_mean_power":0.00034467570344431144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.5932421541601994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001538886928367015,"output_power":0.00013542831265201514,"reference_active":true,"delta_db":-0.5549724320172124},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4161267354242535e-05,"output_power":1.0945029505536018e-05,"reference_active":true,"delta_db":-1.1188518534708698},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.792056845121366e-07,"output_power":2.5388507653230715e-07,"reference_active":true,"delta_db":-0.4128708247866539}],"active_band_count":3,"active_band_mae_db":0.6955650367582454}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":1.7595865566211237,"rms_envelope_nrmse":0.2763655782979944,"envelope_correlation":0.941828167853176,"rms_gain_error_db":-1.3246364304120823,"reference_mean_power":1.611693883749463e-05,"output_mean_power":1.1880067678915452e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.125384681951734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2265606388859267e-06,"output_power":1.6677392236228439e-06,"reference_active":true,"delta_db":-1.255063842803571},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4345900244965858e-06,"output_power":1.0740729900822214e-06,"reference_active":true,"delta_db":-1.2569401118343129},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80760798762104e-07,"output_power":4.131882969621193e-07,"reference_active":true,"delta_db":-0.6578103588636854}],"active_band_count":3,"active_band_mae_db":1.0566047711671898}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":1.3602675827260085,"rms_envelope_nrmse":0.2476264366424238,"envelope_correlation":0.9541926843484063,"rms_gain_error_db":-1.324774095455528,"reference_mean_power":1.6196973858403443e-05,"output_mean_power":1.1938684398902798e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.254945521069739e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.488949340159725e-06,"output_power":1.7718334353365713e-06,"reference_active":true,"delta_db":-1.475931643408741},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.648300260916528e-07,"output_power":5.971932820109642e-07,"reference_active":true,"delta_db":-1.0745001577533013},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.810620519312689e-08,"output_power":3.98438151046594e-08,"reference_active":true,"delta_db":-0.8184018330869423}],"active_band_count":3,"active_band_mae_db":1.1229445447496615}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.7896311283111572,"predicted_mos":3.7896311283111572,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"e72422a8213fbf5adc177ee56472465526c7791604f61080305f4d1c5fa7cb18","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8874261060009303},"started_at":"2026-10-13T07:12:47.990021+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0073164660007023485,"native_input_samples":264600,"native_input_sha256":"6870fe1e50c52299edd63b802efa6629bc6e250453136a2717f44a034bd0f039","output_sha256":"9caec31f10d8ae7366918341973d06a215a9beaab361a230fc73e225f6c47c5b","output_samples":264704,"length_delta_samples":104,"analysis_s":0.2211774930001411,"synthesis_s":37.02228611499959,"total_s":37.24346360799973,"rtf":6.2072439346666215,"output_peak":0.2904132306575775,"output_rms":0.040910440073997155,"finished_at":"2026-10-13T07:13:25.273962+00:00","evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"9caec31f10d8ae7366918341973d06a215a9beaab361a230fc73e225f6c47c5b","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.757947441189774,"output_peak":0.28600848002735524},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.301713375616755,"spectral_convergence":0.4488064080962793,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.263905751431192,"spectral_convergence":0.3925423466262616,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.252378391844195,"spectral_convergence":0.43007759135410434,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":3.3477483235746636e-05,"delta_db":-3.869957693575459,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.0005214054560087347,"delta_db":0.11473159092461571,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0007551093057217963,"delta_db":-0.911128876556705,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.0002958687464242499,"delta_db":-0.7876066594013865,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":6.621483498468721e-05,"delta_db":-1.6434164007674155,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":4.147869857314567e-06,"delta_db":-5.3100104447752825,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.1061419443334777,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7370206022340371,"pitch":null,"output_rms_dbfs":-27.7566802415642,"output_peak":0.28600848002735524,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.272665839630713},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6698497705040249,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04456737119893152,"rms_dbfs":-27.01965963933016,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.1790027369579,"mean":2.3174285888671874e-07,"zero_energy":false},"output":{"rms":0.04094171095878297,"rms_dbfs":-27.7566802415642,"peak":0.28600848002735524,"peak_dbfs":-10.87242180074158,"crest_factor_db":16.88425844082262,"mean":-4.1612494523073174e-05,"zero_energy":false},"rms_difference_db":-0.7370206022340398},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":3.7953962694793584e-07,"output_rms_dbfs":-64.20742873612048,"output_zero_energy":false,"output_peak":0.003582460926556165,"output_energy_fraction":2.075566199691831e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":305,"output_estimated_voiced_frames":294,"both_estimated_voiced_frames":253,"reference_voiced_output_unvoiced_frames":52,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.8295081967213115,"absolute_difference_cents_median":5.887459020574062,"absolute_difference_cents_p95":1894.8416022567585,"difference_over_600_cents_frames":15,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001511941088699898,"log_rms_mae_db":1.607004712762515,"rms_envelope_nrmse":0.11995866773595262,"envelope_correlation":0.9909668510830952,"rms_gain_error_db":-0.7365560162865836,"reference_mean_power":0.0019872702200913737,"output_mean_power":0.0016772636034643293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.92641346613152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000630498462430925,"output_power":0.0005304815836178242,"reference_active":true,"delta_db":-0.7501372043381088},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.948331074534803e-05,"output_power":6.722921422466008e-05,"reference_active":true,"delta_db":-0.7271791338150262},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.312350608813158e-05,"output_power":9.3720878177979e-06,"reference_active":true,"delta_db":-1.4621352771513785}],"active_band_count":3,"active_band_mae_db":0.9798172051015044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013307215895407113,"log_rms_mae_db":1.378488972843585,"rms_envelope_nrmse":0.10232157356107224,"envelope_correlation":0.9953572946001282,"rms_gain_error_db":-0.7353960263477987,"reference_mean_power":0.0019932283004473585,"output_mean_power":0.0016827416420415573},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.959638607230672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005794424432114469,"output_power":0.00048771238960674585,"reference_active":true,"delta_db":-0.7484651505741968},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.108494970400102e-05,"output_power":3.38475656715291e-05,"reference_active":true,"delta_db":-0.8415532011173599},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.934091733283845e-07,"output_power":5.617270460164852e-07,"reference_active":true,"delta_db":-1.499718826656025}],"active_band_count":3,"active_band_mae_db":1.029912392782527}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014956359127228178,"log_rms_mae_db":1.9620915951809548,"rms_envelope_nrmse":0.14890908650161333,"envelope_correlation":0.9817290431044858,"rms_gain_error_db":-0.5341039738652305,"reference_mean_power":0.001265981682968443,"output_mean_power":0.0011194817671858084},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.65319874591582e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005127917421071942,"output_power":0.000396920953046667,"reference_active":true,"delta_db":-1.112369973090984},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.49871434628785e-05,"output_power":5.9548495282476605e-05,"reference_active":true,"delta_db":-1.0011601816712081},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1228212970044594e-05,"output_power":7.346797085749469e-06,"reference_active":true,"delta_db":-1.8421259642941616}],"active_band_count":3,"active_band_mae_db":1.3185520396854513}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012969159209010741,"log_rms_mae_db":1.4818590797715567,"rms_envelope_nrmse":0.12305011547740291,"envelope_correlation":0.9877024309513853,"rms_gain_error_db":-0.5334680559198061,"reference_mean_power":0.0012704534730935044,"output_mean_power":0.0011236005916911399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.90499086389905e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004798268169104881,"output_power":0.0003720955045983722,"reference_active":true,"delta_db":-1.104300932699136},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.776254953569435e-05,"output_power":2.9348270728385152e-05,"reference_active":true,"delta_db":-1.0947879132667548},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.708330714477494e-07,"output_power":4.910081518972242e-07,"reference_active":true,"delta_db":-1.958716366437522}],"active_band_count":3,"active_band_mae_db":1.3859350708011375}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":2.9183148109095223,"rms_envelope_nrmse":0.25058900828363906,"envelope_correlation":0.9475685462602277,"rms_gain_error_db":-0.9328112942368452,"reference_mean_power":0.0004068165018391945,"output_mean_power":0.00032818402086273767},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.83132236545664e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016437493457690725,"output_power":0.00014428504866443863,"reference_active":true,"delta_db":-0.5661426271054525},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7703924674415496e-05,"output_power":2.368481716264507e-05,"reference_active":true,"delta_db":-0.6807126122155838},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.687130809571587e-06,"output_power":3.950368046816068e-06,"reference_active":true,"delta_db":-0.7426951434231004}],"active_band_count":3,"active_band_mae_db":0.6631834609147123}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":2.5330176897310808,"rms_envelope_nrmse":0.2258612418748025,"envelope_correlation":0.9566021581913239,"rms_gain_error_db":-0.9301519958196557,"reference_mean_power":0.0004069857359569261,"output_mean_power":0.00032852164496560915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.5932421541601994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001538886928367015,"output_power":0.00013276918795918769,"reference_active":true,"delta_db":-0.6410941163734973},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4161267354242535e-05,"output_power":1.2264929694137105e-05,"reference_active":true,"delta_db":-0.6243705896350492},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.792056845121366e-07,"output_power":2.3678314771546474e-07,"reference_active":true,"delta_db":-0.7157346646933154}],"active_band_count":3,"active_band_mae_db":0.6603997902339539}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":3.905672497870959,"rms_envelope_nrmse":0.3821151881789955,"envelope_correlation":0.8797885899627328,"rms_gain_error_db":-1.8313377838752967,"reference_mean_power":1.611693883749463e-05,"output_mean_power":1.0571796128834623e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.125384681951734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2265606388859267e-06,"output_power":1.8930864344487556e-06,"reference_active":true,"delta_db":-0.7046408393438154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4345900244965858e-06,"output_power":1.0076114202221704e-06,"reference_active":true,"delta_db":-1.5343472549056616},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80760798762104e-07,"output_power":2.780529752047059e-07,"reference_active":true,"delta_db":-2.378015016162874}],"active_band_count":3,"active_band_mae_db":1.5390010368041172}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":3.578704696915654,"rms_envelope_nrmse":0.3447428493972169,"envelope_correlation":0.9000977279006754,"rms_gain_error_db":-1.8315693014812289,"reference_mean_power":1.6196973858403443e-05,"output_mean_power":1.0623728200692432e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.254945521069739e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.488949340159725e-06,"output_power":1.891783184023683e-06,"reference_active":true,"delta_db":-1.1914469638516854},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.648300260916528e-07,"output_power":5.565740361267426e-07,"reference_active":true,"delta_db":-1.3804198648811838},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.810620519312689e-08,"output_power":2.4326265934547204e-08,"reference_active":true,"delta_db":-2.961256492590849}],"active_band_count":3,"active_band_mae_db":1.8443744404412392}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"golf","status":"measured","mos":{"status":"measured","value":2.98215389251709,"predicted_mos":2.98215389251709,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"dd90357ab65f12ac1f612530da2387c1b5acdb30700cff07ed036ad29e486104","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.46503252399998},"started_at":"2026-10-13T05:02:26.919100+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003705933000674122,"native_input_samples":144000,"native_input_sha256":"b4b699958c0739faff93223b95d862606c6aceae697b0564d1ae7131a9af91de","output_sha256":"279e43c41a386f6dd9c0e01a27fb043a97279e797e9a97622e01bcaf2ad92c21","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5835865040007775,"synthesis_s":0.1395878690000245,"total_s":0.723174373000802,"rtf":0.12052906216680033,"output_peak":0.3056015968322754,"output_rms":0.03710545856026476,"finished_at":"2026-10-13T05:02:27.663922+00:00","analysis_stages":{"external_f0_s":0.1031073069998456},"evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"279e43c41a386f6dd9c0e01a27fb043a97279e797e9a97622e01bcaf2ad92c21","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.607174082818542,"output_peak":0.30532546002055916},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.013684785348367,"spectral_convergence":0.5193680657341319,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.369383224170941,"spectral_convergence":0.357095467353719,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.920414303516883,"spectral_convergence":0.41871798787954917,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.165566259371106e-05,"output_power":6.957307763343318e-05,"delta_db":-0.6954509200910347,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005085027886334004,"output_power":0.0003721508163617045,"delta_db":-1.3557436292880705,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009318492498597469,"output_power":0.000635520729065079,"delta_db":-1.6621593924113032,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00035447813757834735,"output_power":0.00023951132295604646,"delta_db":-1.7026340564111264,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.680655966818438e-05,"output_power":5.5055349380085823e-05,"delta_db":-2.4510526334352214,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4123514362995673e-05,"output_power":6.2946035799626845e-06,"delta_db":-3.509743915219876,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.8961307578094386,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.590061710571565,"pitch":null,"output_rms_dbfs":-28.607174082818542,"output_peak":0.30532546002055916,"reference_rms_dbfs":-27.01711237224698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.76782743767873},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8123356477895592,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04458044316397478,"rms_dbfs":-27.01711237224698,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.176455469874718,"mean":1.1087458997417572e-06,"zero_energy":false},"output":{"rms":0.03712284874543321,"rms_dbfs":-28.607174082818545,"peak":0.30532546002055916,"peak_dbfs":-10.304739597916917,"crest_factor_db":18.302434484901625,"mean":0.00031987211294248004,"zero_energy":false},"rms_difference_db":-1.5900617105715646},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":1.0155254491119455e-06,"output_rms_dbfs":-59.93309188684187,"output_zero_energy":false,"output_peak":0.004294135143231901,"output_energy_fraction":6.766188406890818e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":304,"output_estimated_voiced_frames":241,"both_estimated_voiced_frames":216,"reference_voiced_output_unvoiced_frames":88,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.7105263157894737,"absolute_difference_cents_median":6.271074429241695,"absolute_difference_cents_p95":24.88448786986813,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001511941088699898,"log_rms_mae_db":2.3742782798088395,"rms_envelope_nrmse":0.20628772081282037,"envelope_correlation":0.9818321685231401,"rms_gain_error_db":-1.5907140044923205,"reference_mean_power":0.0019882734160185244,"output_mean_power":0.0013784934453973314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.938116266535944e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006311708849697412,"output_power":0.00045298464679701615,"reference_active":true,"delta_db":-1.4406347473322652},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.965612642881732e-05,"output_power":5.802112533779133e-05,"reference_active":true,"delta_db":-1.376330356003686},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.311152321574422e-05,"output_power":1.3578507486505475e-05,"reference_active":true,"delta_db":0.15198887801162927}],"active_band_count":3,"active_band_mae_db":0.9896513271158601}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013307215895407113,"log_rms_mae_db":2.1462822890334996,"rms_envelope_nrmse":0.1884411346148717,"envelope_correlation":0.9902865266309909,"rms_gain_error_db":-1.5920728934295578,"reference_mean_power":0.0019938766647867837,"output_mean_power":0.0013819457713535644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.970068415782102e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005799846035824634,"output_power":0.0004172508133695857,"reference_active":true,"delta_db":-1.4301927287928957},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.114199148941758e-05,"output_power":2.9746129140315896e-05,"reference_active":true,"delta_db":-1.408548509510816},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.932448696666293e-07,"output_power":7.723572160616914e-07,"reference_active":true,"delta_db":-0.11589063494975411}],"active_band_count":3,"active_band_mae_db":0.9848772910844885}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014956359127228178,"log_rms_mae_db":2.620531395300562,"rms_envelope_nrmse":0.21648291102764347,"envelope_correlation":0.9784288450978069,"rms_gain_error_db":-1.5546378071088938,"reference_mean_power":0.0012668905424660065,"output_mean_power":0.0008856768912372032},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.662605100193672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005133129746147998,"output_power":0.00035064221780336396,"reference_active":true,"delta_db":-1.655180371685533},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.513510966190017e-05,"output_power":4.867933601355556e-05,"reference_active":true,"delta_db":-1.8849827879220817},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1227974215731494e-05,"output_power":9.682322519446427e-06,"reference_active":true,"delta_db":-0.6432186168364873}],"active_band_count":3,"active_band_mae_db":1.3944605921480342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012969159209010741,"log_rms_mae_db":2.248506717961221,"rms_envelope_nrmse":0.1964898379458973,"envelope_correlation":0.9865003240225804,"rms_gain_error_db":-1.5544090224812959,"reference_mean_power":0.001271312110010161,"output_mean_power":0.0008888148083468914},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.913095710357004e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004802509209803863,"output_power":0.00032796628220835034,"reference_active":true,"delta_db":-1.6563900963436196},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.782216446577361e-05,"output_power":2.4306362441724468e-05,"reference_active":true,"delta_db":-1.920264092318103},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.716012687738436e-07,"output_power":6.11750147342142e-07,"reference_active":true,"delta_db":-1.0081885047449732}],"active_band_count":3,"active_band_mae_db":1.528280897802232}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":2.559224715658688,"rms_envelope_nrmse":0.2649257162869069,"envelope_correlation":0.9639177888098055,"rms_gain_error_db":-1.9361425738590754,"reference_mean_power":0.00040647000772115323,"output_mean_power":0.00026026408880144934},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.834720877880788e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016463339733868064,"output_power":0.00010418203423119453,"reference_active":true,"delta_db":-1.987251071846405},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7699764779019658e-05,"output_power":1.6520728640912947e-05,"reference_active":true,"delta_db":-2.244468833245737},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679939847235544e-06,"output_power":4.682656603287919e-06,"reference_active":true,"delta_db":0.002520395280462083}],"active_band_count":3,"active_band_mae_db":1.4114134334575346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":2.219468792116738,"rms_envelope_nrmse":0.23501758368110184,"envelope_correlation":0.9807136199513147,"rms_gain_error_db":-1.9441800016101964,"reference_mean_power":0.000406438099280292,"output_mean_power":0.0002597624737904679},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.5966618776268534e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001541267072298333,"output_power":9.690171415595921e-05,"reference_active":true,"delta_db":-2.0154644059147024},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4122585218746364e-05,"output_power":9.082217337515407e-06,"reference_active":true,"delta_db":-1.9172231368463866},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7954958410543116e-07,"output_power":2.437256052900221e-07,"reference_active":true,"delta_db":-0.5955769285225658}],"active_band_count":3,"active_band_mae_db":1.5094214904278849}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":2.8094387520667663,"rms_envelope_nrmse":0.3037549066023312,"envelope_correlation":0.9314704303010038,"rms_gain_error_db":-1.3394069505777901,"reference_mean_power":1.6144484440307693e-05,"output_mean_power":1.1859967139049167e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.122204979047956e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2272076587708106e-06,"output_power":1.5597921107147864e-06,"reference_active":true,"delta_db":-1.5469399196624538},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4298676713687177e-06,"output_power":1.019596570789711e-06,"reference_active":true,"delta_db":-1.4686748094137836},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.801940103957737e-07,"output_power":5.185302886939174e-07,"reference_active":true,"delta_db":0.33357391059618097}],"active_band_count":3,"active_band_mae_db":1.1163962132241394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":2.537191309524328,"rms_envelope_nrmse":0.24004647308442517,"envelope_correlation":0.9622176001864595,"rms_gain_error_db":-1.339788231109008,"reference_mean_power":1.622384800399466e-05,"output_mean_power":1.1917222442344304e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.2521500978776355e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.487879804979646e-06,"output_power":1.8313847148805608e-06,"reference_active":true,"delta_db":-1.3304980963746633},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.612426501666122e-07,"output_power":5.956312747599301e-07,"reference_active":true,"delta_db":-1.0654561938593785},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7971424538031036e-08,"output_power":4.449088573221677e-08,"reference_active":true,"delta_db":-0.32711563429193097}],"active_band_count":3,"active_band_mae_db":0.907689974841991}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4606988430023193,"predicted_mos":3.4606988430023193,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"44c9e976e6bf1f36d33f900d49de1473089455e54bfa4ad92a399c199ff4322d","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":11.525726576001034},"started_at":"2026-10-13T07:36:36.809475+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003156924998620525,"native_input_samples":144000,"native_input_sha256":"b4b699958c0739faff93223b95d862606c6aceae697b0564d1ae7131a9af91de","output_sha256":"de033b494525c5e9e4a2f5959dc47912e524e79da14acf7c1d4da0736fe228ee","output_samples":144120,"length_delta_samples":120,"analysis_s":1.9360827559976315,"synthesis_s":3.3125532850026502,"total_s":5.248636041000282,"rtf":0.874772673500047,"output_peak":0.3426022231578827,"output_rms":0.0435384086789291,"finished_at":"2026-10-13T07:36:42.072782+00:00","analysis_stages":{"external_f0_s":1.0662541819983744},"evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"de033b494525c5e9e4a2f5959dc47912e524e79da14acf7c1d4da0736fe228ee","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.21905380559578,"output_peak":0.33726356201688046},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.759998271960274,"spectral_convergence":0.4140196239642823,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.145337895181667,"spectral_convergence":0.24028915633608222,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.615907156656153,"spectral_convergence":0.26420788547145874,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":8.555620155153743e-05,"delta_db":0.20502924312735632,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.00048420340947920683,"delta_db":-0.20674569073440247,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0008729954990840608,"delta_db":-0.28110705448002765,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.0003499771414530559,"delta_db":-0.058200779950970546,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":9.832283768317377e-05,"delta_db":0.07357463027236742,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":7.437151917275123e-06,"delta_db":-2.774195132485502,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5998087551751045,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19625209479065905,"pitch":null,"output_rms_dbfs":-27.215911734120823,"output_peak":0.33726356201688046,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.507081107932699},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8981092756205594,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04456737119893152,"rms_dbfs":-27.01965963933016,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.1790027369579,"mean":2.3174285888671874e-07,"zero_energy":false},"output":{"rms":0.04357169082292205,"rms_dbfs":-27.215911734120823,"peak":0.33726356201688046,"peak_dbfs":-9.440611549760405,"crest_factor_db":17.775300184360418,"mean":-3.397090395917585e-05,"zero_energy":false},"rms_difference_db":-0.19625209479066186},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":5.028492088489149e-07,"output_rms_dbfs":-62.985622288265326,"output_zero_energy":false,"output_peak":0.0029185021776828397,"output_energy_fraction":2.4279536050557952e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":305,"output_estimated_voiced_frames":292,"both_estimated_voiced_frames":253,"reference_voiced_output_unvoiced_frames":52,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.8295081967213115,"absolute_difference_cents_median":4.283797083616085,"absolute_difference_cents_p95":17.567376249588904,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001511941088699898,"log_rms_mae_db":1.5775688957057725,"rms_envelope_nrmse":0.12953541376064062,"envelope_correlation":0.9813766008980842,"rms_gain_error_db":-0.19522365013607793,"reference_mean_power":0.0019872702200913737,"output_mean_power":0.0018999166972711564},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.92641346613152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000630498462430925,"output_power":0.0006021598363857097,"reference_active":true,"delta_db":-0.19972246792359036},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.948331074534803e-05,"output_power":7.454143794827769e-05,"reference_active":true,"delta_db":-0.2787818233496481},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.312350608813158e-05,"output_power":1.1630539227132422e-05,"reference_active":true,"delta_db":-0.5245002665348343}],"active_band_count":3,"active_band_mae_db":0.33433485260269097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013307215895407113,"log_rms_mae_db":1.078893411177232,"rms_envelope_nrmse":0.08494015933583146,"envelope_correlation":0.9916793450641095,"rms_gain_error_db":-0.19490771562786297,"reference_mean_power":0.0019932283004473585,"output_mean_power":0.0019057515129473567},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.959638607230672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005794424432114469,"output_power":0.0005536190780120191,"reference_active":true,"delta_db":-0.19799255664839333},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.108494970400102e-05,"output_power":3.865620496670307e-05,"reference_active":true,"delta_db":-0.2646354427902402},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.934091733283845e-07,"output_power":7.137375088387735e-07,"reference_active":true,"delta_db":-0.45958696734790094}],"active_band_count":3,"active_band_mae_db":0.3074049889288448}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014956359127228178,"log_rms_mae_db":1.8135173582830542,"rms_envelope_nrmse":0.14671574922137468,"envelope_correlation":0.9802044532208007,"rms_gain_error_db":-0.2309670613993981,"reference_mean_power":0.001265981682968443,"output_mean_power":0.0012004130716454719},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.65319874591582e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005127917421071942,"output_power":0.00047739174877367075,"reference_active":true,"delta_db":-0.3106611437671698},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.49871434628785e-05,"output_power":6.664716451578232e-05,"reference_active":true,"delta_db":-0.5120513282448181},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1228212970044594e-05,"output_power":1.0499623900280513e-05,"reference_active":true,"delta_db":-0.29136898690940694}],"active_band_count":3,"active_band_mae_db":0.3713604863071316}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012969159209010741,"log_rms_mae_db":1.1309516193087266,"rms_envelope_nrmse":0.09803249658022949,"envelope_correlation":0.990822989335388,"rms_gain_error_db":-0.23119045566617105,"reference_mean_power":0.0012704534730935044,"output_mean_power":0.001204591291800206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.90499086389905e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004798268169104881,"output_power":0.00044585830529267757,"reference_active":true,"delta_db":-0.3188765553725801},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.776254953569435e-05,"output_power":3.398521940964788e-05,"reference_active":true,"delta_db":-0.4577122988900511},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.708330714477494e-07,"output_power":7.10536682524707e-07,"reference_active":true,"delta_db":-0.35373835194702674}],"active_band_count":3,"active_band_mae_db":0.3767757354032193}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":1.7484102900919476,"rms_envelope_nrmse":0.16384514009533255,"envelope_correlation":0.9735525687034624,"rms_gain_error_db":-0.025898876564486778,"reference_mean_power":0.0004068165018391945,"output_mean_power":0.00040439769674158193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.83132236545664e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016437493457690725,"output_power":0.0001636258315612013,"reference_active":true,"delta_db":-0.01983726291664677},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7703924674415496e-05,"output_power":2.8326861909217198e-05,"reference_active":true,"delta_db":0.09657167695522034},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.687130809571587e-06,"output_power":4.061627643058189e-06,"reference_active":true,"delta_db":-0.6220696790971124}],"active_band_count":3,"active_band_mae_db":0.2461595396563265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":1.0882838305946887,"rms_envelope_nrmse":0.09989440846875133,"envelope_correlation":0.9893812237484304,"rms_gain_error_db":-0.023323816787025917,"reference_mean_power":0.0004069857359569261,"output_mean_power":0.00040480587481668377},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.5932421541601994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001538886928367015,"output_power":0.00015330738558997736,"reference_active":true,"delta_db":-0.016436331620563016},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4161267354242535e-05,"output_power":1.4316913278216729e-05,"reference_active":true,"delta_db":0.0474727229623098},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.792056845121366e-07,"output_power":2.3836954271436677e-07,"reference_active":true,"delta_db":-0.6867349276411636}],"active_band_count":3,"active_band_mae_db":0.2502146607413455}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":1.8968517564427718,"rms_envelope_nrmse":0.23539588508454903,"envelope_correlation":0.9523196228559468,"rms_gain_error_db":-0.6883186991477254,"reference_mean_power":1.611693883749463e-05,"output_mean_power":1.375468623362246e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.125384681951734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2265606388859267e-06,"output_power":1.9383018176364642e-06,"reference_active":true,"delta_db":-0.6021312432897905},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4345900244965858e-06,"output_power":1.0939369795517674e-06,"reference_active":true,"delta_db":-1.1773550310488359},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80760798762104e-07,"output_power":2.850969845558794e-07,"reference_active":true,"delta_db":-2.2693642422250933}],"active_band_count":3,"active_band_mae_db":1.3496168388545733}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":1.149396254496701,"rms_envelope_nrmse":0.14822595279293801,"envelope_correlation":0.9833405452057764,"rms_gain_error_db":-0.6883801100177862,"reference_mean_power":1.6196973858403443e-05,"output_mean_power":1.3822795095987876e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.254945521069739e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.488949340159725e-06,"output_power":2.0555723280408476e-06,"reference_active":true,"delta_db":-0.8308329449657049},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.648300260916528e-07,"output_power":5.487220658077555e-07,"reference_active":true,"delta_db":-1.442125044908087},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.810620519312689e-08,"output_power":2.5571710579402792e-08,"reference_active":true,"delta_db":-2.7444131894997037}],"active_band_count":3,"active_band_mae_db":1.6724570597911652}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.3613665103912354,"predicted_mos":3.3613665103912354,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"a888690a35234f80dadeb7db8e26f868157fbaf57d41663ac3e6285c2e2e5923","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.529198191001342},"started_at":"2026-10-13T05:10:35.798292+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027066189995821333,"native_input_samples":144000,"native_input_sha256":"b4b699958c0739faff93223b95d862606c6aceae697b0564d1ae7131a9af91de","output_sha256":"7fd8a9dfe0e0609d26c04d4c7f6b0555c3880ad30a1ae693359c9b7448274013","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9696787490011047,"synthesis_s":0.1869580570000835,"total_s":1.1566368060011882,"rtf":0.19277280100019803,"output_peak":0.3157667815685272,"output_rms":0.04356389448600073,"finished_at":"2026-10-13T05:10:36.969351+00:00","analysis_stages":{"external_f0_s":0.15874283499942976},"evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"7fd8a9dfe0e0609d26c04d4c7f6b0555c3880ad30a1ae693359c9b7448274013","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.214275683094606,"output_peak":0.3120587122846276},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.2139708502026245,"spectral_convergence":0.15797737723546618,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.289644772940407,"spectral_convergence":0.17315487744414376,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.041579849546111,"spectral_convergence":0.1994798558565397,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.1610975793667e-05,"output_power":8.018154759223296e-05,"delta_db":-0.07674136235477456,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005078113602376632,"output_power":0.0005039677145999556,"delta_db":-0.032996968657392496,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000931370997501075,"output_power":0.0008900958264820337,"delta_db":-0.19685945357367993,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000354698831360673,"output_power":0.00032951321675321865,"delta_db":-0.319869187860621,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.667116250090391e-05,"output_power":8.928081703385783e-05,"delta_db":-0.3453878518840084,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4087248189368409e-05,"output_power":9.260091473451705e-06,"delta_db":-1.8221088930629705,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4656606195655744,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.18755207618841988,"pitch":null,"output_rms_dbfs":-27.207211715518582,"output_peak":0.3120587122846276,"reference_rms_dbfs":-27.01965963933016,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.515065157563047},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9382173047866826,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04456737119893152,"rms_dbfs":-27.01965963933016,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.1790027369579,"mean":2.3174285888671874e-07,"zero_energy":false},"output":{"rms":0.043615355254024366,"rms_dbfs":-27.207211715518582,"peak":0.3120587122846276,"peak_dbfs":-10.115273759221973,"crest_factor_db":17.09193795629661,"mean":8.289165068186596e-06,"zero_energy":false},"rms_difference_db":-0.18755207618842107},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":8.354990057247199e-07,"output_rms_dbfs":-60.7805406259707,"output_zero_energy":false,"output_peak":0.005087236222053854,"output_energy_fraction":4.0260442887785525e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":305,"output_estimated_voiced_frames":289,"both_estimated_voiced_frames":268,"reference_voiced_output_unvoiced_frames":37,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.8786885245901639,"absolute_difference_cents_median":0.8423608999095984,"absolute_difference_cents_p95":13.927766184227954,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001511941088699898,"log_rms_mae_db":0.553457603127095,"rms_envelope_nrmse":0.052797293998800225,"envelope_correlation":0.9973323644145833,"rms_gain_error_db":-0.1870313528059503,"reference_mean_power":0.0019872702200913737,"output_mean_power":0.0019035039802232318},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.92641346613152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000630498462430925,"output_power":0.000605856962100131,"reference_active":true,"delta_db":-0.17313928941389223},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.948331074534803e-05,"output_power":6.865787639759861e-05,"reference_active":true,"delta_db":-0.6358558215136507},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.312350608813158e-05,"output_power":9.41276557908867e-06,"reference_active":true,"delta_db":-1.4433263415638575}],"active_band_count":3,"active_band_mae_db":0.7507738174971336}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013307215895407113,"log_rms_mae_db":0.36396783801478755,"rms_envelope_nrmse":0.03642464973522826,"envelope_correlation":0.9989344281329346,"rms_gain_error_db":-0.18652814037516352,"reference_mean_power":0.0019932283004473585,"output_mean_power":0.0019094321501672527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.959638607230672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005794424432114469,"output_power":0.0005574043729983092,"reference_active":true,"delta_db":-0.16839931837191002},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.108494970400102e-05,"output_power":3.511800272355517e-05,"reference_active":true,"delta_db":-0.681529514694722},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.934091733283845e-07,"output_power":5.35254690249739e-07,"reference_active":true,"delta_db":-1.709367358886005}],"active_band_count":3,"active_band_mae_db":0.853098730650879}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014956359127228178,"log_rms_mae_db":0.740661834586327,"rms_envelope_nrmse":0.05437908182036426,"envelope_correlation":0.9974669482184813,"rms_gain_error_db":-0.14089510758392013,"reference_mean_power":0.001265981682968443,"output_mean_power":0.0012255694066407937},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.65319874591582e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005127917421071942,"output_power":0.0004938331612586611,"reference_active":true,"delta_db":-0.16360773010175977},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.49871434628785e-05,"output_power":6.673489857672011e-05,"reference_active":true,"delta_db":-0.5063380526885854},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1228212970044594e-05,"output_power":8.034759890212567e-06,"reference_active":true,"delta_db":-1.4533773858934576}],"active_band_count":3,"active_band_mae_db":0.7077743895612677}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012969159209010741,"log_rms_mae_db":0.41317105099912393,"rms_envelope_nrmse":0.031965460055875486,"envelope_correlation":0.9992239221352552,"rms_gain_error_db":-0.14027915505416103,"reference_mean_power":0.0012704534730935044,"output_mean_power":0.001230072896453088},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.90499086389905e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004798268169104881,"output_power":0.0004614431750542176,"reference_active":true,"delta_db":-0.16966289510631777},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.776254953569435e-05,"output_power":3.3301271080700095e-05,"reference_active":true,"delta_db":-0.546004975079564},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.708330714477494e-07,"output_power":5.186229774436332e-07,"reference_active":true,"delta_db":-1.721085849894846}],"active_band_count":3,"active_band_mae_db":0.8122512400269093}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":0.8467652533372211,"rms_envelope_nrmse":0.08690535728914793,"envelope_correlation":0.9937266695296134,"rms_gain_error_db":-0.332811273793902,"reference_mean_power":0.0004068165018391945,"output_mean_power":0.000376805673844646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.83132236545664e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016437493457690725,"output_power":0.00015146994315622507,"reference_active":true,"delta_db":-0.3550913054696614},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7703924674415496e-05,"output_power":2.1155754192838684e-05,"reference_active":true,"delta_db":-1.1711278537762506},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.687130809571587e-06,"output_power":4.219642532736612e-06,"reference_active":true,"delta_db":-0.45631412846381286}],"active_band_count":3,"active_band_mae_db":0.6608444292365749}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":0.5546867263934282,"rms_envelope_nrmse":0.05813881347279034,"envelope_correlation":0.9978335387523433,"rms_gain_error_db":-0.3330042665118566,"reference_mean_power":0.0004069857359569261,"output_mean_power":0.00037694567240682526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.5932421541601994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001538886928367015,"output_power":0.0001407575295659039,"reference_active":true,"delta_db":-0.387350747443715},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4161267354242535e-05,"output_power":1.095426371886893e-05,"reference_active":true,"delta_db":-1.1151892986187182},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.792056845121366e-07,"output_power":2.3114267033187753e-07,"reference_active":true,"delta_db":-0.82044129834405}],"active_band_count":3,"active_band_mae_db":0.7743271148021611}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":1.049006531804928,"rms_envelope_nrmse":0.14525030602070513,"envelope_correlation":0.9833403781194424,"rms_gain_error_db":-0.5102439114745201,"reference_mean_power":1.611693883749463e-05,"output_mean_power":1.4330395174569239e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.125384681951734e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2265606388859267e-06,"output_power":1.949494866931723e-06,"reference_active":true,"delta_db":-0.577124313299119},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4345900244965858e-06,"output_power":1.095825536731934e-06,"reference_active":true,"delta_db":-1.1698638975641487},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80760798762104e-07,"output_power":2.7150901412797026e-07,"reference_active":true,"delta_db":-2.4814479527724713}],"active_band_count":3,"active_band_mae_db":1.409478721211913}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":0.6650632785687396,"rms_envelope_nrmse":0.09576915175568483,"envelope_correlation":0.9939742231110935,"rms_gain_error_db":-0.5102300209611492,"reference_mean_power":1.6196973858403443e-05,"output_mean_power":1.4401604469944396e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.254945521069739e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.488949340159725e-06,"output_power":2.0798619859763095e-06,"reference_active":true,"delta_db":-0.7798153979951126},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.648300260916528e-07,"output_power":5.395788929888221e-07,"reference_active":true,"delta_db":-1.515099765689928},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.810620519312689e-08,"output_power":2.6327465373105543e-08,"reference_active":true,"delta_db":-2.617920491556733}],"active_band_count":3,"active_band_mae_db":1.637611885080591}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0007","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2004904747009277,"predicted_mos":3.2004904747009277,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287102,"input_sample_rate":48000,"input_seconds":5.9812916666666665,"input_float32_sha256":"02ea5c9f48a96aaf6e7bec8d8945aca03b71ee2180669ed6f136bfa56db8c846","internal_samples":95701,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4444397739989654},"started_at":"2026-10-13T05:14:58.829465+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008643114999358659,"native_input_samples":288000,"native_input_sha256":"e189fda506790e657868c8f73493f5fe6b19e3ad3eb842d3c3f17d76ee4fbd95","output_sha256":"ddc3cfda00928950dc5f696b39c5dc64dafd7f1376d2f0dfc2709126a150593e","output_samples":287102,"length_delta_samples":-898,"analysis_s":3.438207126000634,"synthesis_s":1.0372559459992772,"total_s":4.475463071999911,"rtf":0.7459105119999853,"output_peak":0.33610515482856695,"output_rms":0.04308065147934936,"finished_at":"2026-10-13T05:15:03.325517+00:00","evaluation_input":{"reference":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"ddc3cfda00928950dc5f696b39c5dc64dafd7f1376d2f0dfc2709126a150593e","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95701,"length_delta_samples":-299,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.31264736629981,"output_peak":0.33672180572691657},"common_prefix":{"samples":95701,"sample_rate":16000,"reference_coverage":0.9968854166666666,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95701,"output_samples":95701,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.249634535708457,"spectral_convergence":0.4229819272831228,"spectral_convergence_status":"ok","amplitude_floor":1.0845908458793198e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.420752905177837,"spectral_convergence":0.18931417740344736,"spectral_convergence_status":"ok","amplitude_floor":9.938577161724507e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.746874903905795,"spectral_convergence":0.1894790166464782,"spectral_convergence_status":"ok","amplitude_floor":7.780261052298239e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.329480939844153e-05,"output_power":7.306787371274917e-05,"delta_db":-0.5689146925590678,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005074795986167812,"output_power":0.0004664545157528536,"delta_db":-0.3660928585281703,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000932330601883786,"output_power":0.000853384841686678,"delta_db":-0.384250148979592,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003532699063396942,"output_power":0.00035001601225611786,"delta_db":-0.040187306226828826,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.68071901752544e-05,"output_power":0.00010270975135821795,"delta_db":0.2570406294521133,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.4130832234058335e-05,"output_power":1.1039327550293531e-05,"delta_db":-1.0722512075460424,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.44812280721530245,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2953099670753013,"pitch":null,"output_rms_dbfs":-27.31264736629981,"output_peak":0.33672180572691657,"reference_rms_dbfs":-27.01733739922451,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.805754114930696},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95701,"output_samples":95701,"intelligibility":{"estoi":{"value":0.8999270389578086,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044579288225004406,"rms_dbfs":-27.01733739922451,"peak":0.32208251953125,"peak_dbfs":-9.840656902372261,"crest_factor_db":17.17668049685225,"mean":-3.791538269257897e-07,"zero_energy":false},"output":{"rms":0.04308912069556434,"rms_dbfs":-27.31264736629981,"peak":0.33672180572691657,"peak_dbfs":-9.454575165542037,"crest_factor_db":17.858072200757775,"mean":-0.0003890229968795371,"zero_energy":false},"rms_difference_db":-0.2953099670753012},"full_scale":{"reference":{"sample_count":95701,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95701,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":21,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.2486721462337304e-06,"threshold_dbfs":-56.4807385957586,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":8.650845848023892e-07,"output_mean_power":6.216345717922097e-07,"output_rms_dbfs":-62.06464840537463,"output_zero_energy":false,"output_peak":0.0031892998281126427,"output_energy_fraction":3.078690931201005e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":303,"output_estimated_voiced_frames":248,"both_estimated_voiced_frames":237,"reference_voiced_output_unvoiced_frames":66,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.7821782178217822,"absolute_difference_cents_median":2.295092801245905,"absolute_difference_cents_p95":9.796340173331359,"difference_over_600_cents_frames":0,"uncovered_tail_samples":21,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95701,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1195,"rms_floor":0.0001511941088699898,"log_rms_mae_db":1.2896661820352084,"rms_envelope_nrmse":0.10294290883136313,"envelope_correlation":0.9892246468691359,"rms_gain_error_db":-0.2950569177324814,"reference_mean_power":0.0019884614133866404,"output_mean_power":0.001857853547371438},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":9.950920155725402e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006318155617085767,"output_power":0.0006123998263415487,"reference_active":true,"delta_db":-0.13555259563514663},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.98354515500034e-05,"output_power":7.347066167628746e-05,"reference_active":true,"delta_db":-0.3608183489202374},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3079024220309224e-05,"output_power":1.1839703692564986e-05,"reference_active":true,"delta_db":-0.4323451041113163}],"active_band_count":3,"active_band_mae_db":0.30957201622223346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1189,"rms_floor":0.00013307215895407113,"log_rms_mae_db":0.8682881635599409,"rms_envelope_nrmse":0.07010225048666056,"envelope_correlation":0.995383498301472,"rms_gain_error_db":-0.29519163093489764,"reference_mean_power":0.001994560799134356,"output_mean_power":0.0018634945035427906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":8.980446111512012e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005805025115770793,"output_power":0.0005628280392542986,"reference_active":true,"delta_db":-0.13428377872746777},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.118789181705957e-05,"output_power":3.701010958054402e-05,"reference_active":true,"delta_db":-0.46449192397775807},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.011270008376816e-07,"output_power":7.076806190199393e-07,"reference_active":true,"delta_db":-0.5386406734615375}],"active_band_count":3,"active_band_mae_db":0.3791387920555878}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1195,"rms_floor":0.00014956359127228178,"log_rms_mae_db":1.475677194855195,"rms_envelope_nrmse":0.11089269919471711,"envelope_correlation":0.990146140062049,"rms_gain_error_db":-0.39943210410410795,"reference_mean_power":0.0012677568349194362,"output_mean_power":0.0011563591740638398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":7.672067041117839e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005138079573093241,"output_power":0.00047721561896258563,"reference_active":true,"delta_db":-0.32086176459793114},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.529671441426252e-05,"output_power":6.569761735701197e-05,"reference_active":true,"delta_db":-0.5922640672591641},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.121498373638507e-05,"output_power":9.386599658365621e-06,"reference_active":true,"delta_db":-0.7729035266427613}],"active_band_count":3,"active_band_mae_db":0.5620097861666188}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1189,"rms_floor":0.00012969159209010741,"log_rms_mae_db":1.045384536249095,"rms_envelope_nrmse":0.08051660261263839,"envelope_correlation":0.9953777657219448,"rms_gain_error_db":-0.3993171518876982,"reference_mean_power":0.001272232815003435,"output_mean_power":0.0011604725660685738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":6.921082023511817e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004806548272562197,"output_power":0.0004452422913254407,"reference_active":true,"delta_db":-0.3323689925540424},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.787907607749976e-05,"output_power":3.2162771660476275e-05,"reference_active":true,"delta_db":-0.7104590995353705},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.749815761337162e-07,"output_power":6.917173508744073e-07,"reference_active":true,"delta_db":-0.4936270846246594}],"active_band_count":3,"active_band_mae_db":0.5121517255713574}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1195,"rms_floor":7.983632889928635e-05,"log_rms_mae_db":1.389182603725116,"rms_envelope_nrmse":0.13491964696734346,"envelope_correlation":0.9824371082420845,"rms_gain_error_db":0.02500018372942095,"reference_mean_power":0.00040551884108399136,"output_mean_power":0.00040785994415658283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":2.838771773968452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001648890179322677,"output_power":0.00016890593625973598,"reference_active":true,"delta_db":0.10453181888706232},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.76924654362113e-05,"output_power":2.8565051892439376e-05,"reference_active":true,"delta_db":0.1347339502599384},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.674663632481416e-06,"output_power":5.273174487187256e-06,"reference_active":true,"delta_db":0.5232177571750001}],"active_band_count":3,"active_band_mae_db":0.2541611754406669}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1189,"rms_floor":6.811887222434073e-05,"log_rms_mae_db":0.8455021808087454,"rms_envelope_nrmse":0.07511588325632751,"envelope_correlation":0.994156956588429,"rms_gain_error_db":0.024415059955735546,"reference_mean_power":0.00040588980481770184,"output_mean_power":0.0004081780520803878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":2.6000776284565166e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001543638209252824,"output_power":0.00015882462945604695,"reference_active":true,"delta_db":0.12372330778303547},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4081253735683761e-05,"output_power":1.4007863533098909e-05,"reference_active":true,"delta_db":-0.02269422155983521},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.814778327541048e-07,"output_power":3.3503679793501245e-07,"reference_active":true,"delta_db":0.7564831080628918}],"active_band_count":3,"active_band_mae_db":0.30096687913525416}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1195,"rms_floor":2.4474245023095563e-05,"log_rms_mae_db":1.5331510444974057,"rms_envelope_nrmse":0.20412826150695626,"envelope_correlation":0.9658016152037117,"rms_gain_error_db":0.3150749757158059,"reference_mean_power":1.616962857725981e-05,"output_mean_power":1.7386314915952804e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":6.118527503862653e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2276737657561945e-06,"output_power":2.4386891158515032e-06,"reference_active":true,"delta_db":0.3930484956713082},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4252050652633842e-06,"output_power":1.2957301517005595e-06,"reference_active":true,"delta_db":-0.41362792252656205},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.795674574712449e-07,"output_power":4.2620125270714e-07,"reference_active":true,"delta_db":-0.5123498316015631}],"active_band_count":3,"active_band_mae_db":0.43967541659981113}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":21,"metrics":{"frame_count":1189,"rms_floor":1.9203364541086674e-05,"log_rms_mae_db":0.8831559360584685,"rms_envelope_nrmse":0.12378856621684924,"envelope_correlation":0.9870061928575478,"rms_gain_error_db":0.315059449617317,"reference_mean_power":1.6250174487459307e-05,"output_mean_power":1.747285905043052e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":5.249214177030179e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.48645253897278e-06,"output_power":2.564428344990929e-06,"reference_active":true,"delta_db":0.1341039465639057},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.576700130403763e-07,"output_power":7.062308842962228e-07,"reference_active":true,"delta_db":-0.3053339329815337},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.81086319731659e-08,"output_power":3.7693518611667795e-08,"reference_active":true,"delta_db":-1.0595632756231281}],"active_band_count":3,"active_band_mae_db":0.49966705172285586}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"hnm","status":"measured","mos":{"status":"measured","value":2.7156693935394287,"predicted_mos":2.7156693935394287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"17cd3a386ac1fb2fce09fc4b23e34d29a0317cdb782637c2fa38ec409c006a69","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.707746172998668},"started_at":"2026-10-13T04:39:05.979867+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4205000297806691e-05,"native_input_samples":96000,"native_input_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","output_sha256":"be30244d2d87479a510e52b226ea97e3e21ae2178d503eb0d4581304962beebd","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1284495749996495,"synthesis_s":0.8626340029995845,"total_s":1.991083577999234,"rtf":0.33184726299987233,"output_peak":0.34681114493042037,"output_rms":0.04156079548914815,"finished_at":"2026-10-13T04:39:07.977115+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"be30244d2d87479a510e52b226ea97e3e21ae2178d503eb0d4581304962beebd","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.125019502786032,"spectral_convergence":0.2010668470983575,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1929116257187005,"spectral_convergence":0.19588825775314186,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.830159109126149,"spectral_convergence":0.20610565251795224,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":0.00011229293342314665,"delta_db":0.05422650264323928,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0007687742493421618,"delta_db":-0.153296810281575,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.0005281223211821759,"delta_db":-0.15528007324425555,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.00019465356530700006,"delta_db":-0.25789316389506645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":9.018283303512341e-05,"delta_db":-0.2566347296737713,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":3.327381940118908e-05,"delta_db":-0.4575666107131191,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22248298174183778,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16417752799868446,"pitch":null,"output_rms_dbfs":-27.626322969534375,"output_peak":0.34681114493042037,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.38269674587696},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.125019502786032,"spectral_convergence":0.2010668470983575,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1929116257187005,"spectral_convergence":0.19588825775314186,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.830159109126149,"spectral_convergence":0.20610565251795224,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":0.00011229293342314665,"delta_db":0.05422650264323928,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0007687742493421618,"delta_db":-0.153296810281575,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.0005281223211821759,"delta_db":-0.15528007324425555,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.00019465356530700006,"delta_db":-0.25789316389506645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":9.018283303512341e-05,"delta_db":-0.2566347296737713,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":3.327381940118908e-05,"delta_db":-0.4575666107131191,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22248298174183778,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16417752799868446,"pitch":null,"output_rms_dbfs":-27.626322969534375,"output_peak":0.34681114493042037,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.38269674587696},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9219748130932653,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0423538337852307,"rms_dbfs":-27.46214544153569,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.152971413235022,"mean":-9.95953877766927e-07,"zero_energy":false},"output":{"rms":0.04156079548914815,"rms_dbfs":-27.626322969534375,"peak":0.34681114493042037,"peak_dbfs":-9.198139094504134,"crest_factor_db":18.428183875030243,"mean":1.6311724519936335e-05,"zero_energy":false},"rms_difference_db":-0.16417752799868524},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":7.592524083488438e-07,"output_rms_dbfs":-61.19613821825298,"output_zero_energy":false,"output_peak":0.00526426726322789,"output_energy_fraction":9.523806807225939e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":322,"output_estimated_voiced_frames":303,"both_estimated_voiced_frames":298,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9254658385093167,"absolute_difference_cents_median":1.2716124353880396,"absolute_difference_cents_p95":9.483332811947008,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015211892717126925,"log_rms_mae_db":0.5941722791980905,"rms_envelope_nrmse":0.05532990164405455,"envelope_correlation":0.997120548635358,"rms_gain_error_db":-0.16351106359120032,"reference_mean_power":0.0017946561941472319,"output_mean_power":0.0017283438788451642},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.098233390179555e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003381947765928735,"output_power":0.00032106286653521435,"reference_active":true,"delta_db":-0.22576816777034828},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.872738107665891e-05,"output_power":3.467040416449974e-05,"reference_active":true,"delta_db":-0.4805922458999575},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.329707857377203e-06,"output_power":7.1137423105327275e-06,"reference_active":true,"delta_db":-0.12988535991248407}],"active_band_count":3,"active_band_mae_db":0.27874859119426326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012855597503212083,"log_rms_mae_db":0.3274910255267841,"rms_envelope_nrmse":0.03270677103435198,"envelope_correlation":0.9992017168725866,"rms_gain_error_db":-0.16365986200733126,"reference_mean_power":0.0017996791281982027,"output_mean_power":0.0017331218347162354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.521732580259768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030590519204748624,"output_power":0.00028985555666957794,"reference_active":true,"delta_db":-0.23405217792584868},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7673272930744674e-05,"output_power":1.6213612085915226e-05,"reference_active":true,"delta_db":-0.37437206157293684},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.996266898550174e-07,"output_power":3.379468848598393e-07,"reference_active":true,"delta_db":-0.7280603819320307}],"active_band_count":3,"active_band_mae_db":0.44549487381027203}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013968182745477764,"log_rms_mae_db":1.1143245298694724,"rms_envelope_nrmse":0.06526423851217453,"envelope_correlation":0.9966252012999549,"rms_gain_error_db":-0.17013705411885333,"reference_mean_power":0.0010910330804496425,"output_mean_power":0.0010491176978333432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.057825806205699e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031975947498638787,"output_power":0.0003067845887646038,"reference_active":true,"delta_db":-0.1798988301702975},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9465579420304694e-05,"output_power":3.4411663752012134e-05,"reference_active":true,"delta_db":-0.5951281274170432},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.299925023694004e-06,"output_power":5.707678989243082e-06,"reference_active":true,"delta_db":-0.4287584130465004}],"active_band_count":3,"active_band_mae_db":0.40126179021128044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012087571887341512,"log_rms_mae_db":0.5695764003676098,"rms_envelope_nrmse":0.037008732457617335,"envelope_correlation":0.9991076603348628,"rms_gain_error_db":-0.17016485250752303,"reference_mean_power":0.0010942126232004053,"output_mean_power":0.0010521683539387347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.627912926828162e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002965811192999727,"output_power":0.00028362118029371514,"reference_active":true,"delta_db":-0.19404839865834797},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8971998770340482e-05,"output_power":1.6727097210878423e-05,"reference_active":true,"delta_db":-0.5469250698114207},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113014123121409e-07,"output_power":3.009029754617027e-07,"reference_active":true,"delta_db":-1.3573371851943223}],"active_band_count":3,"active_band_mae_db":0.699436884554697}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":1.2330589880830543,"rms_envelope_nrmse":0.11210462473557016,"envelope_correlation":0.990947309788257,"rms_gain_error_db":-0.2929494646365078,"reference_mean_power":0.00025666306625944267,"output_mean_power":0.0002399210933305522},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2531384738435268e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.15813777859368e-05,"output_power":6.419952047180407e-05,"reference_active":true,"delta_db":-0.4726826917244164},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.324550913609104e-06,"output_power":5.647600569153564e-06,"reference_active":true,"delta_db":0.2558098918637813},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.777204343726204e-06,"output_power":2.6430138177708567e-06,"reference_active":true,"delta_db":1.723620676005475}],"active_band_count":3,"active_band_mae_db":0.8173710865312241}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":0.8110895534898803,"rms_envelope_nrmse":0.06752051876866495,"envelope_correlation":0.997161663661869,"rms_gain_error_db":-0.2932193771323798,"reference_mean_power":0.0002576568649688171,"output_mean_power":0.0002408350988833294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1689575138838286e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.6073744836406e-05,"output_power":5.898219471438332e-05,"reference_active":true,"delta_db":-0.49307993330600997},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.394936823991025e-06,"output_power":2.6524922374452155e-06,"reference_active":true,"delta_db":0.44360059953607456},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.833675392620658e-08,"output_power":1.5103970539320153e-07,"reference_active":true,"delta_db":1.8637526162328615}],"active_band_count":3,"active_band_mae_db":0.9334777163583153}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":1.356889183083505,"rms_envelope_nrmse":0.22751531341761339,"envelope_correlation":0.9610502270784639,"rms_gain_error_db":0.22973363677611885,"reference_mean_power":1.902643050734748e-05,"output_mean_power":2.005998864709095e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1975009508047086e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.298926861031968e-06,"output_power":5.729957291899265e-06,"reference_active":true,"delta_db":0.33963459813165},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5880342223027327e-06,"output_power":1.6365715810811685e-06,"reference_active":true,"delta_db":0.13075148137601492},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.193680029927598e-07,"output_power":8.728021689006862e-07,"reference_active":true,"delta_db":3.183205252738543}],"active_band_count":3,"active_band_mae_db":1.2178637774154026}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":0.9133073623729087,"rms_envelope_nrmse":0.15913045362770264,"envelope_correlation":0.9799568270221514,"rms_gain_error_db":0.2298318700392566,"reference_mean_power":1.9104126133492897e-05,"output_mean_power":2.014236046873278e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1095182205605226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.387890121011253e-06,"output_power":5.946331072720723e-06,"reference_active":true,"delta_db":0.4283035539965553},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.477294263383276e-07,"output_power":7.921183744063693e-07,"reference_active":true,"delta_db":0.2504561521543957},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.44303543402636e-08,"output_power":5.180036171590082e-08,"reference_active":true,"delta_db":3.2640302631465086}],"active_band_count":3,"active_band_mae_db":1.3142633230991532}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.350426435470581,"predicted_mos":3.350426435470581,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c24c2b1f563dc1a637b3d1a44d13aff69609c75a52baab6db0dfb80dbca36265","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8083522519991675},"started_at":"2026-10-13T04:42:03.268681+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1986000319884624e-05,"native_input_samples":96000,"native_input_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","output_sha256":"a4034a1a8f9914d4191e44d9985f9e4c9379142659f5360a6c577aeb8f23fcf5","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3389328390003357,"synthesis_s":0.10258503099976224,"total_s":1.441517870000098,"rtf":0.24025297833334966,"output_peak":0.4255907698986763,"output_rms":0.05108257215322078,"finished_at":"2026-10-13T04:42:04.717958+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"a4034a1a8f9914d4191e44d9985f9e4c9379142659f5360a6c577aeb8f23fcf5","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.96129116897238,"spectral_convergence":0.5573963113134087,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.284038703183226,"spectral_convergence":0.3287649440339218,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.105804332868866,"spectral_convergence":0.3639424463380498,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":0.00012886297610983986,"delta_db":0.6519838044984926,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0012636888568228941,"delta_db":2.0051164698940447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.0007855190623501105,"delta_db":1.568941978829979,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.0002721016897190397,"delta_db":1.1967955957873788,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":0.00011878783656806002,"delta_db":0.9398462555346375,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":4.046875621906322e-05,"delta_db":0.3926054068505287,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1258815852325101,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.627600583694507,"pitch":null,"output_rms_dbfs":-25.834544857841184,"output_peak":0.4255907698986763,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.783711401674824},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.96129116897238,"spectral_convergence":0.5573963113134087,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.284038703183226,"spectral_convergence":0.3287649440339218,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.105804332868866,"spectral_convergence":0.3639424463380498,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":0.00012886297610983986,"delta_db":0.6519838044984926,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0012636888568228941,"delta_db":2.0051164698940447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.0007855190623501105,"delta_db":1.568941978829979,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.0002721016897190397,"delta_db":1.1967955957873788,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":0.00011878783656806002,"delta_db":0.9398462555346375,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":4.046875621906322e-05,"delta_db":0.3926054068505287,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1258815852325101,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.627600583694507,"pitch":null,"output_rms_dbfs":-25.834544857841184,"output_peak":0.4255907698986763,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.783711401674824},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.924634854022209,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0423538337852307,"rms_dbfs":-27.46214544153569,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.152971413235022,"mean":-9.95953877766927e-07,"zero_energy":false},"output":{"rms":0.05108257215322079,"rms_dbfs":-25.834544857841184,"peak":0.4255907698986763,"peak_dbfs":-7.420155989520527,"crest_factor_db":18.414388868320653,"mean":7.4402450920636325e-06,"zero_energy":false},"rms_difference_db":1.6276005836945053},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":8.424362469578112e-07,"output_rms_dbfs":-60.74462955313665,"output_zero_energy":false,"output_peak":0.00836495084956853,"output_energy_fraction":6.994934181819153e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":322,"output_estimated_voiced_frames":347,"both_estimated_voiced_frames":300,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":47,"paired_voiced_fraction":0.9316770186335404,"absolute_difference_cents_median":4.806626625180153,"absolute_difference_cents_p95":1204.9591308084978,"difference_over_600_cents_frames":19,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015211892717126925,"log_rms_mae_db":1.4345972800047708,"rms_envelope_nrmse":0.26207129200183443,"envelope_correlation":0.979254011874303,"rms_gain_error_db":1.6289224777869291,"reference_mean_power":0.0017946561941472319,"output_mean_power":0.0026114006627961325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.098233390179555e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003381947765928735,"output_power":0.0005551279320919339,"reference_active":true,"delta_db":2.152261844082291},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.872738107665891e-05,"output_power":6.47303898727026e-05,"reference_active":true,"delta_db":2.230900938000054},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.329707857377203e-06,"output_power":1.0613387008139043e-05,"reference_active":true,"delta_db":1.607673355323446}],"active_band_count":3,"active_band_mae_db":1.9969453791352638}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012855597503212083,"log_rms_mae_db":1.191033619417687,"rms_envelope_nrmse":0.2441355194117825,"envelope_correlation":0.9862319552021919,"rms_gain_error_db":1.6318245626836767,"reference_mean_power":0.0017996791281982027,"output_mean_power":0.002620460008542931},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.521732580259768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030590519204748624,"output_power":0.0005070111426678324,"reference_active":true,"delta_db":2.1943065578795844},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7673272930744674e-05,"output_power":2.8397695903554667e-05,"reference_active":true,"delta_db":2.059661198625016},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.996266898550174e-07,"output_power":5.381775361871646e-07,"reference_active":true,"delta_db":1.2927108039211588}],"active_band_count":3,"active_band_mae_db":1.8488928534752531}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013968182745477764,"log_rms_mae_db":1.84822310987112,"rms_envelope_nrmse":0.33230370411182275,"envelope_correlation":0.9788016426824915,"rms_gain_error_db":2.119234793433171,"reference_mean_power":0.0010910330804496425,"output_mean_power":0.001777302689443563},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.057825806205699e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031975947498638787,"output_power":0.0005474405536627286,"reference_active":true,"delta_db":2.335135440353184},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9465579420304694e-05,"output_power":7.140605549636494e-05,"reference_active":true,"delta_db":2.575165597556362},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.299925023694004e-06,"output_power":8.67694027790298e-06,"reference_active":true,"delta_db":1.3903122739360254}],"active_band_count":3,"active_band_mae_db":2.100204437281857}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012087571887341512,"log_rms_mae_db":1.5539208684381995,"rms_envelope_nrmse":0.318470048789111,"envelope_correlation":0.9840772218650667,"rms_gain_error_db":2.1229638377823052,"reference_mean_power":0.0010942126232004053,"output_mean_power":0.0017840133684057246},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.627912926828162e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002965811192999727,"output_power":0.0005113868677329388,"reference_active":true,"delta_db":2.3660607139762218},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8971998770340482e-05,"output_power":3.305927111735999e-05,"reference_active":true,"delta_db":2.4118018629228857},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113014123121409e-07,"output_power":4.818464193281294e-07,"reference_active":true,"delta_db":0.6874843524883429}],"active_band_count":3,"active_band_mae_db":1.8217823097958166}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":1.2903308370602566,"rms_envelope_nrmse":0.18054020728081793,"envelope_correlation":0.9896105272756442,"rms_gain_error_db":1.0650199480581832,"reference_mean_power":0.00025666306625944267,"output_mean_power":0.00032799360266854604},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2531384738435268e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.15813777859368e-05,"output_power":9.620817229240868e-05,"reference_active":true,"delta_db":1.2841191089012844},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.324550913609104e-06,"output_power":7.614450266198057e-06,"reference_active":true,"delta_db":1.5535557028929705},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.777204343726204e-06,"output_power":2.534269016844407e-06,"reference_active":true,"delta_db":1.5411534802724414}],"active_band_count":3,"active_band_mae_db":1.459609430688899}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":0.9523834899009538,"rms_envelope_nrmse":0.1456210200041507,"envelope_correlation":0.997103380103012,"rms_gain_error_db":1.0639532494928958,"reference_mean_power":0.0002576568649688171,"output_mean_power":0.00032918273041057547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1689575138838286e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.6073744836406e-05,"output_power":8.92423188408704e-05,"reference_active":true,"delta_db":1.3054192446433788},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.394936823991025e-06,"output_power":3.325812104837102e-06,"reference_active":true,"delta_db":1.4260364801188472},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.833675392620658e-08,"output_power":1.461008891177865e-07,"reference_active":true,"delta_db":1.7193699067424673}],"active_band_count":3,"active_band_mae_db":1.4836085438348976}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":1.3928144404769582,"rms_envelope_nrmse":0.2911977641506235,"envelope_correlation":0.952796609129656,"rms_gain_error_db":1.0591329023150626,"reference_mean_power":1.902643050734748e-05,"output_mean_power":2.4281225901238396e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1975009508047086e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.298926861031968e-06,"output_power":6.568615157870812e-06,"reference_active":true,"delta_db":0.9328589299842525},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5880342223027327e-06,"output_power":2.0516834971948188e-06,"reference_active":true,"delta_db":1.1125050791704403},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.193680029927598e-07,"output_power":7.858827736260687e-07,"reference_active":true,"delta_db":2.7276247771693156}],"active_band_count":3,"active_band_mae_db":1.5909962621080027}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":0.9626033253830495,"rms_envelope_nrmse":0.19830871786804433,"envelope_correlation":0.9833615200618089,"rms_gain_error_db":1.0586427515797012,"reference_mean_power":1.9104126133492897e-05,"output_mean_power":2.4377628367028895e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1095182205605226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.387890121011253e-06,"output_power":6.805140024265912e-06,"reference_active":true,"delta_db":1.014183355113667},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.477294263383276e-07,"output_power":9.958507445203598e-07,"reference_active":true,"delta_db":1.2444978007002452},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.44303543402636e-08,"output_power":4.9850544225819647e-08,"reference_active":true,"delta_db":3.0974013786210564}],"active_band_count":3,"active_band_mae_db":1.785360844811656}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"facodec","status":"measured","mos":{"status":"measured","value":3.5899994373321533,"predicted_mos":3.5899994373321533,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"9eb175bb71575a5f6f354eebb3263b2ec0922d6f17d8a2b09c92c1ed9202ccf4","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7048021120008343},"started_at":"2026-10-13T04:51:39.449404+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.323400076420512e-05,"native_input_samples":96000,"native_input_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","output_sha256":"f0edd34d579f153a8c0b6a2443f267d12c5ef2440b0c389c0b0c3ed43b91ab33","output_samples":96000,"length_delta_samples":0,"analysis_s":1.8472710590003771,"synthesis_s":6.292647143999602,"total_s":8.13991820299998,"rtf":1.3566530338333298,"output_peak":0.31197988986968994,"output_rms":0.03949139864704315,"finished_at":"2026-10-13T04:51:47.596657+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"f0edd34d579f153a8c0b6a2443f267d12c5ef2440b0c389c0b0c3ed43b91ab33","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.457049702022021,"spectral_convergence":0.3262309598134772,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.034798806485905,"spectral_convergence":0.2691298848822781,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.388450308304292,"spectral_convergence":0.3248358324906884,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":9.917311634465497e-05,"delta_db":-0.48535816367603063,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0007537671394220129,"delta_db":-0.2389130851880801,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.000452269630931908,"delta_db":-0.8286510407814804,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.00015436062430889365,"delta_db":-1.2651515272684462,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":7.533637450581499e-05,"delta_db":-1.0378263088626776,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":2.4663681586397013e-05,"delta_db":-1.7580141123631412,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9356523730233093,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6078042544923943,"pitch":null,"output_rms_dbfs":-28.069949696028086,"output_peak":0.31197988986968994,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.960099605604072},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.457049702022021,"spectral_convergence":0.3262309598134772,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.034798806485905,"spectral_convergence":0.2691298848822781,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.388450308304292,"spectral_convergence":0.3248358324906884,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":9.917311634465497e-05,"delta_db":-0.48535816367603063,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0007537671394220129,"delta_db":-0.2389130851880801,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.000452269630931908,"delta_db":-0.8286510407814804,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.00015436062430889365,"delta_db":-1.2651515272684462,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":7.533637450581499e-05,"delta_db":-1.0378263088626776,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":2.4663681586397013e-05,"delta_db":-1.7580141123631412,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9356523730233093,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6078042544923943,"pitch":null,"output_rms_dbfs":-28.069949696028086,"output_peak":0.31197988986968994,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.960099605604072},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8901165891371722,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0423538337852307,"rms_dbfs":-27.46214544153569,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.152971413235022,"mean":-9.95953877766927e-07,"zero_energy":false},"output":{"rms":0.03949139864704315,"rms_dbfs":-28.069949696028083,"peak":0.31197988986968994,"peak_dbfs":-10.11746799143271,"crest_factor_db":17.952481704595375,"mean":7.741086756513009e-06,"zero_energy":false},"rms_difference_db":-0.6078042544923932},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":8.712160044869803e-07,"output_rms_dbfs":-60.59874155065255,"output_zero_energy":false,"output_peak":0.005889192223548889,"output_energy_fraction":0.00012103554120663312,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":322,"output_estimated_voiced_frames":326,"both_estimated_voiced_frames":313,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9720496894409938,"absolute_difference_cents_median":5.891675264135438,"absolute_difference_cents_p95":28.25960309355127,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015211892717126925,"log_rms_mae_db":1.1156074500533852,"rms_envelope_nrmse":0.12457991121163399,"envelope_correlation":0.9876686307719081,"rms_gain_error_db":-0.6082703888950329,"reference_mean_power":0.0017946561941472319,"output_mean_power":0.0015601064175344537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.098233390179555e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003381947765928735,"output_power":0.000285586451749293,"reference_active":true,"delta_db":-0.7342929503178941},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.872738107665891e-05,"output_power":4.0545631671541814e-05,"reference_active":true,"delta_db":0.19925941808270148},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.329707857377203e-06,"output_power":7.541745775626966e-06,"reference_active":true,"delta_db":0.123852235376066}],"active_band_count":3,"active_band_mae_db":0.35246820125888717}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012855597503212083,"log_rms_mae_db":0.9263149008282645,"rms_envelope_nrmse":0.1107746607580175,"envelope_correlation":0.9908486355387442,"rms_gain_error_db":-0.6086127773000636,"reference_mean_power":0.0017996791281982027,"output_mean_power":0.0015643495522902164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.521732580259768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030590519204748624,"output_power":0.00026036228941758667,"reference_active":true,"delta_db":-0.7000876648728622},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7673272930744674e-05,"output_power":1.935529418119991e-05,"reference_active":true,"delta_db":0.39482792172966663},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.996266898550174e-07,"output_power":3.9597094205466994e-07,"reference_active":true,"delta_db":-0.03991168920780553}],"active_band_count":3,"active_band_mae_db":0.3782757586034448}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013968182745477764,"log_rms_mae_db":1.535737167275488,"rms_envelope_nrmse":0.1288446379805624,"envelope_correlation":0.9870217695806776,"rms_gain_error_db":-0.42198249921201664,"reference_mean_power":0.0010910330804496425,"output_mean_power":0.0009900101992745903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.057825806205699e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031975947498638787,"output_power":0.0002695676966972279,"reference_active":true,"delta_db":-0.7415557433049967},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9465579420304694e-05,"output_power":4.2144387903919064e-05,"reference_active":true,"delta_db":0.28521267265291467},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.299925023694004e-06,"output_power":6.282305341386722e-06,"reference_active":true,"delta_db":-0.012163401207729381}],"active_band_count":3,"active_band_mae_db":0.34631060572188027}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012087571887341512,"log_rms_mae_db":1.2162743617461513,"rms_envelope_nrmse":0.1110234097435426,"envelope_correlation":0.9903880624718213,"rms_gain_error_db":-0.4214570407683872,"reference_mean_power":0.0010942126232004053,"output_mean_power":0.0009930154750745946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.627912926828162e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002965811192999727,"output_power":0.00025175602166302154,"reference_active":true,"delta_db":-0.7116363279340929},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8971998770340482e-05,"output_power":2.0996830975108768e-05,"reference_active":true,"delta_db":0.4404066433697723},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113014123121409e-07,"output_power":4.00994120346344e-07,"reference_active":true,"delta_db":-0.1102219603642462}],"active_band_count":3,"active_band_mae_db":0.4207549772227038}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":1.7050349022633124,"rms_envelope_nrmse":0.18442702686744858,"envelope_correlation":0.9841914235165525,"rms_gain_error_db":-1.1721551551750813,"reference_mean_power":0.00025666306625944267,"output_mean_power":0.0001959511708393827},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2531384738435268e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.15813777859368e-05,"output_power":5.999748619514671e-05,"reference_active":true,"delta_db":-0.7666699888305808},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.324550913609104e-06,"output_power":3.6571207399028165e-06,"reference_active":true,"delta_db":-1.6314368532154089},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.777204343726204e-06,"output_power":1.834204832987815e-06,"reference_active":true,"delta_db":0.13710467399400034}],"active_band_count":3,"active_band_mae_db":0.8450705053466634}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":1.3766535525035857,"rms_envelope_nrmse":0.16719946026045407,"envelope_correlation":0.9890702799948876,"rms_gain_error_db":-1.17265217668468,"reference_mean_power":0.0002576568649688171,"output_mean_power":0.00019668738238415735},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1689575138838286e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.6073744836406e-05,"output_power":5.556647057829305e-05,"reference_active":true,"delta_db":-0.7521610931801117},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.394936823991025e-06,"output_power":1.6752197746984697e-06,"reference_active":true,"delta_db":-1.5522227077100466},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.833675392620658e-08,"output_power":1.021906765290402e-07,"reference_active":true,"delta_db":0.16695406075518973}],"active_band_count":3,"active_band_mae_db":0.8237792872151161}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":1.8075025949051986,"rms_envelope_nrmse":0.2353487209255933,"envelope_correlation":0.9569298190085385,"rms_gain_error_db":-0.6814341462486134,"reference_mean_power":1.902643050734748e-05,"output_mean_power":1.626349590218947e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1975009508047086e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.298926861031968e-06,"output_power":4.2111832917362e-06,"reference_active":true,"delta_db":-0.99783780665409},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5880342223027327e-06,"output_power":1.4203569063078954e-06,"reference_active":true,"delta_db":-0.48462369967754604},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.193680029927598e-07,"output_power":3.8053257032729664e-07,"reference_active":true,"delta_db":-0.4220345693950839}],"active_band_count":3,"active_band_mae_db":0.63483202524224}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":1.4556557070666543,"rms_envelope_nrmse":0.19543462892185712,"envelope_correlation":0.9697120313554158,"rms_gain_error_db":-0.681463415618415,"reference_mean_power":1.9104126133492897e-05,"output_mean_power":1.632979885562649e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1095182205605226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.387890121011253e-06,"output_power":4.295230467992289e-06,"reference_active":true,"delta_db":-0.9843225866071865},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.477294263383276e-07,"output_power":6.727844153249617e-07,"reference_active":true,"delta_db":-0.4586854960361926},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.44303543402636e-08,"output_power":2.5987309728007546e-08,"reference_active":true,"delta_db":0.2683155649642228}],"active_band_count":3,"active_band_mae_db":0.5704412158692006}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.6553306579589844,"predicted_mos":3.6553306579589844,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"c65d346dda1a0c23d1f033bad1cb6da64a6a3ee8ee1f919cc66cf6dc05a64c66","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.5549348419990565},"started_at":"2026-10-13T07:13:25.281122+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009867809998468147,"native_input_samples":264600,"native_input_sha256":"4eef261ed584a89d9d85d7d66e3834b57ac4c7d6b71100dd912c5bf2f72db5c4","output_sha256":"1ab538b75dc3f1721252363df86e312a4face149f83cd73b3680cf947b095890","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15505531000053452,"synthesis_s":5.084025692000068,"total_s":5.239081002000603,"rtf":0.8731801670001005,"output_peak":0.291077196598053,"output_rms":0.03902934350461722,"finished_at":"2026-10-13T07:13:30.563458+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"1ab538b75dc3f1721252363df86e312a4face149f83cd73b3680cf947b095890","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.179893959873628,"output_peak":0.2915041584790673},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.865175342383216,"spectral_convergence":0.4439392714547045,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.94915118359311,"spectral_convergence":0.3820640424085171,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.914294273207565,"spectral_convergence":0.4383148253919953,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":4.6007606725920734e-05,"delta_db":-3.821001345264406,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.000772432369495383,"delta_db":-0.13268043917726316,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.00045679398064214197,"delta_db":-0.7854215833041028,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.00015548939340179712,"delta_db":-1.2335090984178956,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":7.421558273491732e-05,"delta_db":-1.102922453853292,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":1.540812858369834e-05,"delta_db":-3.801094220890171,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.8127715234845219,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7184271676637545,"pitch":null,"output_rms_dbfs":-28.180572609199444,"output_peak":0.2915041584790673,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.90954026639463},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7374450437139889,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0423538337852307,"rms_dbfs":-27.46214544153569,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.152971413235022,"mean":-9.95953877766927e-07,"zero_energy":false},"output":{"rms":0.038991628096090816,"rms_dbfs":-28.180572609199444,"peak":0.2915041584790673,"peak_dbfs":-10.70710490786743,"crest_factor_db":17.473467701332016,"mean":-5.5366897748771e-05,"zero_energy":false},"rms_difference_db":-0.718427167663755},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":5.087849217900912e-07,"output_rms_dbfs":-62.93465767800665,"output_zero_energy":false,"output_peak":0.004934052304323096,"output_energy_fraction":7.250761082122786e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":322,"output_estimated_voiced_frames":284,"both_estimated_voiced_frames":278,"reference_voiced_output_unvoiced_frames":44,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.8633540372670807,"absolute_difference_cents_median":7.30386265466502,"absolute_difference_cents_p95":37.268602671492985,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015211892717126925,"log_rms_mae_db":1.4245249337748547,"rms_envelope_nrmse":0.12031733760944092,"envelope_correlation":0.9910525555327667,"rms_gain_error_db":-0.7180919728306563,"reference_mean_power":0.0017946561941472319,"output_mean_power":0.0015211500811581624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.098233390179555e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003381947765928735,"output_power":0.0002903268133968885,"reference_active":true,"delta_db":-0.6627974830381469},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.872738107665891e-05,"output_power":2.9697940760973534e-05,"reference_active":true,"delta_db":-1.1529179237344065},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.329707857377203e-06,"output_power":6.363598501006211e-06,"reference_active":true,"delta_db":-0.6138389430050364}],"active_band_count":3,"active_band_mae_db":0.8098514499258632}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012855597503212083,"log_rms_mae_db":1.2407195944045488,"rms_envelope_nrmse":0.10260810786934599,"envelope_correlation":0.9952256695572393,"rms_gain_error_db":-0.7173853868895524,"reference_mean_power":0.0017996791281982027,"output_mean_power":0.001525655718701473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.521732580259768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030590519204748624,"output_power":0.00026313397413580065,"reference_active":true,"delta_db":-0.6540992324623354},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7673272930744674e-05,"output_power":1.3041739025419783e-05,"reference_active":true,"delta_db":-1.3198147896250403},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.996266898550174e-07,"output_power":3.6489056262279046e-07,"reference_active":true,"delta_db":-0.3949185463562276}],"active_band_count":3,"active_band_mae_db":0.7896108561478677}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013968182745477764,"log_rms_mae_db":2.617393261507754,"rms_envelope_nrmse":0.14764763772173348,"envelope_correlation":0.9825057192309521,"rms_gain_error_db":-0.4391819002718847,"reference_mean_power":0.0010910330804496425,"output_mean_power":0.0009860972069604726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.057825806205699e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031975947498638787,"output_power":0.00026789543160993137,"reference_active":true,"delta_db":-0.7685811448643718},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9465579420304694e-05,"output_power":2.7315996938110913e-05,"reference_active":true,"delta_db":-1.598014279873781},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.299925023694004e-06,"output_power":5.502747593583535e-06,"reference_active":true,"delta_db":-0.5875578833937041}],"active_band_count":3,"active_band_mae_db":0.9847177693772856}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012087571887341512,"log_rms_mae_db":2.143810639005411,"rms_envelope_nrmse":0.11742241770877368,"envelope_correlation":0.9891768980746656,"rms_gain_error_db":-0.4333223707157736,"reference_mean_power":0.0010942126232004053,"output_mean_power":0.000990306166987866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.627912926828162e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002965811192999727,"output_power":0.00024828214753446767,"reference_active":true,"delta_db":-0.7719800668776065},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8971998770340482e-05,"output_power":1.2776793150304975e-05,"reference_active":true,"delta_db":-1.7169122398475074},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113014123121409e-07,"output_power":3.36091714451582e-07,"reference_active":true,"delta_db":-0.8770239468121309}],"active_band_count":3,"active_band_mae_db":1.1219720845124148}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":2.79932235679827,"rms_envelope_nrmse":0.2227778499992369,"envelope_correlation":0.9695219574290198,"rms_gain_error_db":-1.0792719713853824,"reference_mean_power":0.00025666306625944267,"output_mean_power":0.00020018714284408126},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2531384738435268e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.15813777859368e-05,"output_power":6.558804607798245e-05,"reference_active":true,"delta_db":-0.3797536025551017},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.324550913609104e-06,"output_power":1.038499838047523e-05,"reference_active":true,"delta_db":2.901234490076435},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.777204343726204e-06,"output_power":1.7856451787349367e-06,"reference_active":true,"delta_db":0.020577994676930115}],"active_band_count":3,"active_band_mae_db":1.1005220291028224}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":2.3912528444586383,"rms_envelope_nrmse":0.18833556237587076,"envelope_correlation":0.9805706898984016,"rms_gain_error_db":-1.0751495760399106,"reference_mean_power":0.0002576568649688171,"output_mean_power":0.00020115311421284144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1689575138838286e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.6073744836406e-05,"output_power":6.085186780956869e-05,"reference_active":true,"delta_db":-0.3575500868346481},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.394936823991025e-06,"output_power":4.9839229321437665e-06,"reference_active":true,"delta_db":3.1827725642769926},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.833675392620658e-08,"output_power":1.1071829357063388e-07,"reference_active":true,"delta_db":0.5150351543556184}],"active_band_count":3,"active_band_mae_db":1.351785935155753}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":3.9838843807810727,"rms_envelope_nrmse":0.38794045749318,"envelope_correlation":0.889898146810609,"rms_gain_error_db":-1.6888692582516949,"reference_mean_power":1.902643050734748e-05,"output_mean_power":1.2896456375271078e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1975009508047086e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.298926861031968e-06,"output_power":4.290749121747385e-06,"reference_active":true,"delta_db":-0.9165480289892364},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5880342223027327e-06,"output_power":7.778206898116557e-07,"reference_active":true,"delta_db":-3.0998036620700207},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.193680029927598e-07,"output_power":2.0054491910798855e-07,"reference_active":true,"delta_db":-3.203836281041823}],"active_band_count":3,"active_band_mae_db":2.406729324033693}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":3.6115137745155894,"rms_envelope_nrmse":0.34386596868430547,"envelope_correlation":0.9145281581687521,"rms_gain_error_db":-1.6879982647383183,"reference_mean_power":1.9104126133492897e-05,"output_mean_power":1.2951717120264179e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1095182205605226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.387890121011253e-06,"output_power":4.229547437385041e-06,"reference_active":true,"delta_db":-1.0512483002725057},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.477294263383276e-07,"output_power":3.5563980913895e-07,"reference_active":true,"delta_db":-3.227341038200404},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.44303543402636e-08,"output_power":1.2754511832182356e-08,"reference_active":true,"delta_db":-2.822659250476849}],"active_band_count":3,"active_band_mae_db":2.367082862983253}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"golf","status":"measured","mos":{"status":"measured","value":2.960301399230957,"predicted_mos":2.960301399230957,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"40200f792df13a1e5d1a3a6c589de7b75d216069c0839ea34d8aff8f72266fc9","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5439615920004144},"started_at":"2026-10-13T05:02:27.665533+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004298041999390989,"native_input_samples":144000,"native_input_sha256":"72c9645e03c0f9c48dbfd269c00bccdac3898cb1413184b60d399eb42f52f25e","output_sha256":"bd4a8fd40e922562ab14d8c3309fe2aba83b1158840fbe2621b9c1689cf582bc","output_samples":143760,"length_delta_samples":-240,"analysis_s":2.284344305999184,"synthesis_s":0.18624465100037924,"total_s":2.4705889569995634,"rtf":0.4117648261665939,"output_peak":0.36263376474380493,"output_rms":0.03640180777065783,"finished_at":"2026-10-13T05:02:30.162061+00:00","analysis_stages":{"external_f0_s":0.10794469499978732},"evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"bd4a8fd40e922562ab14d8c3309fe2aba83b1158840fbe2621b9c1689cf582bc","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.776874807684244,"output_peak":0.35797203146780343},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.864757318888508,"spectral_convergence":0.5115958259769918,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.29066249279926,"spectral_convergence":0.3083315834049023,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.699861610407316,"spectral_convergence":0.35306648603921736,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011068423537241924,"output_power":8.187855168667753e-05,"delta_db":-1.3091561720650167,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007965614154107522,"output_power":0.0005758774666895061,"delta_db":-1.4088918031560378,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000547347789344374,"output_power":0.00042528983860063764,"delta_db":-1.0957836208184404,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002068852048767748,"output_power":0.00016490956569990787,"delta_db":-0.9848358583368295,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.551473988192054e-05,"output_power":5.9807451487035744e-05,"delta_db":-2.0331510062934317,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.702090606222774e-05,"output_power":1.7532002984702967e-05,"delta_db":-3.2461550770002185,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.679662256278329,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.3151337903911797,"pitch":null,"output_rms_dbfs":-28.776874807684244,"output_peak":0.35797203146780343,"reference_rms_dbfs":-27.461741017293065,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.618427140698361},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8641028740692908,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04235580587060608,"rms_dbfs":-27.46174101729306,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.152566988992398,"mean":-1.5252420619652546e-06,"zero_energy":false},"output":{"rms":0.036404599670212937,"rms_dbfs":-28.776874807684244,"peak":0.35797203146780343,"peak_dbfs":-8.923018073474282,"crest_factor_db":19.853856734209963,"mean":0.00026523306614905935,"zero_energy":false},"rms_difference_db":-1.315133790391183},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":1.2155295697605395e-06,"output_rms_dbfs":-59.15234471762949,"output_zero_energy":false,"output_peak":0.006313977300797504,"output_energy_fraction":0.00019905337196047613,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":321,"output_estimated_voiced_frames":312,"both_estimated_voiced_frames":295,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9190031152647975,"absolute_difference_cents_median":5.768558386716018,"absolute_difference_cents_p95":23.13812090011549,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015211892717126925,"log_rms_mae_db":2.043635317366756,"rms_envelope_nrmse":0.17469496640606677,"envelope_correlation":0.9876137854737059,"rms_gain_error_db":-1.3150384610388306,"reference_mean_power":0.0017948328119552085,"output_mean_power":0.0013259286487412305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.113104227610305e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003381966819816066,"output_power":0.00025964557147589956,"reference_active":true,"delta_db":-1.1478842280491106},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868127397359648e-05,"output_power":3.240196963649037e-05,"reference_active":true,"delta_db":-0.7692935849236913},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.34893026371255e-06,"output_power":4.833224655433571e-06,"reference_active":true,"delta_db":-1.819871439854591}],"active_band_count":3,"active_band_mae_db":1.2456830842757975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012855597503212083,"log_rms_mae_db":1.8752174230141199,"rms_envelope_nrmse":0.15880882668165838,"envelope_correlation":0.9934221610435926,"rms_gain_error_db":-1.3122144691618,"reference_mean_power":0.0018002204838929534,"output_mean_power":0.001330773831059894},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.533547943455634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030583339980221015,"output_power":0.00023555469646709926,"reference_active":true,"delta_db":-1.1339314501009896},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.758196150615533e-05,"output_power":1.5011463654309114e-05,"reference_active":true,"delta_db":-0.6864428579891728},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.007947798512119e-07,"output_power":2.2889154828934856e-07,"reference_active":true,"delta_db":-2.4329229957547556}],"active_band_count":3,"active_band_mae_db":1.417765767948306}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00013968182745477764,"log_rms_mae_db":2.1979316892368854,"rms_envelope_nrmse":0.16800228623535948,"envelope_correlation":0.9868216647653368,"rms_gain_error_db":-1.1143980535610707,"reference_mean_power":0.0010912048483350731,"output_mean_power":0.0008442410816913651},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.07162528783519e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031978000563049003,"output_power":0.00024020648712504744,"reference_active":true,"delta_db":-1.2426657384882664},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9408247550440805e-05,"output_power":3.215800687194152e-05,"reference_active":true,"delta_db":-0.8829799892843806},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.326910376479852e-06,"output_power":2.8821123406843194e-06,"reference_active":true,"delta_db":-3.414807775309949}],"active_band_count":3,"active_band_mae_db":1.8468178343608652}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012087571887341512,"log_rms_mae_db":1.7024292898385152,"rms_envelope_nrmse":0.14624758957600592,"envelope_correlation":0.9930683652936798,"rms_gain_error_db":-1.11374892915023,"reference_mean_power":0.001094776708413282,"output_mean_power":0.0008471311591148011},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.639159559596893e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029648379721976357,"output_power":0.00022327287036206353,"reference_active":true,"delta_db":-1.2316500860401285},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.888635923621955e-05,"output_power":1.6317646864508206e-05,"reference_active":true,"delta_db":-0.6349071592871826},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1237988403468996e-07,"output_power":1.5873356126169667e-07,"reference_active":true,"delta_db":-4.14628712203938}],"active_band_count":3,"active_band_mae_db":2.004281455788897}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":2.0716650618579524,"rms_envelope_nrmse":0.1887021116834264,"envelope_correlation":0.9856456021365614,"rms_gain_error_db":-1.2899712934195249,"reference_mean_power":0.00025683242500623493,"output_mean_power":0.00019083266839930774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.260439897396858e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.158433110521441e-05,"output_power":5.232454427754437e-05,"reference_active":true,"delta_db":-1.3611251667695363},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.2746914925898624e-06,"output_power":3.7898416628518685e-06,"reference_active":true,"delta_db":-1.43575997790763},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7853695134829478e-06,"output_power":1.3728569779999545e-06,"reference_active":true,"delta_db":-1.141028190959398}],"active_band_count":3,"active_band_mae_db":1.3126377785455214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":1.774551598720889,"rms_envelope_nrmse":0.15895629292921837,"envelope_correlation":0.9944780348467128,"rms_gain_error_db":-1.2905898636368505,"reference_mean_power":0.0002579206673408108,"output_mean_power":0.00019161396490219505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.1756288954485439e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.605399936913077e-05,"output_power":4.8184902780580596e-05,"reference_active":true,"delta_db":-1.3698813088485768},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.350327012191534e-06,"output_power":1.6243938052331788e-06,"reference_active":true,"delta_db":-1.6043696733408617},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.852093546492856e-08,"output_power":7.214458570184699e-08,"reference_active":true,"delta_db":-1.353247830782149}],"active_band_count":3,"active_band_mae_db":1.4424996043238625}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":2.8609568247500343,"rms_envelope_nrmse":0.29404108592950123,"envelope_correlation":0.9377110541594627,"rms_gain_error_db":-0.23623911839538297,"reference_mean_power":1.9042983458598197e-05,"output_mean_power":1.8034789528208698e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1990424602282367e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.312556520112214e-06,"output_power":6.144929792249028e-06,"reference_active":true,"delta_db":0.632133617725919},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5945723763159322e-06,"output_power":2.2721088175123774e-06,"reference_active":true,"delta_db":1.5378489079669888},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175107120339286e-07,"output_power":7.270496391620248e-07,"reference_active":true,"delta_db":2.408964406200732}],"active_band_count":3,"active_band_mae_db":1.5263156439645467}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":2.6064462030871636,"rms_envelope_nrmse":0.22892941758114796,"envelope_correlation":0.9611100015316991,"rms_gain_error_db":-0.2366632677589517,"reference_mean_power":1.9125747623005514e-05,"output_mean_power":1.811140298555326e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1106798776055528e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4007581166738e-06,"output_power":6.585394663295483e-06,"reference_active":true,"delta_db":0.8612708038100068},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.498215842579886e-07,"output_power":1.1538524797298844e-06,"reference_active":true,"delta_db":1.8719234992757714},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4455693776853864e-08,"output_power":5.4826241312740365e-08,"reference_active":true,"delta_db":3.5060848554453905}],"active_band_count":3,"active_band_mae_db":2.0797597195103896}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.8956100940704346,"predicted_mos":2.8956100940704346,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"5c8ce36cbf6062462dd94b9147001d5c2597f17e271d45e2d8d34676a7f532cc","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3642929589987034},"started_at":"2026-10-13T07:36:42.076677+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002952736998850014,"native_input_samples":144000,"native_input_sha256":"72c9645e03c0f9c48dbfd269c00bccdac3898cb1413184b60d399eb42f52f25e","output_sha256":"509863cbdbede49f8c1f1060386f9f608dd3e775107bde097d98ea4ad19ee6ef","output_samples":144120,"length_delta_samples":120,"analysis_s":2.0207683900007396,"synthesis_s":3.4353132719988935,"total_s":5.456081661999633,"rtf":0.9093469436666055,"output_peak":0.3757800757884979,"output_rms":0.04099514605444611,"finished_at":"2026-10-13T07:36:47.549522+00:00","analysis_stages":{"external_f0_s":1.1614211819978664},"evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"509863cbdbede49f8c1f1060386f9f608dd3e775107bde097d98ea4ad19ee6ef","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.75075016982749,"output_peak":0.3762667238555278},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.802297020346446,"spectral_convergence":0.4226382719377863,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.370535690751007,"spectral_convergence":0.24741363999641686,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.789128334344639,"spectral_convergence":0.276023882867416,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":0.00011316704526163089,"delta_db":0.08790200038042187,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0007529297129966673,"delta_db":-0.24374072802006344,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.0004920782530080845,"delta_db":-0.46228358560686,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.00020149692515168138,"delta_db":-0.10783255943723197,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":9.907832350280931e-05,"delta_db":0.15191302078968974,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":2.0035965673225588e-05,"delta_db":-2.6604903719310564,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6190270443608873,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.28790058669508795,"pitch":null,"output_rms_dbfs":-27.750046028230777,"output_peak":0.3762667238555278,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.653987015147364},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9048103943775774,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0423538337852307,"rms_dbfs":-27.46214544153569,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.152971413235022,"mean":-9.95953877766927e-07,"zero_energy":false},"output":{"rms":0.0409729938568577,"rms_dbfs":-27.750046028230777,"peak":0.3762667238555278,"peak_dbfs":-8.490083759297066,"crest_factor_db":19.259962268933712,"mean":1.2153258541050302e-05,"zero_energy":false},"rms_difference_db":-0.28790058669508767},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":5.10137797849298e-07,"output_rms_dbfs":-62.92312496917301,"output_zero_energy":false,"output_peak":0.004903689032023247,"output_energy_fraction":6.583914885384902e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":322,"output_estimated_voiced_frames":326,"both_estimated_voiced_frames":305,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.9472049689440993,"absolute_difference_cents_median":5.51543693213788,"absolute_difference_cents_p95":25.30671062737883,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015211892717126925,"log_rms_mae_db":1.6958343092921755,"rms_envelope_nrmse":0.1235641708477198,"envelope_correlation":0.984898510814228,"rms_gain_error_db":-0.2869680343738207,"reference_mean_power":0.0017946561941472319,"output_mean_power":0.0016799039791090303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.098233390179555e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003381947765928735,"output_power":0.00032570451299289096,"reference_active":true,"delta_db":-0.16343119417176677},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.872738107665891e-05,"output_power":3.625765200699361e-05,"reference_active":true,"delta_db":-0.2861845260322095},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.329707857377203e-06,"output_power":6.317152014157811e-06,"reference_active":true,"delta_db":-0.6456533740576675}],"active_band_count":3,"active_band_mae_db":0.36508969808721464}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012855597503212083,"log_rms_mae_db":1.37430522769155,"rms_envelope_nrmse":0.08891352731475093,"envelope_correlation":0.9923666605820838,"rms_gain_error_db":-0.28628390879157356,"reference_mean_power":0.0017996791281982027,"output_mean_power":0.0016848711311074928},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.521732580259768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030590519204748624,"output_power":0.00029540997293419225,"reference_active":true,"delta_db":-0.15161695363429126},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7673272930744674e-05,"output_power":1.683566298621851e-05,"reference_active":true,"delta_db":-0.21086760854808922},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.996266898550174e-07,"output_power":3.352742732712466e-07,"reference_active":true,"delta_db":-0.7625425591120338}],"active_band_count":3,"active_band_mae_db":0.37500904043147143}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013968182745477764,"log_rms_mae_db":1.9804741893592614,"rms_envelope_nrmse":0.1390347677833745,"envelope_correlation":0.9838836794750948,"rms_gain_error_db":-0.3001583024816106,"reference_mean_power":0.0010910330804496425,"output_mean_power":0.0010181742010129888},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.057825806205699e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031975947498638787,"output_power":0.00029785587687934764,"reference_active":true,"delta_db":-0.308172487676004},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9465579420304694e-05,"output_power":3.8171739368221e-05,"reference_active":true,"delta_db":-0.14476533495406244},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.299925023694004e-06,"output_power":5.3799391986640425e-06,"reference_active":true,"delta_db":-0.6855801336496135}],"active_band_count":3,"active_band_mae_db":0.37950598542655994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012087571887341512,"log_rms_mae_db":1.413842165760335,"rms_envelope_nrmse":0.09915950417423859,"envelope_correlation":0.9918660028328231,"rms_gain_error_db":-0.2996830981071211,"reference_mean_power":0.0010942126232004053,"output_mean_power":0.0010212531540558004},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.627912926828162e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002965811192999727,"output_power":0.0002760310527266198,"reference_active":true,"delta_db":-0.3118555814272421},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8971998770340482e-05,"output_power":1.848147229781377e-05,"reference_active":true,"delta_db":-0.11376522173018738},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113014123121409e-07,"output_power":3.5428570673718047e-07,"reference_active":true,"delta_db":-0.648065692179287}],"active_band_count":3,"active_band_mae_db":0.35789549844557217}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":1.6706207661702883,"rms_envelope_nrmse":0.148461528240579,"envelope_correlation":0.9831249050942501,"rms_gain_error_db":-0.020171204023249074,"reference_mean_power":0.00025666306625944267,"output_mean_power":0.00025547373531327634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2531384738435268e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.15813777859368e-05,"output_power":7.09989302419708e-05,"reference_active":true,"delta_db":-0.035482482107267455},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.324550913609104e-06,"output_power":5.102657515136947e-06,"reference_active":true,"delta_db":-0.18486564202203498},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.777204343726204e-06,"output_power":1.7175900979983933e-06,"reference_active":true,"delta_db":-0.14817838340418488}],"active_band_count":3,"active_band_mae_db":0.1228421691778291}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":1.0477142509663793,"rms_envelope_nrmse":0.08712973027105546,"envelope_correlation":0.9939489920057207,"rms_gain_error_db":-0.019854874832002063,"reference_mean_power":0.0002576568649688171,"output_mean_power":0.00025648160973223093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1689575138838286e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.6073744836406e-05,"output_power":6.538513498418426e-05,"reference_active":true,"delta_db":-0.045498971959030785},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.394936823991025e-06,"output_power":2.2853768679053815e-06,"reference_active":true,"delta_db":-0.2033623442168013},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.833675392620658e-08,"output_power":9.446106888098167e-08,"reference_active":true,"delta_db":-0.17463012670307584}],"active_band_count":3,"active_band_mae_db":0.14116381429296931}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":1.6596957320425367,"rms_envelope_nrmse":0.2399435771597701,"envelope_correlation":0.9542937780731584,"rms_gain_error_db":-0.6008009503409224,"reference_mean_power":1.902643050734748e-05,"output_mean_power":1.6568272322858753e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1975009508047086e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.298926861031968e-06,"output_power":4.3815308284121975e-06,"reference_active":true,"delta_db":-0.8256205343055233},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5880342223027327e-06,"output_power":1.2471138117087528e-06,"reference_active":true,"delta_db":-1.0495376824601452},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.193680029927598e-07,"output_power":2.565438306147979e-07,"reference_active":true,"delta_db":-2.1343371635240054}],"active_band_count":3,"active_band_mae_db":1.336498460096558}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":1.0334193532474603,"rms_envelope_nrmse":0.14068664273188922,"envelope_correlation":0.9859206863561901,"rms_gain_error_db":-0.60102297233703,"reference_mean_power":1.9104126133492897e-05,"output_mean_power":1.663507945756693e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1095182205605226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.387890121011253e-06,"output_power":4.361563258769171e-06,"reference_active":true,"delta_db":-0.9177655454114759},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.477294263383276e-07,"output_power":5.655254781612081e-07,"reference_active":true,"delta_db":-1.2129229670131583},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.44303543402636e-08,"output_power":1.4574059286121695e-08,"reference_active":true,"delta_db":-2.2434923420101684}],"active_band_count":3,"active_band_mae_db":1.4580602848116009}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.2022790908813477,"predicted_mos":3.2022790908813477,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"2c3e224bdd3aec2f1e3794259a5a079d1011573de982829412d162b97f252302","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.565073134001068},"started_at":"2026-10-13T05:10:36.972610+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028317539999989094,"native_input_samples":144000,"native_input_sha256":"72c9645e03c0f9c48dbfd269c00bccdac3898cb1413184b60d399eb42f52f25e","output_sha256":"c9c141f6454e4617499725a96ed591463f8b7a7e81a69e139e03430d5e965182","output_samples":144240,"length_delta_samples":240,"analysis_s":1.406693003998953,"synthesis_s":0.18862726500083227,"total_s":1.5953202689997852,"rtf":0.2658867114999642,"output_peak":0.3652859330177307,"output_rms":0.04163157399974327,"finished_at":"2026-10-13T05:10:38.579941+00:00","analysis_stages":{"external_f0_s":0.18044984900006966},"evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"c9c141f6454e4617499725a96ed591463f8b7a7e81a69e139e03430d5e965182","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.618022297583813,"output_peak":0.3700239302166701},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.092229979636769,"spectral_convergence":0.13965632890234284,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.190600194158051,"spectral_convergence":0.13812658619191528,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.8299983448159,"spectral_convergence":0.15136442641456171,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011089954817882694,"output_power":0.0001122962438248549,"delta_db":0.054354531002619,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007963949709957278,"output_power":0.0007844530226135576,"delta_db":-0.06561568174124041,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005473467336544822,"output_power":0.0005273364215066463,"delta_db":-0.16174762979925933,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020656259415557,"output_power":0.00019934202193570757,"delta_db":-0.1545282010926543,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.567254948983353e-05,"output_power":8.804351167997616e-05,"delta_db":-0.3608999085088204,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.697083983250908e-05,"output_power":2.1876491486164364e-05,"delta_db":-2.278816443039608,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.512660399197367,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14899764044167815,"pitch":null,"output_rms_dbfs":-27.611143081977367,"output_peak":0.3700239302166701,"reference_rms_dbfs":-27.46214544153569,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.370942839536907},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9612303407474444,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0423538337852307,"rms_dbfs":-27.46214544153569,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.152971413235022,"mean":-9.95953877766927e-07,"zero_energy":false},"output":{"rms":0.041633492683738504,"rms_dbfs":-27.611143081977367,"peak":0.3700239302166701,"peak_dbfs":-8.635403765958337,"crest_factor_db":18.97573931601903,"mean":1.8180913497887833e-05,"zero_energy":false},"rms_difference_db":-0.14899764044167796},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":6.935870047641413e-07,"output_rms_dbfs":-61.58899052515836,"output_zero_energy":false,"output_peak":0.005080313946242798,"output_energy_fraction":8.669765635274888e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":322,"output_estimated_voiced_frames":308,"both_estimated_voiced_frames":296,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9192546583850931,"absolute_difference_cents_median":0.7327757047085639,"absolute_difference_cents_p95":5.109251425317916,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015211892717126925,"log_rms_mae_db":0.5635943563884142,"rms_envelope_nrmse":0.05028684675291381,"envelope_correlation":0.9976829551926277,"rms_gain_error_db":-0.1486970704351627,"reference_mean_power":0.0017946561941472319,"output_mean_power":0.0017342494090407667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.098233390179555e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003381947765928735,"output_power":0.0003388675624045167,"reference_active":true,"delta_db":0.008631029813338348},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.872738107665891e-05,"output_power":3.5804888979395907e-05,"reference_active":true,"delta_db":-0.3407579754900728},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.329707857377203e-06,"output_power":5.572469583801577e-06,"reference_active":true,"delta_db":-1.1903895851521782}],"active_band_count":3,"active_band_mae_db":0.5132595301518631}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012855597503212083,"log_rms_mae_db":0.46012651055703446,"rms_envelope_nrmse":0.040230551430838014,"envelope_correlation":0.9985864670533725,"rms_gain_error_db":-0.14628151359765676,"reference_mean_power":0.0017996791281982027,"output_mean_power":0.001740070837562152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.521732580259768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030590519204748624,"output_power":0.00030756996578092306,"reference_active":true,"delta_db":0.023570761148870304},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7673272930744674e-05,"output_power":1.6606531959670205e-05,"reference_active":true,"delta_db":-0.27038038741852305},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.996266898550174e-07,"output_power":3.093047623346435e-07,"reference_active":true,"delta_db":-1.1126787889074046}],"active_band_count":3,"active_band_mae_db":0.46887664582493266}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013968182745477764,"log_rms_mae_db":0.7867021998129298,"rms_envelope_nrmse":0.03667482540551315,"envelope_correlation":0.9989310298203287,"rms_gain_error_db":-0.09253479378373113,"reference_mean_power":0.0010910330804496425,"output_mean_power":0.0010680324286003356},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.057825806205699e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031975947498638787,"output_power":0.0003165568594292078,"reference_active":true,"delta_db":-0.043716935539943144},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9465579420304694e-05,"output_power":3.7111384742675474e-05,"reference_active":true,"delta_db":-0.2671132380667863},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.299925023694004e-06,"output_power":4.639315224169288e-06,"reference_active":true,"delta_db":-1.3288149867193224}],"active_band_count":3,"active_band_mae_db":0.5465483867753506}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012087571887341512,"log_rms_mae_db":0.4775292223508008,"rms_envelope_nrmse":0.021711489252307647,"envelope_correlation":0.999671948492087,"rms_gain_error_db":-0.09082481569121556,"reference_mean_power":0.0010942126232004053,"output_mean_power":0.0010715667741529292},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.627912926828162e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002965811192999727,"output_power":0.0002934801565683682,"reference_active":true,"delta_db":-0.04564757785002017},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8971998770340482e-05,"output_power":1.7920669938515705e-05,"reference_active":true,"delta_db":-0.24758846726036932},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113014123121409e-07,"output_power":3.2106627526801725e-07,"reference_active":true,"delta_db":-1.0756551101273462}],"active_band_count":3,"active_band_mae_db":0.4562970517459119}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":0.9021124118389116,"rms_envelope_nrmse":0.06776752426751646,"envelope_correlation":0.997020637926171,"rms_gain_error_db":-0.24264143686443126,"reference_mean_power":0.00025666306625944267,"output_mean_power":0.00024271646295792332},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2531384738435268e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.15813777859368e-05,"output_power":7.050430594402524e-05,"reference_active":true,"delta_db":-0.06584411678161925},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.324550913609104e-06,"output_power":4.986351719341474e-06,"reference_active":true,"delta_db":-0.2850007531476},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.777204343726204e-06,"output_power":1.9237035483845047e-06,"reference_active":true,"delta_db":0.3440078002530602}],"active_band_count":3,"active_band_mae_db":0.23161755672742648}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":0.65404728249199,"rms_envelope_nrmse":0.04219660567619644,"envelope_correlation":0.9991809332053493,"rms_gain_error_db":-0.2411257542094173,"reference_mean_power":0.0002576568649688171,"output_mean_power":0.00024374131104636025},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1689575138838286e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.6073744836406e-05,"output_power":6.507417091565537e-05,"reference_active":true,"delta_db":-0.06620278206565104},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.394936823991025e-06,"output_power":2.232996964060337e-06,"reference_active":true,"delta_db":-0.3040592903946717},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.833675392620658e-08,"output_power":9.766255240847023e-08,"reference_active":true,"delta_db":-0.029877978567485853}],"active_band_count":3,"active_band_mae_db":0.13338001700926955}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":0.9722813808661677,"rms_envelope_nrmse":0.11895977552876572,"envelope_correlation":0.9893318780676352,"rms_gain_error_db":-0.3137727970825866,"reference_mean_power":1.902643050734748e-05,"output_mean_power":1.7700276046950503e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1975009508047086e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.298926861031968e-06,"output_power":4.9176485371685675e-06,"reference_active":true,"delta_db":-0.3243043854416613},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5880342223027327e-06,"output_power":1.2048267858981827e-06,"reference_active":true,"delta_db":-1.1993524301616552},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.193680029927598e-07,"output_power":2.9283792311497964e-07,"reference_active":true,"delta_db":-1.5596797346376374}],"active_band_count":3,"active_band_mae_db":1.027778850080318}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":0.759738758525542,"rms_envelope_nrmse":0.0707684288568445,"envelope_correlation":0.9966752702513526,"rms_gain_error_db":-0.31349796679559666,"reference_mean_power":1.9104126133492897e-05,"output_mean_power":1.7773680956620858e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1095182205605226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.387890121011253e-06,"output_power":4.872775718583007e-06,"reference_active":true,"delta_db":-0.43642307896101507},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.477294263383276e-07,"output_power":5.466423442678374e-07,"reference_active":true,"delta_db":-1.360412021367285},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.44303543402636e-08,"output_power":1.7854760472049188e-08,"reference_active":true,"delta_db":-1.361757377158077}],"active_band_count":3,"active_band_mae_db":1.0528641591621257}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_3729-6852-0017","method":"magphase","status":"measured","mos":{"status":"measured","value":2.8484840393066406,"predicted_mos":2.8484840393066406,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287157,"input_sample_rate":48000,"input_seconds":5.9824375,"input_float32_sha256":"4034e06c421041a2b669e47c15d0e6aeb91d2fa4cf1304e270b4d3b86676c8c8","internal_samples":95719,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.001864118999947},"started_at":"2026-10-13T05:15:03.326185+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008892580999599886,"native_input_samples":288000,"native_input_sha256":"c5ecae1205611a8741dadf28d7c3c0dee746c8e952ab38b9823bea948627ee57","output_sha256":"d612d740f2b0c0daca54ad9e2fdb987ef631c006c25831791f1993e83f765ac8","output_samples":287157,"length_delta_samples":-843,"analysis_s":3.6810314470003505,"synthesis_s":1.037621917999786,"total_s":4.7186533650001365,"rtf":0.7864422275000228,"output_peak":0.40723175856647925,"output_rms":0.04098954851865542,"finished_at":"2026-10-13T05:15:08.064302+00:00","evaluation_input":{"reference":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"output_sha256":"d612d740f2b0c0daca54ad9e2fdb987ef631c006c25831791f1993e83f765ac8","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95719,"length_delta_samples":-281,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.756871186452198,"output_peak":0.40771764030897195},"common_prefix":{"samples":95719,"sample_rate":16000,"reference_coverage":0.9970729166666666,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95719,"output_samples":95719,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.221283757831347,"spectral_convergence":0.5142405334019274,"spectral_convergence_status":"ok","amplitude_floor":1.1114078093309192e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2349370345457995,"spectral_convergence":0.17582858475548963,"spectral_convergence_status":"ok","amplitude_floor":9.683415126458649e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.779346677518052,"spectral_convergence":0.19321910278741158,"spectral_convergence_status":"ok","amplitude_floor":7.055286564887976e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010992281577898044,"output_power":0.00010185775982789218,"delta_db":-0.33093724284858284,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000797277613991692,"output_power":0.0007283837411658226,"delta_db":-0.3924932712051822,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005474354795756432,"output_power":0.000509143001582694,"delta_db":-0.3149316227900565,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020711147216214361,"output_power":0.0002069398970028391,"delta_db":-0.0035992707539111574,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.546201279627045e-05,"output_power":0.00010446168384959363,"delta_db":0.39126434948289845,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.706344070312761e-05,"output_power":2.5363916424099266e-05,"delta_db":-1.6472942014267602,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5134199930845652,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2957560068057594,"pitch":null,"output_rms_dbfs":-27.756871186452198,"output_peak":0.40771764030897195,"reference_rms_dbfs":-27.46111517964644,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.745189156631732},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95719,"output_samples":95719,"intelligibility":{"estoi":{"value":0.9366356212560782,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042358857810472855,"rms_dbfs":-27.46111517964644,"peak":0.384185791015625,"peak_dbfs":-8.30917402830067,"crest_factor_db":19.15194115134577,"mean":-4.91308809020414e-07,"zero_energy":false},"output":{"rms":0.040940810933015735,"rms_dbfs":-27.756871186452198,"peak":0.40771764030897195,"peak_dbfs":-7.792809959178063,"crest_factor_db":19.964061227274136,"mean":-0.00046263668870337205,"zero_energy":false},"rms_difference_db":-0.29575600680575675},"full_scale":{"reference":{"sample_count":95719,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95719,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":39,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.314016800373793e-06,"threshold_dbfs":-56.35633492280058,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":130,"quiet_samples":20800,"reference_mean_power":7.604699707231842e-07,"output_mean_power":5.940369079709988e-07,"output_rms_dbfs":-62.26186571129602,"output_zero_energy":false,"output_peak":0.00407431501492798,"output_energy_fraction":7.701329937911909e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":320,"output_estimated_voiced_frames":281,"both_estimated_voiced_frames":279,"reference_voiced_output_unvoiced_frames":41,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.871875,"absolute_difference_cents_median":3.245501669762413,"absolute_difference_cents_p95":11.557079879778529,"difference_over_600_cents_frames":0,"uncovered_tail_samples":39,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95719,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1195,"rms_floor":0.00015211892717126925,"log_rms_mae_db":1.201025638674188,"rms_envelope_nrmse":0.09762538685465785,"envelope_correlation":0.9911638568046406,"rms_gain_error_db":-0.2955027969087416,"reference_mean_power":0.0017953438244986048,"output_mean_power":0.0016772482695963451},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":8.127764515360079e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033820609098445675,"output_power":0.00032294458622823,"reference_active":true,"delta_db":-0.20053416314718578},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.863825424996463e-05,"output_power":3.465689874758514e-05,"reference_active":true,"delta_db":-0.47227798405923066},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.359586121510318e-06,"output_power":8.562561887487061e-06,"reference_active":true,"delta_db":0.6575033169866595}],"active_band_count":3,"active_band_mae_db":0.4434384880643587}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1189,"rms_floor":0.00012855597503212083,"log_rms_mae_db":0.8471902801125364,"rms_envelope_nrmse":0.05844000014132495,"envelope_correlation":0.9975443690790781,"rms_gain_error_db":-0.2952901853768167,"reference_mean_power":0.0018013151775683,"output_mean_power":0.001682909220089763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":7.544839700011258e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030576986073331693,"output_power":0.00029318644459688087,"reference_active":true,"delta_db":-0.18250788509477398},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7489774833058925e-05,"output_power":1.5583632875816213e-05,"reference_active":true,"delta_db":-0.5011550991120731},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0269163871478593e-07,"output_power":4.364552726520877e-07,"reference_active":true,"delta_db":0.34967132178491334}],"active_band_count":3,"active_band_mae_db":0.3444447686639201}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1195,"rms_floor":0.00013968182745477764,"log_rms_mae_db":1.4888216470850135,"rms_envelope_nrmse":0.1029432426990547,"envelope_correlation":0.9921564545855328,"rms_gain_error_db":-0.3725881069387982,"reference_mean_power":0.0010915350976250062,"output_mean_power":0.001001795079038754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":6.085331844481043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003197926649559035,"output_power":0.0002961736766577636,"reference_active":true,"delta_db":-0.33322041517850814},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.935471611526678e-05,"output_power":3.6604747891282186e-05,"reference_active":true,"delta_db":-0.3145936389251822},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.332648490146207e-06,"output_power":6.108591976958533e-06,"reference_active":true,"delta_db":-0.15644264643738584}],"active_band_count":3,"active_band_mae_db":0.2680855668470254}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1189,"rms_floor":0.00012087571887341512,"log_rms_mae_db":1.0129132030653891,"rms_envelope_nrmse":0.06551069159109683,"envelope_correlation":0.9976391677209508,"rms_gain_error_db":-0.3723446919682951,"reference_mean_power":0.0010957289400323465,"output_mean_power":0.0010057004930572298},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":5.649911848124251e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002963794992589866,"output_power":0.0002751536529276155,"reference_active":true,"delta_db":-0.3227287708663451},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.880085299033583e-05,"output_power":1.7227501585116943e-05,"reference_active":true,"delta_db":-0.37955255003972266},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.146011518297998e-07,"output_power":3.774661082296913e-07,"reference_active":true,"delta_db":-0.40752540837652124}],"active_band_count":3,"active_band_mae_db":0.36993557642752967}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1195,"rms_floor":7.190065013136086e-05,"log_rms_mae_db":1.1366055114490876,"rms_envelope_nrmse":0.11813083297933862,"envelope_correlation":0.9896498938096948,"rms_gain_error_db":0.10963333387027986,"reference_mean_power":0.00025714056584733995,"output_mean_power":0.0002637144508560669},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":2.2674511796866876e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.159241028219422e-05,"output_power":7.318479018712214e-05,"reference_active":true,"delta_db":0.09553848156324962},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.2220574630443455e-06,"output_power":5.869865933136158e-06,"reference_active":true,"delta_db":0.5078653568947957},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7874137320279083e-06,"output_power":2.2531846465546434e-06,"reference_active":true,"delta_db":1.0057169311075937}],"active_band_count":3,"active_band_mae_db":0.536373589855213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1189,"rms_floor":6.863962912697812e-05,"log_rms_mae_db":0.6541605392491093,"rms_envelope_nrmse":0.05522988359477951,"envelope_correlation":0.9977366801376815,"rms_gain_error_db":0.10916568741752927,"reference_mean_power":0.0002582366280430311,"output_mean_power":0.0002648100180847502},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":2.182196049608346e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.603917780966654e-05,"output_power":6.742907113104715e-05,"reference_active":true,"delta_db":0.09045519454390294},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3035091678355327e-06,"output_power":2.6334967790841143e-06,"reference_active":true,"delta_db":0.5814284646756146},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.912467977134815e-08,"output_power":1.207265465568192e-07,"reference_active":true,"delta_db":0.8562098028765154}],"active_band_count":3,"active_band_mae_db":0.5093644873653443}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1195,"rms_floor":3.534983553210223e-05,"log_rms_mae_db":1.1100680570453372,"rms_envelope_nrmse":0.18922976144516981,"envelope_correlation":0.9779568283584363,"rms_gain_error_db":0.6158339697308167,"reference_mean_power":1.9066508062216836e-05,"output_mean_power":2.1971239510247193e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.200564193576383e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.32556149533623e-06,"output_power":6.167746179047262e-06,"reference_active":true,"delta_db":0.6376108897515437},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6017247357323325e-06,"output_power":1.6305539185229385e-06,"reference_active":true,"delta_db":0.0774728194083941},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.153577341046606e-07,"output_power":5.637230563913784e-07,"reference_active":true,"delta_db":1.3264349632895263}],"active_band_count":3,"active_band_mae_db":0.6805062241498213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":39,"metrics":{"frame_count":1189,"rms_floor":2.383126672012766e-05,"log_rms_mae_db":0.6862876761036495,"rms_envelope_nrmse":0.12143808831763167,"envelope_correlation":0.9928888756054828,"rms_gain_error_db":0.6156057213103823,"reference_mean_power":1.915033600968805e-05,"output_mean_power":2.206667865109407e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":1.1118341081926863e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.413385751000457e-06,"output_power":6.2579709975884835e-06,"reference_active":true,"delta_db":0.6296457062030323},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.51963048229585e-07,"output_power":7.77452443031195e-07,"reference_active":true,"delta_db":0.14477332896295375},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.475230021887805e-08,"output_power":3.1524406086137284e-08,"reference_active":true,"delta_db":1.050313493821651}],"active_band_count":3,"active_band_mae_db":0.6082441763292123}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},{"case":"librispeech_4077-13751-0001","method":"hnm","status":"measured","mos":{"status":"measured","value":2.0906758308410645,"predicted_mos":2.0906758308410645,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bce5045ded15b2a17e66e33484467ebd8cb8c69e8f9004f634fc240c46d236ab","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5740870329991594},"started_at":"2026-10-13T04:39:07.979185+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4074000318942126e-05,"native_input_samples":96000,"native_input_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","output_sha256":"18f906b797d0e0cb75a0c35fe692d88a88d492c04f650c771dc0112f24c33573","output_samples":96000,"length_delta_samples":0,"analysis_s":1.126463364999836,"synthesis_s":0.5086694370002078,"total_s":1.6351328020000437,"rtf":0.27252213366667394,"output_peak":0.713193763566102,"output_rms":0.057087130296726026,"finished_at":"2026-10-13T04:39:09.620146+00:00","evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"18f906b797d0e0cb75a0c35fe692d88a88d492c04f650c771dc0112f24c33573","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.385718691451072,"spectral_convergence":0.4243951466922347,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.8554188858881755,"spectral_convergence":0.3752578671284084,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.891583037993778,"spectral_convergence":0.3933863940854157,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0010073929438901862,"delta_db":0.15528296087005636,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.000787577250914993,"delta_db":-0.0007669227740267845,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.000790921290137599,"delta_db":-0.19739742304095073,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.0001435551493116425,"delta_db":-0.28750377423060824,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00015251562809259474,"delta_db":-0.016034121590337132,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00037697818316835774,"delta_db":0.14317092032099638,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.13335935380449593,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.0004013588204409619,"pitch":null,"output_rms_dbfs":-24.869235758405836,"output_peak":0.713193763566102,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.044240205111009},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.385718691451072,"spectral_convergence":0.4243951466922347,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.8554188858881755,"spectral_convergence":0.3752578671284084,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.891583037993778,"spectral_convergence":0.3933863940854157,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0010073929438901862,"delta_db":0.15528296087005636,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.000787577250914993,"delta_db":-0.0007669227740267845,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.000790921290137599,"delta_db":-0.19739742304095073,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.0001435551493116425,"delta_db":-0.28750377423060824,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00015251562809259474,"delta_db":-0.016034121590337132,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00037697818316835774,"delta_db":0.14317092032099638,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.13335935380449593,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.0004013588204409619,"pitch":null,"output_rms_dbfs":-24.869235758405836,"output_peak":0.713193763566102,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.044240205111009},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7395292426520407,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057084492467456845,"rms_dbfs":-24.869637117226276,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.347149433040023,"mean":-0.0066053005854288735,"zero_energy":false},"output":{"rms":0.057087130296726026,"rms_dbfs":-24.869235758405836,"peak":0.713193763566102,"peak_dbfs":-2.9358492622799996,"crest_factor_db":21.93338649612584,"mean":6.995812256706028e-05,"zero_energy":false},"rms_difference_db":0.00040135882044012305},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":5.33711977835094e-06,"output_rms_dbfs":-52.72693050429851,"output_zero_energy":false,"output_peak":0.007040907288483417,"output_energy_fraction":2.729475969045577e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":83,"both_estimated_voiced_frames":74,"reference_voiced_output_unvoiced_frames":39,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.6548672566371682,"absolute_difference_cents_median":2.465730649366586,"absolute_difference_cents_p95":8.072653444139796,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002718271425028404,"log_rms_mae_db":1.80679702951579,"rms_envelope_nrmse":0.12729430308200748,"envelope_correlation":0.9848807276137701,"rms_gain_error_db":0.003721325876206735,"reference_mean_power":0.0032374876277477364,"output_mean_power":0.0032402629127485268},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2735378968683332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666640237447372,"output_power":0.0004849715309775801,"reference_active":true,"delta_db":0.16711923829153694},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002202584731702016,"output_power":0.0002484465093478995,"reference_active":true,"delta_db":0.5230027495173943},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.478672520125922e-05,"output_power":4.419598421761229e-05,"reference_active":true,"delta_db":1.0396926333238685}],"active_band_count":3,"active_band_mae_db":0.5766048737109333}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002399603663054526,"log_rms_mae_db":0.9529140826416245,"rms_envelope_nrmse":0.07251643017738443,"envelope_correlation":0.9945203274507389,"rms_gain_error_db":0.013763606159770786,"reference_mean_power":0.0031195469748277875,"output_mean_power":0.003129449086445344},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1403707959359127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046404851798205753,"output_power":0.00048542667806112657,"reference_active":true,"delta_db":0.19560250574858296},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.861370772885771e-05,"output_power":0.00010864992432691737,"reference_active":true,"delta_db":0.4209214024153088},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.958540928627598e-06,"output_power":2.342245935439438e-06,"reference_active":true,"delta_db":0.7769984232822974}],"active_band_count":3,"active_band_mae_db":0.4645074438153964}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022360065406936124,"log_rms_mae_db":1.9798483079559233,"rms_envelope_nrmse":0.12940038335986975,"envelope_correlation":0.9892189267026655,"rms_gain_error_db":-0.12167262855295728,"reference_mean_power":0.0012597240435666534,"output_mean_power":0.0012249212128175842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2605416603296333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001978275618555469,"output_power":0.0001908931699665366,"reference_active":true,"delta_db":-0.1549640862697396},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.107555923758901e-05,"output_power":7.359567282398734e-05,"reference_active":true,"delta_db":-0.4203767295764139},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1829892137914497e-05,"output_power":1.4574159604750748e-05,"reference_active":true,"delta_db":0.9060273641263041}],"active_band_count":3,"active_band_mae_db":0.49378939332415256}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019906826982741464,"log_rms_mae_db":1.2999100786034243,"rms_envelope_nrmse":0.07251224334620697,"envelope_correlation":0.9965522506472522,"rms_gain_error_db":-0.10949763503205628,"reference_mean_power":0.0011746644654148715,"output_mean_power":0.0011454181687936085},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9316122263011187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021214336491155288,"output_power":0.00020347900809078047,"reference_active":true,"delta_db":-0.18109841243720035},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6617395760512454e-05,"output_power":3.274013008822177e-05,"reference_active":true,"delta_db":-0.48607053231923014},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023494047314193e-06,"output_power":8.529245272371998e-07,"reference_active":true,"delta_db":-0.7010853335993532}],"active_band_count":3,"active_band_mae_db":0.4560847594519279}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.946276793614853e-05,"log_rms_mae_db":2.119558776953792,"rms_envelope_nrmse":0.18689428612226772,"envelope_correlation":0.975854578634764,"rms_gain_error_db":-0.2352677665712026,"reference_mean_power":0.00019501605476176574,"output_mean_power":0.00018473261995158798},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.789505238315191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.753576467028988e-05,"output_power":3.5652836047653556e-05,"reference_active":true,"delta_db":-0.22351185605564677},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6240794510593083e-05,"output_power":1.4369233036724758e-05,"reference_active":true,"delta_db":-0.5317368328328427},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.016250052718377e-06,"output_power":3.94295313347884e-06,"reference_active":true,"delta_db":-0.07999129244458336}],"active_band_count":3,"active_band_mae_db":0.2784133271110243}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.236492865510778e-05,"log_rms_mae_db":1.3599609586523458,"rms_envelope_nrmse":0.12202543008584503,"envelope_correlation":0.9891916299597906,"rms_gain_error_db":-0.2238436452126057,"reference_mean_power":0.00017673870709398643,"output_mean_power":0.00016786003492264448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.939463723397027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.781437846450254e-05,"output_power":3.5523968647049844e-05,"reference_active":true,"delta_db":-0.2713548842839979},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.773513697250703e-06,"output_power":6.913264721809271e-06,"reference_active":true,"delta_db":-0.5093418099702834},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2492076040039023e-07,"output_power":2.7867971584031384e-07,"reference_active":true,"delta_db":-0.6667210189899064}],"active_band_count":3,"active_band_mae_db":0.48247257108139596}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014584429481341945,"log_rms_mae_db":2.0909228359748147,"rms_envelope_nrmse":0.1760662885654409,"envelope_correlation":0.9802480743862702,"rms_gain_error_db":-0.0847291337533925,"reference_mean_power":0.0002465230996620666,"output_mean_power":0.00024176014436269589},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.909673910712007e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.093194518056522e-05,"output_power":7.695132495487585e-05,"reference_active":true,"delta_db":0.3537423219945484},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.832881668922569e-05,"output_power":2.5323818519115423e-05,"reference_active":true,"delta_db":-0.48699241643998975},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4378297018808916e-05,"output_power":9.888399719608893e-06,"reference_active":true,"delta_db":-1.625814371586726}],"active_band_count":3,"active_band_mae_db":0.8221830366737547}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":1.3589617231429179,"rms_envelope_nrmse":0.09015322778038559,"envelope_correlation":0.99457278305258,"rms_gain_error_db":-0.08314198817126277,"reference_mean_power":0.0002460055124763466,"output_mean_power":0.00024134074000261438},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8079454178991006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.471033966073092e-05,"output_power":7.80372458103198e-05,"reference_active":true,"delta_db":0.18921222443666105},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5789672835999793e-05,"output_power":1.3128275718639611e-05,"reference_active":true,"delta_db":-0.801654423184403},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1060211769697672e-06,"output_power":6.965858357565686e-07,"reference_active":true,"delta_db":-2.007888031268851}],"active_band_count":3,"active_band_mae_db":0.999584892963305}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13751-0001","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.958972930908203,"predicted_mos":2.958972930908203,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4b5f68537f7393893d7e00e8f1df95ce8f51c1ea380751be240f8e3da10dbb0c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9679003000001103},"started_at":"2026-10-13T04:42:04.719793+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.533899987611221e-05,"native_input_samples":96000,"native_input_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","output_sha256":"cfdf337344880d65d1c643e80198e09cc0d437f71962316c8f14696552f4e53a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4731489440000587,"synthesis_s":0.08600554699933127,"total_s":1.55915449099939,"rtf":0.25985908183323164,"output_peak":1.0006982926520536,"output_rms":0.05905723848034595,"finished_at":"2026-10-13T04:42:06.286260+00:00","evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"cfdf337344880d65d1c643e80198e09cc0d437f71962316c8f14696552f4e53a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.118312660548824,"spectral_convergence":0.6658888128642284,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.795371578004659,"spectral_convergence":0.42419339925023153,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.937620741924635,"spectral_convergence":0.422724173429619,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0008197056905897861,"delta_db":-0.7401265843310985,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0009089773021512769,"delta_db":0.6218318332183543,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.0010114314299727333,"delta_db":0.8706343668967437,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.00017021712530101033,"delta_db":0.45234099048995724,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00016385177732023017,"delta_db":0.2953339683505871,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00041357409158941617,"delta_db":0.5455419948828458,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5876349563615978,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.2950998266354605,"pitch":null,"output_rms_dbfs":-24.57453729059082,"output_peak":1.0006982926520536,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.283768326826039},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.118312660548824,"spectral_convergence":0.6658888128642284,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.795371578004659,"spectral_convergence":0.42419339925023153,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.937620741924635,"spectral_convergence":0.422724173429619,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0008197056905897861,"delta_db":-0.7401265843310985,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0009089773021512769,"delta_db":0.6218318332183543,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.0010114314299727333,"delta_db":0.8706343668967437,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.00017021712530101033,"delta_db":0.45234099048995724,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00016385177732023017,"delta_db":0.2953339683505871,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00041357409158941617,"delta_db":0.5455419948828458,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5876349563615978,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.2950998266354605,"pitch":null,"output_rms_dbfs":-24.57453729059082,"output_peak":1.0006982926520536,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.283768326826039},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.830817946447904,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057084492467456845,"rms_dbfs":-24.869637117226276,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.347149433040023,"mean":-0.0066053005854288735,"zero_energy":false},"output":{"rms":0.05905723848034595,"rms_dbfs":-24.574537290590822,"peak":1.0006982926520536,"peak_dbfs":0.006063176221394738,"crest_factor_db":24.580600466812214,"mean":1.1202915327012828e-05,"zero_energy":false},"rms_difference_db":0.29509982663545387},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":1,"at_or_above_full_scale_fraction":1.0416666666666666e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":1.0416666666666666e-05,"longest_full_scale_run_samples":1,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":3.0959881515944354e-06,"output_rms_dbfs":-55.0920071003921,"output_zero_energy":false,"output_peak":0.010167157134612028,"output_energy_fraction":1.4794550296469657e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":143,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9380530973451328,"absolute_difference_cents_median":5.552163400115511,"absolute_difference_cents_p95":15.704526566078716,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002718271425028404,"log_rms_mae_db":6.643484664174143,"rms_envelope_nrmse":0.19181807696798459,"envelope_correlation":0.974231639485273,"rms_gain_error_db":0.32671008750325975,"reference_mean_power":0.0032374876277477364,"output_mean_power":0.0034904315546415957},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2735378968683332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666640237447372,"output_power":0.0005558610960371334,"reference_active":true,"delta_db":0.759619580550251},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002202584731702016,"output_power":0.00020343054555840452,"reference_active":true,"delta_db":-0.3451646058684673},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.478672520125922e-05,"output_power":3.615480419389182e-05,"reference_active":true,"delta_db":0.1675246735358574}],"active_band_count":3,"active_band_mae_db":0.42410295331819187}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002399603663054526,"log_rms_mae_db":6.054183914227729,"rms_envelope_nrmse":0.13781044941440634,"envelope_correlation":0.9899108818658773,"rms_gain_error_db":0.3628812076993909,"reference_mean_power":0.0031195469748277875,"output_mean_power":0.0033914049560254214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1403707959359127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046404851798205753,"output_power":0.0005502463595829796,"reference_active":true,"delta_db":0.7399378793601494},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.861370772885771e-05,"output_power":9.268655996805866e-05,"reference_active":true,"delta_db":-0.2692052417793365},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.958540928627598e-06,"output_power":1.8890513536617545e-06,"reference_active":true,"delta_db":-0.1568888731977774}],"active_band_count":3,"active_band_mae_db":0.3886773314457544}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022360065406936124,"log_rms_mae_db":2.220080076563564,"rms_envelope_nrmse":0.20812686005315234,"envelope_correlation":0.981678085502618,"rms_gain_error_db":0.8970627100174634,"reference_mean_power":0.0012597240435666534,"output_mean_power":0.0015487513200461533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2605416603296333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001978275618555469,"output_power":0.0002431285926669676,"reference_active":true,"delta_db":0.8954923768319876},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.107555923758901e-05,"output_power":9.317815891976165e-05,"reference_active":true,"delta_db":0.6042417209529616},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1829892137914497e-05,"output_power":9.82643326003786e-06,"reference_active":true,"delta_db":-0.8058487603295592}],"active_band_count":3,"active_band_mae_db":0.7685276193715028}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019906826982741464,"log_rms_mae_db":1.738676837900632,"rms_envelope_nrmse":0.1502928224642399,"envelope_correlation":0.9951459014222381,"rms_gain_error_db":0.9905233551320265,"reference_mean_power":0.0011746644654148715,"output_mean_power":0.0014755915742082175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9316122263011187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021214336491155288,"output_power":0.0002560594991747166,"reference_active":true,"delta_db":0.817114385901053},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6617395760512454e-05,"output_power":4.1091172617355905e-05,"reference_active":true,"delta_db":0.5006108096690197},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023494047314193e-06,"output_power":7.888635095756542e-07,"reference_active":true,"delta_db":-1.0401726941202618}],"active_band_count":3,"active_band_mae_db":0.7859659632301116}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.946276793614853e-05,"log_rms_mae_db":1.7885648886026688,"rms_envelope_nrmse":0.19827196645778122,"envelope_correlation":0.9770303551705064,"rms_gain_error_db":0.5263297226889053,"reference_mean_power":0.00019501605476176574,"output_mean_power":0.00022014221837017608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.789505238315191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.753576467028988e-05,"output_power":4.409052585271821e-05,"reference_active":true,"delta_db":0.6990001089632238},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6240794510593083e-05,"output_power":1.7845347464266536e-05,"reference_active":true,"delta_db":0.4091773705809352},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.016250052718377e-06,"output_power":3.979967950005588e-06,"reference_active":true,"delta_db":-0.039411694709150624}],"active_band_count":3,"active_band_mae_db":0.3825297247511032}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.236492865510778e-05,"log_rms_mae_db":1.088622125849076,"rms_envelope_nrmse":0.1085319283638137,"envelope_correlation":0.9951847388236083,"rms_gain_error_db":0.5866062884951361,"reference_mean_power":0.00017673870709398643,"output_mean_power":0.00020229833227087312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.939463723397027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.781437846450254e-05,"output_power":4.401023955351351e-05,"reference_active":true,"delta_db":0.658967659756034},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.773513697250703e-06,"output_power":8.620237079693785e-06,"reference_active":true,"delta_db":0.4490184211059264},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2492076040039023e-07,"output_power":3.03979896664878e-07,"reference_active":true,"delta_db":-0.2893259789814567}],"active_band_count":3,"active_band_mae_db":0.46577068661447235}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014584429481341945,"log_rms_mae_db":1.8927916065030277,"rms_envelope_nrmse":0.23610796691588296,"envelope_correlation":0.9677201212018376,"rms_gain_error_db":0.420255678439034,"reference_mean_power":0.0002465230996620666,"output_mean_power":0.00027157087247454986},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.909673910712007e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.093194518056522e-05,"output_power":8.187449205909658e-05,"reference_active":true,"delta_db":0.6230674893592351},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.832881668922569e-05,"output_power":3.19184184038398e-05,"reference_active":true,"delta_db":0.5181292922407874},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4378297018808916e-05,"output_power":1.1414025727976166e-05,"reference_active":true,"delta_db":-1.0026860363760663}],"active_band_count":3,"active_band_mae_db":0.7146276059920296}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":1.134502405565242,"rms_envelope_nrmse":0.12246042529040453,"envelope_correlation":0.9920642626982847,"rms_gain_error_db":0.43166663532696986,"reference_mean_power":0.0002460055124763466,"output_mean_power":0.0002717136787151361},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8079454178991006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.471033966073092e-05,"output_power":8.462552313018577e-05,"reference_active":true,"delta_db":0.5412065550636374},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5789672835999793e-05,"output_power":1.6359630082787107e-05,"reference_active":true,"delta_db":0.1540034789307307},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1060211769697672e-06,"output_power":8.662878165188598e-07,"reference_active":true,"delta_db":-1.0610123595807428}],"active_band_count":3,"active_band_mae_db":0.585407464525037}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13751-0001","method":"facodec","status":"measured","mos":{"status":"measured","value":3.6089601516723633,"predicted_mos":3.6089601516723633,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a865322f07433131f471671da7e8056e0aa0d1b71bac71087ae23f98e2ba7b97","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.932519781999872},"started_at":"2026-10-13T04:51:47.597470+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.843000038410537e-05,"native_input_samples":96000,"native_input_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","output_sha256":"bb63e197cf366265bcc4e49e03a5b4c0859589c5a7ebe4e2d0ed3ac65aa7ee9d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.97098596899923,"synthesis_s":5.744589171000371,"total_s":7.715575139999601,"rtf":1.2859291899999334,"output_peak":0.5464245080947876,"output_rms":0.052812407703309174,"finished_at":"2026-10-13T04:51:55.322247+00:00","evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"bb63e197cf366265bcc4e49e03a5b4c0859589c5a7ebe4e2d0ed3ac65aa7ee9d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.791733699984123,"spectral_convergence":0.37884345644532613,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.151054831313444,"spectral_convergence":0.38239740003386136,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.142436563029009,"spectral_convergence":0.39777069279105964,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0008280866832131005,"delta_db":-0.6959480786632111,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0008669178951482184,"delta_db":0.41608112490609583,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.000630943992959777,"delta_db":-1.1788219849354047,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.00011288513103781286,"delta_db":-1.3313241119192436,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00011487117729879983,"delta_db":-1.2470668763994084,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00023544552776284106,"delta_db":-1.9010867948040555,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.128388161937903,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6756435366298917,"pitch":null,"output_rms_dbfs":-25.545280653856175,"output_peak":0.5464245080947876,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.028408364775525},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.791733699984123,"spectral_convergence":0.37884345644532613,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.151054831313444,"spectral_convergence":0.38239740003386136,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.142436563029009,"spectral_convergence":0.39777069279105964,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0008280866832131005,"delta_db":-0.6959480786632111,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0008669178951482184,"delta_db":0.41608112490609583,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.000630943992959777,"delta_db":-1.1788219849354047,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.00011288513103781286,"delta_db":-1.3313241119192436,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00011487117729879983,"delta_db":-1.2470668763994084,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00023544552776284106,"delta_db":-1.9010867948040555,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.128388161937903,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6756435366298917,"pitch":null,"output_rms_dbfs":-25.545280653856175,"output_peak":0.5464245080947876,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.028408364775525},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7863983015142191,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057084492467456845,"rms_dbfs":-24.869637117226276,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.347149433040023,"mean":-0.0066053005854288735,"zero_energy":false},"output":{"rms":0.052812407703309174,"rms_dbfs":-25.545280653856175,"peak":0.5464245080947876,"peak_dbfs":-5.249396600453256,"crest_factor_db":20.295884053402915,"mean":-0.0013165851415215192,"zero_energy":false},"rms_difference_db":-0.6756435366298987},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":7.774716417104093e-06,"output_rms_dbfs":-51.09315442909862,"output_zero_energy":false,"output_peak":0.009143740870058537,"output_energy_fraction":4.645809225884351e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":131,"both_estimated_voiced_frames":105,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.9292035398230089,"absolute_difference_cents_median":6.749486440481618,"absolute_difference_cents_p95":20.456143900264077,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002718271425028404,"log_rms_mae_db":2.9814659977637925,"rms_envelope_nrmse":0.19026983717646168,"envelope_correlation":0.9685406918724485,"rms_gain_error_db":-0.662442051417126,"reference_mean_power":0.0032374876277477364,"output_mean_power":0.0027794822968797623},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2735378968683332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666640237447372,"output_power":0.00047842255256285457,"reference_active":true,"delta_db":0.10807322487537871},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002202584731702016,"output_power":0.0002252079325090482,"reference_active":true,"delta_db":0.09651059250084797},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.478672520125922e-05,"output_power":3.759398681564181e-05,"reference_active":true,"delta_db":0.33704838294342704}],"active_band_count":3,"active_band_mae_db":0.18054406677321788}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002399603663054526,"log_rms_mae_db":2.5623840706947636,"rms_envelope_nrmse":0.17284817206362743,"envelope_correlation":0.9714234813389597,"rms_gain_error_db":-0.6073670873615691,"reference_mean_power":0.0031195469748277875,"output_mean_power":0.0027124067781391314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1403707959359127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046404851798205753,"output_power":0.0004877940810793368,"reference_active":true,"delta_db":0.2167313623938842},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.861370772885771e-05,"output_power":0.00010047878669991666,"reference_active":true,"delta_db":0.08137094302575774},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.958540928627598e-06,"output_power":2.527803141188993e-06,"reference_active":true,"delta_db":1.1081059753757947}],"active_band_count":3,"active_band_mae_db":0.46873609359847884}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022360065406936124,"log_rms_mae_db":1.8964651401462327,"rms_envelope_nrmse":0.18814098192833034,"envelope_correlation":0.9790632320019833,"rms_gain_error_db":-0.5793040328142727,"reference_mean_power":0.0012597240435666534,"output_mean_power":0.0011024147498027682},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2605416603296333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001978275618555469,"output_power":0.00023995857391384,"reference_active":true,"delta_db":0.838494737140968},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.107555923758901e-05,"output_power":9.835644191218313e-05,"reference_active":true,"delta_db":0.8391285655287727},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1829892137914497e-05,"output_power":1.7392246988390585e-05,"reference_active":true,"delta_db":1.6737490937489887}],"active_band_count":3,"active_band_mae_db":1.1171241321395764}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019906826982741464,"log_rms_mae_db":1.5962934404763414,"rms_envelope_nrmse":0.172888827117369,"envelope_correlation":0.9812606229101061,"rms_gain_error_db":-0.4741830689005795,"reference_mean_power":0.0011746644654148715,"output_mean_power":0.0010531628334875104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9316122263011187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021214336491155288,"output_power":0.00025998217957388025,"reference_active":true,"delta_db":0.8831412723191777},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6617395760512454e-05,"output_power":4.485561288488196e-05,"reference_active":true,"delta_db":0.8812934111874853},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023494047314193e-06,"output_power":1.3852671362858682e-06,"reference_active":true,"delta_db":1.405143944580739}],"active_band_count":3,"active_band_mae_db":1.0565262093624674}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.946276793614853e-05,"log_rms_mae_db":1.980267992516871,"rms_envelope_nrmse":0.21953628422621516,"envelope_correlation":0.9731741066569204,"rms_gain_error_db":-1.0866353087926932,"reference_mean_power":0.00019501605476176574,"output_mean_power":0.00015184721652994533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.789505238315191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.753576467028988e-05,"output_power":3.4076582391410695e-05,"reference_active":true,"delta_db":-0.4198923562692317},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6240794510593083e-05,"output_power":1.420875596832945e-05,"reference_active":true,"delta_db":-0.5805121597671411},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.016250052718377e-06,"output_power":5.294577108642522e-06,"reference_active":true,"delta_db":1.2001053336221075}],"active_band_count":3,"active_band_mae_db":0.7335032832194934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.236492865510778e-05,"log_rms_mae_db":1.6584921576717504,"rms_envelope_nrmse":0.19297904107986252,"envelope_correlation":0.9786420500415375,"rms_gain_error_db":-0.9920563454498056,"reference_mean_power":0.00017673870709398643,"output_mean_power":0.00014064556405103193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.939463723397027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.781437846450254e-05,"output_power":3.463504721939165e-05,"reference_active":true,"delta_db":-0.381411824507921},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.773513697250703e-06,"output_power":7.078509054658028e-06,"reference_active":true,"delta_db":-0.406755762061441},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2492076040039023e-07,"output_power":3.44314653266893e-07,"reference_active":true,"delta_db":0.25178044779241326}],"active_band_count":3,"active_band_mae_db":0.34664934478725845}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014584429481341945,"log_rms_mae_db":2.2932294662941977,"rms_envelope_nrmse":0.2938781045720221,"envelope_correlation":0.9506140707670612,"rms_gain_error_db":-1.3218975931608714,"reference_mean_power":0.0002465230996620666,"output_mean_power":0.00018183097203119478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.909673910712007e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.093194518056522e-05,"output_power":5.5414286718162635e-05,"reference_active":true,"delta_db":-1.0722012230069153},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.832881668922569e-05,"output_power":1.5456358833600148e-05,"reference_active":true,"delta_db":-2.631212424372663},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4378297018808916e-05,"output_power":7.417527655107235e-06,"reference_active":true,"delta_db":-2.874482765885555}],"active_band_count":3,"active_band_mae_db":2.1926321377550444}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":1.9440719275811897,"rms_envelope_nrmse":0.260027419856629,"envelope_correlation":0.9621894138117338,"rms_gain_error_db":-1.3135164135373145,"reference_mean_power":0.0002460055124763466,"output_mean_power":0.00018179971476637728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8079454178991006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.471033966073092e-05,"output_power":5.564529664849218e-05,"reference_active":true,"delta_db":-1.2795224887103485},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5789672835999793e-05,"output_power":7.840659865096944e-06,"reference_active":true,"delta_db":-3.040205172872266},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1060211769697672e-06,"output_power":5.208364180570201e-07,"reference_active":true,"delta_db":-3.270620990062738}],"active_band_count":3,"active_band_mae_db":2.530116217215118}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13751-0001","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.513288736343384,"predicted_mos":2.513288736343384,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"cc75c4b279ca451c2f564c9e6a65239440bb2bef3faf13c905b684ad32b6c174","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":11.314353996000136},"started_at":"2026-10-13T07:13:30.565586+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00843117499971413,"native_input_samples":264600,"native_input_sha256":"6951b77707f75a0a48529f8a20462602e0b44115d284d28ae072e2d43eb242cc","output_sha256":"f452006129b34bc327c83dfee4740dfc6b5edb51331ef292b65591663e782a18","output_samples":264704,"length_delta_samples":104,"analysis_s":0.19037341100010963,"synthesis_s":4.091849555999943,"total_s":4.2822229670000524,"rtf":0.7137038278333421,"output_peak":0.7797968983650208,"output_rms":0.04775340893864055,"finished_at":"2026-10-13T07:13:34.882760+00:00","evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"f452006129b34bc327c83dfee4740dfc6b5edb51331ef292b65591663e782a18","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.424050593394814,"output_peak":0.77017128480469},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.190728801583928,"spectral_convergence":0.5838394764241721,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.018125153621432,"spectral_convergence":0.5018106366183028,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.68421509806483,"spectral_convergence":0.5430159195889988,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0005731801565637725,"delta_db":-2.293794617480222,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0006455884907996408,"delta_db":-0.8641407595750504,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.0005969996367882325,"delta_db":-1.4189894149964408,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.0001163886811001319,"delta_db":-1.1985840578947355,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00012426594933049725,"delta_db":-0.9056561778788824,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00021534014347592538,"delta_db":-2.288741269303384,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.494984382854786,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.5667325522781472,"pitch":null,"output_rms_dbfs":-26.436369669504426,"output_peak":0.77017128480469,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.964356351090062},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.5886198760164965,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057084492467456845,"rms_dbfs":-24.869637117226276,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.347149433040023,"mean":-0.0066053005854288735,"zero_energy":false},"output":{"rms":0.04766301562069065,"rms_dbfs":-26.436369669504426,"peak":0.77017128480469,"peak_dbfs":-2.2682535543942834,"crest_factor_db":24.168116115110145,"mean":0.0012489296643324287,"zero_energy":false},"rms_difference_db":-1.5667325522781503},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":1.1097592664974433e-05,"output_rms_dbfs":-49.5477121992557,"output_zero_energy":false,"output_peak":0.009073247673249721,"output_energy_fraction":8.141688768707647e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":103,"both_estimated_voiced_frames":80,"reference_voiced_output_unvoiced_frames":33,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.7079646017699115,"absolute_difference_cents_median":14.020913440400433,"absolute_difference_cents_p95":44.48564876765504,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002718271425028404,"log_rms_mae_db":2.7162402409578372,"rms_envelope_nrmse":0.21021595380193225,"envelope_correlation":0.9821281425404778,"rms_gain_error_db":-1.557945491938188,"reference_mean_power":0.0032374876277477364,"output_mean_power":0.0022615884014878583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2735378968683332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666640237447372,"output_power":0.0003437394761275941,"reference_active":true,"delta_db":-1.3277491061446225},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002202584731702016,"output_power":0.00017351118282188868,"reference_active":true,"delta_db":-1.036051539987885},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.478672520125922e-05,"output_power":2.8297534927319773e-05,"reference_active":true,"delta_db":-0.896649418459242}],"active_band_count":3,"active_band_mae_db":1.0868166881972499}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002399603663054526,"log_rms_mae_db":2.310216087820748,"rms_envelope_nrmse":0.18824379697213373,"envelope_correlation":0.991388568300746,"rms_gain_error_db":-1.5779076719731457,"reference_mean_power":0.0031195469748277875,"output_mean_power":0.0021692058276887648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1403707959359127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046404851798205753,"output_power":0.0003475547437730608,"reference_active":true,"delta_db":-1.2554016944731183},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.861370772885771e-05,"output_power":8.060470808036214e-05,"reference_active":true,"delta_db":-0.8757687846767273},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.958540928627598e-06,"output_power":1.775613279135026e-06,"reference_active":true,"delta_db":-0.4258426736336861}],"active_band_count":3,"active_band_mae_db":0.8523377175945105}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022360065406936124,"log_rms_mae_db":3.5103998409595216,"rms_envelope_nrmse":0.2378118786160246,"envelope_correlation":0.9692683674637742,"rms_gain_error_db":-1.126787650031886,"reference_mean_power":0.0012597240435666534,"output_mean_power":0.000971844214359887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2605416603296333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001978275618555469,"output_power":0.00020456464866363533,"reference_active":true,"delta_db":0.14543785803619283},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.107555923758901e-05,"output_power":8.39110749352565e-05,"reference_active":true,"delta_db":0.1492933161423682},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1829892137914497e-05,"output_power":9.221653405024494e-06,"reference_active":true,"delta_db":-1.0817198958140406}],"active_band_count":3,"active_band_mae_db":0.4588170233308672}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019906826982741464,"log_rms_mae_db":2.8904937646746998,"rms_envelope_nrmse":0.21195610366496093,"envelope_correlation":0.9767585435879398,"rms_gain_error_db":-1.1346620433346741,"reference_mean_power":0.0011746644654148715,"output_mean_power":0.0009045813443987981},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9316122263011187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021214336491155288,"output_power":0.00022015397956586685,"reference_active":true,"delta_db":0.16097087165577953},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6617395760512454e-05,"output_power":3.729369513279319e-05,"reference_active":true,"delta_db":0.07947962327937493},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023494047314193e-06,"output_power":8.811444414028658e-07,"reference_active":true,"delta_db":-0.5597203092419062}],"active_band_count":3,"active_band_mae_db":0.26672360139235357}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.946276793614853e-05,"log_rms_mae_db":3.3960677005624933,"rms_envelope_nrmse":0.30768607020570227,"envelope_correlation":0.9385868841884893,"rms_gain_error_db":-1.2126480744444057,"reference_mean_power":0.00019501605476176574,"output_mean_power":0.0001475045981747244},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.789505238315191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.753576467028988e-05,"output_power":3.894728461084627e-05,"reference_active":true,"delta_db":0.16031916582350642},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6240794510593083e-05,"output_power":1.581992434086723e-05,"reference_active":true,"delta_db":-0.11402869259269255},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.016250052718377e-06,"output_power":3.5993796190468105e-06,"reference_active":true,"delta_db":-0.4759309106026368}],"active_band_count":3,"active_band_mae_db":0.25009292300627856}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.236492865510778e-05,"log_rms_mae_db":2.902108281007994,"rms_envelope_nrmse":0.24050770697579074,"envelope_correlation":0.961766245681122,"rms_gain_error_db":-1.0212286553684782,"reference_mean_power":0.00017673870709398643,"output_mean_power":0.00013970398942261795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.939463723397027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.781437846450254e-05,"output_power":4.0008523922601376e-05,"reference_active":true,"delta_db":0.2449556226661129},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.773513697250703e-06,"output_power":7.574198895586331e-06,"reference_active":true,"delta_db":-0.11280662801456431},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2492076040039023e-07,"output_power":2.429574384287647e-07,"reference_active":true,"delta_db":-1.262472608865028}],"active_band_count":3,"active_band_mae_db":0.540078286515235}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014584429481341945,"log_rms_mae_db":3.170796365816992,"rms_envelope_nrmse":0.2417330197896727,"envelope_correlation":0.9683316494464158,"rms_gain_error_db":-1.0710566007587605,"reference_mean_power":0.0002465230996620666,"output_mean_power":0.00019264243523587583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.909673910712007e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.093194518056522e-05,"output_power":5.5081249309198296e-05,"reference_active":true,"delta_db":-1.0983808773461632},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.832881668922569e-05,"output_power":2.4226188227138802e-05,"reference_active":true,"delta_db":-0.6794334699056358},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4378297018808916e-05,"output_power":1.216053041025125e-05,"reference_active":true,"delta_db":-0.7275493265114057}],"active_band_count":3,"active_band_mae_db":0.835121224587735}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":2.9961381034746637,"rms_envelope_nrmse":0.1865155547818671,"envelope_correlation":0.9844851593983881,"rms_gain_error_db":-1.0785280511258848,"reference_mean_power":0.0002460055124763466,"output_mean_power":0.00019190753794516354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8079454178991006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.471033966073092e-05,"output_power":5.9768292696354466e-05,"reference_active":true,"delta_db":-0.9690986064987652},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5789672835999793e-05,"output_power":1.4505428803758729e-05,"reference_active":true,"delta_db":-0.3684255969498178},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1060211769697672e-06,"output_power":1.168845132106541e-06,"reference_active":true,"delta_db":0.2399352999535255}],"active_band_count":3,"active_band_mae_db":0.5258198344673696}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13751-0001","method":"golf","status":"measured","mos":{"status":"measured","value":3.2650327682495117,"predicted_mos":3.2650327682495117,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"5730031be5fa764841eb33c94507ee435d3f27577c732deec5f7b557ae5de06b","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.154936724999061},"started_at":"2026-10-13T05:02:30.163804+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030314080004245625,"native_input_samples":144000,"native_input_sha256":"fe204849450e8ed1a14ee23513308f10e05cb68776d01fbbab4461c6b6bf6fc9","output_sha256":"e3cb14a91dd91fd2cf41a90ce0f29fccdaedcbc03c1d047ffdc4ad253b72892a","output_samples":143760,"length_delta_samples":-240,"analysis_s":1.0842254619992673,"synthesis_s":0.12117494900030579,"total_s":1.205400410999573,"rtf":0.20090006849992884,"output_peak":0.7654855251312256,"output_rms":0.054395787457819963,"finished_at":"2026-10-13T05:02:31.384061+00:00","analysis_stages":{"external_f0_s":0.4777266729997791},"evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"e3cb14a91dd91fd2cf41a90ce0f29fccdaedcbc03c1d047ffdc4ad253b72892a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.31168240376892,"output_peak":0.7497203868852556},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.7694884042515495,"spectral_convergence":0.6434590448384809,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.685828871930911,"spectral_convergence":0.4014146289448808,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.741372697340521,"spectral_convergence":0.39218431147480914,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0009696027788320808,"output_power":0.0008877510005157697,"delta_db":-0.3830268105602964,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007786753187283522,"output_power":0.0008520045922874982,"delta_db":0.39085526306304746,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007616481757281707,"output_power":0.0006614673920747982,"delta_db":-0.6124596629540201,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001409581891654615,"output_power":0.00011535318945915604,"delta_db":-0.8706070484032697,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015122741775140468,"output_power":0.00011343116828583346,"delta_db":-1.2489813139966612,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036615374917509504,"output_power":0.0003132738791093205,"delta_db":-0.6773930062883792,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.697220517544279,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.319898263429873,"pitch":null,"output_rms_dbfs":-25.31168240376892,"output_peak":0.7497203868852556,"reference_rms_dbfs":-24.991784140339046,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.065563324507661},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8384828880494086,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.056287348750679,"rms_dbfs":-24.991784140339046,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.46929645615279,"mean":-0.006627300625452414,"zero_energy":false},"output":{"rms":0.05425201583104886,"rms_dbfs":-25.31168240376892,"peak":0.7497203868852556,"peak_dbfs":-2.50201358749584,"crest_factor_db":22.809668816273078,"mean":-0.0003386671047380742,"zero_energy":false},"rms_difference_db":-0.3198982634298737},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":1.151039783829179e-05,"output_rms_dbfs":-49.389096654241726,"output_zero_energy":false,"output_peak":0.01377396492437242,"output_energy_fraction":6.528775762383055e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":118,"both_estimated_voiced_frames":102,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9107142857142857,"absolute_difference_cents_median":5.74619868291275,"absolute_difference_cents_p95":22.890164251452074,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002718271425028404,"log_rms_mae_db":3.5311614420700446,"rms_envelope_nrmse":0.18449783260870106,"envelope_correlation":0.970172723939877,"rms_gain_error_db":-0.345385639806352,"reference_mean_power":0.003156223375000699,"output_mean_power":0.0029149369667358666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.2692767557592677e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004676783638071212,"output_power":0.0005213070413841387,"reference_active":true,"delta_db":0.4714631243903631},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021919638437126605,"output_power":0.0002195213403692054,"reference_active":true,"delta_db":0.006433595663755286},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.442633807371493e-05,"output_power":4.808643774770244e-05,"reference_active":true,"delta_db":1.4513177635005787}],"active_band_count":3,"active_band_mae_db":0.6430714945182324}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00023897139757872587,"log_rms_mae_db":2.610563117280692,"rms_envelope_nrmse":0.12712532634325857,"envelope_correlation":0.9854049955675463,"rms_gain_error_db":-0.31169942557376695,"reference_mean_power":0.0030307012384845466,"output_mean_power":0.0028208057737450993},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1365060696267136e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000465114039345304,"output_power":0.0005259568098253937,"reference_active":true,"delta_db":0.5339063373554199},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.736448893192525e-05,"output_power":0.00010062235283941045,"reference_active":true,"delta_db":0.1429387958723399},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.926106606881102e-06,"output_power":2.829991946838289e-06,"reference_active":true,"delta_db":1.671048787267704}],"active_band_count":3,"active_band_mae_db":0.7826313068318212}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00022360082742982882,"log_rms_mae_db":1.7653339906248164,"rms_envelope_nrmse":0.17422145190191082,"envelope_correlation":0.9802278687818603,"rms_gain_error_db":-0.19433966326382007,"reference_mean_power":0.0012004807545095903,"output_mean_power":0.0011479454096084086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.231298673059606e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019804993035135894,"output_power":0.00023693634830684944,"reference_active":true,"delta_db":0.7785699675887481},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.065386016004797e-05,"output_power":0.00010324207866496006,"reference_active":true,"delta_db":1.0723158208514907},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.179669162858481e-05,"output_power":1.2903180397433824e-05,"reference_active":true,"delta_db":0.38936542232029847}],"active_band_count":3,"active_band_mae_db":0.7467504035868457}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00019889994077163902,"log_rms_mae_db":1.250462762057603,"rms_envelope_nrmse":0.13288336795780828,"envelope_correlation":0.9878815836028988,"rms_gain_error_db":-0.11423551504619371,"reference_mean_power":0.0011126299376596608,"output_mean_power":0.001083745207254147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.9038027808894174e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002123411194253693,"output_power":0.00025769940308972004,"reference_active":true,"delta_db":0.8407931004181691},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6248451512772905e-05,"output_power":4.599976120774861e-05,"reference_active":true,"delta_db":1.0346611839246769},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.9718765246508e-07,"output_power":1.2310524868943075e-06,"reference_active":true,"delta_db":0.9149967753893837}],"active_band_count":3,"active_band_mae_db":0.9301503532440766}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.946276793614847e-05,"log_rms_mae_db":1.8505536966656573,"rms_envelope_nrmse":0.2281326948425237,"envelope_correlation":0.9720074655237192,"rms_gain_error_db":-1.2161354013395762,"reference_mean_power":0.0001844492329845816,"output_mean_power":0.00013940017323204926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.799996273695452e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7664767053441886e-05,"output_power":3.0380154906604376e-05,"reference_active":true,"delta_db":-0.9334530180697913},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6242085801515647e-05,"output_power":1.3502860844943788e-05,"reference_active":true,"delta_db":-0.8021600837609256},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.967884042512802e-06,"output_power":3.357318991834818e-06,"reference_active":true,"delta_db":-0.725663646375164}],"active_band_count":3,"active_band_mae_db":0.8204255827352936}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.02722041337346e-05,"log_rms_mae_db":1.3384574479934204,"rms_envelope_nrmse":0.17807980301803572,"envelope_correlation":0.9863971191322395,"rms_gain_error_db":-1.1805554824424873,"reference_mean_power":0.00016356272460482283,"output_mean_power":0.00012463177725187212},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.95550302469356e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.794937620556857e-05,"output_power":3.1462812595891804e-05,"reference_active":true,"delta_db":-0.8140709808588288},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.703768294107527e-06,"output_power":6.611984915577714e-06,"reference_active":true,"delta_db":-0.663713579456877},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2453605563815325e-07,"output_power":2.669017269979145e-07,"reference_active":true,"delta_db":-0.8491156959293458}],"active_band_count":3,"active_band_mae_db":0.7756334187483506}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014584429481341945,"log_rms_mae_db":2.345848712014964,"rms_envelope_nrmse":0.25411932387693764,"envelope_correlation":0.9605856279631473,"rms_gain_error_db":0.039420756288805515,"reference_mean_power":0.0002456629379748078,"output_mean_power":0.00024790296269371813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.8995616035741734e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.069030363478805e-05,"output_power":7.7664367162735e-05,"reference_active":true,"delta_db":0.40861960726109536},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8194611865607095e-05,"output_power":3.124729874222353e-05,"reference_active":true,"delta_db":0.44646359180181916},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.420805316362498e-05,"output_power":1.3557857598095324e-05,"reference_active":true,"delta_db":-0.20343505395497918}],"active_band_count":3,"active_band_mae_db":0.3528394176726312}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":2.195098317835766,"rms_envelope_nrmse":0.20119875193704503,"envelope_correlation":0.974392268049324,"rms_gain_error_db":0.0476343539878714,"reference_mean_power":0.0002452230882951049,"output_mean_power":0.0002479275523937494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.7984313488718053e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.444708123608024e-05,"output_power":7.90902359763592e-05,"reference_active":true,"delta_db":0.26275195797853745},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5639686787851e-05,"output_power":1.7304448262648265e-05,"reference_active":true,"delta_db":0.43929705444852524},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0939336594431209e-06,"output_power":9.944973278473892e-07,"reference_active":true,"delta_db":-0.41387364884241656}],"active_band_count":3,"active_band_mae_db":0.37197422042315975}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13751-0001","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.836426258087158,"predicted_mos":2.836426258087158,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"b336d666b339330e5fc8ad51064ac155ddc5544361b665cabc80f0d9c1c8a1bb","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":9.182999341999675},"started_at":"2026-10-13T07:36:47.551000+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003259332999732578,"native_input_samples":144000,"native_input_sha256":"fe204849450e8ed1a14ee23513308f10e05cb68776d01fbbab4461c6b6bf6fc9","output_sha256":"8950be49d936efb7eeb42f4e8c3830c47480f81ac10fd715b7157bf4ac68f492","output_samples":144120,"length_delta_samples":120,"analysis_s":2.2066302099992754,"synthesis_s":3.756877728999825,"total_s":5.963507938999101,"rtf":0.9939179898331835,"output_peak":0.7139222621917725,"output_rms":0.05612240160706009,"finished_at":"2026-10-13T07:36:53.529543+00:00","analysis_stages":{"external_f0_s":1.4551154670007236},"evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"8950be49d936efb7eeb42f4e8c3830c47480f81ac10fd715b7157bf4ac68f492","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.031875141013423,"output_peak":0.7181580728779896},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.94535047987425,"spectral_convergence":0.6181958211788317,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.768615256007771,"spectral_convergence":0.45360014585477304,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.553041400231298,"spectral_convergence":0.439900456258773,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0011083791662265378,"delta_db":0.5701774541723544,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0007634173572903764,"delta_db":-0.13607825397383413,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.0007271317702297303,"delta_db":-0.5625988792145803,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.00014156264753934796,"delta_db":-0.3482047631974067,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.0001433335380993976,"delta_db":-0.2856993846137321,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00023807315945547448,"delta_db":-1.8528868952441684,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6259409384026794,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.18617629017777343,"pitch":null,"output_rms_dbfs":-25.055813407404052,"output_peak":0.7181580728779896,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.089002378704439},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8213127028567834,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057084492467456845,"rms_dbfs":-24.869637117226276,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.347149433040023,"mean":-0.0066053005854288735,"zero_energy":false},"output":{"rms":0.055873944185468635,"rms_dbfs":-25.055813407404052,"peak":0.7181580728779896,"peak_dbfs":-2.8755990643784375,"crest_factor_db":22.180214343025618,"mean":0.00022356712739521708,"zero_energy":false},"rms_difference_db":-0.1861762901777766},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":1.0503294824043874e-05,"output_rms_dbfs":-49.78674443840418,"output_zero_energy":false,"output_peak":0.012957660098566451,"output_energy_fraction":5.607323941571032e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":159,"both_estimated_voiced_frames":108,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":51,"paired_voiced_fraction":0.9557522123893806,"absolute_difference_cents_median":7.576962311848088,"absolute_difference_cents_p95":23.035521803107784,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002718271425028404,"log_rms_mae_db":7.408425949354709,"rms_envelope_nrmse":0.21377894512570403,"envelope_correlation":0.961608929629588,"rms_gain_error_db":-0.1815836394307499,"reference_mean_power":0.0032374876277477364,"output_mean_power":0.0031049152717897463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2735378968683332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666640237447372,"output_power":0.0005483165068450583,"reference_active":true,"delta_db":0.7002699897228535},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002202584731702016,"output_power":0.00021268067831387727,"reference_active":true,"delta_db":-0.1520458760953759},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.478672520125922e-05,"output_power":3.569049027037311e-05,"reference_active":true,"delta_db":0.11138967295965804}],"active_band_count":3,"active_band_mae_db":0.32123517959262915}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002399603663054526,"log_rms_mae_db":6.681082476177006,"rms_envelope_nrmse":0.1560246423859644,"envelope_correlation":0.9799365968631996,"rms_gain_error_db":-0.15866332645320144,"reference_mean_power":0.0031195469748277875,"output_mean_power":0.0030076354638786155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1403707959359127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046404851798205753,"output_power":0.0005458172966796838,"reference_active":true,"delta_db":0.7048390411212866},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.861370772885771e-05,"output_power":9.812755494032005e-05,"reference_active":true,"delta_db":-0.021463103396388544},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.958540928627598e-06,"output_power":2.0449137689831398e-06,"reference_active":true,"delta_db":0.18742347515862662}],"active_band_count":3,"active_band_mae_db":0.30457520655876724}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022360065406936124,"log_rms_mae_db":2.3132923739618874,"rms_envelope_nrmse":0.190108790978867,"envelope_correlation":0.9770703405938692,"rms_gain_error_db":-0.433353097651053,"reference_mean_power":0.0012597240435666534,"output_mean_power":0.00114009255267592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2605416603296333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001978275618555469,"output_power":0.00018906236162903492,"reference_active":true,"delta_db":-0.19681720043297854},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.107555923758901e-05,"output_power":7.778677196215108e-05,"reference_active":true,"delta_db":-0.17984203728972598},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1829892137914497e-05,"output_power":1.0152689029726208e-05,"reference_active":true,"delta_db":-0.6639970062447881}],"active_band_count":3,"active_band_mae_db":0.3468854146558309}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019906826982741464,"log_rms_mae_db":1.5930514947188974,"rms_envelope_nrmse":0.12622209456964614,"envelope_correlation":0.9898941121476384,"rms_gain_error_db":-0.3683034704049505,"reference_mean_power":0.0011746644654148715,"output_mean_power":0.001079154147374142},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9316122263011187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021214336491155288,"output_power":0.00020287242798014331,"reference_active":true,"delta_db":-0.1940642625446493},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6617395760512454e-05,"output_power":3.513535564397596e-05,"reference_active":true,"delta_db":-0.1794310004573587},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023494047314193e-06,"output_power":8.190092949688776e-07,"reference_active":true,"delta_db":-0.8773030618279745}],"active_band_count":3,"active_band_mae_db":0.41693277494332753}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.946276793614853e-05,"log_rms_mae_db":2.380215896260826,"rms_envelope_nrmse":0.23560642732908219,"envelope_correlation":0.9612094744590994,"rms_gain_error_db":-0.3492291949095567,"reference_mean_power":0.00019501605476176574,"output_mean_power":0.0001799481751250595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.789505238315191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.753576467028988e-05,"output_power":3.499990709150278e-05,"reference_active":true,"delta_db":-0.3037837614191508},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6240794510593083e-05,"output_power":1.5323833737763455e-05,"reference_active":true,"delta_db":-0.2523984012306964},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.016250052718377e-06,"output_power":3.2720126006877854e-06,"reference_active":true,"delta_db":-0.8900577676748446}],"active_band_count":3,"active_band_mae_db":0.4820799767748973}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.236492865510778e-05,"log_rms_mae_db":1.3988382085369244,"rms_envelope_nrmse":0.1251414833247477,"envelope_correlation":0.9884477698464023,"rms_gain_error_db":-0.21760062600577265,"reference_mean_power":0.00017673870709398643,"output_mean_power":0.0001681015086340776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.939463723397027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.781437846450254e-05,"output_power":3.5476034732004474e-05,"reference_active":true,"delta_db":-0.277218951036758},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.773513697250703e-06,"output_power":7.5069863061371486e-06,"reference_active":true,"delta_db":-0.1515174449496599},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2492076040039023e-07,"output_power":2.5191638409702546e-07,"reference_active":true,"delta_db":-1.1052104693637244}],"active_band_count":3,"active_band_mae_db":0.5113156217833807}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014584429481341945,"log_rms_mae_db":2.424565285644568,"rms_envelope_nrmse":0.32932446560579,"envelope_correlation":0.9295302158233809,"rms_gain_error_db":-0.8237645862332372,"reference_mean_power":0.0002465230996620666,"output_mean_power":0.00020393001964760384},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.909673910712007e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.093194518056522e-05,"output_power":6.0431333345932215e-05,"reference_active":true,"delta_db":-0.6957969320558922},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.832881668922569e-05,"output_power":2.2111485217377127e-05,"reference_active":true,"delta_db":-1.076105192708535},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4378297018808916e-05,"output_power":9.979562487843242e-06,"reference_active":true,"delta_db":-1.5859594889990043}],"active_band_count":3,"active_band_mae_db":1.1192872045878104}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":1.5223472722213163,"rms_envelope_nrmse":0.1867458118453324,"envelope_correlation":0.9801221863029304,"rms_gain_error_db":-0.8209246954487359,"reference_mean_power":0.0002460055124763466,"output_mean_power":0.00020363497391454445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8079454178991006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.471033966073092e-05,"output_power":6.228378975249489e-05,"reference_active":true,"delta_db":-0.790056808639852},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5789672835999793e-05,"output_power":1.1919341252091669e-05,"reference_active":true,"delta_db":-1.221208776121354},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1060211769697672e-06,"output_power":7.309707038366115e-07,"reference_active":true,"delta_db":-1.798634710152126}],"active_band_count":3,"active_band_mae_db":1.2699667649711106}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13751-0001","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.038006544113159,"predicted_mos":3.038006544113159,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"b08438e898f50a84b52d79f314195987cd8cce4ffea4a8303095fa3c8da90c18","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.088199363999593},"started_at":"2026-10-13T05:10:38.581967+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0037312420008674962,"native_input_samples":144000,"native_input_sha256":"fe204849450e8ed1a14ee23513308f10e05cb68776d01fbbab4461c6b6bf6fc9","output_sha256":"d8199e90e1a92c5931b1397da1251d68586c5796fc640a70000396c51bfbbeaa","output_samples":144240,"length_delta_samples":240,"analysis_s":0.8396226370005024,"synthesis_s":0.11449188200094795,"total_s":0.9541145190014504,"rtf":0.15901908650024174,"output_peak":0.7467761635780334,"output_rms":0.055322151896064035,"finished_at":"2026-10-13T05:10:39.550306+00:00","analysis_stages":{"external_f0_s":0.16602582299856294},"evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"d8199e90e1a92c5931b1397da1251d68586c5796fc640a70000396c51bfbbeaa","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.166387366026694,"output_peak":0.7382687543210523},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.324907966143366,"spectral_convergence":0.35923695656076227,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.843840610435041,"spectral_convergence":0.294726911268317,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.193058979805277,"spectral_convergence":0.32258750784345286,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.0009289388389431462,"delta_db":-0.19683487389288729,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0007481547537028268,"delta_db":-0.22378415628321754,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.0008056017962775588,"delta_db":-0.11752582498647003,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.00014091749569941975,"delta_db":-0.36804236084382425,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00012951541230590288,"delta_db":-0.7259630731270577,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.0002940849910467723,"delta_db":-0.9352626463445015,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.427902155912993,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.29133434793454466,"pitch":null,"output_rms_dbfs":-25.16097146516082,"output_peak":0.7382687543210523,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.120602518794562},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8845666141691837,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057084492467456845,"rms_dbfs":-24.869637117226276,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.347149433040023,"mean":-0.0066053005854288735,"zero_energy":false},"output":{"rms":0.055201569615144344,"rms_dbfs":-25.16097146516082,"peak":0.7382687543210523,"peak_dbfs":-2.635710235860696,"crest_factor_db":22.525261229300128,"mean":-0.00018739917508063007,"zero_energy":false},"rms_difference_db":-0.29133434793454427},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":6.9123073289883945e-06,"output_rms_dbfs":-51.60376960876877,"output_zero_energy":false,"output_peak":0.008125805880162772,"output_energy_fraction":3.780671428692195e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":97,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.8230088495575221,"absolute_difference_cents_median":1.4832779145138117,"absolute_difference_cents_p95":5.1315971938538185,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002718271425028404,"log_rms_mae_db":1.889099827935596,"rms_envelope_nrmse":0.10713763541296298,"envelope_correlation":0.9901415753052814,"rms_gain_error_db":-0.2832820784643984,"reference_mean_power":0.0032374876277477364,"output_mean_power":0.0030330523732903167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2735378968683332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666640237447372,"output_power":0.00045330335930675433,"reference_active":true,"delta_db":-0.12615383858650717},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002202584731702016,"output_power":0.0001944483338531708,"reference_active":true,"delta_db":-0.5412839816174555},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.478672520125922e-05,"output_power":3.3002383684744356e-05,"reference_active":true,"delta_db":-0.2286823740133952}],"active_band_count":3,"active_band_mae_db":0.29870673140578596}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002399603663054526,"log_rms_mae_db":1.106386826993186,"rms_envelope_nrmse":0.06931992238461664,"envelope_correlation":0.9963105083585309,"rms_gain_error_db":-0.30100745754813146,"reference_mean_power":0.0031195469748277875,"output_mean_power":0.0029106553512404875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1403707959359127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046404851798205753,"output_power":0.00045121678917299095,"reference_active":true,"delta_db":-0.12178139224250409},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.861370772885771e-05,"output_power":8.750003263487764e-05,"reference_active":true,"delta_db":-0.5192907293425627},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.958540928627598e-06,"output_power":1.3535323124587348e-06,"reference_active":true,"delta_db":-1.6046402364348658}],"active_band_count":3,"active_band_mae_db":0.7485707860066442}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022360065406936124,"log_rms_mae_db":1.205311351690138,"rms_envelope_nrmse":0.07222458216592008,"envelope_correlation":0.9967282629192034,"rms_gain_error_db":-0.12785112797027912,"reference_mean_power":0.0012597240435666534,"output_mean_power":0.001223179815119269},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2605416603296333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001978275618555469,"output_power":0.00018949892541993236,"reference_active":true,"delta_db":-0.1868004694083767},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.107555923758901e-05,"output_power":7.640068916096283e-05,"reference_active":true,"delta_db":-0.2579267692124996},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1829892137914497e-05,"output_power":8.831573189541716e-06,"reference_active":true,"delta_db":-1.2694271245669866}],"active_band_count":3,"active_band_mae_db":0.5713847877292876}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019906826982741464,"log_rms_mae_db":0.7541379994486626,"rms_envelope_nrmse":0.044471671347959714,"envelope_correlation":0.9988731705253362,"rms_gain_error_db":-0.16502541061210063,"reference_mean_power":0.0011746644654148715,"output_mean_power":0.0011308663726540722},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9316122263011187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021214336491155288,"output_power":0.00020203371113879752,"reference_active":true,"delta_db":-0.21205611704332414},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6617395760512454e-05,"output_power":3.4405536100463244e-05,"reference_active":true,"delta_db":-0.27059124589813294},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023494047314193e-06,"output_power":7.290980296674892e-07,"reference_active":true,"delta_db":-1.3823321220763172}],"active_band_count":3,"active_band_mae_db":0.6216598283392581}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.946276793614853e-05,"log_rms_mae_db":1.5483999921181413,"rms_envelope_nrmse":0.12083618522619657,"envelope_correlation":0.9910921874937533,"rms_gain_error_db":-0.4244401059628349,"reference_mean_power":0.00019501605476176574,"output_mean_power":0.00017685867038085462},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.789505238315191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.753576467028988e-05,"output_power":3.263275671014687e-05,"reference_active":true,"delta_db":-0.6079150450741795},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6240794510593083e-05,"output_power":1.2689193000809548e-05,"reference_active":true,"delta_db":-1.071732684131545},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.016250052718377e-06,"output_power":2.616757925634822e-06,"reference_active":true,"delta_db":-1.860571960275736}],"active_band_count":3,"active_band_mae_db":1.1800732298271537}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.236492865510778e-05,"log_rms_mae_db":0.9527845130253219,"rms_envelope_nrmse":0.07345445660858058,"envelope_correlation":0.9976490774154296,"rms_gain_error_db":-0.436403846137122,"reference_mean_power":0.00017673870709398643,"output_mean_power":0.00015984214135876152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.939463723397027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.781437846450254e-05,"output_power":3.2357415202992066e-05,"reference_active":true,"delta_db":-0.6768314479647939},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.773513697250703e-06,"output_power":6.203673215969155e-06,"reference_active":true,"delta_db":-0.979684552500405},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2492076040039023e-07,"output_power":1.8309757564098968e-07,"reference_active":true,"delta_db":-2.4909486699538617}],"active_band_count":3,"active_band_mae_db":1.38248822347302}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014584429481341945,"log_rms_mae_db":1.2579037813874232,"rms_envelope_nrmse":0.15919820639406648,"envelope_correlation":0.9850678178386085,"rms_gain_error_db":-0.5041976015315638,"reference_mean_power":0.0002465230996620666,"output_mean_power":0.0002195016855585133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.909673910712007e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.093194518056522e-05,"output_power":6.464539572725903e-05,"reference_active":true,"delta_db":-0.4030427141862852},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.832881668922569e-05,"output_power":2.3508963931125907e-05,"reference_active":true,"delta_db":-0.8099494451574425},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4378297018808916e-05,"output_power":7.783228552224956e-06,"reference_active":true,"delta_db":-2.665476672142365}],"active_band_count":3,"active_band_mae_db":1.2928229438286973}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":0.8138554967672519,"rms_envelope_nrmse":0.0971756609966324,"envelope_correlation":0.9955060014395274,"rms_gain_error_db":-0.5029513948492281,"reference_mean_power":0.0002460055124763466,"output_mean_power":0.00021910369385825393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8079454178991006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.471033966073092e-05,"output_power":6.51995889196043e-05,"reference_active":true,"delta_db":-0.59135853342449},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5789672835999793e-05,"output_power":1.18397262005866e-05,"reference_active":true,"delta_db":-1.2503147223302138},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1060211769697672e-06,"output_power":5.296232268191342e-07,"reference_active":true,"delta_db":-3.1979641948611315}],"active_band_count":3,"active_band_mae_db":1.6798791502052783}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13751-0001","method":"magphase","status":"measured","mos":{"status":"measured","value":3.29980731010437,"predicted_mos":3.29980731010437,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288157,"input_sample_rate":48000,"input_seconds":6.003270833333334,"input_float32_sha256":"658a3d8a26f201cb918ae3f5f22012c90d6c016b2a5b2805ccce65eed89cfd14","internal_samples":96053,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5181661730002816},"started_at":"2026-10-13T05:15:08.067696+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00984329700077069,"native_input_samples":288000,"native_input_sha256":"0b28c940910c8f7762e9caa19208cc4e9bffbe07bfdd07a95756c43c7586ff9f","output_sha256":"5ab813dd96281c8f5081ea88eb1518958edd337d85468d517e6ccd5008bfdedc","output_samples":288157,"length_delta_samples":157,"analysis_s":3.0596704239997052,"synthesis_s":0.949433947000216,"total_s":4.009104370999921,"rtf":0.6681840618333202,"output_peak":0.8233792154895718,"output_rms":0.052778162319089716,"finished_at":"2026-10-13T05:15:12.097981+00:00","evaluation_input":{"reference":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"5ab813dd96281c8f5081ea88eb1518958edd337d85468d517e6ccd5008bfdedc","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96053,"length_delta_samples":53,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.578027948462164,"output_peak":0.8080149197102098},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9994482212944936,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.121551497048372,"spectral_convergence":0.7249244198657615,"spectral_convergence_status":"ok","amplitude_floor":1.963677544918102e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.435360281665271,"spectral_convergence":0.5772212125235846,"spectral_convergence_status":"ok","amplitude_floor":1.0999213683671372e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.486861715153098,"spectral_convergence":0.5139106167953649,"spectral_convergence_status":"ok","amplitude_floor":9.527550859021228e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.000972009721466654,"output_power":0.000757244054871454,"delta_db":-1.0843473613551826,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00078771634185205,"output_power":0.0007301408229982273,"delta_db":-0.3296322436008422,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008277001014055567,"output_power":0.0006719928097411965,"delta_db":-0.9050838223711622,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015338014805248517,"output_power":0.00012941132640333165,"delta_db":-0.7379686410637541,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000153079755450355,"output_power":0.00014664773813991152,"delta_db":-0.18642390841466408,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000364753212040036,"output_power":0.00033056424320328074,"delta_db":-0.42743250421979617,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6118147468375669,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7118395622130766,"pitch":null,"output_rms_dbfs":-25.581476679439355,"output_peak":0.8080149197102098,"reference_rms_dbfs":-24.869637117226276,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.347924497955581},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.662470039578321,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057084492467456845,"rms_dbfs":-24.869637117226276,"peak":0.941619873046875,"peak_dbfs":-0.5224876841862535,"crest_factor_db":24.347149433040023,"mean":-0.0066053005854288735,"zero_energy":false},"output":{"rms":0.052592784632089995,"rms_dbfs":-25.581476679439355,"peak":0.8080149197102098,"peak_dbfs":-1.8516124011420874,"crest_factor_db":23.729864278297267,"mean":0.0005079396498429991,"zero_energy":false},"rms_difference_db":-0.7118395622130791},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.94081883027684e-06,"threshold_dbfs":-52.26153691557123,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":3.119320026598871e-06,"output_mean_power":4.844615258069864e-06,"output_rms_dbfs":-53.14740707350457,"output_zero_energy":false,"output_peak":0.009425377683409767,"output_energy_fraction":2.919145284836192e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":113,"output_estimated_voiced_frames":92,"both_estimated_voiced_frames":79,"reference_voiced_output_unvoiced_frames":34,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.6991150442477876,"absolute_difference_cents_median":9.625332189506665,"absolute_difference_cents_p95":34.218415961645064,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002718271425028404,"log_rms_mae_db":3.710142621277021,"rms_envelope_nrmse":0.35687522674942146,"envelope_correlation":0.8807518988437638,"rms_gain_error_db":-0.7300707895753329,"reference_mean_power":0.0032374876277477364,"output_mean_power":0.002736535197584014},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2735378968683332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004666640237447372,"output_power":0.000418003461438978,"reference_active":true,"delta_db":-0.4782444323346057},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002202584731702016,"output_power":0.0002527172850596549,"reference_active":true,"delta_db":0.597023229264868},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.478672520125922e-05,"output_power":3.321527819634711e-05,"reference_active":true,"delta_db":-0.20075652234033953}],"active_band_count":3,"active_band_mae_db":0.4253413946466044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002399603663054526,"log_rms_mae_db":2.58413336935314,"rms_envelope_nrmse":0.24596237862561454,"envelope_correlation":0.9410425276491324,"rms_gain_error_db":-0.7934690982724222,"reference_mean_power":0.0031195469748277875,"output_mean_power":0.002598631029270701},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1403707959359127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046404851798205753,"output_power":0.000418693827224023,"reference_active":true,"delta_db":-0.4466683185375112},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.861370772885771e-05,"output_power":0.00011414060384698737,"reference_active":true,"delta_db":0.6350287786580631},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.958540928627598e-06,"output_power":2.995659859136327e-06,"reference_active":true,"delta_db":1.8455984840159414}],"active_band_count":3,"active_band_mae_db":0.975765193737172}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022360065406936124,"log_rms_mae_db":3.758620068470807,"rms_envelope_nrmse":0.4066588695933945,"envelope_correlation":0.888532392851949,"rms_gain_error_db":-0.7412326889629378,"reference_mean_power":0.0012597240435666534,"output_mean_power":0.0010620675741404083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2605416603296333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001978275618555469,"output_power":0.00018510831241532163,"reference_active":true,"delta_db":-0.28860877043100686},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.107555923758901e-05,"output_power":7.609486066609976e-05,"reference_active":true,"delta_db":-0.27534626846287164},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1829892137914497e-05,"output_power":1.2282272732410173e-05,"reference_active":true,"delta_db":0.16297951968574947}],"active_band_count":3,"active_band_mae_db":0.2423115195265427}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019906826982741464,"log_rms_mae_db":2.729921978506434,"rms_envelope_nrmse":0.30800422922357534,"envelope_correlation":0.9349029894457856,"rms_gain_error_db":-0.8434787866114555,"reference_mean_power":0.0011746644654148715,"output_mean_power":0.0009673106125284945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9316122263011187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021214336491155288,"output_power":0.00019868050591693315,"reference_active":true,"delta_db":-0.28474195926061374},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6617395760512454e-05,"output_power":3.4668698984374e-05,"reference_active":true,"delta_db":-0.23749909767159383},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0023494047314193e-06,"output_power":1.0112846052561393e-06,"reference_active":true,"delta_db":0.03854259247523947}],"active_band_count":3,"active_band_mae_db":0.18692788313581568}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.946276793614853e-05,"log_rms_mae_db":3.6611724128216845,"rms_envelope_nrmse":0.42600387079473084,"envelope_correlation":0.8689908406286491,"rms_gain_error_db":-0.6145795667739804,"reference_mean_power":0.00019501605476176574,"output_mean_power":0.0001692826346478837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.789505238315191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.753576467028988e-05,"output_power":3.620835872427422e-05,"reference_active":true,"delta_db":-0.1563642835945887},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6240794510593083e-05,"output_power":1.703432463170848e-05,"reference_active":true,"delta_db":0.20717648154085846},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.016250052718377e-06,"output_power":6.127451863137551e-06,"reference_active":true,"delta_db":1.8345916388813386}],"active_band_count":3,"active_band_mae_db":0.7327108013389285}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.236492865510778e-05,"log_rms_mae_db":2.3591087761227407,"rms_envelope_nrmse":0.2968046723308279,"envelope_correlation":0.9333235138634993,"rms_gain_error_db":-0.7887689563894017,"reference_mean_power":0.00017673870709398643,"output_mean_power":0.00014738550641788773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.939463723397027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.781437846450254e-05,"output_power":3.76167751957255e-05,"reference_active":true,"delta_db":-0.0227540482633907},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.773513697250703e-06,"output_power":8.752729238284176e-06,"reference_active":true,"delta_db":0.5152612582797689},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2492076040039023e-07,"output_power":4.979642863968522e-07,"reference_active":true,"delta_db":1.8542073569397743}],"active_band_count":3,"active_band_mae_db":0.7974075544943112}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014584429481341945,"log_rms_mae_db":3.719079481504163,"rms_envelope_nrmse":0.5836977664871998,"envelope_correlation":0.7877347197768059,"rms_gain_error_db":0.12819306935293492,"reference_mean_power":0.0002465230996620666,"output_mean_power":0.0002539083168922024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.909673910712007e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.093194518056522e-05,"output_power":7.304162996181585e-05,"reference_active":true,"delta_db":0.12728586386648444},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.832881668922569e-05,"output_power":2.8370801920652276e-05,"reference_active":true,"delta_db":0.006431774360997573},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4378297018808916e-05,"output_power":1.2839701868424959e-05,"reference_active":true,"delta_db":-0.4915251103520214}],"active_band_count":3,"active_band_mae_db":0.20841424952650112}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.656067210642838e-05,"log_rms_mae_db":2.4564942504914384,"rms_envelope_nrmse":0.33610433467442064,"envelope_correlation":0.9253454015448949,"rms_gain_error_db":0.12977498286568362,"reference_mean_power":0.0002460055124763466,"output_mean_power":0.0002534675325798746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8079454178991006e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.471033966073092e-05,"output_power":7.619592815824564e-05,"reference_active":true,"delta_db":0.08551052777486684},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5789672835999793e-05,"output_power":1.5574048096997177e-05,"reference_active":true,"delta_db":-0.059716198809293006},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1060211769697672e-06,"output_power":1.0614147803568568e-06,"reference_active":true,"delta_db":-0.17878311527798826}],"active_band_count":3,"active_band_mae_db":0.10800328062071603}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"hnm","status":"measured","mos":{"status":"measured","value":2.259418487548828,"predicted_mos":2.259418487548828,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"989a394a8852f11fccc1a5921eaaf72e0b2fe039a3f1671512a17779c46e63c3","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6100361839999096},"started_at":"2026-10-13T04:39:09.621546+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4104000001680106e-05,"native_input_samples":96000,"native_input_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","output_sha256":"ba7bc9b1f67a4d4281037d057a84e3bc5111d9659a34bfe5404e807ac6782c37","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0597841879998668,"synthesis_s":0.529005911999775,"total_s":1.5887900999996418,"rtf":0.2647983499999403,"output_peak":0.49471824362549516,"output_rms":0.04614650490156708,"finished_at":"2026-10-13T04:39:11.216175+00:00","evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"ba7bc9b1f67a4d4281037d057a84e3bc5111d9659a34bfe5404e807ac6782c37","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.428802656343286,"spectral_convergence":0.45831952191120223,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.808833978669372,"spectral_convergence":0.4258408311916586,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.924086494840156,"spectral_convergence":0.45472203977835834,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.0002037028118104711,"delta_db":0.3070477399194983,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006975476529620124,"delta_db":-0.10489836100279785,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.0004994016893355776,"delta_db":-0.9255064989416196,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00019354103847406005,"delta_db":-0.4180756981739959,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.0003690786925802652,"delta_db":-0.02075013596413717,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00016622802946796817,"delta_db":-0.23029110322597687,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3344282562046709,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.29969898992070615,"pitch":null,"output_rms_dbfs":-26.717223730006886,"output_peak":0.49471824362549516,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.053907709950938},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.428802656343286,"spectral_convergence":0.45831952191120223,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.808833978669372,"spectral_convergence":0.4258408311916586,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.924086494840156,"spectral_convergence":0.45472203977835834,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.0002037028118104711,"delta_db":0.3070477399194983,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006975476529620124,"delta_db":-0.10489836100279785,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.0004994016893355776,"delta_db":-0.9255064989416196,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00019354103847406005,"delta_db":-0.4180756981739959,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.0003690786925802652,"delta_db":-0.02075013596413717,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00016622802946796817,"delta_db":-0.23029110322597687,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3344282562046709,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.29969898992070615,"pitch":null,"output_rms_dbfs":-26.717223730006886,"output_peak":0.49471824362549516,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.053907709950938},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.804417025238634,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04776653768728406,"rms_dbfs":-26.41752474008618,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.40799300030818,"mean":-3.276507059733073e-06,"zero_energy":false},"output":{"rms":0.04614650490156708,"rms_dbfs":-26.717223730006886,"peak":0.49471824362549516,"peak_dbfs":-6.112841478999325,"crest_factor_db":20.604382251007564,"mean":5.049437864879804e-05,"zero_energy":false},"rms_difference_db":-0.29969898992070654},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":1.0493836276461093e-05,"output_rms_dbfs":-49.79065715893748,"output_zero_energy":false,"output_peak":0.07303401711803607,"output_energy_fraction":0.0005174232698758656,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":71,"output_estimated_voiced_frames":76,"both_estimated_voiced_frames":63,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.8873239436619719,"absolute_difference_cents_median":2.188160972191794,"absolute_difference_cents_p95":13.455291383816437,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002801281523487693,"log_rms_mae_db":1.197601279503507,"rms_envelope_nrmse":0.12712499649552378,"envelope_correlation":0.9857328224282565,"rms_gain_error_db":-0.2996964833180404,"reference_mean_power":0.0022835195719053173,"output_mean_power":0.0021312534037734024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.817352658591746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000450774313706133,"output_power":0.00041865949361674995,"reference_active":true,"delta_db":-0.32098217033741916},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021888119814572043,"output_power":0.00021105021223966788,"reference_active":true,"delta_db":-0.1582266412200008},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.753791744185585e-05,"output_power":2.319816334167713e-05,"reference_active":true,"delta_db":-2.090165727827863}],"active_band_count":3,"active_band_mae_db":0.8564581797950943}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022259441477048914,"log_rms_mae_db":0.6543159854009364,"rms_envelope_nrmse":0.0804792423676128,"envelope_correlation":0.9944687547187091,"rms_gain_error_db":-0.29976306402531605,"reference_mean_power":0.002294705859570105,"output_mean_power":0.002141660950898454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.266680044598684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042660839170957063,"output_power":0.0003928907788041409,"reference_active":true,"delta_db":-0.35757558112754007},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012272337901795565,"output_power":0.0001182689541874878,"reference_active":true,"delta_db":-0.16056547957282963},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2413713648325596e-06,"output_power":1.959376296501399e-06,"reference_active":true,"delta_db":-2.1861094124460756}],"active_band_count":3,"active_band_mae_db":0.9014168243821484}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002244694917839089,"log_rms_mae_db":1.8888667316114236,"rms_envelope_nrmse":0.1862189141323043,"envelope_correlation":0.9774653527147029,"rms_gain_error_db":-0.6206813589919484,"reference_mean_power":0.001082154267593255,"output_mean_power":0.0009380393143733165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.600154760198564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003310843525581766,"output_power":0.00029071679625644354,"reference_active":true,"delta_db":-0.5646853196258712},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.047815482126218e-05,"output_power":7.8905759048254e-05,"reference_active":true,"delta_db":-0.5943503320096165},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6938491072881828e-05,"output_power":1.0738341788377541e-05,"reference_active":true,"delta_db":-1.9793749666427214}],"active_band_count":3,"active_band_mae_db":1.046136872759403}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018235367957614062,"log_rms_mae_db":1.1224716257113305,"rms_envelope_nrmse":0.12474510949679453,"envelope_correlation":0.9914887355696211,"rms_gain_error_db":-0.6206906578242198,"reference_mean_power":0.0010875823794426604,"output_mean_power":0.0009427425236076793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.047574386178107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032394555601725244,"output_power":0.0002838882725544802,"reference_active":true,"delta_db":-0.5732457438037768},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.774463594481584e-05,"output_power":4.169931452342983e-05,"reference_active":true,"delta_db":-0.5879567031492436},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5157059923891315e-06,"output_power":8.141654177031893e-07,"reference_active":true,"delta_db":-2.6990231616533196}],"active_band_count":3,"active_band_mae_db":1.2867418695354467}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014040871531833865,"log_rms_mae_db":1.8789556618530343,"rms_envelope_nrmse":0.17417335567813536,"envelope_correlation":0.9756048760718277,"rms_gain_error_db":-0.1740543927765577,"reference_mean_power":0.00036159326722347865,"output_mean_power":0.00034738806705518943},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2796709554971262e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010222442860363177,"output_power":0.00010002469377681778,"reference_active":true,"delta_db":-0.09447461218230951},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.551583786781792e-05,"output_power":4.464425906399878e-05,"reference_active":true,"delta_db":-0.08396922134838168},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1818852015306106e-05,"output_power":9.53847537381164e-06,"reference_active":true,"delta_db":-0.9309633206012041}],"active_band_count":3,"active_band_mae_db":0.36980238471063176}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010845211232047094,"log_rms_mae_db":1.1129482162172932,"rms_envelope_nrmse":0.10237019038166269,"envelope_correlation":0.9908093487783468,"rms_gain_error_db":-0.1740525008606281,"reference_mean_power":0.00036337479954251354,"output_mean_power":0.0003490997639125843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.941433739824423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.608975832570786e-05,"output_power":9.533551605324454e-05,"reference_active":true,"delta_db":-0.03422379339765577},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4640366344588205e-05,"output_power":2.3702423238064564e-05,"reference_active":true,"delta_db":-0.16854411725223728},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.52028436774113e-07,"output_power":7.071247241273286e-07,"reference_active":true,"delta_db":-1.2915389849465986}],"active_band_count":3,"active_band_mae_db":0.49810229853216387}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012084003773539908,"log_rms_mae_db":1.9847987247358132,"rms_envelope_nrmse":0.16618748827943208,"envelope_correlation":0.9798353815007047,"rms_gain_error_db":-0.19566020018397862,"reference_mean_power":0.00030833628390941907,"output_mean_power":0.00029475325782886343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.296444677120579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.747295044804844e-05,"output_power":4.968897561937284e-05,"reference_active":true,"delta_db":0.19813819489108858},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.9782403314319074e-05,"output_power":4.678120596050329e-05,"reference_active":true,"delta_db":-0.2700444576340061},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3740568823166804e-05,"output_power":8.850259941527777e-06,"reference_active":true,"delta_db":-1.9104868516226703}],"active_band_count":3,"active_band_mae_db":0.7928898347159216}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010527157253957593,"log_rms_mae_db":1.1329385178482865,"rms_envelope_nrmse":0.09118101972560612,"envelope_correlation":0.993483762127266,"rms_gain_error_db":-0.19562632287792672,"reference_mean_power":0.0003098592063705906,"output_mean_power":0.0002962114021404809},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.056454468513195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.721882315895358e-05,"output_power":5.8039012384531935e-05,"reference_active":true,"delta_db":0.061810924336026335},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.889894420146288e-05,"output_power":2.5347317297451988e-05,"reference_active":true,"delta_db":-0.5694997510494001},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2532553850736717e-06,"output_power":7.778953628640365e-07,"reference_active":true,"delta_db":-2.071183967848671}],"active_band_count":3,"active_band_mae_db":0.9008315477446991}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.831660270690918,"predicted_mos":2.831660270690918,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"330063512f0afc71508fddc4b20f5ab9c111cc044ed3f03f52742c147007c985","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5323687190011697},"started_at":"2026-10-13T04:42:06.287183+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3116000193112995e-05,"native_input_samples":96000,"native_input_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","output_sha256":"7ba415142d5b9a1c1a221a772de01ce98f2855118909bdca0ee892d9d73ee4ae","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5179570609998336,"synthesis_s":0.0782272770002237,"total_s":1.5961843380000573,"rtf":0.2660307230000096,"output_peak":0.8693246281400739,"output_rms":0.05238804444350793,"finished_at":"2026-10-13T04:42:07.890093+00:00","evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"7ba415142d5b9a1c1a221a772de01ce98f2855118909bdca0ee892d9d73ee4ae","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.092677525947421,"spectral_convergence":0.6510877887828266,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.600227657483598,"spectral_convergence":0.39648491208674974,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.702887424715572,"spectral_convergence":0.4109125542724666,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.00018162011862926116,"delta_db":-0.1912829482120828,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0008540521355473005,"delta_db":0.7742066508843638,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.000837532029947618,"delta_db":1.3200076785117947,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00026824749822156567,"delta_db":0.9995504344975791,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.00040294329352527614,"delta_db":0.36049944794277955,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00020011212474394094,"delta_db":0.5754003629207398,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7034912538282233,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8021684856266718,"pitch":null,"output_rms_dbfs":-25.615356254459506,"output_peak":0.8693246281400739,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.131930869382198},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.092677525947421,"spectral_convergence":0.6510877887828266,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.600227657483598,"spectral_convergence":0.39648491208674974,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.702887424715572,"spectral_convergence":0.4109125542724666,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.00018162011862926116,"delta_db":-0.1912829482120828,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0008540521355473005,"delta_db":0.7742066508843638,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.000837532029947618,"delta_db":1.3200076785117947,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00026824749822156567,"delta_db":0.9995504344975791,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.00040294329352527614,"delta_db":0.36049944794277955,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00020011212474394094,"delta_db":0.5754003629207398,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7034912538282233,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8021684856266718,"pitch":null,"output_rms_dbfs":-25.615356254459506,"output_peak":0.8693246281400739,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.131930869382198},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8680910020377283,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04776653768728406,"rms_dbfs":-26.41752474008618,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.40799300030818,"mean":-3.276507059733073e-06,"zero_energy":false},"output":{"rms":0.05238804444350793,"rms_dbfs":-25.615356254459506,"peak":0.8693246281400739,"peak_dbfs":-1.2163603310251705,"crest_factor_db":24.398995923434338,"mean":1.338482813274928e-06,"zero_energy":false},"rms_difference_db":0.8021684856266731},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":3.368675590711697e-06,"output_rms_dbfs":-54.725408103152596,"output_zero_energy":false,"output_peak":0.02555422907341862,"output_energy_fraction":0.00012887958062032852,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":71,"output_estimated_voiced_frames":97,"both_estimated_voiced_frames":69,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.971830985915493,"absolute_difference_cents_median":3.771215987963046,"absolute_difference_cents_p95":16.525308459042794,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002801281523487693,"log_rms_mae_db":2.1056241368160795,"rms_envelope_nrmse":0.2010625231774701,"envelope_correlation":0.9742144381353924,"rms_gain_error_db":0.8021876277022837,"reference_mean_power":0.0022835195719053173,"output_mean_power":0.0027467776251711555},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.817352658591746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000450774313706133,"output_power":0.0005630263553216407,"reference_active":true,"delta_db":0.9656956384759465},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021888119814572043,"output_power":0.00023490628838326934,"reference_active":true,"delta_db":0.3068618553437597},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.753791744185585e-05,"output_power":3.400098788177148e-05,"reference_active":true,"delta_db":-0.4297863938692143}],"active_band_count":3,"active_band_mae_db":0.5674479625629735}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022259441477048914,"log_rms_mae_db":1.3658227649019188,"rms_envelope_nrmse":0.1386232306873298,"envelope_correlation":0.9912326498511738,"rms_gain_error_db":0.8021622612578477,"reference_mean_power":0.002294705859570105,"output_mean_power":0.0027602171551257535},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.266680044598684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042660839170957063,"output_power":0.000532472033929296,"reference_active":true,"delta_db":0.962674089437775},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012272337901795565,"output_power":0.0001327438507905096,"reference_active":true,"delta_db":0.3408710755089517},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2413713648325596e-06,"output_power":3.0685790139162776e-06,"reference_active":true,"delta_db":-0.23791480559483064}],"active_band_count":3,"active_band_mae_db":0.513819990180519}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002244694917839089,"log_rms_mae_db":2.1462583743433115,"rms_envelope_nrmse":0.22376518018080957,"envelope_correlation":0.9815343285389057,"rms_gain_error_db":1.1530674195522492,"reference_mean_power":0.001082154267593255,"output_mean_power":0.0014112238854289976},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.600154760198564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003310843525581766,"output_power":0.0004295126456188428,"reference_active":true,"delta_db":1.130372987988229},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.047815482126218e-05,"output_power":0.00010876443498104035,"reference_active":true,"delta_db":0.7994317291480195},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6938491072881828e-05,"output_power":1.4078229603057964e-05,"reference_active":true,"delta_db":-0.8032667570633263}],"active_band_count":3,"active_band_mae_db":0.9110238247331915}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018235367957614062,"log_rms_mae_db":1.5302355790559556,"rms_envelope_nrmse":0.1755092567733785,"envelope_correlation":0.9929780873216226,"rms_gain_error_db":1.1530452122395085,"reference_mean_power":0.0010875823794426604,"output_mean_power":0.0014182953660385655},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.047574386178107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032394555601725244,"output_power":0.00041668404069808234,"reference_active":true,"delta_db":1.0933484044153712},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.774463594481584e-05,"output_power":5.8367964922004615e-05,"reference_active":true,"delta_db":0.8724996513732166},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5157059923891315e-06,"output_power":1.1458041841873038e-06,"reference_active":true,"delta_db":-1.2150456376108645}],"active_band_count":3,"active_band_mae_db":1.0602978977998174}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014040871531833865,"log_rms_mae_db":2.017891388291663,"rms_envelope_nrmse":0.21831447715468513,"envelope_correlation":0.9697445307375765,"rms_gain_error_db":0.6755748313644743,"reference_mean_power":0.00036159326722347865,"output_mean_power":0.000422452435824957},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2796709554971262e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010222442860363177,"output_power":0.00011893820886934819,"reference_active":true,"delta_db":0.6576670233600064},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.551583786781792e-05,"output_power":4.831833609207012e-05,"reference_active":true,"delta_db":0.2594942859589724},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1818852015306106e-05,"output_power":9.626584390281355e-06,"reference_active":true,"delta_db":-0.8910307248011942}],"active_band_count":3,"active_band_mae_db":0.6027306780400576}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010845211232047094,"log_rms_mae_db":1.2066082972319003,"rms_envelope_nrmse":0.13172872955116688,"envelope_correlation":0.9912281730184045,"rms_gain_error_db":0.6755676375557791,"reference_mean_power":0.00036337479954251354,"output_mean_power":0.0004245331117086529},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.941433739824423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.608975832570786e-05,"output_power":0.00011241191424344578,"reference_active":true,"delta_db":0.6813524230800243},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4640366344588205e-05,"output_power":2.505924382415222e-05,"reference_active":true,"delta_db":0.07320801308627016},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.52028436774113e-07,"output_power":7.477436262730766e-07,"reference_active":true,"delta_db":-1.0489720099450626}],"active_band_count":3,"active_band_mae_db":0.601177482037119}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012084003773539908,"log_rms_mae_db":2.399792081676581,"rms_envelope_nrmse":0.27354963316353353,"envelope_correlation":0.9522409024328641,"rms_gain_error_db":0.5581933255929139,"reference_mean_power":0.00030833628390941907,"output_mean_power":0.00035062587832002915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.296444677120579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.747295044804844e-05,"output_power":4.7197506081550665e-05,"reference_active":true,"delta_db":-0.025271728979955464},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.9782403314319074e-05,"output_power":4.642327652796346e-05,"reference_active":true,"delta_db":-0.3034006949038648},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3740568823166804e-05,"output_power":1.297920801263812e-05,"reference_active":true,"delta_db":-0.24756518982870146}],"active_band_count":3,"active_band_mae_db":0.19207920457084057}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010527157253957593,"log_rms_mae_db":1.4670422795063214,"rms_envelope_nrmse":0.17652156409319036,"envelope_correlation":0.9799807300699822,"rms_gain_error_db":0.5582366257406165,"reference_mean_power":0.0003098592063705906,"output_mean_power":0.0003523611889883496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.056454468513195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.721882315895358e-05,"output_power":5.597723917622496e-05,"reference_active":true,"delta_db":-0.09527446213080597},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.889894420146288e-05,"output_power":2.7175859728231294e-05,"reference_active":true,"delta_db":-0.2669868426424641},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2532553850736717e-06,"output_power":1.1967952979730953e-06,"reference_active":true,"delta_db":-0.20019705148711112}],"active_band_count":3,"active_band_mae_db":0.1874861187534604}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"facodec","status":"measured","mos":{"status":"measured","value":3.823007822036743,"predicted_mos":3.823007822036743,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"3376ad010b1b71caf69d42149bcc7a812d08e258e3c603748c072aba27be0af1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.764143279000564},"started_at":"2026-10-13T04:51:55.323068+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.22549997488386e-05,"native_input_samples":96000,"native_input_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","output_sha256":"979f8e74483b81a92657b5baa81ef93640e2216d5f102420ce1b953aa212909e","output_samples":96000,"length_delta_samples":0,"analysis_s":1.8130995729998176,"synthesis_s":5.763305946000401,"total_s":7.576405519000218,"rtf":1.262734253166703,"output_peak":0.5866541862487793,"output_rms":0.04510013627563282,"finished_at":"2026-10-13T04:52:02.906501+00:00","evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"979f8e74483b81a92657b5baa81ef93640e2216d5f102420ce1b953aa212909e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.600830016063353,"spectral_convergence":0.37819945711468117,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.812890004373592,"spectral_convergence":0.40780372734070747,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.916103500684311,"spectral_convergence":0.43959667141934333,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.00018188340182064892,"delta_db":-0.1849918147974308,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006838437690750718,"delta_db":-0.1910682373407957,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.000537453816214924,"delta_db":-0.6065950604798185,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00014442891032353874,"delta_db":-1.6892650218373386,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.0003242485119485112,"delta_db":-0.5831599541620089,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00016216388269795637,"delta_db":-0.33779232792199626,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5988120694232315,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4989181768264779,"pitch":null,"output_rms_dbfs":-26.916442916912658,"output_peak":0.5866541862487793,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.776607840373752},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.600830016063353,"spectral_convergence":0.37819945711468117,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.812890004373592,"spectral_convergence":0.40780372734070747,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.916103500684311,"spectral_convergence":0.43959667141934333,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.00018188340182064892,"delta_db":-0.1849918147974308,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006838437690750718,"delta_db":-0.1910682373407957,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.000537453816214924,"delta_db":-0.6065950604798185,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00014442891032353874,"delta_db":-1.6892650218373386,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.0003242485119485112,"delta_db":-0.5831599541620089,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00016216388269795637,"delta_db":-0.33779232792199626,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5988120694232315,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4989181768264779,"pitch":null,"output_rms_dbfs":-26.916442916912658,"output_peak":0.5866541862487793,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.776607840373752},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8485376945062669,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04776653768728406,"rms_dbfs":-26.41752474008618,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.40799300030818,"mean":-3.276507059733073e-06,"zero_energy":false},"output":{"rms":0.04510013627563282,"rms_dbfs":-26.916442916912658,"peak":0.5866541862487793,"peak_dbfs":-4.632356518824141,"crest_factor_db":22.284086398088515,"mean":8.809924165191774e-05,"zero_energy":false},"rms_difference_db":-0.49891817682647854},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":2.82595323134315e-06,"output_rms_dbfs":-55.488350298680935,"output_zero_energy":false,"output_peak":0.016948403790593147,"output_energy_fraction":0.00014588094262600406,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":71,"output_estimated_voiced_frames":82,"both_estimated_voiced_frames":67,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9436619718309859,"absolute_difference_cents_median":5.539975186690886,"absolute_difference_cents_p95":18.981403585151533,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002801281523487693,"log_rms_mae_db":1.3055208178315525,"rms_envelope_nrmse":0.1487545301244264,"envelope_correlation":0.9814192684913237,"rms_gain_error_db":-0.4990638252942608,"reference_mean_power":0.0022835195719053173,"output_mean_power":0.002035627717789289},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.817352658591746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000450774313706133,"output_power":0.00042190039156165507,"reference_active":true,"delta_db":-0.28749232402415004},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021888119814572043,"output_power":0.00022817603042621414,"reference_active":true,"delta_db":0.18061563093864907},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.753791744185585e-05,"output_power":3.915761575509593e-05,"reference_active":true,"delta_db":0.18346065613899715}],"active_band_count":3,"active_band_mae_db":0.21718953703393207}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022259441477048914,"log_rms_mae_db":0.9728962474097764,"rms_envelope_nrmse":0.1338698522594447,"envelope_correlation":0.9838714544943248,"rms_gain_error_db":-0.4996029502351024,"reference_mean_power":0.002294705859570105,"output_mean_power":0.0020453457354016715},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.266680044598684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042660839170957063,"output_power":0.00039922964485575475,"reference_active":true,"delta_db":-0.28806611685138156},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012272337901795565,"output_power":0.00012622566499172138,"reference_active":true,"delta_db":0.12220362891542504},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2413713648325596e-06,"output_power":3.369055485961379e-06,"reference_active":true,"delta_db":0.16779372438813875}],"active_band_count":3,"active_band_mae_db":0.19268782338498178}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002244694917839089,"log_rms_mae_db":1.7860055037976488,"rms_envelope_nrmse":0.15163420856547788,"envelope_correlation":0.9843966644464908,"rms_gain_error_db":-0.34477531354078184,"reference_mean_power":0.001082154267593255,"output_mean_power":0.0009995663908916057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.600154760198564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003310843525581766,"output_power":0.000327666579147491,"reference_active":true,"delta_db":-0.045065088481215154},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.047815482126218e-05,"output_power":9.816742738167628e-05,"reference_active":true,"delta_db":0.35423674670265437},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6938491072881828e-05,"output_power":2.0972872856994546e-05,"reference_active":true,"delta_db":0.9278320447324662}],"active_band_count":3,"active_band_mae_db":0.4423779599721119}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018235367957614062,"log_rms_mae_db":1.292442430576678,"rms_envelope_nrmse":0.1384187116832299,"envelope_correlation":0.9863418964997163,"rms_gain_error_db":-0.34480358173402453,"reference_mean_power":0.0010875823794426604,"output_mean_power":0.0010045737011804116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.047574386178107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032394555601725244,"output_power":0.0003230396522006753,"reference_active":true,"delta_db":-0.012161925212154317},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.774463594481584e-05,"output_power":5.1094422809204134e-05,"reference_active":true,"delta_db":0.29448911312823856},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5157059923891315e-06,"output_power":1.7952331137906882e-06,"reference_active":true,"delta_db":0.7350588277359047}],"active_band_count":3,"active_band_mae_db":0.34723662202543254}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014040871531833865,"log_rms_mae_db":1.5917125796490614,"rms_envelope_nrmse":0.22507846135822057,"envelope_correlation":0.9649927195229105,"rms_gain_error_db":-0.9922265043117494,"reference_mean_power":0.00036159326722347865,"output_mean_power":0.00028773830767986607},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2796709554971262e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010222442860363177,"output_power":8.53102022125037e-05,"reference_active":true,"delta_db":-0.785537203148456},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.551583786781792e-05,"output_power":5.129700362014843e-05,"reference_active":true,"delta_db":0.5192945592948568},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1818852015306106e-05,"output_power":1.205509501436827e-05,"reference_active":true,"delta_db":0.08595342845273332}],"active_band_count":3,"active_band_mae_db":0.46359506363201536}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010845211232047094,"log_rms_mae_db":1.23523703844787,"rms_envelope_nrmse":0.18874759152021497,"envelope_correlation":0.9759151929843963,"rms_gain_error_db":-0.9927842854265024,"reference_mean_power":0.00036337479954251354,"output_mean_power":0.0002891188292867372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.941433739824423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.608975832570786e-05,"output_power":8.344098154851403e-05,"reference_active":true,"delta_db":-0.6129769687323412},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4640366344588205e-05,"output_power":2.772114468304135e-05,"reference_active":true,"delta_db":0.5116399905209353},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.52028436774113e-07,"output_power":9.60133392999597e-07,"reference_active":true,"delta_db":0.03681653707317269}],"active_band_count":3,"active_band_mae_db":0.38714449877548307}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012084003773539908,"log_rms_mae_db":1.9082477458304834,"rms_envelope_nrmse":0.21824048968866527,"envelope_correlation":0.9659542027836556,"rms_gain_error_db":-0.5570810162240671,"reference_mean_power":0.00030833628390941907,"output_mean_power":0.00027121676550921696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.296444677120579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.747295044804844e-05,"output_power":4.4339886043095594e-05,"reference_active":true,"delta_db":-0.2965165138761751},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.9782403314319074e-05,"output_power":4.490344017962023e-05,"reference_active":true,"delta_db":-0.44796244137815305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3740568823166804e-05,"output_power":1.2935524303974022e-05,"reference_active":true,"delta_db":-0.26220675460298265}],"active_band_count":3,"active_band_mae_db":0.33556190328577024}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010527157253957593,"log_rms_mae_db":1.5368737873346972,"rms_envelope_nrmse":0.17749202482039556,"envelope_correlation":0.97609906042624,"rms_gain_error_db":-0.5577568270228608,"reference_mean_power":0.0003098592063705906,"output_mean_power":0.0002725139391860752},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.056454468513195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.721882315895358e-05,"output_power":5.146412791038059e-05,"reference_active":true,"delta_db":-0.4603430346193848},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.889894420146288e-05,"output_power":2.599598904221811e-05,"reference_active":true,"delta_db":-0.45975631231949304},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2532553850736717e-06,"output_power":1.1720818855345987e-06,"reference_active":true,"delta_db":-0.29081625398030064}],"active_band_count":3,"active_band_mae_db":0.4036385336397261}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.1459031105041504,"predicted_mos":3.1459031105041504,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"60295c895c3460f9976791943af660d6378914f51d98f8f560141e5ea2b24594","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.267781119000574},"started_at":"2026-10-13T07:13:34.884171+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006607953999264282,"native_input_samples":264600,"native_input_sha256":"918d5c1b03f4a9387fff0b23a54033627c9162eba6f03f423aa77b362eb0d345","output_sha256":"aec8395e6df1273037280439fbd140d09c035d194b733f6a4b4d3084d9641240","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15718327600006887,"synthesis_s":4.052060217998587,"total_s":4.209243493998656,"rtf":0.7015405823331093,"output_peak":0.6483179926872253,"output_rms":0.04260756637056645,"finished_at":"2026-10-13T07:13:39.126127+00:00","evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"aec8395e6df1273037280439fbd140d09c035d194b733f6a4b4d3084d9641240","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.404927895670006,"output_peak":0.6436787899702683},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.337047713530765,"spectral_convergence":0.5926274036261924,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.902725536165597,"spectral_convergence":0.5170026929009875,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.658673998324462,"spectral_convergence":0.580890243162167,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":8.828924781577844e-05,"delta_db":-3.323844329870723,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006669775787797329,"delta_db":-0.2995248303100812,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.0005461103095704605,"delta_db":-0.537202817044505,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00015293988621055863,"delta_db":-1.44059874079425,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.0002739945821728513,"delta_db":-1.3145201170813419,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":9.00297899112264e-05,"delta_db":-2.893471302371657,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.634860356245426,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9857210045469479,"pitch":null,"output_rms_dbfs":-27.403245744633125,"output_peak":0.6436787899702683,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.966149082673608},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6790942394507725,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04776653768728406,"rms_dbfs":-26.41752474008618,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.40799300030818,"mean":-3.276507059733073e-06,"zero_energy":false},"output":{"rms":0.042642014427798886,"rms_dbfs":-27.403245744633125,"peak":0.6436787899702683,"peak_dbfs":-3.8266160237476448,"crest_factor_db":23.57662972088548,"mean":-0.00020206097543233503,"zero_energy":false},"rms_difference_db":-0.9857210045469458},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":3.6829310144303224e-06,"output_rms_dbfs":-54.33806415908278,"output_zero_energy":false,"output_peak":0.06906230912106699,"output_energy_fraction":0.00021267060063266975,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":71,"output_estimated_voiced_frames":57,"both_estimated_voiced_frames":51,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.7183098591549296,"absolute_difference_cents_median":16.371108050835858,"absolute_difference_cents_p95":51.4621013548674,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002801281523487693,"log_rms_mae_db":2.2251953674214535,"rms_envelope_nrmse":0.19123347208859615,"envelope_correlation":0.9751961286619608,"rms_gain_error_db":-0.985685686390174,"reference_mean_power":0.0022835195719053173,"output_mean_power":0.0018198524162568442},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.817352658591746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000450774313706133,"output_power":0.00041524854396782625,"reference_active":true,"delta_db":-0.3565104253793622},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021888119814572043,"output_power":0.00017935828213173409,"reference_active":true,"delta_db":-0.8648702169508691},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.753791744185585e-05,"output_power":2.116748909174303e-05,"reference_active":true,"delta_db":-2.4880083018392427}],"active_band_count":3,"active_band_mae_db":1.2364629813898247}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022259441477048914,"log_rms_mae_db":1.7494387560536924,"rms_envelope_nrmse":0.14250535545554677,"envelope_correlation":0.9906939717439611,"rms_gain_error_db":-0.9854901866115364,"reference_mean_power":0.002294705859570105,"output_mean_power":0.0018288496602333942},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.266680044598684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042660839170957063,"output_power":0.00038919794475069617,"reference_active":true,"delta_db":-0.3985885588778448},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012272337901795565,"output_power":9.976298970306089e-05,"reference_active":true,"delta_db":-0.8995784883775317},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2413713648325596e-06,"output_power":1.987121222137299e-06,"reference_active":true,"delta_db":-2.125044295506394}],"active_band_count":3,"active_band_mae_db":1.1410704475872568}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002244694917839089,"log_rms_mae_db":3.0315349063401302,"rms_envelope_nrmse":0.1903077236205631,"envelope_correlation":0.9742880473224742,"rms_gain_error_db":-0.24014542044865717,"reference_mean_power":0.001082154267593255,"output_mean_power":0.001023940295833619},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.600154760198564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003310843525581766,"output_power":0.0003329456582785318,"reference_active":true,"delta_db":0.02434699960544687},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.047815482126218e-05,"output_power":0.00010491622764477823,"reference_active":true,"delta_db":0.6429893159580146},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6938491072881828e-05,"output_power":1.0324578008482687e-05,"reference_active":true,"delta_db":-2.1500240959594477}],"active_band_count":3,"active_band_mae_db":0.939120137174303}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018235367957614062,"log_rms_mae_db":2.1865165114954817,"rms_envelope_nrmse":0.12264924738996796,"envelope_correlation":0.9887810636904627,"rms_gain_error_db":-0.2403552699825638,"reference_mean_power":0.0010875823794426604,"output_mean_power":0.0010290266815987156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.047574386178107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032394555601725244,"output_power":0.00032359251208381143,"reference_active":true,"delta_db":-0.0047356303793725},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.774463594481584e-05,"output_power":5.5005266491364816e-05,"reference_active":true,"delta_db":0.6147968693324671},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5157059923891315e-06,"output_power":1.0626878942265647e-06,"reference_active":true,"delta_db":-1.5420923433704368}],"active_band_count":3,"active_band_mae_db":0.7205416143607589}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014040871531833865,"log_rms_mae_db":2.924653496560798,"rms_envelope_nrmse":0.26714425793085517,"envelope_correlation":0.9535465789129298,"rms_gain_error_db":-1.3710789594155803,"reference_mean_power":0.00036159326722347865,"output_mean_power":0.0002637014024274488},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2796709554971262e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010222442860363177,"output_power":7.105638188055053e-05,"reference_active":true,"delta_db":-1.5795160181086065},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.551583786781792e-05,"output_power":3.8148282490498456e-05,"reference_active":true,"delta_db":-0.7668755179771116},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1818852015306106e-05,"output_power":7.27182036324508e-06,"reference_active":true,"delta_db":-2.1093215293911496}],"active_band_count":3,"active_band_mae_db":1.4852376884922893}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010845211232047094,"log_rms_mae_db":2.266528137116118,"rms_envelope_nrmse":0.20513807520344066,"envelope_correlation":0.9782789867998238,"rms_gain_error_db":-1.3712332456113465,"reference_mean_power":0.00036337479954251354,"output_mean_power":0.00026499121756901674},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.941433739824423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.608975832570786e-05,"output_power":6.626816032126216e-05,"reference_active":true,"delta_db":-1.6137218684091104},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4640366344588205e-05,"output_power":2.0134133377716346e-05,"reference_active":true,"delta_db":-0.8771421923271543},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.52028436774113e-07,"output_power":7.397612159328375e-07,"reference_active":true,"delta_db":-1.0955836234684675}],"active_band_count":3,"active_band_mae_db":1.1954825614015774}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012084003773539908,"log_rms_mae_db":3.033643697994832,"rms_envelope_nrmse":0.31072916926113286,"envelope_correlation":0.9575666118413468,"rms_gain_error_db":-2.239090375462926,"reference_mean_power":0.00030833628390941907,"output_mean_power":0.00018412620259996863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.296444677120579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.747295044804844e-05,"output_power":3.212801917542715e-05,"reference_active":true,"delta_db":-1.6956227366223016},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.9782403314319074e-05,"output_power":2.7605836424882434e-05,"reference_active":true,"delta_db":-2.5607494864948843},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3740568823166804e-05,"output_power":7.092188261669965e-06,"reference_active":true,"delta_db":-2.872244563370054}],"active_band_count":3,"active_band_mae_db":2.3762055954957466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010527157253957593,"log_rms_mae_db":2.454012427598112,"rms_envelope_nrmse":0.2694444803282344,"envelope_correlation":0.9779596795194833,"rms_gain_error_db":-2.239090289112013,"reference_mean_power":0.0003098592063705906,"output_mean_power":0.00018503563518592635},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.056454468513195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.721882315895358e-05,"output_power":3.5776579774063195e-05,"reference_active":true,"delta_db":-2.0394010141672334},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.889894420146288e-05,"output_power":1.587743962686601e-05,"reference_active":true,"delta_db":-2.6010150643181125},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2532553850736717e-06,"output_power":6.372248043943289e-07,"reference_active":true,"delta_db":-2.9374690673913433}],"active_band_count":3,"active_band_mae_db":2.52596171529223}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"golf","status":"measured","mos":{"status":"measured","value":2.960852861404419,"predicted_mos":2.960852861404419,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"f4502afe54628ba25daf0eddc52989b3794d45441d98954c3580a416f5ec4120","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2192149349993997},"started_at":"2026-10-13T05:02:31.385032+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002734950000558456,"native_input_samples":144000,"native_input_sha256":"a1110095257a447a07e6d07a86270e6cf31be27a9e984ac5685db560f0a7fc5a","output_sha256":"26f06e50a8988438a012b69c2d50e11d27caf185bb89c234c2fb4184ed6b2bef","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5482784609994269,"synthesis_s":0.11703360200044699,"total_s":0.6653120629998739,"rtf":0.11088534383331232,"output_peak":0.6858320236206055,"output_rms":0.043458358933728354,"finished_at":"2026-10-13T05:02:32.061552+00:00","analysis_stages":{"external_f0_s":0.08375349500056473},"evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"26f06e50a8988438a012b69c2d50e11d27caf185bb89c234c2fb4184ed6b2bef","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.23713794842485,"output_peak":0.6873570537972867},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.799669008247352,"spectral_convergence":0.6318328933007817,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.474264082102926,"spectral_convergence":0.399700451379757,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.606467285167311,"spectral_convergence":0.4200430682197708,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00019001975934330962,"output_power":0.00019136429285272598,"delta_db":0.030621411651803983,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007157859536777457,"output_power":0.0006520999383128371,"delta_db":-0.4046901257703649,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006190490362901089,"output_power":0.0005050074525910633,"delta_db":-0.8842726462371729,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021345522295290577,"output_power":0.00015755511663050325,"delta_db":-1.3187427432313,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003714460323166502,"output_power":0.00023406762472009845,"delta_db":-2.005543756886656,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017555987339336275,"output_power":0.00014914154249686263,"delta_db":-0.7082662850766214,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8920228281423198,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8266003188722608,"pitch":null,"output_rms_dbfs":-27.23713794842485,"output_peak":0.6873570537972867,"reference_rms_dbfs":-26.41053762955259,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9601334585058625},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8292247347131407,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04780497754391359,"rms_dbfs":-26.410537629552586,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.401005889774584,"mean":-3.3905589719845578e-06,"zero_energy":false},"output":{"rms":0.04346534214295444,"rms_dbfs":-27.23713794842485,"peak":0.6873570537972867,"peak_dbfs":-3.2563521231004944,"crest_factor_db":23.98078582532435,"mean":0.0005481462497598258,"zero_energy":false},"rms_difference_db":-0.8266003188722628},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":2.268128373829532e-06,"output_rms_dbfs":-56.44332368444034,"output_zero_energy":false,"output_peak":0.022070670549623948,"output_energy_fraction":0.00012626853841973714,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":70,"output_estimated_voiced_frames":110,"both_estimated_voiced_frames":67,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":43,"paired_voiced_fraction":0.9571428571428572,"absolute_difference_cents_median":4.716140851464905,"absolute_difference_cents_p95":14.474955247588234,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002801281523487693,"log_rms_mae_db":1.9682255670502922,"rms_envelope_nrmse":0.1644148308488779,"envelope_correlation":0.9808695467815088,"rms_gain_error_db":-0.8266496360098129,"reference_mean_power":0.0022871828105930103,"output_mean_power":0.0018907578614952647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.833579368822161e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045222907461210296,"output_power":0.0003762543375554038,"reference_active":true,"delta_db":-0.7987696518449682},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002185437078013187,"output_power":0.0001815657792005306,"reference_active":true,"delta_db":-0.8050430922827676},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7353998043388426e-05,"output_power":3.289108074339565e-05,"reference_active":true,"delta_db":-0.5525894782617686}],"active_band_count":3,"active_band_mae_db":0.7188007407965015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00022259441477048914,"log_rms_mae_db":1.4034907569514938,"rms_envelope_nrmse":0.12437417293713035,"envelope_correlation":0.99162576682245,"rms_gain_error_db":-0.826768132111147,"reference_mean_power":0.002298281527014674,"output_mean_power":0.0018998810593420848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.281069513850263e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042779585985046646,"output_power":0.0003547290197570497,"reference_active":true,"delta_db":-0.8133985842071123},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012243773200075952,"output_power":0.00010147832918174638,"reference_active":true,"delta_db":-0.815419682152187},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.228034325184251e-06,"output_power":2.7951504798455955e-06,"reference_active":true,"delta_db":-0.6253295058010611}],"active_band_count":3,"active_band_mae_db":0.75138259072012}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002244694917839089,"log_rms_mae_db":2.028381095820526,"rms_envelope_nrmse":0.18166258088781506,"envelope_correlation":0.9791931031303616,"rms_gain_error_db":-0.6635634891732585,"reference_mean_power":0.0010839598894747994,"output_mean_power":0.000930372495864491},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.61555520656734e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033218196433655627,"output_power":0.0002915714938775828,"reference_active":true,"delta_db":-0.5663098696588507},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.054413619046297e-05,"output_power":8.549328865392843e-05,"reference_active":true,"delta_db":-0.2492830650353672},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6879300884038657e-05,"output_power":1.6074174362481145e-05,"reference_active":true,"delta_db":-0.21225779850359827}],"active_band_count":3,"active_band_mae_db":0.3426169110659387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00018235367957614062,"log_rms_mae_db":1.4900708741800377,"rms_envelope_nrmse":0.13541117571113898,"envelope_correlation":0.9894944504086676,"rms_gain_error_db":-0.6635134364360116,"reference_mean_power":0.0010893814656256983,"output_mean_power":0.0009350366596039925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.061628019969531e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032488085926962767,"output_power":0.00028710884400255724,"reference_active":true,"delta_db":-0.5367755457748405},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.776217549129917e-05,"output_power":4.5551136139100445e-05,"reference_active":true,"delta_db":-0.20584886353831391},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5133805601764518e-06,"output_power":1.3200322431209606e-06,"reference_active":true,"delta_db":-0.5936361160871985}],"active_band_count":3,"active_band_mae_db":0.44542017513345095}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014040871531833865,"log_rms_mae_db":2.2146255590289563,"rms_envelope_nrmse":0.26479333635488134,"envelope_correlation":0.9652775594654152,"rms_gain_error_db":-1.795174009008479,"reference_mean_power":0.0003621807558570459,"output_mean_power":0.00023955650583508856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.2812718676611613e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010239596396480377,"output_power":6.23847879443376e-05,"reference_active":true,"delta_db":-2.152041356551617},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.53694418159963e-05,"output_power":2.847202273627893e-05,"reference_active":true,"delta_db":-2.023451143369219},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.175098246832859e-05,"output_power":9.481784853268193e-06,"reference_active":true,"delta_db":-0.931840816076121}],"active_band_count":3,"active_band_mae_db":1.702444438665652}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010845211232047094,"log_rms_mae_db":1.8356643748694428,"rms_envelope_nrmse":0.21851160550704898,"envelope_correlation":0.9859109140496686,"rms_gain_error_db":-1.7960260148862317,"reference_mean_power":0.00036394451258683794,"output_mean_power":0.00024067588317152073},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.9428625628949483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.617937994140053e-05,"output_power":5.845639123571232e-05,"reference_active":true,"delta_db":-2.162499719109145},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4514160224375586e-05,"output_power":1.6646543582355685e-05,"reference_active":true,"delta_db":-1.6809294817638407},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.460539236706243e-07,"output_power":8.204319221412883e-07,"reference_active":true,"delta_db":-0.6187334128310494}],"active_band_count":3,"active_band_mae_db":1.487387537901345}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00012084003773539908,"log_rms_mae_db":2.5761626163618043,"rms_envelope_nrmse":0.2095452336210836,"envelope_correlation":0.97588497179205,"rms_gain_error_db":-1.0843246241458098,"reference_mean_power":0.00030883441194739516,"output_mean_power":0.00024059867056894045},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.2960709374052985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7526160415921234e-05,"output_power":3.54003123136327e-05,"reference_active":true,"delta_db":-1.2792563602129987},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.968029437600793e-05,"output_power":3.7771489012968724e-05,"reference_active":true,"delta_db":-1.190200547947225},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.369557528951656e-05,"output_power":1.0488314464372812e-05,"reference_active":true,"delta_db":-1.1587457968212957}],"active_band_count":3,"active_band_mae_db":1.2094009016605065}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010527157253957593,"log_rms_mae_db":2.160057876554168,"rms_envelope_nrmse":0.16377103153394737,"envelope_correlation":0.9884850933623474,"rms_gain_error_db":-1.0845019577451147,"reference_mean_power":0.00031035492024765324,"output_mean_power":0.00024177335614723427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0560999359614099e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.722393320507271e-05,"output_power":4.268146007638213e-05,"reference_active":true,"delta_db":-1.2733843745699747},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8819289555339205e-05,"output_power":2.2068007442891395e-05,"reference_active":true,"delta_db":-1.1592014874941836},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2506381703017274e-06,"output_power":9.517512923819658e-07,"reference_active":true,"delta_db":-1.1860820428615086}],"active_band_count":3,"active_band_mae_db":1.2062226349752223}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.2855136394500732,"predicted_mos":2.2855136394500732,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"280f32f4d3c097438fe7d02e0b11a20a5c076d9e0573ad55782d12b70ca51abb","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1849020439985907},"started_at":"2026-10-13T07:36:53.530724+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003568034000636544,"native_input_samples":144000,"native_input_sha256":"a1110095257a447a07e6d07a86270e6cf31be27a9e984ac5685db560f0a7fc5a","output_sha256":"338a772a80f1634adb9804ccf15dfc0bf98ef21c145ee6b16b5d420dff5a11de","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3147130990000733,"synthesis_s":3.4392447780010116,"total_s":5.753957877001085,"rtf":0.9589929795001808,"output_peak":0.6917146444320679,"output_rms":0.04577143061450581,"finished_at":"2026-10-13T07:36:59.300849+00:00","analysis_stages":{"external_f0_s":1.5666732499994396},"evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"338a772a80f1634adb9804ccf15dfc0bf98ef21c145ee6b16b5d420dff5a11de","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.785552869808097,"output_peak":0.6670108263106369},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.950812498260714,"spectral_convergence":0.643473835196691,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.6338042486223365,"spectral_convergence":0.4476748909418076,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.357530256767092,"spectral_convergence":0.44092994756382103,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.00028627054198502856,"delta_db":1.7848441041034504,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006564934839576244,"delta_db":-0.36833298037005,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.0004967762766065469,"delta_db":-0.9483980893962564,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.0001887515104725787,"delta_db":-0.5269020108068068,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.00033271476410819757,"delta_db":-0.4712191412779172,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00013630091654490625,"delta_db":-1.0923459071220654,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8653403721794244,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.36585417406748094,"pitch":null,"output_rms_dbfs":-26.78337891415366,"output_peak":0.6670108263106369,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.980715667883381},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.82619492120882,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04776653768728406,"rms_dbfs":-26.41752474008618,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.40799300030818,"mean":-3.276507059733073e-06,"zero_energy":false},"output":{"rms":0.04579636987442217,"rms_dbfs":-26.78337891415366,"peak":0.6670108263106369,"peak_dbfs":-3.5173423390959755,"crest_factor_db":23.266036575057683,"mean":0.00015082277170576366,"zero_energy":false},"rms_difference_db":-0.3658541740674792},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":2.625457064054569e-05,"output_rms_dbfs":-45.80795079641882,"output_zero_energy":false,"output_peak":0.27228219316000335,"output_energy_fraction":0.0013144138022541374,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":71,"output_estimated_voiced_frames":79,"both_estimated_voiced_frames":66,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9295774647887324,"absolute_difference_cents_median":6.495538840764819,"absolute_difference_cents_p95":26.12050221896869,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002801281523487693,"log_rms_mae_db":2.9046776145696236,"rms_envelope_nrmse":0.23439370919534877,"envelope_correlation":0.9476589436473096,"rms_gain_error_db":-0.36627377053027754,"reference_mean_power":0.0022835195719053173,"output_mean_power":0.002098830473516406},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.817352658591746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000450774313706133,"output_power":0.00040130610123791175,"reference_active":true,"delta_db":-0.504833982897786},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021888119814572043,"output_power":0.00019349415768279537,"reference_active":true,"delta_db":-0.5354060078757669},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.753791744185585e-05,"output_power":3.2063356555355703e-05,"reference_active":true,"delta_db":-0.6846119026761737}],"active_band_count":3,"active_band_mae_db":0.5749506311499089}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022259441477048914,"log_rms_mae_db":2.1069987537234023,"rms_envelope_nrmse":0.14040860833889846,"envelope_correlation":0.9800082254133993,"rms_gain_error_db":-0.3673171024343954,"reference_mean_power":0.002294705859570105,"output_mean_power":0.0021086053999764946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.266680044598684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042660839170957063,"output_power":0.00037619266912473285,"reference_active":true,"delta_db":-0.5461906590802966},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012272337901795565,"output_power":0.00010936599900782005,"reference_active":true,"delta_db":-0.5004498007710021},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2413713648325596e-06,"output_power":2.663566123070921e-06,"reference_active":true,"delta_db":-0.8526530850218286}],"active_band_count":3,"active_band_mae_db":0.6330978482910424}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002244694917839089,"log_rms_mae_db":2.823667188773014,"rms_envelope_nrmse":0.24957750932753694,"envelope_correlation":0.9567402018921967,"rms_gain_error_db":-0.6963035330830635,"reference_mean_power":0.001082154267593255,"output_mean_power":0.0009218469500765346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.600154760198564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003310843525581766,"output_power":0.0002811959949980672,"reference_active":true,"delta_db":-0.7092952511361224},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.047815482126218e-05,"output_power":7.76539956049724e-05,"reference_active":true,"delta_db":-0.663799283562752},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6938491072881828e-05,"output_power":1.282638168793178e-05,"reference_active":true,"delta_db":-1.2077056005292395}],"active_band_count":3,"active_band_mae_db":0.8602667117427046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018235367957614062,"log_rms_mae_db":1.9551059901723244,"rms_envelope_nrmse":0.1581753706836568,"envelope_correlation":0.9842003538379605,"rms_gain_error_db":-0.6962731199221448,"reference_mean_power":0.0010875823794426604,"output_mean_power":0.0009264774445593924},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.047574386178107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032394555601725244,"output_power":0.00027443704682618,"reference_active":true,"delta_db":-0.7202928923462111},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.774463594481584e-05,"output_power":4.114318039791015e-05,"reference_active":true,"delta_db":-0.6462672623551841},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5157059923891315e-06,"output_power":1.0877382144791393e-06,"reference_active":true,"delta_db":-1.4409058115274356}],"active_band_count":3,"active_band_mae_db":0.9358219887429436}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014040871531833865,"log_rms_mae_db":2.7131377032884583,"rms_envelope_nrmse":0.28322215159877245,"envelope_correlation":0.9344161330876866,"rms_gain_error_db":-0.43028583465011594,"reference_mean_power":0.00036159326722347865,"output_mean_power":0.00032748525595639245},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2796709554971262e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010222442860363177,"output_power":9.148820093056345e-05,"reference_active":true,"delta_db":-0.4818960419293278},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.551583786781792e-05,"output_power":3.908379513491998e-05,"reference_active":true,"delta_db":-0.6616581356866869},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1818852015306106e-05,"output_power":1.0518276932765363e-05,"reference_active":true,"delta_db":-0.5063069380497114}],"active_band_count":3,"active_band_mae_db":0.5499537052219087}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010845211232047094,"log_rms_mae_db":1.6713811982307498,"rms_envelope_nrmse":0.14555308427904992,"envelope_correlation":0.9822148506858012,"rms_gain_error_db":-0.4301340185666054,"reference_mean_power":0.00036337479954251354,"output_mean_power":0.0003291102461485683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.941433739824423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.608975832570786e-05,"output_power":8.557449989263174e-05,"reference_active":true,"delta_db":-0.5033273134591821},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4640366344588205e-05,"output_power":2.0617654055407422e-05,"reference_active":true,"delta_db":-0.7740791218116677},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.52028436774113e-07,"output_power":8.485183457663743e-07,"reference_active":true,"delta_db":-0.4998868419932948}],"active_band_count":3,"active_band_mae_db":0.5924310924213816}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012084003773539908,"log_rms_mae_db":2.8819974784301796,"rms_envelope_nrmse":0.28337089697782236,"envelope_correlation":0.9422964729228717,"rms_gain_error_db":-0.8652560754591189,"reference_mean_power":0.00030833628390941907,"output_mean_power":0.00025263820465298175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.296444677120579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.747295044804844e-05,"output_power":4.0652454616564944e-05,"reference_active":true,"delta_db":-0.6735945034667445},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.9782403314319074e-05,"output_power":4.149649854644717e-05,"reference_active":true,"delta_db":-0.7906440623556925},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3740568823166804e-05,"output_power":1.3512729670869618e-05,"reference_active":true,"delta_db":-0.07261623170477502}],"active_band_count":3,"active_band_mae_db":0.5122849325090707}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010527157253957593,"log_rms_mae_db":1.8584058043091045,"rms_envelope_nrmse":0.17801829708493344,"envelope_correlation":0.9796828845823368,"rms_gain_error_db":-0.8652630349532815,"reference_mean_power":0.0003098592063705906,"output_mean_power":0.0002538856185014162},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.056454468513195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.721882315895358e-05,"output_power":4.901980084723467e-05,"reference_active":true,"delta_db":-0.6716737875195345},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.889894420146288e-05,"output_power":2.4749646091284436e-05,"reference_active":true,"delta_db":-0.6731298336416102},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2532553850736717e-06,"output_power":1.2684439343507048e-06,"reference_active":true,"delta_db":0.05231696624128079}],"active_band_count":3,"active_band_mae_db":0.46570686246747517}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.741586923599243,"predicted_mos":2.741586923599243,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"0af1a1353821e43fd00c794205f70ba0abbdd861ad108622fcbf355840fc15a8","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.433213329000864},"started_at":"2026-10-13T05:10:39.552653+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029652829998667585,"native_input_samples":144000,"native_input_sha256":"a1110095257a447a07e6d07a86270e6cf31be27a9e984ac5685db560f0a7fc5a","output_sha256":"17b225f01e53378532fddf4175db0e798ac78b80a3d6a87df78c7f230241290b","output_samples":144240,"length_delta_samples":240,"analysis_s":1.1427051730006497,"synthesis_s":0.13835387799917953,"total_s":1.2810590509998292,"rtf":0.21350984183330488,"output_peak":0.7564247846603394,"output_rms":0.0452091868956257,"finished_at":"2026-10-13T05:10:40.845227+00:00","analysis_stages":{"external_f0_s":0.1893794000006892},"evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"17b225f01e53378532fddf4175db0e798ac78b80a3d6a87df78c7f230241290b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.893528844885086,"output_peak":0.756950120233251},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.326189355658604,"spectral_convergence":0.318278415107916,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.7051758720333785,"spectral_convergence":0.2943906875933381,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.910883098389268,"spectral_convergence":0.3207422704745164,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.00018099401945273762,"delta_db":-0.20628025025540875,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006676900286510233,"delta_db":-0.2948882739809986,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.0005536283950768875,"delta_db":-0.47782300434905395,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.00018545056218890888,"delta_db":-0.6035248259133359,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.0003051217840017651,"delta_db":-0.8472077213426678,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00015530479204203132,"delta_db":-0.5254851052952438,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4925348635227848,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.46877373320826915,"pitch":null,"output_rms_dbfs":-26.88629847329445,"output_peak":0.756950120233251,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.980749442027084},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8982950509558649,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04776653768728406,"rms_dbfs":-26.41752474008618,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.40799300030818,"mean":-3.276507059733073e-06,"zero_energy":false},"output":{"rms":0.04525692854595143,"rms_dbfs":-26.88629847329445,"peak":0.756950120233251,"peak_dbfs":-2.4186547540190255,"crest_factor_db":24.467643719275426,"mean":7.47138821345709e-05,"zero_energy":false},"rms_difference_db":-0.46877373320826976},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":7.798400288985676e-06,"output_rms_dbfs":-51.07994476401699,"output_zero_energy":false,"output_peak":0.0583444728993138,"output_energy_fraction":0.0003997833197542489,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":71,"output_estimated_voiced_frames":80,"both_estimated_voiced_frames":69,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.971830985915493,"absolute_difference_cents_median":1.864439379001631,"absolute_difference_cents_p95":8.818959984999204,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002801281523487693,"log_rms_mae_db":1.193292804106003,"rms_envelope_nrmse":0.10804506482319935,"envelope_correlation":0.9912795891109712,"rms_gain_error_db":-0.4687509075727591,"reference_mean_power":0.0022835195719053173,"output_mean_power":0.0020498857080133237},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.817352658591746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000450774313706133,"output_power":0.0004091837652613686,"reference_active":true,"delta_db":-0.42040766496219745},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021888119814572043,"output_power":0.0001793300590168416,"reference_active":true,"delta_db":-0.8655536594651124},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.753791744185585e-05,"output_power":2.4714628060239735e-05,"reference_active":true,"delta_db":-1.8151609583789736}],"active_band_count":3,"active_band_mae_db":1.0337074276020946}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022259441477048914,"log_rms_mae_db":0.7133282598134538,"rms_envelope_nrmse":0.07641828753769864,"envelope_correlation":0.9965564054617578,"rms_gain_error_db":-0.4687314202240397,"reference_mean_power":0.002294705859570105,"output_mean_power":0.0020599367355901535},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.266680044598684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042660839170957063,"output_power":0.00038632586145367415,"reference_active":true,"delta_db":-0.430756124077329},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012272337901795565,"output_power":0.00010148840257941939,"reference_active":true,"delta_db":-0.8251088766976054},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2413713648325596e-06,"output_power":1.957685413377924e-06,"reference_active":true,"delta_db":-2.1898588619573185}],"active_band_count":3,"active_band_mae_db":1.148574620910751}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002244694917839089,"log_rms_mae_db":1.4479683838094415,"rms_envelope_nrmse":0.12293845857499441,"envelope_correlation":0.9906112845178827,"rms_gain_error_db":-0.43688473342900014,"reference_mean_power":0.001082154267593255,"output_mean_power":0.0009785898414676202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.600154760198564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003310843525581766,"output_power":0.0003014528944982919,"reference_active":true,"delta_db":-0.4071919771979237},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.047815482126218e-05,"output_power":7.125721729003849e-05,"reference_active":true,"delta_db":-1.0371487672134003},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6938491072881828e-05,"output_power":1.1446316947883602e-05,"reference_active":true,"delta_db":-1.702089522483293}],"active_band_count":3,"active_band_mae_db":1.0488100889648724}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018235367957614062,"log_rms_mae_db":0.8907441947910008,"rms_envelope_nrmse":0.08571434047435184,"envelope_correlation":0.9960542123977331,"rms_gain_error_db":-0.436877128079742,"reference_mean_power":0.0010875823794426604,"output_mean_power":0.000983500193963669},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.047574386178107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032394555601725244,"output_power":0.00029212572692349943,"reference_active":true,"delta_db":-0.44902220442841345},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.774463594481584e-05,"output_power":3.851016523340664e-05,"reference_active":true,"delta_db":-0.9334920414293408},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5157059923891315e-06,"output_power":8.711639195556983e-07,"reference_active":true,"delta_db":-2.405150873905763}],"active_band_count":3,"active_band_mae_db":1.2625550399211725}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014040871531833865,"log_rms_mae_db":1.6239941334848873,"rms_envelope_nrmse":0.16231680689296257,"envelope_correlation":0.982689864029859,"rms_gain_error_db":-0.7225288629186927,"reference_mean_power":0.00036159326722347865,"output_mean_power":0.0003061733944628569},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2796709554971262e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010222442860363177,"output_power":8.858513366664579e-05,"reference_active":true,"delta_db":-0.6219384687063866},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.551583786781792e-05,"output_power":3.315945835436878e-05,"reference_active":true,"delta_db":-1.375551138299784},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1818852015306106e-05,"output_power":6.876205989582511e-06,"reference_active":true,"delta_db":-2.352264165390741}],"active_band_count":3,"active_band_mae_db":1.449917924132304}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010845211232047094,"log_rms_mae_db":1.0751698174688684,"rms_envelope_nrmse":0.10924268050119104,"envelope_correlation":0.9945832546956952,"rms_gain_error_db":-0.7224346761338469,"reference_mean_power":0.00036337479954251354,"output_mean_power":0.0003076885517124723},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.941433739824423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.608975832570786e-05,"output_power":8.294333116862554e-05,"reference_active":true,"delta_db":-0.638956275898913},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4640366344588205e-05,"output_power":1.7267864789041517e-05,"reference_active":true,"delta_db":-1.5440852110128678},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.52028436774113e-07,"output_power":5.141954656388541e-07,"reference_active":true,"delta_db":-2.6752167825227477}],"active_band_count":3,"active_band_mae_db":1.619419423144843}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012084003773539908,"log_rms_mae_db":1.7705886254228547,"rms_envelope_nrmse":0.14887854152094773,"envelope_correlation":0.9866763206591171,"rms_gain_error_db":-0.6360556725073141,"reference_mean_power":0.00030833628390941907,"output_mean_power":0.00026632937302695437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.296444677120579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.747295044804844e-05,"output_power":4.775617812211688e-05,"reference_active":true,"delta_db":0.025833393732519364},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.9782403314319074e-05,"output_power":4.3976607005705954e-05,"reference_active":true,"delta_db":-0.5385414041278614},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3740568823166804e-05,"output_power":9.24869889750425e-06,"reference_active":true,"delta_db":-1.7192407103920675}],"active_band_count":3,"active_band_mae_db":0.7612051694174827}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010527157253957593,"log_rms_mae_db":1.1724269892884986,"rms_envelope_nrmse":0.10352600918020058,"envelope_correlation":0.9950467995278159,"rms_gain_error_db":-0.6359832826016941,"reference_mean_power":0.0003098592063705906,"output_mean_power":0.00026764927784988},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.056454468513195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.721882315895358e-05,"output_power":5.594123865483748e-05,"reference_active":true,"delta_db":-0.09806842942315327},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.889894420146288e-05,"output_power":2.4678200848373896e-05,"reference_active":true,"delta_db":-0.6856848196604983},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2532553850736717e-06,"output_power":7.65495105081457e-07,"reference_active":true,"delta_db":-2.1409716142567454}],"active_band_count":3,"active_band_mae_db":0.9749082877801323}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0002","method":"magphase","status":"measured","mos":{"status":"measured","value":2.210886240005493,"predicted_mos":2.210886240005493,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":290211,"input_sample_rate":48000,"input_seconds":6.0460625,"input_float32_sha256":"4b60b52b4aa9e4648ba3fce1075b3d107db1148e5a1c92c8b050980003ecfefc","internal_samples":96737,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1012301559985644},"started_at":"2026-10-13T05:15:12.098691+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008694575000845362,"native_input_samples":288000,"native_input_sha256":"760aaac0ad7f4359bd39e1588400f2c153c1ecf41f32f32bb7bfd6e18910251e","output_sha256":"1f459afa5a9e4cfbe992211b3ff969d1654b2d3f1a9063291ddb1787ca18522d","output_samples":290211,"length_delta_samples":2211,"analysis_s":3.06946360099937,"synthesis_s":1.0255987040000036,"total_s":4.095062304999374,"rtf":0.6825103841665623,"output_peak":0.8261548100976619,"output_rms":0.045622280729040725,"finished_at":"2026-10-13T05:15:16.215766+00:00","evaluation_input":{"reference":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"1f459afa5a9e4cfbe992211b3ff969d1654b2d3f1a9063291ddb1787ca18522d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96737,"length_delta_samples":737,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.814579868573173,"output_peak":0.823258314764248},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9923814052534191,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":11.569234671302132,"spectral_convergence":0.9056723150448557,"spectral_convergence_status":"ok","amplitude_floor":1.602383167696498e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":11.12436370368987,"spectral_convergence":0.8570407189829892,"spectral_convergence_status":"ok","amplitude_floor":1.3018377539140758e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.457422511060482,"spectral_convergence":0.7947077544189569,"spectral_convergence_status":"ok","amplitude_floor":8.15121073207966e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00018979827012669627,"output_power":0.0001605150437383645,"delta_db":-0.7277650837712644,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007146011617002843,"output_power":0.0006407641111651215,"delta_db":-0.47365538713779903,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006180173350849692,"output_power":0.0005447727925744833,"delta_db":-0.5478524740194379,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00021309861976674557,"output_power":0.0001959984195174141,"delta_db":-0.3632806748347605,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003708463313924785,"output_power":0.0003801323432455277,"delta_db":0.1074083566244853,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001752804045595543,"output_power":0.00017292659899019268,"delta_db":-0.05871566801666805,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.37977960740073585,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3704083942536057,"pitch":null,"output_rms_dbfs":-26.787933134339784,"output_peak":0.823258314764248,"reference_rms_dbfs":-26.41752474008618,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":10.717006962017493},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.3533701037095919,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04776653768728406,"rms_dbfs":-26.41752474008618,"peak":0.79345703125,"peak_dbfs":-2.0095317397779993,"crest_factor_db":24.40799300030818,"mean":-3.276507059733073e-06,"zero_energy":false},"output":{"rms":0.04577236403367324,"rms_dbfs":-26.787933134339784,"peak":0.823258314764248,"peak_dbfs":-1.6892774859197397,"crest_factor_db":25.09865564842005,"mean":0.0006795477477192087,"zero_energy":false},"rms_difference_db":-0.37040839425360517},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.208099222858436e-06,"threshold_dbfs":-52.07041350449042,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":63,"quiet_samples":10080,"reference_mean_power":2.377151995749464e-06,"output_mean_power":7.939032817482051e-05,"output_rms_dbfs":-41.00232402814159,"output_zero_energy":false,"output_peak":0.0998216613164027,"output_energy_fraction":0.003978782596987947,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":71,"output_estimated_voiced_frames":66,"both_estimated_voiced_frames":40,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.5633802816901409,"absolute_difference_cents_median":7.410628000142747,"absolute_difference_cents_p95":92.93723116675484,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002801281523487693,"log_rms_mae_db":7.088784158060475,"rms_envelope_nrmse":0.6656990452134773,"envelope_correlation":0.5757684963895909,"rms_gain_error_db":-0.370971462753407,"reference_mean_power":0.0022835195719053173,"output_mean_power":0.0020965614304109042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.817352658591746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000450774313706133,"output_power":0.000428761127933147,"reference_active":true,"delta_db":-0.21743756107777107},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021888119814572043,"output_power":0.00020006811381598156,"reference_active":true,"delta_db":-0.3903057959927689},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.753791744185585e-05,"output_power":3.2007720914661355e-05,"reference_active":true,"delta_db":-0.6921542318448841}],"active_band_count":3,"active_band_mae_db":0.43329919630514135}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022259441477048914,"log_rms_mae_db":5.5048401350264955,"rms_envelope_nrmse":0.5425692245909407,"envelope_correlation":0.6803229315783964,"rms_gain_error_db":-0.3795162937433608,"reference_mean_power":0.002294705859570105,"output_mean_power":0.0021026907066952305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.266680044598684e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042660839170957063,"output_power":0.0004031761071860357,"reference_active":true,"delta_db":-0.24534606833159858},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012272337901795565,"output_power":0.00011144271656325327,"reference_active":true,"delta_db":-0.41875614419537366},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2413713648325596e-06,"output_power":2.781033418977628e-06,"reference_active":true,"delta_db":-0.6652258342136487}],"active_band_count":3,"active_band_mae_db":0.4431093489135403}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002244694917839089,"log_rms_mae_db":9.072910137440013,"rms_envelope_nrmse":0.7142236826274518,"envelope_correlation":0.6232980972013173,"rms_gain_error_db":-0.5278018722475701,"reference_mean_power":0.001082154267593255,"output_mean_power":0.0009583165518223313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.600154760198564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003310843525581766,"output_power":0.0003043701177344976,"reference_active":true,"delta_db":-0.36536643671089664},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.047815482126218e-05,"output_power":8.066134182662055e-05,"reference_active":true,"delta_db":-0.49878292807176794},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6938491072881828e-05,"output_power":1.2465632944385621e-05,"reference_active":true,"delta_db":-1.3316038481018504}],"active_band_count":3,"active_band_mae_db":0.7319177376281717}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018235367957614062,"log_rms_mae_db":7.410344450800233,"rms_envelope_nrmse":0.613371575863979,"envelope_correlation":0.7011248567924552,"rms_gain_error_db":-0.5318575353451783,"reference_mean_power":0.0010875823794426604,"output_mean_power":0.0009622244967614066},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.047574386178107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032394555601725244,"output_power":0.00029628204282880694,"reference_active":true,"delta_db":-0.3876669614190573},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.774463594481584e-05,"output_power":4.2664307847150424e-05,"reference_active":true,"delta_db":-0.4885988180580422},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5157059923891315e-06,"output_power":1.0284400661316914e-06,"reference_active":true,"delta_db":-1.6843597999735158}],"active_band_count":3,"active_band_mae_db":0.8535418598168718}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014040871531833865,"log_rms_mae_db":9.160673876284829,"rms_envelope_nrmse":0.7597129121977451,"envelope_correlation":0.5297861467030996,"rms_gain_error_db":-0.2524807870510717,"reference_mean_power":0.00036159326722347865,"output_mean_power":0.0003411711164827861},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.2796709554971262e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010222442860363177,"output_power":9.773547059858562e-05,"reference_active":true,"delta_db":-0.1950248324653548},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.551583786781792e-05,"output_power":3.8747606884635794e-05,"reference_active":true,"delta_db":-0.6991765681528445},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1818852015306106e-05,"output_power":1.4771480078689962e-05,"reference_active":true,"delta_db":0.9684871825437749}],"active_band_count":3,"active_band_mae_db":0.6208961943873247}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010845211232047094,"log_rms_mae_db":7.170691064657334,"rms_envelope_nrmse":0.612632294585448,"envelope_correlation":0.6564303911681062,"rms_gain_error_db":-0.27673713452243615,"reference_mean_power":0.00036337479954251354,"output_mean_power":0.0003409424610633369},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.941433739824423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.608975832570786e-05,"output_power":9.13451803081032e-05,"reference_active":true,"delta_db":-0.21991463701679415},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4640366344588205e-05,"output_power":2.145667244695494e-05,"reference_active":true,"delta_db":-0.6008478907763263},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.52028436774113e-07,"output_power":1.163918838097875e-06,"reference_active":true,"delta_db":0.8727277651777593}],"active_band_count":3,"active_band_mae_db":0.5644967643236266}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012084003773539908,"log_rms_mae_db":9.357919229311227,"rms_envelope_nrmse":0.8192707861804288,"envelope_correlation":0.5263526170987397,"rms_gain_error_db":0.16439137194310424,"reference_mean_power":0.00030833628390941907,"output_mean_power":0.0003202312946486651},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.296444677120579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.747295044804844e-05,"output_power":5.0029625055158516e-05,"reference_active":true,"delta_db":0.22781024097137784},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.9782403314319074e-05,"output_power":4.852605928380053e-05,"reference_active":true,"delta_db":-0.11100834491215468},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3740568823166804e-05,"output_power":1.1030035339888931e-05,"reference_active":true,"delta_db":-0.9542780782891218}],"active_band_count":3,"active_band_mae_db":0.4310322213908848}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010527157253957593,"log_rms_mae_db":7.190641959253313,"rms_envelope_nrmse":0.6602847771716472,"envelope_correlation":0.6563448424321727,"rms_gain_error_db":0.157958591538727,"reference_mean_power":0.0003098592063705906,"output_mean_power":0.0003213366511780558},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.056454468513195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.721882315895358e-05,"output_power":5.898089610718417e-05,"reference_active":true,"delta_db":0.1317244531554033},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.889894420146288e-05,"output_power":2.695507220008299e-05,"reference_active":true,"delta_db":-0.30241477025164093},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2532553850736717e-06,"output_power":9.532687862150382e-07,"reference_active":true,"delta_db":-1.1882420665700901}],"active_band_count":3,"active_band_mae_db":0.5407937633257115}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"hnm","status":"measured","mos":{"status":"measured","value":1.9884307384490967,"predicted_mos":1.9884307384490967,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"9600364ca18b72f6b834bdefc49d6c8fcdcd1ad22d0446f9a6711c60efd001e9","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4243302040013077},"started_at":"2026-10-13T04:39:11.217393+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0888000108243432e-05,"native_input_samples":96000,"native_input_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","output_sha256":"13ff6641da35110d85ece58a175252e0d19bff7d06c8a16ab8fb968e9e813b3f","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1713752270006808,"synthesis_s":0.5691751289996319,"total_s":1.7405503560003126,"rtf":0.2900917260000521,"output_peak":0.4960893481116174,"output_rms":0.04306986815813211,"finished_at":"2026-10-13T04:39:12.964574+00:00","evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"13ff6641da35110d85ece58a175252e0d19bff7d06c8a16ab8fb968e9e813b3f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.350497272247688,"spectral_convergence":0.46280386364194076,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.677952382713085,"spectral_convergence":0.4335651021386391,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.854625736572594,"spectral_convergence":0.4595515686527461,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00016012583649373824,"delta_db":0.005919657296989483,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0005259677220362943,"delta_db":-0.07137042856194512,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.0002659499952268973,"delta_db":0.16538750121521797,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.0001528712717625208,"delta_db":-0.1045360460428449,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003595195075774369,"delta_db":0.049516980292256096,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.0003905792100619946,"delta_db":-0.35146825961448613,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12469981217062327,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07227914949145582,"pitch":null,"output_rms_dbfs":-27.31652915322413,"output_peak":0.4960893481116174,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.961025130511122},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.350497272247688,"spectral_convergence":0.46280386364194076,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.677952382713085,"spectral_convergence":0.4335651021386391,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.854625736572594,"spectral_convergence":0.4595515686527461,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00016012583649373824,"delta_db":0.005919657296989483,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0005259677220362943,"delta_db":-0.07137042856194512,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.0002659499952268973,"delta_db":0.16538750121521797,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.0001528712717625208,"delta_db":-0.1045360460428449,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003595195075774369,"delta_db":0.049516980292256096,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.0003905792100619946,"delta_db":-0.35146825961448613,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.12469981217062327,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07227914949145582,"pitch":null,"output_rms_dbfs":-27.31652915322413,"output_peak":0.4960893481116174,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.961025130511122},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8090642696946961,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04342976704277862,"rms_dbfs":-27.244250003732674,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.951942904471995,"mean":2.936585744222005e-05,"zero_energy":false},"output":{"rms":0.04306986815813211,"rms_dbfs":-27.31652915322413,"peak":0.4960893481116174,"peak_dbfs":-6.088801958208387,"crest_factor_db":21.227727195015742,"mean":-0.0001318078341631404,"zero_energy":false},"rms_difference_db":-0.07227914949145742},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":1.9351499746561034e-06,"output_rms_dbfs":-57.13285371398772,"output_zero_energy":false,"output_peak":0.013630237186884336,"output_energy_fraction":0.00014604798841010404,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":64,"output_estimated_voiced_frames":77,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":1.4164113517420889,"absolute_difference_cents_p95":8.872894735571197,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018816516597127566,"log_rms_mae_db":1.2545947307400227,"rms_envelope_nrmse":0.103049808101838,"envelope_correlation":0.9905581826285648,"rms_gain_error_db":-0.07199027324572298,"reference_mean_power":0.0018869472812584292,"output_mean_power":0.0018559263568113887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.448172186664129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039739520206635815,"output_power":0.0003928703738063455,"reference_active":true,"delta_db":-0.049733392920322465},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001684948117890412,"output_power":0.00016285104016090243,"reference_active":true,"delta_db":-0.1479599599495934},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9559595301379592e-05,"output_power":1.8662894212073585e-05,"reference_active":true,"delta_db":-0.20380870422871517}],"active_band_count":3,"active_band_mae_db":0.13383401903287703}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016656172387320294,"log_rms_mae_db":0.6830695230796252,"rms_envelope_nrmse":0.05482109317889502,"envelope_correlation":0.997125510015818,"rms_gain_error_db":-0.07163317657572706,"reference_mean_power":0.0018945898208991615,"output_mean_power":0.001863596482157169},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.514104383617564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003829793712831679,"output_power":0.0003807108076597473,"reference_active":true,"delta_db":-0.0258017606764222},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.831616093630877e-05,"output_power":7.552443801957077e-05,"reference_active":true,"delta_db":-0.1576388781660238},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3812315351816923e-06,"output_power":1.9282186915750203e-06,"reference_active":true,"delta_db":-0.9164533691048614}],"active_band_count":3,"active_band_mae_db":0.3666313359824358}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015092951870698086,"log_rms_mae_db":1.7857380571542716,"rms_envelope_nrmse":0.11530047104794984,"envelope_correlation":0.990678974679058,"rms_gain_error_db":-0.010080655992635974,"reference_mean_power":0.000631981580682194,"output_mean_power":0.0006305163534939529},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.167139764996573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018037000728302603,"output_power":0.00017354982878835007,"reference_active":true,"delta_db":-0.16740133251403327},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6944061726753776e-05,"output_power":2.7589027962443267e-05,"reference_active":true,"delta_db":0.1027333460053888},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.521461134259169e-06,"output_power":4.680712426214741e-06,"reference_active":true,"delta_db":1.2358906022265168}],"active_band_count":3,"active_band_mae_db":0.502008426915313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013588907795973108,"log_rms_mae_db":1.0024727017610955,"rms_envelope_nrmse":0.06426515699199024,"envelope_correlation":0.9969455670765086,"rms_gain_error_db":-0.009607191887881953,"reference_mean_power":0.0006345375227010602,"output_mean_power":0.0006331353897633714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.990151794989553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001715162379513309,"output_power":0.00016604665685833903,"reference_active":true,"delta_db":-0.14075106226375972},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2552077755571314e-05,"output_power":1.2356949688684074e-05,"reference_active":true,"delta_db":-0.06804342434224976},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0572536267135004e-07,"output_power":4.1084897458874607e-07,"reference_active":true,"delta_db":1.2835073852567405}],"active_band_count":3,"active_band_mae_db":0.4974339572875833}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":1.7889130330232768,"rms_envelope_nrmse":0.17157631437477555,"envelope_correlation":0.9765512030536306,"rms_gain_error_db":0.023471868149379408,"reference_mean_power":0.00026216016454011775,"output_mean_power":0.0002635808703894081},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5456728460340465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.502085853898133e-05,"output_power":7.46631674556514e-05,"reference_active":true,"delta_db":-0.020756197379761754},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.151594827944161e-05,"output_power":2.2448327165406595e-05,"reference_active":true,"delta_db":0.1842349156218671},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51387829077428e-06,"output_power":4.341579234927895e-06,"reference_active":true,"delta_db":-2.9247970605968723}],"active_band_count":3,"active_band_mae_db":1.0432627245328339}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":1.0319938457556168,"rms_envelope_nrmse":0.09661383058317143,"envelope_correlation":0.991835311970109,"rms_gain_error_db":0.024211175368072588,"reference_mean_power":0.00026328876508602793,"output_mean_power":0.00026476065392049453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3635149275752304e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.257955032157828e-05,"output_power":7.241666089699356e-05,"reference_active":true,"delta_db":-0.009757773313209015},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405048683009273e-06,"output_power":9.623608936850057e-06,"reference_active":true,"delta_db":0.09976918867193228},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.677553554297771e-07,"output_power":4.187894105758722e-07,"reference_active":true,"delta_db":-2.632271628564234}],"active_band_count":3,"active_band_mae_db":0.9139328635164583}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001568272998494035,"log_rms_mae_db":1.9833256603705456,"rms_envelope_nrmse":0.15422506262393654,"envelope_correlation":0.9856753142205583,"rms_gain_error_db":-0.3302917184751689,"reference_mean_power":0.0005797089728212456,"output_mean_power":0.0005372554758737285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.292445291957658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020605682430557343,"output_power":0.00020514960895760858,"reference_active":true,"delta_db":-0.019163088264438778},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001264586344183223,"output_power":0.00011461280680357399,"reference_active":true,"delta_db":-0.42715339422321613},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.269907486994092e-06,"output_power":1.0976576610162467e-05,"reference_active":true,"delta_db":0.7339151293155091}],"active_band_count":3,"active_band_mae_db":0.3934105372677213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012704534221320317,"log_rms_mae_db":1.0962586531433005,"rms_envelope_nrmse":0.08809227087645695,"envelope_correlation":0.9957245832544522,"rms_gain_error_db":-0.3301018082559212,"reference_mean_power":0.0005824704905358113,"output_mean_power":0.0005398383667103886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.040463159930298e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002294596306039209,"output_power":0.0002269800661112738,"reference_active":true,"delta_db":-0.0471857192835095},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.014148036198554e-05,"output_power":5.499865576038595e-05,"reference_active":true,"delta_db":-0.3882203901172827},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3285241811094983e-06,"output_power":1.161145553657595e-06,"reference_active":true,"delta_db":-0.5848280013785099}],"active_band_count":3,"active_band_mae_db":0.340078036926434}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.476253032684326,"predicted_mos":2.476253032684326,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b009b9a7c85f68b358d4c8270b0e5d887477afd66c35d26e6ae02619cca065c9","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.249025533999884},"started_at":"2026-10-13T04:42:07.891083+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.123299964296166e-05,"native_input_samples":96000,"native_input_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","output_sha256":"129a02f5f83093b515c99923dc150506fde074a73453b1a055e252bc50e748aa","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5338619320000362,"synthesis_s":0.08841259499968146,"total_s":1.6222745269997176,"rtf":0.2703790878332863,"output_peak":0.6271727081135676,"output_rms":0.046580792183748734,"finished_at":"2026-10-13T04:42:09.518742+00:00","evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"129a02f5f83093b515c99923dc150506fde074a73453b1a055e252bc50e748aa","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.0557653941019485,"spectral_convergence":0.6281692878525564,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5619621789443086,"spectral_convergence":0.44792138128872205,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.709188119338274,"spectral_convergence":0.4772630483964902,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00015259277805729525,"delta_db":-0.20335466047720976,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0006755300279882874,"delta_db":1.0154852288004776,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.00032802918403161495,"delta_db":1.0765124691026404,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.00017925173466469877,"delta_db":0.5868388397121014,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003744351165178521,"delta_db":0.22605810217095124,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.00045993135920583826,"delta_db":0.3583707311360373,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5777700052332363,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6083874069438728,"pitch":null,"output_rms_dbfs":-26.635862596788797,"output_peak":0.6271727081135676,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.108971897461511},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.0557653941019485,"spectral_convergence":0.6281692878525564,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5619621789443086,"spectral_convergence":0.44792138128872205,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.709188119338274,"spectral_convergence":0.4772630483964902,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00015259277805729525,"delta_db":-0.20335466047720976,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0006755300279882874,"delta_db":1.0154852288004776,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.00032802918403161495,"delta_db":1.0765124691026404,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.00017925173466469877,"delta_db":0.5868388397121014,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003744351165178521,"delta_db":0.22605810217095124,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.00045993135920583826,"delta_db":0.3583707311360373,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5777700052332363,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6083874069438728,"pitch":null,"output_rms_dbfs":-26.635862596788797,"output_peak":0.6271727081135676,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.108971897461511},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9083288186345835,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04342976704277862,"rms_dbfs":-27.244250003732674,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.951942904471995,"mean":2.936585744222005e-05,"zero_energy":false},"output":{"rms":0.046580792183748734,"rms_dbfs":-26.635862596788797,"peak":0.6271727081135676,"peak_dbfs":-4.052256971190613,"crest_factor_db":22.583605625598185,"mean":1.871507945358566e-06,"zero_energy":false},"rms_difference_db":0.6083874069438764},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":1.31391027835532e-06,"output_rms_dbfs":-58.814342899859525,"output_zero_energy":false,"output_peak":0.014254154727881767,"output_energy_fraction":8.477738284463191e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":64,"output_estimated_voiced_frames":80,"both_estimated_voiced_frames":62,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.96875,"absolute_difference_cents_median":4.633027735871128,"absolute_difference_cents_p95":17.09036964462037,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018816516597127566,"log_rms_mae_db":2.441648059908832,"rms_envelope_nrmse":0.1846568727492661,"envelope_correlation":0.9768280358149992,"rms_gain_error_db":0.6101541377653912,"reference_mean_power":0.0018869472812584292,"output_mean_power":0.002171576736407567},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.448172186664129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039739520206635815,"output_power":0.00044599146759459853,"reference_active":true,"delta_db":0.5010393079502956},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001684948117890412,"output_power":0.00018285566940602334,"reference_active":true,"delta_db":0.3552189727539296},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9559595301379592e-05,"output_power":1.8729898611195207e-05,"reference_active":true,"delta_db":-0.1882443830051066}],"active_band_count":3,"active_band_mae_db":0.34816755456977727}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016656172387320294,"log_rms_mae_db":1.8088546353456054,"rms_envelope_nrmse":0.12379515772455514,"envelope_correlation":0.9914646007605319,"rms_gain_error_db":0.6126106358183075,"reference_mean_power":0.0018945898208991615,"output_mean_power":0.0021816057179567866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.514104383617564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003829793712831679,"output_power":0.00043138899006473614,"reference_active":true,"delta_db":0.5169367496265174},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.831616093630877e-05,"output_power":8.574551561470146e-05,"reference_active":true,"delta_db":0.3936002606902661},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3812315351816923e-06,"output_power":2.2124007573335047e-06,"reference_active":true,"delta_db":-0.31937826914943945}],"active_band_count":3,"active_band_mae_db":0.4099717598220743}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015092951870698086,"log_rms_mae_db":1.921506873265588,"rms_envelope_nrmse":0.21369687681177132,"envelope_correlation":0.9858829399122696,"rms_gain_error_db":1.202856912200774,"reference_mean_power":0.000631981580682194,"output_mean_power":0.0008336622036660174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.167139764996573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018037000728302603,"output_power":0.00023293066713307316,"reference_active":true,"delta_db":1.1106234772796315},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6944061726753776e-05,"output_power":3.2463503240898034e-05,"reference_active":true,"delta_db":0.809323193758615},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.521461134259169e-06,"output_power":2.970875562403746e-06,"reference_active":true,"delta_db":-0.738384381250846}],"active_band_count":3,"active_band_mae_db":0.8861103507630309}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013588907795973108,"log_rms_mae_db":1.2541590225162658,"rms_envelope_nrmse":0.18103282369958157,"envelope_correlation":0.9934907349198921,"rms_gain_error_db":1.2054601254608708,"reference_mean_power":0.0006345375227010602,"output_mean_power":0.0008375356871779244},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.990151794989553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001715162379513309,"output_power":0.00022083904475289607,"reference_active":true,"delta_db":1.0977061764316691},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2552077755571314e-05,"output_power":1.4900791975278592e-05,"reference_active":true,"delta_db":0.7449373084044596},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0572536267135004e-07,"output_power":2.663059413559539e-07,"reference_active":true,"delta_db":-0.59950613086218}],"active_band_count":3,"active_band_mae_db":0.8140498718994363}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":1.8732980038582792,"rms_envelope_nrmse":0.18258089208412023,"envelope_correlation":0.9758137757338029,"rms_gain_error_db":0.3611512237729794,"reference_mean_power":0.00026216016454011775,"output_mean_power":0.0002848930294188986},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5456728460340465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.502085853898133e-05,"output_power":8.230430611790756e-05,"reference_active":true,"delta_db":0.40240527954246086},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.151594827944161e-05,"output_power":1.860959408272518e-05,"reference_active":true,"delta_db":-0.6302359135283365},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51387829077428e-06,"output_power":6.491289964793465e-06,"reference_active":true,"delta_db":-1.1779642805234194}],"active_band_count":3,"active_band_mae_db":0.7368684911980723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":1.1191483610662232,"rms_envelope_nrmse":0.1114405056700466,"envelope_correlation":0.9910876503384299,"rms_gain_error_db":0.36336596234926255,"reference_mean_power":0.00026328876508602793,"output_mean_power":0.00028626544243401516},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3635149275752304e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.257955032157828e-05,"output_power":7.751262126952125e-05,"reference_active":true,"delta_db":0.28558150684303224},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405048683009273e-06,"output_power":8.017086906071826e-06,"reference_active":true,"delta_db":-0.6934445649372595},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.677553554297771e-07,"output_power":5.767238831558756e-07,"reference_active":true,"delta_db":-1.2425491783404616}],"active_band_count":3,"active_band_mae_db":0.7405250833735844}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001568272998494035,"log_rms_mae_db":2.1644631362512547,"rms_envelope_nrmse":0.23198447133549158,"envelope_correlation":0.969163593505445,"rms_gain_error_db":0.34935421191756244,"reference_mean_power":0.0005797089728212456,"output_mean_power":0.000628268728564518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.292445291957658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020605682430557343,"output_power":0.00020856075209804633,"reference_active":true,"delta_db":0.052455819800274264},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001264586344183223,"output_power":0.0001350243280177345,"reference_active":true,"delta_db":0.2846353687238649},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.269907486994092e-06,"output_power":8.573271153937097e-06,"reference_active":true,"delta_db":-0.3392884020391559}],"active_band_count":3,"active_band_mae_db":0.22545986352109834}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012704534221320317,"log_rms_mae_db":1.2883738437049645,"rms_envelope_nrmse":0.1345843595822226,"envelope_correlation":0.9897633633721605,"rms_gain_error_db":0.3502724730112825,"reference_mean_power":0.0005824704905358113,"output_mean_power":0.0006313950533113656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.040463159930298e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002294596306039209,"output_power":0.00023406981514207837,"reference_active":true,"delta_db":0.08639122048959141},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.014148036198554e-05,"output_power":6.618476332434125e-05,"reference_active":true,"delta_db":0.41583906261630227},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3285241811094983e-06,"output_power":1.162229600928425e-06,"reference_active":true,"delta_db":-0.5807753129344716}],"active_band_count":3,"active_band_mae_db":0.3610018653467884}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"facodec","status":"measured","mos":{"status":"measured","value":3.136688232421875,"predicted_mos":3.136688232421875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d6ca7c178df1bc74658e33439245887bd71b9b98332aecd69874d56b7dface7c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1836339969995606},"started_at":"2026-10-13T04:52:02.907223+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.761500026186695e-05,"native_input_samples":96000,"native_input_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","output_sha256":"fee4cb519e514133d0db9ec0fa42ef5657886a5eb8263ae71f4c6cbad2bf96bd","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7649524079997718,"synthesis_s":5.469454298000528,"total_s":7.2344067060003,"rtf":1.20573445100005,"output_peak":0.3799351751804352,"output_rms":0.040073278871170855,"finished_at":"2026-10-13T04:52:10.149237+00:00","evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"fee4cb519e514133d0db9ec0fa42ef5657886a5eb8263ae71f4c6cbad2bf96bd","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5713591787120915,"spectral_convergence":0.44982196668921637,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.686172045484009,"spectral_convergence":0.46680629814083424,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.792359879107411,"spectral_convergence":0.4797386467173954,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00014918932522773467,"delta_db":-0.3013169609893098,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0004555607647306902,"delta_db":-0.6954982242146024,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.00023493238519774987,"delta_db":-0.3731834877005164,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.00014082085066976146,"delta_db":-0.46112519693599885,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003054877413902872,"delta_db":-0.6577697464411048,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.0003198766122704047,"delta_db":-1.2187346082275001,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6179380374181721,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6986524226830698,"pitch":null,"output_rms_dbfs":-27.942902426415742,"output_peak":0.3799351751804352,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.683297034434504},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5713591787120915,"spectral_convergence":0.44982196668921637,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.686172045484009,"spectral_convergence":0.46680629814083424,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.792359879107411,"spectral_convergence":0.4797386467173954,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00014918932522773467,"delta_db":-0.3013169609893098,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0004555607647306902,"delta_db":-0.6954982242146024,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.00023493238519774987,"delta_db":-0.3731834877005164,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.00014082085066976146,"delta_db":-0.46112519693599885,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003054877413902872,"delta_db":-0.6577697464411048,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.0003198766122704047,"delta_db":-1.2187346082275001,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6179380374181721,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6986524226830698,"pitch":null,"output_rms_dbfs":-27.942902426415742,"output_peak":0.3799351751804352,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.683297034434504},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8614300160082305,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04342976704277862,"rms_dbfs":-27.244250003732674,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.951942904471995,"mean":2.936585744222005e-05,"zero_energy":false},"output":{"rms":0.040073278871170855,"rms_dbfs":-27.942902426415742,"peak":0.3799351751804352,"peak_dbfs":-8.405809934139526,"crest_factor_db":19.537092492276216,"mean":-3.386012089324216e-05,"zero_energy":false},"rms_difference_db":-0.6986524226830682},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":1.651973216501503e-06,"output_rms_dbfs":-57.81996998189909,"output_zero_energy":false,"output_peak":0.008664076216518879,"output_energy_fraction":0.00014401949380047678,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":64,"output_estimated_voiced_frames":74,"both_estimated_voiced_frames":60,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9375,"absolute_difference_cents_median":5.128710883046148,"absolute_difference_cents_p95":18.768910590286637,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018816516597127566,"log_rms_mae_db":1.4220266424558792,"rms_envelope_nrmse":0.17305620647263611,"envelope_correlation":0.9771104866220326,"rms_gain_error_db":-0.6987020551786538,"reference_mean_power":0.0018869472812584292,"output_mean_power":0.001606532668337167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.448172186664129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039739520206635815,"output_power":0.0003366927598222789,"reference_active":true,"delta_db":-0.7198884186824055},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001684948117890412,"output_power":0.00012179842996036233,"reference_active":true,"delta_db":-1.409448427563602},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9559595301379592e-05,"output_power":2.373901088621964e-05,"reference_active":true,"delta_db":0.8410275484919459}],"active_band_count":3,"active_band_mae_db":0.9901214649126512}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016656172387320294,"log_rms_mae_db":1.1028660167903435,"rms_envelope_nrmse":0.15736815553956598,"envelope_correlation":0.9805634890499798,"rms_gain_error_db":-0.699292666955184,"reference_mean_power":0.0018945898208991615,"output_mean_power":0.0016128201210983304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.514104383617564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003829793712831679,"output_power":0.0003194413512975186,"reference_active":true,"delta_db":-0.787842475177012},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.831616093630877e-05,"output_power":5.707185896979209e-05,"reference_active":true,"delta_db":-1.3742937135997018},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3812315351816923e-06,"output_power":2.2717182356395984e-06,"reference_active":true,"delta_db":-0.20447161143823736}],"active_band_count":3,"active_band_mae_db":0.7888692667383169}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015092951870698086,"log_rms_mae_db":1.9754855266632076,"rms_envelope_nrmse":0.16801896540338782,"envelope_correlation":0.9818820398861693,"rms_gain_error_db":-0.5452865048629564,"reference_mean_power":0.000631981580682194,"output_mean_power":0.000557411302198491},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.167139764996573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018037000728302603,"output_power":0.00016808896680130757,"reference_active":true,"delta_db":-0.30625115114247686},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6944061726753776e-05,"output_power":2.4573396665231965e-05,"reference_active":true,"delta_db":-0.3999787363938896},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.521461134259169e-06,"output_power":4.322782814686238e-06,"reference_active":true,"delta_db":0.8904051672865185}],"active_band_count":3,"active_band_mae_db":0.5322116849409616}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013588907795973108,"log_rms_mae_db":1.6516248878224837,"rms_envelope_nrmse":0.15437402222525487,"envelope_correlation":0.9844708614395565,"rms_gain_error_db":-0.5448800722547207,"reference_mean_power":0.0006345375227010602,"output_mean_power":0.0005597180359273698},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.990151794989553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001715162379513309,"output_power":0.00015929431832959544,"reference_active":true,"delta_db":-0.3210495650378845},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2552077755571314e-05,"output_power":1.1277109193761284e-05,"reference_active":true,"delta_db":-0.46517835252564443},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0572536267135004e-07,"output_power":3.6508772352514385e-07,"reference_active":true,"delta_db":0.7706576070934369}],"active_band_count":3,"active_band_mae_db":0.518961841552322}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":1.8053291533625684,"rms_envelope_nrmse":0.197197404588548,"envelope_correlation":0.9723590990131021,"rms_gain_error_db":-0.7748610986854247,"reference_mean_power":0.00026216016454011775,"output_mean_power":0.0002193211887153272},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5456728460340465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.502085853898133e-05,"output_power":6.584736757904746e-05,"reference_active":true,"delta_db":-0.5664361232514741},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.151594827944161e-05,"output_power":1.8446445366816942e-05,"reference_active":true,"delta_db":-0.6684780170708371},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51387829077428e-06,"output_power":7.092291907751138e-06,"reference_active":true,"delta_db":-0.7934083529069917}],"active_band_count":3,"active_band_mae_db":0.676107497743101}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":1.4354401352986441,"rms_envelope_nrmse":0.16822289624315073,"envelope_correlation":0.9796197554753214,"rms_gain_error_db":-0.7758580397182981,"reference_mean_power":0.00026328876508602793,"output_mean_power":0.00022021481027326915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3635149275752304e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.257955032157828e-05,"output_power":6.282835049147733e-05,"reference_active":true,"delta_db":-0.6265861532233461},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405048683009273e-06,"output_power":7.91158514351948e-06,"reference_active":true,"delta_db":-0.7509754164674755},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.677553554297771e-07,"output_power":5.446892249446697e-07,"reference_active":true,"delta_db":-1.4907407038561087}],"active_band_count":3,"active_band_mae_db":0.9561007578489767}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001568272998494035,"log_rms_mae_db":1.9123546412319132,"rms_envelope_nrmse":0.21216912121294135,"envelope_correlation":0.9753247374795775,"rms_gain_error_db":-0.8563403100106222,"reference_mean_power":0.0005797089728212456,"output_mean_power":0.0004759660667630574},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.292445291957658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020605682430557343,"output_power":0.00015474546229881983,"reference_active":true,"delta_db":-1.2436907952606422},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001264586344183223,"output_power":9.259418659835331e-05,"reference_active":true,"delta_db":-1.353647667692688},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.269907486994092e-06,"output_power":1.3526836356957293e-05,"reference_active":true,"delta_db":1.6412083617319229}],"active_band_count":3,"active_band_mae_db":1.412848941561751}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012704534221320317,"log_rms_mae_db":1.5960116523381667,"rms_envelope_nrmse":0.17955427075042243,"envelope_correlation":0.9831380420166099,"rms_gain_error_db":-0.8558828724852701,"reference_mean_power":0.0005824704905358113,"output_mean_power":0.00047828376633950057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.040463159930298e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002294596306039209,"output_power":0.00017257925346244465,"reference_active":true,"delta_db":-1.2371770409871083},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.014148036198554e-05,"output_power":4.477314083297834e-05,"reference_active":true,"delta_db":-1.2815655272398683},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3285241811094983e-06,"output_power":1.5095755873553271e-06,"reference_active":true,"delta_db":0.5548540024261052}],"active_band_count":3,"active_band_mae_db":1.024532190217694}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.069185495376587,"predicted_mos":3.069185495376587,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"f0be8faa47ba86bbacaf1b5ffe9da2d651dabc414b8c6c35a03246d0005974a2","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.766384549999202},"started_at":"2026-10-13T07:13:39.127728+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0064462690006621415,"native_input_samples":264600,"native_input_sha256":"0b96f8855c84463e7b4765cedb8fa67e27fdbb1c79b303e55db5805627b092e1","output_sha256":"2afac6cb354b6e6d7e6dc342d727c1a97d57677413bf052ee6d10e3a65336f59","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15083714600041276,"synthesis_s":3.9853372599991417,"total_s":4.136174405999554,"rtf":0.6893624009999257,"output_peak":0.5087265372276306,"output_rms":0.03653791792569879,"finished_at":"2026-10-13T07:13:43.290544+00:00","evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"2afac6cb354b6e6d7e6dc342d727c1a97d57677413bf052ee6d10e3a65336f59","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.739716896021097,"output_peak":0.5048575730986574},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.135492983554556,"spectral_convergence":0.590667976537476,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.798898128804014,"spectral_convergence":0.5354822298106735,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.629836888097909,"spectral_convergence":0.5653269159972926,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":8.957495940053355e-05,"delta_db":-2.516828257774133,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.000405485492897732,"delta_db":-1.2012081471277187,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.00024553428021787825,"delta_db":-0.18149102114060717,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.0001232191121532133,"delta_db":-1.0410140797747105,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.000224258798337829,"delta_db":-2.0002127149568594,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.00024909503649899893,"delta_db":-2.30490874880171,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.540943828262623,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.4938912885079754,"pitch":null,"output_rms_dbfs":-28.73814129224065,"output_peak":0.5048575730986574,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.854742666818826},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6612508914431461,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04342976704277862,"rms_dbfs":-27.244250003732674,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.951942904471995,"mean":2.936585744222005e-05,"zero_energy":false},"output":{"rms":0.03656730342131048,"rms_dbfs":-28.73814129224065,"peak":0.5048575730986574,"peak_dbfs":-5.936622494719398,"crest_factor_db":22.80151879752125,"mean":-0.00018658674470453375,"zero_energy":false},"rms_difference_db":-1.4938912885079745},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":1.1854982526343221e-06,"output_rms_dbfs":-59.260990818138666,"output_zero_energy":false,"output_peak":0.013755309725787023,"output_energy_fraction":0.00012412037615962837,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":64,"output_estimated_voiced_frames":59,"both_estimated_voiced_frames":48,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.75,"absolute_difference_cents_median":8.731128444517019,"absolute_difference_cents_p95":31.892863209071084,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018816516597127566,"log_rms_mae_db":2.3037006231416415,"rms_envelope_nrmse":0.2036410407855592,"envelope_correlation":0.982682405471754,"rms_gain_error_db":-1.4943332325404364,"reference_mean_power":0.0018869472812584292,"output_mean_power":0.0013376005648239964},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.448172186664129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039739520206635815,"output_power":0.00027004727201312956,"reference_active":true,"delta_db":-1.6778282492411976},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001684948117890412,"output_power":9.405804094393386e-05,"reference_active":true,"delta_db":-2.531906038784037},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9559595301379592e-05,"output_power":1.7505276530919087e-05,"reference_active":true,"delta_db":-0.48190889049692703}],"active_band_count":3,"active_band_mae_db":1.563881059507387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016656172387320294,"log_rms_mae_db":1.964195402730596,"rms_envelope_nrmse":0.18221405008264396,"envelope_correlation":0.9905699896109584,"rms_gain_error_db":-1.4949901314277554,"reference_mean_power":0.0018945898208991615,"output_mean_power":0.0013428150079552049},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.514104383617564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003829793712831679,"output_power":0.0002568863643959127,"reference_active":true,"delta_db":-1.7343432945713548},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.831616093630877e-05,"output_power":4.428271540249139e-05,"reference_active":true,"delta_db":-2.4761714634366037},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3812315351816923e-06,"output_power":1.5982182176808005e-06,"reference_active":true,"delta_db":-1.731655486062598}],"active_band_count":3,"active_band_mae_db":1.9807234146901855}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015092951870698086,"log_rms_mae_db":3.4629595524440693,"rms_envelope_nrmse":0.21009425307192947,"envelope_correlation":0.9719577447212314,"rms_gain_error_db":-0.7984611576673749,"reference_mean_power":0.000631981580682194,"output_mean_power":0.0005258456735174824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.167139764996573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018037000728302603,"output_power":0.00014960676892818668,"reference_active":true,"delta_db":-0.8121307927615846},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6944061726753776e-05,"output_power":2.3477742890971757e-05,"reference_active":true,"delta_db":-0.5980672248305514},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.521461134259169e-06,"output_power":4.676360524200347e-06,"reference_active":true,"delta_db":1.231850862070751}],"active_band_count":3,"active_band_mae_db":0.880682959887629}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013588907795973108,"log_rms_mae_db":2.662807225463753,"rms_envelope_nrmse":0.17953204086568172,"envelope_correlation":0.9800283416044749,"rms_gain_error_db":-0.7955782932634443,"reference_mean_power":0.0006345375227010602,"output_mean_power":0.0005283229534591966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.990151794989553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001715162379513309,"output_power":0.0001420349967128919,"reference_active":true,"delta_db":-0.8190987671601375},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2552077755571314e-05,"output_power":1.0932976114673897e-05,"reference_active":true,"delta_db":-0.5997722158356985},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0572536267135004e-07,"output_power":3.4593631318769e-07,"reference_active":true,"delta_db":0.5366468375854717}],"active_band_count":3,"active_band_mae_db":0.6518392735271026}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":3.197306961689701,"rms_envelope_nrmse":0.2810513380663875,"envelope_correlation":0.9481708007481047,"rms_gain_error_db":-1.4261077677226304,"reference_mean_power":0.00026216016454011775,"output_mean_power":0.00018877997497341423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5456728460340465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.502085853898133e-05,"output_power":5.825299641287355e-05,"reference_active":true,"delta_db":-1.0986376043407233},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.151594827944161e-05,"output_power":1.4753614223919865e-05,"reference_active":true,"delta_db":-1.6386206826280534},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51387829077428e-06,"output_power":8.091581830783703e-06,"reference_active":true,"delta_db":-0.22094006935509433}],"active_band_count":3,"active_band_mae_db":0.9860661187746236}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":2.594120801650437,"rms_envelope_nrmse":0.23190429193795567,"envelope_correlation":0.9683865019558636,"rms_gain_error_db":-1.4237223842144588,"reference_mean_power":0.00026328876508602793,"output_mean_power":0.000189696836811883},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3635149275752304e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.257955032157828e-05,"output_power":5.6126191912851426e-05,"reference_active":true,"delta_db":-1.1164869621191678},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405048683009273e-06,"output_power":6.8278791597443495e-06,"reference_active":true,"delta_db":-1.3907522157651833},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.677553554297771e-07,"output_power":6.053425862917373e-07,"reference_active":true,"delta_db":-1.0322162659949117}],"active_band_count":3,"active_band_mae_db":1.1798184812930874}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001568272998494035,"log_rms_mae_db":3.001941040207373,"rms_envelope_nrmse":0.27123797369075087,"envelope_correlation":0.9822009986297494,"rms_gain_error_db":-2.2298851933686388,"reference_mean_power":0.0005797089728212456,"output_mean_power":0.00034691374171098844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.292445291957658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020605682430557343,"output_power":0.00010529898952065257,"reference_active":true,"delta_db":-2.9156279871632576},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001264586344183223,"output_power":6.328277704084799e-05,"reference_active":true,"delta_db":-3.006629586399777},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.269907486994092e-06,"output_power":6.782946035899864e-06,"reference_active":true,"delta_db":-1.3565703804154061}],"active_band_count":3,"active_band_mae_db":2.42627598465948}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012704534221320317,"log_rms_mae_db":2.5840472648179484,"rms_envelope_nrmse":0.24506731926709321,"envelope_correlation":0.9925773564510494,"rms_gain_error_db":-2.2291098708444683,"reference_mean_power":0.0005824704905358113,"output_mean_power":0.00034862854283566625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.040463159930298e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002294596306039209,"output_power":0.00011627077075053845,"reference_active":true,"delta_db":-2.9523573871746045},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.014148036198554e-05,"output_power":3.064607308281418e-05,"reference_active":true,"delta_db":-2.9279928098093677},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3285241811094983e-06,"output_power":9.131577946030824e-07,"reference_active":true,"delta_db":-1.6282363307078576}],"active_band_count":3,"active_band_mae_db":2.5028621758972767}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"golf","status":"measured","mos":{"status":"measured","value":2.768001079559326,"predicted_mos":2.768001079559326,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"3930c328b05e3967d90dd7473840399f1da6102b7c04bfec080389c6f44c9a94","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.353790016000858},"started_at":"2026-10-13T05:02:32.062218+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0025710380004966282,"native_input_samples":144000,"native_input_sha256":"2ffc5080a1cdc4c2a2a803e1780469cf578bb2e2b6b9211311ec993768d988d7","output_sha256":"8db27f61c2169feb341707899c4cdd53ab606720a42fb7108710a9045ac3f3b3","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5827900419999423,"synthesis_s":0.11339587500060588,"total_s":0.6961859170005482,"rtf":0.11603098616675804,"output_peak":0.42509716749191284,"output_rms":0.03897112904444204,"finished_at":"2026-10-13T05:02:32.768597+00:00","analysis_stages":{"external_f0_s":0.10008763200039539},"evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"8db27f61c2169feb341707899c4cdd53ab606720a42fb7108710a9045ac3f3b3","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.182288064392587,"output_peak":0.42048293146040927},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.01166507559031,"spectral_convergence":0.642959201871026,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.4974619288506235,"spectral_convergence":0.43176272432641927,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.656645108014187,"spectral_convergence":0.46268218749982554,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015980019426778817,"output_power":0.00016736785117456765,"delta_db":0.2009473733944147,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346520973716384,"output_power":0.0005314560639135482,"delta_db":-0.026039081274296102,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025620507029200655,"output_power":0.00019084376157210212,"delta_db":-1.2790975216070426,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015672189941518645,"output_power":0.00011144854257508124,"delta_db":-1.4805529298099391,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003558231293077451,"output_power":0.00023692402268231104,"delta_db":-1.7662507696387273,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004240269355018724,"output_power":0.00028170640397048993,"delta_db":-1.7759672537704874,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.088142488249151,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9405348298618819,"pitch":null,"output_rms_dbfs":-28.182288064392587,"output_peak":0.42048293146040927,"reference_rms_dbfs":-27.24175323453071,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.055257370818374},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8498115189383229,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04344225277487618,"rms_dbfs":-27.24175323453071,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.94944613527003,"mean":2.5957773046222873e-05,"zero_energy":false},"output":{"rms":0.038983928045902456,"rms_dbfs":-28.182288064392587,"peak":0.42048293146040927,"peak_dbfs":-7.525032573921076,"crest_factor_db":20.657255490471513,"mean":0.00024172635706093436,"zero_energy":false},"rms_difference_db":-0.9405348298618783},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":1.9495332148519632e-06,"output_rms_dbfs":-57.10069361192055,"output_zero_energy":false,"output_peak":0.0062353999961158885,"output_energy_fraction":0.00017989202431861908,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":63,"output_estimated_voiced_frames":87,"both_estimated_voiced_frames":52,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.8253968253968254,"absolute_difference_cents_median":7.846152977206316,"absolute_difference_cents_p95":32.76411768682987,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00018816516597127566,"log_rms_mae_db":2.3208804546733184,"rms_envelope_nrmse":0.1910931497102384,"envelope_correlation":0.9745626592581661,"rms_gain_error_db":-0.9402132186061053,"reference_mean_power":0.00188847672309588,"output_mean_power":0.0015208637702420309},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.444403568725536e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039742939521633235,"output_power":0.0002963357908401331,"reference_active":true,"delta_db":-1.2747587798309863},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016824970005110536,"output_power":0.0001289288155349937,"reference_active":true,"delta_db":-1.1560430582124852},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9373738111086895e-05,"output_power":1.7089343806830433e-05,"reference_active":true,"delta_db":-0.5448803767686035}],"active_band_count":3,"active_band_mae_db":0.9918940716040251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016656172387320294,"log_rms_mae_db":1.6459731601122871,"rms_envelope_nrmse":0.15370633113333862,"envelope_correlation":0.9860046184230018,"rms_gain_error_db":-0.9409257816472449,"reference_mean_power":0.0018968628010681576,"output_mean_power":0.0015273667837819433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.510981904352539e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038297198441565735,"output_power":0.0002829042486071708,"reference_active":true,"delta_db":-1.315275354814146},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.807042623418066e-05,"output_power":6.260318213562845e-05,"reference_active":true,"delta_db":-0.9589014138004038},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3699663458970633e-06,"output_power":1.5151102424996848e-06,"reference_active":true,"delta_db":-1.9429794482887253}],"active_band_count":3,"active_band_mae_db":1.4057187389677583}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015092951870698086,"log_rms_mae_db":2.142490372220203,"rms_envelope_nrmse":0.17627373217964765,"envelope_correlation":0.9796237726794077,"rms_gain_error_db":-0.4761749492284256,"reference_mean_power":0.0006324384475177303,"output_mean_power":0.000566762058486013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.1687562468727077e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001801985665675859,"output_power":0.00014881115616409123,"reference_active":true,"delta_db":-0.83115841078105},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6984110688336973e-05,"output_power":2.524034768620764e-05,"reference_active":true,"delta_db":-0.2901277662597065},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4821014295052277e-06,"output_power":3.562124638823576e-06,"reference_active":true,"delta_db":0.09867693915773904}],"active_band_count":3,"active_band_mae_db":0.4066543720661652}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013588907795973108,"log_rms_mae_db":1.421658907360541,"rms_envelope_nrmse":0.13269291862120675,"envelope_correlation":0.9888709537299029,"rms_gain_error_db":-0.4759623974054373,"reference_mean_power":0.0006352960699684,"output_mean_power":0.0005693507918434322},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.9919687031967954e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017136868509016892,"output_power":0.00014239003114014737,"reference_active":true,"delta_db":-0.8045187940675179},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.252326429095091e-05,"output_power":1.172190882289646e-05,"reference_active":true,"delta_db":-0.2871920706396685},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.041154771482792e-07,"output_power":2.750758289300621e-07,"reference_active":true,"delta_db":-0.43586092606557875}],"active_band_count":3,"active_band_mae_db":0.5091905969242551}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":2.2206439863164835,"rms_envelope_nrmse":0.2657514328365263,"envelope_correlation":0.965669814773327,"rms_gain_error_db":-1.8244112444732892,"reference_mean_power":0.0002623872723435144,"output_mean_power":0.0001723858602239654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.544942972499959e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.505332810437714e-05,"output_power":4.374155705396601e-05,"reference_active":true,"delta_db":-2.344757165951987},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1417344368137032e-05,"output_power":1.0630045089864766e-05,"reference_active":true,"delta_db":-3.042305130375849},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.497975214801374e-06,"output_power":5.5840710895721675e-06,"reference_active":true,"delta_db":-1.8236452153469527}],"active_band_count":3,"active_band_mae_db":2.4035691705582627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":1.729285283879764,"rms_envelope_nrmse":0.22544906148228386,"envelope_correlation":0.9836591767605789,"rms_gain_error_db":-1.8244037605425742,"reference_mean_power":0.0002636575295638564,"output_mean_power":0.00017322070523903042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.3628007617949963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.255561983685036e-05,"output_power":4.164200412485016e-05,"reference_active":true,"delta_db":-2.4113943376114166},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.34183378763513e-06,"output_power":4.905003492737002e-06,"reference_active":true,"delta_db":-2.7979281497189765},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.645609748403516e-07,"output_power":3.4098459934248727e-07,"reference_active":true,"delta_db":-3.5067736232779922}],"active_band_count":3,"active_band_mae_db":2.905365370202795}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001568272998494035,"log_rms_mae_db":2.6396346531215498,"rms_envelope_nrmse":0.24285185702010673,"envelope_correlation":0.9795112615955972,"rms_gain_error_db":-1.7007579772795132,"reference_mean_power":0.0005804856812112993,"output_mean_power":0.00039238799679017287},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.290066536234327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020623847156725863,"output_power":0.0001446415722651853,"reference_active":true,"delta_db":-1.5407654749417434},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012609574267189502,"output_power":9.649448859593966e-05,"reference_active":true,"delta_db":-1.16197915105547},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.235460765280097e-06,"output_power":8.057591588176042e-06,"reference_active":true,"delta_db":-0.5925331694657429}],"active_band_count":3,"active_band_mae_db":1.0984259318209852}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012704534221320317,"log_rms_mae_db":2.331183843220797,"rms_envelope_nrmse":0.20499028928400737,"envelope_correlation":0.9917593227361772,"rms_gain_error_db":-1.7007106338340627,"reference_mean_power":0.0005833933253300162,"output_mean_power":0.0003943577613749621},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.0384089492870135e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022962978067164174,"output_power":0.00016367606979393788,"reference_active":true,"delta_db":-1.4704302283127026},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9751340105789795e-05,"output_power":4.7990630943582855e-05,"reference_active":true,"delta_db":-0.9519119036905395},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3258099911873628e-06,"output_power":7.798965888946376e-07,"reference_active":true,"delta_db":-2.304442666670514}],"active_band_count":3,"active_band_mae_db":1.575594932891252}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.7142152786254883,"predicted_mos":2.7142152786254883,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"d36113ac33817168111d3970864b5bcc312d2434f5774cc7281ec03667663500","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":30.530634775997896},"started_at":"2026-10-13T07:36:59.302000+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031078239990165457,"native_input_samples":144000,"native_input_sha256":"2ffc5080a1cdc4c2a2a803e1780469cf578bb2e2b6b9211311ec993768d988d7","output_sha256":"ddd6de25d9c6c749e26a1f87431abca58419266b1802349eb588ad77d7adcc96","output_samples":144120,"length_delta_samples":120,"analysis_s":2.36571415599974,"synthesis_s":3.6482512340007816,"total_s":6.013965390000521,"rtf":1.002327565000087,"output_peak":0.43220317363739014,"output_rms":0.04183051692081355,"finished_at":"2026-10-13T07:37:05.331668+00:00","analysis_stages":{"external_f0_s":1.5693443339987425},"evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"ddd6de25d9c6c749e26a1f87431abca58419266b1802349eb588ad77d7adcc96","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.567441178722373,"output_peak":0.43280906516074963},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.937618769427872,"spectral_convergence":0.668330801816626,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.511781624005558,"spectral_convergence":0.46050981240286226,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.426825830989274,"spectral_convergence":0.47264279257539726,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00023144151464657664,"delta_db":1.6057181707390709,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0005121335647855962,"delta_db":-0.18712895685628037,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.0002456085525158044,"delta_db":-0.18017751121055497,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.00014272660429727845,"delta_db":-0.4027454983827259,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003143017530961548,"delta_db":-0.5342395879697913,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.0003024435939748213,"delta_db":-1.4621155849682699,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7286875516877823,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3287072073399092,"pitch":null,"output_rms_dbfs":-27.57295721107258,"output_peak":0.43280906516074963,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.958742074807568},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8677164787030015,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04342976704277862,"rms_dbfs":-27.244250003732674,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.951942904471995,"mean":2.936585744222005e-05,"zero_energy":false},"output":{"rms":0.041816929386508425,"rms_dbfs":-27.57295721107258,"peak":0.43280906516074963,"peak_dbfs":-7.274073030670162,"crest_factor_db":20.29888418040242,"mean":3.8966074956610274e-05,"zero_energy":false},"rms_difference_db":-0.32870720733990666},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":1.6117781024924552e-06,"output_rms_dbfs":-57.926947488190436,"output_zero_energy":false,"output_peak":0.05300409817236345,"output_energy_fraction":0.0001290413827067149,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":64,"output_estimated_voiced_frames":82,"both_estimated_voiced_frames":58,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.90625,"absolute_difference_cents_median":7.314320147644894,"absolute_difference_cents_p95":22.124961311470752,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018816516597127566,"log_rms_mae_db":3.1943526148860784,"rms_envelope_nrmse":0.20723985412885712,"envelope_correlation":0.9614528267873854,"rms_gain_error_db":-0.32842821171061937,"reference_mean_power":0.0018869472812584292,"output_mean_power":0.00174951208011232},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.448172186664129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039739520206635815,"output_power":0.00036450897198699847,"reference_active":true,"delta_db":-0.3751439690392166},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001684948117890412,"output_power":0.00013817238027626473,"reference_active":true,"delta_db":-0.8616529364126274},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9559595301379592e-05,"output_power":1.5452618518199135e-05,"reference_active":true,"delta_db":-1.0235778153631327}],"active_band_count":3,"active_band_mae_db":0.7534582402716589}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016656172387320294,"log_rms_mae_db":2.5154325630290204,"rms_envelope_nrmse":0.14279976162063415,"envelope_correlation":0.980760569486072,"rms_gain_error_db":-0.33020135931952466,"reference_mean_power":0.0018945898208991615,"output_mean_power":0.001755880936305822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.514104383617564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003829793712831679,"output_power":0.0003457055211156412,"reference_active":true,"delta_db":-0.4446906632592185},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.831616093630877e-05,"output_power":6.429365231817246e-05,"reference_active":true,"delta_db":-0.8568329281863478},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3812315351816923e-06,"output_power":1.6516767615309798e-06,"reference_active":true,"delta_db":-1.5887656690751584}],"active_band_count":3,"active_band_mae_db":0.9634297535069082}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015092951870698086,"log_rms_mae_db":2.497337210096247,"rms_envelope_nrmse":0.1770671922864886,"envelope_correlation":0.9779761737311435,"rms_gain_error_db":-0.20147777671952763,"reference_mean_power":0.000631981580682194,"output_mean_power":0.0006033323928072602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.167139764996573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018037000728302603,"output_power":0.00017160434787953953,"reference_active":true,"delta_db":-0.2163603559569288},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6944061726753776e-05,"output_power":2.370890886879193e-05,"reference_active":true,"delta_db":-0.5555149741460018},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.521461134259169e-06,"output_power":3.900115230648836e-06,"reference_active":true,"delta_db":0.44354539116994907}],"active_band_count":3,"active_band_mae_db":0.4051402404242932}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013588907795973108,"log_rms_mae_db":1.7462140088867044,"rms_envelope_nrmse":0.11566986487250702,"envelope_correlation":0.9903333058626659,"rms_gain_error_db":-0.20245183895792243,"reference_mean_power":0.0006345375227010602,"output_mean_power":0.0006056366169377593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.990151794989553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001715162379513309,"output_power":0.000162960011509549,"reference_active":true,"delta_db":-0.22224195607360475},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2552077755571314e-05,"output_power":1.1031227224761472e-05,"reference_active":true,"delta_db":-0.5609179046619979},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0572536267135004e-07,"output_power":2.866216206982403e-07,"reference_active":true,"delta_db":-0.280225214714885}],"active_band_count":3,"active_band_mae_db":0.35446169181682924}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":2.5934943141621454,"rms_envelope_nrmse":0.2635074639642007,"envelope_correlation":0.9436089879521341,"rms_gain_error_db":-0.301990616486822,"reference_mean_power":0.00026216016454011775,"output_mean_power":0.0002445499843152389},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5456728460340465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.502085853898133e-05,"output_power":7.167377588897221e-05,"reference_active":true,"delta_db":-0.1982174551172472},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.151594827944161e-05,"output_power":2.0730187305718065e-05,"reference_active":true,"delta_db":-0.1615726550365362},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51387829077428e-06,"output_power":8.252693954556959e-06,"reference_active":true,"delta_db":-0.1353169757801354}],"active_band_count":3,"active_band_mae_db":0.16503569531130627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":1.6228851217658011,"rms_envelope_nrmse":0.14376260656461842,"envelope_correlation":0.982218748209013,"rms_gain_error_db":-0.30222989470405925,"reference_mean_power":0.00026328876508602793,"output_mean_power":0.000245589241645887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3635149275752304e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.257955032157828e-05,"output_power":6.985085903507942e-05,"reference_active":true,"delta_db":-0.16642521636821353},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405048683009273e-06,"output_power":9.1564755799432e-06,"reference_active":true,"delta_db":-0.11632706422047473},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.677553554297771e-07,"output_power":6.992794316169072e-07,"reference_active":true,"delta_db":-0.4057210029169}],"active_band_count":3,"active_band_mae_db":0.22949109450186275}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001568272998494035,"log_rms_mae_db":2.602399923761504,"rms_envelope_nrmse":0.27576623994602184,"envelope_correlation":0.9587134385247115,"rms_gain_error_db":-1.2634208135527245,"reference_mean_power":0.0005797089728212456,"output_mean_power":0.00043337907784466665},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.292445291957658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020605682430557343,"output_power":0.00016217297845462096,"reference_active":true,"delta_db":-1.0400850931047676},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001264586344183223,"output_power":0.00010116411200878237,"reference_active":true,"delta_db":-0.9692201395051742},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.269907486994092e-06,"output_power":4.365660440839141e-06,"reference_active":true,"delta_db":-3.2702544648687573}],"active_band_count":3,"active_band_mae_db":1.7598532324929}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012704534221320317,"log_rms_mae_db":1.6563930780521978,"rms_envelope_nrmse":0.2077299723239457,"envelope_correlation":0.9813991616256229,"rms_gain_error_db":-1.2638549893661788,"reference_mean_power":0.0005824704905358113,"output_mean_power":0.0004354000041405307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.040463159930298e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002294596306039209,"output_power":0.00018033659757911414,"reference_active":true,"delta_db":-1.0462241854187948},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.014148036198554e-05,"output_power":4.85867610017226e-05,"reference_active":true,"delta_db":-0.9265616577900306},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3285241811094983e-06,"output_power":6.977621918501606e-07,"reference_active":true,"delta_db":-2.7966203005965964}],"active_band_count":3,"active_band_mae_db":1.5898020479351407}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.3918955326080322,"predicted_mos":2.3918955326080322,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"b6a4b68ddd9880ff9e15028ff32dc7eacd83b6da0185c34d3b18ab425f48741a","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8245382710010745},"started_at":"2026-10-13T05:10:40.847819+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028670440005953424,"native_input_samples":144000,"native_input_sha256":"2ffc5080a1cdc4c2a2a803e1780469cf578bb2e2b6b9211311ec993768d988d7","output_sha256":"69287f5aca6a16cd541a894132f136cb99e52f0235e499baa45fae9ae06b84b7","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9742767339994316,"synthesis_s":0.11774389800120844,"total_s":1.09202063200064,"rtf":0.18200343866677335,"output_peak":0.454723596572876,"output_rms":0.04158463544972572,"finished_at":"2026-10-13T05:10:41.953720+00:00","analysis_stages":{"external_f0_s":0.1915762110002106},"evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"69287f5aca6a16cd541a894132f136cb99e52f0235e499baa45fae9ae06b84b7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.617494018447417,"output_peak":0.45282010241685106},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.2254749650967,"spectral_convergence":0.35746676212619566,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.610093611233047,"spectral_convergence":0.3330863210260246,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.999690563456943,"spectral_convergence":0.37252548354562753,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00015596876931987254,"delta_db":-0.10831800431157555,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0005114991051342594,"delta_db":-0.1925125748072816,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.00024463449090534586,"delta_db":-0.19743548742485292,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.00013516654108449065,"delta_db":-0.6391028283920266,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.0003044079363810345,"delta_db":-0.6731479123843481,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.0003819801742759685,"delta_db":-0.44815126163443086,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.37644467815908594,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3661181785562221,"pitch":null,"output_rms_dbfs":-27.610368182288894,"output_peak":0.45282010241685106,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.945086379928896},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9257601458835205,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04342976704277862,"rms_dbfs":-27.244250003732674,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.951942904471995,"mean":2.936585744222005e-05,"zero_energy":false},"output":{"rms":0.04163720712416926,"rms_dbfs":-27.610368182288894,"peak":0.45282010241685106,"peak_dbfs":-6.881486027962783,"crest_factor_db":20.72888215432611,"mean":4.892581389006785e-05,"zero_energy":false},"rms_difference_db":-0.36611817855622064},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":1.782807338996038e-06,"output_rms_dbfs":-57.488955867919856,"output_zero_energy":false,"output_peak":0.018169109087299427,"output_energy_fraction":0.00014396909250066994,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":64,"output_estimated_voiced_frames":72,"both_estimated_voiced_frames":61,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.953125,"absolute_difference_cents_median":1.6969518210351375,"absolute_difference_cents_p95":6.357959745055504,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018816516597127566,"log_rms_mae_db":1.1888758883351507,"rms_envelope_nrmse":0.08845657443666582,"envelope_correlation":0.9943288320009515,"rms_gain_error_db":-0.36527089464097157,"reference_mean_power":0.0018869472812584292,"output_mean_power":0.0017347331481101802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.448172186664129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039739520206635815,"output_power":0.000370009125733663,"reference_active":true,"delta_db":-0.3101018393296977},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001684948117890412,"output_power":0.00014035887529985434,"reference_active":true,"delta_db":-0.793466532852889},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9559595301379592e-05,"output_power":1.2595573433262717e-05,"reference_active":true,"delta_db":-1.9114192053389787}],"active_band_count":3,"active_band_mae_db":1.004995859173855}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016656172387320294,"log_rms_mae_db":0.7514515198034648,"rms_envelope_nrmse":0.058453956152388895,"envelope_correlation":0.998257336834535,"rms_gain_error_db":-0.3646863055252153,"reference_mean_power":0.0018945898208991615,"output_mean_power":0.0017419936560850752},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.514104383617564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003829793712831679,"output_power":0.0003545511965922748,"reference_active":true,"delta_db":-0.3349642642107784},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.831616093630877e-05,"output_power":6.431756477962665e-05,"reference_active":true,"delta_db":-0.8552179757098394},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3812315351816923e-06,"output_power":1.5699289127291106e-06,"reference_active":true,"delta_db":-1.809216376225605}],"active_band_count":3,"active_band_mae_db":0.9997995387154076}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015092951870698086,"log_rms_mae_db":1.186888231366392,"rms_envelope_nrmse":0.07539140371943552,"envelope_correlation":0.9963215212920831,"rms_gain_error_db":-0.20418848489990513,"reference_mean_power":0.000631981580682194,"output_mean_power":0.0006029559321727697},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.167139764996573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018037000728302603,"output_power":0.00017435508442856595,"reference_active":true,"delta_db":-0.14729706308299229},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6944061726753776e-05,"output_power":2.3643721045549352e-05,"reference_active":true,"delta_db":-0.5674723796194319},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.521461134259169e-06,"output_power":2.8806876120249237e-06,"reference_active":true,"delta_db":-0.8722673451986245}],"active_band_count":3,"active_band_mae_db":0.529012262633683}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013588907795973108,"log_rms_mae_db":0.6482513815689954,"rms_envelope_nrmse":0.0480828293054948,"envelope_correlation":0.9986883731481457,"rms_gain_error_db":-0.20391674881069227,"reference_mean_power":0.0006345375227010602,"output_mean_power":0.0006054323653364125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.990151794989553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001715162379513309,"output_power":0.0001658144403914547,"reference_active":true,"delta_db":-0.14682892721814553},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2552077755571314e-05,"output_power":1.102425694660575e-05,"reference_active":true,"delta_db":-0.5636629393369316},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0572536267135004e-07,"output_power":2.358616536085622e-07,"reference_active":true,"delta_db":-1.126741298371372}],"active_band_count":3,"active_band_mae_db":0.6124110549754831}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":1.3964323325557144,"rms_envelope_nrmse":0.14107083294061656,"envelope_correlation":0.9875325383024475,"rms_gain_error_db":-0.6658836833834441,"reference_mean_power":0.00026216016454011775,"output_mean_power":0.00022489424010472587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5456728460340465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.502085853898133e-05,"output_power":6.486728373968119e-05,"reference_active":true,"delta_db":-0.6315631729837208},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.151594827944161e-05,"output_power":1.7485758908899352e-05,"reference_active":true,"delta_db":-0.900760055389494},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51387829077428e-06,"output_power":4.4118636903775135e-06,"reference_active":true,"delta_db":-2.8550535152973398}],"active_band_count":3,"active_band_mae_db":1.4624589145568514}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":0.8635605742649939,"rms_envelope_nrmse":0.095462192276099,"envelope_correlation":0.9964902913345959,"rms_gain_error_db":-0.6652867903781812,"reference_mean_power":0.00026328876508602793,"output_mean_power":0.0002258934552923061},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3635149275752304e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.257955032157828e-05,"output_power":6.182895155175726e-05,"reference_active":true,"delta_db":-0.6962239100424554},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405048683009273e-06,"output_power":7.64558982532098e-06,"reference_active":true,"delta_db":-0.8995005286877404},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.677553554297771e-07,"output_power":4.0077080247118426e-07,"reference_active":true,"delta_db":-2.8232678035212677}],"active_band_count":3,"active_band_mae_db":1.4729974140838211}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001568272998494035,"log_rms_mae_db":1.453180830617711,"rms_envelope_nrmse":0.12910191066593169,"envelope_correlation":0.9912740630802424,"rms_gain_error_db":-0.5154912772413086,"reference_mean_power":0.0005797089728212456,"output_mean_power":0.0005148265014726099},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.292445291957658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020605682430557343,"output_power":0.00019352036827616808,"reference_active":true,"delta_db":-0.2726032047471265},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001264586344183223,"output_power":0.00010534449104466658,"reference_active":true,"delta_db":-0.7933665845976988},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.269907486994092e-06,"output_power":6.384104305136621e-06,"reference_active":true,"delta_db":-1.6197542592314078}],"active_band_count":3,"active_band_mae_db":0.8952413495254111}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012704534221320317,"log_rms_mae_db":0.8865806523530582,"rms_envelope_nrmse":0.08136682747281018,"envelope_correlation":0.9977185332919329,"rms_gain_error_db":-0.5149869918449294,"reference_mean_power":0.0005824704905358113,"output_mean_power":0.0005173390111011252},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.040463159930298e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002294596306039209,"output_power":0.00021121152554151677,"reference_active":true,"delta_db":-0.3598867669923111},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.014148036198554e-05,"output_power":4.94385213608568e-05,"reference_active":true,"delta_db":-0.8510864079043713},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3285241811094983e-06,"output_power":8.017303832657129e-07,"reference_active":true,"delta_db":-2.193411212059472}],"active_band_count":3,"active_band_mae_db":1.1347947956520514}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4077-13754-0012","method":"magphase","status":"measured","mos":{"status":"measured","value":2.462630271911621,"predicted_mos":2.462630271911621,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288776,"input_sample_rate":48000,"input_seconds":6.016166666666667,"input_float32_sha256":"09b0be1958d04ff350f2c4c851244b78105c75b6da1c9e68f1f7e5be6826f4a2","internal_samples":96259,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4995937860003323},"started_at":"2026-10-13T05:15:16.216799+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009938872999555315,"native_input_samples":288000,"native_input_sha256":"8f2ee2c37afade06b0f2f301a86acb337b049961b49af917b5589f87d4d8a532","output_sha256":"02a757af237bdee146850bde20c2bd8f4b7aa43df20f7a1a3cf224f4dbd0947d","output_samples":288776,"length_delta_samples":776,"analysis_s":3.059223013000519,"synthesis_s":0.9883599620006862,"total_s":4.047582975001205,"rtf":0.6745971625002009,"output_peak":0.46660368418645753,"output_rms":0.04248553081694827,"finished_at":"2026-10-13T05:15:20.287527+00:00","evaluation_input":{"reference":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"output_sha256":"02a757af237bdee146850bde20c2bd8f4b7aa43df20f7a1a3cf224f4dbd0947d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96259,"length_delta_samples":259,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.431182084829224,"output_peak":0.4603050658677251},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973093425030387,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.818869435397486,"spectral_convergence":0.8213916007744572,"spectral_convergence_status":"ok","amplitude_floor":1.32832143223519e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.987202647236854,"spectral_convergence":0.720353680056,"spectral_convergence_status":"ok","amplitude_floor":1.1133586355586426e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.940859881557096,"spectral_convergence":0.6562445059890765,"spectral_convergence_status":"ok","amplitude_floor":6.710028542715064e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00015990772541983682,"output_power":0.00013998374430236948,"delta_db":-0.5779183987864124,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005346827040387907,"output_power":0.0004819099603048029,"delta_db":-0.4513023316970726,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025601253900142847,"output_power":0.00023508974206976773,"delta_db":-0.3702755725282075,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015659557404440672,"output_power":0.00014868019005126953,"delta_db":-0.22526375603946658,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003554436518358531,"output_power":0.00037626168073614467,"delta_db":0.24719228921429814,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004235024710497043,"output_power":0.00042869825090227425,"delta_db":0.05295762890799814,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3208183295289092,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.17746830469006425,"pitch":null,"output_rms_dbfs":-27.421718308422736,"output_peak":0.4603050658677251,"reference_rms_dbfs":-27.244250003732674,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.915643988063811},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.547257339775553,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04342976704277862,"rms_dbfs":-27.244250003732674,"peak":0.543731689453125,"peak_dbfs":-5.292307099260677,"crest_factor_db":21.951942904471995,"mean":2.936585744222005e-05,"zero_energy":false},"output":{"rms":0.04255142263622485,"rms_dbfs":-27.421718308422736,"peak":0.4603050658677251,"peak_dbfs":-6.739084909407389,"crest_factor_db":20.682633399015348,"mean":0.0006497906029020526,"zero_energy":false},"rms_difference_db":-0.1774683046900627},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.540612968499772e-06,"threshold_dbfs":-54.509215442476574,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":84,"quiet_samples":13440,"reference_mean_power":1.6902210301763954e-06,"output_mean_power":3.188083827199107e-06,"output_rms_dbfs":-54.96470267824427,"output_zero_energy":false,"output_peak":0.04646275957897234,"output_energy_fraction":0.00024650719432008317,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":64,"output_estimated_voiced_frames":55,"both_estimated_voiced_frames":34,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.53125,"absolute_difference_cents_median":8.731970690969238,"absolute_difference_cents_p95":50.280072821231734,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018816516597127566,"log_rms_mae_db":4.95565193208559,"rms_envelope_nrmse":0.5083133916161952,"envelope_correlation":0.7697396687843218,"rms_gain_error_db":-0.17584679821635052,"reference_mean_power":0.0018869472812584292,"output_mean_power":0.0018120704861652383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.448172186664129e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039739520206635815,"output_power":0.0003756660721689922,"reference_active":true,"delta_db":-0.24420645321060797},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001684948117890412,"output_power":0.00015618447106676727,"reference_active":true,"delta_db":-0.3294868168062981},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9559595301379592e-05,"output_power":2.155635429980608e-05,"reference_active":true,"delta_db":0.4221544828153192}],"active_band_count":3,"active_band_mae_db":0.33194925094407507}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016656172387320294,"log_rms_mae_db":3.7510646873377427,"rms_envelope_nrmse":0.3880375046812477,"envelope_correlation":0.8530012043328565,"rms_gain_error_db":-0.17889961172939808,"reference_mean_power":0.0018945898208991615,"output_mean_power":0.0018181312790423142},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.514104383617564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003829793712831679,"output_power":0.00036468631124006283,"reference_active":true,"delta_db":-0.21255919788305083},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.831616093630877e-05,"output_power":7.186703430123035e-05,"reference_active":true,"delta_db":-0.3732166674437088},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3812315351816923e-06,"output_power":2.2583174276904575e-06,"reference_active":true,"delta_db":-0.2301663933545816}],"active_band_count":3,"active_band_mae_db":0.2719807528937804}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015092951870698086,"log_rms_mae_db":5.682303476784998,"rms_envelope_nrmse":0.44731915403752937,"envelope_correlation":0.8549996626701047,"rms_gain_error_db":-0.4393331154857045,"reference_mean_power":0.000631981580682194,"output_mean_power":0.0005711775236930782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.167139764996573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018037000728302603,"output_power":0.00016362764111439913,"reference_active":true,"delta_db":-0.4230765325133995},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6944061726753776e-05,"output_power":2.4237813771444604e-05,"reference_active":true,"delta_db":-0.45969620458462807},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.521461134259169e-06,"output_power":3.9167110960171555e-06,"reference_active":true,"delta_db":0.4619863882703932}],"active_band_count":3,"active_band_mae_db":0.4482530417894736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013588907795973108,"log_rms_mae_db":4.575161832953513,"rms_envelope_nrmse":0.3529373906202285,"envelope_correlation":0.905501270999094,"rms_gain_error_db":-0.4496582894759229,"reference_mean_power":0.0006345375227010602,"output_mean_power":0.0005721257302813897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.990151794989553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001715162379513309,"output_power":0.0001565033230643739,"reference_active":true,"delta_db":-0.39781678836391915},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2552077755571314e-05,"output_power":1.1145581593335382e-05,"reference_active":true,"delta_db":-0.5161288533242733},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0572536267135004e-07,"output_power":3.277485409547869e-07,"reference_active":true,"delta_db":0.30209298150411035}],"active_band_count":3,"active_band_mae_db":0.4053462077307676}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.324369547066363e-05,"log_rms_mae_db":6.235188114221934,"rms_envelope_nrmse":0.609339987322452,"envelope_correlation":0.7000276182527111,"rms_gain_error_db":-0.16479526188552063,"reference_mean_power":0.00026216016454011775,"output_mean_power":0.0002523987354827943},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5456728460340465e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.502085853898133e-05,"output_power":7.140241139660693e-05,"reference_active":true,"delta_db":-0.21469150927952502},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.151594827944161e-05,"output_power":1.8731244049746928e-05,"reference_active":true,"delta_db":-0.6019386931369133},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.51387829077428e-06,"output_power":1.1430183042494033e-05,"reference_active":true,"delta_db":1.2792574774970342}],"active_band_count":3,"active_band_mae_db":0.6986292266378241}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.873751028776232e-05,"log_rms_mae_db":4.6913507561576955,"rms_envelope_nrmse":0.45546431067267024,"envelope_correlation":0.8143042162320152,"rms_gain_error_db":-0.16654668505382536,"reference_mean_power":0.00026328876508602793,"output_mean_power":0.0002533831080686592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3635149275752304e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.257955032157828e-05,"output_power":6.954127930917022e-05,"reference_active":true,"delta_db":-0.18571597255088573},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405048683009273e-06,"output_power":8.260369239936434e-06,"reference_active":true,"delta_db":-0.5636158712197075},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.677553554297771e-07,"output_power":8.939104501863569e-07,"reference_active":true,"delta_db":0.6607115987130809}],"active_band_count":3,"active_band_mae_db":0.4700144808278914}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001568272998494035,"log_rms_mae_db":6.4735753367832105,"rms_envelope_nrmse":0.7227599377371224,"envelope_correlation":0.6862602939933045,"rms_gain_error_db":0.19932635624149064,"reference_mean_power":0.0005797089728212456,"output_mean_power":0.0006069356645698146},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.292445291957658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020605682430557343,"output_power":0.0001955088885082274,"reference_active":true,"delta_db":-0.22820495601810054},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001264586344183223,"output_power":0.00011555409253761787,"reference_active":true,"delta_db":-0.3916315637281295},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.269907486994092e-06,"output_power":9.183044714508478e-06,"reference_active":true,"delta_db":-0.04088700929584376}],"active_band_count":3,"active_band_mae_db":0.22024117634735796}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012704534221320317,"log_rms_mae_db":5.000662641390767,"rms_envelope_nrmse":0.5595501540120236,"envelope_correlation":0.8013266448447399,"rms_gain_error_db":0.1998559173217564,"reference_mean_power":0.0005824704905358113,"output_mean_power":0.000609901244457695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.040463159930298e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002294596306039209,"output_power":0.00021865097936633602,"reference_active":true,"delta_db":-0.20954863152373174},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.014148036198554e-05,"output_power":5.486912400696835e-05,"reference_active":true,"delta_db":-0.39846087225394955},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3285241811094983e-06,"output_power":1.121178323684699e-06,"reference_active":true,"delta_db":-0.7369477089004963}],"active_band_count":3,"active_band_mae_db":0.4483190708927259}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},{"case":"librispeech_4446-2271-0003","method":"hnm","status":"measured","mos":{"status":"measured","value":2.5355749130249023,"predicted_mos":2.5355749130249023,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59200,"input_sample_rate":16000,"input_seconds":3.7,"input_float32_sha256":"17e1c82af984aeb37c52de8bdf2256edbcfac296bbf0ae4d7e7e16fb57cdb5ad","internal_samples":59200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4184939160004433},"started_at":"2026-10-13T04:39:12.966218+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4740000551682897e-05,"native_input_samples":59200,"native_input_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","output_sha256":"a36c8589e59f3873267e2b6b595a2908a1520a6aca26878f47b6112c43c26650","output_samples":59200,"length_delta_samples":0,"analysis_s":0.7348170440000104,"synthesis_s":0.3167712450003819,"total_s":1.0515882890003923,"rtf":0.2842130510811871,"output_peak":0.47694135518242375,"output_rms":0.07386365095992434,"finished_at":"2026-10-13T04:39:14.022228+00:00","evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"a36c8589e59f3873267e2b6b595a2908a1520a6aca26878f47b6112c43c26650","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.71452946409804,"spectral_convergence":0.20616920930466262,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.70459878624485,"spectral_convergence":0.2058146143347425,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.66980136518324,"spectral_convergence":0.21701748375376198,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.0008021596660065693,"delta_db":-0.08806212675753468,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.0030178586208366265,"delta_db":-0.0966911003146532,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.0013404132131204288,"delta_db":0.0036433773822625456,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.00027493398561358504,"delta_db":-0.06944484257166772,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":9.901117616443461e-06,"delta_db":1.4806493151582811,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":1.0572329935879646e-05,"delta_db":2.2415595738370753,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6633417226702457,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06365766583005214,"pitch":null,"output_rms_dbfs":-22.631384593716632,"output_peak":0.47694135518242375,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.029643205175376},"common_prefix":{"samples":59200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.71452946409804,"spectral_convergence":0.20616920930466262,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.70459878624485,"spectral_convergence":0.2058146143347425,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.66980136518324,"spectral_convergence":0.21701748375376198,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.0008021596660065693,"delta_db":-0.08806212675753468,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.0030178586208366265,"delta_db":-0.0966911003146532,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.0013404132131204288,"delta_db":0.0036433773822625456,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.00027493398561358504,"delta_db":-0.06944484257166772,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":9.901117616443461e-06,"delta_db":1.4806493151582811,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":1.0572329935879646e-05,"delta_db":2.2415595738370753,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6633417226702457,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06365766583005214,"pitch":null,"output_rms_dbfs":-22.631384593716632,"output_peak":0.47694135518242375,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.029643205175376},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59200,"output_samples":59200,"intelligibility":{"estoi":{"value":0.9161391592501423,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07440697583274426,"rms_dbfs":-22.567726927886582,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.84980550831495,"mean":-3.489932498416385e-07,"zero_energy":false},"output":{"rms":0.07386365095992434,"rms_dbfs":-22.631384593716632,"peak":0.47694135518242375,"peak_dbfs":-6.43070037250782,"crest_factor_db":16.20068422120881,"mean":-5.7843553062944246e-05,"zero_energy":false},"rms_difference_db":-0.06365766583004984},"full_scale":{"reference":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":370,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":6.431599327346735e-07,"output_mean_power":1.0706551616321585e-06,"output_rms_dbfs":-59.703503849248314,"output_zero_energy":false,"output_peak":0.012095268472943135,"output_energy_fraction":3.0231605105087245e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":363,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":188,"both_estimated_voiced_frames":185,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9390862944162437,"absolute_difference_cents_median":1.8940343835260491,"absolute_difference_cents_p95":10.951429377699032,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002672933594000982,"log_rms_mae_db":0.7763457457878511,"rms_envelope_nrmse":0.071704047803339,"envelope_correlation":0.993077943670271,"rms_gain_error_db":-0.06365773818588169,"reference_mean_power":0.0055438895852693145,"output_mean_power":0.005463221366876571},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3303693041440345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007860073564586987,"output_power":0.0007534880041042689,"reference_active":true,"delta_db":-0.18350268213574056},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002436436408779341,"output_power":0.0002333247702234777,"reference_active":true,"delta_db":-0.1879423392759637},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.963914121363999e-05,"output_power":4.30777731211563e-05,"reference_active":true,"delta_db":-0.6157101472000915}],"active_band_count":3,"active_band_mae_db":0.32905172287059864}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00019127079187884833,"log_rms_mae_db":0.41774906677868856,"rms_envelope_nrmse":0.03494300466246297,"envelope_correlation":0.9981507215221597,"rms_gain_error_db":-0.0636578740460913,"reference_mean_power":0.005589268172913711,"output_mean_power":0.005507939485909464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.0930617045795466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007030365316665855,"output_power":0.0006753768767674784,"reference_active":true,"delta_db":-0.17431705328126293},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013359627830049331,"output_power":0.00012651662165434453,"reference_active":true,"delta_db":-0.23646773287605222},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.77873293352624e-06,"output_power":2.3676691403949855e-06,"reference_active":true,"delta_db":-0.6952579465605877}],"active_band_count":3,"active_band_mae_db":0.3686809109059676}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002516933783769556,"log_rms_mae_db":1.1108103452669102,"rms_envelope_nrmse":0.09749867777615613,"envelope_correlation":0.989447918788584,"rms_gain_error_db":-0.11994263664923707,"reference_mean_power":0.003609667084718048,"output_mean_power":0.0035113400142944397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0389022511376814e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005770551731550911,"output_power":0.0005463735967026155,"reference_active":true,"delta_db":-0.23727634657200639},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021321841152816334,"output_power":0.0001980639710779995,"reference_active":true,"delta_db":-0.320192213670911},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0182675609825546e-05,"output_power":4.418286653703669e-05,"reference_active":true,"delta_db":-0.5529992403873246}],"active_band_count":3,"active_band_mae_db":0.370155933543414}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00017703062295034008,"log_rms_mae_db":0.5954287697980194,"rms_envelope_nrmse":0.04916372550649012,"envelope_correlation":0.9970608833172562,"rms_gain_error_db":-0.11994268274415074,"reference_mean_power":0.003639213964400563,"output_mean_power":0.0035400820015369892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":8.593451292498865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005240710435115443,"output_power":0.0004964381465551959,"reference_active":true,"delta_db":-0.23525018786018842},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012044100626957098,"output_power":0.00010998033631457944,"reference_active":true,"delta_db":-0.39459331932525443},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7064893260916713e-06,"output_power":2.5162191518553244e-06,"reference_active":true,"delta_db":-0.31657855050752526}],"active_band_count":3,"active_band_mae_db":0.3154740192309894}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":1.5645424359940652,"rms_envelope_nrmse":0.13272576115089085,"envelope_correlation":0.9864257171348992,"rms_gain_error_db":-0.05459970586399503,"reference_mean_power":0.00024301128670702196,"output_mean_power":0.0002399752619294654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.5901124046586961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178189042095749e-05,"output_power":9.112485382440373e-05,"reference_active":true,"delta_db":-0.031201538996227726},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7997292458206196e-05,"output_power":1.5732354615807033e-05,"reference_active":true,"delta_db":-0.5841344697172663},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.335696736133691e-06,"output_power":2.3751546456620522e-06,"reference_active":true,"delta_db":0.07275437812442741}],"active_band_count":3,"active_band_mae_db":0.22936346227930715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":1.0450160766214207,"rms_envelope_nrmse":0.08279693494971697,"envelope_correlation":0.9944727273441383,"rms_gain_error_db":-0.05460047809131479,"reference_mean_power":0.00024500023514433596,"output_mean_power":0.00024193931872042523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.4888167929489614e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.774520748189045e-05,"output_power":8.73957495460182e-05,"reference_active":true,"delta_db":-0.017330938141315158},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.213495209539379e-06,"output_power":7.928014907622011e-06,"reference_active":true,"delta_db":-0.6525995625894955},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5741557318465008e-07,"output_power":1.6596076164079003e-07,"reference_active":true,"delta_db":0.22957724188911321}],"active_band_count":3,"active_band_mae_db":0.2998359142066413}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":2.2099356249469584,"rms_envelope_nrmse":0.5912392169150642,"envelope_correlation":0.8115806845191476,"rms_gain_error_db":1.6353164601309997,"reference_mean_power":7.123238480442997e-06,"output_mean_power":1.038028148219933e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":4.409187309389393e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.616739584019063e-06,"output_power":2.7317236292610963e-06,"reference_active":true,"delta_db":0.18676255174106565},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.009142614437226e-07,"output_power":9.18899171773279e-07,"reference_active":true,"delta_db":1.1760296345911112},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642480310119984e-07,"output_power":3.5345781436405434e-07,"reference_active":true,"delta_db":1.2632582769314615}],"active_band_count":3,"active_band_mae_db":0.8753501544212128}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":1.987241735802386,"rms_envelope_nrmse":0.5017912274650987,"envelope_correlation":0.8526964151192936,"rms_gain_error_db":1.6353311058697353,"reference_mean_power":7.1814170088864135e-06,"output_mean_power":1.0465096962199367e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":3.9660429491134744e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6674623865712373e-06,"output_power":2.81196354843333e-06,"reference_active":true,"delta_db":0.22911382384239293},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2554368602964405e-07,"output_power":5.125696923791886e-07,"reference_active":true,"delta_db":1.971436467486472},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7820276752795495e-08,"output_power":2.851937893728507e-08,"reference_active":true,"delta_db":0.10778617722509748}],"active_band_count":3,"active_band_mae_db":0.7694454895179875}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0003","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.732762098312378,"predicted_mos":2.732762098312378,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59200,"input_sample_rate":16000,"input_seconds":3.7,"input_float32_sha256":"90b6b2436f10a7d66197b6e504a8865765f451f86e8e323898441045c923750c","internal_samples":59200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.423174237001149},"started_at":"2026-10-13T04:42:09.520816+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2531000468006823e-05,"native_input_samples":59200,"native_input_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","output_sha256":"7bd5a3164a6766cbd090e35112c4448d338a055eea2c3c8d517f2dbb2670b364","output_samples":59200,"length_delta_samples":0,"analysis_s":0.9529454700004862,"synthesis_s":0.06931701499979681,"total_s":1.022262485000283,"rtf":0.2762871581081846,"output_peak":0.6223962629342017,"output_rms":0.08400467305516873,"finished_at":"2026-10-13T04:42:10.548437+00:00","evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"7bd5a3164a6766cbd090e35112c4448d338a055eea2c3c8d517f2dbb2670b364","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.015814968485868,"spectral_convergence":0.621533997447968,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.216492980142872,"spectral_convergence":0.272690186954614,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.751680597699806,"spectral_convergence":0.32545825300234493,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.0008261363076910791,"delta_db":0.03984675515887312,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.004238122710261628,"delta_db":1.3780552675176871,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.0016445660809570229,"delta_db":0.8917696622374862,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.00033069748121336844,"delta_db":0.7325797490709343,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":8.679514708419513e-06,"delta_db":0.9087615539703531,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":8.583000274273235e-06,"delta_db":1.3362437587487397,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8812094577840123,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0537958457013157,"pitch":null,"output_rms_dbfs":-21.513931082185266,"output_peak":0.6223962629342017,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.661329515442849},"common_prefix":{"samples":59200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.015814968485868,"spectral_convergence":0.621533997447968,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.216492980142872,"spectral_convergence":0.272690186954614,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.751680597699806,"spectral_convergence":0.32545825300234493,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.0008261363076910791,"delta_db":0.03984675515887312,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.004238122710261628,"delta_db":1.3780552675176871,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.0016445660809570229,"delta_db":0.8917696622374862,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.00033069748121336844,"delta_db":0.7325797490709343,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":8.679514708419513e-06,"delta_db":0.9087615539703531,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":8.583000274273235e-06,"delta_db":1.3362437587487397,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8812094577840123,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0537958457013157,"pitch":null,"output_rms_dbfs":-21.513931082185266,"output_peak":0.6223962629342017,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.661329515442849},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59200,"output_samples":59200,"intelligibility":{"estoi":{"value":0.9462722021845738,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07440697583274426,"rms_dbfs":-22.567726927886582,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.84980550831495,"mean":-3.489932498416385e-07,"zero_energy":false},"output":{"rms":0.08400467305516873,"rms_dbfs":-21.513931082185266,"peak":0.6223962629342017,"peak_dbfs":-4.118660473064238,"crest_factor_db":17.395270609121027,"mean":-1.1651660347291772e-09,"zero_energy":false},"rms_difference_db":1.0537958457013161},"full_scale":{"reference":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":370,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":6.431599327346735e-07,"output_mean_power":4.5958621726244106e-06,"output_rms_dbfs":-53.37633004040805,"output_zero_energy":false,"output_peak":0.017189566660590985,"output_energy_fraction":0.00010033055988306941,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":363,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":207,"both_estimated_voiced_frames":194,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9847715736040609,"absolute_difference_cents_median":6.347946733272162,"absolute_difference_cents_p95":23.998225428562055,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002672933594000982,"log_rms_mae_db":1.6021000485847734,"rms_envelope_nrmse":0.1921379405661403,"envelope_correlation":0.9764553709735659,"rms_gain_error_db":1.053795955078264,"reference_mean_power":0.0055438895852693145,"output_mean_power":0.007066334105318764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3303693041440345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007860073564586987,"output_power":0.0009593028469999158,"reference_active":true,"delta_db":0.8652912262266533},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002436436408779341,"output_power":0.00031459497402617945,"reference_active":true,"delta_db":1.1099669930962404},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.963914121363999e-05,"output_power":6.315151856308485e-05,"reference_active":true,"delta_db":1.0455953886951699}],"active_band_count":3,"active_band_mae_db":1.006951202672688}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00019127079187884833,"log_rms_mae_db":1.2287411008356046,"rms_envelope_nrmse":0.1624209521830042,"envelope_correlation":0.9869566338305003,"rms_gain_error_db":1.0537963613807608,"reference_mean_power":0.005589268172913711,"output_mean_power":0.007124175076281416},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.0930617045795466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007030365316665855,"output_power":0.0008739342161198708,"reference_active":true,"delta_db":0.945008503879487},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013359627830049331,"output_power":0.00017398335735479826,"reference_active":true,"delta_db":1.1471334734824368},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.77873293352624e-06,"output_power":3.0855890790838385e-06,"reference_active":true,"delta_db":0.45491280392186295}],"active_band_count":3,"active_band_mae_db":0.8490182604279289}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002516933783769556,"log_rms_mae_db":1.630524668485485,"rms_envelope_nrmse":0.2461866649471375,"envelope_correlation":0.973725357630209,"rms_gain_error_db":1.4232113926856649,"reference_mean_power":0.003609667084718048,"output_mean_power":0.005009429726090138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0389022511376814e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005770551731550911,"output_power":0.0007922193407736474,"reference_active":true,"delta_db":1.3762810208468932},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021321841152816334,"output_power":0.0003111033430771462,"reference_active":true,"delta_db":1.6407997446401192},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0182675609825546e-05,"output_power":6.0375574164090286e-05,"reference_active":true,"delta_db":0.8030746081607998}],"active_band_count":3,"active_band_mae_db":1.2733851245492707}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00017703062295034008,"log_rms_mae_db":1.3516519058701622,"rms_envelope_nrmse":0.21812644102033332,"envelope_correlation":0.983406960669348,"rms_gain_error_db":1.4232113893666924,"reference_mean_power":0.003639213964400563,"output_mean_power":0.005050434339532018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":8.593451292498865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005240710435115443,"output_power":0.0007304810746846731,"reference_active":true,"delta_db":1.4421880440783563},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012044100626957098,"output_power":0.00017057135714487975,"reference_active":true,"delta_db":1.5113172958678769},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7064893260916713e-06,"output_power":2.9317234640972693e-06,"reference_active":true,"delta_db":0.34716684262023056}],"active_band_count":3,"active_band_mae_db":1.1002240608554879}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":1.2537946974199081,"rms_envelope_nrmse":0.1342715685243582,"envelope_correlation":0.9942884259958288,"rms_gain_error_db":0.8205973100666042,"reference_mean_power":0.00024301128670702196,"output_mean_power":0.00029355276538810336},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.5901124046586961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178189042095749e-05,"output_power":0.00010567590251485843,"reference_active":true,"delta_db":0.6121896700319018},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7997292458206196e-05,"output_power":2.065931438856009e-05,"reference_active":true,"delta_db":0.5990873063277216},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.335696736133691e-06,"output_power":2.5577169700098294e-06,"reference_active":true,"delta_db":0.3943603109486282}],"active_band_count":3,"active_band_mae_db":0.5352124291027506}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":0.8764750898597427,"rms_envelope_nrmse":0.11403630034560576,"envelope_correlation":0.9976503470351081,"rms_gain_error_db":0.8205981284070735,"reference_mean_power":0.00024500023514433596,"output_mean_power":0.00029595543101680697},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.4888167929489614e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.774520748189045e-05,"output_power":0.00010105928773305686,"reference_active":true,"delta_db":0.6135282776051092},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.213495209539379e-06,"output_power":1.0445314618620621e-05,"reference_active":true,"delta_db":0.5449711132958823},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5741557318465008e-07,"output_power":1.5269424092460393e-07,"reference_active":true,"delta_db":-0.13225037726916872}],"active_band_count":3,"active_band_mae_db":0.43024992272338675}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":1.6815642813021503,"rms_envelope_nrmse":0.3190394494558673,"envelope_correlation":0.9527124474332819,"rms_gain_error_db":1.3439291389540573,"reference_mean_power":7.123238480442997e-06,"output_mean_power":9.706672978451415e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":4.409187309389393e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.616739584019063e-06,"output_power":3.672333290119055e-06,"reference_active":true,"delta_db":1.471815854759919},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.009142614437226e-07,"output_power":1.183085780979268e-06,"reference_active":true,"delta_db":2.273513381577349},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642480310119984e-07,"output_power":3.9165385834272897e-07,"reference_active":true,"delta_db":1.7089064937413752}],"active_band_count":3,"active_band_mae_db":1.818078576692881}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":1.1996877634084064,"rms_envelope_nrmse":0.22164216387424643,"envelope_correlation":0.9849923842095234,"rms_gain_error_db":1.3439468519132827,"reference_mean_power":7.1814170088864135e-06,"output_mean_power":9.785991431563711e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":3.9660429491134744e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6674623865712373e-06,"output_power":3.8125009864712285e-06,"reference_active":true,"delta_db":1.5511166049920857},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2554368602964405e-07,"output_power":5.667817417417054e-07,"reference_active":true,"delta_db":2.408065749199962},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7820276752795495e-08,"output_power":3.344207017533145e-08,"reference_active":true,"delta_db":0.799317078842585}],"active_band_count":3,"active_band_mae_db":1.586166477678211}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0003","method":"facodec","status":"measured","mos":{"status":"measured","value":3.7446048259735107,"predicted_mos":3.7446048259735107,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59200,"input_sample_rate":16000,"input_seconds":3.7,"input_float32_sha256":"f8dfbc25b144cb042ed453a625937eff89dbb4ad1f3745333cc8e7d160954627","internal_samples":59200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5208851380011765},"started_at":"2026-10-13T04:52:10.150137+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.8434999876481015e-05,"native_input_samples":59200,"native_input_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","output_sha256":"72abb087d8539809df6db1420ce95b7997fd50cf3ed26ca9b37111815e6ddb4f","output_samples":59200,"length_delta_samples":0,"analysis_s":0.942963941999551,"synthesis_s":6.371219200999803,"total_s":7.3141831429993545,"rtf":1.9768062548646903,"output_peak":0.45517733693122864,"output_rms":0.06935601463756448,"finished_at":"2026-10-13T04:52:17.470855+00:00","evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"72abb087d8539809df6db1420ce95b7997fd50cf3ed26ca9b37111815e6ddb4f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.942268950903812,"spectral_convergence":0.2268505293889722,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.058128956042383,"spectral_convergence":0.22796829053883153,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.416632694843375,"spectral_convergence":0.2539739774073846,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.000653434121022299,"delta_db":-0.9786522506253638,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.002940816248221976,"delta_db":-0.20900111111302314,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.000999444483735107,"delta_db":-1.2711568718760178,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.0002029373560412105,"delta_db":-1.3881091428453285,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":8.55713998128907e-06,"delta_db":0.8470934836150074,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":5.067417404176839e-06,"delta_db":-0.9522807184141497,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9410489297481485,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6105904803979002,"pitch":null,"output_rms_dbfs":-23.178317408284478,"output_peak":0.45517733693122864,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.139010200596523},"common_prefix":{"samples":59200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.942268950903812,"spectral_convergence":0.2268505293889722,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.058128956042383,"spectral_convergence":0.22796829053883153,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.416632694843375,"spectral_convergence":0.2539739774073846,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.000653434121022299,"delta_db":-0.9786522506253638,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.002940816248221976,"delta_db":-0.20900111111302314,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.000999444483735107,"delta_db":-1.2711568718760178,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.0002029373560412105,"delta_db":-1.3881091428453285,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":8.55713998128907e-06,"delta_db":0.8470934836150074,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":5.067417404176839e-06,"delta_db":-0.9522807184141497,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9410489297481485,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6105904803979002,"pitch":null,"output_rms_dbfs":-23.178317408284478,"output_peak":0.45517733693122864,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.139010200596523},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59200,"output_samples":59200,"intelligibility":{"estoi":{"value":0.9063172605592428,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07440697583274426,"rms_dbfs":-22.567726927886582,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.84980550831495,"mean":-3.489932498416385e-07,"zero_energy":false},"output":{"rms":0.06935601463756448,"rms_dbfs":-23.178317408284478,"peak":0.45517733693122864,"peak_dbfs":-6.836387387915757,"crest_factor_db":16.341930020368725,"mean":-1.4928115494921552e-05,"zero_energy":false},"rms_difference_db":-0.6105904803978959},"full_scale":{"reference":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":370,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":6.431599327346735e-07,"output_mean_power":9.880715866145088e-07,"output_rms_dbfs":-60.05211589273308,"output_zero_energy":false,"output_peak":0.01072863768786192,"output_energy_fraction":3.1644138973336695e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":363,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":197,"both_estimated_voiced_frames":191,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9695431472081218,"absolute_difference_cents_median":4.43154369075662,"absolute_difference_cents_p95":20.78765169284009,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002672933594000982,"log_rms_mae_db":1.1150293701612874,"rms_envelope_nrmse":0.10700705766063008,"envelope_correlation":0.9900757332318482,"rms_gain_error_db":-0.6105904356834071,"reference_mean_power":0.0055438895852693145,"output_mean_power":0.004816765776426544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3303693041440345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007860073564586987,"output_power":0.0006664886253572515,"reference_active":true,"delta_db":-0.7163386882292893},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002436436408779341,"output_power":0.00023636278374249644,"reference_active":true,"delta_db":-0.13175984534276025},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.963914121363999e-05,"output_power":5.659082160471483e-05,"reference_active":true,"delta_db":0.5692173987395062}],"active_band_count":3,"active_band_mae_db":0.4724386441038519}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00019127079187884833,"log_rms_mae_db":0.953682414351144,"rms_envelope_nrmse":0.09436836855047595,"envelope_correlation":0.9927743933961113,"rms_gain_error_db":-0.6105906335285293,"reference_mean_power":0.005589268172913711,"output_mean_power":0.004856192391645254},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.0930617045795466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007030365316665855,"output_power":0.0005941678786480604,"reference_active":true,"delta_db":-0.7306872303660249},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013359627830049331,"output_power":0.00012812677171334763,"reference_active":true,"delta_db":-0.18154476030354957},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.77873293352624e-06,"output_power":3.482939232920162e-06,"reference_active":true,"delta_db":0.9809908900141411}],"active_band_count":3,"active_band_mae_db":0.6310742935612385}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002516933783769556,"log_rms_mae_db":1.226717266086273,"rms_envelope_nrmse":0.12109613922144948,"envelope_correlation":0.9868900650708796,"rms_gain_error_db":-0.5418147615118714,"reference_mean_power":0.003609667084718048,"output_mean_power":0.003186292733577609},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0389022511376814e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005770551731550911,"output_power":0.0004999567299700752,"reference_active":true,"delta_db":-0.622849198819863},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021321841152816334,"output_power":0.0002096030686487352,"reference_active":true,"delta_db":-0.07427066989924073},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0182675609825546e-05,"output_power":5.653850174544752e-05,"reference_active":true,"delta_db":0.5179048257431272}],"active_band_count":3,"active_band_mae_db":0.40500823148741033}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00017703062295034008,"log_rms_mae_db":1.018659465270688,"rms_envelope_nrmse":0.0987772239683019,"envelope_correlation":0.9915233560411678,"rms_gain_error_db":-0.5418147728211758,"reference_mean_power":0.003639213964400563,"output_mean_power":0.003212374080036173},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":8.593451292498865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005240710435115443,"output_power":0.00045477910331015605,"reference_active":true,"delta_db":-0.6158966326567775},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012044100626957098,"output_power":0.00011691585483811796,"reference_active":true,"delta_db":-0.12900965951176288},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7064893260916713e-06,"output_power":3.234364752850246e-06,"reference_active":true,"delta_db":0.7738267688774876}],"active_band_count":3,"active_band_mae_db":0.5062443536820093}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":1.7449461087268066,"rms_envelope_nrmse":0.19990750277887417,"envelope_correlation":0.983017853359517,"rms_gain_error_db":-1.3620141764223619,"reference_mean_power":0.00024301128670702196,"output_mean_power":0.0001775926661860946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.5901124046586961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178189042095749e-05,"output_power":6.432876136283608e-05,"reference_active":true,"delta_db":-1.5435180933455317},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7997292458206196e-05,"output_power":1.3881924977057023e-05,"reference_active":true,"delta_db":-1.127574810647293},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.335696736133691e-06,"output_power":1.792418337552913e-06,"reference_active":true,"delta_db":-1.1497707549659244}],"active_band_count":3,"active_band_mae_db":1.2736212196529164}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":1.4512216926499062,"rms_envelope_nrmse":0.18344813052721218,"envelope_correlation":0.9880437090838392,"rms_gain_error_db":-1.3620151897657273,"reference_mean_power":0.00024500023514433596,"output_mean_power":0.00017904614807325584},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.4888167929489614e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.774520748189045e-05,"output_power":6.169989368189546e-05,"reference_active":true,"delta_db":-1.5293898956050058},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.213495209539379e-06,"output_power":7.032843144575006e-06,"reference_active":true,"delta_db":-1.1729348303962162},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5741557318465008e-07,"output_power":1.2415759825871625e-07,"reference_active":true,"delta_db":-1.0307439219806058}],"active_band_count":3,"active_band_mae_db":1.2443562159939425}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":2.2311460053120213,"rms_envelope_nrmse":0.24897147056810615,"envelope_correlation":0.9563852370017909,"rms_gain_error_db":-0.8649683535402991,"reference_mean_power":7.123238480442997e-06,"output_mean_power":5.836878445005922e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":4.409187309389393e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.616739584019063e-06,"output_power":2.164384164316344e-06,"reference_active":true,"delta_db":-0.8242615636270065},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.009142614437226e-07,"output_power":5.799513334077079e-07,"reference_active":true,"delta_db":-0.8227334533850834},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642480310119984e-07,"output_power":2.0239256346673988e-07,"reference_active":true,"delta_db":-1.158172088097878}],"active_band_count":3,"active_band_mae_db":0.9350557017033226}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":1.8416071147749442,"rms_envelope_nrmse":0.18580947231379108,"envelope_correlation":0.9763969918900558,"rms_gain_error_db":-0.8649943726326252,"reference_mean_power":7.1814170088864135e-06,"output_mean_power":5.884515467169441e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":3.9660429491134744e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6674623865712373e-06,"output_power":2.2020260704035013e-06,"reference_active":true,"delta_db":-0.8327584782761814},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2554368602964405e-07,"output_power":2.602598995537594e-07,"reference_active":true,"delta_db":-0.9720201871967017},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7820276752795495e-08,"output_power":1.5523362917599572e-08,"reference_active":true,"delta_db":-2.533756350993937}],"active_band_count":3,"active_band_mae_db":1.4461783388222733}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0003","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.9104628562927246,"predicted_mos":2.9104628562927246,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":163328,"input_sample_rate":44100,"input_seconds":3.7035827664399092,"input_float32_sha256":"282d87e605e4b57c925ea4dc443ace847bad1cd8b401e3f26173e8c08a79a238","internal_samples":59258,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.064023840001028},"started_at":"2026-10-13T07:13:43.291565+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004418864000399481,"native_input_samples":163170,"native_input_sha256":"ef8ed6447924e2e6a1989e82edc3ff0d67cde3a235724e7eb78daf745cc41809","output_sha256":"ffdb7a4838b512d3578977e8ce7cebbfd43a54687d5d7035c15d249f3a331253","output_samples":163328,"length_delta_samples":158,"analysis_s":0.09074437399976887,"synthesis_s":20.155793652000284,"total_s":20.246538026000053,"rtf":5.472037304324338,"output_peak":0.4651215076446533,"output_rms":0.06752554019783201,"finished_at":"2026-10-13T07:14:03.557540+00:00","evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"ffdb7a4838b512d3578977e8ce7cebbfd43a54687d5d7035c15d249f3a331253","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":59200,"output_samples":59258,"length_delta_samples":58,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.40402345760485,"output_peak":0.4623291147499107},"common_prefix":{"samples":59200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990212292011206,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.850851638035435,"spectral_convergence":0.4658377704432703,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.738017899968613,"spectral_convergence":0.38915330570537493,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.908876946511127,"spectral_convergence":0.45441577485368656,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.0002279216906810014,"delta_db":-5.552813760232951,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.0030344987057257106,"delta_db":-0.07281043741041714,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.0011080048549815392,"delta_db":-0.8233269914396668,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.00019142809935388796,"delta_db":-1.6416722482464376,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":5.56897818150391e-06,"delta_db":-1.0184377167078864,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":3.70086353186268e-06,"delta_db":-2.3171167960530297,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.9043629916817315,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8320438146287963,"pitch":null,"output_rms_dbfs":-23.399770742515376,"output_peak":0.4623291147499107,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.832582161505059},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59200,"output_samples":59200,"intelligibility":{"estoi":{"value":0.754136245578905,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07440697583274426,"rms_dbfs":-22.567726927886582,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.84980550831495,"mean":-3.489932498416385e-07,"zero_energy":false},"output":{"rms":0.06761008203260446,"rms_dbfs":-23.399770742515376,"peak":0.4623291147499107,"peak_dbfs":-6.700975128119788,"crest_factor_db":16.69879561439559,"mean":-4.2605034625971865e-05,"zero_energy":false},"rms_difference_db":-0.8320438146287934},"full_scale":{"reference":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":370,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":6.431599327346735e-07,"output_mean_power":1.7577174635699302e-06,"output_rms_dbfs":-57.550509323758654,"output_zero_energy":false,"output_peak":0.017933989740636657,"output_energy_fraction":5.9237848061385964e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":363,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":201,"both_estimated_voiced_frames":188,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9543147208121827,"absolute_difference_cents_median":10.11737607358496,"absolute_difference_cents_p95":33.11662329504119,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002672933594000982,"log_rms_mae_db":1.4122357044381173,"rms_envelope_nrmse":0.13342659303448023,"envelope_correlation":0.9863171915325304,"rms_gain_error_db":-0.8320437666160919,"reference_mean_power":0.0055438895852693145,"output_mean_power":0.0045773086218059205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3303693041440345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007860073564586987,"output_power":0.0007030657369256235,"reference_active":true,"delta_db":-0.48430677114898374},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002436436408779341,"output_power":0.00021870895629419137,"reference_active":true,"delta_db":-0.4688851265481868},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.963914121363999e-05,"output_power":3.756256051418786e-05,"reference_active":true,"delta_db":-1.2106907033530072}],"active_band_count":3,"active_band_mae_db":0.7212942003500592}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00019127079187884833,"log_rms_mae_db":1.0697943777790737,"rms_envelope_nrmse":0.11609363700719988,"envelope_correlation":0.9913565897903288,"rms_gain_error_db":-0.832043707661242,"reference_mean_power":0.005589268172913711,"output_mean_power":0.004614775484118341},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.0930617045795466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007030365316665855,"output_power":0.0006316163481214802,"reference_active":true,"delta_db":-0.4652453039124782},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013359627830049331,"output_power":0.00012034550978586028,"reference_active":true,"delta_db":-0.4536446887390327},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.77873293352624e-06,"output_power":2.3346065921365763e-06,"reference_active":true,"delta_db":-0.7563310086796781}],"active_band_count":3,"active_band_mae_db":0.5584070004437297}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002516933783769556,"log_rms_mae_db":1.7494895892004512,"rms_envelope_nrmse":0.15153719881732133,"envelope_correlation":0.9741916535198011,"rms_gain_error_db":-0.16626619236565943,"reference_mean_power":0.003609667084718048,"output_mean_power":0.0034740857198279794},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0389022511376814e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005770551731550911,"output_power":0.0005616809671685259,"reference_active":true,"delta_db":-0.11727630874484056},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021321841152816334,"output_power":0.00020408884555397417,"reference_active":true,"delta_db":-0.19005434480034022},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0182675609825546e-05,"output_power":3.33262972535212e-05,"reference_active":true,"delta_db":-1.7776674936412489}],"active_band_count":3,"active_band_mae_db":0.6949993823954767}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00017703062295034008,"log_rms_mae_db":1.266013270515062,"rms_envelope_nrmse":0.1164356347423926,"envelope_correlation":0.9824530181518262,"rms_gain_error_db":-0.16626652078494755,"reference_mean_power":0.003639213964400563,"output_mean_power":0.003502522535269769},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":8.593451292498865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005240710435115443,"output_power":0.000510414206588068,"reference_active":true,"delta_db":-0.11467410504987784},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012044100626957098,"output_power":0.00011322924772134479,"reference_active":true,"delta_db":-0.2681575339735096},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7064893260916713e-06,"output_power":1.8570619727649503e-06,"reference_active":true,"delta_db":-1.6357992165280812}],"active_band_count":3,"active_band_mae_db":0.6728769518504896}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":4.335971009877186,"rms_envelope_nrmse":0.33316242492658754,"envelope_correlation":0.9414197767418737,"rms_gain_error_db":-2.176504038026221,"reference_mean_power":0.00024301128670702196,"output_mean_power":0.00014722312812040557},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.5901124046586961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178189042095749e-05,"output_power":4.5769210302491725e-05,"reference_active":true,"delta_db":-3.0218357934672015},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7997292458206196e-05,"output_power":6.6986124857329e-06,"reference_active":true,"delta_db":-4.292223194408063},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.335696736133691e-06,"output_power":9.420911599700626e-07,"reference_active":true,"delta_db":-3.943235252136343}],"active_band_count":3,"active_band_mae_db":3.7524314133372023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":3.77874026776979,"rms_envelope_nrmse":0.30682725643436326,"envelope_correlation":0.9552175750547172,"rms_gain_error_db":-2.1765043560143487,"reference_mean_power":0.00024500023514433596,"output_mean_power":0.0001484280786129323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.4888167929489614e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.774520748189045e-05,"output_power":4.306902072124904e-05,"reference_active":true,"delta_db":-3.0905840810490783},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.213495209539379e-06,"output_power":3.225209484725911e-06,"reference_active":true,"delta_db":-4.558664860659543},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5741557318465008e-07,"output_power":6.503971640455586e-08,"reference_active":true,"delta_db":-3.838690561551639}],"active_band_count":3,"active_band_mae_db":3.8293131677534205}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":5.727415892552129,"rms_envelope_nrmse":0.3979804172723682,"envelope_correlation":0.8971945337236186,"rms_gain_error_db":-1.8375824552538496,"reference_mean_power":7.123238480442997e-06,"output_mean_power":4.665726087301811e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":4.409187309389393e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.616739584019063e-06,"output_power":2.0214403933799075e-06,"reference_active":true,"delta_db":-1.120995643963074},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.009142614437226e-07,"output_power":7.900847594779756e-07,"reference_active":true,"delta_db":0.5200878782687443},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642480310119984e-07,"output_power":1.155980965699837e-07,"reference_active":true,"delta_db":-3.59061076857147}],"active_band_count":3,"active_band_mae_db":1.7438980969344293}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":5.138270280687992,"rms_envelope_nrmse":0.34753817440072354,"envelope_correlation":0.9236812326107875,"rms_gain_error_db":-1.8375283999136105,"reference_mean_power":7.1814170088864135e-06,"output_mean_power":4.703891610883676e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":3.9660429491134744e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6674623865712373e-06,"output_power":2.1427016706212196e-06,"reference_active":true,"delta_db":-0.9513659600545855},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2554368602964405e-07,"output_power":3.5528300173543707e-07,"reference_active":true,"delta_db":0.3796515291762499},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7820276752795495e-08,"output_power":1.0133398636258423e-08,"reference_active":true,"delta_db":-4.386063183538673}],"active_band_count":3,"active_band_mae_db":1.9056935575898362}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0003","method":"golf","status":"measured","mos":{"status":"measured","value":2.4919002056121826,"predicted_mos":2.4919002056121826,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88560,"input_sample_rate":24000,"input_seconds":3.69,"input_float32_sha256":"11e5e2670e6a30c8de2b6194e7465e39ebe28d5aa7313b727557954610f036fe","internal_samples":59040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6886207590014237},"started_at":"2026-10-13T05:02:32.769569+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001909730999614112,"native_input_samples":88800,"native_input_sha256":"a6870d452b2761de81c46d5f98bee27eeeb3017ee484aa5f248c3d68baf8a513","output_sha256":"fb28cbec84ab835058a90e9758bc927d38758ad4ada83320051290c42d9ed262","output_samples":88560,"length_delta_samples":-240,"analysis_s":0.4130165720007426,"synthesis_s":0.07851809199928539,"total_s":0.491534664000028,"rtf":0.13284720648649403,"output_peak":0.5045151710510254,"output_rms":0.07054445534574183,"finished_at":"2026-10-13T05:02:33.267875+00:00","analysis_stages":{"external_f0_s":0.039289153000027},"evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"fb28cbec84ab835058a90e9758bc927d38758ad4ada83320051290c42d9ed262","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59200,"output_samples":59040,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.025188609465634,"output_peak":0.5048295302625394},"common_prefix":{"samples":59040,"sample_rate":16000,"reference_coverage":0.9972972972972973,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59040,"output_samples":59040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.627990864561626,"spectral_convergence":0.534638535879346,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.950454985511855,"spectral_convergence":0.29628152813929176,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.311111190646764,"spectral_convergence":0.3448681526469899,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008072490514636618,"output_power":0.0006775877436006989,"delta_db":-0.7604200163149811,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003107195955572832,"output_power":0.0029159458282179617,"delta_db":-0.27589191440247235,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00134314975009495,"output_power":0.0011403327254544,"delta_db":-0.7109284765766315,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028041769453145745,"output_power":0.00024107384478172348,"delta_db":-0.6565532009903373,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.055753020660351e-06,"output_power":5.417741873504634e-06,"delta_db":-1.1472506098543451,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.333189545876911e-06,"output_power":2.530268756454689e-06,"delta_db":-3.9845583331718073,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2559337585517625,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4692149109017818,"pitch":null,"output_rms_dbfs":-23.025188609465634,"output_peak":0.5048295302625394,"reference_rms_dbfs":-22.55597369856385,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.296519013573415},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59040,"output_samples":59040,"intelligibility":{"estoi":{"value":0.8981179479251148,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07450772707750485,"rms_dbfs":-22.55597369856385,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.838052278992222,"mean":-9.764177689384316e-07,"zero_energy":false},"output":{"rms":0.07058957538817939,"rms_dbfs":-23.025188609465634,"peak":0.5048295302625394,"peak_dbfs":-5.937104974844005,"crest_factor_db":17.088083634621626,"mean":0.0006071286116985682,"zero_energy":false},"rms_difference_db":-0.46921491090178336},"full_scale":{"reference":{"sample_count":59040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":369,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":6.516990002377757e-07,"output_mean_power":7.782251895285259e-07,"output_rms_dbfs":-61.088947160969234,"output_zero_energy":false,"output_peak":0.004959010502031681,"output_energy_fraction":2.370204431071607e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":362,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":199,"both_estimated_voiced_frames":186,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9441624365482234,"absolute_difference_cents_median":6.931388846497553,"absolute_difference_cents_p95":72.69219458006617,"difference_over_600_cents_frames":8,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":0.0002672933594000982,"log_rms_mae_db":1.7996617839573203,"rms_envelope_nrmse":0.1537547723862646,"envelope_correlation":0.9721158186662723,"rms_gain_error_db":-0.4692150494479818,"reference_mean_power":0.005558933671180025,"output_mean_power":0.004989648898587784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":1.3301399574231584e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007866828878829635,"output_power":0.0007334428663571685,"reference_active":true,"delta_db":-0.304334145145523},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024104185039215003,"output_power":0.0001995838656375404,"reference_active":true,"delta_db":-0.81967022603211},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.879809272952036e-05,"output_power":3.4393174971086184e-05,"reference_active":true,"delta_db":-1.5193057883355223}],"active_band_count":3,"active_band_mae_db":0.881103386504385}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":731,"rms_floor":0.00019127079187884833,"log_rms_mae_db":1.4645302162391505,"rms_envelope_nrmse":0.11681967259769416,"envelope_correlation":0.9837116843668339,"rms_gain_error_db":-0.46921649243207825,"reference_mean_power":0.005604559964859168,"output_mean_power":0.005030600978232539},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":731,"frequency_resolution_hz":0.27359781121751026,"power_floor":1.0939229927966163e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007031050656867308,"output_power":0.0006648886268839282,"reference_active":true,"delta_db":-0.24271322644214777},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013134020913854445,"output_power":0.00010450448505838171,"reference_active":true,"delta_db":-0.9926277387046358},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.764508446762904e-06,"output_power":2.1573913854702886e-06,"reference_active":true,"delta_db":-1.076889809843208}],"active_band_count":3,"active_band_mae_db":0.7707435916633306}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":0.0002516933783769556,"log_rms_mae_db":1.7546308401787087,"rms_envelope_nrmse":0.20252496318638072,"envelope_correlation":0.9558581156151882,"rms_gain_error_db":-0.2999329200266475,"reference_mean_power":0.003619462593510694,"output_mean_power":0.0033779312061938886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":1.0382309890148466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005763719641141929,"output_power":0.0004958539510480825,"reference_active":true,"delta_db":-0.6534906958793254},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000211633704279651,"output_power":0.00016265447198116897,"reference_active":true,"delta_db":-1.143188254106307},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9388908760776776e-05,"output_power":2.859307603322309e-05,"reference_active":true,"delta_db":-2.373685514862745}],"active_band_count":3,"active_band_mae_db":1.3901214882827926}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":731,"rms_floor":0.00017703062295034008,"log_rms_mae_db":1.302821501868345,"rms_envelope_nrmse":0.156487432712364,"envelope_correlation":0.9707594194085417,"rms_gain_error_db":-0.2999328645141237,"reference_mean_power":0.003649170714206805,"output_mean_power":0.003405656909373898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":731,"frequency_resolution_hz":0.27359781121751026,"power_floor":8.595703470547666e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005231162615635975,"output_power":0.0004562495310294623,"reference_active":true,"delta_db":-0.5939578963045745},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011877754603364058,"output_power":8.555859496798796e-05,"reference_active":true,"delta_db":-1.4247070439721445},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6948027542655703e-06,"output_power":1.7334178349257842e-06,"reference_active":true,"delta_db":-1.9162372190431987}],"active_band_count":3,"active_band_mae_db":1.311634053106639}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":1.767570976780507,"rms_envelope_nrmse":0.17782184493487,"envelope_correlation":0.9767292448923802,"rms_gain_error_db":-0.4232662325387843,"reference_mean_power":0.00024367065721417694,"output_mean_power":0.00022104292121038398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":1.5844990204196824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.088357418553946e-05,"output_power":8.983287854803967e-05,"reference_active":true,"delta_db":-0.05050082074209413},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.789563717092339e-05,"output_power":1.6409992465355255e-05,"reference_active":true,"delta_db":-0.37638784330838926},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318189450157049e-06,"output_power":2.6008461837055137e-06,"reference_active":true,"delta_db":0.4996574333542191}],"active_band_count":3,"active_band_mae_db":0.3088486991349008}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":731,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":1.3235486785630457,"rms_envelope_nrmse":0.12082421643272072,"envelope_correlation":0.9896544778069793,"rms_gain_error_db":-0.4232665303277322,"reference_mean_power":0.00024567047757912037,"output_mean_power":0.00022285701904594478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":731,"frequency_resolution_hz":0.27359781121751026,"power_floor":1.4837407837384884e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.681784466034025e-05,"output_power":8.522371305367879e-05,"reference_active":true,"delta_db":-0.08048548086514797},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.174061720690315e-06,"output_power":8.496122715365733e-06,"reference_active":true,"delta_db":-0.3334088102788108},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5629145157819176e-07,"output_power":1.6538033998505189e-07,"reference_active":true,"delta_db":0.24548655644081405}],"active_band_count":3,"active_band_mae_db":0.21979361586159094}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":737,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":2.8562029949242733,"rms_envelope_nrmse":0.36913784139963773,"envelope_correlation":0.9295110680672308,"rms_gain_error_db":-2.6290692110318465,"reference_mean_power":7.142520066511995e-06,"output_mean_power":3.898922009837009e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":737,"frequency_resolution_hz":0.27137042062415195,"power_floor":4.426574064670248e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6339845783504805e-06,"output_power":1.2043206933614957e-06,"reference_active":true,"delta_db":-3.398710791675687},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.955023273159077e-07,"output_power":4.536432036351264e-07,"reference_active":true,"delta_db":-1.855841788082437},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6454222889756603e-07,"output_power":1.4916264748927973e-07,"reference_active":true,"delta_db":-2.4883492531398343}],"active_band_count":3,"active_band_mae_db":2.5809672776326527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59040,"uncovered_tail_samples":0,"metrics":{"frame_count":731,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":2.6894810785100183,"rms_envelope_nrmse":0.3249284332538057,"envelope_correlation":0.9565702931633497,"rms_gain_error_db":-2.629082720754256,"reference_mean_power":7.201017761210024e-06,"output_mean_power":3.930842201861427e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":731,"frequency_resolution_hz":0.27359781121751026,"power_floor":3.984026636154084e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6781593686441656e-06,"output_power":1.2858441737387795e-06,"reference_active":true,"delta_db":-3.186480755287012},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.245665915760165e-07,"output_power":2.1249194128390767e-07,"reference_active":true,"delta_db":-1.8396135064595573},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7680389819024547e-08,"output_power":1.6598435870677952e-08,"reference_active":true,"delta_db":-2.2210503707267346}],"active_band_count":3,"active_band_mae_db":2.4157148774911015}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0003","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.9719150066375732,"predicted_mos":2.9719150066375732,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88920,"input_sample_rate":24000,"input_seconds":3.705,"input_float32_sha256":"8cef0152dd054797ed47696cb7e632afca1ce2a2a84da2fa04d09d3a50ad2464","internal_samples":59280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4263388910003414},"started_at":"2026-10-13T07:37:05.332919+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002591458000097191,"native_input_samples":88800,"native_input_sha256":"a6870d452b2761de81c46d5f98bee27eeeb3017ee484aa5f248c3d68baf8a513","output_sha256":"6f1dd048aa365a63a157d9cc8f10612b0d27c6959056d98dd9d4eb8ee842677d","output_samples":88920,"length_delta_samples":120,"analysis_s":1.3378011870008777,"synthesis_s":8.156902593000268,"total_s":9.494703780001146,"rtf":2.566136156757066,"output_peak":0.4308042526245117,"output_rms":0.0717450012109322,"finished_at":"2026-10-13T07:37:14.841151+00:00","analysis_stages":{"external_f0_s":0.8439379070005089},"evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"6f1dd048aa365a63a157d9cc8f10612b0d27c6959056d98dd9d4eb8ee842677d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59200,"output_samples":59280,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.878777077657702,"output_peak":0.4306084632453375},"common_prefix":{"samples":59200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986504723346828,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.761220694529332,"spectral_convergence":0.520463722747013,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.988865081871026,"spectral_convergence":0.26632078875821347,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.140291585623241,"spectral_convergence":0.26595244366093806,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.0008094874060880067,"delta_db":-0.048569373445753704,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.0029127096868389727,"delta_db":-0.25070800183515635,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.001190583191198717,"delta_db":-0.5111461566743895,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.00023602723259119688,"delta_db":-0.7321079784332634,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":7.269826179605418e-06,"delta_db":0.13904739013372974,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":4.624615586547492e-06,"delta_db":-1.3493911195676656,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.505161670014993,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3051852884741095,"pitch":null,"output_rms_dbfs":-22.87291221636069,"output_peak":0.4306084632453375,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.296792454007867},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59200,"output_samples":59200,"intelligibility":{"estoi":{"value":0.9122567686340202,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07440697583274426,"rms_dbfs":-22.567726927886582,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.84980550831495,"mean":-3.489932498416385e-07,"zero_energy":false},"output":{"rms":0.07183802585318617,"rms_dbfs":-22.87291221636069,"peak":0.4306084632453375,"peak_dbfs":-7.318348774754626,"crest_factor_db":15.554563441606064,"mean":-6.270202532398326e-06,"zero_energy":false},"rms_difference_db":-0.30518528847410664},"full_scale":{"reference":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":370,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":6.431599327346735e-07,"output_mean_power":2.991035673292626e-06,"output_rms_dbfs":-55.24178407222193,"output_zero_energy":false,"output_peak":0.06637198265047868,"output_energy_fraction":8.928653020421078e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":363,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":195,"both_estimated_voiced_frames":186,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9441624365482234,"absolute_difference_cents_median":5.415496446314788,"absolute_difference_cents_p95":36.060480015861685,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002672933594000982,"log_rms_mae_db":1.7053784311045588,"rms_envelope_nrmse":0.15778535109706707,"envelope_correlation":0.9667872055783091,"rms_gain_error_db":-0.3051852009838123,"reference_mean_power":0.0055438895852693145,"output_mean_power":0.005167685225121271},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3303693041440345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007860073564586987,"output_power":0.0007480645674591495,"reference_active":true,"delta_db":-0.2148752613594991},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002436436408779341,"output_power":0.0002195786617888161,"reference_active":true,"delta_db":-0.4516494677085402},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.963914121363999e-05,"output_power":4.922998239220649e-05,"reference_active":true,"delta_db":-0.03594578900256081}],"active_band_count":3,"active_band_mae_db":0.2341568393568667}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00019127079187884833,"log_rms_mae_db":1.2691692711456002,"rms_envelope_nrmse":0.10012319947993602,"envelope_correlation":0.985580260849699,"rms_gain_error_db":-0.30518483582365036,"reference_mean_power":0.005589268172913711,"output_mean_power":0.005209984892505875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.0930617045795466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007030365316665855,"output_power":0.0006651445414412802,"reference_active":true,"delta_db":-0.2406186135185837},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013359627830049331,"output_power":0.00012130021676204558,"reference_active":true,"delta_db":-0.4193278286928873},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.77873293352624e-06,"output_power":2.68404683478097e-06,"reference_active":true,"delta_db":-0.15056718652631215}],"active_band_count":3,"active_band_mae_db":0.27017120957926105}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002516933783769556,"log_rms_mae_db":1.9722786868384348,"rms_envelope_nrmse":0.19442714134318714,"envelope_correlation":0.9577799383800404,"rms_gain_error_db":-0.3531958685929784,"reference_mean_power":0.003609667084718048,"output_mean_power":0.003327726019170527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0389022511376814e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005770551731550911,"output_power":0.0005281771951394826,"reference_active":true,"delta_db":-0.38437692775873805},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021321841152816334,"output_power":0.00019097517527517867,"reference_active":true,"delta_db":-0.47847786252099467},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0182675609825546e-05,"output_power":4.522302417703145e-05,"reference_active":true,"delta_db":-0.4519421170731042}],"active_band_count":3,"active_band_mae_db":0.43826563578427896}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00017703062295034008,"log_rms_mae_db":1.3918756023808845,"rms_envelope_nrmse":0.12041155943134917,"envelope_correlation":0.9824326674640973,"rms_gain_error_db":-0.3531957833256394,"reference_mean_power":0.003639213964400563,"output_mean_power":0.003354965140054876},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":8.593451292498865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005240710435115443,"output_power":0.0004772381791997159,"reference_active":true,"delta_db":-0.4065498422089068},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012044100626957098,"output_power":0.0001082947000926386,"reference_active":true,"delta_db":-0.46167172417202135},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7064893260916713e-06,"output_power":2.430028104013685e-06,"reference_active":true,"delta_db":-0.46795022275883996}],"active_band_count":3,"active_band_mae_db":0.4453905963799227}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":1.9662266752109072,"rms_envelope_nrmse":0.1841494259652298,"envelope_correlation":0.97536205453085,"rms_gain_error_db":-0.5592352934423506,"reference_mean_power":0.00024301128670702196,"output_mean_power":0.0002136500090823084},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.5901124046586961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178189042095749e-05,"output_power":7.725545284893973e-05,"reference_active":true,"delta_db":-0.7482785598635511},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7997292458206196e-05,"output_power":1.4640320878729045e-05,"reference_active":true,"delta_db":-0.8965657859192065},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.335696736133691e-06,"output_power":2.3114914228399228e-06,"reference_active":true,"delta_db":-0.04524167604946766}],"active_band_count":3,"active_band_mae_db":0.5633620072774084}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":1.2496677711484694,"rms_envelope_nrmse":0.12661559108075565,"envelope_correlation":0.9894483886702443,"rms_gain_error_db":-0.5592347658295744,"reference_mean_power":0.00024500023514433596,"output_mean_power":0.00021539867358519917},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.4888167929489614e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.774520748189045e-05,"output_power":7.339675099204023e-05,"reference_active":true,"delta_db":-0.7754656954031114},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.213495209539379e-06,"output_power":7.467858146623567e-06,"reference_active":true,"delta_db":-0.9122835452272761},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5741557318465008e-07,"output_power":1.6213112330887348e-07,"reference_active":true,"delta_db":0.12818696604197083}],"active_band_count":3,"active_band_mae_db":0.605312068890786}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":2.16763775345957,"rms_envelope_nrmse":0.28063528309229485,"envelope_correlation":0.9388875907983512,"rms_gain_error_db":-0.4445150194228186,"reference_mean_power":7.123238480442997e-06,"output_mean_power":6.430222229720526e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":4.409187309389393e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.616739584019063e-06,"output_power":2.425254123180485e-06,"reference_active":true,"delta_db":-0.33003252522614257},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.009142614437226e-07,"output_power":6.141906938312549e-07,"reference_active":true,"delta_db":-0.5736166486555038},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642480310119984e-07,"output_power":1.5054480968556962e-07,"reference_active":true,"delta_db":-2.443459729477892}],"active_band_count":3,"active_band_mae_db":1.115702967786513}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":1.4361093412352113,"rms_envelope_nrmse":0.1688987689000233,"envelope_correlation":0.9766931794047088,"rms_gain_error_db":-0.44451905840965067,"reference_mean_power":7.1814170088864135e-06,"output_mean_power":6.482734569688432e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":3.9660429491134744e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6674623865712373e-06,"output_power":2.4138285438575645e-06,"reference_active":true,"delta_db":-0.433918856500215},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2554368602964405e-07,"output_power":2.7321970629161237e-07,"reference_active":true,"delta_db":-0.7609725641784438},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7820276752795495e-08,"output_power":1.4825497009622432e-08,"reference_active":true,"delta_db":-2.733521844286124}],"active_band_count":3,"active_band_mae_db":1.3094710883215943}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0003","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.885375738143921,"predicted_mos":2.885375738143921,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89040,"input_sample_rate":24000,"input_seconds":3.71,"input_float32_sha256":"f05e313cb357952f10a3746c37cae54fdd6ac5aa3b3f688dc1137ab470cfa3f2","internal_samples":59360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.322318129999985},"started_at":"2026-10-13T05:10:41.956375+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00231618699945102,"native_input_samples":88800,"native_input_sha256":"a6870d452b2761de81c46d5f98bee27eeeb3017ee484aa5f248c3d68baf8a513","output_sha256":"309c99690d86986e5af62228e92d1df44e1df8ff30633dc644479739b3d7db89","output_samples":89040,"length_delta_samples":240,"analysis_s":0.5542548550001811,"synthesis_s":0.09073965300012787,"total_s":0.644994508000309,"rtf":0.1743228400000835,"output_peak":0.5107382535934448,"output_rms":0.0729753257610858,"finished_at":"2026-10-13T05:10:42.617598+00:00","analysis_stages":{"external_f0_s":0.0778770760007319},"evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"309c99690d86986e5af62228e92d1df44e1df8ff30633dc644479739b3d7db89","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59200,"output_samples":59360,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.730989286253532,"output_peak":0.5109231640513998},"common_prefix":{"samples":59200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973045822102425,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.688244858189599,"spectral_convergence":0.18514114078761823,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.459443235126768,"spectral_convergence":0.18900708659434548,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.965872751948313,"spectral_convergence":0.2360289589154623,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.0007896950114224067,"delta_db":-0.15607639240234325,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.0029736650183711053,"delta_db":-0.16075956410928594,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.0013086815452087622,"delta_db":-0.10040384389698599,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.00026310336649312317,"delta_db":-0.26046507116488343,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":6.21626357272428e-06,"delta_db":-0.5408986756812442,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":5.184132035593263e-06,"delta_db":-0.8533869640026647,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3453317518762346,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15154053348986402,"pitch":null,"output_rms_dbfs":-22.719267461376443,"output_peak":0.5109231640513998,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.704520281754893},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59200,"output_samples":59200,"intelligibility":{"estoi":{"value":0.9599711882321728,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07440697583274426,"rms_dbfs":-22.567726927886582,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.84980550831495,"mean":-3.489932498416385e-07,"zero_energy":false},"output":{"rms":0.07312007478869066,"rms_dbfs":-22.719267461376447,"peak":0.5109231640513998,"peak_dbfs":-5.832888139674393,"crest_factor_db":16.88637932170205,"mean":-8.97669484915257e-07,"zero_energy":false},"rms_difference_db":-0.15154053348986452},"full_scale":{"reference":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":370,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":6.431599327346735e-07,"output_mean_power":1.1608298448633254e-06,"output_rms_dbfs":-59.35231434742586,"output_zero_energy":false,"output_peak":0.015512411033082013,"output_energy_fraction":3.344786818267373e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":363,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":179,"both_estimated_voiced_frames":176,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.8934010152284264,"absolute_difference_cents_median":1.1898505893252147,"absolute_difference_cents_p95":7.254417312363418,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002672933594000982,"log_rms_mae_db":0.6908203797483818,"rms_envelope_nrmse":0.06582325427408252,"envelope_correlation":0.9944653783476767,"rms_gain_error_db":-0.15154048070478046,"reference_mean_power":0.0055438895852693145,"output_mean_power":0.0053537800371876495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3303693041440345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007860073564586987,"output_power":0.0007596825961728259,"reference_active":true,"delta_db":-0.1479443362160873},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002436436408779341,"output_power":0.00024228299435511476,"reference_active":true,"delta_db":-0.0243214830484576},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.963914121363999e-05,"output_power":4.890646699923893e-05,"reference_active":true,"delta_db":-0.06457968799021327}],"active_band_count":3,"active_band_mae_db":0.07894850241825273}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00019127079187884833,"log_rms_mae_db":0.4473118639722776,"rms_envelope_nrmse":0.041902166757367264,"envelope_correlation":0.9976773106514469,"rms_gain_error_db":-0.15154039489189403,"reference_mean_power":0.005589268172913711,"output_mean_power":0.005397602621345697},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.0930617045795466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007030365316665855,"output_power":0.0006776707176359843,"reference_active":true,"delta_db":-0.1595917266915639},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013359627830049331,"output_power":0.00013768400792525025,"reference_active":true,"delta_db":0.13089139822911838},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.77873293352624e-06,"output_power":3.1549882040493387e-06,"reference_active":true,"delta_db":0.5515093147457342}],"active_band_count":3,"active_band_mae_db":0.28066414655547217}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002516933783769556,"log_rms_mae_db":0.8320663480123704,"rms_envelope_nrmse":0.08358052327817439,"envelope_correlation":0.992435812940894,"rms_gain_error_db":-0.1582333202777889,"reference_mean_power":0.003609667084718048,"output_mean_power":0.0034805174642088205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0389022511376814e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005770551731550911,"output_power":0.0005472390317009324,"reference_active":true,"delta_db":-0.23040272948106183},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021321841152816334,"output_power":0.00021159939546412684,"reference_active":true,"delta_db":-0.03310280950323464},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0182675609825546e-05,"output_power":4.832034299120526e-05,"reference_active":true,"delta_db":-0.16423804682055368}],"active_band_count":3,"active_band_mae_db":0.1425811952682834}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00017703062295034008,"log_rms_mae_db":0.48938445357292976,"rms_envelope_nrmse":0.05037566382032647,"envelope_correlation":0.9970852439581309,"rms_gain_error_db":-0.15823329791104682,"reference_mean_power":0.003639213964400563,"output_mean_power":0.003509007209542018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":8.593451292498865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005240710435115443,"output_power":0.0004922349337917336,"reference_active":true,"delta_db":-0.2721773206716326},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012044100626957098,"output_power":0.00012552933266206002,"reference_active":true,"delta_db":0.1797084470637354},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7064893260916713e-06,"output_power":3.052894754955627e-06,"reference_active":true,"delta_db":0.5230551410174807}],"active_band_count":3,"active_band_mae_db":0.3249803029176162}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":1.0203936396588296,"rms_envelope_nrmse":0.0954715716598766,"envelope_correlation":0.9936784408331499,"rms_gain_error_db":-0.29631741595499606,"reference_mean_power":0.00024301128670702196,"output_mean_power":0.00022698371688823228},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.5901124046586961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178189042095749e-05,"output_power":8.78452125050551e-05,"reference_active":true,"delta_db":-0.1903890079507077},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7997292458206196e-05,"output_power":1.7055006351498058e-05,"reference_active":true,"delta_db":-0.23355288572816452},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.335696736133691e-06,"output_power":1.6092543752213358e-06,"reference_active":true,"delta_db":-1.6179175530543035}],"active_band_count":3,"active_band_mae_db":0.6806198155777253}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":0.6654448079578456,"rms_envelope_nrmse":0.07082951068483029,"envelope_correlation":0.9967403027794001,"rms_gain_error_db":-0.2963165211620102,"reference_mean_power":0.00024500023514433596,"output_mean_power":0.00022884153334142146},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.4888167929489614e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.774520748189045e-05,"output_power":8.367458976009235e-05,"reference_active":true,"delta_db":-0.20629813464944852},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.213495209539379e-06,"output_power":8.731070710595155e-06,"reference_active":true,"delta_db":-0.2335690890394651},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5741557318465008e-07,"output_power":9.906156067881176e-08,"reference_active":true,"delta_db":-2.011425292208526}],"active_band_count":3,"active_band_mae_db":0.8170975052991465}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":1.362170277206766,"rms_envelope_nrmse":0.1623045789058009,"envelope_correlation":0.9831241205969438,"rms_gain_error_db":-0.6906325315045929,"reference_mean_power":7.123238480442997e-06,"output_mean_power":6.075950558948675e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":4.409187309389393e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.616739584019063e-06,"output_power":2.361924348671013e-06,"reference_active":true,"delta_db":-0.444945208742116},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.009142614437226e-07,"output_power":5.449661580333867e-07,"reference_active":true,"delta_db":-1.092953628560824},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642480310119984e-07,"output_power":1.1771102375626255e-07,"reference_active":true,"delta_db":-3.511946230736437}],"active_band_count":3,"active_band_mae_db":1.683281689346459}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":0.985768550960416,"rms_envelope_nrmse":0.11518423662791848,"envelope_correlation":0.9931451892271392,"rms_gain_error_db":-0.6906090077100251,"reference_mean_power":7.1814170088864135e-06,"output_mean_power":6.125608619616397e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":3.9660429491134744e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6674623865712373e-06,"output_power":2.324282587698044e-06,"reference_active":true,"delta_db":-0.5980937551079764},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2554368602964405e-07,"output_power":2.259293163109565e-07,"reference_active":true,"delta_db":-1.5863668845954413},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7820276752795495e-08,"output_power":1.0100445284444234e-08,"reference_active":true,"delta_db":-4.400209256388264}],"active_band_count":3,"active_band_mae_db":2.194889965363894}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0003","method":"magphase","status":"measured","mos":{"status":"measured","value":2.9713354110717773,"predicted_mos":2.9713354110717773,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":177666,"input_sample_rate":48000,"input_seconds":3.701375,"input_float32_sha256":"0e9d5b167cce0592b495642589acafa39a0e7bf5b19a2ce5e0f97ef74ed40550","internal_samples":59222,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.290383427998677},"started_at":"2026-10-13T05:15:20.288349+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006285127999944962,"native_input_samples":177600,"native_input_sha256":"09d5e1fd9236b287d9237170acc21699c4ed827416c10ef92a605b651d08e062","output_sha256":"0350ad61978e986fe3b6150ea512a684a638a117ebe54cce951a9f01167e9087","output_samples":177666,"length_delta_samples":66,"analysis_s":2.1532172330007597,"synthesis_s":0.8521073110005091,"total_s":3.005324544001269,"rtf":0.8122498767570996,"output_peak":0.5299366417938005,"output_rms":0.07198778940484293,"finished_at":"2026-10-13T05:15:23.308203+00:00","evaluation_input":{"reference":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"0350ad61978e986fe3b6150ea512a684a638a117ebe54cce951a9f01167e9087","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":59200,"output_samples":59222,"length_delta_samples":22,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.849281607476325,"output_peak":0.5302811592808054},"common_prefix":{"samples":59200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996285164297052,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59200,"output_samples":59200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.359768630641802,"spectral_convergence":0.4160927889728603,"spectral_convergence_status":"ok","amplitude_floor":2.0296349328313604e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.464847232306344,"spectral_convergence":0.10865098847383023,"spectral_convergence_status":"ok","amplitude_floor":1.6850709272784723e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.977770731916484,"spectral_convergence":0.12657887305314477,"spectral_convergence_status":"ok","amplitude_floor":8.273401155700601e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185911287346583,"output_power":0.000741723391125389,"delta_db":-0.4282505856443178,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030858016031558087,"output_power":0.002859467500326166,"delta_db":-0.3308283545562948,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013392891870332756,"output_power":0.0012996408920661368,"delta_db":-0.13050994918374062,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00027936558999729056,"output_power":0.0002758961040657513,"delta_db":-0.054273443351875506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.040755904577448e-06,"output_power":7.744542229838316e-06,"delta_db":0.41376464195253065,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.309787748977687e-06,"output_power":6.314499848593758e-06,"delta_db":0.0032420664384559417,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2268115068545359,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2799410479584533,"pitch":null,"output_rms_dbfs":-22.847667975845035,"output_peak":0.5302811592808054,"reference_rms_dbfs":-22.567726927886582,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.934128864954877},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59200,"output_samples":59200,"intelligibility":{"estoi":{"value":0.972536378996998,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07440697583274426,"rms_dbfs":-22.567726927886582,"peak":0.517730712890625,"peak_dbfs":-5.717921419571628,"crest_factor_db":16.84980550831495,"mean":-3.489932498416385e-07,"zero_energy":false},"output":{"rms":0.0720471160398657,"rms_dbfs":-22.84766797584503,"peak":0.5302811592808054,"peak_dbfs":-5.5098760590143385,"crest_factor_db":17.3377919168307,"mean":-0.0013771530990809447,"zero_energy":false},"rms_difference_db":-0.27994104795844876},"full_scale":{"reference":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":370,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.144573997939005e-06,"threshold_dbfs":-51.46023661325071,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":6.431599327346735e-07,"output_mean_power":5.80998665927593e-07,"output_rms_dbfs":-62.35824864823008,"output_zero_energy":false,"output_peak":0.0075716409108683895,"output_energy_fraction":1.724308878383745e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":363,"reference_estimated_voiced_frames":197,"output_estimated_voiced_frames":200,"both_estimated_voiced_frames":192,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9746192893401016,"absolute_difference_cents_median":2.442426467306099,"absolute_difference_cents_p95":9.001969063409863,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002672933594000982,"log_rms_mae_db":0.9687158257070663,"rms_envelope_nrmse":0.07834665775257367,"envelope_correlation":0.9930374649943188,"rms_gain_error_db":-0.2799409461816918,"reference_mean_power":0.0055438895852693145,"output_mean_power":0.0051978109232434675},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3303693041440345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007860073564586987,"output_power":0.0007611966378784414,"reference_active":true,"delta_db":-0.13929749351336415},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002436436408779341,"output_power":0.00022184063616023783,"reference_active":true,"delta_db":-0.40713978764108144},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.963914121363999e-05,"output_power":5.236121866909775e-05,"reference_active":true,"delta_db":0.23185486541492561}],"active_band_count":3,"active_band_mae_db":0.25943071552312374}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00019127079187884833,"log_rms_mae_db":0.6892439085170889,"rms_envelope_nrmse":0.04880500154366593,"envelope_correlation":0.9979204898514471,"rms_gain_error_db":-0.2799406845651413,"reference_mean_power":0.005589268172913711,"output_mean_power":0.0052403570571419224},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.0930617045795466e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007030365316665855,"output_power":0.0006818032500249421,"reference_active":true,"delta_db":-0.1331882561792849},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013359627830049331,"output_power":0.00012173307006891045,"reference_active":true,"delta_db":-0.4038578489326692},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.77873293352624e-06,"output_power":2.898819648994833e-06,"reference_active":true,"delta_db":0.1837438806606984}],"active_band_count":3,"active_band_mae_db":0.24026332859088417}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.0002516933783769556,"log_rms_mae_db":0.9500361913837184,"rms_envelope_nrmse":0.07744578527630289,"envelope_correlation":0.9942270751018644,"rms_gain_error_db":-0.2682314287991189,"reference_mean_power":0.003609667084718048,"output_mean_power":0.003393469939452444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0389022511376814e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005770551731550911,"output_power":0.0005558446144558306,"reference_active":true,"delta_db":-0.16263936556264624},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00021321841152816334,"output_power":0.00019269042143796797,"reference_active":true,"delta_db":-0.4396457695233451},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0182675609825546e-05,"output_power":4.8309911773167995e-05,"reference_active":true,"delta_db":-0.16517568699582946}],"active_band_count":3,"active_band_mae_db":0.25582027402727364}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":0.00017703062295034008,"log_rms_mae_db":0.7232685359136409,"rms_envelope_nrmse":0.05089467669555305,"envelope_correlation":0.9978393678318407,"rms_gain_error_db":-0.2682314153494113,"reference_mean_power":0.003639213964400563,"output_mean_power":0.003421247151006718},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":8.593451292498865e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005240710435115443,"output_power":0.0005047388112845347,"reference_active":true,"delta_db":-0.16323463729749643},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012044100626957098,"output_power":0.00010916623785727789,"reference_active":true,"delta_db":-0.42686031646821754},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7064893260916713e-06,"output_power":2.541697927148775e-06,"reference_active":true,"delta_db":-0.2728238390729442}],"active_band_count":3,"active_band_mae_db":0.28763959761288604}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":7.943457135434644e-05,"log_rms_mae_db":0.9418931153907927,"rms_envelope_nrmse":0.08019744593895695,"envelope_correlation":0.9950834609124932,"rms_gain_error_db":-0.05820269811807299,"reference_mean_power":0.00024301128670702196,"output_mean_power":0.00023977625630344884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.5901124046586961e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178189042095749e-05,"output_power":9.321175891619906e-05,"reference_active":true,"delta_db":0.06713704673000012},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7997292458206196e-05,"output_power":1.68064790615192e-05,"reference_active":true,"delta_db":-0.2973043529030032},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.335696736133691e-06,"output_power":2.3077976352538127e-06,"reference_active":true,"delta_db":-0.05218729858020242}],"active_band_count":3,"active_band_mae_db":0.13887623273773528}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":6.19528001588506e-05,"log_rms_mae_db":0.5301375603487501,"rms_envelope_nrmse":0.041093038059388035,"envelope_correlation":0.9986582840529968,"rms_gain_error_db":-0.05820263744911086,"reference_mean_power":0.00024500023514433596,"output_mean_power":0.00024173873071106335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":1.4888167929489614e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.774520748189045e-05,"output_power":8.888100428847968e-05,"reference_active":true,"delta_db":0.05585547921119394},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.213495209539379e-06,"output_power":8.652023074317293e-06,"reference_active":true,"delta_db":-0.2730674537193444},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5741557318465008e-07,"output_power":1.418658379718837e-07,"reference_active":true,"delta_db":-0.45169867384646223}],"active_band_count":3,"active_band_mae_db":0.2602072022590002}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":1.5353574631917567e-05,"log_rms_mae_db":1.1455645108460255,"rms_envelope_nrmse":0.19984878887058347,"envelope_correlation":0.9721058686413395,"rms_gain_error_db":0.33046466943310526,"reference_mean_power":7.123238480442997e-06,"output_mean_power":7.686417167358636e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":4.409187309389393e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.616739584019063e-06,"output_power":2.885894683022543e-06,"reference_active":true,"delta_db":0.4251997390779779},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.009142614437226e-07,"output_power":8.271553991423301e-07,"reference_active":true,"delta_db":0.7192221226059613},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.642480310119984e-07,"output_power":2.0999900442534065e-07,"reference_active":true,"delta_db":-0.9979452412656904}],"active_band_count":3,"active_band_mae_db":0.7141223676498765}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59200,"uncovered_tail_samples":0,"metrics":{"frame_count":733,"rms_floor":1.2828806049648866e-05,"log_rms_mae_db":0.7450677915205399,"rms_envelope_nrmse":0.12555566269337792,"envelope_correlation":0.9884673258597239,"rms_gain_error_db":0.33044914895406785,"reference_mean_power":7.1814170088864135e-06,"output_mean_power":7.74916772304569e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":733,"frequency_resolution_hz":0.2728512960436562,"power_floor":3.9660429491134744e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6674623865712373e-06,"output_power":2.9246903166596677e-06,"reference_active":true,"delta_db":0.3998158298146217},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2554368602964405e-07,"output_power":3.732246689858109e-07,"reference_active":true,"delta_db":0.5936106497736094},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7820276752795495e-08,"output_power":1.804682603251429e-08,"reference_active":true,"delta_db":-1.8796061413742526}],"active_band_count":3,"active_band_mae_db":0.957677540320828}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"hnm","status":"measured","mos":{"status":"measured","value":3.0285823345184326,"predicted_mos":3.0285823345184326,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37200,"input_sample_rate":16000,"input_seconds":2.325,"input_float32_sha256":"821e24a6ae0a575960ea180801faba1a6078869712a643d6bc3d102c73572b91","internal_samples":37200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.742115257000478},"started_at":"2026-10-13T04:39:14.024215+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.1248999726376496e-05,"native_input_samples":37200,"native_input_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","output_sha256":"ec8988b488aa4adbdea2790d357d92c88c00cfae65102a75ed4c93e1d3767102","output_samples":37200,"length_delta_samples":0,"analysis_s":0.5379390799998873,"synthesis_s":0.34725271799925395,"total_s":0.8851917979991413,"rtf":0.38072765505339407,"output_peak":0.4749742704410223,"output_rms":0.06385586600372009,"finished_at":"2026-10-13T04:39:14.912881+00:00","evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"ec8988b488aa4adbdea2790d357d92c88c00cfae65102a75ed4c93e1d3767102","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.524011675500207,"spectral_convergence":0.1506016082297362,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.454950368919398,"spectral_convergence":0.1483837693120489,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.880799909039295,"spectral_convergence":0.14759244450813339,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.0004473661518484708,"delta_db":-0.035681694061147115,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0019080801881227205,"delta_db":-0.06697611063410672,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0013968328962610525,"delta_db":0.056760436869120814,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.0002888885932312363,"delta_db":-0.44091882105443414,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.5095138897030074e-05,"delta_db":0.07668049578529067,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":2.1308654724545927e-05,"delta_db":0.20392834363865991,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1468243170071266,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04749336940765911,"pitch":null,"output_rms_dbfs":-23.89598401780653,"output_peak":0.4749742704410223,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.619920651152967},"common_prefix":{"samples":37200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.524011675500207,"spectral_convergence":0.1506016082297362,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.454950368919398,"spectral_convergence":0.1483837693120489,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.880799909039295,"spectral_convergence":0.14759244450813339,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.0004473661518484708,"delta_db":-0.035681694061147115,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0019080801881227205,"delta_db":-0.06697611063410672,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0013968328962610525,"delta_db":0.056760436869120814,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.0002888885932312363,"delta_db":-0.44091882105443414,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.5095138897030074e-05,"delta_db":0.07668049578529067,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":2.1308654724545927e-05,"delta_db":0.20392834363865991,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1468243170071266,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04749336940765911,"pitch":null,"output_rms_dbfs":-23.89598401780653,"output_peak":0.4749742704410223,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.619920651152967},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37200,"output_samples":37200,"intelligibility":{"estoi":{"value":0.9426332355611222,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0642059782880167,"rms_dbfs":-23.848490648398872,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.062475171735503,"mean":-3.708049815188172e-07,"zero_energy":false},"output":{"rms":0.06385586600372009,"rms_dbfs":-23.89598401780653,"peak":0.4749742704410223,"peak_dbfs":-6.466598313108399,"crest_factor_db":17.429385704698134,"mean":1.0158625851662016e-05,"zero_energy":false},"rms_difference_db":-0.04749336940765758},"full_scale":{"reference":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":4.06875659899601e-07,"output_rms_dbfs":-63.905382897177844,"output_zero_energy":false,"output_peak":0.006029194018894276,"output_energy_fraction":2.7896550938324918e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":100,"both_estimated_voiced_frames":98,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9158878504672897,"absolute_difference_cents_median":1.140317918778974,"absolute_difference_cents_p95":6.151771194437692,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00019668187909197959,"log_rms_mae_db":0.7326557279496709,"rms_envelope_nrmse":0.055201717712744056,"envelope_correlation":0.9971871981810786,"rms_gain_error_db":-0.04749337496788614,"reference_mean_power":0.004131292024738224,"output_mean_power":0.004086359366637919},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.0026327800860698e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011612269501705267,"output_power":0.0011839969653338388,"reference_active":true,"delta_db":0.08433482723196224},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017095886275955477,"output_power":0.00016897564655039124,"reference_active":true,"delta_db":-0.050675033258173324},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217923048678998e-05,"output_power":4.122953055427633e-05,"reference_active":true,"delta_db":-0.09890262369012426}],"active_band_count":3,"active_band_mae_db":0.07797082806008661}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00018115923356648065,"log_rms_mae_db":0.393867482981731,"rms_envelope_nrmse":0.028502511278406622,"envelope_correlation":0.9992277164657468,"rms_gain_error_db":-0.04749330670172974,"reference_mean_power":0.004185412442455258,"output_mean_power":0.0041398912262550465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.7728315536386115e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066252574844359,"output_power":0.0010842565372187133,"reference_active":true,"delta_db":0.07271956350035176},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.440983264620606e-05,"output_power":8.642395789256233e-05,"reference_active":true,"delta_db":0.10241112262619556},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.940091323355611e-06,"output_power":3.3019360778165677e-06,"reference_active":true,"delta_db":0.5040784108416941}],"active_band_count":3,"active_band_mae_db":0.22640303232274714}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.000176021246464811,"log_rms_mae_db":1.0215626299136287,"rms_envelope_nrmse":0.06499308557326788,"envelope_correlation":0.9966452892819535,"rms_gain_error_db":-0.061522363247082414,"reference_mean_power":0.0026395436088586166,"output_mean_power":0.0026024153098454295},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.515899729654865e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009091516678957249,"output_power":0.0009086558466169998,"reference_active":true,"delta_db":-0.0023691446864578533},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001833276731517994,"output_power":0.00016850585117149548,"reference_active":true,"delta_db":-0.36613040437877464},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7517429693718724e-05,"output_power":2.5496010723394406e-05,"reference_active":true,"delta_db":-1.6776084423360984}],"active_band_count":3,"active_band_mae_db":0.6820359971337769}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00016136314602147107,"log_rms_mae_db":0.5054944184225015,"rms_envelope_nrmse":0.03157713521807561,"envelope_correlation":0.9991977847838345,"rms_gain_error_db":-0.0615223620505842,"reference_mean_power":0.002674122622482191,"output_mean_power":0.002636507929519759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.348456388983372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008380618007305507,"output_power":0.000833918395495765,"reference_active":true,"delta_db":-0.021524916592149425},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.026333057603631e-05,"output_power":8.743309389312518e-05,"reference_active":true,"delta_db":-0.1383550773478387},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.498114091574834e-06,"output_power":1.8050120975401834e-06,"reference_active":true,"delta_db":-1.4113215221314481}],"active_band_count":3,"active_band_mae_db":0.5237338386904787}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":1.5042440093668323,"rms_envelope_nrmse":0.1180386212090114,"envelope_correlation":0.9910267425448435,"rms_gain_error_db":-0.39864423457340115,"reference_mean_power":0.0003102146198149335,"output_mean_power":0.00028300743028564153},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.2470137873341424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001843563097177963,"output_power":0.00016583051917234588,"reference_active":true,"delta_db":-0.45993545812596803},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0328682781544793e-05,"output_power":1.452425406310442e-05,"reference_active":true,"delta_db":-1.4601540184893218},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.580926194982453e-06,"output_power":2.750467547217404e-06,"reference_active":true,"delta_db":-2.2154676967850637}],"active_band_count":3,"active_band_mae_db":1.378519057800118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":0.9182687612591806,"rms_envelope_nrmse":0.0806130192496037,"envelope_correlation":0.9964638059004326,"rms_gain_error_db":-0.3986441073249309,"reference_mean_power":0.00031427821970459705,"output_mean_power":0.00028671464295409874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":3.052275331975407e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001756954599460375,"output_power":0.00015776735909199104,"reference_active":true,"delta_db":-0.4674338343701343},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.589536328721272e-06,"output_power":5.88075618448408e-06,"reference_active":true,"delta_db":-1.6453654686491053},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8479520444652186e-07,"output_power":1.4189509403712637e-07,"reference_active":true,"delta_db":-3.025652919448953}],"active_band_count":3,"active_band_mae_db":1.7128174074893974}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":1.7475725381998497,"rms_envelope_nrmse":0.20880223710020124,"envelope_correlation":0.9703265632291882,"rms_gain_error_db":0.06766671236258515,"reference_mean_power":1.9843186142044422e-05,"output_mean_power":2.015478072585438e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.622000222111226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.352365504647437e-06,"output_power":9.195092271129366e-06,"reference_active":true,"delta_db":-0.07365379708875125},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.791249366368626e-07,"output_power":8.392974319524817e-07,"reference_active":true,"delta_db":-0.6692221716683923},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.495952213974191e-07,"output_power":5.179619281547722e-07,"reference_active":true,"delta_db":3.17061572551384}],"active_band_count":3,"active_band_mae_db":1.3044972314236611}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":1.4606054391865253,"rms_envelope_nrmse":0.13933814018728127,"envelope_correlation":0.9863098645335209,"rms_gain_error_db":0.06766496385861123,"reference_mean_power":2.010291979320634e-05,"output_mean_power":2.041858471500353e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.5131001811961107e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.175181026763142e-06,"output_power":8.25108761745742e-06,"reference_active":true,"delta_db":0.04013820648781549},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.283226898133564e-07,"output_power":3.5526276652576393e-07,"reference_active":true,"delta_db":-0.8122138772130244},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.754857048422836e-08,"output_power":3.009511387207431e-08,"reference_active":true,"delta_db":2.34254246535109}],"active_band_count":3,"active_band_mae_db":1.0649648496839765}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.6118626594543457,"predicted_mos":2.6118626594543457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37200,"input_sample_rate":16000,"input_seconds":2.325,"input_float32_sha256":"3fc1eb5674c5b60f9feaeabeabe1b0c1c9d2dd57980c5716bdc522bb94380414","internal_samples":37200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.612900015001287},"started_at":"2026-10-13T04:42:10.549015+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.8433000150253065e-05,"native_input_samples":37200,"native_input_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","output_sha256":"361e1d6155e1599bc650fa59dd7956c585058d8b3e8ca48514a6e17736d24cff","output_samples":37200,"length_delta_samples":0,"analysis_s":0.6325523369996517,"synthesis_s":0.042142072000388,"total_s":0.6746944090000397,"rtf":0.29019114365593107,"output_peak":0.5715276287929965,"output_rms":0.07120226061889418,"finished_at":"2026-10-13T04:42:11.227364+00:00","evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"361e1d6155e1599bc650fa59dd7956c585058d8b3e8ca48514a6e17736d24cff","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.068897819813229,"spectral_convergence":0.543955187993849,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.545547808483695,"spectral_convergence":0.23029249450920972,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.938525945458552,"spectral_convergence":0.28118600730421955,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.00048272159107380045,"delta_db":0.2946543301569126,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0024334310534109907,"delta_db":0.9892481257467213,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0017057154283994197,"delta_db":0.9243816702656792,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.00040822510477261643,"delta_db":1.060774330056535,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.73488679147528e-05,"delta_db":0.6810207660975631,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":2.231987166934841e-05,"delta_db":0.4052849526636266,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7258940291645063,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8983662960590223,"pitch":null,"output_rms_dbfs":-22.950124352339852,"output_peak":0.5715276287929965,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.850990524585158},"common_prefix":{"samples":37200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.068897819813229,"spectral_convergence":0.543955187993849,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.545547808483695,"spectral_convergence":0.23029249450920972,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.938525945458552,"spectral_convergence":0.28118600730421955,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.00048272159107380045,"delta_db":0.2946543301569126,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0024334310534109907,"delta_db":0.9892481257467213,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0017057154283994197,"delta_db":0.9243816702656792,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.00040822510477261643,"delta_db":1.060774330056535,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.73488679147528e-05,"delta_db":0.6810207660975631,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":2.231987166934841e-05,"delta_db":0.4052849526636266,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7258940291645063,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8983662960590223,"pitch":null,"output_rms_dbfs":-22.950124352339852,"output_peak":0.5715276287929965,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.850990524585158},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37200,"output_samples":37200,"intelligibility":{"estoi":{"value":0.9319895399954257,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0642059782880167,"rms_dbfs":-23.848490648398872,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.062475171735503,"mean":-3.708049815188172e-07,"zero_energy":false},"output":{"rms":0.07120226061889418,"rms_dbfs":-22.950124352339852,"peak":0.5715276287929965,"peak_dbfs":-4.8592554018811915,"crest_factor_db":18.09086895045866,"mean":4.3162204843691406e-10,"zero_energy":false},"rms_difference_db":0.8983662960590202},"full_scale":{"reference":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":8.714791844037506e-07,"output_rms_dbfs":-60.59742981711246,"output_zero_energy":false,"output_peak":0.009009645522702427,"output_energy_fraction":4.80573537482626e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":107,"both_estimated_voiced_frames":100,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9345794392523364,"absolute_difference_cents_median":6.074985850746126,"absolute_difference_cents_p95":17.280163822284766,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00019668187909197959,"log_rms_mae_db":1.457003710752327,"rms_envelope_nrmse":0.16251508378927138,"envelope_correlation":0.9879309538745435,"rms_gain_error_db":0.8983663439779016,"reference_mean_power":0.004131292024738224,"output_mean_power":0.005080688033770307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.0026327800860698e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011612269501705267,"output_power":0.001361012534604436,"reference_active":true,"delta_db":0.689450184331804},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017095886275955477,"output_power":0.00024961532002869106,"reference_active":true,"delta_db":1.6437961627546742},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217923048678998e-05,"output_power":5.4976360295885036e-05,"reference_active":true,"delta_db":1.15077331783664}],"active_band_count":3,"active_band_mae_db":1.1613398883077062}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00018115923356648065,"log_rms_mae_db":1.0891494077142967,"rms_envelope_nrmse":0.13453222370315993,"envelope_correlation":0.9945412585721624,"rms_gain_error_db":0.898366824639633,"reference_mean_power":0.004185412442455258,"output_mean_power":0.005147246221993522},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.7728315536386115e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066252574844359,"output_power":0.001245181626019056,"reference_active":true,"delta_db":0.6737261064553978},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.440983264620606e-05,"output_power":0.00011831851031703096,"reference_active":true,"delta_db":1.4665965383788726},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.940091323355611e-06,"output_power":4.052180937868045e-06,"reference_active":true,"delta_db":1.3932800880933571}],"active_band_count":3,"active_band_mae_db":1.1778675776425425}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.000176021246464811,"log_rms_mae_db":1.4358307347281458,"rms_envelope_nrmse":0.19670419688340693,"envelope_correlation":0.9873014022886585,"rms_gain_error_db":1.1560341356679318,"reference_mean_power":0.0026395436088586166,"output_mean_power":0.0034445481114180362},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.515899729654865e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009091516678957249,"output_power":0.001138883430731404,"reference_active":true,"delta_db":0.9784293477189565},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001833276731517994,"output_power":0.00030532802929439466,"reference_active":true,"delta_db":2.215386483376069},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7517429693718724e-05,"output_power":4.149131085172154e-05,"reference_active":true,"delta_db":0.4372407852678004}],"active_band_count":3,"active_band_mae_db":1.2103522054542752}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00016136314602147107,"log_rms_mae_db":1.037140034371455,"rms_envelope_nrmse":0.17366282698367536,"envelope_correlation":0.9928540233599834,"rms_gain_error_db":1.156034021585519,"reference_mean_power":0.002674122622482191,"output_mean_power":0.0034896728950016827},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.348456388983372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008380618007305507,"output_power":0.001052462311199282,"reference_active":true,"delta_db":0.9893050689869142},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.026333057603631e-05,"output_power":0.00014875366935177242,"reference_active":true,"delta_db":2.1695633313769807},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.498114091574834e-06,"output_power":2.9930483001961634e-06,"reference_active":true,"delta_db":0.7850145611622729}],"active_band_count":3,"active_band_mae_db":1.314627653842056}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":1.4162707066095919,"rms_envelope_nrmse":0.17493180015333015,"envelope_correlation":0.9909558686981811,"rms_gain_error_db":1.0455233567351292,"reference_mean_power":0.0003102146198149335,"output_mean_power":0.0003946522626626717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.2470137873341424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001843563097177963,"output_power":0.00023478494428098075,"reference_active":true,"delta_db":1.0501223787095237},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0328682781544793e-05,"output_power":2.4568990873353247e-05,"reference_active":true,"delta_db":0.8227808000947222},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.580926194982453e-06,"output_power":9.229957671410734e-06,"reference_active":true,"delta_db":3.042464146044235}],"active_band_count":3,"active_band_mae_db":1.6384557749494937}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":1.0071848791802498,"rms_envelope_nrmse":0.14032749031417868,"envelope_correlation":0.9977832797295175,"rms_gain_error_db":1.0455243304659785,"reference_mean_power":0.00031427821970459705,"output_mean_power":0.0003998220277138852},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":3.052275331975407e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001756954599460375,"output_power":0.00022384240763314697,"reference_active":true,"delta_db":1.051818291486616},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.589536328721272e-06,"output_power":1.0688358708505382e-05,"reference_active":true,"delta_db":0.9494129975298788},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8479520444652186e-07,"output_power":5.764000376553171e-07,"reference_active":true,"delta_db":3.061913283998568}],"active_band_count":3,"active_band_mae_db":1.6877148576716874}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":1.4488124013095571,"rms_envelope_nrmse":0.18536192320702755,"envelope_correlation":0.9805540290222458,"rms_gain_error_db":0.5452503927720378,"reference_mean_power":1.9843186142044422e-05,"output_mean_power":2.2497613196653825e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.622000222111226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.352365504647437e-06,"output_power":1.0389483812123809e-05,"reference_active":true,"delta_db":0.4567249936810025},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.791249366368626e-07,"output_power":8.355944405381749e-07,"reference_active":true,"delta_db":-0.6884256961999703},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.495952213974191e-07,"output_power":6.349099863351209e-07,"reference_active":true,"delta_db":4.054758916892347}],"active_band_count":3,"active_band_mae_db":1.7333032022577732}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":0.9615727469744232,"rms_envelope_nrmse":0.11216536277477072,"envelope_correlation":0.9944111056230981,"rms_gain_error_db":0.5452601868278085,"reference_mean_power":2.010291979320634e-05,"output_mean_power":2.27921428712476e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.5131001811961107e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.175181026763142e-06,"output_power":9.128571604634972e-06,"reference_active":true,"delta_db":0.4790544830077292},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.283226898133564e-07,"output_power":3.675751943274644e-07,"reference_active":true,"delta_db":-0.6642488566228064},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.754857048422836e-08,"output_power":3.678745401025622e-08,"reference_active":true,"delta_db":3.2145798778980406}],"active_band_count":3,"active_band_mae_db":1.452627739176192}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9486281871795654,"predicted_mos":3.9486281871795654,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37200,"input_sample_rate":16000,"input_seconds":2.325,"input_float32_sha256":"101a46565514e1f9c72cf26052942f4eafac2ddba006758877a8f5b1a93f4567","internal_samples":37200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":13.100212059000114},"started_at":"2026-10-13T04:52:17.471532+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0870999833277892e-05,"native_input_samples":37200,"native_input_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","output_sha256":"6f57dc973c215ad4df018e619f9efa5fc6a2c317b76a9b92b529925ee157e75c","output_samples":37200,"length_delta_samples":0,"analysis_s":0.5553839200001676,"synthesis_s":4.3083936100001665,"total_s":4.863777530000334,"rtf":2.0919473247313265,"output_peak":0.4082970917224884,"output_rms":0.05775814460169206,"finished_at":"2026-10-13T04:52:22.338860+00:00","evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"6f57dc973c215ad4df018e619f9efa5fc6a2c317b76a9b92b529925ee157e75c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.046864104256122,"spectral_convergence":0.22079970708981336,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.254193640310626,"spectral_convergence":0.2233317233124196,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.373972746806688,"spectral_convergence":0.23862027135488326,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.000299205681522713,"delta_db":-1.7826145532104163,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0018395907392075705,"delta_db":-0.2257301868025064,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0009397794835864062,"delta_db":-1.6644245102614388,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.00023116863081382567,"delta_db":-1.408933754351803,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.3850512762281194e-05,"delta_db":-0.29703212143158,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":1.2408219937173238e-05,"delta_db":-2.1445371518554217,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2538787129855278,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9192446777479863,"pitch":null,"output_rms_dbfs":-24.76773532614686,"output_peak":0.4082970917224884,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.225010163791144},"common_prefix":{"samples":37200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.046864104256122,"spectral_convergence":0.22079970708981336,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.254193640310626,"spectral_convergence":0.2233317233124196,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.373972746806688,"spectral_convergence":0.23862027135488326,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.000299205681522713,"delta_db":-1.7826145532104163,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0018395907392075705,"delta_db":-0.2257301868025064,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0009397794835864062,"delta_db":-1.6644245102614388,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.00023116863081382567,"delta_db":-1.408933754351803,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.3850512762281194e-05,"delta_db":-0.29703212143158,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":1.2408219937173238e-05,"delta_db":-2.1445371518554217,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2538787129855278,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9192446777479863,"pitch":null,"output_rms_dbfs":-24.76773532614686,"output_peak":0.4082970917224884,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.225010163791144},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37200,"output_samples":37200,"intelligibility":{"estoi":{"value":0.9277134059817342,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0642059782880167,"rms_dbfs":-23.848490648398872,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.062475171735503,"mean":-3.708049815188172e-07,"zero_energy":false},"output":{"rms":0.05775814460169206,"rms_dbfs":-24.76773532614686,"peak":0.4082970917224884,"peak_dbfs":-7.780474270429277,"crest_factor_db":16.987261055717582,"mean":-7.881493967273347e-06,"zero_energy":false},"rms_difference_db":-0.9192446777479866},"full_scale":{"reference":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":3.8343296772398474e-07,"output_rms_dbfs":-64.16310549034719,"output_zero_energy":false,"output_peak":0.008455311879515648,"output_energy_fraction":3.213315604063331e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":110,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9626168224299065,"absolute_difference_cents_median":1.6549820837721478,"absolute_difference_cents_p95":8.558841328850836,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00019668187909197959,"log_rms_mae_db":1.3191856593678706,"rms_envelope_nrmse":0.11935574951941995,"envelope_correlation":0.9957045219494609,"rms_gain_error_db":-0.9192447707221841,"reference_mean_power":0.004131292024738224,"output_mean_power":0.00334319275940352},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.0026327800860698e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011612269501705267,"output_power":0.0008950169747017142,"reference_active":true,"delta_db":-1.1308583440745925},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017095886275955477,"output_power":0.00015805075967728876,"reference_active":true,"delta_db":-0.34095032548223925},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217923048678998e-05,"output_power":4.6346193884789896e-05,"reference_active":true,"delta_db":0.4091542187341723}],"active_band_count":3,"active_band_mae_db":0.6269876294303347}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00018115923356648065,"log_rms_mae_db":1.1236547987656333,"rms_envelope_nrmse":0.11222654653570052,"envelope_correlation":0.9972610180692152,"rms_gain_error_db":-0.9192449147794807,"reference_mean_power":0.004185412442455258,"output_mean_power":0.003386988870504433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.7728315536386115e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066252574844359,"output_power":0.0008230882733141972,"reference_active":true,"delta_db":-1.124136786732211},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.440983264620606e-05,"output_power":7.439078676218564e-05,"reference_active":true,"delta_db":-0.5487388736196838},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.940091323355611e-06,"output_power":3.063332687938788e-06,"reference_active":true,"delta_db":0.17833344750922403}],"active_band_count":3,"active_band_mae_db":0.617069702620373}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.000176021246464811,"log_rms_mae_db":1.1332208661354362,"rms_envelope_nrmse":0.11569051729591998,"envelope_correlation":0.9956005811531656,"rms_gain_error_db":-0.8341149531691117,"reference_mean_power":0.0026395436088586166,"output_mean_power":0.002178298268269542},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.515899729654865e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009091516678957249,"output_power":0.0007261489605577667,"reference_active":true,"delta_db":-0.9761061977848253},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001833276731517994,"output_power":0.0001654234890925897,"reference_active":true,"delta_db":-0.4463084960880265},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7517429693718724e-05,"output_power":3.9704985449006425e-05,"reference_active":true,"delta_db":0.24611963838090806}],"active_band_count":3,"active_band_mae_db":0.5561781107512533}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00016136314602147107,"log_rms_mae_db":0.8693347408515428,"rms_envelope_nrmse":0.10613893616442947,"envelope_correlation":0.997344650380319,"rms_gain_error_db":-0.8341150525079267,"reference_mean_power":0.002674122622482191,"output_mean_power":0.0022068347440460006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.348456388983372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008380618007305507,"output_power":0.0006657054416059483,"reference_active":true,"delta_db":-0.9999393877647738},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.026333057603631e-05,"output_power":7.682724070660894e-05,"reference_active":true,"delta_db":-0.6999611872984948},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.498114091574834e-06,"output_power":2.799991474987421e-06,"reference_active":true,"delta_db":0.4954443987647314}],"active_band_count":3,"active_band_mae_db":0.7317816579426667}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":1.8201123847350833,"rms_envelope_nrmse":0.1995277006633422,"envelope_correlation":0.986580424681695,"rms_gain_error_db":-1.4633319916853624,"reference_mean_power":0.0003102146198149335,"output_mean_power":0.00022147721932457693},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.2470137873341424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001843563097177963,"output_power":0.00013882459693476668,"reference_active":true,"delta_db":-1.2319158500523522},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0328682781544793e-05,"output_power":1.461281297144903e-05,"reference_active":true,"delta_db":-1.4337541316343827},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.580926194982453e-06,"output_power":2.7225414094157436e-06,"reference_active":true,"delta_db":-2.25978800724328}],"active_band_count":3,"active_band_mae_db":1.6418193296433383}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":1.4641626182677725,"rms_envelope_nrmse":0.18191340322497374,"envelope_correlation":0.9919213407448705,"rms_gain_error_db":-1.463331636318893,"reference_mean_power":0.00031427821970459705,"output_mean_power":0.00022437843816492382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":3.052275331975407e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001756954599460375,"output_power":0.00013167269955686566,"reference_active":true,"delta_db":-1.2526479969316004},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.589536328721272e-06,"output_power":6.038530680934426e-06,"reference_active":true,"delta_db":-1.5303844360595344},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8479520444652186e-07,"output_power":2.3031690115101962e-07,"reference_active":true,"delta_db":-0.9220686351437263}],"active_band_count":3,"active_band_mae_db":1.2350336893782872}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":1.9565591571995387,"rms_envelope_nrmse":0.25138240099569875,"envelope_correlation":0.9707663716342243,"rms_gain_error_db":-1.5668941382551473,"reference_mean_power":1.9843186142044422e-05,"output_mean_power":1.3833178867419324e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.622000222111226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.352365504647437e-06,"output_power":6.6393816817369545e-06,"reference_active":true,"delta_db":-1.4879383548628602},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.791249366368626e-07,"output_power":5.983839339118091e-07,"reference_active":true,"delta_db":-2.1385818686189086},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.495952213974191e-07,"output_power":3.956951650858714e-07,"reference_active":true,"delta_db":2.001244774030398}],"active_band_count":3,"active_band_mae_db":1.8759216658373887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":1.660174977325902,"rms_envelope_nrmse":0.2084434796582043,"envelope_correlation":0.9862010119647324,"rms_gain_error_db":-1.5669018834541928,"reference_mean_power":2.010291979320634e-05,"output_mean_power":1.4014220666293377e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.5131001811961107e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.175181026763142e-06,"output_power":5.696833337444197e-06,"reference_active":true,"delta_db":-1.5686386399151961},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.283226898133564e-07,"output_power":2.2171642026719042e-07,"reference_active":true,"delta_db":-2.859732231116858},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.754857048422836e-08,"output_power":2.5936183913034195e-08,"reference_active":true,"delta_db":1.6966433270402204}],"active_band_count":3,"active_band_mae_db":2.041671399357425}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.4733238220214844,"predicted_mos":3.4733238220214844,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":102912,"input_sample_rate":44100,"input_seconds":2.3336054421768706,"input_float32_sha256":"bdb122b19ab4fb0151663797210156d24428908966cda2a1d1868516c25e4bff","internal_samples":37338,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6190165860025445},"started_at":"2026-10-13T07:14:03.558862+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032189280009333743,"native_input_samples":102533,"native_input_sha256":"ea090abef77247a13282fa47e39bedaa8f0454960132378d88131e8f08d616eb","output_sha256":"f302fa69944c3e1cb7646b9c7d5bf091f2b51cc5939174394ff53e544700cd68","output_samples":102912,"length_delta_samples":379,"analysis_s":0.05191175199979625,"synthesis_s":13.79537498100035,"total_s":13.847286733000146,"rtf":5.955793207311856,"output_peak":0.3794011175632477,"output_rms":0.058334990228550154,"finished_at":"2026-10-13T07:14:17.421729+00:00","evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"f302fa69944c3e1cb7646b9c7d5bf091f2b51cc5939174394ff53e544700cd68","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":37200,"output_samples":37338,"length_delta_samples":138,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.67400310018522,"output_peak":0.3787968009258179},"common_prefix":{"samples":37200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9963040334243933,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.523964505369625,"spectral_convergence":0.5019261000902739,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.445129097954862,"spectral_convergence":0.3811227120773247,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.010413653882322,"spectral_convergence":0.44306327270930285,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.00012539605764305654,"delta_db":-5.559474084530642,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0017295982641774156,"delta_db":-0.4934899252817018,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.001357258280069088,"delta_db":-0.06805910907142486,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.00017930497036911056,"delta_db":-2.5122994770890195,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":7.349150477312647e-06,"delta_db":-3.0493192407955743,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":2.2524358778558827e-05,"delta_db":0.4448923838978985,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.021255703444377,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8094313880948251,"pitch":null,"output_rms_dbfs":-24.657922036493698,"output_peak":0.3787968009258179,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.326502419068937},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37200,"output_samples":37200,"intelligibility":{"estoi":{"value":0.7674177614691605,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0642059782880167,"rms_dbfs":-23.848490648398872,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.062475171735503,"mean":-3.708049815188172e-07,"zero_energy":false},"output":{"rms":0.058493000277935324,"rms_dbfs":-24.657922036493698,"peak":0.3787968009258179,"peak_dbfs":-8.431873948503647,"crest_factor_db":16.22604808799005,"mean":-2.866157880962091e-05,"zero_energy":false},"rms_difference_db":-0.8094313880948256},"full_scale":{"reference":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":3.2646493521508796e-07,"output_rms_dbfs":-64.8616345837173,"output_zero_energy":false,"output_peak":0.004771364600712704,"output_energy_fraction":2.6675903930212702e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":98,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9158878504672897,"absolute_difference_cents_median":7.904516498868912,"absolute_difference_cents_p95":40.85647709640481,"difference_over_600_cents_frames":3,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00019668187909197959,"log_rms_mae_db":1.3580130739354432,"rms_envelope_nrmse":0.12594068571209086,"envelope_correlation":0.9919261055860097,"rms_gain_error_db":-0.8094313594601446,"reference_mean_power":0.004131292024738224,"output_mean_power":0.003428804776397358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.0026327800860698e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011612269501705267,"output_power":0.0009205322678999694,"reference_active":true,"delta_db":-1.0087808987940359},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017095886275955477,"output_power":0.00014006740736903522,"reference_active":true,"delta_db":-0.8655453019668471},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217923048678998e-05,"output_power":2.192997920167632e-05,"reference_active":true,"delta_db":-2.840604324611183}],"active_band_count":3,"active_band_mae_db":1.5716435084573555}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00018115923356648065,"log_rms_mae_db":1.045102042192325,"rms_envelope_nrmse":0.09928869596494817,"envelope_correlation":0.9979185481883283,"rms_gain_error_db":-0.8094309393991611,"reference_mean_power":0.004185412442455258,"output_mean_power":0.003473722863455901},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.7728315536386115e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066252574844359,"output_power":0.000844489826211965,"reference_active":true,"delta_db":-1.0126567105439426},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.440983264620606e-05,"output_power":7.483340674747439e-05,"reference_active":true,"delta_db":-0.522975225458978},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.940091323355611e-06,"output_power":1.9517229393759032e-06,"reference_active":true,"delta_db":-1.7794265382803358}],"active_band_count":3,"active_band_mae_db":1.105019491427752}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.000176021246464811,"log_rms_mae_db":2.059867348335268,"rms_envelope_nrmse":0.1687101764201063,"envelope_correlation":0.9770230264588755,"rms_gain_error_db":-0.1166831419008722,"reference_mean_power":0.0026395436088586166,"output_mean_power":0.0025695704387115915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.515899729654865e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009091516678957249,"output_power":0.0008209925611970184,"reference_active":true,"delta_db":-0.44297117700492805},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001833276731517994,"output_power":0.00012559822829092453,"reference_active":true,"delta_db":-1.6424451306681522},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7517429693718724e-05,"output_power":1.1151584569502996e-05,"reference_active":true,"delta_db":-5.268964950968938}],"active_band_count":3,"active_band_mae_db":2.4514604195473395}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00016136314602147107,"log_rms_mae_db":1.61504578241042,"rms_envelope_nrmse":0.13354041330743258,"envelope_correlation":0.9849241261577135,"rms_gain_error_db":-0.11668336726682793,"reference_mean_power":0.002674122622482191,"output_mean_power":0.0026032326424114635},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.348456388983372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008380618007305507,"output_power":0.0007575896927744691,"reference_active":true,"delta_db":-0.43841988416193944},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.026333057603631e-05,"output_power":6.593515057915006e-05,"reference_active":true,"delta_db":-1.3639435192477702},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.498114091574834e-06,"output_power":8.14657122722522e-07,"reference_active":true,"delta_db":-4.866374101951746}],"active_band_count":3,"active_band_mae_db":2.222912501787152}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":3.6127739310564806,"rms_envelope_nrmse":0.34071276496601477,"envelope_correlation":0.9396263971477401,"rms_gain_error_db":-2.1392394054624915,"reference_mean_power":0.0003102146198149335,"output_mean_power":0.0001895563427182515},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.2470137873341424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001843563097177963,"output_power":0.00010472950879863297,"reference_active":true,"delta_db":-2.4558893960044905},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0328682781544793e-05,"output_power":1.074704539587279e-05,"reference_active":true,"delta_db":-2.7682015562536932},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.580926194982453e-06,"output_power":1.937313601233583e-06,"reference_active":true,"delta_db":-3.7375336723755104}],"active_band_count":3,"active_band_mae_db":2.9872082082112317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":2.9782993299073257,"rms_envelope_nrmse":0.314336926075749,"envelope_correlation":0.9531269798130447,"rms_gain_error_db":-2.1392406995369533,"reference_mean_power":0.00031427821970459705,"output_mean_power":0.0001920393442693179},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":3.052275331975407e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001756954599460375,"output_power":9.810649254191716e-05,"reference_active":true,"delta_db":-2.530627899704477},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.589536328721272e-06,"output_power":4.111541865584197e-06,"reference_active":true,"delta_db":-3.1996500403681134},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8479520444652186e-07,"output_power":1.4932727722650854e-07,"reference_active":true,"delta_db":-2.8039352569550062}],"active_band_count":3,"active_band_mae_db":2.8447377323425322}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":6.230378073138113,"rms_envelope_nrmse":0.2437065507271398,"envelope_correlation":0.9615786172305948,"rms_gain_error_db":-0.12249439960698398,"reference_mean_power":1.9843186142044422e-05,"output_mean_power":1.9291320947873505e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.622000222111226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.352365504647437e-06,"output_power":1.1016179224050593e-05,"reference_active":true,"delta_db":0.7110952154207095},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.791249366368626e-07,"output_power":1.47513554442213e-06,"reference_active":true,"delta_db":1.7799381631932483},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.495952213974191e-07,"output_power":4.2240330253774356e-07,"reference_active":true,"delta_db":2.2849103870279897}],"active_band_count":3,"active_band_mae_db":1.5919812552139827}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":5.6448347871533695,"rms_envelope_nrmse":0.19705881915979265,"envelope_correlation":0.9744582573920157,"rms_gain_error_db":-0.12245514305099293,"reference_mean_power":2.010291979320634e-05,"output_mean_power":1.9544007723956936e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.5131001811961107e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.175181026763142e-06,"output_power":9.93572731585033e-06,"reference_active":true,"delta_db":0.8470228564682004},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.283226898133564e-07,"output_power":6.907412923827372e-07,"reference_active":true,"delta_db":2.075443376565999},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.754857048422836e-08,"output_power":3.214335551481856e-08,"reference_active":true,"delta_db":2.6284946735374954}],"active_band_count":3,"active_band_mae_db":1.850320302190565}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"golf","status":"measured","mos":{"status":"measured","value":3.0507664680480957,"predicted_mos":3.0507664680480957,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":55680,"input_sample_rate":24000,"input_seconds":2.32,"input_float32_sha256":"97e8f65699b0f061c3f27e85e637b51c85932c56140434423d9c3f0dbfe250e8","internal_samples":37120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.914085710999643},"started_at":"2026-10-13T05:02:33.268437+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0015131990003283136,"native_input_samples":55800,"native_input_sha256":"32696c3b4bdde6347e5dcec4db9c29e61a5047052774833a5b61eaeb3f3d0aa0","output_sha256":"1f7d183a3ba3dc757e48d94bb76ba9fae3dc6e06277dd969eda9df52c063da03","output_samples":55680,"length_delta_samples":-120,"analysis_s":0.2629357969999546,"synthesis_s":0.057931899999857706,"total_s":0.3208676969998123,"rtf":0.1380076116128225,"output_peak":0.38044196367263794,"output_rms":0.05846766541462685,"finished_at":"2026-10-13T05:02:33.596348+00:00","analysis_stages":{"external_f0_s":0.026394977000563813},"evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"1f7d183a3ba3dc757e48d94bb76ba9fae3dc6e06277dd969eda9df52c063da03","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":37200,"output_samples":37120,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.65571952703783,"output_peak":0.37955332903570627},"common_prefix":{"samples":37120,"sample_rate":16000,"reference_coverage":0.9978494623655914,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37120,"output_samples":37120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.88648844808007,"spectral_convergence":0.42670467570455706,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.3675646699413715,"spectral_convergence":0.2192644099256365,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.494562446945795,"spectral_convergence":0.25865768835948044,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004504481804257944,"output_power":0.0003766678161368487,"delta_db":-0.7768632284227512,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001943422286786387,"output_power":0.0017483848612974112,"delta_db":-0.4593014140893227,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013817254735594368,"output_power":0.001032489450079502,"delta_db":-1.2653614158556814,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00032045777247844255,"output_power":0.0002491092268065174,"delta_db":-1.0938099580112315,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4846472945594315e-05,"output_power":9.531945915013156e-06,"delta_db":-1.92441721753883,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.0391816583613758e-05,"output_power":6.983388218809205e-06,"delta_db":-4.653897302656778,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6956084227624324,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8165784554508273,"pitch":null,"output_rms_dbfs":-24.65571952703783,"output_peak":0.37955332903570627,"reference_rms_dbfs":-23.839141071587004,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5828718549890795},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37120,"output_samples":37120,"intelligibility":{"estoi":{"value":0.9139555882607285,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.064275127442621,"rms_dbfs":-23.839141071587004,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.053125594923632,"mean":-1.397626153353987e-07,"zero_energy":false},"output":{"rms":0.05850783442009716,"rms_dbfs":-24.65571952703783,"peak":0.37955332903570627,"peak_dbfs":-8.414543901099872,"crest_factor_db":16.24117562593796,"mean":0.0003358529000146211,"zero_energy":false},"rms_difference_db":-0.8165784554508271},"full_scale":{"reference":{"sample_count":37120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":3.9315578649077536e-07,"output_rms_dbfs":-64.05435327978083,"output_zero_energy":false,"output_peak":0.003983026481765674,"output_energy_fraction":3.217821850947154e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":103,"both_estimated_voiced_frames":100,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9345794392523364,"absolute_difference_cents_median":4.780035338486039,"absolute_difference_cents_p95":19.239580809917904,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":0.00019668187909197959,"log_rms_mae_db":2.4234882524387085,"rms_envelope_nrmse":0.14349244232217556,"envelope_correlation":0.9873815290674232,"rms_gain_error_db":-0.8165785857291836,"reference_mean_power":0.004140214728089411,"output_mean_power":0.0034305599044486042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":3.003766410465836e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011632919899859549,"output_power":0.0008307602698417887,"reference_active":true,"delta_db":-1.4621301898294128},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001716356211035032,"output_power":0.0001556065092315544,"reference_active":true,"delta_db":-0.4257966579442881},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.189670046505838e-05,"output_power":3.6131228859697466e-05,"reference_active":true,"delta_db":-0.642970892481344}],"active_band_count":3,"active_band_mae_db":0.843632580085015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":457,"rms_floor":0.00018115923356648065,"log_rms_mae_db":2.2148073193575075,"rms_envelope_nrmse":0.1226884238804608,"envelope_correlation":0.9925325770338027,"rms_gain_error_db":-0.8165818082544746,"reference_mean_power":0.0041945706542992215,"output_mean_power":0.0034755963568787775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":457,"frequency_resolution_hz":0.437636761487965,"power_floor":2.7733647188794523e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010681977427456112,"output_power":0.0007590604359768934,"reference_active":true,"delta_db":-1.4837530034651991},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.460131361958705e-05,"output_power":7.185005018550985e-05,"reference_active":true,"delta_db":-0.7095003064428279},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.881862320932468e-06,"output_power":2.2207600423557615e-06,"reference_active":true,"delta_db":-1.1317159416124865}],"active_band_count":3,"active_band_mae_db":1.1083230838401712}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":0.000176021246464811,"log_rms_mae_db":1.511926723372984,"rms_envelope_nrmse":0.15880474815516402,"envelope_correlation":0.9836149349611549,"rms_gain_error_db":-0.7260400378827965,"reference_mean_power":0.002645244546257576,"output_mean_power":0.0022380089752577683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":2.5185730603939364e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009100767255948098,"output_power":0.0005977688049967383,"reference_active":true,"delta_db":-1.8254476049444732},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018391807330649408,"output_power":0.00014598255632809369,"reference_active":true,"delta_db":-1.003234442984848},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.726609569230994e-05,"output_power":3.264007136371928e-05,"reference_active":true,"delta_db":-0.5756279519977291}],"active_band_count":3,"active_band_mae_db":1.1347699999756835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":457,"rms_floor":0.00016136314602147107,"log_rms_mae_db":1.1772445418512012,"rms_envelope_nrmse":0.13244102588319048,"envelope_correlation":0.9898119307422886,"rms_gain_error_db":-0.7260399529378001,"reference_mean_power":0.00267997405028598,"output_mean_power":0.0022673919142169358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":457,"frequency_resolution_hz":0.437636761487965,"power_floor":2.3505875902630067e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000838941907660698,"output_power":0.0005433098273397104,"reference_active":true,"delta_db":-1.8868432866376825},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.035323466286242e-05,"output_power":6.852624547294032e-05,"reference_active":true,"delta_db":-1.2008676736082942},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447928463952516e-06,"output_power":2.218414373172862e-06,"reference_active":true,"delta_db":-0.4275605184398867}],"active_band_count":3,"active_band_mae_db":1.1717571595619545}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":2.0231228115757043,"rms_envelope_nrmse":0.17922152110547873,"envelope_correlation":0.9859553038842498,"rms_gain_error_db":-1.125601830283387,"reference_mean_power":0.0003108845644954838,"output_mean_power":0.0002399048217192887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":3.2543635903045026e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018497614816052486,"output_power":0.00015041027798190712,"reference_active":true,"delta_db":-0.8983821779580324},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0189748466000555e-05,"output_power":1.3722436897952652e-05,"reference_active":true,"delta_db":-1.6769966594576435},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.561887503648877e-06,"output_power":3.06226253878725e-06,"reference_active":true,"delta_db":-1.731021497640145}],"active_band_count":3,"active_band_mae_db":1.4354667783519404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":457,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":1.568091420980355,"rms_envelope_nrmse":0.14212916288765828,"envelope_correlation":0.9954399101220969,"rms_gain_error_db":-1.1256021033292054,"reference_mean_power":0.00031496584692778333,"output_mean_power":0.00024305426912025295},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":457,"frequency_resolution_hz":0.437636761487965,"power_floor":3.0589929994288476e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017620250076395583,"output_power":0.00014275414518146056,"reference_active":true,"delta_db":-0.9142334006963095},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.46074893649079e-06,"output_power":5.607399535258109e-06,"reference_active":true,"delta_db":-1.7864730672032376},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8544736147468247e-07,"output_power":2.783302344336706e-07,"reference_active":true,"delta_db":-0.10965647569814123}],"active_band_count":3,"active_band_mae_db":0.936787647865896}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":3.223872138159171,"rms_envelope_nrmse":0.3598407088337302,"envelope_correlation":0.9626156772591421,"rms_gain_error_db":-3.1183027480916885,"reference_mean_power":1.9886011102046226e-05,"output_mean_power":9.698786576930011e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":2.6326161724726292e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.387296216809992e-06,"output_power":4.6566084339656515e-06,"reference_active":true,"delta_db":-3.0447080186888043},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.84941796129944e-07,"output_power":1.1012469977928587e-06,"reference_active":true,"delta_db":0.48474170271582634},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5006701971586783e-07,"output_power":2.0423655785177767e-07,"reference_active":true,"delta_db":-0.8792293586534121}],"active_band_count":3,"active_band_mae_db":1.469559693352681}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37120,"uncovered_tail_samples":0,"metrics":{"frame_count":457,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":3.0335325667716657,"rms_envelope_nrmse":0.3374438776484835,"envelope_correlation":0.9768141825756779,"rms_gain_error_db":-3.1183025341638486,"reference_mean_power":2.0146873256287207e-05,"output_mean_power":9.826014504751265e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":457,"frequency_resolution_hz":0.437636761487965,"power_floor":2.5232821098351852e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.204579370828604e-06,"output_power":4.193146813100735e-06,"reference_active":true,"delta_db":-2.9151625225785267},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3065818079453716e-07,"output_power":5.488737330495003e-07,"reference_active":true,"delta_db":1.0533974614295256},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7439512656183384e-08,"output_power":1.3737826765073336e-08,"reference_active":true,"delta_db":-1.0361630889851823}],"active_band_count":3,"active_band_mae_db":1.6682410243310783}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.2677762508392334,"predicted_mos":3.2677762508392334,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":55920,"input_sample_rate":24000,"input_seconds":2.33,"input_float32_sha256":"03ad2e2ae90e52d2a84cd3849b0685dc51f011cb62a6e6f8dee9842ee9a76b72","internal_samples":37280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7937035630020546},"started_at":"2026-10-13T07:37:14.842233+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0016798550022940617,"native_input_samples":55800,"native_input_sha256":"32696c3b4bdde6347e5dcec4db9c29e61a5047052774833a5b61eaeb3f3d0aa0","output_sha256":"d24de892395161947a209f550d3a834aaba5cc725bbf7f95be2eb0d90cd3c2c4","output_samples":55920,"length_delta_samples":120,"analysis_s":0.8695209990000876,"synthesis_s":4.9856517410007655,"total_s":5.855172740000853,"rtf":2.5183538666670335,"output_peak":0.40470200777053833,"output_rms":0.061898369107174084,"finished_at":"2026-10-13T07:37:20.712218+00:00","analysis_stages":{"external_f0_s":0.5956655669979227},"evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"d24de892395161947a209f550d3a834aaba5cc725bbf7f95be2eb0d90cd3c2c4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":37200,"output_samples":37280,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.160547940620077,"output_peak":0.40499652547841275},"common_prefix":{"samples":37200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9978540772532188,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.7466808663086715,"spectral_convergence":0.4814213737693996,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.2115354871592094,"spectral_convergence":0.22862595232480232,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.1931366426026955,"spectral_convergence":0.26500876800172346,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.0004559147714582088,"delta_db":0.04652372126017214,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0018855456660814888,"delta_db":-0.11857178221427175,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0011897654685266,"delta_db":-0.6400705078234394,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.000284153869425885,"delta_db":-0.5126870180070545,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.4985347961106283e-05,"delta_db":0.04497767821034046,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":1.4473963494261046e-05,"delta_db":-1.4757572505199765,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.47309799300587585,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.30272766604743395,"pitch":null,"output_rms_dbfs":-24.151218314446304,"output_peak":0.40499652547841275,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.383784332023526},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37200,"output_samples":37200,"intelligibility":{"estoi":{"value":0.9222885983516048,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0642059782880167,"rms_dbfs":-23.848490648398872,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.062475171735503,"mean":-3.708049815188172e-07,"zero_energy":false},"output":{"rms":0.06200676646098835,"rms_dbfs":-24.151218314446304,"peak":0.40499652547841275,"peak_dbfs":-7.850974052843686,"crest_factor_db":16.30024426160262,"mean":-1.5159475972423877e-05,"zero_energy":false},"rms_difference_db":-0.3027276660474314},"full_scale":{"reference":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":4.1768709345009157e-07,"output_rms_dbfs":-63.791489442465036,"output_zero_energy":false,"output_peak":0.015605978042709282,"output_energy_fraction":3.037129335260641e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":102,"both_estimated_voiced_frames":98,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9158878504672897,"absolute_difference_cents_median":5.207063478600071,"absolute_difference_cents_p95":23.975359419595247,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00019668187909197959,"log_rms_mae_db":1.911450719910361,"rms_envelope_nrmse":0.12608267140834437,"envelope_correlation":0.9858162020970181,"rms_gain_error_db":-0.30272765826366654,"reference_mean_power":0.004131292024738224,"output_mean_power":0.0038531252708078573},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.0026327800860698e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011612269501705267,"output_power":0.0010136983161203423,"reference_active":true,"delta_db":-0.590083814606001},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017095886275955477,"output_power":0.0001611430306876476,"reference_active":true,"delta_db":-0.2568009283099008},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217923048678998e-05,"output_power":3.112564132959668e-05,"reference_active":true,"delta_db":-1.3198034362025612}],"active_band_count":3,"active_band_mae_db":0.7222293930394876}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00018115923356648065,"log_rms_mae_db":1.526862043919516,"rms_envelope_nrmse":0.08766333988709639,"envelope_correlation":0.9933574840745338,"rms_gain_error_db":-0.3027271378295754,"reference_mean_power":0.004185412442455258,"output_mean_power":0.00390360213869394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.7728315536386115e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066252574844359,"output_power":0.0009246138509597712,"reference_active":true,"delta_db":-0.6189969793776546},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.440983264620606e-05,"output_power":7.789496346131492e-05,"reference_active":true,"delta_db":-0.3488366129006719},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.940091323355611e-06,"output_power":2.5542371918780203e-06,"reference_active":true,"delta_db":-0.6109959611082215}],"active_band_count":3,"active_band_mae_db":0.526276517795516}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.000176021246464811,"log_rms_mae_db":1.702091890821284,"rms_envelope_nrmse":0.15272495054451116,"envelope_correlation":0.9818019332355812,"rms_gain_error_db":-0.34161633763126364,"reference_mean_power":0.0026395436088586166,"output_mean_power":0.002439872925231871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.515899729654865e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009091516678957249,"output_power":0.0007659124460355644,"reference_active":true,"delta_db":-0.7445721296392447},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001833276731517994,"output_power":0.000175211633994802,"reference_active":true,"delta_db":-0.19665086482389277},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7517429693718724e-05,"output_power":3.348701934897957e-05,"reference_active":true,"delta_db":-0.49356584280848353}],"active_band_count":3,"active_band_mae_db":0.47826294575720696}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00016136314602147107,"log_rms_mae_db":1.1112490141235987,"rms_envelope_nrmse":0.10569965984355541,"envelope_correlation":0.991519083320326,"rms_gain_error_db":-0.3416162635002522,"reference_mean_power":0.002674122622482191,"output_mean_power":0.0024718362200254057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.348456388983372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008380618007305507,"output_power":0.0007021955453639204,"reference_active":true,"delta_db":-0.7681797571928294},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.026333057603631e-05,"output_power":8.540016611294366e-05,"reference_active":true,"delta_db":-0.24052638826897466},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.498114091574834e-06,"output_power":2.6678608355763477e-06,"reference_active":true,"delta_db":0.2855090241205248}],"active_band_count":3,"active_band_mae_db":0.431405056527443}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":1.8848405871055527,"rms_envelope_nrmse":0.1638791705057893,"envelope_correlation":0.9819802065590156,"rms_gain_error_db":-0.5097934924193925,"reference_mean_power":0.0003102146198149335,"output_mean_power":0.00027585630340591015},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.2470137873341424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001843563097177963,"output_power":0.00015740223260641434,"reference_active":true,"delta_db":-0.6864711807919226},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0328682781544793e-05,"output_power":1.488143882025295e-05,"reference_active":true,"delta_db":-1.3546431573510325},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.580926194982453e-06,"output_power":4.364315810270052e-06,"reference_active":true,"delta_db":-0.21037125315596217}],"active_band_count":3,"active_band_mae_db":0.7504951970996391}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":1.1913430335272412,"rms_envelope_nrmse":0.1091189555189896,"envelope_correlation":0.993086462302042,"rms_gain_error_db":-0.509792980581548,"reference_mean_power":0.00031427821970459705,"output_mean_power":0.00027946986572694935},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":3.052275331975407e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001756954599460375,"output_power":0.0001474235951343083,"reference_active":true,"delta_db":-0.7619354138275668},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.589536328721272e-06,"output_power":6.256286590883593e-06,"reference_active":true,"delta_db":-1.3765308596698422},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8479520444652186e-07,"output_power":2.856079298188911e-07,"reference_active":true,"delta_db":0.012375892142688716}],"active_band_count":3,"active_band_mae_db":0.7169473885466991}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":1.8135398093570547,"rms_envelope_nrmse":0.1914288455238009,"envelope_correlation":0.974843153155894,"rms_gain_error_db":-0.2775943427734045,"reference_mean_power":1.9843186142044422e-05,"output_mean_power":1.8614525582638837e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.622000222111226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.352365504647437e-06,"output_power":8.874223332668965e-06,"reference_active":true,"delta_db":-0.22791117206352896},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.791249366368626e-07,"output_power":1.1063011089213875e-06,"reference_active":true,"delta_db":0.5303523626904239},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.495952213974191e-07,"output_power":2.1484147110715102e-07,"reference_active":true,"delta_db":-0.6511814883948931}],"active_band_count":3,"active_band_mae_db":0.46981500771628193}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":1.1011714233553744,"rms_envelope_nrmse":0.1363882421788703,"envelope_correlation":0.9870639528224375,"rms_gain_error_db":-0.27759102435426675,"reference_mean_power":2.010291979320634e-05,"output_mean_power":1.8858191322032437e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.5131001811961107e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.175181026763142e-06,"output_power":7.903060524666945e-06,"reference_active":true,"delta_db":-0.14702070274066123},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.283226898133564e-07,"output_power":4.840157797662419e-07,"reference_active":true,"delta_db":0.5308843947275367},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.754857048422836e-08,"output_power":1.6566713786299808e-08,"reference_active":true,"delta_db":-0.2500537513489226}],"active_band_count":3,"active_band_mae_db":0.30931961627237353}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.9367005825042725,"predicted_mos":2.9367005825042725,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":56040,"input_sample_rate":24000,"input_seconds":2.335,"input_float32_sha256":"4068f84e064b52b2c42a6f75158fecbf0212d29599663d9e8d1ad1a8a9ecaf79","internal_samples":37360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3615418760000466},"started_at":"2026-10-13T05:10:42.619467+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018260100005136337,"native_input_samples":55800,"native_input_sha256":"32696c3b4bdde6347e5dcec4db9c29e61a5047052774833a5b61eaeb3f3d0aa0","output_sha256":"af6ecaa85017133b2aabe4d0e58d756fa23bb2f5ed911a9acdbbcecfa3791923","output_samples":56040,"length_delta_samples":240,"analysis_s":0.25298225699953036,"synthesis_s":0.06699726400074724,"total_s":0.3199795210002776,"rtf":0.13762560043022692,"output_peak":0.4631173312664032,"output_rms":0.06327599994018945,"finished_at":"2026-10-13T05:10:42.952616+00:00","analysis_stages":{"external_f0_s":0.028384274999552872},"evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"af6ecaa85017133b2aabe4d0e58d756fa23bb2f5ed911a9acdbbcecfa3791923","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":37200,"output_samples":37360,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.969185895850117,"output_peak":0.45389551469864553},"common_prefix":{"samples":37200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9957173447537473,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.860501686975292,"spectral_convergence":0.137671452318508,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.64599907607716,"spectral_convergence":0.16937179175193795,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.151595941552595,"spectral_convergence":0.23234917509915154,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.00044573825182165136,"delta_db":-0.05151385833663576,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0018833019420451105,"delta_db":-0.12374279035443914,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0013584803241598155,"delta_db":-0.06415058090718369,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.0003078192211612153,"delta_db":-0.16526541166163133,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.2922314447821577e-05,"delta_db":-0.5982875910635216,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":1.8401434702789483e-05,"delta_db":-0.4331151305728258,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23934589381603955,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10205598392088511,"pitch":null,"output_rms_dbfs":-23.95054663231976,"output_peak":0.45389551469864553,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.886032234868349},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37200,"output_samples":37200,"intelligibility":{"estoi":{"value":0.9540118769984228,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0642059782880167,"rms_dbfs":-23.848490648398872,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.062475171735503,"mean":-3.708049815188172e-07,"zero_energy":false},"output":{"rms":0.06345599647266131,"rms_dbfs":-23.95054663231976,"peak":0.45389551469864553,"peak_dbfs":-6.8608821768791906,"crest_factor_db":17.089664455440566,"mean":-2.8114297831856456e-05,"zero_energy":false},"rms_difference_db":-0.10205598392088788},"full_scale":{"reference":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":3.6197328031327814e-07,"output_rms_dbfs":-64.41323486485574,"output_zero_energy":false,"output_peak":0.006239537227026443,"output_energy_fraction":2.5131683178492717e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":88,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.794392523364486,"absolute_difference_cents_median":1.0831918308546666,"absolute_difference_cents_p95":7.9620035939702225,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00019668187909197959,"log_rms_mae_db":0.7555527465329372,"rms_envelope_nrmse":0.048209084379310436,"envelope_correlation":0.9979558642202904,"rms_gain_error_db":-0.10205597788499651,"reference_mean_power":0.004131292024738224,"output_mean_power":0.004035341528435559},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.0026327800860698e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011612269501705267,"output_power":0.0011277561956939584,"reference_active":true,"delta_db":-0.12701884824605655},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017095886275955477,"output_power":0.00017294847408355797,"reference_active":true,"delta_db":0.05025114424408766},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217923048678998e-05,"output_power":4.415040203810912e-05,"reference_active":true,"delta_db":0.19836010406576354}],"active_band_count":3,"active_band_mae_db":0.12521003218530258}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00018115923356648065,"log_rms_mae_db":0.46390227391867744,"rms_envelope_nrmse":0.024405790888232026,"envelope_correlation":0.9995504741252409,"rms_gain_error_db":-0.10205566602523539,"reference_mean_power":0.004185412442455258,"output_mean_power":0.00408820527678702},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.7728315536386115e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066252574844359,"output_power":0.0010345976171081445,"reference_active":true,"delta_db":-0.13088619130253948},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.440983264620606e-05,"output_power":8.681637728703182e-05,"reference_active":true,"delta_db":0.12208620221078242},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.940091323355611e-06,"output_power":2.997326473782698e-06,"reference_active":true,"delta_db":0.08373229182323845}],"active_band_count":3,"active_band_mae_db":0.11223489511218677}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.000176021246464811,"log_rms_mae_db":0.7137283766371683,"rms_envelope_nrmse":0.051301744296818295,"envelope_correlation":0.9980400117400191,"rms_gain_error_db":-0.1278371145922196,"reference_mean_power":0.0026395436088586166,"output_mean_power":0.002562979485727904},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.515899729654865e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009091516678957249,"output_power":0.0008772966946932524,"reference_active":true,"delta_db":-0.15489846694959586},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001833276731517994,"output_power":0.00017395261581002182,"reference_active":true,"delta_db":-0.22797062452572914},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7517429693718724e-05,"output_power":3.9368304955238856e-05,"reference_active":true,"delta_db":0.20913638816505029}],"active_band_count":3,"active_band_mae_db":0.1973351598801251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00016136314602147107,"log_rms_mae_db":0.35751076727156794,"rms_envelope_nrmse":0.023951618046077434,"envelope_correlation":0.9996939743325277,"rms_gain_error_db":-0.12783712412149706,"reference_mean_power":0.002674122622482191,"output_mean_power":0.002596555474856507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.348456388983372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008380618007305507,"output_power":0.0008081186423234256,"reference_active":true,"delta_db":-0.15800920218486314},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.026333057603631e-05,"output_power":8.903192886373076e-05,"reference_active":true,"delta_db":-0.05965571863852833},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.498114091574834e-06,"output_power":2.515884380254003e-06,"reference_active":true,"delta_db":0.030784096452591905}],"active_band_count":3,"active_band_mae_db":0.08281633909199447}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":1.0117899662885907,"rms_envelope_nrmse":0.058978063617562484,"envelope_correlation":0.997744579312551,"rms_gain_error_db":-0.17193450543842526,"reference_mean_power":0.0003102146198149335,"output_mean_power":0.0002981733405219197},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.2470137873341424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001843563097177963,"output_power":0.00017660516761631144,"reference_active":true,"delta_db":-0.18654598964444014},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0328682781544793e-05,"output_power":1.911666363580881e-05,"reference_active":true,"delta_db":-0.2669714031813483},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.580926194982453e-06,"output_power":2.6677653587682348e-06,"reference_active":true,"delta_db":-2.3480566586126415}],"active_band_count":3,"active_band_mae_db":0.9338580171461434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":0.6797900937996268,"rms_envelope_nrmse":0.034217067277172386,"envelope_correlation":0.9994155117053067,"rms_gain_error_db":-0.1719331889028709,"reference_mean_power":0.00031427821970459705,"output_mean_power":0.00030207929944083017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":3.052275331975407e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001756954599460375,"output_power":0.0001671463140709326,"reference_active":true,"delta_db":-0.21663735340953938},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.589536328721272e-06,"output_power":8.20628605006205e-06,"reference_active":true,"delta_db":-0.19823069552578376},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8479520444652186e-07,"output_power":1.633847380498883e-07,"reference_active":true,"delta_db":-2.413211859534658}],"active_band_count":3,"active_band_mae_db":0.9426933028233271}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":1.355687675235233,"rms_envelope_nrmse":0.10542582297604068,"envelope_correlation":0.9936953242119916,"rms_gain_error_db":-0.43258715956529,"reference_mean_power":1.9843186142044422e-05,"output_mean_power":1.7961917214191886e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.622000222111226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.352365504647437e-06,"output_power":8.154209706681213e-06,"reference_active":true,"delta_db":-0.5953959510303647},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.791249366368626e-07,"output_power":9.833021010991643e-07,"reference_active":true,"delta_db":0.01848855708686364},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.495952213974191e-07,"output_power":2.511304412069759e-07,"reference_active":true,"delta_db":0.026630933094945318}],"active_band_count":3,"active_band_mae_db":0.21350514707072454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":1.0472074853453637,"rms_envelope_nrmse":0.07147110908803875,"envelope_correlation":0.9979484159560789,"rms_gain_error_db":-0.43257189448939803,"reference_mean_power":2.010291979320634e-05,"output_mean_power":1.819709031080466e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.5131001811961107e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.175181026763142e-06,"output_power":7.190812613491389e-06,"reference_active":true,"delta_db":-0.5571940666658265},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.283226898133564e-07,"output_power":4.5624637657729844e-07,"reference_active":true,"delta_db":0.27428347226443156},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.754857048422836e-08,"output_power":1.959972624465926e-08,"reference_active":true,"delta_db":0.48008261089611026}],"active_band_count":3,"active_band_mae_db":0.43718671660878944}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2271-0022","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2756481170654297,"predicted_mos":3.2756481170654297,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":111647,"input_sample_rate":48000,"input_seconds":2.3259791666666665,"input_float32_sha256":"b047536a7afe7acc22d0a01b6d1eff8bb24b5f11407853d56e42fd47f92a27f3","internal_samples":37216,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.289173025999844},"started_at":"2026-10-13T05:15:23.309205+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005165812000996084,"native_input_samples":111600,"native_input_sha256":"01bda0ed759c6d0e43cb77240b387a2e28c4ce757d23ae7cc78401ac5d69e72b","output_sha256":"5fbeaccc42148f5180689e24d7c0d2027470e2b099507da38768e7511931bb37","output_samples":111647,"length_delta_samples":47,"analysis_s":1.7181128960000933,"synthesis_s":0.9706670749983459,"total_s":2.6887799709984392,"rtf":1.1564645036552426,"output_peak":0.4677944997024764,"output_rms":0.06270455493920915,"finished_at":"2026-10-13T05:15:26.010836+00:00","evaluation_input":{"reference":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"5fbeaccc42148f5180689e24d7c0d2027470e2b099507da38768e7511931bb37","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":37200,"output_samples":37216,"length_delta_samples":16,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.047974399168147,"output_peak":0.46819304245589993},"common_prefix":{"samples":37200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9995700773860705,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37200,"output_samples":37200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.774758629961123,"spectral_convergence":0.50429486785524,"spectral_convergence_status":"ok","amplitude_floor":1.3482074212160965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.9116707766562735,"spectral_convergence":0.11454397545882683,"spectral_convergence_status":"ok","amplitude_floor":1.080252897815134e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.244875459743213,"spectral_convergence":0.12860346587347302,"spectral_convergence_status":"ok","amplitude_floor":8.596818112459857e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004510568589627121,"output_power":0.00040758059057461856,"delta_db":-0.44017797016037796,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019377343288488795,"output_power":0.0018386239309431584,"delta_db":-0.2280132482419029,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013786956670426643,"output_power":0.0013443518227670265,"delta_db":-0.10955470330267203,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031975863120548915,"output_power":0.0003105233900075306,"delta_db":-0.1272795742332179,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.4830953182024825e-05,"output_power":1.5038071540699765e-05,"delta_db":0.06023082632003138,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.03312086795025e-05,"output_power":2.291241887016097e-05,"delta_db":0.5190774221315765,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24738895739829644,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19761621938821994,"pitch":null,"output_rms_dbfs":-24.046106867787095,"output_peak":0.46819304245589993,"reference_rms_dbfs":-23.848490648398872,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.310434955453537},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37200,"output_samples":37200,"intelligibility":{"estoi":{"value":0.960563101549704,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0642059782880167,"rms_dbfs":-23.848490648398872,"peak":0.45782470703125,"peak_dbfs":-6.786015476663371,"crest_factor_db":17.062475171735503,"mean":-3.708049815188172e-07,"zero_energy":false},"output":{"rms":0.06276169392793024,"rms_dbfs":-24.046106867787095,"peak":0.46819304245589993,"peak_dbfs":-6.591500887775562,"crest_factor_db":17.45460598001153,"mean":-0.0009077916170629865,"zero_energy":false},"rms_difference_db":-0.19761621938822316},"full_scale":{"reference":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":232,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.868376156315208e-06,"threshold_dbfs":-54.12471302254709,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":3.3093482041015074e-07,"output_mean_power":2.816699171314755e-07,"output_rms_dbfs":-65.50259533992036,"output_zero_energy":false,"output_peak":0.005431456361526505,"output_energy_fraction":1.999132374042453e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":225,"reference_estimated_voiced_frames":107,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":96,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.897196261682243,"absolute_difference_cents_median":1.992127739246523,"absolute_difference_cents_p95":6.938042378521089,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00019668187909197959,"log_rms_mae_db":1.0232663141361487,"rms_envelope_nrmse":0.07719172709931396,"envelope_correlation":0.9948862523979332,"rms_gain_error_db":-0.19761616592610942,"reference_mean_power":0.004131292024738224,"output_mean_power":0.003947519445596506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.0026327800860698e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011612269501705267,"output_power":0.0011287170003672323,"reference_active":true,"delta_db":-0.12332040242595162},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017095886275955477,"output_power":0.0001653145616833796,"reference_active":true,"delta_db":-0.14580510227592441},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217923048678998e-05,"output_power":3.5906539323864754e-05,"reference_active":true,"delta_db":-0.6992510246129396}],"active_band_count":3,"active_band_mae_db":0.3227921764382719}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00018115923356648065,"log_rms_mae_db":0.6887686511819859,"rms_envelope_nrmse":0.04180182289396811,"envelope_correlation":0.9987894553065273,"rms_gain_error_db":-0.19761579633373813,"reference_mean_power":0.004185412442455258,"output_mean_power":0.003999232760974372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.7728315536386115e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066252574844359,"output_power":0.0010323011591992595,"reference_active":true,"delta_db":-0.14053677912597512},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.440983264620606e-05,"output_power":8.243887266915477e-05,"reference_active":true,"delta_db":-0.10260994943589205},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.940091323355611e-06,"output_power":2.5625703005418437e-06,"reference_active":true,"delta_db":-0.5968503192614493}],"active_band_count":3,"active_band_mae_db":0.2799990159411055}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.000176021246464811,"log_rms_mae_db":1.083247133083711,"rms_envelope_nrmse":0.07648177086056884,"envelope_correlation":0.9956093508767728,"rms_gain_error_db":-0.1836547626348982,"reference_mean_power":0.0026395436088586166,"output_mean_power":0.0025302496023061694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.515899729654865e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009091516678957249,"output_power":0.0008940130113708367,"reference_active":true,"delta_db":-0.07292500288991252},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001833276731517994,"output_power":0.00017579847444865938,"reference_active":true,"delta_db":-0.18212924267147879},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7517429693718724e-05,"output_power":2.848666258595474e-05,"reference_active":true,"delta_db":-1.1959150581083182}],"active_band_count":3,"active_band_mae_db":0.48365643455656987}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":0.00016136314602147107,"log_rms_mae_db":0.7722595532107529,"rms_envelope_nrmse":0.04590911209429949,"envelope_correlation":0.998579070071568,"rms_gain_error_db":-0.18365469211727153,"reference_mean_power":0.002674122622482191,"output_mean_power":0.0025633968649769667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.348456388983372e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008380618007305507,"output_power":0.0008213270195652482,"reference_active":true,"delta_db":-0.08759935506653145},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.026333057603631e-05,"output_power":8.491186629279291e-05,"reference_active":true,"delta_db":-0.2654296784814167},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.498114091574834e-06,"output_power":2.120912284530837e-06,"reference_active":true,"delta_db":-0.7108956160336224}],"active_band_count":3,"active_band_mae_db":0.3546415498605235}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":6.667691377903705e-05,"log_rms_mae_db":1.0919310917068457,"rms_envelope_nrmse":0.0731152635615642,"envelope_correlation":0.9962449367312217,"rms_gain_error_db":-0.10662336736890791,"reference_mean_power":0.0003102146198149335,"output_mean_power":0.000302691290315424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":3.2470137873341424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001843563097177963,"output_power":0.00018327788384546215,"reference_active":true,"delta_db":-0.025479444354944567},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0328682781544793e-05,"output_power":1.9304814819286737e-05,"reference_active":true,"delta_db":-0.2244359895390619},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.580926194982453e-06,"output_power":4.797455598983661e-06,"reference_active":true,"delta_db":0.20057669228150773}],"active_band_count":3,"active_band_mae_db":0.15016404205850473}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":6.131102329922278e-05,"log_rms_mae_db":0.6334524650020592,"rms_envelope_nrmse":0.04653114932391341,"envelope_correlation":0.9984826579318563,"rms_gain_error_db":-0.10662318763737708,"reference_mean_power":0.00031427821970459705,"output_mean_power":0.000306656352411981},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":3.052275331975407e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001756954599460375,"output_power":0.0001746068061535472,"reference_active":true,"delta_db":-0.026993708157612345},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.589536328721272e-06,"output_power":8.252073291086125e-06,"reference_active":true,"delta_db":-0.1740664431758368},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8479520444652186e-07,"output_power":3.0049317820258916e-07,"reference_active":true,"delta_db":0.2330194498161484}],"active_band_count":3,"active_band_mae_db":0.1446932003831992}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":2.0872644535346125e-05,"log_rms_mae_db":1.1678021406123378,"rms_envelope_nrmse":0.17223996842006736,"envelope_correlation":0.9800904143868617,"rms_gain_error_db":0.12494815201591933,"reference_mean_power":1.9843186142044422e-05,"output_mean_power":2.0422373825879054e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":2.622000222111226e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.352365504647437e-06,"output_power":8.520935759488225e-06,"reference_active":true,"delta_db":-0.4043418026085563},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.791249366368626e-07,"output_power":1.7528412161459283e-06,"reference_active":true,"delta_db":2.5290446518720504},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.495952213974191e-07,"output_power":4.643884316584272e-07,"reference_active":true,"delta_db":2.6964512618306316}],"active_band_count":3,"active_band_mae_db":1.876612572103746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37200,"uncovered_tail_samples":0,"metrics":{"frame_count":458,"rms_floor":1.9522833601394486e-05,"log_rms_mae_db":0.6839189230700309,"rms_envelope_nrmse":0.10616329348480583,"envelope_correlation":0.9922196678742167,"rms_gain_error_db":0.12494741554553727,"reference_mean_power":2.010291979320634e-05,"output_mean_power":2.0689685136716615e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":458,"frequency_resolution_hz":0.4366812227074236,"power_floor":2.5131001811961107e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.175181026763142e-06,"output_power":7.66362731260185e-06,"reference_active":true,"delta_db":-0.28063001671079785},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.283226898133564e-07,"output_power":7.778978278172549e-07,"reference_active":true,"delta_db":2.5915147761059574},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.754857048422836e-08,"output_power":3.5500800406433715e-08,"reference_active":true,"delta_db":3.0599640044873686}],"active_band_count":3,"active_band_mae_db":1.9773695991013749}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"hnm","status":"measured","mos":{"status":"measured","value":2.308213949203491,"predicted_mos":2.308213949203491,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"92be33756e4a12b6a69c5f338f5a0f9b906bab45fb168ece4c39e968981fd80d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2045473070011212},"started_at":"2026-10-13T04:39:14.913355+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5325999811466318e-05,"native_input_samples":96000,"native_input_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","output_sha256":"1bb64e8b385f2f6350af3ff0e31993f6025bb8975b2b7d3554c6ff0087801604","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9054855379999935,"synthesis_s":0.5216350099999545,"total_s":1.427120547999948,"rtf":0.23785342466665801,"output_peak":0.4581293407352761,"output_rms":0.04121655337943868,"finished_at":"2026-10-13T04:39:16.344989+00:00","evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"1bb64e8b385f2f6350af3ff0e31993f6025bb8975b2b7d3554c6ff0087801604","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.978500441442182,"spectral_convergence":0.41527531043233806,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.691695009445189,"spectral_convergence":0.3990187363655018,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.840086471435262,"spectral_convergence":0.4151790101523915,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":3.064414620475517e-05,"delta_db":0.06809686717750763,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0003405153598580365,"delta_db":0.053114800844760855,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0005987839496253153,"delta_db":0.010808041613740663,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.0001533783626138226,"delta_db":-0.3276251106528666,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.00012544244039418902,"delta_db":0.2448436308492287,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.0004500400137839988,"delta_db":-0.13740984900462905,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.14031638335712224,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.03438707506692868,"pitch":null,"output_rms_dbfs":-27.698566554325907,"output_peak":0.4581293407352761,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.836760640774211},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.978500441442182,"spectral_convergence":0.41527531043233806,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.691695009445189,"spectral_convergence":0.3990187363655018,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.840086471435262,"spectral_convergence":0.4151790101523915,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":3.064414620475517e-05,"delta_db":0.06809686717750763,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0003405153598580365,"delta_db":0.053114800844760855,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0005987839496253153,"delta_db":0.010808041613740663,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.0001533783626138226,"delta_db":-0.3276251106528666,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.00012544244039418902,"delta_db":0.2448436308492287,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.0004500400137839988,"delta_db":-0.13740984900462905,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.14031638335712224,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.03438707506692868,"pitch":null,"output_rms_dbfs":-27.698566554325907,"output_peak":0.4581293407352761,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.836760640774211},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8808471160401535,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04138005142387099,"rms_dbfs":-27.66417947925898,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.24043625530443,"mean":3.353754679361979e-05,"zero_energy":false},"output":{"rms":0.04121655337943868,"rms_dbfs":-27.698566554325907,"peak":0.4581293407352761,"peak_dbfs":-6.7802378618644195,"crest_factor_db":20.91832869246149,"mean":-4.677069602364079e-06,"zero_energy":false},"rms_difference_db":-0.03438707506692751},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":4.22753596466134e-06,"output_rms_dbfs":-53.739126890672935,"output_zero_energy":false,"output_peak":0.06367828892815304,"output_energy_fraction":0.0003566901916939643,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":136,"both_estimated_voiced_frames":129,"reference_voiced_output_unvoiced_frames":34,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.7914110429447853,"absolute_difference_cents_median":1.8347623061527258,"absolute_difference_cents_p95":10.876202332953936,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014203698766261565,"log_rms_mae_db":1.0902794596115246,"rms_envelope_nrmse":0.09035964970073627,"envelope_correlation":0.9918183917741347,"rms_gain_error_db":-0.03438429788188902,"reference_mean_power":0.0017137270951261038,"output_mean_power":0.0017002126122617127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.499341088840719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004635533203375431,"output_power":0.00044577758461190756,"reference_active":true,"delta_db":-0.16981469466921698},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011665881324257648,"output_power":0.00011230488031403818,"reference_active":true,"delta_db":-0.16518924739836027},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5955876441113554e-05,"output_power":2.135214002467008e-05,"reference_active":true,"delta_db":-0.8479428926293073}],"active_band_count":3,"active_band_mae_db":0.39431561156562817}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274207517780892,"log_rms_mae_db":0.6033594101086625,"rms_envelope_nrmse":0.049552876012370434,"envelope_correlation":0.9972126366963167,"rms_gain_error_db":-0.03534785186191865,"reference_mean_power":0.0017200719561396216,"output_mean_power":0.0017061288627469816},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.335969034308142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042020662443664075,"output_power":0.00040392228786735666,"reference_active":true,"delta_db":-0.17165077057831096},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.86830216281553e-05,"output_power":5.6698892734192524e-05,"reference_active":true,"delta_db":-0.14937890169839876},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5858058061398617e-06,"output_power":1.270331631313043e-06,"reference_active":true,"delta_db":-0.9633289143370949}],"active_band_count":3,"active_band_mae_db":0.42811952887126825}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000125392198184175,"log_rms_mae_db":1.5587071222401874,"rms_envelope_nrmse":0.0931103829673363,"envelope_correlation":0.9938261349500163,"rms_gain_error_db":-0.007408104873582923,"reference_mean_power":0.0008371717774342565,"output_mean_power":0.0008357449645019673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.199403135704403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003464696140419235,"output_power":0.00033305635276038913,"reference_active":true,"delta_db":-0.17147430412143555},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952130707503092e-05,"output_power":8.30509444797796e-05,"reference_active":true,"delta_db":-0.3258183883610766},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19766156805787e-05,"output_power":1.2741477053544e-05,"reference_active":true,"delta_db":0.26885662793072507}],"active_band_count":3,"active_band_mae_db":0.2553831068044124}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011133215930497508,"log_rms_mae_db":0.8479043043084272,"rms_envelope_nrmse":0.05700270235568234,"envelope_correlation":0.9975027322889458,"rms_gain_error_db":-0.007479283323077933,"reference_mean_power":0.0008413676993576814,"output_mean_power":0.0008399199692762093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.575111583882504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003227285828337414,"output_power":0.0003092932396474994,"reference_active":true,"delta_db":-0.18467003355491363},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.553081813418887e-05,"output_power":4.262696582016353e-05,"reference_active":true,"delta_db":-0.2862103338583354},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0551344445715747e-06,"output_power":1.0109576336945777e-06,"reference_active":true,"delta_db":-0.18574844742169222}],"active_band_count":3,"active_band_mae_db":0.2188762716116471}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":1.6470804176138079,"rms_envelope_nrmse":0.13838277630468465,"envelope_correlation":0.9857976267506949,"rms_gain_error_db":-0.15322348573505323,"reference_mean_power":0.00018600338319336193,"output_mean_power":0.00017955541004371502},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4579578491485416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.813551789415928e-05,"output_power":7.964034489779727e-05,"reference_active":true,"delta_db":-0.44017828662278524},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0393966851463274e-05,"output_power":1.851915536863613e-05,"reference_active":true,"delta_db":-0.41880533794342756},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.503164081364988e-06,"output_power":3.752066523099034e-06,"reference_active":true,"delta_db":-0.7924724190229461}],"active_band_count":3,"active_band_mae_db":0.550485347863053}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":0.9999927016687472,"rms_envelope_nrmse":0.07913605544270362,"envelope_correlation":0.995162647711932,"rms_gain_error_db":-0.1534810242173759,"reference_mean_power":0.00018692252747526123,"output_mean_power":0.00018043199135803055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3152307217858676e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.510469005068112e-05,"output_power":7.653436401539363e-05,"reference_active":true,"delta_db":-0.460970167078156},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648548828731209e-06,"output_power":9.054066823113673e-06,"reference_active":true,"delta_db":-0.27618303575341474},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6517442110747655e-07,"output_power":2.63360717608236e-07,"reference_active":true,"delta_db":-1.4194935256882188}],"active_band_count":3,"active_band_mae_db":0.7188822428399299}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":1.7293961974183079,"rms_envelope_nrmse":0.1812572127131552,"envelope_correlation":0.9758083564308914,"rms_gain_error_db":-0.15416474380235753,"reference_mean_power":0.00033891979539590525,"output_mean_power":0.00032709993752289874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5434076582307209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.258901619859076e-05,"output_power":6.127871892273042e-05,"reference_active":true,"delta_db":-0.09188447707406083},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4223152137238314e-05,"output_power":1.2852863989405942e-05,"reference_active":true,"delta_db":-0.4399594366489721},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429050160653033e-06,"output_power":7.070957725615106e-06,"reference_active":true,"delta_db":-0.7630039751732588}],"active_band_count":3,"active_band_mae_db":0.43161596296543053}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":1.02533172052439,"rms_envelope_nrmse":0.10089737296490957,"envelope_correlation":0.9921472381583742,"rms_gain_error_db":-0.15787402932208605,"reference_mean_power":0.00033859490771126504,"output_mean_power":0.00032650739301488964},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4129997856527199e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.165966182107606e-05,"output_power":5.982637414517347e-05,"reference_active":true,"delta_db":-0.13108455802566757},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0839649935296936e-06,"output_power":5.989029283235774e-06,"reference_active":true,"delta_db":-0.7291996982898162},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.531889318612957e-07,"output_power":4.2253948345790707e-07,"reference_active":true,"delta_db":-1.8917152004030056}],"active_band_count":3,"active_band_mae_db":0.9173331522394964}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.5266621112823486,"predicted_mos":3.5266621112823486,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"8ba0fdb7fbda3b13e59ec030632520bf693383cc35265fee862a70544dfe4561","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1828324729995074},"started_at":"2026-10-13T04:42:11.227972+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.331199993932387e-05,"native_input_samples":96000,"native_input_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","output_sha256":"6c36e5aecc60c199301e39e2a759407c0e6c2483744dccfa90790ec00e1020ce","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5607237719996192,"synthesis_s":0.10264680799991766,"total_s":1.6633705799995369,"rtf":0.2772284299999228,"output_peak":0.4508901122679361,"output_rms":0.04552891865911877,"finished_at":"2026-10-13T04:42:12.896741+00:00","evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"6c36e5aecc60c199301e39e2a759407c0e6c2483744dccfa90790ec00e1020ce","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.470916323593152,"spectral_convergence":0.6034273131933194,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.552168411728957,"spectral_convergence":0.42788009360946033,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.768188706922885,"spectral_convergence":0.4462262472654307,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":2.703235923696935e-05,"delta_db":-0.4765388878756046,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0004707358920904742,"delta_db":1.4595208649690392,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0007538690265529583,"delta_db":1.011065538416274,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.0002010972189184286,"delta_db":0.8487945632062013,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.0001331380256248039,"delta_db":0.5034198104716835,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.00048700991184501887,"delta_db":0.2054568604557179,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7507994208990868,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8299261924490798,"pitch":null,"output_rms_dbfs":-26.8342532868099,"output_peak":0.4508901122679361,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.930424480748331},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.470916323593152,"spectral_convergence":0.6034273131933194,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.552168411728957,"spectral_convergence":0.42788009360946033,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.768188706922885,"spectral_convergence":0.4462262472654307,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":2.703235923696935e-05,"delta_db":-0.4765388878756046,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0004707358920904742,"delta_db":1.4595208649690392,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0007538690265529583,"delta_db":1.011065538416274,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.0002010972189184286,"delta_db":0.8487945632062013,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.0001331380256248039,"delta_db":0.5034198104716835,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.00048700991184501887,"delta_db":0.2054568604557179,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7507994208990868,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8299261924490798,"pitch":null,"output_rms_dbfs":-26.8342532868099,"output_peak":0.4508901122679361,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.930424480748331},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9104556478738351,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04138005142387099,"rms_dbfs":-27.66417947925898,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.24043625530443,"mean":3.353754679361979e-05,"zero_energy":false},"output":{"rms":0.045528918659118775,"rms_dbfs":-26.8342532868099,"peak":0.4508901122679361,"peak_dbfs":-6.9185857677899865,"crest_factor_db":19.915667519019912,"mean":9.80179321650084e-09,"zero_energy":false},"rms_difference_db":0.8299261924490793},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":2.2995729707688984e-06,"output_rms_dbfs":-56.38352804702922,"output_zero_energy":false,"output_peak":0.03824713088648404,"output_energy_fraction":0.00015900827451404993,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":169,"both_estimated_voiced_frames":142,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.8711656441717791,"absolute_difference_cents_median":5.765329357215338,"absolute_difference_cents_p95":22.651531119494425,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014203698766261565,"log_rms_mae_db":2.3406084840129107,"rms_envelope_nrmse":0.1767232626565117,"envelope_correlation":0.9809949333802003,"rms_gain_error_db":0.8299387332508233,"reference_mean_power":0.0017137270951261038,"output_mean_power":0.0020746055557126692},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.499341088840719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004635533203375431,"output_power":0.0005636417988919378,"reference_active":true,"delta_db":0.8490349612032949},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011665881324257648,"output_power":0.00015365570097794425,"reference_active":true,"delta_db":1.1963112412399006},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5955876441113554e-05,"output_power":2.337582586484746e-05,"reference_active":true,"delta_db":-0.4546873471832264}],"active_band_count":3,"active_band_mae_db":0.8333445165421406}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274207517780892,"log_rms_mae_db":1.6707939636868725,"rms_envelope_nrmse":0.13599133207951422,"envelope_correlation":0.9914784330749085,"rms_gain_error_db":0.8293651464703231,"reference_mean_power":0.0017200719561396216,"output_mean_power":0.0020820115280941517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.335969034308142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042020662443664075,"output_power":0.0005115266983238377,"reference_active":true,"delta_db":0.8540541129910435},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.86830216281553e-05,"output_power":8.083083352461665e-05,"reference_active":true,"delta_db":1.3906458938931472},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5858058061398617e-06,"output_power":1.5653396386278846e-06,"reference_active":true,"delta_db":-0.05641420717474771}],"active_band_count":3,"active_band_mae_db":0.7670380713529795}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000125392198184175,"log_rms_mae_db":1.863210741919233,"rms_envelope_nrmse":0.2004296546150762,"envelope_correlation":0.9897707180808675,"rms_gain_error_db":1.240292650257159,"reference_mean_power":0.0008371717774342565,"output_mean_power":0.0011138939473448104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.199403135704403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003464696140419235,"output_power":0.0004478481548533272,"reference_active":true,"delta_db":1.1146563697528113},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952130707503092e-05,"output_power":0.0001254421559894561,"reference_active":true,"delta_db":1.4651709509986965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19766156805787e-05,"output_power":1.1779575361845827e-05,"reference_active":true,"delta_db":-0.07204478647988655}],"active_band_count":3,"active_band_mae_db":0.8839573690771315}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011133215930497508,"log_rms_mae_db":1.2769950193195192,"rms_envelope_nrmse":0.17607038582402615,"envelope_correlation":0.9950749752242416,"rms_gain_error_db":1.2402446587031797,"reference_mean_power":0.0008413676993576814,"output_mean_power":0.0011194644357253511},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.575111583882504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003227285828337414,"output_power":0.0004164781402418252,"reference_active":true,"delta_db":1.1075478054835706},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.553081813418887e-05,"output_power":6.509765630133838e-05,"reference_active":true,"delta_db":1.552598988979819},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0551344445715747e-06,"output_power":1.130762139678742e-06,"reference_active":true,"delta_db":0.3006345826698873}],"active_band_count":3,"active_band_mae_db":0.9869271257110922}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":1.6472073090864139,"rms_envelope_nrmse":0.17118383859467734,"envelope_correlation":0.986455967457726,"rms_gain_error_db":0.7975947337097415,"reference_mean_power":0.00018600338319336193,"output_mean_power":0.00022350143557077712},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4579578491485416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.813551789415928e-05,"output_power":0.00010717009530172528,"reference_active":true,"delta_db":0.8492265609464551},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0393966851463274e-05,"output_power":2.5173191448036953e-05,"reference_active":true,"delta_db":0.9143656967388197},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.503164081364988e-06,"output_power":3.621385435814796e-06,"reference_active":true,"delta_db":-0.9464302094891073}],"active_band_count":3,"active_band_mae_db":0.9033408223914607}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":1.03183614850163,"rms_envelope_nrmse":0.1237479571391534,"envelope_correlation":0.9955571881900864,"rms_gain_error_db":0.7976907779452205,"reference_mean_power":0.00018692252747526123,"output_mean_power":0.00022461084541944103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3152307217858676e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.510469005068112e-05,"output_power":0.00010296535808249374,"reference_active":true,"delta_db":0.8273763981713695},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648548828731209e-06,"output_power":1.2280824050063225e-05,"reference_active":true,"delta_db":1.0476551014565485},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6517442110747655e-07,"output_power":3.054534454549346e-07,"reference_active":true,"delta_db":-0.7755532122087918}],"active_band_count":3,"active_band_mae_db":0.8835282372789033}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":1.696274847991685,"rms_envelope_nrmse":0.2229854910107494,"envelope_correlation":0.9658099442202077,"rms_gain_error_db":0.30582051769840074,"reference_mean_power":0.00033891979539590525,"output_mean_power":0.0003636461444407952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5434076582307209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.258901619859076e-05,"output_power":6.330377667169058e-05,"reference_active":true,"delta_db":0.04931495404913434},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4223152137238314e-05,"output_power":1.4010108028146905e-05,"reference_active":true,"delta_db":-0.06554371446994174},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429050160653033e-06,"output_power":6.293539009851676e-06,"reference_active":true,"delta_db":-1.268837098279389}],"active_band_count":3,"active_band_mae_db":0.461231922266155}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":1.1251447002103827,"rms_envelope_nrmse":0.11709475614328482,"envelope_correlation":0.9905237025204239,"rms_gain_error_db":0.29947526687107917,"reference_mean_power":0.00033859490771126504,"output_mean_power":0.0003627671465447786},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4129997856527199e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.165966182107606e-05,"output_power":6.121715060514396e-05,"reference_active":true,"delta_db":-0.03128027270085578},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0839649935296936e-06,"output_power":6.93091406484506e-06,"reference_active":true,"delta_db":-0.09485892250909116},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.531889318612957e-07,"output_power":4.007309529685962e-07,"reference_active":true,"delta_db":-2.1218592791863413}],"active_band_count":3,"active_band_mae_db":0.7493328247987628}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"facodec","status":"measured","mos":{"status":"measured","value":4.388001441955566,"predicted_mos":4.388001441955566,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ea97730f265911e1b13e99d56f36c7f281cdeb577d9b3fd6cf2f10863922adc1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0289044990004186},"started_at":"2026-10-13T04:52:22.339500+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4436999410681892e-05,"native_input_samples":96000,"native_input_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","output_sha256":"e54bc7e714470f94a994a630c9ae8e7ea43da442a88ce2df54aba97824cdf693","output_samples":96000,"length_delta_samples":0,"analysis_s":2.1030412800000704,"synthesis_s":6.516400207000515,"total_s":8.619441487000586,"rtf":1.4365735811667644,"output_peak":0.4281197190284729,"output_rms":0.03968696708397303,"finished_at":"2026-10-13T04:52:30.968043+00:00","evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"e54bc7e714470f94a994a630c9ae8e7ea43da442a88ce2df54aba97824cdf693","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.582938808770031,"spectral_convergence":0.4151821390432466,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.96455262039224,"spectral_convergence":0.4371208760539781,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.040679701081385,"spectral_convergence":0.4488646921035142,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":3.1984156852243126e-05,"delta_db":0.25397067590179384,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0003292530047582263,"delta_db":-0.09295479605721504,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.000551536037057766,"delta_db":-0.34615452160073196,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.0001442081074806452,"delta_db":-0.5953693081648332,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.00011355755199886107,"delta_db":-0.1874410951859455,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.0004045164981766173,"delta_db":-0.600558752367904,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3460748582130706,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.36286230123888813,"pitch":null,"output_rms_dbfs":-28.027041780497868,"output_peak":0.4281197190284729,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.862723710081219},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.582938808770031,"spectral_convergence":0.4151821390432466,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.96455262039224,"spectral_convergence":0.4371208760539781,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.040679701081385,"spectral_convergence":0.4488646921035142,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":3.1984156852243126e-05,"delta_db":0.25397067590179384,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0003292530047582263,"delta_db":-0.09295479605721504,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.000551536037057766,"delta_db":-0.34615452160073196,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.0001442081074806452,"delta_db":-0.5953693081648332,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.00011355755199886107,"delta_db":-0.1874410951859455,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.0004045164981766173,"delta_db":-0.600558752367904,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3460748582130706,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.36286230123888813,"pitch":null,"output_rms_dbfs":-28.027041780497868,"output_peak":0.4281197190284729,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.862723710081219},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.91790509278085,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04138005142387099,"rms_dbfs":-27.66417947925898,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.24043625530443,"mean":3.353754679361979e-05,"zero_energy":false},"output":{"rms":0.03968696708397303,"rms_dbfs":-28.027041780497868,"peak":0.4281197190284729,"peak_dbfs":-7.368695365386828,"crest_factor_db":20.65834641511104,"mean":6.206990163542277e-05,"zero_energy":false},"rms_difference_db":-0.36286230123888785},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":5.577610173266006e-07,"output_rms_dbfs":-62.535518424392095,"output_zero_energy":false,"output_peak":0.008085096254944801,"output_energy_fraction":5.075741972864927e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":170,"both_estimated_voiced_frames":154,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9447852760736196,"absolute_difference_cents_median":3.7780362855654657,"absolute_difference_cents_p95":18.196234335503902,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014203698766261565,"log_rms_mae_db":1.6344284673753482,"rms_envelope_nrmse":0.11777268023393653,"envelope_correlation":0.9872673914338497,"rms_gain_error_db":-0.36284970286351553,"reference_mean_power":0.0017137270951261038,"output_mean_power":0.0015763646708003517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.499341088840719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004635533203375431,"output_power":0.00042395319866912865,"reference_active":true,"delta_db":-0.38781779880101364},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011665881324257648,"output_power":0.00011056823917950048,"reference_active":true,"delta_db":-0.23287160661531625},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5955876441113554e-05,"output_power":2.183501381916173e-05,"reference_active":true,"delta_db":-0.7508222698266677}],"active_band_count":3,"active_band_mae_db":0.45717055841433246}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274207517780892,"log_rms_mae_db":1.2553223355545184,"rms_envelope_nrmse":0.09381136973505858,"envelope_correlation":0.9915239611960693,"rms_gain_error_db":-0.3599256172399352,"reference_mean_power":0.0017200719561396216,"output_mean_power":0.0015832666122055144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.335969034308142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042020662443664075,"output_power":0.0003865123977773435,"reference_active":true,"delta_db":-0.36299465737625},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.86830216281553e-05,"output_power":5.5210344431097455e-05,"reference_active":true,"delta_db":-0.26492011355793305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5858058061398617e-06,"output_power":1.3181265617269498e-06,"reference_active":true,"delta_db":-0.8029289196915007}],"active_band_count":3,"active_band_mae_db":0.4769478968752279}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000125392198184175,"log_rms_mae_db":1.6372004582872162,"rms_envelope_nrmse":0.08363882525245439,"envelope_correlation":0.9953879048280092,"rms_gain_error_db":-0.2272751545394726,"reference_mean_power":0.0008371717774342565,"output_mean_power":0.0007944874907598429},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.199403135704403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003464696140419235,"output_power":0.0003310801536047877,"reference_active":true,"delta_db":-0.1973200425573814},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952130707503092e-05,"output_power":8.681732553748285e-05,"reference_active":true,"delta_db":-0.13320011583169697},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19766156805787e-05,"output_power":1.0700407507555155e-05,"reference_active":true,"delta_db":-0.48933796241363237}],"active_band_count":3,"active_band_mae_db":0.27328604026757025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011133215930497508,"log_rms_mae_db":1.3025727613436149,"rms_envelope_nrmse":0.07186927161185169,"envelope_correlation":0.9964885939413036,"rms_gain_error_db":-0.22726440154806116,"reference_mean_power":0.0008413676993576814,"output_mean_power":0.0007984714551601104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.575111583882504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003227285828337414,"output_power":0.000309263608278917,"reference_active":true,"delta_db":-0.1850861227121051},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.553081813418887e-05,"output_power":4.392172769265652e-05,"reference_active":true,"delta_db":-0.1562603906628923},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0551344445715747e-06,"output_power":1.0906668996942355e-06,"reference_active":true,"delta_db":0.1438433237822963}],"active_band_count":3,"active_band_mae_db":0.1617299457190979}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":1.7194396820896858,"rms_envelope_nrmse":0.1547931970647792,"envelope_correlation":0.9830593132381569,"rms_gain_error_db":-0.4223394489847697,"reference_mean_power":0.00018600338319336193,"output_mean_power":0.00016876675451144138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4579578491485416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.813551789415928e-05,"output_power":7.253748641449772e-05,"reference_active":true,"delta_db":-0.8458845858348024},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0393966851463274e-05,"output_power":1.7651803800020478e-05,"reference_active":true,"delta_db":-0.6271261744674907},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.503164081364988e-06,"output_power":3.416589417120479e-06,"reference_active":true,"delta_db":-1.1992498027184002}],"active_band_count":3,"active_band_mae_db":0.8907535210068978}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":1.45372699588044,"rms_envelope_nrmse":0.12561675910239925,"envelope_correlation":0.9883619721925225,"rms_gain_error_db":-0.4224374433785966,"reference_mean_power":0.00018692252747526123,"output_mean_power":0.00016959689636426742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3152307217858676e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.510469005068112e-05,"output_power":7.092584306110716e-05,"reference_active":true,"delta_db":-0.7914898761545994},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648548828731209e-06,"output_power":8.24666592660561e-06,"reference_active":true,"delta_db":-0.6818359743679077},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6517442110747655e-07,"output_power":2.9083316718642763e-07,"reference_active":true,"delta_db":-0.988564166648117}],"active_band_count":3,"active_band_mae_db":0.8206300057235413}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":2.399830898118661,"rms_envelope_nrmse":0.24639246161199668,"envelope_correlation":0.9559088534052773,"rms_gain_error_db":-0.599747602762111,"reference_mean_power":0.00033891979539590525,"output_mean_power":0.00029520395746045804},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5434076582307209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.258901619859076e-05,"output_power":6.003417409982668e-05,"reference_active":true,"delta_db":-0.18099584828138587},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4223152137238314e-05,"output_power":1.4541124595107828e-05,"reference_active":true,"delta_db":0.09602140222398425},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429050160653033e-06,"output_power":7.428174730981712e-06,"reference_active":true,"delta_db":-0.548965273170518}],"active_band_count":3,"active_band_mae_db":0.27532750789196275}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":2.047884193157085,"rms_envelope_nrmse":0.19303349343484003,"envelope_correlation":0.9721429077409324,"rms_gain_error_db":-0.5864765760522167,"reference_mean_power":0.00033859490771126504,"output_mean_power":0.00029582356375193144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4129997856527199e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.165966182107606e-05,"output_power":5.9584908346939214e-05,"reference_active":true,"delta_db":-0.14864862810153856},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0839649935296936e-06,"output_power":6.017486107153233e-06,"reference_active":true,"delta_db":-0.7086131011974601},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.531889318612957e-07,"output_power":5.559761131226292e-07,"reference_active":true,"delta_db":-0.6998268408769325}],"active_band_count":3,"active_band_mae_db":0.519029523391977}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8904922008514404,"predicted_mos":3.8904922008514404,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"d836147a1bb6f781e813c953d0636f647b03139544f09ce579f94f944ebb4c58","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8521719989985286},"started_at":"2026-10-13T07:14:17.424067+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008981875998870237,"native_input_samples":264600,"native_input_sha256":"b8668ac2575ad73e04ef581a2efd33f5d24bee99740ae3e4d37dbd6715342d12","output_sha256":"206238cd2d27d7b8ac125367c4c627d9c3b45cad98e2ec516eb1ec296c0a6786","output_samples":264704,"length_delta_samples":104,"analysis_s":0.19962211099846172,"synthesis_s":5.4431921190007415,"total_s":5.642814229999203,"rtf":0.9404690383332005,"output_peak":0.3737305700778961,"output_rms":0.0372301621868706,"finished_at":"2026-10-13T07:14:23.100495+00:00","evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"206238cd2d27d7b8ac125367c4c627d9c3b45cad98e2ec516eb1ec296c0a6786","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.61120838903185,"output_peak":0.3723503293881384},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.712405897772545,"spectral_convergence":0.5599792574006138,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.748569775931454,"spectral_convergence":0.5543380775840858,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.478740172718064,"spectral_convergence":0.5996742025856513,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":1.8349452301436847e-05,"delta_db":-2.159147331329651,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0002792468859360986,"delta_db":-0.8083688790885404,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0005110577865263515,"delta_db":-0.67719336796055,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.00017093330563915308,"delta_db":0.1430008320207522,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":9.592775566062868e-05,"delta_db":-0.9201584745684896,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.0003018225128383131,"delta_db":-1.8724048369079394,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0967122869793204,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9454161741711344,"pitch":null,"output_rms_dbfs":-28.609595653430112,"output_peak":0.3723503293881384,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.646571948807354},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7851615180978067,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04138005142387099,"rms_dbfs":-27.66417947925898,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.24043625530443,"mean":3.353754679361979e-05,"zero_energy":false},"output":{"rms":0.03711250057463094,"rms_dbfs":-28.609595653430112,"peak":0.3723503293881384,"peak_dbfs":-8.58096515270671,"crest_factor_db":20.028630500723406,"mean":-2.035222990391858e-05,"zero_energy":false},"rms_difference_db":-0.9454161741711324},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":7.618207284207445e-07,"output_rms_dbfs":-61.18147214772637,"output_zero_energy":false,"output_peak":0.00736137895949494,"output_energy_fraction":7.92792533697607e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":163,"both_estimated_voiced_frames":129,"reference_voiced_output_unvoiced_frames":34,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.7914110429447853,"absolute_difference_cents_median":9.486835183224397,"absolute_difference_cents_p95":43.16909546017093,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014203698766261565,"log_rms_mae_db":1.7420287537097272,"rms_envelope_nrmse":0.1615780022568596,"envelope_correlation":0.9830164140613773,"rms_gain_error_db":-0.9454484691368243,"reference_mean_power":0.0017137270951261038,"output_mean_power":0.0013784684049317159},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.499341088840719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004635533203375431,"output_power":0.00038070594288842284,"reference_active":true,"delta_db":-0.8551003990358286},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011665881324257648,"output_power":0.00010097781945798921,"reference_active":true,"delta_db":-0.6269156589626699},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5955876441113554e-05,"output_power":2.0050684060767493e-05,"reference_active":true,"delta_db":-1.1210650424934048}],"active_band_count":3,"active_band_mae_db":0.8676937001639677}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274207517780892,"log_rms_mae_db":1.312242478354926,"rms_envelope_nrmse":0.13693814918414735,"envelope_correlation":0.9898551141125286,"rms_gain_error_db":-0.9436905579300006,"reference_mean_power":0.0017200719561396216,"output_mean_power":0.0013841321597790428},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.335969034308142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042020662443664075,"output_power":0.00034893831150592454,"reference_active":true,"delta_db":-0.8071423944198539},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.86830216281553e-05,"output_power":5.062860641928502e-05,"reference_active":true,"delta_db":-0.6411649450844736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5858058061398617e-06,"output_power":1.2513397273254015e-06,"reference_active":true,"delta_db":-1.02874770915313}],"active_band_count":3,"active_band_mae_db":0.8256850162191526}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000125392198184175,"log_rms_mae_db":2.9738876686923827,"rms_envelope_nrmse":0.15557581633415393,"envelope_correlation":0.9852112803526212,"rms_gain_error_db":-0.6278998211600699,"reference_mean_power":0.0008371717774342565,"output_mean_power":0.0007244769908045386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.199403135704403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003464696140419235,"output_power":0.0003043085770058426,"reference_active":true,"delta_db":-0.5635095905516919},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952130707503092e-05,"output_power":7.08271385794774e-05,"reference_active":true,"delta_db":-1.017267179458066},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19766156805787e-05,"output_power":1.0959216256490274e-05,"reference_active":true,"delta_db":-0.38554616757202465}],"active_band_count":3,"active_band_mae_db":0.6554409791939275}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011133215930497508,"log_rms_mae_db":2.147388417360275,"rms_envelope_nrmse":0.12961924527954702,"envelope_correlation":0.9901537659774168,"rms_gain_error_db":-0.6278465926346823,"reference_mean_power":0.0008413676993576814,"output_mean_power":0.0007281170081533409},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.575111583882504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003227285828337414,"output_power":0.00027992215867733803,"reference_active":true,"delta_db":-0.6180015226610851},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.553081813418887e-05,"output_power":3.616635177963599e-05,"reference_active":true,"delta_db":-1.0000075188908335},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0551344445715747e-06,"output_power":9.460005192403654e-07,"reference_active":true,"delta_db":-0.4741642591381173}],"active_band_count":3,"active_band_mae_db":0.6973911002300119}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":2.8804530128582915,"rms_envelope_nrmse":0.29943560708984707,"envelope_correlation":0.9351330651383613,"rms_gain_error_db":-0.4591634563948171,"reference_mean_power":0.00018600338319336193,"output_mean_power":0.0001673418238314829},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4579578491485416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.813551789415928e-05,"output_power":6.747538488623134e-05,"reference_active":true,"delta_db":-1.1600559032326638},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0393966851463274e-05,"output_power":2.185156292284247e-05,"reference_active":true,"delta_db":0.29980796066420945},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.503164081364988e-06,"output_power":4.683348817511449e-06,"reference_active":true,"delta_db":0.17038733798277045}],"active_band_count":3,"active_band_mae_db":0.5434170672932146}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":2.217969924229995,"rms_envelope_nrmse":0.2553764308145684,"envelope_correlation":0.9493065365447803,"rms_gain_error_db":-0.45974641566921154,"reference_mean_power":0.00018692252747526123,"output_mean_power":0.00016814617915783886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3152307217858676e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.510469005068112e-05,"output_power":6.819738903144761e-05,"reference_active":true,"delta_db":-0.9618574655212633},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648548828731209e-06,"output_power":1.0499887103072438e-05,"reference_active":true,"delta_db":0.3672263043583101},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6517442110747655e-07,"output_power":3.921897361254749e-07,"reference_active":true,"delta_db":0.3099587422278911}],"active_band_count":3,"active_band_mae_db":0.5463475040358216}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":3.147019389549401,"rms_envelope_nrmse":0.2821602046232978,"envelope_correlation":0.9658992690978883,"rms_gain_error_db":-2.0078022116166268,"reference_mean_power":0.00033891979539590525,"output_mean_power":0.00021346010287387062},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5434076582307209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.258901619859076e-05,"output_power":4.1699988029966594e-05,"reference_active":true,"delta_db":-1.763621948681737},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4223152137238314e-05,"output_power":9.079063074863282e-06,"reference_active":true,"delta_db":-1.949548221952229},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429050160653033e-06,"output_power":4.063423510042156e-06,"reference_active":true,"delta_db":-3.1688654930013316}],"active_band_count":3,"active_band_mae_db":2.294011887878433}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":2.7614052776673668,"rms_envelope_nrmse":0.25738802735539174,"envelope_correlation":0.9762984964563761,"rms_gain_error_db":-2.000926308913321,"reference_mean_power":0.00033859490771126504,"output_mean_power":0.0002135933815240076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4129997856527199e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.165966182107606e-05,"output_power":4.054938304926731e-05,"reference_active":true,"delta_db":-1.8201688756911374},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0839649935296936e-06,"output_power":4.0791459905783395e-06,"reference_active":true,"delta_db":-2.3970715761655543},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.531889318612957e-07,"output_power":2.642131379128352e-07,"reference_active":true,"delta_db":-3.930844081371241}],"active_band_count":3,"active_band_mae_db":2.7160281777426447}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"golf","status":"measured","mos":{"status":"measured","value":3.4340341091156006,"predicted_mos":3.4340341091156006,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"681d8c727a4b20cceb0271cc5b88d4dde3c1d29d3f6dfd63f957a18003b1089e","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.574945062999177},"started_at":"2026-10-13T05:02:33.597045+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004661558999941917,"native_input_samples":144000,"native_input_sha256":"c69356e251583fa61f6d877638b9e895188e7e6c2c7cb50afeceee7564f48e3a","output_sha256":"e671660ee2ac1efafadf2a68a5e89b29016b6fe51ad915f40054d2ec2fb0bdb7","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6229307140001765,"synthesis_s":0.11442308599998796,"total_s":0.7373538000001645,"rtf":0.12289230000002742,"output_peak":0.40862420201301575,"output_rms":0.03727904346403127,"finished_at":"2026-10-13T05:02:34.350040+00:00","analysis_stages":{"external_f0_s":0.08652966000045126},"evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"e671660ee2ac1efafadf2a68a5e89b29016b6fe51ad915f40054d2ec2fb0bdb7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.625810766014684,"output_peak":0.40332176046760265},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.452520162985226,"spectral_convergence":0.5582736337687831,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.646951619105266,"spectral_convergence":0.4456486319312517,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6917188694426155,"spectral_convergence":0.4723616811680338,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.021527927628418e-05,"output_power":3.196500227148962e-05,"delta_db":0.2444812689560878,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00033693028555928427,"output_power":0.00024805069697689974,"delta_db":-1.3299959838188746,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005982910315877378,"output_power":0.00040081147168364727,"delta_db":-1.7397234975215317,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001656712915868069,"output_power":0.0001290296538774918,"delta_db":-1.085577257746528,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011875506324800886,"output_power":9.137578072539523e-05,"delta_db":-1.1382103465247098,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0004652687862066619,"output_power":0.00047097216045066946,"delta_db":0.052913186750686146,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9318169235530697,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9687855996377324,"pitch":null,"output_rms_dbfs":-28.625810766014684,"output_peak":0.40332176046760265,"reference_rms_dbfs":-27.65702516637695,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.930396883844369},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8964709303587746,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041414149000852156,"rms_dbfs":-27.65702516637695,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.2332819424224,"mean":3.3794460392157505e-05,"zero_energy":false},"output":{"rms":0.03704328233277382,"rms_dbfs":-28.625810766014684,"peak":0.40332176046760265,"peak_dbfs":-7.8869669163535425,"crest_factor_db":20.738843849661137,"mean":0.00016348167431894973,"zero_energy":false},"rms_difference_db":-0.9687855996377337},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":7.655905717033101e-07,"output_rms_dbfs":-61.16003423562507,"output_zero_energy":false,"output_peak":0.005196760654894305,"output_energy_fraction":8.010309229630466e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":172,"both_estimated_voiced_frames":146,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8957055214723927,"absolute_difference_cents_median":4.196119208400273,"absolute_difference_cents_p95":17.7460744757589,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014203698766261565,"log_rms_mae_db":2.0978510399155406,"rms_envelope_nrmse":0.19190696462191767,"envelope_correlation":0.9719158189639499,"rms_gain_error_db":-0.968924154449244,"reference_mean_power":0.001716551784129927,"output_mean_power":0.0013732970720181203},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.502827120608417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004637505557808398,"output_power":0.00036640827393076943,"reference_active":true,"delta_db":-1.023191711147729},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001156765295289749,"output_power":8.881830532397101e-05,"reference_active":true,"delta_db":-1.1474276818721727},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5904857657422754e-05,"output_power":2.570504951292037e-05,"reference_active":true,"delta_db":-0.033627655194757286}],"active_band_count":3,"active_band_mae_db":0.734749016071553}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001274207517780892,"log_rms_mae_db":1.5441048396854427,"rms_envelope_nrmse":0.15643111669666676,"envelope_correlation":0.9834232811124424,"rms_gain_error_db":-0.9708455600784477,"reference_mean_power":0.0017188845702174676,"output_mean_power":0.0013745551112978828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.342912449594675e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004201328290131312,"output_power":0.0003339579937264043,"reference_active":true,"delta_db":-0.9969477512484398},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.796380348013076e-05,"output_power":4.16500233632666e-05,"reference_active":true,"delta_db":-1.4354162599685305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.583997755891929e-06,"output_power":1.490469935065606e-06,"reference_active":true,"delta_db":-0.26431341863749613}],"active_band_count":3,"active_band_mae_db":0.8988924766181555}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.000125392198184175,"log_rms_mae_db":2.2112129513268566,"rms_envelope_nrmse":0.22587019635053826,"envelope_correlation":0.9811504703705825,"rms_gain_error_db":-1.6241038042608515,"reference_mean_power":0.0008385669624887034,"output_mean_power":0.0005769356395313027},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.201392931283267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003465464407251047,"output_power":0.0002528931540699282,"reference_active":true,"delta_db":-1.3682436973832959},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.909534886808499e-05,"output_power":5.4648035352753e-05,"reference_active":true,"delta_db":-2.1228047944827684},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1940184448709828e-05,"output_power":9.573074487287806e-06,"reference_active":true,"delta_db":-0.9595959757259462}],"active_band_count":3,"active_band_mae_db":1.4835481558640036}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00011133215930497508,"log_rms_mae_db":1.703773312913599,"rms_envelope_nrmse":0.1944534550053468,"envelope_correlation":0.9919059933832188,"rms_gain_error_db":-1.624186675188404,"reference_mean_power":0.0008427723550437344,"output_mean_power":0.0005798178934416891},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.576998798556237e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032272094520945744,"output_power":0.00023385440264673648,"reference_active":true,"delta_db":-1.398816023229196},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.515895221561515e-05,"output_power":2.7816808554001804e-05,"reference_active":true,"delta_db":-2.1043655526742073},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0532426920592097e-06,"output_power":8.025093442037712e-07,"reference_active":true,"delta_db":-1.1807835653539827}],"active_band_count":3,"active_band_mae_db":1.561321713752462}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":2.0754284428176293,"rms_envelope_nrmse":0.22115534740912351,"envelope_correlation":0.9740654338802162,"rms_gain_error_db":-1.2843405544602051,"reference_mean_power":0.00018631238745332526,"output_mean_power":0.00013861418492574763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.455965259875329e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.807133195791242e-05,"output_power":6.647088995199756e-05,"reference_active":true,"delta_db":-1.222030710932097},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.028427974508902e-05,"output_power":1.752087367666734e-05,"reference_active":true,"delta_db":-0.636038329462914},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.500374827106113e-06,"output_power":4.106096359278248e-06,"reference_active":true,"delta_db":-0.39819549886700956}],"active_band_count":3,"active_band_mae_db":0.7520881797540069}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":1.5796062969568871,"rms_envelope_nrmse":0.18156720853146374,"envelope_correlation":0.9866478715313162,"rms_gain_error_db":-1.2850515583680135,"reference_mean_power":0.00018722163183341342,"output_mean_power":0.00013926784979441547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.313531788366187e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.500977654470149e-05,"output_power":6.471692781049467e-05,"reference_active":true,"delta_db":-1.1845098198151296},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.586830493198115e-06,"output_power":8.340719975276844e-06,"reference_active":true,"delta_db":-0.6047150780457101},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.640223848556034e-07,"output_power":3.3262223015908466e-07,"reference_active":true,"delta_db":-0.3917681949358287}],"active_band_count":3,"active_band_mae_db":0.7269976975988895}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":2.589431197185852,"rms_envelope_nrmse":0.24582566722877627,"envelope_correlation":0.9682946705363499,"rms_gain_error_db":0.698007799169974,"reference_mean_power":0.0003394608503527465,"output_mean_power":0.00039864881246887455},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.5460960424988193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.241684353642754e-05,"output_power":8.342178584153303e-05,"reference_active":true,"delta_db":1.2597768040662638},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4139309606251316e-05,"output_power":2.024504020640854e-05,"reference_active":true,"delta_db":1.5589043934718296},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.444189814427059e-06,"output_power":1.3182861646451023e-05,"reference_active":true,"delta_db":1.9345170707902384}],"active_band_count":3,"active_band_mae_db":1.5843994227761105}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":2.255338185902293,"rms_envelope_nrmse":0.19032746183172825,"envelope_correlation":0.9829108222347122,"rms_gain_error_db":0.7117390864793709,"reference_mean_power":0.0003360198057336414,"output_mean_power":0.00039585741564912786},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.4136382968399707e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.143172895283717e-05,"output_power":8.430653585993003e-05,"reference_active":true,"delta_db":1.3746850618310171},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.035910571444463e-06,"output_power":9.463893421085298e-06,"reference_active":true,"delta_db":1.287495299753857},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.517850534913227e-07,"output_power":7.979648797228246e-07,"reference_active":true,"delta_db":0.8787938031528788}],"active_band_count":3,"active_band_mae_db":1.180324721579251}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4319283962249756,"predicted_mos":3.4319283962249756,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"519a7ff42b32cafcda6dec245ac8301b6b17359a3f4a4d086c062ced1821964b","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6235122839971154},"started_at":"2026-10-13T07:37:28.242754+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030799149972153828,"native_input_samples":144000,"native_input_sha256":"c69356e251583fa61f6d877638b9e895188e7e6c2c7cb50afeceee7564f48e3a","output_sha256":"12effa86c1ac9fae4bf9d1ce771cc63f231b16f3935bb4e1d56bea3fa0ddcff5","output_samples":144120,"length_delta_samples":120,"analysis_s":2.1910007580008823,"synthesis_s":13.687510944000678,"total_s":15.87851170200156,"rtf":2.64641861700026,"output_peak":0.37139272689819336,"output_rms":0.038919470730792036,"finished_at":"2026-10-13T07:37:44.134155+00:00","analysis_stages":{"external_f0_s":1.4858086180029204},"evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"12effa86c1ac9fae4bf9d1ce771cc63f231b16f3935bb4e1d56bea3fa0ddcff5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.233790829469516,"output_peak":0.367311986678955},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.660606522488142,"spectral_convergence":0.624807062704678,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7823891202310005,"spectral_convergence":0.4457096309223955,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.466618262716834,"spectral_convergence":0.44614857836548794,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":3.0348169644043653e-05,"delta_db":0.025946640775951692,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.00032234962190032983,"delta_db":-0.18498061480736364,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0005467302675739247,"delta_db":-0.38416228857807866,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.00015835198424580453,"delta_db":-0.1890309851209666,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.00010598244343038797,"delta_db":-0.4872620325627603,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.00033930877645016,"delta_db":-1.3639702110602223,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4392254621508906,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5660248033601353,"pitch":null,"output_rms_dbfs":-28.230204282619113,"output_peak":0.367311986678955,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.969871301811992},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.888442925297455,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04138005142387099,"rms_dbfs":-27.66417947925898,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.24043625530443,"mean":3.353754679361979e-05,"zero_energy":false},"output":{"rms":0.03876946302497174,"rms_dbfs":-28.230204282619113,"peak":0.367311986678955,"peak_dbfs":-8.699297977412957,"crest_factor_db":19.53090630520616,"mean":0.00020860025739678187,"zero_energy":false},"rms_difference_db":-0.5660248033601327},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":2.17220193787478e-05,"output_rms_dbfs":-46.63099803195504,"output_zero_energy":false,"output_peak":0.3481261817692837,"output_energy_fraction":0.0020714183820973037,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":144,"both_estimated_voiced_frames":132,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.8098159509202454,"absolute_difference_cents_median":5.803584889382996,"absolute_difference_cents_p95":27.217466088112975,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014203698766261565,"log_rms_mae_db":3.3064153658398983,"rms_envelope_nrmse":0.19336180804286268,"envelope_correlation":0.9647752084544451,"rms_gain_error_db":-0.5660189824108433,"reference_mean_power":0.0017137270951261038,"output_mean_power":0.0015043183910903684},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.499341088840719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004635533203375431,"output_power":0.0004370454598272975,"reference_active":true,"delta_db":-0.25573083185156364},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011665881324257648,"output_power":0.00010543257572158243,"reference_active":true,"delta_db":-0.43942737608657245},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5955876441113554e-05,"output_power":1.964644367552584e-05,"reference_active":true,"delta_db":-1.2095175060943046}],"active_band_count":3,"active_band_mae_db":0.6348919046774802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274207517780892,"log_rms_mae_db":2.324951344055642,"rms_envelope_nrmse":0.12840771929382283,"envelope_correlation":0.9849920644080641,"rms_gain_error_db":-0.562508454243337,"reference_mean_power":0.0017200719561396216,"output_mean_power":0.00151110892206732},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.335969034308142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042020662443664075,"output_power":0.0003956263366638025,"reference_active":true,"delta_db":-0.2617769997959339},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.86830216281553e-05,"output_power":5.2701855025411606e-05,"reference_active":true,"delta_db":-0.46686565872190655},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5858058061398617e-06,"output_power":1.2969673682049371e-06,"reference_active":true,"delta_db":-0.8732095426840967}],"active_band_count":3,"active_band_mae_db":0.533950733733979}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000125392198184175,"log_rms_mae_db":2.4195506200965378,"rms_envelope_nrmse":0.19211752723636963,"envelope_correlation":0.9736689202632743,"rms_gain_error_db":-0.3002050482617632,"reference_mean_power":0.0008371717774342565,"output_mean_power":0.0007812572745609647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.199403135704403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003464696140419235,"output_power":0.00033034575113016285,"reference_active":true,"delta_db":-0.20696426940161172},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952130707503092e-05,"output_power":8.40051590472191e-05,"reference_active":true,"delta_db":-0.27620456169870033},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19766156805787e-05,"output_power":9.616644640309187e-06,"reference_active":true,"delta_db":-0.9531054559069889}],"active_band_count":3,"active_band_mae_db":0.4787580956691004}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011133215930497508,"log_rms_mae_db":1.6008771756592384,"rms_envelope_nrmse":0.12042728887681586,"envelope_correlation":0.9893874856431921,"rms_gain_error_db":-0.30017142989424345,"reference_mean_power":0.0008413676993576814,"output_mean_power":0.0007851790298429921},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.575111583882504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003227285828337414,"output_power":0.0003072971477045165,"reference_active":true,"delta_db":-0.21278901699663333},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.553081813418887e-05,"output_power":4.222618206373079e-05,"reference_active":true,"delta_db":-0.3272363824093726},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0551344445715747e-06,"output_power":9.040091235420583e-07,"reference_active":true,"delta_db":-0.6713498715818188}],"active_band_count":3,"active_band_mae_db":0.40379175699594155}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":2.503139617927439,"rms_envelope_nrmse":0.23496522220868432,"envelope_correlation":0.9585488482545272,"rms_gain_error_db":-0.2661091677310124,"reference_mean_power":0.00018600338319336193,"output_mean_power":0.00017494838061309624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4579578491485416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.813551789415928e-05,"output_power":8.565212898313029e-05,"reference_active":true,"delta_db":-0.12412798487953958},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0393966851463274e-05,"output_power":1.822377135490749e-05,"reference_active":true,"delta_db":-0.48863451180631906},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.503164081364988e-06,"output_power":4.077009706215894e-06,"reference_active":true,"delta_db":-0.43176026211654683}],"active_band_count":3,"active_band_mae_db":0.3481742529341352}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":1.5711698919242942,"rms_envelope_nrmse":0.1389212186439906,"envelope_correlation":0.9845536375030949,"rms_gain_error_db":-0.2659366957166451,"reference_mean_power":0.00018692252747526123,"output_mean_power":0.0001758198783131251},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3152307217858676e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.510469005068112e-05,"output_power":8.219089064764259e-05,"reference_active":true,"delta_db":-0.15129807730517822},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648548828731209e-06,"output_power":8.733353220597676e-06,"reference_active":true,"delta_db":-0.4328097347605664},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6517442110747655e-07,"output_power":3.6287421558653025e-07,"reference_active":true,"delta_db":-0.027442394326908218}],"active_band_count":3,"active_band_mae_db":0.20385006879755094}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":2.506968025664297,"rms_envelope_nrmse":0.22453060465256724,"envelope_correlation":0.9645434598431145,"rms_gain_error_db":-0.677414025499281,"reference_mean_power":0.00033891979539590525,"output_mean_power":0.0002899716454661763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5434076582307209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.258901619859076e-05,"output_power":5.537685382240685e-05,"reference_active":true,"delta_db":-0.5316984705006061},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4223152137238314e-05,"output_power":1.1868940892919878e-05,"reference_active":true,"delta_db":-0.7858388840377033},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429050160653033e-06,"output_power":6.298013100821393e-06,"reference_active":true,"delta_db":-1.2657507859894077}],"active_band_count":3,"active_band_mae_db":0.8610960468425723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":1.6749561805258282,"rms_envelope_nrmse":0.137357629050895,"envelope_correlation":0.9885490671246356,"rms_gain_error_db":-0.6615849739730773,"reference_mean_power":0.00033859490771126504,"output_mean_power":0.000290751473832677},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4129997856527199e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.165966182107606e-05,"output_power":5.3609060361296454e-05,"reference_active":true,"delta_db":-0.6076294331538747},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0839649935296936e-06,"output_power":5.902388058710555e-06,"reference_active":true,"delta_db":-0.7924864714226121},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.531889318612957e-07,"output_power":4.616653059690878e-07,"reference_active":true,"delta_db":-1.5071157844821135}],"active_band_count":3,"active_band_mae_db":0.9690772296862001}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.047745704650879,"predicted_mos":3.047745704650879,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"3483a64e64149163d4451857c7e2a7af236104aa1f6f95169a9b2e8032d776b2","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.261878417000844},"started_at":"2026-10-13T05:10:42.956340+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0051277210004627705,"native_input_samples":144000,"native_input_sha256":"c69356e251583fa61f6d877638b9e895188e7e6c2c7cb50afeceee7564f48e3a","output_sha256":"70a85a18dbe243674f4480b3b515342d920ff02becc5c82b37a12c771ae2a547","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7975658110008226,"synthesis_s":0.11369048800042947,"total_s":0.911256299001252,"rtf":0.151876049833542,"output_peak":0.4485328495502472,"output_rms":0.03951036491789118,"finished_at":"2026-10-13T05:10:43.888516+00:00","analysis_stages":{"external_f0_s":0.18157566099944233},"evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"70a85a18dbe243674f4480b3b515342d920ff02becc5c82b37a12c771ae2a547","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.12063531074413,"output_peak":0.4459821752438482},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.0583933845287765,"spectral_convergence":0.3510731988658009,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.690875807376807,"spectral_convergence":0.33185920009273123,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.117465610499979,"spectral_convergence":0.37507291287375044,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":2.9279092191782144e-05,"delta_db":-0.12980231848650822,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.00031099302247363593,"delta_db":-0.34074581439098145,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0005700033181991052,"delta_db":-0.2031196255529547,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.00015781933591913877,"delta_db":-0.20366396623969957,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.00010531903349532301,"delta_db":-0.5145326626518183,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.0003706254212830622,"delta_db":-0.980569098527384,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3954055809748911,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.44923723457681614,"pitch":null,"output_rms_dbfs":-28.113416713835797,"output_peak":0.4459821752438482,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.95557826746852},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9479323929017359,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04138005142387099,"rms_dbfs":-27.66417947925898,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.24043625530443,"mean":3.353754679361979e-05,"zero_energy":false},"output":{"rms":0.03929426451229298,"rms_dbfs":-28.113416713835797,"peak":0.4459821752438482,"peak_dbfs":-7.013649971404413,"crest_factor_db":21.099766742431385,"mean":1.0385555204253397e-05,"zero_energy":false},"rms_difference_db":-0.44923723457681675},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":3.2582963005185116e-06,"output_rms_dbfs":-54.87009424684041,"output_zero_energy":false,"output_peak":0.06193429729804232,"output_energy_fraction":0.0003024680089820395,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":163,"both_estimated_voiced_frames":156,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9570552147239264,"absolute_difference_cents_median":0.9049931803801908,"absolute_difference_cents_p95":6.573226606866689,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014203698766261565,"log_rms_mae_db":1.079561834045014,"rms_envelope_nrmse":0.10351620348405388,"envelope_correlation":0.9914214331895905,"rms_gain_error_db":-0.44923129582052307,"reference_mean_power":0.0017137270951261038,"output_mean_power":0.0015453203852501898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.499341088840719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004635533203375431,"output_power":0.00043923675835648554,"reference_active":true,"delta_db":-0.2340101847037548},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011665881324257648,"output_power":8.793515254542938e-05,"reference_active":true,"delta_db":-1.2275503277034703},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5955876441113554e-05,"output_power":2.1843775391820044e-05,"reference_active":true,"delta_db":-0.7490799583853157}],"active_band_count":3,"active_band_mae_db":0.7368801569308469}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274207517780892,"log_rms_mae_db":0.7077857036817282,"rms_envelope_nrmse":0.08087104373708477,"envelope_correlation":0.9952666286523302,"rms_gain_error_db":-0.4493657969540344,"reference_mean_power":0.0017200719561396216,"output_mean_power":0.0015509937063401898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.335969034308142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042020662443664075,"output_power":0.0003990982078094329,"reference_active":true,"delta_db":-0.22383117106132236},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.86830216281553e-05,"output_power":4.435587457038202e-05,"reference_active":true,"delta_db":-1.2156132115585563},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5858058061398617e-06,"output_power":1.240167198872906e-06,"reference_active":true,"delta_db":-1.0676976308121624}],"active_band_count":3,"active_band_mae_db":0.8357140044773471}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000125392198184175,"log_rms_mae_db":1.04048413771753,"rms_envelope_nrmse":0.09488330953719368,"envelope_correlation":0.9939896671694912,"rms_gain_error_db":-0.2469864280733046,"reference_mean_power":0.0008371717774342565,"output_mean_power":0.0007908897302867165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.199403135704403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003464696140419235,"output_power":0.00033238391890023396,"reference_active":true,"delta_db":-0.1802514829832808},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952130707503092e-05,"output_power":6.682446418618176e-05,"reference_active":true,"delta_db":-1.2699092940607462},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19766156805787e-05,"output_power":1.0455510631142076e-05,"reference_active":true,"delta_db":-0.5898886568836148}],"active_band_count":3,"active_band_mae_db":0.6800164779758807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011133215930497508,"log_rms_mae_db":0.5784958661901705,"rms_envelope_nrmse":0.06997438145265321,"envelope_correlation":0.9967787110192164,"rms_gain_error_db":-0.24698014552903164,"reference_mean_power":0.0008413676993576814,"output_mean_power":0.0007948548354949483},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.575111583882504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003227285828337414,"output_power":0.0003076988682810261,"reference_active":true,"delta_db":-0.2071153199344109},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.553081813418887e-05,"output_power":3.3947952663732195e-05,"reference_active":true,"delta_db":-1.2749186619471629},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0551344445715747e-06,"output_power":7.779397143195683e-07,"reference_active":true,"delta_db":-1.3236185762524377}],"active_band_count":3,"active_band_mae_db":0.9352175193780038}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":1.1194410903562806,"rms_envelope_nrmse":0.10762438044663349,"envelope_correlation":0.9921560718890465,"rms_gain_error_db":-0.33460286187973465,"reference_mean_power":0.00018600338319336193,"output_mean_power":0.0001722108686740079},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4579578491485416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.813551789415928e-05,"output_power":8.341901281130239e-05,"reference_active":true,"delta_db":-0.23885914760167884},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0393966851463274e-05,"output_power":1.7133872161182444e-05,"reference_active":true,"delta_db":-0.7564618683131676},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.503164081364988e-06,"output_power":4.1578842234010935e-06,"reference_active":true,"delta_db":-0.34645379397017484}],"active_band_count":3,"active_band_mae_db":0.4472582699616738}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":0.6848610839690586,"rms_envelope_nrmse":0.06946426863229531,"envelope_correlation":0.997120549206898,"rms_gain_error_db":-0.33453452182993976,"reference_mean_power":0.00018692252747526123,"output_mean_power":0.0001730645799120709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3152307217858676e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.510469005068112e-05,"output_power":8.033659126639594e-05,"reference_active":true,"delta_db":-0.25040093971552857},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648548828731209e-06,"output_power":8.17777271517829e-06,"reference_active":true,"delta_db":-0.718269629960384},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6517442110747655e-07,"output_power":3.220377485072976e-07,"reference_active":true,"delta_db":-0.5459356781591492}],"active_band_count":3,"active_band_mae_db":0.504868749278354}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":1.1119817541499346,"rms_envelope_nrmse":0.12460187175212305,"envelope_correlation":0.989978789757554,"rms_gain_error_db":-0.44495314821531196,"reference_mean_power":0.00033891979539590525,"output_mean_power":0.00030591559758484703},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5434076582307209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.258901619859076e-05,"output_power":5.774319484441855e-05,"reference_active":true,"delta_db":-0.3499731610499865},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4223152137238314e-05,"output_power":1.1822856208977768e-05,"reference_active":true,"delta_db":-0.8027344779793981},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429050160653033e-06,"output_power":6.40715460762347e-06,"reference_active":true,"delta_db":-1.191134344523902}],"active_band_count":3,"active_band_mae_db":0.7812806611844288}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":0.793498122149497,"rms_envelope_nrmse":0.080682608407622,"envelope_correlation":0.9966884212366226,"rms_gain_error_db":-0.44594750943556055,"reference_mean_power":0.00033859490771126504,"output_mean_power":0.0003055523803203086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4129997856527199e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.165966182107606e-05,"output_power":5.6494892251862175e-05,"reference_active":true,"delta_db":-0.3799195377826989},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0839649935296936e-06,"output_power":5.8865045548451454e-06,"reference_active":true,"delta_db":-0.804189219903233},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.531889318612957e-07,"output_power":4.6685631115950355e-07,"reference_active":true,"delta_db":-1.4585558294762646}],"active_band_count":3,"active_band_mae_db":0.8808881957207322}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4446-2273-0010","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2286510467529297,"predicted_mos":3.2286510467529297,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288745,"input_sample_rate":48000,"input_seconds":6.015520833333333,"input_float32_sha256":"7d4c0fb0137063e0e10c3cc8194f5ac7d287524a3e1e949e1172a231c8a60623","internal_samples":96249,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.09504982100043},"started_at":"2026-10-13T05:15:26.011820+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.015974448000633856,"native_input_samples":288000,"native_input_sha256":"105a8c4a870a6d6c23ce7a5b4d99bfbc9b3c2b263da3a21beac5a219c39bd97c","output_sha256":"5d81d0113677bfb57841f6b6f20f6d1c638dd9acdb100fd3fa59a5f1dc1532c5","output_samples":288745,"length_delta_samples":745,"analysis_s":4.337490314999741,"synthesis_s":1.3823603410000942,"total_s":5.719850655999835,"rtf":0.9533084426666392,"output_peak":0.4510690442442886,"output_rms":0.04062241134549755,"finished_at":"2026-10-13T05:15:31.761339+00:00","evaluation_input":{"reference":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"output_sha256":"5d81d0113677bfb57841f6b6f20f6d1c638dd9acdb100fd3fa59a5f1dc1532c5","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96249,"length_delta_samples":249,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.883183013347328,"output_peak":0.4512745569390381},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9974129601346508,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.40630825327047,"spectral_convergence":0.6523332067115177,"spectral_convergence_status":"ok","amplitude_floor":9.383482434461407e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.578994191159281,"spectral_convergence":0.49524525406333564,"spectral_convergence_status":"ok","amplitude_floor":6.715789372234631e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.09548702587272,"spectral_convergence":0.44885604104535093,"spectral_convergence_status":"ok","amplitude_floor":5.217365851457555e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.0167397027219553e-05,"output_power":2.4693818825790445e-05,"delta_db":-0.8694958151719544,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003363761740772986,"output_power":0.0003183089720247668,"delta_db":-0.23976346575757382,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005972956425120644,"output_power":0.0005765910485916725,"delta_db":-0.15321450614159327,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016539661163734624,"output_power":0.00016586359144556403,"delta_db":0.01224456703981409,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011856598981285755,"output_power":0.00012412779522793165,"delta_db":0.19908910239975594,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00046450684077542097,"output_power":0.0004226968723098153,"delta_db":-0.4096307988013401,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3139063758853386,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20786834419580413,"pitch":null,"output_rms_dbfs":-27.872047823454785,"output_peak":0.4512745569390381,"reference_rms_dbfs":-27.66417947925898,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.693596490100823},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8453074982802823,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04138005142387099,"rms_dbfs":-27.66417947925898,"peak":0.4254150390625,"peak_dbfs":-7.423743223954546,"crest_factor_db":20.24043625530443,"mean":3.353754679361979e-05,"zero_energy":false},"output":{"rms":0.040401511090868134,"rms_dbfs":-27.872047823454785,"peak":0.4512745569390381,"peak_dbfs":-6.91118303012748,"crest_factor_db":20.960864793327303,"mean":-8.707664226466659e-05,"zero_energy":false},"rms_difference_db":-0.20786834419580558},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0174505864270036e-06,"threshold_dbfs":-56.951970936839054,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.0081545888381295e-06,"output_mean_power":9.378500000098302e-07,"output_rms_dbfs":-60.278666172470835,"output_zero_energy":false,"output_peak":0.013816323997560769,"output_energy_fraction":8.235412665356e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":163,"output_estimated_voiced_frames":163,"both_estimated_voiced_frames":141,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.8650306748466258,"absolute_difference_cents_median":3.027984827362981,"absolute_difference_cents_p95":26.206341461047163,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014203698766261565,"log_rms_mae_db":2.592717298663017,"rms_envelope_nrmse":0.25292305196222464,"envelope_correlation":0.9356153082965184,"rms_gain_error_db":-0.20811033823189273,"reference_mean_power":0.0017137270951261038,"output_mean_power":0.0016335432199465498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.499341088840719e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004635533203375431,"output_power":0.0004569975964581132,"reference_active":true,"delta_db":-0.061857802468809446},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011665881324257648,"output_power":0.0001119214005139712,"reference_active":true,"delta_db":-0.18004418080314372},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5955876441113554e-05,"output_power":2.9362321634886965e-05,"reference_active":true,"delta_db":0.5355469351651861}],"active_band_count":3,"active_band_mae_db":0.25914963947904646}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274207517780892,"log_rms_mae_db":1.663114475685388,"rms_envelope_nrmse":0.1628320129347081,"envelope_correlation":0.9699459607140747,"rms_gain_error_db":-0.22223425094421012,"reference_mean_power":0.0017200719561396216,"output_mean_power":0.0016342676729842502},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.335969034308142e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042020662443664075,"output_power":0.00041226208441484094,"reference_active":true,"delta_db":-0.08289499909871165},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.86830216281553e-05,"output_power":5.711568104051478e-05,"reference_active":true,"delta_db":-0.11757108224565785},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5858058061398617e-06,"output_power":1.9353925018653593e-06,"reference_active":true,"delta_db":0.865190505567174}],"active_band_count":3,"active_band_mae_db":0.35521886230384786}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000125392198184175,"log_rms_mae_db":2.8928041787390995,"rms_envelope_nrmse":0.23826866797113763,"envelope_correlation":0.9592908795784866,"rms_gain_error_db":-0.18677935701016044,"reference_mean_power":0.0008371717774342565,"output_mean_power":0.0008019303374260863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.199403135704403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003464696140419235,"output_power":0.00034055981062150735,"reference_active":true,"delta_db":-0.07471756607037007},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952130707503092e-05,"output_power":8.604874711217752e-05,"reference_active":true,"delta_db":-0.17181863285466198},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.19766156805787e-05,"output_power":9.817884491736129e-06,"reference_active":true,"delta_db":-0.8631619535642139}],"active_band_count":3,"active_band_mae_db":0.369899384163082}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011133215930497508,"log_rms_mae_db":1.9871138569306728,"rms_envelope_nrmse":0.16364126477205035,"envelope_correlation":0.9794691341196385,"rms_gain_error_db":-0.18674235088849522,"reference_mean_power":0.0008413676993576814,"output_mean_power":0.0008059564960430359},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.575111583882504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003227285828337414,"output_power":0.0003159641773624947,"reference_active":true,"delta_db":-0.09199583906012485},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.553081813418887e-05,"output_power":4.419543246798195e-05,"reference_active":true,"delta_db":-0.1292806614781061},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0551344445715747e-06,"output_power":9.983922523410588e-07,"reference_active":true,"delta_db":-0.2400659852354763}],"active_band_count":3,"active_band_mae_db":0.15378082859123574}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.27719600595036e-05,"log_rms_mae_db":2.723460953935904,"rms_envelope_nrmse":0.30648975120403504,"envelope_correlation":0.930504730702014,"rms_gain_error_db":-0.026134906671344757,"reference_mean_power":0.00018600338319336193,"output_mean_power":0.00018488741608163348},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4579578491485416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.813551789415928e-05,"output_power":8.907964867147455e-05,"reference_active":true,"delta_db":0.04627534698463773},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0393966851463274e-05,"output_power":2.036917606991991e-05,"reference_active":true,"delta_db":-0.005282468515495432},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.503164081364988e-06,"output_power":4.08269571541945e-06,"reference_active":true,"delta_db":-0.42570758571891487}],"active_band_count":3,"active_band_mae_db":0.15908846707301602}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.5906277437135206e-05,"log_rms_mae_db":1.742912690671652,"rms_envelope_nrmse":0.2000257344739804,"envelope_correlation":0.9675038652990404,"rms_gain_error_db":-0.026369337746147168,"reference_mean_power":0.00018692252747526123,"output_mean_power":0.00018579101653739345},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3152307217858676e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.510469005068112e-05,"output_power":8.6098967348456e-05,"reference_active":true,"delta_db":0.05044448293122698},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648548828731209e-06,"output_power":1.0025407948897041e-05,"reference_active":true,"delta_db":0.16640054709734725},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6517442110747655e-07,"output_power":3.405592844378627e-07,"reference_active":true,"delta_db":-0.30307624815790235}],"active_band_count":3,"active_band_mae_db":0.17330709272882552}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.827493119721784e-05,"log_rms_mae_db":2.885557715004378,"rms_envelope_nrmse":0.3320226459471637,"envelope_correlation":0.9214521141865616,"rms_gain_error_db":0.193437820574292,"reference_mean_power":0.00033891979539590525,"output_mean_power":0.00035435675630951724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5434076582307209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.258901619859076e-05,"output_power":6.566557251509011e-05,"reference_active":true,"delta_db":0.2083960990001752},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4223152137238314e-05,"output_power":1.619035050446611e-05,"reference_active":true,"delta_db":0.5626039541204363},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.429050160653033e-06,"output_power":1.3391058311474956e-05,"reference_active":true,"delta_db":2.01036262885886}],"active_band_count":3,"active_band_mae_db":0.9271208939931571}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.761964716205693e-05,"log_rms_mae_db":1.9657973489059537,"rms_envelope_nrmse":0.20089558328749765,"envelope_correlation":0.9688796135764014,"rms_gain_error_db":0.14211409626925517,"reference_mean_power":0.00033859490771126504,"output_mean_power":0.00034985801834088},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4129997856527199e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.165966182107606e-05,"output_power":6.63148576546762e-05,"reference_active":true,"delta_db":0.3160970328320264},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0839649935296936e-06,"output_power":9.375899803867167e-06,"reference_active":true,"delta_db":1.2173655119164843},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.531889318612957e-07,"output_power":1.0534267471481578e-06,"reference_active":true,"delta_db":2.0756552405824285}],"active_band_count":3,"active_band_mae_db":1.2030392617769798}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"hnm","status":"measured","mos":{"status":"measured","value":2.851712465286255,"predicted_mos":2.851712465286255,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"01eb47320bda9777935ec1e60538d2e8ed189a4ec0804b374ffcda3d3f3ea211","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.990984762998778},"started_at":"2026-10-13T04:39:16.346326+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3765000403509475e-05,"native_input_samples":96000,"native_input_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","output_sha256":"8d7f59a9b93d7986dfefe4b4d598b9a2f3e2549737d33703f702c7e22bc0a1a2","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0498074189999897,"synthesis_s":0.49924593400010053,"total_s":1.5490533530000903,"rtf":0.25817555883334836,"output_peak":0.35489879739787733,"output_rms":0.046999119674969225,"finished_at":"2026-10-13T04:39:17.900510+00:00","evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"8d7f59a9b93d7986dfefe4b4d598b9a2f3e2549737d33703f702c7e22bc0a1a2","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.498195616095586,"spectral_convergence":0.18919505876828632,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4772511327915145,"spectral_convergence":0.1780493045849487,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.879366897541151,"spectral_convergence":0.1874836011182091,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":5.958438108780488e-05,"delta_db":0.37955798880238323,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0011273036685778338,"delta_db":-0.11335313711079988,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0009195438413152188,"delta_db":-0.23277867519022188,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":4.771712206128387e-05,"delta_db":-0.09099924583571949,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":2.422370757331941e-05,"delta_db":-0.14918874484669414,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":3.0544529606618425e-05,"delta_db":-0.09610910422704858,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1769978160021445,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15070299181381916,"pitch":null,"output_rms_dbfs":-26.558205532299986,"output_peak":0.35489879739787733,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.618271215476084},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.498195616095586,"spectral_convergence":0.18919505876828632,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4772511327915145,"spectral_convergence":0.1780493045849487,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.879366897541151,"spectral_convergence":0.1874836011182091,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":5.958438108780488e-05,"delta_db":0.37955798880238323,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0011273036685778338,"delta_db":-0.11335313711079988,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0009195438413152188,"delta_db":-0.23277867519022188,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":4.771712206128387e-05,"delta_db":-0.09099924583571949,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":2.422370757331941e-05,"delta_db":-0.14918874484669414,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":3.0544529606618425e-05,"delta_db":-0.09610910422704858,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1769978160021445,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15070299181381916,"pitch":null,"output_rms_dbfs":-26.558205532299986,"output_peak":0.35489879739787733,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.618271215476084},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9004019686958665,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04782168483853718,"rms_dbfs":-26.407502540486163,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.755646813219986,"mean":-4.043451944986979e-05,"zero_energy":false},"output":{"rms":0.046999119674969225,"rms_dbfs":-26.558205532299986,"peak":0.35489879739787733,"peak_dbfs":-8.997909445809524,"crest_factor_db":17.56029608649046,"mean":1.3587228908709801e-05,"zero_energy":false},"rms_difference_db":-0.15070299181382296},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":141,"quiet_samples":22560,"reference_mean_power":6.018758005072885e-07,"output_mean_power":8.914754525636619e-07,"output_rms_dbfs":-60.49890610943951,"output_zero_energy":false,"output_peak":0.019547153687304995,"output_energy_fraction":9.484136688750938e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":265,"both_estimated_voiced_frames":261,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9422382671480144,"absolute_difference_cents_median":1.4485065203748926,"absolute_difference_cents_p95":10.053640819901023,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014140982422473965,"log_rms_mae_db":0.8394704624659322,"rms_envelope_nrmse":0.06191976334664259,"envelope_correlation":0.9964187600111151,"rms_gain_error_db":-0.15070292574674357,"reference_mean_power":0.0022888204000268467,"output_mean_power":0.002210759108728929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0460573675773498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006367944533168534,"output_power":0.0006079659027225613,"reference_active":true,"delta_db":-0.2012004900249595},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21105688014902e-05,"output_power":8.714776621087834e-05,"reference_active":true,"delta_db":-0.24053204504025305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2041191998800965e-05,"output_power":9.458549028366966e-06,"reference_active":true,"delta_db":-1.048449619962764}],"active_band_count":3,"active_band_mae_db":0.4967273850093255}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274322977924003,"log_rms_mae_db":0.4493675648628931,"rms_envelope_nrmse":0.03277600128367624,"envelope_correlation":0.999145368251263,"rms_gain_error_db":-0.15070310013509552,"reference_mean_power":0.0023003276499839287,"output_mean_power":0.00222187380932087},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.201135437818201e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005685810358795958,"output_power":0.0005459991254918362,"reference_active":true,"delta_db":-0.17600423280369826},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6960814797233804e-05,"output_power":4.474351308910128e-05,"reference_active":true,"delta_db":-0.21005543327801438},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.369888538054512e-07,"output_power":7.080460564187222e-07,"reference_active":true,"delta_db":-1.2167291640371882}],"active_band_count":3,"active_band_mae_db":0.534262943372967}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012682596688393775,"log_rms_mae_db":1.5780206554835934,"rms_envelope_nrmse":0.08175660180174386,"envelope_correlation":0.9945512805183558,"rms_gain_error_db":-0.20469861208803844,"reference_mean_power":0.0016135328798776547,"output_mean_power":0.001539245733516321},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.436393406849114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006358276871928121,"output_power":0.0006000373286377544,"reference_active":true,"delta_db":-0.25161166483588393},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.784078725837928e-05,"output_power":8.186198856968861e-05,"reference_active":true,"delta_db":-0.30613929211570284},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0842784949487312e-05,"output_power":8.125302991607897e-06,"reference_active":true,"delta_db":-1.2530127952994792}],"active_band_count":3,"active_band_mae_db":0.603587917417022}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011190972053745482,"log_rms_mae_db":0.8187943170164784,"rms_envelope_nrmse":0.04646406442171532,"envelope_correlation":0.9984378280003993,"rms_gain_error_db":-0.20469876545174254,"reference_mean_power":0.0016216465744125685,"output_mean_power":0.0015469858184601758},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.451026691855673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005862792692297138,"output_power":0.0005548736423120888,"reference_active":true,"delta_db":-0.23910442466404913},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3209611631795496e-05,"output_power":4.094899794469428e-05,"reference_active":true,"delta_db":-0.23337084200758001},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.999199919742322e-07,"output_power":7.090352019905293e-07,"reference_active":true,"delta_db":-1.4929745411091329}],"active_band_count":3,"active_band_mae_db":0.6551499359269207}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":1.4210880240430386,"rms_envelope_nrmse":0.13888425871981078,"envelope_correlation":0.9847908188457885,"rms_gain_error_db":-0.14798247702709216,"reference_mean_power":6.869686019984955e-05,"output_mean_power":6.639549916745022e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5736889079386925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.358608717262281e-05,"output_power":1.3156455727129325e-05,"reference_active":true,"delta_db":-0.1395548821460354},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4455610806681207e-06,"output_power":2.1629355723938753e-06,"reference_active":true,"delta_db":-0.5333493103817086},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901755177746091e-07,"output_power":4.968463862148821e-07,"reference_active":true,"delta_db":-0.7475905436810073}],"active_band_count":3,"active_band_mae_db":0.47349824540291713}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":0.7852006120250753,"rms_envelope_nrmse":0.07832275288256386,"envelope_correlation":0.9950365834987839,"rms_gain_error_db":-0.14799125230291044,"reference_mean_power":6.904136883729848e-05,"output_mean_power":6.67283318540266e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.335283050026158e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2478970323107016e-05,"output_power":1.2098875736348022e-05,"reference_active":true,"delta_db":-0.1343373564598932},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3575517045580523e-06,"output_power":1.1617984106779912e-06,"reference_active":true,"delta_db":-0.6762560135289785},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.985180184805259e-08,"output_power":3.3306469862626415e-08,"reference_active":true,"delta_db":-0.7791935778480851}],"active_band_count":3,"active_band_mae_db":0.5299289826123189}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":1.3634430996052052,"rms_envelope_nrmse":0.18816471469016297,"envelope_correlation":0.9706952698219121,"rms_gain_error_db":-0.1628533805527656,"reference_mean_power":1.4641647362050636e-05,"output_mean_power":1.4102775899221346e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3222330564047251e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0304744672556675e-06,"output_power":5.383669447052676e-06,"reference_active":true,"delta_db":-0.4927309631508869},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1757655511106124e-06,"output_power":2.096200096704373e-06,"reference_active":true,"delta_db":-0.16179359607357807},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.148040946601495e-06,"output_power":5.648851862999433e-07,"reference_active":true,"delta_db":-3.07997192270855}],"active_band_count":3,"active_band_mae_db":1.2448321606443382}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":0.8052020529203783,"rms_envelope_nrmse":0.1038376695565343,"envelope_correlation":0.9906578576610473,"rms_gain_error_db":-0.16282410009243534,"reference_mean_power":1.4714471928104231e-05,"output_mean_power":1.417301578380184e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1752743001544044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9652177739211e-06,"output_power":5.17773980841083e-06,"reference_active":true,"delta_db":-0.6148608071238724},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4021233434065379e-06,"output_power":1.1881853200728351e-06,"reference_active":true,"delta_db":-0.7190203736509827},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593565327256369e-08,"output_power":3.9877128363618355e-08,"reference_active":true,"delta_db":-3.8125616092122665}],"active_band_count":3,"active_band_mae_db":1.7154809299957072}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.6437735557556152,"predicted_mos":2.6437735557556152,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4c36ff0c72a10ac7c784fd68af43a23b32ce66b376bd57d77728620cb77a481e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4226014460000442},"started_at":"2026-10-13T04:42:12.899302+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2497999452752993e-05,"native_input_samples":96000,"native_input_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","output_sha256":"2c03093e9bfff3d25fb820689ba6adc2b233076b8f45757b25fb64e33f6f00e6","output_samples":96000,"length_delta_samples":0,"analysis_s":1.313951413999348,"synthesis_s":0.12027267700068478,"total_s":1.4342240910000328,"rtf":0.23903734850000546,"output_peak":0.3340404822925707,"output_rms":0.055409584603573724,"finished_at":"2026-10-13T04:42:14.339033+00:00","evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"2c03093e9bfff3d25fb820689ba6adc2b233076b8f45757b25fb64e33f6f00e6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.954102907987353,"spectral_convergence":0.49841228590834424,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.537759473035772,"spectral_convergence":0.27420891350498583,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.032678856305323,"spectral_convergence":0.3191977572061302,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":7.518325888692366e-05,"delta_db":1.3894451293609944,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0015077781171263906,"delta_db":1.1496120192818178,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0013610165539918299,"delta_db":1.470130999921795,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":6.296459264140076e-05,"delta_db":1.113222309858418,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":3.013217357645798e-05,"delta_db":0.7987097095447449,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":3.314736971759203e-05,"delta_db":0.2590472063607616,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0300278957214222,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.2792004272605046,"pitch":null,"output_rms_dbfs":-25.128302113225658,"output_peak":0.3340404822925707,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.84151374577615},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.954102907987353,"spectral_convergence":0.49841228590834424,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.537759473035772,"spectral_convergence":0.27420891350498583,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.032678856305323,"spectral_convergence":0.3191977572061302,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":7.518325888692366e-05,"delta_db":1.3894451293609944,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0015077781171263906,"delta_db":1.1496120192818178,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0013610165539918299,"delta_db":1.470130999921795,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":6.296459264140076e-05,"delta_db":1.113222309858418,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":3.013217357645798e-05,"delta_db":0.7987097095447449,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":3.314736971759203e-05,"delta_db":0.2590472063607616,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0300278957214222,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.2792004272605046,"pitch":null,"output_rms_dbfs":-25.128302113225658,"output_peak":0.3340404822925707,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.84151374577615},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9212752626392683,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04782168483853718,"rms_dbfs":-26.407502540486163,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.755646813219986,"mean":-4.043451944986979e-05,"zero_energy":false},"output":{"rms":0.055409584603573724,"rms_dbfs":-25.128302113225658,"peak":0.3340404822925707,"peak_dbfs":-9.524017958924272,"crest_factor_db":15.60428415430139,"mean":-1.051109159231021e-09,"zero_energy":false},"rms_difference_db":1.2792004272605055},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":141,"quiet_samples":22560,"reference_mean_power":6.018758005072885e-07,"output_mean_power":6.97143030220491e-07,"output_rms_dbfs":-61.56678110190529,"output_zero_energy":false,"output_peak":0.007622542018849432,"output_energy_fraction":5.3360508974006344e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":292,"both_estimated_voiced_frames":268,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9675090252707581,"absolute_difference_cents_median":5.473460335587605,"absolute_difference_cents_p95":23.7985785773026,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014140982422473965,"log_rms_mae_db":1.434298575650163,"rms_envelope_nrmse":0.1993595598460685,"envelope_correlation":0.9877044970056235,"rms_gain_error_db":1.2792010269886593,"reference_mean_power":0.0022888204000268467,"output_mean_power":0.003072782482669747},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0460573675773498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006367944533168534,"output_power":0.0009386093041368717,"reference_active":true,"delta_db":1.6848558301912493},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21105688014902e-05,"output_power":0.00014927890017648368,"reference_active":true,"delta_db":2.0968896254803306},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2041191998800965e-05,"output_power":1.2539489921338475e-05,"reference_active":true,"delta_db":0.17610389370059326}],"active_band_count":3,"active_band_mae_db":1.319283116457391}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274322977924003,"log_rms_mae_db":1.0830428309406404,"rms_envelope_nrmse":0.18122179456465295,"envelope_correlation":0.993075411709034,"rms_gain_error_db":1.2792033607627835,"reference_mean_power":0.0023003276499839287,"output_mean_power":0.0030882328318765184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.201135437818201e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005685810358795958,"output_power":0.0008490778463897971,"reference_active":true,"delta_db":1.7415513929657527},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6960814797233804e-05,"output_power":7.820809236278586e-05,"reference_active":true,"delta_db":2.215160692223307},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.369888538054512e-07,"output_power":1.0460894496285768e-06,"reference_active":true,"delta_db":0.47834397370097315}],"active_band_count":3,"active_band_mae_db":1.4783520196300108}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012682596688393775,"log_rms_mae_db":2.0212617313441754,"rms_envelope_nrmse":0.23980363343156624,"envelope_correlation":0.9872316093366268,"rms_gain_error_db":1.5778721393852737,"reference_mean_power":0.0016135328798776547,"output_mean_power":0.002320411630701603},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.436393406849114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006358276871928121,"output_power":0.0010081345403967505,"reference_active":true,"delta_db":2.001790592661638},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.784078725837928e-05,"output_power":0.00015417217431393126,"reference_active":true,"delta_db":2.443097779548358},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0842784949487312e-05,"output_power":1.269391721000587e-05,"reference_active":true,"delta_db":0.6845481723568919}],"active_band_count":3,"active_band_mae_db":1.7098121815222962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011190972053745482,"log_rms_mae_db":1.5909548425870152,"rms_envelope_nrmse":0.22441595282101065,"envelope_correlation":0.9916913770518366,"rms_gain_error_db":1.5778731756849276,"reference_mean_power":0.0016216465744125685,"output_mean_power":0.0023320804409811624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.451026691855673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005862792692297138,"output_power":0.0009347992842317273,"reference_active":true,"delta_db":2.0261383333321166},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3209611631795496e-05,"output_power":7.686323522848721e-05,"reference_active":true,"delta_db":2.501382973902722},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.999199919742322e-07,"output_power":1.1564711422455055e-06,"reference_active":true,"delta_db":0.6316954860187649}],"active_band_count":3,"active_band_mae_db":1.7197389310845346}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":1.4211498300785084,"rms_envelope_nrmse":0.17155405810120358,"envelope_correlation":0.9915800436360516,"rms_gain_error_db":1.0710156086786393,"reference_mean_power":6.869686019984955e-05,"output_mean_power":8.791003417524588e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5736889079386925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.358608717262281e-05,"output_power":1.756008191911175e-05,"reference_active":true,"delta_db":1.114321407573522},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4455610806681207e-06,"output_power":3.3640024397847633e-06,"reference_active":true,"delta_db":1.3847778781232722},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901755177746091e-07,"output_power":7.084111002692006e-07,"reference_active":true,"delta_db":0.7930416792082517}],"active_band_count":3,"active_band_mae_db":1.0973803216350155}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":1.0192547977246105,"rms_envelope_nrmse":0.1470102200594789,"envelope_correlation":0.9968143434080305,"rms_gain_error_db":1.0710381453017779,"reference_mean_power":6.904136883729848e-05,"output_mean_power":8.835135364553448e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.335283050026158e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2478970323107016e-05,"output_power":1.6224117016201537e-05,"reference_active":true,"delta_db":1.139823181094284},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3575517045580523e-06,"output_power":1.8449581324177957e-06,"reference_active":true,"delta_db":1.3323013578915566},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.985180184805259e-08,"output_power":3.9599919888904666e-08,"reference_active":true,"delta_db":-0.027536548997906342}],"active_band_count":3,"active_band_mae_db":0.833220362661249}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":1.2818527093692078,"rms_envelope_nrmse":0.1832973722503598,"envelope_correlation":0.9736634844782657,"rms_gain_error_db":0.2388672747053006,"reference_mean_power":1.4641647362050636e-05,"output_mean_power":1.54695140166871e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3222330564047251e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0304744672556675e-06,"output_power":6.02252253791328e-06,"reference_active":true,"delta_db":-0.005730490986538311},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1757655511106124e-06,"output_power":2.182583949438802e-06,"reference_active":true,"delta_db":0.013588608186697724},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.148040946601495e-06,"output_power":1.024904030210881e-06,"reference_active":true,"delta_db":-0.4927417719914126}],"active_band_count":3,"active_band_mae_db":0.17068695705488288}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":0.8386139785026673,"rms_envelope_nrmse":0.1062934123274103,"envelope_correlation":0.9906944522047102,"rms_gain_error_db":0.23888651755825638,"reference_mean_power":1.4714471928104231e-05,"output_mean_power":1.554652510640499e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1752743001544044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9652177739211e-06,"output_power":5.823252917685041e-06,"reference_active":true,"delta_db":-0.10460650252339401},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4021233434065379e-06,"output_power":1.3227003963686785e-06,"reference_active":true,"delta_db":-0.253247360786335},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593565327256369e-08,"output_power":7.486210322572593e-08,"reference_active":true,"delta_db":-1.0771801291451266}],"active_band_count":3,"active_band_mae_db":0.4783446641516185}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"facodec","status":"measured","mos":{"status":"measured","value":3.7770237922668457,"predicted_mos":3.7770237922668457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c1f643b934074bf4c2f014d690002afc3c3bb1dc57035efc54eb15acd80b15af","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.335625285999413},"started_at":"2026-10-13T04:52:30.970929+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.637999932630919e-05,"native_input_samples":96000,"native_input_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","output_sha256":"b8f006e8c09115ca1df90e9a75561da59dc7326d19b5fe5c71559f57d70f7738","output_samples":96000,"length_delta_samples":0,"analysis_s":2.0794004189992847,"synthesis_s":7.426661600000443,"total_s":9.506062018999728,"rtf":1.584343669833288,"output_peak":0.3345392346382141,"output_rms":0.0453539415414867,"finished_at":"2026-10-13T04:52:40.486092+00:00","evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"b8f006e8c09115ca1df90e9a75561da59dc7326d19b5fe5c71559f57d70f7738","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.290043989925611,"spectral_convergence":0.23276264092037488,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.833312974044517,"spectral_convergence":0.20754691741487818,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.061987262299003,"spectral_convergence":0.2415539979126689,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":5.3228378056403504e-05,"delta_db":-0.11033400929836856,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0011427902489446913,"delta_db":-0.05409708003093219,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0007849796448358173,"delta_db":-0.9199191227014547,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":2.9696139400298997e-05,"delta_db":-2.1507417376901814,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":2.3629843655589765e-05,"delta_db":-0.2569864139021073,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":2.2655758455792023e-05,"delta_db":-1.3936574589196637,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.814289303757118,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4601967405957042,"pitch":null,"output_rms_dbfs":-26.867699281081872,"output_peak":0.3345392346382141,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.728448075423043},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.290043989925611,"spectral_convergence":0.23276264092037488,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.833312974044517,"spectral_convergence":0.20754691741487818,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.061987262299003,"spectral_convergence":0.2415539979126689,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":5.3228378056403504e-05,"delta_db":-0.11033400929836856,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0011427902489446913,"delta_db":-0.05409708003093219,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0007849796448358173,"delta_db":-0.9199191227014547,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":2.9696139400298997e-05,"delta_db":-2.1507417376901814,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":2.3629843655589765e-05,"delta_db":-0.2569864139021073,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":2.2655758455792023e-05,"delta_db":-1.3936574589196637,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.814289303757118,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4601967405957042,"pitch":null,"output_rms_dbfs":-26.867699281081872,"output_peak":0.3345392346382141,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.728448075423043},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9125129745438256,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04782168483853718,"rms_dbfs":-26.407502540486163,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.755646813219986,"mean":-4.043451944986979e-05,"zero_energy":false},"output":{"rms":0.0453539415414867,"rms_dbfs":-26.86769928108187,"peak":0.3345392346382141,"peak_dbfs":-9.511058820246014,"crest_factor_db":17.356640460835855,"mean":8.094544813317838e-06,"zero_energy":false},"rms_difference_db":-0.4601967405957055},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":141,"quiet_samples":22560,"reference_mean_power":6.018758005072885e-07,"output_mean_power":5.666179634092119e-07,"output_rms_dbfs":-62.46709661220749,"output_zero_energy":false,"output_peak":0.004921556916087866,"output_energy_fraction":6.473335692960824e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":286,"both_estimated_voiced_frames":269,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9711191335740073,"absolute_difference_cents_median":4.23266911192594,"absolute_difference_cents_p95":15.457289551010577,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014140982422473965,"log_rms_mae_db":1.170051476145414,"rms_envelope_nrmse":0.10761330087583466,"envelope_correlation":0.9906614141973394,"rms_gain_error_db":-0.46019660847386856,"reference_mean_power":0.0022888204000268467,"output_mean_power":0.0020586952135451643},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0460573675773498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006367944533168534,"output_power":0.000607758058183198,"reference_active":true,"delta_db":-0.2026854609611484},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21105688014902e-05,"output_power":9.096115614524628e-05,"reference_active":true,"delta_db":-0.05453492433088988},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2041191998800965e-05,"output_power":1.091887805127606e-05,"reference_active":true,"delta_db":-0.42491465796117833}],"active_band_count":3,"active_band_mae_db":0.2273783477510722}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274322977924003,"log_rms_mae_db":0.9531398413377579,"rms_envelope_nrmse":0.09620854145629765,"envelope_correlation":0.9924852352982502,"rms_gain_error_db":-0.4601949684430143,"reference_mean_power":0.0023003276499839287,"output_mean_power":0.0020690462698120444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.201135437818201e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005685810358795958,"output_power":0.0005428201673649152,"reference_active":true,"delta_db":-0.20136395786897826},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6960814797233804e-05,"output_power":4.5833788232008764e-05,"reference_active":true,"delta_db":-0.10549869739892723},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.369888538054512e-07,"output_power":9.666300993368335e-07,"reference_active":true,"delta_db":0.13525889603088098}],"active_band_count":3,"active_band_mae_db":0.1473738504329288}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012682596688393775,"log_rms_mae_db":1.512381725745178,"rms_envelope_nrmse":0.10584568075596067,"envelope_correlation":0.9921583086775722,"rms_gain_error_db":-0.4565637290345699,"reference_mean_power":0.0016135328798776547,"output_mean_power":0.00145251774514829},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.436393406849114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006358276871928121,"output_power":0.0005864867455603567,"reference_active":true,"delta_db":-0.350812337695193},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.784078725837928e-05,"output_power":8.389547665008573e-05,"reference_active":true,"delta_db":-0.19957673590841749},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0842784949487312e-05,"output_power":9.57087037464859e-06,"reference_active":true,"delta_db":-0.5418940997457117}],"active_band_count":3,"active_band_mae_db":0.3640943911164407}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011190972053745482,"log_rms_mae_db":1.107671130166547,"rms_envelope_nrmse":0.09277372131093135,"envelope_correlation":0.9940471541607163,"rms_gain_error_db":-0.4565627828279949,"reference_mean_power":0.0016216465744125685,"output_mean_power":0.0014598220886971795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.451026691855673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005862792692297138,"output_power":0.0005362325169982682,"reference_active":true,"delta_db":-0.387513919032096},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3209611631795496e-05,"output_power":4.088073349531273e-05,"reference_active":true,"delta_db":-0.24061683439500858},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.999199919742322e-07,"output_power":9.277993322731133e-07,"reference_active":true,"delta_db":-0.3251119593987316}],"active_band_count":3,"active_band_mae_db":0.3177475709419454}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":1.7036637099369765,"rms_envelope_nrmse":0.22157185938984583,"envelope_correlation":0.9766631673345141,"rms_gain_error_db":-1.4649394039873143,"reference_mean_power":6.869686019984955e-05,"output_mean_power":4.9027861166134825e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5736889079386925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.358608717262281e-05,"output_power":1.047422483507928e-05,"reference_active":true,"delta_db":-1.1297250479524406},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4455610806681207e-06,"output_power":2.192760129635321e-06,"reference_active":true,"delta_db":-0.4738738834708723},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901755177746091e-07,"output_power":5.946995835169631e-07,"reference_active":true,"delta_db":0.03316444633603485}],"active_band_count":3,"active_band_mae_db":0.5455877925864492}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":1.4309519846315906,"rms_envelope_nrmse":0.20698885817451196,"envelope_correlation":0.9813739704867759,"rms_gain_error_db":-1.4649341101828126,"reference_mean_power":6.904136883729848e-05,"output_mean_power":4.927379158660012e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.335283050026158e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2478970323107016e-05,"output_power":9.674630213503774e-06,"reference_active":true,"delta_db":-1.1054437761788058},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3575517045580523e-06,"output_power":1.2323326738859131e-06,"reference_active":true,"delta_db":-0.4202841593217891},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.985180184805259e-08,"output_power":3.965402866797938e-08,"reference_active":true,"delta_db":-0.02160646014967431}],"active_band_count":3,"active_band_mae_db":0.5157781318834231}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":1.8538798365068945,"rms_envelope_nrmse":0.28799266790642386,"envelope_correlation":0.9356149788055245,"rms_gain_error_db":-1.124740885973457,"reference_mean_power":1.4641647362050636e-05,"output_mean_power":1.1300973440450553e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3222330564047251e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0304744672556675e-06,"output_power":3.4659646080394293e-06,"reference_active":true,"delta_db":-2.405273593250644},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1757655511106124e-06,"output_power":2.3303203148581186e-06,"reference_active":true,"delta_db":0.2980352491316104},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.148040946601495e-06,"output_power":9.774759362013811e-07,"reference_active":true,"delta_db":-0.6985130349555396}],"active_band_count":3,"active_band_mae_db":1.1339406257792646}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":1.5449458928166235,"rms_envelope_nrmse":0.24574625694599805,"envelope_correlation":0.9520938793191916,"rms_gain_error_db":-1.1247273298089162,"reference_mean_power":1.4714471928104231e-05,"output_mean_power":1.1357217626766754e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1752743001544044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9652177739211e-06,"output_power":3.4024732570115543e-06,"reference_active":true,"delta_db":-2.438315828175779},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4021233434065379e-06,"output_power":1.4482417561857355e-06,"reference_active":true,"delta_db":0.14054845279294786},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593565327256369e-08,"output_power":7.151562748889232e-08,"reference_active":true,"delta_db":-1.2757908366461843}],"active_band_count":3,"active_band_mae_db":1.2848850392049702}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.6391491889953613,"predicted_mos":3.6391491889953613,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"3e0842c5e0a38da92712471fc3c5f6fd52efba79098256a4cde7e827afa55737","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.763858143000107},"started_at":"2026-10-13T07:14:23.104125+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007483306000722223,"native_input_samples":264600,"native_input_sha256":"37ccd9e4e41b30d429da29fcd94b7444eda85d3466a3ca2235faa716343ed0ce","output_sha256":"6029925e29d86a622c4ee30e3c24c25f75709a0b2de387cee5311a4fe673efda","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15941574899989064,"synthesis_s":5.2425160620005045,"total_s":5.401931811000395,"rtf":0.9003219685000659,"output_peak":0.2796066701412201,"output_rms":0.04403732751775321,"finished_at":"2026-10-13T07:14:28.527826+00:00","evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"6029925e29d86a622c4ee30e3c24c25f75709a0b2de387cee5311a4fe673efda","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.120826378735963,"output_peak":0.27772390887732173},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.956664721505437,"spectral_convergence":0.4210893968895289,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.063548918609067,"spectral_convergence":0.3066171004265605,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.079785620083259,"spectral_convergence":0.36478106081580053,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":3.062357497839902e-05,"delta_db":-2.5112074513213907,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0009836199661358567,"delta_db":-0.7054889829856724,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.000812670004664351,"delta_db":-0.7693607714306283,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":6.250294230806238e-05,"delta_db":1.0812629512563567,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":3.546608623388061e-05,"delta_db":1.5065377612980193,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":1.6402075646465207e-05,"delta_db":-2.796455411653529,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5617188883242659,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7116052542845905,"pitch":null,"output_rms_dbfs":-27.119107794770756,"output_peak":0.27772390887732173,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.033333086732588},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7737604266436907,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04782168483853718,"rms_dbfs":-26.407502540486163,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.755646813219986,"mean":-4.043451944986979e-05,"zero_energy":false},"output":{"rms":0.04406001191519374,"rms_dbfs":-27.119107794770756,"peak":0.27772390887732173,"peak_dbfs":-11.127734616264132,"crest_factor_db":15.991373178506624,"mean":-3.957039746860025e-05,"zero_energy":false},"rms_difference_db":-0.7116052542845921},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":141,"quiet_samples":22560,"reference_mean_power":6.018758005072885e-07,"output_mean_power":4.632765697140146e-07,"output_rms_dbfs":-63.34159663753732,"output_zero_energy":false,"output_peak":0.003596760649319446,"output_energy_fraction":5.6081416954820755e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":265,"both_estimated_voiced_frames":248,"reference_voiced_output_unvoiced_frames":29,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.8953068592057761,"absolute_difference_cents_median":6.103977162961186,"absolute_difference_cents_p95":43.36221553987393,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014140982422473965,"log_rms_mae_db":1.4571255895836859,"rms_envelope_nrmse":0.115558566559371,"envelope_correlation":0.992461071106535,"rms_gain_error_db":-0.7116048751653768,"reference_mean_power":0.0022888204000268467,"output_mean_power":0.0019429034887920498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0460573675773498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006367944533168534,"output_power":0.0005710242573143682,"reference_active":true,"delta_db":-0.4734471427999106},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21105688014902e-05,"output_power":8.670622814897411e-05,"reference_active":true,"delta_db":-0.2625916999625936},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2041191998800965e-05,"output_power":1.1780329177033886e-05,"reference_active":true,"delta_db":-0.09512055261465349}],"active_band_count":3,"active_band_mae_db":0.2770531317923859}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274322977924003,"log_rms_mae_db":1.1171068646689097,"rms_envelope_nrmse":0.09445106070012542,"envelope_correlation":0.9969335067182333,"rms_gain_error_db":-0.7116030414004616,"reference_mean_power":0.0023003276499839287,"output_mean_power":0.001952672434931925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.201135437818201e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005685810358795958,"output_power":0.0005137309908033514,"reference_active":true,"delta_db":-0.44056605077586963},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6960814797233804e-05,"output_power":4.42006246692386e-05,"reference_active":true,"delta_db":-0.26307216467133643},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.369888538054512e-07,"output_power":9.654259358388149e-07,"reference_active":true,"delta_db":0.12984537157101186}],"active_band_count":3,"active_band_mae_db":0.277827862339406}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012682596688393775,"log_rms_mae_db":2.2807513807535775,"rms_envelope_nrmse":0.1450402412816927,"envelope_correlation":0.9872614347239632,"rms_gain_error_db":-0.7929859808722973,"reference_mean_power":0.0016135328798776547,"output_mean_power":0.0013442474509386614},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.436393406849114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006358276871928121,"output_power":0.0005470173048828356,"reference_active":true,"delta_db":-0.6533836994669948},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.784078725837928e-05,"output_power":7.722021524374684e-05,"reference_active":true,"delta_db":-0.5596521157236566},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0842784949487312e-05,"output_power":9.587924183417735e-06,"reference_active":true,"delta_db":-0.5341625308544372}],"active_band_count":3,"active_band_mae_db":0.5823994486816962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011190972053745482,"log_rms_mae_db":1.5603754669152279,"rms_envelope_nrmse":0.12532159679929705,"envelope_correlation":0.9916302714675594,"rms_gain_error_db":-0.7929859973465926,"reference_mean_power":0.0016216465744125685,"output_mean_power":0.0013510070311512352},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.451026691855673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005862792692297138,"output_power":0.000509815235798055,"reference_active":true,"delta_db":-0.6069172765932549},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3209611631795496e-05,"output_power":3.7575104468842985e-05,"reference_active":true,"delta_db":-0.6068016617233436},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.999199919742322e-07,"output_power":9.170261237817392e-07,"reference_active":true,"delta_db":-0.3758354375548698}],"active_band_count":3,"active_band_mae_db":0.5298514586238228}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":2.9073421481903474,"rms_envelope_nrmse":0.3179781518968566,"envelope_correlation":0.9483351139997637,"rms_gain_error_db":1.0906593241509885,"reference_mean_power":6.869686019984955e-05,"output_mean_power":8.830856354150508e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5736889079386925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.358608717262281e-05,"output_power":1.7630110543027008e-05,"reference_active":true,"delta_db":1.1316063855519092},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4455610806681207e-06,"output_power":3.4718314755831533e-06,"reference_active":true,"delta_db":1.521801218559842},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901755177746091e-07,"output_power":8.657320711424371e-07,"reference_active":true,"delta_db":1.6640231658501115}],"active_band_count":3,"active_band_mae_db":1.4391435899872878}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":2.3378078737818764,"rms_envelope_nrmse":0.27952578295370856,"envelope_correlation":0.9610467871307913,"rms_gain_error_db":1.0906823848001852,"reference_mean_power":6.904136883729848e-05,"output_mean_power":8.875189438845944e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.335283050026158e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2478970323107016e-05,"output_power":1.626332027424973e-05,"reference_active":true,"delta_db":1.150304627663866},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3575517045580523e-06,"output_power":1.9186993295856906e-06,"reference_active":true,"delta_db":1.5025054442111734},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.985180184805259e-08,"output_power":5.48506908405721e-08,"reference_active":true,"delta_db":1.3873413961002465}],"active_band_count":3,"active_band_mae_db":1.346717155991762}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":3.7466623707921864,"rms_envelope_nrmse":0.30632698127349617,"envelope_correlation":0.9272557292276123,"rms_gain_error_db":-0.07205282503064647,"reference_mean_power":1.4641647362050636e-05,"output_mean_power":1.4400735056228536e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3222330564047251e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0304744672556675e-06,"output_power":5.69512632159692e-06,"reference_active":true,"delta_db":-0.24848121564513284},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1757655511106124e-06,"output_power":2.3790871241314465e-06,"reference_active":true,"delta_db":0.38798250259778017},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.148040946601495e-06,"output_power":8.454008325185594e-07,"reference_active":true,"delta_db":-1.3289470693209735}],"active_band_count":3,"active_band_mae_db":0.6551369291879622}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":3.1547149213459,"rms_envelope_nrmse":0.2618204692870824,"envelope_correlation":0.9409938161038144,"rms_gain_error_db":-0.07187718376120183,"reference_mean_power":1.4714471928104231e-05,"output_mean_power":1.4472946689690429e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1752743001544044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9652177739211e-06,"output_power":5.171960724781062e-06,"reference_active":true,"delta_db":-0.6197108496611661},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4021233434065379e-06,"output_power":1.4184370191188707e-06,"reference_active":true,"delta_db":0.050238374196130635},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593565327256369e-08,"output_power":6.283220482123857e-08,"reference_active":true,"delta_db":-1.8379773757492892}],"active_band_count":3,"active_band_mae_db":0.8359755332021953}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"golf","status":"measured","mos":{"status":"measured","value":2.9760892391204834,"predicted_mos":2.9760892391204834,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"f6b00a1be6a9f05830642eaf9a2f47b7b318d706ce47e3f07a970130800a8b86","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1928705680002167},"started_at":"2026-10-13T05:02:34.351421+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004755936000037764,"native_input_samples":144000,"native_input_sha256":"2826551247918c5247905fb74a6a5e7fff2475ed17d1902981e13242417eef0c","output_sha256":"632b63f88a5f1fbd8a53d88750e411d77dc0a44ac5df30be301661798eb98184","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5306365920005192,"synthesis_s":0.1151731509999081,"total_s":0.6458097430004273,"rtf":0.10763495716673788,"output_peak":0.27438437938690186,"output_rms":0.04203693359702535,"finished_at":"2026-10-13T05:02:35.012093+00:00","analysis_stages":{"external_f0_s":0.08398161900004197},"evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"632b63f88a5f1fbd8a53d88750e411d77dc0a44ac5df30be301661798eb98184","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.522906537371632,"output_peak":0.2738461653940152},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.805162938364207,"spectral_convergence":0.43915706596649895,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.508097667031034,"spectral_convergence":0.25084010245615385,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.696561348772748,"spectral_convergence":0.2892969597692679,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.468881890087343e-05,"output_power":4.8563634544997285e-05,"delta_db":-0.5158736187603716,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001159043747884044,"output_power":0.0009115899910755441,"delta_db":-1.043002804948627,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009717972857294367,"output_power":0.0007310875446247424,"delta_db":-1.236062967366887,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.880876216664262e-05,"output_power":4.5061791741982896e-05,"delta_db":-0.3468933760386555,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5111983515378267e-05,"output_power":1.702287780179305e-05,"delta_db":-1.6884803601376275,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1279973357889796e-05,"output_power":1.559886065342425e-05,"delta_db":-3.021734960034699,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3086746812144778,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1226466559429162,"pitch":null,"output_rms_dbfs":-27.522906537371632,"output_peak":0.2738461653940152,"reference_rms_dbfs":-26.400259881428717,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.66994065138933},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8923840874581733,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.047861577194595925,"rms_dbfs":-26.400259881428717,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.74840415416254,"mean":-4.047814116056057e-05,"zero_energy":false},"output":{"rms":0.04205858652454317,"rms_dbfs":-27.522906537371632,"peak":0.2738461653940152,"peak_dbfs":-11.249866720402599,"crest_factor_db":16.273039816969032,"mean":0.00025310182299897874,"zero_energy":false},"rms_difference_db":-1.1226466559429156},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":140,"quiet_samples":22400,"reference_mean_power":6.025169360717493e-07,"output_mean_power":5.746259280304935e-07,"output_rms_dbfs":-62.40614781848209,"output_zero_energy":false,"output_peak":0.00579748415129848,"output_energy_fraction":7.592365117446999e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":290,"both_estimated_voiced_frames":263,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9494584837545126,"absolute_difference_cents_median":5.3660687372083,"absolute_difference_cents_p95":21.768651508672765,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014140982422473965,"log_rms_mae_db":2.2349682992972326,"rms_envelope_nrmse":0.17385754338778103,"envelope_correlation":0.9838168931601358,"rms_gain_error_db":-1.1226468997144121,"reference_mean_power":0.0022926437657512415,"output_mean_power":0.0017704019884640099},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.046066534612196e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006373063337700795,"output_power":0.0005005819815158635,"reference_active":true,"delta_db":-1.0487302157692922},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21246005985792e-05,"output_power":7.038824304950217e-05,"reference_active":true,"delta_db":-1.1687549303164988},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2028495225187204e-05,"output_power":8.136981719750514e-06,"reference_active":true,"delta_db":-1.6974795980865238}],"active_band_count":3,"active_band_mae_db":1.3049882480574382}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001274322977924003,"log_rms_mae_db":1.942489116167145,"rms_envelope_nrmse":0.15628762718141542,"envelope_correlation":0.9894618098687056,"rms_gain_error_db":-1.1226507367512233,"reference_mean_power":0.0023041896776318776,"output_mean_power":0.001779316282746458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.199262237623289e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005691383048982246,"output_power":0.000453133107906667,"reference_active":true,"delta_db":-0.9899202124117213},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.681314954422358e-05,"output_power":3.5636288008559116e-05,"reference_active":true,"delta_db":-1.184754008328279},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.344479460569256e-07,"output_power":7.10658853205843e-07,"reference_active":true,"delta_db":-1.1889394303691014}],"active_band_count":3,"active_band_mae_db":1.1212045503697006}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00012682596688393775,"log_rms_mae_db":2.4313297240196086,"rms_envelope_nrmse":0.18234990924703537,"envelope_correlation":0.9807813975547853,"rms_gain_error_db":-1.0366934358832334,"reference_mean_power":0.0016162284432572785,"output_mean_power":0.0012730146507709211},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.44134122144863e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006364860290752885,"output_power":0.0004887053087252048,"reference_active":true,"delta_db":-1.147418185687942},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.78391887160766e-05,"output_power":6.537747791612925e-05,"reference_active":true,"delta_db":-1.282601532979148},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.078626102786349e-05,"output_power":7.3253715122321685e-06,"reference_active":true,"delta_db":-1.6804127083373306}],"active_band_count":3,"active_band_mae_db":1.3701441423348069}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00011190972053745482,"log_rms_mae_db":1.9079053052110844,"rms_envelope_nrmse":0.16471454747407832,"envelope_correlation":0.9855745350047335,"rms_gain_error_db":-1.0366928925143593,"reference_mean_power":0.001624369416522723,"output_mean_power":0.001279427009742138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.454612123344622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005869177434946883,"output_power":0.0004553224621113855,"reference_active":true,"delta_db":-1.1025816356384714},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3068612966271865e-05,"output_power":3.242814147820407e-05,"reference_active":true,"delta_db":-1.232388263439663},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.978731390155141e-07,"output_power":6.42600816508357e-07,"reference_active":true,"delta_db":-1.9113405929817162}],"active_band_count":3,"active_band_mae_db":1.415436830686617}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":2.210772426486852,"rms_envelope_nrmse":0.18875975005411985,"envelope_correlation":0.9765938769985296,"rms_gain_error_db":-0.8114112436392076,"reference_mean_power":6.88113795209706e-05,"output_mean_power":5.708462339773413e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.5764648783079027e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3596149415202431e-05,"output_power":1.1059836460333429e-05,"reference_active":true,"delta_db":-0.8966722343913007},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.443677246884127e-06,"output_power":1.7869899922238455e-06,"reference_active":true,"delta_db":-1.359217248150473},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.90985146421198e-07,"output_power":4.888062972433428e-07,"reference_active":true,"delta_db":-0.8243977339511959}],"active_band_count":3,"active_band_mae_db":1.0267624054976565}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":1.762743282330872,"rms_envelope_nrmse":0.1410393990082376,"envelope_correlation":0.9893313357097324,"rms_gain_error_db":-0.8114067934820632,"reference_mean_power":6.915707383706064e-05,"output_mean_power":5.737146367397084e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.3377570773366808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2488921137376412e-05,"output_power":1.0071965999954011e-05,"reference_active":true,"delta_db":-0.9341067216666609},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3514320119959827e-06,"output_power":9.626765965511221e-07,"reference_active":true,"delta_db":-1.4731378815293898},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.979574360422234e-08,"output_power":3.4610768197479674e-08,"reference_active":true,"delta_db":-0.6062538542152911}],"active_band_count":3,"active_band_mae_db":1.0044994858037806}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":3.3749716550122635,"rms_envelope_nrmse":0.2970858952718973,"envelope_correlation":0.9388675646539404,"rms_gain_error_db":-0.49571238816950014,"reference_mean_power":1.4665961682587433e-05,"output_mean_power":1.3083962995638672e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3239686281549858e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.036163537227795e-06,"output_power":8.677775285086155e-06,"reference_active":true,"delta_db":1.5764740212316275},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.175009648030237e-06,"output_power":2.643758148634165e-06,"reference_active":true,"delta_db":0.847605355478233},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1473348944879346e-06,"output_power":8.934742590924419e-07,"reference_active":true,"delta_db":-1.0860815716810372}],"active_band_count":3,"active_band_mae_db":1.1700536494636327}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":3.1333060649052196,"rms_envelope_nrmse":0.26256337419455555,"envelope_correlation":0.9514853030851964,"rms_gain_error_db":-0.49556547540785045,"reference_mean_power":1.4739059745232852e-05,"output_mean_power":1.3149620879685334e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.176630333435014e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.966103218609535e-06,"output_power":8.753455111687129e-06,"reference_active":true,"delta_db":1.6648874624357946},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3989173651596035e-06,"output_power":1.4411321868874863e-06,"reference_active":true,"delta_db":0.1291175678369117},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.57132987900038e-08,"output_power":6.256792266587959e-08,"reference_active":true,"delta_db":-1.8462054849715568}],"active_band_count":3,"active_band_mae_db":1.213403505081421}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.6026387214660645,"predicted_mos":2.6026387214660645,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"669992fa90a5f3ab85ab3cadedfc1cfc0600ebda1c6d90299f3ffe7f02f5704a","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.26918535099685},"started_at":"2026-10-13T07:37:44.134734+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027232120010012295,"native_input_samples":144000,"native_input_sha256":"2826551247918c5247905fb74a6a5e7fff2475ed17d1902981e13242417eef0c","output_sha256":"19e6aede37a5e7899698c39ba7821a11483789942e82839f9faaa17a6b5fee80","output_samples":144120,"length_delta_samples":120,"analysis_s":2.2900678439982585,"synthesis_s":4.7688409779984795,"total_s":7.058908821996738,"rtf":1.176484803666123,"output_peak":0.2324976921081543,"output_rms":0.046193772028345785,"finished_at":"2026-10-13T07:37:51.211279+00:00","analysis_stages":{"external_f0_s":1.2185250010006712},"evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"19e6aede37a5e7899698c39ba7821a11483789942e82839f9faaa17a6b5fee80","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.707712509256645,"output_peak":0.23330455880494824},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.99210765898403,"spectral_convergence":0.4519236861647337,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.528143459455534,"spectral_convergence":0.23103678093143426,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.676022505658759,"spectral_convergence":0.2442764950259125,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":5.164243077718805e-05,"delta_db":-0.24169957692685082,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0010795332733059737,"delta_db":-0.30140201363406843,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0009058021628378527,"delta_db":-0.29816954211290014,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":4.8136357064758757e-05,"delta_db":-0.05300946992558875,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":2.7584383718520665e-05,"delta_db":0.415037959660418,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":2.3248330678095503e-05,"delta_db":-1.2815257741127204,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4318073893954244,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.29659286290540243,"pitch":null,"output_rms_dbfs":-26.70409540339157,"output_peak":0.23330455880494824,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.732091208032775},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9028994045627541,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04782168483853718,"rms_dbfs":-26.407502540486163,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.755646813219986,"mean":-4.043451944986979e-05,"zero_energy":false},"output":{"rms":0.046216305979409365,"rms_dbfs":-26.70409540339157,"peak":0.23330455880494824,"peak_dbfs":-12.641535498922243,"crest_factor_db":14.062559904469325,"mean":3.6559314279520424e-05,"zero_energy":false},"rms_difference_db":-0.2965928629054062},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":141,"quiet_samples":22560,"reference_mean_power":6.018758005072885e-07,"output_mean_power":1.2265690706930712e-06,"output_rms_dbfs":-59.1130799073012,"output_zero_energy":false,"output_peak":0.027451962731572688,"output_energy_fraction":0.00013494891957904466,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":289,"both_estimated_voiced_frames":261,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9422382671480144,"absolute_difference_cents_median":5.217596810769087,"absolute_difference_cents_p95":21.914142838126565,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014140982422473965,"log_rms_mae_db":2.023645619097862,"rms_envelope_nrmse":0.14150555442409105,"envelope_correlation":0.9808713797391531,"rms_gain_error_db":-0.29659252836043937,"reference_mean_power":0.0022888204000268467,"output_mean_power":0.002137728084304918},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0460573675773498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006367944533168534,"output_power":0.0006219937441164806,"reference_active":true,"delta_db":-0.10213255233043442},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21105688014902e-05,"output_power":9.077202003833182e-05,"reference_active":true,"delta_db":-0.06357463837132733},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2041191998800965e-05,"output_power":9.2417092360367e-06,"reference_active":true,"delta_db":-1.149171807836402}],"active_band_count":3,"active_band_mae_db":0.4382929995127212}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274322977924003,"log_rms_mae_db":1.496719474377867,"rms_envelope_nrmse":0.09355584585019855,"envelope_correlation":0.991709314678821,"rms_gain_error_db":-0.2965907273382674,"reference_mean_power":0.0023003276499839287,"output_mean_power":0.0021484765950603893},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.201135437818201e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005685810358795958,"output_power":0.000558425484540431,"reference_active":true,"delta_db":-0.07827140677895168},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6960814797233804e-05,"output_power":4.633585071284131e-05,"reference_active":true,"delta_db":-0.058184826131548864},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.369888538054512e-07,"output_power":7.239324694761187e-07,"reference_active":true,"delta_db":-1.1203636881995107}],"active_band_count":3,"active_band_mae_db":0.4189399737033371}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012682596688393775,"log_rms_mae_db":2.657963895598476,"rms_envelope_nrmse":0.16129767065440856,"envelope_correlation":0.9777655459991993,"rms_gain_error_db":-0.2577244482039282,"reference_mean_power":0.0016135328798776547,"output_mean_power":0.0015205663483096642},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.436393406849114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006358276871928121,"output_power":0.0006074629563683099,"reference_active":true,"delta_db":-0.1981963598195005},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.784078725837928e-05,"output_power":8.752818901099642e-05,"reference_active":true,"delta_db":-0.015482765602613776},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0842784949487312e-05,"output_power":9.554754311949588e-06,"reference_active":true,"delta_db":-0.5492132005556889}],"active_band_count":3,"active_band_mae_db":0.25429744199260107}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011190972053745482,"log_rms_mae_db":2.002870395175899,"rms_envelope_nrmse":0.10786513221204867,"envelope_correlation":0.9896574924537346,"rms_gain_error_db":-0.2577227767285724,"reference_mean_power":0.0016216465744125685,"output_mean_power":0.001528213146245069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.451026691855673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005862792692297138,"output_power":0.0005625644771222957,"reference_active":true,"delta_db":-0.17932232483633326},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3209611631795496e-05,"output_power":4.272783922612511e-05,"reference_active":true,"delta_db":-0.04869432062141527},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.999199919742322e-07,"output_power":8.718925992573846e-07,"reference_active":true,"delta_db":-0.595022602370866}],"active_band_count":3,"active_band_mae_db":0.2743464159428715}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":1.9011434894150379,"rms_envelope_nrmse":0.17436183792241394,"envelope_correlation":0.9765874863339242,"rms_gain_error_db":0.10616194479751295,"reference_mean_power":6.869686019984955e-05,"output_mean_power":7.039682665877174e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5736889079386925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.358608717262281e-05,"output_power":1.4027281334846978e-05,"reference_active":true,"delta_db":0.13879110567450212},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4455610806681207e-06,"output_power":2.6260852766492314e-06,"reference_active":true,"delta_db":0.30930310499954816},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901755177746091e-07,"output_power":5.668695290041113e-07,"reference_active":true,"delta_db":-0.17498076775883}],"active_band_count":3,"active_band_mae_db":0.20769165947762677}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":1.1849090484544527,"rms_envelope_nrmse":0.10448834277789973,"envelope_correlation":0.991124671030171,"rms_gain_error_db":0.10617200484698303,"reference_mean_power":6.904136883729848e-05,"output_mean_power":7.075002436267016e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.335283050026158e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2478970323107016e-05,"output_power":1.284946666131311e-05,"reference_active":true,"delta_db":0.12706350022221446},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3575517045580523e-06,"output_power":1.4239448815178718e-06,"reference_active":true,"delta_db":0.2073679944069778},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.985180184805259e-08,"output_power":3.599921743135664e-08,"reference_active":true,"delta_db":-0.44154902279972663}],"active_band_count":3,"active_band_mae_db":0.2586601724763063}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":1.8152707017368124,"rms_envelope_nrmse":0.2635518885221052,"envelope_correlation":0.9417354613256869,"rms_gain_error_db":-0.4577106929507173,"reference_mean_power":1.4641647362050636e-05,"output_mean_power":1.3177070622592499e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3222330564047251e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0304744672556675e-06,"output_power":5.175645517593052e-06,"reference_active":true,"delta_db":-0.6638695937489637},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1757655511106124e-06,"output_power":1.8964927189407851e-06,"reference_active":true,"delta_db":-0.596609166021276},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.148040946601495e-06,"output_power":7.888398789460369e-07,"reference_active":true,"delta_db":-1.6296852032650526}],"active_band_count":3,"active_band_mae_db":0.9633879876784307}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":1.1640751985699047,"rms_envelope_nrmse":0.1474752558038443,"envelope_correlation":0.9815419059906797,"rms_gain_error_db":-0.45773469011884665,"reference_mean_power":1.4714471928104231e-05,"output_mean_power":1.3242537510286715e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1752743001544044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9652177739211e-06,"output_power":5.010990819988874e-06,"reference_active":true,"delta_db":-0.7570269608288198},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4021233434065379e-06,"output_power":1.1622504832294611e-06,"reference_active":true,"delta_db":-0.8148648434604551},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593565327256369e-08,"output_power":5.932855515757289e-08,"reference_active":true,"delta_db":-2.0871626523304627}],"active_band_count":3,"active_band_mae_db":1.2196848188732459}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.315438985824585,"predicted_mos":2.315438985824585,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"7d7dcfe876ff0f37bf896ac0f70c5fee3c2353d2bdabfd599e5bcb2ad81aeb81","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6005519880000065},"started_at":"2026-10-13T05:10:43.891167+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00293876500109036,"native_input_samples":144000,"native_input_sha256":"2826551247918c5247905fb74a6a5e7fff2475ed17d1902981e13242417eef0c","output_sha256":"30600421db5b1e811ec9d43d73fc66950b2269672d84a5581e8f794bd2c74b72","output_samples":144240,"length_delta_samples":240,"analysis_s":0.8907661600005667,"synthesis_s":0.16842364199874282,"total_s":1.0591898019993096,"rtf":0.17653163366655159,"output_peak":0.37040773034095764,"output_rms":0.04729500261439542,"finished_at":"2026-10-13T05:10:44.964292+00:00","analysis_stages":{"external_f0_s":0.2229870999999548},"evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"30600421db5b1e811ec9d43d73fc66950b2269672d84a5581e8f794bd2c74b72","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.50233701718557,"output_peak":0.3691061740909764},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.6852472876579485,"spectral_convergence":0.14180681725808966,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.610349727755995,"spectral_convergence":0.1282821949957885,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.304154761658184,"spectral_convergence":0.15121625415713835,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":5.51584848115579e-05,"delta_db":0.0443569397732728,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.001142090567449228,"delta_db":-0.056756893355876345,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0009480542085197949,"delta_db":-0.10017137102595186,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":4.861415953037837e-05,"delta_db":-0.010113851683230233,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":2.3972191081051976e-05,"delta_db":-0.19451758791367302,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":2.3356275978824427e-05,"delta_db":-1.2614075344577267,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27788736303495515,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08760240299734756,"pitch":null,"output_rms_dbfs":-26.495104943483515,"output_peak":0.3691061740909764,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.866583925690709},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9544095368463676,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04782168483853718,"rms_dbfs":-26.407502540486163,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.755646813219986,"mean":-4.043451944986979e-05,"zero_energy":false},"output":{"rms":0.04734179852277304,"rms_dbfs":-26.495104943483515,"peak":0.3691061740909764,"peak_dbfs":-8.656973804771992,"crest_factor_db":17.83813113871152,"mean":-1.90301468442174e-05,"zero_energy":false},"rms_difference_db":-0.08760240299735145},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":141,"quiet_samples":22560,"reference_mean_power":6.018758005072885e-07,"output_mean_power":6.703181172983619e-07,"output_rms_dbfs":-61.73719042375073,"output_zero_energy":false,"output_peak":0.012742896710279767,"output_energy_fraction":7.028446028735582e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":266,"both_estimated_voiced_frames":261,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9422382671480144,"absolute_difference_cents_median":0.7421285206382945,"absolute_difference_cents_p95":4.83612379802865,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014140982422473965,"log_rms_mae_db":0.7726171575711224,"rms_envelope_nrmse":0.0471124369286081,"envelope_correlation":0.9978934662179161,"rms_gain_error_db":-0.08760227316937075,"reference_mean_power":0.0022888204000268467,"output_mean_power":0.0022431147353540876},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0460573675773498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006367944533168534,"output_power":0.0006435840351557534,"reference_active":true,"delta_db":0.04605990482717371},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21105688014902e-05,"output_power":8.569748686808845e-05,"reference_active":true,"delta_db":-0.31341378013830107},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2041191998800965e-05,"output_power":9.748076533412066e-06,"reference_active":true,"delta_db":-0.9175055111851722}],"active_band_count":3,"active_band_mae_db":0.4256597320502156}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274322977924003,"log_rms_mae_db":0.5299670532354984,"rms_envelope_nrmse":0.028854951894483623,"envelope_correlation":0.9992464625457967,"rms_gain_error_db":-0.08760151674324064,"reference_mean_power":0.0023003276499839287,"output_mean_power":0.00225439258863599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.201135437818201e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005685810358795958,"output_power":0.0005773512099868693,"reference_active":true,"delta_db":0.06647709935181001},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6960814797233804e-05,"output_power":4.389725912730485e-05,"reference_active":true,"delta_db":-0.2929821910963249},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.369888538054512e-07,"output_power":8.066908540676443e-07,"reference_active":true,"delta_db":-0.6502729153865141}],"active_band_count":3,"active_band_mae_db":0.336577401944883}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012682596688393775,"log_rms_mae_db":0.9965662034064786,"rms_envelope_nrmse":0.04891807160823204,"envelope_correlation":0.9979996350337544,"rms_gain_error_db":-0.08967762447675012,"reference_mean_power":0.0016135328798776547,"output_mean_power":0.001580556616868683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.436393406849114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006358276871928121,"output_power":0.0006257633907532633,"reference_active":true,"delta_db":-0.06929283518971523},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.784078725837928e-05,"output_power":8.288054253091319e-05,"reference_active":true,"delta_db":-0.25243634139869014},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0842784949487312e-05,"output_power":8.012626894436473e-06,"reference_active":true,"delta_db":-1.3136592385395915}],"active_band_count":3,"active_band_mae_db":0.5451294717093323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011190972053745482,"log_rms_mae_db":0.4948485781374383,"rms_envelope_nrmse":0.027078445143970505,"envelope_correlation":0.9994018673399957,"rms_gain_error_db":-0.08967682796798199,"reference_mean_power":0.0016216465744125685,"output_mean_power":0.0015885047806930407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.451026691855673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005862792692297138,"output_power":0.0005774296001579415,"reference_active":true,"delta_db":-0.06605494961000999},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3209611631795496e-05,"output_power":4.077413291496167e-05,"reference_active":true,"delta_db":-0.25195628595516106},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.999199919742322e-07,"output_power":7.767483161342651e-07,"reference_active":true,"delta_db":-1.0968493111853568}],"active_band_count":3,"active_band_mae_db":0.471620182250176}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":0.9531982323829857,"rms_envelope_nrmse":0.08374345810630897,"envelope_correlation":0.9945628696354588,"rms_gain_error_db":-0.09818514318745279,"reference_mean_power":6.869686019984955e-05,"output_mean_power":6.71611886222949e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5736889079386925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.358608717262281e-05,"output_power":1.2860268498058521e-05,"reference_active":true,"delta_db":-0.23844360904687278},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4455610806681207e-06,"output_power":2.3014319137116857e-06,"reference_active":true,"delta_db":-0.2638038317617518},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901755177746091e-07,"output_power":5.161323375577452e-07,"reference_active":true,"delta_db":-0.5822011967597711}],"active_band_count":3,"active_band_mae_db":0.36148287918946526}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":0.6356990310694495,"rms_envelope_nrmse":0.0499278846564298,"envelope_correlation":0.9980463870152169,"rms_gain_error_db":-0.09818129296845031,"reference_mean_power":6.904136883729848e-05,"output_mean_power":6.749805584359537e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.335283050026158e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2478970323107016e-05,"output_power":1.1850266850453596e-05,"reference_active":true,"delta_db":-0.22450621772998394},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3575517045580523e-06,"output_power":1.2954681551018165e-06,"reference_active":true,"delta_db":-0.20329637658797026},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.985180184805259e-08,"output_power":3.3132125959718e-08,"reference_active":true,"delta_db":-0.8019865847884656}],"active_band_count":3,"active_band_mae_db":0.4099297263688066}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":0.9777049836630354,"rms_envelope_nrmse":0.13669041528456938,"envelope_correlation":0.984882762197883,"rms_gain_error_db":-0.24309650004125893,"reference_mean_power":1.4641647362050636e-05,"output_mean_power":1.3844596183161958e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3222330564047251e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0304744672556675e-06,"output_power":5.975777343602011e-06,"reference_active":true,"delta_db":-0.039570756459898174},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1757655511106124e-06,"output_power":2.1078716332759905e-06,"reference_active":true,"delta_db":-0.13767936931015393},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.148040946601495e-06,"output_power":8.086446706193784e-07,"reference_active":true,"delta_db":-1.5219964943026951}],"active_band_count":3,"active_band_mae_db":0.5664155400242491}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":0.6971096099483624,"rms_envelope_nrmse":0.07833796927532416,"envelope_correlation":0.9949971300705104,"rms_gain_error_db":-0.24307942061639942,"reference_mean_power":1.4714471928104231e-05,"output_mean_power":1.3913511096625912e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1752743001544044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9652177739211e-06,"output_power":5.854848256511071e-06,"reference_active":true,"delta_db":-0.08110659516406465},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4021233434065379e-06,"output_power":1.2355680580511026e-06,"reference_active":true,"delta_db":-0.5491954738972662},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593565327256369e-08,"output_power":5.566528066970018e-08,"reference_active":true,"delta_db":-2.3639563405783863}],"active_band_count":3,"active_band_mae_db":0.9980861365465724}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0016","method":"magphase","status":"measured","mos":{"status":"measured","value":2.541001558303833,"predicted_mos":2.541001558303833,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288504,"input_sample_rate":48000,"input_seconds":6.0105,"input_float32_sha256":"edede427a761bfaff30fe7052386685310cb0073ff9dd30c1330240f75bb7347","internal_samples":96168,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7694991779990232},"started_at":"2026-10-13T05:15:31.765059+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009051966999322758,"native_input_samples":288000,"native_input_sha256":"5a3250c83bc133518dd6c6db4800862ccfd23bfb7256ac53f18abc3c637ea444","output_sha256":"a2043bd4e6c0547918fbfd4a860b4cc6c7f07d80d8e7c7b02ed8afad6649660e","output_samples":288504,"length_delta_samples":504,"analysis_s":3.7935701349997544,"synthesis_s":1.124647537999408,"total_s":4.9182176729991625,"rtf":0.8197029454998604,"output_peak":0.3879115615145112,"output_rms":0.04647802242957979,"finished_at":"2026-10-13T05:15:36.703955+00:00","evaluation_input":{"reference":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"a2043bd4e6c0547918fbfd4a860b4cc6c7f07d80d8e7c7b02ed8afad6649660e","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96168,"length_delta_samples":168,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.655068943506425,"output_peak":0.3781857400986015},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982530571499876,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.74087747493902,"spectral_convergence":0.3060864855100799,"spectral_convergence_status":"ok","amplitude_floor":9.54303147153e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.531381476468239,"spectral_convergence":0.16069222594330237,"spectral_convergence_status":"ok","amplitude_floor":7.553612828866807e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.937386096514804,"spectral_convergence":0.17213581258986016,"spectral_convergence_status":"ok","amplitude_floor":6.230060381303045e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.459798737171228e-05,"output_power":5.0456110470679045e-05,"delta_db":-0.3426286489162514,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011571142287648336,"output_power":0.0010805932829022096,"delta_db":-0.29713970452038796,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009701755098987012,"output_power":0.0009261515933581725,"delta_db":-0.20168229323455744,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.8727504132367385e-05,"output_power":5.1416619295217725e-05,"delta_db":0.2332935087147221,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.5070297613490014e-05,"output_power":2.7340630518168325e-05,"delta_db":0.3764903629703204,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.1228013015272054e-05,"output_power":2.801746812241845e-05,"delta_db":-0.47115465222420944,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.32039819509674144,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23997365389470265,"pitch":null,"output_rms_dbfs":-26.647476194380868,"output_peak":0.3781857400986015,"reference_rms_dbfs":-26.407502540486163,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.736548349307355},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9396004231005346,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04782168483853718,"rms_dbfs":-26.407502540486163,"peak":0.36932373046875,"peak_dbfs":-8.651855727266177,"crest_factor_db":17.755646813219986,"mean":-4.043451944986979e-05,"zero_energy":false},"output":{"rms":0.046518552263230054,"rms_dbfs":-26.647476194380868,"peak":0.3781857400986015,"peak_dbfs":-8.445897014486171,"crest_factor_db":18.201579179894697,"mean":-0.000698982315559327,"zero_energy":false},"rms_difference_db":-0.2399736538947046},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9996738387271764e-06,"threshold_dbfs":-56.99040835132241,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":141,"quiet_samples":22560,"reference_mean_power":6.018758005072885e-07,"output_mean_power":5.417007130725897e-07,"output_rms_dbfs":-62.662405927249864,"output_zero_energy":false,"output_peak":0.007394442139652311,"output_energy_fraction":5.882675452294682e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":273,"both_estimated_voiced_frames":258,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9314079422382672,"absolute_difference_cents_median":3.786446588618187,"absolute_difference_cents_p95":14.910307648215323,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014140982422473965,"log_rms_mae_db":1.1328513488398757,"rms_envelope_nrmse":0.08501918087931593,"envelope_correlation":0.9935598755678224,"rms_gain_error_db":-0.2399733270299922,"reference_mean_power":0.0022888204000268467,"output_mean_power":0.0021657802196820806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0460573675773498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006367944533168534,"output_power":0.0006212188501093976,"reference_active":true,"delta_db":-0.10754646522804023},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.21105688014902e-05,"output_power":8.536538278980321e-05,"reference_active":true,"delta_db":-0.33027672065980296},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2041191998800965e-05,"output_power":1.1635673139604717e-05,"reference_active":true,"delta_db":-0.1487796846758014}],"active_band_count":3,"active_band_mae_db":0.19553429018788152}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001274322977924003,"log_rms_mae_db":0.8837971204865889,"rms_envelope_nrmse":0.056945538396619845,"envelope_correlation":0.9973763185747274,"rms_gain_error_db":-0.23997335196711708,"reference_mean_power":0.0023003276499839287,"output_mean_power":0.0021766688615926033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.201135437818201e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005685810358795958,"output_power":0.0005574665663542697,"reference_active":true,"delta_db":-0.08573544380694118},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6960814797233804e-05,"output_power":4.280868073791606e-05,"reference_active":true,"delta_db":-0.40203779450905025},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.369888538054512e-07,"output_power":9.17933677868778e-07,"reference_active":true,"delta_db":-0.08923120770379835}],"active_band_count":3,"active_band_mae_db":0.1923348153399299}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012682596688393775,"log_rms_mae_db":1.6828158698627897,"rms_envelope_nrmse":0.09026638700566444,"envelope_correlation":0.993494465245313,"rms_gain_error_db":-0.24682203206896308,"reference_mean_power":0.0016135328798776547,"output_mean_power":0.0015243883339677293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.436393406849114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006358276871928121,"output_power":0.0006012003108079127,"reference_active":true,"delta_db":-0.2432023895972045},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.784078725837928e-05,"output_power":8.09176784304707e-05,"reference_active":true,"delta_db":-0.35652805268219356},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0842784949487312e-05,"output_power":1.0129759609008304e-05,"reference_active":true,"delta_db":-0.2954170510499617}],"active_band_count":3,"active_band_mae_db":0.29838249777645326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011190972053745482,"log_rms_mae_db":1.1604237565665867,"rms_envelope_nrmse":0.06060055048790922,"envelope_correlation":0.9972404850236466,"rms_gain_error_db":-0.24682133200847597,"reference_mean_power":0.0016216465744125685,"output_mean_power":0.0015320540096570938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.451026691855673e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005862792692297138,"output_power":0.0005561944557267401,"reference_active":true,"delta_db":-0.22877882448108108},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3209611631795496e-05,"output_power":3.872617611815464e-05,"reference_active":true,"delta_db":-0.47575746685915715},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.999199919742322e-07,"output_power":9.188224711435158e-07,"reference_active":true,"delta_db":-0.36733643608017974}],"active_band_count":3,"active_band_mae_db":0.35729090914013933}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.381727808324488e-05,"log_rms_mae_db":1.1507627976674382,"rms_envelope_nrmse":0.11175735858455428,"envelope_correlation":0.9912338616439661,"rms_gain_error_db":0.2470375441390007,"reference_mean_power":6.869686019984955e-05,"output_mean_power":7.271778512143584e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5736889079386925e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.358608717262281e-05,"output_power":1.470072452193159e-05,"reference_active":true,"delta_db":0.34244342548288575},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4455610806681207e-06,"output_power":2.6888433481346733e-06,"reference_active":true,"delta_db":0.4118698669657348},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.901755177746091e-07,"output_power":5.905976462794143e-07,"reference_active":true,"delta_db":0.003105221115828445}],"active_band_count":3,"active_band_mae_db":0.25247283785481633}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.11127527801434e-05,"log_rms_mae_db":0.781735572539182,"rms_envelope_nrmse":0.06788156309488413,"envelope_correlation":0.9970273989788954,"rms_gain_error_db":0.2470318518855278,"reference_mean_power":6.904136883729848e-05,"output_mean_power":7.308236255099194e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.335283050026158e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2478970323107016e-05,"output_power":1.3538562691671288e-05,"reference_active":true,"delta_db":0.3539380844790851},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3575517045580523e-06,"output_power":1.4571619925740977e-06,"reference_active":true,"delta_db":0.30751455548797146},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.985180184805259e-08,"output_power":3.830607562256909e-08,"reference_active":true,"delta_db":-0.17180300543915228}],"active_band_count":3,"active_band_mae_db":0.2777518818020696}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.6537855240004908e-05,"log_rms_mae_db":1.4380691665417764,"rms_envelope_nrmse":0.2412761690648421,"envelope_correlation":0.9675371714448298,"rms_gain_error_db":0.9624722329541793,"reference_mean_power":1.4641647362050636e-05,"output_mean_power":1.8274149188789125e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3222330564047251e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0304744672556675e-06,"output_power":7.804020632668353e-06,"reference_active":true,"delta_db":1.1196692591407833},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1757655511106124e-06,"output_power":2.439926246476019e-06,"reference_active":true,"delta_db":0.4976460249378677},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.148040946601495e-06,"output_power":1.1074030719557045e-06,"reference_active":true,"delta_db":-0.15651654183823957}],"active_band_count":3,"active_band_mae_db":0.5912772753056301}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.7197117785089623e-05,"log_rms_mae_db":1.1582603916432903,"rms_envelope_nrmse":0.18273950700236202,"envelope_correlation":0.9840344427268061,"rms_gain_error_db":0.9624640900464283,"reference_mean_power":1.4714471928104231e-05,"output_mean_power":1.8365006643704634e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1752743001544044e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9652177739211e-06,"output_power":7.5986569162900865e-06,"reference_active":true,"delta_db":1.051105330990329},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4021233434065379e-06,"output_power":1.4869006549798657e-06,"reference_active":true,"delta_db":0.25495732993758724},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593565327256369e-08,"output_power":8.320359518856589e-08,"reference_active":true,"delta_db":-0.6183794486408422}],"active_band_count":3,"active_band_mae_db":0.6414807031895862}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"hnm","status":"measured","mos":{"status":"measured","value":2.774578809738159,"predicted_mos":2.774578809738159,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b971df5ab58959aa35b262fec7fa9629ac7ff56f75e85c2bda44134d85cb6b1e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.742109127999356},"started_at":"2026-10-13T04:39:17.901674+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4993999684520531e-05,"native_input_samples":96000,"native_input_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","output_sha256":"204e0eda44357bca6c62627e4b1956cb448ace2873e78d7e613dbb471ef2d174","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9753236370006562,"synthesis_s":0.497915381999519,"total_s":1.4732390190001752,"rtf":0.2455398365000292,"output_peak":0.4812303712804796,"output_rms":0.04197479807357494,"finished_at":"2026-10-13T04:39:19.380639+00:00","evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"204e0eda44357bca6c62627e4b1956cb448ace2873e78d7e613dbb471ef2d174","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.403932672249499,"spectral_convergence":0.26215348402798544,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.501946007295421,"spectral_convergence":0.2942510372697704,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.026702664802622,"spectral_convergence":0.2954814181783468,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":5.147227041812338e-05,"delta_db":0.7303406277309281,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.0012435475554170568,"delta_db":-0.060115998158912574,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.0003496907288978534,"delta_db":-0.14924703888434532,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":3.536524813701356e-05,"delta_db":0.742271593065357,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":1.8522335470244567e-05,"delta_db":0.505262745028505,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":6.328553497709821e-05,"delta_db":0.1530791926305427,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.39005286591643173,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.029013423575383886,"pitch":null,"output_rms_dbfs":-27.540227688547812,"output_peak":0.4812303712804796,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.64419378144918},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.403932672249499,"spectral_convergence":0.26215348402798544,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.501946007295421,"spectral_convergence":0.2942510372697704,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.026702664802622,"spectral_convergence":0.2954814181783468,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":5.147227041812338e-05,"delta_db":0.7303406277309281,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.0012435475554170568,"delta_db":-0.060115998158912574,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.0003496907288978534,"delta_db":-0.14924703888434532,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":3.536524813701356e-05,"delta_db":0.742271593065357,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":1.8522335470244567e-05,"delta_db":0.505262745028505,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":6.328553497709821e-05,"delta_db":0.1530791926305427,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.39005286591643173,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.029013423575383886,"pitch":null,"output_rms_dbfs":-27.540227688547812,"output_peak":0.4812303712804796,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.64419378144918},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8713903990893059,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042115240661813005,"rms_dbfs":-27.511214264972427,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.679901434978223,"mean":-3.424930572509765e-05,"zero_energy":false},"output":{"rms":0.04197479807357494,"rms_dbfs":-27.54022768854781,"peak":0.4812303712804796,"peak_dbfs":-6.352939427837228,"crest_factor_db":21.18728826071058,"mean":-8.307268700659901e-06,"zero_energy":false},"rms_difference_db":-0.029013423575381836},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":2.296827609419294e-06,"output_rms_dbfs":-56.3887159996715,"output_zero_energy":false,"output_peak":0.02630887725793925,"output_energy_fraction":0.00014122555012495417,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":278,"output_estimated_voiced_frames":257,"both_estimated_voiced_frames":254,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9136690647482014,"absolute_difference_cents_median":0.9353507310151984,"absolute_difference_cents_p95":10.328712086687585,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018136510927426692,"log_rms_mae_db":0.7658343517362223,"rms_envelope_nrmse":0.07811514613296061,"envelope_correlation":0.9941773191656993,"rms_gain_error_db":-0.028874490930917637,"reference_mean_power":0.0017749427149271913,"output_mean_power":0.0017631809785443237},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.578419076835135e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003061888162786739,"output_power":0.0003038524431034347,"reference_active":true,"delta_db":-0.033265911875138124},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.561280729204581e-05,"output_power":5.278284000929531e-05,"reference_active":true,"delta_db":-0.226820645215329},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1447257663776475e-05,"output_power":8.830063214479544e-06,"reference_active":true,"delta_db":-1.1273764566144198}],"active_band_count":3,"active_band_mae_db":0.4624876712349623}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015531555176710683,"log_rms_mae_db":0.4203790910690266,"rms_envelope_nrmse":0.0353358342209486,"envelope_correlation":0.9987245867692061,"rms_gain_error_db":-0.028881014629343595,"reference_mean_power":0.0017815693662331493,"output_mean_power":0.0017697610596266954},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.067921014076438e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002861407755621019,"output_power":0.00028377220010734785,"reference_active":true,"delta_db":-0.036099026219196625},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5605898871779113e-05,"output_power":2.433943683527224e-05,"reference_active":true,"delta_db":-0.22029500625295984},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.159270454376387e-07,"output_power":7.062088094895527e-07,"reference_active":true,"delta_db":-1.129277522301332}],"active_band_count":3,"active_band_mae_db":0.4618905182578295}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017559797263300874,"log_rms_mae_db":1.4160035814085474,"rms_envelope_nrmse":0.09744593557973429,"envelope_correlation":0.9925654356062094,"rms_gain_error_db":-0.10610478882890659,"reference_mean_power":0.0012870696988326355,"output_mean_power":0.0012560056327247892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.703155295091611e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025376682215949657,"output_power":0.00024848261753122175,"reference_active":true,"delta_db":-0.09138827889455917},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.494529209431561e-05,"output_power":4.0796263303673814e-05,"reference_active":true,"delta_db":-0.4206382109623067},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.800066525463509e-06,"output_power":6.20564775361996e-06,"reference_active":true,"delta_db":-1.5169883485262048}],"active_band_count":3,"active_band_mae_db":0.6763382794610235}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014992083957937532,"log_rms_mae_db":0.8263704310701555,"rms_envelope_nrmse":0.0441744794829555,"envelope_correlation":0.998479148915275,"rms_gain_error_db":-0.10631768408984402,"reference_mean_power":0.0012919551058812144,"output_mean_power":0.0012607113252731694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4067631484646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000244569147219292,"output_power":0.00023811395566107325,"reference_active":true,"delta_db":-0.11616819437566936},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8830927140807284e-05,"output_power":1.715332680196748e-05,"reference_active":true,"delta_db":-0.4052334124332706},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.050384653726171e-07,"output_power":4.826348954001593e-07,"reference_active":true,"delta_db":-1.64594092798307}],"active_band_count":3,"active_band_mae_db":0.7224475115973367}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":1.4012742850789017,"rms_envelope_nrmse":0.1844477414628217,"envelope_correlation":0.9771225759951964,"rms_gain_error_db":0.3343341076458267,"reference_mean_power":4.199153922743894e-05,"output_mean_power":4.535187062167404e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0299126624198237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.222152478948828e-06,"output_power":3.8105216270740705e-06,"reference_active":true,"delta_db":-0.44549482519307215},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.960031981806198e-07,"output_power":7.926388627936435e-07,"reference_active":true,"delta_db":-0.532341980093663},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.633269352164056e-07,"output_power":2.076312035572389e-07,"reference_active":true,"delta_db":-1.0320266313933957}],"active_band_count":3,"active_band_mae_db":0.669954478893377}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":0.8774579310817154,"rms_envelope_nrmse":0.1156071582516033,"envelope_correlation":0.9914321486169919,"rms_gain_error_db":0.33397737223540824,"reference_mean_power":4.2190507757870256e-05,"output_mean_power":4.5563018649752726e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.548028980657063e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9589316603894534e-06,"output_power":3.5461653947236186e-06,"reference_active":true,"delta_db":-0.4781901754303328},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7410748918107177e-07,"output_power":3.2037899112006274e-07,"reference_active":true,"delta_db":-0.6733237288950389},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.800283314283636e-08,"output_power":1.2773167510857277e-08,"reference_active":true,"delta_db":-1.4904224862540272}],"active_band_count":3,"active_band_mae_db":0.8806454635264663}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":1.4100385275228906,"rms_envelope_nrmse":0.3711946964860151,"envelope_correlation":0.9054949760539069,"rms_gain_error_db":0.7237037548012675,"reference_mean_power":2.209560749738961e-05,"output_mean_power":2.6102152462345264e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.718240736767724e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.504526854650136e-06,"output_power":9.824491330414104e-06,"reference_active":true,"delta_db":0.14379571640145003},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.566561169291012e-06,"output_power":2.859167704003481e-06,"reference_active":true,"delta_db":0.4688801024054703},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.915786881504461e-07,"output_power":8.67866009232298e-07,"reference_active":true,"delta_db":1.664401591782132}],"active_band_count":3,"active_band_mae_db":0.7590258035296841}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":0.9800544480945683,"rms_envelope_nrmse":0.28548310363535023,"envelope_correlation":0.9415588465499863,"rms_gain_error_db":0.7236176127848616,"reference_mean_power":2.220381322589701e-05,"output_mean_power":2.6229458618495532e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6122764663239423e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.855212413255221e-06,"output_power":1.0417692208906062e-05,"reference_active":true,"delta_db":0.2410553280451454},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3142139667203487e-06,"output_power":1.5560094890148134e-06,"reference_active":true,"delta_db":0.7334616280602847},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.262111729619265e-08,"output_power":5.8242556354848265e-08,"reference_active":true,"delta_db":1.356155981819731}],"active_band_count":3,"active_band_mae_db":0.7768909793083871}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.6104750633239746,"predicted_mos":2.6104750633239746,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"700838d094954ec6244b6d0d55cae990591ef96f002d365a24be9d8aab881e13","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6320132310011104},"started_at":"2026-10-13T04:42:14.339873+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.296699949511094e-05,"native_input_samples":96000,"native_input_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","output_sha256":"c9e5a4d275f516f824e1633c19b0ef46b57db53594967e2e44e3bca8c0fca54d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4748371929999848,"synthesis_s":0.11254729500069516,"total_s":1.58738448800068,"rtf":0.26456408133344667,"output_peak":0.5229714124040729,"output_rms":0.05075214869572941,"finished_at":"2026-10-13T04:42:15.932498+00:00","evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"c9e5a4d275f516f824e1633c19b0ef46b57db53594967e2e44e3bca8c0fca54d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.893071252666425,"spectral_convergence":0.47089230779515784,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.511281380566299,"spectral_convergence":0.3254215533536163,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.029911580750413,"spectral_convergence":0.35330501835989814,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":5.701532825778222e-05,"delta_db":1.1745236662674965,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.001915182702043304,"delta_db":1.8153621261202386,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.00048029069215252906,"delta_db":1.2289534805287567,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":3.722306442809763e-05,"delta_db":0.9646257370858617,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":2.1369352053314654e-05,"delta_db":1.1262188275919847,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":6.46994582984006e-05,"delta_db":0.24904108068014727,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0931208197124143,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6203029393156765,"pitch":null,"output_rms_dbfs":-25.890911325656752,"output_peak":0.5229714124040729,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.811421404661044},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.893071252666425,"spectral_convergence":0.47089230779515784,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.511281380566299,"spectral_convergence":0.3254215533536163,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.029911580750413,"spectral_convergence":0.35330501835989814,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":5.701532825778222e-05,"delta_db":1.1745236662674965,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.001915182702043304,"delta_db":1.8153621261202386,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.00048029069215252906,"delta_db":1.2289534805287567,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":3.722306442809763e-05,"delta_db":0.9646257370858617,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":2.1369352053314654e-05,"delta_db":1.1262188275919847,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":6.46994582984006e-05,"delta_db":0.24904108068014727,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0931208197124143,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.6203029393156765,"pitch":null,"output_rms_dbfs":-25.890911325656752,"output_peak":0.5229714124040729,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.811421404661044},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9141634363561678,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042115240661813005,"rms_dbfs":-27.511214264972427,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.679901434978223,"mean":-3.424930572509765e-05,"zero_energy":false},"output":{"rms":0.05075214869572941,"rms_dbfs":-25.890911325656752,"peak":0.5229714124040729,"peak_dbfs":-5.630441013266109,"crest_factor_db":20.260470312390638,"mean":5.77951853976087e-06,"zero_energy":false},"rms_difference_db":1.6203029393156747},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":2.06903563305343e-06,"output_rms_dbfs":-56.842320298247344,"output_zero_energy":false,"output_peak":0.019482787977379443,"output_energy_fraction":8.702042679988732e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":278,"output_estimated_voiced_frames":289,"both_estimated_voiced_frames":259,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9316546762589928,"absolute_difference_cents_median":3.871639146032415,"absolute_difference_cents_p95":29.44331651838635,"difference_over_600_cents_frames":9,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018136510927426692,"log_rms_mae_db":1.463142018762133,"rms_envelope_nrmse":0.25214988709549235,"envelope_correlation":0.9844065583490558,"rms_gain_error_db":1.6207779247132845,"reference_mean_power":0.0017749427149271913,"output_mean_power":0.002577876654568146},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.578419076835135e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003061888162786739,"output_power":0.00043326541376450514,"reference_active":true,"delta_db":1.5076469823718721},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.561280729204581e-05,"output_power":8.11493180569634e-05,"reference_active":true,"delta_db":1.6411005607446703},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1447257663776475e-05,"output_power":1.5807164564961033e-05,"reference_active":true,"delta_db":1.4015251629448149}],"active_band_count":3,"active_band_mae_db":1.5167575686871189}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015531555176710683,"log_rms_mae_db":1.1436526558190772,"rms_envelope_nrmse":0.23392798312367635,"envelope_correlation":0.9905638421949743,"rms_gain_error_db":1.618916367408831,"reference_mean_power":0.0017815693662331493,"output_mean_power":0.0025863921484734235},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.067921014076438e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002861407755621019,"output_power":0.0004119209816095947,"reference_active":true,"delta_db":1.5823416380596562},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5605898871779113e-05,"output_power":3.6293958237109615e-05,"reference_active":true,"delta_db":1.5149430920481033},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.159270454376387e-07,"output_power":1.1534993114707631e-06,"reference_active":true,"delta_db":1.0015645659625159}],"active_band_count":3,"active_band_mae_db":1.3662830986900918}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017559797263300874,"log_rms_mae_db":2.0099884269065535,"rms_envelope_nrmse":0.28511918576921597,"envelope_correlation":0.987768846341625,"rms_gain_error_db":1.9202896745766072,"reference_mean_power":0.0012870696988326355,"output_mean_power":0.002002769797226876},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.703155295091611e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025376682215949657,"output_power":0.00039379837299521,"reference_active":true,"delta_db":1.9083907630568588},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.494529209431561e-05,"output_power":7.415267429822166e-05,"reference_active":true,"delta_db":2.1744261116910817},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.800066525463509e-06,"output_power":1.302244722757756e-05,"reference_active":true,"delta_db":1.702066508858996}],"active_band_count":3,"active_band_mae_db":1.9282944612023123}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014992083957937532,"log_rms_mae_db":1.5708309671836973,"rms_envelope_nrmse":0.26914141810673803,"envelope_correlation":0.9927256355953055,"rms_gain_error_db":1.9172732820478833,"reference_mean_power":0.0012919551058812144,"output_mean_power":0.0020089760106699285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4067631484646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000244569147219292,"output_power":0.0003830159200227889,"reference_active":true,"delta_db":1.948151565283787},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8830927140807284e-05,"output_power":3.047992627672534e-05,"reference_active":true,"delta_db":2.091422091853175},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.050384653726171e-07,"output_power":8.784860378233961e-07,"reference_active":true,"delta_db":0.9552205164252665}],"active_band_count":3,"active_band_mae_db":1.6649313911874097}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":1.46719170921008,"rms_envelope_nrmse":0.17710748625874861,"envelope_correlation":0.9904994623547482,"rms_gain_error_db":1.0592291123531665,"reference_mean_power":4.199153922743894e-05,"output_mean_power":5.359011703718709e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0299126624198237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.222152478948828e-06,"output_power":5.720697799553557e-06,"reference_active":true,"delta_db":1.3191509303793683},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.960031981806198e-07,"output_power":1.2169596973586282e-06,"reference_active":true,"delta_db":1.3296663589913487},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.633269352164056e-07,"output_power":3.630491376236282e-07,"reference_active":true,"delta_db":1.3947012494088873}],"active_band_count":3,"active_band_mae_db":1.3478395129265348}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":1.0616125400856045,"rms_envelope_nrmse":0.14830058712605088,"envelope_correlation":0.9964869719121868,"rms_gain_error_db":1.0592674468937673,"reference_mean_power":4.2190507757870256e-05,"output_mean_power":5.3844518390023224e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.548028980657063e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9589316603894534e-06,"output_power":5.4109858794191005e-06,"reference_active":true,"delta_db":1.3569839562621175},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7410748918107177e-07,"output_power":4.983142393058479e-07,"reference_active":true,"delta_db":1.2450689495531229},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.800283314283636e-08,"output_power":1.8911687973991043e-08,"reference_active":true,"delta_db":0.21389437491546567}],"active_band_count":3,"active_band_mae_db":0.938649093576902}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":1.3695855142769957,"rms_envelope_nrmse":0.2055121198830826,"envelope_correlation":0.968033064851488,"rms_gain_error_db":0.35014251508649485,"reference_mean_power":2.209560749738961e-05,"output_mean_power":2.3950809588447963e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.718240736767724e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.504526854650136e-06,"output_power":9.598489976129357e-06,"reference_active":true,"delta_db":0.04272413580514439},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.566561169291012e-06,"output_power":2.1229937113827445e-06,"reference_active":true,"delta_db":-0.8240291161048516},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.915786881504461e-07,"output_power":4.416784606984235e-07,"reference_active":true,"delta_db":-1.2690629941389968}],"active_band_count":3,"active_band_mae_db":0.7119387486829977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":0.8957212689986243,"rms_envelope_nrmse":0.1263793900370172,"envelope_correlation":0.9878176814785741,"rms_gain_error_db":0.3499818642939129,"reference_mean_power":2.220381322589701e-05,"output_mean_power":2.4067210246961004e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6122764663239423e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.855212413255221e-06,"output_power":9.60075098511151e-06,"reference_active":true,"delta_db":-0.11360783702840667},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3142139667203487e-06,"output_power":1.0433779131489904e-06,"reference_active":true,"delta_db":-1.0022443923741273},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.262111729619265e-08,"output_power":3.50320224788963e-08,"reference_active":true,"delta_db":-0.8515961953353162}],"active_band_count":3,"active_band_mae_db":0.6558161415792834}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"facodec","status":"measured","mos":{"status":"measured","value":3.7219696044921875,"predicted_mos":3.7219696044921875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"3ac5a0162195243c64dbdd1ee49f237ac5b37a26fb90192b5f5f472e6d7b31f1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1740181090008264},"started_at":"2026-10-13T04:52:40.487043+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.36530004258384e-05,"native_input_samples":96000,"native_input_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","output_sha256":"6cdc1af922bd65e72a6a431478cc5ebbd86b6a159b143fd73ceabe7bfec479ae","output_samples":96000,"length_delta_samples":0,"analysis_s":2.356642792999992,"synthesis_s":6.074219258000085,"total_s":8.430862051000076,"rtf":1.4051436751666795,"output_peak":0.40218839049339294,"output_rms":0.03827473545829037,"finished_at":"2026-10-13T04:52:48.925722+00:00","evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"6cdc1af922bd65e72a6a431478cc5ebbd86b6a159b143fd73ceabe7bfec479ae","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.360103708718251,"spectral_convergence":0.2699531624072774,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.90880093201062,"spectral_convergence":0.24920610503695442,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.155533447905112,"spectral_convergence":0.29462050252551725,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":3.8564809658019116e-05,"delta_db":-0.5234807024597903,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.0010636443348882602,"delta_db":-0.7387756658826579,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.0002882028542521256,"delta_db":-0.9891054462124554,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":1.7270362867592564e-05,"delta_db":-2.3704808768530166,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":1.2918007341240646e-05,"delta_db":-1.0597394421352706,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":4.435500539487166e-05,"delta_db":-1.390539000444433,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.178686855664604,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8305417812737299,"pitch":null,"output_rms_dbfs":-28.341756046246157,"output_peak":0.40218839049339294,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.808146029544662},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.360103708718251,"spectral_convergence":0.2699531624072774,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.90880093201062,"spectral_convergence":0.24920610503695442,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.155533447905112,"spectral_convergence":0.29462050252551725,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":3.8564809658019116e-05,"delta_db":-0.5234807024597903,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.0010636443348882602,"delta_db":-0.7387756658826579,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.0002882028542521256,"delta_db":-0.9891054462124554,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":1.7270362867592564e-05,"delta_db":-2.3704808768530166,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":1.2918007341240646e-05,"delta_db":-1.0597394421352706,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":4.435500539487166e-05,"delta_db":-1.390539000444433,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.178686855664604,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8305417812737299,"pitch":null,"output_rms_dbfs":-28.341756046246157,"output_peak":0.40218839049339294,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.808146029544662},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8662272765418858,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042115240661813005,"rms_dbfs":-27.511214264972427,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.679901434978223,"mean":-3.424930572509765e-05,"zero_energy":false},"output":{"rms":0.03827473545829037,"rms_dbfs":-28.341756046246157,"peak":0.40218839049339294,"peak_dbfs":-7.9114093966864,"crest_factor_db":20.430346649559755,"mean":6.066890956176962e-06,"zero_energy":false},"rms_difference_db":-0.8305417812737304},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":1.326274273391983e-06,"output_rms_dbfs":-58.77366654586379,"output_zero_energy":false,"output_peak":0.007447241805493832,"output_energy_fraction":9.807787695201157e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":278,"output_estimated_voiced_frames":292,"both_estimated_voiced_frames":275,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9892086330935251,"absolute_difference_cents_median":5.328700185980936,"absolute_difference_cents_p95":33.89282594102784,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018136510927426692,"log_rms_mae_db":1.231393194593349,"rms_envelope_nrmse":0.14767777326293577,"envelope_correlation":0.9862897579405573,"rms_gain_error_db":-0.8305602209552905,"reference_mean_power":0.0017749427149271913,"output_mean_power":0.0014659809235075618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.578419076835135e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003061888162786739,"output_power":0.0003022495955771824,"reference_active":true,"delta_db":-0.05623595369327869},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.561280729204581e-05,"output_power":5.559300020177443e-05,"reference_active":true,"delta_db":-0.0015470613077220325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1447257663776475e-05,"output_power":1.1170189987218376e-05,"reference_active":true,"delta_db":-0.1064089852826346}],"active_band_count":3,"active_band_mae_db":0.05473066676121177}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015531555176710683,"log_rms_mae_db":1.0659644053484045,"rms_envelope_nrmse":0.13900184505513932,"envelope_correlation":0.9882500522350842,"rms_gain_error_db":-0.8307130674351486,"reference_mean_power":0.0017815693662331493,"output_mean_power":0.001471402297163818},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.067921014076438e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002861407755621019,"output_power":0.0002836179920023993,"reference_active":true,"delta_db":-0.03845972020454411},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5605898871779113e-05,"output_power":2.5595286144181847e-05,"reference_active":true,"delta_db":-0.0018003681352313455},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.159270454376387e-07,"output_power":9.988653562373782e-07,"reference_active":true,"delta_db":0.3764606767589493}],"active_band_count":3,"active_band_mae_db":0.13890692169957491}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017559797263300874,"log_rms_mae_db":1.5935180209906656,"rms_envelope_nrmse":0.14732074193841077,"envelope_correlation":0.9885351816193032,"rms_gain_error_db":-0.8130505557539864,"reference_mean_power":0.0012870696988326355,"output_mean_power":0.0010673258068831222},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.703155295091611e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025376682215949657,"output_power":0.0002549801351454271,"reference_active":true,"delta_db":0.02071505789183243},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.494529209431561e-05,"output_power":4.485423168193303e-05,"reference_active":true,"delta_db":-0.008807852538565844},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.800066525463509e-06,"output_power":7.72693592415522e-06,"reference_active":true,"delta_db":-0.5647864441874235}],"active_band_count":3,"active_band_mae_db":0.1981031182059406}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014992083957937532,"log_rms_mae_db":1.2472281861713703,"rms_envelope_nrmse":0.13838414259269505,"envelope_correlation":0.9903266174116264,"rms_gain_error_db":-0.8130367490873519,"reference_mean_power":0.0012919551058812144,"output_mean_power":0.001071380524987273},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4067631484646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000244569147219292,"output_power":0.0002458616119087336,"reference_active":true,"delta_db":0.022890553730564527},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8830927140807284e-05,"output_power":1.851244705434658e-05,"reference_active":true,"delta_db":-0.07407873586871494},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.050384653726171e-07,"output_power":6.716140047076602e-07,"reference_active":true,"delta_db":-0.2109306819833913}],"active_band_count":3,"active_band_mae_db":0.10263332386089026}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":2.0280981823716804,"rms_envelope_nrmse":0.2800710425185089,"envelope_correlation":0.9641095880672952,"rms_gain_error_db":-1.9142237907143702,"reference_mean_power":4.199153922743894e-05,"output_mean_power":2.7023364252843994e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0299126624198237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.222152478948828e-06,"output_power":4.042820886253764e-06,"reference_active":true,"delta_db":-0.18849412667907325},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.960031981806198e-07,"output_power":1.0256154340460242e-06,"reference_active":true,"delta_db":0.5867498790576732},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.633269352164056e-07,"output_power":2.8490888265505213e-07,"reference_active":true,"delta_db":0.34210704984018064}],"active_band_count":3,"active_band_mae_db":0.3724503518589757}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":1.760097527685536,"rms_envelope_nrmse":0.26162471558098666,"envelope_correlation":0.9728186225010701,"rms_gain_error_db":-1.914217474133884,"reference_mean_power":4.2190507757870256e-05,"output_mean_power":2.715144856292294e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.548028980657063e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9589316603894534e-06,"output_power":3.819134117245106e-06,"reference_active":true,"delta_db":-0.15613095125742601},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7410748918107177e-07,"output_power":4.257323548806241e-07,"reference_active":true,"delta_db":0.5614025468064704},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.800283314283636e-08,"output_power":1.709683425077519e-08,"reference_active":true,"delta_db":-0.22425154951230344}],"active_band_count":3,"active_band_mae_db":0.31392834919206664}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":1.8162605605076163,"rms_envelope_nrmse":0.24178145669841242,"envelope_correlation":0.9547504434163543,"rms_gain_error_db":-0.7983240638342004,"reference_mean_power":2.209560749738961e-05,"output_mean_power":1.838541935496536e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.718240736767724e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.504526854650136e-06,"output_power":7.88612711825461e-06,"reference_active":true,"delta_db":-0.8106672881924313},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.566561169291012e-06,"output_power":2.4684355366990323e-06,"reference_active":true,"delta_db":-0.16929829250308903},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.915786881504461e-07,"output_power":1.0583865710208228e-06,"reference_active":true,"delta_db":2.5263180083502936}],"active_band_count":3,"active_band_mae_db":1.1687611963486046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":1.4571153423414556,"rms_envelope_nrmse":0.18477345074400656,"envelope_correlation":0.9742165427247672,"rms_gain_error_db":-0.7983063004920552,"reference_mean_power":2.220381322589701e-05,"output_mean_power":1.847553126569287e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6122764663239423e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.855212413255221e-06,"output_power":8.601046279505614e-06,"reference_active":true,"delta_db":-0.5911470476202314},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3142139667203487e-06,"output_power":1.3609772421319241e-06,"reference_active":true,"delta_db":0.15184784802895407},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.262111729619265e-08,"output_power":8.850397329289103e-08,"reference_active":true,"delta_db":3.1733793793977796}],"active_band_count":3,"active_band_mae_db":1.3054580916823217}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.3541202545166016,"predicted_mos":3.3541202545166016,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"1f63a12f2be88593a4ed59d069d2110ddcf85579a137e34e0118570700076fa5","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.757850113001041},"started_at":"2026-10-13T07:14:28.528530+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0059690540001611225,"native_input_samples":264600,"native_input_sha256":"66a2bdce6e3838dc99ae09e173e4b6ce8dfa95b60ef265fbea7bf47387387732","output_sha256":"c8bb81d10f77d85a707ef4e8bee7d48a99de8298b41d7371a005c4b380a6ad5a","output_samples":264704,"length_delta_samples":104,"analysis_s":0.14149197199913033,"synthesis_s":5.335511159999442,"total_s":5.477003131998572,"rtf":0.9128338553330954,"output_peak":0.35254397988319397,"output_rms":0.0384444702932346,"finished_at":"2026-10-13T07:14:34.028874+00:00","evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"c8bb81d10f77d85a707ef4e8bee7d48a99de8298b41d7371a005c4b380a6ad5a","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.307510699130503,"output_peak":0.35296742581469914},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.70275511274877,"spectral_convergence":0.4106120759507937,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.848760845265176,"spectral_convergence":0.3384488222989658,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.86839247150655,"spectral_convergence":0.3941097128204752,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":2.9405714860703656e-05,"delta_db":-1.7010752074892197,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.0010613712046872886,"delta_db":-0.7480669699798281,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.0002839920892532911,"delta_db":-1.0530257949105983,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":4.365172692240803e-05,"delta_db":1.6565187935906729,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":1.873721320827253e-05,"delta_db":0.5553552736904261,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":3.9902685651095835e-05,"delta_db":-1.8499440957952444,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2606643559093318,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7948024967061389,"pitch":null,"output_rms_dbfs":-28.306016761678563,"output_peak":0.35296742581469914,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.806636143173497},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7131543293558581,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042115240661813005,"rms_dbfs":-27.511214264972427,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.679901434978223,"mean":-3.424930572509765e-05,"zero_energy":false},"output":{"rms":0.03843254655344946,"rms_dbfs":-28.306016761678563,"peak":0.35296742581469914,"peak_dbfs":-9.045307447013437,"crest_factor_db":19.260709314665128,"mean":-6.964271263493179e-05,"zero_energy":false},"rms_difference_db":-0.7948024967061365},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":1.6001928907131282e-06,"output_rms_dbfs":-57.9582766339212,"output_zero_energy":false,"output_peak":0.015986293284838728,"output_energy_fraction":0.0001173643287001481,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":278,"output_estimated_voiced_frames":242,"both_estimated_voiced_frames":238,"reference_voiced_output_unvoiced_frames":40,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.8561151079136691,"absolute_difference_cents_median":7.79613449302934,"absolute_difference_cents_p95":44.20600688783545,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018136510927426692,"log_rms_mae_db":1.440057568201466,"rms_envelope_nrmse":0.12785904653785185,"envelope_correlation":0.9912902020841284,"rms_gain_error_db":-0.7945302172566178,"reference_mean_power":0.0017749427149271913,"output_mean_power":0.0014781936058212068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.578419076835135e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003061888162786739,"output_power":0.00025595663028517925,"reference_active":true,"delta_db":-0.7782293982788004},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.561280729204581e-05,"output_power":4.370415713260776e-05,"reference_active":true,"delta_db":-1.0465206952545165},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1447257663776475e-05,"output_power":1.2001700143129104e-05,"reference_active":true,"delta_db":0.20541313553757334}],"active_band_count":3,"active_band_mae_db":0.6767210763569634}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015531555176710683,"log_rms_mae_db":1.1984782544116228,"rms_envelope_nrmse":0.1077237952795083,"envelope_correlation":0.9959097662622495,"rms_gain_error_db":-0.7952359477604752,"reference_mean_power":0.0017815693662331493,"output_mean_power":0.0014834712762558941},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.067921014076438e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002861407755621019,"output_power":0.00023912368155797853,"reference_active":true,"delta_db":-0.7795716144189686},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5605898871779113e-05,"output_power":2.0413623266137142e-05,"reference_active":true,"delta_db":-0.9841993035586366},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.159270454376387e-07,"output_power":1.0264616020636186e-06,"reference_active":true,"delta_db":0.4948182486828768}],"active_band_count":3,"active_band_mae_db":0.752863055553494}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017559797263300874,"log_rms_mae_db":2.4287559394182203,"rms_envelope_nrmse":0.16668982616697903,"envelope_correlation":0.9861580745148245,"rms_gain_error_db":-1.0068769870134464,"reference_mean_power":0.0012870696988326355,"output_mean_power":0.0010207381980542214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.703155295091611e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025376682215949657,"output_power":0.00020163302591624265,"reference_active":true,"delta_db":-0.9987317353596743},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.494529209431561e-05,"output_power":3.335443715691537e-05,"reference_active":true,"delta_db":-1.2953059070206696},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.800066525463509e-06,"output_power":8.379934799579294e-06,"reference_active":true,"delta_db":-0.21245315681173582}],"active_band_count":3,"active_band_mae_db":0.8354969330640266}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014992083957937532,"log_rms_mae_db":1.9004820558334183,"rms_envelope_nrmse":0.143613946777744,"envelope_correlation":0.9919530557351474,"rms_gain_error_db":-1.0056710056790479,"reference_mean_power":0.0012919551058812144,"output_mean_power":0.0010248972363793895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4067631484646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000244569147219292,"output_power":0.0001941447446834056,"reference_active":true,"delta_db":-1.002760301311976},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8830927140807284e-05,"output_power":1.4136804570333381e-05,"reference_active":true,"delta_db":-1.245204487575032},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.050384653726171e-07,"output_power":6.922749427196026e-07,"reference_active":true,"delta_db":-0.07934199441471419}],"active_band_count":3,"active_band_mae_db":0.7757689277672406}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":3.0903274722723775,"rms_envelope_nrmse":0.38584478703103564,"envelope_correlation":0.9162767961927265,"rms_gain_error_db":1.0499158776529607,"reference_mean_power":4.199153922743894e-05,"output_mean_power":5.347531876190323e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0299126624198237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.222152478948828e-06,"output_power":6.808040343806114e-06,"reference_active":true,"delta_db":2.0748820729279203},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.960031981806198e-07,"output_power":1.067723899384118e-06,"reference_active":true,"delta_db":0.7614940402801487},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.633269352164056e-07,"output_power":4.4666408766543237e-07,"reference_active":true,"delta_db":2.2948575152916457}],"active_band_count":3,"active_band_mae_db":1.7104112094999049}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":2.613781352302201,"rms_envelope_nrmse":0.3349795252667668,"envelope_correlation":0.9367580818217818,"rms_gain_error_db":1.0488491221014389,"reference_mean_power":4.2190507757870256e-05,"output_mean_power":5.3715505154849926e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.548028980657063e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9589316603894534e-06,"output_power":6.308364075494633e-06,"reference_active":true,"delta_db":2.0233874491092103},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7410748918107177e-07,"output_power":4.6960172823718505e-07,"reference_active":true,"delta_db":0.9873328420392748},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.800283314283636e-08,"output_power":2.5112673687848143e-08,"reference_active":true,"delta_db":1.445520971994493}],"active_band_count":3,"active_band_mae_db":1.4854137543809927}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":4.611554941948736,"rms_envelope_nrmse":0.3107130925718728,"envelope_correlation":0.926950233798762,"rms_gain_error_db":-0.7612168745444471,"reference_mean_power":2.209560749738961e-05,"output_mean_power":1.8543181923734898e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.718240736767724e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.504526854650136e-06,"output_power":8.092428189072034e-06,"reference_active":true,"delta_db":-0.6985164787202651},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.566561169291012e-06,"output_power":2.109357814336134e-06,"reference_active":true,"delta_db":-0.8520136315577808},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.915786881504461e-07,"output_power":4.990191913048092e-07,"reference_active":true,"delta_db":-0.7389527191779186}],"active_band_count":3,"active_band_mae_db":0.7631609431519882}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":4.151547460753057,"rms_envelope_nrmse":0.2750120535058642,"envelope_correlation":0.9399389469382231,"rms_gain_error_db":-0.7607957306057724,"reference_mean_power":2.220381322589701e-05,"output_mean_power":1.8635797917811038e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6122764663239423e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.855212413255221e-06,"output_power":8.310845247521606e-06,"reference_active":true,"delta_db":-0.740207936503368},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3142139667203487e-06,"output_power":1.0973096231502856e-06,"reference_active":true,"delta_db":-0.7833689043629197},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.262111729619265e-08,"output_power":4.5695916705346756e-08,"reference_active":true,"delta_db":0.3025256373733506}],"active_band_count":3,"active_band_mae_db":0.6087008260798794}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"golf","status":"measured","mos":{"status":"measured","value":2.9332592487335205,"predicted_mos":2.9332592487335205,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"b161b1857e49e0f6ec7f74187e83a5b51a41ae4d37efcda7695b21c672f33a86","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.897541706000993},"started_at":"2026-10-13T05:02:35.012664+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027644330002658535,"native_input_samples":144000,"native_input_sha256":"c19cd37e07f3dae91cf3926592e8d29ac8413cd470f489f6c4eb631020a9eda2","output_sha256":"9c395e5b333b7ab742551dc8477c0d23a5db9be23f538e9ee85f314c6b0330b2","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5592120960000102,"synthesis_s":0.12725338299969735,"total_s":0.6864654789997076,"rtf":0.11441091316661793,"output_peak":0.38803020119667053,"output_rms":0.03812712813776692,"finished_at":"2026-10-13T05:02:35.708978+00:00","analysis_stages":{"external_f0_s":0.08388760799971351},"evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"9c395e5b333b7ab742551dc8477c0d23a5db9be23f538e9ee85f314c6b0330b2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.374139876162204,"output_peak":0.386027645551565},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.70122760429933,"spectral_convergence":0.3870748843163279,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.432298106048106,"spectral_convergence":0.2794755972310441,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.665975899429517,"spectral_convergence":0.31533533281459836,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.354816191850742e-05,"output_power":3.820051632901591e-05,"delta_db":-0.5690059602350219,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012618831798604149,"output_power":0.0010693768467987508,"delta_db":-0.7188837460222276,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003625272749480512,"output_power":0.00028020480002610465,"delta_db":-1.1186511587070023,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.986355960876659e-05,"output_power":2.424992501674879e-05,"delta_db":-0.9043117246454098,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6518613573778594e-05,"output_power":9.876161706398663e-06,"delta_db":-2.233854014812832,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.119826215640641e-05,"output_power":3.216408745238124e-05,"delta_db":-2.7936785541637574,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3897308597643754,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8674421644209634,"pitch":null,"output_rms_dbfs":-28.374139876162204,"output_peak":0.386027645551565,"reference_rms_dbfs":-27.506697711741243,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.599833869925651},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8638447385992669,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042137145751295554,"rms_dbfs":-27.506697711741243,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.67538488174704,"mean":-2.785907165833029e-05,"zero_energy":false},"output":{"rms":0.038132300446332895,"rms_dbfs":-28.374139876162204,"peak":0.386027645551565,"peak_dbfs":-8.267631840214394,"crest_factor_db":20.106508035947815,"mean":0.00020562951311355867,"zero_energy":false},"rms_difference_db":-0.8674421644209609},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":1.0565691767851286e-06,"output_rms_dbfs":-59.76102063105251,"output_zero_energy":false,"output_peak":0.007862202830040266,"output_energy_fraction":7.884941195120944e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":277,"output_estimated_voiced_frames":300,"both_estimated_voiced_frames":268,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9675090252707581,"absolute_difference_cents_median":5.564745695793103,"absolute_difference_cents_p95":23.30279746025238,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00018136510927426692,"log_rms_mae_db":2.569885546023161,"rms_envelope_nrmse":0.17783418855380584,"envelope_correlation":0.9815956114310124,"rms_gain_error_db":-0.867252991033754,"reference_mean_power":0.0017767851238598797,"output_mean_power":0.0014551561875077324},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.56924223918074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003062529533919575,"output_power":0.0002354842090913734,"reference_active":true,"delta_db":-1.141184956955574},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.4347065172042005e-05,"output_power":4.49376335070935e-05,"reference_active":true,"delta_db":-0.8256589847026191},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1390555497530403e-05,"output_power":1.0883789236918394e-05,"reference_active":true,"delta_db":-0.1976478124394991}],"active_band_count":3,"active_band_mae_db":0.7214972513658974}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015531555176710683,"log_rms_mae_db":2.3742510936644816,"rms_envelope_nrmse":0.15663201079949654,"envelope_correlation":0.9880784735360463,"rms_gain_error_db":-0.8674554249966389,"reference_mean_power":0.001782366337991794,"output_mean_power":0.0014596590656202038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.057688448486708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028571108417895386,"output_power":0.00022283112137625862,"reference_active":true,"delta_db":-1.0795124336162207},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.484813466967698e-05,"output_power":2.0764289544249955e-05,"reference_active":true,"delta_db":-0.7797671594325952},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.143471843563824e-07,"output_power":9.659455290479785e-07,"reference_active":true,"delta_db":0.23841504801874583}],"active_band_count":3,"active_band_mae_db":0.6992315470225207}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017559797263300874,"log_rms_mae_db":2.6390432870390175,"rms_envelope_nrmse":0.18255180100651922,"envelope_correlation":0.9804679347647851,"rms_gain_error_db":-0.7884587748085801,"reference_mean_power":0.0012888055062323038,"output_mean_power":0.0010748342709229947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.694230365481444e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025375932177173447,"output_power":0.00019349617558959045,"reference_active":true,"delta_db":-1.177496191879296},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3727296335471396e-05,"output_power":3.537984417159316e-05,"reference_active":true,"delta_db":-0.9199670993573055},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.743181546305603e-06,"output_power":8.133594334553215e-06,"reference_active":true,"delta_db":-0.31386988274180383}],"active_band_count":3,"active_band_mae_db":0.8037777246594683}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014992083957937532,"log_rms_mae_db":2.1728359742838674,"rms_envelope_nrmse":0.15819250687206465,"envelope_correlation":0.9868054839769985,"rms_gain_error_db":-0.7881455452844304,"reference_mean_power":0.0012922942105162006,"output_mean_power":0.00107782150412357},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.395418088375822e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024412813537535492,"output_power":0.00018838273511520958,"reference_active":true,"delta_db":-1.1257673594572855},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.804652845157765e-05,"output_power":1.5135380804866144e-05,"reference_active":true,"delta_db":-0.7640031830338889},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.016661076914051e-07,"output_power":7.45573050199091e-07,"reference_active":true,"delta_db":0.26359701960563287}],"active_band_count":3,"active_band_mae_db":0.7177891873656024}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":2.13721041720152,"rms_envelope_nrmse":0.217542431643944,"envelope_correlation":0.9801737613508942,"rms_gain_error_db":-1.491918772554449,"reference_mean_power":4.20585885858816e-05,"output_mean_power":2.9830656907426554e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.025852037875375e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.208076515133429e-06,"output_power":3.309631501348863e-06,"reference_active":true,"delta_db":-1.043039864025702},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.786908583651934e-07,"output_power":6.144071644359811e-07,"reference_active":true,"delta_db":-1.5537983691226995},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.622630433041457e-07,"output_power":1.9419749243682822e-07,"reference_active":true,"delta_db":-1.3049347865456575}],"active_band_count":3,"active_band_mae_db":1.3005910065646864}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":1.7381743463215513,"rms_envelope_nrmse":0.1844355203773014,"envelope_correlation":0.9915422598182771,"rms_gain_error_db":-1.4914253443799113,"reference_mean_power":4.224514089976224e-05,"output_mean_power":2.9966376277943378e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.504760112827997e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.942373960402998e-06,"output_power":3.0862390035221987e-06,"reference_active":true,"delta_db":-1.0632826209662474},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.61687711196872e-07,"output_power":2.4533193957166744e-07,"reference_active":true,"delta_db":-1.6857966095798806},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.794575279951317e-08,"output_power":1.2627522335407693e-08,"reference_active":true,"delta_db":-1.5264353574446432}],"active_band_count":3,"active_band_mae_db":1.425171529330257}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":3.319401043271924,"rms_envelope_nrmse":0.27128728018580944,"envelope_correlation":0.9516829685016988,"rms_gain_error_db":-0.23675299458393567,"reference_mean_power":2.2132298785109942e-05,"output_mean_power":2.095806703274945e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.718658788059679e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.494904440240603e-06,"output_power":1.3394872787583607e-05,"reference_active":true,"delta_db":1.494479952103947},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5623893421266805e-06,"output_power":3.4299917809443803e-06,"reference_active":true,"delta_db":1.2664796009683603},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.906354798916517e-07,"output_power":9.94410380838098e-07,"reference_active":true,"delta_db":2.2624611754968957}],"active_band_count":3,"active_band_mae_db":1.6744735761897342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":3.060213665247802,"rms_envelope_nrmse":0.2258796038322955,"envelope_correlation":0.9670961273503627,"rms_gain_error_db":-0.23625588334020173,"reference_mean_power":2.2237457631732985e-05,"output_mean_power":2.1060057146619873e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.612705178396602e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.839697464624309e-06,"output_power":1.4199901135256254e-05,"reference_active":true,"delta_db":1.593035750412233},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3114154255304798e-06,"output_power":1.792604742354324e-06,"reference_active":true,"delta_db":1.3574425326751407},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2515908565206635e-08,"output_power":7.283128173994383e-08,"reference_active":true,"delta_db":2.3376648867679646}],"active_band_count":3,"active_band_mae_db":1.7627143899517794}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.6113698482513428,"predicted_mos":2.6113698482513428,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"0f0ab63da05dbda3ea7dcf79dd7b8ed25665051cd14a4cedd6b706457ff7e026","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.88617584199892},"started_at":"2026-10-13T07:37:51.212665+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032951169996522367,"native_input_samples":144000,"native_input_sha256":"c19cd37e07f3dae91cf3926592e8d29ac8413cd470f489f6c4eb631020a9eda2","output_sha256":"85f3b42cb38e3c19115c5e841cd07b34097bcc4063aaeebf81fbd0472d21e62d","output_samples":144120,"length_delta_samples":120,"analysis_s":2.6114415349984483,"synthesis_s":3.729773228002159,"total_s":6.3412147630006075,"rtf":1.0568691271667678,"output_peak":0.346047043800354,"output_rms":0.0400218690112284,"finished_at":"2026-10-13T07:37:57.567952+00:00","analysis_stages":{"external_f0_s":1.4079691210026795},"evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"85f3b42cb38e3c19115c5e841cd07b34097bcc4063aaeebf81fbd0472d21e62d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.95915964151805,"output_peak":0.3359624688941246},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.762813128649981,"spectral_convergence":0.4170841988369817,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.356993077723223,"spectral_convergence":0.27916123023999234,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.582100963100223,"spectral_convergence":0.2562642713015536,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":3.758466731811355e-05,"delta_db":-0.6352855200139034,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.0011647469165334983,"delta_db":-0.3444242851217828,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.00030886083649627095,"delta_db":-0.6884597888592774,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":2.9407308907899828e-05,"delta_db":-0.05894266704723164,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":1.6310305264458068e-05,"delta_db":-0.04707381835406674,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":4.385790884755393e-05,"delta_db":-1.4394861584776306,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5356120396456489,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.44550194874773497,"pitch":null,"output_rms_dbfs":-27.95671621372016,"output_peak":0.3359624688941246,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.567302389824476},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8922384213997705,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042115240661813005,"rms_dbfs":-27.511214264972427,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.679901434978223,"mean":-3.424930572509765e-05,"zero_energy":false},"output":{"rms":0.04000959814054366,"rms_dbfs":-27.95671621372016,"peak":0.3359624688941246,"peak_dbfs":-9.474184717832696,"crest_factor_db":18.482531495887464,"mean":5.600849302068377e-05,"zero_energy":false},"rms_difference_db":-0.44550194874773297},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":4.295338970418367e-06,"output_rms_dbfs":-53.670025577527966,"output_zero_energy":false,"output_peak":0.04064701838835417,"output_energy_fraction":0.0002906907215314439,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":278,"output_estimated_voiced_frames":287,"both_estimated_voiced_frames":263,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9460431654676259,"absolute_difference_cents_median":4.993577404447876,"absolute_difference_cents_p95":24.923529746019224,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018136510927426692,"log_rms_mae_db":1.8357395968485999,"rms_envelope_nrmse":0.16655148026433972,"envelope_correlation":0.9752058620718825,"rms_gain_error_db":-0.44526467693383004,"reference_mean_power":0.0017749427149271913,"output_mean_power":0.0016019828902916159},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.578419076835135e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003061888162786739,"output_power":0.00028730110539573755,"reference_active":true,"delta_db":-0.2765202684769499},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.561280729204581e-05,"output_power":5.024350629030864e-05,"reference_active":true,"delta_db":-0.4409487904874877},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1447257663776475e-05,"output_power":1.090463525457384e-05,"reference_active":true,"delta_db":-0.21090314777509142}],"active_band_count":3,"active_band_mae_db":0.3094574022465097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015531555176710683,"log_rms_mae_db":1.3522334345735556,"rms_envelope_nrmse":0.10882918403158966,"envelope_correlation":0.9902385414567727,"rms_gain_error_db":-0.4436391832726089,"reference_mean_power":0.0017815693662331493,"output_mean_power":0.0016085657529369122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.067921014076438e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002861407755621019,"output_power":0.00026967536218893513,"reference_active":true,"delta_db":-0.2573847925356631},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5605898871779113e-05,"output_power":2.3058423160842336e-05,"reference_active":true,"delta_db":-0.4551042099537952},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.159270454376387e-07,"output_power":8.843138697310422e-07,"reference_active":true,"delta_db":-0.15254446410487194}],"active_band_count":3,"active_band_mae_db":0.28834448886477676}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017559797263300874,"log_rms_mae_db":2.2598477750413832,"rms_envelope_nrmse":0.18003720511973323,"envelope_correlation":0.9752675410244588,"rms_gain_error_db":-0.4023740756290759,"reference_mean_power":0.0012870696988326355,"output_mean_power":0.0011731800208539074},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.703155295091611e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025376682215949657,"output_power":0.00023858896966564964,"reference_active":true,"delta_db":-0.26784479485098556},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.494529209431561e-05,"output_power":4.2983973134924725e-05,"reference_active":true,"delta_db":-0.19377651077198116},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.800066525463509e-06,"output_power":9.018508967124708e-06,"reference_active":true,"delta_db":0.10648786157012272}],"active_band_count":3,"active_band_mae_db":0.18936972239769648}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014992083957937532,"log_rms_mae_db":1.6429794437552845,"rms_envelope_nrmse":0.11647826869195974,"envelope_correlation":0.9900516472533037,"rms_gain_error_db":-0.40027780374965344,"reference_mean_power":0.0012919551058812144,"output_mean_power":0.0011782016924669586},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4067631484646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000244569147219292,"output_power":0.00023099681679842167,"reference_active":true,"delta_db":-0.24795673994697992},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8830927140807284e-05,"output_power":1.840254600252155e-05,"reference_active":true,"delta_db":-0.0999379098387339},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.050384653726171e-07,"output_power":7.408061581787798e-07,"reference_active":true,"delta_db":0.21491772090510974}],"active_band_count":3,"active_band_mae_db":0.18760412356360787}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":1.7678417325039109,"rms_envelope_nrmse":0.19343228424310177,"envelope_correlation":0.9719030822658937,"rms_gain_error_db":-0.05674533533100663,"reference_mean_power":4.199153922743894e-05,"output_mean_power":4.144644262849132e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0299126624198237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.222152478948828e-06,"output_power":4.369379293625151e-06,"reference_active":true,"delta_db":0.14885832858614656},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.960031981806198e-07,"output_power":8.978615998057494e-07,"reference_active":true,"delta_db":0.008998380040646622},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.633269352164056e-07,"output_power":2.3865321661461944e-07,"reference_active":true,"delta_db":-0.42727992295581907}],"active_band_count":3,"active_band_mae_db":0.19504554386087078}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":1.045479139173002,"rms_envelope_nrmse":0.11052938789736955,"envelope_correlation":0.9904133328354217,"rms_gain_error_db":-0.05632871270881339,"reference_mean_power":4.2190507757870256e-05,"output_mean_power":4.1646823353230464e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.548028980657063e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9589316603894534e-06,"output_power":4.116891683546726e-06,"reference_active":true,"delta_db":0.16991435783821635},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7410748918107177e-07,"output_power":3.7249302731974333e-07,"reference_active":true,"delta_db":-0.01878254715870899},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.800283314283636e-08,"output_power":1.4979383283351544e-08,"reference_active":true,"delta_db":-0.7984692291647123}],"active_band_count":3,"active_band_mae_db":0.32905537805387924}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":1.7628458133488514,"rms_envelope_nrmse":0.249491350987826,"envelope_correlation":0.9505629453560097,"rms_gain_error_db":-0.780964876873088,"reference_mean_power":2.209560749738961e-05,"output_mean_power":1.845905479016784e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.718240736767724e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.504526854650136e-06,"output_power":7.779274871149795e-06,"reference_active":true,"delta_db":-0.8699138507945272},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.566561169291012e-06,"output_power":2.108082813673603e-06,"reference_active":true,"delta_db":-0.8546395168653391},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.915786881504461e-07,"output_power":4.4138811779931853e-07,"reference_active":true,"delta_db":-1.2719188223756244}],"active_band_count":3,"active_band_mae_db":0.9988240633451636}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":1.091339362057688,"rms_envelope_nrmse":0.16305291023400165,"envelope_correlation":0.9813564242760886,"rms_gain_error_db":-0.7807683525512954,"reference_mean_power":2.220381322589701e-05,"output_mean_power":1.8550291145280912e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6122764663239423e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.855212413255221e-06,"output_power":8.018225221633922e-06,"reference_active":true,"delta_db":-0.8958773837623604},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3142139667203487e-06,"output_power":1.0920683088637827e-06,"reference_active":true,"delta_db":-0.8041627397972342},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.262111729619265e-08,"output_power":2.9413856882906177e-08,"reference_active":true,"delta_db":-1.6107285541930503}],"active_band_count":3,"active_band_mae_db":1.1035895592508815}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.35664701461792,"predicted_mos":2.35664701461792,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"5e66cb17098b9de146b6c879edba80c299415ea91b68bf5228f6145d894c9491","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.276225515999613},"started_at":"2026-10-13T05:10:44.968038+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031620249992556637,"native_input_samples":144000,"native_input_sha256":"c19cd37e07f3dae91cf3926592e8d29ac8413cd470f489f6c4eb631020a9eda2","output_sha256":"79f6c5b3fbadc7e55bf15021873e5d86aaf604962c36c090adf673bed0b935aa","output_samples":144240,"length_delta_samples":240,"analysis_s":0.6462643750001007,"synthesis_s":0.1726314999996248,"total_s":0.8188958749997255,"rtf":0.1364826458332876,"output_peak":0.34789854288101196,"output_rms":0.041363150477045627,"finished_at":"2026-10-13T05:10:45.803929+00:00","analysis_stages":{"external_f0_s":0.14587931400092202},"evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"79f6c5b3fbadc7e55bf15021873e5d86aaf604962c36c090adf673bed0b935aa","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.671208972895162,"output_peak":0.3487935633650378},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.577327718784435,"spectral_convergence":0.15632973125146946,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.539214502572498,"spectral_convergence":0.14748944573367143,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.2346134833422635,"spectral_convergence":0.1695394489786858,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.350500895926799e-05,"output_power":4.647723186448243e-05,"delta_db":0.28700991596555026,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012608807022653607,"output_power":0.0012281497866963313,"delta_db":-0.11422660776335969,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00036191686574782775,"output_power":0.0003483183047190692,"delta_db":-0.16632524237604293,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9809147109830135e-05,"output_power":2.7409689117546705e-05,"delta_db":-0.3644544053839885,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.6488056661262093e-05,"output_power":1.5840686618290342e-05,"delta_db":-0.17395468970151456,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.109371525887873e-05,"output_power":4.616457778871045e-05,"delta_db":-1.2168766849823929,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38714125769547475,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1528342895198877,"pitch":null,"output_rms_dbfs":-27.664048554492314,"output_peak":0.3487935633650378,"reference_rms_dbfs":-27.511214264972427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.783718568233065},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.948551139605201,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042115240661813005,"rms_dbfs":-27.511214264972427,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.679901434978223,"mean":-3.424930572509765e-05,"zero_energy":false},"output":{"rms":0.04138067516129275,"rms_dbfs":-27.664048554492314,"peak":0.3487935633650378,"peak_dbfs":-9.148630762681556,"crest_factor_db":18.51541779181076,"mean":-7.178514661208661e-06,"zero_energy":false},"rms_difference_db":-0.15283428951988753},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":1.6059403689350389e-06,"output_rms_dbfs":-57.94270584787888,"output_zero_energy":false,"output_peak":0.01964306738524725,"output_energy_fraction":0.0001016006243884422,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":278,"output_estimated_voiced_frames":278,"both_estimated_voiced_frames":267,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.960431654676259,"absolute_difference_cents_median":0.6988122955890494,"absolute_difference_cents_p95":8.073102568749162,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018136510927426692,"log_rms_mae_db":0.7151099296993013,"rms_envelope_nrmse":0.05527794866277685,"envelope_correlation":0.9973831742777642,"rms_gain_error_db":-0.15276387742778985,"reference_mean_power":0.0017749427149271913,"output_mean_power":0.0017135940807619494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.578419076835135e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003061888162786739,"output_power":0.0003029208213634303,"reference_active":true,"delta_db":-0.04660198063515976},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.561280729204581e-05,"output_power":4.931489536028464e-05,"reference_active":true,"delta_db":-0.521967025391633},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1447257663776475e-05,"output_power":8.44645656359877e-06,"reference_active":true,"delta_db":-1.320269053497032}],"active_band_count":3,"active_band_mae_db":0.6296126865079416}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015531555176710683,"log_rms_mae_db":0.5156666726985164,"rms_envelope_nrmse":0.03940480043043141,"envelope_correlation":0.9987886225836927,"rms_gain_error_db":-0.15223303597604879,"reference_mean_power":0.0017815693662331493,"output_mean_power":0.0017202019390489967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.067921014076438e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002861407755621019,"output_power":0.00028300098129722555,"reference_active":true,"delta_db":-0.04791808510493427},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5605898871779113e-05,"output_power":2.2564907141699776e-05,"reference_active":true,"delta_db":-0.5490647528390438},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.159270454376387e-07,"output_power":6.597724313073756e-07,"reference_active":true,"delta_db":-1.4246671850183117}],"active_band_count":3,"active_band_mae_db":0.6738833409874299}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017559797263300874,"log_rms_mae_db":0.9270193516319988,"rms_envelope_nrmse":0.05461597714550041,"envelope_correlation":0.9978344126012682,"rms_gain_error_db":-0.14328785332406313,"reference_mean_power":0.0012870696988326355,"output_mean_power":0.0012452979713365238},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.703155295091611e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025376682215949657,"output_power":0.0002469831214091521,"reference_active":true,"delta_db":-0.11767566179646194},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.494529209431561e-05,"output_power":3.995432544865532e-05,"reference_active":true,"delta_db":-0.5112040435906049},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.800066525463509e-06,"output_power":6.2989928632114146e-06,"reference_active":true,"delta_db":-1.4521483898640442}],"active_band_count":3,"active_band_mae_db":0.6936760317503703}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014992083957937532,"log_rms_mae_db":0.5023431526980116,"rms_envelope_nrmse":0.03564935016560241,"envelope_correlation":0.9991482307838626,"rms_gain_error_db":-0.1427991497725329,"reference_mean_power":0.0012919551058812144,"output_mean_power":0.0012501654938501675},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4067631484646e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000244569147219292,"output_power":0.000236780637807255,"reference_active":true,"delta_db":-0.140554831578951},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8830927140807284e-05,"output_power":1.6645721561394897e-05,"reference_active":true,"delta_db":-0.5356907726716148},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.050384653726171e-07,"output_power":5.035022739306379e-07,"reference_active":true,"delta_db":-1.462113755380824}],"active_band_count":3,"active_band_mae_db":0.7127864532104633}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":0.8552199673122778,"rms_envelope_nrmse":0.08860305625312634,"envelope_correlation":0.9948289073449692,"rms_gain_error_db":-0.29916598231605124,"reference_mean_power":4.199153922743894e-05,"output_mean_power":3.919631110497057e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0299126624198237e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.222152478948828e-06,"output_power":3.95203864111897e-06,"reference_active":true,"delta_db":-0.2871273112220935},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.960031981806198e-07,"output_power":8.455957528276875e-07,"reference_active":true,"delta_db":-0.25146766854508107},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.633269352164056e-07,"output_power":1.8731827887781578e-07,"reference_active":true,"delta_db":-1.479151258093845}],"active_band_count":3,"active_band_mae_db":0.6725820792870065}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":0.5193977511149234,"rms_envelope_nrmse":0.053851077402041675,"envelope_correlation":0.9985817167076382,"rms_gain_error_db":-0.2991374853355029,"reference_mean_power":4.2190507757870256e-05,"output_mean_power":3.938229341750501e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.548028980657063e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9589316603894534e-06,"output_power":3.6899067413863663e-06,"reference_active":true,"delta_db":-0.3056261501763161},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7410748918107177e-07,"output_power":3.5211194465788356e-07,"reference_active":true,"delta_db":-0.2631564448950226},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.800283314283636e-08,"output_power":1.2758050031626113e-08,"reference_active":true,"delta_db":-1.4955655536120518}],"active_band_count":3,"active_band_mae_db":0.6881160495611303}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":0.8853598306647638,"rms_envelope_nrmse":0.14039069924088826,"envelope_correlation":0.9847701334324609,"rms_gain_error_db":-0.38784477049995336,"reference_mean_power":2.209560749738961e-05,"output_mean_power":2.0207913281432198e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.718240736767724e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.504526854650136e-06,"output_power":8.505657745165097e-06,"reference_active":true,"delta_db":-0.4822259872831742},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.566561169291012e-06,"output_power":2.3910807816017067e-06,"reference_active":true,"delta_db":-0.3075737052502258},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.915786881504461e-07,"output_power":3.689893983338188e-07,"reference_active":true,"delta_db":-2.0499863163032503}],"active_band_count":3,"active_band_mae_db":0.9465953362788834}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":0.5451634841444066,"rms_envelope_nrmse":0.07229200369829182,"envelope_correlation":0.9969129343369055,"rms_gain_error_db":-0.3878284856476663,"reference_mean_power":2.220381322589701e-05,"output_mean_power":2.0306950814406302e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6122764663239423e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.855212413255221e-06,"output_power":8.774412745507251e-06,"reference_active":true,"delta_db":-0.50447929615609},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3142139667203487e-06,"output_power":1.2063145544367576e-06,"reference_active":true,"delta_db":-0.3720551060664594},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.262111729619265e-08,"output_power":2.580113205949568e-08,"reference_active":true,"delta_db":-2.1798606875228366}],"active_band_count":3,"active_band_mae_db":1.018798363248462}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0032","method":"magphase","status":"measured","mos":{"status":"measured","value":2.59660005569458,"predicted_mos":2.59660005569458,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":286872,"input_sample_rate":48000,"input_seconds":5.9765,"input_float32_sha256":"6649e5620209400044104f260a3734cb790b5fc37973c4799703eceabd93755a","internal_samples":95624,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.043595742999969},"started_at":"2026-10-13T05:15:36.708051+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.011956106000070577,"native_input_samples":288000,"native_input_sha256":"00ef40527bb19dee2dfee33e8d71187cb30aa4052829b8602343702d1cd65b19","output_sha256":"e7f7d89044ca0d63241c29357ab8b067a613926f7cef50ba667f026b45910844","output_samples":286872,"length_delta_samples":-1128,"analysis_s":3.8181285860009666,"synthesis_s":1.1191981179999857,"total_s":4.937326704000952,"rtf":0.8228877840001587,"output_peak":0.40784122222222724,"output_rms":0.040512180283755306,"finished_at":"2026-10-13T05:15:41.671932+00:00","evaluation_input":{"reference":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"e7f7d89044ca0d63241c29357ab8b067a613926f7cef50ba667f026b45910844","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95624,"length_delta_samples":-376,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.85434243919496,"output_peak":0.4020657752145294},"common_prefix":{"samples":95624,"sample_rate":16000,"reference_coverage":0.9960833333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95624,"output_samples":95624,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.80907576571287,"spectral_convergence":0.36728179309442555,"spectral_convergence_status":"ok","amplitude_floor":1.3235749190080935e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.305799338181788,"spectral_convergence":0.16072829592193813,"spectral_convergence_status":"ok","amplitude_floor":1.1987597823804542e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.825847382732992,"spectral_convergence":0.1810176563482714,"spectral_convergence_status":"ok","amplitude_floor":9.377214849682438e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.379928047116202e-05,"output_power":4.2334638383670164e-05,"delta_db":-0.14771121586728242,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012628725917770598,"output_power":0.0011429710551421903,"delta_db":-0.433243054615772,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003630773928737578,"output_power":0.00034945770418469456,"delta_db":-0.16604588852969676,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.9941820050848648e-05,"output_power":3.074501623307038e-05,"delta_db":0.11496530743234856,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.656646418437541e-05,"output_power":1.757304784618452e-05,"delta_db":0.2561726557246175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.134531949557712e-05,"output_power":5.586864060094908e-05,"delta_db":-0.40613328463831067,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25404523446800464,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3526912395294952,"pitch":null,"output_rms_dbfs":-27.85434243919496,"output_peak":0.4020657752145294,"reference_rms_dbfs":-27.50165119966546,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.64690749554255},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95624,"output_samples":95624,"intelligibility":{"estoi":{"value":0.9286677625973134,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.042161634595775224,"rms_dbfs":-27.50165119966546,"peak":0.405914306640625,"peak_dbfs":-7.8313128299942045,"crest_factor_db":19.670338369671256,"mean":-2.841507452038714e-05,"zero_energy":false},"output":{"rms":0.040483949835229675,"rms_dbfs":-27.85434243919496,"peak":0.4020657752145294,"peak_dbfs":-7.914057869852749,"crest_factor_db":19.940284569342207,"mean":-0.0005924332089782781,"zero_energy":false},"rms_difference_db":-0.35269123952949855},"full_scale":{"reference":{"sample_count":95624,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95624,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":597,"uncovered_tail_samples":104,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.14341742603574e-06,"threshold_dbfs":-55.02597943581026,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":1.3968061942320603e-06,"output_mean_power":1.2278477999582268e-06,"output_rms_dbfs":-59.10855463597837,"output_zero_energy":false,"output_peak":0.009483642636598077,"output_energy_fraction":8.147891049083975e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":590,"reference_estimated_voiced_frames":275,"output_estimated_voiced_frames":258,"both_estimated_voiced_frames":247,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8981818181818182,"absolute_difference_cents_median":1.9598705889562245,"absolute_difference_cents_p95":9.675235370263437,"difference_over_600_cents_frames":1,"uncovered_tail_samples":104,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95624,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1194,"rms_floor":0.00018136510927426692,"log_rms_mae_db":1.074321181409259,"rms_envelope_nrmse":0.07883078903226329,"envelope_correlation":0.9955271274414536,"rms_gain_error_db":-0.35255031012231286,"reference_mean_power":0.0017783619419178634,"output_mean_power":0.001639702748339927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":5.554413853092248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003062642504754508,"output_power":0.0002874886581025034,"reference_active":true,"delta_db":-0.27475589828405433},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.244417778444042e-05,"output_power":4.871402312239237e-05,"reference_active":true,"delta_db":-0.3204328303780425},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1349776696072239e-05,"output_power":1.0786095839214167e-05,"reference_active":true,"delta_db":-0.2212304211421158}],"active_band_count":3,"active_band_mae_db":0.2721397166014042}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1188,"rms_floor":0.00015531555176710683,"log_rms_mae_db":0.7463796137095399,"rms_envelope_nrmse":0.05440791554075135,"envelope_correlation":0.998599156720628,"rms_gain_error_db":-0.3521332950972472,"reference_mean_power":0.001782254393522938,"output_mean_power":0.0016434495033210233},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":5.041292540117488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002849118305071985,"output_power":0.00026649794422246945,"reference_active":true,"delta_db":-0.29016619788617026},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38306184968876e-05,"output_power":2.2235817948483674e-05,"reference_active":true,"delta_db":-0.30082204439434895},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.189086250451643e-07,"output_power":9.372501280004469e-07,"reference_active":true,"delta_db":0.08583180492125057}],"active_band_count":3,"active_band_mae_db":0.22560668240058993}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1194,"rms_floor":0.00017559797263300874,"log_rms_mae_db":1.5864583816955369,"rms_envelope_nrmse":0.07262958421728466,"envelope_correlation":0.9971344403852948,"rms_gain_error_db":-0.3705244565420128,"reference_mean_power":0.001289570081085261,"output_mean_power":0.001184111238314439},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":4.677893993758691e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000253737799606624,"output_power":0.00023532490947471366,"reference_active":true,"delta_db":-0.3271726908599297},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.185161768523488e-05,"output_power":3.6784582016676904e-05,"reference_active":true,"delta_db":-0.5604642386590577},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.704029693493927e-06,"output_power":7.700792650725805e-06,"reference_active":true,"delta_db":-0.5318493407385946}],"active_band_count":3,"active_band_mae_db":0.4731620900858607}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1188,"rms_floor":0.00014992083957937532,"log_rms_mae_db":1.103315985232486,"rms_envelope_nrmse":0.055146759080793605,"envelope_correlation":0.9988874853912968,"rms_gain_error_db":-0.37012231824595376,"reference_mean_power":0.0012915179541054078,"output_mean_power":0.0011860096319841582},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":4.376981227705439e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002433707143198982,"output_power":0.00022512086153314396,"reference_active":true,"delta_db":-0.33852574678747865},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6937588675304816e-05,"output_power":1.4707250427871425e-05,"reference_active":true,"delta_db":-0.6132009449892735},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.10422224499617e-07,"output_power":6.436788076808717e-07,"reference_active":true,"delta_db":-0.428473284760452}],"active_band_count":3,"active_band_mae_db":0.4600666588457347}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1194,"rms_floor":4.2493337902299445e-05,"log_rms_mae_db":1.0767541496777948,"rms_envelope_nrmse":0.08908161967266039,"envelope_correlation":0.9944964291805652,"rms_gain_error_db":0.14904695253219785,"reference_mean_power":4.2149430056122216e-05,"output_mean_power":4.362107877469407e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":1.0194743649301649e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.188568510342477e-06,"output_power":4.461258381141812e-06,"reference_active":true,"delta_db":0.27391753475240355},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.504191602778564e-07,"output_power":9.014201192218864e-07,"reference_active":true,"delta_db":0.252942106162313},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6177325487269666e-07,"output_power":2.9734503909512757e-07,"reference_active":true,"delta_db":0.5533542410382152}],"active_band_count":3,"active_band_mae_db":0.36007129398431054}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1188,"rms_floor":3.820326553727612e-05,"log_rms_mae_db":0.6464221155057945,"rms_envelope_nrmse":0.04807256778471432,"envelope_correlation":0.9984990418171361,"rms_gain_error_db":0.14941964630257726,"reference_mean_power":4.231019674315293e-05,"output_mean_power":4.3791216457755746e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":9.436675692117976e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.917754105978333e-06,"output_power":4.156625011815096e-06,"reference_active":true,"delta_db":0.2570367213474973},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.430603672590741e-07,"output_power":3.6283022628198e-07,"reference_active":true,"delta_db":0.24332911438735855},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.799576700965379e-08,"output_power":2.303252097823051e-08,"reference_active":true,"delta_db":1.0717111365456438}],"active_band_count":3,"active_band_mae_db":0.5240256574268333}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1194,"rms_floor":3.219318962627015e-05,"log_rms_mae_db":1.3559552055460409,"rms_envelope_nrmse":0.236681781873102,"envelope_correlation":0.958256221847607,"rms_gain_error_db":0.4348822048294146,"reference_mean_power":2.2184098648478868e-05,"output_mean_power":2.4520538765992458e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":1.7192105425908662e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.479296614267929e-06,"output_power":8.675563739130581e-06,"reference_active":true,"delta_db":-0.3847840799337112},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5566286662704926e-06,"output_power":2.91294037327899e-06,"reference_active":true,"delta_db":0.5666394073437702},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.878506168742858e-07,"output_power":7.13557491831348e-07,"reference_active":true,"delta_db":0.841619920029617}],"active_band_count":3,"active_band_mae_db":0.5976811357690328}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":24,"metrics":{"frame_count":1188,"rms_floor":2.1857789632782735e-05,"log_rms_mae_db":1.0302697444528348,"rms_envelope_nrmse":0.1454488396812498,"envelope_correlation":0.9843882579974067,"rms_gain_error_db":0.4348021112438658,"reference_mean_power":2.22832624911078e-05,"output_mean_power":2.4629692362387096e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":1.6132837645867495e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.815927384682043e-06,"output_power":9.089813173137735e-06,"reference_active":true,"delta_db":-0.33376379877786533},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3068788698124376e-06,"output_power":1.4914649279687293e-06,"reference_active":true,"delta_db":0.5737770915792316},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2781739000300166e-08,"output_power":5.8576163328437744e-08,"reference_active":true,"delta_db":1.3646248860693815}],"active_band_count":3,"active_band_mae_db":0.7573885921421595}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"hnm","status":"measured","mos":{"status":"measured","value":2.085975408554077,"predicted_mos":2.085975408554077,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78240,"input_sample_rate":16000,"input_seconds":4.89,"input_float32_sha256":"4aba4be5489bbb375f00a1cd0462d73f01ffa81493749a0b34a542f14da3934c","internal_samples":78240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.6212403690005885},"started_at":"2026-10-13T04:39:19.381947+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4238999938243069e-05,"native_input_samples":78240,"native_input_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","output_sha256":"41127b0970a33f6433696fb5cb7034e136c1cfc6bd3b8b0a2be23d08a644aec6","output_samples":78240,"length_delta_samples":0,"analysis_s":0.7606063310004174,"synthesis_s":0.4038470289997349,"total_s":1.1644533600001523,"rtf":0.2381295214724238,"output_peak":0.49753972714030464,"output_rms":0.05060828174041457,"finished_at":"2026-10-13T04:39:20.550972+00:00","evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"41127b0970a33f6433696fb5cb7034e136c1cfc6bd3b8b0a2be23d08a644aec6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.7216977977784,"spectral_convergence":0.2730878540419999,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.696264197931016,"spectral_convergence":0.2983666839760422,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.044304606033223,"spectral_convergence":0.33662372173530875,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":5.9015237695647334e-05,"delta_db":0.8295688331447941,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0008550744225374163,"delta_db":0.0524590567980699,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.0014510245199957843,"delta_db":-0.3035624178304598,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":6.47930303905178e-05,"delta_db":0.5118565015161461,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":8.046228891839689e-05,"delta_db":0.16448900760596097,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":5.082868117941529e-05,"delta_db":-0.1599650031242993,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3369834700032883,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12738627541031877,"pitch":null,"output_rms_dbfs":-25.915568153379414,"output_peak":0.49753972714030464,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.8207555339142125},"common_prefix":{"samples":78240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.7216977977784,"spectral_convergence":0.2730878540419999,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.696264197931016,"spectral_convergence":0.2983666839760422,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.044304606033223,"spectral_convergence":0.33662372173530875,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":5.9015237695647334e-05,"delta_db":0.8295688331447941,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0008550744225374163,"delta_db":0.0524590567980699,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.0014510245199957843,"delta_db":-0.3035624178304598,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":6.47930303905178e-05,"delta_db":0.5118565015161461,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":8.046228891839689e-05,"delta_db":0.16448900760596097,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":5.082868117941529e-05,"delta_db":-0.1599650031242993,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3369834700032883,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12738627541031877,"pitch":null,"output_rms_dbfs":-25.915568153379414,"output_peak":0.49753972714030464,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.8207555339142125},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78240,"output_samples":78240,"intelligibility":{"estoi":{"value":0.8603543860235295,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051355966406601,"rms_dbfs":-25.788181877969095,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.607283545458053,"mean":-3.586322495786203e-05,"zero_energy":false},"output":{"rms":0.05060828174041457,"rms_dbfs":-25.915568153379414,"peak":0.49753972714030464,"peak_dbfs":-6.06344472694863,"crest_factor_db":19.852123426430783,"mean":-4.214923148080279e-05,"zero_energy":false},"rms_difference_db":-0.12738627541031988},"full_scale":{"reference":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":489,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":7.014925843880822e-07,"output_mean_power":1.0371560953214705e-06,"output_rms_dbfs":-59.841558759782664,"output_zero_energy":false,"output_peak":0.02054543688775676,"output_energy_fraction":9.937412697070314e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":482,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":154,"both_estimated_voiced_frames":152,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9047619047619048,"absolute_difference_cents_median":1.5006348928899897,"absolute_difference_cents_p95":12.356177756739566,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.0002230061513489049,"log_rms_mae_db":0.9303341774932202,"rms_envelope_nrmse":0.09077767195416755,"envelope_correlation":0.99362040874451,"rms_gain_error_db":-0.12738644940584012,"reference_mean_power":0.002640134243154231,"output_mean_power":0.0025638190201439357},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.2388906472962606e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011616255671056531,"output_power":0.0010993132751451185,"reference_active":true,"delta_db":-0.23944689703345995},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014996209998372379,"output_power":0.0001388194882972329,"reference_active":true,"delta_db":-0.3353107418872359},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.935590189308994e-05,"output_power":3.537133782864423e-05,"reference_active":true,"delta_db":-0.46358383102743156}],"active_band_count":3,"active_band_mae_db":0.3461138233160425}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.0001943907377270206,"log_rms_mae_db":0.5417391932483495,"rms_envelope_nrmse":0.055256424672805324,"envelope_correlation":0.9976294400613717,"rms_gain_error_db":-0.12738669386415097,"reference_mean_power":0.0026564436783668454,"output_mean_power":0.002579656872689828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":2.0801208532691145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010876326002892773,"output_power":0.0010240057006228108,"reference_active":true,"delta_db":-0.26179842132784187},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.148356435560037e-05,"output_power":6.694515351579463e-05,"reference_active":true,"delta_db":-0.28487057617743566},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.554313443594102e-06,"output_power":2.1844237053636983e-06,"reference_active":true,"delta_db":-0.6793730826567204}],"active_band_count":3,"active_band_mae_db":0.40868069338733265}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.00021647011759154595,"log_rms_mae_db":1.5790742942053244,"rms_envelope_nrmse":0.10206082767397291,"envelope_correlation":0.9930451006613001,"rms_gain_error_db":-0.21379359750451518,"reference_mean_power":0.0021208071215793162,"output_mean_power":0.002018932565649045},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.0688695374548394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066233690351074,"output_power":0.0009919460610775621,"reference_active":true,"delta_db":-0.3136434381723954},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015588791655839427,"output_power":0.00013849812048793056,"reference_active":true,"delta_db":-0.5136857296839483},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732583628298961e-05,"output_power":3.1145795586708195e-05,"reference_active":true,"delta_db":-0.78610117897074}],"active_band_count":3,"active_band_mae_db":0.5378101156090279}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.00018533448652616975,"log_rms_mae_db":0.8821719471320497,"rms_envelope_nrmse":0.06696339473054984,"envelope_correlation":0.99712837809304,"rms_gain_error_db":-0.21379327121881772,"reference_mean_power":0.0021339103768736657,"output_mean_power":0.002031406548896986},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.9317825081344165e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001010025588563974,"output_power":0.000931750497904304,"reference_active":true,"delta_db":-0.3503274308385622},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.797692317548564e-05,"output_power":7.023784819342031e-05,"reference_active":true,"delta_db":-0.4539489667971705},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604806628732514e-06,"output_power":1.952776821195487e-06,"reference_active":true,"delta_db":-1.1766888881676703}],"active_band_count":3,"active_band_mae_db":0.6603217619344677}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":1.5362629098130227,"rms_envelope_nrmse":0.16585097854886477,"envelope_correlation":0.984310220524849,"rms_gain_error_db":0.2620535590159535,"reference_mean_power":0.00012866293461917059,"output_mean_power":0.00013666547297750786},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.8253387298902018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.359313427297327e-05,"output_power":7.384097319160205e-05,"reference_active":true,"delta_db":0.014601122901338269},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.331803440637784e-06,"output_power":9.140917697415305e-06,"reference_active":true,"delta_db":0.9577898529511236},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.889570973194757e-06,"output_power":5.09388468329898e-06,"reference_active":true,"delta_db":4.306859001971375}],"active_band_count":3,"active_band_mae_db":1.7597499926079456}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":0.9937245724728377,"rms_envelope_nrmse":0.08583858568568917,"envelope_correlation":0.9960986337858672,"rms_gain_error_db":0.2620541625439249,"reference_mean_power":0.00012945711895499175,"output_mean_power":0.0001375090728601326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.7711142240683553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.96962455750884e-05,"output_power":7.027683735899057e-05,"reference_active":true,"delta_db":0.0360282489715401},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2400934899690537e-06,"output_power":3.9340424080615594e-06,"reference_active":true,"delta_db":0.8428149563985337},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5855555874941795e-07,"output_power":2.6130917652458956e-07,"reference_active":true,"delta_db":2.1697318890737245}],"active_band_count":3,"active_band_mae_db":1.0161916981479326}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":1.4702862342252219,"rms_envelope_nrmse":0.21637429238952408,"envelope_correlation":0.9674570069794305,"rms_gain_error_db":-0.11898680552886799,"reference_mean_power":2.594581497195349e-05,"output_mean_power":2.5244608278159065e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.3055463889387911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.094870715433079e-06,"output_power":5.055723408437677e-06,"reference_active":true,"delta_db":-0.03349861790284545},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.046923144145425e-06,"output_power":1.1974371876590823e-06,"reference_active":true,"delta_db":0.5833794070428545},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.366550307348401e-07,"output_power":7.227986702945843e-07,"reference_active":true,"delta_db":2.188788768723748}],"active_band_count":3,"active_band_mae_db":0.9352222645564826}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":0.8979472950310957,"rms_envelope_nrmse":0.12191636934699931,"envelope_correlation":0.9892197545008098,"rms_gain_error_db":-0.11901163148482337,"reference_mean_power":2.6105327616819102e-05,"output_mean_power":2.53996647696754e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.2170827427715928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.759181059225196e-06,"output_power":4.807626707601918e-06,"reference_active":true,"delta_db":0.04398511629103935},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.967295837314226e-07,"output_power":7.079460194366353e-07,"reference_active":true,"delta_db":0.742225750402908},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240630896276417e-08,"output_power":4.6369664191706515e-08,"reference_active":true,"delta_db":1.5560438249242234}],"active_band_count":3,"active_band_mae_db":0.7807515638727236}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.894432306289673,"predicted_mos":2.894432306289673,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78240,"input_sample_rate":16000,"input_seconds":4.89,"input_float32_sha256":"adeb5231d102f7316521b7fc742d2d376d5b57b7cae79decb09d2d794b3f33ac","internal_samples":78240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.829268890000094},"started_at":"2026-10-13T04:42:15.933619+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.152599972760072e-05,"native_input_samples":78240,"native_input_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","output_sha256":"2a1e8a84dca6d493932dd07c6e3d932d21c86a9278afbc5279a5511526328420","output_samples":78240,"length_delta_samples":0,"analysis_s":1.1351825550000285,"synthesis_s":0.09569217199987179,"total_s":1.2308747269999003,"rtf":0.25171262310836406,"output_peak":0.5448144580464758,"output_rms":0.06277278404469665,"finished_at":"2026-10-13T04:42:17.169674+00:00","evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"2a1e8a84dca6d493932dd07c6e3d932d21c86a9278afbc5279a5511526328420","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.898329977304668,"spectral_convergence":0.5103757946774177,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.716513547818421,"spectral_convergence":0.3638182833794131,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.0838244553554865,"spectral_convergence":0.4010966125561006,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":7.164801866595056e-05,"delta_db":1.671969076856796,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0011225503509481479,"delta_db":1.2344781989569442,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.002513657919635325,"delta_db":2.082751816468971,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":7.169083982614346e-05,"delta_db":0.9512102562260232,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":9.880198090350284e-05,"delta_db":1.05622169877404,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":6.20733067430512e-05,"delta_db":0.7079953976533798,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.284104407489359,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.743609688656195,"pitch":null,"output_rms_dbfs":-24.044572189312902,"output_peak":0.5448144580464758,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.899555993492858},"common_prefix":{"samples":78240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.898329977304668,"spectral_convergence":0.5103757946774177,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.716513547818421,"spectral_convergence":0.3638182833794131,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.0838244553554865,"spectral_convergence":0.4010966125561006,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":7.164801866595056e-05,"delta_db":1.671969076856796,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0011225503509481479,"delta_db":1.2344781989569442,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.002513657919635325,"delta_db":2.082751816468971,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":7.169083982614346e-05,"delta_db":0.9512102562260232,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":9.880198090350284e-05,"delta_db":1.05622169877404,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":6.20733067430512e-05,"delta_db":0.7079953976533798,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.284104407489359,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.743609688656195,"pitch":null,"output_rms_dbfs":-24.044572189312902,"output_peak":0.5448144580464758,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.899555993492858},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78240,"output_samples":78240,"intelligibility":{"estoi":{"value":0.9029486625222584,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051355966406601,"rms_dbfs":-25.788181877969095,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.607283545458053,"mean":-3.586322495786203e-05,"zero_energy":false},"output":{"rms":0.06277278404469663,"rms_dbfs":-24.044572189312902,"peak":0.5448144580464758,"peak_dbfs":-5.275027516527765,"crest_factor_db":18.769544672785134,"mean":7.3458360643608935e-09,"zero_energy":false},"rms_difference_db":1.7436096886561927},"full_scale":{"reference":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":489,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":7.014925843880822e-07,"output_mean_power":1.226419667688134e-06,"output_rms_dbfs":-59.11360893460319,"output_zero_energy":false,"output_peak":0.011834257627293376,"output_energy_fraction":7.637807570174619e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":482,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":170,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9107142857142857,"absolute_difference_cents_median":5.485082002821425,"absolute_difference_cents_p95":21.65929342218951,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.0002230061513489049,"log_rms_mae_db":1.535294028977051,"rms_envelope_nrmse":0.2632439577996566,"envelope_correlation":0.9877155891464358,"rms_gain_error_db":1.7436100444169536,"reference_mean_power":0.002640134243154231,"output_mean_power":0.0039444550787868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.2388906472962606e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011616255671056531,"output_power":0.0017677230818833197,"reference_active":true,"delta_db":1.8234807053861537},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014996209998372379,"output_power":0.0002719537660461844,"reference_active":true,"delta_db":2.585135639470778},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.935590189308994e-05,"output_power":6.284418455011141e-05,"reference_active":true,"delta_db":2.032552260283849}],"active_band_count":3,"active_band_mae_db":2.147056201713594}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.0001943907377270206,"log_rms_mae_db":1.163245339060518,"rms_envelope_nrmse":0.24088981339441023,"envelope_correlation":0.9945981851901212,"rms_gain_error_db":1.7436114634263578,"reference_mean_power":0.0026564436783668454,"output_mean_power":0.003968823255658049},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":2.0801208532691145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010876326002892773,"output_power":0.0016648826388241367,"reference_active":true,"delta_db":1.849014080750363},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.148356435560037e-05,"output_power":0.0001335395055790603,"reference_active":true,"delta_db":2.7140356453381624},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.554313443594102e-06,"output_power":3.1724050319774168e-06,"reference_active":true,"delta_db":0.9411444095355092}],"active_band_count":3,"active_band_mae_db":1.8347313785413448}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.00021647011759154595,"log_rms_mae_db":2.1588329146540595,"rms_envelope_nrmse":0.2970106994705884,"envelope_correlation":0.9878070084217857,"rms_gain_error_db":1.9841726634769363,"reference_mean_power":0.0021208071215793162,"output_mean_power":0.0033490253849407492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.0688695374548394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066233690351074,"output_power":0.0016839168378855924,"reference_active":true,"delta_db":1.9846823853633135},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015588791655839427,"output_power":0.00029494196365234207,"reference_active":true,"delta_db":2.769241146079837},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732583628298961e-05,"output_power":6.229884666614888e-05,"reference_active":true,"delta_db":2.224704598262596}],"active_band_count":3,"active_band_mae_db":2.3262093765685825}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.00018533448652616975,"log_rms_mae_db":1.7089133923023236,"rms_envelope_nrmse":0.2750062472986363,"envelope_correlation":0.9944312924359408,"rms_gain_error_db":1.9841731891765693,"reference_mean_power":0.0021339103768736657,"output_mean_power":0.0033697175069273004},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.9317825081344165e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001010025588563974,"output_power":0.0016024376083317911,"reference_active":true,"delta_db":2.0044875247245524},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.797692317548564e-05,"output_power":0.000147690181686458,"reference_active":true,"delta_db":2.773855300192581},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604806628732514e-06,"output_power":3.3381042007516755e-06,"reference_active":true,"delta_db":1.1517838896807395}],"active_band_count":3,"active_band_mae_db":1.9767089048659578}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":1.3516006826284943,"rms_envelope_nrmse":0.16038800284287566,"envelope_correlation":0.9947042521345021,"rms_gain_error_db":1.0316760418187376,"reference_mean_power":0.00012866293461917059,"output_mean_power":0.00016316276523387137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.8253387298902018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.359313427297327e-05,"output_power":9.324170056041355e-05,"reference_active":true,"delta_db":1.0277288609266093},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.331803440637784e-06,"output_power":9.428633647865401e-06,"reference_active":true,"delta_db":1.0923794805074862},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.889570973194757e-06,"output_power":2.321615660783245e-06,"reference_active":true,"delta_db":0.8942711582867564}],"active_band_count":3,"active_band_mae_db":1.0047931665736172}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":0.9390668217961063,"rms_envelope_nrmse":0.13607881266083374,"envelope_correlation":0.9985572593001847,"rms_gain_error_db":1.0316809108998906,"reference_mean_power":0.00012945711895499175,"output_mean_power":0.0001641700871503313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.7711142240683553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.96962455750884e-05,"output_power":8.77666996133616e-05,"reference_active":true,"delta_db":1.0012038343475953},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2400934899690537e-06,"output_power":4.1477731257663735e-06,"reference_active":true,"delta_db":1.072574518341351},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5855555874941795e-07,"output_power":1.9056269391163227e-07,"reference_active":true,"delta_db":0.7985641121290117}],"active_band_count":3,"active_band_mae_db":0.9574474882726527}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":1.397444666514369,"rms_envelope_nrmse":0.271833543991673,"envelope_correlation":0.9589886289535418,"rms_gain_error_db":0.7644197936985826,"reference_mean_power":2.594581497195349e-05,"output_mean_power":3.093921539877661e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.3055463889387911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.094870715433079e-06,"output_power":5.5338196033747514e-06,"reference_active":true,"delta_db":0.35891829427788097},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.046923144145425e-06,"output_power":1.4118594327578785e-06,"reference_active":true,"delta_db":1.2987665895296276},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.366550307348401e-07,"output_power":5.713111448547181e-07,"reference_active":true,"delta_db":1.1673422808263747}],"active_band_count":3,"active_band_mae_db":0.9416757215446278}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":0.9249822006842479,"rms_envelope_nrmse":0.16246043740961946,"envelope_correlation":0.9880048658677685,"rms_gain_error_db":0.7643906205161828,"reference_mean_power":2.6105327616819102e-05,"output_mean_power":3.1129217933856406e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.2170827427715928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.759181059225196e-06,"output_power":5.186685603549267e-06,"reference_active":true,"delta_db":0.3735769607243459},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.967295837314226e-07,"output_power":8.036187277171936e-07,"reference_active":true,"delta_db":1.2927247993911783},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240630896276417e-08,"output_power":4.089705774395742e-08,"reference_active":true,"delta_db":1.010624964319587}],"active_band_count":3,"active_band_mae_db":0.8923089081450369}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"facodec","status":"measured","mos":{"status":"measured","value":4.37321138381958,"predicted_mos":4.37321138381958,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78400,"input_sample_rate":16000,"input_seconds":4.9,"input_float32_sha256":"5af19f023621e4039bf21f23834f861c577d14322c6a2ad66b9506ab5cb52dd7","internal_samples":78400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.2636850100006995},"started_at":"2026-10-13T04:52:48.926381+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0339999537100084e-05,"native_input_samples":78240,"native_input_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","output_sha256":"d95a8dfd7ce55521c0ac5d4de912ceb441ba1cb77f99725c6da222c4b9afd3f1","output_samples":78400,"length_delta_samples":160,"analysis_s":1.7122088079995592,"synthesis_s":11.053620729000613,"total_s":12.765829537000172,"rtf":2.610599087321099,"output_peak":0.3968837857246399,"output_rms":0.045731914604878264,"finished_at":"2026-10-13T04:53:01.698600+00:00","evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"d95a8dfd7ce55521c0ac5d4de912ceb441ba1cb77f99725c6da222c4b9afd3f1","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":78240,"output_samples":78400,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.795612322735085,"output_peak":0.3968837857246399},"common_prefix":{"samples":78240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9979591836734694,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.324570634628536,"spectral_convergence":0.263715903562382,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.787431499494653,"spectral_convergence":0.2670612002139456,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.936115300262276,"spectral_convergence":0.29521879009923674,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":4.724514520189241e-05,"delta_db":-0.13650089281025898,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0008162699939221404,"delta_db":-0.14924177958673915,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.001134038536441259,"delta_db":-1.374031803375844,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":2.4318635516010622e-05,"delta_db":-3.7440343871926727,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":3.7562186005236764e-05,"delta_db":-3.143926231164042,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":3.625038486242028e-05,"delta_db":-1.6279272004785894,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6959437157680244,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9985583175363426,"pitch":null,"output_rms_dbfs":-26.786740195505438,"output_peak":0.3968837857246399,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.6827058114618225},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78240,"output_samples":78240,"intelligibility":{"estoi":{"value":0.8905832727823989,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051355966406601,"rms_dbfs":-25.788181877969095,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.607283545458053,"mean":-3.586322495786203e-05,"zero_energy":false},"output":{"rms":0.045778650940683695,"rms_dbfs":-26.786740195505438,"peak":0.3968837857246399,"peak_dbfs":-8.026732867619247,"crest_factor_db":18.760007327886193,"mean":1.101264449573998e-06,"zero_energy":false},"rms_difference_db":-0.998558317536343},"full_scale":{"reference":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":489,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":7.014925843880822e-07,"output_mean_power":7.013869320268787e-07,"output_rms_dbfs":-61.54042329987859,"output_zero_energy":false,"output_peak":0.007219585590064526,"output_energy_fraction":8.213042619359095e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":482,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":171,"both_estimated_voiced_frames":157,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9345238095238095,"absolute_difference_cents_median":4.4637757082458105,"absolute_difference_cents_p95":30.536416027342938,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.0002230061513489049,"log_rms_mae_db":1.2688206802574429,"rms_envelope_nrmse":0.16443683165547926,"envelope_correlation":0.9867241012139746,"rms_gain_error_db":-0.9985592057057351,"reference_mean_power":0.002640134243154231,"output_mean_power":0.0020978290229421498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.2388906472962606e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011616255671056531,"output_power":0.0008305617785337551,"reference_active":true,"delta_db":-1.4569422066302442},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014996209998372379,"output_power":0.00012374357106450525,"reference_active":true,"delta_db":-0.8345886841479147},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.935590189308994e-05,"output_power":3.312986472995953e-05,"reference_active":true,"delta_db":-0.7479020663407264}],"active_band_count":3,"active_band_mae_db":1.0131443190396283}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.0001943907377270206,"log_rms_mae_db":1.0101550341581083,"rms_envelope_nrmse":0.15565724945727732,"envelope_correlation":0.9886942265186301,"rms_gain_error_db":-0.9985611036877867,"reference_mean_power":0.0026564436783668454,"output_mean_power":0.002110787443969464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":2.0801208532691145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010876326002892773,"output_power":0.0007772217060644621,"reference_active":true,"delta_db":-1.4593729551857288},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.148356435560037e-05,"output_power":5.9141998780384904e-05,"reference_active":true,"delta_db":-0.8231020143866629},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.554313443594102e-06,"output_power":2.3821504230577e-06,"reference_active":true,"delta_db":-0.3030500718481368}],"active_band_count":3,"active_band_mae_db":0.8618416804735096}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.00021647011759154595,"log_rms_mae_db":1.6057417654779536,"rms_envelope_nrmse":0.16657246904872963,"envelope_correlation":0.9883571794972194,"rms_gain_error_db":-1.0411772640646013,"reference_mean_power":0.0021208071215793162,"output_mean_power":0.0016687199061156907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.0688695374548394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066233690351074,"output_power":0.000767921933883603,"reference_active":true,"delta_db":-1.4253532867231535},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015588791655839427,"output_power":0.00013150985857687431,"reference_active":true,"delta_db":-0.738541420075365},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732583628298961e-05,"output_power":3.051267639602376e-05,"reference_active":true,"delta_db":-0.8752924372552693}],"active_band_count":3,"active_band_mae_db":1.0130623813512625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.00018533448652616975,"log_rms_mae_db":1.19900699410664,"rms_envelope_nrmse":0.15606726461261045,"envelope_correlation":0.9906403011517769,"rms_gain_error_db":-1.0411784123322647,"reference_mean_power":0.0021339103768736657,"output_mean_power":0.0016790295289127477},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.9317825081344165e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001010025588563974,"output_power":0.0007288483645228387,"reference_active":true,"delta_db":-1.4169519298810818},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.797692317548564e-05,"output_power":6.506900413705642e-05,"reference_active":true,"delta_db":-0.785919347077576},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604806628732514e-06,"output_power":2.287911500945387e-06,"reference_active":true,"delta_db":-0.48882278892346454}],"active_band_count":3,"active_band_mae_db":0.8972313552940409}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":1.9774921409192505,"rms_envelope_nrmse":0.3801555985726946,"envelope_correlation":0.9758580372215027,"rms_gain_error_db":-3.618121637207029,"reference_mean_power":0.00012866293461917059,"output_mean_power":5.592954536616559e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.8253387298902018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.359313427297327e-05,"output_power":3.1020217816762854e-05,"reference_active":true,"delta_db":-3.751924565727587},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.331803440637784e-06,"output_power":4.41163041437778e-06,"reference_active":true,"delta_db":-2.206116912382291},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.889570973194757e-06,"output_power":1.8114750667454467e-06,"reference_active":true,"delta_db":-0.18330848113396808}],"active_band_count":3,"active_band_mae_db":2.047116653081282}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":1.7031608577224924,"rms_envelope_nrmse":0.37124275884259206,"envelope_correlation":0.9826085406992529,"rms_gain_error_db":-3.618146929612696,"reference_mean_power":0.00012945711895499175,"output_mean_power":5.627444813135995e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.7711142240683553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.96962455750884e-05,"output_power":2.9973277387175372e-05,"reference_active":true,"delta_db":-3.6647515110083004},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2400934899690537e-06,"output_power":2.2206561986822807e-06,"reference_active":true,"delta_db":-1.6407621514905315},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5855555874941795e-07,"output_power":1.598391116109855e-07,"reference_active":true,"delta_db":0.03501584647573657}],"active_band_count":3,"active_band_mae_db":1.7801765029915229}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":1.8040902102083336,"rms_envelope_nrmse":0.3271786695703596,"envelope_correlation":0.9262309835500868,"rms_gain_error_db":-1.007688277628053,"reference_mean_power":2.594581497195349e-05,"output_mean_power":2.0573040872661305e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.3055463889387911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.094870715433079e-06,"output_power":4.595676312959441e-06,"reference_active":true,"delta_db":-0.4478373561850171},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.046923144145425e-06,"output_power":1.723471850912508e-06,"reference_active":true,"delta_db":2.164893937831293},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.366550307348401e-07,"output_power":6.511223501232253e-07,"reference_active":true,"delta_db":1.7352413493123335}],"active_band_count":3,"active_band_mae_db":1.4493242144428813}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":1.471397971295593,"rms_envelope_nrmse":0.26450214383142767,"envelope_correlation":0.9519625438458291,"rms_gain_error_db":-1.0076939046167788,"reference_mean_power":2.6105327616819102e-05,"output_mean_power":2.0699495343663012e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.2170827427715928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.759181059225196e-06,"output_power":4.5288053627803345e-06,"reference_active":true,"delta_db":-0.21548571351975856},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.967295837314226e-07,"output_power":1.0213282231812583e-06,"reference_active":true,"delta_db":2.333877641266357},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240630896276417e-08,"output_power":4.5274685478320844e-08,"reference_active":true,"delta_db":1.4522587370602018}],"active_band_count":3,"active_band_mae_db":1.333874030615439}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.6576058864593506,"predicted_mos":3.6576058864593506,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":216064,"input_sample_rate":44100,"input_seconds":4.899410430839002,"input_float32_sha256":"91ebc50ea2fe00ff1c387bfc133bc94ec7f825009dc21449c491f1d578916548","internal_samples":78391,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.470286384999781},"started_at":"2026-10-13T07:14:42.445627+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.01135454299947014,"native_input_samples":215649,"native_input_sha256":"7aed525d9042a06f9b43d8a0b7c2651a7a7b89f04cbc1b85e6a75bc0140e8c84","output_sha256":"9b54047ce6a3b7ebe5765e8cc956102d177882a3240c12afad68ba50be3b8858","output_samples":216064,"length_delta_samples":415,"analysis_s":0.16242806800073595,"synthesis_s":29.6712056439992,"total_s":29.833633711999937,"rtf":6.100947589366041,"output_peak":0.5238268375396729,"output_rms":0.04924317175488035,"finished_at":"2026-10-13T07:15:12.317862+00:00","evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"9b54047ce6a3b7ebe5765e8cc956102d177882a3240c12afad68ba50be3b8858","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":78240,"output_samples":78391,"length_delta_samples":151,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.150529772712474,"output_peak":0.5259966972852742},"common_prefix":{"samples":78240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998073758467171,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.734606306086537,"spectral_convergence":0.32493331946204485,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.873270935991151,"spectral_convergence":0.3123681101748508,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.738044493711103,"spectral_convergence":0.38085714974645174,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":3.653664182238453e-05,"delta_db":-1.2527864963954052,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0008531439821970289,"delta_db":0.04264321659813479,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.0013823507929937766,"delta_db":-0.5141272716821393,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":5.8998179480839086e-05,"delta_db":0.10495968374296927,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":6.963941071915254e-05,"delta_db":-0.4628839455699995,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":3.0327652555284696e-05,"delta_db":-2.402665457790824,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7966776786299121,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.35397450087488064,"pitch":null,"output_rms_dbfs":-26.142156378843975,"output_peak":0.5259966972852742,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.781973911929597},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78240,"output_samples":78240,"intelligibility":{"estoi":{"value":0.7674211502137845,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051355966406601,"rms_dbfs":-25.788181877969095,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.607283545458053,"mean":-3.586322495786203e-05,"zero_energy":false},"output":{"rms":0.04930513826943868,"rms_dbfs":-26.142156378843975,"peak":0.5259966972852742,"peak_dbfs":-5.580339655148938,"crest_factor_db":20.56181672369504,"mean":-3.34754808839507e-05,"zero_energy":false},"rms_difference_db":-0.35397450087488025},"full_scale":{"reference":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":489,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":7.014925843880822e-07,"output_mean_power":5.478292004998424e-07,"output_rms_dbfs":-62.613548226151615,"output_zero_energy":false,"output_peak":0.006569577858392189,"output_energy_fraction":5.530102769964204e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":482,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":169,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9107142857142857,"absolute_difference_cents_median":10.33605693372813,"absolute_difference_cents_p95":52.17952448800811,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.0002230061513489049,"log_rms_mae_db":1.5693482705004045,"rms_envelope_nrmse":0.10055175691009907,"envelope_correlation":0.9933639207541807,"rms_gain_error_db":-0.3539740079880489,"reference_mean_power":0.002640134243154231,"output_mean_power":0.002433484639415931},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.2388906472962606e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011616255671056531,"output_power":0.0010939427757027075,"reference_active":true,"delta_db":-0.2607155759325607},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014996209998372379,"output_power":0.00013267926743165312,"reference_active":true,"delta_db":-0.5317844844104362},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.935590189308994e-05,"output_power":3.0486032187482618e-05,"reference_active":true,"delta_db":-1.1090896545776547}],"active_band_count":3,"active_band_mae_db":0.6338632383068838}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.0001943907377270206,"log_rms_mae_db":1.24290031382404,"rms_envelope_nrmse":0.06845071143517387,"envelope_correlation":0.9976889602303175,"rms_gain_error_db":-0.35397145401039604,"reference_mean_power":0.0026564436783668454,"output_mean_power":0.00244851893615021},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":2.0801208532691145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010876326002892773,"output_power":0.0010234898958403696,"reference_active":true,"delta_db":-0.2639865693842215},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.148356435560037e-05,"output_power":6.368179751465808e-05,"reference_active":true,"delta_db":-0.5019088590228401},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.554313443594102e-06,"output_power":2.160812862703145e-06,"reference_active":true,"delta_db":-0.7265703269963357}],"active_band_count":3,"active_band_mae_db":0.49748858513446576}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.00021647011759154595,"log_rms_mae_db":2.43008327772001,"rms_envelope_nrmse":0.11531857208519447,"envelope_correlation":0.9918398003936979,"rms_gain_error_db":-0.3905143430741059,"reference_mean_power":0.0021208071215793162,"output_mean_power":0.0019384282871592578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.0688695374548394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066233690351074,"output_power":0.000974832358029818,"reference_active":true,"delta_db":-0.38922464563183723},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015588791655839427,"output_power":0.00013286213338915349,"reference_active":true,"delta_db":-0.6941123104291568},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732583628298961e-05,"output_power":2.743309068751314e-05,"reference_active":true,"delta_db":-1.3373480767353643}],"active_band_count":3,"active_band_mae_db":0.8068950109321195}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.00018533448652616975,"log_rms_mae_db":1.7781844793364925,"rms_envelope_nrmse":0.08279336574151512,"envelope_correlation":0.9963633748287111,"rms_gain_error_db":-0.39051384475141165,"reference_mean_power":0.0021339103768736657,"output_mean_power":0.0019504049516493134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.9317825081344165e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001010025588563974,"output_power":0.0009227433191360327,"reference_active":true,"delta_db":-0.3925146709417283},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.797692317548564e-05,"output_power":6.712540478655487e-05,"reference_active":true,"delta_db":-0.650791770936813},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604806628732514e-06,"output_power":1.9188647379392747e-06,"reference_active":true,"delta_db":-1.2527713822149413}],"active_band_count":3,"active_band_mae_db":0.7653592746978276}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":2.941898623878196,"rms_envelope_nrmse":0.21412167100087698,"envelope_correlation":0.9751488685469794,"rms_gain_error_db":-0.8615882909673735,"reference_mean_power":0.00012866293461917059,"output_mean_power":0.00010551024311258094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.8253387298902018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.359313427297327e-05,"output_power":5.8941396408749274e-05,"reference_active":true,"delta_db":-0.9641687886240041},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.331803440637784e-06,"output_power":5.884398796707117e-06,"reference_active":true,"delta_db":-0.9550871538181736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.889570973194757e-06,"output_power":2.01341256707675e-06,"reference_active":true,"delta_db":0.27569566141913127}],"active_band_count":3,"active_band_mae_db":0.7316505346204364}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":2.452415241673259,"rms_envelope_nrmse":0.18476068557752146,"envelope_correlation":0.9826473966641691,"rms_gain_error_db":-0.861588081872861,"reference_mean_power":0.00012945711895499175,"output_mean_power":0.0001061615203450069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.7711142240683553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.96962455750884e-05,"output_power":5.561740439746626e-05,"reference_active":true,"delta_db":-0.9799866702163217},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2400934899690537e-06,"output_power":2.5963762556065105e-06,"reference_active":true,"delta_db":-0.9618991279701917},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5855555874941795e-07,"output_power":1.859573465779372e-07,"reference_active":true,"delta_db":0.6923186819151311}],"active_band_count":3,"active_band_mae_db":0.8780681600338816}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":4.426653124534434,"rms_envelope_nrmse":0.31014017146139017,"envelope_correlation":0.9363447446722276,"rms_gain_error_db":-1.0071056496748456,"reference_mean_power":2.594581497195349e-05,"output_mean_power":2.0575801035005124e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.3055463889387911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.094870715433079e-06,"output_power":3.7567990586954076e-06,"reference_active":true,"delta_db":-1.323152016557658},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.046923144145425e-06,"output_power":1.7389780947986751e-06,"reference_active":true,"delta_db":2.203793106235265},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.366550307348401e-07,"output_power":7.555711869563702e-07,"reference_active":true,"delta_db":2.3813692006482206}],"active_band_count":3,"active_band_mae_db":1.9694381078137144}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":3.821925951096579,"rms_envelope_nrmse":0.26386314111565773,"envelope_correlation":0.9533751875524915,"rms_gain_error_db":-1.006996386204053,"reference_mean_power":2.6105327616819102e-05,"output_mean_power":2.0702820147284253e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.2170827427715928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.759181059225196e-06,"output_power":3.843068618559914e-06,"reference_active":true,"delta_db":-0.928540885722581},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.967295837314226e-07,"output_power":1.0988231434316592e-06,"reference_active":true,"delta_db":2.651502288082415},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240630896276417e-08,"output_power":7.276498291253648e-08,"reference_active":true,"delta_db":3.512928634342481}],"active_band_count":3,"active_band_mae_db":2.364323936049159}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"golf","status":"measured","mos":{"status":"measured","value":2.883974313735962,"predicted_mos":2.883974313735962,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":117120,"input_sample_rate":24000,"input_seconds":4.88,"input_float32_sha256":"d520972103493454c85127cd5b9796d3e1087e414342d68a22b36df5f2bc8276","internal_samples":78080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":11.67862957699981},"started_at":"2026-10-13T05:02:35.709722+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002402923999397899,"native_input_samples":117360,"native_input_sha256":"8ec1d8c087721217a7fd0321c68b2a96cee8bf7d52c9f92c4b74ab313f3245d1","output_sha256":"1cf9f9b7228c21afd8af1e768f9dd5947a55b8588c1d1da20203ae03d444e31c","output_samples":117120,"length_delta_samples":-240,"analysis_s":0.5436282040000151,"synthesis_s":0.11564632299996447,"total_s":0.6592745269999796,"rtf":0.1348209666666625,"output_peak":0.34280914068222046,"output_rms":0.04201013237983044,"finished_at":"2026-10-13T05:02:36.377943+00:00","analysis_stages":{"external_f0_s":0.044637631999648875},"evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"1cf9f9b7228c21afd8af1e768f9dd5947a55b8588c1d1da20203ae03d444e31c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":78240,"output_samples":78080,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.52867822103073,"output_peak":0.33859603899887936},"common_prefix":{"samples":78080,"sample_rate":16000,"reference_coverage":0.9979550102249489,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78080,"output_samples":78080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.723367759883122,"spectral_convergence":0.4518752838310437,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.573128521199415,"spectral_convergence":0.32934791602803554,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.675278283602852,"spectral_convergence":0.3481303664884556,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8853403025432805e-05,"output_power":4.1128991829654136e-05,"delta_db":-0.7474675711071574,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008465361259075649,"output_power":0.0005976600265270656,"delta_db":-1.5119128658806191,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015592679625199193,"output_power":0.00098152213846528,"delta_db":-2.0102065609378923,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.7707238127460236e-05,"output_power":5.584581682556466e-05,"delta_db":-0.1423964193780993,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.763023406406785e-05,"output_power":6.103116043602056e-05,"delta_db":-1.044792679771958,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.284346664160813e-05,"output_power":2.9387263525270434e-05,"delta_db":-2.548321530536703,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3341829379354049,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.7493843548826469,"pitch":null,"output_rms_dbfs":-27.52867822103073,"output_peak":0.33859603899887936,"reference_rms_dbfs":-25.779293866148084,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.657258188228464},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78080,"output_samples":78080,"intelligibility":{"estoi":{"value":0.8536163569690427,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05140854433152191,"rms_dbfs":-25.779293866148084,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.598395533637046,"mean":-3.5873397451932314e-05,"zero_energy":false},"output":{"rms":0.042030648312973415,"rms_dbfs":-27.52867822103073,"peak":0.33859603899887936,"peak_dbfs":-9.406362534169045,"crest_factor_db":18.122315686861686,"mean":0.00017532332899982572,"zero_energy":false},"rms_difference_db":-1.7493843548826469},"full_scale":{"reference":{"sample_count":78080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":488,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":119,"quiet_samples":19040,"reference_mean_power":7.01499603954809e-07,"output_mean_power":6.298773748768688e-07,"output_rms_dbfs":-62.00743991177988,"output_zero_energy":false,"output_peak":0.005995995975730811,"output_energy_fraction":8.694627296996002e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":481,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":174,"both_estimated_voiced_frames":157,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9345238095238095,"absolute_difference_cents_median":5.607951881956736,"absolute_difference_cents_p95":24.44882960944486,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":975,"rms_floor":0.0002230061513489049,"log_rms_mae_db":2.6473468600965386,"rms_envelope_nrmse":0.228125576191452,"envelope_correlation":0.9826407723434075,"rms_gain_error_db":-1.7493847937667684,"reference_mean_power":0.0026455482771464934,"output_mean_power":0.0017683865853246933},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":975,"frequency_resolution_hz":0.20512820512820512,"power_floor":2.248549190955124e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001156049050776852,"output_power":0.0007992116922003296,"reference_active":true,"delta_db":-1.6031443259818448},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000161993084682787,"output_power":0.00011562233542212958,"reference_active":true,"delta_db":-1.4645473803976936},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.923299597846516e-05,"output_power":2.6401437937388807e-05,"reference_active":true,"delta_db":-1.7202389266407496}],"active_band_count":3,"active_band_mae_db":1.5959768776734293}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":969,"rms_floor":0.0001943907377270206,"log_rms_mae_db":2.3764722450518034,"rms_envelope_nrmse":0.2061208738977248,"envelope_correlation":0.9912422105320464,"rms_gain_error_db":-1.7493868125467174,"reference_mean_power":0.0026619243510543813,"output_mean_power":0.001779332158126036},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":969,"frequency_resolution_hz":0.2063983488132095,"power_floor":2.0890404141386712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001093658348647207,"output_power":0.000762652613808427,"reference_active":true,"delta_db":-1.565549095335136},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.095803703628842e-05,"output_power":5.1481543734152006e-05,"reference_active":true,"delta_db":-1.3935003153185452},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5658228222885596e-06,"output_power":1.5666428484748039e-06,"reference_active":true,"delta_db":-2.142566632033051}],"active_band_count":3,"active_band_mae_db":1.7005386808955774}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":975,"rms_floor":0.00021647011759154595,"log_rms_mae_db":2.855377217451649,"rms_envelope_nrmse":0.2467597076309535,"envelope_correlation":0.9810968410821305,"rms_gain_error_db":-1.8998389220176697,"reference_mean_power":0.0021251570722775506,"output_mean_power":0.001372167543247624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":975,"frequency_resolution_hz":0.20512820512820512,"power_floor":2.0780814743064144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010607038023788995,"output_power":0.000693819951886333,"reference_active":true,"delta_db":-1.84347341192198},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001672509720557198,"output_power":0.0001156118846379738,"reference_active":true,"delta_db":-1.6036616968748474},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.722530089295994e-05,"output_power":2.458666395074775e-05,"reference_active":true,"delta_db":-1.8013861149824606}],"active_band_count":3,"active_band_mae_db":1.7495070745930958}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":969,"rms_floor":0.00018533448652616975,"log_rms_mae_db":2.492545408827689,"rms_envelope_nrmse":0.22300297836885477,"envelope_correlation":0.9901661565348855,"rms_gain_error_db":-1.8998379876419345,"reference_mean_power":0.002138313959280212,"output_mean_power":0.001380662955102979},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":969,"frequency_resolution_hz":0.2063983488132095,"power_floor":1.9402782186487655e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001015430902305757,"output_power":0.0006712620460585194,"reference_active":true,"delta_db":-1.7975828366317426},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.749880813501485e-05,"output_power":5.2932690949878104e-05,"reference_active":true,"delta_db":-1.6557105061363373},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.571189753866295e-06,"output_power":1.5307028828423458e-06,"reference_active":true,"delta_db":-2.2524322864187414}],"active_band_count":3,"active_band_mae_db":1.9019085430622737}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":975,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":2.30558793599174,"rms_envelope_nrmse":0.20579165526741805,"envelope_correlation":0.97703326854627,"rms_gain_error_db":-0.7647912745757174,"reference_mean_power":0.00012892645342902628,"output_mean_power":0.00010810926347132873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":975,"frequency_resolution_hz":0.20512820512820512,"power_floor":1.835174533106084e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.356513215396284e-05,"output_power":6.221816063129852e-05,"reference_active":true,"delta_db":-0.7275485166147511},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.728646735713163e-06,"output_power":8.905406963319254e-06,"reference_active":true,"delta_db":0.615503140047591},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8973084834572131e-06,"output_power":2.1793771854685515e-06,"reference_active":true,"delta_db":0.6019445164288244}],"active_band_count":3,"active_band_mae_db":0.6483320576970555}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":969,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":1.9543549744030646,"rms_envelope_nrmse":0.15365894361077617,"envelope_correlation":0.9890989638761299,"rms_gain_error_db":-0.7647868143891803,"reference_mean_power":0.00012972394867554412,"output_mean_power":0.00010877810236666545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":969,"frequency_resolution_hz":0.2063983488132095,"power_floor":1.7806732455935958e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.004310203864619e-05,"output_power":5.95216173999572e-05,"reference_active":true,"delta_db":-0.7069064776264096},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2177145311653527e-06,"output_power":3.5724838693121286e-06,"reference_active":true,"delta_db":0.4542276473240682},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6304098018001337e-07,"output_power":1.3586712892741862e-07,"reference_active":true,"delta_db":-0.7918237937164829}],"active_band_count":3,"active_band_mae_db":0.6509859728889869}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":975,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":3.465484334215098,"rms_envelope_nrmse":0.30198490049127136,"envelope_correlation":0.9407033113450898,"rms_gain_error_db":-0.7182112291843257,"reference_mean_power":2.599861379947209e-05,"output_mean_power":2.2035812575597722e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":975,"frequency_resolution_hz":0.20512820512820512,"power_floor":1.3105557229772329e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.051041286643557e-06,"output_power":3.340525276174075e-06,"reference_active":true,"delta_db":-1.7956615619812824},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1029636880341872e-06,"output_power":9.045582036348405e-07,"reference_active":true,"delta_db":-0.8612469806051702},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.36011347302708e-07,"output_power":2.561434207605451e-07,"reference_active":true,"delta_db":-2.3101458686647343}],"active_band_count":3,"active_band_mae_db":1.6556848037503957}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78080,"uncovered_tail_samples":0,"metrics":{"frame_count":969,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":3.269201666708982,"rms_envelope_nrmse":0.26279326758971205,"envelope_correlation":0.9547826637807908,"rms_gain_error_db":-0.7181722250327321,"reference_mean_power":2.6158835524822658e-05,"output_mean_power":2.217181186045391e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":969,"frequency_resolution_hz":0.2063983488132095,"power_floor":1.221816281368326e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.768262857965479e-06,"output_power":3.2008239741495698e-06,"reference_active":true,"delta_db":-1.7309839731047283},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.942460814325649e-07,"output_power":4.477081800932051e-07,"reference_active":true,"delta_db":-1.2297129712539963},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3819058142633776e-08,"output_power":1.721443129802266e-08,"reference_active":true,"delta_db":-2.9326882859934655}],"active_band_count":3,"active_band_mae_db":1.96446174345073}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.2630505561828613,"predicted_mos":3.2630505561828613,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":117480,"input_sample_rate":24000,"input_seconds":4.895,"input_float32_sha256":"83d9500545b76f41d8fdcc6a471a283df18809b43b4f71e2b147b8e498aa3dfc","internal_samples":78320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5893119949978427},"started_at":"2026-10-13T07:37:57.569016+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003279705000750255,"native_input_samples":117360,"native_input_sha256":"8ec1d8c087721217a7fd0321c68b2a96cee8bf7d52c9f92c4b74ab313f3245d1","output_sha256":"5e7e6d5b7b677632fa4c4bdd115084655fd1cb4583e6a1339431ae695334f307","output_samples":117480,"length_delta_samples":120,"analysis_s":1.6819477060016652,"synthesis_s":11.251238297998498,"total_s":12.933186004000163,"rtf":2.644823313701465,"output_peak":0.42449548840522766,"output_rms":0.04887032162948658,"finished_at":"2026-10-13T07:38:10.520275+00:00","analysis_stages":{"external_f0_s":1.0898425039995345},"evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"5e7e6d5b7b677632fa4c4bdd115084655fd1cb4583e6a1339431ae695334f307","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":78240,"output_samples":78320,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.21851503909898,"output_peak":0.4248555600241787},"common_prefix":{"samples":78240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9989785495403473,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.951234415038483,"spectral_convergence":0.4200135645073861,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.617006956786267,"spectral_convergence":0.283759054666422,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.713035391901987,"spectral_convergence":0.2762874994330075,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":4.5322975091189146e-05,"delta_db":-0.31688867267102694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0007639805252407224,"delta_db":-0.43675722048868987,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.0014124383899653935,"delta_db":-0.42061480023910797,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":5.979143347388496e-05,"delta_db":0.16296323209295904,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":7.291266598868069e-05,"delta_db":-0.26340503938432686,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":3.662391545533163e-05,"delta_db":-1.5834056900676106,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5306724424906202,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.42589538887223277,"pitch":null,"output_rms_dbfs":-26.21407726684133,"output_peak":0.4248555600241787,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.760425587908912},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78240,"output_samples":78240,"intelligibility":{"estoi":{"value":0.8808318725830799,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051355966406601,"rms_dbfs":-25.788181877969095,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.607283545458053,"mean":-3.586322495786203e-05,"zero_energy":false},"output":{"rms":0.04889856751700609,"rms_dbfs":-26.21407726684133,"peak":0.4248555600241787,"peak_dbfs":-7.43517387647442,"crest_factor_db":18.778903390366906,"mean":4.458178807608715e-05,"zero_energy":false},"rms_difference_db":-0.42589538887223455},"full_scale":{"reference":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":489,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":7.014925843880822e-07,"output_mean_power":8.642936118453481e-07,"output_rms_dbfs":-60.6333869693926,"output_zero_energy":false,"output_peak":0.024758740651875645,"output_energy_fraction":8.870363488799792e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":482,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":175,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9464285714285714,"absolute_difference_cents_median":4.625968741260513,"absolute_difference_cents_p95":18.106920031296116,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.0002230061513489049,"log_rms_mae_db":2.1597127256707562,"rms_envelope_nrmse":0.16564495968848653,"envelope_correlation":0.9796032169075208,"rms_gain_error_db":-0.42589527912626957,"reference_mean_power":0.002640134243154231,"output_mean_power":0.002393516811058469},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.2388906472962606e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011616255671056531,"output_power":0.0010865745839270702,"reference_active":true,"delta_db":-0.29006619949112905},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014996209998372379,"output_power":0.00013730004767274501,"reference_active":true,"delta_db":-0.38310825373654245},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.935590189308994e-05,"output_power":3.5013088287098056e-05,"reference_active":true,"delta_db":-0.5077945041081895}],"active_band_count":3,"active_band_mae_db":0.3936563191119537}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.0001943907377270206,"log_rms_mae_db":1.628456327081032,"rms_envelope_nrmse":0.1061946138549314,"envelope_correlation":0.9924101093210724,"rms_gain_error_db":-0.42589341060319585,"reference_mean_power":0.0026564436783668454,"output_mean_power":0.0024083038026872458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":2.0801208532691145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010876326002892773,"output_power":0.00101756152752035,"reference_active":true,"delta_db":-0.2892153789968748},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.148356435560037e-05,"output_power":6.557389870191844e-05,"reference_active":true,"delta_db":-0.3747519397004906},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.554313443594102e-06,"output_power":2.2850320018294965e-06,"reference_active":true,"delta_db":-0.4838190238362442}],"active_band_count":3,"active_band_mae_db":0.38259544751120317}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.00021647011759154595,"log_rms_mae_db":2.511313265403049,"rms_envelope_nrmse":0.17665703672974953,"envelope_correlation":0.9790359863162912,"rms_gain_error_db":-0.4180456297748633,"reference_mean_power":0.0021208071215793162,"output_mean_power":0.0019261788509788477},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.0688695374548394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066233690351074,"output_power":0.000988338434997876,"reference_active":true,"delta_db":-0.32946716285014943},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015588791655839427,"output_power":0.00014162062341407632,"reference_active":true,"delta_db":-0.4168595092296199},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732583628298961e-05,"output_power":3.426578044734036e-05,"reference_active":true,"delta_db":-0.3714891896335999}],"active_band_count":3,"active_band_mae_db":0.3726052872377898}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.00018533448652616975,"log_rms_mae_db":1.9670489278163934,"rms_envelope_nrmse":0.11529725301100231,"envelope_correlation":0.991588448188725,"rms_gain_error_db":-0.418044255059558,"reference_mean_power":0.0021339103768736657,"output_mean_power":0.0019380802228867733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.9317825081344165e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001010025588563974,"output_power":0.0009368793734281705,"reference_active":true,"delta_db":-0.3264869906735761},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.797692317548564e-05,"output_power":7.103912621423059e-05,"reference_active":true,"delta_db":-0.40468483783013554},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604806628732514e-06,"output_power":2.329126469227908e-06,"reference_active":true,"delta_db":-0.411284293978654}],"active_band_count":3,"active_band_mae_db":0.3808187074941219}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":1.8514482920859643,"rms_envelope_nrmse":0.17208449719274016,"envelope_correlation":0.9818377352980884,"rms_gain_error_db":-0.0897694707961002,"reference_mean_power":0.00012866293461917059,"output_mean_power":0.00012603074565240656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.8253387298902018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.359313427297327e-05,"output_power":7.240198297576834e-05,"reference_active":true,"delta_db":-0.07086838577196981},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.331803440637784e-06,"output_power":7.932139736801223e-06,"reference_active":true,"delta_db":0.34179542980062566},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.889570973194757e-06,"output_power":1.7747503585125787e-06,"reference_active":true,"delta_db":-0.2722593629217842}],"active_band_count":3,"active_band_mae_db":0.22830772616479322}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":1.167373006797295,"rms_envelope_nrmse":0.09823094188447132,"envelope_correlation":0.9939687109372869,"rms_gain_error_db":-0.08976629341969503,"reference_mean_power":0.00012945711895499175,"output_mean_power":0.000126808775324269},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.7711142240683553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.96962455750884e-05,"output_power":6.854717182454294e-05,"reference_active":true,"delta_db":-0.07219842964150183},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2400934899690537e-06,"output_power":3.4342775538647877e-06,"reference_active":true,"delta_db":0.25277849798821145},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5855555874941795e-07,"output_power":1.6349838286260574e-07,"reference_active":true,"delta_db":0.13331989053998997}],"active_band_count":3,"active_band_mae_db":0.15276560605656775}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":1.82197695757568,"rms_envelope_nrmse":0.30776458144101243,"envelope_correlation":0.9440578121622678,"rms_gain_error_db":-1.5530949856360954,"reference_mean_power":2.594581497195349e-05,"output_mean_power":1.8145035328185655e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.3055463889387911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.094870715433079e-06,"output_power":3.6650812241436012e-06,"reference_active":true,"delta_db":-1.4304956431652456},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.046923144145425e-06,"output_power":7.42069981117141e-07,"reference_active":true,"delta_db":-1.4946993730936287},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.366550307348401e-07,"output_power":3.888451079470915e-07,"reference_active":true,"delta_db":-0.5036182858720277}],"active_band_count":3,"active_band_mae_db":1.1429377673769674}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":1.168870119347469,"rms_envelope_nrmse":0.22230479246235252,"envelope_correlation":0.9809663117872041,"rms_gain_error_db":-1.553091990769822,"reference_mean_power":2.6105327616819102e-05,"output_mean_power":1.8256602038565938e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.2170827427715928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.759181059225196e-06,"output_power":3.422567213014611e-06,"reference_active":true,"delta_db":-1.4318024185221443},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.967295837314226e-07,"output_power":4.2067169358807673e-07,"reference_active":true,"delta_db":-1.5183427922456583},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240630896276417e-08,"output_power":2.6493989951183416e-08,"reference_active":true,"delta_db":-0.8748220100658379}],"active_band_count":3,"active_band_mae_db":1.2749890736112135}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.9856927394866943,"predicted_mos":2.9856927394866943,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":117600,"input_sample_rate":24000,"input_seconds":4.9,"input_float32_sha256":"113ee8f7a60c428d2dcff9921450a15de1d8a1163a1bce2b4a788b304b6af0ee","internal_samples":78400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.485936501998367},"started_at":"2026-10-13T05:10:45.806446+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004363796000689035,"native_input_samples":117360,"native_input_sha256":"8ec1d8c087721217a7fd0321c68b2a96cee8bf7d52c9f92c4b74ab313f3245d1","output_sha256":"529bee7685f71b86e853c060ff196f9a076a288f2f616ebcf16bb13bb4e71a2e","output_samples":117600,"length_delta_samples":240,"analysis_s":0.5112326349990326,"synthesis_s":0.22353072900114057,"total_s":0.7347633640001732,"rtf":0.1502583566462522,"output_peak":0.49523937702178955,"output_rms":0.04996960448425336,"finished_at":"2026-10-13T05:10:46.567801+00:00","analysis_stages":{"external_f0_s":0.06354726599965943},"evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"529bee7685f71b86e853c060ff196f9a076a288f2f616ebcf16bb13bb4e71a2e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":78240,"output_samples":78400,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.02489414135289,"output_peak":0.4905405162849529},"common_prefix":{"samples":78240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9979591836734694,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.955791440156237,"spectral_convergence":0.19406275793902072,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.849749822231211,"spectral_convergence":0.2162970181328352,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.462647337213986,"spectral_convergence":0.2510806897445453,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":4.9421400654505713e-05,"delta_db":0.05907772246214193,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0008146171416492097,"delta_db":-0.15804465566126552,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.0014687066935694065,"delta_db":-0.25095919010947354,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":5.487263596740043e-05,"delta_db":-0.20986819057371497,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":7.304631646751236e-05,"delta_db":-0.25545161566132,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":4.197244958018807e-05,"delta_db":-0.9914102570800272,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.32080193859132383,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.22784013979444254,"pitch":null,"output_rms_dbfs":-26.01602201776354,"output_peak":0.4905405162849529,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.089396199867145},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78240,"output_samples":78240,"intelligibility":{"estoi":{"value":0.9336081073639235,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051355966406601,"rms_dbfs":-25.788181877969095,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.607283545458053,"mean":-3.586322495786203e-05,"zero_energy":false},"output":{"rms":0.050026359430686375,"rms_dbfs":-26.01602201776354,"peak":0.4905405162849529,"peak_dbfs":-6.186502323385451,"crest_factor_db":19.82951969437809,"mean":-1.7573893670549034e-05,"zero_energy":false},"rms_difference_db":-0.22784013979444495},"full_scale":{"reference":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":489,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":7.014925843880822e-07,"output_mean_power":7.818841827706487e-07,"output_rms_dbfs":-61.06857572398856,"output_zero_energy":false,"output_peak":0.011173614448027303,"output_energy_fraction":7.666850879587879e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":482,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":155,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9226190476190477,"absolute_difference_cents_median":0.9550823510066736,"absolute_difference_cents_p95":7.5289375946783155,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.0002230061513489049,"log_rms_mae_db":0.9124847624608681,"rms_envelope_nrmse":0.08421934068475373,"envelope_correlation":0.994815448193652,"rms_gain_error_db":-0.2278400966560741,"reference_mean_power":0.002640134243154231,"output_mean_power":0.002505197677333145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.2388906472962606e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011616255671056531,"output_power":0.0011038606794808302,"reference_active":true,"delta_db":-0.22151898525556452},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014996209998372379,"output_power":0.0001428343852572919,"reference_active":true,"delta_db":-0.21148743430093375},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.935590189308994e-05,"output_power":3.4031438662119846e-05,"reference_active":true,"delta_db":-0.6312956021794907}],"active_band_count":3,"active_band_mae_db":0.35476734057866294}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.0001943907377270206,"log_rms_mae_db":0.5848078291910731,"rms_envelope_nrmse":0.04953850236680541,"envelope_correlation":0.9984734528680119,"rms_gain_error_db":-0.22783949689653626,"reference_mean_power":0.0026564436783668454,"output_mean_power":0.0025206738897119254},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":2.0801208532691145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010876326002892773,"output_power":0.0010366220745239112,"reference_active":true,"delta_db":-0.20861763730324362},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.148356435560037e-05,"output_power":6.539159439199829e-05,"reference_active":true,"delta_db":-0.3868427286698193},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.554313443594102e-06,"output_power":1.8509603163617642e-06,"reference_active":true,"delta_db":-1.398770813158551}],"active_band_count":3,"active_band_mae_db":0.6647437263772046}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.00021647011759154595,"log_rms_mae_db":1.0868053083214793,"rms_envelope_nrmse":0.09316896920124018,"envelope_correlation":0.9942871595484457,"rms_gain_error_db":-0.22975284004827962,"reference_mean_power":0.0021208071215793162,"output_mean_power":0.0020115271053279325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.0688695374548394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066233690351074,"output_power":0.001003829781994913,"reference_active":true,"delta_db":-0.2619232469147914},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015588791655839427,"output_power":0.00014718710087573402,"reference_active":true,"delta_db":-0.24942701599364497},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732583628298961e-05,"output_power":3.1939793062495564e-05,"reference_active":true,"delta_db":-0.6767744882306168}],"active_band_count":3,"active_band_mae_db":0.39604158371301773}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.00018533448652616975,"log_rms_mae_db":0.6108315750458325,"rms_envelope_nrmse":0.05304019562775816,"envelope_correlation":0.9983912866797652,"rms_gain_error_db":-0.22975257820111586,"reference_mean_power":0.0021339103768736657,"output_mean_power":0.0020239553038780217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.9317825081344165e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001010025588563974,"output_power":0.000952812939281914,"reference_active":true,"delta_db":-0.2532473031467259},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.797692317548564e-05,"output_power":7.058192101200042e-05,"reference_active":true,"delta_db":-0.4327262045828252},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604806628732514e-06,"output_power":1.7577760841522673e-06,"reference_active":true,"delta_db":-1.633579490492027}],"active_band_count":3,"active_band_mae_db":0.773184332740526}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":0.9539620780558017,"rms_envelope_nrmse":0.10988706464914554,"envelope_correlation":0.9929263358105732,"rms_gain_error_db":-0.24561532098198863,"reference_mean_power":0.00012866293461917059,"output_mean_power":0.00012158833761426028},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.8253387298902018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.359313427297327e-05,"output_power":6.917875900586606e-05,"reference_active":true,"delta_db":-0.2686453261374709},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.331803440637784e-06,"output_power":7.079985262267627e-06,"reference_active":true,"delta_db":-0.15178459722487497},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.889570973194757e-06,"output_power":2.3456217659941754e-06,"reference_active":true,"delta_db":0.9389477410873717}],"active_band_count":3,"active_band_mae_db":0.45312588814990584}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":0.6705503409390621,"rms_envelope_nrmse":0.06402651637038821,"envelope_correlation":0.9978646135403844,"rms_gain_error_db":-0.24561291540974886,"reference_mean_power":0.00012945711895499175,"output_mean_power":0.00012233892108372938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.7711142240683553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.96962455750884e-05,"output_power":6.529593890453718e-05,"reference_active":true,"delta_db":-0.28323212908262274},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2400934899690537e-06,"output_power":3.2612536674960104e-06,"reference_active":true,"delta_db":0.028270389393595625},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5855555874941795e-07,"output_power":1.9102079899008578e-07,"reference_active":true,"delta_db":0.8089918484953869}],"active_band_count":3,"active_band_mae_db":0.3734981223238684}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":0.9978962828241814,"rms_envelope_nrmse":0.13331764669925303,"envelope_correlation":0.9881300529499718,"rms_gain_error_db":-0.2893702149637818,"reference_mean_power":2.594581497195349e-05,"output_mean_power":2.4273382254746332e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.3055463889387911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.094870715433079e-06,"output_power":4.615005154457433e-06,"reference_active":true,"delta_db":-0.42960977648565024},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.046923144145425e-06,"output_power":9.91602598228555e-07,"reference_active":true,"delta_db":-0.2357714473286362},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.366550307348401e-07,"output_power":3.724135882493469e-07,"reference_active":true,"delta_db":-0.6911294936705623}],"active_band_count":3,"active_band_mae_db":0.45217023916161625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":0.6630732372351967,"rms_envelope_nrmse":0.06712284307416608,"envelope_correlation":0.9974002211971132,"rms_gain_error_db":-0.2893527167907722,"reference_mean_power":2.6105327616819102e-05,"output_mean_power":2.4422711328151534e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.2170827427715928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.759181059225196e-06,"output_power":4.324865362823831e-06,"reference_active":true,"delta_db":-0.4155963548760488},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.967295837314226e-07,"output_power":5.520115403166374e-07,"reference_active":true,"delta_db":-0.33829411998615966},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240630896276417e-08,"output_power":2.4631013446617423e-08,"reference_active":true,"delta_db":-1.1914728694564574}],"active_band_count":3,"active_band_mae_db":0.6484544481062219}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4507-16021-0045","method":"magphase","status":"measured","mos":{"status":"measured","value":2.583038806915283,"predicted_mos":2.583038806915283,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":235267,"input_sample_rate":48000,"input_seconds":4.901395833333333,"input_float32_sha256":"82e2bbee0adaf3f1edffd85726ee8797e121fe68600a57a5ebab7f4150f4f653","internal_samples":78423,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.996102917999451},"started_at":"2026-10-13T05:15:41.675313+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008853675000864314,"native_input_samples":234720,"native_input_sha256":"a6400a31d2c6a0cdbaea58641d7d62fd7b476a2784dbd4f4a5877248bcae824b","output_sha256":"cd08449f7593552cb5fef507d231bbde13ddd1624917698bb3170938c03fcbce","output_samples":235267,"length_delta_samples":547,"analysis_s":3.4569187170000077,"synthesis_s":1.0417314799997257,"total_s":4.4986501969997335,"rtf":0.9199693654396184,"output_peak":0.5687428083870851,"output_rms":0.04922482696138522,"finished_at":"2026-10-13T05:15:46.194030+00:00","evaluation_input":{"reference":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"output_sha256":"cd08449f7593552cb5fef507d231bbde13ddd1624917698bb3170938c03fcbce","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":78240,"output_samples":78423,"length_delta_samples":183,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.155853121665913,"output_peak":0.5504087406253474},"common_prefix":{"samples":78240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.997666500898971,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":78240,"output_samples":78240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.963367299167478,"spectral_convergence":0.32705493291100407,"spectral_convergence_status":"ok","amplitude_floor":1.658911626731477e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.653242841149958,"spectral_convergence":0.1686921058653943,"spectral_convergence_status":"ok","amplitude_floor":1.4097603217110054e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.891635707221409,"spectral_convergence":0.2117753843644361,"spectral_convergence_status":"ok","amplitude_floor":8.986287240444542e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.8753665955686144e-05,"output_power":4.583612149910265e-05,"delta_db":-0.2679941540459168,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00084480798498807,"output_power":0.0007830420657879431,"delta_db":-0.32972916679895165,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015560766761226633,"output_power":0.0014025997514418714,"delta_db":-0.4509723556518741,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.758941174067265e-05,"output_power":6.204939944771635e-05,"delta_db":0.3239494012679168,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.7471770650556e-05,"output_power":8.603268746509548e-05,"delta_db":0.45520007415635666,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":5.273577609828158e-05,"output_power":4.944913251363039e-05,"delta_db":-0.2794666461918693,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.35121863301881423,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3575265996327931,"pitch":null,"output_rms_dbfs":-26.145708477601886,"output_peak":0.5504087406253474,"reference_rms_dbfs":-25.788181877969095,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.836081949179615},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":78240,"output_samples":78240,"intelligibility":{"estoi":{"value":0.9234929838686744,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.051355966406601,"rms_dbfs":-25.788181877969095,"peak":0.550750732421875,"peak_dbfs":-5.180898332511042,"crest_factor_db":20.607283545458053,"mean":-3.586322495786203e-05,"zero_energy":false},"output":{"rms":0.04928497903170254,"rms_dbfs":-26.145708477601886,"peak":0.5504087406253474,"peak_dbfs":-5.186293560304978,"crest_factor_db":20.95941491729691,"mean":-0.0005406318287694112,"zero_energy":false},"rms_difference_db":-0.35752659963279143},"full_scale":{"reference":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":78240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":489,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.973174353945069e-06,"threshold_dbfs":-53.033663146208056,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":120,"quiet_samples":19200,"reference_mean_power":7.014925843880822e-07,"output_mean_power":6.566946297128391e-07,"output_rms_dbfs":-61.82636535251338,"output_zero_energy":false,"output_peak":0.006694800941207395,"output_energy_fraction":6.634477100680472e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":482,"reference_estimated_voiced_frames":168,"output_estimated_voiced_frames":151,"both_estimated_voiced_frames":148,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.8809523809523809,"absolute_difference_cents_median":4.061723894529012,"absolute_difference_cents_p95":12.775993512199603,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":78240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.0002230061513489049,"log_rms_mae_db":1.3850739660666183,"rms_envelope_nrmse":0.07781188545154444,"envelope_correlation":0.9964594971494339,"rms_gain_error_db":-0.3575265624190269,"reference_mean_power":0.002640134243154231,"output_mean_power":0.0024314948485950477},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.2388906472962606e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011616255671056531,"output_power":0.0010863052064631932,"reference_active":true,"delta_db":-0.2911430114457305},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014996209998372379,"output_power":0.00015038229931687176,"reference_active":true,"delta_db":0.012152074164013892},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.935590189308994e-05,"output_power":2.8338251158554693e-05,"reference_active":true,"delta_db":-1.4263682420694057}],"active_band_count":3,"active_band_mae_db":0.5765544425597167}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.0001943907377270206,"log_rms_mae_db":1.0210162711393305,"rms_envelope_nrmse":0.056427842499729354,"envelope_correlation":0.9987180081944366,"rms_gain_error_db":-0.35752766877329534,"reference_mean_power":0.0026564436783668454,"output_mean_power":0.0024465147902359334},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":2.0801208532691145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010876326002892773,"output_power":0.0010194528000035303,"reference_active":true,"delta_db":-0.2811509345976753},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.148356435560037e-05,"output_power":7.1414484979916e-05,"reference_active":true,"delta_db":-0.0041989088233930485},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.554313443594102e-06,"output_power":1.948168343201042e-06,"reference_active":true,"delta_db":-1.1764770712755452}],"active_band_count":3,"active_band_mae_db":0.48727563823220454}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":0.00021647011759154595,"log_rms_mae_db":1.7611291711173291,"rms_envelope_nrmse":0.07869359930729461,"envelope_correlation":0.9972357194834679,"rms_gain_error_db":-0.43515921128718504,"reference_mean_power":0.0021208071215793162,"output_mean_power":0.0019186035863556254},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":2.0688695374548394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001066233690351074,"output_power":0.0009642941515546809,"reference_active":true,"delta_db":-0.4364286826588628},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015588791655839427,"output_power":0.00015105112268185984,"reference_active":true,"delta_db":-0.13688495243539855},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732583628298961e-05,"output_power":2.9673362491052022e-05,"reference_active":true,"delta_db":-0.9964278487167864}],"active_band_count":3,"active_band_mae_db":0.5232471612703492}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":0.00018533448652616975,"log_rms_mae_db":1.2808841867836536,"rms_envelope_nrmse":0.06185193643775445,"envelope_correlation":0.998903270167257,"rms_gain_error_db":-0.43515932727085993,"reference_mean_power":0.0021339103768736657,"output_mean_power":0.0019304574899973832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.9317825081344165e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001010025588563974,"output_power":0.000915962985702236,"reference_active":true,"delta_db":-0.4245445251426175},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.797692317548564e-05,"output_power":7.514912400554464e-05,"reference_active":true,"delta_db":-0.1604217224588386},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604806628732514e-06,"output_power":2.0546450385242082e-06,"reference_active":true,"delta_db":-0.9558469656044697}],"active_band_count":3,"active_band_mae_db":0.5136044044019753}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":6.255230128136309e-05,"log_rms_mae_db":1.293605225846539,"rms_envelope_nrmse":0.09236973430503838,"envelope_correlation":0.9959921894487558,"rms_gain_error_db":0.3406904903567031,"reference_mean_power":0.00012866293461917059,"output_mean_power":0.00013916258965028657},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.8253387298902018e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.359313427297327e-05,"output_power":8.009494720734608e-05,"reference_active":true,"delta_db":0.36767819889213493},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.331803440637784e-06,"output_power":7.833042671214421e-06,"reference_active":true,"delta_db":0.2871967902420032},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.889570973194757e-06,"output_power":2.259659531709852e-06,"reference_active":true,"delta_db":0.7767979902233495}],"active_band_count":3,"active_band_mae_db":0.47722432645249585}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":6.0793288183779715e-05,"log_rms_mae_db":0.7567197866169643,"rms_envelope_nrmse":0.05554334375578975,"envelope_correlation":0.9991482843837567,"rms_gain_error_db":0.3406902638205548,"reference_mean_power":0.00012945711895499175,"output_mean_power":0.00014002157681261782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.7711142240683553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.96962455750884e-05,"output_power":7.566205713504182e-05,"reference_active":true,"delta_db":0.3566876093545101},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2400934899690537e-06,"output_power":3.4027422670272996e-06,"reference_active":true,"delta_db":0.2127151413174577},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5855555874941795e-07,"output_power":1.9416595194283926e-07,"reference_active":true,"delta_db":0.8799160392214957}],"active_band_count":3,"active_band_mae_db":0.48310626329782114}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":977,"rms_floor":3.76520185695864e-05,"log_rms_mae_db":1.64074181250734,"rms_envelope_nrmse":0.2767767451612274,"envelope_correlation":0.9580746908507555,"rms_gain_error_db":0.7913277425703741,"reference_mean_power":2.594581497195349e-05,"output_mean_power":3.1131503172216015e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":977,"frequency_resolution_hz":0.2047082906857728,"power_floor":1.3055463889387911e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.094870715433079e-06,"output_power":7.225744899558873e-06,"reference_active":true,"delta_db":1.5174945693504611},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.046923144145425e-06,"output_power":1.8837148006618643e-06,"reference_active":true,"delta_db":2.551003493334065},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.366550307348401e-07,"output_power":7.86817696938697e-07,"reference_active":true,"delta_db":2.55735651355498}],"active_band_count":3,"active_band_mae_db":2.2086181920798356}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":78240,"uncovered_tail_samples":0,"metrics":{"frame_count":971,"rms_floor":2.0830708932641746e-05,"log_rms_mae_db":1.216028490765157,"rms_envelope_nrmse":0.19549070640820207,"envelope_correlation":0.9806561237344795,"rms_gain_error_db":0.7912785074920781,"reference_mean_power":2.6105327616819102e-05,"output_mean_power":3.1322541884046474e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":971,"frequency_resolution_hz":0.20597322348094746,"power_floor":1.2170827427715928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.759181059225196e-06,"output_power":7.0581127865442956e-06,"reference_active":true,"delta_db":1.7115636647793242},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.967295837314226e-07,"output_power":1.0457505619587057e-06,"reference_active":true,"delta_db":2.436505374767572},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240630896276417e-08,"output_power":5.1263892918453184e-08,"reference_active":true,"delta_db":1.9918201464005763}],"active_band_count":3,"active_band_mae_db":2.0466297286491577}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"hnm","status":"measured","mos":{"status":"measured","value":2.199205160140991,"predicted_mos":2.199205160140991,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"7336b31c400435c83ad5ef521231bd7e3a11795e97937d845e763505dcbd17c5","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6190079409989266},"started_at":"2026-10-13T04:39:20.552167+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4439000551647041e-05,"native_input_samples":60000,"native_input_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","output_sha256":"836f1e09414505fcc7ea3be60e03109e56f92fdfd50cead065343fc99e423d28","output_samples":60000,"length_delta_samples":0,"analysis_s":0.6317786899999192,"synthesis_s":0.3136632779996944,"total_s":0.9454419679996136,"rtf":0.2521178581332303,"output_peak":0.6435907585036994,"output_rms":0.05457911509326682,"finished_at":"2026-10-13T04:39:21.501776+00:00","evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"836f1e09414505fcc7ea3be60e03109e56f92fdfd50cead065343fc99e423d28","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.4058214513356,"spectral_convergence":0.3660765759554478,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.975221714884022,"spectral_convergence":0.3771500729794749,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.990634221730427,"spectral_convergence":0.3874372454306109,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":6.214705349404063e-05,"delta_db":0.37064545719684605,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.000648646264914776,"delta_db":-0.055840293087520115,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0009569378119443577,"delta_db":-0.47675193565602286,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0006861038344332187,"delta_db":-0.15309417469414233,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.0005372489609216151,"delta_db":0.15182516542286784,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":8.779587865605772e-05,"delta_db":0.2454095525337765,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24226109643186264,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1654568870436282,"pitch":null,"output_rms_dbfs":-25.25947019898264,"output_peak":0.6435907585036994,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.123892462650017},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.4058214513356,"spectral_convergence":0.3660765759554478,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.975221714884022,"spectral_convergence":0.3771500729794749,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.990634221730427,"spectral_convergence":0.3874372454306109,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":6.214705349404063e-05,"delta_db":0.37064545719684605,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.000648646264914776,"delta_db":-0.055840293087520115,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0009569378119443577,"delta_db":-0.47675193565602286,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0006861038344332187,"delta_db":-0.15309417469414233,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.0005372489609216151,"delta_db":0.15182516542286784,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":8.779587865605772e-05,"delta_db":0.2454095525337765,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24226109643186264,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1654568870436282,"pitch":null,"output_rms_dbfs":-25.25947019898264,"output_peak":0.6435907585036994,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.123892462650017},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9084327652184436,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055628754242498336,"rms_dbfs":-25.09401331193901,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.392072346291222,"mean":0.00027566782633463543,"zero_energy":false},"output":{"rms":0.05457911509326682,"rms_dbfs":-25.25947019898264,"peak":0.6435907585036994,"peak_dbfs":-3.8278040136373965,"crest_factor_db":21.43166618534524,"mean":-0.00010765182455210078,"zero_energy":false},"rms_difference_db":-0.16545688704363215},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":136,"quiet_samples":21760,"reference_mean_power":2.905563365750234e-07,"output_mean_power":5.344267434579729e-07,"output_rms_dbfs":-62.72111817288469,"output_zero_energy":false,"output_peak":0.012798057332481677,"output_energy_fraction":6.506431221007302e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":96,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.8073394495412844,"absolute_difference_cents_median":2.095804641316173,"absolute_difference_cents_p95":9.047706553954297,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.0002227499093660225,"log_rms_mae_db":0.7742494258545011,"rms_envelope_nrmse":0.09143735146451072,"envelope_correlation":0.9938774949926595,"rms_gain_error_db":-0.16545688457974214,"reference_mean_power":0.003098689875324867,"output_mean_power":0.0029828569392562954},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.065525005901086e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018487969313489734,"output_power":0.0017817539453983916,"reference_active":true,"delta_db":-0.16041482482364305},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000254937700664643,"output_power":0.00022129385317867843,"reference_active":true,"delta_db":-0.6146471368903457},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5927620235575996e-05,"output_power":4.016369845612665e-05,"reference_active":true,"delta_db":0.4840524631895144}],"active_band_count":3,"active_band_mae_db":0.4197048083011677}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00020550660315895724,"log_rms_mae_db":0.44676136014550344,"rms_envelope_nrmse":0.05854409171724588,"envelope_correlation":0.9975360207059404,"rms_gain_error_db":-0.16545689199642152,"reference_mean_power":0.003123712899469445,"output_mean_power":0.003006944566060466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":2.8505515590137105e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017535539962877896,"output_power":0.0016926158950328323,"reference_active":true,"delta_db":-0.1536072849564502},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011841366649455359,"output_power":0.00010346168592468721,"reference_active":true,"delta_db":-0.5862227754622228},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.758270588195691e-06,"output_power":2.4731594269689272e-06,"reference_active":true,"delta_db":1.4816640171608917}],"active_band_count":3,"active_band_mae_db":0.7404980258598549}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00016827060195170467,"log_rms_mae_db":1.0487026903552294,"rms_envelope_nrmse":0.16465747113543483,"envelope_correlation":0.9810892579624139,"rms_gain_error_db":-0.31622067980947327,"reference_mean_power":0.001201501254742638,"output_mean_power":0.001117126002559055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.2448894715706734e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007295274990436896,"output_power":0.0006883643257256576,"reference_active":true,"delta_db":-0.2522331191591359},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014476368795830427,"output_power":0.0001223755237821582,"reference_active":true,"delta_db":-0.7296507489059365},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.785748904828138e-05,"output_power":2.389513872333218e-05,"reference_active":true,"delta_db":-0.6663241229212423}],"active_band_count":3,"active_band_mae_db":0.5494026636621049}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012583494842815252,"log_rms_mae_db":0.7502589279587839,"rms_envelope_nrmse":0.1079686355324158,"envelope_correlation":0.99198705904595,"rms_gain_error_db":-0.3162206798422118,"reference_mean_power":0.0012112038213410922,"output_mean_power":0.0011261472078101607},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1451125517965788e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000709760959532951,"output_power":0.0006618985332824995,"reference_active":true,"delta_db":-0.3032068853522993},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.709378020571589e-05,"output_power":5.520007290565258e-05,"reference_active":true,"delta_db":-0.8474261031113985},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7774397828761626e-06,"output_power":1.7098507692648805e-06,"reference_active":true,"delta_db":-0.16836688269731898}],"active_band_count":3,"active_band_mae_db":0.439666623720339}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015339527329010622,"log_rms_mae_db":1.2203127488518664,"rms_envelope_nrmse":0.09989224945069441,"envelope_correlation":0.9939672329665835,"rms_gain_error_db":-0.09198952717511295,"reference_mean_power":0.0009604249251386083,"output_mean_power":0.0009402957425203816},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.606398277844274e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009968434213476986,"output_power":0.0009642952589992477,"reference_active":true,"delta_db":-0.14416915554524953},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073134134931984,"output_power":9.954149727542207e-05,"reference_active":true,"delta_db":-0.462663918003198},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0381145318442426e-05,"output_power":1.9687844902182295e-05,"reference_active":true,"delta_db":2.7795290893103486}],"active_band_count":3,"active_band_mae_db":1.1287873876195986}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00014456690433921235,"log_rms_mae_db":0.7427471903923811,"rms_envelope_nrmse":0.05603013144596249,"envelope_correlation":0.9980917266467768,"rms_gain_error_db":-0.0919895440666474,"reference_mean_power":0.0009681806844812575,"output_mean_power":0.0009478889481507106},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.554834929605817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000971830547751713,"output_power":0.000945147336916114,"reference_active":true,"delta_db":-0.12091031222797297},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3135698812325164e-05,"output_power":4.869470590540528e-05,"reference_active":true,"delta_db":-0.37904649331318996},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.152693877304538e-07,"output_power":1.187572663125131e-06,"reference_active":true,"delta_db":2.855948843358007}],"active_band_count":3,"active_band_mae_db":1.1186352162997233}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":1.2611356548813661,"rms_envelope_nrmse":0.2105055474899255,"envelope_correlation":0.9660489843220204,"rms_gain_error_db":0.1729902627887943,"reference_mean_power":0.00013836378417930354,"output_mean_power":0.00014398639471531339},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":9.868150581916699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.732185415785842e-05,"output_power":5.8918024978353574e-05,"reference_active":true,"delta_db":0.11927950371742067},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6922402540887428e-05,"output_power":1.5806201243125677e-05,"reference_active":true,"delta_db":-0.29634514561180997},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2492402977589626e-06,"output_power":6.389790075284747e-06,"reference_active":true,"delta_db":1.7717529898609574}],"active_band_count":3,"active_band_mae_db":0.7291258797300627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":0.7326800232637382,"rms_envelope_nrmse":0.11775427040018205,"envelope_correlation":0.9885519844683909,"rms_gain_error_db":0.17299028381238662,"reference_mean_power":0.00013948111618043135,"output_mean_power":0.00014514913180520696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":8.397728752995273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.502560023531222e-05,"output_power":5.7776224890271096e-05,"reference_active":true,"delta_db":0.21184372887391828},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.667910609926142e-06,"output_power":7.62393043692155e-06,"reference_active":true,"delta_db":-0.024981163556235014},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.289881205259684e-07,"output_power":3.938556334861563e-07,"reference_active":true,"delta_db":2.3552410920867066}],"active_band_count":3,"active_band_mae_db":0.8640219948389533}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.651090621948242,"predicted_mos":2.651090621948242,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"0145a87b6596340891dc8aba9ea300d08ed73a63ebbfea9b81d2ef1f84fd8f52","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6764033929994184},"started_at":"2026-10-13T04:42:17.170225+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.7098000171245076e-05,"native_input_samples":60000,"native_input_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","output_sha256":"3b210f045acec139c20bd89b0919813ff716b2dcde8df250fa8c8701e17d984b","output_samples":60000,"length_delta_samples":0,"analysis_s":0.9323785319993476,"synthesis_s":0.06869573000039964,"total_s":1.0010742619997473,"rtf":0.2669531365332659,"output_peak":0.8382972587333608,"output_rms":0.061599296951705386,"finished_at":"2026-10-13T04:42:18.175057+00:00","evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"3b210f045acec139c20bd89b0919813ff716b2dcde8df250fa8c8701e17d984b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.683842363785411,"spectral_convergence":0.36467532983496886,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.885957824650545,"spectral_convergence":0.27795593791317535,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.032913268873042,"spectral_convergence":0.3352930559807363,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":6.555131565626817e-05,"delta_db":0.6022541563089452,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.0008122442511318932,"delta_db":0.9209469524510244,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0014544207930701544,"delta_db":1.341311658227766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0008124028765765662,"delta_db":0.5807218854916542,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.000569655593439457,"delta_db":0.4061929840236727,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":8.019855507004176e-05,"delta_db":-0.14766630621112264,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6665156571190309,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8855284215465906,"pitch":null,"output_rms_dbfs":-24.20848489039242,"output_peak":0.8382972587333608,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.200904485769666},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.683842363785411,"spectral_convergence":0.36467532983496886,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.885957824650545,"spectral_convergence":0.27795593791317535,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.032913268873042,"spectral_convergence":0.3352930559807363,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":6.555131565626817e-05,"delta_db":0.6022541563089452,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.0008122442511318932,"delta_db":0.9209469524510244,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0014544207930701544,"delta_db":1.341311658227766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0008124028765765662,"delta_db":0.5807218854916542,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.000569655593439457,"delta_db":0.4061929840236727,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":8.019855507004176e-05,"delta_db":-0.14766630621112264,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6665156571190309,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8855284215465906,"pitch":null,"output_rms_dbfs":-24.20848489039242,"output_peak":0.8382972587333608,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.200904485769666},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9356695989169734,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055628754242498336,"rms_dbfs":-25.09401331193901,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.392072346291222,"mean":0.00027566782633463543,"zero_energy":false},"output":{"rms":0.06159929695170539,"rms_dbfs":-24.20848489039242,"peak":0.8382972587333608,"peak_dbfs":-1.5320390799968822,"crest_factor_db":22.676445810395535,"mean":5.443934305201144e-11,"zero_energy":false},"rms_difference_db":0.8855284215465886},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":136,"quiet_samples":21760,"reference_mean_power":2.905563365750234e-07,"output_mean_power":8.515224324408867e-07,"output_rms_dbfs":-60.698039065305444,"output_zero_energy":false,"output_peak":0.01701815351479736,"output_energy_fraction":8.13864720702883e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":122,"both_estimated_voiced_frames":91,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.8348623853211009,"absolute_difference_cents_median":4.909088418630734,"absolute_difference_cents_p95":18.541385773841135,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.0002227499093660225,"log_rms_mae_db":1.2666993589012818,"rms_envelope_nrmse":0.13521062347876148,"envelope_correlation":0.9954036889906672,"rms_gain_error_db":0.885528433667657,"reference_mean_power":0.003098689875324867,"output_mean_power":0.0037995394361636545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.065525005901086e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018487969313489734,"output_power":0.002277069589208107,"reference_active":true,"delta_db":0.9048709160595205},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000254937700664643,"output_power":0.00030807459680424375,"reference_active":true,"delta_db":0.8222182426628981},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5927620235575996e-05,"output_power":4.232805475212931e-05,"reference_active":true,"delta_db":0.7119985882071198}],"active_band_count":3,"active_band_mae_db":0.8130292489765129}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00020550660315895724,"log_rms_mae_db":0.9281486361251701,"rms_envelope_nrmse":0.12041257819645701,"envelope_correlation":0.9978647758207915,"rms_gain_error_db":0.8855284574712572,"reference_mean_power":0.003123712899469445,"output_mean_power":0.003830222091068381},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":2.8505515590137105e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017535539962877896,"output_power":0.0021687178138604557,"reference_active":true,"delta_db":0.9228390334822824},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011841366649455359,"output_power":0.00014733729003381463,"reference_active":true,"delta_db":0.9491084906438735},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.758270588195691e-06,"output_power":2.4657034577092487e-06,"reference_active":true,"delta_db":1.468551327692674}],"active_band_count":3,"active_band_mae_db":1.1134996172729432}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00016827060195170467,"log_rms_mae_db":1.2914135428183464,"rms_envelope_nrmse":0.199119345137928,"envelope_correlation":0.9919346631882572,"rms_gain_error_db":1.3050368532953556,"reference_mean_power":0.001201501254742638,"output_mean_power":0.001622661432716259},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.2448894715706734e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007295274990436896,"output_power":0.0010153766692098107,"reference_active":true,"delta_db":1.4358551321981683},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014476368795830427,"output_power":0.00020752121043115298,"reference_active":true,"delta_db":1.5640285336416868},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.785748904828138e-05,"output_power":3.683536267021034e-05,"reference_active":true,"delta_db":1.213229816936693}],"active_band_count":3,"active_band_mae_db":1.4043711609255158}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012583494842815252,"log_rms_mae_db":1.0721972660133334,"rms_envelope_nrmse":0.17683068458320475,"envelope_correlation":0.9968138139763727,"rms_gain_error_db":1.3050368545033866,"reference_mean_power":0.0012112038213410922,"output_mean_power":0.0016357650238296385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1451125517965788e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000709760959532951,"output_power":0.0009909176726339898,"reference_active":true,"delta_db":1.4492546669399784},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.709378020571589e-05,"output_power":9.824883188868e-05,"reference_active":true,"delta_db":1.6564513397783576},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7774397828761626e-06,"output_power":2.453365805445132e-06,"reference_active":true,"delta_db":1.3996741149107677}],"active_band_count":3,"active_band_mae_db":1.501793373876368}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015339527329010622,"log_rms_mae_db":1.1373444609620655,"rms_envelope_nrmse":0.07446878199463378,"envelope_correlation":0.9981125841639139,"rms_gain_error_db":0.40197106548249784,"reference_mean_power":0.0009604249251386083,"output_mean_power":0.0010535630453890415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.606398277844274e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009968434213476986,"output_power":0.0010873244788190618,"reference_active":true,"delta_db":0.3773221827389297},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073134134931984,"output_power":0.00011569174827641109,"reference_active":true,"delta_db":0.19031823172412982},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0381145318442426e-05,"output_power":1.035580859705581e-05,"reference_active":true,"delta_db":-0.010612555545118845}],"active_band_count":3,"active_band_mae_db":0.19275099000272614}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00014456690433921235,"log_rms_mae_db":0.7448644285371208,"rms_envelope_nrmse":0.06210695533669286,"envelope_correlation":0.9990577061483513,"rms_gain_error_db":0.40197108442468055,"reference_mean_power":0.0009681806844812575,"output_mean_power":0.0010620709315001494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.554834929605817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000971830547751713,"output_power":0.0010582676561188998,"reference_active":true,"delta_db":0.3700497675765416},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3135698812325164e-05,"output_power":5.547991717185574e-05,"reference_active":true,"delta_db":0.18749407465093437},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.152693877304538e-07,"output_power":5.792275674841531e-07,"reference_active":true,"delta_db":-0.2621608433939942}],"active_band_count":3,"active_band_mae_db":0.27323489520715677}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":1.386873984904999,"rms_envelope_nrmse":0.186283193550236,"envelope_correlation":0.9744655031076345,"rms_gain_error_db":0.3094443266809011,"reference_mean_power":0.00013836378417930354,"output_mean_power":0.00014858222716918702},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":9.868150581916699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.732185415785842e-05,"output_power":5.997685152782798e-05,"reference_active":true,"delta_db":0.19663433992169913},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6922402540887428e-05,"output_power":1.4944815434298777e-05,"reference_active":true,"delta_db":-0.5397146564563481},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2492402977589626e-06,"output_power":3.44159782693553e-06,"reference_active":true,"delta_db":-0.9155117264109935}],"active_band_count":3,"active_band_mae_db":0.5506202409296802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":0.857100575578592,"rms_envelope_nrmse":0.1137099016000962,"envelope_correlation":0.9902006454685661,"rms_gain_error_db":0.3094444010926645,"reference_mean_power":0.00013948111618043135,"output_mean_power":0.0001497820789460852},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":8.397728752995273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.502560023531222e-05,"output_power":5.707897703694611e-05,"reference_active":true,"delta_db":0.15911392472910788},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.667910609926142e-06,"output_power":6.504185575786419e-06,"reference_active":true,"delta_db":-0.7148411742430723},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.289881205259684e-07,"output_power":1.833228137087212e-07,"reference_active":true,"delta_db":-0.965964381995903}],"active_band_count":3,"active_band_mae_db":0.6133064936560277}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"facodec","status":"measured","mos":{"status":"measured","value":3.4240527153015137,"predicted_mos":3.4240527153015137,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"b0ea40c844b5c8fcc57f47bc1b882f22d93a3334b5cf03c911f018beefe71c36","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4752180769992265},"started_at":"2026-10-13T04:53:01.699474+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.7375000324100256e-05,"native_input_samples":60000,"native_input_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","output_sha256":"5795daa9a3b577eedf2c8504258cab9c1244830eaf7ffc8862582180b8d3b623","output_samples":60000,"length_delta_samples":0,"analysis_s":0.9258452390004095,"synthesis_s":2.628769784999349,"total_s":3.5546150239997587,"rtf":0.947897339733269,"output_peak":0.7180344462394714,"output_rms":0.04885086822820203,"finished_at":"2026-10-13T04:53:05.259054+00:00","evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"5795daa9a3b577eedf2c8504258cab9c1244830eaf7ffc8862582180b8d3b623","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.143843251068722,"spectral_convergence":0.2734748447949997,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.1795469390425595,"spectral_convergence":0.29722004579708283,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.141297986130468,"spectral_convergence":0.32180604419842573,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":5.422363270244873e-05,"delta_db":-0.22167387347711617,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.0004142058862076331,"delta_db":-2.0037568466717612,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0008400865896402935,"delta_db":-1.0423485688214065,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0005593271969732462,"delta_db":-1.0403332677833645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.00044807909948083605,"delta_db":-0.6363838081924488,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":7.048492164470116e-05,"delta_db":-0.7083695304129333,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9421443158931716,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.128540950970744,"pitch":null,"output_rms_dbfs":-26.222554262909753,"output_peak":0.7180344462394714,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.1548960587472505},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.143843251068722,"spectral_convergence":0.2734748447949997,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.1795469390425595,"spectral_convergence":0.29722004579708283,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.141297986130468,"spectral_convergence":0.32180604419842573,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":5.422363270244873e-05,"delta_db":-0.22167387347711617,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.0004142058862076331,"delta_db":-2.0037568466717612,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0008400865896402935,"delta_db":-1.0423485688214065,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0005593271969732462,"delta_db":-1.0403332677833645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.00044807909948083605,"delta_db":-0.6363838081924488,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":7.048492164470116e-05,"delta_db":-0.7083695304129333,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9421443158931716,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.128540950970744,"pitch":null,"output_rms_dbfs":-26.222554262909753,"output_peak":0.7180344462394714,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.1548960587472505},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9301549249557811,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055628754242498336,"rms_dbfs":-25.09401331193901,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.392072346291222,"mean":0.00027566782633463543,"zero_energy":false},"output":{"rms":0.048850868228202025,"rms_dbfs":-26.222554262909753,"peak":0.7180344462394714,"peak_dbfs":-2.8770944173020214,"crest_factor_db":23.34545984560773,"mean":2.750189472656596e-05,"zero_energy":false},"rms_difference_db":-1.128540950970745},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":136,"quiet_samples":21760,"reference_mean_power":2.905563365750234e-07,"output_mean_power":3.431004007043243e-07,"output_rms_dbfs":-64.64578774734338,"output_zero_energy":false,"output_peak":0.006295692641288042,"output_energy_fraction":5.214159262163898e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":120,"both_estimated_voiced_frames":104,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9541284403669725,"absolute_difference_cents_median":4.185939090993655,"absolute_difference_cents_p95":15.64312103719992,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.0002227499093660225,"log_rms_mae_db":1.141112526870123,"rms_envelope_nrmse":0.15335971033313642,"envelope_correlation":0.9926163335457677,"rms_gain_error_db":-1.1285409647653986,"reference_mean_power":0.003098689875324867,"output_mean_power":0.002389593436133466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.065525005901086e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018487969313489734,"output_power":0.001425314591154221,"reference_active":true,"delta_db":-1.129784805276271},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000254937700664643,"output_power":0.00020908777269141023,"reference_active":true,"delta_db":-0.8610542818111526},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5927620235575996e-05,"output_power":3.227493824099493e-05,"reference_active":true,"delta_db":-0.4656303156716782}],"active_band_count":3,"active_band_mae_db":0.818823134253034}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00020550660315895724,"log_rms_mae_db":0.9451039057474901,"rms_envelope_nrmse":0.14539126057755591,"envelope_correlation":0.9943790913867976,"rms_gain_error_db":-1.1285410843513475,"reference_mean_power":0.003123712899469445,"output_mean_power":0.0024088901877015374},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":2.8505515590137105e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017535539962877896,"output_power":0.00135310183312182,"reference_active":true,"delta_db":-1.1258866102445826},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011841366649455359,"output_power":9.603121248379968e-05,"reference_active":true,"delta_db":-0.9098941630275541},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.758270588195691e-06,"output_power":1.6562979010444345e-06,"reference_active":true,"delta_db":-0.25947259921525795}],"active_band_count":3,"active_band_mae_db":0.7650844574957983}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00016827060195170467,"log_rms_mae_db":1.1225255152925473,"rms_envelope_nrmse":0.15860725049880334,"envelope_correlation":0.990063306241769,"rms_gain_error_db":-1.0362807296945389,"reference_mean_power":0.001201501254742638,"output_mean_power":0.0009464466878113246},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.2448894715706734e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007295274990436896,"output_power":0.0005774680201448615,"reference_active":true,"delta_db":-1.0151372872835476},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014476368795830427,"output_power":0.00012764061141920272,"reference_active":true,"delta_db":-0.5467076267171815},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.785748904828138e-05,"output_power":2.2730017446145397e-05,"reference_active":true,"delta_db":-0.8834219939227829}],"active_band_count":3,"active_band_mae_db":0.8150889693078374}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012583494842815252,"log_rms_mae_db":0.9622823070025778,"rms_envelope_nrmse":0.14353775687529524,"envelope_correlation":0.993303125862762,"rms_gain_error_db":-1.03628074885212,"reference_mean_power":0.0012112038213410922,"output_mean_power":0.0009540895903279874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1451125517965788e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000709760959532951,"output_power":0.000565966678827524,"reference_active":true,"delta_db":-0.9832124431838689},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.709378020571589e-05,"output_power":5.911101980926923e-05,"reference_active":true,"delta_db":-0.5501380958376354},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7774397828761626e-06,"output_power":1.5381167433833432e-06,"reference_active":true,"delta_db":-0.6280559661495626}],"active_band_count":3,"active_band_mae_db":0.7204688350570224}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015339527329010622,"log_rms_mae_db":1.2352197420676676,"rms_envelope_nrmse":0.13561043318588498,"envelope_correlation":0.994230650483991,"rms_gain_error_db":-0.9049010698175355,"reference_mean_power":0.0009604249251386083,"output_mean_power":0.0007797821965384086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.606398277844274e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009968434213476986,"output_power":0.0007815137564267707,"reference_active":true,"delta_db":-1.0569032009585135},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073134134931984,"output_power":9.995686185051002e-05,"reference_active":true,"delta_db":-0.4445794781723305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0381145318442426e-05,"output_power":9.170792231991831e-06,"reference_active":true,"delta_db":-0.5383841603195737}],"active_band_count":3,"active_band_mae_db":0.6799556131501392}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00014456690433921235,"log_rms_mae_db":0.9985445875901893,"rms_envelope_nrmse":0.12084897226929413,"envelope_correlation":0.9966829674701367,"rms_gain_error_db":-0.9049012686306475,"reference_mean_power":0.0009681806844812575,"output_mean_power":0.0007860791681562919},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.554834929605817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000971830547751713,"output_power":0.0007640669607882958,"reference_active":true,"delta_db":-1.0445912555798984},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3135698812325164e-05,"output_power":4.779968731071393e-05,"reference_active":true,"delta_db":-0.45961340946652723},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.152693877304538e-07,"output_power":5.548634974238901e-07,"reference_active":true,"delta_db":-0.4487915262557708}],"active_band_count":3,"active_band_mae_db":0.6509987304340655}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":1.501987029762528,"rms_envelope_nrmse":0.20308081706265724,"envelope_correlation":0.9679799374216309,"rms_gain_error_db":-0.48778122910932636,"reference_mean_power":0.00013836378417930354,"output_mean_power":0.0001236642899790889},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":9.868150581916699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.732185415785842e-05,"output_power":4.9340881855808204e-05,"reference_active":true,"delta_db":-0.651133225853479},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6922402540887428e-05,"output_power":1.328899878965909e-05,"reference_active":true,"delta_db":-1.0496975972987794},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2492402977589626e-06,"output_power":4.344623237158339e-06,"reference_active":true,"delta_db":0.09640829222515315}],"active_band_count":3,"active_band_mae_db":0.5990797051258039}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":1.1555291138793233,"rms_envelope_nrmse":0.1605145598152331,"envelope_correlation":0.978873996730937,"rms_gain_error_db":-0.48778197743829765,"reference_mean_power":0.00013948111618043135,"output_mean_power":0.00012466289736236507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":8.397728752995273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.502560023531222e-05,"output_power":4.648544517190919e-05,"reference_active":true,"delta_db":-0.7324779428846429},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.667910609926142e-06,"output_power":6.3849820855286205e-06,"reference_active":true,"delta_db":-0.7951735827551645},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.289881205259684e-07,"output_power":2.069932124178533e-07,"reference_active":true,"delta_db":-0.43856847943482713}],"active_band_count":3,"active_band_mae_db":0.6554066683582115}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.4326815605163574,"predicted_mos":3.4326815605163574,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":165376,"input_sample_rate":44100,"input_seconds":3.7500226757369615,"input_float32_sha256":"d0d5e34efb781f52442b35abd39020272a63533ba3b6b365f4c1e583acfa6b91","internal_samples":60001,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5956093260028865},"started_at":"2026-10-13T07:15:12.320445+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006442080000852002,"native_input_samples":165375,"native_input_sha256":"fe4191bf605fabf9efe66f56789254cf9aa7ff78ec37e8a7c4f1eb5b84a24a49","output_sha256":"96b9a85a81090cc54578bd60c0499dfb9e35219d985f0e9e7eaf0561d89e42b4","output_samples":165376,"length_delta_samples":1,"analysis_s":0.10448349800026335,"synthesis_s":24.81797860000006,"total_s":24.922462098000324,"rtf":6.6459898928000865,"output_peak":0.7199612855911255,"output_rms":0.052935562554339614,"finished_at":"2026-10-13T07:15:37.278451+00:00","evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"96b9a85a81090cc54578bd60c0499dfb9e35219d985f0e9e7eaf0561d89e42b4","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60001,"length_delta_samples":1,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.51933973848049,"output_peak":0.7204791545291},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9999833336111065,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.280450996459862,"spectral_convergence":0.4400897992633131,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.16851281494168,"spectral_convergence":0.4023328178159391,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.823293079350611,"spectral_convergence":0.42545989003839024,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":1.6228330835900685e-05,"delta_db":-5.460821442593553,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.0005472594847393259,"delta_db":-0.7939865374179745,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0012307624250840346,"delta_db":0.6161531988570911,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0006245273003961993,"delta_db":-0.5614783620534508,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.0003333288896453061,"delta_db":-1.9212011203085957,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":5.380051639786314e-05,"delta_db":-1.8814673012183194,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.8725179937414975,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.42525404477152423,"pitch":null,"output_rms_dbfs":-25.519267356710532,"output_peak":0.7204791545291,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.090752296917385},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.8040189091847694,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055628754242498336,"rms_dbfs":-25.09401331193901,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.392072346291222,"mean":0.00027566782633463543,"zero_energy":false},"output":{"rms":0.05297081221860421,"rms_dbfs":-25.519267356710532,"peak":0.7204791545291,"peak_dbfs":-2.8475716003710683,"crest_factor_db":22.671695756339467,"mean":-1.9981933690939113e-05,"zero_energy":false},"rms_difference_db":-0.42525404477152406},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":136,"quiet_samples":21760,"reference_mean_power":2.905563365750234e-07,"output_mean_power":2.815293312494561e-07,"output_rms_dbfs":-65.5047635130576,"output_zero_energy":false,"output_peak":0.008211721419920983,"output_energy_fraction":3.6387986507788906e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":102,"both_estimated_voiced_frames":90,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.8256880733944955,"absolute_difference_cents_median":9.281439048297049,"absolute_difference_cents_p95":30.649005937234378,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.0002227499093660225,"log_rms_mae_db":1.3487827197347515,"rms_envelope_nrmse":0.10124997801217057,"envelope_correlation":0.9939778364465984,"rms_gain_error_db":-0.42525405768072566,"reference_mean_power":0.003098689875324867,"output_mean_power":0.0028096531340984065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.065525005901086e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018487969313489734,"output_power":0.0017692920233070139,"reference_active":true,"delta_db":-0.19089692088257765},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000254937700664643,"output_power":0.00022460776508672326,"reference_active":true,"delta_db":-0.5500929798844338},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5927620235575996e-05,"output_power":2.2450290776521754e-05,"reference_active":true,"delta_db":-2.04206481106896}],"active_band_count":3,"active_band_mae_db":0.9276849039453238}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00020550660315895724,"log_rms_mae_db":0.9830424413534079,"rms_envelope_nrmse":0.07667824199190501,"envelope_correlation":0.9972745052861794,"rms_gain_error_db":-0.42525424471165474,"reference_mean_power":0.003123712899469445,"output_mean_power":0.0028323419616594074},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":2.8505515590137105e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017535539962877896,"output_power":0.0016795211624531885,"reference_active":true,"delta_db":-0.18733662981956278},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011841366649455359,"output_power":0.00010391591767234845,"reference_active":true,"delta_db":-0.5671975139514608},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.758270588195691e-06,"output_power":1.2224049798493067e-06,"reference_active":true,"delta_db":-1.5787060094910568}],"active_band_count":3,"active_band_mae_db":0.7777467177540268}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00016827060195170467,"log_rms_mae_db":1.4875384821754751,"rms_envelope_nrmse":0.14957736035109298,"envelope_correlation":0.9861215959388073,"rms_gain_error_db":0.34809338371447623,"reference_mean_power":0.001201501254742638,"output_mean_power":0.0013017679266236009},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.2448894715706734e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007295274990436896,"output_power":0.0008680250640512521,"reference_active":true,"delta_db":0.7549059855081415},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014476368795830427,"output_power":0.00012693418549341917,"reference_active":true,"delta_db":-0.5708103794352137},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.785748904828138e-05,"output_power":1.5830109612528004e-05,"reference_active":true,"delta_db":-2.454580463939742}],"active_band_count":3,"active_band_mae_db":1.2600989429610323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012583494842815252,"log_rms_mae_db":1.0939360717378603,"rms_envelope_nrmse":0.11383195238899693,"envelope_correlation":0.9920760977506164,"rms_gain_error_db":0.3480933615468744,"reference_mean_power":0.0012112038213410922,"output_mean_power":0.001312280176948761},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1451125517965788e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000709760959532951,"output_power":0.000834318320544648,"reference_active":true,"delta_db":0.7021967296971333},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.709378020571589e-05,"output_power":5.634643325652341e-05,"reference_active":true,"delta_db":-0.7581583133475206},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7774397828761626e-06,"output_power":1.13736140899423e-06,"reference_active":true,"delta_db":-1.938964079464744}],"active_band_count":3,"active_band_mae_db":1.1331063741697993}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015339527329010622,"log_rms_mae_db":1.8702706532881488,"rms_envelope_nrmse":0.17878862202115764,"envelope_correlation":0.9842114232429994,"rms_gain_error_db":-0.8247569963203216,"reference_mean_power":0.0009604249251386083,"output_mean_power":0.0007943057799329296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.606398277844274e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009968434213476986,"output_power":0.00078836848245515,"reference_active":true,"delta_db":-1.0189769344575903},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073134134931984,"output_power":0.00010317832592115825,"reference_active":true,"delta_db":-0.3068208367157604},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0381145318442426e-05,"output_power":8.841272223532479e-06,"reference_active":true,"delta_db":-0.6973050773205607}],"active_band_count":3,"active_band_mae_db":0.6743676161646371}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00014456690433921235,"log_rms_mae_db":1.5543487936782894,"rms_envelope_nrmse":0.15451995165480822,"envelope_correlation":0.9893312880003353,"rms_gain_error_db":-0.8247575723786098,"reference_mean_power":0.0009681806844812575,"output_mean_power":0.0008007199642239366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.554834929605817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000971830547751713,"output_power":0.0007669407312685583,"reference_active":true,"delta_db":-1.028287429882896},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3135698812325164e-05,"output_power":5.08947128364787e-05,"reference_active":true,"delta_db":-0.18713728275831482},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.152693877304538e-07,"output_power":5.28306777386045e-07,"reference_active":true,"delta_db":-0.6617912555595196}],"active_band_count":3,"active_band_mae_db":0.6257386560669101}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":2.761141804057494,"rms_envelope_nrmse":0.3105217833880319,"envelope_correlation":0.9416379628118245,"rms_gain_error_db":-1.8557253519981176,"reference_mean_power":0.00013836378417930354,"output_mean_power":9.025055806282325e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":9.868150581916699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.732185415785842e-05,"output_power":4.04497372798643e-05,"reference_active":true,"delta_db":-1.51404524607935},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6922402540887428e-05,"output_power":1.4532340429679647e-05,"reference_active":true,"delta_db":-0.6612645865255504},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2492402977589626e-06,"output_power":1.9929691030883024e-06,"reference_active":true,"delta_db":-3.28810725561473}],"active_band_count":3,"active_band_mae_db":1.8211390294065435}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":2.5997320574363383,"rms_envelope_nrmse":0.2799075494582997,"envelope_correlation":0.9551121627158545,"rms_gain_error_db":-1.8557256990796551,"reference_mean_power":0.00013948111618043135,"output_mean_power":9.097935303766248e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":8.397728752995273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.502560023531222e-05,"output_power":3.926745980734076e-05,"reference_active":true,"delta_db":-1.4653198077954854},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.667910609926142e-06,"output_power":6.583232603824142e-06,"reference_active":true,"delta_db":-0.6623784117904358},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.289881205259684e-07,"output_power":1.16554569931876e-07,"reference_active":true,"delta_db":-2.9328364630830106}],"active_band_count":3,"active_band_mae_db":1.6868448942229772}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"golf","status":"measured","mos":{"status":"measured","value":2.526197910308838,"predicted_mos":2.526197910308838,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89760,"input_sample_rate":24000,"input_seconds":3.74,"input_float32_sha256":"8b034665f547fbd66b139f0f286ca16312374364f7a7c03245b33770a00a2ea5","internal_samples":59840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.756599386000744},"started_at":"2026-10-13T05:02:36.378490+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018676749996302533,"native_input_samples":90000,"native_input_sha256":"963528a5e10232d90658f6fa8ad62b2df7ccee774a6dcd6006015eb648f4028c","output_sha256":"baf22ea33a9ffa6214b5b544d5706190612db5ebb70cd8b8fab1a69aa6499014","output_samples":89760,"length_delta_samples":-240,"analysis_s":0.4048765550005555,"synthesis_s":0.06219402699935017,"total_s":0.4670705819999057,"rtf":0.12455215519997485,"output_peak":0.7186335325241089,"output_rms":0.05210073031567613,"finished_at":"2026-10-13T05:02:36.853575+00:00","analysis_stages":{"external_f0_s":0.04181775500001095},"evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"baf22ea33a9ffa6214b5b544d5706190612db5ebb70cd8b8fab1a69aa6499014","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":59840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.65844091916616,"output_peak":0.7070820951278558},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":0.9973333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.130491261459205,"spectral_convergence":0.4939367339278081,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.034982984704014,"spectral_convergence":0.29546163341358683,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.930599584919319,"spectral_convergence":0.317844221188664,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.736820550777267e-05,"output_power":4.917090974195978e-05,"delta_db":-0.6696302173163958,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000660260325970901,"output_power":0.0005692459672200318,"delta_db":-0.6441523939680249,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010692632186459683,"output_power":0.0010519234468077978,"delta_db":-0.0710049247025717,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000712804450500763,"output_power":0.0006116961414367982,"delta_db":-0.6643466163131979,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005199769063411478,"output_power":0.00037704349112672965,"delta_db":-1.3959260780917941,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.315940899401675e-05,"output_power":5.83346665245781e-05,"delta_db":-1.5398467465459575,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.830817829489657,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.576024258774501,"pitch":null,"output_rms_dbfs":-25.65844091916616,"output_peak":0.7070820951278558,"reference_rms_dbfs":-25.082416660391658,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.365357943694178},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.9292626534788104,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05570307456470037,"rms_dbfs":-25.082416660391658,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.380475694743875,"mean":0.00027646049458712816,"zero_energy":false},"output":{"rms":0.05212882717708019,"rms_dbfs":-25.65844091916616,"peak":0.7070820951278558,"peak_dbfs":-3.0106031981445476,"crest_factor_db":22.647837721021613,"mean":0.0002880921663728451,"zero_energy":false},"rms_difference_db":-0.5760242587745026},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":2.92701232764456e-07,"output_mean_power":3.0504948393873313e-07,"output_rms_dbfs":-65.15629705379838,"output_zero_energy":false,"output_peak":0.004065681149426125,"output_energy_fraction":4.052066394690975e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":117,"both_estimated_voiced_frames":105,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.963302752293578,"absolute_difference_cents_median":5.161815865571897,"absolute_difference_cents_p95":15.761991888134089,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.0002227499093660225,"log_rms_mae_db":2.305201228869061,"rms_envelope_nrmse":0.14675078006924572,"envelope_correlation":0.9863907555846232,"rms_gain_error_db":-0.5760243266057384,"reference_mean_power":0.0031069862311802567,"output_mean_power":0.002721052342441768},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":3.069699409889423e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018471331297864027,"output_power":0.0016357269464680546,"reference_active":true,"delta_db":-0.5278738962005557},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025409104346881055,"output_power":0.00020836750465447052,"reference_active":true,"delta_db":-0.8615936596320966},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.596027761480491e-05,"output_power":3.212289400081222e-05,"reference_active":true,"delta_db":-0.49008371219677765}],"active_band_count":3,"active_band_mae_db":0.6265170893431432}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00020550660315895724,"log_rms_mae_db":2.207275042603078,"rms_envelope_nrmse":0.11286643513134353,"envelope_correlation":0.9929879599845993,"rms_gain_error_db":-0.5760261779989215,"reference_mean_power":0.0031321439709948466,"output_mean_power":0.0027430839475181823},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.853854601346697e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017513381244453068,"output_power":0.001542360538564684,"reference_active":true,"delta_db":-0.5518409658865002},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011713195830246314,"output_power":9.417172595235852e-05,"reference_active":true,"delta_db":-0.9475487409963218},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.74893068387265e-06,"output_power":1.814277026078631e-06,"reference_active":true,"delta_db":0.15931003959473775}],"active_band_count":3,"active_band_mae_db":0.5528999154925199}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00016827060195170467,"log_rms_mae_db":1.1330907468636657,"rms_envelope_nrmse":0.18166293606223938,"envelope_correlation":0.976905076058431,"rms_gain_error_db":-0.27809350743057576,"reference_mean_power":0.0012047181254684356,"output_mean_power":0.001129993886281204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2465776246682866e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007288523412344929,"output_power":0.0006162738070840283,"reference_active":true,"delta_db":-0.7286584346460303},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014497192622568318,"output_power":0.00013507633832399193,"reference_active":true,"delta_db":-0.30704630185483517},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.771387452404623e-05,"output_power":2.1047486271129385e-05,"reference_active":true,"delta_db":-1.1949701143501688}],"active_band_count":3,"active_band_mae_db":0.7435582836170115}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00012583494842815252,"log_rms_mae_db":0.940861673362751,"rms_envelope_nrmse":0.13005980556417418,"envelope_correlation":0.9878373231156378,"rms_gain_error_db":-0.27809351296607926,"reference_mean_power":0.0012144729273954728,"output_mean_power":0.00113914363223191},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.1463439072841361e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007087961236208862,"output_power":0.0005957049552595271,"reference_active":true,"delta_db":-0.7549012103468071},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.689329934021284e-05,"output_power":6.006940691686662e-05,"reference_active":true,"delta_db":-0.46729272891147555},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7632022249311231e-06,"output_power":1.5264835191075877e-06,"reference_active":true,"delta_db":-0.6261000545600972}],"active_band_count":3,"active_band_mae_db":0.61609799793946}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00015339527329010622,"log_rms_mae_db":1.7747297150129857,"rms_envelope_nrmse":0.14185325870127796,"envelope_correlation":0.9899390086016748,"rms_gain_error_db":-0.5987783226608657,"reference_mean_power":0.0009629963433561795,"output_mean_power":0.000838970722942259},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.6097459814270998e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009972354975272095,"output_power":0.0008890391896349929,"reference_active":true,"delta_db":-0.49876823746206567},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011043207417809584,"output_power":8.744240502813957e-05,"reference_active":true,"delta_db":-1.0137313532842507},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0355236394358364e-05,"output_power":9.479661561552083e-06,"reference_active":true,"delta_db":-0.3836718499153219}],"active_band_count":3,"active_band_mae_db":0.6320571468872127}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00014456690433921235,"log_rms_mae_db":1.5627894080330824,"rms_envelope_nrmse":0.1057145007809533,"envelope_correlation":0.9955980417803751,"rms_gain_error_db":-0.5987782304776266,"reference_mean_power":0.0009707938571864756,"output_mean_power":0.0008457640022193393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.5579542142416644e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009719571074691375,"output_power":0.0008647658688210613,"reference_active":true,"delta_db":-0.5074855965743382},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.267308031294983e-05,"output_power":4.1225237139488526e-05,"reference_active":true,"delta_db":-1.064255540577606},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.110119994553358e-07,"output_power":5.530535866276898e-07,"reference_active":true,"delta_db":-0.4328252615717051}],"active_band_count":3,"active_band_mae_db":0.6681887995745498}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":2.612426249928074,"rms_envelope_nrmse":0.2561993287597694,"envelope_correlation":0.9635246483055608,"rms_gain_error_db":-1.524015620218119,"reference_mean_power":0.0001387342355908665,"output_mean_power":9.767469945516253e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":9.853583473102374e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7317032442880536e-05,"output_power":3.957302960037406e-05,"reference_active":true,"delta_db":-1.6088439704615933},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.679285879398322e-05,"output_power":1.4635674827337032e-05,"reference_active":true,"delta_db":-0.5971188429901727},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.249992624438683e-06,"output_power":4.252429355213273e-06,"reference_active":true,"delta_db":0.0024893113063414914}],"active_band_count":3,"active_band_mae_db":0.7361507082527025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":2.363757448922557,"rms_envelope_nrmse":0.21634530223041024,"envelope_correlation":0.9787981581084831,"rms_gain_error_db":-1.5240155489444194,"reference_mean_power":0.00013985758302855673,"output_mean_power":9.846558461109533e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":8.379479110816266e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.503406002749344e-05,"output_power":3.837598405041571e-05,"reference_active":true,"delta_db":-1.5657202823898344},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.520107174694761e-06,"output_power":6.58920284124451e-06,"reference_active":true,"delta_db":-0.5739115306903345},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2851041578288108e-07,"output_power":1.9264109565641505e-07,"reference_active":true,"delta_db":-0.741570608705428}],"active_band_count":3,"active_band_mae_db":0.9604008072618657}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.1722030639648438,"predicted_mos":3.1722030639648438,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90120,"input_sample_rate":24000,"input_seconds":3.755,"input_float32_sha256":"c185e33f70e41510247f2a1eaaafc0132d57124e3e04efce3892116bb97cf3e0","internal_samples":60080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3948007589970075},"started_at":"2026-10-13T07:38:10.522788+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030082810008025263,"native_input_samples":90000,"native_input_sha256":"963528a5e10232d90658f6fa8ad62b2df7ccee774a6dcd6006015eb648f4028c","output_sha256":"d3b2798aea6cfde98d1c7868e388b46185de880902347332b88ef02806276d86","output_samples":90120,"length_delta_samples":120,"analysis_s":1.3737268459990446,"synthesis_s":9.1208635840012,"total_s":10.494590430000244,"rtf":2.7985574480000652,"output_peak":0.6707433462142944,"output_rms":0.053378561824004245,"finished_at":"2026-10-13T07:38:21.032450+00:00","analysis_stages":{"external_f0_s":0.9038352329989721},"evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"d3b2798aea6cfde98d1c7868e388b46185de880902347332b88ef02806276d86","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.449465699983516,"output_peak":0.6678414939044683},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986684420772304,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.844843030847795,"spectral_convergence":0.562676582519738,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.877900705713289,"spectral_convergence":0.34328188447185887,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.510953889650844,"spectral_convergence":0.32030291117468745,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":5.768831955174075e-05,"delta_db":0.047318911911876205,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.000624614145583394,"delta_db":-0.21980135515905502,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0010415541170249718,"delta_db":-0.10877069049211477,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0006199474018857643,"delta_db":-0.5934441967436254,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.00044691256259996547,"delta_db":-0.6477050492421033,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":6.445431081564669e-05,"delta_db":-1.0968120530624388,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.45230870943520224,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3496656665068413,"pitch":null,"output_rms_dbfs":-25.443678978445853,"output_peak":0.6678414939044683,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.07789920873731},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9021580210784278,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055628754242498336,"rms_dbfs":-25.09401331193901,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.392072346291222,"mean":0.00027566782633463543,"zero_energy":false},"output":{"rms":0.05343379883052938,"rms_dbfs":-25.443678978445853,"peak":0.6678414939044683,"peak_dbfs":-3.506532022676681,"crest_factor_db":21.93714695576917,"mean":6.0930826413666586e-05,"zero_energy":false},"rms_difference_db":-0.3496656665068443},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":136,"quiet_samples":21760,"reference_mean_power":2.905563365750234e-07,"output_mean_power":6.346851120909483e-07,"output_rms_dbfs":-61.9744168888091,"output_zero_energy":false,"output_peak":0.03167487946778289,"output_energy_fraction":8.061833966378983e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":120,"both_estimated_voiced_frames":98,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.8990825688073395,"absolute_difference_cents_median":8.12217324552902,"absolute_difference_cents_p95":27.516233373520283,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.0002227499093660225,"log_rms_mae_db":1.9793012643718808,"rms_envelope_nrmse":0.1634040231583778,"envelope_correlation":0.980295868132612,"rms_gain_error_db":-0.3496656724707487,"reference_mean_power":0.003098689875324867,"output_mean_power":0.002858982821642396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.065525005901086e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018487969313489734,"output_power":0.0017546061598709406,"reference_active":true,"delta_db":-0.22709561928776242},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000254937700664643,"output_power":0.00023018151935266275,"reference_active":true,"delta_db":-0.44363612127708063},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5927620235575996e-05,"output_power":3.259531286869236e-05,"reference_active":true,"delta_db":-0.42273297465478465}],"active_band_count":3,"active_band_mae_db":0.3644882384065426}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00020550660315895724,"log_rms_mae_db":1.4323252038247787,"rms_envelope_nrmse":0.10964019648389965,"envelope_correlation":0.9913399153559188,"rms_gain_error_db":-0.3496661425049919,"reference_mean_power":0.003123712899469445,"output_mean_power":0.0028820698140474077},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":2.8505515590137105e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017535539962877896,"output_power":0.0016640539102619851,"reference_active":true,"delta_db":-0.22751751433601897},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011841366649455359,"output_power":0.00010852191471012871,"reference_active":true,"delta_db":-0.37884380928092476},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.758270588195691e-06,"output_power":1.6776743129454023e-06,"reference_active":true,"delta_db":-0.20378056367765318}],"active_band_count":3,"active_band_mae_db":0.27004729576486564}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00016827060195170467,"log_rms_mae_db":1.6207532448620745,"rms_envelope_nrmse":0.1894548077912151,"envelope_correlation":0.9745830555600574,"rms_gain_error_db":-0.07242739700868639,"reference_mean_power":0.001201501254742638,"output_mean_power":0.001181629946619024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.2448894715706734e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007295274990436896,"output_power":0.0007575561777969515,"reference_active":true,"delta_db":0.1637317720083833},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014476368795830427,"output_power":0.0001414171909560766,"reference_active":true,"delta_db":-0.1015743216478699},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.785748904828138e-05,"output_power":2.673999582393676e-05,"reference_active":true,"delta_db":-0.17780633353833158}],"active_band_count":3,"active_band_mae_db":0.14770414239819493}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012583494842815252,"log_rms_mae_db":1.2078103342582924,"rms_envelope_nrmse":0.11639150085230134,"envelope_correlation":0.9898545304661639,"rms_gain_error_db":-0.0724273979829872,"reference_mean_power":0.0012112038213410922,"output_mean_power":0.0011911720447945096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1451125517965788e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000709760959532951,"output_power":0.000736186615868122,"reference_active":true,"delta_db":0.15875810295324805},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.709378020571589e-05,"output_power":6.649434756852217e-05,"reference_active":true,"delta_db":-0.038975324856801954},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7774397828761626e-06,"output_power":1.774400246524132e-06,"reference_active":true,"delta_db":-0.007433072303616156}],"active_band_count":3,"active_band_mae_db":0.06838883337122205}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015339527329010622,"log_rms_mae_db":1.9249199457437434,"rms_envelope_nrmse":0.19378734677252343,"envelope_correlation":0.9787086422729295,"rms_gain_error_db":-0.6283424693070465,"reference_mean_power":0.0009604249251386083,"output_mean_power":0.0008310538685911449},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.606398277844274e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009968434213476986,"output_power":0.0008570974400246234,"reference_active":true,"delta_db":-0.6559674913773662},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073134134931984,"output_power":8.406267961543681e-05,"reference_active":true,"delta_db":-1.196673311600046},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0381145318442426e-05,"output_power":8.975571210037852e-06,"reference_active":true,"delta_db":-0.6318317361454083}],"active_band_count":3,"active_band_mae_db":0.8281575130409401}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00014456690433921235,"log_rms_mae_db":1.3186495375131178,"rms_envelope_nrmse":0.13915715608438659,"envelope_correlation":0.9902673565904094,"rms_gain_error_db":-0.6283424479644972,"reference_mean_power":0.0009681806844812575,"output_mean_power":0.0008377649165772638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.554834929605817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000971830547751713,"output_power":0.0008324017869380463,"reference_active":true,"delta_db":-0.6725754235873196},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3135698812325164e-05,"output_power":4.045188285176191e-05,"reference_active":true,"delta_db":-1.1844765570332274},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.152693877304538e-07,"output_power":4.846892272751035e-07,"reference_active":true,"delta_db":-1.0360194044644313}],"active_band_count":3,"active_band_mae_db":0.9643571283616593}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":2.1498808965080074,"rms_envelope_nrmse":0.22630971923955362,"envelope_correlation":0.9606857694041777,"rms_gain_error_db":-0.6526069710247953,"reference_mean_power":0.00013836378417930354,"output_mean_power":0.00011905886392345218},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":9.868150581916699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.732185415785842e-05,"output_power":4.9390361627401126e-05,"reference_active":true,"delta_db":-0.6467802383270295},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6922402540887428e-05,"output_power":1.408812571568495e-05,"reference_active":true,"delta_db":-0.7960880318526047},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2492402977589626e-06,"output_power":3.5089192499703675e-06,"reference_active":true,"delta_db":-0.8313791746431323}],"active_band_count":3,"active_band_mae_db":0.7580824816075887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":1.4634748346831936,"rms_envelope_nrmse":0.13830018177232747,"envelope_correlation":0.9868766222469599,"rms_gain_error_db":-0.6526069557414126,"reference_mean_power":0.00013948111618043135,"output_mean_power":0.00012002030291494067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":8.397728752995273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.502560023531222e-05,"output_power":4.72951246597019e-05,"reference_active":true,"delta_db":-0.6574841415096924},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.667910609926142e-06,"output_power":6.3127119094489195e-06,"reference_active":true,"delta_db":-0.8446107130112689},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.289881205259684e-07,"output_power":1.913357691008069e-07,"reference_active":true,"delta_db":-0.7801678611589455}],"active_band_count":3,"active_band_mae_db":0.7607542385599689}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.7236487865448,"predicted_mos":2.7236487865448,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90240,"input_sample_rate":24000,"input_seconds":3.76,"input_float32_sha256":"e0b6eb3c1d249f664795e2ca690ff88d935891325a6befd2f8296f802ad1ce9a","internal_samples":60160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.577718751001157},"started_at":"2026-10-13T05:10:46.577166+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008921422000639723,"native_input_samples":90000,"native_input_sha256":"963528a5e10232d90658f6fa8ad62b2df7ccee774a6dcd6006015eb648f4028c","output_sha256":"9e3d5c8b4f9051d73a74e6d2c881ade5b40b41715e4f1c9a054b9a62839a565c","output_samples":90240,"length_delta_samples":240,"analysis_s":0.7548731269998825,"synthesis_s":0.11698881899974367,"total_s":0.8718619459996262,"rtf":0.23249651893323364,"output_peak":0.7039338946342468,"output_rms":0.05443861489805037,"finished_at":"2026-10-13T05:10:47.472810+00:00","analysis_stages":{"external_f0_s":0.260523742001169},"evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"9e3d5c8b4f9051d73a74e6d2c881ade5b40b41715e4f1c9a054b9a62839a565c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.278226738786937,"output_peak":0.6959849632898815},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973404255319149,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.472031893026075,"spectral_convergence":0.18578446208745442,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.9474642820038293,"spectral_convergence":0.18264050567624276,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.194518182603347,"spectral_convergence":0.20589988380371965,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":6.009811500451108e-05,"delta_db":0.22504853392076896,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.0006440951125401998,"delta_db":-0.08641947254301292,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0010454347969704562,"delta_db":-0.09261957615228897,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0006571190938144853,"delta_db":-0.3405517734135024,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.0004918132709554789,"delta_db":-0.2319282422010037,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":7.539125884686425e-05,"delta_db":-0.416121793197062,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23211489857127318,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.17264765521363498,"pitch":null,"output_rms_dbfs":-25.266660967152642,"output_peak":0.6959849632898815,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.204671452544417},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9547452048632012,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055628754242498336,"rms_dbfs":-25.09401331193901,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.392072346291222,"mean":0.00027566782633463543,"zero_energy":false},"output":{"rms":0.054533949500581706,"rms_dbfs":-25.266660967152642,"peak":0.6959849632898815,"peak_dbfs":-3.1480028638454804,"crest_factor_db":22.118658103307162,"mean":2.0066388336298117e-05,"zero_energy":false},"rms_difference_db":-0.172647655213634},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":136,"quiet_samples":21760,"reference_mean_power":2.905563365750234e-07,"output_mean_power":3.6611439278268654e-07,"output_rms_dbfs":-64.36383197687074,"output_zero_energy":false,"output_peak":0.012965788931251072,"output_energy_fraction":4.4646820849487624e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.8532110091743119,"absolute_difference_cents_median":0.826310333940149,"absolute_difference_cents_p95":3.627153655636125,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.0002227499093660225,"log_rms_mae_db":0.733286063516063,"rms_envelope_nrmse":0.06108142256117771,"envelope_correlation":0.9974349549117725,"rms_gain_error_db":-0.17264764848388836,"reference_mean_power":0.003098689875324867,"output_mean_power":0.0029779222063141954},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.065525005901086e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018487969313489734,"output_power":0.0017623294273096995,"reference_active":true,"delta_db":-0.20802118508428508},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000254937700664643,"output_power":0.00023269027622398173,"reference_active":true,"delta_db":-0.3965582934597738},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5927620235575996e-05,"output_power":2.6527525659282465e-05,"reference_active":true,"delta_db":-1.3173170818495181}],"active_band_count":3,"active_band_mae_db":0.6406321867978589}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00020550660315895724,"log_rms_mae_db":0.44597338086835353,"rms_envelope_nrmse":0.03588209195286734,"envelope_correlation":0.9992708833503238,"rms_gain_error_db":-0.17264762386880034,"reference_mean_power":0.003123712899469445,"output_mean_power":0.0030019700055424537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":2.8505515590137105e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017535539962877896,"output_power":0.001667770498444985,"reference_active":true,"delta_db":-0.21782856191320973},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011841366649455359,"output_power":0.00011041710170144738,"reference_active":true,"delta_db":-0.3036548525984746},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.758270588195691e-06,"output_power":1.1955057252779299e-06,"reference_active":true,"delta_db":-1.675340511254183}],"active_band_count":3,"active_band_mae_db":0.7322746419219558}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00016827060195170467,"log_rms_mae_db":0.705022530220815,"rms_envelope_nrmse":0.08703615179246568,"envelope_correlation":0.9947220157650303,"rms_gain_error_db":-0.1155995319494377,"reference_mean_power":0.001201501254742638,"output_mean_power":0.0011699418475256812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.2448894715706734e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007295274990436896,"output_power":0.000710553133888461,"reference_active":true,"delta_db":-0.11445107733677302},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014476368795830427,"output_power":0.0001222930913388924,"reference_active":true,"delta_db":-0.7325771525705003},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.785748904828138e-05,"output_power":2.0345265876334664e-05,"reference_active":true,"delta_db":-1.3647859877416104}],"active_band_count":3,"active_band_mae_db":0.7372714058829613}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012583494842815252,"log_rms_mae_db":0.4650558011326569,"rms_envelope_nrmse":0.05409078118146113,"envelope_correlation":0.9979176059830274,"rms_gain_error_db":-0.11559953077009186,"reference_mean_power":0.0012112038213410922,"output_mean_power":0.0011793895605696188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1451125517965788e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000709760959532951,"output_power":0.0006830304866910606,"reference_active":true,"delta_db":-0.1667201865040669},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.709378020571589e-05,"output_power":5.7360962840431936e-05,"reference_active":true,"delta_db":-0.6806582900493519},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7774397828761626e-06,"output_power":1.1536807557856987e-06,"reference_active":true,"delta_db":-1.8770924802341116}],"active_band_count":3,"active_band_mae_db":0.9081569855958435}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015339527329010622,"log_rms_mae_db":0.7722283384840078,"rms_envelope_nrmse":0.0675960053733121,"envelope_correlation":0.9977668047191958,"rms_gain_error_db":-0.27819383142554166,"reference_mean_power":0.0009604249251386083,"output_mean_power":0.0009008324857550466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.606398277844274e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009968434213476986,"output_power":0.000925643965916276,"reference_active":true,"delta_db":-0.3218297269059228},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073134134931984,"output_power":0.0001056745175375344,"reference_active":true,"delta_db":-0.20300287082145427},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0381145318442426e-05,"output_power":9.69150176488628e-06,"reference_active":true,"delta_db":-0.2985419128823311}],"active_band_count":3,"active_band_mae_db":0.2744581702032361}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00014456690433921235,"log_rms_mae_db":0.4591566877031107,"rms_envelope_nrmse":0.0449949511208532,"envelope_correlation":0.9993350813222203,"rms_gain_error_db":-0.2781937909078722,"reference_mean_power":0.0009681806844812575,"output_mean_power":0.0009081070242656738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.554834929605817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000971830547751713,"output_power":0.0009031704578129067,"reference_active":true,"delta_db":-0.31820822584217245},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3135698812325164e-05,"output_power":5.18063839319525e-05,"reference_active":true,"delta_db":-0.11003116824067684},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.152693877304538e-07,"output_power":4.781267720723887e-07,"reference_active":true,"delta_db":-1.0952224550922591}],"active_band_count":3,"active_band_mae_db":0.5078206163917028}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":0.9270681838224114,"rms_envelope_nrmse":0.10097449415243964,"envelope_correlation":0.9920307558834118,"rms_gain_error_db":-0.14613164203763931,"reference_mean_power":0.00013836378417930354,"output_mean_power":0.00013378556806466674},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":9.868150581916699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.732185415785842e-05,"output_power":5.6166524646968e-05,"reference_active":true,"delta_db":-0.08842677477410431},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6922402540887428e-05,"output_power":1.601032231258686e-05,"reference_active":true,"delta_db":-0.24061946569765225},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2492402977589626e-06,"output_power":4.201396109016028e-06,"reference_active":true,"delta_db":-0.0491766257022569}],"active_band_count":3,"active_band_mae_db":0.12607428872467116}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":0.5872334321647266,"rms_envelope_nrmse":0.0501234006282351,"envelope_correlation":0.9980197694692187,"rms_gain_error_db":-0.14613153304476417,"reference_mean_power":0.00013948111618043135,"output_mean_power":0.00013486593288476628},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":8.397728752995273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.502560023531222e-05,"output_power":5.452525412323954e-05,"reference_active":true,"delta_db":-0.03967090311287142},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.667910609926142e-06,"output_power":7.410153676027603e-06,"reference_active":true,"delta_db":-0.14849826646586992},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.289881205259684e-07,"output_power":2.4049186897321894e-07,"reference_active":true,"delta_db":0.21287444926160626}],"active_band_count":3,"active_band_mae_db":0.13368120628011584}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0004","method":"magphase","status":"measured","mos":{"status":"measured","value":2.959484100341797,"predicted_mos":2.959484100341797,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":180453,"input_sample_rate":48000,"input_seconds":3.7594375,"input_float32_sha256":"113b37c2313bedce40a202108445a45f618f4b670aeffc363e0242450f08a877","internal_samples":60151,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5929564540001593},"started_at":"2026-10-13T05:15:46.196360+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006234006001250236,"native_input_samples":180000,"native_input_sha256":"b4a3cf29faae75a521e792988b0fae3591e5a6a0503c8a3cf2cb8ba98988d16e","output_sha256":"e278fa8d3a1b2edbc1b986c7962ef1e372d7f22302ebfe307c4ccc9d6e62e296","output_samples":180453,"length_delta_samples":453,"analysis_s":2.151060222999149,"synthesis_s":0.7615016119998472,"total_s":2.9125618349989963,"rtf":0.7766831559997324,"output_peak":0.7327128763873777,"output_rms":0.055233870732700856,"finished_at":"2026-10-13T05:15:49.122818+00:00","evaluation_input":{"reference":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"e278fa8d3a1b2edbc1b986c7962ef1e372d7f22302ebfe307c4ccc9d6e62e296","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":60000,"output_samples":60151,"length_delta_samples":151,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.15308470546654,"output_peak":0.7078765307436823},"common_prefix":{"samples":60000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9974896510448704,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":60000,"output_samples":60000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1762823278098535,"spectral_convergence":0.6085212084997417,"spectral_convergence_status":"ok","amplitude_floor":1.4952467880734047e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.908979557673384,"spectral_convergence":0.20756915491518146,"spectral_convergence_status":"ok","amplitude_floor":1.087083783284318e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.782002236157463,"spectral_convergence":0.22054085164482296,"spectral_convergence_status":"ok","amplitude_floor":6.922347653241367e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.706318352359072e-05,"output_power":5.295508038145454e-05,"delta_db":-0.3244836544972556,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006570402135178566,"output_power":0.0005999011191683051,"delta_db":-0.39512278798859524,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010679696452407233,"output_power":0.0010153247501553464,"delta_db":-0.2195393604184408,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007107211898605409,"output_power":0.0007402054560635588,"delta_db":0.17653018013715435,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005187917577577873,"output_power":0.0005655542318726588,"delta_db":0.37481188764880763,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.297230867177819e-05,"output_power":8.649415564188943e-05,"delta_db":0.18053589106865817,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27850396029315194,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.048155381056641945,"pitch":null,"output_rms_dbfs":-25.14216869299565,"output_peak":0.7078765307436823,"reference_rms_dbfs":-25.09401331193901,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.2890880405469005},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":60000,"output_samples":60000,"intelligibility":{"estoi":{"value":0.9623626114590832,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055628754242498336,"rms_dbfs":-25.09401331193901,"peak":0.652984619140625,"peak_dbfs":-3.701940965647786,"crest_factor_db":21.392072346291222,"mean":0.00027566782633463543,"zero_energy":false},"output":{"rms":0.0553211966002473,"rms_dbfs":-25.14216869299565,"peak":0.7078765307436823,"peak_dbfs":-3.000849724534053,"crest_factor_db":22.1413189684616,"mean":0.0008728456012922415,"zero_energy":false},"rms_difference_db":-0.048155381056641744},"full_scale":{"reference":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":60000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":375,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.606015994795598e-06,"threshold_dbfs":-53.366745582433886,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":136,"quiet_samples":21760,"reference_mean_power":2.905563365750234e-07,"output_mean_power":4.2682091000753105e-07,"output_rms_dbfs":-63.69754312587677,"output_zero_energy":false,"output_peak":0.014263617766451346,"output_energy_fraction":5.0578995192380154e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":368,"reference_estimated_voiced_frames":109,"output_estimated_voiced_frames":111,"both_estimated_voiced_frames":104,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9541284403669725,"absolute_difference_cents_median":4.448012121685682,"absolute_difference_cents_p95":17.009166738760083,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":60000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.0002227499093660225,"log_rms_mae_db":1.0340788352210255,"rms_envelope_nrmse":0.094157487814253,"envelope_correlation":0.9933709700451767,"rms_gain_error_db":-0.048155373259202966,"reference_mean_power":0.003098689875324867,"output_mean_power":0.0030645208168935036},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":3.065525005901086e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018487969313489734,"output_power":0.0019079862479793107,"reference_active":true,"delta_db":0.13686028528652566},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000254937700664643,"output_power":0.00023853394450205803,"reference_active":true,"delta_db":-0.28883874578853463},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5927620235575996e-05,"output_power":3.810241437538925e-05,"reference_active":true,"delta_db":0.25524044326968026}],"active_band_count":3,"active_band_mae_db":0.2269798247815802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00020550660315895724,"log_rms_mae_db":0.6197574749313344,"rms_envelope_nrmse":0.04828177420180073,"envelope_correlation":0.9982013804544819,"rms_gain_error_db":-0.04815541114675012,"reference_mean_power":0.003123712899469445,"output_mean_power":0.003089267886773554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":2.8505515590137105e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017535539962877896,"output_power":0.0018093608957023672,"reference_active":true,"delta_db":0.13606056585763182},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011841366649455359,"output_power":0.00011477979678237376,"reference_active":true,"delta_db":-0.13536377081598222},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.758270588195691e-06,"output_power":1.8786383149792025e-06,"reference_active":true,"delta_db":0.2875746414286354}],"active_band_count":3,"active_band_mae_db":0.18633299270074985}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00016827060195170467,"log_rms_mae_db":1.0466201820161167,"rms_envelope_nrmse":0.08834569112128204,"envelope_correlation":0.9948523666785446,"rms_gain_error_db":-0.2301588446957903,"reference_mean_power":0.001201501254742638,"output_mean_power":0.0011394842958861186},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.2448894715706734e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007295274990436896,"output_power":0.0006918356789227668,"reference_active":true,"delta_db":-0.23038711550653923},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014476368795830427,"output_power":0.00013072364347065302,"reference_active":true,"delta_db":-0.4430549472139742},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.785748904828138e-05,"output_power":2.8906285111788054e-05,"reference_active":true,"delta_db":0.16050313492024337}],"active_band_count":3,"active_band_mae_db":0.27798173254691894}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00012583494842815252,"log_rms_mae_db":0.7445230792055546,"rms_envelope_nrmse":0.05014205640337837,"envelope_correlation":0.9985955849961727,"rms_gain_error_db":-0.2301588447410155,"reference_mean_power":0.0012112038213410922,"output_mean_power":0.0011486860526139747},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.1451125517965788e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000709760959532951,"output_power":0.0006719523980785019,"reference_active":true,"delta_db":-0.23773599063935952},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.709378020571589e-05,"output_power":6.097265041749849e-05,"reference_active":true,"delta_db":-0.4154718787612973},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7774397828761626e-06,"output_power":1.805545104466345e-06,"reference_active":true,"delta_db":0.06813445669846098}],"active_band_count":3,"active_band_mae_db":0.24044744203303928}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":0.00015339527329010622,"log_rms_mae_db":1.090423669177949,"rms_envelope_nrmse":0.11575980909750638,"envelope_correlation":0.9929440056437216,"rms_gain_error_db":0.3061226888397813,"reference_mean_power":0.0009604249251386083,"output_mean_power":0.001030565717644232},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":1.606398277844274e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009968434213476986,"output_power":0.0010862657799025198,"reference_active":true,"delta_db":0.37309151293383314},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011073134134931984,"output_power":0.00011205033854288613,"reference_active":true,"delta_db":0.051426122332792486},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0381145318442426e-05,"output_power":1.043333883970601e-05,"reference_active":true,"delta_db":0.02178041561198086}],"active_band_count":3,"active_band_mae_db":0.14876601695953548}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":0.00014456690433921235,"log_rms_mae_db":0.6704641103698399,"rms_envelope_nrmse":0.06217707882311563,"envelope_correlation":0.998455258167012,"rms_gain_error_db":0.30612259131301717,"reference_mean_power":0.0009681806844812575,"output_mean_power":0.0010388878645161538},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":1.554834929605817e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000971830547751713,"output_power":0.0010569390770521676,"reference_active":true,"delta_db":0.36459408682170896},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3135698812325164e-05,"output_power":5.3999545175357924e-05,"reference_active":true,"delta_db":0.07003705323370807},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.152693877304538e-07,"output_power":5.699491466602223e-07,"reference_active":true,"delta_db":-0.33229199867445247}],"active_band_count":3,"active_band_mae_db":0.2556410462432898}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":749,"rms_floor":4.2745097516618104e-05,"log_rms_mae_db":1.3121686464316402,"rms_envelope_nrmse":0.195171995528798,"envelope_correlation":0.9719059024662916,"rms_gain_error_db":0.3072584543789806,"reference_mean_power":0.00013836378417930354,"output_mean_power":0.00014850746221860972},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":749,"frequency_resolution_hz":0.26702269692923897,"power_floor":9.868150581916699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.732185415785842e-05,"output_power":6.0081912930804984e-05,"reference_active":true,"delta_db":0.20423521761295826},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6922402540887428e-05,"output_power":1.6588577909681025e-05,"reference_active":true,"delta_db":-0.08652864774664225},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2492402977589626e-06,"output_power":5.234390418789693e-06,"reference_active":true,"delta_db":0.9055482087726673}],"active_band_count":3,"active_band_mae_db":0.39877069137742255}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":60000,"uncovered_tail_samples":0,"metrics":{"frame_count":743,"rms_floor":3.73494015091435e-05,"log_rms_mae_db":1.0021256598355142,"rms_envelope_nrmse":0.1278682936121527,"envelope_correlation":0.9872717130911907,"rms_gain_error_db":0.30725839915229347,"reference_mean_power":0.00013948111618043135,"output_mean_power":0.0001497067057744268},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":743,"frequency_resolution_hz":0.2691790040376851,"power_floor":8.397728752995273e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.502560023531222e-05,"output_power":5.794033595358778e-05,"reference_active":true,"delta_db":0.22416220368553672},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.667910609926142e-06,"output_power":7.790902641480644e-06,"reference_active":true,"delta_db":0.06910735881518745},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.289881205259684e-07,"output_power":2.892563819848299e-07,"reference_active":true,"delta_db":1.0146999720928114}],"active_band_count":3,"active_band_mae_db":0.43598984486451187}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"hnm","status":"measured","mos":{"status":"measured","value":1.8778634071350098,"predicted_mos":1.8778634071350098,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"37f6048dfe5522144459e19160e772f28b197a1f10b49dd51c71cb9526659cf4","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2752391869998974},"started_at":"2026-10-13T04:39:21.503733+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4969000403652899e-05,"native_input_samples":96000,"native_input_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","output_sha256":"4b606d7a7c3796493aa462462266928f66e07193c5d82e0e2c10eadda7378730","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9651538580001215,"synthesis_s":0.5507268620003742,"total_s":1.5158807200004958,"rtf":0.2526467866667493,"output_peak":0.7136517416988136,"output_rms":0.05762076110669307,"finished_at":"2026-10-13T04:39:23.025687+00:00","evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"4b606d7a7c3796493aa462462266928f66e07193c5d82e0e2c10eadda7378730","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.650962171483894,"spectral_convergence":0.256168390093675,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.158134556283869,"spectral_convergence":0.26340561727634026,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.038220474678467,"spectral_convergence":0.2692148537589734,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.3913164750681255e-05,"delta_db":0.12068698223123654,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.00042876405294818836,"delta_db":-0.0003130621704100723,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0014536364334059035,"delta_db":-0.15016819252995248,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.0009330180023681632,"delta_db":-0.1423881158742889,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.00037931749255544847,"delta_db":-0.38967407090655165,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":8.150296448620754e-05,"delta_db":-0.23071425585393374,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17232411326106223,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15561232303191067,"pitch":null,"output_rms_dbfs":-24.788420189333614,"output_peak":0.7136517416988136,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.282439067482077},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.650962171483894,"spectral_convergence":0.256168390093675,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.158134556283869,"spectral_convergence":0.26340561727634026,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.038220474678467,"spectral_convergence":0.2692148537589734,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.3913164750681255e-05,"delta_db":0.12068698223123654,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.00042876405294818836,"delta_db":-0.0003130621704100723,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0014536364334059035,"delta_db":-0.15016819252995248,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.0009330180023681632,"delta_db":-0.1423881158742889,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.00037931749255544847,"delta_db":-0.38967407090655165,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":8.150296448620754e-05,"delta_db":-0.23071425585393374,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17232411326106223,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15561232303191067,"pitch":null,"output_rms_dbfs":-24.788420189333614,"output_peak":0.7136517416988136,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.282439067482077},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8982290992225471,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0586623702554101,"rms_dbfs":-24.632807866301704,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.232068021857707,"mean":8.900610605875651e-05,"zero_energy":false},"output":{"rms":0.05762076110669307,"rms_dbfs":-24.788420189333614,"peak":0.7136517416988136,"peak_dbfs":-2.9302733991058734,"crest_factor_db":21.858146790227742,"mean":-1.665847430880541e-05,"zero_energy":false},"rms_difference_db":-0.1556123230319102},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":3.025654520998905e-06,"output_rms_dbfs":-55.19180662627924,"output_zero_energy":false,"output_peak":0.0795358943489424,"output_energy_fraction":0.0003143984898881139,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":216,"both_estimated_voiced_frames":211,"reference_voiced_output_unvoiced_frames":42,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.83399209486166,"absolute_difference_cents_median":1.9028305948732052,"absolute_difference_cents_p95":10.511327625407327,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027067935709258154,"log_rms_mae_db":1.1761016585076727,"rms_envelope_nrmse":0.07855316573945093,"envelope_correlation":0.9957252592909407,"rms_gain_error_db":-0.15520208709338643,"reference_mean_power":0.0034432187972516922,"output_mean_power":0.0033223425768678207},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.845747823685581e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011374442611865624,"output_power":0.0010958787352681003,"reference_active":true,"delta_db":-0.16167624032699082},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011129046974770975,"output_power":9.943171729014519e-05,"reference_active":true,"delta_db":-0.4893303533578946},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3561368721227545e-05,"output_power":1.552558797782414e-05,"reference_active":true,"delta_db":-1.8115245921788543}],"active_band_count":3,"active_band_mae_db":0.8208437286212465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026426858408099,"log_rms_mae_db":0.7190789252366355,"rms_envelope_nrmse":0.042208163365580025,"envelope_correlation":0.9988594725147196,"rms_gain_error_db":-0.15462818157754338,"reference_mean_power":0.0034555022621815964,"output_mean_power":0.003334635455257301},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7242545404192642e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010676968681649736,"output_power":0.0010268764537107586,"reference_active":true,"delta_db":-0.169297732316247},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4544672827083896e-05,"output_power":4.0069432507910304e-05,"reference_active":true,"delta_db":-0.4598258177940746},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7651139324996662e-06,"output_power":1.195440531771601e-06,"reference_active":true,"delta_db":-1.692447663257695}],"active_band_count":3,"active_band_mae_db":0.7738570711226722}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001806334260766054,"log_rms_mae_db":1.2498676278191199,"rms_envelope_nrmse":0.09696702737661238,"envelope_correlation":0.9938780596849286,"rms_gain_error_db":-0.12578119826734585,"reference_mean_power":0.0015370357837470275,"output_mean_power":0.0014931583202469451},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.498278279762008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005426794452872498,"output_power":0.0005455399019049421,"reference_active":true,"delta_db":0.022831488690691716},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.715116909567219e-05,"output_power":5.26887652586497e-05,"reference_active":true,"delta_db":-0.35307097671809623},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1787274842814485e-05,"output_power":7.809985306725948e-06,"reference_active":true,"delta_db":-1.7876319323395817}],"active_band_count":3,"active_band_mae_db":0.7211781325827898}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017379577463928091,"log_rms_mae_db":0.830298934160597,"rms_envelope_nrmse":0.060823527472209614,"envelope_correlation":0.9975531626285822,"rms_gain_error_db":-0.12457731861363638,"reference_mean_power":0.0015404251144981946,"output_mean_power":0.001496865775406024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.036061947317895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005182860743704318,"output_power":0.0005192509259240431,"reference_active":true,"delta_db":0.008077395336496243},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1854739923937534e-05,"output_power":1.9694889229780892e-05,"reference_active":true,"delta_db":-0.4519210011555464},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.514034995645174e-07,"output_power":5.438646554581447e-07,"reference_active":true,"delta_db":-1.9464459503641807}],"active_band_count":3,"active_band_mae_db":0.8021481156187411}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018667130520360047,"log_rms_mae_db":1.3083538055470236,"rms_envelope_nrmse":0.10364667409446186,"envelope_correlation":0.9936926553332646,"rms_gain_error_db":-0.2285162449001432,"reference_mean_power":0.0010622650926473529,"output_mean_power":0.0010078160855574068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.902944700544416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004039276312698456,"output_power":0.0003769511506979109,"reference_active":true,"delta_db":-0.300184893485469},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.530941238269666e-05,"output_power":3.595733102416928e-05,"reference_active":true,"delta_db":-1.0040098162402304},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869428114247812e-06,"output_power":4.7891235681211735e-06,"reference_active":true,"delta_db":-0.07221916094861412}],"active_band_count":3,"active_band_mae_db":0.4588046235581045}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017791979118134707,"log_rms_mae_db":0.7908702607930769,"rms_envelope_nrmse":0.05090396974818875,"envelope_correlation":0.9987585774442607,"rms_gain_error_db":-0.22852670695837635,"reference_mean_power":0.0010675566768630386,"output_mean_power":0.0010128339967260995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.67201138357417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000390267321515981,"output_power":0.0003624320254016939,"reference_active":true,"delta_db":-0.3213562171870144},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8376752084991498e-05,"output_power":1.4551709263718925e-05,"reference_active":true,"delta_db":-1.0135474729683565},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.299679027538036e-07,"output_power":3.8411770331636994e-07,"reference_active":true,"delta_db":-0.48971713059656663}],"active_band_count":3,"active_band_mae_db":0.6082069402506458}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":1.5711555244007431,"rms_envelope_nrmse":0.1444651103107548,"envelope_correlation":0.9859597023095643,"rms_gain_error_db":-0.33103352995128005,"reference_mean_power":0.00018921068158653402,"output_mean_power":0.00017532437410049874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9505568897447022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.310547772849948e-05,"output_power":8.132092109132419e-05,"reference_active":true,"delta_db":-0.5877294374856713},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9017487007684164e-05,"output_power":1.5012938323569134e-05,"reference_active":true,"delta_db":-1.026874278008407},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.040878232439128e-06,"output_power":3.3963290178084545e-06,"reference_active":true,"delta_db":-1.7149645105049496}],"active_band_count":3,"active_band_mae_db":1.1098560753330093}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":0.9518281744202676,"rms_envelope_nrmse":0.08869551242438388,"envelope_correlation":0.9950279460695379,"rms_gain_error_db":-0.3318048307344384,"reference_mean_power":0.00019006200677954322,"output_mean_power":0.00017608194524913035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.823509560382403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.910190725222923e-05,"output_power":7.682560815705976e-05,"reference_active":true,"delta_db":-0.6438099348752611},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361035171170283e-06,"output_power":7.574346680220246e-06,"reference_active":true,"delta_db":-0.9197869793774421},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6146687278338854e-07,"output_power":2.200324221514544e-07,"reference_active":true,"delta_db":-2.155818222658714}],"active_band_count":3,"active_band_mae_db":1.2398050456371392}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.7696444988250732,"predicted_mos":2.7696444988250732,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"fee96e003a9c39227939bae64273ab5e4cbb8c5b07ffda0d3c7a5f7921569a56","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.072138257999541},"started_at":"2026-10-13T04:42:18.175559+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6247999812767375e-05,"native_input_samples":96000,"native_input_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","output_sha256":"6a68f6247cfc9eb48bee72ca4a7cc90356a6841338133b052c48d0a911868d96","output_samples":96000,"length_delta_samples":0,"analysis_s":1.412506556000153,"synthesis_s":0.1138310359992829,"total_s":1.526337591999436,"rtf":0.2543895986665727,"output_peak":1.0425958185009225,"output_rms":0.06607669743150489,"finished_at":"2026-10-13T04:42:19.707114+00:00","evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"6a68f6247cfc9eb48bee72ca4a7cc90356a6841338133b052c48d0a911868d96","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.25148977123253,"spectral_convergence":0.6150746051087976,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.024738792669587,"spectral_convergence":0.2638227399385451,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.115301492447593,"spectral_convergence":0.31267041815607993,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.578086506106792e-05,"delta_db":0.30157955801007746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.0005795551222167431,"delta_db":1.3084507484679544,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0020907629973694448,"delta_db":1.4283218660056716,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.0010873170179391135,"delta_db":0.522273503368577,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.0004768527432885444,"delta_db":0.6041400731025137,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":8.586119757973194e-05,"delta_db":-0.0044788943018288655,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6948741072094372,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.033774435382247,"pitch":null,"output_rms_dbfs":-23.599033430919455,"output_peak":1.0425958185009225,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4638433521165695},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.25148977123253,"spectral_convergence":0.6150746051087976,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.024738792669587,"spectral_convergence":0.2638227399385451,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.115301492447593,"spectral_convergence":0.31267041815607993,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.578086506106792e-05,"delta_db":0.30157955801007746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.0005795551222167431,"delta_db":1.3084507484679544,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0020907629973694448,"delta_db":1.4283218660056716,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.0010873170179391135,"delta_db":0.522273503368577,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.0004768527432885444,"delta_db":0.6041400731025137,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":8.586119757973194e-05,"delta_db":-0.0044788943018288655,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6948741072094372,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.033774435382247,"pitch":null,"output_rms_dbfs":-23.599033430919455,"output_peak":1.0425958185009225,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4638433521165695},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9334561245547944,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0586623702554101,"rms_dbfs":-24.632807866301704,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.232068021857707,"mean":8.900610605875651e-05,"zero_energy":false},"output":{"rms":0.06607669743150489,"rms_dbfs":-23.599033430919455,"peak":1.0425958185009225,"peak_dbfs":0.3623195757252978,"crest_factor_db":23.961353006644757,"mean":2.971279566884194e-06,"zero_energy":false},"rms_difference_db":1.0337744353822487},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":5,"at_or_above_full_scale_fraction":5.208333333333334e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":6.25e-05,"longest_full_scale_run_samples":1,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":3.2942515322416694e-06,"output_rms_dbfs":-54.82243243389093,"output_zero_energy":false,"output_peak":0.027915540307087476,"output_energy_fraction":0.0002603030128150792,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":260,"both_estimated_voiced_frames":232,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9169960474308301,"absolute_difference_cents_median":4.8380334422024935,"absolute_difference_cents_p95":23.787710472704926,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027067935709258154,"log_rms_mae_db":1.7694662416977593,"rms_envelope_nrmse":0.18128309139689427,"envelope_correlation":0.9892615353435682,"rms_gain_error_db":1.0348597071275043,"reference_mean_power":0.0034432187972516922,"output_mean_power":0.00436968963246713},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.845747823685581e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011374442611865624,"output_power":0.0013904623204689793,"reference_active":true,"delta_db":0.8722910072652562},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011129046974770975,"output_power":0.00013631742986941548,"reference_active":true,"delta_db":0.8809341376910874},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3561368721227545e-05,"output_power":2.631503130915349e-05,"reference_active":true,"delta_db":0.4800337529346344}],"active_band_count":3,"active_band_mae_db":0.7444196326303261}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026426858408099,"log_rms_mae_db":1.2902919733507496,"rms_envelope_nrmse":0.1533186533616241,"envelope_correlation":0.9949535705435769,"rms_gain_error_db":1.035468393924999,"reference_mean_power":0.0034555022621815964,"output_mean_power":0.004385892886531616},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7242545404192642e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010676968681649736,"output_power":0.0013095351120985589,"reference_active":true,"delta_db":0.8866917867490708},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4544672827083896e-05,"output_power":5.521491972156363e-05,"reference_active":true,"delta_db":0.9326067160533397},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7651139324996662e-06,"output_power":1.7743616826453176e-06,"reference_active":true,"delta_db":0.022694073551569147}],"active_band_count":3,"active_band_mae_db":0.6139975254513265}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001806334260766054,"log_rms_mae_db":1.503426180824865,"rms_envelope_nrmse":0.20918598703975028,"envelope_correlation":0.9925047292963459,"rms_gain_error_db":1.3944138004182052,"reference_mean_power":0.0015370357837470275,"output_mean_power":0.0021189726853827274},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.498278279762008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005426794452872498,"output_power":0.0007443273100174864,"reference_active":true,"delta_db":1.3722058152284022},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.715116909567219e-05,"output_power":7.39729030200067e-05,"reference_active":true,"delta_db":1.1204754381347224},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1787274842814485e-05,"output_power":1.5361025695579688e-05,"reference_active":true,"delta_db":1.1500680557973566}],"active_band_count":3,"active_band_mae_db":1.2142497697201604}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017379577463928091,"log_rms_mae_db":1.349797153597917,"rms_envelope_nrmse":0.19354864115691975,"envelope_correlation":0.9958622620018234,"rms_gain_error_db":1.3960413157294727,"reference_mean_power":0.0015404251144981946,"output_mean_power":0.002124441233773681},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.036061947317895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005182860743704318,"output_power":0.0007090594762196704,"reference_active":true,"delta_db":1.3611312541791434},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1854739923937534e-05,"output_power":2.861066468274414e-05,"reference_active":true,"delta_db":1.169823047890251},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.514034995645174e-07,"output_power":9.8499461637189e-07,"reference_active":true,"delta_db":0.6329842585970245}],"active_band_count":3,"active_band_mae_db":1.0546461868888064}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018667130520360047,"log_rms_mae_db":1.2683317932861484,"rms_envelope_nrmse":0.13946090035207115,"envelope_correlation":0.991901120422884,"rms_gain_error_db":0.6065438131940948,"reference_mean_power":0.0010622650926473529,"output_mean_power":0.0012214826189539587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.902944700544416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004039276312698456,"output_power":0.0004570682416542797,"reference_active":true,"delta_db":0.5367748362918972},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.530941238269666e-05,"output_power":5.206139915110451e-05,"reference_active":true,"delta_db":0.6032740574758256},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869428114247812e-06,"output_power":5.867188704269017e-06,"reference_active":true,"delta_db":0.8095209762574723}],"active_band_count":3,"active_band_mae_db":0.6498566233417317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017791979118134707,"log_rms_mae_db":0.8780800663342065,"rms_envelope_nrmse":0.09311311375826707,"envelope_correlation":0.9979696484948097,"rms_gain_error_db":0.6065861135018824,"reference_mean_power":0.0010675566768630386,"output_mean_power":0.0012275792884568993},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.67201138357417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000390267321515981,"output_power":0.0004424425754851247,"reference_active":true,"delta_db":0.5449472394825505},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8376752084991498e-05,"output_power":2.1554004826261648e-05,"reference_active":true,"delta_db":0.6925921932585455},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.299679027538036e-07,"output_power":3.9560173030771907e-07,"reference_active":true,"delta_db":-0.36177854089050715}],"active_band_count":3,"active_band_mae_db":0.5331059912105344}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":1.628886466443381,"rms_envelope_nrmse":0.16154846228925926,"envelope_correlation":0.9840937387462283,"rms_gain_error_db":0.4055820779844776,"reference_mean_power":0.00018921068158653402,"output_mean_power":0.00020773221789130563},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9505568897447022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.310547772849948e-05,"output_power":9.491953461550626e-05,"reference_active":true,"delta_db":0.08380367400249883},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9017487007684164e-05,"output_power":1.9539177015113078e-05,"reference_active":true,"delta_db":0.1175313917405098},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.040878232439128e-06,"output_power":4.799410280274444e-06,"reference_active":true,"delta_db":-0.21318329311921674}],"active_band_count":3,"active_band_mae_db":0.13817278628740845}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":1.0485185258297707,"rms_envelope_nrmse":0.10609255717390487,"envelope_correlation":0.9935656063621269,"rms_gain_error_db":0.40586002440340796,"reference_mean_power":0.00019006200677954322,"output_mean_power":0.00020868023297880103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.823509560382403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.910190725222923e-05,"output_power":9.090857828773704e-05,"reference_active":true,"delta_db":0.08717865583974942},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361035171170283e-06,"output_power":9.76896281476254e-06,"reference_active":true,"delta_db":0.18524579506631883},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6146687278338854e-07,"output_power":3.258725388948264e-07,"reference_active":true,"delta_db":-0.4502073759098624}],"active_band_count":3,"active_band_mae_db":0.24087727560531022}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"facodec","status":"measured","mos":{"status":"measured","value":3.6139473915100098,"predicted_mos":3.6139473915100098,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1d0e527835af90f24e4fc55c66e8e889532c1845a51582a4b5dee94baf8a84fd","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.907866929999727},"started_at":"2026-10-13T04:53:05.259638+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1597000340989325e-05,"native_input_samples":96000,"native_input_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","output_sha256":"f7f167e3d0e6c82d82cf50189924f3ab7b5792c8eb46a39135ea79ae3ba74c3d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6052435880001212,"synthesis_s":5.828565623000031,"total_s":7.433809211000153,"rtf":1.2389682018333588,"output_peak":0.7519838809967041,"output_rms":0.05184868000071358,"finished_at":"2026-10-13T04:53:12.699028+00:00","evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"f7f167e3d0e6c82d82cf50189924f3ab7b5792c8eb46a39135ea79ae3ba74c3d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.229116178990985,"spectral_convergence":0.2694899991060547,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.328564291344575,"spectral_convergence":0.2878255887755975,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.165027179424422,"spectral_convergence":0.31080789287188076,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.523392017002934e-05,"delta_db":0.2493818833336285,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.00030090028901798375,"delta_db":-1.538270688269681,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0010909566374822347,"delta_db":-1.3966512976223713,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.000842406038354829,"delta_db":-0.5860736382167333,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.0003427296744094264,"delta_db":-0.8301856974046686,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":6.605905838189273e-05,"delta_db":-1.1431245220813226,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9572812878214009,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0724380477230322,"pitch":null,"output_rms_dbfs":-25.705245914024736,"output_peak":0.7519838809967041,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.240902549919994},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.229116178990985,"spectral_convergence":0.2694899991060547,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.328564291344575,"spectral_convergence":0.2878255887755975,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.165027179424422,"spectral_convergence":0.31080789287188076,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.523392017002934e-05,"delta_db":0.2493818833336285,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.00030090028901798375,"delta_db":-1.538270688269681,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0010909566374822347,"delta_db":-1.3966512976223713,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.000842406038354829,"delta_db":-0.5860736382167333,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.0003427296744094264,"delta_db":-0.8301856974046686,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":6.605905838189273e-05,"delta_db":-1.1431245220813226,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9572812878214009,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0724380477230322,"pitch":null,"output_rms_dbfs":-25.705245914024736,"output_peak":0.7519838809967041,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.240902549919994},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.91260039533592,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0586623702554101,"rms_dbfs":-24.632807866301704,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.232068021857707,"mean":8.900610605875651e-05,"zero_energy":false},"output":{"rms":0.05184868000071358,"rms_dbfs":-25.705245914024736,"peak":0.7519838809967041,"peak_dbfs":-2.4758293708589756,"crest_factor_db":23.22941654316576,"mean":1.3026191569071402e-05,"zero_energy":false},"rms_difference_db":-1.0724380477230326},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":9.478054579602087e-07,"output_rms_dbfs":-60.23280794734343,"output_zero_energy":false,"output_peak":0.0385909229516983,"output_energy_fraction":0.00012163621336555648,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":239,"both_estimated_voiced_frames":226,"reference_voiced_output_unvoiced_frames":27,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.8932806324110671,"absolute_difference_cents_median":4.800902542206856,"absolute_difference_cents_p95":15.190551798212105,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027067935709258154,"log_rms_mae_db":1.389057674871924,"rms_envelope_nrmse":0.16099079609106245,"envelope_correlation":0.9899164892359823,"rms_gain_error_db":-1.073570902462425,"reference_mean_power":0.0034432187972516922,"output_mean_power":0.002689103576505296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.845747823685581e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011374442611865624,"output_power":0.0009666548905515424,"reference_active":true,"delta_db":-0.7065867138664192},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011129046974770975,"output_power":9.565251105171548e-05,"reference_active":true,"delta_db":-0.6576160000697298},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3561368721227545e-05,"output_power":1.9498514498269684e-05,"reference_active":true,"delta_db":-0.8219899004312116}],"active_band_count":3,"active_band_mae_db":0.7287308714557869}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026426858408099,"log_rms_mae_db":1.2266997866936005,"rms_envelope_nrmse":0.15472630606864615,"envelope_correlation":0.9910768005048837,"rms_gain_error_db":-1.077854629731131,"reference_mean_power":0.0034555022621815964,"output_mean_power":0.0026960361964069256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7242545404192642e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010676968681649736,"output_power":0.0009083756029119428,"reference_active":true,"delta_db":-0.7018250737204095},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4544672827083896e-05,"output_power":3.809498667435051e-05,"reference_active":true,"delta_db":-0.6792794747262305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7651139324996662e-06,"output_power":1.3424873348369794e-06,"reference_active":true,"delta_db":-1.1886254574609794}],"active_band_count":3,"active_band_mae_db":0.856576668635873}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001806334260766054,"log_rms_mae_db":1.3376536745276144,"rms_envelope_nrmse":0.1826633480256598,"envelope_correlation":0.9900593667203705,"rms_gain_error_db":-1.336538485490365,"reference_mean_power":0.0015370357837470275,"output_mean_power":0.0011298742988130773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.498278279762008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005426794452872498,"output_power":0.0004150780141886071,"reference_active":true,"delta_db":-1.1641364208391594},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.715116909567219e-05,"output_power":4.771430879394668e-05,"reference_active":true,"delta_db":-0.7837648196852082},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1787274842814485e-05,"output_power":1.07594065273351e-05,"reference_active":true,"delta_db":-0.39625093093177144}],"active_band_count":3,"active_band_mae_db":0.7813840571520463}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017379577463928091,"log_rms_mae_db":1.2359778640367542,"rms_envelope_nrmse":0.1757234099723316,"envelope_correlation":0.9917833302502631,"rms_gain_error_db":-1.3458433323934902,"reference_mean_power":0.0015404251144981946,"output_mean_power":0.0011299422748763405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.036061947317895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005182860743704318,"output_power":0.0003969244803251175,"reference_active":true,"delta_db":-1.1586165525069583},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1854739923937534e-05,"output_power":1.84249474048448e-05,"reference_active":true,"delta_db":-0.7413938584721496},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.514034995645174e-07,"output_power":7.771755965663217e-07,"reference_active":true,"delta_db":-0.3961627571808072}],"active_band_count":3,"active_band_mae_db":0.7653910560533049}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018667130520360047,"log_rms_mae_db":1.2637769655432445,"rms_envelope_nrmse":0.14709889088389375,"envelope_correlation":0.9877415399426203,"rms_gain_error_db":-0.44834289045477826,"reference_mean_power":0.0010622650926473529,"output_mean_power":0.0009580730440572367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.902944700544416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004039276312698456,"output_power":0.0004198860827099013,"reference_active":true,"delta_db":0.1682791726876231},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.530941238269666e-05,"output_power":4.5292154729246776e-05,"reference_active":true,"delta_db":-0.001654475522540286},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869428114247812e-06,"output_power":6.502487564257975e-06,"reference_active":true,"delta_db":1.2560157142661992}],"active_band_count":3,"active_band_mae_db":0.47531645415878754}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017791979118134707,"log_rms_mae_db":1.1277416983503226,"rms_envelope_nrmse":0.1341211935592873,"envelope_correlation":0.9897105892145533,"rms_gain_error_db":-0.4485193819813712,"reference_mean_power":0.0010675566768630386,"output_mean_power":0.0009628064763707761},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.67201138357417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000390267321515981,"output_power":0.0004066274855737334,"reference_active":true,"delta_db":0.17834543223153143},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8376752084991498e-05,"output_power":1.800142462630075e-05,"reference_active":true,"delta_db":-0.08961880067011874},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.299679027538036e-07,"output_power":4.929458323475764e-07,"reference_active":true,"delta_db":0.5936316262307338}],"active_band_count":3,"active_band_mae_db":0.2871986197107947}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":1.8480914946313791,"rms_envelope_nrmse":0.23968650009795556,"envelope_correlation":0.9720368904084282,"rms_gain_error_db":-1.450790100887092,"reference_mean_power":0.00018921068158653402,"output_mean_power":0.00013547733347182058},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9505568897447022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.310547772849948e-05,"output_power":6.67054982757464e-05,"reference_active":true,"delta_db":-1.448136002285654},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9017487007684164e-05,"output_power":1.7042438885747197e-05,"reference_active":true,"delta_db":-0.47621382935449486},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.040878232439128e-06,"output_power":4.3457495630358795e-06,"reference_active":true,"delta_db":-0.6444151156579936}],"active_band_count":3,"active_band_mae_db":0.8562549824327141}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":1.5373875409398425,"rms_envelope_nrmse":0.21931300269976353,"envelope_correlation":0.9784143139427997,"rms_gain_error_db":-1.4520526675228917,"reference_mean_power":0.00019006200677954322,"output_mean_power":0.00013604733650902145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.823509560382403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.910190725222923e-05,"output_power":6.504416419786944e-05,"reference_active":true,"delta_db":-1.3667866285664252},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361035171170283e-06,"output_power":8.18634559473681e-06,"reference_active":true,"delta_db":-0.5823380210137454},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6146687278338854e-07,"output_power":3.2847465614702235e-07,"reference_active":true,"delta_db":-0.41566635152686443}],"active_band_count":3,"active_band_mae_db":0.7882636670356784}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.2242581844329834,"predicted_mos":3.2242581844329834,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"8ea8c5a08c30d3ae924ba774c9526e16ddba000f3dcc309e4c0ff125f6fcdf3d","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.250577003000217},"started_at":"2026-10-13T07:15:37.285796+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.025221646999852965,"native_input_samples":264600,"native_input_sha256":"97a2223c1ea0d55f1232167fda51a165a623515edd3f8d1f4725f6ec60073e2c","output_sha256":"d603d2755ba3bd3a28bf762ceb1d8448a315e08ccab53c434fe9f6254ee29eda","output_samples":264704,"length_delta_samples":104,"analysis_s":0.19482214500021655,"synthesis_s":38.78783179099992,"total_s":38.98265393600013,"rtf":6.497108989333356,"output_peak":0.6882467865943909,"output_rms":0.05603770973543924,"finished_at":"2026-10-13T07:16:16.324456+00:00","evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"d603d2755ba3bd3a28bf762ceb1d8448a315e08ccab53c434fe9f6254ee29eda","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.024502904790005,"output_peak":0.6847262537364157},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.630324708631155,"spectral_convergence":0.5223043144928745,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.411110526153932,"spectral_convergence":0.5044017993108457,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.049699084771165,"spectral_convergence":0.6370108715916959,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":3.2896804619573834e-05,"delta_db":-1.1337232343595942,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.0003364827285139872,"delta_db":-1.0528689669288855,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0012767716893336391,"delta_db":-0.7135937445851116,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.0010981341945593337,"delta_db":0.5652658012841489,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.0003072223600585313,"delta_db":-1.3051745707323315,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":9.358014837789166e-05,"delta_db":0.3693889836777245,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.856669216927966,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.390864220910506,"pitch":null,"output_rms_dbfs":-25.02367208721221,"output_peak":0.6847262537364157,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.363711439852084},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7788428604227219,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0586623702554101,"rms_dbfs":-24.632807866301704,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.232068021857707,"mean":8.900610605875651e-05,"zero_energy":false},"output":{"rms":0.0560810834904512,"rms_dbfs":-25.02367208721221,"peak":0.6847262537364157,"peak_dbfs":-3.2896604023654024,"crest_factor_db":21.734011684846806,"mean":-2.898017424746448e-05,"zero_energy":false},"rms_difference_db":-0.39086422091050466},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":1.42160775343054e-06,"output_rms_dbfs":-58.47220216559735,"output_zero_energy":false,"output_peak":0.044852254724553245,"output_energy_fraction":0.0001559430726762087,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":200,"both_estimated_voiced_frames":193,"reference_voiced_output_unvoiced_frames":60,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.7628458498023716,"absolute_difference_cents_median":7.762397336806148,"absolute_difference_cents_p95":46.749319757973126,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027067935709258154,"log_rms_mae_db":1.7124874731646547,"rms_envelope_nrmse":0.15327273311907638,"envelope_correlation":0.984165716911257,"rms_gain_error_db":-0.39040204874308004,"reference_mean_power":0.0034432187972516922,"output_mean_power":0.0031472005298175997},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.845747823685581e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011374442611865624,"output_power":0.0009962510266384515,"reference_active":true,"delta_db":-0.5756134193102376},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011129046974770975,"output_power":0.00010071028938996413,"reference_active":true,"delta_db":-0.43384131626807537},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3561368721227545e-05,"output_power":2.1860149092181815e-05,"reference_active":true,"delta_db":-0.32547396145287577}],"active_band_count":3,"active_band_mae_db":0.4449762323437296}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026426858408099,"log_rms_mae_db":1.3137328093168197,"rms_envelope_nrmse":0.11124974330090298,"envelope_correlation":0.9920410485202901,"rms_gain_error_db":-0.3893217757680345,"reference_mean_power":0.0034555022621815964,"output_mean_power":0.003159213699880912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7242545404192642e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010676968681649736,"output_power":0.0009454681502603585,"reference_active":true,"delta_db":-0.5280106536349514},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4544672827083896e-05,"output_power":3.947026416035678e-05,"reference_active":true,"delta_db":-0.5252574048237407},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7651139324996662e-06,"output_power":1.2876183062981241e-06,"reference_active":true,"delta_db":-1.36985600459151}],"active_band_count":3,"active_band_mae_db":0.807708021016734}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001806334260766054,"log_rms_mae_db":2.1488110589218095,"rms_envelope_nrmse":0.16511811652525737,"envelope_correlation":0.9855880198130539,"rms_gain_error_db":-0.7724177023677726,"reference_mean_power":0.0015370357837470275,"output_mean_power":0.0012865960335985069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.498278279762008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005426794452872498,"output_power":0.00045251604514366444,"reference_active":true,"delta_db":-0.789093895397122},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.715116909567219e-05,"output_power":5.903965507942592e-05,"reference_active":true,"delta_db":0.1411869275739228},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1787274842814485e-05,"output_power":1.2289375564085494e-05,"reference_active":true,"delta_db":0.18116406439710617}],"active_band_count":3,"active_band_mae_db":0.37048162912271704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017379577463928091,"log_rms_mae_db":1.6338258148933342,"rms_envelope_nrmse":0.1254522915861211,"envelope_correlation":0.9934659974951479,"rms_gain_error_db":-0.7637525007383041,"reference_mean_power":0.0015404251144981946,"output_mean_power":0.001292008409723254},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.036061947317895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005182860743704318,"output_power":0.00043736501152143116,"reference_active":true,"delta_db":-0.7372550285461978},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1854739923937534e-05,"output_power":2.216235647331526e-05,"reference_active":true,"delta_db":0.060702934035137227},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.514034995645174e-07,"output_power":8.30203317738058e-07,"reference_active":true,"delta_db":-0.10950966330966526}],"active_band_count":3,"active_band_mae_db":0.30248920863033346}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018667130520360047,"log_rms_mae_db":2.276780064056934,"rms_envelope_nrmse":0.309236594375806,"envelope_correlation":0.9460733182263023,"rms_gain_error_db":0.41488929648075606,"reference_mean_power":0.0010622650926473529,"output_mean_power":0.0011687505813488964},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.902944700544416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004039276312698456,"output_power":0.0003907966925794366,"reference_active":true,"delta_db":-0.1435268323765494},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.530941238269666e-05,"output_power":5.28148289720244e-05,"reference_active":true,"delta_db":0.6656744791607385},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869428114247812e-06,"output_power":1.0404519459600393e-05,"reference_active":true,"delta_db":3.2974406792848505}],"active_band_count":3,"active_band_mae_db":1.3688806636073796}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017791979118134707,"log_rms_mae_db":1.9413176651925126,"rms_envelope_nrmse":0.2582030157949747,"envelope_correlation":0.9614706344378275,"rms_gain_error_db":0.41489238902609166,"reference_mean_power":0.0010675566768630386,"output_mean_power":0.001174573450466129},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.67201138357417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000390267321515981,"output_power":0.00038309668480690664,"reference_active":true,"delta_db":-0.08053794010123813},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8376752084991498e-05,"output_power":2.4328537807273543e-05,"reference_active":true,"delta_db":1.2184725132678977},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.299679027538036e-07,"output_power":6.502474020106073e-07,"reference_active":true,"delta_db":1.7964258913441107}],"active_band_count":3,"active_band_mae_db":1.0318121149044155}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":2.7355134843746516,"rms_envelope_nrmse":0.2825821382080743,"envelope_correlation":0.9511500239480156,"rms_gain_error_db":-1.2452410303092802,"reference_mean_power":0.00018921068158653402,"output_mean_power":0.00014204355959260515},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9505568897447022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.310547772849948e-05,"output_power":7.469973129810977e-05,"reference_active":true,"delta_db":-0.9565619321129658},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9017487007684164e-05,"output_power":1.4405168512387902e-05,"reference_active":true,"delta_db":-1.2063478518211337},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.040878232439128e-06,"output_power":4.205257758057316e-06,"reference_active":true,"delta_db":-0.7871358604037291}],"active_band_count":3,"active_band_mae_db":0.9833485481126095}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":2.2375137338367694,"rms_envelope_nrmse":0.251663525563442,"envelope_correlation":0.9618467985069661,"rms_gain_error_db":-1.2438118202776736,"reference_mean_power":0.00019006200677954322,"output_mean_power":0.00014272962621387228},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.823509560382403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.910190725222923e-05,"output_power":7.168094134664571e-05,"reference_active":true,"delta_db":-0.9448330029230888},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361035171170283e-06,"output_power":6.673303117609322e-06,"reference_active":true,"delta_db":-1.4698302496468627},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6146687278338854e-07,"output_power":2.6349873026979677e-07,"reference_active":true,"delta_db":-1.372899750293897}],"active_band_count":3,"active_band_mae_db":1.2625210009546162}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"golf","status":"measured","mos":{"status":"measured","value":2.863675594329834,"predicted_mos":2.863675594329834,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"389c6d37f0ab12eb937fdd90cc2a6359641417b5cfcea92c91342a7b39bf4bf1","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":11.705507144000876},"started_at":"2026-10-13T05:02:36.854359+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003112628999588196,"native_input_samples":144000,"native_input_sha256":"c6b2b88165dbed40bf46385c411591f623e6aecb55de268285c9381b72d2cac5","output_sha256":"d0684486d578652a0bc275bdac35caeda42318e025241e812ab2eb34e3e728a8","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6623074499993891,"synthesis_s":0.14243684700068115,"total_s":0.8047442970000702,"rtf":0.1341240495000117,"output_peak":0.8749284744262695,"output_rms":0.05427554305120427,"finished_at":"2026-10-13T05:02:37.670989+00:00","analysis_stages":{"external_f0_s":0.08787746199959656},"evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"d0684486d578652a0bc275bdac35caeda42318e025241e812ab2eb34e3e728a8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.302187295376566,"output_peak":0.8753554775157487},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.46248810458069,"spectral_convergence":0.5468590087763348,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.165323345328299,"spectral_convergence":0.2861367521717972,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.093073186003503,"spectral_convergence":0.3520975402496943,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.2687221161777476e-05,"output_power":3.9569553788192976e-05,"delta_db":-0.3293673126047293,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004293086279643765,"output_power":0.00037648221867421397,"delta_db":-0.570251475226855,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001503876604305597,"output_power":0.001274995270558194,"delta_db":-0.7170362928916749,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009657246416819071,"output_power":0.0008888030003389011,"delta_db":-0.3604780113156326,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0004155435384154097,"output_power":0.0003107431076206659,"delta_db":-1.2621502890938006,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.60944371176991e-05,"output_power":5.913009295153351e-05,"delta_db":-1.631665292724585,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.811824778976213,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6718540266342109,"pitch":null,"output_rms_dbfs":-25.302187295376566,"output_peak":0.8753554775157487,"reference_rms_dbfs":-24.630333268742355,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5736282119708305},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9226647740839455,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.058679085461915366,"rms_dbfs":-24.630333268742355,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.229593424298358,"mean":9.924453964615505e-05,"zero_energy":false},"output":{"rms":0.05431135465012543,"rms_dbfs":-25.302187295376566,"peak":0.8753554775157487,"peak_dbfs":-1.1563109264698568,"crest_factor_db":24.14587636890671,"mean":0.0002627400065297788,"zero_energy":false},"rms_difference_db":-0.671854026634211},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":8.879118454781116e-07,"output_rms_dbfs":-60.516301501255356,"output_zero_energy":false,"output_peak":0.01555960497093413,"output_energy_fraction":0.00010402365339474601,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":252,"output_estimated_voiced_frames":255,"both_estimated_voiced_frames":234,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.9285714285714286,"absolute_difference_cents_median":5.672144135014596,"absolute_difference_cents_p95":19.026642450704188,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00027067935709258154,"log_rms_mae_db":2.355297609084749,"rms_envelope_nrmse":0.14554426758235017,"envelope_correlation":0.9881733766167129,"rms_gain_error_db":-0.6724079776512941,"reference_mean_power":0.003445404002172779,"output_mean_power":0.0029512048510478206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.849102573271394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011394548808906818,"output_power":0.0009753595633080807,"reference_active":true,"delta_db":-0.6753238651122012},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011108466767507383,"output_power":0.00010108421821465682,"reference_active":true,"delta_db":-0.4097076350754},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3451445234954316e-05,"output_power":2.074424987100644e-05,"reference_active":true,"delta_db":-0.5327187700112774}],"active_band_count":3,"active_band_mae_db":0.5392500900662929}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00026426858408099,"log_rms_mae_db":2.0735935206435947,"rms_envelope_nrmse":0.1133095380083178,"envelope_correlation":0.9942670230720657,"rms_gain_error_db":-0.6726339333463484,"reference_mean_power":0.003458102746281097,"output_mean_power":0.0029619280148243327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.7274606094385995e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010693051591439456,"output_power":0.0009218724857521555,"reference_active":true,"delta_db":-0.6443080899978858},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.442656219683768e-05,"output_power":4.064377929971055e-05,"reference_active":true,"delta_db":-0.38648623662457116},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7428753185405835e-06,"output_power":1.5139205247019567e-06,"reference_active":true,"delta_db":-0.6116324286267251}],"active_band_count":3,"active_band_mae_db":0.5474755850830606}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001806334260766054,"log_rms_mae_db":1.4198710469196438,"rms_envelope_nrmse":0.1745134558108155,"envelope_correlation":0.9822646851463779,"rms_gain_error_db":-0.6587166257070842,"reference_mean_power":0.0015364190162776657,"output_mean_power":0.0013201947799663474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.511462654432164e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005436563598083986,"output_power":0.00045884847201040893,"reference_active":true,"delta_db":-0.7365518304807124},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.710428277237182e-05,"output_power":5.088468075731962e-05,"reference_active":true,"delta_db":-0.5008162701704568},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1668313079997024e-05,"output_power":1.2824498099460047e-05,"reference_active":true,"delta_db":0.41032304094318844}],"active_band_count":3,"active_band_mae_db":0.5492303805314526}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017379577463928091,"log_rms_mae_db":1.1565238412392433,"rms_envelope_nrmse":0.14266706328763432,"envelope_correlation":0.9888394836859884,"rms_gain_error_db":-0.659101394417877,"reference_mean_power":0.0015402302098788303,"output_mean_power":0.0013233523651221453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.048940950202291e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000519151779908897,"output_power":0.0004428313197490888,"reference_active":true,"delta_db":-0.6905601806403588},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.176177952249787e-05,"output_power":1.8879478519609764e-05,"reference_active":true,"delta_db":-0.61704411728505},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.44438678584171e-07,"output_power":8.253830806482588e-07,"reference_active":true,"delta_db":-0.09912555349229843}],"active_band_count":3,"active_band_mae_db":0.46890995047256906}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00018667130520360047,"log_rms_mae_db":1.4444257526493165,"rms_envelope_nrmse":0.14079081161969884,"envelope_correlation":0.9899798478546489,"rms_gain_error_db":-0.5998997308845444,"reference_mean_power":0.0010640183593011805,"output_mean_power":0.0009267426462138096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.924020106786469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004042863363462807,"output_power":0.00035028379451822775,"reference_active":true,"delta_db":-0.6226901810187331},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.532859219471574e-05,"output_power":3.678468542161896e-05,"reference_active":true,"delta_db":-0.9070518468904653},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8576008033937385e-06,"output_power":4.487647604288347e-06,"reference_active":true,"delta_db":-0.3440307540215521}],"active_band_count":3,"active_band_mae_db":0.6245909273102501}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017791979118134707,"log_rms_mae_db":1.1839195983490494,"rms_envelope_nrmse":0.09943637603814895,"envelope_correlation":0.9963449166875764,"rms_gain_error_db":-0.5998388135770373,"reference_mean_power":0.0010693286224931163,"output_mean_power":0.0009313808630253107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.692448790686184e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039046039509696987,"output_power":0.0003368865073947812,"reference_active":true,"delta_db":-0.6409337194816274},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8418192503824136e-05,"output_power":1.4906761708688641e-05,"reference_active":true,"delta_db":-0.9186369870233547},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2585549303841944e-07,"output_power":3.887465901267685e-07,"reference_active":true,"delta_db":-0.3959566085210087}],"active_band_count":3,"active_band_mae_db":0.6518424383419968}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":2.5766588738530216,"rms_envelope_nrmse":0.23115397794517767,"envelope_correlation":0.9771189921570167,"rms_gain_error_db":-1.5054969092889352,"reference_mean_power":0.00018944508922872025,"output_mean_power":0.00013394720661581732},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.952951260226867e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.298071606255369e-05,"output_power":6.6577046429372e-05,"reference_active":true,"delta_db":-1.4506836188976857},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8977857186681273e-05,"output_power":1.7407637878267667e-05,"reference_active":true,"delta_db":-0.3750733041513535},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.048221599792943e-06,"output_power":2.992384459127133e-06,"reference_active":true,"delta_db":-2.271210201283492}],"active_band_count":3,"active_band_mae_db":1.3656557081108438}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":2.191490188534321,"rms_envelope_nrmse":0.19952708496608113,"envelope_correlation":0.9873638630462358,"rms_gain_error_db":-1.5051347244272653,"reference_mean_power":0.00019032553496065933,"output_mean_power":0.00013458094909666912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8256959526217883e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.893610795484182e-05,"output_power":6.401966051850584e-05,"reference_active":true,"delta_db":-1.4276475296960502},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.322507837481567e-06,"output_power":8.829917972067262e-06,"reference_active":true,"delta_db":-0.2357608804796543},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6008658061731197e-07,"output_power":2.0947140663018725e-07,"reference_active":true,"delta_db":-2.352821877095525}],"active_band_count":3,"active_band_mae_db":1.3387434290904097}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.871249198913574,"predicted_mos":2.871249198913574,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"e22c0fccd7d18765e2ef833488730ae774ef4c8e0345d9882b713341b9d05734","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2910773100011284},"started_at":"2026-10-13T07:38:21.035019+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004455166999832727,"native_input_samples":144000,"native_input_sha256":"c6b2b88165dbed40bf46385c411591f623e6aecb55de268285c9381b72d2cac5","output_sha256":"a292fb2c4f3d4d45c72c7e907ed2be34829d9c8c73f05f2c3f07b404292a7307","output_samples":144120,"length_delta_samples":120,"analysis_s":3.024605487000372,"synthesis_s":4.839346548997128,"total_s":7.8639520359974995,"rtf":1.31065867266625,"output_peak":0.7715322375297546,"output_rms":0.055499321980952934,"finished_at":"2026-10-13T07:38:28.917867+00:00","analysis_stages":{"external_f0_s":1.5352399869989313},"evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"a292fb2c4f3d4d45c72c7e907ed2be34829d9c8c73f05f2c3f07b404292a7307","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.10960452626221,"output_peak":0.7532465455194575},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.379561458273347,"spectral_convergence":0.565745365252944,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.299070850638748,"spectral_convergence":0.30128923325537094,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.656375996462005,"spectral_convergence":0.2833914146582926,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.5306465019793535e-05,"delta_db":0.25634139308591997,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.00041831219520689996,"delta_db":-0.10749147802315956,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0013697765608674574,"delta_db":-0.4082288824191357,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.0008137238407418507,"delta_db":-0.7365179474617995,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.00037466768241091637,"delta_db":-0.44324044211610264,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":6.388519417466889e-05,"delta_db":-1.2884461195356336,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5400443771069585,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.47369462383719757,"pitch":null,"output_rms_dbfs":-25.1065024901389,"output_peak":0.7532465455194575,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.445002768458035},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8896758070003481,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0586623702554101,"rms_dbfs":-24.632807866301704,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.232068021857707,"mean":8.900610605875651e-05,"zero_energy":false},"output":{"rms":0.05554882481584635,"rms_dbfs":-25.1065024901389,"peak":0.7532465455194575,"peak_dbfs":-2.461257027550191,"crest_factor_db":22.64524546258871,"mean":-5.359886788019311e-06,"zero_energy":false},"rms_difference_db":-0.47369462383719707},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":6.086479731440664e-06,"output_rms_dbfs":-52.156338198867445,"output_zero_energy":false,"output_peak":0.1601389777706352,"output_energy_fraction":0.0006805115868607722,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":236,"both_estimated_voiced_frames":216,"reference_voiced_output_unvoiced_frames":37,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.8537549407114624,"absolute_difference_cents_median":4.689559971562474,"absolute_difference_cents_p95":27.958380105700698,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027067935709258154,"log_rms_mae_db":2.6030919509977517,"rms_envelope_nrmse":0.17693313006814523,"envelope_correlation":0.9783326069790734,"rms_gain_error_db":-0.4726872312248182,"reference_mean_power":0.0034432187972516922,"output_mean_power":0.0030881322970157552},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.845747823685581e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011374442611865624,"output_power":0.0010232926224086423,"reference_active":true,"delta_db":-0.45930280853640226},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011129046974770975,"output_power":0.00010321040489872299,"reference_active":true,"delta_db":-0.3273449372716026},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3561368721227545e-05,"output_power":1.8413983612935135e-05,"reference_active":true,"delta_db":-1.0705276344983587}],"active_band_count":3,"active_band_mae_db":0.6190584601021212}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026426858408099,"log_rms_mae_db":1.904603782519926,"rms_envelope_nrmse":0.1201364677655666,"envelope_correlation":0.9906114376243922,"rms_gain_error_db":-0.4729463269962315,"reference_mean_power":0.0034555022621815964,"output_mean_power":0.0030989641267162053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7242545404192642e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010676968681649736,"output_power":0.0009622636692200377,"reference_active":true,"delta_db":-0.45153879742732667},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4544672827083896e-05,"output_power":4.133643926121785e-05,"reference_active":true,"delta_db":-0.3246270987753722},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7651139324996662e-06,"output_power":1.349463300206986e-06,"reference_active":true,"delta_db":-1.1661166491011212}],"active_band_count":3,"active_band_mae_db":0.6474275151012734}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001806334260766054,"log_rms_mae_db":2.260247256600585,"rms_envelope_nrmse":0.19000393686163777,"envelope_correlation":0.9763434652981821,"rms_gain_error_db":-0.36434243922668225,"reference_mean_power":0.0015370357837470275,"output_mean_power":0.0014133500366744398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.498278279762008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005426794452872498,"output_power":0.000509569471412562,"reference_active":true,"delta_db":-0.27339971119009177},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.715116909567219e-05,"output_power":5.223747392509386e-05,"reference_active":true,"delta_db":-0.3904295146749556},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1787274842814485e-05,"output_power":9.853844954168962e-06,"reference_active":true,"delta_db":-0.7780768549067562}],"active_band_count":3,"active_band_mae_db":0.4806353602572678}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017379577463928091,"log_rms_mae_db":1.6850452797249662,"rms_envelope_nrmse":0.12646306957581516,"envelope_correlation":0.9896702904319035,"rms_gain_error_db":-0.3644305532394201,"reference_mean_power":0.0015404251144981946,"output_mean_power":0.001416437888525415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.036061947317895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005182860743704318,"output_power":0.0004856260464174394,"reference_active":true,"delta_db":-0.2826756814938522},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1854739923937534e-05,"output_power":1.9855530332555362e-05,"reference_active":true,"delta_db":-0.41664151288521806},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.514034995645174e-07,"output_power":7.315901717264862e-07,"reference_active":true,"delta_db":-0.6586756846877152}],"active_band_count":3,"active_band_mae_db":0.4526642930222618}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018667130520360047,"log_rms_mae_db":2.280831885555816,"rms_envelope_nrmse":0.1912472491823839,"envelope_correlation":0.9788128252548243,"rms_gain_error_db":-0.5799556071013839,"reference_mean_power":0.0010622650926473529,"output_mean_power":0.0009294742219475967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.902944700544416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004039276312698456,"output_power":0.0003510725969029527,"reference_active":true,"delta_db":-0.6090663096829916},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.530941238269666e-05,"output_power":4.097739393869731e-05,"reference_active":true,"delta_db":-0.43644094970737296},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869428114247812e-06,"output_power":5.243304242484932e-06,"reference_active":true,"delta_db":0.3212709949037099}],"active_band_count":3,"active_band_mae_db":0.4555927514313581}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017791979118134707,"log_rms_mae_db":1.5101345036140397,"rms_envelope_nrmse":0.13619389503869017,"envelope_correlation":0.9902699917759474,"rms_gain_error_db":-0.5799377223150648,"reference_mean_power":0.0010675566768630386,"output_mean_power":0.0009341081663750184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.67201138357417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000390267321515981,"output_power":0.000340067728741957,"reference_active":true,"delta_db":-0.5979676693222795},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8376752084991498e-05,"output_power":1.6487278572384225e-05,"reference_active":true,"delta_db":-0.4711978051279333},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.299679027538036e-07,"output_power":3.9528117511073167e-07,"reference_active":true,"delta_db":-0.3652990458822486}],"active_band_count":3,"active_band_mae_db":0.47815484011082043}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":2.7538741858449844,"rms_envelope_nrmse":0.24032779150483352,"envelope_correlation":0.9616944345338165,"rms_gain_error_db":-0.8200502752040312,"reference_mean_power":0.00018921068158653402,"output_mean_power":0.0001566536876306725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9505568897447022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.310547772849948e-05,"output_power":7.334610144506056e-05,"reference_active":true,"delta_db":-1.0359819803210986},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9017487007684164e-05,"output_power":1.5029819886904228e-05,"reference_active":true,"delta_db":-1.0219935207073496},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.040878232439128e-06,"output_power":3.7628335746392513e-06,"reference_active":true,"delta_db":-1.269911962881609}],"active_band_count":3,"active_band_mae_db":1.1092958213033521}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":1.8060924703635426,"rms_envelope_nrmse":0.1548972466023701,"envelope_correlation":0.9869013000752244,"rms_gain_error_db":-0.8208001851157043,"reference_mean_power":0.00019006200677954322,"output_mean_power":0.00015733135824849017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.823509560382403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.910190725222923e-05,"output_power":6.97588172697135e-05,"reference_active":true,"delta_db":-1.0628789162239185},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361035171170283e-06,"output_power":7.283031838249282e-06,"reference_active":true,"delta_db":-1.0901166846684462},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6146687278338854e-07,"output_power":2.59336016228346e-07,"reference_active":true,"delta_db":-1.4420566665682224}],"active_band_count":3,"active_band_mae_db":1.1983507558201958}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.3254740238189697,"predicted_mos":2.3254740238189697,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"5581da52ebe3cbfda0caa0797c76880784129bd771ca9d2326324e563feaa7ee","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.893176682999183},"started_at":"2026-10-13T05:10:47.474510+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030162809998728335,"native_input_samples":144000,"native_input_sha256":"c6b2b88165dbed40bf46385c411591f623e6aecb55de268285c9381b72d2cac5","output_sha256":"13ae5211434e647bddcdee5a36e4ca203003f2367c147ba70cf30ce4f4a2fd61","output_samples":144240,"length_delta_samples":240,"analysis_s":0.6525191010005074,"synthesis_s":0.13007804799963196,"total_s":0.7825971490001393,"rtf":0.1304328581666899,"output_peak":0.6812079548835754,"output_rms":0.057283757521290515,"finished_at":"2026-10-13T05:10:48.271407+00:00","analysis_stages":{"external_f0_s":0.14935129799960123},"evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"13ae5211434e647bddcdee5a36e4ca203003f2367c147ba70cf30ce4f4a2fd61","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.834325977331694,"output_peak":0.6885276713858354},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.682167083782994,"spectral_convergence":0.1366422552880847,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.1030053821053345,"spectral_convergence":0.1326689395736295,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.2678810926370865,"spectral_convergence":0.14656820973398652,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":4.1973186514595495e-05,"delta_db":-0.07554097737493064,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.00041884045848810433,"delta_db":-0.10201047387277482,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.001451702926085192,"delta_db":-0.15594866482712932,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.0009206743953288115,"delta_db":-0.20022769775075164,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.0003843803367134643,"delta_db":-0.3320911524249508,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":7.300059461165084e-05,"delta_db":-0.7091843350015368,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.262500550208679,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19447828961240654,"pitch":null,"output_rms_dbfs":-24.82728615591411,"output_peak":0.6885276713858354,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.351017852841805},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9661692360575721,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0586623702554101,"rms_dbfs":-24.632807866301704,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.232068021857707,"mean":8.900610605875651e-05,"zero_energy":false},"output":{"rms":0.05736350667228965,"rms_dbfs":-24.82728615591411,"peak":0.6885276713858354,"peak_dbfs":-3.2415720220501076,"crest_factor_db":21.585714133864002,"mean":1.4753755302206553e-06,"zero_energy":false},"rms_difference_db":-0.19447828961240532},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":2.065620476634848e-06,"output_rms_dbfs":-56.84949469865879,"output_zero_energy":false,"output_peak":0.03712331932951081,"output_energy_fraction":0.00021656997220698227,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":246,"both_estimated_voiced_frames":236,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.932806324110672,"absolute_difference_cents_median":1.3716654067917462,"absolute_difference_cents_p95":7.0998205639268175,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027067935709258154,"log_rms_mae_db":0.9260107120929508,"rms_envelope_nrmse":0.04733790667633638,"envelope_correlation":0.9987135744096552,"rms_gain_error_db":-0.19430561386536385,"reference_mean_power":0.0034432187972516922,"output_mean_power":0.00329256273963606},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.845747823685581e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011374442611865624,"output_power":0.0010926533475413313,"reference_active":true,"delta_db":-0.17447723252528025},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011129046974770975,"output_power":0.0001009502384447368,"reference_active":true,"delta_db":-0.4235062645131792},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3561368721227545e-05,"output_power":1.7771931828752163e-05,"reference_active":true,"delta_db":-1.2246587711154298}],"active_band_count":3,"active_band_mae_db":0.607547422717963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026426858408099,"log_rms_mae_db":0.5664704312675631,"rms_envelope_nrmse":0.03231021862667742,"envelope_correlation":0.9995694582821307,"rms_gain_error_db":-0.19424574753276608,"reference_mean_power":0.0034555022621815964,"output_mean_power":0.003304354297950685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7242545404192642e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010676968681649736,"output_power":0.0010272123744993747,"reference_active":true,"delta_db":-0.16787726262526237},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4544672827083896e-05,"output_power":4.094919939423543e-05,"reference_active":true,"delta_db":-0.3655035826167896},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7651139324996662e-06,"output_power":1.194635077311807e-06,"reference_active":true,"delta_db":-1.6953748044680044}],"active_band_count":3,"active_band_mae_db":0.7429185499033522}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001806334260766054,"log_rms_mae_db":0.764320600248591,"rms_envelope_nrmse":0.044905968243821266,"envelope_correlation":0.9988824346203746,"rms_gain_error_db":-0.15971918929399176,"reference_mean_power":0.0015370357837470275,"output_mean_power":0.0014815354903111736},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.498278279762008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005426794452872498,"output_power":0.0005303215171859445,"reference_active":true,"delta_db":-0.1000412392885759},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.715116909567219e-05,"output_power":5.1312778383038085e-05,"reference_active":true,"delta_db":-0.4679958822399912},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1787274842814485e-05,"output_power":9.093250976955099e-06,"reference_active":true,"delta_db":-1.126942321198456}],"active_band_count":3,"active_band_mae_db":0.5649931475756743}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017379577463928091,"log_rms_mae_db":0.4763255208075515,"rms_envelope_nrmse":0.028117385564409094,"envelope_correlation":0.99967994789751,"rms_gain_error_db":-0.15971650529649803,"reference_mean_power":0.0015404251144981946,"output_mean_power":0.0014848033545185385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.036061947317895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005182860743704318,"output_power":0.0005050914729002986,"reference_active":true,"delta_db":-0.11199503414851182},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1854739923937534e-05,"output_power":1.971799188900487e-05,"reference_active":true,"delta_db":-0.446829590285499},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.514034995645174e-07,"output_power":5.738754019074759e-07,"reference_active":true,"delta_db":-1.713178210185018}],"active_band_count":3,"active_band_mae_db":0.7573342782063429}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018667130520360047,"log_rms_mae_db":0.8852293739186992,"rms_envelope_nrmse":0.05632293165191062,"envelope_correlation":0.9983832771384118,"rms_gain_error_db":-0.2131314127817499,"reference_mean_power":0.0010622650926473529,"output_mean_power":0.001011392593589189},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.902944700544416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004039276312698456,"output_power":0.0003845438735920264,"reference_active":true,"delta_db":-0.2135766601534758},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.530941238269666e-05,"output_power":4.424851148006909e-05,"reference_active":true,"delta_db":-0.1028976424188219},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869428114247812e-06,"output_power":4.669768771280396e-06,"reference_active":true,"delta_db":-0.1818258234044603}],"active_band_count":3,"active_band_mae_db":0.16610004199225267}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017791979118134707,"log_rms_mae_db":0.5133489874874241,"rms_envelope_nrmse":0.039087361791644924,"envelope_correlation":0.9993930634186883,"rms_gain_error_db":-0.21311539283950767,"reference_mean_power":0.0010675566768630386,"output_mean_power":0.0010164345100730637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.67201138357417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000390267321515981,"output_power":0.00037255543190945503,"reference_active":true,"delta_db":-0.20171287914356678},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8376752084991498e-05,"output_power":1.81197867404638e-05,"reference_active":true,"delta_db":-0.061156744672502546},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.299679027538036e-07,"output_power":3.6937989910256647e-07,"reference_active":true,"delta_db":-0.6596277823166685}],"active_band_count":3,"active_band_mae_db":0.3074991353775793}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":1.1039576878967419,"rms_envelope_nrmse":0.10374289604820228,"envelope_correlation":0.993579963326875,"rms_gain_error_db":-0.3920944591994743,"reference_mean_power":0.00018921068158653402,"output_mean_power":0.00017287659667185993},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9505568897447022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.310547772849948e-05,"output_power":8.35110775958962e-05,"reference_active":true,"delta_db":-0.4723114513652771},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9017487007684164e-05,"output_power":1.6315766231016623e-05,"reference_active":true,"delta_db":-0.6654565402047643},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.040878232439128e-06,"output_power":4.3098178226157975e-06,"reference_active":true,"delta_db":-0.6804729395677824}],"active_band_count":3,"active_band_mae_db":0.6060803103792746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":0.6758605543441358,"rms_envelope_nrmse":0.06718269300494953,"envelope_correlation":0.9979954673772153,"rms_gain_error_db":-0.39187878657226416,"reference_mean_power":0.00019006200677954322,"output_mean_power":0.00017366305306272238},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.823509560382403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.910190725222923e-05,"output_power":8.003500729653252e-05,"reference_active":true,"delta_db":-0.466070114662376},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361035171170283e-06,"output_power":8.103134813692123e-06,"reference_active":true,"delta_db":-0.6267081254891649},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6146687278338854e-07,"output_power":2.5507363342040544e-07,"reference_active":true,"delta_db":-1.5140293327407217}],"active_band_count":3,"active_band_mae_db":0.8689358576307541}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29093-0005","method":"magphase","status":"measured","mos":{"status":"measured","value":2.6856799125671387,"predicted_mos":2.6856799125671387,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288018,"input_sample_rate":48000,"input_seconds":6.000375,"input_float32_sha256":"8cfebf47a4fff58b19515735752b41c3439d0992e97a21f2f9858d5f41a64afd","internal_samples":96006,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3273292619996937},"started_at":"2026-10-13T05:15:49.123402+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00825059999988298,"native_input_samples":288000,"native_input_sha256":"cb3c7b47784e07391b64069d47262e0b6bdedd2b622f5bd164367bb0ca1aaf75","output_sha256":"b590c091042920bdf7ab4930555526c140a92251382928f26a17db15f7b02c79","output_samples":288018,"length_delta_samples":18,"analysis_s":4.00825914699999,"synthesis_s":1.1326840700003231,"total_s":5.140943217000313,"rtf":0.8568238695000522,"output_peak":0.8030797297008561,"output_rms":0.05818326184107685,"finished_at":"2026-10-13T05:15:54.285052+00:00","evaluation_input":{"reference":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"b590c091042920bdf7ab4930555526c140a92251382928f26a17db15f7b02c79","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96006,"length_delta_samples":6,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.699327228926514,"output_peak":0.8071271482321751},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9999375039060059,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.499453698617689,"spectral_convergence":0.631599433794622,"spectral_convergence_status":"ok","amplitude_floor":1.6417184973675735e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.895504504940246,"spectral_convergence":0.47315165266146325,"spectral_convergence_status":"ok","amplitude_floor":1.2466848673527415e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.71708130071087,"spectral_convergence":0.43163940028879333,"spectral_convergence_status":"ok","amplitude_floor":1.127090317770565e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.270965260159295e-05,"output_power":3.9913545937785046e-05,"delta_db":-0.29405726489693423,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004287949616170093,"output_power":0.00039492350886029444,"delta_db":-0.35736686684765584,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015047786428193433,"output_power":0.0014190616154685066,"delta_db":-0.25471365809549806,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0009641149894002944,"output_power":0.000987911723793155,"delta_db":0.10589304315254476,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00041492564533297956,"output_power":0.00045661434825491806,"delta_db":0.41579276141051746,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.594979221160373e-05,"output_power":9.068112351944273e-05,"delta_db":0.23272061208893152,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2767573677486803,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06634078221010889,"pitch":null,"output_rms_dbfs":-24.69914864851181,"output_peak":0.8071271482321751,"reference_rms_dbfs":-24.632807866301704,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.704013168089602},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.742593370189469,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0586623702554101,"rms_dbfs":-24.632807866301704,"peak":0.676025390625,"peak_dbfs":-3.4007398444439967,"crest_factor_db":21.232068021857707,"mean":8.900610605875651e-05,"zero_energy":false},"output":{"rms":0.0582160275683089,"rms_dbfs":-24.69914864851181,"peak":0.8071271482321751,"peak_dbfs":-1.8611608935151636,"crest_factor_db":22.83798775499665,"mean":0.0005583609748930033,"zero_energy":false},"rms_difference_db":-0.06634078221010498},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.2859012696426365e-06,"threshold_dbfs":-51.37516718125659,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":207,"quiet_samples":33120,"reference_mean_power":7.804382225759508e-07,"output_mean_power":6.048215139716539e-05,"output_rms_dbfs":-42.18372769039596,"output_zero_energy":false,"output_peak":0.13179046240520476,"output_energy_fraction":0.006156887113612364,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":253,"output_estimated_voiced_frames":245,"both_estimated_voiced_frames":216,"reference_voiced_output_unvoiced_frames":37,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.8537549407114624,"absolute_difference_cents_median":5.312621650006724,"absolute_difference_cents_p95":41.48653206188542,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027067935709258154,"log_rms_mae_db":4.568308824880376,"rms_envelope_nrmse":0.3267688701456989,"envelope_correlation":0.9231119071453261,"rms_gain_error_db":-0.06589727573997053,"reference_mean_power":0.0034432187972516922,"output_mean_power":0.0033913678049133033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.845747823685581e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011374442611865624,"output_power":0.001142472122800137,"reference_active":true,"delta_db":0.019154878746138745},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011129046974770975,"output_power":0.00011174581441730528,"reference_active":true,"delta_db":0.017732895152670725},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3561368721227545e-05,"output_power":2.1870572256584037e-05,"reference_active":true,"delta_db":-0.32340369015162046}],"active_band_count":3,"active_band_mae_db":0.12009715468347665}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026426858408099,"log_rms_mae_db":3.6250224010324072,"rms_envelope_nrmse":0.24276066407472047,"envelope_correlation":0.9549514885750204,"rms_gain_error_db":-0.06512313220019368,"reference_mean_power":0.0034555022621815964,"output_mean_power":0.0034040730273324584},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7242545404192642e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010676968681649736,"output_power":0.0010761761177888708,"reference_active":true,"delta_db":0.03435381245285421},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4544672827083896e-05,"output_power":4.465428376483948e-05,"reference_active":true,"delta_db":0.010673543997226032},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7651139324996662e-06,"output_power":1.7011278585313935e-06,"reference_active":true,"delta_db":-0.16035786096079252}],"active_band_count":3,"active_band_mae_db":0.06846173913695759}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001806334260766054,"log_rms_mae_db":4.472971245187149,"rms_envelope_nrmse":0.36482655112765183,"envelope_correlation":0.9106960892826073,"rms_gain_error_db":-0.25109310471530577,"reference_mean_power":0.0015370357837470275,"output_mean_power":0.0014506901690581784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.498278279762008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005426794452872498,"output_power":0.0005166559084479371,"reference_active":true,"delta_db":-0.21341972083340408},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.715116909567219e-05,"output_power":5.294151170277416e-05,"reference_active":true,"delta_db":-0.33228780760377935},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1787274842814485e-05,"output_power":1.066911999780664e-05,"reference_active":true,"delta_db":-0.43284810297957305}],"active_band_count":3,"active_band_mae_db":0.3261852104722522}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017379577463928091,"log_rms_mae_db":3.858527798265711,"rms_envelope_nrmse":0.27215391297811947,"envelope_correlation":0.9479466901736996,"rms_gain_error_db":-0.25002081866030934,"reference_mean_power":0.0015404251144981946,"output_mean_power":0.0014542481122383885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.036061947317895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005182860743704318,"output_power":0.0004942854464365541,"reference_active":true,"delta_db":-0.20591716666551643},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1854739923937534e-05,"output_power":1.9954758962319325e-05,"reference_active":true,"delta_db":-0.3949915644111772},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.514034995645174e-07,"output_power":8.166876488553315e-07,"reference_active":true,"delta_db":-0.18079443344113097}],"active_band_count":3,"active_band_mae_db":0.26056772150594154}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018667130520360047,"log_rms_mae_db":4.011306019542812,"rms_envelope_nrmse":0.3459527655427065,"envelope_correlation":0.9285309833319702,"rms_gain_error_db":0.16394433538820025,"reference_mean_power":0.0010622650926473529,"output_mean_power":0.001103131646606901},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.902944700544416e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004039276312698456,"output_power":0.00041975249516798215,"reference_active":true,"delta_db":0.16689723683273022},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.530941238269666e-05,"output_power":4.863490237637357e-05,"reference_active":true,"delta_db":0.3075961860885707},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869428114247812e-06,"output_power":5.191349023545749e-06,"reference_active":true,"delta_db":0.2780226937236412}],"active_band_count":3,"active_band_mae_db":0.25083870554831406}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017791979118134707,"log_rms_mae_db":3.174385000655481,"rms_envelope_nrmse":0.26325573723892576,"envelope_correlation":0.9571706383040416,"rms_gain_error_db":0.16403511830135842,"reference_mean_power":0.0010675566768630386,"output_mean_power":0.0011086499785890182},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.67201138357417e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000390267321515981,"output_power":0.00040659554850197707,"reference_active":true,"delta_db":0.17800431808552553},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8376752084991498e-05,"output_power":1.994984895711577e-05,"reference_active":true,"delta_db":0.35670855495345966},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.299679027538036e-07,"output_power":4.1140640743875533e-07,"reference_active":true,"delta_db":-0.1916498524298142}],"active_band_count":3,"active_band_mae_db":0.2421209084895998}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.103910936684106e-05,"log_rms_mae_db":4.688609445950614,"rms_envelope_nrmse":0.41819394373428337,"envelope_correlation":0.886563469458525,"rms_gain_error_db":0.4646929976455788,"reference_mean_power":0.00018921068158653402,"output_mean_power":0.00021057894712447778},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9505568897447022e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.310547772849948e-05,"output_power":0.00010248508866965195,"reference_active":true,"delta_db":0.4168544836894472},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9017487007684164e-05,"output_power":1.9356650164487402e-05,"reference_active":true,"delta_db":0.07677072836509356},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.040878232439128e-06,"output_power":5.601207072064889e-06,"reference_active":true,"delta_db":0.4577542174072153}],"active_band_count":3,"active_band_mae_db":0.31712647648725206}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.917454226005244e-05,"log_rms_mae_db":3.619013320385783,"rms_envelope_nrmse":0.3043895197990237,"envelope_correlation":0.9364115608003986,"rms_gain_error_db":0.4657069194115656,"reference_mean_power":0.00019006200677954322,"output_mean_power":0.00021157580522819476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.823509560382403e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.910190725222923e-05,"output_power":9.83711503255581e-05,"reference_active":true,"delta_db":0.42980749604614393},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.361035171170283e-06,"output_power":9.259335024043103e-06,"reference_active":true,"delta_db":-0.04744078816866782},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6146687278338854e-07,"output_power":3.9098420177193275e-07,"reference_active":true,"delta_db":0.34090707680247406}],"active_band_count":3,"active_band_mae_db":0.2727184536724286}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2970845699310303,"predicted_mos":2.2970845699310303,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48640,"input_sample_rate":16000,"input_seconds":3.04,"input_float32_sha256":"9c99e577370d9d7a4eb700d06fd66fa4cb7f8736b25df747e0b6a96fc6b772c9","internal_samples":48640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1848654669993266},"started_at":"2026-10-13T04:39:23.026967+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.56619998961105e-05,"native_input_samples":48640,"native_input_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","output_sha256":"e64e50b366afdf473522371be38c2117f6329828ceffcf9cbb787f1ce0e6d219","output_samples":48640,"length_delta_samples":0,"analysis_s":0.5826491650004755,"synthesis_s":0.2567449589996613,"total_s":0.8393941240001368,"rtf":0.2761164881579397,"output_peak":0.36361484846274966,"output_rms":0.03779135624616199,"finished_at":"2026-10-13T04:39:23.870034+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"e64e50b366afdf473522371be38c2117f6329828ceffcf9cbb787f1ce0e6d219","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.614300472915331,"spectral_convergence":0.26535735257218857,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.161108543305768,"spectral_convergence":0.2682713509396772,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.235586026283735,"spectral_convergence":0.2643858504128024,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":0.00011859566111897013,"delta_db":0.26709984720399116,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.00043338596170391,"delta_db":0.12746034052821112,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.000639693373358866,"delta_db":-0.1760113130066368,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.0001543508269903212,"delta_db":-0.10001917072517266,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":4.9138181762030695e-05,"delta_db":-0.3548365305976991,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":3.302245739192909e-05,"delta_db":0.033476369256602315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17648392855305217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0434950859702323,"pitch":null,"output_rms_dbfs":-28.452150439164335,"output_peak":0.36361484846274966,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.336998347501612},"common_prefix":{"samples":48640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.614300472915331,"spectral_convergence":0.26535735257218857,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.161108543305768,"spectral_convergence":0.2682713509396772,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.235586026283735,"spectral_convergence":0.2643858504128024,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":0.00011859566111897013,"delta_db":0.26709984720399116,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.00043338596170391,"delta_db":0.12746034052821112,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.000639693373358866,"delta_db":-0.1760113130066368,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.0001543508269903212,"delta_db":-0.10001917072517266,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":4.9138181762030695e-05,"delta_db":-0.3548365305976991,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":3.302245739192909e-05,"delta_db":0.033476369256602315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17648392855305217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0434950859702323,"pitch":null,"output_rms_dbfs":-28.452150439164335,"output_peak":0.36361484846274966,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.336998347501612},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48640,"output_samples":48640,"intelligibility":{"estoi":{"value":0.9196663870354667,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03798107322302981,"rms_dbfs":-28.408655353194106,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.085568020654186,"mean":-1.847430279380397e-05,"zero_energy":false},"output":{"rms":0.03779135624616199,"rms_dbfs":-28.452150439164335,"peak":0.36361484846274966,"peak_dbfs":-8.787167807473983,"crest_factor_db":19.664982631690357,"mean":-8.730474769481144e-06,"zero_energy":false},"rms_difference_db":-0.04349508597022833},"full_scale":{"reference":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":304,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.2061320541595871e-07,"output_mean_power":7.173323903344206e-07,"output_rms_dbfs":-61.44279558644679,"output_zero_energy":false,"output_peak":0.03107112624110773,"output_energy_fraction":0.0001817417003027624,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":297,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":86,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.85,"absolute_difference_cents_median":2.8604489274301264,"absolute_difference_cents_p95":12.8195178263667,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00012692463938083332,"log_rms_mae_db":0.9280350763854825,"rms_envelope_nrmse":0.078488762342038,"envelope_correlation":0.9949639291310449,"rms_gain_error_db":-0.04349505382295679,"reference_mean_power":0.0014449383611437946,"output_mean_power":0.0014305393739721685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.0532453868977968e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005746543771154043,"output_power":0.0005546914149457591,"reference_active":true,"delta_db":-0.15355274771422597},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014747895972213286,"output_power":0.00013115214210481825,"reference_active":true,"delta_db":-0.5095467723986837},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536231970329892e-05,"output_power":2.4465730625433198e-05,"reference_active":true,"delta_db":-0.15630783001750898}],"active_band_count":3,"active_band_mae_db":0.2731357833768062}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00010829935323771927,"log_rms_mae_db":0.5802318204007476,"rms_envelope_nrmse":0.03631721984119524,"envelope_correlation":0.9988554044965622,"rms_gain_error_db":-0.043495154068556835,"reference_mean_power":0.0014593630016451321,"output_mean_power":0.0014448202378199802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":9.289432894914325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005498374692309952,"output_power":0.0005298166717085204,"reference_active":true,"delta_db":-0.16108711859152888},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8170998681841746e-05,"output_power":5.098676726208154e-05,"reference_active":true,"delta_db":-0.572490427823457},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4585911877580324e-06,"output_power":1.539362903818526e-06,"reference_active":true,"delta_db":0.23407431151683067}],"active_band_count":3,"active_band_mae_db":0.3225506193106055}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00011599347403670498,"log_rms_mae_db":0.9688191416212826,"rms_envelope_nrmse":0.09569787402317409,"envelope_correlation":0.9932459518585516,"rms_gain_error_db":-0.06015210556957311,"reference_mean_power":0.0008943859828976575,"output_mean_power":0.0008820836528805249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":8.128232732706927e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005209748261573112,"output_power":0.0005083823740526428,"reference_active":true,"delta_db":-0.1062625354347751},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010680702813776804,"output_power":9.271620890995495e-05,"reference_active":true,"delta_db":-0.6144416579457367},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2431239684221885e-05,"output_power":1.9203771601727084e-05,"reference_active":true,"delta_db":-0.6746674386530305}],"active_band_count":3,"active_band_mae_db":0.4651238773445141}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":0.5761098550195083,"rms_envelope_nrmse":0.04580570302258039,"envelope_correlation":0.998366313709408,"rms_gain_error_db":-0.0601521128748568,"reference_mean_power":0.0009033149500834516,"output_mean_power":0.0008908898001167318},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":7.323650960940425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005139241705701183,"output_power":0.0004992180016184005,"reference_active":true,"delta_db":-0.1260880618359404},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.077088099913429e-05,"output_power":3.416797712046123e-05,"reference_active":true,"delta_db":-0.7673082841552257},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2000817341803496e-06,"output_power":1.084982026721248e-06,"reference_active":true,"delta_db":-0.4378828169123556}],"active_band_count":3,"active_band_mae_db":0.44375972096784055}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":1.1787953323160614,"rms_envelope_nrmse":0.12046418404356873,"envelope_correlation":0.9906472018430259,"rms_gain_error_db":-0.09559731588018139,"reference_mean_power":0.00017342262158164613,"output_mean_power":0.00016964693413121544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.6075439307790641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.767193199363684e-05,"output_power":9.411611968540691e-05,"reference_active":true,"delta_db":-0.1610576499815812},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2709798816345444e-05,"output_power":2.0813368083465192e-05,"reference_active":true,"delta_db":-0.37870922096720305},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175922545645166e-06,"output_power":4.611090298943293e-06,"reference_active":true,"delta_db":0.4305119224480215}],"active_band_count":3,"active_band_mae_db":0.3234262644656019}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":0.6393197393554718,"rms_envelope_nrmse":0.05360422500389378,"envelope_correlation":0.9981138810872585,"rms_gain_error_db":-0.09559764400520768,"reference_mean_power":0.0001751536319883132,"output_mean_power":0.0001713402447313102},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.4903701525512372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.99755918258773e-05,"output_power":9.665820205136266e-05,"reference_active":true,"delta_db":-0.14655271183798177},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.52463948098894e-06,"output_power":8.908829125293645e-06,"reference_active":true,"delta_db":-0.2902791700945033},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.337301598444854e-07,"output_power":3.2761327873617363e-07,"reference_active":true,"delta_db":-0.08033960310411224}],"active_band_count":3,"active_band_mae_db":0.17239049501219914}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":1.4012440743523187,"rms_envelope_nrmse":0.14893081992466203,"envelope_correlation":0.984518328818672,"rms_gain_error_db":-0.3654852844011724,"reference_mean_power":3.032942326964182e-05,"output_mean_power":2.7881467187136514e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":3.170053749158531e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1550798308210335e-05,"output_power":1.035852938933849e-05,"reference_active":true,"delta_db":-0.4731389801203384},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.128379941764708e-06,"output_power":5.510136229463803e-06,"reference_active":true,"delta_db":-0.4618334622280802},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7104244593685816e-06,"output_power":9.59066181071266e-07,"reference_active":true,"delta_db":-2.5125532137936997}],"active_band_count":3,"active_band_mae_db":1.1491752187140394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":0.7907561482275234,"rms_envelope_nrmse":0.08059460520391683,"envelope_correlation":0.9960814963290857,"rms_gain_error_db":-0.36548956326926313,"reference_mean_power":3.063204792415604e-05,"output_mean_power":2.8159638579810313e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.7930999787597987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1345569218097283e-05,"output_power":1.021496039053616e-05,"reference_active":true,"delta_db":-0.4558960278427928},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6424993904119246e-06,"output_power":2.281253844854669e-06,"reference_active":true,"delta_db":-0.6384128202705184},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1775222693841555e-07,"output_power":6.754092927504642e-08,"reference_active":true,"delta_db":-2.4140209762063436}],"active_band_count":3,"active_band_mae_db":1.1694432747732184}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.8130176067352295,"predicted_mos":3.8130176067352295,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48640,"input_sample_rate":16000,"input_seconds":3.04,"input_float32_sha256":"1026650a2dba3e02269dee7da976712799ed725e18725b43bdadb69bfc673996","internal_samples":48640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.095904870000595},"started_at":"2026-10-13T04:42:19.707997+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.9600000086938962e-05,"native_input_samples":48640,"native_input_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","output_sha256":"afd84dad4db82ec1abe629ddeee7e637b8a226dd916783e7c21c3169fff6df1f","output_samples":48640,"length_delta_samples":0,"analysis_s":0.7870134689992483,"synthesis_s":0.05245077800009312,"total_s":0.8394642469993414,"rtf":0.2761395549339939,"output_peak":0.4088194960852659,"output_rms":0.04245980024330866,"finished_at":"2026-10-13T04:42:20.551047+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"afd84dad4db82ec1abe629ddeee7e637b8a226dd916783e7c21c3169fff6df1f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.62306941762492,"spectral_convergence":0.6268935901187231,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.033090825286138,"spectral_convergence":0.2649115070226318,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.207574542862142,"spectral_convergence":0.3033214105292998,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":0.00011155985489900355,"delta_db":0.0014912500698683268,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.0005010309215841135,"delta_db":0.7573572414390748,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.0009322875761085686,"delta_db":1.4597691359933864,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.00016563558598570876,"delta_db":0.20642770706570332,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":5.884151478668446e-05,"delta_db":0.42781108962122083,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":3.34785810557226e-05,"delta_db":0.09305291162009766,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.49098488930155854,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9682142861889743,"pitch":null,"output_rms_dbfs":-27.440441067005132,"output_peak":0.4088194960852659,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.621244928591065},"common_prefix":{"samples":48640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.62306941762492,"spectral_convergence":0.6268935901187231,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.033090825286138,"spectral_convergence":0.2649115070226318,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.207574542862142,"spectral_convergence":0.3033214105292998,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":0.00011155985489900355,"delta_db":0.0014912500698683268,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.0005010309215841135,"delta_db":0.7573572414390748,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.0009322875761085686,"delta_db":1.4597691359933864,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.00016563558598570876,"delta_db":0.20642770706570332,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":5.884151478668446e-05,"delta_db":0.42781108962122083,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":3.34785810557226e-05,"delta_db":0.09305291162009766,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.49098488930155854,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9682142861889743,"pitch":null,"output_rms_dbfs":-27.440441067005132,"output_peak":0.4088194960852659,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.621244928591065},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48640,"output_samples":48640,"intelligibility":{"estoi":{"value":0.9626240651933559,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03798107322302981,"rms_dbfs":-28.408655353194106,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.085568020654186,"mean":-1.847430279380397e-05,"zero_energy":false},"output":{"rms":0.04245980024330866,"rms_dbfs":-27.440441067005132,"peak":0.4088194960852659,"peak_dbfs":-7.769368028486677,"crest_factor_db":19.671073038518454,"mean":-1.2339448391139924e-10,"zero_energy":false},"rms_difference_db":0.9682142861889744},"full_scale":{"reference":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":304,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.2061320541595871e-07,"output_mean_power":3.438551484501746e-07,"output_rms_dbfs":-64.63624468699413,"output_zero_energy":false,"output_peak":0.02130790312442669,"output_energy_fraction":6.901424472763533e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":297,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":99,"both_estimated_voiced_frames":91,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.91,"absolute_difference_cents_median":7.265954196914901,"absolute_difference_cents_p95":17.848358371194394,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00012692463938083332,"log_rms_mae_db":1.417294474240573,"rms_envelope_nrmse":0.1767416756695053,"envelope_correlation":0.9873519174125858,"rms_gain_error_db":0.9682145809027849,"reference_mean_power":0.0014449383611437946,"output_mean_power":0.0018058047008575693},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.0532453868977968e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005746543771154043,"output_power":0.0007326566868550033,"reference_active":true,"delta_db":1.05493798656104},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014747895972213286,"output_power":0.00020437527124988244,"reference_active":true,"delta_db":1.4169828086118832},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536231970329892e-05,"output_power":3.711327203498285e-05,"reference_active":true,"delta_db":1.653402722249488}],"active_band_count":3,"active_band_mae_db":1.3751078391408036}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00010829935323771927,"log_rms_mae_db":1.3093454427146272,"rms_envelope_nrmse":0.14106230958362018,"envelope_correlation":0.9952862815619247,"rms_gain_error_db":0.9682153687980731,"reference_mean_power":0.0014593630016451321,"output_mean_power":0.001823832156163299},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":9.289432894914325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005498374692309952,"output_power":0.0007123820427554231,"reference_active":true,"delta_db":1.124786315746507},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8170998681841746e-05,"output_power":8.764247349951424e-05,"reference_active":true,"delta_db":1.7800810614025286},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4585911877580324e-06,"output_power":2.0124684325907683e-06,"reference_active":true,"delta_db":1.3979549139073866}],"active_band_count":3,"active_band_mae_db":1.4342740970188075}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00011599347403670498,"log_rms_mae_db":1.3071696533861945,"rms_envelope_nrmse":0.20649278345213196,"envelope_correlation":0.9884888561167188,"rms_gain_error_db":1.2639976970583553,"reference_mean_power":0.0008943859828976575,"output_mean_power":0.0011965332013480787},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":8.128232732706927e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005209748261573112,"output_power":0.0007107250638922518,"reference_active":true,"delta_db":1.3488489279854092},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010680702813776804,"output_power":0.0001610895787491849,"reference_active":true,"delta_db":1.7846761466473293},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2431239684221885e-05,"output_power":3.305867925802636e-05,"reference_active":true,"delta_db":1.6843222293824542}],"active_band_count":3,"active_band_mae_db":1.6059491013383977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":1.074258180182282,"rms_envelope_nrmse":0.17890314801553617,"envelope_correlation":0.9948880927740822,"rms_gain_error_db":1.2639977199813903,"reference_mean_power":0.0009033149500834516,"output_mean_power":0.0012084786159686047},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":7.323650960940425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005139241705701183,"output_power":0.0007069349305649169,"reference_active":true,"delta_db":1.384803976005008},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.077088099913429e-05,"output_power":6.673788635852922e-05,"reference_active":true,"delta_db":2.1402235206558102},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2000817341803496e-06,"output_power":1.8297163425907638e-06,"reference_active":true,"delta_db":1.8317294148307075}],"active_band_count":3,"active_band_mae_db":1.7855856371638419}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":1.129860977148955,"rms_envelope_nrmse":0.16629202175424732,"envelope_correlation":0.9839721033736253,"rms_gain_error_db":0.35058545765472837,"reference_mean_power":0.00017342262158164613,"output_mean_power":0.00018800278934237376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.6075439307790641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.767193199363684e-05,"output_power":0.00010567836055129834,"reference_active":true,"delta_db":0.3421628886164821},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2709798816345444e-05,"output_power":2.5076639468180328e-05,"reference_active":true,"delta_db":0.43056049188820916},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175922545645166e-06,"output_power":4.8576213974500204e-06,"reference_active":true,"delta_db":0.6567122807435586}],"active_band_count":3,"active_band_mae_db":0.47647855374941656}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":0.7682869117395873,"rms_envelope_nrmse":0.08039210947553363,"envelope_correlation":0.9969443620116433,"rms_gain_error_db":0.35058545139436265,"reference_mean_power":0.0001751536319883132,"output_mean_power":0.00018987933079007128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.4903701525512372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.99755918258773e-05,"output_power":0.00010857043853521496,"reference_active":true,"delta_db":0.3581760831586925},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.52463948098894e-06,"output_power":1.0848098276087728e-05,"reference_active":true,"delta_db":0.5650506484873932},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.337301598444854e-07,"output_power":3.0966865692835943e-07,"reference_active":true,"delta_db":-0.32498205943114666}],"active_band_count":3,"active_band_mae_db":0.41606959702574403}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":1.3656143433849088,"rms_envelope_nrmse":0.1805137834188931,"envelope_correlation":0.978574912572663,"rms_gain_error_db":0.3679015991727636,"reference_mean_power":3.032942326964182e-05,"output_mean_power":3.301066794003255e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":3.170053749158531e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1550798308210335e-05,"output_power":1.2331153736846547e-05,"reference_active":true,"delta_db":0.28391711709054035},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.128379941764708e-06,"output_power":7.31115815189975e-06,"reference_active":true,"delta_db":0.7664049602539652},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7104244593685816e-06,"output_power":1.5408453638081133e-06,"reference_active":true,"delta_db":-0.45344842425627857}],"active_band_count":3,"active_band_mae_db":0.501256833866928}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":0.8532116852993318,"rms_envelope_nrmse":0.09762005256173413,"envelope_correlation":0.994194303916508,"rms_gain_error_db":0.36790564170084894,"reference_mean_power":3.063204792415604e-05,"output_mean_power":3.334007688162182e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.7930999787597987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1345569218097283e-05,"output_power":1.2119320148321047e-05,"reference_active":true,"delta_db":0.2865196836613696},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6424993904119246e-06,"output_power":3.4451506072485742e-06,"reference_active":true,"delta_db":1.1519331636203325},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1775222693841555e-07,"output_power":1.0092138836669724e-07,"reference_active":true,"delta_db":-0.6698591284868515}],"active_band_count":3,"active_band_mae_db":0.7027706585895178}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"facodec","status":"measured","mos":{"status":"measured","value":4.4806318283081055,"predicted_mos":4.4806318283081055,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48800,"input_sample_rate":16000,"input_seconds":3.05,"input_float32_sha256":"a2ffb3903bb76c0fde53df9c8d39a5d9d845d09bb62295066f667eb46b9614b9","internal_samples":48800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.474628867001229},"started_at":"2026-10-13T04:53:12.699549+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0453000615816563e-05,"native_input_samples":48640,"native_input_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","output_sha256":"78022bc5c2d44c2a1e042327f72331142b5b1edf521d9b9a048cd3b0c87ea3f5","output_samples":48800,"length_delta_samples":160,"analysis_s":0.7463391460005369,"synthesis_s":5.577265561999411,"total_s":6.323604707999948,"rtf":2.0801331276315618,"output_peak":0.29376375675201416,"output_rms":0.03380687261203737,"finished_at":"2026-10-13T04:53:19.027125+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"78022bc5c2d44c2a1e042327f72331142b5b1edf521d9b9a048cd3b0c87ea3f5","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":48640,"output_samples":48800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.419900057121865,"output_peak":0.29376375675201416},"common_prefix":{"samples":48640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9967213114754099,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.632837324772458,"spectral_convergence":0.25209488168495686,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.564846421240062,"spectral_convergence":0.27225390676866407,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.588492459794892,"spectral_convergence":0.31130933609827893,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":9.437294403780727e-05,"delta_db":-0.725113123037291,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.0003385111318193318,"delta_db":-0.9455585055676536,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.0005305270617278102,"delta_db":-0.9886544135666675,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.00011770947909599975,"delta_db":-1.2769943997851396,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":4.006644940674687e-05,"delta_db":-1.2412187750240014,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":2.5477124073206967e-05,"delta_db":-1.0931134947404426,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.045108785286866,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9969821479433266,"pitch":null,"output_rms_dbfs":-29.40563750113743,"output_peak":0.29376375675201416,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.595392068602471},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48640,"output_samples":48640,"intelligibility":{"estoi":{"value":0.9283362440030382,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03798107322302981,"rms_dbfs":-28.408655353194106,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.085568020654186,"mean":-1.847430279380397e-05,"zero_energy":false},"output":{"rms":0.03386243036406299,"rms_dbfs":-29.40563750113743,"peak":0.29376375675201416,"peak_dbfs":-10.64003573072101,"crest_factor_db":18.765601770416424,"mean":-4.027245980038691e-06,"zero_energy":false},"rms_difference_db":-0.9969821479433243},"full_scale":{"reference":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":304,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.2061320541595871e-07,"output_mean_power":3.6241377066017864e-07,"output_rms_dbfs":-64.40795308765354,"output_zero_energy":false,"output_peak":0.03752538189291954,"output_energy_fraction":0.00011436352759357044,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":297,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":94,"both_estimated_voiced_frames":83,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.83,"absolute_difference_cents_median":5.627816242019878,"absolute_difference_cents_p95":35.366273905455,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00012692463938083332,"log_rms_mae_db":1.119046677022921,"rms_envelope_nrmse":0.13718729581829933,"envelope_correlation":0.9935404979085392,"rms_gain_error_db":-0.9969822933652229,"reference_mean_power":0.0014449383611437946,"output_mean_power":0.0011485531356530535},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.0532453868977968e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005746543771154043,"output_power":0.0004407053975576403,"reference_active":true,"delta_db":-1.1525834946985474},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014747895972213286,"output_power":0.00010348329657669907,"reference_active":true,"delta_db":-1.5385981034199225},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536231970329892e-05,"output_power":2.5973951831886384e-05,"reference_active":true,"delta_db":0.10349058086222086}],"active_band_count":3,"active_band_mae_db":0.9315573929935635}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00010829935323771927,"log_rms_mae_db":1.259976394630332,"rms_envelope_nrmse":0.12623836797684734,"envelope_correlation":0.9958716736808,"rms_gain_error_db":-0.996984814530597,"reference_mean_power":0.0014593630016451321,"output_mean_power":0.0011600183257833526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":9.289432894914325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005498374692309952,"output_power":0.00042025083460764563,"reference_active":true,"delta_db":-1.1672574734270627},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8170998681841746e-05,"output_power":3.8961819941809146e-05,"reference_active":true,"delta_db":-1.740672848779892},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4585911877580324e-06,"output_power":1.144252610120762e-06,"reference_active":true,"delta_db":-1.0541167363941293}],"active_band_count":3,"active_band_mae_db":1.320682352867028}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00011599347403670498,"log_rms_mae_db":1.0491488425262137,"rms_envelope_nrmse":0.1375817216886004,"envelope_correlation":0.9933916097625749,"rms_gain_error_db":-0.9571549372295819,"reference_mean_power":0.0008943859828976575,"output_mean_power":0.0007174794895195353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":8.128232732706927e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005209748261573112,"output_power":0.0004027698279607812,"reference_active":true,"delta_db":-1.1175980890860198},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010680702813776804,"output_power":8.231151390573746e-05,"reference_active":true,"delta_db":-1.131392416896968},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2431239684221885e-05,"output_power":2.3389139712756547e-05,"reference_active":true,"delta_db":0.1816097215575392}],"active_band_count":3,"active_band_mae_db":0.8102000758468425}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":0.9231992552846459,"rms_envelope_nrmse":0.12429898233356892,"envelope_correlation":0.9959519483913593,"rms_gain_error_db":-0.957155056164094,"reference_mean_power":0.0009033149500834516,"output_mean_power":0.0007246423176406907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":7.323650960940425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005139241705701183,"output_power":0.0003976435580675585,"reference_active":true,"delta_db":-1.114050924818564},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.077088099913429e-05,"output_power":2.9339463479393057e-05,"reference_active":true,"delta_db":-1.4289792823529268},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2000817341803496e-06,"output_power":9.521714506520712e-07,"reference_active":true,"delta_db":-1.0049566992276722}],"active_band_count":3,"active_band_mae_db":1.1826623021330542}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":1.341694756337545,"rms_envelope_nrmse":0.1882616137615954,"envelope_correlation":0.9846253399082509,"rms_gain_error_db":-1.1106414587722184,"reference_mean_power":0.00017342262158164613,"output_mean_power":0.0001342893591064907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.6075439307790641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.767193199363684e-05,"output_power":6.839989230239168e-05,"reference_active":true,"delta_db":-1.5471436042756554},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2709798816345444e-05,"output_power":1.851986384863171e-05,"reference_active":true,"delta_db":-0.8857549728782455},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175922545645166e-06,"output_power":4.0318815019521514e-06,"reference_active":true,"delta_db":-0.1524467524608252}],"active_band_count":3,"active_band_mae_db":0.861781776538242}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":1.0860081144692841,"rms_envelope_nrmse":0.16437370293974404,"envelope_correlation":0.9903068945011242,"rms_gain_error_db":-1.110654186870557,"reference_mean_power":0.0001751536319883132,"output_mean_power":0.0001356293650473024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.4903701525512372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.99755918258773e-05,"output_power":7.077178983203568e-05,"reference_active":true,"delta_db":-1.5003380457354687},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.52463948098894e-06,"output_power":8.123308509221211e-06,"reference_active":true,"delta_db":-0.6911559879093211},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.337301598444854e-07,"output_power":2.9040431742788864e-07,"reference_active":true,"delta_db":-0.6039238775395389}],"active_band_count":3,"active_band_mae_db":0.9318059703947762}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":1.933502382692804,"rms_envelope_nrmse":0.23669337679808627,"envelope_correlation":0.9696385166237094,"rms_gain_error_db":-1.3140433859293796,"reference_mean_power":3.032942326964182e-05,"output_mean_power":2.2410926612815175e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":3.170053749158531e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1550798308210335e-05,"output_power":6.363717883291648e-06,"reference_active":true,"delta_db":-2.5890108234723934},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.128379941764708e-06,"output_power":4.152740458354712e-06,"reference_active":true,"delta_db":-1.6901089341878577},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7104244593685816e-06,"output_power":1.7105207230541695e-06,"reference_active":true,"delta_db":0.0002444165880487793}],"active_band_count":3,"active_band_mae_db":1.4264547247494332}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":1.6559030014595568,"rms_envelope_nrmse":0.20481546578109933,"envelope_correlation":0.9804526990582659,"rms_gain_error_db":-1.3141059324047397,"reference_mean_power":3.063204792415604e-05,"output_mean_power":2.263421513913442e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.7930999787597987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1345569218097283e-05,"output_power":6.090429693148308e-06,"reference_active":true,"delta_db":-2.7017835559828085},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6424993904119246e-06,"output_power":1.8958451470077579e-06,"reference_active":true,"delta_db":-1.4421203459893799},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1775222693841555e-07,"output_power":9.531411115212424e-08,"reference_active":true,"delta_db":-0.918119270549766}],"active_band_count":3,"active_band_mae_db":1.687341057507318}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8659543991088867,"predicted_mos":3.8659543991088867,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":134144,"input_sample_rate":44100,"input_seconds":3.0418140589569163,"input_float32_sha256":"f1b5eab16ac28b89bab792331f7fbc96f53ec93e80a7674eba073541f10ebe03","internal_samples":48670,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.161599417002435},"started_at":"2026-10-13T07:16:16.326265+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005184272000406054,"native_input_samples":134064,"native_input_sha256":"147450e6e2a63a1c0fceb37f40d25e0fe963a89ac5afe94d2332d0807e3fd956","output_sha256":"b3916b70a67ea9f441020edc907ae59228d7171f482685c6831e2cbbfb1b93c9","output_samples":134144,"length_delta_samples":80,"analysis_s":0.09395229000074323,"synthesis_s":19.189644181000403,"total_s":19.283596471001147,"rtf":6.3432883128293245,"output_peak":0.28882089257240295,"output_rms":0.034276995825542445,"finished_at":"2026-10-13T07:16:35.632843+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"b3916b70a67ea9f441020edc907ae59228d7171f482685c6831e2cbbfb1b93c9","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":48640,"output_samples":48670,"length_delta_samples":30,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.296955890591377,"output_peak":0.29081226078890055},"common_prefix":{"samples":48640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9993836038627492,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.990319331865141,"spectral_convergence":0.4407923902736101,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.246909518248552,"spectral_convergence":0.40440641727175886,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.863997722499249,"spectral_convergence":0.42189695847357356,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":4.584249333155266e-05,"delta_db":-3.8609058523574307,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.00041039948432612054,"delta_db":-0.1092199902806158,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.0004743592766377523,"delta_db":-1.4746558502131324,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.00017938753012609353,"delta_db":0.5528137259579471,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":4.921523447447888e-05,"delta_db":-0.34803176957188386,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":1.7242382653638575e-05,"delta_db":-2.7886447257019613,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.522378652347162,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8856227381027636,"pitch":null,"output_rms_dbfs":-29.294278091296867,"output_peak":0.29081226078890055,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.367075524204315},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48640,"output_samples":48640,"intelligibility":{"estoi":{"value":0.8364034055638567,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03798107322302981,"rms_dbfs":-28.408655353194106,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.085568020654186,"mean":-1.847430279380397e-05,"zero_energy":false},"output":{"rms":0.03429936623302908,"rms_dbfs":-29.294278091296867,"peak":0.29081226078890055,"peak_dbfs":-10.727745747132502,"crest_factor_db":18.566532344164365,"mean":-2.6410795127404436e-05,"zero_energy":false},"rms_difference_db":-0.8856227381027608},"full_scale":{"reference":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":304,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.2061320541595871e-07,"output_mean_power":1.902970491300307e-07,"output_rms_dbfs":-67.20567946113823,"output_zero_energy":false,"output_peak":0.024115695764137527,"output_energy_fraction":5.8530059359766624e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":297,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":117,"both_estimated_voiced_frames":76,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.76,"absolute_difference_cents_median":7.22583132587409,"absolute_difference_cents_p95":35.89471266933458,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00012692463938083332,"log_rms_mae_db":1.445280041968043,"rms_envelope_nrmse":0.1284673745390703,"envelope_correlation":0.9935762357150537,"rms_gain_error_db":-0.885622546923497,"reference_mean_power":0.0014449383611437946,"output_mean_power":0.001178384622432667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.0532453868977968e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005746543771154043,"output_power":0.0004649308598590639,"reference_active":true,"delta_db":-0.9201834554757377},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014747895972213286,"output_power":0.00010730694378954403,"reference_active":true,"delta_db":-1.3810223964760726},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536231970329892e-05,"output_power":2.3471826552341798e-05,"reference_active":true,"delta_db":-0.3364208547794733}],"active_band_count":3,"active_band_mae_db":0.8792089022437612}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00010829935323771927,"log_rms_mae_db":1.5080898275813353,"rms_envelope_nrmse":0.10758456458828286,"envelope_correlation":0.9978819672331579,"rms_gain_error_db":-0.88562225792278,"reference_mean_power":0.0014593630016451321,"output_mean_power":0.0011901483692087302},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":9.289432894914325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005498374692309952,"output_power":0.00044368707411612774,"reference_active":true,"delta_db":-0.931575553027765},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8170998681841746e-05,"output_power":4.231087507953482e-05,"reference_active":true,"delta_db":-1.3825451228157557},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4585911877580324e-06,"output_power":1.2225079826138847e-06,"reference_active":true,"delta_db":-0.7668188177732654}],"active_band_count":3,"active_band_mae_db":1.0269798312055953}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00011599347403670498,"log_rms_mae_db":1.9127771514069074,"rms_envelope_nrmse":0.1553606132109187,"envelope_correlation":0.9873711172007456,"rms_gain_error_db":-0.8424431755550775,"reference_mean_power":0.0008943859828976575,"output_mean_power":0.0007366830315951416},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":8.128232732706927e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005209748261573112,"output_power":0.00039584397199001945,"reference_active":true,"delta_db":-1.1929270263226117},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010680702813776804,"output_power":7.599945783233426e-05,"reference_active":true,"delta_db":-1.4778933706030823},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2431239684221885e-05,"output_power":1.980168441846722e-05,"reference_active":true,"delta_db":-0.5415114111828678}],"active_band_count":3,"active_band_mae_db":1.0707772693695208}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":1.6100058320481014,"rms_envelope_nrmse":0.13286027672560632,"envelope_correlation":0.9920575186346686,"rms_gain_error_db":-0.8424433753922085,"reference_mean_power":0.0009033149500834516,"output_mean_power":0.0007440375609826349},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":7.323650960940425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005139241705701183,"output_power":0.00038543575251523887,"reference_active":true,"delta_db":-1.2494704686550546},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.077088099913429e-05,"output_power":2.840921167360319e-05,"reference_active":true,"delta_db":-1.568909133573345},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2000817341803496e-06,"output_power":9.84410584379203e-07,"reference_active":true,"delta_db":-0.8603455103789628}],"active_band_count":3,"active_band_mae_db":1.2262417042024543}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":2.078670701620287,"rms_envelope_nrmse":0.19983281709449505,"envelope_correlation":0.9741523594152652,"rms_gain_error_db":-0.05467674444447256,"reference_mean_power":0.00017342262158164613,"output_mean_power":0.00017125295442608576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.6075439307790641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.767193199363684e-05,"output_power":0.00010319119486038425,"reference_active":true,"delta_db":0.23872862879597806},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2709798816345444e-05,"output_power":2.7206694554974407e-05,"reference_active":true,"delta_db":0.7846249402386616},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175922545645166e-06,"output_power":4.90492564582068e-06,"reference_active":true,"delta_db":0.6987999341624608}],"active_band_count":3,"active_band_mae_db":0.5740511677323669}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":1.8340539636409678,"rms_envelope_nrmse":0.1661842750675095,"envelope_correlation":0.9811569963492067,"rms_gain_error_db":-0.05467791209084817,"reference_mean_power":0.0001751536319883132,"output_mean_power":0.00017296226189102528},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.4903701525512372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.99755918258773e-05,"output_power":0.00010806845157536599,"reference_active":true,"delta_db":0.3380494515347924},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.52463948098894e-06,"output_power":1.2079503538438152e-05,"reference_active":true,"delta_db":1.032005392731127},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.337301598444854e-07,"output_power":3.454526581760465e-07,"reference_active":true,"delta_db":0.14993082359041077}],"active_band_count":3,"active_band_mae_db":0.5066618892854434}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":3.692533249230896,"rms_envelope_nrmse":0.35331486517670935,"envelope_correlation":0.9131092584228306,"rms_gain_error_db":-1.1757312485402296,"reference_mean_power":3.032942326964182e-05,"output_mean_power":2.313614658413187e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":3.170053749158531e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1550798308210335e-05,"output_power":9.63584734391248e-06,"reference_active":true,"delta_db":-0.7872208951796378},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.128379941764708e-06,"output_power":6.562021420087652e-06,"reference_active":true,"delta_db":0.2969196124244774},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7104244593685816e-06,"output_power":9.711020517885121e-07,"reference_active":true,"delta_db":-2.4583902667576423}],"active_band_count":3,"active_band_mae_db":1.1808435914539193}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":3.5786740179932925,"rms_envelope_nrmse":0.3116378788815414,"envelope_correlation":0.9307301338805313,"rms_gain_error_db":-1.175716113034508,"reference_mean_power":3.063204792415604e-05,"output_mean_power":2.3367078712667266e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.7930999787597987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1345569218097283e-05,"output_power":9.169894304670489e-06,"reference_active":true,"delta_db":-0.924619598660193},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6424993904119246e-06,"output_power":3.0643957622121952e-06,"reference_active":true,"delta_db":0.6432995728352429},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1775222693841555e-07,"output_power":8.298082681932688e-08,"reference_active":true,"delta_db":-1.5199137147026212}],"active_band_count":3,"active_band_mae_db":1.0292776287326857}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"golf","status":"measured","mos":{"status":"measured","value":3.259976387023926,"predicted_mos":3.259976387023926,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":72720,"input_sample_rate":24000,"input_seconds":3.03,"input_float32_sha256":"9293094503aeda6296bae11dd86e544085a013f3cb370f584bf4c76cc87ebe80","internal_samples":48480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1738713170016126},"started_at":"2026-10-13T05:02:37.672019+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001795563000086986,"native_input_samples":72960,"native_input_sha256":"2ec19ca970d99d1d3f16a90499abf60f4c4dfb85ed445f7f49c7083f4427d2ca","output_sha256":"0d8a0e95507ff8f8e95068ea5e33cf63f87d711cab953d054cc5a066762a50d3","output_samples":72720,"length_delta_samples":-240,"analysis_s":0.2915568260004875,"synthesis_s":0.0665460609998263,"total_s":0.3581028870003138,"rtf":0.1177970023027348,"output_peak":0.33461490273475647,"output_rms":0.03352736885993539,"finished_at":"2026-10-13T05:02:38.037017+00:00","analysis_stages":{"external_f0_s":0.03811170199969638},"evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"0d8a0e95507ff8f8e95068ea5e33cf63f87d711cab953d054cc5a066762a50d3","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":48640,"output_samples":48480,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.487640148989065,"output_peak":0.33532049147979276},"common_prefix":{"samples":48480,"sample_rate":16000,"reference_coverage":0.9967105263157895,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48480,"output_samples":48480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.01122361118646,"spectral_convergence":0.41746135533419226,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.157889397322587,"spectral_convergence":0.29515009909198847,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.222561761450967,"spectral_convergence":0.3334815826768777,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011188884212108055,"output_power":9.945893115920471e-05,"delta_db":-0.511429918338521,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004222416832991085,"output_power":0.00029481103706545907,"delta_db":-1.5601736572309675,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006683492181474532,"output_power":0.0005475118383391135,"delta_db":-0.8660993037976538,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015846826756657235,"output_power":0.0001300124830996395,"delta_db":-0.8595725715399336,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.349752133186677e-05,"output_power":3.634831684408227e-05,"delta_db":-1.6784935549330293,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.28772665889106e-05,"output_power":1.7073616613741807e-05,"delta_db":-2.845701780313958,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3869117976923437,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0932941831496197,"pitch":null,"output_rms_dbfs":-29.487640148989065,"output_peak":0.33532049147979276,"reference_rms_dbfs":-28.39434596583945,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.4638915899866705},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48480,"output_samples":48480,"intelligibility":{"estoi":{"value":0.9293669757939371,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03804369591739204,"rms_dbfs":-28.39434596583945,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.071258633299525,"mean":-1.8117294059728237e-05,"zero_energy":false},"output":{"rms":0.03354424277161792,"rms_dbfs":-29.487640148989065,"peak":0.33532049147979276,"peak_dbfs":-9.490798119334444,"crest_factor_db":19.996842029654623,"mean":0.0002557581888589527,"zero_energy":false},"rms_difference_db":-1.0932941831496166},"full_scale":{"reference":{"sample_count":48480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":303,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":109,"quiet_samples":17440,"reference_mean_power":1.215680882573948e-07,"output_mean_power":3.8396295560466466e-07,"output_rms_dbfs":-64.15710673946377,"output_zero_energy":false,"output_peak":0.005261307594450338,"output_energy_fraction":0.00012275444027692533,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":296,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":82,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.82,"absolute_difference_cents_median":6.425592128680785,"absolute_difference_cents_p95":21.23657185605663,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":605,"rms_floor":0.00012692463938083332,"log_rms_mae_db":4.198197433792268,"rms_envelope_nrmse":0.16222547655997877,"envelope_correlation":0.9888067037493495,"rms_gain_error_db":-1.0932948417938841,"reference_mean_power":0.001449714962783662,"output_mean_power":0.0011270758318041758},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":605,"frequency_resolution_hz":0.3305785123966942,"power_floor":1.0520824432516861e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005768753370763679,"output_power":0.0004834913682830767,"reference_active":true,"delta_db":-0.7669324705665528},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014572777207439612,"output_power":0.00012507616595083083,"reference_active":true,"delta_db":-0.6636776530545907},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4882879271355222e-05,"output_power":2.6356122075533564e-05,"reference_active":true,"delta_db":0.24980878010941382}],"active_band_count":3,"active_band_mae_db":0.5601396345768525}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":599,"rms_floor":0.00010829935323771927,"log_rms_mae_db":4.093472999449477,"rms_envelope_nrmse":0.13727436552935077,"envelope_correlation":0.9952653042220574,"rms_gain_error_db":-1.093306279925927,"reference_mean_power":0.0014642356119091615,"output_mean_power":0.0011383618615859725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":599,"frequency_resolution_hz":0.333889816360601,"power_floor":9.276603700506952e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005492922093564653,"output_power":0.0004630705689917996,"reference_active":true,"delta_db":-0.7415625967929969},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.309519880282249e-05,"output_power":6.714679923180763e-05,"reference_active":true,"delta_db":-0.3686353616925083},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5165040333372914e-06,"output_power":1.8482788582315006e-06,"reference_active":true,"delta_db":0.8592392609964205}],"active_band_count":3,"active_band_mae_db":0.6564790731606419}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":605,"rms_floor":0.00011599347403670498,"log_rms_mae_db":1.311967742359918,"rms_envelope_nrmse":0.17202020834368825,"envelope_correlation":0.9857545287524043,"rms_gain_error_db":-1.01106785701361,"reference_mean_power":0.0008973426306927495,"output_mean_power":0.0007109703915775149},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":605,"frequency_resolution_hz":0.3305785123966942,"power_floor":8.12056759587837e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005224294572631454,"output_power":0.0004204301020762229,"reference_active":true,"delta_db":-0.9433385350525564},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010564724002149736,"output_power":0.00010120022822636757,"reference_active":true,"delta_db":-0.18676663897788157},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.206996164008084e-05,"output_power":2.05048805621804e-05,"reference_active":true,"delta_db":-0.3194433438135657}],"active_band_count":3,"active_band_mae_db":0.4831828392813346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":599,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":1.1271040713093121,"rms_envelope_nrmse":0.14209919577037503,"envelope_correlation":0.9929156328665891,"rms_gain_error_db":-1.0110678320773705,"reference_mean_power":0.000906331026626923,"output_mean_power":0.0007180919601492623},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":599,"frequency_resolution_hz":0.333889816360601,"power_floor":7.315807228711966e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005059343079332632,"output_power":0.00040787256285843774,"reference_active":true,"delta_db":-0.9356963858456353},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.069981075486612e-05,"output_power":5.369662700288333e-05,"reference_active":true,"delta_db":0.24940667733801436},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2208944153045352e-06,"output_power":1.4865051629026214e-06,"reference_active":true,"delta_db":0.8548831476644059}],"active_band_count":3,"active_band_mae_db":0.6799954036160184}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":605,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":1.5901174183436413,"rms_envelope_nrmse":0.18154817850340318,"envelope_correlation":0.9859053027065112,"rms_gain_error_db":-1.066369266507172,"reference_mean_power":0.0001739959184578738,"output_mean_power":0.00013611379232081208},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":605,"frequency_resolution_hz":0.3305785123966942,"power_floor":1.6179831393589637e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.87533072639564e-05,"output_power":7.744397496826665e-05,"reference_active":true,"delta_db":-1.055640132203387},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.225658006058144e-05,"output_power":1.7916172213136163e-05,"reference_active":true,"delta_db":-0.9421320326412583},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.158652604323761e-06,"output_power":4.008234547877444e-06,"reference_active":true,"delta_db":-0.15999515817804108}],"active_band_count":3,"active_band_mae_db":0.7192557743408955}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":599,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":1.4509246205602682,"rms_envelope_nrmse":0.13698954951714767,"envelope_correlation":0.9959720913368337,"rms_gain_error_db":-1.066364430021982,"reference_mean_power":0.0001757384512176902,"output_mean_power":0.00013747709660454587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":599,"frequency_resolution_hz":0.333889816360601,"power_floor":1.500275823762953e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010232525774287248,"output_power":8.105230331331749e-05,"reference_active":true,"delta_db":-1.0121748627097251},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4531295519825801e-05,"output_power":1.1187850626847229e-05,"reference_active":true,"delta_db":-1.1355767570473736},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5302211087686444e-07,"output_power":3.4263102595324194e-07,"reference_active":true,"delta_db":-0.12975220778595675}],"active_band_count":3,"active_band_mae_db":0.7591679425143517}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":605,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":2.4659729482664976,"rms_envelope_nrmse":0.2831774956718406,"envelope_correlation":0.974386750281681,"rms_gain_error_db":-2.2568922170051455,"reference_mean_power":3.0429685435385824e-05,"output_mean_power":1.809706895551866e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":605,"frequency_resolution_hz":0.3305785123966942,"power_floor":3.155517178855081e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1577287877265916e-05,"output_power":7.291014320948958e-06,"reference_active":true,"delta_db":-2.0081888118407654},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.018853471379837e-06,"output_power":4.395453652503329e-06,"reference_active":true,"delta_db":-1.365100655484308},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7015896950035405e-06,"output_power":1.3562254418965845e-06,"reference_active":true,"delta_db":-0.9852295957602824}],"active_band_count":3,"active_band_mae_db":1.4528396876951186}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48480,"uncovered_tail_samples":0,"metrics":{"frame_count":599,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":2.3683850502796413,"rms_envelope_nrmse":0.25680282935102927,"envelope_correlation":0.9864651276268929,"rms_gain_error_db":-2.256880368431815,"reference_mean_power":3.0734324809447406e-05,"output_mean_power":1.827829321497455e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":599,"frequency_resolution_hz":0.333889816360601,"power_floor":2.7788113605551482e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1707668618361472e-05,"output_power":7.7702925306177e-06,"reference_active":true,"delta_db":-1.7803305224786825},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5665241564059825e-06,"output_power":2.6616743073667965e-06,"reference_active":true,"delta_db":-1.2709025742599567},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.227197548826174e-07,"output_power":1.0592208912670275e-07,"reference_active":true,"delta_db":-0.6392794128273123}],"active_band_count":3,"active_band_mae_db":1.2301708365219837}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.8464882373809814,"predicted_mos":3.8464882373809814,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":73080,"input_sample_rate":24000,"input_seconds":3.045,"input_float32_sha256":"03973ee448b032e8831d6aa0b8ccf73ed19a0c0a38d879d1c22a44858920f10e","internal_samples":48720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0812854890027666},"started_at":"2026-10-13T07:38:28.919297+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002782568000839092,"native_input_samples":72960,"native_input_sha256":"2ec19ca970d99d1d3f16a90499abf60f4c4dfb85ed445f7f49c7083f4427d2ca","output_sha256":"7453ee6f9c88c613d193cb4de971f31a68f619d10442c558104f8fa56a48f352","output_samples":73080,"length_delta_samples":120,"analysis_s":1.2284985939986655,"synthesis_s":7.775524606000545,"total_s":9.00402319999921,"rtf":2.9618497368418453,"output_peak":0.3781753480434418,"output_rms":0.036767954815452655,"finished_at":"2026-10-13T07:38:37.936667+00:00","analysis_stages":{"external_f0_s":0.7605050189995382},"evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"7453ee6f9c88c613d193cb4de971f31a68f619d10442c558104f8fa56a48f352","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":48640,"output_samples":48720,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.688131795102922,"output_peak":0.3787218139453411},"common_prefix":{"samples":48640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983579638752053,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.920622140689357,"spectral_convergence":0.4788638097389384,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.018210736714632,"spectral_convergence":0.308273471668893,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.728237705249358,"spectral_convergence":0.2934477374952613,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":0.00011351663870263958,"delta_db":0.07700707193919241,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.00038808488078091843,"delta_db":-0.3520208184175571,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.0006378919834550097,"delta_db":-0.1882583884647029,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.00013949647758124126,"delta_db":-0.5394763638719173,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":5.232280625555873e-05,"delta_db":-0.08211706821675577,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":2.3565639897094502e-05,"delta_db":-1.4318251455731146,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.44511747608054003,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.27233936399318864,"pitch":null,"output_rms_dbfs":-28.680994717187293,"output_peak":0.3787218139453411,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.222356860884449},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48640,"output_samples":48640,"intelligibility":{"estoi":{"value":0.9221315539779358,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03798107322302981,"rms_dbfs":-28.408655353194106,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.085568020654186,"mean":-1.847430279380397e-05,"zero_energy":false},"output":{"rms":0.03680868175404647,"rms_dbfs":-28.680994717187293,"peak":0.3787218139453411,"peak_dbfs":-8.433593585819626,"crest_factor_db":20.247401131367667,"mean":-8.036033604213452e-05,"zero_energy":false},"rms_difference_db":-0.27233936399318637},"full_scale":{"reference":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":304,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.2061320541595871e-07,"output_mean_power":1.5552090920182217e-06,"output_rms_dbfs":-58.082112134505934,"output_zero_energy":false,"output_peak":0.04783012567184651,"output_energy_fraction":0.0004153434440912638,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":297,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":97,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.85,"absolute_difference_cents_median":6.954941698989354,"absolute_difference_cents_p95":35.8020135389116,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00012692463938083332,"log_rms_mae_db":3.1494006804727284,"rms_envelope_nrmse":0.17954786026928327,"envelope_correlation":0.9738508901196039,"rms_gain_error_db":-0.2723392181983293,"reference_mean_power":0.0014449383611437946,"output_mean_power":0.001357111089581691},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.0532453868977968e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005746543771154043,"output_power":0.0005511732493912519,"reference_active":true,"delta_db":-0.18118587662779928},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014747895972213286,"output_power":0.00014193675677877837,"reference_active":true,"delta_db":-0.16635188065259526},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536231970329892e-05,"output_power":2.8689504369185278e-05,"reference_active":true,"delta_db":0.5353407285461849}],"active_band_count":3,"active_band_mae_db":0.2942928286088598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00010829935323771927,"log_rms_mae_db":2.5562590065459494,"rms_envelope_nrmse":0.11072582021982348,"envelope_correlation":0.9897851987164579,"rms_gain_error_db":-0.27233973087805186,"reference_mean_power":0.0014593630016451321,"output_mean_power":0.0013706587995160006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":9.289432894914325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005498374692309952,"output_power":0.0005318962836757308,"reference_active":true,"delta_db":-0.1440737600878303},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8170998681841746e-05,"output_power":5.465040905478362e-05,"reference_active":true,"delta_db":-0.2711310288193925},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4585911877580324e-06,"output_power":1.5755332412097817e-06,"reference_active":true,"delta_db":0.3349398512348856}],"active_band_count":3,"active_band_mae_db":0.25004821338070277}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00011599347403670498,"log_rms_mae_db":1.8493932754579876,"rms_envelope_nrmse":0.20338108129453025,"envelope_correlation":0.969410767507188,"rms_gain_error_db":-0.26610710565788076,"reference_mean_power":0.0008943859828976575,"output_mean_power":0.0008412290731427784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":8.128232732706927e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005209748261573112,"output_power":0.00048748624141151033,"reference_active":true,"delta_db":-0.2885437553558291},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010680702813776804,"output_power":0.00010546153533645808,"reference_active":true,"delta_db":-0.055057415530752225},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2431239684221885e-05,"output_power":2.6694411077042122e-05,"reference_active":true,"delta_db":0.7556706809135787}],"active_band_count":3,"active_band_mae_db":0.36642395060005334}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":1.3769623779039852,"rms_envelope_nrmse":0.12406622551339687,"envelope_correlation":0.9881801223545766,"rms_gain_error_db":-0.2661071279542442,"reference_mean_power":0.0009033149500834516,"output_mean_power":0.0008496273519980079},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":7.323650960940425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005139241705701183,"output_power":0.00048527469222058115,"reference_active":true,"delta_db":-0.24911400821982072},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.077088099913429e-05,"output_power":3.966906995995247e-05,"reference_active":true,"delta_db":-0.11898077448811181},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2000817341803496e-06,"output_power":1.6171168378818762e-06,"reference_active":true,"delta_db":1.2953057350946313}],"active_band_count":3,"active_band_mae_db":0.5544668392675213}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":1.8088298297113534,"rms_envelope_nrmse":0.20206481695041276,"envelope_correlation":0.9736078544030567,"rms_gain_error_db":-0.32247083981335456,"reference_mean_power":0.00017342262158164613,"output_mean_power":0.0001610121551430915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.6075439307790641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.767193199363684e-05,"output_power":8.798563425812447e-05,"reference_active":true,"delta_db":-0.45358009276633393},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2709798816345444e-05,"output_power":2.1515254816797626e-05,"reference_active":true,"delta_db":-0.23466792835438147},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175922545645166e-06,"output_power":3.511819427903124e-06,"reference_active":true,"delta_db":-0.7522025775752919}],"active_band_count":3,"active_band_mae_db":0.4801501995653357}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":1.1843474157045144,"rms_envelope_nrmse":0.1295890688761475,"envelope_correlation":0.9891182152639032,"rms_gain_error_db":-0.32246955870115923,"reference_mean_power":0.0001751536319883132,"output_mean_power":0.00016261933898562922},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.4903701525512372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.99755918258773e-05,"output_power":9.039862560619845e-05,"reference_active":true,"delta_db":-0.437321560663069},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.52463948098894e-06,"output_power":8.974338846438996e-06,"reference_active":true,"delta_db":-0.25846082816505334},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.337301598444854e-07,"output_power":2.527135085907246e-07,"reference_active":true,"delta_db":-1.207669990752263}],"active_band_count":3,"active_band_mae_db":0.6344841265267951}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":1.963495540856637,"rms_envelope_nrmse":0.2543175621578028,"envelope_correlation":0.9523872311752869,"rms_gain_error_db":-0.46889665557021903,"reference_mean_power":3.032942326964182e-05,"output_mean_power":2.7225413621548715e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":3.170053749158531e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1550798308210335e-05,"output_power":1.0036620178809234e-05,"reference_active":true,"delta_db":-0.6102451135832359},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.128379941764708e-06,"output_power":5.361886750885055e-06,"reference_active":true,"delta_db":-0.5802804550338427},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7104244593685816e-06,"output_power":1.549992788145131e-06,"reference_active":true,"delta_db":-0.42774220920012646}],"active_band_count":3,"active_band_mae_db":0.539422592605735}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":1.1778810640768393,"rms_envelope_nrmse":0.13670991586936537,"envelope_correlation":0.9866768499751591,"rms_gain_error_db":-0.4688991588077587,"reference_mean_power":3.063204792415604e-05,"output_mean_power":2.7497050857265262e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.7930999787597987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1345569218097283e-05,"output_power":9.7922126361775e-06,"reference_active":true,"delta_db":-0.6394545420628274},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6424993904119246e-06,"output_power":2.2913433261089685e-06,"reference_active":true,"delta_db":-0.6192472870673436},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1775222693841555e-07,"output_power":1.0239388243247536e-07,"reference_active":true,"delta_db":-0.6069511898632592}],"active_band_count":3,"active_band_mae_db":0.6218843396644768}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.6779184341430664,"predicted_mos":3.6779184341430664,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":73200,"input_sample_rate":24000,"input_seconds":3.05,"input_float32_sha256":"eeb9ae813c7e927505118a3a0c29793e838a69772258274e075a2f8b1c44e2a6","internal_samples":48800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1992214959991543},"started_at":"2026-10-13T05:10:48.273708+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0022625149995292304,"native_input_samples":72960,"native_input_sha256":"2ec19ca970d99d1d3f16a90499abf60f4c4dfb85ed445f7f49c7083f4427d2ca","output_sha256":"ff9b778efc50ef0a1a5df3ee680342bd8bea7201a4a4c8a1a77b41d03ba5629c","output_samples":73200,"length_delta_samples":240,"analysis_s":0.36940296499960823,"synthesis_s":0.09518780400139804,"total_s":0.46459076900100627,"rtf":0.15282591085559416,"output_peak":0.32070621848106384,"output_rms":0.03680274521953962,"finished_at":"2026-10-13T05:10:48.747196+00:00","analysis_stages":{"external_f0_s":0.05510422899897094},"evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"ff9b778efc50ef0a1a5df3ee680342bd8bea7201a4a4c8a1a77b41d03ba5629c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":48640,"output_samples":48800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.680267919947283,"output_peak":0.3208268528766427},"common_prefix":{"samples":48640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9967213114754099,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.711917390055626,"spectral_convergence":0.1848142667700776,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.112247661601241,"spectral_convergence":0.1996248527102069,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.454719305385045,"spectral_convergence":0.21425710881577656,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":0.00010902270590252128,"delta_db":-0.09841858868656057,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.000418141891920278,"delta_db":-0.028051255217182405,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.0006113984766538856,"delta_db":-0.3724863083250097,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.00014399558787376264,"delta_db":-0.4016169257616959,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":4.9860751138882586e-05,"delta_db":-0.2914391875110321,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":2.714396894948536e-05,"delta_db":-0.8178840172692454,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.33498271379512107,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.25735003771879833,"pitch":null,"output_rms_dbfs":-28.666005390912904,"output_peak":0.3208268528766427,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.4262947856806365},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48640,"output_samples":48640,"intelligibility":{"estoi":{"value":0.9775863948547349,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03798107322302981,"rms_dbfs":-28.408655353194106,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.085568020654186,"mean":-1.847430279380397e-05,"zero_energy":false},"output":{"rms":0.03687225770387257,"rms_dbfs":-28.666005390912904,"peak":0.3208268528766427,"peak_dbfs":-9.874585777328274,"crest_factor_db":18.79141961358463,"mean":-2.7428696426265187e-06,"zero_energy":false},"rms_difference_db":-0.2573500377187976},"full_scale":{"reference":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":304,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.2061320541595871e-07,"output_mean_power":4.720910862442994e-07,"output_rms_dbfs":-63.259741995930206,"output_zero_energy":false,"output_peak":0.02505877861456868,"output_energy_fraction":0.0001256450666498037,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":297,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":98,"both_estimated_voiced_frames":90,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9,"absolute_difference_cents_median":1.4313454379076558,"absolute_difference_cents_p95":6.438156100425358,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00012692463938083332,"log_rms_mae_db":0.715329093055415,"rms_envelope_nrmse":0.06709845944965971,"envelope_correlation":0.9969376383558138,"rms_gain_error_db":-0.257349992237434,"reference_mean_power":0.0014449383611437946,"output_mean_power":0.0013618031108438404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.0532453868977968e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005746543771154043,"output_power":0.0005290122761656547,"reference_active":true,"delta_db":-0.3594096873869491},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014747895972213286,"output_power":0.00012214636131121907,"reference_active":true,"delta_db":-0.818495314935995},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536231970329892e-05,"output_power":2.2508162478950096e-05,"reference_active":true,"delta_db":-0.5184893120099859}],"active_band_count":3,"active_band_mae_db":0.5654647714443101}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00010829935323771927,"log_rms_mae_db":0.7589959214604968,"rms_envelope_nrmse":0.04668591734280657,"envelope_correlation":0.9987937406671014,"rms_gain_error_db":-0.2573497131546975,"reference_mean_power":0.0014593630016451321,"output_mean_power":0.0013753979108340592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":9.289432894914325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005498374692309952,"output_power":0.0005014157990007871,"reference_active":true,"delta_db":-0.40036318025381123},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8170998681841746e-05,"output_power":4.9653377193219144e-05,"reference_active":true,"delta_db":-0.6875772730664421},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4585911877580324e-06,"output_power":1.3799830833027063e-06,"reference_active":true,"delta_db":-0.24059822816176604}],"active_band_count":3,"active_band_mae_db":0.4428462271606732}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00011599347403670498,"log_rms_mae_db":0.620272833560729,"rms_envelope_nrmse":0.08103244855216174,"envelope_correlation":0.9957821645182856,"rms_gain_error_db":-0.2743920796456152,"reference_mean_power":0.0008943859828976575,"output_mean_power":0.0008396258021738715},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":8.128232732706927e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005209748261573112,"output_power":0.0004718789896810018,"reference_active":true,"delta_db":-0.429860975172618},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010680702813776804,"output_power":8.88090876823919e-05,"reference_active":true,"delta_db":-0.8014242250081336},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2431239684221885e-05,"output_power":2.0014704968376957e-05,"reference_active":true,"delta_db":-0.49504083287142064}],"active_band_count":3,"active_band_mae_db":0.5754420110173908}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":0.36295923279591646,"rms_envelope_nrmse":0.056633149319341866,"envelope_correlation":0.9981952420684899,"rms_gain_error_db":-0.2743920682363154,"reference_mean_power":0.0009033149500834516,"output_mean_power":0.0008480080816055925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":7.323650960940425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005139241705701183,"output_power":0.00046125515986107917,"reference_active":true,"delta_db":-0.4695780618239034},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.077088099913429e-05,"output_power":3.520095046177317e-05,"reference_active":true,"delta_db":-0.637957059894209},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2000817341803496e-06,"output_power":1.1433281752120134e-06,"reference_active":true,"delta_db":-0.21039919617139843}],"active_band_count":3,"active_band_mae_db":0.4393114392965036}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":0.7591737248083406,"rms_envelope_nrmse":0.0916265999538599,"envelope_correlation":0.9954299544677684,"rms_gain_error_db":-0.35035668108855167,"reference_mean_power":0.00017342262158164613,"output_mean_power":0.00015998161584688525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.6075439307790641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.767193199363684e-05,"output_power":8.93712247551557e-05,"reference_active":true,"delta_db":-0.3857206873414585},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2709798816345444e-05,"output_power":2.0128692324286022e-05,"reference_active":true,"delta_db":-0.5239772533311946},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175922545645166e-06,"output_power":4.272840950044307e-06,"reference_active":true,"delta_db":0.0996429232844976}],"active_band_count":3,"active_band_mae_db":0.33644695465238356}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":0.5043272529705045,"rms_envelope_nrmse":0.061353677420317825,"envelope_correlation":0.9984483456680742,"rms_gain_error_db":-0.3503579407277741,"reference_mean_power":0.0001751536319883132,"output_mean_power":0.00016157841857072927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.4903701525512372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.99755918258773e-05,"output_power":9.117054058749272e-05,"reference_active":true,"delta_db":-0.4003945379062749},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.52463948098894e-06,"output_power":8.508047812453946e-06,"reference_active":true,"delta_db":-0.4901862426160848},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.337301598444854e-07,"output_power":3.433139828171155e-07,"reference_active":true,"delta_db":0.12296035676563108}],"active_band_count":3,"active_band_mae_db":0.3378470457626636}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":0.9883630369874612,"rms_envelope_nrmse":0.12180618584496385,"envelope_correlation":0.9906397971399715,"rms_gain_error_db":-0.45611688870706557,"reference_mean_power":3.032942326964182e-05,"output_mean_power":2.730564647748259e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":3.170053749158531e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1550798308210335e-05,"output_power":1.0191070636159392e-05,"reference_active":true,"delta_db":-0.5439218880654488},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.128379941764708e-06,"output_power":5.479253102213991e-06,"reference_active":true,"delta_db":-0.48624320266480037},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7104244593685816e-06,"output_power":1.2683292818623588e-06,"reference_active":true,"delta_db":-1.298718792897979}],"active_band_count":3,"active_band_mae_db":0.7762946278760762}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":0.6818666594697271,"rms_envelope_nrmse":0.07267945452671834,"envelope_correlation":0.9977784289602905,"rms_gain_error_db":-0.45610896157886804,"reference_mean_power":3.063204792415604e-05,"output_mean_power":2.757815045758834e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.7930999787597987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1345569218097283e-05,"output_power":9.927614896227897e-06,"reference_active":true,"delta_db":-0.5798136771112219},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6424993904119246e-06,"output_power":2.503711897096863e-06,"reference_active":true,"delta_db":-0.23430542780774663},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1775222693841555e-07,"output_power":7.153951260752585e-08,"reference_active":true,"delta_db":-2.1642315227393762}],"active_band_count":3,"active_band_mae_db":0.9927835425527816}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4970-29095-0008","method":"magphase","status":"measured","mos":{"status":"measured","value":3.5395030975341797,"predicted_mos":3.5395030975341797,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":146402,"input_sample_rate":48000,"input_seconds":3.0500416666666665,"input_float32_sha256":"30929a3fed03eb37f26c7c261c4ee5be36279679312f59ed1d9433e37114c634","internal_samples":48801,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1516142019991094},"started_at":"2026-10-13T05:15:54.287726+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005523240999536938,"native_input_samples":145920,"native_input_sha256":"d9e1007dd90c34928490bc2814f28fe8929430a0686434fa3eb8cfac141749ee","output_sha256":"356883c6c4e5242674b40c661c46a992408ea3147a6f9bd4b545d968b0400408","output_samples":146402,"length_delta_samples":482,"analysis_s":1.8892258819996641,"synthesis_s":0.7037542870002653,"total_s":2.5929801689999294,"rtf":0.852954002960503,"output_peak":0.3693251080616186,"output_rms":0.037173074022015196,"finished_at":"2026-10-13T05:15:56.892804+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"output_sha256":"356883c6c4e5242674b40c661c46a992408ea3147a6f9bd4b545d968b0400408","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":48640,"output_samples":48801,"length_delta_samples":161,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.593996102536092,"output_peak":0.36984679265224535},"common_prefix":{"samples":48640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9967008872769001,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":48640,"output_samples":48640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.367867222301168,"spectral_convergence":0.6791654303402994,"spectral_convergence_status":"ok","amplitude_floor":9.047519636617716e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.1770577012653005,"spectral_convergence":0.40479377978761905,"spectral_convergence_status":"ok","amplitude_floor":7.308511681145121e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.44449255098427,"spectral_convergence":0.30919396961888607,"spectral_convergence_status":"ok","amplitude_floor":4.4367419091404295e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00011152155483091155,"output_power":0.0001048669178932699,"delta_db":-0.26720311724242013,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042085142829602937,"output_power":0.0003904422879786991,"delta_db":-0.32571956374605426,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006661514585812611,"output_power":0.0006329708448449286,"delta_db":-0.22189276627144805,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00015794681682280444,"output_power":0.00016615844043112833,"delta_db":0.22011529456678477,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.332154632901203e-05,"output_power":5.647476133567117e-05,"delta_db":0.24951668754141068,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.27688915881969e-05,"output_power":3.5956037052049714e-05,"delta_db":0.40310067722048265,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2812580177647668,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.17098932743111972,"pitch":null,"output_rms_dbfs":-28.579644680625226,"output_peak":0.36984679265224535,"reference_rms_dbfs":-28.408655353194106,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.329805824850247},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":48640,"output_samples":48640,"intelligibility":{"estoi":{"value":0.8567939115906773,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03798107322302981,"rms_dbfs":-28.408655353194106,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":19.085568020654186,"mean":-1.847430279380397e-05,"zero_energy":false},"output":{"rms":0.03724069402375223,"rms_dbfs":-28.579644680625226,"peak":0.36984679265224535,"peak_dbfs":-8.63956286380392,"crest_factor_db":19.9400818168213,"mean":0.0005696864089793348,"zero_energy":false},"rms_difference_db":-0.17098932743111916},"full_scale":{"reference":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":48640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":304,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.6109864081954585e-06,"threshold_dbfs":-57.92908123684204,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":110,"quiet_samples":17600,"reference_mean_power":1.2061320541595871e-07,"output_mean_power":1.588205998597793e-07,"output_rms_dbfs":-67.99093167997037,"output_zero_energy":false,"output_peak":0.00685084307940207,"output_energy_fraction":4.1437200008676176e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":297,"reference_estimated_voiced_frames":100,"output_estimated_voiced_frames":96,"both_estimated_voiced_frames":86,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.86,"absolute_difference_cents_median":10.271422985339191,"absolute_difference_cents_p95":34.32468100461764,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":48640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00012692463938083332,"log_rms_mae_db":2.2232424245327187,"rms_envelope_nrmse":0.2521978693693357,"envelope_correlation":0.9476466867798696,"rms_gain_error_db":-0.17098915676265702,"reference_mean_power":0.0014449383611437946,"output_mean_power":0.0013891540373784997},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.0532453868977968e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005746543771154043,"output_power":0.0005613699131182705,"reference_active":true,"delta_db":-0.10157586344741965},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014747895972213286,"output_power":0.00014419398032256353,"reference_active":true,"delta_db":-0.09782935309657086},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536231970329892e-05,"output_power":2.8254528687562842e-05,"reference_active":true,"delta_db":0.4689909452701982}],"active_band_count":3,"active_band_mae_db":0.22279872060472958}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":0.00010829935323771927,"log_rms_mae_db":1.6292429957819747,"rms_envelope_nrmse":0.14897913838038412,"envelope_correlation":0.9804334354549854,"rms_gain_error_db":-0.1709896680569105,"reference_mean_power":0.0014593630016451321,"output_mean_power":0.00140302162471485},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":9.289432894914325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005498374692309952,"output_power":0.0005393868248923289,"reference_active":true,"delta_db":-0.08333997819938344},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8170998681841746e-05,"output_power":5.682892724229523e-05,"reference_active":true,"delta_db":-0.1013706188587456},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4585911877580324e-06,"output_power":1.6872134147544822e-06,"reference_active":true,"delta_db":0.6323643435250124}],"active_band_count":3,"active_band_mae_db":0.27235831352771384}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":0.00011599347403670498,"log_rms_mae_db":2.2608378366538497,"rms_envelope_nrmse":0.2724851576159465,"envelope_correlation":0.9444817318488034,"rms_gain_error_db":-0.2281917797293505,"reference_mean_power":0.0008943859828976575,"output_mean_power":0.0008486054297010216},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":8.128232732706927e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005209748261573112,"output_power":0.0004936168841198534,"reference_active":true,"delta_db":-0.23426732132663625},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010680702813776804,"output_power":0.00010073910971096953,"reference_active":true,"delta_db":-0.2540172273620442},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2431239684221885e-05,"output_power":2.1573014486808686e-05,"reference_active":true,"delta_db":-0.16942440839606143}],"active_band_count":3,"active_band_mae_db":0.21923631902824728}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":9.903058317661472e-05,"log_rms_mae_db":1.6124132160075095,"rms_envelope_nrmse":0.1644816902642883,"envelope_correlation":0.9785165497796434,"rms_gain_error_db":-0.2281917753289339,"reference_mean_power":0.0009033149500834516,"output_mean_power":0.0008570773545266327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":7.323650960940425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005139241705701183,"output_power":0.0004870166679926041,"reference_active":true,"delta_db":-0.23355218579045095},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.077088099913429e-05,"output_power":3.8915537783288587e-05,"reference_active":true,"delta_db":-0.20227059574986206},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2000817341803496e-06,"output_power":1.247146677630456e-06,"reference_active":true,"delta_db":0.1670670847703392}],"active_band_count":3,"active_band_mae_db":0.2009632887702174}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.992748424465533e-05,"log_rms_mae_db":2.384332029903682,"rms_envelope_nrmse":0.3170626439216443,"envelope_correlation":0.9374205472596376,"rms_gain_error_db":0.2058037155212492,"reference_mean_power":0.00017342262158164613,"output_mean_power":0.00018183861656394042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.6075439307790641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.767193199363684e-05,"output_power":0.000102572436492439,"reference_active":true,"delta_db":0.21260893474943682},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2709798816345444e-05,"output_power":2.4028415120074296e-05,"reference_active":true,"delta_db":0.245118394362524},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.175922545645166e-06,"output_power":4.819578787191493e-06,"reference_active":true,"delta_db":0.6225664937249171}],"active_band_count":3,"active_band_mae_db":0.36009794094562597}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":6.970659359658015e-05,"log_rms_mae_db":1.54362751064967,"rms_envelope_nrmse":0.19655537076743101,"envelope_correlation":0.9748890628480075,"rms_gain_error_db":0.2058029513047988,"reference_mean_power":0.0001751536319883132,"output_mean_power":0.0001836535985439727},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":1.4903701525512372e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.99755918258773e-05,"output_power":0.00010511399673227643,"reference_active":true,"delta_db":0.21766565801516782},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.52463948098894e-06,"output_power":1.0409548417261427e-05,"reference_active":true,"delta_db":0.385833434424837},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.337301598444854e-07,"output_power":3.492789806734861e-07,"reference_active":true,"delta_db":0.19776994466935857}],"active_band_count":3,"active_band_mae_db":0.2670896790364545}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":2.3755576133519238e-05,"log_rms_mae_db":2.459303811015341,"rms_envelope_nrmse":0.4248296288039062,"envelope_correlation":0.8818460654289062,"rms_gain_error_db":0.614353666585704,"reference_mean_power":3.032942326964182e-05,"output_mean_power":3.493811891898733e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":3.170053749158531e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1550798308210335e-05,"output_power":1.3089137941103094e-05,"reference_active":true,"delta_db":0.5429904400306369},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.128379941764708e-06,"output_power":6.459102257802844e-06,"reference_active":true,"delta_db":0.2282647770972745},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7104244593685816e-06,"output_power":2.1937265370889676e-06,"reference_active":true,"delta_db":1.080785904600805}],"active_band_count":3,"active_band_mae_db":0.6173470405762388}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":48640,"uncovered_tail_samples":0,"metrics":{"frame_count":601,"rms_floor":2.1008424841575618e-05,"log_rms_mae_db":1.6181881364385535,"rms_envelope_nrmse":0.2543210549437421,"envelope_correlation":0.9563635770530082,"rms_gain_error_db":0.6143523777990677,"reference_mean_power":3.063204792415604e-05,"output_mean_power":3.528671831302472e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":601,"frequency_resolution_hz":0.3327787021630616,"power_floor":2.7930999787597987e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1345569218097283e-05,"output_power":1.3220326110106183e-05,"reference_active":true,"delta_db":0.6641587842716067},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6424993904119246e-06,"output_power":2.795188238539374e-06,"reference_active":true,"delta_db":0.24396164445214719},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1775222693841555e-07,"output_power":1.6682341873715516e-07,"reference_active":true,"delta_db":1.5128788772514512}],"active_band_count":3,"active_band_mae_db":0.8069997686584017}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"hnm","status":"measured","mos":{"status":"measured","value":2.5217103958129883,"predicted_mos":2.5217103958129883,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":71920,"input_sample_rate":16000,"input_seconds":4.495,"input_float32_sha256":"2005be3837f554acfdab07e75acdf5bb325dd6a37ea79d0b6f8350dc8bb5310f","internal_samples":71920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.490143143999376},"started_at":"2026-10-13T04:39:23.871135+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.23290001283749e-05,"native_input_samples":71920,"native_input_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","output_sha256":"e06ef0c7e35a463d0930795b7ca3d4fccbadc557d34d31a470a0a8fc1d087d9b","output_samples":71920,"length_delta_samples":0,"analysis_s":0.8843164250001792,"synthesis_s":0.37856749200000195,"total_s":1.262883917000181,"rtf":0.280953040489473,"output_peak":0.32807196276476047,"output_rms":0.03790487367206148,"finished_at":"2026-10-13T04:39:25.138587+00:00","evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"e06ef0c7e35a463d0930795b7ca3d4fccbadc557d34d31a470a0a8fc1d087d9b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.683078178234252,"spectral_convergence":0.2784395992477099,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.604328021201616,"spectral_convergence":0.22882135892808425,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.938659416056214,"spectral_convergence":0.2343022193287057,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.00021241150829496563,"delta_db":0.10341335085260964,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0006517968684534214,"delta_db":-0.013448694533155019,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.0004683624745699021,"delta_db":-0.18499000057122983,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":3.036132387887783e-05,"delta_db":-0.4374989574597703,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":5.730345521329411e-05,"delta_db":-0.27155855373622645,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":1.6543817684478267e-05,"delta_db":-0.39736176986448746,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2347118878362464,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07760540009767622,"pitch":null,"output_rms_dbfs":-28.426098928446415,"output_peak":0.32807196276476047,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.742021871830694},"common_prefix":{"samples":71920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.683078178234252,"spectral_convergence":0.2784395992477099,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.604328021201616,"spectral_convergence":0.22882135892808425,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.938659416056214,"spectral_convergence":0.2343022193287057,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.00021241150829496563,"delta_db":0.10341335085260964,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0006517968684534214,"delta_db":-0.013448694533155019,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.0004683624745699021,"delta_db":-0.18499000057122983,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":3.036132387887783e-05,"delta_db":-0.4374989574597703,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":5.730345521329411e-05,"delta_db":-0.27155855373622645,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":1.6543817684478267e-05,"delta_db":-0.39736176986448746,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2347118878362464,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07760540009767622,"pitch":null,"output_rms_dbfs":-28.426098928446415,"output_peak":0.32807196276476047,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.742021871830694},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71920,"output_samples":71920,"intelligibility":{"estoi":{"value":0.889459514252344,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.038245057973635734,"rms_dbfs":-28.34849352834874,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.396015589762875,"mean":-1.857829703902774e-05,"zero_energy":false},"output":{"rms":0.03790487367206148,"rms_dbfs":-28.426098928446415,"peak":0.32807196276476047,"peak_dbfs":-9.680617662126021,"crest_factor_db":18.745481266320393,"mean":-0.00023120293364680412,"zero_energy":false},"rms_difference_db":-0.07760540009767425},"full_scale":{"reference":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":2.8615459731662007e-07,"output_rms_dbfs":-65.43399272389871,"output_zero_energy":false,"output_peak":0.0032004945531912726,"output_energy_fraction":2.8800118919903962e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":143,"both_estimated_voiced_frames":138,"reference_voiced_output_unvoiced_frames":33,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.8070175438596491,"absolute_difference_cents_median":1.5540392710952147,"absolute_difference_cents_p95":9.59359697751977,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":0.00010701965100918563,"log_rms_mae_db":1.15140605842754,"rms_envelope_nrmse":0.07402440335624257,"envelope_correlation":0.994356204922796,"rms_gain_error_db":-0.07760540681569036,"reference_mean_power":0.0014643132072841782,"output_mean_power":0.0014383793476541935},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":5.509210445243936e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038836857080924806,"output_power":0.00037365709981419576,"reference_active":true,"delta_db":-0.16770838209522246},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.320093144841145e-05,"output_power":2.0171152299691708e-05,"reference_active":true,"delta_db":-0.6077471239235631},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.17358638897755e-06,"output_power":9.296183979670668e-06,"reference_active":true,"delta_db":0.05765555072206824}],"active_band_count":3,"active_band_mae_db":0.2777036855802846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":0.00010161161302678278,"log_rms_mae_db":0.625751848377753,"rms_envelope_nrmse":0.03693708285796823,"envelope_correlation":0.9985486312477239,"rms_gain_error_db":-0.07760516107941808,"reference_mean_power":0.0014741623423244587,"output_mean_power":0.0014480541305778906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":4.939959887293225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035657053350397007,"output_power":0.00034122580365394375,"reference_active":true,"delta_db":-0.19103585538524215},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.172203979326598e-06,"output_power":7.823155138763467e-06,"reference_active":true,"delta_db":-0.18957255443521653},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.984441222185338e-07,"output_power":5.918308179642122e-07,"reference_active":true,"delta_db":-0.7193409077565814}],"active_band_count":3,"active_band_mae_db":0.36664977252568004}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":1.3384644388660916,"rms_envelope_nrmse":0.09090205415967822,"envelope_correlation":0.9933260832327894,"rms_gain_error_db":-0.11890294543462165,"reference_mean_power":0.0006933001790048939,"output_mean_power":0.0006745762049478206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":3.8791214865659925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027195601224651783,"output_power":0.0002540247815556448,"reference_active":true,"delta_db":-0.29622577816779183},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.187110727790427e-05,"output_power":1.7822588078311396e-05,"reference_active":true,"delta_db":-0.889000011554843},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.698791354460839e-06,"output_power":6.2416530787429454e-06,"reference_active":true,"delta_db":-0.3069682493815287}],"active_band_count":3,"active_band_mae_db":0.49739801303472114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":0.8371680750748424,"rms_envelope_nrmse":0.04803666064079019,"envelope_correlation":0.9981552584449145,"rms_gain_error_db":-0.11890298010661843,"reference_mean_power":0.000697963583829786,"output_mean_power":0.0006791136596684403},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.5511880459540643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002583969788516632,"output_power":0.00023956937709221604,"reference_active":true,"delta_db":-0.3285612797106403},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105856462043575e-06,"output_power":7.034906954613283e-06,"reference_active":true,"delta_db":-0.6154055129200449},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388944870715121e-07,"output_power":4.178394319913965e-07,"reference_active":true,"delta_db":-1.8441971823195784}],"active_band_count":3,"active_band_mae_db":0.9293879916500879}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":1.4705696658892908,"rms_envelope_nrmse":0.13634985067687627,"envelope_correlation":0.9871823443419636,"rms_gain_error_db":-0.3203581825831445,"reference_mean_power":8.288074215773489e-05,"output_mean_power":7.698707383689885e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":4.287875385442727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2411523150580798e-05,"output_power":1.9268526594422303e-05,"reference_active":true,"delta_db":-0.6562286667297514},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.113877152498062e-06,"output_power":2.890502750304436e-06,"reference_active":true,"delta_db":-0.32328088007481776},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.881539635700257e-07,"output_power":9.729121510889573e-07,"reference_active":true,"delta_db":-0.06750989471774432}],"active_band_count":3,"active_band_mae_db":0.3490064805074378}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":0.9644428784891441,"rms_envelope_nrmse":0.0740990945659103,"envelope_correlation":0.9967407270583574,"rms_gain_error_db":-0.3203583129964744,"reference_mean_power":8.343814618382183e-05,"output_mean_power":7.750483840924118e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.9683446107037285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0633950863405697e-05,"output_power":1.7787172450197783e-05,"reference_active":true,"delta_db":-0.6447547634410087},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4408129681959182e-06,"output_power":1.4002600117767219e-06,"reference_active":true,"delta_db":-0.12398922212538019},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9896369199766395e-08,"output_power":4.531824219815416e-08,"reference_active":true,"delta_db":-0.41795888419594696}],"active_band_count":3,"active_band_mae_db":0.39556762325411193}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":1.7917014925755315,"rms_envelope_nrmse":0.21276898608036396,"envelope_correlation":0.9698488736231436,"rms_gain_error_db":-0.42085074848198367,"reference_mean_power":1.5917473583289074e-05,"output_mean_power":1.4447378903196747e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":8.500689890351622e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.331755201814065e-06,"output_power":3.912379894188396e-06,"reference_active":true,"delta_db":-0.4422288720241627},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3672340908387865e-06,"output_power":1.3008572546938496e-06,"reference_active":true,"delta_db":-0.21613235322556035},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7907435630616127e-07,"output_power":3.859330021457475e-07,"reference_active":true,"delta_db":1.4079198612197874}],"active_band_count":3,"active_band_mae_db":0.6887603621565036}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":1.5887866305968543,"rms_envelope_nrmse":0.13794042600513565,"envelope_correlation":0.9879522385852132,"rms_gain_error_db":-0.42085169706580616,"reference_mean_power":1.6024474415231717e-05,"output_mean_power":1.4544494251765766e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":7.9315957267754e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.299772492975258e-06,"output_power":3.888836771953732e-06,"reference_active":true,"delta_db":-0.43625762378667476},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.410418312822683e-07,"output_power":7.425488666471418e-07,"reference_active":true,"delta_db":0.008823152760751938},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.910690257092516e-08,"output_power":3.162389210943721e-08,"reference_active":true,"delta_db":2.1882503058845475}],"active_band_count":3,"active_band_mae_db":0.8777770274773248}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.134223699569702,"predicted_mos":3.134223699569702,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":71920,"input_sample_rate":16000,"input_seconds":4.495,"input_float32_sha256":"613eca2641425d0a9fd211bd2ff06a61bcae5afcdd4417ccdb0750f8f2866ea7","internal_samples":71920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1659158099992055},"started_at":"2026-10-13T04:42:20.551739+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0925000171700958e-05,"native_input_samples":71920,"native_input_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","output_sha256":"c9c1bb6b08cc6e6828736f03077f2ad6fcf9213d74bde72183975ae898b733f0","output_samples":71920,"length_delta_samples":0,"analysis_s":1.2482393310001498,"synthesis_s":0.07920189800006483,"total_s":1.3274412290002147,"rtf":0.29531506763074855,"output_peak":0.3777212049937972,"output_rms":0.043684284268624446,"finished_at":"2026-10-13T04:42:21.884435+00:00","evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"c9c1bb6b08cc6e6828736f03077f2ad6fcf9213d74bde72183975ae898b733f0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1408132042127574,"spectral_convergence":0.5493492911478298,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.470379123493776,"spectral_convergence":0.30301407196976643,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8544945009626135,"spectral_convergence":0.3285602634582787,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.0001921013198757407,"delta_db":-0.33306358917566,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0008470094960795302,"delta_db":1.1243114033747303,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.0007325862286935073,"delta_db":1.757776584829184,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":4.581290482703588e-05,"delta_db":1.3491722924571872,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":6.984907504339788e-05,"delta_db":0.5882399478127097,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":2.0957667542777324e-05,"delta_db":0.6297103431167972,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9637123601277113,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1549980192513087,"pitch":null,"output_rms_dbfs":-27.19349550909743,"output_peak":0.3777212049937972,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.821895609556382},"common_prefix":{"samples":71920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1408132042127574,"spectral_convergence":0.5493492911478298,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.470379123493776,"spectral_convergence":0.30301407196976643,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8544945009626135,"spectral_convergence":0.3285602634582787,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.0001921013198757407,"delta_db":-0.33306358917566,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0008470094960795302,"delta_db":1.1243114033747303,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.0007325862286935073,"delta_db":1.757776584829184,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":4.581290482703588e-05,"delta_db":1.3491722924571872,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":6.984907504339788e-05,"delta_db":0.5882399478127097,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":2.0957667542777324e-05,"delta_db":0.6297103431167972,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9637123601277113,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1549980192513087,"pitch":null,"output_rms_dbfs":-27.19349550909743,"output_peak":0.3777212049937972,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.821895609556382},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71920,"output_samples":71920,"intelligibility":{"estoi":{"value":0.9313477405238548,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.038245057973635734,"rms_dbfs":-28.34849352834874,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.396015589762875,"mean":-1.857829703902774e-05,"zero_energy":false},"output":{"rms":0.04368428426862445,"rms_dbfs":-27.19349550909743,"peak":0.3777212049937972,"peak_dbfs":-8.456572670233145,"crest_factor_db":18.736922838864285,"mean":1.829377093574235e-09,"zero_energy":false},"rms_difference_db":1.154998019251309},"full_scale":{"reference":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":7.500810143410044e-08,"output_rms_dbfs":-71.24891827033483,"output_zero_energy":false,"output_peak":0.003200659558495423,"output_energy_fraction":5.683833985038061e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":220,"both_estimated_voiced_frames":166,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":54,"paired_voiced_fraction":0.9707602339181286,"absolute_difference_cents_median":5.137592277064394,"absolute_difference_cents_p95":20.552788454029052,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":0.00010701965100918563,"log_rms_mae_db":2.510092975441095,"rms_envelope_nrmse":0.19264904203850894,"envelope_correlation":0.9868449012950155,"rms_gain_error_db":1.154998154199917,"reference_mean_power":0.0014643132072841782,"output_mean_power":0.0019104417256407622},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":5.509210445243936e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038836857080924806,"output_power":0.0005444516400830904,"reference_active":true,"delta_db":1.4671523353721923},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.320093144841145e-05,"output_power":3.139643014134407e-05,"reference_active":true,"delta_db":1.313748495552135},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.17358638897755e-06,"output_power":9.781980723797854e-06,"reference_active":true,"delta_db":0.278876475434054}],"active_band_count":3,"active_band_mae_db":1.0199257687861272}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":0.00010161161302678278,"log_rms_mae_db":2.106827246469723,"rms_envelope_nrmse":0.17034804822911376,"envelope_correlation":0.993509656207022,"rms_gain_error_db":1.1549986970097799,"reference_mean_power":0.0014741623423244587,"output_mean_power":0.0019232918115709657},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":4.939959887293225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035657053350397007,"output_power":0.0005002753324436872,"reference_active":true,"delta_db":1.470636384097042},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.172203979326598e-06,"output_power":1.0899810084623388e-05,"reference_active":true,"delta_db":1.2507973283037734},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.984441222185338e-07,"output_power":6.654280736761835e-07,"reference_active":true,"delta_db":-0.21030547557023366}],"active_band_count":3,"active_band_mae_db":0.9772463959903498}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":1.75085451368951,"rms_envelope_nrmse":0.22510869258379063,"envelope_correlation":0.989366392235944,"rms_gain_error_db":1.4848701818250996,"reference_mean_power":0.0006933001790048939,"output_mean_power":0.0009759067694160202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":3.8791214865659925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027195601224651783,"output_power":0.0004029653136679247,"reference_active":true,"delta_db":1.7076900028575082},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.187110727790427e-05,"output_power":3.257945320340141e-05,"reference_active":true,"delta_db":1.7307302025418794},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.698791354460839e-06,"output_power":6.005844729916551e-06,"reference_active":true,"delta_db":-0.4742235102070651}],"active_band_count":3,"active_band_mae_db":1.3042145718688174}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":1.4529457896071836,"rms_envelope_nrmse":0.20675863791771393,"envelope_correlation":0.9944642303305418,"rms_gain_error_db":1.4848700916582127,"reference_mean_power":0.000697963583829786,"output_mean_power":0.0009824710749407914},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.5511880459540643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002583969788516632,"output_power":0.0003837716666352408,"reference_active":true,"delta_db":1.7178547655363978},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105856462043575e-06,"output_power":1.2207316914904216e-05,"reference_active":true,"delta_db":1.778213104484857},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388944870715121e-07,"output_power":4.10602217701902e-07,"reference_active":true,"delta_db":-1.9200784997679174}],"active_band_count":3,"active_band_mae_db":1.8053821232630576}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":1.383331745107413,"rms_envelope_nrmse":0.17808688981652016,"envelope_correlation":0.9871581687411959,"rms_gain_error_db":0.8956461588121871,"reference_mean_power":8.288074215773489e-05,"output_mean_power":0.00010186341858167356},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":4.287875385442727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2411523150580798e-05,"output_power":2.8913094294743355e-05,"reference_active":true,"delta_db":1.1062319916358643},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.113877152498062e-06,"output_power":4.607912212535876e-06,"reference_active":true,"delta_db":1.702027216005085},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.881539635700257e-07,"output_power":1.503993998376236e-06,"reference_active":true,"delta_db":1.8242148624975858}],"active_band_count":3,"active_band_mae_db":1.5441580233795118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":1.0112156805850947,"rms_envelope_nrmse":0.14187594349265922,"envelope_correlation":0.9943686518624738,"rms_gain_error_db":0.8956473257944315,"reference_mean_power":8.343814618382183e-05,"output_mean_power":0.00010254851577042326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.9683446107037285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0633950863405697e-05,"output_power":2.6743422228610746e-05,"reference_active":true,"delta_db":1.12634589081311},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4408129681959182e-06,"output_power":2.085936456321807e-06,"reference_active":true,"delta_db":1.606934656772867},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9896369199766395e-08,"output_power":6.05773372281959e-08,"reference_active":true,"delta_db":0.8424123477802847}],"active_band_count":3,"active_band_mae_db":1.191897631788754}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":1.39206757539854,"rms_envelope_nrmse":0.26132125918383264,"envelope_correlation":0.9613866556118409,"rms_gain_error_db":0.6460415534067003,"reference_mean_power":1.5917473583289074e-05,"output_mean_power":1.847048473539861e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":8.500689890351622e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.331755201814065e-06,"output_power":4.983074308582957e-06,"reference_active":true,"delta_db":0.6083345791110478},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3672340908387865e-06,"output_power":1.7425922741134994e-06,"reference_active":true,"delta_db":1.053529056144671},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7907435630616127e-07,"output_power":5.349415819578074e-07,"reference_active":true,"delta_db":2.8258642598695887}],"active_band_count":3,"active_band_mae_db":1.4959092983751026}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":0.9658503242368504,"rms_envelope_nrmse":0.14639111728186496,"envelope_correlation":0.9898797709042503,"rms_gain_error_db":0.6460379939224166,"reference_mean_power":1.6024474415231717e-05,"output_mean_power":1.859463224142846e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":7.9315957267754e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.299772492975258e-06,"output_power":5.191369948537358e-06,"reference_active":true,"delta_db":0.8183650171556145},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.410418312822683e-07,"output_power":9.958870896911951e-07,"reference_active":true,"delta_db":1.2836737809104815},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.910690257092516e-08,"output_power":4.460979010612563e-08,"reference_active":true,"delta_db":3.6823989076833734}],"active_band_count":3,"active_band_mae_db":1.92814590191649}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"facodec","status":"measured","mos":{"status":"measured","value":4.262688636779785,"predicted_mos":4.262688636779785,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":72000,"input_sample_rate":16000,"input_seconds":4.5,"input_float32_sha256":"f9e473225a44ff0280df56bb22ef7290ebcddeaaad268fd6e078b60dfc068e21","internal_samples":72000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.11508268200123},"started_at":"2026-10-13T04:53:19.027742+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.222700004494982e-05,"native_input_samples":71920,"native_input_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","output_sha256":"a09a6bb229d3c74882ac717295b1787eabd0dcc1f22fdb5198afac47e178c6a4","output_samples":72000,"length_delta_samples":80,"analysis_s":1.290618192000693,"synthesis_s":10.294397100999959,"total_s":11.585015293000652,"rtf":2.5773115223583205,"output_peak":0.2742713391780853,"output_rms":0.03623811254216618,"finished_at":"2026-10-13T04:53:30.618711+00:00","evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"a09a6bb229d3c74882ac717295b1787eabd0dcc1f22fdb5198afac47e178c6a4","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":71920,"output_samples":72000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.81668861166442,"output_peak":0.2742713391780853},"common_prefix":{"samples":71920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988888888888889,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.363654497347839,"spectral_convergence":0.24324543896025838,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.645460162687262,"spectral_convergence":0.22710051126990655,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.730251125305541,"spectral_convergence":0.24216300570749932,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.00021444845014975027,"delta_db":0.14486204168456038,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0005622600452160464,"delta_db":-0.6551991583923993,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.00044926889336410444,"delta_db":-0.3657474216382747,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":2.7184314669753368e-05,"delta_db":-0.9175221171857693,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":4.466418980387512e-05,"delta_db":-1.353772042715181,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":1.6835625841237257e-05,"delta_db":-0.32142647107607675,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6264215421153768,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.46336696113710407,"pitch":null,"output_rms_dbfs":-28.811860489485845,"output_peak":0.2742713391780853,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5797885951135475},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71920,"output_samples":71920,"intelligibility":{"estoi":{"value":0.8999292172039117,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.038245057973635734,"rms_dbfs":-28.34849352834874,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.396015589762875,"mean":-1.857829703902774e-05,"zero_energy":false},"output":{"rms":0.0362582613902648,"rms_dbfs":-28.811860489485845,"peak":0.2742713391780853,"peak_dbfs":-11.236391459075124,"crest_factor_db":17.57546903041072,"mean":-0.00017759337675601853,"zero_energy":false},"rms_difference_db":-0.46336696113710474},"full_scale":{"reference":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":5.547759007938626e-07,"output_rms_dbfs":-62.55882412738556,"output_zero_energy":false,"output_peak":0.0031615307088941336,"output_energy_fraction":6.102212072689047e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":183,"both_estimated_voiced_frames":167,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9766081871345029,"absolute_difference_cents_median":5.7152235404392275,"absolute_difference_cents_p95":15.870608410735777,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":0.00010701965100918563,"log_rms_mae_db":1.5092116474276411,"rms_envelope_nrmse":0.11143237709799782,"envelope_correlation":0.9894092248402062,"rms_gain_error_db":-0.46336831276021784,"reference_mean_power":0.0014643132072841782,"output_mean_power":0.0013161250288383736},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":5.509210445243936e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038836857080924806,"output_power":0.0003511591155219243,"reference_active":true,"delta_db":-0.4374013037797785},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.320093144841145e-05,"output_power":2.388048916325224e-05,"reference_active":true,"delta_db":0.12537797678348994},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.17358638897755e-06,"output_power":1.0459899104305475e-05,"reference_active":true,"delta_db":0.5698834023782328}],"active_band_count":3,"active_band_mae_db":0.37755422764716706}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":0.00010161161302678278,"log_rms_mae_db":1.2060621983448085,"rms_envelope_nrmse":0.09665217775402761,"envelope_correlation":0.992270300663644,"rms_gain_error_db":-0.463374207319838,"reference_mean_power":0.0014741623423244587,"output_mean_power":0.0013249756351981369},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":4.939959887293225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035657053350397007,"output_power":0.0003227713213081468,"reference_active":true,"delta_db":-0.43250510745867154},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.172203979326598e-06,"output_power":8.612372406663216e-06,"reference_active":true,"delta_db":0.2278360269293692},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.984441222185338e-07,"output_power":8.638129413728069e-07,"reference_active":true,"delta_db":0.9228803934173035}],"active_band_count":3,"active_band_mae_db":0.5277405092684481}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":1.4136769284824637,"rms_envelope_nrmse":0.10886071592388574,"envelope_correlation":0.9906888500617584,"rms_gain_error_db":-0.2512061021346796,"reference_mean_power":0.0006933001790048939,"output_mean_power":0.0006543358302270922},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":3.8791214865659925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027195601224651783,"output_power":0.0002498674095324748,"reference_active":true,"delta_db":-0.36789050052152417},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.187110727790427e-05,"output_power":2.3134931130445922e-05,"reference_active":true,"delta_db":0.24397440195281972},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.698791354460839e-06,"output_power":8.41023688537148e-06,"reference_active":true,"delta_db":0.9881177730213458}],"active_band_count":3,"active_band_mae_db":0.5333275584985633}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":1.0542553306801226,"rms_envelope_nrmse":0.08836435996334235,"envelope_correlation":0.9936883373391574,"rms_gain_error_db":-0.2512068969364962,"reference_mean_power":0.000697963583829786,"output_mean_power":0.0006587370252348345},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.5511880459540643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002583969788516632,"output_power":0.000238812551783584,"reference_active":true,"delta_db":-0.3423028242667774},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105856462043575e-06,"output_power":8.45665121478227e-06,"reference_active":true,"delta_db":0.18399509948962145},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388944870715121e-07,"output_power":7.957859648796552e-07,"reference_active":true,"delta_db":0.9536713462787471}],"active_band_count":3,"active_band_mae_db":0.4933230900117153}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":1.9678273491328717,"rms_envelope_nrmse":0.2088525677233717,"envelope_correlation":0.9792147707439265,"rms_gain_error_db":-1.2699760603257007,"reference_mean_power":8.288074215773489e-05,"output_mean_power":6.186656810673816e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":4.287875385442727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2411523150580798e-05,"output_power":1.980113901962808e-05,"reference_active":true,"delta_db":-0.53781200547747},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.113877152498062e-06,"output_power":3.1888192793799976e-06,"reference_active":true,"delta_db":0.10328432129968756},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.881539635700257e-07,"output_power":1.0026062832671063e-06,"reference_active":true,"delta_db":0.0630580500453248}],"active_band_count":3,"active_band_mae_db":0.23471812560749414}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":1.7062950885327808,"rms_envelope_nrmse":0.18931855734924452,"envelope_correlation":0.9848749281907015,"rms_gain_error_db":-1.2699861491247568,"reference_mean_power":8.343814618382183e-05,"output_mean_power":6.228249925908138e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.9683446107037285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0633950863405697e-05,"output_power":1.840127137796534e-05,"reference_active":true,"delta_db":-0.49734561743737},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4408129681959182e-06,"output_power":1.4767339365611698e-06,"reference_active":true,"delta_db":0.10694646694804602},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9896369199766395e-08,"output_power":5.401874498362529e-08,"reference_active":true,"delta_db":0.34475545470289126}],"active_band_count":3,"active_band_mae_db":0.31634917969610243}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":2.254116340710732,"rms_envelope_nrmse":0.2951019166121748,"envelope_correlation":0.9436212677073996,"rms_gain_error_db":0.07211423591718445,"reference_mean_power":1.5917473583289074e-05,"output_mean_power":1.6183988507778434e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":8.500689890351622e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.331755201814065e-06,"output_power":3.630978417552389e-06,"reference_active":true,"delta_db":-0.7664023810816496},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3672340908387865e-06,"output_power":1.6905705084185701e-06,"reference_active":true,"delta_db":0.9219040992921244},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7907435630616127e-07,"output_power":4.194459742642043e-07,"reference_active":true,"delta_db":1.7695609880417194}],"active_band_count":3,"active_band_mae_db":1.152622489471831}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":1.9643249286738305,"rms_envelope_nrmse":0.2297894474734354,"envelope_correlation":0.9635947338547078,"rms_gain_error_db":0.07206287631229118,"reference_mean_power":1.6024474415231717e-05,"output_mean_power":1.629258823583289e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":7.9315957267754e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.299772492975258e-06,"output_power":3.698499985835781e-06,"reference_active":true,"delta_db":-0.6541985567945562},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.410418312822683e-07,"output_power":9.647611001290952e-07,"reference_active":true,"delta_db":1.1457705979425492},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.910690257092516e-08,"output_power":3.3992685201694216e-08,"reference_active":true,"delta_db":2.5019518331016672}],"active_band_count":3,"active_band_mae_db":1.4339736626129242}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.565258026123047,"predicted_mos":3.565258026123047,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":198656,"input_sample_rate":44100,"input_seconds":4.504671201814059,"input_float32_sha256":"7597ea4fd85e8ae81a12ed4579db66e7baa1de2ddd7ba42f954047620b71dfd0","internal_samples":72075,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5833535099991423},"started_at":"2026-10-13T07:16:35.633626+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004942420999213937,"native_input_samples":198230,"native_input_sha256":"fc194e15ddc21c515bb7d83a0f4464768923465522cab9413d9429e0acc710eb","output_sha256":"f1637f752a37133cce8cd8846e27a748f6c70675b0d16c7ce758b4da070ba080","output_samples":198656,"length_delta_samples":426,"analysis_s":0.10791948000041884,"synthesis_s":28.260939439998765,"total_s":28.368858919999184,"rtf":6.311187400352944,"output_peak":0.24512827396392822,"output_rms":0.034557787712732164,"finished_at":"2026-10-13T07:17:04.016167+00:00","evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"f1637f752a37133cce8cd8846e27a748f6c70675b0d16c7ce758b4da070ba080","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":71920,"output_samples":72075,"length_delta_samples":155,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.22333621892182,"output_peak":0.2455558186693264},"common_prefix":{"samples":71920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9978494623655914,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.161177799023863,"spectral_convergence":0.4345299125127897,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.939181011607777,"spectral_convergence":0.34660350829055303,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.195017807852182,"spectral_convergence":0.3886181892209513,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":9.68730832979097e-05,"delta_db":-3.306335849307877,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0005843787675974229,"delta_db":-0.4876271082522043,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.00042483188215020897,"delta_db":-0.6086399038366985,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":2.79325511568661e-05,"delta_db":-0.7995999758774629,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":4.8604870247989594e-05,"delta_db":-0.986568765782248,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":1.5777594611382906e-05,"delta_db":-0.6033111930186787,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1320137993458617,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.865493001440555,"pitch":null,"output_rms_dbfs":-29.213986529789292,"output_peak":0.2455558186693264,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.765125539494607},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71920,"output_samples":71920,"intelligibility":{"estoi":{"value":0.768961600316774,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.038245057973635734,"rms_dbfs":-28.34849352834874,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.396015589762875,"mean":-1.857829703902774e-05,"zero_energy":false},"output":{"rms":0.034617896369678214,"rms_dbfs":-29.213986529789292,"peak":0.2455558186693264,"peak_dbfs":-12.196995408123394,"crest_factor_db":17.016991121665896,"mean":-0.00013462864657252203,"zero_energy":false},"rms_difference_db":-0.8654930014405515},"full_scale":{"reference":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":1.63449286630336e-07,"output_rms_dbfs":-67.86616970545127,"output_zero_energy":false,"output_peak":0.0028840433798590502,"output_energy_fraction":1.9722653417302716e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":195,"both_estimated_voiced_frames":160,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.935672514619883,"absolute_difference_cents_median":9.870328685741757,"absolute_difference_cents_p95":34.153928095541126,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":0.00010701965100918563,"log_rms_mae_db":2.0717861983967563,"rms_envelope_nrmse":0.1317198914351861,"envelope_correlation":0.9907829496426241,"rms_gain_error_db":-0.8654928287198496,"reference_mean_power":0.0014643132072841782,"output_mean_power":0.001199733253709479},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":5.509210445243936e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038836857080924806,"output_power":0.00031778220863993694,"reference_active":true,"delta_db":-0.8711449772418203},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.320093144841145e-05,"output_power":2.16268015049917e-05,"reference_active":true,"delta_db":-0.3051312665107323},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.17358638897755e-06,"output_power":8.026869918325818e-06,"reference_active":true,"delta_db":-0.5799293017578268}],"active_band_count":3,"active_band_mae_db":0.5854018485034597}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":0.00010161161302678278,"log_rms_mae_db":1.7033540079068654,"rms_envelope_nrmse":0.10855692948399717,"envelope_correlation":0.9967879245432962,"rms_gain_error_db":-0.8654920548127314,"reference_mean_power":0.0014741623423244587,"output_mean_power":0.0012078030095047884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":4.939959887293225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035657053350397007,"output_power":0.00029089405079325224,"reference_active":true,"delta_db":-0.8841061144289639},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.172203979326598e-06,"output_power":7.945522410776572e-06,"reference_active":true,"delta_db":-0.12216741201104457},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.984441222185338e-07,"output_power":5.000427288685369e-07,"reference_active":true,"delta_db":-1.451245503073979}],"active_band_count":3,"active_band_mae_db":0.8191730098379958}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":2.36646939585967,"rms_envelope_nrmse":0.15452195070075264,"envelope_correlation":0.9817303621099613,"rms_gain_error_db":-0.46696877378504376,"reference_mean_power":0.0006933001790048939,"output_mean_power":0.0006226219609787749},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":3.8791214865659925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027195601224651783,"output_power":0.0002267293025039592,"reference_active":true,"delta_db":-0.7899101239927275},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.187110727790427e-05,"output_power":1.7843341993509045e-05,"reference_active":true,"delta_db":-0.8839457136030477},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.698791354460839e-06,"output_power":5.770390517090988e-06,"reference_active":true,"delta_db":-0.6479124566833215}],"active_band_count":3,"active_band_mae_db":0.7739227647596989}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":1.6886983945244394,"rms_envelope_nrmse":0.12217377223254054,"envelope_correlation":0.9887067790790867,"rms_gain_error_db":-0.4669687970873932,"reference_mean_power":0.000697963583829786,"output_mean_power":0.0006268099534431358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.5511880459540643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002583969788516632,"output_power":0.00021287775644928547,"reference_active":true,"delta_db":-0.8415714715808109},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105856462043575e-06,"output_power":6.477524736865841e-06,"reference_active":true,"delta_db":-0.9738982883859344},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388944870715121e-07,"output_power":3.9887113350334526e-07,"reference_active":true,"delta_db":-2.0459653333819285}],"active_band_count":3,"active_band_mae_db":1.2871450311162247}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":4.271401370413668,"rms_envelope_nrmse":0.24669480032782365,"envelope_correlation":0.9615082166113497,"rms_gain_error_db":-0.9919092717906909,"reference_mean_power":8.288074215773489e-05,"output_mean_power":6.59572749256725e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":4.287875385442727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2411523150580798e-05,"output_power":1.9564216286400778e-05,"reference_active":true,"delta_db":-0.5900891803112522},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.113877152498062e-06,"output_power":2.757110189791869e-06,"reference_active":true,"delta_db":-0.5284735162045086},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.881539635700257e-07,"output_power":7.921522158758116e-07,"reference_active":true,"delta_db":-0.9601597558427634}],"active_band_count":3,"active_band_mae_db":0.6929074841195081}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":3.513616866740866,"rms_envelope_nrmse":0.19520749598193682,"envelope_correlation":0.9781810562664194,"rms_gain_error_db":-0.9919086789515624,"reference_mean_power":8.343814618382183e-05,"output_mean_power":6.640087136168704e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.9683446107037285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0633950863405697e-05,"output_power":1.7937983935113688e-05,"reference_active":true,"delta_db":-0.6080876126715876},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4408129681959182e-06,"output_power":1.1368907475549445e-06,"reference_active":true,"delta_db":-1.0288887669850835},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9896369199766395e-08,"output_power":3.708017099140152e-08,"reference_active":true,"delta_db":-1.2892721636222604}],"active_band_count":3,"active_band_mae_db":0.9754161810929771}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":4.973656844766718,"rms_envelope_nrmse":0.22164306195864195,"envelope_correlation":0.9703726048511044,"rms_gain_error_db":-0.6442927043462143,"reference_mean_power":1.5917473583289074e-05,"output_mean_power":1.372286741399834e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":8.500689890351622e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.331755201814065e-06,"output_power":3.7109219170861315e-06,"reference_active":true,"delta_db":-0.6718208930266302},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3672340908387865e-06,"output_power":1.0726003407918901e-06,"reference_active":true,"delta_db":-1.0540494800417504},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7907435630616127e-07,"output_power":1.543610781097174e-07,"reference_active":true,"delta_db":-2.571821285316479}],"active_band_count":3,"active_band_mae_db":1.4325638861282866}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":4.551166669416632,"rms_envelope_nrmse":0.14818812089960498,"envelope_correlation":0.9889797248527737,"rms_gain_error_db":-0.6442805446662503,"reference_mean_power":1.6024474415231717e-05,"output_mean_power":1.3815154291017014e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":7.9315957267754e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.299772492975258e-06,"output_power":3.624600310670975e-06,"reference_active":true,"delta_db":-0.741853537168877},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.410418312822683e-07,"output_power":5.447191024540101e-07,"reference_active":true,"delta_db":-1.3367011871068166},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.910690257092516e-08,"output_power":1.1971392880249729e-08,"reference_active":true,"delta_db":-2.030456052512814}],"active_band_count":3,"active_band_mae_db":1.3696702589295027}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"golf","status":"measured","mos":{"status":"measured","value":2.2706685066223145,"predicted_mos":2.2706685066223145,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":107760,"input_sample_rate":24000,"input_seconds":4.49,"input_float32_sha256":"9edb89266a063a27f5576bc19b74fab15a83ed8a946f69eb4defa41b454af160","internal_samples":71840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9629319139985455},"started_at":"2026-10-13T05:02:38.037802+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027628999996522907,"native_input_samples":107880,"native_input_sha256":"9d61802c48b0beee2bb122db3f386590e58728908212024524d2a3d09270e30c","output_sha256":"fc1a9320f26cf8256963ac7bbe8cd3799b18b561207530c22cc7de77fe9a2c58","output_samples":107760,"length_delta_samples":-120,"analysis_s":0.392802323000069,"synthesis_s":0.10733463700034918,"total_s":0.5001369600004182,"rtf":0.11126517463858024,"output_peak":0.24795019626617432,"output_rms":0.03391970242459476,"finished_at":"2026-10-13T05:02:38.546989+00:00","analysis_stages":{"external_f0_s":0.040868939000574755},"evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"fc1a9320f26cf8256963ac7bbe8cd3799b18b561207530c22cc7de77fe9a2c58","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":71920,"output_samples":71840,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.386708768117856,"output_peak":0.24652776763329462},"common_prefix":{"samples":71840,"sample_rate":16000,"reference_coverage":0.9988876529477196,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71840,"output_samples":71840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.119212324431357,"spectral_convergence":0.43964289138819945,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.579851199708184,"spectral_convergence":0.32130174326946315,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.672370303023565,"spectral_convergence":0.3348782997912861,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020764128045074382,"output_power":0.00016657876466582961,"delta_db":-0.9569406128716729,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006552104368557998,"output_power":0.0005811348052432092,"delta_db":-0.5210391976892247,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004886268985825736,"output_power":0.000337409133124664,"delta_db":-1.6082053684697362,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.361661304501907e-05,"output_power":2.204448459587017e-05,"delta_db":-1.8325400569360983,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.106882605910341e-05,"output_power":3.385303422531132e-05,"delta_db":-2.562219712316478,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8149106314664746e-05,"output_power":1.0652613463678498e-05,"delta_db":-2.313990760024038,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6324892847178745,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0430484266246918,"pitch":null,"output_rms_dbfs":-29.386708768117856,"output_peak":0.24652776763329462,"reference_rms_dbfs":-28.343660341493162,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.790477942387702},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71840,"output_samples":71840,"intelligibility":{"estoi":{"value":0.8871463709185101,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03826634502154477,"rms_dbfs":-28.343660341493162,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.3911824029073,"mean":-1.837554116026065e-05,"zero_energy":false},"output":{"rms":0.03393630556378468,"rms_dbfs":-29.386708768117856,"peak":0.24652776763329462,"peak_dbfs":-12.162683138222725,"crest_factor_db":17.22402562989513,"mean":0.00027499977090740714,"zero_energy":false},"rms_difference_db":-1.0430484266246935},"full_scale":{"reference":{"sample_count":71840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":3.6808831012453424e-07,"output_rms_dbfs":-64.34047974808222,"output_zero_energy":false,"output_peak":0.0022011515707041455,"output_energy_fraction":4.6268979154371514e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":181,"both_estimated_voiced_frames":161,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9415204678362573,"absolute_difference_cents_median":5.231282110254054,"absolute_difference_cents_p95":19.987734280777364,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":897,"rms_floor":0.00010701965100918563,"log_rms_mae_db":2.4650111672101764,"rms_envelope_nrmse":0.20795084921250329,"envelope_correlation":0.9658976180797174,"rms_gain_error_db":-1.043048701010961,"reference_mean_power":0.0014659455479467923,"output_mean_power":0.001152956624013947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":897,"frequency_resolution_hz":0.22296544035674468,"power_floor":5.509991319317731e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003884194066525961,"output_power":0.00026942612371427235,"reference_active":true,"delta_db":-1.5886121761826055},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3137792863913178e-05,"output_power":2.898934335683826e-05,"reference_active":true,"delta_db":0.9791644935154747},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.181967807605087e-06,"output_power":9.094919973245383e-06,"reference_active":true,"delta_db":-0.04136883743826944}],"active_band_count":3,"active_band_mae_db":0.8697151690454499}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":891,"rms_floor":0.00010161161302678278,"log_rms_mae_db":1.836774498135789,"rms_envelope_nrmse":0.1750364775926278,"envelope_correlation":0.9786062618218291,"rms_gain_error_db":-1.0430517181438685,"reference_mean_power":0.0014758167377078339,"output_mean_power":0.001160719444082309},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":891,"frequency_resolution_hz":0.2244668911335578,"power_floor":4.940385317815735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035653894830067496,"output_power":0.0002518181898890006,"reference_active":true,"delta_db":-1.5101988126603478},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.147721046023141e-06,"output_power":1.1803547259039934e-05,"reference_active":true,"delta_db":1.6097639157891246},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.980135052148376e-07,"output_power":5.598669501591096e-07,"reference_active":true,"delta_db":-0.9577899431064512}],"active_band_count":3,"active_band_mae_db":1.3592508905186413}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":897,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":1.8344750829952077,"rms_envelope_nrmse":0.25113213530024786,"envelope_correlation":0.9548891049734571,"rms_gain_error_db":-1.089531658154739,"reference_mean_power":0.0006940730831021844,"output_mean_power":0.0005400724658562077},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":897,"frequency_resolution_hz":0.22296544035674468,"power_floor":3.879520725038155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002719838403817851,"output_power":0.00017209622703472796,"reference_active":true,"delta_db":-1.9877175253330344},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1836083340439715e-05,"output_power":2.3183639457663803e-05,"reference_active":true,"delta_db":0.2600687102993697},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.711187283645957e-06,"output_power":7.966704609291692e-06,"reference_active":true,"delta_db":0.7447935613764177}],"active_band_count":3,"active_band_mae_db":0.9975265990029406}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":891,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":1.3990866857760527,"rms_envelope_nrmse":0.2065458577031229,"envelope_correlation":0.9716257228978424,"rms_gain_error_db":-1.0895315907930534,"reference_mean_power":0.0006987469280388394,"output_mean_power":0.0005437092886558818},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":891,"frequency_resolution_hz":0.2244668911335578,"power_floor":3.551396390507111e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025835990752335524,"output_power":0.0001672638773416028,"reference_active":true,"delta_db":-1.8882296048117682},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.101716812403624e-06,"output_power":9.25706517069251e-06,"reference_active":true,"delta_db":0.5789626244441258},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.372209470381474e-07,"output_power":5.054846692643489e-07,"reference_active":true,"delta_db":-1.005820552010759}],"active_band_count":3,"active_band_mae_db":1.1576709270888843}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":897,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":1.9605933584926385,"rms_envelope_nrmse":0.2943146952612198,"envelope_correlation":0.9682508424236859,"rms_gain_error_db":-2.260215435293525,"reference_mean_power":8.297310580662603e-05,"output_mean_power":4.930782014185564e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":897,"frequency_resolution_hz":0.22296544035674468,"power_floor":4.285347242256674e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2375580745477273e-05,"output_power":1.1918545968570904e-05,"reference_active":true,"delta_db":-2.735510401554385},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.102750756767275e-06,"output_power":2.1492490664740064e-06,"reference_active":true,"delta_db":-1.5946014343794592},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.87802571564363e-07,"output_power":9.81293649350693e-07,"reference_active":true,"delta_db":-0.028711642085231668}],"active_band_count":3,"active_band_mae_db":1.452941159339692}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":891,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":1.6011827835890171,"rms_envelope_nrmse":0.25992489456182233,"envelope_correlation":0.9853979806324754,"rms_gain_error_db":-2.260215939762076,"reference_mean_power":8.353176966839058e-05,"output_mean_power":4.963980745894023e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":891,"frequency_resolution_hz":0.2244668911335578,"power_floor":3.965952872982214e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.059228618889432e-05,"output_power":1.1140773111532155e-05,"reference_active":true,"delta_db":-2.667892357152626},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.433538132443451e-06,"output_power":1.0079038882302765e-06,"reference_active":true,"delta_db":-1.5299012916883687},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.983537458445364e-08,"output_power":4.056735046218657e-08,"reference_active":true,"delta_db":-0.893610829993126}],"active_band_count":3,"active_band_mae_db":1.69713482627804}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":897,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":2.6965823205433255,"rms_envelope_nrmse":0.26644379676023133,"envelope_correlation":0.973193029614196,"rms_gain_error_db":-1.8581532426403549,"reference_mean_power":1.5935194768483994e-05,"output_mean_power":1.0388241845622814e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":897,"frequency_resolution_hz":0.22296544035674468,"power_floor":8.515320214150546e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.332539415800278e-06,"output_power":2.2843743269667093e-06,"reference_active":true,"delta_db":-2.7797525176634053},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3723569578967091e-06,"output_power":8.59962986144423e-07,"reference_active":true,"delta_db":-2.029873294621952},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7849730108975983e-07,"output_power":3.443521389019603e-07,"reference_active":true,"delta_db":0.9218179410434414}],"active_band_count":3,"active_band_mae_db":1.9104812511095997}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71840,"uncovered_tail_samples":0,"metrics":{"frame_count":891,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":2.5507932843752927,"rms_envelope_nrmse":0.2254319074517024,"envelope_correlation":0.9889027344684647,"rms_gain_error_db":-1.8581470486607754,"reference_mean_power":1.6042442014206544e-05,"output_mean_power":1.0458171835505564e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":891,"frequency_resolution_hz":0.2244668911335578,"power_floor":7.945248321971319e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3000212506479126e-06,"output_power":2.3553916256643516e-06,"reference_active":true,"delta_db":-2.614074752277981},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.435639899288383e-07,"output_power":4.7299729580395687e-07,"reference_active":true,"delta_db":-1.964596913368032},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9118555035980465e-08,"output_power":2.413327612317731e-08,"reference_active":true,"delta_db":1.0116121636826843}],"active_band_count":3,"active_band_mae_db":1.863427943109566}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3775715827941895,"predicted_mos":3.3775715827941895,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":108000,"input_sample_rate":24000,"input_seconds":4.5,"input_float32_sha256":"0462e7baa7ff425331e0c57a5e9ff4879fec1812ba1e4a8594f56e913687bcce","internal_samples":72000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.341152831999352},"started_at":"2026-10-13T07:38:37.938099+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028960719973838422,"native_input_samples":107880,"native_input_sha256":"9d61802c48b0beee2bb122db3f386590e58728908212024524d2a3d09270e30c","output_sha256":"f8aae7a4f417aeaaed6974545a2ce22be83329d7af3664a02a7248f21b85f323","output_samples":108000,"length_delta_samples":120,"analysis_s":2.0118711570030428,"synthesis_s":9.897283097998297,"total_s":11.90915425500134,"rtf":2.649422526140454,"output_peak":0.3388354480266571,"output_rms":0.037638287137051675,"finished_at":"2026-10-13T07:38:49.860709+00:00","analysis_stages":{"external_f0_s":1.4529298169982212},"evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"f8aae7a4f417aeaaed6974545a2ce22be83329d7af3664a02a7248f21b85f323","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":71920,"output_samples":72000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.483327898089485,"output_peak":0.33096954777935894},"common_prefix":{"samples":71920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988888888888889,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.083280156727544,"spectral_convergence":0.5000071625482503,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.56998747017,"spectral_convergence":0.30339326018305296,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.545755899519123,"spectral_convergence":0.2877600384739739,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.00018262027597562366,"delta_db":-0.5528771295768895,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0006361294209028603,"delta_db":-0.11911656923005025,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.000489855271700334,"delta_db":0.009866942136076315,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":3.407252976348054e-05,"delta_db":0.06333779165095507,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":6.29151515771581e-05,"delta_db":0.1341858257147647,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":1.3955153741271395e-05,"delta_db":-1.1363728704127676,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.33595952145358393,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1300062507147776,"pitch":null,"output_rms_dbfs":-28.478499779063515,"output_peak":0.33096954777935894,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.733007842138889},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71920,"output_samples":71920,"intelligibility":{"estoi":{"value":0.8950504788648553,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.038245057973635734,"rms_dbfs":-28.34849352834874,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.396015589762875,"mean":-1.857829703902774e-05,"zero_energy":false},"output":{"rms":0.03767688686264734,"rms_dbfs":-28.478499779063515,"peak":0.33096954777935894,"peak_dbfs":-9.60423926888372,"crest_factor_db":18.874260510179795,"mean":3.0218613316210265e-05,"zero_energy":false},"rms_difference_db":-0.130006250714775},"full_scale":{"reference":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":3.462701907255341e-08,"output_rms_dbfs":-74.60584893840762,"output_zero_energy":false,"output_peak":0.0022732431536068923,"output_energy_fraction":3.5273515441934986e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":214,"both_estimated_voiced_frames":168,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":46,"paired_voiced_fraction":0.9824561403508771,"absolute_difference_cents_median":5.140957428684542,"absolute_difference_cents_p95":23.939064991138924,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":0.00010701965100918563,"log_rms_mae_db":3.3306903281676945,"rms_envelope_nrmse":0.14514508073475602,"envelope_correlation":0.9793765041937167,"rms_gain_error_db":-0.13000607516820595,"reference_mean_power":0.0014643132072841782,"output_mean_power":0.0014211285748783695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":5.509210445243936e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038836857080924806,"output_power":0.00039865743478631054,"reference_active":true,"delta_db":0.11355791048508396},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.320093144841145e-05,"output_power":2.7927401735169876e-05,"reference_active":true,"delta_db":0.8052511151369779},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.17358638897755e-06,"output_power":9.19100514276079e-06,"reference_active":true,"delta_db":0.008238539806549277}],"active_band_count":3,"active_band_mae_db":0.3090158551428704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":0.00010161161302678278,"log_rms_mae_db":2.7955111701363133,"rms_envelope_nrmse":0.09076251035191761,"envelope_correlation":0.9917066851779488,"rms_gain_error_db":-0.1300052884957239,"reference_mean_power":0.0014741623423244587,"output_mean_power":0.0014306875043843625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":4.939959887293225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035657053350397007,"output_power":0.00036329264267223897,"reference_active":true,"delta_db":0.08111151882901242},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.172203979326598e-06,"output_power":1.0223977581656559e-05,"reference_active":true,"delta_db":0.9728069038294817},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.984441222185338e-07,"output_power":7.667298083302419e-07,"reference_active":true,"delta_db":0.40510680869096394}],"active_band_count":3,"active_band_mae_db":0.4863417437831527}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":1.957972513611333,"rms_envelope_nrmse":0.1702345779979827,"envelope_correlation":0.9764269874715626,"rms_gain_error_db":-0.010495747367366519,"reference_mean_power":0.0006933001790048939,"output_mean_power":0.0006916266791153068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":3.8791214865659925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027195601224651783,"output_power":0.0002669923764337651,"reference_active":true,"delta_db":-0.0799980347393017},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.187110727790427e-05,"output_power":2.1819542515403622e-05,"reference_active":true,"delta_db":-0.010251301968967043},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.698791354460839e-06,"output_power":7.1045006271337755e-06,"reference_active":true,"delta_db":0.25537105786202696}],"active_band_count":3,"active_band_mae_db":0.11520679819009856}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":1.392292857049394,"rms_envelope_nrmse":0.10805477144361009,"envelope_correlation":0.9898250934784393,"rms_gain_error_db":-0.010495511292606483,"reference_mean_power":0.000697963583829786,"output_mean_power":0.0006962788651819512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.5511880459540643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002583969788516632,"output_power":0.00025296969650519344,"reference_active":true,"delta_db":-0.09218931912547683},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105856462043575e-06,"output_power":8.28412509375142e-06,"reference_active":true,"delta_db":0.09447739282591892},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388944870715121e-07,"output_power":5.941804382512243e-07,"reference_active":true,"delta_db":-0.3151079092633355}],"active_band_count":3,"active_band_mae_db":0.1672582070715771}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":1.9062291624034815,"rms_envelope_nrmse":0.19783711441925614,"envelope_correlation":0.9729687890408312,"rms_gain_error_db":0.16545904817620072,"reference_mean_power":8.288074215773489e-05,"output_mean_power":8.60992832571981e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":4.287875385442727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2411523150580798e-05,"output_power":2.3043436644912758e-05,"reference_active":true,"delta_db":0.12075875836328234},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.113877152498062e-06,"output_power":3.45395473656242e-06,"reference_active":true,"delta_db":0.4501516697125336},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.881539635700257e-07,"output_power":1.3468444246647413e-06,"reference_active":true,"delta_db":1.3449281585802964}],"active_band_count":3,"active_band_mae_db":0.6386128622187041}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":1.1532323768107275,"rms_envelope_nrmse":0.10696990520617233,"envelope_correlation":0.9918585497448348,"rms_gain_error_db":0.16546017527504348,"reference_mean_power":8.343814618382183e-05,"output_mean_power":8.667835567249577e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.9683446107037285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0633950863405697e-05,"output_power":2.1148405552612236e-05,"reference_active":true,"delta_db":0.10695238064354334},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4408129681959182e-06,"output_power":1.6167189321808002e-06,"reference_active":true,"delta_db":0.5002691529744661},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9896369199766395e-08,"output_power":6.978180111313103e-08,"reference_active":true,"delta_db":1.4567323013270927}],"active_band_count":3,"active_band_mae_db":0.6879846116483673}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":1.9828929296052522,"rms_envelope_nrmse":0.27943147675170765,"envelope_correlation":0.9471272837646147,"rms_gain_error_db":-0.6415997733624695,"reference_mean_power":1.5917473583289074e-05,"output_mean_power":1.3731379194833715e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":8.500689890351622e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.331755201814065e-06,"output_power":3.4997386074425418e-06,"reference_active":true,"delta_db":-0.9262829712390999},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3672340908387865e-06,"output_power":1.0867809106019604e-06,"reference_active":true,"delta_db":-0.9970087722366716},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7907435630616127e-07,"output_power":2.2920356266618166e-07,"reference_active":true,"delta_db":-0.8549856781938103}],"active_band_count":3,"active_band_mae_db":0.9260924738898607}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":1.2915628408126663,"rms_envelope_nrmse":0.16258392182923886,"envelope_correlation":0.9836048436112974,"rms_gain_error_db":-0.6415988661340043,"reference_mean_power":1.6024474415231717e-05,"output_mean_power":1.38236874971033e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":7.9315957267754e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.299772492975258e-06,"output_power":3.429192813557506e-06,"reference_active":true,"delta_db":-0.9825357213464847},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.410418312822683e-07,"output_power":5.910427558235828e-07,"reference_active":true,"delta_db":-0.9822382555809218},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.910690257092516e-08,"output_power":1.599279973355076e-08,"reference_active":true,"delta_db":-0.7726579015774295}],"active_band_count":3,"active_band_mae_db":0.9124772928349453}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.6782140731811523,"predicted_mos":2.6782140731811523,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":108120,"input_sample_rate":24000,"input_seconds":4.505,"input_float32_sha256":"2d8974eed156d0d8c381d6787cccbf5b04b52d4089fd2aa4fac1957dcc20abca","internal_samples":72080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.556398792999971},"started_at":"2026-10-13T05:10:48.748882+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031442899999092333,"native_input_samples":107880,"native_input_sha256":"9d61802c48b0beee2bb122db3f386590e58728908212024524d2a3d09270e30c","output_sha256":"eeac7ea3707c8cf0b4d2e94d259d8409052bfa40cdeb952c4eb8bd26ce8762d4","output_samples":108120,"length_delta_samples":240,"analysis_s":0.6648994159986614,"synthesis_s":0.1133920490010496,"total_s":0.778291464999711,"rtf":0.17314604338147074,"output_peak":0.3208404779434204,"output_rms":0.037194351531058455,"finished_at":"2026-10-13T05:10:49.540000+00:00","analysis_stages":{"external_f0_s":0.07368464399951336},"evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"eeac7ea3707c8cf0b4d2e94d259d8409052bfa40cdeb952c4eb8bd26ce8762d4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":71920,"output_samples":72080,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.58635510204245,"output_peak":0.3186594002661605},"common_prefix":{"samples":71920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9977802441731409,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.907953281303906,"spectral_convergence":0.25463286204490115,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.701706727358253,"spectral_convergence":0.20692945328940265,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.1016462782022876,"spectral_convergence":0.2261622998334539,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.00019578850419514697,"delta_db":-0.25049519276898474,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.0006354836051245358,"delta_db":-0.12352788379236132,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.0004560776566991292,"delta_db":-0.3004229534376971,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":3.0677714029514935e-05,"delta_db":-0.39247605674652325,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":5.345237027907142e-05,"delta_db":-0.5736969646326529,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":1.632877529102695e-05,"delta_db":-0.45418299911689747,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3491336750825194,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2282106474090799,"pitch":null,"output_rms_dbfs":-28.57670417575782,"output_peak":0.3186594002661605,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.903768762288149},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71920,"output_samples":71920,"intelligibility":{"estoi":{"value":0.9365678974676351,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.038245057973635734,"rms_dbfs":-28.34849352834874,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.396015589762875,"mean":-1.857829703902774e-05,"zero_energy":false},"output":{"rms":0.037253303553086745,"rms_dbfs":-28.57670417575782,"peak":0.3186594002661605,"peak_dbfs":-9.93346531123254,"crest_factor_db":18.643238864525276,"mean":4.195725509912968e-05,"zero_energy":false},"rms_difference_db":-0.22821064740907815},"full_scale":{"reference":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":2.3359867315769557e-07,"output_rms_dbfs":-66.31529628348713,"output_zero_energy":false,"output_peak":0.002716764659575692,"output_energy_fraction":2.4340217225850928e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":166,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.8947368421052632,"absolute_difference_cents_median":0.9543290601330265,"absolute_difference_cents_p95":3.5561420279920166,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":0.00010701965100918563,"log_rms_mae_db":1.087481072730948,"rms_envelope_nrmse":0.07800632003858253,"envelope_correlation":0.9942925064724033,"rms_gain_error_db":-0.2282105840001578,"reference_mean_power":0.0014643132072841782,"output_mean_power":0.0013893540170454891},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":5.509210445243936e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038836857080924806,"output_power":0.00036169364715821814,"reference_active":true,"delta_db":-0.3090319604318959},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.320093144841145e-05,"output_power":2.1683938965645696e-05,"reference_active":true,"delta_db":-0.2936724468285358},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.17358638897755e-06,"output_power":8.118994235879606e-06,"reference_active":true,"delta_db":-0.5303692205350703}],"active_band_count":3,"active_band_mae_db":0.37769120926516736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":0.00010161161302678278,"log_rms_mae_db":0.6495669648806153,"rms_envelope_nrmse":0.04944006429987504,"envelope_correlation":0.9980034031623267,"rms_gain_error_db":-0.2282103884942084,"reference_mean_power":0.0014741623423244587,"output_mean_power":0.0013986990311211673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":4.939959887293225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035657053350397007,"output_power":0.00033095761041202653,"reference_active":true,"delta_db":-0.32373078631012525},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.172203979326598e-06,"output_power":8.584718979176925e-06,"reference_active":true,"delta_db":0.21386884952129448},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.984441222185338e-07,"output_power":4.938167263695117e-07,"reference_active":true,"delta_db":-1.5056587078124684}],"active_band_count":3,"active_band_mae_db":0.6810861145479628}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":0.8987732200682786,"rms_envelope_nrmse":0.10259208420164294,"envelope_correlation":0.9918397055794974,"rms_gain_error_db":-0.2580318871205017,"reference_mean_power":0.0006933001790048939,"output_mean_power":0.000653308221580174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":3.8791214865659925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027195601224651783,"output_power":0.0002465657481714807,"reference_active":true,"delta_db":-0.4256591821722671},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.187110727790427e-05,"output_power":1.8030618857116785e-05,"reference_active":true,"delta_db":-0.838601377272533},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.698791354460839e-06,"output_power":7.199696655260653e-06,"reference_active":true,"delta_db":0.3131774754942069}],"active_band_count":3,"active_band_mae_db":0.5258126783130024}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":0.5463937328523865,"rms_envelope_nrmse":0.06279866628674344,"envelope_correlation":0.9972124556443855,"rms_gain_error_db":-0.2580317947738563,"reference_mean_power":0.000697963583829786,"output_mean_power":0.0006577026390357155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.5511880459540643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002583969788516632,"output_power":0.0002327227757440594,"reference_active":true,"delta_db":-0.4544854341938412},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105856462043575e-06,"output_power":7.3283637848957135e-06,"reference_active":true,"delta_db":-0.4379188915213205},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388944870715121e-07,"output_power":4.832005726119685e-07,"reference_active":true,"delta_db":-1.2130170041411035}],"active_band_count":3,"active_band_mae_db":0.7018071099520884}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":0.9786893831075041,"rms_envelope_nrmse":0.1424744925080177,"envelope_correlation":0.9870836756138184,"rms_gain_error_db":-0.5091766107908142,"reference_mean_power":8.288074215773489e-05,"output_mean_power":7.371162241217426e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":4.287875385442727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2411523150580798e-05,"output_power":1.9376011877031035e-05,"reference_active":true,"delta_db":-0.6320698140980208},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.113877152498062e-06,"output_power":2.345798342118021e-06,"reference_active":true,"delta_db":-1.230107999354228},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.881539635700257e-07,"output_power":7.899170205950473e-07,"reference_active":true,"delta_db":-0.9724314516690252}],"active_band_count":3,"active_band_mae_db":0.9448697550404247}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":0.7387404171301034,"rms_envelope_nrmse":0.09633543585251292,"envelope_correlation":0.9951966703569712,"rms_gain_error_db":-0.5091764413428173,"reference_mean_power":8.343814618382183e-05,"output_mean_power":7.420736356989013e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.9683446107037285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0633950863405697e-05,"output_power":1.7838253187047514e-05,"reference_active":true,"delta_db":-0.6323006821487449},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4408129681959182e-06,"output_power":1.1167094329979913e-06,"reference_active":true,"delta_db":-1.106674240456936},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9896369199766395e-08,"output_power":3.682613225945484e-08,"reference_active":true,"delta_db":-1.319128359679821}],"active_band_count":3,"active_band_mae_db":1.0193677607618339}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":1.205956834527408,"rms_envelope_nrmse":0.14207189427452285,"envelope_correlation":0.9868915744309867,"rms_gain_error_db":-0.32881071172040194,"reference_mean_power":1.5917473583289074e-05,"output_mean_power":1.4756829742126895e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":8.500689890351622e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.331755201814065e-06,"output_power":4.043385298249604e-06,"reference_active":true,"delta_db":-0.2991877792954209},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3672340908387865e-06,"output_power":1.2678661938425221e-06,"reference_active":true,"delta_db":-0.3276945657185111},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7907435630616127e-07,"output_power":2.592573187265119e-07,"reference_active":true,"delta_db":-0.31988906513302656}],"active_band_count":3,"active_band_mae_db":0.3155904700489862}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":0.9556484986438971,"rms_envelope_nrmse":0.08204699831318854,"envelope_correlation":0.9960458657574921,"rms_gain_error_db":-0.3288077554026131,"reference_mean_power":1.6024474415231717e-05,"output_mean_power":1.4856038578288086e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":7.9315957267754e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.299772492975258e-06,"output_power":4.005446367791882e-06,"reference_active":true,"delta_db":-0.3079455606788538},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.410418312822683e-07,"output_power":6.935284611629143e-07,"reference_active":true,"delta_db":-0.2877843584387163},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.910690257092516e-08,"output_power":1.797812403648619e-08,"reference_active":true,"delta_db":-0.2644591664554829}],"active_band_count":3,"active_band_mae_db":0.28672969519101765}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-23283-0016","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2549521923065186,"predicted_mos":3.2549521923065186,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":216574,"input_sample_rate":48000,"input_seconds":4.511958333333333,"input_float32_sha256":"05a4603859d11e8f4e62206394d539d0b2653a91ea2d236ec479e5bdfa7e31db","internal_samples":72192,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9390147759986576},"started_at":"2026-10-13T05:15:56.893405+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006924574998265598,"native_input_samples":215760,"native_input_sha256":"76769e13d66af0f967fe7087e8161e519665131b4d4bc0a6475674dd2298b905","output_sha256":"073b7683e3f40951cde646dc875297f484bb65082b92681a312a00e7c48b7167","output_samples":216574,"length_delta_samples":814,"analysis_s":2.785122216999298,"synthesis_s":0.9591683920007199,"total_s":3.744290609000018,"rtf":0.8329901243604044,"output_peak":0.3389447252083677,"output_rms":0.03658561011302284,"finished_at":"2026-10-13T05:16:00.652603+00:00","evaluation_input":{"reference":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"073b7683e3f40951cde646dc875297f484bb65082b92681a312a00e7c48b7167","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":71920,"output_samples":72192,"length_delta_samples":272,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.72990279382629,"output_peak":0.3279729327615412},"common_prefix":{"samples":71920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9962322695035462,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":71920,"output_samples":71920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.221150099523245,"spectral_convergence":0.4271780852619548,"spectral_convergence_status":"ok","amplitude_floor":7.580332405627278e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.593628477465989,"spectral_convergence":0.2099824705475574,"spectral_convergence_status":"ok","amplitude_floor":6.818562526348602e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.551343045094567,"spectral_convergence":0.2062328157455061,"spectral_convergence_status":"ok","amplitude_floor":5.715517707749406e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020741335082478346,"output_power":0.00018777791982746138,"delta_db":-0.4319218400009436,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006538184008046484,"output_power":0.00058662543281366,"delta_db":-0.47096251383075327,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004887436095925812,"output_power":0.00045347658241423665,"delta_db":-0.3252622701001418,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":3.3579219734638266e-05,"output_power":3.305916438507932e-05,"delta_db":-0.06778728415973599,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.100095928785279e-05,"output_power":6.281461028145821e-05,"delta_db":0.12724004953772944,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.8128919162253983e-05,"output_power":2.101948060253528e-05,"delta_db":0.6425006785025813,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3442791060219809,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3650160483173526,"pitch":null,"output_rms_dbfs":-28.71350957666609,"output_peak":0.3279729327615412,"reference_rms_dbfs":-28.34849352834874,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.788707207361267},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":71920,"output_samples":71920,"intelligibility":{"estoi":{"value":0.9124113593176798,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.038245057973635734,"rms_dbfs":-28.34849352834874,"peak":0.317962646484375,"peak_dbfs":-9.952477938585863,"crest_factor_db":18.396015589762875,"mean":-1.857829703902774e-05,"zero_energy":false},"output":{"rms":0.0366711492910221,"rms_dbfs":-28.713509576666087,"peak":0.3279729327615412,"peak_dbfs":-9.683239932922481,"crest_factor_db":19.03026964374361,"mean":-8.98092581191684e-05,"zero_energy":false},"rms_difference_db":-0.36501604831734724},"full_scale":{"reference":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":71920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":449,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1453205702127888e-06,"threshold_dbfs":-59.410729391853494,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":65,"quiet_samples":10400,"reference_mean_power":2.469652547286107e-07,"output_mean_power":2.2714347900829494e-07,"output_rms_dbfs":-66.43699726696,"output_zero_energy":false,"output_peak":0.005023827403336189,"output_energy_fraction":2.4425017949398655e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":442,"reference_estimated_voiced_frames":171,"output_estimated_voiced_frames":173,"both_estimated_voiced_frames":157,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9181286549707602,"absolute_difference_cents_median":3.256878383046031,"absolute_difference_cents_p95":16.75473583232531,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":71920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":0.00010701965100918563,"log_rms_mae_db":1.9528816188128615,"rms_envelope_nrmse":0.10765104565384813,"envelope_correlation":0.9894869702077161,"rms_gain_error_db":-0.3650159672638853,"reference_mean_power":0.0014643132072841782,"output_mean_power":0.0013462706652010955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":5.509210445243936e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00038836857080924806,"output_power":0.00036541799738881877,"reference_active":true,"delta_db":-0.26454143654919865},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.320093144841145e-05,"output_power":2.388773397241732e-05,"reference_active":true,"delta_db":0.12669532981042764},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.17358638897755e-06,"output_power":8.909696655529805e-06,"reference_active":true,"delta_db":-0.12676237071544397}],"active_band_count":3,"active_band_mae_db":0.17266637902502344}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":0.00010161161302678278,"log_rms_mae_db":1.2793963267464892,"rms_envelope_nrmse":0.07279807667154867,"envelope_correlation":0.9958575768790443,"rms_gain_error_db":-0.3650153680859839,"reference_mean_power":0.0014741623423244587,"output_mean_power":0.0013553260198385514},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":4.939959887293225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035657053350397007,"output_power":0.0003335906752288933,"reference_active":true,"delta_db":-0.2893154840885814},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.172203979326598e-06,"output_power":8.496108508218713e-06,"reference_active":true,"delta_db":0.16880852228188695},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.984441222185338e-07,"output_power":6.879931316523182e-07,"reference_active":true,"delta_db":-0.06547564256839111}],"active_band_count":3,"active_band_mae_db":0.17453321631295315}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":9.15514590150253e-05,"log_rms_mae_db":1.5889320262887525,"rms_envelope_nrmse":0.1110250802471486,"envelope_correlation":0.9903731479966974,"rms_gain_error_db":-0.2701146752578988,"reference_mean_power":0.0006933001790048939,"output_mean_power":0.0006514931365561541},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":3.8791214865659925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027195601224651783,"output_power":0.0002537122371080585,"reference_active":true,"delta_db":-0.30157249668903796},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.187110727790427e-05,"output_power":2.1014233671099454e-05,"reference_active":true,"delta_db":-0.17357213652121975},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.698791354460839e-06,"output_power":5.9919951101250905e-06,"reference_active":true,"delta_db":-0.4842500086495789}],"active_band_count":3,"active_band_mae_db":0.31979821395327884}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":8.797925167476037e-05,"log_rms_mae_db":1.2128419604042855,"rms_envelope_nrmse":0.07412359577002743,"envelope_correlation":0.995875292707068,"rms_gain_error_db":-0.2701146961286221,"reference_mean_power":0.000697963583829786,"output_mean_power":0.0006558753279116181},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.5511880459540643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002583969788516632,"output_power":0.000239918904705665,"reference_active":true,"delta_db":-0.32222961545757056},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.105856462043575e-06,"output_power":7.862904891303581e-06,"reference_active":true,"delta_db":-0.1321588648925923},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.388944870715121e-07,"output_power":5.188076971983971e-07,"reference_active":true,"delta_db":-0.9042272989088331}],"active_band_count":3,"active_band_mae_db":0.4528719264196653}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":4.592612274041099e-05,"log_rms_mae_db":1.5422127122649738,"rms_envelope_nrmse":0.11608844916161848,"envelope_correlation":0.9905259921177982,"rms_gain_error_db":0.029806392031741902,"reference_mean_power":8.288074215773489e-05,"output_mean_power":8.345152367454704e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":4.287875385442727e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2411523150580798e-05,"output_power":2.290383706686445e-05,"reference_active":true,"delta_db":0.09436872122897266},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.113877152498062e-06,"output_power":3.0853221154414158e-06,"reference_active":true,"delta_db":-0.04000962768745584},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.881539635700257e-07,"output_power":1.1347748531651253e-06,"reference_active":true,"delta_db":0.6008508620319638}],"active_band_count":3,"active_band_mae_db":0.24507640364946412}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":4.299733639359481e-05,"log_rms_mae_db":0.9296469877293017,"rms_envelope_nrmse":0.06787633029916676,"envelope_correlation":0.9966242977267861,"rms_gain_error_db":0.029808186580199493,"reference_mean_power":8.343814618382183e-05,"output_mean_power":8.401280113499311e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":3.9683446107037285e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0633950863405697e-05,"output_power":2.115256483478863e-05,"reference_active":true,"delta_db":0.10780642880705651},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4408129681959182e-06,"output_power":1.421194877433998e-06,"reference_active":true,"delta_db":-0.05953975298755364},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.9896369199766395e-08,"output_power":6.240516166045078e-08,"reference_active":true,"delta_db":0.971515680223866}],"active_band_count":3,"active_band_mae_db":0.37962062067282537}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":898,"rms_floor":2.431018241105215e-05,"log_rms_mae_db":1.6723202209406938,"rms_envelope_nrmse":0.22447757267009613,"envelope_correlation":0.9740774899224515,"rms_gain_error_db":0.7317803982311124,"reference_mean_power":1.5917473583289074e-05,"output_mean_power":1.8838754110340337e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":898,"frequency_resolution_hz":0.22271714922048996,"power_floor":8.500689890351622e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.331755201814065e-06,"output_power":5.114686992611259e-06,"reference_active":true,"delta_db":0.7215515541891506},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3672340908387865e-06,"output_power":1.8322786618630203e-06,"reference_active":true,"delta_db":1.271486453575175},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7907435630616127e-07,"output_power":6.545073094520127e-07,"reference_active":true,"delta_db":3.7019456935740296}],"active_band_count":3,"active_band_mae_db":1.8983279004461184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":71920,"uncovered_tail_samples":0,"metrics":{"frame_count":892,"rms_floor":1.7655049749551602e-05,"log_rms_mae_db":1.0840234208379371,"rms_envelope_nrmse":0.1361808752221617,"envelope_correlation":0.9930924988713433,"rms_gain_error_db":0.7317855570028048,"reference_mean_power":1.6024474415231717e-05,"output_mean_power":1.8965414974079983e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":892,"frequency_resolution_hz":0.2242152466367713,"power_floor":7.9315957267754e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.299772492975258e-06,"output_power":5.394983429763885e-06,"reference_active":true,"delta_db":0.9854463806798737},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.410418312822683e-07,"output_power":1.038311988510763e-06,"reference_active":true,"delta_db":1.4648514418926402},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.910690257092516e-08,"output_power":5.3333643680666784e-08,"reference_active":true,"delta_db":4.458109660182263}],"active_band_count":3,"active_band_mae_db":2.3028024942515923}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"hnm","status":"measured","mos":{"status":"measured","value":2.3610384464263916,"predicted_mos":2.3610384464263916,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87760,"input_sample_rate":16000,"input_seconds":5.485,"input_float32_sha256":"7c9796819fcf3e7e17114e9cea9f829febac762197d7d4338de7b6b6f07ec710","internal_samples":87760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.061070142999597},"started_at":"2026-10-13T04:39:25.139552+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.546200019220123e-05,"native_input_samples":87760,"native_input_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","output_sha256":"f82fb9eb73ddefa9fb13f5ef3360344764295fafb3c5ea509080e49975665149","output_samples":87760,"length_delta_samples":0,"analysis_s":0.975142723999852,"synthesis_s":0.48461331200087443,"total_s":1.4597560360007265,"rtf":0.26613601385610325,"output_peak":0.5017533371669766,"output_rms":0.06467624537967762,"finished_at":"2026-10-13T04:39:26.603591+00:00","evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"f82fb9eb73ddefa9fb13f5ef3360344764295fafb3c5ea509080e49975665149","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.178773667873693,"spectral_convergence":0.16319704375264268,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.282205346551909,"spectral_convergence":0.16243354465530374,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.037197157024011,"spectral_convergence":0.16880866425741284,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":9.777232679443595e-05,"delta_db":0.3532716895208818,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0019025336232114604,"delta_db":-0.02936654323282086,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0015554555404146274,"delta_db":0.010373424712076893,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.00046809338940158025,"delta_db":-0.24795307923443666,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00014974611273946858,"delta_db":-0.23729763561713546,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":9.415723850698152e-06,"delta_db":-0.045441682576402055,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15395067581562563,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.038832876930999134,"pitch":null,"output_rms_dbfs":-23.785103998632856,"output_peak":0.5017533371669766,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.499392057149872},"common_prefix":{"samples":87760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.178773667873693,"spectral_convergence":0.16319704375264268,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.282205346551909,"spectral_convergence":0.16243354465530374,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.037197157024011,"spectral_convergence":0.16880866425741284,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":9.777232679443595e-05,"delta_db":0.3532716895208818,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0019025336232114604,"delta_db":-0.02936654323282086,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0015554555404146274,"delta_db":0.010373424712076893,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.00046809338940158025,"delta_db":-0.24795307923443666,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00014974611273946858,"delta_db":-0.23729763561713546,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":9.415723850698152e-06,"delta_db":-0.045441682576402055,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15395067581562563,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.038832876930999134,"pitch":null,"output_rms_dbfs":-23.785103998632856,"output_peak":0.5017533371669766,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.499392057149872},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87760,"output_samples":87760,"intelligibility":{"estoi":{"value":0.9080200234607657,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06496604728946706,"rms_dbfs":-23.74627112170186,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.371846874973237,"mean":6.920006890240428e-08,"zero_energy":false},"output":{"rms":0.06467624537967762,"rms_dbfs":-23.785103998632856,"peak":0.5017533371669766,"peak_dbfs":-5.990194606669689,"crest_factor_db":17.794909391963166,"mean":-4.944834996366278e-07,"zero_energy":false},"rms_difference_db":-0.03883287693099646},"full_scale":{"reference":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":9.682365921787958e-07,"output_rms_dbfs":-60.140185082670854,"output_zero_energy":false,"output_peak":0.016062580553140166,"output_energy_fraction":6.330040938303078e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":281,"both_estimated_voiced_frames":277,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9551724137931035,"absolute_difference_cents_median":1.38899749842573,"absolute_difference_cents_p95":11.655407543128494,"difference_over_600_cents_frames":3,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002415922668587691,"log_rms_mae_db":0.8403889192508146,"rms_envelope_nrmse":0.04981563686733226,"envelope_correlation":0.9977039442091215,"rms_gain_error_db":-0.03883297065970448,"reference_mean_power":0.004224437839878649,"output_mean_power":0.004186832889000928},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.932071872304039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008459874532766977,"output_power":0.0008395128294794051,"reference_active":true,"delta_db":-0.033365846891970724},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.267978041368796e-05,"output_power":9.007370025663693e-05,"reference_active":true,"delta_db":-0.12386992077222025},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0604587197812338e-05,"output_power":1.8806193835663e-05,"reference_active":true,"delta_db":-0.39663010009748195}],"active_band_count":3,"active_band_mae_db":0.18462195592055763}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00018058591873130477,"log_rms_mae_db":0.45446932185406286,"rms_envelope_nrmse":0.0252732325024724,"envelope_correlation":0.9993955193359262,"rms_gain_error_db":-0.03883342206954433,"reference_mean_power":0.004247690453399677,"output_mean_power":0.0042098780756531515},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8509413564750568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008019972220007684,"output_power":0.0007959640393924202,"reference_active":true,"delta_db":-0.03279416627359516},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130657727931199e-05,"output_power":4.008224865269467e-05,"reference_active":true,"delta_db":-0.1306713233934908},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0523532353898244e-06,"output_power":9.228526917710733e-07,"reference_active":true,"delta_db":-0.5702915730294709}],"active_band_count":3,"active_band_mae_db":0.24458568756551893}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002374784601360974,"log_rms_mae_db":0.9132516042255108,"rms_envelope_nrmse":0.06147386040979017,"envelope_correlation":0.9969257289906772,"rms_gain_error_db":-0.057195441871745176,"reference_mean_power":0.0028785349502786395,"output_mean_power":0.002840873938652587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.6369199856429154e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008090138264371756,"output_power":0.0008043297595378289,"reference_active":true,"delta_db":-0.025218064407180274},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010508116554565298,"output_power":0.00010328721610067096,"reference_active":true,"delta_db":-0.07478309227510894},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7476383653056017e-05,"output_power":1.4161532744438271e-05,"reference_active":true,"delta_db":-0.9134130919019137}],"active_band_count":3,"active_band_mae_db":0.3378047495280676}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.0001750973468816212,"log_rms_mae_db":0.680290300596276,"rms_envelope_nrmse":0.033285855338585704,"envelope_correlation":0.9990811249522961,"rms_gain_error_db":-0.05719545255130605,"reference_mean_power":0.002894380018498304,"output_mean_power":0.0028565116925514137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.567165950582092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007701013013376472,"output_power":0.0007661584603303966,"reference_active":true,"delta_db":-0.02229255618071644},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.715572645231493e-05,"output_power":4.569906653949794e-05,"reference_active":true,"delta_db":-0.13627110563298075},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1197000044443699e-06,"output_power":8.629104333144421e-07,"reference_active":true,"delta_db":-1.1313595992513867}],"active_band_count":3,"active_band_mae_db":0.429974420355028}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.00012170766433419248,"log_rms_mae_db":1.2024988187723942,"rms_envelope_nrmse":0.09372008878145997,"envelope_correlation":0.9943280012932709,"rms_gain_error_db":-0.26636691342662616,"reference_mean_power":0.0006352125631157456,"output_mean_power":0.0005974236179278576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":2.6162692236926364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010398636594123333,"output_power":9.654725459139324e-05,"reference_active":true,"delta_db":-0.3223647225071351},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0847306945521647e-05,"output_power":9.740245422708376e-06,"reference_active":true,"delta_db":-0.4675202969871505},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.238475994524059e-06,"output_power":3.3468210673516976e-06,"reference_active":true,"delta_db":0.14291811933517448}],"active_band_count":3,"active_band_mae_db":0.31093437960982}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00010363700534457549,"log_rms_mae_db":0.8774792378702125,"rms_envelope_nrmse":0.05753463611321126,"envelope_correlation":0.9982280648980514,"rms_gain_error_db":-0.26636699412370657,"reference_mean_power":0.0006387091058518989,"output_mean_power":0.0006007121393500685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.5308592276752844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945508810428353e-05,"output_power":9.192689096690594e-05,"reference_active":true,"delta_db":-0.3418443424021528},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423063898822289e-06,"output_power":5.220895139529551e-06,"reference_active":true,"delta_db":-0.16499751072054175},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8432082747666131e-07,"output_power":1.7239405600448715e-07,"reference_active":true,"delta_db":-0.29052123783041545}],"active_band_count":3,"active_band_mae_db":0.26578769698437}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":1.619818158902467,"rms_envelope_nrmse":0.23333002676002532,"envelope_correlation":0.9625488258388155,"rms_gain_error_db":0.3548489449729527,"reference_mean_power":1.4688872454843848e-05,"output_mean_power":1.5939450828747192e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":9.038086104701172e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4018145980274595e-06,"output_power":3.75700648275323e-06,"reference_active":true,"delta_db":0.4313130396416937},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2285201909756885e-06,"output_power":1.568769600644413e-06,"reference_active":true,"delta_db":1.0617686646360052},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7360839982104215e-07,"output_power":4.5310569515812137e-07,"reference_active":true,"delta_db":2.190700947686003}],"active_band_count":3,"active_band_mae_db":1.2279275506545673}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":1.4952402990798577,"rms_envelope_nrmse":0.2013985827441779,"envelope_correlation":0.9704992559659537,"rms_gain_error_db":0.35484898447928853,"reference_mean_power":1.476972002493505e-05,"output_mean_power":1.602718172937304e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.51604517212682e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5641156211998283e-06,"output_power":4.132206333059968e-06,"reference_active":true,"delta_db":0.6423021471002545},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790775026435665e-07,"output_power":7.646338452753821e-07,"reference_active":true,"delta_db":1.2071682467449905},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8657206259627095e-08,"output_power":3.437507939522206e-08,"reference_active":true,"delta_db":2.6539709713482127}],"active_band_count":3,"active_band_mae_db":1.5011471217311527}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"world_harvest","status":"measured","mos":{"status":"measured","value":1.936443567276001,"predicted_mos":1.936443567276001,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87760,"input_sample_rate":16000,"input_seconds":5.485,"input_float32_sha256":"cd7b1388f07fcf8dfcf7f8874f36045648373a39d33fba8a77735448985c1484","internal_samples":87760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.096869080000033},"started_at":"2026-10-13T04:42:21.885121+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.098400000249967e-05,"native_input_samples":87760,"native_input_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","output_sha256":"9acf2f212a769d39d042e3cf56602753f3236db1c5a5cbaf9d6b6f3cea26e2cc","output_samples":87760,"length_delta_samples":0,"analysis_s":1.7538500629998452,"synthesis_s":0.10192003199972532,"total_s":1.8557700949995706,"rtf":0.3383354776662845,"output_peak":0.7176128955055253,"output_rms":0.08319008170979275,"finished_at":"2026-10-13T04:42:23.747395+00:00","evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"9acf2f212a769d39d042e3cf56602753f3236db1c5a5cbaf9d6b6f3cea26e2cc","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.681471724344745,"spectral_convergence":0.4597072471127824,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.062936034131753,"spectral_convergence":0.37234707786574395,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.555187702192405,"spectral_convergence":0.4051648911340726,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":9.670209477524608e-05,"delta_db":0.30547100202952543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0033363798011628656,"delta_db":2.4100648820127346,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0025578401986190097,"delta_db":2.1705314867715026,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.0007110881501995579,"delta_db":1.5679562566117227,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00020716022565667164,"delta_db":1.1722105450140357,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":1.1419224468641346e-05,"delta_db":0.7923872893929441,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4031035769720777,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":2.14770213898896,"pitch":null,"output_rms_dbfs":-21.598568982712898,"output_peak":0.7176128955055253,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.433198486889634},"common_prefix":{"samples":87760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.681471724344745,"spectral_convergence":0.4597072471127824,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.062936034131753,"spectral_convergence":0.37234707786574395,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.555187702192405,"spectral_convergence":0.4051648911340726,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":9.670209477524608e-05,"delta_db":0.30547100202952543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0033363798011628656,"delta_db":2.4100648820127346,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0025578401986190097,"delta_db":2.1705314867715026,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.0007110881501995579,"delta_db":1.5679562566117227,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00020716022565667164,"delta_db":1.1722105450140357,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":1.1419224468641346e-05,"delta_db":0.7923872893929441,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4031035769720777,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":2.14770213898896,"pitch":null,"output_rms_dbfs":-21.598568982712898,"output_peak":0.7176128955055253,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.433198486889634},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87760,"output_samples":87760,"intelligibility":{"estoi":{"value":0.9333226967560868,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06496604728946706,"rms_dbfs":-23.74627112170186,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.371846874973237,"mean":6.920006890240428e-08,"zero_energy":false},"output":{"rms":0.08319008170979274,"rms_dbfs":-21.598568982712898,"peak":0.7176128955055253,"peak_dbfs":-2.8821953124183426,"crest_factor_db":18.716373670294555,"mean":-1.1187782149017349e-08,"zero_energy":false},"rms_difference_db":2.1477021389889615},"full_scale":{"reference":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":1.4680240644039693e-06,"output_rms_dbfs":-58.332668252427034,"output_zero_energy":false,"output_peak":0.028201067023639264,"output_energy_fraction":5.801024506762309e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":314,"both_estimated_voiced_frames":282,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9724137931034482,"absolute_difference_cents_median":4.885735280260867,"absolute_difference_cents_p95":1208.8949405376804,"difference_over_600_cents_frames":23,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002415922668587691,"log_rms_mae_db":1.9474048842130975,"rms_envelope_nrmse":0.3221486779449473,"envelope_correlation":0.9830213685601338,"rms_gain_error_db":2.147702460716611,"reference_mean_power":0.004224437839878649,"output_mean_power":0.006926904022156397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.932071872304039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008459874532766977,"output_power":0.0016049889555340517,"reference_active":true,"delta_db":2.781081261003603},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.267978041368796e-05,"output_power":0.00026005884247527587,"reference_active":true,"delta_db":4.480866290221851},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0604587197812338e-05,"output_power":4.932616631499344e-05,"reference_active":true,"delta_db":3.791134448355043}],"active_band_count":3,"active_band_mae_db":3.6843606665268323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00018058591873130477,"log_rms_mae_db":1.6594615442125715,"rms_envelope_nrmse":0.3097020698082092,"envelope_correlation":0.9878660094974665,"rms_gain_error_db":2.1477029719662886,"reference_mean_power":0.004247690453399677,"output_mean_power":0.006965032665998213},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8509413564750568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008019972220007684,"output_power":0.0015508992492862244,"reference_active":true,"delta_db":2.8641072180150444},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130657727931199e-05,"output_power":0.00011916676659011486,"reference_active":true,"delta_db":4.601359453611474},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0523532353898244e-06,"output_power":2.297544713351485e-06,"reference_active":true,"delta_db":3.39102431495222}],"active_band_count":3,"active_band_mae_db":3.61883032885958}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002374784601360974,"log_rms_mae_db":1.9882880831068992,"rms_envelope_nrmse":0.3964370060696239,"envelope_correlation":0.9821670181268473,"rms_gain_error_db":2.6409351611334793,"reference_mean_power":0.0028785349502786395,"output_mean_power":0.005287678275741399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.6369199856429154e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008090138264371756,"output_power":0.0016420425689718415,"reference_active":true,"delta_db":3.0742846777592},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010508116554565298,"output_power":0.0002868826258709259,"reference_active":true,"delta_db":4.361793661108892},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7476383653056017e-05,"output_power":4.4950463629738694e-05,"reference_active":true,"delta_db":4.102826054486171}],"active_band_count":3,"active_band_mae_db":3.8463014644514204}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.0001750973468816212,"log_rms_mae_db":2.2016190868493117,"rms_envelope_nrmse":0.3847588233698001,"envelope_correlation":0.9868343219293084,"rms_gain_error_db":2.6409348854798758,"reference_mean_power":0.002894380018498304,"output_mean_power":0.005316784280374266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.567165950582092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007701013013376472,"output_power":0.0015907720038193594,"reference_active":true,"delta_db":3.1506008190825407},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.715572645231493e-05,"output_power":0.0001303162117429221,"reference_active":true,"delta_db":4.414640069440429},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1197000044443699e-06,"output_power":2.4003078633332363e-06,"reference_active":true,"delta_db":3.3116526793424477}],"active_band_count":3,"active_band_mae_db":3.6256311892884727}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.00012170766433419248,"log_rms_mae_db":1.2409863729367725,"rms_envelope_nrmse":0.21149638607076834,"envelope_correlation":0.9935802204631984,"rms_gain_error_db":1.4637472303802803,"reference_mean_power":0.0006352125631157456,"output_mean_power":0.0008898028694572779},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":2.6162692236926364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010398636594123333,"output_power":0.00013454662181299252,"reference_active":true,"delta_db":1.1189639701684346},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0847306945521647e-05,"output_power":1.687678437599017e-05,"reference_active":true,"delta_db":1.9196777218829806},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.238475994524059e-06,"output_power":4.982614746789412e-06,"reference_active":true,"delta_db":1.871166268856631}],"active_band_count":3,"active_band_mae_db":1.6366026536360154}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00010363700534457549,"log_rms_mae_db":1.0743840706632208,"rms_envelope_nrmse":0.2003013414358937,"envelope_correlation":0.9961904032838009,"rms_gain_error_db":1.463747350666196,"reference_mean_power":0.0006387091058518989,"output_mean_power":0.0008947008354001577},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.5308592276752844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945508810428353e-05,"output_power":0.00012935184932756627,"reference_active":true,"delta_db":1.1414563569622076},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423063898822289e-06,"output_power":8.479884934880496e-06,"reference_active":true,"delta_db":1.9414523760871507},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8432082747666131e-07,"output_power":2.684227212344411e-07,"reference_active":true,"delta_db":1.6324486343123212}],"active_band_count":3,"active_band_mae_db":1.57178578912056}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":1.5017621619929606,"rms_envelope_nrmse":0.18614392715294792,"envelope_correlation":0.9803073008956456,"rms_gain_error_db":0.6658850573039875,"reference_mean_power":1.4688872454843848e-05,"output_mean_power":1.7122886011293846e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":9.038086104701172e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4018145980274595e-06,"output_power":3.733355576907141e-06,"reference_active":true,"delta_db":0.4038871503337265},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2285201909756885e-06,"output_power":1.3569008103808627e-06,"reference_active":true,"delta_db":0.431658032473767},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7360839982104215e-07,"output_power":2.6872236760498793e-07,"reference_active":true,"delta_db":-0.07825608916922594}],"active_band_count":3,"active_band_mae_db":0.3046004239922398}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":1.1383174859318708,"rms_envelope_nrmse":0.13239402834635675,"envelope_correlation":0.9919237605813043,"rms_gain_error_db":0.6658854049567839,"reference_mean_power":1.476972002493505e-05,"output_mean_power":1.7217131772899146e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.51604517212682e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5641156211998283e-06,"output_power":3.8568539842978424e-06,"reference_active":true,"delta_db":0.34281413064085453},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790775026435665e-07,"output_power":6.723567536686188e-07,"reference_active":true,"delta_db":0.6486307884794184},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8657206259627095e-08,"output_power":1.578687164714953e-08,"reference_active":true,"delta_db":-0.7255053476897689}],"active_band_count":3,"active_band_mae_db":0.5723167556033473}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"facodec","status":"measured","mos":{"status":"measured","value":3.2683181762695312,"predicted_mos":3.2683181762695312,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87800,"input_sample_rate":16000,"input_seconds":5.4875,"input_float32_sha256":"5b8547747ee9637e66bbd339eef8f320c2e6e0d4f648e4ed463ce913c91716ae","internal_samples":87800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9498729329989146},"started_at":"2026-10-13T04:53:30.619390+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.250199941045139e-05,"native_input_samples":87760,"native_input_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","output_sha256":"e2c55f31e69334edd02ebfcedfa87f213d1cb8a370dd41cd6249b7f9aba67fc6","output_samples":87800,"length_delta_samples":40,"analysis_s":1.3983392099999037,"synthesis_s":15.07044291800048,"total_s":16.468782128000385,"rtf":3.0025126942571347,"output_peak":0.43177565932273865,"output_rms":0.06129444647024332,"finished_at":"2026-10-13T04:53:47.095236+00:00","evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"e2c55f31e69334edd02ebfcedfa87f213d1cb8a370dd41cd6249b7f9aba67fc6","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":87760,"output_samples":87800,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.251577451422133,"output_peak":0.43177565932273865},"common_prefix":{"samples":87760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9995444191343964,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.92315234524872,"spectral_convergence":0.19167127783718144,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.155119183554116,"spectral_convergence":0.19266341764623965,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.347995990690205,"spectral_convergence":0.21251994895743198,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":6.780068067415531e-05,"delta_db":-1.2365472781463456,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0016541443117756638,"delta_db":-0.6369559923004321,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0014938005342063718,"delta_db":-0.16527649298619573,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.0003824281148103832,"delta_db":-1.1257800319071283,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00015344548348451254,"delta_db":-0.13131210986482986,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":7.102362170824391e-06,"delta_db":-1.2699506654971857,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7609704284503529,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5033274117620128,"pitch":null,"output_rms_dbfs":-24.24959853346387,"output_peak":0.43177565932273865,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.142089173164347},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87760,"output_samples":87760,"intelligibility":{"estoi":{"value":0.8909083433304952,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06496604728946706,"rms_dbfs":-23.74627112170186,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.371846874973237,"mean":6.920006890240428e-08,"zero_energy":false},"output":{"rms":0.061308412857632444,"rms_dbfs":-24.24959853346387,"peak":0.43177565932273865,"peak_dbfs":-7.294836879673427,"crest_factor_db":16.954761653790445,"mean":-6.816449198970625e-07,"zero_energy":false},"rms_difference_db":-0.5033274117620117},"full_scale":{"reference":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":1.1329303412978708e-06,"output_rms_dbfs":-59.45796792095979,"output_zero_energy":false,"output_peak":0.020256685093045235,"output_energy_fraction":8.242855534386213e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":285,"both_estimated_voiced_frames":279,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9620689655172414,"absolute_difference_cents_median":2.5296218749378396,"absolute_difference_cents_p95":12.804636056550839,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002415922668587691,"log_rms_mae_db":1.408145941019439,"rms_envelope_nrmse":0.09472022755885565,"envelope_correlation":0.9943003932670227,"rms_gain_error_db":-0.5033273401901146,"reference_mean_power":0.004224437839878649,"output_mean_power":0.003762150717714326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.932071872304039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008459874532766977,"output_power":0.0006798021869914799,"reference_active":true,"delta_db":-0.9498136470721164},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.267978041368796e-05,"output_power":7.953057154768785e-05,"reference_active":true,"delta_db":-0.6645089261672226},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0604587197812338e-05,"output_power":1.8798071558129192e-05,"reference_active":true,"delta_db":-0.3985061959772101}],"active_band_count":3,"active_band_mae_db":0.670942923072183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00018058591873130477,"log_rms_mae_db":1.2159416952508484,"rms_envelope_nrmse":0.08536239894591278,"envelope_correlation":0.9957637475072231,"rms_gain_error_db":-0.5033283719749383,"reference_mean_power":0.004247690453399677,"output_mean_power":0.0037828578610956333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8509413564750568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008019972220007684,"output_power":0.0006435762859704903,"reference_active":true,"delta_db":-0.9557283079965282},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130657727931199e-05,"output_power":3.505464986025676e-05,"reference_active":true,"delta_db":-0.7127357665599752},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0523532353898244e-06,"output_power":8.875232293234803e-07,"reference_active":true,"delta_db":-0.7398181185523277}],"active_band_count":3,"active_band_mae_db":0.802760731036277}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002374784601360974,"log_rms_mae_db":1.1628030503255438,"rms_envelope_nrmse":0.10464483806462303,"envelope_correlation":0.9919428294081638,"rms_gain_error_db":-0.341035951858168,"reference_mean_power":0.0028785349502786395,"output_mean_power":0.002661141156513774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.6369199856429154e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008090138264371756,"output_power":0.0007022168106137704,"reference_active":true,"delta_db":-0.6148472199698708},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010508116554565298,"output_power":8.916636620845524e-05,"reference_active":true,"delta_db":-0.7132381304899295},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7476383653056017e-05,"output_power":1.6116069866740438e-05,"reference_active":true,"delta_db":-0.3519242857657174}],"active_band_count":3,"active_band_mae_db":0.5600032120751726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.0001750973468816212,"log_rms_mae_db":1.0729105655050712,"rms_envelope_nrmse":0.0928343148306758,"envelope_correlation":0.9936239151941185,"rms_gain_error_db":-0.3410359296117189,"reference_mean_power":0.002894380018498304,"output_mean_power":0.0026757895812681218},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.567165950582092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007701013013376472,"output_power":0.0006683935218836619,"reference_active":true,"delta_db":-0.615156252445508},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.715572645231493e-05,"output_power":3.947453499863293e-05,"reference_active":true,"delta_db":-0.7721741689992698},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1197000044443699e-06,"output_power":8.368533884208881e-07,"reference_active":true,"delta_db":-1.2645230100362512}],"active_band_count":3,"active_band_mae_db":0.8839511438270096}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.00012170766433419248,"log_rms_mae_db":1.5588386581180047,"rms_envelope_nrmse":0.15739445688836873,"envelope_correlation":0.9880535769059717,"rms_gain_error_db":-0.8794917510189169,"reference_mean_power":0.0006352125631157456,"output_mean_power":0.000518764087815829},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":2.6162692236926364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010398636594123333,"output_power":8.033532264212246e-05,"reference_active":true,"delta_db":-1.1206985875041156},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0847306945521647e-05,"output_power":1.055419583819701e-05,"reference_active":true,"delta_db":-0.11896781073122048},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.238475994524059e-06,"output_power":2.724444255540924e-06,"reference_active":true,"delta_db":-0.7506275586285548}],"active_band_count":3,"active_band_mae_db":0.6634313189546303}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00010363700534457549,"log_rms_mae_db":1.4367458363008145,"rms_envelope_nrmse":0.14388454253562885,"envelope_correlation":0.9909527722729279,"rms_gain_error_db":-0.8794918250886196,"reference_mean_power":0.0006387091058518989,"output_mean_power":0.0005216196282400936},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.5308592276752844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945508810428353e-05,"output_power":7.661431481872603e-05,"reference_active":true,"delta_db":-1.133170845260753},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423063898822289e-06,"output_power":5.2834840678444175e-06,"reference_active":true,"delta_db":-0.1132431950417799},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8432082747666131e-07,"output_power":1.7012264441875535e-07,"reference_active":true,"delta_db":-0.3481228658737786}],"active_band_count":3,"active_band_mae_db":0.5315123020587705}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":1.8180588827883242,"rms_envelope_nrmse":0.20367916203901246,"envelope_correlation":0.9731777203712968,"rms_gain_error_db":-0.8535376135821873,"reference_mean_power":1.4688872454843848e-05,"output_mean_power":1.2067983242889017e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":9.038086104701172e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4018145980274595e-06,"output_power":2.5450105622038675e-06,"reference_active":true,"delta_db":-1.2602105141085782},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2285201909756885e-06,"output_power":9.944334870583322e-07,"reference_active":true,"delta_db":-0.9180655805885081},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7360839982104215e-07,"output_power":3.590147733101332e-07,"reference_active":true,"delta_db":1.1798289382331728}],"active_band_count":3,"active_band_mae_db":1.1193683443100861}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":1.6129227619832356,"rms_envelope_nrmse":0.1619049764552214,"envelope_correlation":0.9848929292762778,"rms_gain_error_db":-0.8535400963777087,"reference_mean_power":1.476972002493505e-05,"output_mean_power":1.2134398498831887e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.51604517212682e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5641156211998283e-06,"output_power":2.753384417310932e-06,"reference_active":true,"delta_db":-1.1208493405359223},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790775026435665e-07,"output_power":4.901411523652315e-07,"reference_active":true,"delta_db":-0.7241552526467004},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8657206259627095e-08,"output_power":2.4702105081582186e-08,"reference_active":true,"delta_db":1.2188735206990002}],"active_band_count":3,"active_band_mae_db":1.0212927046272078}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.031790018081665,"predicted_mos":2.031790018081665,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":242176,"input_sample_rate":44100,"input_seconds":5.491519274376417,"input_float32_sha256":"22f23b1e616a4af00144d5a95a145046ee48b62fc6fdb531c040e5576dbdccc9","internal_samples":87865,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.320926807999058},"started_at":"2026-10-13T07:17:04.016793+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0063916369999788,"native_input_samples":241889,"native_input_sha256":"a3b0ad1a6af89d89bcef78c173430998f90ac5349d24fc6630a4787f77e720d5","output_sha256":"3f6fb91dc1638e1639bc8bcc7e9d9e648fa60605016b9e3bbf156d7c41247822","output_samples":242176,"length_delta_samples":287,"analysis_s":0.1359350499988068,"synthesis_s":33.00188314500156,"total_s":33.137818195000364,"rtf":6.041522278398422,"output_peak":0.4598918855190277,"output_rms":0.06252665930335911,"finished_at":"2026-10-13T07:17:37.179730+00:00","evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"3f6fb91dc1638e1639bc8bcc7e9d9e648fa60605016b9e3bbf156d7c41247822","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":87760,"output_samples":87865,"length_delta_samples":105,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.071225808060422,"output_peak":0.45680546208280987},"common_prefix":{"samples":87760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988049849200478,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.238648872208051,"spectral_convergence":0.37807875051647444,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.7752718953723345,"spectral_convergence":0.2735080232840444,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.882029176138888,"spectral_convergence":0.3424977169584162,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":7.030206888564329e-05,"delta_db":-1.079206758834798,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0017704822558197897,"delta_db":-0.34177416542744615,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0015653732346808916,"delta_db":0.037976441110018556,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.00038339832903332046,"delta_db":-1.114776000586461,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00012723698109199646,"delta_db":-0.9447196471342287,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":4.205980428354652e-06,"delta_db":-3.54530633605818,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.177293224858522,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3197617291159857,"pitch":null,"output_rms_dbfs":-24.06603285081784,"output_peak":0.45680546208280987,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.965316647906424},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87760,"output_samples":87760,"intelligibility":{"estoi":{"value":0.7425946411816854,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06496604728946706,"rms_dbfs":-23.74627112170186,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.371846874973237,"mean":6.920006890240428e-08,"zero_energy":false},"output":{"rms":0.06261787963465384,"rms_dbfs":-24.066032850817844,"peak":0.45680546208280987,"peak_dbfs":-6.805374236301878,"crest_factor_db":17.260658614515965,"mean":-5.019531807739682e-05,"zero_energy":false},"rms_difference_db":-0.3197617291159851},"full_scale":{"reference":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":5.257300612018843e-07,"output_rms_dbfs":-62.79237189348634,"output_zero_energy":false,"output_peak":0.007945466947236973,"output_energy_fraction":3.66674512182612e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":285,"both_estimated_voiced_frames":273,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9413793103448276,"absolute_difference_cents_median":8.475218301854955,"absolute_difference_cents_p95":1204.1708409107714,"difference_over_600_cents_frames":20,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002415922668587691,"log_rms_mae_db":1.5556985734407769,"rms_envelope_nrmse":0.08663480660681615,"envelope_correlation":0.9942486866822857,"rms_gain_error_db":-0.31976147899404805,"reference_mean_power":0.004224437839878649,"output_mean_power":0.003924576293952284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.932071872304039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008459874532766977,"output_power":0.0007728440190698924,"reference_active":true,"delta_db":-0.3927207179395767},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.267978041368796e-05,"output_power":9.007587122771369e-05,"reference_active":true,"delta_db":-0.12376524766627943},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0604587197812338e-05,"output_power":2.0887499429462943e-05,"reference_active":true,"delta_db":0.05922532955093283}],"active_band_count":3,"active_band_mae_db":0.191903765052263}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00018058591873130477,"log_rms_mae_db":1.375395307017748,"rms_envelope_nrmse":0.06148133701490495,"envelope_correlation":0.9976905684861009,"rms_gain_error_db":-0.3197609710483118,"reference_mean_power":0.004247690453399677,"output_mean_power":0.003946178838244549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8509413564750568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008019972220007684,"output_power":0.0007337663814171715,"reference_active":true,"delta_db":-0.3861505391804182},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130657727931199e-05,"output_power":4.081294572871746e-05,"reference_active":true,"delta_db":-0.05221268532858063},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0523532353898244e-06,"output_power":1.105120145221382e-06,"reference_active":true,"delta_db":0.2124795512405147}],"active_band_count":3,"active_band_mae_db":0.2169475919165045}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002374784601360974,"log_rms_mae_db":1.7463224028937425,"rms_envelope_nrmse":0.12237190754391034,"envelope_correlation":0.9877436226716341,"rms_gain_error_db":-0.06341368894706197,"reference_mean_power":0.0028785349502786395,"output_mean_power":0.002836809273722502},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.6369199856429154e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008090138264371756,"output_power":0.000729772363826929,"reference_active":true,"delta_db":-0.4476853118791311},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010508116554565298,"output_power":0.00010259731354721274,"reference_active":true,"delta_db":-0.10388892106032083},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7476383653056017e-05,"output_power":1.7692952010427135e-05,"reference_active":true,"delta_db":0.05348729490831883}],"active_band_count":3,"active_band_mae_db":0.20168717594925692}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.0001750973468816212,"log_rms_mae_db":1.6247924017123814,"rms_envelope_nrmse":0.10212722530575724,"envelope_correlation":0.9910584975633367,"rms_gain_error_db":-0.06341374054759127,"reference_mean_power":0.002894380018498304,"output_mean_power":0.0028524246265628807},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.567165950582092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007701013013376472,"output_power":0.0006962977987345376,"reference_active":true,"delta_db":-0.4375283506842329},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.715572645231493e-05,"output_power":4.5893664138180414e-05,"reference_active":true,"delta_db":-0.11781706679489072},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1197000044443699e-06,"output_power":9.756210781682043e-07,"reference_active":true,"delta_db":-0.5982050531277459}],"active_band_count":3,"active_band_mae_db":0.38451682353562316}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.00012170766433419248,"log_rms_mae_db":2.896191918755983,"rms_envelope_nrmse":0.25825561870295427,"envelope_correlation":0.9605428318258663,"rms_gain_error_db":-1.0804024051906662,"reference_mean_power":0.0006352125631157456,"output_mean_power":0.0004953119869426182},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":2.6162692236926364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010398636594123333,"output_power":0.00010509041854558285,"reference_active":true,"delta_db":0.045867207287181265},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0847306945521647e-05,"output_power":7.594212257769953e-06,"reference_active":true,"delta_db":-1.5483919805495803},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.238475994524059e-06,"output_power":2.284034180123182e-06,"reference_active":true,"delta_db":-1.5163808340331932}],"active_band_count":3,"active_band_mae_db":1.0368800072899849}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00010363700534457549,"log_rms_mae_db":2.686205582997971,"rms_envelope_nrmse":0.23750116028790436,"envelope_correlation":0.9675793462475032,"rms_gain_error_db":-1.0804027516380756,"reference_mean_power":0.0006387091058518989,"output_mean_power":0.0004980384038836518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.5308592276752844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945508810428353e-05,"output_power":9.948553935827981e-05,"reference_active":true,"delta_db":0.0013295234713580433},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423063898822289e-06,"output_power":3.60600396262837e-06,"reference_active":true,"delta_db":-1.772185220344968},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8432082747666131e-07,"output_power":1.5388551853867643e-07,"reference_active":true,"delta_db":-0.7837665914805494}],"active_band_count":3,"active_band_mae_db":0.8524271117656251}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":3.7388091090958224,"rms_envelope_nrmse":0.337222540577303,"envelope_correlation":0.9222791996671145,"rms_gain_error_db":-0.6665799028587792,"reference_mean_power":1.4688872454843848e-05,"output_mean_power":1.2598837357826526e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":9.038086104701172e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4018145980274595e-06,"output_power":3.2600993759657924e-06,"reference_active":true,"delta_db":-0.18479801829978534},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2285201909756885e-06,"output_power":1.986987182572854e-06,"reference_active":true,"delta_db":2.0881276703057545},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7360839982104215e-07,"output_power":4.1438669539879145e-07,"reference_active":true,"delta_db":1.8027637702404953}],"active_band_count":3,"active_band_mae_db":1.3585631529486786}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":3.892712813543278,"rms_envelope_nrmse":0.30681026335549133,"envelope_correlation":0.9333588300018786,"rms_gain_error_db":-0.6665790423050312,"reference_mean_power":1.476972002493505e-05,"output_mean_power":1.266818388265111e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.51604517212682e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5641156211998283e-06,"output_power":3.7623809892441045e-06,"reference_active":true,"delta_db":0.23510987054209828},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790775026435665e-07,"output_power":9.738769052641636e-07,"reference_active":true,"delta_db":2.257673742249097},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8657206259627095e-08,"output_power":3.340846427446771e-08,"reference_active":true,"delta_db":2.530098995967531}],"active_band_count":3,"active_band_mae_db":1.6742942029195753}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"golf","status":"measured","mos":{"status":"measured","value":2.374721050262451,"predicted_mos":2.374721050262451,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":131520,"input_sample_rate":24000,"input_seconds":5.48,"input_float32_sha256":"b1dcb31c38eae63966b790a80aa3e4f2db7fa1ef82220eba62f58c54112a1293","internal_samples":87680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7374751810002635},"started_at":"2026-10-13T05:02:38.547870+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00345402200036915,"native_input_samples":131640,"native_input_sha256":"c7a28d78849a6b8bc3a92460fbf68b631a43e50ff04dd41978a6abd3df3721ac","output_sha256":"8fda0b92793ac9703c5d3a9b560563cbdaed71c4fff1f647901358aa50799005","output_samples":131520,"length_delta_samples":-120,"analysis_s":0.5116983210000399,"synthesis_s":0.11429113000031066,"total_s":0.6259894510003505,"rtf":0.11412752069286244,"output_peak":0.4346264898777008,"output_rms":0.05287149921263642,"finished_at":"2026-10-13T05:02:39.184256+00:00","analysis_stages":{"external_f0_s":0.08841322499938542},"evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"8fda0b92793ac9703c5d3a9b560563cbdaed71c4fff1f647901358aa50799005","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":87760,"output_samples":87680,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.52903758048631,"output_peak":0.4349159410650413},"common_prefix":{"samples":87680,"sample_rate":16000,"reference_coverage":0.9990884229717412,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87680,"output_samples":87680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.815262912903679,"spectral_convergence":0.4608239703725593,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.08860567414432,"spectral_convergence":0.3905222854976989,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.381001925623116,"spectral_convergence":0.47488297843928196,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.021482170315269e-05,"output_power":8.157180426353923e-05,"delta_db":-0.43737826687594095,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0019162295616663243,"output_power":0.0012760053734463856,"delta_db":-1.765950324739452,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015540673549060866,"output_power":0.0010588967308212717,"delta_db":-1.6661623015978928,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004960545153586506,"output_power":0.0002849423262246676,"delta_db":-2.407724416800453,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015834763433945367,"output_power":9.461055031583437e-05,"delta_db":-2.236720109159061,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.524012932740633e-06,"output_power":3.5748723413611073e-06,"delta_db":-4.2555943870110715,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.128254967697312,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.7867268842905237,"pitch":null,"output_rms_dbfs":-25.52903758048631,"output_peak":0.4349159410650413,"reference_rms_dbfs":-23.742310696195787,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.428290170890372},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87680,"output_samples":87680,"intelligibility":{"estoi":{"value":0.8483194062729561,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06499567601699677,"rms_dbfs":-23.742310696195787,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.367886449467164,"mean":-5.638512381671989e-08,"zero_energy":false},"output":{"rms":0.05291126210376255,"rms_dbfs":-25.529037580486307,"peak":0.4349159410650413,"peak_dbfs":-7.231893475127205,"crest_factor_db":18.297144105359106,"mean":0.0003292840780180471,"zero_energy":false},"rms_difference_db":-1.7867268842905197},"full_scale":{"reference":{"sample_count":87680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":6.421670193026679e-07,"output_rms_dbfs":-61.923520028845374,"output_zero_energy":false,"output_peak":0.006252136630874184,"output_energy_fraction":6.278594795951503e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":253,"both_estimated_voiced_frames":246,"reference_voiced_output_unvoiced_frames":44,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.8482758620689655,"absolute_difference_cents_median":4.821741851117631,"absolute_difference_cents_p95":20.800343796034312,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1095,"rms_floor":0.0002415922668587691,"log_rms_mae_db":2.544791574698146,"rms_envelope_nrmse":0.2693570542111779,"envelope_correlation":0.956589087916865,"rms_gain_error_db":-1.7867269714363232,"reference_mean_power":0.004228295586927747,"output_mean_power":0.0028021581505843574},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1095,"frequency_resolution_hz":0.182648401826484,"power_floor":1.9308216825387077e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00084476227253018,"output_power":0.0005649384390344821,"reference_active":true,"delta_db":-1.7473338421523346},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.253824683101339e-05,"output_power":6.131481770801545e-05,"reference_active":true,"delta_db":-1.7875582596142725},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0518640643643965e-05,"output_power":1.106029072324766e-05,"reference_active":true,"delta_db":-2.6838204277245516}],"active_band_count":3,"active_band_mae_db":2.072904176497053}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":0.00018058591873130477,"log_rms_mae_db":2.121256055801508,"rms_envelope_nrmse":0.2574990917834618,"envelope_correlation":0.9619706030128582,"rms_gain_error_db":-1.786728376073485,"reference_mean_power":0.0042515907687332485,"output_mean_power":0.00281759532355468},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":1.8497767095975731e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008006648282557012,"output_power":0.0005373954722958996,"reference_active":true,"delta_db":-1.7315674809983705},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.117608439604379e-05,"output_power":2.7971392302313217e-05,"reference_active":true,"delta_db":-1.6793096106411156},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0687725300677789e-06,"output_power":7.494951946506888e-07,"reference_active":true,"delta_db":-1.5411643016289232}],"active_band_count":3,"active_band_mae_db":1.650680464422803}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1095,"rms_floor":0.0002374784601360974,"log_rms_mae_db":1.7459447453072026,"rms_envelope_nrmse":0.28252204549421833,"envelope_correlation":0.9498865721900017,"rms_gain_error_db":-1.6309973945816367,"reference_mean_power":0.0028811637423835717,"output_mean_power":0.0019791021079104136},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1095,"frequency_resolution_hz":0.182648401826484,"power_floor":1.636203782047555e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000807969102407453,"output_power":0.0005680298556281229,"reference_active":true,"delta_db":-1.5302359039930176},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010481470624308955,"output_power":5.951911795863968e-05,"reference_active":true,"delta_db":-2.457657349294845},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.74162535212063e-05,"output_power":9.32435816643427e-06,"reference_active":true,"delta_db":-2.713357906735987}],"active_band_count":3,"active_band_mae_db":2.2337503866746165}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":0.0001750973468816212,"log_rms_mae_db":1.5568405473343339,"rms_envelope_nrmse":0.2696017808952437,"envelope_correlation":0.954813150228393,"rms_gain_error_db":-1.6309973502722472,"reference_mean_power":0.0028970378405243063,"output_mean_power":0.0019900062155549674},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":1.5665049268906835e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007688586670328651,"output_power":0.0005407871546379478,"reference_active":true,"delta_db":-1.528201471309214},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.700537934984004e-05,"output_power":2.75180310509656e-05,"reference_active":true,"delta_db":-2.325302055716651},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1255129496965638e-06,"output_power":7.445566107952467e-07,"reference_active":true,"delta_db":-1.7945277217115754}],"active_band_count":3,"active_band_mae_db":1.8826770829124804}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1095,"rms_floor":0.00012170766433419248,"log_rms_mae_db":2.307589554562204,"rms_envelope_nrmse":0.3385121864929062,"envelope_correlation":0.9502633527329563,"rms_gain_error_db":-2.4635266826204427,"reference_mean_power":0.0006357926619020608,"output_mean_power":0.00036054779415052265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1095,"frequency_resolution_hz":0.182648401826484,"power_floor":2.610370113262895e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001035202957586063,"output_power":4.0134968095475474e-05,"reference_active":true,"delta_db":-4.115025820315919},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.079634211671174e-05,"output_power":5.691775209073655e-06,"reference_active":true,"delta_db":-2.7802889838747564},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.224424078225708e-06,"output_power":2.4704500294154847e-06,"reference_active":true,"delta_db":-1.1567608186464406}],"active_band_count":3,"active_band_mae_db":2.684025207612372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":0.00010363700534457549,"log_rms_mae_db":2.1466177349586286,"rms_envelope_nrmse":0.3202069740022636,"envelope_correlation":0.9604043781902883,"rms_gain_error_db":-2.4635269289417514,"reference_mean_power":0.0006392956121951081,"output_mean_power":0.00036253424036742163},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":2.525182518710934e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.89770603795753e-05,"output_power":3.854855533076271e-05,"reference_active":true,"delta_db":-4.095264443066512},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.391400173977532e-06,"output_power":3.0447957725872818e-06,"reference_active":true,"delta_db":-2.48143400349115},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8686775621302033e-07,"output_power":1.220698469440885e-07,"reference_active":true,"delta_db":-1.8492597086025764}],"active_band_count":3,"active_band_mae_db":2.808652718386746}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1095,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":2.9976987583236476,"rms_envelope_nrmse":0.3851743744171254,"envelope_correlation":0.94813068648823,"rms_gain_error_db":-3.3284803145905064,"reference_mean_power":1.4702286105374141e-05,"output_mean_power":6.831826654609312e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1095,"frequency_resolution_hz":0.182648401826484,"power_floor":9.020864974270513e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.389386182225982e-06,"output_power":1.4272505677188157e-06,"reference_active":true,"delta_db":-3.756208301730352},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2241603178294716e-06,"output_power":4.38156453352118e-07,"reference_active":true,"delta_db":-4.462090848283155},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7235635276653844e-07,"output_power":9.306170275908649e-08,"reference_active":true,"delta_db":-4.6636651513456435}],"active_band_count":3,"active_band_mae_db":4.29398810045305}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87680,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":3.073125693244916,"rms_envelope_nrmse":0.3630063116781551,"envelope_correlation":0.9644845311615028,"rms_gain_error_db":-3.3284824531065054,"reference_mean_power":1.4783281407769319e-05,"output_mean_power":6.869459994747977e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":8.499401967942108e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5507639289789295e-06,"output_power":1.4428267591265395e-06,"reference_active":true,"delta_db":-3.911076110931546},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.753392223957862e-07,"output_power":1.8409301945106328e-07,"reference_active":true,"delta_db":-4.948866610587212},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9158736312143978e-08,"output_power":5.55538742977892e-09,"reference_active":true,"delta_db":-5.376525083303885}],"active_band_count":3,"active_band_mae_db":4.745489268274214}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.5877089500427246,"predicted_mos":2.5877089500427246,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":131760,"input_sample_rate":24000,"input_seconds":5.49,"input_float32_sha256":"b4310dc444570841fa972aa1233b859540e38d9dd2b8e96b51c4bd23f0720e33","internal_samples":87840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1908466669992777},"started_at":"2026-10-13T07:38:49.862498+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003487156001938274,"native_input_samples":131640,"native_input_sha256":"c7a28d78849a6b8bc3a92460fbf68b631a43e50ff04dd41978a6abd3df3721ac","output_sha256":"094757f8d6b5dc644122578630142cfa011399a7258c84bb06b21912217bcf4e","output_samples":131760,"length_delta_samples":120,"analysis_s":2.143129124000552,"synthesis_s":15.440091828000732,"total_s":17.583220952001284,"rtf":3.2056920605289485,"output_peak":0.48665058612823486,"output_rms":0.06331617865759738,"finished_at":"2026-10-13T07:39:07.463364+00:00","analysis_stages":{"external_f0_s":1.429935927997576},"evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"094757f8d6b5dc644122578630142cfa011399a7258c84bb06b21912217bcf4e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":87760,"output_samples":87840,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.963235142170277,"output_peak":0.4792900913843306},"common_prefix":{"samples":87760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990892531876139,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.133097354142622,"spectral_convergence":0.3752389682491333,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.709732921813995,"spectral_convergence":0.2090862709864628,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.88568974165865,"spectral_convergence":0.22567171848880255,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":8.117729485103474e-05,"delta_db":-0.4545420692374508,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0018355123444597945,"delta_db":-0.18511685428066155,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0014565777416634704,"delta_db":-0.27486590607789163,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.0004771069269471716,"delta_db":-0.16512093844117182,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00015985828579245353,"delta_db":0.046498306755295774,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":8.343434828928599e-06,"delta_db":-0.5705300306889158,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2827790175802312,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21300698551694752,"pitch":null,"output_rms_dbfs":-23.959278107218807,"output_peak":0.4792900913843306,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.909506672538423},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87760,"output_samples":87760,"intelligibility":{"estoi":{"value":0.9157065017016461,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06496604728946706,"rms_dbfs":-23.74627112170186,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.371846874973237,"mean":6.920006890240428e-08,"zero_energy":false},"output":{"rms":0.06339223949777176,"rms_dbfs":-23.959278107218807,"peak":0.4792900913843306,"peak_dbfs":-6.388030986264387,"crest_factor_db":17.571247120954418,"mean":1.0140721150432125e-05,"zero_energy":false},"rms_difference_db":-0.21300698551694808},"full_scale":{"reference":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":1.1492472764255522e-06,"output_rms_dbfs":-59.39586516800847,"output_zero_energy":false,"output_peak":0.03829084917004162,"output_energy_fraction":7.820885379828173e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":301,"both_estimated_voiced_frames":283,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9758620689655172,"absolute_difference_cents_median":4.462363988841504,"absolute_difference_cents_p95":32.267343994079106,"difference_over_600_cents_frames":6,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002415922668587691,"log_rms_mae_db":2.299420017405569,"rms_envelope_nrmse":0.12533641928080222,"envelope_correlation":0.9857364654107872,"rms_gain_error_db":-0.21300668084282856,"reference_mean_power":0.004224437839878649,"output_mean_power":0.00402224255062185},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.932071872304039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008459874532766977,"output_power":0.000797109900533079,"reference_active":true,"delta_db":-0.2584571879103382},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.267978041368796e-05,"output_power":9.05339422341331e-05,"reference_active":true,"delta_db":-0.10173564327712846},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0604587197812338e-05,"output_power":2.119254864907864e-05,"reference_active":true,"delta_db":0.12219270702214116}],"active_band_count":3,"active_band_mae_db":0.16079517940320262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00018058591873130477,"log_rms_mae_db":2.111566568864952,"rms_envelope_nrmse":0.0784448295723954,"envelope_correlation":0.9945134091006742,"rms_gain_error_db":-0.2130058511453174,"reference_mean_power":0.004247690453399677,"output_mean_power":0.004044382991341144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8509413564750568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008019972220007684,"output_power":0.0007558514142256571,"reference_active":true,"delta_db":-0.2573643396212356},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130657727931199e-05,"output_power":3.999289621595019e-05,"reference_active":true,"delta_db":-0.1403635409704607},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0523532353898244e-06,"output_power":1.1287852708472162e-06,"reference_active":true,"delta_db":0.30449793207772935}],"active_band_count":3,"active_band_mae_db":0.23407527088980853}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002374784601360974,"log_rms_mae_db":1.3139621826140986,"rms_envelope_nrmse":0.14230977744420853,"envelope_correlation":0.9837600180512159,"rms_gain_error_db":-0.2537484213789535,"reference_mean_power":0.0028785349502786395,"output_mean_power":0.0027151677386079813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.6369199856429154e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008090138264371756,"output_power":0.0007602808762466713,"reference_active":true,"delta_db":-0.2698187739650778},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010508116554565298,"output_power":0.00010314085684051396,"reference_active":true,"delta_db":-0.08094146296635008},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7476383653056017e-05,"output_power":1.743953409741591e-05,"reference_active":true,"delta_db":-0.00916691620778228}],"active_band_count":3,"active_band_mae_db":0.11997571771307004}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.0001750973468816212,"log_rms_mae_db":1.0626462374219763,"rms_envelope_nrmse":0.09276906263889148,"envelope_correlation":0.9931842512448242,"rms_gain_error_db":-0.2537483410657434,"reference_mean_power":0.002894380018498304,"output_mean_power":0.002730113592703581},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.567165950582092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007701013013376472,"output_power":0.000725260657008115,"reference_active":true,"delta_db":-0.26053738212600497},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.715572645231493e-05,"output_power":4.5502537560019345e-05,"reference_active":true,"delta_db":-0.15498822892821099},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1197000044443699e-06,"output_power":1.1789021308736314e-06,"reference_active":true,"delta_db":0.22376072757480464}],"active_band_count":3,"active_band_mae_db":0.21309544620967355}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.00012170766433419248,"log_rms_mae_db":1.4811569115527778,"rms_envelope_nrmse":0.12992435063809493,"envelope_correlation":0.9881516950042817,"rms_gain_error_db":-0.1120241535755369,"reference_mean_power":0.0006352125631157456,"output_mean_power":0.0006190370796127629},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":2.6162692236926364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010398636594123333,"output_power":0.00010018020646826204,"reference_active":true,"delta_db":-0.1619447856314979},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0847306945521647e-05,"output_power":1.0133562727341171e-05,"reference_active":true,"delta_db":-0.2955976938463019},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.238475994524059e-06,"output_power":3.1853653026463532e-06,"reference_active":true,"delta_db":-0.07181437063417423}],"active_band_count":3,"active_band_mae_db":0.17645228337065802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00010363700534457549,"log_rms_mae_db":1.0375604741450768,"rms_envelope_nrmse":0.0806126266174799,"envelope_correlation":0.995423047925054,"rms_gain_error_db":-0.11202419641926926,"reference_mean_power":0.0006387091058518989,"output_mean_power":0.0006224445778728578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.5308592276752844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945508810428353e-05,"output_power":9.576979402280233e-05,"reference_active":true,"delta_db":-0.16398453095079552},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423063898822289e-06,"output_power":5.021030667265432e-06,"reference_active":true,"delta_db":-0.33451847715994243},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8432082747666131e-07,"output_power":1.8460977199703774e-07,"reference_active":true,"delta_db":0.00680274483126951}],"active_band_count":3,"active_band_mae_db":0.16843525098066914}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":1.7115669575168617,"rms_envelope_nrmse":0.22864331437726443,"envelope_correlation":0.9607183010435935,"rms_gain_error_db":-0.16154026676465608,"reference_mean_power":1.4688872454843848e-05,"output_mean_power":1.4152541372447218e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":9.038086104701172e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4018145980274595e-06,"output_power":3.24399728153524e-06,"reference_active":true,"delta_db":-0.20630158878286478},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2285201909756885e-06,"output_power":1.1644217775191075e-06,"reference_active":true,"delta_db":-0.23271979385910863},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7360839982104215e-07,"output_power":2.522568222900895e-07,"reference_active":true,"delta_db":-0.35286505592359296}],"active_band_count":3,"active_band_mae_db":0.26396214618852215}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":1.2069996992930325,"rms_envelope_nrmse":0.13940743210869808,"envelope_correlation":0.9848451237716204,"rms_gain_error_db":-0.16154032688089795,"reference_mean_power":1.476972002493505e-05,"output_mean_power":1.4230436778691426e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.51604517212682e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5641156211998283e-06,"output_power":3.4094168772961257e-06,"reference_active":true,"delta_db":-0.1927167759589704},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790775026435665e-07,"output_power":5.332591856839619e-07,"reference_active":true,"delta_db":-0.35798347625721605},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8657206259627095e-08,"output_power":1.8420405777086815e-08,"reference_active":true,"delta_db":-0.055474198861349217}],"active_band_count":3,"active_band_mae_db":0.20205815035917854}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.573502540588379,"predicted_mos":2.573502540588379,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":131880,"input_sample_rate":24000,"input_seconds":5.495,"input_float32_sha256":"4830d7dbfafd8b130f18f1d5264669d42ce711c4e7c541085101d53ce94aa283","internal_samples":87920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4928292889999284},"started_at":"2026-10-13T05:10:49.542518+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003095498999755364,"native_input_samples":131640,"native_input_sha256":"c7a28d78849a6b8bc3a92460fbf68b631a43e50ff04dd41978a6abd3df3721ac","output_sha256":"3143097847dcc547f1187651e66d3b42af87f2a1e878feb9e5b008c2a72f60f0","output_samples":131880,"length_delta_samples":240,"analysis_s":1.0423567820016615,"synthesis_s":0.19404650099932041,"total_s":1.236403283000982,"rtf":0.22541536608951354,"output_peak":0.47699853777885437,"output_rms":0.06356431706311497,"finished_at":"2026-10-13T05:10:50.794153+00:00","analysis_stages":{"external_f0_s":0.16309498800001165},"evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"3143097847dcc547f1187651e66d3b42af87f2a1e878feb9e5b008c2a72f60f0","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":87760,"output_samples":87920,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.929245649367978,"output_peak":0.4683484791810696},"common_prefix":{"samples":87760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981801637852593,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.064272890290373,"spectral_convergence":0.15915500053136114,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.994102554446176,"spectral_convergence":0.16007668407590125,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.407250643632257,"spectral_convergence":0.17918636965639045,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":9.151409939507337e-05,"delta_db":0.06599228373467532,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.001793428006361598,"delta_db":-0.2858504759505927,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0015389310085796686,"delta_db":-0.036011091881946324,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.00047705109033604803,"delta_db":-0.1656292301850466,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00014461571602271693,"delta_db":-0.388698287103706,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":8.299081024212269e-06,"delta_db":-0.5936787642282984,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25597668884737756,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.17506391142614294,"pitch":null,"output_rms_dbfs":-23.921335033128003,"output_peak":0.4683484791810696,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.155208696122934},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87760,"output_samples":87760,"intelligibility":{"estoi":{"value":0.9398583122863123,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06496604728946706,"rms_dbfs":-23.74627112170186,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.371846874973237,"mean":6.920006890240428e-08,"zero_energy":false},"output":{"rms":0.06366976520860837,"rms_dbfs":-23.921335033128003,"peak":0.4683484791810696,"peak_dbfs":-6.588617713411734,"crest_factor_db":17.332717319716267,"mean":6.726791149822142e-07,"zero_energy":false},"rms_difference_db":-0.1750639114261432},"full_scale":{"reference":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":9.7571979741883e-07,"output_rms_dbfs":-60.10674883057064,"output_zero_energy":false,"output_peak":0.01995400629983066,"output_energy_fraction":6.582232937473759e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":267,"both_estimated_voiced_frames":261,"reference_voiced_output_unvoiced_frames":29,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9,"absolute_difference_cents_median":0.832248197817338,"absolute_difference_cents_p95":5.80052409087309,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002415922668587691,"log_rms_mae_db":0.7772201261037411,"rms_envelope_nrmse":0.062009177565508285,"envelope_correlation":0.9967335685767729,"rms_gain_error_db":-0.17506396163910756,"reference_mean_power":0.004224437839878649,"output_mean_power":0.004057537365936032},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.932071872304039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008459874532766977,"output_power":0.0008212005406596275,"reference_active":true,"delta_db":-0.12914695497002504},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.267978041368796e-05,"output_power":8.046231922035084e-05,"reference_active":true,"delta_db":-0.6139244976277718},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0604587197812338e-05,"output_power":1.7427530661397367e-05,"reference_active":true,"delta_db":-0.7272806259173025}],"active_band_count":3,"active_band_mae_db":0.49011735950503316}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00018058591873130477,"log_rms_mae_db":0.48516505272259963,"rms_envelope_nrmse":0.03965638729678262,"envelope_correlation":0.998835074937672,"rms_gain_error_db":-0.17506389556630703,"reference_mean_power":0.004247690453399677,"output_mean_power":0.004079871369655849},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8509413564750568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008019972220007684,"output_power":0.0007783962969174102,"reference_active":true,"delta_db":-0.12972102769110966},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130657727931199e-05,"output_power":3.6647493672894715e-05,"reference_active":true,"delta_db":-0.519749316863516},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0523532353898244e-06,"output_power":9.659643869108403e-07,"reference_active":true,"delta_db":-0.37200425427412487}],"active_band_count":3,"active_band_mae_db":0.34049153294291684}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002374784601360974,"log_rms_mae_db":0.6243629838182195,"rms_envelope_nrmse":0.0754405124478943,"envelope_correlation":0.9955552715301801,"rms_gain_error_db":-0.16496392740109486,"reference_mean_power":0.0028785349502786395,"output_mean_power":0.002771246222615262},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.6369199856429154e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008090138264371756,"output_power":0.0007833584122608934,"reference_active":true,"delta_db":-0.1399543242763798},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010508116554565298,"output_power":8.772368390336084e-05,"reference_active":true,"delta_db":-0.7840801995056946},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7476383653056017e-05,"output_power":1.373402198800532e-05,"reference_active":true,"delta_db":-1.0465383170095364}],"active_band_count":3,"active_band_mae_db":0.6568576135972036}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.0001750973468816212,"log_rms_mae_db":0.3853880433966263,"rms_envelope_nrmse":0.047149696415315745,"envelope_correlation":0.998371947397171,"rms_gain_error_db":-0.16496394129399047,"reference_mean_power":0.002894380018498304,"output_mean_power":0.0027865007046869875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.567165950582092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007701013013376472,"output_power":0.000743556584512816,"reference_active":true,"delta_db":-0.15233833432113833},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.715572645231493e-05,"output_power":4.01193621161445e-05,"reference_active":true,"delta_db":-0.7018042044403121},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1197000044443699e-06,"output_power":9.573105403890512e-07,"reference_active":true,"delta_db":-0.6804883922917225}],"active_band_count":3,"active_band_mae_db":0.511543643684391}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.00012170766433419248,"log_rms_mae_db":0.7936844997605372,"rms_envelope_nrmse":0.0665982060558656,"envelope_correlation":0.9973384104229374,"rms_gain_error_db":-0.2338503416464121,"reference_mean_power":0.0006352125631157456,"output_mean_power":0.0006019134455860936},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":2.6162692236926364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010398636594123333,"output_power":9.632761618158506e-05,"reference_active":true,"delta_db":-0.3322558804629286},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0847306945521647e-05,"output_power":9.06529121660258e-06,"reference_active":true,"delta_db":-0.7794016944842013},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.238475994524059e-06,"output_power":2.323465867648862e-06,"reference_active":true,"delta_db":-1.4420438520079346}],"active_band_count":3,"active_band_mae_db":0.8512338089850214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00010363700534457549,"log_rms_mae_db":0.6031884892396892,"rms_envelope_nrmse":0.043221817681267566,"envelope_correlation":0.9991661667296237,"rms_gain_error_db":-0.23385025380200863,"reference_mean_power":0.0006387091058518989,"output_mean_power":0.0006052267047757815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.5308592276752844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945508810428353e-05,"output_power":9.18847267341365e-05,"reference_active":true,"delta_db":-0.343836783813165},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423063898822289e-06,"output_power":4.415336084759033e-06,"reference_active":true,"delta_db":-0.8928095504343345},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8432082747666131e-07,"output_power":1.2902398026077086e-07,"reference_active":true,"delta_db":-1.5490397612272404}],"active_band_count":3,"active_band_mae_db":0.9285620318249134}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":1.0660173506290596,"rms_envelope_nrmse":0.13194394458585537,"envelope_correlation":0.9891231566228588,"rms_gain_error_db":-0.540133340023473,"reference_mean_power":1.4688872454843848e-05,"output_mean_power":1.2971049772644773e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":9.038086104701172e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4018145980274595e-06,"output_power":2.929398093652492e-06,"reference_active":true,"delta_db":-0.6493224587927446},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2285201909756885e-06,"output_power":8.732698996695654e-07,"reference_active":true,"delta_db":-1.4823380766792336},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7360839982104215e-07,"output_power":1.4971586770957264e-07,"reference_active":true,"delta_db":-2.6186159446604504}],"active_band_count":3,"active_band_mae_db":1.583425493377476}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":0.8408112857061948,"rms_envelope_nrmse":0.10341836637425794,"envelope_correlation":0.9941939215831597,"rms_gain_error_db":-0.54013300472477,"reference_mean_power":1.476972002493505e-05,"output_mean_power":1.304244345175709e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.51604517212682e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5641156211998283e-06,"output_power":2.955227514493237e-06,"reference_active":true,"delta_db":-0.8136086267208549},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790775026435665e-07,"output_power":4.1720043048433564e-07,"reference_active":true,"delta_db":-1.4239194489651117},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8657206259627095e-08,"output_power":8.601813336896785e-09,"reference_active":true,"delta_db":-3.3625659884271926}],"active_band_count":3,"active_band_mae_db":1.8666980213710531}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41797-0000","method":"magphase","status":"measured","mos":{"status":"measured","value":2.7191991806030273,"predicted_mos":2.7191991806030273,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":263798,"input_sample_rate":48000,"input_seconds":5.495791666666666,"input_float32_sha256":"40fa315806d6cfac19888d95b3b16eb47e84498844fff30423742a2d05cc2a0a","internal_samples":87933,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0780936160008423},"started_at":"2026-10-13T05:16:00.653252+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008537181000065175,"native_input_samples":263280,"native_input_sha256":"bb8ecc976408e83cb601b0933d65b045833b4d91b6673e4e6d4b37381d147be7","output_sha256":"f806562ae228f647047779ba094763f9cc17da58e813b4fd3235f270cf2e32af","output_samples":263798,"length_delta_samples":518,"analysis_s":3.601373539000633,"synthesis_s":1.0656835039990256,"total_s":4.667057042999659,"rtf":0.8508763979944682,"output_peak":0.5034801763615738,"output_rms":0.06279543359957256,"finished_at":"2026-10-13T05:16:05.345204+00:00","evaluation_input":{"reference":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"f806562ae228f647047779ba094763f9cc17da58e813b4fd3235f270cf2e32af","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":87760,"output_samples":87933,"length_delta_samples":173,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.03497945786294,"output_peak":0.5037043123881629},"common_prefix":{"samples":87760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980325929969408,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87760,"output_samples":87760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.424633964973262,"spectral_convergence":0.41734451751590834,"spectral_convergence_status":"ok","amplitude_floor":1.7869460513029852e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.145480608354554,"spectral_convergence":0.3087944341502442,"spectral_convergence_status":"ok","amplitude_floor":1.6383670687778497e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.743698291642721,"spectral_convergence":0.2865220322491659,"spectral_convergence_status":"ok","amplitude_floor":9.631824221799184e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":9.013402842456697e-05,"output_power":8.358955149969342e-05,"delta_db":-0.3273678670667854,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00191544195180521,"output_power":0.0017450158406913483,"delta_db":-0.4046962136112412,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0015517446606579418,"output_power":0.0014639796949415021,"delta_db":-0.25285206511331,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955960266218775,"output_power":0.0004887772359409571,"delta_db":-0.0601684471904557,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000158155871967105,"output_power":0.00016584156867467954,"delta_db":0.20608076127001146,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.514760940569477e-06,"output_power":9.716002648365676e-06,"delta_db":0.09089743449312314,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2236771314574878,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.28015629520747226,"pitch":null,"output_rms_dbfs":-24.02642741690933,"output_peak":0.5037043123881629,"reference_rms_dbfs":-23.74627112170186,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.104604288323512},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87760,"output_samples":87760,"intelligibility":{"estoi":{"value":0.8390840586806335,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06496604728946706,"rms_dbfs":-23.74627112170186,"peak":0.48004150390625,"peak_dbfs":-6.374424246728622,"crest_factor_db":17.371846874973237,"mean":6.920006890240428e-08,"zero_energy":false},"output":{"rms":0.06290405308401475,"rms_dbfs":-24.02642741690933,"peak":0.5037043123881629,"peak_dbfs":-5.956486619599164,"crest_factor_db":18.069940797310164,"mean":-0.0007050458248145868,"zero_energy":false},"rms_difference_db":-0.280156295207469},"full_scale":{"reference":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":548,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.836682340595871e-06,"threshold_dbfs":-52.33833942372626,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":150,"quiet_samples":24000,"reference_mean_power":7.891445808733503e-07,"output_mean_power":9.762681763103694e-07,"output_rms_dbfs":-60.10430867278218,"output_zero_energy":false,"output_peak":0.011878483598487583,"output_energy_fraction":6.747245332449058e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":541,"reference_estimated_voiced_frames":290,"output_estimated_voiced_frames":280,"both_estimated_voiced_frames":268,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9241379310344827,"absolute_difference_cents_median":7.191789169099232,"absolute_difference_cents_p95":29.405054230278335,"difference_over_600_cents_frames":5,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002415922668587691,"log_rms_mae_db":2.3330565450835925,"rms_envelope_nrmse":0.20091117979917433,"envelope_correlation":0.9625752118314638,"rms_gain_error_db":-0.28015596252055147,"reference_mean_power":0.004224437839878649,"output_mean_power":0.003960530187374393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.932071872304039e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008459874532766977,"output_power":0.0007837791146886915,"reference_active":true,"delta_db":-0.33170235445855123},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.267978041368796e-05,"output_power":8.975498740421629e-05,"reference_active":true,"delta_db":-0.1392640583643736},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0604587197812338e-05,"output_power":2.0733360316352844e-05,"reference_active":true,"delta_db":0.027057770917371687}],"active_band_count":3,"active_band_mae_db":0.1660080612467655}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00018058591873130477,"log_rms_mae_db":1.6699874675228952,"rms_envelope_nrmse":0.128888506417527,"envelope_correlation":0.9844680895681028,"rms_gain_error_db":-0.2801553430305418,"reference_mean_power":0.004247690453399677,"output_mean_power":0.003982330739535899},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8509413564750568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008019972220007684,"output_power":0.0007430542193063887,"reference_active":true,"delta_db":-0.33152359364936596},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.130657727931199e-05,"output_power":3.9840837882772704e-05,"reference_active":true,"delta_db":-0.1569074683291245},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0523532353898244e-06,"output_power":1.1056283156717492e-06,"reference_active":true,"delta_db":0.21447612054952175}],"active_band_count":3,"active_band_mae_db":0.23430239417600407}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.0002374784601360974,"log_rms_mae_db":2.0127314576175976,"rms_envelope_nrmse":0.23334689905868222,"envelope_correlation":0.9550190108518581,"rms_gain_error_db":-0.3573302484735433,"reference_mean_power":0.0028785349502786395,"output_mean_power":0.0026511755265306055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":1.6369199856429154e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008090138264371756,"output_power":0.0007281152015670265,"reference_active":true,"delta_db":-0.45755845644995974},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010508116554565298,"output_power":9.67585920501822e-05,"reference_active":true,"delta_db":-0.3583534102761369},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7476383653056017e-05,"output_power":1.580690059093149e-05,"reference_active":true,"delta_db":-0.4360484802239492}],"active_band_count":3,"active_band_mae_db":0.41732011565001526}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.0001750973468816212,"log_rms_mae_db":1.6712807929497437,"rms_envelope_nrmse":0.15490459902908854,"envelope_correlation":0.9799992742815291,"rms_gain_error_db":-0.3573302635968268,"reference_mean_power":0.002894380018498304,"output_mean_power":0.002665769071887741},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.567165950582092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007701013013376472,"output_power":0.0006934758313698788,"reference_active":true,"delta_db":-0.45516527388357153},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.715572645231493e-05,"output_power":4.334591430223134e-05,"reference_active":true,"delta_db":-0.36586271690282396},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1197000044443699e-06,"output_power":1.0199861339528508e-06,"reference_active":true,"delta_db":-0.4050741208094824}],"active_band_count":3,"active_band_mae_db":0.4087007038652926}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":0.00012170766433419248,"log_rms_mae_db":2.1810739278651927,"rms_envelope_nrmse":0.19280952229528672,"envelope_correlation":0.9741407730691615,"rms_gain_error_db":-0.0006738826094113237,"reference_mean_power":0.0006352125631157456,"output_mean_power":0.0006351140066042585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":2.6162692236926364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010398636594123333,"output_power":0.00010249768268980695,"reference_active":true,"delta_db":-0.06262354192178307},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0847306945521647e-05,"output_power":1.0268629617246278e-05,"reference_active":true,"delta_db":-0.23809440110541363},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.238475994524059e-06,"output_power":3.308746820106997e-06,"reference_active":true,"delta_db":0.09322854770587458}],"active_band_count":3,"active_band_mae_db":0.13131549691102376}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":0.00010363700534457549,"log_rms_mae_db":1.6750038233507014,"rms_envelope_nrmse":0.12036247014718825,"envelope_correlation":0.9896778569237906,"rms_gain_error_db":-0.0006738235091806387,"reference_mean_power":0.0006387091058518989,"output_mean_power":0.000638610015524155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.5308592276752844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945508810428353e-05,"output_power":9.786239646785837e-05,"reference_active":true,"delta_db":-0.07011159838514489},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423063898822289e-06,"output_power":5.149815122445201e-06,"reference_active":true,"delta_db":-0.22453083448926608},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8432082747666131e-07,"output_power":1.837165534087691e-07,"reference_active":true,"delta_db":-0.014261221424414511}],"active_band_count":3,"active_band_mae_db":0.10296788476627515}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1096,"rms_floor":2.9435709173680695e-05,"log_rms_mae_db":2.697918258666109,"rms_envelope_nrmse":0.37042967268455956,"envelope_correlation":0.9024150805582469,"rms_gain_error_db":0.3702882369992272,"reference_mean_power":1.4688872454843848e-05,"output_mean_power":1.5996216871252064e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1096,"frequency_resolution_hz":0.18248175182481752,"power_floor":9.038086104701172e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4018145980274595e-06,"output_power":3.6022321931562456e-06,"reference_active":true,"delta_db":0.24861062717127483},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2285201909756885e-06,"output_power":1.3456170433517168e-06,"reference_active":true,"delta_db":0.39539180595484347},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7360839982104215e-07,"output_power":3.647869092125254e-07,"reference_active":true,"delta_db":1.2490981862946136}],"active_band_count":3,"active_band_mae_db":0.6310335398069106}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87760,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":2.1616481165467548e-05,"log_rms_mae_db":1.9701447554705154,"rms_envelope_nrmse":0.23214914697866762,"envelope_correlation":0.9603415593461878,"rms_gain_error_db":0.3702878552430161,"reference_mean_power":1.476972002493505e-05,"output_mean_power":1.6084258652595616e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.51604517212682e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5641156211998283e-06,"output_power":3.80937614780251e-06,"reference_active":true,"delta_db":0.2890207389226331},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790775026435665e-07,"output_power":6.367324517151806e-07,"reference_active":true,"delta_db":0.4122029191273618},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8657206259627095e-08,"output_power":2.1817542495417613e-08,"reference_active":true,"delta_db":0.6795921775935093}],"active_band_count":3,"active_band_mae_db":0.46027194521450143}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"hnm","status":"measured","mos":{"status":"measured","value":2.7264857292175293,"predicted_mos":2.7264857292175293,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e9a05f6cd70e65c70c7dc59ff0729abe5465e01690fa9c63e4dd7b1bb7aadb97","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6335008250007377},"started_at":"2026-10-13T04:39:26.605340+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5761000213387888e-05,"native_input_samples":96000,"native_input_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","output_sha256":"55b03f6738e7b141829c6f216e30e60f6a2a38cfdb8a7c9544122a993ffba685","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1223899249998794,"synthesis_s":0.5139308679999885,"total_s":1.636320792999868,"rtf":0.27272013216664465,"output_peak":0.5976233114832721,"output_rms":0.07866956177943712,"finished_at":"2026-10-13T04:39:28.247006+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"55b03f6738e7b141829c6f216e30e60f6a2a38cfdb8a7c9544122a993ffba685","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.452598857538184,"spectral_convergence":0.25915900056688895,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.57010215667468,"spectral_convergence":0.2806307383958693,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.007403441608762,"spectral_convergence":0.29192313209981857,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.0002423614061474929,"delta_db":0.03866865032532096,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.0024376020714815504,"delta_db":0.07090134622329733,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.002716084883066144,"delta_db":-0.11250113422651574,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.00048821624039368616,"delta_db":-0.06441005471808658,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.0002648752962334261,"delta_db":-0.10811884331504981,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":3.976005324637608e-05,"delta_db":0.39244443940069035,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1311740780348268,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.028113636400025174,"pitch":null,"output_rms_dbfs":-22.083865380322067,"output_peak":0.5976233114832721,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676701485273875},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.452598857538184,"spectral_convergence":0.25915900056688895,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.57010215667468,"spectral_convergence":0.2806307383958693,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.007403441608762,"spectral_convergence":0.29192313209981857,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.0002423614061474929,"delta_db":0.03866865032532096,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.0024376020714815504,"delta_db":0.07090134622329733,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.002716084883066144,"delta_db":-0.11250113422651574,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.00048821624039368616,"delta_db":-0.06441005471808658,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.0002648752962334261,"delta_db":-0.10811884331504981,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":3.976005324637608e-05,"delta_db":0.39244443940069035,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1311740780348268,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.028113636400025174,"pitch":null,"output_rms_dbfs":-22.083865380322067,"output_peak":0.5976233114832721,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.676701485273875},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.888797883889831,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07892460423327397,"rms_dbfs":-22.055751743922045,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.02751756459009,"mean":1.2715657552083334e-08,"zero_energy":false},"output":{"rms":0.07866956177943712,"rms_dbfs":-22.083865380322067,"peak":0.5976233114832721,"peak_dbfs":-4.471449406882163,"crest_factor_db":17.6124159734399,"mean":4.0663628517600256e-05,"zero_energy":false},"rms_difference_db":-0.028113636400021846},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":2.251894236210023e-06,"output_rms_dbfs":-56.47452010671937,"output_zero_energy":false,"output_peak":0.033611457153703124,"output_energy_fraction":5.8824062072695465e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":229,"both_estimated_voiced_frames":215,"reference_voiced_output_unvoiced_frames":40,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.8431372549019608,"absolute_difference_cents_median":1.6107610526852352,"absolute_difference_cents_p95":14.367502479334389,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024330354027503592,"log_rms_mae_db":0.8616546480355214,"rms_envelope_nrmse":0.0789298975807123,"envelope_correlation":0.993976790121826,"rms_gain_error_db":-0.0280268691840416,"reference_mean_power":0.006233969602179056,"output_mean_power":0.006193868679482858},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8072421002875186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017414213578018352,"output_power":0.0017514736039885783,"reference_active":true,"delta_db":0.024997299661565723},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002371798239966289,"output_power":0.00021567868467488937,"reference_active":true,"delta_db":-0.4127051615005586},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.501630810521967e-05,"output_power":6.578832445066143e-05,"reference_active":true,"delta_db":0.05126520756964502}],"active_band_count":3,"active_band_mae_db":0.1629892229105898}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022924128639936837,"log_rms_mae_db":0.4826955381380167,"rms_envelope_nrmse":0.04186800461737243,"envelope_correlation":0.9981756056642476,"rms_gain_error_db":-0.027989356889599235,"reference_mean_power":0.006262494274967633,"output_mean_power":0.00622226360786042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.530420990426951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015992997191427703,"output_power":0.0016116516537026356,"reference_active":true,"delta_db":0.03341317172494653},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010914154654466237,"output_power":9.87023981400883e-05,"reference_active":true,"delta_db":-0.43662398769438976},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.77390165426219e-06,"output_power":4.599389684305456e-06,"reference_active":true,"delta_db":-0.16173261161898847}],"active_band_count":3,"active_band_mae_db":0.2105899236794416}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00021113569197140015,"log_rms_mae_db":1.2782442584491414,"rms_envelope_nrmse":0.09959996411279254,"envelope_correlation":0.9918047456662915,"rms_gain_error_db":-0.04829113136769827,"reference_mean_power":0.0043153114305736935,"output_mean_power":0.004267593357997609},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.429798497423617e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014993372451038876,"output_power":0.0014661978129370194,"reference_active":true,"delta_db":-0.09706762103123177},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023063830674655453,"output_power":0.00020749312223086714,"reference_active":true,"delta_db":-0.4592773523811424},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9359539965921096e-05,"output_power":6.331269812723424e-05,"reference_active":true,"delta_db":0.28000295244451645}],"active_band_count":3,"active_band_mae_db":0.2787826419522969}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001994421712728477,"log_rms_mae_db":0.857624262300184,"rms_envelope_nrmse":0.051617811673567146,"envelope_correlation":0.9976357097704182,"rms_gain_error_db":-0.04827328380834968,"reference_mean_power":0.004336654014858995,"output_mean_power":0.004288717563830619},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.221204429766169e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001403505038426243,"output_power":0.0013705517801760236,"reference_active":true,"delta_db":-0.10318527979666675},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010608827780695942,"output_power":9.27417023080508e-05,"reference_active":true,"delta_db":-0.5839233604345258},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.723378650149635e-06,"output_power":4.482672713487097e-06,"reference_active":true,"delta_db":-0.22715730689538965}],"active_band_count":3,"active_band_mae_db":0.3047553157088607}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011279772332784735,"log_rms_mae_db":1.424289497423381,"rms_envelope_nrmse":0.13253627850362215,"envelope_correlation":0.9873065932331117,"rms_gain_error_db":-0.09384166680128599,"reference_mean_power":0.000725883836364577,"output_mean_power":0.0007103672960888602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4338411984909043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026856860022608727,"output_power":0.0002715883956534637,"reference_active":true,"delta_db":0.048559739636707205},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8973711977024913e-05,"output_power":1.9713648758460862e-05,"reference_active":true,"delta_db":0.16614710728322357},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310520152307557e-06,"output_power":7.178726500713833e-06,"reference_active":true,"delta_db":1.3091034641054602}],"active_band_count":3,"active_band_mae_db":0.507936770341797}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010728134925657118,"log_rms_mae_db":0.9011977798563661,"rms_envelope_nrmse":0.0729411343376791,"envelope_correlation":0.9959987538377181,"rms_gain_error_db":-0.09384488194490447,"reference_mean_power":0.0007295243282152713,"output_mean_power":0.000713929440015802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2203166741702414e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024656486610752886,"output_power":0.0002516410877197264,"reference_active":true,"delta_db":0.08850361232317584},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.896584068544471e-06,"output_power":9.395329525085415e-06,"reference_active":true,"delta_db":0.23688729851061857},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4772410320228666e-07,"output_power":4.7332910614049815e-07,"reference_active":true,"delta_db":1.3392841536654516}],"active_band_count":3,"active_band_mae_db":0.5548916881664153}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":1.6905068641955256,"rms_envelope_nrmse":0.4248210278097514,"envelope_correlation":0.887952664591455,"rms_gain_error_db":0.46465631846264505,"reference_mean_power":3.947976781678989e-05,"output_mean_power":4.393799364801856e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.253181223289244e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6595680637210912e-05,"output_power":1.4678807376516772e-05,"reference_active":true,"delta_db":-0.5330429732881814},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.516181974676607e-06,"output_power":6.538466222965196e-06,"reference_active":true,"delta_db":-1.147690483566834},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.102920350990826e-06,"output_power":1.2087013413630418e-06,"reference_active":true,"delta_db":-2.405038198392433}],"active_band_count":3,"active_band_mae_db":1.3619238850824829}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":1.2859207938704154,"rms_envelope_nrmse":0.3500793077105018,"envelope_correlation":0.9183469995102177,"rms_gain_error_db":0.4646308184437874,"reference_mean_power":3.967785924701113e-05,"output_mean_power":4.415819513791967e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.933756855113043e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8381327852746303e-05,"output_power":1.5501968741677526e-05,"reference_active":true,"delta_db":-0.7399002443887643},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.503290667812916e-06,"output_power":3.4127428566554065e-06,"reference_active":true,"delta_db":-1.2042641339496312},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.651394869607331e-07,"output_power":8.782885077119208e-08,"reference_active":true,"delta_db":-2.742137309343229}],"active_band_count":3,"active_band_mae_db":1.5621005625605413}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9902093410491943,"predicted_mos":2.9902093410491943,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"cb96b094eb2a53db618a7618c93fe8cd9ec16bd16a8133a1bb86d76c3e6b3438","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.628007000999787},"started_at":"2026-10-13T04:42:23.748060+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.873499943234492e-05,"native_input_samples":96000,"native_input_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","output_sha256":"23b2745ab0a4e480c3d46df1dc342af8df742c248d1dffcff694df60d6121d62","output_samples":96000,"length_delta_samples":0,"analysis_s":2.2817584879994683,"synthesis_s":0.09799191300044185,"total_s":2.37975040099991,"rtf":0.3966250668333184,"output_peak":0.7581434197186258,"output_rms":0.09315763944606714,"finished_at":"2026-10-13T04:42:26.133471+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"23b2745ab0a4e480c3d46df1dc342af8df742c248d1dffcff694df60d6121d62","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.734981528055378,"spectral_convergence":0.41895039309162013,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.165000720437733,"spectral_convergence":0.2973785124352879,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.590775228087065,"spectral_convergence":0.3373933650691287,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.00024031905980415665,"delta_db":0.0019161772904466594,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.003289239796950374,"delta_db":1.3722286054849118,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.00408255655616306,"delta_db":1.6573875797440358,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.000674093359037044,"delta_db":1.3366682112972672,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.0003506296098065281,"delta_db":1.1099524864474943,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":4.1507405402283033e-05,"delta_db":0.5792307340520385,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0095639657193658,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4401212491749456,"pitch":null,"output_rms_dbfs":-20.615630494747098,"output_peak":0.7581434197186258,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4969191588600586},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.734981528055378,"spectral_convergence":0.41895039309162013,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.165000720437733,"spectral_convergence":0.2973785124352879,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.590775228087065,"spectral_convergence":0.3373933650691287,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.00024031905980415665,"delta_db":0.0019161772904466594,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.003289239796950374,"delta_db":1.3722286054849118,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.00408255655616306,"delta_db":1.6573875797440358,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.000674093359037044,"delta_db":1.3366682112972672,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.0003506296098065281,"delta_db":1.1099524864474943,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":4.1507405402283033e-05,"delta_db":0.5792307340520385,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0095639657193658,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4401212491749456,"pitch":null,"output_rms_dbfs":-20.615630494747098,"output_peak":0.7581434197186258,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4969191588600586},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9389479027599871,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07892460423327397,"rms_dbfs":-22.055751743922045,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.02751756459009,"mean":1.2715657552083334e-08,"zero_energy":false},"output":{"rms":0.09315763944606714,"rms_dbfs":-20.615630494747098,"peak":0.7581434197186258,"peak_dbfs":-2.4049726023828253,"crest_factor_db":18.21065789236427,"mean":9.73546202023676e-06,"zero_energy":false},"rms_difference_db":1.4401212491749469},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":3.157765208842637e-06,"output_rms_dbfs":-55.00620164478076,"output_zero_energy":false,"output_peak":0.03630409963709374,"output_energy_fraction":5.882519410377406e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":294,"both_estimated_voiced_frames":235,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":59,"paired_voiced_fraction":0.9215686274509803,"absolute_difference_cents_median":9.312863209251113,"absolute_difference_cents_p95":28.33462090567358,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024330354027503592,"log_rms_mae_db":1.5234645103309372,"rms_envelope_nrmse":0.21622614191162925,"envelope_correlation":0.9888937384421971,"rms_gain_error_db":1.4400706877277472,"reference_mean_power":0.006233969602179056,"output_mean_power":0.008685038521081882},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8072421002875186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017414213578018352,"output_power":0.0027938172087220094,"reference_active":true,"delta_db":2.0529412149594863},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002371798239966289,"output_power":0.0004031371602412536,"reference_active":true,"delta_db":2.303750897975126},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.501630810521967e-05,"output_power":7.513062577555652e-05,"reference_active":true,"delta_db":0.6279470134223729}],"active_band_count":3,"active_band_mae_db":1.6615463754523285}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022924128639936837,"log_rms_mae_db":1.2565184063434085,"rms_envelope_nrmse":0.2025036397242179,"envelope_correlation":0.9930017925456835,"rms_gain_error_db":1.440407807616431,"reference_mean_power":0.006262494274967633,"output_mean_power":0.008725455794651737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.530420990426951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015992997191427703,"output_power":0.0026017072990611174,"reference_active":true,"delta_db":2.11328574387284},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010914154654466237,"output_power":0.00018567311994470878,"reference_active":true,"delta_db":2.307589316302515},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.77390165426219e-06,"output_power":5.298740336000701e-06,"reference_active":true,"delta_db":0.4529916950732782}],"active_band_count":3,"active_band_mae_db":1.6246222517495443}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00021113569197140015,"log_rms_mae_db":1.6589027137817507,"rms_envelope_nrmse":0.25802980134306347,"envelope_correlation":0.9873001970849306,"rms_gain_error_db":1.70997603710273,"reference_mean_power":0.0043153114305736935,"output_mean_power":0.006397491939573376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.429798497423617e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014993372451038876,"output_power":0.002520701794375122,"reference_active":true,"delta_db":2.2562214108487346},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023063830674655453,"output_power":0.0004118374884195155,"reference_active":true,"delta_db":2.5179443572158435},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9359539965921096e-05,"output_power":7.864735066429462e-05,"reference_active":true,"delta_db":1.2219357092653547}],"active_band_count":3,"active_band_mae_db":1.9987004924433112}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001994421712728477,"log_rms_mae_db":1.474747928702995,"rms_envelope_nrmse":0.24426500246995778,"envelope_correlation":0.991261518156852,"rms_gain_error_db":1.7100448810955335,"reference_mean_power":0.004336654014858995,"output_mean_power":0.006429234447048746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.221204429766169e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001403505038426243,"output_power":0.002380030868761901,"reference_active":true,"delta_db":2.2936861380704356},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010608827780695942,"output_power":0.00018401303086489697,"reference_active":true,"delta_db":2.391811793087317},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.723378650149635e-06,"output_power":5.473987815617354e-06,"reference_active":true,"delta_db":0.6405106415694242}],"active_band_count":3,"active_band_mae_db":1.7753361909090588}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011279772332784735,"log_rms_mae_db":1.456380592860827,"rms_envelope_nrmse":0.20667883807093257,"envelope_correlation":0.9898280214447677,"rms_gain_error_db":1.3014289419887517,"reference_mean_power":0.000725883836364577,"output_mean_power":0.0009795125844735105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4338411984909043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026856860022608727,"output_power":0.00035302950539140487,"reference_active":true,"delta_db":1.187557685969306},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8973711977024913e-05,"output_power":3.171762518141976e-05,"reference_active":true,"delta_db":2.2314835893405824},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310520152307557e-06,"output_power":6.716388313056047e-06,"reference_active":true,"delta_db":1.019987360202637}],"active_band_count":3,"active_band_mae_db":1.4796762118375086}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010728134925657118,"log_rms_mae_db":1.160199375862126,"rms_envelope_nrmse":0.18846031239122718,"envelope_correlation":0.9939432216754477,"rms_gain_error_db":1.301424673897076,"reference_mean_power":0.0007295243282152713,"output_mean_power":0.0009844241214730595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2203166741702414e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024656486610752886,"output_power":0.00032609442956293337,"reference_active":true,"delta_db":1.2141218760387962},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.896584068544471e-06,"output_power":1.4985202215612363e-05,"reference_active":true,"delta_db":2.264393208569085},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4772410320228666e-07,"output_power":4.6283817276178154e-07,"reference_active":true,"delta_db":1.2419437463589353}],"active_band_count":3,"active_band_mae_db":1.573486276988939}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":1.6043471233264888,"rms_envelope_nrmse":0.2569740070470499,"envelope_correlation":0.9642517037101568,"rms_gain_error_db":0.7369151893322627,"reference_mean_power":3.947976781678989e-05,"output_mean_power":4.6780634590628685e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.253181223289244e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6595680637210912e-05,"output_power":1.86784464333878e-05,"reference_active":true,"delta_db":0.5134568246707565},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.516181974676607e-06,"output_power":9.942648164071691e-06,"reference_active":true,"delta_db":0.6725713859125084},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.102920350990826e-06,"output_power":2.5459199912990283e-06,"reference_active":true,"delta_db":0.8302192737978333}],"active_band_count":3,"active_band_mae_db":0.6720824947936994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":1.195290910714948,"rms_envelope_nrmse":0.1809619380884837,"envelope_correlation":0.9834787858303894,"rms_gain_error_db":0.7369146196374137,"reference_mean_power":3.967785924701113e-05,"output_mean_power":4.70153522660383e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.933756855113043e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8381327852746303e-05,"output_power":2.1179327555695138e-05,"reference_active":true,"delta_db":0.6153528585012832},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.503290667812916e-06,"output_power":5.033537020926656e-06,"reference_active":true,"delta_db":0.4834328744511796},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.651394869607331e-07,"output_power":2.241573172744635e-07,"reference_active":true,"delta_db":1.3270198927737884}],"active_band_count":3,"active_band_mae_db":0.8086018752420837}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"facodec","status":"measured","mos":{"status":"measured","value":4.111084938049316,"predicted_mos":4.111084938049316,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"54514cf69eeda83ce8f2bf7514fa4f0f544e5c0f591c06534ecbf24d3214e2a4","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5076227399986237},"started_at":"2026-10-13T04:53:47.096936+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.803900042636087e-05,"native_input_samples":96000,"native_input_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","output_sha256":"4402ced79f7262ca97c8a6462550df463e0d80ed0b03af77282a43916e788249","output_samples":96000,"length_delta_samples":0,"analysis_s":1.8952152639994893,"synthesis_s":7.4766879430007975,"total_s":9.371903207000287,"rtf":1.561983867833381,"output_peak":0.5228548645973206,"output_rms":0.07203745926243227,"finished_at":"2026-10-13T04:53:56.474292+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"4402ced79f7262ca97c8a6462550df463e0d80ed0b03af77282a43916e788249","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.253456976566262,"spectral_convergence":0.2119872257949954,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.562012565840312,"spectral_convergence":0.2146811555804876,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4729882558758005,"spectral_convergence":0.2299454925270082,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.0001916754444218789,"delta_db":-0.980301196241683,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.0021397287715061593,"delta_db":-0.4951394927226838,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.002252873464921111,"delta_db":-0.9245665202263647,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.00035649448333907645,"delta_db":-1.430004138489739,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.00021856968212323347,"delta_db":-0.9426341910806437,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":3.0053690675130124e-05,"delta_db":-0.8230470122980797,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9326154251765324,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7930805162052734,"pitch":null,"output_rms_dbfs":-22.848832260127313,"output_peak":0.5228548645973206,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.429485932760792},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.253456976566262,"spectral_convergence":0.2119872257949954,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.562012565840312,"spectral_convergence":0.2146811555804876,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4729882558758005,"spectral_convergence":0.2299454925270082,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.0001916754444218789,"delta_db":-0.980301196241683,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.0021397287715061593,"delta_db":-0.4951394927226838,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.002252873464921111,"delta_db":-0.9245665202263647,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.00035649448333907645,"delta_db":-1.430004138489739,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.00021856968212323347,"delta_db":-0.9426341910806437,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":3.0053690675130124e-05,"delta_db":-0.8230470122980797,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9326154251765324,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7930805162052734,"pitch":null,"output_rms_dbfs":-22.848832260127313,"output_peak":0.5228548645973206,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.429485932760792},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9218722062277501,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07892460423327397,"rms_dbfs":-22.055751743922045,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.02751756459009,"mean":1.2715657552083334e-08,"zero_energy":false},"output":{"rms":0.07203745926243227,"rms_dbfs":-22.848832260127313,"peak":0.5228548645973206,"peak_dbfs":-5.632376939745414,"crest_factor_db":17.2164553203819,"mean":-8.72132224017482e-06,"zero_energy":false},"rms_difference_db":-0.7930805162052685},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":1.4083840407536525e-06,"output_rms_dbfs":-58.51278904825862,"output_zero_energy":false,"output_peak":0.009095984511077404,"output_energy_fraction":4.387577544096582e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":255,"both_estimated_voiced_frames":224,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.8784313725490196,"absolute_difference_cents_median":3.873750402749289,"absolute_difference_cents_p95":16.807708712328623,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024330354027503592,"log_rms_mae_db":1.1947864774764594,"rms_envelope_nrmse":0.11107158209919407,"envelope_correlation":0.9950246577365958,"rms_gain_error_db":-0.7930558371917463,"reference_mean_power":0.006233969602179056,"output_mean_power":0.0051934875701532615},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8072421002875186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017414213578018352,"output_power":0.0013777720462037095,"reference_active":true,"delta_db":-1.017264975473687},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002371798239966289,"output_power":0.00020347974869245475,"reference_active":true,"delta_db":-0.6655654986653672},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.501630810521967e-05,"output_power":5.432582578369561e-05,"reference_active":true,"delta_db":-0.7801596825617995}],"active_band_count":3,"active_band_mae_db":0.8209967189002846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022924128639936837,"log_rms_mae_db":0.9910127011345908,"rms_envelope_nrmse":0.10219408646401976,"envelope_correlation":0.9967995381539327,"rms_gain_error_db":-0.7926261385606805,"reference_mean_power":0.006262494274967633,"output_mean_power":0.005217767555917939},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.530420990426951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015992997191427703,"output_power":0.0012562523227634319,"reference_active":true,"delta_db":-1.048529832175207},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010914154654466237,"output_power":9.749643183366486e-05,"reference_active":true,"delta_db":-0.4900138176323366},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.77390165426219e-06,"output_power":4.0768970787210974e-06,"reference_active":true,"delta_db":-0.6854372016549997}],"active_band_count":3,"active_band_mae_db":0.7413269504875144}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00021113569197140015,"log_rms_mae_db":1.3157711836542774,"rms_envelope_nrmse":0.11532350135350594,"envelope_correlation":0.9944391329620059,"rms_gain_error_db":-0.7648079140536699,"reference_mean_power":0.0043153114305736935,"output_mean_power":0.0036185231235979958},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.429798497423617e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014993372451038876,"output_power":0.0011758760986291144,"reference_active":true,"delta_db":-1.0553776665814252},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023063830674655453,"output_power":0.00019623116584795456,"reference_active":true,"delta_db":-0.7016345688865979},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9359539965921096e-05,"output_power":4.700997609211671e-05,"reference_active":true,"delta_db":-1.0130049620003256}],"active_band_count":3,"active_band_mae_db":0.9233390658227828}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001994421712728477,"log_rms_mae_db":0.9948511963623071,"rms_envelope_nrmse":0.10489676763652377,"envelope_correlation":0.9962335543957502,"rms_gain_error_db":-0.7647883969393866,"reference_mean_power":0.004336654014858995,"output_mean_power":0.0036364358877140333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.221204429766169e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001403505038426243,"output_power":0.0010945576507514466,"reference_active":true,"delta_db":-1.079753350864736},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010608827780695942,"output_power":9.264183392288304e-05,"reference_active":true,"delta_db":-0.5886025561897112},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.723378650149635e-06,"output_power":3.823897087564978e-06,"reference_active":true,"delta_db":-0.9174656662925157}],"active_band_count":3,"active_band_mae_db":0.8619405244489876}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011279772332784735,"log_rms_mae_db":1.9154883864553165,"rms_envelope_nrmse":0.17894284652519635,"envelope_correlation":0.9885203204256311,"rms_gain_error_db":-1.2554085856168933,"reference_mean_power":0.000725883836364577,"output_mean_power":0.0005436586060941437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4338411984909043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026856860022608727,"output_power":0.0002205835067306831,"reference_active":true,"delta_db":-0.854821989922493},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8973711977024913e-05,"output_power":1.946141400711981e-05,"reference_active":true,"delta_db":0.11022087936130678},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310520152307557e-06,"output_power":5.785879065428915e-06,"reference_active":true,"delta_db":0.3723229068951338}],"active_band_count":3,"active_band_mae_db":0.44578859205964455}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010728134925657118,"log_rms_mae_db":1.6674584200812046,"rms_envelope_nrmse":0.16533171134970495,"envelope_correlation":0.9919249328339249,"rms_gain_error_db":-1.2553791695754837,"reference_mean_power":0.0007295243282152713,"output_mean_power":0.0005463888929605691},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2203166741702414e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024656486610752886,"output_power":0.00020515230224154383,"reference_active":true,"delta_db":-0.7985479749415672},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.896584068544471e-06,"output_power":9.22064848701006e-06,"reference_active":true,"delta_db":0.15538179014783018},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4772410320228666e-07,"output_power":4.034106299541824e-07,"reference_active":true,"delta_db":0.6451254204710124}],"active_band_count":3,"active_band_mae_db":0.5330183951868032}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":2.0747352148635816,"rms_envelope_nrmse":0.2692303514262657,"envelope_correlation":0.9535136121110483,"rms_gain_error_db":-0.8946316197003302,"reference_mean_power":3.947976781678989e-05,"output_mean_power":3.213005196267714e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.253181223289244e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6595680637210912e-05,"output_power":1.6628578007749713e-05,"reference_active":true,"delta_db":0.008600433385699506},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.516181974676607e-06,"output_power":8.24440618262288e-06,"reference_active":true,"delta_db":-0.14085552607512825},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.102920350990826e-06,"output_power":1.7745382368820558e-06,"reference_active":true,"delta_db":-0.7373746215874039}],"active_band_count":3,"active_band_mae_db":0.2956101936827439}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":1.7656369596050858,"rms_envelope_nrmse":0.22211546466838208,"envelope_correlation":0.9682608560228709,"rms_gain_error_db":-0.8946215210954928,"reference_mean_power":3.967785924701113e-05,"output_mean_power":3.229134096464999e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.933756855113043e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8381327852746303e-05,"output_power":1.8798190400575293e-05,"reference_active":true,"delta_db":0.09739162833992804},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.503290667812916e-06,"output_power":4.184015047453397e-06,"reference_active":true,"delta_db":-0.319367419211593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.651394869607331e-07,"output_power":1.3228727912584742e-07,"reference_active":true,"delta_db":-0.9633284702506051}],"active_band_count":3,"active_band_mae_db":0.4600291726007087}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.680562734603882,"predicted_mos":3.680562734603882,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"d53b651a9ff685644054b6135576da91adeb00ecc278ac9e997b352be9089909","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0292282400005206},"started_at":"2026-10-13T07:17:37.180446+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006780605999665568,"native_input_samples":264600,"native_input_sha256":"9a0e3a9a41cf545ab36029e388f6c7480b0b6ed74b8b326b0dda874bce5302f4","output_sha256":"1c270c7d5ec4796c7ab2dd18f9b09960d09ef7f7edb8e360c756f549a62a90a4","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1453289550008776,"synthesis_s":4.107899975999317,"total_s":4.253228931000194,"rtf":0.7088714885000323,"output_peak":0.5401735901832581,"output_rms":0.07498153470365071,"finished_at":"2026-10-13T07:17:41.459216+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"1c270c7d5ec4796c7ab2dd18f9b09960d09ef7f7edb8e360c756f549a62a90a4","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.493240539503855,"output_peak":0.535969395290114},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.86913270229048,"spectral_convergence":0.3983339295724719,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.67701753394585,"spectral_convergence":0.31290348103773324,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.283805196906222,"spectral_convergence":0.353504150687327,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.0001568371739762339,"delta_db":-1.851475902996477,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.002129108190965598,"delta_db":-0.516749449161451,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.002777811828267859,"delta_db":-0.014906289529106572,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.00035031841672669467,"delta_db":-1.505902581096984,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.00019708078108689666,"delta_db":-1.392090649222932,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":2.319470569597282e-05,"delta_db":-1.948136470080509,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.20487689034791,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.43580919660753664,"pitch":null,"output_rms_dbfs":-22.491560940529574,"output_peak":0.535969395290114,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.60998514438085},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8301053129007069,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07892460423327397,"rms_dbfs":-22.055751743922045,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.02751756459009,"mean":1.2715657552083334e-08,"zero_energy":false},"output":{"rms":0.07506231475726854,"rms_dbfs":-22.491560940529574,"peak":0.535969395290114,"peak_dbfs":-5.417200170178116,"crest_factor_db":17.07436077035146,"mean":-7.342874053006151e-05,"zero_energy":false},"rms_difference_db":-0.43580919660752926},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":7.216896572346777e-07,"output_rms_dbfs":-61.41649518670764,"output_zero_energy":false,"output_peak":0.006315231245699746,"output_energy_fraction":2.0707470878212666e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":263,"both_estimated_voiced_frames":236,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9254901960784314,"absolute_difference_cents_median":6.535622180364518,"absolute_difference_cents_p95":55.191926833100595,"difference_over_600_cents_frames":9,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024330354027503592,"log_rms_mae_db":1.5717859435992965,"rms_envelope_nrmse":0.09764363043983211,"envelope_correlation":0.9930568115918736,"rms_gain_error_db":-0.43565705706859614,"reference_mean_power":0.006233969602179056,"output_mean_power":0.005638959488710005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8072421002875186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017414213578018352,"output_power":0.001659072814103971,"reference_active":true,"delta_db":-0.21038419649271412},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002371798239966289,"output_power":0.00023754674174222837,"reference_active":true,"delta_db":0.006713354400510924},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.501630810521967e-05,"output_power":5.232643837533573e-05,"reference_active":true,"delta_db":-0.9430112954505252}],"active_band_count":3,"active_band_mae_db":0.3867029487812501}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022924128639936837,"log_rms_mae_db":1.1745447699441283,"rms_envelope_nrmse":0.07109600755203058,"envelope_correlation":0.9972876525066484,"rms_gain_error_db":-0.4350641346288615,"reference_mean_power":0.006262494274967633,"output_mean_power":0.00566553502029726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.530420990426951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015992997191427703,"output_power":0.0015166528812382586,"reference_active":true,"delta_db":-0.23043666487519032},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010914154654466237,"output_power":0.00010875242149525635,"reference_active":true,"delta_db":-0.015511676886082718},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.77390165426219e-06,"output_power":4.150651525809794e-06,"reference_active":true,"delta_db":-0.6075719492779529}],"active_band_count":3,"active_band_mae_db":0.284506763679742}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00021113569197140015,"log_rms_mae_db":2.150038761631039,"rms_envelope_nrmse":0.11001296367201775,"envelope_correlation":0.9902410827339234,"rms_gain_error_db":-0.10152261342574397,"reference_mean_power":0.0043153114305736935,"output_mean_power":0.004215604722195417},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.429798497423617e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014993372451038876,"output_power":0.0014849197602480068,"reference_active":true,"delta_db":-0.04196342888621121},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023063830674655453,"output_power":0.00024128856510898937,"reference_active":true,"delta_db":0.1960529979745506},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9359539965921096e-05,"output_power":4.702604196327701e-05,"reference_active":true,"delta_db":-1.0115209946005863}],"active_band_count":3,"active_band_mae_db":0.41651247382044937}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001994421712728477,"log_rms_mae_db":1.6616537525248927,"rms_envelope_nrmse":0.07736333498850363,"envelope_correlation":0.9948622401700687,"rms_gain_error_db":-0.10126492031133766,"reference_mean_power":0.004336654014858995,"output_mean_power":0.004236705560800152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.221204429766169e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001403505038426243,"output_power":0.0013842770927571298,"reference_active":true,"delta_db":-0.059909438617150076},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010608827780695942,"output_power":0.00011088775636802046,"reference_active":true,"delta_db":0.19216197015000133},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.723378650149635e-06,"output_power":4.142866054257411e-06,"reference_active":true,"delta_db":-0.5695187017225338}],"active_band_count":3,"active_band_mae_db":0.2738633701632284}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011279772332784735,"log_rms_mae_db":3.5666587370102922,"rms_envelope_nrmse":0.2433784258599511,"envelope_correlation":0.9732394871230106,"rms_gain_error_db":-1.5656729834834189,"reference_mean_power":0.000725883836364577,"output_mean_power":0.0005061739929609478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4338411984909043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026856860022608727,"output_power":0.00019357524088210439,"reference_active":true,"delta_db":-1.4220542727301835},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8973711977024913e-05,"output_power":1.561090593429035e-05,"reference_active":true,"delta_db":-0.8472419682582522},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310520152307557e-06,"output_power":4.683292582586399e-06,"reference_active":true,"delta_db":-0.5458577062457712}],"active_band_count":3,"active_band_mae_db":0.9383846490780691}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010728134925657118,"log_rms_mae_db":3.0476513602158493,"rms_envelope_nrmse":0.21571743079131744,"envelope_correlation":0.9830507753461013,"rms_gain_error_db":-1.5656200371128453,"reference_mean_power":0.0007295243282152713,"output_mean_power":0.0005087187860616931},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2203166741702414e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024656486610752886,"output_power":0.00017653008719466898,"reference_active":true,"delta_db":-1.4511245680050782},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.896584068544471e-06,"output_power":7.719802160275023e-06,"reference_active":true,"delta_db":-0.6161711641879363},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4772410320228666e-07,"output_power":3.087996206466991e-07,"reference_active":true,"delta_db":-0.5155803758739255}],"active_band_count":3,"active_band_mae_db":0.86095870268898}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":4.145481171102564,"rms_envelope_nrmse":0.28649009519108765,"envelope_correlation":0.9472238353022111,"rms_gain_error_db":-0.9385289840965589,"reference_mean_power":3.947976781678989e-05,"output_mean_power":3.1806925487538636e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.253181223289244e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6595680637210912e-05,"output_power":1.5615148754066477e-05,"reference_active":true,"delta_db":-0.26448943000972264},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.516181974676607e-06,"output_power":7.347699538303116e-06,"reference_active":true,"delta_db":-0.6409354406334853},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.102920350990826e-06,"output_power":1.181027646130103e-06,"reference_active":true,"delta_db":-2.505627599770407}],"active_band_count":3,"active_band_mae_db":1.1370174901378718}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":3.849338892954433,"rms_envelope_nrmse":0.23624707863037267,"envelope_correlation":0.9639414387849147,"rms_gain_error_db":-0.9384892313315034,"reference_mean_power":3.967785924701113e-05,"output_mean_power":3.196681070597586e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.933756855113043e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8381327852746303e-05,"output_power":1.6631363320436654e-05,"reference_active":true,"delta_db":-0.4344903023699493},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.503290667812916e-06,"output_power":3.7295603185206916e-06,"reference_active":true,"delta_db":-0.8187234425707379},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.651394869607331e-07,"output_power":9.636721624873858e-08,"reference_active":true,"delta_db":-2.339216172994537}],"active_band_count":3,"active_band_mae_db":1.1974766393117413}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"golf","status":"measured","mos":{"status":"measured","value":2.3340821266174316,"predicted_mos":2.3340821266174316,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"3ed14df35a4b3185b115e2a17d9ecdc8dd9ca0d4436da4fddbeee8cb488972d9","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4694585729994287},"started_at":"2026-10-13T05:02:39.184804+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002705312999751186,"native_input_samples":144000,"native_input_sha256":"e669d50a84441c33a4d34b0a29184934ef16bd8e1b09fe229e0e31066dab7953","output_sha256":"9d8166714e16525611d5c068c25002057a0f88a730065cb84b7f4f7c77d5f5d1","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5498116090002441,"synthesis_s":0.1163605080000707,"total_s":0.6661721170003148,"rtf":0.11102868616671913,"output_peak":0.5211738348007202,"output_rms":0.05264203973158781,"finished_at":"2026-10-13T05:02:39.861492+00:00","analysis_stages":{"external_f0_s":0.08036137999988568},"evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"9d8166714e16525611d5c068c25002057a0f88a730065cb84b7f4f7c77d5f5d1","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.567262444122342,"output_peak":0.5080571729285461},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.378198730837413,"spectral_convergence":0.6327875114457042,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.613289999916591,"spectral_convergence":0.614051627606779,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.655447227665912,"spectral_convergence":0.7132252959505812,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00023911826357180183,"output_power":0.00033306768139189434,"delta_db":1.4391974548712954,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002402144489188231,"output_power":0.0009104450423333844,"delta_db":-4.213453913223786,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027920114724290703,"output_power":0.0011126130183933217,"delta_db":-3.9957306127875847,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004963280560468133,"output_power":0.00021745243281000938,"delta_db":-3.5840455453261226,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002720066355537253,"output_power":0.00018431078064781018,"delta_db":-1.6902876012311903,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.63848816257074e-05,"output_power":1.717985010698049e-05,"delta_db":-3.259015960074155,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":3.030288514585689,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-3.5177118307392785,"pitch":null,"output_rms_dbfs":-25.567262444122342,"output_peak":0.5080571729285461,"reference_rms_dbfs":-22.049550613383065,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.882311986139972},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.7599367924487009,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07898097111592987,"rms_dbfs":-22.049550613383065,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.021316434051116,"mean":1.1081090554570117e-06,"zero_energy":false},"output":{"rms":0.05267892183486105,"rms_dbfs":-25.567262444122342,"peak":0.5080571729285461,"peak_dbfs":-5.8817482547056645,"crest_factor_db":19.68551418941668,"mean":0.0004496494153075736,"zero_energy":false},"rms_difference_db":-3.5177118307392767},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":9.92179057459183e-07,"output_rms_dbfs":-60.03409944126987,"output_zero_energy":false,"output_peak":0.007228464945152147,"output_energy_fraction":5.789768674113825e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":156,"both_estimated_voiced_frames":131,"reference_voiced_output_unvoiced_frames":123,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.515748031496063,"absolute_difference_cents_median":4.7658657615490725,"absolute_difference_cents_p95":1575.1071856448225,"difference_over_600_cents_frames":15,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00024330354027503592,"log_rms_mae_db":3.5178582162867147,"rms_envelope_nrmse":0.4283139905830449,"envelope_correlation":0.894447433407238,"rms_gain_error_db":-3.517740223175267,"reference_mean_power":0.006242903327296466,"output_mean_power":0.0027772347296477658},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.8103678297126455e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017429005529914874,"output_power":0.0006554483578970857,"reference_active":true,"delta_db":-4.247341279378194},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023587501721917026,"output_power":0.00011626890581592566,"reference_active":true,"delta_db":-3.0721835945310367},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.486758244411879e-05,"output_power":3.3806263441224484e-05,"reference_active":true,"delta_db":-2.830305410111581}],"active_band_count":3,"active_band_mae_db":3.3832767613402708}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00022924128639936837,"log_rms_mae_db":3.2128521618348773,"rms_envelope_nrmse":0.415792472506779,"envelope_correlation":0.9015839089832685,"rms_gain_error_db":-3.517727581964199,"reference_mean_power":0.006269846463979502,"output_mean_power":0.0027892288444719103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.5329682810308056e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015998664584984358,"output_power":0.0006032793749709178,"reference_active":true,"delta_db":-4.23565255603861},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010806179706711597,"output_power":5.2890024726998854e-05,"reference_active":true,"delta_db":-3.1029841551898407},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.749408392487692e-06,"output_power":2.6408222004508005e-06,"reference_active":true,"delta_db":-2.5490035303792165}],"active_band_count":3,"active_band_mae_db":3.295880080535889}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00021113569197140015,"log_rms_mae_db":3.3845198502641827,"rms_envelope_nrmse":0.48626252541932097,"envelope_correlation":0.8697368196560319,"rms_gain_error_db":-4.026229425634641,"reference_mean_power":0.004322448068392772,"output_mean_power":0.0017104360510242563},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.432095263902858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015007095196401276,"output_power":0.0005263102016172411,"reference_active":true,"delta_db":-4.550548493127332},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292924963191579,"output_power":0.00010041214935737011,"reference_active":true,"delta_db":-3.5860357911360112},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9206691853868204e-05,"output_power":3.0408931428592235e-05,"reference_active":true,"delta_db":-2.8936963647753595}],"active_band_count":3,"active_band_mae_db":3.6767602163462336}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001994421712728477,"log_rms_mae_db":3.2030641171240233,"rms_envelope_nrmse":0.4713667471810562,"envelope_correlation":0.8788184279315626,"rms_gain_error_db":-4.026353003689204,"reference_mean_power":0.004343287932440276,"output_mean_power":0.0017186336899902006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.2230459775767276e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014043103904193357,"output_power":0.0004976759386697454,"reference_active":true,"delta_db":-4.50516464932188},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010478951019837992,"output_power":4.7440529455721534e-05,"reference_active":true,"delta_db":-3.4416828311822996},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.688880762177371e-06,"output_power":2.375260709068764e-06,"reference_active":true,"delta_db":-2.9535790398671264}],"active_band_count":3,"active_band_mae_db":3.6334755067904356}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00011279772332784735,"log_rms_mae_db":3.239526427951895,"rms_envelope_nrmse":0.44912400143931097,"envelope_correlation":0.8819979570093005,"rms_gain_error_db":-2.912341183331036,"reference_mean_power":0.0007270906991063193,"output_mean_power":0.00037183859970480067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.440035526430572e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026825795040711167,"output_power":7.695533309001321e-05,"reference_active":true,"delta_db":-5.423138797889953},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8961769403929972e-05,"output_power":6.361749898365355e-06,"reference_active":true,"delta_db":-4.74302269231027},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.295679292221322e-06,"output_power":1.5982345920360867e-06,"reference_active":true,"delta_db":-5.202811499757165}],"active_band_count":3,"active_band_mae_db":5.1229909966524625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010728134925657118,"log_rms_mae_db":3.160963751806957,"rms_envelope_nrmse":0.4266758573282742,"envelope_correlation":0.8960019548346314,"rms_gain_error_db":-2.9124242880577382,"reference_mean_power":0.000730739020551997,"output_mean_power":0.00037369722247322646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.226272850414424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002461286506907591,"output_power":7.129235860502353e-05,"reference_active":true,"delta_db":-5.38119187886661},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.876024712000837e-06,"output_power":3.0150270964228464e-06,"reference_active":true,"delta_db":-4.689272831415263},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.47862005018424e-07,"output_power":9.290377598224995e-08,"reference_active":true,"delta_db":-5.733736300073968}],"active_band_count":3,"active_band_mae_db":5.268067003451947}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":3.244459011924254,"rms_envelope_nrmse":0.36652092129291025,"envelope_correlation":0.9318985825316481,"rms_gain_error_db":-2.2935963899732807,"reference_mean_power":3.954548405172447e-05,"output_mean_power":2.3320467739843826e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.251525458293286e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6611062793121062e-05,"output_power":1.0344859135387482e-05,"reference_active":true,"delta_db":-2.056728386441419},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.487374001807069e-06,"output_power":6.4669114891052425e-06,"reference_active":true,"delta_db":-1.1807642315996372},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1009284904576137e-06,"output_power":1.7796930272592536e-06,"reference_active":true,"delta_db":-0.7206617162482114}],"active_band_count":3,"active_band_mae_db":1.3193847780964225}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":3.152844501009534,"rms_envelope_nrmse":0.33728490257133475,"envelope_correlation":0.9455494646785465,"rms_gain_error_db":-2.293652952161562,"reference_mean_power":3.974390987159025e-05,"output_mean_power":2.343717668752465e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.932114585070644e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8381585740217762e-05,"output_power":1.1871564426563466e-05,"reference_active":true,"delta_db":-1.8987502057257073},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.490081779396795e-06,"output_power":3.7571488230689482e-06,"reference_active":true,"delta_db":-0.7739585294303708},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.644582556776442e-07,"output_power":1.610570319567075e-07,"reference_active":true,"delta_db":-0.09075988189959681}],"active_band_count":3,"active_band_mae_db":0.921156205685225}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.5299980640411377,"predicted_mos":3.5299980640411377,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"883f07fb1378d3a29576619e79b76c740c47570e760d01c9ea2e833afa271451","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.271680547997676},"started_at":"2026-10-13T07:39:07.464956+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003765674999158364,"native_input_samples":144000,"native_input_sha256":"e669d50a84441c33a4d34b0a29184934ef16bd8e1b09fe229e0e31066dab7953","output_sha256":"3be07e7963418cc5c7b62d279655a40cae5ff2fc46f166de0524f639d915976e","output_samples":144120,"length_delta_samples":120,"analysis_s":2.8724213739988045,"synthesis_s":3.5517072600014217,"total_s":6.424128634000226,"rtf":1.0706881056667044,"output_peak":0.5758779048919678,"output_rms":0.07549519284378964,"finished_at":"2026-10-13T07:39:13.905251+00:00","analysis_stages":{"external_f0_s":1.7329078399998252},"evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"3be07e7963418cc5c7b62d279655a40cae5ff2fc46f166de0524f639d915976e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.435333288954308,"output_peak":0.5715388959335563},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.614427052339334,"spectral_convergence":0.43259095547150106,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.052169011777905,"spectral_convergence":0.2817234460364995,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.878267263992345,"spectral_convergence":0.2637439167890466,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.0002498961576626011,"delta_db":0.171629800643678,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.002246501181369661,"delta_db":-0.2836602434525092,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.002503211079197331,"delta_db":-0.46695979484194344,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.00044491732974628634,"delta_db":-0.4677390543416615,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.00024160448852710684,"delta_db":-0.5074834141550139,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":2.6206102503443938e-05,"delta_db":-1.4180007798715746,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5525788478843968,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3761105508299111,"pitch":null,"output_rms_dbfs":-22.431862294751955,"output_peak":0.5715388959335563,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.181621109369861},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9080526362984699,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07892460423327397,"rms_dbfs":-22.055751743922045,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.02751756459009,"mean":1.2715657552083334e-08,"zero_energy":false},"output":{"rms":0.07557999959649664,"rms_dbfs":-22.431862294751955,"peak":0.5715388959335563,"peak_dbfs":-4.859084169258548,"crest_factor_db":17.572778125493404,"mean":-3.7342993645953396e-05,"zero_energy":false},"rms_difference_db":-0.3761105508299103},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":2.7819191043377657e-06,"output_rms_dbfs":-55.55655503046533,"output_zero_energy":false,"output_peak":0.03624501109840954,"output_energy_fraction":7.873198667654627e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":261,"both_estimated_voiced_frames":232,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.9098039215686274,"absolute_difference_cents_median":6.199531229056397,"absolute_difference_cents_p95":36.61804855843725,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024330354027503592,"log_rms_mae_db":2.0426716446590905,"rms_envelope_nrmse":0.1539381984156022,"envelope_correlation":0.9779454383370824,"rms_gain_error_db":-0.3760139572204595,"reference_mean_power":0.006233969602179056,"output_mean_power":0.005716935396274449},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8072421002875186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017414213578018352,"output_power":0.0016688048985777338,"reference_active":true,"delta_db":-0.1849830069166281},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002371798239966289,"output_power":0.00021687145091778428,"reference_active":true,"delta_db":-0.3887535744892592},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.501630810521967e-05,"output_power":5.974533358637565e-05,"reference_active":true,"delta_db":-0.36718314481309994}],"active_band_count":3,"active_band_mae_db":0.3136399087396624}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022924128639936837,"log_rms_mae_db":1.5258941467847082,"rms_envelope_nrmse":0.10395353367709706,"envelope_correlation":0.9901368853570265,"rms_gain_error_db":-0.37561682545378,"reference_mean_power":0.006262494274967633,"output_mean_power":0.005743619473702458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.530420990426951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015992997191427703,"output_power":0.0015388310549374503,"reference_active":true,"delta_db":-0.16738918816732642},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010914154654466237,"output_power":9.914797403896592e-05,"reference_active":true,"delta_db":-0.4170625909502763},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.77390165426219e-06,"output_power":4.7271033393098925e-06,"reference_active":true,"delta_db":-0.042783714626590016}],"active_band_count":3,"active_band_mae_db":0.2090784979147309}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00021113569197140015,"log_rms_mae_db":1.9017341791448634,"rms_envelope_nrmse":0.17441187532931052,"envelope_correlation":0.9753776221518886,"rms_gain_error_db":-0.36644142055224516,"reference_mean_power":0.0043153114305736935,"output_mean_power":0.003966139647447041},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.429798497423617e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014993372451038876,"output_power":0.00141074068905183,"reference_active":true,"delta_db":-0.26452136866037357},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023063830674655453,"output_power":0.00021582753644128707,"reference_active":true,"delta_db":-0.2882458746235767},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9359539965921096e-05,"output_power":5.78494329613884e-05,"reference_active":true,"delta_db":-0.11191420153405941}],"active_band_count":3,"active_band_mae_db":0.22156048160600325}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001994421712728477,"log_rms_mae_db":1.4714288934773847,"rms_envelope_nrmse":0.11598115470797134,"envelope_correlation":0.9890778330616729,"rms_gain_error_db":-0.3662321916006487,"reference_mean_power":0.004336654014858995,"output_mean_power":0.003985947329990751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.221204429766169e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001403505038426243,"output_power":0.00132285634746366,"reference_active":true,"delta_db":-0.2570129050228423},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010608827780695942,"output_power":9.827453657150767e-05,"reference_active":true,"delta_db":-0.3322639479591022},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.723378650149635e-06,"output_power":4.920600371908139e-06,"reference_active":true,"delta_db":0.1776533286928268}],"active_band_count":3,"active_band_mae_db":0.25564339389159046}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011279772332784735,"log_rms_mae_db":1.9920697817420152,"rms_envelope_nrmse":0.18186852339810064,"envelope_correlation":0.9770023937716383,"rms_gain_error_db":-0.4801567167227549,"reference_mean_power":0.000725883836364577,"output_mean_power":0.0006499073583993435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4338411984909043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026856860022608727,"output_power":0.00024250020240678126,"reference_active":true,"delta_db":-0.4434313020894961},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8973711977024913e-05,"output_power":1.5911276198323023e-05,"reference_active":true,"delta_db":-0.7644728908617284},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310520152307557e-06,"output_power":5.252688636664825e-06,"reference_active":true,"delta_db":-0.04755403318672656}],"active_band_count":3,"active_band_mae_db":0.418486075379317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010728134925657118,"log_rms_mae_db":1.372489977084116,"rms_envelope_nrmse":0.11766155010812775,"envelope_correlation":0.991206671167078,"rms_gain_error_db":-0.4801596664272857,"reference_mean_power":0.0007295243282152713,"output_mean_power":0.0006531663652839637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2203166741702414e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024656486610752886,"output_power":0.000223175125154627,"reference_active":true,"delta_db":-0.4328540553009416},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.896584068544471e-06,"output_power":7.3600445870637036e-06,"reference_active":true,"delta_db":-0.8234284169111393},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4772410320228666e-07,"output_power":3.7715118496700343e-07,"reference_active":true,"delta_db":0.3528068082844322}],"active_band_count":3,"active_band_mae_db":0.5363630934988377}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":2.120843920547822,"rms_envelope_nrmse":0.2952350993261809,"envelope_correlation":0.9436293494559893,"rms_gain_error_db":-1.039312411512143,"reference_mean_power":3.947976781678989e-05,"output_mean_power":3.107730490412913e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.253181223289244e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6595680637210912e-05,"output_power":1.2924055408569936e-05,"reference_active":true,"delta_db":-1.0859625751610986},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.516181974676607e-06,"output_power":6.666957286450302e-06,"reference_active":true,"delta_db":-1.0631726019889933},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.102920350990826e-06,"output_power":1.423653260091501e-06,"reference_active":true,"delta_db":-1.694185969428761}],"active_band_count":3,"active_band_mae_db":1.281107048859618}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":1.518807169305989,"rms_envelope_nrmse":0.19442879422614742,"envelope_correlation":0.9795684732007,"rms_gain_error_db":-1.0393504315951578,"reference_mean_power":3.967785924701113e-05,"output_mean_power":3.123296318735959e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.933756855113043e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8381327852746303e-05,"output_power":1.417155621295527e-05,"reference_active":true,"delta_db":-1.1295933748055231},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.503290667812916e-06,"output_power":3.539183086331767e-06,"reference_active":true,"delta_db":-1.0462694987528094},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.651394869607331e-07,"output_power":1.3332697701095924e-07,"reference_active":true,"delta_db":-0.9293289879141974}],"active_band_count":3,"active_band_mae_db":1.0350639538241768}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.553561210632324,"predicted_mos":2.553561210632324,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"af6c171f5bdfff3cd28e2992c74a630d55796efa97e4f71ac5a8e5ec775b1030","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4145437050010514},"started_at":"2026-10-13T05:10:50.798697+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032370560002163984,"native_input_samples":144000,"native_input_sha256":"e669d50a84441c33a4d34b0a29184934ef16bd8e1b09fe229e0e31066dab7953","output_sha256":"4e1d22ed5393ab8a588de73761d8e997ad97782a42a29b3e32d2fbc68a556038","output_samples":144240,"length_delta_samples":240,"analysis_s":1.3127468149996275,"synthesis_s":0.16745710599934682,"total_s":1.4802039209989744,"rtf":0.24670065349982906,"output_peak":0.5917748212814331,"output_rms":0.0751283322315909,"finished_at":"2026-10-13T05:10:52.290654+00:00","analysis_stages":{"external_f0_s":0.17229631999907724},"evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"4e1d22ed5393ab8a588de73761d8e997ad97782a42a29b3e32d2fbc68a556038","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.477544134847275,"output_peak":0.5921544202997115},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.551405987666952,"spectral_convergence":0.2827377243942349,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.5696552402859405,"spectral_convergence":0.29161241474699023,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.818104035796237,"spectral_convergence":0.3304282516112409,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024021305054769385,"output_power":0.00024043081281291574,"delta_db":0.00393526117080808,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0023981297439394636,"output_power":0.0020558581583407735,"delta_db":-0.6687952796766807,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027873624928200164,"output_power":0.0026572358831982172,"delta_db":-0.207633433084016,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004955109173134605,"output_power":0.0004189069915498358,"delta_db":-0.7293561870439164,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00027155220884703173,"output_power":0.00025802760041267783,"delta_db":-0.22187176177832407,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6324739911259635e-05,"output_power":3.144745883319413e-05,"delta_db":-0.6261695611422164,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4096269139826603,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4146210132633953,"pitch":null,"output_rms_dbfs":-22.470372757185437,"output_peak":0.5921544202997115,"reference_rms_dbfs":-22.055751743922045,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.64638842124971},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8979626324873736,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07892460423327397,"rms_dbfs":-22.055751743922045,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.02751756459009,"mean":1.2715657552083334e-08,"zero_energy":false},"output":{"rms":0.07524564376193225,"rms_dbfs":-22.470372757185437,"peak":0.5921544202997115,"peak_dbfs":-4.551300489226183,"crest_factor_db":17.91907226795925,"mean":-9.373776254086316e-06,"zero_energy":false},"rms_difference_db":-0.4146210132633925},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":1.822903255682395e-06,"output_rms_dbfs":-57.39236379418631,"output_zero_energy":false,"output_peak":0.02136303224624166,"output_energy_fraction":5.20500774631341e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":255,"output_estimated_voiced_frames":192,"both_estimated_voiced_frames":171,"reference_voiced_output_unvoiced_frames":84,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.6705882352941176,"absolute_difference_cents_median":1.3566274503364137,"absolute_difference_cents_p95":14.284839673317563,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024330354027503592,"log_rms_mae_db":0.9683685476914,"rms_envelope_nrmse":0.11524788545254706,"envelope_correlation":0.9887352527155518,"rms_gain_error_db":-0.41453671543843074,"reference_mean_power":0.006233969602179056,"output_mean_power":0.005666449318265342},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8072421002875186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017414213578018352,"output_power":0.0016292318048016694,"reference_active":true,"delta_db":-0.28920987085043504},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002371798239966289,"output_power":0.00019432427968894906,"reference_active":true,"delta_db":-0.8655067589054346},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.501630810521967e-05,"output_power":4.849948082511699e-05,"reference_active":true,"delta_db":-1.272852152113079}],"active_band_count":3,"active_band_mae_db":0.8091895939563162}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022924128639936837,"log_rms_mae_db":0.6388810092129702,"rms_envelope_nrmse":0.07529054060652875,"envelope_correlation":0.996144736014239,"rms_gain_error_db":-0.4137106879313678,"reference_mean_power":0.006262494274967633,"output_mean_power":0.005693459990099341},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.530420990426951e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015992997191427703,"output_power":0.0015053853720093065,"reference_active":true,"delta_db":-0.262821693442173},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010914154654466237,"output_power":9.074232125802822e-05,"reference_active":true,"delta_db":-0.8018021882180819},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.77390165426219e-06,"output_power":3.0902082372197756e-06,"reference_active":true,"delta_db":-1.8888572214878432}],"active_band_count":3,"active_band_mae_db":0.9844937010493661}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00021113569197140015,"log_rms_mae_db":1.1193841876698496,"rms_envelope_nrmse":0.13253518644078777,"envelope_correlation":0.9862501530528752,"rms_gain_error_db":-0.3490165140858236,"reference_mean_power":0.0043153114305736935,"output_mean_power":0.003982084689996681},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.429798497423617e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014993372451038876,"output_power":0.0014263556187447722,"reference_active":true,"delta_db":-0.21671512211860705},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023063830674655453,"output_power":0.00018713328212750351,"reference_active":true,"delta_db":-0.9078040623539524},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.9359539965921096e-05,"output_power":4.249391943511319e-05,"reference_active":true,"delta_db":-1.4516373634410198}],"active_band_count":3,"active_band_mae_db":0.8587188493045265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001994421712728477,"log_rms_mae_db":0.8089395142454142,"rms_envelope_nrmse":0.08110722221880926,"envelope_correlation":0.9953290558301564,"rms_gain_error_db":-0.3490995131676896,"reference_mean_power":0.004336654014858995,"output_mean_power":0.0040017027296252086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.221204429766169e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001403505038426243,"output_power":0.0013400634202314526,"reference_active":true,"delta_db":-0.20088623643676107},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010608827780695942,"output_power":8.290840893502538e-05,"reference_active":true,"delta_db":-1.0706881847304839},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.723378650149635e-06,"output_power":2.7434345057422613e-06,"reference_active":true,"delta_db":-2.359581655306083}],"active_band_count":3,"active_band_mae_db":1.2103853588244426}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011279772332784735,"log_rms_mae_db":1.1194219915269996,"rms_envelope_nrmse":0.13540216598828195,"envelope_correlation":0.9886164949754118,"rms_gain_error_db":-0.5253228256181284,"reference_mean_power":0.000725883836364577,"output_mean_power":0.0006431834239350268},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4338411984909043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026856860022608727,"output_power":0.00022476359671939458,"reference_active":true,"delta_db":-0.7732926247855227},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8973711977024913e-05,"output_power":1.3538274557105717e-05,"reference_active":true,"delta_db":-1.4658898628662345},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310520152307557e-06,"output_power":5.6796112389636195e-06,"reference_active":true,"delta_db":0.2918154867511968}],"active_band_count":3,"active_band_mae_db":0.8436659914676513}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010728134925657118,"log_rms_mae_db":0.7415141064799317,"rms_envelope_nrmse":0.0965899112145495,"envelope_correlation":0.9952434429957298,"rms_gain_error_db":-0.5253142189049366,"reference_mean_power":0.0007295243282152713,"output_mean_power":0.0006464104332421419},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2203166741702414e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024656486610752886,"output_power":0.00020777880052130456,"reference_active":true,"delta_db":-0.7432995770705658},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.896584068544471e-06,"output_power":6.465586455858392e-06,"reference_active":true,"delta_db":-1.3861536365009886},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4772410320228666e-07,"output_power":2.6078657194857944e-07,"reference_active":true,"delta_db":-1.2494957016901391}],"active_band_count":3,"active_band_mae_db":1.1263163050872311}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":1.3097763936833122,"rms_envelope_nrmse":0.11804261260308939,"envelope_correlation":0.9915147149116369,"rms_gain_error_db":-0.3707568276906637,"reference_mean_power":3.947976781678989e-05,"output_mean_power":3.624924012665504e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.253181223289244e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6595680637210912e-05,"output_power":1.5438467320701156e-05,"reference_active":true,"delta_db":-0.3139088595284113},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.516181974676607e-06,"output_power":7.707521890292378e-06,"reference_active":true,"delta_db":-0.4333016606678936},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.102920350990826e-06,"output_power":2.021389412485609e-06,"reference_active":true,"delta_db":-0.171728372596031}],"active_band_count":3,"active_band_mae_db":0.306312964264112}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":0.9622998148487677,"rms_envelope_nrmse":0.07931132898112908,"envelope_correlation":0.9966111115090474,"rms_gain_error_db":-0.37080159825050685,"reference_mean_power":3.967785924701113e-05,"output_mean_power":3.643074668617358e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.933756855113043e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8381327852746303e-05,"output_power":1.7282657750624662e-05,"reference_active":true,"delta_db":-0.26766351601995264},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.503290667812916e-06,"output_power":4.109591394173821e-06,"reference_active":true,"delta_db":-0.39731336417795726},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.651394869607331e-07,"output_power":1.663240795969581e-07,"reference_active":true,"delta_db":0.03104197790724496}],"active_band_count":3,"active_band_mae_db":0.2320062860350516}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_4992-41806-0000","method":"magphase","status":"measured","mos":{"status":"measured","value":3.145205020904541,"predicted_mos":3.145205020904541,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":286860,"input_sample_rate":48000,"input_seconds":5.97625,"input_float32_sha256":"aadb04b84907a34572851ac31eef5f61b87786ce5aee21c57324799c4cc298d8","internal_samples":95620,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4322289479987376},"started_at":"2026-10-13T05:16:05.345966+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008823284000754938,"native_input_samples":288000,"native_input_sha256":"4a3d30963879e1a915a6dd678d5a1d9f69971eb93095ee2e50cfd694faffcd4a","output_sha256":"bdba42ecb41176d42c3bf7b97aedba2bb890d1db6bef187c8bee0c8f97f191d2","output_samples":286860,"length_delta_samples":-1140,"analysis_s":3.704044004000025,"synthesis_s":1.1617095279998466,"total_s":4.8657535319998715,"rtf":0.8109589219999785,"output_peak":0.5903186365926014,"output_rms":0.07673736324573192,"finished_at":"2026-10-13T05:16:10.237106+00:00","evaluation_input":{"reference":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"output_sha256":"bdba42ecb41176d42c3bf7b97aedba2bb890d1db6bef187c8bee0c8f97f191d2","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95620,"length_delta_samples":-380,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.2937231963269,"output_peak":0.5699928284079351},"common_prefix":{"samples":95620,"sample_rate":16000,"reference_coverage":0.9960416666666667,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95620,"output_samples":95620,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.892205690700481,"spectral_convergence":0.4035977359671479,"spectral_convergence_status":"ok","amplitude_floor":1.726332693189446e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.948574855174854,"spectral_convergence":0.14725763040605505,"spectral_convergence_status":"ok","amplitude_floor":1.4419419879485463e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.116825382728283,"spectral_convergence":0.1453221543198937,"spectral_convergence_status":"ok","amplitude_floor":1.208114881083552e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00023794537677740494,"output_power":0.00021555479828206362,"delta_db":-0.4291957621965746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0024094545999073457,"output_power":0.002233358351960828,"delta_db":-0.32960334448961637,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00279629984676686,"output_power":0.002626828286408197,"delta_db":-0.2715205465264281,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004975947020814787,"output_power":0.0005047854742364591,"delta_db":0.06231101359235328,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002724319292602541,"output_power":0.0002820008447001524,"delta_db":0.14992403282386502,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.6468397641639184e-05,"output_power":3.4425434471442686e-05,"delta_db":-0.250372515408314,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24882120250619189,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.25265876499677664,"pitch":null,"output_rms_dbfs":-22.2937231963269,"output_peak":0.5699928284079351,"reference_rms_dbfs":-22.041064431330124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.319201976201207},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95620,"output_samples":95620,"intelligibility":{"estoi":{"value":0.9636566532262532,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0790581738496089,"rms_dbfs":-22.041064431330124,"peak":0.560516357421875,"peak_dbfs":-5.028234179331949,"crest_factor_db":17.012830251998178,"mean":-9.499960578004078e-06,"zero_energy":false},"output":{"rms":0.07679162187412858,"rms_dbfs":-22.2937231963269,"peak":0.5699928284079351,"peak_dbfs":-4.882612170846798,"crest_factor_db":17.411111025480103,"mean":-0.0009523829731988419,"zero_energy":false},"rms_difference_db":-0.2526587649967773},"full_scale":{"reference":{"sample_count":95620,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95620,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":597,"uncovered_tail_samples":100,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.919661271036603e-06,"threshold_dbfs":-52.27703143332556,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":97,"quiet_samples":15520,"reference_mean_power":1.27521577560026e-06,"output_mean_power":1.9988415519541084e-06,"output_rms_dbfs":-56.99221631014115,"output_zero_energy":false,"output_peak":0.034545973107288665,"output_energy_fraction":5.501658967462415e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":590,"reference_estimated_voiced_frames":254,"output_estimated_voiced_frames":250,"both_estimated_voiced_frames":232,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9133858267716536,"absolute_difference_cents_median":2.852947129636508,"absolute_difference_cents_p95":15.965616446151552,"difference_over_600_cents_frames":1,"uncovered_tail_samples":100,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95620,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1194,"rms_floor":0.00024330354027503592,"log_rms_mae_db":1.185562804775761,"rms_envelope_nrmse":0.08266855233430392,"envelope_correlation":0.9940756718561918,"rms_gain_error_db":-0.2523725438248343,"reference_mean_power":0.006255493008434519,"output_mean_power":0.005902340890819093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":2.814948387607504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017450619490264716,"output_power":0.0016431667238070683,"reference_active":true,"delta_db":-0.26129217522803955},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023378139414650348,"output_power":0.00022598718518145104,"reference_active":true,"delta_db":-0.14726131440300944},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4604135062635e-05,"output_power":6.39211020665307e-05,"reference_active":true,"delta_db":-0.046160623462851255}],"active_band_count":3,"active_band_mae_db":0.15157137103130008}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1188,"rms_floor":0.00022924128639936837,"log_rms_mae_db":0.7828333726149225,"rms_envelope_nrmse":0.04966784103861222,"envelope_correlation":0.9982308187247444,"rms_gain_error_db":-0.2498253115134193,"reference_mean_power":0.006275100972126789,"output_mean_power":0.00592431561450756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":2.5363163376538044e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016004608648692974,"output_power":0.0015048154001931074,"reference_active":true,"delta_db":-0.26761831952292775},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001064717826660421,"output_power":0.00010267692500661727,"reference_active":true,"delta_db":-0.1576167171930713},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.877441898258618e-06,"output_power":5.001073871933911e-06,"reference_active":true,"delta_db":0.10871165017638526}],"active_band_count":3,"active_band_mae_db":0.1779822289641281}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1194,"rms_floor":0.00021113569197140015,"log_rms_mae_db":1.076529509126209,"rms_envelope_nrmse":0.08683710983537503,"envelope_correlation":0.9946358383554955,"rms_gain_error_db":-0.3138511710889947,"reference_mean_power":0.004332859513824146,"output_mean_power":0.0040307836738035065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":2.4354139487332366e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015028977649290517,"output_power":0.0013807306565388376,"reference_active":true,"delta_db":-0.3682047092578092},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022700470827770596,"output_power":0.00020990986276381358,"reference_active":true,"delta_db":-0.34002020240704006},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.89960172399148e-05,"output_power":5.6794888291952824e-05,"reference_active":true,"delta_db":-0.1651344417946735}],"active_band_count":3,"active_band_mae_db":0.2911197844865076}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1188,"rms_floor":0.0001994421712728477,"log_rms_mae_db":0.746381336200424,"rms_envelope_nrmse":0.05606387971320819,"envelope_correlation":0.9982392469237281,"rms_gain_error_db":-0.3110992645373983,"reference_mean_power":0.004348956983569408,"output_mean_power":0.004048323276377374},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":2.225348219364892e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014053855232869972,"output_power":0.0012893196166239463,"reference_active":true,"delta_db":-0.3743488522841099},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010272026163253022,"output_power":9.498723538104112e-05,"reference_active":true,"delta_db":-0.3399086924494766},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.869022979129171e-06,"output_power":4.656966067863178e-06,"reference_active":true,"delta_db":-0.19338750647036038}],"active_band_count":3,"active_band_mae_db":0.3025483504013156}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1194,"rms_floor":0.00011279772332784735,"log_rms_mae_db":1.1057651988669919,"rms_envelope_nrmse":0.09424910000053768,"envelope_correlation":0.9937398424332868,"rms_gain_error_db":0.07062718458815859,"reference_mean_power":0.0007289068534813915,"output_mean_power":0.0007408576199980219},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":4.4492760140395043e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002677563029421533,"output_power":0.00027461921957784374,"reference_active":true,"delta_db":0.10991225770525873},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.895382263429738e-05,"output_power":2.0863236550632604e-05,"reference_active":true,"delta_db":0.41684869909205663},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.26523975308492e-06,"output_power":5.6383714962401415e-06,"reference_active":true,"delta_db":0.2973553530146301}],"active_band_count":3,"active_band_mae_db":0.27470543660398183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1188,"rms_floor":0.00010728134925657118,"log_rms_mae_db":0.6708009229275015,"rms_envelope_nrmse":0.04709422347544542,"envelope_correlation":0.9983921410573143,"rms_gain_error_db":0.07282304363624864,"reference_mean_power":0.0007320275599190766,"output_mean_power":0.0007444057796314795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":4.2339035759088943e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024544605988726254,"output_power":0.00025150591572527806,"reference_active":true,"delta_db":0.10592139826260863},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.84469125715886e-06,"output_power":9.883857088729015e-06,"reference_active":true,"delta_db":0.4824377975222066},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5247326837556554e-07,"output_power":3.460810252497857e-07,"reference_active":true,"delta_db":-0.07948396943856878}],"active_band_count":3,"active_band_mae_db":0.22261438840779468}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1194,"rms_floor":3.800534128114114e-05,"log_rms_mae_db":1.297767592115047,"rms_envelope_nrmse":0.20307777355841009,"envelope_correlation":0.9734970278417381,"rms_gain_error_db":0.21046963928070056,"reference_mean_power":3.9644365042928945e-05,"output_mean_power":4.161294288499018e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1194,"frequency_resolution_hz":0.16750418760469013,"power_floor":3.248976081445577e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.662729800599465e-05,"output_power":1.8087241731505457e-05,"reference_active":true,"delta_db":0.36550662279092416},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.451138348177121e-06,"output_power":1.0297920973335792e-05,"reference_active":true,"delta_db":0.8583434338985185},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0914516370970276e-06,"output_power":2.5227811589607673e-06,"reference_active":true,"delta_db":0.8143175249487031}],"active_band_count":3,"active_band_mae_db":0.6793891938793819}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95600,"uncovered_tail_samples":20,"metrics":{"frame_count":1188,"rms_floor":3.023522659258422e-05,"log_rms_mae_db":0.9048868552274775,"rms_envelope_nrmse":0.1329921480915862,"envelope_correlation":0.9879951766657923,"rms_gain_error_db":0.21535116862162695,"reference_mean_power":3.978045171217397e-05,"output_mean_power":4.180274752703146e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1188,"frequency_resolution_hz":0.16835016835016833,"power_floor":2.930751135270073e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.837628184244462e-05,"output_power":2.085500955100787e-05,"reference_active":true,"delta_db":0.5495274993617865},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4737237488793695e-06,"output_power":5.6969399614221215e-06,"reference_active":true,"delta_db":1.0497247953096354},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.64172576182332e-07,"output_power":2.1290913624089398e-07,"reference_active":true,"delta_db":1.1289368489379334}],"active_band_count":3,"active_band_mae_db":0.9093963812031184}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},{"case":"librispeech_5105-28240-0011","method":"hnm","status":"measured","mos":{"status":"measured","value":1.847975730895996,"predicted_mos":1.847975730895996,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2f70b189e89c566c067807793cd5ce6a4a2030a86bbc96a33f41e400f74e3272","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.374747030000435},"started_at":"2026-10-13T04:39:28.248442+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4573000044038054e-05,"native_input_samples":96000,"native_input_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","output_sha256":"e773ac33c8c9e897758df08d4a7848a2a27c71c0a823e1b82eba3e1a0fedefe2","output_samples":96000,"length_delta_samples":0,"analysis_s":1.088665602999754,"synthesis_s":0.4875298710003335,"total_s":1.5761954740000874,"rtf":0.26269924566668124,"output_peak":0.6080046910379034,"output_rms":0.05619727891117946,"finished_at":"2026-10-13T04:39:29.830658+00:00","evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"e773ac33c8c9e897758df08d4a7848a2a27c71c0a823e1b82eba3e1a0fedefe2","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.724667393326504,"spectral_convergence":0.2818451906158469,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.431968482554775,"spectral_convergence":0.26013019390797887,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.785406295056166,"spectral_convergence":0.28262286317665636,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0006183777693202412,"delta_db":0.14875432452051013,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005720776633917865,"delta_db":0.13090429049964056,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0010480352554090924,"delta_db":-0.00510775203302823,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0007170652823280852,"delta_db":-0.2595746229719518,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00019351135192113882,"delta_db":-0.31330005127226757,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":9.066834650550928e-06,"delta_db":0.17463796862918005,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1720465016544297,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.02975086253090796,"pitch":null,"output_rms_dbfs":-25.005694251775186,"output_peak":0.6080046910379034,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.647347390312482},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.724667393326504,"spectral_convergence":0.2818451906158469,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.431968482554775,"spectral_convergence":0.26013019390797887,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.785406295056166,"spectral_convergence":0.28262286317665636,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0006183777693202412,"delta_db":0.14875432452051013,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005720776633917865,"delta_db":0.13090429049964056,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0010480352554090924,"delta_db":-0.00510775203302823,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0007170652823280852,"delta_db":-0.2595746229719518,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00019351135192113882,"delta_db":-0.31330005127226757,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":9.066834650550928e-06,"delta_db":0.17463796862918005,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1720465016544297,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.02975086253090796,"pitch":null,"output_rms_dbfs":-25.005694251775186,"output_peak":0.6080046910379034,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.647347390312482},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8007084530632556,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05639009555779642,"rms_dbfs":-24.975943389244275,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.527562802178657,"mean":-2.918211619059245e-05,"zero_energy":false},"output":{"rms":0.05619727891117946,"rms_dbfs":-25.005694251775186,"peak":0.6080046910379034,"peak_dbfs":-4.321861398623705,"crest_factor_db":20.68383285315148,"mean":-0.00017990303942482867,"zero_energy":false},"rms_difference_db":-0.029750862530910638},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.7840253096655944e-06,"output_mean_power":4.595690783109762e-06,"output_rms_dbfs":-53.376492001114435,"output_zero_energy":false,"output_peak":0.014128053171468148,"output_energy_fraction":7.761022687640655e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":114,"both_estimated_voiced_frames":101,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9017857142857143,"absolute_difference_cents_median":1.935842662255806,"absolute_difference_cents_p95":6.646909399249581,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002040524003006053,"log_rms_mae_db":1.1973313748156176,"rms_envelope_nrmse":0.0726946213699829,"envelope_correlation":0.995013078426486,"rms_gain_error_db":-0.029751128589414012,"reference_mean_power":0.0031824932955688964,"output_mean_power":0.0031607662875854318},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.3108768872499092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006941576008860869,"output_power":0.0006892024774209436,"reference_active":true,"delta_db":-0.0311125355638307},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.971136227750716e-05,"output_power":6.330725363888226e-05,"reference_active":true,"delta_db":-0.4185009606430329},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.229406504274779e-05,"output_power":2.44014945804712e-05,"reference_active":true,"delta_db":0.3922716364870616}],"active_band_count":3,"active_band_mae_db":0.2806283775646417}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019294226027584624,"log_rms_mae_db":0.6802693768241425,"rms_envelope_nrmse":0.04345749126105846,"envelope_correlation":0.9981385574747237,"rms_gain_error_db":-0.0297409995086262,"reference_mean_power":0.0031984676882902833,"output_mean_power":0.0031766390313829255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1956491507193048e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000632939746621434,"output_power":0.0006277731916062726,"reference_active":true,"delta_db":-0.03559603184675187},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313010835721604e-05,"output_power":3.081292383686517e-05,"reference_active":true,"delta_db":-0.31489945857667007},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3632381122717e-06,"output_power":1.3740975064472416e-06,"reference_active":true,"delta_db":0.0344583227325359}],"active_band_count":3,"active_band_mae_db":0.1283179377186526}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015457760673902765,"log_rms_mae_db":1.7590703270978705,"rms_envelope_nrmse":0.10004561839924679,"envelope_correlation":0.9925201324163964,"rms_gain_error_db":0.013863523136714276,"reference_mean_power":0.0011955559481795221,"output_mean_power":0.0011993784928148639},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2205605502187216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004134156506493551,"output_power":0.00039508823261401603,"reference_active":true,"delta_db":-0.196928188989771},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135405678849972e-05,"output_power":3.9161343706636476e-05,"reference_active":true,"delta_db":-0.23660534872062322},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1249926033986768e-05,"output_power":1.362479956189215e-05,"reference_active":true,"delta_db":0.8318045491477677}],"active_band_count":3,"active_band_mae_db":0.42177936228605395}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014615313182534672,"log_rms_mae_db":1.2230027385596844,"rms_envelope_nrmse":0.06335122059051136,"envelope_correlation":0.9968761280525622,"rms_gain_error_db":0.013861104316339296,"reference_mean_power":0.0012015673286161978,"output_mean_power":0.0012054084220488432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.16593985496473e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003798908177653278,"output_power":0.000361372622580625,"reference_active":true,"delta_db":-0.21703548945025058},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8446225039409073e-05,"output_power":1.758803179367308e-05,"reference_active":true,"delta_db":-0.20690260602671912},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.60382048073901e-07,"output_power":7.834050863921947e-07,"reference_active":true,"delta_db":0.12954531934448496}],"active_band_count":3,"active_band_mae_db":0.18449447160715157}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011311095186173683,"log_rms_mae_db":1.624543740035956,"rms_envelope_nrmse":0.12482489368594929,"envelope_correlation":0.9883974707256952,"rms_gain_error_db":-0.2252143147864313,"reference_mean_power":0.0008539373095994977,"output_mean_power":0.000810782856753739},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.10125907860188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022995365445444387,"output_power":0.0002363894873453495,"reference_active":true,"delta_db":0.11987842970890766},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.837295392211272e-05,"output_power":4.4912361034967906e-05,"reference_active":true,"delta_db":-1.1384578458399441},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3078079511035852e-05,"output_power":7.515589498788443e-06,"reference_active":true,"delta_db":-2.4058092249450165}],"active_band_count":3,"active_band_mae_db":1.221381833497956}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":1.0412902499693053,"rms_envelope_nrmse":0.07615759888537918,"envelope_correlation":0.9957452554289289,"rms_gain_error_db":-0.2252161237844738,"reference_mean_power":0.0008582264011938478,"output_mean_power":0.000814854856015809},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.58513640339241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002121429018976754,"output_power":0.0002180294834732185,"reference_active":true,"delta_db":0.11886720656101898},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7723754282185392e-05,"output_power":2.1206966210808512e-05,"reference_active":true,"delta_db":-1.1637349658855776},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.120050210892593e-07,"output_power":5.344039307810879e-07,"reference_active":true,"delta_db":-2.3212758545465353}],"active_band_count":3,"active_band_mae_db":1.2012926756643774}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":1.6851687734112923,"rms_envelope_nrmse":0.2044435147640175,"envelope_correlation":0.9655211489887376,"rms_gain_error_db":-0.19262019766540406,"reference_mean_power":3.945170478466555e-05,"output_mean_power":3.7740161552034304e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.222824115041211e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1970652723962296e-05,"output_power":1.0548573656506374e-05,"reference_active":true,"delta_db":-0.5492409210410474},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.324380113074305e-06,"output_power":2.7555556009645658e-06,"reference_active":true,"delta_db":-0.8150149708963255},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0798067724676819e-06,"output_power":8.324961702767921e-07,"reference_active":true,"delta_db":-1.1296380270371822}],"active_band_count":3,"active_band_mae_db":0.8312979729915183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":1.2618725635886563,"rms_envelope_nrmse":0.1289552410615948,"envelope_correlation":0.9856579329379646,"rms_gain_error_db":-0.19262207610587972,"reference_mean_power":3.964923190102766e-05,"output_mean_power":3.792910289445893e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9761262616699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1283346332236671e-05,"output_power":9.587743700295959e-06,"reference_active":true,"delta_db":-0.7072150267744869},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7283331039583754e-06,"output_power":1.3530021289470037e-06,"reference_active":true,"delta_db":-1.0632896840826715},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.035315524247487e-08,"output_power":6.088868889957979e-08,"reference_active":true,"delta_db":-0.6274695711489331}],"active_band_count":3,"active_band_mae_db":0.7993247606686972}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28240-0011","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.333521842956543,"predicted_mos":3.333521842956543,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f0615c17cca54931b96d792b9f27022b90c4e77d6a3a8fbab618dd13deb5e8e1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3244469669989485},"started_at":"2026-10-13T04:42:26.134761+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5647999791544862e-05,"native_input_samples":96000,"native_input_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","output_sha256":"b2e969bf58c68c4ca51a1fa8a14718777739e8401893342139ec8a0612f9057e","output_samples":96000,"length_delta_samples":0,"analysis_s":1.9079649619998236,"synthesis_s":0.10894054100026551,"total_s":2.016905503000089,"rtf":0.3361509171666815,"output_peak":0.7346024154981453,"output_rms":0.05985241590450754,"finished_at":"2026-10-13T04:42:28.158171+00:00","evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"b2e969bf58c68c4ca51a1fa8a14718777739e8401893342139ec8a0612f9057e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.937084327866797,"spectral_convergence":0.7399000176256842,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.316641174279815,"spectral_convergence":0.261972007058337,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.7313964891721,"spectral_convergence":0.3063947392206354,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0005980238486264873,"delta_db":0.0034006773471330024,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0006389876058428047,"delta_db":0.6112787228069595,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0011840234917586947,"delta_db":0.5247365158641886,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.000922066129686528,"delta_db":0.8324591098728372,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00022954670690062992,"delta_db":0.4283461454844607,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":9.663906791002032e-06,"delta_db":0.45160833672837936,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4753049180173264,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5175770915517539,"pitch":null,"output_rms_dbfs":-24.45836629769252,"output_peak":0.7346024154981453,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.995040663772905},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.937084327866797,"spectral_convergence":0.7399000176256842,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.316641174279815,"spectral_convergence":0.261972007058337,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.7313964891721,"spectral_convergence":0.3063947392206354,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0005980238486264873,"delta_db":0.0034006773471330024,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0006389876058428047,"delta_db":0.6112787228069595,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0011840234917586947,"delta_db":0.5247365158641886,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.000922066129686528,"delta_db":0.8324591098728372,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00022954670690062992,"delta_db":0.4283461454844607,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":9.663906791002032e-06,"delta_db":0.45160833672837936,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4753049180173264,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5175770915517539,"pitch":null,"output_rms_dbfs":-24.45836629769252,"output_peak":0.7346024154981453,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.995040663772905},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8410229696458469,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05639009555779642,"rms_dbfs":-24.975943389244275,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.527562802178657,"mean":-2.918211619059245e-05,"zero_energy":false},"output":{"rms":0.059852415904507544,"rms_dbfs":-24.45836629769252,"peak":0.7346024154981453,"peak_dbfs":-2.6789529590796564,"crest_factor_db":21.77941333861287,"mean":-9.175445868453425e-09,"zero_energy":false},"rms_difference_db":0.5175770915517539},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.7840253096655944e-06,"output_mean_power":1.5293282706668248e-06,"output_rms_dbfs":-58.154992831703154,"output_zero_energy":false,"output_peak":0.006638557471058027,"output_energy_fraction":2.2768586740293236e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":125,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9196428571428571,"absolute_difference_cents_median":5.971233361296172,"absolute_difference_cents_p95":16.20417497170858,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002040524003006053,"log_rms_mae_db":3.157550666963025,"rms_envelope_nrmse":0.17312664302476083,"envelope_correlation":0.9773820384450447,"rms_gain_error_db":0.5175784952933622,"reference_mean_power":0.0031824932955688964,"output_mean_power":0.0035852987272351608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.3108768872499092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006941576008860869,"output_power":0.0007638098983745507,"reference_active":true,"delta_db":0.41527198729345083},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.971136227750716e-05,"output_power":7.470220390786385e-05,"reference_active":true,"delta_db":0.3002984514723561},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.229406504274779e-05,"output_power":3.37525856035041e-05,"reference_active":true,"delta_db":1.8011778355418782}],"active_band_count":3,"active_band_mae_db":0.8389160914358951}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019294226027584624,"log_rms_mae_db":2.659269442935345,"rms_envelope_nrmse":0.08472505452104537,"envelope_correlation":0.9972188286868553,"rms_gain_error_db":0.5176068613655708,"reference_mean_power":0.0031984676882902833,"output_mean_power":0.003603318520170473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1956491507193048e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000632939746621434,"output_power":0.0007029171859549848,"reference_active":true,"delta_db":0.45541792862068387},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313010835721604e-05,"output_power":3.579236375393415e-05,"reference_active":true,"delta_db":0.33567524150474365},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3632381122717e-06,"output_power":1.942250920213888e-06,"reference_active":true,"delta_db":1.537336166489586}],"active_band_count":3,"active_band_mae_db":0.7761431122050045}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015457760673902765,"log_rms_mae_db":2.338123836792257,"rms_envelope_nrmse":0.1845073267692129,"envelope_correlation":0.979833086760042,"rms_gain_error_db":0.6106081958843503,"reference_mean_power":0.0011955559481795221,"output_mean_power":0.0013760389403045241},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2205605502187216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004134156506493551,"output_power":0.00044478498798286597,"reference_active":true,"delta_db":0.3176320721076598},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135405678849972e-05,"output_power":4.5254299980897325e-05,"reference_active":true,"delta_db":0.39141731569953986},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1249926033986768e-05,"output_power":1.3825492280573167e-05,"reference_active":true,"delta_db":0.8953093700105756}],"active_band_count":3,"active_band_mae_db":0.5347862526059252}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014615313182534672,"log_rms_mae_db":1.7649083011396187,"rms_envelope_nrmse":0.0976844909223508,"envelope_correlation":0.9969409950584235,"rms_gain_error_db":0.6106055747495925,"reference_mean_power":0.0012015673286161978,"output_mean_power":0.0013829569734471654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.16593985496473e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003798908177653278,"output_power":0.00041087258080469953,"reference_active":true,"delta_db":0.3404836352387117},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8446225039409073e-05,"output_power":2.0073528343201904e-05,"reference_active":true,"delta_db":0.3671621293203578},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.60382048073901e-07,"output_power":9.01771355383013e-07,"reference_active":true,"delta_db":0.7406458087959491}],"active_band_count":3,"active_band_mae_db":0.4827638577850062}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011311095186173683,"log_rms_mae_db":1.6603251708814581,"rms_envelope_nrmse":0.20794336657253792,"envelope_correlation":0.9757159690721955,"rms_gain_error_db":0.752659570700454,"reference_mean_power":0.0008539373095994977,"output_mean_power":0.001015528102188249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.10125907860188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022995365445444387,"output_power":0.0002777647964972729,"reference_active":true,"delta_db":0.8203688721260195},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.837295392211272e-05,"output_power":6.702823152142148e-05,"reference_active":true,"delta_db":0.6004609005040018},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3078079511035852e-05,"output_power":1.2348623886905368e-05,"reference_active":true,"delta_db":-0.2492541029041849}],"active_band_count":3,"active_band_mae_db":0.5566946251780687}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":0.9153081906091675,"rms_envelope_nrmse":0.11818700256041927,"envelope_correlation":0.9957679898139686,"rms_gain_error_db":0.7526600955720817,"reference_mean_power":0.0008582264011938478,"output_mean_power":0.0010206289431171018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.58513640339241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002121429018976754,"output_power":0.00025735352928157336,"reference_active":true,"delta_db":0.8390162317221639},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7723754282185392e-05,"output_power":3.22012492256428e-05,"reference_active":true,"delta_db":0.6502067917681063},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.120050210892593e-07,"output_power":8.043945755312292e-07,"reference_active":true,"delta_db":-0.5452809611358}],"active_band_count":3,"active_band_mae_db":0.6781679948753568}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":1.5367807859791331,"rms_envelope_nrmse":0.21999154005490126,"envelope_correlation":0.962110007491715,"rms_gain_error_db":0.24609065209426584,"reference_mean_power":3.945170478466555e-05,"output_mean_power":4.175176504242108e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.222824115041211e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1970652723962296e-05,"output_power":1.2010427687647355e-05,"reference_active":true,"delta_db":0.014406409377601286},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.324380113074305e-06,"output_power":3.256328696484053e-06,"reference_active":true,"delta_db":-0.08982439162969404},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0798067724676819e-06,"output_power":7.660149452864782e-07,"reference_active":true,"delta_db":-1.4910880401594584}],"active_band_count":3,"active_band_mae_db":0.5317729470555846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":0.8386230661043242,"rms_envelope_nrmse":0.09904646021960989,"envelope_correlation":0.992285080917242,"rms_gain_error_db":0.24610474373084068,"reference_mean_power":3.964923190102766e-05,"output_mean_power":4.196094427044921e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9761262616699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1283346332236671e-05,"output_power":1.122530319964616e-05,"reference_active":true,"delta_db":-0.022398383373332345},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7283331039583754e-06,"output_power":1.6851927475314622e-06,"reference_active":true,"delta_db":-0.10977866957923375},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.035315524247487e-08,"output_power":4.6714060652764766e-08,"reference_active":true,"delta_db":-1.778359594354362}],"active_band_count":3,"active_band_mae_db":0.6368455491023094}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28240-0011","method":"facodec","status":"measured","mos":{"status":"measured","value":4.159567832946777,"predicted_mos":4.159567832946777,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"035be3bde753dba692d54b336d2c374c7d4ac92e03513a6d9d91467ae7237321","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7316708710004605},"started_at":"2026-10-13T04:53:56.475035+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2436000108427834e-05,"native_input_samples":96000,"native_input_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","output_sha256":"98c50655c1d89cd8bb762bb6a2d15daf8f77c7d431ea673f8bf704e697149348","output_samples":96000,"length_delta_samples":0,"analysis_s":1.521614903000227,"synthesis_s":6.048174316999393,"total_s":7.56978921999962,"rtf":1.2616315366666033,"output_peak":0.5382030010223389,"output_rms":0.053978029096871216,"finished_at":"2026-10-13T04:54:04.052245+00:00","evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"98c50655c1d89cd8bb762bb6a2d15daf8f77c7d431ea673f8bf704e697149348","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5148037487459,"spectral_convergence":0.24903283125208583,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.871348336102896,"spectral_convergence":0.26381617059421725,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.9906636109658145,"spectral_convergence":0.2894140203822752,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0005973001920314633,"delta_db":-0.001857814805925474,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005504609979570062,"delta_db":-0.03638008993776543,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.000892010632513731,"delta_db":-0.7051663648318812,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0006605581921524179,"delta_db":-0.616050753700519,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00020536517264052427,"delta_db":-0.055096575629409714,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":7.932437887533101e-06,"delta_db":-0.40585218647376964,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30340063089654506,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.37971614898367917,"pitch":null,"output_rms_dbfs":-25.355659538227954,"output_peak":0.5382030010223389,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.79227189860487},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5148037487459,"spectral_convergence":0.24903283125208583,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.871348336102896,"spectral_convergence":0.26381617059421725,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.9906636109658145,"spectral_convergence":0.2894140203822752,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0005973001920314633,"delta_db":-0.001857814805925474,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005504609979570062,"delta_db":-0.03638008993776543,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.000892010632513731,"delta_db":-0.7051663648318812,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0006605581921524179,"delta_db":-0.616050753700519,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00020536517264052427,"delta_db":-0.055096575629409714,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":7.932437887533101e-06,"delta_db":-0.40585218647376964,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30340063089654506,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.37971614898367917,"pitch":null,"output_rms_dbfs":-25.355659538227954,"output_peak":0.5382030010223389,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.79227189860487},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8255507470442588,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05639009555779642,"rms_dbfs":-24.975943389244275,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.527562802178657,"mean":-2.918211619059245e-05,"zero_energy":false},"output":{"rms":0.053978029096871216,"rms_dbfs":-25.355659538227954,"peak":0.5382030010223389,"peak_dbfs":-5.381077698750817,"crest_factor_db":19.97458183947714,"mean":-6.528695849561927e-05,"zero_energy":false},"rms_difference_db":-0.3797161489836789},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.7840253096655944e-06,"output_mean_power":2.974419994160368e-06,"output_rms_dbfs":-55.26597708218918,"output_zero_energy":false,"output_peak":0.006308732554316521,"output_energy_fraction":5.444612470405867e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":120,"both_estimated_voiced_frames":104,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9285714285714286,"absolute_difference_cents_median":2.7473627955516817,"absolute_difference_cents_p95":9.255492138093823,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002040524003006053,"log_rms_mae_db":1.3569285950592749,"rms_envelope_nrmse":0.11232023541991334,"envelope_correlation":0.9897889681729286,"rms_gain_error_db":-0.37971655762077516,"reference_mean_power":0.0031824932955688964,"output_mean_power":0.00291605587731006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.3108768872499092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006941576008860869,"output_power":0.0005378744661696046,"reference_active":true,"delta_db":-1.107771553971322},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.971136227750716e-05,"output_power":5.82195916736899e-05,"reference_active":true,"delta_db":-0.7823441450032292},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.229406504274779e-05,"output_power":2.2235844555018103e-05,"reference_active":true,"delta_db":-0.011356345068100716}],"active_band_count":3,"active_band_mae_db":0.633824014680884}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019294226027584624,"log_rms_mae_db":1.0230023997251387,"rms_envelope_nrmse":0.09990596170155287,"envelope_correlation":0.9920586848906469,"rms_gain_error_db":-0.37970528300879186,"reference_mean_power":0.0031984676882902833,"output_mean_power":0.0029307005068024165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1956491507193048e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000632939746621434,"output_power":0.00048520727859433725,"reference_active":true,"delta_db":-1.1543506176083347},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313010835721604e-05,"output_power":2.7770195950860925e-05,"reference_active":true,"delta_db":-0.7664391210141523},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3632381122717e-06,"output_power":1.2551744094723675e-06,"reference_active":true,"delta_db":-0.35867642972226604}],"active_band_count":3,"active_band_mae_db":0.7598220561149177}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015457760673902765,"log_rms_mae_db":1.673233043853014,"rms_envelope_nrmse":0.1273334943851882,"envelope_correlation":0.9911428512229654,"rms_gain_error_db":-0.6370896349878296,"reference_mean_power":0.0011955559481795221,"output_mean_power":0.0010324307735106023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2205605502187216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004134156506493551,"output_power":0.00031833613666813477,"reference_active":true,"delta_db":-1.1350097229115161},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135405678849972e-05,"output_power":3.56405548330423e-05,"reference_active":true,"delta_db":-0.6457366351901219},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1249926033986768e-05,"output_power":1.347685765770557e-05,"reference_active":true,"delta_db":0.784389743253122}],"active_band_count":3,"active_band_mae_db":0.8550453671182533}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014615313182534672,"log_rms_mae_db":1.2189273437807777,"rms_envelope_nrmse":0.11007763865684933,"envelope_correlation":0.994332416949254,"rms_gain_error_db":-0.6370875211395376,"reference_mean_power":0.0012015673286161978,"output_mean_power":0.0010376224485395087},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.16593985496473e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003798908177653278,"output_power":0.00029123436499559286,"reference_active":true,"delta_db":-1.154161770410441},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8446225039409073e-05,"output_power":1.6198197542412715e-05,"reference_active":true,"delta_db":-0.5644081159184868},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.60382048073901e-07,"output_power":8.472000494310453e-07,"reference_active":true,"delta_db":0.46954117420801184}],"active_band_count":3,"active_band_mae_db":0.7293703535123132}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011311095186173683,"log_rms_mae_db":1.573753454453557,"rms_envelope_nrmse":0.14752445508239995,"envelope_correlation":0.9850429093060863,"rms_gain_error_db":-0.5002892512716285,"reference_mean_power":0.0008539373095994977,"output_mean_power":0.0007610217405988366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.10125907860188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022995365445444387,"output_power":0.00019440437023108452,"reference_active":true,"delta_db":-0.7293429214478446},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.837295392211272e-05,"output_power":4.889542903832631e-05,"reference_active":true,"delta_db":-0.7694340978589326},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3078079511035852e-05,"output_power":1.2439043480067168e-05,"reference_active":true,"delta_db":-0.21756987545465997}],"active_band_count":3,"active_band_mae_db":0.5721156315871457}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":1.2057981249813414,"rms_envelope_nrmse":0.12763094075668519,"envelope_correlation":0.9887562814750994,"rms_gain_error_db":-0.5002842690988902,"reference_mean_power":0.0008582264011938478,"output_mean_power":0.0007648450203360315},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.58513640339241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002121429018976754,"output_power":0.0001774325667254001,"reference_active":true,"delta_db":-0.7759517019803931},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7723754282185392e-05,"output_power":2.333104390344911e-05,"reference_active":true,"delta_db":-0.7491787009197899},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.120050210892593e-07,"output_power":8.77112137543528e-07,"reference_active":true,"delta_db":-0.16942108515395735}],"active_band_count":3,"active_band_mae_db":0.5648504960180467}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":1.8248532909352917,"rms_envelope_nrmse":0.22281891452679156,"envelope_correlation":0.9624119421621318,"rms_gain_error_db":0.24809737836176854,"reference_mean_power":3.945170478466555e-05,"output_mean_power":4.177106156287977e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.222824115041211e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1970652723962296e-05,"output_power":1.3699758467428727e-05,"reference_active":true,"delta_db":0.5859507859354165},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.324380113074305e-06,"output_power":3.810310280777375e-06,"reference_active":true,"delta_db":0.5924966690392468},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0798067724676819e-06,"output_power":1.409748445668186e-06,"reference_active":true,"delta_db":1.1579557741864246}],"active_band_count":3,"active_band_mae_db":0.7788010763870293}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":1.4478012511121514,"rms_envelope_nrmse":0.16534785590303622,"envelope_correlation":0.9776767081008422,"rms_gain_error_db":0.24812528558111763,"reference_mean_power":3.964923190102766e-05,"output_mean_power":4.198047101400247e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9761262616699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1283346332236671e-05,"output_power":1.2934963724340445e-05,"reference_active":true,"delta_db":0.5932729646254807},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7283331039583754e-06,"output_power":1.9772053398143844e-06,"reference_active":true,"delta_db":0.5842432631593932},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.035315524247487e-08,"output_power":9.433631158772335e-08,"reference_active":true,"delta_db":1.27395312241293}],"active_band_count":3,"active_band_mae_db":0.8171564500659345}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28240-0011","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.0575127601623535,"predicted_mos":3.0575127601623535,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"0f619e857210c42dc2d2c897c4a057b9b9abe600cf5da0aa6701faa5996ecde3","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0693296480021672},"started_at":"2026-10-13T07:17:41.460085+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0066722669998853235,"native_input_samples":264600,"native_input_sha256":"37249343fac07f2427d1d2c8874ed7f94368bb0503dfbd8a41cd511054b45a82","output_sha256":"b3b5aa2323ea6b58b639f70edbd6e5fe556d6750a5938554d8d2738fda7b7dc8","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1482020500006911,"synthesis_s":4.2249676040009945,"total_s":4.373169654001686,"rtf":0.728861609000281,"output_peak":0.6524695754051208,"output_rms":0.05195021462839088,"finished_at":"2026-10-13T07:17:45.859471+00:00","evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"b3b5aa2323ea6b58b639f70edbd6e5fe556d6750a5938554d8d2738fda7b7dc8","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.682442421400225,"output_peak":0.6515965321973507},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.055061348110028,"spectral_convergence":0.46324825462527985,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.773751892681245,"spectral_convergence":0.40806924575404846,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.819648131277315,"spectral_convergence":0.4844928585996913,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0005109719369343563,"delta_db":-0.679813869300449,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005734087900652773,"delta_db":0.14099784031835955,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0008266664816765617,"delta_db":-1.0355633863636613,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0006253976074901564,"delta_db":-0.8535994288047817,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.0001628701848110904,"delta_db":-1.0619486299175043,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":4.192480608394735e-06,"delta_db":-3.175208362988312,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1578552529488446,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7047808531190414,"pitch":null,"output_rms_dbfs":-25.68072424236332,"output_peak":0.6515965321973507,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.882820457356196},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6372387877004093,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05639009555779642,"rms_dbfs":-24.975943389244275,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.527562802178657,"mean":-2.918211619059245e-05,"zero_energy":false},"output":{"rms":0.05199526403035027,"rms_dbfs":-25.68072424236332,"peak":0.6515965321973507,"peak_dbfs":-3.7204247135309245,"crest_factor_db":21.960299528832397,"mean":-0.0003316692313536677,"zero_energy":false},"rms_difference_db":-0.7047808531190434},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.7840253096655944e-06,"output_mean_power":3.7190488822205034e-06,"output_rms_dbfs":-54.29568113348715,"output_zero_energy":false,"output_peak":0.008148272352674566,"output_energy_fraction":7.336738480267794e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":122,"both_estimated_voiced_frames":99,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.8839285714285714,"absolute_difference_cents_median":11.56795831008808,"absolute_difference_cents_p95":29.66729572308417,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002040524003006053,"log_rms_mae_db":1.935738927558456,"rms_envelope_nrmse":0.12876508344526105,"envelope_correlation":0.9897790300094065,"rms_gain_error_db":-0.7047792537321671,"reference_mean_power":0.0031824932955688964,"output_mean_power":0.002705761868119474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.3108768872499092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006941576008860869,"output_power":0.0005993796072929501,"reference_active":true,"delta_db":-0.6375612060957432},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.971136227750716e-05,"output_power":6.487678022352703e-05,"reference_active":true,"delta_db":-0.31214281573631225},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.229406504274779e-05,"output_power":1.3666438698542031e-05,"reference_active":true,"delta_db":-2.1253390619513426}],"active_band_count":3,"active_band_mae_db":1.0250143612611327}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019294226027584624,"log_rms_mae_db":1.6079866674785057,"rms_envelope_nrmse":0.10392163852858391,"envelope_correlation":0.9954044548656557,"rms_gain_error_db":-0.704753932597168,"reference_mean_power":0.0031984676882902833,"output_mean_power":0.0027193591821944524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1956491507193048e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000632939746621434,"output_power":0.0005449021195657339,"reference_active":true,"delta_db":-0.6504387162528397},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313010835721604e-05,"output_power":3.1117172075035705e-05,"reference_active":true,"delta_db":-0.272227348333108},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3632381122717e-06,"output_power":7.532566372855083e-07,"reference_active":true,"delta_db":-2.5762875210677727}],"active_band_count":3,"active_band_mae_db":1.1663178618845735}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015457760673902765,"log_rms_mae_db":2.7991983822595907,"rms_envelope_nrmse":0.18267159143111447,"envelope_correlation":0.9762256992926639,"rms_gain_error_db":-0.54504347249077,"reference_mean_power":0.0011955559481795221,"output_mean_power":0.0010545460693532971},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2205605502187216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004134156506493551,"output_power":0.0003345477097668509,"reference_active":true,"delta_db":-0.9192885251032311},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135405678849972e-05,"output_power":3.2552639452373115e-05,"reference_active":true,"delta_db":-1.0393191174599405},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1249926033986768e-05,"output_power":7.55368143107496e-06,"reference_active":true,"delta_db":-1.7299100211313374}],"active_band_count":3,"active_band_mae_db":1.2295058878981697}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014615313182534672,"log_rms_mae_db":2.128804715622117,"rms_envelope_nrmse":0.15207965646952273,"envelope_correlation":0.9836738883696248,"rms_gain_error_db":-0.5450415874588113,"reference_mean_power":0.0012015673286161978,"output_mean_power":0.0010598488973825068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.16593985496473e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003798908177653278,"output_power":0.0003064158703443402,"reference_active":true,"delta_db":-0.9334754130537857},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8446225039409073e-05,"output_power":1.4324582261176667e-05,"reference_active":true,"delta_db":-1.0982553685390282},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.60382048073901e-07,"output_power":5.049489342291558e-07,"reference_active":true,"delta_db":-1.777843951249567}],"active_band_count":3,"active_band_mae_db":1.2698582442807937}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011311095186173683,"log_rms_mae_db":3.3300715284271702,"rms_envelope_nrmse":0.2038178053999963,"envelope_correlation":0.9777318345652635,"rms_gain_error_db":-1.1237343584012947,"reference_mean_power":0.0008539373095994977,"output_mean_power":0.0006592536652315724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.10125907860188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022995365445444387,"output_power":0.00020084359979272577,"reference_active":true,"delta_db":-0.5878231904161764},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.837295392211272e-05,"output_power":5.174999609209875e-05,"reference_active":true,"delta_db":-0.5230134969861813},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3078079511035852e-05,"output_power":7.911649751338825e-06,"reference_active":true,"delta_db":-2.1827692062077215}],"active_band_count":3,"active_band_mae_db":1.0978686312033596}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":2.884227479334542,"rms_envelope_nrmse":0.1703634600731951,"envelope_correlation":0.9876398182223348,"rms_gain_error_db":-1.1237149606373302,"reference_mean_power":0.0008582264011938478,"output_mean_power":0.0006625678739260091},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.58513640339241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002121429018976754,"output_power":0.0001868225409766515,"reference_active":true,"delta_db":-0.5519923062950048},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7723754282185392e-05,"output_power":2.3627654602619966e-05,"reference_active":true,"delta_db":-0.6943142744357578},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.120050210892593e-07,"output_power":5.178639052469385e-07,"reference_active":true,"delta_db":-2.457815873147775}],"active_band_count":3,"active_band_mae_db":1.2347074846261792}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":3.57255752070967,"rms_envelope_nrmse":0.3343696145699787,"envelope_correlation":0.9221879023813077,"rms_gain_error_db":-1.7573720839541132,"reference_mean_power":3.945170478466555e-05,"output_mean_power":2.6322586787537478e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.222824115041211e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1970652723962296e-05,"output_power":7.953521896848297e-06,"reference_active":true,"delta_db":-1.7755835076613202},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.324380113074305e-06,"output_power":2.172311435839062e-06,"reference_active":true,"delta_db":-1.8478858714262538},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0798067724676819e-06,"output_power":6.355685372329654e-07,"reference_active":true,"delta_db":-2.301836569674805}],"active_band_count":3,"active_band_mae_db":1.975101982920793}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":3.067383961592897,"rms_envelope_nrmse":0.28754185077648187,"envelope_correlation":0.946768699792606,"rms_gain_error_db":-1.7573787350371153,"reference_mean_power":3.964923190102766e-05,"output_mean_power":2.6454338415121397e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9761262616699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1283346332236671e-05,"output_power":7.462087491871351e-06,"reference_active":true,"delta_db":-1.7957758185464807},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7283331039583754e-06,"output_power":1.1220118368316804e-06,"reference_active":true,"delta_db":-1.8763000976930246},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.035315524247487e-08,"output_power":4.5491479326371956e-08,"reference_active":true,"delta_db":-1.8935351994271976}],"active_band_count":3,"active_band_mae_db":1.855203705222234}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28240-0011","method":"golf","status":"measured","mos":{"status":"measured","value":2.4084553718566895,"predicted_mos":2.4084553718566895,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"0e972bc2bc6423171fa71f36baa12f86d09c03b1d7e81df928035dd65523855d","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3913648769994325},"started_at":"2026-10-13T05:02:39.862503+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035241580008005258,"native_input_samples":144000,"native_input_sha256":"fa9e6475159900143180f9cd6f7bb35a3c1c4b662b7655e92d2cf54686842e2d","output_sha256":"c68243724526523a795d420b8213f8806d975d21cf8ce5646eb9880b1543a54d","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5466976239995347,"synthesis_s":0.1186193280000225,"total_s":0.6653169519995572,"rtf":0.11088615866659286,"output_peak":0.5377039909362793,"output_rms":0.05141265405126205,"finished_at":"2026-10-13T05:02:40.540738+00:00","analysis_stages":{"external_f0_s":0.08829876399977366},"evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"c68243724526523a795d420b8213f8806d975d21cf8ce5646eb9880b1543a54d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.77337291205101,"output_peak":0.5438945190714151},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.081017652390399,"spectral_convergence":0.7240947182358052,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.66778348028895,"spectral_convergence":0.29252663072784224,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.855067171561369,"spectral_convergence":0.3068905291691317,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005985532101403517,"output_power":0.0005222951347711066,"delta_db":-0.5918678397691989,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005560163200154676,"output_power":0.0005383136752719593,"delta_db":-0.1405212631004679,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010510200326091811,"output_power":0.0008727995131697448,"delta_db":-0.8069649848960482,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007625015038229956,"output_power":0.0005717115789361361,"delta_db":-1.2506371642900358,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002083340921723558,"output_power":0.0001362812927148653,"delta_db":-1.8432410003245094,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.723895543071297e-06,"output_power":5.042806634765815e-06,"delta_db":-2.3803814128618956,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1689356108736926,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8046705187860588,"pitch":null,"output_rms_dbfs":-25.77337291205101,"output_peak":0.5438945190714151,"reference_rms_dbfs":-24.968702393264948,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.201289434746905},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8315554660091625,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.056437124787708876,"rms_dbfs":-24.968702393264948,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.520321806199334,"mean":-2.8570745146533285e-05,"zero_energy":false},"output":{"rms":0.05144360019962228,"rms_dbfs":-25.77337291205101,"peak":0.5438945190714151,"peak_dbfs":-5.289706352618573,"crest_factor_db":20.483666559432436,"mean":0.0002491977655541893,"zero_energy":false},"rms_difference_db":-0.804670518786061},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":31,"quiet_samples":4960,"reference_mean_power":2.827294426219117e-06,"output_mean_power":4.789378262316557e-06,"output_rms_dbfs":-53.1972086127659,"output_zero_energy":false,"output_peak":0.007467891825493208,"output_energy_fraction":9.365938492370334e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":137,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9196428571428571,"absolute_difference_cents_median":3.8544353935447258,"absolute_difference_cents_p95":12.818071852828986,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002040524003006053,"log_rms_mae_db":2.211002421090573,"rms_envelope_nrmse":0.1724261801273688,"envelope_correlation":0.9775433212737993,"rms_gain_error_db":-0.8046747870373172,"reference_mean_power":0.003187808270241136,"output_mean_power":0.002648650860484027},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.3092505553447405e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006923947617658411,"output_power":0.0005795114805968436,"reference_active":true,"delta_db":-0.7729172949600615},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.942556900734112e-05,"output_power":6.48139573173316e-05,"reference_active":true,"delta_db":-0.2985090909537209},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2304716893330105e-05,"output_power":3.057935453503686e-05,"reference_active":true,"delta_db":1.3703159889164163}],"active_band_count":3,"active_band_mae_db":0.8139141249433995}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00019294226027584624,"log_rms_mae_db":1.492833336656719,"rms_envelope_nrmse":0.10685487069207626,"envelope_correlation":0.9961493893707839,"rms_gain_error_db":-0.8047678639675349,"reference_mean_power":0.0032038287185290363,"output_mean_power":0.0026619047033798774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.19414362993885e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006309549060205941,"output_power":0.0005295416745462177,"reference_active":true,"delta_db":-0.7609817713009238},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.307179302333063e-05,"output_power":3.024536309623477e-05,"reference_active":true,"delta_db":-0.3879893871176738},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3582944459862664e-06,"output_power":1.6854492617565667e-06,"reference_active":true,"delta_db":0.9372175830450618}],"active_band_count":3,"active_band_mae_db":0.6953962471545531}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015457760673902765,"log_rms_mae_db":1.8918744718686085,"rms_envelope_nrmse":0.18764503470700367,"envelope_correlation":0.9751184560360094,"rms_gain_error_db":-0.576515252490648,"reference_mean_power":0.0011975531957500757,"output_mean_power":0.0010486807325896215},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.2195564045769293e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041210993279475357,"output_power":0.00035567286542338163,"reference_active":true,"delta_db":-0.6396234837910313},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.12331570894854e-05,"output_power":3.8768726553333155e-05,"reference_active":true,"delta_db":-0.2676505274316144},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1253600354679287e-05,"output_power":1.9096231077829382e-05,"reference_active":true,"delta_db":2.296561731611343}],"active_band_count":3,"active_band_mae_db":1.0679452476113296}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014615313182534672,"log_rms_mae_db":1.1714830317013194,"rms_envelope_nrmse":0.10642011953804223,"envelope_correlation":0.993943820693421,"rms_gain_error_db":-0.576516375177652,"reference_mean_power":0.0012035847015287148,"output_mean_power":0.0010539621661310593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1649873112506631e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003785795973708492,"output_power":0.0003274178612055237,"reference_active":true,"delta_db":-0.6305483778020242},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8400052102794972e-05,"output_power":1.740453781809196e-05,"reference_active":true,"delta_db":-0.2415655781451553},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.567925066128386e-07,"output_power":1.2193344196529908e-06,"reference_active":true,"delta_db":2.0714601009439084}],"active_band_count":3,"active_band_mae_db":0.9811913522970294}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00011311095186173683,"log_rms_mae_db":2.106083211937034,"rms_envelope_nrmse":0.2358318608808702,"envelope_correlation":0.970032953059931,"rms_gain_error_db":-1.3844105353908307,"reference_mean_power":0.0008553638494417104,"output_mean_power":0.000621884651118556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.095840151397564e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022920082767244302,"output_power":0.00017072835231479042,"reference_active":true,"delta_db":-1.2791053258101284},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.817203945007139e-05,"output_power":4.487015640053802e-05,"reference_active":true,"delta_db":-1.1275670664628077},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3091321992419355e-05,"output_power":1.0517328101039528e-05,"reference_active":true,"delta_db":-0.9507808235811555}],"active_band_count":3,"active_band_mae_db":1.1191510719513638}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":1.5683656045260923,"rms_envelope_nrmse":0.1688691721963158,"envelope_correlation":0.9935846179540725,"rms_gain_error_db":-1.3844548381400008,"reference_mean_power":0.0008596673959037575,"output_mean_power":0.000625007130373739},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.580043668179475e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021130053959696288,"output_power":0.00015842110672235535,"reference_active":true,"delta_db":-1.2508756318392504},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7684120345852087e-05,"output_power":2.1462104590324322e-05,"reference_active":true,"delta_db":-1.1055842153185254},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.088199799404007e-07,"output_power":6.899871728006308e-07,"reference_active":true,"delta_db":-1.1963684914773811}],"active_band_count":3,"active_band_mae_db":1.1842761128783856}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":2.6623900507741807,"rms_envelope_nrmse":0.23907362467460008,"envelope_correlation":0.9649827136428083,"rms_gain_error_db":-1.297177420795762,"reference_mean_power":3.9517499478566866e-05,"output_mean_power":2.931377257583581e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.220198285610132e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1940865548185954e-05,"output_power":8.909382011003255e-06,"reference_active":true,"delta_db":-1.2718822755266233},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3220537121092136e-06,"output_power":2.523465246834782e-06,"reference_active":true,"delta_db":-1.1940932198801644},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0788480777649393e-06,"output_power":8.012244961141155e-07,"reference_active":true,"delta_db":-1.2920607344284925}],"active_band_count":3,"active_band_mae_db":1.2526787432784268}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":2.2643642967079893,"rms_envelope_nrmse":0.17104402536273366,"envelope_correlation":0.9901467226375193,"rms_gain_error_db":-1.2971638391168385,"reference_mean_power":3.971569876050523e-05,"output_mean_power":2.9460887388519175e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.97348668923332e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1247513910888758e-05,"output_power":8.373716525893649e-06,"reference_active":true,"delta_db":-1.2813828427056135},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7301897082506092e-06,"output_power":1.2911678335827533e-06,"reference_active":true,"delta_db":-1.2711102627499131},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.015825622430294e-08,"output_power":4.86955682846077e-08,"reference_active":true,"delta_db":-1.585893476696264}],"active_band_count":3,"active_band_mae_db":1.379462194050597}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28240-0011","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.778398036956787,"predicted_mos":2.778398036956787,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"44c0c0f8b1c749163f038068140ec772d9238c275c0e36b9f225d6ff36c1d802","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.29100476199892},"started_at":"2026-10-13T07:39:22.287673+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031839059993217234,"native_input_samples":144000,"native_input_sha256":"fa9e6475159900143180f9cd6f7bb35a3c1c4b662b7655e92d2cf54686842e2d","output_sha256":"584161cb4b199117affe6b4516a1461abf567e6a89052634f6fbf9e03f459f9c","output_samples":144120,"length_delta_samples":120,"analysis_s":2.961372914000094,"synthesis_s":14.968886709000799,"total_s":17.930259623000893,"rtf":2.988376603833482,"output_peak":0.6243833899497986,"output_rms":0.056292370243854105,"finished_at":"2026-10-13T07:39:40.230996+00:00","analysis_stages":{"external_f0_s":2.080003497998405},"evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"584161cb4b199117affe6b4516a1461abf567e6a89052634f6fbf9e03f459f9c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.986180797413716,"output_peak":0.6052227915730968},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.5434791668757875,"spectral_convergence":0.616848011690128,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.359153947443426,"spectral_convergence":0.31339347695220854,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.518699269273018,"spectral_convergence":0.31934110213619366,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0007792991100014789,"delta_db":1.15325744400504,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005452565384021429,"delta_db":-0.07763680052227517,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0009558810933916526,"delta_db":-0.40482795883235956,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0006897957558228011,"delta_db":-0.4279564045008317,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00019775416386116963,"delta_db":-0.21910815507749135,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":7.012652096025701e-06,"delta_db":-0.9410960513284141,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5373138023777352,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.006620253406500343,"pitch":null,"output_rms_dbfs":-24.982563642650778,"output_peak":0.6052227915730968,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.807110794530743},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8207284821907431,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05639009555779642,"rms_dbfs":-24.975943389244275,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.527562802178657,"mean":-2.918211619059245e-05,"zero_energy":false},"output":{"rms":0.05634713225688838,"rms_dbfs":-24.982563642650778,"peak":0.6052227915730968,"peak_dbfs":-4.361694512248251,"crest_factor_db":20.620869130402525,"mean":2.7001395078402788e-05,"zero_energy":false},"rms_difference_db":-0.006620253406502741},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.7840253096655944e-06,"output_mean_power":1.9473150416432273e-06,"output_rms_dbfs":-57.10563781515915,"output_zero_energy":false,"output_peak":0.025312340324464548,"output_energy_fraction":3.271081092109658e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":141,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":38,"paired_voiced_fraction":0.9196428571428571,"absolute_difference_cents_median":6.194561822108033,"absolute_difference_cents_p95":16.201744569163797,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002040524003006053,"log_rms_mae_db":3.855236533599904,"rms_envelope_nrmse":0.15899708312957542,"envelope_correlation":0.9762316551938767,"rms_gain_error_db":-0.006619935009732916,"reference_mean_power":0.0031824932955688964,"output_mean_power":0.0031776459279514464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.3108768872499092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006941576008860869,"output_power":0.0006406871116232031,"reference_active":true,"delta_db":-0.3481209586026358},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.971136227750716e-05,"output_power":6.420932241380568e-05,"reference_active":true,"delta_db":-0.3570548273453661},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.229406504274779e-05,"output_power":2.0195369335909443e-05,"reference_active":true,"delta_db":-0.4293746382412086}],"active_band_count":3,"active_band_mae_db":0.3781834747297368}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019294226027584624,"log_rms_mae_db":3.3140805345812026,"rms_envelope_nrmse":0.08949060014412266,"envelope_correlation":0.9920796069002159,"rms_gain_error_db":-0.006587926900166102,"reference_mean_power":0.0031984676882902833,"output_mean_power":0.003193619526847691},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1956491507193048e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000632939746621434,"output_power":0.0005813197720857531,"reference_active":true,"delta_db":-0.36947274246639744},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313010835721604e-05,"output_power":2.9792675763126214e-05,"reference_active":true,"delta_db":-0.4611334614967282},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3632381122717e-06,"output_power":1.2351737857939917e-06,"reference_active":true,"delta_db":-0.4284365321644421}],"active_band_count":3,"active_band_mae_db":0.4196809120425226}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015457760673902765,"log_rms_mae_db":2.2046938996289267,"rms_envelope_nrmse":0.18653966543169925,"envelope_correlation":0.9737278752282118,"rms_gain_error_db":-0.2766947164949513,"reference_mean_power":0.0011955559481795221,"output_mean_power":0.0011217612478158286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2205605502187216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004134156506493551,"output_power":0.0003721330426074287,"reference_active":true,"delta_db":-0.45688679852349723},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135405678849972e-05,"output_power":4.032912385403602e-05,"reference_active":true,"delta_db":-0.10899332691769714},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1249926033986768e-05,"output_power":1.3654384469979813e-05,"reference_active":true,"delta_db":0.8412246017069143}],"active_band_count":3,"active_band_mae_db":0.46903490904936956}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014615313182534672,"log_rms_mae_db":1.2026853079701838,"rms_envelope_nrmse":0.09528361246726484,"envelope_correlation":0.9934048359128362,"rms_gain_error_db":-0.2766940058321279,"reference_mean_power":0.0012015673286161978,"output_mean_power":0.0011274017652592608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.16593985496473e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003798908177653278,"output_power":0.0003423880082236711,"reference_active":true,"delta_db":-0.45140250902638124},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8446225039409073e-05,"output_power":1.8011066289940585e-05,"reference_active":true,"delta_db":-0.1036807796647455},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.60382048073901e-07,"output_power":9.918633519934292e-07,"reference_active":true,"delta_db":1.154199889475249}],"active_band_count":3,"active_band_mae_db":0.569761059388792}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011311095186173683,"log_rms_mae_db":2.3023158043299334,"rms_envelope_nrmse":0.2162996341174899,"envelope_correlation":0.9644139549075893,"rms_gain_error_db":-0.3924247602450703,"reference_mean_power":0.0008539373095994977,"output_mean_power":0.0007801597003966933},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.10125907860188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022995365445444387,"output_power":0.00020249424608350766,"reference_active":true,"delta_db":-0.5522762867410989},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.837295392211272e-05,"output_power":5.352648094079733e-05,"reference_active":true,"delta_db":-0.3764297908181064},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3078079511035852e-05,"output_power":1.0270952730632383e-05,"reference_active":true,"delta_db":-1.0493324292525366}],"active_band_count":3,"active_band_mae_db":0.6593461689372473}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":1.3518031173804694,"rms_envelope_nrmse":0.11599843676134444,"envelope_correlation":0.9902777410343914,"rms_gain_error_db":-0.3924156200367528,"reference_mean_power":0.0008582264011938478,"output_mean_power":0.0007840798775887209},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.58513640339241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002121429018976754,"output_power":0.00018695074442665372,"reference_active":true,"delta_db":-0.5490130643183982},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7723754282185392e-05,"output_power":2.5700903257516422e-05,"reference_active":true,"delta_db":-0.3290365413919502},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.120050210892593e-07,"output_power":7.711153556789843e-07,"reference_active":true,"delta_db":-0.7287787778951977}],"active_band_count":3,"active_band_mae_db":0.5356094612018487}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":1.9099936441586902,"rms_envelope_nrmse":0.2471895860306653,"envelope_correlation":0.9500926440726972,"rms_gain_error_db":-0.5579798079443323,"reference_mean_power":3.945170478466555e-05,"output_mean_power":3.4695072062170256e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.222824115041211e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1970652723962296e-05,"output_power":1.0051735139616507e-05,"reference_active":true,"delta_db":-0.75876795299546},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.324380113074305e-06,"output_power":2.837182640474216e-06,"reference_active":true,"delta_db":-0.6882338171371807},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0798067724676819e-06,"output_power":7.166425483034588e-07,"reference_active":true,"delta_db":-1.7804345759881703}],"active_band_count":3,"active_band_mae_db":1.0758121153736038}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":1.1158281720507872,"rms_envelope_nrmse":0.1317880092360291,"envelope_correlation":0.9872199812053916,"rms_gain_error_db":-0.5579681918440267,"reference_mean_power":3.964923190102766e-05,"output_mean_power":3.486887689500086e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9761262616699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1283346332236671e-05,"output_power":9.33153830583558e-06,"reference_active":true,"delta_db":-0.8248467545298228},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7283331039583754e-06,"output_power":1.4620554021533817e-06,"reference_active":true,"delta_db":-0.7266361856878176},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.035315524247487e-08,"output_power":4.840813785251202e-08,"reference_active":true,"delta_db":-1.6236520304669657}],"active_band_count":3,"active_band_mae_db":1.0583783235615354}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28240-0011","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.160325050354004,"predicted_mos":3.160325050354004,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"836a207507160ee3ca00360c74691a64dd17f912e6a9f9559d291eead9d6095d","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.225571906999903},"started_at":"2026-10-13T05:10:52.294728+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0041014939997694455,"native_input_samples":144000,"native_input_sha256":"fa9e6475159900143180f9cd6f7bb35a3c1c4b662b7655e92d2cf54686842e2d","output_sha256":"019ded90cc94493dcee4f4d41fe715d3ee4c56a1edb00c8f5566845117c22679","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9698109239998303,"synthesis_s":0.12865324100130238,"total_s":1.0984641650011326,"rtf":0.1830773608335221,"output_peak":0.6053053736686707,"output_rms":0.054968092086204026,"finished_at":"2026-10-13T05:10:53.405634+00:00","analysis_stages":{"external_f0_s":0.1797459239987802},"evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"019ded90cc94493dcee4f4d41fe715d3ee4c56a1edb00c8f5566845117c22679","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.192651173726524,"output_peak":0.607878019943217},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.819439582800983,"spectral_convergence":0.21136498534587245,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.46077627084972,"spectral_convergence":0.18312231505109156,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.050177185704898,"spectral_convergence":0.19943055208022945,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0005885157704200305,"delta_db":-0.06620330755703502,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005413518718402017,"delta_db":-0.10884919359206166,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0009992923500908657,"delta_db":-0.21194104799664107,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0007019649326552625,"delta_db":-0.3520074129529522,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00019174198213560604,"delta_db":-0.3531923958326221,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":7.2408579805846035e-06,"delta_db":-0.8020186934796099,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.315702008568487,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20947586709964428,"pitch":null,"output_rms_dbfs":-25.18541925634392,"output_peak":0.607878019943217,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.7767976797852},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8976685501066003,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05639009555779642,"rms_dbfs":-24.975943389244275,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.527562802178657,"mean":-2.918211619059245e-05,"zero_energy":false},"output":{"rms":0.05504641464366731,"rms_dbfs":-25.18541925634392,"peak":0.607878019943217,"peak_dbfs":-4.323671196793912,"crest_factor_db":20.861748059550006,"mean":7.544339588408474e-05,"zero_energy":false},"rms_difference_db":-0.20947586709964483},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.7840253096655944e-06,"output_mean_power":4.067995192906011e-06,"output_rms_dbfs":-53.90619568948808,"output_zero_energy":false,"output_peak":0.011562461993022572,"output_energy_fraction":7.160132920648092e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":110,"both_estimated_voiced_frames":101,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9017857142857143,"absolute_difference_cents_median":0.7105588891724977,"absolute_difference_cents_p95":7.34844205044295,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002040524003006053,"log_rms_mae_db":1.125587440621743,"rms_envelope_nrmse":0.0634582273070686,"envelope_correlation":0.9966799868269614,"rms_gain_error_db":-0.2094762939397188,"reference_mean_power":0.0031824932955688964,"output_mean_power":0.0030326330804814157},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.3108768872499092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006941576008860869,"output_power":0.0006798587658943376,"reference_active":true,"delta_db":-0.09039381875194116},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.971136227750716e-05,"output_power":5.9910610997807816e-05,"reference_active":true,"delta_db":-0.6579982090437174},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.229406504274779e-05,"output_power":1.729845001563502e-05,"reference_active":true,"delta_db":-1.101820728438167}],"active_band_count":3,"active_band_mae_db":0.6167375854112752}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019294226027584624,"log_rms_mae_db":0.6353002059087417,"rms_envelope_nrmse":0.038798704751434,"envelope_correlation":0.9990749750640259,"rms_gain_error_db":-0.2094699298963127,"reference_mean_power":0.0031984676882902833,"output_mean_power":0.003047859722341221},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1956491507193048e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000632939746621434,"output_power":0.0006222718488400744,"reference_active":true,"delta_db":-0.07382214617114863},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313010835721604e-05,"output_power":2.9127285012919604e-05,"reference_active":true,"delta_db":-0.559228508766889},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3632381122717e-06,"output_power":1.0182643869805948e-06,"reference_active":true,"delta_db":-1.2671116431916885}],"active_band_count":3,"active_band_mae_db":0.6333874327099087}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015457760673902765,"log_rms_mae_db":1.0909667137178929,"rms_envelope_nrmse":0.08338047757510848,"envelope_correlation":0.9950754480602236,"rms_gain_error_db":-0.20458395450271807,"reference_mean_power":0.0011955559481795221,"output_mean_power":0.0011405425946698143},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2205605502187216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004134156506493551,"output_power":0.000406550872512431,"reference_active":true,"delta_db":-0.07272016204939832},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135405678849972e-05,"output_power":3.6043407737124365e-05,"reference_active":true,"delta_db":-0.5969227502173529},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1249926033986768e-05,"output_power":8.540593681690522e-06,"reference_active":true,"delta_db":-1.1966160623719848}],"active_band_count":3,"active_band_mae_db":0.6220863248795787}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014615313182534672,"log_rms_mae_db":0.5405698650521407,"rms_envelope_nrmse":0.04835904914085083,"envelope_correlation":0.9985632476295854,"rms_gain_error_db":-0.20458340613916975,"reference_mean_power":0.0012015673286161978,"output_mean_power":0.0011462775069423857},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.16593985496473e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003798908177653278,"output_power":0.00037461831402752725,"reference_active":true,"delta_db":-0.06069791486731799},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8446225039409073e-05,"output_power":1.6480807597354806e-05,"reference_active":true,"delta_db":-0.4892901331217622},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.60382048073901e-07,"output_power":5.3709296300514e-07,"reference_active":true,"delta_db":-1.5098239273397152}],"active_band_count":3,"active_band_mae_db":0.6866039917762651}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011311095186173683,"log_rms_mae_db":1.1630375670622188,"rms_envelope_nrmse":0.09524941568110204,"envelope_correlation":0.9940506695406884,"rms_gain_error_db":-0.3365858466008247,"reference_mean_power":0.0008539373095994977,"output_mean_power":0.0007902552763158602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.10125907860188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022995365445444387,"output_power":0.00022420113275641983,"reference_active":true,"delta_db":-0.11002513332284884},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.837295392211272e-05,"output_power":4.7210062773294155e-05,"reference_active":true,"delta_db":-0.9217709313653362},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3078079511035852e-05,"output_power":9.236939530651974e-06,"reference_active":true,"delta_db":-1.5101587288641913}],"active_band_count":3,"active_band_mae_db":0.8473182645174587}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":0.7424593938104278,"rms_envelope_nrmse":0.06040287416054002,"envelope_correlation":0.9982373035713102,"rms_gain_error_db":-0.3365805131290916,"reference_mean_power":0.0008582264011938478,"output_mean_power":0.0007942254859898748},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.58513640339241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002121429018976754,"output_power":0.00020728720133629975,"reference_active":true,"delta_db":-0.10056017255076263},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7723754282185392e-05,"output_power":2.2754140183893803e-05,"reference_active":true,"delta_db":-0.8579161164236695},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.120050210892593e-07,"output_power":6.576595245870695e-07,"reference_active":true,"delta_db":-1.4199611517701405}],"active_band_count":3,"active_band_mae_db":0.792812480248191}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":1.0006388378488111,"rms_envelope_nrmse":0.12388185477824346,"envelope_correlation":0.9892758414239141,"rms_gain_error_db":-0.4793354262627477,"reference_mean_power":3.945170478466555e-05,"output_mean_power":3.532907217883095e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.222824115041211e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1970652723962296e-05,"output_power":1.1739806676952793e-05,"reference_active":true,"delta_db":-0.08456886525648731},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.324380113074305e-06,"output_power":2.7795219656002293e-06,"reference_active":true,"delta_db":-0.7774056516615447},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0798067724676819e-06,"output_power":8.433472469986334e-07,"reference_active":true,"delta_db":-1.0733961543775634}],"active_band_count":3,"active_band_mae_db":0.6451235570985318}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":0.7100060198732413,"rms_envelope_nrmse":0.08093385831958191,"envelope_correlation":0.9966115523648061,"rms_gain_error_db":-0.47932436855382843,"reference_mean_power":3.964923190102766e-05,"output_mean_power":3.550604846769022e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9761262616699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1283346332236671e-05,"output_power":1.0992338726420401e-05,"reference_active":true,"delta_db":-0.11347816020369035},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7283331039583754e-06,"output_power":1.4471730500728151e-06,"reference_active":true,"delta_db":-0.7710698207424225},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.035315524247487e-08,"output_power":5.641138227384655e-08,"reference_active":true,"delta_db":-0.9591683814744566}],"active_band_count":3,"active_band_mae_db":0.6145721208068565}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28240-0011","method":"magphase","status":"measured","mos":{"status":"measured","value":2.8830525875091553,"predicted_mos":2.8830525875091553,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289353,"input_sample_rate":48000,"input_seconds":6.0281875,"input_float32_sha256":"44340c41ea5b01b368a2ae52fd9511f27138f2b7090f83bb163089147410ef65","internal_samples":96451,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.553158426000664},"started_at":"2026-10-13T05:16:10.238007+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008769848000156344,"native_input_samples":288000,"native_input_sha256":"339d7e1f12a79eca432de2ad3b38d4f4d2eac0ee9eb9403940e4add736ae9663","output_sha256":"8d78b5231bf5ec5f36128ca704266f9bdeda9140f6a35c25b8fd58e92a3e2120","output_samples":289353,"length_delta_samples":1353,"analysis_s":3.1574236929991457,"synthesis_s":0.9353604750012892,"total_s":4.092784168000435,"rtf":0.6821306946667391,"output_peak":0.6450920144219049,"output_rms":0.0549916354907597,"finished_at":"2026-10-13T05:16:14.350123+00:00","evaluation_input":{"reference":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"8d78b5231bf5ec5f36128ca704266f9bdeda9140f6a35c25b8fd58e92a3e2120","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96451,"length_delta_samples":451,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.188912004259198,"output_peak":0.6454476298044086},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9953240505541674,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.9610755415895165,"spectral_convergence":0.7294855314733656,"spectral_convergence_status":"ok","amplitude_floor":1.4461531006911672e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.973886354139807,"spectral_convergence":0.4761104904981839,"spectral_convergence_status":"ok","amplitude_floor":7.398993822544586e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4514439337995,"spectral_convergence":0.48353795759404716,"spectral_convergence_status":"ok","amplitude_floor":6.518159095373565e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005975557583734567,"output_power":0.0005288167633004632,"delta_db":-0.5307322217890069,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005550914808547427,"output_power":0.0005399021640514017,"delta_db":-0.12049493916925932,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0010492685783192576,"output_power":0.0010021703280387315,"delta_db":-0.19945127422362008,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007612304608209731,"output_power":0.0007461373724191587,"delta_db":-0.08697364948769751,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00020798712521433106,"output_power":0.00021610108148203516,"delta_db":0.16620488156369004,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.709473434650137e-06,"output_power":8.756669303138349e-06,"delta_db":0.023470499980742653,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1878879110356694,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19262958489594365,"pitch":null,"output_rms_dbfs":-25.168572974140222,"output_peak":0.6454476298044086,"reference_rms_dbfs":-24.975943389244275,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.128801943176274},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6432948044673797,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05639009555779642,"rms_dbfs":-24.975943389244275,"peak":0.599212646484375,"peak_dbfs":-4.448380587065618,"crest_factor_db":20.527562802178657,"mean":-2.918211619059245e-05,"zero_energy":false},"output":{"rms":0.05515328076003212,"rms_dbfs":-25.168572974140222,"peak":0.6454476298044086,"peak_dbfs":-3.802779793269237,"crest_factor_db":21.36579318087098,"mean":0.0013562753618772169,"zero_energy":false},"rms_difference_db":-0.19262958489594695},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.163738206843846e-06,"threshold_dbfs":-53.805165843721916,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":32,"quiet_samples":5120,"reference_mean_power":2.7840253096655944e-06,"output_mean_power":8.81848722186062e-06,"output_rms_dbfs":-50.546059100549414,"output_zero_energy":false,"output_peak":0.021468506510296786,"output_energy_fraction":0.00015461446260376188,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":114,"both_estimated_voiced_frames":89,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.7946428571428571,"absolute_difference_cents_median":7.9265540270049835,"absolute_difference_cents_p95":47.31778369875803,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002040524003006053,"log_rms_mae_db":3.173691696302596,"rms_envelope_nrmse":0.2765734340021022,"envelope_correlation":0.9283064541238497,"rms_gain_error_db":-0.19263005734531302,"reference_mean_power":0.0031824932955688964,"output_mean_power":0.0030444194767027047},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.3108768872499092e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006941576008860869,"output_power":0.0006998000696596162,"reference_active":true,"delta_db":0.03515897918735274},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.971136227750716e-05,"output_power":6.688221806043462e-05,"reference_active":true,"delta_db":-0.17992902192812},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.229406504274779e-05,"output_power":2.5891967736770344e-05,"reference_active":true,"delta_db":0.649757933398313}],"active_band_count":3,"active_band_mae_db":0.28828197817126194}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019294226027584624,"log_rms_mae_db":2.0228937227607235,"rms_envelope_nrmse":0.20124102294618065,"envelope_correlation":0.9602718856731386,"rms_gain_error_db":-0.1926209606671197,"reference_mean_power":0.0031984676882902833,"output_mean_power":0.003059707222451444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.1956491507193048e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000632939746621434,"output_power":0.0006378060694401158,"reference_active":true,"delta_db":0.033262788699926105},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.313010835721604e-05,"output_power":3.074079360068336e-05,"reference_active":true,"delta_db":-0.32507781309198536},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3632381122717e-06,"output_power":1.4390721967979716e-06,"reference_active":true,"delta_db":0.2351086336400485}],"active_band_count":3,"active_band_mae_db":0.19781641181065332}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015457760673902765,"log_rms_mae_db":3.6507371705416864,"rms_envelope_nrmse":0.327190797107934,"envelope_correlation":0.9187343859457883,"rms_gain_error_db":-0.1860672637601846,"reference_mean_power":0.0011955559481795221,"output_mean_power":0.001145415822710534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2205605502187216e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004134156506493551,"output_power":0.0004064028095136488,"reference_active":true,"delta_db":-0.0743021204186844},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135405678849972e-05,"output_power":3.9893912697517504e-05,"reference_active":true,"delta_db":-0.15611486859248527},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1249926033986768e-05,"output_power":1.2063695196520823e-05,"reference_active":true,"delta_db":0.3033068864488451}],"active_band_count":3,"active_band_mae_db":0.17790795848667162}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014615313182534672,"log_rms_mae_db":2.4548315438622366,"rms_envelope_nrmse":0.23770474407093392,"envelope_correlation":0.9551385590894146,"rms_gain_error_db":-0.18606734078069914,"reference_mean_power":0.0012015673286161978,"output_mean_power":0.0011511750729329248},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.16593985496473e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003798908177653278,"output_power":0.0003728527499522297,"reference_active":true,"delta_db":-0.08121445920784871},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8446225039409073e-05,"output_power":1.7423242504434877e-05,"reference_active":true,"delta_db":-0.24778521251363847},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.60382048073901e-07,"output_power":7.916126410805806e-07,"reference_active":true,"delta_db":0.1748086572050414}],"active_band_count":3,"active_band_mae_db":0.1679361096421762}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011311095186173683,"log_rms_mae_db":3.634469504698965,"rms_envelope_nrmse":0.3754373562969013,"envelope_correlation":0.8934488800044177,"rms_gain_error_db":-0.03940077332532355,"reference_mean_power":0.0008539373095994977,"output_mean_power":0.0008462251168697766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.10125907860188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022995365445444387,"output_power":0.0002309636348035732,"reference_active":true,"delta_db":0.019032898297045044},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.837295392211272e-05,"output_power":5.722703580856928e-05,"reference_active":true,"delta_db":-0.086104197123065},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3078079511035852e-05,"output_power":1.4351335392109032e-05,"reference_active":true,"delta_db":0.4034834063964663}],"active_band_count":3,"active_band_mae_db":0.16954016727219212}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.65333606126804e-05,"log_rms_mae_db":2.4624509903615373,"rms_envelope_nrmse":0.2688179742859339,"envelope_correlation":0.9417466433191987,"rms_gain_error_db":-0.03939228060399383,"reference_mean_power":0.0008582264011938478,"output_mean_power":0.0008504771353716032},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.58513640339241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002121429018976754,"output_power":0.0002131636634187751,"reference_active":true,"delta_db":0.02084670133535915},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7723754282185392e-05,"output_power":2.6747230803248126e-05,"reference_active":true,"delta_db":-0.15573215737418936},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.120050210892593e-07,"output_power":9.368304747853352e-07,"reference_active":true,"delta_db":0.1166378038488462}],"active_band_count":3,"active_band_mae_db":0.09773888751946491}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.92616098940908e-05,"log_rms_mae_db":3.1884454527726165,"rms_envelope_nrmse":0.43196564182926317,"envelope_correlation":0.8501828505245472,"rms_gain_error_db":0.14646744898285688,"reference_mean_power":3.945170478466555e-05,"output_mean_power":4.080491894940053e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.222824115041211e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1970652723962296e-05,"output_power":1.2006037122013481e-05,"reference_active":true,"delta_db":0.01281850002992593},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.324380113074305e-06,"output_power":3.3543670357145095e-06,"reference_active":true,"delta_db":0.03899905852118038},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0798067724676819e-06,"output_power":1.1511229483191725e-06,"reference_active":true,"delta_db":0.2777566491611163}],"active_band_count":3,"active_band_mae_db":0.10985806923740754}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.5144247671385576e-05,"log_rms_mae_db":2.1858719218666085,"rms_envelope_nrmse":0.2805541225869638,"envelope_correlation":0.9312840196739938,"rms_gain_error_db":0.1464096010013846,"reference_mean_power":3.964923190102766e-05,"output_mean_power":4.100867511066832e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.9761262616699e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.1283346332236671e-05,"output_power":1.1359274711415783e-05,"reference_active":true,"delta_db":0.029126839720212308},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7283331039583754e-06,"output_power":1.718945547126606e-06,"reference_active":true,"delta_db":-0.023653290732417058},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.035315524247487e-08,"output_power":7.11876318171815e-08,"reference_active":true,"delta_db":0.05120965882050788}],"active_band_count":3,"active_band_mae_db":0.03466326309104575}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"hnm","status":"measured","mos":{"status":"measured","value":2.752824544906616,"predicted_mos":2.752824544906616,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"9e945b465f1b43ab9f1c018b7f4f9049dfbfb6bdfea44329dd22d8da69050095","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.142186248000144},"started_at":"2026-10-13T04:39:29.832314+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.531100042484468e-05,"native_input_samples":96000,"native_input_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","output_sha256":"f18ba0f6f003e605cf6a88df45adb79ea67010cfce70fef451674f2bb5228c13","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0634116040000663,"synthesis_s":0.5080677980004111,"total_s":1.5714794020004774,"rtf":0.26191323366674624,"output_peak":0.6226843441125598,"output_rms":0.05291795051947269,"finished_at":"2026-10-13T04:39:31.409655+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"f18ba0f6f003e605cf6a88df45adb79ea67010cfce70fef451674f2bb5228c13","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.996868545009905,"spectral_convergence":0.33778760880338077,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.59602241947944,"spectral_convergence":0.32823627862982013,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.108609327242153,"spectral_convergence":0.34715461486098537,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":9.394788934181716e-05,"delta_db":0.5383773043389575,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0007164899681575877,"delta_db":-0.1845551002455611,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.001180121081267386,"delta_db":-0.24421269982152632,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.0005451833031284498,"delta_db":-0.11311881613818048,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.00012371841900942008,"delta_db":0.053776914067487434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.00014084882627669887,"delta_db":0.0003404251789290224,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1890635432984403,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15427748519946344,"pitch":null,"output_rms_dbfs":-25.527939682479715,"output_peak":0.6226843441125598,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.900500097243833},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.996868545009905,"spectral_convergence":0.33778760880338077,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.59602241947944,"spectral_convergence":0.32823627862982013,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.108609327242153,"spectral_convergence":0.34715461486098537,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":9.394788934181716e-05,"delta_db":0.5383773043389575,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0007164899681575877,"delta_db":-0.1845551002455611,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.001180121081267386,"delta_db":-0.24421269982152632,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.0005451833031284498,"delta_db":-0.11311881613818048,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.00012371841900942008,"delta_db":0.053776914067487434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.00014084882627669887,"delta_db":0.0003404251789290224,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1890635432984403,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15427748519946344,"pitch":null,"output_rms_dbfs":-25.527939682479715,"output_peak":0.6226843441125598,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.900500097243833},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8000137484906993,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053866268327212564,"rms_dbfs":-25.37366219728025,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.789300578656956,"mean":-0.0032699441909790038,"zero_energy":false},"output":{"rms":0.05291795051947269,"rms_dbfs":-25.52793968247972,"peak":0.6226843441125598,"peak_dbfs":-4.114641068452983,"crest_factor_db":21.413298614026736,"mean":-2.194856970325786e-05,"zero_energy":false},"rms_difference_db":-0.1542774851994686},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":103,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.7586206896551724,"absolute_difference_cents_median":0.8579675178339694,"absolute_difference_cents_p95":4.4042800254821906,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023508705421562093,"log_rms_mae_db":1.224610113822032,"rms_envelope_nrmse":0.1158499696402586,"envelope_correlation":0.987868283111983,"rms_gain_error_db":-0.15431954595728492,"reference_mean_power":0.0029034514839332195,"output_mean_power":0.0028020934753155723},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1197992634528758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005958091728481294,"output_power":0.0006080289813906654,"reference_active":true,"delta_db":0.0881709502985893},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.320217055945433e-05,"output_power":7.990593119318314e-05,"reference_active":true,"delta_db":-0.668470096633666},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9969137080619498e-05,"output_power":1.7082163661567998e-05,"reference_active":true,"delta_db":-0.6781641980087574}],"active_band_count":3,"active_band_mae_db":0.47826841498033756}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019484731874036573,"log_rms_mae_db":0.6318649670272477,"rms_envelope_nrmse":0.059022470028070116,"envelope_correlation":0.9968337912611885,"rms_gain_error_db":-0.15417021472012465,"reference_mean_power":0.002910052669271509,"output_mean_power":0.0028085607867518124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0516863068209195e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005686142820757539,"output_power":0.0005806011287876939,"reference_active":true,"delta_db":0.09060112027401884},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3606311869244326e-05,"output_power":3.529119719932711e-05,"reference_active":true,"delta_db":-0.9188296512771086},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3652354435919545e-06,"output_power":9.351668576578057e-07,"reference_active":true,"delta_db":-1.6431844768260915}],"active_band_count":3,"active_band_mae_db":0.8842050827924064}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020607260287182458,"log_rms_mae_db":1.5825032485965789,"rms_envelope_nrmse":0.15992295538571025,"envelope_correlation":0.9805099391444863,"rms_gain_error_db":-0.19202001476281702,"reference_mean_power":0.001692101437139437,"output_mean_power":0.0016189162891215789},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.505279098297708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003919860661962002,"output_power":0.0004043271750321372,"reference_active":true,"delta_db":0.1346230189722442},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.427449102286661e-05,"output_power":3.8647810764148246e-05,"reference_active":true,"delta_db":-0.5902867926896161},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.188127359577471e-06,"output_power":7.730102017976168e-06,"reference_active":true,"delta_db":-0.7504178090465352}],"active_band_count":3,"active_band_mae_db":0.4917758735694652}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016462000863024626,"log_rms_mae_db":0.9474946295144376,"rms_envelope_nrmse":0.08521938536455265,"envelope_correlation":0.9943857471546802,"rms_gain_error_db":-0.191902129745001,"reference_mean_power":0.001695759255771571,"output_mean_power":0.0016224599427925885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.197777273221207e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003745482187229297,"output_power":0.00038573702875541914,"reference_active":true,"delta_db":0.12783594892934916},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9699889570935678e-05,"output_power":1.6878034778021307e-05,"reference_active":true,"delta_db":-0.6714191426638367},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294406999255858e-07,"output_power":6.466171193872902e-07,"reference_active":true,"delta_db":-1.081381434912399}],"active_band_count":3,"active_band_mae_db":0.6268788421685283}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010781730421004015,"log_rms_mae_db":1.5992170619569048,"rms_envelope_nrmse":0.1303333382275751,"envelope_correlation":0.9874266844628329,"rms_gain_error_db":-0.13905465420131902,"reference_mean_power":0.0005546344869318722,"output_mean_power":0.0005371572044014188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0245820082367756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018054631551466004,"output_power":0.00017076975358983754,"reference_active":true,"delta_db":-0.24177678260040328},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2448913022603264e-05,"output_power":1.979062735696712e-05,"reference_active":true,"delta_db":-0.5473575585995365},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.880443381781099e-06,"output_power":3.614799578895981e-06,"reference_active":true,"delta_db":-1.30375055854541}],"active_band_count":3,"active_band_mae_db":0.6976282999151167}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010092628732279428,"log_rms_mae_db":0.9235781210264209,"rms_envelope_nrmse":0.07114186154503602,"envelope_correlation":0.9961729341342589,"rms_gain_error_db":-0.13949294068096377,"reference_mean_power":0.0005569522649053957,"output_mean_power":0.0005393475127959028},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.880183972268193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017516791642605378,"output_power":0.00016519307981993018,"reference_active":true,"delta_db":-0.2546271405538635},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.918962387005768e-06,"output_power":8.67329758132884e-06,"reference_active":true,"delta_db":-0.5828199609341846},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2471992219001627e-07,"output_power":2.5834416569925883e-07,"reference_active":true,"delta_db":-0.9931027611505259}],"active_band_count":3,"active_band_mae_db":0.6101832875461913}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001308702624055786,"log_rms_mae_db":1.7032956007181006,"rms_envelope_nrmse":0.16982484299370312,"envelope_correlation":0.9825236090293858,"rms_gain_error_db":-0.02730938358320256,"reference_mean_power":0.00016841175600957621,"output_mean_power":0.00016735606885590534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6472123519121686e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.520693965195543e-05,"output_power":9.612766747804279e-05,"reference_active":true,"delta_db":0.0417979901043993},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1413816053513927e-05,"output_power":1.896844803209006e-05,"reference_active":true,"delta_db":-0.5266226866288405},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.377484332308182e-06,"output_power":4.966697876701665e-06,"reference_active":true,"delta_db":-0.3451141096578926}],"active_band_count":3,"active_band_mae_db":0.3045115954637108}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":1.115625587710138,"rms_envelope_nrmse":0.09976555719496775,"envelope_correlation":0.9937204895376371,"rms_gain_error_db":-0.02725252659895042,"reference_mean_power":0.00016921252708073628,"output_mean_power":0.00016815402172798232},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5742070644083455e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.764878819980253e-05,"output_power":9.84000728176747e-05,"reference_active":true,"delta_db":0.033285616591644446},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.033680531591512e-05,"output_power":9.039831128205846e-06,"reference_active":true,"delta_db":-0.5822601926119465},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.467040891819502e-07,"output_power":2.706516862354519e-07,"reference_active":true,"delta_db":-1.0754822673733764}],"active_band_count":3,"active_band_mae_db":0.5636760255256558}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.3599953651428223,"predicted_mos":3.3599953651428223,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"01ff03fdcadae0a4e84c5a275763d7820a457584594b22ad3d54a409218d425d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.985688681999818},"started_at":"2026-10-13T04:42:28.159173+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.592799955891678e-05,"native_input_samples":96000,"native_input_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","output_sha256":"0c8ffa7259a242748f7adb7e343f7baced6f6039723c65909040970900240c20","output_samples":96000,"length_delta_samples":0,"analysis_s":2.274355939999623,"synthesis_s":0.3873270379999667,"total_s":2.6616829779995896,"rtf":0.4436138296665983,"output_peak":0.6832285937387161,"output_rms":0.05863693803383788,"finished_at":"2026-10-13T04:42:30.831771+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"0c8ffa7259a242748f7adb7e343f7baced6f6039723c65909040970900240c20","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.943483607008339,"spectral_convergence":0.6209668394562643,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.292526123253787,"spectral_convergence":0.32130841467136795,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.655673061240849,"spectral_convergence":0.36732154678953294,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":6.841353882072084e-05,"delta_db":-0.8390724142169692,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.000992314076242612,"delta_db":1.2298352790145082,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0014443907014958259,"delta_db":0.6333684542225901,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.0006485294477336292,"delta_db":0.6407527264111206,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.00013864953470504995,"delta_db":0.5486174632883529,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.00014599320298630547,"delta_db":0.15613447032223832,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6746301345792967,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7370878727357815,"pitch":null,"output_rms_dbfs":-24.636574324544473,"output_peak":0.6832285937387161,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.963894263834324},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.943483607008339,"spectral_convergence":0.6209668394562643,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.292526123253787,"spectral_convergence":0.32130841467136795,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.655673061240849,"spectral_convergence":0.36732154678953294,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":6.841353882072084e-05,"delta_db":-0.8390724142169692,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.000992314076242612,"delta_db":1.2298352790145082,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0014443907014958259,"delta_db":0.6333684542225901,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.0006485294477336292,"delta_db":0.6407527264111206,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.00013864953470504995,"delta_db":0.5486174632883529,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.00014599320298630547,"delta_db":0.15613447032223832,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6746301345792967,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7370878727357815,"pitch":null,"output_rms_dbfs":-24.636574324544473,"output_peak":0.6832285937387161,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.963894263834324},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8756126549445917,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053866268327212564,"rms_dbfs":-25.37366219728025,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.789300578656956,"mean":-0.0032699441909790038,"zero_energy":false},"output":{"rms":0.05863693803383788,"rms_dbfs":-24.63657432454447,"peak":0.6832285937387161,"peak_dbfs":-3.3086793263844427,"crest_factor_db":21.327894998160026,"mean":2.338604532472721e-06,"zero_energy":false},"rms_difference_db":0.7370878727357812},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":152,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.8879310344827587,"absolute_difference_cents_median":4.147820990225663,"absolute_difference_cents_p95":13.55528363853879,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023508705421562093,"log_rms_mae_db":3.501049141748807,"rms_envelope_nrmse":0.17259715006936752,"envelope_correlation":0.9820531427551564,"rms_gain_error_db":0.7378345183162014,"reference_mean_power":0.0029034514839332195,"output_mean_power":0.003441105797444533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1197992634528758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005958091728481294,"output_power":0.0007294093020481558,"reference_active":true,"delta_db":0.8786411234513822},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.320217055945433e-05,"output_power":9.77531383578222e-05,"reference_active":true,"delta_db":0.20704682651092943},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9969137080619498e-05,"output_power":2.1093985075845902e-05,"reference_active":true,"delta_db":0.237993361482852}],"active_band_count":3,"active_band_mae_db":0.44122710381505453}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019484731874036573,"log_rms_mae_db":2.8560785120872016,"rms_envelope_nrmse":0.11441514555974269,"envelope_correlation":0.995571729180682,"rms_gain_error_db":0.7391665206986855,"reference_mean_power":0.002910052669271509,"output_mean_power":0.0034499873404637282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0516863068209195e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005686142820757539,"output_power":0.000695000308037682,"reference_active":true,"delta_db":0.8716723326626142},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3606311869244326e-05,"output_power":4.749100542195961e-05,"reference_active":true,"delta_db":0.37062007528579033},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3652354435919545e-06,"output_power":1.1438665598008234e-06,"reference_active":true,"delta_db":-0.7683219083874941}],"active_band_count":3,"active_band_mae_db":0.6702047721119663}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020607260287182458,"log_rms_mae_db":1.8452797503411718,"rms_envelope_nrmse":0.18026068426544567,"envelope_correlation":0.9849853609093185,"rms_gain_error_db":0.853423658904811,"reference_mean_power":0.001692101437139437,"output_mean_power":0.0020595330243912296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.505279098297708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003919860661962002,"output_power":0.000479491188450271,"reference_active":true,"delta_db":0.8751000103726261},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.427449102286661e-05,"output_power":5.402723565940076e-05,"reference_active":true,"delta_db":0.8645916986722755},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.188127359577471e-06,"output_power":6.965211043039977e-06,"reference_active":true,"delta_db":-1.2029272655553311}],"active_band_count":3,"active_band_mae_db":0.980872991533411}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016462000863024626,"log_rms_mae_db":1.2858261267446647,"rms_envelope_nrmse":0.13021074234336932,"envelope_correlation":0.9954292926852529,"rms_gain_error_db":0.8549473083130454,"reference_mean_power":0.001695759255771571,"output_mean_power":0.00206470936215399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.197777273221207e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003745482187229297,"output_power":0.00045919419700182095,"reference_active":true,"delta_db":0.8848865509936505},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9699889570935678e-05,"output_power":2.4071966227805645e-05,"reference_active":true,"delta_db":0.8704777372995841},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294406999255858e-07,"output_power":5.922527968198963e-07,"reference_active":true,"delta_db":-1.4627822171263594}],"active_band_count":3,"active_band_mae_db":1.0727155018065313}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010781730421004015,"log_rms_mae_db":1.5991445597222824,"rms_envelope_nrmse":0.17701884298130866,"envelope_correlation":0.9837630804681579,"rms_gain_error_db":0.7204093381306521,"reference_mean_power":0.0005546344869318722,"output_mean_power":0.0006547082358619294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0245820082367756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018054631551466004,"output_power":0.00020868395330161704,"reference_active":true,"delta_db":0.6290042536662186},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2448913022603264e-05,"output_power":2.5164201906620124e-05,"reference_active":true,"delta_db":0.4958784382741066},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.880443381781099e-06,"output_power":3.822394336398732e-06,"reference_active":true,"delta_db":-1.0612378995355698}],"active_band_count":3,"active_band_mae_db":0.7287068638252983}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010092628732279428,"log_rms_mae_db":1.0313380182119438,"rms_envelope_nrmse":0.10889054769057901,"envelope_correlation":0.9968188996572029,"rms_gain_error_db":0.7191097239660621,"reference_mean_power":0.0005569522649053957,"output_mean_power":0.0006572475059766462},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.880183972268193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017516791642605378,"output_power":0.00020184576567654032,"reference_active":true,"delta_db":0.6156507901696716},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.918962387005768e-06,"output_power":1.1191434339425176e-05,"reference_active":true,"delta_db":0.5241950762284296},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2471992219001627e-07,"output_power":2.4440714082137786e-07,"reference_active":true,"delta_db":-1.2339504374423085}],"active_band_count":3,"active_band_mae_db":0.7912654346134699}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001308702624055786,"log_rms_mae_db":1.7631201408305683,"rms_envelope_nrmse":0.24969771071093624,"envelope_correlation":0.9643281082475742,"rms_gain_error_db":0.2993079658101077,"reference_mean_power":0.00016841175600957621,"output_mean_power":0.00018042769493367656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6472123519121686e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.520693965195543e-05,"output_power":0.00010303762634406939,"reference_active":true,"delta_db":0.3432724003547333},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1413816053513927e-05,"output_power":2.037863157273384e-05,"reference_active":true,"delta_db":-0.2151904999999418},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.377484332308182e-06,"output_power":7.618552178862054e-06,"reference_active":true,"delta_db":1.5129329263028957}],"active_band_count":3,"active_band_mae_db":0.6904652755525236}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":1.0556360341280786,"rms_envelope_nrmse":0.14222517098143403,"envelope_correlation":0.9884724937568389,"rms_gain_error_db":0.2992071426170579,"reference_mean_power":0.00016921252708073628,"output_mean_power":0.0001812813913117084},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5742070644083455e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.764878819980253e-05,"output_power":0.00010606718624736109,"reference_active":true,"delta_db":0.35914189873595226},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.033680531591512e-05,"output_power":9.94982636355488e-06,"reference_active":true,"delta_db":-0.16570834968696363},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.467040891819502e-07,"output_power":4.461273320070285e-07,"reference_active":true,"delta_db":1.0949986717520053}],"active_band_count":3,"active_band_mae_db":0.539949640058307}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"facodec","status":"measured","mos":{"status":"measured","value":3.8095505237579346,"predicted_mos":3.8095505237579346,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f1cc402cf225c108f796ed0d83334df0a6632285060fe29ec64860b33515a37f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9460630159992434},"started_at":"2026-10-13T04:54:04.052995+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.9011999686190393e-05,"native_input_samples":96000,"native_input_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","output_sha256":"ebe591b3be04845a31f40d49713ea84d3d19d7d176b7556d87389c97f1f46476","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3754746230006276,"synthesis_s":5.807862386999659,"total_s":7.183337010000287,"rtf":1.1972228350000478,"output_peak":0.5429461598396301,"output_rms":0.05163984825671391,"finished_at":"2026-10-13T04:54:11.243015+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"ebe591b3be04845a31f40d49713ea84d3d19d7d176b7556d87389c97f1f46476","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.609111518225847,"spectral_convergence":0.2683769815907037,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.769680012805968,"spectral_convergence":0.29346714282470615,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.880682765378793,"spectral_convergence":0.3111664379618955,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":8.796823603108164e-05,"delta_db":0.2527658630047045,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0007649556601538527,"delta_db":0.09970638380657185,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0011270891191858907,"delta_db":-0.4438958133440507,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.00047681887600840736,"delta_db":-0.6950098855575658,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":9.641563676045897e-05,"delta_db":-1.0290919631500028,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.00011342639983674746,"delta_db":-0.9400504199205305,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.576753388130571,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3666386542126666,"pitch":null,"output_rms_dbfs":-25.740300851492922,"output_peak":0.5429461598396301,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.753158098803536},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.609111518225847,"spectral_convergence":0.2683769815907037,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.769680012805968,"spectral_convergence":0.29346714282470615,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.880682765378793,"spectral_convergence":0.3111664379618955,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":8.796823603108164e-05,"delta_db":0.2527658630047045,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0007649556601538527,"delta_db":0.09970638380657185,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0011270891191858907,"delta_db":-0.4438958133440507,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.00047681887600840736,"delta_db":-0.6950098855575658,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":9.641563676045897e-05,"delta_db":-1.0290919631500028,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.00011342639983674746,"delta_db":-0.9400504199205305,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.576753388130571,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3666386542126666,"pitch":null,"output_rms_dbfs":-25.740300851492922,"output_peak":0.5429461598396301,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.753158098803536},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8532595096584812,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053866268327212564,"rms_dbfs":-25.37366219728025,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.789300578656956,"mean":-0.0032699441909790038,"zero_energy":false},"output":{"rms":0.05163984825671391,"rms_dbfs":-25.740300851492922,"peak":0.5429461598396301,"peak_dbfs":-5.304864684239766,"crest_factor_db":20.435436167253155,"mean":-0.00015659787455274453,"zero_energy":false},"rms_difference_db":-0.3666386542126716},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":139,"both_estimated_voiced_frames":108,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.9310344827586207,"absolute_difference_cents_median":2.370004828134684,"absolute_difference_cents_p95":10.332767895715156,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023508705421562093,"log_rms_mae_db":2.6303077476031604,"rms_envelope_nrmse":0.11134479036065707,"envelope_correlation":0.9898774534437376,"rms_gain_error_db":-0.36727937651037856,"reference_mean_power":0.0029034514839332195,"output_mean_power":0.0026680049795431504},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1197992634528758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005958091728481294,"output_power":0.0006063078084002237,"reference_active":true,"delta_db":0.07585976218344348},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.320217055945433e-05,"output_power":8.951977174927483e-05,"reference_active":true,"delta_db":-0.1750706044470774},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9969137080619498e-05,"output_power":2.1288341400181762e-05,"reference_active":true,"delta_db":0.2778252810075308}],"active_band_count":3,"active_band_mae_db":0.17625188254601723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019484731874036573,"log_rms_mae_db":2.2025578929553418,"rms_envelope_nrmse":0.09643514695584025,"envelope_correlation":0.9922570489711096,"rms_gain_error_db":-0.3703947216278082,"reference_mean_power":0.002910052669271509,"output_mean_power":0.0026721533459292436},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0516863068209195e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005686142820757539,"output_power":0.0005797648853671399,"reference_active":true,"delta_db":0.08434144080188255},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3606311869244326e-05,"output_power":4.2608443776159737e-05,"reference_active":true,"delta_db":-0.1005368411732794},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3652354435919545e-06,"output_power":1.3792174899729265e-06,"reference_active":true,"delta_db":0.04425201141060774}],"active_band_count":3,"active_band_mae_db":0.07637676446192324}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020607260287182458,"log_rms_mae_db":1.579000471380091,"rms_envelope_nrmse":0.11084569239182039,"envelope_correlation":0.9910030292420097,"rms_gain_error_db":-0.20783720358635172,"reference_mean_power":0.001692101437139437,"output_mean_power":0.0016130308514964125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.505279098297708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003919860661962002,"output_power":0.0004334364245216481,"reference_active":true,"delta_db":0.4365477558002264},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.427449102286661e-05,"output_power":4.299560898141001e-05,"reference_active":true,"delta_db":-0.1272947267434834},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.188127359577471e-06,"output_power":9.412569865787533e-06,"reference_active":true,"delta_db":0.10481206357400451}],"active_band_count":3,"active_band_mae_db":0.22288484870590478}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016462000863024626,"log_rms_mae_db":1.2730253735076384,"rms_envelope_nrmse":0.09666780567447868,"envelope_correlation":0.9928803729441402,"rms_gain_error_db":-0.20972396581251462,"reference_mean_power":0.001695759255771571,"output_mean_power":0.0016158156107237908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.197777273221207e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003745482187229297,"output_power":0.00041395982013394517,"reference_active":true,"delta_db":0.43450453579230863},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9699889570935678e-05,"output_power":1.9432365193053088e-05,"reference_active":true,"delta_db":-0.05938128117616071},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294406999255858e-07,"output_power":7.878323091884271e-07,"reference_active":true,"delta_db":-0.22351554641829047}],"active_band_count":3,"active_band_mae_db":0.23913378779558658}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010781730421004015,"log_rms_mae_db":1.805046137567278,"rms_envelope_nrmse":0.15139092165778387,"envelope_correlation":0.9885745655865892,"rms_gain_error_db":-0.8603040466730149,"reference_mean_power":0.0005546344869318722,"output_mean_power":0.0004549634050933121},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0245820082367756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018054631551466004,"output_power":0.00014922543283257017,"reference_active":true,"delta_db":-0.8274578277242449},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2448913022603264e-05,"output_power":1.8218577088765183e-05,"reference_active":true,"delta_db":-0.9068086268946162},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.880443381781099e-06,"output_power":5.0235106843983414e-06,"reference_active":true,"delta_db":0.1254805141795884}],"active_band_count":3,"active_band_mae_db":0.6199156562661498}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010092628732279428,"log_rms_mae_db":1.5374178642199225,"rms_envelope_nrmse":0.1297614464151081,"envelope_correlation":0.993110495085316,"rms_gain_error_db":-0.8604887064717803,"reference_mean_power":0.0005569522649053957,"output_mean_power":0.000456845239483166},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.880183972268193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017516791642605378,"output_power":0.00014534909730353426,"reference_active":true,"delta_db":-0.8104222527753479},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.918962387005768e-06,"output_power":8.156636834341483e-06,"reference_active":true,"delta_db":-0.8495511717398864},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2471992219001627e-07,"output_power":3.1921693049889526e-07,"reference_active":true,"delta_db":-0.07423017017571755}],"active_band_count":3,"active_band_mae_db":0.5780678648969839}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001308702624055786,"log_rms_mae_db":1.975523086220079,"rms_envelope_nrmse":0.2301447261016903,"envelope_correlation":0.9719419787075623,"rms_gain_error_db":-0.9887959402281532,"reference_mean_power":0.00016841175600957621,"output_mean_power":0.0001341197731507594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6472123519121686e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.520693965195543e-05,"output_power":7.674877677657888e-05,"reference_active":true,"delta_db":-0.9359714293477706},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1413816053513927e-05,"output_power":1.9284941963972687e-05,"reference_active":true,"delta_db":-0.4547573148914836},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.377484332308182e-06,"output_power":6.083872718162912e-06,"reference_active":true,"delta_db":0.5360096580728347}],"active_band_count":3,"active_band_mae_db":0.6422461341040296}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":1.740844616196495,"rms_envelope_nrmse":0.18508873681922716,"envelope_correlation":0.9840220033317096,"rms_gain_error_db":-0.98948728484781,"reference_mean_power":0.00016921252708073628,"output_mean_power":0.0001347360412405796},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5742070644083455e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.764878819980253e-05,"output_power":8.063338345980281e-05,"reference_active":true,"delta_db":-0.8315197453198651},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.033680531591512e-05,"output_power":9.57690644788524e-06,"reference_active":true,"delta_db":-0.33161091788796},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.467040891819502e-07,"output_power":3.8140964681774865e-07,"reference_active":true,"delta_db":0.41432709309785243}],"active_band_count":3,"active_band_mae_db":0.5258192521018925}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.067488193511963,"predicted_mos":3.067488193511963,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"451628748cdeab79ff8d3e7e7df3ac2c7e7c55191ad7c1f15ab24ac4b1734b84","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8400680549966637},"started_at":"2026-10-13T07:17:45.860226+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007420662999720662,"native_input_samples":264600,"native_input_sha256":"5871943ae2ed898e6910b8fb2d8b5ea6b586ed75ffd8c66d62e66bc4797c8db3","output_sha256":"940ca73c2d028f6fccfa2a33e0755a1992dd7687bc00da78174a4a21034c80a6","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1579169649994583,"synthesis_s":4.221073921000425,"total_s":4.378990885999883,"rtf":0.7298318143333139,"output_peak":0.5767492055892944,"output_rms":0.05021629622576495,"finished_at":"2026-10-13T07:17:50.278628+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"940ca73c2d028f6fccfa2a33e0755a1992dd7687bc00da78174a4a21034c80a6","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.977083028324756,"output_peak":0.5768985745431751},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.642078805749682,"spectral_convergence":0.4422174256942826,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.623269179327874,"spectral_convergence":0.4374111863094844,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.434565690181115,"spectral_convergence":0.5024713062910408,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":0.00011510348777946755,"delta_db":1.4203918635718917,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0005995687047401044,"delta_db":-0.9582666787336915,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0011102980401661165,"delta_db":-0.5090826527498046,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.000509843959500591,"delta_db":-0.40417149946964737,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.0001234703593569186,"delta_db":0.04506042012791409,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":6.714812546911736e-05,"delta_db":-3.216852964952679,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.092304346600938,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6029798469217361,"pitch":null,"output_rms_dbfs":-25.976642044201988,"output_peak":0.5768985745431751,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.56663789175289},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6660036555265536,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053866268327212564,"rms_dbfs":-25.37366219728025,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.789300578656956,"mean":-0.0032699441909790038,"zero_energy":false},"output":{"rms":0.05025368321837033,"rms_dbfs":-25.976642044201988,"peak":0.5768985745431751,"peak_dbfs":-4.778010682747297,"crest_factor_db":21.19863136145469,"mean":-9.502187114624951e-05,"zero_energy":false},"rms_difference_db":-0.6029798469217376},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":162,"both_estimated_voiced_frames":89,"reference_voiced_output_unvoiced_frames":27,"reference_unvoiced_output_voiced_frames":73,"paired_voiced_fraction":0.7672413793103449,"absolute_difference_cents_median":7.659950804755143,"absolute_difference_cents_p95":29.506181208169107,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023508705421562093,"log_rms_mae_db":1.8884113668574227,"rms_envelope_nrmse":0.12887885105485689,"envelope_correlation":0.9888375514333088,"rms_gain_error_db":-0.6029432664688117,"reference_mean_power":0.0029034514839332195,"output_mean_power":0.0025270873094682756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1197992634528758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005958091728481294,"output_power":0.0005257914279831961,"reference_active":true,"delta_db":-0.5429368366798305},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.320217055945433e-05,"output_power":8.00231639285557e-05,"reference_active":true,"delta_db":-0.662103082657737},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9969137080619498e-05,"output_power":1.276366582125729e-05,"reference_active":true,"delta_db":-1.9438387330258164}],"active_band_count":3,"active_band_mae_db":1.0496262174544613}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019484731874036573,"log_rms_mae_db":1.5953535296545442,"rms_envelope_nrmse":0.09734306079361967,"envelope_correlation":0.9953111741478642,"rms_gain_error_db":-0.6022453250973071,"reference_mean_power":0.002910052669271509,"output_mean_power":0.002533239882954078},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0516863068209195e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005686142820757539,"output_power":0.000499973696399951,"reference_active":true,"delta_db":-0.5587060709334589},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3606311869244326e-05,"output_power":3.6785757992452944e-05,"reference_active":true,"delta_db":-0.738696471395506},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3652354435919545e-06,"output_power":9.028206494538964e-07,"reference_active":true,"delta_db":-1.7960607092926753}],"active_band_count":3,"active_band_mae_db":1.0311544172072133}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020607260287182458,"log_rms_mae_db":2.631163513369384,"rms_envelope_nrmse":0.14457487607258718,"envelope_correlation":0.987079246542564,"rms_gain_error_db":-0.642827296331755,"reference_mean_power":0.001692101437139437,"output_mean_power":0.0014592969161130128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.505279098297708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003919860661962002,"output_power":0.00034909416280852793,"reference_active":true,"delta_db":-0.503280425154407},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.427449102286661e-05,"output_power":4.153900260283283e-05,"reference_active":true,"delta_db":-0.2769751282932998},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.188127359577471e-06,"output_power":5.2875757346967145e-06,"reference_active":true,"delta_db":-2.399704055706157}],"active_band_count":3,"active_band_mae_db":1.0599865363846213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016462000863024626,"log_rms_mae_db":1.9907935204197642,"rms_envelope_nrmse":0.10861481786279518,"envelope_correlation":0.9940860013416429,"rms_gain_error_db":-0.6376930333618898,"reference_mean_power":0.001695759255771571,"output_mean_power":0.0014641814246371845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.197777273221207e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003745482187229297,"output_power":0.00033490446992883187,"reference_active":true,"delta_db":-0.48586791939821994},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9699889570935678e-05,"output_power":1.8151645657405002e-05,"reference_active":true,"delta_db":-0.3554778670761915},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294406999255858e-07,"output_power":5.384969680520833e-07,"reference_active":true,"delta_db":-1.876020796054906}],"active_band_count":3,"active_band_mae_db":0.9057888608431058}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010781730421004015,"log_rms_mae_db":3.1275985273512568,"rms_envelope_nrmse":0.25720251984001763,"envelope_correlation":0.9532027693608494,"rms_gain_error_db":-0.7431583341091641,"reference_mean_power":0.0005546344869318722,"output_mean_power":0.000467402506388311},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0245820082367756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018054631551466004,"output_power":0.00013107613486908787,"reference_active":true,"delta_db":-1.3906500345820616},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2448913022603264e-05,"output_power":1.666940113867989e-05,"reference_active":true,"delta_db":-1.2927531955161058},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.880443381781099e-06,"output_power":3.828963635622287e-06,"reference_active":true,"delta_db":-1.0537803707825442}],"active_band_count":3,"active_band_mae_db":1.2457278669602372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010092628732279428,"log_rms_mae_db":2.743860601301279,"rms_envelope_nrmse":0.22486379967813594,"envelope_correlation":0.9635287866266982,"rms_gain_error_db":-0.7457095277480756,"reference_mean_power":0.0005569522649053957,"output_mean_power":0.0004690801135562705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.880183972268193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017516791642605378,"output_power":0.00012356817657206625,"reference_active":true,"delta_db":-1.5154792611140637},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.918962387005768e-06,"output_power":7.64585481777024e-06,"reference_active":true,"delta_db":-1.1304019618538592},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2471992219001627e-07,"output_power":2.5902267378735064e-07,"reference_active":true,"delta_db":-0.9817115213170264}],"active_band_count":3,"active_band_mae_db":1.2091975814283165}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001308702624055786,"log_rms_mae_db":3.148850975875865,"rms_envelope_nrmse":0.26838529769408004,"envelope_correlation":0.9699195662220818,"rms_gain_error_db":-1.677284402460111,"reference_mean_power":0.00016841175600957621,"output_mean_power":0.0001144574231042865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6472123519121686e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.520693965195543e-05,"output_power":6.650783410180707e-05,"reference_active":true,"delta_db":-1.557958005419084},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1413816053513927e-05,"output_power":1.4834167891819127e-05,"reference_active":true,"delta_db":-1.5943087771841586},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.377484332308182e-06,"output_power":2.6445173328643933e-06,"reference_active":true,"delta_db":-3.0823273586781443}],"active_band_count":3,"active_band_mae_db":2.0781980470937955}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":2.975637335045695,"rms_envelope_nrmse":0.22407268007064085,"envelope_correlation":0.9852705576069818,"rms_gain_error_db":-1.6769960292115427,"reference_mean_power":0.00016921252708073628,"output_mean_power":0.00011500928632605409},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5742070644083455e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.764878819980253e-05,"output_power":6.722644510853684e-05,"reference_active":true,"delta_db":-1.6212671151018254},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.033680531591512e-05,"output_power":7.174459334474499e-06,"reference_active":true,"delta_db":-1.5859715850741871},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.467040891819502e-07,"output_power":1.8303651500193922e-07,"reference_active":true,"delta_db":-2.7742122568304413}],"active_band_count":3,"active_band_mae_db":1.9938169856688177}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"golf","status":"measured","mos":{"status":"measured","value":3.2528183460235596,"predicted_mos":3.2528183460235596,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"15a996af8daef4894bf95c7d548d68f2221120b44b40409c734dac1eeb8258b3","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5822069559999363},"started_at":"2026-10-13T05:02:40.542349+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003369119000126375,"native_input_samples":144000,"native_input_sha256":"44c5b73265f29bdd68d89502b8b2fa2269e06027186d02b00495dd8a95796a3f","output_sha256":"a271b3c0f3d9c91f2a8a3a062e9ec02d6ba782023c346b7581019a9a45350ba2","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5422104429999308,"synthesis_s":0.11692735500037088,"total_s":0.6591377980003017,"rtf":0.10985629966671695,"output_peak":0.5545393228530884,"output_rms":0.04748707266093441,"finished_at":"2026-10-13T05:02:41.213633+00:00","analysis_stages":{"external_f0_s":0.09276663600030588},"evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"a271b3c0f3d9c91f2a8a3a062e9ec02d6ba782023c346b7581019a9a45350ba2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.463402339509464,"output_peak":0.5548826673137005},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.699665071117263,"spectral_convergence":0.5380541959591126,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.321612631143193,"spectral_convergence":0.3119865451846349,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.632005362019635,"spectral_convergence":0.3335897363861353,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.291392642648239e-05,"output_power":8.58604345274165e-05,"delta_db":0.15165600207877988,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000747923066696487,"output_power":0.0005964357485642853,"delta_db":-0.982932616411193,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012476229260837946,"output_power":0.0009339984892741158,"delta_db":-1.2573717290490538,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00056048770884392,"output_power":0.000462786617671516,"delta_db":-0.8318530118494496,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012237018422354387,"output_power":8.704321558198271e-05,"delta_db":-1.4794068733622692,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014107153412381215,"output_power":9.154187597553612e-05,"delta_db":-1.8781958101504874,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0969026738168723,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.09095905198838,"pitch":null,"output_rms_dbfs":-26.463402339509464,"output_peak":0.5548826673137005,"reference_rms_dbfs":-25.372443287521087,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.884427688093363},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8555870812395819,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05387382802806981,"rms_dbfs":-25.372443287521087,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.78808166889779,"mean":-0.003269235677830564,"zero_energy":false},"output":{"rms":0.04751490694082071,"rms_dbfs":-26.463402339509464,"peak":0.5548826673137005,"peak_dbfs":-5.115976817740186,"crest_factor_db":21.347425521769278,"mean":0.0004094223808705841,"zero_energy":false},"rms_difference_db":-1.0909590519883778},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":162,"both_estimated_voiced_frames":100,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":62,"paired_voiced_fraction":0.8620689655172413,"absolute_difference_cents_median":3.0899559551904714,"absolute_difference_cents_p95":19.58970963072989,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00023508705421562093,"log_rms_mae_db":3.0639864541904553,"rms_envelope_nrmse":0.1754699692214,"envelope_correlation":0.9836201850022397,"rms_gain_error_db":-1.0908190146260306,"reference_mean_power":0.002903243885914357,"output_mean_power":0.002258403918363199},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.117741208413728e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005936318359465253,"output_power":0.00046382418677634195,"reference_active":true,"delta_db":-1.0716379192163332},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.205535555086432e-05,"output_power":6.928370339091445e-05,"reference_active":true,"delta_db":-1.2341796596511396},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9947617066571042e-05,"output_power":1.5859586462268375e-05,"reference_active":true,"delta_db":-0.9959916356883635}],"active_band_count":3,"active_band_mae_db":1.1006030715186121}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00019484731874036573,"log_rms_mae_db":2.5320992850460766,"rms_envelope_nrmse":0.1392033406598711,"envelope_correlation":0.9946605161432029,"rms_gain_error_db":-1.0909071068333847,"reference_mean_power":0.002908794731930005,"output_mean_power":0.0022626759685350673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0496343291214592e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005660461271296101,"output_power":0.0004429806198587216,"reference_active":true,"delta_db":-1.0646709680041442},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.27939529867297e-05,"output_power":3.1319260736984896e-05,"reference_active":true,"delta_db":-1.355709028515667},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3560121255604701e-06,"output_power":1.0251785806340632e-06,"reference_active":true,"delta_db":-1.2146404928058216}],"active_band_count":3,"active_band_mae_db":1.2116734964418776}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00020607260287182458,"log_rms_mae_db":2.025235946972875,"rms_envelope_nrmse":0.18408404293714548,"envelope_correlation":0.9836890836914637,"rms_gain_error_db":-1.123155273077273,"reference_mean_power":0.0016920435310829546,"output_mean_power":0.0013064596605727422},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.494297987348568e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003903414010175584,"output_power":0.00030555171989063527,"reference_active":true,"delta_db":-1.0635988394305198},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.370573050401974e-05,"output_power":3.6774793037668924e-05,"reference_active":true,"delta_db":-0.7498814622434812},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.179648134064555e-06,"output_power":9.496313779364705e-06,"reference_active":true,"delta_db":0.14729021702270256}],"active_band_count":3,"active_band_mae_db":0.6535901728989012}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016462000863024626,"log_rms_mae_db":1.6507292540749259,"rms_envelope_nrmse":0.14642028185218714,"envelope_correlation":0.9939573840778629,"rms_gain_error_db":-1.123344263281091,"reference_mean_power":0.001694998701495752,"output_mean_power":0.0013086844542812242},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.185922490644251e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037274300394207933,"output_power":0.00029429364880197574,"reference_active":true,"delta_db":-1.026286112540893},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9198281499020533e-05,"output_power":1.6615982384820233e-05,"reference_active":true,"delta_db":-0.6273633220697233},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.243291150024515e-07,"output_power":8.104210878670181e-07,"reference_active":true,"delta_db":-0.07389906014326039}],"active_band_count":3,"active_band_mae_db":0.5758494982512923}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00010781730421004015,"log_rms_mae_db":2.0647443980486835,"rms_envelope_nrmse":0.19449227360097504,"envelope_correlation":0.9806120873839343,"rms_gain_error_db":-1.0956885946089445,"reference_mean_power":0.0005553860948724561,"output_mean_power":0.0004315450532582544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.0184122840218873e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001797459354341711,"output_power":0.00013736765776098576,"reference_active":true,"delta_db":-1.1677458532615819},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.215659714111828e-05,"output_power":1.5323888490005038e-05,"reference_active":true,"delta_db":-1.6013407826024166},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.872879580774364e-06,"output_power":3.124153027965938e-06,"reference_active":true,"delta_db":-1.9305338073670746}],"active_band_count":3,"active_band_mae_db":1.5665401477436909}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010092628732279428,"log_rms_mae_db":1.6213465728380838,"rms_envelope_nrmse":0.1518301014672657,"envelope_correlation":0.9924138382510989,"rms_gain_error_db":-1.0954545305730725,"reference_mean_power":0.0005572695988983137,"output_mean_power":0.0004330319077228606},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.8736468167131697e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000174216607401055,"output_power":0.00013199912920565425,"reference_active":true,"delta_db":-1.2051848608495845},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716794804933479e-06,"output_power":6.869545297561689e-06,"reference_active":true,"delta_db":-1.5059503998207604},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2115171732981056e-07,"output_power":2.0951145429391443e-07,"reference_active":true,"delta_db":-1.8550247724866644}],"active_band_count":3,"active_band_mae_db":1.5220533443856699}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001308702624055786,"log_rms_mae_db":2.2494671509090614,"rms_envelope_nrmse":0.2587240583827825,"envelope_correlation":0.9764628125642968,"rms_gain_error_db":-1.8084990924697375,"reference_mean_power":0.00016867459567950638,"output_mean_power":0.0001112243223836005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.639816951688618e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.480869945539939e-05,"output_power":6.338616845325565e-05,"reference_active":true,"delta_db":-1.7485368868173983},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1261095647073644e-05,"output_power":1.3536765605080447e-05,"reference_active":true,"delta_db":-1.9607073225931915},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.350245562601521e-06,"output_power":3.0300404307903018e-06,"reference_active":true,"delta_db":-2.469252920158275}],"active_band_count":3,"active_band_mae_db":2.0594990431896214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":2.0087114408784066,"rms_envelope_nrmse":0.21425469596124141,"envelope_correlation":0.9918298134362935,"rms_gain_error_db":-1.8084009419829568,"reference_mean_power":0.00016944861228734932,"output_mean_power":0.00011173723551517568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.5669823234562243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.719845147580143e-05,"output_power":6.43345639939486e-05,"reference_active":true,"delta_db":-1.792149839685488},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0246736557479904e-05,"output_power":6.531041674381384e-06,"reference_active":true,"delta_db":-1.956031156948811},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4447019252991003e-07,"output_power":2.009673442051348e-07,"reference_active":true,"delta_db":-2.3402615446010437}],"active_band_count":3,"active_band_mae_db":2.029480847078448}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.1457574367523193,"predicted_mos":3.1457574367523193,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"946ea38ad74d24052b25a64a3f0c09fe9e4fee7476aed6dee74fe383eb0d6aab","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.788269725999271},"started_at":"2026-10-13T07:39:40.231683+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027027859978261404,"native_input_samples":144000,"native_input_sha256":"44c5b73265f29bdd68d89502b8b2fa2269e06027186d02b00495dd8a95796a3f","output_sha256":"1eb3934e4070e1904aeaa306003837f3119b5f12e418f3ac6d23f2e80281cc33","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3362035369973455,"synthesis_s":4.751626993001992,"total_s":7.087830529999337,"rtf":1.1813050883332228,"output_peak":0.5472016334533691,"output_rms":0.05148691376497158,"finished_at":"2026-10-13T07:39:47.334989+00:00","analysis_stages":{"external_f0_s":1.4840254109985835},"evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"1eb3934e4070e1904aeaa306003837f3119b5f12e418f3ac6d23f2e80281cc33","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.760790870327035,"output_peak":0.5482940278213293},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.045846259730395,"spectral_convergence":0.7297760104711619,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.382001930063387,"spectral_convergence":0.3406112538849019,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.348774484987134,"spectral_convergence":0.3307714408386946,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":0.0001030074864251,"delta_db":0.938194925674469,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0007265692403231021,"delta_db":-0.123886162430356,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0010941228892872984,"delta_db":-0.572817348360314,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.0005271960129075477,"delta_db":-0.2588231110328853,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.00011631561472799663,"delta_db":-0.21418649861104772,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":8.801138156252625e-05,"delta_db":-2.041803513482735,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6916185932653011,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3853284021903657,"pitch":null,"output_rms_dbfs":-25.758990599470618,"output_peak":0.5482940278213293,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.925540891593639},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8369149385680353,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053866268327212564,"rms_dbfs":-25.37366219728025,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.789300578656956,"mean":-0.0032699441909790038,"zero_energy":false},"output":{"rms":0.05152885235703946,"rms_dbfs":-25.758990599470618,"peak":0.5482940278213293,"peak_dbfs":-5.21972969106163,"crest_factor_db":20.53926090840899,"mean":1.4162622275391266e-05,"zero_energy":false},"rms_difference_db":-0.38532840219036757},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":134,"both_estimated_voiced_frames":105,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.9051724137931034,"absolute_difference_cents_median":3.5799266127861413,"absolute_difference_cents_p95":13.007398808059335,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023508705421562093,"log_rms_mae_db":4.500907045755347,"rms_envelope_nrmse":0.21046556994159044,"envelope_correlation":0.9604880426882875,"rms_gain_error_db":-0.38507834406639296,"reference_mean_power":0.0029034514839332195,"output_mean_power":0.0026570929007938646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1197992634528758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005958091728481294,"output_power":0.0005713127631804978,"reference_active":true,"delta_db":-0.18233258812544334},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.320217055945433e-05,"output_power":7.936460526184351e-05,"reference_active":true,"delta_db":-0.6979916610587185},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9969137080619498e-05,"output_power":1.3567570025455508e-05,"reference_active":true,"delta_db":-1.6785722650526504}],"active_band_count":3,"active_band_mae_db":0.8529655047456041}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019484731874036573,"log_rms_mae_db":3.5713269751297676,"rms_envelope_nrmse":0.1160791315275477,"envelope_correlation":0.9884868821758866,"rms_gain_error_db":-0.3833195459762577,"reference_mean_power":0.002910052669271509,"output_mean_power":0.002664212703598161},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0516863068209195e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005686142820757539,"output_power":0.0005427949287885728,"reference_active":true,"delta_db":-0.2018198232107757},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3606311869244326e-05,"output_power":3.677469053958788e-05,"reference_active":true,"delta_db":-0.7400032969628341},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3652354435919545e-06,"output_power":9.388345272940625e-07,"reference_active":true,"delta_db":-1.6261850153422488}],"active_band_count":3,"active_band_mae_db":0.8560027118386196}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020607260287182458,"log_rms_mae_db":2.3827736719209622,"rms_envelope_nrmse":0.2193592669941115,"envelope_correlation":0.9634092375417296,"rms_gain_error_db":-0.3993775926888713,"reference_mean_power":0.001692101437139437,"output_mean_power":0.0015434360329975333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.505279098297708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003919860661962002,"output_power":0.00036355090771284497,"reference_active":true,"delta_db":-0.3270539624082594},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.427449102286661e-05,"output_power":4.174478953991705e-05,"reference_active":true,"delta_db":-0.2555130140429439},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.188127359577471e-06,"output_power":7.552941028858501e-06,"reference_active":true,"delta_db":-0.8511091259282648}],"active_band_count":3,"active_band_mae_db":0.4778920341264894}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016462000863024626,"log_rms_mae_db":1.6382543365905518,"rms_envelope_nrmse":0.12233238185097776,"envelope_correlation":0.988998918062838,"rms_gain_error_db":-0.3968722988580241,"reference_mean_power":0.001695759255771571,"output_mean_power":0.0015476650180299588},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.197777273221207e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003745482187229297,"output_power":0.0003482836836644972,"reference_active":true,"delta_db":-0.3157460681361841},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9699889570935678e-05,"output_power":1.842097415012311e-05,"reference_active":true,"delta_db":-0.2915119858551538},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294406999255858e-07,"output_power":7.242284273813383e-07,"reference_active":true,"delta_db":-0.5890977427624633}],"active_band_count":3,"active_band_mae_db":0.3987852655846004}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010781730421004015,"log_rms_mae_db":2.2946263864934147,"rms_envelope_nrmse":0.2297412700354979,"envelope_correlation":0.9604692066499202,"rms_gain_error_db":-0.2784618984759504,"reference_mean_power":0.0005546344869318722,"output_mean_power":0.000520188420490893},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0245820082367756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018054631551466004,"output_power":0.0001729181682808369,"reference_active":true,"delta_db":-0.18748003609286096},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2448913022603264e-05,"output_power":2.0429711960032982e-05,"reference_active":true,"delta_db":-0.4093307376300651},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.880443381781099e-06,"output_power":3.72284338721256e-06,"reference_active":true,"delta_db":-1.1758451219259913}],"active_band_count":3,"active_band_mae_db":0.5908852985496391}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010092628732279428,"log_rms_mae_db":1.3880279159828444,"rms_envelope_nrmse":0.1181790611676543,"envelope_correlation":0.9894657118289929,"rms_gain_error_db":-0.27721908493560776,"reference_mean_power":0.0005569522649053957,"output_mean_power":0.0005225117557656091},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.880183972268193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017516791642605378,"output_power":0.00016756724493212084,"reference_active":true,"delta_db":-0.19265434951207364},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.918962387005768e-06,"output_power":9.10995444430059e-06,"reference_active":true,"delta_db":-0.3695003818406193},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2471992219001627e-07,"output_power":2.7307570474162953e-07,"reference_active":true,"delta_db":-0.7522587110783359}],"active_band_count":3,"active_band_mae_db":0.43813781414367625}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001308702624055786,"log_rms_mae_db":2.2915252299015143,"rms_envelope_nrmse":0.3012558158399186,"envelope_correlation":0.9567522214097319,"rms_gain_error_db":-1.700410297535486,"reference_mean_power":0.00016841175600957621,"output_mean_power":0.00011384956470527415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6472123519121686e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.520693965195543e-05,"output_power":6.646160032778599e-05,"reference_active":true,"delta_db":-1.560978109073764},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1413816053513927e-05,"output_power":1.3052391940942854e-05,"reference_active":true,"delta_db":-2.1500396129905344},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.377484332308182e-06,"output_power":1.7674906256605904e-06,"reference_active":true,"delta_db":-4.8322203462979205}],"active_band_count":3,"active_band_mae_db":2.847746022787406}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":1.489846270068696,"rms_envelope_nrmse":0.23419652685031336,"envelope_correlation":0.9827990086352093,"rms_gain_error_db":-1.700591204356896,"reference_mean_power":0.00016921252708073628,"output_mean_power":0.0001143861363566848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5742070644083455e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.764878819980253e-05,"output_power":6.606344191686525e-05,"reference_active":true,"delta_db":-1.6970566139414056},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.033680531591512e-05,"output_power":6.163899944301061e-06,"reference_active":true,"delta_db":-2.245307564143334},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.467040891819502e-07,"output_power":1.2233464321294e-07,"reference_active":true,"delta_db":-4.524095042490249}],"active_band_count":3,"active_band_mae_db":2.822153073524996}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.7009027004241943,"predicted_mos":3.7009027004241943,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"201e81fab95ed13b0f1bd7f67c836d6c46545000727d65e27531b7350942d3b1","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.898639971999728},"started_at":"2026-10-13T05:10:53.408204+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032109150015458,"native_input_samples":144000,"native_input_sha256":"44c5b73265f29bdd68d89502b8b2fa2269e06027186d02b00495dd8a95796a3f","output_sha256":"862e64f88ead0fc3be76472e36dc84321dee6cf0e65cd21c3d516c35c207fbba","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0693771160003962,"synthesis_s":0.16020966699943529,"total_s":1.2295867829998315,"rtf":0.20493113049997191,"output_peak":0.5907770395278931,"output_rms":0.05185985772156668,"finished_at":"2026-10-13T05:10:54.651366+00:00","analysis_stages":{"external_f0_s":0.17693959200005338},"evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"862e64f88ead0fc3be76472e36dc84321dee6cf0e65cd21c3d516c35c207fbba","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.6979383518577,"output_peak":0.5915203895564529},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.843884832781031,"spectral_convergence":0.23366393274944933,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.426223695437049,"spectral_convergence":0.22505058947774692,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.96402207115584,"spectral_convergence":0.2523154684652134,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":7.729632196336973e-05,"delta_db":-0.30890468018672956,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0006958918069623868,"delta_db":-0.31123900695139367,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0011840939577001353,"delta_db":-0.2296167353457142,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.0005013194939082862,"delta_db":-0.477398356386963,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.00010584315264626467,"delta_db":-0.6239390285934057,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.0001327957260702652,"delta_db":-0.2553509193208264,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3677414544641721,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3171411469761934,"pitch":null,"output_rms_dbfs":-25.690803344256445,"output_peak":0.5915203895564529,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.744710199791307},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9017476541871899,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053866268327212564,"rms_dbfs":-25.37366219728025,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.789300578656956,"mean":-0.0032699441909790038,"zero_energy":false},"output":{"rms":0.05193496374554148,"rms_dbfs":-25.690803344256445,"peak":0.5915203895564529,"peak_dbfs":-4.56060561516774,"crest_factor_db":21.130197729088707,"mean":3.1280009816366665e-07,"zero_energy":false},"rms_difference_db":-0.3171411469761942},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":97,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.7586206896551724,"absolute_difference_cents_median":0.9776347839839903,"absolute_difference_cents_p95":4.393016599799298,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023508705421562093,"log_rms_mae_db":1.273889393893893,"rms_envelope_nrmse":0.08999068720943043,"envelope_correlation":0.9935958647740138,"rms_gain_error_db":-0.31705872604419866,"reference_mean_power":0.0029034514839332195,"output_mean_power":0.002699036146663353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1197992634528758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005958091728481294,"output_power":0.000559843856922506,"reference_active":true,"delta_db":-0.27040268034236503},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.320217055945433e-05,"output_power":8.327633687212051e-05,"reference_active":true,"delta_db":-0.48904413293383187},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9969137080619498e-05,"output_power":1.3106658334520768e-05,"reference_active":true,"delta_db":-1.828673199082045}],"active_band_count":3,"active_band_mae_db":0.8627066707860807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019484731874036573,"log_rms_mae_db":0.7848121544178336,"rms_envelope_nrmse":0.05732618770255341,"envelope_correlation":0.9979731878894542,"rms_gain_error_db":-0.3163225303540598,"reference_mean_power":0.002910052669271509,"output_mean_power":0.0027056311875397845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0516863068209195e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005686142820757539,"output_power":0.0005329338301509781,"reference_active":true,"delta_db":-0.2814447408013218},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3606311869244326e-05,"output_power":3.9536836355934204e-05,"reference_active":true,"delta_db":-0.42547441355312354},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3652354435919545e-06,"output_power":8.54576763502781e-07,"reference_active":true,"delta_db":-2.0345647480663227}],"active_band_count":3,"active_band_mae_db":0.9138279674735893}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020607260287182458,"log_rms_mae_db":1.0345222683840563,"rms_envelope_nrmse":0.089697110093319,"envelope_correlation":0.994441156918256,"rms_gain_error_db":-0.2685565733924654,"reference_mean_power":0.001692101437139437,"output_mean_power":0.0015906357421733611},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.505279098297708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003919860661962002,"output_power":0.00037438758819701747,"reference_active":true,"delta_db":-0.19949187115764433},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.427449102286661e-05,"output_power":3.644868303660102e-05,"reference_active":true,"delta_db":-0.8447173625091939},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.188127359577471e-06,"output_power":5.696195465838157e-06,"reference_active":true,"delta_db":-2.0764212269212203}],"active_band_count":3,"active_band_mae_db":1.0402101535293529}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016462000863024626,"log_rms_mae_db":0.5742953307611568,"rms_envelope_nrmse":0.05735835913338957,"envelope_correlation":0.9980245245863923,"rms_gain_error_db":-0.26788921145549066,"reference_mean_power":0.001695759255771571,"output_mean_power":0.0015943191956914605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.197777273221207e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003745482187229297,"output_power":0.00035523623519374294,"reference_active":true,"delta_db":-0.22990477248220229},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9699889570935678e-05,"output_power":1.5734358374601017e-05,"reference_active":true,"delta_db":-0.9761475402001647},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294406999255858e-07,"output_power":5.067886397586618e-07,"reference_active":true,"delta_db":-2.139584708812585}],"active_band_count":3,"active_band_mae_db":1.1152123404983172}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010781730421004015,"log_rms_mae_db":1.1886944687665817,"rms_envelope_nrmse":0.12853368721211617,"envelope_correlation":0.9894102069043533,"rms_gain_error_db":-0.4868661859080822,"reference_mean_power":0.0005546344869318722,"output_mean_power":0.0004958156739159673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0245820082367756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018054631551466004,"output_power":0.00016316412463280032,"reference_active":true,"delta_db":-0.4396395468895793},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2448913022603264e-05,"output_power":2.1097207453752075e-05,"reference_active":true,"delta_db":-0.26970343872982966},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.880443381781099e-06,"output_power":3.6149755853163494e-06,"reference_active":true,"delta_db":-1.3035391035127855}],"active_band_count":3,"active_band_mae_db":0.6709606963773981}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010092628732279428,"log_rms_mae_db":0.7856510295303311,"rms_envelope_nrmse":0.0828588974995709,"envelope_correlation":0.9967693720125882,"rms_gain_error_db":-0.48669866434207004,"reference_mean_power":0.0005569522649053957,"output_mean_power":0.0004979068577871719},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.880183972268193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017516791642605378,"output_power":0.00015803288369422794,"reference_active":true,"delta_db":-0.4470709925330568},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.918962387005768e-06,"output_power":9.557001321549022e-06,"reference_active":true,"delta_db":-0.16144597344202655},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2471992219001627e-07,"output_power":2.5195516029450574e-07,"reference_active":true,"delta_db":-1.1018567667314518}],"active_band_count":3,"active_band_mae_db":0.570124577568845}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001308702624055786,"log_rms_mae_db":1.1470484815297217,"rms_envelope_nrmse":0.14517850937618026,"envelope_correlation":0.987608920302883,"rms_gain_error_db":-0.3132880140764565,"reference_mean_power":0.00016841175600957621,"output_mean_power":0.000156690837852783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6472123519121686e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.520693965195543e-05,"output_power":8.915201087535528e-05,"reference_active":true,"delta_db":-0.2853746195970681},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1413816053513927e-05,"output_power":1.8802834674592244e-05,"reference_active":true,"delta_db":-0.564707402350445},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.377484332308182e-06,"output_power":3.2459855611483756e-06,"reference_active":true,"delta_db":-2.1923257005482717}],"active_band_count":3,"active_band_mae_db":1.014135907498595}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":0.789018475904945,"rms_envelope_nrmse":0.08035686262851278,"envelope_correlation":0.9966465898118084,"rms_gain_error_db":-0.3130930646836894,"reference_mean_power":0.00016921252708073628,"output_mean_power":0.00015744294508471032},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5742070644083455e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.764878819980253e-05,"output_power":9.012211306804225e-05,"reference_active":true,"delta_db":-0.3483549220463976},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.033680531591512e-05,"output_power":9.17484191343517e-06,"reference_active":true,"delta_db":-0.5178774691564941},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.467040891819502e-07,"output_power":1.9887155261463704e-07,"reference_active":true,"delta_db":-2.413862996077433}],"active_band_count":3,"active_band_mae_db":1.0933651290934414}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0004","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2681632041931152,"predicted_mos":3.2681632041931152,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288518,"input_sample_rate":48000,"input_seconds":6.010791666666667,"input_float32_sha256":"206e112d5881927aee80acde5af170918961214f6e1ac11a1bb77f7272ed8326","internal_samples":96173,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.384264970000004},"started_at":"2026-10-13T05:16:14.350696+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00871683699915593,"native_input_samples":288000,"native_input_sha256":"68ba450b6de4a0b2b0341bbb4472739ea4869b654496e095412896de7d84a310","output_sha256":"39f80475471532b732605021d3d0c20b066546cc17e751470521d26c9ac4de23","output_samples":288518,"length_delta_samples":518,"analysis_s":3.0217220920003456,"synthesis_s":0.9327273119997699,"total_s":3.9544494040001155,"rtf":0.659074900666686,"output_peak":0.6196175680549132,"output_rms":0.05233526478682131,"finished_at":"2026-10-13T05:16:18.326211+00:00","evaluation_input":{"reference":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"39f80475471532b732605021d3d0c20b066546cc17e751470521d26c9ac4de23","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96173,"length_delta_samples":173,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.618747569499195,"output_peak":0.6071181573798993},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982011583292608,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.6439586049037285,"spectral_convergence":0.5848356455451208,"spectral_convergence_status":"ok","amplitude_floor":1.7451878151186278e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.572730004236022,"spectral_convergence":0.41973542350679727,"spectral_convergence_status":"ok","amplitude_floor":9.668633122862388e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.204572593099491,"spectral_convergence":0.3738189553788826,"spectral_convergence_status":"ok","amplitude_floor":6.131486475721294e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.299449699297122e-05,"output_power":7.073237235046462e-05,"delta_db":-0.6943107265534583,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007475936848096536,"output_power":0.0006809580387346731,"delta_db":-0.4054527297309567,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001248382978500818,"output_power":0.0011653724296098714,"delta_db":-0.29883099165540694,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.000559570006980701,"output_power":0.0005535709525611022,"delta_db":-0.04681134824860044,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012219591005283435,"output_power":0.00012727818982005002,"delta_db":0.17697319909969672,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001408377861622848,"output_power":0.00014382900328286781,"delta_db":0.09127281106039972,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.28560863439141976,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2460735589734743,"pitch":null,"output_rms_dbfs":-25.61973575625373,"output_peak":0.6071181573798993,"reference_rms_dbfs":-25.37366219728025,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8070870674130814},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7512731804829206,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053866268327212564,"rms_dbfs":-25.37366219728025,"peak":0.58990478515625,"peak_dbfs":-4.584361618623298,"crest_factor_db":20.789300578656956,"mean":-0.0032699441909790038,"zero_energy":false},"output":{"rms":0.05236163658976894,"rms_dbfs":-25.61973575625373,"peak":0.6071181573798993,"peak_dbfs":-4.334535565503469,"crest_factor_db":21.28520019075026,"mean":-0.000639989463117707,"zero_energy":false},"rms_difference_db":-0.24607355897347816},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":5.526592305977829e-06,"threshold_dbfs":-52.57542571928947,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":116,"output_estimated_voiced_frames":116,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.8017241379310345,"absolute_difference_cents_median":2.630201872008785,"absolute_difference_cents_p95":16.846267928241975,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023508705421562093,"log_rms_mae_db":2.881686904652986,"rms_envelope_nrmse":0.2699176271237342,"envelope_correlation":0.9339704430981648,"rms_gain_error_db":-0.24967987776128758,"reference_mean_power":0.0029034514839332195,"output_mean_power":0.0027412370036918633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1197992634528758e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005958091728481294,"output_power":0.0005678310724767529,"reference_active":true,"delta_db":-0.20888031144992183},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.320217055945433e-05,"output_power":8.5298241543735e-05,"reference_active":true,"delta_db":-0.3848594852431287},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9969137080619498e-05,"output_power":2.043606026513071e-05,"reference_active":true,"delta_db":0.1003787646963894}],"active_band_count":3,"active_band_mae_db":0.23137285379647998}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019484731874036573,"log_rms_mae_db":1.9690709609770174,"rms_envelope_nrmse":0.17611716013443307,"envelope_correlation":0.9700988107238538,"rms_gain_error_db":-0.2506135764067297,"reference_mean_power":0.002910052669271509,"output_mean_power":0.0027468787632779546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0516863068209195e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005686142820757539,"output_power":0.0005388635651201719,"reference_active":true,"delta_db":-0.23338943737994688},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3606311869244326e-05,"output_power":4.124233899806685e-05,"reference_active":true,"delta_db":-0.24206068829592334},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3652354435919545e-06,"output_power":1.5273244761320275e-06,"reference_active":true,"delta_db":0.48723756897850123}],"active_band_count":3,"active_band_mae_db":0.3208958982181238}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020607260287182458,"log_rms_mae_db":3.380990985688953,"rms_envelope_nrmse":0.27105770263301804,"envelope_correlation":0.9437513438676552,"rms_gain_error_db":-0.34516965218918333,"reference_mean_power":0.001692101437139437,"output_mean_power":0.0015628216865756855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.505279098297708e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003919860661962002,"output_power":0.00036132100220739215,"reference_active":true,"delta_db":-0.35377423388982676},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.427449102286661e-05,"output_power":3.809120575691785e-05,"reference_active":true,"delta_db":-0.6532885702509397},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.188127359577471e-06,"output_power":7.968867381813634e-06,"reference_active":true,"delta_db":-0.6183040711915986}],"active_band_count":3,"active_band_mae_db":0.5417889584441217}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016462000863024626,"log_rms_mae_db":2.5438522423857552,"rms_envelope_nrmse":0.1851779313047779,"envelope_correlation":0.9726757168965295,"rms_gain_error_db":-0.3466137753942935,"reference_mean_power":0.001695759255771571,"output_mean_power":0.0015656793320915532},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.197777273221207e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003745482187229297,"output_power":0.0003447911177271703,"reference_active":true,"delta_db":-0.3595166669395543},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9699889570935678e-05,"output_power":1.6636154715042835e-05,"reference_active":true,"delta_db":-0.7341084108216779},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.294406999255858e-07,"output_power":6.920477547829919e-07,"reference_active":true,"delta_db":-0.786492780002045}],"active_band_count":3,"active_band_mae_db":0.6267059525877591}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010781730421004015,"log_rms_mae_db":2.9189117134212426,"rms_envelope_nrmse":0.30575075169734334,"envelope_correlation":0.9310527114758896,"rms_gain_error_db":-0.03805364001478935,"reference_mean_power":0.0005546344869318722,"output_mean_power":0.0005497959119920745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.0245820082367756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018054631551466004,"output_power":0.00017801480305005085,"reference_active":true,"delta_db":-0.061325118942193085},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2448913022603264e-05,"output_power":2.1928247527999173e-05,"reference_active":true,"delta_db":-0.10191392376056048},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.880443381781099e-06,"output_power":5.478644597893786e-06,"reference_active":true,"delta_db":0.5021384960240945}],"active_band_count":3,"active_band_mae_db":0.22179251290894939}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010092628732279428,"log_rms_mae_db":1.940997411817007,"rms_envelope_nrmse":0.2008848340744816,"envelope_correlation":0.9683814047567256,"rms_gain_error_db":-0.043801808760483604,"reference_mean_power":0.0005569522649053957,"output_mean_power":0.0005513632218132312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.880183972268193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017516791642605378,"output_power":0.0001727936974681208,"reference_active":true,"delta_db":-0.059266663511591364},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.918962387005768e-06,"output_power":9.822690019738564e-06,"reference_active":true,"delta_db":-0.04235804424493764},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2471992219001627e-07,"output_power":3.8609313795632485e-07,"reference_active":true,"delta_db":0.7518314872238896}],"active_band_count":3,"active_band_mae_db":0.2844853983268062}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001308702624055786,"log_rms_mae_db":2.964276752783467,"rms_envelope_nrmse":0.5095537054011529,"envelope_correlation":0.8460762553521246,"rms_gain_error_db":0.14421929784864018,"reference_mean_power":0.00016841175600957621,"output_mean_power":0.00017409822125818685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.6472123519121686e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.520693965195543e-05,"output_power":8.942878358233126e-05,"reference_active":true,"delta_db":-0.27191281862480027},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1413816053513927e-05,"output_power":2.0935959080059977e-05,"reference_active":true,"delta_db":-0.09801206937619648},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.377484332308182e-06,"output_power":6.517117959741415e-06,"reference_active":true,"delta_db":0.8347642807784583}],"active_band_count":3,"active_band_mae_db":0.4015630562598183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.575610534879347e-05,"log_rms_mae_db":2.0245773353055885,"rms_envelope_nrmse":0.32340161104829496,"envelope_correlation":0.9353745986909262,"rms_gain_error_db":0.14391660556496583,"reference_mean_power":0.00016921252708073628,"output_mean_power":0.00017491383908812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5742070644083455e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.764878819980253e-05,"output_power":9.252837302632879e-05,"reference_active":true,"delta_db":-0.23391932344595406},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.033680531591512e-05,"output_power":1.0508677714333073e-05,"reference_active":true,"delta_db":0.07161736250367476},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.467040891819502e-07,"output_power":4.470883864019886e-07,"reference_active":true,"delta_db":1.1043442480333097}],"active_band_count":3,"active_band_mae_db":0.4699603113276461}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"hnm","status":"measured","mos":{"status":"measured","value":1.9651844501495361,"predicted_mos":1.9651844501495361,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84640,"input_sample_rate":16000,"input_seconds":5.29,"input_float32_sha256":"a4181bb8fd8dee9a554c9bd9c9730f5626f5a4f2ca3269e9b6f6a8d2393417d9","internal_samples":84640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.330857176999416},"started_at":"2026-10-13T04:39:31.411153+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5001000065240078e-05,"native_input_samples":84640,"native_input_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","output_sha256":"ac6f1853e7bcca7cf7e11e31c5445dd012bb87f53e07f304e0bfe852dca8b440","output_samples":84640,"length_delta_samples":0,"analysis_s":0.9349711920003756,"synthesis_s":0.6301161820001653,"total_s":1.5650873740005409,"rtf":0.29585772665416654,"output_peak":0.5884663073557218,"output_rms":0.031923715313593205,"finished_at":"2026-10-13T04:39:32.981528+00:00","evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"ac6f1853e7bcca7cf7e11e31c5445dd012bb87f53e07f304e0bfe852dca8b440","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.279969259371028,"spectral_convergence":0.39023165646135344,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.747414255441669,"spectral_convergence":0.34511540254321266,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9701265487931545,"spectral_convergence":0.3870557358629913,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":7.86717272210283e-05,"delta_db":0.32983922655321957,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.00025271129274640847,"delta_db":-0.05891832845908673,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.0003768600572720146,"delta_db":-0.09101765355108235,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.00023773584762191427,"delta_db":-0.586036447348104,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.656974244076999e-05,"delta_db":0.05800043772124987,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":2.657493212120966e-05,"delta_db":0.03392334753796738,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19295590686178501,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16391658277535862,"pitch":null,"output_rms_dbfs":-29.917731415163182,"output_peak":0.5884663073557218,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.999170021201951},"common_prefix":{"samples":84640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.279969259371028,"spectral_convergence":0.39023165646135344,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.747414255441669,"spectral_convergence":0.34511540254321266,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9701265487931545,"spectral_convergence":0.3870557358629913,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":7.86717272210283e-05,"delta_db":0.32983922655321957,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.00025271129274640847,"delta_db":-0.05891832845908673,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.0003768600572720146,"delta_db":-0.09101765355108235,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.00023773584762191427,"delta_db":-0.586036447348104,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.656974244076999e-05,"delta_db":0.05800043772124987,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":2.657493212120966e-05,"delta_db":0.03392334753796738,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19295590686178501,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16391658277535862,"pitch":null,"output_rms_dbfs":-29.917731415163182,"output_peak":0.5884663073557218,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.999170021201951},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84640,"output_samples":84640,"intelligibility":{"estoi":{"value":0.8331453451853934,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03253188724556978,"rms_dbfs":-29.753814832387818,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.102693866300093,"mean":-0.0032656952654256262,"zero_energy":false},"output":{"rms":0.031923715313593205,"rms_dbfs":-29.917731415163182,"peak":0.5884663073557218,"peak_dbfs":-4.605567952872292,"crest_factor_db":25.312163462290886,"mean":-0.00021822582263376455,"zero_energy":false},"rms_difference_db":-0.16391658277536436},"full_scale":{"reference":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":529,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":522,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":88,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.8780487804878049,"absolute_difference_cents_median":1.2492889128903273,"absolute_difference_cents_p95":5.798536532432813,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00018085971712132593,"log_rms_mae_db":1.2841603305593785,"rms_envelope_nrmse":0.09248533605115376,"envelope_correlation":0.9925543335158971,"rms_gain_error_db":-0.16399871576777855,"reference_mean_power":0.0010593154039372696,"output_mean_power":0.001020059291281503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":5.491358535179395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002450078451267542,"output_power":0.00022706774672984605,"reference_active":true,"delta_db":-0.3302454032609763},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.385267752666162e-05,"output_power":8.259965571138127e-05,"reference_active":true,"delta_db":-0.06538697644432148},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4530628910939782e-05,"output_power":1.2100571852952786e-05,"reference_active":true,"delta_db":-0.7947851688856045}],"active_band_count":3,"active_band_mae_db":0.3968058495303007}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00014864066644098504,"log_rms_mae_db":0.6433859419793871,"rms_envelope_nrmse":0.054543929172262924,"envelope_correlation":0.9973901382424853,"rms_gain_error_db":-0.1640568169089374,"reference_mean_power":0.0010653029350414017,"output_mean_power":0.0010258112127606803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":4.840611714703517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022860940218938906,"output_power":0.00021087149144688897,"reference_active":true,"delta_db":-0.35076218264772524},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8392485325428115e-05,"output_power":3.848467992230756e-05,"reference_active":true,"delta_db":0.010416519327131194},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1188041388515457e-06,"output_power":9.452079096646465e-07,"reference_active":true,"delta_db":-0.7322671708358442}],"active_band_count":3,"active_band_mae_db":0.3644819576035669}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013930286763496605,"log_rms_mae_db":1.7617659116019706,"rms_envelope_nrmse":0.12244180905406786,"envelope_correlation":0.9889761947425362,"rms_gain_error_db":-0.0851971957323994,"reference_mean_power":0.0005230599489297363,"output_mean_power":0.0005128988761960793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":3.2510596296193775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001486768204336917,"output_power":0.0001334473466480564,"reference_active":true,"delta_db":-0.4693332173440529},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.417268122003869e-05,"output_power":5.389508330627145e-05,"reference_active":true,"delta_db":-0.02231183514036032},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.81005033549118e-06,"output_power":6.296218765530686e-06,"reference_active":true,"delta_db":-0.935740234883773}],"active_band_count":3,"active_band_mae_db":0.47579509578939544}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00011121771745577188,"log_rms_mae_db":1.0663710840193081,"rms_envelope_nrmse":0.0818580562249667,"envelope_correlation":0.9946798762347284,"rms_gain_error_db":-0.08519892330060733,"reference_mean_power":0.0005260459982805144,"output_mean_power":0.000515826712733908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.9069199937213013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001421546332212628,"output_power":0.000128532948466933,"reference_active":true,"delta_db":-0.4374654903727543},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2639933776267854e-05,"output_power":2.3720938695156805e-05,"reference_active":true,"delta_db":0.20256718949941416},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.874281484444507e-07,"output_power":5.668652613217308e-07,"reference_active":true,"delta_db":-0.8374746808585367}],"active_band_count":3,"active_band_mae_db":0.49250245357690176}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010000130422823664,"log_rms_mae_db":1.9979265347204564,"rms_envelope_nrmse":0.1534058442022976,"envelope_correlation":0.985210709637763,"rms_gain_error_db":-0.5151169060353422,"reference_mean_power":0.0002857956004291323,"output_mean_power":0.00025383052517335844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.1492733373341918e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047160855442973e-05,"output_power":8.677257737959856e-05,"reference_active":true,"delta_db":-0.18129814632605235},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0483248697121172e-05,"output_power":1.8268954867566793e-05,"reference_active":true,"delta_db":-0.49685135118417906},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.217073445558823e-06,"output_power":7.752265710055957e-06,"reference_active":true,"delta_db":-0.25288518981200814}],"active_band_count":3,"active_band_mae_db":0.3103448957740798}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":1.2613452049080822,"rms_envelope_nrmse":0.09803382268510166,"envelope_correlation":0.9950818760195022,"rms_gain_error_db":-0.5151181799662726,"reference_mean_power":0.0002874271047267671,"output_mean_power":0.0002552794774704487},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":1.945560884297879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.432816729085125e-05,"output_power":8.033761673949015e-05,"reference_active":true,"delta_db":-0.21053717884394096},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0296492874768448e-05,"output_power":9.169228815378713e-06,"reference_active":true,"delta_db":-0.5035651274249228},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.97290149260777e-07,"output_power":4.82198813443537e-07,"reference_active":true,"delta_db":-0.9295921488879235}],"active_band_count":3,"active_band_mae_db":0.5478981517189291}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":2.20841033044255,"rms_envelope_nrmse":0.18358959952312898,"envelope_correlation":0.9787521077755453,"rms_gain_error_db":-0.12058170881794078,"reference_mean_power":3.470766577035855e-05,"output_mean_power":3.3757263656218245e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.6534989625639412e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5332454345272337e-05,"output_power":1.5007041018961613e-05,"reference_active":true,"delta_db":-0.09316610579459286},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8283604501578115e-06,"output_power":3.1481318001043707e-06,"reference_active":true,"delta_db":-0.8495991469404106},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1450631921352038e-06,"output_power":5.709456912737595e-07,"reference_active":true,"delta_db":-3.022346547333396}],"active_band_count":3,"active_band_mae_db":1.3217039333561331}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":1.48297545408046,"rms_envelope_nrmse":0.10555012331447094,"envelope_correlation":0.9926363877360761,"rms_gain_error_db":-0.1205844227518975,"reference_mean_power":3.4905778716849276e-05,"output_mean_power":3.394993044835989e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.4649629087328264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5574047649212616e-05,"output_power":1.4909554863893168e-05,"reference_active":true,"delta_db":-0.18936821643382104},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.044495719000319e-06,"output_power":1.5318914788555434e-06,"reference_active":true,"delta_db":-1.253582051039352},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0114195457715016e-07,"output_power":4.392900271869897e-08,"reference_active":true,"delta_db":-3.6217999818293634}],"active_band_count":3,"active_band_mae_db":1.6882500831008453}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.797121286392212,"predicted_mos":2.797121286392212,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84640,"input_sample_rate":16000,"input_seconds":5.29,"input_float32_sha256":"39281f53b3215e71e9f401ba3b02c3cdb903c4a94478f9d6467f7d3a2669abc8","internal_samples":84640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.928702011999121},"started_at":"2026-10-13T04:42:30.835769+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3733000489301048e-05,"native_input_samples":84640,"native_input_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","output_sha256":"59de015fe3f914050257c252aaa02c9ee23862557e7418597c94dbacd423a73c","output_samples":84640,"length_delta_samples":0,"analysis_s":1.6661984999991546,"synthesis_s":0.07951148500069394,"total_s":1.7457099849998485,"rtf":0.3300018875236009,"output_peak":0.6772026241035454,"output_rms":0.03442570149905946,"finished_at":"2026-10-13T04:42:32.586620+00:00","evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"59de015fe3f914050257c252aaa02c9ee23862557e7418597c94dbacd423a73c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.533532405894449,"spectral_convergence":0.7177258418245531,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.551548798497923,"spectral_convergence":0.2929370394297771,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.700635619973963,"spectral_convergence":0.3231985614297291,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":5.9867269385900544e-05,"delta_db":-0.8564531252817951,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.0002978796292977193,"delta_db":0.6552432246856955,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.00045371450080213937,"delta_db":0.7150078447920766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.0003028368297993578,"delta_db":0.465103735605063,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.770799134136858e-05,"delta_db":0.16287339508217175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":2.3122703075859515e-05,"delta_db":-0.5704122681026981,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5708489322582501,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.49147080679364546,"pitch":null,"output_rms_dbfs":-29.262344025594174,"output_peak":0.6772026241035454,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.261905608122112},"common_prefix":{"samples":84640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.533532405894449,"spectral_convergence":0.7177258418245531,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.551548798497923,"spectral_convergence":0.2929370394297771,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.700635619973963,"spectral_convergence":0.3231985614297291,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":5.9867269385900544e-05,"delta_db":-0.8564531252817951,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.0002978796292977193,"delta_db":0.6552432246856955,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.00045371450080213937,"delta_db":0.7150078447920766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.0003028368297993578,"delta_db":0.465103735605063,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.770799134136858e-05,"delta_db":0.16287339508217175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":2.3122703075859515e-05,"delta_db":-0.5704122681026981,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5708489322582501,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.49147080679364546,"pitch":null,"output_rms_dbfs":-29.262344025594174,"output_peak":0.6772026241035454,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.261905608122112},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84640,"output_samples":84640,"intelligibility":{"estoi":{"value":0.9030893983166549,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03253188724556978,"rms_dbfs":-29.753814832387818,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.102693866300093,"mean":-0.0032656952654256262,"zero_energy":false},"output":{"rms":0.03442570149905946,"rms_dbfs":-29.262344025594174,"peak":0.6772026241035454,"peak_dbfs":-3.3856273541007447,"crest_factor_db":25.876716671493426,"mean":-2.77398091908178e-09,"zero_energy":false},"rms_difference_db":0.4914708067936431},"full_scale":{"reference":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":529,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":522,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":73,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.8902439024390244,"absolute_difference_cents_median":5.3986331582883675,"absolute_difference_cents_p95":14.674447042343386,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00018085971712132593,"log_rms_mae_db":4.734158274276519,"rms_envelope_nrmse":0.17549019706751293,"envelope_correlation":0.9781104920928335,"rms_gain_error_db":0.49150983787523794,"reference_mean_power":0.0010593154039372696,"output_mean_power":0.001186250125466083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":5.491358535179395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002450078451267542,"output_power":0.00028738755924228654,"reference_active":true,"delta_db":0.6928797332122809},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.385267752666162e-05,"output_power":8.372372202271291e-05,"reference_active":true,"delta_db":-0.006684076633885779},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4530628910939782e-05,"output_power":1.3137596649412873e-05,"reference_active":true,"delta_db":-0.4376848770267486}],"active_band_count":3,"active_band_mae_db":0.3790828956243051}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00014864066644098504,"log_rms_mae_db":3.400499995092839,"rms_envelope_nrmse":0.09417134488538557,"envelope_correlation":0.9957643642930936,"rms_gain_error_db":0.4917536749918597,"reference_mean_power":0.0010653029350414017,"output_mean_power":0.0011930221062477365},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":4.840611714703517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022860940218938906,"output_power":0.0002640309785260026,"reference_active":true,"delta_db":0.6256079727506436},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8392485325428115e-05,"output_power":3.8583728860461704e-05,"reference_active":true,"delta_db":0.02157970062084391},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1188041388515457e-06,"output_power":1.0094669498143795e-06,"reference_active":true,"delta_db":-0.4466195970226996}],"active_band_count":3,"active_band_mae_db":0.364602423464729}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013930286763496605,"log_rms_mae_db":2.0402771844629783,"rms_envelope_nrmse":0.18283402966214762,"envelope_correlation":0.982535519181964,"rms_gain_error_db":0.7238700255485224,"reference_mean_power":0.0005230599489297363,"output_mean_power":0.0006179288464345517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":3.2510596296193775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001486768204336917,"output_power":0.00017584219918198915,"reference_active":true,"delta_db":0.7287984183890872},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.417268122003869e-05,"output_power":5.564928173723625e-05,"reference_active":true,"delta_db":0.11679232218622596},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.81005033549118e-06,"output_power":5.869124000331963e-06,"reference_active":true,"delta_db":-1.2408054770065373}],"active_band_count":3,"active_band_mae_db":0.6954654058606168}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00011121771745577188,"log_rms_mae_db":1.3899886491751268,"rms_envelope_nrmse":0.11332999162858302,"envelope_correlation":0.9961212874834681,"rms_gain_error_db":0.7238699542301188,"reference_mean_power":0.0005260459982805144,"output_mean_power":0.0006214564739971038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.9069199937213013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001421546332212628,"output_power":0.00016581752572611156,"reference_active":true,"delta_db":0.6686941131797285},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2639933776267854e-05,"output_power":2.3491667590707998e-05,"reference_active":true,"delta_db":0.16038694737098944},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.874281484444507e-07,"output_power":4.913524328030158e-07,"reference_active":true,"delta_db":-1.458342007373182}],"active_band_count":3,"active_band_mae_db":0.7624743559746333}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010000130422823664,"log_rms_mae_db":1.8975990678608832,"rms_envelope_nrmse":0.18520028130596428,"envelope_correlation":0.9794215131740998,"rms_gain_error_db":0.46101217574764203,"reference_mean_power":0.0002857956004291323,"output_mean_power":0.00031780209550107764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.1492733373341918e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047160855442973e-05,"output_power":0.00010023699279082071,"reference_active":true,"delta_db":0.4451571692290704},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0483248697121172e-05,"output_power":2.0260581305412256e-05,"reference_active":true,"delta_db":-0.04746936273328428},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.217073445558823e-06,"output_power":6.719686894798268e-06,"reference_active":true,"delta_db":-0.8736813130275339}],"active_band_count":3,"active_band_mae_db":0.45543594832996287}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":1.1894527087063829,"rms_envelope_nrmse":0.08577362157154947,"envelope_correlation":0.9969124985989216,"rms_gain_error_db":0.46101234298517757,"reference_mean_power":0.0002874271047267671,"output_mean_power":0.00031961632567581585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":1.945560884297879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.432816729085125e-05,"output_power":9.201683485700733e-05,"reference_active":true,"delta_db":0.37894628688489596},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0296492874768448e-05,"output_power":9.845119794710313e-06,"reference_active":true,"delta_db":-0.19468318326627077},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.97290149260777e-07,"output_power":4.580682550721116e-07,"reference_active":true,"delta_db":-1.1525515679931155}],"active_band_count":3,"active_band_mae_db":0.5753936793814275}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":2.216881971951358,"rms_envelope_nrmse":0.23576926480064048,"envelope_correlation":0.9647800332703218,"rms_gain_error_db":-0.23916988091118516,"reference_mean_power":3.470766577035855e-05,"output_mean_power":3.284796116062499e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.6534989625639412e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5332454345272337e-05,"output_power":1.410489490117164e-05,"reference_active":true,"delta_db":-0.3624182570822843},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8283604501578115e-06,"output_power":3.579684446285466e-06,"reference_active":true,"delta_db":-0.2916807626928223},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1450631921352038e-06,"output_power":9.901025709078366e-07,"reference_active":true,"delta_db":-0.6314926636422046}],"active_band_count":3,"active_band_mae_db":0.42853056113910376}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":1.3655711643733295,"rms_envelope_nrmse":0.13501333169009197,"envelope_correlation":0.9884869666407708,"rms_gain_error_db":-0.239169872013421,"reference_mean_power":3.4905778716849276e-05,"output_mean_power":3.3035458895675967e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.4649629087328264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5574047649212616e-05,"output_power":1.4209755323852737e-05,"reference_active":true,"delta_db":-0.398148991726949},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.044495719000319e-06,"output_power":1.870382523296663e-06,"reference_active":true,"delta_db":-0.386557696915612},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0114195457715016e-07,"output_power":8.22683810077189e-08,"reference_active":true,"delta_db":-0.8969839142063625}],"active_band_count":3,"active_band_mae_db":0.5605635342829745}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"facodec","status":"measured","mos":{"status":"measured","value":3.674316167831421,"predicted_mos":3.674316167831421,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84800,"input_sample_rate":16000,"input_seconds":5.3,"input_float32_sha256":"2cea92ae298af7802c48b047e7176100b91c2df3b327dddc8df79e2a26e7738e","internal_samples":84800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0642711950004013},"started_at":"2026-10-13T04:54:11.243964+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.8438000299502164e-05,"native_input_samples":84640,"native_input_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","output_sha256":"d16589c4b2c54543111b991c68dd847f75a4dbe15ea17d084d4bd7a79642b972","output_samples":84800,"length_delta_samples":160,"analysis_s":1.335093823000534,"synthesis_s":11.322164878999502,"total_s":12.657258702000036,"rtf":2.3926765032136172,"output_peak":0.5473320484161377,"output_rms":0.031926899359289815,"finished_at":"2026-10-13T04:54:23.907274+00:00","evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"d16589c4b2c54543111b991c68dd847f75a4dbe15ea17d084d4bd7a79642b972","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":84640,"output_samples":84800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.916865134715046,"output_peak":0.5473320484161377},"common_prefix":{"samples":84640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981132075471698,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.517823960734416,"spectral_convergence":0.290669089049409,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.600183508814634,"spectral_convergence":0.3056901081001661,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.618296815613352,"spectral_convergence":0.3390532221251547,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":7.126655374486049e-05,"delta_db":-0.09949004447073036,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.0002858453474405713,"delta_db":0.47614645580664994,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.0003573174906927326,"delta_db":-0.32227599703691134,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.0002274114651109491,"delta_db":-0.7788596146147619,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.4214052845143674e-05,"delta_db":-0.16743466271978175,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":3.519873811247662e-05,"delta_db":1.2544726499254046,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5164465707623734,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15484895905879867,"pitch":null,"output_rms_dbfs":-29.90866379144662,"output_peak":0.5473320484161377,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5787680950541345},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84640,"output_samples":84640,"intelligibility":{"estoi":{"value":0.8943880572093724,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03253188724556978,"rms_dbfs":-29.753814832387818,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.102693866300093,"mean":-0.0032656952654256262,"zero_energy":false},"output":{"rms":0.031957059438357804,"rms_dbfs":-29.90866379144662,"peak":0.5473320484161377,"peak_dbfs":-5.2349824297029866,"crest_factor_db":24.673681361743633,"mean":-0.0002964370044689049,"zero_energy":false},"rms_difference_db":-0.15484895905880336},"full_scale":{"reference":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":529,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":522,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":93,"both_estimated_voiced_frames":76,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.926829268292683,"absolute_difference_cents_median":3.6291964546076496,"absolute_difference_cents_p95":11.208443255542669,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00018085971712132593,"log_rms_mae_db":2.8665540037371913,"rms_envelope_nrmse":0.1093389560332347,"envelope_correlation":0.9894515216318773,"rms_gain_error_db":-0.15481662898849768,"reference_mean_power":0.0010593154039372696,"output_mean_power":0.001022218236799608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":5.491358535179395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002450078451267542,"output_power":0.00024131546623045605,"reference_active":true,"delta_db":-0.06594833339097216},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.385267752666162e-05,"output_power":8.57769461163025e-05,"reference_active":true,"delta_db":0.09853645399003527},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4530628910939782e-05,"output_power":1.8293019534177843e-05,"reference_active":true,"delta_db":1.0000098641202788}],"active_band_count":3,"active_band_mae_db":0.3881648838337621}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00014864066644098504,"log_rms_mae_db":2.173075077154234,"rms_envelope_nrmse":0.09364205478698504,"envelope_correlation":0.9915583826979819,"rms_gain_error_db":-0.15463702779911653,"reference_mean_power":0.0010653029350414017,"output_mean_power":0.0010280385982449913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":4.840611714703517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022860940218938906,"output_power":0.00022548111755309424,"reference_active":true,"delta_db":-0.05983909330433883},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8392485325428115e-05,"output_power":3.965815643146712e-05,"reference_active":true,"delta_db":0.14086294568267943},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1188041388515457e-06,"output_power":1.402870857989437e-06,"reference_active":true,"delta_db":0.9826362935162962}],"active_band_count":3,"active_band_mae_db":0.39444611083443815}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013930286763496605,"log_rms_mae_db":1.3570043121296997,"rms_envelope_nrmse":0.11106575692403022,"envelope_correlation":0.9910028998685424,"rms_gain_error_db":-0.0454233647208424,"reference_mean_power":0.0005230599489297363,"output_mean_power":0.0005176177142138126},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":3.2510596296193775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001486768204336917,"output_power":0.00014919551297239912,"reference_active":true,"delta_db":0.015124971995464611},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.417268122003869e-05,"output_power":5.395072130912124e-05,"reference_active":true,"delta_db":-0.017830755908453406},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.81005033549118e-06,"output_power":1.0366660359091941e-05,"reference_active":true,"delta_db":1.2298503716309988}],"active_band_count":3,"active_band_mae_db":0.42093536651163893}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00011121771745577188,"log_rms_mae_db":1.02807424285056,"rms_envelope_nrmse":0.09561746311981649,"envelope_correlation":0.9928028949851339,"rms_gain_error_db":-0.04542364754542615,"reference_mean_power":0.0005260459982805144,"output_mean_power":0.0005205726609848604},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.9069199937213013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001421546332212628,"output_power":0.0001425157218503325,"reference_active":true,"delta_db":0.011017577376349742},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2639933776267854e-05,"output_power":2.263930365618555e-05,"reference_active":true,"delta_db":-0.00012087556691175667},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.874281484444507e-07,"output_power":8.427301327576037e-07,"reference_active":true,"delta_db":0.8846121145718686}],"active_band_count":3,"active_band_mae_db":0.2985835225050434}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010000130422823664,"log_rms_mae_db":1.5200697449496743,"rms_envelope_nrmse":0.15227490246832978,"envelope_correlation":0.9872949804218762,"rms_gain_error_db":-0.712142215708365,"reference_mean_power":0.0002857956004291323,"output_mean_power":0.00024257236215381785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.1492733373341918e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047160855442973e-05,"output_power":7.86495833819993e-05,"reference_active":true,"delta_db":-0.6081588554760474},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0483248697121172e-05,"output_power":2.0755506995684652e-05,"reference_active":true,"delta_db":0.057345083607338525},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.217073445558823e-06,"output_power":9.803696495063876e-06,"reference_active":true,"delta_db":0.766726890198134}],"active_band_count":3,"active_band_mae_db":0.4774102764271733}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":1.175867389439024,"rms_envelope_nrmse":0.13238537319199883,"envelope_correlation":0.9910297619038496,"rms_gain_error_db":-0.712143946227032,"reference_mean_power":0.0002874271047267671,"output_mean_power":0.00024395702332044236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":1.945560884297879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.432816729085125e-05,"output_power":7.509089462494741e-05,"reference_active":true,"delta_db":-0.5038538338441847},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0296492874768448e-05,"output_power":1.08098141086081e-05,"reference_active":true,"delta_db":0.21128902358481189},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.97290149260777e-07,"output_power":7.824218452019644e-07,"reference_active":true,"delta_db":1.172556151876538}],"active_band_count":3,"active_band_mae_db":0.6292330031018448}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":2.1387432583879953,"rms_envelope_nrmse":0.25481897768455253,"envelope_correlation":0.9752688108078638,"rms_gain_error_db":1.0421795072591329,"reference_mean_power":3.470766577035855e-05,"output_mean_power":4.412079785147346e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.6534989625639412e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5332454345272337e-05,"output_power":2.030400842318167e-05,"reference_active":true,"delta_db":1.2197010473587915},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8283604501578115e-06,"output_power":5.292987258236364e-06,"reference_active":true,"delta_db":1.406880275268875},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1450631921352038e-06,"output_power":1.3878468219614402e-06,"reference_active":true,"delta_db":0.8351208067891285}],"active_band_count":3,"active_band_mae_db":1.1539007098055984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":1.8094669424406615,"rms_envelope_nrmse":0.20905716553422712,"envelope_correlation":0.9862848485869323,"rms_gain_error_db":1.0421760096650652,"reference_mean_power":3.4905778716849276e-05,"output_mean_power":4.4372605651656615e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.4649629087328264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5574047649212616e-05,"output_power":2.0817114293510512e-05,"reference_active":true,"delta_db":1.2601902755310062},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.044495719000319e-06,"output_power":2.7256614392737674e-06,"reference_active":true,"delta_db":1.24885704608563},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0114195457715016e-07,"output_power":1.0960848038678951e-07,"reference_active":true,"delta_db":0.34912814571192047}],"active_band_count":3,"active_band_mae_db":0.9527251557761857}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.154317617416382,"predicted_mos":3.154317617416382,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":233472,"input_sample_rate":44100,"input_seconds":5.294149659863946,"input_float32_sha256":"78753cd417ed10ec49ec4666d304978aac78c7bcc980b7326439b93001ee6963","internal_samples":84707,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.722924552999757},"started_at":"2026-10-13T07:17:50.280547+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.010737923001215677,"native_input_samples":233289,"native_input_sha256":"55e9c0e2bc4ab190de5e9722edc1698b0760a55747bb484952fc3b2d7ae39a7f","output_sha256":"5da397f3330b9c59913866bae766c6a57c5683dc9a3586e1c5f08624175f2b66","output_samples":233472,"length_delta_samples":183,"analysis_s":0.15597173000060138,"synthesis_s":31.581842645000506,"total_s":31.737814375001108,"rtf":5.9995868383745,"output_peak":0.5320368409156799,"output_rms":0.03039350301553908,"finished_at":"2026-10-13T07:18:22.049240+00:00","evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"5da397f3330b9c59913866bae766c6a57c5683dc9a3586e1c5f08624175f2b66","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":84640,"output_samples":84707,"length_delta_samples":67,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.337989424070877,"output_peak":0.5337357521699808},"common_prefix":{"samples":84640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9992090382140791,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.113453376514821,"spectral_convergence":0.4376356082715922,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.063772109976181,"spectral_convergence":0.4347904324253091,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.652022491098169,"spectral_convergence":0.4758948107696031,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":5.808705739201417e-05,"delta_db":-0.9875538617074575,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.00021712363256856714,"delta_db":-0.7180938593308385,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.0003691363704677512,"delta_db":-0.18095037438051575,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.00022081015787052264,"delta_db":-0.9067926938131385,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":3.960080307513435e-05,"delta_db":-0.6459979906527167,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":2.10999394598247e-05,"delta_db":-0.9679861993299381,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7345624965357676,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5807415163181546,"pitch":null,"output_rms_dbfs":-30.334556348705974,"output_peak":0.5337357521699808,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.94308265919639},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84640,"output_samples":84640,"intelligibility":{"estoi":{"value":0.726163831337194,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03253188724556978,"rms_dbfs":-29.753814832387818,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.102693866300093,"mean":-0.0032656952654256262,"zero_energy":false},"output":{"rms":0.03042791417159274,"rms_dbfs":-30.334556348705974,"peak":0.5337357521699808,"peak_dbfs":-5.453474102066088,"crest_factor_db":24.881082246639888,"mean":-5.35124380058249e-05,"zero_energy":false},"rms_difference_db":-0.580741516318156},"full_scale":{"reference":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":529,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":522,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":111,"both_estimated_voiced_frames":69,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":42,"paired_voiced_fraction":0.8414634146341463,"absolute_difference_cents_median":4.932042336550029,"absolute_difference_cents_p95":19.62471515183576,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00018085971712132593,"log_rms_mae_db":2.044598816678687,"rms_envelope_nrmse":0.12077278994490019,"envelope_correlation":0.9910527426068034,"rms_gain_error_db":-0.5807232772215971,"reference_mean_power":0.0010593154039372696,"output_mean_power":0.0009267294402391069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":5.491358535179395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002450078451267542,"output_power":0.00023166902288801995,"reference_active":true,"delta_db":-0.24312023709051672},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.385267752666162e-05,"output_power":7.638462361287604e-05,"reference_active":true,"delta_db":-0.40510991788132383},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4530628910939782e-05,"output_power":1.1373317245727809e-05,"reference_active":true,"delta_db":-1.0639725826190967}],"active_band_count":3,"active_band_mae_db":0.5707342458636457}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00014864066644098504,"log_rms_mae_db":1.5966788926438042,"rms_envelope_nrmse":0.08667819015269286,"envelope_correlation":0.9973376828836269,"rms_gain_error_db":-0.5805488295085814,"reference_mean_power":0.0010653029350414017,"output_mean_power":0.000932004996466261},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":4.840611714703517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022860940218938906,"output_power":0.0002150884794859111,"reference_active":true,"delta_db":-0.2647693853912738},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8392485325428115e-05,"output_power":3.632478131303403e-05,"reference_active":true,"delta_db":-0.24043218559442803},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1188041388515457e-06,"output_power":9.388249938415997e-07,"reference_active":true,"delta_db":-0.7616942123818439}],"active_band_count":3,"active_band_mae_db":0.4222985944558486}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013930286763496605,"log_rms_mae_db":3.04212260119694,"rms_envelope_nrmse":0.1529462339698202,"envelope_correlation":0.982785415540594,"rms_gain_error_db":-0.06301192532886196,"reference_mean_power":0.0005230599489297363,"output_mean_power":0.0005155256452933501},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":3.2510596296193775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001486768204336917,"output_power":0.00014847584315751908,"reference_active":true,"delta_db":-0.0058746460007350135},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.417268122003869e-05,"output_power":5.016699593698881e-05,"reference_active":true,"delta_db":-0.3336223536418873},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.81005033549118e-06,"output_power":6.108633676308504e-06,"reference_active":true,"delta_db":-1.0670975083867749}],"active_band_count":3,"active_band_mae_db":0.46886483600979906}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00011121771745577188,"log_rms_mae_db":2.4156772299464895,"rms_envelope_nrmse":0.11151987057166123,"envelope_correlation":0.990066816747754,"rms_gain_error_db":-0.06301336279158144,"reference_mean_power":0.0005260459982805144,"output_mean_power":0.0005184685111367264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.9069199937213013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001421546332212628,"output_power":0.0001420793670324232,"reference_active":true,"delta_db":-0.002300054970325717},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2639933776267854e-05,"output_power":2.2075582572645955e-05,"reference_active":true,"delta_db":-0.10962978776491722},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.874281484444507e-07,"output_power":5.378952049279509e-07,"reference_active":true,"delta_db":-1.0652963864947682}],"active_band_count":3,"active_band_mae_db":0.39240874307667034}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010000130422823664,"log_rms_mae_db":2.797769465314886,"rms_envelope_nrmse":0.18646851801626685,"envelope_correlation":0.9802992276054425,"rms_gain_error_db":-0.8395998753931905,"reference_mean_power":0.0002857956004291323,"output_mean_power":0.00023555674875383546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.1492733373341918e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047160855442973e-05,"output_power":7.918455222214501e-05,"reference_active":true,"delta_db":-0.5787184680606707},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0483248697121172e-05,"output_power":1.9408656613553278e-05,"reference_active":true,"delta_db":-0.23403361642037993},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.217073445558823e-06,"output_power":5.896974632183065e-06,"reference_active":true,"delta_db":-1.440879092684896}],"active_band_count":3,"active_band_mae_db":0.751210392388649}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":2.2503053971607847,"rms_envelope_nrmse":0.14498228514326514,"envelope_correlation":0.9899890622529272,"rms_gain_error_db":-0.8395996460495136,"reference_mean_power":0.0002874271047267671,"output_mean_power":0.00023690147002582433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":1.945560884297879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.432816729085125e-05,"output_power":7.2819031848574e-05,"reference_active":true,"delta_db":-0.637277613238899},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0296492874768448e-05,"output_power":1.010571301818596e-05,"reference_active":true,"delta_db":-0.08122362304701784},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.97290149260777e-07,"output_power":4.745328784956294e-07,"reference_active":true,"delta_db":-0.9991904391127572}],"active_band_count":3,"active_band_mae_db":0.572563891799558}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":3.8178348649811333,"rms_envelope_nrmse":0.23306646393199723,"envelope_correlation":0.9722797015583176,"rms_gain_error_db":-0.929203281161058,"reference_mean_power":3.470766577035855e-05,"output_mean_power":2.8022383899029252e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.6534989625639412e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5332454345272337e-05,"output_power":1.371999198615657e-05,"reference_active":true,"delta_db":-0.48257822483694557},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8283604501578115e-06,"output_power":3.4210422261142237e-06,"reference_active":true,"delta_db":-0.4885438625313213},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1450631921352038e-06,"output_power":7.768417759596117e-07,"reference_active":true,"delta_db":-1.6849688213302414}],"active_band_count":3,"active_band_mae_db":0.8853636362328361}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":3.719696090683268,"rms_envelope_nrmse":0.19173814115046087,"envelope_correlation":0.9839517086423573,"rms_gain_error_db":-0.9292005947208386,"reference_mean_power":3.4905778716849276e-05,"output_mean_power":2.818235438319969e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.4649629087328264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5574047649212616e-05,"output_power":1.394094930591371e-05,"reference_active":true,"delta_db":-0.48109151140367284},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.044495719000319e-06,"output_power":1.7528350897021406e-06,"reference_active":true,"delta_db":-0.6684514680999477},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0114195457715016e-07,"output_power":6.170774195067604e-08,"reference_active":true,"delta_db":-2.1459168743406036}],"active_band_count":3,"active_band_mae_db":1.0984866179480746}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"golf","status":"measured","mos":{"status":"measured","value":2.7886619567871094,"predicted_mos":2.7886619567871094,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":126720,"input_sample_rate":24000,"input_seconds":5.28,"input_float32_sha256":"690eddb0c29bc6ed6476bb019411dc944c2ed2d9e449cd8ea445517c54871071","internal_samples":84480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.781132461999732},"started_at":"2026-10-13T05:02:41.214239+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0022902099999555503,"native_input_samples":126960,"native_input_sha256":"533a1b28f82a2cbabdefd26f15f8af0712d496c96576c39fcb09314c1c95bee7","output_sha256":"e33f66eec982106c2d95160c9a70f6de5fe4903f574b230010d9c1c1c4e7ef6d","output_samples":126720,"length_delta_samples":-240,"analysis_s":0.45097788700059027,"synthesis_s":0.1006630749998294,"total_s":0.5516409620004197,"rtf":0.10427995500953113,"output_peak":0.5644053220748901,"output_rms":0.029109069439185786,"finished_at":"2026-10-13T05:02:41.774691+00:00","analysis_stages":{"external_f0_s":0.043238517000645516},"evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"e33f66eec982106c2d95160c9a70f6de5fe4903f574b230010d9c1c1c4e7ef6d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":84640,"output_samples":84480,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.715516325313953,"output_peak":0.549424361404721},"common_prefix":{"samples":84480,"sample_rate":16000,"reference_coverage":0.998109640831758,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84480,"output_samples":84480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.873665656890518,"spectral_convergence":0.587406493289583,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.459596837047089,"spectral_convergence":0.321136686214763,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.710735380712009,"spectral_convergence":0.3567403878696638,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.303686067995267e-05,"output_power":6.882373113034301e-05,"delta_db":-0.2580388452341257,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025664781448380726,"output_power":0.00020368826063778343,"delta_db":-1.0037157071599956,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003855706636358305,"output_power":0.0003077357740869659,"delta_db":-0.9792599730606517,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002725964537875834,"output_power":0.00020686701913751802,"delta_db":-1.1982894528750723,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.603894510883052e-05,"output_power":3.4340517337777136e-05,"delta_db":-1.2731853058077252,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6418091526441452e-05,"output_power":2.6647244660475724e-05,"delta_db":0.03750868924704061,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7916663288974352,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9698401033646982,"pitch":null,"output_rms_dbfs":-30.715516325313953,"output_peak":0.549424361404721,"reference_rms_dbfs":-29.745676221949253,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.014665958216539},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84480,"output_samples":84480,"intelligibility":{"estoi":{"value":0.8664260716239639,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03256238365387961,"rms_dbfs":-29.745676221949253,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.09455525586153,"mean":-0.0032658483042861475,"zero_energy":false},"output":{"rms":0.02912220024295663,"rms_dbfs":-30.715516325313953,"peak":0.549424361404721,"peak_dbfs":-5.201841758553774,"crest_factor_db":25.51367456676018,"mean":0.00022835359418992124,"zero_energy":false},"rms_difference_db":-0.9698401033646995},"full_scale":{"reference":{"sample_count":84480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":528,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":521,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":111,"both_estimated_voiced_frames":74,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9024390243902439,"absolute_difference_cents_median":3.8743539113222525,"absolute_difference_cents_p95":28.634150794047613,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":0.00018085971712132593,"log_rms_mae_db":3.280166339020807,"rms_envelope_nrmse":0.1701661618734616,"envelope_correlation":0.9832357993731918,"rms_gain_error_db":-0.969802854834127,"reference_mean_power":0.001061304164288998,"output_mean_power":0.0008489059602412952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":5.495728406123904e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024561341766217757,"output_power":0.00019442273688699815,"reference_active":true,"delta_db":-1.0150503587660786},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.337321150044124e-05,"output_power":7.435440569535413e-05,"reference_active":true,"delta_db":-0.49719823938881164},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4477319894772342e-05,"output_power":1.2043519209423402e-05,"reference_active":true,"delta_db":-0.7993476112280439}],"active_band_count":3,"active_band_mae_db":0.7705320697943114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00014864066644098504,"log_rms_mae_db":2.4257369615979094,"rms_envelope_nrmse":0.13009871680024107,"envelope_correlation":0.994293205890113,"rms_gain_error_db":-0.9696547809123399,"reference_mean_power":0.001067313859792998,"output_mean_power":0.0008537420468506113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":4.844732525446293e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002233543529784973,"output_power":0.00017591139102290994,"reference_active":true,"delta_db":-1.037004580922223},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.367467713573285e-05,"output_power":4.043125801387983e-05,"reference_active":true,"delta_db":-0.33512448586586335},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1252953113956932e-06,"output_power":1.0682940804967358e-06,"reference_active":true,"delta_db":-0.2257568742250934}],"active_band_count":3,"active_band_mae_db":0.5326286470043933}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":0.00013930286763496605,"log_rms_mae_db":2.133332056156763,"rms_envelope_nrmse":0.1742216578467422,"envelope_correlation":0.9834894946906302,"rms_gain_error_db":-0.8997511406113854,"reference_mean_power":0.0005240515252610687,"output_mean_power":0.00042598948112095356},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":3.256745282917119e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014912538037873915,"output_power":0.00011997061474050189,"reference_active":true,"delta_db":-0.9447668022434264},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.396374213263934e-05,"output_power":4.6334013881240227e-05,"reference_active":true,"delta_db":-0.6620213388975844},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.812601322637443e-06,"output_power":7.2905230316932144e-06,"reference_active":true,"delta_db":-0.3003697651807203}],"active_band_count":3,"active_band_mae_db":0.6357193021072437}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":0.00011121771745577188,"log_rms_mae_db":1.635961701160934,"rms_envelope_nrmse":0.13024053429018392,"envelope_correlation":0.9937650632798096,"rms_gain_error_db":-0.8997511087353474,"reference_mean_power":0.0005270489396699482,"output_mean_power":0.00042842601378060587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.9121695800177634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013963291410614106,"output_power":0.00011239456291084895,"reference_active":true,"delta_db":-0.9424249900127095},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.531047880630179e-05,"output_power":2.3764304230041145e-05,"reference_active":true,"delta_db":-0.27375257353939236},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.902961487153752e-07,"output_power":7.115628085458554e-07,"reference_active":true,"delta_db":0.13177790223687846}],"active_band_count":3,"active_band_mae_db":0.44931848859632684}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":0.00010000130422823664,"log_rms_mae_db":2.018629707446355,"rms_envelope_nrmse":0.19683180490095897,"envelope_correlation":0.9830306345474502,"rms_gain_error_db":-1.2255104289482053,"reference_mean_power":0.0002863373694536123,"output_mean_power":0.00021593696245324732},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":2.1473992696978275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047538162035247e-05,"output_power":6.673941192746917e-05,"reference_active":true,"delta_db":-1.3214804805364726},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.039203124985192e-05,"output_power":1.7496909306649352e-05,"reference_active":true,"delta_db":-0.6649914728957791},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.185574973888114e-06,"output_power":4.533692754370268e-06,"reference_active":true,"delta_db":-2.5659710576524524}],"active_band_count":3,"active_band_mae_db":1.5174810036949014}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":1.6350977399950555,"rms_envelope_nrmse":0.1496879897347028,"envelope_correlation":0.9956870945275185,"rms_gain_error_db":-1.2255107645242176,"reference_mean_power":0.00028797508166872494,"output_mean_power":0.00021717200137545705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":1.943540464967835e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.263614627204728e-05,"output_power":6.0607008074200363e-05,"reference_active":true,"delta_db":-1.346472105504398},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1871883777830614e-05,"output_power":9.965015769404015e-06,"reference_active":true,"delta_db":-0.7604164604559015},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.013927956313322e-07,"output_power":3.576732380882088e-07,"reference_active":true,"delta_db":-2.256717746967825}],"active_band_count":3,"active_band_mae_db":1.4545354376427084}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1055,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":2.581156838845135,"rms_envelope_nrmse":0.23160886705745226,"envelope_correlation":0.9682020050887417,"rms_gain_error_db":-0.23900579699684515,"reference_mean_power":3.477345270477926e-05,"output_mean_power":3.2911466530681895e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1055,"frequency_resolution_hz":0.1895734597156398,"power_floor":2.6350756640718963e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.52380830052015e-05,"output_power":1.632158474277534e-05,"reference_active":true,"delta_db":0.29831989306710277},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.777851352222882e-06,"output_power":4.20341563077894e-06,"reference_active":true,"delta_db":0.46357469200191226},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1424382410950618e-06,"output_power":9.029645322923967e-07,"reference_active":true,"delta_db":-1.021620399965545}],"active_band_count":3,"active_band_mae_db":0.5945049950115201}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84480,"uncovered_tail_samples":0,"metrics":{"frame_count":1049,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":2.1999839474167446,"rms_envelope_nrmse":0.1682055202404444,"envelope_correlation":0.9838359988365187,"rms_gain_error_db":-0.2390056054264295,"reference_mean_power":3.4972320833886215e-05,"output_mean_power":3.309968748767752e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1049,"frequency_resolution_hz":0.19065776930409914,"power_floor":2.4468242834533845e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5116299815595994e-05,"output_power":1.6299820400579517e-05,"reference_active":true,"delta_db":0.32737322064876734},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3628758353146943e-06,"output_power":2.361409627678839e-06,"reference_active":true,"delta_db":-0.002695713080883074},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0120190005888306e-07,"output_power":7.65168979027879e-08,"reference_active":true,"delta_db":-1.214313115961295}],"active_band_count":3,"active_band_mae_db":0.5147940165636484}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.595348596572876,"predicted_mos":2.595348596572876,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":127080,"input_sample_rate":24000,"input_seconds":5.295,"input_float32_sha256":"25ce06b5159c2da06f2ffe9a6211130dd016398f1d51a4089d11f095c783beab","internal_samples":84720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6317617899985635},"started_at":"2026-10-13T07:39:47.336551+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003105299998424016,"native_input_samples":126960,"native_input_sha256":"533a1b28f82a2cbabdefd26f15f8af0712d496c96576c39fcb09314c1c95bee7","output_sha256":"fb7dea377e626277cb85e00b2b29f14eb19e0b6338185cde88a2ed7b6fe465a9","output_samples":127080,"length_delta_samples":120,"analysis_s":2.6257210049989226,"synthesis_s":12.326187003000086,"total_s":14.951908007999009,"rtf":2.8264476385631396,"output_peak":0.5907970070838928,"output_rms":0.03182255292473237,"finished_at":"2026-10-13T07:40:02.309103+00:00","analysis_stages":{"external_f0_s":1.7905995969995274},"evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"fb7dea377e626277cb85e00b2b29f14eb19e0b6338185cde88a2ed7b6fe465a9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":84640,"output_samples":84720,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.94112806907348,"output_peak":0.5915519609056251},"common_prefix":{"samples":84640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990557129367328,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.113275971993154,"spectral_convergence":0.643308418051714,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.709224884457698,"spectral_convergence":0.3483993384700862,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.484657688376917,"spectral_convergence":0.35781652272570896,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":8.875350241077074e-05,"delta_db":0.8535073783089148,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.00025283528221534265,"delta_db":-0.0567880423126128,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.0003575380114689242,"delta_db":-0.3195965475819279,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.0002525786912591743,"delta_db":-0.32301609015954325,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.317682550674876e-05,"delta_db":-0.2705308374899338,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":1.9723248670222628e-05,"delta_db":-1.261013788212142,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5140754473441791,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.18321278200605334,"pitch":null,"output_rms_dbfs":-29.937027614393873,"output_peak":0.5915519609056251,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.102386181609256},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84640,"output_samples":84640,"intelligibility":{"estoi":{"value":0.8533986715244378,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03253188724556978,"rms_dbfs":-29.753814832387818,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.102693866300093,"mean":-0.0032656952654256262,"zero_energy":false},"output":{"rms":0.031852873677757604,"rms_dbfs":-29.937027614393873,"peak":0.5915519609056251,"peak_dbfs":-4.560142033621619,"crest_factor_db":25.376885580772253,"mean":1.4556688195376784e-05,"zero_energy":false},"rms_difference_db":-0.18321278200605562},"full_scale":{"reference":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":529,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":522,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":88,"both_estimated_voiced_frames":66,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.8048780487804879,"absolute_difference_cents_median":4.709541731602348,"absolute_difference_cents_p95":18.110173925398342,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00018085971712132593,"log_rms_mae_db":5.93216291401961,"rms_envelope_nrmse":0.20162998068922466,"envelope_correlation":0.9657715125963225,"rms_gain_error_db":-0.18317580964375868,"reference_mean_power":0.0010593154039372696,"output_mean_power":0.001015564956734669},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":5.491358535179395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002450078451267542,"output_power":0.0002497960857607789,"reference_active":true,"delta_db":0.0840563815753223},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.385267752666162e-05,"output_power":8.33904692330848e-05,"reference_active":true,"delta_db":-0.024005171673803533},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4530628910939782e-05,"output_power":1.3018300259195679e-05,"reference_active":true,"delta_db":-0.4773012766671244}],"active_band_count":3,"active_band_mae_db":0.19512094330541674}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00014864066644098504,"log_rms_mae_db":4.410797746856995,"rms_envelope_nrmse":0.11920066291389252,"envelope_correlation":0.9876773621469291,"rms_gain_error_db":-0.18293403240206843,"reference_mean_power":0.0010653029350414017,"output_mean_power":0.0010213620576661607},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":4.840611714703517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022860940218938906,"output_power":0.00023109168960705833,"reference_active":true,"delta_db":0.046902399196649125},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8392485325428115e-05,"output_power":3.906977359441986e-05,"reference_active":true,"delta_db":0.07594667593603936},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1188041388515457e-06,"output_power":1.0881474963058498e-06,"reference_active":true,"delta_db":-0.120662971749024}],"active_band_count":3,"active_band_mae_db":0.08117068229390416}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013930286763496605,"log_rms_mae_db":2.745659118352423,"rms_envelope_nrmse":0.2135713402215947,"envelope_correlation":0.9662477919383372,"rms_gain_error_db":-0.2354798175447052,"reference_mean_power":0.0005230599489297363,"output_mean_power":0.0004954541669193747},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":3.2510596296193775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001486768204336917,"output_power":0.00013997273358403527,"reference_active":true,"delta_db":-0.2619982060924669},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.417268122003869e-05,"output_power":5.25047986007635e-05,"reference_active":true,"delta_db":-0.13581334127891773},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.81005033549118e-06,"output_power":6.912335141218683e-06,"reference_active":true,"delta_db":-0.5302904635975632}],"active_band_count":3,"active_band_mae_db":0.30936733698964924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00011121771745577188,"log_rms_mae_db":1.909553235178969,"rms_envelope_nrmse":0.1251152376289828,"envelope_correlation":0.9877594458735505,"rms_gain_error_db":-0.23548025135388706,"reference_mean_power":0.0005260459982805144,"output_mean_power":0.0004982825703604225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.9069199937213013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001421546332212628,"output_power":0.00013451017050701816,"reference_active":true,"delta_db":-0.24005895932792853},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2639933776267854e-05,"output_power":2.2373413799743607e-05,"reference_active":true,"delta_db":-0.05142897117595721},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.874281484444507e-07,"output_power":6.337401237940299e-07,"reference_active":true,"delta_db":-0.3531610707718077}],"active_band_count":3,"active_band_mae_db":0.21488300042523115}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010000130422823664,"log_rms_mae_db":2.8146578349757254,"rms_envelope_nrmse":0.22892857509837253,"envelope_correlation":0.9633931488039605,"rms_gain_error_db":-0.3007982184549928,"reference_mean_power":0.0002857956004291323,"output_mean_power":0.0002666709555321494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.1492733373341918e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047160855442973e-05,"output_power":8.694320594324042e-05,"reference_active":true,"delta_db":-0.172766616713928},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0483248697121172e-05,"output_power":1.7478827574582943e-05,"reference_active":true,"delta_db":-0.6888653986223119},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.217073445558823e-06,"output_power":6.270710483566113e-06,"reference_active":true,"delta_db":-1.1740041876759886}],"active_band_count":3,"active_band_mae_db":0.6785454010040762}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":1.9130699054201916,"rms_envelope_nrmse":0.12465398847510224,"envelope_correlation":0.9889647359089011,"rms_gain_error_db":-0.30079952575794106,"reference_mean_power":0.0002874271047267671,"output_mean_power":0.0002681932033792733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":1.945560884297879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.432816729085125e-05,"output_power":8.068347909141495e-05,"reference_active":true,"delta_db":-0.19188045052370342},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0296492874768448e-05,"output_power":9.005548112578656e-06,"reference_active":true,"delta_db":-0.5817917249076168},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.97290149260777e-07,"output_power":4.841448947783395e-07,"reference_active":true,"delta_db":-0.9120999559875851}],"active_band_count":3,"active_band_mae_db":0.5619240438063017}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":2.640489981838391,"rms_envelope_nrmse":0.2892404625908077,"envelope_correlation":0.9484022582558773,"rms_gain_error_db":-0.9245829797185902,"reference_mean_power":3.470766577035855e-05,"output_mean_power":2.8052211760258246e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.6534989625639412e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5332454345272337e-05,"output_power":1.2095231556641681e-05,"reference_active":true,"delta_db":-1.0299749307365287},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8283604501578115e-06,"output_power":3.0877934519429776e-06,"reference_active":true,"delta_db":-0.9336457909582835},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1450631921352038e-06,"output_power":1.087692361093832e-06,"reference_active":true,"delta_db":-0.22323376093410469}],"active_band_count":3,"active_band_mae_db":0.7289514942096389}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":1.744297067060866,"rms_envelope_nrmse":0.17440188225387135,"envelope_correlation":0.9851701869823996,"rms_gain_error_db":-0.9245923687453177,"reference_mean_power":3.4905778716849276e-05,"output_mean_power":2.8212274078038196e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.4649629087328264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5574047649212616e-05,"output_power":1.2294188138601508e-05,"reference_active":true,"delta_db":-1.0270164423089838},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.044495719000319e-06,"output_power":1.6806343622447465e-06,"reference_active":true,"delta_db":-0.8511296663413388},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0114195457715016e-07,"output_power":9.5007800991435e-08,"reference_active":true,"delta_db":-0.2717207593680916}],"active_band_count":3,"active_band_mae_db":0.7166222893394715}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.291257381439209,"predicted_mos":3.291257381439209,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":127200,"input_sample_rate":24000,"input_seconds":5.3,"input_float32_sha256":"ada8d8d4621230893d633fed39c19c9cc372b1ea3e38358a9d8678af3b22a7b1","internal_samples":84800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7442281040002854},"started_at":"2026-10-13T05:10:54.654696+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026341410011809785,"native_input_samples":126960,"native_input_sha256":"533a1b28f82a2cbabdefd26f15f8af0712d496c96576c39fcb09314c1c95bee7","output_sha256":"d97d9399dd423afb0bc496ed2edc98b52def649f9196af5a2f7cbe48045c42e5","output_samples":127200,"length_delta_samples":240,"analysis_s":0.7239114149997476,"synthesis_s":0.11804374399980588,"total_s":0.8419551589995535,"rtf":0.15915976540634283,"output_peak":0.5916275978088379,"output_rms":0.03147734322982436,"finished_at":"2026-10-13T05:10:55.508637+00:00","analysis_stages":{"external_f0_s":0.0839838569991116},"evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"d97d9399dd423afb0bc496ed2edc98b52def649f9196af5a2f7cbe48045c42e5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":84640,"output_samples":84800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.035640085916853,"output_peak":0.5868644395099041},"common_prefix":{"samples":84640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981132075471698,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.235320057231677,"spectral_convergence":0.24882826372250103,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.671507905045839,"spectral_convergence":0.2221954745175456,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.053016644056411,"spectral_convergence":0.2600700794272588,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":6.865945883270682e-05,"delta_db":-0.26134386425784567,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.00024008693244196387,"delta_db":-0.2814795950899284,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.0003673126612617827,"delta_db":-0.20245976415824926,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.0002510929290967006,"delta_db":-0.3486383469684343,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.175988574255407e-05,"delta_db":-0.4154449051129058,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":2.478854272675677e-05,"delta_db":-0.26828833012744013,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2962758009524673,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.27363047706737553,"pitch":null,"output_rms_dbfs":-30.027445309455196,"output_peak":0.5868644395099041,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.986614868777976},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84640,"output_samples":84640,"intelligibility":{"estoi":{"value":0.9204558756531636,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03253188724556978,"rms_dbfs":-29.753814832387818,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.102693866300093,"mean":-0.0032656952654256262,"zero_energy":false},"output":{"rms":0.03152301397554594,"rms_dbfs":-30.027445309455196,"peak":0.5868644395099041,"peak_dbfs":-4.629244106968669,"crest_factor_db":25.398201202486526,"mean":-2.8113635099632285e-05,"zero_energy":false},"rms_difference_db":-0.2736304770673783},"full_scale":{"reference":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":529,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":522,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":74,"both_estimated_voiced_frames":67,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.8170731707317073,"absolute_difference_cents_median":1.1816895325846895,"absolute_difference_cents_p95":8.051768643003795,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00018085971712132593,"log_rms_mae_db":1.5307431786127006,"rms_envelope_nrmse":0.08413448592985959,"envelope_correlation":0.9944280261019228,"rms_gain_error_db":-0.27361408134301934,"reference_mean_power":0.0010593154039372696,"output_mean_power":0.0009946353251810227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":5.491358535179395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002450078451267542,"output_power":0.00022595352343752104,"reference_active":true,"delta_db":-0.35160872702052487},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.385267752666162e-05,"output_power":6.847687048012437e-05,"reference_active":true,"delta_db":-0.8797303069948635},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4530628910939782e-05,"output_power":9.998214527644472e-06,"reference_active":true,"delta_db":-1.6236196072922457}],"active_band_count":3,"active_band_mae_db":0.9516528804358781}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00014864066644098504,"log_rms_mae_db":0.8125415982444608,"rms_envelope_nrmse":0.05346068116237912,"envelope_correlation":0.9981383759739627,"rms_gain_error_db":-0.2736111318632548,"reference_mean_power":0.0010653029350414017,"output_mean_power":0.0010002579466745134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":4.840611714703517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022860940218938906,"output_power":0.00020659094440102837,"reference_active":true,"delta_db":-0.43982807015097636},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8392485325428115e-05,"output_power":3.202143383728589e-05,"reference_active":true,"delta_db":-0.7880545233734941},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1188041388515457e-06,"output_power":7.607999449413114e-07,"reference_active":true,"delta_db":-1.6748359181139665}],"active_band_count":3,"active_band_mae_db":0.9675728372128124}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013930286763496605,"log_rms_mae_db":1.2814064840659962,"rms_envelope_nrmse":0.10039253577163192,"envelope_correlation":0.9930303449051175,"rms_gain_error_db":-0.25521297188720127,"reference_mean_power":0.0005230599489297363,"output_mean_power":0.0004932080652914758},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":3.2510596296193775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001486768204336917,"output_power":0.000135244223619775,"reference_active":true,"delta_db":-0.4112453966878098},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.417268122003869e-05,"output_power":4.171980470874288e-05,"reference_active":true,"delta_db":-1.1343806425111307},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.81005033549118e-06,"output_power":5.086136499954233e-06,"reference_active":true,"delta_db":-1.8626582145885298}],"active_band_count":3,"active_band_mae_db":1.1360947512624902}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00011121771745577188,"log_rms_mae_db":0.7948033052604314,"rms_envelope_nrmse":0.06498299058734668,"envelope_correlation":0.9972300246435123,"rms_gain_error_db":-0.25521468076297965,"reference_mean_power":0.0005260459982805144,"output_mean_power":0.0004960235007629379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.9069199937213013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001421546332212628,"output_power":0.0001265009025782656,"reference_active":true,"delta_db":-0.5066739494719564},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2639933776267854e-05,"output_power":1.806090141106503e-05,"reference_active":true,"delta_db":-0.9813573021876575},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.874281484444507e-07,"output_power":4.2189625488252093e-07,"reference_active":true,"delta_db":-2.1202164120210014}],"active_band_count":3,"active_band_mae_db":1.2027492212268718}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010000130422823664,"log_rms_mae_db":1.4428125222440513,"rms_envelope_nrmse":0.10985056590108533,"envelope_correlation":0.992535817445365,"rms_gain_error_db":-0.3680298060592496,"reference_mean_power":0.0002857956004291323,"output_mean_power":0.0002625745067448226},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.1492733373341918e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047160855442973e-05,"output_power":8.28953050223733e-05,"reference_active":true,"delta_db":-0.37982378073125517},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0483248697121172e-05,"output_power":1.67064208628608e-05,"reference_active":true,"delta_db":-0.8851542019720916},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.217073445558823e-06,"output_power":5.023963566960676e-06,"reference_active":true,"delta_db":-2.1367068748540747}],"active_band_count":3,"active_band_mae_db":1.1338949525191404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":0.9077187744310091,"rms_envelope_nrmse":0.06921170533228749,"envelope_correlation":0.9976247701729457,"rms_gain_error_db":-0.36803033107800665,"reference_mean_power":0.0002874271047267671,"output_mean_power":0.00026407341824613773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":1.945560884297879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.432816729085125e-05,"output_power":7.569021792217696e-05,"reference_active":true,"delta_db":-0.46932906251026685},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0296492874768448e-05,"output_power":8.253651153294582e-06,"reference_active":true,"delta_db":-0.9604321414699574},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.97290149260777e-07,"output_power":3.4311866402638676e-07,"reference_active":true,"delta_db":-2.4074100999751686}],"active_band_count":3,"active_band_mae_db":1.2790571013184644}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":1.507422973856662,"rms_envelope_nrmse":0.12108038762017413,"envelope_correlation":0.9907784429328969,"rms_gain_error_db":-0.08376265828434905,"reference_mean_power":3.470766577035855e-05,"output_mean_power":3.4044670934890634e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.6534989625639412e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5332454345272337e-05,"output_power":1.501909289430207e-05,"reference_active":true,"delta_db":-0.08967976735811972},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8283604501578115e-06,"output_power":3.7950060583631898e-06,"reference_active":true,"delta_db":-0.038003474506518496},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1450631921352038e-06,"output_power":7.558643370366906e-07,"reference_active":true,"delta_db":-1.8038559949613626}],"active_band_count":3,"active_band_mae_db":0.6438464122753337}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":1.0436111424088594,"rms_envelope_nrmse":0.06352688834605037,"envelope_correlation":0.997407885270343,"rms_gain_error_db":-0.08376314578717248,"reference_mean_power":3.4905778716849276e-05,"output_mean_power":3.423899563309792e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.4649629087328264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5574047649212616e-05,"output_power":1.5134920070934989e-05,"reference_active":true,"delta_db":-0.1242136735271819},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.044495719000319e-06,"output_power":1.931867951374755e-06,"reference_active":true,"delta_db":-0.2460876767493534},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0114195457715016e-07,"output_power":5.739405757609807e-08,"reference_active":true,"delta_db":-2.4606441309500284}],"active_band_count":3,"active_band_mae_db":0.943648493742188}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5105-28241-0019","method":"magphase","status":"measured","mos":{"status":"measured","value":2.658679962158203,"predicted_mos":2.658679962158203,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":255607,"input_sample_rate":48000,"input_seconds":5.325145833333333,"input_float32_sha256":"5d322e3647c17a5add4de83ab7d46cce61df39c6dc6b12a4b87edefd16bec0bc","internal_samples":85203,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.792532478000794},"started_at":"2026-10-13T05:16:18.326877+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00888380400101596,"native_input_samples":253920,"native_input_sha256":"81533eb41dd3783a42b45c0fd6603a28cec3dd89b88f38712eb554d0b9ff6f44","output_sha256":"b7d0ffc73644034134fd9a5d6f5ef15b10acc87bf0200462577d5a2c37c91d43","output_samples":255607,"length_delta_samples":1687,"analysis_s":2.611596605000159,"synthesis_s":0.9626474769993365,"total_s":3.5742440819994954,"rtf":0.6756605069942335,"output_peak":0.6297809753038758,"output_rms":0.0317955715489944,"finished_at":"2026-10-13T05:16:21.921463+00:00","evaluation_input":{"reference":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"output_sha256":"b7d0ffc73644034134fd9a5d6f5ef15b10acc87bf0200462577d5a2c37c91d43","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":84640,"output_samples":85203,"length_delta_samples":563,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.94837408578721,"output_peak":0.6079896638271078},"common_prefix":{"samples":84640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9933922514465453,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":84640,"output_samples":84640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.411757101952139,"spectral_convergence":0.8539649079461302,"spectral_convergence_status":"ok","amplitude_floor":1.2121984197282631e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.788275619433977,"spectral_convergence":0.5884198505316676,"spectral_convergence_status":"ok","amplitude_floor":7.081048851832571e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.783157306024876,"spectral_convergence":0.5274335340601355,"spectral_convergence_status":"ok","amplitude_floor":5.009180364479501e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.291800237989852e-05,"output_power":5.99910268640587e-05,"delta_db":-0.8474846665867669,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00025616304806087414,"output_power":0.00024146598711784073,"delta_db":-0.2566051753148976,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038484147909669214,"output_power":0.00036897244994143704,"delta_db":-0.18287935231035046,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002720810642199278,"output_power":0.00027637523919321817,"delta_db":0.06800813823096394,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.5951933766776914e-05,"output_power":4.931853639120373e-05,"delta_db":0.3070638745841317,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.6368160234296276e-05,"output_power":2.2544555436166624e-05,"delta_db":-0.6803815342508847,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.39040379021299915,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16585940552048759,"pitch":null,"output_rms_dbfs":-29.919674237908307,"output_peak":0.6079896638271078,"reference_rms_dbfs":-29.753814832387818,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.66106334247033},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":84640,"output_samples":84640,"intelligibility":{"estoi":{"value":0.5924731785566733,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03253188724556978,"rms_dbfs":-29.753814832387818,"peak":0.58538818359375,"peak_dbfs":-4.651120966087727,"crest_factor_db":25.102693866300093,"mean":-0.0032656952654256262,"zero_energy":false},"output":{"rms":0.03191657555164597,"rms_dbfs":-29.919674237908307,"peak":0.6079896638271078,"peak_dbfs":-4.322076078394254,"crest_factor_db":25.597598159514053,"mean":-0.000541718523795211,"zero_energy":false},"rms_difference_db":-0.1658594055204894},"full_scale":{"reference":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":84640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":529,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.9374166316119956e-06,"threshold_dbfs":-55.32034450488605,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":522,"reference_estimated_voiced_frames":82,"output_estimated_voiced_frames":87,"both_estimated_voiced_frames":57,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.6951219512195121,"absolute_difference_cents_median":11.40374449566496,"absolute_difference_cents_p95":35.852942847362364,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":84640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00018085971712132593,"log_rms_mae_db":4.557919370077934,"rms_envelope_nrmse":0.44197353370551634,"envelope_correlation":0.8276945381849347,"rms_gain_error_db":-0.1658404380272574,"reference_mean_power":0.0010593154039372696,"output_mean_power":0.0010196268041806305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":5.491358535179395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002450078451267542,"output_power":0.000246148240588692,"reference_active":true,"delta_db":0.020167452291237106},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.385267752666162e-05,"output_power":8.339129162776052e-05,"reference_active":true,"delta_db":-0.023962341874217315},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4530628910939782e-05,"output_power":1.4677845314827218e-05,"reference_active":true,"delta_db":0.04377894768663733}],"active_band_count":3,"active_band_mae_db":0.029302913950697246}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00014864066644098504,"log_rms_mae_db":3.474095846633592,"rms_envelope_nrmse":0.3127757164929127,"envelope_correlation":0.9054196064421468,"rms_gain_error_db":-0.1656805492870313,"reference_mean_power":0.0010653029350414017,"output_mean_power":0.0010254277560016406},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":4.840611714703517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022860940218938906,"output_power":0.00022987341065708916,"reference_active":true,"delta_db":0.023946515683166858},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8392485325428115e-05,"output_power":4.063255198215855e-05,"reference_active":true,"delta_db":0.2462787268921575},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1188041388515457e-06,"output_power":1.2525456297769417e-06,"reference_active":true,"delta_db":0.4903949170473488}],"active_band_count":3,"active_band_mae_db":0.25354005320755774}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00013930286763496605,"log_rms_mae_db":6.287567957722694,"rms_envelope_nrmse":0.49096739543585843,"envelope_correlation":0.8206787005261064,"rms_gain_error_db":-0.2160983457479891,"reference_mean_power":0.0005230599489297363,"output_mean_power":0.0004976701955218078},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":3.2510596296193775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001486768204336917,"output_power":0.00014332238781148064,"reference_active":true,"delta_db":-0.1592922972313297},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.417268122003869e-05,"output_power":5.188004744067812e-05,"reference_active":true,"delta_db":-0.18779966524852865},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.81005033549118e-06,"output_power":6.519623267570074e-06,"reference_active":true,"delta_db":-0.7843133188169535}],"active_band_count":3,"active_band_mae_db":0.37713509376560395}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":0.00011121771745577188,"log_rms_mae_db":5.2239027451755105,"rms_envelope_nrmse":0.3579489642299405,"envelope_correlation":0.8965881613225581,"rms_gain_error_db":-0.2160983480215806,"reference_mean_power":0.0005260459982805144,"output_mean_power":0.0005005112993574689},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.9069199937213013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001421546332212628,"output_power":0.0001370439122161579,"reference_active":true,"delta_db":-0.15901271127964386},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2639933776267854e-05,"output_power":2.2775112191583685e-05,"reference_active":true,"delta_db":0.02585372841468392},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.874281484444507e-07,"output_power":6.432752573531184e-07,"reference_active":true,"delta_db":-0.2883046420996497}],"active_band_count":3,"active_band_mae_db":0.15772369393132582}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":0.00010000130422823664,"log_rms_mae_db":6.468256310631263,"rms_envelope_nrmse":0.48199648152884605,"envelope_correlation":0.8418882296985657,"rms_gain_error_db":0.09055695925453952,"reference_mean_power":0.0002857956004291323,"output_mean_power":0.00029181743436082036},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.1492733373341918e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.047160855442973e-05,"output_power":9.140754653704198e-05,"reference_active":true,"delta_db":0.044697402403144286},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0483248697121172e-05,"output_power":2.1352102202376297e-05,"reference_active":true,"delta_db":0.18041801532094284},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.217073445558823e-06,"output_power":1.0059555523525316e-05,"reference_active":true,"delta_db":0.8786162327066706}],"active_band_count":3,"active_band_mae_db":0.3679105501435859}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":8.407177238322728e-05,"log_rms_mae_db":5.2347998964622215,"rms_envelope_nrmse":0.3365780487890433,"envelope_correlation":0.9169920877712062,"rms_gain_error_db":0.09055709493250935,"reference_mean_power":0.0002874271047267671,"output_mean_power":0.00029348332431135775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":1.945560884297879e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.432816729085125e-05,"output_power":8.597511955933875e-05,"reference_active":true,"delta_db":0.08400126554703077},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0296492874768448e-05,"output_power":1.1873170450678023e-05,"reference_active":true,"delta_db":0.6187737926189047},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.97290149260777e-07,"output_power":7.682226905174391e-07,"reference_active":true,"delta_db":1.093017782512757}],"active_band_count":3,"active_band_mae_db":0.5985976135595642}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1057,"rms_floor":5.528232384026307e-05,"log_rms_mae_db":6.148646526603919,"rms_envelope_nrmse":0.6600189646542757,"envelope_correlation":0.7156519467967862,"rms_gain_error_db":-0.25373106587758704,"reference_mean_power":3.470766577035855e-05,"output_mean_power":3.273801173371934e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1057,"frequency_resolution_hz":0.1892147587511826,"power_floor":2.6534989625639412e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5332454345272337e-05,"output_power":1.3722897378730063e-05,"reference_active":true,"delta_db":-0.4816586453055283},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8283604501578115e-06,"output_power":3.5201374800103435e-06,"reference_active":true,"delta_db":-0.36453195683402395},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1450631921352038e-06,"output_power":1.0121259313437132e-06,"reference_active":true,"delta_db":-0.535949026536984}],"active_band_count":3,"active_band_mae_db":0.46071320955884537}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":84640,"uncovered_tail_samples":0,"metrics":{"frame_count":1051,"rms_floor":4.431598350735025e-05,"log_rms_mae_db":4.982841425677564,"rms_envelope_nrmse":0.5089521477853592,"envelope_correlation":0.819536004070659,"rms_gain_error_db":-0.25373181397852324,"reference_mean_power":3.4905778716849276e-05,"output_mean_power":3.292487613319326e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1051,"frequency_resolution_hz":0.19029495718363465,"power_floor":2.4649629087328264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5574047649212616e-05,"output_power":1.3904152722256493e-05,"reference_active":true,"delta_db":-0.492569697061827},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.044495719000319e-06,"output_power":1.8506004812436044e-06,"reference_active":true,"delta_db":-0.4327353474714445},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0114195457715016e-07,"output_power":8.937043801922488e-08,"reference_active":true,"delta_db":-0.5373745566795411}],"active_band_count":3,"active_band_mae_db":0.48755986707093757}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},{"case":"librispeech_5142-33396-0019","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2309820652008057,"predicted_mos":2.2309820652008057,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":79760,"input_sample_rate":16000,"input_seconds":4.985,"input_float32_sha256":"d69b020b126967c4031b00d53fd58e15cd28f935cd145e16a3370a1d6adbfc64","internal_samples":79760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8743794450001587},"started_at":"2026-10-13T04:39:32.982680+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.7085999388655182e-05,"native_input_samples":79760,"native_input_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","output_sha256":"e1584c5ff80bb085c297a5131eb7408468cd1ec0e4f0f2e07de493fca137aa19","output_samples":79760,"length_delta_samples":0,"analysis_s":1.1083229859996209,"synthesis_s":0.3768641470005605,"total_s":1.4851871330001813,"rtf":0.2979312202608187,"output_peak":0.36460074517184154,"output_rms":0.04489529674104598,"finished_at":"2026-10-13T04:39:34.475247+00:00","evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"e1584c5ff80bb085c297a5131eb7408468cd1ec0e4f0f2e07de493fca137aa19","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.0180013609390945,"spectral_convergence":0.2116113025643865,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.312913992670058,"spectral_convergence":0.21261853108860093,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.1413099022296915,"spectral_convergence":0.21764404840857568,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":5.6856475594105594e-05,"delta_db":0.2608933064995349,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.0007441134054106948,"delta_db":-0.03839713563935373,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0006003638465128392,"delta_db":-0.05155464579018013,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.0002915106412601284,"delta_db":-0.09619795960903421,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.00021441437216180822,"delta_db":-0.29878451188542166,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":0.00010832892852699745,"delta_db":-0.30707187826564436,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17548323961486148,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0857631197165745,"pitch":null,"output_rms_dbfs":-26.95598307142527,"output_peak":0.36460074517184154,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.490741751946281},"common_prefix":{"samples":79760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.0180013609390945,"spectral_convergence":0.2116113025643865,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.312913992670058,"spectral_convergence":0.21261853108860093,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.1413099022296915,"spectral_convergence":0.21764404840857568,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":5.6856475594105594e-05,"delta_db":0.2608933064995349,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.0007441134054106948,"delta_db":-0.03839713563935373,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0006003638465128392,"delta_db":-0.05155464579018013,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.0002915106412601284,"delta_db":-0.09619795960903421,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.00021441437216180822,"delta_db":-0.29878451188542166,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":0.00010832892852699745,"delta_db":-0.30707187826564436,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17548323961486148,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0857631197165745,"pitch":null,"output_rms_dbfs":-26.95598307142527,"output_peak":0.36460074517184154,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.490741751946281},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79760,"output_samples":79760,"intelligibility":{"estoi":{"value":0.9291252895896104,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04534078160436088,"rms_dbfs":-26.870219951708695,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.39104313787336,"mean":5.436995801858701e-07,"zero_energy":false},"output":{"rms":0.04489529674104598,"rms_dbfs":-26.95598307142527,"peak":0.36460074517184154,"peak_dbfs":-8.76364896136149,"crest_factor_db":18.19233411006378,"mean":-2.23761851966125e-05,"zero_energy":false},"rms_difference_db":-0.08576311971657447},"full_scale":{"reference":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":6.886445701982074e-07,"output_rms_dbfs":-61.62004872475478,"output_zero_energy":false,"output_peak":0.007846559586860449,"output_energy_fraction":3.289800099443952e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":284,"both_estimated_voiced_frames":280,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9090909090909091,"absolute_difference_cents_median":1.0077729971022342,"absolute_difference_cents_p95":6.855420935933783,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.000134521265514355,"log_rms_mae_db":0.6538625758501537,"rms_envelope_nrmse":0.04934216433004115,"envelope_correlation":0.9972545164423787,"rms_gain_error_db":-0.08576196539393349,"reference_mean_power":0.0020578494395295825,"output_mean_power":0.0020176108296304732},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":5.569199266163003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039409678072813294,"output_power":0.00038490373102824247,"reference_active":true,"delta_db":-0.10250766336036034},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.428667833313107e-05,"output_power":5.062375456880507e-05,"reference_active":true,"delta_db":-0.303389171041376},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373716545850661e-06,"output_power":6.878229639238775e-06,"reference_active":true,"delta_db":-0.3020976755949845}],"active_band_count":3,"active_band_mae_db":0.23599816999890696}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00012498981593367327,"log_rms_mae_db":0.32787504571049253,"rms_envelope_nrmse":0.027333586592990632,"envelope_correlation":0.9991731019143797,"rms_gain_error_db":-0.08576262831888437,"reference_mean_power":0.0020703160518770356,"output_mean_power":0.0020298333634997715},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":4.869724313408462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003604621899100366,"output_power":0.0003523316214091621,"reference_active":true,"delta_db":-0.09908094413365082},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.585585870556879e-05,"output_power":2.3590408322980773e-05,"reference_active":true,"delta_db":-0.39823507721958906},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932164327326653e-07,"output_power":3.5906995070612756e-07,"reference_active":true,"delta_db":-0.39452596868452994}],"active_band_count":3,"active_band_mae_db":0.2972806633459233}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.00011089795767449635,"log_rms_mae_db":1.3105037650255833,"rms_envelope_nrmse":0.06510860761960621,"envelope_correlation":0.9961839838904033,"rms_gain_error_db":-0.06271042408788798,"reference_mean_power":0.001090370195617981,"output_mean_power":0.0010747388037698389},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":3.4652335334077796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023090422993351253,"output_power":0.00022626621919427065,"reference_active":true,"delta_db":-0.08812168778553224},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.174073740387618e-05,"output_power":3.693114655754562e-05,"reference_active":true,"delta_db":-0.5316732628203366},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679949697911219e-06,"output_power":4.904514444297968e-06,"reference_active":true,"delta_db":0.2035483279618901}],"active_band_count":3,"active_band_mae_db":0.27444775952258627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00010160775009558359,"log_rms_mae_db":0.8414833742793605,"rms_envelope_nrmse":0.037949259419843735,"envelope_correlation":0.9986488667617401,"rms_gain_error_db":-0.06271032941655415,"reference_mean_power":0.0010969783403958974,"output_mean_power":0.0010812522386953523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":3.031226473561925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021118169575363215,"output_power":0.0002075428554689144,"reference_active":true,"delta_db":-0.07548485208971717},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975380679045621e-05,"output_power":1.697434443200002e-05,"reference_active":true,"delta_db":-0.6585779133061629},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.773024908316542e-07,"output_power":2.6319259069936306e-07,"reference_active":true,"delta_db":-0.22680111569996297}],"active_band_count":3,"active_band_mae_db":0.320287960365281}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":1.1679983548583914,"rms_envelope_nrmse":0.09730174914802202,"envelope_correlation":0.9917232230914779,"rms_gain_error_db":-0.17924622982054744,"reference_mean_power":0.00039491635058927065,"output_mean_power":0.0003789487615481086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":2.1879792460690976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000156993183862756,"output_power":0.00015201404121273726,"reference_active":true,"delta_db":-0.1399709250163499},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3083527056032408e-05,"output_power":1.2271160515533631e-05,"reference_active":true,"delta_db":-0.27839199811932824},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.549781093186999e-06,"output_power":2.841846717252927e-06,"reference_active":true,"delta_db":0.4709775289433875}],"active_band_count":3,"active_band_mae_db":0.29644681735968853}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":0.7157578636414129,"rms_envelope_nrmse":0.054090811541855056,"envelope_correlation":0.9975189595935922,"rms_gain_error_db":-0.17924746908817932,"reference_mean_power":0.0003973070610776642,"output_mean_power":0.0003812427000369721},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":1.9730997112139501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014395632169995678,"output_power":0.00013970083788158452,"reference_active":true,"delta_db":-0.13031730363535288},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.34323278305194e-06,"output_power":5.748588511215558e-06,"reference_active":true,"delta_db":-0.42749426771385624},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5197840492846452e-07,"output_power":1.4301480554109962e-07,"reference_active":true,"delta_db":-0.26400882184958496}],"active_band_count":3,"active_band_mae_db":0.2739401310662647}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":1.2680022730996867,"rms_envelope_nrmse":0.11829267377190335,"envelope_correlation":0.9877212568638194,"rms_gain_error_db":-0.32792698883251803,"reference_mean_power":0.00017446384410387482,"output_mean_power":0.00016177549543844034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":1.0543054012178768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.162385199300118e-05,"output_power":6.566534019520742e-05,"reference_active":true,"delta_db":-0.37721475465524146},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.778561151416548e-06,"output_power":9.320956554293423e-06,"reference_active":true,"delta_db":-0.2081447231865424},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.445912836040226e-06,"output_power":1.787590513538247e-06,"reference_active":true,"delta_db":0.9212592808744806}],"active_band_count":3,"active_band_mae_db":0.5022062529054215}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":0.7726241161657976,"rms_envelope_nrmse":0.06709433755193636,"envelope_correlation":0.9966851368654909,"rms_gain_error_db":-0.3279357921358299,"reference_mean_power":0.0001755196839873955,"output_mean_power":0.0001627542166397347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":9.383868144458726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.640153736830106e-05,"output_power":6.0815260951442244e-05,"reference_active":true,"delta_db":-0.38165559940310617},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.230654239705997e-06,"output_power":4.322786353226817e-06,"reference_active":true,"delta_db":0.09356238726038758},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2239857829028075e-07,"output_power":1.1426223061968444e-07,"reference_active":true,"delta_db":-0.2987367525485123}],"active_band_count":3,"active_band_mae_db":0.2579849130706687}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0019","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.3518853187561035,"predicted_mos":2.3518853187561035,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":79760,"input_sample_rate":16000,"input_seconds":4.985,"input_float32_sha256":"28dd41e752c8658727abdb688dc9ce8954eee7f26652e06542653d9a76449ac6","internal_samples":79760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.772325736999846},"started_at":"2026-10-13T04:42:32.587425+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.12449992634356e-05,"native_input_samples":79760,"native_input_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","output_sha256":"e85451fcdc7418133add7d7735b7af9976abf3d07e6147829355b22f7679bef0","output_samples":79760,"length_delta_samples":0,"analysis_s":1.1935751200007871,"synthesis_s":0.08661661199948867,"total_s":1.2801917320002758,"rtf":0.2568087727182098,"output_peak":0.5619378324939631,"output_rms":0.05143065408783296,"finished_at":"2026-10-13T04:42:33.873499+00:00","evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"e85451fcdc7418133add7d7735b7af9976abf3d07e6147829355b22f7679bef0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6518956148615,"spectral_convergence":0.6060213511278312,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.421605712287289,"spectral_convergence":0.32337578923169524,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.300572859640423,"spectral_convergence":0.3880954200200033,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":6.555655751261766e-05,"delta_db":0.8792553535419982,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.00098138202920788,"delta_db":1.1635925893084214,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0008706753199086582,"delta_db":1.5628623812769604,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.00034435922159600237,"delta_db":0.6273750868446799,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.00026014577690073085,"delta_db":0.5408443625563318,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":0.0001229932747764397,"delta_db":0.24429729616950369,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8363711782829825,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0946609047570461,"pitch":null,"output_rms_dbfs":-25.775559046951653,"output_peak":0.5619378324939631,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.124691395596404},"common_prefix":{"samples":79760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6518956148615,"spectral_convergence":0.6060213511278312,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.421605712287289,"spectral_convergence":0.32337578923169524,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.300572859640423,"spectral_convergence":0.3880954200200033,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":6.555655751261766e-05,"delta_db":0.8792553535419982,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.00098138202920788,"delta_db":1.1635925893084214,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0008706753199086582,"delta_db":1.5628623812769604,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.00034435922159600237,"delta_db":0.6273750868446799,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.00026014577690073085,"delta_db":0.5408443625563318,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":0.0001229932747764397,"delta_db":0.24429729616950369,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8363711782829825,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0946609047570461,"pitch":null,"output_rms_dbfs":-25.775559046951653,"output_peak":0.5619378324939631,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.124691395596404},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79760,"output_samples":79760,"intelligibility":{"estoi":{"value":0.9050155381460655,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04534078160436088,"rms_dbfs":-26.870219951708695,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.39104313787336,"mean":5.436995801858701e-07,"zero_energy":false},"output":{"rms":0.05143065408783296,"rms_dbfs":-25.775559046951653,"peak":0.5619378324939631,"peak_dbfs":-5.006234560441845,"crest_factor_db":20.769324486509806,"mean":-2.019003067539692e-09,"zero_energy":false},"rms_difference_db":1.094660904757042},"full_scale":{"reference":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":8.412874650475633e-07,"output_rms_dbfs":-60.750555819048216,"output_zero_energy":false,"output_peak":0.010371196979329462,"output_energy_fraction":3.06250211652798e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":250,"both_estimated_voiced_frames":243,"reference_voiced_output_unvoiced_frames":65,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.788961038961039,"absolute_difference_cents_median":3.2181087877170262,"absolute_difference_cents_p95":16.208423845930696,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.000134521265514355,"log_rms_mae_db":1.468663331678249,"rms_envelope_nrmse":0.18508762969553116,"envelope_correlation":0.9843069234068661,"rms_gain_error_db":1.094662639124309,"reference_mean_power":0.0020578494395295825,"output_mean_power":0.002647767583315245},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":5.569199266163003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039409678072813294,"output_power":0.0004923235435604523,"reference_active":true,"delta_db":0.96647717559748},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.428667833313107e-05,"output_power":6.791099145015702e-05,"reference_active":true,"delta_db":0.972468018215688},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373716545850661e-06,"output_power":1.0853897165379832e-05,"reference_active":true,"delta_db":1.678992637537064}],"active_band_count":3,"active_band_mae_db":1.2059792771167441}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00012498981593367327,"log_rms_mae_db":1.0839060198803445,"rms_envelope_nrmse":0.16338027527547544,"envelope_correlation":0.9911956218725169,"rms_gain_error_db":1.0946661702870695,"reference_mean_power":0.0020703160518770356,"output_mean_power":0.0026638101316620406},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":4.869724313408462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003604621899100366,"output_power":0.00045490850919593814,"reference_active":true,"delta_db":1.0106434367311614},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.585585870556879e-05,"output_power":3.281572012376815e-05,"reference_active":true,"delta_db":1.0352297324889275},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932164327326653e-07,"output_power":4.970314597002845e-07,"reference_active":true,"delta_db":1.017522194082729}],"active_band_count":3,"active_band_mae_db":1.021131787767606}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.00011089795767449635,"log_rms_mae_db":1.8386586574408963,"rms_envelope_nrmse":0.25469108693219406,"envelope_correlation":0.982146297149138,"rms_gain_error_db":1.578389744684653,"reference_mean_power":0.001090370195617981,"output_mean_power":0.001568241514688507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":3.4652335334077796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023090422993351253,"output_power":0.0003314233603688749,"reference_active":true,"delta_db":1.5695122757134787},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.174073740387618e-05,"output_power":5.967333547502358e-05,"reference_active":true,"delta_db":1.5522019649114083},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679949697911219e-06,"output_power":8.205490661070293e-06,"reference_active":true,"delta_db":2.4386337041470636}],"active_band_count":3,"active_band_mae_db":1.8534493149239835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00010160775009558359,"log_rms_mae_db":1.535681548581164,"rms_envelope_nrmse":0.23563366980114622,"envelope_correlation":0.9881469187091525,"rms_gain_error_db":1.578389602934397,"reference_mean_power":0.0010969783403958974,"output_mean_power":0.0015777457279065026},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":3.031226473561925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021118169575363215,"output_power":0.0003082418561404527,"reference_active":true,"delta_db":1.642353383356486},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975380679045621e-05,"output_power":2.7977371180125258e-05,"reference_active":true,"delta_db":1.5115610306491742},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.773024908316542e-07,"output_power":4.646145934080571e-07,"reference_active":true,"delta_db":2.2413907615460325}],"active_band_count":3,"active_band_mae_db":1.7984350585172308}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":1.4174194956054975,"rms_envelope_nrmse":0.14506972562727338,"envelope_correlation":0.9878925868538273,"rms_gain_error_db":0.673933051052053,"reference_mean_power":0.00039491635058927065,"output_mean_power":0.00046120968674178396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":2.1879792460690976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000156993183862756,"output_power":0.000168991205907124,"reference_active":true,"delta_db":0.3198330790108638},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3083527056032408e-05,"output_power":1.6050169677237034e-05,"reference_active":true,"delta_db":0.8875479114810348},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.549781093186999e-06,"output_power":3.7240473690538823e-06,"reference_active":true,"delta_db":1.6451230000116701}],"active_band_count":3,"active_band_mae_db":0.9508346635011895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":1.099025338497994,"rms_envelope_nrmse":0.11139059937710996,"envelope_correlation":0.9947333000652155,"rms_gain_error_db":0.6739426836007468,"reference_mean_power":0.0003973070610776642,"output_mean_power":0.00046400274724779194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":1.9730997112139501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014395632169995678,"output_power":0.00015558560760963646,"reference_active":true,"delta_db":0.337386790108061},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.34323278305194e-06,"output_power":7.991244375495294e-06,"reference_active":true,"delta_db":1.0030376240420236},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5197840492846452e-07,"output_power":1.9323281768029164e-07,"reference_active":true,"delta_db":1.0429900460122636}],"active_band_count":3,"active_band_mae_db":0.7944714867207828}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":1.476111868218493,"rms_envelope_nrmse":0.15239838054887128,"envelope_correlation":0.9803931658288294,"rms_gain_error_db":0.3331031654168005,"reference_mean_power":0.00017446384410387482,"output_mean_power":0.00018837174197458366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":1.0543054012178768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.162385199300118e-05,"output_power":7.634925744408859e-05,"reference_active":true,"delta_db":0.27747143522741535},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.778561151416548e-06,"output_power":1.0141356987743108e-05,"reference_active":true,"delta_db":0.1582111464634044},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.445912836040226e-06,"output_power":1.9830488492219633e-06,"reference_active":true,"delta_db":1.3719129930932459}],"active_band_count":3,"active_band_mae_db":0.6025318582613552}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":0.92194030006865,"rms_envelope_nrmse":0.08695757747538001,"envelope_correlation":0.9940416344995963,"rms_gain_error_db":0.3330977085769228,"reference_mean_power":0.0001755196839873955,"output_mean_power":0.0001895115131179553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":9.383868144458726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.640153736830106e-05,"output_power":7.107486510463825e-05,"reference_active":true,"delta_db":0.29537909569971577},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.230654239705997e-06,"output_power":4.529873190612814e-06,"reference_active":true,"delta_db":0.2967851150133061},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2239857829028075e-07,"output_power":1.27946894374052e-07,"reference_active":true,"delta_db":0.19253375489482447}],"active_band_count":3,"active_band_mae_db":0.26156598853594876}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0019","method":"facodec","status":"measured","mos":{"status":"measured","value":2.6602530479431152,"predicted_mos":2.6602530479431152,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":79800,"input_sample_rate":16000,"input_seconds":4.9875,"input_float32_sha256":"c52147ce8e8d36a58c17212c678aa3cf780c2ee2add3252acdae1cf5f0b0ddc1","internal_samples":79800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.539803440999094},"started_at":"2026-10-13T04:54:23.908059+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.171300002373755e-05,"native_input_samples":79760,"native_input_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","output_sha256":"16ee814a7adeb1182790809d468e3a08718b7d4d1898aede2556f0696a3c4aec","output_samples":79800,"length_delta_samples":40,"analysis_s":1.3047613980006645,"synthesis_s":9.83599220799988,"total_s":11.140753606000544,"rtf":2.2348552870612926,"output_peak":0.4399140477180481,"output_rms":0.041342757957438324,"finished_at":"2026-10-13T04:54:35.057985+00:00","evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"16ee814a7adeb1182790809d468e3a08718b7d4d1898aede2556f0696a3c4aec","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":79760,"output_samples":79800,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.67201110291932,"output_peak":0.4399140477180481},"common_prefix":{"samples":79760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9994987468671679,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.049794258375758,"spectral_convergence":0.251379918898786,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.539552152964814,"spectral_convergence":0.26492061178456733,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.776281871843949,"spectral_convergence":0.2838723139683268,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":5.058553834767689e-05,"delta_db":-0.24664228337085178,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.0006206445246270169,"delta_db":-0.826359137441111,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0005221480307573941,"delta_db":-0.6577635159685327,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.0002544243211804677,"delta_db":-0.6871558429809013,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.0001728323925515358,"delta_db":-1.2350720186756303,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":8.944583096260865e-05,"delta_db":-1.1389153235284435,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7986513536609117,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7996141490596297,"pitch":null,"output_rms_dbfs":-27.669834100768327,"output_peak":0.4399140477180481,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.455209427728174},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79760,"output_samples":79760,"intelligibility":{"estoi":{"value":0.935415506773845,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04534078160436088,"rms_dbfs":-26.870219951708695,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.39104313787336,"mean":5.436995801858701e-07,"zero_energy":false},"output":{"rms":0.04135312126583313,"rms_dbfs":-27.669834100768327,"peak":0.4399140477180481,"peak_dbfs":-7.132643390650428,"crest_factor_db":20.537190710117898,"mean":-6.233740541825525e-05,"zero_energy":false},"rms_difference_db":-0.7996141490596322},"full_scale":{"reference":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":7.568810573035789e-07,"output_rms_dbfs":-61.20972363849133,"output_zero_energy":false,"output_peak":0.005025843158364296,"output_energy_fraction":4.261741670010919e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":301,"both_estimated_voiced_frames":293,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9512987012987013,"absolute_difference_cents_median":2.5716423568256213,"absolute_difference_cents_p95":14.232211662249272,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.000134521265514355,"log_rms_mae_db":1.1563914917200528,"rms_envelope_nrmse":0.1402038092075354,"envelope_correlation":0.9850269130596131,"rms_gain_error_db":-0.7996149703816181,"reference_mean_power":0.0020578494395295825,"output_mean_power":0.0017117963650682899},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":5.569199266163003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039409678072813294,"output_power":0.00032367218971294376,"reference_active":true,"delta_db":-0.8549750132004612},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.428667833313107e-05,"output_power":4.792380591208807e-05,"reference_active":true,"delta_db":-0.541419684319268},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373716545850661e-06,"output_power":7.016169316555898e-06,"reference_active":true,"delta_db":-0.21586377786484925}],"active_band_count":3,"active_band_mae_db":0.5374194917948595}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00012498981593367327,"log_rms_mae_db":0.9573468526275567,"rms_envelope_nrmse":0.1358822724027494,"envelope_correlation":0.9854495451973061,"rms_gain_error_db":-0.7996217277228024,"reference_mean_power":0.0020703160518770356,"output_mean_power":0.0017221638813325261},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":4.869724313408462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003604621899100366,"output_power":0.00029536484022272777,"reference_active":true,"delta_db":-0.8650092053328119},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.585585870556879e-05,"output_power":2.3179262897083324e-05,"reference_active":true,"delta_db":-0.47459344574778584},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932164327326653e-07,"output_power":3.8184805476740023e-07,"reference_active":true,"delta_db":-0.12741076410443872}],"active_band_count":3,"active_band_mae_db":0.48900447172834544}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.00011089795767449635,"log_rms_mae_db":1.3902412995294597,"rms_envelope_nrmse":0.1563121501336165,"envelope_correlation":0.982593095992272,"rms_gain_error_db":-0.7221713468996606,"reference_mean_power":0.001090370195617981,"output_mean_power":0.0009233297676674148},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":3.4652335334077796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023090422993351253,"output_power":0.0002136898605511085,"reference_active":true,"delta_db":-0.33647973174505996},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.174073740387618e-05,"output_power":3.7337210304385084e-05,"reference_active":true,"delta_db":-0.48418251189904676},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679949697911219e-06,"output_power":4.342904958869835e-06,"reference_active":true,"delta_db":-0.3246085985152597}],"active_band_count":3,"active_band_mae_db":0.38175694738645544}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00010160775009558359,"log_rms_mae_db":1.1329443994339832,"rms_envelope_nrmse":0.15153534785717096,"envelope_correlation":0.9829280405752473,"rms_gain_error_db":-0.7221714457957695,"reference_mean_power":0.0010969783403958974,"output_mean_power":0.0009289255494874081},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":3.031226473561925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021118169575363215,"output_power":0.00019488207298706795,"reference_active":true,"delta_db":-0.34884382242194245},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975380679045621e-05,"output_power":1.753248389624711e-05,"reference_active":true,"delta_db":-0.5180735300528633},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.773024908316542e-07,"output_power":2.7784233934681353e-07,"reference_active":true,"delta_db":0.008446563576057439}],"active_band_count":3,"active_band_mae_db":0.2917879720169544}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":1.447463352000665,"rms_envelope_nrmse":0.154805174479966,"envelope_correlation":0.9836396168644665,"rms_gain_error_db":-0.7797398933548343,"reference_mean_power":0.00039491635058927065,"output_mean_power":0.00033001305897508245},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":2.1879792460690976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000156993183862756,"output_power":0.00012035118723696581,"reference_active":true,"delta_db":-1.1543041830609406},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3083527056032408e-05,"output_power":1.245070823491143e-05,"reference_active":true,"delta_db":-0.21530780686896786},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.549781093186999e-06,"output_power":2.7544121355904086e-06,"reference_active":true,"delta_db":0.33526026674774845}],"active_band_count":3,"active_band_mae_db":0.5682907522258857}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":1.250808036532222,"rms_envelope_nrmse":0.1423719335481563,"envelope_correlation":0.986206039027947,"rms_gain_error_db":-0.779763041271291,"reference_mean_power":0.0003973070610776642,"output_mean_power":0.0003320090939086063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":1.9730997112139501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014395632169995678,"output_power":0.00011038497027127216,"reference_active":true,"delta_db":-1.153207962287637},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.34323278305194e-06,"output_power":5.714814512624824e-06,"reference_active":true,"delta_db":-0.45308510427480087},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5197840492846452e-07,"output_power":1.6858291572009741e-07,"reference_active":true,"delta_db":0.4503167877987682}],"active_band_count":3,"active_band_mae_db":0.685536618120402}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":1.9719189617629451,"rms_envelope_nrmse":0.2230176173749621,"envelope_correlation":0.9685477289027324,"rms_gain_error_db":-1.3665192756930644,"reference_mean_power":0.00017446384410387482,"output_mean_power":0.00012736600002197994},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":1.0543054012178768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.162385199300118e-05,"output_power":4.7627883567140687e-05,"reference_active":true,"delta_db":-1.7719639077464364},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.778561151416548e-06,"output_power":6.096490954520557e-06,"reference_active":true,"delta_db":-2.0519502225066173},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.445912836040226e-06,"output_power":1.3251254428496215e-06,"reference_active":true,"delta_db":-0.37885120517736465}],"active_band_count":3,"active_band_mae_db":1.4009217784768062}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":1.7407474734567598,"rms_envelope_nrmse":0.20678767415192473,"envelope_correlation":0.9738200664718234,"rms_gain_error_db":-1.3665232841271528,"reference_mean_power":0.0001755196839873955,"output_mean_power":0.0001281366896165862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":9.383868144458726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.640153736830106e-05,"output_power":4.3946405167155966e-05,"reference_active":true,"delta_db":-1.7925477912421508},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.230654239705997e-06,"output_power":2.5378700728940494e-06,"reference_active":true,"delta_db":-2.2193814855395737},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2239857829028075e-07,"output_power":9.39925049072774e-08,"reference_active":true,"delta_db":-1.1468314959421693}],"active_band_count":3,"active_band_mae_db":1.7195869242412982}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0019","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.9432661533355713,"predicted_mos":2.9432661533355713,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":220160,"input_sample_rate":44100,"input_seconds":4.992290249433107,"input_float32_sha256":"85402804599398aebfbc0dd296b23527ae12060204e46452454a0aa669b9fba2","internal_samples":79877,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4698769009992247},"started_at":"2026-10-13T07:18:31.495760+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007338154000535724,"native_input_samples":219839,"native_input_sha256":"41ea773bfabd6585b94336e0f8a743492fef8662807c3805bccae6a43d160b99","output_sha256":"b0ef4ea51ffe334ef92dce7233fbbf047738d4d155771ec37d2674d578b963e6","output_samples":220160,"length_delta_samples":321,"analysis_s":0.3992237570000725,"synthesis_s":39.741462576999766,"total_s":40.14068633399984,"rtf":8.05227583517662,"output_peak":0.4338442385196686,"output_rms":0.04160751763187047,"finished_at":"2026-10-13T07:19:11.682065+00:00","evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"b0ef4ea51ffe334ef92dce7233fbbf047738d4d155771ec37d2674d578b963e6","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":79760,"output_samples":79877,"length_delta_samples":117,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.609737611169002,"output_peak":0.43556304833985254},"common_prefix":{"samples":79760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9985352479437135,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.71134079967845,"spectral_convergence":0.44105884246431476,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":10.189644251422932,"spectral_convergence":0.4263081014767215,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":10.33001096771411,"spectral_convergence":0.4779356037184407,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":4.086212762317364e-05,"delta_db":-1.1736962871606924,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.0006444000310734863,"delta_db":-0.6632328941906728,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0005503718904755343,"delta_db":-0.42913751259228444,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.0002478443363455421,"delta_db":-0.800952097602299,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.00015199544690899677,"delta_db":-1.7930176522025154,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":0.00010097796327741989,"delta_db":-0.6122502868062167,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9120477884257802,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7331527312837048,"pitch":null,"output_rms_dbfs":-27.6033726829924,"output_peak":0.43556304833985254,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":10.076998672938496},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79760,"output_samples":79760,"intelligibility":{"estoi":{"value":0.7820030656130795,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04534078160436088,"rms_dbfs":-26.870219951708695,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.39104313787336,"mean":5.436995801858701e-07,"zero_energy":false},"output":{"rms":0.04167075468124081,"rms_dbfs":-27.6033726829924,"peak":0.43556304833985254,"peak_dbfs":-7.218979427857125,"crest_factor_db":20.384393255135276,"mean":-8.488907349006566e-05,"zero_energy":false},"rms_difference_db":-0.7331527312837061},"full_scale":{"reference":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":5.535756268150946e-07,"output_rms_dbfs":-62.568230395280544,"output_zero_energy":false,"output_peak":0.0036771991899156252,"output_energy_fraction":3.069660201058669e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":302,"both_estimated_voiced_frames":278,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9025974025974026,"absolute_difference_cents_median":6.545290275878033,"absolute_difference_cents_p95":55.810571214499696,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.000134521265514355,"log_rms_mae_db":1.3605284460180198,"rms_envelope_nrmse":0.11786470097376349,"envelope_correlation":0.9909842519283797,"rms_gain_error_db":-0.7331510857803959,"reference_mean_power":0.0020578494395295825,"output_mean_power":0.0017381949678775726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":5.569199266163003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039409678072813294,"output_power":0.00033848952432832585,"reference_active":true,"delta_db":-0.6605765468594951},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.428667833313107e-05,"output_power":4.723748739860036e-05,"reference_active":true,"delta_db":-0.6040647999942853},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373716545850661e-06,"output_power":6.514718744754799e-06,"reference_active":true,"delta_db":-0.5379076783987317}],"active_band_count":3,"active_band_mae_db":0.6008496750841706}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00012498981593367327,"log_rms_mae_db":1.0579389758750013,"rms_envelope_nrmse":0.10172225419595146,"envelope_correlation":0.9950100946956009,"rms_gain_error_db":-0.7331485998408006,"reference_mean_power":0.0020703160518770356,"output_mean_power":0.001748726089499321},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":4.869724313408462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003604621899100366,"output_power":0.00031144829715458753,"reference_active":true,"delta_db":-0.6347375612838618},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.585585870556879e-05,"output_power":2.3212309933111796e-05,"reference_active":true,"delta_db":-0.46840605138175034},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932164327326653e-07,"output_power":2.980377517624701e-07,"reference_active":true,"delta_db":-1.2036038029403628}],"active_band_count":3,"active_band_mae_db":0.7689158052019917}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.00011089795767449635,"log_rms_mae_db":3.2848337843067847,"rms_envelope_nrmse":0.17981119949656746,"envelope_correlation":0.9718262807467327,"rms_gain_error_db":-0.4625946121780858,"reference_mean_power":0.001090370195617981,"output_mean_power":0.0009801995768321916},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":3.4652335334077796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023090422993351253,"output_power":0.0001882950319032343,"reference_active":true,"delta_db":-0.8859302736468129},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.174073740387618e-05,"output_power":3.933696819615026e-05,"reference_active":true,"delta_db":-0.2575923251605333},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679949697911219e-06,"output_power":4.477392258179277e-06,"reference_active":true,"delta_db":-0.19216041124895794}],"active_band_count":3,"active_band_mae_db":0.44522767001876806}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00010160775009558359,"log_rms_mae_db":2.917543592303304,"rms_envelope_nrmse":0.15945389083837175,"envelope_correlation":0.9767356708833916,"rms_gain_error_db":-0.4625961190493744,"reference_mean_power":0.0010969783403958974,"output_mean_power":0.0009861396948388614},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":3.031226473561925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021118169575363215,"output_power":0.00017417536952709937,"reference_active":true,"delta_db":-0.8366953227054273},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975380679045621e-05,"output_power":1.887244887084916e-05,"reference_active":true,"delta_db":-0.19822544216163562},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.773024908316542e-07,"output_power":2.3277658306279062e-07,"reference_active":true,"delta_db":-0.7601448171362164}],"active_band_count":3,"active_band_mae_db":0.5983551940010932}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":3.3524289743344564,"rms_envelope_nrmse":0.2733214963018037,"envelope_correlation":0.9484110118999304,"rms_gain_error_db":-1.0677919133036207,"reference_mean_power":0.00039491635058927065,"output_mean_power":0.0003088345811948252},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":2.1879792460690976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000156993183862756,"output_power":0.00014100772183180956,"reference_active":true,"delta_db":-0.46637901114904734},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3083527056032408e-05,"output_power":1.000388300827167e-05,"reference_active":true,"delta_db":-1.1655623266467818},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.549781093186999e-06,"output_power":3.1009385386764563e-06,"reference_active":true,"delta_db":0.8499026203247695}],"active_band_count":3,"active_band_mae_db":0.8272813193735328}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":3.082575687629281,"rms_envelope_nrmse":0.2527331922476889,"envelope_correlation":0.9562568058051651,"rms_gain_error_db":-1.0677925583202397,"reference_mean_power":0.0003973070610776642,"output_mean_power":0.00031070413115786584},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":1.9730997112139501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014395632169995678,"output_power":0.00013095977655567118,"reference_active":true,"delta_db":-0.4109281583816009},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.34323278305194e-06,"output_power":4.217804264038708e-06,"reference_active":true,"delta_db":-1.7722422783260752},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5197840492846452e-07,"output_power":2.0859341660439225e-07,"reference_active":true,"delta_db":1.3751871547154981}],"active_band_count":3,"active_band_mae_db":1.186119197141058}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":3.652962410499147,"rms_envelope_nrmse":0.2868097111317075,"envelope_correlation":0.9355056260309812,"rms_gain_error_db":-1.1409408350324153,"reference_mean_power":0.00017446384410387482,"output_mean_power":0.0001341563870760314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":1.0543054012178768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.162385199300118e-05,"output_power":6.398008392135651e-05,"reference_active":true,"delta_db":-0.49012868635040663},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.778561151416548e-06,"output_power":9.289428626810228e-06,"reference_active":true,"delta_db":-0.22285953741267306},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.445912836040226e-06,"output_power":1.199928452291186e-06,"reference_active":true,"delta_db":-0.8098676187808811}],"active_band_count":3,"active_band_mae_db":0.5076186141813203}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":3.244778222910501,"rms_envelope_nrmse":0.2648553367224226,"envelope_correlation":0.9434962397502408,"rms_gain_error_db":-1.1409221616136813,"reference_mean_power":0.0001755196839873955,"output_mean_power":0.0001349688700909364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":9.383868144458726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.640153736830106e-05,"output_power":5.952682885888929e-05,"reference_active":true,"delta_db":-0.4746538730236711},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.230654239705997e-06,"output_power":4.2607515385928164e-06,"reference_active":true,"delta_db":0.03078676465619446},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2239857829028075e-07,"output_power":9.028360639490889e-08,"reference_active":true,"delta_db":-1.321674746067252}],"active_band_count":3,"active_band_mae_db":0.6090384612490393}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0019","method":"golf","status":"measured","mos":{"status":"measured","value":3.0852363109588623,"predicted_mos":3.0852363109588623,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":119520,"input_sample_rate":24000,"input_seconds":4.98,"input_float32_sha256":"33a042e88efa1d2e824ec30ab19505a74ecbe4d64cde39773aee7f822de63420","internal_samples":79680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8037974779999786},"started_at":"2026-10-13T05:02:41.775380+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002547563999542035,"native_input_samples":119640,"native_input_sha256":"7769d811568403ceb4d22e904a37ed84370cd841a966195e9795bbdf9b241e9e","output_sha256":"b855fb6b75f97a9c200c688b3e932b89bc28fa1e802d82f4bdffa6c7abcf3ff7","output_samples":119520,"length_delta_samples":-120,"analysis_s":0.45471608899970306,"synthesis_s":0.09724807499969756,"total_s":0.5519641639994006,"rtf":0.11072500782335017,"output_peak":0.41796666383743286,"output_rms":0.04178950344267023,"finished_at":"2026-10-13T05:02:42.336775+00:00","analysis_stages":{"external_f0_s":0.0401323809992391},"evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"b855fb6b75f97a9c200c688b3e932b89bc28fa1e802d82f4bdffa6c7abcf3ff7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":79760,"output_samples":79680,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.572264267638765,"output_peak":0.42324839254286883},"common_prefix":{"samples":79680,"sample_rate":16000,"reference_coverage":0.9989969909729187,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79680,"output_samples":79680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.349122801733918,"spectral_convergence":0.45599875642771215,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.556193475951843,"spectral_convergence":0.25461582222473184,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.867075981397557,"spectral_convergence":0.28232715294886135,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.359570980811941e-05,"output_power":5.115815271714344e-05,"delta_db":-0.20215172944331583,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007517362390042488,"output_power":0.0006737412808272953,"delta_db":-0.4757232897541759,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006078813080203947,"output_power":0.000531445132666681,"delta_db":-0.5836035545421661,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00029833849445060176,"output_power":0.00027197060943874303,"delta_db":-0.40187319308769276,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00022991533838867897,"output_power":0.00017047759651323831,"delta_db":-1.2990063158453165,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011638211849961915,"output_power":5.0141842116849865e-05,"delta_db":-3.656859732998914,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1032029692785967,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.706399748101789,"pitch":null,"output_rms_dbfs":-27.572264267638765,"output_peak":0.42324839254286883,"reference_rms_dbfs":-26.865864519536977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9241307530277725},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79680,"output_samples":79680,"intelligibility":{"estoi":{"value":0.9287915117070494,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0453635228809631,"rms_dbfs":-26.865864519536977,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.38668770570164,"mean":1.0318066700395332e-06,"zero_energy":false},"output":{"rms":0.041820265593130214,"rms_dbfs":-27.572264267638765,"peak":0.42324839254286883,"peak_dbfs":-7.468093652500455,"crest_factor_db":20.104170615138308,"mean":0.0002857314846506176,"zero_energy":false},"rms_difference_db":-0.7063997481017878},"full_scale":{"reference":{"sample_count":79680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":8.738665751809839e-07,"output_rms_dbfs":-60.58554871811924,"output_zero_energy":false,"output_peak":0.005457358301600904,"output_energy_fraction":4.815966414306426e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":308,"both_estimated_voiced_frames":296,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.961038961038961,"absolute_difference_cents_median":3.829168350831501,"absolute_difference_cents_p95":13.419969767218177,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":995,"rms_floor":0.000134521265514355,"log_rms_mae_db":2.5689359816558515,"rms_envelope_nrmse":0.1411216730841047,"envelope_correlation":0.9868935954516759,"rms_gain_error_db":-0.7063985336487089,"reference_mean_power":0.002059916609684099,"output_mean_power":0.001750692156811999},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":995,"frequency_resolution_hz":0.20100502512562812,"power_floor":5.564103424351456e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039390337357946307,"output_power":0.00038256127320897054,"reference_active":true,"delta_db":-0.12688696176280476},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.423719072426702e-05,"output_power":4.0034211058008964e-05,"reference_active":true,"delta_db":-1.3186591225633202},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.317742789946324e-06,"output_power":5.823878589978142e-06,"reference_active":true,"delta_db":-0.9916482778827129}],"active_band_count":3,"active_band_mae_db":0.8123981207362793}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":989,"rms_floor":0.00012498981593367327,"log_rms_mae_db":2.3257824499542985,"rms_envelope_nrmse":0.1258853366969112,"envelope_correlation":0.9913857230171061,"rms_gain_error_db":-0.7063958318385197,"reference_mean_power":0.0020724084231720226,"output_mean_power":0.0017613098570241307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":989,"frequency_resolution_hz":0.20222446916076844,"power_floor":4.864298579126761e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003601302826140142,"output_power":0.000350983386966,"reference_active":true,"delta_db":-0.11173081266492262},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5780211543354384e-05,"output_power":1.906581999599227e-05,"reference_active":true,"delta_db":-1.3103098825061386},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.907501038185919e-07,"output_power":3.1758597775201215e-07,"reference_active":true,"delta_db":-0.9003778313317816}],"active_band_count":3,"active_band_mae_db":0.774139508834281}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":995,"rms_floor":0.00011089795767449635,"log_rms_mae_db":1.8422813381897627,"rms_envelope_nrmse":0.13832475812026732,"envelope_correlation":0.9853113894940165,"rms_gain_error_db":-0.43356430783963684,"reference_mean_power":0.0010914659958035562,"output_mean_power":0.000987765331348358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":995,"frequency_resolution_hz":0.20100502512562812,"power_floor":3.46688531807274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023109251281696602,"output_power":0.00022042801880645243,"reference_active":true,"delta_db":-0.20519077712610498},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1736672137492624e-05,"output_power":3.0313167385862498e-05,"reference_active":true,"delta_db":-1.3888650011539174},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6321392318888245e-06,"output_power":3.4833784095580553e-06,"reference_active":true,"delta_db":-1.2378094915237328}],"active_band_count":3,"active_band_mae_db":0.943955089934585}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":989,"rms_floor":0.00010160775009558359,"log_rms_mae_db":1.7132240050329195,"rms_envelope_nrmse":0.12091023838066699,"envelope_correlation":0.9889584047043981,"rms_gain_error_db":-0.43356423822137063,"reference_mean_power":0.0010980874809545051,"output_mean_power":0.0009937577222121053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":989,"frequency_resolution_hz":0.20222446916076844,"power_floor":3.0323039503813747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021127934891118026,"output_power":0.000201539709890021,"reference_active":true,"delta_db":-0.20496420785007266},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.96952055201328e-05,"output_power":1.43087668713521e-05,"reference_active":true,"delta_db":-1.387583091456247},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.759643605703082e-07,"output_power":2.391757420494777e-07,"reference_active":true,"delta_db":-0.6213586865867473}],"active_band_count":3,"active_band_mae_db":0.7379686619643557}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":995,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":2.446182270895226,"rms_envelope_nrmse":0.17949714432766772,"envelope_correlation":0.977600075030947,"rms_gain_error_db":-0.6039164772360963,"reference_mean_power":0.0003953128209860394,"output_mean_power":0.0003439927204085285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":995,"frequency_resolution_hz":0.20100502512562812,"power_floor":2.1836701401935622e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015674507361406674,"output_power":0.00015406815817767714,"reference_active":true,"delta_db":-0.0748100936694627},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3054495639838612e-05,"output_power":1.1427155103856102e-05,"reference_active":true,"delta_db":-0.5782197530834101},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.53078549620576e-06,"output_power":2.1733248770053807e-06,"reference_active":true,"delta_db":-0.66130685684621}],"active_band_count":3,"active_band_mae_db":0.43811223453302767}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":989,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":2.21040718497048,"rms_envelope_nrmse":0.15713780118251888,"envelope_correlation":0.9836079206786666,"rms_gain_error_db":-0.6038969735575851,"reference_mean_power":0.0003977084911717886,"output_mean_power":0.00034607893531252745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":989,"frequency_resolution_hz":0.20222446916076844,"power_floor":1.968764613506404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014367401959746533,"output_power":0.0001417260032558761,"reference_active":true,"delta_db":-0.05928702302310983},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.3233215603502205e-06,"output_power":5.148386064860753e-06,"reference_active":true,"delta_db":-0.8927416145205604},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5169258641164037e-07,"output_power":1.2792154293764367e-07,"reference_active":true,"delta_db":-0.7402066721673337}],"active_band_count":3,"active_band_mae_db":0.5640784365703347}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":995,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":4.3160384445554945,"rms_envelope_nrmse":0.32631903118726213,"envelope_correlation":0.9598012681074597,"rms_gain_error_db":-2.6786764519201887,"reference_mean_power":0.00017463881159096567,"output_mean_power":9.424821249674925e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":995,"frequency_resolution_hz":0.20100502512562812,"power_floor":1.053633581539003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.151837642919676e-05,"output_power":4.435710618569392e-05,"reference_active":true,"delta_db":-2.074544413323159},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.770010244017846e-06,"output_power":4.638413147743969e-06,"reference_active":true,"delta_db":-3.2352559004844723},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4348284796512568e-06,"output_power":6.003055408537318e-07,"reference_active":true,"delta_db":-3.7842763638544428}],"active_band_count":3,"active_band_mae_db":3.0313588925540245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79680,"uncovered_tail_samples":0,"metrics":{"frame_count":989,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":4.23308855889435,"rms_envelope_nrmse":0.314428905907507,"envelope_correlation":0.9667612779369961,"rms_gain_error_db":-2.678652385075794,"reference_mean_power":0.000175696676166044,"output_mean_power":9.481964108375451e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":989,"frequency_resolution_hz":0.20222446916076844,"power_floor":9.37659216498902e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.628019195657781e-05,"output_power":4.1109996836715165e-05,"reference_active":true,"delta_db":-2.074363141630836},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.227280241484103e-06,"output_power":1.9001760409903464e-06,"reference_active":true,"delta_db":-3.4726720187171676},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2137706066954672e-07,"output_power":5.708966037736473e-08,"reference_active":true,"delta_db":-3.275791567541682}],"active_band_count":3,"active_band_mae_db":2.940942242629895}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0019","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.4355125427246094,"predicted_mos":2.4355125427246094,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":119760,"input_sample_rate":24000,"input_seconds":4.99,"input_float32_sha256":"3be5627600dad738368b1cb062d3a7be67779060e3a2330be98bf33770ca0f9c","internal_samples":79840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.507978457997524},"started_at":"2026-10-13T07:40:02.311263+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004036002999782795,"native_input_samples":119640,"native_input_sha256":"7769d811568403ceb4d22e904a37ed84370cd841a966195e9795bbdf9b241e9e","output_sha256":"bad3e714ebb64e22ec29838d5bdbc1a438fc7fba526b5ef45ed9e8e8374247fa","output_samples":119760,"length_delta_samples":120,"analysis_s":1.757124842999474,"synthesis_s":12.273607736002305,"total_s":14.030732579001779,"rtf":2.8145902866603367,"output_peak":0.42323556542396545,"output_rms":0.04404043030025005,"finished_at":"2026-10-13T07:40:16.365603+00:00","analysis_stages":{"external_f0_s":1.0902411169990955},"evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"bad3e714ebb64e22ec29838d5bdbc1a438fc7fba526b5ef45ed9e8e8374247fa","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":79760,"output_samples":79840,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.11671068315006,"output_peak":0.4234295829792064},"common_prefix":{"samples":79760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998997995991984,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.609388692220335,"spectral_convergence":0.4579929478370255,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.357393740544508,"spectral_convergence":0.3002921429243926,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.762003178012982,"spectral_convergence":0.34518251915744,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":5.5096579152358276e-05,"delta_db":0.12434030530307832,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.0007417333109230112,"delta_db":-0.052310586806827916,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0005846449382504277,"delta_db":-0.16677802039296324,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.000282281169523531,"delta_db":-0.23592300666949534,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.000192722685129125,"delta_db":-0.7619950599002522,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":8.782399859061393e-05,"delta_db":-1.2183842895332213,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.426621878100973,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.24214129679900942,"pitch":null,"output_rms_dbfs":-27.112361248507707,"output_peak":0.4234295829792064,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.576261870259276},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79760,"output_samples":79760,"intelligibility":{"estoi":{"value":0.9096733678995,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04534078160436088,"rms_dbfs":-26.870219951708695,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.39104313787336,"mean":5.436995801858701e-07,"zero_energy":false},"output":{"rms":0.044094247715195996,"rms_dbfs":-27.112361248507707,"peak":0.4234295829792064,"peak_dbfs":-7.464376063988588,"crest_factor_db":19.647985184519115,"mean":-5.693704207680937e-05,"zero_energy":false},"rms_difference_db":-0.2421412967990122},"full_scale":{"reference":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":8.584947871865224e-07,"output_rms_dbfs":-60.66262337547912,"output_zero_energy":false,"output_peak":0.016592254644515976,"output_energy_fraction":4.2515751700407136e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":273,"both_estimated_voiced_frames":257,"reference_voiced_output_unvoiced_frames":51,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.8344155844155844,"absolute_difference_cents_median":5.737229602540135,"absolute_difference_cents_p95":29.944600036131423,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.000134521265514355,"log_rms_mae_db":1.65608202434248,"rms_envelope_nrmse":0.13080389024563063,"envelope_correlation":0.9804676775375482,"rms_gain_error_db":-0.24214064770787236,"reference_mean_power":0.0020578494395295825,"output_mean_power":0.001946254062514543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":5.569199266163003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039409678072813294,"output_power":0.00035730813867935086,"reference_active":true,"delta_db":-0.4255997867291495},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.428667833313107e-05,"output_power":5.0455511039154554e-05,"reference_active":true,"delta_db":-0.31784659781045865},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373716545850661e-06,"output_power":7.0705108402884e-06,"reference_active":true,"delta_db":-0.18235646357600502}],"active_band_count":3,"active_band_mae_db":0.30860094937187105}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00012498981593367327,"log_rms_mae_db":1.1845575807044604,"rms_envelope_nrmse":0.09448162081162191,"envelope_correlation":0.9893925061544846,"rms_gain_error_db":-0.24213850186270927,"reference_mean_power":0.0020703160518770356,"output_mean_power":0.001958045588835708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":4.869724313408462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003604621899100366,"output_power":0.00032647652935426043,"reference_active":true,"delta_db":-0.4300775186432812},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.585585870556879e-05,"output_power":2.454694716413145e-05,"reference_active":true,"delta_db":-0.2256147799840347},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932164327326653e-07,"output_power":3.113036648294078e-07,"reference_active":true,"delta_db":-1.0144742549227}],"active_band_count":3,"active_band_mae_db":0.556722184516672}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.00011089795767449635,"log_rms_mae_db":1.9688793619687768,"rms_envelope_nrmse":0.14916823167603477,"envelope_correlation":0.9796491213245877,"rms_gain_error_db":-0.09207235284346192,"reference_mean_power":0.001090370195617981,"output_mean_power":0.0010674971804730115},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":3.4652335334077796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023090422993351253,"output_power":0.00021944356194663654,"reference_active":true,"delta_db":-0.22109044828164057},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.174073740387618e-05,"output_power":4.201259433404192e-05,"reference_active":true,"delta_db":0.028193834642594298},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679949697911219e-06,"output_power":5.217155097636186e-06,"reference_active":true,"delta_db":0.4719256267374728}],"active_band_count":3,"active_band_mae_db":0.24040330322056924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00010160775009558359,"log_rms_mae_db":1.57768397630108,"rms_envelope_nrmse":0.10684125748424592,"envelope_correlation":0.9887910072343584,"rms_gain_error_db":-0.09207229865527979,"reference_mean_power":0.0010969783403958974,"output_mean_power":0.0010739667176614866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":3.031226473561925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021118169575363215,"output_power":0.00020057595081333178,"reference_active":true,"delta_db":-0.22377413253235695},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975380679045621e-05,"output_power":2.040864211426459e-05,"reference_active":true,"delta_db":0.1416330829081055},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.773024908316542e-07,"output_power":2.739752906424604e-07,"reference_active":true,"delta_db":-0.052423742676374604}],"active_band_count":3,"active_band_mae_db":0.1392769860389457}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":1.829507663942362,"rms_envelope_nrmse":0.15997639107232153,"envelope_correlation":0.9772693027926151,"rms_gain_error_db":-0.30782739311826096,"reference_mean_power":0.00039491635058927065,"output_mean_power":0.00036789372112003206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":2.1879792460690976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000156993183862756,"output_power":0.00014187437787005513,"reference_active":true,"delta_db":-0.43976827027099796},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3083527056032408e-05,"output_power":1.2405765795845318e-05,"reference_active":true,"delta_db":-0.23101258827514087},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.549781093186999e-06,"output_power":2.503262886454648e-06,"reference_active":true,"delta_db":-0.07996435940737935}],"active_band_count":3,"active_band_mae_db":0.25024840598450604}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":1.3110483177729977,"rms_envelope_nrmse":0.11146273596272707,"envelope_correlation":0.988659286310804,"rms_gain_error_db":-0.30782518084948973,"reference_mean_power":0.0003973070610776642,"output_mean_power":0.00037012103288576936},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":1.9730997112139501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014395632169995678,"output_power":0.00013009944340093056,"reference_active":true,"delta_db":-0.4395530270548257},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.34323278305194e-06,"output_power":5.99309246399353e-06,"reference_active":true,"delta_db":-0.24659671137692965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5197840492846452e-07,"output_power":1.4394101499582048e-07,"reference_active":true,"delta_db":-0.2359732133428235}],"active_band_count":3,"active_band_mae_db":0.3073743172581929}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":2.0597978935504986,"rms_envelope_nrmse":0.21745873216275133,"envelope_correlation":0.9667705839179593,"rms_gain_error_db":-1.1954999960320931,"reference_mean_power":0.00017446384410387482,"output_mean_power":0.00013248156134839682},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":1.0543054012178768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.162385199300118e-05,"output_power":5.068862491571812e-05,"reference_active":true,"delta_db":-1.5014716422409344},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.778561151416548e-06,"output_power":6.432168586948259e-06,"reference_active":true,"delta_db":-1.819175372824209},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.445912836040226e-06,"output_power":1.0879981370968067e-06,"reference_active":true,"delta_db":-1.2351396142376723}],"active_band_count":3,"active_band_mae_db":1.5185955431009386}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":1.4828928712050826,"rms_envelope_nrmse":0.1748679713736861,"envelope_correlation":0.983478532959226,"rms_gain_error_db":-1.1954980564775042,"reference_mean_power":0.0001755196839873955,"output_mean_power":0.00013328338766285938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":9.383868144458726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.640153736830106e-05,"output_power":4.658573137776809e-05,"reference_active":true,"delta_db":-1.5392521639924492},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.230654239705997e-06,"output_power":2.770195523633314e-06,"reference_active":true,"delta_db":-1.8389710987303172},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2239857829028075e-07,"output_power":8.51665035169474e-08,"reference_active":true,"delta_db":-1.575075555371399}],"active_band_count":3,"active_band_mae_db":1.6510996060313883}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0019","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.475445508956909,"predicted_mos":2.475445508956909,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":119880,"input_sample_rate":24000,"input_seconds":4.995,"input_float32_sha256":"77b1b0f3093439d24075b5341f0e150e3e2fca5fa2d8806db257589034966f64","internal_samples":79920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.928210656998999},"started_at":"2026-10-13T05:10:55.510979+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002899281000281917,"native_input_samples":119640,"native_input_sha256":"7769d811568403ceb4d22e904a37ed84370cd841a966195e9795bbdf9b241e9e","output_sha256":"2bc5ade4adb74de7d343e236f584241661f896eaedee4c57ca3ac66bad1bd457","output_samples":119880,"length_delta_samples":240,"analysis_s":0.6621211809997476,"synthesis_s":0.140940077999403,"total_s":0.8030612589991506,"rtf":0.16109553841507535,"output_peak":0.4121088981628418,"output_rms":0.04474589383518519,"finished_at":"2026-10-13T05:10:56.325515+00:00","analysis_stages":{"external_f0_s":0.07855826500053809},"evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"2bc5ade4adb74de7d343e236f584241661f896eaedee4c57ca3ac66bad1bd457","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":79760,"output_samples":79920,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.978668589859403,"output_peak":0.4019120414280099},"common_prefix":{"samples":79760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.997997997997998,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":3.8812691079736403,"spectral_convergence":0.12608586858421827,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.029891129474085,"spectral_convergence":0.11915772130597092,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.651222731001463,"spectral_convergence":0.13154132147763942,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":5.214337441151468e-05,"delta_db":-0.11491471876165477,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.000744545796811151,"delta_db":-0.03587425862034098,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0005986417775961771,"delta_db":-0.06402974227302369,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.0002886495522525636,"delta_db":-0.13903320598815871,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.00021809816356198827,"delta_db":-0.22480335144053515,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":0.00010702995140384211,"delta_db":-0.35946307182946546,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15635305815219644,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09974588702826859,"pitch":null,"output_rms_dbfs":-26.969965838736965,"output_peak":0.4019120414280099,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.187460989483063},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79760,"output_samples":79760,"intelligibility":{"estoi":{"value":0.9756351611622256,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04534078160436088,"rms_dbfs":-26.870219951708695,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.39104313787336,"mean":5.436995801858701e-07,"zero_energy":false},"output":{"rms":0.04482308128673482,"rms_dbfs":-26.969965838736965,"peak":0.4019120414280099,"peak_dbfs":-7.917379639911534,"crest_factor_db":19.05258619882543,"mean":-3.0934225361898886e-06,"zero_energy":false},"rms_difference_db":-0.09974588702826992},"full_scale":{"reference":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":6.895621083109713e-07,"output_rms_dbfs":-61.61426611167657,"output_zero_energy":false,"output_peak":0.011447665136150153,"output_energy_fraction":3.3048065858377135e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":291,"both_estimated_voiced_frames":287,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9318181818181818,"absolute_difference_cents_median":0.8309353246608242,"absolute_difference_cents_p95":5.695556393582226,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.000134521265514355,"log_rms_mae_db":0.5425464581063764,"rms_envelope_nrmse":0.032691914409299874,"envelope_correlation":0.9988951928580688,"rms_gain_error_db":-0.09974527872789836,"reference_mean_power":0.0020578494395295825,"output_mean_power":0.002011125019954852},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":5.569199266163003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039409678072813294,"output_power":0.00038308733020828124,"reference_active":true,"delta_db":-0.12305098381717637},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.428667833313107e-05,"output_power":5.093639316389166e-05,"reference_active":true,"delta_db":-0.2766508000173055},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373716545850661e-06,"output_power":5.435201936880813e-06,"reference_active":true,"delta_db":-1.3247075452078065}],"active_band_count":3,"active_band_mae_db":0.5748031096807628}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00012498981593367327,"log_rms_mae_db":0.28369362079392024,"rms_envelope_nrmse":0.01904071217783214,"envelope_correlation":0.9997048235078615,"rms_gain_error_db":-0.09974399342051238,"reference_mean_power":0.0020703160518770356,"output_mean_power":0.002023309170929849},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":4.869724313408462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003604621899100366,"output_power":0.0003505674780376522,"reference_active":true,"delta_db":-0.1208809248380087},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.585585870556879e-05,"output_power":2.4930091831409153e-05,"reference_active":true,"delta_db":-0.15835087558153438},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932164327326653e-07,"output_power":3.029949737841571e-07,"reference_active":true,"delta_db":-1.131962346550738}],"active_band_count":3,"active_band_mae_db":0.4703980489900937}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.00011089795767449635,"log_rms_mae_db":0.6547271830037098,"rms_envelope_nrmse":0.034960942014595556,"envelope_correlation":0.9989160613380972,"rms_gain_error_db":-0.04993838817550982,"reference_mean_power":0.001090370195617981,"output_mean_power":0.001077904122936521},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":3.4652335334077796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023090422993351253,"output_power":0.00022555031466376906,"reference_active":true,"delta_db":-0.10188451513095954},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.174073740387618e-05,"output_power":3.926802984008517e-05,"reference_active":true,"delta_db":-0.2652100482200614},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679949697911219e-06,"output_power":3.2084776502881482e-06,"reference_active":true,"delta_db":-1.6394216670155546}],"active_band_count":3,"active_band_mae_db":0.6688387434555252}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00010160775009558359,"log_rms_mae_db":0.4210267107365089,"rms_envelope_nrmse":0.018462851208885293,"envelope_correlation":0.999697265027065,"rms_gain_error_db":-0.04993824988325422,"reference_mean_power":0.0010969783403958974,"output_mean_power":0.0010844367521133122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":3.031226473561925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021118169575363215,"output_power":0.00020596605312709634,"reference_active":true,"delta_db":-0.10860626065109723},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975380679045621e-05,"output_power":1.8972267925089287e-05,"reference_active":true,"delta_db":-0.1753155257213493},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.773024908316542e-07,"output_power":2.0086019220463717e-07,"reference_active":true,"delta_db":-1.4005989664972107}],"active_band_count":3,"active_band_mae_db":0.5615069176232191}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":0.7303446344870189,"rms_envelope_nrmse":0.05757448798856737,"envelope_correlation":0.9973374537380915,"rms_gain_error_db":-0.17698877383029252,"reference_mean_power":0.00039491635058927065,"output_mean_power":0.0003791457897304528},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":2.1879792460690976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000156993183862756,"output_power":0.00015212065169121456,"reference_active":true,"delta_db":-0.13692619874917725},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3083527056032408e-05,"output_power":1.2911081652095618e-05,"reference_active":true,"delta_db":-0.057622091456740195},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.549781093186999e-06,"output_power":1.9524906349712385e-06,"reference_active":true,"delta_db":-1.1591393699100305}],"active_band_count":3,"active_band_mae_db":0.45122922003864935}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":0.4716354426231923,"rms_envelope_nrmse":0.03536449768032178,"envelope_correlation":0.9991682339531539,"rms_gain_error_db":-0.17698761843422287,"reference_mean_power":0.0003973070610776642,"output_mean_power":0.0003814411312383572},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":1.9730997112139501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014395632169995678,"output_power":0.00013970180382587722,"reference_active":true,"delta_db":-0.13028727498039216},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.34323278305194e-06,"output_power":6.325638892907455e-06,"reference_active":true,"delta_db":-0.012062533290364457},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5197840492846452e-07,"output_power":1.2714440181215733e-07,"reference_active":true,"delta_db":-0.7748463923236668}],"active_band_count":3,"active_band_mae_db":0.3057320668648078}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":0.9405947662996357,"rms_envelope_nrmse":0.08709914536011867,"envelope_correlation":0.9938032193360526,"rms_gain_error_db":-0.30310440379869774,"reference_mean_power":0.00017446384410387482,"output_mean_power":0.00016270278880862636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":1.0543054012178768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.162385199300118e-05,"output_power":6.808920153995511e-05,"reference_active":true,"delta_db":-0.21979432661342183},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.778561151416548e-06,"output_power":7.90585828388847e-06,"reference_active":true,"delta_db":-0.9232593086722619},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.445912836040226e-06,"output_power":1.226002119457514e-06,"reference_active":true,"delta_db":-0.7165089220258687}],"active_band_count":3,"active_band_mae_db":0.6198541857705174}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":0.6554224380688635,"rms_envelope_nrmse":0.05553717731074637,"envelope_correlation":0.9979873792073208,"rms_gain_error_db":-0.30310314111199466,"reference_mean_power":0.0001755196839873955,"output_mean_power":0.00016368749940818956},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":9.383868144458726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.640153736830106e-05,"output_power":6.285744055020414e-05,"reference_active":true,"delta_db":-0.23821441234272295},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.230654239705997e-06,"output_power":3.43457324470764e-06,"reference_active":true,"delta_db":-0.9053475058200832},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2239857829028075e-07,"output_power":8.877231406476089e-08,"reference_active":true,"delta_db":-1.3949883236780627}],"active_band_count":3,"active_band_mae_db":0.8461834139469563}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0019","method":"magphase","status":"measured","mos":{"status":"measured","value":3.018548011779785,"predicted_mos":3.018548011779785,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":239523,"input_sample_rate":48000,"input_seconds":4.9900625,"input_float32_sha256":"e04794df2c1af6f5fbd21ff6985c2902de683e512dfbe56824810c855ed46ab2","internal_samples":79841,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9333224410002003},"started_at":"2026-10-13T05:16:21.922081+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.007736340001429198,"native_input_samples":239280,"native_input_sha256":"000438997226232457e704226bf453d7a5d30d806fb0f2ea86257d948a0aa25f","output_sha256":"c880b9724e2f3a0a45ebff876255c6c546ef70f295222c76fc040f113fdaf5e7","output_samples":239523,"length_delta_samples":243,"analysis_s":2.9407593489995634,"synthesis_s":1.0046714820000489,"total_s":3.9454308309996122,"rtf":0.7914605478434528,"output_peak":0.41632035676802515,"output_rms":0.04506686615630579,"finished_at":"2026-10-13T05:16:25.893596+00:00","evaluation_input":{"reference":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"c880b9724e2f3a0a45ebff876255c6c546ef70f295222c76fc040f113fdaf5e7","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":79760,"output_samples":79841,"length_delta_samples":81,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.916673184518185,"output_peak":0.40758177755279656},"common_prefix":{"samples":79760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998985483648752,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":79760,"output_samples":79760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.5988137842539505,"spectral_convergence":0.41847990428183823,"spectral_convergence_status":"ok","amplitude_floor":8.669731759960068e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.896486496537867,"spectral_convergence":0.1745889927591657,"spectral_convergence_status":"ok","amplitude_floor":7.258845275738271e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.399576179616482,"spectral_convergence":0.2049557718944498,"spectral_convergence_status":"ok","amplitude_floor":5.316334054535358e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.354150869521306e-05,"output_power":5.0074834475530494e-05,"delta_db":-0.29071082204851906,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0007507214796955326,"output_power":0.0007024601441267057,"delta_db":-0.28857153640588595,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006075331721980808,"output_power":0.0005812560516104913,"delta_db":-0.19202508554188105,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002980397625803834,"output_power":0.00030489372671505075,"delta_db":0.0987427994581719,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002296848467085225,"output_power":0.0002490395489077162,"delta_db":0.3513597714934657,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001162657066166175,"output_power":0.00014825527222967885,"delta_db":1.0555851119869615,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3794991878224809,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04204587342250944,"pitch":null,"output_rms_dbfs":-26.912265825131207,"output_peak":0.40758177755279656,"reference_rms_dbfs":-26.870219951708698,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.298292153469433},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":79760,"output_samples":79760,"intelligibility":{"estoi":{"value":0.9701623938862616,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04534078160436088,"rms_dbfs":-26.870219951708695,"peak":0.376739501953125,"peak_dbfs":-8.479176813835338,"crest_factor_db":18.39104313787336,"mean":5.436995801858701e-07,"zero_energy":false},"output":{"rms":0.04512183039355976,"rms_dbfs":-26.912265825131207,"peak":0.40758177755279656,"peak_dbfs":-7.795704819375758,"crest_factor_db":19.116561005755447,"mean":-0.0005539862537546031,"zero_energy":false},"rms_difference_db":-0.04204587342251287},"full_scale":{"reference":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":79760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":498,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.770658277673647e-06,"threshold_dbfs":-57.51865245947432,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":48,"quiet_samples":7680,"reference_mean_power":6.115447225359579e-07,"output_mean_power":5.944589565120351e-07,"output_rms_dbfs":-62.25878125192767,"output_zero_energy":false,"output_peak":0.004942720227807722,"output_energy_fraction":2.8114122449616733e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":491,"reference_estimated_voiced_frames":308,"output_estimated_voiced_frames":308,"both_estimated_voiced_frames":301,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9772727272727273,"absolute_difference_cents_median":3.0318390356345466,"absolute_difference_cents_p95":13.359420853699516,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":79760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.000134521265514355,"log_rms_mae_db":0.8102393388577834,"rms_envelope_nrmse":0.06358373866107812,"envelope_correlation":0.995312360021006,"rms_gain_error_db":-0.04204463860711778,"reference_mean_power":0.0020578494395295825,"output_mean_power":0.0020380232445223892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":5.569199266163003e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00039409678072813294,"output_power":0.00040938088980754854,"reference_active":true,"delta_db":0.16524678379035565},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.428667833313107e-05,"output_power":5.519659312188388e-05,"reference_active":true,"delta_db":0.07219003695825724},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373716545850661e-06,"output_power":7.381753248311478e-06,"reference_active":true,"delta_db":0.00473085022063928}],"active_band_count":3,"active_band_mae_db":0.08072255698975073}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00012498981593367327,"log_rms_mae_db":0.5246991835314503,"rms_envelope_nrmse":0.040612733411590995,"envelope_correlation":0.9979365608531271,"rms_gain_error_db":-0.0420469663008704,"reference_mean_power":0.0020703160518770356,"output_mean_power":0.002050368649295683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":4.869724313408462e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003604621899100366,"output_power":0.0003747083659063592,"reference_active":true,"delta_db":0.1683367261726213},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.585585870556879e-05,"output_power":2.5855054638306346e-05,"reference_active":true,"delta_db":-0.00013505929507175477},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.932164327326653e-07,"output_power":3.9378309370536835e-07,"reference_active":true,"delta_db":0.006254076853073049}],"active_band_count":3,"active_band_mae_db":0.05824195410692204}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":0.00011089795767449635,"log_rms_mae_db":0.8513878435160037,"rms_envelope_nrmse":0.061577199491901276,"envelope_correlation":0.9971499875406933,"rms_gain_error_db":-0.24016300602435325,"reference_mean_power":0.001090370195617981,"output_mean_power":0.0010317100742976803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":3.4652335334077796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023090422993351253,"output_power":0.0002156367681841907,"reference_active":true,"delta_db":-0.2970907453074185},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.174073740387618e-05,"output_power":4.00574208863693e-05,"reference_active":true,"delta_db":-0.17877133540825915},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.679949697911219e-06,"output_power":4.288440888272286e-06,"reference_active":true,"delta_db":-0.3794175696338473}],"active_band_count":3,"active_band_mae_db":0.285093216783175}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":0.00010160775009558359,"log_rms_mae_db":0.6308143553134623,"rms_envelope_nrmse":0.04632446006877776,"envelope_correlation":0.9985893875041726,"rms_gain_error_db":-0.24016285112447527,"reference_mean_power":0.0010969783403958974,"output_mean_power":0.0010379627487873385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":3.031226473561925e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021118169575363215,"output_power":0.00019714644159503304,"reference_active":true,"delta_db":-0.29867330209799225},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975380679045621e-05,"output_power":1.8990302649661744e-05,"reference_active":true,"delta_db":-0.17118915501720347},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.773024908316542e-07,"output_power":2.5835057040083297e-07,"reference_active":true,"delta_db":-0.3074434584372337}],"active_band_count":3,"active_band_mae_db":0.2591019718508098}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":7.327188751847547e-05,"log_rms_mae_db":0.8387744781136435,"rms_envelope_nrmse":0.07500761974470282,"envelope_correlation":0.9954646703022757,"rms_gain_error_db":0.1779796383781432,"reference_mean_power":0.00039491635058927065,"output_mean_power":0.00041143674822413494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":2.1879792460690976e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000156993183862756,"output_power":0.00016631044390900572,"reference_active":true,"delta_db":0.25038725604975876},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3083527056032408e-05,"output_power":1.4105128215777659e-05,"reference_active":true,"delta_db":0.3265220141237834},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.549781093186999e-06,"output_power":2.6361225250107284e-06,"reference_active":true,"delta_db":0.14462695602566386}],"active_band_count":3,"active_band_mae_db":0.24051207539973532}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":7.131903461227583e-05,"log_rms_mae_db":0.5538640381330411,"rms_envelope_nrmse":0.048465441521210184,"envelope_correlation":0.9981942264464507,"rms_gain_error_db":0.17797160089348557,"reference_mean_power":0.0003973070610776642,"output_mean_power":0.0004139267024140771},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":1.9730997112139501e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014395632169995678,"output_power":0.00015259671653837048,"reference_active":true,"delta_db":0.2531444764973614},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.34323278305194e-06,"output_power":6.770282061542409e-06,"reference_active":true,"delta_db":0.2829611314688554},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5197840492846452e-07,"output_power":1.6016726423628104e-07,"reference_active":true,"delta_db":0.22791875490179303}],"active_band_count":3,"active_band_mae_db":0.25467478762266993}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":996,"rms_floor":4.2584167929876904e-05,"log_rms_mae_db":1.319793471916838,"rms_envelope_nrmse":0.17130789529569382,"envelope_correlation":0.9843383544050567,"rms_gain_error_db":0.8566101080954246,"reference_mean_power":0.00017446384410387482,"output_mean_power":0.00021250367654490314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":996,"frequency_resolution_hz":0.2008032128514056,"power_floor":1.0543054012178768e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.162385199300118e-05,"output_power":8.78831431347062e-05,"reference_active":true,"delta_db":0.8884790700057121},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.778561151416548e-06,"output_power":1.2915805413861771e-05,"reference_active":true,"delta_db":1.208465374101004},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.445912836040226e-06,"output_power":2.5995549338652505e-06,"reference_active":true,"delta_db":2.547568862148533}],"active_band_count":3,"active_band_mae_db":1.5481711020850832}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":79760,"uncovered_tail_samples":0,"metrics":{"frame_count":990,"rms_floor":3.853974797463784e-05,"log_rms_mae_db":1.0495135288758042,"rms_envelope_nrmse":0.1296128590998195,"envelope_correlation":0.9943769963931419,"rms_gain_error_db":0.8566038861790798,"reference_mean_power":0.0001755196839873955,"output_mean_power":0.00021378942387268815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":990,"frequency_resolution_hz":0.20202020202020202,"power_floor":9.383868144458726e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.640153736830106e-05,"output_power":8.190948989069201e-05,"reference_active":true,"delta_db":0.9115608674565511},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.230654239705997e-06,"output_power":5.412730792289473e-06,"reference_active":true,"delta_db":1.0700889451654736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2239857829028075e-07,"output_power":2.0518086673903067e-07,"reference_active":true,"delta_db":2.2436048673005056}],"active_band_count":3,"active_band_mae_db":1.4084182266408434}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"hnm","status":"measured","mos":{"status":"measured","value":3.3131749629974365,"predicted_mos":3.3131749629974365,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44240,"input_sample_rate":16000,"input_seconds":2.765,"input_float32_sha256":"b2a6a5f91a2c6fd5e2a759aa6ab0f5ffb5e0867fd527762a1d9f83c14dca3256","internal_samples":44240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5127149220006686},"started_at":"2026-10-13T04:39:34.477365+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.0021000131673645e-05,"native_input_samples":44240,"native_input_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","output_sha256":"df54514d7a564f9f517a1995a30ea4ea52f2702773d0be3d252df50af01f7886","output_samples":44240,"length_delta_samples":0,"analysis_s":0.6224656670001423,"synthesis_s":0.32606936800038966,"total_s":0.9485350350005319,"rtf":0.34305064556981263,"output_peak":0.4002194234907295,"output_rms":0.035991362253511336,"finished_at":"2026-10-13T04:39:35.430000+00:00","evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"df54514d7a564f9f517a1995a30ea4ea52f2702773d0be3d252df50af01f7886","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.115532822675523,"spectral_convergence":0.21848512735815565,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.286128061124068,"spectral_convergence":0.22397549799180544,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.83355147239862,"spectral_convergence":0.21911898040618827,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":6.467960461989503e-05,"delta_db":0.12613010715652917,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00019690845263076282,"delta_db":0.0631993370260933,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0005424503674886141,"delta_db":-0.11328127941440864,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.00029687739451246965,"delta_db":-0.13821293979015634,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00015134916789526622,"delta_db":-0.07971363532474665,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":4.311316971647289e-05,"delta_db":-0.6141237656315016,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18911017739057265,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0948146087173639,"pitch":null,"output_rms_dbfs":-28.876034304515237,"output_peak":0.4002194234907295,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.41173745206607},"common_prefix":{"samples":44240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.115532822675523,"spectral_convergence":0.21848512735815565,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.286128061124068,"spectral_convergence":0.22397549799180544,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.83355147239862,"spectral_convergence":0.21911898040618827,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":6.467960461989503e-05,"delta_db":0.12613010715652917,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00019690845263076282,"delta_db":0.0631993370260933,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0005424503674886141,"delta_db":-0.11328127941440864,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.00029687739451246965,"delta_db":-0.13821293979015634,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00015134916789526622,"delta_db":-0.07971363532474665,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":4.311316971647289e-05,"delta_db":-0.6141237656315016,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18911017739057265,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0948146087173639,"pitch":null,"output_rms_dbfs":-28.876034304515237,"output_peak":0.4002194234907295,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.41173745206607},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44240,"output_samples":44240,"intelligibility":{"estoi":{"value":0.9575199959773745,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03638639377946353,"rms_dbfs":-28.781219695797876,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.243327098875167,"mean":1.0381771039359177e-06,"zero_energy":false},"output":{"rms":0.035991362253511336,"rms_dbfs":-28.876034304515237,"peak":0.4002194234907295,"peak_dbfs":-7.954036759266355,"crest_factor_db":20.921997545248882,"mean":-2.552541824229526e-05,"zero_energy":false},"rms_difference_db":-0.09481460871736047},"full_scale":{"reference":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":2.2684388787836406e-07,"output_rms_dbfs":-66.4427291794727,"output_zero_energy":false,"output_peak":0.004167976278740368,"output_energy_fraction":5.0033682481376636e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":122,"both_estimated_voiced_frames":122,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":0,"paired_voiced_fraction":0.9037037037037037,"absolute_difference_cents_median":1.10179893098278,"absolute_difference_cents_p95":4.9250593173492225,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":0.00013943874660832744,"log_rms_mae_db":0.848231518826578,"rms_envelope_nrmse":0.05812750668414989,"envelope_correlation":0.9971479531107569,"rms_gain_error_db":-0.09481478230288892,"reference_mean_power":0.0013263679977567136,"output_mean_power":0.00129772465753421},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":5.047819293813421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030810693936871264,"output_power":0.00030027694654624886,"reference_active":true,"delta_db":-0.11179488900453682},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2023354078518551e-05,"output_power":1.144529629937672e-05,"reference_active":true,"delta_db":-0.21398596628097127},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2193973896229425e-06,"output_power":4.830533947139758e-06,"reference_active":true,"delta_db":1.7622055093150597}],"active_band_count":3,"active_band_mae_db":0.695995454866856}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":0.0001327212418155818,"log_rms_mae_db":0.4558034829427853,"rms_envelope_nrmse":0.025862450094136645,"envelope_correlation":0.9994930891556656,"rms_gain_error_db":-0.09481503682194782,"reference_mean_power":0.0013409420747202037,"output_mean_power":0.001311983925765295},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":4.5645799716966823e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002745054731183271,"output_power":0.0002690371832577291,"reference_active":true,"delta_db":-0.08738700488921014},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.376005868866601e-06,"output_power":4.947199980297122e-06,"reference_active":true,"delta_db":-0.3610026798312277},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7970463694996414e-07,"output_power":2.2639474527073665e-07,"reference_active":true,"delta_db":1.0030705902308572}],"active_band_count":3,"active_band_mae_db":0.48382009165043166}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":1.077349689362437,"rms_envelope_nrmse":0.06298216858610566,"envelope_correlation":0.9970869271713753,"rms_gain_error_db":-0.07191198855375333,"reference_mean_power":0.0005849439303269558,"output_mean_power":0.0005753379744412168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":2.471278629444775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014231281864882344,"output_power":0.00013838139717575797,"reference_active":true,"delta_db":-0.12166309208561218},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.153370236673123e-06,"output_power":8.587791247569531e-06,"reference_active":true,"delta_db":0.22544315376924903},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4861340139926787e-06,"output_power":2.3235383868803006e-06,"reference_active":true,"delta_db":1.9409187755939394}],"active_band_count":3,"active_band_mae_db":0.7626750071496002}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":0.5745175123141184,"rms_envelope_nrmse":0.030016065887931254,"envelope_correlation":0.999352448356359,"rms_gain_error_db":-0.0719120390233615,"reference_mean_power":0.0005913717973798096,"output_mean_power":0.0005816602762286148},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":2.253861652661314e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012623197394568146,"output_power":0.0001234567678529617,"reference_active":true,"delta_db":-0.09654470745945792},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1401147597866345e-06,"output_power":3.4656911934847614e-06,"reference_active":true,"delta_db":0.4284434253176297},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.001702937868675e-08,"output_power":1.2048340568108286e-07,"reference_active":true,"delta_db":1.7774481092931271}],"active_band_count":3,"active_band_mae_db":0.7674787473567383}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":1.337919411116897,"rms_envelope_nrmse":0.09997585104661601,"envelope_correlation":0.9929267253197319,"rms_gain_error_db":-0.1635859617986171,"reference_mean_power":0.0003354294563993913,"output_mean_power":0.00032302981016992655},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":1.775200853272881e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012560860558913709,"output_power":0.00012378665091449594,"reference_active":true,"delta_db":-0.06345581304719654},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.72605183156066e-06,"output_power":7.794488829973097e-06,"reference_active":true,"delta_db":-0.4903014929704933},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2668523428769504e-06,"output_power":2.534618197673138e-06,"reference_active":true,"delta_db":3.0118654969540297}],"active_band_count":3,"active_band_mae_db":1.1885409343239066}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":0.8200023120767264,"rms_envelope_nrmse":0.05809170620981227,"envelope_correlation":0.9976725338994851,"rms_gain_error_db":-0.16358617444840837,"reference_mean_power":0.0003391152422568523,"output_mean_power":0.0003265793295081148},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":1.6261600138727939e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011427350869395267,"output_power":0.00011392355973658662,"reference_active":true,"delta_db":-0.013320155365462128},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1799749558208305e-06,"output_power":3.7528281613181768e-06,"reference_active":true,"delta_db":-0.4681500081387719},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395722090626045e-08,"output_power":1.3211408338058747e-07,"reference_active":true,"delta_db":1.480189533292082}],"active_band_count":3,"active_band_mae_db":0.653886565598772}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":1.4771932224819206,"rms_envelope_nrmse":0.22737861308317814,"envelope_correlation":0.9634476320278301,"rms_gain_error_db":-0.2493355561679909,"reference_mean_power":0.0001243129634607548,"output_mean_power":0.00011737696179765116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":7.893026272799498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.194673883780046e-05,"output_power":3.510991108756535e-05,"reference_active":true,"delta_db":-0.7726847245997599},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.051190096168371e-06,"output_power":6.182719855255986e-06,"reference_active":true,"delta_db":-0.5708285451206841},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.501810525085992e-07,"output_power":2.5801752632061573e-06,"reference_active":true,"delta_db":4.338428412912691}],"active_band_count":3,"active_band_mae_db":1.8939805608777116}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":1.098649378471302,"rms_envelope_nrmse":0.1321724656946064,"envelope_correlation":0.9874243493385944,"rms_gain_error_db":-0.24933699842786397,"reference_mean_power":0.00012567880182247303,"output_mean_power":0.00011866655424063243},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":7.206643574889897e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8301143786829975e-05,"output_power":3.266123918089165e-05,"reference_active":true,"delta_db":-0.691790854707349},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6851867496780295e-06,"output_power":3.345630931693859e-06,"reference_active":true,"delta_db":-0.41981470242176144},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.995617022655136e-08,"output_power":1.5245057631545216e-07,"reference_active":true,"delta_db":4.05295186513713}],"active_band_count":3,"active_band_mae_db":1.7215191407554133}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.6375951766967773,"predicted_mos":2.6375951766967773,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44240,"input_sample_rate":16000,"input_seconds":2.765,"input_float32_sha256":"8e5a219481bb006a37ed677d3f04c4f35d927130b6b7c90d5702e51c507606c6","internal_samples":44240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.841140679000091},"started_at":"2026-10-13T04:42:33.874063+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2477999664261006e-05,"native_input_samples":44240,"native_input_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","output_sha256":"0b60be5fe11b5857eb4f6b18608f7c58350305042b2a2400f8fcc4a6ed5d63d5","output_samples":44240,"length_delta_samples":0,"analysis_s":0.6865269859999898,"synthesis_s":0.04670085399993695,"total_s":0.7332278399999268,"rtf":0.26518185895114893,"output_peak":0.46011113013111804,"output_rms":0.04056636397548155,"finished_at":"2026-10-13T04:42:34.610445+00:00","evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"0b60be5fe11b5857eb4f6b18608f7c58350305042b2a2400f8fcc4a6ed5d63d5","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.535497948803376,"spectral_convergence":0.6806833821414242,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.303124725978127,"spectral_convergence":0.2746719614291644,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9843872519168935,"spectral_convergence":0.3219041465595851,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":7.965111315845934e-05,"delta_db":1.0303750390408453,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00020976469223468074,"delta_db":0.337879635389598,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0007259818301977086,"delta_db":1.1523761566273827,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.00038876300388530346,"delta_db":1.032865053291045,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00018819983112328966,"delta_db":0.8666782826958492,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":5.3269415591805653e-05,"delta_db":0.30455601966165113,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7874550311177285,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9445413553558728,"pitch":null,"output_rms_dbfs":-27.836678340442003,"output_peak":0.46011113013111804,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.941003308899465},"common_prefix":{"samples":44240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.535497948803376,"spectral_convergence":0.6806833821414242,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.303124725978127,"spectral_convergence":0.2746719614291644,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9843872519168935,"spectral_convergence":0.3219041465595851,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":7.965111315845934e-05,"delta_db":1.0303750390408453,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00020976469223468074,"delta_db":0.337879635389598,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0007259818301977086,"delta_db":1.1523761566273827,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.00038876300388530346,"delta_db":1.032865053291045,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00018819983112328966,"delta_db":0.8666782826958492,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":5.3269415591805653e-05,"delta_db":0.30455601966165113,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7874550311177285,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9445413553558728,"pitch":null,"output_rms_dbfs":-27.836678340442003,"output_peak":0.46011113013111804,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.941003308899465},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44240,"output_samples":44240,"intelligibility":{"estoi":{"value":0.9274910365518466,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03638639377946353,"rms_dbfs":-28.781219695797876,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.243327098875167,"mean":1.0381771039359177e-06,"zero_energy":false},"output":{"rms":0.04056636397548155,"rms_dbfs":-27.836678340442003,"peak":0.46011113013111804,"peak_dbfs":-6.742745219682887,"crest_factor_db":21.093933120759118,"mean":-4.814142503654608e-10,"zero_energy":false},"rms_difference_db":0.9445413553558737},"full_scale":{"reference":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":3.140028080643792e-07,"output_rms_dbfs":-65.03066468100633,"output_zero_energy":false,"output_peak":0.004429634239223732,"output_energy_fraction":5.4517172282302485e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":121,"both_estimated_voiced_frames":110,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8148148148148148,"absolute_difference_cents_median":3.8917100051341977,"absolute_difference_cents_p95":20.781374084242987,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":0.00013943874660832744,"log_rms_mae_db":1.6434070855148513,"rms_envelope_nrmse":0.1563152802804347,"envelope_correlation":0.9915482495289031,"rms_gain_error_db":0.9445417058583518,"reference_mean_power":0.0013263679977567136,"output_mean_power":0.0016486110460350104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":5.047819293813421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030810693936871264,"output_power":0.0003796338712056691,"reference_active":true,"delta_db":0.9066347368130562},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2023354078518551e-05,"output_power":1.60586789314569e-05,"reference_active":true,"delta_db":1.2568417820656077},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2193973896229425e-06,"output_power":4.397231405217579e-06,"reference_active":true,"delta_db":1.3540473347875512}],"active_band_count":3,"active_band_mae_db":1.1725079512220715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":0.0001327212418155818,"log_rms_mae_db":1.296018028004741,"rms_envelope_nrmse":0.12897033392422305,"envelope_correlation":0.9974323999331162,"rms_gain_error_db":0.9445436779646328,"reference_mean_power":0.0013409420747202037,"output_mean_power":0.0016667266731996689},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":4.5645799716966823e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002745054731183271,"output_power":0.0003396414855792187,"reference_active":true,"delta_db":0.9246972392041313},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.376005868866601e-06,"output_power":7.596722389310485e-06,"reference_active":true,"delta_db":1.501665219058874},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7970463694996414e-07,"output_power":2.7633372429620165e-07,"reference_active":true,"delta_db":1.868746068506626}],"active_band_count":3,"active_band_mae_db":1.4317028422565439}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":1.5303233214061671,"rms_envelope_nrmse":0.1814043326067319,"envelope_correlation":0.9921126937367517,"rms_gain_error_db":1.1606516686051307,"reference_mean_power":0.0005849439303269558,"output_mean_power":0.0007641513870087509},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":2.471278629444775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014231281864882344,"output_power":0.00018193125175989348,"reference_active":true,"delta_db":1.0666328726810455},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.153370236673123e-06,"output_power":1.1261658500741781e-05,"reference_active":true,"delta_db":1.4026518996574673},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4861340139926787e-06,"output_power":2.2494533668712257e-06,"reference_active":true,"delta_db":1.8001902005348227}],"active_band_count":3,"active_band_mae_db":1.4231583242911119}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":1.1693201930431154,"rms_envelope_nrmse":0.16065935660183728,"envelope_correlation":0.996571936759642,"rms_gain_error_db":1.160651306491003,"reference_mean_power":0.0005913717973798096,"output_mean_power":0.0007725484753280461},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":2.253861652661314e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012623197394568146,"output_power":0.0001632564355195837,"reference_active":true,"delta_db":1.117009365179286},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1401147597866345e-06,"output_power":4.436213776729459e-06,"reference_active":true,"delta_db":1.5006694595170789},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.001702937868675e-08,"output_power":1.5041756706367852e-07,"reference_active":true,"delta_db":2.741161356624666}],"active_band_count":3,"active_band_mae_db":1.786280060440344}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":1.3758574093199805,"rms_envelope_nrmse":0.15720325720739675,"envelope_correlation":0.992284007764147,"rms_gain_error_db":0.9272960168940553,"reference_mean_power":0.0003354294563993913,"output_mean_power":0.00041527023157821746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":1.775200853272881e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012560860558913709,"output_power":0.00014520956176775456,"reference_active":true,"delta_db":0.6297582032240944},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.72605183156066e-06,"output_power":1.2392637441536964e-05,"reference_active":true,"delta_db":1.5234595587299602},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2668523428769504e-06,"output_power":1.8459767031336301e-06,"reference_active":true,"delta_db":1.6350021684565874}],"active_band_count":3,"active_band_mae_db":1.2627399768035474}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":1.0438006110660007,"rms_envelope_nrmse":0.13694234376781414,"envelope_correlation":0.9959385833949216,"rms_gain_error_db":0.9272963950462042,"reference_mean_power":0.0003391152422568523,"output_mean_power":0.0004198333651353582},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":1.6261600138727939e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011427350869395267,"output_power":0.0001322385186681735,"reference_active":true,"delta_db":0.6341241332152541},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1799749558208305e-06,"output_power":6.21951035287027e-06,"reference_active":true,"delta_db":1.725825153518602},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395722090626045e-08,"output_power":1.389288087307611e-07,"reference_active":true,"delta_db":1.698621491876563}],"active_band_count":3,"active_band_mae_db":1.352856926203473}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":1.5084821315462549,"rms_envelope_nrmse":0.19669958100601148,"envelope_correlation":0.9827788772311497,"rms_gain_error_db":0.8672977602649944,"reference_mean_power":0.0001243129634607548,"output_mean_power":0.0001517910606449664},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":7.893026272799498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.194673883780046e-05,"output_power":4.99373885931242e-05,"reference_active":true,"delta_db":0.7572762565987041},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.051190096168371e-06,"output_power":6.407402018783828e-06,"reference_active":true,"delta_db":-0.4158044944446595},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.501810525085992e-07,"output_power":1.3336816153325207e-06,"reference_active":true,"delta_db":1.4724579857261304}],"active_band_count":3,"active_band_mae_db":0.8818462455898314}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":1.0154719024145877,"rms_envelope_nrmse":0.14243761503148178,"envelope_correlation":0.9941578180485416,"rms_gain_error_db":0.8672995506474876,"reference_mean_power":0.00012567880182247303,"output_mean_power":0.00015345886673935583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":7.206643574889897e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8301143786829975e-05,"output_power":4.5619803406594715e-05,"reference_active":true,"delta_db":0.7594166590568276},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6851867496780295e-06,"output_power":3.399549777998273e-06,"reference_active":true,"delta_db":-0.3503809623501208},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.995617022655136e-08,"output_power":1.006587622441035e-07,"reference_active":true,"delta_db":2.2501770179559815}],"active_band_count":3,"active_band_mae_db":1.11999154645431}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"facodec","status":"measured","mos":{"status":"measured","value":4.30551290512085,"predicted_mos":4.30551290512085,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44400,"input_sample_rate":16000,"input_seconds":2.775,"input_float32_sha256":"de8276e8e9aa751dcd0896929c1079002ceb994e9b926dced8c9364e3588a8ad","internal_samples":44400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.721424573001059},"started_at":"2026-10-13T04:54:35.059981+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.7544000153720845e-05,"native_input_samples":44240,"native_input_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","output_sha256":"42822519b2617c2f6e252934cfeccbc5626ee9d370d32671664556df07fc6fe8","output_samples":44400,"length_delta_samples":160,"analysis_s":0.6701391589995183,"synthesis_s":4.629027648000374,"total_s":5.299166806999892,"rtf":1.9165160242314256,"output_peak":0.43084344267845154,"output_rms":0.03535669031091193,"finished_at":"2026-10-13T04:54:40.364161+00:00","evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"42822519b2617c2f6e252934cfeccbc5626ee9d370d32671664556df07fc6fe8","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":44240,"output_samples":44400,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.030567908802137,"output_peak":0.43084344267845154},"common_prefix":{"samples":44240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9963963963963964,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.132791437060096,"spectral_convergence":0.2045519315025601,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.684671171055512,"spectral_convergence":0.24119209438705158,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.8349254702759765,"spectral_convergence":0.25901136497631455,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":6.673644627179402e-05,"delta_db":0.26208730945281944,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00021441743644433407,"delta_db":0.43315673655222753,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0005235798968288087,"delta_db":-0.2670517249762969,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.0002862031558145047,"delta_db":-0.2972400564292332,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00011679524347471012,"delta_db":-1.2052624476911775,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":4.688449980690995e-05,"delta_db":-0.24993043241851182,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.45245478458671107,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23366973750355707,"pitch":null,"output_rms_dbfs":-29.014889433301434,"output_peak":0.43084344267845154,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.550796026130528},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44240,"output_samples":44240,"intelligibility":{"estoi":{"value":0.9572073136950763,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03638639377946353,"rms_dbfs":-28.781219695797876,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.243327098875167,"mean":1.0381771039359177e-06,"zero_energy":false},"output":{"rms":0.03542056858156093,"rms_dbfs":-29.014889433301434,"peak":0.43084344267845154,"peak_dbfs":-7.313610250270953,"crest_factor_db":21.70127918303048,"mean":-3.245250172257712e-05,"zero_energy":false},"rms_difference_db":-0.23366973750355768},"full_scale":{"reference":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":2.7471925383211967e-07,"output_rms_dbfs":-65.61110901773407,"output_zero_energy":false,"output_peak":0.004533193074166775,"output_energy_fraction":6.256190971861096e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":132,"both_estimated_voiced_frames":129,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9555555555555556,"absolute_difference_cents_median":2.262796785208136,"absolute_difference_cents_p95":9.480698750652287,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":0.00013943874660832744,"log_rms_mae_db":1.1109670645506597,"rms_envelope_nrmse":0.07024379519956046,"envelope_correlation":0.9962398791120208,"rms_gain_error_db":-0.23367042205917038,"reference_mean_power":0.0013263679977567136,"output_mean_power":0.0012568891944329751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":5.047819293813421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030810693936871264,"output_power":0.00027598172698711267,"reference_active":true,"delta_db":-0.47821151901073855},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2023354078518551e-05,"output_power":1.2877955355670647e-05,"reference_active":true,"delta_db":0.2982127824489135},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2193973896229425e-06,"output_power":2.859376405133245e-06,"reference_active":true,"delta_db":-0.5150325841548884}],"active_band_count":3,"active_band_mae_db":0.4304856285381802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":0.0001327212418155818,"log_rms_mae_db":0.8287144058743042,"rms_envelope_nrmse":0.06226141940150273,"envelope_correlation":0.9970262099796341,"rms_gain_error_db":-0.23367325016510365,"reference_mean_power":0.0013409420747202037,"output_mean_power":0.0012706990136917317},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":4.5645799716966823e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002745054731183271,"output_power":0.00024626260112770433,"reference_active":true,"delta_db":-0.47152545522115036},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.376005868866601e-06,"output_power":6.012620954394473e-06,"reference_active":true,"delta_db":0.4860409190772085},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7970463694996414e-07,"output_power":1.472529970072559e-07,"reference_active":true,"delta_db":-0.8649514078213605}],"active_band_count":3,"active_band_mae_db":0.6075059273732398}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":1.224657668532095,"rms_envelope_nrmse":0.09410941497418833,"envelope_correlation":0.9936160853601548,"rms_gain_error_db":-0.13534456224703703,"reference_mean_power":0.0005849439303269558,"output_mean_power":0.0005669957226650291},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":2.471278629444775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014231281864882344,"output_power":0.00014048718129455935,"reference_active":true,"delta_db":-0.056073213820108006},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.153370236673123e-06,"output_power":8.655154260950472e-06,"reference_active":true,"delta_db":0.25937649052715434},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4861340139926787e-06,"output_power":1.63291329559065e-06,"reference_active":true,"delta_db":0.40905150947695373}],"active_band_count":3,"active_band_mae_db":0.241500404608072}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":1.0785830590191838,"rms_envelope_nrmse":0.08334575029190028,"envelope_correlation":0.994841874014302,"rms_gain_error_db":-0.13534466021174257,"reference_mean_power":0.0005913717973798096,"output_mean_power":0.0005732263464433456},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":2.253861652661314e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012623197394568146,"output_power":0.0001246951174392536,"reference_active":true,"delta_db":-0.05319924943309379},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1401147597866345e-06,"output_power":3.2129799366757638e-06,"reference_active":true,"delta_db":0.0996249337672548},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.001702937868675e-08,"output_power":8.056624118158832e-08,"reference_active":true,"delta_db":0.029706778865405008}],"active_band_count":3,"active_band_mae_db":0.06084365402191786}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":1.653273462545192,"rms_envelope_nrmse":0.14081839860910986,"envelope_correlation":0.9883045766221046,"rms_gain_error_db":-0.47925952058116755,"reference_mean_power":0.0003354294563993913,"output_mean_power":0.00030038292801754424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":1.775200853272881e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012560860558913709,"output_power":0.00010420812077740845,"reference_active":true,"delta_db":-0.8111783025338752},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.72605183156066e-06,"output_power":9.007338310958553e-06,"reference_active":true,"delta_db":0.13778686617150532},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2668523428769504e-06,"output_power":1.4951524856306605e-06,"reference_active":true,"delta_db":0.7195948823760155}],"active_band_count":3,"active_band_mae_db":0.5561866836937988}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":1.3952755446546836,"rms_envelope_nrmse":0.12699494781493975,"envelope_correlation":0.9906379470379599,"rms_gain_error_db":-0.4792638668085316,"reference_mean_power":0.0003391152422568523,"output_mean_power":0.00030368330961915487},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":1.6261600138727939e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011427350869395267,"output_power":9.503106366137109e-05,"reference_active":true,"delta_db":-0.8007997206307647},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1799749558208305e-06,"output_power":4.721374816046413e-06,"reference_active":true,"delta_db":0.5289479943809562},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395722090626045e-08,"output_power":1.003421343740322e-07,"reference_active":true,"delta_db":0.2855317204444529}],"active_band_count":3,"active_band_mae_db":0.5384264784853913}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":1.8290054752387808,"rms_envelope_nrmse":0.1722987052242364,"envelope_correlation":0.9816904631088886,"rms_gain_error_db":-0.580002856268767,"reference_mean_power":0.0001243129634607548,"output_mean_power":0.00010877175454154999},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":7.893026272799498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.194673883780046e-05,"output_power":3.284667755303759e-05,"reference_active":true,"delta_db":-1.0620675496076193},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.051190096168371e-06,"output_power":3.869589206380613e-06,"reference_active":true,"delta_db":-2.6059756018793383},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.501810525085992e-07,"output_power":8.721010601966517e-07,"reference_active":true,"delta_db":-0.3723955148084835}],"active_band_count":3,"active_band_mae_db":1.346812888765147}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":1.5393878579473939,"rms_envelope_nrmse":0.14981942871304707,"envelope_correlation":0.9865456883930931,"rms_gain_error_db":-0.580007959847843,"reference_mean_power":0.00012567880182247303,"output_mean_power":0.00010996671093501962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":7.206643574889897e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8301143786829975e-05,"output_power":2.954634536579067e-05,"reference_active":true,"delta_db":-1.1270797352819895},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6851867496780295e-06,"output_power":2.2033835624173855e-06,"reference_active":true,"delta_db":-2.233693957561032},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.995617022655136e-08,"output_power":4.9599615497526884e-08,"reference_active":true,"delta_db":-0.8235557417518115}],"active_band_count":3,"active_band_mae_db":1.3947764781982777}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"hifiglot","status":"measured","mos":{"status":"measured","value":4.009105682373047,"predicted_mos":4.009105682373047,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":122368,"input_sample_rate":44100,"input_seconds":2.7747845804988662,"input_float32_sha256":"b5ec10a031771f2d4b7ed69154b7672f420ff590333425d9f7d460ba07328a2a","internal_samples":44397,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.025830468999629},"started_at":"2026-10-13T07:19:11.687054+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006450820999816642,"native_input_samples":121937,"native_input_sha256":"d4bbbf36da760e35e23c7a09bdc16e54467319b91e20d79c7aec5dcce30abd5f","output_sha256":"4f87cff0fc8063d94abfa7cf34e26a5a59e80e92dbcf51ecebc5d6b36c5ad315","output_samples":122368,"length_delta_samples":431,"analysis_s":0.07216408400017826,"synthesis_s":16.648341411999354,"total_s":16.720505495999532,"rtf":6.047174297986496,"output_peak":0.34977996349334717,"output_rms":0.0333356241596443,"finished_at":"2026-10-13T07:19:28.432793+00:00","evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"4f87cff0fc8063d94abfa7cf34e26a5a59e80e92dbcf51ecebc5d6b36c5ad315","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":44240,"output_samples":44397,"length_delta_samples":157,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.533729378190515,"output_peak":0.347962651599719},"common_prefix":{"samples":44240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9964637250264657,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.234380101013166,"spectral_convergence":0.5117874955500122,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.495925767033876,"spectral_convergence":0.44080917590077046,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.176687650798353,"spectral_convergence":0.4788175157610356,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":4.2446730802261716e-05,"delta_db":-1.7031009871394112,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00013513931259817447,"delta_db":-1.5716272016091541,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0005892069168317755,"delta_db":0.2457970078704681,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.0002075365190139249,"delta_db":-1.6930389574633284,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00011212766585680853,"delta_db":-1.3823861950816616,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":3.083197127376655e-05,"delta_db":-2.070210376079772,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4443601208739658,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7371246213889793,"pitch":null,"output_rms_dbfs":-29.518344317186855,"output_peak":0.347962651599719,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.302331172948465},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44240,"output_samples":44240,"intelligibility":{"estoi":{"value":0.8503051995246004,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03638639377946353,"rms_dbfs":-28.781219695797876,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.243327098875167,"mean":1.0381771039359177e-06,"zero_energy":false},"output":{"rms":0.03342587495304666,"rms_dbfs":-29.518344317186855,"peak":0.347962651599719,"peak_dbfs":-9.169347366735774,"crest_factor_db":20.34899695045108,"mean":-6.663820698930998e-06,"zero_energy":false},"rms_difference_db":-0.737124621388979},"full_scale":{"reference":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":1.5334489576677845e-07,"output_rms_dbfs":-68.14330675352936,"output_zero_energy":false,"output_peak":0.0033680595955300975,"output_energy_fraction":3.9213509484473055e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":128,"both_estimated_voiced_frames":121,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.8962962962962963,"absolute_difference_cents_median":8.370721093118277,"absolute_difference_cents_p95":22.93394415048069,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":0.00013943874660832744,"log_rms_mae_db":1.6941038594482454,"rms_envelope_nrmse":0.12847638835591918,"envelope_correlation":0.9908173554715192,"rms_gain_error_db":-0.7371244297760939,"reference_mean_power":0.0013263679977567136,"output_mean_power":0.0011193131134404801},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":5.047819293813421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030810693936871264,"output_power":0.00026642442827212563,"reference_active":true,"delta_db":-0.631274373815786},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2023354078518551e-05,"output_power":1.0600137694758438e-05,"reference_active":true,"delta_db":-0.5471413018977767},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2193973896229425e-06,"output_power":4.065524405093351e-06,"reference_active":true,"delta_db":1.0134198487165158}],"active_band_count":3,"active_band_mae_db":0.7306118414766928}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":0.0001327212418155818,"log_rms_mae_db":1.4914332678512294,"rms_envelope_nrmse":0.10546162683655008,"envelope_correlation":0.9956281910666164,"rms_gain_error_db":-0.7371224572184785,"reference_mean_power":0.0013409420747202037,"output_mean_power":0.0011316125938341407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":4.5645799716966823e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002745054731183271,"output_power":0.00024005357696565068,"reference_active":true,"delta_db":-0.5824282623648757},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.376005868866601e-06,"output_power":4.340450102069451e-06,"reference_active":true,"delta_db":-0.9292496623091033},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7970463694996414e-07,"output_power":2.3089784514757196e-07,"reference_active":true,"delta_db":1.0886059645654345}],"active_band_count":3,"active_band_mae_db":0.8667612964131378}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":2.307403427877352,"rms_envelope_nrmse":0.20626875377623308,"envelope_correlation":0.9696578450954285,"rms_gain_error_db":-0.5470932967769727,"reference_mean_power":0.0005849439303269558,"output_mean_power":0.0005157092357658372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":2.471278629444775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014231281864882344,"output_power":0.00012557342778985264,"reference_active":true,"delta_db":-0.5434627098537558},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.153370236673123e-06,"output_power":7.882556549616241e-06,"reference_active":true,"delta_db":-0.14670068596754407},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4861340139926787e-06,"output_power":3.022076402116481e-06,"reference_active":true,"delta_db":3.082474654464149}],"active_band_count":3,"active_band_mae_db":1.2575460167618162}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":1.920848922883478,"rms_envelope_nrmse":0.17718433027861225,"envelope_correlation":0.9777574017684304,"rms_gain_error_db":-0.5470944208642401,"reference_mean_power":0.0005913717973798096,"output_mean_power":0.0005213761574936627},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":2.253861652661314e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012623197394568146,"output_power":0.00011317070112829553,"reference_active":true,"delta_db":-0.47435367036820336},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1401147597866345e-06,"output_power":2.500743888502229e-06,"reference_active":true,"delta_db":-0.9887630413032961},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.001702937868675e-08,"output_power":1.4733096482453807e-07,"reference_active":true,"delta_db":2.6511160870090222}],"active_band_count":3,"active_band_mae_db":1.3714109328935074}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":2.9501666477595943,"rms_envelope_nrmse":0.22505818416626533,"envelope_correlation":0.9756651461886182,"rms_gain_error_db":-1.2810926613441664,"reference_mean_power":0.0003354294563993913,"output_mean_power":0.0002497421994836908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":1.775200853272881e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012560860558913709,"output_power":0.00011281260295125029,"reference_active":true,"delta_db":-0.46661774514627324},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.72605183156066e-06,"output_power":7.5894223784080515e-06,"reference_active":true,"delta_db":-0.6060906477359284},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2668523428769504e-06,"output_power":1.568035721363099e-06,"reference_active":true,"delta_db":0.9262995317565759}],"active_band_count":3,"active_band_mae_db":0.6663359748795925}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":2.650268775739959,"rms_envelope_nrmse":0.20396488396277032,"envelope_correlation":0.9821416148625451,"rms_gain_error_db":-1.2810935075354655,"reference_mean_power":0.0003391152422568523,"output_mean_power":0.000252486382344815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":1.6261600138727939e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011427350869395267,"output_power":0.00010394792380238933,"reference_active":true,"delta_db":-0.41129742916286083},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1799749558208305e-06,"output_power":4.102815160963893e-06,"reference_active":true,"delta_db":-0.08091728044371123},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395722090626045e-08,"output_power":1.5262580331573604e-07,"reference_active":true,"delta_db":2.1069780017380486}],"active_band_count":3,"active_band_mae_db":0.8663975704482069}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":3.5361454133952264,"rms_envelope_nrmse":0.24835756771226522,"envelope_correlation":0.9675380608559634,"rms_gain_error_db":-1.35518816865812,"reference_mean_power":0.0001243129634607548,"output_mean_power":9.099082500045449e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":7.893026272799498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.194673883780046e-05,"output_power":3.071978176196683e-05,"reference_active":true,"delta_db":-1.3528007604224976},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.051190096168371e-06,"output_power":4.3300654325124204e-06,"reference_active":true,"delta_db":-2.1176796405637126},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.501810525085992e-07,"output_power":6.89495822532619e-07,"reference_active":true,"delta_db":-1.392747266869223}],"active_band_count":3,"active_band_mae_db":1.6210758892851445}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":3.350439278835443,"rms_envelope_nrmse":0.20467666398471918,"envelope_correlation":0.9827834633323256,"rms_gain_error_db":-1.3551841709352677,"reference_mean_power":0.00012567880182247303,"output_mean_power":9.199063453352313e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":7.206643574889897e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8301143786829975e-05,"output_power":2.8184814624223028e-05,"reference_active":true,"delta_db":-1.3319656075053095},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6851867496780295e-06,"output_power":2.186187561239692e-06,"reference_active":true,"delta_db":-2.2677208199381416},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.995617022655136e-08,"output_power":3.94051596743298e-08,"reference_active":true,"delta_db":-1.8228079231533953}],"active_band_count":3,"active_band_mae_db":1.8074981168656155}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"golf","status":"measured","mos":{"status":"measured","value":3.321983814239502,"predicted_mos":3.321983814239502,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66240,"input_sample_rate":24000,"input_seconds":2.76,"input_float32_sha256":"0cbd9effa04041b875345851a1ed5fe55f204084994935bb9a87c125712b5beb","internal_samples":44160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.202681180999207},"started_at":"2026-10-13T05:02:42.337665+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0015644679997421918,"native_input_samples":66360,"native_input_sha256":"83deac5ff8c4cfaf3dbef9f54ef132dcf4b93465fc942416de066077e40172b7","output_sha256":"2fc538d4bf0560fd9666ca574fd6dffda11b5b106569aa80dc3ba1d65895eeb5","output_samples":66240,"length_delta_samples":-120,"analysis_s":0.2520276550003473,"synthesis_s":0.07184526099990762,"total_s":0.3238729160002549,"rtf":0.11713306184457682,"output_peak":0.3847189247608185,"output_rms":0.031250712889673335,"finished_at":"2026-10-13T05:02:42.668167+00:00","analysis_stages":{"external_f0_s":0.043440157000077306},"evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"2fc538d4bf0560fd9666ca574fd6dffda11b5b106569aa80dc3ba1d65895eeb5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44240,"output_samples":44160,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.096123279372744,"output_peak":0.39310746958218584},"common_prefix":{"samples":44160,"sample_rate":16000,"reference_coverage":0.9981916817359855,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44160,"output_samples":44160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.767171584681931,"spectral_convergence":0.5841182243910593,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.705271572108325,"spectral_convergence":0.2795142728243614,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.7696445274774995,"spectral_convergence":0.297583207659675,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.501737676046902e-05,"output_power":5.2839383153154976e-05,"delta_db":-0.900717041255617,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0001924386298667077,"output_power":0.0001690832376402486,"delta_db":-0.561917009362345,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005577000905104827,"output_power":0.00040207515746916726,"delta_db":-1.4209347426963803,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003070332847699586,"output_power":0.00023857132917726246,"delta_db":-1.0956720869549113,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015445943768447196,"output_power":9.096833234621655e-05,"delta_db":-2.2992421624868196,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.971926085746372e-05,"output_power":2.457249939226411e-05,"delta_db":-3.0607533065734676,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5565393915549235,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.322763898079591,"pitch":null,"output_rms_dbfs":-30.096123279372744,"output_peak":0.39310746958218584,"reference_rms_dbfs":-28.773359381293155,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.080695894755919},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44160,"output_samples":44160,"intelligibility":{"estoi":{"value":0.9389713358389887,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03641933662835656,"rms_dbfs":-28.77335938129315,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.235466784370445,"mean":1.2819317803866621e-06,"zero_energy":false},"output":{"rms":0.03127474922646565,"rms_dbfs":-30.096123279372744,"peak":0.39310746958218584,"peak_dbfs":-8.109774078189668,"crest_factor_db":21.986349201183074,"mean":0.00020869343566312973,"zero_energy":false},"rms_difference_db":-1.322763898079593},"full_scale":{"reference":{"sample_count":44160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":4.914333079921733e-07,"output_rms_dbfs":-63.08535411578001,"output_zero_energy":false,"output_peak":0.0031636235909819847,"output_energy_fraction":0.00014381193360902677,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":137,"both_estimated_voiced_frames":131,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9703703703703703,"absolute_difference_cents_median":4.517150356281585,"absolute_difference_cents_p95":17.163529297900364,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":551,"rms_floor":0.00013943874660832744,"log_rms_mae_db":2.7899334996815424,"rms_envelope_nrmse":0.1699842560532001,"envelope_correlation":0.9913513975813861,"rms_gain_error_db":-1.3227645888592385,"reference_mean_power":0.0013287750891698845,"output_mean_power":0.0009798847952398456},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":551,"frequency_resolution_hz":0.3629764065335753,"power_floor":5.050558953987498e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030734863449852884,"output_power":0.00022507248760283616,"reference_active":true,"delta_db":-1.3530887697711749},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1873727578084203e-05,"output_power":9.675211360958986e-06,"reference_active":true,"delta_db":-0.8892661935153144},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1997540284846557e-06,"output_power":3.358169407385933e-06,"reference_active":true,"delta_db":0.20986006411251523}],"active_band_count":3,"active_band_mae_db":0.8174050091330015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":545,"rms_floor":0.0001327212418155818,"log_rms_mae_db":2.5334523923990266,"rms_envelope_nrmse":0.15172631459298733,"envelope_correlation":0.9970133433583597,"rms_gain_error_db":-1.3227740262778842,"reference_mean_power":0.0013434023559712933,"output_mean_power":0.0009906692958685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":545,"frequency_resolution_hz":0.36697247706422015,"power_floor":4.567271260416197e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002737284977215731,"output_power":0.00020111085640492818,"reference_active":true,"delta_db":-1.338844985229374},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.2058905710883025e-06,"output_power":4.39799457344637e-06,"reference_active":true,"delta_db":-0.7324034574589926},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7945800637954934e-07,"output_power":1.887646067023557e-07,"reference_active":true,"delta_db":0.21957728719467262}],"active_band_count":3,"active_band_mae_db":0.7636085766276798}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":551,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":1.9492406115183314,"rms_envelope_nrmse":0.1597216461135454,"envelope_correlation":0.9906304598274944,"rms_gain_error_db":-1.10516850921783,"reference_mean_power":0.0005860055325719512,"output_mean_power":0.0004543440713086814},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":551,"frequency_resolution_hz":0.3629764065335753,"power_floor":2.476604496378982e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014212077651255318,"output_power":0.00011005194068492182,"reference_active":true,"delta_db":-1.110598662805107},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.080530981693378e-06,"output_power":5.654414411526008e-06,"reference_active":true,"delta_db":-1.550522650166205},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.479689452622023e-06,"output_power":1.7090211122536438e-06,"reference_active":true,"delta_db":0.6257684965376542}],"active_band_count":3,"active_band_mae_db":1.0956299365029887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":545,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":1.72542279094164,"rms_envelope_nrmse":0.14030616521185185,"envelope_correlation":0.9953034796226408,"rms_gain_error_db":-1.1051684135720734,"reference_mean_power":0.000592456879484966,"output_mean_power":0.00045934596461371806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":545,"frequency_resolution_hz":0.36697247706422015,"power_floor":2.2592456805786762e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012603513365392867,"output_power":9.808488023981468e-05,"reference_active":true,"delta_db":-1.0888956011764697},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0421085883295177e-06,"output_power":2.1283980751896724e-06,"reference_active":true,"delta_db":-1.551218546728367},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.97651922149237e-08,"output_power":1.0565948924035292e-07,"reference_active":true,"delta_db":1.2209509089244697}],"active_band_count":3,"active_band_mae_db":1.2870216856097687}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":551,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":2.611404714293812,"rms_envelope_nrmse":0.1878709995508235,"envelope_correlation":0.9864176950482378,"rms_gain_error_db":-1.2055931833918008,"reference_mean_power":0.00033603820576576097,"output_mean_power":0.00025458296421021145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":551,"frequency_resolution_hz":0.3629764065335753,"power_floor":1.7744885268479347e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012531956187397092,"output_power":9.033616966124645e-05,"reference_active":true,"delta_db":-1.4215719576785704},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.665500820088321e-06,"output_power":7.3108035766854875e-06,"reference_active":true,"delta_db":-0.7382855206163442},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.262411892321394e-06,"output_power":1.6804882630470122e-06,"reference_active":true,"delta_db":1.2423440651740676}],"active_band_count":3,"active_band_mae_db":1.1340671811563274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":545,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":2.475142361311584,"rms_envelope_nrmse":0.16156669563515474,"envelope_correlation":0.9934779017540509,"rms_gain_error_db":-1.2055938436280274,"reference_mean_power":0.0003397374566311426,"output_mean_power":0.00025738548215506794},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":545,"frequency_resolution_hz":0.36697247706422015,"power_floor":1.625418522895206e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011395527290013323,"output_power":8.295621391701901e-05,"reference_active":true,"delta_db":-1.3788550267448776},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.135204442017973e-06,"output_power":3.5755558652785235e-06,"reference_active":true,"delta_db":-0.6315341773377409},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.378219758440543e-08,"output_power":1.0974367768428069e-07,"reference_active":true,"delta_db":0.6825910471856291}],"active_band_count":3,"active_band_mae_db":0.8976600837560825}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":551,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":3.859144820100214,"rms_envelope_nrmse":0.3192413933294121,"envelope_correlation":0.9754521872217875,"rms_gain_error_db":-2.7704679146758737,"reference_mean_power":0.00012453856537635692,"output_mean_power":6.580472326529494e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":551,"frequency_resolution_hz":0.3629764065335753,"power_floor":7.91331630679602e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.199929543742056e-05,"output_power":2.1318259307762357e-05,"reference_active":true,"delta_db":-2.9449026440728825},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.029733185397958e-06,"output_power":3.229017416376803e-06,"reference_active":true,"delta_db":-3.3786845413785245},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.488948400815819e-07,"output_power":3.9696769165517477e-07,"reference_active":true,"delta_db":-3.7846292311742378}],"active_band_count":3,"active_band_mae_db":3.3694054722085482}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44160,"uncovered_tail_samples":0,"metrics":{"frame_count":545,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":3.900651671392704,"rms_envelope_nrmse":0.30200205319537243,"envelope_correlation":0.9849723489682282,"rms_gain_error_db":-2.7704662243585765,"reference_mean_power":0.00012590939521049525,"output_mean_power":6.652907963137438e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":545,"frequency_resolution_hz":0.36697247706422015,"power_floor":7.224757260417465e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8289232236082604e-05,"output_power":1.9270190362038673e-05,"reference_active":true,"delta_db":-2.9819065320785616},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6952095076479426e-06,"output_power":1.785601353474952e-06,"reference_active":true,"delta_db":-3.1585456021462495},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.939691517064883e-08,"output_power":2.6694203304560264e-08,"reference_active":true,"delta_db":-3.473469263576895}],"active_band_count":3,"active_band_mae_db":3.204640465933902}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3980910778045654,"predicted_mos":3.3980910778045654,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66480,"input_sample_rate":24000,"input_seconds":2.77,"input_float32_sha256":"6e6ffa3e97371387df5e90d452649330a91a9e9c4fb9a9bb8e3b30a9ef6ff884","internal_samples":44320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.866514941000787},"started_at":"2026-10-13T07:40:16.367815+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028609469991351943,"native_input_samples":66360,"native_input_sha256":"83deac5ff8c4cfaf3dbef9f54ef132dcf4b93465fc942416de066077e40172b7","output_sha256":"0fe62d1ebb5da237d8b4c873a767d7d969f74e4a37f9eb2208d138fafda2dd3a","output_samples":66480,"length_delta_samples":120,"analysis_s":0.9954278990007879,"synthesis_s":7.009972871997888,"total_s":8.005400770998676,"rtf":2.8952624849904796,"output_peak":0.36883923411369324,"output_rms":0.03500774149928862,"finished_at":"2026-10-13T07:40:24.388655+00:00","analysis_stages":{"external_f0_s":0.6142634120005823},"evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"0fe62d1ebb5da237d8b4c873a767d7d969f74e4a37f9eb2208d138fafda2dd3a","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44240,"output_samples":44320,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.109975860477594,"output_peak":0.3658638538459451},"common_prefix":{"samples":44240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981949458483754,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.553826946515669,"spectral_convergence":0.33532706011015706,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.5518087927939535,"spectral_convergence":0.27410123335038056,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.697625116934574,"spectral_convergence":0.2999098557354686,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":6.462184356464299e-05,"delta_db":0.12224997903952574,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.0001936913656973707,"delta_db":-0.008341643355686193,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0005205999187781782,"delta_db":-0.29184039677113,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.0002912466624439188,"delta_db":-0.22137466591539184,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00012284401428836863,"delta_db":-0.9859740126930128,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":3.666185240370165e-05,"delta_db":-1.3180792562337642,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.49130999233475176,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3209098644573127,"pitch":null,"output_rms_dbfs":-29.102129560255186,"output_peak":0.3658638538459451,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.601086952081399},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44240,"output_samples":44240,"intelligibility":{"estoi":{"value":0.9431165230200383,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03638639377946353,"rms_dbfs":-28.781219695797876,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.243327098875167,"mean":1.0381771039359177e-06,"zero_energy":false},"output":{"rms":0.03506658890134855,"rms_dbfs":-29.102129560255186,"peak":0.3658638538459451,"peak_dbfs":-8.733609905413072,"crest_factor_db":20.368519654842117,"mean":-2.1270741130947213e-05,"zero_energy":false},"rms_difference_db":-0.3209098644573096},"full_scale":{"reference":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":1.7921673197168487e-07,"output_rms_dbfs":-67.4662144632706,"output_zero_energy":false,"output_peak":0.007841096426925567,"output_energy_fraction":4.164122195697146e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":132,"both_estimated_voiced_frames":124,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9185185185185185,"absolute_difference_cents_median":4.72158007610448,"absolute_difference_cents_p95":31.861288310938804,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":0.00013943874660832744,"log_rms_mae_db":2.1628777344361376,"rms_envelope_nrmse":0.11590255168261686,"envelope_correlation":0.9891038567821987,"rms_gain_error_db":-0.3209097376346569,"reference_mean_power":0.0013263679977567136,"output_mean_power":0.0012318932087646832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":5.047819293813421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030810693936871264,"output_power":0.00028133285941924185,"reference_active":true,"delta_db":-0.3948101954325917},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2023354078518551e-05,"output_power":1.2642340147773053e-05,"reference_active":true,"delta_db":0.21801834100439596},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2193973896229425e-06,"output_power":2.6981809108013277e-06,"reference_active":true,"delta_db":-0.7670352218139947}],"active_band_count":3,"active_band_mae_db":0.45995458608366074}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":0.0001327212418155818,"log_rms_mae_db":1.6792139415024305,"rms_envelope_nrmse":0.08749900907183802,"envelope_correlation":0.994017171805588,"rms_gain_error_db":-0.32090843142316683,"reference_mean_power":0.0013409420747202037,"output_mean_power":0.001245429575218695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":4.5645799716966823e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002745054731183271,"output_power":0.00024936066898157826,"reference_active":true,"delta_db":-0.41723053490469225},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.376005868866601e-06,"output_power":5.858787875008168e-06,"reference_active":true,"delta_db":0.37348039943103417},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7970463694996414e-07,"output_power":1.7425741268145064e-07,"reference_active":true,"delta_db":-0.13368021949697795}],"active_band_count":3,"active_band_mae_db":0.3081303846109014}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":1.939714449066725,"rms_envelope_nrmse":0.12188210005227679,"envelope_correlation":0.989152684074586,"rms_gain_error_db":-0.21905097113233635,"reference_mean_power":0.0005849439303269558,"output_mean_power":0.0005561720267190397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":2.471278629444775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014231281864882344,"output_power":0.00013619613040915043,"reference_active":true,"delta_db":-0.1907925175683438},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.153370236673123e-06,"output_power":8.3447461674723e-06,"reference_active":true,"delta_db":0.10075967075468278},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4861340139926787e-06,"output_power":1.3282861440166382e-06,"reference_active":true,"delta_db":-0.4876633190128793}],"active_band_count":3,"active_band_mae_db":0.259738502445302}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":1.5290798190177568,"rms_envelope_nrmse":0.0886270712313253,"envelope_correlation":0.9942400513722846,"rms_gain_error_db":-0.21905063323799218,"reference_mean_power":0.0005913717973798096,"output_mean_power":0.0005622837670939467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":2.253861652661314e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012623197394568146,"output_power":0.00012038188746741683,"reference_active":true,"delta_db":-0.20608225179843587},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1401147597866345e-06,"output_power":3.2823303955307603e-06,"reference_active":true,"delta_db":0.1923677424216006},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.001702937868675e-08,"output_power":9.415556800303907e-08,"reference_active":true,"delta_db":0.70663583421134}],"active_band_count":3,"active_band_mae_db":0.3683619428104588}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":2.0015388629482858,"rms_envelope_nrmse":0.1439372395172563,"envelope_correlation":0.9853589588044489,"rms_gain_error_db":-0.26337156613266866,"reference_mean_power":0.0003354294563993913,"output_mean_power":0.00031569233819426194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":1.775200853272881e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012560860558913709,"output_power":0.0001168107116013534,"reference_active":true,"delta_db":-0.3153672480561128},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.72605183156066e-06,"output_power":9.15335009785582e-06,"reference_active":true,"delta_db":0.20762285291224455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2668523428769504e-06,"output_power":1.3754182025566662e-06,"reference_active":true,"delta_db":0.3570876865134965}],"active_band_count":3,"active_band_mae_db":0.29335926249395133}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":1.404511218905661,"rms_envelope_nrmse":0.10375956566702073,"envelope_correlation":0.9923441013684303,"rms_gain_error_db":-0.2633723000136416,"reference_mean_power":0.0003391152422568523,"output_mean_power":0.0003191611935456354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":1.6261600138727939e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011427350869395267,"output_power":0.00010633429356173125,"reference_active":true,"delta_db":-0.3127221215985005},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1799749558208305e-06,"output_power":4.522637193651341e-06,"reference_active":true,"delta_db":0.3421807028009942},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395722090626045e-08,"output_power":9.805911524321198e-08,"reference_active":true,"delta_db":0.185578078160113}],"active_band_count":3,"active_band_mae_db":0.2801603008532026}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":2.0812616957825156,"rms_envelope_nrmse":0.225870838294187,"envelope_correlation":0.9751829170924277,"rms_gain_error_db":-1.3555012797880017,"reference_mean_power":0.0001243129634607548,"output_mean_power":9.098426511673198e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":7.893026272799498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.194673883780046e-05,"output_power":2.8140193566307502e-05,"reference_active":true,"delta_db":-1.7337112165307564},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.051190096168371e-06,"output_power":4.7558112337150295e-06,"reference_active":true,"delta_db":-1.710378147751822},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.501810525085992e-07,"output_power":5.787524437503456e-07,"reference_active":true,"delta_db":-2.153135281203588}],"active_band_count":3,"active_band_mae_db":1.8657415484953888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":1.6249869490724165,"rms_envelope_nrmse":0.192213743019372,"envelope_correlation":0.9861409142615201,"rms_gain_error_db":-1.3555060432990467,"reference_mean_power":0.00012567880182247303,"output_mean_power":9.198381700601215e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":7.206643574889897e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8301143786829975e-05,"output_power":2.5587674699644716e-05,"reference_active":true,"delta_db":-1.7518092267302596},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6851867496780295e-06,"output_power":2.4220959714882048e-06,"reference_active":true,"delta_db":-1.8226815362223625},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.995617022655136e-08,"output_power":4.7018643785769815e-08,"reference_active":true,"delta_db":-1.0556378587152513}],"active_band_count":3,"active_band_mae_db":1.5433762072226245}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.822798013687134,"predicted_mos":3.822798013687134,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66600,"input_sample_rate":24000,"input_seconds":2.775,"input_float32_sha256":"99e33135a4e83335b882c900e6ab32ed23f9d531af223ba59b23553ff0c64eb8","internal_samples":44400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.757966872999532},"started_at":"2026-10-13T05:10:56.328224+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0016460349997942103,"native_input_samples":66360,"native_input_sha256":"83deac5ff8c4cfaf3dbef9f54ef132dcf4b93465fc942416de066077e40172b7","output_sha256":"0c6d5453fd71f274ac144b4ed678ca26a358bb7a7657b5fb5c2f41e7c9e1b13c","output_samples":66600,"length_delta_samples":240,"analysis_s":0.38504386999920825,"synthesis_s":0.07671296900116431,"total_s":0.46175683900037257,"rtf":0.16700066509959224,"output_peak":0.4208715558052063,"output_rms":0.03564546605585084,"finished_at":"2026-10-13T05:10:56.797512+00:00","analysis_stages":{"external_f0_s":0.05931896299989603},"evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"0c6d5453fd71f274ac144b4ed678ca26a358bb7a7657b5fb5c2f41e7c9e1b13c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44240,"output_samples":44400,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.953078781052927,"output_peak":0.41353138918741295},"common_prefix":{"samples":44240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9963963963963964,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.276218385592968,"spectral_convergence":0.11043169387908529,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.32140512131178,"spectral_convergence":0.1069518496139612,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.792467329719148,"spectral_convergence":0.11272000563552517,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":6.239469487467101e-05,"delta_db":-0.030066804773794302,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00019122296186763378,"delta_db":-0.06404384934941289,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0005477743442266303,"delta_db":-0.07086448085420972,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.0002956312770005382,"delta_db":-0.15648044482966855,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00013641399585318103,"delta_db":-0.5309247060689979,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":4.37658497959996e-05,"delta_db":-0.5488696475946839,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2335416555784612,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15618058410081997,"pitch":null,"output_rms_dbfs":-28.937400279898696,"output_peak":0.41353138918741295,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.463363612207965},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44240,"output_samples":44240,"intelligibility":{"estoi":{"value":0.9839151812645032,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03638639377946353,"rms_dbfs":-28.781219695797876,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.243327098875167,"mean":1.0381771039359177e-06,"zero_energy":false},"output":{"rms":0.03573797872877891,"rms_dbfs":-28.937400279898693,"peak":0.41353138918741295,"peak_dbfs":-7.669830392918115,"crest_factor_db":21.26756988698058,"mean":-1.5704374502570645e-05,"zero_energy":false},"rms_difference_db":-0.1561805841008166},"full_scale":{"reference":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":1.7759653584248727e-07,"output_rms_dbfs":-67.50565509723887,"output_zero_energy":false,"output_peak":0.002971932787410745,"output_energy_fraction":3.972889389731735e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":133,"both_estimated_voiced_frames":129,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9555555555555556,"absolute_difference_cents_median":0.4271664489792978,"absolute_difference_cents_p95":5.439000326036921,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":0.00013943874660832744,"log_rms_mae_db":0.7986308845512414,"rms_envelope_nrmse":0.033025853729325524,"envelope_correlation":0.9993192402340542,"rms_gain_error_db":-0.1561804781163021,"reference_mean_power":0.0013263679977567136,"output_mean_power":0.0012795167836618833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":5.047819293813421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030810693936871264,"output_power":0.00029172128851447744,"reference_active":true,"delta_db":-0.23733356720639434},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2023354078518551e-05,"output_power":1.0933463616544295e-05,"reference_active":true,"delta_db":-0.41267872856080146},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2193973896229425e-06,"output_power":2.7549331572169365e-06,"reference_active":true,"delta_db":-0.6766352152594801}],"active_band_count":3,"active_band_mae_db":0.442215837008892}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":0.0001327212418155818,"log_rms_mae_db":0.49187602712698314,"rms_envelope_nrmse":0.022545651725492187,"envelope_correlation":0.9998233890164626,"rms_gain_error_db":-0.15617879948131744,"reference_mean_power":0.0013409420747202037,"output_mean_power":0.0012935765614064875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":4.5645799716966823e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002745054731183271,"output_power":0.00026002264148046185,"reference_active":true,"delta_db":-0.2353984205612998},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.376005868866601e-06,"output_power":5.114893025634598e-06,"reference_active":true,"delta_db":-0.21623178974073576},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7970463694996414e-07,"output_power":1.6108656508492787e-07,"reference_active":true,"delta_db":-0.47499962456518274}],"active_band_count":3,"active_band_mae_db":0.3088766116224061}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":0.648101808982306,"rms_envelope_nrmse":0.02640588219220717,"envelope_correlation":0.9995163798325394,"rms_gain_error_db":-0.060333716473845896,"reference_mean_power":0.0005849439303269558,"output_mean_power":0.0005768738694244934},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":2.471278629444775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014231281864882344,"output_power":0.00013929042008121862,"reference_active":true,"delta_db":-0.09322772215757377},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.153370236673123e-06,"output_power":7.349971849599223e-06,"reference_active":true,"delta_db":-0.4505148792604933},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4861340139926787e-06,"output_power":1.1172048190215556e-06,"reference_active":true,"delta_db":-1.2392517389216255}],"active_band_count":3,"active_band_mae_db":0.5943314467798976}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":0.35581034155816993,"rms_envelope_nrmse":0.013363259300411488,"envelope_correlation":0.9998978923389695,"rms_gain_error_db":-0.06033352267213296,"reference_mean_power":0.0005913717973798096,"output_mean_power":0.0005832130817321808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":2.253861652661314e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012623197394568146,"output_power":0.00012373898789416507,"reference_active":true,"delta_db":-0.08662814077247796},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1401147597866345e-06,"output_power":2.929895145384595e-06,"reference_active":true,"delta_db":-0.30093442077528854},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.001702937868675e-08,"output_power":6.253557957488843e-08,"reference_active":true,"delta_db":-1.0705524501067205}],"active_band_count":3,"active_band_mae_db":0.4860383372181623}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":0.8617871135159526,"rms_envelope_nrmse":0.047797596840011704,"envelope_correlation":0.9987461382143134,"rms_gain_error_db":-0.21024481564542313,"reference_mean_power":0.0003354294563993913,"output_mean_power":0.0003195778839617847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":1.775200853272881e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012560860558913709,"output_power":0.00011611382826646098,"reference_active":true,"delta_db":-0.34135450471037604},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.72605183156066e-06,"output_power":8.53769766131418e-06,"reference_active":true,"delta_db":-0.09477016892879885},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2668523428769504e-06,"output_power":1.1685581802419455e-06,"reference_active":true,"delta_db":-0.35075658986595787}],"active_band_count":3,"active_band_mae_db":0.26229375450171094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":0.611201714148562,"rms_envelope_nrmse":0.03333563896481109,"envelope_correlation":0.9995949939487376,"rms_gain_error_db":-0.2102441535547161,"reference_mean_power":0.0003391152422568523,"output_mean_power":0.00032308953788599045},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":1.6261600138727939e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011427350869395267,"output_power":0.0001057352366788132,"reference_active":true,"delta_db":-0.33725820561896414},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1799749558208305e-06,"output_power":4.146714368391374e-06,"reference_active":true,"delta_db":-0.03469558152045149},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395722090626045e-08,"output_power":8.552800271767055e-08,"reference_active":true,"delta_db":-0.4082183227005379}],"active_band_count":3,"active_band_mae_db":0.26005736994665113}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":1.2341009287101168,"rms_envelope_nrmse":0.10928514079663039,"envelope_correlation":0.9945035237240599,"rms_gain_error_db":-0.61295978786034,"reference_mean_power":0.0001243129634607548,"output_mean_power":0.00010794945170437705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":7.893026272799498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.194673883780046e-05,"output_power":3.601218811007041e-05,"reference_active":true,"delta_db":-0.6624869209766611},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.051190096168371e-06,"output_power":5.468237625109182e-06,"reference_active":true,"delta_db":-1.1041504437449745},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.501810525085992e-07,"output_power":9.426592323781649e-07,"reference_active":true,"delta_db":-0.03451640582496127}],"active_band_count":3,"active_band_mae_db":0.6003845901821989}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":1.0260039334842843,"rms_envelope_nrmse":0.08052257191946381,"envelope_correlation":0.9985540150754337,"rms_gain_error_db":-0.6129583097176108,"reference_mean_power":0.00012567880182247303,"output_mean_power":0.000109135539749754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":7.206643574889897e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8301143786829975e-05,"output_power":3.289154243201209e-05,"reference_active":true,"delta_db":-0.6612750351864476},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6851867496780295e-06,"output_power":2.8069558534612135e-06,"reference_active":true,"delta_db":-1.18223918660298},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.995617022655136e-08,"output_power":7.347529548028105e-08,"reference_active":true,"delta_db":0.883074572869116}],"active_band_count":3,"active_band_mae_db":0.9088629315528478}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0030","method":"magphase","status":"measured","mos":{"status":"measured","value":3.693862199783325,"predicted_mos":3.693862199783325,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":132779,"input_sample_rate":48000,"input_seconds":2.7662291666666667,"input_float32_sha256":"84a078e08fb08970b1e0e388b4d769254924b7d09201d7a5c9fec4dcbd8cf948","internal_samples":44260,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2453006069990806},"started_at":"2026-10-13T05:16:25.894226+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.004554156999802217,"native_input_samples":132720,"native_input_sha256":"054235eb35b3a5c550534b31b82aaf9d8458e75990d58dfa2a308bad7efaec22","output_sha256":"ab9cd119b8f286bd21501dc4dc3188847041de96c0ab1757098b6774745ee88d","output_samples":132779,"length_delta_samples":59,"analysis_s":1.658874435999678,"synthesis_s":0.6663751779997256,"total_s":2.3252496139994037,"rtf":0.840958269077542,"output_peak":0.4601366975784145,"output_rms":0.03603522137894875,"finished_at":"2026-10-13T05:16:28.230368+00:00","evaluation_input":{"reference":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"ab9cd119b8f286bd21501dc4dc3188847041de96c0ab1757098b6774745ee88d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":44240,"output_samples":44260,"length_delta_samples":20,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.858698764357264,"output_peak":0.4449577114127075},"common_prefix":{"samples":44240,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9995481247175779,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44240,"output_samples":44240,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.241177413842553,"spectral_convergence":0.27234040399335063,"spectral_convergence_status":"ok","amplitude_floor":7.692579600611731e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.818990925752856,"spectral_convergence":0.13222249119292348,"spectral_convergence_status":"ok","amplitude_floor":5.999960941709396e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.325979680782127,"spectral_convergence":0.1553799041079777,"spectral_convergence_status":"ok","amplitude_floor":5.417405012540354e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":6.282816067608343e-05,"output_power":5.8160779039269284e-05,"delta_db":-0.33524130644497546,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019406375287345015,"output_power":0.00018411362293712267,"delta_db":-0.2285850165133199,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005567857807558596,"output_power":0.0005370752320901646,"delta_db":-0.15653010900585757,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030647737825017173,"output_power":0.0003045897565957218,"delta_db":-0.026831299844159263,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00015415279416012865,"output_power":0.00015843637214445344,"delta_db":0.11903488309413764,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.966178555848869e-05,"output_power":5.8771387316453314e-05,"delta_db":0.731436122024226,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.26627645615444595,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07551615585990695,"pitch":null,"output_rms_dbfs":-28.85673585165778,"output_peak":0.4449577114127075,"reference_rms_dbfs":-28.781219695797876,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.128716006792512},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44240,"output_samples":44240,"intelligibility":{"estoi":{"value":0.989528069227089,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03638639377946353,"rms_dbfs":-28.781219695797876,"peak":0.41986083984375,"peak_dbfs":-7.537892596922708,"crest_factor_db":21.243327098875167,"mean":1.0381771039359177e-06,"zero_energy":false},"output":{"rms":0.03607141735672699,"rms_dbfs":-28.85673585165778,"peak":0.4449577114127075,"peak_dbfs":-7.033625244328111,"crest_factor_db":21.82311060732967,"mean":-0.0004348607762376749,"zero_energy":false},"rms_difference_db":-0.0755161558599049},"full_scale":{"reference":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44240,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":276,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.9443164055701348e-06,"threshold_dbfs":-57.112330593625856,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.2096954333254054e-07,"output_mean_power":1.6003762734056285e-07,"output_rms_dbfs":-67.9577789593675,"output_zero_energy":false,"output_peak":0.0036995840360008856,"output_energy_fraction":3.5142094711339117e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":269,"reference_estimated_voiced_frames":135,"output_estimated_voiced_frames":135,"both_estimated_voiced_frames":133,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9851851851851852,"absolute_difference_cents_median":1.9221477825920164,"absolute_difference_cents_p95":5.426878891913546,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44240,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":0.00013943874660832744,"log_rms_mae_db":0.9721105326051318,"rms_envelope_nrmse":0.041228334984846716,"envelope_correlation":0.9985955978470475,"rms_gain_error_db":-0.07551582814918609,"reference_mean_power":0.0013263679977567136,"output_mean_power":0.0013035042514434282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":5.047819293813421e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030810693936871264,"output_power":0.00030886927593182256,"reference_active":true,"delta_db":0.010732301278020343},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2023354078518551e-05,"output_power":1.2820921339685596e-05,"reference_active":true,"delta_db":0.2789359872929839},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2193973896229425e-06,"output_power":2.898728920610218e-06,"reference_active":true,"delta_db":-0.4556698408674721}],"active_band_count":3,"active_band_mae_db":0.24844604314615878}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":0.0001327212418155818,"log_rms_mae_db":0.658735080313826,"rms_envelope_nrmse":0.021601597973198335,"envelope_correlation":0.999647100505935,"rms_gain_error_db":-0.07551508120722009,"reference_mean_power":0.0013409420747202037,"output_mean_power":0.0013178273291563087},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":4.5645799716966823e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002745054731183271,"output_power":0.00027558915254423996,"reference_active":true,"delta_db":0.017111114404477563},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.376005868866601e-06,"output_power":5.48030997457541e-06,"reference_active":true,"delta_db":0.08345389372537543},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7970463694996414e-07,"output_power":1.9689362694247507e-07,"reference_active":true,"delta_db":0.3967237568214909}],"active_band_count":3,"active_band_mae_db":0.16576292165044795}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":9.713079411954574e-05,"log_rms_mae_db":0.7672605009894689,"rms_envelope_nrmse":0.036160056033087196,"envelope_correlation":0.999248838944266,"rms_gain_error_db":-0.14968404258194073,"reference_mean_power":0.0005849439303269558,"output_mean_power":0.0005651267105895282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":2.471278629444775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014231281864882344,"output_power":0.00013713728992515987,"reference_active":true,"delta_db":-0.16088457588615496},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.153370236673123e-06,"output_power":7.87872686404995e-06,"reference_active":true,"delta_db":-0.14881118839040464},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4861340139926787e-06,"output_power":1.3571709625111916e-06,"reference_active":true,"delta_db":-0.39423415157410824}],"active_band_count":3,"active_band_mae_db":0.23464330528355593}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":9.335285483371196e-05,"log_rms_mae_db":0.5062413802843617,"rms_envelope_nrmse":0.024298213543278643,"envelope_correlation":0.999782474887797,"rms_gain_error_db":-0.14968431902633675,"reference_mean_power":0.0005913717973798096,"output_mean_power":0.0005713367726182057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":2.253861652661314e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012623197394568146,"output_power":0.00012153119754522493,"reference_active":true,"delta_db":-0.16481596145198968},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1401147597866345e-06,"output_power":3.116001433125823e-06,"reference_active":true,"delta_db":-0.03347871505497439},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.001702937868675e-08,"output_power":8.63743597942563e-08,"reference_active":true,"delta_db":0.33202417189780425}],"active_band_count":3,"active_band_mae_db":0.17677294946825609}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":7.293407135281029e-05,"log_rms_mae_db":0.6547465031567428,"rms_envelope_nrmse":0.040521016734928625,"envelope_correlation":0.9988229956299594,"rms_gain_error_db":0.013145648844497492,"reference_mean_power":0.0003354294563993913,"output_mean_power":0.00033644630515675393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":1.775200853272881e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012560860558913709,"output_power":0.00012834439374705473,"reference_active":true,"delta_db":0.09357508432338096},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.72605183156066e-06,"output_power":9.483473057123186e-06,"reference_active":true,"delta_db":0.3614962638373351},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2668523428769504e-06,"output_power":1.2279351242634737e-06,"reference_active":true,"delta_db":-0.1355057668582202}],"active_band_count":3,"active_band_mae_db":0.19685903833964544}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":7.108743338455808e-05,"log_rms_mae_db":0.4563013052947755,"rms_envelope_nrmse":0.027409391342510258,"envelope_correlation":0.9994353292578405,"rms_gain_error_db":0.013144884933443046,"reference_mean_power":0.0003391152422568523,"output_mean_power":0.0003401432045818897},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":1.6261600138727939e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011427350869395267,"output_power":0.00011715226822559456,"reference_active":true,"delta_db":0.10805139224433158},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1799749558208305e-06,"output_power":4.39443525061136e-06,"reference_active":true,"delta_db":0.21729390062532508},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.395722090626045e-08,"output_power":9.73715073045664e-08,"reference_active":true,"delta_db":0.1550173030024266}],"active_band_count":3,"active_band_mae_db":0.16012086529069441}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":552,"rms_floor":4.804880216814636e-05,"log_rms_mae_db":1.0157422667173452,"rms_envelope_nrmse":0.08306896705066552,"envelope_correlation":0.9962680378636857,"rms_gain_error_db":0.31503655018700455,"reference_mean_power":0.0001243129634607548,"output_mean_power":0.0001336657300115764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":552,"frequency_resolution_hz":0.3623188405797101,"power_floor":7.893026272799498e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.194673883780046e-05,"output_power":4.5797640944264936e-05,"reference_active":true,"delta_db":0.3814490577051371},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.051190096168371e-06,"output_power":6.988325945035965e-06,"reference_active":true,"delta_db":-0.03889270303694269},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.501810525085992e-07,"output_power":1.049185924627531e-06,"reference_active":true,"delta_db":0.4304608973724454}],"active_band_count":3,"active_band_mae_db":0.28360088603817507}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44240,"uncovered_tail_samples":0,"metrics":{"frame_count":546,"rms_floor":4.5322864622321904e-05,"log_rms_mae_db":0.7212999276021387,"rms_envelope_nrmse":0.05904261555710839,"envelope_correlation":0.9985201272563508,"rms_gain_error_db":0.31503546496227686,"reference_mean_power":0.00012567880182247303,"output_mean_power":0.0001351342943419271},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":546,"frequency_resolution_hz":0.3663003663003663,"power_floor":7.206643574889897e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8301143786829975e-05,"output_power":4.142890118060744e-05,"reference_active":true,"delta_db":0.3409167115590841},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6851867496780295e-06,"output_power":3.5707681027613546e-06,"reference_active":true,"delta_db":-0.13697854341045754},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.995617022655136e-08,"output_power":6.644741706167622e-08,"reference_active":true,"delta_db":0.44644219775983807}],"active_band_count":3,"active_band_mae_db":0.3081124842431266}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"hnm","status":"measured","mos":{"status":"measured","value":2.176515817642212,"predicted_mos":2.176515817642212,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83120,"input_sample_rate":16000,"input_seconds":5.195,"input_float32_sha256":"5554b08eb14408312986351b794cc61fb574c052f78f7e9a6c48b6faa23de822","internal_samples":83120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7710656929994},"started_at":"2026-10-13T04:39:35.431009+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2448000234144274e-05,"native_input_samples":83120,"native_input_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","output_sha256":"3198f95a050e90cb6c03bac2f8984b388853e23b2f634b331514bd66c5dfb92a","output_samples":83120,"length_delta_samples":0,"analysis_s":0.9193005050001375,"synthesis_s":0.42808766100006324,"total_s":1.3473881660002007,"rtf":0.25936249586144383,"output_peak":0.5754588164586549,"output_rms":0.03233615508222961,"finished_at":"2026-10-13T04:39:36.783326+00:00","evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"3198f95a050e90cb6c03bac2f8984b388853e23b2f634b331514bd66c5dfb92a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.568832983519935,"spectral_convergence":0.3062696026948447,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.558976272321913,"spectral_convergence":0.322119845585967,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.203486763855348,"spectral_convergence":0.33799103404207864,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":3.2061615513136983e-05,"delta_db":0.08085116401586107,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00013085912099077994,"delta_db":0.11107313250995302,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0005190673227396078,"delta_db":-0.1082366635955777,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.00021513504049231862,"delta_db":-0.2070213199012872,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":8.789767224673883e-05,"delta_db":0.28626469005056426,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":6.060615351942116e-05,"delta_db":-0.2239539457818452,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16956681930918138,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07129544638742717,"pitch":null,"output_rms_dbfs":-29.80623241947804,"output_peak":0.5754588164586549,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.777098673232398},"common_prefix":{"samples":83120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.568832983519935,"spectral_convergence":0.3062696026948447,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.558976272321913,"spectral_convergence":0.322119845585967,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.203486763855348,"spectral_convergence":0.33799103404207864,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":3.2061615513136983e-05,"delta_db":0.08085116401586107,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00013085912099077994,"delta_db":0.11107313250995302,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0005190673227396078,"delta_db":-0.1082366635955777,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.00021513504049231862,"delta_db":-0.2070213199012872,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":8.789767224673883e-05,"delta_db":0.28626469005056426,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":6.060615351942116e-05,"delta_db":-0.2239539457818452,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16956681930918138,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07129544638742717,"pitch":null,"output_rms_dbfs":-29.80623241947804,"output_peak":0.5754588164586549,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.777098673232398},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83120,"output_samples":83120,"intelligibility":{"estoi":{"value":0.9113322554030279,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03260266874006855,"rms_dbfs":-29.734936973090612,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.776881716111717,"mean":-5.646779975487248e-07,"zero_energy":false},"output":{"rms":0.03233615508222961,"rms_dbfs":-29.806232419478036,"peak":0.5754588164586549,"peak_dbfs":-4.799715036678476,"crest_factor_db":25.00651738279956,"mean":-1.6134247586132028e-05,"zero_energy":false},"rms_difference_db":-0.07129544638742402},"full_scale":{"reference":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":7.246880351399329e-07,"output_rms_dbfs":-61.398489089806034,"output_zero_energy":false,"output_peak":0.023318065475582458,"output_energy_fraction":0.00017610136094045226,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":164,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.8502673796791443,"absolute_difference_cents_median":1.0080380598865817,"absolute_difference_cents_p95":8.315587203566357,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.00015201529585382603,"log_rms_mae_db":1.0126018135482042,"rms_envelope_nrmse":0.07962617556111848,"envelope_correlation":0.9948598013677236,"rms_gain_error_db":-0.07129557917271678,"reference_mean_power":0.0010639579210132953,"output_mean_power":0.0010466341338265137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.575807359242935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021400202307720183,"output_power":0.00021356296889923982,"reference_active":true,"delta_db":-0.008919293082692474},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3329698566280647e-05,"output_power":2.2740746173549875e-05,"reference_active":true,"delta_db":-0.11104416748420033},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3580980563714554e-06,"output_power":6.019197715948955e-06,"reference_active":true,"delta_db":0.50527951881034}],"active_band_count":3,"active_band_mae_db":0.20841432645907756}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.0001490763524181648,"log_rms_mae_db":0.5471600185853948,"rms_envelope_nrmse":0.041854682290620206,"envelope_correlation":0.9985502267437558,"rms_gain_error_db":-0.07129610121167715,"reference_mean_power":0.0010701430709341084,"output_mean_power":0.0010527184481287336},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":5.309268989226658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001912637761084412,"output_power":0.00019190058900119816,"reference_active":true,"delta_db":0.014435819463091637},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0962051121960504e-05,"output_power":1.0961808692103663e-05,"reference_active":true,"delta_db":-9.60469095969796e-05},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1525134859671347e-07,"output_power":3.639580188859342e-07,"reference_active":true,"delta_db":0.6239433929778982}],"active_band_count":3,"active_band_mae_db":0.21282508645019563}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.0001087945365454221,"log_rms_mae_db":1.41038035390079,"rms_envelope_nrmse":0.09935698772214407,"envelope_correlation":0.993009093133224,"rms_gain_error_db":-0.1021943507533858,"reference_mean_power":0.0005739909463294839,"output_mean_power":0.0005606419712578588},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":3.2669273354505384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013940368795916017,"output_power":0.00013740590852655174,"reference_active":true,"delta_db":-0.06268855265384671},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.236537072685246e-06,"output_power":8.22655463126423e-06,"reference_active":true,"delta_db":-0.5029119156818571},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7105571121059603e-06,"output_power":2.7774938184910067e-06,"reference_active":true,"delta_db":2.1051552206719766}],"active_band_count":3,"active_band_mae_db":0.8902518963358935}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.00010533749157464783,"log_rms_mae_db":0.9329757849304102,"rms_envelope_nrmse":0.04953361620005536,"envelope_correlation":0.9982825165872147,"rms_gain_error_db":-0.10219435399917444,"reference_mean_power":0.0005773280667275119,"output_mean_power":0.0005639014817576928},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":3.141212974645643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001274169753045263,"output_power":0.00012595806370991483,"reference_active":true,"delta_db":-0.050013155181491326},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860144772024442e-06,"output_power":3.393148910908704e-06,"reference_active":true,"delta_db":-0.5600067348191788},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1444270061935377e-07,"output_power":1.6035237539695105e-07,"reference_active":true,"delta_db":1.4648730004231694}],"active_band_count":3,"active_band_mae_db":0.6916309634746133}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":1.3036272553422408,"rms_envelope_nrmse":0.10788020054700388,"envelope_correlation":0.9917780031723368,"rms_gain_error_db":-0.13353077784476708,"reference_mean_power":0.00021157950445783094,"output_mean_power":0.00020517314606449685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.2662436125548275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5392735637289224e-05,"output_power":5.240434690730985e-05,"reference_active":true,"delta_db":-0.24085500905424878},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790041823950811e-06,"output_power":5.47080809223573e-06,"reference_active":true,"delta_db":-0.2463022018984581},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4111913003034656e-06,"output_power":1.2937954568379725e-06,"reference_active":true,"delta_db":-0.3772026867985875}],"active_band_count":3,"active_band_mae_db":0.28811996591709815}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":0.78801559021882,"rms_envelope_nrmse":0.053413470323377066,"envelope_correlation":0.9980529162041162,"rms_gain_error_db":-0.13353102628788113,"reference_mean_power":0.00021280955774068177,"output_mean_power":0.00020636594309243002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":1.2078803728484275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022425345543556e-05,"output_power":4.753905330701856e-05,"reference_active":true,"delta_db":-0.23862961082249934},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7667526770926583e-06,"output_power":2.6443504411232295e-06,"reference_active":true,"delta_db":-0.19651329568535658},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.98389498284838e-08,"output_power":8.460034705872469e-08,"reference_active":true,"delta_db":-0.7192785950544549}],"active_band_count":3,"active_band_mae_db":0.38480716718743696}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":1.5426902319834372,"rms_envelope_nrmse":0.23167832070318797,"envelope_correlation":0.9656283569797595,"rms_gain_error_db":-0.08672916097981627,"reference_mean_power":0.00010955764718773973,"output_mean_power":0.00010739146844708728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":6.937978158746453e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2971450195505062e-05,"output_power":2.1113625894412795e-05,"reference_active":true,"delta_db":-0.36625591057695817},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1375769637538911e-05,"output_power":1.1562244254807846e-05,"reference_active":true,"delta_db":0.07061350804867689},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447046226198746e-06,"output_power":3.3496644640772567e-06,"reference_active":true,"delta_db":1.3635913241551605}],"active_band_count":3,"active_band_mae_db":0.6001535809269318}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":0.9889864773728195,"rms_envelope_nrmse":0.15869279877016526,"envelope_correlation":0.9829735169281411,"rms_gain_error_db":-0.08672938806664052,"reference_mean_power":0.00011019456862153084,"output_mean_power":0.00010801579099356518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":6.599778128222779e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3073624841125744e-05,"output_power":2.233906265205282e-05,"reference_active":true,"delta_db":-0.14050880918714653},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.235055585786142e-06,"output_power":6.344795220615136e-06,"reference_active":true,"delta_db":0.0757728014028649},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9515593840599056e-07,"output_power":2.5815258877227825e-07,"reference_active":true,"delta_db":1.2149471356227912}],"active_band_count":3,"active_band_mae_db":0.47707624873760085}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9643986225128174,"predicted_mos":2.9643986225128174,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83120,"input_sample_rate":16000,"input_seconds":5.195,"input_float32_sha256":"490a82808df035ea7be7a27360ad076c4d8ff3de0fb74dfee88926a4a6374830","internal_samples":83120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6711681979995774},"started_at":"2026-10-13T04:42:34.611261+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.037800004472956e-05,"native_input_samples":83120,"native_input_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","output_sha256":"5bc9f7fb763c90f53be182487710799f9204b930e9dece4094cca9965aaf49be","output_samples":83120,"length_delta_samples":0,"analysis_s":2.9465482280002107,"synthesis_s":0.08523366000008537,"total_s":3.031781888000296,"rtf":0.583596128585235,"output_peak":0.553898957590306,"output_rms":0.037102656616982324,"finished_at":"2026-10-13T04:42:37.649017+00:00","evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"5bc9f7fb763c90f53be182487710799f9204b930e9dece4094cca9965aaf49be","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.246557050986432,"spectral_convergence":0.5642032871467357,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.279429757494152,"spectral_convergence":0.3347041705010309,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.764186334795465,"spectral_convergence":0.37971674596672395,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":3.179162881744473e-05,"delta_db":0.044124939023837684,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00013649405808591368,"delta_db":0.2941706043617269,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0007422638230006736,"delta_db":1.4451093849405225,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.0002927078176851584,"delta_db":1.130210371873446,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":9.948064314264101e-05,"delta_db":0.8238767932024228,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":7.386915730586974e-05,"delta_db":0.6355102848066007,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7288337297014261,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1230371129869638,"pitch":null,"output_rms_dbfs":-28.611899860103648,"output_peak":0.553898957590306,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.763391047758684},"common_prefix":{"samples":83120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.246557050986432,"spectral_convergence":0.5642032871467357,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.279429757494152,"spectral_convergence":0.3347041705010309,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.764186334795465,"spectral_convergence":0.37971674596672395,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":3.179162881744473e-05,"delta_db":0.044124939023837684,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00013649405808591368,"delta_db":0.2941706043617269,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0007422638230006736,"delta_db":1.4451093849405225,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.0002927078176851584,"delta_db":1.130210371873446,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":9.948064314264101e-05,"delta_db":0.8238767932024228,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":7.386915730586974e-05,"delta_db":0.6355102848066007,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7288337297014261,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1230371129869638,"pitch":null,"output_rms_dbfs":-28.611899860103648,"output_peak":0.553898957590306,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.763391047758684},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83120,"output_samples":83120,"intelligibility":{"estoi":{"value":0.9448022517648917,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03260266874006855,"rms_dbfs":-29.734936973090612,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.776881716111717,"mean":-5.646779975487248e-07,"zero_energy":false},"output":{"rms":0.037102656616982324,"rms_dbfs":-28.611899860103648,"peak":0.553898957590306,"peak_dbfs":-5.131389043453904,"crest_factor_db":23.48051081664974,"mean":-5.876197296702926e-12,"zero_energy":false},"rms_difference_db":1.1230371129869638},"full_scale":{"reference":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":4.0483610646640273e-07,"output_rms_dbfs":-63.927207606449656,"output_zero_energy":false,"output_peak":0.011895849985189041,"output_energy_fraction":7.472356863526288e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":183,"both_estimated_voiced_frames":164,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.8770053475935828,"absolute_difference_cents_median":3.4060788246563787,"absolute_difference_cents_p95":23.467841814068347,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.00015201529585382603,"log_rms_mae_db":1.9543973112627229,"rms_envelope_nrmse":0.1999556136362947,"envelope_correlation":0.9856113335981339,"rms_gain_error_db":1.1230374746969332,"reference_mean_power":0.0010639579210132953,"output_mean_power":0.001377933312528156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.575807359242935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021400202307720183,"output_power":0.0002859281700870856,"reference_active":true,"delta_db":1.2583906584235849},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3329698566280647e-05,"output_power":3.052077501107179e-05,"reference_active":true,"delta_db":1.1668642993598513},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3580980563714554e-06,"output_power":6.397918540970575e-06,"reference_active":true,"delta_db":0.770280491049095}],"active_band_count":3,"active_band_mae_db":1.0651784829441773}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.0001490763524181648,"log_rms_mae_db":1.5371808469607517,"rms_envelope_nrmse":0.16322376336185612,"envelope_correlation":0.994815303322848,"rms_gain_error_db":1.1230393190616947,"reference_mean_power":0.0010701430709341084,"output_mean_power":0.001385944296965379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":5.309268989226658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001912637761084412,"output_power":0.00025904102337707014,"reference_active":true,"delta_db":1.3173582137967654},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0962051121960504e-05,"output_power":1.3427519510704184e-05,"reference_active":true,"delta_db":0.8810396900418476},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1525134859671347e-07,"output_power":3.846737879706868e-07,"reference_active":true,"delta_db":0.8643564099152891}],"active_band_count":3,"active_band_mae_db":1.020918104584634}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.0001087945365454221,"log_rms_mae_db":1.4763191614426106,"rms_envelope_nrmse":0.2219495042586295,"envelope_correlation":0.9870927684669712,"rms_gain_error_db":1.3274094300481456,"reference_mean_power":0.0005739909463294839,"output_mean_power":0.0007791946917294225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":3.2669273354505384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013940368795916017,"output_power":0.00019651263998534473,"reference_active":true,"delta_db":1.4911622679476282},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.236537072685246e-06,"output_power":1.7457682442233816e-05,"reference_active":true,"delta_db":2.7647741164783413},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7105571121059603e-06,"output_power":2.394776575992338e-06,"reference_active":true,"delta_db":1.4612742253982858}],"active_band_count":3,"active_band_mae_db":1.9057368699414183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.00010533749157464783,"log_rms_mae_db":1.190964361328296,"rms_envelope_nrmse":0.19492593099965538,"envelope_correlation":0.9936328563692568,"rms_gain_error_db":1.3274092390376893,"reference_mean_power":0.0005773280667275119,"output_mean_power":0.0007837248097237676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":3.141212974645643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001274169753045263,"output_power":0.0001812815382854515,"reference_active":true,"delta_db":1.5312628644808792},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860144772024442e-06,"output_power":7.056905983404238e-06,"reference_active":true,"delta_db":2.620107386376223},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1444270061935377e-07,"output_power":1.4824057608616496e-07,"reference_active":true,"delta_db":1.1237899631345067}],"active_band_count":3,"active_band_mae_db":1.758386737997203}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":1.3936281127429397,"rms_envelope_nrmse":0.18583427092286053,"envelope_correlation":0.9894982915966918,"rms_gain_error_db":1.0797278649213282,"reference_mean_power":0.00021157950445783094,"output_mean_power":0.00027129786878781766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.2662436125548275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5392735637289224e-05,"output_power":7.118559602697147e-05,"reference_active":true,"delta_db":1.089393118405722},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790041823950811e-06,"output_power":6.582260892568456e-06,"reference_active":true,"delta_db":0.5569339103489054},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4111913003034656e-06,"output_power":1.919988275154393e-06,"reference_active":true,"delta_db":1.3371268613132559}],"active_band_count":3,"active_band_mae_db":0.9944846300226278}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":1.0118654380645828,"rms_envelope_nrmse":0.15179104306703356,"envelope_correlation":0.9965686666736994,"rms_gain_error_db":1.0797282650929978,"reference_mean_power":0.00021280955774068177,"output_mean_power":0.0002728751300392803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":1.2078803728484275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022425345543556e-05,"output_power":6.464338146934802e-05,"reference_active":true,"delta_db":1.0961057611311795},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7667526770926583e-06,"output_power":3.0186386982538233e-06,"reference_active":true,"delta_db":0.37840796461107024},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.98389498284838e-08,"output_power":1.2123375925206185e-07,"reference_active":true,"delta_db":0.8432356793378173}],"active_band_count":3,"active_band_mae_db":0.7725831350266891}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":1.6826631600769726,"rms_envelope_nrmse":0.27356071043128816,"envelope_correlation":0.9609329539346889,"rms_gain_error_db":0.7314636515225883,"reference_mean_power":0.00010955764718773973,"output_mean_power":0.00012965493808436503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":6.937978158746453e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2971450195505062e-05,"output_power":2.984248467887579e-05,"reference_active":true,"delta_db":1.136465663518436},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1375769637538911e-05,"output_power":1.3918255797673717e-05,"reference_active":true,"delta_db":0.8760402505284115},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447046226198746e-06,"output_power":3.0284930939145022e-06,"reference_active":true,"delta_db":0.9258441413077524}],"active_band_count":3,"active_band_mae_db":0.9794500184515332}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":1.1512383491826765,"rms_envelope_nrmse":0.18484842363487855,"envelope_correlation":0.9837276752610563,"rms_gain_error_db":0.7314641155752204,"reference_mean_power":0.00011019456862153084,"output_mean_power":0.00013040871053026002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":6.599778128222779e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3073624841125744e-05,"output_power":3.068454351818186e-05,"reference_active":true,"delta_db":1.2380383977368683},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.235055585786142e-06,"output_power":7.247601934423548e-06,"reference_active":true,"delta_db":0.6535400265063249},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9515593840599056e-07,"output_power":2.7496283432040533e-07,"reference_active":true,"delta_db":1.4889222488695992}],"active_band_count":3,"active_band_mae_db":1.1268335577042643}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"facodec","status":"measured","mos":{"status":"measured","value":4.141514301300049,"predicted_mos":4.141514301300049,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83200,"input_sample_rate":16000,"input_seconds":5.2,"input_float32_sha256":"7995a16418c3d92103db0f0745a8fe4eb8ab9ff122a21f0746e6df9d9b343820","internal_samples":83200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3293359500003135},"started_at":"2026-10-13T04:54:40.364829+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7155000680068042e-05,"native_input_samples":83120,"native_input_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","output_sha256":"ead5f9c983acea0ccc098adb712770c81e1dc44c5ba2ac2ea5a7ee77af9f5d86","output_samples":83200,"length_delta_samples":80,"analysis_s":1.9250079429994003,"synthesis_s":9.494791592000183,"total_s":11.419799534999584,"rtf":2.198228976900786,"output_peak":0.4993549585342407,"output_rms":0.029749101237141827,"finished_at":"2026-10-13T04:54:51.789660+00:00","evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"ead5f9c983acea0ccc098adb712770c81e1dc44c5ba2ac2ea5a7ee77af9f5d86","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":83120,"output_samples":83200,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.530523007882387,"output_peak":0.4993549585342407},"common_prefix":{"samples":83120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990384615384615,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.356442379917861,"spectral_convergence":0.24443443496096556,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.628123915783193,"spectral_convergence":0.261681047395615,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6876467688593575,"spectral_convergence":0.27625158671536665,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":2.9058182487868338e-05,"delta_db":-0.34631838450831665,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00012220108410275124,"delta_db":-0.18621626718196585,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.00046514068745448034,"delta_db":-0.5846302520702576,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.0001674830673816294,"delta_db":-1.2944237839277837,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":5.625925106100161e-05,"delta_db":-1.6515695858461368,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":4.571852354922277e-05,"delta_db":-1.4481991948347204,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9185595780615302,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.791408203780911,"pitch":null,"output_rms_dbfs":-30.526345176871523,"output_peak":0.4993549585342407,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.55740435485347},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83120,"output_samples":83120,"intelligibility":{"estoi":{"value":0.9387930544611196,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03260266874006855,"rms_dbfs":-29.734936973090612,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.776881716111717,"mean":-5.646779975487248e-07,"zero_energy":false},"output":{"rms":0.02976341371612057,"rms_dbfs":-30.526345176871523,"peak":0.4993549585342407,"peak_dbfs":-6.031812665493001,"crest_factor_db":24.49453251137852,"mean":-0.0001216129609014155,"zero_energy":false},"rms_difference_db":-0.7914082037809109},"full_scale":{"reference":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":4.163464104432926e-07,"output_rms_dbfs":-63.805451752866745,"output_zero_energy":false,"output_peak":0.0039931791834533215,"output_energy_fraction":0.00011942018007210217,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":182,"both_estimated_voiced_frames":168,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.8983957219251337,"absolute_difference_cents_median":3.700074571376058,"absolute_difference_cents_p95":17.18850932229992,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.00015201529585382603,"log_rms_mae_db":1.7404217791587,"rms_envelope_nrmse":0.12377745776929322,"envelope_correlation":0.9932598101312947,"rms_gain_error_db":-0.791410353557751,"reference_mean_power":0.0010639579210132953,"output_mean_power":0.0008867136965700296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.575807359242935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021400202307720183,"output_power":0.00017450568943078642,"reference_active":true,"delta_db":-0.8860828810664645},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3329698566280647e-05,"output_power":2.6120601730516736e-05,"reference_active":true,"delta_db":0.49074049913137296},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3580980563714554e-06,"output_power":6.572462123275848e-06,"reference_active":true,"delta_db":0.887174347798514}],"active_band_count":3,"active_band_mae_db":0.7546659093321172}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.0001490763524181648,"log_rms_mae_db":1.5527706621553585,"rms_envelope_nrmse":0.11665972086667654,"envelope_correlation":0.9945467644151806,"rms_gain_error_db":-0.7914134232325247,"reference_mean_power":0.0010701430709341084,"output_mean_power":0.0008918678350320131},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":5.309268989226658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001912637761084412,"output_power":0.00015602458575423575,"reference_active":true,"delta_db":-0.884396875716512},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0962051121960504e-05,"output_power":1.3107317781417346e-05,"reference_active":true,"delta_db":0.7762200575381707},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1525134859671347e-07,"output_power":4.404141892563644e-07,"reference_active":true,"delta_db":1.4520434952286587}],"active_band_count":3,"active_band_mae_db":1.0375534761611138}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.0001087945365454221,"log_rms_mae_db":1.3141432003648426,"rms_envelope_nrmse":0.11505833721512261,"envelope_correlation":0.9926525674734087,"rms_gain_error_db":-0.5202902511956607,"reference_mean_power":0.0005739909463294839,"output_mean_power":0.0005091854874586307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":3.2669273354505384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013940368795916017,"output_power":0.00012095548315710795,"reference_active":true,"delta_db":-0.6164870268937057},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.236537072685246e-06,"output_power":9.096365218620971e-06,"reference_active":true,"delta_db":-0.06641288769363898},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7105571121059603e-06,"output_power":1.506997536654751e-06,"reference_active":true,"delta_db":-0.5502503657827563}],"active_band_count":3,"active_band_mae_db":0.41105009345670035}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.00010533749157464783,"log_rms_mae_db":1.1037463780163703,"rms_envelope_nrmse":0.10723767614018438,"envelope_correlation":0.9937427595741498,"rms_gain_error_db":-0.5202910604845259,"reference_mean_power":0.0005773280667275119,"output_mean_power":0.0005121457405914792},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":3.141212974645643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001274169753045263,"output_power":0.00011043698939133564,"reference_active":true,"delta_db":-0.6211273246071852},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860144772024442e-06,"output_power":3.7657559944539947e-06,"reference_active":true,"delta_db":-0.10751416796572975},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1444270061935377e-07,"output_power":1.1894052030604761e-07,"reference_active":true,"delta_db":0.16741736254085235}],"active_band_count":3,"active_band_mae_db":0.29868628503792244}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":1.7080968135969337,"rms_envelope_nrmse":0.17468178026863113,"envelope_correlation":0.9890790914823693,"rms_gain_error_db":-1.2167290051618294,"reference_mean_power":0.00021157950445783094,"output_mean_power":0.0001598824133524192},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.2662436125548275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5392735637289224e-05,"output_power":4.292699188560429e-05,"reference_active":true,"delta_db":-1.1072235753007567},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790041823950811e-06,"output_power":4.001404062765955e-06,"reference_active":true,"delta_db":-1.6046929207678327},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4111913003034656e-06,"output_power":1.4776775196128158e-06,"reference_active":true,"delta_db":0.1999377584724043}],"active_band_count":3,"active_band_mae_db":0.9706180848469979}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":1.5573482516839565,"rms_envelope_nrmse":0.16187101445731547,"envelope_correlation":0.992290759634579,"rms_gain_error_db":-1.216730728582693,"reference_mean_power":0.00021280955774068177,"output_mean_power":0.0001608118530252144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":1.2078803728484275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022425345543556e-05,"output_power":3.888729211066712e-05,"reference_active":true,"delta_db":-1.1110578760924756},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7667526770926583e-06,"output_power":1.8278428786933447e-06,"reference_active":true,"delta_db":-1.8003147777479427},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.98389498284838e-08,"output_power":1.0068410818052417e-07,"reference_active":true,"delta_db":0.0366092344429932}],"active_band_count":3,"active_band_mae_db":0.9826606294278039}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":2.3803341041361135,"rms_envelope_nrmse":0.26543926542442914,"envelope_correlation":0.968871765032546,"rms_gain_error_db":-1.6457841395034192,"reference_mean_power":0.00010955764718773973,"output_mean_power":7.500052147190306e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":6.937978158746453e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2971450195505062e-05,"output_power":1.7215968667275325e-05,"reference_active":true,"delta_db":-1.2525694964003158},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1375769637538911e-05,"output_power":1.3697088054212272e-05,"reference_active":true,"delta_db":0.8064745881853207},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447046226198746e-06,"output_power":4.220361417442156e-06,"reference_active":true,"delta_db":2.367074705436793}],"active_band_count":3,"active_band_mae_db":1.4753729300074765}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":2.167964909642288,"rms_envelope_nrmse":0.2436175134888028,"envelope_correlation":0.9759713334077378,"rms_gain_error_db":-1.6457880519340709,"reference_mean_power":0.00011019456862153084,"output_mean_power":7.543647455707121e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":6.599778128222779e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3073624841125744e-05,"output_power":1.994196380247157e-05,"reference_active":true,"delta_db":-0.6334790345342274},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.235055585786142e-06,"output_power":8.031395899941251e-06,"reference_active":true,"delta_db":1.0995070497768245},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9515593840599056e-07,"output_power":3.414245469262634e-07,"reference_active":true,"delta_db":2.4291297078979577}],"active_band_count":3,"active_band_mae_db":1.3873719307363366}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.543990135192871,"predicted_mos":3.543990135192871,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":229376,"input_sample_rate":44100,"input_seconds":5.201269841269841,"input_float32_sha256":"fd2b67ee43fb82a9058a1326f00d9efdfef407fc8c36c6880f3a1dc4ef55b80c","internal_samples":83221,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.041567096999643},"started_at":"2026-10-13T07:19:28.435136+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007056079000903992,"native_input_samples":229100,"native_input_sha256":"f440f2bdff57076c588ebf87af9c7221e273a32365bf74e1f780638f401fb9e7","output_sha256":"dec5a228723e610f86cd761c87139a8707a6417f74882b4b1eba775417f2c317","output_samples":229376,"length_delta_samples":276,"analysis_s":0.130441855000754,"synthesis_s":39.17163490200073,"total_s":39.30207675700149,"rtf":7.565349563438523,"output_peak":0.4771862030029297,"output_rms":0.030301506819629422,"finished_at":"2026-10-13T07:20:07.768384+00:00","evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"dec5a228723e610f86cd761c87139a8707a6417f74882b4b1eba775417f2c317","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":83120,"output_samples":83221,"length_delta_samples":101,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.36239640813808,"output_peak":0.46481973846437924},"common_prefix":{"samples":83120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987863640186972,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.022682584013873,"spectral_convergence":0.5489139329583086,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":10.040825127655921,"spectral_convergence":0.5148436277009569,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.778755917187897,"spectral_convergence":0.5507064637235144,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":2.7269237635184916e-05,"delta_db":-0.62227288735418,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00010571284349983898,"delta_db":-0.815689306866829,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0005448268002671505,"delta_db":0.10211106953501789,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.00014059330605684507,"delta_db":-2.0544864110292136,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":4.7339221738597205e-05,"delta_db":-2.401297911934999,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":5.531822919917718e-05,"delta_db":-0.6204384677924515,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1027160090854486,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6221855111722835,"pitch":null,"output_rms_dbfs":-30.357122484262895,"output_peak":0.46481973846437924,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.94742120961923},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83120,"output_samples":83120,"intelligibility":{"estoi":{"value":0.7695407095187192,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03260266874006855,"rms_dbfs":-29.734936973090612,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.776881716111717,"mean":-5.646779975487248e-07,"zero_energy":false},"output":{"rms":0.030348964371075243,"rms_dbfs":-30.357122484262895,"peak":0.46481973846437924,"peak_dbfs":-6.65430876019668,"crest_factor_db":23.702813724066214,"mean":-1.9776282268601554e-05,"zero_energy":false},"rms_difference_db":-0.6221855111722832},"full_scale":{"reference":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":2.355457247811572e-07,"output_rms_dbfs":-66.27924773916948,"output_zero_energy":false,"output_peak":0.0043416995902899585,"output_energy_fraction":6.497942348109659e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":150,"both_estimated_voiced_frames":131,"reference_voiced_output_unvoiced_frames":56,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.7005347593582888,"absolute_difference_cents_median":6.805719113932817,"absolute_difference_cents_p95":32.48879481280081,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.00015201529585382603,"log_rms_mae_db":1.7855131109931575,"rms_envelope_nrmse":0.13991361456306153,"envelope_correlation":0.98747936551945,"rms_gain_error_db":-0.6221850935656053,"reference_mean_power":0.0010639579210132953,"output_mean_power":0.0009219469731470648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.575807359242935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021400202307720183,"output_power":0.0001846187932697249,"reference_active":true,"delta_db":-0.6414197104373333},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3329698566280647e-05,"output_power":2.1310610012717586e-05,"reference_active":true,"delta_db":-0.39313245983950984},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3580980563714554e-06,"output_power":4.978313251832148e-06,"reference_active":true,"delta_db":-0.31928436822867157}],"active_band_count":3,"active_band_mae_db":0.4512788461685049}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.0001490763524181648,"log_rms_mae_db":1.4897597069416957,"rms_envelope_nrmse":0.10956921258474525,"envelope_correlation":0.9936676799544257,"rms_gain_error_db":-0.6221833945566377,"reference_mean_power":0.0010701430709341084,"output_mean_power":0.0009273069278129153},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":5.309268989226658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001912637761084412,"output_power":0.00016618544757450766,"reference_active":true,"delta_db":-0.6103973469300733},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0962051121960504e-05,"output_power":1.0170346559210607e-05,"reference_active":true,"delta_db":-0.3255607113246642},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1525134859671347e-07,"output_power":2.9807968438025e-07,"reference_active":true,"delta_db":-0.2432457539149383}],"active_band_count":3,"active_band_mae_db":0.39306793738989193}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.0001087945365454221,"log_rms_mae_db":2.7199785129833804,"rms_envelope_nrmse":0.1545397736572707,"envelope_correlation":0.9832250802398065,"rms_gain_error_db":-0.22429487753904864,"reference_mean_power":0.0005739909463294839,"output_mean_power":0.000545099212349686},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":3.2669273354505384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013940368795916017,"output_power":0.00011869997217925868,"reference_active":true,"delta_db":-0.6982364611702492},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.236537072685246e-06,"output_power":7.890544574061969e-06,"reference_active":true,"delta_db":-0.6840220017658677},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7105571121059603e-06,"output_power":1.8074224241511264e-06,"reference_active":true,"delta_db":0.2392208714654517}],"active_band_count":3,"active_band_mae_db":0.5404931114671895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.00010533749157464783,"log_rms_mae_db":2.314880370023316,"rms_envelope_nrmse":0.11511511508596951,"envelope_correlation":0.9905924620583229,"rms_gain_error_db":-0.22429497259547232,"reference_mean_power":0.0005773280667275119,"output_mean_power":0.0005482683473753827},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":3.141212974645643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001274169753045263,"output_power":0.00010830311629300833,"reference_active":true,"delta_db":-0.7058633823202467},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860144772024442e-06,"output_power":3.434631770249706e-06,"reference_active":true,"delta_db":-0.5072341007140605},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1444270061935377e-07,"output_power":1.074238508016892e-07,"reference_active":true,"delta_db":-0.27487381349699197}],"active_band_count":3,"active_band_mae_db":0.49599043217709965}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":3.3363066444324163,"rms_envelope_nrmse":0.33004690323186753,"envelope_correlation":0.9373499245006198,"rms_gain_error_db":-1.8831607046784007,"reference_mean_power":0.00021157950445783094,"output_mean_power":0.00013713790958626662},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.2662436125548275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5392735637289224e-05,"output_power":3.2394467260018094e-05,"reference_active":true,"delta_db":-2.3298197164145664},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790041823950811e-06,"output_power":7.282008277084922e-06,"reference_active":true,"delta_db":0.9956946740471397},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4111913003034656e-06,"output_power":1.4575659762335827e-06,"reference_active":true,"delta_db":0.14042331607426692}],"active_band_count":3,"active_band_mae_db":1.1553125688453243}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":3.0310088693524317,"rms_envelope_nrmse":0.2903028072152963,"envelope_correlation":0.9568334927002515,"rms_gain_error_db":-1.883160573099965,"reference_mean_power":0.00021280955774068177,"output_mean_power":0.00013793518822872875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":1.2078803728484275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022425345543556e-05,"output_power":3.044582043468762e-05,"reference_active":true,"delta_db":-2.1738580837682324},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7667526770926583e-06,"output_power":3.3258821875831962e-06,"reference_active":true,"delta_db":0.7993652234340066},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.98389498284838e-08,"output_power":8.25179638703667e-08,"reference_active":true,"delta_db":-0.8275150094544057}],"active_band_count":3,"active_band_mae_db":1.2669127722188815}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":3.499866617548972,"rms_envelope_nrmse":0.2216869535951639,"envelope_correlation":0.9796273612491991,"rms_gain_error_db":-1.3499285950995397,"reference_mean_power":0.00010955764718773973,"output_mean_power":8.028785170091367e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":6.937978158746453e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2971450195505062e-05,"output_power":1.6306827063305198e-05,"reference_active":true,"delta_db":-1.4881894776192812},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1375769637538911e-05,"output_power":8.066011298663333e-06,"reference_active":true,"delta_db":-1.4932196287145603},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447046226198746e-06,"output_power":1.9428307740592392e-06,"reference_active":true,"delta_db":-1.002071995223852}],"active_band_count":3,"active_band_mae_db":1.3278270338525644}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":3.3167257263969434,"rms_envelope_nrmse":0.1949378523905413,"envelope_correlation":0.9872569677792195,"rms_gain_error_db":-1.3499273718332474,"reference_mean_power":0.00011019456862153084,"output_mean_power":8.075463377342871e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":6.599778128222779e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3073624841125744e-05,"output_power":1.6476245810597835e-05,"reference_active":true,"delta_db":-1.4625756445580462},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.235055585786142e-06,"output_power":4.427645597090992e-06,"reference_active":true,"delta_db":-1.4866747830656515},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9515593840599056e-07,"output_power":1.6158272861197396e-07,"reference_active":true,"delta_db":-0.819868332859386}],"active_band_count":3,"active_band_mae_db":1.2563729201610279}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"golf","status":"measured","mos":{"status":"measured","value":3.428478717803955,"predicted_mos":3.428478717803955,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":124560,"input_sample_rate":24000,"input_seconds":5.19,"input_float32_sha256":"e86c13e59e1e85710f5e3dc41479df04e431250c37025ecb199f57c7505b7666","internal_samples":83040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.221489682000538},"started_at":"2026-10-13T05:02:42.668722+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00261125599990919,"native_input_samples":124680,"native_input_sha256":"312eb46371fbaed8212515695273ac42234bc2ce5d779b9b311c9ec91f1d6344","output_sha256":"23c7f1299e557c1121c0d2843422ae325572ad6e5b92d32965ae867598a0990d","output_samples":124560,"length_delta_samples":-120,"analysis_s":0.47608629600017593,"synthesis_s":0.10428812299960555,"total_s":0.5803744189997815,"rtf":0.11171788623672405,"output_peak":0.5450625419616699,"output_rms":0.027485698067146645,"finished_at":"2026-10-13T05:02:43.260651+00:00","analysis_stages":{"external_f0_s":0.04279541699997935},"evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"23c7f1299e557c1121c0d2843422ae325572ad6e5b92d32965ae867598a0990d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":83120,"output_samples":83040,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.21115019023891,"output_peak":0.49966420711818305},"common_prefix":{"samples":83040,"sample_rate":16000,"reference_coverage":0.9990375360923965,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83040,"output_samples":83040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.643688578544103,"spectral_convergence":0.5356109584042638,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.653551868758957,"spectral_convergence":0.3595566695869477,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.768448415119349,"spectral_convergence":0.40012245444601247,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.163145597226362e-05,"output_power":2.7935830890500872e-05,"delta_db":-0.539575895934176,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012754954811613938,"output_power":0.00011289972248181045,"delta_db":-0.5298604993615177,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000532676093292526,"output_power":0.00036096800456574554,"delta_db":-1.6899449724444726,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022585595153731738,"output_power":0.00017198410497476445,"delta_db":-1.1834322853162844,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.236999454739008e-05,"output_power":5.235132755088437e-05,"delta_db":-1.968413380258843,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.387490421215087e-05,"output_power":3.0493490648628017e-05,"delta_db":-3.2112312013098347,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5204097057708548,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.4803948208404591,"pitch":null,"output_rms_dbfs":-31.21115019023891,"output_peak":0.49966420711818305,"reference_rms_dbfs":-29.73075536939845,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.021896287474136},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83040,"output_samples":83040,"intelligibility":{"estoi":{"value":0.9000880654449804,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03261836825590433,"rms_dbfs":-29.73075536939845,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.772700112419557,"mean":-3.6015446245785163e-07,"zero_energy":false},"output":{"rms":0.027506953322975154,"rms_dbfs":-31.211150190238914,"peak":0.49966420711818305,"peak_dbfs":-6.026435192769756,"crest_factor_db":25.184714997469154,"mean":0.0001278551599016773,"zero_energy":false},"rms_difference_db":-1.4803948208404627},"full_scale":{"reference":{"sample_count":83040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":4.5214893060353365e-07,"output_rms_dbfs":-63.4471849198973,"output_zero_energy":false,"output_peak":0.004674794141638143,"output_energy_fraction":0.00015198583030295537,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":177,"both_estimated_voiced_frames":158,"reference_voiced_output_unvoiced_frames":29,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.8449197860962567,"absolute_difference_cents_median":4.2480303213528074,"absolute_difference_cents_p95":16.875002648472652,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":0.00015201529585382603,"log_rms_mae_db":3.4319320039479106,"rms_envelope_nrmse":0.2522771876072401,"envelope_correlation":0.965974496605176,"rms_gain_error_db":-1.4803955119255054,"reference_mean_power":0.001064983834157785,"output_mean_power":0.0007573619185949658},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":5.556612266072114e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021326811788506428,"output_power":0.0001766874683961731,"reference_active":true,"delta_db":-0.8172018811556435},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3206922278415e-05,"output_power":2.1337778321181088e-05,"reference_active":true,"delta_db":-0.36468348856872734},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3431743252264855e-06,"output_power":2.5847676082018382e-06,"reference_active":true,"delta_db":-3.153778410591352}],"active_band_count":3,"active_band_mae_db":1.4452212601052412}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1031,"rms_floor":0.0001490763524181648,"log_rms_mae_db":3.2376669115079504,"rms_envelope_nrmse":0.23517128766043663,"envelope_correlation":0.9726145073376011,"rms_gain_error_db":-1.4804030097431564,"reference_mean_power":0.0010711809448856284,"output_mean_power":0.0007617676709758524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1031,"frequency_resolution_hz":0.19398642095053345,"power_floor":5.291101115290615e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001905478061525919,"output_power":0.00015919765922924655,"reference_active":true,"delta_db":-0.7806727516594503},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.089605333994107e-05,"output_power":9.163713682664151e-06,"reference_active":true,"delta_db":-0.7519770925966097},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.136986885662709e-07,"output_power":1.7825326596036853e-07,"reference_active":true,"delta_db":-2.4547520736069233}],"active_band_count":3,"active_band_mae_db":1.3291339726209943}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":0.0001087945365454221,"log_rms_mae_db":2.2804575444608344,"rms_envelope_nrmse":0.27580847648669626,"envelope_correlation":0.9589508332308097,"rms_gain_error_db":-1.477652541273825,"reference_mean_power":0.0005745444556266076,"output_mean_power":0.00040884471102479525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":3.254753972361981e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001389486718870661,"output_power":0.00011267653376823509,"reference_active":true,"delta_db":-0.9102092155822659},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.171306455849933e-06,"output_power":1.1666013624951185e-05,"reference_active":true,"delta_db":1.0449127383534265},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.702013202159027e-06,"output_power":1.4419263174727617e-06,"reference_active":true,"delta_db":-0.720198560181896}],"active_band_count":3,"active_band_mae_db":0.8917735047058627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1031,"rms_floor":0.00010533749157464783,"log_rms_mae_db":2.0826312538540077,"rms_envelope_nrmse":0.2544246416212921,"envelope_correlation":0.9669680735711474,"rms_gain_error_db":-1.477652478163432,"reference_mean_power":0.0005778880337425662,"output_mean_power":0.00041122400064907966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1031,"frequency_resolution_hz":0.19398642095053345,"power_floor":3.1295279798906464e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012697002472569506,"output_power":0.00010390094626908278,"reference_active":true,"delta_db":-0.8708170128703832},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.83107811849647e-06,"output_power":4.93719434691872e-06,"reference_active":true,"delta_db":1.1015921539685922},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1389404327603784e-07,"output_power":9.869765357386824e-08,"reference_active":true,"delta_db":-0.6219418289424639}],"active_band_count":3,"active_band_mae_db":0.8647836652604797}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":2.488139915496164,"rms_envelope_nrmse":0.2095711288861033,"envelope_correlation":0.9786047192888105,"rms_gain_error_db":-1.0936187068830492,"reference_mean_power":0.00021178352755146868,"output_mean_power":0.00016463808545669008},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":1.2610055739271758e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.515338656426855e-05,"output_power":4.610938599799997e-05,"reference_active":true,"delta_db":-0.7778284532451032},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.762171675169058e-06,"output_power":5.520445336794963e-06,"reference_active":true,"delta_db":-0.1861207941025435},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4060397616718376e-06,"output_power":8.946678382987994e-07,"reference_active":true,"delta_db":-1.9633577683792391}],"active_band_count":3,"active_band_mae_db":0.9757690052422953}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1031,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":2.266376239876392,"rms_envelope_nrmse":0.1808356207932372,"envelope_correlation":0.9869588870190524,"rms_gain_error_db":-1.0936188668467361,"reference_mean_power":0.000213015961222312,"output_mean_power":0.0001655961591587041},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1031,"frequency_resolution_hz":0.19398642095053345,"power_floor":1.2028800026943469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.9996444083284706e-05,"output_power":4.220957883584815e-05,"reference_active":true,"delta_db":-0.7352809812070569},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7525281948476266e-06,"output_power":2.437884125224995e-06,"reference_active":true,"delta_db":-0.5271871688108822},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.929036640910416e-08,"output_power":7.425929066448713e-08,"reference_active":true,"delta_db":-1.2615631690012699}],"active_band_count":3,"active_band_mae_db":0.8413437730064031}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":3.7954263210794106,"rms_envelope_nrmse":0.3540522757166111,"envelope_correlation":0.9588199214793801,"rms_gain_error_db":-2.7982985518660692,"reference_mean_power":0.00010966329119893038,"output_mean_power":5.757466511301533e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":6.915006284659968e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2942106271538924e-05,"output_power":1.2806766165995323e-05,"reference_active":true,"delta_db":-2.5319380717977378},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1326187270790092e-05,"output_power":4.461820329002766e-06,"reference_active":true,"delta_db":-4.045716602187094},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4330009782866417e-06,"output_power":8.279780928846244e-07,"reference_active":true,"delta_db":-4.681234374301073}],"active_band_count":3,"active_band_mae_db":3.752963016095302}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83040,"uncovered_tail_samples":0,"metrics":{"frame_count":1031,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":3.8929042468981403,"rms_envelope_nrmse":0.3408850170933071,"envelope_correlation":0.9655666218550604,"rms_gain_error_db":-2.7982991784766007,"reference_mean_power":0.00011030144595655165,"output_mean_power":5.79096963689931e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1031,"frequency_resolution_hz":0.19398642095053345,"power_floor":6.577895559651072e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3035765585475087e-05,"output_power":1.2003475874571003e-05,"reference_active":true,"delta_db":-2.830956265233437},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.189761540176797e-06,"output_power":2.405324113676905e-06,"reference_active":true,"delta_db":-4.105003130416086},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9470846451511378e-07,"output_power":7.091806714684563e-08,"reference_active":true,"delta_db":-4.38627941402922}],"active_band_count":3,"active_band_mae_db":3.7740796032262476}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.0003674030303955,"predicted_mos":3.0003674030303955,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":124800,"input_sample_rate":24000,"input_seconds":5.2,"input_float32_sha256":"ec71f1ae9728f222bc59e37b4e590f4164f08492274d54e3ee31bfe780c91486","internal_samples":83200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.622086314000626},"started_at":"2026-10-13T07:40:24.390307+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003500238999549765,"native_input_samples":124680,"native_input_sha256":"312eb46371fbaed8212515695273ac42234bc2ce5d779b9b311c9ec91f1d6344","output_sha256":"7ce83e9964a31c9e29916eca7e9d16be7dd15844b3773c5dbf7ae04fb5dad264","output_samples":124800,"length_delta_samples":120,"analysis_s":1.8888223989997641,"synthesis_s":16.003245443000196,"total_s":17.89206784199996,"rtf":3.444093906063515,"output_peak":0.4252324402332306,"output_rms":0.031173708328731457,"finished_at":"2026-10-13T07:40:42.302855+00:00","analysis_stages":{"external_f0_s":1.1897174779987836},"evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"7ce83e9964a31c9e29916eca7e9d16be7dd15844b3773c5dbf7ae04fb5dad264","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":83120,"output_samples":83200,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.117349981984972,"output_peak":0.4266073684530078},"common_prefix":{"samples":83120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990384615384615,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.21182619587485,"spectral_convergence":0.48451643981680825,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.566322878165482,"spectral_convergence":0.3288112424193505,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.605217839514634,"spectral_convergence":0.35515671980254776,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":3.376606819587822e-05,"delta_db":0.30580207964602635,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00012030560383345867,"delta_db":-0.25410828228201693,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0004919343787204091,"delta_db":-0.3414018142159695,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.00022372734901121273,"delta_db":-0.03694207979520051,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":6.601039617742019e-05,"delta_db":-0.957385657186052,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":3.853395038990059e-05,"delta_db":-2.1906858224975987,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6810542892704774,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.37823519834445873,"pitch":null,"output_rms_dbfs":-30.113172171435068,"output_peak":0.4266073684530078,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.794455637851655},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83120,"output_samples":83120,"intelligibility":{"estoi":{"value":0.9201095596561781,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03260266874006855,"rms_dbfs":-29.734936973090612,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.776881716111717,"mean":-5.646779975487248e-07,"zero_energy":false},"output":{"rms":0.03121342253467697,"rms_dbfs":-30.113172171435068,"peak":0.4266073684530078,"peak_dbfs":-7.3994329519069035,"crest_factor_db":22.713739219528165,"mean":0.00012642197143193006,"zero_energy":false},"rms_difference_db":-0.3782351983444556},"full_scale":{"reference":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":1.377215134851726e-06,"output_rms_dbfs":-58.60998213278958,"output_zero_energy":false,"output_peak":0.046216761153897765,"output_energy_fraction":0.0003591760608697063,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":173,"both_estimated_voiced_frames":156,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.8342245989304813,"absolute_difference_cents_median":5.221836342464519,"absolute_difference_cents_p95":30.66403372125169,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.00015201529585382603,"log_rms_mae_db":2.4543189036891464,"rms_envelope_nrmse":0.15941138792496437,"envelope_correlation":0.9799079515367758,"rms_gain_error_db":-0.3782349697962819,"reference_mean_power":0.0010639579210132953,"output_mean_power":0.0009752163081352968},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.575807359242935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021400202307720183,"output_power":0.00020461555102178485,"reference_active":true,"delta_db":-0.1947924146959075},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3329698566280647e-05,"output_power":2.2760958703822108e-05,"reference_active":true,"delta_db":-0.10718576658973672},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3580980563714554e-06,"output_power":3.5971268755228273e-06,"reference_active":true,"delta_db":-1.7305490094988918}],"active_band_count":3,"active_band_mae_db":0.6775090635948454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.0001490763524181648,"log_rms_mae_db":1.8979756891175668,"rms_envelope_nrmse":0.11711313084241744,"envelope_correlation":0.9892483874684196,"rms_gain_error_db":-0.3782333602017471,"reference_mean_power":0.0010701430709341084,"output_mean_power":0.000980885936361884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":5.309268989226658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001912637761084412,"output_power":0.00018416344764058692,"reference_active":true,"delta_db":-0.16429289180553885},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0962051121960504e-05,"output_power":1.0399613881880532e-05,"reference_active":true,"delta_db":-0.22874608030292873},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1525134859671347e-07,"output_power":2.2252891941677334e-07,"reference_active":true,"delta_db":-1.512704940602697}],"active_band_count":3,"active_band_mae_db":0.6352479709037215}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.0001087945365454221,"log_rms_mae_db":1.9252186285847785,"rms_envelope_nrmse":0.14970470739594896,"envelope_correlation":0.9844940582377775,"rms_gain_error_db":-0.3328175031882064,"reference_mean_power":0.0005739909463294839,"output_mean_power":0.0005316469074009772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":3.2669273354505384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013940368795916017,"output_power":0.00013465721995909276,"reference_active":true,"delta_db":-0.1504461920476185},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.236537072685246e-06,"output_power":1.0086166653228523e-05,"reference_active":true,"delta_db":0.3821696200704435},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7105571121059603e-06,"output_power":1.6122159193054302e-06,"reference_active":true,"delta_db":-0.25714373855098327}],"active_band_count":3,"active_band_mae_db":0.2632531835563484}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.00010533749157464783,"log_rms_mae_db":1.4829166542710794,"rms_envelope_nrmse":0.11616582539476313,"envelope_correlation":0.9907060711169084,"rms_gain_error_db":-0.3328173671484974,"reference_mean_power":0.0005773280667275119,"output_mean_power":0.000534737860951893},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":3.141212974645643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001274169753045263,"output_power":0.00012367041673509623,"reference_active":true,"delta_db":-0.12961467099817392},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860144772024442e-06,"output_power":4.163601290243531e-06,"reference_active":true,"delta_db":0.32865541558778655},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1444270061935377e-07,"output_power":1.1006318317802209e-07,"reference_active":true,"delta_db":-0.16946028689161052}],"active_band_count":3,"active_band_mae_db":0.209243457825857}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":1.9547955467623492,"rms_envelope_nrmse":0.1624919629867633,"envelope_correlation":0.9810995762748793,"rms_gain_error_db":-0.11989006761391166,"reference_mean_power":0.00021157950445783094,"output_mean_power":0.00020581858536207167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.2662436125548275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5392735637289224e-05,"output_power":5.417157879682782e-05,"reference_active":true,"delta_db":-0.09681320821862449},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790041823950811e-06,"output_power":5.236411149753703e-06,"reference_active":true,"delta_db":-0.4364796191241457},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4111913003034656e-06,"output_power":1.1815069083627773e-06,"reference_active":true,"delta_db":-0.7714962513275212}],"active_band_count":3,"active_band_mae_db":0.43492969289009714}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":1.3047663271438616,"rms_envelope_nrmse":0.10040817991905376,"envelope_correlation":0.9925153039820634,"rms_gain_error_db":-0.1198899289448694,"reference_mean_power":0.00021280955774068177,"output_mean_power":0.00020701515317601146},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":1.2078803728484275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022425345543556e-05,"output_power":4.916654938132961e-05,"reference_active":true,"delta_db":-0.09243760418487011},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7667526770926583e-06,"output_power":2.366076191019633e-06,"reference_active":true,"delta_db":-0.6794161342205923},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.98389498284838e-08,"output_power":7.617727940494628e-08,"reference_active":true,"delta_db":-1.1747454595785418}],"active_band_count":3,"active_band_mae_db":0.6488663993280014}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":2.251710656119745,"rms_envelope_nrmse":0.3241883566410892,"envelope_correlation":0.9459607296912611,"rms_gain_error_db":-1.8365881116323841,"reference_mean_power":0.00010955764718773973,"output_mean_power":7.177676584985913e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":6.937978158746453e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2971450195505062e-05,"output_power":1.42527966755994e-05,"reference_active":true,"delta_db":-2.072883234989167},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1375769637538911e-05,"output_power":7.5235726359825915e-06,"reference_active":true,"delta_db":-1.7955667068814962},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447046226198746e-06,"output_power":1.5928083841241578e-06,"reference_active":true,"delta_db":-1.8647864048023624}],"active_band_count":3,"active_band_mae_db":1.911078782224342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":1.66423453678252,"rms_envelope_nrmse":0.2541460531117622,"envelope_correlation":0.9768565880144255,"rms_gain_error_db":-1.836588846026025,"reference_mean_power":0.00011019456862153084,"output_mean_power":7.219403314527231e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":6.599778128222779e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3073624841125744e-05,"output_power":1.4305392441195745e-05,"reference_active":true,"delta_db":-2.0761605072041474},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.235055585786142e-06,"output_power":4.107352491083381e-06,"reference_active":true,"delta_db":-1.8127835418183813},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9515593840599056e-07,"output_power":1.2544395107628427e-07,"reference_active":true,"delta_db":-1.919320465595017}],"active_band_count":3,"active_band_mae_db":1.9360881715391818}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.043696641921997,"predicted_mos":3.043696641921997,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":124920,"input_sample_rate":24000,"input_seconds":5.205,"input_float32_sha256":"2e0afe8fecb50f265129511ace97f38ff9e14488231b82ad5fb556c86faea0c2","internal_samples":83280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.680631207998886},"started_at":"2026-10-13T05:10:56.799369+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002712520999921253,"native_input_samples":124680,"native_input_sha256":"312eb46371fbaed8212515695273ac42234bc2ce5d779b9b311c9ec91f1d6344","output_sha256":"e1151c9bc29154ceeb86e6d9c4bcfcd57d4d03f96cca0588dcf853099d69626d","output_samples":124920,"length_delta_samples":240,"analysis_s":0.5826162260000274,"synthesis_s":0.13011965099940426,"total_s":0.7127358769994316,"rtf":0.1371965114532111,"output_peak":0.5503891110420227,"output_rms":0.031884909836821476,"finished_at":"2026-10-13T05:10:57.525070+00:00","analysis_stages":{"external_f0_s":0.06740415900094376},"evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"e1151c9bc29154ceeb86e6d9c4bcfcd57d4d03f96cca0588dcf853099d69626d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":83120,"output_samples":83280,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.92140913534236,"output_peak":0.5507622871121516},"common_prefix":{"samples":83120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980787704130644,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.693119404274282,"spectral_convergence":0.22307704556932792,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.492661268514029,"spectral_convergence":0.22643552315559337,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.969346010564028,"spectral_convergence":0.286963461898377,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":2.9651250237864102e-05,"delta_db":-0.25857275286868314,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.0001263744625348016,"delta_db":-0.04037363903432147,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0005180164513332633,"delta_db":-0.11703803182785033,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.0002102457449073967,"delta_db":-0.30686069585193465,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":7.734300086955197e-05,"delta_db":-0.2692988653027477,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":5.859010366992156e-05,"delta_db":-0.37087849714562315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22717041367186008,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.17812038214413928,"pitch":null,"output_rms_dbfs":-29.913057355234752,"output_peak":0.5507622871121516,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.718375561117447},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83120,"output_samples":83120,"intelligibility":{"estoi":{"value":0.9677103880049032,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03260266874006855,"rms_dbfs":-29.734936973090612,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.776881716111717,"mean":-5.646779975487248e-07,"zero_energy":false},"output":{"rms":0.03194089875931483,"rms_dbfs":-29.913057355234752,"peak":0.5507622871121516,"peak_dbfs":-5.180716105405654,"crest_factor_db":24.732341249829094,"mean":6.925803220473322e-05,"zero_energy":false},"rms_difference_db":-0.17812038214414017},"full_scale":{"reference":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":3.923974207559015e-07,"output_rms_dbfs":-64.06273855917881,"output_zero_energy":false,"output_peak":0.01297846097419319,"output_energy_fraction":9.772828080667729e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":166,"both_estimated_voiced_frames":165,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.8823529411764706,"absolute_difference_cents_median":0.9342179881881171,"absolute_difference_cents_p95":7.890775134309614,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.00015201529585382603,"log_rms_mae_db":0.9528807297558959,"rms_envelope_nrmse":0.0667424426603945,"envelope_correlation":0.9966435541446348,"rms_gain_error_db":-0.17812041641409068,"reference_mean_power":0.0010639579210132953,"output_mean_power":0.0010212037725986463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.575807359242935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021400202307720183,"output_power":0.00020712907725030003,"reference_active":true,"delta_db":-0.1417680857094098},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3329698566280647e-05,"output_power":2.0624059672520612e-05,"reference_active":true,"delta_db":-0.5353497090949383},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3580980563714554e-06,"output_power":3.9583332674407e-06,"reference_active":true,"delta_db":-1.3149830084371246}],"active_band_count":3,"active_band_mae_db":0.6640336010804909}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.0001490763524181648,"log_rms_mae_db":0.5928064812420064,"rms_envelope_nrmse":0.0386356016277007,"envelope_correlation":0.9990481939098945,"rms_gain_error_db":-0.17811959677049058,"reference_mean_power":0.0010701430709341084,"output_mean_power":0.0010271405719390933},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":5.309268989226658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001912637761084412,"output_power":0.00018547282310053855,"reference_active":true,"delta_db":-0.13352443315874407},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0962051121960504e-05,"output_power":9.492511766539546e-06,"reference_active":true,"delta_db":-0.6251067894414568},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1525134859671347e-07,"output_power":2.138344516526121e-07,"reference_active":true,"delta_db":-1.685792758075192}],"active_band_count":3,"active_band_mae_db":0.8148079935584643}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.0001087945365454221,"log_rms_mae_db":0.7177049137325082,"rms_envelope_nrmse":0.06897502622022943,"envelope_correlation":0.9967481709771182,"rms_gain_error_db":-0.1377826315343802,"reference_mean_power":0.0005739909463294839,"output_mean_power":0.0005560665603279156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":3.2669273354505384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013940368795916017,"output_power":0.00013545634939134352,"reference_active":true,"delta_db":-0.1247489628299297},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.236537072685246e-06,"output_power":8.332811154681565e-06,"reference_active":true,"delta_db":-0.44717638118539305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7105571121059603e-06,"output_power":1.3849089851115264e-06,"reference_active":true,"delta_db":-0.9171634607167132}],"active_band_count":3,"active_band_mae_db":0.4963629349106786}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.00010533749157464783,"log_rms_mae_db":0.43579377922165263,"rms_envelope_nrmse":0.03903270353653734,"envelope_correlation":0.9990416001570488,"rms_gain_error_db":-0.1377825084318471,"reference_mean_power":0.0005773280667275119,"output_mean_power":0.0005592994861655165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":3.141212974645643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001274169753045263,"output_power":0.00012385455485137937,"reference_active":true,"delta_db":-0.12315308596701946},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860144772024442e-06,"output_power":3.5008470722573215e-06,"reference_active":true,"delta_db":-0.42430453023626624},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1444270061935377e-07,"output_power":8.838576354866256e-08,"reference_active":true,"delta_db":-1.1220577967179044}],"active_band_count":3,"active_band_mae_db":0.5565051376403968}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":0.9668820720041365,"rms_envelope_nrmse":0.075185434877912,"envelope_correlation":0.9963840677756456,"rms_gain_error_db":-0.23308470122246047,"reference_mean_power":0.00021157950445783094,"output_mean_power":0.000200523430367798},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.2662436125548275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5392735637289224e-05,"output_power":5.325190819217618e-05,"reference_active":true,"delta_db":-0.17117639236174712},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790041823950811e-06,"output_power":5.383377915547078e-06,"reference_active":true,"delta_db":-0.31626832280219236},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4111913003034656e-06,"output_power":1.0839419658126013e-06,"reference_active":true,"delta_db":-1.145798597377075}],"active_band_count":3,"active_band_mae_db":0.5444144375136716}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":0.6624646857025018,"rms_envelope_nrmse":0.04886692621475287,"envelope_correlation":0.9987115659768608,"rms_gain_error_db":-0.2330843851582629,"reference_mean_power":0.00021280955774068177,"output_mean_power":0.00020168922196966865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":1.2078803728484275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022425345543556e-05,"output_power":4.8421717277309484e-05,"reference_active":true,"delta_db":-0.15873302375145254},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7667526770926583e-06,"output_power":2.547071429310488e-06,"reference_active":true,"delta_db":-0.359292144801993},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.98389498284838e-08,"output_power":7.003608660497958e-08,"reference_active":true,"delta_db":-1.5397813308111032}],"active_band_count":3,"active_band_mae_db":0.685935499788183}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":1.1702138570955296,"rms_envelope_nrmse":0.10774753347549759,"envelope_correlation":0.9932345896005207,"rms_gain_error_db":-0.32072513866419233,"reference_mean_power":0.00010955764718773973,"output_mean_power":0.00010175837970328024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":6.937978158746453e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2971450195505062e-05,"output_power":2.1422489332245052e-05,"reference_active":true,"delta_db":-0.30318477982323316},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1375769637538911e-05,"output_power":1.0057365621963419e-05,"reference_active":true,"delta_db":-0.5349655028663745},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447046226198746e-06,"output_power":1.4457396481501582e-06,"reference_active":true,"delta_db":-2.2855208218552}],"active_band_count":3,"active_band_mae_db":1.0412237015149357}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":0.9261787061014645,"rms_envelope_nrmse":0.06317846941713252,"envelope_correlation":0.9981412808953306,"rms_gain_error_db":-0.32072466014562373,"reference_mean_power":0.00011019456862153084,"output_mean_power":0.00010234997079957192},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":6.599778128222779e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3073624841125744e-05,"output_power":2.1215638789552622e-05,"reference_active":true,"delta_db":-0.36459714433627727},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.235055585786142e-06,"output_power":5.188897913546219e-06,"reference_active":true,"delta_db":-0.7976520312266628},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9515593840599056e-07,"output_power":1.2204510043599737e-07,"reference_active":true,"delta_db":-2.0386142185292373}],"active_band_count":3,"active_band_mae_db":1.0669544646973925}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5142-33396-0065","method":"magphase","status":"measured","mos":{"status":"measured","value":3.4230215549468994,"predicted_mos":3.4230215549468994,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":249793,"input_sample_rate":48000,"input_seconds":5.204020833333334,"input_float32_sha256":"e0bb29b0e5e1fc8adef072be3ec7d46828ce3c41a0a04eb5c1f90d0a0cff2ad7","internal_samples":83265,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.252437828999973},"started_at":"2026-10-13T05:16:28.231056+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.007699542000409565,"native_input_samples":249360,"native_input_sha256":"86837e007590808b10315775fd46f748afcde8a594c5ca96ac6376f966b0e431","output_sha256":"49f28b57a839cbd6ea8118d98fc38e415d657647c97de2ef0438ae9529391695","output_samples":249793,"length_delta_samples":433,"analysis_s":3.058354425998914,"synthesis_s":1.1614204200013774,"total_s":4.219774846000291,"rtf":0.8122761974976499,"output_peak":0.6084040301162488,"output_rms":0.032144539529907874,"finished_at":"2026-10-13T05:16:32.468864+00:00","evaluation_input":{"reference":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"output_sha256":"49f28b57a839cbd6ea8118d98fc38e415d657647c97de2ef0438ae9529391695","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":83120,"output_samples":83265,"length_delta_samples":145,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.85105930122224,"output_peak":0.5829752869590812},"common_prefix":{"samples":83120,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982585720290639,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83120,"output_samples":83120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.9030139480409,"spectral_convergence":0.4934593503721427,"spectral_convergence_status":"ok","amplitude_floor":8.933262458231008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.744059243555252,"spectral_convergence":0.34430339035339314,"spectral_convergence_status":"ok","amplitude_floor":6.540934930277788e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.537578417325725,"spectral_convergence":0.29530586320120306,"spectral_convergence_status":"ok","amplitude_floor":6.044182898288457e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.147025669043969e-05,"output_power":2.7605870292693957e-05,"delta_db":-0.5689884219310494,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00012755476455897642,"output_power":0.00012203425525318472,"delta_db":-0.19214930375631267,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005321662836872518,"output_power":0.0005024281336652378,"delta_db":-0.24973405706979104,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022563854189948133,"output_power":0.00022227822238463782,"delta_db":-0.06516369640759234,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.229072811272526e-05,"output_power":9.195128207594173e-05,"delta_db":0.4820688397877524,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.381343402576783e-05,"output_power":7.03972377125017e-05,"delta_db":0.42643502210758977,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3307565568433479,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10855298181222163,"pitch":null,"output_rms_dbfs":-29.843489954902832,"output_peak":0.5829752869590812,"reference_rms_dbfs":-29.734936973090612,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.061550536307293},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83120,"output_samples":83120,"intelligibility":{"estoi":{"value":0.9432024050238873,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03260266874006855,"rms_dbfs":-29.734936973090612,"peak":0.5650634765625,"peak_dbfs":-4.958055256978897,"crest_factor_db":24.776881716111717,"mean":-5.646779975487248e-07,"zero_energy":false},"output":{"rms":0.03219774838997594,"rms_dbfs":-29.843489954902832,"peak":0.5829752869590812,"peak_dbfs":-4.686997102582526,"crest_factor_db":25.156492852320305,"mean":-0.00017743340757376294,"zero_energy":false},"rms_difference_db":-0.1085529818122204},"full_scale":{"reference":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":519,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.309279075125232e-06,"threshold_dbfs":-56.36523579658352,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":132,"quiet_samples":21120,"reference_mean_power":3.116785647404013e-07,"output_mean_power":3.956347646489597e-07,"output_rms_dbfs":-64.02705553698088,"output_zero_energy":false,"output_peak":0.010738899223486132,"output_energy_fraction":9.696875532309806e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":512,"reference_estimated_voiced_frames":187,"output_estimated_voiced_frames":189,"both_estimated_voiced_frames":170,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9090909090909091,"absolute_difference_cents_median":3.876029811014325,"absolute_difference_cents_p95":22.269082386402772,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.00015201529585382603,"log_rms_mae_db":2.0168019011089418,"rms_envelope_nrmse":0.16221899776963716,"envelope_correlation":0.9785309045253568,"rms_gain_error_db":-0.10855283225252366,"reference_mean_power":0.0010639579210132953,"output_mean_power":0.0010376936734257413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.575807359242935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021400202307720183,"output_power":0.00020772912966814723,"reference_active":true,"delta_db":-0.1292047746161266},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3329698566280647e-05,"output_power":2.678728165246245e-05,"reference_active":true,"delta_db":0.6001951657415534},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3580980563714554e-06,"output_power":7.7695463159312e-06,"reference_active":true,"delta_db":1.6138500276260759}],"active_band_count":3,"active_band_mae_db":0.781083322661252}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.0001490763524181648,"log_rms_mae_db":1.3750171809596006,"rms_envelope_nrmse":0.11166763866026697,"envelope_correlation":0.9892630576774828,"rms_gain_error_db":-0.10855204394388454,"reference_mean_power":0.0010701430709341084,"output_mean_power":0.0010437263297779522},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":5.309268989226658e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001912637761084412,"output_power":0.0001864046642923551,"reference_active":true,"delta_db":-0.1117595052806229},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0962051121960504e-05,"output_power":1.3226067862267755e-05,"reference_active":true,"delta_db":0.8153892362566785},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1525134859671347e-07,"output_power":4.841966485623838e-07,"reference_active":true,"delta_db":1.8636482598961945}],"active_band_count":3,"active_band_mae_db":0.9302656671444987}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":0.0001087945365454221,"log_rms_mae_db":1.8818300439179154,"rms_envelope_nrmse":0.15498118301479655,"envelope_correlation":0.9829875231961248,"rms_gain_error_db":-0.23159472982884305,"reference_mean_power":0.0005739909463294839,"output_mean_power":0.0005441837502395381},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":3.2669273354505384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013940368795916017,"output_power":0.00013352610943842064,"reference_active":true,"delta_db":-0.18708068170898984},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.236537072685246e-06,"output_power":8.74493840067472e-06,"reference_active":true,"delta_db":-0.23752423045243334},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7105571121059603e-06,"output_power":1.6991804373208553e-06,"reference_active":true,"delta_db":-0.02898079599676106}],"active_band_count":3,"active_band_mae_db":0.15119523605272808}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":0.00010533749157464783,"log_rms_mae_db":1.3892490278243441,"rms_envelope_nrmse":0.11035220694111211,"envelope_correlation":0.9912693336766905,"rms_gain_error_db":-0.2315948050231291,"reference_mean_power":0.0005773280667275119,"output_mean_power":0.0005473475653923554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":3.141212974645643e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001274169753045263,"output_power":0.00012204318248508215,"reference_active":true,"delta_db":-0.1871376725217449},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860144772024442e-06,"output_power":3.634274462712629e-06,"reference_active":true,"delta_db":-0.26185870482088625},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1444270061935377e-07,"output_power":1.1989977160840513e-07,"reference_active":true,"delta_db":0.20230258094680112}],"active_band_count":3,"active_band_mae_db":0.2170996527631441}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":8.38950411228042e-05,"log_rms_mae_db":1.9053840371934017,"rms_envelope_nrmse":0.20554309899080367,"envelope_correlation":0.9698954295606622,"rms_gain_error_db":-0.017761337041709513,"reference_mean_power":0.00021157950445783094,"output_mean_power":0.00021071597496350323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.2662436125548275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5392735637289224e-05,"output_power":5.503482991344505e-05,"reference_active":true,"delta_db":-0.02815185219466327},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.790041823950811e-06,"output_power":6.483206323619168e-06,"reference_active":true,"delta_db":0.49108142114763503},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4111913003034656e-06,"output_power":2.085066917353882e-06,"reference_active":true,"delta_db":1.695341071551225}],"active_band_count":3,"active_band_mae_db":0.7381914482978411}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":7.722064559288107e-05,"log_rms_mae_db":1.3149752710566065,"rms_envelope_nrmse":0.13289613442638012,"envelope_correlation":0.986776696177971,"rms_gain_error_db":-0.01776168933393589,"reference_mean_power":0.00021280955774068177,"output_mean_power":0.00021194099077909912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":1.2078803728484275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.022425345543556e-05,"output_power":5.018362477579508e-05,"reference_active":true,"delta_db":-0.003514627068851653},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7667526770926583e-06,"output_power":3.2774126176221072e-06,"reference_active":true,"delta_db":0.7356078251730649},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.98389498284838e-08,"output_power":1.505735459388075e-07,"reference_active":true,"delta_db":1.784486737921461}],"active_band_count":3,"active_band_mae_db":0.8412030633877925}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.54681310079566e-05,"log_rms_mae_db":1.9692181804975655,"rms_envelope_nrmse":0.25227307357040346,"envelope_correlation":0.9641308853516551,"rms_gain_error_db":0.4930583014774733,"reference_mean_power":0.00010955764718773973,"output_mean_power":0.0001227293762813416},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":6.937978158746453e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.2971450195505062e-05,"output_power":2.6416077305505593e-05,"reference_active":true,"delta_db":0.6067991364493788},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1375769637538911e-05,"output_power":1.4483003025959534e-05,"reference_active":true,"delta_db":1.0487783252577936},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.447046226198746e-06,"output_power":3.537673831984224e-06,"reference_active":true,"delta_db":1.600756156273756}],"active_band_count":3,"active_band_mae_db":1.085444539326976}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83120,"uncovered_tail_samples":0,"metrics":{"frame_count":1032,"rms_floor":5.7621495731062106e-05,"log_rms_mae_db":1.290289656709159,"rms_envelope_nrmse":0.17679170248102355,"envelope_correlation":0.9824879352183853,"rms_gain_error_db":0.4930582542653827,"reference_mean_power":0.00011019456862153084,"output_mean_power":0.00012344287118832003},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1032,"frequency_resolution_hz":0.19379844961240308,"power_floor":6.599778128222779e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3073624841125744e-05,"output_power":2.7620465308001178e-05,"reference_active":true,"delta_db":0.781151635765164},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.235055585786142e-06,"output_power":7.996691614059138e-06,"reference_active":true,"delta_db":1.080700185393229},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9515593840599056e-07,"output_power":2.8733447391386654e-07,"reference_active":true,"delta_db":1.680059640716549}],"active_band_count":3,"active_band_mae_db":1.1806371539583138}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},{"case":"librispeech_5639-40744-0013","method":"hnm","status":"measured","mos":{"status":"measured","value":2.786190986633301,"predicted_mos":2.786190986633301,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"0e609a2823383d54ee50d2bc1981c62381348487d90fb5f60dd61297af58d58a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.986663759998919},"started_at":"2026-10-13T04:39:36.784932+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4304000615084078e-05,"native_input_samples":96000,"native_input_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","output_sha256":"4ce05c2de503ed24c1e49598365b5e112d9ca9b97a905f97c561804c2d24fc2a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1291076470006374,"synthesis_s":0.4981233179996707,"total_s":1.627230965000308,"rtf":0.2712051608333847,"output_peak":0.2944960879581819,"output_rms":0.024666122338574865,"finished_at":"2026-10-13T04:39:38.418334+00:00","evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"4ce05c2de503ed24c1e49598365b5e112d9ca9b97a905f97c561804c2d24fc2a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.669713794658225,"spectral_convergence":0.24548293302304824,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.539561740836287,"spectral_convergence":0.24418220760357506,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.373834291650238,"spectral_convergence":0.24768317269767917,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00011387345841080127,"delta_db":0.2404894393708598,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.00025028002279460114,"delta_db":-0.18993837223079624,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.0001516315840692547,"delta_db":-0.1937047731616316,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":6.308226196576369e-05,"delta_db":-0.09622575342312882,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.7067220593124853e-05,"delta_db":0.29398083813441395,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":1.248304338799631e-05,"delta_db":0.056200590601806336,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17842329448710612,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08556873475659588,"pitch":null,"output_rms_dbfs":-32.157982376499405,"output_peak":0.2944960879581819,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.8610366090482495},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.669713794658225,"spectral_convergence":0.24548293302304824,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.539561740836287,"spectral_convergence":0.24418220760357506,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.373834291650238,"spectral_convergence":0.24768317269767917,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00011387345841080127,"delta_db":0.2404894393708598,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.00025028002279460114,"delta_db":-0.18993837223079624,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.0001516315840692547,"delta_db":-0.1937047731616316,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":6.308226196576369e-05,"delta_db":-0.09622575342312882,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.7067220593124853e-05,"delta_db":0.29398083813441395,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":1.248304338799631e-05,"delta_db":0.056200590601806336,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17842329448710612,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08556873475659588,"pitch":null,"output_rms_dbfs":-32.157982376499405,"output_peak":0.2944960879581819,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.8610366090482495},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8932494776187272,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024910320651486587,"rms_dbfs":-32.072413641742806,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.23336111718971,"mean":-9.009043375651042e-07,"zero_energy":false},"output":{"rms":0.024666122338574865,"rms_dbfs":-32.157982376499405,"peak":0.2944960879581819,"peak_dbfs":-10.618409398825131,"crest_factor_db":21.53957297767427,"mean":7.716144093176362e-05,"zero_energy":false},"rms_difference_db":-0.08556873475659899},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":3.862138958069595e-07,"output_rms_dbfs":-64.13172104550108,"output_zero_energy":false,"output_peak":0.010193220811061622,"output_energy_fraction":0.00014282645309794112,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":141,"both_estimated_voiced_frames":111,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9487179487179487,"absolute_difference_cents_median":1.4120405770753048,"absolute_difference_cents_p95":10.144245192052033,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001113800968506657,"log_rms_mae_db":0.9013270388345159,"rms_envelope_nrmse":0.08054769388026038,"envelope_correlation":0.9940456415128456,"rms_gain_error_db":-0.08557166415140026,"reference_mean_power":0.0006210401213146882,"output_mean_power":0.0006089231587330538},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.9949527250696107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014723003466329858,"output_power":0.00014238359683206697,"reference_active":true,"delta_db":-0.14536454639028504},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0749733198322775e-05,"output_power":2.9861841444910552e-05,"reference_active":true,"delta_db":-0.127247667546057},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.678489010984287e-06,"output_power":5.914230958712328e-06,"reference_active":true,"delta_db":0.17665490633703038}],"active_band_count":3,"active_band_mae_db":0.14975570675779082}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":0.4990853920582688,"rms_envelope_nrmse":0.04694536665176626,"envelope_correlation":0.9978612696402592,"rms_gain_error_db":-0.08556032387167323,"reference_mean_power":0.0006241534506116662,"output_mean_power":0.000611977342610869},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.688739338888854e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013414910111714758,"output_power":0.000129547803779798,"reference_active":true,"delta_db":-0.1515771217078601},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.37620762285614e-05,"output_power":1.377861600195729e-05,"reference_active":true,"delta_db":0.005216378093347621},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.14998789435472e-07,"output_power":4.326786274694324e-07,"reference_active":true,"delta_db":0.18118613518186805}],"active_band_count":3,"active_band_mae_db":0.11265987832769193}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":1.8203508048555153,"rms_envelope_nrmse":0.11289238194486058,"envelope_correlation":0.9911431253116508,"rms_gain_error_db":-0.22485122305436334,"reference_mean_power":0.000315142987429675,"output_mean_power":0.00029924198389016905},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.94905977338963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010605229421605894,"output_power":0.00010216969923550311,"reference_active":true,"delta_db":-0.161979530744632},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.109806584151368e-05,"output_power":1.9482091115553065e-05,"reference_active":true,"delta_db":-0.34607073152889917},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3612767263836796e-06,"output_power":2.9965731506065455e-06,"reference_active":true,"delta_db":-0.49879384526208764}],"active_band_count":3,"active_band_mae_db":0.335614702511873}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":1.3282751791232654,"rms_envelope_nrmse":0.07174675449725061,"envelope_correlation":0.9964547736848843,"rms_gain_error_db":-0.2248511664888175,"reference_mean_power":0.00031672629889435445,"output_mean_power":0.0003007454109602853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7988521248624547e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.967318213226487e-05,"output_power":9.534234313836308e-05,"reference_active":true,"delta_db":-0.1929250265499735},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.491202888187961e-06,"output_power":9.2611028463149e-06,"reference_active":true,"delta_db":-0.10658550017124054},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5805315595038676e-07,"output_power":3.115903975671935e-07,"reference_active":true,"delta_db":-0.603634406931101}],"active_band_count":3,"active_band_mae_db":0.30104831121743836}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":1.8537957846795488,"rms_envelope_nrmse":0.15712643942149263,"envelope_correlation":0.9826296753964496,"rms_gain_error_db":-0.12947045618553266,"reference_mean_power":6.366451195739098e-05,"output_mean_power":6.179457768967526e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5126235269128304e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.399199855439123e-05,"output_power":1.1662431890780824e-05,"reference_active":true,"delta_db":-0.7909063114148419},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.395771779201436e-06,"output_power":2.275254581631116e-06,"reference_active":true,"delta_db":-0.22415447301881233},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.685934249147904e-07,"output_power":6.468924458922541e-07,"reference_active":true,"delta_db":-0.7486458520371269}],"active_band_count":3,"active_band_mae_db":0.5879022121569271}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":1.1515579670680827,"rms_envelope_nrmse":0.09699928421795066,"envelope_correlation":0.9930289365500823,"rms_gain_error_db":-0.12942961458841776,"reference_mean_power":6.398182095924489e-05,"output_mean_power":6.210315081316306e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.275432891493173e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2825234350232915e-05,"output_power":1.066097225921446e-05,"reference_active":true,"delta_db":-0.8026849626141266},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0681397696196026e-06,"output_power":1.0443065755984537e-06,"reference_active":true,"delta_db":-0.09800072680963716},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7830475792130745e-08,"output_power":5.156370503572783e-08,"reference_active":true,"delta_db":-0.4981265000848401}],"active_band_count":3,"active_band_mae_db":0.46627072983620127}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":1.8973828253828533,"rms_envelope_nrmse":0.2025998662232813,"envelope_correlation":0.9732986194430236,"rms_gain_error_db":0.2787024001612432,"reference_mean_power":1.1507041647425961e-05,"output_mean_power":1.2269699457681e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.59330844916253e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.586447862335723e-06,"output_power":5.280818275892233e-06,"reference_active":true,"delta_db":0.6122476184166047},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.405946680666402e-07,"output_power":1.4592063307397791e-06,"reference_active":true,"delta_db":2.395300742401587},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329768801508534e-07,"output_power":5.271456837953396e-07,"reference_active":true,"delta_db":1.995165750800453}],"active_band_count":3,"active_band_mae_db":1.667571370539548}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":1.2614290814632447,"rms_envelope_nrmse":0.13040304141569764,"envelope_correlation":0.9887968562731904,"rms_gain_error_db":0.27909121972555556,"reference_mean_power":1.1561945751119366e-05,"output_mean_power":1.232934625221992e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.107379849793497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.414350374295126e-06,"output_power":5.341016340458461e-06,"reference_active":true,"delta_db":0.8275710579864594},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6495486661033e-07,"output_power":8.141883557479868e-07,"reference_active":true,"delta_db":2.43314089187694},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1590180837860975e-08,"output_power":4.114078497921958e-08,"reference_active":true,"delta_db":2.8001629394415364}],"active_band_count":3,"active_band_mae_db":2.020291629768312}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0013","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.2454609870910645,"predicted_mos":3.2454609870910645,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"991e5e08eb569459b1dfd58c771c0d87e42d14fce4da161bf527704075d53f7a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.623380470000484},"started_at":"2026-10-13T04:42:37.649967+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.467700051056454e-05,"native_input_samples":96000,"native_input_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","output_sha256":"256ba4996455bda3b970ce3d5bf72a06b3539fd3f9c2fe4cb0cc2ab696982e32","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3847658220001904,"synthesis_s":0.0827511620000223,"total_s":1.4675169840002127,"rtf":0.24458616400003544,"output_peak":0.3689477954403849,"output_rms":0.027544440994255764,"finished_at":"2026-10-13T04:42:39.124215+00:00","evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"256ba4996455bda3b970ce3d5bf72a06b3539fd3f9c2fe4cb0cc2ab696982e32","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.445474931565582,"spectral_convergence":0.6690943746989142,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.045577152719057,"spectral_convergence":0.26777549100633286,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.674077837476051,"spectral_convergence":0.28931648731359144,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00010769474933985844,"delta_db":-0.0017903694223437604,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.00032488301910195644,"delta_db":0.9430698926877062,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.00020699208241182197,"delta_db":1.1579358440123875,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":8.778733600337685e-05,"delta_db":1.3390203716573577,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.909598614764167e-05,"delta_db":0.7817737301003793,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":1.2243056681382069e-05,"delta_db":-0.0281056081592536,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7086159693399048,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8730929003270973,"pitch":null,"output_rms_dbfs":-31.199320741415704,"output_peak":0.3689477954403849,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.72170997392023},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.445474931565582,"spectral_convergence":0.6690943746989142,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.045577152719057,"spectral_convergence":0.26777549100633286,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.674077837476051,"spectral_convergence":0.28931648731359144,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00010769474933985844,"delta_db":-0.0017903694223437604,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.00032488301910195644,"delta_db":0.9430698926877062,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.00020699208241182197,"delta_db":1.1579358440123875,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":8.778733600337685e-05,"delta_db":1.3390203716573577,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.909598614764167e-05,"delta_db":0.7817737301003793,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":1.2243056681382069e-05,"delta_db":-0.0281056081592536,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7086159693399048,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8730929003270973,"pitch":null,"output_rms_dbfs":-31.199320741415704,"output_peak":0.3689477954403849,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.72170997392023},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9334566718891966,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024910320651486587,"rms_dbfs":-32.072413641742806,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.23336111718971,"mean":-9.009043375651042e-07,"zero_energy":false},"output":{"rms":0.027544440994255764,"rms_dbfs":-31.199320741415704,"peak":0.3689477954403849,"peak_dbfs":-8.660701606688843,"crest_factor_db":22.538619134726865,"mean":8.248182613298666e-09,"zero_energy":false},"rms_difference_db":0.8730929003271015},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":6.004856827242225e-07,"output_rms_dbfs":-62.21497342945491,"output_zero_energy":false,"output_peak":0.02030555392163082,"output_energy_fraction":0.000178080862045223,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":120,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.7948717948717948,"absolute_difference_cents_median":6.5945873846306124,"absolute_difference_cents_p95":22.664801248177046,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001113800968506657,"log_rms_mae_db":1.4866295144778374,"rms_envelope_nrmse":0.1817173983105851,"envelope_correlation":0.981951096854631,"rms_gain_error_db":0.8730533144894576,"reference_mean_power":0.0006210401213146882,"output_mean_power":0.0007593202629142115},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.9949527250696107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014723003466329858,"output_power":0.00018561665316544864,"reference_active":true,"delta_db":1.006205233522826},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0749733198322775e-05,"output_power":3.813286256493634e-05,"reference_active":true,"delta_db":0.9345805624881043},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.678489010984287e-06,"output_power":5.3811859462198295e-06,"reference_active":true,"delta_db":-0.23354790375751014}],"active_band_count":3,"active_band_mae_db":0.7247778999228135}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":0.9724368406964478,"rms_envelope_nrmse":0.12717976684871354,"envelope_correlation":0.9956440095018453,"rms_gain_error_db":0.8730577328856832,"reference_mean_power":0.0006241534506116662,"output_mean_power":0.0007631275792301114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.688739338888854e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013414910111714758,"output_power":0.00017055110972797176,"reference_active":true,"delta_db":1.0426678261118072},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.37620762285614e-05,"output_power":1.778436324922716e-05,"reference_active":true,"delta_db":1.1135436124695994},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.14998789435472e-07,"output_power":3.809025084899005e-07,"reference_active":true,"delta_db":-0.372329970822548}],"active_band_count":3,"active_band_mae_db":0.8428471364679848}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":1.9443064366184912,"rms_envelope_nrmse":0.21087465062410674,"envelope_correlation":0.9841473494123752,"rms_gain_error_db":1.126422308459393,"reference_mean_power":0.000315142987429675,"output_mean_power":0.0004084603249210502},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.94905977338963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010605229421605894,"output_power":0.0001429758435534782,"reference_active":true,"delta_db":1.297425995589803},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.109806584151368e-05,"output_power":3.082872644173718e-05,"reference_active":true,"delta_db":1.6471294074611953},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3612767263836796e-06,"output_power":3.056584481172934e-06,"reference_active":true,"delta_db":-0.41267864643051266}],"active_band_count":3,"active_band_mae_db":1.119078016493837}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":1.3423899927185636,"rms_envelope_nrmse":0.16327528580067693,"envelope_correlation":0.994971690192084,"rms_gain_error_db":1.1264253260337047,"reference_mean_power":0.00031672629889435445,"output_mean_power":0.0004105127577225929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7988521248624547e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.967318213226487e-05,"output_power":0.00013485241738338052,"reference_active":true,"delta_db":1.3128041246173376},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.491202888187961e-06,"output_power":1.4298671088293474e-05,"reference_active":true,"delta_db":1.7797441904874294},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5805315595038676e-07,"output_power":2.6153460905027223e-07,"reference_active":true,"delta_db":-1.3641833854928844}],"active_band_count":3,"active_band_mae_db":1.4855772335325506}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":1.9053770628422178,"rms_envelope_nrmse":0.2647250402514183,"envelope_correlation":0.9752132686732565,"rms_gain_error_db":1.3326737029552214,"reference_mean_power":6.366451195739098e-05,"output_mean_power":8.652961760841757e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5126235269128304e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.399199855439123e-05,"output_power":1.7017316259957164e-05,"reference_active":true,"delta_db":0.850113185196278},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.395771779201436e-06,"output_power":2.645124357677599e-06,"reference_active":true,"delta_db":0.4300064986765878},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.685934249147904e-07,"output_power":7.756169371137619e-07,"reference_active":true,"delta_db":0.039506192944146486}],"active_band_count":3,"active_band_mae_db":0.43987529227233746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":1.3095550028506062,"rms_envelope_nrmse":0.21154203376826775,"envelope_correlation":0.9897987800078266,"rms_gain_error_db":1.332724076926202,"reference_mean_power":6.398182095924489e-05,"output_mean_power":8.696189678753094e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.275432891493173e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2825234350232915e-05,"output_power":1.5524384460004694e-05,"reference_active":true,"delta_db":0.8294907992467195},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0681397696196026e-06,"output_power":1.2368912053380516e-06,"reference_active":true,"delta_db":0.6370341629039976},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7830475792130745e-08,"output_power":5.210556618963782e-08,"reference_active":true,"delta_db":-0.45272646025079794}],"active_band_count":3,"active_band_mae_db":0.6397504741338383}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":2.023863851023313,"rms_envelope_nrmse":0.21338170439382356,"envelope_correlation":0.9727318238790522,"rms_gain_error_db":0.5181283043392161,"reference_mean_power":1.1507041647425961e-05,"output_mean_power":1.2965119241460163e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.59330844916253e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.586447862335723e-06,"output_power":5.2251142897690515e-06,"reference_active":true,"delta_db":0.5661933336655206},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.405946680666402e-07,"output_power":1.0789183861546173e-06,"reference_active":true,"delta_db":1.0840196316915236},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329768801508534e-07,"output_power":3.3058755565242106e-07,"reference_active":true,"delta_db":-0.03127578495098734}],"active_band_count":3,"active_band_mae_db":0.5604962501026772}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":1.2506295403256702,"rms_envelope_nrmse":0.1263165748883307,"envelope_correlation":0.9916000561765352,"rms_gain_error_db":0.5175627280736599,"reference_mean_power":1.1561945751119366e-05,"output_mean_power":1.302528396433809e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.107379849793497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.414350374295126e-06,"output_power":5.048678706846823e-06,"reference_active":true,"delta_db":0.5831093261798173},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6495486661033e-07,"output_power":5.933866828014274e-07,"reference_active":true,"delta_db":1.0592699758559945},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1590180837860975e-08,"output_power":2.094002102620068e-08,"reference_active":true,"delta_db":-0.13279166545114757}],"active_band_count":3,"active_band_mae_db":0.5917236558289865}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0013","method":"facodec","status":"measured","mos":{"status":"measured","value":3.6063148975372314,"predicted_mos":3.6063148975372314,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7a99f64075117b73dcbfebd5576e456c00ebb952fd88f8ff2e26f954508f40db","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6989120339985675},"started_at":"2026-10-13T04:54:51.791112+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4762999601080082e-05,"native_input_samples":96000,"native_input_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","output_sha256":"6be4e384cc8e681c6c6f2e597b0264019de5c486a2a272f67df87ab3eca5ee7d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4333463570001186,"synthesis_s":5.868915856000058,"total_s":7.302262213000176,"rtf":1.217043702166696,"output_peak":0.26670724153518677,"output_rms":0.024430212028393534,"finished_at":"2026-10-13T04:54:59.099422+00:00","evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"6be4e384cc8e681c6c6f2e597b0264019de5c486a2a272f67df87ab3eca5ee7d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5764872144577895,"spectral_convergence":0.24416880434663396,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.80346724882901,"spectral_convergence":0.24860926059462285,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.106831236739139,"spectral_convergence":0.26278691597040277,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00011435181089447505,"delta_db":0.25869479839282783,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.0002550489296089291,"delta_db":-0.10796517861524425,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.0001442000301213023,"delta_db":-0.41194798212299705,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":5.3641545798163775e-05,"delta_db":-0.8002854844557741,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.4235823262173018e-05,"delta_db":-0.49382130351793263,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":1.5357120067220843e-05,"delta_db":0.9560935878995879,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5048013891673939,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1690416341288395,"pitch":null,"output_rms_dbfs":-32.241455275871644,"output_peak":0.26670724153518677,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.828928566675313},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5764872144577895,"spectral_convergence":0.24416880434663396,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.80346724882901,"spectral_convergence":0.24860926059462285,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.106831236739139,"spectral_convergence":0.26278691597040277,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00011435181089447505,"delta_db":0.25869479839282783,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.0002550489296089291,"delta_db":-0.10796517861524425,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.0001442000301213023,"delta_db":-0.41194798212299705,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":5.3641545798163775e-05,"delta_db":-0.8002854844557741,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.4235823262173018e-05,"delta_db":-0.49382130351793263,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":1.5357120067220843e-05,"delta_db":0.9560935878995879,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5048013891673939,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1690416341288395,"pitch":null,"output_rms_dbfs":-32.241455275871644,"output_peak":0.26670724153518677,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.828928566675313},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9086833330446105,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024910320651486587,"rms_dbfs":-32.072413641742806,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.23336111718971,"mean":-9.009043375651042e-07,"zero_energy":false},"output":{"rms":0.024430212028393534,"rms_dbfs":-32.241455275871644,"peak":0.26670724153518677,"peak_dbfs":-11.479303846976654,"crest_factor_db":20.76215142889499,"mean":1.1425309390321085e-05,"zero_energy":false},"rms_difference_db":-0.16904163412883833},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":1.759388780009185e-07,"output_rms_dbfs":-67.54638181943295,"output_zero_energy":false,"output_peak":0.006650588475167751,"output_energy_fraction":6.632692506578485e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":129,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8803418803418803,"absolute_difference_cents_median":4.8705143323409645,"absolute_difference_cents_p95":22.688250026560663,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001113800968506657,"log_rms_mae_db":1.2122886676990474,"rms_envelope_nrmse":0.11641286726661043,"envelope_correlation":0.9879648153511088,"rms_gain_error_db":-0.16904813203578078,"reference_mean_power":0.0006210401213146882,"output_mean_power":0.0005973307120548953},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.9949527250696107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014723003466329858,"output_power":0.00013006501970661629,"reference_active":true,"delta_db":-0.5383590325500782},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0749733198322775e-05,"output_power":2.8668252814276245e-05,"reference_active":true,"delta_db":-0.3044012625877049},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.678489010984287e-06,"output_power":5.8163761308250714e-06,"reference_active":true,"delta_db":0.104196938474853}],"active_band_count":3,"active_band_mae_db":0.315652411204212}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":1.0443472694350004,"rms_envelope_nrmse":0.10436110289627458,"envelope_correlation":0.9896457045670987,"rms_gain_error_db":-0.16904167223625044,"reference_mean_power":0.0006241534506116662,"output_mean_power":0.0006003260769157137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.688739338888854e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013414910111714758,"output_power":0.00011777434503719291,"reference_active":true,"delta_db":-0.5653706924031809},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.37620762285614e-05,"output_power":1.2367902963305771e-05,"reference_active":true,"delta_db":-0.463878899086012},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.14998789435472e-07,"output_power":3.745077114421624e-07,"reference_active":true,"delta_db":-0.4458606523152706}],"active_band_count":3,"active_band_mae_db":0.4917034146014878}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":1.7747005170535837,"rms_envelope_nrmse":0.11303917515969143,"envelope_correlation":0.9915283024781398,"rms_gain_error_db":-0.277848077840825,"reference_mean_power":0.000315142987429675,"output_mean_power":0.00029561253101271855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.94905977338963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010605229421605894,"output_power":0.00010270582615970487,"reference_active":true,"delta_db":-0.1392498761592747},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.109806584151368e-05,"output_power":1.9509449971065463e-05,"reference_active":true,"delta_db":-0.33997617789033147},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3612767263836796e-06,"output_power":3.6725976978809606e-06,"reference_active":true,"delta_db":0.3846908921795983}],"active_band_count":3,"active_band_mae_db":0.28797231540973484}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":1.351679436396219,"rms_envelope_nrmse":0.09628536571018184,"envelope_correlation":0.9937400951040158,"rms_gain_error_db":-0.27784310955870783,"reference_mean_power":0.00031672629889435445,"output_mean_power":0.0002970980592900875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7988521248624547e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.967318213226487e-05,"output_power":9.594172440097983e-05,"reference_active":true,"delta_db":-0.16570803663378197},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.491202888187961e-06,"output_power":8.572494590918322e-06,"reference_active":true,"delta_db":-0.44214037413304924},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5805315595038676e-07,"output_power":3.6346667686138314e-07,"reference_active":true,"delta_db":0.0651709425267174}],"active_band_count":3,"active_band_mae_db":0.22433978443118285}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":1.563615408394785,"rms_envelope_nrmse":0.16737858062147246,"envelope_correlation":0.9830458229885354,"rms_gain_error_db":-0.6745215316597969,"reference_mean_power":6.366451195739098e-05,"output_mean_power":5.450611912199143e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5126235269128304e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.399199855439123e-05,"output_power":1.1573828364998361e-05,"reference_active":true,"delta_db":-0.8240271392972891},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.395771779201436e-06,"output_power":2.1251219255951745e-06,"reference_active":true,"delta_db":-0.520615927939353},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.685934249147904e-07,"output_power":7.936906647451306e-07,"reference_active":true,"delta_db":0.13954607351940346}],"active_band_count":3,"active_band_mae_db":0.4947297135853485}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":1.2210763602679737,"rms_envelope_nrmse":0.1484485118640623,"envelope_correlation":0.9867496447346509,"rms_gain_error_db":-0.6745841985233201,"reference_mean_power":6.398182095924489e-05,"output_mean_power":5.477699154939505e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.275432891493173e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2825234350232915e-05,"output_power":1.0723977505270403e-05,"reference_active":true,"delta_db":-0.7770941520976211},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0681397696196026e-06,"output_power":8.562715802265463e-07,"reference_active":true,"delta_db":-0.9601655531309219},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7830475792130745e-08,"output_power":5.111506040644366e-08,"reference_active":true,"delta_db":-0.5360788704650498}],"active_band_count":3,"active_band_mae_db":0.7577795252311975}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":2.3000518145078948,"rms_envelope_nrmse":0.2531197069334941,"envelope_correlation":0.9548354726864555,"rms_gain_error_db":-0.26226757991812694,"reference_mean_power":1.1507041647425961e-05,"output_mean_power":1.0832705261847993e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.59330844916253e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.586447862335723e-06,"output_power":4.48059686508288e-06,"reference_active":true,"delta_db":-0.10140590231281685},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.405946680666402e-07,"output_power":7.634735168120191e-07,"reference_active":true,"delta_db":-0.41792653976980243},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329768801508534e-07,"output_power":3.4741314551042215e-07,"reference_active":true,"delta_db":0.18432167533179514}],"active_band_count":3,"active_band_mae_db":0.23455137247147148}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":2.048227400832526,"rms_envelope_nrmse":0.20968279672465723,"envelope_correlation":0.96735885670902,"rms_gain_error_db":-0.2619292836076765,"reference_mean_power":1.1561945751119366e-05,"output_mean_power":1.0885239751072333e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.107379849793497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.414350374295126e-06,"output_power":4.364474458547729e-06,"reference_active":true,"delta_db":-0.04934844575065228},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6495486661033e-07,"output_power":3.859291725419443e-07,"reference_active":true,"delta_db":-0.8090318945693958},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1590180837860975e-08,"output_power":1.8842853322305987e-08,"reference_active":true,"delta_db":-0.591096124948362}],"active_band_count":3,"active_band_mae_db":0.48315882175613667}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0013","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.2942094802856445,"predicted_mos":3.2942094802856445,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"e53aa5a78a35abb4e2b2fe27df883885e8bf7d80346f37fa4196e491066c9b07","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.468239113000891},"started_at":"2026-10-13T07:20:07.769285+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0066833790006057825,"native_input_samples":264600,"native_input_sha256":"60dabd3d6d3f6cfeb5fa1928a5f9ca9ddea39521092eb99209e36eb339b661ea","output_sha256":"758ee64e5e97ff0f2bf4cff69511fb58e11199756ceccf69544f4bcbb0cdb46f","output_samples":264704,"length_delta_samples":104,"analysis_s":0.182860029000949,"synthesis_s":37.12494391299879,"total_s":37.30780394199974,"rtf":6.217967323666623,"output_peak":0.30116087198257446,"output_rms":0.023023977053460473,"finished_at":"2026-10-13T07:20:45.107382+00:00","evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"758ee64e5e97ff0f2bf4cff69511fb58e11199756ceccf69544f4bcbb0cdb46f","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-32.756397048572,"output_peak":0.3019450598995159},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.997076753727967,"spectral_convergence":0.44426624017314936,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.653460586946757,"spectral_convergence":0.36842157447024393,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.56245572145042,"spectral_convergence":0.39750252494165933,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":8.666620420856401e-05,"delta_db":-0.9452379089255063,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.00025504964192381145,"delta_db":-0.10795304941355441,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.00012208490375385068,"delta_db":-1.134981841543739,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":4.638248661053433e-05,"delta_db":-1.4317580506977212,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.2156957436255222e-05,"delta_db":-1.1793982179339428,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":7.970705514617003e-06,"delta_db":-1.89203656957207,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1152276063477558,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6822808258737947,"pitch":null,"output_rms_dbfs":-32.7546944676166,"output_peak":0.3019450598995159,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.737664354041714},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7589304582773917,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024910320651486587,"rms_dbfs":-32.072413641742806,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.23336111718971,"mean":-9.009043375651042e-07,"zero_energy":false},"output":{"rms":0.023028480181020036,"rms_dbfs":-32.7546944676166,"peak":0.3019450598995159,"peak_dbfs":-10.401441429139847,"crest_factor_db":22.353253038476755,"mean":-1.2535443151005212e-05,"zero_energy":false},"rms_difference_db":-0.6822808258737965},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":2.2342869942464332e-07,"output_rms_dbfs":-66.50861042495157,"output_zero_energy":false,"output_peak":0.014774420942124498,"output_energy_fraction":9.479619865046535e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":142,"both_estimated_voiced_frames":99,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":43,"paired_voiced_fraction":0.8461538461538461,"absolute_difference_cents_median":5.5238549473586005,"absolute_difference_cents_p95":40.174169441133834,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001113800968506657,"log_rms_mae_db":1.7399695439479568,"rms_envelope_nrmse":0.12908861228748905,"envelope_correlation":0.9892278263174863,"rms_gain_error_db":-0.6822925720085765,"reference_mean_power":0.0006210401213146882,"output_mean_power":0.0005307504863362517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.9949527250696107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014723003466329858,"output_power":0.00014004202897468493,"reference_active":true,"delta_db":-0.217380199966209},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0749733198322775e-05,"output_power":2.6211941956946102e-05,"reference_active":true,"delta_db":-0.6934215435429882},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.678489010984287e-06,"output_power":4.778950569793964e-06,"reference_active":true,"delta_db":-0.7490025117600583}],"active_band_count":3,"active_band_mae_db":0.5532680850897519}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":1.4527737772907396,"rms_envelope_nrmse":0.10432246535509025,"envelope_correlation":0.9944441919313627,"rms_gain_error_db":-0.6822624525494264,"reference_mean_power":0.0006241534506116662,"output_mean_power":0.0005334148850314168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.688739338888854e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013414910111714758,"output_power":0.00012850706176570085,"reference_active":true,"delta_db":-0.18660773145450765},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.37620762285614e-05,"output_power":1.1563981660653162e-05,"reference_active":true,"delta_db":-0.755765648369783},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.14998789435472e-07,"output_power":3.358129371016177e-07,"reference_active":true,"delta_db":-0.9194940666273771}],"active_band_count":3,"active_band_mae_db":0.6206224821505559}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":2.666872823194986,"rms_envelope_nrmse":0.15599298391127922,"envelope_correlation":0.9839164445204562,"rms_gain_error_db":-0.4970512517179076,"reference_mean_power":0.000315142987429675,"output_mean_power":0.00028106225249158303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.94905977338963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010605229421605894,"output_power":0.00010242328534673375,"reference_active":true,"delta_db":-0.1512136576706396},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.109806584151368e-05,"output_power":1.6634664733916915e-05,"reference_active":true,"delta_db":-1.0322859108025502},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3612767263836796e-06,"output_power":3.295535973604386e-06,"reference_active":true,"delta_db":-0.08578211734054517}],"active_band_count":3,"active_band_mae_db":0.4230938952712449}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":2.1004861408762334,"rms_envelope_nrmse":0.12157255890362612,"envelope_correlation":0.9904763095492244,"rms_gain_error_db":-0.49704467726203905,"reference_mean_power":0.00031672629889435445,"output_mean_power":0.00028247476638169927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7988521248624547e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.967318213226487e-05,"output_power":9.563802212564521e-05,"reference_active":true,"delta_db":-0.17947737657675072},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.491202888187961e-06,"output_power":7.3152395323507865e-06,"reference_active":true,"delta_db":-1.1309270580861386},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5805315595038676e-07,"output_power":3.289164165996081e-07,"reference_active":true,"delta_db":-0.36861955871010216}],"active_band_count":3,"active_band_mae_db":0.5596746644576639}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":3.8716618975344645,"rms_envelope_nrmse":0.31730118165488946,"envelope_correlation":0.9447411093501774,"rms_gain_error_db":-1.8302253187985282,"reference_mean_power":6.366451195739098e-05,"output_mean_power":4.177100095097076e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5126235269128304e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.399199855439123e-05,"output_power":8.674411087960817e-06,"reference_active":true,"delta_db":-2.076397517032805},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.395771779201436e-06,"output_power":2.287754331859628e-06,"reference_active":true,"delta_db":-0.20036058512033456},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.685934249147904e-07,"output_power":5.50634642974729e-07,"reference_active":true,"delta_db":-1.4483313349232816}],"active_band_count":3,"active_band_mae_db":1.2416964790254739}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":3.693318923949208,"rms_envelope_nrmse":0.29627199369636154,"envelope_correlation":0.9528033647417858,"rms_gain_error_db":-1.830270972138809,"reference_mean_power":6.398182095924489e-05,"output_mean_power":4.197874966311057e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.275432891493173e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2825234350232915e-05,"output_power":8.228457401757253e-06,"reference_active":true,"delta_db":-1.9274688437692153},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0681397696196026e-06,"output_power":1.1364615945226207e-06,"reference_active":true,"delta_db":0.2692667852215436},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7830475792130745e-08,"output_power":3.597093524055636e-08,"reference_active":true,"delta_db":-2.062050358388091}],"active_band_count":3,"active_band_mae_db":1.4195953291262835}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":5.5988425686848,"rms_envelope_nrmse":0.2695237636434385,"envelope_correlation":0.9567355526678998,"rms_gain_error_db":-1.1557189088831583,"reference_mean_power":1.1507041647425961e-05,"output_mean_power":8.818440603106515e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.59330844916253e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.586447862335723e-06,"output_power":3.3023477208176875e-06,"reference_active":true,"delta_db":-1.426536605268161},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.405946680666402e-07,"output_power":4.3376290749545503e-07,"reference_active":true,"delta_db":-2.873342196221333},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329768801508534e-07,"output_power":1.850449321785942e-07,"reference_active":true,"delta_db":-2.5513688425743055}],"active_band_count":3,"active_band_mae_db":2.283749214687933}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":5.585714152424788,"rms_envelope_nrmse":0.22554411538180738,"envelope_correlation":0.9722651441831635,"rms_gain_error_db":-1.1547276953872951,"reference_mean_power":1.1561945751119366e-05,"output_mean_power":8.862538968555215e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.107379849793497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.414350374295126e-06,"output_power":3.116557203743857e-06,"reference_active":true,"delta_db":-1.5119169818159384},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6495486661033e-07,"output_power":2.2510779317612668e-07,"reference_active":true,"delta_db":-3.150202673360236},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1590180837860975e-08,"output_power":1.0272227030850282e-08,"reference_active":true,"delta_db":-3.225916706124412}],"active_band_count":3,"active_band_mae_db":2.629345453766862}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0013","method":"golf","status":"measured","mos":{"status":"measured","value":3.142965078353882,"predicted_mos":3.142965078353882,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"a38bedd7d14490daefdc904d93b7a11337c1e474a2c8d28f60fc2f72e1d69b6a","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.617228488001274},"started_at":"2026-10-13T05:02:43.261265+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030498449996230192,"native_input_samples":144000,"native_input_sha256":"e92d1ee9674cb056f32676a0ef28ff32b1976e5c4a4a930f7f4dd37561267ddb","output_sha256":"2f1c98999e0237a261a18ccea709d541c03e3b77c4857367d0726aecaba14a05","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5493743690003612,"synthesis_s":0.1316737329998432,"total_s":0.6810481020002044,"rtf":0.11350801700003406,"output_peak":0.2620125412940979,"output_rms":0.0226993552261679,"finished_at":"2026-10-13T05:02:43.954854+00:00","analysis_stages":{"external_f0_s":0.08351640100045188},"evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"2f1c98999e0237a261a18ccea709d541c03e3b77c4857367d0726aecaba14a05","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-32.877098081846476,"output_peak":0.2621914320867216},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.270338400142849,"spectral_convergence":0.5482167649551407,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.226206837192091,"spectral_convergence":0.29102693262318474,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.5691787630187966,"spectral_convergence":0.29899854739030307,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00010791898913801197,"output_power":8.82889175969649e-05,"delta_db":-0.871916762120996,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002619024994739356,"output_power":0.00023216601604042716,"delta_db":-0.5234099417940186,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0001588117434581943,"output_power":0.00013314443857229417,"delta_db":-0.7655958279873397,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.460207798152873e-05,"output_power":4.586727810847051e-05,"delta_db":-1.4874351901889424,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.596706789752364e-05,"output_power":8.375989315842044e-06,"delta_db":-2.8018905736768236,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2342969985923176e-05,"output_power":7.73039131556267e-06,"delta_db":-2.0321819429216985,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4137383731149697,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.81182628434266,"pitch":null,"output_rms_dbfs":-32.877098081846476,"output_peak":0.2621914320867216,"reference_rms_dbfs":-32.06527179750382,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.688574666784579},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8490806798510234,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024930811216948345,"rms_dbfs":-32.06527179750382,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.226219272950722,"mean":-9.090952164740713e-07,"zero_energy":false},"output":{"rms":0.022706233306067334,"rms_dbfs":-32.877098081846476,"peak":0.2621914320867216,"peak_dbfs":-11.627630086500316,"crest_factor_db":21.249467995346162,"mean":0.000253985502310622,"zero_energy":false},"rms_difference_db":-0.8118262843426578},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":3.112275183269541e-07,"output_rms_dbfs":-65.06922010238958,"output_zero_energy":false,"output_peak":0.004272908924760431,"output_energy_fraction":0.00013604880894504135,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":165,"both_estimated_voiced_frames":112,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":53,"paired_voiced_fraction":0.9572649572649573,"absolute_difference_cents_median":3.9375692889070937,"absolute_difference_cents_p95":16.96090454133866,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001113800968506657,"log_rms_mae_db":2.742371286528702,"rms_envelope_nrmse":0.16634996457511572,"envelope_correlation":0.9816086289714713,"rms_gain_error_db":-0.8118291185027182,"reference_mean_power":0.0006220645586350807,"output_mean_power":0.0005160033804597506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.9988361550726657e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014769075238419378,"output_power":0.00012794677956028775,"reference_active":true,"delta_db":-0.6232394383057276},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.03091730862518e-05,"output_power":2.1853027850488672e-05,"reference_active":true,"delta_db":-1.4206246860532983},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.670313798166069e-06,"output_power":3.564452252824049e-06,"reference_active":true,"delta_db":-2.0161429209403696}],"active_band_count":3,"active_band_mae_db":1.353335681766465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":2.4874545170598403,"rms_envelope_nrmse":0.12977826945680584,"envelope_correlation":0.9917422687958393,"rms_gain_error_db":-0.811846537828955,"reference_mean_power":0.0006251979929448878,"output_mean_power":0.0005186004883120871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.692875856250024e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001343362376484674,"output_power":0.00011775122338208581,"reference_active":true,"delta_db":-0.5722775291057705},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3586869751951917e-05,"output_power":9.731243202992256e-06,"reference_active":true,"delta_db":-1.449510855682962},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1069923932456125e-07,"output_power":2.631542341576622e-07,"reference_active":true,"delta_db":-1.9331353625940961}],"active_band_count":3,"active_band_mae_db":1.318307915794276}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":2.434680686991046,"rms_envelope_nrmse":0.17643497051727297,"envelope_correlation":0.9809597792659646,"rms_gain_error_db":-0.6657441985955415,"reference_mean_power":0.0003156674005948115,"output_mean_power":0.0002708041491599996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.952073017749223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001062176103596908,"output_power":9.09237454144813e-05,"reference_active":true,"delta_db":-0.6751920933475102},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0892163461453977e-05,"output_power":1.5141231799680537e-05,"reference_active":true,"delta_db":-1.3982220693055831},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.350269902888054e-06,"output_power":2.0412742726209426e-06,"reference_active":true,"delta_db":-2.151784339419677}],"active_band_count":3,"active_band_mae_db":1.4083995006909233}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":2.1067763705892797,"rms_envelope_nrmse":0.13589169361549952,"envelope_correlation":0.9901089838734243,"rms_gain_error_db":-0.6657445511066571,"reference_mean_power":0.00031725753518382834,"output_mean_power":0.00027216826870493156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8019599938106461e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.97060338867444e-05,"output_power":8.555502855249228e-05,"reference_active":true,"delta_db":-0.6647590063641438},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.405782537010948e-06,"output_power":6.749874102153975e-06,"reference_active":true,"delta_db":-1.4409926108010127},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.534065762824191e-07,"output_power":1.9474660265729336e-07,"reference_active":true,"delta_db":-2.588047365420445}],"active_band_count":3,"active_band_mae_db":1.564599660861867}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":3.1998543033519464,"rms_envelope_nrmse":0.23045981324491507,"envelope_correlation":0.9801037099522708,"rms_gain_error_db":-1.571898030998156,"reference_mean_power":6.376708927527328e-05,"output_mean_power":4.440243533359774e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.5155783675301928e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4010956256524555e-05,"output_power":8.223319025153133e-06,"reference_active":true,"delta_db":-2.3142063804805932},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.37395555125899e-06,"output_power":1.4553438480281902e-06,"reference_active":true,"delta_db":-2.1250696879149484},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.697410766901598e-07,"output_power":3.91603254882176e-07,"reference_active":true,"delta_db":-2.9349837013265354}],"active_band_count":3,"active_band_mae_db":2.4580865899073587}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":3.049772152593706,"rms_envelope_nrmse":0.20120419413695914,"envelope_correlation":0.9901372065661274,"rms_gain_error_db":-1.5718973864724992,"reference_mean_power":6.408828706484192e-05,"output_mean_power":4.462609908495374e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.2787644375591154e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2827529766002323e-05,"output_power":7.730267487532558e-06,"reference_active":true,"delta_db":-2.1994850919282554},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0642106838647739e-06,"output_power":6.633990162905243e-07,"reference_active":true,"delta_db":-2.052527913983689},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7637652604815225e-08,"output_power":2.7029840327487617e-08,"reference_active":true,"delta_db":-3.2886280488003825}],"active_band_count":3,"active_band_mae_db":2.513547018237442}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":3.260811928037625,"rms_envelope_nrmse":0.2809534785219233,"envelope_correlation":0.9541571172500691,"rms_gain_error_db":-1.1805520442566169,"reference_mean_power":1.1522333474294859e-05,"output_mean_power":8.77981238903424e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.594690205591191e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.573106924359828e-06,"output_power":2.7848422944676743e-06,"reference_active":true,"delta_db":-2.15410749667475},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408152873310453e-07,"output_power":6.191895588219963e-07,"reference_active":true,"delta_db":-1.328769748619305},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3159049810941635e-07,"output_power":2.4417180980672667e-07,"reference_active":true,"delta_db":-1.3290655491931036}],"active_band_count":3,"active_band_mae_db":1.6039809314957196}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":3.303859186756274,"rms_envelope_nrmse":0.24282752432989904,"envelope_correlation":0.968142938796588,"rms_gain_error_db":-1.1805532063282793,"reference_mean_power":1.1580365810381607e-05,"output_mean_power":8.82402963192317e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.108607599298726e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.397340308988013e-06,"output_power":2.7942621611908094e-06,"reference_active":true,"delta_db":-1.9692292710957935},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6428856932851695e-07,"output_power":3.1002961620106547e-07,"reference_active":true,"delta_db":-1.7538480890581842},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1458421848197725e-08,"output_power":1.2062686977868708e-08,"reference_active":true,"delta_db":-2.5015372058278262}],"active_band_count":3,"active_band_mae_db":2.0748715219939347}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0013","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.621272563934326,"predicted_mos":3.621272563934326,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"d31811b5f9cf0226b9c701fbd8674f22e7615fee936fbd7b51d23d5298874a6e","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.091568798998196},"started_at":"2026-10-13T07:40:42.305390+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003853976999380393,"native_input_samples":144000,"native_input_sha256":"e92d1ee9674cb056f32676a0ef28ff32b1976e5c4a4a930f7f4dd37561267ddb","output_sha256":"bd8345fb5df15214fc75f3599befdc8f909e492ecdf347b64d017ce68fabe051","output_samples":144120,"length_delta_samples":120,"analysis_s":2.0841659600009734,"synthesis_s":4.680915207998623,"total_s":6.765081167999597,"rtf":1.1275135279999329,"output_peak":0.2879234552383423,"output_rms":0.0250431223211117,"finished_at":"2026-10-13T07:40:49.089680+00:00","analysis_stages":{"external_f0_s":1.2837999430012133},"evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"bd8345fb5df15214fc75f3599befdc8f909e492ecdf347b64d017ce68fabe051","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-32.02710361090935,"output_peak":0.2899713625842885},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.987931333408419,"spectral_convergence":0.522876895873958,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.098867926411078,"spectral_convergence":0.3010797087102744,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.2340250281104055,"spectral_convergence":0.27889571747779873,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00013610222331336685,"delta_db":1.0149165300997536,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.00025260324574297274,"delta_db":-0.14981096502156005,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.00015462855425042967,"delta_db":-0.10870453985301563,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":5.978919350481502e-05,"delta_db":-0.3290713767795209,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.557772943509463e-05,"delta_db":-0.1026056151005357,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":8.851668918399538e-06,"delta_db":-1.4367525944256252,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5236436035466685,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.048915079017279844,"pitch":null,"output_rms_dbfs":-32.02349856272552,"output_peak":0.2899713625842885,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.440274762643301},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8799118921649985,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024910320651486587,"rms_dbfs":-32.072413641742806,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.23336111718971,"mean":-9.009043375651042e-07,"zero_energy":false},"output":{"rms":0.02505100028272481,"rms_dbfs":-32.02349856272552,"peak":0.2899713625842885,"peak_dbfs":-10.752897813451119,"crest_factor_db":21.270600749274404,"mean":-4.0138945540033465e-05,"zero_energy":false},"rms_difference_db":0.04891507901728431},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":1.1191087780169626e-06,"output_rms_dbfs":-59.51127697746462,"output_zero_energy":false,"output_peak":0.03686478151300147,"output_energy_fraction":0.0004012404202754862,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":136,"both_estimated_voiced_frames":108,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9230769230769231,"absolute_difference_cents_median":6.8613789333975035,"absolute_difference_cents_p95":25.34985342382006,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001113800968506657,"log_rms_mae_db":2.253370499036109,"rms_envelope_nrmse":0.1648178985763457,"envelope_correlation":0.9751045388039682,"rms_gain_error_db":0.048920941383646444,"reference_mean_power":0.0006210401213146882,"output_mean_power":0.0006280753544769164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.9949527250696107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014723003466329858,"output_power":0.0001514941924964129,"reference_active":true,"delta_db":0.1239957021821706},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0749733198322775e-05,"output_power":2.9870470111389785e-05,"reference_active":true,"delta_db":-0.12599294220007792},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.678489010984287e-06,"output_power":4.8864142775330475e-06,"reference_active":true,"delta_db":-0.6524250533800195}],"active_band_count":3,"active_band_mae_db":0.30080456592075605}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":1.526090725934615,"rms_envelope_nrmse":0.09420458761630406,"envelope_correlation":0.9910673951017145,"rms_gain_error_db":0.04892579548157639,"reference_mean_power":0.0006241534506116662,"output_mean_power":0.0006312246575404319},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.688739338888854e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013414910111714758,"output_power":0.00013808629268861287,"reference_active":true,"delta_db":0.12562802951225113},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.37620762285614e-05,"output_power":1.3435000484028139e-05,"reference_active":true,"delta_db":-0.10446272671773557},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.14998789435472e-07,"output_power":3.709209010146723e-07,"reference_active":true,"delta_db":-0.487655237937915}],"active_band_count":3,"active_band_mae_db":0.23924866472263387}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":2.5944146912975805,"rms_envelope_nrmse":0.1927752128599043,"envelope_correlation":0.9733615542625633,"rms_gain_error_db":-0.09307368700259905,"reference_mean_power":0.000315142987429675,"output_mean_power":0.0003084610118922945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.94905977338963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010605229421605894,"output_power":0.00010122062880127149,"reference_active":true,"delta_db":-0.20251037093772223},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.109806584151368e-05,"output_power":1.9029020761229013e-05,"reference_active":true,"delta_db":-0.44826203368208994},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3612767263836796e-06,"output_power":3.3771759191263862e-06,"reference_active":true,"delta_db":0.02049415452380351}],"active_band_count":3,"active_band_mae_db":0.22375551971453858}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":1.9449527828948416,"rms_envelope_nrmse":0.11552171522789134,"envelope_correlation":0.989753770679177,"rms_gain_error_db":-0.09306910890165054,"reference_mean_power":0.00031672629889435445,"output_mean_power":0.00031001107920712714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7988521248624547e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.967318213226487e-05,"output_power":9.48104612292987e-05,"reference_active":true,"delta_db":-0.21722064304925862},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.491202888187961e-06,"output_power":8.79525936235905e-06,"reference_active":true,"delta_db":-0.3307260632805743},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5805315595038676e-07,"output_power":3.3674050254677e-07,"reference_active":true,"delta_db":-0.2665215034050788}],"active_band_count":3,"active_band_mae_db":0.27148940324497056}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":2.5691772909516115,"rms_envelope_nrmse":0.19890579990292004,"envelope_correlation":0.9720983200014254,"rms_gain_error_db":-0.2717471806462291,"reference_mean_power":6.366451195739098e-05,"output_mean_power":5.9802962178225e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5126235269128304e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.399199855439123e-05,"output_power":1.3458925556041369e-05,"reference_active":true,"delta_db":-0.16869360636031824},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.395771779201436e-06,"output_power":2.128008187938484e-06,"reference_active":true,"delta_db":-0.5147215021024515},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.685934249147904e-07,"output_power":7.265049224369218e-07,"reference_active":true,"delta_db":-0.24458103677803986}],"active_band_count":3,"active_band_mae_db":0.30933204841360323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":1.5852656805222722,"rms_envelope_nrmse":0.11401826292528253,"envelope_correlation":0.9907106034307087,"rms_gain_error_db":-0.271766317696448,"reference_mean_power":6.398182095924489e-05,"output_mean_power":6.010076007550315e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.275432891493173e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2825234350232915e-05,"output_power":1.228138800680917e-05,"reference_active":true,"delta_db":-0.18817857260655158},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0681397696196026e-06,"output_power":9.561422334256373e-07,"reference_active":true,"delta_db":-0.4810558359210829},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7830475792130745e-08,"output_power":5.198962129673725e-08,"reference_active":true,"delta_db":-0.46240111461881594}],"active_band_count":3,"active_band_mae_db":0.3772118410488168}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":2.396950429350484,"rms_envelope_nrmse":0.2342540155533099,"envelope_correlation":0.9630569095773246,"rms_gain_error_db":-0.6886887037906144,"reference_mean_power":1.1507041647425961e-05,"output_mean_power":9.819623002456122e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.59330844916253e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.586447862335723e-06,"output_power":3.7056069643964546e-06,"reference_active":true,"delta_db":-0.9261710697943477},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.405946680666402e-07,"output_power":6.091907424844455e-07,"reference_active":true,"delta_db":-1.398333359320609},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329768801508534e-07,"output_power":1.8613307773845302e-07,"reference_active":true,"delta_db":-2.5259052130838326}],"active_band_count":3,"active_band_mae_db":1.6168032140662632}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":1.5220045302691385,"rms_envelope_nrmse":0.13738372019056672,"envelope_correlation":0.9896917885818508,"rms_gain_error_db":-0.6887647634006775,"reference_mean_power":1.1561945751119366e-05,"output_mean_power":9.866303050300195e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.107379849793497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.414350374295126e-06,"output_power":3.505638334992087e-06,"reference_active":true,"delta_db":-1.000996915090739},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6495486661033e-07,"output_power":3.276967833202489e-07,"reference_active":true,"delta_db":-1.5193861940646636},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1590180837860975e-08,"output_power":1.1233918275343459e-08,"reference_active":true,"delta_db":-2.837250198295968}],"active_band_count":3,"active_band_mae_db":1.785877769150457}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0013","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.1565444469451904,"predicted_mos":3.1565444469451904,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"7e882534f7f3d5c6d6f63c8c1c201537371b6145a6ad789aeb0ade3e0447ca30","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.336659721000615},"started_at":"2026-10-13T05:10:57.527307+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028050140008417657,"native_input_samples":144000,"native_input_sha256":"e92d1ee9674cb056f32676a0ef28ff32b1976e5c4a4a930f7f4dd37561267ddb","output_sha256":"99dedd93cc01f99aaea923cd7754ecc5bf3fd4508635139e8f989d3c16289e7e","output_samples":144240,"length_delta_samples":240,"analysis_s":1.3248720629999298,"synthesis_s":0.6637796070008335,"total_s":1.9886516700007633,"rtf":0.3314419450001272,"output_peak":0.28181418776512146,"output_rms":0.024244673784965268,"finished_at":"2026-10-13T05:10:59.534441+00:00","analysis_stages":{"external_f0_s":0.16920727900105703},"evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"99dedd93cc01f99aaea923cd7754ecc5bf3fd4508635139e8f989d3c16289e7e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-32.307457854230336,"output_peak":0.2819225159139088},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.449523505662695,"spectral_convergence":0.16134733360154335,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.101319163814524,"spectral_convergence":0.15706286034432793,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.5138433108096505,"spectral_convergence":0.16629880304657949,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00010587248293516261,"delta_db":-0.07590468388845048,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.0002463993196148896,"delta_db":-0.2578051881055289,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.00015302900957709978,"delta_db":-0.15386381861849993,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":6.007516925294217e-05,"delta_db":-0.30834829909992006,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.4074311986615681e-05,"delta_db":-0.5433754410333437,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":9.362201676992522e-06,"delta_db":-1.1932242865716147,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.42208695288622633,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.22781618343163568,"pitch":null,"output_rms_dbfs":-32.30022982517444,"output_peak":0.2819225159139088,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.35489532676229},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9520431245799487,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024910320651486587,"rms_dbfs":-32.072413641742806,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.23336111718971,"mean":-9.009043375651042e-07,"zero_energy":false},"output":{"rms":0.024265458887968768,"rms_dbfs":-32.30022982517444,"peak":0.2819225159139088,"peak_dbfs":-10.997404750984357,"crest_factor_db":21.302825074190082,"mean":6.004733720027358e-06,"zero_energy":false},"rms_difference_db":-0.227816183431635},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":3.7445558635934744e-07,"output_rms_dbfs":-64.26599685956259,"output_zero_energy":false,"output_peak":0.010887684548780318,"output_energy_fraction":0.00014308885704709755,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":127,"both_estimated_voiced_frames":106,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.905982905982906,"absolute_difference_cents_median":0.963739299010419,"absolute_difference_cents_p95":5.525788935133573,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001113800968506657,"log_rms_mae_db":0.9284295178083507,"rms_envelope_nrmse":0.0671824185264266,"envelope_correlation":0.9963560298985066,"rms_gain_error_db":-0.22781566302258255,"reference_mean_power":0.0006210401213146882,"output_mean_power":0.0005893022397167443},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.9949527250696107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014723003466329858,"output_power":0.00014414203201629295,"reference_active":true,"delta_db":-0.09205774180900085},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0749733198322775e-05,"output_power":2.832389776000857e-05,"reference_active":true,"delta_db":-0.3568833389217456},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.678489010984287e-06,"output_power":3.751220595337855e-06,"reference_active":true,"delta_db":-1.8006018552839096}],"active_band_count":3,"active_band_mae_db":0.7498476453382187}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":0.6361834138383907,"rms_envelope_nrmse":0.042382259552342365,"envelope_correlation":0.9988611225170541,"rms_gain_error_db":-0.2277784181977545,"reference_mean_power":0.0006241534506116662,"output_mean_power":0.000592261543370306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.688739338888854e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013414910111714758,"output_power":0.00013166506659505732,"reference_active":true,"delta_db":-0.08117203837176913},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.37620762285614e-05,"output_power":1.281426202364322e-05,"reference_active":true,"delta_db":-0.3099035896814792},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.14998789435472e-07,"output_power":2.637164480201496e-07,"reference_active":true,"delta_db":-1.9690961223661294}],"active_band_count":3,"active_band_mae_db":0.7867239168064591}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":1.3046200104765453,"rms_envelope_nrmse":0.07951898389641716,"envelope_correlation":0.9958811396028803,"rms_gain_error_db":-0.23328938211181674,"reference_mean_power":0.000315142987429675,"output_mean_power":0.00029866113377223746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.94905977338963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010605229421605894,"output_power":0.00010129395589000188,"reference_active":true,"delta_db":-0.199365357806951},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.109806584151368e-05,"output_power":1.9173961749847517e-05,"reference_active":true,"delta_db":-0.4153078665212008},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3612767263836796e-06,"output_power":2.0432861083837495e-06,"reference_active":true,"delta_db":-2.1617508604330156}],"active_band_count":3,"active_band_mae_db":0.9254746949203891}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":0.8154772801088171,"rms_envelope_nrmse":0.047252220930315496,"envelope_correlation":0.9987907320867538,"rms_gain_error_db":-0.23328053166514204,"reference_mean_power":0.00031672629889435445,"output_mean_power":0.0003001622503704656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7988521248624547e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.967318213226487e-05,"output_power":9.470852965221473e-05,"reference_active":true,"delta_db":-0.22189229368668986},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.491202888187961e-06,"output_power":8.800001551256489e-06,"reference_active":true,"delta_db":-0.3283850846156048},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5805315595038676e-07,"output_power":1.9082205146267606e-07,"reference_active":true,"delta_db":-2.733189455460214}],"active_band_count":3,"active_band_mae_db":1.0944889445875028}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":1.4514272914902278,"rms_envelope_nrmse":0.0984108247747637,"envelope_correlation":0.9941221202815104,"rms_gain_error_db":-0.3537670838676198,"reference_mean_power":6.366451195739098e-05,"output_mean_power":5.868413477128642e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5126235269128304e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.399199855439123e-05,"output_power":1.2359724429572971e-05,"reference_active":true,"delta_db":-0.5387096369984256},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.395771779201436e-06,"output_power":2.1074141468559524e-06,"reference_active":true,"delta_db":-0.5569555377232795},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.685934249147904e-07,"output_power":6.375248040578008e-07,"reference_active":true,"delta_db":-0.8119957843722416}],"active_band_count":3,"active_band_mae_db":0.6358869863646489}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":0.9833116958217027,"rms_envelope_nrmse":0.06499547098767827,"envelope_correlation":0.9979859975482859,"rms_gain_error_db":-0.35365102435493295,"reference_mean_power":6.398182095924489e-05,"output_mean_power":5.897819727031681e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.275432891493173e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2825234350232915e-05,"output_power":1.135659478087461e-05,"reference_active":true,"delta_db":-0.5281717971241561},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0681397696196026e-06,"output_power":9.243778497507855e-07,"reference_active":true,"delta_db":-0.6277855505644786},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7830475792130745e-08,"output_power":4.500869575350045e-08,"reference_active":true,"delta_db":-1.0886033713050296}],"active_band_count":3,"active_band_mae_db":0.7481869063312215}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":1.3809339527779307,"rms_envelope_nrmse":0.1510231971404905,"envelope_correlation":0.9858882358286859,"rms_gain_error_db":-0.5412802176208866,"reference_mean_power":1.1507041647425961e-05,"output_mean_power":1.0158642175996858e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.59330844916253e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.586447862335723e-06,"output_power":4.044962397199613e-06,"reference_active":true,"delta_db":-0.5456197223735636},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.405946680666402e-07,"output_power":8.780040504386889e-07,"reference_active":true,"delta_db":0.18909888486189452},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329768801508534e-07,"output_power":3.477402656226758e-07,"reference_active":true,"delta_db":0.18840901685619613}],"active_band_count":3,"active_band_mae_db":0.3077092080305514}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":0.9858000003316157,"rms_envelope_nrmse":0.09462759821311259,"envelope_correlation":0.9958823977790984,"rms_gain_error_db":-0.5407508370729712,"reference_mean_power":1.1561945751119366e-05,"output_mean_power":1.0208356861605605e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.107379849793497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.414350374295126e-06,"output_power":3.987665227177666e-06,"reference_active":true,"delta_db":-0.4414810962412412},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6495486661033e-07,"output_power":4.658032823554894e-07,"reference_active":true,"delta_db":0.007917467688950526},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1590180837860975e-08,"output_power":1.8745551117416826e-08,"reference_active":true,"delta_db":-0.6135806680834384}],"active_band_count":3,"active_band_mae_db":0.35432641067121}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0013","method":"magphase","status":"measured","mos":{"status":"measured","value":2.946514129638672,"predicted_mos":2.946514129638672,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289263,"input_sample_rate":48000,"input_seconds":6.0263125,"input_float32_sha256":"17c5101c7a5be82359055c1d4177dd89a2d58fa8072e98f3613c7a2b1448c9a1","internal_samples":96421,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5317961599994305},"started_at":"2026-10-13T05:16:32.469719+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.012873245001173927,"native_input_samples":288000,"native_input_sha256":"f5d10c197eb1877975474567620f2d37c801bc9b87808cd8b0b7933812c5b55e","output_sha256":"ac5b155c23f26518835499092fd8f30e098311d997d62aa5d750af078ec7ae32","output_samples":289263,"length_delta_samples":1263,"analysis_s":3.375277092998658,"synthesis_s":0.982805814999665,"total_s":4.358082907998323,"rtf":0.7263471513330538,"output_peak":0.29308004439047336,"output_rms":0.023766111784777885,"finished_at":"2026-10-13T05:16:36.852292+00:00","evaluation_input":{"reference":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"ac5b155c23f26518835499092fd8f30e098311d997d62aa5d750af078ec7ae32","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96421,"length_delta_samples":421,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-32.48256619934261,"output_peak":0.2916560498036773},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9956337312411196,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.420010050838638,"spectral_convergence":0.6345726826413315,"spectral_convergence_status":"ok","amplitude_floor":7.181372005990482e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.561013585615801,"spectral_convergence":0.38939001146519,"spectral_convergence_status":"ok","amplitude_floor":5.417622086660404e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.323550143023212,"spectral_convergence":0.34224289279081965,"spectral_convergence_status":"ok","amplitude_floor":3.166881446630363e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001077391554152542,"output_power":0.00010034050146374402,"delta_db":-0.3089729958211957,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002614688899274717,"output_power":0.0002337342410246678,"delta_db":-0.486976838235172,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00015854777413607497,"output_power":0.0001453234592015021,"delta_db":-0.37824422287458587,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.449556222199797e-05,"output_power":6.13369762665493e-05,"delta_db":-0.21807470221397962,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.5950148027800597e-05,"output_power":1.597169984384767e-05,"delta_db":0.00586421954267789,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.2322545231279567e-05,"output_power":1.0362462902195597e-05,"delta_db":-0.7523743217710288,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3584178834097733,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.39122468567342505,"pitch":null,"output_rms_dbfs":-32.46363832741623,"output_peak":0.2916560498036773,"reference_rms_dbfs":-32.072413641742806,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.768191259825883},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7623182064752158,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.024910320651486587,"rms_dbfs":-32.072413641742806,"peak":0.287109375,"peak_dbfs":-10.839052524553093,"crest_factor_db":21.23336111718971,"mean":-9.009043375651042e-07,"zero_energy":false},"output":{"rms":0.02381321777296186,"rms_dbfs":-32.46363832741623,"peak":0.2916560498036773,"peak_dbfs":-10.702580211930917,"crest_factor_db":21.761058115485312,"mean":-0.00017116864795549072,"zero_energy":false},"rms_difference_db":-0.3912246856734214},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2342336686560883e-06,"threshold_dbfs":-59.086026106423326,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":135,"quiet_samples":21600,"reference_mean_power":1.6482844431367186e-07,"output_mean_power":3.2149615528106807e-07,"output_rms_dbfs":-64.92824216364022,"output_zero_energy":false,"output_peak":0.008552372339645299,"output_energy_fraction":0.00012756223929974247,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":95,"both_estimated_voiced_frames":78,"reference_voiced_output_unvoiced_frames":39,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.6666666666666666,"absolute_difference_cents_median":7.134316237945338,"absolute_difference_cents_p95":32.753171870279004,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001113800968506657,"log_rms_mae_db":2.9890373532414416,"rms_envelope_nrmse":0.2730868817269772,"envelope_correlation":0.9302131439385183,"rms_gain_error_db":-0.39121442462873546,"reference_mean_power":0.0006210401213146882,"output_mean_power":0.0005675422734301886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.9949527250696107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014723003466329858,"output_power":0.00013655913504243567,"reference_active":true,"delta_db":-0.3267565701601677},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0749733198322775e-05,"output_power":2.811793308061932e-05,"reference_active":true,"delta_db":-0.3885795895362635},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.678489010984287e-06,"output_power":5.8529968019865726e-06,"reference_active":true,"delta_db":0.13145497146445473}],"active_band_count":3,"active_band_mae_db":0.28226371038696196}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.537149969198292e-05,"log_rms_mae_db":2.156802405503205,"rms_envelope_nrmse":0.18037606854436997,"envelope_correlation":0.967607331984395,"rms_gain_error_db":-0.39115054748183686,"reference_mean_power":0.0006241534506116662,"output_mean_power":0.0005703958027294957},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.688739338888854e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013414910111714758,"output_power":0.00012436273598191017,"reference_active":true,"delta_db":-0.328974990122832},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.37620762285614e-05,"output_power":1.2786224102857857e-05,"reference_active":true,"delta_db":-0.3194164708740914},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.14998789435472e-07,"output_power":4.429357565933567e-07,"reference_active":true,"delta_db":0.282939108513239}],"active_band_count":3,"active_band_mae_db":0.31044352317005414}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.440030149355929e-05,"log_rms_mae_db":3.9005653398348206,"rms_envelope_nrmse":0.3039956878041851,"envelope_correlation":0.9328782191780948,"rms_gain_error_db":-0.45879331494401016,"reference_mean_power":0.000315142987429675,"output_mean_power":0.0002835491280691905},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.94905977338963e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010605229421605894,"output_power":9.470397541802038e-05,"reference_active":true,"delta_db":-0.49151858093091694},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.109806584151368e-05,"output_power":1.8522905133011294e-05,"reference_active":true,"delta_db":-0.5653354086587601},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3612767263836796e-06,"output_power":3.0554584556341003e-06,"reference_active":true,"delta_db":-0.41427885339432635}],"active_band_count":3,"active_band_mae_db":0.49037761432800114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.828793840107377e-05,"log_rms_mae_db":2.9157561675463586,"rms_envelope_nrmse":0.20031628490079545,"envelope_correlation":0.9697626718936251,"rms_gain_error_db":-0.45877407897218664,"reference_mean_power":0.00031672629889435445,"output_mean_power":0.00028497497088801496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7988521248624547e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.967318213226487e-05,"output_power":8.926803308985415e-05,"reference_active":true,"delta_db":-0.47882357880640536},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.491202888187961e-06,"output_power":8.300294703383977e-06,"reference_active":true,"delta_db":-0.5822774481905681},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5805315595038676e-07,"output_power":3.1678561129480143e-07,"reference_active":true,"delta_db":-0.531820587405846}],"active_band_count":3,"active_band_mae_db":0.5309738714676064}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5893418086145424e-05,"log_rms_mae_db":3.8624456926855997,"rms_envelope_nrmse":0.31522331020297617,"envelope_correlation":0.9294666810502518,"rms_gain_error_db":-0.24101624033197513,"reference_mean_power":6.366451195739098e-05,"output_mean_power":6.0227632280757304e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.5126235269128304e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.399199855439123e-05,"output_power":1.3273171767997398e-05,"reference_active":true,"delta_db":-0.2290503696957097},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.395771779201436e-06,"output_power":2.286649498289263e-06,"reference_active":true,"delta_db":-0.20245844595718762},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.685934249147904e-07,"output_power":9.77358103234662e-07,"reference_active":true,"delta_db":1.0435705314587689}],"active_band_count":3,"active_band_mae_db":0.4916931157038887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.638203463734759e-05,"log_rms_mae_db":2.740917071407662,"rms_envelope_nrmse":0.20960867327809082,"envelope_correlation":0.9669299690596715,"rms_gain_error_db":-0.24082502434484548,"reference_mean_power":6.398182095924489e-05,"output_mean_power":6.053047664460535e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.275432891493173e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2825234350232915e-05,"output_power":1.2214190075391592e-05,"reference_active":true,"delta_db":-0.21200635402884083},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0681397696196026e-06,"output_power":1.0695469784653953e-06,"reference_active":true,"delta_db":0.00571779869467191},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.7830475792130745e-08,"output_power":6.697494088422535e-08,"reference_active":true,"delta_db":0.6375557353291905}],"active_band_count":3,"active_band_mae_db":0.2850932960175678}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.0860899025143115e-05,"log_rms_mae_db":3.6231651507570484,"rms_envelope_nrmse":0.32864917746606354,"envelope_correlation":0.9251642378626008,"rms_gain_error_db":-0.021310928354404568,"reference_mean_power":1.1507041647425961e-05,"output_mean_power":1.1450714646022673e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.59330844916253e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.586447862335723e-06,"output_power":3.8742961739535955e-06,"reference_active":true,"delta_db":-0.7328364324079835},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.405946680666402e-07,"output_power":1.1157578689791674e-06,"reference_active":true,"delta_db":1.2298332749600107},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.329768801508534e-07,"output_power":6.50009043357253e-07,"reference_active":true,"delta_db":2.9050531903978465}],"active_band_count":3,"active_band_mae_db":1.6225742992552803}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.4734203651083813e-05,"log_rms_mae_db":2.4401359379792598,"rms_envelope_nrmse":0.182583537727142,"envelope_correlation":0.9754991046340293,"rms_gain_error_db":-0.020210074433923808,"reference_mean_power":1.1561945751119366e-05,"output_mean_power":1.1508266751051473e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.107379849793497e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.414350374295126e-06,"output_power":3.8873701103256685e-06,"reference_active":true,"delta_db":-0.5521090977884201},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6495486661033e-07,"output_power":6.558506168324216e-07,"reference_active":true,"delta_db":1.493941335728126},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1590180837860975e-08,"output_power":3.7248268769320596e-08,"reference_active":true,"delta_db":2.3684981237865337}],"active_band_count":3,"active_band_mae_db":1.4715161857676933}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"hnm","status":"measured","mos":{"status":"measured","value":3.279989242553711,"predicted_mos":3.279989242553711,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4f8c1c41f21df802cc494aa84673e91d63fee2cf878983f6eb60a5d529980bf9","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.132209589000922},"started_at":"2026-10-13T04:39:38.420017+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.623300017352449e-05,"native_input_samples":96000,"native_input_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","output_sha256":"70597e2941879c06f6df9367af4d84aa921110b769d031bbb1c4325ad0f4f70d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0847453240003233,"synthesis_s":0.49511670599986246,"total_s":1.5798620300001858,"rtf":0.2633103383333643,"output_peak":0.24340173916946148,"output_rms":0.030536173157116573,"finished_at":"2026-10-13T04:39:40.007087+00:00","evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"70597e2941879c06f6df9367af4d84aa921110b769d031bbb1c4325ad0f4f70d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.254206405456743,"spectral_convergence":0.18834472659933918,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.255389153730749,"spectral_convergence":0.1958863411895917,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.75043434826696,"spectral_convergence":0.20033125628182907,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00012712485917522784,"delta_db":0.13258816648602442,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.00038921076665065796,"delta_db":-0.047541759912082746,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00034352152383952206,"delta_db":-0.18190563191829923,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":5.3605428675827446e-05,"delta_db":-0.4512534222190887,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":9.443126446663979e-06,"delta_db":-0.45152721644005145,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":9.552166293507342e-06,"delta_db":0.07118690609595175,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22266718384524972,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10084130567494962,"pitch":null,"output_rms_dbfs":-30.303707807205473,"output_peak":0.24340173916946148,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.420009969151484},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.254206405456743,"spectral_convergence":0.18834472659933918,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.255389153730749,"spectral_convergence":0.1958863411895917,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.75043434826696,"spectral_convergence":0.20033125628182907,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00012712485917522784,"delta_db":0.13258816648602442,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.00038921076665065796,"delta_db":-0.047541759912082746,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00034352152383952206,"delta_db":-0.18190563191829923,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":5.3605428675827446e-05,"delta_db":-0.4512534222190887,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":9.443126446663979e-06,"delta_db":-0.45152721644005145,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":9.552166293507342e-06,"delta_db":0.07118690609595175,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.22266718384524972,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10084130567494962,"pitch":null,"output_rms_dbfs":-30.303707807205473,"output_peak":0.24340173916946148,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.420009969151484},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9102567077314064,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030892757470915053,"rms_dbfs":-30.202866501530522,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.8508064339276,"mean":-8.862813313802083e-07,"zero_energy":false},"output":{"rms":0.030536173157116573,"rms_dbfs":-30.303707807205473,"peak":0.24340173916946148,"peak_dbfs":-12.273526458930764,"crest_factor_db":18.03018134827471,"mean":-5.0675240491980075e-05,"zero_energy":false},"rms_difference_db":-0.10084130567495109},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":127,"quiet_samples":20320,"reference_mean_power":1.5799149570209303e-07,"output_mean_power":2.2287309824982286e-07,"output_rms_dbfs":-66.51942349584986,"output_zero_energy":false,"output_peak":0.00811161959359522,"output_energy_fraction":5.059188973492399e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":226,"both_estimated_voiced_frames":213,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9551569506726457,"absolute_difference_cents_median":1.5912784161152365,"absolute_difference_cents_p95":6.362086276318756,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":0.757588229527955,"rms_envelope_nrmse":0.06281437416876767,"envelope_correlation":0.9961601538895933,"rms_gain_error_db":-0.100841216517711,"reference_mean_power":0.0009551583992700257,"output_mean_power":0.0009332355569956512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1194144575506277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002166069295810575,"output_power":0.00020554671633415448,"reference_active":true,"delta_db":-0.22761803047513057},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38201712334803e-05,"output_power":2.1279841019538764e-05,"reference_active":true,"delta_db":-0.4897650007741837},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.302931519380677e-06,"output_power":6.89716790399437e-06,"reference_active":true,"delta_db":-0.8056065743877627}],"active_band_count":3,"active_band_mae_db":0.5076632018790257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":0.4457909469791467,"rms_envelope_nrmse":0.03636263007566115,"envelope_correlation":0.9987197486770476,"rms_gain_error_db":-0.10084101832227252,"reference_mean_power":0.0009599620519447541,"output_mean_power":0.0009379289988028571},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.73215395386014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001986653950491428,"output_power":0.00018844552491443397,"reference_active":true,"delta_db":-0.22936396640794154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916927878865493e-05,"output_power":9.718695112781876e-06,"reference_active":true,"delta_db":-0.5049248321441028},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.81959177520348e-07,"output_power":4.5763057768905286e-07,"reference_active":true,"delta_db":-1.0437748623449572}],"active_band_count":3,"active_band_mae_db":0.5926878869656672}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":1.2992026161025907,"rms_envelope_nrmse":0.08490046061840423,"envelope_correlation":0.9944915991523962,"rms_gain_error_db":-0.1354988174497198,"reference_mean_power":0.0006127358569067375,"output_mean_power":0.0005939137964920907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.366962347376557e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019364231078177667,"output_power":0.00018191406915476585,"reference_active":true,"delta_db":-0.27133968035456246},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9545968551344243e-05,"output_power":1.6673677720670017e-05,"reference_active":true,"delta_db":-0.6902579274771998},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.863910891753386e-06,"output_power":4.85224701232139e-06,"reference_active":true,"delta_db":-1.506287355103975}],"active_band_count":3,"active_band_mae_db":0.8226283209785791}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":0.788605779938473,"rms_envelope_nrmse":0.05243379275998193,"envelope_correlation":0.9978997923319161,"rms_gain_error_db":-0.13549882428118398,"reference_mean_power":0.000615817475654304,"output_mean_power":0.0005969007529330503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.097820742933078e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018026482035346189,"output_power":0.0001690299109686917,"reference_active":true,"delta_db":-0.2794741737838535},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.739141870245849e-06,"output_power":7.607523874678688e-06,"reference_active":true,"delta_db":-0.6022546571436636},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.76389261167761e-07,"output_power":3.3771252234283634e-07,"reference_active":true,"delta_db":-2.3216871625831583}],"active_band_count":3,"active_band_mae_db":1.0678053311702251}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":1.6452500193984134,"rms_envelope_nrmse":0.16020950295461125,"envelope_correlation":0.9823581868097605,"rms_gain_error_db":-0.4334099270668393,"reference_mean_power":5.214627438247733e-05,"output_mean_power":4.719351149238125e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.666296423769005e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7235548052453483e-05,"output_power":2.3421216731286826e-05,"reference_active":true,"delta_db":-0.6552666596608571},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163900077209868e-06,"output_power":3.7340274640649723e-06,"reference_active":true,"delta_db":-0.4732279112884629},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1939677607224186e-06,"output_power":7.21292505806058e-07,"reference_active":true,"delta_db":-2.188811803042971}],"active_band_count":3,"active_band_mae_db":1.1057687913307637}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":0.9466121028798112,"rms_envelope_nrmse":0.09836749084067553,"envelope_correlation":0.9941733625276111,"rms_gain_error_db":-0.4334106900316318,"reference_mean_power":5.2408520364582374e-05,"output_mean_power":4.743084147237112e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.333048787706058e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.583453225376263e-05,"output_power":2.2304855754992657e-05,"reference_active":true,"delta_db":-0.6380118383508313},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.076540066109676e-06,"output_power":1.8932667105176725e-06,"reference_active":true,"delta_db":-0.4012851636406539},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.119613733354675e-08,"output_power":3.6794060268331016e-08,"reference_active":true,"delta_db":-2.209462951939246}],"active_band_count":3,"active_band_mae_db":1.082919984643577}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":1.7141199465354586,"rms_envelope_nrmse":0.20447129238019182,"envelope_correlation":0.967318348008361,"rms_gain_error_db":-0.36584980736651085,"reference_mean_power":6.886139457761318e-06,"output_mean_power":6.329812316035236e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.997833564030977e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1829042042252137e-06,"output_power":1.9870222927101204e-06,"reference_active":true,"delta_db":-0.40831937754159564},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11137684368614e-07,"output_power":6.354811839577482e-07,"reference_active":true,"delta_db":-1.0599188260810812},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6537364416157555e-07,"output_power":1.9607960783016778e-07,"reference_active":true,"delta_db":0.7396613304492644}],"active_band_count":3,"active_band_mae_db":0.7359665113573137}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":1.1373838739212505,"rms_envelope_nrmse":0.12156005101372729,"envelope_correlation":0.9885434107919249,"rms_gain_error_db":-0.36584990491630504,"reference_mean_power":6.9207717017349125e-06,"output_mean_power":6.361646498569307e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.44334903190685e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.931899455537554e-06,"output_power":1.790015187436955e-06,"reference_active":true,"delta_db":-0.33127804311187203},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0565449907563406e-07,"output_power":3.4434128188545067e-07,"reference_active":true,"delta_db":-0.7116720517010513},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.9695465540608e-09,"output_power":9.579823042316487e-09,"reference_active":true,"delta_db":0.7992387500001421}],"active_band_count":3,"active_band_mae_db":0.6140629482710218}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.2355141639709473,"predicted_mos":3.2355141639709473,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"30e7c352ecbd194565d59d705c2afa89bb92cc30b83d60bbe6d3a6e49c9a2d02","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9370022830007656},"started_at":"2026-10-13T04:42:39.125518+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2194999473867938e-05,"native_input_samples":96000,"native_input_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","output_sha256":"02d2d6c20ad784f8f1351ae5a7c8134e6b3ddb67d2b796ff38ea8069dda89e19","output_samples":96000,"length_delta_samples":0,"analysis_s":2.7701325500001985,"synthesis_s":0.11876864199984993,"total_s":2.8889011920000485,"rtf":0.4814835320000081,"output_peak":0.27661458404557177,"output_rms":0.03595499384609935,"finished_at":"2026-10-13T04:42:42.031387+00:00","evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"02d2d6c20ad784f8f1351ae5a7c8134e6b3ddb67d2b796ff38ea8069dda89e19","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.132128597063173,"spectral_convergence":0.6374444545937361,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.709182566174027,"spectral_convergence":0.29586942862132876,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.072835778252388,"spectral_convergence":0.31970425526445945,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00012322327099324731,"delta_db":-0.0027893511668967406,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.0005508566121691764,"delta_db":1.4609954527807032,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.0005215656160450321,"delta_db":1.6316443622805012,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":7.626024845069093e-05,"delta_db":1.0796410025971828,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":1.1137567977098813e-05,"delta_db":0.26521841009127967,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":9.71826683779581e-06,"delta_db":0.14605635596624036,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7643908224804674,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.318050876678189,"pitch":null,"output_rms_dbfs":-28.884815624852337,"output_peak":0.27661458404557177,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.304715647163197},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.132128597063173,"spectral_convergence":0.6374444545937361,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.709182566174027,"spectral_convergence":0.29586942862132876,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.072835778252388,"spectral_convergence":0.31970425526445945,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00012322327099324731,"delta_db":-0.0027893511668967406,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.0005508566121691764,"delta_db":1.4609954527807032,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.0005215656160450321,"delta_db":1.6316443622805012,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":7.626024845069093e-05,"delta_db":1.0796410025971828,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":1.1137567977098813e-05,"delta_db":0.26521841009127967,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":9.71826683779581e-06,"delta_db":0.14605635596624036,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7643908224804674,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.318050876678189,"pitch":null,"output_rms_dbfs":-28.884815624852337,"output_peak":0.27661458404557177,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.304715647163197},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9418607997442632,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030892757470915053,"rms_dbfs":-30.202866501530522,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.8508064339276,"mean":-8.862813313802083e-07,"zero_energy":false},"output":{"rms":0.03595499384609934,"rms_dbfs":-28.884815624852337,"peak":0.27661458404557177,"peak_dbfs":-11.162498523323467,"crest_factor_db":17.722317101528866,"mean":3.454105952445703e-08,"zero_energy":false},"rms_difference_db":1.3180508766781855},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":127,"quiet_samples":20320,"reference_mean_power":1.5799149570209303e-07,"output_mean_power":3.0635722814346975e-07,"output_rms_dbfs":-65.13771868536563,"output_zero_energy":false,"output_peak":0.010060851302172625,"output_energy_fraction":5.0160535530703713e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":243,"both_estimated_voiced_frames":211,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9461883408071748,"absolute_difference_cents_median":4.8683575058963084,"absolute_difference_cents_p95":17.686855286947406,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":1.4895143870416716,"rms_envelope_nrmse":0.2263353816737956,"envelope_correlation":0.9798983385160904,"rms_gain_error_db":1.3180510117788904,"reference_mean_power":0.0009551583992700257,"output_mean_power":0.0012938397815716158},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1194144575506277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002166069295810575,"output_power":0.00029141730885109274,"reference_active":true,"delta_db":1.2884299705528686},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38201712334803e-05,"output_power":3.06939421028713e-05,"reference_active":true,"delta_db":1.1010779046100811},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.302931519380677e-06,"output_power":1.0296146826131056e-05,"reference_active":true,"delta_db":0.9344327112995519}],"active_band_count":3,"active_band_mae_db":1.1079801954875006}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":1.1039690329993828,"rms_envelope_nrmse":0.18792477128489335,"envelope_correlation":0.9928687923539935,"rms_gain_error_db":1.3180512780775144,"reference_mean_power":0.0009599620519447541,"output_mean_power":0.0013003467997696532},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.73215395386014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001986653950491428,"output_power":0.00026966197947481016,"reference_active":true,"delta_db":1.3269749297135482},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916927878865493e-05,"output_power":1.3804015410421983e-05,"reference_active":true,"delta_db":1.0190499424357513},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.81959177520348e-07,"output_power":5.775799733187339e-07,"reference_active":true,"delta_db":-0.032803950791429705}],"active_band_count":3,"active_band_mae_db":0.792942940980243}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":1.8850368452777206,"rms_envelope_nrmse":0.27974044426928957,"envelope_correlation":0.9819327049569947,"rms_gain_error_db":1.745805416582188,"reference_mean_power":0.0006127358569067375,"output_mean_power":0.0009159121862573201},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.366962347376557e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019364231078177667,"output_power":0.00027107994176586996,"reference_active":true,"delta_db":1.460971271280125},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9545968551344243e-05,"output_power":2.9178940005416832e-05,"reference_active":true,"delta_db":1.7401231538824524},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.863910891753386e-06,"output_power":8.895288745934203e-06,"reference_active":true,"delta_db":1.1258841327915956}],"active_band_count":3,"active_band_mae_db":1.4423261859847243}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":1.543462966338008,"rms_envelope_nrmse":0.2490151672275162,"envelope_correlation":0.9918060610236799,"rms_gain_error_db":1.74580553385209,"reference_mean_power":0.000615817475654304,"output_mean_power":0.0009205185877968537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.097820742933078e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018026482035346189,"output_power":0.0002535853143255797,"reference_active":true,"delta_db":1.4821311891045146},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.739141870245849e-06,"output_power":1.260030142305516e-05,"reference_active":true,"delta_db":1.5891214469539325},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.76389261167761e-07,"output_power":5.30512607360507e-07,"reference_active":true,"delta_db":-0.3602017163745705}],"active_band_count":3,"active_band_mae_db":1.1438181174776725}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":1.5580193361164447,"rms_envelope_nrmse":0.21348338719128995,"envelope_correlation":0.9798754526265925,"rms_gain_error_db":0.9766426605409304,"reference_mean_power":5.214627438247733e-05,"output_mean_power":6.529614645602306e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.666296423769005e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7235548052453483e-05,"output_power":3.685018949610938e-05,"reference_active":true,"delta_db":1.313036066198712},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163900077209868e-06,"output_power":5.716816594325081e-06,"reference_active":true,"delta_db":1.3765396047237877},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1939677607224186e-06,"output_power":1.2041776955382127e-06,"reference_active":true,"delta_db":0.03697978477755119}],"active_band_count":3,"active_band_mae_db":0.9088518185666835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":1.0085839507271332,"rms_envelope_nrmse":0.14607178039543553,"envelope_correlation":0.9951885268082362,"rms_gain_error_db":0.9766430904048246,"reference_mean_power":5.2408520364582374e-05,"output_mean_power":6.56245302374944e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.333048787706058e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.583453225376263e-05,"output_power":3.529099373937519e-05,"reference_active":true,"delta_db":1.3546328473879796},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.076540066109676e-06,"output_power":2.825109080358854e-06,"reference_active":true,"delta_db":1.3369490594861984},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.119613733354675e-08,"output_power":7.168026649789466e-08,"reference_active":true,"delta_db":0.6867560072502155}],"active_band_count":3,"active_band_mae_db":1.1261126380414643}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":1.7190619490826275,"rms_envelope_nrmse":0.1999528808410819,"envelope_correlation":0.9702472241455548,"rms_gain_error_db":0.24665114765103457,"reference_mean_power":6.886139457761318e-06,"output_mean_power":7.288546494334855e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.997833564030977e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1829042042252137e-06,"output_power":2.2206981906082473e-06,"reference_active":true,"delta_db":0.07454861402361111},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11137684368614e-07,"output_power":7.920665543174801e-07,"reference_active":true,"delta_db":-0.10332903384712969},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6537364416157555e-07,"output_power":1.6501440083308938e-07,"reference_active":true,"delta_db":-0.009444497709429654}],"active_band_count":3,"active_band_mae_db":0.062440715193390145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":1.0564351835245935,"rms_envelope_nrmse":0.10036004757366515,"envelope_correlation":0.9924622798946545,"rms_gain_error_db":0.24665122124478284,"reference_mean_power":6.9207717017349125e-06,"output_mean_power":7.325202675454698e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.44334903190685e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.931899455537554e-06,"output_power":1.9845533063251957e-06,"reference_active":true,"delta_db":0.11678248726235581},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0565449907563406e-07,"output_power":4.0572542357132016e-07,"reference_active":true,"delta_db":0.0007592526343564686},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.9695465540608e-09,"output_power":8.691821919155975e-09,"reference_active":true,"delta_db":0.3767720787299035}],"active_band_count":3,"active_band_mae_db":0.16477127287553858}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"facodec","status":"measured","mos":{"status":"measured","value":4.281975269317627,"predicted_mos":4.281975269317627,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"fc483462702cb7b21d2ad5b7f8944639eb098639e4d6af811864ec994035bc8b","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.677625062999141},"started_at":"2026-10-13T04:54:59.100278+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4164999558706768e-05,"native_input_samples":96000,"native_input_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","output_sha256":"3733cdd5b1384874a5ea1a327eaadb692afdb6c6b01e813643fe49e6c95aab44","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4425415770001564,"synthesis_s":5.352901788000054,"total_s":6.79544336500021,"rtf":1.1325738941667016,"output_peak":0.24298909306526184,"output_rms":0.02985186699075391,"finished_at":"2026-10-13T04:55:05.901290+00:00","evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"3733cdd5b1384874a5ea1a327eaadb692afdb6c6b01e813643fe49e6c95aab44","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6101241791161724,"spectral_convergence":0.2153337813922564,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.866823759839877,"spectral_convergence":0.21848548812187027,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.609454308517254,"spectral_convergence":0.2319476216256983,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00011386222357620959,"delta_db":-0.345920074392687,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.00037777438888536016,"delta_db":-0.17706509840684737,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00033207620614983624,"delta_db":-0.329067580489079,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":4.78608383970725e-05,"delta_db":-0.9435381353127552,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":9.97215711174689e-06,"delta_db":-0.21479414506298627,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":9.588148713437568e-06,"delta_db":0.08751577510864798,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.34965013479550056,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.297703540634781,"pitch":null,"output_rms_dbfs":-30.500570042165304,"output_peak":0.24298909306526184,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.695467415824434},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6101241791161724,"spectral_convergence":0.2153337813922564,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.866823759839877,"spectral_convergence":0.21848548812187027,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.609454308517254,"spectral_convergence":0.2319476216256983,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00011386222357620959,"delta_db":-0.345920074392687,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.00037777438888536016,"delta_db":-0.17706509840684737,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00033207620614983624,"delta_db":-0.329067580489079,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":4.78608383970725e-05,"delta_db":-0.9435381353127552,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":9.97215711174689e-06,"delta_db":-0.21479414506298627,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":9.588148713437568e-06,"delta_db":0.08751577510864798,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.34965013479550056,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.297703540634781,"pitch":null,"output_rms_dbfs":-30.500570042165304,"output_peak":0.24298909306526184,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.695467415824434},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9041015341699468,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030892757470915053,"rms_dbfs":-30.202866501530522,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.8508064339276,"mean":-8.862813313802083e-07,"zero_energy":false},"output":{"rms":0.02985186699075391,"rms_dbfs":-30.500570042165304,"peak":0.24298909306526184,"peak_dbfs":-12.288264398641113,"crest_factor_db":18.212305643524193,"mean":1.1700475001275616e-05,"zero_energy":false},"rms_difference_db":-0.29770354063478166},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":127,"quiet_samples":20320,"reference_mean_power":1.5799149570209303e-07,"output_mean_power":1.8685204522820676e-07,"output_rms_dbfs":-67.2850214412709,"output_zero_energy":false,"output_peak":0.006667034700512886,"output_energy_fraction":4.438204717003491e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":230,"both_estimated_voiced_frames":211,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9461883408071748,"absolute_difference_cents_median":2.397152993799591,"absolute_difference_cents_p95":18.205054529256845,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":1.3287596631460519,"rms_envelope_nrmse":0.10261655930564322,"envelope_correlation":0.9903061133717498,"rms_gain_error_db":-0.29770354979281405,"reference_mean_power":0.0009551583992700257,"output_mean_power":0.0008918771637068974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1194144575506277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002166069295810575,"output_power":0.00019481976078782474,"reference_active":true,"delta_db":-0.46039340489004044},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38201712334803e-05,"output_power":2.436042022481588e-05,"reference_active":true,"delta_db":0.09739896616573787},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.302931519380677e-06,"output_power":7.795893817266004e-06,"reference_active":true,"delta_db":-0.27365540732819}],"active_band_count":3,"active_band_mae_db":0.27714925946132274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":1.1835239605686423,"rms_envelope_nrmse":0.09171055670382303,"envelope_correlation":0.9919798121956525,"rms_gain_error_db":-0.2977038037488622,"reference_mean_power":0.0009599620519447541,"output_mean_power":0.0008963625119604907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.73215395386014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001986653950491428,"output_power":0.0001785779016205515,"reference_active":true,"delta_db":-0.462945097154516},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916927878865493e-05,"output_power":1.0913576754266335e-05,"reference_active":true,"delta_db":-0.001333340623567221},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.81959177520348e-07,"output_power":5.762909514713641e-07,"reference_active":true,"delta_db":-0.042507207823336945}],"active_band_count":3,"active_band_mae_db":0.1689285485338067}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":1.6622162876289694,"rms_envelope_nrmse":0.11352819886733775,"envelope_correlation":0.990191519412283,"rms_gain_error_db":-0.21704675226466383,"reference_mean_power":0.0006127358569067375,"output_mean_power":0.0005828658629493843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.366962347376557e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019364231078177667,"output_power":0.0001846014194914374,"reference_active":true,"delta_db":-0.20765220336283446},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9545968551344243e-05,"output_power":2.104447391486381e-05,"reference_active":true,"delta_db":0.3208087773598137},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.863910891753386e-06,"output_power":6.4456892194171925e-06,"reference_active":true,"delta_db":-0.27302274746169647}],"active_band_count":3,"active_band_mae_db":0.26716124272811487}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":1.4184969801370158,"rms_envelope_nrmse":0.09881304226176481,"envelope_correlation":0.9922549064048991,"rms_gain_error_db":-0.21704730009044176,"reference_mean_power":0.000615817475654304,"output_mean_power":0.0005857971833107106},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.097820742933078e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018026482035346189,"output_power":0.0001725457016275073,"reference_active":true,"delta_db":-0.1900683525938619},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.739141870245849e-06,"output_power":9.361759467976567e-06,"reference_active":true,"delta_db":0.2988868889867781},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.76389261167761e-07,"output_power":6.173818839568557e-07,"reference_active":true,"delta_db":0.29838000901314177}],"active_band_count":3,"active_band_mae_db":0.2624450835312606}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":1.81492102196234,"rms_envelope_nrmse":0.1800025989358923,"envelope_correlation":0.9814389925071483,"rms_gain_error_db":-0.8664328815781884,"reference_mean_power":5.214627438247733e-05,"output_mean_power":4.2714959342102135e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.666296423769005e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7235548052453483e-05,"output_power":1.9368686122322564e-05,"reference_active":true,"delta_db":-1.4803595758160775},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163900077209868e-06,"output_power":3.5692393226120484e-06,"reference_active":true,"delta_db":-0.669246300242206},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1939677607224186e-06,"output_power":1.5303834830533285e-06,"reference_active":true,"delta_db":1.0780766962171904}],"active_band_count":3,"active_band_mae_db":1.0758941907584914}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":1.4418600881606831,"rms_envelope_nrmse":0.15120399275663857,"envelope_correlation":0.9886138303171588,"rms_gain_error_db":-0.866435782820416,"reference_mean_power":5.2408520364582374e-05,"output_mean_power":4.2929746133930663e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.333048787706058e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.583453225376263e-05,"output_power":1.836874847467945e-05,"reference_active":true,"delta_db":-1.4812103549566427},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.076540066109676e-06,"output_power":1.6827425055398384e-06,"reference_active":true,"delta_db":-0.9132265007785494},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.119613733354675e-08,"output_power":6.715461937783856e-08,"reference_active":true,"delta_db":0.40351881348713126}],"active_band_count":3,"active_band_mae_db":0.9326518897407743}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":2.083878782749033,"rms_envelope_nrmse":0.2513544340258492,"envelope_correlation":0.9552601587125769,"rms_gain_error_db":0.29805080714315724,"reference_mean_power":6.886139457761318e-06,"output_mean_power":7.375320446250735e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.997833564030977e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1829042042252137e-06,"output_power":2.422657061895507e-06,"reference_active":true,"delta_db":0.4525726482467618},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11137684368614e-07,"output_power":8.755097673820791e-07,"reference_active":true,"delta_db":0.33166416923915876},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6537364416157555e-07,"output_power":2.0286849990681645e-07,"reference_active":true,"delta_db":0.8874832127538684}],"active_band_count":3,"active_band_mae_db":0.5572400100799296}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":1.9315410951070808,"rms_envelope_nrmse":0.21295931148096145,"envelope_correlation":0.9659813603119175,"rms_gain_error_db":0.2980505555914115,"reference_mean_power":6.9207717017349125e-06,"output_mean_power":7.412412483398835e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.44334903190685e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.931899455537554e-06,"output_power":2.1829655818142037e-06,"reference_active":true,"delta_db":0.5306236827673684},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0565449907563406e-07,"output_power":4.022801404865495e-07,"reference_active":true,"delta_db":-0.03627703936453569},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.9695465540608e-09,"output_power":9.49202516671092e-09,"reference_active":true,"delta_db":0.7592526911330442}],"active_band_count":3,"active_band_mae_db":0.4420511377549827}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5278944969177246,"predicted_mos":3.5278944969177246,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"19289145d1cb603966f2246fe4a95f7b38d912e34bcd72018c8f94ab1570fde9","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.800270712999918},"started_at":"2026-10-13T07:20:45.108644+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00692232399887871,"native_input_samples":264600,"native_input_sha256":"e3dc57fb693fd861eead4ab68169f36ad814af3ae5eae65b17d37520fbebfdd1","output_sha256":"1d36eef7e2411159188d3aaef77a5c94c77a90ed34ade0100106fde921fd0b2b","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1699507490011456,"synthesis_s":5.371028432999083,"total_s":5.540979182000228,"rtf":0.9234965303333714,"output_peak":0.2166864573955536,"output_rms":0.028555288712974348,"finished_at":"2026-10-13T07:20:50.672469+00:00","evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"1d36eef7e2411159188d3aaef77a5c94c77a90ed34ade0100106fde921fd0b2b","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.884439790151184,"output_peak":0.21628510380974},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.69158478446933,"spectral_convergence":0.4516823642671151,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.711493779602689,"spectral_convergence":0.3768536029350036,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.162223858283436,"spectral_convergence":0.45578238892169365,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00010634747689153783,"delta_db":-0.6424447752032563,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.0003719906593840812,"delta_db":-0.24406986579562348,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00027613305516525157,"delta_db":-1.1302611878398108,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":4.676178309331875e-05,"delta_db":-1.0444305251003276,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":9.472620725623916e-06,"delta_db":-0.43798377788155457,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":5.365311147198986e-06,"delta_db":-2.4338827106122927,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9888454737388109,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6798545466730306,"pitch":null,"output_rms_dbfs":-30.882721048203553,"output_peak":0.21628510380974,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.521767474118485},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7651603519958053,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030892757470915053,"rms_dbfs":-30.202866501530522,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.8508064339276,"mean":-8.862813313802083e-07,"zero_energy":false},"output":{"rms":0.028566954797580583,"rms_dbfs":-30.882721048203553,"peak":0.21628510380974,"peak_dbfs":-13.299467813345508,"crest_factor_db":17.583253234858045,"mean":-1.3079668073581666e-05,"zero_energy":false},"rms_difference_db":-0.679854546673031},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":127,"quiet_samples":20320,"reference_mean_power":1.5799149570209303e-07,"output_mean_power":1.420406533076126e-07,"output_rms_dbfs":-68.47587338853957,"output_zero_energy":false,"output_peak":0.005310615836433224,"output_energy_fraction":3.6841494263224047e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":226,"both_estimated_voiced_frames":190,"reference_voiced_output_unvoiced_frames":33,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.852017937219731,"absolute_difference_cents_median":12.149810511916918,"absolute_difference_cents_p95":34.97510496457466,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":1.568199899934362,"rms_envelope_nrmse":0.12482964217524846,"envelope_correlation":0.9894642523240634,"rms_gain_error_db":-0.6798544094024381,"reference_mean_power":0.0009551583992700257,"output_mean_power":0.000816751532637818},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1194144575506277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002166069295810575,"output_power":0.00017912667725652361,"reference_active":true,"delta_db":-0.8251207629417612},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38201712334803e-05,"output_power":1.9761642086058428e-05,"reference_active":true,"delta_db":-0.8112184983941062},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.302931519380677e-06,"output_power":7.336761344714533e-06,"reference_active":true,"delta_db":-0.5372706370864679}],"active_band_count":3,"active_band_mae_db":0.7245366328074452}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":1.3287070635847675,"rms_envelope_nrmse":0.09293976269769125,"envelope_correlation":0.9966577000742876,"rms_gain_error_db":-0.679853679990433,"reference_mean_power":0.0009599620519447541,"output_mean_power":0.0008208592517043469},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.73215395386014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001986653950491428,"output_power":0.00016591276243128674,"reference_active":true,"delta_db":-0.7824243085277628},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916927878865493e-05,"output_power":9.046001034856682e-06,"reference_active":true,"delta_db":-0.8164380811761189},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.81959177520348e-07,"output_power":5.067963862444631e-07,"reference_active":true,"delta_db":-0.6005901197257536}],"active_band_count":3,"active_band_mae_db":0.7331508364765451}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":2.376520730953004,"rms_envelope_nrmse":0.17770189414327756,"envelope_correlation":0.9821951653203507,"rms_gain_error_db":-0.9691233901587049,"reference_mean_power":0.0006127358569067375,"output_mean_power":0.0004901860601647025},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.366962347376557e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019364231078177667,"output_power":0.00013035009810542515,"reference_active":true,"delta_db":-1.7188889417603481},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9545968551344243e-05,"output_power":1.3006306713879154e-05,"reference_active":true,"delta_db":-1.769032043760919},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.863910891753386e-06,"output_power":4.805782448613447e-06,"reference_active":true,"delta_db":-1.5480752982717116}],"active_band_count":3,"active_band_mae_db":1.6786654279309932}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":1.9636413722233153,"rms_envelope_nrmse":0.14970594058029044,"envelope_correlation":0.9896278036931718,"rms_gain_error_db":-0.9691231312759037,"reference_mean_power":0.000615817475654304,"output_mean_power":0.0004926513713260194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.097820742933078e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018026482035346189,"output_power":0.0001221471242358108,"reference_active":true,"delta_db":-1.6902773345278472},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.739141870245849e-06,"output_power":6.3449662010839434e-06,"reference_active":true,"delta_db":-1.390394766929258},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.76389261167761e-07,"output_power":3.7900397213927387e-07,"reference_active":true,"delta_db":-1.8207211918181305}],"active_band_count":3,"active_band_mae_db":1.6337977644250785}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":3.453186191950063,"rms_envelope_nrmse":0.31739633368933867,"envelope_correlation":0.9280656647158657,"rms_gain_error_db":-0.8427833636197474,"reference_mean_power":5.214627438247733e-05,"output_mean_power":4.2948198245882014e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.666296423769005e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7235548052453483e-05,"output_power":2.1714003550105382e-05,"reference_active":true,"delta_db":-0.9839621437340609},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163900077209868e-06,"output_power":4.561208223651519e-06,"reference_active":true,"delta_db":0.3957959964104225},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1939677607224186e-06,"output_power":1.0157512680945794e-06,"reference_active":true,"delta_db":-0.7020522703418418}],"active_band_count":3,"active_band_mae_db":0.6939368034954417}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":3.1835011441475305,"rms_envelope_nrmse":0.27213384928916595,"envelope_correlation":0.94608316071185,"rms_gain_error_db":-0.8427821017427566,"reference_mean_power":5.2408520364582374e-05,"output_mean_power":4.316419922706119e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.333048787706058e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.583453225376263e-05,"output_power":2.0757839374113937e-05,"reference_active":true,"delta_db":-0.9501845582737332},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.076540066109676e-06,"output_power":2.189748387297884e-06,"reference_active":true,"delta_db":0.2305390008027747},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.119613733354675e-08,"output_power":6.365156868800233e-08,"reference_active":true,"delta_db":0.1708510069321812}],"active_band_count":3,"active_band_mae_db":0.4505248553362297}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":5.486501101228893,"rms_envelope_nrmse":0.3660563708623539,"envelope_correlation":0.9075558225800171,"rms_gain_error_db":-0.4834899540352166,"reference_mean_power":6.886139457761318e-06,"output_mean_power":6.1606540015665774e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.997833564030977e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1829042042252137e-06,"output_power":2.656219235706641e-06,"reference_active":true,"delta_db":0.8522924012280763},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11137684368614e-07,"output_power":8.445253105771106e-07,"reference_active":true,"delta_db":0.17518091378060455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6537364416157555e-07,"output_power":1.8570704115458116e-07,"reference_active":true,"delta_db":0.5036207390840004}],"active_band_count":3,"active_band_mae_db":0.5103646846975604}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":5.383417749537605,"rms_envelope_nrmse":0.31516670672157454,"envelope_correlation":0.9300788251302166,"rms_gain_error_db":-0.4834896554512816,"reference_mean_power":6.9207717017349125e-06,"output_mean_power":6.191638010102794e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.44334903190685e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.931899455537554e-06,"output_power":2.4483035926641702e-06,"reference_active":true,"delta_db":1.0288074977282269},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0565449907563406e-07,"output_power":3.125195965158058e-07,"reference_active":true,"delta_db":-1.1327904212556141},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.9695465540608e-09,"output_power":1.4372919502938944e-08,"reference_active":true,"delta_db":2.5611138137071157}],"active_band_count":3,"active_band_mae_db":1.574237244230319}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"golf","status":"measured","mos":{"status":"measured","value":3.533273458480835,"predicted_mos":3.533273458480835,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"a5b6c02decdf13deb16710affcab5bfe8fe590b221a0abeb21378fc6b392274e","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7771397790002084},"started_at":"2026-10-13T05:02:43.955937+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035195520003981073,"native_input_samples":144000,"native_input_sha256":"cc6c79e28e64fced7a0d657ed91b5e08b9958cb5d9d7385f47867c5755ecc07a","output_sha256":"05651e69f2d470585c6aff01fdcb1eb16c68a4cd874e72fec5ad627406c1b0d2","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5869764009994469,"synthesis_s":0.1244944560003205,"total_s":0.7114708569997674,"rtf":0.11857847616662791,"output_peak":0.22166851162910461,"output_rms":0.024434583182885328,"finished_at":"2026-10-13T05:02:44.679246+00:00","analysis_stages":{"external_f0_s":0.08836085699931573},"evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"05651e69f2d470585c6aff01fdcb1eb16c68a4cd874e72fec5ad627406c1b0d2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-32.23534704941767,"output_peak":0.21626658593113968},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.03730020450895,"spectral_convergence":0.5473090682002054,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.1708838760313265,"spectral_convergence":0.3689635145759731,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.056843053787395,"spectral_convergence":0.3778315065899565,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00012350822291436943,"output_power":9.557688160564e-05,"delta_db":-1.1134301630363157,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00039415167446603215,"output_power":0.00026709896697388346,"delta_db":-1.6899116754988295,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003588136748531669,"output_power":0.00019109130742206264,"delta_db":-2.7362805438700115,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.957424369701976e-05,"output_power":3.4382767222007886e-05,"delta_db":-2.3871771062547578,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0495253614878532e-05,"output_power":6.251311743141956e-06,"delta_db":-2.2502178062075915,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.412557189863337e-06,"output_power":3.274046935398745e-06,"delta_db":-4.58622726962815,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.4605407607492755,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-2.039724387333038,"pitch":null,"output_rms_dbfs":-32.23534704941767,"output_peak":0.21626658593113968,"reference_rms_dbfs":-30.195622662084634,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.421675711442558},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8556334967070245,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030918532092182677,"rms_dbfs":-30.195622662084634,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.843562594481707,"mean":-8.826661786571369e-07,"zero_energy":false},"output":{"rms":0.024447398264480714,"rms_dbfs":-32.23534704941767,"peak":0.21626658593113968,"peak_dbfs":-13.300211512902194,"crest_factor_db":18.935135536515478,"mean":0.0002848111988364717,"zero_energy":false},"rms_difference_db":-2.0397243873330346},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":126,"quiet_samples":20160,"reference_mean_power":1.5878444722306635e-07,"output_mean_power":3.0901206542435447e-07,"output_rms_dbfs":-65.10024563148238,"output_zero_energy":false,"output_peak":0.0034843726159593134,"output_energy_fraction":0.00010875615988854719,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":252,"both_estimated_voiced_frames":213,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.9551569506726457,"absolute_difference_cents_median":4.310422683579437,"absolute_difference_cents_p95":21.10011757482995,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":3.109409161215495,"rms_envelope_nrmse":0.2578112268437341,"envelope_correlation":0.9718521571221063,"rms_gain_error_db":-2.039726278654147,"reference_mean_power":0.0009567542251115067,"output_mean_power":0.0005981743149840085},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.1181987535005683e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021668665405953318,"output_power":0.0001641069427169094,"reference_active":true,"delta_db":-1.2070520879976232},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3646928116667708e-05,"output_power":1.6916217737463312e-05,"reference_active":true,"delta_db":-1.4547146463656755},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.30245324490828e-06,"output_power":6.004571095012536e-06,"reference_active":true,"delta_db":-1.40724447254701}],"active_band_count":3,"active_band_mae_db":1.356337068970103}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":2.9727159841129587,"rms_envelope_nrmse":0.23871902601123005,"envelope_correlation":0.9825094481383441,"rms_gain_error_db":-2.0397358368209755,"reference_mean_power":0.0009615739723009672,"output_mean_power":0.0006011863561072734},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.7310436867658744e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019866948333910694,"output_power":0.00015292043497676535,"reference_active":true,"delta_db":-1.1366563766855562},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0810004820357939e-05,"output_power":7.525130595266983e-06,"reference_active":true,"delta_db":-1.5731184629510078},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.816622272333975e-07,"output_power":4.2057612389324806e-07,"reference_active":true,"delta_db":-1.4082624776080213}],"active_band_count":3,"active_band_mae_db":1.372679105748195}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":2.8652827896230786,"rms_envelope_nrmse":0.3127778212039747,"envelope_correlation":0.9665997907779555,"rms_gain_error_db":-2.571186297912662,"reference_mean_power":0.0006137596258340648,"output_mean_power":0.00033953119861077427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.365648836364774e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001936920207538399,"output_power":0.000134776830765184,"reference_active":true,"delta_db":-1.5749649023716383},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.946690369482364e-05,"output_power":1.3341295142567438e-05,"reference_active":true,"delta_db":-1.6409888842838347},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.861625969171953e-06,"output_power":4.663379187768945e-06,"reference_active":true,"delta_db":-1.677263105879845}],"active_band_count":3,"active_band_mae_db":1.6310722975117729}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":2.6993625738881897,"rms_envelope_nrmse":0.2933815624876586,"envelope_correlation":0.9775910781926856,"rms_gain_error_db":-2.5711866181663097,"reference_mean_power":0.0006168515634429444,"output_mean_power":0.0003412416301768438},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.0966279994385797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018028642529894354,"output_power":0.00012799660100444991,"reference_active":true,"delta_db":-1.4876459071131816},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.686846596178631e-06,"output_power":6.0552021315174264e-06,"reference_active":true,"delta_db":-1.5673350712754364},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.760645205818156e-07,"output_power":3.5310111589900035e-07,"reference_active":true,"delta_db":-2.125720380931399}],"active_band_count":3,"active_band_mae_db":1.726900453106672}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":3.501151505225125,"rms_envelope_nrmse":0.3174714908946698,"envelope_correlation":0.9594805502638425,"rms_gain_error_db":-2.3483558760371945,"reference_mean_power":5.223339519529336e-05,"output_mean_power":3.0416740211718466e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.66120516599218e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.723381755549791e-05,"output_power":1.600832632172928e-05,"reference_active":true,"delta_db":-2.307625951209547},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.158662056003102e-06,"output_power":2.0618264139890485e-06,"reference_active":true,"delta_db":-3.0470153081201934},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1956487192645045e-06,"output_power":6.634347236870404e-07,"reference_active":true,"delta_db":-2.5580540455227316}],"active_band_count":3,"active_band_mae_db":2.637565101617491}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":3.4570433217795338,"rms_envelope_nrmse":0.27393838380460445,"envelope_correlation":0.9825656619847483,"rms_gain_error_db":-2.3483559265840372,"reference_mean_power":5.249651403096864e-05,"output_mean_power":3.056996017100003e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.3280727933917817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5831414183246877e-05,"output_power":1.530525888103556e-05,"reference_active":true,"delta_db":-2.2730750311360293},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.072279885308017e-06,"output_power":9.656954062776127e-07,"reference_active":true,"delta_db":-3.3160824600313275},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.104684991603001e-08,"output_power":4.09726482632318e-08,"reference_active":true,"delta_db":-1.7316922334764617}],"active_band_count":3,"active_band_mae_db":2.4402832415479394}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":3.2740628162144128,"rms_envelope_nrmse":0.3028884638391051,"envelope_correlation":0.9509552083310234,"rms_gain_error_db":-1.8226816028595538,"reference_mean_power":6.897644729377568e-06,"output_mean_power":4.533489993932354e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.008016672179218e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1826652525735066e-06,"output_power":1.5822287973568008e-06,"reference_active":true,"delta_db":-1.3971785002585135},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.131748323789655e-07,"output_power":7.33360684301796e-07,"reference_active":true,"delta_db":-0.44866305250429644},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6548084630445296e-07,"output_power":1.5476867575450324e-07,"reference_active":true,"delta_db":-0.29064666644223713}],"active_band_count":3,"active_band_mae_db":0.7121627397350156}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":3.249978368753844,"rms_envelope_nrmse":0.2571916472713275,"envelope_correlation":0.9727995206025959,"rms_gain_error_db":-1.822681725483235,"reference_mean_power":6.932393213320881e-06,"output_mean_power":4.5563283718235875e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.4519485137569075e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9312369318097166e-06,"output_power":1.4501688292420056e-06,"reference_active":true,"delta_db":-1.2441699210058912},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0644178486729243e-07,"output_power":3.562047437673463e-07,"reference_active":true,"delta_db":-0.5729865098184902},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.984940696021735e-09,"output_power":8.077146422336266e-09,"reference_active":true,"delta_db":0.04986260842746683}],"active_band_count":3,"active_band_mae_db":0.6223396797506161}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3822848796844482,"predicted_mos":3.3822848796844482,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"9688f1c5d28dd0af4e9a3e05a71719f61d85deacc0d3f03155e6108acaf881e1","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":31.503420659002586},"started_at":"2026-10-13T07:40:49.090605+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028078569994249847,"native_input_samples":144000,"native_input_sha256":"cc6c79e28e64fced7a0d657ed91b5e08b9958cb5d9d7385f47867c5755ecc07a","output_sha256":"9e28227a8bd44c819e39caa8abed9f53f2d80fe9d37c2e52c742c374774b3508","output_samples":144120,"length_delta_samples":120,"analysis_s":2.031929011998727,"synthesis_s":3.5199496050008747,"total_s":5.551878616999602,"rtf":0.925313102833267,"output_peak":0.22140547633171082,"output_rms":0.02960482022281043,"finished_at":"2026-10-13T07:40:54.653298+00:00","analysis_stages":{"external_f0_s":1.225320676996489},"evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"9e28227a8bd44c819e39caa8abed9f53f2d80fe9d37c2e52c742c374774b3508","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.570717188032628,"output_peak":0.2220855159834029},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.711860005034207,"spectral_convergence":0.5337212288583687,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.812384180492203,"spectral_convergence":0.2589360804080158,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.585596424251003,"spectral_convergence":0.26440441758879846,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00015061029714954797,"delta_db":0.868829971877805,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.0003471144629854871,"delta_db":-0.5446631203985034,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00030606889519360636,"delta_db":-0.6832532091507977,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":5.6616950317337594e-05,"delta_db":-0.2138764327678527,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":1.0908425487676486e-05,"delta_db":0.174935427175804,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":6.2676502440177924e-06,"delta_db":-1.758784304883837,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7073904110424333,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.36423326264881045,"pitch":null,"output_rms_dbfs":-30.567099764179332,"output_peak":0.2220855159834029,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.036613536592472},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9053367008352783,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030892757470915053,"rms_dbfs":-30.202866501530522,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.8508064339276,"mean":-8.862813313802083e-07,"zero_energy":false},"output":{"rms":0.029624089545126502,"rms_dbfs":-30.567099764179332,"peak":0.2220855159834029,"peak_dbfs":-13.069595288607413,"crest_factor_db":17.497504475571922,"mean":1.2505244345275258e-05,"zero_energy":false},"rms_difference_db":-0.36423326264880984},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":127,"quiet_samples":20320,"reference_mean_power":1.5799149570209303e-07,"output_mean_power":6.297628773493732e-07,"output_rms_dbfs":-62.00822943314414,"output_zero_energy":false,"output_peak":0.029101955566209863,"output_energy_fraction":0.00015189360990493938,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":251,"both_estimated_voiced_frames":214,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9596412556053812,"absolute_difference_cents_median":5.540264982012647,"absolute_difference_cents_p95":26.112782037225948,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":1.9430735415807139,"rms_envelope_nrmse":0.15151341724496498,"envelope_correlation":0.9778143381780678,"rms_gain_error_db":-0.36423320224918726,"reference_mean_power":0.0009551583992700257,"output_mean_power":0.0008783185979592946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1194144575506277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002166069295810575,"output_power":0.00021247575321334966,"reference_active":true,"delta_db":-0.08362968782573538},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38201712334803e-05,"output_power":2.2245469700024363e-05,"reference_active":true,"delta_db":-0.297032990698712},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.302931519380677e-06,"output_power":6.180281806506404e-06,"reference_active":true,"delta_db":-1.282231776189271}],"active_band_count":3,"active_band_mae_db":0.5542981515712394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":1.4729824239398042,"rms_envelope_nrmse":0.09876144555855804,"envelope_correlation":0.991055069293646,"rms_gain_error_db":-0.3642328550399588,"reference_mean_power":0.0009599620519447541,"output_mean_power":0.0008827358808872721},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.73215395386014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001986653950491428,"output_power":0.00019443568535661151,"reference_active":true,"delta_db":-0.09346249940385319},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916927878865493e-05,"output_power":1.031422458762375e-05,"reference_active":true,"delta_db":-0.24663857468915987},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.81959177520348e-07,"output_power":4.10869743678008e-07,"reference_active":true,"delta_db":-1.5118836029376763}],"active_band_count":3,"active_band_mae_db":0.6173282256768965}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":2.127357212262865,"rms_envelope_nrmse":0.17611154401450096,"envelope_correlation":0.9784162945969742,"rms_gain_error_db":-0.6633385188048974,"reference_mean_power":0.0006127358569067375,"output_mean_power":0.0005259439258974921},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.366962347376557e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019364231078177667,"output_power":0.00018085262315596708,"reference_active":true,"delta_db":-0.29675444226231373},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9545968551344243e-05,"output_power":1.7365243259252352e-05,"reference_active":true,"delta_db":-0.5137632421328128},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.863910891753386e-06,"output_power":5.083963763239323e-06,"reference_active":true,"delta_db":-1.3036919041585229}],"active_band_count":3,"active_band_mae_db":0.704736529517883}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":1.7566761679903902,"rms_envelope_nrmse":0.1226959163369304,"envelope_correlation":0.9914191644899364,"rms_gain_error_db":-0.6633383201727556,"reference_mean_power":0.000615817475654304,"output_mean_power":0.0005285890681025889},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.097820742933078e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018026482035346189,"output_power":0.00016821726211656046,"reference_active":true,"delta_db":-0.3004041992049701},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.739141870245849e-06,"output_power":7.828002107014255e-06,"reference_active":true,"delta_db":-0.4781785581000981},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.76389261167761e-07,"output_power":4.076192987685875e-07,"reference_active":true,"delta_db":-1.504611432550317}],"active_band_count":3,"active_band_mae_db":0.7610647299517951}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":2.191545899363904,"rms_envelope_nrmse":0.2179997595431269,"envelope_correlation":0.965506642941803,"rms_gain_error_db":-0.14277197677715242,"reference_mean_power":5.214627438247733e-05,"output_mean_power":5.0459865409570725e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.666296423769005e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7235548052453483e-05,"output_power":2.607323482043811e-05,"reference_active":true,"delta_db":-0.18941202871541332},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163900077209868e-06,"output_power":4.255786935252563e-06,"reference_active":true,"delta_db":0.0947957797942637},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1939677607224186e-06,"output_power":1.297668965098137e-06,"reference_active":true,"delta_db":0.36171318002306246}],"active_band_count":3,"active_band_mae_db":0.21530699617757984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":1.349455775298966,"rms_envelope_nrmse":0.11481474545877286,"envelope_correlation":0.9900297969050117,"rms_gain_error_db":-0.14277165773426237,"reference_mean_power":5.2408520364582374e-05,"output_mean_power":5.071363408990446e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.333048787706058e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.583453225376263e-05,"output_power":2.47922061528855e-05,"reference_active":true,"delta_db":-0.17885428344464988},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.076540066109676e-06,"output_power":2.1571105184829006e-06,"reference_active":true,"delta_db":0.1653208144811323},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.119613733354675e-08,"output_power":6.910771683245566e-08,"reference_active":true,"delta_db":0.5280253448549963}],"active_band_count":3,"active_band_mae_db":0.29073348092692614}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":2.0932336082458445,"rms_envelope_nrmse":0.22964491134850698,"envelope_correlation":0.9584813295425572,"rms_gain_error_db":-0.3796388244885761,"reference_mean_power":6.886139457761318e-06,"output_mean_power":6.30974678932836e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.997833564030977e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1829042042252137e-06,"output_power":1.99835334345929e-06,"reference_active":true,"delta_db":-0.38362395847155417},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11137684368614e-07,"output_power":7.69922140014773e-07,"reference_active":true,"delta_db":-0.2264777015255024},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6537364416157555e-07,"output_power":1.3622210478892855e-07,"reference_active":true,"delta_db":-0.8421871023746526}],"active_band_count":3,"active_band_mae_db":0.48409625412390306}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":1.2795323261935823,"rms_envelope_nrmse":0.1254562441942578,"envelope_correlation":0.9877858851646955,"rms_gain_error_db":-0.3796388407983187,"reference_mean_power":6.9207717017349125e-06,"output_mean_power":6.341480176019918e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.44334903190685e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.931899455537554e-06,"output_power":1.799301673241871e-06,"reference_active":true,"delta_db":-0.308805362672013},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0565449907563406e-07,"output_power":3.7707541589576374e-07,"reference_active":true,"delta_db":-0.31728078334255777},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.9695465540608e-09,"output_power":8.686348197028131e-09,"reference_active":true,"delta_db":0.3740362240589972}],"active_band_count":3,"active_band_mae_db":0.333374123357856}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.2157483100891113,"predicted_mos":3.2157483100891113,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"6b4ec8b3aa37cb0be5cea4c4f8c078bb967e502cb4e729568f81f73761fbdea0","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.647753171999284},"started_at":"2026-10-13T05:10:59.543691+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003835003999483888,"native_input_samples":144000,"native_input_sha256":"cc6c79e28e64fced7a0d657ed91b5e08b9958cb5d9d7385f47867c5755ecc07a","output_sha256":"1ed58b01b7adb3b1ebc75c4d763aa0be08c7f325611b796a830167b41caf88cc","output_samples":144240,"length_delta_samples":240,"analysis_s":1.7310144120001496,"synthesis_s":0.2034074999992299,"total_s":1.9344219119993795,"rtf":0.32240365199989657,"output_peak":0.2481878101825714,"output_rms":0.030295787264596006,"finished_at":"2026-10-13T05:11:01.501542+00:00","analysis_stages":{"external_f0_s":0.29123600700040697},"evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"1ed58b01b7adb3b1ebc75c4d763aa0be08c7f325611b796a830167b41caf88cc","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.36930413742793,"output_peak":0.24838826559337854},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.180171551637806,"spectral_convergence":0.1361564411494667,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.981182672157506,"spectral_convergence":0.1308540022567313,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.016495998341712,"spectral_convergence":0.14117711330971894,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00012175962482539296,"delta_db":-0.05468366936373133,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.0003792199120714143,"delta_db":-0.16047888354565817,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00034668523779381925,"delta_db":-0.14209167436661216,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":5.675616360585606e-05,"delta_db":-0.20321083643626311,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":9.095515977392147e-06,"delta_db":-0.614411856312995,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":6.494084053367174e-06,"delta_db":-1.6046527843798697,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.46325495073418826,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15920547815351754,"pitch":null,"output_rms_dbfs":-30.36207197968404,"output_peak":0.24838826559337854,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.0592834073790085},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.962953724358206,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030892757470915053,"rms_dbfs":-30.202866501530522,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.8508064339276,"mean":-8.862813313802083e-07,"zero_energy":false},"output":{"rms":0.030331675494888872,"rms_dbfs":-30.36207197968404,"peak":0.24838826559337854,"peak_dbfs":-12.09737850069979,"crest_factor_db":18.26469347898425,"mean":1.0985849449269882e-05,"zero_energy":false},"rms_difference_db":-0.15920547815351682},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":127,"quiet_samples":20320,"reference_mean_power":1.5799149570209303e-07,"output_mean_power":2.2265005689782776e-07,"output_rms_dbfs":-66.52377189549559,"output_zero_energy":false,"output_peak":0.007048789745487444,"output_energy_fraction":5.122506038071477e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":237,"both_estimated_voiced_frames":218,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9775784753363229,"absolute_difference_cents_median":0.7991618603679977,"absolute_difference_cents_p95":4.794506865479456,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":0.7726422487073625,"rms_envelope_nrmse":0.05826336344730836,"envelope_correlation":0.996918781081407,"rms_gain_error_db":-0.15920542647932145,"reference_mean_power":0.0009551583992700257,"output_mean_power":0.0009207778350046823},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1194144575506277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002166069295810575,"output_power":0.00021605053514111758,"reference_active":true,"delta_db":-0.011169998806728356},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38201712334803e-05,"output_power":1.9620394083307425e-05,"reference_active":true,"delta_db":-0.842371530171291},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.302931519380677e-06,"output_power":4.868853479208333e-06,"reference_active":true,"delta_db":-2.3180475067359065}],"active_band_count":3,"active_band_mae_db":1.0571963452379753}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":0.5276180241129304,"rms_envelope_nrmse":0.03505092953572315,"envelope_correlation":0.9990798177858933,"rms_gain_error_db":-0.15920525746149444,"reference_mean_power":0.0009599620519447541,"output_mean_power":0.0009254086180383943},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.73215395386014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001986653950491428,"output_power":0.0001987723745243295,"reference_active":true,"delta_db":0.0023380061196714617},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916927878865493e-05,"output_power":9.07170607742371e-06,"reference_active":true,"delta_db":-0.8041147054141785},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.81959177520348e-07,"output_power":3.1441489104098303e-07,"reference_active":true,"delta_db":-2.673894148936112}],"active_band_count":3,"active_band_mae_db":1.160115620156654}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":1.0032877427702362,"rms_envelope_nrmse":0.0660821207810539,"envelope_correlation":0.9967675874906249,"rms_gain_error_db":-0.1475944123325585,"reference_mean_power":0.0006127358569067375,"output_mean_power":0.0005922619803879792},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.366962347376557e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019364231078177667,"output_power":0.00018690181407306962,"reference_active":true,"delta_db":-0.1538673987269458},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9545968551344243e-05,"output_power":1.6572385455444994e-05,"reference_active":true,"delta_db":-0.7167216971759743},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.863910891753386e-06,"output_power":3.8313685361399625e-06,"reference_active":true,"delta_db":-2.532177083899602}],"active_band_count":3,"active_band_mae_db":1.1342553932675072}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":0.6306447068454213,"rms_envelope_nrmse":0.036754020195464844,"envelope_correlation":0.9991092505659561,"rms_gain_error_db":-0.14759437751209917,"reference_mean_power":0.000615817475654304,"output_mean_power":0.0005952406350985239},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.097820742933078e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018026482035346189,"output_power":0.0001734817167884778,"reference_active":true,"delta_db":-0.1665726878429102},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.739141870245849e-06,"output_power":7.37214207838258e-06,"reference_active":true,"delta_db":-0.7387509319642104},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.76389261167761e-07,"output_power":2.580456330485171e-07,"reference_active":true,"delta_db":-3.490193669569574}],"active_band_count":3,"active_band_mae_db":1.4651724297922317}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":1.2433206852692344,"rms_envelope_nrmse":0.10800224536599964,"envelope_correlation":0.992198877123905,"rms_gain_error_db":-0.2956393781392141,"reference_mean_power":5.214627438247733e-05,"output_mean_power":4.87146231960105e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.666296423769005e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7235548052453483e-05,"output_power":2.6660371710646383e-05,"reference_active":true,"delta_db":-0.09269918624691212},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163900077209868e-06,"output_power":3.6396786356308745e-06,"reference_active":true,"delta_db":-0.5843725958709021},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1939677607224186e-06,"output_power":7.870557698748151e-07,"reference_active":true,"delta_db":-1.8098709315791464}],"active_band_count":3,"active_band_mae_db":0.8289809045656535}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":0.8388126579765541,"rms_envelope_nrmse":0.06672736108836723,"envelope_correlation":0.9974603938543511,"rms_gain_error_db":-0.29563915169231364,"reference_mean_power":5.2408520364582374e-05,"output_mean_power":4.8959613801268645e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.333048787706058e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.583453225376263e-05,"output_power":2.532240490595723e-05,"reference_active":true,"delta_db":-0.08695653960682423},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.076540066109676e-06,"output_power":1.8090755434099973e-06,"reference_active":true,"delta_db":-0.5988361257042621},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.119613733354675e-08,"output_power":4.4967210939300996e-08,"reference_active":true,"delta_db":-1.3382805899918595}],"active_band_count":3,"active_band_mae_db":0.674691085100982}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":1.2645567658874328,"rms_envelope_nrmse":0.14548158255731047,"envelope_correlation":0.9867707944701846,"rms_gain_error_db":-0.664621349182561,"reference_mean_power":6.886139457761318e-06,"output_mean_power":5.908995766781429e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.997833564030977e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1829042042252137e-06,"output_power":1.7895285869149422e-06,"reference_active":true,"delta_db":-0.8629603686759484},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11137684368614e-07,"output_power":7.706693456217757e-07,"reference_active":true,"delta_db":-0.2222649390840723},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6537364416157555e-07,"output_power":1.118966065380761e-07,"reference_active":true,"delta_db":-1.6964938060547214}],"active_band_count":3,"active_band_mae_db":0.9272397046049141}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":1.002510644777641,"rms_envelope_nrmse":0.10298332773459291,"envelope_correlation":0.9953284221839989,"rms_gain_error_db":-0.6646212638502275,"reference_mean_power":6.9207717017349125e-06,"output_mean_power":5.938713809497656e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.44334903190685e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.931899455537554e-06,"output_power":1.588966771581825e-06,"reference_active":true,"delta_db":-0.8486970473997358},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0565449907563406e-07,"output_power":3.887804026086083e-07,"reference_active":true,"delta_db":-0.1845193188928573},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.9695465540608e-09,"output_power":6.593235511086637e-09,"reference_active":true,"delta_db":-0.8233502283486083}],"active_band_count":3,"active_band_mae_db":0.6188555315470671}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0023","method":"magphase","status":"measured","mos":{"status":"measured","value":2.922083854675293,"predicted_mos":2.922083854675293,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288682,"input_sample_rate":48000,"input_seconds":6.014208333333333,"input_float32_sha256":"4a4fe5ad0791b30ac65fab01de0a36ba91282cd3638c854a5c1abe01bb363253","internal_samples":96228,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7287785799999256},"started_at":"2026-10-13T05:16:36.853067+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009149264998995932,"native_input_samples":288000,"native_input_sha256":"6e69393479dc968b2f9ce5cd42d8f1be11968513849150e1b78ed7aabb38f276","output_sha256":"12eeef4672d41c8a556fbce4b85af4ff81345e49f086f90db30ffbf0f2081f36","output_samples":288682,"length_delta_samples":682,"analysis_s":3.082770042001357,"synthesis_s":0.9995324310002616,"total_s":4.082302473001619,"rtf":0.6803837455002698,"output_peak":0.261781508502317,"output_rms":0.02923666140875902,"finished_at":"2026-10-13T05:16:40.955860+00:00","evaluation_input":{"reference":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"12eeef4672d41c8a556fbce4b85af4ff81345e49f086f90db30ffbf0f2081f36","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96228,"length_delta_samples":228,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.679569765984525,"output_peak":0.26177635216665046},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9976306272602569,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.559702041100752,"spectral_convergence":0.5394248904605699,"spectral_convergence_status":"ok","amplitude_floor":7.733796592545008e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.470888944277183,"spectral_convergence":0.28318446719588053,"spectral_convergence_status":"ok","amplitude_floor":4.871370963739547e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.092527261158308,"spectral_convergence":0.25380328569993604,"spectral_convergence_status":"ok","amplitude_floor":4.7483837933045734e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001233024392514864,"output_power":0.00011359281925969336,"delta_db":-0.3562078975438092,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000393494821723947,"output_power":0.00035168770470304683,"delta_db":-0.4878183626189512,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000358215617892152,"output_power":0.00032013374241106243,"delta_db":-0.4881306516897121,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":5.947495410559829e-05,"output_power":5.397966860013553e-05,"delta_db":-0.42103901637479135,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":1.0477761937191392e-05,"output_power":1.0504448105854333e-05,"delta_db":0.01104713185008616,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":9.39686924640271e-06,"output_power":7.2839755597884225e-06,"delta_db":-1.1061470394721136,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.47839834992491054,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.46640125443001623,"pitch":null,"output_rms_dbfs":-30.66926775596054,"output_peak":0.26177635216665046,"reference_rms_dbfs":-30.202866501530522,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.707706082178748},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8764995075813958,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.030892757470915053,"rms_dbfs":-30.202866501530522,"peak":0.2412109375,"peak_dbfs":-12.352060067602924,"crest_factor_db":17.8508064339276,"mean":-8.862813313802083e-07,"zero_energy":false},"output":{"rms":0.02927767679717059,"rms_dbfs":-30.66926775596054,"peak":0.26177635216665046,"peak_dbfs":-11.641391768951994,"crest_factor_db":19.027875987008546,"mean":-0.00022882283823639327,"zero_energy":false},"rms_difference_db":-0.4664012544300178},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":127,"quiet_samples":20320,"reference_mean_power":1.5799149570209303e-07,"output_mean_power":1.840304567098137e-07,"output_rms_dbfs":-67.35110296086054,"output_zero_energy":false,"output_peak":0.004025048048561536,"output_energy_fraction":4.544320463936328e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":223,"output_estimated_voiced_frames":217,"both_estimated_voiced_frames":202,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.905829596412556,"absolute_difference_cents_median":3.7162289539905293,"absolute_difference_cents_p95":26.02797718007502,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.933633747331572e-05,"log_rms_mae_db":1.741086028464226,"rms_envelope_nrmse":0.17314482904251577,"envelope_correlation":0.9715136137984162,"rms_gain_error_db":-0.46640115785072134,"reference_mean_power":0.0009551583992700257,"output_mean_power":0.0008578972649469221},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.1194144575506277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002166069295810575,"output_power":0.00019989789017313677,"reference_active":true,"delta_db":-0.3486413587316295},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.38201712334803e-05,"output_power":2.1975651466477303e-05,"reference_active":true,"delta_db":-0.35003120560837536},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.302931519380677e-06,"output_power":6.848496604344686e-06,"reference_active":true,"delta_db":-0.8363621121667281}],"active_band_count":3,"active_band_mae_db":0.5116782255022443}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.652181544798992e-05,"log_rms_mae_db":1.1010923581574417,"rms_envelope_nrmse":0.10790658785125429,"envelope_correlation":0.990047585721326,"rms_gain_error_db":-0.46640115094725443,"reference_mean_power":0.0009599620519447541,"output_mean_power":0.0008622117763448769},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.73215395386014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001986653950491428,"output_power":0.0001838283490319541,"reference_active":true,"delta_db":-0.3370973839323659},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916927878865493e-05,"output_power":1.0083128025271274e-05,"reference_active":true,"delta_db":-0.3450515975818016},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.81959177520348e-07,"output_power":5.016142941107351e-07,"reference_active":true,"delta_db":-0.6452261767096928}],"active_band_count":3,"active_band_mae_db":0.44245838607462007}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.027389787958093e-05,"log_rms_mae_db":2.1975559774754725,"rms_envelope_nrmse":0.19270050083859286,"envelope_correlation":0.9720996514822547,"rms_gain_error_db":-0.5091685963081416,"reference_mean_power":0.0006127358569067375,"output_mean_power":0.0005449497228902633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.366962347376557e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019364231078177667,"output_power":0.00017468959531013282,"reference_active":true,"delta_db":-0.44733217811229564},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9545968551344243e-05,"output_power":1.7476439631880484e-05,"reference_active":true,"delta_db":-0.48604234521087886},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.863910891753386e-06,"output_power":5.482934150799984e-06,"reference_active":true,"delta_db":-0.975586067012904}],"active_band_count":3,"active_band_mae_db":0.6363201967786928}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.943169310162246e-05,"log_rms_mae_db":1.4719181064937965,"rms_envelope_nrmse":0.12094242738656645,"envelope_correlation":0.9900858021400165,"rms_gain_error_db":-0.509168717396543,"reference_mean_power":0.000615817475654304,"output_mean_power":0.0005476904110788605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.097820742933078e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018026482035346189,"output_power":0.00016289620539194596,"reference_active":true,"delta_db":-0.44000012400222865},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.739141870245849e-06,"output_power":7.763650866584553e-06,"reference_active":true,"delta_db":-0.5140279287360071},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.76389261167761e-07,"output_power":4.5315444040165653e-07,"reference_active":true,"delta_db":-1.0446964083829062}],"active_band_count":3,"active_band_mae_db":0.6662414870403807}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.055544735899722e-05,"log_rms_mae_db":2.2175278369363443,"rms_envelope_nrmse":0.24695455329130142,"envelope_correlation":0.9555513309927326,"rms_gain_error_db":-0.3464675453970322,"reference_mean_power":5.214627438247733e-05,"output_mean_power":4.8147809211080236e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.666296423769005e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.7235548052453483e-05,"output_power":2.5139610602335808e-05,"reference_active":true,"delta_db":-0.3477757243173476},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.163900077209868e-06,"output_power":4.096980905711866e-06,"reference_active":true,"delta_db":-0.0703635912793138},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1939677607224186e-06,"output_power":1.2763711662271179e-06,"reference_active":true,"delta_db":0.28984384514072026}],"active_band_count":3,"active_band_mae_db":0.23599438691246055}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.291047760123657e-05,"log_rms_mae_db":1.2862146688287481,"rms_envelope_nrmse":0.14953443895470372,"envelope_correlation":0.9835269852699335,"rms_gain_error_db":-0.34646842198708,"reference_mean_power":5.2408520364582374e-05,"output_mean_power":4.8389936963109056e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.333048787706058e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.583453225376263e-05,"output_power":2.3957267111452793e-05,"reference_active":true,"delta_db":-0.32763327784033747},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.076540066109676e-06,"output_power":2.0782018534055138e-06,"reference_active":true,"delta_db":0.003474127178755295},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.119613733354675e-08,"output_power":6.150329096294563e-08,"reference_active":true,"delta_db":0.021743443606623558}],"active_band_count":3,"active_band_mae_db":0.11761694954190544}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.3449414786854301e-05,"log_rms_mae_db":2.2364792785411884,"rms_envelope_nrmse":0.24988651817209193,"envelope_correlation":0.9529856992166784,"rms_gain_error_db":0.14177377855812068,"reference_mean_power":6.886139457761318e-06,"output_mean_power":7.11464429047893e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.997833564030977e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1829042042252137e-06,"output_power":2.3660619700120342e-06,"reference_active":true,"delta_db":0.349914378280963},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.11137684368614e-07,"output_power":8.698570869035588e-07,"reference_active":true,"delta_db":0.3035332749565902},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6537364416157555e-07,"output_power":2.0848028220512868e-07,"reference_active":true,"delta_db":1.005986896553871}],"active_band_count":3,"active_band_mae_db":0.5531448499304747}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.029139065396281e-05,"log_rms_mae_db":1.3693564435474166,"rms_envelope_nrmse":0.14363007556531127,"envelope_correlation":0.9834876969752778,"rms_gain_error_db":0.1417737791093736,"reference_mean_power":6.9207717017349125e-06,"output_mean_power":7.150425747507863e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.44334903190685e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.931899455537554e-06,"output_power":2.113163908812298e-06,"reference_active":true,"delta_db":0.3894866457685344},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.0565449907563406e-07,"output_power":4.2266908940600766e-07,"reference_active":true,"delta_db":0.17844191149725966},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.9695465540608e-09,"output_power":1.1099871005971516e-08,"reference_active":true,"delta_db":1.4388431988325943}],"active_band_count":3,"active_band_mae_db":0.6689239186994628}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"hnm","status":"measured","mos":{"status":"measured","value":3.2204713821411133,"predicted_mos":3.2204713821411133,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a43f69d46584aa27434ae0914edfd8b1aeba57f1b9539947ce49cd1649efad0a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.085926498000845},"started_at":"2026-10-13T04:39:40.008429+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6164999578904826e-05,"native_input_samples":96000,"native_input_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","output_sha256":"d5024fea1835be060eff22db66af11b9faa25f6e1ae5ed932a02d91fde747ef7","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9768900520002717,"synthesis_s":0.49154000699945755,"total_s":1.4684300589997292,"rtf":0.24473834316662155,"output_peak":0.20481438213340805,"output_rms":0.021972741902886752,"finished_at":"2026-10-13T04:39:41.484338+00:00","evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"d5024fea1835be060eff22db66af11b9faa25f6e1ae5ed932a02d91fde747ef7","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.39239903997546,"spectral_convergence":0.2382348183825864,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.154069469730499,"spectral_convergence":0.2299577865215379,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.55206506578145,"spectral_convergence":0.2447000376656896,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.748956021680024e-05,"delta_db":0.16377998585189735,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.00029835716516061876,"delta_db":-0.1350713756939683,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":5.256171269217274e-05,"delta_db":-0.06741076026392628,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":2.0451161379590112e-05,"delta_db":-0.22040675749232852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":2.6084982304491982e-05,"delta_db":-0.5208763650889635,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":7.856804977201333e-06,"delta_db":-0.08802010308021907,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19926089124521715,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10600006406840333,"pitch":null,"output_rms_dbfs":-33.1623149116644,"output_peak":0.20481438213340805,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.366177858495803},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.39239903997546,"spectral_convergence":0.2382348183825864,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.154069469730499,"spectral_convergence":0.2299577865215379,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.55206506578145,"spectral_convergence":0.2447000376656896,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.748956021680024e-05,"delta_db":0.16377998585189735,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.00029835716516061876,"delta_db":-0.1350713756939683,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":5.256171269217274e-05,"delta_db":-0.06741076026392628,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":2.0451161379590112e-05,"delta_db":-0.22040675749232852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":2.6084982304491982e-05,"delta_db":-0.5208763650889635,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":7.856804977201333e-06,"delta_db":-0.08802010308021907,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19926089124521715,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10600006406840333,"pitch":null,"output_rms_dbfs":-33.1623149116644,"output_peak":0.20481438213340805,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.366177858495803},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9007770630205475,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.022242533719115894,"rms_dbfs":-33.056314847596,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.991136769433584,"mean":-1.7630259195963542e-06,"zero_energy":false},"output":{"rms":0.021972741902886752,"rms_dbfs":-33.1623149116644,"peak":0.20481438213340805,"peak_dbfs":-13.772791006475792,"crest_factor_db":19.38952390518861,"mean":-1.3574485187709653e-05,"zero_energy":false},"rms_difference_db":-0.10600006406840379},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":1.7553132705150872e-07,"output_rms_dbfs":-67.5564536380513,"output_zero_energy":false,"output_peak":0.007849580101206807,"output_energy_fraction":9.937536230432732e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":139,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9144736842105263,"absolute_difference_cents_median":1.7853870619101186,"absolute_difference_cents_p95":8.746751959087296,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":0.7445658418615381,"rms_envelope_nrmse":0.07036898277458262,"envelope_correlation":0.9956306128228241,"rms_gain_error_db":-0.10595524203239172,"reference_mean_power":0.0004951134756771775,"output_mean_power":0.0004831803039059556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.645854738282274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010486873457313227,"output_power":0.00010056963997700606,"reference_active":true,"delta_db":-0.18179132077135873},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5071002541192514e-05,"output_power":2.5855260777907913e-05,"reference_active":true,"delta_db":0.13377221546827145},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059292105678353e-06,"output_power":5.053234927556807e-06,"reference_active":true,"delta_db":-0.005202654766793935}],"active_band_count":3,"active_band_mae_db":0.10692206366880803}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":0.42282471814415185,"rms_envelope_nrmse":0.03667787310219672,"envelope_correlation":0.9988384324735742,"rms_gain_error_db":-0.1057477942079197,"reference_mean_power":0.0004970274687625529,"output_mean_power":0.0004850713358038274},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4721291240539383e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.182773781582673e-05,"output_power":8.86977806611716e-05,"reference_active":true,"delta_db":-0.15061132236649932},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2533326164459844e-05,"output_power":1.2959333226806297e-05,"reference_active":true,"delta_db":0.1451631534071063},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.472370238880856e-07,"output_power":2.491389070367326e-07,"reference_active":true,"delta_db":0.03328047744576025}],"active_band_count":3,"active_band_mae_db":0.10968498440645529}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":1.457548746067171,"rms_envelope_nrmse":0.09514465145105595,"envelope_correlation":0.9934688260044046,"rms_gain_error_db":-0.1074814799353696,"reference_mean_power":0.00026104952136561866,"output_mean_power":0.0002546682202685374},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.392840173310503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.633554932680893e-05,"output_power":9.217604979745555e-05,"reference_active":true,"delta_db":-0.19168485661380655},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.160826072139944e-05,"output_power":2.2313608713303695e-05,"reference_active":true,"delta_db":0.13950001783372287},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769645916048805e-06,"output_power":4.476083387008239e-06,"reference_active":true,"delta_db":-0.27587970937772754}],"active_band_count":3,"active_band_mae_db":0.20235486127508565}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":0.8453368447115174,"rms_envelope_nrmse":0.05246309505598904,"envelope_correlation":0.9979796402264046,"rms_gain_error_db":-0.10686933377601583,"reference_mean_power":0.00026200159904973233,"output_mean_power":0.0002556330540279782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2678124187221727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604417147270571e-05,"output_power":8.307363743519077e-05,"reference_active":true,"delta_db":-0.15258230323009364},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1040946271380912e-05,"output_power":1.1359051054214446e-05,"reference_active":true,"delta_db":0.12335755046393283},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9369342076961927e-07,"output_power":2.561675110828306e-07,"reference_active":true,"delta_db":-0.5937016899698346}],"active_band_count":3,"active_band_mae_db":0.2898805145546204}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":1.529818370089994,"rms_envelope_nrmse":0.17244780278692232,"envelope_correlation":0.9786425383347112,"rms_gain_error_db":-0.3905205113500152,"reference_mean_power":4.057559706591337e-05,"output_mean_power":3.708624542523905e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.7364679678483124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.645857845256653e-05,"output_power":1.4807993598492915e-05,"reference_active":true,"delta_db":-0.45896103954550727},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1440018553357532e-06,"output_power":2.301391663476418e-06,"reference_active":true,"delta_db":-1.3549225807351997},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.079905260013692e-07,"output_power":7.411083026985378e-07,"reference_active":true,"delta_db":-0.8819963837234861}],"active_band_count":3,"active_band_mae_db":0.8986266680013978}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":0.8915599378075254,"rms_envelope_nrmse":0.10693544885849726,"envelope_correlation":0.9922464744022932,"rms_gain_error_db":-0.39365785222471195,"reference_mean_power":4.071142209359619e-05,"output_mean_power":3.7183518944027845e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5164519386807125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5218014411010497e-05,"output_power":1.3696053506769677e-05,"reference_active":true,"delta_db":-0.4576254702153634},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2444589918408902e-06,"output_power":1.053975797895292e-06,"reference_active":true,"delta_db":-0.7214995162058052},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241409422416778e-08,"output_power":4.3069974382407165e-08,"reference_active":true,"delta_db":-0.8527347164063317}],"active_band_count":3,"active_band_mae_db":0.6772865676091667}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":1.6216910312944854,"rms_envelope_nrmse":0.17736454487514258,"envelope_correlation":0.975255986286772,"rms_gain_error_db":-0.2981038260051573,"reference_mean_power":9.786447369332921e-06,"output_mean_power":9.137232635116504e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.593718108385403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.378843387910107e-06,"output_power":2.192344873473102e-06,"reference_active":true,"delta_db":-0.3545697783068257},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.926849454092712e-07,"output_power":7.849479886013069e-07,"reference_active":true,"delta_db":-0.5585732978826672},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.698308970455951e-07,"output_power":1.1597240110885968e-07,"reference_active":true,"delta_db":-1.656620546015659}],"active_band_count":3,"active_band_mae_db":0.856587874068384}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":0.9599711968742607,"rms_envelope_nrmse":0.09535500652629425,"envelope_correlation":0.9928241083132254,"rms_gain_error_db":-0.3004106040522582,"reference_mean_power":9.831708602072054e-06,"output_mean_power":9.174616881200978e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.220673114788623e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.414723334287998e-06,"output_power":2.2052371858213083e-06,"reference_active":true,"delta_db":-0.39412071704453144},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7015119113313933e-07,"output_power":3.655489663683709e-07,"reference_active":true,"delta_db":-0.05433591039714661},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1816168225696594e-08,"output_power":7.086013929967577e-09,"reference_active":true,"delta_db":-2.2207466360703125}],"active_band_count":3,"active_band_mae_db":0.8897344211706635}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.705435276031494,"predicted_mos":3.705435276031494,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f4d18b66dc5fe9f3b128f907f6a4883d5fdb8dc43365f5a1b698f0280b21849e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.51932597799896},"started_at":"2026-10-13T04:42:42.045271+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0008656640002300264,"native_input_samples":96000,"native_input_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","output_sha256":"fee5eca946fa1cbe7fadc58439d3e6250504912eb3dcd1933323d4481a566163","output_samples":96000,"length_delta_samples":0,"analysis_s":2.216710771999715,"synthesis_s":0.08069381499990413,"total_s":2.297404586999619,"rtf":0.3829007644999365,"output_peak":0.2932275318321913,"output_rms":0.026037367791006242,"finished_at":"2026-10-13T04:42:44.362822+00:00","evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"fee5eca946fa1cbe7fadc58439d3e6250504912eb3dcd1933323d4481a566163","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.214908021236943,"spectral_convergence":0.6481573926002068,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.787786387121878,"spectral_convergence":0.30438398159007674,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.324238877486447,"spectral_convergence":0.33470810698960635,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.167683669702996e-05,"delta_db":-0.17486367101316885,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.000467597303885869,"delta_db":1.8162838878161045,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":7.181678976471411e-05,"delta_db":1.2881540407442957,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":2.5458748379266522e-05,"delta_db":0.7307839726692493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":3.2998615405648273e-05,"delta_db":0.5001753467524355,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":8.396227351601416e-06,"delta_db":0.20036206357385467,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7851038304281848,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.368256402798674,"pitch":null,"output_rms_dbfs":-31.688058444797328,"output_peak":0.2932275318321913,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.442311095281756},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.214908021236943,"spectral_convergence":0.6481573926002068,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.787786387121878,"spectral_convergence":0.30438398159007674,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.324238877486447,"spectral_convergence":0.33470810698960635,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.167683669702996e-05,"delta_db":-0.17486367101316885,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.000467597303885869,"delta_db":1.8162838878161045,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":7.181678976471411e-05,"delta_db":1.2881540407442957,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":2.5458748379266522e-05,"delta_db":0.7307839726692493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":3.2998615405648273e-05,"delta_db":0.5001753467524355,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":8.396227351601416e-06,"delta_db":0.20036206357385467,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7851038304281848,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.368256402798674,"pitch":null,"output_rms_dbfs":-31.688058444797328,"output_peak":0.2932275318321913,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.442311095281756},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9429131875840753,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.022242533719115894,"rms_dbfs":-33.056314847596,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.991136769433584,"mean":-1.7630259195963542e-06,"zero_energy":false},"output":{"rms":0.026037367791006242,"rms_dbfs":-31.688058444797328,"peak":0.2932275318321913,"peak_dbfs":-10.655905103438943,"crest_factor_db":21.03215334135838,"mean":1.162836055483807e-06,"zero_energy":false},"rms_difference_db":1.3682564027986714},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":2.1307394384393214e-07,"output_rms_dbfs":-66.71469655572865,"output_zero_energy":false,"output_peak":0.010512429526635433,"output_energy_fraction":8.59070461840215e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":149,"both_estimated_voiced_frames":129,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.8486842105263158,"absolute_difference_cents_median":5.846987685125802,"absolute_difference_cents_p95":19.916722652244722,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":1.4124749319160945,"rms_envelope_nrmse":0.2365979391230172,"envelope_correlation":0.9802583697217748,"rms_gain_error_db":1.3684609633706475,"reference_mean_power":0.0004951134756771775,"output_mean_power":0.0006785015485245766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.645854738282274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010486873457313227,"output_power":0.0001499437306128532,"reference_active":true,"delta_db":1.5528228442393133},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5071002541192514e-05,"output_power":3.85888452025157e-05,"reference_active":true,"delta_db":1.8729008130073814},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059292105678353e-06,"output_power":6.638363608308361e-06,"reference_active":true,"delta_db":1.1797128191337893}],"active_band_count":3,"active_band_mae_db":1.5351454921268282}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":0.9958606363850314,"rms_envelope_nrmse":0.1999052903213684,"envelope_correlation":0.991938091299439,"rms_gain_error_db":1.3687372159967874,"reference_mean_power":0.0004970274687625529,"output_mean_power":0.0006811678044551124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4721291240539383e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.182773781582673e-05,"output_power":0.00013233635954466185,"reference_active":true,"delta_db":1.5870529785058265},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2533326164459844e-05,"output_power":2.0448334551262617e-05,"reference_active":true,"delta_db":2.125916001466141},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.472370238880856e-07,"output_power":2.9210490944781636e-07,"reference_active":true,"delta_db":0.7242534914458305}],"active_band_count":3,"active_band_mae_db":1.479074157139266}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":1.7050032964285295,"rms_envelope_nrmse":0.3088786435143806,"envelope_correlation":0.9807064693054531,"rms_gain_error_db":1.919140559864694,"reference_mean_power":0.00026104952136561866,"output_mean_power":0.00040610371019371194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.392840173310503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.633554932680893e-05,"output_power":0.00015483307006167039,"reference_active":true,"delta_db":2.0607714698608954},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.160826072139944e-05,"output_power":3.517515067881398e-05,"reference_active":true,"delta_db":2.1161615515559666},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769645916048805e-06,"output_power":6.5053473006827275e-06,"reference_active":true,"delta_db":1.347843476361027}],"active_band_count":3,"active_band_mae_db":1.841592165925963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":1.2814382202031045,"rms_envelope_nrmse":0.2804518378129295,"envelope_correlation":0.9898850514898668,"rms_gain_error_db":1.9192069867883486,"reference_mean_power":0.00026200159904973233,"output_mean_power":0.00040759105142448393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2678124187221727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604417147270571e-05,"output_power":0.0001387397768528574,"reference_active":true,"delta_db":2.0747953456973223},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1040946271380912e-05,"output_power":1.8368023745314615e-05,"reference_active":true,"delta_db":2.2105613567825344},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9369342076961927e-07,"output_power":2.88314919741786e-07,"reference_active":true,"delta_db":-0.08027100826646727}],"active_band_count":3,"active_band_mae_db":1.4552092369154412}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":1.5671670609577353,"rms_envelope_nrmse":0.1858789804138408,"envelope_correlation":0.9811311226582946,"rms_gain_error_db":0.6846697750890441,"reference_mean_power":4.057559706591337e-05,"output_mean_power":4.750418774393922e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.7364679678483124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.645857845256653e-05,"output_power":1.9361519173197105e-05,"reference_active":true,"delta_db":0.705471086350984},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1440018553357532e-06,"output_power":3.25561040668077e-06,"reference_active":true,"delta_db":0.15149634397096368},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.079905260013692e-07,"output_power":7.952297393855638e-07,"reference_active":true,"delta_db":-0.575887040024607}],"active_band_count":3,"active_band_mae_db":0.4776181567821849}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":0.904238782380119,"rms_envelope_nrmse":0.10382482188450075,"envelope_correlation":0.9970693601509989,"rms_gain_error_db":0.6848357997008527,"reference_mean_power":4.071142209359619e-05,"output_mean_power":4.766502805586264e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5164519386807125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5218014411010497e-05,"output_power":1.7824955484512885e-05,"reference_active":true,"delta_db":0.6867046290299856},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2444589918408902e-06,"output_power":1.3480947155430509e-06,"reference_active":true,"delta_db":0.34739816240643534},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241409422416778e-08,"output_power":4.5193931106254746e-08,"reference_active":true,"delta_db":-0.643679658536439}],"active_band_count":3,"active_band_mae_db":0.55926081665762}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":1.7305357089911462,"rms_envelope_nrmse":0.24443617257060016,"envelope_correlation":0.9533414500797472,"rms_gain_error_db":0.08866843968098434,"reference_mean_power":9.786447369332921e-06,"output_mean_power":9.988307609035698e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.593718108385403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.378843387910107e-06,"output_power":2.4282956760244675e-06,"reference_active":true,"delta_db":0.0893571549461174},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.926849454092712e-07,"output_power":8.255915544660014e-07,"reference_active":true,"delta_db":-0.33932969108239425},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.698308970455951e-07,"output_power":1.2003514363607568e-07,"reference_active":true,"delta_db":-1.5070828706377553}],"active_band_count":3,"active_band_mae_db":0.645256572222089}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":1.0693529930027115,"rms_envelope_nrmse":0.12753201625852767,"envelope_correlation":0.9860191470768744,"rms_gain_error_db":0.08858694424788466,"reference_mean_power":9.831708602072054e-06,"output_mean_power":1.003431412713471e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.220673114788623e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.414723334287998e-06,"output_power":2.423144034438485e-06,"reference_active":true,"delta_db":0.01511850996965885},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7015119113313933e-07,"output_power":3.80697586688934e-07,"reference_active":true,"delta_db":0.12200972487187865},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1816168225696594e-08,"output_power":6.929336569367118e-09,"reference_active":true,"delta_db":-2.3178500910953157}],"active_band_count":3,"active_band_mae_db":0.8183261086456177}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"facodec","status":"measured","mos":{"status":"measured","value":3.999415159225464,"predicted_mos":3.999415159225464,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f19aeb02b23c69dbfea5cec7864214ce95ce0122d2f01f9bbaf3ad5f467ae7a6","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2041468669995083},"started_at":"2026-10-13T04:55:05.901961+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6570000045467168e-05,"native_input_samples":96000,"native_input_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","output_sha256":"0b9af27903f6da4a044b11c6ff75faaa972647895804beb31ad6d8e56b802c2f","output_samples":96000,"length_delta_samples":0,"analysis_s":1.462546165999811,"synthesis_s":6.231077018000178,"total_s":7.693623183999989,"rtf":1.2822705306666649,"output_peak":0.17767316102981567,"output_rms":0.02175646917819631,"finished_at":"2026-10-13T04:55:13.603960+00:00","evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"0b9af27903f6da4a044b11c6ff75faaa972647895804beb31ad6d8e56b802c2f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.768907819528586,"spectral_convergence":0.2425499265174925,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.793221396552462,"spectral_convergence":0.24640879540019103,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.86167050976549,"spectral_convergence":0.26494163136107896,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.745113404213598e-05,"delta_db":0.16162583570570088,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.000291678857248019,"delta_db":-0.23338659805033024,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":5.078177915127235e-05,"delta_db":-0.21702671625315859,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":1.704066108544546e-05,"delta_db":-1.0127221245400357,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":2.4747912010394297e-05,"delta_db":-0.749396197063943,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":1.1643607564538719e-05,"delta_db":1.6203957611433741,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6657588721260904,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.191916836336135,"pitch":null,"output_rms_dbfs":-33.24823168393213,"output_peak":0.17767316102981567,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.807933241948845},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.768907819528586,"spectral_convergence":0.2425499265174925,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.793221396552462,"spectral_convergence":0.24640879540019103,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.86167050976549,"spectral_convergence":0.26494163136107896,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.745113404213598e-05,"delta_db":0.16162583570570088,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.000291678857248019,"delta_db":-0.23338659805033024,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":5.078177915127235e-05,"delta_db":-0.21702671625315859,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":1.704066108544546e-05,"delta_db":-1.0127221245400357,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":2.4747912010394297e-05,"delta_db":-0.749396197063943,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":1.1643607564538719e-05,"delta_db":1.6203957611433741,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6657588721260904,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.191916836336135,"pitch":null,"output_rms_dbfs":-33.24823168393213,"output_peak":0.17767316102981567,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.807933241948845},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.909227753751584,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.022242533719115894,"rms_dbfs":-33.056314847596,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.991136769433584,"mean":-1.7630259195963542e-06,"zero_energy":false},"output":{"rms":0.021756469178196305,"rms_dbfs":-33.24823168393213,"peak":0.17767316102981567,"peak_dbfs":-15.007563418792627,"crest_factor_db":18.24066826513951,"mean":1.1614065537481012e-05,"zero_energy":false},"rms_difference_db":-0.19191683633613366},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":1.3690768087358322e-07,"output_rms_dbfs":-68.63572186143298,"output_zero_energy":false,"output_peak":0.019015034660696983,"output_energy_fraction":7.905759159910385e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":182,"both_estimated_voiced_frames":137,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":45,"paired_voiced_fraction":0.9013157894736842,"absolute_difference_cents_median":3.939930224360008,"absolute_difference_cents_p95":28.15794219887226,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":1.276744406972973,"rms_envelope_nrmse":0.13189910814692177,"envelope_correlation":0.984872595211393,"rms_gain_error_db":-0.191988469517166,"reference_mean_power":0.0004951134756771775,"output_mean_power":0.00047370274337764964},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.645854738282274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010486873457313227,"output_power":9.498454333467971e-05,"reference_active":true,"delta_db":-0.4299308843250632},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5071002541192514e-05,"output_power":2.340991822106904e-05,"reference_active":true,"delta_db":-0.297718043482243},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059292105678353e-06,"output_power":5.120548923096808e-06,"reference_active":true,"delta_db":0.05226765097521388}],"active_band_count":3,"active_band_mae_db":0.2599721929275067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":1.15100681225703,"rms_envelope_nrmse":0.12313355624593515,"envelope_correlation":0.9862696532273576,"rms_gain_error_db":-0.19377565995741117,"reference_mean_power":0.0004970274687625529,"output_mean_power":0.00047533831808642895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4721291240539383e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.182773781582673e-05,"output_power":8.357550804656333e-05,"reference_active":true,"delta_db":-0.40894860264803323},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2533326164459844e-05,"output_power":1.1181036553833311e-05,"reference_active":true,"delta_db":-0.4958427449620332},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.472370238880856e-07,"output_power":2.719084589532516e-07,"reference_active":true,"delta_db":0.41309211371041826}],"active_band_count":3,"active_band_mae_db":0.4392944871068282}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":1.4587648273662004,"rms_envelope_nrmse":0.11935842876504975,"envelope_correlation":0.9902679817644462,"rms_gain_error_db":-0.2596996618540067,"reference_mean_power":0.00026104952136561866,"output_mean_power":0.0002458968328607839},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.392840173310503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.633554932680893e-05,"output_power":8.410537255355178e-05,"reference_active":true,"delta_db":-0.589628392570961},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.160826072139944e-05,"output_power":1.942489907326531e-05,"reference_active":true,"delta_db":-0.4626104033296799},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769645916048805e-06,"output_power":4.66028520850455e-06,"reference_active":true,"delta_db":-0.10073643303252354}],"active_band_count":3,"active_band_mae_db":0.38432507631105484}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":1.131587108869184,"rms_envelope_nrmse":0.10772806699650472,"envelope_correlation":0.9918767628882028,"rms_gain_error_db":-0.261677173782403,"reference_mean_power":0.00026200159904973233,"output_mean_power":0.00024668129776459776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2678124187221727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604417147270571e-05,"output_power":7.587749675166008e-05,"reference_active":true,"delta_db":-0.5460846228768053},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1040946271380912e-05,"output_power":9.801261332596527e-06,"reference_active":true,"delta_db":-0.5172432746278671},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9369342076961927e-07,"output_power":2.9605932858889803e-07,"reference_active":true,"delta_db":0.034845322005494034}],"active_band_count":3,"active_band_mae_db":0.3660577398367222}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":1.6252044274987754,"rms_envelope_nrmse":0.18291775794318507,"envelope_correlation":0.9807505302858223,"rms_gain_error_db":-0.9042286260432041,"reference_mean_power":4.057559706591337e-05,"output_mean_power":3.294898620878793e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.7364679678483124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.645857845256653e-05,"output_power":1.269439340964242e-05,"reference_active":true,"delta_db":-1.1278036862562737},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1440018553357532e-06,"output_power":2.659206760881952e-06,"reference_active":true,"delta_db":-0.7273068737850223},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.079905260013692e-07,"output_power":8.217303337050797e-07,"reference_active":true,"delta_db":-0.4335199810208197}],"active_band_count":3,"active_band_mae_db":0.7628768470207051}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":1.3125882688190205,"rms_envelope_nrmse":0.15779724703270553,"envelope_correlation":0.987072302340607,"rms_gain_error_db":-0.9055929929819848,"reference_mean_power":4.071142209359619e-05,"output_mean_power":3.304889731515244e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5164519386807125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5218014411010497e-05,"output_power":1.2003077116773565e-05,"reference_active":true,"delta_db":-1.0306539468754854},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2444589918408902e-06,"output_power":1.0582609164120913e-06,"reference_active":true,"delta_db":-0.7038783294643792},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241409422416778e-08,"output_power":4.885930165895212e-08,"reference_active":true,"delta_db":-0.3050082973451377}],"active_band_count":3,"active_band_mae_db":0.6798468578950008}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":2.091858754366723,"rms_envelope_nrmse":0.2704016978547033,"envelope_correlation":0.9440071813757922,"rms_gain_error_db":0.06625403188017881,"reference_mean_power":9.786447369332921e-06,"output_mean_power":9.936889674354652e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.593718108385403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.378843387910107e-06,"output_power":2.442043498838689e-06,"reference_active":true,"delta_db":0.11387544506939519},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.926849454092712e-07,"output_power":8.051485196508368e-07,"reference_active":true,"delta_db":-0.4482221195020645},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.698308970455951e-07,"output_power":2.131032896080616e-07,"reference_active":true,"delta_db":0.9857345028177358}],"active_band_count":3,"active_band_mae_db":0.5159440224630651}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":1.8578073232557173,"rms_envelope_nrmse":0.22428325259409962,"envelope_correlation":0.9571621718874105,"rms_gain_error_db":0.06694253869549043,"reference_mean_power":9.831708602072054e-06,"output_mean_power":9.984429437648166e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.220673114788623e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.414723334287998e-06,"output_power":2.5151381329946463e-06,"reference_active":true,"delta_db":0.17694462911660405},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7015119113313933e-07,"output_power":2.955358233419442e-07,"reference_active":true,"delta_db":-0.9776901989900726},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1816168225696594e-08,"output_power":1.4763613717125876e-08,"reference_active":true,"delta_db":0.9671600820701873}],"active_band_count":3,"active_band_mae_db":0.7072649700589547}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.6076571941375732,"predicted_mos":3.6076571941375732,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"52d72b58afaa9abb4051c526cfa87f3f19cb0daa85298773c323f07496313b91","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8482343040013802},"started_at":"2026-10-13T07:20:50.673501+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.013095684998916113,"native_input_samples":264600,"native_input_sha256":"a209bf7fdf88f2689731bd99e7456baa367479322ce99ef4bb0795b7b80a4183","output_sha256":"8d33a92c0fae00b7327252aa50d8cf00452db0c4c1b5f22ab81fed1ec3a906ca","output_samples":264704,"length_delta_samples":104,"analysis_s":0.44229496699881565,"synthesis_s":4.57114290500067,"total_s":5.013437871999486,"rtf":0.835572978666581,"output_peak":0.20897096395492554,"output_rms":0.020883448779984118,"finished_at":"2026-10-13T07:20:55.717256+00:00","evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"8d33a92c0fae00b7327252aa50d8cf00452db0c4c1b5f22ab81fed1ec3a906ca","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-33.60322070791368,"output_peak":0.20738786044791963},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.813664815191444,"spectral_convergence":0.44402069270737327,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.614697143937144,"spectral_convergence":0.34911364872328815,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.311349547331506,"spectral_convergence":0.39895953482624336,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":6.668999096040678e-05,"delta_db":-0.488045389741973,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.0002831823089744321,"delta_db":-0.36177491093925584,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":3.712724172906538e-05,"delta_db":-1.5771789837982462,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":1.7760149911432215e-05,"delta_db":-0.8331202419845944,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":2.565914460578538e-05,"delta_db":-0.5923600866661842,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":5.877610329957968e-06,"delta_db":-1.3484719334973163,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8668252577712616,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.545868394227619,"pitch":null,"output_rms_dbfs":-33.60218324182362,"output_peak":0.20738786044791963,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.579903835486698},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7536522773322147,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.022242533719115894,"rms_dbfs":-33.056314847596,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.991136769433584,"mean":-1.7630259195963542e-06,"zero_energy":false},"output":{"rms":0.020887710418116198,"rms_dbfs":-33.60218324182362,"peak":0.20738786044791963,"peak_dbfs":-13.664333376951319,"crest_factor_db":19.9378498648723,"mean":-8.569839421741281e-06,"zero_energy":false},"rms_difference_db":-0.5458683942276181},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":1.0690680151064887e-07,"output_rms_dbfs":-69.70994663691336,"output_zero_energy":false,"output_peak":0.005781679369389334,"output_energy_fraction":6.697554528940819e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":170,"both_estimated_voiced_frames":131,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.8618421052631579,"absolute_difference_cents_median":15.557734771324775,"absolute_difference_cents_p95":43.84156874151585,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":1.4789270127791694,"rms_envelope_nrmse":0.11826569062097422,"envelope_correlation":0.9904076851161939,"rms_gain_error_db":-0.5457274723515306,"reference_mean_power":0.0004951134756771775,"output_mean_power":0.00043664852716521755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.645854738282274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010486873457313227,"output_power":0.00010178462719079865,"reference_active":true,"delta_db":-0.12963837228228028},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5071002541192514e-05,"output_power":2.161305422768152e-05,"reference_active":true,"delta_db":-0.6445555778565359},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059292105678353e-06,"output_power":4.9678619471745e-06,"reference_active":true,"delta_db":-0.07920236116453191}],"active_band_count":3,"active_band_mae_db":0.28446543710111605}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":1.1401188615512832,"rms_envelope_nrmse":0.08822394138851081,"envelope_correlation":0.9959905405142244,"rms_gain_error_db":-0.5442200042515684,"reference_mean_power":0.0004970274687625529,"output_mean_power":0.000438488684647872},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4721291240539383e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.182773781582673e-05,"output_power":8.935179711313973e-05,"reference_active":true,"delta_db":-0.11870593767710741},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2533326164459844e-05,"output_power":1.0361318754744906e-05,"reference_active":true,"delta_db":-0.8265130729053853},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.472370238880856e-07,"output_power":2.590855904728411e-07,"reference_active":true,"delta_db":0.20329752423812186}],"active_band_count":3,"active_band_mae_db":0.38283884494020487}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":2.3075393757400975,"rms_envelope_nrmse":0.16084074439620957,"envelope_correlation":0.9819520200565659,"rms_gain_error_db":-0.41487490109794645,"reference_mean_power":0.00026104952136561866,"output_mean_power":0.00023726594672414698},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.392840173310503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.633554932680893e-05,"output_power":9.426774995498862e-05,"reference_active":true,"delta_db":-0.0942343699123637},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.160826072139944e-05,"output_power":1.697771085146571e-05,"reference_active":true,"delta_db":-1.0474067851998268},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769645916048805e-06,"output_power":3.877586936162867e-06,"reference_active":true,"delta_db":-0.8992459598891185}],"active_band_count":3,"active_band_mae_db":0.6802957050004363}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":1.7417330053665747,"rms_envelope_nrmse":0.12811798995899504,"envelope_correlation":0.9884957667994051,"rms_gain_error_db":-0.4115783793274228,"reference_mean_power":0.00026200159904973233,"output_mean_power":0.0002383121056608542},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2678124187221727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604417147270571e-05,"output_power":8.347116920947052e-05,"reference_active":true,"delta_db":-0.13184960224572698},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1040946271380912e-05,"output_power":8.3807669357537e-06,"reference_active":true,"delta_db":-1.1972253312851848},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9369342076961927e-07,"output_power":2.241870050834619e-07,"reference_active":true,"delta_db":-1.1728378242925577}],"active_band_count":3,"active_band_mae_db":0.8339709192744899}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":2.946645640234763,"rms_envelope_nrmse":0.23753304036717118,"envelope_correlation":0.9609750828674232,"rms_gain_error_db":-0.702714873213342,"reference_mean_power":4.057559706591337e-05,"output_mean_power":3.4513851949628825e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.7364679678483124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.645857845256653e-05,"output_power":1.359534386216467e-05,"reference_active":true,"delta_db":-0.8300212543632708},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1440018553357532e-06,"output_power":3.0559500943907203e-06,"reference_active":true,"delta_db":-0.12336535996574508},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.079905260013692e-07,"output_power":6.955995338186547e-07,"reference_active":true,"delta_db":-1.1572203483139472}],"active_band_count":3,"active_band_mae_db":0.703535654214321}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":2.5299929671050765,"rms_envelope_nrmse":0.19376657359851068,"envelope_correlation":0.9739953847494606,"rms_gain_error_db":-0.7028286259761983,"reference_mean_power":4.071142209359619e-05,"output_mean_power":3.462847853223497e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5164519386807125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5218014411010497e-05,"output_power":1.28943547928441e-05,"reference_active":true,"delta_db":-0.719583752192664},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2444589918408902e-06,"output_power":1.0903258634476652e-06,"reference_active":true,"delta_db":-0.5742427600929819},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241409422416778e-08,"output_power":5.059414146226105e-08,"reference_active":true,"delta_db":-0.15347854374299544}],"active_band_count":3,"active_band_mae_db":0.48243501867621386}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":4.74948328851087,"rms_envelope_nrmse":0.2959229363738851,"envelope_correlation":0.9375459576517824,"rms_gain_error_db":-0.7256462402781427,"reference_mean_power":9.786447369332921e-06,"output_mean_power":8.280573965473184e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.593718108385403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.378843387910107e-06,"output_power":2.6647656282707373e-06,"reference_active":true,"delta_db":0.49293166900751034},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.926849454092712e-07,"output_power":1.1018454180503334e-06,"reference_active":true,"delta_db":0.91422459295474},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.698308970455951e-07,"output_power":1.2212966321826323e-07,"reference_active":true,"delta_db":-1.4319554404969337}],"active_band_count":3,"active_band_mae_db":0.9463705674863947}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":4.342170228370856,"rms_envelope_nrmse":0.24116815852474857,"envelope_correlation":0.9576233965821416,"rms_gain_error_db":-0.724191266586153,"reference_mean_power":9.831708602072054e-06,"output_mean_power":8.32165815645228e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.220673114788623e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.414723334287998e-06,"output_power":2.714577881613575e-06,"reference_active":true,"delta_db":0.5083492727303095},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7015119113313933e-07,"output_power":4.898802095383321e-07,"reference_active":true,"delta_db":1.2171074366875512},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1816168225696594e-08,"output_power":1.0328749807220076e-08,"reference_active":true,"delta_db":-0.5842890769971312}],"active_band_count":3,"active_band_mae_db":0.7699152621383307}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"golf","status":"measured","mos":{"status":"measured","value":3.5740253925323486,"predicted_mos":3.5740253925323486,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"a73dd88a3daecfd5780d331a2b7d00ace989e5ca439bfca63e3673e1ff0f5850","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2592404069982877},"started_at":"2026-10-13T05:02:44.680101+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003209022000191908,"native_input_samples":144000,"native_input_sha256":"8e02dec65a4c9f4670dd2e090b12026fa3d5c9016e85324b58ef3b5936de6d3c","output_sha256":"165c44423968abef73c02670ae3cf6fc0d4ccfa8f75b02582097813ae282337b","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5491976140001498,"synthesis_s":0.13002889800009143,"total_s":0.6792265120002412,"rtf":0.11320441866670687,"output_peak":0.1844191551208496,"output_rms":0.019481857689935423,"finished_at":"2026-10-13T05:02:45.370801+00:00","analysis_stages":{"external_f0_s":0.08375616400007857},"evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"165c44423968abef73c02670ae3cf6fc0d4ccfa8f75b02582097813ae282337b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-34.20410102214137,"output_peak":0.18552267226128943},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.254537845465373,"spectral_convergence":0.503314644312226,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.26466073897717,"spectral_convergence":0.2995349764612599,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.400263371248639,"spectral_convergence":0.30967519592899395,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.470636967202408e-05,"output_power":5.959931667571857e-05,"delta_db":-0.9811635209413057,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00030812201313514735,"output_power":0.0002383001383907679,"delta_db":-1.1159843213024963,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.3455498385521636e-05,"output_power":4.512239521095714e-05,"delta_db":-0.7360023795297527,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1538713741263306e-05,"output_power":1.6164184909733026e-05,"delta_db":-1.246659544418894,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.9446988475243514e-05,"output_power":1.7520943502530537e-05,"delta_db":-2.2548339716129755,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.030718464642014e-06,"output_power":3.123576044622203e-06,"delta_db":-4.101023175070745,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.7392778188126947,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1527869488890916,"pitch":null,"output_rms_dbfs":-34.20410102214137,"output_peak":0.18552267226128943,"reference_rms_dbfs":-33.051314073252286,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.639820651897061},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.839405232387093,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.02225534322076031,"rms_dbfs":-33.051314073252286,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.98613599508987,"mean":-4.0885403080655257e-07,"zero_energy":false},"output":{"rms":0.01948924202564916,"rms_dbfs":-34.20410102214137,"peak":0.18552267226128943,"peak_dbfs":-14.6320601752869,"crest_factor_db":19.572040846854478,"mean":0.00022429915035291018,"zero_energy":false},"rms_difference_db":-1.1527869488890872},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":2.0495185660465466e-07,"output_rms_dbfs":-66.88348143170829,"output_zero_energy":false,"output_peak":0.0031627368768722946,"output_energy_fraction":0.00014773349923397338,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":186,"both_estimated_voiced_frames":144,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":42,"paired_voiced_fraction":0.9473684210526315,"absolute_difference_cents_median":5.830527450052171,"absolute_difference_cents_p95":19.992727353067338,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":3.363404812210574,"rms_envelope_nrmse":0.18547462055188957,"envelope_correlation":0.981771899319333,"rms_gain_error_db":-1.1530146537825183,"reference_mean_power":0.000495656975690873,"output_mean_power":0.0003800841481948097},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.646323101943136e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010481926430920747,"output_power":9.6675504942567e-05,"reference_active":true,"delta_db":-0.35124658112023854},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.503010734386605e-05,"output_power":1.6009225844343974e-05,"reference_active":true,"delta_db":-1.9409238080713263},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.057163790642674e-06,"output_power":4.130173545387307e-06,"reference_active":true,"delta_db":-0.8793871910296465}],"active_band_count":3,"active_band_mae_db":1.0571858600737372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":3.3070048411566773,"rms_envelope_nrmse":0.15897207367324806,"envelope_correlation":0.9903625308006219,"rms_gain_error_db":-1.1529045694165416,"reference_mean_power":0.0004974331108447674,"output_mean_power":0.0003814558091239744},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.4729219548938384e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.178632449351614e-05,"output_power":8.589628180933203e-05,"reference_active":true,"delta_db":-0.28803614305282893},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2500068690826628e-05,"output_power":7.726274071134438e-06,"reference_active":true,"delta_db":-2.089422899228794},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4514087928255473e-07,"output_power":2.0002357222929455e-07,"reference_active":true,"delta_db":-0.8833456041853877}],"active_band_count":3,"active_band_mae_db":1.0869348821556704}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":2.516713420420949,"rms_envelope_nrmse":0.2258686580152222,"envelope_correlation":0.975921028308908,"rms_gain_error_db":-1.3794456441335268,"reference_mean_power":0.00026131183019191843,"output_mean_power":0.0001902017494826712},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.3931593398031703e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.628134433124284e-05,"output_power":7.970592043198564e-05,"reference_active":true,"delta_db":-0.8205156418155033},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.157615987702236e-05,"output_power":1.4143518161580427e-05,"reference_active":true,"delta_db":-1.8341669906471358},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7698338843080635e-06,"output_power":4.0182608140402e-06,"reference_active":true,"delta_db":-0.7446513226724749}],"active_band_count":3,"active_band_mae_db":1.1331113183783712}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":2.1923913166199682,"rms_envelope_nrmse":0.20026428294865461,"envelope_correlation":0.9842434871102317,"rms_gain_error_db":-1.3766948378204946,"reference_mean_power":0.0002621853406044688,"output_mean_power":0.00019095846801174047},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.268285448938086e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.599322690839022e-05,"output_power":7.198856566345683e-05,"reference_active":true,"delta_db":-0.7720072567093995},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1005502827753051e-05,"output_power":6.865289296707356e-06,"reference_active":true,"delta_db":-2.0495104700669566},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9252970760948466e-07,"output_power":2.1590970337018687e-07,"reference_active":true,"delta_db":-1.3189781632273383}],"active_band_count":3,"active_band_mae_db":1.3801652966678981}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":3.3640012105806454,"rms_envelope_nrmse":0.2593253746834774,"envelope_correlation":0.9729391554174285,"rms_gain_error_db":-1.749956528824673,"reference_mean_power":4.061556762817391e-05,"output_mean_power":2.7145439296733998e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.736691969478702e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6453684585240387e-05,"output_power":1.2327951906029348e-05,"reference_active":true,"delta_db":-1.2537223624917948},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1368406859882844e-06,"output_power":1.976834265828258e-06,"reference_active":true,"delta_db":-2.005222019050612},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.061884518548568e-07,"output_power":5.061863128606268e-07,"reference_active":true,"delta_db":-2.529081256060884}],"active_band_count":3,"active_band_mae_db":1.929341879201097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":3.2617340424632486,"rms_envelope_nrmse":0.22467629882727447,"envelope_correlation":0.9868946452376499,"rms_gain_error_db":-1.7459883818839894,"reference_mean_power":4.07290699472098e-05,"output_mean_power":2.724618209653955e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.5170070592088446e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5208678073906827e-05,"output_power":1.1583522163707073e-05,"reference_active":true,"delta_db":-1.1825083319041423},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.239245778658589e-06,"output_power":7.504961176069877e-07,"reference_active":true,"delta_db":-2.1780899827246865},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.221112060279402e-08,"output_power":3.17738649921789e-08,"reference_active":true,"delta_db":-2.1569296865169822}],"active_band_count":3,"active_band_mae_db":1.839176000381937}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":3.2519165937988874,"rms_envelope_nrmse":0.30492674762662736,"envelope_correlation":0.9588279973255951,"rms_gain_error_db":-2.1903166210020957,"reference_mean_power":9.802275751755874e-06,"output_mean_power":5.919639419892827e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.58883797637495e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3790602188292784e-06,"output_power":1.515704722379271e-06,"reference_active":true,"delta_db":-1.9579083133290807},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.846885554073932e-07,"output_power":5.249329428514249e-07,"reference_active":true,"delta_db":-2.266865808060532},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6902164873918997e-07,"output_power":1.0657602925404433e-07,"reference_active":true,"delta_db":-2.002827982383611}],"active_band_count":3,"active_band_mae_db":2.075867367924408}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":3.1801870954187823,"rms_envelope_nrmse":0.269528560632357,"envelope_correlation":0.9764553348635109,"rms_gain_error_db":-2.1893572051863224,"reference_mean_power":9.843620895577495e-06,"output_mean_power":5.945921332585636e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.216543586123948e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4129699675829806e-06,"output_power":1.519605767835652e-06,"reference_active":true,"delta_db":-2.008209833069448},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.650413705206513e-07,"output_power":2.1316860990012384e-07,"reference_active":true,"delta_db":-2.3361883317602947},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.169829527778327e-08,"output_power":7.817767627825836e-09,"reference_active":true,"delta_db":-1.7503982174003285}],"active_band_count":3,"active_band_mae_db":2.0315987940766904}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.7824153900146484,"predicted_mos":3.7824153900146484,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"eca3a2746464ab49d97a0aab567178a4b17ce3285fb00b22b63aac1e331ee383","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8104059270008293},"started_at":"2026-10-13T07:40:54.654369+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003137160001642769,"native_input_samples":144000,"native_input_sha256":"8e02dec65a4c9f4670dd2e090b12026fa3d5c9016e85324b58ef3b5936de6d3c","output_sha256":"fdc133ac67f0d61b020c4cffb6244d1c25a403ade2256012b9b4cc0f447a92a4","output_samples":144120,"length_delta_samples":120,"analysis_s":1.9778975859990169,"synthesis_s":3.5660320130009495,"total_s":5.543929598999966,"rtf":0.9239882664999944,"output_peak":0.20323188602924347,"output_rms":0.021930648101227464,"finished_at":"2026-10-13T07:41:00.209739+00:00","analysis_stages":{"external_f0_s":1.2073388780008827},"evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"fdc133ac67f0d61b020c4cffb6244d1c25a403ade2256012b9b4cc0f447a92a4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-33.17907858034845,"output_peak":0.2033721286216429},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.788460107855755,"spectral_convergence":0.48703126091603033,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.902660690397308,"spectral_convergence":0.2678603316413569,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.907115117107143,"spectral_convergence":0.2741565417892286,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":8.797435086684691e-05,"delta_db":0.7149087348033613,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.0002883115975614397,"delta_db":-0.28381497043949944,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":4.8988989624048974e-05,"delta_db":-0.37312101616598803,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":2.0436383828062438e-05,"delta_db":-0.2235460065381472,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":3.0193196396058688e-05,"delta_db":0.11430908995119138,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":5.418746909401538e-06,"delta_db":-1.7014911634434318,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5685318302236032,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11931661047680428,"pitch":null,"output_rms_dbfs":-33.1756314580728,"output_peak":0.2033721286216429,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.199411971786735},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9081960256790984,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.022242533719115894,"rms_dbfs":-33.056314847596,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.991136769433584,"mean":-1.7630259195963542e-06,"zero_energy":false},"output":{"rms":0.021939080773493182,"rms_dbfs":-33.17563145807281,"peak":0.2033721286216429,"peak_dbfs":-13.83417131491403,"crest_factor_db":19.341460143158777,"mean":-2.5403390953360513e-05,"zero_energy":false},"rms_difference_db":-0.1193166104768082},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":4.805089077742845e-07,"output_rms_dbfs":-63.18298556879184,"output_zero_energy":false,"output_peak":0.02157451922798146,"output_energy_fraction":0.00027287087693047445,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":160,"both_estimated_voiced_frames":133,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":7.495704807584949,"absolute_difference_cents_p95":24.43189893804887,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":1.9252859508135531,"rms_envelope_nrmse":0.14673873889580788,"envelope_correlation":0.9805802336208771,"rms_gain_error_db":-0.11938351704113793,"reference_mean_power":0.0004951134756771775,"output_mean_power":0.00048168862998316885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.645854738282274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010486873457313227,"output_power":0.00010082146891892495,"reference_active":true,"delta_db":-0.17093006900490687},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5071002541192514e-05,"output_power":2.5241550173179537e-05,"reference_active":true,"delta_db":0.029443220497397456},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059292105678353e-06,"output_power":3.646934884245621e-06,"reference_active":true,"delta_db":-1.4216174567682138}],"active_band_count":3,"active_band_mae_db":0.5406635820901727}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":1.3701531688871302,"rms_envelope_nrmse":0.09180550135156791,"envelope_correlation":0.9920075882581428,"rms_gain_error_db":-0.11857181555810645,"reference_mean_power":0.0004970274687625529,"output_mean_power":0.00048364111039417724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4721291240539383e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.182773781582673e-05,"output_power":8.880726963838732e-05,"reference_active":true,"delta_db":-0.14525367596891609},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2533326164459844e-05,"output_power":1.2552963266987431e-05,"reference_active":true,"delta_db":0.006799161775998785},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.472370238880856e-07,"output_power":2.0634612950960811e-07,"reference_active":true,"delta_db":-0.7851718002774383}],"active_band_count":3,"active_band_mae_db":0.3124082126741177}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":2.0654968763377495,"rms_envelope_nrmse":0.1798051481967605,"envelope_correlation":0.9765035701775633,"rms_gain_error_db":-0.2627934170553375,"reference_mean_power":0.00026104952136561866,"output_mean_power":0.0002457217273187284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.392840173310503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.633554932680893e-05,"output_power":8.827312021075228e-05,"reference_active":true,"delta_db":-0.3795810020049827},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.160826072139944e-05,"output_power":2.0482941861220585e-05,"reference_active":true,"delta_db":-0.23227479029484274},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769645916048805e-06,"output_power":4.16263213447106e-06,"reference_active":true,"delta_db":-0.5911810699332111}],"active_band_count":3,"active_band_mae_db":0.4010122874110122}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":1.461550508245507,"rms_envelope_nrmse":0.1146136386266858,"envelope_correlation":0.9902949414966878,"rms_gain_error_db":-0.26073771727753947,"reference_mean_power":0.00026200159904973233,"output_mean_power":0.00024673466510558755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2678124187221727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604417147270571e-05,"output_power":7.8949452292994e-05,"reference_active":true,"delta_db":-0.37372335617784963},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1040946271380912e-05,"output_power":1.0541375193594307e-05,"reference_active":true,"delta_db":-0.2010902525438949},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9369342076961927e-07,"output_power":2.7138484368363757e-07,"reference_active":true,"delta_db":-0.3430862820156143}],"active_band_count":3,"active_band_mae_db":0.3059666302457863}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":2.090128413625835,"rms_envelope_nrmse":0.20014304088362478,"envelope_correlation":0.970504175173385,"rms_gain_error_db":-0.018709178726997778,"reference_mean_power":4.057559706591337e-05,"output_mean_power":4.0401175489012694e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.7364679678483124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.645857845256653e-05,"output_power":1.5872389242232633e-05,"reference_active":true,"delta_db":-0.1575001670996659},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1440018553357532e-06,"output_power":2.851746214030772e-06,"reference_active":true,"delta_db":-0.42371920001233654},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.079905260013692e-07,"output_power":9.214381213835805e-07,"reference_active":true,"delta_db":0.06384858648708933}],"active_band_count":3,"active_band_mae_db":0.2150226511996973}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":1.2587239531215733,"rms_envelope_nrmse":0.10861142901750881,"envelope_correlation":0.990711237235194,"rms_gain_error_db":-0.01699048678836723,"reference_mean_power":4.071142209359619e-05,"output_mean_power":4.05524618439356e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5164519386807125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5218014411010497e-05,"output_power":1.4704023217047649e-05,"reference_active":true,"delta_db":-0.14921811253088846},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2444589918408902e-06,"output_power":1.1217032421060215e-06,"reference_active":true,"delta_db":-0.4510261491286049},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241409422416778e-08,"output_power":4.825451275488041e-08,"reference_active":true,"delta_db":-0.3591015035460061}],"active_band_count":3,"active_band_mae_db":0.3197819217351665}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":2.2755529414424926,"rms_envelope_nrmse":0.25522123230486476,"envelope_correlation":0.94778997577583,"rms_gain_error_db":-0.42050574961029374,"reference_mean_power":9.786447369332921e-06,"output_mean_power":8.883303290489123e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.593718108385403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.378843387910107e-06,"output_power":2.221239224067002e-06,"reference_active":true,"delta_db":-0.2977051711163323},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.926849454092712e-07,"output_power":7.915351792328972e-07,"reference_active":true,"delta_db":-0.5222798917750573},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.698308970455951e-07,"output_power":1.1030470135549099e-07,"reference_active":true,"delta_db":-1.8742268041934174}],"active_band_count":3,"active_band_mae_db":0.8980706223616023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":1.431820423437576,"rms_envelope_nrmse":0.14655559392304018,"envelope_correlation":0.9823091634060437,"rms_gain_error_db":-0.4199016598508497,"reference_mean_power":9.831708602072054e-06,"output_mean_power":8.925629022276739e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.220673114788623e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.414723334287998e-06,"output_power":2.222426981991006e-06,"reference_active":true,"delta_db":-0.3603987776782912},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7015119113313933e-07,"output_power":3.3187702910185693e-07,"reference_active":true,"delta_db":-0.474019575468131},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1816168225696594e-08,"output_power":8.732488475095518e-09,"reference_active":true,"delta_db":-1.3133864421853436}],"active_band_count":3,"active_band_mae_db":0.7159349317772552}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.6331264972686768,"predicted_mos":3.6331264972686768,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"8c5eb2a935c83475bebb0d264b7bc11e0d99986d886f8bf4e35829c9e8c50401","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.420620429998962},"started_at":"2026-10-13T05:11:01.504037+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031170570000540465,"native_input_samples":144000,"native_input_sha256":"8e02dec65a4c9f4670dd2e090b12026fa3d5c9016e85324b58ef3b5936de6d3c","output_sha256":"e27cb75988b6b2944db415bd01c14313583bf23402f98dc7c934b60c477aa23f","output_samples":144240,"length_delta_samples":240,"analysis_s":1.4999681429999328,"synthesis_s":0.20405889299945557,"total_s":1.7040270359993883,"rtf":0.28400450599989807,"output_peak":0.2476450651884079,"output_rms":0.021807114658165608,"finished_at":"2026-10-13T05:11:03.220527+00:00","analysis_stages":{"external_f0_s":0.21334997300073155},"evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"e27cb75988b6b2944db415bd01c14313583bf23402f98dc7c934b60c477aa23f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-33.22713409957698,"output_peak":0.24773927288058956},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.355024464449393,"spectral_convergence":0.15530852088954747,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.983978303376292,"spectral_convergence":0.15882011609579075,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.236022079727428,"spectral_convergence":0.17509610212714466,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.382224556590223e-05,"delta_db":-0.04677945855410423,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.00029538983582371397,"delta_db":-0.17848062185394398,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":5.5399122808250144e-05,"delta_db":0.16092304111772682,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":1.9832446453107406e-05,"delta_db":-0.35382361107188237,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":2.6021660224008792e-05,"delta_db":-0.5314318115117784,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":5.967853264268666e-06,"delta_db":-1.282298471661345,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.42562283596179684,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16366530478838529,"pitch":null,"output_rms_dbfs":-33.21998015238439,"output_peak":0.24773927288058956,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.191674949184371},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9580024951659318,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.022242533719115894,"rms_dbfs":-33.056314847596,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.991136769433584,"mean":-1.7630259195963542e-06,"zero_energy":false},"output":{"rms":0.0218273489947646,"rms_dbfs":-33.21998015238439,"peak":0.24773927288058956,"peak_dbfs":-12.120102828044601,"crest_factor_db":21.099877324339786,"mean":-3.1627749250821995e-06,"zero_energy":false},"rms_difference_db":-0.16366530478838826},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":1.8359931215180791e-07,"output_rms_dbfs":-67.3612895020032,"output_zero_energy":false,"output_peak":0.010321233071436208,"output_energy_fraction":0.00010533232311572074,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":160,"both_estimated_voiced_frames":145,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9539473684210527,"absolute_difference_cents_median":1.5354464259592984,"absolute_difference_cents_p95":8.18363304836232,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":0.6908122195247881,"rms_envelope_nrmse":0.06425749145188726,"envelope_correlation":0.9965752704595943,"rms_gain_error_db":-0.16367197375882828,"reference_mean_power":0.0004951134756771775,"output_mean_power":0.00047680143023320817},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.645854738282274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010486873457313227,"output_power":0.00010262666321004643,"reference_active":true,"delta_db":-0.09385818964380041},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5071002541192514e-05,"output_power":2.2978648671622854e-05,"reference_active":true,"delta_db":-0.37847215793644223},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059292105678353e-06,"output_power":4.1971683192542464e-06,"reference_active":true,"delta_db":-0.8113336866937647}],"active_band_count":3,"active_band_mae_db":0.4278880114246691}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":0.5052216994678602,"rms_envelope_nrmse":0.03787210598623969,"envelope_correlation":0.9990026188515947,"rms_gain_error_db":-0.1632478998703467,"reference_mean_power":0.0004970274687625529,"output_mean_power":0.00047869137354011876},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4721291240539383e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.182773781582673e-05,"output_power":9.004986728848492e-05,"reference_active":true,"delta_db":-0.08490808455072779},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2533326164459844e-05,"output_power":1.1835388370837171e-05,"reference_active":true,"delta_db":-0.24883828198549035},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.472370238880856e-07,"output_power":2.580828130397153e-07,"reference_active":true,"delta_db":0.18645576636579714}],"active_band_count":3,"active_band_mae_db":0.17340071096733842}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":0.8916008808529663,"rms_envelope_nrmse":0.08210251702858085,"envelope_correlation":0.9952101207143603,"rms_gain_error_db":-0.13313920615724373,"reference_mean_power":0.00026104952136561866,"output_mean_power":0.0002531680990500298},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.392840173310503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.633554932680893e-05,"output_power":9.309050135101624e-05,"reference_active":true,"delta_db":-0.14881208908793803},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.160826072139944e-05,"output_power":1.964997621228411e-05,"reference_active":true,"delta_db":-0.4125778236344011},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769645916048805e-06,"output_power":3.7729519103967167e-06,"reference_active":true,"delta_db":-1.0180486977067744}],"active_band_count":3,"active_band_mae_db":0.5264795368097045}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":0.5175194200784416,"rms_envelope_nrmse":0.04299601228534983,"envelope_correlation":0.9987399772665511,"rms_gain_error_db":-0.13243737635064814,"reference_mean_power":0.00026200159904973233,"output_mean_power":0.0002541324973528968},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2678124187221727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604417147270571e-05,"output_power":8.268217265150071e-05,"reference_active":true,"delta_db":-0.17309576956505882},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1040946271380912e-05,"output_power":1.0242821015581673e-05,"reference_active":true,"delta_db":-0.325867126116606},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9369342076961927e-07,"output_power":2.6325326319113294e-07,"reference_active":true,"delta_db":-0.47520454452938277}],"active_band_count":3,"active_band_mae_db":0.3247224800703492}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":1.248196426365385,"rms_envelope_nrmse":0.11936410684712975,"envelope_correlation":0.9907724581702975,"rms_gain_error_db":-0.41576885615230436,"reference_mean_power":4.057559706591337e-05,"output_mean_power":3.687126463074358e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.7364679678483124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.645857845256653e-05,"output_power":1.4408182198769421e-05,"reference_active":true,"delta_db":-0.5778313023346949},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1440018553357532e-06,"output_power":3.230746773150926e-06,"reference_active":true,"delta_db":0.11820125573909247},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.079905260013692e-07,"output_power":7.327961068516994e-07,"reference_active":true,"delta_db":-0.9309816372757357}],"active_band_count":3,"active_band_mae_db":0.5423380651165077}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":0.8882877365357348,"rms_envelope_nrmse":0.07818252727518941,"envelope_correlation":0.9968426283667114,"rms_gain_error_db":-0.4159476308408977,"reference_mean_power":4.071142209359619e-05,"output_mean_power":3.699316673581534e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5164519386807125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5218014411010497e-05,"output_power":1.3542059874072085e-05,"reference_active":true,"delta_db":-0.5067326139889249},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2444589918408902e-06,"output_power":1.350597579806665e-06,"reference_active":true,"delta_db":0.35545377105929693},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241409422416778e-08,"output_power":3.722960760992022e-08,"reference_active":true,"delta_db":-1.4855962619412102}],"active_band_count":3,"active_band_mae_db":0.7825942156631441}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":1.2212769693115646,"rms_envelope_nrmse":0.15810081535454934,"envelope_correlation":0.984055964926206,"rms_gain_error_db":-0.7216529266406875,"reference_mean_power":9.786447369332921e-06,"output_mean_power":8.288191408800264e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.593718108385403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.378843387910107e-06,"output_power":2.156480289824179e-06,"reference_active":true,"delta_db":-0.426203579709823},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.926849454092712e-07,"output_power":7.043281674398409e-07,"reference_active":true,"delta_db":-1.029231537099696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.698308970455951e-07,"output_power":1.633731228567061e-07,"reference_active":true,"delta_db":-0.16836092940102856}],"active_band_count":3,"active_band_mae_db":0.5412653487368492}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":0.932288045804106,"rms_envelope_nrmse":0.11091847994997206,"envelope_correlation":0.9943644888681171,"rms_gain_error_db":-0.7221655509271805,"reference_mean_power":9.831708602072054e-06,"output_mean_power":8.325540601658945e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.220673114788623e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.414723334287998e-06,"output_power":2.1833525806899033e-06,"reference_active":true,"delta_db":-0.43743505051587817},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7015119113313933e-07,"output_power":3.1857555696083877e-07,"reference_active":true,"delta_db":-0.6516670021441329},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1816168225696594e-08,"output_power":1.1776824847035019e-08,"reference_active":true,"delta_db":-0.01448449379688926}],"active_band_count":3,"active_band_mae_db":0.3678621821523001}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5639-40744-0030","method":"magphase","status":"measured","mos":{"status":"measured","value":3.5262980461120605,"predicted_mos":3.5262980461120605,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288078,"input_sample_rate":48000,"input_seconds":6.001625,"input_float32_sha256":"b71f003d5f80c5db065afd2ca16f2317366c17b6c85ed53f5c3241d8590d8d05","internal_samples":96026,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.305247882999538},"started_at":"2026-10-13T05:16:40.957669+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008808257000055164,"native_input_samples":288000,"native_input_sha256":"25210cf7b2dbe4e933104b3b45cf78b747c874d789deb627b1287bd4df258cce","output_sha256":"fbacd2eb35b9317a95946a5758cf2957b45b7cf46d66b2a28683c452d70373fa","output_samples":288078,"length_delta_samples":78,"analysis_s":3.410232049000115,"synthesis_s":1.0555785880005715,"total_s":4.465810637000686,"rtf":0.7443017728334477,"output_peak":0.22915362298488803,"output_rms":0.021327276086250735,"finished_at":"2026-10-13T05:16:45.444721+00:00","evaluation_input":{"reference":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"output_sha256":"fbacd2eb35b9317a95946a5758cf2957b45b7cf46d66b2a28683c452d70373fa","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96026,"length_delta_samples":26,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-33.420965517659916,"output_peak":0.22881232377449132},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9997292399975006,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.203584983163687,"spectral_convergence":0.5900368088071997,"spectral_convergence_status":"ok","amplitude_floor":6.716189184049609e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.141467763063536,"spectral_convergence":0.33563016763718295,"spectral_convergence_status":"ok","amplitude_floor":4.979433430044932e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.877827495724432,"spectral_convergence":0.31600135419454334,"spectral_convergence_status":"ok","amplitude_floor":3.630042469105695e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.462171010234058e-05,"output_power":7.029218372859943e-05,"delta_db":-0.25958162011829744,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0003077822779549395,"output_power":0.00027592830924197506,"delta_db":-0.4744734982073167,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":5.338393512143839e-05,"output_power":5.1088969525900544e-05,"delta_db":-0.19083441066852053,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.1515857114073014e-05,"output_power":2.100461162980658e-05,"delta_db":-0.10443995539274885,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.940885952831811e-05,"output_power":2.983241754163222e-05,"delta_db":0.06210265933885389,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.017666424897867e-06,"output_power":6.850290376295314e-06,"delta_db":-0.6833900231420309,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.29580369447796134,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3636019011160618,"pitch":null,"output_rms_dbfs":-33.41991674871206,"output_peak":0.22881232377449132,"reference_rms_dbfs":-33.056314847596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.407626747317219},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8129286535640018,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.022242533719115894,"rms_dbfs":-33.056314847596,"peak":0.222198486328125,"peak_dbfs":-13.065178078162416,"crest_factor_db":19.991136769433584,"mean":-1.7630259195963542e-06,"zero_energy":false},"output":{"rms":0.021330653577520994,"rms_dbfs":-33.41991674871206,"peak":0.22881232377449132,"peak_dbfs":-12.810411765139484,"crest_factor_db":20.609504983572577,"mean":-0.00014525989040608335,"zero_energy":false},"rms_difference_db":-0.3636019011160627},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":164,"quiet_samples":26240,"reference_mean_power":9.42413263568064e-08,"output_mean_power":2.109572676248872e-07,"output_rms_dbfs":-66.75805508276808,"output_zero_energy":false,"output_peak":0.007724648174121777,"output_energy_fraction":0.00012672980431145087,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":152,"output_estimated_voiced_frames":128,"both_estimated_voiced_frames":112,"reference_voiced_output_unvoiced_frames":40,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.7368421052631579,"absolute_difference_cents_median":8.350377559361148,"absolute_difference_cents_p95":34.09427495670238,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.650394450462669e-05,"log_rms_mae_db":2.1768506559023977,"rms_envelope_nrmse":0.21181974055335673,"envelope_correlation":0.9598877874780016,"rms_gain_error_db":-0.3649223266805034,"reference_mean_power":0.0004951134756771775,"output_mean_power":0.0004552107548301793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.645854738282274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010486873457313227,"output_power":9.841295718184262e-05,"reference_active":true,"delta_db":-0.27593745515053303},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5071002541192514e-05,"output_power":2.3664323937268688e-05,"reference_active":true,"delta_db":-0.25077599223382574},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.059292105678353e-06,"output_power":4.83406075863455e-06,"reference_active":true,"delta_db":-0.1977765007114056}],"active_band_count":3,"active_band_mae_db":0.2414966493652548}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.020301765617229e-05,"log_rms_mae_db":1.42283673001945,"rms_envelope_nrmse":0.13319665993627292,"envelope_correlation":0.984142151360349,"rms_gain_error_db":-0.36593054816867415,"reference_mean_power":0.0004970274687625529,"output_mean_power":0.0004568644192067427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4721291240539383e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.182773781582673e-05,"output_power":8.620777579219044e-05,"reference_active":true,"delta_db":-0.274274453649072},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2533326164459844e-05,"output_power":1.1819588927608849e-05,"reference_active":true,"delta_db":-0.25463969277948123},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.472370238880856e-07,"output_power":2.6815725213099785e-07,"reference_active":true,"delta_db":0.35276039519567126}],"active_band_count":3,"active_band_mae_db":0.2938915138747415}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.432622539351727e-05,"log_rms_mae_db":2.7781453839304593,"rms_envelope_nrmse":0.24075729323120704,"envelope_correlation":0.9578884411024576,"rms_gain_error_db":-0.44331578399609595,"reference_mean_power":0.00026104952136561866,"output_mean_power":0.00023571722672722888},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.392840173310503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.633554932680893e-05,"output_power":8.757352265702145e-05,"reference_active":true,"delta_db":-0.4141375855395293},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.160826072139944e-05,"output_power":2.0094280700566907e-05,"reference_active":true,"delta_db":-0.3154734662791664},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.769645916048805e-06,"output_power":4.162286840178178e-06,"reference_active":true,"delta_db":-0.5915413362771211}],"active_band_count":3,"active_band_mae_db":0.4403841293652723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.77451695930137e-05,"log_rms_mae_db":2.1323980677591847,"rms_envelope_nrmse":0.15694398323552833,"envelope_correlation":0.9822175694559021,"rms_gain_error_db":-0.4434400751402202,"reference_mean_power":0.00026200159904973233,"output_mean_power":0.0002365701440938861},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2678124187221727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.604417147270571e-05,"output_power":7.788012380616296e-05,"reference_active":true,"delta_db":-0.4329482383388463},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1040946271380912e-05,"output_power":1.0274629372832994e-05,"reference_active":true,"delta_db":-0.3124013154642948},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9369342076961927e-07,"output_power":2.695457533304701e-07,"reference_active":true,"delta_db":-0.3726172375196743}],"active_band_count":3,"active_band_mae_db":0.37265559710760515}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.588832427367551e-05,"log_rms_mae_db":2.732236003395097,"rms_envelope_nrmse":0.27576217944859116,"envelope_correlation":0.9446983586723798,"rms_gain_error_db":-0.027380326411892496,"reference_mean_power":4.057559706591337e-05,"output_mean_power":4.032059075295103e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.7364679678483124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.645857845256653e-05,"output_power":1.639805263428913e-05,"reference_active":true,"delta_db":-0.016000458814019257},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.1440018553357532e-06,"output_power":2.7652259181205514e-06,"reference_active":true,"delta_db":-0.5575217483500187},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.079905260013692e-07,"output_power":8.783084984401354e-07,"reference_active":true,"delta_db":-0.14434232134091335}],"active_band_count":3,"active_band_mae_db":0.23928817616831707}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.0684503225281635e-05,"log_rms_mae_db":1.6600686550102912,"rms_envelope_nrmse":0.15643511533538135,"envelope_correlation":0.981050374414837,"rms_gain_error_db":-0.026831509677170302,"reference_mean_power":4.071142209359619e-05,"output_mean_power":4.0460674839439696e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.5164519386807125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5218014411010497e-05,"output_power":1.5112270551272856e-05,"reference_active":true,"delta_db":-0.030282709778248443},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2444589918408902e-06,"output_power":1.084346110038944e-06,"reference_active":true,"delta_db":-0.5981266422184562},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241409422416778e-08,"output_power":5.041824259332344e-08,"reference_active":true,"delta_db":-0.16860381498220495}],"active_band_count":3,"active_band_mae_db":0.2656710556596365}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.7354442023779843e-05,"log_rms_mae_db":3.148757968162786,"rms_envelope_nrmse":0.42781638055664106,"envelope_correlation":0.8569977568679419,"rms_gain_error_db":0.07389262127631867,"reference_mean_power":9.786447369332921e-06,"output_mean_power":9.954382553988238e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.593718108385403e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.378843387910107e-06,"output_power":2.4199855608403056e-06,"reference_active":true,"delta_db":0.0744692368304428},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.926849454092712e-07,"output_power":8.805004410572453e-07,"reference_active":true,"delta_db":-0.05968632873674649},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.698308970455951e-07,"output_power":1.740901214931849e-07,"reference_active":true,"delta_db":0.10757424816971928}],"active_band_count":3,"active_band_mae_db":0.08057660457896952}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.2351945663576383e-05,"log_rms_mae_db":1.9896835061756475,"rms_envelope_nrmse":0.25825414200274377,"envelope_correlation":0.9421310148286888,"rms_gain_error_db":0.07177311257837957,"reference_mean_power":9.831708602072054e-06,"output_mean_power":9.995541104725942e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.220673114788623e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.414723334287998e-06,"output_power":2.457831543164557e-06,"reference_active":true,"delta_db":0.07684734679940146},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7015119113313933e-07,"output_power":3.6382161743031696e-07,"reference_active":true,"delta_db":-0.07490651007032309},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1816168225696594e-08,"output_power":1.1262329023632913e-08,"reference_active":true,"delta_db":-0.20848454459496413}],"active_band_count":3,"active_band_mae_db":0.12007946715489622}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},{"case":"librispeech_5683-32865-0017","method":"hnm","status":"measured","mos":{"status":"measured","value":2.3027150630950928,"predicted_mos":2.3027150630950928,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87280,"input_sample_rate":16000,"input_seconds":5.455,"input_float32_sha256":"6c88bd68e5c7ccc959ac18ba2ee70f5aa613480b24f59b6120c9c404d4f2fe2b","internal_samples":87280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7999007660000643},"started_at":"2026-10-13T04:39:41.485509+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6031000086513814e-05,"native_input_samples":87280,"native_input_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","output_sha256":"f80c509bff3a3c7014c9251f5fe75b44ef536bb881b3dc65c76aab0bf4e2bea6","output_samples":87280,"length_delta_samples":0,"analysis_s":0.8308657899997343,"synthesis_s":0.44706573100029345,"total_s":1.2779315210000277,"rtf":0.23426792318973927,"output_peak":0.13522516010487368,"output_rms":0.017479522694952662,"finished_at":"2026-10-13T04:39:42.770320+00:00","evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"f80c509bff3a3c7014c9251f5fe75b44ef536bb881b3dc65c76aab0bf4e2bea6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.9162872287059995,"spectral_convergence":0.34605098420884983,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.712700102998662,"spectral_convergence":0.3408853524838566,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9049171035385335,"spectral_convergence":0.3568013974107628,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00012390320475122284,"delta_db":0.012877957695700316,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":7.299676446661409e-05,"delta_db":0.044720071321911206,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":4.225199498685534e-05,"delta_db":-0.10606551997284347,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":3.610186357980362e-05,"delta_db":-1.189529839312487,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":9.74929419265041e-06,"delta_db":0.3910150313879708,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":2.0530591666218716e-05,"delta_db":-0.13770399845661452,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.31365206969125453,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15535031442758687,"pitch":null,"output_rms_dbfs":-35.14940861226849,"output_peak":0.13522516010487368,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.844634811747731},"common_prefix":{"samples":87280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.9162872287059995,"spectral_convergence":0.34605098420884983,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.712700102998662,"spectral_convergence":0.3408853524838566,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9049171035385335,"spectral_convergence":0.3568013974107628,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00012390320475122284,"delta_db":0.012877957695700316,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":7.299676446661409e-05,"delta_db":0.044720071321911206,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":4.225199498685534e-05,"delta_db":-0.10606551997284347,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":3.610186357980362e-05,"delta_db":-1.189529839312487,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":9.74929419265041e-06,"delta_db":0.3910150313879708,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":2.0530591666218716e-05,"delta_db":-0.13770399845661452,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.31365206969125453,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15535031442758687,"pitch":null,"output_rms_dbfs":-35.14940861226849,"output_peak":0.13522516010487368,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.844634811747731},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87280,"output_samples":87280,"intelligibility":{"estoi":{"value":0.830147405884757,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.01779496282666337,"rms_dbfs":-34.9940582978409,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.990945259100272,"mean":-2.6739244609661434e-05,"zero_energy":false},"output":{"rms":0.017479522694952662,"rms_dbfs":-35.14940861226849,"peak":0.13522516010487368,"peak_dbfs":-17.378849913728757,"crest_factor_db":17.770558698539734,"mean":-4.285554747401209e-05,"zero_energy":false},"rms_difference_db":-0.1553503144275865},"full_scale":{"reference":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":5.00087841516101e-07,"output_rms_dbfs":-63.00953704193894,"output_zero_energy":false,"output_peak":0.005547924953177054,"output_energy_fraction":0.00034205603202720845,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":151,"both_estimated_voiced_frames":143,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9166666666666666,"absolute_difference_cents_median":1.576144424232396,"absolute_difference_cents_p95":11.640477690536677,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":1.1278709263183144,"rms_envelope_nrmse":0.09962546310864664,"envelope_correlation":0.9903986765306055,"rms_gain_error_db":-0.15535037459984086,"reference_mean_power":0.0003169510345489487,"output_mean_power":0.00030581384009693017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8938855835285014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.978895728357633e-05,"output_power":5.5323619842952375e-05,"reference_active":true,"delta_db":-0.33710391011985635},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1930505918996815e-05,"output_power":2.141751607014355e-05,"reference_active":true,"delta_db":-0.10279549178845929},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0863861439295835e-06,"output_power":3.0624489338284237e-06,"reference_active":true,"delta_db":-0.03381404979925117}],"active_band_count":3,"active_band_mae_db":0.1579044839025223}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":0.6200352004023112,"rms_envelope_nrmse":0.058304908367079064,"envelope_correlation":0.996605828851036,"rms_gain_error_db":-0.15534795424843406,"reference_mean_power":0.00031870366172679027,"output_mean_power":0.00030750505390100596},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":1.711395707762808e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.23137186055323e-05,"output_power":4.878894960817778e-05,"reference_active":true,"delta_db":-0.3029412383423393},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1932036565608145e-05,"output_power":1.1390473565175167e-05,"reference_active":true,"delta_db":-0.20172795097880042},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.569030812932547e-07,"output_power":2.284319097028169e-07,"reference_active":true,"delta_db":-0.5101254272022937}],"active_band_count":3,"active_band_mae_db":0.3382648721744778}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":1.8074063606093451,"rms_envelope_nrmse":0.12505811557466087,"envelope_correlation":0.9875525857257076,"rms_gain_error_db":-0.07308604550905541,"reference_mean_power":0.000101477720422738,"output_mean_power":9.978427315411618e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.361033224981284e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.73677738976729e-05,"output_power":2.7450876436509162e-05,"reference_active":true,"delta_db":0.01316741710196877},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1422451661963211e-05,"output_power":1.0816509191846739e-05,"reference_active":true,"delta_db":-0.23672205185711637},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1019925401091896e-06,"output_power":1.1473955765723745e-06,"reference_active":true,"delta_db":0.17534516674013384}],"active_band_count":3,"active_band_mae_db":0.14174487856640633}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":1.2065688144428215,"rms_envelope_nrmse":0.06546579998302075,"envelope_correlation":0.9963933783534016,"rms_gain_error_db":-0.07308691203595058,"reference_mean_power":0.00010203935243697765,"output_mean_power":0.00010033651270519871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":7.765913024651851e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5071474654741866e-05,"output_power":2.5195396922505245e-05,"reference_active":true,"delta_db":0.021413254661218368},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067872704197783e-06,"output_power":5.66968433709227e-06,"reference_active":true,"delta_db":-0.29477581323453234},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0845180908068795e-07,"output_power":8.087097956634878e-08,"reference_active":true,"delta_db":-1.274440972375321}],"active_band_count":3,"active_band_mae_db":0.5302100134236906}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":1.7642098155563188,"rms_envelope_nrmse":0.22161014192785775,"envelope_correlation":0.9683771781173225,"rms_gain_error_db":-0.7358389458805621,"reference_mean_power":4.068228265632885e-05,"output_mean_power":3.434167060847998e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.8091442271155458e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.649297963098176e-06,"output_power":6.256976997739453e-06,"reference_active":true,"delta_db":-1.406163001465679},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5842183345769264e-06,"output_power":2.2521084653867986e-06,"reference_active":true,"delta_db":-0.5973990035417734},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3501889520111636e-06,"output_power":1.6541903519866035e-06,"reference_active":true,"delta_db":0.8819093345689202}],"active_band_count":3,"active_band_mae_db":0.9618237798587908}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":1.253250887757049,"rms_envelope_nrmse":0.14562622612581708,"envelope_correlation":0.9888724902939598,"rms_gain_error_db":-0.7358428920766701,"reference_mean_power":4.090728037801193e-05,"output_mean_power":3.453156951913339e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":2.6173037083412677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.994446901443647e-06,"output_power":5.903042292977662e-06,"reference_active":true,"delta_db":-1.3171252752702758},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1581434874555988e-06,"output_power":1.0316054092118478e-06,"reference_active":true,"delta_db":-0.5024875867497995},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.731926648355597e-08,"output_power":1.0773931297382039e-07,"reference_active":true,"delta_db":0.9126412302671973}],"active_band_count":3,"active_band_mae_db":0.9107513640957575}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":1.6283848963954015,"rms_envelope_nrmse":0.2240491757297991,"envelope_correlation":0.9656341788251901,"rms_gain_error_db":-0.04283994514782775,"reference_mean_power":6.700122358829964e-06,"output_mean_power":6.634355501099015e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":5.0013650649342945e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1848288835729857e-06,"output_power":2.3516972385963203e-06,"reference_active":true,"delta_db":0.31963980641295353},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069865846115692e-06,"output_power":1.0824144008250365e-06,"reference_active":true,"delta_db":0.05064238104212963},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241288004170751e-07,"output_power":3.2798508643955216e-07,"reference_active":true,"delta_db":-2.035839277719745}],"active_band_count":3,"active_band_mae_db":0.8020404883916094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":1.0659375795758939,"rms_envelope_nrmse":0.12724011489136264,"envelope_correlation":0.9878256819358339,"rms_gain_error_db":-0.042845675100667585,"reference_mean_power":6.736890999373383e-06,"output_mean_power":6.670754427934613e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":4.500371991965456e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.395916806498351e-06,"output_power":2.539748821106371e-06,"reference_active":true,"delta_db":0.2531903344988446},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.187368503320663e-07,"output_power":5.509806620684917e-07,"reference_active":true,"delta_db":-0.5036962566499329},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.878047962583858e-08,"output_power":2.7858890235907404e-08,"reference_active":true,"delta_db":-2.432822538951863}],"active_band_count":3,"active_band_mae_db":1.0632363767002133}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32865-0017","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.375115394592285,"predicted_mos":2.375115394592285,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87280,"input_sample_rate":16000,"input_seconds":5.455,"input_float32_sha256":"c02159266825d2a1e9c5d49e8d37e4246d989ffcba9f0e72a8aad0dd4076461c","internal_samples":87280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.621178871999291},"started_at":"2026-10-13T04:42:44.365121+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2512000214192085e-05,"native_input_samples":87280,"native_input_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","output_sha256":"f6e266aab00c8ccbf45179bab2636092d827a296900bc737c37ef7eb80dff116","output_samples":87280,"length_delta_samples":0,"analysis_s":2.1117919090002033,"synthesis_s":0.12374458900012542,"total_s":2.2355364980003287,"rtf":0.40981420678282837,"output_peak":0.170615043779736,"output_rms":0.01942634575330791,"finished_at":"2026-10-13T04:42:46.608685+00:00","evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"f6e266aab00c8ccbf45179bab2636092d827a296900bc737c37ef7eb80dff116","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.3385481789226095,"spectral_convergence":0.5804902461738741,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.897342354600703,"spectral_convergence":0.3396612913416568,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.13625175946145,"spectral_convergence":0.3687529534162222,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00013746064461344555,"delta_db":0.4638363241530605,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":9.047086717954488e-05,"delta_db":0.9767714952870166,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":5.319044448649817e-05,"delta_db":0.8937984793544624,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":6.186503882888666e-05,"delta_db":1.1496268512605252,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":1.0716833304119191e-05,"delta_db":0.8019480286698899,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":2.367908091456954e-05,"delta_db":0.4819297634401212,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7946518236941792,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7618805801094769,"pitch":null,"output_rms_dbfs":-34.232177717731425,"output_peak":0.170615043779736,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.124047430994921},"common_prefix":{"samples":87280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.3385481789226095,"spectral_convergence":0.5804902461738741,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.897342354600703,"spectral_convergence":0.3396612913416568,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.13625175946145,"spectral_convergence":0.3687529534162222,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00013746064461344555,"delta_db":0.4638363241530605,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":9.047086717954488e-05,"delta_db":0.9767714952870166,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":5.319044448649817e-05,"delta_db":0.8937984793544624,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":6.186503882888666e-05,"delta_db":1.1496268512605252,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":1.0716833304119191e-05,"delta_db":0.8019480286698899,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":2.367908091456954e-05,"delta_db":0.4819297634401212,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7946518236941792,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7618805801094769,"pitch":null,"output_rms_dbfs":-34.232177717731425,"output_peak":0.170615043779736,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.124047430994921},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87280,"output_samples":87280,"intelligibility":{"estoi":{"value":0.8710915954515431,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.01779496282666337,"rms_dbfs":-34.9940582978409,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.990945259100272,"mean":-2.6739244609661434e-05,"zero_energy":false},"output":{"rms":0.01942634575330791,"rms_dbfs":-34.232177717731425,"peak":0.170615043779736,"peak_dbfs":-15.359653561555252,"crest_factor_db":18.872524156176173,"mean":1.7746624744193801e-09,"zero_energy":false},"rms_difference_db":0.7618805801094766},"full_scale":{"reference":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":2.7100644560677376e-07,"output_rms_dbfs":-65.67020379759671,"output_zero_energy":false,"output_peak":0.005090378944769608,"output_energy_fraction":0.00015007470157348053,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":193,"both_estimated_voiced_frames":145,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":48,"paired_voiced_fraction":0.9294871794871795,"absolute_difference_cents_median":5.442636952308737,"absolute_difference_cents_p95":20.715646948112255,"difference_over_600_cents_frames":4,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":2.5529654022204333,"rms_envelope_nrmse":0.16440355771849025,"envelope_correlation":0.983376460750159,"rms_gain_error_db":0.7618821732063019,"reference_mean_power":0.0003169510345489487,"output_mean_power":0.0003777290540079756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8938855835285014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.978895728357633e-05,"output_power":7.231103158782298e-05,"reference_active":true,"delta_db":0.8258357780286092},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1930505918996815e-05,"output_power":2.548305189924297e-05,"reference_active":true,"delta_db":0.6520278810463815},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0863861439295835e-06,"output_power":3.031320392555142e-06,"reference_active":true,"delta_db":-0.07818419435712942}],"active_band_count":3,"active_band_mae_db":0.5186826178107067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":2.286149812360442,"rms_envelope_nrmse":0.1261375005891333,"envelope_correlation":0.9926664393698758,"rms_gain_error_db":0.7618960189539304,"reference_mean_power":0.00031870366172679027,"output_mean_power":0.00037981897304759646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":1.711395707762808e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.23137186055323e-05,"output_power":6.402252576707677e-05,"reference_active":true,"delta_db":0.8771721160352015},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1932036565608145e-05,"output_power":1.3666427701813664e-05,"reference_active":true,"delta_db":0.5894043261020031},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.569030812932547e-07,"output_power":2.038049500113671e-07,"reference_active":true,"delta_db":-1.005545852857698}],"active_band_count":3,"active_band_mae_db":0.8240407649983008}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":2.155335832608401,"rms_envelope_nrmse":0.20677904230988475,"envelope_correlation":0.9811160950929392,"rms_gain_error_db":1.027000268792992,"reference_mean_power":0.000101477720422738,"output_mean_power":0.00012854960000363127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.361033224981284e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.73677738976729e-05,"output_power":3.550996917919527e-05,"reference_active":true,"delta_db":1.1311082214376593},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1422451661963211e-05,"output_power":1.361695621723558e-05,"reference_active":true,"delta_db":0.7632071220743335},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1019925401091896e-06,"output_power":9.273819449985275e-07,"reference_active":true,"delta_db":-0.749200181326506}],"active_band_count":3,"active_band_mae_db":0.8811718416128329}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":1.8581292737703246,"rms_envelope_nrmse":0.17817412435592536,"envelope_correlation":0.9880366187859078,"rms_gain_error_db":1.026999412677411,"reference_mean_power":0.00010203935243697765,"output_mean_power":0.00012926103680699426},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":7.765913024651851e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5071474654741866e-05,"output_power":3.2356766728264354e-05,"reference_active":true,"delta_db":1.1078523888412553},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067872704197783e-06,"output_power":7.229064790907805e-06,"reference_active":true,"delta_db":0.760456558570432},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0845180908068795e-07,"output_power":8.850855467379274e-08,"reference_active":true,"delta_db":-0.8825155199299579}],"active_band_count":3,"active_band_mae_db":0.9169414891138817}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":1.7072296782210696,"rms_envelope_nrmse":0.2527703983642792,"envelope_correlation":0.9735599405977903,"rms_gain_error_db":1.1433195878698137,"reference_mean_power":4.068228265632885e-05,"output_mean_power":5.293431186335632e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.8091442271155458e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.649297963098176e-06,"output_power":1.1915118546084158e-05,"reference_active":true,"delta_db":1.391175090448914},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5842183345769264e-06,"output_power":3.348530544676122e-06,"reference_active":true,"delta_db":1.1252506144364707},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3501889520111636e-06,"output_power":1.2799474375220925e-06,"reference_active":true,"delta_db":-0.23202414807167493}],"active_band_count":3,"active_band_mae_db":0.9161499509856865}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":1.2611277480470804,"rms_envelope_nrmse":0.18976128774806492,"envelope_correlation":0.9896872742656474,"rms_gain_error_db":1.1433239301133398,"reference_mean_power":4.090728037801193e-05,"output_mean_power":5.3227123963377876e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":2.6173037083412677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.994446901443647e-06,"output_power":1.1059257518254153e-05,"reference_active":true,"delta_db":1.4093754855253797},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1581434874555988e-06,"output_power":1.4917838596216625e-06,"reference_active":true,"delta_db":1.099435346202393},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.731926648355597e-08,"output_power":7.402761278582938e-08,"reference_active":true,"delta_db":-0.7171633416115618}],"active_band_count":3,"active_band_mae_db":1.0753247244464448}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":1.6474546379430914,"rms_envelope_nrmse":0.3599892148636286,"envelope_correlation":0.9471876488705203,"rms_gain_error_db":1.4819041489766438,"reference_mean_power":6.700122358829964e-06,"output_mean_power":9.424821828430299e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":5.0013650649342945e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1848288835729857e-06,"output_power":3.144767509948469e-06,"reference_active":true,"delta_db":1.581711153679716},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069865846115692e-06,"output_power":1.8254253396649266e-06,"reference_active":true,"delta_db":2.3203475136872047},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241288004170751e-07,"output_power":8.416291956607352e-07,"reference_active":true,"delta_db":2.056827680527418}],"active_band_count":3,"active_band_mae_db":1.9862954492981129}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":1.1671355080893158,"rms_envelope_nrmse":0.2674500549107192,"envelope_correlation":0.9782462294692288,"rms_gain_error_db":1.4819757820167825,"reference_mean_power":6.736890999373383e-06,"output_mean_power":9.476699264906926e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":4.500371991965456e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.395916806498351e-06,"output_power":3.656016829668469e-06,"reference_active":true,"delta_db":1.8353645228856412},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.187368503320663e-07,"output_power":1.0626542426986934e-06,"reference_active":true,"delta_db":2.3488599828009074},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.878047962583858e-08,"output_power":6.636447551749901e-08,"reference_active":true,"delta_db":1.3368960036157849}],"active_band_count":3,"active_band_mae_db":1.8403735031007777}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32865-0017","method":"facodec","status":"measured","mos":{"status":"measured","value":3.3434979915618896,"predicted_mos":3.3434979915618896,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87400,"input_sample_rate":16000,"input_seconds":5.4625,"input_float32_sha256":"ac5f0a6ed0bd4ba08a972a0b35d6bcded4b30af4a2afb9a0575e1676a4cbc7d4","internal_samples":87400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.763761105999947},"started_at":"2026-10-13T04:55:13.604782+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.391599991824478e-05,"native_input_samples":87280,"native_input_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","output_sha256":"f7e5ce5174cbe38bc4aa3c095e65fe3bb486f86002b6d1dfbbdfeef35b1a7d4e","output_samples":87400,"length_delta_samples":120,"analysis_s":1.2603516600001967,"synthesis_s":10.403124172000389,"total_s":11.663475832000586,"rtf":2.138125725389658,"output_peak":0.173705592751503,"output_rms":0.016964556907139956,"finished_at":"2026-10-13T04:55:25.273883+00:00","evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"f7e5ce5174cbe38bc4aa3c095e65fe3bb486f86002b6d1dfbbdfeef35b1a7d4e","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":87280,"output_samples":87400,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-35.40914958191233,"output_peak":0.173705592751503},"common_prefix":{"samples":87280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986270022883296,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.712419280996198,"spectral_convergence":0.4320220417733235,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.16232331006401,"spectral_convergence":0.39964771118821435,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.171109229925716,"spectral_convergence":0.43685122684769045,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00011364976124801961,"delta_db":-0.362262160511881,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":7.217006875660929e-05,"delta_db":-0.004744845472847188,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":3.6044894764949717e-05,"delta_db":-0.7961000967178113,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":3.653859692470953e-05,"delta_db":-1.1373073820621733,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":6.69857528523947e-06,"delta_db":-1.2388922986085504,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":2.3089975605518913e-05,"delta_db":0.3725160886119766,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6519704786642065,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.40912441278664646,"pitch":null,"output_rms_dbfs":-35.40318271062755,"output_peak":0.173705592751503,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.015283940328641},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87280,"output_samples":87280,"intelligibility":{"estoi":{"value":0.8131537933123986,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.01779496282666337,"rms_dbfs":-34.9940582978409,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.990945259100272,"mean":-2.6739244609661434e-05,"zero_energy":false},"output":{"rms":0.01697621490748296,"rms_dbfs":-35.40318271062755,"peak":0.173705592751503,"peak_dbfs":-15.20372396934074,"crest_factor_db":20.199458741286808,"mean":2.6052973991017532e-05,"zero_energy":false},"rms_difference_db":-0.409124412786646},"full_scale":{"reference":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":4.3414588882171833e-07,"output_rms_dbfs":-63.623643072072255,"output_zero_energy":false,"output_peak":0.00490319449454546,"output_energy_fraction":0.0003148212654497962,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":168,"both_estimated_voiced_frames":143,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.9166666666666666,"absolute_difference_cents_median":9.139004541647676,"absolute_difference_cents_p95":36.87203780182101,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":2.07691034131562,"rms_envelope_nrmse":0.19165194036448596,"envelope_correlation":0.9642153568253826,"rms_gain_error_db":-0.4091240310391281,"reference_mean_power":0.0003169510345489487,"output_mean_power":0.0002884561286447151},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8938855835285014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.978895728357633e-05,"output_power":4.311878253279259e-05,"reference_active":true,"delta_db":-1.419544894494912},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1930505918996815e-05,"output_power":1.962846817303492e-05,"reference_active":true,"delta_db":-0.48162242505177083},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0863861439295835e-06,"output_power":3.3224725551404533e-06,"reference_active":true,"delta_db":0.320111416198519}],"active_band_count":3,"active_band_mae_db":0.7404262452484006}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":1.7404530666060314,"rms_envelope_nrmse":0.1769067131920261,"envelope_correlation":0.9665009067220941,"rms_gain_error_db":-0.40912294680965033,"reference_mean_power":0.00031870366172679027,"output_mean_power":0.00029005126147812435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":1.711395707762808e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.23137186055323e-05,"output_power":3.6738762166932505e-05,"reference_active":true,"delta_db":-1.5349107238815853},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1932036565608145e-05,"output_power":1.1109440868244643e-05,"reference_active":true,"delta_db":-0.31022373883262844},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.569030812932547e-07,"output_power":2.800951533026257e-07,"reference_active":true,"delta_db":0.3753628074239452}],"active_band_count":3,"active_band_mae_db":0.7401657567127197}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":2.2284658598291633,"rms_envelope_nrmse":0.17040246025458208,"envelope_correlation":0.9779216951367585,"rms_gain_error_db":-0.46289736997313047,"reference_mean_power":0.000101477720422738,"output_mean_power":9.121808818667195e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.361033224981284e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.73677738976729e-05,"output_power":2.197876303024472e-05,"reference_active":true,"delta_db":-0.9523622662944425},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1422451661963211e-05,"output_power":8.925592398651863e-06,"reference_active":true,"delta_db":-1.0712227865471609},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1019925401091896e-06,"output_power":9.105793158094686e-07,"reference_active":true,"delta_db":-0.8286087368485076}],"active_band_count":3,"active_band_mae_db":0.950731263230037}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":1.954920712111146,"rms_envelope_nrmse":0.15112653435487516,"envelope_correlation":0.9817905988680203,"rms_gain_error_db":-0.4628982724968733,"reference_mean_power":0.00010203935243697765,"output_mean_power":9.172291884402151e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":7.765913024651851e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5071474654741866e-05,"output_power":2.021704896098226e-05,"reference_active":true,"delta_db":-0.9346211620254425},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067872704197783e-06,"output_power":4.780965297809156e-06,"reference_active":true,"delta_db":-1.0352086987534614},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0845180908068795e-07,"output_power":7.091137803472468e-08,"reference_active":true,"delta_db":-1.8452087567467792}],"active_band_count":3,"active_band_mae_db":1.2716795391752278}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":2.0108853520764467,"rms_envelope_nrmse":0.23529487129119037,"envelope_correlation":0.9676796927549638,"rms_gain_error_db":-1.078898743502844,"reference_mean_power":4.068228265632885e-05,"output_mean_power":3.173331469345582e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.8091442271155458e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.649297963098176e-06,"output_power":6.698806933863722e-06,"reference_active":true,"delta_db":-1.1098339736677287},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5842183345769264e-06,"output_power":2.1368206354669977e-06,"reference_active":true,"delta_db":-0.8256113434651524},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3501889520111636e-06,"output_power":1.1797621923938392e-06,"reference_active":true,"delta_db":-0.586000757243768}],"active_band_count":3,"active_band_mae_db":0.8404820247922163}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":1.583252162647803,"rms_envelope_nrmse":0.20738586219600078,"envelope_correlation":0.975650180470401,"rms_gain_error_db":-1.0789104376488956,"reference_mean_power":4.090728037801193e-05,"output_mean_power":3.190873326928594e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":2.6173037083412677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.994446901443647e-06,"output_power":6.268163427567514e-06,"reference_active":true,"delta_db":-1.0564811116083952},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1581434874555988e-06,"output_power":1.0446385162757235e-06,"reference_active":true,"delta_db":-0.4479633494297052},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.731926648355597e-08,"output_power":8.290374928557447e-08,"reference_active":true,"delta_db":-0.22535907035445868}],"active_band_count":3,"active_band_mae_db":0.5766011771308531}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":2.1578333567965924,"rms_envelope_nrmse":0.3984268177574602,"envelope_correlation":0.8876540807990564,"rms_gain_error_db":-0.6107209066475945,"reference_mean_power":6.700122358829964e-06,"output_mean_power":5.82117483571935e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":5.0013650649342945e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1848288835729857e-06,"output_power":1.61845060153405e-06,"reference_active":true,"delta_db":-1.3031798019615026},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069865846115692e-06,"output_power":7.892288785526584e-07,"reference_active":true,"delta_db":-1.3212635545801743},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241288004170751e-07,"output_power":4.999748123392607e-07,"reference_active":true,"delta_db":-0.20489898397666292}],"active_band_count":3,"active_band_mae_db":0.9431141135061133}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":1.8022331615465728,"rms_envelope_nrmse":0.36184117554990103,"envelope_correlation":0.8980132443936548,"rms_gain_error_db":-0.6106202279853372,"reference_mean_power":6.736890999373383e-06,"output_mean_power":5.853255715507642e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":4.500371991965456e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.395916806498351e-06,"output_power":1.8203517214209149e-06,"reference_active":true,"delta_db":-1.1931642513970715},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.187368503320663e-07,"output_power":4.843386855756143e-07,"reference_active":true,"delta_db":-1.063568233944507},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.878047962583858e-08,"output_power":4.9124368951309336e-08,"reference_active":true,"delta_db":0.030509183725575883}],"active_band_count":3,"active_band_mae_db":0.7624138896890514}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32865-0017","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.892976760864258,"predicted_mos":2.892976760864258,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":240640,"input_sample_rate":44100,"input_seconds":5.456689342403628,"input_float32_sha256":"1be6dc1964362f93fa96a2456f7ea5c517f0174831d647a1dcabc70a27f13252","internal_samples":87308,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.557947501001763},"started_at":"2026-10-13T07:20:55.717972+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007270574000358465,"native_input_samples":240566,"native_input_sha256":"8dd6e5a949535c5433105527b0715d6539be9d5eb06081d6102215595a12e98a","output_sha256":"30ad49272972233e05a03794fc060a5428884c3e98a23769a46af2c2d0fd8dc7","output_samples":240640,"length_delta_samples":74,"analysis_s":0.12736191499971028,"synthesis_s":29.71745692500008,"total_s":29.84481883999979,"rtf":5.471082824854679,"output_peak":0.17064327001571655,"output_rms":0.015863147549173732,"finished_at":"2026-10-13T07:21:25.595234+00:00","evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"30ad49272972233e05a03794fc060a5428884c3e98a23769a46af2c2d0fd8dc7","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":87280,"output_samples":87308,"length_delta_samples":28,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-36.00256767981036,"output_peak":0.16496457140857873},"common_prefix":{"samples":87280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996792962844184,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.528755173187983,"spectral_convergence":0.5556640151923584,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.627541516137526,"spectral_convergence":0.4618440633889179,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.227344843938145,"spectral_convergence":0.505525527515084,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":8.662265703795433e-05,"delta_db":-1.5416324274083715,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":7.186527598901905e-05,"delta_db":-0.023125058315867378,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":4.076871069877957e-05,"delta_db":-0.26126794381685464,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":2.815766483050215e-05,"delta_db":-2.2688596963226164,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":7.999837999094301e-06,"delta_db":-0.46790480237718046,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":1.5706538758532763e-05,"delta_db":-1.300913748824187,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.977283946177513,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0071168232309817,"pitch":null,"output_rms_dbfs":-36.00117512107188,"output_peak":0.16496457140857873,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.461213844421218},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87280,"output_samples":87280,"intelligibility":{"estoi":{"value":0.6740064262446115,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.01779496282666337,"rms_dbfs":-34.9940582978409,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.990945259100272,"mean":-2.6739244609661434e-05,"zero_energy":false},"output":{"rms":0.015846787854763573,"rms_dbfs":-36.00117512107188,"peak":0.16496457140857873,"peak_dbfs":-15.652186339221393,"crest_factor_db":20.348988781850494,"mean":-1.864291471264553e-05,"zero_energy":false},"rms_difference_db":-1.0071168232309802},"full_scale":{"reference":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":4.3715577231097083e-07,"output_rms_dbfs":-63.593637827079306,"output_zero_energy":false,"output_peak":0.00663794803310518,"output_energy_fraction":0.00036380094748811564,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":169,"both_estimated_voiced_frames":140,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.8974358974358975,"absolute_difference_cents_median":7.332944374517477,"absolute_difference_cents_p95":26.26446051669747,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":1.9323814648605806,"rms_envelope_nrmse":0.16055194596082153,"envelope_correlation":0.9850965999563537,"rms_gain_error_db":-1.0071155114346422,"reference_mean_power":0.0003169510345489487,"output_mean_power":0.0002513510029657175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8938855835285014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.978895728357633e-05,"output_power":4.9868029596220823e-05,"reference_active":true,"delta_db":-0.7879877085680527},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1930505918996815e-05,"output_power":1.5767975995582262e-05,"reference_active":true,"delta_db":-1.4327270049219023},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0863861439295835e-06,"output_power":3.2285016131615935e-06,"reference_active":true,"delta_db":0.1955074716691417}],"active_band_count":3,"active_band_mae_db":0.8054073950530323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":1.5574334621962078,"rms_envelope_nrmse":0.1357483935882791,"envelope_correlation":0.9923759394429346,"rms_gain_error_db":-1.00710518697777,"reference_mean_power":0.00031870366172679027,"output_mean_power":0.0002527414859988431},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":1.711395707762808e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.23137186055323e-05,"output_power":4.336482348773486e-05,"reference_active":true,"delta_db":-0.8147800901514364},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1932036565608145e-05,"output_power":8.626537417902796e-06,"reference_active":true,"delta_db":-1.408780651584662},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.569030812932547e-07,"output_power":1.9812748441453945e-07,"reference_active":true,"delta_db":-1.1282458780115405}],"active_band_count":3,"active_band_mae_db":1.1172688732492129}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":4.017275518072368,"rms_envelope_nrmse":0.24374918842829876,"envelope_correlation":0.9524214103016552,"rms_gain_error_db":-0.4047363412528216,"reference_mean_power":0.000101477720422738,"output_mean_power":9.244790388031427e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.361033224981284e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.73677738976729e-05,"output_power":2.86903285219663e-05,"reference_active":true,"delta_db":0.20496048001279332},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1422451661963211e-05,"output_power":8.501119607924561e-06,"reference_active":true,"delta_db":-1.28283202265664},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1019925401091896e-06,"output_power":1.1178915011593357e-06,"reference_active":true,"delta_db":0.062209998260762754}],"active_band_count":3,"active_band_mae_db":0.5166675003100654}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":3.6184149234525593,"rms_envelope_nrmse":0.20949265725693553,"envelope_correlation":0.9626769721263032,"rms_gain_error_db":-0.4047427749047339,"reference_mean_power":0.00010203935243697765,"output_mean_power":9.295942234670155e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":7.765913024651851e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5071474654741866e-05,"output_power":2.5626808220229226e-05,"reference_active":true,"delta_db":0.09514639783668655},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067872704197783e-06,"output_power":4.538961711233612e-06,"reference_active":true,"delta_db":-1.260799420633276},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0845180908068795e-07,"output_power":8.817069397365823e-08,"reference_active":true,"delta_db":-0.8991254181155885}],"active_band_count":3,"active_band_mae_db":0.7516904121951837}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":3.682289698386977,"rms_envelope_nrmse":0.3297661498355233,"envelope_correlation":0.9466771877861987,"rms_gain_error_db":-2.1471703446764803,"reference_mean_power":4.068228265632885e-05,"output_mean_power":2.4813514377504686e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.8091442271155458e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.649297963098176e-06,"output_power":5.639037745578913e-06,"reference_active":true,"delta_db":-1.857758569479941},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5842183345769264e-06,"output_power":1.7279801073201274e-06,"reference_active":true,"delta_db":-1.7479046487040795},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3501889520111636e-06,"output_power":1.09958904978132e-06,"reference_active":true,"delta_db":-0.8916414374423394}],"active_band_count":3,"active_band_mae_db":1.4991015518754534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":3.074713745703476,"rms_envelope_nrmse":0.28481396701105605,"envelope_correlation":0.9688948595036204,"rms_gain_error_db":-2.1471840138152793,"reference_mean_power":4.090728037801193e-05,"output_mean_power":2.4950669645780945e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":2.6173037083412677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.994446901443647e-06,"output_power":5.2375554978496396e-06,"reference_active":true,"delta_db":-1.8365978446655218},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1581434874555988e-06,"output_power":9.025433712921767e-07,"reference_active":true,"delta_db":-1.0829428847148828},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.731926648355597e-08,"output_power":6.754792943133885e-08,"reference_active":true,"delta_db":-1.1149803781465326}],"active_band_count":3,"active_band_mae_db":1.3448403691756459}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":4.730046491689702,"rms_envelope_nrmse":0.40478350633137494,"envelope_correlation":0.908051066852236,"rms_gain_error_db":0.6098230028512901,"reference_mean_power":6.700122358829964e-06,"output_mean_power":7.710189180585435e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":5.0013650649342945e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1848288835729857e-06,"output_power":3.0175764764148767e-06,"reference_active":true,"delta_db":1.402408569485104},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069865846115692e-06,"output_power":1.2765861316880497e-06,"reference_active":true,"delta_db":0.7672079858940478},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241288004170751e-07,"output_power":3.5117827277217427e-07,"reference_active":true,"delta_db":-1.739103859802381}],"active_band_count":3,"active_band_mae_db":1.302906805060511}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":4.221489054246495,"rms_envelope_nrmse":0.35074131496297845,"envelope_correlation":0.9267321679949014,"rms_gain_error_db":0.609976166530716,"reference_mean_power":6.736890999373383e-06,"output_mean_power":7.752774236105704e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":4.500371991965456e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.395916806498351e-06,"output_power":3.1952111324196604e-06,"reference_active":true,"delta_db":1.2502782669571564},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.187368503320663e-07,"output_power":6.679952926903045e-07,"reference_active":true,"delta_db":0.33267419825478406},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.878047962583858e-08,"output_power":3.221328144685042e-08,"reference_active":true,"delta_db":-1.8021109909353568}],"active_band_count":3,"active_band_mae_db":1.1283544853824325}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32865-0017","method":"golf","status":"measured","mos":{"status":"measured","value":2.3541061878204346,"predicted_mos":2.3541061878204346,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":130800,"input_sample_rate":24000,"input_seconds":5.45,"input_float32_sha256":"68d6bcc96b4cec7692e581012abd43db5c2c5a052d49cc19afe6e6bd00eea0f4","internal_samples":87200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.442002268999204},"started_at":"2026-10-13T05:02:45.371861+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031549850000374136,"native_input_samples":130920,"native_input_sha256":"f0c367061358b083702e5b76462f1d9af60724ae244f54f83d9ff7b0a394a036","output_sha256":"e1edfc7849597f29a3b3c8793bef56e11ba9b76ac510508dd216e0238ac13cdd","output_samples":130800,"length_delta_samples":-120,"analysis_s":0.7289418450000085,"synthesis_s":0.12994419899951026,"total_s":0.8588860439995187,"rtf":0.15744932062319317,"output_peak":0.1272381842136383,"output_rms":0.015012793641514114,"finished_at":"2026-10-13T05:02:46.241802+00:00","analysis_stages":{"external_f0_s":0.13749718100007158},"evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"e1edfc7849597f29a3b3c8793bef56e11ba9b76ac510508dd216e0238ac13cdd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":87280,"output_samples":87200,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-36.471005904439416,"output_peak":0.1268344029240506},"common_prefix":{"samples":87200,"sample_rate":16000,"reference_coverage":0.999083409715857,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87200,"output_samples":87200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.370765050620033,"spectral_convergence":0.5359060447376125,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.920282484613471,"spectral_convergence":0.36741379285195735,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.984395067078704,"spectral_convergence":0.3881703280479609,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001238134264151971,"output_power":9.846050896480857e-05,"delta_db":-0.9950566602772988,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.220948683224596e-05,"output_power":5.294790368996052e-05,"delta_db":-1.3474548840121183,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.334860686294624e-05,"output_power":2.7835718792764713e-05,"delta_db":-1.9237270425147295,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.745223562906422e-05,"output_power":2.915128458303227e-05,"delta_db":-2.115989810950628,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.916201293865722e-06,"output_power":5.5791376433908125e-06,"delta_db":-2.0361278857462866,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1210983342042276e-05,"output_power":1.1397161336899779e-05,"delta_db":-2.697641067525263,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.8526662251710542,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.4809263508080361,"pitch":null,"output_rms_dbfs":-36.471005904439416,"output_peak":0.1268344029240506,"reference_rms_dbfs":-34.990079553631375,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.091814200770735},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87200,"output_samples":87200,"intelligibility":{"estoi":{"value":0.8424388590742207,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.017803116029935926,"rms_dbfs":-34.990079553631375,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.986966514890746,"mean":-2.6778824832461295e-05,"zero_energy":false},"output":{"rms":0.015012385387101436,"rms_dbfs":-36.471005904439416,"peak":0.1268344029240506,"peak_dbfs":-17.935258624105657,"crest_factor_db":18.535747280333755,"mean":7.359143632833399e-05,"zero_energy":false},"rms_difference_db":-1.4809263508080406},"full_scale":{"reference":{"sample_count":87200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":5.479288826661835e-07,"output_rms_dbfs":-62.612758062418024,"output_zero_energy":false,"output_peak":0.004508301522415967,"output_energy_fraction":0.0005085493847836741,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":164,"both_estimated_voiced_frames":137,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.8782051282051282,"absolute_difference_cents_median":4.224019367212973,"absolute_difference_cents_p95":20.332787618656095,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":2.527347995042686,"rms_envelope_nrmse":0.2145187466914207,"envelope_correlation":0.9755148999551794,"rms_gain_error_db":-1.4809249860473284,"reference_mean_power":0.0003172417909451704,"output_mean_power":0.00022557859863622383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":1.894082739141712e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.9845447639507514e-05,"output_power":4.511741796613354e-05,"reference_active":true,"delta_db":-1.2268688245841426},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.169219528182916e-05,"output_power":1.4869433867543915e-05,"reference_active":true,"delta_db":-1.6400907177922863},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2744128584126e-06,"output_power":3.4961785848916274e-06,"reference_active":true,"delta_db":0.28460171214635616}],"active_band_count":3,"active_band_mae_db":1.0505204181742616}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1083,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":2.024283384786426,"rms_envelope_nrmse":0.19068061471678657,"envelope_correlation":0.9852529902959383,"rms_gain_error_db":-1.4809690297575946,"reference_mean_power":0.0003189976957516138,"output_mean_power":0.0002268248555677123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1083,"frequency_resolution_hz":0.18467220683287164,"power_floor":1.7115776286008471e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.234125913315479e-05,"output_power":3.979744050439164e-05,"reference_active":true,"delta_db":-1.1898902375565166},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1906687650500704e-05,"output_power":8.492876240277801e-06,"reference_active":true,"delta_db":-1.4673616536596428},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.562939193894861e-07,"output_power":2.391498899366955e-07,"reference_active":true,"delta_db":-0.3006811727286305}],"active_band_count":3,"active_band_mae_db":0.9859776879815966}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":2.2755951645644514,"rms_envelope_nrmse":0.2486910040821429,"envelope_correlation":0.9669444735995331,"rms_gain_error_db":-1.5576043353871059,"reference_mean_power":0.0001015708994859399,"output_mean_power":7.095922521961639e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":8.357055360766472e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.742564688812439e-05,"output_power":2.0135327650899613e-05,"reference_active":true,"delta_db":-1.3419817903026257},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1265731738744465e-05,"output_power":6.8242593576342395e-06,"reference_active":true,"delta_db":-2.177038815982129},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1851401410246642e-06,"output_power":1.291888403849715e-06,"reference_active":true,"delta_db":0.3745529192396596}],"active_band_count":3,"active_band_mae_db":1.2978578418414715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1083,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":1.8806256194090514,"rms_envelope_nrmse":0.21594714697613526,"envelope_correlation":0.9799592257259986,"rms_gain_error_db":-1.557604241489214,"reference_mean_power":0.00010213356861079314,"output_mean_power":7.135231735574406e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1083,"frequency_resolution_hz":0.18467220683287164,"power_floor":7.762310486275802e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5118046470477567e-05,"output_power":1.830810856884972e-05,"reference_active":true,"delta_db":-1.373423804130835},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.01371909256978e-06,"output_power":3.732648747994094e-06,"reference_active":true,"delta_db":-2.071260142626638},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0780733460945745e-07,"output_power":1.1329985538111412e-07,"reference_active":true,"delta_db":0.21581046690006653}],"active_band_count":3,"active_band_mae_db":1.2201648045525133}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":2.574403163733409,"rms_envelope_nrmse":0.31133387453642875,"envelope_correlation":0.9509011131349329,"rms_gain_error_db":-1.9075038781875024,"reference_mean_power":4.071959655612039e-05,"output_mean_power":2.6245392897343112e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":2.8082189878144656e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.654844846214226e-06,"output_power":5.1528421160171845e-06,"reference_active":true,"delta_db":-2.252124509590779},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517253264640107e-06,"output_power":2.152470202521265e-06,"reference_active":true,"delta_db":-0.6798976470760016},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3897105990527486e-06,"output_power":9.88528676634167e-07,"reference_active":true,"delta_db":-1.4793509734855}],"active_band_count":3,"active_band_mae_db":1.4704577100507603}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1083,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":2.2366661615998353,"rms_envelope_nrmse":0.2654664181922957,"envelope_correlation":0.9715205462710708,"rms_gain_error_db":-1.907509885270056,"reference_mean_power":4.094502196523588e-05,"output_mean_power":2.6390651999597632e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1083,"frequency_resolution_hz":0.18467220683287164,"power_floor":2.6165823588530447e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.991877506241595e-06,"output_power":4.857652921965969e-06,"reference_active":true,"delta_db":-2.1622233736486915},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.139590735761854e-06,"output_power":1.092669550568713e-06,"reference_active":true,"delta_db":-0.1826006925204352},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.693214774137235e-08,"output_power":6.354329250154215e-08,"reference_active":true,"delta_db":-1.3611069533321343}],"active_band_count":3,"active_band_mae_db":1.2353103398337535}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1089,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":2.75438478934581,"rms_envelope_nrmse":0.2664733418169732,"envelope_correlation":0.9573091302879668,"rms_gain_error_db":-0.8118242411725209,"reference_mean_power":6.7061457476675e-06,"output_mean_power":5.562763082711734e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1089,"frequency_resolution_hz":0.18365472910927455,"power_floor":5.0007302992020595e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.184632530396472e-06,"output_power":1.4980862668096145e-06,"reference_active":true,"delta_db":-1.6384157346293602},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.064669119483919e-06,"output_power":9.591468392500698e-07,"reference_active":true,"delta_db":-0.45329557655148556},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.276430565159006e-07,"output_power":4.207773624203559e-07,"reference_active":true,"delta_db":-0.9828786046223055}],"active_band_count":3,"active_band_mae_db":1.024863305267717}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87200,"uncovered_tail_samples":0,"metrics":{"frame_count":1083,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":2.5161670840778787,"rms_envelope_nrmse":0.21919353708938033,"envelope_correlation":0.9714119222721265,"rms_gain_error_db":-0.8117411739116515,"reference_mean_power":6.743032967300065e-06,"output_mean_power":5.5934680998483606e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1083,"frequency_resolution_hz":0.18467220683287164,"power_floor":4.499376380010927e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.394691563500266e-06,"output_power":1.7299583651857082e-06,"reference_active":true,"delta_db":-1.4121393306951728},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.167260222040222e-07,"output_power":5.553952174472566e-07,"reference_active":true,"delta_db":-0.4549013779518893},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.881500260243624e-08,"output_power":3.581605266968423e-08,"reference_active":true,"delta_db":-1.3447559676791518}],"active_band_count":3,"active_band_mae_db":1.0705988921087382}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32865-0017","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.4557502269744873,"predicted_mos":2.4557502269744873,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":131040,"input_sample_rate":24000,"input_seconds":5.46,"input_float32_sha256":"000d70b5d30d56971f9c04678fab7e0007159420c6d3b1c6764ab26dcbbaa101","internal_samples":87360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.762748914999975},"started_at":"2026-10-13T07:41:00.210483+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027399150021665264,"native_input_samples":130920,"native_input_sha256":"f0c367061358b083702e5b76462f1d9af60724ae244f54f83d9ff7b0a394a036","output_sha256":"6e44013db2b862adb15792bc9dd0aefe2d5a391216c95df68320303369ee44ec","output_samples":131040,"length_delta_samples":120,"analysis_s":2.2649764830021013,"synthesis_s":11.905073744997935,"total_s":14.170050228000036,"rtf":2.597626072960593,"output_peak":0.13879702985286713,"output_rms":0.01677560382152265,"finished_at":"2026-10-13T07:41:14.391020+00:00","analysis_stages":{"external_f0_s":1.6219297909992747},"evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"6e44013db2b862adb15792bc9dd0aefe2d5a391216c95df68320303369ee44ec","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":87280,"output_samples":87360,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-35.51436193667699,"output_peak":0.1378228911237419},"common_prefix":{"samples":87280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990842490842491,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.324495288460956,"spectral_convergence":0.5593474639076711,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.875215842147333,"spectral_convergence":0.36218922914253215,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.824103841767845,"spectral_convergence":0.37169889182655846,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00011662309673752687,"delta_db":-0.25010174535425694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":6.573788148698795e-05,"delta_db":-0.4101589949254296,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":3.6074244910153844e-05,"delta_db":-0.7925652218243101,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":4.14627662468773e-05,"delta_db":-0.5882433142324481,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":6.839581172031956e-06,"delta_db":-1.1484216451107292,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":1.4427317179995744e-05,"delta_db":-1.6698628550060302,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.809892296075534,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5163308578732211,"pitch":null,"output_rms_dbfs":-35.51038915571412,"output_peak":0.1378228911237419,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.007938324125377},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87280,"output_samples":87280,"intelligibility":{"estoi":{"value":0.8261191362261275,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.01779496282666337,"rms_dbfs":-34.9940582978409,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.990945259100272,"mean":-2.6739244609661434e-05,"zero_energy":false},"output":{"rms":0.01676797208172693,"rms_dbfs":-35.51038915571412,"peak":0.1378228911237419,"peak_dbfs":-17.21357288179214,"crest_factor_db":18.296816273921984,"mean":4.714958800989711e-05,"zero_energy":false},"rms_difference_db":-0.5163308578732213},"full_scale":{"reference":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":2.8332648486454396e-07,"output_rms_dbfs":-65.4771282649296,"output_zero_energy":false,"output_peak":0.017428722410427,"output_energy_fraction":0.00021058924399206672,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":152,"both_estimated_voiced_frames":137,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.8782051282051282,"absolute_difference_cents_median":5.9932544630710645,"absolute_difference_cents_p95":25.219919854201347,"difference_over_600_cents_frames":4,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":3.5455690519792262,"rms_envelope_nrmse":0.18315025287579184,"envelope_correlation":0.9695613352152733,"rms_gain_error_db":-0.5163307481956294,"reference_mean_power":0.0003169510345489487,"output_mean_power":0.00028142268280539526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8938855835285014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.978895728357633e-05,"output_power":5.9848546012447915e-05,"reference_active":true,"delta_db":0.004326245125419716},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1930505918996815e-05,"output_power":1.7027345632233695e-05,"reference_active":true,"delta_db":-1.099016989012509},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0863861439295835e-06,"output_power":2.9058374565781132e-06,"reference_active":true,"delta_db":-0.2617894300377106}],"active_band_count":3,"active_band_mae_db":0.4550442213918798}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":3.206398035210298,"rms_envelope_nrmse":0.1252395585880536,"envelope_correlation":0.9867447534261545,"rms_gain_error_db":-0.5163217735399318,"reference_mean_power":0.00031870366172679027,"output_mean_power":0.00028297943553128995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":1.711395707762808e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.23137186055323e-05,"output_power":5.2845787685093223e-05,"reference_active":true,"delta_db":0.04394783605562495},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1932036565608145e-05,"output_power":8.876278347818494e-06,"reference_active":true,"delta_db":-1.2848366293511537},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.569030812932547e-07,"output_power":2.0610783562921503e-07,"reference_active":true,"delta_db":-0.9567481049710996}],"active_band_count":3,"active_band_mae_db":0.7618441901259594}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":2.6088013036405955,"rms_envelope_nrmse":0.20476688120550326,"envelope_correlation":0.9684933872011163,"rms_gain_error_db":-0.615647997090414,"reference_mean_power":0.000101477720422738,"output_mean_power":8.806551984181118e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.361033224981284e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.73677738976729e-05,"output_power":2.3239965810709442e-05,"reference_active":true,"delta_db":-0.7100398840311467},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1422451661963211e-05,"output_power":8.770478307701759e-06,"reference_active":true,"delta_db":-1.1473605012483858},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1019925401091896e-06,"output_power":1.1245030070380964e-06,"reference_active":true,"delta_db":0.08781966533840077}],"active_band_count":3,"active_band_mae_db":0.6484066835393111}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":2.1581586496672838,"rms_envelope_nrmse":0.13875722840249902,"envelope_correlation":0.9867470337117391,"rms_gain_error_db":-0.61564756253395,"reference_mean_power":0.00010203935243697765,"output_mean_power":8.855293042065305e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":7.765913024651851e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5071474654741866e-05,"output_power":2.1329119858587023e-05,"reference_active":true,"delta_db":-0.7020694430178959},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067872704197783e-06,"output_power":4.479702410526057e-06,"reference_active":true,"delta_db":-1.3178729679822851},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0845180908068795e-07,"output_power":1.1520434086635644e-07,"reference_active":true,"delta_db":0.2623204262527911}],"active_band_count":3,"active_band_mae_db":0.760754279084324}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":2.170643776706316,"rms_envelope_nrmse":0.28516203495094006,"envelope_correlation":0.9436975249762206,"rms_gain_error_db":-0.6527033360888496,"reference_mean_power":4.068228265632885e-05,"output_mean_power":3.5005394735764265e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.8091442271155458e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.649297963098176e-06,"output_power":7.7808807358305e-06,"reference_active":true,"delta_db":-0.4595209998270696},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5842183345769264e-06,"output_power":2.2436079141205287e-06,"reference_active":true,"delta_db":-0.6138224012651998},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3501889520111636e-06,"output_power":1.259474928051154e-06,"reference_active":true,"delta_db":-0.3020502346670071}],"active_band_count":3,"active_band_mae_db":0.4584645452530922}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":1.4643615018441605,"rms_envelope_nrmse":0.17865925309665062,"envelope_correlation":0.9789531286933905,"rms_gain_error_db":-0.6527099968300488,"reference_mean_power":4.090728037801193e-05,"output_mean_power":3.5198941836366124e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":2.6173037083412677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.994446901443647e-06,"output_power":7.2493052970869795e-06,"reference_active":true,"delta_db":-0.4249203230279694},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1581434874555988e-06,"output_power":1.034988709145405e-06,"reference_active":true,"delta_db":-0.4882675733330974},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.731926648355597e-08,"output_power":8.219494286849487e-08,"reference_active":true,"delta_db":-0.26264980875733923}],"active_band_count":3,"active_band_mae_db":0.3919459017061353}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":2.0975591723270077,"rms_envelope_nrmse":0.3913823419068427,"envelope_correlation":0.8906207447689077,"rms_gain_error_db":-0.7887859711820416,"reference_mean_power":6.700122358829964e-06,"output_mean_power":5.5873276110846586e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":5.0013650649342945e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1848288835729857e-06,"output_power":1.9619994698985174e-06,"reference_active":true,"delta_db":-0.46718542865154045},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069865846115692e-06,"output_power":9.443244626378259e-07,"reference_active":true,"delta_db":-0.5420808334133632},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241288004170751e-07,"output_power":4.4583936946441854e-07,"reference_active":true,"delta_db":-0.7025960861793396}],"active_band_count":3,"active_band_mae_db":0.570620782748081}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":1.377264174012467,"rms_envelope_nrmse":0.23702974710092783,"envelope_correlation":0.9598256554175524,"rms_gain_error_db":-0.7888380695811491,"reference_mean_power":6.736890999373383e-06,"output_mean_power":5.617922111373877e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":4.500371991965456e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.395916806498351e-06,"output_power":2.1386236110777982e-06,"reference_active":true,"delta_db":-0.49337376695086177},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.187368503320663e-07,"output_power":5.175341320700883e-07,"reference_active":true,"delta_db":-0.7756698484515917},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.878047962583858e-08,"output_power":3.6473900792437195e-08,"reference_active":true,"delta_db":-1.2626385362001553}],"active_band_count":3,"active_band_mae_db":0.8438940505342029}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32865-0017","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.1410791873931885,"predicted_mos":2.1410791873931885,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":131160,"input_sample_rate":24000,"input_seconds":5.465,"input_float32_sha256":"166f9ffc5813fcfc5cf09f747f9808ce0ebefa5b004f873b1efa980af6968357","internal_samples":87440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3809849379995285},"started_at":"2026-10-13T05:11:03.222607+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004140342000027886,"native_input_samples":130920,"native_input_sha256":"f0c367061358b083702e5b76462f1d9af60724ae244f54f83d9ff7b0a394a036","output_sha256":"430019c417304df8613c68a5549f362261ba77c4788bc9740e35530ac9df22f2","output_samples":131160,"length_delta_samples":240,"analysis_s":0.8172154060011962,"synthesis_s":0.14439596900047036,"total_s":0.9616113750016666,"rtf":0.17628072868958142,"output_peak":0.16967004537582397,"output_rms":0.017492446198307702,"finished_at":"2026-10-13T05:11:04.197648+00:00","analysis_stages":{"external_f0_s":0.20916071699866734},"evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"430019c417304df8613c68a5549f362261ba77c4788bc9740e35530ac9df22f2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":87280,"output_samples":87440,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-35.14756598542819,"output_peak":0.17560481989479332},"common_prefix":{"samples":87280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981701738334858,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.235575751641262,"spectral_convergence":0.285017011691579,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.949906221230561,"spectral_convergence":0.2704585642886531,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4223263290259665,"spectral_convergence":0.30860719923563396,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00012312301558112894,"delta_db":-0.014554998820525064,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":6.999328734431676e-05,"delta_db":-0.1377521228605003,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":4.3861570074032216e-05,"delta_db":0.05630402090591676,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":4.2608112798389426e-05,"delta_db":-0.4699030589214797,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":8.558138755060737e-06,"delta_db":-0.17492349107442773,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":1.8079552384486846e-05,"delta_db":-0.689841912628748,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25721326753526624,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1455540329935505,"pitch":null,"output_rms_dbfs":-35.13961233083445,"output_peak":0.17560481989479332,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.202602767299264},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87280,"output_samples":87280,"intelligibility":{"estoi":{"value":0.901127170399241,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.01779496282666337,"rms_dbfs":-34.9940582978409,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.990945259100272,"mean":-2.6739244609661434e-05,"zero_energy":false},"output":{"rms":0.017499247896335862,"rms_dbfs":-35.13961233083445,"peak":0.17560481989479332,"peak_dbfs":-15.109271360285723,"crest_factor_db":20.03034097054873,"mean":4.8420662266245825e-05,"zero_energy":false},"rms_difference_db":-0.14555403299355163},"full_scale":{"reference":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":3.670704282867343e-07,"output_rms_dbfs":-64.35250601464949,"output_zero_energy":false,"output_peak":0.0037435749602548543,"output_energy_fraction":0.0002505074972273628,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":151,"both_estimated_voiced_frames":134,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.8589743589743589,"absolute_difference_cents_median":1.7357717932604915,"absolute_difference_cents_p95":8.438078800156385,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":1.233777560482364,"rms_envelope_nrmse":0.07948603791489724,"envelope_correlation":0.9940185363766013,"rms_gain_error_db":-0.14555345609201245,"reference_mean_power":0.0003169510345489487,"output_mean_power":0.00030650448094066043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8938855835285014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.978895728357633e-05,"output_power":5.867519502596359e-05,"reference_active":true,"delta_db":-0.08166437621136147},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1930505918996815e-05,"output_power":2.0870476089561378e-05,"reference_active":true,"delta_db":-0.21516294471659245},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0863861439295835e-06,"output_power":2.733227223025464e-06,"reference_active":true,"delta_db":-0.5277452291740801}],"active_band_count":3,"active_band_mae_db":0.274857516700678}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":0.8193809614889377,"rms_envelope_nrmse":0.044143451079108796,"envelope_correlation":0.9982177063163865,"rms_gain_error_db":-0.14554381025235238,"reference_mean_power":0.00031870366172679027,"output_mean_power":0.00030820002689273707},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":1.711395707762808e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.23137186055323e-05,"output_power":5.211500791490913e-05,"reference_active":true,"delta_db":-0.01652783931493413},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1932036565608145e-05,"output_power":1.110787844077139e-05,"reference_active":true,"delta_db":-0.31083457186940383},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.569030812932547e-07,"output_power":1.9617616868442604e-07,"reference_active":true,"delta_db":-1.1712306470105998}],"active_band_count":3,"active_band_mae_db":0.4995310193983126}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":1.0904789522367904,"rms_envelope_nrmse":0.0923707504216659,"envelope_correlation":0.9932627086796186,"rms_gain_error_db":-0.08512917109743348,"reference_mean_power":0.000101477720422738,"output_mean_power":9.95079514586073e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.361033224981284e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.73677738976729e-05,"output_power":2.7012724742997483e-05,"reference_active":true,"delta_db":-0.056710800227726746},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1422451661963211e-05,"output_power":1.0821544315736298e-05,"reference_active":true,"delta_db":-0.23470086583903235},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1019925401091896e-06,"output_power":1.1781804618063772e-06,"reference_active":true,"delta_db":0.2903316180887373}],"active_band_count":3,"active_band_mae_db":0.19391442805183212}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":0.5839783338652742,"rms_envelope_nrmse":0.043519033975157474,"envelope_correlation":0.9984571906723025,"rms_gain_error_db":-0.08512868463040169,"reference_mean_power":0.00010203935243697765,"output_mean_power":0.00010005869292509847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":7.765913024651851e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5071474654741866e-05,"output_power":2.4932282928057364e-05,"reference_active":true,"delta_db":-0.02417832494307787},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067872704197783e-06,"output_power":5.660551253308247e-06,"reference_active":true,"delta_db":-0.3017773424041639},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0845180908068795e-07,"output_power":1.0190586092675627e-07,"reference_active":true,"delta_db":-0.2703763849402266}],"active_band_count":3,"active_band_mae_db":0.19877735076248945}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":1.2652113467702033,"rms_envelope_nrmse":0.1358134653887748,"envelope_correlation":0.9880329976939658,"rms_gain_error_db":-0.36317278574967504,"reference_mean_power":4.068228265632885e-05,"output_mean_power":3.741864306281151e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.8091442271155458e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.649297963098176e-06,"output_power":7.5428334699942085e-06,"reference_active":true,"delta_db":-0.5944633904166103},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5842183345769264e-06,"output_power":2.453743027321339e-06,"reference_active":true,"delta_db":-0.22500124903596125},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3501889520111636e-06,"output_power":1.073183174320929e-06,"reference_active":true,"delta_db":-0.9972069497333538}],"active_band_count":3,"active_band_mae_db":0.6055571963953085}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":0.8594141372623387,"rms_envelope_nrmse":0.08081452428345855,"envelope_correlation":0.9963034648558502,"rms_gain_error_db":-0.3631696398132276,"reference_mean_power":4.090728037801193e-05,"output_mean_power":3.762561813134047e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":2.6173037083412677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.994446901443647e-06,"output_power":7.013499688875174e-06,"reference_active":true,"delta_db":-0.5685364024852608},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1581434874555988e-06,"output_power":1.0764658468722262e-06,"reference_active":true,"delta_db":-0.3176211388690051},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.731926648355597e-08,"output_power":6.92326488790881e-08,"reference_active":true,"delta_db":-1.0079913053245502}],"active_band_count":3,"active_band_mae_db":0.6313829488929387}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":1.353647000643905,"rms_envelope_nrmse":0.1663114001575062,"envelope_correlation":0.9812519679677207,"rms_gain_error_db":-0.06634647897033183,"reference_mean_power":6.700122358829964e-06,"output_mean_power":6.59854352993589e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":5.0013650649342945e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1848288835729857e-06,"output_power":2.223171911381105e-06,"reference_active":true,"delta_db":0.07555618149911332},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069865846115692e-06,"output_power":1.1651133952684646e-06,"reference_active":true,"delta_db":0.37038871824268155},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241288004170751e-07,"output_power":3.7770430603281597e-07,"reference_active":true,"delta_db":-1.4228608841138266}],"active_band_count":3,"active_band_mae_db":0.6229352612852072}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":1.025837238828255,"rms_envelope_nrmse":0.09475609098878292,"envelope_correlation":0.9935102245713049,"rms_gain_error_db":-0.06631813450506967,"reference_mean_power":6.736890999373383e-06,"output_mean_power":6.6347980328213755e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":4.500371991965456e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.395916806498351e-06,"output_power":2.4568765158104028e-06,"reference_active":true,"delta_db":0.10911595152522822},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.187368503320663e-07,"output_power":6.087887101287147e-07,"reference_active":true,"delta_db":-0.07039392293868983},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.878047962583858e-08,"output_power":2.8698357625225146e-08,"reference_active":true,"delta_db":-2.303890228662252}],"active_band_count":3,"active_band_mae_db":0.82780003437539}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32865-0017","method":"magphase","status":"measured","mos":{"status":"measured","value":2.41167950630188,"predicted_mos":2.41167950630188,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":262545,"input_sample_rate":48000,"input_seconds":5.4696875,"input_float32_sha256":"98a9cc1e4553f668dc75648b626463bd57c1ddf82e1386aed93e27f4405f159a","internal_samples":87515,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.328691680999327},"started_at":"2026-10-13T05:16:45.445945+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008728392000193708,"native_input_samples":261840,"native_input_sha256":"db1665e768090fb5155a20322a01813c645896901a3cbd866facf543c85c267d","output_sha256":"7688bbdad95a1e0bf8e1e860e4d235136504687b1e0a8fe8f58f3512968491d2","output_samples":262545,"length_delta_samples":705,"analysis_s":3.2801157679987227,"synthesis_s":0.9624936800009891,"total_s":4.242609447999712,"rtf":0.7777469198899564,"output_peak":0.13794524363178115,"output_rms":0.017055714086884798,"finished_at":"2026-10-13T05:16:49.708179+00:00","evaluation_input":{"reference":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"7688bbdad95a1e0bf8e1e860e4d235136504687b1e0a8fe8f58f3512968491d2","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":87280,"output_samples":87515,"length_delta_samples":235,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-35.366591096894936,"output_peak":0.13786100424194977},"common_prefix":{"samples":87280,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973147460435354,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":87280,"output_samples":87280,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.28919874111052,"spectral_convergence":0.5408116798829953,"spectral_convergence_status":"ok","amplitude_floor":4.3919411473735467e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.863319578719059,"spectral_convergence":0.356127114835461,"spectral_convergence_status":"ok","amplitude_floor":3.1709236310047874e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.872894328838409,"spectral_convergence":0.32634668612900647,"spectral_convergence_status":"ok","amplitude_floor":2.989229099554426e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0001235363438035874,"output_power":0.00011076956410358651,"delta_db":-0.47374297065280446,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":7.224896060742283e-05,"output_power":6.893955852467204e-05,"delta_db":-0.2036310565404289,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.329659782840476e-05,"output_power":4.2346740429385905e-05,"delta_db":-0.09633784735993516,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.747692740945149e-05,"output_power":4.0577778761428954e-05,"delta_db":-0.6819433476528922,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.909876260278539e-06,"output_power":8.198011504770696e-06,"delta_db":-0.3616314917528364,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.1191996093186054e-05,"output_power":2.0580671715825272e-05,"delta_db":-0.12712319983042594,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.32406831896488714,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.360862542729397,"pitch":null,"output_rms_dbfs":-35.3549208405703,"output_peak":0.13786100424194977,"reference_rms_dbfs":-34.9940582978409,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.00847088288933},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":87280,"output_samples":87280,"intelligibility":{"estoi":{"value":0.8641806149830638,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.01779496282666337,"rms_dbfs":-34.9940582978409,"peak":0.2510986328125,"peak_dbfs":-12.003113038740631,"crest_factor_db":22.990945259100272,"mean":-2.6739244609661434e-05,"zero_energy":false},"output":{"rms":0.017070803292161433,"rms_dbfs":-35.3549208405703,"peak":0.13786100424194977,"peak_dbfs":-17.211171244654896,"crest_factor_db":18.143749595915402,"mean":0.00031818287705312926,"zero_energy":false},"rms_difference_db":-0.3608625427294001},"full_scale":{"reference":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":87280,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":545,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":114,"quiet_samples":18240,"reference_mean_power":4.392627344809865e-07,"output_mean_power":3.901681865462414e-07,"output_rms_dbfs":-64.08748144912495,"output_zero_energy":false,"output_peak":0.005447753337396142,"output_energy_fraction":0.0002798040752533522,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":538,"reference_estimated_voiced_frames":156,"output_estimated_voiced_frames":152,"both_estimated_voiced_frames":143,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9166666666666666,"absolute_difference_cents_median":3.8852620300348892,"absolute_difference_cents_p95":21.168763758133473,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":87280,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":5.579882181126775e-05,"log_rms_mae_db":2.1942935579834226,"rms_envelope_nrmse":0.20250358986539388,"envelope_correlation":0.9601074950300096,"rms_gain_error_db":-0.3608612637435816,"reference_mean_power":0.0003169510345489487,"output_mean_power":0.00029167959433675205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":1.8938855835285014e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.978895728357633e-05,"output_power":5.736427593157704e-05,"reference_active":true,"delta_db":-0.17979463246900906},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1930505918996815e-05,"output_power":2.008475437643409e-05,"reference_active":true,"delta_db":-0.38182125655466015},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0863861439295835e-06,"output_power":3.64653570731766e-06,"reference_active":true,"delta_db":0.7243021000809031}],"active_band_count":3,"active_band_mae_db":0.42863932970152413}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":5.258028604719465e-05,"log_rms_mae_db":1.4904761278095835,"rms_envelope_nrmse":0.12330523209139949,"envelope_correlation":0.9848184053682492,"rms_gain_error_db":-0.3608551314960129,"reference_mean_power":0.00031870366172679027,"output_mean_power":0.00029329289351366964},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":1.711395707762808e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.23137186055323e-05,"output_power":5.049849363497735e-05,"reference_active":true,"delta_db":-0.15337168641092783},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1932036565608145e-05,"output_power":1.1136880462949746e-05,"reference_active":true,"delta_db":-0.29951017371590016},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.569030812932547e-07,"output_power":2.6163064764014537e-07,"reference_active":true,"delta_db":0.07919303041338308}],"active_band_count":3,"active_band_mae_db":0.177358296846737}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":4.2439491809696055e-05,"log_rms_mae_db":2.0925808593530038,"rms_envelope_nrmse":0.24036779822956603,"envelope_correlation":0.9539307409870631,"rms_gain_error_db":-0.17158502511568194,"reference_mean_power":0.000101477720422738,"output_mean_power":9.754661429266031e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":8.361033224981284e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.73677738976729e-05,"output_power":2.777190970082776e-05,"reference_active":true,"delta_db":0.06366271283935672},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1422451661963211e-05,"output_power":1.0548791135305477e-05,"reference_active":true,"delta_db":-0.3455663542533376},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1019925401091896e-06,"output_power":9.730631814997552e-07,"reference_active":true,"delta_db":-0.5403761443697561}],"active_band_count":3,"active_band_mae_db":0.31653507048748347}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.7895186903062846e-05,"log_rms_mae_db":1.4330828441488113,"rms_envelope_nrmse":0.14586637127824317,"envelope_correlation":0.9819819331026576,"rms_gain_error_db":-0.17158613108756174,"reference_mean_power":0.00010203935243697765,"output_mean_power":9.808646448305756e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":7.765913024651851e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5071474654741866e-05,"output_power":2.5469360597265196e-05,"reference_active":true,"delta_db":0.06838163177127383},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.067872704197783e-06,"output_power":5.54999633067153e-06,"reference_active":true,"delta_db":-0.38743765303559285},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0845180908068795e-07,"output_power":8.650203236146026e-08,"reference_active":true,"delta_db":-0.9821048954660461}],"active_band_count":3,"active_band_mae_db":0.47930806009097093}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":3.774086107813168e-05,"log_rms_mae_db":1.912875362908112,"rms_envelope_nrmse":0.3167209957065948,"envelope_correlation":0.9294753413890396,"rms_gain_error_db":-0.6320577933075118,"reference_mean_power":4.068228265632885e-05,"output_mean_power":3.517219996418941e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":2.8091442271155458e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.649297963098176e-06,"output_power":7.205798970984449e-06,"reference_active":true,"delta_db":-0.7929871663903838},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5842183345769264e-06,"output_power":2.1740769869474163e-06,"reference_active":true,"delta_db":-0.7505428443478329},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3501889520111636e-06,"output_power":1.3926250757185677e-06,"reference_active":true,"delta_db":0.13439660820209598}],"active_band_count":3,"active_band_mae_db":0.5593088729801042}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":3.332899304538304e-05,"log_rms_mae_db":1.321467323996988,"rms_envelope_nrmse":0.19547591171301718,"envelope_correlation":0.9737607435441774,"rms_gain_error_db":-0.632057219870287,"reference_mean_power":4.090728037801193e-05,"output_mean_power":3.536672825314239e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":2.6173037083412677e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.994446901443647e-06,"output_power":6.73244890969429e-06,"reference_active":true,"delta_db":-0.746153559283075},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1581434874555988e-06,"output_power":9.712631458305194e-07,"reference_active":true,"delta_db":-0.7642545943737812},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.731926648355597e-08,"output_power":8.062276556415213e-08,"reference_active":true,"delta_db":-0.34652387338295076}],"active_band_count":3,"active_band_mae_db":0.6189773423466023}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1090,"rms_floor":1.8891386922351935e-05,"log_rms_mae_db":1.7915751259050492,"rms_envelope_nrmse":0.4857028427253646,"envelope_correlation":0.8607053262158221,"rms_gain_error_db":0.7670008803610627,"reference_mean_power":6.700122358829964e-06,"output_mean_power":7.994343811716342e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1090,"frequency_resolution_hz":0.18348623853211007,"power_floor":5.0013650649342945e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1848288835729857e-06,"output_power":2.5305294494210628e-06,"reference_active":true,"delta_db":0.6379396727271156},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069865846115692e-06,"output_power":1.4147985561400159e-06,"reference_active":true,"delta_db":1.2136528439601177},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.241288004170751e-07,"output_power":5.616639570439389e-07,"reference_active":true,"delta_db":0.30038530877224223}],"active_band_count":3,"active_band_mae_db":0.7173259418198251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":87280,"uncovered_tail_samples":0,"metrics":{"frame_count":1084,"rms_floor":1.232461634599244e-05,"log_rms_mae_db":1.1823557951323431,"rms_envelope_nrmse":0.2863952130386065,"envelope_correlation":0.950776936139595,"rms_gain_error_db":0.7670595674287208,"reference_mean_power":6.736890999373383e-06,"output_mean_power":8.038323447262402e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1084,"frequency_resolution_hz":0.18450184501845018,"power_floor":4.500371991965456e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.395916806498351e-06,"output_power":2.836162584596698e-06,"reference_active":true,"delta_db":0.7325938944038317},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.187368503320663e-07,"output_power":7.760934174378143e-07,"reference_active":true,"delta_db":0.9840801769056531},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.878047962583858e-08,"output_power":4.728322506336257e-08,"reference_active":true,"delta_db":-0.13538975167840892}],"active_band_count":3,"active_band_mae_db":0.6173546076626312}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"hnm","status":"measured","mos":{"status":"measured","value":2.412008762359619,"predicted_mos":2.412008762359619,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":82160,"input_sample_rate":16000,"input_seconds":5.135,"input_float32_sha256":"e835e6249f1b0271cfeb7b0fcbc529db7773d420a82d0b46dfdb1437c469a69d","internal_samples":82160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.9316426499990484},"started_at":"2026-10-13T04:39:42.772279+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5438000446010847e-05,"native_input_samples":82160,"native_input_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","output_sha256":"6a1599de024916b08fd9095eb1f140fd0c08a2d0a080e295537f190ed573bd99","output_samples":82160,"length_delta_samples":0,"analysis_s":0.779703411999435,"synthesis_s":0.33441032800055837,"total_s":1.1141137399999934,"rtf":0.21696470107107954,"output_peak":0.7434766822581035,"output_rms":0.07800331901629545,"finished_at":"2026-10-13T04:39:43.893318+00:00","evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"6a1599de024916b08fd9095eb1f140fd0c08a2d0a080e295537f190ed573bd99","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.440793654397106,"spectral_convergence":0.22324182396197845,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.369342730053801,"spectral_convergence":0.2163831686602846,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.663173997774121,"spectral_convergence":0.21968048348385344,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":9.266209262831308e-05,"delta_db":0.4431338062449061,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.001592198333319752,"delta_db":-0.03444889799853097,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0012967669111394848,"delta_db":-0.10293250050628786,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.00258239217272184,"delta_db":-0.16333531219145087,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.00031588422640920036,"delta_db":0.09972855584672687,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00020461404133937077,"delta_db":-0.07142308344292861,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1525003593718052,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09180329203235867,"pitch":null,"output_rms_dbfs":-22.157738356499156,"output_peak":0.7434766822581035,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.491103460741676},"common_prefix":{"samples":82160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.440793654397106,"spectral_convergence":0.22324182396197845,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.369342730053801,"spectral_convergence":0.2163831686602846,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.663173997774121,"spectral_convergence":0.21968048348385344,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":9.266209262831308e-05,"delta_db":0.4431338062449061,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.001592198333319752,"delta_db":-0.03444889799853097,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0012967669111394848,"delta_db":-0.10293250050628786,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.00258239217272184,"delta_db":-0.16333531219145087,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.00031588422640920036,"delta_db":0.09972855584672687,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00020461404133937077,"delta_db":-0.07142308344292861,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1525003593718052,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09180329203235867,"pitch":null,"output_rms_dbfs":-22.157738356499156,"output_peak":0.7434766822581035,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.491103460741676},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82160,"output_samples":82160,"intelligibility":{"estoi":{"value":0.8790942771540567,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07883212739732415,"rms_dbfs":-22.065935064466796,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.49939156382414,"mean":5.828649415013389e-06,"zero_energy":false},"output":{"rms":0.07800331901629545,"rms_dbfs":-22.157738356499156,"peak":0.7434766822581035,"peak_dbfs":-2.574652955091727,"crest_factor_db":19.583085401407427,"mean":1.128141666535818e-06,"zero_energy":false},"rms_difference_db":-0.0918032920323597},"full_scale":{"reference":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":1.564205619136803e-05,"output_rms_dbfs":-48.05706158317786,"output_zero_energy":false,"output_peak":0.13201819580360694,"output_energy_fraction":0.0005557125542848365,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":219,"both_estimated_voiced_frames":214,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.8770491803278688,"absolute_difference_cents_median":1.0894770663535909,"absolute_difference_cents_p95":8.754399327305455,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.0003747956436237944,"log_rms_mae_db":0.8400189585346443,"rms_envelope_nrmse":0.06121932549831111,"envelope_correlation":0.9969107056399371,"rms_gain_error_db":-0.091803093168322,"reference_mean_power":0.006220557891926456,"output_mean_power":0.006090445017801217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.814448626587233e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017308539978626828,"output_power":0.0016442439715863605,"reference_active":true,"delta_db":-0.22294177277849458},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0006400481816254359,"output_power":0.0005976530849688276,"reference_active":true,"delta_db":-0.29763502498016364},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4706618104713675e-05,"output_power":4.6377009884052236e-05,"reference_active":true,"delta_db":0.15930926168041495}],"active_band_count":3,"active_band_mae_db":0.22662868647969103}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.00031163820648484283,"log_rms_mae_db":0.4565886307367638,"rms_envelope_nrmse":0.03289436897488693,"envelope_correlation":0.9991442139940991,"rms_gain_error_db":-0.0917938293744463,"reference_mean_power":0.006257088451954998,"output_mean_power":0.006126224550630071},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":6.546091034740867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001677353066756132,"output_power":0.0015906620894116361,"reference_active":true,"delta_db":-0.23046556493529563},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035132010859926165,"output_power":0.0003268778501122669,"reference_active":true,"delta_db":-0.31317515401355944},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3827329510863972e-06,"output_power":2.782197603662809e-06,"reference_active":true,"delta_db":0.6731260144780289}],"active_band_count":3,"active_band_mae_db":0.4055889111422946}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00015152624573583172,"log_rms_mae_db":1.252678431605545,"rms_envelope_nrmse":0.07887497324384184,"envelope_correlation":0.9952414968494391,"rms_gain_error_db":-0.04775150320931298,"reference_mean_power":0.001672933915423699,"output_mean_power":0.0016546404436528364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":1.623496524404657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006256720367801862,"output_power":0.0006117479296203935,"reference_active":true,"delta_db":-0.09774237232362651},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001572550903934534,"output_power":0.0001264077556914198,"reference_active":true,"delta_db":-0.9483099187686915},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4941700779965039e-05,"output_power":1.1789170497373222e-05,"reference_active":true,"delta_db":-1.0291678645139548}],"active_band_count":3,"active_band_mae_db":0.6917400518687575}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0001385613366624899,"log_rms_mae_db":0.8104644872215726,"rms_envelope_nrmse":0.042821559057523884,"envelope_correlation":0.9985451188789449,"rms_gain_error_db":-0.047751852642863024,"reference_mean_power":0.0016827742497549935,"output_mean_power":0.0016643730403690983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":1.5301470087365332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005852793815004823,"output_power":0.000569628403330956,"reference_active":true,"delta_db":-0.11771588495514672},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151264176170612e-05,"output_power":6.755097679958126e-05,"reference_active":true,"delta_db":-0.8159333513231841},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3333886770336671e-06,"output_power":8.7337727985828e-07,"reference_active":true,"delta_db":-1.837548728886107}],"active_band_count":3,"active_band_mae_db":0.9237326550548127}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00030489717691300735,"log_rms_mae_db":1.2532136110329624,"rms_envelope_nrmse":0.069516839655024,"envelope_correlation":0.9972549466563926,"rms_gain_error_db":-0.1819728998923448,"reference_mean_power":0.002589850567098419,"output_mean_power":0.0024835757819763236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":4.739471399174343e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014413951986732994,"output_power":0.0013590074468745811,"reference_active":true,"delta_db":-0.2556123456168151},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004245147064825656,"output_power":0.000380043183110379,"reference_active":true,"delta_db":-0.48059793189080124},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.389197181736509e-05,"output_power":2.6919188866262e-05,"reference_active":true,"delta_db":0.518099757236521}],"active_band_count":3,"active_band_mae_db":0.41810334491471246}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0002474836728453423,"log_rms_mae_db":0.7790540781022303,"rms_envelope_nrmse":0.04398973250400079,"envelope_correlation":0.9990391896231052,"rms_gain_error_db":-0.18194996719553763,"reference_mean_power":0.002605032676611575,"output_mean_power":0.002498148083367492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":4.638343817264176e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014427582782272214,"output_power":0.0013535760725152236,"reference_active":true,"delta_db":-0.27710906236927435},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020918188683306623,"output_power":0.00018769939091072354,"reference_active":true,"delta_db":-0.47061212714183276},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0082267328987973e-06,"output_power":2.0342117234402583e-06,"reference_active":true,"delta_db":0.055834089070006684}],"active_band_count":3,"active_band_mae_db":0.26785175952703794}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":1.4531988083315537,"rms_envelope_nrmse":0.17707195048502253,"envelope_correlation":0.9792859566440463,"rms_gain_error_db":0.27928974471075907,"reference_mean_power":0.00014375800668184295,"output_mean_power":0.0001533066581451819},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.640530401894145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.05643435973393e-05,"output_power":2.2928548292510245e-05,"reference_active":true,"delta_db":0.47261706879005405},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8444282129891133e-05,"output_power":1.7230345491068835e-05,"reference_active":true,"delta_db":-0.2956777100103565},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6443800103937953e-06,"output_power":3.00009980341038e-06,"reference_active":true,"delta_db":-0.8448795334271085}],"active_band_count":3,"active_band_mae_db":0.5377247707425064}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":0.9112733637339807,"rms_envelope_nrmse":0.09472365336939591,"envelope_correlation":0.994188986357191,"rms_gain_error_db":0.2792728344437974,"reference_mean_power":0.0001445994240238279,"output_mean_power":0.00015420336343114404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":5.765239137842011e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0474191785118637e-05,"output_power":2.2470738556277896e-05,"reference_active":true,"delta_db":0.40410579736063557},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1083544144974644e-05,"output_power":1.0520830538627416e-05,"reference_active":true,"delta_db":-0.22628630003119404},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8424426027278453e-07,"output_power":2.3388554058251102e-07,"reference_active":true,"delta_db":-0.8468833035077205}],"active_band_count":3,"active_band_mae_db":0.49242513363318335}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.7994840145111084,"predicted_mos":2.7994840145111084,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":82160,"input_sample_rate":16000,"input_seconds":5.135,"input_float32_sha256":"06b4b6970f9505b64570d0dd5483530d0996b43b93cc54f24f811b99a51a6afa","internal_samples":82160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4620523720004712},"started_at":"2026-10-13T04:42:46.609779+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.872400000342168e-05,"native_input_samples":82160,"native_input_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","output_sha256":"8f6a13e7c92050fb8c6386fbfbb2c0e991e5a012604e6a02215364778399fe0f","output_samples":82160,"length_delta_samples":0,"analysis_s":2.3061841499993534,"synthesis_s":0.1857250520006346,"total_s":2.491909201999988,"rtf":0.4852792993184008,"output_peak":0.9695785399929581,"output_rms":0.09050736748758137,"finished_at":"2026-10-13T04:42:49.108412+00:00","evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"8f6a13e7c92050fb8c6386fbfbb2c0e991e5a012604e6a02215364778399fe0f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.650076918263706,"spectral_convergence":0.3360817048537929,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.463268279007218,"spectral_convergence":0.2755591478735853,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.762622761147911,"spectral_convergence":0.29766026600908396,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":0.00010267572824035202,"delta_db":0.8887906860970953,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.0022265436265175966,"delta_db":1.421891542701896,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0016852049456074509,"delta_db":1.0349755460184642,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.0035927933743408365,"delta_db":1.2707651324155236,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.0003674527212322616,"delta_db":0.7564638347960986,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00021691317359360287,"delta_db":0.18208186866360235,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9258281017821134,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.199613726796077,"pitch":null,"output_rms_dbfs":-20.86632133767072,"output_peak":0.9695785399929581,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.625322652806278},"common_prefix":{"samples":82160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.650076918263706,"spectral_convergence":0.3360817048537929,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.463268279007218,"spectral_convergence":0.2755591478735853,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.762622761147911,"spectral_convergence":0.29766026600908396,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":0.00010267572824035202,"delta_db":0.8887906860970953,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.0022265436265175966,"delta_db":1.421891542701896,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0016852049456074509,"delta_db":1.0349755460184642,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.0035927933743408365,"delta_db":1.2707651324155236,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.0003674527212322616,"delta_db":0.7564638347960986,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00021691317359360287,"delta_db":0.18208186866360235,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9258281017821134,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.199613726796077,"pitch":null,"output_rms_dbfs":-20.86632133767072,"output_peak":0.9695785399929581,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.625322652806278},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82160,"output_samples":82160,"intelligibility":{"estoi":{"value":0.9073019677691304,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07883212739732415,"rms_dbfs":-22.065935064466796,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.49939156382414,"mean":5.828649415013389e-06,"zero_energy":false},"output":{"rms":0.09050736748758137,"rms_dbfs":-20.86632133767072,"peak":0.9695785399929581,"peak_dbfs":-0.26834010913705264,"crest_factor_db":20.59798122853367,"mean":3.940018617150302e-08,"zero_energy":false},"rms_difference_db":1.199613726796077},"full_scale":{"reference":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":5.174955633860087e-06,"output_rms_dbfs":-52.86093369166183,"output_zero_energy":false,"output_peak":0.06160949017131128,"output_energy_fraction":0.00013655930425308536,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":266,"both_estimated_voiced_frames":234,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9590163934426229,"absolute_difference_cents_median":4.079785124847543,"absolute_difference_cents_p95":19.018462457684752,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.0003747956436237944,"log_rms_mae_db":1.5682592979819132,"rms_envelope_nrmse":0.1768125447684388,"envelope_correlation":0.9932212954777772,"rms_gain_error_db":1.1996110127689261,"reference_mean_power":0.006220557891926456,"output_mean_power":0.008199557910834583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.814448626587233e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017308539978626828,"output_power":0.0022863736257498534,"reference_active":true,"delta_db":1.208867661862291},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0006400481816254359,"output_power":0.0008002699618365816,"reference_active":true,"delta_db":0.9702384785231171},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4706618104713675e-05,"output_power":7.6849363275458e-05,"reference_active":true,"delta_db":2.3526845527416853}],"active_band_count":3,"active_band_mae_db":1.5105968977090312}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.00031163820648484283,"log_rms_mae_db":1.3693977748629658,"rms_envelope_nrmse":0.16348514433917705,"envelope_correlation":0.9963708230182586,"rms_gain_error_db":1.1996013932218303,"reference_mean_power":0.006257088451954998,"output_mean_power":0.0082476919861572},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":6.546091034740867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001677353066756132,"output_power":0.0022265582371347272,"reference_active":true,"delta_db":1.2300957182383274},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035132010859926165,"output_power":0.0004350216955093572,"reference_active":true,"delta_db":0.9280790840811288},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3827329510863972e-06,"output_power":4.559465332494358e-06,"reference_active":true,"delta_db":2.8183854716680754}],"active_band_count":3,"active_band_mae_db":1.6588534246625104}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00015152624573583172,"log_rms_mae_db":1.7436284568014635,"rms_envelope_nrmse":0.20173709422721708,"envelope_correlation":0.9905290342404657,"rms_gain_error_db":1.306799953588529,"reference_mean_power":0.001672933915423699,"output_mean_power":0.0022602619866159543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":1.623496524404657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006256720367801862,"output_power":0.000795926266391441,"reference_active":true,"delta_db":1.0452609165966336},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001572550903934534,"output_power":0.0001832832773278883,"reference_active":true,"delta_db":0.6651812937949714},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4941700779965039e-05,"output_power":1.5914664961334877e-05,"reference_active":true,"delta_db":0.27397465149630257}],"active_band_count":3,"active_band_mae_db":0.6614722872959692}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0001385613366624899,"log_rms_mae_db":1.5101584630512228,"rms_envelope_nrmse":0.19033245581018238,"envelope_correlation":0.9931109713734342,"rms_gain_error_db":1.30679991515948,"reference_mean_power":0.0016827742497549935,"output_mean_power":0.002273557012655988},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":1.5301470087365332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005852793815004823,"output_power":0.0007461710006646481,"reference_active":true,"delta_db":1.0547514167554142},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151264176170612e-05,"output_power":9.076741084121717e-05,"reference_active":true,"delta_db":0.46704978769943667},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3333886770336671e-06,"output_power":1.3597981838686123e-06,"reference_active":true,"delta_db":0.08517693967645884}],"active_band_count":3,"active_band_mae_db":0.5356593813771032}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00030489717691300735,"log_rms_mae_db":1.458703181621052,"rms_envelope_nrmse":0.1927008261102114,"envelope_correlation":0.9954830678170375,"rms_gain_error_db":1.3516184076188267,"reference_mean_power":0.002589850567098419,"output_mean_power":0.0035353836329257387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":4.739471399174343e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014413951986732994,"output_power":0.002037440064440474,"reference_active":true,"delta_db":1.503017708406076},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004245147064825656,"output_power":0.0005860796172838117,"reference_active":true,"delta_db":1.400638775973111},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.389197181736509e-05,"output_power":5.0861443001081944e-05,"reference_active":true,"delta_db":3.2813668369020244}],"active_band_count":3,"active_band_mae_db":2.0616744404270704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0002474836728453423,"log_rms_mae_db":1.1289863045665978,"rms_envelope_nrmse":0.18030221277678735,"envelope_correlation":0.9978508172564624,"rms_gain_error_db":1.3515932647369415,"reference_mean_power":0.002605032676611575,"output_mean_power":0.0035560880177691783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":4.638343817264176e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014427582782272214,"output_power":0.002040195569208405,"reference_active":true,"delta_db":1.5047822517858869},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020918188683306623,"output_power":0.0002869649373482163,"reference_active":true,"delta_db":1.3730475990379343},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0082267328987973e-06,"output_power":3.8679490305011335e-06,"reference_active":true,"delta_db":2.8466799860304617}],"active_band_count":3,"active_band_mae_db":1.9081699456180943}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":1.4591688214761476,"rms_envelope_nrmse":0.20274886945878148,"envelope_correlation":0.9731895832844755,"rms_gain_error_db":0.3488779167465231,"reference_mean_power":0.00014375800668184295,"output_mean_power":0.00015578291783788095},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.640530401894145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.05643435973393e-05,"output_power":2.427243148781243e-05,"reference_active":true,"delta_db":0.7199843228416674},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8444282129891133e-05,"output_power":2.1012196651257366e-05,"reference_active":true,"delta_db":0.5660969996163809},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6443800103937953e-06,"output_power":3.7482529160121435e-06,"reference_active":true,"delta_db":0.12205231685751963}],"active_band_count":3,"active_band_mae_db":0.46937787977185597}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":1.030806568003928,"rms_envelope_nrmse":0.12371526392749504,"envelope_correlation":0.9900991587157686,"rms_gain_error_db":0.3489097538248026,"reference_mean_power":0.0001445994240238279,"output_mean_power":0.00015669586582476786},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":5.765239137842011e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0474191785118637e-05,"output_power":2.373815199606156e-05,"reference_active":true,"delta_db":0.6424013923633858},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1083544144974644e-05,"output_power":1.269559935464611e-05,"reference_active":true,"delta_db":0.5897455319669174},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8424426027278453e-07,"output_power":2.526835187300488e-07,"reference_active":true,"delta_db":-0.5111478775833271}],"active_band_count":3,"active_band_mae_db":0.5810982673045434}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"facodec","status":"measured","mos":{"status":"measured","value":4.519413948059082,"predicted_mos":4.519413948059082,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":82200,"input_sample_rate":16000,"input_seconds":5.1375,"input_float32_sha256":"b24fdf88e89909e40c6abd5989f08460b551948120d0a8de94382912cb135ac6","internal_samples":82200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2588656230000197},"started_at":"2026-10-13T04:55:25.274596+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7521999982127454e-05,"native_input_samples":82160,"native_input_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","output_sha256":"a744e450decbb49d3b5250bd9df2cdb45ab093dc065390bf83a7194c9f9bdc0a","output_samples":82200,"length_delta_samples":40,"analysis_s":1.1883163869997588,"synthesis_s":9.848529652000252,"total_s":11.036846039000011,"rtf":2.1493371059396322,"output_peak":0.6789238452911377,"output_rms":0.07141360164648854,"finished_at":"2026-10-13T04:55:36.341605+00:00","evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"a744e450decbb49d3b5250bd9df2cdb45ab093dc065390bf83a7194c9f9bdc0a","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":82160,"output_samples":82200,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.92438126659549,"output_peak":0.6789238452911377},"common_prefix":{"samples":82160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9995133819951338,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.330131337529843,"spectral_convergence":0.2834765410543337,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.83454378845598,"spectral_convergence":0.282318725466857,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.937857349194258,"spectral_convergence":0.30252337356447556,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":5.06543228600643e-05,"delta_db":-2.1797220907271773,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.0012102623382006376,"delta_db":-1.2256253595710016,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0010662009902842307,"delta_db":-0.9531608882773179,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.002341220613896765,"delta_db":-0.5891338858882765,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.00022891031995448535,"delta_db":-1.298897120919418,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00020513667732442744,"delta_db":-0.060344247599314976,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0511472654970844,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.856332460475101,"pitch":null,"output_rms_dbfs":-22.9222675249419,"output_peak":0.6789238452911377,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.70084415839336},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82160,"output_samples":82160,"intelligibility":{"estoi":{"value":0.8733586149171215,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07883212739732415,"rms_dbfs":-22.065935064466796,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.49939156382414,"mean":5.828649415013389e-06,"zero_energy":false},"output":{"rms":0.07143098251123675,"rms_dbfs":-22.9222675249419,"peak":0.6789238452911377,"peak_dbfs":-3.363578753703289,"crest_factor_db":19.558688771238607,"mean":0.00010815622316420203,"zero_energy":false},"rms_difference_db":-0.8563324604751053},"full_scale":{"reference":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":2.911741462568597e-06,"output_rms_dbfs":-55.35847189236914,"output_zero_energy":false,"output_peak":0.011680211871862411,"output_energy_fraction":0.0001233565157606051,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":253,"both_estimated_voiced_frames":230,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.9426229508196722,"absolute_difference_cents_median":3.6465493550114045,"absolute_difference_cents_p95":23.72755787888239,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.0003747956436237944,"log_rms_mae_db":1.499712758406447,"rms_envelope_nrmse":0.13989569281591027,"envelope_correlation":0.9900762686313286,"rms_gain_error_db":-0.8563312073966557,"reference_mean_power":0.006220557891926456,"output_mean_power":0.005107356996973121},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.814448626587233e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017308539978626828,"output_power":0.001494956231080255,"reference_active":true,"delta_db":-0.636319578344934},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0006400481816254359,"output_power":0.0005626151128481409,"reference_active":true,"delta_db":-0.5600127417532973},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4706618104713675e-05,"output_power":4.56117474873272e-05,"reference_active":true,"delta_db":0.08704893054704188}],"active_band_count":3,"active_band_mae_db":0.4277937502150911}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.00031163820648484283,"log_rms_mae_db":1.281685170617762,"rms_envelope_nrmse":0.1322347394286732,"envelope_correlation":0.9917113281622995,"rms_gain_error_db":-0.8563304876063504,"reference_mean_power":0.006257088451954998,"output_mean_power":0.005137351076462764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":6.546091034740867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001677353066756132,"output_power":0.0014689459658446115,"reference_active":true,"delta_db":-0.5761866619380974},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035132010859926165,"output_power":0.00030841857846461445,"reference_active":true,"delta_db":-0.5656247720596731},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3827329510863972e-06,"output_power":2.50588518848941e-06,"reference_active":true,"delta_db":0.21885798383819796}],"active_band_count":3,"active_band_mae_db":0.45355647261198945}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00015152624573583172,"log_rms_mae_db":1.791042270813823,"rms_envelope_nrmse":0.18357281996080813,"envelope_correlation":0.981590907658929,"rms_gain_error_db":-1.021651999029308,"reference_mean_power":0.001672933915423699,"output_mean_power":0.0013222500308057578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":1.623496524404657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006256720367801862,"output_power":0.0004555357014227279,"reference_active":true,"delta_db":-1.3782432610629907},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001572550903934534,"output_power":0.00013514140423659626,"reference_active":true,"delta_db":-0.6581628527538524},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4941700779965039e-05,"output_power":1.694410061055895e-05,"reference_active":true,"delta_db":0.5461848647116896}],"active_band_count":3,"active_band_mae_db":0.8608636595095108}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0001385613366624899,"log_rms_mae_db":1.45944305561501,"rms_envelope_nrmse":0.17491277401057032,"envelope_correlation":0.9834951511360133,"rms_gain_error_db":-1.0216519337520193,"reference_mean_power":0.0016827742497549935,"output_mean_power":0.0013300276337921393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":1.5301470087365332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005852793815004823,"output_power":0.00043175065993128107,"reference_active":true,"delta_db":-1.321302147922482},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151264176170612e-05,"output_power":6.954010130276814e-05,"reference_active":true,"delta_db":-0.6898964950647175},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3333886770336671e-06,"output_power":1.401823170321516e-06,"reference_active":true,"delta_db":0.21736471200249508}],"active_band_count":3,"active_band_mae_db":0.7428544516632316}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00030489717691300735,"log_rms_mae_db":1.8026722003094233,"rms_envelope_nrmse":0.1459048365602061,"envelope_correlation":0.9899446065106252,"rms_gain_error_db":-0.6913827458639488,"reference_mean_power":0.002589850567098419,"output_mean_power":0.0022086984768163297},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":4.739471399174343e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014413951986732994,"output_power":0.0013705710722210196,"reference_active":true,"delta_db":-0.2188150988061504},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004245147064825656,"output_power":0.0004065092014382407,"reference_active":true,"delta_db":-0.188223597179488},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.389197181736509e-05,"output_power":3.702667792285281e-05,"reference_active":true,"delta_db":1.9026275459733621}],"active_band_count":3,"active_band_mae_db":0.7698887473196668}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0002474836728453423,"log_rms_mae_db":1.5758406759528154,"rms_envelope_nrmse":0.13342104728907778,"envelope_correlation":0.9919922102845553,"rms_gain_error_db":-0.6913790069807988,"reference_mean_power":0.002605032676611575,"output_mean_power":0.0022216481257373803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":4.638343817264176e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014427582782272214,"output_power":0.0013878482204781482,"reference_active":true,"delta_db":-0.16851601991611245},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020918188683306623,"output_power":0.00019912137193406806,"reference_active":true,"delta_db":-0.21406200167974926},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0082267328987973e-06,"output_power":3.2449225156034768e-06,"reference_active":true,"delta_db":2.083915869156484}],"active_band_count":3,"active_band_mae_db":0.8221646302507818}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":1.6976727669156422,"rms_envelope_nrmse":0.25203564041614634,"envelope_correlation":0.9553027698655882,"rms_gain_error_db":-0.25799972958462253,"reference_mean_power":0.00014375800668184295,"output_mean_power":0.00013546654963065346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.640530401894145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.05643435973393e-05,"output_power":1.976888353395794e-05,"reference_active":true,"delta_db":-0.17132708800446167},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8444282129891133e-05,"output_power":1.6142203286137278e-05,"reference_active":true,"delta_db":-0.5789894443531652},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6443800103937953e-06,"output_power":6.069171158111951e-06,"reference_active":true,"delta_db":2.215057294576685}],"active_band_count":3,"active_band_mae_db":0.9884579423114372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":1.3775126081473767,"rms_envelope_nrmse":0.21196890017960776,"envelope_correlation":0.9659063314539669,"rms_gain_error_db":-0.2579549867908101,"reference_mean_power":0.0001445994240238279,"output_mean_power":0.0001362608407824841},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":5.765239137842011e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0474191785118637e-05,"output_power":2.1639622359351833e-05,"reference_active":true,"delta_db":0.24042910397578918},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1083544144974644e-05,"output_power":1.0157516827454251e-05,"reference_active":true,"delta_db":-0.37891104909014145},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8424426027278453e-07,"output_power":5.105521329620998e-07,"reference_active":true,"delta_db":2.543483911422712}],"active_band_count":3,"active_band_mae_db":1.0542746881628808}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.708690643310547,"predicted_mos":3.708690643310547,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":226816,"input_sample_rate":44100,"input_seconds":5.143219954648526,"input_float32_sha256":"bd6d32abfb24f552aacee14354e667b43ba9a1db02254641654b32813f14705a","internal_samples":82292,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7021929590009677},"started_at":"2026-10-13T07:21:25.600032+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008838570998705109,"native_input_samples":226454,"native_input_sha256":"643a2ba22866177fef6df5f6406e627a6e9943be43e4f95b4973d74e6e5f912e","output_sha256":"9373ba15b6b6835bcbdb7706ed50a47d4b61d1e484c2458d387e6f9b751141bd","output_samples":226816,"length_delta_samples":362,"analysis_s":0.12922397200054547,"synthesis_s":29.823850701000993,"total_s":29.953074673001538,"rtf":5.83310779707741,"output_peak":0.7823904156684875,"output_rms":0.07534629073618863,"finished_at":"2026-10-13T07:21:55.596062+00:00","evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"9373ba15b6b6835bcbdb7706ed50a47d4b61d1e484c2458d387e6f9b751141bd","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":82160,"output_samples":82292,"length_delta_samples":132,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.457113428821458,"output_peak":0.7841067674342819},"common_prefix":{"samples":82160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983959558644826,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.21790438813407,"spectral_convergence":0.5332701053498267,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.35825124189997,"spectral_convergence":0.518768532018068,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.271137557481241,"spectral_convergence":0.5292771549401462,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":7.414514666305048e-05,"delta_db":-0.5250599463951277,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.0013025281709182397,"delta_db":-0.9065492983222805,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0013928860894371938,"delta_db":0.20760430779666073,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.002649002386041642,"delta_db":-0.052733784281419324,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.00015965263665441802,"delta_db":-2.8637898920409373,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00011011922181929893,"delta_db":-2.7621460786819774,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2196472179197337,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3842143136426346,"pitch":null,"output_rms_dbfs":-22.45014937810943,"output_peak":0.7841067674342819,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.282431062505092},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82160,"output_samples":82160,"intelligibility":{"estoi":{"value":0.7677587879631346,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07883212739732415,"rms_dbfs":-22.065935064466796,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.49939156382414,"mean":5.828649415013389e-06,"zero_energy":false},"output":{"rms":0.07542104249832299,"rms_dbfs":-22.45014937810943,"peak":0.7841067674342819,"peak_dbfs":-2.1124959567565864,"crest_factor_db":20.337653421352844,"mean":-7.898362535098444e-05,"zero_energy":false},"rms_difference_db":-0.3842143136426337},"full_scale":{"reference":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":2.083221069480143e-06,"output_rms_dbfs":-56.812646406806955,"output_zero_energy":false,"output_peak":0.01128506145888666,"output_energy_fraction":7.916492925197214e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":252,"both_estimated_voiced_frames":236,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9672131147540983,"absolute_difference_cents_median":6.164641813708717,"absolute_difference_cents_p95":25.267606889162167,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.0003747956436237944,"log_rms_mae_db":1.5155914871698157,"rms_envelope_nrmse":0.11545405169224947,"envelope_correlation":0.9910662988980559,"rms_gain_error_db":-0.3842144158629609,"reference_mean_power":0.006220557891926456,"output_mean_power":0.005693874553798041},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.814448626587233e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017308539978626828,"output_power":0.0017940405601793398,"reference_active":true,"delta_db":0.15571821946269931},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0006400481816254359,"output_power":0.0005518724731655328,"reference_active":true,"delta_db":-0.6437393564673821},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4706618104713675e-05,"output_power":3.0123658001774394e-05,"reference_active":true,"delta_db":-1.714641099414052}],"active_band_count":3,"active_band_mae_db":0.8380328917813777}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.00031163820648484283,"log_rms_mae_db":1.3359787636201435,"rms_envelope_nrmse":0.09779923895624922,"envelope_correlation":0.9941809839428482,"rms_gain_error_db":-0.38420530522100754,"reference_mean_power":0.006257088451954998,"output_mean_power":0.005727324152444731},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":6.546091034740867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001677353066756132,"output_power":0.0017340271700558379,"reference_active":true,"delta_db":0.14431410992074126},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035132010859926165,"output_power":0.00030220326645463805,"reference_active":true,"delta_db":-0.6540385409768946},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3827329510863972e-06,"output_power":1.5442837323256823e-06,"reference_active":true,"delta_db":-1.8834827421074471}],"active_band_count":3,"active_band_mae_db":0.8939451310016944}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00015152624573583172,"log_rms_mae_db":3.3113586224737745,"rms_envelope_nrmse":0.19734794469641234,"envelope_correlation":0.9703882805140993,"rms_gain_error_db":-0.2902376639990869,"reference_mean_power":0.001672933915423699,"output_mean_power":0.0015647862430082833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":1.623496524404657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006256720367801862,"output_power":0.0005727729989953591,"reference_active":true,"delta_db":-0.38364208674324635},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001572550903934534,"output_power":0.000192802155222047,"reference_active":true,"delta_db":0.8850717173211351},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4941700779965039e-05,"output_power":2.2452177535448342e-05,"reference_active":true,"delta_db":1.7685843259474396}],"active_band_count":3,"active_band_mae_db":1.0124327100039403}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0001385613366624899,"log_rms_mae_db":3.0330267922558547,"rms_envelope_nrmse":0.1757103096845348,"envelope_correlation":0.9753427488081932,"rms_gain_error_db":-0.2902446284430136,"reference_mean_power":0.0016827742497549935,"output_mean_power":0.0015739879197869293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":1.5301470087365332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005852793815004823,"output_power":0.0005402689964708072,"reference_active":true,"delta_db":-0.34753178741366153},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151264176170612e-05,"output_power":0.00010479541262107785,"reference_active":true,"delta_db":1.0911730347565234},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3333886770336671e-06,"output_power":1.6238797531627607e-06,"reference_active":true,"delta_db":0.8559710414751422}],"active_band_count":3,"active_band_mae_db":0.7648919545484424}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00030489717691300735,"log_rms_mae_db":2.737040659480887,"rms_envelope_nrmse":0.2476692619431493,"envelope_correlation":0.9742385275274494,"rms_gain_error_db":-1.525800285184364,"reference_mean_power":0.002589850567098419,"output_mean_power":0.0018226138986737134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":4.739471399174343e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014413951986732994,"output_power":0.0010480257803449937,"reference_active":true,"delta_db":-1.3841110510706975},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004245147064825656,"output_power":0.00023147646050637852,"reference_active":true,"delta_db":-2.63385907154733},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.389197181736509e-05,"output_power":1.0650934028888795e-05,"reference_active":true,"delta_db":-3.508642990828097}],"active_band_count":3,"active_band_mae_db":2.508871037815375}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0002474836728453423,"log_rms_mae_db":2.467981925906984,"rms_envelope_nrmse":0.235484974424948,"envelope_correlation":0.9779542730086558,"rms_gain_error_db":-1.5258026260635214,"reference_mean_power":0.002605032676611575,"output_mean_power":0.0018332973585505183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":4.638343817264176e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014427582782272214,"output_power":0.001022298770763751,"reference_active":true,"delta_db":-1.4961573628444018},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020918188683306623,"output_power":0.00010938002369118847,"reference_active":true,"delta_db":-2.8158606293649346},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0082267328987973e-06,"output_power":9.281320523134304e-07,"reference_active":true,"delta_db":-3.3520297302432027}],"active_band_count":3,"active_band_mae_db":2.554682574150846}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":3.8704588545004888,"rms_envelope_nrmse":0.3426064946622044,"envelope_correlation":0.9227009313800397,"rms_gain_error_db":-1.4586775340468197,"reference_mean_power":0.00014375800668184295,"output_mean_power":0.00010274584989501667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.640530401894145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.05643435973393e-05,"output_power":1.4289719364945458e-05,"reference_active":true,"delta_db":-1.580911518348183},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8444282129891133e-05,"output_power":1.1077344610841166e-05,"reference_active":true,"delta_db":-2.2142609009782506},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6443800103937953e-06,"output_power":2.607830164896365e-06,"reference_active":true,"delta_db":-1.4534435132016585}],"active_band_count":3,"active_band_mae_db":1.7495386441760308}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":3.660538405853531,"rms_envelope_nrmse":0.31529039442803186,"envelope_correlation":0.933103974834996,"rms_gain_error_db":-1.4586124735387143,"reference_mean_power":0.0001445994240238279,"output_mean_power":0.00010334877081911856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":5.765239137842011e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0474191785118637e-05,"output_power":1.4746492778530167e-05,"reference_active":true,"delta_db":-1.425180245971298},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1083544144974644e-05,"output_power":6.213786422437408e-06,"reference_active":true,"delta_db":-2.5132233358761673},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8424426027278453e-07,"output_power":2.1209009921243454e-07,"reference_active":true,"delta_db":-1.2717130850625658}],"active_band_count":3,"active_band_mae_db":1.736705555636677}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"golf","status":"measured","mos":{"status":"measured","value":2.650346279144287,"predicted_mos":2.650346279144287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":123120,"input_sample_rate":24000,"input_seconds":5.13,"input_float32_sha256":"6bf161fd2d232cc5357cb73f0a2fe83379bd5731d937d3132a10863cd8f80238","internal_samples":82080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0349990609993256},"started_at":"2026-10-13T05:02:46.242563+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026969110003847163,"native_input_samples":123240,"native_input_sha256":"d8a4be041ab24f460fe177307feeaf69719f94900a30669098cc2c46fb6b44fe","output_sha256":"e91edd797431e773071512236d1b6e93616d2378b504eb06a2a29f690137195d","output_samples":123120,"length_delta_samples":-120,"analysis_s":0.6975041839996265,"synthesis_s":0.15305159899980936,"total_s":0.8505557829994359,"rtf":0.1656389061342621,"output_peak":0.8284342288970947,"output_rms":0.07167001999790028,"finished_at":"2026-10-13T05:02:47.102184+00:00","analysis_stages":{"external_f0_s":0.043296345000271685},"evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"e91edd797431e773071512236d1b6e93616d2378b504eb06a2a29f690137195d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":82160,"output_samples":82080,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.891170148665097,"output_peak":0.8187076794217193},"common_prefix":{"samples":82080,"sample_rate":16000,"reference_coverage":0.9990262901655307,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82080,"output_samples":82080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.820228800234177,"spectral_convergence":0.3665486762524237,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.414115671089186,"spectral_convergence":0.27141592426504596,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.496984366310854,"spectral_convergence":0.29991223207192036,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.375404082282796e-05,"output_power":7.425809128307037e-05,"delta_db":-0.5226198731294756,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001605788364377437,"output_power":0.0014284704543816155,"delta_db":-0.5081704454401144,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0013298173098334272,"output_power":0.0010729241885171616,"delta_db":-0.9322294538881226,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002683972937964128,"output_power":0.002140124280078256,"delta_db":-0.98339138419432,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00030901259918202133,"output_power":0.0002591214398547237,"delta_db":-0.7647283876560101,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002082094129597234,"output_power":0.00016415320238869798,"delta_db":-1.0325100008243484,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7906082575220652,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8294612605521661,"pitch":null,"output_rms_dbfs":-22.891170148665097,"output_peak":0.8187076794217193,"reference_rms_dbfs":-22.061708888112932,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.577109612544739},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82080,"output_samples":82080,"intelligibility":{"estoi":{"value":0.8790081394734237,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07887049301673447,"rms_dbfs":-22.06170888811293,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.495165387470273,"mean":4.952042191116898e-06,"zero_energy":false},"output":{"rms":0.0716871793294267,"rms_dbfs":-22.891170148665097,"peak":0.8187076794217193,"peak_dbfs":-1.7374227187793545,"crest_factor_db":21.153747429885744,"mean":0.000381537961003388,"zero_energy":false},"rms_difference_db":-0.8294612605521685},"full_scale":{"reference":{"sample_count":82080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":4.0135714303699034e-06,"output_rms_dbfs":-53.964690033887315,"output_zero_energy":false,"output_peak":0.03955650474196192,"output_energy_fraction":0.00016898712805193283,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":236,"both_estimated_voiced_frames":228,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9344262295081968,"absolute_difference_cents_median":5.217569878437977,"absolute_difference_cents_p95":20.942564962128206,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1025,"rms_floor":0.0003747956436237944,"log_rms_mae_db":1.773475243893656,"rms_envelope_nrmse":0.13881577366379738,"envelope_correlation":0.9900115493274643,"rms_gain_error_db":-0.8294590046133807,"reference_mean_power":0.006226617597832866,"output_mean_power":0.005144063183138387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1025,"frequency_resolution_hz":0.1951219512195122,"power_floor":6.81616373783378e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017208603354636886,"output_power":0.0014420249761292153,"reference_active":true,"delta_db":-0.7677284203467731},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000649322351029643,"output_power":0.0006663948036316608,"reference_active":true,"delta_db":0.11271249447840297},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.464129324853342e-05,"output_power":6.079890472342771e-05,"reference_active":true,"delta_db":1.341589881120053}],"active_band_count":3,"active_band_mae_db":0.740676931981743}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1019,"rms_floor":0.00031163820648484283,"log_rms_mae_db":1.437760347674555,"rms_envelope_nrmse":0.12082851896865066,"envelope_correlation":0.9939808951438388,"rms_gain_error_db":-0.829447957703683,"reference_mean_power":0.006263207975473741,"output_mean_power":0.005174305150568865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1019,"frequency_resolution_hz":0.19627085377821393,"power_floor":6.5475613648078524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001675881250620287,"output_power":0.0014291479113607024,"reference_active":true,"delta_db":-0.6916606326113378},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035146204409910585,"output_power":0.0003660161206940185,"reference_active":true,"delta_db":0.1762178318407884},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3914866941013442e-06,"output_power":4.1773164519996525e-06,"reference_active":true,"delta_db":2.422294075658596}],"active_band_count":3,"active_band_mae_db":1.0967241800369074}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1025,"rms_floor":0.00015152624573583172,"log_rms_mae_db":1.6650852783825438,"rms_envelope_nrmse":0.1854732165324441,"envelope_correlation":0.9779733747000529,"rms_gain_error_db":-0.8188150079884191,"reference_mean_power":0.0016745661776214878,"output_mean_power":0.00138682229345527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1025,"frequency_resolution_hz":0.1951219512195122,"power_floor":1.6237580940776693e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006172068441155169,"output_power":0.0004955747076955397,"reference_active":true,"delta_db":-0.9532159976340768},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00016533224711667437,"output_power":0.0001858562605819251,"reference_active":true,"delta_db":0.5081962641225609},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495640508585152e-05,"output_power":1.4837497412294342e-05,"reference_active":true,"delta_db":-0.034665630438325656}],"active_band_count":3,"active_band_mae_db":0.49869263073165454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1019,"rms_floor":0.0001385613366624899,"log_rms_mae_db":1.3607624714891946,"rms_envelope_nrmse":0.15970050661405932,"envelope_correlation":0.9845498563999731,"rms_gain_error_db":-0.8188145105271537,"reference_mean_power":0.001684425552885097,"output_mean_power":0.0013949876733453746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1019,"frequency_resolution_hz":0.19627085377821393,"power_floor":1.5303545721680128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000585009025473071,"output_power":0.00048005578247896515,"reference_active":true,"delta_db":-0.8587086107090303},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.13039615137169e-05,"output_power":9.737100136703341e-05,"reference_active":true,"delta_db":0.7831792933826904},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.333911681912081e-06,"output_power":1.1585143829082367e-06,"reference_active":true,"delta_db":-0.6122564605382581}],"active_band_count":3,"active_band_mae_db":0.7513814548766596}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1025,"rms_floor":0.00030489717691300735,"log_rms_mae_db":2.260780461838561,"rms_envelope_nrmse":0.16342799832908858,"envelope_correlation":0.9891441430561001,"rms_gain_error_db":-0.9327207304401132,"reference_mean_power":0.0025923703790156206,"output_mean_power":0.0020913416052764453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1025,"frequency_resolution_hz":0.1951219512195122,"power_floor":4.741077701483886e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001436231681102908,"output_power":0.0012394012981464036,"reference_active":true,"delta_db":-0.6401255568446218},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004284338592907288,"output_power":0.00036975034428435576,"reference_active":true,"delta_db":-0.6397519892187603},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.388295567836678e-05,"output_power":3.3799547104756165e-05,"reference_active":true,"delta_db":1.5082280832839352}],"active_band_count":3,"active_band_mae_db":0.9293685431157724}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1019,"rms_floor":0.0002474836728453423,"log_rms_mae_db":1.9627449655140905,"rms_envelope_nrmse":0.14132085884442358,"envelope_correlation":0.9934861298157099,"rms_gain_error_db":-0.9326948779678345,"reference_mean_power":0.0026075707473310695,"output_mean_power":0.002103616712759151},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1019,"frequency_resolution_hz":0.19627085377821393,"power_floor":4.639838984491208e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014408631481782226,"output_power":0.0012549779526567389,"reference_active":true,"delta_db":-0.5998663765631715},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020933546014559692,"output_power":0.0001810284608442619,"reference_active":true,"delta_db":-0.6309594246209066},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0176124943621174e-06,"output_power":3.1771588691940515e-06,"reference_active":true,"delta_db":1.9720117296229165}],"active_band_count":3,"active_band_mae_db":1.0676125102689982}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1025,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":2.610200527686244,"rms_envelope_nrmse":0.2259449903984888,"envelope_correlation":0.9647152388238867,"rms_gain_error_db":-0.26662714393030057,"reference_mean_power":0.00014389696139422293,"output_mean_power":0.00013532838806539518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1025,"frequency_resolution_hz":0.1951219512195122,"power_floor":6.640680217258388e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9632781609138226e-05,"output_power":1.9409061169085645e-05,"reference_active":true,"delta_db":-0.049773068670173094},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9397332859095146e-05,"output_power":1.7069415457041896e-05,"reference_active":true,"delta_db":-0.5552336943923135},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6304728430260785e-06,"output_power":3.3787036651251207e-06,"reference_active":true,"delta_db":-0.3121308953972857}],"active_band_count":3,"active_band_mae_db":0.3057125528199241}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82080,"uncovered_tail_samples":0,"metrics":{"frame_count":1019,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":2.452239123553018,"rms_envelope_nrmse":0.1779337601480293,"envelope_correlation":0.9769022494071036,"rms_gain_error_db":-0.266620119645281,"reference_mean_power":0.000144739951171322,"output_mean_power":0.00013612140083882565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1019,"frequency_resolution_hz":0.19627085377821393,"power_floor":5.7648476780544145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0483374750369076e-05,"output_power":2.057982210142247e-05,"reference_active":true,"delta_db":0.020401056221965985},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1069540153932542e-05,"output_power":9.540225738260028e-06,"reference_active":true,"delta_db":-0.6457092899410145},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8868748946505785e-07,"output_power":2.8242442232709634e-07,"reference_active":true,"delta_db":-0.09525714622654265}],"active_band_count":3,"active_band_mae_db":0.25378916412984104}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3643081188201904,"predicted_mos":3.3643081188201904,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":123360,"input_sample_rate":24000,"input_seconds":5.14,"input_float32_sha256":"8ede5a6de0d4f7e5e1d4bd48c4ad2c80f79ea70199b9986677ee7e876c8b323c","internal_samples":82240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4578397479999694},"started_at":"2026-10-13T07:41:21.916712+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002723654000874376,"native_input_samples":123240,"native_input_sha256":"d8a4be041ab24f460fe177307feeaf69719f94900a30669098cc2c46fb6b44fe","output_sha256":"23903a636248245cac392a5755d3cc0725be3b7665eec4c1ce6f9f725974a381","output_samples":123360,"length_delta_samples":120,"analysis_s":2.070807768002851,"synthesis_s":9.997580723997089,"total_s":12.06838849199994,"rtf":2.3502217121713613,"output_peak":0.685910701751709,"output_rms":0.07368774487630943,"finished_at":"2026-10-13T07:41:33.993752+00:00","analysis_stages":{"external_f0_s":1.444822593999561},"evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"23903a636248245cac392a5755d3cc0725be3b7665eec4c1ce6f9f725974a381","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":82160,"output_samples":82240,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.65203032273822,"output_peak":0.6856758329987941},"common_prefix":{"samples":82160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990272373540856,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.665396333236932,"spectral_convergence":0.4337588141001993,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.256128744490233,"spectral_convergence":0.26878532099037983,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.195922911269791,"spectral_convergence":0.26569567675333716,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":8.022670570234613e-05,"delta_db":-0.18269763801009387,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.0015300618924641277,"delta_db":-0.20733055973438158,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0012281665385928532,"delta_db":-0.338979095375488,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.002193797463714547,"delta_db":-0.8715919785645554,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.00027183177215535793,"delta_db":-0.5525486805428153,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00013116443291577736,"delta_db":-2.0026165563162674,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6926274180906002,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5818706428426362,"pitch":null,"output_rms_dbfs":-22.64780570730943,"output_peak":0.6856758329987941,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.372482662998984},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82160,"output_samples":82160,"intelligibility":{"estoi":{"value":0.884278275929448,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07883212739732415,"rms_dbfs":-22.065935064466796,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.49939156382414,"mean":5.828649415013389e-06,"zero_energy":false},"output":{"rms":0.07372413991051377,"rms_dbfs":-22.647805707309434,"peak":0.6856758329987941,"peak_dbfs":-3.27762314399357,"crest_factor_db":19.37018256331586,"mean":-2.413977360950038e-06,"zero_energy":false},"rms_difference_db":-0.5818706428426381},"full_scale":{"reference":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":5.4828410298628797e-05,"output_rms_dbfs":-42.60994345919219,"output_zero_energy":false,"output_peak":0.31071708989365476,"output_energy_fraction":0.0021805635874629476,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":254,"both_estimated_voiced_frames":235,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9631147540983607,"absolute_difference_cents_median":5.457228507213062,"absolute_difference_cents_p95":29.689657369176732,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.0003747956436237944,"log_rms_mae_db":1.90982110386846,"rms_envelope_nrmse":0.15173834868065175,"envelope_correlation":0.9830126870742973,"rms_gain_error_db":-0.5818691748095755,"reference_mean_power":0.006220557891926456,"output_mean_power":0.0054405451495737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.814448626587233e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017308539978626828,"output_power":0.0014588452168384923,"reference_active":true,"delta_db":-0.7425121973870525},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0006400481816254359,"output_power":0.0005539640966340884,"reference_active":true,"delta_db":-0.62731049819518},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4706618104713675e-05,"output_power":4.124187334472266e-05,"reference_active":true,"delta_db":-0.3503343411403812}],"active_band_count":3,"active_band_mae_db":0.5733856789075379}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.00031163820648484283,"log_rms_mae_db":1.5884210808825512,"rms_envelope_nrmse":0.11104660498605864,"envelope_correlation":0.992306314283827,"rms_gain_error_db":-0.5818520237253105,"reference_mean_power":0.006257088451954998,"output_mean_power":0.005472516654843243},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":6.546091034740867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001677353066756132,"output_power":0.0014104179232708159,"reference_active":true,"delta_db":-0.7527666881547803},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035132010859926165,"output_power":0.00030789309314453885,"reference_active":true,"delta_db":-0.5730306168828471},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3827329510863972e-06,"output_power":2.045487877045142e-06,"reference_active":true,"delta_db":-0.6627846085221192}],"active_band_count":3,"active_band_mae_db":0.6628606378532488}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00015152624573583172,"log_rms_mae_db":1.6431257899459262,"rms_envelope_nrmse":0.15093779370760524,"envelope_correlation":0.9828037708333633,"rms_gain_error_db":-0.2980695622137553,"reference_mean_power":0.001672933915423699,"output_mean_power":0.0015619669110999586},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":1.623496524404657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006256720367801862,"output_power":0.0005744827977576524,"reference_active":true,"delta_db":-0.3706971668832312},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001572550903934534,"output_power":0.0001429220065186087,"reference_active":true,"delta_db":-0.4150560784022849},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4941700779965039e-05,"output_power":1.3505349119360356e-05,"reference_active":true,"delta_db":-0.4389421967391957}],"active_band_count":3,"active_band_mae_db":0.40823181400823727}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0001385613366624899,"log_rms_mae_db":1.1848579525109022,"rms_envelope_nrmse":0.10721151721243073,"envelope_correlation":0.9913021437677126,"rms_gain_error_db":-0.2980693550643055,"reference_mean_power":0.0016827742497549935,"output_mean_power":0.0015711546033626105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":1.5301470087365332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005852793815004823,"output_power":0.000536184801072972,"reference_active":true,"delta_db":-0.3804872573025694},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151264176170612e-05,"output_power":7.48561499375664e-05,"reference_active":true,"delta_db":-0.36997482148549177},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3333886770336671e-06,"output_power":1.1510685193827049e-06,"reference_active":true,"delta_db":-0.638555862880621}],"active_band_count":3,"active_band_mae_db":0.4630059805562274}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00030489717691300735,"log_rms_mae_db":1.919810610042246,"rms_envelope_nrmse":0.17956513957891138,"envelope_correlation":0.9844352398460737,"rms_gain_error_db":-0.8793530670424228,"reference_mean_power":0.002589850567098419,"output_mean_power":0.002115141369384309},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":4.739471399174343e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014413951986732994,"output_power":0.0011316011451497486,"reference_active":true,"delta_db":-1.050896927996309},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004245147064825656,"output_power":0.00034399339335790084,"reference_active":true,"delta_db":-0.9134263841259946},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.389197181736509e-05,"output_power":1.7932291347325887e-05,"reference_active":true,"delta_db":-1.2461620752050955}],"active_band_count":3,"active_band_mae_db":1.0701617957757996}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0002474836728453423,"log_rms_mae_db":1.3635160980896925,"rms_envelope_nrmse":0.1373653830864482,"envelope_correlation":0.9933525914135185,"rms_gain_error_db":-0.8793217821646669,"reference_mean_power":0.002605032676611575,"output_mean_power":0.002127555985282907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":4.638343817264176e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014427582782272214,"output_power":0.00113219835064666,"reference_active":true,"delta_db":-1.052710569917514},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020918188683306623,"output_power":0.0001703425574819856,"reference_active":true,"delta_db":-0.8920091269039532},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0082267328987973e-06,"output_power":1.5619914658179428e-06,"reference_active":true,"delta_db":-1.091340872622783}],"active_band_count":3,"active_band_mae_db":1.01202018981475}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":2.0356682184595223,"rms_envelope_nrmse":0.2838118678182304,"envelope_correlation":0.9441345455892985,"rms_gain_error_db":-0.9002327937524485,"reference_mean_power":0.00014375800668184295,"output_mean_power":0.00011684463141337929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.640530401894145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.05643435973393e-05,"output_power":1.695929449100341e-05,"reference_active":true,"delta_db":-0.8370707003686645},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8444282129891133e-05,"output_power":1.553759480337899e-05,"reference_active":true,"delta_db":-0.7447796518505178},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6443800103937953e-06,"output_power":2.53364274729668e-06,"reference_active":true,"delta_db":-1.5787827805884602}],"active_band_count":3,"active_band_mae_db":1.0535443776025475}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":1.4485243829817211,"rms_envelope_nrmse":0.17160496586199683,"envelope_correlation":0.9833532640329025,"rms_gain_error_db":-0.9001995484949737,"reference_mean_power":0.0001445994240238279,"output_mean_power":0.00011752942413095074},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":5.765239137842011e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0474191785118637e-05,"output_power":1.6420955904090267e-05,"reference_active":true,"delta_db":-0.9580833221206108},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1083544144974644e-05,"output_power":9.41887364217714e-06,"reference_active":true,"delta_db":-0.7067968467367519},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8424426027278453e-07,"output_power":2.0554724609235036e-07,"reference_active":true,"delta_db":-1.4078004125781143}],"active_band_count":3,"active_band_mae_db":1.0242268604784923}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0828564167022705,"predicted_mos":3.0828564167022705,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":123480,"input_sample_rate":24000,"input_seconds":5.145,"input_float32_sha256":"2321f3178758689abcf5d9e83f5341884d1e40254c318cad9bb0ba6ac7929ae5","internal_samples":82320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.09881757400035},"started_at":"2026-10-13T05:11:04.199987+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0037311710002541076,"native_input_samples":123240,"native_input_sha256":"d8a4be041ab24f460fe177307feeaf69719f94900a30669098cc2c46fb6b44fe","output_sha256":"eef9195542e7f1d630ee5a58571e7108c3b96abd17b1f0b38b83edd5cfc8debe","output_samples":123480,"length_delta_samples":240,"analysis_s":0.6238414130002639,"synthesis_s":0.18205135700009123,"total_s":0.8058927700003551,"rtf":0.15694114313541482,"output_peak":0.8278371095657349,"output_rms":0.07799093985228903,"finished_at":"2026-10-13T05:11:05.017206+00:00","analysis_stages":{"external_f0_s":0.11567941800058179},"evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"eef9195542e7f1d630ee5a58571e7108c3b96abd17b1f0b38b83edd5cfc8debe","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":82160,"output_samples":82320,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.158640169269226,"output_peak":0.8290893393878828},"common_prefix":{"samples":82160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980563654033042,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.602368639872085,"spectral_convergence":0.16217183655401074,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.457435505091596,"spectral_convergence":0.151343780187341,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.946128010336503,"spectral_convergence":0.17472931631802366,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":8.679022945515702e-05,"delta_db":0.15882113338208598,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.0015827914840614405,"delta_db":-0.06018349819544934,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0013133864099164664,"delta_db":-0.047626522152875335,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.0026380444101024593,"delta_db":-0.07073625515459706,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.00028932514114161115,"delta_db":-0.281689108308155,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00018476185479029695,"delta_db":-0.5146542816793739,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18895179981208945,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08425691762074831,"pitch":null,"output_rms_dbfs":-22.150191982087545,"output_peak":0.8290893393878828,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.668644051766728},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82160,"output_samples":82160,"intelligibility":{"estoi":{"value":0.9246179562544888,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07883212739732415,"rms_dbfs":-22.065935064466796,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.49939156382414,"mean":5.828649415013389e-06,"zero_energy":false},"output":{"rms":0.0780711184079454,"rms_dbfs":-22.150191982087545,"peak":0.8290893393878828,"peak_dbfs":-1.6279733814516308,"crest_factor_db":20.522218600635913,"mean":3.290188347771777e-05,"zero_energy":false},"rms_difference_db":-0.08425691762074905},"full_scale":{"reference":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":1.4058246554416443e-05,"output_rms_dbfs":-48.52068844268534,"output_zero_energy":false,"output_peak":0.1743304698672928,"output_energy_fraction":0.0004985777401588964,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":241,"both_estimated_voiced_frames":234,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9590163934426229,"absolute_difference_cents_median":0.5864724968607729,"absolute_difference_cents_p95":6.827998629897037,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.0003747956436237944,"log_rms_mae_db":0.6260029081926833,"rms_envelope_nrmse":0.03931204125443042,"envelope_correlation":0.998771873660741,"rms_gain_error_db":-0.08425614087048067,"reference_mean_power":0.006220557891926456,"output_mean_power":0.006101037889760625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.814448626587233e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017308539978626828,"output_power":0.0017265932123219628,"reference_active":true,"delta_db":-0.01070406269806196},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0006400481816254359,"output_power":0.0006130283432873924,"reference_active":true,"delta_db":-0.18732113555009705},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4706618104713675e-05,"output_power":3.634848057576748e-05,"reference_active":true,"delta_db":-0.8988555691572553}],"active_band_count":3,"active_band_mae_db":0.36562692246847145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.00031163820648484283,"log_rms_mae_db":0.37547332617213547,"rms_envelope_nrmse":0.022010941757205263,"envelope_correlation":0.9996654299516391,"rms_gain_error_db":-0.0842550416127113,"reference_mean_power":0.006257088451954998,"output_mean_power":0.00613686811548645},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":6.546091034740867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001677353066756132,"output_power":0.0016707214372356526,"reference_active":true,"delta_db":-0.017204419255354153},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035132010859926165,"output_power":0.0003373108764284758,"reference_active":true,"delta_db":-0.17672663294624472},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3827329510863972e-06,"output_power":2.1703605580362677e-06,"reference_active":true,"delta_db":-0.4054348237785294}],"active_band_count":3,"active_band_mae_db":0.1997886253267094}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00015152624573583172,"log_rms_mae_db":0.6212022802997295,"rms_envelope_nrmse":0.03904010149352588,"envelope_correlation":0.9988398920593164,"rms_gain_error_db":-0.036589506319194776,"reference_mean_power":0.001672933915423699,"output_mean_power":0.0016588985787944841},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":1.623496524404657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006256720367801862,"output_power":0.0006246417313368451,"reference_active":true,"delta_db":-0.007157500483605164},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001572550903934534,"output_power":0.0001485616908233842,"reference_active":true,"delta_db":-0.24697878994341},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4941700779965039e-05,"output_power":1.5315045461797916e-05,"reference_active":true,"delta_db":0.1071825527118721}],"active_band_count":3,"active_band_mae_db":0.12043961437962908}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0001385613366624899,"log_rms_mae_db":0.3229336805942118,"rms_envelope_nrmse":0.019833798647530935,"envelope_correlation":0.999694178173493,"rms_gain_error_db":-0.03658992992671291,"reference_mean_power":0.0016827742497549935,"output_mean_power":0.0016686561933665434},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":1.5301470087365332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005852793815004823,"output_power":0.0005827225408479977,"reference_active":true,"delta_db":-0.01901407076419834},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151264176170612e-05,"output_power":7.729717970307152e-05,"reference_active":true,"delta_db":-0.23061320155096945},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3333886770336671e-06,"output_power":1.2385401636952628e-06,"reference_active":true,"delta_db":-0.32046668269868706}],"active_band_count":3,"active_band_mae_db":0.19003131833795162}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00030489717691300735,"log_rms_mae_db":0.8462021160187185,"rms_envelope_nrmse":0.04850649169430304,"envelope_correlation":0.9986246108721056,"rms_gain_error_db":-0.09301765356316159,"reference_mean_power":0.002589850567098419,"output_mean_power":0.002534970682019549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":4.739471399174343e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014413951986732994,"output_power":0.0014399431975534451,"reference_active":true,"delta_db":-0.004377105641430284},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004245147064825656,"output_power":0.0004031239472856221,"reference_active":true,"delta_db":-0.22454142256379245},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.389197181736509e-05,"output_power":1.8537128853656425e-05,"reference_active":true,"delta_db":-1.102095250133945}],"active_band_count":3,"active_band_mae_db":0.44367125944638924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0002474836728453423,"log_rms_mae_db":0.5604096347386207,"rms_envelope_nrmse":0.02561103966231313,"envelope_correlation":0.9996698109911517,"rms_gain_error_db":-0.09301636060685203,"reference_mean_power":0.002605032676611575,"output_mean_power":0.0025498318361749286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":4.638343817264176e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014427582782272214,"output_power":0.0014344870614811703,"reference_active":true,"delta_db":-0.024969394420059684},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020918188683306623,"output_power":0.00019666928128904103,"reference_active":true,"delta_db":-0.2678754534660093},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0082267328987973e-06,"output_power":1.6759740843242602e-06,"reference_active":true,"delta_db":-0.7854544514726413}],"active_band_count":3,"active_band_mae_db":0.3594330997862367}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":0.943505928483515,"rms_envelope_nrmse":0.12255612710642708,"envelope_correlation":0.9894522711071598,"rms_gain_error_db":-0.16640042500845448,"reference_mean_power":0.00014375800668184295,"output_mean_power":0.00013835408923547693},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.640530401894145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.05643435973393e-05,"output_power":1.934662459226972e-05,"reference_active":true,"delta_db":-0.2650964706498212},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8444282129891133e-05,"output_power":1.6897909988657257e-05,"reference_active":true,"delta_db":-0.3802876431810398},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6443800103937953e-06,"output_power":2.7644166350196347e-06,"reference_active":true,"delta_db":-1.2002015815128784}],"active_band_count":3,"active_band_mae_db":0.6151952317812465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":0.6549063845399562,"rms_envelope_nrmse":0.0679977192287296,"envelope_correlation":0.9967060524123167,"rms_gain_error_db":-0.16640974467351027,"reference_mean_power":0.0001445994240238279,"output_mean_power":0.00013916357874773646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":5.765239137842011e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0474191785118637e-05,"output_power":1.9058709636221218e-05,"reference_active":true,"delta_db":-0.311132735475301},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1083544144974644e-05,"output_power":9.949846462871508e-06,"reference_active":true,"delta_db":-0.4686227622863359},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8424426027278453e-07,"output_power":2.0341855319485982e-07,"reference_active":true,"delta_db":-1.4530114267837855}],"active_band_count":3,"active_band_mae_db":0.7442556415151408}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0009","method":"magphase","status":"measured","mos":{"status":"measured","value":3.0397298336029053,"predicted_mos":3.0397298336029053,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":246774,"input_sample_rate":48000,"input_seconds":5.141125,"input_float32_sha256":"bd3c88a0aef3d6ea39e76450b03c0becc3327d52be8f407d09f4decfeda60d25","internal_samples":82258,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.113955815999361},"started_at":"2026-10-13T05:16:49.710623+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.007968654001160758,"native_input_samples":246480,"native_input_sha256":"668de4e7a09d5afcc0364213527469fa88f9f3c98b7cd057a81250ef1fc8b398","output_sha256":"94dfaa4c2f18d60d9521a49867aa18286fe122b169084e848ff3286a92a60b0d","output_samples":246774,"length_delta_samples":294,"analysis_s":3.5467378760004067,"synthesis_s":1.07330082199951,"total_s":4.620038697999917,"rtf":0.8997154231742779,"output_peak":0.7684038803902015,"output_rms":0.07675940432106053,"finished_at":"2026-10-13T05:16:54.348811+00:00","evaluation_input":{"reference":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"94dfaa4c2f18d60d9521a49867aa18286fe122b169084e848ff3286a92a60b0d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":82160,"output_samples":82258,"length_delta_samples":98,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.298054441527135,"output_peak":0.7550955258806252},"common_prefix":{"samples":82160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988086265165698,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":82160,"output_samples":82160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.356094819817768,"spectral_convergence":0.35749697196039165,"spectral_convergence_status":"ok","amplitude_floor":2.2676184880983433e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.956430580709632,"spectral_convergence":0.17724068833038797,"spectral_convergence_status":"ok","amplitude_floor":1.9761500134157035e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.240796763209384,"spectral_convergence":0.18720110338286094,"spectral_convergence_status":"ok","amplitude_floor":1.298774161363914e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.367365175011865e-05,"output_power":7.896448290309213e-05,"delta_db":-0.25156927463566797,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016048781148481498,"output_power":0.001435343985527898,"delta_db":-0.48486060940251363,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001327868811244639,"output_power":0.0012959851719180735,"delta_db":-0.10555137828216724,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0026813637139198997,"output_power":0.0025478205747420817,"delta_db":-0.22186887503942104,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003087131182449005,"output_power":0.000329108925219687,"delta_db":0.2778457574954463,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00020800689998023804,"output_power":0.00021087677601266144,"delta_db":0.059510116913199455,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2335343352947359,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.22694368141814833,"pitch":null,"output_rms_dbfs":-22.29287874588494,"output_peak":0.7550955258806252,"reference_rms_dbfs":-22.065935064466796,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.184440721245594},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":82160,"output_samples":82160,"intelligibility":{"estoi":{"value":0.9242276416827203,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07883212739732415,"rms_dbfs":-22.065935064466796,"peak":0.744171142578125,"peak_dbfs":-2.5665435006426556,"crest_factor_db":19.49939156382414,"mean":5.828649415013389e-06,"zero_energy":false},"output":{"rms":0.07679908799148263,"rms_dbfs":-22.29287874588494,"peak":0.7550955258806252,"peak_dbfs":-2.4399620604350853,"crest_factor_db":19.85291668544986,"mean":0.0011253998151653703,"zero_energy":false},"rms_difference_db":-0.22694368141814536},"full_scale":{"reference":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":82160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":513,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4047177447937428e-05,"threshold_dbfs":-48.52410931412396,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":111,"quiet_samples":17760,"reference_mean_power":2.7924773979697142e-06,"output_mean_power":2.6495727179289313e-06,"output_rms_dbfs":-55.76824156707532,"output_zero_energy":false,"output_peak":0.031155245500245204,"output_energy_fraction":9.710604108446085e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":506,"reference_estimated_voiced_frames":244,"output_estimated_voiced_frames":241,"both_estimated_voiced_frames":239,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9795081967213115,"absolute_difference_cents_median":2.5637463237621043,"absolute_difference_cents_p95":11.76336368000445,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":82160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.0003747956436237944,"log_rms_mae_db":1.0378040646030662,"rms_envelope_nrmse":0.07952342790027556,"envelope_correlation":0.9951333071453673,"rms_gain_error_db":-0.22694349280647186,"reference_mean_power":0.006220557891926456,"output_mean_power":0.005903845543447926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.814448626587233e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017308539978626828,"output_power":0.0016548074154996853,"reference_active":true,"delta_db":-0.19512977143837007},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0006400481816254359,"output_power":0.0006315425375228289,"reference_active":true,"delta_db":-0.0581006032338935},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.4706618104713675e-05,"output_power":4.4090411060014416e-05,"reference_active":true,"delta_db":-0.0602767046792326}],"active_band_count":3,"active_band_mae_db":0.10450235978383204}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.00031163820648484283,"log_rms_mae_db":0.7104274958453859,"rms_envelope_nrmse":0.04977674497902546,"envelope_correlation":0.9983685506355447,"rms_gain_error_db":-0.22694434375356054,"reference_mean_power":0.006257088451954998,"output_mean_power":0.005938515029634845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":6.546091034740867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001677353066756132,"output_power":0.0016102853761333388,"reference_active":true,"delta_db":-0.1772163817025654},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00035132010859926165,"output_power":0.0003452962098667112,"reference_active":true,"delta_db":-0.07511196972704486},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3827329510863972e-06,"output_power":2.391102961682665e-06,"reference_active":true,"delta_db":0.015229066129154789}],"active_band_count":3,"active_band_mae_db":0.08918580585292168}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00015152624573583172,"log_rms_mae_db":1.0264977062718745,"rms_envelope_nrmse":0.0712822087894152,"envelope_correlation":0.9964928611761037,"rms_gain_error_db":-0.21328580615723194,"reference_mean_power":0.001672933915423699,"output_mean_power":0.0015927595036271266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":1.623496524404657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006256720367801862,"output_power":0.0006112550215806211,"reference_active":true,"delta_db":-0.10124305485492043},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001572550903934534,"output_power":0.00015426728231948285,"reference_active":true,"delta_db":-0.08330883830534638},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4941700779965039e-05,"output_power":1.2263835048794587e-05,"reference_active":true,"delta_db":-0.8577373451865576}],"active_band_count":3,"active_band_mae_db":0.3474297461156082}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0001385613366624899,"log_rms_mae_db":0.6780656032739418,"rms_envelope_nrmse":0.050058555481171725,"envelope_correlation":0.9984280570675561,"rms_gain_error_db":-0.21328567945928603,"reference_mean_power":0.0016827742497549935,"output_mean_power":0.001602128292239969},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":1.5301470087365332e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005852793815004823,"output_power":0.0005721100329135717,"reference_active":true,"delta_db":-0.09883660933685867},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151264176170612e-05,"output_power":7.983411210497222e-05,"reference_active":true,"delta_db":-0.09036468974891519},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3333886770336671e-06,"output_power":1.0064908908956397e-06,"reference_active":true,"delta_db":-1.2214691411065073}],"active_band_count":3,"active_band_mae_db":0.4702234800640937}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":0.00030489717691300735,"log_rms_mae_db":0.9972135031239003,"rms_envelope_nrmse":0.08674008451881565,"envelope_correlation":0.9955858693184584,"rms_gain_error_db":-0.18900932330826795,"reference_mean_power":0.002589850567098419,"output_mean_power":0.002479555159496804},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":4.739471399174343e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014413951986732994,"output_power":0.0013594041506840098,"reference_active":true,"delta_db":-0.2543447945457223},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0004245147064825656,"output_power":0.00041197053819904257,"reference_active":true,"delta_db":-0.13026581263771353},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.389197181736509e-05,"output_power":2.5328052505491693e-05,"reference_active":true,"delta_db":0.253498040315437}],"active_band_count":3,"active_band_mae_db":0.2127028824996243}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":0.0002474836728453423,"log_rms_mae_db":0.6273453267269395,"rms_envelope_nrmse":0.05438277367945926,"envelope_correlation":0.9984053610313461,"rms_gain_error_db":-0.1890115696550375,"reference_mean_power":0.002605032676611575,"output_mean_power":0.0024940894100175374},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":4.638343817264176e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014427582782272214,"output_power":0.0013619784560246443,"reference_active":true,"delta_db":-0.2502333697616148},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020918188683306623,"output_power":0.0002008109895161629,"reference_active":true,"delta_db":-0.17736599860914087},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0082267328987973e-06,"output_power":2.2283689062444213e-06,"reference_active":true,"delta_db":0.45174345885751677}],"active_band_count":3,"active_band_mae_db":0.29311427574275745}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1026,"rms_floor":8.624763428459013e-05,"log_rms_mae_db":1.2110776793137144,"rms_envelope_nrmse":0.15963796467330718,"envelope_correlation":0.985045671388383,"rms_gain_error_db":0.47320885128853346,"reference_mean_power":0.00014375800668184295,"output_mean_power":0.00016030716266091414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1026,"frequency_resolution_hz":0.19493177387914232,"power_floor":6.640530401894145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.05643435973393e-05,"output_power":2.3277828306295516e-05,"reference_active":true,"delta_db":0.5382760895183069},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8444282129891133e-05,"output_power":2.131528374988145e-05,"reference_active":true,"delta_db":0.6282936165514508},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6443800103937953e-06,"output_power":6.486636707904305e-06,"reference_active":true,"delta_db":2.5039591956727096}],"active_band_count":3,"active_band_mae_db":1.2235096339141558}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":82160,"uncovered_tail_samples":0,"metrics":{"frame_count":1020,"rms_floor":6.581694648038092e-05,"log_rms_mae_db":0.8245527508858644,"rms_envelope_nrmse":0.10607060638986354,"envelope_correlation":0.9941584588060365,"rms_gain_error_db":0.4731886365229052,"reference_mean_power":0.0001445994240238279,"output_mean_power":0.00016124469187706506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1020,"frequency_resolution_hz":0.196078431372549,"power_floor":5.765239137842011e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0474191785118637e-05,"output_power":2.426539738545998e-05,"reference_active":true,"delta_db":0.7378064091244901},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1083544144974644e-05,"output_power":1.3076756890527694e-05,"reference_active":true,"delta_db":0.7182139447860516},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8424426027278453e-07,"output_power":4.482574556099785e-07,"reference_active":true,"delta_db":1.9783581790009492}],"active_band_count":3,"active_band_mae_db":1.1447928443038304}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"hnm","status":"measured","mos":{"status":"measured","value":2.24654483795166,"predicted_mos":2.24654483795166,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ed3a3e054046d4e06f44af9006bb812bf7fe83629a96cdf3130883526f089674","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0368700069993793},"started_at":"2026-10-13T04:39:43.893918+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5410999367304612e-05,"native_input_samples":96000,"native_input_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","output_sha256":"5548611de2f427e9cafa9bb3d4b70cbb0e5e13e56dd06161da95b4f97d303fd4","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9785815669993099,"synthesis_s":0.48814681700059737,"total_s":1.4667283839999072,"rtf":0.24445473066665122,"output_peak":0.7551253835662786,"output_rms":0.078889788070597,"finished_at":"2026-10-13T04:39:45.366872+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"5548611de2f427e9cafa9bb3d4b70cbb0e5e13e56dd06161da95b4f97d303fd4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.443585136553119,"spectral_convergence":0.2642987846010118,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4534368269474065,"spectral_convergence":0.26418996078438434,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.724884788493518,"spectral_convergence":0.27735183592589896,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":4.4113164781533575e-05,"delta_db":1.3751576779528558,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.0021464181801068008,"delta_db":-0.08198992432256802,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.001797624896172728,"delta_db":-0.14206212475630997,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.0019628825863377717,"delta_db":-0.13074350920924022,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.0002291727196787447,"delta_db":-0.04221296425530745,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":4.338711474613021e-05,"delta_db":0.14127582335784042,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.31890700397568694,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10312405546448383,"pitch":null,"output_rms_dbfs":-22.0595842125217,"output_peak":0.7551253835662786,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.540635583998014},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.443585136553119,"spectral_convergence":0.2642987846010118,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4534368269474065,"spectral_convergence":0.26418996078438434,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.724884788493518,"spectral_convergence":0.27735183592589896,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":4.4113164781533575e-05,"delta_db":1.3751576779528558,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.0021464181801068008,"delta_db":-0.08198992432256802,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.001797624896172728,"delta_db":-0.14206212475630997,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.0019628825863377717,"delta_db":-0.13074350920924022,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.0002291727196787447,"delta_db":-0.04221296425530745,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":4.338711474613021e-05,"delta_db":0.14127582335784042,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.31890700397568694,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10312405546448383,"pitch":null,"output_rms_dbfs":-22.0595842125217,"output_peak":0.7551253835662786,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.540635583998014},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8689430652650397,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0798319967885378,"rms_dbfs":-21.95646015705721,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.013818800436425,"mean":-1.4187812805175782e-05,"zero_energy":false},"output":{"rms":0.078889788070597,"rms_dbfs":-22.0595842125217,"peak":0.7551253835662786,"peak_dbfs":-2.4396186132468705,"crest_factor_db":19.619965599274828,"mean":2.8382643292843914e-05,"zero_energy":false},"rms_difference_db":-0.10312405546448744},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":4.613770208279608e-05,"output_rms_dbfs":-43.35944039585481,"output_zero_energy":false,"output_peak":0.17157105806329234,"output_energy_fraction":0.0009143129079187885,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":339,"output_estimated_voiced_frames":304,"both_estimated_voiced_frames":300,"reference_voiced_output_unvoiced_frames":39,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.8849557522123894,"absolute_difference_cents_median":1.3257986059559719,"absolute_difference_cents_p95":8.185723161761207,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003217977906320387,"log_rms_mae_db":0.808538570635909,"rms_envelope_nrmse":0.07724769665265463,"envelope_correlation":0.9947322837591313,"rms_gain_error_db":-0.10305817588845262,"reference_mean_power":0.006378025340653872,"output_mean_power":0.006228456316067608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.934428693189062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026409373563640536,"output_power":0.002557351679877502,"reference_active":true,"delta_db":-0.13967644664753606},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022353254716419522,"output_power":0.00019944004330567852,"reference_active":true,"delta_db":-0.495284071552518},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.171045723198584e-05,"output_power":2.8748110123896266e-05,"reference_active":true,"delta_db":-0.42593204332074636}],"active_band_count":3,"active_band_mae_db":0.3536308538402668}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00029448546471938644,"log_rms_mae_db":0.45799751908102626,"rms_envelope_nrmse":0.03770515226839406,"envelope_correlation":0.9987663649643889,"rms_gain_error_db":-0.10300496395866979,"reference_mean_power":0.006407121641291136,"output_mean_power":0.006256946951470594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7183045790225314e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025278777543949896,"output_power":0.0024442316148043674,"reference_active":true,"delta_db":-0.14613710915435354},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010431651037251879,"output_power":9.215354909709605e-05,"reference_active":true,"delta_db":-0.538409846789067},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4474914310652014e-06,"output_power":1.8189751738212591e-06,"reference_active":true,"delta_db":-1.288944083049167}],"active_band_count":3,"active_band_mae_db":0.6578303463308625}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002169046697405618,"log_rms_mae_db":1.2912949969071499,"rms_envelope_nrmse":0.07905545615143683,"envelope_correlation":0.9951379706568556,"rms_gain_error_db":-0.13554566251270508,"reference_mean_power":0.00266513165861851,"output_mean_power":0.002583236104610985},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9887000298118345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014141653900070267,"output_power":0.0013510571571260347,"reference_active":true,"delta_db":-0.19826481755534173},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010731466350302803,"output_power":8.819139136477737e-05,"reference_active":true,"delta_db":-0.8523287376004269},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0173327828420224e-05,"output_power":1.7054283113585492e-05,"reference_active":true,"delta_db":-0.7294407780741246}],"active_band_count":3,"active_band_mae_db":0.5933447777432977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020963310569053025,"log_rms_mae_db":0.8330877035242842,"rms_envelope_nrmse":0.04810534457856274,"envelope_correlation":0.9982140433966435,"rms_gain_error_db":-0.13557092214196428,"reference_mean_power":0.0026772178601939312,"output_mean_power":0.002594935822262785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.883939001913369e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013521293519634118,"output_power":0.001289189660353647,"reference_active":true,"delta_db":-0.2070142684192537},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.792726117783695e-05,"output_power":3.9021503176266066e-05,"reference_active":true,"delta_db":-0.8927861661155111},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.685082668516172e-06,"output_power":1.2956689178005246e-06,"reference_active":true,"delta_db":-1.1412717132626835}],"active_band_count":3,"active_band_mae_db":0.747024049265816}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002744641361828467,"log_rms_mae_db":1.2942885000981812,"rms_envelope_nrmse":0.116270982034705,"envelope_correlation":0.9906335851978245,"rms_gain_error_db":-0.13887935165590742,"reference_mean_power":0.0019381811578355007,"output_mean_power":0.0018771822199447274},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1196235236571852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006735796402683561,"output_power":0.0006608886172732302,"reference_active":true,"delta_db":-0.08260679709949578},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013823323993445422,"output_power":0.00013269835246670198,"reference_active":true,"delta_db":-0.17746956348942794},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8792319908524532e-05,"output_power":1.2018615905693638e-05,"reference_active":true,"delta_db":-1.9412594089425723}],"active_band_count":3,"active_band_mae_db":0.733778589843832}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002445180716567874,"log_rms_mae_db":0.7683822308257776,"rms_envelope_nrmse":0.0576467755868335,"envelope_correlation":0.9977143941833715,"rms_gain_error_db":-0.13863370544260814,"reference_mean_power":0.0019467823981069803,"output_mean_power":0.001885619411397012},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0571732642454034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000664982480329528,"output_power":0.0006489455255756458,"reference_active":true,"delta_db":-0.10601961148621827},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927173921912536e-05,"output_power":6.615575582244488e-05,"reference_active":true,"delta_db":-0.1998845611950712},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7481239585012511e-06,"output_power":1.079810654084867e-06,"reference_active":true,"delta_db":-2.092246167949709}],"active_band_count":3,"active_band_mae_db":0.7993834468769995}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":1.487460096753752,"rms_envelope_nrmse":0.20649353738399068,"envelope_correlation":0.9655514681083132,"rms_gain_error_db":0.19397716799999087,"reference_mean_power":6.564824297759047e-05,"output_mean_power":6.864688323987991e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.001422573299626e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.567613726314845e-05,"output_power":1.7159809680436297e-05,"reference_active":true,"delta_db":0.39273409191121855},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.796434791027312e-06,"output_power":6.538382585518373e-06,"reference_active":true,"delta_db":0.5230937471811312},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.958485457614675e-07,"output_power":7.683258705862412e-07,"reference_active":true,"delta_db":-1.1264783677707124}],"active_band_count":3,"active_band_mae_db":0.6807687356210207}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":0.9016466573957677,"rms_envelope_nrmse":0.1181295541888398,"envelope_correlation":0.9878151795032717,"rms_gain_error_db":0.19467807292321332,"reference_mean_power":6.585532886735057e-05,"output_mean_power":6.887454299109182e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6817222442955515e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.548077264503126e-05,"output_power":1.738295386078278e-05,"reference_active":true,"delta_db":0.5033094491507704},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9799921502867964e-06,"output_power":3.1594046565595137e-06,"reference_active":true,"delta_db":0.253901338206314},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.719812912358677e-08,"output_power":7.504260461781938e-08,"reference_active":true,"delta_db":-0.6519926769068278}],"active_band_count":3,"active_band_mae_db":0.46973448808797075}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.721027374267578,"predicted_mos":2.721027374267578,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bf48571a0bc6d8f9034039e24738481303c1b4143444146185d3e4f8b30cc473","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7904576059991086},"started_at":"2026-10-13T04:42:49.109023+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5026000002981164e-05,"native_input_samples":96000,"native_input_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","output_sha256":"65abac1d0df6319df3a601e6f8ced88df9406112677cf125e727811c759955ef","output_samples":96000,"length_delta_samples":0,"analysis_s":2.574252143999729,"synthesis_s":0.15889120799965895,"total_s":2.733143351999388,"rtf":0.455523891999898,"output_peak":0.9941412019993691,"output_rms":0.09543421399528676,"finished_at":"2026-10-13T04:42:51.850064+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"65abac1d0df6319df3a601e6f8ced88df9406112677cf125e727811c759955ef","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.697058389654536,"spectral_convergence":0.4467591403251173,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.569033188351573,"spectral_convergence":0.3426292701295577,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.991341248737145,"spectral_convergence":0.37654128732580877,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":3.29415385037931e-05,"delta_db":0.10691430223716054,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.003670328397228741,"delta_db":2.2479159327501885,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.0025551329939206453,"delta_db":1.3850822327731676,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.002531165808190005,"delta_db":0.9735192214251597,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.0002713820902017113,"delta_db":0.6919696796631426,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":4.6738372853290965e-05,"delta_db":0.4644040075607334,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9783008960682588,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5505421764026992,"pitch":null,"output_rms_dbfs":-20.405917980654515,"output_peak":0.9941412019993691,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.752477608914418},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.697058389654536,"spectral_convergence":0.4467591403251173,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.569033188351573,"spectral_convergence":0.3426292701295577,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.991341248737145,"spectral_convergence":0.37654128732580877,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":3.29415385037931e-05,"delta_db":0.10691430223716054,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.003670328397228741,"delta_db":2.2479159327501885,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.0025551329939206453,"delta_db":1.3850822327731676,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.002531165808190005,"delta_db":0.9735192214251597,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.0002713820902017113,"delta_db":0.6919696796631426,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":4.6738372853290965e-05,"delta_db":0.4644040075607334,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9783008960682588,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5505421764026992,"pitch":null,"output_rms_dbfs":-20.405917980654515,"output_peak":0.9941412019993691,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.752477608914418},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.89229575397091,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0798319967885378,"rms_dbfs":-21.95646015705721,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.013818800436425,"mean":-1.4187812805175782e-05,"zero_energy":false},"output":{"rms":0.09543421399528676,"rms_dbfs":-20.405917980654515,"peak":0.9941412019993691,"peak_dbfs":-0.051038531491243974,"crest_factor_db":20.354879449163267,"mean":2.291213702946909e-06,"zero_energy":false},"rms_difference_db":1.550542176402697},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":1.2433792302172923e-05,"output_rms_dbfs":-49.05396391493645,"output_zero_energy":false,"output_peak":0.10228755079247928,"output_energy_fraction":0.000168374328194038,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":339,"output_estimated_voiced_frames":388,"both_estimated_voiced_frames":327,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":61,"paired_voiced_fraction":0.9646017699115044,"absolute_difference_cents_median":3.500038119821023,"absolute_difference_cents_p95":15.546335833870517,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003217977906320387,"log_rms_mae_db":1.5319135546443512,"rms_envelope_nrmse":0.24732217596735256,"envelope_correlation":0.9831905414531128,"rms_gain_error_db":1.5505847955026528,"reference_mean_power":0.006378025340653872,"output_mean_power":0.00911474913257364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.934428693189062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026409373563640536,"output_power":0.004448287755594026,"reference_active":true,"delta_db":2.264347738234673},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022353254716419522,"output_power":0.00028384155464521914,"reference_active":true,"delta_db":1.0373520990582616},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.171045723198584e-05,"output_power":3.7309263397868964e-05,"reference_active":true,"delta_db":0.7061417063796428}],"active_band_count":3,"active_band_mae_db":1.3359471812241923}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00029448546471938644,"log_rms_mae_db":1.283497362903463,"rms_envelope_nrmse":0.232795386670899,"envelope_correlation":0.9876477121370572,"rms_gain_error_db":1.55059417748681,"reference_mean_power":0.006407121641291136,"output_mean_power":0.009156350039737145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7183045790225314e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025278777543949896,"output_power":0.0042613973148679955,"reference_active":true,"delta_db":2.2679595981247918},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010431651037251879,"output_power":0.00013223769897079723,"reference_active":true,"delta_db":1.0300223311856174},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4474914310652014e-06,"output_power":2.4360615096279177e-06,"reference_active":true,"delta_db":-0.02032930095418655}],"active_band_count":3,"active_band_mae_db":1.106103743421532}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002169046697405618,"log_rms_mae_db":1.6980686663639357,"rms_envelope_nrmse":0.28747947000132745,"envelope_correlation":0.9837480002451593,"rms_gain_error_db":1.8495059621921597,"reference_mean_power":0.00266513165861851,"output_mean_power":0.004080085503966375},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9887000298118345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014141653900070267,"output_power":0.0025504546997426256,"reference_active":true,"delta_db":2.561174099494318},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010731466350302803,"output_power":0.0001518671872950097,"reference_active":true,"delta_db":1.5080488140093327},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0173327828420224e-05,"output_power":2.166687271510266e-05,"reference_active":true,"delta_db":0.31018685891835324}],"active_band_count":3,"active_band_mae_db":1.4598032574740012}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020963310569053025,"log_rms_mae_db":1.4215113008699083,"rms_envelope_nrmse":0.27745764278577073,"envelope_correlation":0.9865899197179271,"rms_gain_error_db":1.849265850305199,"reference_mean_power":0.0026772178601939312,"output_mean_power":0.004098361835017108},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.883939001913369e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013521293519634118,"output_power":0.0024353555414889424,"reference_active":true,"delta_db":2.555441329940116},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.792726117783695e-05,"output_power":6.681956507829591e-05,"reference_active":true,"delta_db":1.4432103276087649},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.685082668516172e-06,"output_power":1.697137382277771e-06,"reference_active":true,"delta_db":0.0309578783007098}],"active_band_count":3,"active_band_mae_db":1.3432031786165302}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002744641361828467,"log_rms_mae_db":1.4700318417673996,"rms_envelope_nrmse":0.17990152041234556,"envelope_correlation":0.9894970744707445,"rms_gain_error_db":1.01534506528815,"reference_mean_power":0.0019381811578355007,"output_mean_power":0.002448662181260952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1196235236571852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006735796402683561,"output_power":0.00092817446087913,"reference_active":true,"delta_db":1.392406628154788},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013823323993445422,"output_power":0.00016618809043928863,"reference_active":true,"delta_db":0.7998741044773696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8792319908524532e-05,"output_power":2.0683789423868812e-05,"reference_active":true,"delta_db":0.4164971070046587}],"active_band_count":3,"active_band_mae_db":0.8695926132122721}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002445180716567874,"log_rms_mae_db":1.1856509445336798,"rms_envelope_nrmse":0.1519106539213926,"envelope_correlation":0.9949550692148174,"rms_gain_error_db":1.0153625549717973,"reference_mean_power":0.0019467823981069803,"output_mean_power":0.002459538733839663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0571732642454034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000664982480329528,"output_power":0.0009128053892654165,"reference_active":true,"delta_db":1.3756799201808902},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927173921912536e-05,"output_power":8.086620796538825e-05,"reference_active":true,"delta_db":0.6721098665806348},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7481239585012511e-06,"output_power":1.8554694824586829e-06,"reference_active":true,"delta_db":0.25881590782279645}],"active_band_count":3,"active_band_mae_db":0.7688685648614405}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":1.4044125235858969,"rms_envelope_nrmse":0.19029105386755407,"envelope_correlation":0.9740971564660318,"rms_gain_error_db":0.5019943636754448,"reference_mean_power":6.564824297759047e-05,"output_mean_power":7.369237329755815e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.001422573299626e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.567613726314845e-05,"output_power":1.7232590740604322e-05,"reference_active":true,"delta_db":0.41111516424989863},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.796434791027312e-06,"output_power":6.2378604112420605e-06,"reference_active":true,"delta_db":0.31874697548867925},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.958485457614675e-07,"output_power":1.2131105012360126e-06,"reference_active":true,"delta_db":0.8570706869248366}],"active_band_count":3,"active_band_mae_db":0.5289776088878048}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":1.0148002162596788,"rms_envelope_nrmse":0.11337337590546166,"envelope_correlation":0.9917560080101838,"rms_gain_error_db":0.5014176238092964,"reference_mean_power":6.585532886735057e-05,"output_mean_power":7.391501770191728e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6817222442955515e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.548077264503126e-05,"output_power":1.7026941234654774e-05,"reference_active":true,"delta_db":0.4134400462895354},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9799921502867964e-06,"output_power":3.15906931882209e-06,"reference_active":true,"delta_db":0.25344035560310546},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.719812912358677e-08,"output_power":9.341074532378444e-08,"reference_active":true,"delta_db":0.2988967028408521}],"active_band_count":3,"active_band_mae_db":0.321925701577831}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"facodec","status":"measured","mos":{"status":"measured","value":4.109444618225098,"predicted_mos":4.109444618225098,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"22d45ad24e97412a41e00f6f2d1096575a8eb2ce26f825819c86c6fec5b609f7","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4262268519996724},"started_at":"2026-10-13T04:55:36.342721+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.4425000194460154e-05,"native_input_samples":96000,"native_input_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","output_sha256":"59e14e30d01916aa4fc7f6b74df9495cc579d63099b6a4aff393d5f075701c9a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6176753689996985,"synthesis_s":51.73666363600023,"total_s":53.35433900499993,"rtf":8.892389834166655,"output_peak":0.7532517313957214,"output_rms":0.07235791645907087,"finished_at":"2026-10-13T04:56:29.724638+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"59e14e30d01916aa4fc7f6b74df9495cc579d63099b6a4aff393d5f075701c9a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.522008028502885,"spectral_convergence":0.2624754193501107,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.076501216408956,"spectral_convergence":0.28358773482383526,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.175211836598572,"spectral_convergence":0.31928431810496394,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":2.3172107951551365e-05,"delta_db":-1.4208690532898494,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.0015884524196668607,"delta_db":-1.3893912002707594,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.001742363152318934,"delta_db":-0.27766608678773375,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.0016462098264481292,"delta_db":-0.8948348340130581,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.00019453778025289734,"delta_db":-0.7538025903007256,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":4.093278765950689e-05,"delta_db":-0.11161865506421056,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8080304032877228,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8538187851452655,"pitch":null,"output_rms_dbfs":-22.81027894220248,"output_peak":0.7532517313957214,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9245736938368045},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.522008028502885,"spectral_convergence":0.2624754193501107,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.076501216408956,"spectral_convergence":0.28358773482383526,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.175211836598572,"spectral_convergence":0.31928431810496394,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":2.3172107951551365e-05,"delta_db":-1.4208690532898494,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.0015884524196668607,"delta_db":-1.3893912002707594,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.001742363152318934,"delta_db":-0.27766608678773375,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.0016462098264481292,"delta_db":-0.8948348340130581,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.00019453778025289734,"delta_db":-0.7538025903007256,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":4.093278765950689e-05,"delta_db":-0.11161865506421056,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8080304032877228,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8538187851452655,"pitch":null,"output_rms_dbfs":-22.81027894220248,"output_peak":0.7532517313957214,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9245736938368045},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8603199877843658,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0798319967885378,"rms_dbfs":-21.95646015705721,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.013818800436425,"mean":-1.4187812805175782e-05,"zero_energy":false},"output":{"rms":0.07235791645907087,"rms_dbfs":-22.81027894220248,"peak":0.7532517313957214,"peak_dbfs":-2.461197228014279,"crest_factor_db":20.349081714188202,"mean":7.897068615962747e-05,"zero_energy":false},"rms_difference_db":-0.8538187851452683},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":4.121378963457765e-06,"output_rms_dbfs":-53.84957449981207,"output_zero_energy":false,"output_peak":0.017083758488297462,"output_energy_fraction":9.708472696892449e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":339,"output_estimated_voiced_frames":347,"both_estimated_voiced_frames":307,"reference_voiced_output_unvoiced_frames":32,"reference_unvoiced_output_voiced_frames":40,"paired_voiced_fraction":0.9056047197640118,"absolute_difference_cents_median":4.534859702437758,"absolute_difference_cents_p95":32.8204204504553,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003217977906320387,"log_rms_mae_db":1.4754884784158149,"rms_envelope_nrmse":0.14262888115399922,"envelope_correlation":0.9886346521056645,"rms_gain_error_db":-0.8541078405161722,"reference_mean_power":0.006378025340653872,"output_mean_power":0.005239326416373861},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.934428693189062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026409373563640536,"output_power":0.0020877534885128395,"reference_active":true,"delta_db":-1.0207888172733055},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022353254716419522,"output_power":0.0001897510820226897,"reference_active":true,"delta_db":-0.7115650588842428},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.171045723198584e-05,"output_power":2.5911337916320942e-05,"reference_active":true,"delta_db":-0.8771266605066528}],"active_band_count":3,"active_band_mae_db":0.8698268455547337}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00029448546471938644,"log_rms_mae_db":1.279595329376156,"rms_envelope_nrmse":0.13107558789064572,"envelope_correlation":0.9912354365423622,"rms_gain_error_db":-0.8553601305200876,"reference_mean_power":0.006407121641291136,"output_mean_power":0.005261710582052498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7183045790225314e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025278777543949896,"output_power":0.0019970343127213006,"reference_active":true,"delta_db":-1.0237054116485178},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010431651037251879,"output_power":8.847222230732958e-05,"reference_active":true,"delta_db":-0.715461141585127},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4474914310652014e-06,"output_power":2.0516634285512275e-06,"reference_active":true,"delta_db":-0.7661506285891112}],"active_band_count":3,"active_band_mae_db":0.8351057272742519}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002169046697405618,"log_rms_mae_db":1.7915035405981186,"rms_envelope_nrmse":0.14896970352275152,"envelope_correlation":0.9872276966962733,"rms_gain_error_db":-0.7949018605313724,"reference_mean_power":0.00266513165861851,"output_mean_power":0.0022193637191580775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9887000298118345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014141653900070267,"output_power":0.0010990959742921461,"reference_active":true,"delta_db":-1.0946458704400945},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010731466350302803,"output_power":8.089859159862844e-05,"reference_active":true,"delta_db":-1.227181072869244},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0173327828420224e-05,"output_power":1.279912683333407e-05,"reference_active":true,"delta_db":-1.9759720337923974}],"active_band_count":3,"active_band_mae_db":1.4325996590339123}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020963310569053025,"log_rms_mae_db":1.507592433804246,"rms_envelope_nrmse":0.13811257200839577,"envelope_correlation":0.9893209515437752,"rms_gain_error_db":-0.7972971630491248,"reference_mean_power":0.0026772178601939312,"output_mean_power":0.002228199114442377},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.883939001913369e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013521293519634118,"output_power":0.0010465095741325307,"reference_active":true,"delta_db":-1.112750346442115},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.792726117783695e-05,"output_power":3.609092333909421e-05,"reference_active":true,"delta_db":-1.231846187139107},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.685082668516172e-06,"output_power":1.0370882506955986e-06,"reference_active":true,"delta_db":-2.1080549768208483}],"active_band_count":3,"active_band_mae_db":1.4842171701340234}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002744641361828467,"log_rms_mae_db":1.7984125332005045,"rms_envelope_nrmse":0.1688484653023192,"envelope_correlation":0.9853207661334555,"rms_gain_error_db":-0.8897853487165,"reference_mean_power":0.0019381811578355007,"output_mean_power":0.001579122539149148},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1196235236571852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006735796402683561,"output_power":0.0005960481925913003,"reference_active":true,"delta_db":-0.5310757623668682},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013823323993445422,"output_power":0.00013223609619407233,"reference_active":true,"delta_db":-0.19262467473780287},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8792319908524532e-05,"output_power":1.742804691779612e-05,"reference_active":true,"delta_db":-0.3273167656278271}],"active_band_count":3,"active_band_mae_db":0.3503390675774994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002445180716567874,"log_rms_mae_db":1.5382724794389007,"rms_envelope_nrmse":0.15000428877420863,"envelope_correlation":0.9892992685061337,"rms_gain_error_db":-0.889335442347525,"reference_mean_power":0.0019467823981069803,"output_mean_power":0.0015862946761210524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0571732642454034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000664982480329528,"output_power":0.0005953032282210432,"reference_active":true,"delta_db":-0.48071965845888054},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927173921912536e-05,"output_power":6.327219834400866e-05,"reference_active":true,"delta_db":-0.3934316774034575},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7481239585012511e-06,"output_power":1.5581345218956678e-06,"reference_active":true,"delta_db":-0.4996727510401746}],"active_band_count":3,"active_band_mae_db":0.45794136230083754}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":1.9854438738024283,"rms_envelope_nrmse":0.23758841551690882,"envelope_correlation":0.9517883422539698,"rms_gain_error_db":-0.3655951902669359,"reference_mean_power":6.564824297759047e-05,"output_mean_power":6.034809815101637e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.001422573299626e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.567613726314845e-05,"output_power":1.3089283607175914e-05,"reference_active":true,"delta_db":-0.7832317986296894},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.796434791027312e-06,"output_power":6.170228312515937e-06,"reference_active":true,"delta_db":0.2714027973254427},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.958485457614675e-07,"output_power":1.8656214838925579e-06,"reference_active":true,"delta_db":2.726302407998213}],"active_band_count":3,"active_band_mae_db":1.260312334651115}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":1.717895783934334,"rms_envelope_nrmse":0.18917875995273592,"envelope_correlation":0.9666355504055312,"rms_gain_error_db":-0.36809303597486515,"reference_mean_power":6.585532886735057e-05,"output_mean_power":6.050365615357718e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6817222442955515e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.548077264503126e-05,"output_power":1.3218162205085571e-05,"reference_active":true,"delta_db":-0.686215555565586},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9799921502867964e-06,"output_power":3.4153163518836695e-06,"reference_active":true,"delta_db":0.5921581736491637},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.719812912358677e-08,"output_power":1.3877993293659267e-07,"reference_active":true,"delta_db":2.0181950624530742}],"active_band_count":3,"active_band_mae_db":1.0988562638892747}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.2978641986846924,"predicted_mos":3.2978641986846924,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"17e9449ff97fedd3c44e95961091c4afece038676a922da37c6aaf62055ffa12","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.680161140000564},"started_at":"2026-10-13T07:21:55.601321+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.011515185000462225,"native_input_samples":264600,"native_input_sha256":"4d416eef22a33e5b61b80ce629b8e9c42662680a477f2a584b95b16b4accf595","output_sha256":"a924a718645d3a9f2b2712995358dfef26cca2d2e1f8720ddd3fe21261a5dd11","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1946583399985684,"synthesis_s":4.117670409999846,"total_s":4.312328749998414,"rtf":0.7187214583330691,"output_peak":0.7995373010635376,"output_rms":0.07793193060784188,"finished_at":"2026-10-13T07:21:59.960675+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"a924a718645d3a9f2b2712995358dfef26cca2d2e1f8720ddd3fe21261a5dd11","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.159416116781227,"output_peak":0.7988954341458568},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.372032407246607,"spectral_convergence":0.40553146740812557,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.579114685355155,"spectral_convergence":0.38334914731654457,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.241322548023712,"spectral_convergence":0.41040184643355393,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":1.9505859312225665e-05,"delta_db":-2.1688736089905913,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.0018805784206183543,"delta_db":-0.6562188224383393,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.002179117536359569,"delta_db":0.6937536899623608,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.0018439488644932179,"delta_db":-0.4021979988571893,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.0001366600528353352,"delta_db":-2.287426308461194,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":2.455362325949762e-05,"delta_db":-2.331176001377803,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4232744050145794,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20138770786497798,"pitch":null,"output_rms_dbfs":-22.15784786492219,"output_peak":0.7988954341458568,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.397489880208491},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7722531434778108,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0798319967885378,"rms_dbfs":-21.95646015705721,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.013818800436425,"mean":-1.4187812805175782e-05,"zero_energy":false},"output":{"rms":0.0780023355860464,"rms_dbfs":-22.15784786492219,"peak":0.7988954341458568,"peak_dbfs":-1.9502012183581796,"crest_factor_db":20.20764664656401,"mean":-7.611133281338317e-05,"zero_energy":false},"rms_difference_db":-0.20138770786497773},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":2.199366553157556e-06,"output_rms_dbfs":-56.57702383760083,"output_zero_energy":false,"output_peak":0.013220355923708387,"output_energy_fraction":4.458234127877614e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":339,"output_estimated_voiced_frames":349,"both_estimated_voiced_frames":313,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.9233038348082596,"absolute_difference_cents_median":5.528999232169269,"absolute_difference_cents_p95":22.669457584610253,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003217977906320387,"log_rms_mae_db":1.4662597831693513,"rms_envelope_nrmse":0.1185565760976328,"envelope_correlation":0.988276398612585,"rms_gain_error_db":-0.20135451083985856,"reference_mean_power":0.006378025340653872,"output_mean_power":0.00608906751265899},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.934428693189062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026409373563640536,"output_power":0.0027425876432831727,"reference_active":true,"delta_db":0.1640241520044653},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022353254716419522,"output_power":0.00022663802105245194,"reference_active":true,"delta_db":0.05992002424075552},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.171045723198584e-05,"output_power":3.08878936039122e-05,"reference_active":true,"delta_db":-0.11414211496621036}],"active_band_count":3,"active_band_mae_db":0.1126954304038104}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00029448546471938644,"log_rms_mae_db":1.0845223309927519,"rms_envelope_nrmse":0.08693717541492801,"envelope_correlation":0.9936893607888194,"rms_gain_error_db":-0.2012235549526073,"reference_mean_power":0.006407121641291136,"output_mean_power":0.006117030047408231},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7183045790225314e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025278777543949896,"output_power":0.002623728549690521,"reference_active":true,"delta_db":0.16162832950299666},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010431651037251879,"output_power":0.00010436923036006551,"reference_active":true,"delta_db":0.002194304228382071},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4474914310652014e-06,"output_power":2.0649935593467755e-06,"reference_active":true,"delta_db":-0.7380247852240861}],"active_band_count":3,"active_band_mae_db":0.3006158063184883}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002169046697405618,"log_rms_mae_db":2.3051185166507158,"rms_envelope_nrmse":0.13637377923538804,"envelope_correlation":0.9850590044492895,"rms_gain_error_db":-0.09206483453401113,"reference_mean_power":0.00266513165861851,"output_mean_power":0.0026092289277286863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9887000298118345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014141653900070267,"output_power":0.0014290028044794288,"reference_active":true,"delta_db":0.04532876909216177},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010731466350302803,"output_power":9.489532088233225e-05,"reference_active":true,"delta_db":-0.5341426945439519},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0173327828420224e-05,"output_power":1.5140402269132356e-05,"reference_active":true,"delta_db":-1.2464013192490133}],"active_band_count":3,"active_band_mae_db":0.608624260961709}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020963310569053025,"log_rms_mae_db":1.7137395556403217,"rms_envelope_nrmse":0.10494562026722726,"envelope_correlation":0.9905728784651925,"rms_gain_error_db":-0.09263527950973618,"reference_mean_power":0.0026772178601939312,"output_mean_power":0.002620717360660751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.883939001913369e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013521293519634118,"output_power":0.001360087646088961,"reference_active":true,"delta_db":0.02548655322022604},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.792726117783695e-05,"output_power":4.0450958449510845e-05,"reference_active":true,"delta_db":-0.7365379548015806},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.685082668516172e-06,"output_power":1.2703719885724147e-06,"reference_active":true,"delta_db":-1.2269030270552541}],"active_band_count":3,"active_band_mae_db":0.662975845025687}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002744641361828467,"log_rms_mae_db":2.9904848333186433,"rms_envelope_nrmse":0.22241626896146569,"envelope_correlation":0.9730322588400069,"rms_gain_error_db":-1.083625534868077,"reference_mean_power":0.0019381811578355007,"output_mean_power":0.0015101907774036738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1196235236571852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006735796402683561,"output_power":0.0005393840762938481,"reference_active":true,"delta_db":-0.9649083057501168},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013823323993445422,"output_power":0.00012716530773689463,"reference_active":true,"delta_db":-0.3624384062421566},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8792319908524532e-05,"output_power":1.9409263570268645e-05,"reference_active":true,"delta_db":0.14028660636424325}],"active_band_count":3,"active_band_mae_db":0.4892111061188389}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002445180716567874,"log_rms_mae_db":2.5271104931504946,"rms_envelope_nrmse":0.19141437464494004,"envelope_correlation":0.9819228495445048,"rms_gain_error_db":-1.08336068740806,"reference_mean_power":0.0019467823981069803,"output_mean_power":0.0015169851944627493},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0571732642454034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000664982480329528,"output_power":0.0005360442625610383,"reference_active":true,"delta_db":-0.9360955151202424},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927173921912536e-05,"output_power":6.845896379156274e-05,"reference_active":true,"delta_db":-0.05125770406052245},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7481239585012511e-06,"output_power":1.5132833294718172e-06,"reference_active":true,"delta_db":-0.6265197711579322}],"active_band_count":3,"active_band_mae_db":0.5379576634462323}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":4.294940870422815,"rms_envelope_nrmse":0.32602436281003183,"envelope_correlation":0.9291829508148918,"rms_gain_error_db":-1.8680477521830139,"reference_mean_power":6.564824297759047e-05,"output_mean_power":4.2699061718763074e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.001422573299626e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.567613726314845e-05,"output_power":1.1259615669155846e-05,"reference_active":true,"delta_db":-1.4371549083668929},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.796434791027312e-06,"output_power":4.3631945406692535e-06,"reference_active":true,"delta_db":-1.233563772855433},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.958485457614675e-07,"output_power":7.875856220730948e-07,"reference_active":true,"delta_db":-1.0189551431379995}],"active_band_count":3,"active_band_mae_db":1.229891274786775}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":3.782179512968459,"rms_envelope_nrmse":0.29368255631779144,"envelope_correlation":0.9436151660597947,"rms_gain_error_db":-1.8638736924305288,"reference_mean_power":6.585532886735057e-05,"output_mean_power":4.2874942797109266e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6817222442955515e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.548077264503126e-05,"output_power":1.1157307583405797e-05,"reference_active":true,"delta_db":-1.422332266907982},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9799921502867964e-06,"output_power":2.2243019764739107e-06,"reference_active":true,"delta_db":-1.2702137233156516},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.719812912358677e-08,"output_power":6.692354097639184e-08,"reference_active":true,"delta_db":-1.1492825528670818}],"active_band_count":3,"active_band_mae_db":1.2806095143635718}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"golf","status":"measured","mos":{"status":"measured","value":2.8086016178131104,"predicted_mos":2.8086016178131104,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"ca6ce7b18beb8a7194db33fbd141cb24f3e1dfcebfc5562530439f6f1df0b5df","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1349305769999773},"started_at":"2026-10-13T05:02:47.102875+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005136818999744719,"native_input_samples":144000,"native_input_sha256":"cd7b4b7f559ae268b6c2b9b6024e77abd0bd35c06865bf41574a87f595f7c12c","output_sha256":"795bcd7890ab36b1125cc877c2761e18126f2c608a481dcf0663ab49c2f2ce1e","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.767699920999803,"synthesis_s":0.13968674399984593,"total_s":0.9073866649996489,"rtf":0.1512311108332748,"output_peak":1.0511112213134766,"output_rms":0.07356322707293345,"finished_at":"2026-10-13T05:02:48.026629+00:00","analysis_stages":{"external_f0_s":0.12392412300050637},"evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"795bcd7890ab36b1125cc877c2761e18126f2c608a481dcf0663ab49c2f2ce1e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.661921575984316,"output_peak":1.0516062868702298},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.119188005442241,"spectral_convergence":0.45622701354391565,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.769404006795237,"spectral_convergence":0.32854884727846706,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.8049053461292806,"spectral_convergence":0.354450250126128,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2196813462216405e-05,"output_power":2.8768641108607682e-05,"delta_db":-0.48893543034372383,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0021897767639359146,"output_power":0.0021366194893049834,"delta_db":-0.10672657570593497,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018604852153017329,"output_power":0.0012900375859982343,"delta_db":-1.590238592610094,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0020261989943474246,"output_power":0.0017821756251362381,"delta_db":-0.5573159595278623,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002311682557361608,"output_power":0.00015305173317065163,"delta_db":-1.7908994412738104,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.205784090713682e-05,"output_power":2.6958226836269368e-05,"delta_db":-1.9315565055897848,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0776120841752017,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7114105568042937,"pitch":null,"output_rms_dbfs":-22.661921575984316,"output_peak":1.0516062868702298,"reference_rms_dbfs":-21.950511019180023,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.897832452788919},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8495038415683254,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07988669403405417,"rms_dbfs":-21.950511019180023,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.007869662559237,"mean":-1.3964403053754957e-05,"zero_energy":false},"output":{"rms":0.07360442446996637,"rms_dbfs":-22.661921575984316,"peak":1.0516062868702298,"peak_dbfs":0.43706347612826224,"crest_factor_db":23.098985052112578,"mean":0.00021958524633425825,"zero_energy":false},"rms_difference_db":-0.7114105568042923},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":2,"at_or_above_full_scale_fraction":2.0868113522537563e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":2.0868113522537563e-05,"longest_full_scale_run_samples":1,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":4.144663981590634e-06,"output_rms_dbfs":-53.825106730404485,"output_zero_energy":false,"output_peak":0.030152949009092267,"output_energy_fraction":9.451187560433181e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":338,"output_estimated_voiced_frames":322,"both_estimated_voiced_frames":294,"reference_voiced_output_unvoiced_frames":44,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.8698224852071006,"absolute_difference_cents_median":5.127187192166025,"absolute_difference_cents_p95":21.233113878725888,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0003217977906320387,"log_rms_mae_db":2.2542728485593635,"rms_envelope_nrmse":0.21737764911144195,"envelope_correlation":0.9636981917650952,"rms_gain_error_db":-0.7115134905050482,"reference_mean_power":0.006386691431487493,"output_mean_power":0.005421563951569433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.932821789520232e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026399066000462506,"output_power":0.002398943190925794,"reference_active":true,"delta_db":-0.41568598177091687},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231583888695557,"output_power":0.0002235637188614185,"reference_active":true,"delta_db":0.007881079717211015},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141776554756477e-05,"output_power":3.6341836847147124e-05,"reference_active":true,"delta_db":0.632315798428156}],"active_band_count":3,"active_band_mae_db":0.351960953305428}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00029448546471938644,"log_rms_mae_db":2.053968305872051,"rms_envelope_nrmse":0.19060120154187335,"envelope_correlation":0.9723103591557813,"rms_gain_error_db":-0.7116422866786604,"reference_mean_power":0.006415710203195347,"output_mean_power":0.005446036027458776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.7167701761502867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025267017543213234,"output_power":0.0023205100701963344,"reference_active":true,"delta_db":-0.369705244453506},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010394624185061282,"output_power":0.00010031651069719575,"reference_active":true,"delta_db":-0.1543637443504895},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4292380634974333e-06,"output_power":2.39946127886177e-06,"reference_active":true,"delta_db":-0.05356331494058158}],"active_band_count":3,"active_band_mae_db":0.19254410124819235}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002169046697405618,"log_rms_mae_db":2.289048544113829,"rms_envelope_nrmse":0.22269392864993584,"envelope_correlation":0.9641845551466968,"rms_gain_error_db":-0.6374937698330685,"reference_mean_power":0.002668565027510296,"output_mean_power":0.002304243723316509},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.989513456930054e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014140558775643387,"output_power":0.001374821709357639,"reference_active":true,"delta_db":-0.12220189977765886},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010734684470986247,"output_power":0.00011408027058444732,"reference_active":true,"delta_db":0.2642125903958326},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0015362627903167e-05,"output_power":1.6657452059412866e-05,"reference_active":true,"delta_db":-0.7975489080238474}],"active_band_count":3,"active_band_mae_db":0.39465446606577964}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00020963310569053025,"log_rms_mae_db":1.8828465726761683,"rms_envelope_nrmse":0.19579580003393238,"envelope_correlation":0.9719024412485655,"rms_gain_error_db":-0.6378421077763125,"reference_mean_power":0.0026810274253976155,"output_mean_power":0.002314819039584496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8846642066290474e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013519373358716119,"output_power":0.0013218657925017354,"reference_active":true,"delta_db":-0.09769197960476492},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.7759625960080496e-05,"output_power":4.9974497498414636e-05,"reference_active":true,"delta_db":0.19687519048712368},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.68381696272099e-06,"output_power":1.1878215540893503e-06,"reference_active":true,"delta_db":-1.5154367859879658}],"active_band_count":3,"active_band_mae_db":0.6033346520266182}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002744641361828467,"log_rms_mae_db":2.821252532470716,"rms_envelope_nrmse":0.2680487051575992,"envelope_correlation":0.9558495843652686,"rms_gain_error_db":-0.8005885141289272,"reference_mean_power":0.0019407425527460374,"output_mean_power":0.0016140206130352862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1163233704495926e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006727204982975257,"output_power":0.0005630822180177105,"reference_active":true,"delta_db":-0.7726284850589953},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013765991470112172,"output_power":0.00011021641000775709,"reference_active":true,"delta_db":-0.9656123512579771},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8671154576785563e-05,"output_power":1.7633296302103298e-05,"reference_active":true,"delta_db":-0.24837669181511818}],"active_band_count":3,"active_band_mae_db":0.6622058427106968}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002445180716567874,"log_rms_mae_db":2.5303825489474647,"rms_envelope_nrmse":0.23319597716574647,"envelope_correlation":0.9674342074087687,"rms_gain_error_db":-0.7999646318897001,"reference_mean_power":0.001949060376774701,"output_mean_power":0.0016211710115451865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0539791697157927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006640756663058943,"output_power":0.0005595343187378206,"reference_active":true,"delta_db":-0.7439083779688693},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.892990231280358e-05,"output_power":5.4339774411647285e-05,"reference_active":true,"delta_db":-1.0328983187631475},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.732756112376923e-06,"output_power":1.4744272480019875e-06,"reference_active":true,"delta_db":-0.7011409129566035}],"active_band_count":3,"active_band_mae_db":0.8259825365628735}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":2.72417430455067,"rms_envelope_nrmse":0.24122335343686366,"envelope_correlation":0.9681115685789626,"rms_gain_error_db":-1.4629049509571777,"reference_mean_power":6.571934982681283e-05,"output_mean_power":4.692483596521659e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.001987642584391e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5672935644135047e-05,"output_power":1.0196283094394467e-05,"reference_active":true,"delta_db":-1.8670846538579255},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7944900363763546e-06,"output_power":4.473641472542428e-06,"reference_active":true,"delta_db":-1.1235404459076412},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.937759392888437e-07,"output_power":9.358428449686451e-07,"reference_active":true,"delta_db":-0.2608555337802159}],"active_band_count":3,"active_band_mae_db":1.0838268778485942}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":2.4649424287192687,"rms_envelope_nrmse":0.20748561787395342,"envelope_correlation":0.9805730160667446,"rms_gain_error_db":-1.4655405855568207,"reference_mean_power":6.586240798407476e-05,"output_mean_power":4.6998451196214044e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.6820698528098298e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.547232411509048e-05,"output_power":1.0215194489536702e-05,"reference_active":true,"delta_db":-1.803089147748066},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9785303189771038e-06,"output_power":2.3093212899488e-06,"reference_active":true,"delta_db":-1.1051766581799536},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.680441525335973e-08,"output_power":8.830257855228382e-08,"reference_active":true,"delta_db":0.07431569876140401}],"active_band_count":3,"active_band_mae_db":0.9941938348964746}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.0963449478149414,"predicted_mos":3.0963449478149414,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"e34beee64340d0241cb445e5d75af138a02071df616bed3331cedce6a70ee680","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9703639409999596},"started_at":"2026-10-13T07:41:33.994378+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026466589988558553,"native_input_samples":144000,"native_input_sha256":"cd7b4b7f559ae268b6c2b9b6024e77abd0bd35c06865bf41574a87f595f7c12c","output_sha256":"b8d5be6bba704d8b4d6ec0246acfee784ed7f7ad91b0f83470aef37432e57486","output_samples":144120,"length_delta_samples":120,"analysis_s":2.409870403000241,"synthesis_s":13.690614586001175,"total_s":16.100484989001416,"rtf":2.6834141648335694,"output_peak":0.7493392825126648,"output_rms":0.07462932409212827,"finished_at":"2026-10-13T07:41:50.104444+00:00","analysis_stages":{"external_f0_s":1.6420421080001688},"evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"b8d5be6bba704d8b4d6ec0246acfee784ed7f7ad91b0f83470aef37432e57486","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.53711328032223,"output_peak":0.7289532429377811},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.737838535565002,"spectral_convergence":0.41370830297892375,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.466174230887403,"spectral_convergence":0.2969537306857816,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.330302614250272,"spectral_convergence":0.3049029175214565,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":2.912704889828637e-05,"delta_db":-0.4275596350353808,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.0020963646560869477,"delta_db":-0.184465015227933,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.001574011732186594,"delta_db":-0.7189732155265616,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.0016518822538408935,"delta_db":-0.8798958554472627,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.0001978037895205481,"delta_db":-0.6814960759442718,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":3.000908580167695e-05,"delta_db":-1.4598042280604004,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.725365670873635,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.577821658613343,"pitch":null,"output_rms_dbfs":-22.534281815670557,"output_peak":0.7289532429377811,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.511438460234225},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8519444317277161,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0798319967885378,"rms_dbfs":-21.95646015705721,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.013818800436425,"mean":-1.4187812805175782e-05,"zero_energy":false},"output":{"rms":0.07469403300354686,"rms_dbfs":-22.534281815670557,"peak":0.7289532429377811,"peak_dbfs":-2.7460065525801447,"crest_factor_db":19.788275263090412,"mean":2.5029533699195925e-06,"zero_energy":false},"rms_difference_db":-0.5778216586133453},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":3.069059129863551e-05,"output_rms_dbfs":-45.129947441761686,"output_zero_energy":false,"output_peak":0.1367120107655187,"output_energy_fraction":0.0006784438449048149,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":339,"output_estimated_voiced_frames":328,"both_estimated_voiced_frames":300,"reference_voiced_output_unvoiced_frames":39,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.8849557522123894,"absolute_difference_cents_median":5.95273837607545,"absolute_difference_cents_p95":31.656976836425574,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003217977906320387,"log_rms_mae_db":1.9159943023419563,"rms_envelope_nrmse":0.1681708756409988,"envelope_correlation":0.9768552354559498,"rms_gain_error_db":-0.5776813339745456,"reference_mean_power":0.006378025340653872,"output_mean_power":0.005583648964183663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.934428693189062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026409373563640536,"output_power":0.002319891503034761,"reference_active":true,"delta_db":-0.562904255358136},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022353254716419522,"output_power":0.00019257564827072828,"reference_active":true,"delta_db":-0.6473939844337708},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.171045723198584e-05,"output_power":2.2456933440482607e-05,"reference_active":true,"delta_db":-1.4985205234868815}],"active_band_count":3,"active_band_mae_db":0.9029395877595962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00029448546471938644,"log_rms_mae_db":1.3900616328916526,"rms_envelope_nrmse":0.11659305978871472,"envelope_correlation":0.9902726332423527,"rms_gain_error_db":-0.5778022026901479,"reference_mean_power":0.006407121641291136,"output_mean_power":0.0056089652454831},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7183045790225314e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025278777543949896,"output_power":0.0022200064556703136,"reference_active":true,"delta_db":-0.5640183071775589},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010431651037251879,"output_power":8.817086974468556e-05,"reference_active":true,"delta_db":-0.7302792565209117},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4474914310652014e-06,"output_power":1.8455410388502455e-06,"reference_active":true,"delta_db":-1.2259747302489459}],"active_band_count":3,"active_band_mae_db":0.8400907646491388}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002169046697405618,"log_rms_mae_db":2.1458348962914364,"rms_envelope_nrmse":0.16927977865438204,"envelope_correlation":0.9779061635779204,"rms_gain_error_db":-0.43771285808610777,"reference_mean_power":0.00266513165861851,"output_mean_power":0.0024096134707347174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9887000298118345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014141653900070267,"output_power":0.0012733118101747806,"reference_active":true,"delta_db":-0.45565436955582733},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010731466350302803,"output_power":9.196297688906556e-05,"reference_active":true,"delta_db":-0.6704604696800315},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0173327828420224e-05,"output_power":1.3371872261705425e-05,"reference_active":true,"delta_db":-1.7858532689176452}],"active_band_count":3,"active_band_mae_db":0.9706560360511679}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020963310569053025,"log_rms_mae_db":1.6532306412714075,"rms_envelope_nrmse":0.12273023652606567,"envelope_correlation":0.988615387684421,"rms_gain_error_db":-0.43811463325022737,"reference_mean_power":0.0026772178601939312,"output_mean_power":0.002420316994587156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.883939001913369e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013521293519634118,"output_power":0.0012145054471663747,"reference_active":true,"delta_db":-0.4661877354272549},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.792726117783695e-05,"output_power":4.070253541698455e-05,"reference_active":true,"delta_db":-0.7096114890294319},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.685082668516172e-06,"output_power":1.1405923519267348e-06,"reference_active":true,"delta_db":-1.694907566332704}],"active_band_count":3,"active_band_mae_db":0.9569022635964636}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002744641361828467,"log_rms_mae_db":2.280372583391071,"rms_envelope_nrmse":0.23788515108012792,"envelope_correlation":0.96358517030008,"rms_gain_error_db":-0.8944641048127983,"reference_mean_power":0.0019381811578355007,"output_mean_power":0.0015774222295358885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1196235236571852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006735796402683561,"output_power":0.0005633118768103836,"reference_active":true,"delta_db":-0.7764004367921311},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013823323993445422,"output_power":0.00011299528374307072,"reference_active":true,"delta_db":-0.8755217015952316},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8792319908524532e-05,"output_power":1.4395404882474395e-05,"reference_active":true,"delta_db":-1.157565127145173}],"active_band_count":3,"active_band_mae_db":0.9364957551775119}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002445180716567874,"log_rms_mae_db":1.6453999141118538,"rms_envelope_nrmse":0.1629598388300602,"envelope_correlation":0.986060030936088,"rms_gain_error_db":-0.8943547775879299,"reference_mean_power":0.0019467823981069803,"output_mean_power":0.001584462383570709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0571732642454034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000664982480329528,"output_power":0.0005568101127347114,"reference_active":true,"delta_db":-0.7710308921286578},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927173921912536e-05,"output_power":5.4449389262682275e-05,"reference_active":true,"delta_db":-1.045630788811004},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7481239585012511e-06,"output_power":1.5020791767959713e-06,"reference_active":true,"delta_db":-0.6587939940087153}],"active_band_count":3,"active_band_mae_db":0.8251518916494591}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":2.110366774392161,"rms_envelope_nrmse":0.27653020664759065,"envelope_correlation":0.9352344101496147,"rms_gain_error_db":-0.6213591194356199,"reference_mean_power":6.564824297759047e-05,"output_mean_power":5.6896715325996284e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.001422573299626e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.567613726314845e-05,"output_power":1.365228924529126e-05,"reference_active":true,"delta_db":-0.600335766672798},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.796434791027312e-06,"output_power":5.3830130608754605e-06,"reference_active":true,"delta_db":-0.32135520932670253},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.958485457614675e-07,"output_power":8.023111095179448e-07,"reference_active":true,"delta_db":-0.9385048762808188}],"active_band_count":3,"active_band_mae_db":0.6200652840934398}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":1.4344776915136803,"rms_envelope_nrmse":0.15969033544443906,"envelope_correlation":0.9790671907740134,"rms_gain_error_db":-0.6195466469246662,"reference_mean_power":6.585532886735057e-05,"output_mean_power":5.710001971230249e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6817222442955515e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.548077264503126e-05,"output_power":1.3565779644882023e-05,"reference_active":true,"delta_db":-0.5734787417225328},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9799921502867964e-06,"output_power":2.7149722547545727e-06,"reference_active":true,"delta_db":-0.4044972434454413},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.719812912358677e-08,"output_power":7.555771818047283e-08,"reference_active":true,"delta_db":-0.622283331525303}],"active_band_count":3,"active_band_mae_db":0.5334197722310924}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.9507293701171875,"predicted_mos":2.9507293701171875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"1c7ef9500011220c975cb40d66221ce96b208227426c020457519d353025aa3b","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.607151404999968},"started_at":"2026-10-13T05:11:05.019082+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029516169997805264,"native_input_samples":144000,"native_input_sha256":"cd7b4b7f559ae268b6c2b9b6024e77abd0bd35c06865bf41574a87f595f7c12c","output_sha256":"fcf10c7e5551938705ab8ac69eeaf276d255acc856c069997b6b6d09d70dba25","output_samples":144240,"length_delta_samples":240,"analysis_s":0.8163328120008373,"synthesis_s":0.2590446809990681,"total_s":1.0753774929999054,"rtf":0.17922958216665089,"output_peak":0.8807238340377808,"output_rms":0.07795438622505906,"finished_at":"2026-10-13T05:11:06.107126+00:00","analysis_stages":{"external_f0_s":0.16764038899964362},"evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"fcf10c7e5551938705ab8ac69eeaf276d255acc856c069997b6b6d09d70dba25","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.158382612743303,"output_peak":0.8810424936264559},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.7013121707874586,"spectral_convergence":0.1996120509208182,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.658593971946046,"spectral_convergence":0.2007422405811704,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.169778730252281,"spectral_convergence":0.2511408392617565,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.2140486659240935e-05,"output_power":3.60829118135504e-05,"delta_db":0.5024912851488074,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002187325069648938,"output_power":0.0021595405372929976,"delta_db":-0.05551969860931019,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018573993391017246,"output_power":0.0017505551602923918,"delta_db":-0.2572948667170824,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002022873257906857,"output_power":0.0018922449520649095,"delta_db":-0.2899131777876251,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023141111513331255,"output_power":0.0002196837769840361,"delta_db":-0.22586228434682584,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.19984427950361e-05,"output_power":3.5584744491882467e-05,"delta_db":-0.7196933624116831,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.34179577917022236,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19473478952868575,"pitch":null,"output_rms_dbfs":-22.151194946585903,"output_peak":0.8810424936264559,"reference_rms_dbfs":-21.95646015705721,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.8432282909952615},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8931477469822431,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0798319967885378,"rms_dbfs":-21.95646015705721,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.013818800436425,"mean":-1.4187812805175782e-05,"zero_energy":false},"output":{"rms":0.07806210401302137,"rms_dbfs":-22.151194946585896,"peak":0.8810424936264559,"peak_dbfs":-1.1000628918627295,"crest_factor_db":21.051132054723173,"mean":-2.305086144007823e-05,"zero_energy":false},"rms_difference_db":-0.19473478952868462},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":1.3508758984825315e-05,"output_rms_dbfs":-48.693845466699216,"output_zero_energy":false,"output_peak":0.06665684554397588,"output_energy_fraction":0.00027341064368177897,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":339,"output_estimated_voiced_frames":305,"both_estimated_voiced_frames":294,"reference_voiced_output_unvoiced_frames":45,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8672566371681416,"absolute_difference_cents_median":0.4744810614944721,"absolute_difference_cents_p95":5.38217694390177,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003217977906320387,"log_rms_mae_db":0.6749370102813178,"rms_envelope_nrmse":0.06640389982307722,"envelope_correlation":0.9964178056732339,"rms_gain_error_db":-0.1946075520588853,"reference_mean_power":0.006378025340653872,"output_mean_power":0.006098534502837951},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.934428693189062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026409373563640536,"output_power":0.0025816221853554,"reference_active":true,"delta_db":-0.09865415212354006},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022353254716419522,"output_power":0.00020894406523887254,"reference_active":true,"delta_db":-0.2931072681875994},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.171045723198584e-05,"output_power":2.892552142689738e-05,"reference_active":true,"delta_db":-0.399213076307221}],"active_band_count":3,"active_band_mae_db":0.2636581655394535}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00029448546471938644,"log_rms_mae_db":0.44338191744746513,"rms_envelope_nrmse":0.04779589256230192,"envelope_correlation":0.9982461312854691,"rms_gain_error_db":-0.19442273329598292,"reference_mean_power":0.006407121641291136,"output_mean_power":0.0061266164965834554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7183045790225314e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025278777543949896,"output_power":0.002470631427347759,"reference_active":true,"delta_db":-0.09948106571738634},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010431651037251879,"output_power":9.725644173963556e-05,"reference_active":true,"delta_db":-0.30434674217061153},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4474914310652014e-06,"output_power":2.076240857066898e-06,"reference_active":true,"delta_db":-0.7144344696070649}],"active_band_count":3,"active_band_mae_db":0.3727540924983543}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002169046697405618,"log_rms_mae_db":0.8227679017605352,"rms_envelope_nrmse":0.07006826814641925,"envelope_correlation":0.9963068502548231,"rms_gain_error_db":-0.16755676908351447,"reference_mean_power":0.00266513165861851,"output_mean_power":0.0025642655116533382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9887000298118345e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014141653900070267,"output_power":0.0013870122817020612,"reference_active":true,"delta_db":-0.08419897527564471},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010731466350302803,"output_power":0.00010474124729849057,"reference_active":true,"delta_db":-0.10541326789866917},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0173327828420224e-05,"output_power":1.8222006848804895e-05,"reference_active":true,"delta_db":-0.44181340587855966}],"active_band_count":3,"active_band_mae_db":0.21047521635095787}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020963310569053025,"log_rms_mae_db":0.5358062298336025,"rms_envelope_nrmse":0.04835922725633161,"envelope_correlation":0.9982907703101711,"rms_gain_error_db":-0.16737428707586677,"reference_mean_power":0.0026772178601939312,"output_mean_power":0.0025760025279955496},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.883939001913369e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013521293519634118,"output_power":0.00132490035060603,"reference_active":true,"delta_db":-0.08835025517668431},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.792726117783695e-05,"output_power":4.7340531559564036e-05,"reference_active":true,"delta_db":-0.05349481709086564},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.685082668516172e-06,"output_power":1.4946703262915947e-06,"reference_active":true,"delta_db":-0.520757992420105}],"active_band_count":3,"active_band_mae_db":0.22086768822921832}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002744641361828467,"log_rms_mae_db":0.9840877204350549,"rms_envelope_nrmse":0.0888935974378617,"envelope_correlation":0.9951513326321494,"rms_gain_error_db":-0.30203635109716914,"reference_mean_power":0.0019381811578355007,"output_mean_power":0.0018079679695159063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1196235236571852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006735796402683561,"output_power":0.0006496898834783547,"reference_active":true,"delta_db":-0.15682847347195814},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013823323993445422,"output_power":0.00012413948263600408,"reference_active":true,"delta_db":-0.4670255612193164},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8792319908524532e-05,"output_power":1.615460302692654e-05,"reference_active":true,"delta_db":-0.6568410658122028}],"active_band_count":3,"active_band_mae_db":0.4268983668344924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002445180716567874,"log_rms_mae_db":0.6630626610877471,"rms_envelope_nrmse":0.0653565800799809,"envelope_correlation":0.9976269983583447,"rms_gain_error_db":-0.30181296315309203,"reference_mean_power":0.0019467823981069803,"output_mean_power":0.0018160847625527828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0571732642454034e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000664982480329528,"output_power":0.0006392905990175323,"reference_active":true,"delta_db":-0.17111885469002985},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927173921912536e-05,"output_power":6.200850133430527e-05,"reference_active":true,"delta_db":-0.48104856555407693},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7481239585012511e-06,"output_power":1.3585190286469299e-06,"reference_active":true,"delta_db":-1.095064990533027}],"active_band_count":3,"active_band_mae_db":0.582410803592378}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":1.050123423020474,"rms_envelope_nrmse":0.12679395761336185,"envelope_correlation":0.9866425233956635,"rms_gain_error_db":-0.19099071034405646,"reference_mean_power":6.564824297759047e-05,"output_mean_power":6.282377600676898e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.001422573299626e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.567613726314845e-05,"output_power":1.5239012637442422e-05,"reference_active":true,"delta_db":-0.12282228377080351},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.796434791027312e-06,"output_power":5.005250521172075e-06,"reference_active":true,"delta_db":-0.6373513495847314},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.958485457614675e-07,"output_power":8.112720416284007e-07,"reference_active":true,"delta_db":-0.8902678409149213}],"active_band_count":3,"active_band_mae_db":0.550147158090152}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":0.7097083678453092,"rms_envelope_nrmse":0.07732800811180951,"envelope_correlation":0.99475121531932,"rms_gain_error_db":-0.1913068706406548,"reference_mean_power":6.585532886735057e-05,"output_mean_power":6.301736443688576e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6817222442955515e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.548077264503126e-05,"output_power":1.5040848200874385e-05,"reference_active":true,"delta_db":-0.12520304334123475},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9799921502867964e-06,"output_power":2.5061648115733496e-06,"reference_active":true,"delta_db":-0.7520549221325206},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.719812912358677e-08,"output_power":7.159490727139241e-08,"reference_active":true,"delta_db":-0.8562503611159383}],"active_band_count":3,"active_band_mae_db":0.5778361088632312}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_5683-32879-0019","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1593337059020996,"predicted_mos":3.1593337059020996,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287549,"input_sample_rate":48000,"input_seconds":5.9906041666666665,"input_float32_sha256":"00aa16abea0c5cacafd51b66d5f48717566df8cafd01aa7de0647b38607345b1","internal_samples":95850,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.359061446999476},"started_at":"2026-10-13T05:16:54.349431+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009171466999760014,"native_input_samples":288000,"native_input_sha256":"26380fafeafc65c607369470e36a47596a80c221c641c5171e80a6bea12e2836","output_sha256":"002a879dd0c0eb4bc171999491f4cf466ba796cdc3fba8d518fb9815963d269b","output_samples":287549,"length_delta_samples":-451,"analysis_s":4.144459334000203,"synthesis_s":1.3201335249996191,"total_s":5.464592858999822,"rtf":0.9107654764999703,"output_peak":0.9283900696401161,"output_rms":0.07746872647736222,"finished_at":"2026-10-13T05:16:59.835742+00:00","evaluation_input":{"reference":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"output_sha256":"002a879dd0c0eb4bc171999491f4cf466ba796cdc3fba8d518fb9815963d269b","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95850,"length_delta_samples":-150,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.212807678716743,"output_peak":0.8961893981266849},"common_prefix":{"samples":95850,"sample_rate":16000,"reference_coverage":0.9984375,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95850,"output_samples":95850,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.663427066889756,"spectral_convergence":0.3872902280850941,"spectral_convergence_status":"ok","amplitude_floor":2.005860984254184e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.612778507389854,"spectral_convergence":0.2785741288105882,"spectral_convergence_status":"ok","amplitude_floor":1.6673734947378395e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.708753437018567,"spectral_convergence":0.2504757853822447,"spectral_convergence_status":"ok","amplitude_floor":1.455049241978443e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":3.23333328468019e-05,"output_power":3.412984193806507e-05,"delta_db":0.23483803969501946,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0021896719319428916,"output_power":0.001946497302803747,"delta_db":-0.5112524518484693,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018604849728624528,"output_power":0.0017553655848509147,"delta_db":-0.25258587112942166,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.002025611049041462,"output_power":0.0019873145123853993,"delta_db":-0.08289453285380576,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00023122569381749315,"output_power":0.00024128162085088227,"delta_db":0.18488150380194687,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.205346026402217e-05,"output_power":4.326322842790658e-05,"delta_db":0.123171881876871,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23160404686758904,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2619540473731037,"pitch":null,"output_rms_dbfs":-22.212807678716743,"output_peak":0.8961893981266849,"reference_rms_dbfs":-21.95085363134364,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.66165300376606},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95850,"output_samples":95850,"intelligibility":{"estoi":{"value":0.8720275471816302,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07988354299087594,"rms_dbfs":-21.95085363134364,"peak":0.799591064453125,"peak_dbfs":-1.9426413566207843,"crest_factor_db":20.008212274722858,"mean":-1.6285593334311425e-05,"zero_energy":false},"output":{"rms":0.077510335383463,"rms_dbfs":-22.212807678716743,"peak":0.8961893981266849,"peak_dbfs":-0.9520039614680588,"crest_factor_db":21.260803717248685,"mean":0.0012349964579699245,"zero_energy":false},"rms_difference_db":-0.26195404737310213},"full_scale":{"reference":{"sample_count":95850,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95850,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":10,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.726337239262648e-06,"threshold_dbfs":-50.12050676304782,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":74,"quiet_samples":11840,"reference_mean_power":3.557419640364478e-06,"output_mean_power":3.7871846578825394e-06,"output_rms_dbfs":-54.216835187650275,"output_zero_energy":false,"output_peak":0.027804940749027916,"output_energy_fraction":7.786760806932744e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":338,"output_estimated_voiced_frames":330,"both_estimated_voiced_frames":323,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9556213017751479,"absolute_difference_cents_median":3.1614155841900327,"absolute_difference_cents_p95":17.663140540504862,"difference_over_600_cents_frames":0,"uncovered_tail_samples":10,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95850,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1197,"rms_floor":0.0003217977906320387,"log_rms_mae_db":1.4338382197389803,"rms_envelope_nrmse":0.17662862608155855,"envelope_correlation":0.9721626811062608,"rms_gain_error_db":-0.2618933225753972,"reference_mean_power":0.006386691431487493,"output_mean_power":0.006012936282618162},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.932821789520232e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0026399066000462506,"output_power":0.002432378244923515,"reference_active":true,"delta_db":-0.3555745134467881},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002231583888695557,"output_power":0.00021504259545385813,"reference_active":true,"delta_db":-0.16088724168733956},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.141776554756477e-05,"output_power":3.0250356345955568e-05,"reference_active":true,"delta_db":-0.16444799528053045}],"active_band_count":3,"active_band_mae_db":0.226969916804886}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1191,"rms_floor":0.00029448546471938644,"log_rms_mae_db":1.0017162788987128,"rms_envelope_nrmse":0.11064643867341742,"envelope_correlation":0.988793741229102,"rms_gain_error_db":-0.2625080334590193,"reference_mean_power":0.006415710203195347,"output_mean_power":0.006039401956678446},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.7167701761502867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0025267017543213234,"output_power":0.002328608238262816,"reference_active":true,"delta_db":-0.35457552295172456},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010394624185061282,"output_power":0.00010115515058211629,"reference_active":true,"delta_db":-0.1182079122471711},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4292380634974333e-06,"output_power":2.2755653095404317e-06,"reference_active":true,"delta_db":-0.28380772993624725}],"active_band_count":3,"active_band_mae_db":0.25219705504504764}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1197,"rms_floor":0.0002169046697405618,"log_rms_mae_db":1.7424268462637171,"rms_envelope_nrmse":0.15199004993161339,"envelope_correlation":0.9822338598029907,"rms_gain_error_db":-0.3841439601707977,"reference_mean_power":0.0026685302149359645,"output_mean_power":0.0024426302350909076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.989513583946499e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014140558776058602,"output_power":0.0012526763085212243,"reference_active":true,"delta_db":-0.526277074871668},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010734684472157074,"output_power":0.00010133358044058611,"reference_active":true,"delta_db":-0.25035895669920727},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0015362617653695e-05,"output_power":1.909189104722295e-05,"reference_active":true,"delta_db":-0.20514515322634325}],"active_band_count":3,"active_band_mae_db":0.32726039493240616}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1191,"rms_floor":0.00020963310569053025,"log_rms_mae_db":1.3419105444137285,"rms_envelope_nrmse":0.10400731645888979,"envelope_correlation":0.991984061391848,"rms_gain_error_db":-0.38363483475184507,"reference_mean_power":0.002681018816143429,"output_mean_power":0.00245434934220594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.884664208440078e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0013519373358876403,"output_power":0.001199654212685896,"reference_active":true,"delta_db":-0.5190047852770872},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.775962595882716e-05,"output_power":4.5255788147682825e-05,"reference_active":true,"delta_db":-0.2338678414241112},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.683816962375765e-06,"output_power":1.6027797298129257e-06,"reference_active":true,"delta_db":-0.21421038867197062}],"active_band_count":3,"active_band_mae_db":0.32236100512438964}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1197,"rms_floor":0.0002744641361828467,"log_rms_mae_db":1.7316813975123038,"rms_envelope_nrmse":0.27347949329721727,"envelope_correlation":0.9478409538541501,"rms_gain_error_db":-0.04427227129457582,"reference_mean_power":0.0019407491945753763,"output_mean_power":0.0019210655654610402},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1163242061173237e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006727204983011425,"output_power":0.0006543922681310047,"reference_active":true,"delta_db":-0.11996501809567292},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013765991469580441,"output_power":0.00013498318118854876,"reference_active":true,"delta_db":-0.08527837024568259},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8671154577584244e-05,"output_power":1.9319553077206804e-05,"reference_active":true,"delta_db":0.1482590111518603}],"active_band_count":3,"active_band_mae_db":0.11783413316440527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1191,"rms_floor":0.0002445180716567874,"log_rms_mae_db":1.263709571540206,"rms_envelope_nrmse":0.16909153921196438,"envelope_correlation":0.9787926097734851,"rms_gain_error_db":-0.045697057604670535,"reference_mean_power":0.001949062045596807,"output_mean_power":0.0019286612670819335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0539793623455465e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006640756663061234,"output_power":0.0006461292593399748,"reference_active":true,"delta_db":-0.11898158605841015},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.892990231203228e-05,"output_power":6.835351259021126e-05,"reference_active":true,"delta_db":-0.036468257535819316},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.732756112398842e-06,"output_power":1.7997813217890788e-06,"reference_active":true,"delta_db":0.1648230085460866}],"active_band_count":3,"active_band_mae_db":0.10675761738010536}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1197,"rms_floor":5.688273685761601e-05,"log_rms_mae_db":1.6993329700941178,"rms_envelope_nrmse":0.27582733702545603,"envelope_correlation":0.940630219204486,"rms_gain_error_db":0.37593095123560505,"reference_mean_power":6.571011866433656e-05,"output_mean_power":7.165150989237518e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.0019288313072804e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.567293563491629e-05,"output_power":1.7750689194134398e-05,"reference_active":true,"delta_db":0.5406486972876802},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.794490041999584e-06,"output_power":6.457697056845475e-06,"reference_active":true,"delta_db":0.4706244670506729},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.937759381432871e-07,"output_power":1.1600173605823612e-06,"reference_active":true,"delta_db":0.6717601168102603}],"active_band_count":3,"active_band_mae_db":0.5610110937162044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":10,"metrics":{"frame_count":1191,"rms_floor":3.766393520730195e-05,"log_rms_mae_db":1.293331813794516,"rms_envelope_nrmse":0.1597577813047432,"envelope_correlation":0.9788056606611016,"rms_gain_error_db":0.3707988871665066,"reference_mean_power":6.586008856730648e-05,"output_mean_power":7.173022599284104e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.682056148541952e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5472324113389566e-05,"output_power":1.7529195674081387e-05,"reference_active":true,"delta_db":0.5420643456801733},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9785303196163117e-06,"output_power":3.3664403784583132e-06,"reference_active":true,"delta_db":0.5316890196445595},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.680441530161322e-08,"output_power":9.416325013203108e-08,"reference_active":true,"delta_db":0.3533962406919566}],"active_band_count":3,"active_band_mae_db":0.4757165353388964}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},{"case":"librispeech_61-70968-0042","method":"hnm","status":"measured","mos":{"status":"measured","value":2.0247488021850586,"predicted_mos":2.0247488021850586,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44560,"input_sample_rate":16000,"input_seconds":2.785,"input_float32_sha256":"f0f467c95c9091a36109e21abdb3acdd611b40bdde085653856a8e41771c95d4","internal_samples":44560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.484574228999918},"started_at":"2026-10-13T04:39:45.368265+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.383299968438223e-05,"native_input_samples":44560,"native_input_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","output_sha256":"951ef11f251b323613fd0fceba64b7356de96fd01e762448b8088288ca590857","output_samples":44560,"length_delta_samples":0,"analysis_s":0.6269537759999366,"synthesis_s":0.22687680099988938,"total_s":0.853830576999826,"rtf":0.3065818947934743,"output_peak":0.3479316517959658,"output_rms":0.043742039048373295,"finished_at":"2026-10-13T04:39:46.225469+00:00","evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"951ef11f251b323613fd0fceba64b7356de96fd01e762448b8088288ca590857","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.995885943582176,"spectral_convergence":0.3216829854189462,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.575508161314728,"spectral_convergence":0.30194082003508055,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.699257639055105,"spectral_convergence":0.3076854860611755,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0006048701395318235,"delta_db":0.10126069048331879,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.0005654144664958374,"delta_db":-0.24537290530100764,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.0005197738168176318,"delta_db":-0.2797171813141233,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":0.00010760359688498982,"delta_db":-0.2491199047384407,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":9.949221883709383e-05,"delta_db":-0.5348265647691901,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.6211741542038946e-05,"delta_db":-0.07203311503107951,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24705506027286003,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1629762434108378,"pitch":null,"output_rms_dbfs":-27.182019521158992,"output_peak":0.3479316517959658,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.756883914650669},"common_prefix":{"samples":44560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.995885943582176,"spectral_convergence":0.3216829854189462,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.575508161314728,"spectral_convergence":0.30194082003508055,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.699257639055105,"spectral_convergence":0.3076854860611755,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0006048701395318235,"delta_db":0.10126069048331879,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.0005654144664958374,"delta_db":-0.24537290530100764,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.0005197738168176318,"delta_db":-0.2797171813141233,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":0.00010760359688498982,"delta_db":-0.2491199047384407,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":9.949221883709383e-05,"delta_db":-0.5348265647691901,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.6211741542038946e-05,"delta_db":-0.07203311503107951,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24705506027286003,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1629762434108378,"pitch":null,"output_rms_dbfs":-27.182019521158992,"output_peak":0.3479316517959658,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.756883914650669},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44560,"output_samples":44560,"intelligibility":{"estoi":{"value":0.8403637274048065,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044570533867492254,"rms_dbfs":-27.019043277748153,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.196863425855337,"mean":-0.0008769427424487349,"zero_energy":false},"output":{"rms":0.043742039048373295,"rms_dbfs":-27.182019521158992,"peak":0.3479316517959658,"peak_dbfs":-9.170121222400539,"crest_factor_db":18.011898298758453,"mean":-4.2857682648162024e-05,"zero_energy":false},"rms_difference_db":-0.1629762434108386},"full_scale":{"reference":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":62,"both_estimated_voiced_frames":42,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":1.6218686796051536,"absolute_difference_cents_p95":7.198705878650369,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":0.00011061650326899306,"log_rms_mae_db":1.1502258367752372,"rms_envelope_nrmse":0.09257385833585911,"envelope_correlation":0.9882894663811006,"rms_gain_error_db":-0.16295617543744323,"reference_mean_power":0.00199005080866686,"output_mean_power":0.0019167635723816002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":8.439637043999617e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970864143567599,"output_power":0.0005782300648805987,"reference_active":true,"delta_db":-0.13936520773523625},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010123102438186566,"output_power":9.598399381393108e-05,"reference_active":true,"delta_db":-0.23114815000874114},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240115070482855e-05,"output_power":2.6145279770083176e-05,"reference_active":true,"delta_db":-0.9316714075317255}],"active_band_count":3,"active_band_mae_db":0.4340615884252343}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":0.48727216377162097,"rms_envelope_nrmse":0.042532193741649515,"envelope_correlation":0.9975161059574734,"rms_gain_error_db":-0.16296276861055153,"reference_mean_power":0.0020114637230104894,"output_mean_power":0.0019373849760417829},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":6.839204647135985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005172552173341314,"output_power":0.0005004526954422182,"reference_active":true,"delta_db":-0.14341847198442423},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.664160204550464e-05,"output_power":5.440383307355635e-05,"reference_active":true,"delta_db":-0.1750602899289433},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529860845744804e-06,"output_power":2.086846186706939e-06,"reference_active":true,"delta_db":-0.8360619349924783}],"active_band_count":3,"active_band_mae_db":0.3848468989686153}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":1.7809146759812011,"rms_envelope_nrmse":0.12532416363984006,"envelope_correlation":0.9857392718381356,"rms_gain_error_db":-0.275590662814091,"reference_mean_power":0.000839400894204417,"output_mean_power":0.0007877898141708492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.694834819211473e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042315114528166547,"output_power":0.0003955041474834005,"reference_active":true,"delta_db":-0.29344478535675333},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.939611213191948e-05,"output_power":5.864091384213741e-05,"reference_active":true,"delta_db":-0.7313441049662086},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4277196064029688e-05,"output_power":1.9127799624137038e-05,"reference_active":true,"delta_db":-1.0353351210050585}],"active_band_count":3,"active_band_mae_db":0.6867080037760068}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":1.097201301008649,"rms_envelope_nrmse":0.06558659000936626,"envelope_correlation":0.9963157382921948,"rms_gain_error_db":-0.2756047756343545,"reference_mean_power":0.0008485551352947845,"output_mean_power":0.0007963786132061378},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":5.733703583211585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003764127527272362,"output_power":0.00035330208391619577,"reference_active":true,"delta_db":-0.2751812965968434},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8963437868750806e-05,"output_power":3.338098602731627e-05,"reference_active":true,"delta_db":-0.6715810810838312},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2966305004996595e-06,"output_power":1.7148271354284542e-06,"reference_active":true,"delta_db":-1.2687078102375116}],"active_band_count":3,"active_band_mae_db":0.7384900626393954}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":1.926349156153344,"rms_envelope_nrmse":0.22244518734948504,"envelope_correlation":0.9468526174853364,"rms_gain_error_db":-0.16997366349709248,"reference_mean_power":0.00013682778387507752,"output_mean_power":0.00013157607490002813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.073610600879068e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.932287229550181e-05,"output_power":3.925968579530088e-05,"reference_active":true,"delta_db":-0.006984133616613738},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.861650334659597e-06,"output_power":8.86208991309005e-06,"reference_active":true,"delta_db":1.1110756907447652},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.31740167391869e-06,"output_power":4.503824900123884e-06,"reference_active":true,"delta_db":0.18359042082979604}],"active_band_count":3,"active_band_mae_db":0.43388341506372496}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":1.0833685541757287,"rms_envelope_nrmse":0.12855576356069107,"envelope_correlation":0.9794321113412443,"rms_gain_error_db":-0.17004884435127665,"reference_mean_power":0.00013831808192096793,"output_mean_power":0.00013300686998394352},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":4.8587078894174116e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5019378943818874e-05,"output_power":3.605716557230208e-05,"reference_active":true,"delta_db":0.12683144394989468},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.168829067205391e-06,"output_power":5.25563496151778e-06,"reference_active":true,"delta_db":1.0061110469331782},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163313597471507e-07,"output_power":3.2912135957244284e-07,"reference_active":true,"delta_db":0.17213820228142768}],"active_band_count":3,"active_band_mae_db":0.4350268977215002}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":2.073016752250715,"rms_envelope_nrmse":0.1959306078241507,"envelope_correlation":0.9774259630607505,"rms_gain_error_db":-0.5218283878678414,"reference_mean_power":5.106513987055756e-05,"output_mean_power":4.5283677315465615e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":2.3044643706212166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.045287512435104e-05,"output_power":9.063061586549565e-06,"reference_active":true,"delta_db":-0.619608309322395},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8556283872264966e-06,"output_power":2.404441361861044e-06,"reference_active":true,"delta_db":-0.7468750032810696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6522678437727898e-06,"output_power":6.706817412434033e-07,"reference_active":true,"delta_db":-3.915639675660355}],"active_band_count":3,"active_band_mae_db":1.7607076627546066}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":1.4772441465357316,"rms_envelope_nrmse":0.13043458041100198,"envelope_correlation":0.9912812287290071,"rms_gain_error_db":-0.5218647520361055,"reference_mean_power":5.1621714358619315e-05,"output_mean_power":4.577685458653547e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":2.0666699185875215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0052154991285664e-05,"output_power":8.436939072604475e-06,"reference_active":true,"delta_db":-0.760742633749571},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8908095519422077e-06,"output_power":1.3382621603729678e-06,"reference_active":true,"delta_db":-1.501065892213277},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.81443132062933e-07,"output_power":6.334497652036419e-08,"reference_active":true,"delta_db":-4.570283545449828}],"active_band_count":3,"active_band_mae_db":2.277364023804225}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0042","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.908222198486328,"predicted_mos":2.908222198486328,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44560,"input_sample_rate":16000,"input_seconds":2.785,"input_float32_sha256":"823056638bda704b83c075363b6eaaab64d3718cbd8f0aae227289e11a66b5b9","internal_samples":44560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4334737449989916},"started_at":"2026-10-13T04:42:51.854581+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.477700026152888e-05,"native_input_samples":44560,"native_input_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","output_sha256":"55f4704a1fabdb106c0cbbeed573c53b970c4ee50eb6dbcd0f88be42353b4a9e","output_samples":44560,"length_delta_samples":0,"analysis_s":0.9670662929993341,"synthesis_s":0.2197099969998817,"total_s":1.1867762899992158,"rtf":0.4261315224413701,"output_peak":0.4545788436980868,"output_rms":0.04586811263369409,"finished_at":"2026-10-13T04:42:53.048744+00:00","evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"55f4704a1fabdb106c0cbbeed573c53b970c4ee50eb6dbcd0f88be42353b4a9e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.9900066393041556,"spectral_convergence":0.8725202439449806,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.547365155796667,"spectral_convergence":0.38033207916588935,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.666599030942695,"spectral_convergence":0.31280931580843074,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0005849310605059295,"delta_db":-0.0443139277695263,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.0006687780782524796,"delta_db":0.48377822477541493,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.0005825084639720704,"delta_db":0.21516123518372773,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":0.00013046618593845198,"delta_db":0.5875918707475651,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":0.00012077038007588735,"delta_db":0.306886601501541,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.64295878324306e-05,"delta_db":-0.01406314361606038,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27529916726563924,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.24926068355520362,"pitch":null,"output_rms_dbfs":-26.76978259419295,"output_peak":0.4545788436980868,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.067990275347839},"common_prefix":{"samples":44560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.9900066393041556,"spectral_convergence":0.8725202439449806,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.547365155796667,"spectral_convergence":0.38033207916588935,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.666599030942695,"spectral_convergence":0.31280931580843074,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0005849310605059295,"delta_db":-0.0443139277695263,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.0006687780782524796,"delta_db":0.48377822477541493,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.0005825084639720704,"delta_db":0.21516123518372773,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":0.00013046618593845198,"delta_db":0.5875918707475651,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":0.00012077038007588735,"delta_db":0.306886601501541,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.64295878324306e-05,"delta_db":-0.01406314361606038,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27529916726563924,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.24926068355520362,"pitch":null,"output_rms_dbfs":-26.76978259419295,"output_peak":0.4545788436980868,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.067990275347839},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44560,"output_samples":44560,"intelligibility":{"estoi":{"value":0.9025561675191314,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044570533867492254,"rms_dbfs":-27.019043277748153,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.196863425855337,"mean":-0.0008769427424487349,"zero_energy":false},"output":{"rms":0.04586811263369409,"rms_dbfs":-26.76978259419295,"peak":0.4545788436980868,"peak_dbfs":-6.847815607987936,"crest_factor_db":19.921966986205014,"mean":3.067558680182731e-07,"zero_energy":false},"rms_difference_db":0.24926068355520314},"full_scale":{"reference":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":38,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9047619047619048,"absolute_difference_cents_median":4.641882128539793,"absolute_difference_cents_p95":12.948941352536032,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":0.00011061650326899306,"log_rms_mae_db":2.519360913731302,"rms_envelope_nrmse":0.17933037538737986,"envelope_correlation":0.9597023931264116,"rms_gain_error_db":0.24936530516999234,"reference_mean_power":0.00199005080866686,"output_mean_power":0.0021076606881414986},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":8.439637043999617e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970864143567599,"output_power":0.0006452000623105873,"reference_active":true,"delta_db":0.33657210966288265},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010123102438186566,"output_power":9.52642408988942e-05,"reference_active":true,"delta_db":-0.2638372041676516},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240115070482855e-05,"output_power":3.0364319847293178e-05,"reference_active":true,"delta_db":-0.28196876667236437}],"active_band_count":3,"active_band_mae_db":0.2941260268342995}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":1.7038308029659388,"rms_envelope_nrmse":0.08857553236650624,"envelope_correlation":0.9900274431891286,"rms_gain_error_db":0.2498444167900171,"reference_mean_power":0.0020114637230104894,"output_mean_power":0.0021305741138408444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":6.839204647135985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005172552173341314,"output_power":0.0005594917079668752,"reference_active":true,"delta_db":0.34088774479511286},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.664160204550464e-05,"output_power":5.698198689808205e-05,"reference_active":true,"delta_db":0.026020601319311748},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529860845744804e-06,"output_power":2.146277161933535e-06,"reference_active":true,"delta_db":-0.7141082923044177}],"active_band_count":3,"active_band_mae_db":0.3603388794729474}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":2.75075508435376,"rms_envelope_nrmse":0.196140611482424,"envelope_correlation":0.9661775004325195,"rms_gain_error_db":0.30473865070672285,"reference_mean_power":0.000839400894204417,"output_mean_power":0.0009004161844698713},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.694834819211473e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042315114528166547,"output_power":0.0004435659564013132,"reference_active":true,"delta_db":0.2046268608694919},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.939611213191948e-05,"output_power":6.309530298297904e-05,"reference_active":true,"delta_db":-0.4133810992409373},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4277196064029688e-05,"output_power":1.4889167794136213e-05,"reference_active":true,"delta_db":-2.123281014757489}],"active_band_count":3,"active_band_mae_db":0.9137629916226393}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":1.9245163135809964,"rms_envelope_nrmse":0.10841805816544227,"envelope_correlation":0.9892418796032274,"rms_gain_error_db":0.3046234975927332,"reference_mean_power":0.0008485551352947845,"output_mean_power":0.0009102117044796911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":5.733703583211585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003764127527272362,"output_power":0.00039351415503569194,"reference_active":true,"delta_db":0.19296030170810363},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8963437868750806e-05,"output_power":3.671137192422097e-05,"reference_active":true,"delta_db":-0.2585665445682914},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2966305004996595e-06,"output_power":1.3514156263142927e-06,"reference_active":true,"delta_db":-2.3030219186210945}],"active_band_count":3,"active_band_mae_db":0.9181829216324965}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":1.686082587369346,"rms_envelope_nrmse":0.19216768542088952,"envelope_correlation":0.9668656341634113,"rms_gain_error_db":0.5130061129959712,"reference_mean_power":0.00013682778387507752,"output_mean_power":0.0001539837543951839},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.073610600879068e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.932287229550181e-05,"output_power":4.15929197177302e-05,"reference_active":true,"delta_db":0.24374174932378478},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.861650334659597e-06,"output_power":7.475860966509703e-06,"reference_active":true,"delta_db":0.37232633009300264},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.31740167391869e-06,"output_power":3.1845312962766475e-06,"reference_active":true,"delta_db":-1.3217693425915098}],"active_band_count":3,"active_band_mae_db":0.6459458073360991}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":0.8697497320428452,"rms_envelope_nrmse":0.1031541428897783,"envelope_correlation":0.9919069816690532,"rms_gain_error_db":0.5130197037423923,"reference_mean_power":0.00013831808192096793,"output_mean_power":0.0001556613986120773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":4.8587078894174116e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5019378943818874e-05,"output_power":3.695071851396525e-05,"reference_active":true,"delta_db":0.2331444799613021},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.168829067205391e-06,"output_power":4.56596832991472e-06,"reference_active":true,"delta_db":0.3951880655563068},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163313597471507e-07,"output_power":2.406451672219015e-07,"reference_active":true,"delta_db":-1.1876510381673946}],"active_band_count":3,"active_band_mae_db":0.6053278612283345}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":1.8968108053600234,"rms_envelope_nrmse":0.1940490151631267,"envelope_correlation":0.9765493802767173,"rms_gain_error_db":-0.02442956213279785,"reference_mean_power":5.106513987055756e-05,"output_mean_power":5.077869899544988e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":2.3044643706212166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.045287512435104e-05,"output_power":1.0547174261291084e-05,"reference_active":true,"delta_db":0.03900359407173466},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8556283872264966e-06,"output_power":2.9027123463859292e-06,"reference_active":true,"delta_db":0.07102309613828929},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6522678437727898e-06,"output_power":7.228595692276056e-07,"reference_active":true,"delta_db":-3.5902651616608603}],"active_band_count":3,"active_band_mae_db":1.2334306172902947}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":1.090147141227874,"rms_envelope_nrmse":0.09396365177166274,"envelope_correlation":0.9942861655387107,"rms_gain_error_db":-0.02442449785510674,"reference_mean_power":5.1621714358619315e-05,"output_mean_power":5.133221133540788e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":2.0666699185875215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0052154991285664e-05,"output_power":9.975590488646425e-06,"reference_active":true,"delta_db":-0.03320563767705338},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8908095519422077e-06,"output_power":1.6018588766095981e-06,"reference_active":true,"delta_db":-0.720235354021369},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.81443132062933e-07,"output_power":7.043744682239915e-08,"reference_active":true,"delta_db":-4.109369285342849}],"active_band_count":3,"active_band_mae_db":1.6209367590137571}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0042","method":"facodec","status":"measured","mos":{"status":"measured","value":4.039346694946289,"predicted_mos":4.039346694946289,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44600,"input_sample_rate":16000,"input_seconds":2.7875,"input_float32_sha256":"724983abacacd9c14427fd06620508ee1f20d68d2ef2f7653b5984d130d57136","internal_samples":44600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.278601604999494},"started_at":"2026-10-13T04:56:29.726857+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.0625000110594556e-05,"native_input_samples":44560,"native_input_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","output_sha256":"83651dc96eeb3455a039e678ea1f72975427afc22654177e9341c2058aa0a423","output_samples":44600,"length_delta_samples":40,"analysis_s":0.8435902139999598,"synthesis_s":4.843475543999375,"total_s":5.687065757999335,"rtf":2.0420343834827053,"output_peak":0.33226537704467773,"output_rms":0.04214789533234743,"finished_at":"2026-10-13T04:56:35.425068+00:00","evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"83651dc96eeb3455a039e678ea1f72975427afc22654177e9341c2058aa0a423","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":44560,"output_samples":44600,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.504482142430774,"output_peak":0.33226537704467773},"common_prefix":{"samples":44560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991031390134529,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.506063621988128,"spectral_convergence":0.2786558158053279,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.926138259750259,"spectral_convergence":0.29264393795489096,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.987354379208315,"spectral_convergence":0.3259877466403606,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0005851879446075655,"delta_db":-0.04240705591915973,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.0005708504605724545,"delta_db":-0.2038185074442574,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.00044014973430317645,"delta_db":-1.001856725867861,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":9.49492328224465e-05,"delta_db":-0.7924731924361443,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":7.235177312635452e-05,"delta_db":-1.9182259406951807,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.4547019198961382e-05,"delta_db":-0.5425897081571154,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7502285217532866,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.48155082103961855,"pitch":null,"output_rms_dbfs":-27.500594098787772,"output_peak":0.33226537704467773,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.806518753648901},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44560,"output_samples":44560,"intelligibility":{"estoi":{"value":0.8764009638033803,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044570533867492254,"rms_dbfs":-27.019043277748153,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.196863425855337,"mean":-0.0008769427424487349,"zero_energy":false},"output":{"rms":0.04216676611540133,"rms_dbfs":-27.500594098787772,"peak":0.33226537704467773,"peak_dbfs":-9.570298220579348,"crest_factor_db":17.930295878208426,"mean":3.7527155221899226e-05,"zero_energy":false},"rms_difference_db":-0.4815508210396189},"full_scale":{"reference":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":39,"both_estimated_voiced_frames":27,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.6428571428571429,"absolute_difference_cents_median":3.735077402138782,"absolute_difference_cents_p95":9.192740477228059,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":0.00011061650326899306,"log_rms_mae_db":1.1189559428983085,"rms_envelope_nrmse":0.12248714784843226,"envelope_correlation":0.9821198769799794,"rms_gain_error_db":-0.481461079187486,"reference_mean_power":0.00199005080866686,"output_mean_power":0.0017812220256699753},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":8.439637043999617e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970864143567599,"output_power":0.0004741007454999781,"reference_active":true,"delta_db":-1.0016655146680014},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010123102438186566,"output_power":9.928988069661502e-05,"reference_active":true,"delta_db":-0.08408642742105163},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240115070482855e-05,"output_power":2.7739024022756413e-05,"reference_active":true,"delta_db":-0.6746925748080985}],"active_band_count":3,"active_band_mae_db":0.5868148389657172}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":0.7894700889454543,"rms_envelope_nrmse":0.10618300429037088,"envelope_correlation":0.9851958445874455,"rms_gain_error_db":-0.4813126449527513,"reference_mean_power":0.0020114637230104894,"output_mean_power":0.0018004494808570447},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":6.839204647135985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005172552173341314,"output_power":0.000414276224059251,"reference_active":true,"delta_db":-0.9641487070964326},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.664160204550464e-05,"output_power":5.208447671347772e-05,"reference_active":true,"delta_db":-0.3642722318326531},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529860845744804e-06,"output_power":1.9337884688712e-06,"reference_active":true,"delta_db":-1.166876673745946}],"active_band_count":3,"active_band_mae_db":0.8317658708916772}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":1.727188226134554,"rms_envelope_nrmse":0.15730882811553165,"envelope_correlation":0.9819013318247476,"rms_gain_error_db":-0.7602326934431834,"reference_mean_power":0.000839400894204417,"output_mean_power":0.0007046057097977652},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.694834819211473e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042315114528166547,"output_power":0.00031432629669756753,"reference_active":true,"delta_db":-1.2911480482000597},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.939611213191948e-05,"output_power":6.921617675785729e-05,"reference_active":true,"delta_db":-0.011275332739221142},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4277196064029688e-05,"output_power":2.0341059014704338e-05,"reference_active":true,"delta_db":-0.7682496591179898}],"active_band_count":3,"active_band_mae_db":0.6902243466857568}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":1.18375394598296,"rms_envelope_nrmse":0.13671673024420777,"envelope_correlation":0.9864755354492225,"rms_gain_error_db":-0.7602240650913944,"reference_mean_power":0.0008485551352947845,"output_mean_power":0.0007122913323163551},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":5.733703583211585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003764127527272362,"output_power":0.00028002648054215376,"reference_active":true,"delta_db":-1.2846522665302043},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8963437868750806e-05,"output_power":3.7569744977678185e-05,"reference_active":true,"delta_db":-0.15819021993208976},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2966305004996595e-06,"output_power":1.6349242575778503e-06,"reference_active":true,"delta_db":-1.4759349063332359}],"active_band_count":3,"active_band_mae_db":0.97292579759851}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":1.7483669079454671,"rms_envelope_nrmse":0.1988665177949957,"envelope_correlation":0.9630169709611383,"rms_gain_error_db":-0.7302137866907926,"reference_mean_power":0.00013682778387507752,"output_mean_power":0.0001156519378936678},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.073610600879068e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.932287229550181e-05,"output_power":3.670138902870151e-05,"reference_active":true,"delta_db":-0.29962731673837595},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.861650334659597e-06,"output_power":6.890501880766322e-06,"reference_active":true,"delta_db":0.018222726822431628},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.31740167391869e-06,"output_power":4.4591882197274515e-06,"reference_active":true,"delta_db":0.14033348481541819}],"active_band_count":3,"active_band_mae_db":0.15272784279207527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":1.4117217093835903,"rms_envelope_nrmse":0.15414278725809416,"envelope_correlation":0.9769171556465645,"rms_gain_error_db":-0.730186860174572,"reference_mean_power":0.00013831808192096793,"output_mean_power":0.00011691231815634183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":4.8587078894174116e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5019378943818874e-05,"output_power":3.414778557931769e-05,"reference_active":true,"delta_db":-0.1094589402069292},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.168829067205391e-06,"output_power":3.926974955412671e-06,"reference_active":true,"delta_db":-0.25955956759684506},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163313597471507e-07,"output_power":2.7543688481862284e-07,"reference_active":true,"delta_db":-0.6012015029066952}],"active_band_count":3,"active_band_mae_db":0.3234066702368232}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":1.7569262019910954,"rms_envelope_nrmse":0.274758574159963,"envelope_correlation":0.9725528688997906,"rms_gain_error_db":-1.8860629557799855,"reference_mean_power":5.106513987055756e-05,"output_mean_power":3.3076399613172285e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":2.3044643706212166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.045287512435104e-05,"output_power":6.198009769969621e-06,"reference_active":true,"delta_db":-2.269835056267037},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8556283872264966e-06,"output_power":3.4043562759389558e-06,"reference_active":true,"delta_db":0.7633331344920791},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6522678437727898e-06,"output_power":2.3214678029614963e-06,"reference_active":true,"delta_db":1.4768221394228869}],"active_band_count":3,"active_band_mae_db":1.5033301100606675}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":1.3423166847127925,"rms_envelope_nrmse":0.24873667775072156,"envelope_correlation":0.984289865529363,"rms_gain_error_db":-1.886075569599612,"reference_mean_power":5.1621714358619315e-05,"output_mean_power":3.3436812235564644e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":2.0666699185875215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0052154991285664e-05,"output_power":6.380540115026774e-06,"reference_active":true,"delta_db":-1.9740173277171345},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8908095519422077e-06,"output_power":2.3103333410697687e-06,"reference_active":true,"delta_db":0.8702685798780465},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.81443132062933e-07,"output_power":2.3888514640260345e-07,"reference_active":true,"delta_db":1.1944861262063795}],"active_band_count":3,"active_band_mae_db":1.3462573446005202}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0042","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.611051559448242,"predicted_mos":2.611051559448242,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":122880,"input_sample_rate":44100,"input_seconds":2.786394557823129,"input_float32_sha256":"5fb53466b7fd1c93d22ba751d05668d225e35910007b655a6dce5782df624ac9","internal_samples":44583,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8516271240005153},"started_at":"2026-10-13T07:21:59.961781+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0047618820008210605,"native_input_samples":122819,"native_input_sha256":"f409fc56effb6a8127134777a6a441e41edce04417a507cc7a54f1b3ae1c3f8c","output_sha256":"ba311f352bd7b2d8c8e62c8f954278beca815ffd64b5e907a6061567c8b10f7e","output_samples":122880,"length_delta_samples":61,"analysis_s":0.0740968520003662,"synthesis_s":13.990472654999394,"total_s":14.06456950699976,"rtf":5.050094165061509,"output_peak":0.28773266077041626,"output_rms":0.03722418809202702,"finished_at":"2026-10-13T07:22:14.042370+00:00","evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"ba311f352bd7b2d8c8e62c8f954278beca815ffd64b5e907a6061567c8b10f7e","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":44560,"output_samples":44583,"length_delta_samples":23,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.577856931633058,"output_peak":0.2883241631003222},"common_prefix":{"samples":44560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9994841082923984,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.348686871629846,"spectral_convergence":0.6388283273244421,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.966421387214343,"spectral_convergence":0.47655192498713583,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.878587146480381,"spectral_convergence":0.5678519286941281,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.00036141475503409683,"delta_db":-2.1353019729292684,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.00044279232091782276,"delta_db":-1.3070412596899326,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.00043652275209663227,"delta_db":-1.037792320288715,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":6.085508227183431e-05,"delta_db":-2.7244192539112695,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":7.334693376772701e-05,"delta_db":-1.8588980982570797,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.3220405881577123e-05,"delta_db":-0.9578819413584732,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6702224744057894,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.556585708790133,"pitch":null,"output_rms_dbfs":-28.575628986538284,"output_peak":0.2883241631003222,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":10.064565135108191},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44560,"output_samples":44560,"intelligibility":{"estoi":{"value":0.6976281530462963,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044570533867492254,"rms_dbfs":-27.019043277748153,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.196863425855337,"mean":-0.0008769427424487349,"zero_energy":false},"output":{"rms":0.03725791526601682,"rms_dbfs":-28.575628986538284,"peak":0.2883241631003222,"peak_dbfs":-10.8023791979243,"crest_factor_db":17.773249788613988,"mean":-0.0005806950431614369,"zero_energy":false},"rms_difference_db":-1.5565857087901307},"full_scale":{"reference":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":31,"both_estimated_voiced_frames":19,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.4523809523809524,"absolute_difference_cents_median":17.791102773752275,"absolute_difference_cents_p95":60.61185940380614,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":0.00011061650326899306,"log_rms_mae_db":2.381258795574287,"rms_envelope_nrmse":0.21942501751756927,"envelope_correlation":0.9661901572047029,"rms_gain_error_db":-1.5565423568444294,"reference_mean_power":0.00199005080866686,"output_mean_power":0.0013906246680342313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":8.439637043999617e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970864143567599,"output_power":0.0003824904525994526,"reference_active":true,"delta_db":-1.9341659055651723},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010123102438186566,"output_power":9.293313620678232e-05,"reference_active":true,"delta_db":-0.371430381075582},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240115070482855e-05,"output_power":1.8237820699802e-05,"reference_active":true,"delta_db":-2.4958749240881004}],"active_band_count":3,"active_band_mae_db":1.6004904035762848}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":1.9812164858814008,"rms_envelope_nrmse":0.18663727861474708,"envelope_correlation":0.9854403682766899,"rms_gain_error_db":-1.556375754903903,"reference_mean_power":0.0020114637230104894,"output_mean_power":0.00140564168836085},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":6.839204647135985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005172552173341314,"output_power":0.00033305101483910185,"reference_active":true,"delta_db":-1.911941185714538},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.664160204550464e-05,"output_power":5.441861476286476e-05,"reference_active":true,"delta_db":-0.17388045869614757},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529860845744804e-06,"output_power":1.2345503587852988e-06,"reference_active":true,"delta_db":-3.115878235714567}],"active_band_count":3,"active_band_mae_db":1.7338999600417508}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":3.1244837452584444,"rms_envelope_nrmse":0.22089867377852393,"envelope_correlation":0.9576944472969873,"rms_gain_error_db":-0.7986156488923789,"reference_mean_power":0.000839400894204417,"output_mean_power":0.0006984058406363368},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.694834819211473e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042315114528166547,"output_power":0.00031715575100052664,"reference_active":true,"delta_db":-1.2522292976668659},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.939611213191948e-05,"output_power":6.026793241549925e-05,"reference_active":true,"delta_db":-0.6124884751479505},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4277196064029688e-05,"output_power":1.5754144666915107e-05,"reference_active":true,"delta_db":-1.878036965369856}],"active_band_count":3,"active_band_mae_db":1.2475849127282241}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":2.1182583743396597,"rms_envelope_nrmse":0.1528065205539828,"envelope_correlation":0.9814915120545106,"rms_gain_error_db":-0.7986056964750289,"reference_mean_power":0.0008485551352947845,"output_mean_power":0.0007060240521383907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":5.733703583211585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003764127527272362,"output_power":0.00028210652165141034,"reference_active":true,"delta_db":-1.2525120258657934},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8963437868750806e-05,"output_power":3.444167696047996e-05,"reference_active":true,"delta_db":-0.535729800538675},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2966305004996595e-06,"output_power":1.164978243185886e-06,"reference_active":true,"delta_db":-2.947733135204224}],"active_band_count":3,"active_band_mae_db":1.578658320536231}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":4.027119777177269,"rms_envelope_nrmse":0.3356073620213835,"envelope_correlation":0.9160874374317809,"rms_gain_error_db":-2.0059187727785868,"reference_mean_power":0.00013682778387507752,"output_mean_power":8.621491725269208e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.073610600879068e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.932287229550181e-05,"output_power":1.668128031939365e-05,"reference_active":true,"delta_db":-3.724158523835671},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.861650334659597e-06,"output_power":5.256461506189153e-06,"reference_active":true,"delta_db":-1.157350945516932},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.31740167391869e-06,"output_power":2.109218026599172e-06,"reference_active":true,"delta_db":-3.1110098116357428}],"active_band_count":3,"active_band_mae_db":2.664173093662782}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":3.621420619477345,"rms_envelope_nrmse":0.29456694563613073,"envelope_correlation":0.9403540375120136,"rms_gain_error_db":-2.005865335661961,"reference_mean_power":0.00013831808192096793,"output_mean_power":8.71550234969087e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":4.8587078894174116e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5019378943818874e-05,"output_power":1.4784794772062818e-05,"reference_active":true,"delta_db":-3.7449313930179065},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.168829067205391e-06,"output_power":2.914835604636327e-06,"reference_active":true,"delta_db":-1.5540002251195024},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163313597471507e-07,"output_power":1.558280292449909e-07,"reference_active":true,"delta_db":-3.074966699737228}],"active_band_count":3,"active_band_mae_db":2.791299439291546}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":4.525385794731676,"rms_envelope_nrmse":0.3224763878242546,"envelope_correlation":0.9561795091143331,"rms_gain_error_db":-2.154462131363883,"reference_mean_power":5.106513987055756e-05,"output_mean_power":3.1094123034782284e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":2.3044643706212166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.045287512435104e-05,"output_power":6.757204827934057e-06,"reference_active":true,"delta_db":-1.8946867842573503},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8556283872264966e-06,"output_power":1.94834514384727e-06,"reference_active":true,"delta_db":-1.6603579714162149},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6522678437727898e-06,"output_power":1.0630811602081758e-06,"reference_active":true,"delta_db":-1.9151402905311656}],"active_band_count":3,"active_band_mae_db":1.8233950154015768}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":3.8518213463510147,"rms_envelope_nrmse":0.2822010380672136,"envelope_correlation":0.975236241356661,"rms_gain_error_db":-2.1544324189056816,"reference_mean_power":5.1621714358619315e-05,"output_mean_power":3.143324238852921e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":2.0666699185875215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0052154991285664e-05,"output_power":6.378015559694904e-06,"reference_active":true,"delta_db":-1.9757360182839225},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8908095519422077e-06,"output_power":1.2127785376970603e-06,"reference_active":true,"delta_db":-1.9286628486184947},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.81443132062933e-07,"output_power":9.432445494706135e-08,"reference_active":true,"delta_db":-2.841162297348524}],"active_band_count":3,"active_band_mae_db":2.248520388083647}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0042","method":"golf","status":"measured","mos":{"status":"measured","value":1.4648687839508057,"predicted_mos":1.4648687839508057,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66720,"input_sample_rate":24000,"input_seconds":2.78,"input_float32_sha256":"b77600e12fa30383aae262e727b48913f359d92dfb88260d6be14632aa1ea3e4","internal_samples":44480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7039149859992904},"started_at":"2026-10-13T05:02:48.027527+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0019718859994100058,"native_input_samples":66840,"native_input_sha256":"d9ce927a0f6389019f5fd5eb32ea5689cd29942151955eafad42f89af16da52a","output_sha256":"8ca64d899d33d69292b7795f2c68db2d7e0cff1a1e4d34bd0daa7fb2a5d1867f","output_samples":66720,"length_delta_samples":-120,"analysis_s":0.25878236900007323,"synthesis_s":0.06548485999974218,"total_s":0.3242672289998154,"rtf":0.11643347540388345,"output_peak":0.30644506216049194,"output_rms":0.04140834213881772,"finished_at":"2026-10-13T05:02:48.358987+00:00","analysis_stages":{"external_f0_s":0.042402460000630526},"evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"8ca64d899d33d69292b7795f2c68db2d7e0cff1a1e4d34bd0daa7fb2a5d1867f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44560,"output_samples":44480,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.653797745776302,"output_peak":0.3054017185385563},"common_prefix":{"samples":44480,"sample_rate":16000,"reference_coverage":0.9982046678635548,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44480,"output_samples":44480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.497877907874589,"spectral_convergence":0.7973317176582457,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.733230289335998,"spectral_convergence":0.39356447600078964,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.0183206551167245,"spectral_convergence":0.4209371941302799,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005919232133471678,"output_power":0.0005342533917042738,"delta_db":-0.44518083994497837,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005990579422215635,"output_power":0.0005527390913807036,"delta_db":-0.34948650065203357,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005556454645885725,"output_power":0.00045376528957491387,"delta_db":-0.8796650200569485,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011416212543221813,"output_power":8.315937590325257e-05,"delta_db":-1.3761082436446148,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011273293208980102,"output_power":7.911244172582828e-05,"delta_db":-1.5380601400493532,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6512386224312142e-05,"output_power":1.3377205702523588e-05,"delta_db":-0.9144443289644149,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9171575122187239,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6424028497666658,"pitch":null,"output_rms_dbfs":-27.653797745776302,"output_peak":0.3054017185385563,"reference_rms_dbfs":-27.011394896009637,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.4164762841091045},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44480,"output_samples":44480,"intelligibility":{"estoi":{"value":0.8237483918110916,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04460979784647802,"rms_dbfs":-27.011394896009637,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.189215044116825,"mean":-0.0008894769407862382,"zero_energy":false},"output":{"rms":0.041429540137340355,"rms_dbfs":-27.653797745776302,"peak":0.3054017185385563,"peak_dbfs":-10.302570468728034,"crest_factor_db":17.35122727704827,"mean":0.000510407160795228,"zero_energy":false},"rms_difference_db":-0.6424028497666647},"full_scale":{"reference":{"sample_count":44480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":10,"both_estimated_voiced_frames":5,"reference_voiced_output_unvoiced_frames":37,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.11904761904761904,"absolute_difference_cents_median":7.251859045719598,"absolute_difference_cents_p95":9.443125900159819,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":555,"rms_floor":0.00011061650326899306,"log_rms_mae_db":2.0552600126984073,"rms_envelope_nrmse":0.17848948812402196,"envelope_correlation":0.9610356290218,"rms_gain_error_db":-0.6424293420051235,"reference_mean_power":0.001993592579247061,"output_mean_power":0.0017194655309145576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":555,"frequency_resolution_hz":0.3603603603603604,"power_floor":8.44420451913152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000597751822113587,"output_power":0.0005569824620393103,"reference_active":true,"delta_db":-0.3067938807935498},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010092864916858658,"output_power":8.812536283896657e-05,"reference_active":true,"delta_db":-0.5891354255875485},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2456741641339754e-05,"output_power":3.338602839917224e-05,"reference_active":true,"delta_db":0.1225983996726134}],"active_band_count":3,"active_band_mae_db":0.3395092353512372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":549,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":1.1873963265149345,"rms_envelope_nrmse":0.11194475305008829,"envelope_correlation":0.9869985999628712,"rms_gain_error_db":-0.6425124371762859,"reference_mean_power":0.0020150795455835206,"output_mean_power":0.0017379646989751757},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":549,"frequency_resolution_hz":0.36429872495446264,"power_floor":6.841540667962234e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005173125727379392,"output_power":0.0004882572481574245,"reference_active":true,"delta_db":-0.2510433451764155},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6736875816503644e-05,"output_power":4.939257043487072e-05,"reference_active":true,"delta_db":-0.6020379010727316},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.508344842536959e-06,"output_power":1.979713311432319e-06,"reference_active":true,"delta_db":-1.0278493900012486}],"active_band_count":3,"active_band_mae_db":0.6269768787501319}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":555,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":2.1276906592259466,"rms_envelope_nrmse":0.2157675104789048,"envelope_correlation":0.9578958188785606,"rms_gain_error_db":-0.5521597914806534,"reference_mean_power":0.00084091262358723,"output_mean_power":0.0007405167610611302},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":555,"frequency_resolution_hz":0.3603603603603604,"power_floor":6.700241126586565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000423686950876099,"output_power":0.00038814847815012436,"reference_active":true,"delta_db":-0.38047200658015023},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.922875692323124e-05,"output_power":6.183286504685904e-05,"reference_active":true,"delta_db":-0.49067163309236195},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.432737138600163e-05,"output_power":2.6783341832012932e-05,"reference_active":true,"delta_db":0.4176957898075901}],"active_band_count":3,"active_band_mae_db":0.42961314316003407}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":549,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":1.3421904615983107,"rms_envelope_nrmse":0.13448596573728358,"envelope_correlation":0.9837075355627821,"rms_gain_error_db":-0.5521557921834817,"reference_mean_power":0.0008501002633786055,"output_mean_power":0.0007486081855099937},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":549,"frequency_resolution_hz":0.36429872495446264,"power_floor":5.73707978038134e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037664017009863106,"output_power":0.00035135347044125404,"reference_active":true,"delta_db":-0.3018238972682805},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.895401622604751e-05,"output_power":3.281462182451122e-05,"reference_active":true,"delta_db":-0.7448483727448886},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.280766629034113e-06,"output_power":1.544616780389779e-06,"reference_active":true,"delta_db":-1.692601013703937}],"active_band_count":3,"active_band_mae_db":0.9130910945723687}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":555,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":2.102982399806825,"rms_envelope_nrmse":0.25814728425698297,"envelope_correlation":0.9459591392142177,"rms_gain_error_db":-1.4102397059081653,"reference_mean_power":0.00013707354848228024,"output_mean_power":9.906715362000564e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":555,"frequency_resolution_hz":0.3603603603603604,"power_floor":6.080578574048733e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.938341951810108e-05,"output_power":2.403273959469305e-05,"reference_active":true,"delta_db":-2.1451014092889875},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.839169133440182e-06,"output_power":4.454246002988451e-06,"reference_active":true,"delta_db":-1.8622914506304182},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.320121038017744e-06,"output_power":3.987641819552726e-06,"reference_active":true,"delta_db":-0.3477977281113764}],"active_band_count":3,"active_band_mae_db":1.4517301960102609}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":549,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":1.623450756410725,"rms_envelope_nrmse":0.1838292192108838,"envelope_correlation":0.9834773515796503,"rms_gain_error_db":-1.4102666966094213,"reference_mean_power":0.00013856958971342336,"output_mean_power":0.0001001477649630727},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":549,"frequency_resolution_hz":0.36429872495446264,"power_floor":4.864800061370029e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.504502242225172e-05,"output_power":2.128819954058116e-05,"reference_active":true,"delta_db":-2.164874098539469},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.169892883370514e-06,"output_power":2.9431778116867973e-06,"reference_active":true,"delta_db":-1.51308398182319},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.150098457063471e-07,"output_power":2.683638510274837e-07,"reference_active":true,"delta_db":-0.6960011259283149}],"active_band_count":3,"active_band_mae_db":1.4579864020969913}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":555,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":3.0914322618518644,"rms_envelope_nrmse":0.25831788732204186,"envelope_correlation":0.9685905761087624,"rms_gain_error_db":-1.3684475255514694,"reference_mean_power":5.115705480936164e-05,"output_mean_power":3.7330239906094566e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":555,"frequency_resolution_hz":0.3603603603603604,"power_floor":2.3155714108159272e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0506094157706596e-05,"output_power":7.757890485060942e-06,"reference_active":true,"delta_db":-1.3169764416680485},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8558720257200903e-06,"output_power":1.922740363131191e-06,"reference_active":true,"delta_db":-1.7181809907349783},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6495571973173027e-06,"output_power":1.1181905935032206e-06,"reference_active":true,"delta_db":-1.6885154447459862}],"active_band_count":3,"active_band_mae_db":1.5745576257163378}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44480,"uncovered_tail_samples":0,"metrics":{"frame_count":549,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":2.8054030190351766,"rms_envelope_nrmse":0.1780789627148782,"envelope_correlation":0.9934048884974898,"rms_gain_error_db":-1.36845590805846,"reference_mean_power":5.171567163506652e-05,"output_mean_power":3.773779999620356e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":549,"frequency_resolution_hz":0.36429872495446264,"power_floor":2.0775442768525136e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0102760506846079e-05,"output_power":7.457079393969393e-06,"reference_active":true,"delta_db":-1.3187129089229521},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8828389244852283e-06,"output_power":1.1738241082868705e-06,"reference_active":true,"delta_db":-2.0521014310994126},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7993601392734107e-07,"output_power":1.1262884039611984e-07,"reference_active":true,"delta_db":-2.0346848271671414}],"active_band_count":3,"active_band_mae_db":1.8018330557298352}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0042","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.5540692806243896,"predicted_mos":2.5540692806243896,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66960,"input_sample_rate":24000,"input_seconds":2.79,"input_float32_sha256":"c60a7ecf1d8970cfa60d0fefed99799f656b188e98aa8c9fada8871cfa0927da","internal_samples":44640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1339797340006044},"started_at":"2026-10-13T07:41:50.105102+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0014305889999377541,"native_input_samples":66840,"native_input_sha256":"d9ce927a0f6389019f5fd5eb32ea5689cd29942151955eafad42f89af16da52a","output_sha256":"b1edfdd6b0e1218b9e73a2912aa9ca51d03b050beeace7e042dbe6e67fd06db3","output_samples":66960,"length_delta_samples":120,"analysis_s":1.1390326149994507,"synthesis_s":5.505673187002685,"total_s":6.644705802002136,"rtf":2.3858907727117185,"output_peak":0.3627506196498871,"output_rms":0.045849950480529186,"finished_at":"2026-10-13T07:41:56.755729+00:00","analysis_stages":{"external_f0_s":0.7507137430002331},"evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"b1edfdd6b0e1218b9e73a2912aa9ca51d03b050beeace7e042dbe6e67fd06db3","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44560,"output_samples":44640,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.769485384530483,"output_peak":0.36263240830796445},"common_prefix":{"samples":44560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982078853046595,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.933767453585961,"spectral_convergence":0.8126134499269504,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.400947927966295,"spectral_convergence":0.41027610697846356,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.603042331299616,"spectral_convergence":0.4300976941936854,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0008535055069758548,"delta_db":1.5967025132382442,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.0005779870430036242,"delta_db":-0.14986103777480142,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.00047095401685085973,"delta_db":-0.7080761119249368,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":9.656160521139578e-05,"delta_db":-0.719343025558775,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":9.607823905126927e-05,"delta_db":-0.6864673832102418,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.2709927457695545e-05,"delta_db":-1.1288991101544439,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.831558196976907,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.25732973538961157,"pitch":null,"output_rms_dbfs":-26.761713542358542,"output_peak":0.36263240830796445,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.979252570950624},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44560,"output_samples":44560,"intelligibility":{"estoi":{"value":0.8606060909203059,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044570533867492254,"rms_dbfs":-27.019043277748153,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.196863425855337,"mean":-0.0008769427424487349,"zero_energy":false},"output":{"rms":0.04591074317140486,"rms_dbfs":-26.761713542358542,"peak":0.36263240830796445,"peak_dbfs":-8.810667715766034,"crest_factor_db":17.951045826592505,"mean":5.6378141518598314e-06,"zero_energy":false},"rms_difference_db":0.2573297353896109},"full_scale":{"reference":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":36,"both_estimated_voiced_frames":17,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.40476190476190477,"absolute_difference_cents_median":10.486699512369503,"absolute_difference_cents_p95":17.613539518014615,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":0.00011061650326899306,"log_rms_mae_db":2.884521388310395,"rms_envelope_nrmse":0.2029007691597883,"envelope_correlation":0.9462396697352244,"rms_gain_error_db":0.25743265731841125,"reference_mean_power":0.00199005080866686,"output_mean_power":0.0021115794676524365},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":8.439637043999617e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970864143567599,"output_power":0.000627984370244742,"reference_active":true,"delta_db":0.21911645125622703},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010123102438186566,"output_power":9.504629995073566e-05,"reference_active":true,"delta_db":-0.2737841668655667},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240115070482855e-05,"output_power":2.6380110029901116e-05,"reference_active":true,"delta_db":-0.8928383152253289}],"active_band_count":3,"active_band_mae_db":0.4619129777823742}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":2.2148451971158063,"rms_envelope_nrmse":0.1259998796601136,"envelope_correlation":0.9763455179731656,"rms_gain_error_db":0.25798259019682596,"reference_mean_power":0.0020114637230104894,"output_mean_power":0.0021345703049248186},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":6.839204647135985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005172552173341314,"output_power":0.0005410348811017683,"reference_active":true,"delta_db":0.19520385538178694},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.664160204550464e-05,"output_power":5.438148574942437e-05,"reference_active":true,"delta_db":-0.17684459681839318},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529860845744804e-06,"output_power":2.215886683479257e-06,"reference_active":true,"delta_db":-0.575490860183877}],"active_band_count":3,"active_band_mae_db":0.3158464374613524}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":2.5422362905253113,"rms_envelope_nrmse":0.24686914668958482,"envelope_correlation":0.9433353579846525,"rms_gain_error_db":-0.603871239242592,"reference_mean_power":0.000839400894204417,"output_mean_power":0.000730436225489025},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.694834819211473e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042315114528166547,"output_power":0.0003545942596318376,"reference_active":true,"delta_db":-0.7676381993202537},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.939611213191948e-05,"output_power":6.306511335186744e-05,"reference_active":true,"delta_db":-0.415459594355907},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4277196064029688e-05,"output_power":2.0094408757386697e-05,"reference_active":true,"delta_db":-0.8212329335599096}],"active_band_count":3,"active_band_mae_db":0.6681102424120233}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":1.5625586657964747,"rms_envelope_nrmse":0.1435709148229545,"envelope_correlation":0.9814188038951639,"rms_gain_error_db":-0.6038727908392647,"reference_mean_power":0.0008485551352947845,"output_mean_power":0.0007384018685030471},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":5.733703583211585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003764127527272362,"output_power":0.0003141160285748998,"reference_active":true,"delta_db":-0.7857423079142649},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8963437868750806e-05,"output_power":3.606707493977664e-05,"reference_active":true,"delta_db":-0.3354634687060031},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2966305004996595e-06,"output_power":1.7907780408634447e-06,"reference_active":true,"delta_db":-1.080493679179023}],"active_band_count":3,"active_band_mae_db":0.7338998185997636}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":2.2498921665520886,"rms_envelope_nrmse":0.2567508130309772,"envelope_correlation":0.9302729997440653,"rms_gain_error_db":-0.5948474199815822,"reference_mean_power":0.00013682778387507752,"output_mean_power":0.00011931349046457023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.073610600879068e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.932287229550181e-05,"output_power":3.344925704524583e-05,"reference_active":true,"delta_db":-0.7025875695299073},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.861650334659597e-06,"output_power":5.278645433081346e-06,"reference_active":true,"delta_db":-1.1390609151802937},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.31740167391869e-06,"output_power":3.444783803443093e-06,"reference_active":true,"delta_db":-0.9806048503659924}],"active_band_count":3,"active_band_mae_db":0.9407511116920645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":1.3646488377021784,"rms_envelope_nrmse":0.13754770999094051,"envelope_correlation":0.9806047297235188,"rms_gain_error_db":-0.5948479092940713,"reference_mean_power":0.00013831808192096793,"output_mean_power":0.00012061301309997392},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":4.8587078894174116e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5019378943818874e-05,"output_power":2.9750228965232463e-05,"reference_active":true,"delta_db":-0.7081812724134545},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.168829067205391e-06,"output_power":3.138639633646656e-06,"reference_active":true,"delta_db":-1.2327263372723758},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163313597471507e-07,"output_power":2.2941446552779675e-07,"reference_active":true,"delta_db":-1.3952144967385225}],"active_band_count":3,"active_band_mae_db":1.112040702141451}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":2.0584070098064284,"rms_envelope_nrmse":0.28374251786083554,"envelope_correlation":0.9551630224904605,"rms_gain_error_db":-1.2559343573895287,"reference_mean_power":5.106513987055756e-05,"output_mean_power":3.824116285790459e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":2.3044643706212166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.045287512435104e-05,"output_power":7.212161977576094e-06,"reference_active":true,"delta_db":-1.611702901547807},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8556283872264966e-06,"output_power":2.5692916381614653e-06,"reference_active":true,"delta_db":-0.45888287071412615},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6522678437727898e-06,"output_power":1.3277007944340188e-06,"reference_active":true,"delta_db":-0.9498023564182354}],"active_band_count":3,"active_band_mae_db":1.0067960428933895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":1.2509273457003847,"rms_envelope_nrmse":0.20483234688342575,"envelope_correlation":0.9824890361629398,"rms_gain_error_db":-1.2559506306873456,"reference_mean_power":5.1621714358619315e-05,"output_mean_power":3.865782006716108e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":2.0666699185875215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0052154991285664e-05,"output_power":6.972707480923958e-06,"reference_active":true,"delta_db":-1.5885773017207747},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8908095519422077e-06,"output_power":1.6767158911950638e-06,"reference_active":true,"delta_db":-0.5218830718175533},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.81443132062933e-07,"output_power":1.428206699063775e-07,"reference_active":true,"delta_db":-1.039494682286114}],"active_band_count":3,"active_band_mae_db":1.0499850186081474}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0042","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.173046350479126,"predicted_mos":2.173046350479126,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":67080,"input_sample_rate":24000,"input_seconds":2.795,"input_float32_sha256":"e559efb28f8949579d54a9d30c2c5cb6c45b393f1c20413820f6bc67ab72a5b5","internal_samples":44720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.81879188899984},"started_at":"2026-10-13T05:11:06.109784+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0016728650007280521,"native_input_samples":66840,"native_input_sha256":"d9ce927a0f6389019f5fd5eb32ea5689cd29942151955eafad42f89af16da52a","output_sha256":"ee6e14d693dec902952fbfcf58ed52f62f3f9937b890b846dcddbbf9068fa5df","output_samples":67080,"length_delta_samples":240,"analysis_s":0.5907558669987338,"synthesis_s":0.08091941900056554,"total_s":0.6716752859992994,"rtf":0.24117604524211825,"output_peak":0.36079463362693787,"output_rms":0.04314171580680026,"finished_at":"2026-10-13T05:11:06.787982+00:00","analysis_stages":{"external_f0_s":0.0861575350008934},"evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"ee6e14d693dec902952fbfcf58ed52f62f3f9937b890b846dcddbbf9068fa5df","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44560,"output_samples":44720,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.297584483898113,"output_peak":0.359222200788966},"common_prefix":{"samples":44560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9964221824686941,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.114257063010443,"spectral_convergence":0.2977551251132841,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.608809898106167,"spectral_convergence":0.26323335521390595,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.135006503347011,"spectral_convergence":0.29583670842105275,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0005652193659397876,"delta_db":-0.19319042752907034,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.000572331354436664,"delta_db":-0.1925666770736244,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.0005202372155916812,"delta_db":-0.27584700037301246,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":9.78500780479208e-05,"delta_db":-0.6617760288801986,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":9.905034108141606e-05,"delta_db":-0.554157975415107,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.5122980105626665e-05,"delta_db":-0.37395601986198446,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.37524902152216616,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.262978839848246,"pitch":null,"output_rms_dbfs":-27.2820221175964,"output_peak":0.359222200788966,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.952691154821207},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44560,"output_samples":44560,"intelligibility":{"estoi":{"value":0.9119747808127425,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044570533867492254,"rms_dbfs":-27.019043277748153,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.196863425855337,"mean":-0.0008769427424487349,"zero_energy":false},"output":{"rms":0.04324131514192295,"rms_dbfs":-27.2820221175964,"peak":0.359222200788966,"peak_dbfs":-8.89273661480534,"crest_factor_db":18.38928550279106,"mean":0.00019720886657110119,"zero_energy":false},"rms_difference_db":-0.2629788398482482},"full_scale":{"reference":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":36,"both_estimated_voiced_frames":29,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.6904761904761905,"absolute_difference_cents_median":1.2785517432987843,"absolute_difference_cents_p95":2.787025842985121,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":0.00011061650326899306,"log_rms_mae_db":1.1849728051275874,"rms_envelope_nrmse":0.09666662744283712,"envelope_correlation":0.9878368626036491,"rms_gain_error_db":-0.26298055477393767,"reference_mean_power":0.00199005080866686,"output_mean_power":0.0018731221918095356},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":8.439637043999617e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970864143567599,"output_power":0.000567275502640542,"reference_active":true,"delta_db":-0.22243160398857492},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010123102438186566,"output_power":8.649968787284546e-05,"reference_active":true,"delta_db":-0.6829909125343145},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240115070482855e-05,"output_power":1.960269856703865e-05,"reference_active":true,"delta_db":-2.1824457238380295}],"active_band_count":3,"active_band_mae_db":1.0292894134536397}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":0.5684558064689936,"rms_envelope_nrmse":0.05583853284535339,"envelope_correlation":0.9962074592914143,"rms_gain_error_db":-0.26315633210943024,"reference_mean_power":0.0020114637230104894,"output_mean_power":0.001893200328765982},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":6.839204647135985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005172552173341314,"output_power":0.0004950169452277583,"reference_active":true,"delta_db":-0.1908481411620947},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.664160204550464e-05,"output_power":4.455853812818716e-05,"reference_active":true,"delta_db":-1.0420459428596462},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529860845744804e-06,"output_power":1.4680242729202786e-06,"reference_active":true,"delta_db":-2.363633971480015}],"active_band_count":3,"active_band_mae_db":1.198842685167252}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":1.3237095002143133,"rms_envelope_nrmse":0.1231216473373409,"envelope_correlation":0.9862211778908463,"rms_gain_error_db":-0.2679123054827524,"reference_mean_power":0.000839400894204417,"output_mean_power":0.0007891838641478433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.694834819211473e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042315114528166547,"output_power":0.00039667590759483537,"reference_active":true,"delta_db":-0.28059696576363646},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.939611213191948e-05,"output_power":6.01034350764849e-05,"reference_active":true,"delta_db":-0.6243584626014352},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4277196064029688e-05,"output_power":1.5444475863370267e-05,"reference_active":true,"delta_db":-1.9642535141386028}],"active_band_count":3,"active_band_mae_db":0.9564029808345582}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":0.7070875186895536,"rms_envelope_nrmse":0.07122477397841658,"envelope_correlation":0.9953965850765782,"rms_gain_error_db":-0.26792309965728905,"reference_mean_power":0.0008485551352947845,"output_mean_power":0.0007977884712994476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":5.733703583211585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003764127527272362,"output_power":0.00035566711898140274,"reference_active":true,"delta_db":-0.24620611730498343},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8963437868750806e-05,"output_power":3.0519871208910044e-05,"reference_active":true,"delta_db":-1.060745724850897},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2966305004996595e-06,"output_power":1.2536620293793979e-06,"reference_active":true,"delta_db":-2.6291065594018947}],"active_band_count":3,"active_band_mae_db":1.312019467185925}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":1.3262547623800074,"rms_envelope_nrmse":0.16738852385906827,"envelope_correlation":0.9750889150268158,"rms_gain_error_db":-0.7200952577261804,"reference_mean_power":0.00013682778387507752,"output_mean_power":0.00011592170687182376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.073610600879068e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.932287229550181e-05,"output_power":3.20150876443446e-05,"reference_active":true,"delta_db":-0.8929053779223626},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.861650334659597e-06,"output_power":5.1669048129434775e-06,"reference_active":true,"delta_db":-1.2319812206317753},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.31740167391869e-06,"output_power":3.0500379759017727e-06,"reference_active":true,"delta_db":-1.509172088072956}],"active_band_count":3,"active_band_mae_db":1.2113528955423647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":0.8207210178493194,"rms_envelope_nrmse":0.11000214046808912,"envelope_correlation":0.9922119467095385,"rms_gain_error_db":-0.7201088870617772,"reference_mean_power":0.00013831808192096793,"output_mean_power":0.00011718393278217673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":4.8587078894174116e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5019378943818874e-05,"output_power":2.8522026245703685e-05,"reference_active":true,"delta_db":-0.8912806457869881},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.168829067205391e-06,"output_power":2.9704222333792963e-06,"reference_active":true,"delta_db":-1.4719590140152166},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163313597471507e-07,"output_power":2.259452015213991e-07,"reference_active":true,"delta_db":-1.4613912566111773}],"active_band_count":3,"active_band_mae_db":1.274876972137794}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":1.190459338776959,"rms_envelope_nrmse":0.1328101996066677,"envelope_correlation":0.9896789917598166,"rms_gain_error_db":-0.3664969229580836,"reference_mean_power":5.106513987055756e-05,"output_mean_power":4.6932623701831844e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":2.3044643706212166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.045287512435104e-05,"output_power":9.728552465302566e-06,"reference_active":true,"delta_db":-0.3118753675634422},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8556283872264966e-06,"output_power":3.0798053740238083e-06,"reference_active":true,"delta_db":0.32821581913612996},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6522678437727898e-06,"output_power":1.2651616035649912e-06,"reference_active":true,"delta_db":-1.1593444773413168}],"active_band_count":3,"active_band_mae_db":0.5998118880136297}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":0.7310594910653925,"rms_envelope_nrmse":0.07321810863382594,"envelope_correlation":0.9975347991588871,"rms_gain_error_db":-0.36650655801777593,"reference_mean_power":5.1621714358619315e-05,"output_mean_power":4.74440513821669e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":2.0666699185875215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0052154991285664e-05,"output_power":9.391858820412046e-06,"reference_active":true,"delta_db":-0.2950762064979753},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8908095519422077e-06,"output_power":1.8599222527768146e-06,"reference_active":true,"delta_db":-0.07152997083374028},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.81443132062933e-07,"output_power":1.214348111450468e-07,"reference_active":true,"delta_db":-1.7439733230906727}],"active_band_count":3,"active_band_mae_db":0.7035265001407961}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0042","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1998233795166016,"predicted_mos":3.1998233795166016,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":135477,"input_sample_rate":48000,"input_seconds":2.8224375,"input_float32_sha256":"33a0fdb844ca242bdc3289d1f3d7efdefe6a71c6ae986046a28d709e87c18b27","internal_samples":45159,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.684405596999568},"started_at":"2026-10-13T05:16:59.836465+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.004606189000696759,"native_input_samples":133680,"native_input_sha256":"f45e52843151f50551103f3e0f33477b1abe4ae637b6abec2ec0fdb5750a1ce5","output_sha256":"120344e2873d88a75c51cfdbac21ca87dd5eb0c26bcb36acb677dcbcc60b1db9","output_samples":135477,"length_delta_samples":1797,"analysis_s":1.2572394760009047,"synthesis_s":0.5725721590006287,"total_s":1.8298116350015334,"rtf":0.657023926391933,"output_peak":0.3734897102514392,"output_rms":0.04311205209561156,"finished_at":"2026-10-13T05:17:01.678410+00:00","evaluation_input":{"reference":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"120344e2873d88a75c51cfdbac21ca87dd5eb0c26bcb36acb677dcbcc60b1db9","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":44560,"output_samples":45159,"length_delta_samples":599,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.303409937493722,"output_peak":0.3666808818373143},"common_prefix":{"samples":44560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9867357558847627,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44560,"output_samples":44560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.895909518218357,"spectral_convergence":0.845536716890379,"spectral_convergence_status":"ok","amplitude_floor":9.278908430047933e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.983516574356813,"spectral_convergence":0.558002374822352,"spectral_convergence_status":"ok","amplitude_floor":4.74996132181228e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.095179216564239,"spectral_convergence":0.5113424001928134,"spectral_convergence_status":"ok","amplitude_floor":3.301834166389807e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005909300513360037,"output_power":0.0005335904613586355,"delta_db":-0.44328018899902244,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.000598279616407713,"output_power":0.0005761826240146404,"delta_db":-0.1634404944320852,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005543526355345429,"output_power":0.0005462869643718814,"delta_db":-0.06365279215785422,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00011395641530102644,"output_power":0.00010899926113612063,"delta_db":-0.19315225251335544,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011253089498164016,"output_power":0.00010396723679495436,"delta_db":-0.34375271320743295,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6482875672349115e-05,"output_power":1.652574259103522e-05,"delta_db":0.011280009698618131,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20309307516806138,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.22657091431761398,"pitch":null,"output_rms_dbfs":-27.245614192065766,"output_peak":0.3666808818373143,"reference_rms_dbfs":-27.019043277748153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.99153510304647},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44560,"output_samples":44560,"intelligibility":{"estoi":{"value":0.573556723635585,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.044570533867492254,"rms_dbfs":-27.019043277748153,"peak":0.362152099609375,"peak_dbfs":-8.822179851892814,"crest_factor_db":18.196863425855337,"mean":-0.0008769427424487349,"zero_energy":false},"output":{"rms":0.04342294658665239,"rms_dbfs":-27.245614192065766,"peak":0.3666808818373143,"peak_dbfs":-8.714234657564884,"crest_factor_db":18.531379534500886,"mean":-0.00047190711387401285,"zero_energy":false},"rms_difference_db":-0.22657091431761245},"full_scale":{"reference":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":278,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1e-06,"threshold_dbfs":-60.0,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":271,"reference_estimated_voiced_frames":42,"output_estimated_voiced_frames":55,"both_estimated_voiced_frames":26,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.6190476190476191,"absolute_difference_cents_median":10.772366957685616,"absolute_difference_cents_p95":31.234754933659378,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":0.00011061650326899306,"log_rms_mae_db":4.048810875981899,"rms_envelope_nrmse":0.39339300869673727,"envelope_correlation":0.7874269825096389,"rms_gain_error_db":-0.22650023434732783,"reference_mean_power":0.00199005080866686,"output_mean_power":0.0018889225064369767},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":8.439637043999617e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005970864143567599,"output_power":0.0006392673041638753,"reference_active":true,"delta_db":0.2964530297133595},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010123102438186566,"output_power":9.607433741393779e-05,"reference_active":true,"delta_db":-0.2270623361207127},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.240115070482855e-05,"output_power":2.8978453167719986e-05,"reference_active":true,"delta_db":-0.48485234476463906}],"active_band_count":3,"active_band_mae_db":0.3361225701995704}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":9.74750993349889e-05,"log_rms_mae_db":2.9622916765166116,"rms_envelope_nrmse":0.28829694031366426,"envelope_correlation":0.8641339627402193,"rms_gain_error_db":-0.22610578603376455,"reference_mean_power":0.0020114637230104894,"output_mean_power":0.001909420697306453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":6.839204647135985e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005172552173341314,"output_power":0.000552477845136459,"reference_active":true,"delta_db":0.2860998719880744},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.664160204550464e-05,"output_power":5.267589675808722e-05,"reference_active":true,"delta_db":-0.31523590595681933},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529860845744804e-06,"output_power":2.205879263703676e-06,"reference_active":true,"delta_db":-0.5951489545251073}],"active_band_count":3,"active_band_mae_db":0.398828244156667}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":8.308955229004579e-05,"log_rms_mae_db":5.590661682979192,"rms_envelope_nrmse":0.49401317528048994,"envelope_correlation":0.7707665898324575,"rms_gain_error_db":-0.08525861122040226,"reference_mean_power":0.000839400894204417,"output_mean_power":0.0008230828759824637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.694834819211473e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042315114528166547,"output_power":0.0004429901060830687,"reference_active":true,"delta_db":0.1989850597162615},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.939611213191948e-05,"output_power":6.75321045252174e-05,"reference_active":true,"delta_db":-0.11824856102273669},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4277196064029688e-05,"output_power":1.944031440766765e-05,"reference_active":true,"delta_db":-0.9649524124151323}],"active_band_count":3,"active_band_mae_db":0.4273953443847101}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":8.008999734930271e-05,"log_rms_mae_db":4.320052173654857,"rms_envelope_nrmse":0.3578871196990763,"envelope_correlation":0.8617496112874752,"rms_gain_error_db":-0.08525282801124237,"reference_mean_power":0.0008485551352947845,"output_mean_power":0.0008320602659171952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":5.733703583211585e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003764127527272362,"output_power":0.0003938934667982405,"reference_active":true,"delta_db":0.19714448823539707},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8963437868750806e-05,"output_power":3.500108574265924e-05,"reference_active":true,"delta_db":-0.4657575260297559},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2966305004996595e-06,"output_power":1.8711737090774366e-06,"reference_active":true,"delta_db":-0.8897702140338969}],"active_band_count":3,"active_band_mae_db":0.5175574094330165}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.2528786526255e-05,"log_rms_mae_db":4.984636353148962,"rms_envelope_nrmse":0.43644543532690394,"envelope_correlation":0.7964485415029743,"rms_gain_error_db":-0.24560341729961677,"reference_mean_power":0.00013682778387507752,"output_mean_power":0.00012930459288800132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":6.073610600879068e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.932287229550181e-05,"output_power":4.029592588108211e-05,"reference_active":true,"delta_db":0.10615906147833976},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.861650334659597e-06,"output_power":6.408042487599921e-06,"reference_active":true,"delta_db":-0.2970320030069099},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.31740167391869e-06,"output_power":3.247791099615341e-06,"reference_active":true,"delta_db":-1.2363436825380738}],"active_band_count":3,"active_band_mae_db":0.5465115823411079}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":3.2687349322598056e-05,"log_rms_mae_db":3.4601193816260936,"rms_envelope_nrmse":0.2904285501551169,"envelope_correlation":0.89451681129365,"rms_gain_error_db":-0.24556486136176947,"reference_mean_power":0.00013831808192096793,"output_mean_power":0.00013071411044969183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":4.8587078894174116e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.5019378943818874e-05,"output_power":3.589882684074811e-05,"reference_active":true,"delta_db":0.10771816482707691},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.168829067205391e-06,"output_power":3.7854775208766475e-06,"reference_active":true,"delta_db":-0.41893416714795656},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163313597471507e-07,"output_power":2.2014677374708758e-07,"reference_active":true,"delta_db":-1.5742992290719415}],"active_band_count":3,"active_band_mae_db":0.7003171870156583}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":556,"rms_floor":4.9606629692154145e-05,"log_rms_mae_db":4.559204310635997,"rms_envelope_nrmse":0.43439876670252536,"envelope_correlation":0.8812195343344444,"rms_gain_error_db":-0.12250672672373288,"reference_mean_power":5.106513987055756e-05,"output_mean_power":4.9644810025588455e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":556,"frequency_resolution_hz":0.3597122302158273,"power_floor":2.3044643706212166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.045287512435104e-05,"output_power":1.0706090620066173e-05,"reference_active":true,"delta_db":0.1039515296999163},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8556283872264966e-06,"output_power":3.330013707027469e-06,"reference_active":true,"delta_db":0.6674433061424624},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6522678437727898e-06,"output_power":1.9827507909865603e-06,"reference_active":true,"delta_db":0.7918768104556112}],"active_band_count":3,"active_band_mae_db":0.5210905487659966}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44560,"uncovered_tail_samples":0,"metrics":{"frame_count":550,"rms_floor":4.1718716562101094e-05,"log_rms_mae_db":3.203075062333379,"rms_envelope_nrmse":0.2909818167612776,"envelope_correlation":0.9442594776108676,"rms_gain_error_db":-0.12250892412794077,"reference_mean_power":5.1621714358619315e-05,"output_mean_power":5.018587851414941e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":550,"frequency_resolution_hz":0.36363636363636365,"power_floor":2.0666699185875215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0052154991285664e-05,"output_power":1.0575089280198018e-05,"reference_active":true,"delta_db":0.22024866360172232},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8908095519422077e-06,"output_power":2.2692159296966493e-06,"reference_active":true,"delta_db":0.7922803601859361},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.81443132062933e-07,"output_power":1.973260090212821e-07,"reference_active":true,"delta_db":0.36443798187598186}],"active_band_count":3,"active_band_mae_db":0.4589890018878801}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"hnm","status":"measured","mos":{"status":"measured","value":1.981201171875,"predicted_mos":1.981201171875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c2f2094c773e822dbe969f002d415515755f9baea2edadea665a03ce088bc36a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8720714610008145},"started_at":"2026-10-13T04:39:46.227426+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.451800017093774e-05,"native_input_samples":96000,"native_input_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","output_sha256":"0a76e878224253e749a6d7b5b9c31d00f725ec411f06aea387fe071aab2d1eef","output_samples":96000,"length_delta_samples":0,"analysis_s":0.9793304000004355,"synthesis_s":0.49170025499915937,"total_s":1.4710306549995948,"rtf":0.2451717758332658,"output_peak":0.49863895985609774,"output_rms":0.04122845755541206,"finished_at":"2026-10-13T04:39:47.704250+00:00","evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"0a76e878224253e749a6d7b5b9c31d00f725ec411f06aea387fe071aab2d1eef","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.275687321928733,"spectral_convergence":0.49994200944133854,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.780329722497737,"spectral_convergence":0.42387657836330545,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.087316512812587,"spectral_convergence":0.46329161790696516,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.0003228943135469158,"delta_db":0.35473500364645255,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0006525963313760851,"delta_db":-0.2334949931960599,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.0003576746380840471,"delta_db":-0.29178251069203975,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00017095081793754594,"delta_db":-0.5279155843681703,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":0.00011244836659579097,"delta_db":0.13841921523665207,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":8.322124485802838e-05,"delta_db":0.12674527569946673,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27884876380647355,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13025289006720653,"pitch":null,"output_rms_dbfs":-27.6960582555633,"output_peak":0.49863895985609774,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.047777852413019},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.275687321928733,"spectral_convergence":0.49994200944133854,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.780329722497737,"spectral_convergence":0.42387657836330545,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.087316512812587,"spectral_convergence":0.46329161790696516,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.0003228943135469158,"delta_db":0.35473500364645255,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0006525963313760851,"delta_db":-0.2334949931960599,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.0003576746380840471,"delta_db":-0.29178251069203975,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00017095081793754594,"delta_db":-0.5279155843681703,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":0.00011244836659579097,"delta_db":0.13841921523665207,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":8.322124485802838e-05,"delta_db":0.12674527569946673,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27884876380647355,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13025289006720653,"pitch":null,"output_rms_dbfs":-27.6960582555633,"output_peak":0.49863895985609774,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.047777852413019},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7541467391368752,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041851375067044644,"rms_dbfs":-27.565805365496097,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.213327125010572,"mean":-0.000893246332804362,"zero_energy":false},"output":{"rms":0.04122845755541206,"rms_dbfs":-27.6960582555633,"peak":0.49863895985609774,"peak_dbfs":-6.044275840772475,"crest_factor_db":21.65178241479083,"mean":-6.583976499326191e-05,"zero_energy":false},"rms_difference_db":-0.13025289006720442},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":32,"both_estimated_voiced_frames":20,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.8695652173913043,"absolute_difference_cents_median":1.8313747389803532,"absolute_difference_cents_p95":4.026703124105044,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017313921740261034,"log_rms_mae_db":1.3992768279909948,"rms_envelope_nrmse":0.1396826992893491,"envelope_correlation":0.978753996841953,"rms_gain_error_db":-0.12366431988333354,"reference_mean_power":0.0017487207983142375,"output_mean_power":0.0016996286504744765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.577112847045013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002769005691966373,"output_power":0.0002742405628850231,"reference_active":true,"delta_db":-0.04192156838589289},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.377497445156709e-05,"output_power":3.168221011691158e-05,"reference_active":true,"delta_db":-0.2777955929015041},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495313749243886e-05,"output_power":1.4359203318936515e-05,"reference_active":true,"delta_db":-0.17601981684803764}],"active_band_count":3,"active_band_mae_db":0.1652456593784782}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015508669364693045,"log_rms_mae_db":0.7186029747858197,"rms_envelope_nrmse":0.06811785727643552,"envelope_correlation":0.9946886830438042,"rms_gain_error_db":-0.12189869441262668,"reference_mean_power":0.0017487025727625826,"output_mean_power":0.0017003020543667259},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.09499315514524e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002529678773025127,"output_power":0.00025100502903927733,"reference_active":true,"delta_db":-0.03382953603903212},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6687984786311574e-05,"output_power":1.5789522987122014e-05,"reference_active":true,"delta_db":-0.24034885298870096},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969938613307946e-07,"output_power":6.857067300070185e-07,"reference_active":true,"delta_db":-0.07090541174059724}],"active_band_count":3,"active_band_mae_db":0.11502793358944342}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014635391485956695,"log_rms_mae_db":2.1086187563351926,"rms_envelope_nrmse":0.18565156168070868,"envelope_correlation":0.9740064660180818,"rms_gain_error_db":-0.2441391427119673,"reference_mean_power":0.0008142561216173605,"output_mean_power":0.0007697454279806356},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2354521678013805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001985819153870851,"output_power":0.00018244922848057796,"reference_active":true,"delta_db":-0.3679766407909558},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3653932189965355e-05,"output_power":2.1142677678842874e-05,"reference_active":true,"delta_db":-0.4874335845798832},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.646479803369055e-06,"output_power":5.397231045848639e-06,"reference_active":true,"delta_db":-0.9042067956398399}],"active_band_count":3,"active_band_mae_db":0.5865390070035597}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012544182666407365,"log_rms_mae_db":1.3486290488761927,"rms_envelope_nrmse":0.09386244525716969,"envelope_correlation":0.993503115756249,"rms_gain_error_db":-0.2405197642067713,"reference_mean_power":0.0008126713020357493,"output_mean_power":0.0007688877598668505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.0135181965386276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018492598004057708,"output_power":0.0001693609001585112,"reference_active":true,"delta_db":-0.38184775676854477},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0725232186915858e-05,"output_power":9.937887987676263e-06,"reference_active":true,"delta_db":-0.33112605452922195},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.105637817013121e-07,"output_power":3.5951318098714846e-07,"reference_active":true,"delta_db":-0.5766581726084592}],"active_band_count":3,"active_band_mae_db":0.42987732796874195}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":1.932223045188509,"rms_envelope_nrmse":0.2265394153479174,"envelope_correlation":0.959353961379537,"rms_gain_error_db":-0.41440711598694174,"reference_mean_power":0.00024251974980034777,"output_mean_power":0.00022044812022265826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1105776313023488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.895741122613483e-05,"output_power":3.127676853700807e-05,"reference_active":true,"delta_db":-0.9536821356434934},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.645635870343922e-06,"output_power":4.874251281110022e-06,"reference_active":true,"delta_db":-0.6380494928075375},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.039589927108398e-06,"output_power":2.9375027998469804e-06,"reference_active":true,"delta_db":-1.3835899137468872}],"active_band_count":3,"active_band_mae_db":0.9917738473993061}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":1.0868891050940641,"rms_envelope_nrmse":0.1167717635816527,"envelope_correlation":0.9900108732117373,"rms_gain_error_db":-0.4160635697946375,"reference_mean_power":0.00024219152470715302,"output_mean_power":0.0002200658148891828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0514876259433796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707790652686e-05,"output_power":2.93962918145227e-05,"reference_active":true,"delta_db":-1.0082265532553343},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.355351339185193e-06,"output_power":2.620916471013534e-06,"reference_active":true,"delta_db":-1.0728482176677778},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8063796091292895e-07,"output_power":1.7122266066162327e-07,"reference_active":true,"delta_db":-2.1458517480573027}],"active_band_count":3,"active_band_mae_db":1.408975506326805}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":1.9430184320732684,"rms_envelope_nrmse":0.2370391716864447,"envelope_correlation":0.9631965566172674,"rms_gain_error_db":0.19754770053836687,"reference_mean_power":0.00010238046641023414,"output_mean_power":0.00010714499126368931},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5067174911165995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.703714317319113e-05,"output_power":2.7568067254995434e-05,"reference_active":true,"delta_db":0.08445518791598897},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.303852533122451e-06,"output_power":4.468764646331344e-06,"reference_active":true,"delta_db":0.16330100263306652},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8801329619345514e-06,"output_power":2.480771852393038e-06,"reference_active":true,"delta_db":1.2039826220863896}],"active_band_count":3,"active_band_mae_db":0.48391293754514836}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":1.5097171203293502,"rms_envelope_nrmse":0.17259082217501956,"envelope_correlation":0.9796503464786794,"rms_gain_error_db":0.1974501805708508,"reference_mean_power":0.00010289230869716874,"output_mean_power":0.00010767823545990396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2184431092042676e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.696472252022106e-05,"output_power":2.7380089406563427e-05,"reference_active":true,"delta_db":0.06638906388398019},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.97826680457688e-06,"output_power":2.390266367699385e-06,"reference_active":true,"delta_db":0.82161437199455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5223765481589494e-07,"output_power":1.935132454930839e-07,"reference_active":true,"delta_db":1.0418861160367203}],"active_band_count":3,"active_band_mae_db":0.6432965173050835}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9246792793273926,"predicted_mos":2.9246792793273926,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"204003bf8a32f19bcbae0239289a00b6b77a432a4bc8751abf29b05b2d1bd421","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.553435396999703},"started_at":"2026-10-13T04:42:53.050626+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00010793400088005001,"native_input_samples":96000,"native_input_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","output_sha256":"e99ff4cfff43134178d9000da60c458daded1e6325c041dccce39214a9906895","output_samples":96000,"length_delta_samples":0,"analysis_s":2.8459317590004503,"synthesis_s":0.09660375400017074,"total_s":2.942535513000621,"rtf":0.4904225855001035,"output_peak":0.5715834860490729,"output_rms":0.043010243385843104,"finished_at":"2026-10-13T04:42:56.004474+00:00","evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"e99ff4cfff43134178d9000da60c458daded1e6325c041dccce39214a9906895","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.9658122420916415,"spectral_convergence":0.790335697292477,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.80968873222118,"spectral_convergence":0.41495629202386713,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.968197613272243,"spectral_convergence":0.4390275777840078,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.00026136382873588626,"delta_db":-0.5634141288939718,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0007795983099044921,"delta_db":0.5387676158098497,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.00039584203280116433,"delta_db":0.14855510773404035,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00020267049587447176,"delta_db":0.211277885708184,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":0.00012131890520717767,"delta_db":0.46817252731095127,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":8.908746358626811e-05,"delta_db":0.42256913997683393,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3921260675723052,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.2372433683238359,"pitch":null,"output_rms_dbfs":-27.32856199717226,"output_peak":0.5715834860490729,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.2478995291950215},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.9658122420916415,"spectral_convergence":0.790335697292477,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.80968873222118,"spectral_convergence":0.41495629202386713,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.968197613272243,"spectral_convergence":0.4390275777840078,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.00026136382873588626,"delta_db":-0.5634141288939718,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0007795983099044921,"delta_db":0.5387676158098497,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.00039584203280116433,"delta_db":0.14855510773404035,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00020267049587447176,"delta_db":0.211277885708184,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":0.00012131890520717767,"delta_db":0.46817252731095127,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":8.908746358626811e-05,"delta_db":0.42256913997683393,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3921260675723052,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.2372433683238359,"pitch":null,"output_rms_dbfs":-27.32856199717226,"output_peak":0.5715834860490729,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.2478995291950215},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.80738385539494,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041851375067044644,"rms_dbfs":-27.565805365496097,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.213327125010572,"mean":-0.000893246332804362,"zero_energy":false},"output":{"rms":0.043010243385843104,"rms_dbfs":-27.328561997172258,"peak":0.5715834860490729,"peak_dbfs":-4.8584065430854855,"crest_factor_db":22.470155454086775,"mean":1.3486091138381408e-06,"zero_energy":false},"rms_difference_db":0.2372433683238384},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":54,"both_estimated_voiced_frames":22,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9565217391304348,"absolute_difference_cents_median":5.546112767146585,"absolute_difference_cents_p95":8.915422730118374,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017313921740261034,"log_rms_mae_db":2.7847815454459037,"rms_envelope_nrmse":0.2092383167570692,"envelope_correlation":0.958087775686118,"rms_gain_error_db":0.24571767352173482,"reference_mean_power":0.0017487207983142375,"output_mean_power":0.0018505134358307784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.577112847045013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002769005691966373,"output_power":0.00033059320370038714,"reference_active":true,"delta_db":0.7697007266587732},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.377497445156709e-05,"output_power":3.5963576623244915e-05,"reference_active":true,"delta_db":0.2726784716487712},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495313749243886e-05,"output_power":1.5529502816166724e-05,"reference_active":true,"delta_db":0.16425225203314506}],"active_band_count":3,"active_band_mae_db":0.40221048344689647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015508669364693045,"log_rms_mae_db":1.9246451965638682,"rms_envelope_nrmse":0.11155728980148004,"envelope_correlation":0.9890655632416251,"rms_gain_error_db":0.25589658520405917,"reference_mean_power":0.0017487025727625826,"output_mean_power":0.0018548363890996233},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.09499315514524e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002529678773025127,"output_power":0.00030321211171825723,"reference_active":true,"delta_db":0.7868116854364298},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6687984786311574e-05,"output_power":1.7627934650102555e-05,"reference_active":true,"delta_db":0.2379753667796573},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969938613307946e-07,"output_power":6.896957535388387e-07,"reference_active":true,"delta_db":-0.04571401087874935}],"active_band_count":3,"active_band_mae_db":0.3568336876982788}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014635391485956695,"log_rms_mae_db":3.5039488404477566,"rms_envelope_nrmse":0.2338572551831651,"envelope_correlation":0.9633111406595823,"rms_gain_error_db":0.4744985199977981,"reference_mean_power":0.0008142561216173605,"output_mean_power":0.0009082614558401089},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2354521678013805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001985819153870851,"output_power":0.00023423996874666236,"reference_active":true,"delta_db":0.7172130614159009},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3653932189965355e-05,"output_power":2.1897088001617414e-05,"reference_active":true,"delta_db":-0.3351698371271095},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.646479803369055e-06,"output_power":5.4103723639690494e-06,"reference_active":true,"delta_db":-0.8936453334156949}],"active_band_count":3,"active_band_mae_db":0.6486760773195684}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012544182666407365,"log_rms_mae_db":2.829614009891701,"rms_envelope_nrmse":0.15624199483389184,"envelope_correlation":0.9840767997680081,"rms_gain_error_db":0.4879642791842836,"reference_mean_power":0.0008126713020357493,"output_mean_power":0.0009093087111729296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.0135181965386276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018492598004057708,"output_power":0.00021523564279681839,"reference_active":true,"delta_db":0.6591626267760309},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0725232186915858e-05,"output_power":9.905505147501966e-06,"reference_active":true,"delta_db":-0.3453007485292911},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.105637817013121e-07,"output_power":2.777151103924249e-07,"reference_active":true,"delta_db":-1.6978112477769733}],"active_band_count":3,"active_band_mae_db":0.9007582076940984}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":1.8094749703380022,"rms_envelope_nrmse":0.26470970838437585,"envelope_correlation":0.9472324090548311,"rms_gain_error_db":0.28080056939459797,"reference_mean_power":0.00024251974980034777,"output_mean_power":0.00025871831165761347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1105776313023488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.895741122613483e-05,"output_power":4.185767548948846e-05,"reference_active":true,"delta_db":0.3118501700774117},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.645635870343922e-06,"output_power":5.4918137934317175e-06,"reference_active":true,"delta_db":-0.11997060249796965},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.039589927108398e-06,"output_power":3.4340061482587067e-06,"reference_active":true,"delta_db":-0.7053621220924773}],"active_band_count":3,"active_band_mae_db":0.3790609648892862}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":1.0338884150162424,"rms_envelope_nrmse":0.12000460739709984,"envelope_correlation":0.9892457777488181,"rms_gain_error_db":0.3005782498575591,"reference_mean_power":0.00024219152470715302,"output_mean_power":0.0002595474499394971},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0514876259433796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707790652686e-05,"output_power":3.951247756328233e-05,"reference_active":true,"delta_db":0.27619056979021367},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.355351339185193e-06,"output_power":2.7418517885782406e-06,"reference_active":true,"delta_db":-0.8769402670505144},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8063796091292895e-07,"output_power":2.0764190869214883e-07,"reference_active":true,"delta_db":-1.308314037977264}],"active_band_count":3,"active_band_mae_db":0.8204816249393306}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":1.7243682254578565,"rms_envelope_nrmse":0.24248288191732684,"envelope_correlation":0.9642810097538582,"rms_gain_error_db":0.4545014349086284,"reference_mean_power":0.00010238046641023414,"output_mean_power":0.00011367559791635241},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5067174911165995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.703714317319113e-05,"output_power":3.235587042081594e-05,"reference_active":true,"delta_db":0.779922866623989},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.303852533122451e-06,"output_power":5.576962277581558e-06,"reference_active":true,"delta_db":1.1254032466565014},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8801329619345514e-06,"output_power":2.088574918831545e-06,"reference_active":true,"delta_db":0.45661494957873644}],"active_band_count":3,"active_band_mae_db":0.7873136876197423}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":1.0312659054741866,"rms_envelope_nrmse":0.14088661341968736,"envelope_correlation":0.9885364228107103,"rms_gain_error_db":0.45456100663551435,"reference_mean_power":0.00010289230869716874,"output_mean_power":0.00011424547631902322},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2184431092042676e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.696472252022106e-05,"output_power":3.273589924050569e-05,"reference_active":true,"delta_db":0.8422831974592773},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.97826680457688e-06,"output_power":2.3981469835571273e-06,"reference_active":true,"delta_db":0.8359093412044921},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5223765481589494e-07,"output_power":1.2543789546475219e-07,"reference_active":true,"delta_db":-0.8409332610695468}],"active_band_count":3,"active_band_mae_db":0.8397085999111052}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"facodec","status":"measured","mos":{"status":"measured","value":4.174256324768066,"predicted_mos":4.174256324768066,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c9e27818bb73dd0372b8c256956e9eea8156dbe42a5741a0096ea41aa0164076","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.387710140999843},"started_at":"2026-10-13T06:56:15.620241+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.238499994040467e-05,"native_input_samples":96000,"native_input_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","output_sha256":"d96e067a3e2ec8eb436289e674ec4fdbcb8624f6293a850e99817f4c9c5d8f63","output_samples":96000,"length_delta_samples":0,"analysis_s":1.9646230899998045,"synthesis_s":13.250902951998796,"total_s":15.2155260419986,"rtf":2.5359210069997666,"output_peak":0.45278966426849365,"output_rms":0.040127445416730356,"finished_at":"2026-10-13T06:56:30.846771+00:00","evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"d96e067a3e2ec8eb436289e674ec4fdbcb8624f6293a850e99817f4c9c5d8f63","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.48363687008839,"spectral_convergence":0.3017979566823595,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.794274059916019,"spectral_convergence":0.3190423542574073,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.933788064744493,"spectral_convergence":0.3535857225958439,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.00030581359334409005,"delta_db":0.11869889383715748,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0006772951606743135,"delta_db":-0.07216154806016828,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.00034996263892630997,"delta_db":-0.3864471474098199,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00012991415173262426,"delta_db":-1.7200627955252557,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":8.301006473328531e-05,"delta_db":-1.17980477278114,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":6.42162662620506e-05,"delta_db":-0.9991462959106971,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7460535755873732,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3653643811335373,"pitch":null,"output_rms_dbfs":-27.93116974662963,"output_peak":0.45278966426849365,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.737232998249635},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.48363687008839,"spectral_convergence":0.3017979566823595,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.794274059916019,"spectral_convergence":0.3190423542574073,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.933788064744493,"spectral_convergence":0.3535857225958439,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.00030581359334409005,"delta_db":0.11869889383715748,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0006772951606743135,"delta_db":-0.07216154806016828,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.00034996263892630997,"delta_db":-0.3864471474098199,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00012991415173262426,"delta_db":-1.7200627955252557,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":8.301006473328531e-05,"delta_db":-1.17980477278114,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":6.42162662620506e-05,"delta_db":-0.9991462959106971,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7460535755873732,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3653643811335373,"pitch":null,"output_rms_dbfs":-27.93116974662963,"output_peak":0.45278966426849365,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.737232998249635},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.814033407050615,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041851375067044644,"rms_dbfs":-27.565805365496097,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.213327125010572,"mean":-0.000893246332804362,"zero_energy":false},"output":{"rms":0.040127445416730356,"rms_dbfs":-27.93116974662963,"peak":0.45278966426849365,"peak_dbfs":-6.882069905052552,"crest_factor_db":21.049099841577082,"mean":3.825059285389898e-05,"zero_energy":false},"rms_difference_db":-0.365364381133535},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":27,"both_estimated_voiced_frames":20,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.8695652173913043,"absolute_difference_cents_median":5.005224935600257,"absolute_difference_cents_p95":10.438526885980613,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017313921740261034,"log_rms_mae_db":1.3832811080583498,"rms_envelope_nrmse":0.11414402847803841,"envelope_correlation":0.9872458137186357,"rms_gain_error_db":-0.36180373441967867,"reference_mean_power":0.0017487207983142375,"output_mean_power":0.0016089409374233934},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.577112847045013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002769005691966373,"output_power":0.00025434021659928656,"reference_active":true,"delta_db":-0.3690881167936262},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.377497445156709e-05,"output_power":3.173046478856221e-05,"reference_active":true,"delta_db":-0.27118595478096774},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495313749243886e-05,"output_power":1.733561938292328e-05,"reference_active":true,"delta_db":0.642070365217611}],"active_band_count":3,"active_band_mae_db":0.4274481455974016}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015508669364693045,"log_rms_mae_db":1.0872326482340937,"rms_envelope_nrmse":0.09027970222738634,"envelope_correlation":0.9919603768131141,"rms_gain_error_db":-0.35666534991096904,"reference_mean_power":0.0017487025727625826,"output_mean_power":0.0016108289047685653},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.09499315514524e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002529678773025127,"output_power":0.00023176442261641568,"reference_active":true,"delta_db":-0.38018606881255185},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6687984786311574e-05,"output_power":1.524785074369185e-05,"reference_active":true,"delta_db":-0.39195263009788106},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969938613307946e-07,"output_power":7.634795096526899e-07,"reference_active":true,"delta_db":0.3956843277447835}],"active_band_count":3,"active_band_mae_db":0.38927434221840546}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014635391485956695,"log_rms_mae_db":1.6792335122784547,"rms_envelope_nrmse":0.11487876023286672,"envelope_correlation":0.9903892150878907,"rms_gain_error_db":-0.22418118564957748,"reference_mean_power":0.0008142561216173605,"output_mean_power":0.0007732909253761176},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2354521678013805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001985819153870851,"output_power":0.00018696761364495005,"reference_active":true,"delta_db":-0.2617331032368891},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3653932189965355e-05,"output_power":2.339315020750608e-05,"reference_active":true,"delta_db":-0.04814638023788211},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.646479803369055e-06,"output_power":8.174742518463176e-06,"reference_active":true,"delta_db":0.8988239308520078}],"active_band_count":3,"active_band_mae_db":0.4029011381089263}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012544182666407365,"log_rms_mae_db":1.2161034207270707,"rms_envelope_nrmse":0.08547865141403146,"envelope_correlation":0.9946361640644699,"rms_gain_error_db":-0.21526261590519008,"reference_mean_power":0.0008126713020357493,"output_mean_power":0.0007733723878213628},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.0135181965386276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018492598004057708,"output_power":0.00017368204683366356,"reference_active":true,"delta_db":-0.2724300036933453},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0725232186915858e-05,"output_power":1.0490186465182727e-05,"reference_active":true,"delta_db":-0.09623494918267322},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.105637817013121e-07,"output_power":4.996139054844369e-07,"reference_active":true,"delta_db":0.8525388235650178}],"active_band_count":3,"active_band_mae_db":0.4070679254803455}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":1.7519848109604936,"rms_envelope_nrmse":0.23461656718617363,"envelope_correlation":0.9713891460821261,"rms_gain_error_db":-1.4671767548034937,"reference_mean_power":0.00024251974980034777,"output_mean_power":0.00017299336083708892},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1105776313023488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.895741122613483e-05,"output_power":2.563321621285546e-05,"reference_active":true,"delta_db":-1.8178698894071008},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.645635870343922e-06,"output_power":5.731812938550607e-06,"reference_active":true,"delta_db":0.06579144687133107},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.039589927108398e-06,"output_power":3.720830190140231e-06,"reference_active":true,"delta_db":-0.3569743029317759}],"active_band_count":3,"active_band_mae_db":0.7468785464034026}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":1.3315153778239168,"rms_envelope_nrmse":0.20592270078024516,"envelope_correlation":0.9823772156624887,"rms_gain_error_db":-1.4626939860387498,"reference_mean_power":0.00024219152470715302,"output_mean_power":0.00017293764582470024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0514876259433796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707790652686e-05,"output_power":2.4736510721033412e-05,"reference_active":true,"delta_db":-1.7577676629105357},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.355351339185193e-06,"output_power":3.1246012372289664e-06,"reference_active":true,"delta_db":-0.3094340151277661},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8063796091292895e-07,"output_power":2.66752947794717e-07,"reference_active":true,"delta_db":-0.22037188796197898}],"active_band_count":3,"active_band_mae_db":0.7625245220000937}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":1.857680309758454,"rms_envelope_nrmse":0.22615170192587639,"envelope_correlation":0.9728949156021377,"rms_gain_error_db":-1.1792189829870836,"reference_mean_power":0.00010238046641023414,"output_mean_power":7.803603690019774e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5067174911165995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.703714317319113e-05,"output_power":2.0954343119287075e-05,"reference_active":true,"delta_db":-1.1068674977983821},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.303852533122451e-06,"output_power":3.5413219037566277e-06,"reference_active":true,"delta_db":-0.8469197665401901},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8801329619345514e-06,"output_power":1.7518162858322925e-06,"reference_active":true,"delta_db":-0.30700003946904647}],"active_band_count":3,"active_band_mae_db":0.7535957679358729}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":1.4308878385813997,"rms_envelope_nrmse":0.171223823251778,"envelope_correlation":0.9892727396609441,"rms_gain_error_db":-1.1792562530832154,"reference_mean_power":0.00010289230869716874,"output_mean_power":7.842549828590659e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2184431092042676e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.696472252022106e-05,"output_power":2.118350064253706e-05,"reference_active":true,"delta_db":-1.0479822527005191},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.97826680457688e-06,"output_power":1.581739133594399e-06,"reference_active":true,"delta_db":-0.971500039938952},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5223765481589494e-07,"output_power":1.2526979691724926e-07,"reference_active":true,"delta_db":-0.8467571176766503}],"active_band_count":3,"active_band_mae_db":0.9554131367720404}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5159382820129395,"predicted_mos":3.5159382820129395,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"4d31f1bf8879c8b2dff94b442dd5af1f43c90f0e1ddf47a96009844896461172","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4580029380013},"started_at":"2026-10-13T07:22:29.524928+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.010616606999974465,"native_input_samples":264600,"native_input_sha256":"c352e22d760df0aa614a72cfaab032b67ce91de012c9e52c747a53838aae86c0","output_sha256":"e3d89c967a67a662a2b31825c4ef4e791aa3c22b373aaf18eb1ae61a6a77cb32","output_samples":264704,"length_delta_samples":104,"analysis_s":0.2505628800008708,"synthesis_s":37.06229410599917,"total_s":37.31285698600004,"rtf":6.2188094976666735,"output_peak":0.5203120708465576,"output_rms":0.038106979439707096,"finished_at":"2026-10-13T07:23:06.868927+00:00","evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"e3d89c967a67a662a2b31825c4ef4e791aa3c22b373aaf18eb1ae61a6a77cb32","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.3740750253994,"output_peak":0.5198273692604279},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.986843992443717,"spectral_convergence":0.5653084767309347,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.689053690056443,"spectral_convergence":0.4762743204561119,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.542948966898658,"spectral_convergence":0.5447333402859895,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.00020803476918896378,"delta_db":-1.5545097104039955,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.000524838942275714,"delta_db":-1.179680757827992,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.00042007548352460144,"delta_db":0.40660938501360055,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00014783892094454574,"delta_db":-1.1587395748232323,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":9.83036929476895e-05,"delta_db":-0.44541396652874626,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":5.506019697837729e-05,"delta_db":-1.6672191954051987,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0686954316671275,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8080965675152421,"pitch":null,"output_rms_dbfs":-28.37390193301134,"output_peak":0.5198273692604279,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.739615549799607},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6630962593119102,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041851375067044644,"rms_dbfs":-27.565805365496097,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.213327125010572,"mean":-0.000893246332804362,"zero_energy":false},"output":{"rms":0.03813334506517743,"rms_dbfs":-28.37390193301134,"peak":0.5198273692604279,"peak_dbfs":-5.682817166732414,"crest_factor_db":22.691084766278923,"mean":-0.0006072995772962424,"zero_energy":false},"rms_difference_db":-0.8080965675152427},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":20,"both_estimated_voiced_frames":11,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.4782608695652174,"absolute_difference_cents_median":7.008108770694572,"absolute_difference_cents_p95":21.39552681802624,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017313921740261034,"log_rms_mae_db":2.331602583681755,"rms_envelope_nrmse":0.18545857544121513,"envelope_correlation":0.97309669526518,"rms_gain_error_db":-0.8013069468820158,"reference_mean_power":0.0017487207983142375,"output_mean_power":0.001454084953974926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.577112847045013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002769005691966373,"output_power":0.00023060590900682475,"reference_active":true,"delta_db":-0.7945341709232099},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.377497445156709e-05,"output_power":2.423114932246623e-05,"reference_active":true,"delta_db":-1.4422101526715605},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495313749243886e-05,"output_power":1.2120823269662983e-05,"reference_active":true,"delta_db":-0.9120020766996612}],"active_band_count":3,"active_band_mae_db":1.0495821334314772}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015508669364693045,"log_rms_mae_db":1.9660550647626924,"rms_envelope_nrmse":0.13757214807613888,"envelope_correlation":0.9903200936237194,"rms_gain_error_db":-0.8060970776651271,"reference_mean_power":0.0017487025727625826,"output_mean_power":0.0014524668903227297},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.09499315514524e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002529678773025127,"output_power":0.00021134470159244294,"reference_active":true,"delta_db":-0.7807401198037907},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6687984786311574e-05,"output_power":1.1153669506963884e-05,"reference_active":true,"delta_db":-1.7498612333703456},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969938613307946e-07,"output_power":6.874834181078052e-07,"reference_active":true,"delta_db":-0.05966725554123719}],"active_band_count":3,"active_band_mae_db":0.863422869571791}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014635391485956695,"log_rms_mae_db":3.136321009808662,"rms_envelope_nrmse":0.1965506184599409,"envelope_correlation":0.971535693008601,"rms_gain_error_db":0.03421493747247334,"reference_mean_power":0.0008142561216173605,"output_mean_power":0.0008206963955723414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2354521678013805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001985819153870851,"output_power":0.00018910666431585905,"reference_active":true,"delta_db":-0.21232861186499877},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3653932189965355e-05,"output_power":2.026842092762136e-05,"reference_active":true,"delta_db":-0.6708343246947239},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.646479803369055e-06,"output_power":7.013425324547352e-06,"reference_active":true,"delta_db":0.23338487799049498}],"active_band_count":3,"active_band_mae_db":0.37218260485007254}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012544182666407365,"log_rms_mae_db":2.2667398101622083,"rms_envelope_nrmse":0.12958374057303607,"envelope_correlation":0.9870335967632327,"rms_gain_error_db":0.03563284340992098,"reference_mean_power":0.0008126713020357493,"output_mean_power":0.0008193665081596194},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.0135181965386276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018492598004057708,"output_power":0.00017494145026629706,"reference_active":true,"delta_db":-0.24105206473485719},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0725232186915858e-05,"output_power":8.953259433894388e-06,"reference_active":true,"delta_db":-0.7842553385429564},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.105637817013121e-07,"output_power":5.750450749166187e-07,"reference_active":true,"delta_db":1.4632125320527027}],"active_band_count":3,"active_band_mae_db":0.8295066451101721}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":3.0854531883064125,"rms_envelope_nrmse":0.2648541270833971,"envelope_correlation":0.9488227052539638,"rms_gain_error_db":-0.7958447966367098,"reference_mean_power":0.00024251974980034777,"output_mean_power":0.00020191223302941098},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1105776313023488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.895741122613483e-05,"output_power":3.456586443018066e-05,"reference_active":true,"delta_db":-0.5194266741718191},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.645635870343922e-06,"output_power":4.611852008783824e-06,"reference_active":true,"delta_db":-0.878375011564482},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.039589927108398e-06,"output_power":3.662863931266855e-06,"reference_active":true,"delta_db":-0.42516494883388606}],"active_band_count":3,"active_band_mae_db":0.607655544856729}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":2.468571856509711,"rms_envelope_nrmse":0.18998893459389057,"envelope_correlation":0.9763641897272001,"rms_gain_error_db":-0.8019168217162381,"reference_mean_power":0.00024219152470715302,"output_mean_power":0.00020135724436249198},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0514876259433796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707790652686e-05,"output_power":3.327782607255059e-05,"reference_active":true,"delta_db":-0.46960257698822777},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.355351339185193e-06,"output_power":2.7050082621813854e-06,"reference_active":true,"delta_db":-0.9356940595172278},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8063796091292895e-07,"output_power":2.2934099579380884e-07,"reference_active":true,"delta_db":-0.8766472228398536}],"active_band_count":3,"active_band_mae_db":0.760647953115103}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":3.26353420013826,"rms_envelope_nrmse":0.2798228229692669,"envelope_correlation":0.9605412680701794,"rms_gain_error_db":-1.6380650044115308,"reference_mean_power":0.00010238046641023414,"output_mean_power":7.021188022200133e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5067174911165995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.703714317319113e-05,"output_power":1.9559375719477303e-05,"reference_active":true,"delta_db":-1.4060581160890209},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.303852533122451e-06,"output_power":3.404394580677122e-06,"reference_active":true,"delta_db":-1.018174917724041},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8801329619345514e-06,"output_power":1.2059261999981828e-06,"reference_active":true,"delta_db":-1.928678326478762}],"active_band_count":3,"active_band_mae_db":1.4509704534306078}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":2.654862599318945,"rms_envelope_nrmse":0.2432088820977102,"envelope_correlation":0.9750864792215207,"rms_gain_error_db":-1.6383828832290255,"reference_mean_power":0.00010289230869716874,"output_mean_power":7.055773383141542e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2184431092042676e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.696472252022106e-05,"output_power":1.9240077901117782e-05,"reference_active":true,"delta_db":-1.465891294315314},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.97826680457688e-06,"output_power":1.5582003890300485e-06,"reference_active":true,"delta_db":-1.0366155512966098},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5223765481589494e-07,"output_power":1.0012130184965971e-07,"reference_active":true,"delta_db":-1.819955971452537}],"active_band_count":3,"active_band_mae_db":1.440820939021487}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"golf","status":"measured","mos":{"status":"measured","value":2.3777449131011963,"predicted_mos":2.3777449131011963,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"014d532a0b775123769106392493c75dc13204df85429851cf19a24da9bc2512","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4719462459997885},"started_at":"2026-10-13T05:02:48.359535+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029890669993619667,"native_input_samples":144000,"native_input_sha256":"2b50ebbeb56ed445c59855edf3cb9c1655381e940ae9800b9fc234871432b99b","output_sha256":"0ba2fc0a1c2fcaa4cb34eb990f57b1a5a315d9acfdef3e799e1145b7ad8741ed","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6418956019997495,"synthesis_s":0.1243313049999415,"total_s":0.766226906999691,"rtf":0.1277044844999485,"output_peak":0.35936856269836426,"output_rms":0.03764217979329721,"finished_at":"2026-10-13T05:02:49.139186+00:00","analysis_stages":{"external_f0_s":0.08596168799977022},"evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"0ba2fc0a1c2fcaa4cb34eb990f57b1a5a315d9acfdef3e799e1145b7ad8741ed","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.48481291640015,"output_peak":0.35631570834420506},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.842491343595547,"spectral_convergence":0.7375247836541875,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.769733940613419,"spectral_convergence":0.4822269405094406,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.99737333695464,"spectral_convergence":0.565496575479655,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002974217621808005,"output_power":0.0002578833685098902,"delta_db":-0.6194940803472073,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006866532113111554,"output_power":0.0004977713052432646,"delta_db":-1.397075980364958,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003794278594185899,"output_power":0.0003465146102431024,"delta_db":-0.39407665036231915,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019214457347547766,"output_power":0.0001483228490647222,"delta_db":-1.1242006463816165,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010898977272596691,"output_power":9.529555000968293e-05,"delta_db":-0.5831312597755668,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.095726078275736e-05,"output_power":7.169808892903989e-05,"delta_db":-0.5274822528393546,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7742434783451705,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9042464156228565,"pitch":null,"output_rms_dbfs":-28.48481291640015,"output_peak":0.35631570834420506,"reference_rms_dbfs":-27.580566500777294,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.203199540387868},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.757492312915111,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04178031162993818,"rms_dbfs":-27.580566500777287,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.228088260291766,"mean":-0.0008814150980597545,"zero_energy":false},"output":{"rms":0.03764951224119248,"rms_dbfs":-28.48481291640015,"peak":0.35631570834420506,"peak_dbfs":-8.963300623042596,"crest_factor_db":19.521512293357553,"mean":0.00013838777246689922,"zero_energy":false},"rms_difference_db":-0.9042464156228611},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":10,"both_estimated_voiced_frames":4,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.17391304347826086,"absolute_difference_cents_median":19.082239122232494,"absolute_difference_cents_p95":21.061285073779043,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017313921740261034,"log_rms_mae_db":2.510857319838651,"rms_envelope_nrmse":0.23306552416582124,"envelope_correlation":0.9474447524157882,"rms_gain_error_db":-0.9033201166106819,"reference_mean_power":0.0017460623238552376,"output_mean_power":0.001418168157005591},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.543318310730322e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027552481044059514,"output_power":0.00019218381400107543,"reference_active":true,"delta_db":-1.564439043075564},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.374155085888462e-05,"output_power":3.6559389853620815e-05,"reference_active":true,"delta_db":0.3483389896439142},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4975828026604466e-05,"output_power":1.8520412521948434e-05,"reference_active":true,"delta_db":0.9225981164238063}],"active_band_count":3,"active_band_mae_db":0.9451253830477615}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015508669364693045,"log_rms_mae_db":1.6687852526648668,"rms_envelope_nrmse":0.16216577047278613,"envelope_correlation":0.9788279813874774,"rms_gain_error_db":-0.9148974437995643,"reference_mean_power":0.0017457267595059178,"output_mean_power":0.001414120847013432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.060529197096095e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002516708665469197,"output_power":0.00017744941434497737,"reference_active":true,"delta_db":-1.5175837413276554},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6683379290059148e-05,"output_power":1.751951213603783e-05,"reference_active":true,"delta_db":0.21237984826072032},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.927980262282698e-07,"output_power":1.0030697118134431e-06,"reference_active":true,"delta_db":1.607244751604435}],"active_band_count":3,"active_band_mae_db":1.1124027803976035}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014635391485956695,"log_rms_mae_db":2.276854654178853,"rms_envelope_nrmse":0.2957042845294311,"envelope_correlation":0.9359209897285822,"rms_gain_error_db":-0.9710238574870864,"reference_mean_power":0.0008116898727607626,"output_mean_power":0.0006490643284899498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.2106827910543474e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001975400912222581,"output_power":0.00011529292627074195,"reference_active":true,"delta_db":-2.3385258775478723},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.360835797788981e-05,"output_power":2.871218809911045e-05,"reference_active":true,"delta_db":0.8500050865262266},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.6513052986697664e-06,"output_power":1.0557049138624883e-05,"reference_active":true,"delta_db":2.006356603641006}],"active_band_count":3,"active_band_mae_db":1.7316291892383682}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012544182666407365,"log_rms_mae_db":1.4540395811972713,"rms_envelope_nrmse":0.20891889871746488,"envelope_correlation":0.9712269257151396,"rms_gain_error_db":-0.9874542892813237,"reference_mean_power":0.0008104000322263929,"output_mean_power":0.0006455858764236838},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.9889208096730425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001839586922112079,"output_power":0.00011201997871005788,"reference_active":true,"delta_db":-2.1542482765811095},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.069940312827464e-05,"output_power":1.3406821609190454e-05,"reference_active":true,"delta_db":0.9796627968856895},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.098353824400392e-07,"output_power":6.472165558166245e-07,"reference_active":true,"delta_db":1.9844016825213031}],"active_band_count":3,"active_band_mae_db":1.7061042519960339}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":2.0441264066246068,"rms_envelope_nrmse":0.2875565990022847,"envelope_correlation":0.9376955175138414,"rms_gain_error_db":-1.0113702696765694,"reference_mean_power":0.0002421606757808928,"output_mean_power":0.0001918521157352163},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.103626242844057e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8581775965734575e-05,"output_power":3.276591995495469e-05,"reference_active":true,"delta_db":-0.7095984890053252},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.641751375325106e-06,"output_power":7.03669394043413e-06,"reference_active":true,"delta_db":0.9595471844258017},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.025017300061596e-06,"output_power":3.112907266187938e-06,"reference_active":true,"delta_db":-1.1160156816344315}],"active_band_count":3,"active_band_mae_db":0.9283871183551861}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":1.3798075383168613,"rms_envelope_nrmse":0.1978845737083437,"envelope_correlation":0.9749492023060465,"rms_gain_error_db":-1.0430178337216534,"reference_mean_power":0.00024143910391486912,"output_mean_power":0.00018989163209074346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0440929074854008e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.672001863510633e-05,"output_power":3.174160470045106e-05,"reference_active":true,"delta_db":-0.6327401414915147},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3420423866414493e-06,"output_power":4.0134450203263176e-06,"reference_active":true,"delta_db":0.7950536432202013},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779662134289111e-07,"output_power":1.9164396037133517e-07,"reference_active":true,"delta_db":-1.6149687389369962}],"active_band_count":3,"active_band_mae_db":1.0142541745495708}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":2.45884542713525,"rms_envelope_nrmse":0.21046189460123157,"envelope_correlation":0.9716167309755221,"rms_gain_error_db":-0.42100770538084775,"reference_mean_power":0.0001025493261739622,"output_mean_power":9.30747847537124e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.497426003438051e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6982197020894056e-05,"output_power":2.3269771076753026e-05,"reference_active":true,"delta_db":-0.6428519832394826},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.268081189168408e-06,"output_power":4.600948844990661e-06,"reference_active":true,"delta_db":0.3261473249241359},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8791350036844864e-06,"output_power":2.489468982266733e-06,"reference_active":true,"delta_db":1.2214873705087954}],"active_band_count":3,"active_band_mae_db":0.730162226224138}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":2.002086088400444,"rms_envelope_nrmse":0.12236036357198603,"envelope_correlation":0.9913425540929133,"rms_gain_error_db":-0.4209933945151234,"reference_mean_power":0.00010306355149288537,"output_mean_power":9.354180898774373e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.2100771065555124e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.6899249836426427e-05,"output_power":2.3918168409354144e-05,"reference_active":true,"delta_db":-0.510122491538957},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9632239467400065e-06,"output_power":2.2447436014533485e-06,"reference_active":true,"delta_db":0.5819689950354647},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5153570415602866e-07,"output_power":1.8061989837282974e-07,"reference_active":true,"delta_db":0.7625062218942046}],"active_band_count":3,"active_band_mae_db":0.6181992361562086}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.8690409660339355,"predicted_mos":2.8690409660339355,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"cac74217a910e3d2edf9928f59996ddb9c49cba9b5d3f4208f4a86d9d7fca2d3","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.36627769000188},"started_at":"2026-10-13T07:41:56.756377+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00296085599984508,"native_input_samples":144000,"native_input_sha256":"2b50ebbeb56ed445c59855edf3cb9c1655381e940ae9800b9fc234871432b99b","output_sha256":"7d7c921ef17db888574da11367945ba13e11b5143d82d62b888f70ce80a39a03","output_samples":144120,"length_delta_samples":120,"analysis_s":2.1969573339993076,"synthesis_s":3.663960235000559,"total_s":5.8609175689998665,"rtf":0.976819594833311,"output_peak":0.47479531168937683,"output_rms":0.040566904178083424,"finished_at":"2026-10-13T07:42:02.630581+00:00","analysis_stages":{"external_f0_s":1.5138031200003752},"evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"7d7c921ef17db888574da11367945ba13e11b5143d82d62b888f70ce80a39a03","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.834544301631375,"output_peak":0.4824875000587157},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.807387664907058,"spectral_convergence":0.7941824921832085,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.566721070064489,"spectral_convergence":0.45569945741873163,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.597361898365484,"spectral_convergence":0.4787808091887788,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.0003828753494874397,"delta_db":1.0947050982572033,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0006276861566305977,"delta_db":-0.40251576792534705,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.0003236673598856505,"delta_db":-0.7256749210516656,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00015528677972395882,"delta_db":-0.9452825769522338,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":9.273956875440332e-05,"delta_db":-0.698461577481303,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":6.527462981121129e-05,"delta_db":-0.9281526278415133,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7991320949182109,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2658617305567189,"pitch":null,"output_rms_dbfs":-27.83166709605281,"output_peak":0.4824875000587157,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.990490211112344},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7821766371123555,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041851375067044644,"rms_dbfs":-27.565805365496097,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.213327125010572,"mean":-0.000893246332804362,"zero_energy":false},"output":{"rms":0.04058977512001343,"rms_dbfs":-27.83166709605281,"peak":0.4824875000587157,"peak_dbfs":-6.3302786713092285,"crest_factor_db":21.501388424743585,"mean":0.0001409230814252698,"zero_energy":false},"rms_difference_db":-0.2658617305567148},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":28,"both_estimated_voiced_frames":14,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.6086956521739131,"absolute_difference_cents_median":3.1723498788952025,"absolute_difference_cents_p95":8.234117213131205,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017313921740261034,"log_rms_mae_db":3.4292184137246733,"rms_envelope_nrmse":0.24209200183521373,"envelope_correlation":0.9397429178328391,"rms_gain_error_db":-0.256164218135572,"reference_mean_power":0.0017487207983142375,"output_mean_power":0.0016485573254734383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.577112847045013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002769005691966373,"output_power":0.0003236606288958326,"reference_active":true,"delta_db":0.6776602517699016},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.377497445156709e-05,"output_power":3.5898092351592276e-05,"reference_active":true,"delta_db":0.2647634131460094},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495313749243886e-05,"output_power":1.0304245163472194e-05,"reference_active":true,"delta_db":-1.6171614354150479}],"active_band_count":3,"active_band_mae_db":0.853195033443653}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015508669364693045,"log_rms_mae_db":2.648076752757497,"rms_envelope_nrmse":0.14652191372869283,"envelope_correlation":0.9780983453311578,"rms_gain_error_db":-0.24278371811147192,"reference_mean_power":0.0017487025727625826,"output_mean_power":0.001653627085494368},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.09499315514524e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002529678773025127,"output_power":0.0002957183214154335,"reference_active":true,"delta_db":0.6781285582099804},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6687984786311574e-05,"output_power":1.6998216959074996e-05,"reference_active":true,"delta_db":0.0799947296087402},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969938613307946e-07,"output_power":4.939879294203869e-07,"reference_active":true,"delta_db":-1.4951261604994084}],"active_band_count":3,"active_band_mae_db":0.7510831494393764}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014635391485956695,"log_rms_mae_db":2.763016488344019,"rms_envelope_nrmse":0.2569757109280452,"envelope_correlation":0.9510999789508979,"rms_gain_error_db":-0.6289399039834664,"reference_mean_power":0.0008142561216173605,"output_mean_power":0.0007044773620767201},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2354521678013805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001985819153870851,"output_power":0.00018625599524619882,"reference_active":true,"delta_db":-0.2782943448598412},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3653932189965355e-05,"output_power":2.0993499693235304e-05,"reference_active":true,"delta_db":-0.5181850436348909},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.646479803369055e-06,"output_power":5.199163192857682e-06,"reference_active":true,"delta_db":-1.0665823999046402}],"active_band_count":3,"active_band_mae_db":0.6210205961331241}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012544182666407365,"log_rms_mae_db":1.7363928610492532,"rms_envelope_nrmse":0.15674895042224266,"envelope_correlation":0.9831698567217614,"rms_gain_error_db":-0.612366413562915,"reference_mean_power":0.0008126713020357493,"output_mean_power":0.0007057945205311551},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.0135181965386276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018492598004057708,"output_power":0.00017342323508569066,"reference_active":true,"delta_db":-0.2789064556911371},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0725232186915858e-05,"output_power":9.522829240095664e-06,"reference_active":true,"delta_db":-0.5164070606302182},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.105637817013121e-07,"output_power":3.4236467305604527e-07,"reference_active":true,"delta_db":-0.7889168941440008}],"active_band_count":3,"active_band_mae_db":0.5280768034884521}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":2.4605591767032338,"rms_envelope_nrmse":0.3285472161971319,"envelope_correlation":0.9136492907399576,"rms_gain_error_db":-0.7914016024098351,"reference_mean_power":0.00024251974980034777,"output_mean_power":0.00020211891176564296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1105776313023488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.895741122613483e-05,"output_power":3.7071851936532704e-05,"reference_active":true,"delta_db":-0.21545807801450562},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.645635870343922e-06,"output_power":4.326821251640138e-06,"reference_active":true,"delta_db":-1.1554390966198045},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.039589927108398e-06,"output_power":3.0236272926097757e-06,"reference_active":true,"delta_db":-1.258090237744661}],"active_band_count":3,"active_band_mae_db":0.8763291374596571}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":1.475355507295284,"rms_envelope_nrmse":0.18406480380267126,"envelope_correlation":0.9755563618734081,"rms_gain_error_db":-0.7696989572917246,"reference_mean_power":0.00024219152470715302,"output_mean_power":0.00020285655487655804},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0514876259433796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707790652686e-05,"output_power":3.4894410444659686e-05,"reference_active":true,"delta_db":-0.2635934011846131},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.355351339185193e-06,"output_power":2.3959818756320336e-06,"reference_active":true,"delta_db":-1.4625447339028355},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8063796091292895e-07,"output_power":1.7492612787711228e-07,"reference_active":true,"delta_db":-2.0529173334915325}],"active_band_count":3,"active_band_mae_db":1.2596851561929936}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":2.070334590738926,"rms_envelope_nrmse":0.2602788837931479,"envelope_correlation":0.9566452940857662,"rms_gain_error_db":-0.9224779836329983,"reference_mean_power":0.00010238046641023414,"output_mean_power":8.278836492794528e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5067174911165995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.703714317319113e-05,"output_power":2.490381002891295e-05,"reference_active":true,"delta_db":-0.3569500601283816},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.303852533122451e-06,"output_power":3.684465570257679e-06,"reference_active":true,"delta_db":-0.6748287969692435},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8801329619345514e-06,"output_power":1.291983992146335e-06,"reference_active":true,"delta_db":-1.629314306827263}],"active_band_count":3,"active_band_mae_db":0.8870310546416293}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":1.2877332157519197,"rms_envelope_nrmse":0.17719761525957342,"envelope_correlation":0.9832405719115209,"rms_gain_error_db":-0.9225069115760618,"reference_mean_power":0.00010289230869716874,"output_mean_power":8.320170399754362e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2184431092042676e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.696472252022106e-05,"output_power":2.4622197935149972e-05,"reference_active":true,"delta_db":-0.39469137317199576},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.97826680457688e-06,"output_power":1.6078399402225826e-06,"reference_active":true,"delta_db":-0.9004205083671828},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5223765481589494e-07,"output_power":9.260848342959013e-08,"reference_active":true,"delta_db":-2.158713129410187}],"active_band_count":3,"active_band_mae_db":1.1512750036497885}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.4504239559173584,"predicted_mos":2.4504239559173584,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"3b9c20f2b8a29e4eadcbac47dc68d94811672e15d7faf7eac88a1bab2407da81","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.680487510999228},"started_at":"2026-10-13T05:11:06.789618+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0029552000014518853,"native_input_samples":144000,"native_input_sha256":"2b50ebbeb56ed445c59855edf3cb9c1655381e940ae9800b9fc234871432b99b","output_sha256":"606469aa7a821a93ef2c383b3b45de5da4caf7fa998f787a7d390516be611f50","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7274303290014359,"synthesis_s":0.08419631799915805,"total_s":0.811626647000594,"rtf":0.13527110783343232,"output_peak":0.4685687720775604,"output_rms":0.04071874405068118,"finished_at":"2026-10-13T05:11:07.616397+00:00","analysis_stages":{"external_f0_s":0.175887037999928},"evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"606469aa7a821a93ef2c383b3b45de5da4caf7fa998f787a7d390516be611f50","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.801433186102074,"output_peak":0.4688376358401116},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.566607673739891,"spectral_convergence":0.462595936982737,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.89338363368505,"spectral_convergence":0.3972266710464584,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.59986068116294,"spectral_convergence":0.5321219710583717,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.00029852801128775697,"delta_db":0.013981914052235384,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.00063340055471874,"delta_db":-0.3631568797426489,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.00039234304710748184,"delta_db":0.10999563229518035,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.000173422771005915,"delta_db":-0.4655662027320904,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":9.23295900962567e-05,"delta_db":-0.7177032232851045,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":7.08785040837371e-05,"delta_db":-0.570451372584073,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.37347587078188865,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23075330324465823,"pitch":null,"output_rms_dbfs":-27.796558668740758,"output_peak":0.4688376358401116,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.35328399619596},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.826203003652962,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041851375067044644,"rms_dbfs":-27.565805365496097,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.213327125010572,"mean":-0.000893246332804362,"zero_energy":false},"output":{"rms":0.04075417129938833,"rms_dbfs":-27.796558668740758,"peak":0.4688376358401116,"peak_dbfs":-6.579550653894131,"crest_factor_db":21.217008014846623,"mean":-0.000163444408677621,"zero_energy":false},"rms_difference_db":-0.23075330324466137},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":14,"both_estimated_voiced_frames":13,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.5652173913043478,"absolute_difference_cents_median":1.3791495589894778,"absolute_difference_cents_p95":3.6908269464863155,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017313921740261034,"log_rms_mae_db":1.275802141758905,"rms_envelope_nrmse":0.12677522640704553,"envelope_correlation":0.9829981440808834,"rms_gain_error_db":-0.2287492020504027,"reference_mean_power":0.0017487207983142375,"output_mean_power":0.001658996825966271},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.577112847045013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002769005691966373,"output_power":0.00027184258263270803,"reference_active":true,"delta_db":-0.08006360935041404},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.377497445156709e-05,"output_power":3.1996181059914007e-05,"reference_active":true,"delta_db":-0.23496883472440522},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495313749243886e-05,"output_power":1.409770430689893e-05,"reference_active":true,"delta_db":-0.25583929426725033}],"active_band_count":3,"active_band_mae_db":0.19029057944735653}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015508669364693045,"log_rms_mae_db":0.5999818529220619,"rms_envelope_nrmse":0.059524968612700506,"envelope_correlation":0.9966349137757504,"rms_gain_error_db":-0.22760257852758886,"reference_mean_power":0.0017487025727625826,"output_mean_power":0.0016594175968469062},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.09499315514524e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002529678773025127,"output_power":0.00024967509003315306,"reference_active":true,"delta_db":-0.05690161372145734},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6687984786311574e-05,"output_power":1.4946114233270555e-05,"reference_active":true,"delta_db":-0.4787559790677342},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969938613307946e-07,"output_power":7.524322938388115e-07,"reference_active":true,"delta_db":0.3323847374588907}],"active_band_count":3,"active_band_mae_db":0.2893474434160274}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014635391485956695,"log_rms_mae_db":1.470616531127156,"rms_envelope_nrmse":0.1768791980874558,"envelope_correlation":0.9764142010953024,"rms_gain_error_db":-0.1814866892351633,"reference_mean_power":0.0008142561216173605,"output_mean_power":0.0007809304611946993},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2354521678013805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001985819153870851,"output_power":0.0001838614213593968,"reference_active":true,"delta_db":-0.33449082139932207},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3653932189965355e-05,"output_power":2.262623969344883e-05,"reference_active":true,"delta_db":-0.19290963902878483},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.646479803369055e-06,"output_power":8.045364637555066e-06,"reference_active":true,"delta_db":0.829540428741844}],"active_band_count":3,"active_band_mae_db":0.45231362972331696}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012544182666407365,"log_rms_mae_db":0.7757197169303885,"rms_envelope_nrmse":0.08509454249093817,"envelope_correlation":0.9944554721854105,"rms_gain_error_db":-0.17835630270110536,"reference_mean_power":0.0008126713020357493,"output_mean_power":0.0007799725048518037},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.0135181965386276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018492598004057708,"output_power":0.00017229688583386406,"reference_active":true,"delta_db":-0.3072050108471749},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0725232186915858e-05,"output_power":9.923879498973104e-06,"reference_active":true,"delta_db":-0.33725220655115734},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.105637817013121e-07,"output_power":5.193869589475028e-07,"reference_active":true,"delta_db":1.0211040592829215}],"active_band_count":3,"active_band_mae_db":0.5551870922270846}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":1.3088234060631776,"rms_envelope_nrmse":0.18410548051504586,"envelope_correlation":0.9740390293967666,"rms_gain_error_db":-0.44848913243808697,"reference_mean_power":0.00024251974980034777,"output_mean_power":0.00021872488571018822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1105776313023488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.895741122613483e-05,"output_power":3.7726101691135836e-05,"reference_active":true,"delta_db":-0.13948158640492858},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.645635870343922e-06,"output_power":5.731068856818029e-06,"reference_active":true,"delta_db":0.06522762601586665},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.039589927108398e-06,"output_power":2.880282523095444e-06,"reference_active":true,"delta_db":-1.4690219137868135}],"active_band_count":3,"active_band_mae_db":0.5579103754025362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":0.7442631223721848,"rms_envelope_nrmse":0.09493438822418557,"envelope_correlation":0.9943373661408507,"rms_gain_error_db":-0.45000563519063447,"reference_mean_power":0.00024219152470715302,"output_mean_power":0.00021835260512501262},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0514876259433796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707790652686e-05,"output_power":3.5771236379191496e-05,"reference_active":true,"delta_db":-0.1558125411295292},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.355351339185193e-06,"output_power":3.228020760774177e-06,"reference_active":true,"delta_db":-0.16801682714563917},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8063796091292895e-07,"output_power":1.9353630130641224e-07,"reference_active":true,"delta_db":-1.6138397921393324}],"active_band_count":3,"active_band_mae_db":0.6458897201381669}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":1.0911426830285753,"rms_envelope_nrmse":0.14157846329807414,"envelope_correlation":0.9884621156413483,"rms_gain_error_db":-0.5666060633992627,"reference_mean_power":0.00010238046641023414,"output_mean_power":8.98579481184875e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5067174911165995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.703714317319113e-05,"output_power":2.492104151964206e-05,"reference_active":true,"delta_db":-0.35394612076704196},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.303852533122451e-06,"output_power":3.7209561427760393e-06,"reference_active":true,"delta_db":-0.6320283106055766},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8801329619345514e-06,"output_power":1.3297539821616394e-06,"reference_active":true,"delta_db":-1.504172638496106}],"active_band_count":3,"active_band_mae_db":0.8300490232895749}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":0.6456217413304846,"rms_envelope_nrmse":0.08994489646648426,"envelope_correlation":0.9968963551251555,"rms_gain_error_db":-0.5666597882482842,"reference_mean_power":0.00010289230869716874,"output_mean_power":9.030606801081048e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2184431092042676e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.696472252022106e-05,"output_power":2.459966285778306e-05,"reference_active":true,"delta_db":-0.39866800480511416},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.97826680457688e-06,"output_power":1.7354376977825075e-06,"reference_active":true,"delta_db":-0.5687583645694696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5223765481589494e-07,"output_power":9.83328798082378e-08,"reference_active":true,"delta_db":-1.8982332689367603}],"active_band_count":3,"active_band_mae_db":0.9552198794371147}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0043","method":"magphase","status":"measured","mos":{"status":"measured","value":2.711303472518921,"predicted_mos":2.711303472518921,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":290870,"input_sample_rate":48000,"input_seconds":6.0597916666666665,"input_float32_sha256":"9b4b0b0c6ea9fde28f89bebc71dfff7fba9740e1646efa9ff31af44e7dff8c8e","internal_samples":96957,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5584056970001257},"started_at":"2026-10-13T05:17:01.679380+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008480192000206443,"native_input_samples":288000,"native_input_sha256":"de1c7c05c5e85f053a144b734ff7bbbdbbd2004c6ef2030d1661d01d245c3bf0","output_sha256":"69a9c189eb9e1cf61b28639305c46501879ab101bf14f63cae816eb955a1e60d","output_samples":290870,"length_delta_samples":2870,"analysis_s":3.027855855998496,"synthesis_s":1.09436093900149,"total_s":4.122216794999986,"rtf":0.6870361324999976,"output_peak":0.578180720720031,"output_rms":0.040084313730602554,"finished_at":"2026-10-13T05:17:05.822265+00:00","evaluation_input":{"reference":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"69a9c189eb9e1cf61b28639305c46501879ab101bf14f63cae816eb955a1e60d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96957,"length_delta_samples":957,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.938058787567286,"output_peak":0.5614041785870082},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9901296451004054,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.27713599661491,"spectral_convergence":0.9861087686657968,"spectral_convergence_status":"ok","amplitude_floor":1.4385235391470026e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.515696389784178,"spectral_convergence":0.8419059823398582,"spectral_convergence_status":"ok","amplitude_floor":9.243439915833013e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.89712955073491,"spectral_convergence":0.8744029375173048,"spectral_convergence_status":"ok","amplitude_floor":6.315989368278398e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00029756845932849967,"output_power":0.00025076549801750236,"delta_db":-0.7431911326171546,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0006886429842704082,"output_power":0.0006057843278348496,"delta_db":-0.5567609378835906,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003825307900717409,"output_power":0.0003678451678888701,"delta_db":-0.17001342041490097,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019304685978074288,"output_power":0.00017240775289089477,"delta_db":-0.4910595008483618,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010892090215881134,"output_power":0.00011313611046153343,"delta_db":0.1649001411537337,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.08275993922433e-05,"output_power":7.846011058001794e-05,"delta_db":-0.12910764413849463,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3758387961760394,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4245986731350277,"pitch":null,"output_rms_dbfs":-27.990404038631127,"output_peak":0.5614041785870082,"reference_rms_dbfs":-27.565805365496097,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.563320645711332},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.3506135428727612,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.041851375067044644,"rms_dbfs":-27.565805365496097,"peak":0.53997802734375,"peak_dbfs":-5.3524782404855245,"crest_factor_db":22.213327125010572,"mean":-0.000893246332804362,"zero_energy":false},"output":{"rms":0.03985472327935133,"rms_dbfs":-27.990404038631127,"peak":0.5614041785870082,"peak_dbfs":-5.014487183852535,"crest_factor_db":22.97591685477859,"mean":-4.314163121891074e-05,"zero_energy":false},"rms_difference_db":-0.42459867313502997},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":2.8062215517275037e-06,"threshold_dbfs":-55.51878044321783,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":23,"output_estimated_voiced_frames":35,"both_estimated_voiced_frames":5,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.21739130434782608,"absolute_difference_cents_median":10.351234767472512,"absolute_difference_cents_p95":15.551336633806942,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017313921740261034,"log_rms_mae_db":5.247030440796383,"rms_envelope_nrmse":0.5958901336126727,"envelope_correlation":0.6133674759744178,"rms_gain_error_db":-0.4140454785165024,"reference_mean_power":0.0017487207983142375,"output_mean_power":0.001589702757182712},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.577112847045013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002769005691966373,"output_power":0.00025850192667177036,"reference_active":true,"delta_db":-0.29860064147547766},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.377497445156709e-05,"output_power":3.759209601423322e-05,"reference_active":true,"delta_db":0.4650151207041632},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.495313749243886e-05,"output_power":1.3632548734840628e-05,"reference_active":true,"delta_db":-0.40155267703841124}],"active_band_count":3,"active_band_mae_db":0.38838947973935073}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015508669364693045,"log_rms_mae_db":3.924108284202791,"rms_envelope_nrmse":0.509271538819258,"envelope_correlation":0.6926451045611532,"rms_gain_error_db":-0.41012572179667445,"reference_mean_power":0.0017487025727625826,"output_mean_power":0.0015911216195704737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.09499315514524e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002529678773025127,"output_power":0.00023807542959580408,"reference_active":true,"delta_db":-0.26350799861049934},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6687984786311574e-05,"output_power":1.8405803180106476e-05,"reference_active":true,"delta_db":0.4255087849199269},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.969938613307946e-07,"output_power":6.970986932771892e-07,"reference_active":true,"delta_db":0.0006531551406926165}],"active_band_count":3,"active_band_mae_db":0.2298899795570396}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014635391485956695,"log_rms_mae_db":8.621816645769925,"rms_envelope_nrmse":0.7714693380301823,"envelope_correlation":0.5385540087190671,"rms_gain_error_db":-0.4708956084841525,"reference_mean_power":0.0008142561216173605,"output_mean_power":0.000730586211536151},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.2354521678013805e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001985819153870851,"output_power":0.00017916282142248395,"reference_active":true,"delta_db":-0.44691802276723325},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3653932189965355e-05,"output_power":2.349015330100462e-05,"reference_active":true,"delta_db":-0.030174963892810013},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.646479803369055e-06,"output_power":6.04083550292605e-06,"reference_active":true,"delta_db":-0.4149467971701868}],"active_band_count":3,"active_band_mae_db":0.2973465946100767}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012544182666407365,"log_rms_mae_db":7.214068781714741,"rms_envelope_nrmse":0.6921615688510523,"envelope_correlation":0.6063906536627175,"rms_gain_error_db":-0.4596448872942228,"reference_mean_power":0.0008126713020357493,"output_mean_power":0.0007310556450711302},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.0135181965386276e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018492598004057708,"output_power":0.0001682825094454313,"reference_active":true,"delta_db":-0.4095894934764427},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0725232186915858e-05,"output_power":1.0926690936949566e-05,"reference_active":true,"delta_db":0.08081956314943745},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.105637817013121e-07,"output_power":3.69016866769834e-07,"reference_active":true,"delta_db":-0.46334417950931217}],"active_band_count":3,"active_band_mae_db":0.31791774537839745}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.111876173726921e-05,"log_rms_mae_db":6.323409470433936,"rms_envelope_nrmse":0.6916462988685199,"envelope_correlation":0.6181479377367202,"rms_gain_error_db":-0.26605975563136386,"reference_mean_power":0.00024251974980034777,"output_mean_power":0.0002281083252385987},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1105776313023488e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.895741122613483e-05,"output_power":3.659735069931765e-05,"reference_active":true,"delta_db":-0.2714044196071355},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.645635870343922e-06,"output_power":6.066896991008753e-06,"reference_active":true,"delta_db":0.3125375739558389},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.039589927108398e-06,"output_power":3.4639123109653096e-06,"reference_active":true,"delta_db":-0.6677039128966165}],"active_band_count":3,"active_band_mae_db":0.41721530215319697}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.178958171420994e-05,"log_rms_mae_db":4.929223137816694,"rms_envelope_nrmse":0.5678874904583131,"envelope_correlation":0.7241890424299797,"rms_gain_error_db":-0.2865869873043885,"reference_mean_power":0.00024219152470715302,"output_mean_power":0.00022672543434663626},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0514876259433796e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707790652686e-05,"output_power":3.517965963411598e-05,"reference_active":true,"delta_db":-0.22823571949821195},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.355351339185193e-06,"output_power":2.96803324970479e-06,"reference_active":true,"delta_db":-0.5326924004219533},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8063796091292895e-07,"output_power":2.077307495545875e-07,"reference_active":true,"delta_db":-1.3064562798342108}],"active_band_count":3,"active_band_mae_db":0.6891281332514586}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.3400740982798506e-05,"log_rms_mae_db":6.564203968175948,"rms_envelope_nrmse":0.9120709970203127,"envelope_correlation":0.4342437920966788,"rms_gain_error_db":-0.03952512515400569,"reference_mean_power":0.00010238046641023414,"output_mean_power":0.00010145292931229154},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.5067174911165995e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.703714317319113e-05,"output_power":2.8222779773021508e-05,"reference_active":true,"delta_db":0.18638986085924658},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.303852533122451e-06,"output_power":4.883080872607355e-06,"reference_active":true,"delta_db":0.5483653476191451},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8801329619345514e-06,"output_power":2.428586901876882e-06,"reference_active":true,"delta_db":1.1116508499347852}],"active_band_count":3,"active_band_mae_db":0.6154686861377257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.152733550445532e-05,"log_rms_mae_db":5.785841631406549,"rms_envelope_nrmse":0.8356738527725177,"envelope_correlation":0.4959837101898485,"rms_gain_error_db":-0.039462478558741694,"reference_mean_power":0.00010289230869716874,"output_mean_power":0.00010196160523426568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.2184431092042676e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.696472252022106e-05,"output_power":2.8532173730169516e-05,"reference_active":true,"delta_db":0.2453890419070304},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.97826680457688e-06,"output_power":2.38855447278236e-06,"reference_active":true,"delta_db":0.8185028657352402},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5223765481589494e-07,"output_power":2.0096685718393015e-07,"reference_active":true,"delta_db":1.2060235572907132}],"active_band_count":3,"active_band_mae_db":0.7566384883109946}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2266907691955566,"predicted_mos":2.2266907691955566,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d994160fd611883538c8428ff97969119d9c3a6b39c1c292ea3a7159d7196b5b","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.423065906999909},"started_at":"2026-10-13T04:39:47.705363+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4539999938278925e-05,"native_input_samples":96000,"native_input_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","output_sha256":"2b0bf42589adf07c5838f2ebc76eab37c500b4fc40006f774c12d28ff67aeff9","output_samples":96000,"length_delta_samples":0,"analysis_s":1.099310272999901,"synthesis_s":0.49535580100018706,"total_s":1.594666074000088,"rtf":0.26577767900001464,"output_peak":0.43652958742844233,"output_rms":0.04935492115893842,"finished_at":"2026-10-13T04:39:49.307180+00:00","evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"2b0bf42589adf07c5838f2ebc76eab37c500b4fc40006f774c12d28ff67aeff9","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.289483396651079,"spectral_convergence":0.41045804597071195,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.783694828167174,"spectral_convergence":0.35510745253728654,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.907307397656477,"spectral_convergence":0.3905459773221125,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0006222613487550557,"delta_db":0.05967375036483224,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0009392650353297065,"delta_db":-0.10199333176390069,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0005450068294484088,"delta_db":0.020869295456246988,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00013503319428087986,"delta_db":-0.06401827073652026,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":8.465376045720157e-05,"delta_db":-0.0787872348677916,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":0.00010968807433377428,"delta_db":-0.10545662952299023,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.07179975211871367,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.031029316485022638,"pitch":null,"output_rms_dbfs":-26.133390750214463,"output_peak":0.43652958742844233,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.993495207491577},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.289483396651079,"spectral_convergence":0.41045804597071195,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.783694828167174,"spectral_convergence":0.35510745253728654,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.907307397656477,"spectral_convergence":0.3905459773221125,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0006222613487550557,"delta_db":0.05967375036483224,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0009392650353297065,"delta_db":-0.10199333176390069,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0005450068294484088,"delta_db":0.020869295456246988,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00013503319428087986,"delta_db":-0.06401827073652026,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":8.465376045720157e-05,"delta_db":-0.0787872348677916,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":0.00010968807433377428,"delta_db":-0.10545662952299023,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.07179975211871367,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.031029316485022638,"pitch":null,"output_rms_dbfs":-26.133390750214463,"output_peak":0.43652958742844233,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.993495207491577},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7944348340454123,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04953155110178405,"rms_dbfs":-26.102361433729442,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43919705387114,"mean":-0.0009115762710571289,"zero_energy":false},"output":{"rms":0.04935492115893842,"rms_dbfs":-26.133390750214463,"peak":0.43652958742844233,"peak_dbfs":-7.199726300500209,"crest_factor_db":18.933664449714254,"mean":-4.9592629593456277e-05,"zero_energy":false},"rms_difference_db":-0.031029316485021496},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":81,"both_estimated_voiced_frames":63,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9402985074626866,"absolute_difference_cents_median":1.4143379708558788,"absolute_difference_cents_p95":6.170151933030751,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013602681282863616,"log_rms_mae_db":1.202530254245828,"rms_envelope_nrmse":0.11419248044553688,"envelope_correlation":0.9808496815375812,"rms_gain_error_db":-0.030677355480920567,"reference_mean_power":0.0024543453947075094,"output_mean_power":0.0024370696680970478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.153929939453048e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000440684439069062,"output_power":0.0004318978420521389,"reference_active":true,"delta_db":-0.08746681353970125},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001235413948928314,"output_power":0.0001147373727670297,"reference_active":true,"delta_db":-0.3210759932868329},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6241583124510377e-05,"output_power":2.790632638838906e-05,"reference_active":true,"delta_db":0.26712637390533994}],"active_band_count":3,"active_band_mae_db":0.22522306024395802}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011635584990380617,"log_rms_mae_db":0.5661019735127868,"rms_envelope_nrmse":0.05565039710952709,"envelope_correlation":0.9945084335280017,"rms_gain_error_db":-0.03018384384494085,"reference_mean_power":0.0024645680451820984,"output_mean_power":0.002447498469352907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.594353848464586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003918219819660693,"output_power":0.0003879162917896582,"reference_active":true,"delta_db":-0.04350777523442893},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9871492701275195e-05,"output_power":5.537066916411958e-05,"reference_active":true,"delta_db":-0.33940313935846755},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.328965955891303e-06,"output_power":1.362543315085361e-06,"reference_active":true,"delta_db":0.10836461569132179}],"active_band_count":3,"active_band_mae_db":0.16375851009473943}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010777091024361473,"log_rms_mae_db":2.1006719690859197,"rms_envelope_nrmse":0.15266145835010378,"envelope_correlation":0.978661519876237,"rms_gain_error_db":-0.054770621711888186,"reference_mean_power":0.0010415487104655132,"output_mean_power":0.0010284958020167227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.64672159263247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003552137892650671,"output_power":0.00033387580765653694,"reference_active":true,"delta_db":-0.26904864978170406},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.132330818484315e-05,"output_power":6.360688072807215e-05,"reference_active":true,"delta_db":-0.4972738046389161},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2237174874701013e-05,"output_power":1.697989837529577e-05,"reference_active":true,"delta_db":1.4225392031946849}],"active_band_count":3,"active_band_mae_db":0.729620552538435}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":1.2851884122492758,"rms_envelope_nrmse":0.0803297008999456,"envelope_correlation":0.9932621230078191,"rms_gain_error_db":-0.05483512330987725,"reference_mean_power":0.0010464471526710588,"output_mean_power":0.0010333175089210344},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.816907801707229e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300382500563707,"output_power":0.00031060115771074174,"reference_active":true,"delta_db":-0.2636120556228784},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.204099572239432e-05,"output_power":2.9388166870821598e-05,"reference_active":true,"delta_db":-0.3753350680518603},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.619198216109296e-07,"output_power":8.247038304301972e-07,"reference_active":true,"delta_db":-0.1916885661956105}],"active_band_count":3,"active_band_mae_db":0.2768785632901164}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":2.1175362152776405,"rms_envelope_nrmse":0.2433405530034974,"envelope_correlation":0.9424344035428043,"rms_gain_error_db":-0.058009627380074744,"reference_mean_power":0.0001796467459408233,"output_mean_power":0.00017726312129827553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.25198062486055e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.755898706119916e-05,"output_power":4.474565414376139e-05,"reference_active":true,"delta_db":-0.26481734706663895},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1481328556544465e-05,"output_power":9.098957723835103e-06,"reference_active":true,"delta_db":-1.0100049796391024},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593622469144874e-06,"output_power":8.558367454328485e-06,"reference_active":true,"delta_db":-0.4959169496741477}],"active_band_count":3,"active_band_mae_db":0.590246425459963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":1.307823189723149,"rms_envelope_nrmse":0.14780749933339526,"envelope_correlation":0.9757375471013221,"rms_gain_error_db":-0.05811483057044625,"reference_mean_power":0.0001804894756897876,"output_mean_power":0.00017809035527245644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.074348636306739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4204323184460245e-05,"output_power":4.225424355115742e-05,"reference_active":true,"delta_db":-0.19594414209008218},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.849625664941812e-06,"output_power":4.2596975820168915e-06,"reference_active":true,"delta_db":-1.3774930778373315},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.792037562074968e-07,"output_power":4.4010728234736866e-07,"reference_active":true,"delta_db":-1.1927281486449468}],"active_band_count":3,"active_band_mae_db":0.9220551228574535}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":2.343369061574166,"rms_envelope_nrmse":0.20518367416778066,"envelope_correlation":0.972687231450927,"rms_gain_error_db":-0.1341150620479428,"reference_mean_power":0.00012676783245330934,"output_mean_power":0.0001229129262827121},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.277972501807568e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.154032818666027e-05,"output_power":4.6007823847089475e-05,"reference_active":true,"delta_db":-0.4931548784075219},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0074595754794449e-05,"output_power":9.136973645862484e-06,"reference_active":true,"delta_db":-0.4242525674757559},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.777776179485967e-06,"output_power":3.441238874787066e-06,"reference_active":true,"delta_db":-0.4052140349376338}],"active_band_count":3,"active_band_mae_db":0.44087382694030386}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":1.732524513401771,"rms_envelope_nrmse":0.13840728603557406,"envelope_correlation":0.9873287081138805,"rms_gain_error_db":-0.13436340838284527,"reference_mean_power":0.00012731487400123045,"output_mean_power":0.00012343627398024308},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.770926544449125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2932423481689414e-05,"output_power":4.618139646527089e-05,"reference_active":true,"delta_db":-0.592547170996487},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6914627477314786e-06,"output_power":4.406709226270186e-06,"reference_active":true,"delta_db":-0.2719387735963089},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8562019106765283e-07,"output_power":2.636675163546826e-07,"reference_active":true,"delta_db":-0.347322770163648}],"active_band_count":3,"active_band_mae_db":0.4039362382521479}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9834558963775635,"predicted_mos":2.9834558963775635,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ed712b2f90ebe56e6ac420a8defa4607bc2e354209a9872203583bf20cc3f580","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1018308929997147},"started_at":"2026-10-13T04:42:56.005991+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0434000362001825e-05,"native_input_samples":96000,"native_input_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","output_sha256":"f633ea62bdb78c430fe430a6a46f365672c48be6434bee015e793996d928a7f7","output_samples":96000,"length_delta_samples":0,"analysis_s":1.9330784550002136,"synthesis_s":0.09554709599979105,"total_s":2.0286255510000046,"rtf":0.33810425850000075,"output_peak":0.5147063685391089,"output_rms":0.05151160259455153,"finished_at":"2026-10-13T04:42:58.042122+00:00","evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"f633ea62bdb78c430fe430a6a46f365672c48be6434bee015e793996d928a7f7","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.0162710641150134,"spectral_convergence":0.710110864005918,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6235563691622295,"spectral_convergence":0.36205897651352764,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.754163857607573,"spectral_convergence":0.3403075722457641,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0005892831067705275,"delta_db":-0.17681459777076775,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0010928217118638328,"delta_db":0.5556182590256681,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0005895637117422101,"delta_db":0.3621572951430667,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00015152353200127222,"delta_db":0.43637716829861645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":9.221326076539004e-05,"delta_db":0.2926840101409726,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":0.0001380398787157746,"delta_db":0.8929949394311127,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4527743783017007,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.34046266513781065,"pitch":null,"output_rms_dbfs":-25.761898768591628,"output_peak":0.5147063685391089,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.1313304302949385},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.0162710641150134,"spectral_convergence":0.710110864005918,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6235563691622295,"spectral_convergence":0.36205897651352764,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.754163857607573,"spectral_convergence":0.3403075722457641,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0005892831067705275,"delta_db":-0.17681459777076775,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0010928217118638328,"delta_db":0.5556182590256681,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0005895637117422101,"delta_db":0.3621572951430667,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00015152353200127222,"delta_db":0.43637716829861645,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":9.221326076539004e-05,"delta_db":0.2926840101409726,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":0.0001380398787157746,"delta_db":0.8929949394311127,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4527743783017007,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.34046266513781065,"pitch":null,"output_rms_dbfs":-25.761898768591628,"output_peak":0.5147063685391089,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.1313304302949385},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8807136774873267,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04953155110178405,"rms_dbfs":-26.102361433729442,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43919705387114,"mean":-0.0009115762710571289,"zero_energy":false},"output":{"rms":0.05151160259455153,"rms_dbfs":-25.761898768591628,"peak":0.5147063685391089,"peak_dbfs":-5.768809162516262,"crest_factor_db":19.99308960607537,"mean":1.3371823936569739e-06,"zero_energy":false},"rms_difference_db":0.34046266513781376},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":96,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.9701492537313433,"absolute_difference_cents_median":3.9599764575426435,"absolute_difference_cents_p95":13.445349445866995,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013602681282863616,"log_rms_mae_db":2.3307703023092214,"rms_envelope_nrmse":0.16525527690720013,"envelope_correlation":0.9652456574075051,"rms_gain_error_db":0.3423583077761059,"reference_mean_power":0.0024543453947075094,"output_mean_power":0.002655654121032043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.153929939453048e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000440684439069062,"output_power":0.000490164065873473,"reference_active":true,"delta_db":0.4621375429399927},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001235413948928314,"output_power":0.0001268186823749573,"reference_active":true,"delta_db":0.11370736024413304},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6241583124510377e-05,"output_power":2.707136801426749e-05,"reference_active":true,"delta_db":0.1352017081778255}],"active_band_count":3,"active_band_mae_db":0.2370155371206504}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011635584990380617,"log_rms_mae_db":1.5013328563842416,"rms_envelope_nrmse":0.09103873579687745,"envelope_correlation":0.989703066618108,"rms_gain_error_db":0.3450011059521386,"reference_mean_power":0.0024645680451820984,"output_mean_power":0.002668338508574965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.594353848464586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003918219819660693,"output_power":0.0004376489979095286,"reference_active":true,"delta_db":0.48037141297930136},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9871492701275195e-05,"output_power":6.312938087580846e-05,"reference_active":true,"delta_db":0.23011444120042612},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.328965955891303e-06,"output_power":1.446094642114348e-06,"reference_active":true,"delta_db":0.3668286125948872}],"active_band_count":3,"active_band_mae_db":0.3591048222582049}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010777091024361473,"log_rms_mae_db":2.877381967976547,"rms_envelope_nrmse":0.17902222123824146,"envelope_correlation":0.9756240813696886,"rms_gain_error_db":0.5007428141357666,"reference_mean_power":0.0010415487104655132,"output_mean_power":0.0011688367740959862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.64672159263247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003552137892650671,"output_power":0.00038723972800668016,"reference_active":true,"delta_db":0.37490089826051504},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.132330818484315e-05,"output_power":6.567311061768032e-05,"reference_active":true,"delta_db":-0.3584389162181934},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2237174874701013e-05,"output_power":7.369275648482796e-06,"reference_active":true,"delta_db":-2.202563646888827}],"active_band_count":3,"active_band_mae_db":0.9786344871225118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":2.1628765901441955,"rms_envelope_nrmse":0.10930333072501845,"envelope_correlation":0.9919406246196841,"rms_gain_error_db":0.5018758319782463,"reference_mean_power":0.0010464471526710588,"output_mean_power":0.0011746402651880378},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.816907801707229e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300382500563707,"output_power":0.00035820036474308,"reference_active":true,"delta_db":0.355617480268139},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.204099572239432e-05,"output_power":3.00496248790185e-05,"reference_active":true,"delta_db":-0.27866949071506875},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.619198216109296e-07,"output_power":5.097808145592544e-07,"reference_active":true,"delta_db":-2.2808338141274485}],"active_band_count":3,"active_band_mae_db":0.9717069283702188}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":1.8178600755764107,"rms_envelope_nrmse":0.2050355895716815,"envelope_correlation":0.9626855500536251,"rms_gain_error_db":0.34543577533201275,"reference_mean_power":0.0001796467459408233,"output_mean_power":0.0001945194050069529},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.25198062486055e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.755898706119916e-05,"output_power":5.196681736053932e-05,"reference_active":true,"delta_db":0.38493523652075656},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1481328556544465e-05,"output_power":9.805432787405784e-06,"reference_active":true,"delta_db":-0.6852537804717695},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593622469144874e-06,"output_power":5.953688146353525e-06,"reference_active":true,"delta_db":-2.071965413901279}],"active_band_count":3,"active_band_mae_db":1.047384810297935}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":1.0199131600013411,"rms_envelope_nrmse":0.10539700134100274,"envelope_correlation":0.9899839973474038,"rms_gain_error_db":0.34661109059917733,"reference_mean_power":0.0001804894756897876,"output_mean_power":0.00019548479914595023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.074348636306739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4204323184460245e-05,"output_power":4.748959286323153e-05,"reference_active":true,"delta_db":0.3113370087097813},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.849625664941812e-06,"output_power":4.62139541260717e-06,"reference_active":true,"delta_db":-1.0235494625768273},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.792037562074968e-07,"output_power":3.0150460662483606e-07,"reference_active":true,"delta_db":-2.835374176312846}],"active_band_count":3,"active_band_mae_db":1.3900868825331516}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":2.018734160633856,"rms_envelope_nrmse":0.2945270964485734,"envelope_correlation":0.9577383228273569,"rms_gain_error_db":0.9320931659595786,"reference_mean_power":0.00012676783245330934,"output_mean_power":0.00015711526460939856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.277972501807568e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.154032818666027e-05,"output_power":6.896945189588124e-05,"reference_active":true,"delta_db":1.2650959501563546},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0074595754794449e-05,"output_power":1.873573502906996e-05,"reference_active":true,"delta_db":2.694431067352698},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.777776179485967e-06,"output_power":6.548243392072856e-06,"reference_active":true,"delta_db":2.3888858949619944}],"active_band_count":3,"active_band_mae_db":2.116137637490349}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":1.1425853194391324,"rms_envelope_nrmse":0.20994455219765232,"envelope_correlation":0.9814210769164794,"rms_gain_error_db":0.9350727908855073,"reference_mean_power":0.00012731487400123045,"output_mean_power":0.00015790156108715886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.770926544449125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2932423481689414e-05,"output_power":7.419212841898302e-05,"reference_active":true,"delta_db":1.4663605189292739},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6914627477314786e-06,"output_power":8.703099857378456e-06,"reference_active":true,"delta_db":2.68365694359428},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8562019106765283e-07,"output_power":4.5205890008718957e-07,"reference_active":true,"delta_db":1.994061186641773}],"active_band_count":3,"active_band_mae_db":2.048026216388442}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"facodec","status":"measured","mos":{"status":"measured","value":3.869527816772461,"predicted_mos":3.869527816772461,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"0b6c2cbfbb26177c61e1c815d180e700c8ae1b4b22c3c128726886a33e47a7fc","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.696948548000364},"started_at":"2026-10-13T06:56:30.847576+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.108699977223296e-05,"native_input_samples":96000,"native_input_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","output_sha256":"eba67554e92a521b613cebee81883a2c8101f73763be2fa25ab5d0a2e6f45bd6","output_samples":96000,"length_delta_samples":0,"analysis_s":1.9183070690014574,"synthesis_s":7.453214351999122,"total_s":9.37152142100058,"rtf":1.56192023683343,"output_peak":0.3603735566139221,"output_rms":0.047925554917246864,"finished_at":"2026-10-13T06:56:40.230439+00:00","evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"eba67554e92a521b613cebee81883a2c8101f73763be2fa25ab5d0a2e6f45bd6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.585532868026513,"spectral_convergence":0.2794647563437498,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.897163483873238,"spectral_convergence":0.29744530415090775,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.942898547462242,"spectral_convergence":0.3207649683246352,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.000597645183715758,"delta_db":-0.11562026970311182,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0009542541422848505,"delta_db":-0.03323435119695344,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0004924542806862451,"delta_db":-0.4194909795324627,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00010894861400171837,"delta_db":-0.9962065861023545,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":6.812275327619593e-05,"delta_db":-1.022327861616123,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":7.54338401612773e-05,"delta_db":-1.7313885789761732,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7197114378545297,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2862955630815978,"pitch":null,"output_rms_dbfs":-26.388656996811036,"output_peak":0.3603735566139221,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.808531633120665},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.585532868026513,"spectral_convergence":0.2794647563437498,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.897163483873238,"spectral_convergence":0.29744530415090775,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.942898547462242,"spectral_convergence":0.3207649683246352,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.000597645183715758,"delta_db":-0.11562026970311182,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0009542541422848505,"delta_db":-0.03323435119695344,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0004924542806862451,"delta_db":-0.4194909795324627,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00010894861400171837,"delta_db":-0.9962065861023545,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":6.812275327619593e-05,"delta_db":-1.022327861616123,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":7.54338401612773e-05,"delta_db":-1.7313885789761732,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7197114378545297,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2862955630815978,"pitch":null,"output_rms_dbfs":-26.388656996811036,"output_peak":0.3603735566139221,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.808531633120665},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8668933960598071,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04953155110178405,"rms_dbfs":-26.102361433729442,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43919705387114,"mean":-0.0009115762710571289,"zero_energy":false},"output":{"rms":0.047925554917246864,"rms_dbfs":-26.388656996811036,"peak":0.3603735566139221,"peak_dbfs":-8.864941681161874,"crest_factor_db":17.523715315649163,"mean":-6.315612506140648e-05,"zero_energy":false},"rms_difference_db":-0.28629556308159465},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":72,"both_estimated_voiced_frames":59,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.8805970149253731,"absolute_difference_cents_median":2.309118831152786,"absolute_difference_cents_p95":7.005185579204488,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013602681282863616,"log_rms_mae_db":1.323050672849633,"rms_envelope_nrmse":0.11271938411151193,"envelope_correlation":0.982776837536757,"rms_gain_error_db":-0.28592039756635546,"reference_mean_power":0.0024543453947075094,"output_mean_power":0.0022979662198224396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.153929939453048e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000440684439069062,"output_power":0.00040394271322402445,"reference_active":true,"delta_db":-0.3780793724207055},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001235413948928314,"output_power":0.0001115542633640488,"reference_active":true,"delta_db":-0.44326327864267606},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6241583124510377e-05,"output_power":2.4701802087537335e-05,"reference_active":true,"delta_db":-0.26261394181150743}],"active_band_count":3,"active_band_mae_db":0.3613188642916296}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011635584990380617,"log_rms_mae_db":1.104686948158063,"rms_envelope_nrmse":0.09438688757309303,"envelope_correlation":0.9859341133926072,"rms_gain_error_db":-0.2856675208000954,"reference_mean_power":0.0024645680451820984,"output_mean_power":0.002307671896760155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.594353848464586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003918219819660693,"output_power":0.0003603645446832408,"reference_active":true,"delta_db":-0.36346741776459646},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9871492701275195e-05,"output_power":5.254679673001274e-05,"reference_active":true,"delta_db":-0.5667383956654513},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.328965955891303e-06,"output_power":1.4196647962193685e-06,"reference_active":true,"delta_db":0.28671957392104047}],"active_band_count":3,"active_band_mae_db":0.405641795783696}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010777091024361473,"log_rms_mae_db":1.6672437469757306,"rms_envelope_nrmse":0.13167187578861772,"envelope_correlation":0.9847304072634526,"rms_gain_error_db":-0.2805312361069232,"reference_mean_power":0.0010415487104655132,"output_mean_power":0.0009763970575065291},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.64672159263247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003552137892650671,"output_power":0.00032747282812314464,"reference_active":true,"delta_db":-0.35314546103267197},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.132330818484315e-05,"output_power":6.655850706721236e-05,"reference_active":true,"delta_db":-0.30027906825648487},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2237174874701013e-05,"output_power":1.2628125064642348e-05,"reference_active":true,"delta_db":0.1365770800240941}],"active_band_count":3,"active_band_mae_db":0.26333386977108364}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":1.245221408896139,"rms_envelope_nrmse":0.11224974726170059,"envelope_correlation":0.9874608593338728,"rms_gain_error_db":-0.2801138625622478,"reference_mean_power":0.0010464471526710588,"output_mean_power":0.0009810833703919769},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.816907801707229e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300382500563707,"output_power":0.0003048760804106855,"reference_active":true,"delta_db":-0.3444092334928076},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.204099572239432e-05,"output_power":3.009323052816399e-05,"reference_active":true,"delta_db":-0.2723719193763483},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.619198216109296e-07,"output_power":9.310733900836214e-07,"reference_active":true,"delta_db":0.33517046447325116}],"active_band_count":3,"active_band_mae_db":0.31731720578080236}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":1.7543794425593797,"rms_envelope_nrmse":0.2000616599892376,"envelope_correlation":0.9664826519806924,"rms_gain_error_db":-0.8170658247505714,"reference_mean_power":0.0001796467459408233,"output_mean_power":0.00014883763908643685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.25198062486055e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.755898706119916e-05,"output_power":3.772572278974628e-05,"reference_active":true,"delta_db":-1.0059502698469538},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1481328556544465e-05,"output_power":9.748426512299888e-06,"reference_active":true,"delta_db":-0.7105762300949298},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593622469144874e-06,"output_power":8.766568030770567e-06,"reference_active":true,"delta_db":-0.3915301660101886}],"active_band_count":3,"active_band_mae_db":0.7026855553173573}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":1.3194243036598698,"rms_envelope_nrmse":0.14800751379246857,"envelope_correlation":0.9831807440072177,"rms_gain_error_db":-0.8165447536199294,"reference_mean_power":0.0001804894756897876,"output_mean_power":0.00014955378468350022},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.074348636306739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4204323184460245e-05,"output_power":3.5498422722800996e-05,"reference_active":true,"delta_db":-0.9525568866942457},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.849625664941812e-06,"output_power":5.553433245578961e-06,"reference_active":true,"delta_db":-0.22566519320452397},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.792037562074968e-07,"output_power":5.783074485807451e-07,"reference_active":true,"delta_db":-0.006725836131462324}],"active_band_count":3,"active_band_mae_db":0.39498263867674405}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":2.1175429720135193,"rms_envelope_nrmse":0.2666384304672127,"envelope_correlation":0.974033087186124,"rms_gain_error_db":-1.901568003675921,"reference_mean_power":0.00012676783245330934,"output_mean_power":8.181864148698854e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.277972501807568e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.154032818666027e-05,"output_power":3.437924503353182e-05,"reference_active":true,"delta_db":-1.7585084425324373},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0074595754794449e-05,"output_power":5.987835453893465e-06,"reference_active":true,"delta_db":-2.259577716137876},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.777776179485967e-06,"output_power":2.5318229268163335e-06,"reference_active":true,"delta_db":-1.7380289558054127}],"active_band_count":3,"active_band_mae_db":1.918705038158575}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":1.6727815457336974,"rms_envelope_nrmse":0.23469012828416164,"envelope_correlation":0.9864546549035791,"rms_gain_error_db":-1.9030010193921296,"reference_mean_power":0.00012731487400123045,"output_mean_power":8.214460443778382e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.770926544449125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2932423481689414e-05,"output_power":3.490543971986944e-05,"reference_active":true,"delta_db":-1.8082866499466754},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6914627477314786e-06,"output_power":3.060485369427614e-06,"reference_active":true,"delta_db":-1.8551796448064684},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8562019106765283e-07,"output_power":1.735762310738616e-07,"reference_active":true,"delta_db":-2.162986512084276}],"active_band_count":3,"active_band_mae_db":1.9421509356124733}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.081138849258423,"predicted_mos":3.081138849258423,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"2410f19bfa5f984c18beb861ab0d249da3ebb9c17ecd4bab8418e2ae03c1172c","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.704606265997427},"started_at":"2026-10-13T07:23:06.869596+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00640999200004444,"native_input_samples":264600,"native_input_sha256":"9721106ad8859ecceb9f7f0ee5bf75da83a87a74cbb48b537f21136d1f90054d","output_sha256":"b9f322099dfeb7f0109ab3dec7b11eda2c53f7f79a50538940c5ec9ceb10dede","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1326866550007253,"synthesis_s":3.931035564999547,"total_s":4.0637222200002725,"rtf":0.6772870366667121,"output_peak":0.33023419976234436,"output_rms":0.04330064097971373,"finished_at":"2026-10-13T07:23:10.952375+00:00","evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"b9f322099dfeb7f0109ab3dec7b11eda2c53f7f79a50538940c5ec9ceb10dede","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.265399749397297,"output_peak":0.32609336723416493},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.890549194938933,"spectral_convergence":0.5504305212093207,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.628994349800772,"spectral_convergence":0.4310327290200214,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.460087508515407,"spectral_convergence":0.4783341560079112,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0004284734134899665,"delta_db":-1.5608157155453761,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0007845474431702993,"delta_db":-0.8836827757446006,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0004323035068437692,"delta_db":-0.9852625634114299,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":9.261721317166622e-05,"delta_db":-1.7015065927233655,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":7.547911089953089e-05,"delta_db":-0.5769820264072625,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":6.40110793787271e-05,"delta_db":-2.4444992460457375,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3587914866462956,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.161996945853985,"pitch":null,"output_rms_dbfs":-27.264358379583427,"output_peak":0.32609336723416493,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.659877017751704},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.711249884328151,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04953155110178405,"rms_dbfs":-26.102361433729442,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43919705387114,"mean":-0.0009115762710571289,"zero_energy":false},"output":{"rms":0.04332934071681635,"rms_dbfs":-27.264358379583427,"peak":0.32609336723416493,"peak_dbfs":-9.733160693791763,"crest_factor_db":17.53119768579166,"mean":-0.0007505816847842532,"zero_energy":false},"rms_difference_db":-1.1619969458539856},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":56,"both_estimated_voiced_frames":42,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.6268656716417911,"absolute_difference_cents_median":10.628669022459304,"absolute_difference_cents_p95":30.384359483189566,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013602681282863616,"log_rms_mae_db":2.1963312775664687,"rms_envelope_nrmse":0.18460162958478613,"envelope_correlation":0.9716930867373207,"rms_gain_error_db":-1.1604254732142871,"reference_mean_power":0.0024543453947075094,"output_mean_power":0.0018788544281930812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.153929939453048e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000440684439069062,"output_power":0.00038667630978010285,"reference_active":true,"delta_db":-0.5678015020454159},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001235413948928314,"output_power":8.927569287485906e-05,"reference_active":true,"delta_db":-1.4107927113488172},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6241583124510377e-05,"output_power":1.9465078833269222e-05,"reference_active":true,"delta_db":-1.2973386503657283}],"active_band_count":3,"active_band_mae_db":1.0919776212533205}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011635584990380617,"log_rms_mae_db":1.7683405021469223,"rms_envelope_nrmse":0.15495519779145597,"envelope_correlation":0.9855900791116393,"rms_gain_error_db":-1.1588652433843003,"reference_mean_power":0.0024645680451820984,"output_mean_power":0.0018873580112778196},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.594353848464586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003918219819660693,"output_power":0.0003467920143116171,"reference_active":true,"delta_db":-0.5301970881293283},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9871492701275195e-05,"output_power":4.181017838947614e-05,"reference_active":true,"delta_db":-1.5593806546573559},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.328965955891303e-06,"output_power":9.629991720828217e-07,"reference_active":true,"delta_db":-1.398879420178234}],"active_band_count":3,"active_band_mae_db":1.1628190543216392}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010777091024361473,"log_rms_mae_db":3.0085263090562195,"rms_envelope_nrmse":0.2171770298125177,"envelope_correlation":0.9587978350299442,"rms_gain_error_db":-0.6363425898386517,"reference_mean_power":0.0010415487104655132,"output_mean_power":0.0008995914654563467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.64672159263247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003552137892650671,"output_power":0.0003559231878975965,"reference_active":true,"delta_db":0.008664660217810993},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.132330818484315e-05,"output_power":5.630604539055444e-05,"reference_active":true,"delta_db":-1.0267645273482102},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2237174874701013e-05,"output_power":1.1514180053921935e-05,"reference_active":true,"delta_db":-0.26448149856145947}],"active_band_count":3,"active_band_mae_db":0.4333035620424936}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":2.145518297706825,"rms_envelope_nrmse":0.17357421527724232,"envelope_correlation":0.9717227081233089,"rms_gain_error_db":-0.6356042378294274,"reference_mean_power":0.0010464471526710588,"output_mean_power":0.000903975951019768},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.816907801707229e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300382500563707,"output_power":0.00033035460566609425,"reference_active":true,"delta_db":0.00416090225510855},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.204099572239432e-05,"output_power":2.5041903794046957e-05,"reference_active":true,"delta_db":-1.0703866123762165},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.619198216109296e-07,"output_power":7.163584842243478e-07,"reference_active":true,"delta_db":-0.8033645947455752}],"active_band_count":3,"active_band_mae_db":0.6259707031256334}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":3.441015904129147,"rms_envelope_nrmse":0.2900618611070805,"envelope_correlation":0.9323949272940355,"rms_gain_error_db":-1.436743030509038,"reference_mean_power":0.0001796467459408233,"output_mean_power":0.0001290461499090965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.25198062486055e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.755898706119916e-05,"output_power":3.105107019419029e-05,"reference_active":true,"delta_db":-1.851560231938158},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1481328556544465e-05,"output_power":1.1322423036599114e-05,"reference_active":true,"delta_db":-0.06052767874048689},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593622469144874e-06,"output_power":7.246082221618014e-06,"reference_active":true,"delta_db":-1.2187936630559375}],"active_band_count":3,"active_band_mae_db":1.043627191244861}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":2.6691920913956,"rms_envelope_nrmse":0.22781329248189056,"envelope_correlation":0.9632278735755341,"rms_gain_error_db":-1.4358577361259715,"reference_mean_power":0.0001804894756897876,"output_mean_power":0.00012967794203771538},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.074348636306739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4204323184460245e-05,"output_power":2.948388855312806e-05,"reference_active":true,"delta_db":-1.758799845123089},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.849625664941812e-06,"output_power":5.459673413191391e-06,"reference_active":true,"delta_db":-0.2996141031808687},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.792037562074968e-07,"output_power":4.47759602283083e-07,"reference_active":true,"delta_db":-1.1178646150912421}],"active_band_count":3,"active_band_mae_db":1.0587595211317333}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":4.026385733308076,"rms_envelope_nrmse":0.2926850264105895,"envelope_correlation":0.9608604018744777,"rms_gain_error_db":-1.8433897374429484,"reference_mean_power":0.00012676783245330934,"output_mean_power":8.29220615804544e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.277972501807568e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.154032818666027e-05,"output_power":3.272595134728878e-05,"reference_active":true,"delta_db":-1.9725489922588944},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0074595754794449e-05,"output_power":7.495956848081355e-06,"reference_active":true,"delta_db":-1.2840055122112908},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.777776179485967e-06,"output_power":1.271468563897506e-06,"reference_active":true,"delta_db":-4.729305968222415}],"active_band_count":3,"active_band_mae_db":2.6619534908975333}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":3.263688946645419,"rms_envelope_nrmse":0.24205636270206635,"envelope_correlation":0.9818435608431398,"rms_gain_error_db":-1.8403634233495094,"reference_mean_power":0.00012731487400123045,"output_mean_power":8.333794789684693e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.770926544449125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2932423481689414e-05,"output_power":3.283470890209123e-05,"reference_active":true,"delta_db":-2.073886079995814},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6914627477314786e-06,"output_power":3.442960454656706e-06,"reference_active":true,"delta_db":-1.3437623769706946},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8562019106765283e-07,"output_power":8.529190319829215e-08,"reference_active":true,"delta_db":-5.24881100003994}],"active_band_count":3,"active_band_mae_db":2.8888198190021495}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"golf","status":"measured","mos":{"status":"measured","value":2.2293269634246826,"predicted_mos":2.2293269634246826,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"65cc7b524c595f74cee38758a7d0452137edcdeb2f628399af907d825324906b","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1103296740002406},"started_at":"2026-10-13T05:02:49.139918+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026611450002747006,"native_input_samples":144000,"native_input_sha256":"74f969fd5fe8e497fdd246996e2ff534f4ba92ec002d5099d2caf01eb13de886","output_sha256":"adcc482bf95006f47b5cb2921820e5edf902855bc4a4b7a15423307ea6525194","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6619391799995356,"synthesis_s":0.11658754200016119,"total_s":0.7785267219996967,"rtf":0.12975445366661612,"output_peak":0.34059572219848633,"output_rms":0.04683156944186385,"finished_at":"2026-10-13T05:02:49.931525+00:00","analysis_stages":{"external_f0_s":0.08292923599947244},"evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"adcc482bf95006f47b5cb2921820e5edf902855bc4a4b7a15423307ea6525194","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.58680504311553,"output_peak":0.34079857502425803},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.142811044129603,"spectral_convergence":0.7494380049696366,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.624788476130752,"spectral_convergence":0.4048602844725961,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.797627795548569,"spectral_convergence":0.4517918157116544,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006141022524296969,"output_power":0.000555410069430229,"delta_db":-0.43626941184498047,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009619800606719209,"output_power":0.0008964762892924718,"delta_db":-0.3062726285426989,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005430803616702298,"output_power":0.0004831477743689415,"delta_db":-0.5078411526044766,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013719627407782464,"output_power":0.00010361860873919085,"delta_db":-1.219045603047653,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.626542487362229e-05,"output_power":6.369283790332886e-05,"delta_db":-1.3174616586413381,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011237982092236106,"output_power":9.207312119613714e-05,"delta_db":-0.8655547011563804,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7754075259729213,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.48732742811234747,"pitch":null,"output_rms_dbfs":-26.58680504311553,"output_peak":0.34079857502425803,"reference_rms_dbfs":-26.09947761500318,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.188409105269641},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8244128922070718,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04954799889648073,"rms_dbfs":-26.09947761500318,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43631323514488,"mean":-0.0008977409197213455,"zero_energy":false},"output":{"rms":0.04684462296710583,"rms_dbfs":-26.586805043115525,"peak":0.34079857502425803,"peak_dbfs":-9.350044596205072,"crest_factor_db":17.236760446910456,"mean":0.000475530721705016,"zero_energy":false},"rms_difference_db":-0.4873274281123443},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":44,"both_estimated_voiced_frames":35,"reference_voiced_output_unvoiced_frames":32,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.5223880597014925,"absolute_difference_cents_median":6.012294348686926,"absolute_difference_cents_p95":20.880502726817827,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00013602681282863616,"log_rms_mae_db":2.1391508958286938,"rms_envelope_nrmse":0.18239631292821953,"envelope_correlation":0.955441132339627,"rms_gain_error_db":-0.48675592481767704,"reference_mean_power":0.0024565842739830016,"output_mean_power":0.0021961200394437458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.163860887816869e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044026055475507623,"output_power":0.0004039221286170985,"reference_active":true,"delta_db":-0.3741213012035275},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012354660544529653,"output_power":0.00011681504230887095,"reference_active":true,"delta_db":-0.24332046657234815},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6264013948338786e-05,"output_power":3.0317325114451194e-05,"reference_active":true,"delta_db":0.6232978065348861}],"active_band_count":3,"active_band_mae_db":0.4135798581035872}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00011635584990380617,"log_rms_mae_db":1.293807863742161,"rms_envelope_nrmse":0.11146882781630386,"envelope_correlation":0.9832722722849955,"rms_gain_error_db":-0.4865904773141605,"reference_mean_power":0.002467702642159361,"output_mean_power":0.002206143603490564},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.601056954850817e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000391506947262984,"output_power":0.00035844711490019723,"reference_active":true,"delta_db":-0.3831438383754362},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.96201552423696e-05,"output_power":5.977240507245588e-05,"reference_active":true,"delta_db":0.011076284274960371},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3251002824061527e-06,"output_power":1.6766293646330178e-06,"reference_active":true,"delta_db":1.0218832162560618}],"active_band_count":3,"active_band_mae_db":0.4720344463021528}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00010777091024361473,"log_rms_mae_db":2.031345936065252,"rms_envelope_nrmse":0.20035708981450526,"envelope_correlation":0.9634446647284856,"rms_gain_error_db":-0.30001103504225246,"reference_mean_power":0.0010429188486234607,"output_mean_power":0.0009733060277716142},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.655256812660283e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035521661769756226,"output_power":0.00033493692695591104,"reference_active":true,"delta_db":-0.2553024333648003},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.131913639702657e-05,"output_power":6.276453894321208e-05,"reference_active":true,"delta_db":-0.5549173274104215},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2225665440141096e-05,"output_power":1.5220152440804e-05,"reference_active":true,"delta_db":0.9514649524163863}],"active_band_count":3,"active_band_mae_db":0.5872282377305361}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":1.2338945531708871,"rms_envelope_nrmse":0.12185809143129453,"envelope_correlation":0.985161816116248,"rms_gain_error_db":-0.2997811984730903,"reference_mean_power":0.0010480646630427403,"output_mean_power":0.00097816013365761},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.823670231786802e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300384947037951,"output_power":0.00031145747640158215,"reference_active":true,"delta_db":-0.25165837256944995},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.192826855998665e-05,"output_power":2.8872946288262404e-05,"reference_active":true,"delta_db":-0.4368426484906318},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.609594281025732e-07,"output_power":9.992245689742353e-07,"reference_active":true,"delta_db":0.6468041772788167}],"active_band_count":3,"active_band_mae_db":0.4451017327796328}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":2.0757594129469408,"rms_envelope_nrmse":0.268909733506559,"envelope_correlation":0.9418435959631911,"rms_gain_error_db":-1.3453453570633422,"reference_mean_power":0.000179879130007188,"output_mean_power":0.0001319611959064323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.2539241310014821e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.756330994035622e-05,"output_power":3.0406245564297093e-05,"reference_active":true,"delta_db":-1.9430927104611113},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1527485897652932e-05,"output_power":9.339885381304416e-06,"reference_active":true,"delta_db":-0.9139305297170466},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.586205280089969e-06,"output_power":7.592527685414788e-06,"reference_active":true,"delta_db":-1.0126034058093012}],"active_band_count":3,"active_band_mae_db":1.2898755486624864}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":1.4760636699163725,"rms_envelope_nrmse":0.19905581153042706,"envelope_correlation":0.9753935609305351,"rms_gain_error_db":-1.3454412608582043,"reference_mean_power":0.0001807748900928601,"output_mean_power":0.00013261540626432634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0757510037910425e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4192944851896775e-05,"output_power":2.8665286749223654e-05,"reference_active":true,"delta_db":-1.8799665174292668},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.852722912093809e-06,"output_power":5.287545958096031e-06,"reference_active":true,"delta_db":-0.44103808431577807},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.775862435307944e-07,"output_power":4.773804706852565e-07,"reference_active":true,"delta_db":-0.8275219268137892}],"active_band_count":3,"active_band_mae_db":1.0495088428529447}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":2.487243574505817,"rms_envelope_nrmse":0.22219259512112885,"envelope_correlation":0.9706428749445118,"rms_gain_error_db":-0.7447338669659374,"reference_mean_power":0.00012686274095820004,"output_mean_power":0.00010687120438562873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.270508348910756e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.15630368783769e-05,"output_power":4.5411583068601115e-05,"reference_active":true,"delta_db":-0.5517184595612027},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.953062497667927e-06,"output_power":1.0456878922861018e-05,"reference_active":true,"delta_db":0.21445348362125516},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.773337838922395e-06,"output_power":5.911700233853216e-06,"reference_active":true,"delta_db":1.9498671335370996}],"active_band_count":3,"active_band_mae_db":0.905346358906519}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":1.9843585120042557,"rms_envelope_nrmse":0.14234853035015063,"envelope_correlation":0.9903134213484627,"rms_gain_error_db":-0.7446975487197872,"reference_mean_power":0.0001274978900893334,"output_mean_power":0.00010740716238187792},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.764109985573904e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.293603988654347e-05,"output_power":4.875483022923479e-05,"reference_active":true,"delta_db":-0.35733800364234436},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.585718024305607e-06,"output_power":5.365677426450091e-06,"reference_active":true,"delta_db":0.682172137675545},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8539672670263197e-07,"output_power":4.3830336158285255e-07,"reference_active":true,"delta_db":1.8632581374481552}],"active_band_count":3,"active_band_mae_db":0.9675894262553482}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.595773935317993,"predicted_mos":2.595773935317993,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"1577bb40fac2c121f98523bbae5e1960a5f4d037be1541a5b576c30399fa1673","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8128480599989416},"started_at":"2026-10-13T07:42:02.631265+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002583140001661377,"native_input_samples":144000,"native_input_sha256":"74f969fd5fe8e497fdd246996e2ff534f4ba92ec002d5099d2caf01eb13de886","output_sha256":"61efe12d8574344198c44e6d6cff49514c7c08611d301959b9a6d4e3b7985547","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3767363510014548,"synthesis_s":3.4320592260010017,"total_s":5.808795577002456,"rtf":0.968132596167076,"output_peak":0.43988192081451416,"output_rms":0.05011211702463022,"finished_at":"2026-10-13T07:42:08.452250+00:00","analysis_stages":{"external_f0_s":1.5658618649977143},"evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"61efe12d8574344198c44e6d6cff49514c7c08611d301959b9a6d4e3b7985547","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.998549047696137,"output_peak":0.44074600496102756},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.862085442137434,"spectral_convergence":0.7228982127455015,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.4833567119546505,"spectral_convergence":0.4226826488901196,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.458669540308082,"spectral_convergence":0.4447505268594161,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0008855555331255694,"delta_db":1.5921035001062571,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0008812608176656867,"delta_db":-0.37883027866865177,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.00046148215438757253,"delta_db":-0.7016010345637858,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00011499273265476368,"delta_db":-0.7617197376872846,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":7.713242667994018e-05,"delta_db":-0.48287983414813546,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":8.782153882946829e-05,"delta_db":-1.0710403245370583,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8313624516185288,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.09606133766648281,"pitch":null,"output_rms_dbfs":-26.006300096062954,"output_peak":0.44074600496102756,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.934703898133389},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8372817996563847,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04953155110178405,"rms_dbfs":-26.102361433729442,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43919705387114,"mean":-0.0009115762710571289,"zero_energy":false},"output":{"rms":0.050082384161928645,"rms_dbfs":-26.006300096062954,"peak":0.44074600496102756,"peak_dbfs":-7.116232311429032,"crest_factor_db":18.890067784633924,"mean":0.00015497354400953347,"zero_energy":false},"rms_difference_db":0.09606133766648739},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":67,"both_estimated_voiced_frames":53,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.7910447761194029,"absolute_difference_cents_median":6.199175510869296,"absolute_difference_cents_p95":27.396109841162048,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013602681282863616,"log_rms_mae_db":2.9960971141789794,"rms_envelope_nrmse":0.20717221700319052,"envelope_correlation":0.9387701259218866,"rms_gain_error_db":0.09589805101968261,"reference_mean_power":0.0024543453947075094,"output_mean_power":0.0025091434158101616},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.153929939453048e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000440684439069062,"output_power":0.0004557575217925192,"reference_active":true,"delta_db":0.1460612946968256},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001235413948928314,"output_power":0.0001091701876396002,"reference_active":true,"delta_db":-0.537084438389804},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6241583124510377e-05,"output_power":2.2736620905368418e-05,"reference_active":true,"delta_db":-0.6226411125906589}],"active_band_count":3,"active_band_mae_db":0.4352622818924295}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011635584990380617,"log_rms_mae_db":2.1439146589600737,"rms_envelope_nrmse":0.12512791541946924,"envelope_correlation":0.9733331408251199,"rms_gain_error_db":0.09208522407687271,"reference_mean_power":0.0024645680451820984,"output_mean_power":0.0025173832353502152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.594353848464586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003918219819660693,"output_power":0.00040541323511651694,"reference_active":true,"delta_db":0.1480912546471554},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9871492701275195e-05,"output_power":5.364750966972488e-05,"reference_active":true,"delta_db":-0.4767051928768494},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.328965955891303e-06,"output_power":1.2150468988454137e-06,"reference_active":true,"delta_db":-0.3892081444492423}],"active_band_count":3,"active_band_mae_db":0.33800153065774907}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010777091024361473,"log_rms_mae_db":2.6066186518454346,"rms_envelope_nrmse":0.23970025703186182,"envelope_correlation":0.9480964332871367,"rms_gain_error_db":-0.5778089220755451,"reference_mean_power":0.0010415487104655132,"output_mean_power":0.000911798121968415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.64672159263247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003552137892650671,"output_power":0.0003118393975273346,"reference_active":true,"delta_db":-0.5655883386259526},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.132330818484315e-05,"output_power":5.63917458744878e-05,"reference_active":true,"delta_db":-1.020159383800098},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2237174874701013e-05,"output_power":9.820363835606153e-06,"reference_active":true,"delta_db":-0.9555358804031597}],"active_band_count":3,"active_band_mae_db":0.8470945342764035}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":1.6885057446559504,"rms_envelope_nrmse":0.14818215541542704,"envelope_correlation":0.9800526453434124,"rms_gain_error_db":-0.5785656985188038,"reference_mean_power":0.0010464471526710588,"output_mean_power":0.0009159267243455186},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.816907801707229e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300382500563707,"output_power":0.0002898911918982016,"reference_active":true,"delta_db":-0.5632925583324072},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.204099572239432e-05,"output_power":2.6131507884075177e-05,"reference_active":true,"delta_db":-0.8854153314103324},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.619198216109296e-07,"output_power":7.260954154818325e-07,"reference_active":true,"delta_db":-0.7447317365783257}],"active_band_count":3,"active_band_mae_db":0.7311465421070218}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":2.540567076969095,"rms_envelope_nrmse":0.2857507386094215,"envelope_correlation":0.9204817108343296,"rms_gain_error_db":-0.5496101444875546,"reference_mean_power":0.0001796467459408233,"output_mean_power":0.00015829177187737418},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.25198062486055e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.755898706119916e-05,"output_power":3.976692366247798e-05,"reference_active":true,"delta_db":-0.777106005797975},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1481328556544465e-05,"output_power":8.900662422454823e-06,"reference_active":true,"delta_db":-1.1056981539333186},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593622469144874e-06,"output_power":6.506838901363048e-06,"reference_active":true,"delta_db":-1.6861256954155102}],"active_band_count":3,"active_band_mae_db":1.1896432850489347}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":1.4433019998048657,"rms_envelope_nrmse":0.14388189282159144,"envelope_correlation":0.9801183743496238,"rms_gain_error_db":-0.5508484620194789,"reference_mean_power":0.0001804894756897876,"output_mean_power":0.0001589889851335312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.074348636306739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4204323184460245e-05,"output_power":3.6606784552554645e-05,"reference_active":true,"delta_db":-0.8190316221573652},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.849625664941812e-06,"output_power":4.327978953134079e-06,"reference_active":true,"delta_db":-1.3084293502816},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.792037562074968e-07,"output_power":3.453794245348974e-07,"reference_active":true,"delta_db":-2.245349080965035}],"active_band_count":3,"active_band_mae_db":1.4576033511346669}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":2.3634133052288853,"rms_envelope_nrmse":0.25283927855038707,"envelope_correlation":0.9617945421240077,"rms_gain_error_db":-0.9670334513282419,"reference_mean_power":0.00012676783245330934,"output_mean_power":0.00010146253749136778},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.277972501807568e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.154032818666027e-05,"output_power":3.837808342760119e-05,"reference_active":true,"delta_db":-1.280638970205927},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0074595754794449e-05,"output_power":6.9198416179727766e-06,"reference_active":true,"delta_db":-1.6313147461329314},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.777776179485967e-06,"output_power":1.7748315870572563e-06,"reference_active":true,"delta_db":-3.2807907451772444}],"active_band_count":3,"active_band_mae_db":2.064248153838701}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":1.2999439955418564,"rms_envelope_nrmse":0.17496429906974534,"envelope_correlation":0.9852898111882161,"rms_gain_error_db":-0.9685962289864984,"reference_mean_power":0.00012731487400123045,"output_mean_power":0.00010186371750933283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.770926544449125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2932423481689414e-05,"output_power":3.822928343592692e-05,"reference_active":true,"delta_db":-1.4132562063541962},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6914627477314786e-06,"output_power":3.31131149194419e-06,"reference_active":true,"delta_db":-1.5130823584170108},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8562019106765283e-07,"output_power":1.4130211084995082e-07,"reference_active":true,"delta_db":-3.0564025571177496}],"active_band_count":3,"active_band_mae_db":1.9942470406296522}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.757441520690918,"predicted_mos":2.757441520690918,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"7af73ebcef136292b68619529ef30aac47eee3c60eb2a6f1512d56e39277f6e4","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8861848169999575},"started_at":"2026-10-13T05:11:07.619479+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002822487000230467,"native_input_samples":144000,"native_input_sha256":"74f969fd5fe8e497fdd246996e2ff534f4ba92ec002d5099d2caf01eb13de886","output_sha256":"fef05e9ffd1dad9eb8e5101a46e9f45f36f2b19e7b87e93fb8249ab281fd9574","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0528253590000531,"synthesis_s":0.16300959399995918,"total_s":1.2158349530000123,"rtf":0.2026391588333354,"output_peak":0.46604523062705994,"output_rms":0.04786797571641253,"finished_at":"2026-10-13T05:11:08.848174+00:00","analysis_stages":{"external_f0_s":0.1867594769992138},"evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"fef05e9ffd1dad9eb8e5101a46e9f45f36f2b19e7b87e93fb8249ab281fd9574","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.395931465916448,"output_peak":0.464632473883733},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.320890218641747,"spectral_convergence":0.34312638372444626,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.714858274364643,"spectral_convergence":0.2876873408378906,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.2101684211685,"spectral_convergence":0.35458757337345864,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0006008883873057398,"delta_db":-0.09211639972857957,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0008859732799969076,"delta_db":-0.35566864727853953,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0005100062219714268,"delta_db":-0.267395404496682,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.00011863446763307083,"delta_db":-0.6263148259269081,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":7.762881409806007e-05,"delta_db":-0.4550202674035348,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":0.00010341050902452919,"delta_db":-0.3614039920794453,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.35965325615228155,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2868952477311074,"pitch":null,"output_rms_dbfs":-26.389256681460548,"output_peak":0.464632473883733,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.081972304724964},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8865684252527025,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04953155110178405,"rms_dbfs":-26.102361433729442,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43919705387114,"mean":-0.0009115762710571289,"zero_energy":false},"output":{"rms":0.04792224619140608,"rms_dbfs":-26.389256681460548,"peak":0.464632473883733,"peak_dbfs":-6.657808799211083,"crest_factor_db":19.731447882249466,"mean":0.00011813605262249742,"zero_energy":false},"rms_difference_db":-0.2868952477311062},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":59,"both_estimated_voiced_frames":49,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.7313432835820896,"absolute_difference_cents_median":2.640865797723313,"absolute_difference_cents_p95":6.040764108416849,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013602681282863616,"log_rms_mae_db":1.159670992823871,"rms_envelope_nrmse":0.11089723388028636,"envelope_correlation":0.9829246156954646,"rms_gain_error_db":-0.28630160361964174,"reference_mean_power":0.0024543453947075094,"output_mean_power":0.002297764522535214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.153929939453048e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000440684439069062,"output_power":0.0004307909427441502,"reference_active":true,"delta_db":-0.09861151679206664},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001235413948928314,"output_power":0.00010995086555007107,"reference_active":true,"delta_db":-0.5061384805943545},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6241583124510377e-05,"output_power":2.0996612927827582e-05,"reference_active":true,"delta_db":-0.9684078988416975}],"active_band_count":3,"active_band_mae_db":0.5243859654093729}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011635584990380617,"log_rms_mae_db":0.6229111313330086,"rms_envelope_nrmse":0.06291862283788226,"envelope_correlation":0.994689544497121,"rms_gain_error_db":-0.2858651526755775,"reference_mean_power":0.0024645680451820984,"output_mean_power":0.0023075668852605693},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.594353848464586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003918219819660693,"output_power":0.00038573689366887964,"reference_active":true,"delta_db":-0.06797618302268058},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9871492701275195e-05,"output_power":5.4913371250145386e-05,"reference_active":true,"delta_db":-0.37541979201910797},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.328965955891303e-06,"output_power":1.0926422687775944e-06,"reference_active":true,"delta_db":-0.8503585861710559}],"active_band_count":3,"active_band_mae_db":0.4312515204042815}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010777091024361473,"log_rms_mae_db":1.474083421337209,"rms_envelope_nrmse":0.14739567953189986,"envelope_correlation":0.9809367530561951,"rms_gain_error_db":-0.3596338709114349,"reference_mean_power":0.0010415487104655132,"output_mean_power":0.0009587738898992539},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.64672159263247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003552137892650671,"output_power":0.00034123630437066693,"reference_active":true,"delta_db":-0.1743458673223915},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.132330818484315e-05,"output_power":5.9275976725631083e-05,"reference_active":true,"delta_db":-0.8035275998766622},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2237174874701013e-05,"output_power":1.2821251606432933e-05,"reference_active":true,"delta_db":0.20249256595597104}],"active_band_count":3,"active_band_mae_db":0.39345534438500823}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":0.8419563485739393,"rms_envelope_nrmse":0.09056342611988534,"envelope_correlation":0.9928623543395884,"rms_gain_error_db":-0.3594195602156936,"reference_mean_power":0.0010464471526710588,"output_mean_power":0.0009633305752218366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.816907801707229e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300382500563707,"output_power":0.0003184331271214202,"reference_active":true,"delta_db":-0.15546033929105507},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.204099572239432e-05,"output_power":2.733257943909261e-05,"reference_active":true,"delta_db":-0.6902538497259982},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.619198216109296e-07,"output_power":8.843138250838474e-07,"reference_active":true,"delta_db":0.1113954637415866}],"active_band_count":3,"active_band_mae_db":0.31903655091954664}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":1.3591744566990749,"rms_envelope_nrmse":0.18408626701433453,"envelope_correlation":0.9701195979453466,"rms_gain_error_db":-0.6305085539340001,"reference_mean_power":0.0001796467459408233,"output_mean_power":0.00015537047717856264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.25198062486055e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.755898706119916e-05,"output_power":4.021405864965407e-05,"reference_active":true,"delta_db":-0.7285468921022762},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1481328556544465e-05,"output_power":1.0160194144031707e-05,"reference_active":true,"delta_db":-0.5309013849568214},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593622469144874e-06,"output_power":6.466023244873198e-06,"reference_active":true,"delta_db":-1.7134536245617689}],"active_band_count":3,"active_band_mae_db":0.9909673005402887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":0.8444622391473496,"rms_envelope_nrmse":0.10882219406466444,"envelope_correlation":0.9918003804229172,"rms_gain_error_db":-0.6307018191219287,"reference_mean_power":0.0001804894756897876,"output_mean_power":0.00015609237960886494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.074348636306739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4204323184460245e-05,"output_power":3.735114993043272e-05,"reference_active":true,"delta_db":-0.7316076845457722},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.849625664941812e-06,"output_power":5.387912266920438e-06,"reference_active":true,"delta_db":-0.35707559813235445},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.792037562074968e-07,"output_power":2.9397095756373654e-07,"reference_active":true,"delta_db":-2.9452694261609302}],"active_band_count":3,"active_band_mae_db":1.3446509029463523}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":1.3017539675086884,"rms_envelope_nrmse":0.14436738033821778,"envelope_correlation":0.9866192779856108,"rms_gain_error_db":-0.22166617040019665,"reference_mean_power":0.00012676783245330934,"output_mean_power":0.00012045988592061457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.277972501807568e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.154032818666027e-05,"output_power":4.936221794747144e-05,"reference_active":true,"delta_db":-0.18752514403270956},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0074595754794449e-05,"output_power":1.0741954776945627e-05,"reference_active":true,"delta_db":0.27855690678561673},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.777776179485967e-06,"output_power":3.262650946844995e-06,"reference_active":true,"delta_db":-0.636656103776352}],"active_band_count":3,"active_band_mae_db":0.3675793848648928}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":0.7846833380471482,"rms_envelope_nrmse":0.06808458710722456,"envelope_correlation":0.9972166791327819,"rms_gain_error_db":-0.22298824727630823,"reference_mean_power":0.00012731487400123045,"output_mean_power":0.00012094288380013424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.770926544449125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2932423481689414e-05,"output_power":5.087855595759432e-05,"reference_active":true,"delta_db":-0.17187001771191815},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6914627477314786e-06,"output_power":4.80758446652739e-06,"reference_active":true,"delta_db":0.10618651097725493},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8562019106765283e-07,"output_power":2.585621360046207e-07,"reference_active":true,"delta_db":-0.43223978488281434}],"active_band_count":3,"active_band_mae_db":0.23676543785732915}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_61-70968-0054","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2230019569396973,"predicted_mos":3.2230019569396973,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289987,"input_sample_rate":48000,"input_seconds":6.0413958333333335,"input_float32_sha256":"a7eefdda9509c93bc983dac7680a203c220d5f91a7b1b255c42b75f9235d2381","internal_samples":96663,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.387671419000981},"started_at":"2026-10-13T05:17:05.824200+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010539307000726694,"native_input_samples":288000,"native_input_sha256":"a4b091361c372150e0a28c0db226b913e605539c616eb3f1b5cdcc325a3e591a","output_sha256":"a0374ac51453b26f3bb077c828379afcf8ac197a0bf6c7aae918a39882a60b41","output_samples":289987,"length_delta_samples":1987,"analysis_s":2.9227552320007817,"synthesis_s":1.069947164000041,"total_s":3.9927023960008228,"rtf":0.6654503993334705,"output_peak":0.42648058969090175,"output_rms":0.047850484222052694,"finished_at":"2026-10-13T05:17:09.840299+00:00","evaluation_input":{"reference":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"output_sha256":"a0374ac51453b26f3bb077c828379afcf8ac197a0bf6c7aae918a39882a60b41","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96663,"length_delta_samples":663,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.399098333741,"output_peak":0.4261505684963853},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9931411191458986,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.764347674657097,"spectral_convergence":0.8982950288107842,"spectral_convergence_status":"ok","amplitude_floor":1.3149451373049662e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.986748208750212,"spectral_convergence":0.7070097453447168,"spectral_convergence_status":"ok","amplitude_floor":7.412465414213296e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.061553829782108,"spectral_convergence":0.7238208925723174,"spectral_convergence_status":"ok","amplitude_floor":4.876012126794439e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006137697087039973,"output_power":0.0005341370937002619,"delta_db":-0.6035271190228952,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0009615845769051578,"output_power":0.0009031159904676752,"delta_db":-0.2724395704321937,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000542394172591287,"output_power":0.0005323340112585316,"delta_db":-0.08130800392203368,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013703842807085808,"output_power":0.0001350522221932632,"delta_db":-0.06340633711405541,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":8.620351571877408e-05,"output_power":8.118285299213566e-05,"delta_db":-0.2606066885460031,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011238415255857088,"output_power":0.00011872901498352406,"delta_db":0.2385178966456778,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25330093594714315,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2717749809371903,"pitch":null,"output_rms_dbfs":-26.37413641466663,"output_peak":0.4261505684963853,"reference_rms_dbfs":-26.102361433729442,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.937549904396471},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.48916439589054256,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04953155110178405,"rms_dbfs":-26.102361433729442,"peak":0.413848876953125,"peak_dbfs":-7.663164379858301,"crest_factor_db":18.43919705387114,"mean":-0.0009115762710571289,"zero_energy":false},"output":{"rms":0.04800574117327418,"rms_dbfs":-26.37413641466663,"peak":0.4261505684963853,"peak_dbfs":-7.408738557208209,"crest_factor_db":18.96539785745842,"mean":-0.00037350945453105305,"zero_energy":false},"rms_difference_db":-0.2717749809371881},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"no_reference_quiet_blocks","threshold_power":1.8503293808316815e-06,"threshold_dbfs":-57.3275095508632,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":0,"quiet_samples":0,"reference_mean_power":null,"output_mean_power":null,"output_rms_dbfs":null,"output_zero_energy":null,"output_peak":null,"output_energy_fraction":null,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":67,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":32,"reference_voiced_output_unvoiced_frames":35,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.47761194029850745,"absolute_difference_cents_median":18.199109995816443,"absolute_difference_cents_p95":43.3219942507874,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013602681282863616,"log_rms_mae_db":5.22765191827095,"rms_envelope_nrmse":0.52605023502034,"envelope_correlation":0.5989974472762567,"rms_gain_error_db":-0.2703032456401119,"reference_mean_power":0.0024543453947075094,"output_mean_power":0.002306244540692256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.153929939453048e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000440684439069062,"output_power":0.000461051453466746,"reference_active":true,"delta_db":0.19621679991445498},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001235413948928314,"output_power":0.00012164491271577564,"reference_active":true,"delta_db":-0.06718549486575624},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6241583124510377e-05,"output_power":2.6448398799460023e-05,"reference_active":true,"delta_db":0.03409352761131351}],"active_band_count":3,"active_band_mae_db":0.09916527413050824}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011635584990380617,"log_rms_mae_db":3.880242354932935,"rms_envelope_nrmse":0.4212425781115061,"envelope_correlation":0.6881156449872833,"rms_gain_error_db":-0.27114291724394535,"reference_mean_power":0.0024645680451820984,"output_mean_power":0.002315402626122961},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.594353848464586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003918219819660693,"output_power":0.00040833378340989087,"reference_active":true,"delta_db":0.1792651557517122},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.9871492701275195e-05,"output_power":5.966789682999546e-05,"reference_active":true,"delta_db":-0.01479355872208165},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.328965955891303e-06,"output_power":1.4925314305421805e-06,"reference_active":true,"delta_db":0.5040962974067353}],"active_band_count":3,"active_band_mae_db":0.23271833729350974}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010777091024361473,"log_rms_mae_db":7.500241981913052,"rms_envelope_nrmse":0.6501182166978771,"envelope_correlation":0.612788336449951,"rms_gain_error_db":-0.20147786583034652,"reference_mean_power":0.0010415487104655132,"output_mean_power":0.0009943328761865247},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.64672159263247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003552137892650671,"output_power":0.00035749504782351657,"reference_active":true,"delta_db":0.027802135617314008},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.132330818484315e-05,"output_power":6.445350287982692e-05,"reference_active":true,"delta_db":-0.43984953722576653},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2237174874701013e-05,"output_power":1.1907859398698076e-05,"reference_active":true,"delta_db":-0.11847468243307535}],"active_band_count":3,"active_band_mae_db":0.19537545175871862}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.988831023316543e-05,"log_rms_mae_db":6.080253357862843,"rms_envelope_nrmse":0.5360940580158275,"envelope_correlation":0.6986558851979335,"rms_gain_error_db":-0.2027179154104533,"reference_mean_power":0.0010464471526710588,"output_mean_power":0.0009987240522127065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.816907801707229e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003300382500563707,"output_power":0.0003339273392759537,"reference_active":true,"delta_db":0.05087701324162897},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.204099572239432e-05,"output_power":2.9369262487234306e-05,"reference_active":true,"delta_db":-0.37812963202401734},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.619198216109296e-07,"output_power":8.588795854788328e-07,"reference_active":true,"delta_db":-0.015345881038990533}],"active_band_count":3,"active_band_mae_db":0.14811750876821228}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.5701896659278796e-05,"log_rms_mae_db":6.326543745209106,"rms_envelope_nrmse":0.6211936402203104,"envelope_correlation":0.6222193308012922,"rms_gain_error_db":-0.22224645551294525,"reference_mean_power":0.0001796467459408233,"output_mean_power":0.00017068474568773497},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.25198062486055e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.755898706119916e-05,"output_power":4.634656463799211e-05,"reference_active":true,"delta_db":-0.11215047872725614},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1481328556544465e-05,"output_power":1.059946121296742e-05,"reference_active":true,"delta_db":-0.3470835518729355},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.593622469144874e-06,"output_power":7.298485600109362e-06,"reference_active":true,"delta_db":-1.1874986849313047}],"active_band_count":3,"active_band_mae_db":0.5489109051771655}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.308273569685667e-05,"log_rms_mae_db":4.77108964701635,"rms_envelope_nrmse":0.4826834963177132,"envelope_correlation":0.7375302616732371,"rms_gain_error_db":-0.22128450593059237,"reference_mean_power":0.0001804894756897876,"output_mean_power":0.00017152342208046322},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.074348636306739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.4204323184460245e-05,"output_power":4.2435017643186075e-05,"reference_active":true,"delta_db":-0.17740358363666733},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.849625664941812e-06,"output_power":5.10774961170552e-06,"reference_active":true,"delta_db":-0.5889847571456202},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.792037562074968e-07,"output_power":3.17786071770238e-07,"reference_active":true,"delta_db":-2.6069651103380673}],"active_band_count":3,"active_band_mae_db":1.1244511503734516}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.825215268339244e-05,"log_rms_mae_db":6.090868207608215,"rms_envelope_nrmse":0.7047690171664507,"envelope_correlation":0.6968988687379895,"rms_gain_error_db":0.377849339499618,"reference_mean_power":0.00012676783245330934,"output_mean_power":0.00013829102593841523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.277972501807568e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.154032818666027e-05,"output_power":6.1007889747574805e-05,"reference_active":true,"delta_db":0.7323882345635985},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0074595754794449e-05,"output_power":1.2962571656119198e-05,"reference_active":true,"delta_db":1.0946354114264278},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.777776179485967e-06,"output_power":5.87602418900139e-06,"reference_active":true,"delta_db":1.9184735105806148}],"active_band_count":3,"active_band_mae_db":1.2484990521902137}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.580061124327404e-05,"log_rms_mae_db":4.65057604139779,"rms_envelope_nrmse":0.5704096846418627,"envelope_correlation":0.791114951771457,"rms_gain_error_db":0.3799539915395372,"reference_mean_power":0.00012731487400123045,"output_mean_power":0.0001389551168388281},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.770926544449125e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2932423481689414e-05,"output_power":6.450281118247926e-05,"reference_active":true,"delta_db":0.85856864232218},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6914627477314786e-06,"output_power":5.771763154340997e-06,"reference_active":true,"delta_db":0.9000022922375199},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8562019106765283e-07,"output_power":4.5269521849358186e-07,"reference_active":true,"delta_db":2.000170020246809}],"active_band_count":3,"active_band_mae_db":1.2529136516021697}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"hnm","status":"measured","mos":{"status":"measured","value":2.522932529449463,"predicted_mos":2.522932529449463,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76160,"input_sample_rate":16000,"input_seconds":4.76,"input_float32_sha256":"6a3f847620d14a007469bbce0d855456dd249a2fbd85288a80be08c05b14488a","internal_samples":76160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.89428033600052},"started_at":"2026-10-13T04:39:49.308555+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3349999790079892e-05,"native_input_samples":76160,"native_input_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","output_sha256":"2e7872bfb82156fd8af66f338691c81ad9bbb68b749ba0f0da84e941a549f79b","output_samples":76160,"length_delta_samples":0,"analysis_s":0.8686276290000023,"synthesis_s":0.47254751899981784,"total_s":1.3411751479998202,"rtf":0.28175948487391184,"output_peak":0.3374406162831328,"output_rms":0.03142014030736412,"finished_at":"2026-10-13T04:39:50.654638+00:00","evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"2e7872bfb82156fd8af66f338691c81ad9bbb68b749ba0f0da84e941a549f79b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.160186172166213,"spectral_convergence":0.4875585876734439,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.824233712832519,"spectral_convergence":0.4622370196017725,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.222303386314481,"spectral_convergence":0.5161545792233538,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.00020664110175033605,"delta_db":-0.23280030025201973,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.00018555658808912544,"delta_db":-0.12900668814235736,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":4.861213896230324e-05,"delta_db":0.02252239332167729,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":4.258649993086586e-05,"delta_db":-0.6707922296330476,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":0.00011840535991561771,"delta_db":-0.18130908003080948,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.00038542352828619905,"delta_db":0.19998854475721572,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2394032060228545,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.050458848242133994,"pitch":null,"output_rms_dbfs":-30.055837598792998,"output_peak":0.3374406162831328,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.068907757104405},"common_prefix":{"samples":76160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.160186172166213,"spectral_convergence":0.4875585876734439,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.824233712832519,"spectral_convergence":0.4622370196017725,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.222303386314481,"spectral_convergence":0.5161545792233538,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.00020664110175033605,"delta_db":-0.23280030025201973,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.00018555658808912544,"delta_db":-0.12900668814235736,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":4.861213896230324e-05,"delta_db":0.02252239332167729,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":4.258649993086586e-05,"delta_db":-0.6707922296330476,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":0.00011840535991561771,"delta_db":-0.18130908003080948,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.00038542352828619905,"delta_db":0.19998854475721572,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2394032060228545,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.050458848242133994,"pitch":null,"output_rms_dbfs":-30.055837598792998,"output_peak":0.3374406162831328,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.068907757104405},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76160,"output_samples":76160,"intelligibility":{"estoi":{"value":0.734891844031261,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031603200210495876,"rms_dbfs":-30.00537875055086,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.32404640060221,"mean":-3.006077614151129e-06,"zero_energy":false},"output":{"rms":0.03142014030736412,"rms_dbfs":-30.055837598792998,"peak":0.3374406162831328,"peak_dbfs":-9.436052888642305,"crest_factor_db":20.61978471015069,"mean":-8.751160064513051e-05,"zero_energy":false},"rms_difference_db":-0.05045884824213687},"full_scale":{"reference":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":476,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":1.7876799185095088e-06,"output_mean_power":2.016260502053748e-06,"output_rms_dbfs":-56.9545335749654,"output_zero_energy":false,"output_peak":0.006572878225826913,"output_energy_fraction":0.00024027659932072723,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":469,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":27,"both_estimated_voiced_frames":21,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":1.6253363209360472,"absolute_difference_cents_p95":11.676272670338596,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018341452353274373,"log_rms_mae_db":1.5861523233698913,"rms_envelope_nrmse":0.14261759862200785,"envelope_correlation":0.9821801338786015,"rms_gain_error_db":-0.05056316506815689,"reference_mean_power":0.0009998027766238483,"output_mean_power":0.0009882299732811436},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":5.754282242130778e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018894414816227905,"output_power":0.0001992500867661149,"reference_active":true,"delta_db":0.23065073372107298},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013084274193993133,"output_power":0.00015088530916983507,"reference_active":true,"delta_db":0.6189732046968706},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.690431882761885e-05,"output_power":3.1008030565281594e-05,"reference_active":true,"delta_db":-0.7560301015407074}],"active_band_count":3,"active_band_mae_db":0.5352180133195503}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.0001230241515816465,"log_rms_mae_db":0.8604684605397941,"rms_envelope_nrmse":0.07744471335043987,"envelope_correlation":0.9943756220449913,"rms_gain_error_db":-0.050698680104414647,"reference_mean_power":0.001006100998286567,"output_mean_power":0.0009944242624931967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":5.061730772699165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019045538591950462,"output_power":0.00020718244463987978,"reference_active":true,"delta_db":0.36559694544976934},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.588390935994399e-05,"output_power":8.422783580380287e-05,"reference_active":true,"delta_db":0.45305945214492793},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5269059669432984e-06,"output_power":2.1879453953076706e-06,"reference_active":true,"delta_db":-0.6255260185786413}],"active_band_count":3,"active_band_mae_db":0.4813941387244462}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":2.1500811891413787,"rms_envelope_nrmse":0.1620894145179952,"envelope_correlation":0.9818711881000339,"rms_gain_error_db":-0.19937248877313335,"reference_mean_power":0.0001499835837122031,"output_mean_power":0.00014325390770319937},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.245105068973029e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.560041827470393e-05,"output_power":6.0330175258774575e-05,"reference_active":true,"delta_db":-0.36372021547781364},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.138220090606147e-05,"output_power":1.717208417993873e-05,"reference_active":true,"delta_db":-0.9522939705386471},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.449241698586229e-06,"output_power":2.015341872596641e-06,"reference_active":true,"delta_db":-0.8468291644612859}],"active_band_count":3,"active_band_mae_db":0.7209477834925822}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":1.334070905696113,"rms_envelope_nrmse":0.07686684483834198,"envelope_correlation":0.9960962250474665,"rms_gain_error_db":-0.19937909165328094,"reference_mean_power":0.00015093488949896852,"output_mean_power":0.00014416230977349538},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":1.1354747567630586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.429099822482352e-05,"output_power":5.839223467583616e-05,"reference_active":true,"delta_db":-0.417950729025343},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916365304226932e-05,"output_power":8.898803108178339e-06,"reference_active":true,"delta_db":-0.8874646260274767},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3162457402511593e-07,"output_power":1.8956333510615217e-07,"reference_active":true,"delta_db":-0.8703029271906171}],"active_band_count":3,"active_band_mae_db":0.7252394274144788}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":1.8236457480187493,"rms_envelope_nrmse":0.1922522091683995,"envelope_correlation":0.97321332358037,"rms_gain_error_db":-0.43574557192541596,"reference_mean_power":0.0001020893317881212,"output_mean_power":9.234338791828985e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":9.602503308503008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0812258101146943e-05,"output_power":2.6152635531768572e-05,"reference_active":true,"delta_db":-0.7120806567024751},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3389822486480089e-05,"output_power":1.0992402058373178e-05,"reference_active":true,"delta_db":-0.8568221468578889},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310273392307661e-06,"output_power":2.3140923599084995e-06,"reference_active":true,"delta_db":-3.6073619223066835}],"active_band_count":3,"active_band_mae_db":1.725421575289016}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":1.1836706153323042,"rms_envelope_nrmse":0.11806775568783037,"envelope_correlation":0.9905963291335173,"rms_gain_error_db":-0.43577742355342286,"reference_mean_power":0.00010273559922291735,"output_mean_power":9.292727798101967e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":8.714815910547829e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.937520326415114e-05,"output_power":2.467442463929736e-05,"reference_active":true,"delta_db":-0.7573384588925668},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.005890574709939e-06,"output_power":5.481274566743663e-06,"reference_active":true,"delta_db":-1.065817925751737},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.153574933783463e-07,"output_power":2.0895729152761752e-07,"reference_active":true,"delta_db":-3.920510647644857}],"active_band_count":3,"active_band_mae_db":1.9145556774297203}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018081886412701817,"log_rms_mae_db":1.706876357255869,"rms_envelope_nrmse":0.1806625962845091,"envelope_correlation":0.9819788386052198,"rms_gain_error_db":0.15074242461062592,"reference_mean_power":0.00040250829399997306,"output_mean_power":0.00041672454035561853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":3.596628676311722e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013671994273759655,"output_power":0.00016125332622367945,"reference_active":true,"delta_db":0.7167681391300079},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.299777768885513e-05,"output_power":0.00010509944477495115,"reference_active":true,"delta_db":0.5312785111214064},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4959378812525703e-05,"output_power":2.5019105880648296e-05,"reference_active":true,"delta_db":0.010380126013644094}],"active_band_count":3,"active_band_mae_db":0.41947559208835283}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.00011978792663441868,"log_rms_mae_db":1.1371320524968465,"rms_envelope_nrmse":0.10285784172538814,"envelope_correlation":0.9941364681724861,"rms_gain_error_db":0.15074389655105977,"reference_mean_power":0.0004050621320160613,"output_mean_power":0.00041936871986667624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":3.4505381724992573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016854041012359926,"output_power":0.00019706332686206593,"reference_active":true,"delta_db":0.6790176396904049},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.211674623356502e-05,"output_power":5.84712674853692e-05,"reference_active":true,"delta_db":0.4996521426895322},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0452544406168717e-06,"output_power":1.8513444398921012e-06,"reference_active":true,"delta_db":-0.43260118150568827}],"active_band_count":3,"active_band_mae_db":0.5370903212952084}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9875564575195312,"predicted_mos":2.9875564575195312,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76160,"input_sample_rate":16000,"input_seconds":4.76,"input_float32_sha256":"81b4972f3ae72084d96554d3df46bd75e4cbc643ff2fa85fd93bfcc3b4620bd2","internal_samples":76160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.520033204998981},"started_at":"2026-10-13T04:42:58.046110+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.539099998306483e-05,"native_input_samples":76160,"native_input_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","output_sha256":"07ad632e22757156dcc57c02a109ad3aaf61e5dc006c312bc419b4e9e2c927c6","output_samples":76160,"length_delta_samples":0,"analysis_s":2.096230065000782,"synthesis_s":0.07422574699921825,"total_s":2.170455812,"rtf":0.45597811176470593,"output_peak":0.5514247813968873,"output_rms":0.0335769159794188,"finished_at":"2026-10-13T04:43:00.225802+00:00","evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"07ad632e22757156dcc57c02a109ad3aaf61e5dc006c312bc419b4e9e2c927c6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.863714841682292,"spectral_convergence":0.7046990964898545,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.811430151886153,"spectral_convergence":0.5822701802776994,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.0903176188070764,"spectral_convergence":0.5868476727667499,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.00019994553788530193,"delta_db":-0.3758502214917645,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.00022947469056198796,"delta_db":0.7935774572977495,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":5.8097632829728395e-05,"delta_db":0.7966594618320922,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":5.563849592610635e-05,"delta_db":0.4902421019213887,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":0.00012890878443055722,"delta_db":0.1878024289795423,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.00045534414505526733,"delta_db":0.9240038838552684,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5946892592296342,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5261948292186558,"pitch":null,"output_rms_dbfs":-29.479183921332208,"output_peak":0.5514247813968873,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.255154204125174},"common_prefix":{"samples":76160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.863714841682292,"spectral_convergence":0.7046990964898545,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.811430151886153,"spectral_convergence":0.5822701802776994,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.0903176188070764,"spectral_convergence":0.5868476727667499,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.00019994553788530193,"delta_db":-0.3758502214917645,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.00022947469056198796,"delta_db":0.7935774572977495,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":5.8097632829728395e-05,"delta_db":0.7966594618320922,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":5.563849592610635e-05,"delta_db":0.4902421019213887,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":0.00012890878443055722,"delta_db":0.1878024289795423,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.00045534414505526733,"delta_db":0.9240038838552684,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5946892592296342,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.5261948292186558,"pitch":null,"output_rms_dbfs":-29.479183921332208,"output_peak":0.5514247813968873,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.255154204125174},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76160,"output_samples":76160,"intelligibility":{"estoi":{"value":0.80669471844344,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031603200210495876,"rms_dbfs":-30.00537875055086,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.32404640060221,"mean":-3.006077614151129e-06,"zero_energy":false},"output":{"rms":0.033576915979418795,"rms_dbfs":-29.479183921332208,"peak":0.5514247813968873,"peak_dbfs":-5.170274406168806,"crest_factor_db":24.308909515163396,"mean":-3.652485885490919e-09,"zero_energy":false},"rms_difference_db":0.5261948292186531},"full_scale":{"reference":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":476,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":1.7876799185095088e-06,"output_mean_power":1.1089686627155323e-06,"output_rms_dbfs":-59.550807259897134,"output_zero_energy":false,"output_peak":0.005754791708405218,"output_energy_fraction":0.0001157227486382597,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":469,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":36,"both_estimated_voiced_frames":21,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":5.149568270748566,"absolute_difference_cents_p95":22.103995929819686,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018341452353274373,"log_rms_mae_db":3.6430167811811924,"rms_envelope_nrmse":0.2723098372020689,"envelope_correlation":0.9458837409548719,"rms_gain_error_db":0.526235258609078,"reference_mean_power":0.0009998027766238483,"output_mean_power":0.0011285943308397147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":5.754282242130778e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018894414816227905,"output_power":0.00020927565613280537,"reference_active":true,"delta_db":0.44385266412027546},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013084274193993133,"output_power":0.00010312813111799487,"reference_active":true,"delta_db":-1.0337248905997316},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.690431882761885e-05,"output_power":1.6315009488579517e-05,"reference_active":true,"delta_db":-3.5448986291483453}],"active_band_count":3,"active_band_mae_db":1.6741587279561176}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.0001230241515816465,"log_rms_mae_db":3.1989250809380203,"rms_envelope_nrmse":0.18774306771385496,"envelope_correlation":0.9756424183285175,"rms_gain_error_db":0.526419746432556,"reference_mean_power":0.001006100998286567,"output_mean_power":0.0011357521158715917},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":5.061730772699165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019045538591950462,"output_power":0.00019911288365680334,"reference_active":true,"delta_db":0.1930610343038575},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.588390935994399e-05,"output_power":5.502695351823677e-05,"reference_active":true,"delta_db":-1.3957422718218793},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5269059669432984e-06,"output_power":1.0703573976991602e-06,"reference_active":true,"delta_db":-3.7306026593464674}],"active_band_count":3,"active_band_mae_db":1.773135321824068}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":2.3392725846803617,"rms_envelope_nrmse":0.19881894356329272,"envelope_correlation":0.9845032226562818,"rms_gain_error_db":0.9830211191801236,"reference_mean_power":0.0001499835837122031,"output_mean_power":0.0001880813954588761},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.245105068973029e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.560041827470393e-05,"output_power":8.323614258354233e-05,"reference_active":true,"delta_db":1.034053369876838},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.138220090606147e-05,"output_power":2.4579045932469108e-05,"reference_active":true,"delta_db":0.6051261533601229},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.449241698586229e-06,"output_power":1.3624325275109227e-06,"reference_active":true,"delta_db":-2.5471664105510747}],"active_band_count":3,"active_band_mae_db":1.395448644596012}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":1.6062236717628218,"rms_envelope_nrmse":0.15000835359771214,"envelope_correlation":0.9947417302583762,"rms_gain_error_db":0.9830112045714698,"reference_mean_power":0.00015093488949896852,"output_mean_power":0.00018927391338523686},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":1.1354747567630586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.429099822482352e-05,"output_power":8.038019329867344e-05,"reference_active":true,"delta_db":0.9699887729173385},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916365304226932e-05,"output_power":1.3007069663176475e-05,"reference_active":true,"delta_db":0.7610140581121929},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3162457402511593e-07,"output_power":1.3922587206269732e-07,"reference_active":true,"delta_db":-2.210646867870067}],"active_band_count":3,"active_band_mae_db":1.3138832329665329}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":1.6812241777151111,"rms_envelope_nrmse":0.21675461352340825,"envelope_correlation":0.9669385921139079,"rms_gain_error_db":0.28712770503453267,"reference_mean_power":0.0001020893317881212,"output_mean_power":0.00010906694076433407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":9.602503308503008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0812258101146943e-05,"output_power":3.117410753898991e-05,"reference_active":true,"delta_db":0.0507050212139886},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3389822486480089e-05,"output_power":1.0247704112736746e-05,"reference_active":true,"delta_db":-1.1614824214925896},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310273392307661e-06,"output_power":3.966668310734622e-06,"reference_active":true,"delta_db":-1.2669099405105895}],"active_band_count":3,"active_band_mae_db":0.8263657944057226}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":0.8782714191430534,"rms_envelope_nrmse":0.09720821909398865,"envelope_correlation":0.9935853124431615,"rms_gain_error_db":0.28713070626641446,"reference_mean_power":0.00010273559922291735,"output_mean_power":0.00010975745518101827},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":8.714815910547829e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.937520326415114e-05,"output_power":2.9628053014818608e-05,"reference_active":true,"delta_db":0.037222326562446145},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.005890574709939e-06,"output_power":5.428279920420776e-06,"reference_active":true,"delta_db":-1.1080111491794775},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.153574933783463e-07,"output_power":2.7809120271270315e-07,"reference_active":true,"delta_db":-2.679213447621737}],"active_band_count":3,"active_band_mae_db":1.2748156411212201}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018081886412701817,"log_rms_mae_db":1.689252449947162,"rms_envelope_nrmse":0.3824384891393759,"envelope_correlation":0.9306717894810874,"rms_gain_error_db":0.8645673487143879,"reference_mean_power":0.00040250829399997306,"output_mean_power":0.0004911697012273341},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":3.596628676311722e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013671994273759655,"output_power":0.00012784800739351377,"reference_active":true,"delta_db":-0.2913790420529097},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.299777768885513e-05,"output_power":5.325303865596665e-05,"reference_active":true,"delta_db":-2.4212817664027257},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4959378812525703e-05,"output_power":1.1610841227772507e-05,"reference_active":true,"delta_db":-3.323700861019352}],"active_band_count":3,"active_band_mae_db":2.0121205564916624}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.00011978792663441868,"log_rms_mae_db":0.9843335821252666,"rms_envelope_nrmse":0.2729609764826403,"envelope_correlation":0.9666358226151328,"rms_gain_error_db":0.8645703127248172,"reference_mean_power":0.0004050621320160613,"output_mean_power":0.0004942864162319723},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":3.4505381724992573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016854041012359926,"output_power":0.00013676630214386516,"reference_active":true,"delta_db":-0.9072494166685535},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.211674623356502e-05,"output_power":2.771163630153276e-05,"reference_active":true,"delta_db":-2.7431512337713357},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0452544406168717e-06,"output_power":7.539182827686643e-07,"reference_active":true,"delta_db":-4.3342306904081465}],"active_band_count":3,"active_band_mae_db":2.661543780282679}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"facodec","status":"measured","mos":{"status":"measured","value":3.494910955429077,"predicted_mos":3.494910955429077,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76200,"input_sample_rate":16000,"input_seconds":4.7625,"input_float32_sha256":"6943fa9503e9629abc5b0ad6b9da86e015673e0d176475845133fa8bd91c708b","internal_samples":76200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.972444331000588},"started_at":"2026-10-13T06:56:40.231270+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.346400030248333e-05,"native_input_samples":76160,"native_input_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","output_sha256":"4dd722d7506e72b995dc2573b5e3e658186d0b1c965c32f09021e64abe93d66f","output_samples":76200,"length_delta_samples":40,"analysis_s":1.3050765360003425,"synthesis_s":14.038789044998339,"total_s":15.343865580998681,"rtf":3.2235011724787146,"output_peak":0.3134331703186035,"output_rms":0.0285210291194274,"finished_at":"2026-10-13T06:56:55.585304+00:00","evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"4dd722d7506e72b995dc2573b5e3e658186d0b1c965c32f09021e64abe93d66f","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":76160,"output_samples":76200,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.896696159275116,"output_peak":0.3134331703186035},"common_prefix":{"samples":76160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9994750656167979,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.605132017747356,"spectral_convergence":0.4489332297781573,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.825433999813653,"spectral_convergence":0.5162765233920519,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.049849273594558,"spectral_convergence":0.5262527462006402,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.00021276588321824782,"delta_db":-0.10594748178572391,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.0001970719258833272,"delta_db":0.1324771357635139,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":4.916212281163109e-05,"delta_db":0.07138136005171937,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":4.1612799732918324e-05,"delta_db":-0.7712423465947317,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":9.533485674747931e-05,"delta_db":-1.1225055194133085,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.00021792842635749482,"delta_db":-2.2762548541791503,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7466347829646912,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8890387510566915,"pitch":null,"output_rms_dbfs":-30.894417501607553,"output_peak":0.3134331703186035,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.826805097051856},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76160,"output_samples":76160,"intelligibility":{"estoi":{"value":0.784296523553116,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031603200210495876,"rms_dbfs":-30.00537875055086,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.32404640060221,"mean":-3.006077614151129e-06,"zero_energy":false},"output":{"rms":0.028528512312265754,"rms_dbfs":-30.894417501607553,"peak":0.3134331703186035,"peak_dbfs":-10.0771008897011,"crest_factor_db":20.817316611906453,"mean":-0.00012760259015190517,"zero_energy":false},"rms_difference_db":-0.8890387510566917},"full_scale":{"reference":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":476,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":1.7876799185095088e-06,"output_mean_power":2.568947718622771e-06,"output_rms_dbfs":-55.90244734092813,"output_zero_energy":false,"output_peak":0.007548477966338396,"output_energy_fraction":0.0003713454357784622,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":469,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":23,"both_estimated_voiced_frames":18,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.75,"absolute_difference_cents_median":3.0531981376545856,"absolute_difference_cents_p95":11.476553056746571,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018341452353274373,"log_rms_mae_db":1.7017697940886292,"rms_envelope_nrmse":0.19673332975072175,"envelope_correlation":0.9714267031529201,"rms_gain_error_db":-0.8890150385534217,"reference_mean_power":0.0009998027766238483,"output_mean_power":0.0008147283612729829},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":5.754282242130778e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018894414816227905,"output_power":0.00015273494923653867,"reference_active":true,"delta_db":-0.9239502091435916},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013084274193993133,"output_power":7.555812845022772e-05,"reference_active":true,"delta_db":-2.384684445687819},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.690431882761885e-05,"output_power":2.0553066552334422e-05,"reference_active":true,"delta_db":-2.542005650996951}],"active_band_count":3,"active_band_mae_db":1.9502134352761207}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.0001230241515816465,"log_rms_mae_db":1.3552080950586094,"rms_envelope_nrmse":0.17625305051303955,"envelope_correlation":0.9771959813147203,"rms_gain_error_db":-0.8889462858101118,"reference_mean_power":0.001006100998286567,"output_mean_power":0.000819873692546113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":5.061730772699165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019045538591950462,"output_power":0.00014320084308607292,"reference_active":true,"delta_db":-1.2384768379527764},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.588390935994399e-05,"output_power":4.147598162337343e-05,"reference_active":true,"delta_db":-2.6235302323582443},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5269059669432984e-06,"output_power":1.2816068897817378e-06,"reference_active":true,"delta_db":-2.948342474723299}],"active_band_count":3,"active_band_mae_db":2.27011651501144}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":2.0103134932580233,"rms_envelope_nrmse":0.14442829210452607,"envelope_correlation":0.9863982645916413,"rms_gain_error_db":0.19916274646795393,"reference_mean_power":0.0001499835837122031,"output_mean_power":0.00015702181791396732},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.245105068973029e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.560041827470393e-05,"output_power":6.610785267559059e-05,"reference_active":true,"delta_db":0.03346442087951023},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.138220090606147e-05,"output_power":1.933390498176208e-05,"reference_active":true,"delta_db":-0.437328259559732},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.449241698586229e-06,"output_power":2.8147280107933574e-06,"reference_active":true,"delta_db":0.6040479029059463}],"active_band_count":3,"active_band_mae_db":0.3582801944483962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":1.5894513986106595,"rms_envelope_nrmse":0.12332255968260196,"envelope_correlation":0.989624655873574,"rms_gain_error_db":0.1991569253497279,"reference_mean_power":0.00015093488949896852,"output_mean_power":0.0001580175535383641},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":1.1354747567630586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.429099822482352e-05,"output_power":6.392311028498701e-05,"reference_active":true,"delta_db":-0.024922708203811074},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916365304226932e-05,"output_power":9.782599990838683e-06,"reference_active":true,"delta_db":-0.4762376478358621},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3162457402511593e-07,"output_power":2.2288381957012907e-07,"reference_active":true,"delta_db":-0.16706091959810593}],"active_band_count":3,"active_band_mae_db":0.222740425212593}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":2.228531560557146,"rms_envelope_nrmse":0.20334288516457039,"envelope_correlation":0.9729059495295589,"rms_gain_error_db":-0.8153414333007186,"reference_mean_power":0.0001020893317881212,"output_mean_power":8.461477771465822e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":9.602503308503008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0812258101146943e-05,"output_power":2.271694727206104e-05,"reference_active":true,"delta_db":-1.3237355710624163},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3389822486480089e-05,"output_power":1.2258963724646415e-05,"reference_active":true,"delta_db":-0.3832105952718986},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310273392307661e-06,"output_power":4.576046993185436e-06,"reference_active":true,"delta_db":-0.6462640498149068}],"active_band_count":3,"active_band_mae_db":0.784403405383074}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":1.8367304745594881,"rms_envelope_nrmse":0.15116678688465696,"envelope_correlation":0.9869868933734998,"rms_gain_error_db":-0.815346719986439,"reference_mean_power":0.00010273559922291735,"output_mean_power":8.515032038520376e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":8.714815910547829e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.937520326415114e-05,"output_power":2.210261559576048e-05,"reference_active":true,"delta_db":-1.2353720983716934},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.005890574709939e-06,"output_power":6.057753124320759e-06,"reference_active":true,"delta_db":-0.6315177950851153},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.153574933783463e-07,"output_power":4.1578645470263516e-07,"reference_active":true,"delta_db":-0.9323825815479253}],"active_band_count":3,"active_band_mae_db":0.933090825001578}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018081886412701817,"log_rms_mae_db":2.491076372520185,"rms_envelope_nrmse":0.27312990460070474,"envelope_correlation":0.9802315132758814,"rms_gain_error_db":-2.1257810784912854,"reference_mean_power":0.00040250829399997306,"output_mean_power":0.00024671566522989775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":3.596628676311722e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013671994273759655,"output_power":6.762164266892592e-05,"reference_active":true,"delta_db":-3.057461513521714},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.299777768885513e-05,"output_power":3.899295840134456e-05,"reference_active":true,"delta_db":-3.7748638418509257},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4959378812525703e-05,"output_power":1.1806617813418438e-05,"reference_active":true,"delta_db":-3.251082673298751}],"active_band_count":3,"active_band_mae_db":3.3611360095571303}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.00011978792663441868,"log_rms_mae_db":2.085690098028759,"rms_envelope_nrmse":0.24868872517869559,"envelope_correlation":0.9892418018160553,"rms_gain_error_db":-2.125787088750183,"reference_mean_power":0.0004050621320160613,"output_mean_power":0.0002482806852627853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":3.4505381724992573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016854041012359926,"output_power":7.890286817668832e-05,"reference_active":true,"delta_db":-3.296112559750284},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.211674623356502e-05,"output_power":2.1667689062193146e-05,"reference_active":true,"delta_db":-3.811646991949539},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0452544406168717e-06,"output_power":9.31580196457005e-07,"reference_active":true,"delta_db":-3.4152709654266684}],"active_band_count":3,"active_band_mae_db":3.5076768390421638}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.171398878097534,"predicted_mos":3.171398878097534,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":209920,"input_sample_rate":44100,"input_seconds":4.760090702947846,"input_float32_sha256":"255d7ce0b94c2bcad8a97146c219131c3a1cbba78b8f2196f414aebf79b2dc06","internal_samples":76162,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.20900822099793},"started_at":"2026-10-13T07:23:10.953374+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0056031030017038574,"native_input_samples":209916,"native_input_sha256":"118debe68d425f53e6fb77ec77c3667c62cc3f749b45acf3575ea3cd21d78e39","output_sha256":"183278bccc8e73a8362a2fdf00aa688706fd767789ec73894f84a4c0c9604649","output_samples":209920,"length_delta_samples":4,"analysis_s":0.10441315800017037,"synthesis_s":27.281170394000583,"total_s":27.385583552000753,"rtf":5.753273855462344,"output_peak":0.3264494836330414,"output_rms":0.025429075890113017,"finished_at":"2026-10-13T07:23:38.357217+00:00","evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"183278bccc8e73a8362a2fdf00aa688706fd767789ec73894f84a4c0c9604649","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":76160,"output_samples":76162,"length_delta_samples":2,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.89337974649007,"output_peak":0.322481025071276},"common_prefix":{"samples":76160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9999737401853943,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.141724747361739,"spectral_convergence":0.6574237541888579,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.046071782459432,"spectral_convergence":0.6627807926072878,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.934705212821376,"spectral_convergence":0.6826924482226506,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":8.04443983652978e-05,"delta_db":-4.330009307378729,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.0001992763187713116,"delta_db":0.1807864471650274,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":3.27843104322358e-05,"delta_db":-1.6882643799258465,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":3.169090385269809e-05,"delta_db":-1.9541654518512852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":0.00012343922162392633,"delta_db":-0.0004909574993967143,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.00017902101618551502,"delta_db":-3.130353502742886,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.8806783410938614,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.8878869975815624,"pitch":null,"output_rms_dbfs":-31.89326574813242,"output_peak":0.322481025071276,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.040833914214183},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76160,"output_samples":76160,"intelligibility":{"estoi":{"value":0.5217228712739348,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031603200210495876,"rms_dbfs":-30.00537875055086,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.32404640060221,"mean":-3.006077614151129e-06,"zero_energy":false},"output":{"rms":0.025429435094610042,"rms_dbfs":-31.89326574813242,"peak":0.322481025071276,"peak_dbfs":-9.829916687196162,"crest_factor_db":22.06334906093626,"mean":-0.0001289098244323125,"zero_energy":false},"rms_difference_db":-1.8878869975815604},"full_scale":{"reference":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":476,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":1.7876799185095088e-06,"output_mean_power":1.2455966139444096e-06,"output_rms_dbfs":-59.046225810334306,"output_zero_energy":false,"output_peak":0.004660680895093864,"output_energy_fraction":0.0002266131293317374,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":469,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":18,"both_estimated_voiced_frames":14,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.5833333333333334,"absolute_difference_cents_median":15.881446264662358,"absolute_difference_cents_p95":73.07492198788015,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018341452353274373,"log_rms_mae_db":2.603876818324701,"rms_envelope_nrmse":0.23252537738485124,"envelope_correlation":0.9828299562187558,"rms_gain_error_db":-1.887884934390169,"reference_mean_power":0.0009998027766238483,"output_mean_power":0.0006473301648064182},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":5.754282242130778e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018894414816227905,"output_power":0.0001299692515306308,"reference_active":true,"delta_db":-1.624928278939628},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013084274193993133,"output_power":9.384868850852871e-05,"reference_active":true,"delta_db":-1.4432142868380902},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.690431882761885e-05,"output_power":2.0007895096069476e-05,"reference_active":true,"delta_db":-2.658757919509614}],"active_band_count":3,"active_band_mae_db":1.9089668284291106}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.0001230241515816465,"log_rms_mae_db":2.4604219969734324,"rms_envelope_nrmse":0.21217504508662358,"envelope_correlation":0.992096462506911,"rms_gain_error_db":-1.887866798214199,"reference_mean_power":0.001006100998286567,"output_mean_power":0.0006514107182085865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":5.061730772699165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019045538591950462,"output_power":0.00013060742958682974,"reference_active":true,"delta_db":-1.638253762289616},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.588390935994399e-05,"output_power":5.23628996301458e-05,"reference_active":true,"delta_db":-1.6112600873150758},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5269059669432984e-06,"output_power":1.4853740868503023e-06,"reference_active":true,"delta_db":-2.307532377946217}],"active_band_count":3,"active_band_mae_db":1.8523487425169698}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":3.6087403467084034,"rms_envelope_nrmse":0.23179985319694874,"envelope_correlation":0.9624501294496745,"rms_gain_error_db":-0.15496264516504232,"reference_mean_power":0.0001499835837122031,"output_mean_power":0.00014472630087382374},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.245105068973029e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.560041827470393e-05,"output_power":5.3686906578596963e-05,"reference_active":true,"delta_db":-0.870382277057677},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.138220090606147e-05,"output_power":1.7577214157975904e-05,"reference_active":true,"delta_db":-0.8510236170411373},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.449241698586229e-06,"output_power":2.932430440021033e-06,"reference_active":true,"delta_db":0.7819607430661648}],"active_band_count":3,"active_band_mae_db":0.8344555457216597}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":2.8879391042184586,"rms_envelope_nrmse":0.18467724650640344,"envelope_correlation":0.974893457511632,"rms_gain_error_db":-0.1549996952374748,"reference_mean_power":0.00015093488949896852,"output_mean_power":0.00014564301862096308},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":1.1354747567630586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.429099822482352e-05,"output_power":5.311338376131888e-05,"reference_active":true,"delta_db":-0.8294619851223313},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916365304226932e-05,"output_power":8.870913743256875e-06,"reference_active":true,"delta_db":-0.9010970404725214},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3162457402511593e-07,"output_power":2.5870156431776813e-07,"reference_active":true,"delta_db":0.48014421190572265}],"active_band_count":3,"active_band_mae_db":0.7369010791668584}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":2.8211549949491133,"rms_envelope_nrmse":0.28201699011727316,"envelope_correlation":0.9491532652314619,"rms_gain_error_db":-1.308809384287043,"reference_mean_power":0.0001020893317881212,"output_mean_power":7.552651102734058e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":9.602503308503008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0812258101146943e-05,"output_power":2.52838492090336e-05,"reference_active":true,"delta_db":-0.8588033554011107},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3389822486480089e-05,"output_power":1.1216438097671102e-05,"reference_active":true,"delta_db":-0.7691985559633558},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310273392307661e-06,"output_power":3.930837178027884e-06,"reference_active":true,"delta_db":-1.306318257509329}],"active_band_count":3,"active_band_mae_db":0.9781067229579318}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":2.334444671495328,"rms_envelope_nrmse":0.24018664607507406,"envelope_correlation":0.9652351971676871,"rms_gain_error_db":-1.3088664591222954,"reference_mean_power":0.00010273559922291735,"output_mean_power":7.600362604420692e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":8.714815910547829e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.937520326415114e-05,"output_power":2.461032606178871e-05,"reference_active":true,"delta_db":-0.7686351272573895},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.005890574709939e-06,"output_power":5.813630465174715e-06,"reference_active":true,"delta_db":-0.8101592674639397},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.153574933783463e-07,"output_power":3.8993191501057723e-07,"reference_active":true,"delta_db":-1.211198125123142}],"active_band_count":3,"active_band_mae_db":0.9299975066148237}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018081886412701817,"log_rms_mae_db":3.2354208813840275,"rms_envelope_nrmse":0.2659101858932058,"envelope_correlation":0.9883713587602739,"rms_gain_error_db":-2.3196288710690136,"reference_mean_power":0.00040250829399997306,"output_mean_power":0.00023594563465978414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":3.596628676311722e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013671994273759655,"output_power":8.967564880846209e-05,"reference_active":true,"delta_db":-1.8315734031233828},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.299777768885513e-05,"output_power":5.8381685955399696e-05,"reference_active":true,"delta_db":-2.0219593815108343},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4959378812525703e-05,"output_power":1.1710077595929915e-05,"reference_active":true,"delta_db":-3.2867399954978413}],"active_band_count":3,"active_band_mae_db":2.380090926710686}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.00011978792663441868,"log_rms_mae_db":2.7180670492431562,"rms_envelope_nrmse":0.24605808530827342,"envelope_correlation":0.9958000543069868,"rms_gain_error_db":-2.319627512431643,"reference_mean_power":0.0004050621320160613,"output_mean_power":0.00023744273879216615},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":3.4505381724992573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016854041012359926,"output_power":0.00010834276210270901,"reference_active":true,"delta_db":-1.9190414304663488},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.211674623356502e-05,"output_power":3.165661403127974e-05,"reference_active":true,"delta_db":-2.165128327151148},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0452544406168717e-06,"output_power":9.441916692156113e-07,"reference_active":true,"delta_db":-3.3568718001759104}],"active_band_count":3,"active_band_mae_db":2.480347185931136}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"golf","status":"measured","mos":{"status":"measured","value":2.419565439224243,"predicted_mos":2.419565439224243,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":114000,"input_sample_rate":24000,"input_seconds":4.75,"input_float32_sha256":"015061a012525d8d4f4214d76849651b80f2f7fae896279929531dac370fb052","internal_samples":76000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.9864719029992557},"started_at":"2026-10-13T05:02:49.932245+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002310892000423337,"native_input_samples":114240,"native_input_sha256":"5f72cac752e0b1f3dff668f140c621a89002c50a22e2b517de7ff74764f8c926","output_sha256":"d4e7feebc5ae9275605b7e391603cec0540a832ee4a44b6a2fd69fff3f86610e","output_samples":114000,"length_delta_samples":-240,"analysis_s":0.8935766040003728,"synthesis_s":0.10493189499993605,"total_s":0.9985084990003088,"rtf":0.20977069306729176,"output_peak":0.3533763289451599,"output_rms":0.02891075750414108,"finished_at":"2026-10-13T05:02:50.940372+00:00","analysis_stages":{"external_f0_s":0.04568006800036528},"evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"d4e7feebc5ae9275605b7e391603cec0540a832ee4a44b6a2fd69fff3f86610e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":76160,"output_samples":76000,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.782372081385112,"output_peak":0.3455876976963917},"common_prefix":{"samples":76000,"sample_rate":16000,"reference_coverage":0.9978991596638656,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76000,"output_samples":76000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.03503940919115,"spectral_convergence":0.6555962722296791,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.010205258015197,"spectral_convergence":0.48574334790637136,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.106105981369478,"spectral_convergence":0.4802320304544903,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021847703008738965,"output_power":0.00019052930477922187,"delta_db":-0.5944400073196368,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0001915547492476723,"output_power":0.0002018349710109914,"delta_db":0.22703492629382846,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.846195231290236e-05,"output_power":4.4426169806618674e-05,"delta_db":-0.37762033672028794,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.980371698911952e-05,"output_power":3.7410714786230445e-05,"delta_db":-1.2426575046090491,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00012371261561332144,"output_power":6.521543573145221e-05,"delta_db":-2.78063588808168,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003688522737538365,"output_power":0.0002957301471149819,"delta_db":-0.9595686564262672,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0303262199084582,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7861155539154562,"pitch":null,"output_rms_dbfs":-30.782372081385112,"output_peak":0.3455876976963917,"reference_rms_dbfs":-29.996256527469658,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.383783549525275},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76000,"output_samples":76000,"intelligibility":{"estoi":{"value":0.7935055224322773,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03163640842453899,"rms_dbfs":-29.996256527469658,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.314924177521007,"mean":-1.79290771484375e-06,"zero_energy":false},"output":{"rms":0.02889890557148309,"rms_dbfs":-30.782372081385112,"peak":0.3455876976963917,"peak_dbfs":-9.228834520487199,"crest_factor_db":21.55353756089792,"mean":0.00017205778570434318,"zero_energy":false},"rms_difference_db":-0.7861155539154545},"full_scale":{"reference":{"sample_count":76000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":475,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.7978760032829912e-06,"output_mean_power":2.4277684166238945e-06,"output_rms_dbfs":-56.14792742711732,"output_zero_energy":false,"output_peak":0.0054398706017354075,"output_energy_fraction":0.0003365995610555611,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":468,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":31,"both_estimated_voiced_frames":18,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.75,"absolute_difference_cents_median":9.05300257194477,"absolute_difference_cents_p95":23.874880664876653,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":0.00018341452353274373,"log_rms_mae_db":2.9153882239605626,"rms_envelope_nrmse":0.2309127874486526,"envelope_correlation":0.9562868765819087,"rms_gain_error_db":-0.7860814919317467,"reference_mean_power":0.001001907428276349,"output_mean_power":0.0008360253519358909},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":5.760814830601667e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018462389516850198,"output_power":0.0001543326131384461,"reference_active":true,"delta_db":-0.7783019974764722},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013360643113818836,"output_power":0.00010519989753748773,"reference_active":true,"delta_db":-1.0381204655944936},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.76578664931019e-05,"output_power":3.01235061865787e-05,"reference_active":true,"delta_db":-0.969501918025591}],"active_band_count":3,"active_band_mae_db":0.9286414603655189}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":943,"rms_floor":0.0001230241515816465,"log_rms_mae_db":1.900799143242051,"rms_envelope_nrmse":0.15509122655844843,"envelope_correlation":0.9828738846401464,"rms_gain_error_db":-0.7861961759998254,"reference_mean_power":0.001008232483146937,"output_mean_power":0.0008412809750751426},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":943,"frequency_resolution_hz":0.21208907741251326,"power_floor":5.067688870152086e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001893872625630233,"output_power":0.00015619223777879653,"reference_active":true,"delta_db":-0.8369131959929311},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.587539815397842e-05,"output_power":5.90160544622209e-05,"reference_active":true,"delta_db":-1.0913081180345852},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.529966052867134e-06,"output_power":2.1852804997649243e-06,"reference_active":true,"delta_db":-0.6360750348351607}],"active_band_count":3,"active_band_mae_db":0.8547654496208924}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":1.8816585136349049,"rms_envelope_nrmse":0.1641666106855826,"envelope_correlation":0.9814679549298884,"rms_gain_error_db":-0.10448312199609505,"reference_mean_power":0.00015029911528718248,"output_mean_power":0.0001467263492990779},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":1.2408199472363458e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.437918401027878e-05,"output_power":6.026831773280549e-05,"reference_active":true,"delta_db":-0.28656398620580675},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2200464133462377e-05,"output_power":2.1079311684924834e-05,"reference_active":true,"delta_db":-0.225056286118681},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4810265070122936e-06,"output_power":2.9766452812959735e-06,"reference_active":true,"delta_db":0.7909567978948473}],"active_band_count":3,"active_band_mae_db":0.43419235673977835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":943,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":1.1086408600226325,"rms_envelope_nrmse":0.09935733927276451,"envelope_correlation":0.9929467263755569,"rms_gain_error_db":-0.10447712704448664,"reference_mean_power":0.0001512546810287846,"output_mean_power":0.0001476594040782993},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":943,"frequency_resolution_hz":0.21208907741251326,"power_floor":1.1312849500294457e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.402262786707175e-05,"output_power":6.03971836099295e-05,"reference_active":true,"delta_db":-0.2531680874394203},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0839278625025134e-05,"output_power":1.0260285524411396e-05,"reference_active":true,"delta_db":-0.23840933484643784},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3052057419147192e-07,"output_power":2.950669737717571e-07,"reference_active":true,"delta_db":1.0721090987682684}],"active_band_count":3,"active_band_mae_db":0.5212288403513755}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":1.9641063787157016,"rms_envelope_nrmse":0.25530894724322845,"envelope_correlation":0.966925563147868,"rms_gain_error_db":-1.5976038780021518,"reference_mean_power":0.00010230396344704328,"output_mean_power":7.081611077837482e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":9.572214881010142e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0487889051245066e-05,"output_power":1.8977998770882444e-05,"reference_active":true,"delta_db":-2.058769411498645},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3420596781637454e-05,"output_power":9.498510048759355e-06,"reference_active":true,"delta_db":-1.501163417775571},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.3053809269342525e-06,"output_power":3.916853268865951e-06,"reference_active":true,"delta_db":-1.3177926921141172}],"active_band_count":3,"active_band_mae_db":1.6259085071294443}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":943,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":1.557808946118161,"rms_envelope_nrmse":0.2007232223284599,"envelope_correlation":0.9884020637374136,"rms_gain_error_db":-1.5976265349560141,"reference_mean_power":0.00010295298978675851,"output_mean_power":7.126500331685336e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":943,"frequency_resolution_hz":0.21208907741251326,"power_floor":8.686542795585821e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9128301303699414e-05,"output_power":1.8254602005005e-05,"reference_active":true,"delta_db":-2.029427895929673},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.984088290321644e-06,"output_power":5.079499662974879e-06,"reference_active":true,"delta_db":-1.3828878508137168},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.134828259238867e-07,"output_power":3.230805278978708e-07,"reference_active":true,"delta_db":-2.012151388274847}],"active_band_count":3,"active_band_mae_db":1.8081557116727458}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":0.00018081886412701817,"log_rms_mae_db":3.5192739331376353,"rms_envelope_nrmse":0.3060752273624878,"envelope_correlation":0.9508901762520784,"rms_gain_error_db":-1.241535380752967,"reference_mean_power":0.0004033561027738837,"output_mean_power":0.00030306451858706145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":3.615847505291845e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013490099609723453,"output_power":9.379166943846168e-05,"reference_active":true,"delta_db":-1.5785089035298592},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.497167879490613e-05,"output_power":6.503952235044178e-05,"reference_active":true,"delta_db":-1.6441677186259054},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5686293651866598e-05,"output_power":1.9848626652563357e-05,"reference_active":true,"delta_db":-1.1197098036008641}],"active_band_count":3,"active_band_mae_db":1.4474621419188762}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76000,"uncovered_tail_samples":0,"metrics":{"frame_count":943,"rms_floor":0.00011978792663441868,"log_rms_mae_db":3.22559137088866,"rms_envelope_nrmse":0.20674828489690503,"envelope_correlation":0.9835581297049089,"rms_gain_error_db":-1.2415259179040052,"reference_mean_power":0.0004059207470853323,"output_mean_power":0.0003049921471450187},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":943,"frequency_resolution_hz":0.21208907741251326,"power_floor":3.4691037853330727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016913830544530855,"output_power":0.00011886365649975829,"reference_active":true,"delta_db":-1.53192889235609},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.214706191685512e-05,"output_power":3.6513170741101996e-05,"reference_active":true,"delta_db":-1.5478029634569521},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0527126968089196e-06,"output_power":1.7344699495352643e-06,"reference_active":true,"delta_db":-0.731613887021795}],"active_band_count":3,"active_band_mae_db":1.2704485809449457}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.849255084991455,"predicted_mos":2.849255084991455,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":114360,"input_sample_rate":24000,"input_seconds":4.765,"input_float32_sha256":"53a23e275b87ca2e7493f256a72435612297babadba230573f9e31fbe0e568c9","internal_samples":76240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.79535437800223},"started_at":"2026-10-13T07:42:08.453156+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002507994999177754,"native_input_samples":114240,"native_input_sha256":"5f72cac752e0b1f3dff668f140c621a89002c50a22e2b517de7ff74764f8c926","output_sha256":"0c7e47f805cdec93e3124eb4e5af7c50ab7b7f5965f77b1a4637ddaeedb15aaa","output_samples":114360,"length_delta_samples":120,"analysis_s":2.044255893997615,"synthesis_s":11.558093010000448,"total_s":13.602348903998063,"rtf":2.857636324369341,"output_peak":0.32740575075149536,"output_rms":0.029893636814192565,"finished_at":"2026-10-13T07:42:22.070618+00:00","analysis_stages":{"external_f0_s":1.3483223310031462},"evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"0c7e47f805cdec93e3124eb4e5af7c50ab7b7f5965f77b1a4637ddaeedb15aaa","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":76160,"output_samples":76240,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.490856444460782,"output_peak":0.3073823836207496},"common_prefix":{"samples":76160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9989506820566632,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.86308440606845,"spectral_convergence":0.6951132761965177,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.703506456748073,"spectral_convergence":0.5460462723897834,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.8435375413377635,"spectral_convergence":0.5359451198503083,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.0003103185371483529,"delta_db":1.5331098058988133,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.00018471607708753462,"delta_db":-0.14872350503308102,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":4.334182156300082e-05,"delta_db":-0.47585331546884996,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":4.920509536669454e-05,"delta_db":-0.043410931458815326,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":9.801833868501994e-05,"delta_db":-1.0019493289178851,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.00020846714482743444,"delta_db":-2.4690174995109344,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9453440643813965,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4809205244194523,"pitch":null,"output_rms_dbfs":-30.486299274970314,"output_peak":0.3073823836207496,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.1367094680514285},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76160,"output_samples":76160,"intelligibility":{"estoi":{"value":0.784002473119107,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031603200210495876,"rms_dbfs":-30.00537875055086,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.32404640060221,"mean":-3.006077614151129e-06,"zero_energy":false},"output":{"rms":0.029900953407509222,"rms_dbfs":-30.486299274970314,"peak":0.3073823836207496,"peak_dbfs":-10.24642051912819,"crest_factor_db":20.23987875584212,"mean":0.0002991840775333445,"zero_energy":false},"rms_difference_db":-0.480920524419453},"full_scale":{"reference":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":476,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":1.7876799185095088e-06,"output_mean_power":8.876076691279535e-07,"output_rms_dbfs":-60.51778954022766,"output_zero_energy":false,"output_peak":0.004491368347529698,"output_energy_fraction":0.0001167970968034387,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":469,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":50,"both_estimated_voiced_frames":21,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":16.1210474548507,"absolute_difference_cents_p95":29.86275770579018,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018341452353274373,"log_rms_mae_db":4.5304508020500895,"rms_envelope_nrmse":0.2754070074338273,"envelope_correlation":0.9318175090232679,"rms_gain_error_db":-0.48088225778814303,"reference_mean_power":0.0009998027766238483,"output_mean_power":0.0008950063420400953},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":5.754282242130778e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018894414816227905,"output_power":0.0001832736076682984,"reference_active":true,"delta_db":-0.13233516964932682},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013084274193993133,"output_power":6.700946338574906e-05,"reference_active":true,"delta_db":-2.9061349651794854},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.690431882761885e-05,"output_power":1.2588804046957604e-05,"reference_active":true,"delta_db":-4.67092720101257}],"active_band_count":3,"active_band_mae_db":2.5697991119471273}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.0001230241515816465,"log_rms_mae_db":4.163419210743182,"rms_envelope_nrmse":0.21472929271091892,"envelope_correlation":0.9564571158333443,"rms_gain_error_db":-0.4806938293600359,"reference_mean_power":0.001006100998286567,"output_mean_power":0.000900683479659724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":5.061730772699165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019045538591950462,"output_power":0.00016667683312387643,"reference_active":true,"delta_db":-0.5791801843042195},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.588390935994399e-05,"output_power":3.5903365540124216e-05,"reference_active":true,"delta_db":-3.250145358689328},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5269059669432984e-06,"output_power":7.593128447962184e-07,"reference_active":true,"delta_db":-5.221683343252119}],"active_band_count":3,"active_band_mae_db":3.017002962081889}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":2.27973250480166,"rms_envelope_nrmse":0.19648490056779663,"envelope_correlation":0.9733278216635552,"rms_gain_error_db":-0.25859088194158214,"reference_mean_power":0.0001499835837122031,"output_mean_power":0.00014131382059351798},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.245105068973029e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.560041827470393e-05,"output_power":6.059687847500997e-05,"reference_active":true,"delta_db":-0.34456355543469314},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.138220090606147e-05,"output_power":1.8482802544989988e-05,"reference_active":true,"delta_db":-0.6328458194428024},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.449241698586229e-06,"output_power":1.9016026186597517e-06,"reference_active":true,"delta_db":-1.0991187795489923}],"active_band_count":3,"active_band_mae_db":0.6921760514754959}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":1.427155645946798,"rms_envelope_nrmse":0.12578910049586395,"envelope_correlation":0.9888773867963703,"rms_gain_error_db":-0.25859292598437883,"reference_mean_power":0.00015093488949896852,"output_mean_power":0.00014221006945755094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":1.1354747567630586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.429099822482352e-05,"output_power":5.929265585333018e-05,"reference_active":true,"delta_db":-0.35149265137661223},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916365304226932e-05,"output_power":9.220782661367215e-06,"reference_active":true,"delta_db":-0.7331027483424861},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3162457402511593e-07,"output_power":1.7373439013604786e-07,"reference_active":true,"delta_db":-1.2489883954440624}],"active_band_count":3,"active_band_mae_db":0.777861265054387}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":2.0080917845591437,"rms_envelope_nrmse":0.28057451363471664,"envelope_correlation":0.9404526121453951,"rms_gain_error_db":-0.217821657052355,"reference_mean_power":0.0001020893317881212,"output_mean_power":9.709529722051802e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":9.602503308503008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0812258101146943e-05,"output_power":2.9056978977840984e-05,"reference_active":true,"delta_db":-0.25473067914800995},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3389822486480089e-05,"output_power":1.1908336463146685e-05,"reference_active":true,"delta_db":-0.5092372257064289},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310273392307661e-06,"output_power":4.307702392630722e-06,"reference_active":true,"delta_db":-0.9087118928117642}],"active_band_count":3,"active_band_mae_db":0.557559932555401}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":1.1064768202300337,"rms_envelope_nrmse":0.1449354606747775,"envelope_correlation":0.9833071530289373,"rms_gain_error_db":-0.21783040763655953,"reference_mean_power":0.00010273559922291735,"output_mean_power":9.77097534884301e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":8.714815910547829e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.937520326415114e-05,"output_power":2.7374893078782838e-05,"reference_active":true,"delta_db":-0.3062844884842305},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.005890574709939e-06,"output_power":6.675101561959115e-06,"reference_active":true,"delta_db":-0.21005471875238627},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.153574933783463e-07,"output_power":4.137844367098296e-07,"reference_active":true,"delta_db":-0.9533444321601987}],"active_band_count":3,"active_band_mae_db":0.4898945464656051}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018081886412701817,"log_rms_mae_db":1.9974962082626178,"rms_envelope_nrmse":0.35988130688448766,"envelope_correlation":0.9478508323006206,"rms_gain_error_db":-2.4383665494966027,"reference_mean_power":0.00040250829399997306,"output_mean_power":0.0002295821818026508},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":3.596628676311722e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013671994273759655,"output_power":6.427953053434245e-05,"reference_active":true,"delta_db":-3.277591715847699},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.299777768885513e-05,"output_power":2.517422199778383e-05,"reference_active":true,"delta_db":-5.67516512922886},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4959378812525703e-05,"output_power":3.902394167394824e-06,"reference_active":true,"delta_db":-8.059026385867728}],"active_band_count":3,"active_band_mae_db":5.670594410314763}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.00011978792663441868,"log_rms_mae_db":1.4565291602206165,"rms_envelope_nrmse":0.3050612089627524,"envelope_correlation":0.9747451739969554,"rms_gain_error_db":-2.4383689625016833,"reference_mean_power":0.0004050621320160613,"output_mean_power":0.0002310387083960612},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":3.4505381724992573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016854041012359926,"output_power":6.835834831999799e-05,"reference_active":true,"delta_db":-3.91912485675227},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.211674623356502e-05,"output_power":1.2700793132707126e-05,"reference_active":true,"delta_db":-6.131464514943689},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0452544406168717e-06,"output_power":3.970917186285198e-07,"reference_active":true,"delta_db":-7.118565143451318}],"active_band_count":3,"active_band_mae_db":5.723051505049092}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0777153968811035,"predicted_mos":3.0777153968811035,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":114480,"input_sample_rate":24000,"input_seconds":4.77,"input_float32_sha256":"a792367b00fdf80b6449b9524ea2cde56cf53a20dae9917a36dd593915dbaced","internal_samples":76320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.333854397000323},"started_at":"2026-10-13T05:11:08.853972+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005913969000175712,"native_input_samples":114240,"native_input_sha256":"5f72cac752e0b1f3dff668f140c621a89002c50a22e2b517de7ff74764f8c926","output_sha256":"28a9ee5aa257d78613f588c8d50a6ffeb549e4198ef65439b62f24bed5b61b3e","output_samples":114480,"length_delta_samples":240,"analysis_s":0.8668303289996402,"synthesis_s":0.1601293359999545,"total_s":1.0269596649995947,"rtf":0.21574782878142745,"output_peak":0.3904895484447479,"output_rms":0.02932825304288575,"finished_at":"2026-10-13T05:11:09.895814+00:00","analysis_stages":{"external_f0_s":0.09601241100062907},"evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"28a9ee5aa257d78613f588c8d50a6ffeb549e4198ef65439b62f24bed5b61b3e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":76160,"output_samples":76320,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.655705652331648,"output_peak":0.3747410279732098},"common_prefix":{"samples":76160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9979035639412998,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.191582205940492,"spectral_convergence":0.3982136363927156,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.7869925592592395,"spectral_convergence":0.40464670567470623,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4515567449029625,"spectral_convergence":0.4551903444903863,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.00020079883571328929,"delta_db":-0.35735548406885476,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.00017780612649077936,"delta_db":-0.31430326220356575,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":4.398892752445687e-05,"delta_db":-0.4114911761702636,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":4.062694501179382e-05,"delta_db":-0.8753700471727586,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":0.00010290320928614974,"delta_db":-0.7907335076369771,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.0002955455571418068,"delta_db":-0.9531493318017437,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6170671348423605,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.641213531990029,"pitch":null,"output_rms_dbfs":-30.646592282540887,"output_peak":0.3747410279732098,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.143377170034231},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76160,"output_samples":76160,"intelligibility":{"estoi":{"value":0.8644906271100847,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031603200210495876,"rms_dbfs":-30.00537875055086,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.32404640060221,"mean":-3.006077614151129e-06,"zero_energy":false},"output":{"rms":0.029354209258099184,"rms_dbfs":-30.646592282540887,"peak":0.3747410279732098,"peak_dbfs":-8.525375124143146,"crest_factor_db":22.121217158397744,"mean":0.00016424720004791016,"zero_energy":false},"rms_difference_db":-0.6412135319900258},"full_scale":{"reference":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":476,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":1.7876799185095088e-06,"output_mean_power":1.614147713476232e-06,"output_rms_dbfs":-57.92056724748862,"output_zero_energy":false,"output_peak":0.004977761287733752,"output_energy_fraction":0.00022038578445814078,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":469,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":29,"both_estimated_voiced_frames":24,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":2.336775856240725,"absolute_difference_cents_p95":7.605977796965925,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018341452353274373,"log_rms_mae_db":1.583560506562998,"rms_envelope_nrmse":0.14724393634220426,"envelope_correlation":0.9842032970083089,"rms_gain_error_db":-0.6412289367050032,"reference_mean_power":0.0009998027766238483,"output_mean_power":0.0008625642311830497},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":5.754282242130778e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018894414816227905,"output_power":0.00016451901002603788,"reference_active":true,"delta_db":-0.6011735828683574},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013084274193993133,"output_power":8.62448361898654e-05,"reference_active":true,"delta_db":-1.8101653497925718},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.690431882761885e-05,"output_power":1.5531898711422797e-05,"reference_active":true,"delta_db":-3.758526439110661}],"active_band_count":3,"active_band_mae_db":2.05662179059053}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.0001230241515816465,"log_rms_mae_db":1.0631318992918972,"rms_envelope_nrmse":0.10544575148815455,"envelope_correlation":0.993772416102455,"rms_gain_error_db":-0.6411623584899568,"reference_mean_power":0.001006100998286567,"output_mean_power":0.0008680112302447692},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":5.061730772699165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019045538591950462,"output_power":0.000155324298833165,"reference_active":true,"delta_db":-0.8855385686428346},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.588390935994399e-05,"output_power":4.8831967367433405e-05,"reference_active":true,"delta_db":-1.9144547490365236},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5269059669432984e-06,"output_power":9.514639928213326e-07,"reference_active":true,"delta_db":-4.2419672343459816}],"active_band_count":3,"active_band_mae_db":2.3473201840084466}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":1.4010801249291094,"rms_envelope_nrmse":0.12884538276723104,"envelope_correlation":0.9894717766674834,"rms_gain_error_db":-0.400698368236435,"reference_mean_power":0.0001499835837122031,"output_mean_power":0.00013676465983017484},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.245105068973029e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.560041827470393e-05,"output_power":6.314577304094404e-05,"reference_active":true,"delta_db":-0.16562324166396106},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.138220090606147e-05,"output_power":1.8788516366605053e-05,"reference_active":true,"delta_db":-0.5615991840597063},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.449241698586229e-06,"output_power":2.0212298233971455e-06,"reference_active":true,"delta_db":-0.8341594707285072}],"active_band_count":3,"active_band_mae_db":0.5204606321507249}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":0.7070991342179019,"rms_envelope_nrmse":0.07787878075812002,"envelope_correlation":0.9969167403951178,"rms_gain_error_db":-0.4007004878282222,"reference_mean_power":0.00015093488949896852,"output_mean_power":0.00013763205434375844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":1.1354747567630586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.429099822482352e-05,"output_power":6.190223404109189e-05,"reference_active":true,"delta_db":-0.16443846042089466},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916365304226932e-05,"output_power":9.781008574353316e-06,"reference_active":true,"delta_db":-0.47694420807726784},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3162457402511593e-07,"output_power":1.566419225364045e-07,"reference_active":true,"delta_db":-1.6987662884543877}],"active_band_count":3,"active_band_mae_db":0.7800496523175168}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":1.2526035101070896,"rms_envelope_nrmse":0.1623896974359946,"envelope_correlation":0.9841639237531687,"rms_gain_error_db":-0.7479480833040707,"reference_mean_power":0.0001020893317881212,"output_mean_power":8.593806145576485e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":9.602503308503008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0812258101146943e-05,"output_power":2.5781244370665694e-05,"reference_active":true,"delta_db":-0.7741965170224587},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3389822486480089e-05,"output_power":9.300813635795034e-06,"reference_active":true,"delta_db":-1.5825387713167602},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310273392307661e-06,"output_power":2.4908474118622716e-06,"reference_active":true,"delta_db":-3.2876975705009364}],"active_band_count":3,"active_band_mae_db":1.8814776196133849}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":0.8166783375608684,"rms_envelope_nrmse":0.10773368692276283,"envelope_correlation":0.9958252231179987,"rms_gain_error_db":-0.7479523579136298,"reference_mean_power":0.00010273559922291735,"output_mean_power":8.6481999588991e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":8.714815910547829e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.937520326415114e-05,"output_power":2.4456737899721046e-05,"reference_active":true,"delta_db":-0.7958235113736842},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.005890574709939e-06,"output_power":4.897702603379529e-06,"reference_active":true,"delta_db":-1.5547093926296982},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.153574933783463e-07,"output_power":2.2130477124671266e-07,"reference_active":true,"delta_db":-3.671178178853709}],"active_band_count":3,"active_band_mae_db":2.0072370276190306}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018081886412701817,"log_rms_mae_db":1.2117510711414476,"rms_envelope_nrmse":0.19482408091876138,"envelope_correlation":0.9826912727055291,"rms_gain_error_db":-0.9405579350475141,"reference_mean_power":0.00040250829399997306,"output_mean_power":0.0003241298589828962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":3.596628676311722e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013671994273759655,"output_power":9.872699212181187e-05,"reference_active":true,"delta_db":-1.4139596208846528},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.299777768885513e-05,"output_power":5.0590727333959094e-05,"reference_active":true,"delta_db":-2.6440164737913427},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4959378812525703e-05,"output_power":9.238158482871832e-06,"reference_active":true,"delta_db":-4.316483640480538}],"active_band_count":3,"active_band_mae_db":2.791486578385511}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.00011978792663441868,"log_rms_mae_db":0.8648392053376796,"rms_envelope_nrmse":0.14358789375503805,"envelope_correlation":0.9934648447354782,"rms_gain_error_db":-0.9405552242399722,"reference_mean_power":0.0004050621320160613,"output_mean_power":0.00032618660444513165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":3.4505381724992573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016854041012359926,"output_power":0.00011217064553132503,"reference_active":true,"delta_db":-1.7682482718767591},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.211674623356502e-05,"output_power":2.6818228217589896e-05,"reference_active":true,"delta_db":-2.8854721168659827},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0452544406168717e-06,"output_power":6.888254825971422e-07,"reference_active":true,"delta_db":-4.726381391152271}],"active_band_count":3,"active_band_mae_db":3.126700593298338}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0003","method":"magphase","status":"measured","mos":{"status":"measured","value":3.0951569080352783,"predicted_mos":3.0951569080352783,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":229965,"input_sample_rate":48000,"input_seconds":4.7909375,"input_float32_sha256":"22cfc041f0f0c258d2f51d9fa242eb57419c3fa277db194f59d2891d5f397849","internal_samples":76655,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.963049010000759},"started_at":"2026-10-13T05:17:09.840875+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00824807200115174,"native_input_samples":228480,"native_input_sha256":"2a7ba5874c5bbaba69e70c7e6ea878568e712a08851e709de77e825bf98fba29","output_sha256":"9ea280bfa7140333c0d416edabc84482821538242583ee98fc30498cad23f0a0","output_samples":229965,"length_delta_samples":1485,"analysis_s":2.741736101999777,"synthesis_s":0.9114927089995035,"total_s":3.6532288109992805,"rtf":0.76748504432758,"output_peak":0.5850501844508809,"output_rms":0.032638457358158735,"finished_at":"2026-10-13T05:17:13.513679+00:00","evaluation_input":{"reference":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"9ea280bfa7140333c0d416edabc84482821538242583ee98fc30498cad23f0a0","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":76160,"output_samples":76655,"length_delta_samples":495,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.725876626947738,"output_peak":0.5484497141252593},"common_prefix":{"samples":76160,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9935424955971561,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76160,"output_samples":76160,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.719710741161658,"spectral_convergence":0.7664135539316107,"spectral_convergence_status":"ok","amplitude_floor":1.3356779608968942e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.077096942233931,"spectral_convergence":0.6772263568036289,"spectral_convergence_status":"ok","amplitude_floor":9.962006407647762e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.712190810975431,"spectral_convergence":0.6029440074259614,"spectral_convergence_status":"ok","amplitude_floor":4.291398708237255e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021802020252140593,"output_power":0.00017901420107124653,"delta_db":-0.8560925410496222,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00019115120810434436,"output_power":0.00017667321062180523,"delta_db":-0.342063457740103,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":4.836069030265609e-05,"output_power":4.644141764642224e-05,"delta_db":-0.17587023056774034,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":4.9699402895957116e-05,"output_power":4.595939241091718e-05,"delta_db":-0.339768923908019,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001234531768639194,"output_power":0.0001335750837332189,"delta_db":0.3422318483393508,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00036807758285640346,"output_power":0.000490407839436238,"delta_db":1.2461803571945473,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5503678931332305,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.3076148207308679,"pitch":null,"output_rms_dbfs":-29.697763929819992,"output_peak":0.5484497141252593,"reference_rms_dbfs":-30.00537875055086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.16966616479034},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76160,"output_samples":76160,"intelligibility":{"estoi":{"value":0.599662943864064,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031603200210495876,"rms_dbfs":-30.00537875055086,"peak":0.368072509765625,"peak_dbfs":-8.68133234994865,"crest_factor_db":21.32404640060221,"mean":-3.006077614151129e-06,"zero_energy":false},"output":{"rms":0.03274249753637997,"rms_dbfs":-29.697763929819992,"peak":0.5484497141252593,"peak_dbfs":-5.217263711067474,"crest_factor_db":24.480500218752518,"mean":0.0010577666906205886,"zero_energy":false},"rms_difference_db":0.3076148207308691},"full_scale":{"reference":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76160,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":476,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.6521784957731147e-06,"threshold_dbfs":-55.763972505837614,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":56,"quiet_samples":8960,"reference_mean_power":1.7876799185095088e-06,"output_mean_power":1.7973983149517694e-06,"output_rms_dbfs":-57.45355669793198,"output_zero_energy":false,"output_peak":0.005978148601906188,"output_energy_fraction":0.00019724308996699362,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":469,"reference_estimated_voiced_frames":24,"output_estimated_voiced_frames":21,"both_estimated_voiced_frames":16,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.6666666666666666,"absolute_difference_cents_median":14.36771079735853,"absolute_difference_cents_p95":29.934973552633405,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76160,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018341452353274373,"log_rms_mae_db":3.423978514675289,"rms_envelope_nrmse":0.4151752176144249,"envelope_correlation":0.8662316878702845,"rms_gain_error_db":0.30765389736632914,"reference_mean_power":0.0009998027766238483,"output_mean_power":0.001073197687676275},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":5.754282242130778e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018894414816227905,"output_power":0.0002262832294418911,"reference_active":true,"delta_db":0.7831892240914649},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013084274193993133,"output_power":0.0001761781715106155,"reference_active":true,"delta_db":1.2920246166699247},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.690431882761885e-05,"output_power":4.167407251055813e-05,"reference_active":true,"delta_db":0.5278874942209023}],"active_band_count":3,"active_band_mae_db":0.8677004449940973}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.0001230241515816465,"log_rms_mae_db":2.442828610213988,"rms_envelope_nrmse":0.27911211154173426,"envelope_correlation":0.9369335820331601,"rms_gain_error_db":0.30778201776511416,"reference_mean_power":0.001006100998286567,"output_mean_power":0.0010799901180591908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":5.061730772699165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019045538591950462,"output_power":0.00023482443583563916,"reference_active":true,"delta_db":0.9095002896342588},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.588390935994399e-05,"output_power":0.0001002212426867665,"reference_active":true,"delta_db":1.2081008682611167},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5269059669432984e-06,"output_power":2.9387104712854995e-06,"reference_active":true,"delta_db":0.6556771951996987}],"active_band_count":3,"active_band_mae_db":0.9244261176983581}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":5.489840871101384e-05,"log_rms_mae_db":4.020321067480057,"rms_envelope_nrmse":0.4371528185583438,"envelope_correlation":0.8646777090038515,"rms_gain_error_db":-0.3495640296183281,"reference_mean_power":0.0001499835837122031,"output_mean_power":0.00013838446002511075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.245105068973029e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.560041827470393e-05,"output_power":6.0090315020200086e-05,"reference_active":true,"delta_db":-0.38102127704371735},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.138220090606147e-05,"output_power":1.9969300245522836e-05,"reference_active":true,"delta_db":-0.2968955903630987},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.449241698586229e-06,"output_power":2.1257575402821083e-06,"reference_active":true,"delta_db":-0.6151791649505044}],"active_band_count":3,"active_band_mae_db":0.4310320107857735}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":4.550505877934166e-05,"log_rms_mae_db":2.8597442887805693,"rms_envelope_nrmse":0.3203298040530478,"envelope_correlation":0.9235892751903977,"rms_gain_error_db":-0.3495548426864023,"reference_mean_power":0.00015093488949896852,"output_mean_power":0.00013926249026159716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":1.1354747567630586e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.429099822482352e-05,"output_power":5.87936946392426e-05,"reference_active":true,"delta_db":-0.3881941652600711},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0916365304226932e-05,"output_power":1.0164148539470313e-05,"reference_active":true,"delta_db":-0.31007057200876087},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3162457402511593e-07,"output_power":2.1344997643422592e-07,"reference_active":true,"delta_db":-0.3548852243568001}],"active_band_count":3,"active_band_mae_db":0.351049987208544}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":4.066436664003698e-05,"log_rms_mae_db":3.314964125673089,"rms_envelope_nrmse":0.49773609284334575,"envelope_correlation":0.8116490787979912,"rms_gain_error_db":-0.26184750780508975,"reference_mean_power":0.0001020893317881212,"output_mean_power":9.611598266768909e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":9.602503308503008e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0812258101146943e-05,"output_power":2.884941898671092e-05,"reference_active":true,"delta_db":-0.28586456004229593},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3389822486480089e-05,"output_power":1.2004436843698343e-05,"reference_active":true,"delta_db":-0.4743302835995531},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.310273392307661e-06,"output_power":2.856211625394131e-06,"reference_active":true,"delta_db":-2.693264982328434}],"active_band_count":3,"active_band_mae_db":1.1511532753234277}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":3.8037796302268924e-05,"log_rms_mae_db":2.2094014398805317,"rms_envelope_nrmse":0.330315370874355,"envelope_correlation":0.9112556283206475,"rms_gain_error_db":-0.2618346959776007,"reference_mean_power":0.00010273559922291735,"output_mean_power":9.672472168779563e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":8.714815910547829e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.937520326415114e-05,"output_power":2.701285275568906e-05,"reference_active":true,"delta_db":-0.3641042912972728},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.005890574709939e-06,"output_power":5.481713751008885e-06,"reference_active":true,"delta_db":-1.065469963503674},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.153574933783463e-07,"output_power":2.558448326334516e-07,"reference_active":true,"delta_db":-3.0413194530259697}],"active_band_count":3,"active_band_mae_db":1.4902979026089722}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.00018081886412701817,"log_rms_mae_db":3.045315360467625,"rms_envelope_nrmse":0.5120800700958769,"envelope_correlation":0.886240428800531,"rms_gain_error_db":1.2955410831954417,"reference_mean_power":0.00040250829399997306,"output_mean_power":0.0005424115666861472},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":3.596628676311722e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013671994273759655,"output_power":0.0001916792692877598,"reference_active":true,"delta_db":1.4674327729468164},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.299777768885513e-05,"output_power":0.00014039898425959523,"reference_active":true,"delta_db":1.788913952199178},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4959378812525703e-05,"output_power":3.249649734806015e-05,"reference_active":true,"delta_db":1.1460278039316627}],"active_band_count":3,"active_band_mae_db":1.4674581763592192}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76160,"uncovered_tail_samples":0,"metrics":{"frame_count":945,"rms_floor":0.00011978792663441868,"log_rms_mae_db":2.1324927748539784,"rms_envelope_nrmse":0.34247252056000543,"envelope_correlation":0.9549138382734127,"rms_gain_error_db":1.2955404372127188,"reference_mean_power":0.0004050621320160613,"output_mean_power":0.0005458529829740451},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":945,"frequency_resolution_hz":0.21164021164021163,"power_floor":3.4505381724992573e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016854041012359926,"output_power":0.00023537465241201019,"reference_active":true,"delta_db":1.4505564529172759},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.211674623356502e-05,"output_power":7.646794743669653e-05,"reference_active":true,"delta_db":1.6650213905300153},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0452544406168717e-06,"output_power":2.6726805883298877e-06,"reference_active":true,"delta_db":1.1619971512094094}],"active_band_count":3,"active_band_mae_db":1.4258583315522335}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"hnm","status":"measured","mos":{"status":"measured","value":3.101025104522705,"predicted_mos":3.101025104522705,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59840,"input_sample_rate":16000,"input_seconds":3.74,"input_float32_sha256":"c708c18116955dedd25bbec877e6149197242002011e4aef7247dbff5c9877de","internal_samples":59840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2447773100011545},"started_at":"2026-10-13T04:39:50.656511+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3421000403468497e-05,"native_input_samples":59840,"native_input_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","output_sha256":"4b807dc9304fbed2479d8d6cf695026272e6f87f2ac8c15d1d6334b4446d469b","output_samples":59840,"length_delta_samples":0,"analysis_s":0.8751523850005469,"synthesis_s":0.311292472999412,"total_s":1.186444857999959,"rtf":0.31723124545453446,"output_peak":0.30519236042503584,"output_rms":0.038297309149055374,"finished_at":"2026-10-13T04:39:51.847050+00:00","evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"4b807dc9304fbed2479d8d6cf695026272e6f87f2ac8c15d1d6334b4446d469b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.522197786653838,"spectral_convergence":0.2740490781739609,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.214976247644672,"spectral_convergence":0.2531255457040616,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.658199598075056,"spectral_convergence":0.2601553478585203,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.00025394582317242155,"delta_db":0.0013449780399047036,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.00048330983211131964,"delta_db":-0.05524113947360961,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0005427629371012559,"delta_db":0.004919255810065732,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.585624404922106e-05,"delta_db":0.06910393188529594,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":9.819942478434779e-05,"delta_db":-0.40884285286727506,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":7.260962683975473e-05,"delta_db":-0.08612037649256307,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1042620890947857,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04826631952755105,"pitch":null,"output_rms_dbfs":-28.336634788393,"output_peak":0.30519236042503584,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.465124544124522},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.522197786653838,"spectral_convergence":0.2740490781739609,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.214976247644672,"spectral_convergence":0.2531255457040616,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.658199598075056,"spectral_convergence":0.2601553478585203,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.00025394582317242155,"delta_db":0.0013449780399047036,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.00048330983211131964,"delta_db":-0.05524113947360961,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0005427629371012559,"delta_db":0.004919255810065732,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.585624404922106e-05,"delta_db":0.06910393188529594,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":9.819942478434779e-05,"delta_db":-0.40884285286727506,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":7.260962683975473e-05,"delta_db":-0.08612037649256307,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1042620890947857,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04826631952755105,"pitch":null,"output_rms_dbfs":-28.336634788393,"output_peak":0.30519236042503584,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.465124544124522},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.8218131314279402,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03851071452419941,"rms_dbfs":-28.28836846886545,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.589436133538992,"mean":-9.952381970410678e-06,"zero_energy":false},"output":{"rms":0.038297309149055374,"rms_dbfs":-28.336634788393,"peak":0.30519236042503584,"peak_dbfs":-10.308526836805566,"crest_factor_db":18.028107951587433,"mean":-0.00017099795344109863,"zero_energy":false},"rms_difference_db":-0.04826631952754923},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":1.8467326336293835e-06,"output_rms_dbfs":-57.335959763218085,"output_zero_energy":false,"output_peak":0.006490347369946448,"output_energy_fraction":3.3666339629528635e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":172,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9814814814814815,"absolute_difference_cents_median":1.2912116371653315,"absolute_difference_cents_p95":8.55332171612936,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.000110539633783132,"log_rms_mae_db":1.0738340862649933,"rms_envelope_nrmse":0.07479130043177423,"envelope_correlation":0.994304326930876,"rms_gain_error_db":-0.0482697195962676,"reference_mean_power":0.00148505750597689,"output_mean_power":0.0014686432014936878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.0996976338580339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028388367321682665,"output_power":0.000280878143391521,"reference_active":true,"delta_db":-0.04622470225698463},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.612183760153141e-05,"output_power":4.426454072542617e-05,"reference_active":true,"delta_db":-0.17850639792196382},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3216272322760606e-05,"output_power":1.1617122457822705e-05,"reference_active":true,"delta_db":-0.56010411410316}],"active_band_count":3,"active_band_mae_db":0.26161173809403615}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00010014294879668002,"log_rms_mae_db":0.5679763209136658,"rms_envelope_nrmse":0.03971428926953841,"envelope_correlation":0.9983331911714712,"rms_gain_error_db":-0.048286011255491014,"reference_mean_power":0.0014970488171277764,"output_mean_power":0.0014804964191876455},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.0259449632693504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002533237576934496,"output_power":0.00025073551702395406,"reference_active":true,"delta_db":-0.04460064822355194},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.399735365000864e-05,"output_power":2.2734807737820995e-05,"reference_active":true,"delta_db":-0.23472065926782026},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.21988770898296e-07,"output_power":4.76582268919621e-07,"reference_active":true,"delta_db":-1.1564466367373056}],"active_band_count":3,"active_band_mae_db":0.47858931474289257}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00010594974849403162,"log_rms_mae_db":1.317633214088713,"rms_envelope_nrmse":0.07584440672807806,"envelope_correlation":0.99608001487296,"rms_gain_error_db":-0.04235955838497441,"reference_mean_power":0.0008075829245111437,"output_mean_power":0.0007997443338867976},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2363198879746264e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030947682521494476,"output_power":0.0003103518546177118,"reference_active":true,"delta_db":0.012262120355091173},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.465844834587485e-05,"output_power":3.216489622051133e-05,"reference_active":true,"delta_db":-0.324269610119861},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.693327955984244e-06,"output_power":7.635744207802582e-06,"reference_active":true,"delta_db":0.5720926655197627}],"active_band_count":3,"active_band_mae_db":0.30287479866490497}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":0.8460229603740046,"rms_envelope_nrmse":0.04387878326538515,"envelope_correlation":0.9986663997795343,"rms_gain_error_db":-0.04236008178658228,"reference_mean_power":0.0008141211968758593,"output_mean_power":0.0008062190470718868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.204812078547339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002902058778585621,"output_power":0.0002912009537227377,"reference_active":true,"delta_db":0.014865885826350891},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7747321249654287e-05,"output_power":1.6508872914766425e-05,"reference_active":true,"delta_db":-0.31415386324086253},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.319074590753967e-07,"output_power":3.1303441706793916e-07,"reference_active":true,"delta_db":-0.25424923203348115}],"active_band_count":3,"active_band_mae_db":0.19442299370023153}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":1.3881981288404788,"rms_envelope_nrmse":0.2024985437024836,"envelope_correlation":0.9649321349021116,"rms_gain_error_db":-0.17407697025926125,"reference_mean_power":4.5961113590202697e-05,"output_mean_power":4.4155300592493076e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.8606659711815018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3098445296769507e-05,"output_power":1.2071995304765474e-05,"reference_active":true,"delta_db":-0.3544069283126915},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.217362950062473e-06,"output_power":3.0941061713503385e-06,"reference_active":true,"delta_db":-0.1696484431138689},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2063471779341695e-06,"output_power":1.5986819643283524e-06,"reference_active":true,"delta_db":1.2228976287425157}],"active_band_count":3,"active_band_mae_db":0.5823176667230253}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":0.8888676413813531,"rms_envelope_nrmse":0.1314609527038421,"envelope_correlation":0.9844065006550348,"rms_gain_error_db":-0.17411976792993725,"reference_mean_power":4.6330782152537655e-05,"output_mean_power":4.451000624296209e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.575546676565828e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2064714839118608e-05,"output_power":1.1283289740178398e-05,"reference_active":true,"delta_db":-0.2908132100594561},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6378421988809907e-06,"output_power":1.5567562590810506e-06,"reference_active":true,"delta_db":-0.22051435968665586},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.209021145285469e-07,"output_power":1.0812011141727468e-07,"reference_active":true,"delta_db":-0.4852741199472965}],"active_band_count":3,"active_band_mae_db":0.33220056323113617}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":1.4321293405544553,"rms_envelope_nrmse":0.18172178545146256,"envelope_correlation":0.9790231683200247,"rms_gain_error_db":-0.35318220434751946,"reference_mean_power":0.00012461308874590145,"output_mean_power":0.00011488026810347998},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.481947168400202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.119164755141344e-05,"output_power":5.6792224606082784e-05,"reference_active":true,"delta_db":-0.3240326574445497},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.192933297758675e-05,"output_power":1.6268643146556737e-05,"reference_active":true,"delta_db":-1.2967408899495316},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.860956874662954e-06,"output_power":3.4241868326185258e-06,"reference_active":true,"delta_db":-3.018272366287435}],"active_band_count":3,"active_band_mae_db":1.546348637893839}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":1.0044247098698877,"rms_envelope_nrmse":0.11043690038394137,"envelope_correlation":0.9925635986464364,"rms_gain_error_db":-0.35319884974463134,"reference_mean_power":0.00012561601350824357,"output_mean_power":0.00011580441625901172},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.3825123163038373e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7375597371987975e-05,"output_power":5.20392272706419e-05,"reference_active":true,"delta_db":-0.4239638116286121},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1774546568382313e-05,"output_power":8.21021503931016e-06,"reference_active":true,"delta_db":-1.565896595003363},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.943011691356466e-07,"output_power":2.4952357171428394e-07,"reference_active":true,"delta_db":-3.7689500627867787}],"active_band_count":3,"active_band_mae_db":1.919603489806251}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.6653614044189453,"predicted_mos":3.6653614044189453,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59840,"input_sample_rate":16000,"input_seconds":3.74,"input_float32_sha256":"846f0c51f474e72696120951d0aff9b50a85145ad58c0eea62caa5bfb99e19d6","internal_samples":59840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.297369805000926},"started_at":"2026-10-13T04:43:00.228061+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.549700002418831e-05,"native_input_samples":59840,"native_input_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","output_sha256":"7fdd49bb850ee3b1b806bb5d626adf80143c9698f2e8bf83e913d7ed5da8c54b","output_samples":59840,"length_delta_samples":0,"analysis_s":2.766764124000474,"synthesis_s":0.04937030299970502,"total_s":2.816134427000179,"rtf":0.7529771195187644,"output_peak":0.34579067780434886,"output_rms":0.0419876494996743,"finished_at":"2026-10-13T04:43:03.050319+00:00","evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"7fdd49bb850ee3b1b806bb5d626adf80143c9698f2e8bf83e913d7ed5da8c54b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.677768929805292,"spectral_convergence":0.7314352658580539,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.243378902843287,"spectral_convergence":0.3139516812068553,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.815548787735265,"spectral_convergence":0.3086125290826318,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.0002537088676739589,"delta_db":-0.002709292501393711,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.0006027136278729542,"delta_db":0.903612671962733,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0006872096867777784,"delta_db":1.0297101424404145,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.7516419131500563e-05,"delta_db":0.5015540002016312,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":0.0001288491483895824,"delta_db":0.7708832334197536,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":7.296496066172516e-05,"delta_db":-0.06491889795916556,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.545564706414182,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7507997325728157,"pitch":null,"output_rms_dbfs":-27.537568736292634,"output_peak":0.34579067780434886,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.912232206794616},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.677768929805292,"spectral_convergence":0.7314352658580539,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.243378902843287,"spectral_convergence":0.3139516812068553,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.815548787735265,"spectral_convergence":0.3086125290826318,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.0002537088676739589,"delta_db":-0.002709292501393711,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.0006027136278729542,"delta_db":0.903612671962733,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0006872096867777784,"delta_db":1.0297101424404145,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.7516419131500563e-05,"delta_db":0.5015540002016312,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":0.0001288491483895824,"delta_db":0.7708832334197536,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":7.296496066172516e-05,"delta_db":-0.06491889795916556,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.545564706414182,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7507997325728157,"pitch":null,"output_rms_dbfs":-27.537568736292634,"output_peak":0.34579067780434886,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.912232206794616},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.8510893077550719,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03851071452419941,"rms_dbfs":-28.28836846886545,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.589436133538992,"mean":-9.952381970410678e-06,"zero_energy":false},"output":{"rms":0.0419876494996743,"rms_dbfs":-27.537568736292634,"peak":0.34579067780434886,"peak_dbfs":-9.22373438160647,"crest_factor_db":18.313834354686165,"mean":-1.914560690262817e-08,"zero_energy":false},"rms_difference_db":0.7507997325728155},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":6.733195571714676e-07,"output_rms_dbfs":-61.71778770875442,"output_zero_energy":false,"output_peak":0.002799457040724373,"output_energy_fraction":1.0211898759124694e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":177,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9444444444444444,"absolute_difference_cents_median":4.877672101937844,"absolute_difference_cents_p95":16.521647210451405,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.000110539633783132,"log_rms_mae_db":2.3598871324437733,"rms_envelope_nrmse":0.16982325292848147,"envelope_correlation":0.981796296812407,"rms_gain_error_db":0.7508055309090023,"reference_mean_power":0.00148505750597689,"output_mean_power":0.001765321555814838},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.0996976338580339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028388367321682665,"output_power":0.00035153357959710237,"reference_active":true,"delta_db":0.9282640051723732},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.612183760153141e-05,"output_power":6.799831021456174e-05,"reference_active":true,"delta_db":1.6859151817529006},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3216272322760606e-05,"output_power":1.9105220708267432e-05,"reference_active":true,"delta_db":1.6004308037682784}],"active_band_count":3,"active_band_mae_db":1.4048699968978504}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00010014294879668002,"log_rms_mae_db":2.1467343318692707,"rms_envelope_nrmse":0.11753566785661701,"envelope_correlation":0.9951620156414368,"rms_gain_error_db":0.750869252951311,"reference_mean_power":0.0014970488171277764,"output_mean_power":0.0017796020105775002},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.0259449632693504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002533237576934496,"output_power":0.0003172247399726236,"reference_active":true,"delta_db":0.9769112852933594},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.399735365000864e-05,"output_power":3.5360550278180675e-05,"reference_active":true,"delta_db":1.6835566307981507},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.21988770898296e-07,"output_power":1.1831363170025291e-06,"reference_active":true,"delta_db":2.7925224125891064}],"active_band_count":3,"active_band_mae_db":1.817663442893539}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00010594974849403162,"log_rms_mae_db":1.6570648431878814,"rms_envelope_nrmse":0.1959937688794855,"envelope_correlation":0.985897857002455,"rms_gain_error_db":1.0070435388167391,"reference_mean_power":0.0008075829245111437,"output_mean_power":0.0010183369014036855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2363198879746264e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030947682521494476,"output_power":0.0003859055090662275,"reference_active":true,"delta_db":0.9585284549315082},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.465844834587485e-05,"output_power":5.3015082770272596e-05,"reference_active":true,"delta_db":1.845903282830369},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.693327955984244e-06,"output_power":6.334608571286877e-06,"reference_active":true,"delta_db":-0.23922320761711355}],"active_band_count":3,"active_band_mae_db":1.0145516484596635}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":1.290010264878721,"rms_envelope_nrmse":0.14682349087929278,"envelope_correlation":0.9960013475692071,"rms_gain_error_db":1.0070410561578742,"reference_mean_power":0.0008141211968758593,"output_mean_power":0.0010265808722573124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.204812078547339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002902058778585621,"output_power":0.00036205456105549503,"reference_active":true,"delta_db":0.9606781852934},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7747321249654287e-05,"output_power":2.7607595186096265e-05,"reference_active":true,"delta_db":1.9189576754067061},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.319074590753967e-07,"output_power":3.359324882349929e-07,"reference_active":true,"delta_db":0.05234994181942072}],"active_band_count":3,"active_band_mae_db":0.9773286008398423}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":1.228749718455366,"rms_envelope_nrmse":0.15697843324644797,"envelope_correlation":0.9837722262560213,"rms_gain_error_db":0.5420133123111351,"reference_mean_power":4.5961113590202697e-05,"output_mean_power":5.207051509863138e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.8606659711815018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3098445296769507e-05,"output_power":1.4823780810198653e-05,"reference_active":true,"delta_db":0.5373923403649682},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.217362950062473e-06,"output_power":3.4716016755314558e-06,"reference_active":true,"delta_db":0.33029832965141476},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2063471779341695e-06,"output_power":7.623692145984964e-07,"reference_active":true,"delta_db":-1.9930696194054849}],"active_band_count":3,"active_band_mae_db":0.9535867631406226}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":0.7249906788847322,"rms_envelope_nrmse":0.09016695324185103,"envelope_correlation":0.9966095650806328,"rms_gain_error_db":0.5420225549282777,"reference_mean_power":4.6330782152537655e-05,"output_mean_power":5.2489433726807086e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.575546676565828e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2064714839118608e-05,"output_power":1.3605177632822e-05,"reference_active":true,"delta_db":0.5218715500958901},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6378421988809907e-06,"output_power":1.690594134295407e-06,"reference_active":true,"delta_db":0.13767301322848843},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.209021145285469e-07,"output_power":7.119532112835438e-08,"reference_active":true,"delta_db":-2.299824433011444}],"active_band_count":3,"active_band_mae_db":0.986456332111941}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":1.3985723865197965,"rms_envelope_nrmse":0.2386180428264367,"envelope_correlation":0.9664093733948518,"rms_gain_error_db":0.38276190233559493,"reference_mean_power":0.00012461308874590145,"output_mean_power":0.00013609427335124983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.481947168400202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.119164755141344e-05,"output_power":5.614942832037979e-05,"reference_active":true,"delta_db":-0.3734680759837872},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.192933297758675e-05,"output_power":2.9182875959457634e-05,"reference_active":true,"delta_db":1.2410266720979937},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.860956874662954e-06,"output_power":8.267248282708177e-06,"reference_active":true,"delta_db":0.8097629106583397}],"active_band_count":3,"active_band_mae_db":0.8080858862467069}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":0.8707967114763945,"rms_envelope_nrmse":0.15170818874707578,"envelope_correlation":0.9864883001293033,"rms_gain_error_db":0.3827523889825597,"reference_mean_power":0.00012561601350824357,"output_mean_power":0.00013718930172823828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.3825123163038373e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7375597371987975e-05,"output_power":5.3984166792394075e-05,"reference_active":true,"delta_db":-0.2646081754017282},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1774546568382313e-05,"output_power":1.4836484660029654e-05,"reference_active":true,"delta_db":1.003868202161311},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.943011691356466e-07,"output_power":7.852763579588812e-07,"reference_active":true,"delta_db":1.2101593787437948}],"active_band_count":3,"active_band_mae_db":0.8262119187689446}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"facodec","status":"measured","mos":{"status":"measured","value":4.398837089538574,"predicted_mos":4.398837089538574,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"d4cd7bcff731771be88eae09d370d78443648f33e360f01ccb39748b7a02e3ca","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7441452349994506},"started_at":"2026-10-13T06:56:55.586260+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.543200157058891e-05,"native_input_samples":59840,"native_input_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","output_sha256":"bbb80a0cb933ea0a7f9be39db45890239f060f238006c82e9a0b1e38d2582eb2","output_samples":60000,"length_delta_samples":160,"analysis_s":0.9931974549999723,"synthesis_s":7.811605333999978,"total_s":8.80480278899995,"rtf":2.354225344652393,"output_peak":0.28123074769973755,"output_rms":0.036025073567503854,"finished_at":"2026-10-13T06:57:04.398763+00:00","evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"bbb80a0cb933ea0a7f9be39db45890239f060f238006c82e9a0b1e38d2582eb2","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":59840,"output_samples":60000,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.867902472637077,"output_peak":0.28123074769973755},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973333333333333,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.556849390754222,"spectral_convergence":0.2874034693598973,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.824577475559307,"spectral_convergence":0.29558480680941135,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.046767030368102,"spectral_convergence":0.3118297826004527,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.00022673756056351956,"delta_db":-0.4908310645458004,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.0004839133177687106,"delta_db":-0.04982169671201982,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.000440219185237729,"delta_db":-0.9044929269277914,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.1162807258686552e-05,"delta_db":-1.4551650227772592,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":9.085731155715769e-05,"delta_db":-0.7463334713128411,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":4.838553013569392e-05,"delta_db":-1.848907387006132,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.915925261546974,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5679382908026034,"pitch":null,"output_rms_dbfs":-28.85630675966805,"output_peak":0.28123074769973755,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.809397965560543},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.7980372536147957,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03851071452419941,"rms_dbfs":-28.28836846886545,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.589436133538992,"mean":-9.952381970410678e-06,"zero_energy":false},"output":{"rms":0.03607319936630929,"rms_dbfs":-28.85630675966805,"peak":0.28123074769973755,"peak_dbfs":-11.01874396989272,"crest_factor_db":17.837562789775333,"mean":-0.00018583992106606484,"zero_energy":false},"rms_difference_db":-0.5679382908026014},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":9.762177588506804e-07,"output_rms_dbfs":-60.104532961461,"output_zero_energy":false,"output_peak":0.0035595514345914125,"output_energy_fraction":2.0058838309581803e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":174,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9814814814814815,"absolute_difference_cents_median":3.1871457697763774,"absolute_difference_cents_p95":12.6998893456858,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.000110539633783132,"log_rms_mae_db":1.410806619261212,"rms_envelope_nrmse":0.1276848994910264,"envelope_correlation":0.986646342341148,"rms_gain_error_db":-0.5679437367247053,"reference_mean_power":0.00148505750597689,"output_mean_power":0.0013030134466926573},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.0996976338580339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028388367321682665,"output_power":0.0002404066635596804,"reference_active":true,"delta_db":-0.7219391478606753},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.612183760153141e-05,"output_power":4.2117056065109834e-05,"reference_active":true,"delta_db":-0.39448595393879304},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3216272322760606e-05,"output_power":1.0681243299744856e-05,"reference_active":true,"delta_db":-0.9248717111900874}],"active_band_count":3,"active_band_mae_db":0.6804322709965186}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00010014294879668002,"log_rms_mae_db":1.2457353585594895,"rms_envelope_nrmse":0.11544017216600916,"envelope_correlation":0.9893692192863967,"rms_gain_error_db":-0.5679117939712822,"reference_mean_power":0.0014970488171277764,"output_mean_power":0.0013135444780181791},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.0259449632693504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002533237576934496,"output_power":0.0002132180346437203,"reference_active":true,"delta_db":-0.7485198565226387},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.399735365000864e-05,"output_power":2.234238882771481e-05,"reference_active":true,"delta_db":-0.3103374614166401},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.21988770898296e-07,"output_power":4.915505118984471e-07,"reference_active":true,"delta_db":-1.0221439149643927}],"active_band_count":3,"active_band_mae_db":0.6936670776345571}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00010594974849403162,"log_rms_mae_db":1.652597776895445,"rms_envelope_nrmse":0.11934582041998587,"envelope_correlation":0.9919260338955276,"rms_gain_error_db":-0.4700173285276774,"reference_mean_power":0.0008075829245111437,"output_mean_power":0.0007247452786343145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2363198879746264e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030947682521494476,"output_power":0.00027244276102108243,"reference_active":true,"delta_db":-0.5535286006408371},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.465844834587485e-05,"output_power":3.3454273246871874e-05,"reference_active":true,"delta_db":-0.15357515704886662},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.693327955984244e-06,"output_power":5.534565417056392e-06,"reference_active":true,"delta_db":-0.8255857979084193}],"active_band_count":3,"active_band_mae_db":0.5108965185327077}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":1.3464739693863201,"rms_envelope_nrmse":0.10566549370899526,"envelope_correlation":0.9940287672543473,"rms_gain_error_db":-0.4700217810506011,"reference_mean_power":0.0008141211968758593,"output_mean_power":0.0007306121400618344},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.204812078547339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002902058778585621,"output_power":0.00025349675853808556,"reference_active":true,"delta_db":-0.5873379405093482},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7747321249654287e-05,"output_power":1.7252557146519685e-05,"reference_active":true,"delta_db":-0.12279336054647669},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.319074590753967e-07,"output_power":2.689954609398599e-07,"reference_active":true,"delta_db":-0.9127206084443539}],"active_band_count":3,"active_band_mae_db":0.5409506365000596}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":1.667186384793445,"rms_envelope_nrmse":0.18078426147566748,"envelope_correlation":0.9741649929563744,"rms_gain_error_db":-0.5400703551982802,"reference_mean_power":4.5961113590202697e-05,"output_mean_power":4.058667810759687e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.8606659711815018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3098445296769507e-05,"output_power":1.0998814975789696e-05,"reference_active":true,"delta_db":-0.7587385436408146},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.217362950062473e-06,"output_power":3.011529081658639e-06,"reference_active":true,"delta_db":-0.28712994927179314},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2063471779341695e-06,"output_power":2.4893305789535255e-06,"reference_active":true,"delta_db":3.1461026143960873}],"active_band_count":3,"active_band_mae_db":1.3973237024362317}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":1.3220615111347955,"rms_envelope_nrmse":0.13299712651031562,"envelope_correlation":0.9864217904125042,"rms_gain_error_db":-0.540053512897312,"reference_mean_power":4.6330782152537655e-05,"output_mean_power":4.0913278361239386e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.575546676565828e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2064714839118608e-05,"output_power":1.0765437503922243e-05,"reference_active":true,"delta_db":-0.49485377301205646},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6378421988809907e-06,"output_power":1.5776883989941655e-06,"reference_active":true,"delta_db":-0.16250824381206308},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.209021145285469e-07,"output_power":2.4989066736991497e-07,"reference_active":true,"delta_db":3.1531614033692716}],"active_band_count":3,"active_band_mae_db":1.270174473397797}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":2.0652999988436753,"rms_envelope_nrmse":0.22201799223680935,"envelope_correlation":0.9812441673768042,"rms_gain_error_db":-1.4819148865568694,"reference_mean_power":0.00012461308874590145,"output_mean_power":8.858744420119807e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.481947168400202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.119164755141344e-05,"output_power":4.466866024479846e-05,"reference_active":true,"delta_db":-1.3668921969956158},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.192933297758675e-05,"output_power":1.4888674710587933e-05,"reference_active":true,"delta_db":-1.6816938047815588},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.860956874662954e-06,"output_power":3.799802879793002e-06,"reference_active":true,"delta_db":-2.5662362197284665}],"active_band_count":3,"active_band_mae_db":1.871607407168547}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":1.7448945032565244,"rms_envelope_nrmse":0.1876333915070094,"envelope_correlation":0.991550931435758,"rms_gain_error_db":-1.4818322292122297,"reference_mean_power":0.00012561601350824357,"output_mean_power":8.930212304158172e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.3825123163038373e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7375597371987975e-05,"output_power":4.157259397207154e-05,"reference_active":true,"delta_db":-1.3992009659200222},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1774546568382313e-05,"output_power":7.484462798239245e-06,"reference_active":true,"delta_db":-1.9678355761662192},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.943011691356466e-07,"output_power":3.286041417728634e-07,"reference_active":true,"delta_db":-2.573335515801076}],"active_band_count":3,"active_band_mae_db":1.9801240192957723}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.544311046600342,"predicted_mos":3.544311046600342,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":165376,"input_sample_rate":44100,"input_seconds":3.7500226757369615,"input_float32_sha256":"0ff601b770b46573a4d005f08847a84e500bf8737ba3e71c0fa8c7b57e63c229","internal_samples":60001,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.592306806000124},"started_at":"2026-10-13T07:23:38.358831+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005516381999768782,"native_input_samples":164934,"native_input_sha256":"7581e937a61730fd480451c750d8207d128c636055de913b786a86ce704c506d","output_sha256":"95a0d02bf400a2122af2160aeb4b09581510488c7fe287167bf6589396e71ec1","output_samples":165376,"length_delta_samples":442,"analysis_s":0.09792359600032796,"synthesis_s":20.227178896000623,"total_s":20.32510249200095,"rtf":5.434519382887955,"output_peak":0.3557721674442291,"output_rms":0.03594736335716197,"finished_at":"2026-10-13T07:23:58.705668+00:00","evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"95a0d02bf400a2122af2160aeb4b09581510488c7fe287167bf6589396e71ec1","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":59840,"output_samples":60001,"length_delta_samples":161,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.88211538500415,"output_peak":0.3568239153737397},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973167113881435,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.797700867696335,"spectral_convergence":0.44466904248027,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.605689169390615,"spectral_convergence":0.38766129729826754,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.643683721240526,"spectral_convergence":0.4410437048356315,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.00019942177324678347,"delta_db":-1.0483400261093765,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.00046652733426656457,"delta_db":-0.2087264961199755,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0004916142620601996,"delta_db":-0.4249378478314964,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.3932496772084333e-05,"delta_db":-0.49260977278671486,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":8.468362307895744e-05,"delta_db":-1.05193798674942,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":4.086234052782653e-05,"delta_db":-2.582830037726777,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9682303612206268,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5820917068814776,"pitch":null,"output_rms_dbfs":-28.870460175746928,"output_peak":0.3568239153737397,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.682357919442493},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.6022289977968722,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03851071452419941,"rms_dbfs":-28.28836846886545,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.589436133538992,"mean":-9.952381970410678e-06,"zero_energy":false},"output":{"rms":0.036014466953606515,"rms_dbfs":-28.870460175746928,"peak":0.3568239153737397,"peak_dbfs":-8.950920912892531,"crest_factor_db":19.919539262854396,"mean":-0.0002292955666077556,"zero_energy":false},"rms_difference_db":-0.5820917068814779},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":9.609559120189e-07,"output_rms_dbfs":-60.17296536999646,"output_zero_energy":false,"output_peak":0.003728227460816916,"output_energy_fraction":1.980969907785778e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":168,"both_estimated_voiced_frames":142,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8765432098765432,"absolute_difference_cents_median":13.349535054590156,"absolute_difference_cents_p95":40.65962927025033,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.000110539633783132,"log_rms_mae_db":1.7267578790979938,"rms_envelope_nrmse":0.11755202039867567,"envelope_correlation":0.9906686831020594,"rms_gain_error_db":-0.5820918588873181,"reference_mean_power":0.00148505750597689,"output_mean_power":0.0012987754933288006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.0996976338580339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028388367321682665,"output_power":0.0002777219826130901,"reference_active":true,"delta_db":-0.09530159044988532},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.612183760153141e-05,"output_power":3.282439847301988e-05,"reference_active":true,"delta_db":-1.4770982614692376},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3216272322760606e-05,"output_power":1.0776877237531583e-05,"reference_active":true,"delta_db":-0.8861604304104957}],"active_band_count":3,"active_band_mae_db":0.8195200941098729}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00010014294879668002,"log_rms_mae_db":1.6078131574064805,"rms_envelope_nrmse":0.09872905267725011,"envelope_correlation":0.9948572846005298,"rms_gain_error_db":-0.5820557264999842,"reference_mean_power":0.0014970488171277764,"output_mean_power":0.0013092735363200165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.0259449632693504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002533237576934496,"output_power":0.0002511418926752469,"reference_active":true,"delta_db":-0.03756758644010872},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.399735365000864e-05,"output_power":1.652448355985013e-05,"reference_active":true,"delta_db":-1.6203545641923682},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.21988770898296e-07,"output_power":4.662429702084084e-07,"reference_active":true,"delta_db":-1.2517024742825822}],"active_band_count":3,"active_band_mae_db":0.9698748749716865}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00010594974849403162,"log_rms_mae_db":2.6448081527861813,"rms_envelope_nrmse":0.11016858960716987,"envelope_correlation":0.9922519722053358,"rms_gain_error_db":-0.2972045270109249,"reference_mean_power":0.0008075829245111437,"output_mean_power":0.0007541655237013824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2363198879746264e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030947682521494476,"output_power":0.0003148217917281765,"reference_active":true,"delta_db":0.07436653291076},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.465844834587485e-05,"output_power":2.7240476136342257e-05,"reference_active":true,"delta_db":-1.0459442116560627},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.693327955984244e-06,"output_power":5.373143056966204e-06,"reference_active":true,"delta_db":-0.9541370127251723}],"active_band_count":3,"active_band_mae_db":0.6914825857639982}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":2.0936985509876394,"rms_envelope_nrmse":0.0874649038522737,"envelope_correlation":0.9952964253091905,"rms_gain_error_db":-0.29720664163090554,"reference_mean_power":0.0008141211968758593,"output_mean_power":0.000760270953247848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.204812078547339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002902058778585621,"output_power":0.00029245634589186633,"reference_active":true,"delta_db":0.03354844947544588},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7747321249654287e-05,"output_power":1.3918501100240232e-05,"reference_active":true,"delta_db":-1.0554034256173248},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.319074590753967e-07,"output_power":2.5665515358969673e-07,"reference_active":true,"delta_db":-1.1166702333391267}],"active_band_count":3,"active_band_mae_db":0.735207369477299}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":2.906628880145538,"rms_envelope_nrmse":0.3067247788814833,"envelope_correlation":0.9275480926690226,"rms_gain_error_db":-1.3183127609060628,"reference_mean_power":4.5961113590202697e-05,"output_mean_power":3.3928078677425636e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.8606659711815018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3098445296769507e-05,"output_power":1.1558599373626183e-05,"reference_active":true,"delta_db":-0.5431453953913047},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.217362950062473e-06,"output_power":3.0448828998856756e-06,"reference_active":true,"delta_db":-0.23929461146349582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2063471779341695e-06,"output_power":1.2440080064990488e-06,"reference_active":true,"delta_db":0.1335086292281525}],"active_band_count":3,"active_band_mae_db":0.30531621202765097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":2.5077840536415885,"rms_envelope_nrmse":0.26663613031302913,"envelope_correlation":0.9458227806976356,"rms_gain_error_db":-1.3186518134055638,"reference_mean_power":4.6330782152537655e-05,"output_mean_power":3.4198294713632134e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.575546676565828e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2064714839118608e-05,"output_power":1.090846793818402e-05,"reference_active":true,"delta_db":-0.4375330190219282},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6378421988809907e-06,"output_power":1.4991395224637401e-06,"reference_active":true,"delta_db":-0.3843000267986195},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.209021145285469e-07,"output_power":8.575419060141751e-08,"reference_active":true,"delta_db":-1.4917854439803704}],"active_band_count":3,"active_band_mae_db":0.7712061632669727}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":3.1667788606307394,"rms_envelope_nrmse":0.25539243671566175,"envelope_correlation":0.9778654395384948,"rms_gain_error_db":-1.864397091317584,"reference_mean_power":0.00012461308874590145,"output_mean_power":8.111925468254718e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.481947168400202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.119164755141344e-05,"output_power":4.099792660866239e-05,"reference_active":true,"delta_db":-1.7393025282741637},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.192933297758675e-05,"output_power":1.1204144601226072e-05,"reference_active":true,"delta_db":-2.916467167718566},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.860956874662954e-06,"output_power":3.726506729440826e-06,"reference_active":true,"delta_db":-2.650827796975001}],"active_band_count":3,"active_band_mae_db":2.43553249765591}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":2.7706090593704498,"rms_envelope_nrmse":0.22451979210040482,"envelope_correlation":0.9891111159551451,"rms_gain_error_db":-1.8642689821335379,"reference_mean_power":0.00012561601350824357,"output_mean_power":8.177453975591651e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.3825123163038373e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7375597371987975e-05,"output_power":3.788753569182457e-05,"reference_active":true,"delta_db":-1.8023086176942171},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1774546568382313e-05,"output_power":5.979770335771084e-06,"reference_active":true,"delta_db":-2.942596872506498},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.943011691356466e-07,"output_power":2.956260256793179e-07,"reference_active":true,"delta_db":-3.032639197461479}],"active_band_count":3,"active_band_mae_db":2.592514895887398}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"golf","status":"measured","mos":{"status":"measured","value":3.8867077827453613,"predicted_mos":3.8867077827453613,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89520,"input_sample_rate":24000,"input_seconds":3.73,"input_float32_sha256":"c73a92fb7e997a4c1605a8a9513c415400e428af5468f824abd5772aee5f3478","internal_samples":59680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2405222759989556},"started_at":"2026-10-13T05:02:50.941521+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001972797999769682,"native_input_samples":89760,"native_input_sha256":"eeb21de4b306024f8c565c16b70a6ef253064d47c4d6619d74f9bd9217ac3142","output_sha256":"0e5a0e51274f459f020558fcf2dac38e95a9d8074f2399ef36c803d5511777d8","output_samples":89520,"length_delta_samples":-240,"analysis_s":0.3690367259996492,"synthesis_s":0.0796544550003091,"total_s":0.4486911809999583,"rtf":0.11997090401068403,"output_peak":0.2795604467391968,"output_rms":0.034390750070295205,"finished_at":"2026-10-13T05:02:51.400079+00:00","analysis_stages":{"external_f0_s":0.04187555499993323},"evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"0e5a0e51274f459f020558fcf2dac38e95a9d8074f2399ef36c803d5511777d8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59840,"output_samples":59680,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.267702159791074,"output_peak":0.27120398352386355},"common_prefix":{"samples":59680,"sample_rate":16000,"reference_coverage":0.9973262032085561,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59680,"output_samples":59680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.19399627762432,"spectral_convergence":0.7070645791852562,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.791507844452232,"spectral_convergence":0.34035435803335184,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.037392934658768,"spectral_convergence":0.3259535273888141,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002544704903871102,"output_power":0.0002268365298213475,"delta_db":-0.49924431967484967,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004906914619526937,"output_power":0.0004382645440992684,"delta_db":-0.4907216395352523,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005437740916040561,"output_power":0.00036115329948372603,"delta_db":-1.7772692464730542,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.566079110581574e-05,"output_power":1.358713622989563e-05,"delta_db":-0.6168576685893284,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010816832255987807,"output_power":6.981025159003697e-05,"delta_db":-1.9018089168993315,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.427618726471757e-05,"output_power":7.401590416965996e-05,"delta_db":-0.015245536954137433,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8835245546876589,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9909325955225943,"pitch":null,"output_rms_dbfs":-29.267702159791074,"output_peak":0.27120398352386355,"reference_rms_dbfs":-28.276769564268477,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.34096568557844},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59680,"output_samples":59680,"intelligibility":{"estoi":{"value":0.8067306304957341,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03856217505372682,"rms_dbfs":-28.276769564268477,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.577837228942023,"mean":-5.9884610828062165e-06,"zero_energy":false},"output":{"rms":0.03440447159009907,"rms_dbfs":-29.267702159791074,"peak":0.27120398352386355,"peak_dbfs":-11.33407871426122,"crest_factor_db":17.933623445529854,"mean":0.0003545763528424773,"zero_energy":false},"rms_difference_db":-0.9909325955225974},"full_scale":{"reference":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":373,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":1.4846713797718052e-06,"output_rms_dbfs":-58.28369663346987,"output_zero_energy":false,"output_peak":0.0036081485914271883,"output_energy_fraction":3.3627278508656365e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":366,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":168,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9444444444444444,"absolute_difference_cents_median":3.6210571653733057,"absolute_difference_cents_p95":14.61349396070485,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.000110539633783132,"log_rms_mae_db":2.324740146363089,"rms_envelope_nrmse":0.19524936437261664,"envelope_correlation":0.9694269339155718,"rms_gain_error_db":-0.9909252862200149,"reference_mean_power":0.0014890321545559138,"output_mean_power":0.0011852543216290803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":1.0979273965708723e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028323131374301757,"output_power":0.00022905025819720674,"reference_active":true,"delta_db":-0.9221048112176834},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.5726290840766905e-05,"output_power":3.734911646130717e-05,"reference_active":true,"delta_db":-0.8788564182701641},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3241297729953963e-05,"output_power":1.1886570920153932e-05,"reference_active":true,"delta_db":-0.46873964878744345}],"active_band_count":3,"active_band_mae_db":0.756566959425097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":0.00010014294879668002,"log_rms_mae_db":1.6804357260103202,"rms_envelope_nrmse":0.15702053093181007,"envelope_correlation":0.9844165682800396,"rms_gain_error_db":-0.9909066038723784,"reference_mean_power":0.001501086277658254,"output_mean_power":0.0011948544199534734},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.0240131533756941e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025266208287539894,"output_power":0.00020280975432933711,"reference_active":true,"delta_db":-0.9545123304806274},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3625439652804924e-05,"output_power":2.052805977505284e-05,"reference_active":true,"delta_db":-0.6103199552915392},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.271613193724876e-07,"output_power":6.277534363460537e-07,"reference_active":true,"delta_db":0.004098336594938348}],"active_band_count":3,"active_band_mae_db":0.5229768741223683}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":0.00010594974849403162,"log_rms_mae_db":1.8546736829860493,"rms_envelope_nrmse":0.2392055243707849,"envelope_correlation":0.9744024736987159,"rms_gain_error_db":-1.5166668572584021,"reference_mean_power":0.0008097505272007746,"output_mean_power":0.0005710636984914713},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":1.235811152405605e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030835453412536774,"output_power":0.00025163888406139537,"reference_active":true,"delta_db":-0.882725882294146},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4464038776684894e-05,"output_power":2.9066781084572183e-05,"reference_active":true,"delta_db":-0.7396923049341252},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.768528595461299e-06,"output_power":3.006934655718235e-06,"reference_active":true,"delta_db":-3.523702775810651}],"active_band_count":3,"active_band_mae_db":1.7153736543463074}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":1.3209017638612706,"rms_envelope_nrmse":0.20071937346707996,"envelope_correlation":0.9880745427023335,"rms_gain_error_db":-1.516667318851313,"reference_mean_power":0.0008163242316416158,"output_mean_power":0.0005756996379532465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.2041398641164121e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028903835839107396,"output_power":0.0002365621190479359,"reference_active":true,"delta_db":-0.8701028016126824},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7474723603516325e-05,"output_power":1.5192324951744015e-05,"reference_active":true,"delta_db":-0.6078607422477342},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3608312124866867e-07,"output_power":1.5675130913680258e-07,"reference_active":true,"delta_db":-3.3123552529278912}],"active_band_count":3,"active_band_mae_db":1.596772932262769}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":1.9331791672383118,"rms_envelope_nrmse":0.2411265733817856,"envelope_correlation":0.9579310830362389,"rms_gain_error_db":-1.0617285368527198,"reference_mean_power":4.60838929150879e-05,"output_mean_power":3.608912106230003e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":2.8637727301813914e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3093489932613049e-05,"output_power":1.002279077326854e-05,"reference_active":true,"delta_db":-1.160667541132295},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.189795408821595e-06,"output_power":2.6752149175325377e-06,"reference_active":true,"delta_db":-0.7640415113699115},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2108923855177567e-06,"output_power":2.0014609060238606e-06,"reference_active":true,"delta_db":2.182415633519002}],"active_band_count":3,"active_band_mae_db":1.3690415620070695}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":1.5087145389118999,"rms_envelope_nrmse":0.16840604118561575,"envelope_correlation":0.9847086505234854,"rms_gain_error_db":-1.0617914524564795,"reference_mean_power":4.645555778084685e-05,"output_mean_power":3.637965143791777e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":2.5781119490074983e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2038087175825736e-05,"output_power":9.57560190029132e-06,"reference_active":true,"delta_db":-0.9939140176643108},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6192573703259313e-06,"output_power":1.549406853726825e-06,"reference_active":true,"delta_db":-0.1915040978004895},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2393464269378905e-07,"output_power":1.9147311967589853e-07,"reference_active":true,"delta_db":1.8891509428341196}],"active_band_count":3,"active_band_mae_db":1.0248563527663066}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":745,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":3.2469761993568182,"rms_envelope_nrmse":0.2714151730015224,"envelope_correlation":0.9573874878171488,"rms_gain_error_db":-0.9233275301875515,"reference_mean_power":0.0001249473972287349,"output_mean_power":0.00010101699862175852},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":745,"frequency_resolution_hz":0.2684563758389262,"power_floor":1.475095742378174e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0815424070071423e-05,"output_power":5.1848942665103415e-05,"reference_active":true,"delta_db":-0.6927383503314438},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1841381159756728e-05,"output_power":1.6423548608438837e-05,"reference_active":true,"delta_db":-1.2381309771218534},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.832473511911846e-06,"output_power":7.4986593818630485e-06,"reference_active":true,"delta_db":0.40405670013536776}],"active_band_count":3,"active_band_mae_db":0.7783086758628883}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59680,"uncovered_tail_samples":0,"metrics":{"frame_count":739,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":2.986422416023226,"rms_envelope_nrmse":0.2065035433452496,"envelope_correlation":0.9778841231732606,"rms_gain_error_db":-0.9232297320346604,"reference_mean_power":0.00012595573530693987,"output_mean_power":0.00010183450914426263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":739,"frequency_resolution_hz":0.2706359945872801,"power_floor":1.3755879245213433e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7009424723841386e-05,"output_power":4.755717446630112e-05,"reference_active":true,"delta_db":-0.7873061480701316},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1696758041681807e-05,"output_power":9.476939737383369e-06,"reference_active":true,"delta_db":-0.913973872033586},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.986552270441604e-07,"output_power":5.954485888280499e-07,"reference_active":true,"delta_db":-0.023325084666566798}],"active_band_count":3,"active_band_mae_db":0.5748683682567615}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.101590871810913,"predicted_mos":3.101590871810913,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":89880,"input_sample_rate":24000,"input_seconds":3.745,"input_float32_sha256":"90116650cae813e6d5295f2f08e96afe45075c4095b716617a4668c07a5b78c2","internal_samples":59920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.480335637999815},"started_at":"2026-10-13T07:42:22.072849+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004226212000503438,"native_input_samples":89760,"native_input_sha256":"eeb21de4b306024f8c565c16b70a6ef253064d47c4d6619d74f9bd9217ac3142","output_sha256":"1f56d2118a4f3c6c15a8ac8e5992edb8477cb8e683d145c8314fc3a7ea4c1de7","output_samples":89880,"length_delta_samples":120,"analysis_s":1.3897337130001688,"synthesis_s":9.14781466400018,"total_s":10.537548377000348,"rtf":2.817526304010788,"output_peak":0.2839796543121338,"output_rms":0.03815814384081016,"finished_at":"2026-10-13T07:42:32.626509+00:00","analysis_stages":{"external_f0_s":0.8867423110023083},"evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"1f56d2118a4f3c6c15a8ac8e5992edb8477cb8e683d145c8314fc3a7ea4c1de7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59840,"output_samples":59920,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.364136683922034,"output_peak":0.2828014667097028},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986648865153538,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.296369609059125,"spectral_convergence":0.5271283123099372,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.193358816795729,"spectral_convergence":0.3174606636486521,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.635816037156832,"spectral_convergence":0.32136495785283237,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.00036312364802186395,"delta_db":1.5544795650765808,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.0004597632398419646,"delta_db":-0.27215499647899266,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0004662983463979512,"delta_db":-0.6545438300687481,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.5205660490634561e-05,"delta_db":-0.11284638981806462,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":9.476242608351341e-05,"delta_db":-0.563570582559758,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":6.021930951670499e-05,"delta_db":-0.8987047093489763,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6760500122251867,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06996956049680604,"pitch":null,"output_rms_dbfs":-28.358338029362255,"output_peak":0.2828014667097028,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.708514821003895},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.8207136241958961,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03851071452419941,"rms_dbfs":-28.28836846886545,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.589436133538992,"mean":-9.952381970410678e-06,"zero_energy":false},"output":{"rms":0.03820173595993555,"rms_dbfs":-28.358338029362255,"peak":0.2828014667097028,"peak_dbfs":-10.97036684925109,"crest_factor_db":17.387971180111165,"mean":0.00010183723832071446,"zero_energy":false},"rms_difference_db":-0.06996956049680492},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":5.297397098612432e-07,"output_rms_dbfs":-62.75937470660366,"output_zero_energy":false,"output_peak":0.0027613187700574237,"output_energy_fraction":9.705652326679137e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":178,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9814814814814815,"absolute_difference_cents_median":6.353949995202698,"absolute_difference_cents_p95":19.97009855697347,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.000110539633783132,"log_rms_mae_db":3.023562830615044,"rms_envelope_nrmse":0.1435056890770882,"envelope_correlation":0.9787902450480852,"rms_gain_error_db":-0.06996653277974595,"reference_mean_power":0.00148505750597689,"output_mean_power":0.0014613243396611908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.0996976338580339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028388367321682665,"output_power":0.000253934135536112,"reference_active":true,"delta_db":-0.4841933041365506},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.612183760153141e-05,"output_power":4.082706705588123e-05,"reference_active":true,"delta_db":-0.529584201740566},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3216272322760606e-05,"output_power":7.826176648843622e-06,"reference_active":true,"delta_db":-2.275593323848922}],"active_band_count":3,"active_band_mae_db":1.096456943242013}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00010014294879668002,"log_rms_mae_db":2.830532123033179,"rms_envelope_nrmse":0.0929510783202257,"envelope_correlation":0.9906828780048916,"rms_gain_error_db":-0.06989821540733257,"reference_mean_power":0.0014970488171277764,"output_mean_power":0.0014731471873254266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.0259449632693504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002533237576934496,"output_power":0.00022671504044358175,"reference_active":true,"delta_db":-0.4819558903720538},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.399735365000864e-05,"output_power":2.0866894361656335e-05,"reference_active":true,"delta_db":-0.6070553432827062},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.21988770898296e-07,"output_power":2.889882026408283e-07,"reference_active":true,"delta_db":-3.3290243028287496}],"active_band_count":3,"active_band_mae_db":1.4726785121611699}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00010594974849403162,"log_rms_mae_db":1.7700968980169995,"rms_envelope_nrmse":0.1343822428487297,"envelope_correlation":0.9898228637863269,"rms_gain_error_db":-0.5733133765403908,"reference_mean_power":0.0008075829245111437,"output_mean_power":0.0007077107459342461},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2363198879746264e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030947682521494476,"output_power":0.00028529478692270425,"reference_active":true,"delta_db":-0.35334296933978815},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.465844834587485e-05,"output_power":3.10847399918755e-05,"reference_active":true,"delta_db":-0.47261876441167855},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.693327955984244e-06,"output_power":3.7041437887374615e-06,"reference_active":true,"delta_db":-2.569542679150097}],"active_band_count":3,"active_band_mae_db":1.131834804300521}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":1.2334533912439345,"rms_envelope_nrmse":0.10036428624867116,"envelope_correlation":0.9955329425592891,"rms_gain_error_db":-0.5733130733850009,"reference_mean_power":0.0008141211968758593,"output_mean_power":0.0007134404929241957},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.204812078547339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002902058778585621,"output_power":0.00026688430331270286,"reference_active":true,"delta_db":-0.3638317274164555},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7747321249654287e-05,"output_power":1.5943789663697123e-05,"reference_active":true,"delta_db":-0.46541254312072516},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.319074590753967e-07,"output_power":2.0435130607797828e-07,"reference_active":true,"delta_db":-2.1063959479527488}],"active_band_count":3,"active_band_mae_db":0.9785467394966432}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":1.6121971775381219,"rms_envelope_nrmse":0.2474486225371854,"envelope_correlation":0.947802941978366,"rms_gain_error_db":-0.13881980560200197,"reference_mean_power":4.5961113590202697e-05,"output_mean_power":4.451522389828545e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.8606659711815018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3098445296769507e-05,"output_power":1.259779249190466e-05,"reference_active":true,"delta_db":-0.16925300217431158},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.217362950062473e-06,"output_power":3.3899907767161534e-06,"reference_active":true,"delta_db":0.2269846026416902},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2063471779341695e-06,"output_power":1.056248912871346e-06,"reference_active":true,"delta_db":-0.5770603761566743}],"active_band_count":3,"active_band_mae_db":0.32443266032422535}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":0.8905286604775409,"rms_envelope_nrmse":0.13167872764177913,"envelope_correlation":0.98425188887365,"rms_gain_error_db":-0.13880138379623289,"reference_mean_power":4.6330782152537655e-05,"output_mean_power":4.4873453408007166e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.575546676565828e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2064714839118608e-05,"output_power":1.1543520542049238e-05,"reference_active":true,"delta_db":-0.1917878128704708},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6378421988809907e-06,"output_power":1.6637830204773828e-06,"reference_active":true,"delta_db":0.06824631361571777},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.209021145285469e-07,"output_power":1.0598121042299193e-07,"reference_active":true,"delta_db":-0.5720502122870434}],"active_band_count":3,"active_band_mae_db":0.277361446257744}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":1.679960593493253,"rms_envelope_nrmse":0.29084452270449884,"envelope_correlation":0.9467633194175799,"rms_gain_error_db":-0.9223928796821574,"reference_mean_power":0.00012461308874590145,"output_mean_power":0.00010076840232395202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.481947168400202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.119164755141344e-05,"output_power":4.955078115622692e-05,"reference_active":true,"delta_db":-0.91641641069823},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.192933297758675e-05,"output_power":1.6754709374058405e-05,"reference_active":true,"delta_db":-1.1688852298804033},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.860956874662954e-06,"output_power":4.030079963967422e-06,"reference_active":true,"delta_db":-2.310710261238898}],"active_band_count":3,"active_band_mae_db":1.4653373006058434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":1.1375761434301614,"rms_envelope_nrmse":0.18904793941614786,"envelope_correlation":0.9808765717191177,"rms_gain_error_db":-0.9223955896159729,"reference_mean_power":0.00012561601350824357,"output_mean_power":0.00010157935427520886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.3825123163038373e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7375597371987975e-05,"output_power":4.650939790516076e-05,"reference_active":true,"delta_db":-0.9118650295161241},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1774546568382313e-05,"output_power":7.954986428375415e-06,"reference_active":true,"delta_db":-1.7030474861601195},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.943011691356466e-07,"output_power":3.8296163101634516e-07,"reference_active":true,"delta_db":-1.9085132050396014}],"active_band_count":3,"active_band_mae_db":1.5078085735719482}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.7092888355255127,"predicted_mos":3.7092888355255127,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90000,"input_sample_rate":24000,"input_seconds":3.75,"input_float32_sha256":"32dc0e8f3852c18bb067bce0fb3df1d621db5daab9da8ec9f6cf9240b0d0c0a8","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.357006089998322},"started_at":"2026-10-13T05:11:09.901833+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00549186600073881,"native_input_samples":89760,"native_input_sha256":"eeb21de4b306024f8c565c16b70a6ef253064d47c4d6619d74f9bd9217ac3142","output_sha256":"7c5a915478afbefef19d4312ca2ef8b2ad94ccc328ef41e2ca2e0567dae072eb","output_samples":90000,"length_delta_samples":240,"analysis_s":0.9668827770001371,"synthesis_s":0.18560070300009102,"total_s":1.1524834800002282,"rtf":0.3081506631016653,"output_peak":0.2974832057952881,"output_rms":0.03783327997150036,"finished_at":"2026-10-13T05:11:11.070725+00:00","analysis_stages":{"external_f0_s":0.11228432099960628},"evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"7c5a915478afbefef19d4312ca2ef8b2ad94ccc328ef41e2ca2e0567dae072eb","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":59840,"output_samples":60000,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.43759466942265,"output_peak":0.29147716923308614},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973333333333333,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.647479426147129,"spectral_convergence":0.23187974044147572,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.294621946317612,"spectral_convergence":0.1987143244968344,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.987005549780839,"spectral_convergence":0.21832036428449536,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.00024384482203967302,"delta_db":-0.17493038336507752,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.0004738817256805659,"delta_db":-0.14079782846718447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0005352007609490905,"delta_db":-0.056015361145461746,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.5586364285039907e-05,"delta_db":-0.005451060212015284,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":9.617297716095023e-05,"delta_db":-0.4994016873495506,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":7.212581267492924e-05,"delta_db":-0.11515522868587338,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1652919248708605,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13763066740142313,"pitch":null,"output_rms_dbfs":-28.42599913626687,"output_peak":0.29147716923308614,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.6430356407485265},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.9002404117842203,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03851071452419941,"rms_dbfs":-28.28836846886545,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.589436133538992,"mean":-9.952381970410678e-06,"zero_energy":false},"output":{"rms":0.037905309163628366,"rms_dbfs":-28.42599913626687,"peak":0.29147716923308614,"peak_dbfs":-10.707909138126723,"crest_factor_db":17.71808999814015,"mean":-2.248588339099142e-06,"zero_energy":false},"rms_difference_db":-0.13763066740142094},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":1.260808608192678e-06,"output_rms_dbfs":-58.99350834691277,"output_zero_energy":false,"output_peak":0.003871327371825609,"output_energy_fraction":2.3462672398203587e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":153,"both_estimated_voiced_frames":150,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9259259259259259,"absolute_difference_cents_median":0.9651998149431692,"absolute_difference_cents_p95":5.7726496987024225,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.000110539633783132,"log_rms_mae_db":0.970749625241845,"rms_envelope_nrmse":0.05622158398472967,"envelope_correlation":0.9970094330026664,"rms_gain_error_db":-0.1376336512806573,"reference_mean_power":0.00148505750597689,"output_mean_power":0.001438732009466221},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.0996976338580339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028388367321682665,"output_power":0.00028033712143681224,"reference_active":true,"delta_db":-0.054598064032045965},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.612183760153141e-05,"output_power":4.267709856265692e-05,"reference_active":true,"delta_db":-0.33711716380381734},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3216272322760606e-05,"output_power":1.0689943035078958e-05,"reference_active":true,"delta_db":-0.9213358782240315}],"active_band_count":3,"active_band_mae_db":0.4376837020199649}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00010014294879668002,"log_rms_mae_db":0.6058886196643302,"rms_envelope_nrmse":0.030520596791736142,"envelope_correlation":0.9992815242076374,"rms_gain_error_db":-0.13763794687142558,"reference_mean_power":0.0014970488171277764,"output_mean_power":0.0014503478241755655},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.0259449632693504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002533237576934496,"output_power":0.00024968349038888656,"reference_active":true,"delta_db":-0.06286094724308279},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.399735365000864e-05,"output_power":2.212280614897205e-05,"reference_active":true,"delta_db":-0.3532313792174908},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.21988770898296e-07,"output_power":4.771339974050419e-07,"reference_active":true,"delta_db":-1.1514218160035532}],"active_band_count":3,"active_band_mae_db":0.5225047141547089}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00010594974849403162,"log_rms_mae_db":0.8195964740736117,"rms_envelope_nrmse":0.043344376726173105,"envelope_correlation":0.9988051196957701,"rms_gain_error_db":-0.10473692457376718,"reference_mean_power":0.0008075829245111437,"output_mean_power":0.0007883397680030575},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2363198879746264e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030947682521494476,"output_power":0.00030538548092779897,"reference_active":true,"delta_db":-0.05779747611160926},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.465844834587485e-05,"output_power":3.102194488901854e-05,"reference_active":true,"delta_db":-0.48140093528645456},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.693327955984244e-06,"output_power":6.144050739868356e-06,"reference_active":true,"delta_db":-0.3718731124136735}],"active_band_count":3,"active_band_mae_db":0.3036905079372458}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":0.45100475644041677,"rms_envelope_nrmse":0.025377986070918283,"envelope_correlation":0.9996483803153833,"rms_gain_error_db":-0.10473725845407378,"reference_mean_power":0.0008141211968758593,"output_mean_power":0.0007947221847451453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.204812078547339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002902058778585621,"output_power":0.0002859057296277725,"reference_active":true,"delta_db":-0.06483345610696663},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7747321249654287e-05,"output_power":1.5503070286831524e-05,"reference_active":true,"delta_db":-0.5871509465631277},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.319074590753967e-07,"output_power":2.8745309073951877e-07,"reference_active":true,"delta_db":-0.6245002996629367}],"active_band_count":3,"active_band_mae_db":0.4254949007776771}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":0.9496859010430636,"rms_envelope_nrmse":0.13087820100113987,"envelope_correlation":0.9868966476707365,"rms_gain_error_db":-0.42705787328349076,"reference_mean_power":4.5961113590202697e-05,"output_mean_power":4.165668970791072e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.8606659711815018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3098445296769507e-05,"output_power":1.2092521282930362e-05,"reference_active":true,"delta_db":-0.3470289026325203},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.217362950062473e-06,"output_power":2.939782029668259e-06,"reference_active":true,"delta_db":-0.39184925525139935},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2063471779341695e-06,"output_power":6.281742151014313e-07,"reference_active":true,"delta_db":-2.833922067997867}],"active_band_count":3,"active_band_mae_db":1.1909334086272623}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":0.6343644662723451,"rms_envelope_nrmse":0.08256625843050223,"envelope_correlation":0.9957205105727026,"rms_gain_error_db":-0.4270498484947658,"reference_mean_power":4.6330782152537655e-05,"output_mean_power":4.199181506888565e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.575546676565828e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2064714839118608e-05,"output_power":1.1150289792682906e-05,"reference_active":true,"delta_db":-0.3423090669359625},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6378421988809907e-06,"output_power":1.4950652492803344e-06,"reference_active":true,"delta_db":-0.3961190945530281},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.209021145285469e-07,"output_power":5.385012872615146e-08,"reference_active":true,"delta_db":-3.512471507647201}],"active_band_count":3,"active_band_mae_db":1.4169665563787304}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":1.0663790005975267,"rms_envelope_nrmse":0.14897448963810273,"envelope_correlation":0.9860020687419744,"rms_gain_error_db":-0.23862694615013277,"reference_mean_power":0.00012461308874590145,"output_mean_power":0.00011795082048644228},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.481947168400202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.119164755141344e-05,"output_power":6.137700991030354e-05,"reference_active":true,"delta_db":0.013135806927732266},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.192933297758675e-05,"output_power":2.1281529491194605e-05,"reference_active":true,"delta_db":-0.13022584747438365},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.860956874662954e-06,"output_power":4.855978528957127e-06,"reference_active":true,"delta_db":-1.5010793171831174}],"active_band_count":3,"active_band_mae_db":0.5481469905284111}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":0.7872854115078517,"rms_envelope_nrmse":0.07812282508399931,"envelope_correlation":0.9964082949214796,"rms_gain_error_db":-0.23863537434495777,"reference_mean_power":0.00012561601350824357,"output_mean_power":0.0001188998945039284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.3825123163038373e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7375597371987975e-05,"output_power":5.771484245762681e-05,"reference_active":true,"delta_db":0.025602942456149545},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1774546568382313e-05,"output_power":1.088708981188347e-05,"reference_active":true,"delta_db":-0.3403238606915747},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.943011691356466e-07,"output_power":3.895321359341221e-07,"reference_active":true,"delta_db":-1.8346329236186756}],"active_band_count":3,"active_band_mae_db":0.7335199089221333}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0044","method":"magphase","status":"measured","mos":{"status":"measured","value":3.5663561820983887,"predicted_mos":3.5663561820983887,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":180454,"input_sample_rate":48000,"input_seconds":3.7594583333333333,"input_float32_sha256":"195c2a1c7b18fe215a82bf7f3185d1a30d4c7b7a18349c2fa8818a017dbdc991","internal_samples":60152,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1237184410001646},"started_at":"2026-10-13T05:17:13.514388+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.007070771000144305,"native_input_samples":179520,"native_input_sha256":"d79f6a5a0208b95da69cb919adfd1a08c1d4efe49bf31f168fb9d1767820e642","output_sha256":"9af06fe95c285d3858ea17b9271532c81c7b459b8fc7dd59ffc75d4d3f158b7d","output_samples":180454,"length_delta_samples":934,"analysis_s":1.9314585909996822,"synthesis_s":0.7258675649991346,"total_s":2.657326155998817,"rtf":0.7105150149729457,"output_peak":0.3204064660907524,"output_rms":0.0375774325442071,"finished_at":"2026-10-13T05:17:16.188157+00:00","evaluation_input":{"reference":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"9af06fe95c285d3858ea17b9271532c81c7b459b8fc7dd59ffc75d4d3f158b7d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":59840,"output_samples":60152,"length_delta_samples":312,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.496804195048313,"output_peak":0.3182501551110875},"common_prefix":{"samples":59840,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9948131400452188,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":59840,"output_samples":59840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.059834450381265,"spectral_convergence":0.5731016888464379,"spectral_convergence_status":"ok","amplitude_floor":9.91771656423115e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.055275580240266,"spectral_convergence":0.3373561170269769,"spectral_convergence_status":"ok","amplitude_floor":5.25712991724442e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.88978647540088,"spectral_convergence":0.29385296871366884,"spectral_convergence_status":"ok","amplitude_floor":4.1776466342690114e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00025386719019706026,"output_power":0.0002245081509532904,"delta_db":-0.5337446322476754,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004894966727833924,"output_power":0.00045282932818941854,"delta_db":-0.3381519731136872,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005421484972995941,"output_power":0.0005091795698256288,"delta_db":-0.2724728853065304,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.5605939839496278e-05,"output_power":1.6315267264123832e-05,"delta_db":0.19304264070891805,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000107892990107672,"output_power":0.00013653936847264752,"delta_db":1.0226466065825959,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.406384293716839e-05,"output_power":8.15685952017338e-05,"delta_db":0.4191674036767595,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4632043569393611,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.18587327580030588,"pitch":null,"output_rms_dbfs":-28.474241744665754,"output_peak":0.3182501551110875,"reference_rms_dbfs":-28.28836846886545,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.334965502007471},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":59840,"output_samples":59840,"intelligibility":{"estoi":{"value":0.767417290787036,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03851071452419941,"rms_dbfs":-28.28836846886545,"peak":0.291778564453125,"peak_dbfs":-10.698932335326457,"crest_factor_db":17.589436133538992,"mean":-9.952381970410678e-06,"zero_energy":false},"output":{"rms":0.03769536151712625,"rms_dbfs":-28.474241744665754,"peak":0.3182501551110875,"peak_dbfs":-9.944627520092709,"crest_factor_db":18.529614224573045,"mean":0.0011500150675202378,"zero_energy":false},"rms_difference_db":-0.18587327580030433},"full_scale":{"reference":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":59840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":374,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2219010636908935e-06,"threshold_dbfs":-59.12963957131581,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":10,"quiet_samples":1600,"reference_mean_power":8.768751285970211e-07,"output_mean_power":4.627734710732266e-06,"output_rms_dbfs":-53.34631545348505,"output_zero_energy":false,"output_peak":0.008728062606481029,"output_energy_fraction":8.708052263829472e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":367,"reference_estimated_voiced_frames":162,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":148,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9135802469135802,"absolute_difference_cents_median":5.778760607057769,"absolute_difference_cents_p95":14.248966666489162,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":59840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.000110539633783132,"log_rms_mae_db":2.0042840038271352,"rms_envelope_nrmse":0.1819596485872446,"envelope_correlation":0.9665086410183983,"rms_gain_error_db":-0.18587403224982113,"reference_mean_power":0.00148505750597689,"output_mean_power":0.0014228393514828983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.0996976338580339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028388367321682665,"output_power":0.00027239895309145594,"reference_active":true,"delta_db":-0.1793498187612303},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.612183760153141e-05,"output_power":5.7056713188765325e-05,"reference_active":true,"delta_db":0.9240014778486221},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3216272322760606e-05,"output_power":2.1284785939002787e-05,"reference_active":true,"delta_db":2.06960308082864}],"active_band_count":3,"active_band_mae_db":1.0576514591461643}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":0.00010014294879668002,"log_rms_mae_db":1.315747266786006,"rms_envelope_nrmse":0.11407874969152061,"envelope_correlation":0.9863101377108585,"rms_gain_error_db":-0.1858291616000762,"reference_mean_power":0.0014970488171277764,"output_mean_power":0.0014343430925057702},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.0259449632693504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002533237576934496,"output_power":0.00023916711057687782,"reference_active":true,"delta_db":-0.24974464769988794},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.399735365000864e-05,"output_power":3.211290449310001e-05,"reference_active":true,"delta_db":1.2651623589667615},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.21988770898296e-07,"output_power":1.3646207771613076e-06,"reference_active":true,"delta_db":3.4122943517847855}],"active_band_count":3,"active_band_mae_db":1.642400452817145}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":0.00010594974849403162,"log_rms_mae_db":2.1210660755098956,"rms_envelope_nrmse":0.15269307436103616,"envelope_correlation":0.9844132541668049,"rms_gain_error_db":-0.2988611376901073,"reference_mean_power":0.0008075829245111437,"output_mean_power":0.0007538779030983911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.2363198879746264e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030947682521494476,"output_power":0.0002963025278757534,"reference_active":true,"delta_db":-0.1889277633821673},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.465844834587485e-05,"output_power":3.0319911865353234e-05,"reference_active":true,"delta_db":-0.5808118096266213},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.693327955984244e-06,"output_power":3.8019618995189578e-06,"reference_active":true,"delta_db":-2.4563434447623504}],"active_band_count":3,"active_band_mae_db":1.075361005923713}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":9.562043532359368e-05,"log_rms_mae_db":1.500358407973993,"rms_envelope_nrmse":0.1048263004220043,"envelope_correlation":0.992907257379783,"rms_gain_error_db":-0.2988596154357972,"reference_mean_power":0.0008141211968758593,"output_mean_power":0.0007599816406052939},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.204812078547339e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002902058778585621,"output_power":0.00027727492658487425,"reference_active":true,"delta_db":-0.1979560545916451},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7747321249654287e-05,"output_power":1.5523927497358207e-05,"reference_active":true,"delta_db":-0.5813120491828077},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.319074590753967e-07,"output_power":2.5004929482493584e-07,"reference_active":true,"delta_db":-1.2299137846161046}],"active_band_count":3,"active_band_mae_db":0.6697272961301858}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":2.870494430586826e-05,"log_rms_mae_db":1.9060669711130664,"rms_envelope_nrmse":0.28124360545469945,"envelope_correlation":0.93450208422315,"rms_gain_error_db":0.06517827819906587,"reference_mean_power":4.5961113590202697e-05,"output_mean_power":4.665609326681323e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":2.8606659711815018e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3098445296769507e-05,"output_power":1.3935182921081668e-05,"reference_active":true,"delta_db":0.2689292319535766},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.217362950062473e-06,"output_power":3.2511496934642904e-06,"reference_active":true,"delta_db":0.04536909943264769},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2063471779341695e-06,"output_power":9.958366226786821e-07,"reference_active":true,"delta_db":-0.8328421882656984}],"active_band_count":3,"active_band_mae_db":0.38238017321730755}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":2.131135367573917e-05,"log_rms_mae_db":1.1707019867997246,"rms_envelope_nrmse":0.15767407223191948,"envelope_correlation":0.9778754725095739,"rms_gain_error_db":0.06518746708393561,"reference_mean_power":4.6330782152537655e-05,"output_mean_power":4.7031451110905044e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":2.575546676565828e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2064714839118608e-05,"output_power":1.2800716014344408e-05,"reference_active":true,"delta_db":0.25717201395083483},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6378421988809907e-06,"output_power":1.6957814005399748e-06,"reference_active":true,"delta_db":0.15097811093556246},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.209021145285469e-07,"output_power":9.50873090681348e-08,"reference_active":true,"delta_db":-1.043113393437075}],"active_band_count":3,"active_band_mae_db":0.4837545061078241}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":747,"rms_floor":6.327125594838895e-05,"log_rms_mae_db":2.04956774730751,"rms_envelope_nrmse":0.4716194560756199,"envelope_correlation":0.886203917291809,"rms_gain_error_db":1.064939981091151,"reference_mean_power":0.00012461308874590145,"output_mean_power":0.0001592420130226421},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":747,"frequency_resolution_hz":0.2677376171352075,"power_floor":1.481947168400202e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.119164755141344e-05,"output_power":6.739104381873967e-05,"reference_active":true,"delta_db":0.4191003670985842},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.192933297758675e-05,"output_power":3.4821403579486884e-05,"reference_active":true,"delta_db":2.0082085067063207},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.860956874662954e-06,"output_power":1.6674933697910207e-05,"reference_active":true,"delta_db":3.8567942624380427}],"active_band_count":3,"active_band_mae_db":2.094701045414316}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":59840,"uncovered_tail_samples":0,"metrics":{"frame_count":741,"rms_floor":5.46294552803286e-05,"log_rms_mae_db":1.367597749448224,"rms_envelope_nrmse":0.2941516054229143,"envelope_correlation":0.9596116328052583,"rms_gain_error_db":1.0649859069269767,"reference_mean_power":0.00012561601350824357,"output_mean_power":0.0001605253396175914},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":741,"frequency_resolution_hz":0.2699055330634278,"power_floor":1.3825123163038373e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.7375597371987975e-05,"output_power":6.479868252983305e-05,"reference_active":true,"delta_db":0.5283895572557937},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1774546568382313e-05,"output_power":2.0364740303104914e-05,"reference_active":true,"delta_db":2.379346845419161},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.943011691356466e-07,"output_power":1.4741707643328332e-06,"reference_active":true,"delta_db":3.9454120942428403}],"active_band_count":3,"active_band_mae_db":2.2843828323059316}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"hnm","status":"measured","mos":{"status":"measured","value":2.4533474445343018,"predicted_mos":2.4533474445343018,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83600,"input_sample_rate":16000,"input_seconds":5.225,"input_float32_sha256":"4c310820c8d1a99d6e31e4329277489f420a42350a067b21ddbefcdd2927ec57","internal_samples":83600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.045679270000619},"started_at":"2026-10-13T04:39:51.848400+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4227000065147877e-05,"native_input_samples":83600,"native_input_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","output_sha256":"70c0aee9f78939b5bf1c0f7e4606c8a737137eac1182846827068a59927e69c6","output_samples":83600,"length_delta_samples":0,"analysis_s":1.0573853539999618,"synthesis_s":0.4101858320000247,"total_s":1.4675711859999865,"rtf":0.2808748681339687,"output_peak":0.31592758749225236,"output_rms":0.03350378592273226,"finished_at":"2026-10-13T04:39:53.322206+00:00","evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"70c0aee9f78939b5bf1c0f7e4606c8a737137eac1182846827068a59927e69c6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.635394729659678,"spectral_convergence":0.35057431520021637,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.265443154467399,"spectral_convergence":0.33570483182655103,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.6554230687599665,"spectral_convergence":0.35592825203337114,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00021217881230552394,"delta_db":0.02748646312504046,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.0004708707052672988,"delta_db":-0.049073912603191217,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":9.446294937102097e-05,"delta_db":-0.06465298948274015,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":6.33486750101576e-05,"delta_db":-0.6815190922599033,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.0001233858336397803,"delta_db":-0.3022264245580997,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.0001582566955624909,"delta_db":0.28833527747717874,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23554902658435894,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05734119406776289,"pitch":null,"output_rms_dbfs":-29.498122299591124,"output_peak":0.31592758749225236,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.518753650962348},"common_prefix":{"samples":83600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.635394729659678,"spectral_convergence":0.35057431520021637,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.265443154467399,"spectral_convergence":0.33570483182655103,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.6554230687599665,"spectral_convergence":0.35592825203337114,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00021217881230552394,"delta_db":0.02748646312504046,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.0004708707052672988,"delta_db":-0.049073912603191217,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":9.446294937102097e-05,"delta_db":-0.06465298948274015,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":6.33486750101576e-05,"delta_db":-0.6815190922599033,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.0001233858336397803,"delta_db":-0.3022264245580997,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.0001582566955624909,"delta_db":0.28833527747717874,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23554902658435894,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05734119406776289,"pitch":null,"output_rms_dbfs":-29.498122299591124,"output_peak":0.31592758749225236,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.518753650962348},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83600,"output_samples":83600,"intelligibility":{"estoi":{"value":0.8406397644982715,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.033725697841694866,"rms_dbfs":-29.44078110552336,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.117693772983436,"mean":-5.154404343600478e-06,"zero_energy":false},"output":{"rms":0.03350378592273226,"rms_dbfs":-29.498122299591124,"peak":0.31592758749225236,"peak_dbfs":-10.008248977779788,"crest_factor_db":19.489873321811334,"mean":-0.0001509686170198807,"zero_energy":false},"rms_difference_db":-0.05734119406776372},"full_scale":{"reference":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":1.4487330821969302e-06,"output_rms_dbfs":-58.39011622539249,"output_zero_energy":false,"output_peak":0.0038683142365259636,"output_energy_fraction":1.2350493657186427e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":163,"both_estimated_voiced_frames":157,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.8820224719101124,"absolute_difference_cents_median":1.0940136532716362,"absolute_difference_cents_p95":8.141018762403062,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011323786448874551,"log_rms_mae_db":1.0179722814247536,"rms_envelope_nrmse":0.07992713591685549,"envelope_correlation":0.9926179328482382,"rms_gain_error_db":-0.057414841845520134,"reference_mean_power":0.0011384943575262435,"output_mean_power":0.0011235422240711136},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":5.62759551802579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002928658523286973,"output_power":0.0002913476852378085,"reference_active":true,"delta_db":-0.02257164656214845},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4610498849864094e-05,"output_power":4.4289766712185035e-05,"reference_active":true,"delta_db":-0.031336869700298677},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2101686887119984e-05,"output_power":1.0148791243376034e-05,"reference_active":true,"delta_db":-0.7643159269393971}],"active_band_count":3,"active_band_mae_db":0.2727414810672814}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":0.5716956963338626,"rms_envelope_nrmse":0.03892116863302404,"envelope_correlation":0.9981591359489046,"rms_gain_error_db":-0.057482840876672325,"reference_mean_power":0.0011449801518647084,"output_mean_power":0.0011299251470589442},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.095010942049215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002727818973842863,"output_power":0.00027243468887422774,"reference_active":true,"delta_db":-0.005531407670752435},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1267992599082604e-05,"output_power":2.0792188533287256e-05,"reference_active":true,"delta_db":-0.09826296036228827},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.757168583210849e-07,"output_power":4.435111041422231e-07,"reference_active":true,"delta_db":-1.1330444909770534}],"active_band_count":3,"active_band_mae_db":0.41227961967003135}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":1.4295202768146749,"rms_envelope_nrmse":0.08097189624611402,"envelope_correlation":0.9946324170066806,"rms_gain_error_db":-0.09897037735586417,"reference_mean_power":0.00038382289364078325,"output_mean_power":0.0003751749536214976},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":2.8601579771498626e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014700129208541836,"output_power":0.0001458900119348845,"reference_active":true,"delta_db":-0.032955922318348374},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.957621157356578e-05,"output_power":1.851765826059288e-05,"reference_active":true,"delta_db":-0.2414258506405495},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3113428408893907e-06,"output_power":2.5059803347030863e-06,"reference_active":true,"delta_db":0.3511328956601822}],"active_band_count":3,"active_band_mae_db":0.20850488953969334}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":0.8278661884717204,"rms_envelope_nrmse":0.043260276090788655,"envelope_correlation":0.9984546039753156,"rms_gain_error_db":-0.0989687545119327,"reference_mean_power":0.00038603855676420007,"output_mean_power":0.0003773408364889599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":2.6676069300397395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013762626674571402,"output_power":0.000137311071222278,"reference_active":true,"delta_db":-0.009957740736915527},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.746730510582487e-06,"output_power":9.492731055454759e-06,"reference_active":true,"delta_db":-0.11467781553376341},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.594967364659404e-07,"output_power":1.5995030168950619e-07,"reference_active":true,"delta_db":0.012332623729046189}],"active_band_count":3,"active_band_mae_db":0.04565605999990838}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":1.3887585488523662,"rms_envelope_nrmse":0.15797233745553263,"envelope_correlation":0.981094770087102,"rms_gain_error_db":-0.42397312452974356,"reference_mean_power":0.00014337292372884635,"output_mean_power":0.0001300378646136256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.145508332328126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5890331523222755e-05,"output_power":4.596186811452495e-05,"reference_active":true,"delta_db":-0.8493901349555895},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4083924473542483e-05,"output_power":1.335477153981532e-05,"reference_active":true,"delta_db":-0.23087224514948404},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.499188380560451e-06,"output_power":3.901833700714853e-06,"reference_active":true,"delta_db":-1.490298416837399}],"active_band_count":3,"active_band_mae_db":0.8568535989808241}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":0.8113124058000265,"rms_envelope_nrmse":0.09057800875808067,"envelope_correlation":0.9947725686400793,"rms_gain_error_db":-0.4238962146155979,"reference_mean_power":0.00014418079784839883,"output_mean_power":0.00013077291453804696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.0427751966205221e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.204944813792335e-05,"output_power":4.326307254284574e-05,"reference_active":true,"delta_db":-0.802987694118153},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.197217517258329e-06,"output_power":5.790093803849958e-06,"reference_active":true,"delta_db":-0.2951114011791708},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536643030241646e-07,"output_power":2.3831665158943655e-07,"reference_active":true,"delta_db":-0.271049670161934}],"active_band_count":3,"active_band_mae_db":0.4563829218197526}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011141137328067573,"log_rms_mae_db":1.5475107903846075,"rms_envelope_nrmse":0.13829708841724672,"envelope_correlation":0.9891363948297595,"rms_gain_error_db":0.06956051385093638,"reference_mean_power":0.00018590462562984242,"output_mean_power":0.00018890621538280935},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.8616455925800917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.396584441083371e-05,"output_power":9.22753107561022e-05,"reference_active":true,"delta_db":-0.0788450439284531},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7617019162087808e-05,"output_power":1.6992927372313537e-05,"reference_active":true,"delta_db":-0.15664225479617944},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489954892102077e-06,"output_power":5.577639492851789e-06,"reference_active":true,"delta_db":-0.6579123779067536}],"active_band_count":3,"active_band_mae_db":0.29779989221046205}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":1.00772132445085,"rms_envelope_nrmse":0.07472542389048181,"envelope_correlation":0.9967730055452301,"rms_gain_error_db":0.06955990973993761,"reference_mean_power":0.00018697687803543296,"output_mean_power":0.00018999575379512757},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.7928912528322635e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.176227386223459e-05,"output_power":9.072325018688338e-05,"reference_active":true,"delta_db":-0.049455662571560664},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.193273720945635e-06,"output_power":7.580257865774273e-06,"reference_active":true,"delta_db":-0.3377348426959612},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7870970633055527e-07,"output_power":2.8725610878222975e-07,"reference_active":true,"delta_db":-1.2003716307630343}],"active_band_count":3,"active_band_mae_db":0.5291873786768521}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.3703770637512207,"predicted_mos":3.3703770637512207,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83600,"input_sample_rate":16000,"input_seconds":5.225,"input_float32_sha256":"dddade6cfe2bbd38d4a337a85200ade2bf78778625717bdd4a5df99f76ece897","internal_samples":83600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.18559086499954},"started_at":"2026-10-13T04:43:03.051143+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.388000019796891e-05,"native_input_samples":83600,"native_input_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","output_sha256":"0584c3c0eac93240ee93d2d13ed256cdd60fc5215c269251e7b7d9ade34b40b6","output_samples":83600,"length_delta_samples":0,"analysis_s":1.3943020970000362,"synthesis_s":0.07369786300023407,"total_s":1.4679999600002702,"rtf":0.2809569301435924,"output_peak":0.4622516378628635,"output_rms":0.03635986489233189,"finished_at":"2026-10-13T04:43:04.525994+00:00","evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"0584c3c0eac93240ee93d2d13ed256cdd60fc5215c269251e7b7d9ade34b40b6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.946337796252332,"spectral_convergence":0.692054547031791,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.209153756112912,"spectral_convergence":0.39822150036024506,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.771653839810262,"spectral_convergence":0.40872396099737895,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00019960844485897734,"delta_db":-0.23774456667024302,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.0005887264647500712,"delta_db":0.9210449552853016,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":0.0001195223155549338,"delta_db":0.957221981508968,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":8.466858562497969e-05,"delta_db":0.5783286010043839,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.00014175562840353974,"delta_db":0.300523694154602,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.00018775833579612731,"delta_db":1.0307066165407879,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6709284025273811,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6532263211513603,"pitch":null,"output_rms_dbfs":-28.787554784371995,"output_peak":0.4622516378628635,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9757151307251695},"common_prefix":{"samples":83600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.946337796252332,"spectral_convergence":0.692054547031791,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.209153756112912,"spectral_convergence":0.39822150036024506,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.771653839810262,"spectral_convergence":0.40872396099737895,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00019960844485897734,"delta_db":-0.23774456667024302,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.0005887264647500712,"delta_db":0.9210449552853016,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":0.0001195223155549338,"delta_db":0.957221981508968,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":8.466858562497969e-05,"delta_db":0.5783286010043839,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.00014175562840353974,"delta_db":0.300523694154602,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.00018775833579612731,"delta_db":1.0307066165407879,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6709284025273811,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6532263211513603,"pitch":null,"output_rms_dbfs":-28.787554784371995,"output_peak":0.4622516378628635,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9757151307251695},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83600,"output_samples":83600,"intelligibility":{"estoi":{"value":0.8798319955359589,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.033725697841694866,"rms_dbfs":-29.44078110552336,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.117693772983436,"mean":-5.154404343600478e-06,"zero_energy":false},"output":{"rms":0.03635986489233189,"rms_dbfs":-28.787554784371995,"peak":0.4622516378628635,"peak_dbfs":-6.70243082723545,"crest_factor_db":22.08512395713655,"mean":-1.0475228714301307e-08,"zero_energy":false},"rms_difference_db":0.6532263211513651},"full_scale":{"reference":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":7.104044353423196e-07,"output_rms_dbfs":-61.48494335744825,"output_zero_energy":false,"output_peak":0.0031588430678449555,"output_energy_fraction":5.142151315324502e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":211,"both_estimated_voiced_frames":175,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.9831460674157303,"absolute_difference_cents_median":4.679896909989188,"absolute_difference_cents_p95":17.368831806865064,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011323786448874551,"log_rms_mae_db":2.297147121393734,"rms_envelope_nrmse":0.18028056714454727,"envelope_correlation":0.9737303788937444,"rms_gain_error_db":0.6532791487441901,"reference_mean_power":0.0011384943575262435,"output_mean_power":0.0013233014775965682},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":5.62759551802579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002928658523286973,"output_power":0.00035995140866248434,"reference_active":true,"delta_db":0.8957514088673978},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4610498849864094e-05,"output_power":5.24198342250269e-05,"reference_active":true,"delta_db":0.7005856348097519},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2101686887119984e-05,"output_power":8.687611129513316e-06,"reference_active":true,"delta_db":-1.4394553901932041}],"active_band_count":3,"active_band_mae_db":1.011930811290118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":1.9599567953916952,"rms_envelope_nrmse":0.12056288219222712,"envelope_correlation":0.9919210650169661,"rms_gain_error_db":0.6535243688582268,"reference_mean_power":0.0011449801518647084,"output_mean_power":0.001330915231031508},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.095010942049215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002727818973842863,"output_power":0.00033988767239881824,"reference_active":true,"delta_db":0.9551986729628374},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1267992599082604e-05,"output_power":2.5099261330608352e-05,"reference_active":true,"delta_db":0.7193443990013992},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.757168583210849e-07,"output_power":5.362734344882204e-07,"reference_active":true,"delta_db":-0.30822663047046406}],"active_band_count":3,"active_band_mae_db":0.6609232341449003}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":1.8707251731854866,"rms_envelope_nrmse":0.19856612918097394,"envelope_correlation":0.9874242814802875,"rms_gain_error_db":1.1960526443155253,"reference_mean_power":0.00038382289364078325,"output_mean_power":0.0005055174360985351},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":2.8601579771498626e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014700129208541836,"output_power":0.00019109534271438144,"reference_active":true,"delta_db":1.139289507154004},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.957621157356578e-05,"output_power":2.418114831772644e-05,"reference_active":true,"delta_db":0.9174827076318836},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3113428408893907e-06,"output_power":1.6178848702554378e-06,"reference_active":true,"delta_db":-1.5491675534936444}],"active_band_count":3,"active_band_mae_db":1.201979922759844}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":1.4257727400741382,"rms_envelope_nrmse":0.17426298000440069,"envelope_correlation":0.993580418552747,"rms_gain_error_db":1.196024351938569,"reference_mean_power":0.00038603855676420007,"output_mean_power":0.000508432283161248},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":2.6676069300397395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013762626674571402,"output_power":0.00017853710279087882,"reference_active":true,"delta_db":1.1302715363498645},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.746730510582487e-06,"output_power":1.2038954303373593e-05,"reference_active":true,"delta_db":0.91729807661329},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.594967364659404e-07,"output_power":1.0390972468327784e-07,"reference_active":true,"delta_db":-1.8609560706302917}],"active_band_count":3,"active_band_mae_db":1.3028418945311488}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":1.4802314047123766,"rms_envelope_nrmse":0.18586033949696354,"envelope_correlation":0.9764694420339868,"rms_gain_error_db":0.4851189844945004,"reference_mean_power":0.00014337292372884635,"output_mean_power":0.00016031680168761948},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.145508332328126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5890331523222755e-05,"output_power":6.302779822069078e-05,"reference_active":true,"delta_db":0.5219545026322594},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4083924473542483e-05,"output_power":1.3612574933536232e-05,"reference_active":true,"delta_db":-0.14783404012827195},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.499188380560451e-06,"output_power":3.6020523696030588e-06,"reference_active":true,"delta_db":-1.8374857449700042}],"active_band_count":3,"active_band_mae_db":0.8357580959101786}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":0.8574513421307287,"rms_envelope_nrmse":0.10960471539384087,"envelope_correlation":0.9929796453829773,"rms_gain_error_db":0.4852446586169713,"reference_mean_power":0.00014418079784839883,"output_mean_power":0.0001612248161230057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.0427751966205221e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.204944813792335e-05,"output_power":5.841892166548121e-05,"reference_active":true,"delta_db":0.5013740702239058},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.197217517258329e-06,"output_power":5.774168991134307e-06,"reference_active":true,"delta_db":-0.30707249680984544},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536643030241646e-07,"output_power":1.80531667339434e-07,"reference_active":true,"delta_db":-1.4770596212455014}],"active_band_count":3,"active_band_mae_db":0.7618353960930841}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011141137328067573,"log_rms_mae_db":1.5425383858567419,"rms_envelope_nrmse":0.2821821090170228,"envelope_correlation":0.9641929925583786,"rms_gain_error_db":0.8448860626345793,"reference_mean_power":0.00018590462562984242,"output_mean_power":0.00022582852719889537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.8616455925800917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.396584441083371e-05,"output_power":0.00011136619183908446,"reference_active":true,"delta_db":0.7378334839932477},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7617019162087808e-05,"output_power":2.4980893296474045e-05,"reference_active":true,"delta_db":1.5167553761171813},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489954892102077e-06,"output_power":4.043249555195587e-06,"reference_active":true,"delta_db":-2.0551113083986547}],"active_band_count":3,"active_band_mae_db":1.4365667228363612}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":0.820254777871506,"rms_envelope_nrmse":0.17927815115761755,"envelope_correlation":0.988596626730988,"rms_gain_error_db":0.8449031345210557,"reference_mean_power":0.00018697687803543296,"output_mean_power":0.00022713194374466257},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.7928912528322635e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.176227386223459e-05,"output_power":0.00011003163487355633,"reference_active":true,"delta_db":0.7885339903642818},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.193273720945635e-06,"output_power":1.0677014402113365e-05,"reference_active":true,"delta_db":1.1499236452436075},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7870970633055527e-07,"output_power":2.3621802216071852e-07,"reference_active":true,"delta_db":-2.0499340736440024}],"active_band_count":3,"active_band_mae_db":1.329463903083964}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"facodec","status":"measured","mos":{"status":"measured","value":3.872835397720337,"predicted_mos":3.872835397720337,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83600,"input_sample_rate":16000,"input_seconds":5.225,"input_float32_sha256":"4e7fdacb878841f9f89d3aadf6d7e01ba91834f3b284223f77e8e942a1dabbd0","internal_samples":83600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4296508699990227},"started_at":"2026-10-13T06:57:04.399785+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.9023000024608336e-05,"native_input_samples":83600,"native_input_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","output_sha256":"653563a6e5ca99ea044e5e08fba0bb082307ab8dd00bf0258df599855205f544","output_samples":83600,"length_delta_samples":0,"analysis_s":1.4316057800006092,"synthesis_s":12.838803111999368,"total_s":14.270408891999978,"rtf":2.731178735311001,"output_peak":0.294864296913147,"output_rms":0.03322467776216561,"finished_at":"2026-10-13T06:57:18.682969+00:00","evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"653563a6e5ca99ea044e5e08fba0bb082307ab8dd00bf0258df599855205f544","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.413302222031099,"spectral_convergence":0.33685566456565585,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.832131969445029,"spectral_convergence":0.3729885379586538,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.012778944286095,"spectral_convergence":0.38279286013137914,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00021204551722339422,"delta_db":0.024757278752086165,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.000522402098906094,"delta_db":0.4019584981540902,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":9.949846041077166e-05,"delta_db":0.16089560575529105,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":5.865160386096233e-05,"delta_db":-1.0160955227296087,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":9.634667450255711e-05,"delta_db":-1.3765121298138905,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.00011493485749596269,"delta_db":-1.100768038724698,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6801645123216108,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1300033440424893,"pitch":null,"output_rms_dbfs":-29.57078444956585,"output_peak":0.294864296913147,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.752737711920741},"common_prefix":{"samples":83600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.413302222031099,"spectral_convergence":0.33685566456565585,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.832131969445029,"spectral_convergence":0.3729885379586538,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.012778944286095,"spectral_convergence":0.38279286013137914,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00021204551722339422,"delta_db":0.024757278752086165,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.000522402098906094,"delta_db":0.4019584981540902,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":9.949846041077166e-05,"delta_db":0.16089560575529105,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":5.865160386096233e-05,"delta_db":-1.0160955227296087,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":9.634667450255711e-05,"delta_db":-1.3765121298138905,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.00011493485749596269,"delta_db":-1.100768038724698,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6801645123216108,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1300033440424893,"pitch":null,"output_rms_dbfs":-29.57078444956585,"output_peak":0.294864296913147,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.752737711920741},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83600,"output_samples":83600,"intelligibility":{"estoi":{"value":0.8273503942363007,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.033725697841694866,"rms_dbfs":-29.44078110552336,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.117693772983436,"mean":-5.154404343600478e-06,"zero_energy":false},"output":{"rms":0.03322467776216561,"rms_dbfs":-29.57078444956585,"peak":0.294864296913147,"peak_dbfs":-10.607556199849471,"crest_factor_db":18.963228249716376,"mean":-0.00020426102437384405,"zero_energy":false},"rms_difference_db":-0.13000334404248903},"full_scale":{"reference":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":9.54596209769853e-07,"output_rms_dbfs":-60.201802943352746,"output_zero_energy":false,"output_peak":0.0032324299681931734,"output_energy_fraction":8.275264047831488e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":200,"both_estimated_voiced_frames":176,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9887640449438202,"absolute_difference_cents_median":3.0381592951592875,"absolute_difference_cents_p95":12.877390620489763,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011323786448874551,"log_rms_mae_db":1.2180566037908374,"rms_envelope_nrmse":0.13929409601269885,"envelope_correlation":0.9772719930930591,"rms_gain_error_db":-0.12995473310695863,"reference_mean_power":0.0011384943575262435,"output_mean_power":0.0011049316383708902},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":5.62759551802579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002928658523286973,"output_power":0.00026690872402582394,"reference_active":true,"delta_db":-0.4030596749880458},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4610498849864094e-05,"output_power":4.607873840622671e-05,"reference_active":true,"delta_db":0.14063500293376743},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2101686887119984e-05,"output_power":1.414066250854606e-05,"reference_active":true,"delta_db":0.6762384517478637}],"active_band_count":3,"active_band_mae_db":0.406644376556559}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":1.0114593137726633,"rms_envelope_nrmse":0.11857468117296324,"envelope_correlation":0.9823486825864601,"rms_gain_error_db":-0.12981465480445628,"reference_mean_power":0.0011449801518647084,"output_mean_power":0.0011112620743452635},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.095010942049215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002727818973842863,"output_power":0.0002512259207691039,"reference_active":true,"delta_db":-0.3575109923637731},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1267992599082604e-05,"output_power":2.1864171383329575e-05,"reference_active":true,"delta_db":0.1200652242083425},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.757168583210849e-07,"output_power":7.648513835446382e-07,"reference_active":true,"delta_db":1.23368109833812}],"active_band_count":3,"active_band_mae_db":0.5704191049700785}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":1.654346965218445,"rms_envelope_nrmse":0.13589778730270383,"envelope_correlation":0.9884991863328408,"rms_gain_error_db":0.514665066663029,"reference_mean_power":0.00038382289364078325,"output_mean_power":0.0004321130495387964},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":2.8601579771498626e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014700129208541836,"output_power":0.00017289956025383582,"reference_active":true,"delta_db":0.7047273665907893},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.957621157356578e-05,"output_power":2.399045191639784e-05,"reference_active":true,"delta_db":0.8830977891159242},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3113428408893907e-06,"output_power":2.8849697348922573e-06,"reference_active":true,"delta_db":0.9627689245325033}],"active_band_count":3,"active_band_mae_db":0.8501980267464054}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":1.3257121644027121,"rms_envelope_nrmse":0.12202366585497225,"envelope_correlation":0.9907478397241116,"rms_gain_error_db":0.5146864995606713,"reference_mean_power":0.00038603855676420007,"output_mean_power":0.0004346096181475679},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":2.6676069300397395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013762626674571402,"output_power":0.00016303707049657035,"reference_active":true,"delta_db":0.7358503383660129},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.746730510582487e-06,"output_power":1.17391442580763e-05,"reference_active":true,"delta_db":0.8077748120966856},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.594967364659404e-07,"output_power":1.9963390900955894e-07,"reference_active":true,"delta_db":0.9748250953329674}],"active_band_count":3,"active_band_mae_db":0.8394834152652219}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":1.673825952102414,"rms_envelope_nrmse":0.19122459232733102,"envelope_correlation":0.9782332442743057,"rms_gain_error_db":-1.0051155223177708,"reference_mean_power":0.00014337292372884635,"output_mean_power":0.00011375109590326938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.145508332328126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5890331523222755e-05,"output_power":3.863713894668988e-05,"reference_active":true,"delta_db":-1.6033172598754306},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4083924473542483e-05,"output_power":1.1674147719926805e-05,"reference_active":true,"delta_db":-0.814985030616681},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.499188380560451e-06,"output_power":6.699833914532959e-06,"reference_active":true,"delta_db":0.8576543973830555}],"active_band_count":3,"active_band_mae_db":1.0919855626250559}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":1.2692919254954604,"rms_envelope_nrmse":0.15983096695320007,"envelope_correlation":0.9877360188835537,"rms_gain_error_db":-1.0047960666763325,"reference_mean_power":0.00014418079784839883,"output_mean_power":0.00011440047238710513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.0427751966205221e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.204944813792335e-05,"output_power":3.648970199949123e-05,"reference_active":true,"delta_db":-1.5424581257293806},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.197217517258329e-06,"output_power":5.076854347211425e-06,"reference_active":true,"delta_db":-0.8660203593810798},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536643030241646e-07,"output_power":3.18875376629246e-07,"reference_active":true,"delta_db":0.9936162925460095}],"active_band_count":3,"active_band_mae_db":1.1340315925521567}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011141137328067573,"log_rms_mae_db":1.9349228199812225,"rms_envelope_nrmse":0.2622551910326944,"envelope_correlation":0.9658328482386771,"rms_gain_error_db":-1.2189284078758607,"reference_mean_power":0.00018590462562984242,"output_mean_power":0.0001404097787851112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.8616455925800917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.396584441083371e-05,"output_power":5.939368238984672e-05,"reference_active":true,"delta_db":-1.9922976860441528},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7617019162087808e-05,"output_power":1.8992441364712722e-05,"reference_active":true,"delta_db":0.3264836756044873},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489954892102077e-06,"output_power":6.975849995974368e-06,"reference_active":true,"delta_db":0.3135545496327216}],"active_band_count":3,"active_band_mae_db":0.8774453037604539}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":1.668789107311212,"rms_envelope_nrmse":0.21370787686812112,"envelope_correlation":0.9808180949957407,"rms_gain_error_db":-1.218914542668061,"reference_mean_power":0.00018697687803543296,"output_mean_power":0.00014122007889879237},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.7928912528322635e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.176227386223459e-05,"output_power":6.096313096389858e-05,"reference_active":true,"delta_db":-1.7759690327243616},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.193273720945635e-06,"output_power":9.316114667043632e-06,"reference_active":true,"delta_db":0.5577736137784085},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7870970633055527e-07,"output_power":5.364623779093767e-07,"reference_active":true,"delta_db":1.5123283407884414}],"active_band_count":3,"active_band_mae_db":1.2820236624304038}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.656759023666382,"predicted_mos":2.656759023666382,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":230912,"input_sample_rate":44100,"input_seconds":5.23609977324263,"input_float32_sha256":"341700796aeb4b45388863becd2985045eed31128964ef201c5e8bb764361901","internal_samples":83778,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6319663300018874},"started_at":"2026-10-13T07:23:58.708550+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00712579600076424,"native_input_samples":230423,"native_input_sha256":"1d959e9a1966d4cba16cafcc2ccf90eaac885e022320c59609688ccfa1bf50de","output_sha256":"aac47406e5976bae96e7642750c504e5427cec3578be370a0be30b8e6a529e14","output_samples":230912,"length_delta_samples":489,"analysis_s":0.14958125699922675,"synthesis_s":29.32791903600082,"total_s":29.47750029300005,"rtf":5.641614608443176,"output_peak":0.38116851449012756,"output_rms":0.03038061248434324,"finished_at":"2026-10-13T07:24:28.216199+00:00","evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"aac47406e5976bae96e7642750c504e5427cec3578be370a0be30b8e6a529e14","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":83600,"output_samples":83778,"length_delta_samples":178,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.34608204343076,"output_peak":0.3816934874609434},"common_prefix":{"samples":83600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9978753372006971,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.553015000637846,"spectral_convergence":0.5774184139756237,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.418866415918169,"spectral_convergence":0.563308222191809,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.457624180582743,"spectral_convergence":0.5855820858802672,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00015521969978656445,"delta_db":-1.3300652863340225,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.0004135125414366829,"delta_db":-0.6132037754139059,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":7.858752234623628e-05,"delta_db":-0.8637320331818659,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":5.276984250308021e-05,"delta_db":-1.475036469294222,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.00014320143095444707,"delta_db":0.3445941558110457,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":8.207766823418226e-05,"delta_db":-2.5630355560805054,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1982778793525946,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8960708129717565,"pitch":null,"output_rms_dbfs":-30.336851918495114,"output_peak":0.3816934874609434,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.476501865712919},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83600,"output_samples":83600,"intelligibility":{"estoi":{"value":0.6264964220540026,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.033725697841694866,"rms_dbfs":-29.44078110552336,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.117693772983436,"mean":-5.154404343600478e-06,"zero_energy":false},"output":{"rms":0.030419873524740255,"rms_dbfs":-30.336851918495114,"peak":0.3816934874609434,"peak_dbfs":-8.36570500033185,"crest_factor_db":21.971146918163264,"mean":-0.000125280863276844,"zero_energy":false},"rms_difference_db":-0.8960708129717538},"full_scale":{"reference":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":8.676061654300181e-07,"output_rms_dbfs":-60.616773704398625,"output_zero_energy":false,"output_peak":0.003525446265052218,"output_energy_fraction":8.972046814015056e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":193,"both_estimated_voiced_frames":174,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9775280898876404,"absolute_difference_cents_median":9.675161676909537,"absolute_difference_cents_p95":33.98881553548046,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011323786448874551,"log_rms_mae_db":1.7543874001750759,"rms_envelope_nrmse":0.1522900288576724,"envelope_correlation":0.9835262506401307,"rms_gain_error_db":-0.8960580243031555,"reference_mean_power":0.0011384943575262435,"output_mean_power":0.0009262433012914945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":5.62759551802579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002928658523286973,"output_power":0.0002453206413915611,"reference_active":true,"delta_db":-0.7693464513804872},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4610498849864094e-05,"output_power":4.2088179838320506e-05,"reference_active":true,"delta_db":-0.2527693521104066},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2101686887119984e-05,"output_power":1.0642861427090413e-05,"reference_active":true,"delta_db":-0.5578750447039592}],"active_band_count":3,"active_band_mae_db":0.526663616064951}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":1.555733086473787,"rms_envelope_nrmse":0.12884923945022483,"envelope_correlation":0.9909737523213922,"rms_gain_error_db":-0.8958778847209914,"reference_mean_power":0.0011449801518647084,"output_mean_power":0.0009315585791544432},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.095010942049215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002727818973842863,"output_power":0.00022973937386255524,"reference_active":true,"delta_db":-0.7458011277501263},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1267992599082604e-05,"output_power":2.0436921593874367e-05,"reference_active":true,"delta_db":-0.1731102174672277},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.757168583210849e-07,"output_power":5.89112077168758e-07,"reference_active":true,"delta_db":0.09988979349602845}],"active_band_count":3,"active_band_mae_db":0.3396003795711275}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":3.2044571875044685,"rms_envelope_nrmse":0.34016056406732237,"envelope_correlation":0.9067723090201699,"rms_gain_error_db":-1.3261646944733465,"reference_mean_power":0.00038382289364078325,"output_mean_power":0.0002828227923533188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":2.8601579771498626e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014700129208541836,"output_power":8.052477018301817e-05,"reference_active":true,"delta_db":-2.6139165802729645},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.957621157356578e-05,"output_power":1.862580124719842e-05,"reference_active":true,"delta_db":-0.2161368572784611},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3113428408893907e-06,"output_power":1.96392545827942e-06,"reference_active":true,"delta_db":-0.7073936913497174}],"active_band_count":3,"active_band_mae_db":1.1791490429670477}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":2.6474501509561565,"rms_envelope_nrmse":0.3233510116200991,"envelope_correlation":0.9123875255062264,"rms_gain_error_db":-1.3262803738444666,"reference_mean_power":0.00038603855676420007,"output_mean_power":0.0002844478437279427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":2.6676069300397395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013762626674571402,"output_power":7.75191191848605e-05,"reference_active":true,"delta_db":-2.4929249997404614},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.746730510582487e-06,"output_power":9.770731748788529e-06,"reference_active":true,"delta_db":0.010681317506899693},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.594967364659404e-07,"output_power":1.394616165543576e-07,"reference_active":true,"delta_db":-0.5829710621453708}],"active_band_count":3,"active_band_mae_db":1.028859126464244}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":3.103028775773411,"rms_envelope_nrmse":0.2629389773377653,"envelope_correlation":0.9546092686727836,"rms_gain_error_db":-1.1746379415926262,"reference_mean_power":0.00014337292372884635,"output_mean_power":0.00010939647984427996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.145508332328126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5890331523222755e-05,"output_power":3.8679636403392204e-05,"reference_active":true,"delta_db":-1.598543027448935},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4083924473542483e-05,"output_power":1.0764306288112629e-05,"reference_active":true,"delta_db":-1.1673764075210966},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.499188380560451e-06,"output_power":3.5492012206547302e-06,"reference_active":true,"delta_db":-1.9016797494603208}],"active_band_count":3,"active_band_mae_db":1.5558663948101177}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":2.7482059315795495,"rms_envelope_nrmse":0.22422667799190765,"envelope_correlation":0.9693066523622839,"rms_gain_error_db":-1.1743490287467842,"reference_mean_power":0.00014418079784839883,"output_mean_power":0.00011002022317074987},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.0427751966205221e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.204944813792335e-05,"output_power":3.605698717082144e-05,"reference_active":true,"delta_db":-1.5942669382696417},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.197217517258329e-06,"output_power":4.8811746489260835e-06,"reference_active":true,"delta_db":-1.0367239276243898},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536643030241646e-07,"output_power":2.1594595179867687e-07,"reference_active":true,"delta_db":-0.6991428829872953}],"active_band_count":3,"active_band_mae_db":1.1100445829604424}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011141137328067573,"log_rms_mae_db":3.470747061482541,"rms_envelope_nrmse":0.21982916107681394,"envelope_correlation":0.9785481625066479,"rms_gain_error_db":-1.1247981085571404,"reference_mean_power":0.00018590462562984242,"output_mean_power":0.00014348628289286486},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.8616455925800917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.396584441083371e-05,"output_power":7.238847481516356e-05,"reference_active":true,"delta_db":-1.133005944570057},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7617019162087808e-05,"output_power":1.4676953035135997e-05,"reference_active":true,"delta_db":-0.7929652218965965},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489954892102077e-06,"output_power":4.9931851560885775e-06,"reference_active":true,"delta_db":-1.138640075257492}],"active_band_count":3,"active_band_mae_db":1.021537080574715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":3.0827091415759393,"rms_envelope_nrmse":0.16969208873956368,"envelope_correlation":0.9910291537290271,"rms_gain_error_db":-1.124785456720323,"reference_mean_power":0.00018697687803543296,"output_mean_power":0.00014431429708850743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.7928912528322635e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.176227386223459e-05,"output_power":6.988269119139757e-05,"reference_active":true,"delta_db":-1.1829454544198397},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.193273720945635e-06,"output_power":7.109794872818573e-06,"reference_active":true,"delta_db":-0.6160039308649244},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7870970633055527e-07,"output_power":3.534779652880692e-07,"reference_active":true,"delta_db":-0.2994408981631664}],"active_band_count":3,"active_band_mae_db":0.6994634278159767}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"golf","status":"measured","mos":{"status":"measured","value":3.010061502456665,"predicted_mos":3.010061502456665,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":125280,"input_sample_rate":24000,"input_seconds":5.22,"input_float32_sha256":"8e6e98deb0a202c7fdcbff5306eaa8d921f0ea78af37199cd70fe1a69449cda4","internal_samples":83520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.127551362000304},"started_at":"2026-10-13T05:02:51.401415+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007933212999887473,"native_input_samples":125400,"native_input_sha256":"25b4e60ca2813e5c08638e2a9966e0b7a14332042b2f2fd4801cd0ca917b921c","output_sha256":"203894a4cd94ad6467b59e0e419bef011d5f06e989741081528ab7dfd6b24843","output_samples":125280,"length_delta_samples":-120,"analysis_s":0.5584069970000201,"synthesis_s":0.12158773200008,"total_s":0.6799947290001,"rtf":0.13014253186604788,"output_peak":0.36038142442703247,"output_rms":0.03151278137576997,"finished_at":"2026-10-13T05:02:52.098801+00:00","analysis_stages":{"external_f0_s":0.06186824599990359},"evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"203894a4cd94ad6467b59e0e419bef011d5f06e989741081528ab7dfd6b24843","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":83600,"output_samples":83520,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.03101951989963,"output_peak":0.36337606500862935},"common_prefix":{"samples":83520,"sample_rate":16000,"reference_coverage":0.999043062200957,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83520,"output_samples":83520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.269122473523145,"spectral_convergence":0.6848308730850784,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.707363063933487,"spectral_convergence":0.37140369960736996,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.000774481198323,"spectral_convergence":0.36509514763066686,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021098937699707616,"output_power":0.00018102339598116116,"delta_db":-0.6652588180665582,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004766042115248086,"output_power":0.0004406060996167007,"delta_db":-0.34107370584286256,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.604908562483909e-05,"output_power":7.443875145326751e-05,"delta_db":-1.1069415460603762,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.419606291827959e-05,"output_power":5.66900233033288e-05,"delta_db":-1.168742253138758,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013240770712643333,"output_power":9.239033073602012e-05,"delta_db":-1.5628674326545102,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014823694563435527,"output_power":0.00014773433759005268,"delta_db":-0.014750087079105946,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8099389738070285,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5942865074470401,"pitch":null,"output_rms_dbfs":-30.03101951989963,"output_peak":0.36337606500862935,"reference_rms_dbfs":-29.43673301245259,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.325753339551652},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83520,"output_samples":83520,"intelligibility":{"estoi":{"value":0.8470804967278501,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0337414194992711,"rms_dbfs":-29.43673301245259,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.11364567991267,"mean":-1.0247796888095666e-05,"zero_energy":false},"output":{"rms":0.03151004504408921,"rms_dbfs":-30.03101951989963,"peak":0.36337606500862935,"peak_dbfs":-8.792873647164786,"crest_factor_db":21.238145872734844,"mean":0.0001419078475051487,"zero_energy":false},"rms_difference_db":-0.5942865074470411},"full_scale":{"reference":{"sample_count":83520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":1.739385362174005e-06,"output_rms_dbfs":-57.59604189058304,"output_zero_energy":false,"output_peak":0.0038306079687105816,"output_energy_fraction":1.6780205079673934e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":174,"both_estimated_voiced_frames":158,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.8876404494382022,"absolute_difference_cents_median":3.7960597383731782,"absolute_difference_cents_p95":17.69559922013648,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00011323786448874551,"log_rms_mae_db":2.0403135190484676,"rms_envelope_nrmse":0.1624004382820397,"envelope_correlation":0.9737820464748617,"rms_gain_error_db":-0.5943122452205345,"reference_mean_power":0.0011395572655238675,"output_mean_power":0.000993813586898387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":5.627199012584147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00029309958753001694,"output_power":0.0002557894139730318,"reference_active":true,"delta_db":-0.5913264034138994},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.444085188273949e-05,"output_power":4.5014611416507416e-05,"reference_active":true,"delta_db":0.05571130198253431},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.207970031547442e-05,"output_power":1.4255936238422116e-05,"reference_active":true,"delta_db":0.7193958422283088}],"active_band_count":3,"active_band_mae_db":0.45547784920824747}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":1.3564295127414967,"rms_envelope_nrmse":0.10555587967067394,"envelope_correlation":0.9913536114164495,"rms_gain_error_db":-0.5944312434974512,"reference_mean_power":0.0011460726356755525,"output_mean_power":0.0009994682876002907},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":5.094165660573185e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027297534569854207,"output_power":0.00024097737542251943,"reference_active":true,"delta_db":-0.5414715465703638},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1104687750394886e-05,"output_power":2.0822461584458726e-05,"reference_active":true,"delta_db":-0.05846861508877613},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.748702295881884e-07,"output_power":7.786062477675945e-07,"reference_active":true,"delta_db":1.3174806558723757}],"active_band_count":3,"active_band_mae_db":0.6391402725105052}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":1.9066154258112566,"rms_envelope_nrmse":0.170394528283119,"envelope_correlation":0.9791672604328757,"rms_gain_error_db":-0.6706253588890883,"reference_mean_power":0.0003841907081305083,"output_mean_power":0.00032921856767435944},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":2.8620360956160757e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001470401653373578,"output_power":0.0001309576702726352,"reference_active":true,"delta_db":-0.5030504197338785},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9490925657334733e-05,"output_power":2.0104399680788804e-05,"reference_active":true,"delta_db":0.13458644565847963},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3092759996365147e-06,"output_power":2.670206845531939e-06,"reference_active":true,"delta_db":0.6306906300811235}],"active_band_count":3,"active_band_mae_db":0.42277583182449385}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":1.3080898296478871,"rms_envelope_nrmse":0.13174224262797682,"envelope_correlation":0.9888086160323885,"rms_gain_error_db":-0.6706109176905863,"reference_mean_power":0.00038641064139086344,"output_mean_power":0.00033112196160133825},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":2.6693038751012205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013764649939715083,"output_power":0.00012426029804628883,"reference_active":true,"delta_db":-0.44432780003361266},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.652189881002628e-06,"output_power":9.945712410059444e-06,"reference_active":true,"delta_db":0.1301004012836862},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5949370452332684e-07,"output_power":1.8740994001681644e-07,"reference_active":true,"delta_db":0.700490762437424}],"active_band_count":3,"active_band_mae_db":0.42497298791824095}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":2.0043935276314255,"rms_envelope_nrmse":0.24142166211306057,"envelope_correlation":0.9692564957069142,"rms_gain_error_db":-1.518325843596381,"reference_mean_power":0.00014350905709118752,"output_mean_power":0.00010116882979674173},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.1475009616861659e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.602368526533868e-05,"output_power":3.9530882666219896e-05,"reference_active":true,"delta_db":-1.514351619244915},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4054479078547208e-05,"output_power":9.235446091973652e-06,"reference_active":true,"delta_db":-1.8235687571479837},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.5067950222375016e-06,"output_power":4.0554310016273485e-06,"reference_active":true,"delta_db":-1.3286189437855034}],"active_band_count":3,"active_band_mae_db":1.555513106726134}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":1.5785109459329782,"rms_envelope_nrmse":0.19410142299592026,"envelope_correlation":0.9878834810502877,"rms_gain_error_db":-1.5189443889920051,"reference_mean_power":0.00014431836354818775,"output_mean_power":0.00010172487310765016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":1.0446000931960287e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.2157766199105866e-05,"output_power":3.7427927599707446e-05,"reference_active":true,"delta_db":-1.4412320254759416},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.160172755884807e-06,"output_power":4.070656193938523e-06,"reference_active":true,"delta_db":-1.7992846810678056},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.538564525317293e-07,"output_power":1.7423049661352883e-07,"reference_active":true,"delta_db":-1.634640322874589}],"active_band_count":3,"active_band_mae_db":1.6250523431394452}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1043,"rms_floor":0.00011141137328067573,"log_rms_mae_db":2.911902937658002,"rms_envelope_nrmse":0.2353778031496007,"envelope_correlation":0.9692177172403488,"rms_gain_error_db":-0.21100957924669272,"reference_mean_power":0.00018608265837041923,"output_mean_power":0.0001772576322409571},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1043,"frequency_resolution_hz":0.19175455417066156,"power_floor":1.8556533945760504e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.373468913830471e-05,"output_power":8.409584638159549e-05,"reference_active":true,"delta_db":-0.47125797551438414},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7518066031945642e-05,"output_power":1.6495952992475943e-05,"reference_active":true,"delta_db":-0.26108748664620907},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.471614550176077e-06,"output_power":8.162778408342762e-06,"reference_active":true,"delta_db":1.0082536427789388}],"active_band_count":3,"active_band_mae_db":0.5801997016465107}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83520,"uncovered_tail_samples":0,"metrics":{"frame_count":1037,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":2.6236440472813642,"rms_envelope_nrmse":0.13887267888828192,"envelope_correlation":0.9899806028998159,"rms_gain_error_db":-0.21099960426179182,"reference_mean_power":0.00018715700716380733,"output_mean_power":0.00017828143920000159},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1037,"frequency_resolution_hz":0.1928640308582449,"power_floor":1.786890676004092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.151085791359478e-05,"output_power":8.421299636650212e-05,"reference_active":true,"delta_db":-0.3609350669949092},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.136081981527805e-06,"output_power":7.69352141465769e-06,"reference_active":true,"delta_db":-0.2429014904703898},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7657770388053054e-07,"output_power":4.183374215357307e-07,"reference_active":true,"delta_db":0.45672112188597264}],"active_band_count":3,"active_band_mae_db":0.3535192264504239}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.918530225753784,"predicted_mos":2.918530225753784,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":125520,"input_sample_rate":24000,"input_seconds":5.23,"input_float32_sha256":"42599978e3ecc7490837bd59a47fbbca3b63f2ebd37799f1453035a9d5063e71","internal_samples":83680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.710153847998299},"started_at":"2026-10-13T07:42:32.628770+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003737412000191398,"native_input_samples":125400,"native_input_sha256":"25b4e60ca2813e5c08638e2a9966e0b7a14332042b2f2fd4801cd0ca917b921c","output_sha256":"15068fe98afa0b7fabbfc666de054a77873fa4f5245ef3ddcd738f8bc0f25597","output_samples":125520,"length_delta_samples":120,"analysis_s":2.0206865449981706,"synthesis_s":11.40244485400035,"total_s":13.423131398998521,"rtf":2.569020363444693,"output_peak":0.3865481913089752,"output_rms":0.03396368668614271,"finished_at":"2026-10-13T07:42:46.073432+00:00","analysis_stages":{"external_f0_s":1.3140246370021487},"evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"15068fe98afa0b7fabbfc666de054a77873fa4f5245ef3ddcd738f8bc0f25597","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":83600,"output_samples":83680,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.37950642201278,"output_peak":0.3684837869585476},"common_prefix":{"samples":83600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990439770554493,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.57105566486052,"spectral_convergence":0.6816357473650736,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.189969186813099,"spectral_convergence":0.41511921316756945,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.558729642685073,"spectral_convergence":0.3949855673223041,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00030156318608980413,"delta_db":1.554249553362714,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.0004543709177793763,"delta_db":-0.2039853698915664,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":9.266946374693184e-05,"delta_db":-0.14790150283453843,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":6.724501498729055e-05,"delta_db":-0.4222935019753554,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.0001283291275452528,"delta_db":-0.13162700448793538,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.0001105072643827219,"delta_db":-1.2713773752123478,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6219057179607429,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.06541625097370231,"pitch":null,"output_rms_dbfs":-29.37536485454966,"output_peak":0.3684837869585476,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.7732514981195635},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83600,"output_samples":83600,"intelligibility":{"estoi":{"value":0.8357660260027898,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.033725697841694866,"rms_dbfs":-29.44078110552336,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.117693772983436,"mean":-5.154404343600478e-06,"zero_energy":false},"output":{"rms":0.033980655887304145,"rms_dbfs":-29.37536485454966,"peak":0.3684837869585476,"peak_dbfs":-8.671632321056,"crest_factor_db":20.703732533493657,"mean":-3.7650167391454966e-05,"zero_energy":false},"rms_difference_db":0.0654162509737013},"full_scale":{"reference":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":5.856397584964661e-07,"output_rms_dbfs":-62.323694471315385,"output_zero_energy":false,"output_peak":0.0023495921458275984,"output_energy_fraction":4.853452100692333e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":212,"both_estimated_voiced_frames":170,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":42,"paired_voiced_fraction":0.9550561797752809,"absolute_difference_cents_median":4.319258824531355,"absolute_difference_cents_p95":19.27324055914051,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011323786448874551,"log_rms_mae_db":2.717945155246626,"rms_envelope_nrmse":0.17371816366390508,"envelope_correlation":0.9653564750009623,"rms_gain_error_db":0.06546804637405267,"reference_mean_power":0.0011384943575262435,"output_mean_power":0.0011557866856226338},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":5.62759551802579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002928658523286973,"output_power":0.000286842550194974,"reference_active":true,"delta_db":-0.09025161645072},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4610498849864094e-05,"output_power":3.9993711558128535e-05,"reference_active":true,"delta_db":-0.47445369627860456},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2101686887119984e-05,"output_power":1.038861288905037e-05,"reference_active":true,"delta_db":-0.6628834857511752}],"active_band_count":3,"active_band_mae_db":0.40919626616016663}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":2.3539328628657006,"rms_envelope_nrmse":0.10737697413228763,"envelope_correlation":0.9856984337185347,"rms_gain_error_db":0.06578525545247195,"reference_mean_power":0.0011449801518647084,"output_mean_power":0.0011624558939766344},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.095010942049215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002727818973842863,"output_power":0.0002641923306782828,"reference_active":true,"delta_db":-0.13895339679616736},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1267992599082604e-05,"output_power":1.8299713885398006e-05,"reference_active":true,"delta_db":-0.6528220089161731},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.757168583210849e-07,"output_power":5.598924293587489e-07,"reference_active":true,"delta_db":-0.12104351609573247}],"active_band_count":3,"active_band_mae_db":0.3042729739360243}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":1.944916734972941,"rms_envelope_nrmse":0.167174555961719,"envelope_correlation":0.9769179045861967,"rms_gain_error_db":-0.21540931734837104,"reference_mean_power":0.00038382289364078325,"output_mean_power":0.0003652497636051051},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":2.8601579771498626e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014700129208541836,"output_power":0.0001433195725928168,"reference_active":true,"delta_db":-0.11015647703832059},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.957621157356578e-05,"output_power":1.8769446452807435e-05,"reference_active":true,"delta_db":-0.18277185459943487},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3113428408893907e-06,"output_power":2.2844954820631296e-06,"reference_active":true,"delta_db":-0.050740655367873325}],"active_band_count":3,"active_band_mae_db":0.11455632900187625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":1.2682249233505607,"rms_envelope_nrmse":0.10670448167312409,"envelope_correlation":0.9903301516827735,"rms_gain_error_db":-0.21539331530158692,"reference_mean_power":0.00038603855676420007,"output_mean_power":0.0003673595647084017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":2.6676069300397395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013762626674571402,"output_power":0.00013439150991105786,"reference_active":true,"delta_db":-0.10329496013976042},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.746730510582487e-06,"output_power":9.25617711679501e-06,"reference_active":true,"delta_db":-0.22427302089594336},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.594967364659404e-07,"output_power":1.6559596214405885e-07,"reference_active":true,"delta_db":0.16297941645255168}],"active_band_count":3,"active_band_mae_db":0.16351579916275183}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":1.764990109792694,"rms_envelope_nrmse":0.24595375757752252,"envelope_correlation":0.9519164068221914,"rms_gain_error_db":-0.05313018805374534,"reference_mean_power":0.00014337292372884635,"output_mean_power":0.0001416296307449167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.145508332328126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5890331523222755e-05,"output_power":5.497997964847899e-05,"reference_active":true,"delta_db":-0.07132110968810766},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4083924473542483e-05,"output_power":1.4264298445768959e-05,"reference_active":true,"delta_db":0.05526729361960403},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.499188380560451e-06,"output_power":5.1986171661366445e-06,"reference_active":true,"delta_db":-0.24410760634549777}],"active_band_count":3,"active_band_mae_db":0.12356533655106983}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":0.9691519105325781,"rms_envelope_nrmse":0.12414757629396539,"envelope_correlation":0.9868639758851921,"rms_gain_error_db":-0.05269544979030865,"reference_mean_power":0.00014418079784839883,"output_mean_power":0.0001424419398370104},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.0427751966205221e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.204944813792335e-05,"output_power":5.1624052422082014e-05,"reference_active":true,"delta_db":-0.03564036091307186},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.197217517258329e-06,"output_power":6.3802092352884444e-06,"reference_active":true,"delta_db":0.12638181594433967},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536643030241646e-07,"output_power":2.5443637920479604e-07,"reference_active":true,"delta_db":0.013198513391496611}],"active_band_count":3,"active_band_mae_db":0.05840689674963604}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011141137328067573,"log_rms_mae_db":1.8100864998818644,"rms_envelope_nrmse":0.2509766544276395,"envelope_correlation":0.9686067449032737,"rms_gain_error_db":-1.1341808077878464,"reference_mean_power":0.00018590462562984242,"output_mean_power":0.00014317662310055375},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.8616455925800917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.396584441083371e-05,"output_power":7.444884077521268e-05,"reference_active":true,"delta_db":-1.0111208099194082},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7617019162087808e-05,"output_power":1.1474755272604148e-05,"reference_active":true,"delta_db":-1.8618899481291624},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489954892102077e-06,"output_power":4.023466066899095e-06,"reference_active":true,"delta_db":-2.076413353563714}],"active_band_count":3,"active_band_mae_db":1.6498080372040949}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":1.0326004520815124,"rms_envelope_nrmse":0.17294395505776966,"envelope_correlation":0.9901668260085942,"rms_gain_error_db":-1.1341623912958079,"reference_mean_power":0.00018697687803543296,"output_mean_power":0.00014400304149084098},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.7928912528322635e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.176227386223459e-05,"output_power":7.218942523595078e-05,"reference_active":true,"delta_db":-1.041905827876025},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.193273720945635e-06,"output_power":4.835025809249979e-06,"reference_active":true,"delta_db":-2.2905866733626756},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7870970633055527e-07,"output_power":2.476382966722385e-07,"reference_active":true,"delta_db":-1.844886280837386}],"active_band_count":3,"active_band_mae_db":1.7257929273586956}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.1958134174346924,"predicted_mos":3.1958134174346924,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":125640,"input_sample_rate":24000,"input_seconds":5.235,"input_float32_sha256":"fa4ad588c725473aced0cbb47d53b55089a5b573eb00ba7e9e76e6da179f7efe","internal_samples":83760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.248153435999484},"started_at":"2026-10-13T05:11:11.074973+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0055872020002425415,"native_input_samples":125400,"native_input_sha256":"25b4e60ca2813e5c08638e2a9966e0b7a14332042b2f2fd4801cd0ca917b921c","output_sha256":"957dd98db0b3293e343a5e7ee4421873c4ad4dbeb47784dcb9e954d982d06eb5","output_samples":125640,"length_delta_samples":240,"analysis_s":1.0128485999994155,"synthesis_s":0.1182069770002272,"total_s":1.1310555769996427,"rtf":0.21646996688988376,"output_peak":0.33973607420921326,"output_rms":0.032304023317983606,"finished_at":"2026-10-13T05:11:12.219027+00:00","analysis_stages":{"external_f0_s":0.12830400799975905},"evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"957dd98db0b3293e343a5e7ee4421873c4ad4dbeb47784dcb9e954d982d06eb5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":83600,"output_samples":83760,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.814403544129043,"output_peak":0.32735466381684253},"common_prefix":{"samples":83600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980897803247374,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.7091174448920246,"spectral_convergence":0.3060671550024367,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.334427146997205,"spectral_convergence":0.27663848482136855,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.976921126512656,"spectral_convergence":0.2858943216048669,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.000203642612456125,"delta_db":-0.1508470795190568,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.0004545262552793645,"delta_db":-0.20250088469931238,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":8.989785442780924e-05,"delta_db":-0.2797747352072937,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":6.593783073968458e-05,"delta_db":-0.5075478953787478,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.00011756050004709328,"delta_db":-0.5122651745389787,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.00011408896479823764,"delta_db":-1.1328492604720306,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.46429750496923666,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3653387894934067,"pitch":null,"output_rms_dbfs":-29.806119895016767,"output_peak":0.32735466381684253,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.673488572800629},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83600,"output_samples":83600,"intelligibility":{"estoi":{"value":0.912995412243793,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.033725697841694866,"rms_dbfs":-29.44078110552336,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.117693772983436,"mean":-5.154404343600478e-06,"zero_energy":false},"output":{"rms":0.03233657399521962,"rms_dbfs":-29.806119895016764,"peak":0.32735466381684253,"peak_dbfs":-9.699629346222565,"crest_factor_db":20.1064905487942,"mean":-2.5962394521414784e-05,"zero_energy":false},"rms_difference_db":-0.3653387894934035},"full_scale":{"reference":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":8.762206004225322e-07,"output_rms_dbfs":-60.5738654053851,"output_zero_energy":false,"output_peak":0.0025862096583508147,"output_energy_fraction":8.018795879062458e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":165,"both_estimated_voiced_frames":162,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9101123595505618,"absolute_difference_cents_median":0.6988645357578651,"absolute_difference_cents_p95":4.585139024974313,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011323786448874551,"log_rms_mae_db":1.0261924014791683,"rms_envelope_nrmse":0.1000499874206665,"envelope_correlation":0.9899046613236749,"rms_gain_error_db":-0.36538854800813836,"reference_mean_power":0.0011384943575262435,"output_mean_power":0.0010466272256797087},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":5.62759551802579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002928658523286973,"output_power":0.00026909535807068433,"reference_active":true,"delta_db":-0.36762530426172535},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4610498849864094e-05,"output_power":3.2587752253703954e-05,"reference_active":true,"delta_db":-1.363826737430682},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2101686887119984e-05,"output_power":7.759231412131668e-06,"reference_active":true,"delta_db":-1.9302720752221862}],"active_band_count":3,"active_band_mae_db":1.220574705638198}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":0.656306531673797,"rms_envelope_nrmse":0.06857972282478254,"envelope_correlation":0.9960471739052777,"rms_gain_error_db":-0.3654098531126262,"reference_mean_power":0.0011449801518647084,"output_mean_power":0.0010525845060961442},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.095010942049215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002727818973842863,"output_power":0.00024888279878402516,"reference_active":true,"delta_db":-0.39820663932159506},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1267992599082604e-05,"output_power":1.5215462310027636e-05,"reference_active":true,"delta_db":-1.4544134791887395},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.757168583210849e-07,"output_power":4.0434749950481996e-07,"reference_active":true,"delta_db":-1.5345418502130272}],"active_band_count":3,"active_band_mae_db":1.129053989574454}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":0.9340150644210837,"rms_envelope_nrmse":0.07214133747992045,"envelope_correlation":0.9963316186954689,"rms_gain_error_db":-0.26294266786780535,"reference_mean_power":0.00038382289364078325,"output_mean_power":0.00036127391754933844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":2.8601579771498626e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014700129208541836,"output_power":0.0001340888046974052,"reference_active":true,"delta_db":-0.3992863266866045},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.957621157356578e-05,"output_power":1.6127950973293358e-05,"reference_active":true,"delta_db":-0.841494555147023},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3113428408893907e-06,"output_power":1.4629347988893922e-06,"reference_active":true,"delta_db":-1.98639398432393}],"active_band_count":3,"active_band_mae_db":1.0757249553858526}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":0.5188398250372818,"rms_envelope_nrmse":0.047262082108124256,"envelope_correlation":0.9987860338550377,"rms_gain_error_db":-0.26293931646599716,"reference_mean_power":0.00038603855676420007,"output_mean_power":0.0003633596944374782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":2.6676069300397395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013762626674571402,"output_power":0.00012497478833330247,"reference_active":true,"delta_db":-0.4187891947903756},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.746730510582487e-06,"output_power":7.978339374179377e-06,"reference_active":true,"delta_db":-0.8694645240345501},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.594967364659404e-07,"output_power":9.405888851012379e-08,"reference_active":true,"delta_db":-2.293519587894775}],"active_band_count":3,"active_band_mae_db":1.1939244355732335}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":0.946060727109715,"rms_envelope_nrmse":0.121209514957224,"envelope_correlation":0.9908327100310644,"rms_gain_error_db":-0.557508405815223,"reference_mean_power":0.00014337292372884635,"output_mean_power":0.00012610035264241083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.145508332328126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5890331523222755e-05,"output_power":4.7971632972075794e-05,"reference_active":true,"delta_db":-0.6635218345497358},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4083924473542483e-05,"output_power":1.1669997165541978e-05,"reference_active":true,"delta_db":-0.8165293689456345},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.499188380560451e-06,"output_power":4.254300572149786e-06,"reference_active":true,"delta_db":-1.1147042700779017}],"active_band_count":3,"active_band_mae_db":0.8649184911910908}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":0.6073302627072976,"rms_envelope_nrmse":0.08127476109935437,"envelope_correlation":0.9975067401560123,"rms_gain_error_db":-0.5575726365693485,"reference_mean_power":0.00014418079784839883,"output_mean_power":0.0001268090242373858},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.0427751966205221e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.204944813792335e-05,"output_power":4.444960869935844e-05,"reference_active":true,"delta_db":-0.6854818707477717},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.197217517258329e-06,"output_power":4.9085575838797656e-06,"reference_active":true,"delta_db":-1.0124284958795806},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536643030241646e-07,"output_power":2.617520326824181e-07,"reference_active":true,"delta_db":0.13630707639465386}],"active_band_count":3,"active_band_mae_db":0.6114058143406688}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011141137328067573,"log_rms_mae_db":1.0904519111354123,"rms_envelope_nrmse":0.19658867351387815,"envelope_correlation":0.9816493047695708,"rms_gain_error_db":-0.9069575912780827,"reference_mean_power":0.00018590462562984242,"output_mean_power":0.00015086706333733899},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.8616455925800917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.396584441083371e-05,"output_power":7.916250976302946e-05,"reference_active":true,"delta_db":-0.7445046624260006},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7617019162087808e-05,"output_power":1.004985929684127e-05,"reference_active":true,"delta_db":-2.4377244526765067},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489954892102077e-06,"output_power":4.0234677405902684e-06,"reference_active":true,"delta_db":-2.07641154697537}],"active_band_count":3,"active_band_mae_db":1.752880220692626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":0.8042922571072298,"rms_envelope_nrmse":0.13472273206185567,"envelope_correlation":0.9946016584197643,"rms_gain_error_db":-0.9069653512681164,"reference_mean_power":0.00018697687803543296,"output_mean_power":0.0001517369565299331},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.7928912528322635e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.176227386223459e-05,"output_power":7.596604978185655e-05,"reference_active":true,"delta_db":-0.8204462303871786},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.193273720945635e-06,"output_power":4.130919725614611e-06,"reference_active":true,"delta_db":-2.9741070841403334},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7870970633055527e-07,"output_power":2.9271215852209695e-07,"reference_active":true,"delta_db":-1.118656739300699}],"active_band_count":3,"active_band_mae_db":1.6377366846094035}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_672-122797-0056","method":"magphase","status":"measured","mos":{"status":"measured","value":3.336557626724243,"predicted_mos":3.336557626724243,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":252178,"input_sample_rate":48000,"input_seconds":5.253708333333333,"input_float32_sha256":"21b89246adb27c590448db3273aa1e7934108aa97d92fe997ac4bd3743f513e5","internal_samples":84060,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2087911979997443},"started_at":"2026-10-13T05:17:16.188830+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008735076999073499,"native_input_samples":250800,"native_input_sha256":"5480c997b30f5f9924e3cf339ab6fdcd73dfa16f56c68dd17514f981b96b2478","output_sha256":"dc4e51a414872e6d7d8828366682fca646bf46b085903ad0a6f4838f231d824f","output_samples":252178,"length_delta_samples":1378,"analysis_s":2.6429682739999407,"synthesis_s":0.9453695789998164,"total_s":3.588337852999757,"rtf":0.686763225454499,"output_peak":0.3894544438228975,"output_rms":0.03266196819052742,"finished_at":"2026-10-13T05:17:19.796390+00:00","evaluation_input":{"reference":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"output_sha256":"dc4e51a414872e6d7d8828366682fca646bf46b085903ad0a6f4838f231d824f","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":83600,"output_samples":84060,"length_delta_samples":460,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.718956118269695,"output_peak":0.3692138398162722},"common_prefix":{"samples":83600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9945277182964549,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":83600,"output_samples":83600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.022258874117245,"spectral_convergence":0.7462204405940962,"spectral_convergence_status":"ok","amplitude_floor":9.102684030105167e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.159668996671338,"spectral_convergence":0.6062560964305314,"spectral_convergence_status":"ok","amplitude_floor":4.6649711782186323e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.788562594722691,"spectral_convergence":0.5446759893862412,"spectral_convergence_status":"ok","amplitude_floor":4.1057122515023654e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00021084017487574348,"output_power":0.00017978901823515286,"delta_db":-0.6919020686441735,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00047622157098200406,"output_power":0.00043555503614045533,"delta_db":-0.3876602369300549,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":9.587972951197478e-05,"output_power":9.028877206264968e-05,"delta_db":-0.2609305355726983,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":7.411215555329107e-05,"output_power":7.209240162559006e-05,"delta_db":-0.11999951427385674,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013227809814888984,"output_power":0.00013928227108226431,"delta_db":0.22407897460123816,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014809096583739848,"output_power":0.0001556836382222371,"delta_db":0.2171440664752129,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3169525660828724,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.25447196974711733,"pitch":null,"output_rms_dbfs":-29.69525307527047,"output_peak":0.3692138398162722,"reference_rms_dbfs":-29.44078110552336,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.323496821837092},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":83600,"output_samples":83600,"intelligibility":{"estoi":{"value":0.5850518264775539,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.033725697841694866,"rms_dbfs":-29.44078110552336,"peak":0.34185791015625,"peak_dbfs":-9.32308733253992,"crest_factor_db":20.117693772983436,"mean":-5.154404343600478e-06,"zero_energy":false},"output":{"rms":0.032751963870405534,"rms_dbfs":-29.69525307527048,"peak":0.3692138398162722,"peak_dbfs":-8.654440560117274,"crest_factor_db":21.0408125151532,"mean":0.0010570961367627175,"zero_energy":false},"rms_difference_db":-0.2544719697471187},"full_scale":{"reference":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":83600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":522,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.1876817041775212e-06,"threshold_dbfs":-59.252999336316485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":9.875744581222535e-07,"output_mean_power":1.0553784637429656e-06,"output_rms_dbfs":-59.765917723656614,"output_zero_energy":false,"output_peak":0.0034601430334643916,"output_energy_fraction":9.414933237253394e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":515,"reference_estimated_voiced_frames":178,"output_estimated_voiced_frames":164,"both_estimated_voiced_frames":141,"reference_voiced_output_unvoiced_frames":37,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.7921348314606742,"absolute_difference_cents_median":11.719332621628787,"absolute_difference_cents_p95":43.824518469945865,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":83600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011323786448874551,"log_rms_mae_db":2.9425175430338655,"rms_envelope_nrmse":0.37430565362941026,"envelope_correlation":0.8379653664926567,"rms_gain_error_db":-0.25443288667504516,"reference_mean_power":0.0011384943575262435,"output_mean_power":0.0010737114733933466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":5.62759551802579e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002928658523286973,"output_power":0.000293427028407028,"reference_active":true,"delta_db":0.008313788436811463},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4610498849864094e-05,"output_power":4.419287704948806e-05,"reference_active":true,"delta_db":-0.040848037290250196},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2101686887119984e-05,"output_power":1.4013016022299311e-05,"reference_active":true,"delta_db":0.6368570654671541}],"active_band_count":3,"active_band_mae_db":0.22867296373140525}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.563924945576787e-05,"log_rms_mae_db":2.151549746270469,"rms_envelope_nrmse":0.2634675935522245,"envelope_correlation":0.9135209976702593,"rms_gain_error_db":-0.25419244861023954,"reference_mean_power":0.0011449801518647084,"output_mean_power":0.0010798879955648984},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":5.095010942049215e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002727818973842863,"output_power":0.0002739588122943041,"reference_active":true,"delta_db":0.018697288550644064},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1267992599082604e-05,"output_power":2.0810193086638874e-05,"reference_active":true,"delta_db":-0.09450390698435943},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.757168583210849e-07,"output_power":7.519094429643446e-07,"reference_active":true,"delta_db":1.1595659222227037}],"active_band_count":3,"active_band_mae_db":0.4242557059192357}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.979916609312644e-05,"log_rms_mae_db":4.014177407905065,"rms_envelope_nrmse":0.3666214262849628,"envelope_correlation":0.8857784561607813,"rms_gain_error_db":-0.39800086950540225,"reference_mean_power":0.00038382289364078325,"output_mean_power":0.00035021181067824183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":2.8601579771498626e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014700129208541836,"output_power":0.00013531027909815076,"reference_active":true,"delta_db":-0.3599036220564216},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.957621157356578e-05,"output_power":1.8046924482529202e-05,"reference_active":true,"delta_db":-0.3532544905795493},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3113428408893907e-06,"output_power":2.0002355538765894e-06,"reference_active":true,"delta_db":-0.627832265195689}],"active_band_count":3,"active_band_mae_db":0.4469967926105533}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":6.59156861629162e-05,"log_rms_mae_db":3.1315903591149326,"rms_envelope_nrmse":0.2880304803791428,"envelope_correlation":0.9257501293667795,"rms_gain_error_db":-0.39800254539328006,"reference_mean_power":0.00038603855676420007,"output_mean_power":0.00035223331392105487},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":2.6676069300397395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013762626674571402,"output_power":0.00012627959640033865,"reference_active":true,"delta_db":-0.37368144167904926},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.746730510582487e-06,"output_power":9.014460453160047e-06,"reference_active":true,"delta_db":-0.33919220530811095},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.594967364659404e-07,"output_power":1.2479267331211788e-07,"reference_active":true,"delta_db":-1.0656271290756234}],"active_band_count":3,"active_band_mae_db":0.5928335920209279}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":6.34102616150739e-05,"log_rms_mae_db":3.677801215885031,"rms_envelope_nrmse":0.5225456172842117,"envelope_correlation":0.7839918463029096,"rms_gain_error_db":-0.033675068359716814,"reference_mean_power":0.00014337292372884635,"output_mean_power":0.00014226551319328838},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.145508332328126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5890331523222755e-05,"output_power":5.589015623583748e-05,"reference_active":true,"delta_db":-1.3620685630708243e-05},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4083924473542483e-05,"output_power":1.4455979379495217e-05,"reference_active":true,"delta_db":0.11323832604098222},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.499188380560451e-06,"output_power":5.7926532833856444e-06,"reference_active":true,"delta_db":0.22578937622114814}],"active_band_count":3,"active_band_mae_db":0.11301377431592036}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":5.482542724448805e-05,"log_rms_mae_db":2.5465917725695686,"rms_envelope_nrmse":0.33183222219495867,"envelope_correlation":0.9061516002518187,"rms_gain_error_db":-0.033525679138871264,"reference_mean_power":0.00014418079784839883,"output_mean_power":0.00014307206863230508},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.0427751966205221e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.204944813792335e-05,"output_power":5.237568138222663e-05,"reference_active":true,"delta_db":0.027135567731810185},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.197217517258329e-06,"output_power":6.566691499759967e-06,"reference_active":true,"delta_db":0.25149874007878237},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.536643030241646e-07,"output_power":2.3777826852172988e-07,"reference_active":true,"delta_db":-0.2808719507721681}],"active_band_count":3,"active_band_mae_db":0.18650208619425354}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1044,"rms_floor":0.00011141137328067573,"log_rms_mae_db":3.7606917326000433,"rms_envelope_nrmse":0.6531108738823928,"envelope_correlation":0.76558357988138,"rms_gain_error_db":0.3396087682707635,"reference_mean_power":0.00018590462562984242,"output_mean_power":0.00020102546377846067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1044,"frequency_resolution_hz":0.19157088122605365,"power_floor":1.8616455925800917e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.396584441083371e-05,"output_power":9.851219290589119e-05,"reference_active":true,"delta_db":0.2051996584161149},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7617019162087808e-05,"output_power":1.7488247672852946e-05,"reference_active":true,"delta_db":-0.031861314881944436},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.489954892102077e-06,"output_power":7.82478971524513e-06,"reference_active":true,"delta_db":0.8123099678737317}],"active_band_count":3,"active_band_mae_db":0.3497903137239304}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":83600,"uncovered_tail_samples":0,"metrics":{"frame_count":1038,"rms_floor":9.054702623382986e-05,"log_rms_mae_db":2.588092353145416,"rms_envelope_nrmse":0.4735877116958821,"envelope_correlation":0.8739129623002989,"rms_gain_error_db":0.33960258400847587,"reference_mean_power":0.00018697687803543296,"output_mean_power":0.00020218464157281823},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1038,"frequency_resolution_hz":0.19267822736030826,"power_floor":1.7928912528322635e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.176227386223459e-05,"output_power":9.677637059350236e-05,"reference_active":true,"delta_db":0.23105163910761473},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.193273720945635e-06,"output_power":7.603722877390857e-06,"reference_active":true,"delta_db":-0.3243118367691847},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7870970633055527e-07,"output_power":5.640814542469583e-07,"reference_active":true,"delta_db":1.7303538508428609}],"active_band_count":3,"active_band_mae_db":0.76190577557322}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},{"case":"librispeech_6829-68769-0014","method":"hnm","status":"measured","mos":{"status":"measured","value":2.50319242477417,"predicted_mos":2.50319242477417,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":58480,"input_sample_rate":16000,"input_seconds":3.655,"input_float32_sha256":"1fa44d2fba22b6b3e0c90d7499922da90050e74afda73fbf3855b2d0e3e5e3f7","internal_samples":58480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.281489471999521},"started_at":"2026-10-13T04:39:53.323501+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3939999917056412e-05,"native_input_samples":58480,"native_input_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","output_sha256":"c4abaa8d73d241efb5a2e20e2256aacc45cb7e1ff6aa4d6d2043e16a5d723af6","output_samples":58480,"length_delta_samples":0,"analysis_s":0.5759753120000823,"synthesis_s":0.2877630200000567,"total_s":0.863738332000139,"rtf":0.23631691709990127,"output_peak":0.5123225870076662,"output_rms":0.0656403076305978,"finished_at":"2026-10-13T04:39:54.191787+00:00","evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"c4abaa8d73d241efb5a2e20e2256aacc45cb7e1ff6aa4d6d2043e16a5d723af6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.834792649759766,"spectral_convergence":0.28129816608444597,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4803702077815135,"spectral_convergence":0.2602870619510084,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.701015608818534,"spectral_convergence":0.27349410670781205,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.00031174759578423496,"delta_db":0.23175093122498852,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.0014839976055347117,"delta_db":0.05259598432231897,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.002056351041836031,"delta_db":-0.10785262366480765,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.0001753739098884444,"delta_db":-0.083058881577482,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00016138603513608676,"delta_db":0.38671257807641296,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":0.0001197937976600072,"delta_db":0.10389165657154217,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16097710923959205,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.00448185328975999,"pitch":null,"output_rms_dbfs":-23.65658784489625,"output_peak":0.5123225870076662,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.672059488786605},"common_prefix":{"samples":58480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.834792649759766,"spectral_convergence":0.28129816608444597,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4803702077815135,"spectral_convergence":0.2602870619510084,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.701015608818534,"spectral_convergence":0.27349410670781205,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.00031174759578423496,"delta_db":0.23175093122498852,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.0014839976055347117,"delta_db":0.05259598432231897,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.002056351041836031,"delta_db":-0.10785262366480765,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.0001753739098884444,"delta_db":-0.083058881577482,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00016138603513608676,"delta_db":0.38671257807641296,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":0.0001197937976600072,"delta_db":0.10389165657154217,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16097710923959205,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.00448185328975999,"pitch":null,"output_rms_dbfs":-23.65658784489625,"output_peak":0.5123225870076662,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.672059488786605},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58480,"output_samples":58480,"intelligibility":{"estoi":{"value":0.8884223641856455,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06567418627216229,"rms_dbfs":-23.65210599160649,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.882286904150295,"mean":5.537834115295828e-06,"zero_energy":false},"output":{"rms":0.0656403076305978,"rms_dbfs":-23.65658784489625,"peak":0.5123225870076662,"peak_dbfs":-5.80912993511605,"crest_factor_db":17.8474579097802,"mean":1.1748696533737522e-05,"zero_energy":false},"rms_difference_db":-0.0044818532897608065},"full_scale":{"reference":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":2.3189345688321815e-06,"output_rms_dbfs":-56.347115052628425,"output_zero_energy":false,"output_peak":0.02530604231506393,"output_energy_fraction":7.215326331737016e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":121,"both_estimated_voiced_frames":117,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9285714285714286,"absolute_difference_cents_median":0.9008807898771091,"absolute_difference_cents_p95":6.676378221091781,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.0002136685864137289,"log_rms_mae_db":0.8730500784300653,"rms_envelope_nrmse":0.07713792833248442,"envelope_correlation":0.9940600349580181,"rms_gain_error_db":-0.0044810292496766995,"reference_mean_power":0.004318997149502425,"output_mean_power":0.004314543127555379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.9590597233654055e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007496792113024279,"output_power":0.0007165744937434853,"reference_active":true,"delta_db":-0.19614122451235103},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031885669741703006,"output_power":0.00029736809904773773,"reference_active":true,"delta_db":-0.3030116695467876},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.908796664605564e-05,"output_power":5.820059492985899e-05,"reference_active":true,"delta_db":-0.744749869316846}],"active_band_count":3,"active_band_mae_db":0.4146342544586615}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00017810705079567896,"log_rms_mae_db":0.4642293271189296,"rms_envelope_nrmse":0.041935284715552856,"envelope_correlation":0.9980499472162577,"rms_gain_error_db":-0.004476293672506006,"reference_mean_power":0.0043546897567722925,"output_mean_power":0.004350203669849767},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.684242583835521e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006964729892166044,"output_power":0.0006563355134201166,"reference_active":true,"delta_db":-0.25778374198049187},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001781987747792959,"output_power":0.00016373279160846333,"reference_active":true,"delta_db":-0.36769047169851404},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.211993913505217e-06,"output_power":4.520053972121675e-06,"reference_active":true,"delta_db":-0.6186027930203127}],"active_band_count":3,"active_band_mae_db":0.4146923355664396}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.00018825872292702335,"log_rms_mae_db":1.6328632563435035,"rms_envelope_nrmse":0.08691806439619408,"envelope_correlation":0.9939406749961565,"rms_gain_error_db":-0.05664504093148861,"reference_mean_power":0.003284629583795886,"output_mean_power":0.0032420663310312754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":2.043578979045215e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008901806972697356,"output_power":0.0008610922418187184,"reference_active":true,"delta_db":-0.1442849639522801},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027181233736104136,"output_power":0.00022850571033999773,"reference_active":true,"delta_db":-0.7537210762237689},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0768894288562865e-05,"output_power":4.338939386360282e-05,"reference_active":true,"delta_db":-1.462977510659792}],"active_band_count":3,"active_band_mae_db":0.7869945169452803}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00015953911997549416,"log_rms_mae_db":1.0331210896763254,"rms_envelope_nrmse":0.05397972139856654,"envelope_correlation":0.997455592761269,"rms_gain_error_db":-0.056645601150307395,"reference_mean_power":0.0033118454961761057,"output_mean_power":0.003268929149469592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.8056141501282585e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008648498098004976,"output_power":0.0008091588032447266,"reference_active":true,"delta_db":-0.2890693084329907},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013850439438479803,"output_power":0.00011545642835097313,"reference_active":true,"delta_db":-0.7904543420129028},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4525026177578845e-06,"output_power":4.0743162372523e-06,"reference_active":true,"delta_db":-1.9967345468869184}],"active_band_count":3,"active_band_mae_db":1.0254193991109373}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":1.5081332417165083,"rms_envelope_nrmse":0.1809619120238298,"envelope_correlation":0.9740835266018466,"rms_gain_error_db":-0.09213345649656306,"reference_mean_power":0.00023215995390589003,"output_mean_power":0.0002272866686511824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.4445004803909088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.349295087979124e-05,"output_power":6.765030656783064e-05,"reference_active":true,"delta_db":-0.35975916456832446},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7764968189144563e-05,"output_power":2.3140654867476004e-05,"reference_active":true,"delta_db":-0.791215351509384},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.892866752425831e-06,"output_power":5.50421441736382e-06,"reference_active":true,"delta_db":-0.2963127752843567}],"active_band_count":3,"active_band_mae_db":0.4824290971206884}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":0.864543441439185,"rms_envelope_nrmse":0.111143787269638,"envelope_correlation":0.9893464751353921,"rms_gain_error_db":-0.09224932938024698,"reference_mean_power":0.0002340443801325015,"output_mean_power":0.00022912542544744764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.2566052761381825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.967443215044191e-05,"output_power":6.340507949348746e-05,"reference_active":true,"delta_db":-0.40949386504010393},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5348913604759922e-05,"output_power":1.2972457089717947e-05,"reference_active":true,"delta_db":-0.7305539875162158},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.927931708651049e-07,"output_power":4.5930538013441526e-07,"reference_active":true,"delta_db":-0.30563148125190015}],"active_band_count":3,"active_band_mae_db":0.48189311126940665}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":1.6679032068519692,"rms_envelope_nrmse":0.16503251663530455,"envelope_correlation":0.9812076447286039,"rms_gain_error_db":0.3824185056311825,"reference_mean_power":0.00012513898539343645,"output_mean_power":0.00013665781734514145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":8.947237162820929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5394872382848794e-05,"output_power":5.1586209453451094e-05,"reference_active":true,"delta_db":0.5552681763763547},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2730550619395117e-05,"output_power":1.2950128632359158e-05,"reference_active":true,"delta_db":0.07426894159623496},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6846121221132666e-06,"output_power":1.5943553146701424e-06,"reference_active":true,"delta_db":-0.23914807657462767}],"active_band_count":3,"active_band_mae_db":0.2895617315157391}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":1.0321001670197805,"rms_envelope_nrmse":0.09995887219180227,"envelope_correlation":0.9936503440639001,"rms_gain_error_db":0.3828680178647992,"reference_mean_power":0.00012613106271167372,"output_mean_power":0.00013775547120931534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":8.199333001241022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.113815724379872e-05,"output_power":4.703328057457939e-05,"reference_active":true,"delta_db":0.5816043800881527},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.547615689423206e-06,"output_power":7.176562282475545e-06,"reference_active":true,"delta_db":0.3983327758256133},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0611878817701006e-07,"output_power":1.186952945659598e-07,"reference_active":true,"delta_db":0.4864122065771856}],"active_band_count":3,"active_band_mae_db":0.4887831208303172}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0014","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.1230132579803467,"predicted_mos":3.1230132579803467,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":58480,"input_sample_rate":16000,"input_seconds":3.655,"input_float32_sha256":"b5ea7359ffc54f53c6fbd25fe3ef7b68cabd667f822d1cbcd109274ed15216d9","internal_samples":58480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.422735067000758},"started_at":"2026-10-13T04:43:04.526531+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.9583000721468125e-05,"native_input_samples":58480,"native_input_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","output_sha256":"28b85991ddac771f60007366b72ae59dcf166daa3888d80bd3613e6f750a50d0","output_samples":58480,"length_delta_samples":0,"analysis_s":1.0094638349992238,"synthesis_s":0.06325096399996255,"total_s":1.0727147989991863,"rtf":0.29349242106680884,"output_peak":0.6085572133565414,"output_rms":0.07435943490098641,"finished_at":"2026-10-13T04:43:05.603404+00:00","evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"28b85991ddac771f60007366b72ae59dcf166daa3888d80bd3613e6f750a50d0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.214671698079062,"spectral_convergence":0.5246596916243569,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.486464151752095,"spectral_convergence":0.38013286411335045,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.929076094945893,"spectral_convergence":0.4316565254816777,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.00022502217143727203,"delta_db":-1.1840270847629535,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.001784210065031066,"delta_db":0.8527238320206848,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.0030409639491647423,"delta_db":1.5912875370030184,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.00020603939838296847,"delta_db":0.6167940044904446,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00015643279863103508,"delta_db":0.2513312068312435,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":0.00011665717614695308,"delta_db":-0.01133707818037244,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7512501238814528,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0788276049119743,"pitch":null,"output_rms_dbfs":-22.57327838669451,"output_peak":0.6085572133565414,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.876737314925684},"common_prefix":{"samples":58480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.214671698079062,"spectral_convergence":0.5246596916243569,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.486464151752095,"spectral_convergence":0.38013286411335045,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.929076094945893,"spectral_convergence":0.4316565254816777,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.00022502217143727203,"delta_db":-1.1840270847629535,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.001784210065031066,"delta_db":0.8527238320206848,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.0030409639491647423,"delta_db":1.5912875370030184,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.00020603939838296847,"delta_db":0.6167940044904446,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00015643279863103508,"delta_db":0.2513312068312435,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":0.00011665717614695308,"delta_db":-0.01133707818037244,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7512501238814528,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0788276049119743,"pitch":null,"output_rms_dbfs":-22.57327838669451,"output_peak":0.6085572133565414,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.876737314925684},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58480,"output_samples":58480,"intelligibility":{"estoi":{"value":0.8923407940358875,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06567418627216229,"rms_dbfs":-23.65210599160649,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.882286904150295,"mean":5.537834115295828e-06,"zero_energy":false},"output":{"rms":0.07435943490098641,"rms_dbfs":-22.57327838669451,"peak":0.6085572133565414,"peak_dbfs":-4.313971708514403,"crest_factor_db":18.25930667818011,"mean":-5.46363812066105e-09,"zero_energy":false},"rms_difference_db":1.0788276049119787},"full_scale":{"reference":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":2.1119660655775184e-06,"output_rms_dbfs":-56.75313064192595,"output_zero_energy":false,"output_peak":0.02827720274624699,"output_energy_fraction":5.120630907226036e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":148,"both_estimated_voiced_frames":114,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9047619047619048,"absolute_difference_cents_median":4.144351963473142,"absolute_difference_cents_p95":19.446814711203935,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.0002136685864137289,"log_rms_mae_db":1.4665842409352317,"rms_envelope_nrmse":0.19635081176296967,"envelope_correlation":0.9821659058072739,"rms_gain_error_db":1.078830771288045,"reference_mean_power":0.004318997149502425,"output_mean_power":0.005536891261754448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.9590597233654055e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007496792113024279,"output_power":0.0009711320151101908,"reference_active":true,"delta_db":1.12402803682248},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031885669741703006,"output_power":0.00043930310417902087,"reference_active":true,"delta_db":1.3916872851755155},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.908796664605564e-05,"output_power":7.023566081170292e-05,"reference_active":true,"delta_db":0.0715526184568826}],"active_band_count":3,"active_band_mae_db":0.8624226468182927}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00017810705079567896,"log_rms_mae_db":1.0746207453312904,"rms_envelope_nrmse":0.16560027699365945,"envelope_correlation":0.9908842676321277,"rms_gain_error_db":1.0788566306127985,"reference_mean_power":0.0043546897567722925,"output_mean_power":0.005582681903971919},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.684242583835521e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006964729892166044,"output_power":0.0009094850190315551,"reference_active":true,"delta_db":1.1589127161929467},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001781987747792959,"output_power":0.0002383391747879807,"reference_active":true,"delta_db":1.2629071765218853},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.211993913505217e-06,"output_power":5.2662476176453425e-06,"reference_active":true,"delta_db":0.0449737579030007}],"active_band_count":3,"active_band_mae_db":0.8222645502059441}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.00018825872292702335,"log_rms_mae_db":2.193856788347967,"rms_envelope_nrmse":0.23444838528282216,"envelope_correlation":0.9830651608231689,"rms_gain_error_db":1.3955514959343525,"reference_mean_power":0.003284629583795886,"output_mean_power":0.0045294091150286046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":2.043578979045215e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008901806972697356,"output_power":0.0012202419449076779,"reference_active":true,"delta_db":1.369677766805071},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027181233736104136,"output_power":0.0003636859391311999,"reference_active":true,"delta_db":1.2645734548534189},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0768894288562865e-05,"output_power":5.178242878542987e-05,"reference_active":true,"delta_db":-0.6949891770086156}],"active_band_count":3,"active_band_mae_db":1.1097467995557018}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00015953911997549416,"log_rms_mae_db":1.7688608865982816,"rms_envelope_nrmse":0.20641245731378752,"envelope_correlation":0.9907293806696189,"rms_gain_error_db":1.3955540289714263,"reference_mean_power":0.0033118454961761057,"output_mean_power":0.0045669417340580875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.8056141501282585e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008648498098004976,"output_power":0.0011706233698172681,"reference_active":true,"delta_db":1.3147649568994018},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013850439438479803,"output_power":0.00018169323962272736,"reference_active":true,"delta_db":1.1787521587912821},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4525026177578845e-06,"output_power":4.644460606048619e-06,"reference_active":true,"delta_db":-1.4279290582917805}],"active_band_count":3,"active_band_mae_db":1.3071487246608215}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":1.5405357486032447,"rms_envelope_nrmse":0.1481055288229766,"envelope_correlation":0.986372913024835,"rms_gain_error_db":0.4991179529765536,"reference_mean_power":0.00023215995390589003,"output_mean_power":0.00026043485324587996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.4445004803909088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.349295087979124e-05,"output_power":8.530917650330902e-05,"reference_active":true,"delta_db":0.6475006422747692},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7764968189144563e-05,"output_power":2.9813024654656667e-05,"reference_active":true,"delta_db":0.30908859009908424},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.892866752425831e-06,"output_power":5.664642855037141e-06,"reference_active":true,"delta_db":-0.17154087460385073}],"active_band_count":3,"active_band_mae_db":0.376043368992568}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":1.0194306223945808,"rms_envelope_nrmse":0.11048904268669968,"envelope_correlation":0.9929341930563846,"rms_gain_error_db":0.4992356151775791,"reference_mean_power":0.0002340443801325015,"output_mean_power":0.00026255589812512503},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.2566052761381825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.967443215044191e-05,"output_power":7.993387360225534e-05,"reference_active":true,"delta_db":0.5965742154561672},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5348913604759922e-05,"output_power":1.6585473339584902e-05,"reference_active":true,"delta_db":0.33650228977574315},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.927931708651049e-07,"output_power":4.679227738916627e-07,"reference_active":true,"delta_db":-0.22490497836076162}],"active_band_count":3,"active_band_mae_db":0.38599382786422404}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":1.5243240841825305,"rms_envelope_nrmse":0.18962651803267852,"envelope_correlation":0.9741632846131116,"rms_gain_error_db":0.3101976058912283,"reference_mean_power":0.00012513898539343645,"output_mean_power":0.00013440406068817348},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":8.947237162820929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5394872382848794e-05,"output_power":4.865800751875342e-05,"reference_active":true,"delta_db":0.3014752165463473},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2730550619395117e-05,"output_power":1.4280742512222487e-05,"reference_active":true,"delta_db":0.49903600629478906},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6846121221132666e-06,"output_power":2.9386722607457647e-06,"reference_active":true,"delta_db":2.416512320569575}],"active_band_count":3,"active_band_mae_db":1.0723411811369037}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":0.8887604397495393,"rms_envelope_nrmse":0.10710239291976238,"envelope_correlation":0.9918779899631104,"rms_gain_error_db":0.3103055050568248,"reference_mean_power":0.00012613106271167372,"output_mean_power":0.0001354729554483826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":8.199333001241022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.113815724379872e-05,"output_power":4.529340539840009e-05,"reference_active":true,"delta_db":0.41790140608686077},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.547615689423206e-06,"output_power":8.068734081169115e-06,"reference_active":true,"delta_db":0.9072222210255261},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0611878817701006e-07,"output_power":2.0368811724855236e-07,"reference_active":true,"delta_db":2.8317341196546253}],"active_band_count":3,"active_band_mae_db":1.3856192489223373}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0014","method":"facodec","status":"measured","mos":{"status":"measured","value":3.7274997234344482,"predicted_mos":3.7274997234344482,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":58600,"input_sample_rate":16000,"input_seconds":3.6625,"input_float32_sha256":"31e801a0c33df37a1cbcefd965ad88a6497f297bece2054e6da0abdac242c004","internal_samples":58600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":14.716249025001161},"started_at":"2026-10-13T06:57:18.685193+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":9.184200098388828e-05,"native_input_samples":58480,"native_input_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","output_sha256":"b6d46ea52b5076b4dc65e5c74de982b4aa37239764e065ec5a8889a3d3286864","output_samples":58600,"length_delta_samples":120,"analysis_s":1.1405285660002846,"synthesis_s":8.352606147000188,"total_s":9.493134713000472,"rtf":2.597300879069897,"output_peak":0.5722711682319641,"output_rms":0.05982935805176779,"finished_at":"2026-10-13T06:57:28.190975+00:00","evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"b6d46ea52b5076b4dc65e5c74de982b4aa37239764e065ec5a8889a3d3286864","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":58480,"output_samples":58600,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.46171313923816,"output_peak":0.5722711682319641},"common_prefix":{"samples":58480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9979522184300341,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.530295695107587,"spectral_convergence":0.28280915928686523,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.875934232808447,"spectral_convergence":0.27628528689066856,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.978678803987442,"spectral_convergence":0.3335257673004011,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.0002852539843969214,"delta_db":-0.15396300425297543,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.0014037938383137517,"delta_db":-0.18870269939616377,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.0015228511105050406,"delta_db":-1.4122507354098637,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.00014848902739903573,"delta_db":-0.8057651002379845,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00012032071436319829,"delta_db":-0.8885429262151973,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":0.00010618854940075236,"delta_db":-0.4196747920446434,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.644816542926138,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8007046785579092,"pitch":null,"output_rms_dbfs":-24.4528106701644,"output_peak":0.5722711682319641,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.794969577301159},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58480,"output_samples":58480,"intelligibility":{"estoi":{"value":0.9079551249119345,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06567418627216229,"rms_dbfs":-23.65210599160649,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.882286904150295,"mean":5.537834115295828e-06,"zero_energy":false},"output":{"rms":0.059890710668506014,"rms_dbfs":-24.4528106701644,"peak":0.5722711682319641,"peak_dbfs":-4.847962677260505,"crest_factor_db":19.604847992903892,"mean":0.00019178207627417842,"zero_energy":false},"rms_difference_db":-0.8007046785579099},"full_scale":{"reference":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":1.116032640339415e-06,"output_rms_dbfs":-59.52323103505967,"output_zero_energy":false,"output_peak":0.007839307188987732,"output_energy_fraction":4.1712542508534166e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":131,"both_estimated_voiced_frames":120,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9523809523809523,"absolute_difference_cents_median":5.400797266107087,"absolute_difference_cents_p95":22.650721827204205,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.0002136685864137289,"log_rms_mae_db":1.3116595309921328,"rms_envelope_nrmse":0.12917532545572238,"envelope_correlation":0.9901898780081438,"rms_gain_error_db":-0.8007086542002602,"reference_mean_power":0.004318997149502425,"output_mean_power":0.0035917992216081248},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.9590597233654055e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007496792113024279,"output_power":0.0006213827080689406,"reference_active":true,"delta_db":-0.8151630449797382},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031885669741703006,"output_power":0.00025138000231199895,"reference_active":true,"delta_db":-1.0326481765423632},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.908796664605564e-05,"output_power":5.510051635120441e-05,"reference_active":true,"delta_db":-0.9824674230959507}],"active_band_count":3,"active_band_mae_db":0.9434262148726841}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00017810705079567896,"log_rms_mae_db":1.0967101348318486,"rms_envelope_nrmse":0.11631978185416925,"envelope_correlation":0.9929556217036868,"rms_gain_error_db":-0.8007550118265817,"reference_mean_power":0.0043546897567722925,"output_mean_power":0.003621443539000448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.684242583835521e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006964729892166044,"output_power":0.0005666843706162168,"reference_active":true,"delta_db":-0.8956304338378401},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001781987747792959,"output_power":0.0001450193514691947,"reference_active":true,"delta_db":-0.8947875506484595},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.211993913505217e-06,"output_power":3.6208579452245918e-06,"reference_active":true,"delta_db":-1.581924131207121}],"active_band_count":3,"active_band_mae_db":1.1241140385644737}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.00018825872292702335,"log_rms_mae_db":1.558735197908381,"rms_envelope_nrmse":0.13911824102752596,"envelope_correlation":0.9919289967069783,"rms_gain_error_db":-0.933742308984393,"reference_mean_power":0.003284629583795886,"output_mean_power":0.0026491842795189863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":2.043578979045215e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008901806972697356,"output_power":0.0007153513638391299,"reference_active":true,"delta_db":-0.9495876329570268},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027181233736104136,"output_power":0.00021537451907103163,"reference_active":true,"delta_db":-1.0107484446701784},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0768894288562865e-05,"output_power":4.78342370569639e-05,"reference_active":true,"delta_db":-1.0394248280279865}],"active_band_count":3,"active_band_mae_db":0.999920301885064}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00015953911997549416,"log_rms_mae_db":1.193778360355468,"rms_envelope_nrmse":0.12953842446862124,"envelope_correlation":0.9936776034764114,"rms_gain_error_db":-0.9337433519180396,"reference_mean_power":0.0033118454961761057,"output_mean_power":0.0026711343528650214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.8056141501282585e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008648498098004976,"output_power":0.0006920415582301106,"reference_active":true,"delta_db":-0.9680851890596341},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013850439438479803,"output_power":0.00010712832731985123,"reference_active":true,"delta_db":-1.1155922867657395},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4525026177578845e-06,"output_power":5.0953970716251565e-06,"reference_active":true,"delta_db":-1.0255015625732815}],"active_band_count":3,"active_band_mae_db":1.0363930127995518}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":1.669529036475895,"rms_envelope_nrmse":0.17937042792384872,"envelope_correlation":0.9808305108152191,"rms_gain_error_db":-0.9254352229468411,"reference_mean_power":0.00023215995390589003,"output_mean_power":0.00018760473116849276},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.4445004803909088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.349295087979124e-05,"output_power":5.851461594920174e-05,"reference_active":true,"delta_db":-0.9898132650921956},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7764968189144563e-05,"output_power":2.3731994037420556e-05,"reference_active":true,"delta_db":-0.6816294967889109},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.892866752425831e-06,"output_power":5.934076438230134e-06,"reference_active":true,"delta_db":0.030265151844171563}],"active_band_count":3,"active_band_mae_db":0.5672359712417594}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":1.3321700168868311,"rms_envelope_nrmse":0.1481685300883284,"envelope_correlation":0.988839655105972,"rms_gain_error_db":-0.9258372716040146,"reference_mean_power":0.0002340443801325015,"output_mean_power":0.00018910999805370267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.2566052761381825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.967443215044191e-05,"output_power":5.476652613888835e-05,"reference_active":true,"delta_db":-1.0455824353944547},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5348913604759922e-05,"output_power":1.3470663455592283e-05,"reference_active":true,"delta_db":-0.5668865548963152},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.927931708651049e-07,"output_power":4.3899820188897e-07,"reference_active":true,"delta_db":-0.5020193933795682}],"active_band_count":3,"active_band_mae_db":0.7048294612234461}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":1.8709065493959827,"rms_envelope_nrmse":0.19789833063278242,"envelope_correlation":0.9710994422805064,"rms_gain_error_db":-0.6458419227172361,"reference_mean_power":0.00012513898539343645,"output_mean_power":0.00010784709149322912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":8.947237162820929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5394872382848794e-05,"output_power":3.6258337693333526e-05,"reference_active":true,"delta_db":-0.9759891004967165},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2730550619395117e-05,"output_power":1.23424609273834e-05,"reference_active":true,"delta_db":-0.13445427039012853},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6846121221132666e-06,"output_power":2.495799491002481e-06,"reference_active":true,"delta_db":1.707097703498564}],"active_band_count":3,"active_band_mae_db":0.9391803581284698}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":1.5414941666124296,"rms_envelope_nrmse":0.15877587701427823,"envelope_correlation":0.9815791088641403,"rms_gain_error_db":-0.6465048289482525,"reference_mean_power":0.00012613106271167372,"output_mean_power":0.00010868549107220977},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":8.199333001241022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.113815724379872e-05,"output_power":3.252267460368292e-05,"reference_active":true,"delta_db":-1.0205857991065606},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.547615689423206e-06,"output_power":6.795228591386019e-06,"reference_active":true,"delta_db":0.1612089028615367},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0611878817701006e-07,"output_power":1.9594774569144163e-07,"reference_active":true,"delta_db":2.6634798952569505}],"active_band_count":3,"active_band_mae_db":1.281758199075016}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0014","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.577418327331543,"predicted_mos":3.577418327331543,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":161280,"input_sample_rate":44100,"input_seconds":3.657142857142857,"input_float32_sha256":"7ee9d5609218d79c1399c1087aae50d99276c02ae33d5d89f30cb3e6413c0238","internal_samples":58515,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2877608249982586},"started_at":"2026-10-13T07:24:28.219818+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005998602000545361,"native_input_samples":161186,"native_input_sha256":"b39f97b0e68bc189e87cb264c2c19461d00794445b8313fb49ac6f26d00936a8","output_sha256":"e339408e210cdc045b11e98b2d1bb1268d2b9055753076b1a180bf575a357f96","output_samples":161280,"length_delta_samples":94,"analysis_s":0.09411449099934543,"synthesis_s":20.12487417599914,"total_s":20.218988666998484,"rtf":5.53185388442317,"output_peak":0.47315484285354614,"output_rms":0.060764790674941935,"finished_at":"2026-10-13T07:24:48.459077+00:00","evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"e339408e210cdc045b11e98b2d1bb1268d2b9055753076b1a180bf575a357f96","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":58480,"output_samples":58515,"length_delta_samples":35,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.322676844223615,"output_peak":0.47275248954186333},"common_prefix":{"samples":58480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9994018627702299,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.667669787329999,"spectral_convergence":0.47360460949644617,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.418942510474954,"spectral_convergence":0.3663613602901106,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.366382807580662,"spectral_convergence":0.41878731516028567,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.00023005892756400447,"delta_db":-1.087889287709105,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.0013758587352286715,"delta_db":-0.27599756283196974,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.0016751788193303514,"delta_db":-0.9982134465802783,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.00020168866676973873,"delta_db":0.5241062249112431,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00012515801390503971,"delta_db":-0.7173603151203978,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":9.027079050742676e-05,"delta_db":-1.1249792170039057,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7880910090261498,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6679736798808036,"pitch":null,"output_rms_dbfs":-24.320079671487292,"output_peak":0.47275248954186333,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.484331701795206},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58480,"output_samples":58480,"intelligibility":{"estoi":{"value":0.8034921532614181,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06567418627216229,"rms_dbfs":-23.65210599160649,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.882286904150295,"mean":5.537834115295828e-06,"zero_energy":false},"output":{"rms":0.06081294231744779,"rms_dbfs":-24.320079671487292,"peak":0.47275248954186333,"peak_dbfs":-6.507323509142457,"crest_factor_db":17.812756162344836,"mean":-0.0001265219981787273,"zero_energy":false},"rms_difference_db":-0.6679736798808023},"full_scale":{"reference":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":1.3437490516653515e-06,"output_rms_dbfs":-58.71681829238292,"output_zero_energy":false,"output_peak":0.005986069572506023,"output_energy_fraction":4.871187388333931e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":126,"both_estimated_voiced_frames":107,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.8492063492063492,"absolute_difference_cents_median":4.668109948049593,"absolute_difference_cents_p95":39.4375722515517,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.0002136685864137289,"log_rms_mae_db":1.6794116013029357,"rms_envelope_nrmse":0.1230778154214347,"envelope_correlation":0.9899738938253384,"rms_gain_error_db":-0.6679759581630446,"reference_mean_power":0.004318997149502425,"output_mean_power":0.003703269527783789},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.9590597233654055e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007496792113024279,"output_power":0.0006126599137971041,"reference_active":true,"delta_db":-0.8765600250373738},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031885669741703006,"output_power":0.0002688314850933271,"reference_active":true,"delta_db":-0.7411541234300562},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.908796664605564e-05,"output_power":5.93559286459946e-05,"reference_active":true,"delta_db":-0.6593830691106789}],"active_band_count":3,"active_band_mae_db":0.7590324058593696}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00017810705079567896,"log_rms_mae_db":1.1865064835125867,"rms_envelope_nrmse":0.0987208613904946,"envelope_correlation":0.9951835534279188,"rms_gain_error_db":-0.6679585683114704,"reference_mean_power":0.0043546897567722925,"output_mean_power":0.003733888653911546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.684242583835521e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006964729892166044,"output_power":0.0005718285601401381,"reference_active":true,"delta_db":-0.8563843564812936},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001781987747792959,"output_power":0.0001428979043769311,"reference_active":true,"delta_db":-0.958788538505868},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.211993913505217e-06,"output_power":4.241487252430692e-06,"reference_active":true,"delta_db":-0.89485733787553}],"active_band_count":3,"active_band_mae_db":0.9033434109542305}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.00018825872292702335,"log_rms_mae_db":2.8662709319211532,"rms_envelope_nrmse":0.152049757741928,"envelope_correlation":0.9889104102867021,"rms_gain_error_db":-0.9481449479141237,"reference_mean_power":0.003284629583795886,"output_mean_power":0.002640413261601434},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":2.043578979045215e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008901806972697356,"output_power":0.0006867629502573308,"reference_active":true,"delta_db":-1.1267131514081477},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027181233736104136,"output_power":0.00020451711076883544,"reference_active":true,"delta_db":-1.235395163677184},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0768894288562865e-05,"output_power":4.817638879681347e-05,"reference_active":true,"delta_db":-1.0084709121843551}],"active_band_count":3,"active_band_mae_db":1.1235264090898955}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00015953911997549416,"log_rms_mae_db":2.1949721617782676,"rms_envelope_nrmse":0.13080245600241022,"envelope_correlation":0.993741073343665,"rms_gain_error_db":-0.9481568967222174,"reference_mean_power":0.0033118454961761057,"output_mean_power":0.002662283976370542},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.8056141501282585e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008648498098004976,"output_power":0.0006655034135832547,"reference_active":true,"delta_db":-1.1379040678753025},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013850439438479803,"output_power":0.00010286614295913657,"reference_active":true,"delta_db":-1.2919109670294233},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4525026177578845e-06,"output_power":4.569944362931885e-06,"reference_active":true,"delta_db":-1.4981727665049214}],"active_band_count":3,"active_band_mae_db":1.309329267136549}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":2.932004446247809,"rms_envelope_nrmse":0.2578401509403214,"envelope_correlation":0.9490147139992672,"rms_gain_error_db":0.18838817711986983,"reference_mean_power":0.00023215995390589003,"output_mean_power":0.0002424521985158177},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.4445004803909088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.349295087979124e-05,"output_power":7.33215165050427e-05,"reference_active":true,"delta_db":-0.010142465947285077},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7764968189144563e-05,"output_power":2.9633620017613107e-05,"reference_active":true,"delta_db":0.2828752753294099},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.892866752425831e-06,"output_power":7.162806677614582e-06,"reference_active":true,"delta_db":0.8475660873435817}],"active_band_count":3,"active_band_mae_db":0.38019460954009227}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":2.230436509955827,"rms_envelope_nrmse":0.1994050492609866,"envelope_correlation":0.9666657319526861,"rms_gain_error_db":0.1886693643939162,"reference_mean_power":0.0002340443801325015,"output_mean_power":0.0002444359918596813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.2566052761381825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.967443215044191e-05,"output_power":7.054828992623476e-05,"reference_active":true,"delta_db":0.05413053141588148},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5348913604759922e-05,"output_power":1.6687219164529415e-05,"reference_active":true,"delta_db":0.3630632830335453},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.927931708651049e-07,"output_power":5.39116702942214e-07,"reference_active":true,"delta_db":0.3901810663283948}],"active_band_count":3,"active_band_mae_db":0.26912496025927385}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":3.341373156854195,"rms_envelope_nrmse":0.215574475464242,"envelope_correlation":0.9715074298722981,"rms_gain_error_db":-1.0338971870518192,"reference_mean_power":0.00012513898539343645,"output_mean_power":9.862860959707069e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":8.947237162820929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5394872382848794e-05,"output_power":3.564513563465667e-05,"reference_active":true,"delta_db":-1.050065279153539},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2730550619395117e-05,"output_power":1.1354009483791248e-05,"reference_active":true,"delta_db":-0.49697935436034657},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6846121221132666e-06,"output_power":1.6243819130215455e-06,"reference_active":true,"delta_db":-0.15811776344862283}],"active_band_count":3,"active_band_mae_db":0.5683874656541694}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":2.9322001508424385,"rms_envelope_nrmse":0.17794452384962242,"envelope_correlation":0.9828958267257452,"rms_gain_error_db":-1.0333968422807842,"reference_mean_power":0.00012613106271167372,"output_mean_power":9.942197146897587e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":8.199333001241022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.113815724379872e-05,"output_power":3.2574891764245615e-05,"reference_active":true,"delta_db":-1.0136185249603544},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.547615689423206e-06,"output_power":5.347418825155558e-06,"reference_active":true,"delta_db":-0.8793898006785016},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0611878817701006e-07,"output_power":1.324804551501743e-07,"reference_active":true,"delta_db":0.9635952959168131}],"active_band_count":3,"active_band_mae_db":0.9522012071852229}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0014","method":"golf","status":"measured","mos":{"status":"measured","value":2.5837717056274414,"predicted_mos":2.5837717056274414,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87600,"input_sample_rate":24000,"input_seconds":3.65,"input_float32_sha256":"056fe1f77b6c75ae112377a3ffc2ade8a2d3899c2cee25b66df87173d572f32d","internal_samples":58400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.371215695000501},"started_at":"2026-10-13T05:02:52.099777+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001984335000088322,"native_input_samples":87720,"native_input_sha256":"4245c8b92bb195f6cc812c529da80e476ccfa2ba1b40c071343dcd7c3f4758e0","output_sha256":"b519117e61c02e5e7b34bd9a085ba62df12731e12fea2c2d9803f93b9ef465b2","output_samples":87600,"length_delta_samples":-120,"analysis_s":0.4534106569999494,"synthesis_s":0.14794176200030051,"total_s":0.6013524190002499,"rtf":0.16452870560882352,"output_peak":0.48656749725341797,"output_rms":0.05828398872194704,"finished_at":"2026-10-13T05:02:52.863145+00:00","analysis_stages":{"external_f0_s":0.041584122000131174},"evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"b519117e61c02e5e7b34bd9a085ba62df12731e12fea2c2d9803f93b9ef465b2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":58480,"output_samples":58400,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.685846404639,"output_peak":0.4819540344683076},"common_prefix":{"samples":58400,"sample_rate":16000,"reference_coverage":0.9986320109439124,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58400,"output_samples":58400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.551826173443257,"spectral_convergence":0.5569457683884621,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.69488198012308,"spectral_convergence":0.3427013593385362,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.8277903226865355,"spectral_convergence":0.37662943048809533,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002959511117568644,"output_power":0.00026796271178379446,"delta_db":-0.4314561155906563,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014634244437815745,"output_power":0.0012208770225425006,"delta_db":-0.786983846459059,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002115664440231604,"output_power":0.0015381443977258147,"delta_db":-1.3844967853966483,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001790011839359678,"output_power":0.00016490924194922094,"delta_db":-0.35610908131100427,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001478381389869573,"output_power":0.00011613503140053031,"delta_db":-1.04823245238712,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001171216913966858,"output_power":9.147405268065079e-05,"delta_db":-1.0733941500168962,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8467787385268973,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0396794608035242,"pitch":null,"output_rms_dbfs":-24.685846404639,"output_peak":0.4819540344683076,"reference_rms_dbfs":-23.64616694383548,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.024832825417625},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58400,"output_samples":58400,"intelligibility":{"estoi":{"value":0.8724746006596972,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06571910688749241,"rms_dbfs":-23.64616694383548,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.876347856379283,"mean":4.943429607234589e-06,"zero_energy":false},"output":{"rms":0.05830525240561533,"rms_dbfs":-24.685846404639,"peak":0.4819540344683076,"peak_dbfs":-6.3398875974093,"crest_factor_db":18.345958807229703,"mean":0.0003266366201607475,"zero_energy":false},"rms_difference_db":-1.039679460803523},"full_scale":{"reference":{"sample_count":58400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":1.1027608212464298e-06,"output_rms_dbfs":-59.57518671854199,"output_zero_energy":false,"output_peak":0.0070797038190004245,"output_energy_fraction":4.354809725823267e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":126,"both_estimated_voiced_frames":111,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.8809523809523809,"absolute_difference_cents_median":3.7654758318541166,"absolute_difference_cents_p95":15.742779860417313,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":729,"rms_floor":0.0002136685864137289,"log_rms_mae_db":1.921620291334909,"rms_envelope_nrmse":0.18819401921630455,"envelope_correlation":0.9744712162521808,"rms_gain_error_db":-1.0396786546957362,"reference_mean_power":0.004324917015720181,"output_mean_power":0.0034041596009325077},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":729,"frequency_resolution_hz":0.27434842249657065,"power_floor":1.9596809680457874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007487639114696899,"output_power":0.0005581258874531896,"reference_active":true,"delta_db":-1.2761273754969782},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003185182704763509,"output_power":0.000214593091118109,"reference_active":true,"delta_db":-1.7151861325509643},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.856546958076139e-05,"output_power":6.623861482129911e-05,"reference_active":true,"delta_db":-0.14994212758563505}],"active_band_count":3,"active_band_mae_db":1.047085211877859}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00017810705079567896,"log_rms_mae_db":1.501664093098221,"rms_envelope_nrmse":0.14874916297434612,"envelope_correlation":0.9880694162829199,"rms_gain_error_db":-1.0396412095982352,"reference_mean_power":0.004360704132998852,"output_mean_power":0.0034323573789063672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.6847028114943126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006955921575645038,"output_power":0.0005155692997187283,"reference_active":true,"delta_db":-1.3006762812600992},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017747868466474556,"output_power":0.00012556065743023518,"reference_active":true,"delta_db":-1.502926204190891},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.203847993782119e-06,"output_power":5.194806006177287e-06,"reference_active":true,"delta_db":-0.007552681937344813}],"active_band_count":3,"active_band_mae_db":0.9370517224627783}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":729,"rms_floor":0.00018825872292702335,"log_rms_mae_db":2.321652353657233,"rms_envelope_nrmse":0.20764825880253535,"envelope_correlation":0.9732116049347687,"rms_gain_error_db":-1.0879279059342115,"reference_mean_power":0.0032891350707873043,"output_mean_power":0.0025602885729599308},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":729,"frequency_resolution_hz":0.27434842249657065,"power_floor":2.044909297180059e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008880115790309605,"output_power":0.0006506229600042574,"reference_active":true,"delta_db":-1.3508924355442526},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002718135609231881,"output_power":0.00018017899752318615,"reference_active":true,"delta_db":-1.7856695385061758},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.040515234699662e-05,"output_power":5.51940024302642e-05,"reference_active":true,"delta_db":-0.3918209562849822}],"active_band_count":3,"active_band_mae_db":1.1761276434451369}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00015953911997549416,"log_rms_mae_db":1.7329302586383148,"rms_envelope_nrmse":0.16554709432892842,"envelope_correlation":0.9862904199551401,"rms_gain_error_db":-1.0879258611338913,"reference_mean_power":0.003316425733484736,"output_mean_power":0.002581533054622533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.8066658405103616e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008626863176781035,"output_power":0.0006351009926859038,"reference_active":true,"delta_db":-1.3301011856552625},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001382487828619263,"output_power":9.421879145935645e-05,"reference_active":true,"delta_db":-1.6652378721806225},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.443795608556995e-06,"output_power":5.325095435408076e-06,"reference_active":true,"delta_db":-0.8281436100390649}],"active_band_count":3,"active_band_mae_db":1.2744942226249834}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":729,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":2.024489401534108,"rms_envelope_nrmse":0.20995304323957986,"envelope_correlation":0.9693962666617645,"rms_gain_error_db":-0.8471209708434722,"reference_mean_power":0.0002324755704916027,"output_mean_power":0.0001912780893632963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":729,"frequency_resolution_hz":0.27434842249657065,"power_floor":1.4418310510097085e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.333176165667788e-05,"output_power":5.6693374226108795e-05,"reference_active":true,"delta_db":-1.1175981250817961},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7710650481310886e-05,"output_power":2.265456737225256e-05,"reference_active":true,"delta_db":-0.8749094755631693},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8736681775387674e-06,"output_power":5.099620093549247e-06,"reference_active":true,"delta_db":-0.6137158452115112}],"active_band_count":3,"active_band_mae_db":0.8687411486188257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":1.3971342228447043,"rms_envelope_nrmse":0.14064139275850704,"envelope_correlation":0.9893967925652195,"rms_gain_error_db":-0.8469201518157792,"reference_mean_power":0.00023436216790128763,"output_mean_power":0.0001928392756456931},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.2539352328002064e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.958277283992106e-05,"output_power":5.428412903374479e-05,"reference_active":true,"delta_db":-1.0782885703635634},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5227299246545024e-05,"output_power":1.2967604342307004e-05,"reference_active":true,"delta_db":-0.6976313138640128},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.930263438227798e-07,"output_power":3.987882346134675e-07,"reference_active":true,"delta_db":-0.9212778859997601}],"active_band_count":3,"active_band_mae_db":0.8990659234091121}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":729,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":2.716878462437561,"rms_envelope_nrmse":0.19180923286546464,"envelope_correlation":0.9736610525660785,"rms_gain_error_db":-0.35202992819119955,"reference_mean_power":0.00012530976595709264,"output_mean_power":0.00011555318647958161},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":729,"frequency_resolution_hz":0.27434842249657065,"power_floor":8.940845580846751e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.53666588626671e-05,"output_power":4.82854720117513e-05,"reference_active":true,"delta_db":0.27079685405762105},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2691028994013653e-05,"output_power":1.2601315825356929e-05,"reference_active":true,"delta_db":-0.030809399272141012},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6744029160317476e-06,"output_power":3.979891081405755e-06,"reference_active":true,"delta_db":3.7601121512893543}],"active_band_count":3,"active_band_mae_db":1.3539061348730386}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58400,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":2.3150225720149296,"rms_envelope_nrmse":0.14170013174012736,"envelope_correlation":0.9851823583437115,"rms_gain_error_db":-0.35114077277236966,"reference_mean_power":0.0001263040947362902,"output_mean_power":0.00011649394514525225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.19390253963545e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.1107747408701164e-05,"output_power":4.502612430312275e-05,"reference_active":true,"delta_db":0.39540886676571235},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.489814321264054e-06,"output_power":6.428660408170713e-06,"reference_active":true,"delta_db":-0.04111786561371257},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0573378186506846e-07,"output_power":2.7053708720875795e-07,"reference_active":true,"delta_db":4.080130435535499}],"active_band_count":3,"active_band_mae_db":1.5055523893049747}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0014","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.270075798034668,"predicted_mos":3.270075798034668,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87840,"input_sample_rate":24000,"input_seconds":3.66,"input_float32_sha256":"d0231b7433265585d771a21e00286eac0b6115c6abc5f5f17a7be18a90d36ae5","internal_samples":58560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.373900782000419},"started_at":"2026-10-13T07:42:46.076024+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00362729599874001,"native_input_samples":87720,"native_input_sha256":"4245c8b92bb195f6cc812c529da80e476ccfa2ba1b40c071343dcd7c3f4758e0","output_sha256":"bea9a11d13c83ac4354113329265ca551002bdf96992087f355a8bf644e6271e","output_samples":87840,"length_delta_samples":120,"analysis_s":1.308458266001253,"synthesis_s":10.263976322999952,"total_s":11.572434589001205,"rtf":3.166192774008538,"output_peak":0.5084723830223083,"output_rms":0.06389355247368116,"finished_at":"2026-10-13T07:42:57.665345+00:00","analysis_stages":{"external_f0_s":0.8697422179975547},"evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"bea9a11d13c83ac4354113329265ca551002bdf96992087f355a8bf644e6271e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":58480,"output_samples":58560,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.888244020533588,"output_peak":0.5089859371504354},"common_prefix":{"samples":58480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986338797814208,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.39049433724326,"spectral_convergence":0.5459495787379425,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.426871082582255,"spectral_convergence":0.33396090984205035,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.31405918243102,"spectral_convergence":0.3207745758138509,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.00024513927024669564,"delta_db":-0.8121513082091419,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.0014665278538487357,"delta_db":0.0011671420832484137,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.001975127664204572,"delta_db":-0.2828734598811419,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.00017542529154337873,"delta_db":-0.08178665693856965,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00013853554373745173,"delta_db":-0.2763348074493343,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":8.967186743690485e-05,"delta_db":-1.1538895318912639,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.43470048440878334,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23020702296993908,"pitch":null,"output_rms_dbfs":-23.882313014576425,"output_peak":0.5089859371504354,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.710474867418846},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58480,"output_samples":58480,"intelligibility":{"estoi":{"value":0.8890907776027919,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06567418627216229,"rms_dbfs":-23.65210599160649,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.882286904150295,"mean":5.537834115295828e-06,"zero_energy":false},"output":{"rms":0.06395644995633934,"rms_dbfs":-23.882313014576425,"peak":0.5089859371504354,"peak_dbfs":-5.8658843337102615,"crest_factor_db":18.016428680866166,"mean":9.964852500448257e-05,"zero_energy":false},"rms_difference_db":-0.23020702296993534},"full_scale":{"reference":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":7.777437061435905e-06,"output_rms_dbfs":-51.0916349471385,"output_zero_energy":false,"output_peak":0.14253641800942402,"output_energy_fraction":0.00025490391485146965,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":134,"both_estimated_voiced_frames":120,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9523809523809523,"absolute_difference_cents_median":5.150382390407863,"absolute_difference_cents_p95":20.134915746663413,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.0002136685864137289,"log_rms_mae_db":2.2010659086750826,"rms_envelope_nrmse":0.1905236831677188,"envelope_correlation":0.9640096920008848,"rms_gain_error_db":-0.23021709725047468,"reference_mean_power":0.004318997149502425,"output_mean_power":0.004096011880968351},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.9590597233654055e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007496792113024279,"output_power":0.0007155932773998496,"reference_active":true,"delta_db":-0.20209216013395243},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031885669741703006,"output_power":0.0003053762548843616,"reference_active":true,"delta_db":-0.1876027891206863},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.908796664605564e-05,"output_power":6.653078517622708e-05,"reference_active":true,"delta_db":-0.16379762099096137}],"active_band_count":3,"active_band_mae_db":0.1844975234152}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00017810705079567896,"log_rms_mae_db":1.5010582667485384,"rms_envelope_nrmse":0.10744287892542938,"envelope_correlation":0.9878103875068032,"rms_gain_error_db":-0.23021355067786914,"reference_mean_power":0.0043546897567722925,"output_mean_power":0.004129865089126249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.684242583835521e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006964729892166044,"output_power":0.0006706233983673714,"reference_active":true,"delta_db":-0.16425576076351894},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001781987747792959,"output_power":0.00016515273262254329,"reference_active":true,"delta_db":-0.3301894974512836},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.211993913505217e-06,"output_power":5.425353483796582e-06,"reference_active":true,"delta_db":0.17424139612631195}],"active_band_count":3,"active_band_mae_db":0.22289555144703818}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.00018825872292702335,"log_rms_mae_db":2.6387566230102326,"rms_envelope_nrmse":0.21079198838636143,"envelope_correlation":0.9637112994510605,"rms_gain_error_db":-0.19447910121661383,"reference_mean_power":0.003284629583795886,"output_mean_power":0.0031407870386459475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":2.043578979045215e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008901806972697356,"output_power":0.0008541082095505864,"reference_active":true,"delta_db":-0.1796527652636579},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027181233736104136,"output_power":0.00024589682052601785,"reference_active":true,"delta_db":-0.4351625185155745},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0768894288562865e-05,"output_power":5.5310943453181475e-05,"reference_active":true,"delta_db":-0.40870267871590976}],"active_band_count":3,"active_band_mae_db":0.34117265416504744}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00015953911997549416,"log_rms_mae_db":1.942896099904469,"rms_envelope_nrmse":0.1222741291157176,"envelope_correlation":0.9868313925626349,"rms_gain_error_db":-0.19447930148395082,"reference_mean_power":0.0033118454961761057,"output_mean_power":0.0031668109488029606},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.8056141501282585e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008648498098004976,"output_power":0.0008228271153673831,"reference_active":true,"delta_db":-0.21632099283431797},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013850439438479803,"output_power":0.00012478898885063432,"reference_active":true,"delta_db":-0.4528728695940046},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4525026177578845e-06,"output_power":6.089816357818038e-06,"reference_active":true,"delta_db":-0.25123992996304223}],"active_band_count":3,"active_band_mae_db":0.30681126413045495}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":2.2761048422044827,"rms_envelope_nrmse":0.231984479388824,"envelope_correlation":0.9575485662012008,"rms_gain_error_db":-0.11760705696777826,"reference_mean_power":0.00023215995390589003,"output_mean_power":0.00022595741785995494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.4445004803909088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.349295087979124e-05,"output_power":7.189701792920374e-05,"reference_active":true,"delta_db":-0.09534807932384345},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7764968189144563e-05,"output_power":2.477904431589364e-05,"reference_active":true,"delta_db":-0.4941262783975273},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.892866752425831e-06,"output_power":5.486923798117241e-06,"reference_active":true,"delta_db":-0.3099769228454453}],"active_band_count":3,"active_band_mae_db":0.29981709352227204}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":1.359789972032066,"rms_envelope_nrmse":0.12090185141646358,"envelope_correlation":0.9874314205512399,"rms_gain_error_db":-0.11760013033129083,"reference_mean_power":0.0002340443801325015,"output_mean_power":0.00022779186184095508},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.2566052761381825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.967443215044191e-05,"output_power":6.762490009386304e-05,"reference_active":true,"delta_db":-0.12966801374441794},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5348913604759922e-05,"output_power":1.3879769884381156e-05,"reference_active":true,"delta_db":-0.4369537564292926},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.927931708651049e-07,"output_power":4.2258744749378726e-07,"reference_active":true,"delta_db":-0.6674808808206402}],"active_band_count":3,"active_band_mae_db":0.4113675503314502}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":2.3522580222629914,"rms_envelope_nrmse":0.24148143308311035,"envelope_correlation":0.9541955454903595,"rms_gain_error_db":-0.5387492206357387,"reference_mean_power":0.00012513898539343645,"output_mean_power":0.00011053955385379187},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":8.947237162820929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5394872382848794e-05,"output_power":3.808812678971588e-05,"reference_active":true,"delta_db":-0.7621718534722695},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2730550619395117e-05,"output_power":1.1321802492569121e-05,"reference_active":true,"delta_db":-0.5093161368197339},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6846121221132666e-06,"output_power":1.5842685103362712e-06,"reference_active":true,"delta_db":-0.26671131404994886}],"active_band_count":3,"active_band_mae_db":0.5127331014473174}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":1.4673580680886713,"rms_envelope_nrmse":0.15326692695723398,"envelope_correlation":0.981737572113605,"rms_gain_error_db":-0.5387544046468649,"reference_mean_power":0.00012613106271167372,"output_mean_power":0.00011141575675094597},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":8.199333001241022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.113815724379872e-05,"output_power":3.4746048063978154e-05,"reference_active":true,"delta_db":-0.7333941778984538},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.547615689423206e-06,"output_power":5.768269183740329e-06,"reference_active":true,"delta_db":-0.5503766158998037},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0611878817701006e-07,"output_power":9.096951556481279e-08,"reference_active":true,"delta_db":-0.6689639992354955}],"active_band_count":3,"active_band_mae_db":0.6509115976779176}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0014","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.5484368801116943,"predicted_mos":2.5484368801116943,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87960,"input_sample_rate":24000,"input_seconds":3.665,"input_float32_sha256":"e5a0da5e622bf2ab70e64b1aee6024cf8e677689a68026c49b3666bf38bf9dfb","internal_samples":58640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.348097014000814},"started_at":"2026-10-13T05:11:12.220754+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0017819990007410524,"native_input_samples":87720,"native_input_sha256":"4245c8b92bb195f6cc812c529da80e476ccfa2ba1b40c071343dcd7c3f4758e0","output_sha256":"1c3536c082de2b7c12cbe9ad7da253d63ad10371510375583514fd9d5f304d97","output_samples":87960,"length_delta_samples":240,"analysis_s":0.4051608320005471,"synthesis_s":0.08802181799910613,"total_s":0.4931826499996532,"rtf":0.13493369357035656,"output_peak":0.49935823678970337,"output_rms":0.06386770205628678,"finished_at":"2026-10-13T05:11:12.721967+00:00","analysis_stages":{"external_f0_s":0.07509767200099304},"evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"1c3536c082de2b7c12cbe9ad7da253d63ad10371510375583514fd9d5f304d97","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":58480,"output_samples":58640,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.89172420170551,"output_peak":0.49980337875180364},"common_prefix":{"samples":58480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9972714870395635,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.019442027508389,"spectral_convergence":0.2835421366810688,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.597977159245212,"spectral_convergence":0.29416816474958213,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.223295840928682,"spectral_convergence":0.33958558169978625,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.00029797916119372,"delta_db":0.035578735511279456,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.0013431989868911316,"delta_db":-0.38033246203738885,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.002045967391935502,"delta_db":-0.12983810118089645,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.00017014719897753247,"delta_db":-0.21446068675816135,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00013705718253202218,"delta_db":-0.3229289451825147,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":9.8389077606934e-05,"delta_db":-0.7509827822419166,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30568695215202624,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.227753510020131,"pitch":null,"output_rms_dbfs":-23.87985950162662,"output_peak":0.49980337875180364,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.946905009227428},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58480,"output_samples":58480,"intelligibility":{"estoi":{"value":0.9216871079490816,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06567418627216229,"rms_dbfs":-23.65210599160649,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.882286904150295,"mean":5.537834115295828e-06,"zero_energy":false},"output":{"rms":0.06397451835799034,"rms_dbfs":-23.87985950162662,"peak":0.49980337875180364,"peak_dbfs":-6.024016245971921,"crest_factor_db":17.8558432556547,"mean":0.00013237608077149238,"zero_energy":false},"rms_difference_db":-0.2277535100201291},"full_scale":{"reference":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":1.641410314602983e-06,"output_rms_dbfs":-57.84782841802815,"output_zero_energy":false,"output_peak":0.017971409544319486,"output_energy_fraction":5.376650503371148e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":114,"both_estimated_voiced_frames":110,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.873015873015873,"absolute_difference_cents_median":0.630928877506145,"absolute_difference_cents_p95":4.446667569861643,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.0002136685864137289,"log_rms_mae_db":0.7884325282493766,"rms_envelope_nrmse":0.09230635917075163,"envelope_correlation":0.9919351508055706,"rms_gain_error_db":-0.22774979291096747,"reference_mean_power":0.004318997149502425,"output_mean_power":0.004098339559440799},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.9590597233654055e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007496792113024279,"output_power":0.0007279746715841686,"reference_active":true,"delta_db":-0.12759198801086494},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031885669741703006,"output_power":0.00028113996436938607,"reference_active":true,"delta_db":-0.5467295808998813},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.908796664605564e-05,"output_power":6.017476756238275e-05,"reference_active":true,"delta_db":-0.5998798959056462}],"active_band_count":3,"active_band_mae_db":0.42473382160546413}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00017810705079567896,"log_rms_mae_db":0.525251210183631,"rms_envelope_nrmse":0.05899015061644742,"envelope_correlation":0.9968177743222231,"rms_gain_error_db":-0.22772840976197614,"reference_mean_power":0.0043546897567722925,"output_mean_power":0.0041322289768003385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.684242583835521e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006964729892166044,"output_power":0.0006721344721249459,"reference_active":true,"delta_db":-0.15448108304814395},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001781987747792959,"output_power":0.00015821059721937518,"reference_active":true,"delta_db":-0.5166914375147383},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.211993913505217e-06,"output_power":4.962086268659419e-06,"reference_active":true,"delta_db":-0.2133958942217281}],"active_band_count":3,"active_band_mae_db":0.2948561382615368}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.00018825872292702335,"log_rms_mae_db":1.223745614864962,"rms_envelope_nrmse":0.1113075754266374,"envelope_correlation":0.9902843828933672,"rms_gain_error_db":-0.23151890708728456,"reference_mean_power":0.003284629583795886,"output_mean_power":0.003114114017668709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":2.043578979045215e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008901806972697356,"output_power":0.000854677241924216,"reference_active":true,"delta_db":-0.17676032924937451},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027181233736104136,"output_power":0.00022378138996922682,"reference_active":true,"delta_db":-0.8444519810322438},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0768894288562865e-05,"output_power":5.39744814646467e-05,"reference_active":true,"delta_db":-0.5149285553910308}],"active_band_count":3,"active_band_mae_db":0.5120469552242164}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00015953911997549416,"log_rms_mae_db":0.7638086097181164,"rms_envelope_nrmse":0.06863133344804753,"envelope_correlation":0.9963384515560532,"rms_gain_error_db":-0.23151801792411333,"reference_mean_power":0.0033118454961761057,"output_mean_power":0.0031399177083548642},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.8056141501282585e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008648498098004976,"output_power":0.0008191974914920354,"reference_active":true,"delta_db":-0.23552080483763654},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013850439438479803,"output_power":0.00011526130620852191,"reference_active":true,"delta_db":-0.7978001570304378},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4525026177578845e-06,"output_power":5.439882456168438e-06,"reference_active":true,"delta_db":-0.7413867375161037}],"active_band_count":3,"active_band_mae_db":0.5915692331280593}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":1.1081212507593323,"rms_envelope_nrmse":0.10769905341360096,"envelope_correlation":0.9914299658751293,"rms_gain_error_db":-0.2793163424009177,"reference_mean_power":0.00023215995390589003,"output_mean_power":0.00021769861929145438},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.4445004803909088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.349295087979124e-05,"output_power":7.097860310861616e-05,"reference_active":true,"delta_db":-0.1511823748599666},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7764968189144563e-05,"output_power":2.593761261364174e-05,"reference_active":true,"delta_db":-0.2956718058758261},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.892866752425831e-06,"output_power":4.693794130759396e-06,"reference_active":true,"delta_db":-0.9880258350445381}],"active_band_count":3,"active_band_mae_db":0.4782933385934436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":0.635007101558568,"rms_envelope_nrmse":0.06184597657659729,"envelope_correlation":0.9974943381607533,"rms_gain_error_db":-0.2791722827018451,"reference_mean_power":0.0002340443801325015,"output_mean_power":0.00021947294387911055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.2566052761381825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.967443215044191e-05,"output_power":6.686778744452015e-05,"reference_active":true,"delta_db":-0.17856484601276568},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5348913604759922e-05,"output_power":1.4379190168978035e-05,"reference_active":true,"delta_db":-0.28343214141614104},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.927931708651049e-07,"output_power":3.580141987007514e-07,"reference_active":true,"delta_db":-1.3876442980697372}],"active_band_count":3,"active_band_mae_db":0.6165470951662146}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":1.0989205663773949,"rms_envelope_nrmse":0.10509865112123705,"envelope_correlation":0.9917129792777564,"rms_gain_error_db":-0.24869026573259623,"reference_mean_power":0.00012513898539343645,"output_mean_power":0.00011817445358622541},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":8.947237162820929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5394872382848794e-05,"output_power":4.3678886915819756e-05,"reference_active":true,"delta_db":-0.16735236956676192},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2730550619395117e-05,"output_power":1.1697610199944723e-05,"reference_active":true,"delta_db":-0.36750042832141455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6846121221132666e-06,"output_power":1.395008385823325e-06,"reference_active":true,"delta_db":-0.8192310316337647}],"active_band_count":3,"active_band_mae_db":0.45136127650731367}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":0.6921216328004004,"rms_envelope_nrmse":0.06448151985150163,"envelope_correlation":0.9970297134130189,"rms_gain_error_db":-0.2483227767263715,"reference_mean_power":0.00012613106271167372,"output_mean_power":0.00011912139678791284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":8.199333001241022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.113815724379872e-05,"output_power":3.972091793322547e-05,"reference_active":true,"delta_db":-0.1522555756709595},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.547615689423206e-06,"output_power":6.0320390709787886e-06,"reference_active":true,"delta_db":-0.3561903478640823},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0611878817701006e-07,"output_power":8.5794176347991e-08,"reference_active":true,"delta_db":-0.9233447268818435}],"active_band_count":3,"active_band_mae_db":0.47726355013896177}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0014","method":"magphase","status":"measured","mos":{"status":"measured","value":3.3916661739349365,"predicted_mos":3.3916661739349365,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":175845,"input_sample_rate":48000,"input_seconds":3.6634375,"input_float32_sha256":"3811baf8d5a86b23e0a393476da10ce112e09185a0f1fb4df6fa20655a24ec2e","internal_samples":58615,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2620655780010566},"started_at":"2026-10-13T05:17:19.797373+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005984809999063145,"native_input_samples":175440,"native_input_sha256":"f80bf4b4663b35cf14e34a9e918c2563874251a8a30be691bf0714f8c67d921c","output_sha256":"dd4739a60d3aa5c2b719165a6808326657ce5ca74f41941330100c3354feefaa","output_samples":175845,"length_delta_samples":405,"analysis_s":2.147643951999271,"synthesis_s":0.9841494020001846,"total_s":3.1317933539994556,"rtf":0.8568518068397963,"output_peak":0.4945267954111787,"output_rms":0.06354599058113848,"finished_at":"2026-10-13T05:17:22.944604+00:00","evaluation_input":{"reference":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"dd4739a60d3aa5c2b719165a6808326657ce5ca74f41941330100c3354feefaa","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":58480,"output_samples":58615,"length_delta_samples":135,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.936504803318247,"output_peak":0.48023851442990184},"common_prefix":{"samples":58480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9976968352810714,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":58480,"output_samples":58480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.938313151932805,"spectral_convergence":0.5527372522535228,"spectral_convergence_status":"ok","amplitude_floor":1.5267125480811954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.786335314621889,"spectral_convergence":0.26587727761568514,"spectral_convergence_status":"ok","amplitude_floor":1.0684354905475814e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.74847848292001,"spectral_convergence":0.26027078469660164,"spectral_convergence_status":"ok","amplitude_floor":7.412945744843815e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0002955479965644196,"output_power":0.0002510703382421946,"delta_db":-0.7083261207590619,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00146613378566028,"output_power":0.0013842281877081335,"delta_db":-0.2496591302352286,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0021080578166581473,"output_power":0.0019473775917571935,"delta_db":-0.34432349559630304,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001787602167832836,"output_power":0.00018678140901969248,"delta_db":0.19062774598879373,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00014763682412872298,"output_power":0.00015914446596967298,"delta_db":0.325968468731094,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001169621027158164,"output_power":0.00012042035433098614,"delta_db":0.1265473344024076,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.32424204928548145,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2743922714228607,"pitch":null,"output_rms_dbfs":-23.92649826302935,"output_peak":0.48023851442990184,"reference_rms_dbfs":-23.65210599160649,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.157708983158234},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":58480,"output_samples":58480,"intelligibility":{"estoi":{"value":0.8976439693618944,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06567418627216229,"rms_dbfs":-23.65210599160649,"peak":0.45867919921875,"peak_dbfs":-6.769819087456196,"crest_factor_db":16.882286904150295,"mean":5.537834115295828e-06,"zero_energy":false},"output":{"rms":0.06363192867600254,"rms_dbfs":-23.92649826302935,"peak":0.48023851442990184,"peak_dbfs":-6.3708602619381205,"crest_factor_db":17.55563800109123,"mean":2.1424570771783932e-05,"zero_energy":false},"rms_difference_db":-0.2743922714228617},"full_scale":{"reference":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":58480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":365,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.906763410195709e-06,"threshold_dbfs":-54.08182888431633,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":49,"quiet_samples":7840,"reference_mean_power":1.377575645907497e-06,"output_mean_power":1.5722940231620315e-06,"output_rms_dbfs":-58.03466236479406,"output_zero_energy":false,"output_peak":0.012091795289669409,"output_energy_fraction":5.2058576506796066e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":358,"reference_estimated_voiced_frames":126,"output_estimated_voiced_frames":135,"both_estimated_voiced_frames":123,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9761904761904762,"absolute_difference_cents_median":4.335334096944815,"absolute_difference_cents_p95":10.000978884017252,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":58480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.0002136685864137289,"log_rms_mae_db":1.5823152344595488,"rms_envelope_nrmse":0.15980101869818722,"envelope_correlation":0.9749116883608725,"rms_gain_error_db":-0.27439031116450063,"reference_mean_power":0.004318997149502425,"output_mean_power":0.004054561444672715},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.9590597233654055e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007496792113024279,"output_power":0.0007004229476368485,"reference_active":true,"delta_db":-0.2951510173712619},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00031885669741703006,"output_power":0.00029457767567792807,"reference_active":true,"delta_db":-0.34395712435351045},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.908796664605564e-05,"output_power":6.29394427939078e-05,"reference_active":true,"delta_db":-0.40479517224604744}],"active_band_count":3,"active_band_mae_db":0.3479677713236066}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00017810705079567896,"log_rms_mae_db":1.0253367017567994,"rms_envelope_nrmse":0.09048874736127217,"envelope_correlation":0.9918122028243902,"rms_gain_error_db":-0.27441573397261054,"reference_mean_power":0.0043546897567722925,"output_mean_power":0.00408804479913978},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.684242583835521e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006964729892166044,"output_power":0.0006482132171870862,"reference_active":true,"delta_db":-0.3118639604218448},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001781987747792959,"output_power":0.0001664039057031739,"reference_active":true,"delta_db":-0.2974119818583286},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.211993913505217e-06,"output_power":4.798668750694898e-06,"reference_active":true,"delta_db":-0.3588312800541838}],"active_band_count":3,"active_band_mae_db":0.32270240744478573}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":0.00018825872292702335,"log_rms_mae_db":2.1804904437859203,"rms_envelope_nrmse":0.15817259979624565,"envelope_correlation":0.980305436481114,"rms_gain_error_db":-0.3283214562759177,"reference_mean_power":0.003284629583795886,"output_mean_power":0.0030454695004195325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":2.043578979045215e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008901806972697356,"output_power":0.0008327618065929869,"reference_active":true,"delta_db":-0.28957374106141376},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00027181233736104136,"output_power":0.0002419577055556776,"reference_active":true,"delta_db":-0.5052970763617074},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.0768894288562865e-05,"output_power":5.064525066183426e-05,"reference_active":true,"delta_db":-0.791426093215618}],"active_band_count":3,"active_band_mae_db":0.5287656368795797}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":0.00015953911997549416,"log_rms_mae_db":1.6114842328929988,"rms_envelope_nrmse":0.09825100913075634,"envelope_correlation":0.9924555294120012,"rms_gain_error_db":-0.3283211908743778,"reference_mean_power":0.0033118454961761057,"output_mean_power":0.003070703958472808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.8056141501282585e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008648498098004976,"output_power":0.000802816061786958,"reference_active":true,"delta_db":-0.3232464149268979},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013850439438479803,"output_power":0.00012414709503141204,"reference_active":true,"delta_db":-0.47526990879800307},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.4525026177578845e-06,"output_power":5.289785147135888e-06,"reference_active":true,"delta_db":-0.8629015656122012}],"active_band_count":3,"active_band_mae_db":0.5538059631123674}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":6.60450817247952e-05,"log_rms_mae_db":1.8490676112677507,"rms_envelope_nrmse":0.1810863781762078,"envelope_correlation":0.9757050928196118,"rms_gain_error_db":0.2370475595785947,"reference_mean_power":0.00023215995390589003,"output_mean_power":0.00024518396548170484},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":1.4445004803909088e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.349295087979124e-05,"output_power":7.745306767230693e-05,"reference_active":true,"delta_db":0.22792938006238342},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7764968189144563e-05,"output_power":2.892121085719897e-05,"reference_active":true,"delta_db":0.17719291574012738},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.892866752425831e-06,"output_power":5.791536395318898e-06,"reference_active":true,"delta_db":-0.07532831188871059}],"active_band_count":3,"active_band_mae_db":0.16015020256374046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.621650617144381e-05,"log_rms_mae_db":1.1139304031462298,"rms_envelope_nrmse":0.1093866107079477,"envelope_correlation":0.9906484010804802,"rms_gain_error_db":0.236906479160104,"reference_mean_power":0.0002340443801325015,"output_mean_power":0.0002471660773937523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":1.2566052761381825e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.967443215044191e-05,"output_power":7.324331397457937e-05,"reference_active":true,"delta_db":0.21694548325945706},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5348913604759922e-05,"output_power":1.5639741835500493e-05,"reference_active":true,"delta_db":0.08151938341888418},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.927931708651049e-07,"output_power":4.547515066167533e-07,"reference_active":true,"delta_db":-0.34890534159443964}],"active_band_count":3,"active_band_mae_db":0.2157900694242603}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":730,"rms_floor":5.9166379423348476e-05,"log_rms_mae_db":2.041215844514149,"rms_envelope_nrmse":0.19422928114507848,"envelope_correlation":0.9722020077398884,"rms_gain_error_db":0.2981941281542822,"reference_mean_power":0.00012513898539343645,"output_mean_power":0.0001340330938107624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":730,"frequency_resolution_hz":0.273972602739726,"power_floor":8.947237162820929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5394872382848794e-05,"output_power":4.674988330146744e-05,"reference_active":true,"delta_db":0.12773731592411308},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2730550619395117e-05,"output_power":1.2835670798656754e-05,"reference_active":true,"delta_db":0.035713819665074545},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6846121221132666e-06,"output_power":3.6400906636044243e-06,"reference_active":true,"delta_db":3.346122792822822}],"active_band_count":3,"active_band_mae_db":1.1698579761373364}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":58480,"uncovered_tail_samples":0,"metrics":{"frame_count":724,"rms_floor":5.172390189127307e-05,"log_rms_mae_db":1.2426684269598087,"rms_envelope_nrmse":0.1094301598349345,"envelope_correlation":0.991275150813426,"rms_gain_error_db":0.29800279593807233,"reference_mean_power":0.00012613106271167372,"output_mean_power":0.00013508973024828867},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":724,"frequency_resolution_hz":0.27624309392265195,"power_floor":8.199333001241022e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.113815724379872e-05,"output_power":4.2662801717170156e-05,"reference_active":true,"delta_db":0.1580453882358605},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.547615689423206e-06,"output_power":7.07497689932927e-06,"reference_active":true,"delta_db":0.3364184548584722},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0611878817701006e-07,"output_power":2.2806506460915276e-07,"reference_active":true,"delta_db":3.3226648249609196}],"active_band_count":3,"active_band_mae_db":1.272376222685084}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"hnm","status":"measured","mos":{"status":"measured","value":2.636467218399048,"predicted_mos":2.636467218399048,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":40080,"input_sample_rate":16000,"input_seconds":2.505,"input_float32_sha256":"dd1e148d8a43f0c0f64c63fbc3f617dad21c6c4915a01cf7c3b072f0cdd8b75d","internal_samples":40080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2809592239991616},"started_at":"2026-10-13T04:39:54.193075+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2362999768811278e-05,"native_input_samples":40080,"native_input_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","output_sha256":"b1a448dde87b78ad54b6613b74be812d4295f6c13a15e8aea417ac10e5395300","output_samples":40080,"length_delta_samples":0,"analysis_s":0.3488930519997666,"synthesis_s":0.21066987999984121,"total_s":0.5595629319996078,"rtf":0.22337841596790733,"output_peak":0.3920935910531185,"output_rms":0.061182125709596684,"finished_at":"2026-10-13T04:39:54.756108+00:00","evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"b1a448dde87b78ad54b6613b74be812d4295f6c13a15e8aea417ac10e5395300","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.9347295964530735,"spectral_convergence":0.30415234637465854,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.685881277936386,"spectral_convergence":0.29866676739314607,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.929182927421702,"spectral_convergence":0.30561145407565643,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.00026450861626878096,"delta_db":1.0944274935447695,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.0025062027370098394,"delta_db":-0.2523365478982331,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.000610006711628791,"delta_db":-0.3348310163239773,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":6.538833742794834e-05,"delta_db":0.16432161368129472,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.0001788598924940608,"delta_db":-0.07364751626102044,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":0.00011828621151547096,"delta_db":-0.24851601324580647,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.36134670015918363,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16784183950268122,"pitch":null,"output_rms_dbfs":-24.267508759543883,"output_peak":0.3920935910531185,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.849931267270388},"common_prefix":{"samples":40080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.9347295964530735,"spectral_convergence":0.30415234637465854,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.685881277936386,"spectral_convergence":0.29866676739314607,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.929182927421702,"spectral_convergence":0.30561145407565643,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.00026450861626878096,"delta_db":1.0944274935447695,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.0025062027370098394,"delta_db":-0.2523365478982331,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.000610006711628791,"delta_db":-0.3348310163239773,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":6.538833742794834e-05,"delta_db":0.16432161368129472,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.0001788598924940608,"delta_db":-0.07364751626102044,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":0.00011828621151547096,"delta_db":-0.24851601324580647,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.36134670015918363,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16784183950268122,"pitch":null,"output_rms_dbfs":-24.267508759543883,"output_peak":0.3920935910531185,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.849931267270388},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40080,"output_samples":40080,"intelligibility":{"estoi":{"value":0.8752288624879712,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06237587544550748,"rms_dbfs":-24.099666920041205,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.680906994925625,"mean":-1.5409549553237275e-05,"zero_energy":false},"output":{"rms":0.061182125709596684,"rms_dbfs":-24.267508759543883,"peak":0.3920935910531185,"peak_dbfs":-8.132205127625708,"crest_factor_db":16.135303631918177,"mean":0.00029916732196474325,"zero_energy":false},"rms_difference_db":-0.16784183950267817},"full_scale":{"reference":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":6.172951329620234e-07,"output_rms_dbfs":-62.095071471962946,"output_zero_energy":false,"output_peak":0.005387190545448815,"output_energy_fraction":3.028264673203725e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":60,"both_estimated_voiced_frames":57,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9193548387096774,"absolute_difference_cents_median":1.6369607523980816,"absolute_difference_cents_p95":12.839851822822167,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.0001593748762509214,"log_rms_mae_db":0.9139342904261315,"rms_envelope_nrmse":0.08755057932790529,"envelope_correlation":0.9931073540048595,"rms_gain_error_db":-0.167831021552308,"reference_mean_power":0.003898496329889168,"output_mean_power":0.003750714673836689},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.3730898670333144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011160072986886684,"output_power":0.0010949494964658515,"reference_active":true,"delta_db":-0.08272946690866083},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003924601048841299,"output_power":0.0003689323259473657,"reference_active":true,"delta_db":-0.26848805689245553},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.521600297301381e-05,"output_power":8.997473307142136e-05,"reference_active":true,"delta_db":0.23599406992320904}],"active_band_count":3,"active_band_mae_db":0.19573719790810848}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00014354118626608257,"log_rms_mae_db":0.4700764183621952,"rms_envelope_nrmse":0.047383690585783865,"envelope_correlation":0.9980595536733371,"rms_gain_error_db":-0.16781842830287375,"reference_mean_power":0.00394559944963645,"output_mean_power":0.0037960432466253554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":2.0339309551628037e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010452646972319804,"output_power":0.0010343032226456186,"reference_active":true,"delta_db":-0.04578404936866319},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020675655639960477,"output_power":0.00019591414063318258,"reference_active":true,"delta_db":-0.23393506729023833},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.033282667633225e-06,"output_power":6.154268480877522e-06,"reference_active":true,"delta_db":0.08622765396159163}],"active_band_count":3,"active_band_mae_db":0.12198225687349773}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.00014555177137634006,"log_rms_mae_db":1.555630530287412,"rms_envelope_nrmse":0.10913948466427097,"envelope_correlation":0.9915791721408828,"rms_gain_error_db":-0.3321868938653654,"reference_mean_power":0.0025948122953778335,"output_mean_power":0.002403738711782998},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.091667884405576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010721637846984168,"output_power":0.0010256055647286581,"reference_active":true,"delta_db":-0.19280765062686428},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045198280410970506,"output_power":0.00037915524909631356,"reference_active":true,"delta_db":-0.7630483937722744},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681659745338991e-05,"output_power":5.3621492212395734e-05,"reference_active":true,"delta_db":-0.2513632616011785}],"active_band_count":3,"active_band_mae_db":0.40240643533343906}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00013266563413733552,"log_rms_mae_db":0.8911321192604661,"rms_envelope_nrmse":0.06802945771770488,"envelope_correlation":0.9971747658784741,"rms_gain_error_db":-0.33218600344936133,"reference_mean_power":0.002626297782108314,"output_mean_power":0.0024329062079302133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":1.815151890487515e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010442677388382842,"output_power":0.0009976756433789722,"reference_active":true,"delta_db":-0.19822491545445714},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292871431701824,"output_power":0.0001962170087679869,"reference_active":true,"delta_db":-0.6764305237448753},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.238038641621491e-06,"output_power":5.5452127311954406e-06,"reference_active":true,"delta_db":-1.1570268568099948}],"active_band_count":3,"active_band_mae_db":0.6772274320031091}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":1.5223030011192213,"rms_envelope_nrmse":0.20171144304993013,"envelope_correlation":0.9628350521038813,"rms_gain_error_db":-0.049928087849628794,"reference_mean_power":0.00015194739829742755,"output_mean_power":0.0001502105580355829},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":8.081548208249737e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5556653827814036e-05,"output_power":4.6074141871035454e-05,"reference_active":true,"delta_db":0.04905437943890535},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1818343242486448e-05,"output_power":1.2279732678432738e-05,"reference_active":true,"delta_db":0.16632313480077843},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7000790662746814e-06,"output_power":6.169741332312591e-06,"reference_active":true,"delta_db":2.22055952007491}],"active_band_count":3,"active_band_mae_db":0.811979011438198}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":0.7838416498782951,"rms_envelope_nrmse":0.10033424566016301,"envelope_correlation":0.9897280236464436,"rms_gain_error_db":-0.04987387808529537,"reference_mean_power":0.00015370989511122725,"output_mean_power":0.0001519548052929368},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":6.838335178235257e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.320123857018819e-05,"output_power":4.510136878827775e-05,"reference_active":true,"delta_db":0.1869352442454483},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064245816738489e-06,"output_power":6.552729393538822e-06,"reference_active":true,"delta_db":0.3364543608445848},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0299327724151418e-07,"output_power":3.1845118285360433e-07,"reference_active":true,"delta_db":1.9556121123741401}],"active_band_count":3,"active_band_mae_db":0.8263339058213911}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":1.7294420966363728,"rms_envelope_nrmse":0.19447046118045758,"envelope_correlation":0.9717356313012232,"rms_gain_error_db":-0.13420265356318206,"reference_mean_power":0.00013710154501992005,"output_mean_power":0.00013292971805094157},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":1.0763544842397469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510390122735278e-05,"output_power":4.3778873870872305e-05,"reference_active":true,"delta_db":-0.12949521503305522},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.111326790734689e-05,"output_power":2.356000168229056e-05,"reference_active":true,"delta_db":-1.2077031145629336},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.435692144349248e-06,"output_power":5.042735883359915e-06,"reference_active":true,"delta_db":-0.32588629778475386}],"active_band_count":3,"active_band_mae_db":0.5543615424602476}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":1.0968901142184742,"rms_envelope_nrmse":0.13038033758063952,"envelope_correlation":0.9864358246240718,"rms_gain_error_db":-0.13375399545234432,"reference_mean_power":0.00013864857515666153,"output_mean_power":0.0001344435623036979},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":9.213681129362451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3816318671200005e-05,"output_power":4.275721704170236e-05,"reference_active":true,"delta_db":-0.10626456172038629},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.911163628753254e-05,"output_power":1.411052583461878e-05,"reference_active":true,"delta_db":-1.3175467356623032},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5242776229186793e-07,"output_power":3.9815100437966896e-07,"reference_active":true,"delta_db":0.5297770323659009}],"active_band_count":3,"active_band_mae_db":0.6511961099161968}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.2752292156219482,"predicted_mos":2.2752292156219482,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":40080,"input_sample_rate":16000,"input_seconds":2.505,"input_float32_sha256":"013b5cc85ca65908ceb36244ccb898165a4259c074d8751e60be8a2763ed19e2","internal_samples":40080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1163254210005107},"started_at":"2026-10-13T04:43:05.604039+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.7762999959813897e-05,"native_input_samples":40080,"native_input_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","output_sha256":"d6811601c609e96757b986525030431d8af0c198d985f0cc6525d215a51f31ba","output_samples":40080,"length_delta_samples":0,"analysis_s":0.6056054259997836,"synthesis_s":0.04490397700010362,"total_s":0.6505094029998872,"rtf":0.25968439241512464,"output_peak":0.41688863541978854,"output_rms":0.07004520721796631,"finished_at":"2026-10-13T04:43:06.257754+00:00","evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"d6811601c609e96757b986525030431d8af0c198d985f0cc6525d215a51f31ba","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.362139219541681,"spectral_convergence":0.6319299465823969,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6724881828188,"spectral_convergence":0.33315973520786624,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.970394773597237,"spectral_convergence":0.3796104121332174,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.00019167705098753138,"delta_db":-0.30426955224989694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.0034140571864182817,"delta_db":1.0902093666498895,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.0009048042571219563,"delta_db":1.3773692027053839,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":7.336459642937927e-05,"delta_db":0.6641839916792946,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.0001945104770283216,"delta_db":0.2906528193773492,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":0.00012791748622236965,"delta_db":0.09144192553676893,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6363544763664305,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.007235408408477,"pitch":null,"output_rms_dbfs":-23.09243151163273,"output_peak":0.41688863541978854,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.001674058652573},"common_prefix":{"samples":40080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.362139219541681,"spectral_convergence":0.6319299465823969,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6724881828188,"spectral_convergence":0.33315973520786624,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.970394773597237,"spectral_convergence":0.3796104121332174,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.00019167705098753138,"delta_db":-0.30426955224989694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.0034140571864182817,"delta_db":1.0902093666498895,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.0009048042571219563,"delta_db":1.3773692027053839,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":7.336459642937927e-05,"delta_db":0.6641839916792946,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.0001945104770283216,"delta_db":0.2906528193773492,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":0.00012791748622236965,"delta_db":0.09144192553676893,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6363544763664305,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.007235408408477,"pitch":null,"output_rms_dbfs":-23.09243151163273,"output_peak":0.41688863541978854,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.001674058652573},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40080,"output_samples":40080,"intelligibility":{"estoi":{"value":0.892678131194247,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06237587544550748,"rms_dbfs":-24.099666920041205,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.680906994925625,"mean":-1.5409549553237275e-05,"zero_energy":false},"output":{"rms":0.07004520721796631,"rms_dbfs":-23.092431511632725,"peak":0.41688863541978854,"peak_dbfs":-7.5995988756821315,"crest_factor_db":15.492832635950595,"mean":-1.9432631392189782e-09,"zero_energy":false},"rms_difference_db":1.00723540840848},"full_scale":{"reference":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":7.034384687423926e-07,"output_rms_dbfs":-61.527738852290284,"output_zero_energy":false,"output_peak":0.008951279114370556,"output_energy_fraction":2.6328090820254982e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":53,"both_estimated_voiced_frames":42,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.6774193548387096,"absolute_difference_cents_median":4.322462484009735,"absolute_difference_cents_p95":17.65345234562248,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.0001593748762509214,"log_rms_mae_db":1.590857030153498,"rms_envelope_nrmse":0.1813584794174944,"envelope_correlation":0.9855865998791956,"rms_gain_error_db":1.0072593597563317,"reference_mean_power":0.003898496329889168,"output_mean_power":0.004916126683523401},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.3730898670333144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011160072986886684,"output_power":0.0015723636924327865,"reference_active":true,"delta_db":1.4888597204045846},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003924601048841299,"output_power":0.0005258968385711018,"reference_active":true,"delta_db":1.2710504448007995},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.521600297301381e-05,"output_power":6.758586784219711e-05,"reference_active":true,"delta_db":-1.006652651490715}],"active_band_count":3,"active_band_mae_db":1.2555209388986996}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00014354118626608257,"log_rms_mae_db":1.0649547084741153,"rms_envelope_nrmse":0.14561512396728615,"envelope_correlation":0.994751782014968,"rms_gain_error_db":1.0073381142275792,"reference_mean_power":0.00394559944963645,"output_mean_power":0.004975615427811103},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":2.0339309551628037e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010452646972319804,"output_power":0.0014744018665112098,"reference_active":true,"delta_db":1.4938958922167291},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020675655639960477,"output_power":0.00027463827621126654,"reference_active":true,"delta_db":1.2330177430842133},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.033282667633225e-06,"output_power":5.41879950533504e-06,"reference_active":true,"delta_db":-0.4665059060260512}],"active_band_count":3,"active_band_mae_db":1.064473180442331}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.00014555177137634006,"log_rms_mae_db":2.3080406296280245,"rms_envelope_nrmse":0.23086459180385988,"envelope_correlation":0.9833306562074815,"rms_gain_error_db":1.3451099419483041,"reference_mean_power":0.0025948122953778335,"output_mean_power":0.003536852441798735},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.091667884405576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010721637846984168,"output_power":0.0015986150270760006,"reference_active":true,"delta_db":1.734827574089473},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045198280410970506,"output_power":0.0005810516012738229,"reference_active":true,"delta_db":1.0909279016608737},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681659745338991e-05,"output_power":4.7222978020731434e-05,"reference_active":true,"delta_db":-0.8032185029192547}],"active_band_count":3,"active_band_mae_db":1.2096579928898672}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00013266563413733552,"log_rms_mae_db":1.6345448909844436,"rms_envelope_nrmse":0.20335928497371988,"envelope_correlation":0.9904519671688626,"rms_gain_error_db":1.3451358147772439,"reference_mean_power":0.002626297782108314,"output_mean_power":0.0035797899824493553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":1.815151890487515e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010442677388382842,"output_power":0.0015492242568266308,"reference_active":true,"delta_db":1.7130242698759748},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292871431701824,"output_power":0.00028542946783527797,"reference_active":true,"delta_db":0.9511910460743873},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.238038641621491e-06,"output_power":6.473888642004609e-06,"reference_active":true,"delta_db":-0.4845567255189942}],"active_band_count":3,"active_band_mae_db":1.0495906804897854}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":1.6009320532578308,"rms_envelope_nrmse":0.20052428308302747,"envelope_correlation":0.9685198856152161,"rms_gain_error_db":0.49640461383183265,"reference_mean_power":0.00015194739829742755,"output_mean_power":0.00017034670195421788},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":8.081548208249737e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5556653827814036e-05,"output_power":5.304720528097061e-05,"reference_active":true,"delta_db":0.6611069132676342},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1818343242486448e-05,"output_power":1.43766764764867e-05,"reference_active":true,"delta_db":0.8510190063647547},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7000790662746814e-06,"output_power":3.3905660237754386e-06,"reference_active":true,"delta_db":-0.37938798801017326}],"active_band_count":3,"active_band_mae_db":0.630504635880854}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":0.9880095618144066,"rms_envelope_nrmse":0.10674916448184728,"envelope_correlation":0.9924547081066237,"rms_gain_error_db":0.4973829561023275,"reference_mean_power":0.00015370989511122725,"output_mean_power":0.00017236144319891737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":6.838335178235257e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.320123857018819e-05,"output_power":4.9976301964385256e-05,"reference_active":true,"delta_db":0.6326791889643666},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064245816738489e-06,"output_power":7.646613504563262e-06,"reference_active":true,"delta_db":1.0069234219005982},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0299327724151418e-07,"output_power":1.8206854779552264e-07,"reference_active":true,"delta_db":-0.47246726898924185}],"active_band_count":3,"active_band_mae_db":0.7040232932847356}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":1.687520000075382,"rms_envelope_nrmse":0.19850463364721613,"envelope_correlation":0.9718148081136652,"rms_gain_error_db":0.19087335088409108,"reference_mean_power":0.00013710154501992005,"output_mean_power":0.00014326156266710258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":1.0763544842397469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510390122735278e-05,"output_power":4.738889723269656e-05,"reference_active":true,"delta_db":0.21462495060061246},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.111326790734689e-05,"output_power":3.530317751370391e-05,"reference_active":true,"delta_db":0.5486816786996994},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.435692144349248e-06,"output_power":5.075647937902488e-06,"reference_active":true,"delta_db":-0.297633615661843}],"active_band_count":3,"active_band_mae_db":0.35364674832071835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":1.1811602286697072,"rms_envelope_nrmse":0.11385417380353186,"envelope_correlation":0.9903066289894726,"rms_gain_error_db":0.19084697694237462,"reference_mean_power":0.00013864857515666153,"output_mean_power":0.00014487722156163635},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":9.213681129362451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3816318671200005e-05,"output_power":4.635424225177593e-05,"reference_active":true,"delta_db":0.2445359959385951},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.911163628753254e-05,"output_power":2.1138349547588106e-05,"reference_active":true,"delta_db":0.43773203409926137},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5242776229186793e-07,"output_power":3.2081932671318806e-07,"reference_active":true,"delta_db":-0.4080958932889834}],"active_band_count":3,"active_band_mae_db":0.3634546411089466}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"facodec","status":"measured","mos":{"status":"measured","value":3.666386604309082,"predicted_mos":3.666386604309082,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":40200,"input_sample_rate":16000,"input_seconds":2.5125,"input_float32_sha256":"e874ec7d4c4094e1d66d850e9b4d7e2757373b0fa415ca412e6d481e5f520748","internal_samples":40200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1388390729989624},"started_at":"2026-10-13T06:57:28.192781+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.694400038511958e-05,"native_input_samples":40080,"native_input_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","output_sha256":"bfbb82372cf1907b9a801ed10146959f2cac3f4d2579862f7f5a26a4dabe945c","output_samples":40200,"length_delta_samples":120,"analysis_s":0.6804804250004963,"synthesis_s":5.022657634999632,"total_s":5.7031380600001285,"rtf":2.276701820359333,"output_peak":0.43894627690315247,"output_rms":0.05918777294446285,"finished_at":"2026-10-13T06:57:33.902249+00:00","evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"bfbb82372cf1907b9a801ed10146959f2cac3f4d2579862f7f5a26a4dabe945c","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":40080,"output_samples":40200,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.555360017973573,"output_peak":0.43894627690315247},"common_prefix":{"samples":40080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9970149253731343,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.598904114503573,"spectral_convergence":0.2530930922770819,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.94095808360349,"spectral_convergence":0.2502046945181593,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.110445458119355,"spectral_convergence":0.2809544605167591,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.00018447359840928729,"delta_db":-0.47062854873217425,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.0025130151487398255,"delta_db":-0.2405474815262336,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.0005109180229193352,"delta_db":-1.1046649208230683,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":5.246331005284588e-05,"delta_db":-0.7921244577095754,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.00014318226339244137,"delta_db":-1.0398849364058687,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":0.0001096286781929238,"delta_db":-0.5786154591038383,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7044109673834598,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.44270975082613595,"pitch":null,"output_rms_dbfs":-24.542376670867338,"output_peak":0.43894627690315247,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.883435885408805},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40080,"output_samples":40080,"intelligibility":{"estoi":{"value":0.9012478814110901,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06237587544550748,"rms_dbfs":-24.099666920041205,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.680906994925625,"mean":-1.5409549553237275e-05,"zero_energy":false},"output":{"rms":0.05927631079703475,"rms_dbfs":-24.542376670867338,"peak":0.43894627690315247,"peak_dbfs":-7.151772605282659,"crest_factor_db":17.390604065584682,"mean":0.00013510598764511118,"zero_energy":false},"rms_difference_db":-0.4427097508261326},"full_scale":{"reference":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":4.638263320592165e-07,"output_rms_dbfs":-63.33644599502702,"output_zero_energy":false,"output_peak":0.007811259012669325,"output_energy_fraction":2.4240589050999985e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":71,"both_estimated_voiced_frames":61,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9838709677419355,"absolute_difference_cents_median":2.193213445276181,"absolute_difference_cents_p95":12.010116498389618,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.0001593748762509214,"log_rms_mae_db":1.3284635356172834,"rms_envelope_nrmse":0.11046854129054846,"envelope_correlation":0.9905109807157902,"rms_gain_error_db":-0.4426870971924504,"reference_mean_power":0.003898496329889168,"output_mean_power":0.0035206951336555605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.3730898670333144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011160072986886684,"output_power":0.0010377921905401254,"reference_active":true,"delta_db":-0.3155663662271201},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003924601048841299,"output_power":0.00040723089859326405,"reference_active":true,"delta_db":0.16045207038225967},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.521600297301381e-05,"output_power":8.254501884520845e-05,"reference_active":true,"delta_db":-0.13830288893272907}],"active_band_count":3,"active_band_mae_db":0.20477377518070294}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00014354118626608257,"log_rms_mae_db":1.1005096604132838,"rms_envelope_nrmse":0.09580323286495071,"envelope_correlation":0.9928140985595013,"rms_gain_error_db":-0.4425595786518024,"reference_mean_power":0.00394559944963645,"output_mean_power":0.0035633381412516433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":2.0339309551628037e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010452646972319804,"output_power":0.0009838191289155169,"reference_active":true,"delta_db":-0.2631102025983387},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020675655639960477,"output_power":0.00021171261294818434,"reference_active":true,"delta_db":0.10287441996777685},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.033282667633225e-06,"output_power":6.275927309410248e-06,"reference_active":true,"delta_db":0.17124231661701356}],"active_band_count":3,"active_band_mae_db":0.17907564639437637}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.00014555177137634006,"log_rms_mae_db":1.8243606377115715,"rms_envelope_nrmse":0.12737223608375603,"envelope_correlation":0.9889873426011013,"rms_gain_error_db":-0.4645789916763748,"reference_mean_power":0.0025948122953778335,"output_mean_power":0.002331567795112754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.091667884405576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010721637846984168,"output_power":0.0009954869886418486,"reference_active":true,"delta_db":-0.32225545616052154},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045198280410970506,"output_power":0.0004540831737143752,"reference_active":true,"delta_db":0.020134969980457523},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681659745338991e-05,"output_power":5.853441481036587e-05,"reference_active":true,"delta_db":0.1293605912026683}],"active_band_count":3,"active_band_mae_db":0.1572503391145491}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00013266563413733552,"log_rms_mae_db":1.3930007767745454,"rms_envelope_nrmse":0.11153889878360435,"envelope_correlation":0.9913339524723629,"rms_gain_error_db":-0.4645401675607932,"reference_mean_power":0.002626297782108314,"output_mean_power":0.0023598801658436465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":1.815151890487515e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010442677388382842,"output_power":0.0009814195749379812,"reference_active":true,"delta_db":-0.26957145336023236},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292871431701824,"output_power":0.000231183883608464,"reference_active":true,"delta_db":0.03577851881997168},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.238038641621491e-06,"output_power":6.981391603503641e-06,"reference_active":true,"delta_db":-0.15678898000472022}],"active_band_count":3,"active_band_mae_db":0.15404631739497476}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":1.8205479831220202,"rms_envelope_nrmse":0.1892285705181884,"envelope_correlation":0.9736489901370204,"rms_gain_error_db":-0.8929604619595872,"reference_mean_power":0.00015194739829742755,"output_mean_power":0.0001237078394745357},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":8.081548208249737e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5556653827814036e-05,"output_power":3.942704894794038e-05,"reference_active":true,"delta_db":-0.6275754526923665},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1818343242486448e-05,"output_power":9.903648802071528e-06,"reference_active":true,"delta_db":-0.7676136789849697},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7000790662746814e-06,"output_power":4.301345278058368e-06,"reference_active":true,"delta_db":0.6539330116648011}],"active_band_count":3,"active_band_mae_db":0.6830407144473791}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":1.358920660343035,"rms_envelope_nrmse":0.1400696660251677,"envelope_correlation":0.988619399565027,"rms_gain_error_db":-0.8919628089667915,"reference_mean_power":0.00015370989511122725,"output_mean_power":0.0001251715255527857},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":6.838335178235257e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.320123857018819e-05,"output_power":3.785203217416061e-05,"reference_active":true,"delta_db":-0.574069975536996},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064245816738489e-06,"output_power":4.968560330925958e-06,"reference_active":true,"delta_db":-0.8654622966431982},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0299327724151418e-07,"output_power":2.591851888021402e-07,"reference_active":true,"delta_db":1.0612852491984113}],"active_band_count":3,"active_band_mae_db":0.8336058404595352}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":1.9478390025064167,"rms_envelope_nrmse":0.18723998240520098,"envelope_correlation":0.9783696845234326,"rms_gain_error_db":-0.831447714117337,"reference_mean_power":0.00013710154501992005,"output_mean_power":0.00011321333334698305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":1.0763544842397469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510390122735278e-05,"output_power":3.6662047046239946e-05,"reference_active":true,"delta_db":-0.8999739708829407},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.111326790734689e-05,"output_power":2.8283669681418125e-05,"reference_active":true,"delta_db":-0.41409872007419385},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.435692144349248e-06,"output_power":4.6360024325883115e-06,"reference_active":true,"delta_db":-0.6911119689819873}],"active_band_count":3,"active_band_mae_db":0.6683948866463739}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":1.4882910796253486,"rms_envelope_nrmse":0.13315226032327537,"envelope_correlation":0.9917265012041844,"rms_gain_error_db":-0.8302055049953165,"reference_mean_power":0.00013864857515666153,"output_mean_power":0.00011452356543403045},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":9.213681129362451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3816318671200005e-05,"output_power":3.721603884277603e-05,"reference_active":true,"delta_db":-0.7090574018118991},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.911163628753254e-05,"output_power":1.6896919142502004e-05,"reference_active":true,"delta_db":-0.5349034595322313},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5242776229186793e-07,"output_power":3.2268412207720176e-07,"reference_active":true,"delta_db":-0.38292516250203823}],"active_band_count":3,"active_band_mae_db":0.5422953412820563}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.472341775894165,"predicted_mos":3.472341775894165,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":110592,"input_sample_rate":44100,"input_seconds":2.507755102040816,"input_float32_sha256":"8ef80671f793459a02024497a94b727154ccaed83022f2c3eff8bb457ea6b517","internal_samples":40125,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1213737789985316},"started_at":"2026-10-13T07:24:48.460095+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00467584100078966,"native_input_samples":110471,"native_input_sha256":"374eaaa58146533aa645d05b89cac6f0eeb987b0ee48548b27d2d501bdd9af51","output_sha256":"5e8a0162bac8bb6f6d8280620eab8aa2a4b36d0ceb82c44b3c3732f47befa8e8","output_samples":110592,"length_delta_samples":121,"analysis_s":0.05627781600014714,"synthesis_s":13.150791368998398,"total_s":13.207069184998545,"rtf":5.272259245036578,"output_peak":0.37188780307769775,"output_rms":0.05881006742279088,"finished_at":"2026-10-13T07:25:01.679812+00:00","evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"5e8a0162bac8bb6f6d8280620eab8aa2a4b36d0ceb82c44b3c3732f47befa8e8","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":40080,"output_samples":40125,"length_delta_samples":45,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.60796248821626,"output_peak":0.3704039697727873},"common_prefix":{"samples":40080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988785046728972,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.630792440550925,"spectral_convergence":0.48162835361409356,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.654903714690848,"spectral_convergence":0.3735500193492042,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.448099108241108,"spectral_convergence":0.43748882964684566,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":6.793823245962357e-05,"delta_db":-4.808828305301626,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.0026264164425457565,"delta_db":-0.04886266121217817,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.0004811373180347035,"delta_db":-1.3654867200223795,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":5.400812129581587e-05,"delta_db":-0.6660906335355703,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.00015318819959734615,"delta_db":-0.7465240501814479,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":8.221468910122913e-05,"delta_db":-1.8283630488039846,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5773592365095308,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5034222657662949,"pitch":null,"output_rms_dbfs":-24.603089185807498,"output_peak":0.3704039697727873,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.577931754494294},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40080,"output_samples":40080,"intelligibility":{"estoi":{"value":0.7549464400788534,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06237587544550748,"rms_dbfs":-24.099666920041205,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.680906994925625,"mean":-1.5409549553237275e-05,"zero_energy":false},"output":{"rms":0.058863426701428745,"rms_dbfs":-24.603089185807498,"peak":0.3704039697727873,"peak_dbfs":-8.626487349025759,"crest_factor_db":15.97660183678174,"mean":-6.929111435391274e-05,"zero_energy":false},"rms_difference_db":-0.503422265766293},"full_scale":{"reference":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":3.939418244679467e-07,"output_rms_dbfs":-64.04567908067811,"output_zero_energy":false,"output_peak":0.004021809762967623,"output_energy_fraction":2.0878106662668995e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":75,"both_estimated_voiced_frames":59,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9516129032258065,"absolute_difference_cents_median":21.43876793174551,"absolute_difference_cents_p95":88.21980173144699,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.0001593748762509214,"log_rms_mae_db":1.5254773246422997,"rms_envelope_nrmse":0.13017440712422296,"envelope_correlation":0.986676845341449,"rms_gain_error_db":-0.5034030636339564,"reference_mean_power":0.003898496329889168,"output_mean_power":0.003471816983703549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.3730898670333144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011160072986886684,"output_power":0.0010713263391292477,"reference_active":true,"delta_db":-0.1774525249203154},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003924601048841299,"output_power":0.0003521824042757709,"reference_active":true,"delta_db":-0.47027861611428473},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.521600297301381e-05,"output_power":7.033217458652387e-05,"reference_active":true,"delta_db":-0.8336711442475798}],"active_band_count":3,"active_band_mae_db":0.4938007617607267}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00014354118626608257,"log_rms_mae_db":0.9946974161627022,"rms_envelope_nrmse":0.08720693941874548,"envelope_correlation":0.995359053720226,"rms_gain_error_db":-0.5033263849754447,"reference_mean_power":0.00394559944963645,"output_mean_power":0.003513826840107916},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":2.0339309551628037e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010452646972319804,"output_power":0.0009996453994430927,"reference_active":true,"delta_db":-0.19380311158517438},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020675655639960477,"output_power":0.00018470107432820476,"reference_active":true,"delta_db":-0.4898986867905058},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.033282667633225e-06,"output_power":5.306424091191468e-06,"reference_active":true,"delta_db":-0.5575171709602367}],"active_band_count":3,"active_band_mae_db":0.4137396564453056}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.00014555177137634006,"log_rms_mae_db":2.639209152879543,"rms_envelope_nrmse":0.14152705488064693,"envelope_correlation":0.9845849665079899,"rms_gain_error_db":-0.13980350880639228,"reference_mean_power":0.0025948122953778335,"output_mean_power":0.0025126129642785338},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.091667884405576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010721637846984168,"output_power":0.0010248715623210303,"reference_active":true,"delta_db":-0.1959169094890782},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045198280410970506,"output_power":0.00045734736400670166,"reference_active":true,"delta_db":0.05124268108268249},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681659745338991e-05,"output_power":5.187724807301364e-05,"reference_active":true,"delta_db":-0.3949829174457301}],"active_band_count":3,"active_band_mae_db":0.21404750267249692}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00013266563413733552,"log_rms_mae_db":1.9136927548033684,"rms_envelope_nrmse":0.10189001201308177,"envelope_correlation":0.9913688922360686,"rms_gain_error_db":-0.1398380206739449,"reference_mean_power":0.002626297782108314,"output_mean_power":0.002543080834181075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":1.815151890487515e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010442677388382842,"output_power":0.0009983407784117492,"reference_active":true,"delta_db":-0.1953305055488655},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292871431701824,"output_power":0.00022680128329332345,"reference_active":true,"delta_db":-0.047341956308307884},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.238038641621491e-06,"output_power":6.79417413325878e-06,"reference_active":true,"delta_db":-0.2748422378741258}],"active_band_count":3,"active_band_mae_db":0.17250489991043305}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":3.183280456380121,"rms_envelope_nrmse":0.2993765321389519,"envelope_correlation":0.923609095090979,"rms_gain_error_db":-0.8859873044549544,"reference_mean_power":0.00015194739829742755,"output_mean_power":0.00012390662789833753},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":8.081548208249737e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5556653827814036e-05,"output_power":4.5748358682311686e-05,"reference_active":true,"delta_db":0.01823700156545119},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1818343242486448e-05,"output_power":1.0121713704436384e-05,"reference_active":true,"delta_db":-0.673025501245683},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7000790662746814e-06,"output_power":2.0353697226085526e-06,"reference_active":true,"delta_db":-2.5956769469231595}],"active_band_count":3,"active_band_mae_db":1.0956464832447645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":2.60541719608289,"rms_envelope_nrmse":0.22641595309476004,"envelope_correlation":0.9566829349786301,"rms_gain_error_db":-0.885291553415014,"reference_mean_power":0.00015370989511122725,"output_mean_power":0.00012536395096107426},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":6.838335178235257e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.320123857018819e-05,"output_power":4.3483789855413324e-05,"reference_active":true,"delta_db":0.02831190133669416},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064245816738489e-06,"output_power":5.050446259911877e-06,"reference_active":true,"delta_db":-0.7944704300126105},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0299327724151418e-07,"output_power":1.0222482382927802e-07,"reference_active":true,"delta_db":-2.979252843497365}],"active_band_count":3,"active_band_mae_db":1.2673450582822232}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":2.9788404161149273,"rms_envelope_nrmse":0.24284140398336865,"envelope_correlation":0.9638474928512493,"rms_gain_error_db":-1.0906450034370978,"reference_mean_power":0.00013710154501992005,"output_mean_power":0.00010665417203902209},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":1.0763544842397469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510390122735278e-05,"output_power":3.85544636599915e-05,"reference_active":true,"delta_db":-0.6813944153826672},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.111326790734689e-05,"output_power":1.8792463640443792e-05,"reference_active":true,"delta_db":-2.18961909933597},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.435692144349248e-06,"output_power":4.995123256057246e-06,"reference_active":true,"delta_db":-0.3670864303993424}],"active_band_count":3,"active_band_mae_db":1.0793666483726598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":2.6405454518602918,"rms_envelope_nrmse":0.1994944087101336,"envelope_correlation":0.9778964975483098,"rms_gain_error_db":-1.0887759049720938,"reference_mean_power":0.00013864857515666153,"output_mean_power":0.00010790406859264844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":9.213681129362451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3816318671200005e-05,"output_power":3.9014669397309816e-05,"reference_active":true,"delta_db":-0.5040795536704611},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.911163628753254e-05,"output_power":1.0379459311347149e-05,"reference_active":true,"delta_db":-2.6512314103102694},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5242776229186793e-07,"output_power":4.515403610896293e-07,"reference_active":true,"delta_db":1.0762646356276813}],"active_band_count":3,"active_band_mae_db":1.4105251998694708}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"golf","status":"measured","mos":{"status":"measured","value":2.2526674270629883,"predicted_mos":2.2526674270629883,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":24000,"input_seconds":2.5,"input_float32_sha256":"6d7f0078f389c590f0ab6d28256be00a36266f12d0b1db7331bc05e2e08ee8f4","internal_samples":40000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.137684671999523},"started_at":"2026-10-13T05:02:52.866734+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018444469997120905,"native_input_samples":60120,"native_input_sha256":"5ede71dd1f2b35c0cba977e27693dbd1bc8000e6862704dd4648fab0f8ae6b8d","output_sha256":"5afadb5ca69f703cbf0699d2abc046c79caeed8dd1e6c6e18bb47b30973083fe","output_samples":60000,"length_delta_samples":-120,"analysis_s":0.42777442900023743,"synthesis_s":0.08527693799987901,"total_s":0.5130513670001164,"rtf":0.2048109249501463,"output_peak":0.3391585350036621,"output_rms":0.05794033911540612,"finished_at":"2026-10-13T05:02:53.391117+00:00","analysis_stages":{"external_f0_s":0.022520023000652145},"evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"5afadb5ca69f703cbf0699d2abc046c79caeed8dd1e6c6e18bb47b30973083fe","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":40080,"output_samples":40000,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.739013241943777,"output_peak":0.33925251442470844},"common_prefix":{"samples":40000,"sample_rate":16000,"reference_coverage":0.998003992015968,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40000,"output_samples":40000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.489807016988563,"spectral_convergence":0.48452890437889906,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.784613116056059,"spectral_convergence":0.2940954985488583,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.839374750701435,"spectral_convergence":0.30856556913120853,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020599522572520152,"output_power":0.0001708933457118054,"delta_db":-0.8113200262207211,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0026576780163762416,"output_power":0.0023967503776537976,"delta_db":-0.44879559573144484,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006639860110156221,"output_power":0.00048672536320337306,"delta_db":-1.3487495179584204,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.30858248148848e-05,"output_power":6.106700244152036e-05,"delta_db":-0.14125183038499534,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018228318083359283,"output_power":0.00013665283297097623,"delta_db":-1.25127958848435,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012550277819814402,"output_power":0.00010605013466550154,"delta_db":-0.7314211537114694,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7888029520819001,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6480232026765687,"pitch":null,"output_rms_dbfs":-24.739013241943777,"output_peak":0.33925251442470844,"reference_rms_dbfs":-24.090990039267204,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.037931627915352},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40000,"output_samples":40000,"intelligibility":{"estoi":{"value":0.8844718928206624,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06243821775934741,"rms_dbfs":-24.090990039267204,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.672230114151628,"mean":-1.542816162109375e-05,"zero_energy":false},"output":{"rms":0.05794945260006322,"rms_dbfs":-24.739013241943777,"peak":0.33925251442470844,"peak_dbfs":-9.389538496254454,"crest_factor_db":15.34947474568932,"mean":0.00041805645379986773,"zero_energy":false},"rms_difference_db":-0.6480232026765727},"full_scale":{"reference":{"sample_count":40000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":4.567564428083192e-07,"output_rms_dbfs":-63.403153179625804,"output_zero_energy":false,"output_peak":0.00394735095207961,"output_energy_fraction":2.5026713920729046e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":72,"both_estimated_voiced_frames":61,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9838709677419355,"absolute_difference_cents_median":5.732999221691588,"absolute_difference_cents_p95":19.549745837473377,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":499,"rms_floor":0.0001593748762509214,"log_rms_mae_db":1.9529788555859766,"rms_envelope_nrmse":0.15782389535908636,"envelope_correlation":0.980518648921806,"rms_gain_error_db":-0.6480339079134931,"reference_mean_power":0.003906308675351888,"output_mean_power":0.003364830309251924},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":499,"frequency_resolution_hz":0.4008016032064128,"power_floor":2.3749434367429194e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011186815136079668,"output_power":0.0010719398716522183,"reference_active":true,"delta_db":-0.18536036204427242},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003927756184098066,"output_power":0.00033197876977768254,"reference_active":true,"delta_db":-0.7303420984887442},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.523503788148054e-05,"output_power":6.532424708428096e-05,"reference_active":true,"delta_db":-1.1554374567874797}],"active_band_count":3,"active_band_mae_db":0.6903799724401655}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":493,"rms_floor":0.00014354118626608257,"log_rms_mae_db":1.5955226472201562,"rms_envelope_nrmse":0.11387194986209949,"envelope_correlation":0.9916251583234948,"rms_gain_error_db":-0.6479646761285909,"reference_mean_power":0.0039536021574539445,"output_mean_power":0.0034056224293886454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":493,"frequency_resolution_hz":0.40567951318458423,"power_floor":2.034838565237499e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010477497132038737,"output_power":0.0009897059778987357,"reference_active":true,"delta_db":-0.2475135718394733},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002065992476877692,"output_power":0.00017333569536204084,"reference_active":true,"delta_db":-0.7624072869867854},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.003941766840419e-06,"output_power":4.115487869231449e-06,"reference_active":true,"delta_db":-1.6401514539776696}],"active_band_count":3,"active_band_mae_db":0.8833574376013095}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":499,"rms_floor":0.00014555177137634006,"log_rms_mae_db":2.793584350091297,"rms_envelope_nrmse":0.16456464960051312,"envelope_correlation":0.9810623960170595,"rms_gain_error_db":-0.5669452628945648,"reference_mean_power":0.0026000121727472805,"output_mean_power":0.0022818172289642506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":499,"frequency_resolution_hz":0.4008016032064128,"power_floor":2.0948925533713536e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010746199976486288,"output_power":0.0010001046719495565,"reference_active":true,"delta_db":-0.3120946204095462},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045260265907262186,"output_power":0.00038850082018226803,"reference_active":true,"delta_db":-0.6632516113967966},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681089999216211e-05,"output_power":4.5954098096652296e-05,"reference_active":true,"delta_db":-0.9210742239596228}],"active_band_count":3,"active_band_mae_db":0.6321401519219886}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":493,"rms_floor":0.00013266563413733552,"log_rms_mae_db":2.367546258166411,"rms_envelope_nrmse":0.1176477054951228,"envelope_correlation":0.9911972780912464,"rms_gain_error_db":-0.5669136680558774,"reference_mean_power":0.0026316248108262308,"output_mean_power":0.0023095778481534153},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":493,"frequency_resolution_hz":0.40567951318458423,"power_floor":1.8173706149638225e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010463355507461024,"output_power":0.0009674005396637541,"reference_active":true,"delta_db":-0.3406465602249265},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002295109529469857,"output_power":0.00019718304770916992,"reference_active":true,"delta_db":-0.6593384130437474},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.2055635302877385e-06,"output_power":4.7782743603595695e-06,"reference_active":true,"delta_db":-1.783968690571247}],"active_band_count":3,"active_band_mae_db":0.9279845546133071}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":499,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":2.2724500797388445,"rms_envelope_nrmse":0.25870690174903316,"envelope_correlation":0.9441408143633446,"rms_gain_error_db":-0.9815307216211598,"reference_mean_power":0.00015225181533373982,"output_mean_power":0.00012145332457565981},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":499,"frequency_resolution_hz":0.4008016032064128,"power_floor":8.083123076318004e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5687642635029655e-05,"output_power":3.9285423017217725e-05,"reference_active":true,"delta_db":-0.6556731625617253},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1813407654926244e-05,"output_power":1.276977672658479e-05,"reference_active":true,"delta_db":0.33808113122097616},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6870291276886195e-06,"output_power":3.944718866452469e-06,"reference_active":true,"delta_db":0.2933948863873739}],"active_band_count":3,"active_band_mae_db":0.4290497267233584}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":493,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":1.8766041477864945,"rms_envelope_nrmse":0.1847086864723426,"envelope_correlation":0.9743930570409081,"rms_gain_error_db":-0.9811185554533877,"reference_mean_power":0.00015402156608666738,"output_mean_power":0.00012287673963242926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":493,"frequency_resolution_hz":0.40567951318458423,"power_floor":6.837450093231805e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3320162140028636e-05,"output_power":3.751017801609169e-05,"reference_active":true,"delta_db":-0.6254094821329261},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.036418583199957e-06,"output_power":6.3850859768998014e-06,"reference_active":true,"delta_db":0.24387402113306605},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0213081933347455e-07,"output_power":2.18375986126802e-07,"reference_active":true,"delta_db":0.33572342746610906}],"active_band_count":3,"active_band_mae_db":0.4016689769107004}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":499,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":2.8187107704194716,"rms_envelope_nrmse":0.19359713552795796,"envelope_correlation":0.9753298696925192,"rms_gain_error_db":-0.21553503004143743,"reference_mean_power":0.00013737628647497862,"output_mean_power":0.0001307248852995327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":499,"frequency_resolution_hz":0.4008016032064128,"power_floor":1.0704353027727627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.483719818774276e-05,"output_power":4.3447991033672924e-05,"reference_active":true,"delta_db":-0.1366876617147111},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.108944080779335e-05,"output_power":3.444883499750556e-05,"reference_active":true,"delta_db":0.445616287244397},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.414683169789874e-06,"output_power":7.224216061649751e-06,"reference_active":true,"delta_db":1.252176766035009}],"active_band_count":3,"active_band_mae_db":0.6114935716647056}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40000,"uncovered_tail_samples":0,"metrics":{"frame_count":493,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":2.5287979023955245,"rms_envelope_nrmse":0.1380449376351849,"envelope_correlation":0.9875951653082424,"rms_gain_error_db":-0.21440007293629285,"reference_mean_power":0.00013892959647180544,"output_mean_power":0.00013223754166968372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":493,"frequency_resolution_hz":0.40567951318458423,"power_floor":9.154904580101656e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.355830105372237e-05,"output_power":4.411923779095997e-05,"reference_active":true,"delta_db":0.05557068702388385},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9066905911441762e-05,"output_power":2.1155763649463546e-05,"reference_active":true,"delta_db":0.45148482989344924},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5137060160066173e-07,"output_power":4.7672504481769144e-07,"reference_active":true,"delta_db":1.3250254602573652}],"active_band_count":3,"active_band_mae_db":0.6106936590582328}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.9060556888580322,"predicted_mos":2.9060556888580322,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60240,"input_sample_rate":24000,"input_seconds":2.51,"input_float32_sha256":"9e7d7ec6d0f2efb10dfa53f14ec02126d470e7c81dba95f7c497f16413bbfd68","internal_samples":40160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.209132894000504},"started_at":"2026-10-13T07:42:57.667414+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0025456900002609473,"native_input_samples":60120,"native_input_sha256":"5ede71dd1f2b35c0cba977e27693dbd1bc8000e6862704dd4648fab0f8ae6b8d","output_sha256":"c4ad22475cacdd26a23aa7f6cb6bc7fad8381fa1810e887a529f16dd55d1f8a5","output_samples":60240,"length_delta_samples":120,"analysis_s":0.9453824369993526,"synthesis_s":6.533305453001958,"total_s":7.478687890001311,"rtf":2.985504147705114,"output_peak":0.4094373881816864,"output_rms":0.060382556400313286,"finished_at":"2026-10-13T07:43:05.156829+00:00","analysis_stages":{"external_f0_s":0.6510015829990152},"evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"c4ad22475cacdd26a23aa7f6cb6bc7fad8381fa1810e887a529f16dd55d1f8a5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":40080,"output_samples":40160,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.37984002793967,"output_peak":0.396206668345816},"common_prefix":{"samples":40080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.99800796812749,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.2909489000005,"spectral_convergence":0.5135580268848539,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.58550978837959,"spectral_convergence":0.32922746076454146,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.507204068903026,"spectral_convergence":0.3180132517420358,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.00020394276109381748,"delta_db":-0.03488779492951882,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.002536173332023316,"delta_db":-0.20070923118739123,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.0005870790260648896,"delta_db":-0.5012114992854189,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":5.861771824180941e-05,"delta_db":-0.3103922466095697,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.00017484967473809675,"delta_db":-0.1721288824923286,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":9.429188955717966e-05,"delta_db":-1.23311384861037,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4087405838524329,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2715134481562661,"pitch":null,"output_rms_dbfs":-24.371180368197468,"output_peak":0.396206668345816,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.794554252427705},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40080,"output_samples":40080,"intelligibility":{"estoi":{"value":0.8905605590565798,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06237587544550748,"rms_dbfs":-24.099666920041205,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.680906994925625,"mean":-1.5409549553237275e-05,"zero_energy":false},"output":{"rms":0.06045621888374354,"rms_dbfs":-24.371180368197468,"peak":0.396206668345816,"peak_dbfs":-8.041564387084193,"crest_factor_db":16.329615981113278,"mean":7.883745025442552e-05,"zero_energy":false},"rms_difference_db":-0.2715134481562629},"full_scale":{"reference":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":3.3070386079133173e-07,"output_rms_dbfs":-64.80560734900075,"output_zero_energy":false,"output_peak":0.0052187440107905915,"output_energy_fraction":1.6615270014993203e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":66,"both_estimated_voiced_frames":60,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.967741935483871,"absolute_difference_cents_median":8.024247803673909,"absolute_difference_cents_p95":34.73471910781186,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.0001593748762509214,"log_rms_mae_db":2.163077903250452,"rms_envelope_nrmse":0.19489708893732263,"envelope_correlation":0.9649976135152102,"rms_gain_error_db":-0.2715214608344804,"reference_mean_power":0.003898496329889168,"output_mean_power":0.0036622246679697988},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.3730898670333144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011160072986886684,"output_power":0.0011142735324058699,"reference_active":true,"delta_db":-0.0067522015228495755},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003924601048841299,"output_power":0.00038811917696843044,"reference_active":true,"delta_db":-0.04830413819503497},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.521600297301381e-05,"output_power":7.301077533317035e-05,"reference_active":true,"delta_db":-0.6713419948328005}],"active_band_count":3,"active_band_mae_db":0.24213277818356171}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00014354118626608257,"log_rms_mae_db":1.5741191988304712,"rms_envelope_nrmse":0.1089461689773802,"envelope_correlation":0.9887567762520925,"rms_gain_error_db":-0.2715197610285615,"reference_mean_power":0.00394559944963645,"output_mean_power":0.0037064745140659343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":2.0339309551628037e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010452646972319804,"output_power":0.0010467014804237701,"reference_active":true,"delta_db":0.005965556390573107},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020675655639960477,"output_power":0.00020395164744807336,"reference_active":true,"delta_db":-0.059320724924180446},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.033282667633225e-06,"output_power":5.700834408027617e-06,"reference_active":true,"delta_db":-0.24615246841832933}],"active_band_count":3,"active_band_mae_db":0.1038129165776943}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.00014555177137634006,"log_rms_mae_db":3.0876033258501856,"rms_envelope_nrmse":0.22469339286957205,"envelope_correlation":0.9606034901603518,"rms_gain_error_db":-0.22623793231862435,"reference_mean_power":0.0025948122953778335,"output_mean_power":0.002463100622090111},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.091667884405576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010721637846984168,"output_power":0.0010352264178955161,"reference_active":true,"delta_db":-0.1522578743094945},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045198280410970506,"output_power":0.0004281336706972657,"reference_active":true,"delta_db":-0.23542527802101143},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681659745338991e-05,"output_power":6.27638253122889e-05,"reference_active":true,"delta_db":0.4323418318827751}],"active_band_count":3,"active_band_mae_db":0.273341661404427}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00013266563413733552,"log_rms_mae_db":2.416712301064162,"rms_envelope_nrmse":0.13325763237594482,"envelope_correlation":0.9851856953085134,"rms_gain_error_db":-0.22622827984479077,"reference_mean_power":0.002626297782108314,"output_mean_power":0.0024929934584658597},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":1.815151890487515e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010442677388382842,"output_power":0.0010099157994188743,"reference_active":true,"delta_db":-0.14526694819356284},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292871431701824,"output_power":0.00021722023309603852,"reference_active":true,"delta_db":-0.2347942780055021},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.238038641621491e-06,"output_power":7.875390030001941e-06,"reference_active":true,"delta_db":0.3665117416016231}],"active_band_count":3,"active_band_mae_db":0.24885765593356268}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":2.2010415153400418,"rms_envelope_nrmse":0.27231127083043966,"envelope_correlation":0.9328995689297496,"rms_gain_error_db":-0.12066407080763167,"reference_mean_power":0.00015194739829742755,"output_mean_power":0.000147783810806033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":8.081548208249737e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5556653827814036e-05,"output_power":4.4535725162804946e-05,"reference_active":true,"delta_db":-0.09843289100641689},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1818343242486448e-05,"output_power":1.1796078848345925e-05,"reference_active":true,"delta_db":-0.008189322902813916},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7000790662746814e-06,"output_power":3.430808173940177e-06,"reference_active":true,"delta_db":-0.3281456840255075}],"active_band_count":3,"active_band_mae_db":0.14492263264491276}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":1.4496748825412777,"rms_envelope_nrmse":0.1309050499603672,"envelope_correlation":0.9829119381094367,"rms_gain_error_db":-0.11997899010635005,"reference_mean_power":0.00015370989511122725,"output_mean_power":0.00014952159707639094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":6.838335178235257e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.320123857018819e-05,"output_power":4.207356754878477e-05,"reference_active":true,"delta_db":-0.11486859362885365},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064245816738489e-06,"output_power":6.1358739894765394e-06,"reference_active":true,"delta_db":0.05099634773549481},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0299327724151418e-07,"output_power":1.9604310894294045e-07,"reference_active":true,"delta_db":-0.1513007398928591}],"active_band_count":3,"active_band_mae_db":0.10572189375240253}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":2.413606733360262,"rms_envelope_nrmse":0.2285839844344187,"envelope_correlation":0.9625957851092126,"rms_gain_error_db":-0.6680799000544758,"reference_mean_power":0.00013710154501992005,"output_mean_power":0.00011755317376720153},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":1.0763544842397469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510390122735278e-05,"output_power":3.974457015831452e-05,"reference_active":true,"delta_db":-0.5493630306592594},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.111326790734689e-05,"output_power":2.6094564021713365e-05,"reference_active":true,"delta_db":-0.7639558336080041},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.435692144349248e-06,"output_power":4.64854059729194e-06,"reference_active":true,"delta_db":-0.6793822391792803}],"active_band_count":3,"active_band_mae_db":0.6642337011488479}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":1.6889513272847072,"rms_envelope_nrmse":0.1538970590924183,"envelope_correlation":0.9842910846356567,"rms_gain_error_db":-0.6687164080255449,"reference_mean_power":0.00013864857515666153,"output_mean_power":0.00011886220157701554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":9.213681129362451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3816318671200005e-05,"output_power":3.8489952141366066e-05,"reference_active":true,"delta_db":-0.562885154169546},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.911163628753254e-05,"output_power":1.5932417294241946e-05,"reference_active":true,"delta_db":-0.7901619906749242},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5242776229186793e-07,"output_power":2.8695526520402276e-07,"reference_active":true,"delta_db":-0.8925591456616255}],"active_band_count":3,"active_band_mae_db":0.7485354301686987}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.7724897861480713,"predicted_mos":2.7724897861480713,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60360,"input_sample_rate":24000,"input_seconds":2.515,"input_float32_sha256":"5f6773fba673ebec12675d0a38b3383634bb3f85ee247c24cc45ffbabdad5dab","internal_samples":40240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0016688230007276},"started_at":"2026-10-13T05:11:12.723476+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0012991000003239606,"native_input_samples":60120,"native_input_sha256":"5ede71dd1f2b35c0cba977e27693dbd1bc8000e6862704dd4648fab0f8ae6b8d","output_sha256":"1d29582107e32fd2f91c2c10def9e759a03c53270a20c0fa752e4333c38e1015","output_samples":60360,"length_delta_samples":240,"analysis_s":0.25792562999959046,"synthesis_s":0.06738040600066597,"total_s":0.3253060360002564,"rtf":0.12986268902205847,"output_peak":0.3794121742248535,"output_rms":0.06104311767086528,"finished_at":"2026-10-13T05:11:13.055488+00:00","analysis_stages":{"external_f0_s":0.044129170000815066},"evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"1d29582107e32fd2f91c2c10def9e759a03c53270a20c0fa752e4333c38e1015","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":40080,"output_samples":40240,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.285472662028383,"output_peak":0.378070061147317},"common_prefix":{"samples":40080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9960238568588469,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.100382614743113,"spectral_convergence":0.24458287657637806,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.752866688961304,"spectral_convergence":0.22973914372539597,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.323656066618177,"spectral_convergence":0.2665990351507745,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.0002124979768254963,"delta_db":0.1435772529983628,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.002594383219318904,"delta_db":-0.10215728107329405,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.0006087692641586415,"delta_db":-0.3436499757271385,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":5.901002035519648e-05,"delta_db":-0.2814236926253609,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.00016508411029226837,"delta_db":-0.421724435265095,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":0.00010293790359119933,"delta_db":-0.8521040443972684,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.35743944701441993,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16850322109680838,"pitch":null,"output_rms_dbfs":-24.26817014113801,"output_peak":0.378070061147317,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.058968456774198},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40080,"output_samples":40080,"intelligibility":{"estoi":{"value":0.9451664850496292,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06237587544550748,"rms_dbfs":-24.099666920041205,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.680906994925625,"mean":-1.5409549553237275e-05,"zero_energy":false},"output":{"rms":0.061177467212542455,"rms_dbfs":-24.26817014113801,"peak":0.378070061147317,"peak_dbfs":-8.448554249274942,"crest_factor_db":15.81961589186307,"mean":3.085929840818949e-05,"zero_energy":false},"rms_difference_db":-0.16850322109680604},"full_scale":{"reference":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":5.020759542575628e-07,"output_rms_dbfs":-62.99230577636447,"output_zero_energy":false,"output_peak":0.005577977405966303,"output_energy_fraction":2.463409080439138e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":59,"both_estimated_voiced_frames":58,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.9354838709677419,"absolute_difference_cents_median":1.243197820262326,"absolute_difference_cents_p95":5.193548210166329,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.0001593748762509214,"log_rms_mae_db":0.8551415575726731,"rms_envelope_nrmse":0.08955766428520304,"envelope_correlation":0.9927790673659973,"rms_gain_error_db":-0.1685011238225416,"reference_mean_power":0.003898496329889168,"output_mean_power":0.0037501359953982555},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.3730898670333144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011160072986886684,"output_power":0.0011177078583583032,"reference_active":true,"delta_db":0.006612694738863261},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003924601048841299,"output_power":0.0003733848617825171,"reference_active":true,"delta_db":-0.21638809298368178},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.521600297301381e-05,"output_power":5.6889516982068805e-05,"reference_active":true,"delta_db":-1.7548891348519384}],"active_band_count":3,"active_band_mae_db":0.6592966408581612}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00014354118626608257,"log_rms_mae_db":0.5214762496608834,"rms_envelope_nrmse":0.054871272137954605,"envelope_correlation":0.9972899212734357,"rms_gain_error_db":-0.16849795914302138,"reference_mean_power":0.00394559944963645,"output_mean_power":0.0037954493347134837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":2.0339309551628037e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010452646972319804,"output_power":0.0010509634151488504,"reference_active":true,"delta_db":0.023613153813242697},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020675655639960477,"output_power":0.00019265297725882646,"reference_active":true,"delta_db":-0.30683565247202294},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.033282667633225e-06,"output_power":3.6181378385863894e-06,"reference_active":true,"delta_db":-2.220685651274209}],"active_band_count":3,"active_band_mae_db":0.8503781525198248}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.00014555177137634006,"log_rms_mae_db":1.2273079651161862,"rms_envelope_nrmse":0.11176130263148157,"envelope_correlation":0.9904986573441009,"rms_gain_error_db":-0.18791939609148955,"reference_mean_power":0.0025948122953778335,"output_mean_power":0.002484929131923959},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.091667884405576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010721637846984168,"output_power":0.0010656308234118007,"reference_active":true,"delta_db":-0.026543596457745983},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045198280410970506,"output_power":0.00042667128849890055,"reference_active":true,"delta_db":-0.25028492832959687},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681659745338991e-05,"output_power":3.548074929291467e-05,"reference_active":true,"delta_db":-2.0448243897807257}],"active_band_count":3,"active_band_mae_db":0.7738843048560229}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00013266563413733552,"log_rms_mae_db":0.6978876838558374,"rms_envelope_nrmse":0.0751477793185163,"envelope_correlation":0.9955182274600991,"rms_gain_error_db":-0.18790375182333657,"reference_mean_power":0.002626297782108314,"output_mean_power":0.002515090354802163},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":1.815151890487515e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010442677388382842,"output_power":0.0010349352295774875,"reference_active":true,"delta_db":-0.03898690560466333},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292871431701824,"output_power":0.0002170335720457169,"reference_active":true,"delta_db":-0.2385278489344758},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.238038641621491e-06,"output_power":4.2544040900551805e-06,"reference_active":true,"delta_db":-2.3078215983534767}],"active_band_count":3,"active_band_mae_db":0.8617787842975386}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":0.9971482196616328,"rms_envelope_nrmse":0.13091809245910535,"envelope_correlation":0.985575877013583,"rms_gain_error_db":-0.39679277398187707,"reference_mean_power":0.00015194739829742755,"output_mean_power":0.00013868005043307012},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":8.081548208249737e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5556653827814036e-05,"output_power":4.014055303399105e-05,"reference_active":true,"delta_db":-0.5496846564789639},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1818343242486448e-05,"output_power":1.0400432052488479e-05,"reference_active":true,"delta_db":-0.5550521808704983},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7000790662746814e-06,"output_power":2.200743429796535e-06,"reference_active":true,"delta_db":-2.2564159055405026}],"active_band_count":3,"active_band_mae_db":1.1203842476299883}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":0.5787230620608595,"rms_envelope_nrmse":0.07553152098638923,"envelope_correlation":0.9960549959439808,"rms_gain_error_db":-0.3970396251748042,"reference_mean_power":0.00015370989511122725,"output_mean_power":0.00014028068040989223},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":6.838335178235257e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.320123857018819e-05,"output_power":3.744346621233451e-05,"reference_active":true,"delta_db":-0.6212015273561523},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064245816738489e-06,"output_power":5.157043492629218e-06,"reference_active":true,"delta_db":-0.7037600318525683},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0299327724151418e-07,"output_power":1.1379436815804374e-07,"reference_active":true,"delta_db":-2.513608863772613}],"active_band_count":3,"active_band_mae_db":1.2795234743271113}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":1.0934989896830294,"rms_envelope_nrmse":0.10636433742942793,"envelope_correlation":0.9930470828667601,"rms_gain_error_db":-0.43123488228146534,"reference_mean_power":0.00013710154501992005,"output_mean_power":0.0001241420351255264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":1.0763544842397469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510390122735278e-05,"output_power":4.2027942215706596e-05,"reference_active":true,"delta_db":-0.3067598099716481},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.111326790734689e-05,"output_power":2.595025562479237e-05,"reference_active":true,"delta_db":-0.7880398832849205},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.435692144349248e-06,"output_power":3.819634914293613e-06,"reference_active":true,"delta_db":-1.5323299761125817}],"active_band_count":3,"active_band_mae_db":0.8757098897897168}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":0.7323177390003351,"rms_envelope_nrmse":0.06921780408762565,"envelope_correlation":0.9979592200244504,"rms_gain_error_db":-0.43157220834571774,"reference_mean_power":0.00013864857515666153,"output_mean_power":0.00012553308158371707},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":9.213681129362451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3816318671200005e-05,"output_power":4.098543032413652e-05,"reference_active":true,"delta_db":-0.29006387211382956},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.911163628753254e-05,"output_power":1.5379355591907444e-05,"reference_active":true,"delta_db":-0.9435973324867164},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5242776229186793e-07,"output_power":2.4507573669423526e-07,"reference_active":true,"delta_db":-1.5776979558754318}],"active_band_count":3,"active_band_mae_db":0.9371197201586593}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0038","method":"magphase","status":"measured","mos":{"status":"measured","value":2.524766683578491,"predicted_mos":2.524766683578491,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":120744,"input_sample_rate":48000,"input_seconds":2.5155,"input_float32_sha256":"9df1789726049e3218c9d2e69790382291442916ddf96e64e7b1ec0b18767183","internal_samples":40248,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.412802456999998},"started_at":"2026-10-13T05:17:22.945255+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008874048999132356,"native_input_samples":120240,"native_input_sha256":"fea047940e05100489cd8e4c83e549aa47416505a290250500857b9de8cd795c","output_sha256":"1059b992f8cd773e489b5b14cd36bd8625b8340d6262f8b9b545ee7bf42e59a2","output_samples":120744,"length_delta_samples":504,"analysis_s":1.6536738310005603,"synthesis_s":0.754987981999875,"total_s":2.4086618130004354,"rtf":0.9615416419163415,"output_peak":0.47431085280473706,"output_rms":0.060199295249186525,"finished_at":"2026-10-13T05:17:25.371354+00:00","evaluation_input":{"reference":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"1059b992f8cd773e489b5b14cd36bd8625b8340d6262f8b9b545ee7bf42e59a2","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":40080,"output_samples":40248,"length_delta_samples":168,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.407766830661586,"output_peak":0.4604933895473595},"common_prefix":{"samples":40080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9958258795468098,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":40080,"output_samples":40080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.992299070697587,"spectral_convergence":0.6062387166115596,"spectral_convergence_status":"ok","amplitude_floor":1.4347408585118857e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.992570401944097,"spectral_convergence":0.3144577242701159,"spectral_convergence_status":"ok","amplitude_floor":9.356428468567288e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.772313168703905,"spectral_convergence":0.2592390863463088,"spectral_convergence_status":"ok","amplitude_floor":6.30553407186755e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00020558767458207222,"output_power":0.00019311004091211078,"delta_db":-0.27192218368842774,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002656133227213937,"output_power":0.0024298501550101613,"delta_db":-0.3867036255444916,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000658897280424305,"output_power":0.0006022385496460168,"delta_db":-0.3904916350246384,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.296049498403336e-05,"output_power":6.506030066278668e-05,"delta_db":0.14247932243858902,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018191885630194454,"output_power":0.00021075910834715021,"delta_db":0.639086355423259,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012525230408717123,"output_power":0.0001384663940812829,"delta_db":0.4355865882068776,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3777116183877139,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2899342806229767,"pitch":null,"output_rms_dbfs":-24.38960120066418,"output_peak":0.4604933895473595,"reference_rms_dbfs":-24.099666920041205,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.252394213781862},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":40080,"output_samples":40080,"intelligibility":{"estoi":{"value":0.9205052248902139,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06237587544550748,"rms_dbfs":-24.099666920041205,"peak":0.4256591796875,"peak_dbfs":-7.41875992511558,"crest_factor_db":16.680906994925625,"mean":-1.5409549553237275e-05,"zero_energy":false},"output":{"rms":0.060328140603366096,"rms_dbfs":-24.38960120066418,"peak":0.4604933895473595,"peak_dbfs":-6.735531995714438,"crest_factor_db":17.654069204949742,"mean":-0.0007052764575897416,"zero_energy":false},"rms_difference_db":-0.2899342806229761},"full_scale":{"reference":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":40080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":250,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.5400351179996505e-06,"threshold_dbfs":-55.95160278872963,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":46,"quiet_samples":7360,"reference_mean_power":4.887923790627848e-07,"output_mean_power":7.398771755221173e-07,"output_rms_dbfs":-61.308403700261906,"output_zero_energy":false,"output_peak":0.008436199820244072,"output_energy_fraction":3.7331019583032765e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":243,"reference_estimated_voiced_frames":62,"output_estimated_voiced_frames":64,"both_estimated_voiced_frames":59,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9516129032258065,"absolute_difference_cents_median":5.261462950441154,"absolute_difference_cents_p95":24.60320906411882,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":40080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.0001593748762509214,"log_rms_mae_db":1.8277851848099445,"rms_envelope_nrmse":0.1894331521441834,"envelope_correlation":0.9667503306041556,"rms_gain_error_db":-0.2899224281152984,"reference_mean_power":0.003898496329889168,"output_mean_power":0.0036467407236150215},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.3730898670333144e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011160072986886684,"output_power":0.0010634773686328566,"reference_active":true,"delta_db":-0.20938782578871284},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003924601048841299,"output_power":0.000370922397694482,"reference_active":true,"delta_db":-0.24512457119363337},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.521600297301381e-05,"output_power":7.122700288820294e-05,"reference_active":true,"delta_db":-0.7787648955863169}],"active_band_count":3,"active_band_mae_db":0.4110924308562211}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00014354118626608257,"log_rms_mae_db":1.0082287379846926,"rms_envelope_nrmse":0.10374913081710348,"envelope_correlation":0.9898346707655644,"rms_gain_error_db":-0.289989545901508,"reference_mean_power":0.00394559944963645,"output_mean_power":0.0036907449972515645},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":2.0339309551628037e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010452646972319804,"output_power":0.0009976286911504502,"reference_active":true,"delta_db":-0.20257352106363175},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020675655639960477,"output_power":0.0001914015415998653,"reference_active":true,"delta_db":-0.33513858854050005},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.033282667633225e-06,"output_power":5.191718964109028e-06,"reference_active":true,"delta_db":-0.6524249770283823}],"active_band_count":3,"active_band_mae_db":0.396712362210838}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":0.00014555177137634006,"log_rms_mae_db":2.4373851942392992,"rms_envelope_nrmse":0.19270549773855472,"envelope_correlation":0.9717891887758731,"rms_gain_error_db":-0.4063442530532232,"reference_mean_power":0.0025948122953778335,"output_mean_power":0.0023630424413705305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":2.091667884405576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010721637846984168,"output_power":0.0009800242637087384,"reference_active":true,"delta_db":-0.39024305430811707},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00045198280410970506,"output_power":0.0004161015306390929,"reference_active":true,"delta_db":-0.3592259883744229},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.681659745338991e-05,"output_power":4.930379311881254e-05,"reference_active":true,"delta_db":-0.6159488938837429}],"active_band_count":3,"active_band_mae_db":0.45513931218876086}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":0.00013266563413733552,"log_rms_mae_db":1.5455691601030632,"rms_envelope_nrmse":0.11356616536343418,"envelope_correlation":0.9904750695910568,"rms_gain_error_db":-0.4063402171831425,"reference_mean_power":0.002626297782108314,"output_mean_power":0.002391717852232018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":1.815151890487515e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010442677388382842,"output_power":0.0009579260522475737,"reference_active":true,"delta_db":-0.3747987658797327},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002292871431701824,"output_power":0.00021136714730362173,"reference_active":true,"delta_db":-0.35342217166212464},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.238038641621491e-06,"output_power":6.404393246744294e-06,"reference_active":true,"delta_db":-0.5314290644067219}],"active_band_count":3,"active_band_mae_db":0.41988333398285976}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.97711960335137e-05,"log_rms_mae_db":1.9839175314274593,"rms_envelope_nrmse":0.3039308449951509,"envelope_correlation":0.9224636782427097,"rms_gain_error_db":0.36818235419818385,"reference_mean_power":0.00015194739829742755,"output_mean_power":0.00016539085934849057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":8.081548208249737e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5556653827814036e-05,"output_power":4.996144173773293e-05,"reference_active":true,"delta_db":0.40083145016350785},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1818343242486448e-05,"output_power":1.3030354796015308e-05,"reference_active":true,"delta_db":0.42399641899962404},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7000790662746814e-06,"output_power":4.723128858679138e-06,"reference_active":true,"delta_db":1.0601878984879036}],"active_band_count":3,"active_band_mae_db":0.6283385892170118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.109627865595406e-05,"log_rms_mae_db":1.1174799927841248,"rms_envelope_nrmse":0.14227475848430976,"envelope_correlation":0.9824134968295566,"rms_gain_error_db":0.3674732783813677,"reference_mean_power":0.00015370989511122725,"output_mean_power":0.00016728197759965666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":6.838335178235257e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.320123857018819e-05,"output_power":4.752801118686031e-05,"reference_active":true,"delta_db":0.4145344346780349},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.064245816738489e-06,"output_power":6.590426363709266e-06,"reference_active":true,"delta_db":0.3613671463666591},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0299327724151418e-07,"output_power":2.588214186545349e-07,"reference_active":true,"delta_db":1.055185582071311}],"active_band_count":3,"active_band_mae_db":0.610362387705335}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":500,"rms_floor":4.882145378793849e-05,"log_rms_mae_db":2.2992970663373717,"rms_envelope_nrmse":0.29741309220985795,"envelope_correlation":0.9453061855623516,"rms_gain_error_db":0.741858533075396,"reference_mean_power":0.00013710154501992005,"output_mean_power":0.00016264031331220436},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":500,"frequency_resolution_hz":0.4,"power_floor":1.0763544842397469e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510390122735278e-05,"output_power":5.048974582390488e-05,"reference_active":true,"delta_db":0.48989076899604106},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.111326790734689e-05,"output_power":3.758610008697571e-05,"reference_active":true,"delta_db":0.8208163731349511},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.435692144349248e-06,"output_power":7.436662253116252e-06,"reference_active":true,"delta_db":1.3612320574539778}],"active_band_count":3,"active_band_mae_db":0.8906463998616566}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":40080,"uncovered_tail_samples":0,"metrics":{"frame_count":494,"rms_floor":4.395074730684816e-05,"log_rms_mae_db":1.5580142220667879,"rms_envelope_nrmse":0.18860765931556797,"envelope_correlation":0.9797257730583724,"rms_gain_error_db":0.7423614895161584,"reference_mean_power":0.00013864857515666153,"output_mean_power":0.00016449456750897444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":494,"frequency_resolution_hz":0.4048582995951417,"power_floor":9.213681129362451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3816318671200005e-05,"output_power":4.973850445145005e-05,"reference_active":true,"delta_db":0.5505683615842079},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.911163628753254e-05,"output_power":2.2135331109803815e-05,"reference_active":true,"delta_db":0.6378815094033328},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5242776229186793e-07,"output_power":4.369184968885531e-07,"reference_active":true,"delta_db":0.9333031851734592}],"active_band_count":3,"active_band_mae_db":0.7072510187203332}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"hnm","status":"measured","mos":{"status":"measured","value":2.706559181213379,"predicted_mos":2.706559181213379,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":42560,"input_sample_rate":16000,"input_seconds":2.66,"input_float32_sha256":"c17d5fe64175c6cea9a77e4642cbfccaec1b17112b485eaf9d4eeb3836111d9d","internal_samples":42560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8090566089995264},"started_at":"2026-10-13T04:39:54.757150+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2118000086047687e-05,"native_input_samples":42560,"native_input_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","output_sha256":"2c3020675cd53fc91090c7a141b2fd6231590df2ec87b46bd7cc69e80f711084","output_samples":42560,"length_delta_samples":0,"analysis_s":0.39461671500066586,"synthesis_s":0.2275633969993578,"total_s":0.6221801120000237,"rtf":0.23390229774436977,"output_peak":0.49201988154814175,"output_rms":0.06581647624383849,"finished_at":"2026-10-13T04:39:55.382740+00:00","evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"2c3020675cd53fc91090c7a141b2fd6231590df2ec87b46bd7cc69e80f711084","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.738213892901784,"spectral_convergence":0.3071478479507867,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.40645687830132,"spectral_convergence":0.2955239182757622,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.627597714795621,"spectral_convergence":0.3007004384689766,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.0004465290369458996,"delta_db":0.23778448998421672,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.0016791602460432295,"delta_db":-0.14096002087474768,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0015834041504884124,"delta_db":-0.40704151460676974,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.00030635508510511617,"delta_db":-0.38088037215464876,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.00018403155959962125,"delta_db":0.2813387454813413,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":0.0001323284669734783,"delta_db":0.18612138875983336,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27235442197692633,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19427005075310239,"pitch":null,"output_rms_dbfs":-23.633307463264543,"output_peak":0.49201988154814175,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5907561619995745},"common_prefix":{"samples":42560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.738213892901784,"spectral_convergence":0.3071478479507867,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.40645687830132,"spectral_convergence":0.2955239182757622,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.627597714795621,"spectral_convergence":0.3007004384689766,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.0004465290369458996,"delta_db":0.23778448998421672,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.0016791602460432295,"delta_db":-0.14096002087474768,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0015834041504884124,"delta_db":-0.40704151460676974,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.00030635508510511617,"delta_db":-0.38088037215464876,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.00018403155959962125,"delta_db":0.2813387454813413,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":0.0001323284669734783,"delta_db":0.18612138875983336,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27235442197692633,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19427005075310239,"pitch":null,"output_rms_dbfs":-23.633307463264543,"output_peak":0.49201988154814175,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5907561619995745},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42560,"output_samples":42560,"intelligibility":{"estoi":{"value":0.9079053226846008,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06730512410226613,"rms_dbfs":-23.43903741251144,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.24609129701517,"mean":6.90625126200511e-05,"zero_energy":false},"output":{"rms":0.06581647624383849,"rms_dbfs":-23.633307463264543,"peak":0.49201988154814175,"peak_dbfs":-6.160346957978124,"crest_factor_db":17.47296050528642,"mean":0.00024806974738949785,"zero_energy":false},"rms_difference_db":-0.19427005075310433},"full_scale":{"reference":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":266,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.5796519462661497e-06,"output_mean_power":1.6629741422026372e-06,"output_rms_dbfs":-57.79114503628356,"output_zero_energy":false,"output_peak":0.013949100192487683,"output_energy_fraction":8.370715847900216e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":259,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":99,"both_estimated_voiced_frames":94,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9591836734693877,"absolute_difference_cents_median":1.0290742378416047,"absolute_difference_cents_p95":7.360739271498465,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00020029577381909937,"log_rms_mae_db":0.894898730624989,"rms_envelope_nrmse":0.10124083875265635,"envelope_correlation":0.9900876917900701,"rms_gain_error_db":-0.1942690315570779,"reference_mean_power":0.004538506592765079,"output_mean_power":0.004339963404808376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.8045928558777289e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008527253032911584,"output_power":0.000806174627428608,"reference_active":true,"delta_db":-0.2438002461523883},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002893325022312121,"output_power":0.0002606331458340721,"reference_active":true,"delta_db":-0.4536757697672599},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.722708282094668e-05,"output_power":6.252129234773128e-05,"reference_active":true,"delta_db":-1.9175930894205082}],"active_band_count":3,"active_band_mae_db":0.8716897017800521}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00017887965336918488,"log_rms_mae_db":0.4905347820723073,"rms_envelope_nrmse":0.06698615082637234,"envelope_correlation":0.9954710479888094,"rms_gain_error_db":-0.19427023770951565,"reference_mean_power":0.004590346759257262,"output_mean_power":0.004389534533112479},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.4591224648550268e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007420820227306516,"output_power":0.0006873436712863756,"reference_active":true,"delta_db":-0.3327797247983982},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015779167985385838,"output_power":0.0001437793178840391,"reference_active":true,"delta_db":-0.40387680798502473},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.56670863086132e-06,"output_power":4.0607487849426205e-06,"reference_active":true,"delta_db":-2.087416237545194}],"active_band_count":3,"active_band_mae_db":0.941357590109539}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00017771570730226456,"log_rms_mae_db":1.3191936829069908,"rms_envelope_nrmse":0.12336278305650773,"envelope_correlation":0.9877292776019053,"rms_gain_error_db":-0.2789531724291949,"reference_mean_power":0.003163634497093627,"output_mean_power":0.002966818546873453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.820001870942524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009255585566014783,"output_power":0.0008907905036812171,"reference_active":true,"delta_db":-0.1662832156551985},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002598152375676334,"output_power":0.0002261432764685294,"reference_active":true,"delta_db":-0.6028093772053131},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.760488032618479e-05,"output_power":4.430138161020601e-05,"reference_active":true,"delta_db":-2.4347176291754185}],"active_band_count":3,"active_band_mae_db":1.0679367406786433}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00016184521030355344,"log_rms_mae_db":0.842195086810443,"rms_envelope_nrmse":0.09351903956368181,"envelope_correlation":0.9926329635949539,"rms_gain_error_db":-0.2789543010807581,"reference_mean_power":0.0031997824540876,"output_mean_power":0.003000716888239582},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.5556561049407657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008585656786559776,"output_power":0.000814544893772002,"reference_active":true,"delta_db":-0.228584979629799},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001566221417610138,"output_power":0.00013786756433323475,"reference_active":true,"delta_db":-0.5539105541838902},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.505375823689366e-06,"output_power":4.157769494521034e-06,"reference_active":true,"delta_db":-2.565120354481934}],"active_band_count":3,"active_band_mae_db":1.1158719627652076}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.0001137540358892563,"log_rms_mae_db":1.4798588596789148,"rms_envelope_nrmse":0.2136239201011557,"envelope_correlation":0.9636154208217352,"rms_gain_error_db":-0.30797160662502765,"reference_mean_power":0.0003949518827357507,"output_mean_power":0.0003679146046187435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.625495481863056e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.926011413944637e-05,"output_power":8.038192406637129e-05,"reference_active":true,"delta_db":0.061037040848956885},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0135009616582808e-05,"output_power":1.9458009485419515e-05,"reference_active":true,"delta_db":-0.14853430597558706},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.142336744382214e-06,"output_power":4.723129598634868e-06,"reference_active":true,"delta_db":-1.7961045989387217}],"active_band_count":3,"active_band_mae_db":0.6685586485877552}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":0.8174610953828697,"rms_envelope_nrmse":0.11759384144825653,"envelope_correlation":0.9887632542640423,"rms_gain_error_db":-0.3079868531513047,"reference_mean_power":0.000399452081398279,"output_mean_power":0.00037210542619965404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.3661376304062826e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.270190105592022e-05,"output_power":7.372517558732745e-05,"reference_active":true,"delta_db":0.06070048343056796},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.983801499648588e-06,"output_power":9.518126965304296e-06,"reference_active":true,"delta_db":-0.20744443983709504},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.054433356878422e-07,"output_power":3.106439387688515e-07,"reference_active":true,"delta_db":-2.1140959055695845}],"active_band_count":3,"active_band_mae_db":0.7940802762790825}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":1.5921166658437562,"rms_envelope_nrmse":0.18805696687443035,"envelope_correlation":0.9792168648039978,"rms_gain_error_db":0.4593849167687405,"reference_mean_power":0.00015129473773616958,"output_mean_power":0.00016817534006335249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":2.0948865520528097e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010579126587837111,"output_power":0.00011406738538014363,"reference_active":true,"delta_db":0.3271167303948548},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5885080610072743e-05,"output_power":1.8566319441142847e-05,"reference_active":true,"delta_db":0.6773639536654013},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.603064633980169e-06,"output_power":4.658377833318437e-06,"reference_active":true,"delta_db":-0.8019092175302294}],"active_band_count":3,"active_band_mae_db":0.6021299671968284}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":1.0247661540293003,"rms_envelope_nrmse":0.13279021898652374,"envelope_correlation":0.9900106510521606,"rms_gain_error_db":0.4594261289341498,"reference_mean_power":0.00015301852862864023,"output_mean_power":0.0001700930757761162},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.9708112174484797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.707695341873246e-05,"output_power":0.00010694423109367783,"reference_active":true,"delta_db":0.42041223994801413},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.3675521469727354e-06,"output_power":8.668052858211415e-06,"reference_active":true,"delta_db":0.7059833256881032},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.245758891603436e-07,"output_power":3.072138139993148e-07,"reference_active":true,"delta_db":-0.23875515463897118}],"active_band_count":3,"active_band_mae_db":0.4550502400916962}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.2808353900909424,"predicted_mos":3.2808353900909424,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":42560,"input_sample_rate":16000,"input_seconds":2.66,"input_float32_sha256":"378f302268e63ed31ceab3217fceffdd2eedb1a07f4777e5354dd926862a3399","internal_samples":42560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.033376164999936},"started_at":"2026-10-13T04:43:06.258355+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":7.073500000842614e-05,"native_input_samples":42560,"native_input_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","output_sha256":"f3bbfa55667294d028f8173032143c8a9331c049f794aff5709bbc13789fa721","output_samples":42560,"length_delta_samples":0,"analysis_s":0.693739602000278,"synthesis_s":0.044460013000389154,"total_s":0.7381996150006671,"rtf":0.2775186522558899,"output_peak":0.6444663938578207,"output_rms":0.07565680531281467,"finished_at":"2026-10-13T04:43:07.000139+00:00","evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"f3bbfa55667294d028f8173032143c8a9331c049f794aff5709bbc13789fa721","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.161190571177855,"spectral_convergence":0.5576398967175682,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.279030847853786,"spectral_convergence":0.3126541508769584,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.775178223269717,"spectral_convergence":0.3537500180067617,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.00037830988726999685,"delta_db":-0.48223564610825004,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.002118242881011667,"delta_db":0.8678760960137601,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0025345076837147047,"delta_db":1.635976817283026,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.00038258624322545,"delta_db":0.584162192281921,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.0001803723430634142,"delta_db":0.194115147921061,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":0.00012993315185590902,"delta_db":0.10678831138344544,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6451923684985773,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0159973886777922,"pitch":null,"output_rms_dbfs":-22.42304002383365,"output_peak":0.6444663938578207,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.738466547433785},"common_prefix":{"samples":42560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.161190571177855,"spectral_convergence":0.5576398967175682,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.279030847853786,"spectral_convergence":0.3126541508769584,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.775178223269717,"spectral_convergence":0.3537500180067617,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.00037830988726999685,"delta_db":-0.48223564610825004,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.002118242881011667,"delta_db":0.8678760960137601,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0025345076837147047,"delta_db":1.635976817283026,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.00038258624322545,"delta_db":0.584162192281921,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.0001803723430634142,"delta_db":0.194115147921061,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":0.00012993315185590902,"delta_db":0.10678831138344544,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6451923684985773,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0159973886777922,"pitch":null,"output_rms_dbfs":-22.42304002383365,"output_peak":0.6444663938578207,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.738466547433785},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42560,"output_samples":42560,"intelligibility":{"estoi":{"value":0.9397997609474291,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06730512410226613,"rms_dbfs":-23.43903741251144,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.24609129701517,"mean":6.90625126200511e-05,"zero_energy":false},"output":{"rms":0.07565680531281467,"rms_dbfs":-22.42304002383365,"peak":0.6444663938578207,"peak_dbfs":-3.8159944860760184,"crest_factor_db":18.607045537757628,"mean":3.647784335389975e-08,"zero_energy":false},"rms_difference_db":1.015997388677789},"full_scale":{"reference":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":266,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.5796519462661497e-06,"output_mean_power":2.007453795514072e-06,"output_rms_dbfs":-56.973544418571045,"output_zero_energy":false,"output_peak":0.01209553725077071,"output_energy_fraction":7.647084998392273e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":259,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":130,"both_estimated_voiced_frames":95,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9693877551020408,"absolute_difference_cents_median":4.673530446826713,"absolute_difference_cents_p95":22.3177350041701,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00020029577381909937,"log_rms_mae_db":1.3817324164481588,"rms_envelope_nrmse":0.17496609624460066,"envelope_correlation":0.9867094158026206,"rms_gain_error_db":1.016000348889201,"reference_mean_power":0.004538506592765079,"output_mean_power":0.005734730397196413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.8045928558777289e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008527253032911584,"output_power":0.0011682200362135994,"reference_active":true,"delta_db":1.367155003941325},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002893325022312121,"output_power":0.00038180746284112644,"reference_active":true,"delta_db":1.2044718994150845},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.722708282094668e-05,"output_power":8.853673619372906e-05,"reference_active":true,"delta_db":-0.40663747310643794}],"active_band_count":3,"active_band_mae_db":0.9927547921542824}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00017887965336918488,"log_rms_mae_db":1.017375285721715,"rms_envelope_nrmse":0.14866554327918288,"envelope_correlation":0.9936872354348937,"rms_gain_error_db":1.016005333093569,"reference_mean_power":0.004590346759257262,"output_mean_power":0.005800240842948018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.4591224648550268e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007420820227306516,"output_power":0.001008143003469817,"reference_active":true,"delta_db":1.3307022971754399},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015779167985385838,"output_power":0.00022507851879091122,"reference_active":true,"delta_db":1.542499487576592},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.56670863086132e-06,"output_power":6.476624142008345e-06,"reference_active":true,"delta_db":-0.0599905241678518}],"active_band_count":3,"active_band_mae_db":0.9777307696399613}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00017771570730226456,"log_rms_mae_db":1.721503975720022,"rms_envelope_nrmse":0.21211078691592886,"envelope_correlation":0.9886658226097395,"rms_gain_error_db":1.3678118282691456,"reference_mean_power":0.003163634497093627,"output_mean_power":0.004334784236508464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.820001870942524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009255585566014783,"output_power":0.0013144557955015012,"reference_active":true,"delta_db":1.5234208544611243},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002598152375676334,"output_power":0.00036625680504317446,"reference_active":true,"delta_db":1.4912108475984143},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.760488032618479e-05,"output_power":6.674834323619429e-05,"reference_active":true,"delta_db":-0.654485430009639}],"active_band_count":3,"active_band_mae_db":1.2230390440230592}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00016184521030355344,"log_rms_mae_db":1.2933271156012378,"rms_envelope_nrmse":0.19229284809201888,"envelope_correlation":0.9938607861186051,"rms_gain_error_db":1.3678136717778566,"reference_mean_power":0.0031997824540876,"output_mean_power":0.0043843157111612225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.5556561049407657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008585656786559776,"output_power":0.0011954654157866756,"reference_active":true,"delta_db":1.4376349323561683},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001566221417610138,"output_power":0.00022380706793005367,"reference_active":true,"delta_db":1.5502063910882702},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.505375823689366e-06,"output_power":5.895534218525675e-06,"reference_active":true,"delta_db":-1.0484927953906684}],"active_band_count":3,"active_band_mae_db":1.345444706278369}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.0001137540358892563,"log_rms_mae_db":1.5448760217124715,"rms_envelope_nrmse":0.18220083667762807,"envelope_correlation":0.9781536520167222,"rms_gain_error_db":0.5201680144495343,"reference_mean_power":0.0003949518827357507,"output_mean_power":0.0004452059807737642},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.625495481863056e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.926011413944637e-05,"output_power":8.971303255982593e-05,"reference_active":true,"delta_db":0.5380084392326635},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0135009616582808e-05,"output_power":1.9844500130612358e-05,"reference_active":true,"delta_db":-0.06311677564796125},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.142336744382214e-06,"output_power":5.208109490828249e-06,"reference_active":true,"delta_db":-1.3716021641213625}],"active_band_count":3,"active_band_mae_db":0.6575757930006624}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":0.978414414774115,"rms_envelope_nrmse":0.09854518864616198,"envelope_correlation":0.9951604030165421,"rms_gain_error_db":0.5202089300953783,"reference_mean_power":0.000399452081398279,"output_mean_power":0.0004502830316807082},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.3661376304062826e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.270190105592022e-05,"output_power":8.19290197387382e-05,"reference_active":true,"delta_db":0.5189199144208942},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.983801499648588e-06,"output_power":9.919809840762743e-06,"reference_active":true,"delta_db":-0.027925907137139033},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.054433356878422e-07,"output_power":3.382508719937794e-07,"reference_active":true,"delta_db":-1.7443354984897383}],"active_band_count":3,"active_band_mae_db":0.7637271066825906}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":1.5083338390767154,"rms_envelope_nrmse":0.20617298704244533,"envelope_correlation":0.9726004471098987,"rms_gain_error_db":0.2504264464226748,"reference_mean_power":0.00015129473773616958,"output_mean_power":0.00016027525166868636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":2.0948865520528097e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010579126587837111,"output_power":0.00010662450012860518,"reference_active":true,"delta_db":0.03407194295266698},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5885080610072743e-05,"output_power":1.5341087265171107e-05,"reference_active":true,"delta_db":-0.15133282678150095},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.603064633980169e-06,"output_power":3.102333889049257e-06,"reference_active":true,"delta_db":-2.567370952994816}],"active_band_count":3,"active_band_mae_db":0.917591907576328}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":0.8938744980176593,"rms_envelope_nrmse":0.09836879914050066,"envelope_correlation":0.9938881402593243,"rms_gain_error_db":0.25043892802210505,"reference_mean_power":0.00015301852862864023,"output_mean_power":0.0001621018287744399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.9708112174484797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.707695341873246e-05,"output_power":9.76952050965947e-05,"reference_active":true,"delta_db":0.027571106257150212},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.3675521469727354e-06,"output_power":7.103853895186008e-06,"reference_active":true,"delta_db":-0.15829197699192774},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.245758891603436e-07,"output_power":1.4991010207620628e-07,"reference_active":true,"delta_db":-3.354853556173572}],"active_band_count":3,"active_band_mae_db":1.1802388798075498}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"facodec","status":"measured","mos":{"status":"measured","value":3.664762496948242,"predicted_mos":3.664762496948242,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":42600,"input_sample_rate":16000,"input_seconds":2.6625,"input_float32_sha256":"16bfbdd944422934c2f9b72d46e09f3bdcb1a954a8a4465450b10ff77c8eb99c","internal_samples":42600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.857250938999641},"started_at":"2026-10-13T06:57:33.903383+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6726000214694068e-05,"native_input_samples":42560,"native_input_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","output_sha256":"dda48c7ebcc76a2c9581900f64754b922aa6fa5bd3c78ee00fc4e2ebc3546884","output_samples":42600,"length_delta_samples":40,"analysis_s":0.7398483179986215,"synthesis_s":5.387861340001109,"total_s":6.1277096579997306,"rtf":2.30365024736832,"output_peak":0.6739097833633423,"output_rms":0.06189691682101384,"finished_at":"2026-10-13T06:57:40.035786+00:00","evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"dda48c7ebcc76a2c9581900f64754b922aa6fa5bd3c78ee00fc4e2ebc3546884","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":42560,"output_samples":42600,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.166619666114233,"output_peak":0.6739097833633423},"common_prefix":{"samples":42560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990610328638497,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.4938033558625,"spectral_convergence":0.27017540504939924,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7736912522874935,"spectral_convergence":0.2863869093137305,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.86975255409677,"spectral_convergence":0.31348934862235744,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.0003918701760750715,"delta_db":-0.32929044460920653,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.001489947884400202,"delta_db":-0.6601706803711237,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0014620748559832691,"delta_db":-0.7532632205721773,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.000248082456486515,"delta_db":-1.2971707709073301,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.0001302894491919291,"delta_db":-1.2184918399926925,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":0.00011256421011736101,"delta_db":-0.5164081416853243,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7957991830229757,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7235025213364563,"pitch":null,"output_rms_dbfs":-24.162539933847896,"output_peak":0.6739097833633423,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.712415720748922},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42560,"output_samples":42560,"intelligibility":{"estoi":{"value":0.9083616775765567,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06730512410226613,"rms_dbfs":-23.43903741251144,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.24609129701517,"mean":6.90625126200511e-05,"zero_energy":false},"output":{"rms":0.0619259964171296,"rms_dbfs":-24.162539933847896,"peak":0.6739097833633423,"peak_dbfs":-3.427964775819468,"crest_factor_db":20.73457515802843,"mean":0.00018726338146947955,"zero_energy":false},"rms_difference_db":-0.7235025213364565},"full_scale":{"reference":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":266,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.5796519462661497e-06,"output_mean_power":1.904251357834511e-06,"output_rms_dbfs":-57.20275726058944,"output_zero_energy":false,"output_peak":0.008269786834716797,"output_energy_fraction":0.00010827411042106593,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":259,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":100,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9489795918367347,"absolute_difference_cents_median":3.94693249928367,"absolute_difference_cents_p95":17.6900479764734,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00020029577381909937,"log_rms_mae_db":1.265586831844314,"rms_envelope_nrmse":0.11204775086546727,"envelope_correlation":0.9933015939561608,"rms_gain_error_db":-0.7235045053918748,"reference_mean_power":0.004538506592765079,"output_mean_power":0.003842045644710428},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.8045928558777289e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008527253032911584,"output_power":0.0007602465255756633,"reference_active":true,"delta_db":-0.4985470614565233},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002893325022312121,"output_power":0.00022927478725273865,"reference_active":true,"delta_db":-1.01040923825874},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.722708282094668e-05,"output_power":7.960061030080201e-05,"reference_active":true,"delta_db":-0.868708579733287}],"active_band_count":3,"active_band_mae_db":0.7925549598161834}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00017887965336918488,"log_rms_mae_db":1.0178315560027924,"rms_envelope_nrmse":0.09910453944039527,"envelope_correlation":0.9958472156663195,"rms_gain_error_db":-0.7235129484145909,"reference_mean_power":0.004590346759257262,"output_mean_power":0.0038859230734357856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.4591224648550268e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007420820227306516,"output_power":0.0006575807003709806,"reference_active":true,"delta_db":-0.5250285232127863},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015779167985385838,"output_power":0.00012697262817928842,"reference_active":true,"delta_db":-0.9437399065350647},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.56670863086132e-06,"output_power":5.378003449996952e-06,"reference_active":true,"delta_db":-0.8672667029817703}],"active_band_count":3,"active_band_mae_db":0.7786783775765405}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00017771570730226456,"log_rms_mae_db":1.5805496547317974,"rms_envelope_nrmse":0.12249517133494804,"envelope_correlation":0.9921243952943172,"rms_gain_error_db":-0.7206293488981026,"reference_mean_power":0.003163634497093627,"output_mean_power":0.0026799294896593783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.820001870942524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009255585566014783,"output_power":0.0007979221413734731,"reference_active":true,"delta_db":-0.644433837467326},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002598152375676334,"output_power":0.00022125849836815057,"reference_active":true,"delta_db":-0.6976465725288896},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.760488032618479e-05,"output_power":6.646807876956809e-05,"reference_active":true,"delta_db":-0.6727590766178502}],"active_band_count":3,"active_band_mae_db":0.6716131622046886}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00016184521030355344,"log_rms_mae_db":1.302989173566221,"rms_envelope_nrmse":0.10676523205243266,"envelope_correlation":0.9949421169647656,"rms_gain_error_db":-0.7206295873152981,"reference_mean_power":0.0031997824540876,"output_mean_power":0.002710550443271679},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.5556561049407657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008585656786559776,"output_power":0.0007387667692187777,"reference_active":true,"delta_db":-0.6526617145358069},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001566221417610138,"output_power":0.00013787197735872165,"reference_active":true,"delta_db":-0.5537715423754377},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.505375823689366e-06,"output_power":6.166606528962427e-06,"reference_active":true,"delta_db":-0.8532620555996716}],"active_band_count":3,"active_band_mae_db":0.6865651041703055}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.0001137540358892563,"log_rms_mae_db":1.9229843158317437,"rms_envelope_nrmse":0.17984183445976365,"envelope_correlation":0.9854483355546346,"rms_gain_error_db":-1.1835806210566877,"reference_mean_power":0.0003949518827357507,"output_mean_power":0.00030073648980959774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.625495481863056e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.926011413944637e-05,"output_power":6.017886190818639e-05,"reference_active":true,"delta_db":-1.1961072322683695},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0135009616582808e-05,"output_power":1.4089470139825198e-05,"reference_active":true,"delta_db":-1.5505718038292968},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.142336744382214e-06,"output_power":6.372543005606792e-06,"reference_active":true,"delta_db":-0.49527547281445483}],"active_band_count":3,"active_band_mae_db":1.0806515029707071}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":1.5972066547802573,"rms_envelope_nrmse":0.16006267331342486,"envelope_correlation":0.9906939683176827,"rms_gain_error_db":-1.1836377308183568,"reference_mean_power":0.000399452081398279,"output_mean_power":0.00030415917067420033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.3661376304062826e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.270190105592022e-05,"output_power":5.501212082879846e-05,"reference_active":true,"delta_db":-1.2108737901271962},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.983801499648588e-06,"output_power":6.578448188312289e-06,"reference_active":true,"delta_db":-1.8117247912877135},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.054433356878422e-07,"output_power":4.6777425045437184e-07,"reference_active":true,"delta_db":-0.33636163164856014}],"active_band_count":3,"active_band_mae_db":1.11965340435449}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":1.7322161977906207,"rms_envelope_nrmse":0.20891825618563784,"envelope_correlation":0.9758647046950173,"rms_gain_error_db":-0.9922246618550462,"reference_mean_power":0.00015129473773616958,"output_mean_power":0.00012039303331053183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":2.0948865520528097e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010579126587837111,"output_power":7.258154748832964e-05,"reference_active":true,"delta_db":-1.6362359042924388},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5885080610072743e-05,"output_power":1.6440377343399645e-05,"reference_active":true,"delta_db":0.14922358317546258},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.603064633980169e-06,"output_power":4.401278065800503e-06,"reference_active":true,"delta_db":-1.0484682481760135}],"active_band_count":3,"active_band_mae_db":0.9446425785479716}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":1.365262493702425,"rms_envelope_nrmse":0.16724379885693716,"envelope_correlation":0.9873103120604233,"rms_gain_error_db":-0.992398163638933,"reference_mean_power":0.00015301852862864023,"output_mean_power":0.00012175987827705088},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.9708112174484797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.707695341873246e-05,"output_power":6.6815073696855e-05,"reference_active":true,"delta_db":-1.6224168657195694},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.3675521469727354e-06,"output_power":8.6403389265144e-06,"reference_active":true,"delta_db":0.6920756004141286},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.245758891603436e-07,"output_power":2.6396524140302754e-07,"reference_active":true,"delta_db":-0.8976951219123513}],"active_band_count":3,"active_band_mae_db":1.07072919601535}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.3876209259033203,"predicted_mos":3.3876209259033203,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":117760,"input_sample_rate":44100,"input_seconds":2.670294784580499,"input_float32_sha256":"37894994a14d1fdacafebbf3744edabb3563271ccc095f02d225788ea0d4ce61","internal_samples":42725,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":10.029765834999125},"started_at":"2026-10-13T07:25:01.680538+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004123480997805018,"native_input_samples":117306,"native_input_sha256":"53890309785cb73bb6aa80b43abfa55893e1f09630118a227c2d726b555fc1b4","output_sha256":"d3068d93bd104de253444cba208f4c0424c2f03cbb916887f761ce000ea4dd8a","output_samples":117760,"length_delta_samples":454,"analysis_s":0.06061330500233453,"synthesis_s":15.621634756000276,"total_s":15.68224806100261,"rtf":5.89558197782053,"output_peak":0.44069594144821167,"output_rms":0.06196581969200768,"finished_at":"2026-10-13T07:25:17.375883+00:00","evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"d3068d93bd104de253444cba208f4c0424c2f03cbb916887f761ce000ea4dd8a","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":42560,"output_samples":42725,"length_delta_samples":165,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.150930207551088,"output_peak":0.43579263760949616},"common_prefix":{"samples":42560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9961380924517261,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.57813667697102,"spectral_convergence":0.5158860430354224,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.516084829625274,"spectral_convergence":0.4075303454557724,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.412477498192706,"spectral_convergence":0.5154862178801509,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.00026490403013273945,"delta_db":-2.0298269098851613,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.0015735323934690756,"delta_db":-0.4231245797429262,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0015448199828651768,"delta_db":-0.5141805187687727,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.0002383700791804341,"delta_db":-1.4706138994022322,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.00014716472400037067,"delta_db":-0.6895471180042181,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":9.120974019164997e-05,"delta_db":-1.4299992369248724,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0928820437880304,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6950884720923057,"pitch":null,"output_rms_dbfs":-24.134125884603748,"output_peak":0.43579263760949616,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.502233001596332},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42560,"output_samples":42560,"intelligibility":{"estoi":{"value":0.7952296689425837,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06730512410226613,"rms_dbfs":-23.43903741251144,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.24609129701517,"mean":6.90625126200511e-05,"zero_energy":false},"output":{"rms":0.06212890591213921,"rms_dbfs":-24.134125884603748,"peak":0.43579263760949616,"peak_dbfs":-7.214402222161418,"crest_factor_db":16.91972366244233,"mean":-7.711850012267215e-05,"zero_energy":false},"rms_difference_db":-0.6950884720923085},"full_scale":{"reference":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":266,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.5796519462661497e-06,"output_mean_power":1.1665632901094447e-06,"output_rms_dbfs":-59.33091694247291,"output_zero_energy":false,"output_peak":0.005315072781313802,"output_energy_fraction":6.589724392937837e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":259,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":96,"both_estimated_voiced_frames":81,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.826530612244898,"absolute_difference_cents_median":11.694686071128126,"absolute_difference_cents_p95":44.67158385233583,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00020029577381909937,"log_rms_mae_db":1.4498057236270026,"rms_envelope_nrmse":0.13953215224880747,"envelope_correlation":0.9855891908141554,"rms_gain_error_db":-0.6950881983025557,"reference_mean_power":0.004538506592765079,"output_mean_power":0.0038672669428729267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.8045928558777289e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008527253032911584,"output_power":0.0007782176145871994,"reference_active":true,"delta_db":-0.39708093555832613},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002893325022312121,"output_power":0.0002559309710244199,"reference_active":true,"delta_db":-0.5327437852589161},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.722708282094668e-05,"output_power":6.67288897912127e-05,"reference_active":true,"delta_db":-1.63473356183865}],"active_band_count":3,"active_band_mae_db":0.8548527608852975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00017887965336918488,"log_rms_mae_db":1.168015244516422,"rms_envelope_nrmse":0.09130009290367995,"envelope_correlation":0.9972633360699689,"rms_gain_error_db":-0.6950855294184505,"reference_mean_power":0.004590346759257262,"output_mean_power":0.003911442414990799},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.4591224648550268e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007420820227306516,"output_power":0.000675939252542693,"reference_active":true,"delta_db":-0.40544243599597457},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015779167985385838,"output_power":0.00014088286067237703,"reference_active":true,"delta_db":-0.4922593816502601},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.56670863086132e-06,"output_power":5.076723351167087e-06,"reference_active":true,"delta_db":-1.1176424899800894}],"active_band_count":3,"active_band_mae_db":0.6717814358754414}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00017771570730226456,"log_rms_mae_db":1.9076469890262935,"rms_envelope_nrmse":0.13343297445640953,"envelope_correlation":0.9864693504320348,"rms_gain_error_db":-0.4302831300057233,"reference_mean_power":0.003163634497093627,"output_mean_power":0.002865220102694175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.820001870942524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009255585566014783,"output_power":0.0007919524964396464,"reference_active":true,"delta_db":-0.6770476797908037},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002598152375676334,"output_power":0.00025118596267599254,"reference_active":true,"delta_db":-0.14669252310915826},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.760488032618479e-05,"output_power":6.55007426314445e-05,"reference_active":true,"delta_db":-0.7364280958273217}],"active_band_count":3,"active_band_mae_db":0.5200560995757613}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00016184521030355344,"log_rms_mae_db":1.295190018349077,"rms_envelope_nrmse":0.09245806458553914,"envelope_correlation":0.9940281419064614,"rms_gain_error_db":-0.4302849062103432,"reference_mean_power":0.0031997824540876,"output_mean_power":0.0028979571661974014},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.5556561049407657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008585656786559776,"output_power":0.0007370706720125228,"reference_active":true,"delta_db":-0.6626439247869197},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001566221417610138,"output_power":0.00014989652927946105,"reference_active":true,"delta_db":-0.19061581256356583},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.505375823689366e-06,"output_power":5.727950396587611e-06,"reference_active":true,"delta_db":-1.173731955825109}],"active_band_count":3,"active_band_mae_db":0.6756638977251982}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.0001137540358892563,"log_rms_mae_db":2.8612523310320572,"rms_envelope_nrmse":0.3376331769313168,"envelope_correlation":0.9097133793667435,"rms_gain_error_db":-1.105349020808729,"reference_mean_power":0.0003949518827357507,"output_mean_power":0.0003062028907289019},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.625495481863056e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.926011413944637e-05,"output_power":8.92359452194956e-05,"reference_active":true,"delta_db":0.5148513480024567},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0135009616582808e-05,"output_power":1.6809532214544836e-05,"reference_active":true,"delta_db":-0.7839621355696446},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.142336744382214e-06,"output_power":5.755726619996785e-06,"reference_active":true,"delta_db":-0.937401642529381}],"active_band_count":3,"active_band_mae_db":0.7454050420338274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":2.425664278184445,"rms_envelope_nrmse":0.2520264571477364,"envelope_correlation":0.9508165975544327,"rms_gain_error_db":-1.1053337337298184,"reference_mean_power":0.000399452081398279,"output_mean_power":0.00030969294721676827},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.3661376304062826e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.270190105592022e-05,"output_power":8.153040688465238e-05,"reference_active":true,"delta_db":0.49773842496106163},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.983801499648588e-06,"output_power":8.265556440061384e-06,"reference_active":true,"delta_db":-0.8202384193309721},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.054433356878422e-07,"output_power":3.9291317745783044e-07,"reference_active":true,"delta_db":-1.0937588035616395}],"active_band_count":3,"active_band_mae_db":0.803911882617891}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":3.3699862397438256,"rms_envelope_nrmse":0.22027765251105658,"envelope_correlation":0.9739101499699822,"rms_gain_error_db":-1.0559745511997072,"reference_mean_power":0.00015129473773616958,"output_mean_power":0.00011863869686694899},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":2.0948865520528097e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010579126587837111,"output_power":8.793883320006904e-05,"reference_active":true,"delta_db":-0.8026911488149318},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5885080610072743e-05,"output_power":9.475843743688318e-06,"reference_active":true,"delta_db":-2.243715324198834},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.603064633980169e-06,"output_power":2.7147864710701834e-06,"reference_active":true,"delta_db":-3.146899560354988}],"active_band_count":3,"active_band_mae_db":2.064435344456251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":2.944010637197102,"rms_envelope_nrmse":0.1901716619834698,"envelope_correlation":0.9824917652087808,"rms_gain_error_db":-1.0559228697160492,"reference_mean_power":0.00015301852862864023,"output_mean_power":0.00011999184597826115},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.9708112174484797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.707695341873246e-05,"output_power":7.935732258043339e-05,"reference_active":true,"delta_db":-0.8752913151629133},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.3675521469727354e-06,"output_power":4.375135524553927e-06,"reference_active":true,"delta_db":-2.2633170812819925},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.245758891603436e-07,"output_power":1.8169569836860995e-07,"reference_active":true,"delta_db":-2.5197160995583374}],"active_band_count":3,"active_band_mae_db":1.8861081653344145}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"golf","status":"measured","mos":{"status":"measured","value":2.986516237258911,"predicted_mos":2.986516237258911,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":63600,"input_sample_rate":24000,"input_seconds":2.65,"input_float32_sha256":"49a21ba4f88c748b7c9b7ffade15d9d9d47106d6f2f5beed78ac5922a778eecb","internal_samples":42400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.34631126199929},"started_at":"2026-10-13T05:02:53.392160+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018869330006054952,"native_input_samples":63840,"native_input_sha256":"29f42c6325d732070e76922d87b4623407303841acd7da65542745abe9878306","output_sha256":"da77e62f7e8c1faf48ff764949ba7545db442af7d907df7e74e6822ce2f59c4c","output_samples":63600,"length_delta_samples":-240,"analysis_s":0.2448675830000866,"synthesis_s":0.06373262600027374,"total_s":0.30860020900036034,"rtf":0.11601511616554899,"output_peak":0.4474928379058838,"output_rms":0.06160972573001304,"finished_at":"2026-10-13T05:02:53.706926+00:00","analysis_stages":{"external_f0_s":0.021868126999834203},"evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"da77e62f7e8c1faf48ff764949ba7545db442af7d907df7e74e6822ce2f59c4c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":42560,"output_samples":42400,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.205480721061754,"output_peak":0.43535240376078443},"common_prefix":{"samples":42400,"sample_rate":16000,"reference_coverage":0.9962406015037594,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42400,"output_samples":42400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.793183779101128,"spectral_convergence":0.4456214683635866,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.797175893338929,"spectral_convergence":0.2761646104827472,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.853643868834227,"spectral_convergence":0.28942422230318643,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042433119840148726,"output_power":0.00037942853987198795,"delta_db":-0.48574969317498295,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017411030130216618,"output_power":0.0014476714747167988,"delta_db":-0.80154450038733,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017455472735384361,"output_power":0.0014614362363366577,"delta_db":-0.7715174354680676,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003356976985374756,"output_power":0.00026352872023124147,"delta_db":-1.0512041086346353,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017313880172661195,"output_power":0.000140363915544303,"delta_db":-0.9113893284608428,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001272546500776051,"output_power":0.00010466956364369537,"delta_db":-0.8485324743280749,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8116562567423222,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7828005547854144,"pitch":null,"output_rms_dbfs":-24.205480721061754,"output_peak":0.43535240376078443,"reference_rms_dbfs":-23.422680166276344,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.148001180424761},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42400,"output_samples":42400,"intelligibility":{"estoi":{"value":0.9257780339025876,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0674319923697457,"rms_dbfs":-23.422680166276344,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.22973405078007,"mean":6.946203843602595e-05,"zero_energy":false},"output":{"rms":0.061620605848355825,"rms_dbfs":-24.205480721061754,"peak":0.43535240376078443,"peak_dbfs":-7.223181065565829,"crest_factor_db":16.982299655495925,"mean":0.0004983417659714883,"zero_energy":false},"rms_difference_db":-0.7828005547854104},"full_scale":{"reference":{"sample_count":42400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":265,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":57,"quiet_samples":9120,"reference_mean_power":1.605511990196088e-06,"output_mean_power":8.403574039399916e-07,"output_rms_dbfs":-60.75535969239252,"output_zero_energy":false,"output_peak":0.005167224408595108,"output_energy_fraction":4.760374110540423e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":258,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":109,"both_estimated_voiced_frames":97,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9897959183673469,"absolute_difference_cents_median":4.45745589598879,"absolute_difference_cents_p95":14.803392573481778,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":529,"rms_floor":0.00020029577381909937,"log_rms_mae_db":2.2654891351468693,"rms_envelope_nrmse":0.15484424328429933,"envelope_correlation":0.9823045156145973,"rms_gain_error_db":-0.7827984041190648,"reference_mean_power":0.004555664835742011,"output_mean_power":0.003804275187952216},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":529,"frequency_resolution_hz":0.3780718336483932,"power_floor":1.805399659524261e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008555148962458395,"output_power":0.0007311092243266702,"reference_active":true,"delta_db":-0.6824531250300963},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00028377760730250904,"output_power":0.00026019977655828954,"reference_active":true,"delta_db":-0.37671203308626394},{"low_hz":20.0,"high_hz":50.0,"reference_power":0.00010486405714358021,"output_power":6.181002537220203e-05,"reference_active":true,"delta_db":-2.295677344984461}],"active_band_count":3,"active_band_mae_db":1.1182808343669404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":523,"rms_floor":0.00017887965336918488,"log_rms_mae_db":1.8366847080693867,"rms_envelope_nrmse":0.12089451926350396,"envelope_correlation":0.9917593558624179,"rms_gain_error_db":-0.7827871567256582,"reference_mean_power":0.004607899669665989,"output_mean_power":0.0038479046226175513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":523,"frequency_resolution_hz":0.3824091778202677,"power_floor":1.4576497238632447e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007482237369526647,"output_power":0.0006331334994017296,"reference_active":true,"delta_db":-0.7253618895275644},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015441568127822457,"output_power":0.00014919976328745002,"reference_active":true,"delta_db":-0.14923267764448025},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.567330734850084e-06,"output_power":3.922740961614529e-06,"reference_active":true,"delta_db":-2.237992577581987}],"active_band_count":3,"active_band_mae_db":1.0375290482513437}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":529,"rms_floor":0.00017771570730226456,"log_rms_mae_db":3.0689700260690995,"rms_envelope_nrmse":0.1760419974978462,"envelope_correlation":0.9795768937822591,"rms_gain_error_db":-0.8410255415816905,"reference_mean_power":0.003175595064984652,"output_mean_power":0.002616510995886068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":529,"frequency_resolution_hz":0.3780718336483932,"power_floor":1.8235234574835932e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009273368637224185,"output_power":0.0008132533465428788,"reference_active":true,"delta_db":-0.5701166522464617},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025491535046489775,"output_power":0.0002314575862082292,"reference_active":true,"delta_db":-0.41924568852223965},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.42451085250081e-05,"output_power":4.3774474371649985e-05,"reference_active":true,"delta_db":-2.843237541533737}],"active_band_count":3,"active_band_mae_db":1.2775332941008128}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":523,"rms_floor":0.00016184521030355344,"log_rms_mae_db":2.554862631069368,"rms_envelope_nrmse":0.14075115308246922,"envelope_correlation":0.9889193000827773,"rms_gain_error_db":-0.8410174923888696,"reference_mean_power":0.0032120182594939377,"output_mean_power":0.002646526556263403},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":523,"frequency_resolution_hz":0.3824091778202677,"power_floor":1.557567565967724e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008652200591362492,"output_power":0.000754621160495947,"reference_active":true,"delta_db":-0.5939760032232558},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015208528841434982,"output_power":0.00014389459867167537,"reference_active":true,"delta_db":-0.2404271344561014},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.490845151424151e-06,"output_power":2.9325016447449163e-06,"reference_active":true,"delta_db":-4.072925551850737}],"active_band_count":3,"active_band_mae_db":1.635776229843365}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":529,"rms_floor":0.0001137540358892563,"log_rms_mae_db":2.6244630689697233,"rms_envelope_nrmse":0.23539598537385065,"envelope_correlation":0.9629013562973331,"rms_gain_error_db":-1.0742353010115637,"reference_mean_power":0.0003964449001330799,"output_mean_power":0.0003095703123434064},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":529,"frequency_resolution_hz":0.3780718336483932,"power_floor":1.6244495474358405e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.923844342466735e-05,"output_power":6.384913808407939e-05,"reference_active":true,"delta_db":-0.9378089659916639},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9902052856072902e-05,"output_power":1.4252752643602624e-05,"reference_active":true,"delta_db":-1.4499912728098827},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.487269236019753e-06,"output_power":4.303671165822889e-06,"reference_active":true,"delta_db":-2.404843698529854}],"active_band_count":3,"active_band_mae_db":1.5975479791104668}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":523,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":2.1818225216105582,"rms_envelope_nrmse":0.14842426337851922,"envelope_correlation":0.9918811015632347,"rms_gain_error_db":-1.0741944387587792,"reference_mean_power":0.0004009793180296583,"output_mean_power":0.00031311403079249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":523,"frequency_resolution_hz":0.3824091778202677,"power_floor":1.3640477813277365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.268627237768137e-05,"output_power":5.8148101652506006e-05,"reference_active":true,"delta_db":-0.9691685620699088},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0002493739627306e-05,"output_power":6.4484639279158005e-06,"reference_active":true,"delta_db":-1.9065201345626202},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.085281256989185e-07,"output_power":3.105965289374497e-07,"reference_active":true,"delta_db":-2.1411837992968157}],"active_band_count":3,"active_band_mae_db":1.6722908319764482}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":529,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":3.4579340023017693,"rms_envelope_nrmse":0.17160245393255674,"envelope_correlation":0.9821652633342118,"rms_gain_error_db":-0.4818116201764914,"reference_mean_power":0.0001518666931819062,"output_mean_power":0.00013591937676905332},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":529,"frequency_resolution_hz":0.3780718336483932,"power_floor":2.1055673764881167e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010629561739015354,"output_power":9.36022747432522e-05,"reference_active":true,"delta_db":-0.5522895557398119},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5677508808890717e-05,"output_power":1.978129593511726e-05,"reference_active":true,"delta_db":1.0097768672595824},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8242922009987255e-06,"output_power":6.5978300785022226e-06,"reference_active":true,"delta_db":0.5415797134187293}],"active_band_count":3,"active_band_mae_db":0.7012153788060412}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42400,"uncovered_tail_samples":0,"metrics":{"frame_count":523,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":3.1593093447454126,"rms_envelope_nrmse":0.1369560569747009,"envelope_correlation":0.9890391687259322,"rms_gain_error_db":-0.48171126496526073,"reference_mean_power":0.00015360357785963318,"output_mean_power":0.00013747705026544025},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":523,"frequency_resolution_hz":0.3824091778202677,"power_floor":1.9810033123551193e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.750588952142242e-05,"output_power":8.716501067687905e-05,"reference_active":true,"delta_db":-0.4868866092414797},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.332882538846719e-06,"output_power":9.254281385173877e-06,"reference_active":true,"delta_db":1.0106797215668917},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.228991255361575e-07,"output_power":3.721881248479989e-07,"reference_active":true,"delta_db":0.6169564340073534}],"active_band_count":3,"active_band_mae_db":0.7048409216052415}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4123058319091797,"predicted_mos":3.4123058319091797,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":63960,"input_sample_rate":24000,"input_seconds":2.665,"input_float32_sha256":"2d175c9a3555dfa26619e1a57d66cb1776b618b7d0c9129d5fced076412cd2eb","internal_samples":42640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.413658551999106},"started_at":"2026-10-13T07:43:13.011200+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001754999000695534,"native_input_samples":63840,"native_input_sha256":"29f42c6325d732070e76922d87b4623407303841acd7da65542745abe9878306","output_sha256":"844fafa311e771e829c55859485452b77ad94eda84a8656b0f56647c4983b69b","output_samples":63960,"length_delta_samples":120,"analysis_s":0.9868990720024158,"synthesis_s":5.0149485069996445,"total_s":6.00184757900206,"rtf":2.256333676316564,"output_peak":0.5937476754188538,"output_rms":0.06555763459524197,"finished_at":"2026-10-13T07:43:19.020254+00:00","analysis_stages":{"external_f0_s":0.6619824119989062},"evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"844fafa311e771e829c55859485452b77ad94eda84a8656b0f56647c4983b69b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":42560,"output_samples":42640,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.665196894316445,"output_peak":0.5960696395523061},"common_prefix":{"samples":42560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.99812382739212,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.059293528514547,"spectral_convergence":0.5671270701553267,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.3265291522196385,"spectral_convergence":0.3249018787358606,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.295483573903972,"spectral_convergence":0.28517961615968945,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.0004557444778830989,"delta_db":0.32650158644901783,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.0017003125741618234,"delta_db":-0.08659379375142469,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.001589778021145139,"delta_db":-0.389594419370649,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.00030216535163812754,"delta_db":-0.44068467020483437,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.000165811840016846,"delta_db":-0.17142893372177292,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":9.438787296358613e-05,"delta_db":-1.2812494299794093,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.44934213891285135,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21800388466699167,"pitch":null,"output_rms_dbfs":-23.65704129717843,"output_peak":0.5960696395523061,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.56043541821272},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42560,"output_samples":42560,"intelligibility":{"estoi":{"value":0.9295427191780946,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06730512410226613,"rms_dbfs":-23.43903741251144,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.24609129701517,"mean":6.90625126200511e-05,"zero_energy":false},"output":{"rms":0.06563688092687389,"rms_dbfs":-23.65704129717843,"peak":0.5960696395523061,"peak_dbfs":-4.4940599626954825,"crest_factor_db":19.16298133448295,"mean":-2.453464584956271e-05,"zero_energy":false},"rms_difference_db":-0.2180038846669916},"full_scale":{"reference":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":266,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.5796519462661497e-06,"output_mean_power":1.4221051136913385e-06,"output_rms_dbfs":-58.47068301912175,"output_zero_energy":false,"output_peak":0.017392206448374057,"output_energy_fraction":7.197508471840571e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":259,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":103,"both_estimated_voiced_frames":91,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9285714285714286,"absolute_difference_cents_median":4.474680617669107,"absolute_difference_cents_p95":20.305653136242412,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00020029577381909937,"log_rms_mae_db":1.9900917142365089,"rms_envelope_nrmse":0.21868937648896083,"envelope_correlation":0.9518070186210598,"rms_gain_error_db":-0.21800468929812422,"reference_mean_power":0.004538506592765079,"output_mean_power":0.004316308740738875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.8045928558777289e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008527253032911584,"output_power":0.0008182430226790342,"reference_active":true,"delta_db":-0.17926839504332479},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002893325022312121,"output_power":0.00026915253037957944,"reference_active":true,"delta_db":-0.3139875585897549},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.722708282094668e-05,"output_power":9.379614274474873e-05,"reference_active":true,"delta_db":-0.15602276575280133}],"active_band_count":3,"active_band_mae_db":0.2164262397952937}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00017887965336918488,"log_rms_mae_db":1.4083198469842366,"rms_envelope_nrmse":0.10874906827321827,"envelope_correlation":0.9872581913789915,"rms_gain_error_db":-0.2179970719829016,"reference_mean_power":0.004590346759257262,"output_mean_power":0.004365618554633036},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.4591224648550268e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007420820227306516,"output_power":0.0007232787938535515,"reference_active":true,"delta_db":-0.11146178735611677},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015779167985385838,"output_power":0.00014027162516638472,"reference_active":true,"delta_db":-0.5111427101391673},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.56670863086132e-06,"output_power":7.008262426233612e-06,"reference_active":true,"delta_db":0.2826260894975545}],"active_band_count":3,"active_band_mae_db":0.30174352899761286}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00017771570730226456,"log_rms_mae_db":2.4722961819893823,"rms_envelope_nrmse":0.22059438384059551,"envelope_correlation":0.9588593463490499,"rms_gain_error_db":-0.24451211340209744,"reference_mean_power":0.003163634497093627,"output_mean_power":0.002990439986478594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.820001870942524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009255585566014783,"output_power":0.000883631805696554,"reference_active":true,"delta_db":-0.2013256053231321},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002598152375676334,"output_power":0.00024315023434415392,"reference_active":true,"delta_db":-0.2878992538876329},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.760488032618479e-05,"output_power":7.553792322671782e-05,"reference_active":true,"delta_db":-0.11723992982551279}],"active_band_count":3,"active_band_mae_db":0.20215492967875925}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00016184521030355344,"log_rms_mae_db":1.8698162862089664,"rms_envelope_nrmse":0.12025304281187535,"envelope_correlation":0.9870558730033906,"rms_gain_error_db":-0.24450757886355456,"reference_mean_power":0.0031997824540876,"output_mean_power":0.003024612166299725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.5556561049407657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008585656786559776,"output_power":0.0008226571348310021,"reference_active":true,"delta_db":-0.18554654821958752},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001566221417610138,"output_power":0.00014993213451846549,"reference_active":true,"delta_db":-0.18958434621186426},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.505375823689366e-06,"output_power":7.036274799464319e-06,"reference_active":true,"delta_db":-0.28029651607877304}],"active_band_count":3,"active_band_mae_db":0.21847580350340826}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.0001137540358892563,"log_rms_mae_db":2.175818397388884,"rms_envelope_nrmse":0.33897333990589557,"envelope_correlation":0.9063229743175009,"rms_gain_error_db":-0.31975012691959986,"reference_mean_power":0.0003949518827357507,"output_mean_power":0.0003669181336323555},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.625495481863056e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.926011413944637e-05,"output_power":7.743745600915678e-05,"reference_active":true,"delta_db":-0.10103616404197299},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0135009616582808e-05,"output_power":1.8936123728382472e-05,"reference_active":true,"delta_db":-0.2666075876004036},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.142336744382214e-06,"output_power":5.748099716038269e-06,"reference_active":true,"delta_db":-0.9431602878561596}],"active_band_count":3,"active_band_mae_db":0.43693467983284534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":1.4790793925183903,"rms_envelope_nrmse":0.15774487838064558,"envelope_correlation":0.9786329430961402,"rms_gain_error_db":-0.31970271186528204,"reference_mean_power":0.000399452081398279,"output_mean_power":0.0003711029590352427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.3661376304062826e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.270190105592022e-05,"output_power":7.04036542094723e-05,"reference_active":true,"delta_db":-0.1395056600460337},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.983801499648588e-06,"output_power":9.140476835084126e-06,"reference_active":true,"delta_db":-0.383270853294885},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.054433356878422e-07,"output_power":3.5702082516566004e-07,"reference_active":true,"delta_db":-1.509789252898538}],"active_band_count":3,"active_band_mae_db":0.6775219220798189}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":2.0545879717561175,"rms_envelope_nrmse":0.20472657331013075,"envelope_correlation":0.9742545318364876,"rms_gain_error_db":-0.7582848875378958,"reference_mean_power":0.00015129473773616958,"output_mean_power":0.00012705604547062066},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":2.0948865520528097e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010579126587837111,"output_power":8.412896625637522e-05,"reference_active":true,"delta_db":-0.9950426133971241},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5885080610072743e-05,"output_power":1.2586316152432297e-05,"reference_active":true,"delta_db":-1.010907865501029},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.603064633980169e-06,"output_power":2.7443977730964718e-06,"reference_active":true,"delta_db":-3.0997857403485507}],"active_band_count":3,"active_band_mae_db":1.7019120730822346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":1.3697564499457193,"rms_envelope_nrmse":0.1326533471373362,"envelope_correlation":0.9918564476392856,"rms_gain_error_db":-0.7582904387922156,"reference_mean_power":0.00015301852862864023,"output_mean_power":0.00012850350627577963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.9708112174484797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.707695341873246e-05,"output_power":7.626596865055754e-05,"reference_active":true,"delta_db":-1.0478534775607784},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.3675521469727354e-06,"output_power":5.916572778793548e-06,"reference_active":true,"delta_db":-0.9525300736611912},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.245758891603436e-07,"output_power":1.640621553038486e-07,"reference_active":true,"delta_db":-2.963078428676591}],"active_band_count":3,"active_band_mae_db":1.6544873266328537}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.112987518310547,"predicted_mos":3.112987518310547,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64080,"input_sample_rate":24000,"input_seconds":2.67,"input_float32_sha256":"7fcd7f6d4aa916f3f21a3eec208e6e67a5acfc690973531046c862315fcdb073","internal_samples":42720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.041311188999316},"started_at":"2026-10-13T05:11:13.056719+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008647201000712812,"native_input_samples":63840,"native_input_sha256":"29f42c6325d732070e76922d87b4623407303841acd7da65542745abe9878306","output_sha256":"f39d0d77c702371270a29d15d317d2858c5bf15c9de16ea119ff8bdcb8e5b2de","output_samples":64080,"length_delta_samples":240,"analysis_s":0.35007330599910347,"synthesis_s":0.07510868900135392,"total_s":0.4251819950004574,"rtf":0.15984285526332984,"output_peak":0.5061424970626831,"output_rms":0.06584938968613493,"finished_at":"2026-10-13T05:11:13.497633+00:00","analysis_stages":{"external_f0_s":0.0491593679998914},"evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"f39d0d77c702371270a29d15d317d2858c5bf15c9de16ea119ff8bdcb8e5b2de","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":42560,"output_samples":42720,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.6266840657768,"output_peak":0.5068519024199409},"common_prefix":{"samples":42560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9962546816479401,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.818249203549858,"spectral_convergence":0.21758447017365626,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.455439948577354,"spectral_convergence":0.1968450724466566,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.969895280045133,"spectral_convergence":0.21929580967487358,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.00040232685388218056,"delta_db":-0.214922353541338,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.0017260489403508562,"delta_db":-0.021350403743388292,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0016620014628549733,"delta_db":-0.19664528753460792,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.0003185418896324275,"delta_db":-0.21146578655305415,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.00014894110454295715,"delta_db":-0.6374386183899712,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":9.686958034938682e-05,"delta_db":-1.1685372402037713,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.40839328166102185,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1713504229365826,"pitch":null,"output_rms_dbfs":-23.610387835448027,"output_peak":0.5068519024199409,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.7478614773907815},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42560,"output_samples":42560,"intelligibility":{"estoi":{"value":0.9591040275359216,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06730512410226613,"rms_dbfs":-23.43903741251144,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.24609129701517,"mean":6.90625126200511e-05,"zero_energy":false},"output":{"rms":0.06599037681065915,"rms_dbfs":-23.610387835448027,"peak":0.5068519024199409,"peak_dbfs":-5.902378381674645,"crest_factor_db":17.70800945377338,"mean":0.00013129508663321015,"zero_energy":false},"rms_difference_db":-0.17135042293658742},"full_scale":{"reference":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":266,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.5796519462661497e-06,"output_mean_power":1.478780043702735e-06,"output_rms_dbfs":-58.30096418910577,"output_zero_energy":false,"output_peak":0.005851886541577807,"output_energy_fraction":7.404380383557584e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":259,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":102,"both_estimated_voiced_frames":94,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9591836734693877,"absolute_difference_cents_median":0.5753640923395398,"absolute_difference_cents_p95":3.492142028857532,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00020029577381909937,"log_rms_mae_db":0.7868438215753625,"rms_envelope_nrmse":0.07717680938618462,"envelope_correlation":0.9943315286617869,"rms_gain_error_db":-0.1713502799704999,"reference_mean_power":0.004538506592765079,"output_mean_power":0.004362926961542912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.8045928558777289e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008527253032911584,"output_power":0.0008413052740909332,"reference_active":true,"delta_db":-0.05855538695259063},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002893325022312121,"output_power":0.0002555699864414069,"reference_active":true,"delta_db":-0.5388737303233069},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.722708282094668e-05,"output_power":6.208967838041205e-05,"reference_active":true,"delta_db":-1.9476784523451274}],"active_band_count":3,"active_band_mae_db":0.8483691898736749}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00017887965336918488,"log_rms_mae_db":0.530728837481929,"rms_envelope_nrmse":0.0476592065514792,"envelope_correlation":0.9979028339084636,"rms_gain_error_db":-0.17134677052934727,"reference_mean_power":0.004590346759257262,"output_mean_power":0.004412765171345128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.4591224648550268e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007420820227306516,"output_power":0.0007348941701820933,"reference_active":true,"delta_db":-0.04227108545196111},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015779167985385838,"output_power":0.000140007142936773,"reference_active":true,"delta_db":-0.5193390646295747},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.56670863086132e-06,"output_power":3.9311667252557566e-06,"reference_active":true,"delta_db":-2.2282628365205523}],"active_band_count":3,"active_band_mae_db":0.929957662200696}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00017771570730226456,"log_rms_mae_db":1.057416973221493,"rms_envelope_nrmse":0.07761230961426709,"envelope_correlation":0.995123331293258,"rms_gain_error_db":-0.14715690053312092,"reference_mean_power":0.003163634497093627,"output_mean_power":0.0030582332973764672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.820001870942524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009255585566014783,"output_power":0.0009043045691584426,"reference_active":true,"delta_db":-0.10089174936886254},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002598152375676334,"output_power":0.00022942247143988917,"reference_active":true,"delta_db":-0.5402866399562732},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.760488032618479e-05,"output_power":4.568581583479768e-05,"reference_active":true,"delta_db":-2.3010764879334946}],"active_band_count":3,"active_band_mae_db":0.9807516257528768}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00016184521030355344,"log_rms_mae_db":0.6100258987438543,"rms_envelope_nrmse":0.043589307783394295,"envelope_correlation":0.9984875616462151,"rms_gain_error_db":-0.14715470533038127,"reference_mean_power":0.0031997824540876,"output_mean_power":0.0030931784947829883},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.5556561049407657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008585656786559776,"output_power":0.0008287654761336225,"reference_active":true,"delta_db":-0.15341872095264047},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001566221417610138,"output_power":0.00013763244982865154,"reference_active":true,"delta_db":-0.5613231822416778},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.505375823689366e-06,"output_power":3.8038271635100416e-06,"reference_active":true,"delta_db":-2.951516685092442}],"active_band_count":3,"active_band_mae_db":1.2220861960955869}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.0001137540358892563,"log_rms_mae_db":1.0586143542766113,"rms_envelope_nrmse":0.12921198962228475,"envelope_correlation":0.9870598210045929,"rms_gain_error_db":-0.2095257526075919,"reference_mean_power":0.0003949518827357507,"output_mean_power":0.000376349733802586},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.625495481863056e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.926011413944637e-05,"output_power":7.458087707983691e-05,"reference_active":true,"delta_db":-0.2642720700020601},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0135009616582808e-05,"output_power":1.7502920557330484e-05,"reference_active":true,"delta_db":-0.6084131975816929},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.142336744382214e-06,"output_power":5.453256440084485e-06,"reference_active":true,"delta_db":-1.1718440138223072}],"active_band_count":3,"active_band_mae_db":0.6815097604686867}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":0.7106156969474331,"rms_envelope_nrmse":0.08006852788044955,"envelope_correlation":0.9948860357823878,"rms_gain_error_db":-0.20950621309514542,"reference_mean_power":0.000399452081398279,"output_mean_power":0.0003806396866207136},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.3661376304062826e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.270190105592022e-05,"output_power":6.844323198233969e-05,"reference_active":true,"delta_db":-0.2621525788970453},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.983801499648588e-06,"output_power":8.508046487304232e-06,"reference_active":true,"delta_db":-0.694660834880199},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.054433356878422e-07,"output_power":4.179898037837783e-07,"reference_active":true,"delta_db":-0.8250678674890367}],"active_band_count":3,"active_band_mae_db":0.5939604270887604}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":1.0690475139476137,"rms_envelope_nrmse":0.15240264323889632,"envelope_correlation":0.9907684946176322,"rms_gain_error_db":-0.9335871512638045,"reference_mean_power":0.00015129473773616958,"output_mean_power":0.00012202957752863473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":2.0948865520528097e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010579126587837111,"output_power":7.866564821128415e-05,"reference_active":true,"delta_db":-1.286646882134626},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5885080610072743e-05,"output_power":1.2220245550236516e-05,"reference_active":true,"delta_db":-1.139094904422713},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.603064633980169e-06,"output_power":2.115069914280075e-06,"reference_active":true,"delta_db":-4.231009046024554}],"active_band_count":3,"active_band_mae_db":2.2189169441939645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":0.6905631401293827,"rms_envelope_nrmse":0.12357973244776091,"envelope_correlation":0.9965402708599831,"rms_gain_error_db":-0.9336118609996931,"reference_mean_power":0.00015301852862864023,"output_mean_power":0.00012341923084120405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.9708112174484797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.707695341873246e-05,"output_power":7.132549726245825e-05,"reference_active":true,"delta_db":-1.338713302162006},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.3675521469727354e-06,"output_power":5.76981212509048e-06,"reference_active":true,"delta_db":-1.0616154641812343},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.245758891603436e-07,"output_power":1.54772425186486e-07,"reference_active":true,"delta_db":-3.216226677303587}],"active_band_count":3,"active_band_mae_db":1.872185147882276}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6829-68769-0041","method":"magphase","status":"measured","mos":{"status":"measured","value":3.333143949508667,"predicted_mos":3.333143949508667,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":127852,"input_sample_rate":48000,"input_seconds":2.6635833333333334,"input_float32_sha256":"33e553ada6a93c065e41dfa74b02870a504742ae6fabde8ed0abcab9dc943e11","internal_samples":42618,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.116474767000909},"started_at":"2026-10-13T05:17:25.372109+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006272451999393525,"native_input_samples":127680,"native_input_sha256":"c2a92689d83cdb6a98be5f256752c950b4166a194fa404000eaf1d9f41331b77","output_sha256":"b94e9f58591b776d414ac48479a10e78e19214a9b3c27a3a3dfcd7280b2f3447","output_samples":127852,"length_delta_samples":172,"analysis_s":1.6472414700001536,"synthesis_s":0.7447905030003312,"total_s":2.392031973000485,"rtf":0.8992601402257462,"output_peak":0.6254887132491384,"output_rms":0.06508348914920245,"finished_at":"2026-10-13T05:17:27.777219+00:00","evaluation_input":{"reference":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"output_sha256":"b94e9f58591b776d414ac48479a10e78e19214a9b3c27a3a3dfcd7280b2f3447","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":42560,"output_samples":42618,"length_delta_samples":58,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.7286604535811,"output_peak":0.6208305729559328},"common_prefix":{"samples":42560,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986390726922897,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":42560,"output_samples":42560,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.543336247331818,"spectral_convergence":0.5559485791485298,"spectral_convergence_status":"ok","amplitude_floor":1.6028753096638316e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.226585031112567,"spectral_convergence":0.2258720147711597,"spectral_convergence_status":"ok","amplitude_floor":1.0331614165752405e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.555949117114188,"spectral_convergence":0.2488056759288582,"spectral_convergence_status":"ok","amplitude_floor":7.831529815953997e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00042273796895084704,"output_power":0.0004107156867466707,"delta_db":-0.1252996602191833,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0017345552824327432,"output_power":0.0016192363981050453,"delta_db":-0.29877888318910345,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017389853785610774,"output_power":0.0015627803986140742,"delta_db":-0.4639797506836962,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00033443608489121105,"output_power":0.00032227190097960966,"delta_db":-0.16090689832335328,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00017248781948922851,"output_power":0.00019598355686322106,"delta_db":0.55461203246374,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00012677719609633915,"output_power":0.00013252370872850948,"delta_db":0.19252438937569008,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.29935026904246104,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.28370859932685827,"pitch":null,"output_rms_dbfs":-23.722746011838296,"output_peak":0.6208305729559328,"reference_rms_dbfs":-23.43903741251144,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.775290131852858},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":42560,"output_samples":42560,"intelligibility":{"estoi":{"value":0.9558897555616984,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06730512410226613,"rms_dbfs":-23.43903741251144,"peak":0.617095947265625,"peak_dbfs":-4.1929461154962695,"crest_factor_db":19.24609129701517,"mean":6.90625126200511e-05,"zero_energy":false},"output":{"rms":0.06514224167187624,"rms_dbfs":-23.722746011838296,"peak":0.6208305729559328,"peak_dbfs":-4.140538085670016,"crest_factor_db":19.582207926168284,"mean":-0.0005775955239273194,"zero_energy":false},"rms_difference_db":-0.2837085993268573},"full_scale":{"reference":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":42560,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":266,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.011839700979181e-06,"threshold_dbfs":-53.96656428186765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":58,"quiet_samples":9280,"reference_mean_power":1.5796519462661497e-06,"output_mean_power":1.437055248593965e-06,"output_rms_dbfs":-58.42526534789748,"output_zero_energy":false,"output_peak":0.006796034147667162,"output_energy_fraction":7.384046506643375e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":259,"reference_estimated_voiced_frames":98,"output_estimated_voiced_frames":103,"both_estimated_voiced_frames":93,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9489795918367347,"absolute_difference_cents_median":2.3155855486717423,"absolute_difference_cents_p95":10.879624463754917,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":42560,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00020029577381909937,"log_rms_mae_db":1.1686409047463435,"rms_envelope_nrmse":0.1072219095462298,"envelope_correlation":0.9893040721693976,"rms_gain_error_db":-0.28370606660650455,"reference_mean_power":0.004538506592765079,"output_mean_power":0.004251501767751156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.8045928558777289e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008527253032911584,"output_power":0.000821948967812305,"reference_active":true,"delta_db":-0.15964295765770423},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002893325022312121,"output_power":0.00027465695579809624,"reference_active":true,"delta_db":-0.22606620776530495},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.722708282094668e-05,"output_power":9.135540408711481e-05,"reference_active":true,"delta_db":-0.2705301253576316}],"active_band_count":3,"active_band_mae_db":0.21874643026021356}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00017887965336918488,"log_rms_mae_db":0.7867064994672028,"rms_envelope_nrmse":0.06857902159953462,"envelope_correlation":0.99586022966426,"rms_gain_error_db":-0.28371622643458744,"reference_mean_power":0.004590346759257262,"output_mean_power":0.004300053620627795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.4591224648550268e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007420820227306516,"output_power":0.0007232010816797917,"reference_active":true,"delta_db":-0.11192843703225802},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015779167985385838,"output_power":0.00013884925154100187,"reference_active":true,"delta_db":-0.555405566518338},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.56670863086132e-06,"output_power":6.529255219509436e-06,"reference_active":true,"delta_db":-0.02484101882295666}],"active_band_count":3,"active_band_mae_db":0.23072500745785085}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.00017771570730226456,"log_rms_mae_db":1.6529166854665713,"rms_envelope_nrmse":0.10268758694707789,"envelope_correlation":0.9924301715844337,"rms_gain_error_db":-0.36693129978861955,"reference_mean_power":0.003163634497093627,"output_mean_power":0.00290732225387244},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.820001870942524e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009255585566014783,"output_power":0.0008792805587854404,"reference_active":true,"delta_db":-0.22276429278636323},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002598152375676334,"output_power":0.00024201263892978438,"reference_active":true,"delta_db":-0.3082657058047393},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.760488032618479e-05,"output_power":6.350628992851102e-05,"reference_active":true,"delta_db":-0.8707229176354074}],"active_band_count":3,"active_band_mae_db":0.4672509720755033}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":0.00016184521030355344,"log_rms_mae_db":1.1867987816308951,"rms_envelope_nrmse":0.07228526997551747,"envelope_correlation":0.9967644920577833,"rms_gain_error_db":-0.36693523446718146,"reference_mean_power":0.0031997824540876,"output_mean_power":0.0029405389012532846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.5556561049407657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008585656786559776,"output_power":0.000820976079528555,"reference_active":true,"delta_db":-0.19443020048232074},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001566221417610138,"output_power":0.00014272025820634722,"reference_active":true,"delta_db":-0.4036753577062136},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.505375823689366e-06,"output_power":6.269322469728334e-06,"reference_active":true,"delta_db":-0.7815183511288604}],"active_band_count":3,"active_band_mae_db":0.45987463643913157}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":0.0001137540358892563,"log_rms_mae_db":1.3357309938908255,"rms_envelope_nrmse":0.17334542759298877,"envelope_correlation":0.9762013650187955,"rms_gain_error_db":-0.03010947244134527,"reference_mean_power":0.0003949518827357507,"output_mean_power":0.0003922231662202138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":1.625495481863056e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.926011413944637e-05,"output_power":8.575759346339258e-05,"reference_active":true,"delta_db":0.3421789190296528},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0135009616582808e-05,"output_power":2.1359131134230085e-05,"reference_active":true,"delta_db":0.2563174073314402},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.142336744382214e-06,"output_power":4.62469238752136e-06,"reference_active":true,"delta_db":-1.8875747146474755}],"active_band_count":3,"active_band_mae_db":0.8286903470028562}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":8.700603836695893e-05,"log_rms_mae_db":0.8634135102427272,"rms_envelope_nrmse":0.11009517810722654,"envelope_correlation":0.9894039700668699,"rms_gain_error_db":-0.030164392238852276,"reference_mean_power":0.000399452081398279,"output_mean_power":0.0003966872566391567},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.3661376304062826e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.270190105592022e-05,"output_power":7.907490152664936e-05,"reference_active":true,"delta_db":0.3649289253555553},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.983801499648588e-06,"output_power":9.36779822439871e-06,"reference_active":true,"delta_db":-0.2765840990058064},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.054433356878422e-07,"output_power":3.2259563005179836e-07,"reference_active":true,"delta_db":-1.950139946609999}],"active_band_count":3,"active_band_mae_db":0.8638843236571202}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":531,"rms_floor":6.751108516993362e-05,"log_rms_mae_db":1.4260868602411116,"rms_envelope_nrmse":0.21490218924630172,"envelope_correlation":0.9779614206966323,"rms_gain_error_db":0.8353481209191526,"reference_mean_power":0.00015129473773616958,"output_mean_power":0.00018338281496095056},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":531,"frequency_resolution_hz":0.3766478342749529,"power_floor":2.0948865520528097e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010579126587837111,"output_power":0.0001245975484294801,"reference_active":true,"delta_db":0.7105968341833417},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5885080610072743e-05,"output_power":2.0401951961607714e-05,"reference_active":true,"delta_db":1.086822976118487},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.603064633980169e-06,"output_power":1.2803097823578757e-05,"reference_active":true,"delta_db":3.588894314944203}],"active_band_count":3,"active_band_mae_db":1.7954380417486773}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":42560,"uncovered_tail_samples":0,"metrics":{"frame_count":525,"rms_floor":5.7873331101151596e-05,"log_rms_mae_db":1.1284919913824998,"rms_envelope_nrmse":0.14081888965018363,"envelope_correlation":0.9937244094375272,"rms_gain_error_db":0.8353456356764615,"reference_mean_power":0.00015301852862864023,"output_mean_power":0.00018547209825763887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":525,"frequency_resolution_hz":0.38095238095238093,"power_floor":1.9708112174484797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.707695341873246e-05,"output_power":0.00011821384214422304,"reference_active":true,"delta_db":0.8555219450523839},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.3675521469727354e-06,"output_power":9.741428739710472e-06,"reference_active":true,"delta_db":1.2129943949666295},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.245758891603436e-07,"output_power":7.564592628106724e-07,"reference_active":true,"delta_db":3.674692896665342}],"active_band_count":3,"active_band_mae_db":1.9144030788947852}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},{"case":"librispeech_6930-75918-0013","method":"hnm","status":"measured","mos":{"status":"measured","value":1.9538862705230713,"predicted_mos":1.9538862705230713,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":47040,"input_sample_rate":16000,"input_seconds":2.94,"input_float32_sha256":"0fdd31b8739c060275289901ae3e62b5889ccab808310d0e6da9a0c8c5b83aec","internal_samples":47040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4404944039997645},"started_at":"2026-10-13T04:39:55.383851+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.312099993810989e-05,"native_input_samples":47040,"native_input_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","output_sha256":"945319dcef24628a59402740221049aae28b841b4c6279a36c152e862b3f56ad","output_samples":47040,"length_delta_samples":0,"analysis_s":0.4503349159995196,"synthesis_s":0.2062982440002088,"total_s":0.6566331599997284,"rtf":0.22334461224480556,"output_peak":0.27726386590273416,"output_rms":0.030844344223620562,"finished_at":"2026-10-13T04:39:56.043931+00:00","evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"945319dcef24628a59402740221049aae28b841b4c6279a36c152e862b3f56ad","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.177519427120054,"spectral_convergence":0.5697612034518823,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.957742158967163,"spectral_convergence":0.5698050907810357,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.438994076182746,"spectral_convergence":0.5869418070100197,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":4.8298057741743025e-05,"delta_db":0.08465565890387342,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00014052602462013224,"delta_db":-0.5561577648604061,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":4.089092231496508e-05,"delta_db":0.1582480982569567,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":6.44122715968702e-05,"delta_db":-0.06198749944900703,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.00047585475517637006,"delta_db":-0.37811962664781906,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.0001813915391351143,"delta_db":0.15896908842178203,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23302295608997406,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2419110092738435,"pitch":null,"output_rms_dbfs":-30.2164891756657,"output_peak":0.27726386590273416,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.191418554089988},"common_prefix":{"samples":47040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.177519427120054,"spectral_convergence":0.5697612034518823,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.957742158967163,"spectral_convergence":0.5698050907810357,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.438994076182746,"spectral_convergence":0.5869418070100197,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":4.8298057741743025e-05,"delta_db":0.08465565890387342,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00014052602462013224,"delta_db":-0.5561577648604061,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":4.089092231496508e-05,"delta_db":0.1582480982569567,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":6.44122715968702e-05,"delta_db":-0.06198749944900703,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.00047585475517637006,"delta_db":-0.37811962664781906,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.0001813915391351143,"delta_db":0.15896908842178203,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23302295608997406,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2419110092738435,"pitch":null,"output_rms_dbfs":-30.2164891756657,"output_peak":0.27726386590273416,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.191418554089988},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":47040,"output_samples":47040,"intelligibility":{"estoi":{"value":0.7826110077986957,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03171546561641813,"rms_dbfs":-29.974578166391865,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.207027209586595,"mean":-2.270977513319781e-05,"zero_energy":false},"output":{"rms":0.030844344223620562,"rms_dbfs":-30.2164891756657,"peak":0.27726386590273416,"peak_dbfs":-11.142134513353144,"crest_factor_db":19.07435466231256,"mean":0.00010021793924959072,"zero_energy":false},"rms_difference_db":-0.24191100927383502},"full_scale":{"reference":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":294,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":3.2013646916459974e-07,"output_rms_dbfs":-64.9466484927112,"output_zero_energy":false,"output_peak":0.0020653805724725398,"output_energy_fraction":1.373466224391973e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":287,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":32,"both_estimated_voiced_frames":31,"reference_voiced_output_unvoiced_frames":43,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.4189189189189189,"absolute_difference_cents_median":1.999063969767814,"absolute_difference_cents_p95":7.225182042166667,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":47040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.0001280716783512989,"log_rms_mae_db":1.2277428490452076,"rms_envelope_nrmse":0.14485479226386802,"envelope_correlation":0.9777123522089809,"rms_gain_error_db":-0.2418565094866253,"reference_mean_power":0.0010075555569028208,"output_mean_power":0.0009529790462719078},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":5.126669358054452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020945501880404742,"output_power":0.0001897288387187065,"reference_active":true,"delta_db":-0.429574225414353},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010240530560263733,"output_power":8.349895478393709e-05,"reference_active":true,"delta_db":-0.8864141880755373},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779418868232472e-05,"output_power":2.0552990408534016e-05,"reference_active":true,"delta_db":-1.3107898185389255}],"active_band_count":3,"active_band_mae_db":0.8755927440096052}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":0.7289907829586759,"rms_envelope_nrmse":0.08005336244048449,"envelope_correlation":0.9930333906355959,"rms_gain_error_db":-0.24199465423789546,"reference_mean_power":0.001017821671515535,"output_mean_power":0.0009626584520254938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":4.336511567866126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018218362838522706,"output_power":0.00016287334397268902,"reference_active":true,"delta_db":-0.4865933425208001},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6513405903536415e-05,"output_power":4.406001435603341e-05,"reference_active":true,"delta_db":-1.0810684746927854},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.038210274571303e-06,"output_power":8.572928830011676e-07,"reference_active":true,"delta_db":-3.7611976817384694}],"active_band_count":3,"active_band_mae_db":1.7762864996506849}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":2.0811644145893267,"rms_envelope_nrmse":0.25584381095794695,"envelope_correlation":0.9491972413190809,"rms_gain_error_db":-0.8244678851499178,"reference_mean_power":0.00014923818668565985,"output_mean_power":0.00012343353730756936},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":7.990988889403065e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7339937941838396e-05,"output_power":3.449283437235284e-05,"reference_active":true,"delta_db":-1.3749880125267768},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462900675807451e-06,"output_power":4.982338970834429e-06,"reference_active":true,"delta_db":-1.1299421037515114},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2961437099516356e-06,"output_power":3.7786412650009743e-06,"reference_active":true,"delta_db":4.646825064027932}],"active_band_count":3,"active_band_mae_db":2.383918393435407}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":1.3743939010311548,"rms_envelope_nrmse":0.14657219213184308,"envelope_correlation":0.9885711911484073,"rms_gain_error_db":-0.8244753031352714,"reference_mean_power":0.00015077911609921791,"output_mean_power":0.0001247078129039742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":7.46020344966024e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.50390789043777e-05,"output_power":3.4064382931800884e-05,"reference_active":true,"delta_db":-1.2128897395860008},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3028613860270352e-06,"output_power":2.6601850230397724e-06,"reference_active":true,"delta_db":-0.9397850363700797},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.84375886973483e-08,"output_power":1.9452675831693824e-07,"reference_active":true,"delta_db":3.4234245679492794}],"active_band_count":3,"active_band_mae_db":1.8586997813017867}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":1.6331828627102425,"rms_envelope_nrmse":0.192131976756531,"envelope_correlation":0.966657582348528,"rms_gain_error_db":-0.15589193422636452,"reference_mean_power":0.00016484395516825918,"output_mean_power":0.000159031748085146},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":1.0813916284336478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.944970033970201e-05,"output_power":4.9059851600167004e-05,"reference_active":true,"delta_db":-0.03437433972933468},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1861966099880329e-05,"output_power":1.0698268122830766e-05,"reference_active":true,"delta_db":-0.44843200437175684},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.815426369634167e-06,"output_power":3.561953203986482e-06,"reference_active":true,"delta_db":-1.3094653714227946}],"active_band_count":3,"active_band_mae_db":0.5974239051746287}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":1.091091963297401,"rms_envelope_nrmse":0.09850565122147659,"envelope_correlation":0.990499883730537,"rms_gain_error_db":-0.1559456613840584,"reference_mean_power":0.00016654362747028518,"output_mean_power":0.00016066950423239553},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":9.377377349368442e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3529785361500974e-05,"output_power":4.332912330080481e-05,"reference_active":true,"delta_db":-0.020066238930567117},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5942380285921235e-06,"output_power":4.723303238345965e-06,"reference_active":true,"delta_db":-0.7349511278777993},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1488780558632e-07,"output_power":1.8186318777802786e-07,"reference_active":true,"delta_db":-2.3841104305670657}],"active_band_count":3,"active_band_mae_db":1.0463759324584774}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.00011951234204294975,"log_rms_mae_db":1.730802485849249,"rms_envelope_nrmse":0.18983144919382314,"envelope_correlation":0.9720797508250361,"rms_gain_error_db":-0.26025397655066623,"reference_mean_power":0.0005186016960204269,"output_mean_power":0.0004884369774631222},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":3.869293609534798e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017018269823121555,"output_power":0.0001560078935827798,"reference_active":true,"delta_db":-0.3776883202354441},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010605303684284745,"output_power":8.449447662586977e-05,"reference_active":true,"delta_db":-0.9869478891053564},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.258811740016227e-05,"output_power":1.6778118566865033e-05,"reference_active":true,"delta_db":-1.2913677716154495}],"active_band_count":3,"active_band_mae_db":0.8853346603187499}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":1.1000537240490948,"rms_envelope_nrmse":0.1105013013662112,"envelope_correlation":0.9900323723697857,"rms_gain_error_db":-0.2604082180078918,"reference_mean_power":0.0005239212873916427,"output_mean_power":0.0004934296275627784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":3.3612492298744625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001684428838944303,"output_power":0.00015067104661668348,"reference_active":true,"delta_db":-0.48422863275349737},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8728944802248194e-05,"output_power":4.540785199496846e-05,"reference_active":true,"delta_db":-1.1172123962001903},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9018720843055647e-06,"output_power":8.826536627714922e-07,"reference_active":true,"delta_db":-3.333909761476477}],"active_band_count":3,"active_band_mae_db":1.6451169301433886}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-75918-0013","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.3008646965026855,"predicted_mos":2.3008646965026855,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":47040,"input_sample_rate":16000,"input_seconds":2.94,"input_float32_sha256":"6d586ea31280657ee3db07697458956e72ea1504ec4b2a87167c77090f2d5996","internal_samples":47040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.224206476999825},"started_at":"2026-10-13T04:43:07.000844+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6601000424998347e-05,"native_input_samples":47040,"native_input_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","output_sha256":"67677862577b89bcac8eb20fabb4ec9ef961327561c96e8b0bb07af91ebab454","output_samples":47040,"length_delta_samples":0,"analysis_s":0.6846917620005115,"synthesis_s":0.04599224199955643,"total_s":0.7306840040000679,"rtf":0.24853197414968298,"output_peak":0.3310802397938809,"output_rms":0.03331344029887172,"finished_at":"2026-10-13T04:43:07.735438+00:00","evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"67677862577b89bcac8eb20fabb4ec9ef961327561c96e8b0bb07af91ebab454","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.761848355335853,"spectral_convergence":0.7227221148652864,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.418780277452055,"spectral_convergence":0.5382488041830461,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8890668784731846,"spectral_convergence":0.5925746805959087,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":3.4660370268637376e-05,"delta_db":-1.3563090262335888,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00019283697817628305,"delta_db":0.8181778027365937,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":4.645108405558248e-05,"delta_db":0.7119375726144205,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":8.220854502471897e-05,"delta_db":0.9974959644615744,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.0005514150373555368,"delta_db":0.2619222915041105,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.0002022132896657314,"delta_db":0.6308957744203986,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7961230719951145,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.426967863544718,"pitch":null,"output_rms_dbfs":-29.547610302847144,"output_peak":0.3310802397938809,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.023231837087031},"common_prefix":{"samples":47040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.761848355335853,"spectral_convergence":0.7227221148652864,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.418780277452055,"spectral_convergence":0.5382488041830461,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8890668784731846,"spectral_convergence":0.5925746805959087,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":3.4660370268637376e-05,"delta_db":-1.3563090262335888,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00019283697817628305,"delta_db":0.8181778027365937,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":4.645108405558248e-05,"delta_db":0.7119375726144205,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":8.220854502471897e-05,"delta_db":0.9974959644615744,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.0005514150373555368,"delta_db":0.2619222915041105,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.0002022132896657314,"delta_db":0.6308957744203986,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7961230719951145,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.426967863544718,"pitch":null,"output_rms_dbfs":-29.547610302847144,"output_peak":0.3310802397938809,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.023231837087031},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":47040,"output_samples":47040,"intelligibility":{"estoi":{"value":0.897191836699331,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03171546561641813,"rms_dbfs":-29.974578166391865,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.207027209586595,"mean":-2.270977513319781e-05,"zero_energy":false},"output":{"rms":0.03331344029887172,"rms_dbfs":-29.547610302847144,"peak":0.3310802397938809,"peak_dbfs":-9.601334778469377,"crest_factor_db":19.946275524377764,"mean":9.95464202340821e-09,"zero_energy":false},"rms_difference_db":0.4269678635447214},"full_scale":{"reference":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":294,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":1.0857278300267061e-07,"output_rms_dbfs":-69.64279029932914,"output_zero_energy":false,"output_peak":0.002470758207003542,"output_energy_fraction":3.993152680270211e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":287,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":54,"both_estimated_voiced_frames":43,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.581081081081081,"absolute_difference_cents_median":4.1042205972898795,"absolute_difference_cents_p95":11.765523400526089,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":47040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.0001280716783512989,"log_rms_mae_db":2.8432308050866015,"rms_envelope_nrmse":0.2075808185672143,"envelope_correlation":0.9593224027141201,"rms_gain_error_db":0.42703861324592146,"reference_mean_power":0.0010075555569028208,"output_mean_power":0.0011116622649712277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":5.126669358054452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020945501880404742,"output_power":0.00023206469532143352,"reference_active":true,"delta_db":0.44518303935135384},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010240530560263733,"output_power":0.00010775183017881296,"reference_active":true,"delta_db":0.22102197492605738},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779418868232472e-05,"output_power":2.808943733207419e-05,"reference_active":true,"delta_db":0.045890383852265515}],"active_band_count":3,"active_band_mae_db":0.23736513270989223}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":2.1605870035802757,"rms_envelope_nrmse":0.12336057579316398,"envelope_correlation":0.9855349744815894,"rms_gain_error_db":0.42745842095838726,"reference_mean_power":0.001017821671515535,"output_mean_power":0.0011230976945760158},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":4.336511567866126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018218362838522706,"output_power":0.0001979077480150893,"reference_active":true,"delta_db":0.3595344973499518},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6513405903536415e-05,"output_power":5.968289483948638e-05,"reference_active":true,"delta_db":0.23698398062274273},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.038210274571303e-06,"output_power":1.8538830757690808e-06,"reference_active":true,"delta_db":-0.4116664676756625}],"active_band_count":3,"active_band_mae_db":0.3360616485494523}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":2.0346627801684503,"rms_envelope_nrmse":0.24107384393347836,"envelope_correlation":0.968606678582316,"rms_gain_error_db":0.9723107977571511,"reference_mean_power":0.00014923818668565985,"output_mean_power":0.0001866856960186874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":7.990988889403065e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7339937941838396e-05,"output_power":5.651104861432758e-05,"reference_active":true,"delta_db":0.7690568183039986},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462900675807451e-06,"output_power":7.198439081172514e-06,"reference_active":true,"delta_db":0.4681085208426327},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2961437099516356e-06,"output_power":3.4764830046015936e-06,"reference_active":true,"delta_db":4.284869539910266}],"active_band_count":3,"active_band_mae_db":1.840678293018966}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":1.5750578391708845,"rms_envelope_nrmse":0.16815988407184634,"envelope_correlation":0.9891416363425006,"rms_gain_error_db":0.9723066443724893,"reference_mean_power":0.00015077911609921791,"output_mean_power":0.00018861310191120814},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":7.46020344966024e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.50390789043777e-05,"output_power":5.4216603395621775e-05,"reference_active":true,"delta_db":0.8054280591072172},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3028613860270352e-06,"output_power":3.757816698641355e-06,"reference_active":true,"delta_db":0.5604524429665204},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.84375886973483e-08,"output_power":1.9650743594369758e-07,"reference_active":true,"delta_db":3.467420961770222}],"active_band_count":3,"active_band_mae_db":1.6111004879479864}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":1.6176894119820182,"rms_envelope_nrmse":0.22071338267930757,"envelope_correlation":0.966546056722182,"rms_gain_error_db":0.7326035805188948,"reference_mean_power":0.00016484395516825918,"output_mean_power":0.00019513419639717307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":1.0813916284336478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.944970033970201e-05,"output_power":6.087730433371102e-05,"reference_active":true,"delta_db":0.9029174933174154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1861966099880329e-05,"output_power":1.3054897477883845e-05,"reference_active":true,"delta_db":0.4161678705508204},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.815426369634167e-06,"output_power":2.9083046268084673e-06,"reference_active":true,"delta_db":-2.189948524456979}],"active_band_count":3,"active_band_mae_db":1.1696779627750715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":1.0939865929098866,"rms_envelope_nrmse":0.14514054516072625,"envelope_correlation":0.9879823811090012,"rms_gain_error_db":0.7326149697678724,"reference_mean_power":0.00016654362747028518,"output_mean_power":0.00019714670219140757},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":9.377377349368442e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3529785361500974e-05,"output_power":5.342677329405148e-05,"reference_active":true,"delta_db":0.8897242006725412},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5942380285921235e-06,"output_power":5.816797848270347e-06,"reference_active":true,"delta_db":0.16943029808398502},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1488780558632e-07,"output_power":2.0340709899510757e-07,"reference_active":true,"delta_db":-1.8978973646443411}],"active_band_count":3,"active_band_mae_db":0.9856839544669557}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.00011951234204294975,"log_rms_mae_db":1.9115895887272207,"rms_envelope_nrmse":0.2803025396240835,"envelope_correlation":0.9425441540180799,"rms_gain_error_db":0.3370647145106859,"reference_mean_power":0.0005186016960204269,"output_mean_power":0.0005604545563175123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":3.869293609534798e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017018269823121555,"output_power":0.0001840678277023435,"reference_active":true,"delta_db":0.340624818954883},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010605303684284745,"output_power":9.478570583702474e-05,"reference_active":true,"delta_db":-0.48780260635503697},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.258811740016227e-05,"output_power":2.7804911851198214e-05,"reference_active":true,"delta_db":0.902414863027081}],"active_band_count":3,"active_band_mae_db":0.576947429445667}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":1.22592060994152,"rms_envelope_nrmse":0.16469638648218338,"envelope_correlation":0.9787878336727076,"rms_gain_error_db":0.3370657834664026,"reference_mean_power":0.0005239212873916427,"output_mean_power":0.0005662035955485695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":3.3612492298744625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001684428838944303,"output_power":0.00017800233556828696,"reference_active":true,"delta_db":0.23973032397327082},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8728944802248194e-05,"output_power":5.335072885194937e-05,"reference_active":true,"delta_db":-0.41711840911697007},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9018720843055647e-06,"output_power":1.8772072421142457e-06,"reference_active":true,"delta_db":-0.05669082884077303}],"active_band_count":3,"active_band_mae_db":0.2378465206436713}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-75918-0013","method":"facodec","status":"measured","mos":{"status":"measured","value":4.137192249298096,"predicted_mos":4.137192249298096,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":47200,"input_sample_rate":16000,"input_seconds":2.95,"input_float32_sha256":"eccfdf129ef1e51fdfe28e53d386073ce72e8eefdcd4bae8992a15388808f960","internal_samples":47200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4628694359998917},"started_at":"2026-10-13T06:57:40.036513+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.626899913593661e-05,"native_input_samples":47040,"native_input_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","output_sha256":"2a50d2c21b35a1a18f5f7e366e2dd0baa2378b309d5239c9910daf8af1995d08","output_samples":47200,"length_delta_samples":160,"analysis_s":0.8054277139999613,"synthesis_s":5.456476258999828,"total_s":6.261903972999789,"rtf":2.129899310544146,"output_peak":0.3585033714771271,"output_rms":0.028112724444736475,"finished_at":"2026-10-13T06:57:46.303338+00:00","evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"2a50d2c21b35a1a18f5f7e366e2dd0baa2378b309d5239c9910daf8af1995d08","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":47040,"output_samples":47200,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.02194128494824,"output_peak":0.3585033714771271},"common_prefix":{"samples":47040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9966101694915255,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.779641034824316,"spectral_convergence":0.46226115438825793,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.182686900389283,"spectral_convergence":0.5248848344153058,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.4011905683282055,"spectral_convergence":0.5693843911770513,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":4.977508424274614e-05,"delta_db":0.21547903237141608,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00017446076470071592,"delta_db":0.38325234711889633,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":3.1855286065824063e-05,"delta_db":-0.9262058709279699,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":4.193483897999109e-05,"delta_db":-1.9259238566022905,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.0003797180359576672,"delta_db":-1.358251503267345,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.00011526916529101111,"delta_db":-1.810069687561541,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1031970496415766,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0326178055063875,"pitch":null,"output_rms_dbfs":-31.00719597189825,"output_peak":0.3585033714771271,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.121172834513935},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":47040,"output_samples":47040,"intelligibility":{"estoi":{"value":0.8907670248595073,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03171546561641813,"rms_dbfs":-29.974578166391865,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.207027209586595,"mean":-2.270977513319781e-05,"zero_energy":false},"output":{"rms":0.028160489612894794,"rms_dbfs":-31.00719597189825,"peak":0.3585033714771271,"peak_dbfs":-8.910135114733821,"crest_factor_db":22.097060857164426,"mean":-0.00017531541915401732,"zero_energy":false},"rms_difference_db":-1.0326178055063835},"full_scale":{"reference":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":294,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":6.017193693793599e-07,"output_rms_dbfs":-62.20606008320242,"output_zero_energy":false,"output_peak":0.0019601313397288322,"output_energy_fraction":3.0970449202199495e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":287,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":77,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.972972972972973,"absolute_difference_cents_median":5.655934797019215,"absolute_difference_cents_p95":11.333248494666252,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":47040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.0001280716783512989,"log_rms_mae_db":1.7334311922092631,"rms_envelope_nrmse":0.19209093310755454,"envelope_correlation":0.9705801464835067,"rms_gain_error_db":-1.0325401554497624,"reference_mean_power":0.0010075555569028208,"output_mean_power":0.0007943556466385299},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":5.126669358054452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020945501880404742,"output_power":0.00014493516353892725,"reference_active":true,"delta_db":-1.5991700616741205},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010240530560263733,"output_power":0.00010139579623948658,"reference_active":true,"delta_db":-0.04302507961876621},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779418868232472e-05,"output_power":2.5513287270463488e-05,"reference_active":true,"delta_db":-0.3718758234657381}],"active_band_count":3,"active_band_mae_db":0.6713569882528749}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":1.4562099644574302,"rms_envelope_nrmse":0.1634982868297339,"envelope_correlation":0.980584201866317,"rms_gain_error_db":-1.0322716215744387,"reference_mean_power":0.001017821671515535,"output_mean_power":0.0008024990583927521},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":4.336511567866126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018218362838522706,"output_power":0.00012861106901545664,"reference_active":true,"delta_db":-1.5123099917058556},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6513405903536415e-05,"output_power":5.7777132568846e-05,"reference_active":true,"delta_db":0.09604502578369324},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.038210274571303e-06,"output_power":2.066020203774499e-06,"reference_active":true,"delta_db":0.05885577676630608}],"active_band_count":3,"active_band_mae_db":0.5557369314186184}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":2.0568690268541907,"rms_envelope_nrmse":0.1114386033752318,"envelope_correlation":0.9905090354681082,"rms_gain_error_db":-0.15446686936235152,"reference_mean_power":0.00014923818668565985,"output_mean_power":0.0001440234720845083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":7.990988889403065e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7339937941838396e-05,"output_power":4.510855097222919e-05,"reference_active":true,"delta_db":-0.20968807850215046},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462900675807451e-06,"output_power":7.820539333093058e-06,"reference_active":true,"delta_db":0.8280922305448071},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2961437099516356e-06,"output_power":1.3209617734941957e-06,"reference_active":true,"delta_db":0.08237093402797961}],"active_band_count":3,"active_band_mae_db":0.37338374769164573}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":1.752452462129078,"rms_envelope_nrmse":0.09863116385004986,"envelope_correlation":0.992336288576535,"rms_gain_error_db":-0.15446470825534653,"reference_mean_power":0.00015077911609921791,"output_mean_power":0.0001455106304006598},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":7.46020344966024e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.50390789043777e-05,"output_power":4.311540366356769e-05,"reference_active":true,"delta_db":-0.18957043670508655},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3028613860270352e-06,"output_power":4.17066823215028e-06,"reference_active":true,"delta_db":1.013152963404073},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.84375886973483e-08,"output_power":8.319063038770552e-08,"reference_active":true,"delta_db":-0.2656247754238973}],"active_band_count":3,"active_band_mae_db":0.48944939184435227}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":2.1333814998389684,"rms_envelope_nrmse":0.23783056189542864,"envelope_correlation":0.9690867046086509,"rms_gain_error_db":-1.5775208230767483,"reference_mean_power":0.00016484395516825918,"output_mean_power":0.00011463597890469526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":1.0813916284336478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.944970033970201e-05,"output_power":2.8797205316951482e-05,"reference_active":true,"delta_db":-2.3481332134190684},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1861966099880329e-05,"output_power":1.2214443530923657e-05,"reference_active":true,"delta_db":0.12717007513221382},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.815426369634167e-06,"output_power":3.6315736171264574e-06,"reference_active":true,"delta_db":-1.2253989430999426}],"active_band_count":3,"active_band_mae_db":1.2335674105504084}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":1.8335582698863588,"rms_envelope_nrmse":0.21454922855427827,"envelope_correlation":0.9779526109293437,"rms_gain_error_db":-1.5775166451529268,"reference_mean_power":0.00016654362747028518,"output_mean_power":0.00011581807845100682},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":9.377377349368442e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3529785361500974e-05,"output_power":2.5024029297424678e-05,"reference_active":true,"delta_db":-2.4042928588542574},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5942380285921235e-06,"output_power":6.394788220916753e-06,"reference_active":true,"delta_db":0.580852251520262},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1488780558632e-07,"output_power":2.4701136049359706e-07,"reference_active":true,"delta_db":-1.0543891468177344}],"active_band_count":3,"active_band_mae_db":1.3465114190640846}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.00011951234204294975,"log_rms_mae_db":2.4702814322955433,"rms_envelope_nrmse":0.24760461111827856,"envelope_correlation":0.9658899688253376,"rms_gain_error_db":-1.4322034205864098,"reference_mean_power":0.0005186016960204269,"output_mean_power":0.0003729182097703659},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":3.869293609534798e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017018269823121555,"output_power":0.0001183296203940178,"reference_active":true,"delta_db":-1.5782193379034966},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010605303684284745,"output_power":9.835050985220491e-05,"reference_active":true,"delta_db":-0.32746493453274284},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.258811740016227e-05,"output_power":2.1104952479896773e-05,"reference_active":true,"delta_db":-0.29495657655822194}],"active_band_count":3,"active_band_mae_db":0.7335469496648205}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":2.088865090140889,"rms_envelope_nrmse":0.21222464137673436,"envelope_correlation":0.9780608818675176,"rms_gain_error_db":-1.4320239858654424,"reference_mean_power":0.0005239212873916427,"output_mean_power":0.00037675900909056283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":3.3612492298744625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001684428838944303,"output_power":0.00012174146249910412,"reference_active":true,"delta_db":-1.4101415363055931},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8728944802248194e-05,"output_power":5.573884585532223e-05,"reference_active":true,"delta_db":-0.226942259915288},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9018720843055647e-06,"output_power":2.0420427015412218e-06,"reference_active":true,"delta_db":0.3088351555903698}],"active_band_count":3,"active_band_mae_db":0.6486396506037503}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-75918-0013","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.745922565460205,"predicted_mos":2.745922565460205,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":130048,"input_sample_rate":44100,"input_seconds":2.948934240362812,"input_float32_sha256":"5b485a3c95c85ccc503f2dddd764b558e67cae3bba36b6ea1f8c0f65fa7113ae","internal_samples":47183,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.212408344999858},"started_at":"2026-10-13T07:25:17.376679+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0041244820022257045,"native_input_samples":129654,"native_input_sha256":"90d35804e1e01393cb05830d11252483b3624e7ef77978b8caef91b92205510c","output_sha256":"8a184f2e10e015bffaf0d4c6dc752474f19f615fc781134522d28031a254b84d","output_samples":130048,"length_delta_samples":394,"analysis_s":0.097819063001225,"synthesis_s":18.045682721996855,"total_s":18.14350178499808,"rtf":6.171259110543565,"output_peak":0.3500747084617615,"output_rms":0.026147233015569333,"finished_at":"2026-10-13T07:25:35.532176+00:00","evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"8a184f2e10e015bffaf0d4c6dc752474f19f615fc781134522d28031a254b84d","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":47040,"output_samples":47183,"length_delta_samples":143,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.647319576554768,"output_peak":0.35103444923285654},"common_prefix":{"samples":47040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9969692473984274,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.98595096648021,"spectral_convergence":0.6718813263751368,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.180590035447052,"spectral_convergence":0.7008979601262163,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.163468527254073,"spectral_convergence":0.7382744655334647,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":1.3734129258163635e-05,"delta_db":-5.376629700713603,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00014071353049320063,"delta_db":-0.5503667744692426,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":4.93740067375325e-05,"delta_db":0.9769627540544789,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":3.453826216629894e-05,"delta_db":-2.7686688368903996,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.00037707776890961824,"delta_db":-1.3885544741474576,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":7.096669373034467e-05,"delta_db":-3.9166554469155757,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.49630633119846,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.659621300977356,"pitch":null,"output_rms_dbfs":-31.634199467369214,"output_peak":0.35103444923285654,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.11000317639378},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":47040,"output_samples":47040,"intelligibility":{"estoi":{"value":0.6110666881697486,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03171546561641813,"rms_dbfs":-29.974578166391865,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.207027209586595,"mean":-2.270977513319781e-05,"zero_energy":false},"output":{"rms":0.02619932043575098,"rms_dbfs":-31.634199467369214,"peak":0.35103444923285654,"peak_dbfs":-9.093005227515594,"crest_factor_db":22.54119423985362,"mean":-3.3268072701511756e-05,"zero_energy":false},"rms_difference_db":-1.6596213009773493},"full_scale":{"reference":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":294,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":2.3818768689525926e-07,"output_rms_dbfs":-66.23080693112105,"output_zero_energy":false,"output_peak":0.002323385811745981,"output_energy_fraction":1.4163584218253676e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":287,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":66,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.7567567567567568,"absolute_difference_cents_median":10.578693739469431,"absolute_difference_cents_p95":134.59838406106235,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":47040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.0001280716783512989,"log_rms_mae_db":2.2381293125386077,"rms_envelope_nrmse":0.22651811509047529,"envelope_correlation":0.9730926229801973,"rms_gain_error_db":-1.6595779957250392,"reference_mean_power":0.0010075555569028208,"output_mean_power":0.0006875609500756471},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":5.126669358054452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020945501880404742,"output_power":0.0001439828873704194,"reference_active":true,"delta_db":-1.6277989259186252},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010240530560263733,"output_power":7.991372702898296e-05,"reference_active":true,"delta_db":-1.0770107211763105},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779418868232472e-05,"output_power":2.8770779529158082e-05,"reference_active":true,"delta_db":0.14997627609853142}],"active_band_count":3,"active_band_mae_db":0.9515953077311557}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":1.8305432977303397,"rms_envelope_nrmse":0.19587570355245418,"envelope_correlation":0.9881385185331188,"rms_gain_error_db":-1.6597066664283942,"reference_mean_power":0.001017821671515535,"output_mean_power":0.0006945460200367074},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":4.336511567866126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018218362838522706,"output_power":0.00012653557171719927,"reference_active":true,"delta_db":-1.5829671560386571},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6513405903536415e-05,"output_power":4.714925740958489e-05,"reference_active":true,"delta_db":-0.7867662470500096},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.038210274571303e-06,"output_power":2.20303168299574e-06,"reference_active":true,"delta_db":0.3377175649289204}],"active_band_count":3,"active_band_mae_db":0.9024836560058623}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":3.700192512409917,"rms_envelope_nrmse":0.274090454342618,"envelope_correlation":0.9391720270801651,"rms_gain_error_db":-0.5967015537128963,"reference_mean_power":0.00014923818668565985,"output_mean_power":0.00013007978431468195},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":7.990988889403065e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7339937941838396e-05,"output_power":4.414478728851314e-05,"reference_active":true,"delta_db":-0.30348255760574905},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462900675807451e-06,"output_power":7.611662574153112e-06,"reference_active":true,"delta_db":0.7105204613964837},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2961437099516356e-06,"output_power":1.4281229712891763e-06,"reference_active":true,"delta_db":0.42112448205308456}],"active_band_count":3,"active_band_mae_db":0.47837583368510583}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":3.007058973388581,"rms_envelope_nrmse":0.23871664862698458,"envelope_correlation":0.9516163800474916,"rms_gain_error_db":-0.596721797625172,"reference_mean_power":0.00015077911609921791,"output_mean_power":0.00013142228482496783},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":7.46020344966024e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.50390789043777e-05,"output_power":4.175216580458554e-05,"reference_active":true,"delta_db":-0.3291049165969792},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3028613860270352e-06,"output_power":3.4825296143812027e-06,"reference_active":true,"delta_db":0.23004470550363365},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.84375886973483e-08,"output_power":1.0221332204341831e-07,"reference_active":true,"delta_db":0.6287061078218521}],"active_band_count":3,"active_band_mae_db":0.395951909974155}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":3.505983944169177,"rms_envelope_nrmse":0.3733304273644786,"envelope_correlation":0.8797459337623794,"rms_gain_error_db":-0.8881697004675457,"reference_mean_power":0.00016484395516825918,"output_mean_power":0.000134355687675868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":1.0813916284336478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.944970033970201e-05,"output_power":4.260240470450661e-05,"reference_active":true,"delta_db":-0.6472955049288589},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1861966099880329e-05,"output_power":2.069310594048499e-05,"reference_active":true,"delta_db":2.4166900263616746},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.815426369634167e-06,"output_power":1.170513969744226e-05,"reference_active":true,"delta_db":3.857418545828759}],"active_band_count":3,"active_band_mae_db":2.3071346923730975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":3.1259938926219917,"rms_envelope_nrmse":0.3204097607801939,"envelope_correlation":0.9034095392900544,"rms_gain_error_db":-0.8882042737232411,"reference_mean_power":0.00016654362747028518,"output_mean_power":0.00013573992108751055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":9.377377349368442e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3529785361500974e-05,"output_power":3.942487211720782e-05,"reference_active":true,"delta_db":-0.43016232446625535},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5942380285921235e-06,"output_power":1.2583653306461288e-05,"reference_active":true,"delta_db":3.5206580331296147},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1488780558632e-07,"output_power":9.463708189499792e-07,"reference_active":true,"delta_db":4.779054981113954}],"active_band_count":3,"active_band_mae_db":2.909958446236608}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.00011951234204294975,"log_rms_mae_db":3.2979205308397725,"rms_envelope_nrmse":0.3337817014986573,"envelope_correlation":0.9645490538154358,"rms_gain_error_db":-2.852659885108256,"reference_mean_power":0.0005186016960204269,"output_mean_power":0.0002688858474995882},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":3.869293609534798e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017018269823121555,"output_power":8.265229161144706e-05,"reference_active":true,"delta_db":-3.1366050578316997},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010605303684284745,"output_power":4.847679338575512e-05,"reference_active":true,"delta_db":-3.399892244571248},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.258811740016227e-05,"output_power":1.159759015492072e-05,"reference_active":true,"delta_db":-2.8951227905440873}],"active_band_count":3,"active_band_mae_db":3.1438733643156787}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":3.147101994580378,"rms_envelope_nrmse":0.30903597036233255,"envelope_correlation":0.9793932472847369,"rms_gain_error_db":-2.8527217808023253,"reference_mean_power":0.0005239212873916427,"output_mean_power":0.00027164009050968397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":3.3612492298744625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001684428838944303,"output_power":8.192616117085283e-05,"reference_active":true,"delta_db":-3.1303006280595076},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8728944802248194e-05,"output_power":2.7060287109330138e-05,"reference_active":true,"delta_db":-3.3651979768922233},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9018720843055647e-06,"output_power":9.447177343772912e-07,"reference_active":true,"delta_db":-3.0387923582962073}],"active_band_count":3,"active_band_mae_db":3.178096987749313}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-75918-0013","method":"golf","status":"measured","mos":{"status":"measured","value":3.3684537410736084,"predicted_mos":3.3684537410736084,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":70320,"input_sample_rate":24000,"input_seconds":2.93,"input_float32_sha256":"4c0c2e3ae60cefcc034f04b68ca334ca55eb883589bda1c9214bfa6fe0685bad","internal_samples":46880,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.877284539999891},"started_at":"2026-10-13T05:02:53.707469+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001407316000040737,"native_input_samples":70560,"native_input_sha256":"5fc35a6bfadec42d29f0dcc7adb69d09937d839d95641de631769be7d67af8e3","output_sha256":"2a2cecc14c8eec747e642c23a5627131cf429259d4d30f944c202f3a8b82f040","output_samples":70320,"length_delta_samples":-240,"analysis_s":0.277238405999924,"synthesis_s":0.07416601200020523,"total_s":0.3514044180001292,"rtf":0.11952531224494192,"output_peak":0.2787007689476013,"output_rms":0.026088721740950626,"finished_at":"2026-10-13T05:02:54.064708+00:00","analysis_stages":{"external_f0_s":0.03429148899977008},"evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"2a2cecc14c8eec747e642c23a5627131cf429259d4d30f944c202f3a8b82f040","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":47040,"output_samples":46880,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-31.66653818404757,"output_peak":0.2789683572912012},"common_prefix":{"samples":46880,"sample_rate":16000,"reference_coverage":0.9965986394557823,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":46880,"output_samples":46880,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.59727096876432,"spectral_convergence":0.6358806726192588,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.715557445922557,"spectral_convergence":0.5055685214344643,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.861847609852871,"spectral_convergence":0.533027986791062,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.761943867999873e-05,"output_power":5.1729417484429926e-05,"delta_db":0.3595331612606444,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0001604742499068295,"output_power":0.00015896013757634125,"delta_db":-0.04117124395997284,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.924783403863831e-05,"output_power":3.6203077073286086e-05,"delta_db":-0.35070209556530174,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.55658191303647e-05,"output_power":5.102233570349904e-05,"delta_db":-1.089171554427504,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005212264409649209,"output_power":0.00030082136891981077,"delta_db":-2.3871775541762434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017508630943452875,"output_power":8.257588809070407e-05,"delta_db":-3.26398935361756,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2486241605012045,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.706397065298293,"pitch":null,"output_rms_dbfs":-31.66653818404757,"output_peak":0.2789683572912012,"reference_rms_dbfs":-29.960141118749277,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.05822534151325},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":46880,"output_samples":46880,"intelligibility":{"estoi":{"value":0.8696081072702535,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.031768224567250855,"rms_dbfs":-29.960141118749277,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.192590161944015,"mean":-1.753588992174168e-05,"zero_energy":false},"output":{"rms":0.026101958256959785,"rms_dbfs":-31.66653818404757,"peak":0.2789683572912012,"peak_dbfs":-11.088901098196128,"crest_factor_db":20.57763708585144,"mean":-0.00022641686839536555,"zero_energy":false},"rms_difference_db":-1.7063970652982938},"full_scale":{"reference":{"sample_count":46880,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":46880,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":293,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":4.553776173513505e-07,"output_rms_dbfs":-63.416283195699954,"output_zero_energy":false,"output_peak":0.002564184746899239,"output_energy_fraction":2.7374054308067387e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":286,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":67,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":1,"paired_voiced_fraction":0.9054054054054054,"absolute_difference_cents_median":5.493494075589533,"absolute_difference_cents_p95":23.835940562711595,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":46880,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":585,"rms_floor":0.0001280716783512989,"log_rms_mae_db":2.421727422972992,"rms_envelope_nrmse":0.24725135053763692,"envelope_correlation":0.9615993031812182,"rms_gain_error_db":-1.7062496721085076,"reference_mean_power":0.0010109053656626016,"output_mean_power":0.0006824730942659775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":585,"frequency_resolution_hz":0.34188034188034183,"power_floor":5.155641916025045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002129504966218599,"output_power":0.00013640780293660575,"reference_active":true,"delta_db":-1.9343944324748548},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010066340492935683,"output_power":6.624801331485996e-05,"reference_active":true,"delta_db":-1.8169875735508234},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7865965231184505e-05,"output_power":2.048319785527363e-05,"reference_active":true,"delta_db":-1.336763308568273}],"active_band_count":3,"active_band_mae_db":1.696048438197984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":579,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":1.8856146090187007,"rms_envelope_nrmse":0.22150623833399924,"envelope_correlation":0.9741747945510799,"rms_gain_error_db":-1.7060155895627371,"reference_mean_power":0.0010212657063286019,"output_mean_power":0.0006895046347258171},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":579,"frequency_resolution_hz":0.3454231433506045,"power_floor":4.3603162130386704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018463756707563043,"output_power":0.00012107749353509787,"reference_active":true,"delta_db":-1.8325664701535889},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.518512893254967e-05,"output_power":3.588333978129764e-05,"reference_active":true,"delta_db":-1.8692920407871636},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0675918854207115e-06,"output_power":1.551446805847973e-06,"reference_active":true,"delta_db":-1.2472792943833193}],"active_band_count":3,"active_band_mae_db":1.6497126017746906}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":585,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":2.2365186810343434,"rms_envelope_nrmse":0.17638153716795416,"envelope_correlation":0.9771243732907682,"rms_gain_error_db":-0.0007060617104342594,"reference_mean_power":0.00014974825974451895,"output_mean_power":0.00014972391614299053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":585,"frequency_resolution_hz":0.34188034188034183,"power_floor":8.062337877182929e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7769585184437016e-05,"output_power":5.105879100548916e-05,"reference_active":true,"delta_db":0.28919057129707293},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.451999280897251e-06,"output_power":7.481391784484773e-06,"reference_active":true,"delta_db":0.6428808809896163},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.286631777388002e-06,"output_power":1.1072978744131476e-06,"reference_active":true,"delta_db":-0.6518980721727413}],"active_band_count":3,"active_band_mae_db":0.5279898414864769}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":579,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":1.6781249919896577,"rms_envelope_nrmse":0.13678963015585865,"envelope_correlation":0.9859409883685091,"rms_gain_error_db":-0.0007029896508095141,"reference_mean_power":0.0001512998219881135,"output_mean_power":0.00015127533316646545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":579,"frequency_resolution_hz":0.3454231433506045,"power_floor":7.530040572489559e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5429249523722676e-05,"output_power":4.8599562287525724e-05,"reference_active":true,"delta_db":0.2929679533031864},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2845478542159055e-06,"output_power":3.573602837720585e-06,"reference_active":true,"delta_db":0.36630690624968276},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.883251210480151e-08,"output_power":8.601028242857465e-08,"reference_active":true,"delta_db":-0.1402157000193689}],"active_band_count":3,"active_band_mae_db":0.26649685319074606}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":585,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":2.075360979515273,"rms_envelope_nrmse":0.26199872439530597,"envelope_correlation":0.9483103510300542,"rms_gain_error_db":-1.2309615070590314,"reference_mean_power":0.00016540482526779837,"output_mean_power":0.0001245810606838366},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":585,"frequency_resolution_hz":0.34188034188034183,"power_floor":1.0865890479035355e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.994405140847851e-05,"output_power":3.358698457346963e-05,"reference_active":true,"delta_db":-1.72312754151403},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.156576077381852e-05,"output_power":1.097540429900736e-05,"reference_active":true,"delta_db":-0.2275367787688613},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8394681831155415e-06,"output_power":4.121756234779748e-06,"reference_active":true,"delta_db":-0.6971533540568465}],"active_band_count":3,"active_band_mae_db":0.8826058914465792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":579,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":1.7214276171095628,"rms_envelope_nrmse":0.20063817688676516,"envelope_correlation":0.9745706642386657,"rms_gain_error_db":-1.230960143383398,"reference_mean_power":0.00016711790131044703,"output_mean_power":0.00012587136988389267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":579,"frequency_resolution_hz":0.3454231433506045,"power_floor":9.423219098674636e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.386326597612678e-05,"output_power":2.8877768373832396e-05,"reference_active":true,"delta_db":-1.8153733665893512},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.383312429759054e-06,"output_power":5.903921021069792e-06,"reference_active":true,"delta_db":0.40090953564918197},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2528823198301986e-07,"output_power":3.3470042989866667e-07,"reference_active":true,"delta_db":0.12387918169898275}],"active_band_count":3,"active_band_mae_db":0.780054027979172}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":585,"rms_floor":0.00011951234204294975,"log_rms_mae_db":3.4258010339016844,"rms_envelope_nrmse":0.36525839162217977,"envelope_correlation":0.9523310573258172,"rms_gain_error_db":-3.0755175752953505,"reference_mean_power":0.0005203389153280662,"output_mean_power":0.0002562917047122032},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":585,"frequency_resolution_hz":0.34188034188034183,"power_floor":3.8922283023323295e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017339629787951351,"output_power":9.166563560766895e-05,"reference_active":true,"delta_db":-2.768332666025346},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010472268867475174,"output_power":6.66674024423182e-05,"reference_active":true,"delta_db":-1.961272497962829},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2509572806578883e-05,"output_power":1.3213868400124346e-05,"reference_active":true,"delta_db":-2.313372756500525}],"active_band_count":3,"active_band_mae_db":2.347659306829567}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":46880,"uncovered_tail_samples":0,"metrics":{"frame_count":579,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":3.3524647852472644,"rms_envelope_nrmse":0.3449486790673064,"envelope_correlation":0.9642264799212948,"rms_gain_error_db":-3.075504297994801,"reference_mean_power":0.000525715491091994,"output_mean_power":0.0002589407160509754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":579,"frequency_resolution_hz":0.3454231433506045,"power_floor":3.3803383800508367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001709388384078204,"output_power":0.0001013476451698018,"reference_active":true,"delta_db":-2.270270862043777},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.74742711303404e-05,"output_power":3.512565746140238e-05,"reference_active":true,"delta_db":-2.1384901053958516},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.937984834408187e-06,"output_power":1.3434493290872218e-06,"reference_active":true,"delta_db":-1.5912908339890977}],"active_band_count":3,"active_band_mae_db":2.0000172671429084}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-75918-0013","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.641866683959961,"predicted_mos":2.641866683959961,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":70680,"input_sample_rate":24000,"input_seconds":2.945,"input_float32_sha256":"914f10d49f272575c9400a7f16c8007a598c30b8bb2a3cb78a34d5042a441f4b","internal_samples":47120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1942522900026233},"started_at":"2026-10-13T07:43:19.020959+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0016991759985103272,"native_input_samples":70560,"native_input_sha256":"5fc35a6bfadec42d29f0dcc7adb69d09937d839d95641de631769be7d67af8e3","output_sha256":"289052ac29007f50afb2923a68b1f63fcc693439d7ce06859140aa818a841818","output_samples":70680,"length_delta_samples":120,"analysis_s":1.1149110049991577,"synthesis_s":6.129197636000754,"total_s":7.244108640999912,"rtf":2.463982530952351,"output_peak":0.3255367577075958,"output_rms":0.02971222272481235,"finished_at":"2026-10-13T07:43:26.272253+00:00","analysis_stages":{"external_f0_s":0.7378123640010017},"evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"289052ac29007f50afb2923a68b1f63fcc693439d7ce06859140aa818a841818","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":47040,"output_samples":47120,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.53745624793484,"output_peak":0.3260702034315648},"common_prefix":{"samples":47040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.99830220713073,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.229121880290703,"spectral_convergence":0.604113641250611,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.52968985537811,"spectral_convergence":0.5566832510215075,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.651307533203211,"spectral_convergence":0.5778574064653543,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":6.974596697231598e-05,"delta_db":1.6805499930533194,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00016865772081061617,"delta_db":0.23633690064228438,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":3.889677667051171e-05,"delta_db":-0.05888483214705795,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":5.831645305054997e-05,"delta_db":-0.49376264167782735,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.00043173842774428577,"delta_db":-0.8006567057433989,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.00011772114974758129,"delta_db":-1.7186562195790702,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8314745488071598,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5556137603142447,"pitch":null,"output_rms_dbfs":-30.530191926706106,"output_peak":0.3260702034315648,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.803373089624007},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":47040,"output_samples":47040,"intelligibility":{"estoi":{"value":0.9049301271765287,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03171546561641813,"rms_dbfs":-29.974578166391865,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.207027209586595,"mean":-2.270977513319781e-05,"zero_energy":false},"output":{"rms":0.02975023520908467,"rms_dbfs":-30.530191926706106,"peak":0.3260702034315648,"peak_dbfs":-9.733777711491364,"crest_factor_db":20.796414215214742,"mean":0.00015186164836696492,"zero_energy":false},"rms_difference_db":-0.555613760314241},"full_scale":{"reference":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":294,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":7.524390010045205e-08,"output_rms_dbfs":-71.235287018483,"output_zero_energy":false,"output_peak":0.0022214370009570853,"output_energy_fraction":3.469959504405526e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":287,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":66,"both_estimated_voiced_frames":61,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.8243243243243243,"absolute_difference_cents_median":6.827779227391535,"absolute_difference_cents_p95":26.11512146698366,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":47040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.0001280716783512989,"log_rms_mae_db":3.1449248995040646,"rms_envelope_nrmse":0.21235264207916854,"envelope_correlation":0.9528180954475974,"rms_gain_error_db":-0.5555270126894539,"reference_mean_power":0.0010075555569028208,"output_mean_power":0.0008865766751907077},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":5.126669358054452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020945501880404742,"output_power":0.00018164684075789084,"reference_active":true,"delta_db":-0.6186292214197826},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010240530560263733,"output_power":8.226276432231649e-05,"reference_active":true,"delta_db":-0.9511915868881526},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779418868232472e-05,"output_power":1.5407414020783186e-05,"reference_active":true,"delta_db":-2.562242485392722}],"active_band_count":3,"active_band_mae_db":1.3773544312335524}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":2.5610850582642257,"rms_envelope_nrmse":0.14107341242487204,"envelope_correlation":0.9790741596537836,"rms_gain_error_db":-0.5551566826078593,"reference_mean_power":0.001017821671515535,"output_mean_power":0.000895686493654232},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":4.336511567866126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018218362838522706,"output_power":0.00015700145734237446,"reference_active":true,"delta_db":-0.6460566357142185},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6513405903536415e-05,"output_power":4.3263190008624336e-05,"reference_active":true,"delta_db":-1.1603294282948073},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.038210274571303e-06,"output_power":1.0401085985656794e-06,"reference_active":true,"delta_db":-2.921702998293969}],"active_band_count":3,"active_band_mae_db":1.5760296874343318}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":2.2531066112503297,"rms_envelope_nrmse":0.19244831153341735,"envelope_correlation":0.9725412455631556,"rms_gain_error_db":0.3258667560817477,"reference_mean_power":0.00014923818668565985,"output_mean_power":0.0001608668801236882},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":7.990988889403065e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7339937941838396e-05,"output_power":5.109113142342163e-05,"reference_active":true,"delta_db":0.33117836053123173},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462900675807451e-06,"output_power":7.344448414283741e-06,"reference_active":true,"delta_db":0.5553170321284228},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2961437099516356e-06,"output_power":1.3660634444102728e-06,"reference_active":true,"delta_db":0.22817713245220828}],"active_band_count":3,"active_band_mae_db":0.3715575083706209}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":1.6227684897615975,"rms_envelope_nrmse":0.13143170144439917,"envelope_correlation":0.9869668889277015,"rms_gain_error_db":0.32586480679478325,"reference_mean_power":0.00015077911609921791,"output_mean_power":0.0001625278063650271},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":7.46020344966024e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.50390789043777e-05,"output_power":4.85116854593937e-05,"reference_active":true,"delta_db":0.3225686357770141},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3028613860270352e-06,"output_power":3.741162870744818e-06,"reference_active":true,"delta_db":0.5411626785881334},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.84375886973483e-08,"output_power":9.882041045221582e-08,"reference_active":true,"delta_db":0.482097606317887}],"active_band_count":3,"active_band_mae_db":0.4486096402276782}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":1.9783517987892356,"rms_envelope_nrmse":0.22255194714198318,"envelope_correlation":0.9555406610158164,"rms_gain_error_db":-0.36542235573723886,"reference_mean_power":0.00016484395516825918,"output_mean_power":0.00015154122395812615},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":1.0813916284336478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.944970033970201e-05,"output_power":4.557779239617311e-05,"reference_active":true,"delta_db":-0.3541037827301864},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1861966099880329e-05,"output_power":1.0907708150844493e-05,"reference_active":true,"delta_db":-0.3642316919426264},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.815426369634167e-06,"output_power":3.0838568647663374e-06,"reference_active":true,"delta_db":-1.9354053431163942}],"active_band_count":3,"active_band_mae_db":0.8845802725964024}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":1.3367941632974123,"rms_envelope_nrmse":0.11924879764011587,"envelope_correlation":0.9870502344010338,"rms_gain_error_db":-0.3654361165212433,"reference_mean_power":0.00016654362747028518,"output_mean_power":0.00015310324940561663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":9.377377349368442e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3529785361500974e-05,"output_power":3.946020551381317e-05,"reference_active":true,"delta_db":-0.4262718294170965},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5942380285921235e-06,"output_power":5.133977385570987e-06,"reference_active":true,"delta_db":-0.3728698983957486},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1488780558632e-07,"output_power":2.1824894392115848e-07,"reference_active":true,"delta_db":-1.5920369149257567}],"active_band_count":3,"active_band_mae_db":0.797059547579534}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.00011951234204294975,"log_rms_mae_db":2.3348686612280987,"rms_envelope_nrmse":0.275984518614107,"envelope_correlation":0.9480023653241326,"rms_gain_error_db":-1.2549294698973368,"reference_mean_power":0.0005186016960204269,"output_mean_power":0.00038845524141877343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":3.869293609534798e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017018269823121555,"output_power":0.00012699707728241452,"reference_active":true,"delta_db":-1.271216788805578},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010605303684284745,"output_power":7.629871836212209e-05,"reference_active":true,"delta_db":-1.4300586621424074},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.258811740016227e-05,"output_power":1.2719606970807641e-05,"reference_active":true,"delta_db":-2.494063442382109}],"active_band_count":3,"active_band_mae_db":1.7317796311100313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":1.6322452831963639,"rms_envelope_nrmse":0.20386224094372232,"envelope_correlation":0.9765202408730361,"rms_gain_error_db":-1.2550158781634633,"reference_mean_power":0.0005239212873916427,"output_mean_power":0.0003924320388812327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":3.3612492298744625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001684428838944303,"output_power":0.00012566535458097196,"reference_active":true,"delta_db":-1.2723710733229932},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8728944802248194e-05,"output_power":4.0839878820616404e-05,"reference_active":true,"delta_db":-1.5776775304895827},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9018720843055647e-06,"output_power":9.442448970211936e-07,"reference_active":true,"delta_db":-3.0409665745084973}],"active_band_count":3,"active_band_mae_db":1.9636717261070242}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-75918-0013","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.2405803203582764,"predicted_mos":3.2405803203582764,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":70800,"input_sample_rate":24000,"input_seconds":2.95,"input_float32_sha256":"a0d544d17238be3e61960c4a033f2b2a002d95140af8f6f3bf07b36c79cfcea4","internal_samples":47200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.998762631999853},"started_at":"2026-10-13T05:11:13.500423+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0021519029996852623,"native_input_samples":70560,"native_input_sha256":"5fc35a6bfadec42d29f0dcc7adb69d09937d839d95641de631769be7d67af8e3","output_sha256":"5d42ed3398f6933ef5be4a808067a5d06d7e0a7847ccb57d4998babe892c17dd","output_samples":70800,"length_delta_samples":240,"analysis_s":0.42018961099893204,"synthesis_s":0.08477847100039071,"total_s":0.5049680819993227,"rtf":0.17175785102017782,"output_peak":0.32021868228912354,"output_rms":0.030432541889642957,"finished_at":"2026-10-13T05:11:14.013803+00:00","analysis_stages":{"external_f0_s":0.07864076199984993},"evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"5d42ed3398f6933ef5be4a808067a5d06d7e0a7847ccb57d4998babe892c17dd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":47040,"output_samples":47200,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.328232207396763,"output_peak":0.3202393193514826},"common_prefix":{"samples":47040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9966101694915255,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.694449355544618,"spectral_convergence":0.35644034207651326,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.407560551724569,"spectral_convergence":0.32344439036626266,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.8100646056060725,"spectral_convergence":0.354363652063159,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":4.254404770542978e-05,"delta_db":-0.46625293522295574,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00015491775127331352,"delta_db":-0.13271352734755024,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":3.882143924730595e-05,"delta_db":-0.0673046439095735,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":5.831186316371489e-05,"delta_db":-0.4941044733342842,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.00047020792429263916,"delta_db":-0.4299643229874461,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.00016554707447934098,"delta_db":-0.23798606656409232,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30472099489431703,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.33895774638665477,"pitch":null,"output_rms_dbfs":-30.31353591277852,"output_peak":0.3202393193514826,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.637358170958421},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":47040,"output_samples":47040,"intelligibility":{"estoi":{"value":0.9541188229045146,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03171546561641813,"rms_dbfs":-29.974578166391865,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.207027209586595,"mean":-2.270977513319781e-05,"zero_energy":false},"output":{"rms":0.030501640942115632,"rms_dbfs":-30.31353591277852,"peak":0.3202393193514826,"peak_dbfs":-9.890506919351141,"crest_factor_db":20.423028993427373,"mean":9.405098087089386e-05,"zero_energy":false},"rms_difference_db":-0.33895774638665443},"full_scale":{"reference":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":294,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":1.8029614299530294e-07,"output_rms_dbfs":-67.44013563869798,"output_zero_energy":false,"output_peak":0.002017718272902584,"output_energy_fraction":7.909953729705464e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":287,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":64,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.8648648648648649,"absolute_difference_cents_median":0.6836798410759821,"absolute_difference_cents_p95":3.868530370979847,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":47040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.0001280716783512989,"log_rms_mae_db":1.0399105229499452,"rms_envelope_nrmse":0.10771941266114023,"envelope_correlation":0.9887291876078486,"rms_gain_error_db":-0.3389030943150082,"reference_mean_power":0.0010075555569028208,"output_mean_power":0.0009319201306951833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":5.126669358054452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020945501880404742,"output_power":0.0001888401276781471,"reference_active":true,"delta_db":-0.44996485667032926},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010240530560263733,"output_power":0.00010269307100599751,"reference_active":true,"delta_db":0.012186835491945701},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779418868232472e-05,"output_power":1.7149912604335205e-05,"reference_active":true,"delta_db":-2.0969209019365036}],"active_band_count":3,"active_band_mae_db":0.8530241980329262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":0.6195536585076522,"rms_envelope_nrmse":0.06554914019915553,"envelope_correlation":0.9964071727649437,"rms_gain_error_db":-0.33882979455117435,"reference_mean_power":0.001017821671515535,"output_mean_power":0.0009414314753599599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":4.336511567866126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018218362838522706,"output_power":0.00016444628993931057,"reference_active":true,"delta_db":-0.4448526742922759},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6513405903536415e-05,"output_power":5.416109406644361e-05,"reference_active":true,"delta_db":-0.18464053176609427},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.038210274571303e-06,"output_power":1.2499908691862178e-06,"reference_active":true,"delta_db":-2.1234214590088705}],"active_band_count":3,"active_band_mae_db":0.9176382216890802}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":0.8436852614565215,"rms_envelope_nrmse":0.10122792060393823,"envelope_correlation":0.9919226623108055,"rms_gain_error_db":-0.1554460594669047,"reference_mean_power":0.00014923818668565985,"output_mean_power":0.000143991003225848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":7.990988889403065e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7339937941838396e-05,"output_power":4.645523058132022e-05,"reference_active":true,"delta_db":-0.08193064228826094},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462900675807451e-06,"output_power":6.522179851169659e-06,"reference_active":true,"delta_db":0.039652888631288255},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2961437099516356e-06,"output_power":1.0084643627167682e-06,"reference_active":true,"delta_db":-1.0899260096234824}],"active_band_count":3,"active_band_mae_db":0.40383651351434385}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":0.4755226384670559,"rms_envelope_nrmse":0.05832139942843003,"envelope_correlation":0.9973439213179314,"rms_gain_error_db":-0.1554459596853622,"reference_mean_power":0.00015077911609921791,"output_mean_power":0.00014547775722564151},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":7.46020344966024e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.50390789043777e-05,"output_power":4.4274127681899975e-05,"reference_active":true,"delta_db":-0.07439486974233255},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3028613860270352e-06,"output_power":3.2822499482805212e-06,"reference_active":true,"delta_db":-0.027186975148471623},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.84375886973483e-08,"output_power":6.412660966400914e-08,"reference_active":true,"delta_db":-1.3959861314335478}],"active_band_count":3,"active_band_mae_db":0.4991893254414506}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":1.0547493782648611,"rms_envelope_nrmse":0.12334540274401222,"envelope_correlation":0.9879357072479924,"rms_gain_error_db":-0.42651076273786337,"reference_mean_power":0.00016484395516825918,"output_mean_power":0.00014942454769537714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":1.0813916284336478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.944970033970201e-05,"output_power":4.1651068014146125e-05,"reference_active":true,"delta_db":-0.7453752212131748},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1861966099880329e-05,"output_power":1.1283626943036694e-05,"reference_active":true,"delta_db":-0.21707959370506516},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.815426369634167e-06,"output_power":4.205175538244977e-06,"reference_active":true,"delta_db":-0.5885061722680878}],"active_band_count":3,"active_band_mae_db":0.5169869957287759}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":0.6993967917493985,"rms_envelope_nrmse":0.08349446811741731,"envelope_correlation":0.9951616854913807,"rms_gain_error_db":-0.42653705595395097,"reference_mean_power":0.00016654362747028518,"output_mean_power":0.00015096431966457523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":9.377377349368442e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3529785361500974e-05,"output_power":3.64459824865688e-05,"reference_active":true,"delta_db":-0.7713686358079443},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5942380285921235e-06,"output_power":5.583729327527409e-06,"reference_active":true,"delta_db":-0.008165835672175112},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1488780558632e-07,"output_power":2.6692648737152213e-07,"reference_active":true,"delta_db":-0.7176417099548467}],"active_band_count":3,"active_band_mae_db":0.4990587271449887}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.00011951234204294975,"log_rms_mae_db":1.2126203902620458,"rms_envelope_nrmse":0.13900107402333753,"envelope_correlation":0.9859719020759831,"rms_gain_error_db":-0.38610339298466273,"reference_mean_power":0.0005186016960204269,"output_mean_power":0.000474486208254262},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":3.869293609534798e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017018269823121555,"output_power":0.00015997918639446647,"reference_active":true,"delta_db":-0.26851921306579396},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010605303684284745,"output_power":0.00010188828197764331,"reference_active":true,"delta_db":-0.1739886980607219},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.258811740016227e-05,"output_power":1.2718063612085017e-05,"reference_active":true,"delta_db":-2.494590434172077}],"active_band_count":3,"active_band_mae_db":0.9790327817661977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":0.7912191810277964,"rms_envelope_nrmse":0.08842028968569442,"envelope_correlation":0.9946590444972602,"rms_gain_error_db":-0.38613089680828705,"reference_mean_power":0.0005239212873916427,"output_mean_power":0.00047935024634746415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":3.3612492298744625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001684428838944303,"output_power":0.00015846637591116096,"reference_active":true,"delta_db":-0.2651554250009489},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8728944802248194e-05,"output_power":5.398339913670121e-05,"reference_active":true,"delta_db":-0.3659197082209508},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9018720843055647e-06,"output_power":1.0742371220935885e-06,"reference_active":true,"delta_db":-2.4808114781648776}],"active_band_count":3,"active_band_mae_db":1.0372955371289259}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-75918-0013","method":"magphase","status":"measured","mos":{"status":"measured","value":2.884704828262329,"predicted_mos":2.884704828262329,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":141794,"input_sample_rate":48000,"input_seconds":2.9540416666666665,"input_float32_sha256":"7a7f0f77b01943ceb4e0f6c31d1e47986686fdb90b49e64546b1b7f007f6c080","internal_samples":47265,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.498233889999028},"started_at":"2026-10-13T05:17:27.778260+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.005365660999814281,"native_input_samples":141120,"native_input_sha256":"9dbaf6293d3fa23368b9dd87eeed3531c8f826f8abc0c0463efd80107d9aa4d9","output_sha256":"2f2f497bf5ca22eb2ddefaf3e2ac242d42ddd3e3c8a2c2d671ab889055236adc","output_samples":141794,"length_delta_samples":674,"analysis_s":1.8139304220003396,"synthesis_s":0.7036929549994966,"total_s":2.517623376999836,"rtf":0.8563344819727334,"output_peak":0.37284323264982067,"output_rms":0.031500127891994856,"finished_at":"2026-10-13T05:17:30.308360+00:00","evaluation_input":{"reference":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"2f2f497bf5ca22eb2ddefaf3e2ac242d42ddd3e3c8a2c2d671ab889055236adc","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":47040,"output_samples":47265,"length_delta_samples":225,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.02885992136548,"output_peak":0.3473355885386619},"common_prefix":{"samples":47040,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9952396064741352,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":47040,"output_samples":47040,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.293258138809961,"spectral_convergence":0.6504860693139134,"spectral_convergence_status":"ok","amplitude_floor":8.904239335448439e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.361264010942457,"spectral_convergence":0.4901395387851444,"spectral_convergence_status":"ok","amplitude_floor":4.6801685305432975e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.882799276320512,"spectral_convergence":0.47911929246237006,"spectral_convergence_status":"ok","amplitude_floor":2.078367899935203e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":4.736571532646871e-05,"output_power":3.586775850695989e-05,"delta_db":-1.2075986344508518,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00015972486768670453,"output_power":0.00014866976171875785,"delta_db":-0.31149891681803066,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":3.942775929744416e-05,"output_power":3.808855507497901e-05,"delta_db":-0.15007598960433727,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":6.533823001233142e-05,"output_power":6.274485937876796e-05,"delta_db":-0.17589215113071704,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0005191422513822901,"output_power":0.0005347302827375055,"delta_db":0.12848403472271577,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017487193556096138,"output_power":0.0001778832636998728,"delta_db":0.0741497200382238,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3412832411274794,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.034183953978139,"pitch":null,"output_rms_dbfs":-30.008762120369997,"output_peak":0.3473355885386619,"reference_rms_dbfs":-29.974578166391858,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.512440475357643},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":47040,"output_samples":47040,"intelligibility":{"estoi":{"value":0.85342815981854,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03171546561641813,"rms_dbfs":-29.974578166391865,"peak":0.408905029296875,"peak_dbfs":-7.767550956805262,"crest_factor_db":22.207027209586595,"mean":-2.270977513319781e-05,"zero_energy":false},"output":{"rms":0.03159089237607642,"rms_dbfs":-30.008762120369997,"peak":0.3473355885386619,"peak_dbfs":-9.185014318575144,"crest_factor_db":20.823747801794855,"mean":0.000672745076273936,"zero_energy":false},"rms_difference_db":-0.034183953978132564},"full_scale":{"reference":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":47040,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":294,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.2130772462114693e-06,"threshold_dbfs":-59.16111543292334,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":12,"quiet_samples":1920,"reference_mean_power":2.6578006024161974e-07,"output_mean_power":2.822186878646839e-07,"output_rms_dbfs":-65.49414231659657,"output_zero_energy":false,"output_peak":0.004000520198243644,"output_energy_fraction":1.1542394030050276e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":287,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":62,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.8378378378378378,"absolute_difference_cents_median":4.126558773728532,"absolute_difference_cents_p95":40.93935977097942,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":47040,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.0001280716783512989,"log_rms_mae_db":2.3808426456678307,"rms_envelope_nrmse":0.2153858210442363,"envelope_correlation":0.9516310917485479,"rms_gain_error_db":-0.0341731656900433,"reference_mean_power":0.0010075555569028208,"output_mean_power":0.000999658552762775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":5.126669358054452e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020945501880404742,"output_power":0.0002299451206250207,"reference_active":true,"delta_db":0.40533427332631405},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010240530560263733,"output_power":0.00011423363458783869,"reference_active":true,"delta_db":0.47471537072940934},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.779418868232472e-05,"output_power":2.3097133706168864e-05,"reference_active":true,"delta_db":-0.8039591299311204}],"active_band_count":3,"active_band_mae_db":0.5613362579956146}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":9.465799671406843e-05,"log_rms_mae_db":1.572236252880629,"rms_envelope_nrmse":0.12881299898979415,"envelope_correlation":0.9804529306120016,"rms_gain_error_db":-0.03375844804782929,"reference_mean_power":0.001017821671515535,"output_mean_power":0.001009940640689035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":4.336511567866126e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018218362838522706,"output_power":0.00020228479259972061,"reference_active":true,"delta_db":0.45453887261682147},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.6513405903536415e-05,"output_power":6.0943482043741914e-05,"reference_active":true,"delta_db":0.3277578249101048},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.038210274571303e-06,"output_power":1.574811946336901e-06,"reference_active":true,"delta_db":-1.1202028589053528}],"active_band_count":3,"active_band_mae_db":0.6341665188107597}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.812061629421404e-05,"log_rms_mae_db":2.5037104586966215,"rms_envelope_nrmse":0.18767084173216633,"envelope_correlation":0.9716625055123519,"rms_gain_error_db":-0.2798964393785109,"reference_mean_power":0.00014923818668565985,"output_mean_power":0.00013992339035239916},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":7.990988889403065e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.7339937941838396e-05,"output_power":4.50480174959217e-05,"reference_active":true,"delta_db":-0.21552001268686308},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.462900675807451e-06,"output_power":6.039437844370159e-06,"reference_active":true,"delta_db":-0.2943096554939766},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2961437099516356e-06,"output_power":1.240599239389075e-06,"reference_active":true,"delta_db":-0.1902164604761824}],"active_band_count":3,"active_band_mae_db":0.2333487095523407}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.488253980944051e-05,"log_rms_mae_db":1.7444514863259366,"rms_envelope_nrmse":0.11790791759711468,"envelope_correlation":0.9886011419381332,"rms_gain_error_db":-0.2798922059035057,"reference_mean_power":0.00015077911609921791,"output_mean_power":0.00014136827948145548},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":7.46020344966024e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.50390789043777e-05,"output_power":4.289253543316487e-05,"reference_active":true,"delta_db":-0.21207781458410316},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3028613860270352e-06,"output_power":3.15553573286151e-06,"reference_active":true,"delta_db":-0.19817245239750644},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.84375886973483e-08,"output_power":8.078726979681761e-08,"reference_active":true,"delta_db":-0.39293961363563545}],"active_band_count":3,"active_band_mae_db":0.26772996020574835}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":4.5882621361725573e-05,"log_rms_mae_db":2.4590201719599993,"rms_envelope_nrmse":0.2611308191602057,"envelope_correlation":0.9391941225260745,"rms_gain_error_db":-0.027276010987062914,"reference_mean_power":0.00016484395516825918,"output_mean_power":0.00016381189151193912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":1.0813916284336478e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.944970033970201e-05,"output_power":4.9519139013684513e-05,"reference_active":true,"delta_db":0.006094208659367137},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1861966099880329e-05,"output_power":1.2015560218537519e-05,"reference_active":true,"delta_db":0.05587345804749716},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.815426369634167e-06,"output_power":4.423913406270337e-06,"reference_active":true,"delta_db":-0.3682812916431365}],"active_band_count":3,"active_band_mae_db":0.14341631945000025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":4.018253092677302e-05,"log_rms_mae_db":1.6155474196326465,"rms_envelope_nrmse":0.1390547595115555,"envelope_correlation":0.9807465810235828,"rms_gain_error_db":-0.027224309590524867,"reference_mean_power":0.00016654362747028518,"output_mean_power":0.00016550289266504262},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":9.377377349368442e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3529785361500974e-05,"output_power":4.330499957601579e-05,"reference_active":true,"delta_db":-0.022484870407797702},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5942380285921235e-06,"output_power":5.550307056676482e-06,"reference_active":true,"delta_db":-0.03423931145490158},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1488780558632e-07,"output_power":3.0022805534748766e-07,"reference_active":true,"delta_db":-0.20704569151570662}],"active_band_count":3,"active_band_mae_db":0.0879232911261353}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":587,"rms_floor":0.00011951234204294975,"log_rms_mae_db":2.9324657007648685,"rms_envelope_nrmse":0.26689363834899316,"envelope_correlation":0.9463874977099416,"rms_gain_error_db":0.1085671294219433,"reference_mean_power":0.0005186016960204269,"output_mean_power":0.0005317293657252125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":587,"frequency_resolution_hz":0.34071550255536626,"power_floor":3.869293609534798e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017018269823121555,"output_power":0.0001937042105048407,"reference_active":true,"delta_db":0.5622365590825305},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010605303684284745,"output_power":0.00011243702946361562,"reference_active":true,"delta_db":0.25386254108607265},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.258811740016227e-05,"output_power":1.9356019598993383e-05,"reference_active":true,"delta_db":-0.6706398310300177}],"active_band_count":3,"active_band_mae_db":0.4955796437328736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":47040,"uncovered_tail_samples":0,"metrics":{"frame_count":581,"rms_floor":8.50019565596005e-05,"log_rms_mae_db":1.9290377674966566,"rms_envelope_nrmse":0.16818146633174244,"envelope_correlation":0.97626818595247,"rms_gain_error_db":0.1087227635299737,"reference_mean_power":0.0005239212873916427,"output_mean_power":0.0005372028659398031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":581,"frequency_resolution_hz":0.3442340791738382,"power_floor":3.3612492298744625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001684428838944303,"output_power":0.0001918559668705384,"reference_active":true,"delta_db":0.565226423162658},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8728944802248194e-05,"output_power":6.023297239296457e-05,"reference_active":true,"delta_db":0.10982097545963096},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9018720843055647e-06,"output_power":1.3789660435402984e-06,"reference_active":true,"delta_db":-1.396277319125836}],"active_band_count":3,"active_band_mae_db":0.6904415725827083}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"hnm","status":"measured","mos":{"status":"measured","value":1.9852142333984375,"predicted_mos":1.9852142333984375,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64800,"input_sample_rate":16000,"input_seconds":4.05,"input_float32_sha256":"4dfd2a63e5c961b5b939244fc11e3092553a482d62891f5effd94550033a28c8","internal_samples":64800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.596502013000645},"started_at":"2026-10-13T04:39:56.045104+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.1609000466705766e-05,"native_input_samples":64800,"native_input_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","output_sha256":"3033fac6c8e3836ce8be811f57b97a8cb632ded09b6509e6699ee0f90e9a5039","output_samples":64800,"length_delta_samples":0,"analysis_s":0.6639302319999842,"synthesis_s":0.2960884320000332,"total_s":0.9600186640000175,"rtf":0.2370416454321031,"output_peak":0.4546380206545812,"output_rms":0.03532806623522811,"finished_at":"2026-10-13T04:39:57.008736+00:00","evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"3033fac6c8e3836ce8be811f57b97a8cb632ded09b6509e6699ee0f90e9a5039","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.157415436104894,"spectral_convergence":0.6189817202198741,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.841528665804622,"spectral_convergence":0.5801058377589344,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4515800716466,"spectral_convergence":0.611153603228024,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":0.000162764903948214,"delta_db":-1.7079388930668593,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.00010627720286796804,"delta_db":0.6307153914786889,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":6.535612469151741e-05,"delta_db":-0.5601997157012925,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":2.6190389360312212e-05,"delta_db":0.351467810633044,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0006338514288643961,"delta_db":0.08865683085337933,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00025363208290910135,"delta_db":0.42260808947105977,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6269311218673873,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11907262181862005,"pitch":null,"output_rms_dbfs":-29.037602680883708,"output_peak":0.4546380206545812,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.150174724518705},"common_prefix":{"samples":64800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.157415436104894,"spectral_convergence":0.6189817202198741,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.841528665804622,"spectral_convergence":0.5801058377589344,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4515800716466,"spectral_convergence":0.611153603228024,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":0.000162764903948214,"delta_db":-1.7079388930668593,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.00010627720286796804,"delta_db":0.6307153914786889,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":6.535612469151741e-05,"delta_db":-0.5601997157012925,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":2.6190389360312212e-05,"delta_db":0.351467810633044,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0006338514288643961,"delta_db":0.08865683085337933,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00025363208290910135,"delta_db":0.42260808947105977,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6269311218673873,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11907262181862005,"pitch":null,"output_rms_dbfs":-29.037602680883708,"output_peak":0.4546380206545812,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.150174724518705},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64800,"output_samples":64800,"intelligibility":{"estoi":{"value":0.7584728670627029,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03581570440383133,"rms_dbfs":-28.918530059065088,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.359480807956004,"mean":-3.073421525366512e-06,"zero_energy":false},"output":{"rms":0.03532806623522811,"rms_dbfs":-29.037602680883708,"peak":0.4546380206545812,"peak_dbfs":-6.846684954710144,"crest_factor_db":22.190917726173566,"mean":7.117710537887365e-05,"zero_energy":false},"rms_difference_db":-0.11907262181862066},"full_scale":{"reference":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":405,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":1.1886033790484819e-06,"output_rms_dbfs":-59.24963039431987,"output_zero_energy":false,"output_peak":0.0051033926967387995,"output_energy_fraction":1.1757424802170572e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":398,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":32,"both_estimated_voiced_frames":32,"reference_voiced_output_unvoiced_frames":36,"reference_unvoiced_output_voiced_frames":0,"paired_voiced_fraction":0.47058823529411764,"absolute_difference_cents_median":1.9483760804864554,"absolute_difference_cents_p95":3.726872133861332,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.0001629307662014554,"log_rms_mae_db":1.1902424003080259,"rms_envelope_nrmse":0.1538283069355371,"envelope_correlation":0.9739023859454373,"rms_gain_error_db":-0.118636819029666,"reference_mean_power":0.0012840985730840112,"output_mean_power":0.0012494954571324684},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.3397819245367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003947037290166488,"output_power":0.0003778806858018224,"reference_active":true,"delta_db":-0.18916534600477813},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980542214464163e-05,"output_power":5.685301937266227e-05,"reference_active":true,"delta_db":-0.8913562349045752},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.603685703735387e-05,"output_power":2.5020819725821497e-05,"reference_active":true,"delta_db":-0.17287024641141063}],"active_band_count":3,"active_band_mae_db":0.417797275773588}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.000144561944924798,"log_rms_mae_db":0.6103789064135359,"rms_envelope_nrmse":0.09208868601961599,"envelope_correlation":0.9898518894856272,"rms_gain_error_db":-0.11856282214637157,"reference_mean_power":0.0012928029739092497,"output_mean_power":0.0012579867309136994},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":6.614408925582191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037454976712341086,"output_power":0.00035193154809281055,"reference_active":true,"delta_db":-0.27051331475698254},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9110632304140427e-05,"output_power":2.3260285164528615e-05,"reference_active":true,"delta_db":-0.9743660398099286},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1315453531682465e-06,"output_power":1.351144448990521e-06,"reference_active":true,"delta_db":-1.979927962932666}],"active_band_count":3,"active_band_mae_db":1.074935772499859}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":1.9239259805664122,"rms_envelope_nrmse":0.17875934887504258,"envelope_correlation":0.9698617262686222,"rms_gain_error_db":-0.2420237881865062,"reference_mean_power":0.00013419941652038683,"output_mean_power":0.00012692531466976427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.237716340453255e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0020888510118014e-05,"output_power":1.7482275877015256e-05,"reference_active":true,"delta_db":-0.5888537790565179},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.868636044401025e-06,"output_power":6.386429874216455e-06,"reference_active":true,"delta_db":-0.9064131013859871},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3701828329102724e-06,"output_power":1.5042600269953459e-06,"reference_active":true,"delta_db":0.40544393144285645}],"active_band_count":3,"active_band_mae_db":0.6335702706284538}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":1.2491060955526052,"rms_envelope_nrmse":0.09891819790955658,"envelope_correlation":0.9906068307727042,"rms_gain_error_db":-0.24202418192482988,"reference_mean_power":0.00013520032575819994,"output_mean_power":0.00012787195935328903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":4.6212903051863556e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8157545163089018e-05,"output_power":1.565176099602577e-05,"reference_active":true,"delta_db":-0.6449392549271162},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868319726245688e-06,"output_power":2.9504622242291166e-06,"reference_active":true,"delta_db":-1.1763230382274938},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2646920446881674e-07,"output_power":1.2117181375741247e-07,"reference_active":true,"delta_db":-0.1858317805270003}],"active_band_count":3,"active_band_mae_db":0.6690313578938701}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":1.805055764013792,"rms_envelope_nrmse":0.2022253262800224,"envelope_correlation":0.9666506328618263,"rms_gain_error_db":0.38293861507056826,"reference_mean_power":0.0001750632057906084,"output_mean_power":0.00019120037450448416},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":9.028764248682411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.277369850023164e-05,"output_power":4.998347700091453e-05,"reference_active":true,"delta_db":0.6764965978315574},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.603092418254851e-06,"output_power":7.729522013101932e-06,"reference_active":true,"delta_db":0.6840526256535008},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4970630790605323e-06,"output_power":4.739330403591214e-06,"reference_active":true,"delta_db":1.3201352063791587}],"active_band_count":3,"active_band_mae_db":0.8935614766214056}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":1.1316350604384426,"rms_envelope_nrmse":0.1203873160384416,"envelope_correlation":0.9882671645338439,"rms_gain_error_db":0.38308201320295915,"reference_mean_power":0.0001762447957551007,"output_mean_power":0.00019249723826032675},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":7.899096535479964e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7672515224695035e-05,"output_power":4.540775216202054e-05,"reference_active":true,"delta_db":0.8110538631707698},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.050253319949333e-06,"output_power":3.830904747237568e-06,"reference_active":true,"delta_db":0.9896544519902402},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5566787812291305e-07,"output_power":2.803820800140412e-07,"reference_active":true,"delta_db":0.40074086138430715}],"active_band_count":3,"active_band_mae_db":0.7338163921817724}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00015276264453937597,"log_rms_mae_db":1.5775113920046893,"rms_envelope_nrmse":0.1478983156334786,"envelope_correlation":0.9838099411190169,"rms_gain_error_db":0.11554293243329089,"reference_mean_power":0.0006070254152411959,"output_mean_power":0.0006233919185946284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.389907119886333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002863305566547833,"output_power":0.00029918542874948436,"reference_active":true,"delta_db":0.1907274065105175},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.608243104449436e-05,"output_power":3.608198023170672e-05,"reference_active":true,"delta_db":-5.426095843735021e-05},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0810295137794258e-05,"output_power":1.2284010836202354e-05,"reference_active":true,"delta_db":0.5550263985228545}],"active_band_count":3,"active_band_mae_db":0.24860268866393645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.00013215266595876848,"log_rms_mae_db":0.881932674044988,"rms_envelope_nrmse":0.06856241941435667,"envelope_correlation":0.996426540794476,"rms_gain_error_db":0.11628281460822673,"reference_mean_power":0.0006109513084873902,"output_mean_power":0.0006275305606398267},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":5.083168365591418e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026729578490804684,"output_power":0.00028225946415662924,"reference_active":true,"delta_db":0.23656402479486718},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4768224147952764e-05,"output_power":1.4855619081909706e-05,"reference_active":true,"delta_db":0.025624796474229508},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.446503663842911e-07,"output_power":6.655298116583993e-07,"reference_active":true,"delta_db":-1.035094608021743}],"active_band_count":3,"active_band_mae_db":0.43242780976361317}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.287475109100342,"predicted_mos":2.287475109100342,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64800,"input_sample_rate":16000,"input_seconds":4.05,"input_float32_sha256":"83c1dce810f86f56fd683385c933237c47fe7b9e5f9b42699952408772b18600","internal_samples":64800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.423531197000557},"started_at":"2026-10-13T04:43:07.736208+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.030199993896531e-05,"native_input_samples":64800,"native_input_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","output_sha256":"78a81a9ca52a30739677a23cb9e407126fde1662b58d3e3c6c229515d9fda824","output_samples":64800,"length_delta_samples":0,"analysis_s":0.939645214999473,"synthesis_s":0.0613295740004105,"total_s":1.0009747889998835,"rtf":0.24715426888886013,"output_peak":0.489526404671685,"output_rms":0.035186353649113454,"finished_at":"2026-10-13T04:43:08.741019+00:00","evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"78a81a9ca52a30739677a23cb9e407126fde1662b58d3e3c6c229515d9fda824","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.010349366926082,"spectral_convergence":0.7450110861457179,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.637423334776295,"spectral_convergence":0.6504404691247228,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.042375991866162,"spectral_convergence":0.6626341712933816,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":5.6163722050252546e-05,"delta_db":-6.328987741628598,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.00011065971415039766,"delta_db":0.8062096758279786,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":9.664853194404761e-05,"delta_db":1.1388899738667682,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":3.018566414434199e-05,"delta_db":0.9680556176901292,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0007244925102492666,"delta_db":0.6691210953398686,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00021992931939256134,"delta_db":-0.19660249104186125,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6846444325658674,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1539846736370159,"pitch":null,"output_rms_dbfs":-29.072514732702107,"output_peak":0.489526404671685,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.230049564522846},"common_prefix":{"samples":64800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.010349366926082,"spectral_convergence":0.7450110861457179,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.637423334776295,"spectral_convergence":0.6504404691247228,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.042375991866162,"spectral_convergence":0.6626341712933816,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":5.6163722050252546e-05,"delta_db":-6.328987741628598,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.00011065971415039766,"delta_db":0.8062096758279786,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":9.664853194404761e-05,"delta_db":1.1388899738667682,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":3.018566414434199e-05,"delta_db":0.9680556176901292,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0007244925102492666,"delta_db":0.6691210953398686,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00021992931939256134,"delta_db":-0.19660249104186125,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6846444325658674,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1539846736370159,"pitch":null,"output_rms_dbfs":-29.072514732702107,"output_peak":0.489526404671685,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.230049564522846},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64800,"output_samples":64800,"intelligibility":{"estoi":{"value":0.8465789416638694,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03581570440383133,"rms_dbfs":-28.918530059065088,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.359480807956004,"mean":-3.073421525366512e-06,"zero_energy":false},"output":{"rms":0.035186353649113454,"rms_dbfs":-29.072514732702107,"peak":0.489526404671685,"peak_dbfs":-6.204477554482145,"crest_factor_db":22.86803717821996,"mean":5.755880241679392e-09,"zero_energy":false},"rms_difference_db":-0.1539846736370194},"full_scale":{"reference":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":405,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":7.900384886670201e-07,"output_rms_dbfs":-61.02351750470913,"output_zero_energy":false,"output_peak":0.004147930329754143,"output_energy_fraction":7.877976916230502e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":398,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":76,"both_estimated_voiced_frames":51,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.75,"absolute_difference_cents_median":3.8304339338811597,"absolute_difference_cents_p95":15.23896347719247,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.0001629307662014554,"log_rms_mae_db":2.4864408151224997,"rms_envelope_nrmse":0.26386965217345076,"envelope_correlation":0.9234289822125634,"rms_gain_error_db":-0.15348335028947188,"reference_mean_power":0.0012840985730840112,"output_mean_power":0.0012395099813651725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.3397819245367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003947037290166488,"output_power":0.0003448793398262208,"reference_active":true,"delta_db":-0.5860405095762412},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980542214464163e-05,"output_power":7.08871297322408e-05,"reference_active":true,"delta_db":0.06678234021154886},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.603685703735387e-05,"output_power":3.52592729165677e-05,"reference_active":true,"delta_db":1.3168479396639001}],"active_band_count":3,"active_band_mae_db":0.65655692981723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.000144561944924798,"log_rms_mae_db":1.7735464752066146,"rms_envelope_nrmse":0.18173058363957176,"envelope_correlation":0.9599242163924376,"rms_gain_error_db":-0.15307747342374786,"reference_mean_power":0.0012928029739092497,"output_mean_power":0.0012480287646846587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":6.614408925582191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037454976712341086,"output_power":0.00033031787097450677,"reference_active":true,"delta_db":-0.5457746070320662},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9110632304140427e-05,"output_power":3.125660903580786e-05,"reference_active":true,"delta_db":0.30890221804215356},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1315453531682465e-06,"output_power":2.360387070325337e-06,"reference_active":true,"delta_db":0.4428864938328803}],"active_band_count":3,"active_band_mae_db":0.4325211063023667}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":2.606092989553254,"rms_envelope_nrmse":0.221461013169567,"envelope_correlation":0.9698441990685982,"rms_gain_error_db":0.9474988459140234,"reference_mean_power":0.00013419941652038683,"output_mean_power":0.00016691697757215173},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.237716340453255e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0020888510118014e-05,"output_power":2.649502219182982e-05,"reference_active":true,"delta_db":1.2168094041579773},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.868636044401025e-06,"output_power":9.721915979210702e-06,"reference_active":true,"delta_db":0.9185240546976926},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3701828329102724e-06,"output_power":1.458338059800809e-06,"reference_active":true,"delta_db":0.27079688267514235}],"active_band_count":3,"active_band_mae_db":0.8020434471769375}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":2.1161864633894885,"rms_envelope_nrmse":0.16651851987121433,"envelope_correlation":0.9865249878494052,"rms_gain_error_db":0.9475063871031759,"reference_mean_power":0.00013520032575819994,"output_mean_power":0.00016816219856806118},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":4.6212903051863556e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8157545163089018e-05,"output_power":2.3769478742538907e-05,"reference_active":true,"delta_db":1.1696252483501064},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868319726245688e-06,"output_power":4.7126128823969784e-06,"reference_active":true,"delta_db":0.8573940368949566},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2646920446881674e-07,"output_power":1.3354595589920747e-07,"reference_active":true,"delta_db":0.23645954331255986}],"active_band_count":3,"active_band_mae_db":0.754492942852541}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":1.7439051163130468,"rms_envelope_nrmse":0.21454343692218883,"envelope_correlation":0.9631825873038911,"rms_gain_error_db":0.46870613733044125,"reference_mean_power":0.0001750632057906084,"output_mean_power":0.0001950138654958274},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":9.028764248682411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.277369850023164e-05,"output_power":4.40385410081652e-05,"reference_active":true,"delta_db":0.12656118523666926},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.603092418254851e-06,"output_power":1.1607069507687211e-05,"reference_active":true,"delta_db":2.4497520938303827},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4970630790605323e-06,"output_power":6.511025222825652e-06,"reference_active":true,"delta_db":2.6994591168374478}],"active_band_count":3,"active_band_mae_db":1.7585907986348335}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":1.096074726694233,"rms_envelope_nrmse":0.13560061940972,"envelope_correlation":0.9855838222076319,"rms_gain_error_db":0.46915538516882355,"reference_mean_power":0.0001762447957551007,"output_mean_power":0.00019635042258895724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":7.899096535479964e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7672515224695035e-05,"output_power":3.993995910383084e-05,"reference_active":true,"delta_db":0.2538299877902044},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.050253319949333e-06,"output_power":6.358578111868316e-06,"reference_active":true,"delta_db":3.190241022511703},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5566787812291305e-07,"output_power":5.035600102868061e-07,"reference_active":true,"delta_db":2.943750666500443}],"active_band_count":3,"active_band_mae_db":2.12927389226745}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00015276264453937597,"log_rms_mae_db":1.8143019993279148,"rms_envelope_nrmse":0.25296693704636947,"envelope_correlation":0.956216672685451,"rms_gain_error_db":0.4586583567444469,"reference_mean_power":0.0006070254152411959,"output_mean_power":0.0006746409678846791},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.389907119886333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002863305566547833,"output_power":0.0003057942662258653,"reference_active":true,"delta_db":0.2856164036464899},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.608243104449436e-05,"output_power":5.4722670227376415e-05,"reference_active":true,"delta_db":1.8087149056913134},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0810295137794258e-05,"output_power":2.4125692020442154e-05,"reference_active":true,"delta_db":3.486422284541539}],"active_band_count":3,"active_band_mae_db":1.8602511979597807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.00013215266595876848,"log_rms_mae_db":1.0592939982332017,"rms_envelope_nrmse":0.14636897603793156,"envelope_correlation":0.9858700183060243,"rms_gain_error_db":0.4595608484472332,"reference_mean_power":0.0006109513084873902,"output_mean_power":0.0006791452759260378},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":5.083168365591418e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026729578490804684,"output_power":0.0002977713307824817,"reference_active":true,"delta_db":0.46890771606251647},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4768224147952764e-05,"output_power":2.181106962983189e-05,"reference_active":true,"delta_db":1.6934868886920105},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.446503663842911e-07,"output_power":1.4901188691797125e-06,"reference_active":true,"delta_db":2.465439393829397}],"active_band_count":3,"active_band_mae_db":1.542611332861308}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"facodec","status":"measured","mos":{"status":"measured","value":4.027527809143066,"predicted_mos":4.027527809143066,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64800,"input_sample_rate":16000,"input_seconds":4.05,"input_float32_sha256":"889bd44061b04934285b0aa710564ecf071343bf0e67985c3769071979b34d20","internal_samples":64800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.905518944000505},"started_at":"2026-10-13T06:57:46.304712+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.279699947393965e-05,"native_input_samples":64800,"native_input_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","output_sha256":"b06f09816e5500b0e32be3dbe88b5066763cf9384e0af2f2b1625d594b0e0e48","output_samples":64800,"length_delta_samples":0,"analysis_s":1.1198054569995293,"synthesis_s":7.630043737999586,"total_s":8.749849194999115,"rtf":2.160456591357806,"output_peak":0.4245581030845642,"output_rms":0.03275356508621737,"finished_at":"2026-10-13T06:57:55.062814+00:00","evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"b06f09816e5500b0e32be3dbe88b5066763cf9384e0af2f2b1625d594b0e0e48","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.535649594600546,"spectral_convergence":0.472161337122755,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.877619837689522,"spectral_convergence":0.513121119856412,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.158408719780504,"spectral_convergence":0.5585324967554203,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":0.00020186857354044795,"delta_db":-0.7728594136790706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.00010651665897907723,"delta_db":0.6404895936766155,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":6.793483813889148e-05,"delta_db":-0.39213719260572266,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":2.2509212720167047e-05,"delta_db":-0.3063486843508527,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0005280785208498016,"delta_db":-0.7042328730720409,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00014588822159009057,"delta_db":-1.9792314916267606,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7992165415018437,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7762983807875025,"pitch":null,"output_rms_dbfs":-29.694828439852593,"output_peak":0.4245581030845642,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8572260506901905},"common_prefix":{"samples":64800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.535649594600546,"spectral_convergence":0.472161337122755,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.877619837689522,"spectral_convergence":0.513121119856412,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.158408719780504,"spectral_convergence":0.5585324967554203,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":0.00020186857354044795,"delta_db":-0.7728594136790706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.00010651665897907723,"delta_db":0.6404895936766155,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":6.793483813889148e-05,"delta_db":-0.39213719260572266,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":2.2509212720167047e-05,"delta_db":-0.3063486843508527,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0005280785208498016,"delta_db":-0.7042328730720409,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00014588822159009057,"delta_db":-1.9792314916267606,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7992165415018437,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7762983807875025,"pitch":null,"output_rms_dbfs":-29.694828439852593,"output_peak":0.4245581030845642,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8572260506901905},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64800,"output_samples":64800,"intelligibility":{"estoi":{"value":0.862763850062548,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03581570440383133,"rms_dbfs":-28.918530059065088,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.359480807956004,"mean":-3.073421525366512e-06,"zero_energy":false},"output":{"rms":0.03275356508621737,"rms_dbfs":-29.694828439852593,"peak":0.4245581030845642,"peak_dbfs":-7.441257315833583,"crest_factor_db":22.25357112401901,"mean":-0.00020632250335691222,"zero_energy":false},"rms_difference_db":-0.7762983807875052},"full_scale":{"reference":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":405,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":2.087672741786125e-06,"output_rms_dbfs":-56.80337579221283,"output_zero_energy":false,"output_peak":0.004831833299249411,"output_energy_fraction":2.4024825718685886e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":398,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":85,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9558823529411765,"absolute_difference_cents_median":3.370933576187122,"absolute_difference_cents_p95":12.985255872008018,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.0001629307662014554,"log_rms_mae_db":1.4759849122798134,"rms_envelope_nrmse":0.17480474706348695,"envelope_correlation":0.9720536454505122,"rms_gain_error_db":-0.77594706319705,"reference_mean_power":0.0012840985730840112,"output_mean_power":0.0010739984561068353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.3397819245367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003947037290166488,"output_power":0.00030008058999057933,"reference_active":true,"delta_db":-1.1903332445754609},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980542214464163e-05,"output_power":7.897101815857599e-05,"reference_active":true,"delta_db":0.5357857951674715},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.603685703735387e-05,"output_power":2.499043009972619e-05,"reference_active":true,"delta_db":-0.17814827829054733}],"active_band_count":3,"active_band_mae_db":0.6347557726778265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.000144561944924798,"log_rms_mae_db":1.200773424619858,"rms_envelope_nrmse":0.14675972090593165,"envelope_correlation":0.9816786389313911,"rms_gain_error_db":-0.7755949761071195,"reference_mean_power":0.0012928029739092497,"output_mean_power":0.0010813663345530385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":6.614408925582191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037454976712341086,"output_power":0.00029202432440902096,"reference_active":true,"delta_db":-1.0809050350895577},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9110632304140427e-05,"output_power":3.277297710088004e-05,"reference_active":true,"delta_db":0.5146425579955584},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1315453531682465e-06,"output_power":2.183579328347595e-06,"reference_active":true,"delta_db":0.10474396624074755}],"active_band_count":3,"active_band_mae_db":0.5667638531086211}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":1.8305742067290858,"rms_envelope_nrmse":0.16559032002189297,"envelope_correlation":0.9747655385349548,"rms_gain_error_db":0.11942486567189568,"reference_mean_power":0.00013419941652038683,"output_mean_power":0.00013794091871542193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.237716340453255e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0020888510118014e-05,"output_power":2.3469477591186327e-05,"reference_active":true,"delta_db":0.6902007552043615},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.868636044401025e-06,"output_power":8.99083656258606e-06,"reference_active":true,"delta_db":0.5790064510223683},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3701828329102724e-06,"output_power":1.345382558232314e-06,"reference_active":true,"delta_db":-0.0793272880406156}],"active_band_count":3,"active_band_mae_db":0.44951149808911506}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":1.3976054414256187,"rms_envelope_nrmse":0.1375383819796663,"envelope_correlation":0.9807013579193682,"rms_gain_error_db":0.11944470327822704,"reference_mean_power":0.00013520032575819994,"output_mean_power":0.00013897036825461255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":4.6212903051863556e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8157545163089018e-05,"output_power":2.1884241574191616e-05,"reference_active":true,"delta_db":0.8107436713927553},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868319726245688e-06,"output_power":4.505155115807045e-06,"reference_active":true,"delta_db":0.661873862377519},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2646920446881674e-07,"output_power":1.0679309060227876e-07,"reference_active":true,"delta_db":-0.7344163150049274}],"active_band_count":3,"active_band_mae_db":0.7356779495917339}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":1.8707039454371743,"rms_envelope_nrmse":0.19560609840568466,"envelope_correlation":0.9699154548187833,"rms_gain_error_db":-0.7650923617629439,"reference_mean_power":0.0001750632057906084,"output_mean_power":0.00014678633968967996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":9.028764248682411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.277369850023164e-05,"output_power":3.101193310713531e-05,"reference_active":true,"delta_db":-1.3964796536425506},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.603092418254851e-06,"output_power":7.369326881842404e-06,"reference_active":true,"delta_db":0.4768044519488395},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4970630790605323e-06,"output_power":3.478048841650213e-06,"reference_active":true,"delta_db":-0.023677896308374198}],"active_band_count":3,"active_band_mae_db":0.6323206672999214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":1.587263588705823,"rms_envelope_nrmse":0.16226013516011897,"envelope_correlation":0.9793685262174002,"rms_gain_error_db":-0.7639519427224718,"reference_mean_power":0.0001762447957551007,"output_mean_power":0.00014781588486103062},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":7.899096535479964e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7672515224695035e-05,"output_power":2.75784105206424e-05,"reference_active":true,"delta_db":-1.354553850284932},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.050253319949333e-06,"output_power":3.543982029280428e-06,"reference_active":true,"delta_db":0.6515560252265057},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5566787812291305e-07,"output_power":2.6651369854566867e-07,"reference_active":true,"delta_db":0.18043369157079545}],"active_band_count":3,"active_band_mae_db":0.7288478556940777}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00015276264453937597,"log_rms_mae_db":2.2501304580807897,"rms_envelope_nrmse":0.248114023978215,"envelope_correlation":0.9603058527784972,"rms_gain_error_db":-1.1721988834687327,"reference_mean_power":0.0006070254152411959,"output_mean_power":0.00046343303275739545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.389907119886333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002863305566547833,"output_power":0.00018592518081798128,"reference_active":true,"delta_db":-1.8752948505497316},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.608243104449436e-05,"output_power":4.248820288945607e-05,"reference_active":true,"delta_db":0.7097257192242019},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0810295137794258e-05,"output_power":1.3636757612281232e-05,"reference_active":true,"delta_db":1.0087357013752976}],"active_band_count":3,"active_band_mae_db":1.1979187570497436}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.00013215266595876848,"log_rms_mae_db":1.886724550607306,"rms_envelope_nrmse":0.21146724890649587,"envelope_correlation":0.9738316779368246,"rms_gain_error_db":-1.1724083408015566,"reference_mean_power":0.0006109513084873902,"output_mean_power":0.0004664077575122418},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":5.083168365591418e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026729578490804684,"output_power":0.0001810145627923115,"reference_active":true,"delta_db":-1.692785945752887},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4768224147952764e-05,"output_power":1.811591010992298e-05,"reference_active":true,"delta_db":0.8873188174130543},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.446503663842911e-07,"output_power":1.0920462940995595e-06,"reference_active":true,"delta_db":1.115640746188906}],"active_band_count":3,"active_band_mae_db":1.2319151697849493}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.3106608390808105,"predicted_mos":2.3106608390808105,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":178688,"input_sample_rate":44100,"input_seconds":4.051882086167801,"input_float32_sha256":"439c9b9a8c2750b34d24f75020cb7e9983143751616a39cae8a511b27b4a9846","internal_samples":64831,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1151555700016615},"started_at":"2026-10-13T07:25:35.532860+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005663190997438505,"native_input_samples":178605,"native_input_sha256":"702e5e6a3c76b1171acce87dff72d0975be67a611bfd4ea92332d2be62cb71b2","output_sha256":"30d1c48ca50ef4e8a66a582f1eccc229ae9c9506775271f561500091a809bb56","output_samples":178688,"length_delta_samples":83,"analysis_s":0.09187877700242097,"synthesis_s":23.459224357000494,"total_s":23.551103134002915,"rtf":5.8150871935809665,"output_peak":0.45377567410469055,"output_rms":0.029614562382323045,"finished_at":"2026-10-13T07:25:59.112304+00:00","evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"30d1c48ca50ef4e8a66a582f1eccc229ae9c9506775271f561500091a809bb56","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":64800,"output_samples":64831,"length_delta_samples":31,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-30.564260173214464,"output_peak":0.4326523171809331},"common_prefix":{"samples":64800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9995218336906727,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.768234456423034,"spectral_convergence":0.6939134182724024,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.961353154863216,"spectral_convergence":0.7218933557393733,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.963747106675152,"spectral_convergence":0.7488408351386806,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":3.4796690769526916e-05,"delta_db":-8.408167118636321,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.0001323610185202022,"delta_db":1.583915239507458,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":6.465989788305863e-05,"delta_db":-0.6067124985547571,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":1.4230776814852716e-05,"delta_db":-2.2976656631647003,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0005079216092799561,"delta_db":-0.8732510052075622,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.0001245973498743816,"delta_db":-2.6643457414939844,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.739009544427464,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.6437193762837659,"pitch":null,"output_rms_dbfs":-30.562249435348857,"output_peak":0.4326523171809331,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.897778239320466},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64800,"output_samples":64800,"intelligibility":{"estoi":{"value":0.5487248697171724,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03581570440383133,"rms_dbfs":-28.918530059065088,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.359480807956004,"mean":-3.073421525366512e-06,"zero_energy":false},"output":{"rms":0.029640636693133916,"rms_dbfs":-30.562249435348857,"peak":0.4326523171809331,"peak_dbfs":-7.2772193193350105,"crest_factor_db":23.285030116013843,"mean":-0.00017621253621206184,"zero_energy":false},"rms_difference_db":-1.6437193762837694},"full_scale":{"reference":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":405,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":1.6432677706919183e-06,"output_rms_dbfs":-57.84291662457478,"output_zero_energy":false,"output_peak":0.005526648057220645,"output_energy_fraction":2.3091293543598794e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":398,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":75,"both_estimated_voiced_frames":48,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.7058823529411765,"absolute_difference_cents_median":12.943028363983391,"absolute_difference_cents_p95":83.40815701613789,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.0001629307662014554,"log_rms_mae_db":2.221348340177915,"rms_envelope_nrmse":0.26403331734860735,"envelope_correlation":0.9482020947421179,"rms_gain_error_db":-1.6432427902841875,"reference_mean_power":0.0012840985730840112,"output_mean_power":0.0008795774452600263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.3397819245367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003947037290166488,"output_power":0.00027279088435192147,"reference_active":true,"delta_db":-1.6044137573609119},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980542214464163e-05,"output_power":4.291021659260407e-05,"reference_active":true,"delta_db":-2.1132845113713152},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.603685703735387e-05,"output_power":1.0313983963891481e-05,"reference_active":true,"delta_db":-4.021621066452625}],"active_band_count":3,"active_band_mae_db":2.5797731117282843}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.000144561944924798,"log_rms_mae_db":2.0722334944285388,"rms_envelope_nrmse":0.2294771158175373,"envelope_correlation":0.9678846424219226,"rms_gain_error_db":-1.6435176866123637,"reference_mean_power":0.0012928029739092497,"output_mean_power":0.000885483705513376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":6.614408925582191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037454976712341086,"output_power":0.00025674541470961795,"reference_active":true,"delta_db":-1.640068352359903},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9110632304140427e-05,"output_power":1.7515275750934298e-05,"reference_active":true,"delta_db":-2.206346597311484},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1315453531682465e-06,"output_power":8.674277196081533e-07,"reference_active":true,"delta_db":-3.9046128122875174}],"active_band_count":3,"active_band_mae_db":2.583675920652968}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":4.215479392830289,"rms_envelope_nrmse":0.37193739248210805,"envelope_correlation":0.8775495606573369,"rms_gain_error_db":0.3976134409114563,"reference_mean_power":0.00013419941652038683,"output_mean_power":0.00014706589547497332},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.237716340453255e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0020888510118014e-05,"output_power":1.5899332591918543e-05,"reference_active":true,"delta_db":-1.0010445291867978},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.868636044401025e-06,"output_power":7.474504430268275e-06,"reference_active":true,"delta_db":-0.22317054293079341},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3701828329102724e-06,"output_power":1.5896757162228876e-06,"reference_active":true,"delta_db":0.6453001811195884}],"active_band_count":3,"active_band_mae_db":0.6231717510790599}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":3.7387556977791485,"rms_envelope_nrmse":0.3146110773282474,"envelope_correlation":0.9026710916710359,"rms_gain_error_db":0.39714747875824274,"reference_mean_power":0.00013520032575819994,"output_mean_power":0.00014814687191561473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":4.6212903051863556e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8157545163089018e-05,"output_power":1.4563467790751687e-05,"reference_active":true,"delta_db":-0.9579233341305884},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868319726245688e-06,"output_power":3.4269534015865146e-06,"reference_active":true,"delta_db":-0.5261416331521471},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2646920446881674e-07,"output_power":1.4341127048850819e-07,"reference_active":true,"delta_db":0.5459849654322039}],"active_band_count":3,"active_band_mae_db":0.6766833109049798}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":3.228281423294613,"rms_envelope_nrmse":0.3195307480276802,"envelope_correlation":0.9107916538985206,"rms_gain_error_db":-0.5807712809431231,"reference_mean_power":0.0001750632057906084,"output_mean_power":0.0001531502637228214},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":9.028764248682411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.277369850023164e-05,"output_power":4.905072407887437e-05,"reference_active":true,"delta_db":0.5946861886056639},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.603092418254851e-06,"output_power":6.249715713948822e-06,"reference_active":true,"delta_db":-0.2388711311478112},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4970630790605323e-06,"output_power":2.9526784808790635e-06,"reference_active":true,"delta_db":-0.7348730713907775}],"active_band_count":3,"active_band_mae_db":0.5228101303814175}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":2.8022811201247406,"rms_envelope_nrmse":0.267085992258381,"envelope_correlation":0.934524922893977,"rms_gain_error_db":-0.5813846751985048,"reference_mean_power":0.0001762447957551007,"output_mean_power":0.00015416217689006984},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":7.899096535479964e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7672515224695035e-05,"output_power":4.473939197370248e-05,"reference_active":true,"delta_db":0.7466546050047769},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.050253319949333e-06,"output_power":2.778797424375438e-06,"reference_active":true,"delta_db":-0.40479020828659756},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5566787812291305e-07,"output_power":2.2163922527121345e-07,"reference_active":true,"delta_db":-0.6202954365848353}],"active_band_count":3,"active_band_mae_db":0.5905800832920699}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00015276264453937597,"log_rms_mae_db":3.2779255051699687,"rms_envelope_nrmse":0.2863649284800974,"envelope_correlation":0.9591596984555039,"rms_gain_error_db":-1.8233551294114232,"reference_mean_power":0.0006070254152411959,"output_mean_power":0.00039890672850209457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.389907119886333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002863305566547833,"output_power":0.000213856249023757,"reference_active":true,"delta_db":-1.26745752423735},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.608243104449436e-05,"output_power":2.8911863714038985e-05,"reference_active":true,"delta_db":-0.9621970230841242},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0810295137794258e-05,"output_power":6.020548598653239e-06,"reference_active":true,"delta_db":-2.542014846008888}],"active_band_count":3,"active_band_mae_db":1.590556464443454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.00013215266595876848,"log_rms_mae_db":3.0245935269547966,"rms_envelope_nrmse":0.24754629443539983,"envelope_correlation":0.9767968563231583,"rms_gain_error_db":-1.8235673741636225,"reference_mean_power":0.0006109513084873902,"output_mean_power":0.00040146700846133805},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":5.083168365591418e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026729578490804684,"output_power":0.00020528139300151285,"reference_active":true,"delta_db":-1.1464252416793164},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4768224147952764e-05,"output_power":1.1119722148957803e-05,"reference_active":true,"delta_db":-1.232343397653644},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.446503663842911e-07,"output_power":5.575305222009947e-07,"reference_active":true,"delta_db":-1.8040832680837358}],"active_band_count":3,"active_band_mae_db":1.3942839691388986}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"golf","status":"measured","mos":{"status":"measured","value":2.1090970039367676,"predicted_mos":2.1090970039367676,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96960,"input_sample_rate":24000,"input_seconds":4.04,"input_float32_sha256":"b1dd5e00cfefb658946d043e218e351b49f6ee225068327b756a9258d2c0b7d4","internal_samples":64640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7281315399995947},"started_at":"2026-10-13T05:02:54.065580+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0020110770001338096,"native_input_samples":97200,"native_input_sha256":"f28a006a950d6646148bad5a2212551ae12b0fc0580f9df55943c49fa26febae","output_sha256":"88d69078e603c9841433434ae68851fa38634bc771c14b6a0022220e88f5e909","output_samples":96960,"length_delta_samples":-240,"analysis_s":0.37677928999983124,"synthesis_s":0.0676178660005462,"total_s":0.44439715600037744,"rtf":0.10972769283959938,"output_peak":0.5014899969100952,"output_rms":0.032195894444562126,"finished_at":"2026-10-13T05:02:54.517999+00:00","analysis_stages":{"external_f0_s":0.04417921200001729},"evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"88d69078e603c9841433434ae68851fa38634bc771c14b6a0022220e88f5e909","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":64800,"output_samples":64640,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.840330748834827,"output_peak":0.4203016554808679},"common_prefix":{"samples":64640,"sample_rate":16000,"reference_coverage":0.9975308641975309,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64640,"output_samples":64640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.333832098185037,"spectral_convergence":0.6743433299391369,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.702186791309364,"spectral_convergence":0.597659187334852,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.931794426297493,"spectral_convergence":0.6020237379663931,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024178533577542895,"output_power":0.00027393638778799303,"delta_db":0.5421976725421163,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.213856013392515e-05,"output_power":0.0001041030818699207,"delta_db":0.5302216539356747,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.453851756477215e-05,"output_power":5.833617128075858e-05,"delta_db":-1.0644282906996787,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.421431282643702e-05,"output_power":2.225916274897311e-05,"delta_db":-0.3656332394045314,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006225212883616579,"output_power":0.00040069368702186305,"delta_db":-1.9134170746710322,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023059157259207807,"output_power":0.00017812091230450398,"delta_db":-1.1212852028414642,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9228638556824161,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9320297926407215,"pitch":null,"output_rms_dbfs":-29.840330748834827,"output_peak":0.4203016554808679,"reference_rms_dbfs":-28.908300956194104,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.989271105263964},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64640,"output_samples":64640,"intelligibility":{"estoi":{"value":0.8367379731111666,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03585790829446552,"rms_dbfs":-28.908300956194104,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.349251705085017,"mean":-3.1617608400854734e-06,"zero_energy":false},"output":{"rms":0.03220946138969127,"rms_dbfs":-29.840330748834827,"peak":0.4203016554808679,"peak_dbfs":-7.528777987909823,"crest_factor_db":22.311552760925007,"mean":-7.774853784418342e-05,"zero_energy":false},"rms_difference_db":-0.9320297926407228},"full_scale":{"reference":{"sample_count":64640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":404,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":3.1636227002127584e-06,"output_rms_dbfs":-54.99815316889379,"output_zero_energy":false,"output_peak":0.004103090307235401,"output_energy_fraction":3.774039117089648e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":397,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":69,"both_estimated_voiced_frames":62,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9117647058823529,"absolute_difference_cents_median":4.899502304923011,"absolute_difference_cents_p95":23.930487637024537,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":0.0001629307662014554,"log_rms_mae_db":2.5267879575541716,"rms_envelope_nrmse":0.2610446760605536,"envelope_correlation":0.9333492465268479,"rms_gain_error_db":-0.9319465396028253,"reference_mean_power":0.0012871498373582904,"output_mean_power":0.0010385668403134977},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":7.372833279359238e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003960941443464865,"output_power":0.0003440659000452166,"reference_active":true,"delta_db":-0.6115678971051814},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.988260714266831e-05,"output_power":6.791950658030048e-05,"reference_active":true,"delta_db":-0.12374576840597765},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6074936473557895e-05,"output_power":3.745320337145907e-05,"reference_active":true,"delta_db":1.572657097922656}],"active_band_count":3,"active_band_mae_db":0.7693235878112716}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":0.000144561944924798,"log_rms_mae_db":1.8916034249901803,"rms_envelope_nrmse":0.17809005990560045,"envelope_correlation":0.9733004072171475,"rms_gain_error_db":-0.9310764734220043,"reference_mean_power":0.0012959267050108511,"output_mean_power":0.0010458581668899801},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":6.645351671909195e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037585267113454435,"output_power":0.0003378761433003241,"reference_active":true,"delta_db":-0.462601127971916},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.904572954043021e-05,"output_power":2.9877116701606168e-05,"reference_active":true,"delta_db":0.12256394364850878},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1239324153519507e-06,"output_power":2.1835182188087595e-06,"reference_active":true,"delta_db":0.12016126782539324}],"active_band_count":3,"active_band_mae_db":0.23510877981527267}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":2.3149024504552314,"rms_envelope_nrmse":0.3160556749680801,"envelope_correlation":0.9089911469899332,"rms_gain_error_db":-0.187778157475087,"reference_mean_power":0.00013453193771591283,"output_mean_power":0.0001288390701532972},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":5.2423948777343856e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0047394262670393e-05,"output_power":1.7583572779219736e-05,"reference_active":true,"delta_db":-0.5695080819090418},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.850487466744024e-06,"output_power":9.5029268850676e-06,"reference_active":true,"delta_db":0.8296076325736557},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.350052020385118e-06,"output_power":2.445816515797846e-06,"reference_active":true,"delta_db":2.580733702089474}],"active_band_count":3,"active_band_mae_db":1.3266164721907239}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":1.6891319371632645,"rms_envelope_nrmse":0.26036681976884213,"envelope_correlation":0.9310742801791424,"rms_gain_error_db":-0.1877493529906149,"reference_mean_power":0.00013553784133615325,"output_mean_power":0.00012980326876504465},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":4.626025199865149e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8176534296205136e-05,"output_power":1.702235909736334e-05,"reference_active":true,"delta_db":-0.2849133224150853},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8436927090084355e-06,"output_power":4.770651123681747e-06,"reference_active":true,"delta_db":0.9382899796846288},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2693428333146716e-07,"output_power":1.611953685427469e-07,"reference_active":true,"delta_db":1.0377362418106775}],"active_band_count":3,"active_band_mae_db":0.7536465146367973}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":2.2500515235456793,"rms_envelope_nrmse":0.26016347511010907,"envelope_correlation":0.9494568516056944,"rms_gain_error_db":-1.2563015771485564,"reference_mean_power":0.00017547959358801547,"output_mean_power":0.00013140033189820875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":9.044433958704804e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2795647993321395e-05,"output_power":2.8251001945501043e-05,"reference_active":true,"delta_db":-1.8036575161606732},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.580957661437357e-06,"output_power":4.6989082215316784e-06,"reference_active":true,"delta_db":-1.4629213425523515},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5034181276391228e-06,"output_power":3.7173313879881694e-06,"reference_active":true,"delta_db":0.25739305961737774}],"active_band_count":3,"active_band_mae_db":1.1746573061101342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":1.747041760946224,"rms_envelope_nrmse":0.2014160017050762,"envelope_correlation":0.9748141018155216,"rms_gain_error_db":-1.2565843324513462,"reference_mean_power":0.00017667234117128199,"output_mean_power":0.00013228485660210377},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":7.913096629030374e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7674862471946175e-05,"output_power":2.491747967241081e-05,"reference_active":true,"delta_db":-1.795475630067644},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.042869679743245e-06,"output_power":2.3187332585295274e-06,"reference_active":true,"delta_db":-1.1803256143521197},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.532154624752839e-07,"output_power":2.5385170602300563e-07,"reference_active":true,"delta_db":0.010898643181809838}],"active_band_count":3,"active_band_mae_db":0.9955666292005244}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":807,"rms_floor":0.00015276264453937597,"log_rms_mae_db":3.312993131245223,"rms_envelope_nrmse":0.26650850831787054,"envelope_correlation":0.9694443675806748,"rms_gain_error_db":-1.772294152795789,"reference_mean_power":0.0006084209435601983,"output_mean_power":0.0004045523609095876},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":807,"frequency_resolution_hz":0.24783147459727384,"power_floor":5.417361809487122e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028753828158727807,"output_power":0.00022444475053059914,"reference_active":true,"delta_db":-1.0758622066380432},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.619361613819712e-05,"output_power":3.06249729081409e-05,"reference_active":true,"delta_db":-0.7255626286415017},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0811105176246507e-05,"output_power":8.304726708244986e-06,"reference_active":true,"delta_db":-1.1454474714901286}],"active_band_count":3,"active_band_mae_db":0.9822907689232245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64640,"uncovered_tail_samples":0,"metrics":{"frame_count":801,"rms_floor":0.00013215266595876848,"log_rms_mae_db":3.054032612787093,"rms_envelope_nrmse":0.23273870329418592,"envelope_correlation":0.9834526334519023,"rms_gain_error_db":-1.7708415222647895,"reference_mean_power":0.0006123930798728107,"output_mean_power":0.0004073297421039283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":801,"frequency_resolution_hz":0.24968789013732834,"power_floor":5.109927545308598e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026841635475720625,"output_power":0.00021282704083152318,"reference_active":true,"delta_db":-1.0078216750498696},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4787739775008156e-05,"output_power":1.2044279161200772e-05,"reference_active":true,"delta_db":-0.891209865423481},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.431673527944988e-07,"output_power":5.70289883434283e-07,"reference_active":true,"delta_db":-1.6981811487612186}],"active_band_count":3,"active_band_mae_db":1.199070896411523}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.4714365005493164,"predicted_mos":2.4714365005493164,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":97320,"input_sample_rate":24000,"input_seconds":4.055,"input_float32_sha256":"e191c14456038e8e4d5a6faed24e9120d1eb086d61583b587927b2c49b5c332c","internal_samples":64880,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.892628244000662},"started_at":"2026-10-13T07:43:26.273029+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018877410002460238,"native_input_samples":97200,"native_input_sha256":"f28a006a950d6646148bad5a2212551ae12b0fc0580f9df55943c49fa26febae","output_sha256":"359b3eceba3518057c3732de215311c81842160bf0ab9c5644ca439fd424f0dd","output_samples":97320,"length_delta_samples":120,"analysis_s":1.572447505001037,"synthesis_s":9.549714532000507,"total_s":11.122162037001544,"rtf":2.7462128486423567,"output_peak":0.4643138349056244,"output_rms":0.031902989493922146,"finished_at":"2026-10-13T07:43:37.403582+00:00","analysis_stages":{"external_f0_s":1.06819985599941},"evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"359b3eceba3518057c3732de215311c81842160bf0ab9c5644ca439fd424f0dd","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":64800,"output_samples":64880,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.920835023257677,"output_peak":0.467587790066474},"common_prefix":{"samples":64800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998766954377312,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.615091089274415,"spectral_convergence":0.740432810241553,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.500296735787848,"spectral_convergence":0.6569614870347643,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.588000256340298,"spectral_convergence":0.6674462058191086,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":0.00011016831871674357,"delta_db":-3.402979338585756,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":0.00010165762853541693,"delta_db":0.4377139782965732,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":7.646247601939369e-05,"delta_db":0.12142093046716257,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":2.1433468480850944e-05,"delta_db":-0.519027172312028,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0005384860470435091,"delta_db":-0.6194733601347487,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00017133446681810016,"delta_db":-1.2809864192131801,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.063600199834908,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9974145155836195,"pitch":null,"output_rms_dbfs":-29.91594457464871,"output_peak":0.467587790066474,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.901129360467521},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64800,"output_samples":64800,"intelligibility":{"estoi":{"value":0.846311029997404,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03581570440383133,"rms_dbfs":-28.918530059065088,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.359480807956004,"mean":-3.073421525366512e-06,"zero_energy":false},"output":{"rms":0.03193028325976877,"rms_dbfs":-29.91594457464871,"peak":0.467587790066474,"peak_dbfs":-6.6027367584155785,"crest_factor_db":23.31320781623313,"mean":0.0005814222560466216,"zero_energy":false},"rms_difference_db":-0.9974145155836212},"full_scale":{"reference":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":405,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":1.1674660845856357e-06,"output_rms_dbfs":-59.327557270284494,"output_zero_energy":false,"output_peak":0.004194185203456915,"output_energy_fraction":1.4136884557989511e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":398,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":85,"both_estimated_voiced_frames":54,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.7941176470588235,"absolute_difference_cents_median":4.881237953955729,"absolute_difference_cents_p95":19.767595211633818,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.0001629307662014554,"log_rms_mae_db":2.9217340165591805,"rms_envelope_nrmse":0.2733817741782208,"envelope_correlation":0.9216595730145377,"rms_gain_error_db":-0.9973752544939462,"reference_mean_power":0.0012840985730840112,"output_mean_power":0.0010206123939110873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.3397819245367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003947037290166488,"output_power":0.0002517091357790664,"reference_active":true,"delta_db":-1.953722509248037},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980542214464163e-05,"output_power":3.855989604562231e-05,"reference_active":true,"delta_db":-2.5775330340060294},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.603685703735387e-05,"output_power":1.461068126289073e-05,"reference_active":true,"delta_db":-2.509180919065712}],"active_band_count":3,"active_band_mae_db":2.3468121541065927}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.000144561944924798,"log_rms_mae_db":2.16764268833289,"rms_envelope_nrmse":0.2113684178645179,"envelope_correlation":0.9547017946111025,"rms_gain_error_db":-0.9975769186025867,"reference_mean_power":0.0012928029739092497,"output_mean_power":0.0010274830130663757},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":6.614408925582191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037454976712341086,"output_power":0.0002305782142413664,"reference_active":true,"delta_db":-2.1069126000279064},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9110632304140427e-05,"output_power":1.7104436656775886e-05,"reference_active":true,"delta_db":-2.309428636843535},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1315453531682465e-06,"output_power":9.249900501358875e-07,"reference_active":true,"delta_db":-3.625575164441389}],"active_band_count":3,"active_band_mae_db":2.6806388004376096}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":2.3503163913443825,"rms_envelope_nrmse":0.22494861172138564,"envelope_correlation":0.9538200191899648,"rms_gain_error_db":0.17293038658254964,"reference_mean_power":0.00013419941652038683,"output_mean_power":0.0001396508766315854},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.237716340453255e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0020888510118014e-05,"output_power":2.2808077749863435e-05,"reference_active":true,"delta_db":0.5660533756590667},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.868636044401025e-06,"output_power":8.40732167730297e-06,"reference_active":true,"delta_db":0.28758206588790974},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3701828329102724e-06,"output_power":1.4988945268610338e-06,"reference_active":true,"delta_db":0.38992551879673926}],"active_band_count":3,"active_band_mae_db":0.4145203201145719}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":1.6991575779370844,"rms_envelope_nrmse":0.14296445391092227,"envelope_correlation":0.9794808924547918,"rms_gain_error_db":0.17289526163831842,"reference_mean_power":0.00013520032575819994,"output_mean_power":0.0001406913069993937},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":4.6212903051863556e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8157545163089018e-05,"output_power":2.0715501015983025e-05,"reference_active":true,"delta_db":0.5723830836507707},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868319726245688e-06,"output_power":4.106923610479562e-06,"reference_active":true,"delta_db":0.2599426255955504},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2646920446881674e-07,"output_power":1.3727233909574019e-07,"reference_active":true,"delta_db":0.3559824732660096}],"active_band_count":3,"active_band_mae_db":0.39610272750411024}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":2.162985989316545,"rms_envelope_nrmse":0.23584134976027452,"envelope_correlation":0.9492885216297797,"rms_gain_error_db":-0.34968333280127795,"reference_mean_power":0.0001750632057906084,"output_mean_power":0.0001615200884676398},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":9.028764248682411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.277369850023164e-05,"output_power":3.732410347376545e-05,"reference_active":true,"delta_db":-0.5918741913366319},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.603092418254851e-06,"output_power":6.394154838131121e-06,"reference_active":true,"delta_db":-0.1396422703817614},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4970630790605323e-06,"output_power":3.0662635594784167e-06,"reference_active":true,"delta_db":-0.570939843299562}],"active_band_count":3,"active_band_mae_db":0.4341521016726518}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":1.3334944188950213,"rms_envelope_nrmse":0.11619480488322352,"envelope_correlation":0.9876475941718601,"rms_gain_error_db":-0.3497735158612211,"reference_mean_power":0.0001762447957551007,"output_mean_power":0.0001626068924604826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":7.899096535479964e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7672515224695035e-05,"output_power":3.2929933475083394e-05,"reference_active":true,"delta_db":-0.5843376371102134},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.050253319949333e-06,"output_power":3.069185928193647e-06,"reference_active":true,"delta_db":0.026872898609421207},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5566787812291305e-07,"output_power":2.2817558561102635e-07,"reference_active":true,"delta_db":-0.4940699327711453}],"active_band_count":3,"active_band_mae_db":0.36842682283026}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00015276264453937597,"log_rms_mae_db":2.3383499591035646,"rms_envelope_nrmse":0.24422901583750106,"envelope_correlation":0.9586537292926551,"rms_gain_error_db":-0.9636634448440335,"reference_mean_power":0.0006070254152411959,"output_mean_power":0.00048622863329427895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.389907119886333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002863305566547833,"output_power":0.00021903944810996556,"reference_active":true,"delta_db":-1.16345361033094},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.608243104449436e-05,"output_power":2.8388370778907048e-05,"reference_active":true,"delta_db":-1.0415532144488506},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0810295137794258e-05,"output_power":9.589876968432081e-06,"reference_active":true,"delta_db":-0.5202451552192819}],"active_band_count":3,"active_band_mae_db":0.9084173266663575}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.00013215266595876848,"log_rms_mae_db":1.459752599059603,"rms_envelope_nrmse":0.16117642627829973,"envelope_correlation":0.9865697484489468,"rms_gain_error_db":-0.9634255385083533,"reference_mean_power":0.0006109513084873902,"output_mean_power":0.0004894000906171058},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":5.083168365591418e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026729578490804684,"output_power":0.00020603504249310417,"reference_active":true,"delta_db":-1.1305101868199372},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4768224147952764e-05,"output_power":1.200189462785777e-05,"reference_active":true,"delta_db":-0.9007846582753564},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.446503663842911e-07,"output_power":6.419469311242169e-07,"reference_active":true,"delta_db":-1.1917784775952125}],"active_band_count":3,"active_band_mae_db":1.0743577742301686}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.684890031814575,"predicted_mos":2.684890031814575,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":97440,"input_sample_rate":24000,"input_seconds":4.06,"input_float32_sha256":"61755f7b34602fcbede21baa26e1709758f50e45acf4b47e42a90b033956cd23","internal_samples":64960,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.485039573000904},"started_at":"2026-10-13T05:11:14.015106+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0018947000007756287,"native_input_samples":97200,"native_input_sha256":"f28a006a950d6646148bad5a2212551ae12b0fc0580f9df55943c49fa26febae","output_sha256":"cc4288d8f895805b0ce2454ec80dc437aa2078dcd81d18ed9e1e09b102b919f5","output_samples":97440,"length_delta_samples":240,"analysis_s":0.4557464939989586,"synthesis_s":0.085923503000231,"total_s":0.5416699969991896,"rtf":0.13374567827140485,"output_peak":0.44672033190727234,"output_rms":0.033913573519249354,"finished_at":"2026-10-13T05:11:14.565333+00:00","analysis_stages":{"external_f0_s":0.1038521070004208},"evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"cc4288d8f895805b0ce2454ec80dc437aa2078dcd81d18ed9e1e09b102b919f5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":64800,"output_samples":64960,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.387934521005235,"output_peak":0.4433877061201406},"common_prefix":{"samples":64800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975369458128078,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.117139987702585,"spectral_convergence":0.49995702727384433,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.65941814300741,"spectral_convergence":0.45937000247485893,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.281966868966219,"spectral_convergence":0.5205261153876987,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":0.00018263989132742876,"delta_db":-1.20759015482628,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":8.88629061542637e-05,"delta_db":-0.14648064238269426,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":6.812936517340884e-05,"delta_db":-0.37971922032872085,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":2.1827124733353095e-05,"delta_db":-0.439986435910358,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0005718645885008955,"delta_db":-0.35828585975124666,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.00022085829360860782,"delta_db":-0.1782966693130995,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.45172649708539997,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4587263327964315,"pitch":null,"output_rms_dbfs":-29.37725639186152,"output_peak":0.4433877061201406,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.019508333225405},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64800,"output_samples":64800,"intelligibility":{"estoi":{"value":0.902668987558832,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03581570440383133,"rms_dbfs":-28.918530059065088,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.359480807956004,"mean":-3.073421525366512e-06,"zero_energy":false},"output":{"rms":0.03397325668204365,"rms_dbfs":-29.37725639186152,"peak":0.4433877061201406,"peak_dbfs":-7.0643270561041955,"crest_factor_db":22.312929335757325,"mean":-0.00036812955545523435,"zero_energy":false},"rms_difference_db":-0.4587263327964308},"full_scale":{"reference":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":405,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":1.3855711535801576e-06,"output_rms_dbfs":-58.583711668765034,"output_zero_energy":false,"output_peak":0.0033320533273125644,"output_energy_fraction":1.4820725152105944e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":398,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":62,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9117647058823529,"absolute_difference_cents_median":1.2693122758638602,"absolute_difference_cents_p95":3.513624691677709,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.0001629307662014554,"log_rms_mae_db":1.1445324844581617,"rms_envelope_nrmse":0.13949825561042434,"envelope_correlation":0.9804631803205621,"rms_gain_error_db":-0.4586064255558927,"reference_mean_power":0.0012840985730840112,"output_mean_power":0.0011554142535264353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.3397819245367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003947037290166488,"output_power":0.0003338656211076955,"reference_active":true,"delta_db":-0.7269952839132537},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980542214464163e-05,"output_power":5.775635839767574e-05,"reference_active":true,"delta_db":-0.8228935513296666},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.603685703735387e-05,"output_power":2.2608212694987868e-05,"reference_active":true,"delta_db":-0.6132232817145102}],"active_band_count":3,"active_band_mae_db":0.7210373723191434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.000144561944924798,"log_rms_mae_db":0.7067136961279722,"rms_envelope_nrmse":0.089028804535047,"envelope_correlation":0.9932066069467248,"rms_gain_error_db":-0.4583705007403458,"reference_mean_power":0.0012928029739092497,"output_mean_power":0.0011633095473552837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":6.614408925582191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037454976712341086,"output_power":0.00031304513078475853,"reference_active":true,"delta_db":-0.7790257838128214},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9110632304140427e-05,"output_power":2.289978529668495e-05,"reference_active":true,"delta_db":-1.0422022822368353},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1315453531682465e-06,"output_power":1.558626086364108e-06,"reference_active":true,"delta_db":-1.3595263697151365}],"active_band_count":3,"active_band_mae_db":1.0602514785882644}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":1.146729735857643,"rms_envelope_nrmse":0.12752924735864207,"envelope_correlation":0.9856090360223174,"rms_gain_error_db":-0.3419276004334727,"reference_mean_power":0.00013419941652038683,"output_mean_power":0.00012403888896574404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.237716340453255e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0020888510118014e-05,"output_power":1.8535654458868756e-05,"reference_active":true,"delta_db":-0.33475422440908503},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.868636044401025e-06,"output_power":6.313204097343854e-06,"reference_active":true,"delta_db":-0.9564962822664571},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3701828329102724e-06,"output_power":1.1790521590051386e-06,"reference_active":true,"delta_db":-0.6524550404540113}],"active_band_count":3,"active_band_mae_db":0.6479018490431845}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":0.6887160973954953,"rms_envelope_nrmse":0.07458279027128278,"envelope_correlation":0.995592227616685,"rms_gain_error_db":-0.34193217559296774,"reference_mean_power":0.00013520032575819994,"output_mean_power":0.00012496388555854939},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":4.6212903051863556e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8157545163089018e-05,"output_power":1.6812729056814022e-05,"reference_active":true,"delta_db":-0.3341891883149669},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868319726245688e-06,"output_power":3.005777729706459e-06,"reference_active":true,"delta_db":-1.0956550004514949},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2646920446881674e-07,"output_power":7.958885489261635e-08,"reference_active":true,"delta_db":-2.0113253050659643}],"active_band_count":3,"active_band_mae_db":1.147056497944142}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":1.1913162096845753,"rms_envelope_nrmse":0.13235599404617698,"envelope_correlation":0.9855870444664997,"rms_gain_error_db":-0.4262987784497521,"reference_mean_power":0.0001750632057906084,"output_mean_power":0.000158695641505143},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":9.028764248682411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.277369850023164e-05,"output_power":3.810059111729063e-05,"reference_active":true,"delta_db":-0.5024509023353263},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.603092418254851e-06,"output_power":5.835457605790391e-06,"reference_active":true,"delta_db":-0.5367245750479367},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4970630790605323e-06,"output_power":3.1979894568665875e-06,"reference_active":true,"delta_db":-0.3882643843384436}],"active_band_count":3,"active_band_mae_db":0.47581328724056887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":0.7124921880464097,"rms_envelope_nrmse":0.075541408709055,"envelope_correlation":0.9964370496826128,"rms_gain_error_db":-0.42580815971445524,"reference_mean_power":0.0001762447957551007,"output_mean_power":0.00015978480824696664},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":7.899096535479964e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7672515224695035e-05,"output_power":3.3555623819373795e-05,"reference_active":true,"delta_db":-0.5025930003902196},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.050253319949333e-06,"output_power":2.69254173104056e-06,"reference_active":true,"delta_db":-0.5417346544499374},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5566787812291305e-07,"output_power":2.1225202295036963e-07,"reference_active":true,"delta_db":-0.8082432894643219}],"active_band_count":3,"active_band_mae_db":0.617523648101493}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00015276264453937597,"log_rms_mae_db":1.047886920793365,"rms_envelope_nrmse":0.12530337253768803,"envelope_correlation":0.9883986923440287,"rms_gain_error_db":-0.24852430728422303,"reference_mean_power":0.0006070254152411959,"output_mean_power":0.0005732637020850598},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.389907119886333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002863305566547833,"output_power":0.0002584899894338672,"reference_active":true,"delta_db":-0.44423968787118484},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.608243104449436e-05,"output_power":3.693554024341045e-05,"reference_active":true,"delta_db":0.10148665344617061},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0810295137794258e-05,"output_power":1.2182748835971917e-05,"reference_active":true,"delta_db":0.5190773970039829}],"active_band_count":3,"active_band_mae_db":0.3549345794404461}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.00013215266595876848,"log_rms_mae_db":0.6435360344155877,"rms_envelope_nrmse":0.06892639602690553,"envelope_correlation":0.9967307191873583,"rms_gain_error_db":-0.24827552951980553,"reference_mean_power":0.0006109513084873902,"output_mean_power":0.0005770042955820447},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":5.083168365591418e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026729578490804684,"output_power":0.00024195943734414108,"reference_active":true,"delta_db":-0.4324954432867635},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4768224147952764e-05,"output_power":1.512764334648543e-05,"reference_active":true,"delta_db":0.10443001579174341},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.446503663842911e-07,"output_power":8.948209399138833e-07,"reference_active":true,"delta_db":0.2505916378934492}],"active_band_count":3,"active_band_mae_db":0.262505698990652}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-76324-0024","method":"magphase","status":"measured","mos":{"status":"measured","value":2.481335163116455,"predicted_mos":2.481335163116455,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":195240,"input_sample_rate":48000,"input_seconds":4.0675,"input_float32_sha256":"03667680f5f5fd21d8978eba53d3210a6cd8bee526e907a9d0271b334a86730c","internal_samples":65080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":10.37818300500112},"started_at":"2026-10-13T05:17:30.308949+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006884481999804848,"native_input_samples":194400,"native_input_sha256":"08af17282e9b5eee363efcf29b517c7f490e51dcdb3cceecbb5f02d340ec6dfe","output_sha256":"0b53bcdbe10abe0d8681acc2b4d26429a64d3ef1688062060f4d466918eba788","output_samples":195240,"length_delta_samples":840,"analysis_s":2.3543333810011973,"synthesis_s":0.8348072999997385,"total_s":3.1891406810009357,"rtf":0.7874421434570212,"output_peak":0.48031125118432855,"output_rms":0.035085840023769396,"finished_at":"2026-10-13T05:17:33.515582+00:00","evaluation_input":{"reference":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"0b53bcdbe10abe0d8681acc2b4d26429a64d3ef1688062060f4d466918eba788","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":64800,"output_samples":65080,"length_delta_samples":280,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-29.092491646507096,"output_peak":0.4414782828741337},"common_prefix":{"samples":64800,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9956976029502151,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":64800,"output_samples":64800,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.549310702292366,"spectral_convergence":0.743017838321924,"spectral_convergence_status":"ok","amplitude_floor":1.7331847370499167e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.341329061791958,"spectral_convergence":0.5838273492453951,"spectral_convergence_status":"ok","amplitude_floor":1.0846095570828337e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.978181968451933,"spectral_convergence":0.5637442746714918,"spectral_convergence_status":"ok","amplitude_floor":4.5533844599041527e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00024118742205032635,"output_power":0.00013777928986512058,"delta_db":-2.431707135515279,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":9.191122965591724e-05,"output_power":8.331426023012152e-05,"delta_db":-0.42649233999374025,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":7.435432994216677e-05,"output_power":6.766354507653585e-05,"delta_db":-0.4095151634004387,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":2.4154339925686723e-05,"output_power":2.13995130136148e-05,"delta_db":-0.5259128359411804,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006210431709364446,"output_power":0.0006561749660977045,"delta_db":0.23897866892284592,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023011417890304625,"output_power":0.0002712461156070033,"delta_db":0.7142014822984277,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7911346043453188,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1557452122602773,"pitch":null,"output_rms_dbfs":-29.074275271325366,"output_peak":0.4414782828741337,"reference_rms_dbfs":-28.918530059065088,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.622940577512085},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":64800,"output_samples":64800,"intelligibility":{"estoi":{"value":0.8088499316317141,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.03581570440383133,"rms_dbfs":-28.918530059065088,"peak":0.373291015625,"peak_dbfs":-8.559049251109084,"crest_factor_db":20.359480807956004,"mean":-3.073421525366512e-06,"zero_energy":false},"output":{"rms":0.03517922246743609,"rms_dbfs":-29.074275271325366,"peak":0.4414782828741337,"peak_dbfs":-7.101813106049986,"crest_factor_db":21.97246216527538,"mean":0.0013092982936846482,"zero_energy":false},"rms_difference_db":-0.15574521226027827},"full_scale":{"reference":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":64800,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":405,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.3123664717422797e-06,"threshold_dbfs":-56.35943336322904,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":5,"quiet_samples":800,"reference_mean_power":1.3342907186597585e-06,"output_mean_power":3.6456911481197825e-06,"output_rms_dbfs":-54.382201262320024,"output_zero_energy":false,"output_peak":0.006541730537652349,"output_energy_fraction":3.636824817737211e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":398,"reference_estimated_voiced_frames":68,"output_estimated_voiced_frames":70,"both_estimated_voiced_frames":62,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9117647058823529,"absolute_difference_cents_median":4.238687999432768,"absolute_difference_cents_p95":14.427962216242703,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":64800,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.0001629307662014554,"log_rms_mae_db":2.3624411211961593,"rms_envelope_nrmse":0.26221264372507147,"envelope_correlation":0.9264080107884026,"rms_gain_error_db":-0.15520905958435283,"reference_mean_power":0.0012840985730840112,"output_mean_power":0.0012390175484518185},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":7.3397819245367e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003947037290166488,"output_power":0.00038666203735199136,"reference_active":true,"delta_db":-0.08939694526676083},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.980542214464163e-05,"output_power":4.696776190155181e-05,"reference_active":true,"delta_db":-1.7208929204795083},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.603685703735387e-05,"output_power":2.6254572040061193e-05,"reference_active":true,"delta_db":0.03616385029914231}],"active_band_count":3,"active_band_mae_db":0.6154845720151372}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.000144561944924798,"log_rms_mae_db":1.3876022384475049,"rms_envelope_nrmse":0.15583874420542246,"envelope_correlation":0.9718880091848513,"rms_gain_error_db":-0.1555519730462806,"reference_mean_power":0.0012928029739092497,"output_mean_power":0.0012473178721450144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":6.614408925582191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037454976712341086,"output_power":0.0003578374340842767,"reference_active":true,"delta_db":-0.19823760368482812},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9110632304140427e-05,"output_power":2.0024232452666446e-05,"reference_active":true,"delta_db":-1.6249576056727977},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1315453531682465e-06,"output_power":1.7488824725046727e-06,"reference_active":true,"delta_db":-0.859339523956048}],"active_band_count":3,"active_band_mae_db":0.8941782444378913}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":3.733123977436472e-05,"log_rms_mae_db":2.5141825324758447,"rms_envelope_nrmse":0.24826897968120895,"envelope_correlation":0.9419619714168324,"rms_gain_error_db":-0.4210245003348861,"reference_mean_power":0.00013419941652038683,"output_mean_power":0.0001218002493967821},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.237716340453255e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0020888510118014e-05,"output_power":1.815872845307755e-05,"reference_active":true,"delta_db":-0.4239791298312452},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.868636044401025e-06,"output_power":7.01359298167161e-06,"reference_active":true,"delta_db":-0.49958898974966204},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3701828329102724e-06,"output_power":1.5936166774688896e-06,"reference_active":true,"delta_db":0.656053441488246}],"active_band_count":3,"active_band_mae_db":0.5265405203563844}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.100732475570721e-05,"log_rms_mae_db":1.8156988462955554,"rms_envelope_nrmse":0.1567408486308795,"envelope_correlation":0.9757107347965838,"rms_gain_error_db":-0.42102154240509004,"reference_mean_power":0.00013520032575819994,"output_mean_power":0.000122708764602987},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":4.6212903051863556e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8157545163089018e-05,"output_power":1.6396016644307538e-05,"reference_active":true,"delta_db":-0.4431878253545125},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.868319726245688e-06,"output_power":3.5438212359016538e-06,"reference_active":true,"delta_db":-0.3805055614529085},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2646920446881674e-07,"output_power":1.3563941128825174e-07,"reference_active":true,"delta_db":0.3040110943694214}],"active_band_count":3,"active_band_mae_db":0.37590149372561416}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":4.662083951365485e-05,"log_rms_mae_db":2.753953034666112,"rms_envelope_nrmse":0.2886573639156572,"envelope_correlation":0.9272523457701026,"rms_gain_error_db":0.07590237226891572,"reference_mean_power":0.0001750632057906084,"output_mean_power":0.0001781497077809016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":9.028764248682411e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.277369850023164e-05,"output_power":4.518550085514337e-05,"reference_active":true,"delta_db":0.23822296658740505},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.603092418254851e-06,"output_power":8.57984512776432e-06,"reference_active":true,"delta_db":1.1373207280733009},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4970630790605323e-06,"output_power":7.1861743278597355e-06,"reference_active":true,"delta_db":3.1279428238885805}],"active_band_count":3,"active_band_mae_db":1.501162172849762}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":3.913562930844258e-05,"log_rms_mae_db":1.8434548230557752,"rms_envelope_nrmse":0.1690239227431632,"envelope_correlation":0.9725183662831541,"rms_gain_error_db":0.07596190504341314,"reference_mean_power":0.0001762447957551007,"output_mean_power":0.000179354588668608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":7.899096535479964e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.7672515224695035e-05,"output_power":4.178173326429199e-05,"reference_active":true,"delta_db":0.44961835131485317},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.050253319949333e-06,"output_power":4.850168621914307e-06,"reference_active":true,"delta_db":2.014209291416415},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5566787812291305e-07,"output_power":5.680109348778419e-07,"reference_active":true,"delta_db":3.4668052937607636}],"active_band_count":3,"active_band_mae_db":1.9768776454973438}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":809,"rms_floor":0.00015276264453937597,"log_rms_mae_db":2.9454729221764713,"rms_envelope_nrmse":0.2687176028893593,"envelope_correlation":0.9506994348060253,"rms_gain_error_db":0.4359391800491897,"reference_mean_power":0.0006070254152411959,"output_mean_power":0.0006711209446942385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":809,"frequency_resolution_hz":0.24721878862793573,"power_floor":5.389907119886333e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002863305566547833,"output_power":0.00032718712729396963,"reference_active":true,"delta_db":0.5792851106535345},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.608243104449436e-05,"output_power":3.631532681818922e-05,"reference_active":true,"delta_db":0.027941665946507908},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0810295137794258e-05,"output_power":1.1121930314509173e-05,"reference_active":true,"delta_db":0.12342618609938379}],"active_band_count":3,"active_band_mae_db":0.24355098756647542}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":64800,"uncovered_tail_samples":0,"metrics":{"frame_count":803,"rms_floor":0.00013215266595876848,"log_rms_mae_db":1.8305502006136007,"rms_envelope_nrmse":0.14891256678850004,"envelope_correlation":0.9853881238911395,"rms_gain_error_db":0.43583664180915577,"reference_mean_power":0.0006109513084873902,"output_mean_power":0.0006754454235117311},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":803,"frequency_resolution_hz":0.24906600249066005,"power_floor":5.083168365591418e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026729578490804684,"output_power":0.00030569589536225854,"reference_active":true,"delta_db":0.5829749715877152},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4768224147952764e-05,"output_power":1.4572412761870978e-05,"reference_active":true,"delta_db":-0.05796811247343987},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.446503663842911e-07,"output_power":7.226011752959402e-07,"reference_active":true,"delta_db":-0.6777831117718864}],"active_band_count":3,"active_band_mae_db":0.43957539861101386}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"hnm","status":"measured","mos":{"status":"measured","value":2.8805973529815674,"predicted_mos":2.8805973529815674,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49200,"input_sample_rate":16000,"input_seconds":3.075,"input_float32_sha256":"2b214903b054e4f235956f1b851c34818a0331b35d91b2a4b362b5d83fa16231","internal_samples":49200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6058961629987607},"started_at":"2026-10-13T04:39:57.009698+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.58180000653374e-05,"native_input_samples":49200,"native_input_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","output_sha256":"780f90b9a3fddc3419e431d52474d9b36c4299d55cf698b68d9147e020283830","output_samples":49200,"length_delta_samples":0,"analysis_s":0.4838923409997733,"synthesis_s":0.20867694899970957,"total_s":0.6925692899994829,"rtf":0.22522578536568547,"output_peak":0.4595306165309517,"output_rms":0.04344328630485304,"finished_at":"2026-10-13T04:39:57.706476+00:00","evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"780f90b9a3fddc3419e431d52474d9b36c4299d55cf698b68d9147e020283830","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.518675739943631,"spectral_convergence":0.43657805448646425,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.368127539170453,"spectral_convergence":0.407143836287345,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.392852290565386,"spectral_convergence":0.44908212602384406,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":7.243855958164896e-05,"delta_db":-0.1344149934493152,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.00028877448286268717,"delta_db":0.10181116161477519,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.000273573277298369,"delta_db":0.09254280506119839,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.00028041910933917504,"delta_db":-0.09015863448806766,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0005741989682450433,"delta_db":-0.31081897234847183,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.0003979147276385082,"delta_db":0.20370865006493227,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1555758695044601,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04593503149499824,"pitch":null,"output_rms_dbfs":-27.241546591919928,"output_peak":0.4595306165309517,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.759885189893157},"common_prefix":{"samples":49200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.518675739943631,"spectral_convergence":0.43657805448646425,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.368127539170453,"spectral_convergence":0.407143836287345,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.392852290565386,"spectral_convergence":0.44908212602384406,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":7.243855958164896e-05,"delta_db":-0.1344149934493152,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.00028877448286268717,"delta_db":0.10181116161477519,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.000273573277298369,"delta_db":0.09254280506119839,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.00028041910933917504,"delta_db":-0.09015863448806766,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0005741989682450433,"delta_db":-0.31081897234847183,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.0003979147276385082,"delta_db":0.20370865006493227,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1555758695044601,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04593503149499824,"pitch":null,"output_rms_dbfs":-27.241546591919928,"output_peak":0.4595306165309517,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.759885189893157},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49200,"output_samples":49200,"intelligibility":{"estoi":{"value":0.8505219157408774,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04367364322520635,"rms_dbfs":-27.195611560424933,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.948879491442757,"mean":-0.00013451925138147867,"zero_energy":false},"output":{"rms":0.04344328630485304,"rms_dbfs":-27.241546591919928,"peak":0.4595306165309517,"peak_dbfs":-6.7537109631421455,"crest_factor_db":20.487835628777784,"mean":-7.291269330400715e-05,"zero_energy":false},"rms_difference_db":-0.045935031494995116},"full_scale":{"reference":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":1.0814275022238572e-06,"output_rms_dbfs":-59.66002589899859,"output_zero_energy":false,"output_peak":0.0022446246750628383,"output_energy_fraction":1.8634036814894537e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":70,"both_estimated_voiced_frames":68,"reference_voiced_output_unvoiced_frames":29,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.7010309278350515,"absolute_difference_cents_median":1.2095204904199566,"absolute_difference_cents_p95":4.80187079090623,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":0.00010992699921133752,"log_rms_mae_db":0.8661046981919748,"rms_envelope_nrmse":0.09680617405093497,"envelope_correlation":0.9867150499664621,"rms_gain_error_db":-0.04602411674130105,"reference_mean_power":0.0019104810338505282,"output_mean_power":0.0018903417182215927},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":4.70123588122935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030753312258320766,"output_power":0.0003002630175961667,"reference_active":true,"delta_db":-0.1039005305758113},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8607043791082874e-05,"output_power":2.7416741198810257e-05,"reference_active":true,"delta_db":-0.18457148255577543},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.145843039363973e-05,"output_power":1.5243535447818782e-05,"reference_active":true,"delta_db":1.2396057413351238}],"active_band_count":3,"active_band_mae_db":0.5093592514889035}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":0.4318048520456661,"rms_envelope_nrmse":0.05494663062358986,"envelope_correlation":0.9952500133752431,"rms_gain_error_db":-0.04600646813731596,"reference_mean_power":0.0019290839559002245,"output_mean_power":0.0019087562944410119},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":4.127760176364045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002739478020838986,"output_power":0.0002676216968084081,"reference_active":true,"delta_db":-0.10146500446903926},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7486141765201905e-05,"output_power":1.6482468276387406e-05,"reference_active":true,"delta_db":-0.25671746482381047},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8391125647823e-07,"output_power":7.791596242346521e-07,"reference_active":true,"delta_db":1.2527959193341607}],"active_band_count":3,"active_band_mae_db":0.5369927962090034}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":1.2365096595702794,"rms_envelope_nrmse":0.08463097164756136,"envelope_correlation":0.9927863591426538,"rms_gain_error_db":0.06454061408632898,"reference_mean_power":0.0004974887922363912,"output_mean_power":0.0005049371942734059},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.1951630543463553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015634740722861155,"output_power":0.00015721762157102262,"reference_active":true,"delta_db":0.02410538309644155},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160597194093242e-06,"output_power":8.43077332033083e-06,"reference_active":true,"delta_db":0.14145470835529272},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0536955073753773e-06,"output_power":3.0556982996589193e-06,"reference_active":true,"delta_db":1.7257441947338903}],"active_band_count":3,"active_band_mae_db":0.6304347620618749}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":0.724107545788371,"rms_envelope_nrmse":0.042093338303546474,"envelope_correlation":0.9981287336653669,"rms_gain_error_db":0.0645347299340154,"reference_mean_power":0.00050239739349007,"output_mean_power":0.0005099185962264651},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.0265825460069485e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014180308081562638,"output_power":0.00014235590101406288,"reference_active":true,"delta_db":0.016898080294019232},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469176320839401e-06,"output_power":4.520793098329078e-06,"reference_active":true,"delta_db":0.0498714220886988},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.310040798443835e-08,"output_power":1.332802504708419e-07,"reference_active":true,"delta_db":1.5581421607419368}],"active_band_count":3,"active_band_mae_db":0.5416372210415515}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":1.1158036995209577,"rms_envelope_nrmse":0.11626996297565684,"envelope_correlation":0.9875441258006941,"rms_gain_error_db":-0.05575698513306042,"reference_mean_power":0.0004940955013379761,"output_mean_power":0.0004877925928076051},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":1.347463406037001e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.364408948380081e-05,"output_power":9.069729682723617e-05,"reference_active":true,"delta_db":-0.13886027899939565},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.996833569277437e-06,"output_power":6.492474191578519e-06,"reference_active":true,"delta_db":-0.32491311727489136},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2622375016010212e-06,"output_power":3.343548548324746e-06,"reference_active":true,"delta_db":1.6966943603379874}],"active_band_count":3,"active_band_mae_db":0.7201559188707581}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":0.6139876737511339,"rms_envelope_nrmse":0.04989331024251948,"envelope_correlation":0.9976015140292217,"rms_gain_error_db":-0.0557584139226112,"reference_mean_power":0.0004989699589215842,"output_mean_power":0.0004926047075150638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":1.2432388404774247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.827683631948742e-05,"output_power":8.547636320992335e-05,"reference_active":true,"delta_db":-0.1400072462849335},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.844195926765935e-06,"output_power":3.919262287833842e-06,"reference_active":true,"delta_db":0.08398814428429988},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.483409815447234e-08,"output_power":1.5371142776999692e-07,"reference_active":true,"delta_db":2.0974163805069064}],"active_band_count":3,"active_band_mae_db":0.7738039236920465}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":1.187346180538631,"rms_envelope_nrmse":0.164147900529032,"envelope_correlation":0.9716180284310129,"rms_gain_error_db":-0.14228898171142815,"reference_mean_power":0.0007033343720650755,"output_mean_power":0.000680664254249135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.960535239443851e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011696049836275019,"output_power":0.0001089421124895788,"reference_active":true,"delta_db":-0.30843417729025735},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1203242070906313e-05,"output_power":1.9308225551284e-05,"reference_active":true,"delta_db":-0.4065990814438455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0967126516322994e-05,"output_power":1.3330807102623282e-05,"reference_active":true,"delta_db":0.8476359070572846}],"active_band_count":3,"active_band_mae_db":0.5208897219304625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":0.6874561719662858,"rms_envelope_nrmse":0.10284941839611685,"envelope_correlation":0.9879222691120259,"rms_gain_error_db":-0.14228964147892362,"reference_mean_power":0.0007102740547296545,"output_mean_power":0.0006873801502264581},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.637352177112327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010173798550296787,"output_power":9.483296187221628e-05,"reference_active":true,"delta_db":-0.30523819105814515},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907094272462648e-05,"output_power":1.1675843782389268e-05,"reference_active":true,"delta_db":-0.4354020643402624},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.316189457175963e-07,"output_power":7.155235224465417e-07,"reference_active":true,"delta_db":0.5416876711937725}],"active_band_count":3,"active_band_mae_db":0.4274426421973933}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.1761155128479004,"predicted_mos":2.1761155128479004,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49200,"input_sample_rate":16000,"input_seconds":3.075,"input_float32_sha256":"caae20a9bb87db8f8c725bffa9bb0adf579381c4a14a16ee4661dfe08274758d","internal_samples":49200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2475309879991983},"started_at":"2026-10-13T04:43:08.741525+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3392000002786517e-05,"native_input_samples":49200,"native_input_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","output_sha256":"7549262e2f11da579be94eb3fd219fb87d06bad93aebf20fccfe88b4c6a62634","output_samples":49200,"length_delta_samples":0,"analysis_s":0.6481078389997492,"synthesis_s":0.04086094600006618,"total_s":0.6889687849998154,"rtf":0.22405488943083426,"output_peak":0.5004980393644866,"output_rms":0.04792050295538567,"finished_at":"2026-10-13T04:43:09.434438+00:00","evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"7549262e2f11da579be94eb3fd219fb87d06bad93aebf20fccfe88b4c6a62634","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6796747425196665,"spectral_convergence":0.6676745423384182,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4425260348087035,"spectral_convergence":0.5309518553762255,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.243041837128292,"spectral_convergence":0.6301489097298001,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":5.156287104764157e-05,"delta_db":-1.6107421416274703,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.00042941152935458153,"delta_db":1.8249600208361758,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.0003696956976175612,"delta_db":1.400250038159265,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.0003313145037002265,"delta_db":0.6341697904712531,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0006840050104696972,"delta_db":0.44914977993485616,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.00043038499130741904,"delta_db":0.5443796933022857,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0772752440552178,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8060389123248053,"pitch":null,"output_rms_dbfs":-26.389572648100128,"output_peak":0.5004980393644866,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.121747538152221},"common_prefix":{"samples":49200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6796747425196665,"spectral_convergence":0.6676745423384182,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4425260348087035,"spectral_convergence":0.5309518553762255,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.243041837128292,"spectral_convergence":0.6301489097298001,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":5.156287104764157e-05,"delta_db":-1.6107421416274703,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.00042941152935458153,"delta_db":1.8249600208361758,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.0003696956976175612,"delta_db":1.400250038159265,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.0003313145037002265,"delta_db":0.6341697904712531,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0006840050104696972,"delta_db":0.44914977993485616,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.00043038499130741904,"delta_db":0.5443796933022857,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0772752440552178,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8060389123248053,"pitch":null,"output_rms_dbfs":-26.389572648100128,"output_peak":0.5004980393644866,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.121747538152221},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49200,"output_samples":49200,"intelligibility":{"estoi":{"value":0.8110543821742516,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04367364322520635,"rms_dbfs":-27.195611560424933,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.948879491442757,"mean":-0.00013451925138147867,"zero_energy":false},"output":{"rms":0.04792050295538568,"rms_dbfs":-26.38957264810013,"peak":0.5004980393644866,"peak_dbfs":-6.011952389461576,"crest_factor_db":20.377620258638554,"mean":8.27078985736812e-09,"zero_energy":false},"rms_difference_db":0.8060389123248015},"full_scale":{"reference":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":2.7796122089517953e-07,"output_rms_dbfs":-65.560157894272,"output_zero_energy":false,"output_peak":0.001973965673440562,"output_energy_fraction":3.9363740060694857e-07,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":50,"both_estimated_voiced_frames":45,"reference_voiced_output_unvoiced_frames":52,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.4639175257731959,"absolute_difference_cents_median":3.1919860211693316,"absolute_difference_cents_p95":10.039126151398388,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":0.00010992699921133752,"log_rms_mae_db":2.0834930749601517,"rms_envelope_nrmse":0.20798071525963652,"envelope_correlation":0.9591691653199482,"rms_gain_error_db":0.8060657276890384,"reference_mean_power":0.0019104810338505282,"output_mean_power":0.0023001136928519164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":4.70123588122935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030753312258320766,"output_power":0.00041400364285872234,"reference_active":true,"delta_db":1.2911226460355996},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8607043791082874e-05,"output_power":4.542575735444937e-05,"reference_active":true,"delta_db":2.008291959682872},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.145843039363973e-05,"output_power":1.998381256079923e-05,"reference_active":true,"delta_db":2.415532166726595}],"active_band_count":3,"active_band_mae_db":1.9049822574816888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":1.5159283879273027,"rms_envelope_nrmse":0.1423444187876752,"envelope_correlation":0.9858760414902747,"rms_gain_error_db":0.8066109854759596,"reference_mean_power":0.0019290839559002245,"output_mean_power":0.002322802194067226},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":4.127760176364045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002739478020838986,"output_power":0.0003704975096543572,"reference_active":true,"delta_db":1.3111747277219186},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7486141765201905e-05,"output_power":2.3124800171480468e-05,"reference_active":true,"delta_db":1.2138399353747582},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8391125647823e-07,"output_power":9.466095063665499e-07,"reference_active":true,"delta_db":2.0982401403503306}],"active_band_count":3,"active_band_mae_db":1.5410849344823359}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":2.2275538852409538,"rms_envelope_nrmse":0.37136883697411976,"envelope_correlation":0.9256326819379478,"rms_gain_error_db":1.6272544786957113,"reference_mean_power":0.0004974887922363912,"output_mean_power":0.0007236169795969065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.1951630543463553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015634740722861155,"output_power":0.00023438295555310379,"reference_active":true,"delta_db":1.758353428201531},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160597194093242e-06,"output_power":3.5514151442314946e-05,"reference_active":true,"delta_db":6.386795006414275},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0536955073753773e-06,"output_power":1.6045471811191627e-05,"reference_active":true,"delta_db":8.928164388345156}],"active_band_count":3,"active_band_mae_db":5.69110427432032}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":1.72814349057943,"rms_envelope_nrmse":0.28141072362595276,"envelope_correlation":0.9689348701801332,"rms_gain_error_db":1.6272353749275634,"reference_mean_power":0.00050239739349007,"output_mean_power":0.0007307535183814458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.0265825460069485e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014180308081562638,"output_power":0.00022565810779942808,"reference_active":true,"delta_db":2.0176527571447194},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469176320839401e-06,"output_power":1.7648350700384065e-05,"reference_active":true,"delta_db":5.964766362281732},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.310040798443835e-08,"output_power":7.443530162500268e-07,"reference_active":true,"delta_db":9.02827368396697}],"active_band_count":3,"active_band_mae_db":5.670230934464474}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":1.547882113436191,"rms_envelope_nrmse":0.19684833645748656,"envelope_correlation":0.9701595290163855,"rms_gain_error_db":0.5508770161681166,"reference_mean_power":0.0004940955013379761,"output_mean_power":0.0005609169983802603},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":1.347463406037001e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.364408948380081e-05,"output_power":9.428759311627657e-05,"reference_active":true,"delta_db":0.029741783295395716},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.996833569277437e-06,"output_power":1.0716290925875542e-05,"reference_active":true,"delta_db":1.8514295166745853},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2622375016010212e-06,"output_power":5.553772511099884e-06,"reference_active":true,"delta_db":3.900498890740745}],"active_band_count":3,"active_band_mae_db":1.9272233969035752}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":0.9924262761875932,"rms_envelope_nrmse":0.11086316323163264,"envelope_correlation":0.9927289158543885,"rms_gain_error_db":0.5508776006355031,"reference_mean_power":0.0004989699589215842,"output_mean_power":0.0005664507540511414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":1.2432388404774247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.827683631948742e-05,"output_power":8.926634274300203e-05,"reference_active":true,"delta_db":0.04840981572927291},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.844195926765935e-06,"output_power":5.645830926362985e-06,"reference_active":true,"delta_db":1.669223541414859},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.483409815447234e-08,"output_power":2.3559557492493867e-07,"reference_active":true,"delta_db":3.9520261051549133}],"active_band_count":3,"active_band_mae_db":1.8898864874330152}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":1.5584980350359472,"rms_envelope_nrmse":0.21630833609692018,"envelope_correlation":0.9563015278896367,"rms_gain_error_db":0.4869005977990759,"reference_mean_power":0.0007033343720650755,"output_mean_power":0.0007867774459002975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.960535239443851e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011696049836275019,"output_power":0.00013584232838091855,"reference_active":true,"delta_db":0.6499590678539436},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1203242070906313e-05,"output_power":3.053788415614302e-05,"reference_active":true,"delta_db":1.584366717097636},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0967126516322994e-05,"output_power":1.4889536614716305e-05,"reference_active":true,"delta_db":1.3278832854798595}],"active_band_count":3,"active_band_mae_db":1.1874030234771464}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":0.9157123662327339,"rms_envelope_nrmse":0.12173208282460937,"envelope_correlation":0.987275336286112,"rms_gain_error_db":0.4869014813048527,"reference_mean_power":0.0007102740547296545,"output_mean_power":0.000794540609061392},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.637352177112327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010173798550296787,"output_power":0.00012064857851080996,"reference_active":true,"delta_db":0.7403907540263643},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907094272462648e-05,"output_power":1.6076970898468657e-05,"reference_active":true,"delta_db":0.9537574350981768},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.316189457175963e-07,"output_power":8.294445415837477e-07,"reference_active":true,"delta_db":1.1833220507761906}],"active_band_count":3,"active_band_mae_db":0.9591567466335773}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"facodec","status":"measured","mos":{"status":"measured","value":4.629208564758301,"predicted_mos":4.629208564758301,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49200,"input_sample_rate":16000,"input_seconds":3.075,"input_float32_sha256":"482965d5947dc36085c696973b9f02b8e731306a17f8e03c7cd49585dba60936","internal_samples":49200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.599615200000699},"started_at":"2026-10-13T06:58:02.146776+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.211699913663324e-05,"native_input_samples":49200,"native_input_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","output_sha256":"9866729e1a30fb174f37246acee86f8a33a72141e8713d27995adcf754a66cda","output_samples":49200,"length_delta_samples":0,"analysis_s":0.8499691930010158,"synthesis_s":5.551211869998951,"total_s":6.401181062999967,"rtf":2.0816848985365746,"output_peak":0.43974271416664124,"output_rms":0.03838292411093581,"finished_at":"2026-10-13T06:58:08.556165+00:00","evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"9866729e1a30fb174f37246acee86f8a33a72141e8713d27995adcf754a66cda","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6312118410509715,"spectral_convergence":0.3974074848079168,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.170451527056601,"spectral_convergence":0.4515213311362712,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.494575125516861,"spectral_convergence":0.5004074509486275,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":5.914198107720434e-05,"delta_db":-1.0151543727810444,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.0002314739322992591,"delta_db":-0.8587560909150221,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.00021505460610734889,"delta_db":-0.9527064378142818,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.00023652818452074765,"delta_db":-0.8294057066503142,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0004726305513774813,"delta_db":-1.1562251353581077,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.0002584196079238161,"delta_db":-1.670936853331744,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0805307661417525,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1216272927274609,"pitch":null,"output_rms_dbfs":-28.31723885315239,"output_peak":0.43974271416664124,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.0987461645414776},"common_prefix":{"samples":49200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.6312118410509715,"spectral_convergence":0.3974074848079168,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.170451527056601,"spectral_convergence":0.4515213311362712,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.494575125516861,"spectral_convergence":0.5004074509486275,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":5.914198107720434e-05,"delta_db":-1.0151543727810444,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.0002314739322992591,"delta_db":-0.8587560909150221,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.00021505460610734889,"delta_db":-0.9527064378142818,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.00023652818452074765,"delta_db":-0.8294057066503142,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0004726305513774813,"delta_db":-1.1562251353581077,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.0002584196079238161,"delta_db":-1.670936853331744,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0805307661417525,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1216272927274609,"pitch":null,"output_rms_dbfs":-28.31723885315239,"output_peak":0.43974271416664124,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.0987461645414776},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49200,"output_samples":49200,"intelligibility":{"estoi":{"value":0.8440066429272327,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04367364322520635,"rms_dbfs":-27.195611560424933,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.948879491442757,"mean":-0.00013451925138147867,"zero_energy":false},"output":{"rms":0.03838292411093581,"rms_dbfs":-28.31723885315239,"peak":0.43974271416664124,"peak_dbfs":-7.136026947515149,"crest_factor_db":21.18121190563724,"mean":-0.0004038498008243018,"zero_energy":false},"rms_difference_db":-1.1216272927274566},"full_scale":{"reference":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":5.995011340589871e-07,"output_rms_dbfs":-62.22209991022023,"output_zero_energy":false,"output_peak":0.0019252729834988713,"output_energy_fraction":1.3233318772477574e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":110,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9072164948453608,"absolute_difference_cents_median":2.7286185256614655,"absolute_difference_cents_p95":12.279186651244151,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":0.00010992699921133752,"log_rms_mae_db":1.4841270534502946,"rms_envelope_nrmse":0.1757383709693017,"envelope_correlation":0.9738571231102052,"rms_gain_error_db":-1.121620974135741,"reference_mean_power":0.0019104810338505282,"output_mean_power":0.0014756407275489378},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":4.70123588122935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030753312258320766,"output_power":0.00022051795558855483,"reference_active":true,"delta_db":-1.4444794043148574},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8607043791082874e-05,"output_power":2.7883683912446116e-05,"reference_active":true,"delta_db":-0.11122829825329562},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.145843039363973e-05,"output_power":9.03769009977373e-06,"reference_active":true,"delta_db":-1.0306768549215237}],"active_band_count":3,"active_band_mae_db":0.8621281858298923}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":1.292187656101244,"rms_envelope_nrmse":0.16382963577503723,"envelope_correlation":0.9785503710371816,"rms_gain_error_db":-1.1212429705416704,"reference_mean_power":0.0019290839559002245,"output_mean_power":0.001490139174104116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":4.127760176364045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002739478020838986,"output_power":0.00019750745917545857,"reference_active":true,"delta_db":-1.4208431831119472},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7486141765201905e-05,"output_power":1.702901033321644e-05,"reference_active":true,"delta_db":-0.11504586004480828},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8391125647823e-07,"output_power":4.860306102842452e-07,"reference_active":true,"delta_db":-0.7968322548140505}],"active_band_count":3,"active_band_mae_db":0.7775737659902687}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":1.7109705440631882,"rms_envelope_nrmse":0.17584125749286186,"envelope_correlation":0.9782865081959009,"rms_gain_error_db":-0.9860253478746706,"reference_mean_power":0.0004974887922363912,"output_mean_power":0.00039644301133212226},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.1951630543463553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015634740722861155,"output_power":0.00012319587262113772,"reference_active":true,"delta_db":-1.0349452547109366},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160597194093242e-06,"output_power":5.29903090179415e-06,"reference_active":true,"delta_db":-1.8752548952105832},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0536955073753773e-06,"output_power":2.000898700294525e-06,"reference_active":true,"delta_db":-0.11310950933018489}],"active_band_count":3,"active_band_mae_db":1.0077698864172349}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":1.4224517762184787,"rms_envelope_nrmse":0.16724932895428574,"envelope_correlation":0.9803181228284279,"rms_gain_error_db":-0.9860206054185936,"reference_mean_power":0.00050239739349007,"output_mean_power":0.00040035505555675996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.0265825460069485e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014180308081562638,"output_power":0.00011184864008005836,"reference_active":true,"delta_db":-1.0305495837400453},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469176320839401e-06,"output_power":2.8579779126403636e-06,"reference_active":true,"delta_db":-1.9416862096316154},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.310040798443835e-08,"output_power":8.511864318264076e-08,"reference_active":true,"delta_db":-0.3892689193150549}],"active_band_count":3,"active_band_mae_db":1.120501570895572}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":2.104889841710261,"rms_envelope_nrmse":0.17974618898395442,"envelope_correlation":0.978243210654253,"rms_gain_error_db":-0.9110875897677455,"reference_mean_power":0.0004940955013379761,"output_mean_power":0.0004005918789417358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":1.347463406037001e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.364408948380081e-05,"output_power":7.056704705952512e-05,"reference_active":true,"delta_db":-1.2287842691620252},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.996833569277437e-06,"output_power":5.576488295765094e-06,"reference_active":true,"delta_db":-0.9854074863903796},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2622375016010212e-06,"output_power":1.7861569797657598e-06,"reference_active":true,"delta_db":-1.0261857248645347}],"active_band_count":3,"active_band_mae_db":1.0801258268056466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":1.874939137203511,"rms_envelope_nrmse":0.16158563953976773,"envelope_correlation":0.9835065438186824,"rms_gain_error_db":-0.9110831311605833,"reference_mean_power":0.0004989699589215842,"output_mean_power":0.0004045442997261995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":1.2432388404774247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.827683631948742e-05,"output_power":6.665942200015959e-05,"reference_active":true,"delta_db":-1.2198521684833306},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.844195926765935e-06,"output_power":2.8640720923403304e-06,"reference_active":true,"delta_db":-1.2782156868755439},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.483409815447234e-08,"output_power":8.974071109487667e-08,"reference_active":true,"delta_db":-0.239750120286417}],"active_band_count":3,"active_band_mae_db":0.9126059918817638}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":1.891403587297346,"rms_envelope_nrmse":0.24898997580654894,"envelope_correlation":0.9523348900567513,"rms_gain_error_db":-1.4764919296336845,"reference_mean_power":0.0007033343720650755,"output_mean_power":0.0005006251332353591},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.960535239443851e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011696049836275019,"output_power":8.424368020336417e-05,"reference_active":true,"delta_db":-1.4250187928439955},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1203242070906313e-05,"output_power":2.362745667641623e-05,"reference_active":true,"delta_db":0.47014703928940593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0967126516322994e-05,"output_power":8.534845497303221e-06,"reference_active":true,"delta_db":-1.088971898294975}],"active_band_count":3,"active_band_mae_db":0.9947125768094587}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":1.5287796787784602,"rms_envelope_nrmse":0.22508489657064065,"envelope_correlation":0.9633220729519513,"rms_gain_error_db":-1.4764943222283702,"reference_mean_power":0.0007102740547296545,"output_mean_power":0.0005055644392064122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.637352177112327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010173798550296787,"output_power":7.374511323959865e-05,"reference_active":true,"delta_db":-1.397498871457441},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907094272462648e-05,"output_power":1.4394571046019724e-05,"reference_active":true,"delta_db":0.47370245437850345},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.316189457175963e-07,"output_power":5.425468988020545e-07,"reference_active":true,"delta_db":-0.6601786300868885}],"active_band_count":3,"active_band_mae_db":0.8437933186409444}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.3103108406066895,"predicted_mos":2.3103108406066895,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":135680,"input_sample_rate":44100,"input_seconds":3.076643990929705,"input_float32_sha256":"30a3656f29789fa7c1372994320e4438bef97b07f8441367cc18284e7aec560b","internal_samples":49227,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3747830979991704},"started_at":"2026-10-13T07:26:11.448652+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00540552899838076,"native_input_samples":135608,"native_input_sha256":"f9b874502ba59792092239a74353265b961623abc93239fe09b47d59f0490abf","output_sha256":"b9e8a1dc1d32a96297841affa517d707676db138750dd3576676eed2ddd3f1ab","output_samples":135680,"length_delta_samples":72,"analysis_s":0.14038501899995026,"synthesis_s":24.63427965200026,"total_s":24.77466467100021,"rtf":8.056771812806835,"output_peak":0.510014533996582,"output_rms":0.03821509375575617,"finished_at":"2026-10-13T07:26:36.250501+00:00","evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"b9e8a1dc1d32a96297841affa517d707676db138750dd3576676eed2ddd3f1ab","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":49200,"output_samples":49227,"length_delta_samples":27,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.34957264026082,"output_peak":0.5093112774867404},"common_prefix":{"samples":49200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9994515205070388,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.96954931240918,"spectral_convergence":0.6709840509638478,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.287850909292208,"spectral_convergence":0.700841803346528,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.443131476170814,"spectral_convergence":0.7476270344627689,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":0.00012308689269380118,"delta_db":2.1680050287845396,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.00027389258460208384,"delta_db":-0.12797424354457662,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.0002065558960127024,"delta_db":-1.1278179646531004,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.00015161335199679525,"delta_db":-2.760860194244841,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.000573465676817538,"delta_db":-0.3163687549491647,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.00013450698876900865,"delta_db":-4.5067429839046405,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.834628195013477,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1515848419756092,"pitch":null,"output_rms_dbfs":-28.34719640240054,"output_peak":0.5093112774867404,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.2335105659574},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49200,"output_samples":49200,"intelligibility":{"estoi":{"value":0.52227101012874,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04367364322520635,"rms_dbfs":-27.195611560424933,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.948879491442757,"mean":-0.00013451925138147867,"zero_energy":false},"output":{"rms":0.03825076980783432,"rms_dbfs":-28.34719640240054,"peak":0.5093112774867404,"peak_dbfs":-5.860334145982183,"crest_factor_db":22.486862256418362,"mean":-4.382635407632354e-05,"zero_energy":false},"rms_difference_db":-1.151584841975609},"full_scale":{"reference":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":4.596044958570173e-07,"output_rms_dbfs":-63.376157316888346,"output_zero_energy":false,"output_peak":0.002066227673836884,"output_energy_fraction":1.0215480248727288e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":105,"both_estimated_voiced_frames":77,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.7938144329896907,"absolute_difference_cents_median":12.901976728466227,"absolute_difference_cents_p95":55.548973774216904,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":0.00010992699921133752,"log_rms_mae_db":1.8158143536275537,"rms_envelope_nrmse":0.1751740666904042,"envelope_correlation":0.9771106433303026,"rms_gain_error_db":-1.1515782643943442,"reference_mean_power":0.0019104810338505282,"output_mean_power":0.0014654969002704152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":4.70123588122935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030753312258320766,"output_power":0.00026436093310243494,"reference_active":true,"delta_db":-0.65694621823919},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8607043791082874e-05,"output_power":2.8373783311193816e-05,"reference_active":true,"delta_db":-0.03555733052871013},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.145843039363973e-05,"output_power":7.276544892864577e-06,"reference_active":true,"delta_db":-1.971999178060104}],"active_band_count":3,"active_band_mae_db":0.8881675756093347}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":1.57666310811589,"rms_envelope_nrmse":0.14477189618489048,"envelope_correlation":0.9914619029591869,"rms_gain_error_db":-1.1514476275374146,"reference_mean_power":0.0019290839559002245,"output_mean_power":0.0014798113918865295},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":4.127760176364045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002739478020838986,"output_power":0.00023844324095065363,"reference_active":true,"delta_db":-0.6028280420492445},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7486141765201905e-05,"output_power":1.6142504335480166e-05,"reference_active":true,"delta_db":-0.34723083298727303},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8391125647823e-07,"output_power":3.07268434100622e-07,"reference_active":true,"delta_db":-2.7882890036527286}],"active_band_count":3,"active_band_mae_db":1.246115959563082}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":3.665616366500541,"rms_envelope_nrmse":0.2607715962174702,"envelope_correlation":0.9434765284317951,"rms_gain_error_db":-1.3101049974448817,"reference_mean_power":0.0004974887922363912,"output_mean_power":0.0003679364394772387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.1951630543463553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015634740722861155,"output_power":0.00013355265718195078,"reference_active":true,"delta_db":-0.6843815033430674},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160597194093242e-06,"output_power":7.73024085479796e-06,"reference_active":true,"delta_db":-0.23528916022391438},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0536955073753773e-06,"output_power":2.1546202944179775e-06,"reference_active":true,"delta_db":0.2083469310454626}],"active_band_count":3,"active_band_mae_db":0.37600586487081483}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":3.1678668017701446,"rms_envelope_nrmse":0.23820978868897802,"envelope_correlation":0.9530628508249865,"rms_gain_error_db":-1.3101096722990502,"reference_mean_power":0.00050239739349007,"output_mean_power":0.0003715663791239132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.0265825460069485e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014180308081562638,"output_power":0.00012283119815058435,"reference_active":true,"delta_db":-0.6237697826561049},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469176320839401e-06,"output_power":4.008234537775416e-06,"reference_active":true,"delta_db":-0.4727436316211855},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.310040798443835e-08,"output_power":1.1536114813155153e-07,"reference_active":true,"delta_db":0.9310798556165725}],"active_band_count":3,"active_band_mae_db":0.6758644232979544}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":3.28090216365715,"rms_envelope_nrmse":0.31218880974342733,"envelope_correlation":0.9462481095623305,"rms_gain_error_db":-2.2595905366448834,"reference_mean_power":0.0004940955013379761,"output_mean_power":0.0002936647681727411},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":1.347463406037001e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.364408948380081e-05,"output_power":5.450997635227126e-05,"reference_active":true,"delta_db":-2.350043776735721},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.996833569277437e-06,"output_power":6.651819795285955e-06,"reference_active":true,"delta_db":-0.2196106843793069},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2622375016010212e-06,"output_power":1.9885476112282514e-06,"reference_active":true,"delta_db":-0.5600220386837069}],"active_band_count":3,"active_band_mae_db":1.0432254999329116}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":3.0290540272467923,"rms_envelope_nrmse":0.2876194114796864,"envelope_correlation":0.9616340000344733,"rms_gain_error_db":-2.259596655326911,"reference_mean_power":0.0004989699589215842,"output_mean_power":0.0002965614754078624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":1.2432388404774247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.827683631948742e-05,"output_power":5.098395940268713e-05,"reference_active":true,"delta_db":-2.384132007942365},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.844195926765935e-06,"output_power":3.678077727954465e-06,"reference_active":true,"delta_db":-0.19184611363372184},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.483409815447234e-08,"output_power":8.766020020508983e-08,"reference_active":true,"delta_db":-0.3416206034377817}],"active_band_count":3,"active_band_mae_db":0.9725329083379562}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":3.2893791521278084,"rms_envelope_nrmse":0.3004717811462698,"envelope_correlation":0.9269179513143572,"rms_gain_error_db":-1.6426049914831538,"reference_mean_power":0.0007033343720650755,"output_mean_power":0.0004818383276463866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.960535239443851e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011696049836275019,"output_power":7.116884147901436e-05,"reference_active":true,"delta_db":-2.1574931389663616},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1203242070906313e-05,"output_power":1.380940968393696e-05,"reference_active":true,"delta_db":-1.862271575573977},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0967126516322994e-05,"output_power":6.1668000017040295e-06,"reference_active":true,"delta_db":-2.5003298964935805}],"active_band_count":3,"active_band_mae_db":2.1733648703446398}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":3.0596785677724267,"rms_envelope_nrmse":0.2549915080419038,"envelope_correlation":0.9538408257683231,"rms_gain_error_db":-1.6426017416264573,"reference_mean_power":0.0007102740547296545,"output_mean_power":0.00048659290999157774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.637352177112327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010173798550296787,"output_power":6.346616400112738e-05,"reference_active":true,"delta_db":-2.049408843496327},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907094272462648e-05,"output_power":6.716523779824809e-06,"reference_active":true,"delta_db":-2.836839254150521},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.316189457175963e-07,"output_power":3.0542080259941196e-07,"reference_active":true,"delta_db":-3.1555653432844433}],"active_band_count":3,"active_band_mae_db":2.6806044803104303}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"golf","status":"measured","mos":{"status":"measured","value":2.791105270385742,"predicted_mos":2.791105270385742,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":73680,"input_sample_rate":24000,"input_seconds":3.07,"input_float32_sha256":"472a08cd8ed8ac29e269363536bdb8330d3e00655ce2a627fb12c52921cf543b","internal_samples":49120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.003783895999732},"started_at":"2026-10-13T05:02:54.518782+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001513680999778444,"native_input_samples":73800,"native_input_sha256":"38e6da0cc8ddcfe1e82b5bd6764ca0586dd5946b57117bea3a315cef1c7178cd","output_sha256":"2194323f546ba0836df9b6940d70b968ffa064025a091c405d86a1a5244432ab","output_samples":73680,"length_delta_samples":-120,"analysis_s":0.31440269500035356,"synthesis_s":0.09091012999942905,"total_s":0.4053128249997826,"rtf":0.1318090487804171,"output_peak":0.4453928768634796,"output_rms":0.03765500937222047,"finished_at":"2026-10-13T05:02:54.930747+00:00","analysis_stages":{"external_f0_s":0.04236284299986437},"evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"2194323f546ba0836df9b6940d70b968ffa064025a091c405d86a1a5244432ab","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":49200,"output_samples":49120,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.478630362633254,"output_peak":0.44523211880227337},"common_prefix":{"samples":49120,"sample_rate":16000,"reference_coverage":0.9983739837398374,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49120,"output_samples":49120,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.102230575263173,"spectral_convergence":0.7432851251198614,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.683078197238678,"spectral_convergence":0.4887313071635327,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.07420094118037,"spectral_convergence":0.5495855630008802,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.481093096722808e-05,"output_power":0.00010881330706533495,"delta_db":1.6271695037005993,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002827374966518303,"output_power":0.0003505542160071234,"delta_db":0.9337178615375035,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026805275850349576,"output_power":0.00025463960362793504,"delta_db":-0.22294331273790846,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028681323953908455,"output_power":0.00018282078264240568,"delta_db":-1.955736310618683,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006178117503428883,"output_power":0.0003473842312548041,"delta_db":-2.500460633050033,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0003802432988283916,"output_power":0.00017529298077253078,"delta_db":-3.362970429956402,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.7671663419335217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.290031392573786,"pitch":null,"output_rms_dbfs":-28.478630362633254,"output_peak":0.44523211880227337,"reference_rms_dbfs":-27.18859897005947,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.286503237894074},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49120,"output_samples":49120,"intelligibility":{"estoi":{"value":0.8336935272758954,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04370891756647514,"rms_dbfs":-27.18859897005947,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.941866901077294,"mean":-0.00012906661639384417,"zero_energy":false},"output":{"rms":0.03767632043300054,"rms_dbfs":-28.478630362633254,"peak":0.44523211880227337,"peak_dbfs":-7.028270268802145,"crest_factor_db":21.450360093831108,"mean":-2.4825439661575278e-06,"zero_energy":false},"rms_difference_db":-1.290031392573784},"full_scale":{"reference":{"sample_count":49120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49120,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":3.7261776329384907e-07,"output_rms_dbfs":-64.28736445423097,"output_zero_energy":false,"output_peak":0.001609371048328355,"output_energy_fraction":8.550435108535924e-07,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":91,"both_estimated_voiced_frames":80,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.8247422680412371,"absolute_difference_cents_median":5.608822907177933,"absolute_difference_cents_p95":23.20299247333823,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49120,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00010992699921133752,"log_rms_mae_db":2.237682005338705,"rms_envelope_nrmse":0.22940414093921238,"envelope_correlation":0.946013012150831,"rms_gain_error_db":-1.2899885659948933,"reference_mean_power":0.0019135496138422605,"output_mean_power":0.0014218077276973576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":4.687884958919096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003072220100062265,"output_power":0.0002535156464172019,"reference_active":true,"delta_db":-0.8344755803113879},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.847915955984182e-05,"output_power":2.162042944734428e-05,"reference_active":true,"delta_db":-1.1966285270693007},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.150556761595462e-05,"output_power":8.362821670567288e-06,"reference_active":true,"delta_db":-1.385552130721694}],"active_band_count":3,"active_band_mae_db":1.1388854127007941}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":1.7957542052491204,"rms_envelope_nrmse":0.17805666125897432,"envelope_correlation":0.9774705984540821,"rms_gain_error_db":-1.2896008809233797,"reference_mean_power":0.001932164023539126,"output_mean_power":0.001435766788759319},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":4.1162832554748395e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00027369116890182824,"output_power":0.00023046908214535708,"reference_active":true,"delta_db":-0.7464811218318348},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7386872608465675e-05,"output_power":1.1002483205725921e-05,"reference_active":true,"delta_db":-1.9873075778516547},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8339526086099e-07,"output_power":4.212775625694214e-07,"reference_active":true,"delta_db":-1.4139456783740632}],"active_band_count":3,"active_band_mae_db":1.3825781260191843}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":2.2825843742076652,"rms_envelope_nrmse":0.23518208600425577,"envelope_correlation":0.951413897779325,"rms_gain_error_db":0.4252197941098955,"reference_mean_power":0.000498300190626467,"output_mean_power":0.0005495573652516316},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.195856306468127e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001563551450741453,"output_power":0.00018941625589818923,"reference_active":true,"delta_db":0.8330507101444179},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.134540900150013e-06,"output_power":1.2185050610644782e-05,"reference_active":true,"delta_db":1.7549429025583185},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0573304136365163e-06,"output_power":3.468546842702783e-06,"reference_active":true,"delta_db":2.2684351768414217}],"active_band_count":3,"active_band_mae_db":1.6188095965147193}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":1.6866200095150248,"rms_envelope_nrmse":0.1752604135415791,"envelope_correlation":0.9723254612829995,"rms_gain_error_db":0.42522261164697733,"reference_mean_power":0.000503224821958991,"output_mean_power":0.0005549889241524658},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":2.0272365870415756e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014178964012327475,"output_power":0.00017385984228613434,"reference_active":true,"delta_db":0.8855478109796164},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.432665014355366e-06,"output_power":5.753690698962528e-06,"reference_active":true,"delta_db":1.132815998128964},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.31021748940936e-08,"output_power":1.8816341561797172e-07,"reference_active":true,"delta_db":3.055753616094112}],"active_band_count":3,"active_band_mae_db":1.691372475067564}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":2.3315500931586457,"rms_envelope_nrmse":0.2866208773923718,"envelope_correlation":0.9511298214990054,"rms_gain_error_db":-1.9303889797207057,"reference_mean_power":0.0004949011336985692,"output_mean_power":0.00031730692524024765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.3425669983316813e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.326255278232453e-05,"output_power":6.0574237184003605e-05,"reference_active":true,"delta_db":-1.8741934504807731},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.978169825560327e-06,"output_power":3.6289531967265374e-06,"reference_active":true,"delta_db":-2.8396016746989217},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2606396746930688e-06,"output_power":2.751893140294112e-06,"reference_active":true,"delta_db":0.8540022036342564}],"active_band_count":3,"active_band_mae_db":1.855932442937984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":2.007406358310029,"rms_envelope_nrmse":0.220774507271404,"envelope_correlation":0.9889847077411681,"rms_gain_error_db":-1.9303924388900218,"reference_mean_power":0.0004997917217289496,"output_mean_power":0.00032044228103069453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":1.2389248583605091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.790720813590849e-05,"output_power":5.741594747100611e-05,"reference_active":true,"delta_db":-1.849919515258852},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.831198852857888e-06,"output_power":1.8204807534030644e-06,"reference_active":true,"delta_db":-3.231486021778057},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.393137842745331e-08,"output_power":1.3537938236725773e-07,"reference_active":true,"delta_db":1.5874183282859575}],"active_band_count":3,"active_band_mae_db":2.2229412884409556}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":3.689788348398418,"rms_envelope_nrmse":0.38284942816917017,"envelope_correlation":0.9297827339273897,"rms_gain_error_db":-3.3040756262485758,"reference_mean_power":0.0007044815465097788,"output_mean_power":0.00032920169213834116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.9485119825934036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011672333830879762,"output_power":4.8419603747498785e-05,"reference_active":true,"delta_db":-3.821364689889086},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1085531651685335e-05,"output_power":1.0117167921373638e-05,"reference_active":true,"delta_db":-3.1892559748390195},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0999946048098634e-05,"output_power":7.142688881834848e-06,"reference_active":true,"delta_db":-1.8752882133633322}],"active_band_count":3,"active_band_mae_db":2.961969626030479}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49120,"uncovered_tail_samples":0,"metrics":{"frame_count":607,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":3.6046376654691357,"rms_envelope_nrmse":0.33986962237593243,"envelope_correlation":0.9744284929498346,"rms_gain_error_db":-3.304077328304356,"reference_mean_power":0.0007114440794496708,"output_mean_power":0.0003324551284382729},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":607,"frequency_resolution_hz":0.32948929159802304,"power_floor":2.6265301466478817e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001015593905932603,"output_power":4.368413252961378e-05,"reference_active":true,"delta_db":-3.6639637022287546},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2820661442587352e-05,"output_power":5.394369705533334e-06,"reference_active":true,"delta_db":-3.759697241457236},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.305188289482774e-07,"output_power":3.260228144687606e-07,"reference_active":true,"delta_db":-2.8645006802710133}],"active_band_count":3,"active_band_mae_db":3.429387207985668}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.532458543777466,"predicted_mos":2.532458543777466,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":73920,"input_sample_rate":24000,"input_seconds":3.08,"input_float32_sha256":"42108abb2d8c48a024d6f03b143c0adc8bf15ae209442148ecff2d4250b522ec","internal_samples":49280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.439797705003002},"started_at":"2026-10-13T07:43:37.404305+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0015911369991954416,"native_input_samples":73800,"native_input_sha256":"38e6da0cc8ddcfe1e82b5bd6764ca0586dd5946b57117bea3a315cef1c7178cd","output_sha256":"dbaa3665a3afff65e015341b1ea6c4f9daaa2347093a1dc1b9e3b9e50adafeea","output_samples":73920,"length_delta_samples":120,"analysis_s":1.1411171250001644,"synthesis_s":6.138980559000629,"total_s":7.280097684000793,"rtf":2.3675114419514776,"output_peak":0.490461140871048,"output_rms":0.04085950370473935,"finished_at":"2026-10-13T07:43:44.693385+00:00","analysis_stages":{"external_f0_s":0.7892560530017363},"evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"dbaa3665a3afff65e015341b1ea6c4f9daaa2347093a1dc1b9e3b9e50adafeea","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":49200,"output_samples":49280,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.769136125123673,"output_peak":0.4651498059071199},"common_prefix":{"samples":49200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983766233766234,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.68257985468173,"spectral_convergence":0.7193260451599095,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.158034373839035,"spectral_convergence":0.48049619531995674,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.8000378987626,"spectral_convergence":0.5775110881524108,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":8.21263025591858e-05,"delta_db":0.4107096546208945,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.00027668233484049864,"delta_db":-0.08396266726126769,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.00026655245813126397,"delta_db":-0.020367007093787175,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.00026694808064200716,"delta_db":-0.3039666640078806,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0005040702374988692,"delta_db":-0.8765324947143178,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.0002777591609247464,"delta_db":-1.3575075701530228,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5088410096418617,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5664744066358465,"pitch":null,"output_rms_dbfs":-27.762085967060777,"output_peak":0.4651498059071199,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.880217375761123},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49200,"output_samples":49200,"intelligibility":{"estoi":{"value":0.8639190763268536,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04367364322520635,"rms_dbfs":-27.195611560424933,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.948879491442757,"mean":-0.00013451925138147867,"zero_energy":false},"output":{"rms":0.0409162385196461,"rms_dbfs":-27.762085967060777,"peak":0.4651498059071199,"peak_dbfs":-6.648143118496909,"crest_factor_db":21.113942848563866,"mean":-0.00010381933777625334,"zero_energy":false},"rms_difference_db":-0.5664744066358445},"full_scale":{"reference":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":3.2223006947399173e-07,"output_rms_dbfs":-64.91833935061129,"output_zero_energy":false,"output_peak":0.0015246562760941927,"output_energy_fraction":6.259354397879466e-07,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":75,"both_estimated_voiced_frames":69,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.711340206185567,"absolute_difference_cents_median":9.256611402960715,"absolute_difference_cents_p95":27.477811868282018,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":0.00010992699921133752,"log_rms_mae_db":2.388167216720467,"rms_envelope_nrmse":0.1928679431784863,"envelope_correlation":0.9508255256217698,"rms_gain_error_db":-0.5664481511543669,"reference_mean_power":0.0019104810338505282,"output_mean_power":0.0016768642872431038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":4.70123588122935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030753312258320766,"output_power":0.0002493030758283633,"reference_active":true,"delta_db":-0.91164261198959},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8607043791082874e-05,"output_power":1.778899911541547e-05,"reference_active":true,"delta_db":-2.063214672641746},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.145843039363973e-05,"output_power":7.012323327962068e-06,"reference_active":true,"delta_db":-2.1326319832286282}],"active_band_count":3,"active_band_mae_db":1.7024964226199881}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":1.8118174111550254,"rms_envelope_nrmse":0.11755671326169291,"envelope_correlation":0.9840754902852521,"rms_gain_error_db":-0.5659065969853332,"reference_mean_power":0.0019290839559002245,"output_mean_power":0.0016934035634984144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":4.127760176364045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002739478020838986,"output_power":0.0002251351409635979,"reference_active":true,"delta_db":-0.8522453180481935},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7486141765201905e-05,"output_power":9.686283568460729e-06,"reference_active":true,"delta_db":-2.565368160373593},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8391125647823e-07,"output_power":2.909388436426479e-07,"reference_active":true,"delta_db":-3.025451391718405}],"active_band_count":3,"active_band_mae_db":2.1476882900467307}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":1.7822310912421184,"rms_envelope_nrmse":0.17799118034650854,"envelope_correlation":0.9674963540570697,"rms_gain_error_db":-0.006349982510558167,"reference_mean_power":0.0004974887922363912,"output_mean_power":0.000496761926732123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.1951630543463553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015634740722861155,"output_power":0.00015083816683808134,"reference_active":true,"delta_db":-0.15579437849239325},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160597194093242e-06,"output_power":7.191899929359026e-06,"reference_active":true,"delta_db":-0.5487830583156224},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0536955073753773e-06,"output_power":3.008184757525465e-06,"reference_active":true,"delta_db":1.6576845330353809}],"active_band_count":3,"active_band_mae_db":0.7874206566144655}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":1.1287033852898158,"rms_envelope_nrmse":0.11453255383156015,"envelope_correlation":0.985461401752244,"rms_gain_error_db":-0.006348593061107096,"reference_mean_power":0.00050239739349007,"output_mean_power":0.0005016635166786299},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.0265825460069485e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014180308081562638,"output_power":0.00013770307473137458,"reference_active":true,"delta_db":-0.12742028838568703},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469176320839401e-06,"output_power":3.817748837491499e-06,"reference_active":true,"delta_db":-0.684201354845015},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.310040798443835e-08,"output_power":1.4571652573323161e-07,"reference_active":true,"delta_db":1.9455722381602405}],"active_band_count":3,"active_band_mae_db":0.9190646271303141}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":1.7444996693464905,"rms_envelope_nrmse":0.2020401638579747,"envelope_correlation":0.9629776989853304,"rms_gain_error_db":-0.39352307409318477,"reference_mean_power":0.0004940955013379761,"output_mean_power":0.0004512929948861691},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":1.347463406037001e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.364408948380081e-05,"output_power":8.274207057563129e-05,"reference_active":true,"delta_db":-0.5375398661365087},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.996833569277437e-06,"output_power":6.421402873057022e-06,"reference_active":true,"delta_db":-0.372716256381869},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2622375016010212e-06,"output_power":1.553799283007612e-06,"reference_active":true,"delta_db":-1.6314328079447595}],"active_band_count":3,"active_band_mae_db":0.8472296434877125}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":1.0714995670963243,"rms_envelope_nrmse":0.09228805063655932,"envelope_correlation":0.9934178858031881,"rms_gain_error_db":-0.39352663915203473,"reference_mean_power":0.0004989699589215842,"output_mean_power":0.00045574481383049156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":1.2432388404774247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.827683631948742e-05,"output_power":7.839829498671765e-05,"reference_active":true,"delta_db":-0.5154014273675821},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.844195926765935e-06,"output_power":3.5192472661484606e-06,"reference_active":true,"delta_db":-0.3835573229505324},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.483409815447234e-08,"output_power":5.568032672746396e-08,"reference_active":true,"delta_db":-2.3126274352450875}],"active_band_count":3,"active_band_mae_db":1.0705287285210672}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":2.089203055902271,"rms_envelope_nrmse":0.2629639332414597,"envelope_correlation":0.9380467125846332,"rms_gain_error_db":-1.2880846045646175,"reference_mean_power":0.0007033343720650755,"output_mean_power":0.0005228214310683808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.960535239443851e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011696049836275019,"output_power":7.110471731077625e-05,"reference_active":true,"delta_db":-2.161407959758794},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1203242070906313e-05,"output_power":1.2532984006488723e-05,"reference_active":true,"delta_db":-2.283477861217033},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0967126516322994e-05,"output_power":4.726156655397258e-06,"reference_active":true,"delta_db":-3.6558474063936512}],"active_band_count":3,"active_band_mae_db":2.7002444091231594}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":1.5772491196523617,"rms_envelope_nrmse":0.1857760623714429,"envelope_correlation":0.9789426728909592,"rms_gain_error_db":-1.2880870089607064,"reference_mean_power":0.0007102740547296545,"output_mean_power":0.0005279797304156901},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.637352177112327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010173798550296787,"output_power":6.14409048657699e-05,"reference_active":true,"delta_db":-2.1902553088114347},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907094272462648e-05,"output_power":7.247539688357469e-06,"reference_active":true,"delta_db":-2.5063787984147368},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.316189457175963e-07,"output_power":2.628774374868443e-07,"reference_active":true,"delta_db":-3.807018357703113}],"active_band_count":3,"active_band_mae_db":2.834550821643095}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8337371349334717,"predicted_mos":2.8337371349334717,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":74040,"input_sample_rate":24000,"input_seconds":3.085,"input_float32_sha256":"db55d50e01e1a6f67160de160c65bcc3373c0072eadcffd0c8d73f9d535106b5","internal_samples":49360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.818235939999795},"started_at":"2026-10-13T05:11:14.567958+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002428621000944986,"native_input_samples":73800,"native_input_sha256":"38e6da0cc8ddcfe1e82b5bd6764ca0586dd5946b57117bea3a315cef1c7178cd","output_sha256":"c950f890d6b9760c3edd72de41a6556e01aaaff73e9de3d3d9f8aac899c67513","output_samples":74040,"length_delta_samples":240,"analysis_s":0.3691929329997947,"synthesis_s":0.07627747300102783,"total_s":0.44547040600082255,"rtf":0.14486842471571465,"output_peak":0.470053106546402,"output_rms":0.04240367880562876,"finished_at":"2026-10-13T05:11:15.025476+00:00","analysis_stages":{"external_f0_s":0.047046858999237884},"evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"c950f890d6b9760c3edd72de41a6556e01aaaff73e9de3d3d9f8aac899c67513","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":49200,"output_samples":49360,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.44666446907263,"output_peak":0.4267806008887419},"common_prefix":{"samples":49200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9967585089141004,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.680202668821397,"spectral_convergence":0.3696931593188358,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.3308602074396845,"spectral_convergence":0.3304610517418079,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.1342121280721456,"spectral_convergence":0.40061943346298895,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":7.050814580269335e-05,"delta_db":-0.2517201135084297,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.00027573377132349966,"delta_db":-0.09887737743913876,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.0002594062994985539,"delta_db":-0.13838874216759744,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.0002710668380204389,"delta_db":-0.23747079242640318,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0005694774231635579,"delta_db":-0.3466779380492602,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.0003599136773134331,"delta_db":-0.23220797691835673,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.217557156751531,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23695571275199587,"pitch":null,"output_rms_dbfs":-27.43256727317693,"output_peak":0.4267806008887419,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.715091668111076},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49200,"output_samples":49200,"intelligibility":{"estoi":{"value":0.9073991803710316,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04367364322520635,"rms_dbfs":-27.195611560424933,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.948879491442757,"mean":-0.00013451925138147867,"zero_energy":false},"output":{"rms":0.04249830767362598,"rms_dbfs":-27.43256727317693,"peak":0.4267806008887419,"peak_dbfs":-7.395906589002238,"crest_factor_db":20.03666068417469,"mean":-0.00023023357250830537,"zero_energy":false},"rms_difference_db":-0.23695571275199612},"full_scale":{"reference":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":5.70366310313157e-07,"output_rms_dbfs":-62.438461347641756,"output_zero_energy":false,"output_peak":0.0022993278918615233,"output_energy_fraction":1.0269882445037174e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":88,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.8762886597938144,"absolute_difference_cents_median":0.7436319099197972,"absolute_difference_cents_p95":3.9823277033365008,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":0.00010992699921133752,"log_rms_mae_db":0.8175042006621495,"rms_envelope_nrmse":0.0933699969226577,"envelope_correlation":0.9884550357355829,"rms_gain_error_db":-0.23693328575899206,"reference_mean_power":0.0019104810338505282,"output_mean_power":0.0018090451331912987},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":4.70123588122935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030753312258320766,"output_power":0.0002920052783879335,"reference_active":true,"delta_db":-0.2250119595802833},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8607043791082874e-05,"output_power":2.586256739427866e-05,"reference_active":true,"delta_db":-0.438013453981423},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.145843039363973e-05,"output_power":1.3757595018549205e-05,"reference_active":true,"delta_db":0.7941739011365452}],"active_band_count":3,"active_band_mae_db":0.4857331048994172}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":0.5009161128725667,"rms_envelope_nrmse":0.05061640084495998,"envelope_correlation":0.9970519818351041,"rms_gain_error_db":-0.2366963361173459,"reference_mean_power":0.0019290839559002245,"output_mean_power":0.0018267600084134717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":4.127760176364045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002739478020838986,"output_power":0.00026134100156091893,"reference_active":true,"delta_db":-0.20460269234133488},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7486141765201905e-05,"output_power":1.5484119761741468e-05,"reference_active":true,"delta_db":-0.5280747331200052},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8391125647823e-07,"output_power":6.146226582852636e-07,"reference_active":true,"delta_db":0.22261719097896854}],"active_band_count":3,"active_band_mae_db":0.3184315388134362}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":0.8441985700939107,"rms_envelope_nrmse":0.07507981759059969,"envelope_correlation":0.9943844041060269,"rms_gain_error_db":-0.12754170660258649,"reference_mean_power":0.0004974887922363912,"output_mean_power":0.00048309120446226174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.1951630543463553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015634740722861155,"output_power":0.0001513142986928947,"reference_active":true,"delta_db":-0.14210714228223503},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160597194093242e-06,"output_power":6.809322350027472e-06,"reference_active":true,"delta_db":-0.7861804771588279},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0536955073753773e-06,"output_power":2.7182248265133546e-06,"reference_active":true,"delta_db":1.2174932166613708}],"active_band_count":3,"active_band_mae_db":0.7152602787008112}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":0.39778251042925633,"rms_envelope_nrmse":0.04382767831899271,"envelope_correlation":0.9980912381445576,"rms_gain_error_db":-0.12754319953539495,"reference_mean_power":0.00050239739349007,"output_mean_power":0.0004878575805031311},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.0265825460069485e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014180308081562638,"output_power":0.00013822866118581647,"reference_active":true,"delta_db":-0.11087564757284729},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469176320839401e-06,"output_power":3.542901531728717e-06,"reference_active":true,"delta_db":-1.0086840696136365},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.310040798443835e-08,"output_power":1.3396441861299412e-07,"reference_active":true,"delta_db":1.5803787949831065}],"active_band_count":3,"active_band_mae_db":0.89997950405653}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":0.8423994517734436,"rms_envelope_nrmse":0.10482722237534697,"envelope_correlation":0.9904644742576364,"rms_gain_error_db":-0.26026812183588177,"reference_mean_power":0.0004940955013379761,"output_mean_power":0.00046535468175458173},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":1.347463406037001e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.364408948380081e-05,"output_power":8.970259338881427e-05,"reference_active":true,"delta_db":-0.18675372194497125},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.996833569277437e-06,"output_power":6.164863820803816e-06,"reference_active":true,"delta_db":-0.5497805599675977},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2622375016010212e-06,"output_power":2.240836876246801e-06,"reference_active":true,"delta_db":-0.04127954678818438}],"active_band_count":3,"active_band_mae_db":0.2592712762335844}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":0.5308341487785309,"rms_envelope_nrmse":0.04987558215513644,"envelope_correlation":0.9983996183489409,"rms_gain_error_db":-0.26026737983081166,"reference_mean_power":0.0004989699589215842,"output_mean_power":0.00046994567949313346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":1.2432388404774247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.827683631948742e-05,"output_power":8.43935981744429e-05,"reference_active":true,"delta_db":-0.19537256748267703},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.844195926765935e-06,"output_power":3.394715855917621e-06,"reference_active":true,"delta_db":-0.5400208533058027},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.483409815447234e-08,"output_power":9.232512619064576e-08,"reference_active":true,"delta_db":-0.11644608617664673}],"active_band_count":3,"active_band_mae_db":0.28394650232170887}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":0.9619277247243562,"rms_envelope_nrmse":0.1479687887318286,"envelope_correlation":0.9773534684995164,"rms_gain_error_db":-0.30917861912858363,"reference_mean_power":0.0007033343720650755,"output_mean_power":0.0006550040420748759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.960535239443851e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011696049836275019,"output_power":0.00010023616712253479,"reference_active":true,"delta_db":-0.6701475857652226},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1203242070906313e-05,"output_power":2.203217404120898e-05,"reference_active":true,"delta_db":0.16655082015015626},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0967126516322994e-05,"output_power":1.3151832771954321e-05,"reference_active":true,"delta_db":0.7889342459229429}],"active_band_count":3,"active_band_mae_db":0.5418775506127739}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":0.5684140915814022,"rms_envelope_nrmse":0.08168938134076789,"envelope_correlation":0.9934048401897919,"rms_gain_error_db":-0.30917752979582375,"reference_mean_power":0.0007102740547296545,"output_mean_power":0.0006614670233679056},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.637352177112327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010173798550296787,"output_power":8.747952037733627e-05,"reference_active":true,"delta_db":-0.6557674071499365},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907094272462648e-05,"output_power":1.3367728952842767e-05,"reference_active":true,"delta_db":0.1522914899811995},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.316189457175963e-07,"output_power":6.328626232149142e-07,"reference_active":true,"delta_db":0.00854298649799533}],"active_band_count":3,"active_band_mae_db":0.2722006278763771}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_6930-81414-0026","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2715024948120117,"predicted_mos":3.2715024948120117,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":147914,"input_sample_rate":48000,"input_seconds":3.081541666666667,"input_float32_sha256":"7ffe791e65bf0191b6448d00f98e5146b2d2bc0f62cd51ea575bb450508310bf","internal_samples":49305,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5523330830001214},"started_at":"2026-10-13T05:17:33.516607+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.006166573999507818,"native_input_samples":147600,"native_input_sha256":"38903b66eaff51d731980be2cb7b28675d4a335385d098c7816c30e4d83574ac","output_sha256":"ceaa7868fc152724257bb9921f52d4b4846dfb46e6c8a7e54403df470578f643","output_samples":147914,"length_delta_samples":314,"analysis_s":1.6822869480001827,"synthesis_s":0.7185532950006746,"total_s":2.4008402430008573,"rtf":0.7807610546344251,"output_peak":0.5656013856445774,"output_rms":0.043231840713294244,"finished_at":"2026-10-13T05:17:35.934480+00:00","evaluation_input":{"reference":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"output_sha256":"ceaa7868fc152724257bb9921f52d4b4846dfb46e6c8a7e54403df470578f643","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":49200,"output_samples":49305,"length_delta_samples":105,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.278769413157804,"output_peak":0.5662759619273144},"common_prefix":{"samples":49200,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9978703985397018,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49200,"output_samples":49200,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.434936318108338,"spectral_convergence":0.5762513079531735,"spectral_convergence_status":"ok","amplitude_floor":6.39186387874099e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.183664167665441,"spectral_convergence":0.37405518805549165,"spectral_convergence_status":"ok","amplitude_floor":4.1641566181737356e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.515693819823125,"spectral_convergence":0.39177848709074825,"spectral_convergence_status":"ok","amplitude_floor":3.471921787916526e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.471560290376807e-05,"output_power":6.440782555874408e-05,"delta_db":-0.6447266786263093,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0002820835097806379,"output_power":0.0002597039620779665,"delta_db":-0.35899123297232577,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026780543872196867,"output_power":0.0002549774429934648,"delta_db":-0.21317631082658556,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00028630139823372175,"output_power":0.00027916512949529605,"delta_db":-0.10962299393287502,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0006167999147673381,"output_power":0.0006263597099408029,"delta_db":0.06679509292852033,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00037968124815517825,"output_power":0.0003905848948525956,"delta_db":0.12296295176821434,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25271254350913835,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07391489343351404,"pitch":null,"output_rms_dbfs":-27.269526453858447,"output_peak":0.5662759619273144,"reference_rms_dbfs":-27.195611560424933,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.711431435198968},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49200,"output_samples":49200,"intelligibility":{"estoi":{"value":0.9357606955521756,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04367364322520635,"rms_dbfs":-27.195611560424933,"peak":0.434173583984375,"peak_dbfs":-7.246732068982174,"crest_factor_db":19.948879491442757,"mean":-0.00013451925138147867,"zero_energy":false},"output":{"rms":0.04330356757726631,"rms_dbfs":-27.269526453858447,"peak":0.5662759619273144,"peak_dbfs":-4.939437470414754,"crest_factor_db":22.33008898344369,"mean":0.0008693802173009802,"zero_energy":false},"rms_difference_db":-0.0739148934335141},"full_scale":{"reference":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49200,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":307,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0349273029714824e-06,"threshold_dbfs":-59.850901555473264,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":1,"quiet_samples":160,"reference_mean_power":6.552261766046286e-07,"output_mean_power":3.524585039470408e-07,"output_rms_dbfs":-64.52892006524054,"output_zero_energy":false,"output_peak":0.002335226768485368,"output_energy_fraction":6.112452802844502e-07,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":300,"reference_estimated_voiced_frames":97,"output_estimated_voiced_frames":90,"both_estimated_voiced_frames":87,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.8969072164948454,"absolute_difference_cents_median":3.1843269571714146,"absolute_difference_cents_p95":14.424242558059003,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49200,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":0.00010992699921133752,"log_rms_mae_db":1.2570217858000967,"rms_envelope_nrmse":0.13859057483448559,"envelope_correlation":0.9734772517626188,"rms_gain_error_db":-0.07391319445894624,"reference_mean_power":0.0019104810338505282,"output_mean_power":0.0018782414094531965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":4.70123588122935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030753312258320766,"output_power":0.00029507273360838466,"reference_active":true,"delta_db":-0.17962817849522392},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8607043791082874e-05,"output_power":3.114550202972565e-05,"reference_active":true,"delta_db":0.3692235481004377},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.145843039363973e-05,"output_power":1.4064476963554448e-05,"reference_active":true,"delta_db":0.8899845516316749}],"active_band_count":3,"active_band_mae_db":0.4796120927424455}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":9.464213695185308e-05,"log_rms_mae_db":0.7687296085647183,"rms_envelope_nrmse":0.06998907161789657,"envelope_correlation":0.9926416000404557,"rms_gain_error_db":-0.07365006303581895,"reference_mean_power":0.0019290839559002245,"output_mean_power":0.001896645315629235},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":4.127760176364045e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002739478020838986,"output_power":0.00026255710717990955,"reference_active":true,"delta_db":-0.18444041652736137},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7486141765201905e-05,"output_power":1.914513459069949e-05,"reference_active":true,"delta_db":0.3936442877634427},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.8391125647823e-07,"output_power":7.463330075789712e-07,"reference_active":true,"delta_db":1.0658580176170802}],"active_band_count":3,"active_band_mae_db":0.547980907302628}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":6.478025356347348e-05,"log_rms_mae_db":1.3334264983613089,"rms_envelope_nrmse":0.11042797729334018,"envelope_correlation":0.9882748223262395,"rms_gain_error_db":-0.2936244342282799,"reference_mean_power":0.0004974887922363912,"output_mean_power":0.00046496563687240265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.1951630543463553e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00015634740722861155,"output_power":0.00014547601747772937,"reference_active":true,"delta_db":-0.3129928017333005},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.160597194093242e-06,"output_power":7.428702933274332e-06,"reference_active":true,"delta_db":-0.4080894999669987},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0536955073753773e-06,"output_power":2.0361418411701972e-06,"reference_active":true,"delta_db":-0.037280246291545804}],"active_band_count":3,"active_band_mae_db":0.25278751599728166}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":5.647638237474489e-05,"log_rms_mae_db":0.7659335448371198,"rms_envelope_nrmse":0.0655590614346912,"envelope_correlation":0.996352222206431,"rms_gain_error_db":-0.2936246644113422,"reference_mean_power":0.00050239739349007,"output_mean_power":0.00046955331515308375},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.0265825460069485e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014180308081562638,"output_power":0.00013208517754752976,"reference_active":true,"delta_db":-0.30831582134588037},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.469176320839401e-06,"output_power":4.0195606893941156e-06,"reference_active":true,"delta_db":-0.4604889881636332},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.310040798443835e-08,"output_power":9.422531956959272e-08,"reference_active":true,"delta_db":0.052160349042539515}],"active_band_count":3,"active_band_mae_db":0.27365505285068437}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":7.855020595588601e-05,"log_rms_mae_db":1.3615964925336679,"rms_envelope_nrmse":0.1585529661336056,"envelope_correlation":0.9770524728701779,"rms_gain_error_db":-0.04782203093426147,"reference_mean_power":0.0004940955013379761,"output_mean_power":0.0004886846488152021},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":1.347463406037001e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.364408948380081e-05,"output_power":9.113731855730006e-05,"reference_active":true,"delta_db":-0.11784124674290108},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.996833569277437e-06,"output_power":6.584674320299559e-06,"reference_active":true,"delta_db":-0.263672441445027},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2622375016010212e-06,"output_power":2.7690813302201732e-06,"reference_active":true,"delta_db":0.8779751401903767}],"active_band_count":3,"active_band_mae_db":0.4198296094594349}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":6.305842206002451e-05,"log_rms_mae_db":0.7871712983096902,"rms_envelope_nrmse":0.0743520059468629,"envelope_correlation":0.9947086388745011,"rms_gain_error_db":-0.04782606647930819,"reference_mean_power":0.0004989699589215842,"output_mean_power":0.0004935052675141632},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":1.2432388404774247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.827683631948742e-05,"output_power":8.593690491720845e-05,"reference_active":true,"delta_db":-0.11667052263347993},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.844195926765935e-06,"output_power":3.67426900492703e-06,"reference_active":true,"delta_db":-0.19634564990960776},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.483409815447234e-08,"output_power":1.1228174435714073e-07,"reference_active":true,"delta_db":0.7334463247469343}],"active_band_count":3,"active_band_mae_db":0.34882083243000733}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":614,"rms_floor":9.512970769085293e-05,"log_rms_mae_db":1.4585545500931005,"rms_envelope_nrmse":0.19100091784967915,"envelope_correlation":0.96226443974096,"rms_gain_error_db":0.10395585306374747,"reference_mean_power":0.0007033343720650755,"output_mean_power":0.0007203730007267766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":614,"frequency_resolution_hz":0.3257328990228013,"power_floor":2.960535239443851e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011696049836275019,"output_power":0.00010777820798966957,"reference_active":true,"delta_db":-0.3550825174653291},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1203242070906313e-05,"output_power":2.3773464113310477e-05,"reference_active":true,"delta_db":0.496901973012588},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0967126516322994e-05,"output_power":1.307897157987924e-05,"reference_active":true,"delta_db":0.7648074256995887}],"active_band_count":3,"active_band_mae_db":0.5389306387258352}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49200,"uncovered_tail_samples":0,"metrics":{"frame_count":608,"rms_floor":7.035164526339387e-05,"log_rms_mae_db":0.9485144958331708,"rms_envelope_nrmse":0.10422447130354295,"envelope_correlation":0.9877124931473542,"rms_gain_error_db":0.10395296100311352,"reference_mean_power":0.0007102740547296545,"output_mean_power":0.0007274803162468188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":608,"frequency_resolution_hz":0.32894736842105265,"power_floor":2.637352177112327e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010173798550296787,"output_power":9.368805322682918e-05,"reference_active":true,"delta_db":-0.3579891928360509},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907094272462648e-05,"output_power":1.4647077161766375e-05,"reference_active":true,"delta_db":0.5492248733600353},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.316189457175963e-07,"output_power":7.478845783678847e-07,"reference_active":true,"delta_db":0.7337942961122063}],"active_band_count":3,"active_band_mae_db":0.5470027874360975}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"hnm","status":"measured","mos":{"status":"measured","value":2.531749725341797,"predicted_mos":2.531749725341797,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"3a1fb88d92649ab8e926e07c2b304e5d0cdba0aa7a63e79b641fdbbc83aad035","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9123326029985037},"started_at":"2026-10-13T04:39:57.707577+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4827999621047638e-05,"native_input_samples":96000,"native_input_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","output_sha256":"3d7473b32ad73091e9c22779d30915c69dd4dd3ee46474d19a20914d884a4b2b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0587184239993803,"synthesis_s":0.5279064180003843,"total_s":1.5866248419997646,"rtf":0.26443747366662745,"output_peak":0.5376258399908612,"output_rms":0.06423281802908444,"finished_at":"2026-10-13T04:39:59.302426+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"3d7473b32ad73091e9c22779d30915c69dd4dd3ee46474d19a20914d884a4b2b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.735293205364369,"spectral_convergence":0.3106172394175445,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.231766097002522,"spectral_convergence":0.3334575763874893,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.25233821193176,"spectral_convergence":0.35248710797212546,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.002216872708784896,"delta_db":-0.08782994986494307,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.0013803493349359088,"delta_db":-0.31115652685471606,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.00026783462177684865,"delta_db":0.11847445719916547,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00019555418975042457,"delta_db":-0.19370494687754275,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":3.7575532370752684e-05,"delta_db":-0.3562234581313364,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":2.7668524338644657e-05,"delta_db":-0.05229413683229432,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18661391262666635,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15821189319927648,"pitch":null,"output_rms_dbfs":-23.844860483076452,"output_peak":0.5376258399908612,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4064658380995505},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.735293205364369,"spectral_convergence":0.3106172394175445,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.231766097002522,"spectral_convergence":0.3334575763874893,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.25233821193176,"spectral_convergence":0.35248710797212546,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.002216872708784896,"delta_db":-0.08782994986494307,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.0013803493349359088,"delta_db":-0.31115652685471606,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.00026783462177684865,"delta_db":0.11847445719916547,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00019555418975042457,"delta_db":-0.19370494687754275,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":3.7575532370752684e-05,"delta_db":-0.3562234581313364,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":2.7668524338644657e-05,"delta_db":-0.05229413683229432,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18661391262666635,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15821189319927648,"pitch":null,"output_rms_dbfs":-23.844860483076452,"output_peak":0.5376258399908612,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.4064658380995505},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.871711677059177,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06541352763725852,"rms_dbfs":-23.686648589877173,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.82005406670475,"mean":-7.879892985026041e-06,"zero_energy":false},"output":{"rms":0.06423281802908444,"rms_dbfs":-23.84486048307645,"peak":0.5376258399908612,"peak_dbfs":-5.390397317843819,"crest_factor_db":18.45446316523263,"mean":-0.00021783835646793995,"zero_energy":false},"rms_difference_db":-0.15821189319927598},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":2.2104303774200233e-06,"output_rms_dbfs":-56.555231596445495,"output_zero_energy":false,"output_peak":0.04351941268830927,"output_energy_fraction":0.00016340401674310153,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":140,"both_estimated_voiced_frames":119,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.9596774193548387,"absolute_difference_cents_median":1.8617718000916539,"absolute_difference_cents_p95":9.824113043344337,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022610981418459577,"log_rms_mae_db":0.9140403907447994,"rms_envelope_nrmse":0.09689488845114815,"envelope_correlation":0.9916351361332028,"rms_gain_error_db":-0.1581288356917777,"reference_mean_power":0.00428070632223443,"output_mean_power":0.00412764701484682},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.20587323032393e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001500467490243704,"output_power":0.0014266932002760914,"reference_active":true,"delta_db":-0.21895998888850565},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013253274491379155,"output_power":0.00012700865767135985,"reference_active":true,"delta_db":-0.18489866747166045},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.272532196136399e-05,"output_power":3.202154331775036e-05,"reference_active":true,"delta_db":-0.09441668515367219}],"active_band_count":3,"active_band_mae_db":0.16609178050461276}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019856877416737668,"log_rms_mae_db":0.5217024354801417,"rms_envelope_nrmse":0.047026356523282825,"envelope_correlation":0.9981447526856431,"rms_gain_error_db":-0.1583257009765977,"reference_mean_power":0.004296662851053013,"output_mean_power":0.004142845209044397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.009968768189794e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014002280466658247,"output_power":0.0013346573270990598,"reference_active":true,"delta_db":-0.20828997327900067},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3360463722840526e-05,"output_power":5.152084980613652e-05,"reference_active":true,"delta_db":-0.15236577267000861},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4980780249701638e-06,"output_power":2.1467262610976705e-06,"reference_active":true,"delta_db":-0.6582932991667187}],"active_band_count":3,"active_band_mae_db":0.33964968170524273}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000212457290437723,"log_rms_mae_db":1.5944543649980676,"rms_envelope_nrmse":0.1811911505660728,"envelope_correlation":0.9792325947603187,"rms_gain_error_db":-0.5997103564824972,"reference_mean_power":0.0012391682631200216,"output_mean_power":0.0010793424210657808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.921692896601052e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036532110371103373,"output_power":0.000321020968975444,"reference_active":true,"delta_db":-0.5614135977940757},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3180096390035205e-05,"output_power":3.66964224607256e-05,"reference_active":true,"delta_db":-0.7065988057887228},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1750484667690233e-05,"output_power":9.946364756333627e-06,"reference_active":true,"delta_db":-0.7239139837630393}],"active_band_count":3,"active_band_mae_db":0.6639754624486126}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018128574308840577,"log_rms_mae_db":1.0578235589777996,"rms_envelope_nrmse":0.1275814574661858,"envelope_correlation":0.991068309947381,"rms_gain_error_db":-0.6003177752341586,"reference_mean_power":0.0012444705639415084,"output_mean_power":0.0010838092442518133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.573537659747316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003540640123239248,"output_power":0.0003118387219357002,"reference_active":true,"delta_db":-0.5515174473152497},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0107771514146375e-05,"output_power":1.7383216492178056e-05,"reference_active":true,"delta_db":-0.6323380272327324},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973740035567709e-07,"output_power":5.149923925728406e-07,"reference_active":true,"delta_db":-1.898612587155286}],"active_band_count":3,"active_band_mae_db":1.0274893539010892}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013864336468723026,"log_rms_mae_db":1.8784965928110309,"rms_envelope_nrmse":0.2030090100123162,"envelope_correlation":0.9722310614193305,"rms_gain_error_db":-0.22593010251876505,"reference_mean_power":0.00021050674147203496,"output_mean_power":0.00019983566424164282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.902091296877062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.015318920526363e-05,"output_power":8.458816385730507e-05,"reference_active":true,"delta_db":0.23388790773223617},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860315173964141e-05,"output_power":3.821875273499471e-05,"reference_active":true,"delta_db":-0.043462541319140055},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0682238748218518e-05,"output_power":2.2603470919593207e-05,"reference_active":true,"delta_db":0.38577586037810224}],"active_band_count":3,"active_band_mae_db":0.2210421031431595}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":1.510906769429778,"rms_envelope_nrmse":0.1267387616746077,"envelope_correlation":0.9887924611957538,"rms_gain_error_db":-0.22954483607306597,"reference_mean_power":0.00021105284570735848,"output_mean_power":0.0002001873952151319},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7773203833688672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041444617967505e-05,"output_power":9.714904300483919e-05,"reference_active":true,"delta_db":0.31200700375217677},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4052463298203542e-05,"output_power":2.405691637806105e-05,"reference_active":true,"delta_db":0.0008039796098502879},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.463215485080669e-06,"output_power":1.615999687784227e-06,"reference_active":true,"delta_db":0.43132983938685415}],"active_band_count":3,"active_band_mae_db":0.24804694091629376}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":1.8339759840595589,"rms_envelope_nrmse":0.19378140926940196,"envelope_correlation":0.9764151216490776,"rms_gain_error_db":0.010418399998661234,"reference_mean_power":2.4404568133397866e-05,"output_mean_power":2.4463183146588728e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1035922908644866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.10428529724051e-05,"output_power":1.1729877453757677e-05,"reference_active":true,"delta_db":0.26212185020686124},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8700267025557585e-06,"output_power":4.168283253350011e-06,"reference_active":true,"delta_db":0.3224326194148374},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.131168166122287e-06,"output_power":8.683454171745738e-07,"reference_active":true,"delta_db":-1.1483465792780267}],"active_band_count":3,"active_band_mae_db":0.5776336829665751}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":1.4727172153716297,"rms_envelope_nrmse":0.10914815662920622,"envelope_correlation":0.992201757244525,"rms_gain_error_db":0.00932613360863318,"reference_mean_power":2.4522755214627912e-05,"output_mean_power":2.4575472492265448e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9657184130662422e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.186833957836496e-05,"output_power":1.2600769464208063e-05,"reference_active":true,"delta_db":0.26007102149806116},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.99155667709149e-06,"output_power":2.075883862510603e-06,"reference_active":true,"delta_db":0.18010382425291918},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.546174499350968e-08,"output_power":6.031693123874742e-08,"reference_active":true,"delta_db":-1.5133251847447138}],"active_band_count":3,"active_band_mae_db":0.651166676831898}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.068368911743164,"predicted_mos":3.068368911743164,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"816ff0d8ed05d9712994b1dc62ca4b887e84ab17cef0458d5a8f402ec20ad580","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8648083320003934},"started_at":"2026-10-13T04:43:09.435009+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3178999981610104e-05,"native_input_samples":96000,"native_input_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","output_sha256":"2db56137324c6aad303091b151f7d801743f61863fd11af5c6aa15e36b791f45","output_samples":96000,"length_delta_samples":0,"analysis_s":1.577466921999985,"synthesis_s":0.09361828900000546,"total_s":1.6710852109999905,"rtf":0.27851420183333175,"output_peak":0.6586828823428932,"output_rms":0.07339802401935872,"finished_at":"2026-10-13T04:43:11.113116+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"2db56137324c6aad303091b151f7d801743f61863fd11af5c6aa15e36b791f45","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.202090568366254,"spectral_convergence":0.6334113377590362,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.9095305119031725,"spectral_convergence":0.30870955036742903,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.944388007820064,"spectral_convergence":0.31187067585556877,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.002637091529321741,"delta_db":0.6660145179284449,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.002062395032348805,"delta_db":1.4326719090822997,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.0003635516987805258,"delta_db":1.4454690844832543,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00024555978750967115,"delta_db":0.7951962906114325,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":4.458378825385737e-05,"delta_db":0.386494792088774,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":3.40880937317584e-05,"delta_db":0.8538730440617942,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9299532730426666,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0003359546726545,"pitch":null,"output_rms_dbfs":-22.686312635204523,"output_peak":0.6586828823428932,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.352003029363163},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.202090568366254,"spectral_convergence":0.6334113377590362,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.9095305119031725,"spectral_convergence":0.30870955036742903,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.944388007820064,"spectral_convergence":0.31187067585556877,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.002637091529321741,"delta_db":0.6660145179284449,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.002062395032348805,"delta_db":1.4326719090822997,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.0003635516987805258,"delta_db":1.4454690844832543,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00024555978750967115,"delta_db":0.7951962906114325,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":4.458378825385737e-05,"delta_db":0.386494792088774,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":3.40880937317584e-05,"delta_db":0.8538730440617942,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9299532730426666,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0003359546726545,"pitch":null,"output_rms_dbfs":-22.686312635204523,"output_peak":0.6586828823428932,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.352003029363163},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9135670355247597,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06541352763725852,"rms_dbfs":-23.686648589877173,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.82005406670475,"mean":-7.879892985026041e-06,"zero_energy":false},"output":{"rms":0.07339802401935873,"rms_dbfs":-22.686312635204523,"peak":0.6586828823428932,"peak_dbfs":-3.6264724545831584,"crest_factor_db":19.05984018062136,"mean":1.2409857121461753e-06,"zero_energy":false},"rms_difference_db":1.00033595467265},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":3.2282727439395933e-06,"output_rms_dbfs":-54.91029780584397,"output_zero_energy":false,"output_peak":0.026722047488257433,"output_energy_fraction":0.00018276848936570354,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":148,"both_estimated_voiced_frames":118,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9516129032258065,"absolute_difference_cents_median":8.755284476960185,"absolute_difference_cents_p95":44.2617917291409,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022610981418459577,"log_rms_mae_db":1.6253652440096198,"rms_envelope_nrmse":0.20026323947906335,"envelope_correlation":0.9800816754970383,"rms_gain_error_db":1.0021506352312801,"reference_mean_power":0.00428070632223443,"output_mean_power":0.0053917593188353334},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.20587323032393e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001500467490243704,"output_power":0.001864596185292822,"reference_active":true,"delta_db":0.9435820112209993},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013253274491379155,"output_power":0.00014566617826795137,"reference_active":true,"delta_db":0.41035533296285964},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.272532196136399e-05,"output_power":2.728296834131573e-05,"reference_active":true,"delta_db":-0.7899230889279705}],"active_band_count":3,"active_band_mae_db":0.7146201443706097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019856877416737668,"log_rms_mae_db":1.1978004541481349,"rms_envelope_nrmse":0.16398655229434067,"envelope_correlation":0.9901113065892118,"rms_gain_error_db":1.0053942022321345,"reference_mean_power":0.004296662851053013,"output_mean_power":0.005415900752148473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.009968768189794e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014002280466658247,"output_power":0.0017372916244695895,"reference_active":true,"delta_db":0.9367395356058249},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3360463722840526e-05,"output_power":6.037980849412686e-05,"reference_active":true,"delta_db":0.5367213613289998},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4980780249701638e-06,"output_power":1.5420845497864527e-06,"reference_active":true,"delta_db":-2.0949781303287427}],"active_band_count":3,"active_band_mae_db":1.1894796757545225}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000212457290437723,"log_rms_mae_db":1.906051609621235,"rms_envelope_nrmse":0.31337987198700434,"envelope_correlation":0.9728950255022106,"rms_gain_error_db":1.726192113132556,"reference_mean_power":0.0012391682631200216,"output_mean_power":0.001843951497189167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.921692896601052e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036532110371103373,"output_power":0.0004965533201188267,"reference_active":true,"delta_db":1.332911290003809},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3180096390035205e-05,"output_power":6.097694806370772e-05,"reference_active":true,"delta_db":1.4988207638343054},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1750484667690233e-05,"output_power":1.301610720364999e-05,"reference_active":true,"delta_db":0.44425336739194327}],"active_band_count":3,"active_band_mae_db":1.0919951404100192}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018128574308840577,"log_rms_mae_db":1.4832983207196813,"rms_envelope_nrmse":0.27139175887037803,"envelope_correlation":0.9859930246767367,"rms_gain_error_db":1.7283006862059298,"reference_mean_power":0.0012444705639415084,"output_mean_power":0.0018527409352644161},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.573537659747316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003540640123239248,"output_power":0.00048209527426686994,"reference_active":true,"delta_db":1.3405108777531622},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0107771514146375e-05,"output_power":2.749111206789704e-05,"reference_active":true,"delta_db":1.3582836656964896},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973740035567709e-07,"output_power":6.740445886520016e-07,"reference_active":true,"delta_db":-0.7297344598640301}],"active_band_count":3,"active_band_mae_db":1.1428430011045607}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013864336468723026,"log_rms_mae_db":1.6534785222981274,"rms_envelope_nrmse":0.22973607414066827,"envelope_correlation":0.9731055426442282,"rms_gain_error_db":0.774408302438437,"reference_mean_power":0.00021050674147203496,"output_mean_power":0.00025159779978958547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.902091296877062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.015318920526363e-05,"output_power":0.00010224380858876577,"reference_active":true,"delta_db":1.0571621149137367},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860315173964141e-05,"output_power":4.814002329475596e-05,"reference_active":true,"delta_db":0.9588353216524746},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0682238748218518e-05,"output_power":1.6165795687476707e-05,"reference_active":true,"delta_db":-1.0700046152111402}],"active_band_count":3,"active_band_mae_db":1.0286673505924504}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":1.27734303093213,"rms_envelope_nrmse":0.14038037858776659,"envelope_correlation":0.9929211652050494,"rms_gain_error_db":0.7785710788044969,"reference_mean_power":0.00021105284570735848,"output_mean_power":0.00025249240566688956},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7773203833688672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041444617967505e-05,"output_power":0.00011208593117820824,"reference_active":true,"delta_db":0.9331327772758695},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4052463298203542e-05,"output_power":2.7795697418719595e-05,"reference_active":true,"delta_db":0.628180147171965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.463215485080669e-06,"output_power":1.133088016750862e-06,"reference_active":true,"delta_db":-1.1104464200886934}],"active_band_count":3,"active_band_mae_db":0.8905864481788427}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":1.700556996520735,"rms_envelope_nrmse":0.25762211775020427,"envelope_correlation":0.9714531154174022,"rms_gain_error_db":0.9895585267001534,"reference_mean_power":2.4404568133397866e-05,"output_mean_power":3.064975302936667e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1035922908644866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.10428529724051e-05,"output_power":1.3688728472239513e-05,"reference_active":true,"delta_db":0.9328181908322672},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8700267025557585e-06,"output_power":5.17791784938255e-06,"reference_active":true,"delta_db":1.264411942334174},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.131168166122287e-06,"output_power":1.9689154748509457e-06,"reference_active":true,"delta_db":2.4069989786177626}],"active_band_count":3,"active_band_mae_db":1.534743037261401}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":1.2421378065280901,"rms_envelope_nrmse":0.17952020553606488,"envelope_correlation":0.9899783062152681,"rms_gain_error_db":0.9887979541162656,"reference_mean_power":2.4522755214627912e-05,"output_mean_power":3.079279129373473e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9657184130662422e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.186833957836496e-05,"output_power":1.5206440317045022e-05,"reference_active":true,"delta_db":1.076375978367941},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.99155667709149e-06,"output_power":2.7741379746591807e-06,"reference_active":true,"delta_db":1.4393538699290973},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.546174499350968e-08,"output_power":1.5159886471692714e-07,"reference_active":true,"delta_db":2.489241928164377}],"active_band_count":3,"active_band_mae_db":1.6683239254871385}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"facodec","status":"measured","mos":{"status":"measured","value":3.989194631576538,"predicted_mos":3.989194631576538,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"dea478732c125737661888103005299fdf45ace8cf47580c29ba7a12b02c6ff9","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.737328470997454},"started_at":"2026-10-13T06:58:08.557375+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.4570999812567607e-05,"native_input_samples":96000,"native_input_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","output_sha256":"e2377a6f6d6fe98de578ed2f185bda33d0317f7556a201ceef23776816ad90d4","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6281497479994869,"synthesis_s":12.227641030000086,"total_s":13.855790777999573,"rtf":2.3092984629999287,"output_peak":0.4845239818096161,"output_rms":0.06137206564794856,"finished_at":"2026-10-13T06:58:22.423720+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"e2377a6f6d6fe98de578ed2f185bda33d0317f7556a201ceef23776816ad90d4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.279274311175278,"spectral_convergence":0.24136021881686714,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.275648743923861,"spectral_convergence":0.2459927480828328,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.115400191659111,"spectral_convergence":0.2608763934414672,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.0018596930821820415,"delta_db":-0.8508247638939437,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.0014704285965283542,"delta_db":-0.03660722850745233,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.00023548569703791655,"delta_db":-0.4405473585001627,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.0001438419541684101,"delta_db":-1.5275204529148787,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":3.392584296191514e-05,"delta_db":-0.7999684074296199,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":2.3155269017469726e-05,"delta_db":-0.8256557997976546,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7468540018406187,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5539365916958578,"pitch":null,"output_rms_dbfs":-24.24058518157303,"output_peak":0.4845239818096161,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.223441082252751},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.279274311175278,"spectral_convergence":0.24136021881686714,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.275648743923861,"spectral_convergence":0.2459927480828328,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.115400191659111,"spectral_convergence":0.2608763934414672,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.0018596930821820415,"delta_db":-0.8508247638939437,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.0014704285965283542,"delta_db":-0.03660722850745233,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.00023548569703791655,"delta_db":-0.4405473585001627,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.0001438419541684101,"delta_db":-1.5275204529148787,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":3.392584296191514e-05,"delta_db":-0.7999684074296199,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":2.3155269017469726e-05,"delta_db":-0.8256557997976546,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7468540018406187,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5539365916958578,"pitch":null,"output_rms_dbfs":-24.24058518157303,"output_peak":0.4845239818096161,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.223441082252751},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9034291837924631,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06541352763725852,"rms_dbfs":-23.686648589877173,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.82005406670475,"mean":-7.879892985026041e-06,"zero_energy":false},"output":{"rms":0.06137206564794856,"rms_dbfs":-24.24058518157303,"peak":0.4845239818096161,"peak_dbfs":-6.293694448226112,"crest_factor_db":17.94689073334692,"mean":6.94532552590689e-05,"zero_energy":false},"rms_difference_db":-0.5539365916958587},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":4.3840267242485157e-07,"output_rms_dbfs":-63.581268071335664,"output_zero_energy":false,"output_peak":0.005649163853377104,"output_energy_fraction":3.5500261355186986e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":131,"both_estimated_voiced_frames":119,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9596774193548387,"absolute_difference_cents_median":3.915741972139696,"absolute_difference_cents_p95":25.017056871384405,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022610981418459577,"log_rms_mae_db":1.211042987317129,"rms_envelope_nrmse":0.13177272237566823,"envelope_correlation":0.986919472720133,"rms_gain_error_db":-0.5544678826711127,"reference_mean_power":0.00428070632223443,"output_mean_power":0.0037676334656772865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.20587323032393e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001500467490243704,"output_power":0.001176261914103886,"reference_active":true,"delta_db":-1.057225549239227},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013253274491379155,"output_power":0.00012088588404821275,"reference_active":true,"delta_db":-0.3994760195709173},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.272532196136399e-05,"output_power":3.344762997529821e-05,"reference_active":true,"delta_db":0.09481422091609056}],"active_band_count":3,"active_band_mae_db":0.5171719299087449}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019856877416737668,"log_rms_mae_db":1.0194699481089053,"rms_envelope_nrmse":0.12169494885606301,"envelope_correlation":0.9886737871995569,"rms_gain_error_db":-0.5564823819901084,"reference_mean_power":0.004296662851053013,"output_mean_power":0.003779923746384066},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.009968768189794e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014002280466658247,"output_power":0.0010982652647349648,"reference_active":true,"delta_db":-1.054915241349548},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3360463722840526e-05,"output_power":4.764018066558753e-05,"reference_active":true,"delta_db":-0.4924619532369207},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4980780249701638e-06,"output_power":2.602854566487139e-06,"reference_active":true,"delta_db":0.1784390374347975}],"active_band_count":3,"active_band_mae_db":0.5752720773404221}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000212457290437723,"log_rms_mae_db":1.3729902496760313,"rms_envelope_nrmse":0.14745293956017488,"envelope_correlation":0.9851502979415583,"rms_gain_error_db":-0.05183739977884869,"reference_mean_power":0.0012391682631200216,"output_mean_power":0.001224465468610135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.921692896601052e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036532110371103373,"output_power":0.00037814689788246623,"reference_active":true,"delta_db":0.14985780952259034},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3180096390035205e-05,"output_power":4.6191839826952514e-05,"reference_active":true,"delta_db":0.2928165313053674},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1750484667690233e-05,"output_power":1.276296931693831e-05,"reference_active":true,"delta_db":0.3589594501074767}],"active_band_count":3,"active_band_mae_db":0.26721126364514486}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018128574308840577,"log_rms_mae_db":1.1199339465095857,"rms_envelope_nrmse":0.12300704689549985,"envelope_correlation":0.9892472091591314,"rms_gain_error_db":-0.0524388617568151,"reference_mean_power":0.0012444705639415084,"output_mean_power":0.001229534565191341},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.573537659747316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003540640123239248,"output_power":0.0003665971375242607,"reference_active":true,"delta_db":0.15107283004964386},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0107771514146375e-05,"output_power":2.039175912638695e-05,"reference_active":true,"delta_db":0.060907508050509755},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973740035567709e-07,"output_power":8.315580753636501e-07,"reference_active":true,"delta_db":0.1823051296051188}],"active_band_count":3,"active_band_mae_db":0.1314284892350908}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013864336468723026,"log_rms_mae_db":1.4106460624260204,"rms_envelope_nrmse":0.23994342940636706,"envelope_correlation":0.9717741854661547,"rms_gain_error_db":-1.3698769622234979,"reference_mean_power":0.00021050674147203496,"output_mean_power":0.0001535600738972749},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.902091296877062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.015318920526363e-05,"output_power":4.499403155213884e-05,"reference_active":true,"delta_db":-2.50765898589494},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860315173964141e-05,"output_power":2.115130826111996e-05,"reference_active":true,"delta_db":-2.612855291893921},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0682238748218518e-05,"output_power":1.11659751054178e-05,"reference_active":true,"delta_db":-2.6770089190322155}],"active_band_count":3,"active_band_mae_db":2.599174398940359}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":1.1634777887150585,"rms_envelope_nrmse":0.22169292983958927,"envelope_correlation":0.9769948899657046,"rms_gain_error_db":-1.3671798307866339,"reference_mean_power":0.00021105284570735848,"output_mean_power":0.00015405408867733982},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7773203833688672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041444617967505e-05,"output_power":4.8930252299487963e-05,"reference_active":true,"delta_db":-2.666603713142503},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4052463298203542e-05,"output_power":1.2613295198798244e-05,"reference_active":true,"delta_db":-2.8033100064470706},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.463215485080669e-06,"output_power":7.008894043426663e-07,"reference_active":true,"delta_db":-3.196587939829185}],"active_band_count":3,"active_band_mae_db":2.88883388647292}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":1.7983390127016035,"rms_envelope_nrmse":0.21571453119921097,"envelope_correlation":0.9730966202122534,"rms_gain_error_db":-0.7803369229389308,"reference_mean_power":2.4404568133397866e-05,"output_mean_power":2.0390948812496388e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1035922908644866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.10428529724051e-05,"output_power":8.840788034717293e-06,"reference_active":true,"delta_db":-0.9659031176692111},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8700267025557585e-06,"output_power":3.6140642068510375e-06,"reference_active":true,"delta_db":-0.29718097695171414},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.131168166122287e-06,"output_power":1.2681536228074874e-06,"reference_active":true,"delta_db":0.49644692212257985}],"active_band_count":3,"active_band_mae_db":0.5865103389145017}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":1.5315679816942134,"rms_envelope_nrmse":0.15117186996180615,"envelope_correlation":0.9887610786299775,"rms_gain_error_db":-0.7802797828673448,"reference_mean_power":2.4522755214627912e-05,"output_mean_power":2.04899682175282e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9657184130662422e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.186833957836496e-05,"output_power":9.829099510693661e-06,"reference_active":true,"delta_db":-0.8187623198202625},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.99155667709149e-06,"output_power":1.8718334598411447e-06,"reference_active":true,"delta_db":-0.269254641594062},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.546174499350968e-08,"output_power":9.844274372455783e-08,"reference_active":true,"delta_db":0.6141195336204274}],"active_band_count":3,"active_band_mae_db":0.5673788316782505}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.3075954914093018,"predicted_mos":3.3075954914093018,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"c5eec2da4284c1168188ef5a0cb910b55a583d81469051099e44080a54e6daaa","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4894754939996346},"started_at":"2026-10-13T07:26:36.253805+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0084322440015967,"native_input_samples":264600,"native_input_sha256":"d4d8b3c9cb395334a6116b8673273da900fc7028b4760437d055952e19a67032","output_sha256":"9e93e3a933a6f8e281e1c455838e27a3433b1d34f2d76a07b14210e3d40b9ea4","output_samples":264704,"length_delta_samples":104,"analysis_s":0.24651892499969108,"synthesis_s":70.77146992299822,"total_s":71.01798884799791,"rtf":11.836331474666318,"output_peak":0.5083937048912048,"output_rms":0.05858988568473661,"finished_at":"2026-10-13T07:27:47.309715+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"9e93e3a933a6f8e281e1c455838e27a3433b1d34f2d76a07b14210e3d40b9ea4","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.640645414730677,"output_peak":0.5068955726209743},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.714257775692753,"spectral_convergence":0.48084575702402704,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.1248149996152,"spectral_convergence":0.38834153927599474,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.7252362181093455,"spectral_convergence":0.4544237987148467,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.0014684338616002554,"delta_db":-1.8766936130943013,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.001519510792253883,"delta_db":0.10599125957429287,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.000293062039493841,"delta_db":0.5094029777113687,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00010462318533814455,"delta_db":-2.9100968149694433,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":3.347354678967201e-05,"delta_db":-0.8582575644151659,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":1.716869613327316e-05,"delta_db":-2.124780967946322,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3975371996184824,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9524759230427887,"pitch":null,"output_rms_dbfs":-24.639124512919963,"output_peak":0.5068955726209743,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.188102997805766},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7823394421312669,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06541352763725852,"rms_dbfs":-23.686648589877173,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.82005406670475,"mean":-7.879892985026041e-06,"zero_energy":false},"output":{"rms":0.05861972468042706,"rms_dbfs":-24.639124512919963,"peak":0.5068955726209743,"peak_dbfs":-5.901630040384095,"crest_factor_db":18.737494472535868,"mean":-0.00013560582234002698,"zero_energy":false},"rms_difference_db":-0.9524759230427904},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":3.225473786116178e-07,"output_rms_dbfs":-64.9140648332856,"output_zero_energy":false,"output_peak":0.011668582790499378,"output_energy_fraction":2.8628975527839577e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":116,"both_estimated_voiced_frames":94,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.7580645161290323,"absolute_difference_cents_median":13.681411826354804,"absolute_difference_cents_p95":41.87627772963921,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022610981418459577,"log_rms_mae_db":1.4745638446644227,"rms_envelope_nrmse":0.15449997196759488,"envelope_correlation":0.9871226474388884,"rms_gain_error_db":-0.9510099139878645,"reference_mean_power":0.00428070632223443,"output_mean_power":0.003438859581493307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.20587323032393e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001500467490243704,"output_power":0.0012393778770648939,"reference_active":true,"delta_db":-0.8302285054456516},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013253274491379155,"output_power":0.00010214611370437009,"reference_active":true,"delta_db":-1.1310134490173416},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.272532196136399e-05,"output_power":2.734287571920436e-05,"reference_active":true,"delta_db":-0.7803973948099743}],"active_band_count":3,"active_band_mae_db":0.9138797830909892}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019856877416737668,"log_rms_mae_db":1.2708313319992965,"rms_envelope_nrmse":0.12791297843550403,"envelope_correlation":0.9941584769202303,"rms_gain_error_db":-0.9495116842637055,"reference_mean_power":0.004296662851053013,"output_mean_power":0.003452869053857727},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.009968768189794e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014002280466658247,"output_power":0.0011594441944458174,"reference_active":true,"delta_db":-0.8194892203207677},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3360463722840526e-05,"output_power":3.835079190654913e-05,"reference_active":true,"delta_db":-1.4344525897385816},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4980780249701638e-06,"output_power":2.1159380274589575e-06,"reference_active":true,"delta_db":-0.7210305526597166}],"active_band_count":3,"active_band_mae_db":0.9916574542396885}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000212457290437723,"log_rms_mae_db":2.270850022276278,"rms_envelope_nrmse":0.20325710574183542,"envelope_correlation":0.9718957408107021,"rms_gain_error_db":0.05169075095798738,"reference_mean_power":0.0012391682631200216,"output_mean_power":0.0012540052568553275},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.921692896601052e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036532110371103373,"output_power":0.0003872193250084264,"reference_active":true,"delta_db":0.2528226250021456},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3180096390035205e-05,"output_power":5.078564440266035e-05,"reference_active":true,"delta_db":0.7045735997969274},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1750484667690233e-05,"output_power":1.0478322599759347e-05,"reference_active":true,"delta_db":-0.4976401506059187}],"active_band_count":3,"active_band_mae_db":0.48501212513499725}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018128574308840577,"log_rms_mae_db":1.826135737386477,"rms_envelope_nrmse":0.1592761173420946,"envelope_correlation":0.9818078029074919,"rms_gain_error_db":0.05253004208058791,"reference_mean_power":0.0012444705639415084,"output_mean_power":0.0012596144458864463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.573537659747316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003540640123239248,"output_power":0.0003730838909599889,"reference_active":true,"delta_db":0.22724710888730165},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0107771514146375e-05,"output_power":2.2872570962620605e-05,"reference_active":true,"delta_db":0.5595104200750001},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973740035567709e-07,"output_power":7.767585864477302e-07,"reference_active":true,"delta_db":-0.11376009706873609}],"active_band_count":3,"active_band_mae_db":0.30017254201034593}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013864336468723026,"log_rms_mae_db":2.5484779527490993,"rms_envelope_nrmse":0.36673284067498996,"envelope_correlation":0.9421097140121174,"rms_gain_error_db":-2.68712405974719,"reference_mean_power":0.00021050674147203496,"output_mean_power":0.00011338447739758771},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.902091296877062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.015318920526363e-05,"output_power":3.294369876731112e-05,"reference_active":true,"delta_db":-3.8614844896041163},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860315173964141e-05,"output_power":1.90206766451148e-05,"reference_active":true,"delta_db":-3.0739680137762058},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0682238748218518e-05,"output_power":7.070015289137626e-06,"reference_active":true,"delta_db":-4.661771941837475}],"active_band_count":3,"active_band_mae_db":3.8657414817392657}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":2.374587142897651,"rms_envelope_nrmse":0.3287783144769309,"envelope_correlation":0.9637497238950281,"rms_gain_error_db":-2.685963012856206,"reference_mean_power":0.00021105284570735848,"output_mean_power":0.00011370901853876466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7773203833688672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041444617967505e-05,"output_power":3.632476017439221e-05,"reference_active":true,"delta_db":-3.9603507078311564},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4052463298203542e-05,"output_power":1.1548001970120307e-05,"reference_active":true,"delta_db":-3.1865271132761928},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.463215485080669e-06,"output_power":4.690047959681419e-07,"reference_active":true,"delta_db":-4.941310048291473}],"active_band_count":3,"active_band_mae_db":4.029395956466274}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":2.955563869403045,"rms_envelope_nrmse":0.23803736707459633,"envelope_correlation":0.9638300321230399,"rms_gain_error_db":-0.3656848699697484,"reference_mean_power":2.4404568133397866e-05,"output_mean_power":2.243378946019097e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1035922908644866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.10428529724051e-05,"output_power":9.063075237935664e-06,"reference_active":true,"delta_db":-0.8580570454544387},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8700267025557585e-06,"output_power":3.252088145520736e-06,"reference_active":true,"delta_db":-0.7555165326810305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.131168166122287e-06,"output_power":1.05137715568677e-06,"reference_active":true,"delta_db":-0.3176863806255773}],"active_band_count":3,"active_band_mae_db":0.6437533195870154}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":2.7561512523484817,"rms_envelope_nrmse":0.16726823762472728,"envelope_correlation":0.9818063760319271,"rms_gain_error_db":-0.3664615762602086,"reference_mean_power":2.4522755214627912e-05,"output_mean_power":2.2538401201892184e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9657184130662422e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.186833957836496e-05,"output_power":9.485395756835485e-06,"reference_active":true,"delta_db":-0.9733450834411699},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.99155667709149e-06,"output_power":1.7578460253425614e-06,"reference_active":true,"delta_db":-0.5421183903917711},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.546174499350968e-08,"output_power":9.06510732218645e-08,"reference_active":true,"delta_db":0.2560119391522771}],"active_band_count":3,"active_band_mae_db":0.590491804328406}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"golf","status":"measured","mos":{"status":"measured","value":3.2727272510528564,"predicted_mos":3.2727272510528564,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"3c59e0ddf9ef9b01bf3417479d9b5e2d28738d80cbc6f30338cc355bbc5c6c58","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8534381989993562},"started_at":"2026-10-13T05:02:54.931667+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027496080001583323,"native_input_samples":144000,"native_input_sha256":"729be1fd14fe8a47de44c43192f9e85a9ee8211fe00d0838329954c95a472c70","output_sha256":"0ba726f54af74940e970c7c549bd113ed87cef76643d79ab362615c13abb5d07","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5668356389996916,"synthesis_s":0.12090914500004146,"total_s":0.687744783999733,"rtf":0.11462413066662218,"output_peak":0.5104880332946777,"output_rms":0.06193267766445641,"finished_at":"2026-10-13T05:02:55.629639+00:00","analysis_stages":{"external_f0_s":0.08711805399980221},"evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"0ba726f54af74940e970c7c549bd113ed87cef76643d79ab362615c13abb5d07","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.159686499124604,"output_peak":0.5109238873107558},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.199813108523429,"spectral_convergence":0.5956367000779533,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.955642770156788,"spectral_convergence":0.29491526838258375,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8908463120200985,"spectral_convergence":0.2972241326712096,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0022636510116587805,"output_power":0.0018130590390579224,"delta_db":-0.9639752589732171,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014845545807864492,"output_power":0.0016705678902640517,"delta_db":0.5126796024575448,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002606411042999767,"output_power":0.0001938496977193814,"delta_db":-1.2857777888922093,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020459129843142628,"output_power":0.00011073221854581729,"delta_db":-2.666131573927202,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.08198593522311e-05,"output_power":2.3195114218338418e-05,"delta_db":-2.454749885588257,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8040897623340483e-05,"output_power":2.5945488964101775e-05,"delta_db":-0.33730052068150906,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.37010243841999,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4764561223687942,"pitch":null,"output_rms_dbfs":-24.159686499124604,"output_peak":0.5109238873107558,"reference_rms_dbfs":-23.683230376755805,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.348767396900105},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8917474599391306,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06543927530277367,"rms_dbfs":-23.683230376755805,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.816635853583378,"mean":4.8409718304922265e-06,"zero_energy":false},"output":{"rms":0.061946343304263025,"rms_dbfs":-24.159686499124604,"peak":0.5109238873107558,"peak_dbfs":-5.832875843996802,"crest_factor_db":18.326810655127797,"mean":0.0007245107273386531,"zero_energy":false},"rms_difference_db":-0.4764561223687984},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":7.206453190574409e-07,"output_rms_dbfs":-61.42278429976848,"output_zero_energy":false,"output_peak":0.01151568033450688,"output_energy_fraction":5.73739152128312e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":146,"both_estimated_voiced_frames":117,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.9435483870967742,"absolute_difference_cents_median":5.463275010443153,"absolute_difference_cents_p95":24.032409096737045,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00022610981418459577,"log_rms_mae_db":2.4729743602642644,"rms_envelope_nrmse":0.1340609038469911,"envelope_correlation":0.9861389312369558,"rms_gain_error_db":-0.47670434829553693,"reference_mean_power":0.004284168753688167,"output_mean_power":0.0038388057315264694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.2061395013535486e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015013448850666811,"output_power":0.0013671059701151039,"reference_active":true,"delta_db":-0.4067828885122133},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001326639835719793,"output_power":0.0001264167047998711,"reference_active":true,"delta_db":-0.20948568206798357},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2386318171994066e-05,"output_power":2.521427617489084e-05,"reference_active":true,"delta_db":-1.087150727884145}],"active_band_count":3,"active_band_mae_db":0.5678064328214473}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00019856877416737668,"log_rms_mae_db":2.3351863803002506,"rms_envelope_nrmse":0.09601826290897925,"envelope_correlation":0.9939751166970486,"rms_gain_error_db":-0.47821520325287076,"reference_mean_power":0.004299729728452919,"output_mean_power":0.003851408967945289},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.010088899343981e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001400890511393977,"output_power":0.0012782821937361987,"reference_active":true,"delta_db":-0.39777454393862555},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.328857187223474e-05,"output_power":5.087027454406625e-05,"reference_active":true,"delta_db":-0.2016999985228984},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5018639163665303e-06,"output_power":2.0824630721007777e-06,"reference_active":true,"delta_db":-0.7968637457323924}],"active_band_count":3,"active_band_mae_db":0.46544609606463877}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.000212457290437723,"log_rms_mae_db":1.799954148959027,"rms_envelope_nrmse":0.17299469795434066,"envelope_correlation":0.9801080291478351,"rms_gain_error_db":-0.32188477528084797,"reference_mean_power":0.0012406298393014506,"output_mean_power":0.0011520033578385413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.91928035457325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036510709303612295,"output_power":0.0003903106624205881,"reference_active":true,"delta_db":0.28990145213888685},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3229729469038054e-05,"output_power":4.3557460536168354e-05,"reference_active":true,"delta_db":0.03280034082393663},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1589397039157059e-05,"output_power":8.119013744796836e-06,"reference_active":true,"delta_db":-1.5455756490969865}],"active_band_count":3,"active_band_mae_db":0.62275914735327}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00018128574308840577,"log_rms_mae_db":1.4371433391800008,"rms_envelope_nrmse":0.12245127987319894,"envelope_correlation":0.990149907520655,"rms_gain_error_db":-0.3218261149642984,"reference_mean_power":0.0012458178477013433,"output_mean_power":0.0011568363774494318},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.570964792299031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003538275485239925,"output_power":0.0003782776493523293,"reference_active":true,"delta_db":0.29019038173527295},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0021527213158215e-05,"output_power":2.025223857023456e-05,"reference_active":true,"delta_db":0.049758329473763865},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.928011173242173e-07,"output_power":5.338688400530245e-07,"reference_active":true,"delta_db":-1.7172968010793113}],"active_band_count":3,"active_band_mae_db":0.6857485040961161}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00013864336468723026,"log_rms_mae_db":1.921140985153701,"rms_envelope_nrmse":0.34236432445444565,"envelope_correlation":0.9544599099712007,"rms_gain_error_db":-2.5926018903710575,"reference_mean_power":0.00021056893674005817,"output_mean_power":0.00011591352552307733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.9025488622563334e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.006439307578228e-05,"output_power":3.613797294822047e-05,"reference_active":true,"delta_db":-3.454756273945711},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.861288884018985e-05,"output_power":1.4769239080181957e-05,"reference_active":true,"delta_db":-4.1737417398831145},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0688851654363862e-05,"output_power":4.171752364395582e-06,"reference_active":true,"delta_db":-6.954178648309782}],"active_band_count":3,"active_band_mae_db":4.860892220712869}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":1.781186294445194,"rms_envelope_nrmse":0.30860564263376355,"envelope_correlation":0.9729620604351491,"rms_gain_error_db":-2.5984660789726437,"reference_mean_power":0.00021083146468861384,"output_mean_power":0.00011590143631150202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.7777493375724432e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.031116109928719e-05,"output_power":3.726049200166496e-05,"reference_active":true,"delta_db":-3.8449284070004075},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4046523839998478e-05,"output_power":8.09073898870844e-06,"reference_active":true,"delta_db":-4.7306411289494354},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4618519806290058e-06,"output_power":2.3116821277039422e-07,"reference_active":true,"delta_db":-8.009752851381688}],"active_band_count":3,"active_band_mae_db":5.528440795777176}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":2.356699322771582,"rms_envelope_nrmse":0.2494832787562895,"envelope_correlation":0.9668313853920824,"rms_gain_error_db":0.20480104077333083,"reference_mean_power":2.4444187936330816e-05,"output_mean_power":2.5624518775736433e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0955943284884355e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.0999691706719759e-05,"output_power":1.1773539388642465e-05,"reference_active":true,"delta_db":0.29526527920359225},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.866688940307464e-06,"output_power":5.309822443340437e-06,"reference_active":true,"delta_db":1.377407626072886},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1280054685240527e-06,"output_power":2.338596960529542e-06,"reference_active":true,"delta_db":3.166441757998884}],"active_band_count":3,"active_band_mae_db":1.6130382210917873}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":2.3056991666712245,"rms_envelope_nrmse":0.18623681246082588,"envelope_correlation":0.9825728360044077,"rms_gain_error_db":0.205572373798177,"reference_mean_power":2.4553157706240946e-05,"output_mean_power":2.574332210595301e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.9580557323996584e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.182828305738263e-05,"output_power":1.3654272338377634e-05,"reference_active":true,"delta_db":0.6234685184345298},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.986249909442059e-06,"output_power":3.0822366088698233e-06,"reference_active":true,"delta_db":1.9083208400069676},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.527573416085312e-08,"output_power":1.9109167265292707e-07,"reference_active":true,"delta_db":3.5041629483755106}],"active_band_count":3,"active_band_mae_db":2.011984102272336}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.714480400085449,"predicted_mos":3.714480400085449,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"f7dac8abbf871b19b8498f4d29095e9bff63a9fa3e77a6c3e41e96d430361328","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2044454059978307},"started_at":"2026-10-13T07:43:44.694671+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003223914998670807,"native_input_samples":144000,"native_input_sha256":"729be1fd14fe8a47de44c43192f9e85a9ee8211fe00d0838329954c95a472c70","output_sha256":"7c85316419830d85fa0e24c563725f30faee97f375b653d9e2c199b1e16cdf7d","output_samples":144120,"length_delta_samples":120,"analysis_s":2.235451254000509,"synthesis_s":15.650117180000962,"total_s":17.88556843400147,"rtf":2.9809280723335783,"output_peak":0.4698544442653656,"output_rms":0.06756976319563272,"finished_at":"2026-10-13T07:44:02.597940+00:00","analysis_stages":{"external_f0_s":1.4330090090006706},"evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"7c85316419830d85fa0e24c563725f30faee97f375b653d9e2c199b1e16cdf7d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.403553389643207,"output_peak":0.4731423251431839},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.124685874750626,"spectral_convergence":0.6142324403204653,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.755955367995611,"spectral_convergence":0.33287786717736273,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.545032100577918,"spectral_convergence":0.31488700301797395,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.002798491862400063,"delta_db":0.9240029643626004,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.001297752254953139,"delta_db":-0.5791287085105652,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.00024594232115668753,"delta_db":-0.2518600237114151,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00016545760716772055,"delta_db":-0.9195088052808325,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":3.8011373281188483e-05,"delta_db":-0.30613928057437934,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":2.2551390704925547e-05,"delta_db":-0.9404208174099573,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6535100999749583,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.284106158416801,"pitch":null,"output_rms_dbfs":-23.402542431460375,"output_peak":0.4731423251431839,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.1418911144413855},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8915534261514246,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06541352763725852,"rms_dbfs":-23.686648589877173,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.82005406670475,"mean":-7.879892985026041e-06,"zero_energy":false},"output":{"rms":0.06758851092947472,"rms_dbfs":-23.402542431460375,"peak":0.4731423251431839,"peak_dbfs":-6.500164004408519,"crest_factor_db":16.902378427051854,"mean":0.00017121849480072071,"zero_energy":false},"rms_difference_db":0.2841061584167974},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":1.0191900187269814e-05,"output_rms_dbfs":-49.91744838180841,"output_zero_energy":false,"output_peak":0.11052677432755671,"output_energy_fraction":0.0006804704092077036,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":160,"both_estimated_voiced_frames":121,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.9758064516129032,"absolute_difference_cents_median":8.225371784684144,"absolute_difference_cents_p95":39.4247002246563,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022610981418459577,"log_rms_mae_db":2.2175361707027976,"rms_envelope_nrmse":0.19700637250825573,"envelope_correlation":0.9669231020785429,"rms_gain_error_db":0.28402901685663207,"reference_mean_power":0.00428070632223443,"output_mean_power":0.004570022473718076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.20587323032393e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001500467490243704,"output_power":0.0016691281070187337,"reference_active":true,"delta_db":0.46263080176824023},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013253274491379155,"output_power":0.00013737583288767367,"reference_active":true,"delta_db":0.15587145670332112},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.272532196136399e-05,"output_power":3.0176031512408904e-05,"reference_active":true,"delta_db":-0.352218033758005}],"active_band_count":3,"active_band_mae_db":0.3235734307431888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019856877416737668,"log_rms_mae_db":1.8708968816010936,"rms_envelope_nrmse":0.14418066994927595,"envelope_correlation":0.9814525994313923,"rms_gain_error_db":0.279595606641117,"reference_mean_power":0.004296662851053013,"output_mean_power":0.0045823772231910675},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.009968768189794e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014002280466658247,"output_power":0.0015518538123389432,"reference_active":true,"delta_db":0.44652035108384724},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3360463722840526e-05,"output_power":5.3093470738941725e-05,"reference_active":true,"delta_db":-0.021784788572004266},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4980780249701638e-06,"output_power":2.399275677120938e-06,"reference_active":true,"delta_db":-0.17525847670634362}],"active_band_count":3,"active_band_mae_db":0.21452120545406506}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000212457290437723,"log_rms_mae_db":2.116553107799212,"rms_envelope_nrmse":0.22181828828522304,"envelope_correlation":0.9680813528359957,"rms_gain_error_db":-0.7297347597859895,"reference_mean_power":0.0012391682631200216,"output_mean_power":0.0010475066916870528},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.921692896601052e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036532110371103373,"output_power":0.0003040706459920715,"reference_active":true,"delta_db":-0.797002643048032},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3180096390035205e-05,"output_power":2.9638515305447737e-05,"reference_active":true,"delta_db":-1.6342716290895776},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1750484667690233e-05,"output_power":6.731956811139208e-06,"reference_active":true,"delta_db":-2.419144590564914}],"active_band_count":3,"active_band_mae_db":1.616806287567508}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018128574308840577,"log_rms_mae_db":1.6445461802302856,"rms_envelope_nrmse":0.15528711261422304,"envelope_correlation":0.9861542159586218,"rms_gain_error_db":-0.7298195421729413,"reference_mean_power":0.0012444705639415084,"output_mean_power":0.0010519683515529282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.573537659747316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003540640123239248,"output_power":0.00029156787286235955,"reference_active":true,"delta_db":-0.8434211810709914},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0107771514146375e-05,"output_power":1.3130841622971964e-05,"reference_active":true,"delta_db":-1.8507137852704825},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973740035567709e-07,"output_power":4.061880896569664e-07,"reference_active":true,"delta_db":-2.929348876699539}],"active_band_count":3,"active_band_mae_db":1.8744946143470045}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013864336468723026,"log_rms_mae_db":2.1387333509610533,"rms_envelope_nrmse":0.28868063752249035,"envelope_correlation":0.9441474357686135,"rms_gain_error_db":-0.7358127649288648,"reference_mean_power":0.00021050674147203496,"output_mean_power":0.00017769889703168323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.902091296877062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.015318920526363e-05,"output_power":6.4594395020395e-05,"reference_active":true,"delta_db":-0.9372597205859162},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860315173964141e-05,"output_power":3.115263628980471e-05,"reference_active":true,"delta_db":-0.931279592346749},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0682238748218518e-05,"output_power":1.0319823514476713e-05,"reference_active":true,"delta_db":-3.019252769346333}],"active_band_count":3,"active_band_mae_db":1.629264027426333}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":1.5874377661497125,"rms_envelope_nrmse":0.17885665547104387,"envelope_correlation":0.9803603370898576,"rms_gain_error_db":-0.7360286111671746,"reference_mean_power":0.00021105284570735848,"output_mean_power":0.0001781510355615726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7773203833688672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041444617967505e-05,"output_power":7.009337014762144e-05,"reference_active":true,"delta_db":-1.1056088471040557},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4052463298203542e-05,"output_power":1.776308825683087e-05,"reference_active":true,"delta_db":-1.3164108700630157},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.463215485080669e-06,"output_power":7.811174865122307e-07,"reference_active":true,"delta_db":-2.7259192833079586}],"active_band_count":3,"active_band_mae_db":1.71597966682501}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":2.1090851876208623,"rms_envelope_nrmse":0.28460876643089095,"envelope_correlation":0.9486199949308329,"rms_gain_error_db":-0.6861820703353876,"reference_mean_power":2.4404568133397866e-05,"output_mean_power":2.083785058675003e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1035922908644866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.10428529724051e-05,"output_power":9.184800364870315e-06,"reference_active":true,"delta_db":-0.8001156869947413},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8700267025557585e-06,"output_power":2.8586596562993165e-06,"reference_active":true,"delta_db":-1.31551508992253},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.131168166122287e-06,"output_power":9.128798578301119e-07,"reference_active":true,"delta_db":-0.931135497757733}],"active_band_count":3,"active_band_mae_db":1.0155887582250014}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":1.560161261745503,"rms_envelope_nrmse":0.172965688574755,"envelope_correlation":0.982631645978367,"rms_gain_error_db":-0.6861430986113243,"reference_mean_power":2.4522755214627912e-05,"output_mean_power":2.0938952571239632e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9657184130662422e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.186833957836496e-05,"output_power":9.699986935482484e-06,"reference_active":true,"delta_db":-0.8761881457564614},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.99155667709149e-06,"output_power":1.4740982917063455e-06,"reference_active":true,"delta_db":-1.3066622747212597},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.546174499350968e-08,"output_power":6.94111114657593e-08,"reference_active":true,"delta_db":-0.9034275745167855}],"active_band_count":3,"active_band_mae_db":1.0287593316648356}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.2064802646636963,"predicted_mos":3.2064802646636963,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"9cec2fcf5170a97ff0ceb3c62dedaf860e0603604bf088732b4e6208261ca97e","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.075388324001324},"started_at":"2026-10-13T05:11:15.027290+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0042922019983961945,"native_input_samples":144000,"native_input_sha256":"729be1fd14fe8a47de44c43192f9e85a9ee8211fe00d0838329954c95a472c70","output_sha256":"6dc8647adcf43a194f27defa4e025a3b0d687eb8320f29e99e40bba6bb108523","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9556829299999663,"synthesis_s":0.11617356199894857,"total_s":1.0718564919989149,"rtf":0.1786427486664858,"output_peak":0.48828184604644775,"output_rms":0.06400156221170634,"finished_at":"2026-10-13T05:11:16.110706+00:00","analysis_stages":{"external_f0_s":0.17438164899976982},"evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"6dc8647adcf43a194f27defa4e025a3b0d687eb8320f29e99e40bba6bb108523","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.874410894122377,"output_peak":0.48532520380887817},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.6562641824926745,"spectral_convergence":0.17294202007673842,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.023112339088889,"spectral_convergence":0.16064467974326313,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.115244799291964,"spectral_convergence":0.16796814320780812,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.0022093050440864996,"delta_db":-0.1026806784412859,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.0014083872129707915,"delta_db":-0.22382589656024368,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.0002520516837820396,"delta_db":-0.14529666875746905,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00017600370198255137,"delta_db":-0.6511581717446052,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":3.507965431025385e-05,"delta_db":-0.6547218321330464,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":2.381263553655825e-05,"delta_db":-0.7040794602351764,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.41362711797863777,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1806007507781596,"pitch":null,"output_rms_dbfs":-23.867249340655334,"output_peak":0.48532520380887817,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.264873773624509},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9453533106204228,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06541352763725852,"rms_dbfs":-23.686648589877173,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.82005406670475,"mean":-7.879892985026041e-06,"zero_energy":false},"output":{"rms":0.06406746391631789,"rms_dbfs":-23.867249340655334,"peak":0.48532520380887817,"peak_dbfs":-6.2793430881433165,"crest_factor_db":17.587906252512017,"mean":0.00012446434504386066,"zero_energy":false},"rms_difference_db":-0.1806007507781615},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":3.2180639923731197e-06,"output_rms_dbfs":-54.92405324047367,"output_zero_energy":false,"output_peak":0.06666067802187231,"output_energy_fraction":0.00023912195314916646,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":122,"both_estimated_voiced_frames":115,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9274193548387096,"absolute_difference_cents_median":1.2487992260145722,"absolute_difference_cents_p95":9.335053437873748,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022610981418459577,"log_rms_mae_db":0.8702263527422257,"rms_envelope_nrmse":0.07260147711318783,"envelope_correlation":0.9955199255260341,"rms_gain_error_db":-0.18096105070075505,"reference_mean_power":0.00428070632223443,"output_mean_power":0.004106003630383445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.20587323032393e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001500467490243704,"output_power":0.0014491008825763454,"reference_active":true,"delta_db":-0.15127969298398591},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013253274491379155,"output_power":0.00012702121238995483,"reference_active":true,"delta_db":-0.18446939157502906},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.272532196136399e-05,"output_power":2.4779319894374293e-05,"reference_active":true,"delta_db":-1.2079454562120902}],"active_band_count":3,"active_band_mae_db":0.5145648469237017}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019856877416737668,"log_rms_mae_db":0.5623626673429595,"rms_envelope_nrmse":0.04038616179491792,"envelope_correlation":0.998806744980356,"rms_gain_error_db":-0.1811123818238411,"reference_mean_power":0.004296662851053013,"output_mean_power":0.004121165341261647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.009968768189794e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014002280466658247,"output_power":0.001352030821235036,"reference_active":true,"delta_db":-0.15212180347105758},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3360463722840526e-05,"output_power":5.252204189675998e-05,"reference_active":true,"delta_db":-0.06877993319856049},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4980780249701638e-06,"output_power":1.8934803141417797e-06,"reference_active":true,"delta_db":-1.203452047328172}],"active_band_count":3,"active_band_mae_db":0.47478459466593}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000212457290437723,"log_rms_mae_db":1.1364979874588688,"rms_envelope_nrmse":0.10466448111341324,"envelope_correlation":0.9929198760975577,"rms_gain_error_db":-0.2696386010136427,"reference_mean_power":0.0012391682631200216,"output_mean_power":0.0011645722086334718},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.921692896601052e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036532110371103373,"output_power":0.0003451071791863645,"reference_active":true,"delta_db":-0.2472076715675609},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3180096390035205e-05,"output_power":4.513612989622815e-05,"reference_active":true,"delta_db":0.1924071119329674},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1750484667690233e-05,"output_power":9.151000881991643e-06,"reference_active":true,"delta_db":-1.0858718294453773}],"active_band_count":3,"active_band_mae_db":0.5084955376486352}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018128574308840577,"log_rms_mae_db":0.810151122745321,"rms_envelope_nrmse":0.06414631056715321,"envelope_correlation":0.9976391438330593,"rms_gain_error_db":-0.27009856799167653,"reference_mean_power":0.0012444705639415084,"output_mean_power":0.0011694314563908631},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.573537659747316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003540640123239248,"output_power":0.00033567073154633234,"reference_active":true,"delta_db":-0.23168311510329576},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0107771514146375e-05,"output_power":2.2183345764442578e-05,"reference_active":true,"delta_db":0.42663106792189986},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973740035567709e-07,"output_power":5.809633313098619e-07,"reference_active":true,"delta_db":-1.375133505701955}],"active_band_count":3,"active_band_mae_db":0.6778158962423836}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013864336468723026,"log_rms_mae_db":1.2487428608468276,"rms_envelope_nrmse":0.17373110965552033,"envelope_correlation":0.9822215313824865,"rms_gain_error_db":-0.6669478396345109,"reference_mean_power":0.00021050674147203496,"output_mean_power":0.0001805390798622387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.902091296877062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.015318920526363e-05,"output_power":6.591724397058966e-05,"reference_active":true,"delta_db":-0.8492176610061315},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860315173964141e-05,"output_power":2.6783079095683346e-05,"reference_active":true,"delta_db":-1.5876226005639038},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0682238748218518e-05,"output_power":9.371922668625683e-06,"reference_active":true,"delta_db":-3.437688507503365}],"active_band_count":3,"active_band_mae_db":1.9581762563578}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":0.9597159942521065,"rms_envelope_nrmse":0.11519175515674267,"envelope_correlation":0.9940363413679703,"rms_gain_error_db":-0.6679834003872571,"reference_mean_power":0.00021105284570735848,"output_mean_power":0.00018096428542411957},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7773203833688672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041444617967505e-05,"output_power":7.127343395087001e-05,"reference_active":true,"delta_db":-1.0331014284658762},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4052463298203542e-05,"output_power":1.5636954867091176e-05,"reference_active":true,"delta_db":-1.8700737796443017},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.463215485080669e-06,"output_power":5.932484793042187e-07,"reference_active":true,"delta_db":-3.9207165494526484}],"active_band_count":3,"active_band_mae_db":2.2746305858542755}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":1.1840109222332629,"rms_envelope_nrmse":0.1399411714806055,"envelope_correlation":0.9882332203350627,"rms_gain_error_db":-0.3579941653520846,"reference_mean_power":2.4404568133397866e-05,"output_mean_power":2.247355153625984e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1035922908644866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.10428529724051e-05,"output_power":1.0200979128242603e-05,"reference_active":true,"delta_db":-0.3443943091202889},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8700267025557585e-06,"output_power":3.2509002856520576e-06,"reference_active":true,"delta_db":-0.7571031293586628},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.131168166122287e-06,"output_power":7.049561978877488e-07,"reference_active":true,"delta_db":-2.0536504134873406}],"active_band_count":3,"active_band_mae_db":1.0517159506554308}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":0.8468325134096875,"rms_envelope_nrmse":0.07547648788867625,"envelope_correlation":0.9972053713295433,"rms_gain_error_db":-0.3580555250403493,"reference_mean_power":2.4522755214627912e-05,"output_mean_power":2.2582067984736352e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9657184130662422e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.186833957836496e-05,"output_power":1.0754833839724641e-05,"reference_active":true,"delta_db":-0.4278625890613363},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.99155667709149e-06,"output_power":1.5882830030267325e-06,"reference_active":true,"delta_db":-0.9826478204082023},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.546174499350968e-08,"output_power":5.2807752195075293e-08,"reference_active":true,"delta_db":-2.0907407439329884}],"active_band_count":3,"active_band_mae_db":1.1670837178008424}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0002","method":"magphase","status":"measured","mos":{"status":"measured","value":3.8764901161193848,"predicted_mos":3.8764901161193848,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288864,"input_sample_rate":48000,"input_seconds":6.018,"input_float32_sha256":"a3777311e5d1e81c9b2e8597f9dee88732e8afe34586288100396b4bab2758d3","internal_samples":96288,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.611417787000391},"started_at":"2026-10-13T05:17:35.935128+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009727930000735796,"native_input_samples":288000,"native_input_sha256":"16bece45e51e2fcbff437ab8b7ed819d2e99730d389cc13008405015664fcce2","output_sha256":"ef4fbad6e66f78837afe948c5ef640159143e1c94b141bc7b752bf5bddffb0d5","output_samples":288864,"length_delta_samples":864,"analysis_s":3.4000951210000494,"synthesis_s":0.9793288700002449,"total_s":4.379423991000294,"rtf":0.7299039985000491,"output_peak":0.5941605679583554,"output_rms":0.06259312656524983,"finished_at":"2026-10-13T05:17:40.336451+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"ef4fbad6e66f78837afe948c5ef640159143e1c94b141bc7b752bf5bddffb0d5","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96288,"length_delta_samples":288,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.06785349606489,"output_peak":0.5859209008287649},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9970089730807578,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.299364722292509,"spectral_convergence":0.7516151507641913,"spectral_convergence_status":"ok","amplitude_floor":1.8465213976940698e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.2326915883985405,"spectral_convergence":0.6021687533046111,"spectral_convergence_status":"ok","amplitude_floor":1.4089022879434144e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4775286433653285,"spectral_convergence":0.6077624128564536,"spectral_convergence_status":"ok","amplitude_floor":1.0260349239920231e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002262162256014784,"output_power":0.0020256731602545214,"delta_db":-0.47954378120715174,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0014828754084214857,"output_power":0.001376923469242366,"delta_db":-0.32194860611372655,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00026062691889949753,"output_power":0.0002619825638829907,"delta_db":0.022531181529500975,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020447377727008734,"output_power":0.00019503381915800353,"delta_db":-0.20527694629323295,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":4.078753769921038e-05,"output_power":4.132350931474439e-05,"delta_db":0.05669707779859715,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.800369964531977e-05,"output_power":2.4738242188908566e-05,"delta_db":-0.5384657398848496,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.27074388880450984,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3742082487030517,"pitch":null,"output_rms_dbfs":-24.06085683858023,"output_peak":0.5859209008287649,"reference_rms_dbfs":-23.686648589877173,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.336528318018793},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6840843542418401,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06541352763725852,"rms_dbfs":-23.686648589877173,"peak":0.571044921875,"peak_dbfs":-4.866594523172426,"crest_factor_db":18.82005406670475,"mean":-7.879892985026041e-06,"zero_energy":false},"output":{"rms":0.06265520540259631,"rms_dbfs":-24.06085683858023,"peak":0.5859209008287649,"peak_dbfs":-4.643220193367343,"crest_factor_db":19.417636645212884,"mean":-0.000946243979969148,"zero_energy":false},"rms_difference_db":-0.3742082487030558},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.106411339947954e-06,"threshold_dbfs":-52.91884204127492,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":183,"quiet_samples":29280,"reference_mean_power":3.6498337673644226e-07,"output_mean_power":1.6806689835370482e-06,"output_rms_dbfs":-57.74517814686259,"output_zero_energy":false,"output_peak":0.03694375014004144,"output_energy_fraction":0.00013057730728845886,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":124,"output_estimated_voiced_frames":124,"both_estimated_voiced_frames":94,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.7580645161290323,"absolute_difference_cents_median":10.22221306201335,"absolute_difference_cents_p95":72.71363801201386,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00022610981418459577,"log_rms_mae_db":3.3521998099440142,"rms_envelope_nrmse":0.405979996023813,"envelope_correlation":0.847441567676834,"rms_gain_error_db":-0.3735425689285135,"reference_mean_power":0.00428070632223443,"output_mean_power":0.003927906824875182},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.20587323032393e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001500467490243704,"output_power":0.0013895640551811806,"reference_active":true,"delta_db":-0.3334801886964366},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013253274491379155,"output_power":0.00013290437410689584,"reference_active":true,"delta_db":0.012160817140453912},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.272532196136399e-05,"output_power":3.433261253188928e-05,"reference_active":true,"delta_db":0.20822924256340883}],"active_band_count":3,"active_band_mae_db":0.1846234161334331}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00019856877416737668,"log_rms_mae_db":2.65235085979164,"rms_envelope_nrmse":0.314859564493308,"envelope_correlation":0.9021979802472068,"rms_gain_error_db":-0.3786447336411687,"reference_mean_power":0.004296662851053013,"output_mean_power":0.003937919224977871},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.009968768189794e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014002280466658247,"output_power":0.001302388772243318,"reference_active":true,"delta_db":-0.3145812879773787},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.3360463722840526e-05,"output_power":5.491735190351706e-05,"reference_active":true,"delta_db":0.12489992423752734},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4980780249701638e-06,"output_power":2.7342272504782682e-06,"reference_active":true,"delta_db":0.39228608348219385}],"active_band_count":3,"active_band_mae_db":0.2772557652323666}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.000212457290437723,"log_rms_mae_db":3.9355613824176614,"rms_envelope_nrmse":0.6461067848987693,"envelope_correlation":0.7029635330142733,"rms_gain_error_db":-0.3810549868731255,"reference_mean_power":0.0012391682631200216,"output_mean_power":0.001135075587659738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.921692896601052e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00036532110371103373,"output_power":0.00033275243999383153,"reference_active":true,"delta_db":-0.40553512336149006},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3180096390035205e-05,"output_power":4.0020998378123324e-05,"reference_active":true,"delta_db":-0.3299568896524431},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1750484667690233e-05,"output_power":1.0160492551858955e-05,"reference_active":true,"delta_db":-0.6314101833065205}],"active_band_count":3,"active_band_mae_db":0.4556340654401512}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018128574308840577,"log_rms_mae_db":3.471100973729483,"rms_envelope_nrmse":0.5104187741709292,"envelope_correlation":0.8052522648194267,"rms_gain_error_db":-0.3849521155682506,"reference_mean_power":0.0012444705639415084,"output_mean_power":0.0011389100281105479},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.573537659747316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003540640123239248,"output_power":0.00032421517510491737,"reference_active":true,"delta_db":-0.3824844813447733},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0107771514146375e-05,"output_power":1.8511740150681033e-05,"reference_active":true,"delta_db":-0.35916696210677457},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.973740035567709e-07,"output_power":6.911722786449577e-07,"reference_active":true,"delta_db":-0.6207576118827763}],"active_band_count":3,"active_band_mae_db":0.45413635177810807}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013864336468723026,"log_rms_mae_db":4.2163455956770255,"rms_envelope_nrmse":0.6634068092109093,"envelope_correlation":0.7027064145622055,"rms_gain_error_db":-0.18142399779728957,"reference_mean_power":0.00021050674147203496,"output_mean_power":0.0002018940912815263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.902091296877062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.015318920526363e-05,"output_power":7.571618341459969e-05,"reference_active":true,"delta_db":-0.24732092530578198},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.860315173964141e-05,"output_power":3.650350470311802e-05,"reference_active":true,"delta_db":-0.24288200836229834},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0682238748218518e-05,"output_power":2.0135553874258652e-05,"reference_active":true,"delta_db":-0.11633966791195116}],"active_band_count":3,"active_band_mae_db":0.20218086719334383}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.950462169412165e-05,"log_rms_mae_db":3.5181995349510684,"rms_envelope_nrmse":0.4569719634273234,"envelope_correlation":0.8491741962312488,"rms_gain_error_db":-0.23495320111435786,"reference_mean_power":0.00021105284570735848,"output_mean_power":0.00019993825259690217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7773203833688672e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041444617967505e-05,"output_power":8.587098839213593e-05,"reference_active":true,"delta_db":-0.22391364693142268},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4052463298203542e-05,"output_power":2.308470931721237e-05,"reference_active":true,"delta_db":-0.17835150345051637},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.463215485080669e-06,"output_power":1.5390065803569553e-06,"reference_active":true,"delta_db":0.21932188162897548}],"active_band_count":3,"active_band_mae_db":0.2071956773369715}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.6125957709065377e-05,"log_rms_mae_db":4.558501994761003,"rms_envelope_nrmse":0.593713049883471,"envelope_correlation":0.7751686393808032,"rms_gain_error_db":-0.09685516906202227,"reference_mean_power":2.4404568133397866e-05,"output_mean_power":2.3866328268686724e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1035922908644866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.10428529724051e-05,"output_power":1.1501026642976756e-05,"reference_active":true,"delta_db":0.17655319637570166},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8700267025557585e-06,"output_power":4.450728977061665e-06,"reference_active":true,"delta_db":0.6071718754466403},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.131168166122287e-06,"output_power":1.2110740371993372e-06,"reference_active":true,"delta_db":0.29643519397074625}],"active_band_count":3,"active_band_mae_db":0.3600534219310294}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.642002234154929e-05,"log_rms_mae_db":3.876649819289768,"rms_envelope_nrmse":0.4416613445609408,"envelope_correlation":0.8687356840512843,"rms_gain_error_db":-0.10969171524185548,"reference_mean_power":2.4522755214627912e-05,"output_mean_power":2.3911129498343854e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9657184130662422e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.186833957836496e-05,"output_power":1.2644415853849128e-05,"reference_active":true,"delta_db":0.27508806718325457},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.99155667709149e-06,"output_power":2.3150428838146527e-06,"reference_active":true,"delta_db":0.653663699193353},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.546174499350968e-08,"output_power":9.60916017255897e-08,"reference_active":true,"delta_db":0.5091367640675013}],"active_band_count":3,"active_band_mae_db":0.479296176814703}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"hnm","status":"measured","mos":{"status":"measured","value":3.0488877296447754,"predicted_mos":3.0488877296447754,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"585c84b25d7af6c2d2cdb2be53d7a0aa52805cbc362b540c6e41bb87c895828c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.668575695999607},"started_at":"2026-10-13T04:39:59.304187+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5241999790305272e-05,"native_input_samples":96000,"native_input_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","output_sha256":"460fb56f5add0cd95da947a6932b59c6e8566aa3fb39e04aa60680628eb5283e","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1187966439993033,"synthesis_s":0.49139414000001125,"total_s":1.6101907839993146,"rtf":0.26836513066655243,"output_peak":0.5247168935688253,"output_rms":0.06050027713926589,"finished_at":"2026-10-13T04:40:00.921146+00:00","evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"460fb56f5add0cd95da947a6932b59c6e8566aa3fb39e04aa60680628eb5283e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.515855329325584,"spectral_convergence":0.3147950294778251,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.237179573320257,"spectral_convergence":0.31685717698686156,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.561292555438582,"spectral_convergence":0.3445481797652698,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.0017091878174282649,"delta_db":-0.12440583419920612,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.001211810357754823,"delta_db":0.016671512128753452,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.0004189431530563499,"delta_db":0.28033455983927485,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.00022730533412252969,"delta_db":-1.4949903625560903,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":5.244481489890768e-05,"delta_db":-0.13615399788934227,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":4.059205666710301e-05,"delta_db":-0.25609046754645703,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38477445569318736,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13592992441920956,"pitch":null,"output_rms_dbfs":-24.364852718594182,"output_peak":0.5247168935688253,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.438109152694807},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.515855329325584,"spectral_convergence":0.3147950294778251,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.237179573320257,"spectral_convergence":0.31685717698686156,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.561292555438582,"spectral_convergence":0.3445481797652698,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.0017091878174282649,"delta_db":-0.12440583419920612,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.001211810357754823,"delta_db":0.016671512128753452,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.0004189431530563499,"delta_db":0.28033455983927485,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.00022730533412252969,"delta_db":-1.4949903625560903,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":5.244481489890768e-05,"delta_db":-0.13615399788934227,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":4.059205666710301e-05,"delta_db":-0.25609046754645703,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38477445569318736,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13592992441920956,"pitch":null,"output_rms_dbfs":-24.364852718594182,"output_peak":0.5247168935688253,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.438109152694807},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8967152366755053,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06145452415925838,"rms_dbfs":-24.228922794174977,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.813876120914777,"mean":2.243359883626302e-05,"zero_energy":false},"output":{"rms":0.06050027713926589,"rms_dbfs":-24.364852718594186,"peak":0.5247168935688253,"peak_dbfs":-5.601499064208818,"crest_factor_db":18.763353654385366,"mean":-0.00018907581181886542,"zero_energy":false},"rms_difference_db":-0.1359299244192087},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":3.5456966270620464e-07,"output_rms_dbfs":-64.50298425765335,"output_zero_energy":false,"output_peak":0.013991396660313432,"output_energy_fraction":2.841503946249569e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":217,"both_estimated_voiced_frames":201,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9305555555555556,"absolute_difference_cents_median":1.9276064710603784,"absolute_difference_cents_p95":16.07507704229682,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002198976712894503,"log_rms_mae_db":0.6981364087499674,"rms_envelope_nrmse":0.12164750217850866,"envelope_correlation":0.9855781465128293,"rms_gain_error_db":-0.1350858622297878,"reference_mean_power":0.0037766562988771884,"output_mean_power":0.003660992816184911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.069519300574134e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006799849644412746,"output_power":0.0005841501577164414,"reference_active":true,"delta_db":-0.6597481157850313},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022275189153607582,"output_power":0.00020774033134981186,"reference_active":true,"delta_db":-0.30300580617621736},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.238302101062001e-05,"output_power":4.3429152707030615e-05,"reference_active":true,"delta_db":0.10589446992585194}],"active_band_count":3,"active_band_mae_db":0.3562161306290335}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018113949472991762,"log_rms_mae_db":0.41869768000113816,"rms_envelope_nrmse":0.07090713454593874,"envelope_correlation":0.9949148960512011,"rms_gain_error_db":-0.13632972981138342,"reference_mean_power":0.003780325824366865,"output_mean_power":0.0036635005416916157},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9093794448028363e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006511405512347742,"output_power":0.0005577029336198827,"reference_active":true,"delta_db":-0.6727181388340896},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011667254760432671,"output_power":0.00010426213756347923,"reference_active":true,"delta_db":-0.48842056502056147},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.399065253982207e-06,"output_power":2.9307998211005306e-06,"reference_active":true,"delta_db":-0.6437334572884028}],"active_band_count":3,"active_band_mae_db":0.6016240537143512}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013526658282762358,"log_rms_mae_db":1.2758890164131702,"rms_envelope_nrmse":0.1640339139835274,"envelope_correlation":0.9799772627679147,"rms_gain_error_db":-0.15335542409446226,"reference_mean_power":0.001047203990993015,"output_mean_power":0.0010108710201284232},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.584657585149468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003504540347149448,"output_power":0.00032273914135501667,"reference_active":true,"delta_db":-0.3577942507351657},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.5576057457372e-05,"output_power":9.472838439687864e-05,"reference_active":true,"delta_db":-0.038689813416827805},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9623586773434205e-05,"output_power":2.5615323048355083e-05,"reference_active":true,"delta_db":1.157214472891038}],"active_band_count":3,"active_band_mae_db":0.5178995123476772}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012031399971827572,"log_rms_mae_db":0.9515052831304955,"rms_envelope_nrmse":0.07662690486139732,"envelope_correlation":0.9955520460993749,"rms_gain_error_db":-0.15365954250143174,"reference_mean_power":0.0010510881038568295,"output_mean_power":0.0010145493256818031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.061243701565634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003432746241782782,"output_power":0.00032224010141737467,"reference_active":true,"delta_db":-0.2746211479040155},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.625199841508276e-05,"output_power":4.548200953333935e-05,"reference_active":true,"delta_db":-0.07290857199899531},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5417688576161993e-06,"output_power":1.4583776352181065e-06,"reference_active":true,"delta_db":-0.24149273429520168}],"active_band_count":3,"active_band_mae_db":0.19634081806607084}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013601220330575493,"log_rms_mae_db":1.5890235890458806,"rms_envelope_nrmse":0.2632282740762697,"envelope_correlation":0.9586441813857827,"rms_gain_error_db":-0.838745490549244,"reference_mean_power":0.0002492356992147181,"output_mean_power":0.0002054639814024751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.707381052427382e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017017591719688496,"output_power":0.0001242580637720663,"reference_active":true,"delta_db":-1.3657351798160997},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.646281785953888e-05,"output_power":1.0360462877426664e-05,"reference_active":true,"delta_db":-4.072569286658708},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0978410742643585e-06,"output_power":5.046202865345512e-06,"reference_active":true,"delta_db":2.1190557208024776}],"active_band_count":3,"active_band_mae_db":2.519120062425762}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011511226838452263,"log_rms_mae_db":1.3054423860134732,"rms_envelope_nrmse":0.1839656530951572,"envelope_correlation":0.9824648586184773,"rms_gain_error_db":-0.8427343024208861,"reference_mean_power":0.0002496383892619269,"output_mean_power":0.00020560702136186007},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6223780118670245e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016948627965662866,"output_power":0.00012169274167570584,"reference_active":true,"delta_db":-1.4386987107603368},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1767373976886582e-05,"output_power":3.7798804788989575e-06,"reference_active":true,"delta_db":-4.932014883955055},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8952968721807653e-07,"output_power":4.218730256440298e-07,"reference_active":true,"delta_db":3.4750451180357373}],"active_band_count":3,"active_band_mae_db":3.2819195709170432}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":1.6117248932096242,"rms_envelope_nrmse":0.1606627094805184,"envelope_correlation":0.9843996607216396,"rms_gain_error_db":-0.1447046786660928,"reference_mean_power":4.1047668613925756e-05,"output_mean_power":3.9702515836155475e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.890895346126464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.549789437651058e-06,"output_power":9.642738282638517e-06,"reference_active":true,"delta_db":0.042065835897711235},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.275656140671785e-06,"output_power":3.080199934740874e-06,"reference_active":true,"delta_db":-0.2671939844851899},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.939546621487426e-07,"output_power":5.414726700320174e-07,"reference_active":true,"delta_db":-0.4017675440957671}],"active_band_count":3,"active_band_mae_db":0.2370091214928894}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":1.396331206566422,"rms_envelope_nrmse":0.08974464275148406,"envelope_correlation":0.995184382753728,"rms_gain_error_db":-0.14550487597239145,"reference_mean_power":4.122071638798199e-05,"output_mean_power":3.986254730720957e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7789380940114685e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.936665760227186e-06,"output_power":9.985574838522203e-06,"reference_active":true,"delta_db":0.021323892146127153},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7465237791634247e-06,"output_power":1.6541650445370568e-06,"reference_active":true,"delta_db":-0.23595663825496466},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8043700755065895e-08,"output_power":3.567040049197134e-08,"reference_active":true,"delta_db":-1.2932846782238625}],"active_band_count":3,"active_band_mae_db":0.5168550695416515}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.4564480781555176,"predicted_mos":3.4564480781555176,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2e45a8422295ec6ddae3ad5af6ccbe54585a9348e14edf9b5b511e21306fe379","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.545665804998862},"started_at":"2026-10-13T04:43:11.118560+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.099699941027211e-05,"native_input_samples":96000,"native_input_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","output_sha256":"1a67ce0f10f1e629c2b807f7215c38385df01f473d71974ff93cee1526e80a7b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5692023980000158,"synthesis_s":0.10147484699973575,"total_s":1.6706772449997516,"rtf":0.2784462074999586,"output_peak":0.5841181660826203,"output_rms":0.06710222055316768,"finished_at":"2026-10-13T04:43:12.796108+00:00","evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"1a67ce0f10f1e629c2b807f7215c38385df01f473d71974ff93cee1526e80a7b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.046519029608891,"spectral_convergence":0.6238685387023335,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.96570995404591,"spectral_convergence":0.2623431599012371,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.242780645348987,"spectral_convergence":0.2966919714311337,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.002051048131839002,"delta_db":0.6674547999686725,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.001493841199223243,"delta_db":0.9253692381710443,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.00043553205661689675,"delta_db":0.44898485052744125,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.00040611266365091073,"delta_db":1.0253786854471045,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":6.807332633368948e-05,"delta_db":0.9965901575951913,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":4.810062550221639e-05,"delta_db":0.48100621077031325,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7574639904132945,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7636606366101807,"pitch":null,"output_rms_dbfs":-23.465262157564798,"output_peak":0.5841181660826203,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.418336543001263},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.046519029608891,"spectral_convergence":0.6238685387023335,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.96570995404591,"spectral_convergence":0.2623431599012371,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.242780645348987,"spectral_convergence":0.2966919714311337,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.002051048131839002,"delta_db":0.6674547999686725,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.001493841199223243,"delta_db":0.9253692381710443,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.00043553205661689675,"delta_db":0.44898485052744125,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.00040611266365091073,"delta_db":1.0253786854471045,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":6.807332633368948e-05,"delta_db":0.9965901575951913,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":4.810062550221639e-05,"delta_db":0.48100621077031325,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7574639904132945,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7636606366101807,"pitch":null,"output_rms_dbfs":-23.465262157564798,"output_peak":0.5841181660826203,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.418336543001263},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9326563023506349,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06145452415925838,"rms_dbfs":-24.228922794174977,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.813876120914777,"mean":2.243359883626302e-05,"zero_energy":false},"output":{"rms":0.06710222055316768,"rms_dbfs":-23.465262157564798,"peak":0.5841181660826203,"peak_dbfs":-4.66998573972391,"crest_factor_db":18.795276417840885,"mean":1.2002473965396071e-05,"zero_energy":false},"rms_difference_db":0.7636606366101795},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":5.011993648091043e-07,"output_rms_dbfs":-62.99989488077081,"output_zero_energy":false,"output_peak":0.012912993285447616,"output_energy_fraction":3.265112466556388e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":231,"both_estimated_voiced_frames":200,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.9259259259259259,"absolute_difference_cents_median":8.919308147999306,"absolute_difference_cents_p95":31.47311351953795,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002198976712894503,"log_rms_mae_db":1.490180855962405,"rms_envelope_nrmse":0.16707480131208471,"envelope_correlation":0.9827015504783597,"rms_gain_error_db":0.7672157352494606,"reference_mean_power":0.0037766562988771884,"output_mean_power":0.0045063927179202455},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.069519300574134e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006799849644412746,"output_power":0.0007693275559764297,"reference_active":true,"delta_db":0.5361197852564131},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022275189153607582,"output_power":0.00027284516656925624,"reference_active":true,"delta_db":0.880948640835866},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.238302101062001e-05,"output_power":3.875465134720002e-05,"reference_active":true,"delta_db":-0.38868075314400113}],"active_band_count":3,"active_band_mae_db":0.6019163930787601}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018113949472991762,"log_rms_mae_db":1.1774236404509009,"rms_envelope_nrmse":0.12217508607437302,"envelope_correlation":0.9940118846078757,"rms_gain_error_db":0.7735115333852778,"reference_mean_power":0.003780325824366865,"output_mean_power":0.004517315111354535},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9093794448028363e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006511405512347742,"output_power":0.0007340087504104657,"reference_active":true,"delta_db":0.5202649449457506},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011667254760432671,"output_power":0.00014706299695510633,"reference_active":true,"delta_db":1.005347310787614},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.399065253982207e-06,"output_power":2.4120849752924627e-06,"reference_active":true,"delta_db":-1.4896689859688872}],"active_band_count":3,"active_band_mae_db":1.005093747234084}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013526658282762358,"log_rms_mae_db":1.8159737602596713,"rms_envelope_nrmse":0.22541587143425532,"envelope_correlation":0.9779168494511362,"rms_gain_error_db":1.0667673459617761,"reference_mean_power":0.001047203990993015,"output_mean_power":0.0013387763238130147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.584657585149468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003504540347149448,"output_power":0.0004167520023714169,"reference_active":true,"delta_db":0.7524663085951288},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.5576057457372e-05,"output_power":0.00012697382299454258,"reference_active":true,"delta_db":1.2336508398330588},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9623586773434205e-05,"output_power":2.042904579233129e-05,"reference_active":true,"delta_db":0.17469691839080662}],"active_band_count":3,"active_band_mae_db":0.7202713556063314}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012031399971827572,"log_rms_mae_db":1.55641756746777,"rms_envelope_nrmse":0.18490911006220115,"envelope_correlation":0.9881071015974153,"rms_gain_error_db":1.0676131658762187,"reference_mean_power":0.0010510881038568295,"output_mean_power":0.001344003616657562},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.061243701565634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003432746241782782,"output_power":0.00040602718448383147,"reference_active":true,"delta_db":0.7291341124935969},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.625199841508276e-05,"output_power":6.288889656374421e-05,"reference_active":true,"delta_db":1.3344347266056515},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5417688576161993e-06,"output_power":1.309397790315097e-06,"reference_active":true,"delta_db":-0.7094766536520057}],"active_band_count":3,"active_band_mae_db":0.9243484975837514}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013601220330575493,"log_rms_mae_db":1.502999591608869,"rms_envelope_nrmse":0.22364623654910368,"envelope_correlation":0.9817995690441084,"rms_gain_error_db":1.0552075837552566,"reference_mean_power":0.0002492356992147181,"output_mean_power":0.0003177832533547665},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.707381052427382e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017017591719688496,"output_power":0.00021767862262211918,"reference_active":true,"delta_db":1.0691768075008903},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.646281785953888e-05,"output_power":3.558234862785651e-05,"reference_active":true,"delta_db":1.2859852287358622},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0978410742643585e-06,"output_power":8.102529762151112e-06,"reference_active":true,"delta_db":4.175615011762749}],"active_band_count":3,"active_band_mae_db":2.1769256826665004}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011511226838452263,"log_rms_mae_db":1.219570738560091,"rms_envelope_nrmse":0.16385403411814542,"envelope_correlation":0.9942922114734017,"rms_gain_error_db":1.0583687851921844,"reference_mean_power":0.0002496383892619269,"output_mean_power":0.0003185284661128075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6223780118670245e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016948627965662866,"output_power":0.0002196340503041887,"reference_active":true,"delta_db":1.1256512381168913},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1767373976886582e-05,"output_power":1.6357160364092844e-05,"reference_active":true,"delta_db":1.4302835557290936},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8952968721807653e-07,"output_power":4.915939488862978e-07,"reference_active":true,"delta_db":4.13929282554335}],"active_band_count":3,"active_band_mae_db":2.231742539796445}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":1.3253910574969918,"rms_envelope_nrmse":0.19307246287654123,"envelope_correlation":0.9797145446976626,"rms_gain_error_db":0.3879232152034393,"reference_mean_power":4.1047668613925756e-05,"output_mean_power":4.488289038806271e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.890895346126464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.549789437651058e-06,"output_power":1.0580203551073868e-05,"reference_active":true,"delta_db":0.44500227137192694},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.275656140671785e-06,"output_power":3.126167615668156e-06,"reference_active":true,"delta_db":-0.20286045906307168},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.939546621487426e-07,"output_power":5.228394392614921e-07,"reference_active":true,"delta_db":-0.5538495539549564}],"active_band_count":3,"active_band_mae_db":0.4005707614633183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":0.8769607010753847,"rms_envelope_nrmse":0.11220671897462149,"envelope_correlation":0.9937458622409103,"rms_gain_error_db":0.3886439529024127,"reference_mean_power":4.122071638798199e-05,"output_mean_power":4.507958720452641e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7789380940114685e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.936665760227186e-06,"output_power":1.0888215331157935e-05,"reference_active":true,"delta_db":0.3971601942379343},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7465237791634247e-06,"output_power":1.629345025374423e-06,"reference_active":true,"delta_db":-0.3016144400728945},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8043700755065895e-08,"output_power":3.502584130557055e-08,"reference_active":true,"delta_db":-1.3724787761159858}],"active_band_count":3,"active_band_mae_db":0.6904178034756049}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"facodec","status":"measured","mos":{"status":"measured","value":4.218729496002197,"predicted_mos":4.218729496002197,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d0b7894a9adbb2976b3c08dc4e9226737751ca59e960cd68f7fef97f19ae477b","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5553765789991303},"started_at":"2026-10-13T06:58:22.424696+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3436999981640838e-05,"native_input_samples":96000,"native_input_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","output_sha256":"09840ddbef130f06b5997ceb8284d7e0ffa5968a96027314c1d9c22e8e9ce08d","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6518650769994565,"synthesis_s":7.290661491999344,"total_s":8.942526568998801,"rtf":1.4904210948331336,"output_peak":0.48633068799972534,"output_rms":0.05835530769569827,"finished_at":"2026-10-13T06:58:31.380085+00:00","evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"09840ddbef130f06b5997ceb8284d7e0ffa5968a96027314c1d9c22e8e9ce08d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.32361506498316,"spectral_convergence":0.2441858105492049,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.5085286482773705,"spectral_convergence":0.24165634338866857,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.5802209999893675,"spectral_convergence":0.2550290311024412,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.0014775308491991514,"delta_db":-0.7569381482370248,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.0012431357905780325,"delta_db":0.1275106119066936,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.00038684688275358804,"delta_db":-0.06582539370282457,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.00021563537474670775,"delta_db":-1.72388655697285,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":5.350249987473571e-05,"delta_db":-0.04943882055719461,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":2.8690539107405268e-05,"delta_db":-1.7631139401386717,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7477855785858766,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4494699388587345,"pitch":null,"output_rms_dbfs":-24.67839273303371,"output_peak":0.48633068799972534,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.470788237749967},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.32361506498316,"spectral_convergence":0.2441858105492049,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.5085286482773705,"spectral_convergence":0.24165634338866857,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.5802209999893675,"spectral_convergence":0.2550290311024412,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.0014775308491991514,"delta_db":-0.7569381482370248,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.0012431357905780325,"delta_db":0.1275106119066936,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.00038684688275358804,"delta_db":-0.06582539370282457,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.00021563537474670775,"delta_db":-1.72388655697285,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":5.350249987473571e-05,"delta_db":-0.04943882055719461,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":2.8690539107405268e-05,"delta_db":-1.7631139401386717,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7477855785858766,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4494699388587345,"pitch":null,"output_rms_dbfs":-24.67839273303371,"output_peak":0.48633068799972534,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.470788237749967},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9001983385583271,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06145452415925838,"rms_dbfs":-24.228922794174977,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.813876120914777,"mean":2.243359883626302e-05,"zero_energy":false},"output":{"rms":0.05835530769569826,"rms_dbfs":-24.67839273303371,"peak":0.48633068799972534,"peak_dbfs":-6.261366502182061,"crest_factor_db":18.41702623085165,"mean":4.171458290341423e-05,"zero_energy":false},"rms_difference_db":-0.44946993885873354},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":2.828245740294486e-07,"output_rms_dbfs":-65.48482858306099,"output_zero_energy":false,"output_peak":0.010686328634619713,"output_energy_fraction":2.4362274509138543e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":204,"both_estimated_voiced_frames":197,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9120370370370371,"absolute_difference_cents_median":3.6454751193027723,"absolute_difference_cents_p95":16.825644665681214,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002198976712894503,"log_rms_mae_db":1.1096131913625715,"rms_envelope_nrmse":0.12192325565902988,"envelope_correlation":0.987345183936198,"rms_gain_error_db":-0.44910389183044913,"reference_mean_power":0.0037766562988771884,"output_mean_power":0.003405626948484938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.069519300574134e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006799849644412746,"output_power":0.0005366231395742909,"reference_active":true,"delta_db":-1.0282991403555835},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022275189153607582,"output_power":0.0002127042010567669,"reference_active":true,"delta_db":-0.20045333003746643},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.238302101062001e-05,"output_power":3.735447193204996e-05,"reference_active":true,"delta_db":-0.5484930815541965}],"active_band_count":3,"active_band_mae_db":0.5924151839824154}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018113949472991762,"log_rms_mae_db":1.0078153645878036,"rms_envelope_nrmse":0.11163102065877027,"envelope_correlation":0.989051820893868,"rms_gain_error_db":-0.4492177286253287,"reference_mean_power":0.003780325824366865,"output_mean_power":0.0034088466159706713},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9093794448028363e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006511405512347742,"output_power":0.0005084142042497086,"reference_active":true,"delta_db":-1.0745706733067928},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011667254760432671,"output_power":0.00011146696339668536,"reference_active":true,"delta_db":-0.19822510851117825},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.399065253982207e-06,"output_power":3.1622433523631845e-06,"reference_active":true,"delta_db":-0.3136421380735738}],"active_band_count":3,"active_band_mae_db":0.5288126399638483}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013526658282762358,"log_rms_mae_db":1.3519658516208466,"rms_envelope_nrmse":0.13126294647764644,"envelope_correlation":0.987222035673664,"rms_gain_error_db":-0.04148366079647113,"reference_mean_power":0.001047203990993015,"output_mean_power":0.0010372487559115393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.584657585149468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003504540347149448,"output_power":0.0003230944023236543,"reference_active":true,"delta_db":-0.35301630460997435},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.5576057457372e-05,"output_power":9.39190555079076e-05,"reference_active":true,"delta_db":-0.07595395303739629},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9623586773434205e-05,"output_power":2.027411421658661e-05,"reference_active":true,"delta_db":0.1416349873937494}],"active_band_count":3,"active_band_mae_db":0.19020174834704004}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012031399971827572,"log_rms_mae_db":1.1161386582356836,"rms_envelope_nrmse":0.11018339946205666,"envelope_correlation":0.9904182118413842,"rms_gain_error_db":-0.041636034383570705,"reference_mean_power":0.0010510881038568295,"output_mean_power":0.0010410594179538345},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.061243701565634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003432746241782782,"output_power":0.00031329648991004695,"reference_active":true,"delta_db":-0.39686171157844},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.625199841508276e-05,"output_power":4.620324669307315e-05,"reference_active":true,"delta_db":-0.0045800767256864525},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5417688576161993e-06,"output_power":1.4164398861887964e-06,"reference_active":true,"delta_db":-0.3682112129626496}],"active_band_count":3,"active_band_mae_db":0.2565510004222587}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013601220330575493,"log_rms_mae_db":1.3412732545547583,"rms_envelope_nrmse":0.2185992587084665,"envelope_correlation":0.9786504489039153,"rms_gain_error_db":-1.2045049732829372,"reference_mean_power":0.0002492356992147181,"output_mean_power":0.0001888685956867742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.707381052427382e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017017591719688496,"output_power":0.00011421466484392795,"reference_active":true,"delta_db":-1.7317623031091565},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.646281785953888e-05,"output_power":1.712917452894581e-05,"reference_active":true,"delta_db":-1.8889965320032571},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0978410742643585e-06,"output_power":3.862250481968591e-06,"reference_active":true,"delta_db":0.9578130226526355}],"active_band_count":3,"active_band_mae_db":1.526190619255016}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011511226838452263,"log_rms_mae_db":1.212878494642642,"rms_envelope_nrmse":0.19041416884233792,"envelope_correlation":0.9867478735716807,"rms_gain_error_db":-1.2025297551666223,"reference_mean_power":0.0002496383892619269,"output_mean_power":0.0001892598084516733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6223780118670245e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016948627965662866,"output_power":0.00011306410953739062,"reference_active":true,"delta_db":-1.758097800264633},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1767373976886582e-05,"output_power":7.746682512888233e-06,"reference_active":true,"delta_db":-1.8156379854087545},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8952968721807653e-07,"output_power":2.9046810362182217e-07,"reference_active":true,"delta_db":1.8542120346612827}],"active_band_count":3,"active_band_mae_db":1.809315940111557}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":1.5866397791780125,"rms_envelope_nrmse":0.21074571063445785,"envelope_correlation":0.9768870780889166,"rms_gain_error_db":-0.9107757546295563,"reference_mean_power":4.1047668613925756e-05,"output_mean_power":3.3282115242558576e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.890895346126464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.549789437651058e-06,"output_power":7.821944232965962e-06,"reference_active":true,"delta_db":-0.8667908067601986},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.275656140671785e-06,"output_power":2.282942673554054e-06,"reference_active":true,"delta_db":-1.5680329961375048},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.939546621487426e-07,"output_power":5.273318240597596e-07,"reference_active":true,"delta_db":-0.516693141520567}],"active_band_count":3,"active_band_mae_db":0.9838389814727568}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":1.354378808239589,"rms_envelope_nrmse":0.18068193655550746,"envelope_correlation":0.9842798948894456,"rms_gain_error_db":-0.9101295451837741,"reference_mean_power":4.122071638798199e-05,"output_mean_power":3.342739865072642e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7789380940114685e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.936665760227186e-06,"output_power":7.895912894224274e-06,"reference_active":true,"delta_db":-0.9983833303816191},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7465237791634247e-06,"output_power":1.1929296700279928e-06,"reference_active":true,"delta_db":-1.655596626920741},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8043700755065895e-08,"output_power":4.287652753283659e-08,"reference_active":true,"delta_db":-0.49416847620367793}],"active_band_count":3,"active_band_mae_db":1.0493828111686794}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.867391347885132,"predicted_mos":3.867391347885132,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"8527aa090cda9055dfbec96c934ce50a34e17ea611b1c62cb5082a2e9c939f40","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2303400400014652},"started_at":"2026-10-13T07:27:47.313559+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008713361999980407,"native_input_samples":264600,"native_input_sha256":"617c63bcea429b3c5bc0c994e7d696b5d9ed8433221c300af87a889765f9eca5","output_sha256":"5d3571fec97b0a7505030ecbe39b7ec6f4bab31d69ff7a75c430f7bba4610f92","output_samples":264704,"length_delta_samples":104,"analysis_s":0.16057684000043082,"synthesis_s":4.298562100000709,"total_s":4.459138940001139,"rtf":0.7431898233335232,"output_peak":0.4848850965499878,"output_rms":0.05570708317366532,"finished_at":"2026-10-13T07:27:51.807146+00:00","evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"5d3571fec97b0a7505030ecbe39b7ec6f4bab31d69ff7a75c430f7bba4610f92","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.079163095073923,"output_peak":0.4823255193684851},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.830415833797758,"spectral_convergence":0.4697685392734703,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.453692279641684,"spectral_convergence":0.3807788560167969,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.190688979590725,"spectral_convergence":0.4210860471036822,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.0011725803685168157,"delta_db":-1.7608775315833405,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.0012781077446359672,"delta_db":0.24799957416490304,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.000411499640097228,"delta_db":0.20247818714390114,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.0001742470589975699,"delta_db":-2.649432069312214,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":4.629992023294242e-05,"delta_db":-0.6773771391918617,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":2.311002800457576e-05,"delta_db":-2.702496311596075,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3734434688320494,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8492797599923058,"pitch":null,"output_rms_dbfs":-25.078202554167284,"output_peak":0.4823255193684851,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.491599031010056},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7724123238679106,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06145452415925838,"rms_dbfs":-24.228922794174977,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.813876120914777,"mean":2.243359883626302e-05,"zero_energy":false},"output":{"rms":0.055730106410136156,"rms_dbfs":-25.07820255416728,"peak":0.4823255193684851,"peak_dbfs":-6.333195187533056,"crest_factor_db":18.745007366634226,"mean":-6.613544886956576e-05,"zero_energy":false},"rms_difference_db":-0.849279759992303},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":1.318267515832026e-07,"output_rms_dbfs":-68.79996449483578,"output_zero_energy":false,"output_peak":0.005573014598082394,"output_energy_fraction":1.2450455011913212e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":200,"both_estimated_voiced_frames":186,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.8611111111111112,"absolute_difference_cents_median":10.125867617208128,"absolute_difference_cents_p95":37.77699467505739,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002198976712894503,"log_rms_mae_db":1.3633530822981843,"rms_envelope_nrmse":0.1390012893952159,"envelope_correlation":0.9889277479712741,"rms_gain_error_db":-0.8492745086930319,"reference_mean_power":0.0037766562988771884,"output_mean_power":0.0031058466731799838},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.069519300574134e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006799849644412746,"output_power":0.0005773180863096423,"reference_active":true,"delta_db":-0.7108414650854165},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022275189153607582,"output_power":0.00018651632366757453,"reference_active":true,"delta_db":-0.7710455391123227},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.238302101062001e-05,"output_power":2.900578643051608e-05,"reference_active":true,"delta_db":-1.6470726452589959}],"active_band_count":3,"active_band_mae_db":1.0429865498189115}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018113949472991762,"log_rms_mae_db":1.183747497398889,"rms_envelope_nrmse":0.11173770557087874,"envelope_correlation":0.995774266068976,"rms_gain_error_db":-0.8443048568992039,"reference_mean_power":0.003780325824366865,"output_mean_power":0.003112423941714636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9093794448028363e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006511405512347742,"output_power":0.0005518955119785813,"reference_active":true,"delta_db":-0.7181788045843707},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011667254760432671,"output_power":9.743560480896196e-05,"reference_active":true,"delta_db":-0.782509957292516},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.399065253982207e-06,"output_power":2.0061150669714565e-06,"reference_active":true,"delta_db":-2.290036623758664}],"active_band_count":3,"active_band_mae_db":1.2635751285451835}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013526658282762358,"log_rms_mae_db":2.0805019646638057,"rms_envelope_nrmse":0.20098816495006608,"envelope_correlation":0.9700128531870781,"rms_gain_error_db":-0.2328030874680559,"reference_mean_power":0.001047203990993015,"output_mean_power":0.0009925467689475582},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.584657585149468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003504540347149448,"output_power":0.0003226960427407885,"reference_active":true,"delta_db":-0.358374246707682},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.5576057457372e-05,"output_power":9.138785042250466e-05,"reference_active":true,"delta_db":-0.19460649571560448},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9623586773434205e-05,"output_power":1.994033558709199e-05,"reference_active":true,"delta_db":0.06954072947905926}],"active_band_count":3,"active_band_mae_db":0.20750715730078192}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012031399971827572,"log_rms_mae_db":1.6984490761954552,"rms_envelope_nrmse":0.15251960461624564,"envelope_correlation":0.9818252546645254,"rms_gain_error_db":-0.23488343609329668,"reference_mean_power":0.0010510881038568295,"output_mean_power":0.0009957510595244956},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.061243701565634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003432746241782782,"output_power":0.0003203697296496555,"reference_active":true,"delta_db":-0.2999022560461966},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.625199841508276e-05,"output_power":4.697807979412421e-05,"reference_active":true,"delta_db":0.06764759142350549},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5417688576161993e-06,"output_power":1.7197950078217651e-06,"reference_active":true,"delta_db":0.47457414900937733}],"active_band_count":3,"active_band_mae_db":0.28070799882635983}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013601220330575493,"log_rms_mae_db":2.605292117921385,"rms_envelope_nrmse":0.31509457294052473,"envelope_correlation":0.9647184266435771,"rms_gain_error_db":-2.3394302000271674,"reference_mean_power":0.0002492356992147181,"output_mean_power":0.00014543442845035145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.707381052427382e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017017591719688496,"output_power":9.675932059604453e-05,"reference_active":true,"delta_db":-2.4520528973505584},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.646281785953888e-05,"output_power":1.56137282585257e-05,"reference_active":true,"delta_db":-2.291294709808081},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0978410742643585e-06,"output_power":2.970449366520567e-06,"reference_active":true,"delta_db":-0.18236979869019856}],"active_band_count":3,"active_band_mae_db":1.6419058019496129}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011511226838452263,"log_rms_mae_db":2.477330453211221,"rms_envelope_nrmse":0.284296079103359,"envelope_correlation":0.9789263072082426,"rms_gain_error_db":-2.3394725674260393,"reference_mean_power":0.0002496383892619269,"output_mean_power":0.00014566798575009647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6223780118670245e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016948627965662866,"output_power":9.695933925556862e-05,"reference_active":true,"delta_db":-2.4254489945908784},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1767373976886582e-05,"output_power":6.51778832522911e-06,"reference_active":true,"delta_db":-2.565793039057078},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8952968721807653e-07,"output_power":1.5765721099610954e-07,"reference_active":true,"delta_db":-0.7996340639450055}],"active_band_count":3,"active_band_mae_db":1.9302920325309874}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":2.9811466036374323,"rms_envelope_nrmse":0.2098068612219051,"envelope_correlation":0.9786129175994193,"rms_gain_error_db":-1.0281485136522612,"reference_mean_power":4.1047668613925756e-05,"output_mean_power":3.239467626145581e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.890895346126464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.549789437651058e-06,"output_power":6.201550442126978e-06,"reference_active":true,"delta_db":-1.8749351546242867},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.275656140671785e-06,"output_power":2.173774543255005e-06,"reference_active":true,"delta_db":-1.7808380728157531},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.939546621487426e-07,"output_power":6.325789381055672e-07,"reference_active":true,"delta_db":0.27361432259037033}],"active_band_count":3,"active_band_mae_db":1.3097958500101368}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":2.6621645689851,"rms_envelope_nrmse":0.164390382341642,"envelope_correlation":0.9897607226582448,"rms_gain_error_db":-1.0268534870728114,"reference_mean_power":4.122071638798199e-05,"output_mean_power":3.2540946925799975e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7789380940114685e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.936665760227186e-06,"output_power":6.506628367094127e-06,"reference_active":true,"delta_db":-1.8388467943866729},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7465237791634247e-06,"output_power":1.1903596336168713e-06,"reference_active":true,"delta_db":-1.6649631193778214},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8043700755065895e-08,"output_power":5.113319465307187e-08,"reference_active":true,"delta_db":0.2706647376803554}],"active_band_count":3,"active_band_mae_db":1.2581582171482832}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"golf","status":"measured","mos":{"status":"measured","value":3.5734026432037354,"predicted_mos":3.5734026432037354,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"3c64233e58f7a4be2b34ec3a2e9d1ea80f27202674029990eda6e11aed3d50f6","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.515920748001008},"started_at":"2026-10-13T05:02:55.630264+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026869080002143164,"native_input_samples":144000,"native_input_sha256":"6e1166e08d7c2a9e543eae0f1fdf98b944e0ac0284ca510b3eaa042fa436a047","output_sha256":"ead9d7897da4085d320698db775770c15e3993ca3e3f687dbad44e9f6b64ace9","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.572084243000063,"synthesis_s":0.11427655700026662,"total_s":0.6863608000003296,"rtf":0.11439346666672161,"output_peak":0.49322962760925293,"output_rms":0.0557647676398837,"finished_at":"2026-10-13T05:02:56.326990+00:00","analysis_stages":{"external_f0_s":0.08847487499951967},"evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"ead9d7897da4085d320698db775770c15e3993ca3e3f687dbad44e9f6b64ace9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.071354896240546,"output_peak":0.4941097141345538},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.185590779348539,"spectral_convergence":0.6321442664934243,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.091214793221035,"spectral_convergence":0.29824952741188926,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.17561326271721,"spectral_convergence":0.3096341006301841,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017498816356154704,"output_power":0.0013209537747818985,"delta_db":-1.221210531500069,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012084384696377324,"output_power":0.0011631680064915027,"delta_db":-0.16582094030642874,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000392899732403401,"output_power":0.00034145491143567003,"delta_db":-0.6094836909679945,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00032110561620833536,"output_power":0.00020530393497203572,"delta_db":-1.9425062815407457,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.395819158246405e-05,"output_power":3.8411399384910935e-05,"delta_db":-1.4759725634085654,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.311799628076386e-05,"output_power":4.145367942957943e-05,"delta_db":-0.17095486135889235,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9309914781804493,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.834079049969383,"pitch":null,"output_rms_dbfs":-25.071354896240546,"output_peak":0.4941097141345538,"reference_rms_dbfs":-24.237275846271164,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.484139611762262},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.889287929936323,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06139545294016623,"rms_dbfs":-24.237275846271164,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.822229173010964,"mean":6.641012201325125e-06,"zero_energy":false},"output":{"rms":0.055774059440707714,"rms_dbfs":-25.071354896240546,"peak":0.4941097141345538,"peak_dbfs":-6.1235321570401435,"crest_factor_db":18.9478227392004,"mean":0.00040273377542664044,"zero_energy":false},"rms_difference_db":-0.8340790499693824},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":3.8746957518393655e-07,"output_rms_dbfs":-64.11762393411821,"output_zero_energy":false,"output_peak":0.00797516060820303,"output_energy_fraction":3.659813389096375e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":225,"both_estimated_voiced_frames":202,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.9351851851851852,"absolute_difference_cents_median":4.888323277663213,"absolute_difference_cents_p95":21.148971472818573,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002198976712894503,"log_rms_mae_db":2.395150563159886,"rms_envelope_nrmse":0.15979855526547002,"envelope_correlation":0.9818955510603258,"rms_gain_error_db":-0.833166335990418,"reference_mean_power":0.0037704431277963526,"output_mean_power":0.0031122592100289183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0694315650265903e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006783406203528856,"output_power":0.0005430387454326322,"reference_active":true,"delta_db":-0.9661700693994293},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022318834313268102,"output_power":0.0001621884488644125,"reference_active":true,"delta_db":-1.3865158783800937},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.217218804780056e-05,"output_power":2.3859517574439453e-05,"reference_active":true,"delta_db":-2.4736447614870225}],"active_band_count":3,"active_band_mae_db":1.6087769030888486}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00018113949472991762,"log_rms_mae_db":2.4579043993923273,"rms_envelope_nrmse":0.12054090977001868,"envelope_correlation":0.9930499580377923,"rms_gain_error_db":-0.8314900428813132,"reference_mean_power":0.003779129294821913,"output_mean_power":0.0031206333563236132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.909252292211284e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006497490768951187,"output_power":0.0005157868622088735,"reference_active":true,"delta_db":-1.0027539529390543},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011667797889371582,"output_power":8.209384783224685e-05,"reference_active":true,"delta_db":-1.526782856068295},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3935141261510565e-06,"output_power":1.6946236290883056e-06,"reference_active":true,"delta_db":-3.0157640364024765}],"active_band_count":3,"active_band_mae_db":1.8484336151366085}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00013526658282762358,"log_rms_mae_db":1.7683453242104263,"rms_envelope_nrmse":0.19377988443937644,"envelope_correlation":0.9749043687626129,"rms_gain_error_db":-0.6629492068590509,"reference_mean_power":0.001048287496888658,"output_mean_power":0.0008998818349813629},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.584509549157379e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035046448441086665,"output_power":0.0003009938366507982,"reference_active":true,"delta_db":-0.6608641094330249},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.549170208470694e-05,"output_power":7.561264786022955e-05,"reference_active":true,"delta_db":-1.0137118768183593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.959235135531538e-05,"output_power":1.1312373531764338e-05,"reference_active":true,"delta_db":-2.385328235297993}],"active_band_count":3,"active_band_mae_db":1.3533014071831257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012031399971827572,"log_rms_mae_db":1.5099693758546167,"rms_envelope_nrmse":0.15377624195306647,"envelope_correlation":0.9850416409310006,"rms_gain_error_db":-0.665899997112303,"reference_mean_power":0.0010518597719747903,"output_mean_power":0.0009023350894446393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.061633696679896e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00034330089836852556,"output_power":0.0002914731620432153,"reference_active":true,"delta_db":-0.7107636748365892},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.61348297743637e-05,"output_power":3.598560529238814e-05,"reference_active":true,"delta_db":-1.0790011042220642},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5416441845160533e-06,"output_power":8.012506468689854e-07,"reference_active":true,"delta_db":-2.8421575584213183}],"active_band_count":3,"active_band_mae_db":1.5439741124933242}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00013601220330575493,"log_rms_mae_db":1.9769811309556469,"rms_envelope_nrmse":0.250647476435876,"envelope_correlation":0.9730755232783839,"rms_gain_error_db":-1.522089499135872,"reference_mean_power":0.00024922254104881177,"output_mean_power":0.00017554091986962068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.700023422478146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016980363385657858,"output_power":0.00011753104124345328,"reference_active":true,"delta_db":-1.597943963426828},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.628860238183316e-05,"output_power":1.5329713328186025e-05,"reference_active":true,"delta_db":-2.3423346425946074},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.088133303721018e-06,"output_power":2.8888473557212898e-06,"reference_active":true,"delta_db":-0.28971444321866435}],"active_band_count":3,"active_band_mae_db":1.4099976830800331}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00011511226838452263,"log_rms_mae_db":1.8150744510647767,"rms_envelope_nrmse":0.19153200898576145,"envelope_correlation":0.9919872018679863,"rms_gain_error_db":-1.5289610513014071,"reference_mean_power":0.00024912893169203086,"output_mean_power":0.00017519756289065153},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.615198416776522e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016905896243671892,"output_power":0.0001164873561669031,"reference_active":true,"delta_db":-1.6175941076260933},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1666918286370517e-05,"output_power":6.553224020517667e-06,"reference_active":true,"delta_db":-2.5050114166282906},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.891827337842046e-07,"output_power":1.5817376838429472e-07,"reference_active":true,"delta_db":-0.7774703542496955}],"active_band_count":3,"active_band_mae_db":1.6333586261680264}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":2.295714263930029,"rms_envelope_nrmse":0.23172772900179187,"envelope_correlation":0.9739090098910973,"rms_gain_error_db":0.42596718203199346,"reference_mean_power":4.109741607471854e-05,"output_mean_power":4.5332663383007536e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.8926814674506913e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.563245279683278e-06,"output_power":8.686857012156104e-06,"reference_active":true,"delta_db":-0.4174262179937702},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2813206247585914e-06,"output_power":3.3989629401005075e-06,"reference_active":true,"delta_db":0.15297761078338537},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.89519509083114e-07,"output_power":9.395808534883315e-07,"reference_active":true,"delta_db":2.024359764059854}],"active_band_count":3,"active_band_mae_db":0.8649211976123365}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":2.198660610363674,"rms_envelope_nrmse":0.17560793810414807,"envelope_correlation":0.9861884229934568,"rms_gain_error_db":0.4267562266005376,"reference_mean_power":4.1263855832095826e-05,"output_mean_power":4.552452573653083e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.7804274875484423e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.954146303870926e-06,"output_power":9.365661259466301e-06,"reference_active":true,"delta_db":-0.264655736060732},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7418516393724393e-06,"output_power":1.8982252078617484e-06,"reference_active":true,"delta_db":0.3733657478996439},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8131201617425546e-08,"output_power":7.883355583330719e-08,"reference_active":true,"delta_db":2.1428441193670036}],"active_band_count":3,"active_band_mae_db":0.9269552011091265}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.389188289642334,"predicted_mos":3.389188289642334,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"53fa0cffa6e7bc354bbf179e5950d61e41227aaa65ccb050517f30e1221db135","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1522379540001566},"started_at":"2026-10-13T07:44:02.602678+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004648539998015622,"native_input_samples":144000,"native_input_sha256":"6e1166e08d7c2a9e543eae0f1fdf98b944e0ac0284ca510b3eaa042fa436a047","output_sha256":"fe4f368f6097fa30f156101ccd9b4e67742f7c0338dc78c0d01ab63d7f9b76b8","output_samples":144120,"length_delta_samples":120,"analysis_s":2.2523413759990945,"synthesis_s":3.669565527001396,"total_s":5.92190690300049,"rtf":0.9869844838334151,"output_peak":0.48162224888801575,"output_rms":0.0635211244743278,"finished_at":"2026-10-13T07:44:08.545225+00:00","analysis_stages":{"external_f0_s":1.4665282940004545},"evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"fe4f368f6097fa30f156101ccd9b4e67742f7c0338dc78c0d01ab63d7f9b76b8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.94032560169392,"output_peak":0.4784930523612216},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.849304675331477,"spectral_convergence":0.5464017010495683,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.867244597936265,"spectral_convergence":0.3043498714656005,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.878378262407961,"spectral_convergence":0.30444421610850536,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.0022331478304132126,"delta_db":1.0368710148815343,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.001109673448510491,"delta_db":-0.36572314360965885,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.00034542460366604173,"delta_db":-0.5576837303289229,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.000262850752026297,"delta_db":-0.8639943970015701,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":5.2904220694603645e-05,"delta_db":-0.09827635323381387,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":3.3460545334155445e-05,"delta_db":-1.0951708817407408,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6696199201327069,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.290006627895735,"pitch":null,"output_rms_dbfs":-23.938916166279242,"output_peak":0.4784930523612216,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.198309178558567},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9044520886101829,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06145452415925838,"rms_dbfs":-24.228922794174977,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.813876120914777,"mean":2.243359883626302e-05,"zero_energy":false},"output":{"rms":0.06354102140070461,"rms_dbfs":-23.938916166279242,"peak":0.4784930523612216,"peak_dbfs":-6.4024872744862895,"crest_factor_db":17.53642889179295,"mean":3.871437598153298e-05,"zero_energy":false},"rms_difference_db":0.2900066278957354},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":1.1923660125714992e-06,"output_rms_dbfs":-59.23590411675819,"output_zero_energy":false,"output_peak":0.04912123689417744,"output_energy_fraction":8.662886460415812e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":223,"both_estimated_voiced_frames":202,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.9351851851851852,"absolute_difference_cents_median":7.320501425837596,"absolute_difference_cents_p95":36.70812456644101,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002198976712894503,"log_rms_mae_db":1.8186808558885021,"rms_envelope_nrmse":0.1839860239981144,"envelope_correlation":0.9691857651676201,"rms_gain_error_db":0.2909052166858279,"reference_mean_power":0.0037766562988771884,"output_mean_power":0.004038294472989479},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.069519300574134e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006799849644412746,"output_power":0.0006641610901715074,"reference_active":true,"delta_db":-0.1022588096707113},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022275189153607582,"output_power":0.0002196993558277648,"reference_active":true,"delta_db":-0.05992617067182396},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.238302101062001e-05,"output_power":4.0473504298381093e-05,"reference_active":true,"delta_db":-0.20021101130637403}],"active_band_count":3,"active_band_mae_db":0.12079866388296977}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018113949472991762,"log_rms_mae_db":1.6135923254344673,"rms_envelope_nrmse":0.12758760109994957,"envelope_correlation":0.9846146613701002,"rms_gain_error_db":0.28444578481293975,"reference_mean_power":0.003780325824366865,"output_mean_power":0.004036210534290929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9093794448028363e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006511405512347742,"output_power":0.0006375077886730605,"reference_active":true,"delta_db":-0.09189247771869789},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011667254760432671,"output_power":0.0001127187363664759,"reference_active":true,"delta_db":-0.14972569541417752},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.399065253982207e-06,"output_power":3.1839728019238616e-06,"reference_active":true,"delta_db":-0.2839015282712642}],"active_band_count":3,"active_band_mae_db":0.17517323380137986}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013526658282762358,"log_rms_mae_db":1.7837323313634346,"rms_envelope_nrmse":0.2075217201525085,"envelope_correlation":0.969094742477846,"rms_gain_error_db":-0.5590215381872597,"reference_mean_power":0.001047203990993015,"output_mean_power":0.0009207233026171637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.584657585149468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003504540347149448,"output_power":0.00030469442263291607,"reference_active":true,"delta_db":-0.6076655967681902},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.5576057457372e-05,"output_power":8.240872089543098e-05,"reference_active":true,"delta_db":-0.643759384660459},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9623586773434205e-05,"output_power":1.7262848892134892e-05,"reference_active":true,"delta_db":-0.5566592105725752}],"active_band_count":3,"active_band_mae_db":0.6026947306670748}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012031399971827572,"log_rms_mae_db":1.4461519017437057,"rms_envelope_nrmse":0.14083298776046227,"envelope_correlation":0.9865994416480733,"rms_gain_error_db":-0.5582591530600369,"reference_mean_power":0.0010510881038568295,"output_mean_power":0.0009243005373001475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.061243701565634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003432746241782782,"output_power":0.00029857294925466076,"reference_active":true,"delta_db":-0.605912423090587},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.625199841508276e-05,"output_power":3.9627201952342116e-05,"reference_active":true,"delta_db":-0.6713709388694379},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5417688576161993e-06,"output_power":1.4090461871009307e-06,"reference_active":true,"delta_db":-0.390940399805158}],"active_band_count":3,"active_band_mae_db":0.5560745872550609}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013601220330575493,"log_rms_mae_db":1.8225718468271548,"rms_envelope_nrmse":0.26414541291338944,"envelope_correlation":0.9570298997611391,"rms_gain_error_db":-0.6762792224624353,"reference_mean_power":0.0002492356992147181,"output_mean_power":0.00021329581100347085},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.707381052427382e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017017591719688496,"output_power":0.00013842561588932258,"reference_active":true,"delta_db":-0.8968163552893},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.646281785953888e-05,"output_power":2.1917136077365864e-05,"reference_active":true,"delta_db":-0.8185228355750633},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0978410742643585e-06,"output_power":4.588302988142672e-06,"reference_active":true,"delta_db":1.705929549920311}],"active_band_count":3,"active_band_mae_db":1.1404229135948916}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011511226838452263,"log_rms_mae_db":1.3652474805275845,"rms_envelope_nrmse":0.16770820140436687,"envelope_correlation":0.9844265774736567,"rms_gain_error_db":-0.67251916010436,"reference_mean_power":0.0002496383892619269,"output_mean_power":0.00021382548004915492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6223780118670245e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016948627965662866,"output_power":0.00013896469494222342,"reference_active":true,"delta_db":-0.8623006833588829},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1767373976886582e-05,"output_power":9.839300251860802e-06,"reference_active":true,"delta_db":-0.7771534239466904},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8952968721807653e-07,"output_power":2.932714773929783e-07,"reference_active":true,"delta_db":1.8959258111149775}],"active_band_count":3,"active_band_mae_db":1.1784599728068503}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":1.8001725103311983,"rms_envelope_nrmse":0.2275778221610371,"envelope_correlation":0.9704928532704277,"rms_gain_error_db":-0.7570479677448929,"reference_mean_power":4.1047668613925756e-05,"output_mean_power":3.448130537388687e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.890895346126464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.549789437651058e-06,"output_power":7.631346372643838e-06,"reference_active":true,"delta_db":-0.9739263015181581},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.275656140671785e-06,"output_power":2.5640712319169636e-06,"reference_active":true,"delta_db":-1.0636821971562016},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.939546621487426e-07,"output_power":4.595465320368774e-07,"reference_active":true,"delta_db":-1.1142380255060065}],"active_band_count":3,"active_band_mae_db":1.050615508060122}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":1.265323034522628,"rms_envelope_nrmse":0.15566994828992992,"envelope_correlation":0.9882447713394289,"rms_gain_error_db":-0.756350768527584,"reference_mean_power":4.122071638798199e-05,"output_mean_power":3.4632230106996516e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7789380940114685e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.936665760227186e-06,"output_power":7.870533032831345e-06,"reference_active":true,"delta_db":-1.012365356928381},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7465237791634247e-06,"output_power":1.368585821802611e-06,"reference_active":true,"delta_db":-1.059024664891311},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8043700755065895e-08,"output_power":3.33107721954195e-08,"reference_active":true,"delta_db":-1.590517528321853}],"active_band_count":3,"active_band_mae_db":1.2206358500471817}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.5097665786743164,"predicted_mos":3.5097665786743164,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"44187dc64bad32b621e991ab0bb4705deceebe3bc85f0776be7770cbdbe365b7","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.407336395999664},"started_at":"2026-10-13T05:11:16.112898+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002888165001422749,"native_input_samples":144000,"native_input_sha256":"6e1166e08d7c2a9e543eae0f1fdf98b944e0ac0284ca510b3eaa042fa436a047","output_sha256":"50a539edd90e7cabb93bac3712ef2ea2e738a3154023de9b62d0014861ddd258","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9339391720004642,"synthesis_s":0.1474481579989515,"total_s":1.0813873299994157,"rtf":0.18023122166656927,"output_peak":0.5265323519706726,"output_rms":0.06011277660089509,"finished_at":"2026-10-13T05:11:17.206768+00:00","analysis_stages":{"external_f0_s":0.14785453300100926},"evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"50a539edd90e7cabb93bac3712ef2ea2e738a3154023de9b62d0014861ddd258","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.419753063495623,"output_peak":0.5261813964417145},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.504636080555352,"spectral_convergence":0.18217722912438616,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.121474921440062,"spectral_convergence":0.17178997946605873,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.425004505639254,"spectral_convergence":0.18651905542355268,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017588563541266742,"output_power":0.0016969550267912211,"delta_db":-0.15560039388761468,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012071674288590822,"output_power":0.0011710335109007476,"delta_db":-0.13198185780228935,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003927549262280726,"output_power":0.0003734880860127879,"delta_db":-0.2184488828747352,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003207071656425,"output_power":0.00029695359203117395,"delta_db":-0.33420080601267754,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.4115036625953016e-05,"output_power":4.5603689663253254e-05,"delta_db":-0.7431797511836933,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.305762815858983e-05,"output_power":3.390516856599071e-05,"delta_db":-1.0378419501818434,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4368756069904756,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.18646471509099655,"pitch":null,"output_rms_dbfs":-24.415387509265972,"output_peak":0.5261813964417145,"reference_rms_dbfs":-24.228922794174977,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.3503718358782235},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.957068640002302,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06145452415925838,"rms_dbfs":-24.228922794174977,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.813876120914777,"mean":2.243359883626302e-05,"zero_energy":false},"output":{"rms":0.06014930651275354,"rms_dbfs":-24.415387509265972,"peak":0.5261813964417145,"peak_dbfs":-5.577290216056589,"crest_factor_db":18.838097293209383,"mean":5.78974427413339e-05,"zero_energy":false},"rms_difference_db":-0.18646471509099527},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":2.587770716684574e-07,"output_rms_dbfs":-65.87074205936074,"output_zero_energy":false,"output_peak":0.012844558700315735,"output_energy_fraction":2.0980989306587246e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":217,"both_estimated_voiced_frames":207,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9583333333333334,"absolute_difference_cents_median":1.4574699426174054,"absolute_difference_cents_p95":9.288234346118594,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002198976712894503,"log_rms_mae_db":0.635125848519798,"rms_envelope_nrmse":0.07636906040616132,"envelope_correlation":0.9946712355855966,"rms_gain_error_db":-0.18975974803005488,"reference_mean_power":0.0037766562988771884,"output_mean_power":0.003615193005767511},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.069519300574134e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006799849644412746,"output_power":0.0006279902134998091,"reference_active":true,"delta_db":-0.345464340511076},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022275189153607582,"output_power":0.000185956212854154,"reference_active":true,"delta_db":-0.7841070776166781},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.238302101062001e-05,"output_power":2.948922312362766e-05,"reference_active":true,"delta_db":-1.575285779971098}],"active_band_count":3,"active_band_mae_db":0.9016190660329507}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018113949472991762,"log_rms_mae_db":0.5057048238990245,"rms_envelope_nrmse":0.048084720003751476,"envelope_correlation":0.9980243335011745,"rms_gain_error_db":-0.1853854779239002,"reference_mean_power":0.003780325824366865,"output_mean_power":0.003622352291350148},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9093794448028363e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006511405512347742,"output_power":0.0005952371803414433,"reference_active":true,"delta_db":-0.3898469209818365},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011667254760432671,"output_power":9.630009526437893e-05,"reference_active":true,"delta_db":-0.8334196426286511},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.399065253982207e-06,"output_power":2.066395586651688e-06,"reference_active":true,"delta_db":-2.1614603647338564}],"active_band_count":3,"active_band_mae_db":1.1282423094481147}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013526658282762358,"log_rms_mae_db":0.881907522522346,"rms_envelope_nrmse":0.08186350441522905,"envelope_correlation":0.9953984004990677,"rms_gain_error_db":-0.219532574737772,"reference_mean_power":0.001047203990993015,"output_mean_power":0.000995584281394053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.584657585149468e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003504540347149448,"output_power":0.0003307577179137379,"reference_active":true,"delta_db":-0.25121077480963805},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.5576057457372e-05,"output_power":8.250327260550541e-05,"reference_active":true,"delta_db":-0.6387793599130008},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9623586773434205e-05,"output_power":1.2363759472390347e-05,"reference_active":true,"delta_db":-2.006278424725994}],"active_band_count":3,"active_band_mae_db":0.9654228531495442}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012031399971827572,"log_rms_mae_db":0.591918932300098,"rms_envelope_nrmse":0.04914578952828588,"envelope_correlation":0.9985520476433032,"rms_gain_error_db":-0.21772132526944132,"reference_mean_power":0.0010510881038568295,"output_mean_power":0.0009996937760765963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.061243701565634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003432746241782782,"output_power":0.00032086030739815724,"reference_active":true,"delta_db":-0.29325705110016176},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.625199841508276e-05,"output_power":3.953222797712029e-05,"reference_active":true,"delta_db":-0.6817921087013081},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5417688576161993e-06,"output_power":7.924780675765892e-07,"reference_active":true,"delta_db":-2.890320174232013}],"active_band_count":3,"active_band_mae_db":1.2884564446778277}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00013601220330575493,"log_rms_mae_db":0.9840294074858861,"rms_envelope_nrmse":0.15014033301243743,"envelope_correlation":0.9866030666654968,"rms_gain_error_db":-0.355440571436525,"reference_mean_power":0.0002492356992147181,"output_mean_power":0.00022964986336602766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.707381052427382e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017017591719688496,"output_power":0.00016085277901995148,"reference_active":true,"delta_db":-0.24469531481170662},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.646281785953888e-05,"output_power":2.0537057294645136e-05,"reference_active":true,"delta_db":-1.1009787289857011},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0978410742643585e-06,"output_power":5.5947336066777756e-06,"reference_active":true,"delta_db":2.56720278640469}],"active_band_count":3,"active_band_mae_db":1.3042922767340326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011511226838452263,"log_rms_mae_db":0.7748665903937837,"rms_envelope_nrmse":0.10084628667639481,"envelope_correlation":0.9943539640677391,"rms_gain_error_db":-0.3544288904064438,"reference_mean_power":0.0002496383892619269,"output_mean_power":0.00023007449777419126},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.6223780118670245e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00016948627965662866,"output_power":0.0001599783966332881,"reference_active":true,"delta_db":-0.2507320692101938},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1767373976886582e-05,"output_power":8.774163103782582e-06,"reference_active":true,"delta_db":-1.2747385262433555},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8952968721807653e-07,"output_power":2.334321640881118e-07,"reference_active":true,"delta_db":0.9048345039520828}],"active_band_count":3,"active_band_mae_db":0.8101016998018773}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":1.0060047234458835,"rms_envelope_nrmse":0.10811078578022615,"envelope_correlation":0.9943501213970918,"rms_gain_error_db":-0.4726173147375278,"reference_mean_power":4.1047668613925756e-05,"output_mean_power":3.6815166083521284e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.890895346126464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.549789437651058e-06,"output_power":8.451975900305478e-06,"reference_active":true,"delta_db":-0.5303554592525735},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.275656140671785e-06,"output_power":2.7288023330751847e-06,"reference_active":true,"delta_db":-0.7932622805933003},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.939546621487426e-07,"output_power":3.7807061448425895e-07,"reference_active":true,"delta_db":-1.9618037244209279}],"active_band_count":3,"active_band_mae_db":1.095140488088934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":0.7877802253404464,"rms_envelope_nrmse":0.06929797771747344,"envelope_correlation":0.9986946336338743,"rms_gain_error_db":-0.4734382971005993,"reference_mean_power":4.122071638798199e-05,"output_mean_power":3.696338242435961e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7789380940114685e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.936665760227186e-06,"output_power":8.690555359575047e-06,"reference_active":true,"delta_db":-0.581931512895403},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7465237791634247e-06,"output_power":1.4142980292614229e-06,"reference_active":true,"delta_db":-0.9163356678203726},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8043700755065895e-08,"output_power":2.9173416849335943e-08,"reference_active":true,"delta_db":-2.1664915571537784}],"active_band_count":3,"active_band_mae_db":1.2215862459565179}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-79740-0011","method":"magphase","status":"measured","mos":{"status":"measured","value":3.475034236907959,"predicted_mos":3.475034236907959,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287913,"input_sample_rate":48000,"input_seconds":5.9981875,"input_float32_sha256":"19b28f08e7f20ef274d049639ca81b16e1809313ec58e76de46d4dcc2b07504e","internal_samples":95971,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.333112334001271},"started_at":"2026-10-13T05:17:40.337025+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008747426998525043,"native_input_samples":288000,"native_input_sha256":"7157ea3c7546c5155d33e1679e32cb581099b27a6073ac2f5c13a0e5f27d1ef8","output_sha256":"d8c38e05df151cf1e37f018a7827a33c5ad385434bfbbbdfc6d881311441309e","output_samples":287913,"length_delta_samples":-87,"analysis_s":3.375019319000785,"synthesis_s":1.0165598900002806,"total_s":4.3915792090010655,"rtf":0.7319298681668442,"output_peak":0.5705448750103377,"output_rms":0.05913211076407054,"finished_at":"2026-10-13T05:17:44.749818+00:00","evaluation_input":{"reference":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"d8c38e05df151cf1e37f018a7827a33c5ad385434bfbbbdfc6d881311441309e","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95971,"length_delta_samples":-29,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.562993731202248,"output_peak":0.5501334036852155},"common_prefix":{"samples":95971,"sample_rate":16000,"reference_coverage":0.9996979166666666,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95971,"output_samples":95971,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.172577554992758,"spectral_convergence":0.6579508897912821,"spectral_convergence_status":"ok","amplitude_floor":1.6898412267444875e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.103959795261514,"spectral_convergence":0.41223795953908104,"spectral_convergence_status":"ok","amplitude_floor":1.0945626100175482e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.618754969286394,"spectral_convergence":0.3436548112463582,"spectral_convergence_status":"ok","amplitude_floor":6.58943811507865e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017577943883139184,"output_power":0.0015635364022807386,"delta_db":-0.5086007672312058,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0012082728373421924,"output_power":0.0011483701739185988,"delta_db":-0.22083108164946144,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003922395543126594,"output_power":0.00037994912392945955,"delta_db":-0.13825939295971565,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00032070382907719165,"output_power":0.00030987538989669434,"delta_db":-0.1491705904254091,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.409319055280006e-05,"output_power":5.739793268996299e-05,"delta_db":0.2575365264815602,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.306874575656626e-05,"output_power":3.7911194497304286e-05,"delta_db":-0.5539475617514817,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30472432008313904,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.33351203498282345,"pitch":null,"output_rms_dbfs":-24.562993731202248,"output_peak":0.5501334036852155,"reference_rms_dbfs":-24.229481696219427,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.298430773180222},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95971,"output_samples":95971,"intelligibility":{"estoi":{"value":0.8177875407989849,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06145056993515462,"rms_dbfs":-24.229481696219427,"peak":0.536102294921875,"peak_dbfs":-5.415046673260199,"crest_factor_db":18.814435022959227,"mean":6.80175257206604e-07,"zero_energy":false},"output":{"rms":0.05913577781016123,"rms_dbfs":-24.562993731202248,"peak":0.5501334036852155,"peak_dbfs":-5.190639684326722,"crest_factor_db":19.372354046875525,"mean":-0.0010072942773678877,"zero_energy":false},"rms_difference_db":-0.3335120349828209},"full_scale":{"reference":{"sample_count":95971,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95971,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":131,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.751598841045052e-06,"threshold_dbfs":-54.257836069760515,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":176,"quiet_samples":28160,"reference_mean_power":2.2082481750244784e-07,"output_mean_power":1.0996640316512702e-06,"output_rms_dbfs":-59.58739979829684,"output_zero_energy":false,"output_peak":0.020829187020856878,"output_energy_fraction":9.226819476058012e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":216,"output_estimated_voiced_frames":218,"both_estimated_voiced_frames":201,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9305555555555556,"absolute_difference_cents_median":8.475291703435202,"absolute_difference_cents_p95":67.38423586141593,"difference_over_600_cents_frames":0,"uncovered_tail_samples":131,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95971,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1198,"rms_floor":0.0002198976712894503,"log_rms_mae_db":2.200417167995457,"rms_envelope_nrmse":0.2708181040061603,"envelope_correlation":0.928158601338758,"rms_gain_error_db":-0.3253699947071794,"reference_mean_power":0.003773027848050415,"output_mean_power":0.003500684818835217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1198,"frequency_resolution_hz":0.1669449081803005,"power_floor":2.069476579203875e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006791587507821139,"output_power":0.0006380327584767278,"reference_active":true,"delta_db":-0.2712832358614934},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022297180203788942,"output_power":0.0002160661426852648,"reference_active":true,"delta_db":-0.13663225022581574},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.2278690818506895e-05,"output_power":4.048278330334466e-05,"reference_active":true,"delta_db":-0.18851166885399298}],"active_band_count":3,"active_band_mae_db":0.1988090516471007}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1192,"rms_floor":0.00018113949472991762,"log_rms_mae_db":1.7576332300016928,"rms_envelope_nrmse":0.1854075608285528,"envelope_correlation":0.9644382141661355,"rms_gain_error_db":-0.31692182354883774,"reference_mean_power":0.003779354150579491,"output_mean_power":0.0035133822895093217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1192,"frequency_resolution_hz":0.16778523489932887,"power_floor":1.909318902524677e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006504406623269672,"output_power":0.0006135093067246072,"reference_active":true,"delta_db":-0.2538652816304859},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001166798549357764,"output_power":0.00011102544845810426,"reference_active":true,"delta_db":-0.21573344438378605},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3941722398312818e-06,"output_power":3.388752416372119e-06,"reference_active":true,"delta_db":-0.006940368768478543}],"active_band_count":3,"active_band_mae_db":0.1588463649275835}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1198,"rms_floor":0.00013526658282762358,"log_rms_mae_db":2.7436385964772834,"rms_envelope_nrmse":0.33344819169714063,"envelope_correlation":0.9167381447360048,"rms_gain_error_db":-0.21700766643660435,"reference_mean_power":0.0010477571466537459,"output_mean_power":0.0009966894589381044},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1198,"frequency_resolution_hz":0.1669449081803005,"power_floor":7.584573503858926e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035045950197451226,"output_power":0.0003305009440640905,"reference_active":true,"delta_db":-0.25465135093352537},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.553211532075457e-05,"output_power":9.370285258963559e-05,"reference_active":true,"delta_db":-0.08396581935122932},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9606684996561682e-05,"output_power":1.9416739979527037e-05,"reference_active":true,"delta_db":-0.04227856693529472}],"active_band_count":3,"active_band_mae_db":0.12696524574001647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1192,"rms_floor":0.00012031399971827572,"log_rms_mae_db":2.221030169273968,"rms_envelope_nrmse":0.2271360265857488,"envelope_correlation":0.9591224106703434,"rms_gain_error_db":-0.21537223408219158,"reference_mean_power":0.0010514563564835658,"output_mean_power":0.0010005850908743844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1192,"frequency_resolution_hz":0.16778523489932887,"power_floor":7.061444078188054e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003432882986904638,"output_power":0.00032497850123942287,"reference_active":true,"delta_db":-0.23804368901304548},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.61923258017295e-05,"output_power":4.589944655992095e-05,"reference_active":true,"delta_db":-0.027623807025135973},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5413227355485603e-06,"output_power":1.577232637182837e-06,"reference_active":true,"delta_db":0.1000217062143554}],"active_band_count":3,"active_band_mae_db":0.12189640075084562}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1198,"rms_floor":0.00013601220330575493,"log_rms_mae_db":2.7423629183853664,"rms_envelope_nrmse":0.4161936474115963,"envelope_correlation":0.892293198707007,"rms_gain_error_db":-0.09341507818386577,"reference_mean_power":0.00024926839171775814,"output_mean_power":0.00024396397664835538},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1198,"frequency_resolution_hz":0.1669449081803005,"power_floor":2.703706028755951e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001699908620754683,"output_power":0.00016380480083199494,"reference_active":true,"delta_db":-0.16098950315879285},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6375919014519467e-05,"output_power":2.5718236001209694e-05,"reference_active":true,"delta_db":-0.10966423396444185},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.092973727972727e-06,"output_power":5.565916115460631e-06,"reference_active":true,"delta_db":2.5516042566881074}],"active_band_count":3,"active_band_mae_db":0.9407526646037807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1192,"rms_floor":0.00011511226838452263,"log_rms_mae_db":2.200371552323587,"rms_envelope_nrmse":0.29493630851736546,"envelope_correlation":0.9436667416460389,"rms_gain_error_db":-0.10933383030939323,"reference_mean_power":0.00024939414815172933,"output_mean_power":0.0002431940111252994},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1192,"frequency_resolution_hz":0.16778523489932887,"power_floor":2.618796110279332e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001692734685758238,"output_power":0.0001648182467626779,"reference_active":true,"delta_db":-0.11583603341391333},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1716913058596465e-05,"output_power":1.1497162647458697e-05,"reference_active":true,"delta_db":-0.08222532139609603},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.892390963664183e-07,"output_power":3.4302937263589905e-07,"reference_active":true,"delta_db":2.5832044345775933}],"active_band_count":3,"active_band_mae_db":0.9270885964625343}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1198,"rms_floor":4.380082655552893e-05,"log_rms_mae_db":2.681975270694679,"rms_envelope_nrmse":0.24857291657366964,"envelope_correlation":0.9634829869321201,"rms_gain_error_db":0.1391779338490212,"reference_mean_power":4.107277313510432e-05,"output_mean_power":4.241034640365378e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1198,"frequency_resolution_hz":0.1669449081803005,"power_floor":1.891788651764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.556543521291482e-06,"output_power":9.282983481971346e-06,"reference_active":true,"delta_db":-0.12613264279656028},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2784249973981726e-06,"output_power":2.8144806223427986e-06,"reference_active":true,"delta_db":-0.6626698959121962},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.916779896858499e-07,"output_power":5.685489083220695e-07,"reference_active":true,"delta_db":-0.17317583926371674}],"active_band_count":3,"active_band_mae_db":0.32065945932415774}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95920,"uncovered_tail_samples":51,"metrics":{"frame_count":1192,"rms_floor":3.772714769720091e-05,"log_rms_mae_db":2.021589311065602,"rms_envelope_nrmse":0.17055585112434102,"envelope_correlation":0.9823015236316928,"rms_gain_error_db":0.13901552045406043,"reference_mean_power":4.124177512563452e-05,"output_mean_power":4.2583259583301044e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1192,"frequency_resolution_hz":0.16778523489932887,"power_floor":1.7796731486797748e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945270514472732e-06,"output_power":9.535377807193416e-06,"reference_active":true,"delta_db":-0.18278695389657149},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.74427672315164e-06,"output_power":1.4423292552800082e-06,"reference_active":true,"delta_db":-0.8255097274773621},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.805574097919838e-08,"output_power":4.5063279220823406e-08,"reference_active":true,"delta_db":-0.2792248627565328}],"active_band_count":3,"active_band_mae_db":0.4291738480434888}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"hnm","status":"measured","mos":{"status":"measured","value":3.001762628555298,"predicted_mos":3.001762628555298,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44400,"input_sample_rate":16000,"input_seconds":2.775,"input_float32_sha256":"42013615992ffe31ac4e71850de584dc61f04fa5ebe98e19cc70546fe437fdf4","internal_samples":44400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.186960049000845},"started_at":"2026-10-13T04:40:00.922737+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.0721000762714539e-05,"native_input_samples":44400,"native_input_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","output_sha256":"6901a4227acc4e5e30b01132711c048edf9d074ac763e5a918d300955cd706c0","output_samples":44400,"length_delta_samples":0,"analysis_s":0.47400447499967413,"synthesis_s":0.22718828699998994,"total_s":0.7011927619996641,"rtf":0.2526820763962753,"output_peak":0.4032771281173782,"output_rms":0.05229166585810932,"finished_at":"2026-10-13T04:40:01.627015+00:00","evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"6901a4227acc4e5e30b01132711c048edf9d074ac763e5a918d300955cd706c0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.727404506738651,"spectral_convergence":0.1746043466059582,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4995734052726695,"spectral_convergence":0.16875454801886744,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.523391480235755,"spectral_convergence":0.17045220958529966,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.0018253567751846122,"delta_db":-0.10126238364117955,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0005084386319537655,"delta_db":-0.14382478621299924,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.000260062505824243,"delta_db":0.10629319445381065,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.00012620092817021238,"delta_db":-0.6650048707574939,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":8.379380706127643e-06,"delta_db":-0.5554799095171071,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":5.980022160979628e-06,"delta_db":-0.06563422027051188,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2729165608088504,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.119075239144832,"pitch":null,"output_rms_dbfs":-25.63135045219613,"output_peak":0.4032771281173782,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.583456464082359},"common_prefix":{"samples":44400,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.727404506738651,"spectral_convergence":0.1746043466059582,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4995734052726695,"spectral_convergence":0.16875454801886744,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.523391480235755,"spectral_convergence":0.17045220958529966,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.0018253567751846122,"delta_db":-0.10126238364117955,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0005084386319537655,"delta_db":-0.14382478621299924,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.000260062505824243,"delta_db":0.10629319445381065,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.00012620092817021238,"delta_db":-0.6650048707574939,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":8.379380706127643e-06,"delta_db":-0.5554799095171071,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":5.980022160979628e-06,"delta_db":-0.06563422027051188,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2729165608088504,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.119075239144832,"pitch":null,"output_rms_dbfs":-25.63135045219613,"output_peak":0.4032771281173782,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.583456464082359},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44400,"output_samples":44400,"intelligibility":{"estoi":{"value":0.8875360857667918,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05301347090551401,"rms_dbfs":-25.512275213051296,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.796429980110695,"mean":-1.35583276147241e-05,"zero_energy":false},"output":{"rms":0.05229166585810932,"rms_dbfs":-25.63135045219613,"peak":0.4032771281173782,"peak_dbfs":-7.8879281665467165,"crest_factor_db":17.743422285649412,"mean":-0.00015079244704122803,"zero_energy":false},"rms_difference_db":-0.1190752391448342},"full_scale":{"reference":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":2.7791894946550897e-07,"output_rms_dbfs":-65.56081840528363,"output_zero_energy":false,"output_peak":0.014065713315553363,"output_energy_fraction":3.1498412666903486e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":71,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9848484848484849,"absolute_difference_cents_median":1.4243246160357153,"absolute_difference_cents_p95":7.983351911625459,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":0.00014622417526559514,"log_rms_mae_db":0.8855634412120855,"rms_envelope_nrmse":0.05598126670171155,"envelope_correlation":0.9974661064176399,"rms_gain_error_db":-0.11907440085526702,"reference_mean_power":0.0028155003489198547,"output_mean_power":0.00273935391624351},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.019039396656961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008418483111337408,"output_power":0.00082351632374032,"reference_active":true,"delta_db":-0.09561632800516642},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001428627689371398,"output_power":0.000138548556134708,"reference_active":true,"delta_db":-0.13317059077605625},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9192224262112578e-05,"output_power":1.5390809321396392e-05,"reference_active":true,"delta_db":-0.9586385205683436}],"active_band_count":3,"active_band_mae_db":0.3958084797831887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":0.00013893685565863478,"log_rms_mae_db":0.5599435874126666,"rms_envelope_nrmse":0.0333863989947688,"envelope_correlation":0.9991489391470759,"rms_gain_error_db":-0.11907385699178029,"reference_mean_power":0.0028463248833832306,"output_mean_power":0.0027693451345555766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.8422021834325858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007582285194051116,"output_power":0.0007415216681445662,"reference_active":true,"delta_db":-0.09676269454411066},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661169958011629e-05,"output_power":6.600629651864085e-05,"reference_active":true,"delta_db":-0.039651486686113374},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.710179663683521e-06,"output_power":1.353863805063235e-06,"reference_active":true,"delta_db":-1.0146676006005546}],"active_band_count":3,"active_band_mae_db":0.3836939272769262}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":1.199387242980994,"rms_envelope_nrmse":0.12175826692103799,"envelope_correlation":0.9897111842606955,"rms_gain_error_db":-0.16895673646752432,"reference_mean_power":0.0005201241069420424,"output_mean_power":0.0005002778899932115},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":5.057137989146479e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002701505367125327,"output_power":0.0002576431100685394,"reference_active":true,"delta_db":-0.20587301713177134},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.404522229736366e-05,"output_power":5.12101699337428e-05,"reference_active":true,"delta_db":-0.23401090528464374},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.56355116704778e-06,"output_power":5.878086549933732e-06,"reference_active":true,"delta_db":-1.094897728154687}],"active_band_count":3,"active_band_mae_db":0.5115938835237007}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":0.771769503079965,"rms_envelope_nrmse":0.07162042731288673,"envelope_correlation":0.9964135588854115,"rms_gain_error_db":-0.1689569528833475,"reference_mean_power":0.0005258188868038473,"output_mean_power":0.000505755350673729},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":4.6537380366659205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025071266719842557,"output_power":0.00023919488638419936,"reference_active":true,"delta_db":-0.20424386265943295},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3806665370442115e-05,"output_power":2.2465124152050988e-05,"reference_active":true,"delta_db":-0.25189744454697854},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.079732275271239e-06,"output_power":9.668747919045009e-07,"reference_active":true,"delta_db":-0.4794584589021261}],"active_band_count":3,"active_band_mae_db":0.31186658870284584}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":1.4457940588797913,"rms_envelope_nrmse":0.1721732507587154,"envelope_correlation":0.9822552098979535,"rms_gain_error_db":-0.5603567710542077,"reference_mean_power":0.00010733282194912887,"output_mean_power":9.434021696158464e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":1.1016693525679888e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.607161402154181e-05,"output_power":5.811558351491441e-05,"reference_active":true,"delta_db":-0.5572231330654218},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0221078826476041e-05,"output_power":7.838313434077595e-06,"reference_active":true,"delta_db":-1.1527411183065692},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3761448978787783e-06,"output_power":1.5027265633837882e-06,"reference_active":true,"delta_db":-3.5154111405754374}],"active_band_count":3,"active_band_mae_db":1.7417917973158097}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":1.084759001279324,"rms_envelope_nrmse":0.11387401522859036,"envelope_correlation":0.9934653695584856,"rms_gain_error_db":-0.5603572413850167,"reference_mean_power":0.00010850799286328448,"output_mean_power":9.537312347040382e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.0153781665834753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.269855261867839e-05,"output_power":5.3958370428353146e-05,"reference_active":true,"delta_db":-0.6519869008831748},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.20737764206477e-06,"output_power":3.690641898699118e-06,"reference_active":true,"delta_db":-0.5690958694049156},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2468166736228132e-07,"output_power":1.0030355675832053e-07,"reference_active":true,"delta_db":-3.5025130476402073}],"active_band_count":3,"active_band_mae_db":1.5745319393094326}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":1.7571572695325854,"rms_envelope_nrmse":0.20399077810118302,"envelope_correlation":0.9693320947153304,"rms_gain_error_db":-0.41567067502398414,"reference_mean_power":5.352135557413052e-06,"output_mean_power":4.863624495292296e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.7121423089901894e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7985008947946144e-06,"output_power":1.5992781504369303e-06,"reference_active":true,"delta_db":-0.5098665433299733},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2650926706890826e-07,"output_power":2.791947915210109e-07,"reference_active":true,"delta_db":-1.8402115972249793},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.085430385821354e-07,"output_power":7.67084675808201e-08,"reference_active":true,"delta_db":-1.5075866831481257}],"active_band_count":3,"active_band_mae_db":1.2858882745676927}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":1.2760365151624935,"rms_envelope_nrmse":0.11991874331936338,"envelope_correlation":0.9899552711545219,"rms_gain_error_db":-0.41567335978926356,"reference_mean_power":5.410729753830904e-06,"output_mean_power":4.916867522825582e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":2.394714459074953e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.814627629314413e-06,"output_power":1.5562551248754002e-06,"reference_active":true,"delta_db":-0.6670672450069917},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0097168111890158e-07,"output_power":1.352257193469396e-07,"reference_active":true,"delta_db":-1.7207556498556902},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.411981302222172e-09,"output_power":4.292270934393848e-09,"reference_active":true,"delta_db":-3.409939282789267}],"active_band_count":3,"active_band_mae_db":1.9325873925506496}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.3874621391296387,"predicted_mos":3.3874621391296387,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44400,"input_sample_rate":16000,"input_seconds":2.775,"input_float32_sha256":"f17cde6d4e2872a6ded7fd7cbdc6bbd8311ea797de8ec5efe03d0f45fc4f7724","internal_samples":44400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4690705740013073},"started_at":"2026-10-13T04:43:12.797219+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.06419998803176e-05,"native_input_samples":44400,"native_input_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","output_sha256":"656b1669af35679acbff1e2ceff3f2d90e6469a2fbccd40de9ae6d4cb9447dce","output_samples":44400,"length_delta_samples":0,"analysis_s":0.8315682289994584,"synthesis_s":0.04137083300065569,"total_s":0.8729390620001141,"rtf":0.31457263495499604,"output_peak":0.5308888280545736,"output_rms":0.05716434140755608,"finished_at":"2026-10-13T04:43:13.674987+00:00","evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"656b1669af35679acbff1e2ceff3f2d90e6469a2fbccd40de9ae6d4cb9447dce","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.515129971226217,"spectral_convergence":0.7012848971555037,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.48882888434245,"spectral_convergence":0.24866968926304073,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4862990053073055,"spectral_convergence":0.228735902397188,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.002163392881756805,"delta_db":0.6366139590913066,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0006392077862880949,"delta_db":0.8502103642233987,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.00027226119380066756,"delta_db":0.3053732104635761,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.00017623980035470045,"delta_db":0.7854095607001371,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":1.0725775410821052e-05,"delta_db":0.516687855238334,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":5.9326109329425715e-06,"delta_db":-0.1002034767411175,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.532416404409645,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6547792997361906,"pitch":null,"output_rms_dbfs":-24.857495913315102,"output_peak":0.5308888280545736,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.830085953625324},"common_prefix":{"samples":44400,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.515129971226217,"spectral_convergence":0.7012848971555037,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.48882888434245,"spectral_convergence":0.24866968926304073,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.4862990053073055,"spectral_convergence":0.228735902397188,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.002163392881756805,"delta_db":0.6366139590913066,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0006392077862880949,"delta_db":0.8502103642233987,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.00027226119380066756,"delta_db":0.3053732104635761,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.00017623980035470045,"delta_db":0.7854095607001371,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":1.0725775410821052e-05,"delta_db":0.516687855238334,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":5.9326109329425715e-06,"delta_db":-0.1002034767411175,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.532416404409645,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.6547792997361906,"pitch":null,"output_rms_dbfs":-24.857495913315102,"output_peak":0.5308888280545736,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.830085953625324},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44400,"output_samples":44400,"intelligibility":{"estoi":{"value":0.9389630332595061,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05301347090551401,"rms_dbfs":-25.512275213051296,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.796429980110695,"mean":-1.35583276147241e-05,"zero_energy":false},"output":{"rms":0.05716434140755608,"rms_dbfs":-24.857495913315102,"peak":0.5308888280545736,"peak_dbfs":-5.4999282758215,"crest_factor_db":19.357567637493606,"mean":7.440727657869063e-09,"zero_energy":false},"rms_difference_db":0.6547792997361945},"full_scale":{"reference":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":1.8473528210305376e-06,"output_rms_dbfs":-57.33450151883308,"output_zero_energy":false,"output_peak":0.027950747824275864,"output_energy_fraction":0.0001752003232958115,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":77,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9848484848484849,"absolute_difference_cents_median":9.436815328704972,"absolute_difference_cents_p95":33.10797571543456,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":0.00014622417526559514,"log_rms_mae_db":2.0717547883091045,"rms_envelope_nrmse":0.1516567710195404,"envelope_correlation":0.9867139342258037,"rms_gain_error_db":0.6547803559776912,"reference_mean_power":0.0028155003489198547,"output_mean_power":0.0032736603708197116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.019039396656961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008418483111337408,"output_power":0.000987061317542304,"reference_active":true,"delta_db":0.691102874731887},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001428627689371398,"output_power":0.00014289175575935118,"reference_active":true,"delta_db":0.0008810930984116879},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9192224262112578e-05,"output_power":1.4082229831729416e-05,"reference_active":true,"delta_db":-1.3445388171608343}],"active_band_count":3,"active_band_mae_db":0.6788409283303777}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":0.00013893685565863478,"log_rms_mae_db":1.6991262765126556,"rms_envelope_nrmse":0.10543413088244412,"envelope_correlation":0.9958380556748793,"rms_gain_error_db":0.6547833469558996,"reference_mean_power":0.0028463248833832306,"output_mean_power":0.003309503191504895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.8422021834325858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007582285194051116,"output_power":0.0008877528510592993,"reference_active":true,"delta_db":0.6849196012746183},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661169958011629e-05,"output_power":6.882943392379125e-05,"reference_active":true,"delta_db":0.14223683077334712},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.710179663683521e-06,"output_power":1.0818125910565668e-06,"reference_active":true,"delta_db":-1.9888970596379743}],"active_band_count":3,"active_band_mae_db":0.9386844972286466}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":1.7089222448875552,"rms_envelope_nrmse":0.21709973708610775,"envelope_correlation":0.9766740875718447,"rms_gain_error_db":0.8438144956782996,"reference_mean_power":0.0005201241069420424,"output_mean_power":0.0006316673547711799},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":5.057137989146479e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002701505367125327,"output_power":0.0003519348874268789,"reference_active":true,"delta_db":1.1485648603501972},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.404522229736366e-05,"output_power":5.8003590784270314e-05,"reference_active":true,"delta_db":0.30697572423650077},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.56355116704778e-06,"output_power":5.488521584507366e-06,"reference_active":true,"delta_db":-1.3927037274442848}],"active_band_count":3,"active_band_mae_db":0.9494147706769942}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":1.608889562285275,"rms_envelope_nrmse":0.15656784875983087,"envelope_correlation":0.9904984175151305,"rms_gain_error_db":0.8438141547734674,"reference_mean_power":0.0005258188868038473,"output_mean_power":0.0006385833588774184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":4.6537380366659205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025071266719842557,"output_power":0.0003242723540289426,"reference_active":true,"delta_db":1.1173364716011598},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3806665370442115e-05,"output_power":2.7624699059981385e-05,"reference_active":true,"delta_db":0.6459898807848659},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.079732275271239e-06,"output_power":6.472828593404286e-07,"reference_active":true,"delta_db":-2.222219768192883}],"active_band_count":3,"active_band_mae_db":1.328515373526303}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":1.5234721972160365,"rms_envelope_nrmse":0.2127448739554232,"envelope_correlation":0.9765633849797192,"rms_gain_error_db":0.6687917422753178,"reference_mean_power":0.00010733282194912887,"output_mean_power":0.00012520213134255119},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":1.1016693525679888e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.607161402154181e-05,"output_power":7.798308465336317e-05,"reference_active":true,"delta_db":0.7198549384840294},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0221078826476041e-05,"output_power":9.75334409234635e-06,"reference_active":true,"delta_db":-0.20343191499368582},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3761448978787783e-06,"output_power":2.124072288422522e-06,"reference_active":true,"delta_db":-2.0125178448894103}],"active_band_count":3,"active_band_mae_db":0.9786015661223751}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":1.0566909476997188,"rms_envelope_nrmse":0.12226018478084553,"envelope_correlation":0.9945539195453178,"rms_gain_error_db":0.6687917402809357,"reference_mean_power":0.00010850799286328448,"output_mean_power":0.00012657295057783637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.0153781665834753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.269855261867839e-05,"output_power":7.26192407032605e-05,"reference_active":true,"delta_db":0.6379418831696553},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.20737764206477e-06,"output_power":4.098518362547247e-06,"reference_active":true,"delta_db":-0.11384609592972003},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2468166736228132e-07,"output_power":1.0874566130942685e-07,"reference_active":true,"delta_db":-3.1515569944859854}],"active_band_count":3,"active_band_mae_db":1.30111499119512}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":1.770383596715788,"rms_envelope_nrmse":0.2578675077353954,"envelope_correlation":0.9519602799341802,"rms_gain_error_db":0.2470724931873922,"reference_mean_power":5.352135557413052e-06,"output_mean_power":5.6654492543583205e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.7121423089901894e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7985008947946144e-06,"output_power":1.911074367973929e-06,"reference_active":true,"delta_db":0.26366929390274363},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2650926706890826e-07,"output_power":4.318187618370999e-07,"reference_active":true,"delta_db":0.05373035981496621},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.085430385821354e-07,"output_power":1.0921778262997854e-07,"reference_active":true,"delta_db":0.026913799541380994}],"active_band_count":3,"active_band_mae_db":0.11477115108636361}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":1.059595561460745,"rms_envelope_nrmse":0.14600664641860311,"envelope_correlation":0.9837597427919628,"rms_gain_error_db":0.24707162799993052,"reference_mean_power":5.410729753830904e-06,"output_mean_power":5.7274724105379585e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":2.394714459074953e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.814627629314413e-06,"output_power":1.9158002486752784e-06,"reference_active":true,"delta_db":0.23562706185402738},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0097168111890158e-07,"output_power":2.0155641259823297e-07,"reference_active":true,"delta_db":0.012617545828816105},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.411981302222172e-09,"output_power":7.031890660820469e-09,"reference_active":true,"delta_db":-1.2660894648431786}],"active_band_count":3,"active_band_mae_db":0.5047780241753407}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"facodec","status":"measured","mos":{"status":"measured","value":3.631312608718872,"predicted_mos":3.631312608718872,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44400,"input_sample_rate":16000,"input_seconds":2.775,"input_float32_sha256":"5efd4f7fff934d6ec4908ba7b0a0e1df787b6b9aaf528c161b603546e1a83195","internal_samples":44400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2255878760006453},"started_at":"2026-10-13T06:58:31.383565+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.9380000341916457e-05,"native_input_samples":44400,"native_input_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","output_sha256":"56d1ebc2fdb40f84ba29853315098becf09af710bfe72e675570ec9f147d0bdb","output_samples":44400,"length_delta_samples":0,"analysis_s":0.8469542540005932,"synthesis_s":4.5930407289997675,"total_s":5.439994983000361,"rtf":1.9603585524325624,"output_peak":0.42527705430984497,"output_rms":0.05091517160516468,"finished_at":"2026-10-13T06:58:36.840745+00:00","evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"56d1ebc2fdb40f84ba29853315098becf09af710bfe72e675570ec9f147d0bdb","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.338818053519247,"spectral_convergence":0.19623118262792266,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.78106464291088,"spectral_convergence":0.1963354110084014,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.677135324425665,"spectral_convergence":0.21193615140227168,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.0015999918336268106,"delta_db":-0.6735623448085883,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0005530060631062773,"delta_db":0.22108872770989718,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.0002893332518017162,"delta_db":0.569499246584985,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.0001331236814796403,"delta_db":-0.4330771675740946,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":1.1126088557145115e-05,"delta_db":0.6758259892965122,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":5.77378101166122e-06,"delta_db":-0.21805907534506136,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4651854252198564,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3507805497905014,"pitch":null,"output_rms_dbfs":-25.863055762841796,"output_peak":0.42527705430984497,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.599006006951931},"common_prefix":{"samples":44400,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.338818053519247,"spectral_convergence":0.19623118262792266,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.78106464291088,"spectral_convergence":0.1963354110084014,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.677135324425665,"spectral_convergence":0.21193615140227168,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.0015999918336268106,"delta_db":-0.6735623448085883,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0005530060631062773,"delta_db":0.22108872770989718,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.0002893332518017162,"delta_db":0.569499246584985,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.0001331236814796403,"delta_db":-0.4330771675740946,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":1.1126088557145115e-05,"delta_db":0.6758259892965122,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":5.77378101166122e-06,"delta_db":-0.21805907534506136,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4651854252198564,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3507805497905014,"pitch":null,"output_rms_dbfs":-25.863055762841796,"output_peak":0.42527705430984497,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.599006006951931},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44400,"output_samples":44400,"intelligibility":{"estoi":{"value":0.9132276208251598,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05301347090551401,"rms_dbfs":-25.512275213051296,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.796429980110695,"mean":-1.35583276147241e-05,"zero_energy":false},"output":{"rms":0.05091517160516468,"rms_dbfs":-25.863055762841796,"peak":0.42527705430984497,"peak_dbfs":-7.42656097752989,"crest_factor_db":18.436494785311904,"mean":6.793537793228794e-05,"zero_energy":false},"rms_difference_db":-0.35078054979049966},"full_scale":{"reference":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":2.5701342438684453e-07,"output_rms_dbfs":-65.90044191902689,"output_zero_energy":false,"output_peak":0.002950668567791581,"output_energy_fraction":3.072535066677627e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":62,"both_estimated_voiced_frames":60,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9090909090909091,"absolute_difference_cents_median":2.9463440563521854,"absolute_difference_cents_p95":16.859876009015252,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":0.00014622417526559514,"log_rms_mae_db":1.6599252551149157,"rms_envelope_nrmse":0.075277343355631,"envelope_correlation":0.9964106703373659,"rms_gain_error_db":-0.35078047030015025,"reference_mean_power":0.0028155003489198547,"output_mean_power":0.002597033420430111},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.019039396656961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008418483111337408,"output_power":0.00071994682213493,"reference_active":true,"delta_db":-0.6793342591525443},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001428627689371398,"output_power":0.00011956823027795915,"reference_active":true,"delta_db":-0.7730326195133526},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9192224262112578e-05,"output_power":2.0201175002898906e-05,"reference_active":true,"delta_db":0.22251321210825076}],"active_band_count":3,"active_band_mae_db":0.5582933635913826}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":0.00013893685565863478,"log_rms_mae_db":1.574278337471598,"rms_envelope_nrmse":0.06288179023776064,"envelope_correlation":0.9977977606462444,"rms_gain_error_db":-0.35078185570205805,"reference_mean_power":0.0028463248833832306,"output_mean_power":0.0026254653075542347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.8422021834325858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007582285194051116,"output_power":0.0006479033061439222,"reference_active":true,"delta_db":-0.6828991952099142},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661169958011629e-05,"output_power":5.6912579275900895e-05,"reference_active":true,"delta_db":-0.6834224644496516},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.710179663683521e-06,"output_power":1.6059128458620005e-06,"reference_active":true,"delta_db":-0.27319765705218385}],"active_band_count":3,"active_band_mae_db":0.5465064389039166}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":1.1062456577396809,"rms_envelope_nrmse":0.123169809735706,"envelope_correlation":0.9914983798936869,"rms_gain_error_db":0.4156306195718498,"reference_mean_power":0.0005201241069420424,"output_mean_power":0.000572361014874447},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":5.057137989146479e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002701505367125327,"output_power":0.0002846255405070078,"reference_active":true,"delta_db":0.2266803365785226},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.404522229736366e-05,"output_power":5.419217687308293e-05,"reference_active":true,"delta_db":0.01179289248078653},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.56355116704778e-06,"output_power":9.416552466819223e-06,"reference_active":true,"delta_db":0.9516618130934655}],"active_band_count":3,"active_band_mae_db":0.3967116807175916}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":0.8571936021235448,"rms_envelope_nrmse":0.09717113741045577,"envelope_correlation":0.9950287810255042,"rms_gain_error_db":0.41563061843315063,"reference_mean_power":0.0005258188868038473,"output_mean_power":0.0005786277305655207},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":4.6537380366659205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025071266719842557,"output_power":0.00026608729812203295,"reference_active":true,"delta_db":0.2584786653246427},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3806665370442115e-05,"output_power":2.2020086643058457e-05,"reference_active":true,"delta_db":-0.338795440207586},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.079732275271239e-06,"output_power":1.1597262781737505e-06,"reference_active":true,"delta_db":0.3103941446567281}],"active_band_count":3,"active_band_mae_db":0.30255608339631895}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":1.0553221084652404,"rms_envelope_nrmse":0.12757789018176965,"envelope_correlation":0.9893731700863027,"rms_gain_error_db":-0.0462571920080574,"reference_mean_power":0.00010733282194912887,"output_mean_power":0.00010619567472476384},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":1.1016693525679888e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.607161402154181e-05,"output_power":6.924224692717823e-05,"reference_active":true,"delta_db":0.2035623634770235},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0221078826476041e-05,"output_power":1.0723381688349723e-05,"reference_active":true,"delta_db":0.2083502690979917},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3761448978787783e-06,"output_power":4.072919328225202e-06,"reference_active":true,"delta_db":0.8148473069030872}],"active_band_count":3,"active_band_mae_db":0.4089199798260341}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":0.7223360507314784,"rms_envelope_nrmse":0.07249793980553648,"envelope_correlation":0.9963951807614121,"rms_gain_error_db":-0.04625719138462758,"reference_mean_power":0.00010850799286328448,"output_mean_power":0.00010735839520050945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.0153781665834753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.269855261867839e-05,"output_power":6.628041829120085e-05,"reference_active":true,"delta_db":0.24127725198216873},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.20737764206477e-06,"output_power":4.4776834583387855e-06,"reference_active":true,"delta_db":0.27041893887447366},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2468166736228132e-07,"output_power":3.054011669178254e-07,"reference_active":true,"delta_db":1.3330305401960543}],"active_band_count":3,"active_band_mae_db":0.6149089103508989}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":1.7809984675150186,"rms_envelope_nrmse":0.23562252930138244,"envelope_correlation":0.959254266925161,"rms_gain_error_db":0.04348322850472226,"reference_mean_power":5.352135557413052e-06,"output_mean_power":5.405992358038413e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.7121423089901894e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7985008947946144e-06,"output_power":1.579634440384347e-06,"reference_active":true,"delta_db":-0.5635406424147598},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2650926706890826e-07,"output_power":3.221931923074288e-07,"reference_active":true,"delta_db":-1.2181211195568244},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.085430385821354e-07,"output_power":1.0998438790111696e-07,"reference_active":true,"delta_db":0.0572906699525657}],"active_band_count":3,"active_band_mae_db":0.6129841439747167}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":1.3952884420256626,"rms_envelope_nrmse":0.17594205397831864,"envelope_correlation":0.9751879072300969,"rms_gain_error_db":0.04348342701635317,"reference_mean_power":5.410729753830904e-06,"output_mean_power":5.465176418618913e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":2.394714459074953e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.814627629314413e-06,"output_power":1.5747180196156699e-06,"reference_active":true,"delta_db":-0.6158472192974819},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0097168111890158e-07,"output_power":1.3926946063204564e-07,"reference_active":true,"delta_db":-1.5927897173402503},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.411981302222172e-09,"output_power":7.981000573821443e-09,"reference_active":true,"delta_db":-0.7162371370482566}],"active_band_count":3,"active_band_mae_db":0.9749580245619963}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5988516807556152,"predicted_mos":3.5988516807556152,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":122880,"input_sample_rate":44100,"input_seconds":2.786394557823129,"input_float32_sha256":"658f2e1b7c08b16c7b9f62547aece4fbcb7eebdaaa819c963c87f4e9b1b805bb","internal_samples":44583,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.28665593900223},"started_at":"2026-10-13T07:27:51.808210+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00383895100094378,"native_input_samples":122378,"native_input_sha256":"b67961acfe7bbd3bd8288fa0ea0aa9d4f056f17b53f40e3c57fad2c2ba7498bc","output_sha256":"73af22088dc09bcb5754c72666450f1976d1d82116cf96847e0702decf46317f","output_samples":122880,"length_delta_samples":502,"analysis_s":0.07940636800049106,"synthesis_s":14.958988197999133,"total_s":15.038394565999624,"rtf":5.419219143641695,"output_peak":0.3643970787525177,"output_rms":0.04721706524729331,"finished_at":"2026-10-13T07:28:06.861017+00:00","evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"73af22088dc09bcb5754c72666450f1976d1d82116cf96847e0702decf46317f","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":44400,"output_samples":44583,"length_delta_samples":183,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.51498972881882,"output_peak":0.3649592053145772},"common_prefix":{"samples":44400,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9958952964134311,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.886948851710798,"spectral_convergence":0.4321320585350886,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.037811546977373,"spectral_convergence":0.3540082650498604,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.182738423461739,"spectral_convergence":0.3722926673032544,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.0012206918596346812,"delta_db":-1.8486795207414457,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0005760441071752886,"delta_db":0.3983471810778757,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.0003576292917330773,"delta_db":1.489846591809696,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":7.3963772675759e-05,"delta_db":-2.9854398095841588,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":9.440267907160304e-06,"delta_db":-0.03775594019750827,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":2.433513496935006e-06,"delta_db":-3.9703245613961964,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.7883989341344801,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9848513937912623,"pitch":null,"output_rms_dbfs":-26.497126606842556,"output_peak":0.3649592053145772,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.702499607383304},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44400,"output_samples":44400,"intelligibility":{"estoi":{"value":0.7954887773792226,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05301347090551401,"rms_dbfs":-25.512275213051296,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.796429980110695,"mean":-1.35583276147241e-05,"zero_energy":false},"output":{"rms":0.04733078087852512,"rms_dbfs":-26.497126606842556,"peak":0.3649592053145772,"peak_dbfs":-8.75511355453736,"crest_factor_db":17.742013052305197,"mean":-0.00011866973798233977,"zero_energy":false},"rms_difference_db":-0.98485139379126},"full_scale":{"reference":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":7.09647153291968e-08,"output_rms_dbfs":-71.48957535039162,"output_zero_energy":false,"output_peak":0.004788478431486057,"output_energy_fraction":9.81726670109408e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":70,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9848484848484849,"absolute_difference_cents_median":9.40759258237279,"absolute_difference_cents_p95":49.64441296943832,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":0.00014622417526559514,"log_rms_mae_db":1.890257497895635,"rms_envelope_nrmse":0.15466330820084234,"envelope_correlation":0.9884461458203825,"rms_gain_error_db":-0.984850433874402,"reference_mean_power":0.0028155003489198547,"output_mean_power":0.002244246425359828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.019039396656961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008418483111337408,"output_power":0.0006626694764656083,"reference_active":true,"delta_db":-1.0393687816989352},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001428627689371398,"output_power":0.00012024353083120863,"reference_active":true,"delta_db":-0.7485734288757405},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9192224262112578e-05,"output_power":1.2746842933080407e-05,"reference_active":true,"delta_db":-1.7772267528783683}],"active_band_count":3,"active_band_mae_db":1.188389654484348}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":0.00013893685565863478,"log_rms_mae_db":1.8249464203904628,"rms_envelope_nrmse":0.14166684147214098,"envelope_correlation":0.9915921860433867,"rms_gain_error_db":-0.9848480428498428,"reference_mean_power":0.0028463248833832306,"output_mean_power":0.0022688180323807774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.8422021834325858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007582285194051116,"output_power":0.0005941617361104413,"reference_active":true,"delta_db":-1.0589543570939357},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661169958011629e-05,"output_power":5.845603618139292e-05,"reference_active":true,"delta_db":-0.5672115156446448},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.710179663683521e-06,"output_power":1.2157562377110094e-06,"reference_active":true,"delta_db":-1.4819523129691095}],"active_band_count":3,"active_band_mae_db":1.0360393952358966}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":1.9961771730147284,"rms_envelope_nrmse":0.24050277702362688,"envelope_correlation":0.9671896490077472,"rms_gain_error_db":0.6361453904112127,"reference_mean_power":0.0005201241069420424,"output_mean_power":0.0006021733373068786},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":5.057137989146479e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002701505367125327,"output_power":0.00033588079454953857,"reference_active":true,"delta_db":0.945793371729185},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.404522229736366e-05,"output_power":6.108349064939092e-05,"reference_active":true,"delta_db":0.5316653957820493},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.56355116704778e-06,"output_power":7.462060851826084e-06,"reference_active":true,"delta_db":-0.05866962907797841}],"active_band_count":3,"active_band_mae_db":0.5120427988630709}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":1.528560397828417,"rms_envelope_nrmse":0.19735526600793735,"envelope_correlation":0.9779301682067641,"rms_gain_error_db":0.636145314783042,"reference_mean_power":0.0005258188868038473,"output_mean_power":0.0006087664542860101},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":4.6537380366659205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025071266719842557,"output_power":0.0003102832545765925,"reference_active":true,"delta_db":0.9258206099369579},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3806665370442115e-05,"output_power":2.8922154753068475e-05,"reference_active":true,"delta_db":0.845320780650159},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.079732275271239e-06,"output_power":9.736954913809534e-07,"reference_active":true,"delta_db":-0.44892924417575225}],"active_band_count":3,"active_band_mae_db":0.7400235449209563}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":3.061421274923379,"rms_envelope_nrmse":0.40808413920802505,"envelope_correlation":0.9241198322176233,"rms_gain_error_db":-2.9057368090865703,"reference_mean_power":0.00010733282194912887,"output_mean_power":5.4974193517076386e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":1.1016693525679888e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.607161402154181e-05,"output_power":3.720799252698262e-05,"reference_active":true,"delta_db":-2.493786767644261},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0221078826476041e-05,"output_power":6.908889138501188e-06,"reference_active":true,"delta_db":-1.7008851368538087},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3761448978787783e-06,"output_power":1.3055382382067278e-06,"reference_active":true,"delta_db":-4.126314809722597}],"active_band_count":3,"active_band_mae_db":2.7736622380735554}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":2.730482866047614,"rms_envelope_nrmse":0.38078734791697544,"envelope_correlation":0.9402730542937918,"rms_gain_error_db":-2.905737079888783,"reference_mean_power":0.00010850799286328448,"output_mean_power":5.557609422297317e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.0153781665834753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.269855261867839e-05,"output_power":3.5418587037640875e-05,"reference_active":true,"delta_db":-2.48026283576998},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.20737764206477e-06,"output_power":3.296245703781133e-06,"reference_active":true,"delta_db":-1.0599191802818086},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2468166736228132e-07,"output_power":1.1601008469504634e-07,"reference_active":true,"delta_db":-2.8707189425811097}],"active_band_count":3,"active_band_mae_db":2.1369669862109664}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":4.619249398429356,"rms_envelope_nrmse":0.32017214888785156,"envelope_correlation":0.9204613639212849,"rms_gain_error_db":-0.49788706604098815,"reference_mean_power":5.352135557413052e-06,"output_mean_power":4.772417153111537e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.7121423089901894e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7985008947946144e-06,"output_power":1.8594012413348564e-06,"reference_active":true,"delta_db":0.14462458331841027},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2650926706890826e-07,"output_power":3.8043127253507196e-07,"reference_active":true,"delta_db":-0.4965226185262778},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.085430385821354e-07,"output_power":1.0662979687978027e-07,"reference_active":true,"delta_db":-0.0772339325076925}],"active_band_count":3,"active_band_mae_db":0.2394603781174602}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":4.3786896870965455,"rms_envelope_nrmse":0.2626240918189213,"envelope_correlation":0.9422135356147311,"rms_gain_error_db":-0.4978838334203742,"reference_mean_power":5.410729753830904e-06,"output_mean_power":4.824668290169634e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":2.394714459074953e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.814627629314413e-06,"output_power":1.8007713400007411e-06,"reference_active":true,"delta_db":-0.03328949033784359},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0097168111890158e-07,"output_power":2.0375276320200187e-07,"reference_active":true,"delta_db":0.059686417761520505},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.411981302222172e-09,"output_power":8.340496506507805e-09,"reference_active":true,"delta_db":-0.5248915094977782}],"active_band_count":3,"active_band_mae_db":0.20595580586571408}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"golf","status":"measured","mos":{"status":"measured","value":2.917147636413574,"predicted_mos":2.917147636413574,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66480,"input_sample_rate":24000,"input_seconds":2.77,"input_float32_sha256":"cb2c7f1eada48cf7524fe474baf8e8451ec9951a8228a102767a98fbe6ce8c0d","internal_samples":44320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0890727730002254},"started_at":"2026-10-13T05:02:56.327835+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001415149999957066,"native_input_samples":66600,"native_input_sha256":"a969ccef6bc8cbe61d3a5ba2768511295038e7d8c2fd8d36a249bf20e92aa0c5","output_sha256":"4ae7107b457c4685c96118d610a04288f0f6dc6774a3b04ba687da6995222d1f","output_samples":66480,"length_delta_samples":-120,"analysis_s":0.2579108890004136,"synthesis_s":0.07033203299943125,"total_s":0.32824292199984484,"rtf":0.11828573765760175,"output_peak":0.4138917922973633,"output_rms":0.04816982622922409,"finished_at":"2026-10-13T05:02:56.662848+00:00","analysis_stages":{"external_f0_s":0.043189750999772514},"evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"4ae7107b457c4685c96118d610a04288f0f6dc6774a3b04ba687da6995222d1f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44400,"output_samples":44320,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.342628791173347,"output_peak":0.413270005054339},"common_prefix":{"samples":44320,"sample_rate":16000,"reference_coverage":0.9981981981981982,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44320,"output_samples":44320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.286920777666679,"spectral_convergence":0.5820184355298317,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.617150875759257,"spectral_convergence":0.27278350312640787,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.483383320708909,"spectral_convergence":0.2599434843252273,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0018717837155662071,"output_power":0.0015077525569556272,"delta_db":-0.9392559096273077,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005262517430208731,"output_power":0.0005359441808105354,"delta_db":0.07926012497744343,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002544933630932735,"output_power":0.00017984408979479177,"delta_db":-1.5078029070218544,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014734886854492608,"output_power":8.92054391597738e-05,"delta_db":-2.1795546976138698,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.5398831151018e-06,"output_power":5.691893225010342e-06,"delta_db":-2.242863091581267,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.082043765271772e-06,"output_power":2.893105521420281e-06,"delta_db":-3.226852667857942,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.695931566446614,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8381836802598773,"pitch":null,"output_rms_dbfs":-26.342628791173347,"output_peak":0.413270005054339,"reference_rms_dbfs":-25.504445110913466,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.795818324711615},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44320,"output_samples":44320,"intelligibility":{"estoi":{"value":0.9056749588470481,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.053061282709106344,"rms_dbfs":-25.504445110913466,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.788599877972867,"mean":-1.202662498942351e-05,"zero_energy":false},"output":{"rms":0.048180195782421636,"rms_dbfs":-26.342628791173347,"peak":0.413270005054339,"peak_dbfs":-7.675322289338392,"crest_factor_db":18.667306501834954,"mean":9.166049298802425e-05,"zero_energy":false},"rms_difference_db":-0.8381836802598812},"full_scale":{"reference":{"sample_count":44320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":3.3566114012968766e-07,"output_rms_dbfs":-64.74098934605784,"output_zero_energy":false,"output_peak":0.003717646669927989,"output_energy_fraction":4.489341331073953e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":80,"both_estimated_voiced_frames":64,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9696969696969697,"absolute_difference_cents_median":5.754834269238549,"absolute_difference_cents_p95":20.7757952344355,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":553,"rms_floor":0.00014622417526559514,"log_rms_mae_db":3.569763293157878,"rms_envelope_nrmse":0.13115506389634843,"envelope_correlation":0.9919299444014076,"rms_gain_error_db":-0.8381833811849936,"reference_mean_power":0.002820590715249074,"output_mean_power":0.00232552886185004},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":553,"frequency_resolution_hz":0.3616636528028933,"power_floor":2.0198429021951397e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008395201628757108,"output_power":0.0006779978366755328,"reference_active":true,"delta_db":-0.928028229696356},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001430537256343165,"output_power":0.00012005559975691059,"reference_active":true,"delta_db":-0.7611675120535283},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9186069436097123e-05,"output_power":1.4169072669055572e-05,"reference_active":true,"delta_db":-1.3164458421503085}],"active_band_count":3,"active_band_mae_db":1.0018805279667309}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":547,"rms_floor":0.00013893685565863478,"log_rms_mae_db":3.3916580782087196,"rms_envelope_nrmse":0.11099563842889275,"envelope_correlation":0.996218732747015,"rms_gain_error_db":-0.8381859705596896,"reference_mean_power":0.0028515278971561406,"output_mean_power":0.002351034638099204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":547,"frequency_resolution_hz":0.3656307129798903,"power_floor":1.842587288648236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007559557512273005,"output_power":0.0006105078237516198,"reference_active":true,"delta_db":-0.9280514158088504},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.651387292318911e-05,"output_power":5.6966757383847394e-05,"reference_active":true,"delta_db":-0.6729073676775955},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7107329498036905e-06,"output_power":1.2786109512162057e-06,"reference_active":true,"delta_db":-1.2644380043259065}],"active_band_count":3,"active_band_mae_db":0.9551322626041174}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":553,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":1.5405603921532873,"rms_envelope_nrmse":0.20652534628940217,"envelope_correlation":0.9851104328152412,"rms_gain_error_db":-1.4846584979469548,"reference_mean_power":0.0005210646554952586,"output_mean_power":0.00037019092288045685},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":553,"frequency_resolution_hz":0.3616636528028933,"power_floor":5.064195066132361e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00026958609451077425,"output_power":0.000189044723004593,"reference_active":true,"delta_db":-1.5413292817819153},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.395265569330312e-05,"output_power":3.873943795263089e-05,"reference_active":true,"delta_db":-1.4385951116576625},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.557056391073109e-06,"output_power":4.658564214858663e-06,"reference_active":true,"delta_db":-2.101005766594772}],"active_band_count":3,"active_band_mae_db":1.6936433866781166}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":547,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":1.3080827947072713,"rms_envelope_nrmse":0.17663402810556728,"envelope_correlation":0.9942604299050458,"rms_gain_error_db":-1.484658488283824,"reference_mean_power":0.000526780162646489,"output_mean_power":0.00037425151167424456},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":547,"frequency_resolution_hz":0.3656307129798903,"power_floor":4.659728756431183e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024986724132439924,"output_power":0.00017616089537524543,"reference_active":true,"delta_db":-1.5179981294691403},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3737092092163315e-05,"output_power":1.6877786604952376e-05,"reference_active":true,"delta_db":-1.4811202315442653},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.082151663572804e-06,"output_power":6.437902576354697e-07,"reference_active":true,"delta_db":-2.2554373111723303}],"active_band_count":3,"active_band_mae_db":1.7515185573952454}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":553,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":2.0820879929598988,"rms_envelope_nrmse":0.26651440539285465,"envelope_correlation":0.9766982596383437,"rms_gain_error_db":-2.0041332651429125,"reference_mean_power":0.0001075269129383616,"output_mean_power":6.778035685596897e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":553,"frequency_resolution_hz":0.3616636528028933,"power_floor":1.1046696034350563e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.612221677550072e-05,"output_power":4.404583174655529e-05,"reference_active":true,"delta_db":-1.7644258958165415},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0226362802396073e-05,"output_power":7.285163922624376e-06,"reference_active":true,"delta_db":-1.4728186797686413},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3852176721061298e-06,"output_power":1.1584539802105918e-06,"reference_active":true,"delta_db":-4.657078133455192}],"active_band_count":3,"active_band_mae_db":2.6314409030134582}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":547,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":1.9151183118289172,"rms_envelope_nrmse":0.23656406886144885,"envelope_correlation":0.9883884729707199,"rms_gain_error_db":-2.0041334951557763,"reference_mean_power":0.00010870636114505681,"output_mean_power":6.852382682012023e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":547,"frequency_resolution_hz":0.3656307129798903,"power_floor":1.0180500984590584e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.271280550550892e-05,"output_power":4.1697613193729385e-05,"reference_active":true,"delta_db":-1.7724503340139584},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.184428026925556e-06,"output_power":3.072395717427419e-06,"reference_active":true,"delta_db":-1.3415895121716237},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2550934448513313e-07,"output_power":9.993739830842467e-08,"reference_active":true,"delta_db":-3.5343650337911625}],"active_band_count":3,"active_band_mae_db":2.216134959992248}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":553,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":2.7674279663831194,"rms_envelope_nrmse":0.317354480772444,"envelope_correlation":0.9371219570969063,"rms_gain_error_db":-1.1999545250644887,"reference_mean_power":5.361812686383591e-06,"output_mean_power":4.0673934548142955e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":553,"frequency_resolution_hz":0.3616636528028933,"power_floor":2.7124411518121564e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7973649586734093e-06,"output_power":1.4658177060078525e-06,"reference_active":true,"delta_db":-0.8855630710718176},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.280073234299806e-07,"output_power":4.5078683264808267e-07,"reference_active":true,"delta_db":0.22520021860865658},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0830313758303051e-07,"output_power":6.039567263541637e-08,"reference_active":true,"delta_db":-2.536352160526996}],"active_band_count":3,"active_band_mae_db":1.2157051500691567}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44320,"uncovered_tail_samples":0,"metrics":{"frame_count":547,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":2.6347612987564584,"rms_envelope_nrmse":0.28369092645410954,"envelope_correlation":0.950745411083825,"rms_gain_error_db":-1.1999543635722767,"reference_mean_power":5.420620177598792e-06,"output_mean_power":4.112004118912372e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":547,"frequency_resolution_hz":0.3656307129798903,"power_floor":2.3953546370031994e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.8170631058075638e-06,"output_power":1.5022765613387867e-06,"reference_active":true,"delta_db":-0.8262011903303833},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9829120285115512e-07,"output_power":2.0970984518173638e-07,"reference_active":true,"delta_db":0.2431537238313914},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.420366275832653e-09,"output_power":5.382164712193861e-09,"reference_active":true,"delta_db":-2.431108045524941}],"active_band_count":3,"active_band_mae_db":1.1668209865622385}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.5604138374328613,"predicted_mos":3.5604138374328613,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66720,"input_sample_rate":24000,"input_seconds":2.78,"input_float32_sha256":"707f590f5e4bd81d22ee2e7df7253ffa7b37f63de1470b4a973b5ebc93117fbb","internal_samples":44480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.699373277002451},"started_at":"2026-10-13T07:44:08.546532+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003165609999996377,"native_input_samples":66600,"native_input_sha256":"a969ccef6bc8cbe61d3a5ba2768511295038e7d8c2fd8d36a249bf20e92aa0c5","output_sha256":"339d6f9c6a8a330d4edb825e719105a00b39752d98cd2c21d6b08095d973df10","output_samples":66720,"length_delta_samples":120,"analysis_s":1.0902218100018217,"synthesis_s":6.680463065000367,"total_s":7.770684875002189,"rtf":2.8002468018025906,"output_peak":0.36877158284187317,"output_rms":0.057984781249031316,"finished_at":"2026-10-13T07:44:16.330750+00:00","analysis_stages":{"external_f0_s":0.7092116940002597},"evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"339d6f9c6a8a330d4edb825e719105a00b39752d98cd2c21d6b08095d973df10","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44400,"output_samples":44480,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.73252819232162,"output_peak":0.3683859217673024},"common_prefix":{"samples":44400,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982014388489209,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.406719608057882,"spectral_convergence":0.7175326709331918,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.350343223883843,"spectral_convergence":0.3213786385591402,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.193132641960003,"spectral_convergence":0.2950575901904712,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.002529341310823876,"delta_db":1.315334367556298,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0005008902567093557,"delta_db":-0.20878436225013852,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.000218019366234068,"delta_db":-0.6595335057004765,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.0001082144562119686,"delta_db":-1.3327775468719665,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":9.141353484165412e-06,"delta_db":-0.17749410552706374,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":3.610266873331878e-06,"delta_db":-2.257269090128122,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9918654963390109,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7875650603644104,"pitch":null,"output_rms_dbfs":-24.724710152686885,"output_peak":0.3683859217673024,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.650065157967242},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44400,"output_samples":44400,"intelligibility":{"estoi":{"value":0.9057888856926393,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05301347090551401,"rms_dbfs":-25.512275213051296,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.796429980110695,"mean":-1.35583276147241e-05,"zero_energy":false},"output":{"rms":0.058044956807766286,"rms_dbfs":-24.724710152686885,"peak":0.3683859217673024,"peak_dbfs":-8.673939503032667,"crest_factor_db":16.05077064965422,"mean":0.000359095196169884,"zero_energy":false},"rms_difference_db":0.7875650603644111},"full_scale":{"reference":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":2.0235577902767028e-06,"output_rms_dbfs":-56.938843881887024,"output_zero_energy":false,"output_peak":0.07777955948057774,"output_energy_fraction":0.00018613244871702901,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":83,"both_estimated_voiced_frames":64,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9696969696969697,"absolute_difference_cents_median":7.131615188675488,"absolute_difference_cents_p95":22.782832857137752,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":0.00014622417526559514,"log_rms_mae_db":2.4957867685167097,"rms_envelope_nrmse":0.19215203788331434,"envelope_correlation":0.9784062170503856,"rms_gain_error_db":0.7875660879886158,"reference_mean_power":0.0028155003489198547,"output_mean_power":0.0033752985613554856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.019039396656961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008418483111337408,"output_power":0.001021145644638995,"reference_active":true,"delta_db":0.8385384436975868},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001428627689371398,"output_power":0.00016570990021014923,"reference_active":true,"delta_db":0.6442939250578527},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9192224262112578e-05,"output_power":1.795244381082452e-05,"reference_active":true,"delta_db":-0.2900173363509008}],"active_band_count":3,"active_band_mae_db":0.5909499017021135}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":0.00013893685565863478,"log_rms_mae_db":2.202752101850894,"rms_envelope_nrmse":0.1453042123053263,"envelope_correlation":0.9900252567752496,"rms_gain_error_db":0.7875690642475404,"reference_mean_power":0.0028463248833832306,"output_mean_power":0.0034122541919127296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.8422021834325858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007582285194051116,"output_power":0.0009164149034576623,"reference_active":true,"delta_db":0.8229202780693257},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661169958011629e-05,"output_power":7.893694503256067e-05,"reference_active":true,"delta_db":0.7372979988742601},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.710179663683521e-06,"output_power":1.559340644698506e-06,"reference_active":true,"delta_db":-0.4010073873005823}],"active_band_count":3,"active_band_mae_db":0.6537418880813893}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":1.7188687588664864,"rms_envelope_nrmse":0.22181546300754054,"envelope_correlation":0.9655599186773436,"rms_gain_error_db":-0.42176069859149684,"reference_mean_power":0.0005201241069420424,"output_mean_power":0.000471987952340031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":5.057137989146479e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002701505367125327,"output_power":0.00024078139852946976,"reference_active":true,"delta_db":-0.4998290206067699},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.404522229736366e-05,"output_power":4.5420576319701376e-05,"reference_active":true,"delta_db":-0.7550466703929416},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.56355116704778e-06,"output_power":5.609404322582494e-06,"reference_active":true,"delta_db":-1.298090043940698}],"active_band_count":3,"active_band_mae_db":0.8509885783134697}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":1.3434881123426134,"rms_envelope_nrmse":0.13667612175079552,"envelope_correlation":0.9872009880795908,"rms_gain_error_db":-0.42176067331062095,"reference_mean_power":0.0005258188868038473,"output_mean_power":0.0004771556976816074},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":4.6537380366659205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025071266719842557,"output_power":0.00022275130118236093,"reference_active":true,"delta_db":-0.5135602752912507},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3806665370442115e-05,"output_power":1.873951142414652e-05,"reference_active":true,"delta_db":-1.039403037010662},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.079732275271239e-06,"output_power":8.070231522329293e-07,"reference_active":true,"delta_db":-1.264300893319807}],"active_band_count":3,"active_band_mae_db":0.9390880685405731}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":1.8704321175850243,"rms_envelope_nrmse":0.2745150328872019,"envelope_correlation":0.9542067822575517,"rms_gain_error_db":-1.067451968776524,"reference_mean_power":0.00010733282194912887,"output_mean_power":8.39435537113181e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":1.1016693525679888e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.607161402154181e-05,"output_power":5.088337114643253e-05,"reference_active":true,"delta_db":-1.134390393886871},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0221078826476041e-05,"output_power":7.094200756459587e-06,"reference_active":true,"delta_db":-1.5859326337704769},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3761448978787783e-06,"output_power":1.912911397414536e-06,"reference_active":true,"delta_db":-2.467262227068243}],"active_band_count":3,"active_band_mae_db":1.7291950849085305}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":1.3540280749990856,"rms_envelope_nrmse":0.1751347424190758,"envelope_correlation":0.9865833266024623,"rms_gain_error_db":-1.067452133224156,"reference_mean_power":0.00010850799286328448,"output_mean_power":8.486263583421053e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.0153781665834753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.269855261867839e-05,"output_power":4.764139745841335e-05,"reference_active":true,"delta_db":-1.1927302326403872},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.20737764206477e-06,"output_power":3.2097411169326748e-06,"reference_active":true,"delta_db":-1.1754148917794194},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2468166736228132e-07,"output_power":1.3784371842910568e-07,"reference_active":true,"delta_db":-2.1218065812437144}],"active_band_count":3,"active_band_mae_db":1.496650568554507}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":2.057300105964659,"rms_envelope_nrmse":0.32902737966569906,"envelope_correlation":0.9168073333980526,"rms_gain_error_db":-0.7705051687965171,"reference_mean_power":5.352135557413052e-06,"output_mean_power":4.482048871624814e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.7121423089901894e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7985008947946144e-06,"output_power":1.455596786440331e-06,"reference_active":true,"delta_db":-0.9186957011404996},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2650926706890826e-07,"output_power":2.9793942196795844e-07,"reference_active":true,"delta_db":-1.5580050097320244},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.085430385821354e-07,"output_power":1.1520133773275012e-07,"reference_active":true,"delta_db":0.2585554703320781}],"active_band_count":3,"active_band_mae_db":0.911752060401534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":1.343052244683558,"rms_envelope_nrmse":0.19349983438722398,"envelope_correlation":0.9728014160348241,"rms_gain_error_db":-0.7705071157837783,"reference_mean_power":5.410729753830904e-06,"output_mean_power":4.53111548757261e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":2.394714459074953e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.814627629314413e-06,"output_power":1.454216316279511e-06,"reference_active":true,"delta_db":-0.9615850599907215},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0097168111890158e-07,"output_power":1.4277042775916147e-07,"reference_active":true,"delta_db":-1.484966046550824},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.411981302222172e-09,"output_power":8.119819354150548e-09,"reference_active":true,"delta_db":-0.6413468836522689}],"active_band_count":3,"active_band_mae_db":1.029299330064605}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.6410486698150635,"predicted_mos":3.6410486698150635,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":66840,"input_sample_rate":24000,"input_seconds":2.785,"input_float32_sha256":"533de0b0665443a148661000dbfd2fca86e51e43284067da48ce608b14a01c5c","internal_samples":44560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.686882173000413},"started_at":"2026-10-13T05:11:17.210575+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0022214599994185846,"native_input_samples":66600,"native_input_sha256":"a969ccef6bc8cbe61d3a5ba2768511295038e7d8c2fd8d36a249bf20e92aa0c5","output_sha256":"5a09cf71da00a72819d26a5dc062e53e24098bb1cd978f2a7606313b1d8ad9fc","output_samples":66840,"length_delta_samples":240,"analysis_s":0.4975378520011873,"synthesis_s":0.063072580998778,"total_s":0.5606104329999653,"rtf":0.20202177765764515,"output_peak":0.4090871214866638,"output_rms":0.051814732569462583,"finished_at":"2026-10-13T05:11:17.779711+00:00","analysis_stages":{"external_f0_s":0.06095976300093753},"evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"5a09cf71da00a72819d26a5dc062e53e24098bb1cd978f2a7606313b1d8ad9fc","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":44400,"output_samples":44560,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.709402717221543,"output_peak":0.3998951489079042},"common_prefix":{"samples":44400,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9964093357271095,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.823115551462993,"spectral_convergence":0.15906344154995597,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.5365358249324945,"spectral_convergence":0.14790742869663145,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.6627894069262945,"spectral_convergence":0.1509896564681001,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.0018350824949670732,"delta_db":-0.07818408040044933,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.00048383809369806184,"delta_db":-0.35920961737224255,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.0002400317848807099,"delta_db":-0.24179668855045894,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.00012457468348283187,"delta_db":-0.7213324370537993,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":7.713322352515402e-06,"delta_db":-0.9151843179105884,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":4.1514060076378896e-06,"delta_db":-1.650710059728401,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6610695335026566,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1815054693614024,"pitch":null,"output_rms_dbfs":-25.6937806824127,"output_peak":0.3998951489079042,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.6741469277739265},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44400,"output_samples":44400,"intelligibility":{"estoi":{"value":0.9545539780057882,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05301347090551401,"rms_dbfs":-25.512275213051296,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.796429980110695,"mean":-1.35583276147241e-05,"zero_energy":false},"output":{"rms":0.05191716459576091,"rms_dbfs":-25.6937806824127,"peak":0.3998951489079042,"peak_dbfs":-7.961077284436699,"crest_factor_db":17.732703397976,"mean":0.00017446051830588682,"zero_energy":false},"rms_difference_db":-0.18150546936140444},"full_scale":{"reference":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":1.4356801420996902e-07,"output_rms_dbfs":-68.42942306607813,"output_zero_energy":false,"output_peak":0.005195704150348321,"output_energy_fraction":1.6507116844398123e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":64,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9696969696969697,"absolute_difference_cents_median":1.4047936947987805,"absolute_difference_cents_p95":8.068096375252573,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":0.00014622417526559514,"log_rms_mae_db":1.0486010708686069,"rms_envelope_nrmse":0.06409519692679955,"envelope_correlation":0.9968125729654981,"rms_gain_error_db":-0.1815060604296054,"reference_mean_power":0.0028155003489198547,"output_mean_power":0.0027002562468221624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.019039396656961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008418483111337408,"output_power":0.0008102767972962014,"reference_active":true,"delta_db":-0.16600442143263744},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001428627689371398,"output_power":0.00011399298975594481,"reference_active":true,"delta_db":-0.9804091899291989},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9192224262112578e-05,"output_power":9.394885530618573e-06,"reference_active":true,"delta_db":-3.102338168145251}],"active_band_count":3,"active_band_mae_db":1.4162505931690292}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":0.00013893685565863478,"log_rms_mae_db":0.7577449278932173,"rms_envelope_nrmse":0.03668456435180188,"envelope_correlation":0.9991555884215448,"rms_gain_error_db":-0.18150568900690614,"reference_mean_power":0.0028463248833832306,"output_mean_power":0.0027298193044460522},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.8422021834325858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007582285194051116,"output_power":0.0007298690034731781,"reference_active":true,"delta_db":-0.16555195448923712},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661169958011629e-05,"output_power":5.5814923888692057e-05,"reference_active":true,"delta_db":-0.7680017785099902},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.710179663683521e-06,"output_power":7.588793354265585e-07,"reference_active":true,"delta_db":-3.5286901080519617}],"active_band_count":3,"active_band_mae_db":1.4874146136837296}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":0.8925573188561721,"rms_envelope_nrmse":0.09703367903455078,"envelope_correlation":0.9939555259047669,"rms_gain_error_db":-0.28432404276247975,"reference_mean_power":0.0005201241069420424,"output_mean_power":0.0004871633170018143},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":5.057137989146479e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002701505367125327,"output_power":0.00026326412202076776,"reference_active":true,"delta_db":-0.11214157778002962},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.404522229736366e-05,"output_power":4.375194726750903e-05,"reference_active":true,"delta_db":-0.9175992058143536},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.56355116704778e-06,"output_power":4.2456374032909676e-06,"reference_active":true,"delta_db":-2.5078284853070536}],"active_band_count":3,"active_band_mae_db":1.179189756300479}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":0.5962900432133623,"rms_envelope_nrmse":0.058561600989158184,"envelope_correlation":0.9981690587850157,"rms_gain_error_db":-0.28432412093130566,"reference_mean_power":0.0005258188868038473,"output_mean_power":0.0004924972040480336},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":4.6537380366659205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025071266719842557,"output_power":0.00024306131616781934,"reference_active":true,"delta_db":-0.13460431877141107},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3806665370442115e-05,"output_power":2.170371260008192e-05,"reference_active":true,"delta_db":-0.4016453761367266},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.079732275271239e-06,"output_power":4.778074229886547e-07,"reference_active":true,"delta_db":-3.5406319100410726}],"active_band_count":3,"active_band_mae_db":1.35896053498307}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":1.070558883331384,"rms_envelope_nrmse":0.16055394837706588,"envelope_correlation":0.9862030083923488,"rms_gain_error_db":-0.7136005594818795,"reference_mean_power":0.00010733282194912887,"output_mean_power":9.106940347292586e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":1.1016693525679888e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.607161402154181e-05,"output_power":6.009308705057281e-05,"reference_active":true,"delta_db":-0.41190401251999076},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0221078826476041e-05,"output_power":6.8932371772365865e-06,"reference_active":true,"delta_db":-1.7107351602003558},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3761448978787783e-06,"output_power":1.7246317079295787e-06,"reference_active":true,"delta_db":-2.9172471098045665}],"active_band_count":3,"active_band_mae_db":1.6799620941749709}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":0.8434892956448671,"rms_envelope_nrmse":0.11847996272145239,"envelope_correlation":0.9941519413200817,"rms_gain_error_db":-0.7136006575519234,"reference_mean_power":0.00010850799286328448,"output_mean_power":9.206650658679966e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.0153781665834753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.269855261867839e-05,"output_power":5.554295226997624e-05,"reference_active":true,"delta_db":-0.5262855531394738},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.20737764206477e-06,"output_power":3.093585552215238e-06,"reference_active":true,"delta_db":-1.3354936397421693},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2468166736228132e-07,"output_power":1.2156049684246163e-07,"reference_active":true,"delta_db":-2.6677517165332967}],"active_band_count":3,"active_band_mae_db":1.5098436364716463}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":1.7195140632851569,"rms_envelope_nrmse":0.17725643899637272,"envelope_correlation":0.9782158261755048,"rms_gain_error_db":-0.5834475750019431,"reference_mean_power":5.352135557413052e-06,"output_mean_power":4.67931570330634e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.7121423089901894e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7985008947946144e-06,"output_power":1.5844851497240233e-06,"reference_active":true,"delta_db":-0.5502248502021992},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2650926706890826e-07,"output_power":3.3584952782251437e-07,"reference_active":true,"delta_db":-1.0378372977789225},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.085430385821354e-07,"output_power":5.7723023377262517e-08,"reference_active":true,"delta_db":-2.742529049684296}],"active_band_count":3,"active_band_mae_db":1.443530399221806}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":1.4018981045891417,"rms_envelope_nrmse":0.10936072123382828,"envelope_correlation":0.9932462991973925,"rms_gain_error_db":-0.5834460425484974,"reference_mean_power":5.410729753830904e-06,"output_mean_power":4.730545660840611e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":2.394714459074953e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.814627629314413e-06,"output_power":1.5752214986039741e-06,"reference_active":true,"delta_db":-0.6144588869376247},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0097168111890158e-07,"output_power":1.526852435714025e-07,"reference_active":true,"delta_db":-1.1933779914934601},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.411981302222172e-09,"output_power":5.123235217910855e-09,"reference_active":true,"delta_db":-2.641367600815163}],"active_band_count":3,"active_band_mae_db":1.4830681597487494}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7021-85628-0025","method":"magphase","status":"measured","mos":{"status":"measured","value":4.003424644470215,"predicted_mos":4.003424644470215,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":134350,"input_sample_rate":48000,"input_seconds":2.798958333333333,"input_float32_sha256":"b402bbb4bcc974e0c13dfe6bbf92903a352d85086104e266ebd275cd78923b6b","internal_samples":44784,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4864862699996593},"started_at":"2026-10-13T05:17:44.750515+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.004803277999599231,"native_input_samples":133200,"native_input_sha256":"b6d5b799bd357e58975ecd38f6b58122158048a19b5a002dd4a3114696ca5cfb","output_sha256":"b0aaa3baa83c401c6263c3c4a3b9c023512a4a499a2a35becda4d74d2bdee78d","output_samples":134350,"length_delta_samples":1150,"analysis_s":1.4793410920010501,"synthesis_s":0.7465517249984259,"total_s":2.225892816999476,"rtf":0.8021235376574689,"output_peak":0.4236408297951847,"output_rms":0.050650925332637126,"finished_at":"2026-10-13T05:17:46.987659+00:00","evaluation_input":{"reference":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"output_sha256":"b0aaa3baa83c401c6263c3c4a3b9c023512a4a499a2a35becda4d74d2bdee78d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":44400,"output_samples":44784,"length_delta_samples":384,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.90682523439998,"output_peak":0.4193113121105666},"common_prefix":{"samples":44400,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9914255091103966,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":44400,"output_samples":44400,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.720175655431164,"spectral_convergence":0.7046400196956322,"spectral_convergence_status":"ok","amplitude_floor":1.585777823535548e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.961600876537528,"spectral_convergence":0.37534567411647635,"spectral_convergence_status":"ok","amplitude_floor":7.694479165248034e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.201310215589927,"spectral_convergence":0.3362999879336746,"spectral_convergence_status":"ok","amplitude_floor":6.411814451459315e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00186841781908955,"output_power":0.0016638987171746472,"delta_db":-0.5034711358726174,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005255584463676,"output_power":0.0005147116335763384,"delta_db":-0.09057035784537856,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00025377475865110864,"output_power":0.00025543375180442737,"delta_db":0.02829858914112811,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00014708318852606447,"output_power":0.0001413738421926517,"delta_db":-0.17193974889360117,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":9.522695850195022e-06,"output_power":8.966974683403087e-06,"delta_db":-0.2611396980808372,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.071083630879648e-06,"output_power":4.168865004120215e-06,"delta_db":-1.632483833804042,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.44798389393960075,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3571528554595559,"pitch":null,"output_rms_dbfs":-25.86942806851085,"output_peak":0.4193113121105666,"reference_rms_dbfs":-25.512275213051296,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.961028915852872},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":44400,"output_samples":44400,"intelligibility":{"estoi":{"value":0.7035999026839811,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05301347090551401,"rms_dbfs":-25.512275213051296,"peak":0.411346435546875,"peak_dbfs":-7.715845232940603,"crest_factor_db":17.796429980110695,"mean":-1.35583276147241e-05,"zero_energy":false},"output":{"rms":0.05087783195824234,"rms_dbfs":-25.86942806851085,"peak":0.4193113121105666,"peak_dbfs":-7.549268422287377,"crest_factor_db":18.320159646223473,"mean":-0.0004621560930168486,"zero_energy":false},"rms_difference_db":-0.3571528554595531},"full_scale":{"reference":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":44400,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":277,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.1225224231602626e-06,"threshold_dbfs":-56.73147713002347,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":86,"quiet_samples":13760,"reference_mean_power":1.4844733604424914e-07,"output_mean_power":2.8582864414540646e-07,"output_rms_dbfs":-65.43894250804199,"output_zero_energy":false,"output_peak":0.008840235407662241,"output_energy_fraction":3.422031633181077e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":270,"reference_estimated_voiced_frames":66,"output_estimated_voiced_frames":72,"both_estimated_voiced_frames":61,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9242424242424242,"absolute_difference_cents_median":8.964014983097975,"absolute_difference_cents_p95":31.36502954049195,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":44400,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":0.00014622417526559514,"log_rms_mae_db":3.009102509768767,"rms_envelope_nrmse":0.27070759494769736,"envelope_correlation":0.9370271306410013,"rms_gain_error_db":-0.35715179039468553,"reference_mean_power":0.0028155003489198547,"output_mean_power":0.002593226234168236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.019039396656961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008418483111337408,"output_power":0.0007795002031838821,"reference_active":true,"delta_db":-0.33417612248981654},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001428627689371398,"output_power":0.0001291397408572609,"reference_active":true,"delta_db":-0.43859152698828346},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9192224262112578e-05,"output_power":1.5522359587228393e-05,"reference_active":true,"delta_db":-0.9216756976477879}],"active_band_count":3,"active_band_mae_db":0.5648144490419625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":0.00013893685565863478,"log_rms_mae_db":2.4433424177510754,"rms_envelope_nrmse":0.19695045044255313,"envelope_correlation":0.9651791893919661,"rms_gain_error_db":-0.3571489344509813,"reference_mean_power":0.0028463248833832306,"output_mean_power":0.0026216190011418973},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.8422021834325858e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007582285194051116,"output_power":0.0007009860972382269,"reference_active":true,"delta_db":-0.34090710980359074},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661169958011629e-05,"output_power":5.8667960187683885e-05,"reference_active":true,"delta_db":-0.551495261109351},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.710179663683521e-06,"output_power":1.6332720450159346e-06,"reference_active":true,"delta_db":-0.19983209016790823}],"active_band_count":3,"active_band_mae_db":0.3640781536936166}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":8.753945313428826e-05,"log_rms_mae_db":2.8954833954509627,"rms_envelope_nrmse":0.3711907703601819,"envelope_correlation":0.9026760072743503,"rms_gain_error_db":-0.15944523190405238,"reference_mean_power":0.0005201241069420424,"output_mean_power":0.0005013747517150914},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":5.057137989146479e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002701505367125327,"output_power":0.0002568426687020051,"reference_active":true,"delta_db":-0.21938660928143255},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.404522229736366e-05,"output_power":5.2443291849897737e-05,"reference_active":true,"delta_db":-0.13067363100719095},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.56355116704778e-06,"output_power":7.553706080758105e-06,"reference_active":true,"delta_db":-0.005656670629930749}],"active_band_count":3,"active_band_mae_db":0.11857230363951808}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":7.440703161891069e-05,"log_rms_mae_db":2.455690349383567,"rms_envelope_nrmse":0.26237703938535445,"envelope_correlation":0.94846132393553,"rms_gain_error_db":-0.15944525260760073,"reference_mean_power":0.0005258188868038473,"output_mean_power":0.0005068642445964954},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":4.6537380366659205e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00025071266719842557,"output_power":0.0002372748633541398,"reference_active":true,"delta_db":-0.23924545178721346},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3806665370442115e-05,"output_power":2.223792986440439e-05,"reference_active":true,"delta_db":-0.2960421130324701},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.079732275271239e-06,"output_power":1.0667645489009388e-06,"reference_active":true,"delta_db":-0.05247508833796325}],"active_band_count":3,"active_band_mae_db":0.19592088438588226}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":5.329471255433994e-05,"log_rms_mae_db":3.2965628065267216,"rms_envelope_nrmse":0.4121527123555383,"envelope_correlation":0.8878816760455197,"rms_gain_error_db":-0.16165239803220668,"reference_mean_power":0.00010733282194912887,"output_mean_power":0.00010341113576068529},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":1.1016693525679888e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.607161402154181e-05,"output_power":6.45357319464571e-05,"reference_active":true,"delta_db":-0.1021467597522879},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0221078826476041e-05,"output_power":1.0322802208636105e-05,"reference_active":true,"delta_db":0.043008684217669346},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3761448978787783e-06,"output_power":3.152096993447926e-06,"reference_active":true,"delta_db":-0.29821504731459153}],"active_band_count":3,"active_band_mae_db":0.14779016376151624}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":4.1357257962539876e-05,"log_rms_mae_db":2.627498205373857,"rms_envelope_nrmse":0.26800699892217855,"envelope_correlation":0.9499164924007882,"rms_gain_error_db":-0.16165236974323452,"reference_mean_power":0.00010850799286328448,"output_mean_power":0.00010454336940394135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":1.0153781665834753e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.269855261867839e-05,"output_power":6.084498034600109e-05,"reference_active":true,"delta_db":-0.13032760205434574},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.20737764206477e-06,"output_power":4.325359168295921e-06,"reference_active":true,"delta_db":0.12010681370353754},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2468166736228132e-07,"output_power":2.177929065645173e-07,"reference_active":true,"delta_db":-0.13523907277375896}],"active_band_count":3,"active_band_mae_db":0.12855782951054742}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":554,"rms_floor":1.1026487361641338e-05,"log_rms_mae_db":3.569601529506847,"rms_envelope_nrmse":0.42437273611569376,"envelope_correlation":0.8592937493862042,"rms_gain_error_db":-0.397463093656799,"reference_mean_power":5.352135557413052e-06,"output_mean_power":4.884057803338879e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":554,"frequency_resolution_hz":0.36101083032490977,"power_floor":2.7121423089901894e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7985008947946144e-06,"output_power":1.60039162320343e-06,"reference_active":true,"delta_db":-0.5068438870437239},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.2650926706890826e-07,"output_power":3.9916710269783494e-07,"reference_active":true,"delta_db":-0.2877373005999666},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.085430385821354e-07,"output_power":1.1536305987628202e-07,"reference_active":true,"delta_db":0.2646479153742503}],"active_band_count":3,"active_band_mae_db":0.35307636767264694}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":44400,"uncovered_tail_samples":0,"metrics":{"frame_count":548,"rms_floor":9.954651742580567e-06,"log_rms_mae_db":2.680695429356678,"rms_envelope_nrmse":0.28066304804033376,"envelope_correlation":0.9334860030222556,"rms_gain_error_db":-0.39746326179602715,"reference_mean_power":5.410729753830904e-06,"output_mean_power":4.937527379373976e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":548,"frequency_resolution_hz":0.36496350364963503,"power_floor":2.394714459074953e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.814627629314413e-06,"output_power":1.626915887571266e-06,"reference_active":true,"delta_db":-0.4742241885176674},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0097168111890158e-07,"output_power":1.7935428914102836e-07,"reference_active":true,"delta_db":-0.49423098373500757},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.411981302222172e-09,"output_power":8.48606869448751e-09,"reference_active":true,"delta_db":-0.44974512691938473}],"active_band_count":3,"active_band_mae_db":0.4727334330573532}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"hnm","status":"measured","mos":{"status":"measured","value":2.9365837574005127,"predicted_mos":2.9365837574005127,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"47bb6c3861b5fffbdf3ddec31a4b0a5a47881cef13a871cc819d3e7da381600b","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.596045125999808},"started_at":"2026-10-13T04:40:01.628143+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.3992999811307527e-05,"native_input_samples":96000,"native_input_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","output_sha256":"5d51670d1eebe2df61d1744632f60625a6490e880329f722d53d41c934c67487","output_samples":96000,"length_delta_samples":0,"analysis_s":1.066324741000244,"synthesis_s":0.5076899939995201,"total_s":1.5740147349997642,"rtf":0.26233578916662736,"output_peak":0.49564794536866247,"output_rms":0.04931605913450811,"finished_at":"2026-10-13T04:40:03.208732+00:00","evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"5d51670d1eebe2df61d1744632f60625a6490e880329f722d53d41c934c67487","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.232750546231531,"spectral_convergence":0.3414379501133128,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.81287539250454,"spectral_convergence":0.32834258671461863,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.988413039019585,"spectral_convergence":0.3521334322578276,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.0003482247049692358,"delta_db":0.1816932170237706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0008313781635616136,"delta_db":-0.09640646387906832,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0006266894727564931,"delta_db":-0.0010201719754167507,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0003748981331553372,"delta_db":0.062145869705164045,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.00013043512169384565,"delta_db":-0.13372921533425838,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":0.0001204480924217748,"delta_db":0.23390447688982757,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11814990246791761,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.005400078237815243,"pitch":null,"output_rms_dbfs":-26.140232706618733,"output_peak":0.49564794536866247,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.011346325918552},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.232750546231531,"spectral_convergence":0.3414379501133128,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.81287539250454,"spectral_convergence":0.32834258671461863,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.988413039019585,"spectral_convergence":0.3521334322578276,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.0003482247049692358,"delta_db":0.1816932170237706,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0008313781635616136,"delta_db":-0.09640646387906832,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0006266894727564931,"delta_db":-0.0010201719754167507,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0003748981331553372,"delta_db":0.062145869705164045,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.00013043512169384565,"delta_db":-0.13372921533425838,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":0.0001204480924217748,"delta_db":0.23390447688982757,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11814990246791761,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.005400078237815243,"pitch":null,"output_rms_dbfs":-26.140232706618733,"output_peak":0.49564794536866247,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.011346325918552},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.770949503186439,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04928540852502758,"rms_dbfs":-26.14563278485655,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.384597483135295,"mean":-0.0027082255681355792,"zero_energy":false},"output":{"rms":0.04931605913450811,"rms_dbfs":-26.14023270661874,"peak":0.49564794536866247,"peak_dbfs":-6.096533795640773,"crest_factor_db":20.04369891097796,"mean":-0.00021005761915996894,"zero_energy":false},"rms_difference_db":0.0054000782378089696},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":2.0987513112473254e-06,"output_rms_dbfs":-56.7803901951748,"output_zero_energy":false,"output_peak":0.005057495162057023,"output_energy_fraction":2.8764908469137656e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":60,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9491525423728814,"absolute_difference_cents_median":1.3253987591164238,"absolute_difference_cents_p95":6.645380443724738,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018099058448312497,"log_rms_mae_db":1.383831774075584,"rms_envelope_nrmse":0.09534115364976299,"envelope_correlation":0.9913944586854667,"rms_gain_error_db":0.005397999083878995,"reference_mean_power":0.002431072908198481,"output_mean_power":0.002434096452989853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.757963657265947e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033408940482734,"output_power":0.0003484494045143769,"reference_active":true,"delta_db":0.18277023843583573},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010057920506755965,"output_power":9.776168670802707e-05,"reference_active":true,"delta_db":-0.1233951287202441},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.918423836914529e-05,"output_power":1.9702011281856206e-05,"reference_active":true,"delta_db":0.11566001411695988}],"active_band_count":3,"active_band_mae_db":0.14060846042434658}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016835992502135956,"log_rms_mae_db":0.7427490014897594,"rms_envelope_nrmse":0.05178240159127817,"envelope_correlation":0.9972442316723726,"rms_gain_error_db":0.005508448019861467,"reference_mean_power":0.002443196853891242,"output_mean_power":0.0024462976903111885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.921742652983223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003035779346246488,"output_power":0.000318805085960118,"reference_active":true,"delta_db":0.2125503922102702},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.673255237859031e-05,"output_power":4.434475661225022e-05,"reference_active":true,"delta_db":-0.22777225811134916},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2224669143851135e-06,"output_power":1.1755063393472334e-06,"reference_active":true,"delta_db":-0.17012138376320163}],"active_band_count":3,"active_band_mae_db":0.2034813446949403}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015437420870825604,"log_rms_mae_db":1.6555776334095038,"rms_envelope_nrmse":0.13891218980446318,"envelope_correlation":0.9853587358306026,"rms_gain_error_db":0.03388839619232706,"reference_mean_power":0.0008784594297426283,"output_mean_power":0.0008853409427520734},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.115238295382225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019746522774839338,"output_power":0.00020330410313024956,"reference_active":true,"delta_db":0.1265551329678864},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173113277871683e-05,"output_power":5.03586624843829e-05,"reference_active":true,"delta_db":-0.11677801627272487},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0925495372248409e-05,"output_power":1.1111265537661908e-05,"reference_active":true,"delta_db":0.07322389102971677}],"active_band_count":3,"active_band_mae_db":0.10551901342344268}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012941193693393203,"log_rms_mae_db":1.1417646036356304,"rms_envelope_nrmse":0.08226031442697875,"envelope_correlation":0.9944994332594529,"rms_gain_error_db":0.03389171849602682,"reference_mean_power":0.000882875660341548,"output_mean_power":0.0008897924490823824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7308162041776265e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001828938795712491,"output_power":0.00019000657562798318,"reference_active":true,"delta_db":0.1656945867534263},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517691716707475e-05,"output_power":2.45666191821695e-05,"reference_active":true,"delta_db":-0.10657157359946798},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.42204663895331e-07,"output_power":7.623477616451187e-07,"reference_active":true,"delta_db":-0.9199212016755348}],"active_band_count":3,"active_band_mae_db":0.3973957873428097}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":1.7611545672672728,"rms_envelope_nrmse":0.1757875102539933,"envelope_correlation":0.9773765630095297,"rms_gain_error_db":-0.02745821571338666,"reference_mean_power":0.0004246628880558568,"output_mean_power":0.00042198643196548406},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.038477949048031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001401263951276247,"output_power":0.0001381495949062061,"reference_active":true,"delta_db":-0.061703337712213265},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6627820389923164e-05,"output_power":3.2726553287686125e-05,"reference_active":true,"delta_db":-0.4891080742878495},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.517673552823718e-06,"output_power":8.01398423403649e-06,"reference_active":true,"delta_db":-0.2647250763939906}],"active_band_count":3,"active_band_mae_db":0.27184549613135117}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":1.1681534768931658,"rms_envelope_nrmse":0.09887205669760539,"envelope_correlation":0.9922497058417548,"rms_gain_error_db":-0.027464785284826553,"reference_mean_power":0.00042679805438924104,"output_mean_power":0.000424107499773567},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.813635389636041e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013656501467579054,"output_power":0.00013454322150246092,"reference_active":true,"delta_db":-0.06477633502779917},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7671441991318864e-05,"output_power":1.5243532800542473e-05,"reference_active":true,"delta_db":-0.6418635982252303},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.379326657430826e-07,"output_power":4.473486630823726e-07,"reference_active":true,"delta_db":-0.8008177366245738}],"active_band_count":3,"active_band_mae_db":0.5024858899592011}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":2.061823100332807,"rms_envelope_nrmse":0.16284455783771906,"envelope_correlation":0.9833040027315413,"rms_gain_error_db":0.13887408431082793,"reference_mean_power":0.00015588925116025805,"output_mean_power":0.00016095466899545635},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103456000035227e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.437460419122184e-05,"output_power":5.8258220914116194e-05,"reference_active":true,"delta_db":0.29961109419993054},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.583240107798151e-05,"output_power":1.5409604875535138e-05,"reference_active":true,"delta_db":-0.11755280278518958},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.90382854367363e-06,"output_power":3.3737038874853725e-06,"reference_active":true,"delta_db":-0.6338377360600332}],"active_band_count":3,"active_band_mae_db":0.3503338776817178}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":1.3176821509464527,"rms_envelope_nrmse":0.09494507916847587,"envelope_correlation":0.994182171197942,"rms_gain_error_db":0.13886251160586258,"reference_mean_power":0.00015667279701186946,"output_mean_power":0.00016176324409393884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1374269590951092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4195005275221475e-05,"output_power":5.7913848846480965e-05,"reference_active":true,"delta_db":0.28823165397548756},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661491941713124e-06,"output_power":6.810016503952874e-06,"reference_active":true,"delta_db":0.09576657547490544},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7948760162946994e-07,"output_power":2.1727496573704775e-07,"reference_active":true,"delta_db":-1.0935285676971238}],"active_band_count":3,"active_band_mae_db":0.4925089323825056}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.7261340618133545,"predicted_mos":2.7261340618133545,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bd4bf05fec8e3e395b26d4c688fc6ac3c221690dea57c5a08a0830d4e0112a53","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6911773459996766},"started_at":"2026-10-13T04:43:13.675486+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.715399932640139e-05,"native_input_samples":96000,"native_input_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","output_sha256":"e79d53703742e4b778230c574f93bfd1b81305504c14f57089456c4f5bf9b687","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7215551480003342,"synthesis_s":0.09527429799982201,"total_s":1.8168294460001562,"rtf":0.3028049076666927,"output_peak":0.593167130357251,"output_rms":0.05616220245582691,"finished_at":"2026-10-13T04:43:15.500469+00:00","evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"e79d53703742e4b778230c574f93bfd1b81305504c14f57089456c4f5bf9b687","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.032570978961366,"spectral_convergence":0.634083315303526,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.47202198755321,"spectral_convergence":0.3349734172418003,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.666906470062424,"spectral_convergence":0.3612778198060473,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.0003632480477154485,"delta_db":0.3651303144825352,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.001144430281435912,"delta_db":1.2915008083403352,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0009208162889856383,"delta_db":1.6701857663777777,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0004566079525296144,"delta_db":0.9184478028490439,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.0001424641869874242,"delta_db":0.24938234755247612,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":0.0001266262270352527,"delta_db":0.45114188167776315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8242981535466553,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1345153996479,"pitch":null,"output_rms_dbfs":-25.011117385208653,"output_peak":0.593167130357251,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.057166478859},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.032570978961366,"spectral_convergence":0.634083315303526,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.47202198755321,"spectral_convergence":0.3349734172418003,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.666906470062424,"spectral_convergence":0.3612778198060473,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.0003632480477154485,"delta_db":0.3651303144825352,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.001144430281435912,"delta_db":1.2915008083403352,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0009208162889856383,"delta_db":1.6701857663777777,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0004566079525296144,"delta_db":0.9184478028490439,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.0001424641869874242,"delta_db":0.24938234755247612,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":0.0001266262270352527,"delta_db":0.45114188167776315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8242981535466553,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1345153996479,"pitch":null,"output_rms_dbfs":-25.011117385208653,"output_peak":0.593167130357251,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.057166478859},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.880783151106253,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04928540852502758,"rms_dbfs":-26.14563278485655,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.384597483135295,"mean":-0.0027082255681355792,"zero_energy":false},"output":{"rms":0.05616220245582691,"rms_dbfs":-25.011117385208653,"peak":0.593167130357251,"peak_dbfs":-4.536458457660016,"crest_factor_db":20.47465892754864,"mean":6.11471985329191e-09,"zero_energy":false},"rms_difference_db":1.1345153996478956},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":4.4038978145045274e-07,"output_rms_dbfs":-63.56162766701121,"output_zero_energy":false,"output_peak":0.0043619489421111305,"output_energy_fraction":4.654014339939889e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":65,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9491525423728814,"absolute_difference_cents_median":6.858586046295581,"absolute_difference_cents_p95":19.01252376725107,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018099058448312497,"log_rms_mae_db":4.621088130133147,"rms_envelope_nrmse":0.20610808518296697,"envelope_correlation":0.9822859411839143,"rms_gain_error_db":1.1345231818196027,"reference_mean_power":0.002431072908198481,"output_mean_power":0.003156823506382121},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.757963657265947e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033408940482734,"output_power":0.0004269504870851053,"reference_active":true,"delta_db":1.0651481053469651},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010057920506755965,"output_power":0.00011872629342934559,"reference_active":true,"delta_db":0.7203871074397614},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.918423836914529e-05,"output_power":1.726529364940638e-05,"reference_active":true,"delta_db":-0.4577059274871803}],"active_band_count":3,"active_band_mae_db":0.747747046757969}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016835992502135956,"log_rms_mae_db":3.997645379583143,"rms_envelope_nrmse":0.16766084439356677,"envelope_correlation":0.9936815045605981,"rms_gain_error_db":1.1346959176702671,"reference_mean_power":0.002443196853891242,"output_mean_power":0.0031726930134294605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.921742652983223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003035779346246488,"output_power":0.00038970266122315367,"reference_active":true,"delta_db":1.0846316959363398},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.673255237859031e-05,"output_power":5.705998266736221e-05,"reference_active":true,"delta_db":0.8671213398812533},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2224669143851135e-06,"output_power":1.0144117188782388e-06,"reference_active":true,"delta_db":-0.8102285622999262}],"active_band_count":3,"active_band_mae_db":0.9206605327058397}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015437420870825604,"log_rms_mae_db":2.019840132063495,"rms_envelope_nrmse":0.26534088036329384,"envelope_correlation":0.9811282191042672,"rms_gain_error_db":1.5892829539235678,"reference_mean_power":0.0008784594297426283,"output_mean_power":0.0012666306837826682},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.115238295382225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019746522774839338,"output_power":0.000282170296885339,"reference_active":true,"delta_db":1.5502066464928084},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173113277871683e-05,"output_power":6.906593895423087e-05,"reference_active":true,"delta_db":1.2551193199402022},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0925495372248409e-05,"output_power":9.46194249428995e-06,"reference_active":true,"delta_db":-0.6246083314843606}],"active_band_count":3,"active_band_mae_db":1.1433114326391238}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012941193693393203,"log_rms_mae_db":1.6374307017128236,"rms_envelope_nrmse":0.2343138315062027,"envelope_correlation":0.9902942711807217,"rms_gain_error_db":1.5892824547133495,"reference_mean_power":0.000882875660341548,"output_mean_power":0.0012729981999715688},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7308162041776265e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001828938795712491,"output_power":0.00026007711035323057,"reference_active":true,"delta_db":1.529029588581281},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517691716707475e-05,"output_power":3.352453898102537e-05,"reference_active":true,"delta_db":1.2436026322480074},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.42204663895331e-07,"output_power":6.131384038686188e-07,"reference_active":true,"delta_db":-1.8658673068217086}],"active_band_count":3,"active_band_mae_db":1.546166509216999}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":1.621590385848325,"rms_envelope_nrmse":0.19596383577392343,"envelope_correlation":0.9811546152292869,"rms_gain_error_db":0.8268981885928534,"reference_mean_power":0.0004246628880558568,"output_mean_power":0.000513729053829522},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.038477949048031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001401263951276247,"output_power":0.00015898968103636972,"reference_active":true,"delta_db":0.5484898851539193},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6627820389923164e-05,"output_power":4.02880910845383e-05,"reference_active":true,"delta_db":0.41365614403960704},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.517673552823718e-06,"output_power":7.229483192921728e-06,"reference_active":true,"delta_db":-0.7121373893231088}],"active_band_count":3,"active_band_mae_db":0.5580944728388784}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":1.0125027854794326,"rms_envelope_nrmse":0.13079746836428305,"envelope_correlation":0.9948834632065261,"rms_gain_error_db":0.8269008123815659,"reference_mean_power":0.00042679805438924104,"output_mean_power":0.0005163123486643242},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.813635389636041e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013656501467579054,"output_power":0.00015362942245112344,"reference_active":true,"delta_db":0.5113494216647936},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7671441991318864e-05,"output_power":1.9777127705458652e-05,"reference_active":true,"delta_db":0.4889122846252258},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.379326657430826e-07,"output_power":4.671892580949051e-07,"reference_active":true,"delta_db":-0.6123506878784681}],"active_band_count":3,"active_band_mae_db":0.5375374647228292}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":1.9376318186348178,"rms_envelope_nrmse":0.2183149693453127,"envelope_correlation":0.9722005872696626,"rms_gain_error_db":0.41898682660531406,"reference_mean_power":0.00015588925116025805,"output_mean_power":0.00017167808232776237},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103456000035227e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.437460419122184e-05,"output_power":6.16811307171536e-05,"reference_active":true,"delta_db":0.5475621783549979},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.583240107798151e-05,"output_power":1.8319557608129916e-05,"reference_active":true,"delta_db":0.6336819864632726},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.90382854367363e-06,"output_power":2.2053295824379027e-06,"reference_active":true,"delta_db":-2.4801723201676276}],"active_band_count":3,"active_band_mae_db":1.220472161661966}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":1.1043542211653845,"rms_envelope_nrmse":0.12929837404542532,"envelope_correlation":0.9908393349931188,"rms_gain_error_db":0.41898693352524174,"reference_mean_power":0.00015667279701186946,"output_mean_power":0.00017254099179977281},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1374269590951092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4195005275221475e-05,"output_power":6.164514758066771e-05,"reference_active":true,"delta_db":0.5593963373828397},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661491941713124e-06,"output_power":8.132448311766219e-06,"reference_active":true,"delta_db":0.8664980480133223},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7948760162946994e-07,"output_power":1.2863255265461962e-07,"reference_active":true,"delta_db":-3.3701165877920762}],"active_band_count":3,"active_band_mae_db":1.5986703243960794}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"facodec","status":"measured","mos":{"status":"measured","value":3.8810207843780518,"predicted_mos":3.8810207843780518,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"be5baf682317fdcb983960b9f9cdda0e0752128b03c97c3b20bc98991a6ab9b0","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.809301532000973},"started_at":"2026-10-13T06:58:36.842467+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.20329995197244e-05,"native_input_samples":96000,"native_input_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","output_sha256":"6e1b76bfd83d72fabebd164bbc5d76fac2a88c662966291f23f251ef924318c8","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7065524049994565,"synthesis_s":5.637999024000237,"total_s":7.344551428999694,"rtf":1.2240919048332823,"output_peak":0.4734083116054535,"output_rms":0.044539289235093045,"finished_at":"2026-10-13T06:58:44.200683+00:00","evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"6e1b76bfd83d72fabebd164bbc5d76fac2a88c662966291f23f251ef924318c8","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.568591694566162,"spectral_convergence":0.32529411929040364,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.78080555786106,"spectral_convergence":0.33660269901944595,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.763319867753767,"spectral_convergence":0.3558272015915047,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.00021866105243081044,"delta_db":-1.8391882333969498,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0006613534936394799,"delta_db":-1.0900560773666423,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0005781602307607901,"delta_db":-0.351062015515422,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0003427436484050517,"delta_db":-0.3272927548544968,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":9.929855062661606e-05,"delta_db":-1.3182455972085763,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":8.353130970452643e-05,"delta_db":-1.355601877760286,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0469077593503955,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8795015708552658,"pitch":null,"output_rms_dbfs":-27.025134355711813,"output_peak":0.4734083116054535,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.70423904006033},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.568591694566162,"spectral_convergence":0.32529411929040364,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.78080555786106,"spectral_convergence":0.33660269901944595,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.763319867753767,"spectral_convergence":0.3558272015915047,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.00021866105243081044,"delta_db":-1.8391882333969498,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0006613534936394799,"delta_db":-1.0900560773666423,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0005781602307607901,"delta_db":-0.351062015515422,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0003427436484050517,"delta_db":-0.3272927548544968,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":9.929855062661606e-05,"delta_db":-1.3182455972085763,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":8.353130970452643e-05,"delta_db":-1.355601877760286,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0469077593503955,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8795015708552658,"pitch":null,"output_rms_dbfs":-27.025134355711813,"output_peak":0.4734083116054535,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.70423904006033},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8599654520100599,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04928540852502758,"rms_dbfs":-26.14563278485655,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.384597483135295,"mean":-0.0027082255681355792,"zero_energy":false},"output":{"rms":0.044539289235093045,"rms_dbfs":-27.025134355711813,"peak":0.4734083116054535,"peak_dbfs":-6.49528242908099,"crest_factor_db":20.52985192663083,"mean":-0.0001468186096423286,"zero_energy":false},"rms_difference_db":-0.8795015708552647},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":7.126778752779278e-07,"output_rms_dbfs":-61.47106723445022,"output_zero_energy":false,"output_peak":0.004731113091111183,"output_energy_fraction":1.1975273954244479e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":59,"both_estimated_voiced_frames":53,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.8983050847457628,"absolute_difference_cents_median":3.80947770544153,"absolute_difference_cents_p95":11.342758525904067,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018099058448312497,"log_rms_mae_db":3.0289008695730835,"rms_envelope_nrmse":0.1808694280025854,"envelope_correlation":0.9755283177151229,"rms_gain_error_db":-0.8794988132587673,"reference_mean_power":0.002431072908198481,"output_mean_power":0.0019854003872941854},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.757963657265947e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033408940482734,"output_power":0.00028812729533913997,"reference_active":true,"delta_db":-0.6427830030779664},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010057920506755965,"output_power":8.433269892280116e-05,"reference_active":true,"delta_db":-0.7651219943525571},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.918423836914529e-05,"output_power":1.9679217634526956e-05,"reference_active":true,"delta_db":0.1106326665021975}],"active_band_count":3,"active_band_mae_db":0.506179221310907}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016835992502135956,"log_rms_mae_db":2.5884949435177758,"rms_envelope_nrmse":0.1706701581904071,"envelope_correlation":0.9777836461582038,"rms_gain_error_db":-0.8794043100058331,"reference_mean_power":0.002443196853891242,"output_mean_power":0.001995345148785613},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.921742652983223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003035779346246488,"output_power":0.00026166651945233246,"reference_active":true,"delta_db":-0.6452204425309743},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.673255237859031e-05,"output_power":3.66873332015643e-05,"reference_active":true,"delta_db":-1.0510335732839124},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2224669143851135e-06,"output_power":1.173056723508661e-06,"reference_active":true,"delta_db":-0.17918100872360454}],"active_band_count":3,"active_band_mae_db":0.6251450081794971}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015437420870825604,"log_rms_mae_db":1.4607924280814426,"rms_envelope_nrmse":0.14953657982350246,"envelope_correlation":0.9831958080210975,"rms_gain_error_db":-0.2840222364232029,"reference_mean_power":0.0008784594297426283,"output_mean_power":0.0008228477490717166},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.115238295382225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019746522774839338,"output_power":0.00021364877626491198,"reference_active":true,"delta_db":0.3420977915223594},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173113277871683e-05,"output_power":4.8536707334628035e-05,"reference_active":true,"delta_db":-0.2768167740718776},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0925495372248409e-05,"output_power":1.0634329249699938e-05,"reference_active":true,"delta_db":-0.11731035022408315}],"active_band_count":3,"active_band_mae_db":0.24540830527277335}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012941193693393203,"log_rms_mae_db":1.1690869503017585,"rms_envelope_nrmse":0.1314274883986438,"envelope_correlation":0.9862451255319487,"rms_gain_error_db":-0.28402038457790507,"reference_mean_power":0.000882875660341548,"output_mean_power":0.0008269847587670651},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7308162041776265e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001828938795712491,"output_power":0.0001977559352299779,"reference_active":true,"delta_db":0.3393035447642714},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517691716707475e-05,"output_power":2.2412213791384296e-05,"reference_active":true,"delta_db":-0.5051779448292189},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.42204663895331e-07,"output_power":9.449558759878314e-07,"reference_active":true,"delta_db":0.012662802676821378}],"active_band_count":3,"active_band_mae_db":0.2857147640901039}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":1.6430197245487903,"rms_envelope_nrmse":0.16087931774412523,"envelope_correlation":0.9821374817638632,"rms_gain_error_db":-0.4503030165275761,"reference_mean_power":0.0004246628880558568,"output_mean_power":0.00038283709079090086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.038477949048031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001401263951276247,"output_power":0.00011115344266570308,"reference_active":true,"delta_db":-1.0059703125588058},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6627820389923164e-05,"output_power":3.1782192933174585e-05,"reference_active":true,"delta_db":-0.6162721618773539},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.517673552823718e-06,"output_power":8.276414533055151e-06,"reference_active":true,"delta_db":-0.12478756705456212}],"active_band_count":3,"active_band_mae_db":0.5823433471635739}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":1.2874973493475919,"rms_envelope_nrmse":0.13709161365412847,"envelope_correlation":0.9864625350324489,"rms_gain_error_db":-0.4503013039418747,"reference_mean_power":0.00042679805438924104,"output_mean_power":0.00038476211255016583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.813635389636041e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013656501467579054,"output_power":0.00010821523525660514,"reference_active":true,"delta_db":-1.010510477949411},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7671441991318864e-05,"output_power":1.484686661033148e-05,"reference_active":true,"delta_db":-0.7563718273485632},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.379326657430826e-07,"output_power":5.729705247833067e-07,"reference_active":true,"delta_db":0.27404363772362494}],"active_band_count":3,"active_band_mae_db":0.6803086476738663}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":2.053224645229422,"rms_envelope_nrmse":0.23190132835132019,"envelope_correlation":0.9737608487596563,"rms_gain_error_db":-1.281804277641751,"reference_mean_power":0.00015588925116025805,"output_mean_power":0.00011604748774978546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103456000035227e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.437460419122184e-05,"output_power":3.782031865443499e-05,"reference_active":true,"delta_db":-1.5767092495158286},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.583240107798151e-05,"output_power":1.1594548623038434e-05,"reference_active":true,"delta_db":-1.3529293702905805},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.90382854367363e-06,"output_power":4.920084444193747e-06,"reference_active":true,"delta_db":1.0048182151655525}],"active_band_count":3,"active_band_mae_db":1.3114856116573206}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":1.6987120710191126,"rms_envelope_nrmse":0.19983371244043768,"envelope_correlation":0.9833211888226541,"rms_gain_error_db":-1.2817995904526815,"reference_mean_power":0.00015667279701186946,"output_mean_power":0.0001166309028960695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1374269590951092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4195005275221475e-05,"output_power":3.856198870341127e-05,"reference_active":true,"delta_db":-1.477998398278296},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661491941713124e-06,"output_power":5.026081186644375e-06,"reference_active":true,"delta_db":-1.2234200736880423},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7948760162946994e-07,"output_power":3.5751990782525226e-07,"reference_active":true,"delta_db":1.0693768280421292}],"active_band_count":3,"active_band_mae_db":1.2569317666694892}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.951828956604004,"predicted_mos":2.951828956604004,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"66ccc4e736ed0b9cada52f47772c1af807b533b0c51976b62dfcd1e2cba2ce76","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1324118099983025},"started_at":"2026-10-13T07:28:06.861773+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006433198002923746,"native_input_samples":264600,"native_input_sha256":"f3fad961fd5b622667ee44a16fc63bc6521ba6b37c049072c6f6461406610bf0","output_sha256":"87cb996a2b727dd29ffee1b87c55e76c1f5f3c488bfb7eb6a780c4784029c992","output_samples":264704,"length_delta_samples":104,"analysis_s":0.22152374500001315,"synthesis_s":4.478827790997457,"total_s":4.70035153599747,"rtf":0.783391922666245,"output_peak":0.5265930891036987,"output_rms":0.043351561903498516,"finished_at":"2026-10-13T07:28:11.582501+00:00","evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"87cb996a2b727dd29ffee1b87c55e76c1f5f3c488bfb7eb6a780c4784029c992","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.255758657752644,"output_peak":0.4878113397647321},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.68139081594599,"spectral_convergence":0.5287993601471732,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.326670187110588,"spectral_convergence":0.4875703251848734,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.868046667284784,"spectral_convergence":0.5385491308558523,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.0002539996303487378,"delta_db":-1.188571728208539,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0006951845377708567,"delta_db":-0.8733915580037739,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0004822458886996682,"delta_db":-1.138858827688139,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.00024980030867871665,"delta_db":-1.7010572014442793,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.00013153737348782786,"delta_db":-0.09718303545120621,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":6.912709949506704e-05,"delta_db":-2.177611439309772,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1961122983509516,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.108413704058409,"pitch":null,"output_rms_dbfs":-27.254046488914963,"output_peak":0.4878113397647321,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.292035890113787},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.616377837372369,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04928540852502758,"rms_dbfs":-26.14563278485655,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.384597483135295,"mean":-0.0027082255681355792,"zero_energy":false},"output":{"rms":0.04338081186977572,"rms_dbfs":-27.254046488914963,"peak":0.4878113397647321,"peak_dbfs":-6.234962164086299,"crest_factor_db":21.01908432482866,"mean":-9.765507776121581e-05,"zero_energy":false},"rms_difference_db":-1.1084137040584139},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":1.0266265777154174e-06,"output_rms_dbfs":-59.885874967421785,"output_zero_energy":false,"output_peak":0.0041999948955684695,"output_energy_fraction":1.8184271099585693e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":93,"both_estimated_voiced_frames":57,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.9661016949152542,"absolute_difference_cents_median":15.603445826037476,"absolute_difference_cents_p95":32.77080737945205,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018099058448312497,"log_rms_mae_db":2.4620437005148434,"rms_envelope_nrmse":0.1752975962655919,"envelope_correlation":0.9833075985233297,"rms_gain_error_db":-1.108409388085154,"reference_mean_power":0.002431072908198481,"output_mean_power":0.0018834627906426339},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.757963657265947e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033408940482734,"output_power":0.0002698274365409318,"reference_active":true,"delta_db":-0.9277659539343653},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010057920506755965,"output_power":8.573691734886715e-05,"reference_active":true,"delta_db":-0.6934033433460078},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.918423836914529e-05,"output_power":1.5060225042094328e-05,"reference_active":true,"delta_db":-1.0511310051658465}],"active_band_count":3,"active_band_mae_db":0.8907667674820733}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016835992502135956,"log_rms_mae_db":2.2147019169526483,"rms_envelope_nrmse":0.15097451122722405,"envelope_correlation":0.9909662453956379,"rms_gain_error_db":-1.1083075855353557,"reference_mean_power":0.002443196853891242,"output_mean_power":0.0018929001337536881},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.921742652983223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003035779346246488,"output_power":0.00025007721323846565,"reference_active":true,"delta_db":-0.8419608084725896},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.673255237859031e-05,"output_power":4.060494283837813e-05,"reference_active":true,"delta_db":-0.61040597962503},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2224669143851135e-06,"output_power":8.785845026570132e-07,"reference_active":true,"delta_db":-1.434535754542203}],"active_band_count":3,"active_band_mae_db":0.9623008475466076}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015437420870825604,"log_rms_mae_db":2.6585381170299267,"rms_envelope_nrmse":0.20394516385632813,"envelope_correlation":0.9743959259187907,"rms_gain_error_db":-0.9899282684586668,"reference_mean_power":0.0008784594297426283,"output_mean_power":0.0006994052407913856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.115238295382225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019746522774839338,"output_power":0.0001484409962961124,"reference_active":true,"delta_db":-1.2393676990650622},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173113277871683e-05,"output_power":3.7847705728236055e-05,"reference_active":true,"delta_db":-1.357124300981404},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0925495372248409e-05,"output_power":8.00055485784408e-06,"reference_active":true,"delta_db":-1.3532102999974496}],"active_band_count":3,"active_band_mae_db":1.316567433347972}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012941193693393203,"log_rms_mae_db":2.1178247926736775,"rms_envelope_nrmse":0.16272305016139982,"envelope_correlation":0.9863264359171683,"rms_gain_error_db":-0.989924086849811,"reference_mean_power":0.000882875660341548,"output_mean_power":0.0007029219989590788},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7308162041776265e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001828938795712491,"output_power":0.0001363228370045646,"reference_active":true,"delta_db":-1.2763055674036858},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517691716707475e-05,"output_power":1.8249795197913143e-05,"reference_active":true,"delta_db":-1.3974455595951287},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.42204663895331e-07,"output_power":6.177201827025583e-07,"reference_active":true,"delta_db":-1.833534584324132}],"active_band_count":3,"active_band_mae_db":1.5024285704409823}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":3.2601018275312725,"rms_envelope_nrmse":0.27776631149396613,"envelope_correlation":0.9567518719760105,"rms_gain_error_db":-1.541039177347063,"reference_mean_power":0.0004246628880558568,"output_mean_power":0.00029781076435390403},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.038477949048031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001401263951276247,"output_power":0.00012701706220676385,"reference_active":true,"delta_db":-0.4265788588223208},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6627820389923164e-05,"output_power":3.509393863444334e-05,"reference_active":true,"delta_db":-0.1857896355441387},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.517673552823718e-06,"output_power":6.378486228523582e-06,"reference_active":true,"delta_db":-1.2560336914641752}],"active_band_count":3,"active_band_mae_db":0.6228007286102116}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":2.8917346002342517,"rms_envelope_nrmse":0.248149488362042,"envelope_correlation":0.9675860617208576,"rms_gain_error_db":-1.541034778164304,"reference_mean_power":0.00042679805438924104,"output_mean_power":0.00029930843294396266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.813635389636041e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013656501467579054,"output_power":0.00012625956026952068,"reference_active":true,"delta_db":-0.34075183264853925},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7671441991318864e-05,"output_power":1.7494533642702803e-05,"reference_active":true,"delta_db":-0.04369619574720014},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.379326657430826e-07,"output_power":4.315982444788457e-07,"reference_active":true,"delta_db":-0.9564824791258546}],"active_band_count":3,"active_band_mae_db":0.44697683584053133}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":3.9770567659411538,"rms_envelope_nrmse":0.24913013691101926,"envelope_correlation":0.9728009149023915,"rms_gain_error_db":-1.4969961399947471,"reference_mean_power":0.00015588925116025805,"output_mean_power":0.00011043749745651885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103456000035227e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.437460419122184e-05,"output_power":3.875322598058435e-05,"reference_active":true,"delta_db":-1.4708824778068876},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.583240107798151e-05,"output_power":1.326924848791401e-05,"reference_active":true,"delta_db":-0.7670045617900167},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.90382854367363e-06,"output_power":2.050386077251849e-06,"reference_active":true,"delta_db":-2.7965509098627765}],"active_band_count":3,"active_band_mae_db":1.678145983153227}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":3.711440935992617,"rms_envelope_nrmse":0.2188399595519189,"envelope_correlation":0.9826708498619741,"rms_gain_error_db":-1.4969905910283157,"reference_mean_power":0.00015667279701186946,"output_mean_power":0.00011099273105843872},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1374269590951092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4195005275221475e-05,"output_power":3.8638891977019087e-05,"reference_active":true,"delta_db":-1.4693459896215466},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661491941713124e-06,"output_power":6.510658833525776e-06,"reference_active":true,"delta_db":-0.09946568489568457},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7948760162946994e-07,"output_power":1.4988119771625693e-07,"reference_active":true,"delta_db":-2.7061539195150637}],"active_band_count":3,"active_band_mae_db":1.4249885313440984}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"golf","status":"measured","mos":{"status":"measured","value":3.3529703617095947,"predicted_mos":3.3529703617095947,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"4966e17c21f498824288e4a84c8cdd146e13afd91df9a667a60216d966b86a79","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.551490483001544},"started_at":"2026-10-13T05:02:56.663485+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031868599999143044,"native_input_samples":144000,"native_input_sha256":"8865d014222c86c1b168a71abeb42f14b2da215030045b4ce7a9d524d7ac01bb","output_sha256":"a26ec13f0f7f7e40099c4a24b9307d5ef0a041c5bb441ecda53abb23b639aea4","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5697950010007844,"synthesis_s":0.12810619099946052,"total_s":0.6979011920002449,"rtf":0.11631686533337415,"output_peak":0.4729345440864563,"output_rms":0.045483448981209776,"finished_at":"2026-10-13T05:02:57.373845+00:00","analysis_stages":{"external_f0_s":0.08490610000080778},"evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"a26ec13f0f7f7e40099c4a24b9307d5ef0a041c5bb441ecda53abb23b639aea4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.841176105588566,"output_peak":0.4635369869271445},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.659388564135084,"spectral_convergence":0.5254489477007981,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.454627526080373,"spectral_convergence":0.32641398165715135,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.5460885401558455,"spectral_convergence":0.3627124318475457,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003344824434143825,"output_power":0.00027928536363273443,"delta_db":-0.78325150655899,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008514578141100089,"output_power":0.0008089919275696772,"delta_db":-0.22218947423352486,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006278843760908538,"output_power":0.0005107924485312889,"delta_db":-0.8963520835878029,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003701884440916829,"output_power":0.0002667773775879097,"delta_db":-1.4227385843003764,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001347384315650688,"output_power":0.00010339069751338949,"delta_db":-1.1501002234817341,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011432295280743207,"output_power":0.00010034299835831547,"delta_db":-0.5664635908878304,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.840182577175043,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7027301088932837,"pitch":null,"output_rms_dbfs":-26.841176105588566,"output_peak":0.4635369869271445,"reference_rms_dbfs":-26.138445996695282,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.886701543457101},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8405112385029017,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.049326204618634795,"rms_dbfs":-26.138445996695282,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.377410694974024,"mean":-0.0027059784317653448,"zero_energy":false},"output":{"rms":0.0454926457044797,"rms_dbfs":-26.841176105588566,"peak":0.4635369869271445,"peak_dbfs":-6.6783121310378135,"crest_factor_db":20.162863974550753,"mean":0.0005057265906101435,"zero_energy":false},"rms_difference_db":-0.7027301088932845},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":1.9344515419237387e-06,"output_rms_dbfs":-57.13442144897873,"output_zero_energy":false,"output_peak":0.004082318527380895,"output_energy_fraction":3.120891274087526e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":62,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9491525423728814,"absolute_difference_cents_median":7.273379494561938,"absolute_difference_cents_p95":25.906767178037107,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00018099058448312497,"log_rms_mae_db":2.521208152978585,"rms_envelope_nrmse":0.15138112307644008,"envelope_correlation":0.983243967945652,"rms_gain_error_db":-0.7028852034117928,"reference_mean_power":0.002435091313603159,"output_mean_power":0.0020712223849941444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.77155347973889e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003341442561591366,"output_power":0.0002766110769024019,"reference_active":true,"delta_db":-0.8206443266765685},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010065706534741455,"output_power":9.021495769492189e-05,"reference_active":true,"delta_db":-0.47565714606494386},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9195010270165833e-05,"output_power":1.9000136081051214e-05,"reference_active":true,"delta_db":-0.04431637391420669}],"active_band_count":3,"active_band_mae_db":0.4468726155519063}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016835992502135956,"log_rms_mae_db":1.749524527407028,"rms_envelope_nrmse":0.11972751423315722,"envelope_correlation":0.9911744177609723,"rms_gain_error_db":-0.7032778312344843,"reference_mean_power":0.002447220631244881,"output_mean_power":0.002081351077486968},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":7.936402440377387e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00030366742062761686,"output_power":0.0002503329151184232,"reference_active":true,"delta_db":-0.8388024380556653},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.6767818194012726e-05,"output_power":4.1651231642761374e-05,"reference_active":true,"delta_db":-0.5031926154742027},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2202380178393209e-06,"output_power":1.2056646572669782e-06,"reference_active":true,"delta_db":-0.05218021549582666}],"active_band_count":3,"active_band_mae_db":0.4647250896752315}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015437420870825604,"log_rms_mae_db":1.800596541298505,"rms_envelope_nrmse":0.17779373762884315,"envelope_correlation":0.9769970467347773,"rms_gain_error_db":-0.47931662312971446,"reference_mean_power":0.0008799267302469161,"output_mean_power":0.0007879793719592377},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.1185996179244935e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001973965123649926,"output_power":0.00016899460289200217,"reference_active":true,"delta_db":-0.6746664025401307},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173439825362333e-05,"output_power":4.750984724468697e-05,"reference_active":true,"delta_db":-0.36995767912565725},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.09451923470098e-05,"output_power":1.0613085058279337e-05,"reference_active":true,"delta_db":-0.13381753226567186}],"active_band_count":3,"active_band_mae_db":0.39281387131048656}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012941193693393203,"log_rms_mae_db":1.1872043855293732,"rms_envelope_nrmse":0.14614302325928777,"envelope_correlation":0.9840580141964568,"rms_gain_error_db":-0.47931556827197214,"reference_mean_power":0.0008843561292687433,"output_mean_power":0.0007919461162327746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.734784477690062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018284161166124237,"output_power":0.00015587246321886286,"reference_active":true,"delta_db":-0.6930564221024537},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.518353699713111e-05,"output_power":2.3168791412313934e-05,"reference_active":true,"delta_db":-0.36213346547159453},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.409030902016336e-07,"output_power":8.593120652395628e-07,"reference_active":true,"delta_db":-0.39393985303667745}],"active_band_count":3,"active_band_mae_db":0.4830432468702419}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":1.9957907340462968,"rms_envelope_nrmse":0.2308422752262716,"envelope_correlation":0.9730008878841004,"rms_gain_error_db":-1.383389774061969,"reference_mean_power":0.0004253721305789108,"output_mean_power":0.0003093357077534583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.039292907601217e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014032935339980504,"output_power":0.00010628899157279558,"reference_active":true,"delta_db":-1.2066023744064989},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.654499638912836e-05,"output_power":2.7491684332299007e-05,"reference_active":true,"delta_db":-1.2362657474184853},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.516756984102733e-06,"output_power":5.155594021374285e-06,"reference_active":true,"delta_db":-2.1799554393410743}],"active_band_count":3,"active_band_mae_db":1.5409411870553527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":1.4974647440197077,"rms_envelope_nrmse":0.18980561708787502,"envelope_correlation":0.9866932551943811,"rms_gain_error_db":-1.3833982057179135,"reference_mean_power":0.0004275130300472027,"output_mean_power":0.00031089199190179913},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.8146653776801975e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013675774277153603,"output_power":0.00010414818899586714,"reference_active":true,"delta_db":-1.1830020149770608},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577299867408626e-05,"output_power":1.3186802592701763e-05,"reference_active":true,"delta_db":-1.2481265700150574},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.383206428043633e-07,"output_power":3.45026308505348e-07,"reference_active":true,"delta_db":-1.9318882227849512}],"active_band_count":3,"active_band_mae_db":1.4543389359256897}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":2.5354919496394155,"rms_envelope_nrmse":0.1778004826012101,"envelope_correlation":0.9845589798674588,"rms_gain_error_db":-0.8016699677335087,"reference_mean_power":0.00015614947435906831,"output_mean_power":0.00012982954343731065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.205418740891517e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4128438289339e-05,"output_power":4.796523364261844e-05,"reference_active":true,"delta_db":-0.5249893273635227},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.576204766966827e-05,"output_power":1.4339367426519949e-05,"reference_active":true,"delta_db":-0.410826435355662},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.865636428169866e-06,"output_power":3.103719218701471e-06,"reference_active":true,"delta_db":-0.9533857979148964}],"active_band_count":3,"active_band_mae_db":0.6297338535446937}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":2.160975866577481,"rms_envelope_nrmse":0.12479956584232278,"envelope_correlation":0.9954451090162795,"rms_gain_error_db":-0.8016467459916115,"reference_mean_power":0.00015693388932727054,"output_mean_power":0.0001304824382437325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1325825390878352e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.39456208596342e-05,"output_power":4.8555430620487095e-05,"reference_active":true,"delta_db":-0.45718385677753665},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.580314927444137e-06,"output_power":6.014557271622205e-06,"reference_active":true,"delta_db":-0.39043014355580047},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7874842533245406e-07,"output_power":2.1414965687066738e-07,"reference_active":true,"delta_db":-1.1449503983755813}],"active_band_count":3,"active_band_mae_db":0.6641881329029728}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.4918482303619385,"predicted_mos":2.4918482303619385,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"592501136544d8525af42d80484057f10ac762b972e633b8ff230e89f3c37d40","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8625861030013766},"started_at":"2026-10-13T07:44:16.332453+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004502879000938265,"native_input_samples":144000,"native_input_sha256":"8865d014222c86c1b168a71abeb42f14b2da215030045b4ce7a9d524d7ac01bb","output_sha256":"62573f0ed913d33071649eba95eae4f9cecea629be225d3a9441c003b0bd897d","output_samples":144120,"length_delta_samples":120,"analysis_s":2.298884817002545,"synthesis_s":3.6934183269986534,"total_s":5.992303144001198,"rtf":0.9987171906668664,"output_peak":0.5001639723777771,"output_rms":0.04864589567925742,"finished_at":"2026-10-13T07:44:22.346408+00:00","analysis_stages":{"external_f0_s":1.5140880559993093},"evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"62573f0ed913d33071649eba95eae4f9cecea629be225d3a9441c003b0bd897d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.255908119342696,"output_peak":0.5047672043944789},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.814394667594403,"spectral_convergence":0.540091274504576,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.56817522250128,"spectral_convergence":0.35252845977083386,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.3195234712272494,"spectral_convergence":0.34628043105984774,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.00041332291323342577,"delta_db":0.925992243728795,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0007990100486321582,"delta_db":-0.2688701862246178,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0006000382958647856,"delta_db":-0.18975447630931466,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0003490898836793826,"delta_db":-0.24761427320483342,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.00012202795126390032,"delta_db":-0.42308149317579147,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":8.663381933029463e-05,"delta_db":-1.1972201754131222,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5420888080094124,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10665852879373773,"pitch":null,"output_rms_dbfs":-26.25229131365029,"output_peak":0.5047672043944789,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.900697787107643},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8403778235632884,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04928540852502758,"rms_dbfs":-26.14563278485655,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.384597483135295,"mean":-0.0027082255681355792,"zero_energy":false},"output":{"rms":0.048683908142259355,"rms_dbfs":-26.25229131365029,"peak":0.5047672043944789,"peak_dbfs":-5.938177394343147,"crest_factor_db":20.314113919307143,"mean":-1.0512970334448215e-05,"zero_energy":false},"rms_difference_db":-0.10665852879374071},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":1.837937602026033e-07,"output_rms_dbfs":-67.35669236995489,"output_zero_energy":false,"output_peak":0.003304870965123624,"output_energy_fraction":2.584869604184438e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":65,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9491525423728814,"absolute_difference_cents_median":4.551549607952804,"absolute_difference_cents_p95":21.12691100057238,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018099058448312497,"log_rms_mae_db":5.218455434030575,"rms_envelope_nrmse":0.17415140188704456,"envelope_correlation":0.9719815027587929,"rms_gain_error_db":-0.10665105313366016,"reference_mean_power":0.002431072908198481,"output_mean_power":0.0023720993705479637},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.757963657265947e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033408940482734,"output_power":0.0003383930266222414,"reference_active":true,"delta_db":0.055587019865666526},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010057920506755965,"output_power":9.725939483649808e-05,"reference_active":true,"delta_db":-0.1457663584089477},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.918423836914529e-05,"output_power":1.5768100170041324e-05,"reference_active":true,"delta_db":-0.8516519178648468}],"active_band_count":3,"active_band_mae_db":0.35100176537982036}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016835992502135956,"log_rms_mae_db":4.430707392941007,"rms_envelope_nrmse":0.10250085028766183,"envelope_correlation":0.9898604349272297,"rms_gain_error_db":-0.10647305976285402,"reference_mean_power":0.002443196853891242,"output_mean_power":0.002384026916839147},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.921742652983223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003035779346246488,"output_power":0.0003073079394634129,"reference_active":true,"delta_db":0.05303578685893755},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.673255237859031e-05,"output_power":4.527976453651983e-05,"reference_active":true,"delta_db":-0.1371534163962629},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2224669143851135e-06,"output_power":1.1700239537026953e-06,"reference_active":true,"delta_db":-0.19042360860474059}],"active_band_count":3,"active_band_mae_db":0.126870937286647}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015437420870825604,"log_rms_mae_db":2.1901417776462115,"rms_envelope_nrmse":0.1900614591682575,"envelope_correlation":0.9723229910541569,"rms_gain_error_db":-0.160036031884967,"reference_mean_power":0.0008784594297426283,"output_mean_power":0.0008466776698462694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.115238295382225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019746522774839338,"output_power":0.00019184917602940238,"reference_active":true,"delta_db":-0.12530692168525162},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173113277871683e-05,"output_power":5.098636293156387e-05,"reference_active":true,"delta_db":-0.06297955396737129},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0925495372248409e-05,"output_power":1.0705418125486967e-05,"reference_active":true,"delta_db":-0.08837503047253817}],"active_band_count":3,"active_band_mae_db":0.09222050204172035}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012941193693393203,"log_rms_mae_db":1.546653705252808,"rms_envelope_nrmse":0.11451199535232832,"envelope_correlation":0.9893572398933642,"rms_gain_error_db":-0.16003032570486136,"reference_mean_power":0.000882875660341548,"output_mean_power":0.0008509352437981864},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7308162041776265e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001828938795712491,"output_power":0.00017854522263578224,"reference_active":true,"delta_db":-0.10450938133077549},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517691716707475e-05,"output_power":2.5117086440030982e-05,"reference_active":true,"delta_db":-0.010332908496969074},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.42204663895331e-07,"output_power":9.022161744023014e-07,"reference_active":true,"delta_db":-0.18834641048856557}],"active_band_count":3,"active_band_mae_db":0.10106290010543671}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":2.2159320028282656,"rms_envelope_nrmse":0.2432119602980381,"envelope_correlation":0.9563365369284248,"rms_gain_error_db":-0.23643758967015424,"reference_mean_power":0.0004246628880558568,"output_mean_power":0.0004021615536567286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.038477949048031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001401263951276247,"output_power":0.00013137670673623737,"reference_active":true,"delta_db":-0.2800157848758761},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6627820389923164e-05,"output_power":3.242543989529197e-05,"reference_active":true,"delta_db":-0.5292519929832902},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.517673552823718e-06,"output_power":7.39327058435879e-06,"reference_active":true,"delta_db":-0.6148439022904583}],"active_band_count":3,"active_band_mae_db":0.4747038933832082}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":1.4036651173250039,"rms_envelope_nrmse":0.1379910977809388,"envelope_correlation":0.9851683889378987,"rms_gain_error_db":-0.23643496596943564,"reference_mean_power":0.00042679805438924104,"output_mean_power":0.0004041838295088846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.813635389636041e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013656501467579054,"output_power":0.00012836255042457604,"reference_active":true,"delta_db":-0.2690111825048606},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7671441991318864e-05,"output_power":1.569576284463954e-05,"reference_active":true,"delta_db":-0.5148956131650615},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.379326657430826e-07,"output_power":4.6167446504377474e-07,"reference_active":true,"delta_db":-0.6639206280826582}],"active_band_count":3,"active_band_mae_db":0.48260914125086013}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":2.426384709337962,"rms_envelope_nrmse":0.2537267972988458,"envelope_correlation":0.9615332685700505,"rms_gain_error_db":-0.9804647612110302,"reference_mean_power":0.00015588925116025805,"output_mean_power":0.00012438548237558346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103456000035227e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.437460419122184e-05,"output_power":4.142347592704465e-05,"reference_active":true,"delta_db":-1.1814957000804418},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.583240107798151e-05,"output_power":1.1772424474781483e-05,"reference_active":true,"delta_db":-1.286808702677563},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.90382854367363e-06,"output_power":2.2317881015416936e-06,"reference_active":true,"delta_db":-2.4283777730703764}],"active_band_count":3,"active_band_mae_db":1.6322273919427939}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":1.3913233724972975,"rms_envelope_nrmse":0.17120199311786236,"envelope_correlation":0.9864623778689192,"rms_gain_error_db":-0.980464102364561,"reference_mean_power":0.00015667279701186946,"output_mean_power":0.00012501069985780848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1374269590951092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4195005275221475e-05,"output_power":4.086029340513222e-05,"reference_active":true,"delta_db":-1.2265778205495754},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661491941713124e-06,"output_power":5.055537761622816e-06,"reference_active":true,"delta_db":-1.198041482335305},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7948760162946994e-07,"output_power":1.6536608095852803e-07,"reference_active":true,"delta_db":-2.2791611278272703}],"active_band_count":3,"active_band_mae_db":1.5679268102373836}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.873199701309204,"predicted_mos":2.873199701309204,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"7ce596a97b2976b71f34cdb987791b133b7a8ff45d4f05fbbb26928ff8a87f5d","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.725262013000247},"started_at":"2026-10-13T05:11:17.780885+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00309326900060114,"native_input_samples":144000,"native_input_sha256":"8865d014222c86c1b168a71abeb42f14b2da215030045b4ce7a9d524d7ac01bb","output_sha256":"b2a137b5e248f5745c980e5f26ffdb44de46a4c8e9fc97d74f3200113c9bc90e","output_samples":144240,"length_delta_samples":240,"analysis_s":0.985887498000011,"synthesis_s":0.1484626030014624,"total_s":1.1343501010014734,"rtf":0.18905835016691222,"output_peak":0.5418850779533386,"output_rms":0.047777232461172665,"finished_at":"2026-10-13T05:11:18.927678+00:00","analysis_stages":{"external_f0_s":0.15300233000016306},"evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"b2a137b5e248f5745c980e5f26ffdb44de46a4c8e9fc97d74f3200113c9bc90e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.41244582935308,"output_peak":0.5217470760764196},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.068301517044168,"spectral_convergence":0.21290656886474835,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.642087966830916,"spectral_convergence":0.19831400449437647,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.941839105597165,"spectral_convergence":0.21417858164722228,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.0003177386276470058,"delta_db":-0.21620241964556278,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0008209744718543333,"delta_db":-0.15109606839506895,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0005988975668339081,"delta_db":-0.19801867939744527,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0003328086110519896,"delta_db":-0.45504136191513417,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.00012085048240251828,"delta_db":-0.46519080915379984,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":9.684921911717472e-05,"delta_db":-0.713133552188992,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.36644714844933385,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.25958118115697615,"pitch":null,"output_rms_dbfs":-26.405213966013527,"output_peak":0.5217470760764196,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.884076196490749},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9138737980097529,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04928540852502758,"rms_dbfs":-26.14563278485655,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.384597483135295,"mean":-0.0027082255681355792,"zero_energy":false},"output":{"rms":0.04783428664574115,"rms_dbfs":-26.405213966013527,"peak":0.5217470760764196,"peak_dbfs":-5.650799521721069,"crest_factor_db":20.754414444292458,"mean":-0.00015406800884836165,"zero_energy":false},"rms_difference_db":-0.25958118115697815},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":1.6828836203911315e-06,"output_rms_dbfs":-57.73945916566984,"output_zero_energy":false,"output_peak":0.00574866798706069,"output_energy_fraction":2.4516260385420328e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":61,"both_estimated_voiced_frames":56,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9491525423728814,"absolute_difference_cents_median":1.1762048210919622,"absolute_difference_cents_p95":10.151640119888363,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018099058448312497,"log_rms_mae_db":1.3691386296428982,"rms_envelope_nrmse":0.07022400359192513,"envelope_correlation":0.9961037957872839,"rms_gain_error_db":-0.25958305490960126,"reference_mean_power":0.002431072908198481,"output_mean_power":0.0022900221239963705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.757963657265947e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033408940482734,"output_power":0.0003204913205058234,"reference_active":true,"delta_db":-0.18046430300574265},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010057920506755965,"output_power":8.6755136632476e-05,"reference_active":true,"delta_db":-0.6421300072796898},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.918423836914529e-05,"output_power":1.2315858022675499e-05,"reference_active":true,"delta_db":-1.9247988830081584}],"active_band_count":3,"active_band_mae_db":0.9157977310978636}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016835992502135956,"log_rms_mae_db":0.8001498292067821,"rms_envelope_nrmse":0.04659281130008774,"envelope_correlation":0.9987227332732607,"rms_gain_error_db":-0.25952795348149366,"reference_mean_power":0.002443196853891242,"output_mean_power":0.0023014718385948657},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.921742652983223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003035779346246488,"output_power":0.0002907993310213011,"reference_active":true,"delta_db":-0.186767988420322},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.673255237859031e-05,"output_power":4.1733533609441256e-05,"reference_active":true,"delta_db":-0.4913434310441732},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2224669143851135e-06,"output_power":7.939192257722408e-07,"reference_active":true,"delta_db":-1.8746079486560852}],"active_band_count":3,"active_band_mae_db":0.8509064560401934}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015437420870825604,"log_rms_mae_db":1.056200438726117,"rms_envelope_nrmse":0.07725996264071615,"envelope_correlation":0.995720780532906,"rms_gain_error_db":-0.18949009330740466,"reference_mean_power":0.0008784594297426283,"output_mean_power":0.0008409548890385512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.115238295382225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019746522774839338,"output_power":0.00019198398680751432,"reference_active":true,"delta_db":-0.12225624315555952},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173113277871683e-05,"output_power":4.595999338525524e-05,"reference_active":true,"delta_db":-0.5137203093416466},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0925495372248409e-05,"output_power":6.000219536094892e-06,"reference_active":true,"delta_db":-2.602739967642869}],"active_band_count":3,"active_band_mae_db":1.079572173380025}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012941193693393203,"log_rms_mae_db":0.6667424375986112,"rms_envelope_nrmse":0.04671062379805162,"envelope_correlation":0.9985751486694238,"rms_gain_error_db":-0.1894897658397249,"reference_mean_power":0.000882875660341548,"output_mean_power":0.0008451826388589387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7308162041776265e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001828938795712491,"output_power":0.0001772725314705614,"reference_active":true,"delta_db":-0.1355772581275814},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517691716707475e-05,"output_power":2.1862268341332927e-05,"reference_active":true,"delta_db":-0.6130733045184965},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.42204663895331e-07,"output_power":4.33705972054135e-07,"reference_active":true,"delta_db":-3.369498473092907}],"active_band_count":3,"active_band_mae_db":1.3727163452463282}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":1.211011263931442,"rms_envelope_nrmse":0.11390206074840752,"envelope_correlation":0.9922842282713262,"rms_gain_error_db":-0.4856025687192945,"reference_mean_power":0.0004246628880558568,"output_mean_power":0.00037973799421718373},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.038477949048031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001401263951276247,"output_power":0.00012792182656138132,"reference_active":true,"delta_db":-0.3957529759438643},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6627820389923164e-05,"output_power":3.0390468490737102e-05,"reference_active":true,"delta_db":-0.8107368080046146},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.517673552823718e-06,"output_power":5.146228475135759e-06,"reference_active":true,"delta_db":-2.1883192783832426}],"active_band_count":3,"active_band_mae_db":1.1316030207772405}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":0.7431066355207248,"rms_envelope_nrmse":0.07787893518725611,"envelope_correlation":0.9974122034313083,"rms_gain_error_db":-0.4856019244071186,"reference_mean_power":0.00042679805438924104,"output_mean_power":0.00038164733889556145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.813635389636041e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013656501467579054,"output_power":0.00012328268773674944,"reference_active":true,"delta_db":-0.44437361672938797},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7671441991318864e-05,"output_power":1.4964916460084714e-05,"reference_active":true,"delta_db":-0.7219769263880339},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.379326657430826e-07,"output_power":3.2237161939276853e-07,"reference_active":true,"delta_db":-2.2237111657531066}],"active_band_count":3,"active_band_mae_db":1.1300205696235095}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":1.353904557980536,"rms_envelope_nrmse":0.11780030887075811,"envelope_correlation":0.9925367425926843,"rms_gain_error_db":-0.48461055847425927,"reference_mean_power":0.00015588925116025805,"output_mean_power":0.00013942964287549096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103456000035227e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.437460419122184e-05,"output_power":4.995927638212553e-05,"reference_active":true,"delta_db":-0.3677996925729146},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.583240107798151e-05,"output_power":1.3350810652527626e-05,"reference_active":true,"delta_db":-0.7403914662433206},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.90382854367363e-06,"output_power":2.058183262252492e-06,"reference_active":true,"delta_db":-2.780066931106799}],"active_band_count":3,"active_band_mae_db":1.2960860299743446}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":0.922676917035342,"rms_envelope_nrmse":0.0782718259486056,"envelope_correlation":0.9977448675445738,"rms_gain_error_db":-0.48460913614123197,"reference_mean_power":0.00015667279701186946,"output_mean_power":0.00014013050372209084},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1374269590951092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4195005275221475e-05,"output_power":4.9513601799007246e-05,"reference_active":true,"delta_db":-0.3923474326032961},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661491941713124e-06,"output_power":5.580328985115304e-06,"reference_active":true,"delta_db":-0.7691170359543493},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7948760162946994e-07,"output_power":1.207190867179838e-07,"reference_active":true,"delta_db":-3.6458660568889982}],"active_band_count":3,"active_band_mae_db":1.6024435084822146}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0002","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1476991176605225,"predicted_mos":3.1476991176605225,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289589,"input_sample_rate":48000,"input_seconds":6.033104166666667,"input_float32_sha256":"6e86670e8b216ef1866f7edaec4f23791545280c8509f78595fd3d91d6fa5fa4","internal_samples":96530,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4943821760007268},"started_at":"2026-10-13T05:17:46.988443+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009610799001166015,"native_input_samples":288000,"native_input_sha256":"f59a3385a414fd6432ee9cae83b3771f93c4a99ca5f15f6713217f4a986aa6a0","output_sha256":"6e1dc3e89d2c6cda193a86c9cc72586b720e99f694897915904bf74b47697c9d","output_samples":289589,"length_delta_samples":1589,"analysis_s":3.712583529000767,"synthesis_s":1.0086234849986795,"total_s":4.721207013999447,"rtf":0.7868678356665745,"output_peak":0.5404825526719973,"output_rms":0.04690425486810045,"finished_at":"2026-10-13T05:17:51.730416+00:00","evaluation_input":{"reference":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"6e1dc3e89d2c6cda193a86c9cc72586b720e99f694897915904bf74b47697c9d","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96530,"length_delta_samples":530,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.573086751986814,"output_peak":0.5406476013963798},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.994509478918471,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.801043576577756,"spectral_convergence":0.649248771317167,"spectral_convergence_status":"ok","amplitude_floor":1.3425996786840756e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.052842603052955,"spectral_convergence":0.49912246103353913,"spectral_convergence_status":"ok","amplitude_floor":1.1106411258473627e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.148781986458179,"spectral_convergence":0.4579304162774104,"spectral_convergence_status":"ok","amplitude_floor":1.0520318003394391e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033395677419724207,"output_power":0.0002902831803975543,"delta_db":-0.6086838475467222,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008500397987721807,"output_power":0.0007599852369277317,"delta_db":-0.4863410382412012,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006268367014596879,"output_power":0.0005615536204404412,"delta_db":-0.4776318503392527,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003695716856827319,"output_power":0.0003505179309217374,"delta_db":-0.22988451416944622,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013451399439061329,"output_power":0.00014352836864607084,"delta_db":0.28170279408551474,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00011413253897640505,"output_power":0.00010753268855295748,"delta_db":-0.25868974374153786,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.39048896468727917,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.40376614613240996,"pitch":null,"output_rms_dbfs":-26.549398930988964,"output_peak":0.5406476013963798,"reference_rms_dbfs":-26.14563278485655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.000889388696297},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.583913499950646,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04928540852502758,"rms_dbfs":-26.14563278485655,"peak":0.515167236328125,"peak_dbfs":-5.7610353017212566,"crest_factor_db":20.384597483135295,"mean":-0.0027082255681355792,"zero_energy":false},"output":{"rms":0.04704679612775447,"rms_dbfs":-26.549398930988964,"peak":0.5406476013963798,"peak_dbfs":-5.341714388666245,"crest_factor_db":21.207684542322717,"mean":-2.0667325882019212e-05,"zero_energy":false},"rms_difference_db":-0.40376614613241557},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.227729081409052e-06,"threshold_dbfs":-54.91102924813279,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":20,"quiet_samples":3200,"reference_mean_power":1.877044269349426e-06,"output_mean_power":2.8304519507175477e-06,"output_rms_dbfs":-55.48144213227842,"output_zero_energy":false,"output_peak":0.007495066930556033,"output_energy_fraction":4.262598474194885e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":59,"output_estimated_voiced_frames":54,"both_estimated_voiced_frames":50,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.847457627118644,"absolute_difference_cents_median":5.978458852075365,"absolute_difference_cents_p95":32.11401550011268,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018099058448312497,"log_rms_mae_db":3.971663838526495,"rms_envelope_nrmse":0.3237191921206227,"envelope_correlation":0.8996624218954502,"rms_gain_error_db":-0.4037642892817276,"reference_mean_power":0.002431072908198481,"output_mean_power":0.0022152439271497034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.757963657265947e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033408940482734,"output_power":0.0003191297250043225,"reference_active":true,"delta_db":-0.19895444838999962},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010057920506755965,"output_power":9.435519581869734e-05,"reference_active":true,"delta_db":-0.27742378557012193},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.918423836914529e-05,"output_power":1.8318888927752604e-05,"reference_active":true,"delta_db":-0.20045432589737913}],"active_band_count":3,"active_band_mae_db":0.22561085328583355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016835992502135956,"log_rms_mae_db":2.705793809138217,"rms_envelope_nrmse":0.22965025439448922,"envelope_correlation":0.9462753928675463,"rms_gain_error_db":-0.403844291494399,"reference_mean_power":0.002443196853891242,"output_mean_power":0.0022262505067191457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":7.921742652983223e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003035779346246488,"output_power":0.0002915761380519874,"reference_active":true,"delta_db":-0.1751822254573185},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.673255237859031e-05,"output_power":4.4393219076956784e-05,"reference_active":true,"delta_db":-0.22302863331474954},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2224669143851135e-06,"output_power":1.115426243314626e-06,"reference_active":true,"delta_db":-0.39796255747866727}],"active_band_count":3,"active_band_mae_db":0.2653911387502451}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015437420870825604,"log_rms_mae_db":4.402322462993594,"rms_envelope_nrmse":0.36865067773840704,"envelope_correlation":0.8937784383281863,"rms_gain_error_db":-0.4574113385291161,"reference_mean_power":0.0008784594297426283,"output_mean_power":0.000790643263976741},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.115238295382225e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019746522774839338,"output_power":0.0001798644197406785,"reference_active":true,"delta_db":-0.4054536949761456},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.173113277871683e-05,"output_power":4.5171060015321345e-05,"reference_active":true,"delta_db":-0.5889170630951158},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0925495372248409e-05,"output_power":9.166070763001846e-06,"reference_active":true,"delta_db":-0.7625793169021945}],"active_band_count":3,"active_band_mae_db":0.5856500249911519}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012941193693393203,"log_rms_mae_db":3.4682975743205025,"rms_envelope_nrmse":0.2725657752466557,"envelope_correlation":0.9385276905562321,"rms_gain_error_db":-0.4574261122845913,"reference_mean_power":0.000882875660341548,"output_mean_power":0.0007946153180996759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.7308162041776265e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001828938795712491,"output_power":0.00016671464874289043,"reference_active":true,"delta_db":-0.4022541059985379},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.517691716707475e-05,"output_power":2.1833315992964795e-05,"reference_active":true,"delta_db":-0.6188285070938017},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.42204663895331e-07,"output_power":7.749810283028811e-07,"reference_active":true,"delta_db":-0.8485417864004301}],"active_band_count":3,"active_band_mae_db":0.6232081331642566}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.92340600003817e-05,"log_rms_mae_db":4.607232754761174,"rms_envelope_nrmse":0.5174845875202037,"envelope_correlation":0.8020863395824106,"rms_gain_error_db":-0.19533846500811122,"reference_mean_power":0.0004246628880558568,"output_mean_power":0.00040598544363172287},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.038477949048031e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001401263951276247,"output_power":0.0001393381478990027,"reference_active":true,"delta_db":-0.02449916013510256},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6627820389923164e-05,"output_power":3.491987689781574e-05,"reference_active":true,"delta_db":-0.20738371862493363},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.517673552823718e-06,"output_power":8.32036392173022e-06,"reference_active":true,"delta_db":-0.10178669173278918}],"active_band_count":3,"active_band_mae_db":0.11122319016427512}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.482146577973155e-05,"log_rms_mae_db":3.3873362670379925,"rms_envelope_nrmse":0.38114596089821245,"envelope_correlation":0.8835651431326002,"rms_gain_error_db":-0.19577804013957087,"reference_mean_power":0.00042679805438924104,"output_mean_power":0.0004079854047694698},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.813635389636041e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013656501467579054,"output_power":0.00013730944748363124,"reference_active":true,"delta_db":0.023609638830403935},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7671441991318864e-05,"output_power":1.6211743812587895e-05,"reference_active":true,"delta_db":-0.37442257273377955},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.379326657430826e-07,"output_power":5.910677202870053e-07,"reference_active":true,"delta_db":0.4090932463021149}],"active_band_count":3,"active_band_mae_db":0.26904181928876614}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.51121938209953e-05,"log_rms_mae_db":4.8606395434038365,"rms_envelope_nrmse":0.578381234634875,"envelope_correlation":0.7845390650830437,"rms_gain_error_db":0.04338190200328912,"reference_mean_power":0.00015588925116025805,"output_mean_power":0.0001574542403761023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2103456000035227e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.437460419122184e-05,"output_power":5.2997273281959665e-05,"reference_active":true,"delta_db":-0.11142582960370623},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.583240107798151e-05,"output_power":1.5432668100804772e-05,"reference_active":true,"delta_db":-0.11105766972925758},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.90382854367363e-06,"output_power":4.818008864010295e-06,"reference_active":true,"delta_db":0.9137685948320129}],"active_band_count":3,"active_band_mae_db":0.37875069805499223}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.091623719145541e-05,"log_rms_mae_db":3.5549554106738177,"rms_envelope_nrmse":0.42354911308217885,"envelope_correlation":0.8781848517121089,"rms_gain_error_db":0.04215435710416683,"reference_mean_power":0.00015667279701186946,"output_mean_power":0.00015820093009647157},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1374269590951092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.4195005275221475e-05,"output_power":5.311005713183857e-05,"reference_active":true,"delta_db":-0.08782494295963174},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.661491941713124e-06,"output_power":6.433811073109056e-06,"reference_active":true,"delta_db":-0.15103203042668623},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.7948760162946994e-07,"output_power":3.342068088809403e-07,"reference_active":true,"delta_db":0.7765274676069198}],"active_band_count":3,"active_band_mae_db":0.33846148033107926}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"hnm","status":"measured","mos":{"status":"measured","value":1.7998404502868652,"predicted_mos":1.7998404502868652,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bfece6ccfde3e42024b7068b6af7e158c24d4db9a7398a8be720ad4bd22b068a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.840480075999949},"started_at":"2026-10-13T04:40:03.210169+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4100000043981709e-05,"native_input_samples":96000,"native_input_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","output_sha256":"b06580d8ff4699ad51919f05684be3626b647b064537839ae30ce59df994875b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1136678560005748,"synthesis_s":0.49418611799956125,"total_s":1.607853974000136,"rtf":0.267975662333356,"output_peak":0.4417002118673188,"output_rms":0.041002169856524176,"finished_at":"2026-10-13T04:40:04.825320+00:00","evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"b06580d8ff4699ad51919f05684be3626b647b064537839ae30ce59df994875b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.010339410142771,"spectral_convergence":0.36288799086597345,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.546588970332742,"spectral_convergence":0.34649700077674944,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.865188966926183,"spectral_convergence":0.3823329328254011,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.0003487641584913587,"delta_db":0.1384830342909462,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0005054794389001199,"delta_db":-0.11150692098886007,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.000421076264430183,"delta_db":-0.2565700127027862,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.00018483919130014926,"delta_db":-0.45851847985165195,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00017373067482784132,"delta_db":-0.4001256775070945,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":4.72882049936074e-05,"delta_db":0.36033836542630365,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2875904151279404,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15581104386034772,"pitch":null,"output_rms_dbfs":-27.743863191566497,"output_peak":0.4417002118673188,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.807372449133898},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.010339410142771,"spectral_convergence":0.36288799086597345,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.546588970332742,"spectral_convergence":0.34649700077674944,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.865188966926183,"spectral_convergence":0.3823329328254011,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.0003487641584913587,"delta_db":0.1384830342909462,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0005054794389001199,"delta_db":-0.11150692098886007,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.000421076264430183,"delta_db":-0.2565700127027862,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.00018483919130014926,"delta_db":-0.45851847985165195,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00017373067482784132,"delta_db":-0.4001256775070945,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":4.72882049936074e-05,"delta_db":0.36033836542630365,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2875904151279404,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15581104386034772,"pitch":null,"output_rms_dbfs":-27.743863191566497,"output_peak":0.4417002118673188,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.807372449133898},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8309637521027217,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04174432016039388,"rms_dbfs":-27.58805214770615,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.45916060622689,"mean":-0.0027310686111450197,"zero_energy":false},"output":{"rms":0.041002169856524176,"rms_dbfs":-27.743863191566497,"peak":0.4417002118673188,"peak_dbfs":-7.097447848529919,"crest_factor_db":20.646415343036576,"mean":-0.00019744498611070653,"zero_energy":false},"rms_difference_db":-0.1558110438603464},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":8.324866808132284e-07,"output_rms_dbfs":-60.79622706164695,"output_zero_energy":false,"output_peak":0.003165103543437723,"output_energy_fraction":4.9518059004961084e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":96,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.8571428571428571,"absolute_difference_cents_median":1.8143289224762016,"absolute_difference_cents_p95":9.560915124360747,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001220285373493014,"log_rms_mae_db":1.28575061553603,"rms_envelope_nrmse":0.09634029793985592,"envelope_correlation":0.9898893520170052,"rms_gain_error_db":-0.15573601080244687,"reference_mean_power":0.0017439355594579833,"output_mean_power":0.0016825068153564259},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.14297331662316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004136999026359357,"output_power":0.00040242472948789015,"reference_active":true,"delta_db":-0.12000757893785066},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.498679838909322e-05,"output_power":7.412802831469254e-05,"reference_active":true,"delta_db":-0.5936902012001348},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.720897303077822e-05,"output_power":1.6838022999287334e-05,"reference_active":true,"delta_db":-0.09463855585488068}],"active_band_count":3,"active_band_mae_db":0.2694454453309554}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011225274665232883,"log_rms_mae_db":0.6817010412349582,"rms_envelope_nrmse":0.05270830314053403,"envelope_correlation":0.9969393641667018,"rms_gain_error_db":-0.15623822462705697,"reference_mean_power":0.0017512823148994152,"output_mean_power":0.0016893994154099782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.129560410778096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037542654743209525,"output_power":0.0003658660029127302,"reference_active":true,"delta_db":-0.1120292376812668},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.989892754916159e-05,"output_power":3.535161151990718e-05,"reference_active":true,"delta_db":-0.525520062566712},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1683754663794268e-06,"output_power":9.014979352547245e-07,"reference_active":true,"delta_db":-1.1261769268132003}],"active_band_count":3,"active_band_mae_db":0.5879087423537264}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":1.6174118204433159,"rms_envelope_nrmse":0.13638603672841082,"envelope_correlation":0.9846412509739251,"rms_gain_error_db":-0.06890675947601098,"reference_mean_power":0.0005656706566344367,"output_mean_power":0.0005567663443338888},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3997327154946766e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018576385507894965,"output_power":0.00018033106156912048,"reference_active":true,"delta_db":-0.12890675926789494},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.516179020835253e-05,"output_power":3.251140726878484e-05,"reference_active":true,"delta_db":-0.340352098918921},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.50699704718427e-06,"output_power":8.901814999563805e-06,"reference_active":true,"delta_db":0.7401231934981447}],"active_band_count":3,"active_band_mae_db":0.4031273505616535}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":1.0697286201697045,"rms_envelope_nrmse":0.07721273249308538,"envelope_correlation":0.9948228535610475,"rms_gain_error_db":-0.0699742834609244,"reference_mean_power":0.0005675319915431874,"output_mean_power":0.0005584610894439476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.076532458167812e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017606095149491153,"output_power":0.00017259443366235012,"reference_active":true,"delta_db":-0.08636259418384659},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577844172032785e-05,"output_power":1.6227581321145717e-05,"reference_active":true,"delta_db":-0.3471181593832181},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.080515782562601e-07,"output_power":5.544698410193135e-07,"reference_active":true,"delta_db":0.3797012394764947}],"active_band_count":3,"active_band_mae_db":0.27106066434785314}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":1.5803618866763014,"rms_envelope_nrmse":0.1420001784502403,"envelope_correlation":0.9852713982976798,"rms_gain_error_db":-0.38713894788796244,"reference_mean_power":0.0002582156143082965,"output_mean_power":0.00023619386111998168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9030353181623002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.270493557085618e-05,"output_power":8.373113018919382e-05,"reference_active":true,"delta_db":-0.44215903127020445},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6852701179097834e-05,"output_power":1.3596896892380478e-05,"reference_active":true,"delta_db":-0.9322971600011605},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066518069212081e-06,"output_power":3.702646029573735e-06,"reference_active":true,"delta_db":-0.40710510610790784}],"active_band_count":3,"active_band_mae_db":0.5938537657930909}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":0.9875997204299581,"rms_envelope_nrmse":0.09094795225681208,"envelope_correlation":0.9944928223766928,"rms_gain_error_db":-0.38761588769042077,"reference_mean_power":0.0002594441947231778,"output_mean_power":0.00023729160217780837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.757465163940661e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.903046567125773e-05,"output_power":7.99379915863339e-05,"reference_active":true,"delta_db":-0.4678541238027787},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408774632836403e-06,"output_power":6.662395549908581e-06,"reference_active":true,"delta_db":-1.011022995127093},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6645360300144784e-07,"output_power":2.2653606908185292e-07,"reference_active":true,"delta_db":-0.7048423693354318}],"active_band_count":3,"active_band_mae_db":0.7279064960884346}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":1.8955109112450579,"rms_envelope_nrmse":0.18692674560126007,"envelope_correlation":0.9755299263983275,"rms_gain_error_db":-0.2718549233898603,"reference_mean_power":0.00014927695528659852,"output_mean_power":0.0001402191325620335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.071172575013412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0274139443533366e-05,"output_power":5.966810796834626e-05,"reference_active":true,"delta_db":-0.04388751603730059},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2776930907508548e-05,"output_power":1.0671916658866664e-05,"reference_active":true,"delta_db":-0.7818412129849784},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1605300538171703e-06,"output_power":2.03811621747366e-06,"reference_active":true,"delta_db":-0.25331367176470465}],"active_band_count":3,"active_band_mae_db":0.3596808002623279}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":1.1727331128693577,"rms_envelope_nrmse":0.10097550321260562,"envelope_correlation":0.992862955742515,"rms_gain_error_db":-0.27206187385494096,"reference_mean_power":0.00014998968981604915,"output_mean_power":0.00014088190632713276},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.934560164313176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8952201525784146e-05,"output_power":5.7967685317297834e-05,"reference_active":true,"delta_db":-0.07314069045002844},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.694634972457906e-06,"output_power":4.9106474066325255e-06,"reference_active":true,"delta_db":-0.6432713879958212},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8905823102195034e-07,"output_power":1.448699620098974e-07,"reference_active":true,"delta_db":-1.1561724374222873}],"active_band_count":3,"active_band_mae_db":0.6241948386227123}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.6047604084014893,"predicted_mos":2.6047604084014893,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c041ae59f7307934612774b0930676a5a82253a36563ef568edd5aa715a1bd01","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6301872780004487},"started_at":"2026-10-13T04:43:15.501716+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2107999939180445e-05,"native_input_samples":96000,"native_input_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","output_sha256":"452cedb48be484419dadc5ab605bc8880a55d2c846656148d7dec500f3164d56","output_samples":96000,"length_delta_samples":0,"analysis_s":1.698520070000086,"synthesis_s":0.0873325639995528,"total_s":1.7858526339996388,"rtf":0.2976421056666065,"output_peak":0.5784166230173017,"output_rms":0.047317435236720815,"finished_at":"2026-10-13T04:43:17.294195+00:00","evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"452cedb48be484419dadc5ab605bc8880a55d2c846656148d7dec500f3164d56","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.946962015393973,"spectral_convergence":0.6515601352047161,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2313933858095885,"spectral_convergence":0.32685001125821994,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.648904543095775,"spectral_convergence":0.3584908572688901,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.0003552985410393318,"delta_db":0.21909880245765728,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0007362419139085785,"delta_db":1.5216635116128194,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.0006749680269302758,"delta_db":1.7926543817985505,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.00022559121779964055,"delta_db":0.4067628095185493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00019855936620796725,"delta_db":0.18001303558051185,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":4.828061149547426e-05,"delta_db":0.4505376953155343,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7617883727139372,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0884760753081442,"pitch":null,"output_rms_dbfs":-26.499576072398007,"output_peak":0.5784166230173017,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.942419981433112},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.946962015393973,"spectral_convergence":0.6515601352047161,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2313933858095885,"spectral_convergence":0.32685001125821994,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.648904543095775,"spectral_convergence":0.3584908572688901,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.0003552985410393318,"delta_db":0.21909880245765728,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0007362419139085785,"delta_db":1.5216635116128194,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.0006749680269302758,"delta_db":1.7926543817985505,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.00022559121779964055,"delta_db":0.4067628095185493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00019855936620796725,"delta_db":0.18001303558051185,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":4.828061149547426e-05,"delta_db":0.4505376953155343,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7617883727139372,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0884760753081442,"pitch":null,"output_rms_dbfs":-26.499576072398007,"output_peak":0.5784166230173017,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.942419981433112},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.889335888551534,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04174432016039388,"rms_dbfs":-27.58805214770615,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.45916060622689,"mean":-0.0027310686111450197,"zero_energy":false},"output":{"rms":0.047317435236720815,"rms_dbfs":-26.499576072398007,"peak":0.5784166230173017,"peak_dbfs":-4.75518468837325,"crest_factor_db":21.74439138402476,"mean":4.0638427239543125e-07,"zero_energy":false},"rms_difference_db":1.0884760753081437},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":6.203668626663055e-07,"output_rms_dbfs":-62.07351408385067,"output_zero_energy":false,"output_peak":0.004762484373616432,"output_energy_fraction":2.7708065069082403e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":138,"both_estimated_voiced_frames":103,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9196428571428571,"absolute_difference_cents_median":7.438122590908951,"absolute_difference_cents_p95":24.166805154500285,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001220285373493014,"log_rms_mae_db":3.879589153656763,"rms_envelope_nrmse":0.2003416070032942,"envelope_correlation":0.9796625409809631,"rms_gain_error_db":1.0887386107911021,"reference_mean_power":0.0017439355594579833,"output_mean_power":0.0022408061835538402},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.14297331662316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004136999026359357,"output_power":0.0005850921099245304,"reference_active":true,"delta_db":1.5053882292414424},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.498679838909322e-05,"output_power":0.00011442495847777984,"reference_active":true,"delta_db":1.291692946371116},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.720897303077822e-05,"output_power":2.010771503537832e-05,"reference_active":true,"delta_db":0.6760776784151139}],"active_band_count":3,"active_band_mae_db":1.1577196180092242}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011225274665232883,"log_rms_mae_db":3.3763989151139393,"rms_envelope_nrmse":0.1615463898344579,"envelope_correlation":0.9924175192475195,"rms_gain_error_db":1.0897953418529076,"reference_mean_power":0.0017512823148994152,"output_mean_power":0.002250793727602648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.129560410778096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037542654743209525,"output_power":0.0005338584965654288,"reference_active":true,"delta_db":1.5290117956497276},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.989892754916159e-05,"output_power":5.5962738675610077e-05,"reference_active":true,"delta_db":1.4693773725905293},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1683754663794268e-06,"output_power":1.332129170402582e-06,"reference_active":true,"delta_db":0.5696390934798468}],"active_band_count":3,"active_band_mae_db":1.1893427539067012}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":2.1426688003426593,"rms_envelope_nrmse":0.259141556807034,"envelope_correlation":0.9854309448922746,"rms_gain_error_db":1.6675644301289785,"reference_mean_power":0.0005656706566344367,"output_mean_power":0.0008304626290859314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3997327154946766e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018576385507894965,"output_power":0.00028862729378402647,"reference_active":true,"delta_db":1.9137618215361059},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.516179020835253e-05,"output_power":5.0652514202831216e-05,"reference_active":true,"delta_db":1.5853002859891763},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.50699704718427e-06,"output_power":9.11010242970567e-06,"reference_active":true,"delta_db":0.8405701496303463}],"active_band_count":3,"active_band_mae_db":1.4465440857185428}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":1.8319093927180812,"rms_envelope_nrmse":0.23560917387062652,"envelope_correlation":0.9929434121413081,"rms_gain_error_db":1.6699132781124182,"reference_mean_power":0.0005675319915431874,"output_mean_power":0.000833646008994359},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.076532458167812e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017606095149491153,"output_power":0.00027327886060808795,"reference_active":true,"delta_db":1.9094299368648182},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577844172032785e-05,"output_power":2.555379454815266e-05,"reference_active":true,"delta_db":1.6248978863169956},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.080515782562601e-07,"output_power":5.683814206459422e-07,"reference_active":true,"delta_db":0.48732068358239955}],"active_band_count":3,"active_band_mae_db":1.3405495022547378}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":1.3834976609064773,"rms_envelope_nrmse":0.1625859121358975,"envelope_correlation":0.9819416105169599,"rms_gain_error_db":0.37879377232723044,"reference_mean_power":0.0002582156143082965,"output_mean_power":0.00028174867203314264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9030353181623002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.270493557085618e-05,"output_power":9.780087973401355e-05,"reference_active":true,"delta_db":0.23239904939845135},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6852701179097834e-05,"output_power":1.8059800969818683e-05,"reference_active":true,"delta_db":0.30043439578375214},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066518069212081e-06,"output_power":4.125425117909753e-06,"reference_active":true,"delta_db":0.06246002096345144}],"active_band_count":3,"active_band_mae_db":0.19843115538188497}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":0.8154725851317567,"rms_envelope_nrmse":0.07603438303092155,"envelope_correlation":0.9970401230659365,"rms_gain_error_db":0.3786724351625203,"reference_mean_power":0.0002594441947231778,"output_mean_power":0.00028308131277775294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.757465163940661e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.903046567125773e-05,"output_power":9.358966736962289e-05,"reference_active":true,"delta_db":0.21689258790743332},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408774632836403e-06,"output_power":9.144811520698213e-06,"reference_active":true,"delta_db":0.3644204594661183},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6645360300144784e-07,"output_power":2.7623360671907113e-07,"reference_active":true,"delta_db":0.15654916768894733}],"active_band_count":3,"active_band_mae_db":0.24595407168749964}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":1.8671853737366153,"rms_envelope_nrmse":0.23080094961935363,"envelope_correlation":0.9643945523824619,"rms_gain_error_db":0.24730349548197794,"reference_mean_power":0.00014927695528659852,"output_mean_power":0.00015802402447625231},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.071172575013412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0274139443533366e-05,"output_power":5.731077644602044e-05,"reference_active":true,"delta_db":-0.21894726192379704},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2776930907508548e-05,"output_power":1.1179939819316147e-05,"reference_active":true,"delta_db":-0.5798708052457229},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1605300538171703e-06,"output_power":2.0393119689782823e-06,"reference_active":true,"delta_db":-0.2507664372270235}],"active_band_count":3,"active_band_mae_db":0.3498615014655145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":1.1259408072183739,"rms_envelope_nrmse":0.1317737969667128,"envelope_correlation":0.9879748923442638,"rms_gain_error_db":0.24751418901017846,"reference_mean_power":0.00014998968981604915,"output_mean_power":0.00015878622573806307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.934560164313176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8952201525784146e-05,"output_power":5.520017370850173e-05,"reference_active":true,"delta_db":-0.2855958373051388},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.694634972457906e-06,"output_power":5.412657574266658e-06,"reference_active":true,"delta_db":-0.22055338116168088},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8905823102195034e-07,"output_power":1.408378623528995e-07,"reference_active":true,"delta_db":-1.2787616527904586}],"active_band_count":3,"active_band_mae_db":0.5949702904190928}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"facodec","status":"measured","mos":{"status":"measured","value":3.422826051712036,"predicted_mos":3.422826051712036,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"cd736d0d1a7c6b8268ccddf4f993df6f39dc24bcbd3630c08fd761721495010a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.828707532000408},"started_at":"2026-10-13T06:58:44.203057+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0002290009997523157,"native_input_samples":96000,"native_input_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","output_sha256":"6da5b8c36e6e10079aae97142cd57905fc421611d51633e37c6c15f11e6a9035","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7802718800012372,"synthesis_s":5.92078805500023,"total_s":7.701059935001467,"rtf":1.2835099891669113,"output_peak":0.4191438853740692,"output_rms":0.04035290644057678,"finished_at":"2026-10-13T06:58:51.918236+00:00","evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"6da5b8c36e6e10079aae97142cd57905fc421611d51633e37c6c15f11e6a9035","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.393522152951675,"spectral_convergence":0.2967199570061756,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.566400437995708,"spectral_convergence":0.3020076828972456,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.718656588294718,"spectral_convergence":0.3325684692103508,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.00030955671906069977,"delta_db":-0.37943308907601664,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0005233866550925391,"delta_db":0.039684583054355654,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.00043717864069940703,"delta_db":-0.09358827308268597,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.0001681835257292012,"delta_db":-0.868624551234075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00015016401085358652,"delta_db":-1.0332321465621193,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":3.988750676650897e-05,"delta_db":-0.3788210158666367,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.46556394314598154,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2944514425589731,"pitch":null,"output_rms_dbfs":-27.882503590265124,"output_peak":0.4191438853740692,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5595263930806995},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.393522152951675,"spectral_convergence":0.2967199570061756,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.566400437995708,"spectral_convergence":0.3020076828972456,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.718656588294718,"spectral_convergence":0.3325684692103508,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.00030955671906069977,"delta_db":-0.37943308907601664,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0005233866550925391,"delta_db":0.039684583054355654,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.00043717864069940703,"delta_db":-0.09358827308268597,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.0001681835257292012,"delta_db":-0.868624551234075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00015016401085358652,"delta_db":-1.0332321465621193,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":3.988750676650897e-05,"delta_db":-0.3788210158666367,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.46556394314598154,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2944514425589731,"pitch":null,"output_rms_dbfs":-27.882503590265124,"output_peak":0.4191438853740692,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5595263930806995},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8612815379902635,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04174432016039388,"rms_dbfs":-27.58805214770615,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.45916060622689,"mean":-0.0027310686111450197,"zero_energy":false},"output":{"rms":0.04035290644057678,"rms_dbfs":-27.882503590265124,"peak":0.4191438853740692,"peak_dbfs":-7.5527373021496995,"crest_factor_db":20.329766288115426,"mean":-0.0001357514471125114,"zero_energy":false},"rms_difference_db":-0.29445144255897304},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":2.0630616891782903e-06,"output_rms_dbfs":-56.85487785662734,"output_zero_energy":false,"output_peak":0.004758368246257305,"output_energy_fraction":1.2669590362793987e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":131,"both_estimated_voiced_frames":105,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.9375,"absolute_difference_cents_median":3.055303431537613,"absolute_difference_cents_p95":15.423320356422318,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001220285373493014,"log_rms_mae_db":2.705352742540728,"rms_envelope_nrmse":0.1254943869381212,"envelope_correlation":0.9832678431998221,"rms_gain_error_db":-0.29464951545407164,"reference_mean_power":0.0017439355594579833,"output_mean_power":0.0016295417115987133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.14297331662316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004136999026359357,"output_power":0.0003768314320520259,"reference_active":true,"delta_db":-0.4053829792511118},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.498679838909322e-05,"output_power":7.475385430162741e-05,"reference_active":true,"delta_db":-0.5571787911065178},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.720897303077822e-05,"output_power":1.6090867991268308e-05,"reference_active":true,"delta_db":-0.29175482057418606}],"active_band_count":3,"active_band_mae_db":0.4181055303106052}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011225274665232883,"log_rms_mae_db":2.3456773366551853,"rms_envelope_nrmse":0.11299791531184829,"envelope_correlation":0.9853628621623091,"rms_gain_error_db":-0.29751691076056985,"reference_mean_power":0.0017512823148994152,"output_mean_power":0.001635326486933123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.129560410778096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037542654743209525,"output_power":0.0003392777154496561,"reference_active":true,"delta_db":-0.43969644405923886},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.989892754916159e-05,"output_power":3.633530331640948e-05,"reference_active":true,"delta_db":-0.4063243245308755},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1683754663794268e-06,"output_power":1.1286803119574574e-06,"reference_active":true,"delta_db":-0.15011479524245686}],"active_band_count":3,"active_band_mae_db":0.3320451879441904}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":1.4775884656629306,"rms_envelope_nrmse":0.10228388890678598,"envelope_correlation":0.9914551866580227,"rms_gain_error_db":0.02043246598255749,"reference_mean_power":0.0005656706566344367,"output_mean_power":0.0005683382654748731},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3997327154946766e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018576385507894965,"output_power":0.00018121507870090618,"reference_active":true,"delta_db":-0.10766883223277365},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.516179020835253e-05,"output_power":3.455086450689853e-05,"reference_active":true,"delta_db":-0.07612059948245058},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.50699704718427e-06,"output_power":7.974525421295089e-06,"reference_active":true,"delta_db":0.262386017929789}],"active_band_count":3,"active_band_mae_db":0.14872514988167107}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":1.1314977832308033,"rms_envelope_nrmse":0.08219790660916453,"envelope_correlation":0.9941212036384571,"rms_gain_error_db":0.013895212086548177,"reference_mean_power":0.0005675319915431874,"output_mean_power":0.0005693507128806393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.076532458167812e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017606095149491153,"output_power":0.00017189976940541388,"reference_active":true,"delta_db":-0.10387750501739992},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577844172032785e-05,"output_power":1.7192541203446483e-05,"reference_active":true,"delta_db":-0.09625536282858978},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.080515782562601e-07,"output_power":5.093194658278495e-07,"reference_active":true,"delta_db":0.010824700865703835}],"active_band_count":3,"active_band_mae_db":0.07031918957056452}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":1.5461075075268023,"rms_envelope_nrmse":0.15790241050438814,"envelope_correlation":0.9867781041259287,"rms_gain_error_db":-0.8749453985359291,"reference_mean_power":0.0002582156143082965,"output_mean_power":0.00021109986748001305},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9030353181623002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.270493557085618e-05,"output_power":7.11136697438265e-05,"reference_active":true,"delta_db":-1.1514976589682142},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6852701179097834e-05,"output_power":1.451526752508789e-05,"reference_active":true,"delta_db":-0.6484447568229051},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066518069212081e-06,"output_power":3.478986733614731e-06,"reference_active":true,"delta_db":-0.6776993372948674}],"active_band_count":3,"active_band_mae_db":0.8258805843619955}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":1.20465661204798,"rms_envelope_nrmse":0.1357356517523034,"envelope_correlation":0.9919938505546184,"rms_gain_error_db":-0.8753999676483736,"reference_mean_power":0.0002594441947231778,"output_mean_power":0.00021208207344652137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.757465163940661e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.903046567125773e-05,"output_power":6.728706759042461e-05,"reference_active":true,"delta_db":-1.2160704303802052},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408774632836403e-06,"output_power":7.614651292372614e-06,"reference_active":true,"delta_db":-0.4308269300831957},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6645360300144784e-07,"output_power":2.6376880584613487e-07,"reference_active":true,"delta_db":-0.04398163998243992}],"active_band_count":3,"active_band_mae_db":0.563626333481947}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":2.0585825717395108,"rms_envelope_nrmse":0.20672177378539083,"envelope_correlation":0.9752534752615095,"rms_gain_error_db":-0.9540536113054496,"reference_mean_power":0.00014927695528659852,"output_mean_power":0.00011983602842453812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.071172575013412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0274139443533366e-05,"output_power":4.910729700833318e-05,"reference_active":true,"delta_db":-0.8898498840223947},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2776930907508548e-05,"output_power":1.1381744362207897e-05,"reference_active":true,"delta_db":-0.5021771932802088},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1605300538171703e-06,"output_power":1.8331654652437003e-06,"reference_active":true,"delta_db":-0.7135864486039023}],"active_band_count":3,"active_band_mae_db":0.7018711753021686}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":1.7327572107073068,"rms_envelope_nrmse":0.167718292329646,"envelope_correlation":0.9855796628248112,"rms_gain_error_db":-0.9541941154047378,"reference_mean_power":0.00014998968981604915,"output_mean_power":0.00012040429952514969},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.934560164313176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8952201525784146e-05,"output_power":4.824261413179033e-05,"reference_active":true,"delta_db":-0.8706919516773701},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.694634972457906e-06,"output_power":5.033231086293935e-06,"reference_active":true,"delta_db":-0.536190206872696},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8905823102195034e-07,"output_power":1.8505414964547754e-07,"reference_active":true,"delta_db":-0.09296761835463144}],"active_band_count":3,"active_band_mae_db":0.4999499256348992}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.3838820457458496,"predicted_mos":2.3838820457458496,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"6cc6d2da163b2188031d79e2f2cf4d833132f55231f8152cfc7b1d68ddd8d739","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.560289812001429},"started_at":"2026-10-13T07:28:11.583311+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009785957998246886,"native_input_samples":264600,"native_input_sha256":"3fa490ef63c3c9483e12f24049a6f3de8173cfb5b363ec91d93798e06a7bfaf4","output_sha256":"f54232b2ebe4d3faf8af4023e4bcce516dcbcf4604a3a245a055baa8d19cce90","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1474748680011544,"synthesis_s":4.159090470999217,"total_s":4.306565339000372,"rtf":0.7177608898333953,"output_peak":0.43363624811172485,"output_rms":0.03732922397197593,"finished_at":"2026-10-13T07:28:15.918997+00:00","evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"f54232b2ebe4d3faf8af4023e4bcce516dcbcf4604a3a245a055baa8d19cce90","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.554266243465896,"output_peak":0.4301283464196071},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.296112031438671,"spectral_convergence":0.5672800432968294,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.986469664374615,"spectral_convergence":0.5408123722329545,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.735896889029553,"spectral_convergence":0.5855385825088221,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.00040448923502298595,"delta_db":0.7822342398965999,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0003073088526114192,"delta_db":-2.272791157266615,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.0003137661591216959,"delta_db":-1.5341166125686123,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.0001876195004278112,"delta_db":-0.393679324935099,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00015208567384438247,"delta_db":-0.978007680129386,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":3.0216901192061354e-05,"delta_db":-1.5846906800462568,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2575866158070947,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9646920109648355,"pitch":null,"output_rms_dbfs":-28.552744158670986,"output_peak":0.4301283464196071,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.006159528280946},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.646054731955181,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04174432016039388,"rms_dbfs":-27.58805214770615,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.45916060622689,"mean":-0.0027310686111450197,"zero_energy":false},"output":{"rms":0.037356208616779574,"rms_dbfs":-28.552744158670986,"peak":0.4301283464196071,"peak_dbfs":-7.328038710510487,"crest_factor_db":21.2247054481605,"mean":-0.00015481356055698655,"zero_energy":false},"rms_difference_db":-0.9646920109648356},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":9.796937151415935e-07,"output_rms_dbfs":-60.08909677995902,"output_zero_energy":false,"output_peak":0.0033630917320314037,"output_energy_fraction":7.020446560757429e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":175,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":27,"reference_unvoiced_output_voiced_frames":90,"paired_voiced_fraction":0.7589285714285714,"absolute_difference_cents_median":12.318326153380426,"absolute_difference_cents_p95":29.789994984145817,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001220285373493014,"log_rms_mae_db":2.35772966478342,"rms_envelope_nrmse":0.1631728868821323,"envelope_correlation":0.9824816274539349,"rms_gain_error_db":-0.9647907264639468,"reference_mean_power":0.0017439355594579833,"output_mean_power":0.0013965335081451615},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.14297331662316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004136999026359357,"output_power":0.0003360578792098267,"reference_active":true,"delta_db":-0.9027133636409177},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.498679838909322e-05,"output_power":5.4356013751138695e-05,"reference_active":true,"delta_db":-1.9410386902845045},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.720897303077822e-05,"output_power":1.2930960706872631e-05,"reference_active":true,"delta_db":-1.2412416195201212}],"active_band_count":3,"active_band_mae_db":1.3616645578151811}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011225274665232883,"log_rms_mae_db":2.1019973748754595,"rms_envelope_nrmse":0.1382443326692994,"envelope_correlation":0.9907705655296362,"rms_gain_error_db":-0.9638081754648395,"reference_mean_power":0.0017512823148994152,"output_mean_power":0.0014027340670854122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.129560410778096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037542654743209525,"output_power":0.00030358829166496153,"reference_active":true,"delta_db":-0.922399611827034},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.989892754916159e-05,"output_power":2.6280812576184242e-05,"reference_active":true,"delta_db":-1.8132243331778177},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1683754663794268e-06,"output_power":8.151707958975185e-07,"reference_active":true,"delta_db":-1.5633381670181374}],"active_band_count":3,"active_band_mae_db":1.43298737067433}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":3.4213078769325813,"rms_envelope_nrmse":0.26930814153651594,"envelope_correlation":0.9451490208093256,"rms_gain_error_db":-1.0760791890571457,"reference_mean_power":0.0005656706566344367,"output_mean_power":0.0004415254400087842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3997327154946766e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018576385507894965,"output_power":0.00011137073914445922,"reference_active":true,"delta_db":-2.2219011322560593},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.516179020835253e-05,"output_power":1.971942716779086e-05,"reference_active":true,"delta_db":-2.511766834873553},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.50699704718427e-06,"output_power":4.486934321284907e-06,"reference_active":true,"delta_db":-2.23516532563625}],"active_band_count":3,"active_band_mae_db":2.3229444309219542}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":2.8061821960814144,"rms_envelope_nrmse":0.24165827945502308,"envelope_correlation":0.9550581621538523,"rms_gain_error_db":-1.0715259918659532,"reference_mean_power":0.0005675319915431874,"output_mean_power":0.0004434429435005583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.076532458167812e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017606095149491153,"output_power":0.00010382020704813925,"reference_active":true,"delta_db":-2.2938115395066965},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577844172032785e-05,"output_power":9.887738403210648e-06,"reference_active":true,"delta_db":-2.4986864221929714},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.080515782562601e-07,"output_power":2.6038186194493453e-07,"reference_active":true,"delta_db":-2.902970765943469}],"active_band_count":3,"active_band_mae_db":2.565156242547712}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":3.5536063240384683,"rms_envelope_nrmse":0.22799210392323802,"envelope_correlation":0.964309957648087,"rms_gain_error_db":-0.039233014229272,"reference_mean_power":0.0002582156143082965,"output_mean_power":0.0002558934674023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9030353181623002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.270493557085618e-05,"output_power":0.0001046807859461834,"reference_active":true,"delta_db":0.527641182636015},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6852701179097834e-05,"output_power":1.4828781528351216e-05,"reference_active":true,"delta_db":-0.5556405344147931},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066518069212081e-06,"output_power":3.105223064564425e-06,"reference_active":true,"delta_db":-1.1712990309106848}],"active_band_count":3,"active_band_mae_db":0.7515269159871641}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":2.975988065164568,"rms_envelope_nrmse":0.17295420688536153,"envelope_correlation":0.9795798304710144,"rms_gain_error_db":-0.038805739440926836,"reference_mean_power":0.0002594441947231778,"output_mean_power":0.0002571362958918595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.757465163940661e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.903046567125773e-05,"output_power":9.940567315974226e-05,"reference_active":true,"delta_db":0.4787252568463083},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408774632836403e-06,"output_power":7.144878302068991e-06,"reference_active":true,"delta_db":-0.707378770243047},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6645360300144784e-07,"output_power":2.415062294773329e-07,"reference_active":true,"delta_db":-0.42693259630484404}],"active_band_count":3,"active_band_mae_db":0.5376788744647331}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":4.093092271538797,"rms_envelope_nrmse":0.3342627630253977,"envelope_correlation":0.9521560181396829,"rms_gain_error_db":-2.389785165523433,"reference_mean_power":0.00014927695528659852,"output_mean_power":8.610220071898315e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.071172575013412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0274139443533366e-05,"output_power":4.0403269002177486e-05,"reference_active":true,"delta_db":-1.7371451349291456},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2776930907508548e-05,"output_power":7.879157389810796e-06,"reference_active":true,"delta_db":-2.0994677051134305},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1605300538171703e-06,"output_power":1.0886008980315422e-06,"reference_active":true,"delta_db":-2.976916236466513}],"active_band_count":3,"active_band_mae_db":2.2711763588363634}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":3.825369111414072,"rms_envelope_nrmse":0.3001313682035756,"envelope_correlation":0.9697886317849519,"rms_gain_error_db":-2.3895655938368803,"reference_mean_power":0.00014998968981604915,"output_mean_power":8.651767650020657e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.934560164313176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8952201525784146e-05,"output_power":3.919410938527096e-05,"reference_active":true,"delta_db":-1.7727922779163507},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.694634972457906e-06,"output_power":3.194633934309066e-06,"reference_active":true,"delta_db":-2.5104479037286165},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8905823102195034e-07,"output_power":1.1132863438660058e-07,"reference_active":true,"delta_db":-2.2998870814955144}],"active_band_count":3,"active_band_mae_db":2.1943757543801605}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"golf","status":"measured","mos":{"status":"measured","value":2.3312952518463135,"predicted_mos":2.3312952518463135,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"61bb10773e376ec692137c65af142144bcb8e3733c40ee5470e8300c6eb1a7c9","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0144663709997985},"started_at":"2026-10-13T05:02:57.374875+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002947559000858746,"native_input_samples":144000,"native_input_sha256":"a9f58857d50e8a8a26fbf99818441ff142a38f16bce78ba3cecab264a459b957","output_sha256":"2b338e9c18bac8aedb298212053b131085a723ae6ec2fddedb06817b6e375c44","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5471639299994422,"synthesis_s":0.11703607800063764,"total_s":0.6642000080000798,"rtf":0.11070000133334663,"output_peak":0.4186374843120575,"output_rms":0.03827575906460516,"finished_at":"2026-10-13T05:02:58.049932+00:00","analysis_stages":{"external_f0_s":0.0938100140001552},"evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"2b338e9c18bac8aedb298212053b131085a723ae6ec2fddedb06817b6e375c44","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.338232214575743,"output_peak":0.4306910928831607},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.025882928479469,"spectral_convergence":0.6091485429009729,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.358341333190809,"spectral_convergence":0.32694833374289156,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.65669450960495,"spectral_convergence":0.3654811065799044,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00033824122780413485,"output_power":0.0003172549813024558,"delta_db":-0.2781809231418909,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005193208344850619,"output_power":0.0004714465536345912,"delta_db":-0.4200328097325827,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00044705981190819213,"output_power":0.0003403359693034752,"delta_db":-1.1845777998234084,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00020575139850446082,"output_power":0.00016280207730038737,"delta_db":-1.0168285377793924,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001907970370285226,"output_power":0.00013529252839276134,"delta_db":-1.4929781297079168,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.359513980610697e-05,"output_power":3.901240023531184e-05,"delta_db":-0.4823540367654289,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.81249203949177,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.755602740601866,"pitch":null,"output_rms_dbfs":-28.338232214575743,"output_peak":0.4306910928831607,"reference_rms_dbfs":-27.582629473973878,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.013639590425076},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8347011185652757,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04177038962634271,"rms_dbfs":-27.582629473973874,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.453737932494622,"mean":-0.002728297913412817,"zero_energy":false},"output":{"rms":0.038290266519952335,"rms_dbfs":-28.338232214575743,"peak":0.4306910928831607,"peak_dbfs":-7.316682195660139,"crest_factor_db":21.021550018915605,"mean":0.00037097339619000433,"zero_energy":false},"rms_difference_db":-0.7556027406018693},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":2.3976298281303418e-06,"output_rms_dbfs":-56.20217867112865,"output_zero_energy":false,"output_peak":0.004114280929781708,"output_energy_fraction":1.6380599131859205e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":117,"both_estimated_voiced_frames":92,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.8214285714285714,"absolute_difference_cents_median":8.695413444781678,"absolute_difference_cents_p95":31.989677936335735,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001220285373493014,"log_rms_mae_db":2.4264165332762473,"rms_envelope_nrmse":0.15104089449980745,"envelope_correlation":0.9815736481607813,"rms_gain_error_db":-0.7558661794923216,"reference_mean_power":0.0017461178280207357,"output_mean_power":0.00146719192667878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.144964997289727e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343440472553885,"output_power":0.00035514933831387065,"reference_active":true,"delta_db":-0.6599560456567051},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.503360060606474e-05,"output_power":7.827507826388626e-05,"reference_active":true,"delta_db":-0.3596705842019921},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.718676313482109e-05,"output_power":1.4175332418263711e-05,"reference_active":true,"delta_db":-0.8366083953868779}],"active_band_count":3,"active_band_mae_db":0.6187450084151916}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00011225274665232883,"log_rms_mae_db":1.7049155950861818,"rms_envelope_nrmse":0.111120565306471,"envelope_correlation":0.9931955066647697,"rms_gain_error_db":-0.7549202189956186,"reference_mean_power":0.0017531213854547272,"output_mean_power":0.0014733976255023086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.132284993487634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003751449521962352,"output_power":0.000321306643520548,"reference_active":true,"delta_db":-0.6727940176250601},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9944943956318604e-05,"output_power":3.854105781853288e-05,"reference_active":true,"delta_db":-0.15538185807663213},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1666284542264562e-06,"output_power":9.482705348009328e-07,"reference_active":true,"delta_db":-0.9000030859976699}],"active_band_count":3,"active_band_mae_db":0.5760596538997874}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":1.910585693204052,"rms_envelope_nrmse":0.17481988305344867,"envelope_correlation":0.978386313335775,"rms_gain_error_db":-0.7185925975947387,"reference_mean_power":0.0005661181313916627,"output_mean_power":0.00047978625796215023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.3988286384590637e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001858549145768432,"output_power":0.00014286604275552943,"reference_active":true,"delta_db":-1.1424503428811084},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.508765448985161e-05,"output_power":2.797872499656257e-05,"reference_active":true,"delta_db":-0.9832641822545501},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.501910471777156e-06,"output_power":6.589880125654018e-06,"reference_active":true,"delta_db":-0.5629436239326248}],"active_band_count":3,"active_band_mae_db":0.896219383022761}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":1.3508507680144397,"rms_envelope_nrmse":0.13602688440141314,"envelope_correlation":0.9882668035832358,"rms_gain_error_db":-0.714924245108597,"reference_mean_power":0.0005677156328664187,"output_mean_power":0.00048154671971491523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.076191400447036e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001761527325734457,"output_power":0.0001353618793316878,"reference_active":true,"delta_db":-1.1439300919632103},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7491215234323364e-05,"output_power":1.4206779774525862e-05,"reference_active":true,"delta_db":-0.9032433560857367},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.070906611517388e-07,"output_power":3.9994682250092686e-07,"reference_active":true,"delta_db":-1.0308336167224539}],"active_band_count":3,"active_band_mae_db":1.0260023549238004}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":1.9256563963419546,"rms_envelope_nrmse":0.19394545624612708,"envelope_correlation":0.979188857337072,"rms_gain_error_db":-1.0653694734186268,"reference_mean_power":0.0002586283219726026,"output_mean_power":0.00020236673961159074},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.899756863502067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.25131855112258e-05,"output_power":7.153064106260383e-05,"reference_active":true,"delta_db":-1.1171151805823512},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.685107163243189e-05,"output_power":1.3962355900405437e-05,"reference_active":true,"delta_db":-0.8166882077122866},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.061605684902155e-06,"output_power":3.175577254961827e-06,"reference_active":true,"delta_db":-1.0687507561144618}],"active_band_count":3,"active_band_mae_db":1.0008513814697}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":1.4419237590744451,"rms_envelope_nrmse":0.1418665437316201,"envelope_correlation":0.9938557051015917,"rms_gain_error_db":-1.0644160067101218,"reference_mean_power":0.00025977609970183053,"output_mean_power":0.00020330946208717304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.7544975048841952e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.883932271429609e-05,"output_power":6.900357664365622e-05,"reference_active":true,"delta_db":-1.097336369629147},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.409761804986657e-06,"output_power":6.950226983839292e-06,"reference_active":true,"delta_db":-0.8278470695549165},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.658344605530017e-07,"output_power":2.3938899301184494e-07,"reference_active":true,"delta_db":-0.4550710067251936}],"active_band_count":3,"active_band_mae_db":0.793418148636419}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":2.428797855332907,"rms_envelope_nrmse":0.2274064866225497,"envelope_correlation":0.9753022688524245,"rms_gain_error_db":-1.352704691114512,"reference_mean_power":0.00014951016619371758,"output_mean_power":0.00010949650447192423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0729524628063423e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.023318608226086e-05,"output_power":4.8813958053611306e-05,"reference_active":true,"delta_db":-0.9129181251059957},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2777017198178084e-05,"output_power":1.0316824797340828e-05,"reference_active":true,"delta_db":-0.9288342404792981},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.162101527603024e-06,"output_power":1.1611145629161954e-06,"reference_active":true,"delta_db":-2.700010115273389}],"active_band_count":3,"active_band_mae_db":1.5139208269528943}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":2.051337129290308,"rms_envelope_nrmse":0.1826598510208038,"envelope_correlation":0.989310384127904,"rms_gain_error_db":-1.3524793506913904,"reference_mean_power":0.00015020653063552844,"output_mean_power":0.00011001220771190215},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.952904897645382e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.887337101923156e-05,"output_power":4.7372614869131054e-05,"reference_active":true,"delta_db":-0.9439154617437104},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.690080330140722e-06,"output_power":4.749045919311862e-06,"reference_active":true,"delta_db":-0.7851202875602125},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8914503004453116e-07,"output_power":1.0792608263859197e-07,"reference_active":true,"delta_db":-2.4366852041418334}],"active_band_count":3,"active_band_mae_db":1.3885736511485856}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.4499659538269043,"predicted_mos":2.4499659538269043,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"6eba660be87039b65fa4b2bf8bc81852d5ff4295f7236cd6ffe3120dd168e4b6","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5464125010003045},"started_at":"2026-10-13T07:44:22.347970+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003198212998540839,"native_input_samples":144000,"native_input_sha256":"a9f58857d50e8a8a26fbf99818441ff142a38f16bce78ba3cecab264a459b957","output_sha256":"6a688c39ac11e2b6b2be953b07a4f6f8c07e15f87fe7001c32fbf43eedf2d420","output_samples":144120,"length_delta_samples":120,"analysis_s":2.4284815450009773,"synthesis_s":3.327307782998105,"total_s":5.755789327999082,"rtf":0.9592982213331803,"output_peak":0.46753597259521484,"output_rms":0.04214459816434149,"finished_at":"2026-10-13T07:44:28.126289+00:00","analysis_stages":{"external_f0_s":1.588920952999615},"evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"6a688c39ac11e2b6b2be953b07a4f6f8c07e15f87fe7001c32fbf43eedf2d420","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.501255827819534,"output_peak":0.466384268163452},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.006364197275957,"spectral_convergence":0.6329835344104444,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.30990803215763,"spectral_convergence":0.33491370736712234,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.401687091568187,"spectral_convergence":0.34359726818927055,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.0004295201292571891,"delta_db":1.0429997774324669,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.00048351897047842234,"delta_db":-0.30440668886903044,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.0004517505416727416,"delta_db":0.04880919259860602,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.00019992632161488236,"delta_db":-0.11775932124267174,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00018014563789451931,"delta_db":-0.24265323914457074,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":3.430382725728986e-05,"delta_db":-1.0337641574284642,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4650653961193017,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.09021545768193379,"pitch":null,"output_rms_dbfs":-27.49783669002422,"output_peak":0.466384268163452,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9059864403339235},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8604543399676771,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04174432016039388,"rms_dbfs":-27.58805214770615,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.45916060622689,"mean":-0.0027310686111450197,"zero_energy":false},"output":{"rms":0.042180154435173,"rms_dbfs":-27.49783669002422,"peak":0.466384268163452,"peak_dbfs":-6.625122148043997,"crest_factor_db":20.872714541980223,"mean":6.11713308818256e-05,"zero_energy":false},"rms_difference_db":0.09021545768193207},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":3.9290501198582363e-07,"output_rms_dbfs":-64.0571243119353,"output_zero_energy":false,"output_peak":0.0032990378040031842,"output_energy_fraction":2.208366944207323e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":129,"both_estimated_voiced_frames":102,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9107142857142857,"absolute_difference_cents_median":6.683262954817282,"absolute_difference_cents_p95":30.803903390300984,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001220285373493014,"log_rms_mae_db":4.4081799955379335,"rms_envelope_nrmse":0.17603686754763323,"envelope_correlation":0.9677070736426457,"rms_gain_error_db":0.09031672058002652,"reference_mean_power":0.0017439355594579833,"output_mean_power":0.0017805825185826887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.14297331662316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004136999026359357,"output_power":0.00043750663692674203,"reference_active":true,"delta_db":0.2429922692247727},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.498679838909322e-05,"output_power":7.51006550486713e-05,"reference_active":true,"delta_db":-0.5370774380944484},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.720897303077822e-05,"output_power":1.6269166848612003e-05,"reference_active":true,"delta_db":-0.24389640888528655}],"active_band_count":3,"active_band_mae_db":0.3413220387348359}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011225274665232883,"log_rms_mae_db":3.74189550840766,"rms_envelope_nrmse":0.10154392745204233,"envelope_correlation":0.9888419893244943,"rms_gain_error_db":0.09138334400923512,"reference_mean_power":0.0017512823148994152,"output_mean_power":0.001788522864016672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.129560410778096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037542654743209525,"output_power":0.00039772360524736595,"reference_active":true,"delta_db":0.2505638799712069},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.989892754916159e-05,"output_power":3.4123415518089884e-05,"reference_active":true,"delta_db":-0.6790872781223573},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1683754663794268e-06,"output_power":1.0791311763309006e-06,"reference_active":true,"delta_db":-0.345081894747114}],"active_band_count":3,"active_band_mae_db":0.42491101761355937}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":2.137884713353796,"rms_envelope_nrmse":0.1784213236232643,"envelope_correlation":0.9742449629303852,"rms_gain_error_db":0.00011788791881635674,"reference_mean_power":0.0005656706566344367,"output_mean_power":0.0005656860118011071},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3997327154946766e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018576385507894965,"output_power":0.0001923531011090396,"reference_active":true,"delta_db":0.1513797715936199},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.516179020835253e-05,"output_power":3.398336481459628e-05,"reference_active":true,"delta_db":-0.148046006855391},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.50699704718427e-06,"output_power":7.214189211250123e-06,"reference_active":true,"delta_db":-0.17278716383625936}],"active_band_count":3,"active_band_mae_db":0.15740431409509006}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":1.610788139517291,"rms_envelope_nrmse":0.11270149134914761,"envelope_correlation":0.9890724608289366,"rms_gain_error_db":0.0026513077113316744,"reference_mean_power":0.0005675319915431874,"output_mean_power":0.0005678785677493475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.076532458167812e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017606095149491153,"output_power":0.00018281615508698688,"reference_active":true,"delta_db":0.16351526196464836},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577844172032785e-05,"output_power":1.6932756709888926e-05,"reference_active":true,"delta_db":-0.1623794165221491},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.080515782562601e-07,"output_power":5.236990353095478e-07,"reference_active":true,"delta_db":0.1317396906813156}],"active_band_count":3,"active_band_mae_db":0.15254478972270435}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":2.075684419263029,"rms_envelope_nrmse":0.21960171224273708,"envelope_correlation":0.9625788973704142,"rms_gain_error_db":-0.09885225761747152,"reference_mean_power":0.0002582156143082965,"output_mean_power":0.00025240460536601005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9030353181623002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.270493557085618e-05,"output_power":9.144130756429653e-05,"reference_active":true,"delta_db":-0.059604287889089294},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6852701179097834e-05,"output_power":1.5981094237084033e-05,"reference_active":true,"delta_db":-0.23063007778145317},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066518069212081e-06,"output_power":3.6562551165464e-06,"reference_active":true,"delta_db":-0.4618621521307644}],"active_band_count":3,"active_band_mae_db":0.2506988392671023}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":1.3715985138083193,"rms_envelope_nrmse":0.11517373315674173,"envelope_correlation":0.989024775636286,"rms_gain_error_db":-0.09846267633059566,"reference_mean_power":0.0002594441947231778,"output_mean_power":0.00025362828776820574},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.757465163940661e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.903046567125773e-05,"output_power":8.773963458618309e-05,"reference_active":true,"delta_db":-0.06342823602833159},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408774632836403e-06,"output_power":7.843145557853106e-06,"reference_active":true,"delta_db":-0.30242437938529815},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6645360300144784e-07,"output_power":2.42723424596514e-07,"reference_active":true,"delta_db":-0.40509906204727164}],"active_band_count":3,"active_band_mae_db":0.2569838924869671}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":2.2854831701412173,"rms_envelope_nrmse":0.25051394997073145,"envelope_correlation":0.9559040485042192,"rms_gain_error_db":-0.5113489866111907,"reference_mean_power":0.00014927695528659852,"output_mean_power":0.0001326960116518162},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.071172575013412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0274139443533366e-05,"output_power":5.354463138535892e-05,"reference_active":true,"delta_db":-0.5141508538749779},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2776930907508548e-05,"output_power":1.1066130652646896e-05,"reference_active":true,"delta_db":-0.6243075288935045},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1605300538171703e-06,"output_power":1.777814431332573e-06,"reference_active":true,"delta_db":-0.8467388465031321}],"active_band_count":3,"active_band_mae_db":0.6617324097572048}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":1.3752487493607302,"rms_envelope_nrmse":0.14840019086597883,"envelope_correlation":0.9850993583106376,"rms_gain_error_db":-0.5111219657288388,"reference_mean_power":0.00014998968981604915,"output_mean_power":0.00013333654895017233},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.934560164313176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8952201525784146e-05,"output_power":5.198531663352681e-05,"reference_active":true,"delta_db":-0.5461933460754772},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.694634972457906e-06,"output_power":4.950070111175662e-06,"reference_active":true,"delta_db":-0.6085454073889691},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8905823102195034e-07,"output_power":1.6124428206718364e-07,"reference_active":true,"delta_db":-0.6911126704378032}],"active_band_count":3,"active_band_mae_db":0.6152838079674166}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.2419095039367676,"predicted_mos":2.2419095039367676,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"68abf97728acae9ad28b9f436b05f17c2a9020d27dec1ae0f4b9f9dcf35e7b80","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0576651739993395},"started_at":"2026-10-13T05:11:18.930178+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002904941999076982,"native_input_samples":144000,"native_input_sha256":"a9f58857d50e8a8a26fbf99818441ff142a38f16bce78ba3cecab264a459b957","output_sha256":"857e41c56bef08f2cb768e38be8cd55930e57c3e093bff9ade6ad27e84cd120c","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0892303830005403,"synthesis_s":0.13533319899943308,"total_s":1.2245635819999734,"rtf":0.2040939303333289,"output_peak":0.493619441986084,"output_rms":0.04070325455513076,"finished_at":"2026-10-13T05:11:20.166467+00:00","analysis_stages":{"external_f0_s":0.1655301499995403},"evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"857e41c56bef08f2cb768e38be8cd55930e57c3e093bff9ade6ad27e84cd120c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.80337289960224,"output_peak":0.49388953104007677},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.93354150320596,"spectral_convergence":0.23610625532319057,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.453083479936042,"spectral_convergence":0.21708225541617676,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.924605575425696,"spectral_convergence":0.25063679423012164,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.0003280804464406988,"delta_db":-0.12703196652936716,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.0005070935040956127,"delta_db":-0.09766139595496638,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.00042109714794267026,"delta_db":-0.2563546272701759,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.00019311199129422942,"delta_db":-0.268366655868343,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.00017066714698222852,"delta_db":-0.4773914584084271,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":4.098887868292487e-05,"delta_db":-0.2605295481087647,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24788927535667402,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2081492555401776,"pitch":null,"output_rms_dbfs":-27.79620140324633,"output_peak":0.49388953104007677,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.770410186189233},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9093944443218756,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04174432016039388,"rms_dbfs":-27.58805214770615,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.45916060622689,"mean":-0.0027310686111450197,"zero_energy":false},"output":{"rms":0.04075584762262177,"rms_dbfs":-27.79620140324633,"peak":0.49388953104007677,"peak_dbfs":-6.1274035893308,"crest_factor_db":21.668797813915532,"mean":-3.028720011165507e-05,"zero_energy":false},"rms_difference_db":-0.20814925554017805},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":1.0372255894926047e-06,"output_rms_dbfs":-59.84126777249444,"output_zero_energy":false,"output_peak":0.0034039248716513652,"output_energy_fraction":6.244438074047826e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":92,"both_estimated_voiced_frames":86,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.7678571428571429,"absolute_difference_cents_median":1.009588817937582,"absolute_difference_cents_p95":6.504495460574661,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001220285373493014,"log_rms_mae_db":1.2217135454474932,"rms_envelope_nrmse":0.07562358002133243,"envelope_correlation":0.9941664081456639,"rms_gain_error_db":-0.20832919281672205,"reference_mean_power":0.0017439355594579833,"output_mean_power":0.0016622544868677752},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.14297331662316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004136999026359357,"output_power":0.00039975624840666885,"reference_active":true,"delta_db":-0.14890157955246938},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.498679838909322e-05,"output_power":7.62832448322162e-05,"reference_active":true,"delta_db":-0.46922310675762513},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.720897303077822e-05,"output_power":1.2521189406556006e-05,"reference_active":true,"delta_db":-1.3810936886016365}],"active_band_count":3,"active_band_mae_db":0.6664061249705769}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011225274665232883,"log_rms_mae_db":0.7274397121345534,"rms_envelope_nrmse":0.046465890410297826,"envelope_correlation":0.9980074241050189,"rms_gain_error_db":-0.20778062895365446,"reference_mean_power":0.0017512823148994152,"output_mean_power":0.0016694680004819102},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.129560410778096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037542654743209525,"output_power":0.0003615544288876909,"reference_active":true,"delta_db":-0.1635129369017864},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.989892754916159e-05,"output_power":3.59014105324991e-05,"reference_active":true,"delta_db":-0.4584971042652872},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1683754663794268e-06,"output_power":7.996919471821698e-07,"reference_active":true,"delta_db":-1.6465970631742612}],"active_band_count":3,"active_band_mae_db":0.7562023681137783}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":1.0782737599307164,"rms_envelope_nrmse":0.08044679368829179,"envelope_correlation":0.9951753451512265,"rms_gain_error_db":-0.24103421967871413,"reference_mean_power":0.0005656706566344367,"output_mean_power":0.0005351311462294464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3997327154946766e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018576385507894965,"output_power":0.0001730871717017347,"reference_active":true,"delta_db":-0.30696333741481985},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.516179020835253e-05,"output_power":3.0093278765440464e-05,"reference_active":true,"delta_db":-0.6760147021460029},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.50699704718427e-06,"output_power":6.06030297140149e-06,"reference_active":true,"delta_db":-0.9297190874214595}],"active_band_count":3,"active_band_mae_db":0.637565708994094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":0.6505988629802913,"rms_envelope_nrmse":0.050335050901951006,"envelope_correlation":0.9983830200686453,"rms_gain_error_db":-0.23987366096468277,"reference_mean_power":0.0005675319915431874,"output_mean_power":0.000537035483139666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.076532458167812e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017606095149491153,"output_power":0.00016352123973535312,"reference_active":true,"delta_db":-0.3208887353750324},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577844172032785e-05,"output_power":1.5106274457049275e-05,"reference_active":true,"delta_db":-0.6580823924852576},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.080515782562601e-07,"output_power":3.993709174845353e-07,"reference_active":true,"delta_db":-1.0453136882485004}],"active_band_count":3,"active_band_mae_db":0.6747616053695967}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":1.0722721663010588,"rms_envelope_nrmse":0.09832506600448404,"envelope_correlation":0.9932596009424638,"rms_gain_error_db":-0.30666945735382245,"reference_mean_power":0.0002582156143082965,"output_mean_power":0.0002406110425041421},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9030353181623002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.270493557085618e-05,"output_power":8.800664537715058e-05,"reference_active":true,"delta_db":-0.2258738946947304},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6852701179097834e-05,"output_power":1.470177068998455e-05,"reference_active":true,"delta_db":-0.5929987564814856},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066518069212081e-06,"output_power":3.330867790252474e-06,"reference_active":true,"delta_db":-0.866653114957106}],"active_band_count":3,"active_band_mae_db":0.5618419220444406}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":0.6481285574147809,"rms_envelope_nrmse":0.060285537946137065,"envelope_correlation":0.9979375873347106,"rms_gain_error_db":-0.30654198022217666,"reference_mean_power":0.0002594441947231778,"output_mean_power":0.00024176295730259946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.757465163940661e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.903046567125773e-05,"output_power":8.429350110309552e-05,"reference_active":true,"delta_db":-0.2374455245559728},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408774632836403e-06,"output_power":7.314993494240178e-06,"reference_active":true,"delta_db":-0.6051876871227073},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6645360300144784e-07,"output_power":2.2946535114808322e-07,"reference_active":true,"delta_db":-0.6490448002780992}],"active_band_count":3,"active_band_mae_db":0.49722600398559313}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":1.3680207219849019,"rms_envelope_nrmse":0.16535110791327298,"envelope_correlation":0.98158981478235,"rms_gain_error_db":-0.39981196781208195,"reference_mean_power":0.00014927695528659852,"output_mean_power":0.0001361480958293135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.071172575013412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0274139443533366e-05,"output_power":6.21899350193171e-05,"reference_active":true,"delta_db":0.13589084521734798},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2776930907508548e-05,"output_power":1.3575604552969468e-05,"reference_active":true,"delta_db":0.2633263251579362},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1605300538171703e-06,"output_power":1.5489212847249776e-06,"reference_active":true,"delta_db":-1.4453096416711053}],"active_band_count":3,"active_band_mae_db":0.6148422706821298}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":0.91914970958944,"rms_envelope_nrmse":0.10082711926171999,"envelope_correlation":0.9936703335480991,"rms_gain_error_db":-0.39934226249153326,"reference_mean_power":0.00014998968981604915,"output_mean_power":0.00013681294160815282},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.934560164313176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8952201525784146e-05,"output_power":6.109972412034323e-05,"reference_active":true,"delta_db":0.1553922116913262},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.694634972457906e-06,"output_power":6.3128191944815695e-06,"reference_active":true,"delta_db":0.4475745995165946},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8905823102195034e-07,"output_power":1.2239821523651219e-07,"reference_active":true,"delta_db":-1.8882050483329418}],"active_band_count":3,"active_band_mae_db":0.8303906198469542}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75946-0029","method":"magphase","status":"measured","mos":{"status":"measured","value":2.1846070289611816,"predicted_mos":2.1846070289611816,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288166,"input_sample_rate":48000,"input_seconds":6.003458333333334,"input_float32_sha256":"b1ed0f423d20d74da31debb013c218b4521ffba98750f9d470d8bb04be6b93c8","internal_samples":96056,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.737391867000042},"started_at":"2026-10-13T05:17:51.732426+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008693154999491526,"native_input_samples":288000,"native_input_sha256":"cabeb66ac0427ef13929137747765f2ea0082ce08842a17fe3cc06c15f538f23","output_sha256":"6c7032b5d1763e67f947845a1935a4ed27a19701c69a2c10513734cd407849b7","output_samples":288166,"length_delta_samples":166,"analysis_s":2.951384128000427,"synthesis_s":0.9534649480010557,"total_s":3.904849076001483,"rtf":0.6508081793335805,"output_peak":0.5332150263812707,"output_rms":0.04047025900465068,"finished_at":"2026-10-13T05:17:55.659369+00:00","evaluation_input":{"reference":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"6c7032b5d1763e67f947845a1935a4ed27a19701c69a2c10513734cd407849b7","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96056,"length_delta_samples":56,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.853372942137177,"output_peak":0.5031802887105841},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9994170067460648,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.032688932620246,"spectral_convergence":0.6698396059351958,"spectral_convergence_status":"ok","amplitude_floor":1.016885378852613e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.669143802890835,"spectral_convergence":0.410361463023887,"spectral_convergence_status":"ok","amplitude_floor":6.009137305261044e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.349844008300717,"spectral_convergence":0.37834088784056696,"spectral_convergence_status":"ok","amplitude_floor":5.349766097366943e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003378185892142299,"output_power":0.0003021098960419091,"delta_db":-0.4851859246247626,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0005186258824964237,"output_power":0.00047115206702983336,"delta_db":-0.4169308604807669,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004467018619566432,"output_power":0.0004110134816408727,"delta_db":-0.3616169532545854,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002054215054829419,"output_power":0.00020621891873880062,"delta_db":0.016825976732514927,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00019049740954433386,"output_power":0.0002013168106988365,"delta_db":0.23990967248061043,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.3523016958893975e-05,"output_power":4.843186796318957e-05,"delta_db":0.46412227162018993,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3307652765322384,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2628658106686508,"pitch":null,"output_rms_dbfs":-27.8509179583748,"output_peak":0.5031802887105841,"reference_rms_dbfs":-27.58805214770615,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.017225581270599},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7595372702355393,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04174432016039388,"rms_dbfs":-27.58805214770615,"peak":0.493804931640625,"peak_dbfs":-6.1288915414792555,"crest_factor_db":21.45916060622689,"mean":-0.0027310686111450197,"zero_energy":false},"output":{"rms":0.040499914100074856,"rms_dbfs":-27.850917958374804,"peak":0.5031802887105841,"peak_dbfs":-5.965527600536133,"crest_factor_db":21.88539035783867,"mean":-0.0002574126500783736,"zero_energy":false},"rms_difference_db":-0.26286581066865367},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.4890963927609847e-06,"threshold_dbfs":-58.270771884202745,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":6,"quiet_samples":960,"reference_mean_power":9.472006543849905e-07,"output_mean_power":4.966393670375903e-07,"output_rms_dbfs":-63.03958858261986,"output_zero_energy":false,"output_peak":0.002241272872500833,"output_energy_fraction":3.027840108363903e-06,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":112,"output_estimated_voiced_frames":101,"both_estimated_voiced_frames":91,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.8125,"absolute_difference_cents_median":3.706340906572137,"absolute_difference_cents_p95":17.224663500408603,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001220285373493014,"log_rms_mae_db":2.9168165483563087,"rms_envelope_nrmse":0.24819096898944476,"envelope_correlation":0.9330124489408169,"rms_gain_error_db":-0.2634163791638652,"reference_mean_power":0.0017439355594579833,"output_mean_power":0.001641303120778961},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.14297331662316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004136999026359357,"output_power":0.00040198297929961953,"reference_active":true,"delta_db":-0.12477754039733947},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.498679838909322e-05,"output_power":7.951274268175843e-05,"reference_active":true,"delta_db":-0.289147347648388},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.720897303077822e-05,"output_power":1.935509091403451e-05,"reference_active":true,"delta_db":0.5104026163401243}],"active_band_count":3,"active_band_mae_db":0.3081091681286173}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00011225274665232883,"log_rms_mae_db":2.024825246201491,"rms_envelope_nrmse":0.15812708310730453,"envelope_correlation":0.970875956171471,"rms_gain_error_db":-0.26566367762015414,"reference_mean_power":0.0017512823148994152,"output_mean_power":0.0016473648466159845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.129560410778096e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037542654743209525,"output_power":0.0003658820093211973,"reference_active":true,"delta_db":-0.1118392407217332},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.989892754916159e-05,"output_power":3.65153357231293e-05,"reference_active":true,"delta_db":-0.38485924447148856},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1683754663794268e-06,"output_power":1.26643533016245e-06,"reference_active":true,"delta_db":0.3500058862437023}],"active_band_count":3,"active_band_mae_db":0.28223479047897465}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.832439224759049e-05,"log_rms_mae_db":2.9646434618099233,"rms_envelope_nrmse":0.28277931802408945,"envelope_correlation":0.933448088342185,"rms_gain_error_db":-0.3601726582185372,"reference_mean_power":0.0005656706566344367,"output_mean_power":0.0005206506142427169},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3997327154946766e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018576385507894965,"output_power":0.00016868647467120865,"reference_active":true,"delta_db":-0.4188095308641318},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.516179020835253e-05,"output_power":3.0468179055526384e-05,"reference_active":true,"delta_db":-0.6222447920625784},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.50699704718427e-06,"output_power":7.4309913614298285e-06,"reference_active":true,"delta_db":-0.0441948860428467}],"active_band_count":3,"active_band_mae_db":0.36174973632318563}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.746258281389908e-05,"log_rms_mae_db":2.2850804143304706,"rms_envelope_nrmse":0.20639952861076266,"envelope_correlation":0.9626915725794003,"rms_gain_error_db":-0.36459218995625314,"reference_mean_power":0.0005675319915431874,"output_mean_power":0.0005218325058472784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.076532458167812e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017606095149491153,"output_power":0.000159356681948691,"reference_active":true,"delta_db":-0.43292766111326697},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7577844172032785e-05,"output_power":1.5374893211362698e-05,"reference_active":true,"delta_db":-0.5815350212315435},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.080515782562601e-07,"output_power":4.823486475103454e-07,"reference_active":true,"delta_db":-0.2254673971848571}],"active_band_count":3,"active_band_mae_db":0.4133100265098892}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.679296131568703e-05,"log_rms_mae_db":3.0668617989521327,"rms_envelope_nrmse":0.3232667219990197,"envelope_correlation":0.9214122372607116,"rms_gain_error_db":0.07255256781385581,"reference_mean_power":0.0002582156143082965,"output_mean_power":0.00026256555820431576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9030353181623002e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.270493557085618e-05,"output_power":9.649016497254683e-05,"reference_active":true,"delta_db":0.17380192519833607},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6852701179097834e-05,"output_power":1.7110105180818642e-05,"reference_active":true,"delta_db":0.06583159052879603},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.066518069212081e-06,"output_power":4.29137582146399e-06,"reference_active":true,"delta_db":0.2337384357573058}],"active_band_count":3,"active_band_mae_db":0.15779065049481264}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.121002038281381e-05,"log_rms_mae_db":2.1632676078789417,"rms_envelope_nrmse":0.21491611051510812,"envelope_correlation":0.9629421553683633,"rms_gain_error_db":0.06717496320307073,"reference_mean_power":0.0002594441947231778,"output_mean_power":0.0002634883717779237},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.757465163940661e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.903046567125773e-05,"output_power":9.294022489468729e-05,"reference_active":true,"delta_db":0.18665074093518458},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.408774632836403e-06,"output_power":8.673913150454884e-06,"reference_active":true,"delta_db":0.13482356437530243},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6645360300144784e-07,"output_power":2.7182257287862724e-07,"reference_active":true,"delta_db":0.0866392170519052}],"active_band_count":3,"active_band_mae_db":0.1360378407874641}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.318455452770038e-05,"log_rms_mae_db":3.653147888348009,"rms_envelope_nrmse":0.38656438540087834,"envelope_correlation":0.9010637969751777,"rms_gain_error_db":0.30068835855202436,"reference_mean_power":0.00014927695528659852,"output_mean_power":0.00015997849406481588},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.071172575013412e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.0274139443533366e-05,"output_power":6.5330068646017e-05,"reference_active":true,"delta_db":0.34982096019196673},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2776930907508548e-05,"output_power":1.6837345038842527e-05,"reference_active":true,"delta_db":1.1984706544464525},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1605300538171703e-06,"output_power":3.015230198730707e-06,"reference_active":true,"delta_db":1.4476016220025563}],"active_band_count":3,"active_band_mae_db":0.9986310788803251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.047245550868427e-05,"log_rms_mae_db":2.605339310835941,"rms_envelope_nrmse":0.2751717731285632,"envelope_correlation":0.9466020508319063,"rms_gain_error_db":0.2991562517576565,"reference_mean_power":0.00014998968981604915,"output_mean_power":0.00016068562715398704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.934560164313176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8952201525784146e-05,"output_power":6.546203301634797e-05,"reference_active":true,"delta_db":0.45489460677082705},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.694634972457906e-06,"output_power":7.121990797406168e-06,"reference_active":true,"delta_db":0.9713551724909448},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8905823102195034e-07,"output_power":2.6316428007380775e-07,"reference_active":true,"delta_db":1.4363135112380563}],"active_band_count":3,"active_band_mae_db":0.9541877634999428}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"hnm","status":"measured","mos":{"status":"measured","value":1.9943938255310059,"predicted_mos":1.9943938255310059,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":57920,"input_sample_rate":16000,"input_seconds":3.62,"input_float32_sha256":"9fc12b527353b6027e40b991ea47f152a14fd635d07d5cf5c3a58759f217a339","internal_samples":57920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.099028425000142},"started_at":"2026-10-13T04:40:04.826712+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5599000107613392e-05,"native_input_samples":57920,"native_input_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","output_sha256":"e32beeda875008e2bfeb7a3a3d74ab190337ce4330904d306fe0252286ae5c78","output_samples":57920,"length_delta_samples":0,"analysis_s":0.6081783839999844,"synthesis_s":0.309312370999578,"total_s":0.9174907549995623,"rtf":0.25345048480650895,"output_peak":0.4889510168376783,"output_rms":0.05375438943374775,"finished_at":"2026-10-13T04:40:05.748024+00:00","evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"e32beeda875008e2bfeb7a3a3d74ab190337ce4330904d306fe0252286ae5c78","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.172343533635333,"spectral_convergence":0.40130124726676036,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.59023450178705,"spectral_convergence":0.3837924210215586,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.959370032341219,"spectral_convergence":0.41413604313065794,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.00039504395168052473,"delta_db":-0.32380577357426443,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.0012993531697463524,"delta_db":-0.07514781449161652,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0005468933909464569,"delta_db":-0.5845986315484398,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.00027439025140270606,"delta_db":-0.38387791204507726,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.0003637050895126796,"delta_db":-0.37064521401467276,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":1.0148506672563279e-05,"delta_db":1.1602267460415365,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4830503486192679,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2729629708833871,"pitch":null,"output_rms_dbfs":-25.391721333727567,"output_peak":0.4889510168376783,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.907316022587867},"common_prefix":{"samples":57920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.172343533635333,"spectral_convergence":0.40130124726676036,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.59023450178705,"spectral_convergence":0.3837924210215586,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.959370032341219,"spectral_convergence":0.41413604313065794,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.00039504395168052473,"delta_db":-0.32380577357426443,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.0012993531697463524,"delta_db":-0.07514781449161652,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0005468933909464569,"delta_db":-0.5845986315484398,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.00027439025140270606,"delta_db":-0.38387791204507726,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.0003637050895126796,"delta_db":-0.37064521401467276,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":1.0148506672563279e-05,"delta_db":1.1602267460415365,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4830503486192679,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2729629708833871,"pitch":null,"output_rms_dbfs":-25.391721333727567,"output_peak":0.4889510168376783,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.907316022587867},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57920,"output_samples":57920,"intelligibility":{"estoi":{"value":0.8038787165827951,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055470500168298326,"rms_dbfs":-25.11875836284418,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.685647510165401,"mean":-0.00030342070437267997,"zero_energy":false},"output":{"rms":0.05375438943374775,"rms_dbfs":-25.391721333727567,"peak":0.4889510168376783,"peak_dbfs":-6.214692927246959,"crest_factor_db":19.17702840648061,"mean":-0.00012691215150405485,"zero_energy":false},"rms_difference_db":-0.2729629708833876},"full_scale":{"reference":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":362,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":5.039671273196785e-07,"output_mean_power":7.298555605310419e-07,"output_rms_dbfs":-61.367630788809606,"output_zero_energy":false,"output_peak":0.02474447330750932,"output_energy_fraction":7.186835571012823e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":355,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":42,"both_estimated_voiced_frames":36,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.8,"absolute_difference_cents_median":1.7010445749976417,"absolute_difference_cents_p95":13.306714090539659,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00020370994921487513,"log_rms_mae_db":1.1149952721525482,"rms_envelope_nrmse":0.10609887756172169,"envelope_correlation":0.9902079269884586,"rms_gain_error_db":-0.27296325672708655,"reference_mean_power":0.003081231895334411,"output_mean_power":0.002893530464140216},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.3803128421607647e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003568424263155193,"output_power":0.00033759482045876677,"reference_active":true,"delta_db":-0.24080708799619518},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014048200587458577,"output_power":9.86691435587228e-05,"reference_active":true,"delta_db":-1.5343934111416475},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7025084230927582e-05,"output_power":1.8849869790892582e-05,"reference_active":true,"delta_db":0.4421908533453657}],"active_band_count":3,"active_band_mae_db":0.7391304508277362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.00013932647455664127,"log_rms_mae_db":0.6104406596737783,"rms_envelope_nrmse":0.06680534479579965,"envelope_correlation":0.9964840612217318,"rms_gain_error_db":-0.27296527579763497,"reference_mean_power":0.0031070091895842884,"output_mean_power":0.0029177361095119882},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":1.2870957120852427e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031478645736461515,"output_power":0.00029645887432214494,"reference_active":true,"delta_db":-0.26051584794548943},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.928842911180188e-05,"output_power":4.842428193880205e-05,"reference_active":true,"delta_db":-1.5559752575241135},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4070650648317698e-06,"output_power":1.0633977626526977e-06,"reference_active":true,"delta_db":-1.216184380969037}],"active_band_count":3,"active_band_mae_db":1.0108918288128799}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.0001235612100218032,"log_rms_mae_db":1.394696511726699,"rms_envelope_nrmse":0.16538080619475862,"envelope_correlation":0.9785872231179803,"rms_gain_error_db":-0.28970721818323153,"reference_mean_power":0.0011039001556015875,"output_mean_power":0.0010326640844659095},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":4.707425731708226e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018855725502559516,"output_power":0.00014930976650953721,"reference_active":true,"delta_db":-1.0135503088527547},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8855557263471545e-05,"output_power":5.9335629754553335e-05,"reference_active":true,"delta_db":0.03528079980677399},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.265121532076036e-06,"output_power":7.2068461759852455e-06,"reference_active":true,"delta_db":2.278138418710234}],"active_band_count":3,"active_band_mae_db":1.1089898424565874}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.0001123986305082509,"log_rms_mae_db":0.874873050731013,"rms_envelope_nrmse":0.09864224765679808,"envelope_correlation":0.9926985708193817,"rms_gain_error_db":-0.28970791287460634,"reference_mean_power":0.0011131375427710934,"output_mean_power":0.0010413052048099086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":4.401493945076042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017713633289459936,"output_power":0.00014212906010070385,"reference_active":true,"delta_db":-0.9562476554186898},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.789895589253068e-05,"output_power":2.812449046207575e-05,"reference_active":true,"delta_db":0.034967126994467745},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0644420897956184e-07,"output_power":3.978889601427604e-07,"reference_active":true,"delta_db":-0.09239050755845304}],"active_band_count":3,"active_band_mae_db":0.3612017633238702}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":1.6471252632908864,"rms_envelope_nrmse":0.16017549681863524,"envelope_correlation":0.9811178077548076,"rms_gain_error_db":-0.4512885312454363,"reference_mean_power":0.0005065055850916313,"output_mean_power":0.00045651535050829817},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":2.2793822474038483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569673324300418,"output_power":8.85804315826486e-05,"reference_active":true,"delta_db":-0.7672377289563581},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5368696588833676e-05,"output_power":1.64686291087947e-05,"reference_active":true,"delta_db":0.30020412108097666},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905327294990816e-06,"output_power":6.408310892382234e-06,"reference_active":true,"delta_db":2.150861361292428}],"active_band_count":3,"active_band_mae_db":1.072767737109921}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":0.9915992846770507,"rms_envelope_nrmse":0.08744347410705171,"envelope_correlation":0.9956363682308518,"rms_gain_error_db":-0.451293554166973,"reference_mean_power":0.0005107405079369303,"output_mean_power":0.00046033176966964477},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":2.126988285753868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.701418164823099e-05,"output_power":8.386132354327175e-05,"reference_active":true,"delta_db":-0.6327351220547416},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.796784971798844e-06,"output_power":7.376244600456884e-06,"reference_active":true,"delta_db":0.35531779263096697},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1410692673718095e-07,"output_power":3.664853923514683e-07,"reference_active":true,"delta_db":2.3342595117310028}],"active_band_count":3,"active_band_mae_db":1.1074374754722371}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":1.9146319899565234,"rms_envelope_nrmse":0.1992821208556031,"envelope_correlation":0.9705701267594836,"rms_gain_error_db":-0.2642512113372033,"reference_mean_power":0.00014114001095811158,"output_mean_power":0.0001328082413147264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.582618674096986e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4831777511599814e-05,"output_power":3.0411869189724854e-05,"reference_active":true,"delta_db":-0.5893252385249135},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4244491686748363e-05,"output_power":1.4859404762465864e-05,"reference_active":true,"delta_db":0.1835445672215436},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318870090030596e-06,"output_power":3.2847603878708213e-06,"reference_active":true,"delta_db":1.5122727588527163}],"active_band_count":3,"active_band_mae_db":0.7617141881997244}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":1.191604757247861,"rms_envelope_nrmse":0.11786867917552965,"envelope_correlation":0.9894501788173857,"rms_gain_error_db":-0.26426568908310843,"reference_mean_power":0.00014231959462201218,"output_mean_power":0.00013391774542614347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":7.830084455082623e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3431441965989596e-05,"output_power":3.015842888538636e-05,"reference_active":true,"delta_db":-0.4474639594574787},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2151972914162375e-06,"output_power":8.08058551103707e-06,"reference_active":true,"delta_db":0.491946195690671},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7745463336154605e-07,"output_power":2.757244695992503e-07,"reference_active":true,"delta_db":1.9138796664695292}],"active_band_count":3,"active_band_mae_db":0.951096607205893}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.172847270965576,"predicted_mos":3.172847270965576,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":57920,"input_sample_rate":16000,"input_seconds":3.62,"input_float32_sha256":"c5317b122c0756001cb59744c1be01ef5023028a7ac6fe06d179e666c1928a78","internal_samples":57920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.019980578999821},"started_at":"2026-10-13T04:43:17.296963+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5457999981881585e-05,"native_input_samples":57920,"native_input_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","output_sha256":"77a7de1be38ebfdc8af4e9810d0cfb1d4fa3818297b3a7014edd93107114b42b","output_samples":57920,"length_delta_samples":0,"analysis_s":1.0185585809995246,"synthesis_s":0.05337915800009796,"total_s":1.0719377389996225,"rtf":0.29611539751370786,"output_peak":0.5878861636006804,"output_rms":0.06548945291514893,"finished_at":"2026-10-13T04:43:18.374058+00:00","evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"77a7de1be38ebfdc8af4e9810d0cfb1d4fa3818297b3a7014edd93107114b42b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.9507857849647,"spectral_convergence":0.6435109021070435,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.193520192573219,"spectral_convergence":0.35318337440753317,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.628128886496384,"spectral_convergence":0.419612181965551,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.00043373407623298023,"delta_db":0.08197549474157974,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.002064422481405536,"delta_db":1.9355658819368282,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0009504175670529891,"delta_db":1.8155191676393159,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.0003711419135109139,"delta_db":0.9278353372289839,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.0004616837578190213,"delta_db":0.6653069639720771,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":7.468644843209805e-06,"delta_db":-0.1713765909138522,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9329299060721064,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4421856116631924,"pitch":null,"output_rms_dbfs":-23.676572751180988,"output_peak":0.5878861636006804,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.924144954678101},"common_prefix":{"samples":57920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.9507857849647,"spectral_convergence":0.6435109021070435,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.193520192573219,"spectral_convergence":0.35318337440753317,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.628128886496384,"spectral_convergence":0.419612181965551,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.00043373407623298023,"delta_db":0.08197549474157974,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.002064422481405536,"delta_db":1.9355658819368282,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0009504175670529891,"delta_db":1.8155191676393159,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.0003711419135109139,"delta_db":0.9278353372289839,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.0004616837578190213,"delta_db":0.6653069639720771,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":7.468644843209805e-06,"delta_db":-0.1713765909138522,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9329299060721064,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4421856116631924,"pitch":null,"output_rms_dbfs":-23.676572751180988,"output_peak":0.5878861636006804,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.924144954678101},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57920,"output_samples":57920,"intelligibility":{"estoi":{"value":0.9158642014842708,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055470500168298326,"rms_dbfs":-25.11875836284418,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.685647510165401,"mean":-0.00030342070437267997,"zero_energy":false},"output":{"rms":0.06548945291514893,"rms_dbfs":-23.676572751180988,"peak":0.5878861636006804,"peak_dbfs":-4.6141352235899475,"crest_factor_db":19.06243752759104,"mean":1.0048692156514647e-09,"zero_energy":false},"rms_difference_db":1.4421856116631915},"full_scale":{"reference":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":362,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":5.039671273196785e-07,"output_mean_power":4.256555386545841e-07,"output_rms_dbfs":-63.70941711169828,"output_zero_energy":false,"output_peak":0.007568487588659098,"output_energy_fraction":2.823866869259639e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":355,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":56,"both_estimated_voiced_frames":41,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9111111111111111,"absolute_difference_cents_median":8.241875779554958,"absolute_difference_cents_p95":51.22717584056129,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00020370994921487513,"log_rms_mae_db":2.084677646249134,"rms_envelope_nrmse":0.2355041157980418,"envelope_correlation":0.9828163028103001,"rms_gain_error_db":1.442186038737375,"reference_mean_power":0.003081231895334411,"output_mean_power":0.0042948004375211795},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.3803128421607647e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003568424263155193,"output_power":0.0005558110593919404,"reference_active":true,"delta_db":1.9245070036160294},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014048200587458577,"output_power":0.00016628958151659114,"reference_active":true,"delta_db":0.7324434068343149},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7025084230927582e-05,"output_power":1.5912208836129192e-05,"reference_active":true,"delta_db":-0.29358799292891385}],"active_band_count":3,"active_band_mae_db":0.9835128011264193}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.00013932647455664127,"log_rms_mae_db":1.9172555935061253,"rms_envelope_nrmse":0.2045613584395862,"envelope_correlation":0.9929872166902689,"rms_gain_error_db":1.4421905053569821,"reference_mean_power":0.0031070091895842884,"output_mean_power":0.0043307347853320565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":1.2870957120852427e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031478645736461515,"output_power":0.00049035246181386,"reference_active":true,"delta_db":1.9249231999621987},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.928842911180188e-05,"output_power":8.425761123354813e-05,"reference_active":true,"delta_db":0.8494842715114282},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4070650648317698e-06,"output_power":1.2712756875358403e-06,"reference_active":true,"delta_db":-0.44074438916071385}],"active_band_count":3,"active_band_mae_db":1.0717172868781135}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.0001235612100218032,"log_rms_mae_db":1.6272880845216584,"rms_envelope_nrmse":0.2816895104876401,"envelope_correlation":0.984568627692716,"rms_gain_error_db":1.8129672633409801,"reference_mean_power":0.0011039001556015875,"output_mean_power":0.0016758167266904803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":4.707425731708226e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018855725502559516,"output_power":0.00029363851826753743,"reference_active":true,"delta_db":1.9236977670670286},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8855557263471545e-05,"output_power":9.857491509579491e-05,"reference_active":true,"delta_db":2.239789355814346},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.265121532076036e-06,"output_power":5.152759128509357e-06,"reference_active":true,"delta_db":0.82108430713863}],"active_band_count":3,"active_band_mae_db":1.661523810006668}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.0001123986305082509,"log_rms_mae_db":1.2160563516747047,"rms_envelope_nrmse":0.25816139374650615,"envelope_correlation":0.9921747849253814,"rms_gain_error_db":1.8129675159404264,"reference_mean_power":0.0011131375427710934,"output_mean_power":0.0016898399844547352},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":4.401493945076042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017713633289459936,"output_power":0.00027732482973269674,"reference_active":true,"delta_db":1.9468110542086432},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.789895589253068e-05,"output_power":5.008548448655603e-05,"reference_active":true,"delta_db":2.541239288674226},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0644420897956184e-07,"output_power":3.52185853914248e-07,"reference_active":true,"delta_db":-0.6222903222511924}],"active_band_count":3,"active_band_mae_db":1.7034468883780205}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":1.4931053832843255,"rms_envelope_nrmse":0.16456804070560768,"envelope_correlation":0.9867859302913825,"rms_gain_error_db":0.7633653593281315,"reference_mean_power":0.0005065055850916313,"output_mean_power":0.0006038384650236865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":2.2793822474038483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569673324300418,"output_power":0.00012414943825367022,"reference_active":true,"delta_db":0.6988319399266683},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5368696588833676e-05,"output_power":1.5162154284367544e-05,"reference_active":true,"delta_db":-0.0587612522932399},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905327294990816e-06,"output_power":2.669595311422765e-06,"reference_active":true,"delta_db":-1.6521200589120197}],"active_band_count":3,"active_band_mae_db":0.8032377503773093}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":0.9997749549163688,"rms_envelope_nrmse":0.11028247366918643,"envelope_correlation":0.99725178465994,"rms_gain_error_db":0.7633692087830662,"reference_mean_power":0.0005107405079369303,"output_mean_power":0.0006088877334749179},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":2.126988285753868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.701418164823099e-05,"output_power":0.00011401282028231473,"reference_active":true,"delta_db":0.7011846415736354},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.796784971798844e-06,"output_power":6.524255778341944e-06,"reference_active":true,"delta_db":-0.17772551595456237},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1410692673718095e-07,"output_power":1.7343120345871692e-07,"reference_active":true,"delta_db":-0.9150347999143217}],"active_band_count":3,"active_band_mae_db":0.5979816524808398}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":1.9970055792421475,"rms_envelope_nrmse":0.22191509691297243,"envelope_correlation":0.971465944039054,"rms_gain_error_db":0.6936505663682991,"reference_mean_power":0.00014114001095811158,"output_mean_power":0.00016558279284298213},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.582618674096986e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4831777511599814e-05,"output_power":4.075537224437989e-05,"reference_active":true,"delta_db":0.6820922653099405},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4244491686748363e-05,"output_power":1.723854965596209e-05,"reference_active":true,"delta_db":0.8285376809985777},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318870090030596e-06,"output_power":2.217255193374954e-06,"reference_active":true,"delta_db":-0.19460738057589128}],"active_band_count":3,"active_band_mae_db":0.5684124422948031}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":1.0926963135121741,"rms_envelope_nrmse":0.12810895658347593,"envelope_correlation":0.9933390978015407,"rms_gain_error_db":0.6936747902783859,"reference_mean_power":0.00014231959462201218,"output_mean_power":0.00016696758940221946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":7.830084455082623e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3431441965989596e-05,"output_power":3.9135391271377966e-05,"reference_active":true,"delta_db":0.6841457120692973},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2151972914162375e-06,"output_power":9.058298816198416e-06,"reference_active":true,"delta_db":0.9879843225563967},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7745463336154605e-07,"output_power":1.512682502341182e-07,"reference_active":true,"delta_db":-0.6933956000201392}],"active_band_count":3,"active_band_mae_db":0.7885085448819443}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"facodec","status":"measured","mos":{"status":"measured","value":4.219052791595459,"predicted_mos":4.219052791595459,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":58000,"input_sample_rate":16000,"input_seconds":3.625,"input_float32_sha256":"c5119120390f5e3d3beb25f00ae110d2b2982e38834c8e5ee6c95d3d35e4d84c","internal_samples":58000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.265851323001698},"started_at":"2026-10-13T06:58:51.920803+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.657499994209502e-05,"native_input_samples":57920,"native_input_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","output_sha256":"86723f8a8db480644db82ad2c8b107322b02c0efd58ea7a106723f024bbc16f2","output_samples":58000,"length_delta_samples":80,"analysis_s":1.0245292960007646,"synthesis_s":6.122902601000533,"total_s":7.147431897001297,"rtf":1.9744287008290875,"output_peak":0.4408506751060486,"output_rms":0.05145610721987632,"finished_at":"2026-10-13T06:58:59.077922+00:00","evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"86723f8a8db480644db82ad2c8b107322b02c0efd58ea7a106723f024bbc16f2","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":57920,"output_samples":58000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.771261446413508,"output_peak":0.4408506751060486},"common_prefix":{"samples":57920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986206896551724,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.2941259124765,"spectral_convergence":0.3037764718424351,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.419987225466673,"spectral_convergence":0.3138752999837016,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.684343277070814,"spectral_convergence":0.3482344648782759,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.00033134756769297174,"delta_db":-1.0874220734177624,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.001181149456738577,"delta_db":-0.4893713712429676,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0005762333955896385,"delta_db":-0.357641140361421,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.00023869954330756394,"delta_db":-0.9890488071063303,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.0003156575957855363,"delta_db":-0.9859765542717098,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":8.300487252449158e-06,"delta_db":0.2872411993806545,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6994501909634744,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6465086975719504,"pitch":null,"output_rms_dbfs":-25.76526706041613,"output_peak":0.4408506751060486,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.466152138337996},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57920,"output_samples":57920,"intelligibility":{"estoi":{"value":0.8804954429976753,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055470500168298326,"rms_dbfs":-25.11875836284418,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.685647510165401,"mean":-0.00030342070437267997,"zero_energy":false},"output":{"rms":0.051491630838106456,"rms_dbfs":-25.76526706041613,"peak":0.4408506751060486,"peak_dbfs":-7.114169796102283,"crest_factor_db":18.651097264313847,"mean":-8.618460283767777e-05,"zero_energy":false},"rms_difference_db":-0.6465086975719494},"full_scale":{"reference":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":362,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":5.039671273196785e-07,"output_mean_power":4.2135425219349125e-07,"output_rms_dbfs":-63.753526189378256,"output_zero_energy":false,"output_peak":0.009872635826468468,"output_energy_fraction":4.521710373384732e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":355,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":47,"both_estimated_voiced_frames":42,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9333333333333333,"absolute_difference_cents_median":5.3989696921458865,"absolute_difference_cents_p95":18.900317653945603,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00020370994921487513,"log_rms_mae_db":1.6001735583866623,"rms_envelope_nrmse":0.14364517379851527,"envelope_correlation":0.9847955360619173,"rms_gain_error_db":-0.6465085014764438,"reference_mean_power":0.003081231895334411,"output_mean_power":0.002655055077337727},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.3803128421607647e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003568424263155193,"output_power":0.000292581565954328,"reference_active":true,"delta_db":-0.8622952364027865},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014048200587458577,"output_power":9.897310513274634e-05,"reference_active":true,"delta_db":-1.5210350387024205},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7025084230927582e-05,"output_power":1.4764295297054876e-05,"reference_active":true,"delta_db":-0.618765464167542}],"active_band_count":3,"active_band_mae_db":1.000698579757583}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.00013932647455664127,"log_rms_mae_db":1.2598419249543948,"rms_envelope_nrmse":0.13163868375161536,"envelope_correlation":0.9874706262601013,"rms_gain_error_db":-0.6465088712928915,"reference_mean_power":0.0031070091895842884,"output_mean_power":0.0026772667887079984},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":1.2870957120852427e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031478645736461515,"output_power":0.0002562207736864825,"reference_active":true,"delta_db":-0.8940170176346514},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.928842911180188e-05,"output_power":4.658254613742105e-05,"reference_active":true,"delta_db":-1.7243749245804465},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4070650648317698e-06,"output_power":1.1448685667140845e-06,"reference_active":true,"delta_db":-0.89558548707086}],"active_band_count":3,"active_band_mae_db":1.171325809761986}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.0001235612100218032,"log_rms_mae_db":1.2554756882188494,"rms_envelope_nrmse":0.10240828095124584,"envelope_correlation":0.9924642964509289,"rms_gain_error_db":-0.2977587280585412,"reference_mean_power":0.0011039001556015875,"output_mean_power":0.0010307513724856399},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":4.707425731708226e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018855725502559516,"output_power":0.00017331602716803517,"reference_active":true,"delta_db":-0.3660452182734907},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8855557263471545e-05,"output_power":6.000560497691242e-05,"reference_active":true,"delta_db":0.08404342565576249},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.265121532076036e-06,"output_power":3.979007503124369e-06,"reference_active":true,"delta_db":-0.3015665255380566}],"active_band_count":3,"active_band_mae_db":0.2505517231557699}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.0001123986305082509,"log_rms_mae_db":0.9197098907880605,"rms_envelope_nrmse":0.0813261985050123,"envelope_correlation":0.9954526708579314,"rms_gain_error_db":-0.29775869180715764,"reference_mean_power":0.0011131375427710934,"output_mean_power":0.0010393766625828762},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":4.401493945076042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017713633289459936,"output_power":0.0001638574117898967,"reference_active":true,"delta_db":-0.3384155892116377},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.789895589253068e-05,"output_power":2.7908439202464095e-05,"reference_active":true,"delta_db":0.001475986903489482},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0644420897956184e-07,"output_power":4.011658409340603e-07,"reference_active":true,"delta_db":-0.056769941318466875}],"active_band_count":3,"active_band_mae_db":0.132220505811198}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":1.4704199213722255,"rms_envelope_nrmse":0.19460142066841155,"envelope_correlation":0.9796579361691188,"rms_gain_error_db":-1.1222987114555796,"reference_mean_power":0.0005065055850916313,"output_mean_power":0.0003911599369175847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":2.2793822474038483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569673324300418,"output_power":8.322778869516999e-05,"reference_active":true,"delta_db":-1.037932089423739},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5368696588833676e-05,"output_power":1.4136356661119443e-05,"reference_active":true,"delta_db":-0.3629954285978538},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905327294990816e-06,"output_power":4.817462576115827e-06,"reference_active":true,"delta_db":0.9116091278306503}],"active_band_count":3,"active_band_mae_db":0.7708455486174143}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":1.2039859656103735,"rms_envelope_nrmse":0.17473520286525396,"envelope_correlation":0.9857301324106762,"rms_gain_error_db":-1.1223055108406268,"reference_mean_power":0.0005107405079369303,"output_mean_power":0.00039442983052049164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":2.126988285753868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.701418164823099e-05,"output_power":7.676983763775456e-05,"reference_active":true,"delta_db":-1.016446024810708},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.796784971798844e-06,"output_power":5.864508496513633e-06,"reference_active":true,"delta_db":-0.640719106446559},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1410692673718095e-07,"output_power":2.844558336964326e-07,"reference_active":true,"delta_db":1.2338412710754048}],"active_band_count":3,"active_band_mae_db":0.9636688007775572}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":1.9519888843563293,"rms_envelope_nrmse":0.2507733521977613,"envelope_correlation":0.9573555405183614,"rms_gain_error_db":-0.9824063762230131,"reference_mean_power":0.00014114001095811158,"output_mean_power":0.000112566589758213},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.582618674096986e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4831777511599814e-05,"output_power":2.6686255463254593e-05,"reference_active":true,"delta_db":-1.1568799850434186},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4244491686748363e-05,"output_power":1.0151093497415256e-05,"reference_active":true,"delta_db":-1.4713412820122695},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318870090030596e-06,"output_power":2.6141054090247894e-06,"reference_active":true,"delta_db":0.520466768726327}],"active_band_count":3,"active_band_mae_db":1.049562678594005}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":1.5189696691990666,"rms_envelope_nrmse":0.21793772029376735,"envelope_correlation":0.9679195246052871,"rms_gain_error_db":-0.9824111129348394,"reference_mean_power":0.00014231959462201218,"output_mean_power":0.00011350724603642139},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":7.830084455082623e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3431441965989596e-05,"output_power":2.6101464447680567e-05,"reference_active":true,"delta_db":-1.0749023446956048},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2151972914162375e-06,"output_power":4.76502633978105e-06,"reference_active":true,"delta_db":-1.8018290523989842},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7745463336154605e-07,"output_power":1.7813359488959213e-07,"reference_active":true,"delta_db":0.016584891173313594}],"active_band_count":3,"active_band_mae_db":0.9644387627559675}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.057785987854004,"predicted_mos":3.057785987854004,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":159744,"input_sample_rate":44100,"input_seconds":3.622312925170068,"input_float32_sha256":"dd2dbba72697b8585f9caa8b974c558c8c2b1159a6d78e9c09ab0bc1b57dedf0","internal_samples":57958,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.263990976000059},"started_at":"2026-10-13T07:28:15.919983+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004369761001726147,"native_input_samples":159642,"native_input_sha256":"3376d54341577adf240b4ad63dc12c72b70707ada5d8e670f93e3d3d0cb5c6ad","output_sha256":"47d54fb3a981b0a716bb0f380f2a577059795ba09532b2c148b7b9a8497a01e4","output_samples":159744,"length_delta_samples":102,"analysis_s":0.09602179700232227,"synthesis_s":21.322690978999162,"total_s":21.418712776001485,"rtf":5.916771485083283,"output_peak":0.4315236210823059,"output_rms":0.05020946553491592,"finished_at":"2026-10-13T07:28:37.356174+00:00","evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"47d54fb3a981b0a716bb0f380f2a577059795ba09532b2c148b7b9a8497a01e4","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":57920,"output_samples":57958,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.979512219372282,"output_peak":0.39278613811146185},"common_prefix":{"samples":57920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9993443528072052,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.306377447653901,"spectral_convergence":0.5378631451416611,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.915790629938021,"spectral_convergence":0.5354657438429903,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.936992272058145,"spectral_convergence":0.5967012687110397,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.0005586597329725144,"delta_db":1.1812137517232588,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.0008955624214658734,"delta_db":-1.69146129727762,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.00045821977146837383,"delta_db":-1.352887138468267,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.0002569021132248374,"delta_db":-0.6698879210365943,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.00035119476432084846,"delta_db":-0.5226586706852393,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":4.881089435357951e-06,"delta_db":-2.0186270176476837,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2394559661397773,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8579056658531914,"pitch":null,"output_rms_dbfs":-25.97666402869737,"output_peak":0.39278613811146185,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.053053449883356},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57920,"output_samples":57920,"intelligibility":{"estoi":{"value":0.6607919388897299,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055470500168298326,"rms_dbfs":-25.11875836284418,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.685647510165401,"mean":-0.00030342070437267997,"zero_energy":false},"output":{"rms":0.050253556023515854,"rms_dbfs":-25.97666402869737,"peak":0.39278613811146185,"peak_dbfs":-8.116876947637024,"crest_factor_db":17.859787081060347,"mean":-8.662077741698449e-05,"zero_energy":false},"rms_difference_db":-0.8579056658531918},"full_scale":{"reference":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":362,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":5.039671273196785e-07,"output_mean_power":8.704277264732712e-07,"output_rms_dbfs":-60.60267283474757,"output_zero_energy":false,"output_peak":0.041971468220744455,"output_energy_fraction":9.80681027850363e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":355,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":45,"both_estimated_voiced_frames":33,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.7333333333333333,"absolute_difference_cents_median":14.755912834164278,"absolute_difference_cents_p95":28.282869029200786,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00020370994921487513,"log_rms_mae_db":1.964845586350163,"rms_envelope_nrmse":0.16814195027629283,"envelope_correlation":0.9808086253888358,"rms_gain_error_db":-0.8579056553764695,"reference_mean_power":0.003081231895334411,"output_mean_power":0.0025289125942724545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.3803128421607647e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003568424263155193,"output_power":0.00030827267234800274,"reference_active":true,"delta_db":-0.635414564087263},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014048200587458577,"output_power":9.969825645714944e-05,"reference_active":true,"delta_db":-1.4893313633688976},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7025084230927582e-05,"output_power":2.2682818394145943e-05,"reference_active":true,"delta_db":1.2460774645130692}],"active_band_count":3,"active_band_mae_db":1.1236077973230767}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.00013932647455664127,"log_rms_mae_db":1.7757503979670213,"rms_envelope_nrmse":0.13013754711270734,"envelope_correlation":0.9916979023792153,"rms_gain_error_db":-0.8579036069880419,"reference_mean_power":0.0031070091895842884,"output_mean_power":0.0025500704403162273},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":1.2870957120852427e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031478645736461515,"output_power":0.0002720561502283428,"reference_active":true,"delta_db":-0.6335749180344368},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.928842911180188e-05,"output_power":5.141977744896382e-05,"reference_active":true,"delta_db":-1.2953052260834383},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4070650648317698e-06,"output_power":1.3449966579699257e-06,"reference_active":true,"delta_db":-0.1959297512689165}],"active_band_count":3,"active_band_mae_db":0.7082699651289306}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.0001235612100218032,"log_rms_mae_db":2.8213139001864165,"rms_envelope_nrmse":0.27309368180773896,"envelope_correlation":0.942098604228878,"rms_gain_error_db":-0.8162395528834819,"reference_mean_power":0.0011039001556015875,"output_mean_power":0.0009147572058828231},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":4.707425731708226e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018855725502559516,"output_power":0.00010769981935805544,"reference_active":true,"delta_db":-2.432282723318138},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8855557263471545e-05,"output_power":4.460400751951522e-05,"reference_active":true,"delta_db":-1.2041359574057375},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.265121532076036e-06,"output_power":1.0619060926810983e-05,"reference_active":true,"delta_db":3.9615470193039752}],"active_band_count":3,"active_band_mae_db":2.532655233342617}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.0001123986305082509,"log_rms_mae_db":2.492753131562358,"rms_envelope_nrmse":0.23475179473156801,"envelope_correlation":0.9570054131273381,"rms_gain_error_db":-0.8162454075174288,"reference_mean_power":0.0011131375427710934,"output_mean_power":0.0009224106098182823},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":4.401493945076042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017713633289459936,"output_power":9.986729551975451e-05,"reference_active":true,"delta_db":-2.48884360579577},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.789895589253068e-05,"output_power":2.2333281455080784e-05,"reference_active":true,"delta_db":-0.9663541112290013},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0644420897956184e-07,"output_power":6.747146865668489e-07,"reference_active":true,"delta_db":2.2011922347801662}],"active_band_count":3,"active_band_mae_db":1.8854633172683126}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":2.337659040821119,"rms_envelope_nrmse":0.2097684776363855,"envelope_correlation":0.9672922901639271,"rms_gain_error_db":-0.5226499398845423,"reference_mean_power":0.0005065055850916313,"output_mean_power":0.0004490753784645019},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":2.2793822474038483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569673324300418,"output_power":9.450225245422138e-05,"reference_active":true,"delta_db":-0.48619404814759565},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5368696588833676e-05,"output_power":1.3772797499274467e-05,"reference_active":true,"delta_db":-0.4761487464889552},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905327294990816e-06,"output_power":2.576006429466172e-06,"reference_active":true,"delta_db":-1.8071049414587845}],"active_band_count":3,"active_band_mae_db":0.9231492453651118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":1.910260490807428,"rms_envelope_nrmse":0.16048275547954258,"envelope_correlation":0.9810850049074571,"rms_gain_error_db":-0.5226373637984271,"reference_mean_power":0.0005107405079369303,"output_mean_power":0.00045283143527562067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":2.126988285753868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.701418164823099e-05,"output_power":8.639501778518255e-05,"reference_active":true,"delta_db":-0.5034652621726806},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.796784971798844e-06,"output_power":5.595671280137075e-06,"reference_active":true,"delta_db":-0.844513367000823},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1410692673718095e-07,"output_power":2.063453713732795e-07,"reference_active":true,"delta_db":-0.16035986249175677}],"active_band_count":3,"active_band_mae_db":0.5027794972217535}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":3.2218527698139234,"rms_envelope_nrmse":0.3089846389452056,"envelope_correlation":0.9472013506026211,"rms_gain_error_db":-1.8663389571156606,"reference_mean_power":0.00014114001095811158,"output_mean_power":9.183669607213547e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.582618674096986e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4831777511599814e-05,"output_power":2.8583934500438865e-05,"reference_active":true,"delta_db":-0.858536296269306},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4244491686748363e-05,"output_power":9.176283187320194e-06,"reference_active":true,"delta_db":-1.9098014832765515},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318870090030596e-06,"output_power":1.970820041907406e-06,"reference_active":true,"delta_db":-0.7062944874993106}],"active_band_count":3,"active_band_mae_db":1.1582107556817227}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":2.7666018204461325,"rms_envelope_nrmse":0.2590574574106159,"envelope_correlation":0.9709782309061479,"rms_gain_error_db":-1.8663237230244587,"reference_mean_power":0.00014231959462201218,"output_mean_power":9.260455001388633e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":7.830084455082623e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3431441965989596e-05,"output_power":2.7012232388979765e-05,"reference_active":true,"delta_db":-0.9259463165581976},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2151972914162375e-06,"output_power":4.73355398151419e-06,"reference_active":true,"delta_db":-1.830608767018639},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7745463336154605e-07,"output_power":1.7135124498968361e-07,"reference_active":true,"delta_db":-0.15200079073912398}],"active_band_count":3,"active_band_mae_db":0.9695186247719869}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"golf","status":"measured","mos":{"status":"measured","value":2.407331943511963,"predicted_mos":2.407331943511963,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":86640,"input_sample_rate":24000,"input_seconds":3.61,"input_float32_sha256":"819cd0e7944b703742e0da9c27a479a43cec215ae1137366ddd5a992f5a39770","internal_samples":57760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6611371869985305},"started_at":"2026-10-13T05:02:58.050491+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0016531709998162114,"native_input_samples":86880,"native_input_sha256":"ed81f5ba0ebb3fc0d124045c3ac1909a0e16714f9d7ad8f46708b5f3a6861fc6","output_sha256":"e4fe01e6b77e1f10ecadafbaa2f9ad93bd8854eec5c9dda384faaf920ac45d80","output_samples":86640,"length_delta_samples":-240,"analysis_s":0.33573170399995433,"synthesis_s":0.09388615300031233,"total_s":0.42961785700026667,"rtf":0.11867896602217311,"output_peak":0.3584272861480713,"output_rms":0.048959014928421835,"finished_at":"2026-10-13T05:02:58.486833+00:00","analysis_stages":{"external_f0_s":0.04318143099953886},"evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"e4fe01e6b77e1f10ecadafbaa2f9ad93bd8854eec5c9dda384faaf920ac45d80","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":57920,"output_samples":57760,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.198830809365564,"output_peak":0.3588327689060995},"common_prefix":{"samples":57760,"sample_rate":16000,"reference_coverage":0.9972375690607734,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57760,"output_samples":57760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.7450014801481615,"spectral_convergence":0.5790335177908357,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.362947210066852,"spectral_convergence":0.3504523454982478,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.655870803044668,"spectral_convergence":0.42911607899926846,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004267978729339312,"output_power":0.00048343236301550557,"delta_db":0.5411347423156564,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013256977709908044,"output_power":0.0010124246354796173,"delta_db":-1.1708182169888646,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006274290557913588,"output_power":0.00043162400193520354,"delta_db":-1.624590395758379,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00030057801530011265,"output_power":0.00021042373042533835,"delta_db":-1.548624969266913,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00039720547160585704,"output_power":0.0002572464488814405,"delta_db":-1.8866583405562756,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.790770058601158e-06,"output_power":4.3276270277271166e-06,"delta_db":-2.5533056204456677,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5541887142219595,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0920848484816748,"pitch":null,"output_rms_dbfs":-26.198830809365564,"output_peak":0.3588327689060995,"reference_rms_dbfs":-25.106745960883888,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.921273164419894},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57760,"output_samples":57760,"intelligibility":{"estoi":{"value":0.8472650135387405,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0555472677704801,"rms_dbfs":-25.106745960883888,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.673635108205108,"mean":-0.00030333764665344746,"zero_energy":false},"output":{"rms":0.04898447519758639,"rms_dbfs":-26.198830809365564,"peak":0.3588327689060995,"peak_dbfs":-8.902158073692263,"crest_factor_db":17.2966727356733,"mean":0.00022748006983430456,"zero_energy":false},"rms_difference_db":-1.0920848484816759},"full_scale":{"reference":{"sample_count":57760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":361,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":102,"quiet_samples":16320,"reference_mean_power":5.056785976550743e-07,"output_mean_power":4.966069822199518e-07,"output_rms_dbfs":-63.03987178623019,"output_zero_energy":false,"output_peak":0.0048694032011985374,"output_energy_fraction":5.8477510039114516e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":354,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":47,"both_estimated_voiced_frames":34,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.7555555555555555,"absolute_difference_cents_median":7.90940546675734,"absolute_difference_cents_p95":42.01513683326648,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":721,"rms_floor":0.00020370994921487513,"log_rms_mae_db":2.379145349562047,"rms_envelope_nrmse":0.18119566751551647,"envelope_correlation":0.9811836870695165,"rms_gain_error_db":-1.0920844356891641,"reference_mean_power":0.003089777942315447,"output_mean_power":0.0024028066481105717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":721,"frequency_resolution_hz":0.27739251040221913,"power_floor":1.3789937941271744e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00035342922272743604,"output_power":0.00030495948640952403,"reference_active":true,"delta_db":-0.6406030804500398},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014093336723514544,"output_power":0.00016034477200536695,"reference_active":true,"delta_db":0.5604097606391576},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7146424085778286e-05,"output_power":2.5219798341258312e-05,"reference_active":true,"delta_db":1.6756804859918315}],"active_band_count":3,"active_band_mae_db":0.9588977756936763}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":715,"rms_floor":0.00013932647455664127,"log_rms_mae_db":1.9190686500644247,"rms_envelope_nrmse":0.1481786022884241,"envelope_correlation":0.9917794012639559,"rms_gain_error_db":-1.0920814234583944,"reference_mean_power":0.003115698995538401,"output_mean_power":0.0024229661782437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":715,"frequency_resolution_hz":0.2797202797202797,"power_floor":1.2858950649495452e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031174780806562225,"output_power":0.0002851482188814088,"reference_active":true,"delta_db":-0.3873274538717361},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.927600748627717e-05,"output_power":8.004681778437245e-05,"reference_active":true,"delta_db":0.6276122099076538},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4055719571330271e-06,"output_power":2.2963795469236533e-06,"reference_active":true,"delta_db":2.1319058622716205}],"active_band_count":3,"active_band_mae_db":1.0489485086836703}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":721,"rms_floor":0.0001235612100218032,"log_rms_mae_db":1.886837665603308,"rms_envelope_nrmse":0.1918487360907905,"envelope_correlation":0.9811246610467892,"rms_gain_error_db":-1.1641333733318724,"reference_mean_power":0.0011069622040577641,"output_mean_power":0.0008466803005958339},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":721,"frequency_resolution_hz":0.27739251040221913,"power_floor":4.706962359209797e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018819251631096287,"output_power":0.00013747979322931484,"reference_active":true,"delta_db":-1.363634789381746},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.802073985102316e-05,"output_power":4.288212048722753e-05,"reference_active":true,"delta_db":-1.3130700966750155},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.252539585686752e-06,"output_power":4.391431283065036e-06,"reference_active":true,"delta_db":0.13957726291469713}],"active_band_count":3,"active_band_mae_db":0.9387607163238195}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":715,"rms_floor":0.0001123986305082509,"log_rms_mae_db":1.5160390088632887,"rms_envelope_nrmse":0.1582235017614251,"envelope_correlation":0.9912439962792731,"rms_gain_error_db":-1.1641327643693047,"reference_mean_power":0.0011162510454186575,"output_mean_power":0.0008537851606251082},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":715,"frequency_resolution_hz":0.2797202797202797,"power_floor":4.4024298517167966e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017672751230513565,"output_power":0.00012687005208791648,"reference_active":true,"delta_db":-1.4394504600584193},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.713560547889103e-05,"output_power":1.9733611914567986e-05,"reference_active":true,"delta_db":-1.3833293344828403},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0547235125564333e-07,"output_power":4.011098090581898e-07,"reference_active":true,"delta_db":-0.046979627668043776}],"active_band_count":3,"active_band_mae_db":0.9565864740697677}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":721,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":2.0344553244659624,"rms_envelope_nrmse":0.23436010472517532,"envelope_correlation":0.9780113746909841,"rms_gain_error_db":-1.69407628802846,"reference_mean_power":0.0005079102174443075,"output_mean_power":0.00034385814870801064},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":721,"frequency_resolution_hz":0.27739251040221913,"power_floor":2.2834302519239191e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569679395747767,"output_power":8.019449045663217e-05,"reference_active":true,"delta_db":-1.1991728199914935},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5208971722397499e-05,"output_power":1.2435244940068018e-05,"reference_active":true,"delta_db":-0.8744550832841449},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905062389566477e-06,"output_power":2.7220084020837675e-06,"reference_active":true,"delta_db":-1.5673851539623507}],"active_band_count":3,"active_band_mae_db":1.2136710190793296}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":715,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":1.7874371661833912,"rms_envelope_nrmse":0.19771162789290553,"envelope_correlation":0.9924654290582763,"rms_gain_error_db":-1.694063653354292,"reference_mean_power":0.0005121688268744508,"output_mean_power":0.00034674226062699013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":715,"frequency_resolution_hz":0.2797202797202797,"power_floor":2.1312724445357945e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.6958821608135e-05,"output_power":7.455476962670467e-05,"reference_active":true,"delta_db":-1.1411189606189904},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.690004626425245e-06,"output_power":5.5867960707523965e-06,"reference_active":true,"delta_db":-0.7826359907024505},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1396984629381625e-07,"output_power":1.713113802147271e-07,"reference_active":true,"delta_db":-0.9656636059449824}],"active_band_count":3,"active_band_mae_db":0.9631395190888078}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":721,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":2.574243545465657,"rms_envelope_nrmse":0.2729080233036387,"envelope_correlation":0.9745175292964545,"rms_gain_error_db":-2.129969743786587,"reference_mean_power":0.00014153149575043002,"output_mean_power":8.666747065271865e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":721,"frequency_resolution_hz":0.27739251040221913,"power_floor":8.579157385710667e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4754058284960415e-05,"output_power":2.1041537577896447e-05,"reference_active":true,"delta_db":-2.179280530829069},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4184058443688752e-05,"output_power":9.259047589648096e-06,"reference_active":true,"delta_db":-1.8523419576456814},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313777863216646e-06,"output_power":1.6510412569445867e-06,"reference_active":true,"delta_db":-1.4656373595889003}],"active_band_count":3,"active_band_mae_db":1.8324199493545503}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57760,"uncovered_tail_samples":0,"metrics":{"frame_count":715,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":2.3070412854876534,"rms_envelope_nrmse":0.23464732980280437,"envelope_correlation":0.9922215827919585,"rms_gain_error_db":-2.1299431581565225,"reference_mean_power":0.00014271766362184893,"output_mean_power":8.739436106259985e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":715,"frequency_resolution_hz":0.2797202797202797,"power_floor":7.828599203460401e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.335070301397212e-05,"output_power":2.0400070678364787e-05,"reference_active":true,"delta_db":-2.134733209401771},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.167303308802194e-06,"output_power":4.901933201680376e-06,"reference_active":true,"delta_db":-1.6498839465146697},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7519218764133424e-07,"output_power":1.4035567723766912e-07,"reference_active":true,"delta_db":-0.9628475166167493}],"active_band_count":3,"active_band_mae_db":1.58248822417773}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.1229443550109863,"predicted_mos":3.1229443550109863,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87000,"input_sample_rate":24000,"input_seconds":3.625,"input_float32_sha256":"0d51f1a8e4889add475d88d30ec622f935627787066bcff4954e3be3bac17371","internal_samples":58000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.536939776997315},"started_at":"2026-10-13T07:44:28.127786+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002854065998690203,"native_input_samples":86880,"native_input_sha256":"ed81f5ba0ebb3fc0d124045c3ac1909a0e16714f9d7ad8f46708b5f3a6861fc6","output_sha256":"bd8f7e5af65f6067ef8bcf43932b0795be7c7a11dde5d63d6744404914a810de","output_samples":87000,"length_delta_samples":120,"analysis_s":1.4012121480009228,"synthesis_s":8.646696169998904,"total_s":10.047908317999827,"rtf":2.7756652812154217,"output_peak":0.4593145251274109,"output_rms":0.054307700545875026,"finished_at":"2026-10-13T07:44:38.187394+00:00","analysis_stages":{"external_f0_s":0.9023129170018365},"evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"bd8f7e5af65f6067ef8bcf43932b0795be7c7a11dde5d63d6744404914a810de","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":57920,"output_samples":58000,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.298027269010266,"output_peak":0.4602407203764588},"common_prefix":{"samples":57920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986206896551724,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.535412485333505,"spectral_convergence":0.5101412663822665,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.182967022037858,"spectral_convergence":0.33202555964529096,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.298914257374629,"spectral_convergence":0.37266234467812054,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.0005065257682611005,"delta_db":0.7557554965301315,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.0011700558826706372,"delta_db":-0.5303538722069614,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0005977894306284384,"delta_db":-0.19814306119387756,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.00028921068738950006,"delta_db":-0.15542131119508648,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.00038626396911285667,"delta_db":-0.10929700967226758,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":6.7820029115978415e-06,"delta_db":-0.5902149510680265,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3898642836443918,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1732746559831361,"pitch":null,"output_rms_dbfs":-25.292033018827315,"output_peak":0.4602407203764588,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.672431254915331},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57920,"output_samples":57920,"intelligibility":{"estoi":{"value":0.8862617071432329,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055470500168298326,"rms_dbfs":-25.11875836284418,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.685647510165401,"mean":-0.00030342070437267997,"zero_energy":false},"output":{"rms":0.054374884700111946,"rms_dbfs":-25.292033018827315,"peak":0.4602407203764588,"peak_dbfs":-6.740299184341039,"crest_factor_db":18.551733834486278,"mean":-7.624968962492754e-05,"zero_energy":false},"rms_difference_db":-0.17327465598313552},"full_scale":{"reference":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":362,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":5.039671273196785e-07,"output_mean_power":3.601049043191334e-07,"output_rms_dbfs":-64.43570963871055,"output_zero_energy":false,"output_peak":0.01095088905182428,"output_energy_fraction":3.465460812368088e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":355,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":50,"both_estimated_voiced_frames":38,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.8444444444444444,"absolute_difference_cents_median":6.11868748600439,"absolute_difference_cents_p95":25.6396197052011,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00020370994921487513,"log_rms_mae_db":2.4195819956209172,"rms_envelope_nrmse":0.13867771739501286,"envelope_correlation":0.9824710830174106,"rms_gain_error_db":-0.17327429787045767,"reference_mean_power":0.003081231895334411,"output_mean_power":0.002960717393117207},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.3803128421607647e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003568424263155193,"output_power":0.0003613830207854347,"reference_active":true,"delta_db":0.054912601680078874},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014048200587458577,"output_power":0.00014852211189053183,"reference_active":true,"delta_db":0.24170416303927444},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7025084230927582e-05,"output_power":1.3204343171053202e-05,"reference_active":true,"delta_db":-1.1037246644946876}],"active_band_count":3,"active_band_mae_db":0.46678047640468034}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.00013932647455664127,"log_rms_mae_db":2.2093125825179842,"rms_envelope_nrmse":0.0797140382778189,"envelope_correlation":0.9941800510305325,"rms_gain_error_db":-0.1732722780303959,"reference_mean_power":0.0031070091895842884,"output_mean_power":0.002985487862960397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":1.2870957120852427e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031478645736461515,"output_power":0.00031920991691344587,"reference_active":true,"delta_db":0.06060335120586739},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.928842911180188e-05,"output_power":7.344143563004689e-05,"reference_active":true,"delta_db":0.25280442606109294},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4070650648317698e-06,"output_power":1.0789341141825905e-06,"reference_active":true,"delta_db":-1.1531925532140546}],"active_band_count":3,"active_band_mae_db":0.48886677682700497}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.0001235612100218032,"log_rms_mae_db":1.8883705036356542,"rms_envelope_nrmse":0.14023487251047673,"envelope_correlation":0.9851835886458927,"rms_gain_error_db":-0.33803365861241935,"reference_mean_power":0.0011039001556015875,"output_mean_power":0.001021236735561317},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":4.707425731708226e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018855725502559516,"output_power":0.00017836301811075816,"reference_active":true,"delta_db":-0.24138434690184132},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8855557263471545e-05,"output_power":5.568988159006259e-05,"reference_active":true,"delta_db":-0.24011181621616007},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.265121532076036e-06,"output_power":3.2407426633870153e-06,"reference_active":true,"delta_db":-1.1928686412149176}],"active_band_count":3,"active_band_mae_db":0.5581216014443063}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.0001123986305082509,"log_rms_mae_db":1.3921314296213823,"rms_envelope_nrmse":0.09154734956083914,"envelope_correlation":0.994086596069906,"rms_gain_error_db":-0.33803297958874434,"reference_mean_power":0.0011131375427710934,"output_mean_power":0.0010297825599335985},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":4.401493945076042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017713633289459936,"output_power":0.00016747640215500378,"reference_active":true,"delta_db":-0.24354027097147637},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.789895589253068e-05,"output_power":2.630687164043906e-05,"reference_active":true,"delta_db":-0.2551874453570765},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0644420897956184e-07,"output_power":2.683192082756313e-07,"reference_active":true,"delta_db":-1.8034917630233596}],"active_band_count":3,"active_band_mae_db":0.7674064931173041}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":2.0537628423408667,"rms_envelope_nrmse":0.1622208103428004,"envelope_correlation":0.9796257297265232,"rms_gain_error_db":-0.050769692188201446,"reference_mean_power":0.0005065055850916313,"output_mean_power":0.0005006189319605945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":2.2793822474038483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569673324300418,"output_power":0.00010821859522583826,"reference_active":true,"delta_db":0.10240327277974132},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5368696588833676e-05,"output_power":1.5867365431776523e-05,"reference_active":true,"delta_db":0.13867787057226702},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905327294990816e-06,"output_power":3.6765131969608444e-06,"reference_active":true,"delta_db":-0.26221307568191676}],"active_band_count":3,"active_band_mae_db":0.16776473967797503}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":1.31521887068533,"rms_envelope_nrmse":0.08946203278180814,"envelope_correlation":0.9935178560173747,"rms_gain_error_db":-0.050763076462705656,"reference_mean_power":0.0005107405079369303,"output_mean_power":0.0005048054051376527},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":2.126988285753868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.701418164823099e-05,"output_power":9.943518961144643e-05,"reference_active":true,"delta_db":0.10704881633740837},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.796784971798844e-06,"output_power":6.938250196438267e-06,"reference_active":true,"delta_db":0.0894642617100235},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1410692673718095e-07,"output_power":2.1785814428303287e-07,"reference_active":true,"delta_db":0.07543082274167462}],"active_band_count":3,"active_band_mae_db":0.09064796692970216}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":2.418185648734669,"rms_envelope_nrmse":0.2240682838741248,"envelope_correlation":0.9646290160783915,"rms_gain_error_db":-0.7067339743262169,"reference_mean_power":0.00014114001095811158,"output_mean_power":0.00011994350878480648},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.582618674096986e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4831777511599814e-05,"output_power":2.92515798284934e-05,"reference_active":true,"delta_db":-0.7582631108047911},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4244491686748363e-05,"output_power":1.2369387331546399e-05,"reference_active":true,"delta_db":-0.612987669868703},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318870090030596e-06,"output_power":2.354080104706979e-06,"reference_active":true,"delta_db":0.06544818087725189}],"active_band_count":3,"active_band_mae_db":0.47889965385024863}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":1.3749784744380886,"rms_envelope_nrmse":0.1316422616377734,"envelope_correlation":0.9903805542682056,"rms_gain_error_db":-0.7067578162698868,"reference_mean_power":0.00014231959462201218,"output_mean_power":0.00012094527780889041},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":7.830084455082623e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3431441965989596e-05,"output_power":2.8380002918905507e-05,"reference_active":true,"delta_db":-0.711426732106051},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2151972914162375e-06,"output_power":6.3849981737322065e-06,"reference_active":true,"delta_db":-0.5308743350323583},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7745463336154605e-07,"output_power":1.9093843612096314e-07,"reference_active":true,"delta_db":0.31806017847418494}],"active_band_count":3,"active_band_mae_db":0.5201204152041982}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0932328701019287,"predicted_mos":3.0932328701019287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87120,"input_sample_rate":24000,"input_seconds":3.63,"input_float32_sha256":"ea85fddbb3aebd1d404c338bb77d94f40a2a2cf5c226832ee811f7f898698444","internal_samples":58080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.240837653998824},"started_at":"2026-10-13T05:11:20.168964+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0020873920002486557,"native_input_samples":86880,"native_input_sha256":"ed81f5ba0ebb3fc0d124045c3ac1909a0e16714f9d7ad8f46708b5f3a6861fc6","output_sha256":"b738e04fc527104f6177831b0fad8f30ab0b9fcaa7620b723e2d4348ca119a42","output_samples":87120,"length_delta_samples":240,"analysis_s":0.5622666389990627,"synthesis_s":0.08035056600056123,"total_s":0.6426172049996239,"rtf":0.17751856491702317,"output_peak":0.3954603672027588,"output_rms":0.05404912640129906,"finished_at":"2026-10-13T05:11:20.819589+00:00","analysis_stages":{"external_f0_s":0.07532156199886231},"evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"b738e04fc527104f6177831b0fad8f30ab0b9fcaa7620b723e2d4348ca119a42","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":57920,"output_samples":58080,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.33947611264867,"output_peak":0.38649477180383646},"common_prefix":{"samples":57920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9972451790633609,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.861651641439548,"spectral_convergence":0.25131125297663043,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.374176594811087,"spectral_convergence":0.21564944019677124,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.812232852505267,"spectral_convergence":0.2205088694597472,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.00045195329862184996,"delta_db":0.26067566201849407,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.0012746496571376853,"delta_db":-0.1585115817249977,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0005825109729945144,"delta_db":-0.31058427478239975,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.0002646569373118817,"delta_db":-0.5407318639596672,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.00035210852922669064,"delta_db":-0.5113735456673046,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":6.704330046423102e-06,"delta_db":-0.6402408103278914,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4036862897467925,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20873738224888144,"pitch":null,"output_rms_dbfs":-25.327495745093064,"output_peak":0.38649477180383646,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.6826870295853},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57920,"output_samples":57920,"intelligibility":{"estoi":{"value":0.9164118038707766,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055470500168298326,"rms_dbfs":-25.11875836284418,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.685647510165401,"mean":-0.00030342070437267997,"zero_energy":false},"output":{"rms":0.05415333564846251,"rms_dbfs":-25.327495745093064,"peak":0.38649477180383646,"peak_dbfs":-8.257127529976584,"crest_factor_db":17.070368215116474,"mean":0.00030229867364098745,"zero_energy":false},"rms_difference_db":-0.20873738224888427},"full_scale":{"reference":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":362,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":5.039671273196785e-07,"output_mean_power":5.199150388901446e-07,"output_rms_dbfs":-62.840676201264486,"output_zero_energy":false,"output_peak":0.01613627539699218,"output_energy_fraction":5.04441267839359e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":355,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":43,"both_estimated_voiced_frames":35,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.7777777777777778,"absolute_difference_cents_median":1.080924501262109,"absolute_difference_cents_p95":5.740391594033105,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00020370994921487513,"log_rms_mae_db":1.1459266438391325,"rms_envelope_nrmse":0.08955708030999925,"envelope_correlation":0.9930366227075595,"rms_gain_error_db":-0.2087374136970929,"reference_mean_power":0.003081231895334411,"output_mean_power":0.00293663954974454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.3803128421607647e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003568424263155193,"output_power":0.00033048300615020254,"reference_active":true,"delta_db":-0.33327351220652424},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014048200587458577,"output_power":0.00015335809011708738,"reference_active":true,"delta_db":0.3808599162285885},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7025084230927582e-05,"output_power":1.1352651198493531e-05,"reference_active":true,"delta_db":-1.7599197455077937}],"active_band_count":3,"active_band_mae_db":0.8246843913143022}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.00013932647455664127,"log_rms_mae_db":0.8621415643383806,"rms_envelope_nrmse":0.04968392016692501,"envelope_correlation":0.9981561640915151,"rms_gain_error_db":-0.20873602241699363,"reference_mean_power":0.0031070091895842884,"output_mean_power":0.0029612081467587825},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":1.2870957120852427e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031478645736461515,"output_power":0.00029403848632043037,"reference_active":true,"delta_db":-0.29611861639692755},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.928842911180188e-05,"output_power":7.579908627320047e-05,"reference_active":true,"delta_db":0.3900325518226959},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4070650648317698e-06,"output_power":9.92407686299215e-07,"reference_active":true,"delta_db":-1.5162406106896167}],"active_band_count":3,"active_band_mae_db":0.7341305929697467}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.0001235612100218032,"log_rms_mae_db":1.0054879458051091,"rms_envelope_nrmse":0.07909996972940989,"envelope_correlation":0.9952866925684513,"rms_gain_error_db":-0.16980169632526934,"reference_mean_power":0.0011039001556015875,"output_mean_power":0.0010615724152027509},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":4.707425731708226e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018855725502559516,"output_power":0.00017778318027926037,"reference_active":true,"delta_db":-0.2555257637837254},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8855557263471545e-05,"output_power":5.230321274879328e-05,"reference_active":true,"delta_db":-0.5125910967492349},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.265121532076036e-06,"output_power":4.026150567748171e-06,"reference_active":true,"delta_db":-0.2504139827949455}],"active_band_count":3,"active_band_mae_db":0.33951028110930187}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.0001123986305082509,"log_rms_mae_db":0.62591587277551,"rms_envelope_nrmse":0.044296428909613716,"envelope_correlation":0.9986650561499576,"rms_gain_error_db":-0.16980154621820054,"reference_mean_power":0.0011131375427710934,"output_mean_power":0.001070455642730816},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":4.401493945076042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017713633289459936,"output_power":0.0001652969591324817,"reference_active":true,"delta_db":-0.3004278540552118},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.789895589253068e-05,"output_power":2.4168296104893493e-05,"reference_active":true,"delta_db":-0.6234191709457486},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0644420897956184e-07,"output_power":3.141016899530806e-07,"reference_active":true,"delta_db":-1.1193066707379908}],"active_band_count":3,"active_band_mae_db":0.6810512319129837}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":1.1388297135343621,"rms_envelope_nrmse":0.12327163165551193,"envelope_correlation":0.9898629734333669,"rms_gain_error_db":-0.46242028439083804,"reference_mean_power":0.0005065055850916313,"output_mean_power":0.00045534671743688976},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":2.2793822474038483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569673324300418,"output_power":9.5608887861044e-05,"reference_active":true,"delta_db":-0.43563298389357585},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5368696588833676e-05,"output_power":1.3368976987482204e-05,"reference_active":true,"delta_db":-0.605388610503458},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905327294990816e-06,"output_power":3.5131432729640816e-06,"reference_active":true,"delta_db":-0.45961575247378145}],"active_band_count":3,"active_band_mae_db":0.5002124489569385}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":0.7793601652839955,"rms_envelope_nrmse":0.07587130869161834,"envelope_correlation":0.9974322749855128,"rms_gain_error_db":-0.462416566806178,"reference_mean_power":0.0005107405079369303,"output_mean_power":0.00045915429103357},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":2.126988285753868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.701418164823099e-05,"output_power":8.856839303561803e-05,"reference_active":true,"delta_db":-0.3955645957885775},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.796784971798844e-06,"output_power":5.726225162127038e-06,"reference_active":true,"delta_db":-0.7443510946499056},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1410692673718095e-07,"output_power":1.8253424935811467e-07,"reference_active":true,"delta_db":-0.6928635351608435}],"active_band_count":3,"active_band_mae_db":0.6109264085331089}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":1.5179491497802364,"rms_envelope_nrmse":0.16382070871316173,"envelope_correlation":0.9842768525054093,"rms_gain_error_db":-0.7778510553725695,"reference_mean_power":0.00014114001095811158,"output_mean_power":0.00011799539033507746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.582618674096986e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4831777511599814e-05,"output_power":3.0462930565896395e-05,"reference_active":true,"delta_db":-0.5820395704982568},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4244491686748363e-05,"output_power":1.0672212292806807e-05,"reference_active":true,"delta_db":-1.2539250040571392},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318870090030596e-06,"output_power":2.227888721092897e-06,"reference_active":true,"delta_db":-0.17382924016516338}],"active_band_count":3,"active_band_mae_db":0.6699312715735198}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":1.0415523002595002,"rms_envelope_nrmse":0.10821021450967225,"envelope_correlation":0.9962185332187944,"rms_gain_error_db":-0.7778477786054205,"reference_mean_power":0.00014231959462201218,"output_mean_power":0.00011898163161658666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":7.830084455082623e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3431441965989596e-05,"output_power":2.9132511035003256e-05,"reference_active":true,"delta_db":-0.5977718936862176},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2151972914162375e-06,"output_power":5.492078185702021e-06,"reference_active":true,"delta_db":-1.1851149962100977},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7745463336154605e-07,"output_power":1.760909050537487e-07,"reference_active":true,"delta_db":-0.03350417708134357}],"active_band_count":3,"active_band_mae_db":0.605463688992553}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7127-75947-0011","method":"magphase","status":"measured","mos":{"status":"measured","value":3.2893471717834473,"predicted_mos":3.2893471717834473,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":174530,"input_sample_rate":48000,"input_seconds":3.636041666666667,"input_float32_sha256":"f6bd5dcf444c8b2833cb21fa84cd99e0e8983fb4733053ca42f129b25ee26302","internal_samples":58177,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3366792679989885},"started_at":"2026-10-13T05:17:55.661514+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00635515100111661,"native_input_samples":173760,"native_input_sha256":"a86e389ad53c8f3a9814bc94fa53ab3b579e5e2a9bfe7755b955755071e09264","output_sha256":"db118b8a232c9d7f5c4efef93491e8f1de862c933b8ba3845b1da329f44f5ddb","output_samples":174530,"length_delta_samples":770,"analysis_s":1.70099108300019,"synthesis_s":0.6795466410003428,"total_s":2.3805377240005328,"rtf":0.6576071060774952,"output_peak":0.3813620880012448,"output_rms":0.052753643599238974,"finished_at":"2026-10-13T05:17:58.056031+00:00","evaluation_input":{"reference":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"output_sha256":"db118b8a232c9d7f5c4efef93491e8f1de862c933b8ba3845b1da329f44f5ddb","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":57920,"output_samples":58177,"length_delta_samples":257,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.550160269220726,"output_peak":0.38178760340357265},"common_prefix":{"samples":57920,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.995582446671365,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":57920,"output_samples":57920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.679319378271997,"spectral_convergence":0.591598626312381,"spectral_convergence_status":"ok","amplitude_floor":1.9579592332350965e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.275723003484921,"spectral_convergence":0.36941866048875915,"spectral_convergence_status":"ok","amplitude_floor":1.206925486771671e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.820294823802717,"spectral_convergence":0.3330421724641809,"spectral_convergence_status":"ok","amplitude_floor":6.395923898397892e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004256238876774699,"output_power":0.00037668612084314695,"delta_db":-0.5304637553897193,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001322032073723835,"output_power":0.0011927991728667565,"delta_db":-0.4467466237583111,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006256947892797981,"output_power":0.0005533748682881343,"delta_db":-0.53343106615905,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002997481206139804,"output_power":0.00027919152430126194,"delta_db":-0.3085423896772994,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003961082609243847,"output_power":0.0003878295331682039,"delta_db":-0.09173022335197649,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":7.769256628656105e-06,"output_power":8.498395455100717e-06,"delta_db":0.3895746921435781,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.38341479174665577,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.41217535204596784,"pitch":null,"output_rms_dbfs":-25.530933714890146,"output_peak":0.38178760340357265,"reference_rms_dbfs":-25.11875836284418,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.591779068519879},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":57920,"output_samples":57920,"intelligibility":{"estoi":{"value":0.8118479234624033,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.055470500168298326,"rms_dbfs":-25.11875836284418,"peak":0.337554931640625,"peak_dbfs":-9.43311085267878,"crest_factor_db":15.685647510165401,"mean":-0.00030342070437267997,"zero_energy":false},"output":{"rms":0.0528997128072946,"rms_dbfs":-25.530933714890146,"peak":0.38178760340357265,"peak_dbfs":-8.363563544034418,"crest_factor_db":17.16737017085573,"mean":0.00037811339667391585,"zero_energy":false},"rms_difference_db":-0.412175352045967},"full_scale":{"reference":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":57920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":362,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.1497743409127e-06,"threshold_dbfs":-53.819755189893485,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":103,"quiet_samples":16480,"reference_mean_power":5.039671273196785e-07,"output_mean_power":7.523977200668007e-07,"output_rms_dbfs":-61.23552529128979,"output_zero_energy":false,"output_peak":0.018111706244461682,"output_energy_fraction":7.650142000692795e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":355,"reference_estimated_voiced_frames":45,"output_estimated_voiced_frames":40,"both_estimated_voiced_frames":34,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.7555555555555555,"absolute_difference_cents_median":9.803455337167012,"absolute_difference_cents_p95":24.92764875243008,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":57920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.00020370994921487513,"log_rms_mae_db":2.426158375685147,"rms_envelope_nrmse":0.2151087011750454,"envelope_correlation":0.9576083859435893,"rms_gain_error_db":-0.41217536805439,"reference_mean_power":0.003081231895334411,"output_mean_power":0.002802249807487836},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":1.3803128421607647e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003568424263155193,"output_power":0.0003320794289903023,"reference_active":true,"delta_db":-0.3123451004775817},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014048200587458577,"output_power":0.00011840876070802799,"reference_active":true,"delta_db":-0.7423686396907437},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7025084230927582e-05,"output_power":1.6023307224280415e-05,"reference_active":true,"delta_db":-0.26337109493019994}],"active_band_count":3,"active_band_mae_db":0.43936161169950844}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.00013932647455664127,"log_rms_mae_db":1.7296390419095438,"rms_envelope_nrmse":0.13353263497137546,"envelope_correlation":0.9840956615375259,"rms_gain_error_db":-0.4121752643379845,"reference_mean_power":0.0031070091895842884,"output_mean_power":0.002825693231491866},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":1.2870957120852427e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031478645736461515,"output_power":0.0002932958782465918,"reference_active":true,"delta_db":-0.30710080235437315},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.928842911180188e-05,"output_power":5.855296488587032e-05,"reference_active":true,"delta_db":-0.7311182435476202},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4070650648317698e-06,"output_power":1.231010629483383e-06,"reference_active":true,"delta_db":-0.5805237736057347}],"active_band_count":3,"active_band_mae_db":0.5395809398359094}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":0.0001235612100218032,"log_rms_mae_db":2.4106625119311187,"rms_envelope_nrmse":0.22042520484352318,"envelope_correlation":0.9620410510291615,"rms_gain_error_db":-0.46007215169958254,"reference_mean_power":0.0011039001556015875,"output_mean_power":0.0009929390238422665},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":4.707425731708226e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018855725502559516,"output_power":0.00017162874802394983,"reference_active":true,"delta_db":-0.40853212740717343},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8855557263471545e-05,"output_power":5.172966378961014e-05,"reference_active":true,"delta_db":-0.5604782032262169},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.265121532076036e-06,"output_power":3.8555529370418644e-06,"reference_active":true,"delta_db":-0.43844740241367075}],"active_band_count":3,"active_band_mae_db":0.46915257768235374}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":0.0001123986305082509,"log_rms_mae_db":1.6884745973478106,"rms_envelope_nrmse":0.1573710905425797,"envelope_correlation":0.9808178671966785,"rms_gain_error_db":-0.46007193067547514,"reference_mean_power":0.0011131375427710934,"output_mean_power":0.0010012479441761815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":4.401493945076042e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00017713633289459936,"output_power":0.00016147458506240149,"reference_active":true,"delta_db":-0.40203472379398875},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.789895589253068e-05,"output_power":2.456734280295677e-05,"reference_active":true,"delta_db":-0.5522976442883488},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.0644420897956184e-07,"output_power":3.23094868402445e-07,"reference_active":true,"delta_db":-0.9967087976969612}],"active_band_count":3,"active_band_mae_db":0.6503470552597662}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":8.851421864498589e-05,"log_rms_mae_db":2.6025833678178465,"rms_envelope_nrmse":0.21335697264906345,"envelope_correlation":0.9646153186833145,"rms_gain_error_db":-0.12313320447512047,"reference_mean_power":0.0005065055850916313,"output_mean_power":0.0004923465716119046},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":2.2793822474038483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010569673324300418,"output_power":0.00010041387151284871,"reference_active":true,"delta_db":-0.2226785295801353},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5368696588833676e-05,"output_power":1.4507433680753495e-05,"reference_active":true,"delta_db":-0.2504644286394989},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.905327294990816e-06,"output_power":3.999345246617663e-06,"reference_active":true,"delta_db":0.103314597712926}],"active_band_count":3,"active_band_mae_db":0.19215251864418673}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":8.086137863012641e-05,"log_rms_mae_db":1.6289665588289413,"rms_envelope_nrmse":0.12799717502177574,"envelope_correlation":0.9867093931937446,"rms_gain_error_db":-0.12312484791263842,"reference_mean_power":0.0005107405079369303,"output_mean_power":0.0004964640653968536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":2.126988285753868e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.701418164823099e-05,"output_power":9.197752135921074e-05,"reference_active":true,"delta_db":-0.2315352270599042},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.796784971798844e-06,"output_power":6.378927292708576e-06,"reference_active":true,"delta_db":-0.27555878401509076},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1410692673718095e-07,"output_power":2.1149422588301305e-07,"reference_active":true,"delta_db":-0.05332202748183251}],"active_band_count":3,"active_band_mae_db":0.18680534618560918}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":723,"rms_floor":5.208111658419099e-05,"log_rms_mae_db":3.2514590969564217,"rms_envelope_nrmse":0.3185292051320922,"envelope_correlation":0.9228773269118759,"rms_gain_error_db":-0.4982046876554863,"reference_mean_power":0.00014114001095811158,"output_mean_power":0.0001258431782246132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":723,"frequency_resolution_hz":0.276625172890733,"power_floor":8.582618674096986e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.4831777511599814e-05,"output_power":2.9376862276558137e-05,"reference_active":true,"delta_db":-0.7397023042279718},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4244491686748363e-05,"output_power":1.257789879024242e-05,"reference_active":true,"delta_db":-0.5403886029963144},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.318870090030596e-06,"output_power":2.1733515481261226e-06,"reference_active":true,"delta_db":-0.2814643802025717}],"active_band_count":3,"active_band_mae_db":0.5205184291422861}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":57920,"uncovered_tail_samples":0,"metrics":{"frame_count":717,"rms_floor":4.677450134765528e-05,"log_rms_mae_db":1.909125761461891,"rms_envelope_nrmse":0.19763723911740447,"envelope_correlation":0.9697278865899503,"rms_gain_error_db":-0.49824740089290254,"reference_mean_power":0.00014231959462201218,"output_mean_power":0.00012689366993989996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":717,"frequency_resolution_hz":0.2789400278940028,"power_floor":7.830084455082623e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3431441965989596e-05,"output_power":2.8451550611426715e-05,"reference_active":true,"delta_db":-0.7004916854494652},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.2151972914162375e-06,"output_power":6.4278803177407426e-06,"reference_active":true,"delta_db":-0.5018042895117466},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7745463336154605e-07,"output_power":1.7940882748205494e-07,"reference_active":true,"delta_db":0.04756464584499156}],"active_band_count":3,"active_band_mae_db":0.4166202069354011}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"hnm","status":"measured","mos":{"status":"measured","value":2.9683303833007812,"predicted_mos":2.9683303833007812,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"347606a6cc46ec9271c3db04af6f5f0ed7cb5ec7d69e9ddbf97567e13f85ea28","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4489514580000105},"started_at":"2026-10-13T04:40:05.749357+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.336799959972268e-05,"native_input_samples":96000,"native_input_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","output_sha256":"7e54afcb23efa15fc90f9286490ceb0f3357f7b1f7967029264b3b6fe6f21e34","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1424364510003215,"synthesis_s":0.4960836999998719,"total_s":1.6385201510001934,"rtf":0.2730866918333656,"output_peak":0.6547071696621664,"output_rms":0.0828299917087497,"finished_at":"2026-10-13T04:40:07.394077+00:00","evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"7e54afcb23efa15fc90f9286490ceb0f3357f7b1f7967029264b3b6fe6f21e34","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.268535307199305,"spectral_convergence":0.23193179952197146,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5500849044688945,"spectral_convergence":0.22121776283814304,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.563769022093351,"spectral_convergence":0.24572858504240702,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.002328480275391046,"delta_db":-0.021635045168318347,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.0032998345792126455,"delta_db":-0.15347462557568814,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.000966499337115711,"delta_db":-0.0019343884891735504,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.0001527874869275682,"delta_db":-0.37165723593069855,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":7.170452456639419e-05,"delta_db":-0.28455072372394213,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":4.150132325817823e-05,"delta_db":0.07780621448876275,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15184303889609727,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09301484720912424,"pitch":null,"output_rms_dbfs":-21.636247642275656,"output_peak":0.6547071696621664,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.794129744587184},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.268535307199305,"spectral_convergence":0.23193179952197146,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5500849044688945,"spectral_convergence":0.22121776283814304,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.563769022093351,"spectral_convergence":0.24572858504240702,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.002328480275391046,"delta_db":-0.021635045168318347,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.0032998345792126455,"delta_db":-0.15347462557568814,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.000966499337115711,"delta_db":-0.0019343884891735504,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.0001527874869275682,"delta_db":-0.37165723593069855,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":7.170452456639419e-05,"delta_db":-0.28455072372394213,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":4.150132325817823e-05,"delta_db":0.07780621448876275,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15184303889609727,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09301484720912424,"pitch":null,"output_rms_dbfs":-21.636247642275656,"output_peak":0.6547071696621664,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.794129744587184},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8957919683375861,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08372176207031103,"rms_dbfs":-21.54323279506653,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.494273826613313,"mean":-2.5322596232096355e-05,"zero_energy":false},"output":{"rms":0.0828299917087497,"rms_dbfs":-21.636247642275656,"peak":0.6547071696621664,"peak_dbfs":-3.6790580623400686,"crest_factor_db":17.957189579935587,"mean":0.0001411502237529396,"zero_energy":false},"rms_difference_db":-0.09301484720912612},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":8.157722793901835e-06,"output_rms_dbfs":-50.88431056450184,"output_zero_energy":false,"output_peak":0.06429524180494219,"output_energy_fraction":0.00012683012818225935,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":154,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.9625,"absolute_difference_cents_median":1.0822435987304977,"absolute_difference_cents_p95":5.912548546790793,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019906122792766575,"log_rms_mae_db":0.990261667551667,"rms_envelope_nrmse":0.07575872112930529,"envelope_correlation":0.9940121003100458,"rms_gain_error_db":-0.09259337711270799,"reference_mean_power":0.007012794178240996,"output_mean_power":0.006864861110889599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.192818590388021e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015356197181510699,"output_power":0.0015290578349359376,"reference_active":true,"delta_db":-0.01859767611293924},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003622112164404606,"output_power":0.0003663513621806556,"reference_active":true,"delta_db":0.049359158886926824},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6887620873910397e-05,"output_power":4.413969077030479e-05,"reference_active":true,"delta_db":-0.2622891045253853}],"active_band_count":3,"active_band_mae_db":0.11008197984175044}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018229956701751592,"log_rms_mae_db":0.5597813783920854,"rms_envelope_nrmse":0.046405053271748674,"envelope_correlation":0.9975567586828756,"rms_gain_error_db":-0.08958291482919935,"reference_mean_power":0.0070179714231847195,"output_mean_power":0.006874692922892258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8443307374696498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014276197130393435,"output_power":0.0014231172631131719,"reference_active":true,"delta_db":-0.013718493149301755},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018503527061359352,"output_power":0.00019112136299422614,"reference_active":true,"delta_db":0.14054714344108762},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.327921114523829e-06,"output_power":3.668576895843077e-06,"reference_active":true,"delta_db":-0.7178170962807078}],"active_band_count":3,"active_band_mae_db":0.2906942442903657}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015227518148493586,"log_rms_mae_db":1.6849852727050096,"rms_envelope_nrmse":0.1247297065364731,"envelope_correlation":0.9870917148925845,"rms_gain_error_db":-0.1727506992944844,"reference_mean_power":0.0024461505543329222,"output_mean_power":0.0023507590314477095},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3707367109300066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00065188242251555,"output_power":0.000625919123373351,"reference_active":true,"delta_db":-0.1765105027335604},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000159373128948768,"output_power":0.00014496373343703142,"reference_active":true,"delta_db":-0.41155733741779377},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8816561344170712e-05,"output_power":1.4305599475694086e-05,"reference_active":true,"delta_db":-1.1903419903961252}],"active_band_count":3,"active_band_mae_db":0.5928032768491598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013579343349222632,"log_rms_mae_db":1.0272821112791828,"rms_envelope_nrmse":0.07178509780084248,"envelope_correlation":0.9955569884951523,"rms_gain_error_db":-0.1693516016644394,"reference_mean_power":0.0024424650949128645,"output_mean_power":0.0023490551107415865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.250982515304928e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006212562593255464,"output_power":0.0005954096059202473,"reference_active":true,"delta_db":-0.18454940429029246},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.001783099426803e-05,"output_power":7.18493225922531e-05,"reference_active":true,"delta_db":-0.46764097099527585},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7444705804029888e-06,"output_power":1.187470866209985e-06,"reference_active":true,"delta_db":-1.6704068638544998}],"active_band_count":3,"active_band_mae_db":0.7741990797133561}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":1.9975684098101316,"rms_envelope_nrmse":0.19805269707805717,"envelope_correlation":0.968189929878265,"rms_gain_error_db":-0.38022206512105705,"reference_mean_power":0.00019825791077017447,"output_mean_power":0.0001816386723206644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.446126492709316e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.480334346914859e-05,"output_power":3.054398226890752e-05,"reference_active":true,"delta_db":-0.5669530858548911},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0911626587448623e-05,"output_power":8.58277501633151e-06,"reference_active":true,"delta_db":-1.0426176700377352},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.788733059913985e-06,"output_power":2.0023040942352317e-06,"reference_active":true,"delta_db":-1.438769097987736}],"active_band_count":3,"active_band_mae_db":1.0161132846267875}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":1.2168056212968403,"rms_envelope_nrmse":0.1072718123720188,"envelope_correlation":0.9908960748024865,"rms_gain_error_db":-0.39203205508705263,"reference_mean_power":0.00019869265760544442,"output_mean_power":0.00018154262599741358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.64113687643322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.340875190885912e-05,"output_power":2.8938889928214416e-05,"reference_active":true,"delta_db":-0.6237838355024237},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7578298997729145e-06,"output_power":4.475658245017679e-06,"reference_active":true,"delta_db":-1.0940191361652858},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2467549734271305e-07,"output_power":1.5795489453129218e-07,"reference_active":true,"delta_db":-1.5302262383402656}],"active_band_count":3,"active_band_mae_db":1.0826764033359917}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":2.3850254392678205,"rms_envelope_nrmse":0.212571929145699,"envelope_correlation":0.9682225926716441,"rms_gain_error_db":0.17037038425554518,"reference_mean_power":3.3091629565415505e-05,"output_mean_power":3.441558788341772e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.646064632792655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.725047198052175e-06,"output_power":7.003267404424344e-06,"reference_active":true,"delta_db":0.17605372576279504},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.390363033635178e-06,"output_power":2.216281363151219e-06,"reference_active":true,"delta_db":-0.328389694698289},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2031554280127926e-06,"output_power":8.485861993557015e-07,"reference_active":true,"delta_db":-1.5162577019908459}],"active_band_count":3,"active_band_mae_db":0.6735670408173099}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":1.7065748017240658,"rms_envelope_nrmse":0.15224055051657623,"envelope_correlation":0.9822196869845571,"rms_gain_error_db":0.16875728948506297,"reference_mean_power":3.324742134736325e-05,"output_mean_power":3.456477198044828e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5226373785787838e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9685546500066154e-06,"output_power":7.1348574916683125e-06,"reference_active":true,"delta_db":0.10242592850432448},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4829248214027976e-06,"output_power":1.2646881427129605e-06,"reference_active":true,"delta_db":-0.6913568774333899},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.259963132949634e-08,"output_power":6.313412781259779e-08,"reference_active":true,"delta_db":-1.6634507247229904}],"active_band_count":3,"active_band_mae_db":0.8190778435535683}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9395482540130615,"predicted_mos":2.9395482540130615,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6d6e78479866992aa1fa2c066ee99207a2007b9effaf560b7858a69b2b319936","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.598763742000301},"started_at":"2026-10-13T04:43:18.374560+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.998999960051151e-05,"native_input_samples":96000,"native_input_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","output_sha256":"b5348f6fee4f8d903fa1c9d77882528d9f7f7a24a4044f4f4ba0560a6ea76742","output_samples":96000,"length_delta_samples":0,"analysis_s":1.98617611499958,"synthesis_s":0.08309816099972522,"total_s":2.0692742759993052,"rtf":0.3448790459998842,"output_peak":0.7665293988151443,"output_rms":0.08792679860955356,"finished_at":"2026-10-13T04:43:20.452843+00:00","evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"b5348f6fee4f8d903fa1c9d77882528d9f7f7a24a4044f4f4ba0560a6ea76742","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.19099166512416,"spectral_convergence":0.7502853744954884,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.294618166109782,"spectral_convergence":0.25863235542297197,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.467839743138967,"spectral_convergence":0.23223651228048292,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.002377329651024989,"delta_db":0.06853339211236367,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.003989425595575746,"delta_db":0.6707073792611167,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0010591012597066336,"delta_db":0.39542485244457837,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.00018746946762309468,"delta_db":0.5167703487254312,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":8.04201125965292e-05,"delta_db":0.21363043872180998,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":3.73758271979993e-05,"delta_db":-0.3769050964297454,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3736619179491742,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.42565801314352336,"pitch":null,"output_rms_dbfs":-21.117574781923008,"output_peak":0.7665293988151443,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9844831914576355},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.19099166512416,"spectral_convergence":0.7502853744954884,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.294618166109782,"spectral_convergence":0.25863235542297197,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.467839743138967,"spectral_convergence":0.23223651228048292,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.002377329651024989,"delta_db":0.06853339211236367,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.003989425595575746,"delta_db":0.6707073792611167,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0010591012597066336,"delta_db":0.39542485244457837,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.00018746946762309468,"delta_db":0.5167703487254312,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":8.04201125965292e-05,"delta_db":0.21363043872180998,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":3.73758271979993e-05,"delta_db":-0.3769050964297454,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3736619179491742,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.42565801314352336,"pitch":null,"output_rms_dbfs":-21.117574781923008,"output_peak":0.7665293988151443,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.9844831914576355},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9234571210634849,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08372176207031103,"rms_dbfs":-21.54323279506653,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.494273826613313,"mean":-2.5322596232096355e-05,"zero_energy":false},"output":{"rms":0.08792679860955357,"rms_dbfs":-21.117574781923008,"peak":0.7665293988151443,"peak_dbfs":-2.309423678596636,"crest_factor_db":18.80815110332637,"mean":1.8647653880366847e-06,"zero_energy":false},"rms_difference_db":0.42565801314352214},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":4.442072405915009e-06,"output_rms_dbfs":-53.524143667166165,"output_zero_energy":false,"output_peak":0.05519967946466258,"output_energy_fraction":6.128748995534063e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":193,"both_estimated_voiced_frames":146,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":47,"paired_voiced_fraction":0.9125,"absolute_difference_cents_median":5.035140352125072,"absolute_difference_cents_p95":18.489497856549125,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019906122792766575,"log_rms_mae_db":1.7036528653901362,"rms_envelope_nrmse":0.14611571786209923,"envelope_correlation":0.9810616811690349,"rms_gain_error_db":0.42712900790576613,"reference_mean_power":0.007012794178240996,"output_mean_power":0.007737559355738536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.192818590388021e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015356197181510699,"output_power":0.0017074103224306453,"reference_active":true,"delta_db":0.4605422264565294},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003622112164404606,"output_power":0.00034834917103999805,"reference_active":true,"delta_db":-0.16947113409143652},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6887620873910397e-05,"output_power":2.9501885839474552e-05,"reference_active":true,"delta_db":-2.0120841861866983}],"active_band_count":3,"active_band_mae_db":0.8806991822448881}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018229956701751592,"log_rms_mae_db":1.0897116299145806,"rms_envelope_nrmse":0.0923309774693818,"envelope_correlation":0.9933190778097908,"rms_gain_error_db":0.4324224941046212,"reference_mean_power":0.0070179714231847195,"output_mean_power":0.007752715461317712},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8443307374696498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014276197130393435,"output_power":0.0015812417709963082,"reference_active":true,"delta_db":0.44385742191350874},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018503527061359352,"output_power":0.00017734839368400493,"reference_active":true,"delta_db":-0.1842726036230274},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.327921114523829e-06,"output_power":2.227485192008198e-06,"reference_active":true,"delta_db":-2.8846451064608636}],"active_band_count":3,"active_band_mae_db":1.1709250439991332}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015227518148493586,"log_rms_mae_db":2.1068130064849497,"rms_envelope_nrmse":0.15143095502769488,"envelope_correlation":0.9861595978398494,"rms_gain_error_db":0.601684771653656,"reference_mean_power":0.0024461505543329222,"output_mean_power":0.0028096463608649567},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3707367109300066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00065188242251555,"output_power":0.0007404417961586432,"reference_active":true,"delta_db":0.5532165485465561},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000159373128948768,"output_power":0.00016177645493697365,"reference_active":true,"delta_db":0.06500215178786557},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8816561344170712e-05,"output_power":1.2242691762188743e-05,"reference_active":true,"delta_db":-1.8666334538666949}],"active_band_count":3,"active_band_mae_db":0.8282840514003723}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013579343349222632,"log_rms_mae_db":1.4289418006651626,"rms_envelope_nrmse":0.11135105694114082,"envelope_correlation":0.9938615891526484,"rms_gain_error_db":0.6036397380115803,"reference_mean_power":0.0024424650949128645,"output_mean_power":0.0028066763798650874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.250982515304928e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006212562593255464,"output_power":0.0006970508759005083,"reference_active":true,"delta_db":0.49993699862559726},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.001783099426803e-05,"output_power":8.026087108605956e-05,"reference_active":true,"delta_db":0.013170939154837184},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7444705804029888e-06,"output_power":9.899532071915659e-07,"reference_active":true,"delta_db":-2.460489827107138}],"active_band_count":3,"active_band_mae_db":0.991199254962524}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":1.9711400542601356,"rms_envelope_nrmse":0.1651127753596673,"envelope_correlation":0.9813351549456506,"rms_gain_error_db":0.4555737292166102,"reference_mean_power":0.00019825791077017447,"output_mean_power":0.00022018508552186896},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.446126492709316e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.480334346914859e-05,"output_power":3.940081643225866e-05,"reference_active":true,"delta_db":0.5388425349951622},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0911626587448623e-05,"output_power":1.0399101125933167e-05,"reference_active":true,"delta_db":-0.20893693827090828},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.788733059913985e-06,"output_power":1.44022731443517e-06,"reference_active":true,"delta_db":-2.86975901973545}],"active_band_count":3,"active_band_mae_db":1.20584616433384}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":1.2563281137080293,"rms_envelope_nrmse":0.09662353424001661,"envelope_correlation":0.9945484364127606,"rms_gain_error_db":0.4570744122821094,"reference_mean_power":0.00019869265760544442,"output_mean_power":0.00022074417890810427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.64113687643322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.340875190885912e-05,"output_power":3.676668356077564e-05,"reference_active":true,"delta_db":0.4159420579963775},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7578298997729145e-06,"output_power":5.172192920740038e-06,"reference_active":true,"delta_db":-0.46584115099727513},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2467549734271305e-07,"output_power":1.0342477849096653e-07,"reference_active":true,"delta_db":-3.3693111229752337}],"active_band_count":3,"active_band_mae_db":1.4170314439896288}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":2.1861740438135446,"rms_envelope_nrmse":0.24679761023278343,"envelope_correlation":0.9552827942979721,"rms_gain_error_db":-0.1534633317575728,"reference_mean_power":3.3091629565415505e-05,"output_mean_power":3.194271458074531e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.646064632792655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.725047198052175e-06,"output_power":5.827142711808372e-06,"reference_active":true,"delta_db":-0.6223968220060224},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.390363033635178e-06,"output_power":2.116613478020597e-06,"reference_active":true,"delta_db":-0.5282230666317044},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2031554280127926e-06,"output_power":5.890284994828481e-07,"reference_active":true,"delta_db":-3.101854265649272}],"active_band_count":3,"active_band_mae_db":1.417491384762333}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":1.379615167895732,"rms_envelope_nrmse":0.15532604003747438,"envelope_correlation":0.9807387155339228,"rms_gain_error_db":-0.15372093936914968,"reference_mean_power":3.324742134736325e-05,"output_mean_power":3.209119380751541e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5226373785787838e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9685546500066154e-06,"output_power":5.871910953471209e-06,"reference_active":true,"delta_db":-0.7436324933431494},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4829248214027976e-06,"output_power":1.1415438441636428e-06,"reference_active":true,"delta_db":-1.1362653808992371},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.259963132949634e-08,"output_power":3.440057792061127e-08,"reference_active":true,"delta_db":-4.300435189466554}],"active_band_count":3,"active_band_mae_db":2.0601110212363136}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"facodec","status":"measured","mos":{"status":"measured","value":4.25779390335083,"predicted_mos":4.25779390335083,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"e43da2b8385f5a437bedb76b171865ac2161879d97851da0e9d3cc2e6a7d3a69","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.559688979999919},"started_at":"2026-10-13T06:58:59.079503+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.361799933074508e-05,"native_input_samples":96000,"native_input_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","output_sha256":"3bb61af13699c766c0a82b11841f5f686260029ed4f6c016b4f4b54211229c47","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4693515640010446,"synthesis_s":6.263737323999521,"total_s":7.733088888000566,"rtf":1.2888481480000944,"output_peak":0.6335831880569458,"output_rms":0.08136589177494681,"finished_at":"2026-10-13T06:59:06.829226+00:00","evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"3bb61af13699c766c0a82b11841f5f686260029ed4f6c016b4f4b54211229c47","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.470423771594144,"spectral_convergence":0.19695565562581288,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.6239373506532715,"spectral_convergence":0.1913538495855574,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.679458166501383,"spectral_convergence":0.20515086270742552,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.0020461680698139,"delta_db":-0.5829476460884675,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.003386250778151619,"delta_db":-0.0412051394726172,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0009279572618739978,"delta_db":-0.17867024594242276,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.00014200215913349806,"delta_db":-0.6895856335191766,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":7.751686140425269e-05,"delta_db":0.053945472096893245,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":4.0513213955088384e-05,"delta_db":-0.026846252631811746,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.26220006495856485,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2479194310486123,"pitch":null,"output_rms_dbfs":-21.791152226115145,"output_peak":0.6335831880569458,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5912730962496004},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.470423771594144,"spectral_convergence":0.19695565562581288,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.6239373506532715,"spectral_convergence":0.1913538495855574,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.679458166501383,"spectral_convergence":0.20515086270742552,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.0020461680698139,"delta_db":-0.5829476460884675,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.003386250778151619,"delta_db":-0.0412051394726172,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0009279572618739978,"delta_db":-0.17867024594242276,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.00014200215913349806,"delta_db":-0.6895856335191766,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":7.751686140425269e-05,"delta_db":0.053945472096893245,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":4.0513213955088384e-05,"delta_db":-0.026846252631811746,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.26220006495856485,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2479194310486123,"pitch":null,"output_rms_dbfs":-21.791152226115145,"output_peak":0.6335831880569458,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.5912730962496004},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9090013633737191,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08372176207031103,"rms_dbfs":-21.54323279506653,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.494273826613313,"mean":-2.5322596232096355e-05,"zero_energy":false},"output":{"rms":0.08136589177494681,"rms_dbfs":-21.791152226115145,"peak":0.6335831880569458,"peak_dbfs":-3.9639271028336016,"crest_factor_db":17.827225123281544,"mean":1.97126826329234e-05,"zero_energy":false},"rms_difference_db":-0.24791943104861502},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":1.4516560709438947e-06,"output_rms_dbfs":-58.381362653052975,"output_zero_energy":false,"output_peak":0.009524867869913578,"output_energy_fraction":2.3388786035618895e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":147,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.91875,"absolute_difference_cents_median":3.408790842668746,"absolute_difference_cents_p95":11.867303236313496,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019906122792766575,"log_rms_mae_db":0.9922887882259122,"rms_envelope_nrmse":0.07097086470265129,"envelope_correlation":0.9954074855101261,"rms_gain_error_db":-0.24702474609856673,"reference_mean_power":0.007012794178240996,"output_mean_power":0.0066250417300206905},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.192818590388021e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015356197181510699,"output_power":0.0013943365035152427,"reference_active":true,"delta_db":-0.41916082767074697},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003622112164404606,"output_power":0.0003506161510530764,"reference_active":true,"delta_db":-0.14129976903630392},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6887620873910397e-05,"output_power":4.8263526966360585e-05,"reference_active":true,"delta_db":0.12560859034302896}],"active_band_count":3,"active_band_mae_db":0.22868972901669327}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018229956701751592,"log_rms_mae_db":0.7225877549521728,"rms_envelope_nrmse":0.0595052022923982,"envelope_correlation":0.9966345438221476,"rms_gain_error_db":-0.24549444587913938,"reference_mean_power":0.0070179714231847195,"output_mean_power":0.0066322692795145145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8443307374696498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014276197130393435,"output_power":0.001289048568042103,"reference_active":true,"delta_db":-0.4434325544183177},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018503527061359352,"output_power":0.00018236433061278524,"reference_active":true,"delta_db":-0.06314622747460906},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.327921114523829e-06,"output_power":4.57681198500469e-06,"reference_active":true,"delta_db":0.24283735779099302}],"active_band_count":3,"active_band_mae_db":0.2498053798946399}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015227518148493586,"log_rms_mae_db":1.5160436539020352,"rms_envelope_nrmse":0.0913148633051415,"envelope_correlation":0.993030475468832,"rms_gain_error_db":-0.040274783752165214,"reference_mean_power":0.0024461505543329222,"output_mean_power":0.0024235707639330435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3707367109300066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00065188242251555,"output_power":0.0006291309075413243,"reference_active":true,"delta_db":-0.15428249346193743},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000159373128948768,"output_power":0.00016901228755803556,"reference_active":true,"delta_db":0.25503180745940135},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8816561344170712e-05,"output_power":1.96770342837943e-05,"reference_active":true,"delta_db":0.1941938164584744}],"active_band_count":3,"active_band_mae_db":0.2011693724599377}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013579343349222632,"log_rms_mae_db":0.992454403733754,"rms_envelope_nrmse":0.07242275028574328,"envelope_correlation":0.9951708718760445,"rms_gain_error_db":-0.03751188321262603,"reference_mean_power":0.0024424650949128645,"output_mean_power":0.0024214593214192007},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.250982515304928e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006212562593255464,"output_power":0.0006014612673837968,"reference_active":true,"delta_db":-0.14063112311126336},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.001783099426803e-05,"output_power":8.632953495587279e-05,"reference_active":true,"delta_db":0.32972626452724907},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7444705804029888e-06,"output_power":1.8526303675748908e-06,"reference_active":true,"delta_db":0.2612512886154646}],"active_band_count":3,"active_band_mae_db":0.24386955875132568}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":1.7626517028946358,"rms_envelope_nrmse":0.16687329164628195,"envelope_correlation":0.9781875567790029,"rms_gain_error_db":-0.4088366404038519,"reference_mean_power":0.00019825791077017447,"output_mean_power":0.00018044583458824576},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.446126492709316e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.480334346914859e-05,"output_power":3.14380309219495e-05,"reference_active":true,"delta_db":-0.4416563074422494},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0911626587448623e-05,"output_power":1.0381283271320815e-05,"reference_active":true,"delta_db":-0.21638453662706586},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.788733059913985e-06,"output_power":2.9585336525753567e-06,"reference_active":true,"delta_db":0.2566956844206724}],"active_band_count":3,"active_band_mae_db":0.3049121761633293}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":1.348981305821046,"rms_envelope_nrmse":0.13467216692841677,"envelope_correlation":0.9846993540616856,"rms_gain_error_db":-0.40974277448574675,"reference_mean_power":0.00019869265760544442,"output_mean_power":0.00018080379472551583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.64113687643322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.340875190885912e-05,"output_power":3.0433201489004536e-05,"reference_active":true,"delta_db":-0.40512610131964466},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7578298997729145e-06,"output_power":5.501424278188913e-06,"reference_active":true,"delta_db":-0.19783690930881553},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2467549734271305e-07,"output_power":2.1392078196284024e-07,"reference_active":true,"delta_db":-0.2130273428561099}],"active_band_count":3,"active_band_mae_db":0.2719967844948567}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":1.9002170359621273,"rms_envelope_nrmse":0.1903307891022313,"envelope_correlation":0.9747011903621248,"rms_gain_error_db":0.1551774790467823,"reference_mean_power":3.3091629565415505e-05,"output_mean_power":3.429540232501124e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.646064632792655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.725047198052175e-06,"output_power":7.527831894793598e-06,"reference_active":true,"delta_db":0.489745755568988},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.390363033635178e-06,"output_power":2.6790306540903358e-06,"reference_active":true,"delta_db":0.49513819002873},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2031554280127926e-06,"output_power":1.0105265138768668e-06,"reference_active":true,"delta_db":-0.7577402182861459}],"active_band_count":3,"active_band_mae_db":0.5808747212946214}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":1.4253208592328557,"rms_envelope_nrmse":0.11289306135773379,"envelope_correlation":0.9903432118242033,"rms_gain_error_db":0.154596624366999,"reference_mean_power":3.324742134736325e-05,"output_mean_power":3.4452253153337785e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5226373785787838e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9685546500066154e-06,"output_power":7.725022513619905e-06,"reference_active":true,"delta_db":0.44757043505644073},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4829248214027976e-06,"output_power":1.606242494783743e-06,"reference_active":true,"delta_db":0.34691976903133626},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.259963132949634e-08,"output_power":9.027912634783346e-08,"reference_active":true,"delta_db":-0.11021909954867928}],"active_band_count":3,"active_band_mae_db":0.3015697678788188}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.587162971496582,"predicted_mos":3.587162971496582,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"f0b3278b356b4a085410b07b8a64a5f835d85f110994ec2167c5fc22ff7471df","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.132229656999698},"started_at":"2026-10-13T07:28:37.356849+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008068500999797834,"native_input_samples":264600,"native_input_sha256":"2af60273630e26aa21c782121c72b6f1a4a944a73b68528bd4e9c3dfcc90be62","output_sha256":"a01f7da078c7ce01a4fbf1cb38affcdcc371cff95fa09d3e61e9019346cfc257","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13699108400032856,"synthesis_s":3.9466744549972645,"total_s":4.083665538997593,"rtf":0.6806109231662655,"output_peak":0.43665385246276855,"output_rms":0.07505229249977707,"finished_at":"2026-10-13T07:28:41.469988+00:00","evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"a01f7da078c7ce01a4fbf1cb38affcdcc371cff95fa09d3e61e9019346cfc257","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.48938147146295,"output_peak":0.4360638551650057},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.73437786677654,"spectral_convergence":0.4894814898134621,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.724729533102963,"spectral_convergence":0.3345769569886962,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.5602585506177,"spectral_convergence":0.38666639930858904,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.0016081248521098497,"delta_db":-1.6291630426799244,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.003032492806869076,"delta_db":-0.5203985240322309,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0007887328840469432,"delta_db":-0.8846505143120488,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.0001232378839175957,"delta_db":-1.305092786414592,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":5.7896486568534205e-05,"delta_db":-1.2134942345999475,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":2.8754694749372635e-05,"delta_db":-1.515725612159195,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.178087452366323,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9445617257373597,"pitch":null,"output_rms_dbfs":-22.487794520803888,"output_peak":0.4360638551650057,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.673121983499067},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7958197497397748,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08372176207031103,"rms_dbfs":-21.54323279506653,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.494273826613313,"mean":-2.5322596232096355e-05,"zero_energy":false},"output":{"rms":0.07509487071872067,"rms_dbfs":-22.487794520803888,"peak":0.4360638551650057,"peak_dbfs":-7.208998200167949,"crest_factor_db":15.27879632063594,"mean":-0.00012039142811215625,"zero_energy":false},"rms_difference_db":-0.9445617257373584},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":1.570205189818466e-06,"output_rms_dbfs":-58.04043591550545,"output_zero_energy":false,"output_peak":0.04098050630649938,"output_energy_fraction":2.9700556318846428e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":147,"both_estimated_voiced_frames":122,"reference_voiced_output_unvoiced_frames":38,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.7625,"absolute_difference_cents_median":11.08210758958821,"absolute_difference_cents_p95":29.198400735210733,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019906122792766575,"log_rms_mae_db":1.8015076601421962,"rms_envelope_nrmse":0.14813872338536616,"envelope_correlation":0.986891455140146,"rms_gain_error_db":-0.9434108635868016,"reference_mean_power":0.007012794178240996,"output_mean_power":0.005643519193577893},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.192818590388021e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015356197181510699,"output_power":0.0012770797973677649,"reference_active":true,"delta_db":-0.8006564537249777},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003622112164404606,"output_power":0.000287029515777563,"reference_active":true,"delta_db":-1.0103533646094804},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6887620873910397e-05,"output_power":4.608515167081404e-05,"reference_active":true,"delta_db":-0.07497175614813283}],"active_band_count":3,"active_band_mae_db":0.6286605248275304}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018229956701751592,"log_rms_mae_db":1.4148310452657926,"rms_envelope_nrmse":0.12486561665072118,"envelope_correlation":0.9933902959564234,"rms_gain_error_db":-0.9361095275522892,"reference_mean_power":0.0070179714231847195,"output_mean_power":0.005657188406722461},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8443307374696498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014276197130393435,"output_power":0.0011879070916085298,"reference_active":true,"delta_db":-0.798300611993227},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018503527061359352,"output_power":0.00014518431392571927,"reference_active":true,"delta_db":-1.053348229646533},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.327921114523829e-06,"output_power":4.0313939560120894e-06,"reference_active":true,"delta_db":-0.308240958727454}],"active_band_count":3,"active_band_mae_db":0.7199632667890713}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015227518148493586,"log_rms_mae_db":2.8194225503253105,"rms_envelope_nrmse":0.180754127191461,"envelope_correlation":0.9733769701058739,"rms_gain_error_db":-0.39739571157066894,"reference_mean_power":0.0024461505543329222,"output_mean_power":0.002232254011462452},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3707367109300066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00065188242251555,"output_power":0.0006029572072128015,"reference_active":true,"delta_db":-0.33882780101168963},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000159373128948768,"output_power":0.0001377046852773806,"reference_active":true,"delta_db":-0.6346638217629781},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8816561344170712e-05,"output_power":2.0991795820077985e-05,"reference_active":true,"delta_db":0.4750933282544967}],"active_band_count":3,"active_band_mae_db":0.4828616503430548}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013579343349222632,"log_rms_mae_db":1.872404042359219,"rms_envelope_nrmse":0.15317954262381656,"envelope_correlation":0.9793982021279632,"rms_gain_error_db":-0.38301086780593074,"reference_mean_power":0.0024424650949128645,"output_mean_power":0.0022362856613247486},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.250982515304928e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006212562593255464,"output_power":0.0005735325966025047,"reference_active":true,"delta_db":-0.347126714255256},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.001783099426803e-05,"output_power":6.881967932879201e-05,"reference_active":true,"delta_db":-0.6547413032897069},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7444705804029888e-06,"output_power":2.0152523161401435e-06,"reference_active":true,"delta_db":0.6266577926875728}],"active_band_count":3,"active_band_mae_db":0.5428419367441785}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":3.7552606427699677,"rms_envelope_nrmse":0.273679390786934,"envelope_correlation":0.9510092472199903,"rms_gain_error_db":-1.4111228835475056,"reference_mean_power":0.00019825791077017447,"output_mean_power":0.00014325778664017955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.446126492709316e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.480334346914859e-05,"output_power":2.6446968620340193e-05,"reference_active":true,"delta_db":-1.1924506759415072},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0911626587448623e-05,"output_power":6.78196098941001e-06,"reference_active":true,"delta_db":-2.0653420800592843},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.788733059913985e-06,"output_power":1.8068825100836827e-06,"reference_active":true,"delta_db":-1.884770310796521}],"active_band_count":3,"active_band_mae_db":1.7141876889324374}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":3.169901519560883,"rms_envelope_nrmse":0.2273258119117036,"envelope_correlation":0.9697380178533306,"rms_gain_error_db":-1.4084428919027212,"reference_mean_power":0.00019869265760544442,"output_mean_power":0.00014366055156117116},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.64113687643322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.340875190885912e-05,"output_power":2.5112816782578153e-05,"reference_active":true,"delta_db":-1.2396482333122976},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7578298997729145e-06,"output_power":3.5006465523756408e-06,"reference_active":true,"delta_db":-2.1611056678668445},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2467549734271305e-07,"output_power":1.8619693627083292e-07,"reference_active":true,"delta_db":-0.8158318100173511}],"active_band_count":3,"active_band_mae_db":1.405528570398831}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":3.3226600662380785,"rms_envelope_nrmse":0.2644387264768702,"envelope_correlation":0.9487305102765553,"rms_gain_error_db":-0.17308769765562165,"reference_mean_power":3.3091629565415505e-05,"output_mean_power":3.179870138308395e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.646064632792655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.725047198052175e-06,"output_power":8.748144262434933e-06,"reference_active":true,"delta_db":1.1422059954376795},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.390363033635178e-06,"output_power":2.532606138053567e-06,"reference_active":true,"delta_db":0.2510379116846984},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2031554280127926e-06,"output_power":7.720551873180727e-07,"reference_active":true,"delta_db":-1.926733894390968}],"active_band_count":3,"active_band_mae_db":1.1066592671711153}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":2.8215724893861003,"rms_envelope_nrmse":0.20699787493133717,"envelope_correlation":0.9655300586687894,"rms_gain_error_db":-0.17299504170459462,"reference_mean_power":3.324742134736325e-05,"output_mean_power":3.194908782100173e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5226373785787838e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9685546500066154e-06,"output_power":8.733598390499919e-06,"reference_active":true,"delta_db":0.9805050697198281},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4829248214027976e-06,"output_power":1.3898858165665569e-06,"reference_active":true,"delta_db":-0.2814001144296068},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.259963132949634e-08,"output_power":6.306730889515969e-08,"reference_active":true,"delta_db":-1.668049577255701}],"active_band_count":3,"active_band_mae_db":0.9766515871350453}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"golf","status":"measured","mos":{"status":"measured","value":3.100785255432129,"predicted_mos":3.100785255432129,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"4bc3cdcaf14d1434d71c1a80b7c6ee874bde6a1a3731a843a020390d7e8d0614","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.451826346999951},"started_at":"2026-10-13T05:02:58.487401+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003053100000215636,"native_input_samples":144000,"native_input_sha256":"b605bd4e22fe6db639d190db942b95229f42fde0a57b6ab68aa7fa4937633726","output_sha256":"f2e6cb1f897a313ea9a5ff417788f61f50a163e57d99a681b766fe4af6d4c39e","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5882391179993647,"synthesis_s":0.13219571900026494,"total_s":0.7204348369996296,"rtf":0.1200724728332716,"output_peak":0.5095789432525635,"output_rms":0.0760789019403414,"finished_at":"2026-10-13T05:02:59.218671+00:00","analysis_stages":{"external_f0_s":0.08076629700008198},"evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"f2e6cb1f897a313ea9a5ff417788f61f50a163e57d99a681b766fe4af6d4c39e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.37192722450038,"output_peak":0.5100190749505955},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.171492588057008,"spectral_convergence":0.7382229039497834,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.378677362598892,"spectral_convergence":0.2900636713639606,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.558659877755137,"spectral_convergence":0.259665834689434,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002337132363306887,"output_power":0.0019860973289741717,"delta_db":-0.7068278198046919,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0034194385843663595,"output_power":0.0029031374159645138,"delta_db":-0.7108721484818649,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009700388492272881,"output_power":0.0006965218411257154,"delta_db":-1.4385438841095421,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016690750715079368,"output_power":0.00012411107248470231,"delta_db":-1.2866534226335047,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.681018385472877e-05,"output_power":5.17014225900312e-05,"delta_db":-1.7191631156575704,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.0915198940206604e-05,"output_power":3.0147181699567116e-05,"delta_db":-1.3263794868695953,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1980733129261283,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8298772247221511,"pitch":null,"output_rms_dbfs":-22.37192722450038,"output_peak":0.5100190749505955,"reference_rms_dbfs":-21.54204999977823,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.036276609470345},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8874152283756541,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08373316360228045,"rms_dbfs":-21.54204999977823,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.493091031325008,"mean":-5.73860384984088e-06,"zero_energy":false},"output":{"rms":0.07610332622716764,"rms_dbfs":-22.37192722450038,"peak":0.5100190749505955,"peak_dbfs":-5.8482716156542605,"crest_factor_db":16.52365560884612,"mean":0.0012073878500537482,"zero_energy":false},"rms_difference_db":-0.82987722472215},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":1.525190627335989e-06,"output_rms_dbfs":-58.16675872234433,"output_zero_energy":false,"output_peak":0.01675644350030295,"output_energy_fraction":2.813649538703201e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":217,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":58,"paired_voiced_fraction":0.99375,"absolute_difference_cents_median":6.797218444305585,"absolute_difference_cents_p95":24.72530749456814,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019906122792766575,"log_rms_mae_db":1.7789972454841476,"rms_envelope_nrmse":0.15445832469988138,"envelope_correlation":0.9819723238543759,"rms_gain_error_db":-0.8282993503795224,"reference_mean_power":0.007010483632493242,"output_mean_power":0.005793193632387584},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.1910467820492943e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015327113507304895,"output_power":0.0012846859122940199,"reference_active":true,"delta_db":-0.7666341191980852},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003610053853452887,"output_power":0.00028419000750231705,"reference_active":true,"delta_db":-1.0390487710923193},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.634541734908187e-05,"output_power":2.9071447063116487e-05,"reference_active":true,"delta_db":-2.025401476227796}],"active_band_count":3,"active_band_mae_db":1.2770281221727335}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00018229956701751592,"log_rms_mae_db":1.390298934325574,"rms_envelope_nrmse":0.11803921712840733,"envelope_correlation":0.9927955773628581,"rms_gain_error_db":-0.8331940086048968,"reference_mean_power":0.006998793337320929,"output_mean_power":0.005777018640304487},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.8433113043467917e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014245174438005939,"output_power":0.0011923146125443643,"reference_active":true,"delta_db":-0.7727690519070198},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001840617931001038,"output_power":0.00014659743883867134,"reference_active":true,"delta_db":-0.9883726559900272},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.297904318098217e-06,"output_power":2.5207538154506833e-06,"reference_active":true,"delta_db":-2.3172630958535407}],"active_band_count":3,"active_band_mae_db":1.3594682679168626}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00015227518300081126,"log_rms_mae_db":2.262604593518879,"rms_envelope_nrmse":0.17548835101219468,"envelope_correlation":0.9832953535356685,"rms_gain_error_db":-1.0549383542305208,"reference_mean_power":0.0024428256038924547,"output_mean_power":0.0019160137785537146},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3687862358249396e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006502760822609981,"output_power":0.0004847225754801327,"reference_active":true,"delta_db":-1.276045335253502},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00015897144897183705,"output_power":0.00012379955562880673,"reference_active":true,"delta_db":-1.0860004688915776},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8476953899642917e-05,"output_power":1.5174201856745773e-05,"reference_active":true,"delta_db":-0.8552451821421091}],"active_band_count":3,"active_band_mae_db":1.0724303287623962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013579253971476415,"log_rms_mae_db":1.662672585333177,"rms_envelope_nrmse":0.14905253751893952,"envelope_correlation":0.9908653861907861,"rms_gain_error_db":-1.0701092199764375,"reference_mean_power":0.002427988920869854,"output_mean_power":0.0018977359233480144},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2493030716017832e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006195192673440398,"output_power":0.000461296084832256,"reference_active":true,"delta_db":-1.2807504905976095},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.96271943966253e-05,"output_power":6.46554623364217e-05,"reference_active":true,"delta_db":-0.9045619236525299},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7339891146116926e-06,"output_power":1.1967088817515055e-06,"reference_active":true,"delta_db":-1.6105785250249418}],"active_band_count":3,"active_band_mae_db":1.2652969797583602}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":2.2565869163967736,"rms_envelope_nrmse":0.24279069801770733,"envelope_correlation":0.9689928498928216,"rms_gain_error_db":-1.5569760636535301,"reference_mean_power":0.00019837628346948354,"output_mean_power":0.0001386092275402859},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.423177483321167e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.471794920861932e-05,"output_power":2.522213739147161e-05,"reference_active":true,"delta_db":-1.3877217638646906},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0843806534070874e-05,"output_power":8.085140747429829e-06,"reference_active":true,"delta_db":-1.2749417609018563},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.782957183237755e-06,"output_power":1.8649499122248433e-06,"reference_active":true,"delta_db":-1.7383935234953207}],"active_band_count":3,"active_band_mae_db":1.4670190160872891}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":1.7686236983932337,"rms_envelope_nrmse":0.196639504589488,"envelope_correlation":0.9871411100594479,"rms_gain_error_db":-1.5527098418302083,"reference_mean_power":0.0001980679017138676,"output_mean_power":0.0001385297710972979},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.6153747051322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.329138251567497e-05,"output_power":2.420665605567092e-05,"reference_active":true,"delta_db":-1.3839703131272536},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.740535417542642e-06,"output_power":4.34049904144473e-06,"reference_active":true,"delta_db":-1.2141273611976156},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2374969282136017e-07,"output_power":1.5035715283274776e-07,"reference_active":true,"delta_db":-1.7263835445248146}],"active_band_count":3,"active_band_mae_db":1.4414937396165612}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":2.607870675665674,"rms_envelope_nrmse":0.2136978870624568,"envelope_correlation":0.9704502846149309,"rms_gain_error_db":-0.5677446873269714,"reference_mean_power":3.314456970877984e-05,"output_mean_power":2.908291383087593e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.6391928608277906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.686424807769384e-06,"output_power":6.966339513117596e-06,"reference_active":true,"delta_db":0.17810671404636835},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.368787901056478e-06,"output_power":2.282578013290526e-06,"reference_active":true,"delta_db":-0.16100546510895433},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1944979633461615e-06,"output_power":9.675100641615499e-07,"reference_active":true,"delta_db":-0.9152992230143213}],"active_band_count":3,"active_band_mae_db":0.418137134056548}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":2.265290080798057,"rms_envelope_nrmse":0.15162582946232117,"envelope_correlation":0.9864782148422536,"rms_gain_error_db":-0.5667635188137101,"reference_mean_power":3.328630494256307e-05,"output_mean_power":2.9213879638934276e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.516694082048402e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9253567194374316e-06,"output_power":7.222150125525796e-06,"reference_active":true,"delta_db":0.18224363320570244},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4690456895958632e-06,"output_power":1.3879476058113945e-06,"reference_active":true,"delta_db":-0.24662231158845438},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.182022733897719e-08,"output_power":8.609070175895562e-08,"reference_active":true,"delta_db":-0.2798211577074009}],"active_band_count":3,"active_band_mae_db":0.2362290341671859}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.0363566875457764,"predicted_mos":3.0363566875457764,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"b8721d20e7363ab025a6659c501564c8e422ba449c336e5eeec0fc465a9beeee","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.185219031998713},"started_at":"2026-10-13T07:44:46.005244+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003809933001321042,"native_input_samples":144000,"native_input_sha256":"b605bd4e22fe6db639d190db942b95229f42fde0a57b6ab68aa7fa4937633726","output_sha256":"0289f12238b72851cd1a3ec4ca4690f7b49ca8eb15fbef92e0998c4eddad6710","output_samples":144120,"length_delta_samples":120,"analysis_s":2.2352690680017986,"synthesis_s":13.015490919999138,"total_s":15.250759988000937,"rtf":2.5417933313334893,"output_peak":0.5941647291183472,"output_rms":0.0863645455848048,"finished_at":"2026-10-13T07:45:01.270273+00:00","analysis_stages":{"external_f0_s":1.4224726779975754},"evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"0289f12238b72851cd1a3ec4ca4690f7b49ca8eb15fbef92e0998c4eddad6710","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.271066769096976,"output_peak":0.5964258750535685},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.886257363795929,"spectral_convergence":0.685600451036903,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.400325555173566,"spectral_convergence":0.30948658195050915,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.285545967022942,"spectral_convergence":0.2886643716427762,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.003261309559282136,"delta_db":1.4415595554977314,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.0031293102283467987,"delta_db":-0.38391012144577136,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0008358130667733663,"delta_db":-0.632858428977872,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.0001362216498034095,"delta_db":-0.8700737541196801,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":7.393684589124872e-05,"delta_db":-0.15140713105082937,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":2.869944048168944e-05,"delta_db":-1.5240789294123314,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8339813200840359,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.27370008685452746,"pitch":null,"output_rms_dbfs":-21.269532708212004,"output_peak":0.5964258750535685,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.857376295330813},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8993557927112895,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08372176207031103,"rms_dbfs":-21.54323279506653,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.494273826613313,"mean":-2.5322596232096355e-05,"zero_energy":false},"output":{"rms":0.08640191427612383,"rms_dbfs":-21.269532708212004,"peak":0.5964258750535685,"peak_dbfs":-4.488870471744329,"crest_factor_db":16.780662236467673,"mean":-0.0002126158343389715,"zero_energy":false},"rms_difference_db":0.2737000868545252},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":3.773661963051037e-05,"output_rms_dbfs":-44.23237005643472,"output_zero_energy":false,"output_peak":0.20914758350547488,"output_energy_fraction":0.000539193922939261,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":207,"both_estimated_voiced_frames":154,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":53,"paired_voiced_fraction":0.9625,"absolute_difference_cents_median":7.073284225315204,"absolute_difference_cents_p95":24.955290755834326,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019906122792766575,"log_rms_mae_db":2.3879540299677195,"rms_envelope_nrmse":0.16510770696088273,"envelope_correlation":0.9731331540339297,"rms_gain_error_db":0.2743879959589539,"reference_mean_power":0.007012794178240996,"output_mean_power":0.007470159802150292},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.192818590388021e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015356197181510699,"output_power":0.0016237879080249807,"reference_active":true,"delta_db":0.24245622943494569},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003622112164404606,"output_power":0.00033696886948495294,"reference_active":true,"delta_db":-0.3137211391349758},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6887620873910397e-05,"output_power":3.1226973943285154e-05,"reference_active":true,"delta_db":-1.7652829588158143}],"active_band_count":3,"active_band_mae_db":0.7738201091285787}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018229956701751592,"log_rms_mae_db":1.7819484194175474,"rms_envelope_nrmse":0.1096531763183263,"envelope_correlation":0.9874017694123779,"rms_gain_error_db":0.27617870680837603,"reference_mean_power":0.0070179714231847195,"output_mean_power":0.007478757754189357},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8443307374696498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014276197130393435,"output_power":0.0014931255923799752,"reference_active":true,"delta_db":0.19483803190103535},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018503527061359352,"output_power":0.00017171059101912998,"reference_active":true,"delta_db":-0.32457436564879893},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.327921114523829e-06,"output_power":2.830891479715749e-06,"reference_active":true,"delta_db":-1.8435611493622261}],"active_band_count":3,"active_band_mae_db":0.7876578489706868}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015227518148493586,"log_rms_mae_db":2.8501705822830377,"rms_envelope_nrmse":0.17935755329406397,"envelope_correlation":0.9744957079197365,"rms_gain_error_db":-0.5034304304901009,"reference_mean_power":0.0024461505543329222,"output_mean_power":0.002178412599406911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3707367109300066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00065188242251555,"output_power":0.0005795171830244058,"reference_active":true,"delta_db":-0.5110295322051345},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000159373128948768,"output_power":0.0001314645841161755,"reference_active":true,"delta_db":-0.8360632728259212},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8816561344170712e-05,"output_power":1.1383237171588374e-05,"reference_active":true,"delta_db":-2.1827447611855693}],"active_band_count":3,"active_band_mae_db":1.1766125220722083}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013579343349222632,"log_rms_mae_db":2.2084474824072875,"rms_envelope_nrmse":0.12701816744284888,"envelope_correlation":0.9874313746207878,"rms_gain_error_db":-0.4940463356515674,"reference_mean_power":0.0024424650949128645,"output_mean_power":0.002179835556895947},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.250982515304928e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006212562593255464,"output_power":0.0005514261337888277,"reference_active":true,"delta_db":-0.517834325829712},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.001783099426803e-05,"output_power":6.514460781677702e-05,"reference_active":true,"delta_db":-0.8930830109770846},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7444705804029888e-06,"output_power":1.0427088608954967e-06,"reference_active":true,"delta_db":-2.2350058552073797}],"active_band_count":3,"active_band_mae_db":1.2153077306713922}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":2.6676846317385454,"rms_envelope_nrmse":0.22600487636221855,"envelope_correlation":0.9590516338402874,"rms_gain_error_db":-0.5569006056014624,"reference_mean_power":0.00019825791077017447,"output_mean_power":0.00017439758422311608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.446126492709316e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.480334346914859e-05,"output_power":3.0302727351848576e-05,"reference_active":true,"delta_db":-0.6013924923325413},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0911626587448623e-05,"output_power":8.67595056934001e-06,"reference_active":true,"delta_db":-0.9957242636042368},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.788733059913985e-06,"output_power":1.6931171776112538e-06,"reference_active":true,"delta_db":-2.1671992929613153}],"active_band_count":3,"active_band_mae_db":1.2547720162993645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":1.841983249141659,"rms_envelope_nrmse":0.14539825377521348,"envelope_correlation":0.9831452079824192,"rms_gain_error_db":-0.5516311363954003,"reference_mean_power":0.00019869265760544442,"output_mean_power":0.0001749922056106079},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.64113687643322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.340875190885912e-05,"output_power":2.8565693836737977e-05,"reference_active":true,"delta_db":-0.6801547430380848},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7578298997729145e-06,"output_power":4.350401533647746e-06,"reference_active":true,"delta_db":-1.217294872763914},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2467549734271305e-07,"output_power":1.26961377814894e-07,"reference_active":true,"delta_db":-2.4788408488122364}],"active_band_count":3,"active_band_mae_db":1.458763488204745}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":2.669164526821297,"rms_envelope_nrmse":0.2829098049444254,"envelope_correlation":0.9431487761589847,"rms_gain_error_db":-0.8740677155549375,"reference_mean_power":3.3091629565415505e-05,"output_mean_power":2.7058977563427492e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.646064632792655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.725047198052175e-06,"output_power":5.258448693630309e-06,"reference_active":true,"delta_db":-1.068376957812318},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.390363033635178e-06,"output_power":2.088877807537355e-06,"reference_active":true,"delta_db":-0.5855082800355844},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2031554280127926e-06,"output_power":6.637585287316401e-07,"reference_active":true,"delta_db":-2.583116202719658}],"active_band_count":3,"active_band_mae_db":1.41233381352252}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":1.8235044934955937,"rms_envelope_nrmse":0.1752291672852542,"envelope_correlation":0.9810586138862919,"rms_gain_error_db":-0.8742105288367658,"reference_mean_power":3.324742134736325e-05,"output_mean_power":2.718547429675464e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5226373785787838e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9685546500066154e-06,"output_power":5.387660310054857e-06,"reference_active":true,"delta_db":-1.1174250447919343},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4829248214027976e-06,"output_power":1.20096843470679e-06,"reference_active":true,"delta_db":-0.9158754161866685},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.259963132949634e-08,"output_power":4.982894872221245e-08,"reference_active":true,"delta_db":-2.6912753297262832}],"active_band_count":3,"active_band_mae_db":1.5748585969016287}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.584935188293457,"predicted_mos":3.584935188293457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"9ac0c8d8a4f4f5755ea0be74252b0c4a6358978635cc3a96ed26be9b217bb579","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3405975599998783},"started_at":"2026-10-13T05:11:20.821597+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030626750012743287,"native_input_samples":144000,"native_input_sha256":"b605bd4e22fe6db639d190db942b95229f42fde0a57b6ab68aa7fa4937633726","output_sha256":"4d321b382dfb55cb774a6f845ac451bdac555282f91be5b4acf55d18bd024556","output_samples":144240,"length_delta_samples":240,"analysis_s":1.2101878829998896,"synthesis_s":0.1459040770005231,"total_s":1.3560919600004127,"rtf":0.22601532666673543,"output_peak":0.6611707210540771,"output_rms":0.08240194467938715,"finished_at":"2026-10-13T05:11:22.190058+00:00","analysis_stages":{"external_f0_s":0.18577623199962545},"evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"4d321b382dfb55cb774a6f845ac451bdac555282f91be5b4acf55d18bd024556","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.67847999341295,"output_peak":0.6662138500562023},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.012553647838827,"spectral_convergence":0.14959648183577937,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.427049424264546,"spectral_convergence":0.14793270705457007,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.640667905411199,"spectral_convergence":0.15461705022427447,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.0023325792418716113,"delta_db":-0.013996614565482363,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.0032672073089741677,"delta_db":-0.19662939832819748,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0009585622406430734,"delta_db":-0.03774681935338839,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.0001489759106763065,"delta_db":-0.4813746229629541,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":6.467505828819685e-05,"delta_db":-0.7326480415730083,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":3.3243100475493275e-05,"delta_db":-0.8857980061294497,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3913655838187467,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12833091868141988,"pitch":null,"output_rms_dbfs":-21.67156371374795,"output_peak":0.6662138500562023,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.693423659171525},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9457843103443614,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08372176207031103,"rms_dbfs":-21.54323279506653,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.494273826613313,"mean":-2.5322596232096355e-05,"zero_energy":false},"output":{"rms":0.08249389590102318,"rms_dbfs":-21.67156371374795,"peak":0.6662138500562023,"peak_dbfs":-3.527726858280848,"crest_factor_db":18.1438368554671,"mean":0.0002258121931701288,"zero_energy":false},"rms_difference_db":-0.12833091868142077},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":9.7167009435354e-06,"output_rms_dbfs":-50.12481163594232,"output_zero_energy":false,"output_peak":0.07195707870336387,"output_energy_fraction":0.0001523014125762311,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":168,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.95625,"absolute_difference_cents_median":1.1093983333713853,"absolute_difference_cents_p95":7.405472317999918,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019906122792766575,"log_rms_mae_db":0.9938549278608225,"rms_envelope_nrmse":0.06761402994376065,"envelope_correlation":0.9953484543917058,"rms_gain_error_db":-0.12826886363757573,"reference_mean_power":0.007012794178240996,"output_mean_power":0.006808700115697146},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.192818590388021e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015356197181510699,"output_power":0.001491883262046556,"reference_active":true,"delta_db":-0.1254883856736917},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003622112164404606,"output_power":0.0003065813160683117,"reference_active":true,"delta_db":-0.7241621060658252},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6887620873910397e-05,"output_power":4.038542411165231e-05,"reference_active":true,"delta_db":-0.6483354872027391}],"active_band_count":3,"active_band_mae_db":0.4993286596474187}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018229956701751592,"log_rms_mae_db":0.6321384166099966,"rms_envelope_nrmse":0.03864187375929709,"envelope_correlation":0.9984600566529804,"rms_gain_error_db":-0.12747399839983187,"reference_mean_power":0.0070179714231847195,"output_mean_power":0.0068149738797003935},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8443307374696498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014276197130393435,"output_power":0.0013787538378189988,"reference_active":true,"delta_db":-0.15125801878307188},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018503527061359352,"output_power":0.00015810676014607681,"reference_active":true,"delta_db":-0.6830408021703138},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.327921114523829e-06,"output_power":3.361808089236065e-06,"reference_active":true,"delta_db":-1.0970641838041675}],"active_band_count":3,"active_band_mae_db":0.6437876682525178}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015227518148493586,"log_rms_mae_db":1.3160118299959482,"rms_envelope_nrmse":0.09003139838616692,"envelope_correlation":0.9933657901965929,"rms_gain_error_db":-0.14971445693431293,"reference_mean_power":0.0024461505543329222,"output_mean_power":0.002363261269451805},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3707367109300066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00065188242251555,"output_power":0.0006174594369493862,"reference_active":true,"delta_db":-0.23560838238224233},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000159373128948768,"output_power":0.00012889497267056963,"reference_active":true,"delta_db":-0.9217912039630515},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8816561344170712e-05,"output_power":1.4143713587676218e-05,"reference_active":true,"delta_db":-1.2397680742918928}],"active_band_count":3,"active_band_mae_db":0.7990558868790621}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013579343349222632,"log_rms_mae_db":0.8065377380037607,"rms_envelope_nrmse":0.05559962284677969,"envelope_correlation":0.9973674767013603,"rms_gain_error_db":-0.146440484703193,"reference_mean_power":0.0024424650949128645,"output_mean_power":0.0023614802486108155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.250982515304928e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006212562593255464,"output_power":0.0005826733058318538,"reference_active":true,"delta_db":-0.2784565530175808},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.001783099426803e-05,"output_power":6.495120678917683e-05,"reference_active":true,"delta_db":-0.9059955033905138},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7444705804029888e-06,"output_power":1.1322664244786233e-06,"reference_active":true,"delta_db":-1.8771502048522481}],"active_band_count":3,"active_band_mae_db":1.020534087086781}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":1.4854736773349144,"rms_envelope_nrmse":0.14739200934803412,"envelope_correlation":0.9841746407965624,"rms_gain_error_db":-0.5183216850629745,"reference_mean_power":0.00019825791077017447,"output_mean_power":0.00017595368098521414},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.446126492709316e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.480334346914859e-05,"output_power":3.10123433006945e-05,"reference_active":true,"delta_db":-0.5008638464205936},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0911626587448623e-05,"output_power":8.565593099230507e-06,"reference_active":true,"delta_db":-1.051320555621109},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.788733059913985e-06,"output_power":1.7918246078620046e-06,"reference_active":true,"delta_db":-1.9211144854949582}],"active_band_count":3,"active_band_mae_db":1.1577662958455537}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":1.0008365297172053,"rms_envelope_nrmse":0.09845559477110785,"envelope_correlation":0.9939388631270556,"rms_gain_error_db":-0.5203812726932446,"reference_mean_power":0.00019869265760544442,"output_mean_power":0.000176255911330808},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.64113687643322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.340875190885912e-05,"output_power":2.9326733012394167e-05,"reference_active":true,"delta_db":-0.5659656605248197},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7578298997729145e-06,"output_power":4.557736299912266e-06,"reference_active":true,"delta_db":-1.0150963632542556},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2467549734271305e-07,"output_power":1.2817537058995873e-07,"reference_active":true,"delta_db":-2.4375112994194077}],"active_band_count":3,"active_band_mae_db":1.339524441066161}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":1.670797011798134,"rms_envelope_nrmse":0.14598884226419687,"envelope_correlation":0.9868305212626346,"rms_gain_error_db":-0.5841579873583621,"reference_mean_power":3.3091629565415505e-05,"output_mean_power":2.8926930713841107e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.646064632792655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.725047198052175e-06,"output_power":6.288542405120044e-06,"reference_active":true,"delta_db":-0.29145342859315376},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.390363033635178e-06,"output_power":2.052220797090527e-06,"reference_active":true,"delta_db":-0.6623977951978368},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2031554280127926e-06,"output_power":6.25036089045577e-07,"reference_active":true,"delta_db":-2.844166408540066}],"active_band_count":3,"active_band_mae_db":1.2660058774436855}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":1.1077652013492987,"rms_envelope_nrmse":0.09287960407477282,"envelope_correlation":0.9962764739540334,"rms_gain_error_db":-0.5847802053203269,"reference_mean_power":3.324742134736325e-05,"output_mean_power":2.9058951945144113e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5226373785787838e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9685546500066154e-06,"output_power":6.346798426111139e-06,"reference_active":true,"delta_db":-0.405880049208629},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4829248214027976e-06,"output_power":1.1768477618117814e-06,"reference_active":true,"delta_db":-1.0039884880895933},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.259963132949634e-08,"output_power":3.988727733839315e-08,"reference_active":true,"delta_db":-3.657748647532364}],"active_band_count":3,"active_band_mae_db":1.6892057282768622}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0002","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1963255405426025,"predicted_mos":3.1963255405426025,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289466,"input_sample_rate":48000,"input_seconds":6.030541666666666,"input_float32_sha256":"6979e9010ad123a7fdc1f4220032c8bf0aabcd2953712b79b8b86fa71e6653d0","internal_samples":96489,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.524375222999879},"started_at":"2026-10-13T05:17:58.056646+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009463352998864138,"native_input_samples":288000,"native_input_sha256":"58c13ebd3f5285254aa64d7a3423f110d3862b0232c504a3abf8e9d2648d1c2e","output_sha256":"057f238af371cba35ebabf63c3cadfc2e2bc082c6c90ce18157dbbbab6f5939f","output_samples":289466,"length_delta_samples":1466,"analysis_s":2.9562624390000565,"synthesis_s":0.9089780949998385,"total_s":3.865240533999895,"rtf":0.6442067556666492,"output_peak":0.6476626004477708,"output_rms":0.08142226714193261,"finished_at":"2026-10-13T05:18:01.944466+00:00","evaluation_input":{"reference":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"057f238af371cba35ebabf63c3cadfc2e2bc082c6c90ce18157dbbbab6f5939f","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96489,"length_delta_samples":489,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.782574575945645,"output_peak":0.6481564921090432},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9949320647949507,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":11.54584116308106,"spectral_convergence":0.9083471944444306,"spectral_convergence_status":"ok","amplitude_floor":1.9341972091070217e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":10.719468045282186,"spectral_convergence":0.6613863820637563,"spectral_convergence_status":"ok","amplitude_floor":1.0469896576426271e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.11970702670884,"spectral_convergence":0.6486331252480023,"spectral_convergence_status":"ok","amplitude_floor":8.978224937278799e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002340108897553237,"output_power":0.002151729528718737,"delta_db":-0.36448388047959296,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003418531859998669,"output_power":0.003151769496253314,"delta_db":-0.3528518381012086,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009669299209010629,"output_power":0.0009314732095000546,"delta_db":-0.16224630865628256,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00016643839195516676,"output_power":0.00016359598437181436,"delta_db":-0.07480871933514598,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.655994855693747e-05,"output_power":7.759754702237795e-05,"delta_db":0.05846359807069799,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.076442519269658e-05,"output_power":3.9613563720233337e-05,"delta_db":-0.12437408628280172,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18953807182095497,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.31710352893356336,"pitch":null,"output_rms_dbfs":-21.860336324000095,"output_peak":0.6481564921090432,"reference_rms_dbfs":-21.54323279506653,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":10.461672078357362},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.5407193464808104,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08372176207031103,"rms_dbfs":-21.54323279506653,"peak":0.627410888671875,"peak_dbfs":-4.048958968453217,"crest_factor_db":17.494273826613313,"mean":-2.5322596232096355e-05,"zero_energy":false},"output":{"rms":0.08072037741231475,"rms_dbfs":-21.860336324000095,"peak":0.6481564921090432,"peak_dbfs":-3.7664024920288024,"crest_factor_db":18.09393383197129,"mean":0.0003815236503236515,"zero_energy":false},"rms_difference_db":-0.31710352893356486},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.9514004508964714e-06,"threshold_dbfs":-54.03248954929599,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":64,"quiet_samples":10240,"reference_mean_power":1.2614219485840295e-06,"output_mean_power":4.5590516371497504e-05,"output_rms_dbfs":-43.41125488824097,"output_zero_energy":false,"output_peak":0.09860297637378375,"output_energy_fraction":0.0007463402560117581,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":160,"output_estimated_voiced_frames":169,"both_estimated_voiced_frames":127,"reference_voiced_output_unvoiced_frames":33,"reference_unvoiced_output_voiced_frames":42,"paired_voiced_fraction":0.79375,"absolute_difference_cents_median":16.389096412339388,"absolute_difference_cents_p95":52.828571547740175,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019906122792766575,"log_rms_mae_db":6.566693456740331,"rms_envelope_nrmse":0.5576302494403467,"envelope_correlation":0.6667528324606423,"rms_gain_error_db":-0.3202937037693079,"reference_mean_power":0.007012794178240996,"output_mean_power":0.006514209512488564},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.192818590388021e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015356197181510699,"output_power":0.0014665109969880798,"reference_active":true,"delta_db":-0.1999835605564759},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003622112164404606,"output_power":0.0003475157972203385,"reference_active":true,"delta_db":-0.1798734345601679},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6887620873910397e-05,"output_power":4.171290455664076e-05,"reference_active":true,"delta_db":-0.507877650190595}],"active_band_count":3,"active_band_mae_db":0.2959115484357463}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018229956701751592,"log_rms_mae_db":5.381736733807133,"rms_envelope_nrmse":0.4650579699018707,"envelope_correlation":0.7399795868813367,"rms_gain_error_db":-0.33441042826615075,"reference_mean_power":0.0070179714231847195,"output_mean_power":0.006497863031840318},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.8443307374696498e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0014276197130393435,"output_power":0.001367034324336954,"reference_active":true,"delta_db":-0.18833116967937438},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018503527061359352,"output_power":0.00017735546007039787,"reference_active":true,"delta_db":-0.18409956391199106},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.327921114523829e-06,"output_power":3.910902289000587e-06,"reference_active":true,"delta_db":-0.4400237067772653}],"active_band_count":3,"active_band_mae_db":0.2708181467895436}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00015227518148493586,"log_rms_mae_db":7.770848731129934,"rms_envelope_nrmse":0.6365758619270712,"envelope_correlation":0.652137988408343,"rms_gain_error_db":-0.3713921860126824,"reference_mean_power":0.0024461505543329222,"output_mean_power":0.0022456597998407853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3707367109300066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00065188242251555,"output_power":0.0006218582499198043,"reference_active":true,"delta_db":-0.20477870500932177},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.000159373128948768,"output_power":0.0001492349595335809,"reference_active":true,"delta_db":-0.28544527003194164},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8816561344170712e-05,"output_power":1.7104275187004698e-05,"reference_active":true,"delta_db":-0.4143558546734171}],"active_band_count":3,"active_band_mae_db":0.3015266099048935}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013579343349222632,"log_rms_mae_db":6.773254853231181,"rms_envelope_nrmse":0.538685835135268,"envelope_correlation":0.7265786840168585,"rms_gain_error_db":-0.37766929916983244,"reference_mean_power":0.0024424650949128645,"output_mean_power":0.002239037854330896},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.250982515304928e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006212562593255464,"output_power":0.0005964006839448506,"reference_active":true,"delta_db":-0.17732644573564885},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.001783099426803e-05,"output_power":7.476175402257472e-05,"reference_active":true,"delta_db":-0.29507292961681547},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7444705804029888e-06,"output_power":1.5936815924601653e-06,"reference_active":true,"delta_db":-0.3926209326726825}],"active_band_count":3,"active_band_mae_db":0.28834010267504895}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":7.560305347439797e-05,"log_rms_mae_db":8.016058906964988,"rms_envelope_nrmse":0.7592363245877715,"envelope_correlation":0.5313726222674948,"rms_gain_error_db":-0.12446270758339964,"reference_mean_power":0.00019825791077017447,"output_mean_power":0.0001926567610041873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.446126492709316e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.480334346914859e-05,"output_power":3.4284924346413375e-05,"reference_active":true,"delta_db":-0.06517772040527176},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0911626587448623e-05,"output_power":1.1382265755292012e-05,"reference_active":true,"delta_db":0.18339226035055944},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.788733059913985e-06,"output_power":2.777363386480665e-06,"reference_active":true,"delta_db":-0.017742392171294457}],"active_band_count":3,"active_band_mae_db":0.08877079097570856}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.513956828007689e-05,"log_rms_mae_db":6.506691191487898,"rms_envelope_nrmse":0.6219842708900681,"envelope_correlation":0.6460282766246995,"rms_gain_error_db":-0.15905545766731757,"reference_mean_power":0.00019869265760544442,"output_mean_power":0.00019154740508745192},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.64113687643322e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.340875190885912e-05,"output_power":3.3033617879641366e-05,"reference_active":true,"delta_db":-0.04904110752076185},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7578298997729145e-06,"output_power":6.069788859123491e-06,"reference_active":true,"delta_db":0.2291475354866397},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2467549734271305e-07,"output_power":2.003633915756773e-07,"reference_active":true,"delta_db":-0.49737337284402294}],"active_band_count":3,"active_band_mae_db":0.25852067195047485}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.7519381942341847e-05,"log_rms_mae_db":8.300689074061587,"rms_envelope_nrmse":0.8348557033964685,"envelope_correlation":0.5111310710112391,"rms_gain_error_db":0.2147927327653022,"reference_mean_power":3.3091629565415505e-05,"output_mean_power":3.476941849809016e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.646064632792655e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.725047198052175e-06,"output_power":6.8459085464176205e-06,"reference_active":true,"delta_db":0.07735756486625946},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.390363033635178e-06,"output_power":2.6257498282695547e-06,"reference_active":true,"delta_db":0.4078948182082145},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2031554280127926e-06,"output_power":1.4668105520126952e-06,"reference_active":true,"delta_db":0.8605229083699166}],"active_band_count":3,"active_band_mae_db":0.4485917638147969}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.2421508904297867e-05,"log_rms_mae_db":6.813162551743456,"rms_envelope_nrmse":0.6894745639892771,"envelope_correlation":0.6326948504207905,"rms_gain_error_db":0.19855151998750215,"reference_mean_power":3.324742134736325e-05,"output_mean_power":3.480271453588876e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.5226373785787838e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":6.9685546500066154e-06,"output_power":7.329899425261694e-06,"reference_active":true,"delta_db":0.21955305366547578},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4829248214027976e-06,"output_power":1.6707068485059116e-06,"reference_active":true,"delta_db":0.5178111833557906},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.259963132949634e-08,"output_power":1.1848893972193873e-07,"reference_active":true,"delta_db":1.0706855567168447}],"active_band_count":3,"active_band_mae_db":0.6026832645793704}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"hnm","status":"measured","mos":{"status":"measured","value":3.0623464584350586,"predicted_mos":3.0623464584350586,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a18132775a21bf6d656c36b9b89d5f564477233479c8cb4f52fda127018032a5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.78726338100023},"started_at":"2026-10-13T04:40:07.395383+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6089999917312525e-05,"native_input_samples":96000,"native_input_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","output_sha256":"a24d6d7345fca69c1a9f47fd67169f9f7f96059e8108e4afb76a9c164ea59e25","output_samples":96000,"length_delta_samples":0,"analysis_s":1.101817922999544,"synthesis_s":0.4867825580004137,"total_s":1.5886004809999577,"rtf":0.2647667468333263,"output_peak":0.5889633010791506,"output_rms":0.08989375438375226,"finished_at":"2026-10-13T04:40:08.990952+00:00","evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"a24d6d7345fca69c1a9f47fd67169f9f7f96059e8108e4afb76a9c164ea59e25","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.291826743189387,"spectral_convergence":0.22249741225105793,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.549341600279901,"spectral_convergence":0.21158588972725914,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.485688326930659,"spectral_convergence":0.22265879068341507,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.0029782049526686704,"delta_db":0.08747182350457114,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.003681625783624045,"delta_db":-0.03617525061360835,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0009082751290423498,"delta_db":-0.38784128551200037,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.00031559532500197835,"delta_db":-0.4118970145014478,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.0001307827004422478,"delta_db":-0.06115206360556631,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":6.640318642708663e-05,"delta_db":0.13433624782947345,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18647894759444458,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04676430112593434,"pitch":null,"output_rms_dbfs":-20.92540962048499,"output_peak":0.5889633010791506,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.775618890133316},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.291826743189387,"spectral_convergence":0.22249741225105793,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.549341600279901,"spectral_convergence":0.21158588972725914,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.485688326930659,"spectral_convergence":0.22265879068341507,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.0029782049526686704,"delta_db":0.08747182350457114,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.003681625783624045,"delta_db":-0.03617525061360835,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0009082751290423498,"delta_db":-0.38784128551200037,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.00031559532500197835,"delta_db":-0.4118970145014478,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.0001307827004422478,"delta_db":-0.06115206360556631,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":6.640318642708663e-05,"delta_db":0.13433624782947345,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18647894759444458,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.04676430112593434,"pitch":null,"output_rms_dbfs":-20.92540962048499,"output_peak":0.5889633010791506,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.775618890133316},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9084358271243489,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09037904209327785,"rms_dbfs":-20.87864531935906,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.554125087496356,"mean":1.0741869608561199e-05,"zero_energy":false},"output":{"rms":0.08989375438375226,"rms_dbfs":-20.92540962048499,"peak":0.5889633010791506,"peak_dbfs":-4.598235314287422,"crest_factor_db":16.32717430619757,"mean":-0.00014741471824280953,"zero_energy":false},"rms_difference_db":-0.04676430112593266},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":1.6203386850654382e-05,"output_rms_dbfs":-47.90394199234165,"output_zero_energy":false,"output_peak":0.09175967633243792,"output_energy_fraction":0.000294088596386815,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":184,"output_estimated_voiced_frames":196,"both_estimated_voiced_frames":181,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9836956521739131,"absolute_difference_cents_median":2.014591941235078,"absolute_difference_cents_p95":6.925076974534632,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023582673937264833,"log_rms_mae_db":0.9793757752517839,"rms_envelope_nrmse":0.06500225338486935,"envelope_correlation":0.9953105225297812,"rms_gain_error_db":-0.04676292670656287,"reference_mean_power":0.008175179133340451,"output_mean_power":0.008087624607155112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.411874313885781e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017924334098852407,"output_power":0.0017358943021961264,"reference_active":true,"delta_db":-0.13919752650772382},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00033647515979494424,"output_power":0.0003256628428405906,"reference_active":true,"delta_db":-0.14184798362336154},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.450579873055231e-05,"output_power":5.672273734740403e-05,"reference_active":true,"delta_db":0.1731447283077435}],"active_band_count":3,"active_band_mae_db":0.1513967461462763}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002175646104262597,"log_rms_mae_db":0.501682921543735,"rms_envelope_nrmse":0.034039980121859964,"envelope_correlation":0.9985768783446667,"rms_gain_error_db":-0.046749529042427226,"reference_mean_power":0.008216199965234223,"output_mean_power":0.008128231189134532},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0004980946791747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016116003385924513,"output_power":0.0015760826283394297,"reference_active":true,"delta_db":-0.09678367713404257},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017199808112094875,"output_power":0.0001647023342928017,"reference_active":true,"delta_db":-0.1882384739143807},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542016550968783e-06,"output_power":3.3780562415423885e-06,"reference_active":true,"delta_db":-0.20583710309153358}],"active_band_count":3,"active_band_mae_db":0.1636197513799856}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019245421130728438,"log_rms_mae_db":1.632003612858976,"rms_envelope_nrmse":0.10953542872743098,"envelope_correlation":0.9896230287551361,"rms_gain_error_db":-0.16613066958166794,"reference_mean_power":0.0024199693169718968,"output_mean_power":0.0023291464286512528},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.323542749940927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006913978889533925,"output_power":0.0006250200338406186,"reference_active":true,"delta_db":-0.4383411111691133},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018527392767596855,"output_power":0.0001676276848375366,"reference_active":true,"delta_db":-0.4346856147946083},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604035138941518e-05,"output_power":2.9096014618072392e-05,"reference_active":true,"delta_db":0.5552509169046749}],"active_band_count":3,"active_band_mae_db":0.47609254762279885}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017723694424605393,"log_rms_mae_db":0.9152460179466243,"rms_envelope_nrmse":0.062071687324639836,"envelope_correlation":0.9965335464277757,"rms_gain_error_db":-0.16613580139568823,"reference_mean_power":0.0024321306730885443,"output_mean_power":0.002340848595785999},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1677626980235007e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000654046958746599,"output_power":0.0005950929677000887,"reference_active":true,"delta_db":-0.4102411240139911},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.291953991407944e-05,"output_power":8.46628931125901e-05,"reference_active":true,"delta_db":-0.40413945583391503},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.049791938534598e-06,"output_power":2.1070965870735467e-06,"reference_active":true,"delta_db":0.11974662866791304}],"active_band_count":3,"active_band_mae_db":0.31137573617193975}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":2.024422432119929,"rms_envelope_nrmse":0.1862367053052408,"envelope_correlation":0.9717630363443271,"rms_gain_error_db":-0.3666966599840974,"reference_mean_power":0.0003849600012580848,"output_mean_power":0.0003537903159315717},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.553182623332075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001221299679169956,"output_power":0.00011192104186681739,"reference_active":true,"delta_db":-0.37910498685818006},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7738426055120482e-05,"output_power":1.969354160254044e-05,"reference_active":true,"delta_db":-1.4875798969566536},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.9827801273792e-06,"output_power":5.970383384275848e-06,"reference_active":true,"delta_db":-0.6802614730203026}],"active_band_count":3,"active_band_mae_db":0.8489821189450454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":1.2453156843110258,"rms_envelope_nrmse":0.11230571090266485,"envelope_correlation":0.9895901230640235,"rms_gain_error_db":-0.36669920963707797,"reference_mean_power":0.0003868956457514565,"output_mean_power":0.00035556902518773506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2949024312838274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011569002203634688,"output_power":0.00010508239954140595,"reference_active":true,"delta_db":-0.4176592251787075},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.480008250335501e-05,"output_power":1.0151751852044092e-05,"reference_active":true,"delta_db":-1.6372314300014612},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.80314470487776e-07,"output_power":3.711209954067779e-07,"reference_active":true,"delta_db":-1.9414787573114332}],"active_band_count":3,"active_band_mae_db":1.3321231374972007}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":2.207970572611655,"rms_envelope_nrmse":0.19059609047993295,"envelope_correlation":0.9750494952131864,"rms_gain_error_db":-0.1317478574192753,"reference_mean_power":8.262304385157993e-05,"output_mean_power":8.015422229816226e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.222558052994835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.126031645648209e-05,"output_power":1.999453400860765e-05,"reference_active":true,"delta_db":-0.26658437694032194},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.562486047803936e-06,"output_power":8.91261248956208e-06,"reference_active":true,"delta_db":-0.30565790207729326},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5273847358800203e-06,"output_power":2.1303061608330807e-06,"reference_active":true,"delta_db":-2.190108078985559}],"active_band_count":3,"active_band_mae_db":0.9207834526677248}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":1.380296977731945,"rms_envelope_nrmse":0.12173024088755105,"envelope_correlation":0.9890287030242703,"rms_gain_error_db":-0.1317503449382754,"reference_mean_power":8.303850395551146e-05,"output_mean_power":8.055722208692744e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.7676217824640817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1974657580034156e-05,"output_power":2.0531998025584885e-05,"reference_active":true,"delta_db":-0.29490902581717915},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.865612603723309e-06,"output_power":4.8101748914702166e-06,"reference_active":true,"delta_db":-0.861525094501179},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8280548950663716e-07,"output_power":1.7697718617781098e-07,"reference_active":true,"delta_db":-2.035705494283758}],"active_band_count":3,"active_band_mae_db":1.0640465382007054}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.595987558364868,"predicted_mos":3.595987558364868,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bb2f191293a49370d034fab1756c11090877aede361c6608a1e52b8b152b8bfa","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.604795802000808},"started_at":"2026-10-13T04:43:20.454285+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.9644000278494786e-05,"native_input_samples":96000,"native_input_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","output_sha256":"36d88baa7f2562ab0e2f29786214e3d2f6dc64f16fd04982a5898f62ad180c5a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.981480821000332,"synthesis_s":0.20586244900005113,"total_s":2.187343270000383,"rtf":0.36455721166673055,"output_peak":0.7748992106208388,"output_rms":0.0949364857683621,"finished_at":"2026-10-13T04:43:22.651091+00:00","evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"36d88baa7f2562ab0e2f29786214e3d2f6dc64f16fd04982a5898f62ad180c5a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.33585812791282,"spectral_convergence":0.7385788047007079,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4904335815152985,"spectral_convergence":0.26282686174464176,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.5450741284934155,"spectral_convergence":0.24728031465647415,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.002988723743995254,"delta_db":0.1027837469205292,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.004290262756559785,"delta_db":0.6282672340358091,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0011398159653602806,"delta_db":0.5983318504796932,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.00038684033302193387,"delta_db":0.4721148577705644,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.0001408714170979962,"delta_db":0.2615737648676477,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":6.6422114011169e-05,"delta_db":0.1355739857229701,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.36644090663286893,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.42730835033098236,"pitch":null,"output_rms_dbfs":-20.451336969028073,"output_peak":0.7748992106208388,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.123788612640511},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.33585812791282,"spectral_convergence":0.7385788047007079,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.4904335815152985,"spectral_convergence":0.26282686174464176,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.5450741284934155,"spectral_convergence":0.24728031465647415,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.002988723743995254,"delta_db":0.1027837469205292,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.004290262756559785,"delta_db":0.6282672340358091,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0011398159653602806,"delta_db":0.5983318504796932,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.00038684033302193387,"delta_db":0.4721148577705644,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.0001408714170979962,"delta_db":0.2615737648676477,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":6.6422114011169e-05,"delta_db":0.1355739857229701,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.36644090663286893,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.42730835033098236,"pitch":null,"output_rms_dbfs":-20.451336969028073,"output_peak":0.7748992106208388,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.123788612640511},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.922655711248538,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09037904209327785,"rms_dbfs":-20.87864531935906,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.554125087496356,"mean":1.0741869608561199e-05,"zero_energy":false},"output":{"rms":0.0949364857683621,"rms_dbfs":-20.451336969028073,"peak":0.7748992106208388,"peak_dbfs":-2.2150956303321343,"crest_factor_db":18.23624133869594,"mean":6.605208505184891e-07,"zero_energy":false},"rms_difference_db":0.4273083503309856},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":1.61153082137051e-05,"output_rms_dbfs":-47.927613839629196,"output_zero_energy":false,"output_peak":0.13184886260724088,"output_energy_fraction":0.00026224289748176977,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":184,"output_estimated_voiced_frames":192,"both_estimated_voiced_frames":167,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.907608695652174,"absolute_difference_cents_median":5.402370133350121,"absolute_difference_cents_p95":17.209027446702915,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023582673937264833,"log_rms_mae_db":1.8388092499695385,"rms_envelope_nrmse":0.15163748143602565,"envelope_correlation":0.9781838643809762,"rms_gain_error_db":0.4273101558277476,"reference_mean_power":0.008175179133340451,"output_mean_power":0.009020451861843592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.411874313885781e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017924334098852407,"output_power":0.0018808700414069777,"reference_active":true,"delta_db":0.20915758732586004},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00033647515979494424,"output_power":0.00032099752337113696,"reference_active":true,"delta_db":-0.20451326392334393},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.450579873055231e-05,"output_power":5.711054113734847e-05,"reference_active":true,"delta_db":0.20273567040304422}],"active_band_count":3,"active_band_mae_db":0.2054688405507494}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002175646104262597,"log_rms_mae_db":1.076311409684914,"rms_envelope_nrmse":0.08757434822071207,"envelope_correlation":0.9938537550880158,"rms_gain_error_db":0.4272739916561486,"reference_mean_power":0.008216199965234223,"output_mean_power":0.00906563855219471},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0004980946791747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016116003385924513,"output_power":0.0016926384955245723,"reference_active":true,"delta_db":0.2130686381121082},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017199808112094875,"output_power":0.00017157922929515913,"reference_active":true,"delta_db":-0.010588890551005672},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542016550968783e-06,"output_power":3.4861045030456283e-06,"reference_active":true,"delta_db":-0.06910184383554702}],"active_band_count":3,"active_band_mae_db":0.09758645749955364}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019245421130728438,"log_rms_mae_db":2.1386918145048193,"rms_envelope_nrmse":0.17568617000093026,"envelope_correlation":0.9810281568196921,"rms_gain_error_db":0.7189174831330942,"reference_mean_power":0.0024199693169718968,"output_mean_power":0.00285562784360735},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.323542749940927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006913978889533925,"output_power":0.000784872125769519,"reference_active":true,"delta_db":0.550708564908422},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018527392767596855,"output_power":0.00019945991960513776,"reference_active":true,"delta_db":0.3204133130957098},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604035138941518e-05,"output_power":2.2567820221159536e-05,"reference_active":true,"delta_db":-0.548188010470275}],"active_band_count":3,"active_band_mae_db":0.4731032961581356}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017723694424605393,"log_rms_mae_db":1.4180234688411533,"rms_envelope_nrmse":0.12633513615317288,"envelope_correlation":0.9922696946897328,"rms_gain_error_db":0.7189153956662587,"reference_mean_power":0.0024321306730885443,"output_mean_power":0.0028699771862259154},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1677626980235007e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000654046958746599,"output_power":0.0007364585975137675,"reference_active":true,"delta_db":0.5153940597066528},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.291953991407944e-05,"output_power":0.00010371124961358773,"reference_active":true,"delta_db":0.47718816345548987},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.049791938534598e-06,"output_power":1.694785849993273e-06,"reference_active":true,"delta_db":-0.825949514381489}],"active_band_count":3,"active_band_mae_db":0.6061772458478772}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":2.0241434551065813,"rms_envelope_nrmse":0.17780436424495666,"envelope_correlation":0.9771552118878356,"rms_gain_error_db":0.4153622953293462,"reference_mean_power":0.0003849600012580848,"output_mean_power":0.0004235959865229063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.553182623332075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001221299679169956,"output_power":0.00013057626623167484,"reference_active":true,"delta_db":0.2904200298303541},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7738426055120482e-05,"output_power":2.9826189468316482e-05,"reference_active":true,"delta_db":0.31515957824964497},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.9827801273792e-06,"output_power":6.356467705798371e-06,"reference_active":true,"delta_db":-0.40812522376288773}],"active_band_count":3,"active_band_mae_db":0.3379016106142956}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":1.2666680654748024,"rms_envelope_nrmse":0.09978448597387392,"envelope_correlation":0.9934258053266269,"rms_gain_error_db":0.41536277886125206,"reference_mean_power":0.0003868956457514565,"output_mean_power":0.0004257259467339413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2949024312838274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011569002203634688,"output_power":0.00012433650350765943,"reference_active":true,"delta_db":0.31302746522713054},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.480008250335501e-05,"output_power":1.637747311979991e-05,"reference_active":true,"delta_db":0.4398275891596126},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.80314470487776e-07,"output_power":5.087265864955824e-07,"reference_active":true,"delta_db":-0.5717896525847936}],"active_band_count":3,"active_band_mae_db":0.44154823565717893}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":2.346315060364379,"rms_envelope_nrmse":0.30293599591742754,"envelope_correlation":0.9416753415125078,"rms_gain_error_db":0.4036287371682157,"reference_mean_power":8.262304385157993e-05,"output_mean_power":9.067009362096539e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.222558052994835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.126031645648209e-05,"output_power":2.2106305659604155e-05,"reference_active":true,"delta_db":0.16946445992919706},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.562486047803936e-06,"output_power":1.165250424124039e-05,"reference_active":true,"delta_db":0.8584845517527193},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5273847358800203e-06,"output_power":3.234842213821863e-06,"reference_active":true,"delta_db":-0.37599729344030913}],"active_band_count":3,"active_band_mae_db":0.46798210170740845}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":1.4856102180147521,"rms_envelope_nrmse":0.18734449255840252,"envelope_correlation":0.9767708792340674,"rms_gain_error_db":0.40362830359602503,"reference_mean_power":8.303850395551146e-05,"output_mean_power":9.112600825572822e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.7676217824640817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1974657580034156e-05,"output_power":2.3560393033713524e-05,"reference_active":true,"delta_db":0.3026041465451864},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.865612603723309e-06,"output_power":6.4389389196943465e-06,"reference_active":true,"delta_db":0.40500928885868315},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8280548950663716e-07,"output_power":2.469692592764141e-07,"reference_active":true,"delta_db":-0.5884493606885773}],"active_band_count":3,"active_band_mae_db":0.43202093203081565}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"facodec","status":"measured","mos":{"status":"measured","value":4.424135208129883,"predicted_mos":4.424135208129883,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"baa53f671c8a840b47cdaf809d68b8dc0c5f29b75650a328ffc6751886f7b7a5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.074329510000098},"started_at":"2026-10-13T06:59:06.831778+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7755999326473102e-05,"native_input_samples":96000,"native_input_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","output_sha256":"67989f59a108745ed8f3d16ab1716991e4ce692a16ee43beb9c8853071d3022a","output_samples":96000,"length_delta_samples":0,"analysis_s":5.735432084999047,"synthesis_s":7.689099606999662,"total_s":13.424531691998709,"rtf":2.2374219486664515,"output_peak":0.5822151899337769,"output_rms":0.08634854205529283,"finished_at":"2026-10-13T06:59:20.284310+00:00","evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"67989f59a108745ed8f3d16ab1716991e4ce692a16ee43beb9c8853071d3022a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5755438691838135,"spectral_convergence":0.22252020156192337,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7532563392330545,"spectral_convergence":0.21968291628665046,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.831180991151991,"spectral_convergence":0.23862525638821483,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.0025195949478662486,"delta_db":-0.6387667030317413,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.0035000987804742253,"delta_db":-0.25576866634958445,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0009557343632227025,"delta_db":-0.16664349525452343,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.0003017830460973439,"delta_db":-0.6062542486694885,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.000125342430589182,"delta_db":-0.24567394821970734,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":5.351714682497034e-05,"delta_db":-0.8026234345443889,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4526217493449057,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.396254497507495,"pitch":null,"output_rms_dbfs":-21.27489981686655,"output_peak":0.5822151899337769,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.719993733189621},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5755438691838135,"spectral_convergence":0.22252020156192337,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7532563392330545,"spectral_convergence":0.21968291628665046,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.831180991151991,"spectral_convergence":0.23862525638821483,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.0025195949478662486,"delta_db":-0.6387667030317413,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.0035000987804742253,"delta_db":-0.25576866634958445,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0009557343632227025,"delta_db":-0.16664349525452343,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.0003017830460973439,"delta_db":-0.6062542486694885,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.000125342430589182,"delta_db":-0.24567394821970734,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":5.351714682497034e-05,"delta_db":-0.8026234345443889,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4526217493449057,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.396254497507495,"pitch":null,"output_rms_dbfs":-21.27489981686655,"output_peak":0.5822151899337769,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.719993733189621},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9053908939774147,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09037904209327785,"rms_dbfs":-20.87864531935906,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.554125087496356,"mean":1.0741869608561199e-05,"zero_energy":false},"output":{"rms":0.08634854205529283,"rms_dbfs":-21.27489981686655,"peak":0.5822151899337769,"peak_dbfs":-4.698329361030291,"crest_factor_db":16.57657045583626,"mean":2.4925505026646516e-05,"zero_energy":false},"rms_difference_db":-0.396254497507492},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":1.9635780847286515e-06,"output_rms_dbfs":-57.069518236563646,"output_zero_energy":false,"output_peak":0.0079045994207263,"output_energy_fraction":3.862509670738612e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":184,"output_estimated_voiced_frames":172,"both_estimated_voiced_frames":161,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":2.8181062672394224,"absolute_difference_cents_p95":10.681169340452128,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023582673937264833,"log_rms_mae_db":1.2116061684436168,"rms_envelope_nrmse":0.089926874862817,"envelope_correlation":0.9928914452951251,"rms_gain_error_db":-0.3962559299819602,"reference_mean_power":0.008175179133340451,"output_mean_power":0.007462282474437512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.411874313885781e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017924334098852407,"output_power":0.0016256322132985677,"reference_active":true,"delta_db":-0.4242073369821646},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00033647515979494424,"output_power":0.00035734224117831535,"reference_active":true,"delta_db":0.2613134883795325},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.450579873055231e-05,"output_power":5.73855539093993e-05,"reference_active":true,"delta_db":0.22359869791427556}],"active_band_count":3,"active_band_mae_db":0.3030398410919909}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002175646104262597,"log_rms_mae_db":0.9525347602641384,"rms_envelope_nrmse":0.0785723612502819,"envelope_correlation":0.994466328186594,"rms_gain_error_db":-0.39623911303823894,"reference_mean_power":0.008216199965234223,"output_mean_power":0.007499755225053668},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0004980946791747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016116003385924513,"output_power":0.0014687716143484426,"reference_active":true,"delta_db":-0.40303079247881546},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017199808112094875,"output_power":0.0001840405885726425,"reference_active":true,"delta_db":0.29390011721834164},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542016550968783e-06,"output_power":3.8091363944161657e-06,"reference_active":true,"delta_db":0.31575937588511466}],"active_band_count":3,"active_band_mae_db":0.3375634285274239}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019245421130728438,"log_rms_mae_db":1.6713424890686042,"rms_envelope_nrmse":0.10904912767933038,"envelope_correlation":0.9899216614235357,"rms_gain_error_db":-0.21273441973379825,"reference_mean_power":0.0024199693169718968,"output_mean_power":0.00230428619289651},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.323542749940927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006913978889533925,"output_power":0.0007034546220508579,"reference_active":true,"delta_db":0.07508038339392695},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018527392767596855,"output_power":0.00021062668761683814,"reference_active":true,"delta_db":0.55699089606265},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604035138941518e-05,"output_power":2.5994793225420497e-05,"reference_active":true,"delta_db":0.06577952655349162}],"active_band_count":3,"active_band_mae_db":0.2326169353366895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017723694424605393,"log_rms_mae_db":1.2111655569785456,"rms_envelope_nrmse":0.08887516070112446,"envelope_correlation":0.9926520226777558,"rms_gain_error_db":-0.21274789178350723,"reference_mean_power":0.0024321306730885443,"output_mean_power":0.0023158590090877687},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1677626980235007e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000654046958746599,"output_power":0.0006731379694643082,"reference_active":true,"delta_db":0.12495157773044759},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.291953991407944e-05,"output_power":0.00010631895937577727,"reference_active":true,"delta_db":0.5850366643766904},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.049791938534598e-06,"output_power":2.254704861140321e-06,"reference_active":true,"delta_db":0.4137992037702014}],"active_band_count":3,"active_band_mae_db":0.3745958152924465}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":1.7858393902790866,"rms_envelope_nrmse":0.1774930089554963,"envelope_correlation":0.9756054473683312,"rms_gain_error_db":-0.5065366984026408,"reference_mean_power":0.0003849600012580848,"output_mean_power":0.0003425799465249562},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.553182623332075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001221299679169956,"output_power":0.0001064880494693135,"reference_active":true,"delta_db":-0.5952137083093781},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7738426055120482e-05,"output_power":2.5819852747838937e-05,"reference_active":true,"delta_db":-0.31128053394859584},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.9827801273792e-06,"output_power":8.616205879077421e-06,"reference_active":true,"delta_db":0.9128770035655962}],"active_band_count":3,"active_band_mae_db":0.6064570819411901}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":1.365750035575309,"rms_envelope_nrmse":0.14808592102400325,"envelope_correlation":0.9818004521044869,"rms_gain_error_db":-0.5065474042278058,"reference_mean_power":0.0003868956457514565,"output_mean_power":0.00034430164813891546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2949024312838274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011569002203634688,"output_power":0.00010189605309524977,"reference_active":true,"delta_db":-0.5513854169574137},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.480008250335501e-05,"output_power":1.3458307561656485e-05,"reference_active":true,"delta_db":-0.4127368740690068},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.80314470487776e-07,"output_power":7.256406369628173e-07,"reference_active":true,"delta_db":0.970581957829802}],"active_band_count":3,"active_band_mae_db":0.6449014162854074}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":1.856794270831677,"rms_envelope_nrmse":0.18039578472403164,"envelope_correlation":0.9785541068135426,"rms_gain_error_db":-0.45734404330379796,"reference_mean_power":8.262304385157993e-05,"output_mean_power":7.436469244039931e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.222558052994835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.126031645648209e-05,"output_power":1.7769764500980644e-05,"reference_active":true,"delta_db":-0.7788805241240604},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.562486047803936e-06,"output_power":8.03057134959228e-06,"reference_active":true,"delta_db":-0.7582436945442168},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5273847358800203e-06,"output_power":3.7449766216192966e-06,"reference_active":true,"delta_db":0.25996279672509176}],"active_band_count":3,"active_band_mae_db":0.599029005131123}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":1.4993253945953648,"rms_envelope_nrmse":0.12749983904619366,"envelope_correlation":0.9893452622443247,"rms_gain_error_db":-0.4573458915987164,"reference_mean_power":8.303850395551146e-05,"output_mean_power":7.473859460331618e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.7676217824640817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1974657580034156e-05,"output_power":1.838841543108202e-05,"reference_active":true,"delta_db":-0.7737780963740231},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.865612603723309e-06,"output_power":5.091599503763392e-06,"reference_active":true,"delta_db":-0.6145914098541396},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8280548950663716e-07,"output_power":3.195315444919242e-07,"reference_active":true,"delta_db":0.5302590337159714}],"active_band_count":3,"active_band_mae_db":0.6395428466480447}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.6027770042419434,"predicted_mos":3.6027770042419434,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"5f15e14af5ac2e43ddbe683974269866cc04ed4870cb79157d2a26d9eb4ad00c","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.9117124379990855},"started_at":"2026-10-13T07:28:41.470686+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006677047000266612,"native_input_samples":264600,"native_input_sha256":"32dc087a551900f3fa94a1d13af3f6c2fd3515df661ae871379141b0315ec6a1","output_sha256":"437a40ba133df54927fbb6afe9c61d7fe992e72fd9e0ddd24b6b653d57b451fc","output_samples":264704,"length_delta_samples":104,"analysis_s":0.14103282200085232,"synthesis_s":3.9909059819983668,"total_s":4.131938803999219,"rtf":0.6886564673332032,"output_peak":0.5941495895385742,"output_rms":0.08101368073642505,"finished_at":"2026-10-13T07:28:45.631182+00:00","evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"437a40ba133df54927fbb6afe9c61d7fe992e72fd9e0ddd24b6b653d57b451fc","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.82603033767097,"output_peak":0.5957143072906066},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.688565909854773,"spectral_convergence":0.5012136790272362,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.561339250869421,"spectral_convergence":0.3406141370087403,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.428851359323906,"spectral_convergence":0.3836368136819734,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.0020272033718869587,"delta_db":-1.5831007925136742,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.0032901490827271715,"delta_db":-0.5244159082190922,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0008780375687158614,"delta_db":-0.5348845225707481,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.00022814313400138322,"delta_db":-1.8211285956086543,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.00010377622582312392,"delta_db":-1.0656763488461425,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":4.27356175517258e-05,"delta_db":-1.7796531104498334,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.218143213034691,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9456712384807842,"pitch":null,"output_rms_dbfs":-21.824316557839843,"output_peak":0.5957143072906066,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.559585506682701},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7785337335668618,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09037904209327785,"rms_dbfs":-20.87864531935906,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.554125087496356,"mean":1.0741869608561199e-05,"zero_energy":false},"output":{"rms":0.08105581410797269,"rms_dbfs":-21.824316557839843,"peak":0.5957143072906066,"peak_dbfs":-4.499239386213639,"crest_factor_db":17.325077171626205,"mean":-0.0002420004685692,"zero_energy":false},"rms_difference_db":-0.9456712384807844},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":1.6943273822493598e-06,"output_rms_dbfs":-57.71002670411885,"output_zero_energy":false,"output_peak":0.0084725522482155,"output_energy_fraction":3.782338619748585e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":184,"output_estimated_voiced_frames":145,"both_estimated_voiced_frames":130,"reference_voiced_output_unvoiced_frames":54,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.7065217391304348,"absolute_difference_cents_median":7.565058388504083,"absolute_difference_cents_p95":22.788183751043675,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023582673937264833,"log_rms_mae_db":1.836827553929802,"rms_envelope_nrmse":0.15204326269264376,"envelope_correlation":0.9852888789471316,"rms_gain_error_db":-0.9456705786305059,"reference_mean_power":0.008175179133340451,"output_mean_power":0.006575521767274151},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.411874313885781e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017924334098852407,"output_power":0.0016149928253659006,"reference_active":true,"delta_db":-0.4527243299497088},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00033647515979494424,"output_power":0.0002972433250493637,"reference_active":true,"delta_db":-0.5384089724462823},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.450579873055231e-05,"output_power":4.957638487036257e-05,"reference_active":true,"delta_db":-0.4116785353740488}],"active_band_count":3,"active_band_mae_db":0.4676039459233466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002175646104262597,"log_rms_mae_db":1.504031758469449,"rms_envelope_nrmse":0.1297377061710743,"envelope_correlation":0.992003348894198,"rms_gain_error_db":-0.9456747013109056,"reference_mean_power":0.008216199965234223,"output_mean_power":0.006608509678966442},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0004980946791747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016116003385924513,"output_power":0.0014615226715535595,"reference_active":true,"delta_db":-0.42451793315636144},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017199808112094875,"output_power":0.00014992573613284308,"reference_active":true,"delta_db":-0.5964741187882452},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542016550968783e-06,"output_power":3.2582695781575936e-06,"reference_active":true,"delta_db":-0.3626357270038632}],"active_band_count":3,"active_band_mae_db":0.4612092596494899}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019245421130728438,"log_rms_mae_db":2.6190737397265007,"rms_envelope_nrmse":0.19380619994048046,"envelope_correlation":0.9679816073501603,"rms_gain_error_db":-0.30781689270253254,"reference_mean_power":0.0024199693169718968,"output_mean_power":0.0022543854369337962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.323542749940927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006913978889533925,"output_power":0.0007324435349311483,"reference_active":true,"delta_db":0.2504610078781465},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018527392767596855,"output_power":0.0002206888408935631,"reference_active":true,"delta_db":0.7596606531836585},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604035138941518e-05,"output_power":2.829657707807753e-05,"reference_active":true,"delta_db":0.43425489313844867}],"active_band_count":3,"active_band_mae_db":0.4814588514000846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017723694424605393,"log_rms_mae_db":1.8240720050486887,"rms_envelope_nrmse":0.15450723881299283,"envelope_correlation":0.9775800627814488,"rms_gain_error_db":-0.3078200880026948,"reference_mean_power":0.0024321306730885443,"output_mean_power":0.002265712997936864},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1677626980235007e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000654046958746599,"output_power":0.0007098335574648434,"reference_active":true,"delta_db":0.35547596243932267},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.291953991407944e-05,"output_power":0.00010881756982310478,"reference_active":true,"delta_db":0.6859197201406441},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.049791938534598e-06,"output_power":2.5067280150223936e-06,"reference_active":true,"delta_db":0.8739743390278258}],"active_band_count":3,"active_band_mae_db":0.6384566738692642}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":3.1856027446844757,"rms_envelope_nrmse":0.25007723476929,"envelope_correlation":0.9643105449361074,"rms_gain_error_db":-1.4991427806788185,"reference_mean_power":0.0003849600012580848,"output_mean_power":0.00027258460804445607},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.553182623332075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001221299679169956,"output_power":8.961574460894091e-05,"reference_active":true,"delta_db":-1.3443792535773402},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7738426055120482e-05,"output_power":1.985906931226899e-05,"reference_active":true,"delta_db":-1.451229229344442},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.9827801273792e-06,"output_power":6.274630213849714e-06,"reference_active":true,"delta_db":-0.46440230866375726}],"active_band_count":3,"active_band_mae_db":1.0866702638618466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":2.6070154097472744,"rms_envelope_nrmse":0.2166132917182336,"envelope_correlation":0.976780933410671,"rms_gain_error_db":-1.499139803798863,"reference_mean_power":0.0003868956457514565,"output_mean_power":0.00027395539769229857},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2949024312838274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011569002203634688,"output_power":8.610999374322418e-05,"reference_active":true,"delta_db":-1.2824234612174823},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.480008250335501e-05,"output_power":1.058862526741724e-05,"reference_active":true,"delta_db":-1.4542455753622254},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.80314470487776e-07,"output_power":5.60577848284743e-07,"reference_active":true,"delta_db":-0.1502746784794862}],"active_band_count":3,"active_band_mae_db":0.962314571686398}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":3.2333534500135337,"rms_envelope_nrmse":0.2894051793079018,"envelope_correlation":0.945062990570688,"rms_gain_error_db":-1.0041039027570269,"reference_mean_power":8.262304385157993e-05,"output_mean_power":6.556782840784116e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.222558052994835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.126031645648209e-05,"output_power":1.8676000989587876e-05,"reference_active":true,"delta_db":-0.5628583636997103},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.562486047803936e-06,"output_power":5.837177472302949e-06,"reference_active":true,"delta_db":-2.1436791679447103},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5273847358800203e-06,"output_power":2.659891894712301e-06,"reference_active":true,"delta_db":-1.225888451727258}],"active_band_count":3,"active_band_mae_db":1.310808661123893}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":2.871878306923906,"rms_envelope_nrmse":0.23891307314848034,"envelope_correlation":0.9623643424083335,"rms_gain_error_db":-1.0041012910733742,"reference_mean_power":8.303850395551146e-05,"output_mean_power":6.589756802703446e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.7676217824640817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1974657580034156e-05,"output_power":1.8339008215575872e-05,"reference_active":true,"delta_db":-0.7854627134114263},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.865612603723309e-06,"output_power":3.56061862121163e-06,"reference_active":true,"delta_db":-2.167879177023289},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8280548950663716e-07,"output_power":2.106829447708579e-07,"reference_active":true,"delta_db":-1.2785845527377984}],"active_band_count":3,"active_band_mae_db":1.410642147724171}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"golf","status":"measured","mos":{"status":"measured","value":2.947385549545288,"predicted_mos":2.947385549545288,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"e5c211a455fa2c4147bc526da6aefe504b5fc1feddd034b06b53f310a78170d5","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.3566397430004145},"started_at":"2026-10-13T05:02:59.219679+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003106776999629801,"native_input_samples":144000,"native_input_sha256":"aefb14bdcf832d1ad6e1871c426a954eba41c8f3c45424345324561bc347a2fe","output_sha256":"82e0b558e78fabd15c17bfd5aa67e7013c0e8e2973ff3c735ba5c317a0546046","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5709844969996993,"synthesis_s":0.12293588000011368,"total_s":0.693920376999813,"rtf":0.1156533961666355,"output_peak":0.5832095146179199,"output_rms":0.08402370925958293,"finished_at":"2026-10-13T05:02:59.925942+00:00","analysis_stages":{"external_f0_s":0.09026581599937344},"evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"82e0b558e78fabd15c17bfd5aa67e7013c0e8e2973ff3c735ba5c317a0546046","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.510070044607403,"output_peak":0.5835757161646207},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.051491898305159,"spectral_convergence":0.749364926570595,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.255833562472896,"spectral_convergence":0.30196044756993196,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.437340815034358,"spectral_convergence":0.27565417055191954,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029236876060789864,"output_power":0.0025719127163715315,"delta_db":-0.5567474096663807,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003718611252604551,"output_power":0.00335446241830896,"delta_db":-0.44757848521314203,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009947805044625932,"output_power":0.0007565488498311868,"delta_db":-1.1889029009651422,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003475711156445355,"output_power":0.00023983874628423791,"delta_db":-1.6112433260015773,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013285860243863903,"output_power":9.54795999790872e-05,"delta_db":-1.4347908889462355,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.448803723597661e-05,"output_power":4.481929528674312e-05,"delta_db":-1.5801413525855668,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1365673938963408,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6386632667334574,"pitch":null,"output_rms_dbfs":-21.510070044607403,"output_peak":0.5835757161646207,"reference_rms_dbfs":-20.871406777873943,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.914888758604138},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8874863126340109,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09045439247745397,"rms_dbfs":-20.87140677787394,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.54688654601124,"mean":1.0873797739885486e-05,"zero_energy":false},"output":{"rms":0.08404202297697115,"rms_dbfs":-21.510070044607403,"peak":0.5835757161646207,"peak_dbfs":-4.67805576652201,"crest_factor_db":16.83201427808539,"mean":0.0013239745164211275,"zero_energy":false},"rms_difference_db":-0.6386632667334631},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":3.210129929094771e-06,"output_rms_dbfs":-54.934773892939006,"output_zero_energy":false,"output_peak":0.04558532258372365,"output_energy_fraction":6.677062861355402e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":183,"output_estimated_voiced_frames":211,"both_estimated_voiced_frames":174,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9508196721311475,"absolute_difference_cents_median":5.556195983535998,"absolute_difference_cents_p95":21.672696449502094,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00023582673937264833,"log_rms_mae_db":1.8114607731991157,"rms_envelope_nrmse":0.15423038547680581,"envelope_correlation":0.9775199527079613,"rms_gain_error_db":-0.6386640889454704,"reference_mean_power":0.008188830573425326,"output_mean_power":0.007068959228174936},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.4135607304709467e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017974165581169654,"output_power":0.001560387395949992,"reference_active":true,"delta_db":-0.614163045356304},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003352953555022715,"output_power":0.00025258795179883086,"reference_active":true,"delta_db":-1.2301490635453711},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.4503000255596816e-05,"output_power":6.668339860536997e-05,"reference_active":true,"delta_db":0.8759731632451602}],"active_band_count":3,"active_band_mae_db":0.9067617573822785}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002175646104262597,"log_rms_mae_db":1.2033894822748246,"rms_envelope_nrmse":0.10520490791262965,"envelope_correlation":0.9918101877175379,"rms_gain_error_db":-0.6385455173912358,"reference_mean_power":0.008229630304583593,"output_mean_power":0.0071043733148647885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.0017952111630574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016166193353564513,"output_power":0.0013979039007734236,"reference_active":true,"delta_db":-0.631304520525318},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017023114679964348,"output_power":0.00012917282657081802,"reference_active":true,"delta_db":-1.1986786192643961},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.532695240105371e-06,"output_power":4.358862593958109e-06,"reference_active":true,"delta_db":0.9126700563795591}],"active_band_count":3,"active_band_mae_db":0.9142177320564245}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019245421130728438,"log_rms_mae_db":2.1283485383239213,"rms_envelope_nrmse":0.17508654466159623,"envelope_correlation":0.9765991291676077,"rms_gain_error_db":-0.7003778593427561,"reference_mean_power":0.0024240105149714135,"output_mean_power":0.0020629880555316704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.32389979202062e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006930110951774774,"output_power":0.0005982881110760788,"reference_active":true,"delta_db":-0.6382981495118263},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018492060037791357,"output_power":0.0001590968688321918,"reference_active":true,"delta_db":-0.6532366235408517},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.556483749032591e-05,"output_power":2.8058223998424025e-05,"reference_active":true,"delta_db":0.4041714165863103}],"active_band_count":3,"active_band_mae_db":0.5652353965463294}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017723694424605393,"log_rms_mae_db":1.3198270103102225,"rms_envelope_nrmse":0.1264184172299751,"envelope_correlation":0.9893327581670733,"rms_gain_error_db":-0.7004112445326294,"reference_mean_power":0.002436197767926035,"output_mean_power":0.002073344249309474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1679563951674205e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006559410037167136,"output_power":0.0005683741434043742,"reference_active":true,"delta_db":-0.622304673686143},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.208551922919727e-05,"output_power":7.90282309804445e-05,"reference_active":true,"delta_db":-0.6640908071928635},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0400901152656994e-06,"output_power":2.3637093069357766e-06,"reference_active":true,"delta_db":0.6394471361396025}],"active_band_count":3,"active_band_mae_db":0.6419475390062029}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":2.275945144468466,"rms_envelope_nrmse":0.2545765928871878,"envelope_correlation":0.9664313646636983,"rms_gain_error_db":-1.6886889886166676,"reference_mean_power":0.000385603031150758,"output_mean_power":0.00026137951049206134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.5514740627049277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012212637270461992,"output_power":7.922048218914318e-05,"reference_active":true,"delta_db":-1.87971976754474},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7711994854361194e-05,"output_power":2.0164682654607205e-05,"reference_active":true,"delta_db":-1.3807639808794403},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.965479580829498e-06,"output_power":5.365516976713235e-06,"reference_active":true,"delta_db":-1.133394501037019}],"active_band_count":3,"active_band_mae_db":1.464626083153733}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":1.7216328502610612,"rms_envelope_nrmse":0.20440789767387813,"envelope_correlation":0.988289644628348,"rms_gain_error_db":-1.6886950012612334,"reference_mean_power":0.0003875448511767381,"output_mean_power":0.00026269540192974975},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.2930406616783493e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011575718857401511,"output_power":7.508248445124203e-05,"reference_active":true,"delta_db":-1.8800933558358648},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4650732617980191e-05,"output_power":1.0368033972587005e-05,"reference_active":true,"delta_db":-1.501629308017616},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.798271360693588e-07,"output_power":4.2856360821716245e-07,"reference_active":true,"delta_db":-1.3128324674828384}],"active_band_count":3,"active_band_mae_db":1.5648517104454396}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":2.575920832166527,"rms_envelope_nrmse":0.2066371586494362,"envelope_correlation":0.9732532405377773,"rms_gain_error_db":-0.5955105177839422,"reference_mean_power":8.27610489495666e-05,"output_mean_power":7.21564128381677e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.2191138266404684e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.121026355120556e-05,"output_power":1.95386352440635e-05,"reference_active":true,"delta_db":-0.35651839572359423},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.608610444628603e-06,"output_power":7.550323668324471e-06,"reference_active":true,"delta_db":-1.0469501700953219},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.526221838032985e-06,"output_power":3.4992865401538186e-06,"reference_active":true,"delta_db":-0.033301244333505634}],"active_band_count":3,"active_band_mae_db":0.47892327005080726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":2.1996855237512123,"rms_envelope_nrmse":0.1421478609043304,"envelope_correlation":0.9884046945795021,"rms_gain_error_db":-0.5955111647834386,"reference_mean_power":8.317788961585088e-05,"output_mean_power":7.25198305745577e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.76312993185785e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1942334982313343e-05,"output_power":2.0228555014491583e-05,"reference_active":true,"delta_db":-0.353179799826426},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.882415506722356e-06,"output_power":4.7254674241858825e-06,"reference_active":true,"delta_db":-0.9511092418601728},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.834076248063091e-07,"output_power":2.8569928215763835e-07,"reference_active":true,"delta_db":0.034976188454295384}],"active_band_count":3,"active_band_mae_db":0.4464217433802981}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.5622596740722656,"predicted_mos":3.5622596740722656,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"46b00d6aa50287bb715de3455ec76e533ceca1045877db488b463df50f4e1a53","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3197515630017733},"started_at":"2026-10-13T07:45:01.270886+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028724130024784245,"native_input_samples":144000,"native_input_sha256":"aefb14bdcf832d1ad6e1871c426a954eba41c8f3c45424345324561bc347a2fe","output_sha256":"3457217dcaedce4078f48a2e2b170cc84db386ff42233299ac50bb6344606b9f","output_samples":144120,"length_delta_samples":120,"analysis_s":2.3091679429999203,"synthesis_s":3.003195995999704,"total_s":5.312363938999624,"rtf":0.8853939898332707,"output_peak":0.5827421545982361,"output_rms":0.09435692121251378,"finished_at":"2026-10-13T07:45:06.596733+00:00","analysis_stages":{"external_f0_s":1.437943835000624},"evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"3457217dcaedce4078f48a2e2b170cc84db386ff42233299ac50bb6344606b9f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.503231891060526,"output_peak":0.5786537515623412},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.403261388642377,"spectral_convergence":0.8047961173210365,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.426031498482208,"spectral_convergence":0.34866052206684583,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.221195288698306,"spectral_convergence":0.29871884913396723,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.004179023706784541,"delta_db":1.5586743565648289,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.003399959394656373,"delta_db":-0.38183437560562367,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0008505400978802662,"delta_db":-0.6730675795985563,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.0003200461261109824,"delta_db":-0.35107687920366326,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.00012008962216029763,"delta_db":-0.43160028528706496,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":4.3634213166148005e-05,"delta_db":-1.6892814620578216,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8475891563862598,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.3790272276373002,"pitch":null,"output_rms_dbfs":-20.499618091721757,"output_peak":0.5786537515623412,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.016829391940964},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8915353806107569,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09037904209327785,"rms_dbfs":-20.87864531935906,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.554125087496356,"mean":1.0741869608561199e-05,"zero_energy":false},"output":{"rms":0.09441023864369061,"rms_dbfs":-20.499618091721757,"peak":0.5786537515623412,"peak_dbfs":-4.751624537761267,"crest_factor_db":15.747993553960491,"mean":0.00013579658193950658,"zero_energy":false},"rms_difference_db":0.3790272276373017},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":8.434939349627489e-05,"output_rms_dbfs":-40.739180358035824,"output_zero_energy":false,"output_peak":0.2508063572435309,"output_energy_fraction":0.0013879543908551005,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":184,"output_estimated_voiced_frames":216,"both_estimated_voiced_frames":178,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":38,"paired_voiced_fraction":0.967391304347826,"absolute_difference_cents_median":5.819392157123978,"absolute_difference_cents_p95":23.11636853662471,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023582673937264833,"log_rms_mae_db":2.348933198658326,"rms_envelope_nrmse":0.1822186094452953,"envelope_correlation":0.9665828234649315,"rms_gain_error_db":0.3790277377177753,"reference_mean_power":0.008175179133340451,"output_mean_power":0.00892072294317138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.411874313885781e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017924334098852407,"output_power":0.0020080672058820194,"reference_active":true,"delta_db":0.49335213354393626},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00033647515979494424,"output_power":0.0003253541769304326,"reference_active":true,"delta_db":-0.145966214253501},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.450579873055231e-05,"output_power":5.015382620007814e-05,"reference_active":true,"delta_db":-0.36138637557559616}],"active_band_count":3,"active_band_mae_db":0.3335682411243444}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002175646104262597,"log_rms_mae_db":1.68394961198642,"rms_envelope_nrmse":0.11603609515900687,"envelope_correlation":0.9861270028229189,"rms_gain_error_db":0.3790652202451186,"reference_mean_power":0.008216199965234223,"output_mean_power":0.008965562090203206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0004980946791747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016116003385924513,"output_power":0.0017967770536685493,"reference_active":true,"delta_db":0.4723684269215579},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017199808112094875,"output_power":0.0001677618115134627,"reference_active":true,"delta_db":-0.10830494722600219},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542016550968783e-06,"output_power":3.2384716202929175e-06,"reference_active":true,"delta_db":-0.3891049065709602}],"active_band_count":3,"active_band_mae_db":0.32325942690617343}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019245421130728438,"log_rms_mae_db":2.9833700361565407,"rms_envelope_nrmse":0.21892685002553497,"envelope_correlation":0.959423936161625,"rms_gain_error_db":-0.5663578125997403,"reference_mean_power":0.0024199693169718968,"output_mean_power":0.002124095692841462},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.323542749940927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006913978889533925,"output_power":0.0006133123402077082,"reference_active":true,"delta_db":-0.5204634613844626},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018527392767596855,"output_power":0.00015450391171127587,"reference_active":true,"delta_db":-0.7887482903180238},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604035138941518e-05,"output_power":2.3380743332470764e-05,"reference_active":true,"delta_db":-0.39450100193167664}],"active_band_count":3,"active_band_mae_db":0.5679042512113877}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017723694424605393,"log_rms_mae_db":2.2604250516754454,"rms_envelope_nrmse":0.1462876345900734,"envelope_correlation":0.9820492173161756,"rms_gain_error_db":-0.5663526659582561,"reference_mean_power":0.0024321306730885443,"output_mean_power":0.0021347726903022494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1677626980235007e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000654046958746599,"output_power":0.0005790701425387844,"reference_active":true,"delta_db":-0.528777577419635},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.291953991407944e-05,"output_power":7.73905152308462e-05,"reference_active":true,"delta_db":-0.7941931274418363},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.049791938534598e-06,"output_power":1.983416129772702e-06,"reference_active":true,"delta_db":-0.14295940063168736}],"active_band_count":3,"active_band_mae_db":0.4886433684977196}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":2.62390290896293,"rms_envelope_nrmse":0.2428682069494707,"envelope_correlation":0.9509680265492667,"rms_gain_error_db":-0.3312542186864893,"reference_mean_power":0.0003849600012580848,"output_mean_power":0.0003566893851567695},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.553182623332075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001221299679169956,"output_power":0.00011423039497911853,"reference_active":true,"delta_db":-0.2904056453682249},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7738426055120482e-05,"output_power":2.465614013005973e-05,"reference_active":true,"delta_db":-0.5115672488575749},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.9827801273792e-06,"output_power":7.201376169262011e-06,"reference_active":true,"delta_db":0.13387129946205942}],"active_band_count":3,"active_band_mae_db":0.31194806456261975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":1.7736150004866478,"rms_envelope_nrmse":0.139136028666534,"envelope_correlation":0.9828763357312231,"rms_gain_error_db":-0.33125700562490284,"reference_mean_power":0.0003868956457514565,"output_mean_power":0.0003584826501301184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2949024312838274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011569002203634688,"output_power":0.00010850208719262526,"reference_active":true,"delta_db":-0.2785781125727493},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.480008250335501e-05,"output_power":1.3675641547843833e-05,"reference_active":true,"delta_db":-0.34316427401199434},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.80314470487776e-07,"output_power":5.992128594283392e-07,"reference_active":true,"delta_db":0.13917724891465347}],"active_band_count":3,"active_band_mae_db":0.253639878499799}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":2.581530634261673,"rms_envelope_nrmse":0.2771902196140238,"envelope_correlation":0.9492337208253078,"rms_gain_error_db":-0.9180576711775519,"reference_mean_power":8.262304385157993e-05,"output_mean_power":6.687987047230659e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.222558052994835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.126031645648209e-05,"output_power":1.7344452747170234e-05,"reference_active":true,"delta_db":-0.8840912312244742},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.562486047803936e-06,"output_power":7.364664665931816e-06,"reference_active":true,"delta_db":-1.134178375330371},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5273847358800203e-06,"output_power":2.1282492402055386e-06,"reference_active":true,"delta_db":-2.1943034422653493}],"active_band_count":3,"active_band_mae_db":1.4041910162733984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":1.7878960097485423,"rms_envelope_nrmse":0.17686534414447103,"envelope_correlation":0.9825204154291999,"rms_gain_error_db":-0.918059507473561,"reference_mean_power":8.303850395551146e-05,"output_mean_power":6.721613948942609e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.7676217824640817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1974657580034156e-05,"output_power":1.7551009080113362e-05,"reference_active":true,"delta_db":-0.9762002550152657},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.865612603723309e-06,"output_power":4.2418093383788435e-06,"reference_active":true,"delta_db":-1.407622326010025},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8280548950663716e-07,"output_power":1.6328182946057523e-07,"reference_active":true,"delta_db":-2.3854997754350666}],"active_band_count":3,"active_band_mae_db":1.589774118820119}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0830936431884766,"predicted_mos":3.0830936431884766,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"9ee5d559a0021b3856ad9b9bc33b52a03ab7c1dacfdd0927d341d5ad86d178e7","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.793327725999916},"started_at":"2026-10-13T05:11:22.192862+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002762056999927154,"native_input_samples":144000,"native_input_sha256":"aefb14bdcf832d1ad6e1871c426a954eba41c8f3c45424345324561bc347a2fe","output_sha256":"3b715d06f9335a07e2b098a6ce713651c8310f2bf0c9bfbe51394666d941b488","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0838000269995973,"synthesis_s":0.13101107900001807,"total_s":1.2148111059996154,"rtf":0.20246851766660257,"output_peak":0.5947518348693848,"output_rms":0.08796658034619159,"finished_at":"2026-10-13T05:11:23.418689+00:00","analysis_stages":{"external_f0_s":0.15362043999994057},"evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"3b715d06f9335a07e2b098a6ce713651c8310f2bf0c9bfbe51394666d941b488","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.11233930817637,"output_peak":0.5910095683528498},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.143097911139309,"spectral_convergence":0.1692678807952017,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.517985036045575,"spectral_convergence":0.16009487374767484,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.728703457005938,"spectral_convergence":0.17314557772931277,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.002859239861429717,"delta_db":-0.08956809342209728,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.0035222282705630577,"delta_db":-0.22839669356328338,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0009068660852758212,"delta_db":-0.3945839020164851,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.0003045339719483607,"delta_db":-0.5668451580578482,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.00011087542385295756,"delta_db":-0.7783021387107546,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":4.960203470922802e-05,"delta_db":-1.1325580325397806,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5317090030517082,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.22646322611669098,"pitch":null,"output_rms_dbfs":-21.10510854547575,"output_peak":0.5910095683528498,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.796595468063607},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9488034819973873,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09037904209327785,"rms_dbfs":-20.87864531935906,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.554125087496356,"mean":1.0741869608561199e-05,"zero_energy":false},"output":{"rms":0.088053084260457,"rms_dbfs":-21.105108545475744,"peak":0.5910095683528498,"peak_dbfs":-4.568109758036312,"crest_factor_db":16.536998787439437,"mean":5.676719922428044e-05,"zero_energy":false},"rms_difference_db":-0.22646322611668523},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":1.5330836100834303e-05,"output_rms_dbfs":-48.144341592958625,"output_zero_energy":false,"output_peak":0.10778461169035133,"output_energy_fraction":0.0002900067571173521,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":184,"output_estimated_voiced_frames":188,"both_estimated_voiced_frames":179,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9728260869565217,"absolute_difference_cents_median":1.4248990147047038,"absolute_difference_cents_p95":6.46596602959233,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023582673937264833,"log_rms_mae_db":1.044566907251596,"rms_envelope_nrmse":0.07262234207497849,"envelope_correlation":0.9947203494515287,"rms_gain_error_db":-0.22646245620056193,"reference_mean_power":0.008175179133340451,"output_mean_power":0.007759809006295493},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.411874313885781e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017924334098852407,"output_power":0.0016893115634236677,"reference_active":true,"delta_db":-0.2573327546679804},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00033647515979494424,"output_power":0.0002850557432128421,"reference_active":true,"delta_db":-0.7202321259231602},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.450579873055231e-05,"output_power":3.35421096097303e-05,"reference_active":true,"delta_db":-2.1085233439270095}],"active_band_count":3,"active_band_mae_db":1.0286960748393834}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002175646104262597,"log_rms_mae_db":0.6073119901218752,"rms_envelope_nrmse":0.0448710523203916,"envelope_correlation":0.9982680299330731,"rms_gain_error_db":-0.22644587308673125,"reference_mean_power":0.008216199965234223,"output_mean_power":0.007798775402345566},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0004980946791747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016116003385924513,"output_power":0.001534586837055728,"reference_active":true,"delta_db":-0.2126588124782745},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017199808112094875,"output_power":0.0001483023824754497,"reference_active":true,"delta_db":-0.6437547375401041},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542016550968783e-06,"output_power":1.9937607930540134e-06,"reference_active":true,"delta_db":-2.495775347859211}],"active_band_count":3,"active_band_mae_db":1.11739629929253}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019245421130728438,"log_rms_mae_db":1.3638546776806242,"rms_envelope_nrmse":0.10980250467101701,"envelope_correlation":0.9903120800501677,"rms_gain_error_db":-0.3345262456774301,"reference_mean_power":0.0024199693169718968,"output_mean_power":0.0022405633874550733},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.323542749940927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006913978889533925,"output_power":0.0006383334757569098,"reference_active":true,"delta_db":-0.34680428692340615},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018527392767596855,"output_power":0.00015060255891823679,"reference_active":true,"delta_db":-0.8998195723492384},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604035138941518e-05,"output_power":1.4248324156138504e-05,"reference_active":true,"delta_db":-2.54544627576081}],"active_band_count":3,"active_band_mae_db":1.2640233783444847}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017723694424605393,"log_rms_mae_db":0.7797372197844556,"rms_envelope_nrmse":0.06954342838288066,"envelope_correlation":0.9965099404800384,"rms_gain_error_db":-0.334534443323808,"reference_mean_power":0.0024321306730885443,"output_mean_power":0.002251818903389303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1677626980235007e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000654046958746599,"output_power":0.0006062457993794258,"reference_active":true,"delta_db":-0.329601881408296},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.291953991407944e-05,"output_power":7.599462523037953e-05,"reference_active":true,"delta_db":-0.8732417314322517},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.049791938534598e-06,"output_power":1.010580879388126e-06,"reference_active":true,"delta_db":-3.071387038490423}],"active_band_count":3,"active_band_mae_db":1.424743550443657}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":1.5225792477073998,"rms_envelope_nrmse":0.15185082091124064,"envelope_correlation":0.9834517412813424,"rms_gain_error_db":-0.5799502809401437,"reference_mean_power":0.0003849600012580848,"output_mean_power":0.0003368376113869574},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.553182623332075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001221299679169956,"output_power":9.508464313367602e-05,"reference_active":true,"delta_db":-1.087118621506887},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7738426055120482e-05,"output_power":2.585121203445844e-05,"reference_active":true,"delta_db":-0.30600904714460997},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.9827801273792e-06,"output_power":5.4928891073754515e-06,"reference_active":true,"delta_db":-1.0422753586324272}],"active_band_count":3,"active_band_mae_db":0.8118010090946415}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":1.005460695801446,"rms_envelope_nrmse":0.10396196440299159,"envelope_correlation":0.9934906741236862,"rms_gain_error_db":-0.5799497657224335,"reference_mean_power":0.0003868956457514565,"output_mean_power":0.0003385313284638424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2949024312838274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011569002203634688,"output_power":9.172417252710426e-05,"reference_active":true,"delta_db":-1.008121012596713},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.480008250335501e-05,"output_power":1.3690455269056435e-05,"reference_active":true,"delta_db":-0.3384624577094922},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.80314470487776e-07,"output_power":3.745555342597256e-07,"reference_active":true,"delta_db":-1.9014718170245226}],"active_band_count":3,"active_band_mae_db":1.0826850957769094}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":1.696587748981903,"rms_envelope_nrmse":0.14812767154021475,"envelope_correlation":0.9882018406053193,"rms_gain_error_db":-0.6930972797141238,"reference_mean_power":8.262304385157993e-05,"output_mean_power":7.043547739132451e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.222558052994835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.126031645648209e-05,"output_power":1.833785570272896e-05,"reference_active":true,"delta_db":-0.6422117361850892},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.562486047803936e-06,"output_power":7.985445701033339e-06,"reference_active":true,"delta_db":-0.7827165356491954},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5273847358800203e-06,"output_power":1.8490583274823562e-06,"reference_active":true,"delta_db":-2.8050222030079732}],"active_band_count":3,"active_band_mae_db":1.4099834916140859}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":1.206142821574973,"rms_envelope_nrmse":0.1052357035266629,"envelope_correlation":0.9958327535760885,"rms_gain_error_db":-0.693096921017758,"reference_mean_power":8.303850395551146e-05,"output_mean_power":7.0789659620244e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.7676217824640817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1974657580034156e-05,"output_power":1.8825088353954554e-05,"reference_active":true,"delta_db":-0.6718509320564194},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.865612603723309e-06,"output_power":4.515158125845515e-06,"reference_active":true,"delta_db":-1.1364041208221418},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8280548950663716e-07,"output_power":1.380608056723648e-07,"reference_active":true,"delta_db":-3.1141743180626253}],"active_band_count":3,"active_band_mae_db":1.6408097903137289}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0004","method":"magphase","status":"measured","mos":{"status":"measured","value":3.356645107269287,"predicted_mos":3.356645107269287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":290099,"input_sample_rate":48000,"input_seconds":6.043729166666667,"input_float32_sha256":"10b6e9e80e0c15583a76547c889cd0d381a80645e87bc5d6260ea1d2b9e97439","internal_samples":96700,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.274779834000583},"started_at":"2026-10-13T05:18:01.946482+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009638101000746246,"native_input_samples":288000,"native_input_sha256":"109a10d036aec5f82dcb7e04035ad48f5e9331c0b47f7c1f982775d6286da7a7","output_sha256":"54a7ef392c0aaf2a703b671cd32d184d5aa3cb033b34d234e9c9a46098c27e6f","output_samples":290099,"length_delta_samples":2099,"analysis_s":2.816230948001248,"synthesis_s":0.8706727699991461,"total_s":3.6869037180003943,"rtf":0.6144839530000658,"output_peak":0.6202622996525893,"output_rms":0.08812549601113219,"finished_at":"2026-10-13T05:18:05.656427+00:00","evaluation_input":{"reference":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"54a7ef392c0aaf2a703b671cd32d184d5aa3cb033b34d234e9c9a46098c27e6f","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96700,"length_delta_samples":700,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.097029976257694,"output_peak":0.6205141340849688},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9927611168562565,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.453946952509343,"spectral_convergence":0.8719325089484521,"spectral_convergence_status":"ok","amplitude_floor":2.016382236479878e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.563136055106336,"spectral_convergence":0.6062173508606071,"spectral_convergence_status":"ok","amplitude_floor":1.164576113083389e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.381115916400889,"spectral_convergence":0.568620670908138,"spectral_convergence_status":"ok","amplitude_floor":9.482830334434221e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0029188205763026483,"output_power":0.00272832880290473,"delta_db":-0.293106912850668,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003712420548895462,"output_power":0.0035500372645135263,"delta_db":-0.19424256061886966,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0009931195946472759,"output_power":0.0009830998284696489,"delta_db":-0.04403930452527362,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003469926396566426,"output_power":0.0003540211879375888,"delta_db":0.08708992330821266,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013263724937728663,"output_power":0.00014761184067583498,"delta_db":0.4645568886409013,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.438064081917431e-05,"output_power":5.987829796670368e-05,"delta_db":-0.31485848085922785,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23298234513385882,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1876340231645181,"pitch":null,"output_rms_dbfs":-21.06627934252358,"output_peak":0.6205141340849688,"reference_rms_dbfs":-20.87864531935906,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.466066308005523},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.5658617226460833,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09037904209327785,"rms_dbfs":-20.87864531935906,"peak":0.607818603515625,"peak_dbfs":-4.324520231862702,"crest_factor_db":16.554125087496356,"mean":1.0741869608561199e-05,"zero_energy":false},"output":{"rms":0.08844759591118366,"rms_dbfs":-21.066279342523572,"peak":0.6205141340849688,"peak_dbfs":-4.144966433661244,"crest_factor_db":16.921312908862333,"mean":0.0007457852224749901,"zero_energy":false},"rms_difference_db":-0.18763402316451305},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.5614251003135e-06,"threshold_dbfs":-52.54813907357131,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":1.8976366184440188e-06,"output_mean_power":2.8212017023990003e-05,"output_rms_dbfs":-45.49565862760844,"output_zero_energy":false,"output_peak":0.071249000365819,"output_energy_fraction":0.0005289242674729998,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":184,"output_estimated_voiced_frames":174,"both_estimated_voiced_frames":141,"reference_voiced_output_unvoiced_frames":43,"reference_unvoiced_output_voiced_frames":33,"paired_voiced_fraction":0.7663043478260869,"absolute_difference_cents_median":9.997226458855602,"absolute_difference_cents_p95":42.971994449108294,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023582673937264833,"log_rms_mae_db":5.097291571121939,"rms_envelope_nrmse":0.47527452508543977,"envelope_correlation":0.7453670614668086,"rms_gain_error_db":-0.18794559770126196,"reference_mean_power":0.008175179133340451,"output_mean_power":0.007828935550344968},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.411874313885781e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017924334098852407,"output_power":0.001788988052771007,"reference_active":true,"delta_db":-0.008355900360170265},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00033647515979494424,"output_power":0.00031892637969143907,"reference_active":true,"delta_db":-0.23262565065599483},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.450579873055231e-05,"output_power":5.730253582413874e-05,"reference_active":true,"delta_db":0.21731133142177206}],"active_band_count":3,"active_band_mae_db":0.15276429414597906}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002175646104262597,"log_rms_mae_db":3.706809430722745,"rms_envelope_nrmse":0.35850513404361184,"envelope_correlation":0.8363659726267852,"rms_gain_error_db":-0.22896440112693092,"reference_mean_power":0.008216199965234223,"output_mean_power":0.007794254106015045},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0004980946791747e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016116003385924513,"output_power":0.001616850078248718,"reference_active":true,"delta_db":0.014124020376808257},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017199808112094875,"output_power":0.00015444550406525513,"reference_active":true,"delta_db":-0.4674833133682304},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.542016550968783e-06,"output_power":3.838307042328953e-06,"reference_active":true,"delta_db":0.3488912662217223}],"active_band_count":3,"active_band_mae_db":0.276832866655587}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019245421130728438,"log_rms_mae_db":6.483232067040493,"rms_envelope_nrmse":0.5954093262320055,"envelope_correlation":0.6863617699566577,"rms_gain_error_db":-0.12857316723954187,"reference_mean_power":0.0024199693169718968,"output_mean_power":0.002349376080689721},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.323542749940927e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006913978889533925,"output_power":0.0006913978563186193,"reference_active":true,"delta_db":-2.049919806504336e-07},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018527392767596855,"output_power":0.00017252424399417108,"reference_active":true,"delta_db":-0.30964175354579077},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5604035138941518e-05,"output_power":2.358187746803292e-05,"reference_active":true,"delta_db":-0.3573003611643806}],"active_band_count":3,"active_band_mae_db":0.222314106567384}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017723694424605393,"log_rms_mae_db":5.073655452114218,"rms_envelope_nrmse":0.4536052799787479,"envelope_correlation":0.7953461005584799,"rms_gain_error_db":-0.18636635330470055,"reference_mean_power":0.0024321306730885443,"output_mean_power":0.0023299696801358763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1677626980235007e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000654046958746599,"output_power":0.0006597149923977971,"reference_active":true,"delta_db":0.03747423205665605},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.291953991407944e-05,"output_power":8.106508128826758e-05,"reference_active":true,"delta_db":-0.5927322822065699},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.049791938534598e-06,"output_power":1.8836729509417274e-06,"reference_active":true,"delta_db":-0.36704279330276146}],"active_band_count":3,"active_band_mae_db":0.33241643585532915}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.001235221787968e-05,"log_rms_mae_db":6.7120380405698645,"rms_envelope_nrmse":0.5757885412380046,"envelope_correlation":0.7346082907203355,"rms_gain_error_db":0.12460450452995475,"reference_mean_power":0.0003849600012580848,"output_mean_power":0.000396164957364633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.553182623332075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001221299679169956,"output_power":0.00012846182721677367,"reference_active":true,"delta_db":0.21951852032890007},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7738426055120482e-05,"output_power":2.793607914031562e-05,"reference_active":true,"delta_db":0.030836378181429036},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.9827801273792e-06,"output_power":7.060940322464255e-06,"reference_active":true,"delta_db":0.048341736837484744}],"active_band_count":3,"active_band_mae_db":0.09956554511593796}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.393121399816876e-05,"log_rms_mae_db":5.305282911794895,"rms_envelope_nrmse":0.43258561172267285,"envelope_correlation":0.8325875220514902,"rms_gain_error_db":0.11597086554952944,"reference_mean_power":0.0003868956457514565,"output_mean_power":0.00039736620493267035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2949024312838274e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011569002203634688,"output_power":0.00012249588525317507,"reference_active":true,"delta_db":0.2482559680692736},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.480008250335501e-05,"output_power":1.3059967346276075e-05,"reference_active":true,"delta_db":-0.5432204530729575},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.80314470487776e-07,"output_power":5.346742940200763e-07,"reference_active":true,"delta_db":-0.3557409542584139}],"active_band_count":3,"active_band_mae_db":0.38240579180021506}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.31472680502341e-05,"log_rms_mae_db":6.632643081922546,"rms_envelope_nrmse":0.6839693155636642,"envelope_correlation":0.7065419903582081,"rms_gain_error_db":0.568550403436092,"reference_mean_power":8.262304385157993e-05,"output_mean_power":9.417946754679958e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.222558052994835e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.126031645648209e-05,"output_power":2.489625230415121e-05,"reference_active":true,"delta_db":0.6856425192788651},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.562486047803936e-06,"output_power":1.2027923136575975e-05,"reference_active":true,"delta_db":0.9961982961278081},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.5273847358800203e-06,"output_power":4.359590379271157e-06,"reference_active":true,"delta_db":0.9199285426594281}],"active_band_count":3,"active_band_mae_db":0.8672564526887004}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.6400870566820567e-05,"log_rms_mae_db":5.38643297893197,"rms_envelope_nrmse":0.4816371831677576,"envelope_correlation":0.8433464955639836,"rms_gain_error_db":0.5640656314948086,"reference_mean_power":8.303850395551146e-05,"output_mean_power":9.455534408448099e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.7676217824640817e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.1974657580034156e-05,"output_power":2.6328404207276377e-05,"reference_active":true,"delta_db":0.7850242044351157},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.865612603723309e-06,"output_power":7.090946105214804e-06,"reference_active":true,"delta_db":0.8239080805632917},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8280548950663716e-07,"output_power":3.546522330667072e-07,"reference_active":true,"delta_db":0.9831486347735028}],"active_band_count":3,"active_band_mae_db":0.8640269732573035}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"hnm","status":"measured","mos":{"status":"measured","value":2.410280227661133,"predicted_mos":2.410280227661133,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88480,"input_sample_rate":16000,"input_seconds":5.53,"input_float32_sha256":"213bbcc0607f050c499f661ab0facee94566c2fba9fee65c45ef679c78b80708","internal_samples":88480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7181257080010255},"started_at":"2026-10-13T04:40:08.992684+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5132000044104643e-05,"native_input_samples":88480,"native_input_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","output_sha256":"ef854bb25f598e50ec3b8118818c5455501f7a66488f15a3326bbd78a950170f","output_samples":88480,"length_delta_samples":0,"analysis_s":1.1005111819995363,"synthesis_s":0.4826475039999423,"total_s":1.5831586859994786,"rtf":0.2862854766726001,"output_peak":0.6088429249160301,"output_rms":0.07555924304298293,"finished_at":"2026-10-13T04:40:10.583131+00:00","evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"ef854bb25f598e50ec3b8118818c5455501f7a66488f15a3326bbd78a950170f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.658291421188941,"spectral_convergence":0.2821392019963134,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.858046362125468,"spectral_convergence":0.25042692022002006,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.675725453830279,"spectral_convergence":0.289394881208201,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.002211242690830804,"delta_db":0.02021944095904135,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.0021770804230613495,"delta_db":-0.04804498082276489,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0010775614236301083,"delta_db":-0.2985495692384758,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.00017145660053140163,"delta_db":-0.6210873910070757,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":5.45891187740117e-05,"delta_db":-0.2647414740749917,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.7268952400889813e-05,"delta_db":0.16356254447247506,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2360342334291374,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09017399952427668,"pitch":null,"output_rms_dbfs":-22.434248030330238,"output_peak":0.6088429249160301,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.064021079048229},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.658291421188941,"spectral_convergence":0.2821392019963134,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.858046362125468,"spectral_convergence":0.25042692022002006,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.675725453830279,"spectral_convergence":0.289394881208201,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.002211242690830804,"delta_db":0.02021944095904135,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.0021770804230613495,"delta_db":-0.04804498082276489,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0010775614236301083,"delta_db":-0.2985495692384758,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.00017145660053140163,"delta_db":-0.6210873910070757,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":5.45891187740117e-05,"delta_db":-0.2647414740749917,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.7268952400889813e-05,"delta_db":0.16356254447247506,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2360342334291374,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09017399952427668,"pitch":null,"output_rms_dbfs":-22.434248030330238,"output_peak":0.6088429249160301,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.064021079048229},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.8851855943566838,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07634775979506618,"rms_dbfs":-22.34407403080596,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.31326298831702,"mean":7.506261897992484e-06,"zero_energy":false},"output":{"rms":0.07555924304298293,"rms_dbfs":-22.434248030330238,"peak":0.6088429249160301,"peak_dbfs":-4.309894726687997,"crest_factor_db":18.12435330364224,"mean":8.532757581698256e-05,"zero_energy":false},"rms_difference_db":-0.09017399952427851},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":1.378517517658012e-05,"output_rms_dbfs":-48.6058771067981,"output_zero_energy":false,"output_peak":0.07765114533732814,"output_energy_fraction":0.0002969072987123101,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":164,"both_estimated_voiced_frames":130,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.9848484848484849,"absolute_difference_cents_median":2.1711234658102843,"absolute_difference_cents_p95":8.884825063903357,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00024519594682554,"log_rms_mae_db":1.0706863137923064,"rms_envelope_nrmse":0.07962094274336759,"envelope_correlation":0.9936226188784953,"rms_gain_error_db":-0.09016800685789984,"reference_mean_power":0.00583423429533283,"output_mean_power":0.005714353000735304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.472320009859619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016537460501756134,"output_power":0.0016619522028874405,"reference_active":true,"delta_db":0.021497095604476164},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003778384775464376,"output_power":0.00036229623591452285,"reference_active":true,"delta_db":-0.18242360691110973},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8517105494706513e-05,"output_power":5.2321924219603046e-05,"reference_active":true,"delta_db":0.32788824276137707}],"active_band_count":3,"active_band_mae_db":0.17726964842565432}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020002070942310007,"log_rms_mae_db":0.5573862082609471,"rms_envelope_nrmse":0.040958105476694984,"envelope_correlation":0.9981731086657407,"rms_gain_error_db":-0.09015534736530861,"reference_mean_power":0.005865946925483087,"output_mean_power":0.005745430750412889},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.084203911038529e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015941324207372327,"output_power":0.0015979610292306981,"reference_active":true,"delta_db":0.01041789261790697},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018248750335661006,"output_power":0.0001781727890873821,"reference_active":true,"delta_db":-0.10391751140698488},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.552526527817181e-06,"output_power":4.71289295174634e-06,"reference_active":true,"delta_db":0.1503509021171379}],"active_band_count":3,"active_band_mae_db":0.08822876871400992}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00018690652054173922,"log_rms_mae_db":1.8770849056225845,"rms_envelope_nrmse":0.11761584417212033,"envelope_correlation":0.9893653981577238,"rms_gain_error_db":-0.17099888091662963,"reference_mean_power":0.0020929267743018004,"output_mean_power":0.0020121212411442608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":8.903808595053404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005237826202169962,"output_power":0.0005128645116355858,"reference_active":true,"delta_db":-0.09148435374986594},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020282121471265822,"output_power":0.00020225313354679534,"reference_active":true,"delta_db":-0.012181204680593739},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.57818052122161e-05,"output_power":2.464239080236646e-05,"reference_active":true,"delta_db":-0.19630482134900787}],"active_band_count":3,"active_band_mae_db":0.09999012659315586}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00016950242057071555,"log_rms_mae_db":1.0925292895352712,"rms_envelope_nrmse":0.07606222391840073,"envelope_correlation":0.9952600019550498,"rms_gain_error_db":-0.17099330932235476,"reference_mean_power":0.002104315784196243,"output_mean_power":0.0020230731296742416},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":7.681995008528509e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000538193996495516,"output_power":0.0005312520892901079,"reference_active":true,"delta_db":-0.05638197899530596},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010305287620923108,"output_power":0.00010470575517880833,"reference_active":true,"delta_db":0.06910435985049729},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.715218896542221e-06,"output_power":2.2942526445900783e-06,"reference_active":true,"delta_db":-0.731636064639024}],"active_band_count":3,"active_band_mae_db":0.2857074678282758}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":2.3764792293530155,"rms_envelope_nrmse":0.194970002219374,"envelope_correlation":0.972461034792148,"rms_gain_error_db":-0.527446240944424,"reference_mean_power":0.0002131533037612313,"output_mean_power":0.00018877628894631992},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.6207032137826844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010202948507054384,"output_power":9.016404424571865e-05,"reference_active":true,"delta_db":-0.5369231137755942},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692530062160889e-05,"output_power":3.165807817705074e-05,"reference_active":true,"delta_db":-0.6683949252449993},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.752197568216406e-06,"output_power":4.642065636960818e-06,"reference_active":true,"delta_db":-1.6273386427583894}],"active_band_count":3,"active_band_mae_db":0.9442188939263275}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":1.5377584096252679,"rms_envelope_nrmse":0.1248137715155868,"envelope_correlation":0.9891540303756016,"rms_gain_error_db":-0.5271740173565497,"reference_mean_power":0.00021425147153668843,"output_mean_power":0.00018976076029070864},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.4064708995422746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010396272320640845,"output_power":9.201382785905808e-05,"reference_active":true,"delta_db":-0.5302454879216274},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.972311051816857e-05,"output_power":1.7094922518861522e-05,"reference_active":true,"delta_db":-0.6210827155223524},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.307963255185345e-07,"output_power":3.610007449694426e-07,"reference_active":true,"delta_db":-2.4238105670827017}],"active_band_count":3,"active_band_mae_db":1.191712923508894}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":2.708483647691662,"rms_envelope_nrmse":0.24828691052713903,"envelope_correlation":0.955146213618468,"rms_gain_error_db":-0.1559339806279864,"reference_mean_power":2.6417005173666296e-05,"output_mean_power":2.5485325405560574e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.6945603106309378e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5028414095885614e-05,"output_power":1.3595823989593751e-05,"reference_active":true,"delta_db":-0.43507619584829293},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.702646590810387e-06,"output_power":5.68802597285484e-06,"reference_active":true,"delta_db":-0.7128475021745386},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.011540321330317e-07,"output_power":6.052899353308888e-07,"reference_active":true,"delta_db":-1.7283557766546873}],"active_band_count":3,"active_band_mae_db":0.9587598248925063}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":1.8367093997288264,"rms_envelope_nrmse":0.16267458614452907,"envelope_correlation":0.9787275151396087,"rms_gain_error_db":-0.15598672108194653,"reference_mean_power":2.6545935258029172e-05,"output_mean_power":2.5609397359306695e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.4345030346700565e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.623830647966854e-05,"output_power":1.4189707309705718e-05,"reference_active":true,"delta_db":-0.5856729658186931},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.69303691046191e-06,"output_power":3.1760434470762697e-06,"reference_active":true,"delta_db":-0.6549721349605953},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.89764420950003e-08,"output_power":5.923576635293224e-08,"reference_active":true,"delta_db":-1.2491355334364733}],"active_band_count":3,"active_band_mae_db":0.8299268780719204}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0728940963745117,"predicted_mos":3.0728940963745117,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88480,"input_sample_rate":16000,"input_seconds":5.53,"input_float32_sha256":"219530ba1039bddfa4029aa198e11e980679abc9786392da411a6f74a9747862","internal_samples":88480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.550608278001164},"started_at":"2026-10-13T04:43:22.652539+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2836999960418325e-05,"native_input_samples":88480,"native_input_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","output_sha256":"796a28f7b8e2755b2e3e5b2453a517daa5b6c7a4032224dc3f3e407517d16d88","output_samples":88480,"length_delta_samples":0,"analysis_s":1.8712293290000162,"synthesis_s":0.11257323599966185,"total_s":1.983802564999678,"rtf":0.35873464104876635,"output_peak":0.7719099371953577,"output_rms":0.0798776011829014,"finished_at":"2026-10-13T04:43:24.644426+00:00","evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"796a28f7b8e2755b2e3e5b2453a517daa5b6c7a4032224dc3f3e407517d16d88","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.743863339369689,"spectral_convergence":0.8182027336320188,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6225727730072395,"spectral_convergence":0.29751304514919935,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.576547913523784,"spectral_convergence":0.272183712009663,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.0023476978004870193,"delta_db":0.2802772679273119,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.0024538185811573427,"delta_db":0.4716348016566441,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0012857878866515304,"delta_db":0.4687233766918403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.00021081586180944726,"delta_db":0.27640336312437475,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":6.455705023082819e-05,"delta_db":0.46363447189416673,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.7753990398484288e-05,"delta_db":0.2838624267903276,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.37408928468077757,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.39257431134611287,"pitch":null,"output_rms_dbfs":-21.951499719459846,"output_peak":0.7719099371953577,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.314328008633571},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.743863339369689,"spectral_convergence":0.8182027336320188,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6225727730072395,"spectral_convergence":0.29751304514919935,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.576547913523784,"spectral_convergence":0.272183712009663,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.0023476978004870193,"delta_db":0.2802772679273119,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.0024538185811573427,"delta_db":0.4716348016566441,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0012857878866515304,"delta_db":0.4687233766918403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.00021081586180944726,"delta_db":0.27640336312437475,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":6.455705023082819e-05,"delta_db":0.46363447189416673,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.7753990398484288e-05,"delta_db":0.2838624267903276,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.37408928468077757,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.39257431134611287,"pitch":null,"output_rms_dbfs":-21.951499719459846,"output_peak":0.7719099371953577,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.314328008633571},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.9138315364894387,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07634775979506618,"rms_dbfs":-22.34407403080596,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.31326298831702,"mean":7.506261897992484e-06,"zero_energy":false},"output":{"rms":0.0798776011829014,"rms_dbfs":-21.951499719459846,"peak":0.7719099371953577,"peak_dbfs":-2.2486673627359925,"crest_factor_db":19.702832356723857,"mean":-1.1392308319469726e-07,"zero_energy":false},"rms_difference_db":0.3925743113461131},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":1.4600307315157176e-05,"output_rms_dbfs":-48.35638002854444,"output_zero_energy":false,"output_peak":0.05130678808850675,"output_energy_fraction":0.0002813816360803135,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":117,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.8863636363636364,"absolute_difference_cents_median":6.167616105712052,"absolute_difference_cents_p95":28.10059386475956,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00024519594682554,"log_rms_mae_db":1.9276043214365828,"rms_envelope_nrmse":0.14951181876430444,"envelope_correlation":0.9803575600338952,"rms_gain_error_db":0.3925840726620886,"reference_mean_power":0.00583423429533283,"output_mean_power":0.006386196436488751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.472320009859619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016537460501756134,"output_power":0.001759175525451692,"reference_active":true,"delta_db":0.26840354344833894},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003778384775464376,"output_power":0.00035725498072876897,"reference_active":true,"delta_db":-0.2432789034330228},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8517105494706513e-05,"output_power":3.1269310676666564e-05,"reference_active":true,"delta_db":-1.9077657573801987}],"active_band_count":3,"active_band_mae_db":0.8064827347538536}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020002070942310007,"log_rms_mae_db":1.2248420811936724,"rms_envelope_nrmse":0.08846566536397618,"envelope_correlation":0.9937224322412491,"rms_gain_error_db":0.392602538873304,"reference_mean_power":0.005865946925483087,"output_mean_power":0.0064209366200968865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.084203911038529e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015941324207372327,"output_power":0.0016865724348692182,"reference_active":true,"delta_db":0.24480603639714385},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018248750335661006,"output_power":0.00017816714214155603,"reference_active":true,"delta_db":-0.1040551573566608},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.552526527817181e-06,"output_power":2.298001514467367e-06,"reference_active":true,"delta_db":-2.9690217456328982}],"active_band_count":3,"active_band_mae_db":1.1059609797955676}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00018690652054173922,"log_rms_mae_db":2.4879264064834263,"rms_envelope_nrmse":0.1853190104851667,"envelope_correlation":0.9782573937020049,"rms_gain_error_db":0.5615201015343522,"reference_mean_power":0.0020929267743018004,"output_mean_power":0.0023818041300946433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":8.903808595053404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005237826202169962,"output_power":0.0005785268709570584,"reference_active":true,"delta_db":0.4317245164459028},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020282121471265822,"output_power":0.0002051798240353772,"reference_active":true,"delta_db":0.0502127361144126},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.57818052122161e-05,"output_power":1.3406146260236671e-05,"reference_active":true,"delta_db":-2.8400936967010644}],"active_band_count":3,"active_band_mae_db":1.1073436497537932}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00016950242057071555,"log_rms_mae_db":1.7386900252543467,"rms_envelope_nrmse":0.12081625249070957,"envelope_correlation":0.9918425915585536,"rms_gain_error_db":0.5615499850495356,"reference_mean_power":0.002104315784196243,"output_mean_power":0.002394781592547703},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":7.681995008528509e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000538193996495516,"output_power":0.0005854222363901679,"reference_active":true,"delta_db":0.3653036549918238},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010305287620923108,"output_power":0.000105050178735654,"reference_active":true,"delta_db":0.08336678206850254},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.715218896542221e-06,"output_power":1.2544110166783785e-06,"reference_active":true,"delta_db":-3.353649879689024}],"active_band_count":3,"active_band_mae_db":1.2674401055831168}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":2.1872106503257243,"rms_envelope_nrmse":0.19671424512817626,"envelope_correlation":0.9728884785113266,"rms_gain_error_db":0.32243053282501577,"reference_mean_power":0.0002131533037612313,"output_mean_power":0.00022958056568477868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.6207032137826844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010202948507054384,"output_power":0.00011184275235245146,"reference_active":true,"delta_db":0.3988215131048807},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692530062160889e-05,"output_power":3.483097836741738e-05,"reference_active":true,"delta_db":-0.25358366211464806},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.752197568216406e-06,"output_power":4.218379755727451e-06,"reference_active":true,"delta_db":-2.042994670687124}],"active_band_count":3,"active_band_mae_db":0.8984666153022175}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":1.3689331094622474,"rms_envelope_nrmse":0.10621576143290425,"envelope_correlation":0.9918860429458592,"rms_gain_error_db":0.32277578053718287,"reference_mean_power":0.00021425147153668843,"output_mean_power":0.0002307817124096822},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.4064708995422746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010396272320640845,"output_power":0.00011185654729336779,"reference_active":true,"delta_db":0.3178376279640329},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.972311051816857e-05,"output_power":1.821103752685888e-05,"reference_active":true,"delta_db":-0.34640718983231505},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.307963255185345e-07,"output_power":3.311756779838128e-07,"reference_active":true,"delta_db":-2.798307207597977}],"active_band_count":3,"active_band_mae_db":1.154184008464775}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":2.706868117434355,"rms_envelope_nrmse":0.28259936245308587,"envelope_correlation":0.9493890515357721,"rms_gain_error_db":0.5767201644191631,"reference_mean_power":2.6417005173666296e-05,"output_mean_power":3.016863062453408e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.6945603106309378e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5028414095885614e-05,"output_power":1.833898237447644e-05,"reference_active":true,"delta_db":0.864620799189181},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.702646590810387e-06,"output_power":7.329519578117534e-06,"reference_active":true,"delta_db":0.388291882035541},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.011540321330317e-07,"output_power":6.927390745267715e-07,"reference_active":true,"delta_db":-1.1422934518421919}],"active_band_count":3,"active_band_mae_db":0.798402044355638}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":1.8260586788027515,"rms_envelope_nrmse":0.17674554733147288,"envelope_correlation":0.9801415997992434,"rms_gain_error_db":0.5773455715932176,"reference_mean_power":2.6545935258029172e-05,"output_mean_power":3.032023674654236e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.4345030346700565e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.623830647966854e-05,"output_power":1.9655648015553544e-05,"reference_active":true,"delta_db":0.8294663242879423},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.69303691046191e-06,"output_power":3.898584062932499e-06,"reference_active":true,"delta_db":0.2352325482408263},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.89764420950003e-08,"output_power":4.6369790370695105e-08,"reference_active":true,"delta_db":-2.3126243225060215}],"active_band_count":3,"active_band_mae_db":1.12577439834493}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"facodec","status":"measured","mos":{"status":"measured","value":4.628403663635254,"predicted_mos":4.628403663635254,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88600,"input_sample_rate":16000,"input_seconds":5.5375,"input_float32_sha256":"6ee96eaba2d09ae05737ce5e699997598714d7e1f01484956479321477d8126b","internal_samples":88600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.343443992001994},"started_at":"2026-10-13T06:59:20.288475+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.950800030201208e-05,"native_input_samples":88480,"native_input_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","output_sha256":"f5cfe3533960f26c4e8bc4a22130d51164ab870564680e83d37f7fc6f0a7eeae","output_samples":88600,"length_delta_samples":120,"analysis_s":2.7255909719988267,"synthesis_s":12.993890482999632,"total_s":15.719481454998459,"rtf":2.842582541591041,"output_peak":0.6640409231185913,"output_rms":0.07382802831958554,"finished_at":"2026-10-13T06:59:36.034086+00:00","evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"f5cfe3533960f26c4e8bc4a22130d51164ab870564680e83d37f7fc6f0a7eeae","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":88480,"output_samples":88600,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.63557459724437,"output_peak":0.6640409231185913},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986455981941309,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.592044303198344,"spectral_convergence":0.22024436519963553,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.680332648596559,"spectral_convergence":0.22279340492862956,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.834769085035928,"spectral_convergence":0.2474124788188192,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.0018785652271359156,"delta_db":-0.6878818714183986,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.002193244127893722,"delta_db":-0.01591995330035113,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0011102638488925162,"delta_db":-0.16870793249112787,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.00020336271745503343,"delta_db":0.12008388729283731,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":5.8526338490447275e-05,"delta_db":0.03771123742533124,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.4007652504737208e-05,"delta_db":-0.7454437855945424,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.29595811125376476,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.28561459672493666,"pitch":null,"output_rms_dbfs":-22.629688627530896,"output_peak":0.6640409231185913,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.702382012276945},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.894929253956958,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07634775979506618,"rms_dbfs":-22.34407403080596,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.31326298831702,"mean":7.506261897992484e-06,"zero_energy":false},"output":{"rms":0.07387807463904546,"rms_dbfs":-22.629688627530896,"peak":0.6640409231185913,"peak_dbfs":-3.5561031073103426,"crest_factor_db":19.073585520220558,"mean":2.2055641324581092e-05,"zero_energy":false},"rms_difference_db":-0.28561459672493683},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":4.39545864952737e-06,"output_rms_dbfs":-53.569958012349886,"output_zero_energy":false,"output_peak":0.09288337826728821,"output_energy_fraction":9.902773433978081e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":110,"both_estimated_voiced_frames":102,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.7727272727272727,"absolute_difference_cents_median":3.7510771821039253,"absolute_difference_cents_p95":17.684591716812008,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00024519594682554,"log_rms_mae_db":1.3222748701371805,"rms_envelope_nrmse":0.09444752851028886,"envelope_correlation":0.9917587777232306,"rms_gain_error_db":-0.2856040330546318,"reference_mean_power":0.00583423429533283,"output_mean_power":0.005462902664678266},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.472320009859619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016537460501756134,"output_power":0.0015753655540617568,"reference_active":true,"delta_db":-0.21087474692288974},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003778384775464376,"output_power":0.00036779720306555184,"reference_active":true,"delta_db":-0.11697760338214058},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8517105494706513e-05,"output_power":5.5880091908792514e-05,"reference_active":true,"delta_db":0.6136222856431622}],"active_band_count":3,"active_band_mae_db":0.3138248786493975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020002070942310007,"log_rms_mae_db":1.0254257688241601,"rms_envelope_nrmse":0.08109501513093431,"envelope_correlation":0.9934015138090273,"rms_gain_error_db":-0.2855383397992934,"reference_mean_power":0.005865946925483087,"output_mean_power":0.0054926799644189304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.084203911038529e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015941324207372327,"output_power":0.001528227421166245,"reference_active":true,"delta_db":-0.1833640633952867},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018248750335661006,"output_power":0.0001770326534538551,"reference_active":true,"delta_db":-0.13179750734496157},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.552526527817181e-06,"output_power":5.3406956994345035e-06,"reference_active":true,"delta_db":0.693453484410475}],"active_band_count":3,"active_band_mae_db":0.33620501838357447}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00018690652054173922,"log_rms_mae_db":1.7864929831519862,"rms_envelope_nrmse":0.10623111055864166,"envelope_correlation":0.9913145266131999,"rms_gain_error_db":-0.12138296125368137,"reference_mean_power":0.0020929267743018004,"output_mean_power":0.002035240506938753},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":8.903808595053404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005237826202169962,"output_power":0.0005379500698388451,"reference_active":true,"delta_db":0.11590884354734106},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020282121471265822,"output_power":0.00020000094611169732,"reference_active":true,"delta_db":-0.06081329298071051},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.57818052122161e-05,"output_power":2.6176465174444582e-05,"reference_active":true,"delta_db":0.06597676902385166}],"active_band_count":3,"active_band_mae_db":0.08089963518396774}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00016950242057071555,"log_rms_mae_db":1.313531582067584,"rms_envelope_nrmse":0.08746272010370894,"envelope_correlation":0.9935672731722868,"rms_gain_error_db":-0.12132454986126974,"reference_mean_power":0.002104315784196243,"output_mean_power":0.0020463431299514367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":7.681995008528509e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000538193996495516,"output_power":0.0005539189827884657,"reference_active":true,"delta_db":0.12507399721998025},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010305287620923108,"output_power":0.00010293616118057391,"reference_active":true,"delta_db":-0.004921494752122547},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.715218896542221e-06,"output_power":2.747619018851614e-06,"reference_active":true,"delta_db":0.05151666421994898}],"active_band_count":3,"active_band_mae_db":0.060504052064017266}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":1.8024624890213237,"rms_envelope_nrmse":0.18598279442429577,"envelope_correlation":0.9755302314654779,"rms_gain_error_db":0.16209417647462,"reference_mean_power":0.0002131533037612313,"output_mean_power":0.0002212592755132664},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.6207032137826844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010202948507054384,"output_power":0.0001005486341198032,"reference_active":true,"delta_db":-0.06349519432921938},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692530062160889e-05,"output_power":4.0045083813652674e-05,"reference_active":true,"delta_db":0.3522516723114238},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.752197568216406e-06,"output_power":6.129502471696878e-06,"reference_active":true,"delta_db":-0.4201991668260247}],"active_band_count":3,"active_band_mae_db":0.27864867782222263}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":1.391795286014302,"rms_envelope_nrmse":0.1483132035250992,"envelope_correlation":0.982480408858179,"rms_gain_error_db":0.16299718172666008,"reference_mean_power":0.00021425147153668843,"output_mean_power":0.00022244545241272557},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.4064708995422746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010396272320640845,"output_power":0.0001038087366848663,"reference_active":true,"delta_db":-0.00643741049638695},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.972311051816857e-05,"output_power":2.0365533003405693e-05,"reference_active":true,"delta_db":0.1392037256488881},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.307963255185345e-07,"output_power":6.029962827627712e-07,"reference_active":true,"delta_db":-0.1957451994286611}],"active_band_count":3,"active_band_mae_db":0.11379544519131206}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":2.378318531796698,"rms_envelope_nrmse":0.2697127542089406,"envelope_correlation":0.9493347664143393,"rms_gain_error_db":-0.8762507824651085,"reference_mean_power":2.6417005173666296e-05,"output_mean_power":2.1590291352165502e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.6945603106309378e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5028414095885614e-05,"output_power":1.1474858416835992e-05,"reference_active":true,"delta_db":-1.1716581751571638},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.702646590810387e-06,"output_power":5.712379443037853e-06,"reference_active":true,"delta_db":-0.6942927343079726},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.011540321330317e-07,"output_power":1.36884331608396e-06,"reference_active":true,"delta_db":1.8155470943853447}],"active_band_count":3,"active_band_mae_db":1.2271660012834937}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":1.9178546099472134,"rms_envelope_nrmse":0.2109601235739816,"envelope_correlation":0.968938702290297,"rms_gain_error_db":-0.8751571585970459,"reference_mean_power":2.6545935258029172e-05,"output_mean_power":2.1701128321996835e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.4345030346700565e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.623830647966854e-05,"output_power":1.2552285774268313e-05,"reference_active":true,"delta_db":-1.1181791581529947},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.69303691046191e-06,"output_power":3.1701646505516434e-06,"reference_active":true,"delta_db":-0.6630182928611825},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.89764420950003e-08,"output_power":1.1675857922447973e-07,"reference_active":true,"delta_db":1.697912368551155}],"active_band_count":3,"active_band_mae_db":1.159703273188444}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.671604633331299,"predicted_mos":3.671604633331299,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":244224,"input_sample_rate":44100,"input_seconds":5.537959183673469,"input_float32_sha256":"bb1bbfdb0afbef4e63c3dd2c251f2d6b065c9097f3279d8e0f3845051ffbb3df","internal_samples":88608,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.910635821001051},"started_at":"2026-10-13T07:28:45.632762+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008709008998266654,"native_input_samples":243873,"native_input_sha256":"17b124643d18137636b98ff165eb6a17d879eb2c922ca2b5f30a8da9a7c73fcb","output_sha256":"7eb0470825ece456cd0067488462260fe06814024239135df9b7f7eba1f21978","output_samples":244224,"length_delta_samples":351,"analysis_s":0.13497193999864976,"synthesis_s":31.541600395001296,"total_s":31.676572334999946,"rtf":5.728132429475577,"output_peak":0.5811663866043091,"output_rms":0.0687556506582158,"finished_at":"2026-10-13T07:29:17.345220+00:00","evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"7eb0470825ece456cd0067488462260fe06814024239135df9b7f7eba1f21978","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":88480,"output_samples":88608,"length_delta_samples":128,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.24986142635753,"output_peak":0.5752713766109108},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9985554351751534,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.849111685434309,"spectral_convergence":0.5470591716511453,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.612713737398726,"spectral_convergence":0.36087834701171434,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.567098879028316,"spectral_convergence":0.4041030330412222,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.0016361998859241002,"delta_db":-1.2877810823595368,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.001942136696458862,"delta_db":-0.5439917632036153,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0009753429007457741,"delta_db":-0.7313966510017649,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.00013335164958126034,"delta_db":-1.7126455547227697,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":4.5216773269589706e-05,"delta_db":-1.0828066373273149,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":6.257977697174768e-06,"delta_db":-4.244757269654503,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6005631597115844,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8995117270632841,"pitch":null,"output_rms_dbfs":-23.243585757869244,"output_peak":0.5752713766109108,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.676308100620451},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.7499359983522311,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07634775979506618,"rms_dbfs":-22.34407403080596,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.31326298831702,"mean":7.506261897992484e-06,"zero_energy":false},"output":{"rms":0.0688368061699318,"rms_dbfs":-23.243585757869244,"peak":0.5752713766109108,"peak_dbfs":-4.802544686678946,"crest_factor_db":18.4410410711903,"mean":-0.00017961761071314325,"zero_energy":false},"rms_difference_db":-0.8995117270632846},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":4.350872803991291e-06,"output_rms_dbfs":-53.61423612945426,"output_zero_energy":false,"output_peak":0.06608525919811778,"output_energy_fraction":0.00011290644784077746,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":112,"both_estimated_voiced_frames":92,"reference_voiced_output_unvoiced_frames":40,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.696969696969697,"absolute_difference_cents_median":16.192334468272893,"absolute_difference_cents_p95":69.5246978301043,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00024519594682554,"log_rms_mae_db":2.010207381944062,"rms_envelope_nrmse":0.1435730184827887,"envelope_correlation":0.9882691867784779,"rms_gain_error_db":-0.8995109787297704,"reference_mean_power":0.00583423429533283,"output_mean_power":0.004742777686606213},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.472320009859619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016537460501756134,"output_power":0.0013622107658797854,"reference_active":true,"delta_db":-0.8422451162925887},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003778384775464376,"output_power":0.00031663858448793906,"reference_active":true,"delta_db":-0.7674234725448391},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8517105494706513e-05,"output_power":3.4604537680927244e-05,"reference_active":true,"delta_db":-1.467618317738375}],"active_band_count":3,"active_band_mae_db":1.0257623021919342}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020002070942310007,"log_rms_mae_db":1.5488296662029506,"rms_envelope_nrmse":0.11818213974210934,"envelope_correlation":0.9950595609833942,"rms_gain_error_db":-0.8994813948157225,"reference_mean_power":0.005865946925483087,"output_mean_power":0.004768590066103472},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.084203911038529e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015941324207372327,"output_power":0.0013220192664878283,"reference_active":true,"delta_db":-0.8128660995467436},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018248750335661006,"output_power":0.00015340807268585807,"reference_active":true,"delta_db":-0.7538491577514205},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.552526527817181e-06,"output_power":3.1252538275100844e-06,"reference_active":true,"delta_db":-1.633671894002742}],"active_band_count":3,"active_band_mae_db":1.066795717100302}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00018690652054173922,"log_rms_mae_db":2.8752747992954157,"rms_envelope_nrmse":0.182558380844113,"envelope_correlation":0.974121406657585,"rms_gain_error_db":-0.20045516009227737,"reference_mean_power":0.0020929267743018004,"output_mean_power":0.0019985200877005076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":8.903808595053404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005237826202169962,"output_power":0.0005346090304122483,"reference_active":true,"delta_db":0.08885206542406163},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020282121471265822,"output_power":0.00020294819212624083,"reference_active":true,"delta_db":0.002718075385472508},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.57818052122161e-05,"output_power":2.766287666438697e-05,"reference_active":true,"delta_db":0.3058401755384689}],"active_band_count":3,"active_band_mae_db":0.13247010544933432}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00016950242057071555,"log_rms_mae_db":2.1514727180578515,"rms_envelope_nrmse":0.14353209225297642,"envelope_correlation":0.9823768586460366,"rms_gain_error_db":-0.20044382793961513,"reference_mean_power":0.002104315784196243,"output_mean_power":0.0020094006110455286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":7.681995008528509e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000538193996495516,"output_power":0.0005540184393955597,"reference_active":true,"delta_db":0.12585370652306832},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010305287620923108,"output_power":0.00010160282226185585,"reference_active":true,"delta_db":-0.061543457725694584},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.715218896542221e-06,"output_power":2.6763468817100794e-06,"reference_active":true,"delta_db":-0.06262445737658627}],"active_band_count":3,"active_band_mae_db":0.08334054054178307}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":3.4971760453351197,"rms_envelope_nrmse":0.28628049028144237,"envelope_correlation":0.9540418653112637,"rms_gain_error_db":-1.6881864987720452,"reference_mean_power":0.0002131533037612313,"output_mean_power":0.00014450185374644128},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.6207032137826844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010202948507054384,"output_power":6.799886247208847e-05,"reference_active":true,"delta_db":-1.762240472158506},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692530062160889e-05,"output_power":2.826118148817102e-05,"reference_active":true,"delta_db":-1.1613372571366511},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.752197568216406e-06,"output_power":5.507459144327008e-06,"reference_active":true,"delta_db":-0.8849385705776279}],"active_band_count":3,"active_band_mae_db":1.2695054332909284}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":2.784131457241475,"rms_envelope_nrmse":0.23321016849369094,"envelope_correlation":0.9761131101348705,"rms_gain_error_db":-1.6878322216276318,"reference_mean_power":0.00021425147153668843,"output_mean_power":0.00014525817762199372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.4064708995422746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010396272320640845,"output_power":7.067947243252205e-05,"reference_active":true,"delta_db":-1.675843477410481},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.972311051816857e-05,"output_power":1.4992520347375398e-05,"reference_active":true,"delta_db":-1.1910076136918322},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.307963255185345e-07,"output_power":5.16418346029649e-07,"reference_active":true,"delta_db":-0.8688749242972269}],"active_band_count":3,"active_band_mae_db":1.24524200513318}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":3.6759452384177416,"rms_envelope_nrmse":0.4293898285732863,"envelope_correlation":0.8604182863108847,"rms_gain_error_db":-1.1895135092906242,"reference_mean_power":2.6417005173666296e-05,"output_mean_power":2.0087793272567676e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.6945603106309378e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5028414095885614e-05,"output_power":8.787449996501801e-06,"reference_active":true,"delta_db":-2.330502864364093},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.702646590810387e-06,"output_power":4.213286549099705e-06,"reference_active":true,"delta_db":-2.0162532417583057},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.011540321330317e-07,"output_power":8.916840300580214e-07,"reference_active":true,"delta_db":-0.045880417000718474}],"active_band_count":3,"active_band_mae_db":1.4642121743743726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":3.208024171508293,"rms_envelope_nrmse":0.35765943689736546,"envelope_correlation":0.8954434742944237,"rms_gain_error_db":-1.1888003015583717,"reference_mean_power":2.6545935258029172e-05,"output_mean_power":2.0189148418398044e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.4345030346700565e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.623830647966854e-05,"output_power":9.069308694411188e-06,"reference_active":true,"delta_db":-2.529665496209632},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.69303691046191e-06,"output_power":2.459903260185809e-06,"reference_active":true,"delta_db":-1.7646562019484842},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.89764420950003e-08,"output_power":8.483231981570342e-08,"reference_active":true,"delta_db":0.3106377855005038}],"active_band_count":3,"active_band_mae_db":1.5349864945528733}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"golf","status":"measured","mos":{"status":"measured","value":2.5562145709991455,"predicted_mos":2.5562145709991455,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":132480,"input_sample_rate":24000,"input_seconds":5.52,"input_float32_sha256":"f769d3b38d051a4d0a9d373c4d036485cf4b9eb81ea480df5e7e4093f7524d90","internal_samples":88320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5421743799997785},"started_at":"2026-10-13T05:02:59.926662+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002777310000055877,"native_input_samples":132720,"native_input_sha256":"11f38103be40a6f5aeccf7e90a5051921ebfce1ae1a3f7de4b6631a4886c0878","output_sha256":"4427be40e3f4b6bb27dae185cebfd3be24cb1d96bc23b45b8085289abe88d932","output_samples":132480,"length_delta_samples":-240,"analysis_s":0.5252189929997257,"synthesis_s":0.11884721200021886,"total_s":0.6440662049999446,"rtf":0.11646766817358853,"output_peak":0.54095458984375,"output_rms":0.07021541796295992,"finished_at":"2026-10-13T05:03:00.582557+00:00","analysis_stages":{"external_f0_s":0.08613954599968565},"evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"4427be40e3f4b6bb27dae185cebfd3be24cb1d96bc23b45b8085289abe88d932","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":88480,"output_samples":88320,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.068070521113846,"output_peak":0.5411047007654923},"common_prefix":{"samples":88320,"sample_rate":16000,"reference_coverage":0.9981916817359855,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88320,"output_samples":88320,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.055012717450035,"spectral_convergence":0.7162162284733272,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.357494290982323,"spectral_convergence":0.3244415791081586,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6267273819762575,"spectral_convergence":0.3132599384418436,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002211592157702101,"output_power":0.001974485946039604,"delta_db":-0.492509943925992,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0022002916201661754,"output_power":0.0019351721767885746,"delta_db":-0.5576063347912626,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011546940723108348,"output_power":0.0008448392420207451,"delta_db":-1.3569285809879836,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019816786027308027,"output_power":0.00012935267548517236,"delta_db":-1.8525780414976543,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.8121778147445474e-05,"output_power":3.8924804279280323e-05,"delta_db":-1.7411245487160563,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6659703596085356e-05,"output_power":1.1154749039957776e-05,"delta_db":-1.7420746636317574,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.290470352258451,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7318473721471779,"pitch":null,"output_rms_dbfs":-23.068070521113846,"output_peak":0.5411047007654923,"reference_rms_dbfs":-22.336223148966667,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.013078130136205},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88320,"output_samples":88320,"intelligibility":{"estoi":{"value":0.8664898276281178,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.076416799149112,"rms_dbfs":-22.33622314896667,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.30541210647773,"mean":7.055116736370584e-06,"zero_energy":false},"output":{"rms":0.0702419361468157,"rms_dbfs":-23.068070521113846,"peak":0.5411047007654923,"peak_dbfs":-5.334373863664155,"crest_factor_db":17.73369665744969,"mean":0.000993238175829144,"zero_energy":false},"rms_difference_db":-0.7318473721471754},"full_scale":{"reference":{"sample_count":88320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88320,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":552,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":7.046924784605354e-06,"output_rms_dbfs":-51.52000363922794,"output_zero_energy":false,"output_peak":0.04273681579338889,"output_energy_fraction":0.00017594483535076765,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":545,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":189,"both_estimated_voiced_frames":128,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":61,"paired_voiced_fraction":0.9696969696969697,"absolute_difference_cents_median":9.325499490587251,"absolute_difference_cents_p95":38.700943930586504,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88320,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1103,"rms_floor":0.00024519594682554,"log_rms_mae_db":1.8381775608721755,"rms_envelope_nrmse":0.15073446876537885,"envelope_correlation":0.9821473825414215,"rms_gain_error_db":-0.7318522898254597,"reference_mean_power":0.0058448001771915364,"output_mean_power":0.004938379248861437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1103,"frequency_resolution_hz":0.1813236627379873,"power_floor":2.4707518810758626e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016543746840648216,"output_power":0.0014815296288301198,"reference_active":true,"delta_db":-0.47923534432499937},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00037561762545314367,"output_power":0.00028522645387616476,"reference_active":true,"delta_db":-1.1955616034871404},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.80488975294345e-05,"output_power":3.6226147124759405e-05,"reference_active":true,"delta_db":-1.2266128075375364}],"active_band_count":3,"active_band_mae_db":0.9671365851165588}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1097,"rms_floor":0.00020002070942310007,"log_rms_mae_db":1.3843901093751727,"rms_envelope_nrmse":0.11366754525241199,"envelope_correlation":0.9919946003253993,"rms_gain_error_db":-0.731827312233227,"reference_mean_power":0.00587662370746434,"output_mean_power":0.004965296092172475},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1097,"frequency_resolution_hz":0.18231540565177756,"power_floor":2.0829672176948064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015925584329258132,"output_power":0.0014185802216180278,"reference_active":true,"delta_db":-0.5024147564112234},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018132912087283863,"output_power":0.00013465357858531788,"reference_active":true,"delta_db":-1.2924965617810413},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.545205135527191e-06,"output_power":3.374296818265656e-06,"reference_active":true,"delta_db":-1.293702062840667}],"active_band_count":3,"active_band_mae_db":1.029537793677644}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1103,"rms_floor":0.00018690652054173922,"log_rms_mae_db":2.3181152697483127,"rms_envelope_nrmse":0.18512514165271193,"envelope_correlation":0.98003120433522,"rms_gain_error_db":-0.9695968716000926,"reference_mean_power":0.002096717717909672,"output_mean_power":0.0016771823289828832},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1103,"frequency_resolution_hz":0.1813236627379873,"power_floor":8.899977709994408e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005245505492940695,"output_power":0.0004349581716056491,"reference_active":true,"delta_db":-0.8133985175267356},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020121675296914303,"output_power":0.00016030930500794831,"reference_active":true,"delta_db":-0.9870540520453907},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.556689114478124e-05,"output_power":2.152132158708666e-05,"reference_active":true,"delta_db":-0.7480898528013733}],"active_band_count":3,"active_band_mae_db":0.8495141407911665}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1097,"rms_floor":0.00016950242057071555,"log_rms_mae_db":1.6926024616180575,"rms_envelope_nrmse":0.14485586460484468,"envelope_correlation":0.990593799598869,"rms_gain_error_db":-0.9695604218268995,"reference_mean_power":0.0021081447481821453,"output_mean_power":0.0016863370608002463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1097,"frequency_resolution_hz":0.18231540565177756,"power_floor":7.680431708586589e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005377912730410487,"output_power":0.00044850715785989834,"reference_active":true,"delta_db":-0.7884437199930341},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010246668354542539,"output_power":8.267961576378035e-05,"reference_active":true,"delta_db":-0.9318423031576986},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.70302857288894e-06,"output_power":2.1373763912724033e-06,"reference_active":true,"delta_db":-1.0196962850441689}],"active_band_count":3,"active_band_mae_db":0.9133274360649671}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1103,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":2.3778753577978136,"rms_envelope_nrmse":0.284572782172783,"envelope_correlation":0.956953873660366,"rms_gain_error_db":-1.795335734241112,"reference_mean_power":0.00021353432640854218,"output_mean_power":0.0001412323306301448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1103,"frequency_resolution_hz":0.1813236627379873,"power_floor":1.6241260108833537e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010234780450320082,"output_power":6.424468447457252e-05,"reference_active":true,"delta_db":-2.0224133039949006},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.689555652336319e-05,"output_power":2.3487737183559663e-05,"reference_active":true,"delta_db":-1.9613288680234584},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.7221756989532275e-06,"output_power":5.864763342274463e-06,"reference_active":true,"delta_db":-0.5925936758295842}],"active_band_count":3,"active_band_mae_db":1.525445282615981}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1097,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":1.8665371070031473,"rms_envelope_nrmse":0.23634789071003465,"envelope_correlation":0.9777391278178927,"rms_gain_error_db":-1.7954812269545026,"reference_mean_power":0.00021463523108481707,"output_mean_power":0.00014195571700296747},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1097,"frequency_resolution_hz":0.18231540565177756,"power_floor":1.4096411150253293e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010420738619312387,"output_power":6.48333425222103e-05,"reference_active":true,"delta_db":-2.061000902423623},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.975859194449227e-05,"output_power":1.3520040497680993e-05,"reference_active":true,"delta_db":-1.6477799976401484},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.295986025312372e-07,"output_power":5.543686593394374e-07,"reference_active":true,"delta_db":-0.5526508547697213}],"active_band_count":3,"active_band_mae_db":1.4204772516111641}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1103,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":2.6897520367317034,"rms_envelope_nrmse":0.25344572502816715,"envelope_correlation":0.9633567293074453,"rms_gain_error_db":-1.2484393244864331,"reference_mean_power":2.6463447144942018e-05,"output_mean_power":1.9851918453230577e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1103,"frequency_resolution_hz":0.1813236627379873,"power_floor":2.698737164592067e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5049939487118587e-05,"output_power":1.137012413774288e-05,"reference_active":true,"delta_db":-1.2176954742750552},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.700182541109718e-06,"output_power":5.7177559197389655e-06,"reference_active":true,"delta_db":-0.6886102266557794},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.96149303093097e-07,"output_power":1.116982702364882e-06,"reference_active":true,"delta_db":0.9566607627793349}],"active_band_count":3,"active_band_mae_db":0.9543221545700565}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88320,"uncovered_tail_samples":0,"metrics":{"frame_count":1097,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":2.372059497428872,"rms_envelope_nrmse":0.2004049543287966,"envelope_correlation":0.980859057635727,"rms_gain_error_db":-1.2481519093572166,"reference_mean_power":2.6593395022841922e-05,"output_mean_power":1.995072092978952e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1097,"frequency_resolution_hz":0.18231540565177756,"power_floor":2.4386259655756482e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.629086654236559e-05,"output_power":1.2716792915273257e-05,"reference_active":true,"delta_db":-1.0756658673604609},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6746889506644716e-06,"output_power":3.235358740774785e-06,"reference_active":true,"delta_db":-0.5529814068416888},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.814838347157667e-08,"output_power":9.149618071151357e-08,"reference_active":true,"delta_db":0.6848296695892038}],"active_band_count":3,"active_band_mae_db":0.7711589812637846}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3802289962768555,"predicted_mos":3.3802289962768555,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":132840,"input_sample_rate":24000,"input_seconds":5.535,"input_float32_sha256":"e364b3ce9f788ba75972359b5b856e9415c96079b34ea9350d77229fd1c1858e","internal_samples":88560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.18826466199971},"started_at":"2026-10-13T07:45:06.597966+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0024370450009882916,"native_input_samples":132720,"native_input_sha256":"11f38103be40a6f5aeccf7e90a5051921ebfce1ae1a3f7de4b6631a4886c0878","output_sha256":"f3adb09dc9b244c90a1c128ddd3cffbcd9003ece6d975ab4c96521fa3c7e4b66","output_samples":132840,"length_delta_samples":120,"analysis_s":2.1334358530002646,"synthesis_s":11.70681714300008,"total_s":13.840252996000345,"rtf":2.5027582271248363,"output_peak":0.6352573037147522,"output_rms":0.07978276484994876,"finished_at":"2026-10-13T07:45:20.449310+00:00","analysis_stages":{"external_f0_s":1.3640551950011286},"evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"f3adb09dc9b244c90a1c128ddd3cffbcd9003ece6d975ab4c96521fa3c7e4b66","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":88480,"output_samples":88560,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.959123679321447,"output_peak":0.6481360086103312},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.999096657633243,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.26885032052388,"spectral_convergence":0.7533866874245935,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.589116981807282,"spectral_convergence":0.3639271734950991,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.3541497144915216,"spectral_convergence":0.3372408258427547,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.003179517934095272,"delta_db":1.597468126856887,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.0020383312832452545,"delta_db":-0.33404200745653767,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.000913222185312395,"delta_db":-1.0172053885203765,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.0001746935916168683,"delta_db":-0.5398597401694001,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":5.630850104899886e-05,"delta_db":-0.13006264571677434,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.2912490502557128e-05,"delta_db":-1.0989972290108634,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7862725229551399,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.38886637474879426,"pitch":null,"output_rms_dbfs":-21.95520765605717,"output_peak":0.6481360086103312,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.070705672274227},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.8704826721227837,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07634775979506618,"rms_dbfs":-22.34407403080596,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.31326298831702,"mean":7.506261897992484e-06,"zero_energy":false},"output":{"rms":0.0798435093531174,"rms_dbfs":-21.95520765605717,"peak":0.6481360086103312,"peak_dbfs":-3.766676993978217,"crest_factor_db":18.188530662078946,"mean":5.804387964027963e-05,"zero_energy":false},"rms_difference_db":0.3888663747487904},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":8.913595546672984e-05,"output_rms_dbfs":-40.49947075862715,"output_zero_energy":false,"output_peak":0.17603305478460995,"output_energy_fraction":0.001719322992934681,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":196,"both_estimated_voiced_frames":122,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":74,"paired_voiced_fraction":0.9242424242424242,"absolute_difference_cents_median":7.894947796172954,"absolute_difference_cents_p95":32.69955796086297,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00024519594682554,"log_rms_mae_db":2.4527544255039015,"rms_envelope_nrmse":0.19548972335092363,"envelope_correlation":0.9647617820059619,"rms_gain_error_db":0.3888602773466846,"reference_mean_power":0.00583423429533283,"output_mean_power":0.006380723031463447},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.472320009859619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016537460501756134,"output_power":0.0019101819195165724,"reference_active":true,"delta_db":0.6260591008877164},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003778384775464376,"output_power":0.00036009086807139664,"reference_active":true,"delta_db":-0.20894074862394701},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8517105494706513e-05,"output_power":4.2116323413801404e-05,"reference_active":true,"delta_db":-0.6144443118348555}],"active_band_count":3,"active_band_mae_db":0.483148053782173}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020002070942310007,"log_rms_mae_db":1.8305011328217384,"rms_envelope_nrmse":0.13609662433426553,"envelope_correlation":0.9817957164613164,"rms_gain_error_db":0.38889499928810956,"reference_mean_power":0.005865946925483087,"output_mean_power":0.006415457453511561},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.084203911038529e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015941324207372327,"output_power":0.001816500108301694,"reference_active":true,"delta_db":0.5671103373763542},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018248750335661006,"output_power":0.0001697840631451469,"reference_active":true,"delta_db":-0.31336206983802717},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.552526527817181e-06,"output_power":4.212315635522839e-06,"reference_active":true,"delta_db":-0.33731579021351377}],"active_band_count":3,"active_band_mae_db":0.40592939914263165}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00018690652054173922,"log_rms_mae_db":3.057985971413476,"rms_envelope_nrmse":0.22701912723339535,"envelope_correlation":0.9622622069390319,"rms_gain_error_db":-0.7485799335614527,"reference_mean_power":0.0020929267743018004,"output_mean_power":0.0017615543227836334},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":8.903808595053404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005237826202169962,"output_power":0.00047885606208366617,"reference_active":true,"delta_db":-0.38946094130484693},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020282121471265822,"output_power":0.00017439824651280453,"reference_active":true,"delta_db":-0.6557126540807243},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.57818052122161e-05,"output_power":2.486955166823108e-05,"reference_active":true,"delta_db":-0.15645366712053102}],"active_band_count":3,"active_band_mae_db":0.4005424208353674}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00016950242057071555,"log_rms_mae_db":2.273152360058758,"rms_envelope_nrmse":0.15737142968394388,"envelope_correlation":0.9831813759806006,"rms_gain_error_db":-0.7485492998846393,"reference_mean_power":0.002104315784196243,"output_mean_power":0.0017711526073221628},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":7.681995008528509e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000538193996495516,"output_power":0.0004916094697507177,"reference_active":true,"delta_db":-0.3931860893841255},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010305287620923108,"output_power":8.855558403466002e-05,"reference_active":true,"delta_db":-0.6584416583384674},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.715218896542221e-06,"output_power":2.7251654528025745e-06,"reference_active":true,"delta_db":0.015880272545122212}],"active_band_count":3,"active_band_mae_db":0.355836006755905}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":2.9641808289833076,"rms_envelope_nrmse":0.28403987486369797,"envelope_correlation":0.9376435316626068,"rms_gain_error_db":-0.3737531248040703,"reference_mean_power":0.0002131533037612313,"output_mean_power":0.00019557653876226507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.6207032137826844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010202948507054384,"output_power":9.474571891648427e-05,"reference_active":true,"delta_db":-0.32166099335013015},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692530062160889e-05,"output_power":3.3558889207357784e-05,"reference_active":true,"delta_db":-0.41516462407054744},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.752197568216406e-06,"output_power":5.00813142700506e-06,"reference_active":true,"delta_db":-1.2976942376865375}],"active_band_count":3,"active_band_mae_db":0.6781732850357383}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":1.9961885278639597,"rms_envelope_nrmse":0.1637230215207031,"envelope_correlation":0.9775127503777012,"rms_gain_error_db":-0.37353895270594656,"reference_mean_power":0.00021425147153668843,"output_mean_power":0.00019659384565951434},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.4064708995422746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010396272320640845,"output_power":9.590087610054153e-05,"reference_active":true,"delta_db":-0.3505507223882103},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.972311051816857e-05,"output_power":1.779903834354486e-05,"reference_active":true,"delta_db":-0.44578869653968634},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.307963255185345e-07,"output_power":4.898466255367802e-07,"reference_active":true,"delta_db":-1.0982903421445587}],"active_band_count":3,"active_band_mae_db":0.6315432536908184}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":2.869182603587589,"rms_envelope_nrmse":0.32300282617882325,"envelope_correlation":0.9226181393594632,"rms_gain_error_db":-0.47715227655041076,"reference_mean_power":2.6417005173666296e-05,"output_mean_power":2.3668370204272654e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.6945603106309378e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5028414095885614e-05,"output_power":1.3022006552882626e-05,"reference_active":true,"delta_db":-0.6223524363712906},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.702646590810387e-06,"output_power":5.6966881406932226e-06,"reference_active":true,"delta_db":-0.7062387592900965},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.011540321330317e-07,"output_power":5.82509032903449e-07,"reference_active":true,"delta_db":-1.8949636594837953}],"active_band_count":3,"active_band_mae_db":1.074518285048394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":2.0023820660397713,"rms_envelope_nrmse":0.1750781281764938,"envelope_correlation":0.9763257424654634,"rms_gain_error_db":-0.4770254587184981,"reference_mean_power":2.6545935258029172e-05,"output_mean_power":2.3784579899593407e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.4345030346700565e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.623830647966854e-05,"output_power":1.3808554620759936e-05,"reference_active":true,"delta_db":-0.703925117994269},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.69303691046191e-06,"output_power":3.0610379618304594e-06,"reference_active":true,"delta_db":-0.815149326904534},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.89764420950003e-08,"output_power":5.672864216098367e-08,"reference_active":true,"delta_db":-1.4369517623289307}],"active_band_count":3,"active_band_mae_db":0.9853420690759113}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.1239218711853027,"predicted_mos":3.1239218711853027,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":132960,"input_sample_rate":24000,"input_seconds":5.54,"input_float32_sha256":"da28d65f60d00abc5320abc310b83bb3175e62e79e0280afe7fc9f1145c093f7","internal_samples":88640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5635151140013477},"started_at":"2026-10-13T05:11:23.421506+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031264549998013536,"native_input_samples":132720,"native_input_sha256":"11f38103be40a6f5aeccf7e90a5051921ebfce1ae1a3f7de4b6631a4886c0878","output_sha256":"0ccf16e3e65d00476f40188084d89c10f011c5c1623e452b112f6c248b743352","output_samples":132960,"length_delta_samples":240,"analysis_s":1.177652100999694,"synthesis_s":0.14042402300037793,"total_s":1.318076124000072,"rtf":0.23835011283907268,"output_peak":0.6326834559440613,"output_rms":0.07449195463724875,"finished_at":"2026-10-13T05:11:24.751033+00:00","analysis_stages":{"external_f0_s":0.16801672799920198},"evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"0ccf16e3e65d00476f40188084d89c10f011c5c1623e452b112f6c248b743352","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":88480,"output_samples":88640,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.55448441794616,"output_peak":0.6300530549879979},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981949458483754,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.283300150903851,"spectral_convergence":0.188232945955572,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.6233544331932235,"spectral_convergence":0.17952576869355155,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.713313929355466,"spectral_convergence":0.18626614252266557,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.0021503538300264996,"delta_db":-0.1010453948993776,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.002075684169701634,"delta_db":-0.25517697429634156,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0011019523244577654,"delta_db":-0.20134186968258097,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.0001725399557107982,"delta_db":-0.5937326560066438,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":4.96973837601621e-05,"delta_db":-0.6724670441135043,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.3118360522573745e-05,"delta_db":-1.0303017585855008,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4756776162639915,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2025652426632436,"pitch":null,"output_rms_dbfs":-22.546639273469204,"output_peak":0.6300530549879979,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.873322837817514},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.9334387506979062,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07634775979506618,"rms_dbfs":-22.34407403080596,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.31326298831702,"mean":7.506261897992484e-06,"zero_energy":false},"output":{"rms":0.0745878409942226,"rms_dbfs":-22.546639273469204,"peak":0.6300530549879979,"peak_dbfs":-4.012457565900799,"crest_factor_db":18.534181707568404,"mean":0.00013336393479664073,"zero_energy":false},"rms_difference_db":-0.20256524266324405},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":2.3001802077586313e-05,"output_rms_dbfs":-46.38238137821833,"output_zero_energy":false,"output_peak":0.06461305363671903,"output_energy_fraction":0.0005084047165595917,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":121,"both_estimated_voiced_frames":111,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.8409090909090909,"absolute_difference_cents_median":1.9449858750212419,"absolute_difference_cents_p95":11.221226301701554,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00024519594682554,"log_rms_mae_db":1.2209672194585974,"rms_envelope_nrmse":0.09070403431050647,"envelope_correlation":0.9920239368475252,"rms_gain_error_db":-0.20256246191754604,"reference_mean_power":0.00583423429533283,"output_mean_power":0.005568364033320217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.472320009859619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016537460501756134,"output_power":0.0015378475526991492,"reference_active":true,"delta_db":-0.31555534069723945},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003778384775464376,"output_power":0.00030761432710259376,"reference_active":true,"delta_db":-0.8929962381913193},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8517105494706513e-05,"output_power":3.072149253211573e-05,"reference_active":true,"delta_db":-1.9845257212708982}],"active_band_count":3,"active_band_mae_db":1.0643591000531523}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020002070942310007,"log_rms_mae_db":0.7549803379980912,"rms_envelope_nrmse":0.05561918671480938,"envelope_correlation":0.9969788134363674,"rms_gain_error_db":-0.20255045480615053,"reference_mean_power":0.005865946925483087,"output_mean_power":0.005598646974836272},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.084203911038529e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015941324207372327,"output_power":0.0014778783856359604,"reference_active":true,"delta_db":-0.3288569684498506},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018248750335661006,"output_power":0.00015658107138419506,"reference_active":true,"delta_db":-0.6649386922618082},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.552526527817181e-06,"output_power":2.640555026777978e-06,"reference_active":true,"delta_db":-2.3655726290872576}],"active_band_count":3,"active_band_mae_db":1.119789429932972}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00018690652054173922,"log_rms_mae_db":1.6825019385393765,"rms_envelope_nrmse":0.11374730943220289,"envelope_correlation":0.9903294741858946,"rms_gain_error_db":-0.2457127130533955,"reference_mean_power":0.0020929267743018004,"output_mean_power":0.0019778017760145703},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":8.903808595053404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005237826202169962,"output_power":0.0004566448490279761,"reference_active":true,"delta_db":-0.5957252072458789},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020282121471265822,"output_power":0.00016328005441076823,"reference_active":true,"delta_db":-0.9418024298210957},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.57818052122161e-05,"output_power":1.7220176933980262e-05,"reference_active":true,"delta_db":-1.7527571344774107}],"active_band_count":3,"active_band_mae_db":1.0967615905147952}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00016950242057071555,"log_rms_mae_db":0.9711750602688005,"rms_envelope_nrmse":0.070226013881135,"envelope_correlation":0.9963262408488736,"rms_gain_error_db":-0.2456842868487373,"reference_mean_power":0.002104315784196243,"output_mean_power":0.0019885773299843884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":7.681995008528509e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000538193996495516,"output_power":0.00046298580419454577,"reference_active":true,"delta_db":-0.653711738062668},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010305287620923108,"output_power":8.710015688099018e-05,"reference_active":true,"delta_db":-0.7304118021736155},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.715218896542221e-06,"output_power":1.3880226456309574e-06,"reference_active":true,"delta_db":-2.914082957486836}],"active_band_count":3,"active_band_mae_db":1.43273549924104}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":1.951732178233474,"rms_envelope_nrmse":0.16633465138898507,"envelope_correlation":0.9819643193045505,"rms_gain_error_db":-0.6609252493103526,"reference_mean_power":0.0002131533037612313,"output_mean_power":0.00018306256508594958},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.6207032137826844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010202948507054384,"output_power":8.495816526064705e-05,"reference_active":true,"delta_db":-0.7952056992763528},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692530062160889e-05,"output_power":2.7598562352250984e-05,"reference_active":true,"delta_db":-1.2643758009267243},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.752197568216406e-06,"output_power":3.7500558106969466e-06,"reference_active":true,"delta_db":-2.554074099699553}],"active_band_count":3,"active_band_mae_db":1.5378851999675434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":1.3233289422400105,"rms_envelope_nrmse":0.10846683234568434,"envelope_correlation":0.994102022607696,"rms_gain_error_db":-0.6610856603915173,"reference_mean_power":0.00021425147153668843,"output_mean_power":0.0001839989087876083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.4064708995422746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010396272320640845,"output_power":8.482639551888443e-05,"reference_active":true,"delta_db":-0.8834663375770564},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.972311051816857e-05,"output_power":1.4834717912088549e-05,"reference_active":true,"delta_db":-1.236961158367044},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.307963255185345e-07,"output_power":2.800027597372154e-07,"reference_active":true,"delta_db":-3.527268430186548}],"active_band_count":3,"active_band_mae_db":1.882565308710216}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":2.0172397651249447,"rms_envelope_nrmse":0.15820279539218407,"envelope_correlation":0.9845361066054844,"rms_gain_error_db":-0.6332709066051803,"reference_mean_power":2.6417005173666296e-05,"output_mean_power":2.2832658997658134e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.6945603106309378e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5028414095885614e-05,"output_power":1.343063758795836e-05,"reference_active":true,"delta_db":-0.4881652292394344},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.702646590810387e-06,"output_power":5.789943497602208e-06,"reference_active":true,"delta_db":-0.6357199544087},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.011540321330317e-07,"output_power":6.328296582925728e-07,"reference_active":true,"delta_db":-1.5351220557647858}],"active_band_count":3,"active_band_mae_db":0.8863357464709735}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":1.4228291612482624,"rms_envelope_nrmse":0.10451973117562044,"envelope_correlation":0.9948362215349571,"rms_gain_error_db":-0.6329732942986461,"reference_mean_power":2.6545935258029172e-05,"output_mean_power":2.2945667785462515e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.4345030346700565e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.623830647966854e-05,"output_power":1.4436097746585523e-05,"reference_active":true,"delta_db":-0.5109091996551665},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.69303691046191e-06,"output_power":3.1986505147440916e-06,"reference_active":true,"delta_db":-0.6241685671048682},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.89764420950003e-08,"output_power":4.767169462952472e-08,"reference_active":true,"delta_db":-2.19236974318885}],"active_band_count":3,"active_band_mae_db":1.1091491699829614}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7176-88083-0026","method":"magphase","status":"measured","mos":{"status":"measured","value":3.105912446975708,"predicted_mos":3.105912446975708,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":267594,"input_sample_rate":48000,"input_seconds":5.574875,"input_float32_sha256":"917303a8e21ed8ef685322940fd6a91823eb404f1064810d57f89feea1ee4409","internal_samples":89198,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.765405398000439},"started_at":"2026-10-13T05:18:05.658358+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010079985999254859,"native_input_samples":265440,"native_input_sha256":"4dbb4e0df07feada5b7cece3b609524c52ab344fbdb2d4828f9b59befddca02e","output_sha256":"ad8d6b566352b09b24cb1f6a35f01832f5a61354269a5d4b1a3595ac9f33e67f","output_samples":267594,"length_delta_samples":2154,"analysis_s":2.5492330380002386,"synthesis_s":0.8818742660005228,"total_s":3.4311073040007614,"rtf":0.6204534003618013,"output_peak":0.6450969187892771,"output_rms":0.07401234044103334,"finished_at":"2026-10-13T05:18:09.110884+00:00","evaluation_input":{"reference":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"output_sha256":"ad8d6b566352b09b24cb1f6a35f01832f5a61354269a5d4b1a3595ac9f33e67f","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":88480,"output_samples":89198,"length_delta_samples":718,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.610579026051603,"output_peak":0.6446422468992673},"common_prefix":{"samples":88480,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9919504921635014,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":88480,"output_samples":88480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.580698914716974,"spectral_convergence":0.8591662783042624,"spectral_convergence_status":"ok","amplitude_floor":2.1848847242256272e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.797200727648589,"spectral_convergence":0.6136471014044996,"spectral_convergence_status":"ok","amplitude_floor":1.2036494394056804e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.468085486854983,"spectral_convergence":0.578733391150899,"spectral_convergence_status":"ok","amplitude_floor":7.153950066191215e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.002200971739696214,"output_power":0.001975516146572964,"delta_db":-0.46933882612053607,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002201298666996261,"output_power":0.0021214564973791655,"delta_db":-0.1604484005472901,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0011542423166681864,"output_power":0.0011483919038736612,"delta_db":-0.02206870234247573,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00019781669780676847,"output_power":0.00019826884913841177,"delta_db":0.009915378893776413,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.802033486419172e-05,"output_power":6.675746822288043e-05,"delta_db":0.6091962669720524,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.6630669693501258e-05,"output_power":1.6078968360188376e-05,"delta_db":-0.1465155737922522,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23624719144473047,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23144793167857441,"pitch":null,"output_rms_dbfs":-22.575521962484537,"output_peak":0.6446422468992673,"reference_rms_dbfs":-22.34407403080596,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.615328376406849},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":88480,"output_samples":88480,"intelligibility":{"estoi":{"value":0.5722115124800389,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07634775979506618,"rms_dbfs":-22.34407403080596,"peak":0.62872314453125,"peak_dbfs":-4.0308110424889385,"crest_factor_db":18.31326298831702,"mean":7.506261897992484e-06,"zero_energy":false},"output":{"rms":0.07434023024949056,"rms_dbfs":-22.575521962484537,"peak":0.6446422468992673,"peak_dbfs":-3.813624724410158,"crest_factor_db":18.761897238074376,"mean":0.0006870000836743895,"zero_energy":false},"rms_difference_db":-0.23144793167857713},"full_scale":{"reference":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":88480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":553,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.012105233967304e-06,"threshold_dbfs":-52.20973426266879,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.2731874624322004e-06,"output_mean_power":8.475393940336753e-05,"output_rms_dbfs":-40.71840106441894,"output_zero_energy":false,"output_peak":0.1471681778662109,"output_energy_fraction":0.001885801041293285,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":546,"reference_estimated_voiced_frames":132,"output_estimated_voiced_frames":116,"both_estimated_voiced_frames":99,"reference_voiced_output_unvoiced_frames":33,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.75,"absolute_difference_cents_median":11.61374675814765,"absolute_difference_cents_p95":47.327188837884265,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":88480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00024519594682554,"log_rms_mae_db":5.705764848420236,"rms_envelope_nrmse":0.5111590933975975,"envelope_correlation":0.7336714948795997,"rms_gain_error_db":-0.2314405987087381,"reference_mean_power":0.00583423429533283,"output_mean_power":0.0055314603792282506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.472320009859619e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0016537460501756134,"output_power":0.001620449997381422,"reference_active":true,"delta_db":-0.08833185478479338},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0003778384775464376,"output_power":0.0003720052778307545,"reference_active":true,"delta_db":-0.06757081150669847},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.8517105494706513e-05,"output_power":4.269617509584587e-05,"reference_active":true,"delta_db":-0.5550591235538208}],"active_band_count":3,"active_band_mae_db":0.2369872632817709}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00020002070942310007,"log_rms_mae_db":4.684668369255152,"rms_envelope_nrmse":0.4127491165886022,"envelope_correlation":0.8027496691888891,"rms_gain_error_db":-0.23144500302162452,"reference_mean_power":0.005865946925483087,"output_mean_power":0.005561521607932196},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.084203911038529e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015941324207372327,"output_power":0.0015601597392270756,"reference_active":true,"delta_db":-0.09355327850210782},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00018248750335661006,"output_power":0.00018302068133751822,"reference_active":true,"delta_db":0.012670382164885178},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.552526527817181e-06,"output_power":4.338535258623661e-06,"reference_active":true,"delta_db":-0.20909353883542459}],"active_band_count":3,"active_band_mae_db":0.10510573316747253}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":0.00018690652054173922,"log_rms_mae_db":6.85476269434573,"rms_envelope_nrmse":0.6113195881963492,"envelope_correlation":0.7093084328445739,"rms_gain_error_db":-0.08970755693451418,"reference_mean_power":0.0020929267743018004,"output_mean_power":0.002050138862208915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":8.903808595053404e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005237826202169962,"output_power":0.0005172510737926339,"reference_active":true,"delta_db":-0.05449682960389172},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020282121471265822,"output_power":0.00018955687272727878,"reference_active":true,"delta_db":-0.29373844232762614},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.57818052122161e-05,"output_power":2.6376642795231284e-05,"reference_active":true,"delta_db":0.09906195082148031}],"active_band_count":3,"active_band_mae_db":0.1490990742509994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":0.00016950242057071555,"log_rms_mae_db":5.636030661896071,"rms_envelope_nrmse":0.49360482743325734,"envelope_correlation":0.7897860596754164,"rms_gain_error_db":-0.08968664152610253,"reference_mean_power":0.002104315784196243,"output_mean_power":0.0020613049616670497},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":7.681995008528509e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000538193996495516,"output_power":0.0005255513178151551,"reference_active":true,"delta_db":-0.10323719525496362},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010305287620923108,"output_power":0.00010202544528019874,"reference_active":true,"delta_db":-0.04351618572189148},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.715218896542221e-06,"output_power":2.693181681451575e-06,"reference_active":true,"delta_db":-0.03539195641553052}],"active_band_count":3,"active_band_mae_db":0.06071511246412855}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":7.550458878286076e-05,"log_rms_mae_db":6.936211264614079,"rms_envelope_nrmse":0.6377074524216125,"envelope_correlation":0.6969435950861101,"rms_gain_error_db":0.0427109523146273,"reference_mean_power":0.0002131533037612313,"output_mean_power":0.00021525991457763756},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":1.6207032137826844e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010202948507054384,"output_power":0.000107288210690206,"reference_active":true,"delta_db":0.218263075316675},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.692530062160889e-05,"output_power":3.6408139547349026e-05,"reference_active":true,"delta_db":-0.06125552647518563},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.752197568216406e-06,"output_power":5.553522547066798e-06,"reference_active":true,"delta_db":-0.8487660176588459}],"active_band_count":3,"active_band_mae_db":0.37609487315023554}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":5.938009261210463e-05,"log_rms_mae_db":5.537322222844121,"rms_envelope_nrmse":0.4934210832842966,"envelope_correlation":0.794630703129125,"rms_gain_error_db":0.04270759676519618,"reference_mean_power":0.00021425147153668843,"output_mean_power":0.00021636876845528293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":1.4064708995422746e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010396272320640845,"output_power":0.00011009221335850776,"reference_active":true,"delta_db":0.24878956206148162},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.972311051816857e-05,"output_power":1.9201822143167996e-05,"reference_active":true,"delta_db":-0.11632965567094271},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.307963255185345e-07,"output_power":5.029808491411429e-07,"reference_active":true,"delta_db":-0.9833770510835221}],"active_band_count":3,"active_band_mae_db":0.44949875627198216}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1105,"rms_floor":4.429193400462834e-05,"log_rms_mae_db":7.347751733241986,"rms_envelope_nrmse":0.7414658851538515,"envelope_correlation":0.6528546707634904,"rms_gain_error_db":0.775839400531868,"reference_mean_power":2.6417005173666296e-05,"output_mean_power":3.158402838529315e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1105,"frequency_resolution_hz":0.18099547511312217,"power_floor":2.6945603106309378e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5028414095885614e-05,"output_power":1.824222287671961e-05,"reference_active":true,"delta_db":0.8416460428027926},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.702646590810387e-06,"output_power":1.0176121441156316e-05,"reference_active":true,"delta_db":1.8133596014300977},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.011540321330317e-07,"output_power":1.947929642784102e-06,"reference_active":true,"delta_db":3.3477423630034213}],"active_band_count":3,"active_band_mae_db":2.000916002412104}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":88480,"uncovered_tail_samples":0,"metrics":{"frame_count":1099,"rms_floor":3.252486753767026e-05,"log_rms_mae_db":6.012673559334026,"rms_envelope_nrmse":0.557177500474998,"envelope_correlation":0.7856857608112999,"rms_gain_error_db":0.7748287718081269,"reference_mean_power":2.6545935258029172e-05,"output_mean_power":3.173079169574673e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1099,"frequency_resolution_hz":0.18198362147406733,"power_floor":2.4345030346700565e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.623830647966854e-05,"output_power":2.097684368180066e-05,"reference_active":true,"delta_db":1.1119940791558816},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.69303691046191e-06,"output_power":5.8268434231653175e-06,"reference_active":true,"delta_db":1.9804969976584241},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.89764420950003e-08,"output_power":1.4939889745034785e-07,"reference_active":true,"delta_db":2.7684982765397597}],"active_band_count":3,"active_band_mae_db":1.9536631177846886}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"hnm","status":"measured","mos":{"status":"measured","value":2.724931478500366,"predicted_mos":2.724931478500366,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"713e18f86a60698c4ada61a7c82caa31fcce9a27f583dad3ddaecd25d628dec8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.796577418999732},"started_at":"2026-10-13T04:40:10.584352+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.649899968469981e-05,"native_input_samples":96000,"native_input_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","output_sha256":"9a424e1039e5b064f2664872e5b0b77a56da54bb28258ffe7e5aa26b6d452062","output_samples":96000,"length_delta_samples":0,"analysis_s":1.148630108999896,"synthesis_s":0.4933965119998902,"total_s":1.6420266209997862,"rtf":0.2736711034999644,"output_peak":0.7320522824075955,"output_rms":0.06987815673565852,"finished_at":"2026-10-13T04:40:12.232453+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"9a424e1039e5b064f2664872e5b0b77a56da54bb28258ffe7e5aa26b6d452062","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.833145929777386,"spectral_convergence":0.2440358739642662,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6806485594169915,"spectral_convergence":0.24543981422405914,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.095825703027619,"spectral_convergence":0.2634339705792726,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0005952164245969882,"delta_db":-0.005779397121914363,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.00303453240236147,"delta_db":-0.03132303141833542,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.000819166275467463,"delta_db":-0.22676274089642803,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00029770317397166224,"delta_db":-0.2597807987152415,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":0.00010121886130210401,"delta_db":-0.52988445490163,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.5119651073569814e-05,"delta_db":0.11442453099465384,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19465915900803388,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08563522301591411,"pitch":null,"output_rms_dbfs":-23.113171189434752,"output_peak":0.7320522824075955,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.869873397407332},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.833145929777386,"spectral_convergence":0.2440358739642662,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6806485594169915,"spectral_convergence":0.24543981422405914,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.095825703027619,"spectral_convergence":0.2634339705792726,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0005952164245969882,"delta_db":-0.005779397121914363,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.00303453240236147,"delta_db":-0.03132303141833542,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.000819166275467463,"delta_db":-0.22676274089642803,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00029770317397166224,"delta_db":-0.2597807987152415,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":0.00010121886130210401,"delta_db":-0.52988445490163,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.5119651073569814e-05,"delta_db":0.11442453099465384,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19465915900803388,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08563522301591411,"pitch":null,"output_rms_dbfs":-23.113171189434752,"output_peak":0.7320522824075955,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.869873397407332},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8416830537069871,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705705011713163,"rms_dbfs":-23.02753596641884,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.096170390751162,"mean":1.1076291402180989e-05,"zero_energy":false},"output":{"rms":0.06987815673565852,"rms_dbfs":-23.113171189434752,"peak":0.7320522824075955,"peak_dbfs":-2.70915801953484,"crest_factor_db":20.40401316989992,"mean":-3.541741333176098e-05,"zero_energy":false},"rms_difference_db":-0.0856352230159132},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":8.132749086323823e-07,"output_rms_dbfs":-60.89762626458732,"output_zero_energy":false,"output_peak":0.007487515047566794,"output_energy_fraction":2.886932179160819e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":118,"output_estimated_voiced_frames":139,"both_estimated_voiced_frames":115,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9745762711864406,"absolute_difference_cents_median":1.0189199464039642,"absolute_difference_cents_p95":4.642986153272284,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023539858364162597,"log_rms_mae_db":0.8327023672385901,"rms_envelope_nrmse":0.07236631472236475,"envelope_correlation":0.9945672619672976,"rms_gain_error_db":-0.08565677197169255,"reference_mean_power":0.004983368761309487,"output_mean_power":0.004886043715115598},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0209735395978665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009552928696150586,"output_power":0.000924131491024301,"reference_active":true,"delta_db":-0.14402766482597007},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014624552237472608,"output_power":0.00014952189353304585,"reference_active":true,"delta_db":0.0962221018937891},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.510798757145359e-05,"output_power":3.267648231196395e-05,"reference_active":true,"delta_db":1.1442339271425983}],"active_band_count":3,"active_band_mae_db":0.4614945646207858}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002083747898533542,"log_rms_mae_db":0.46155320167268626,"rms_envelope_nrmse":0.03996357516872756,"envelope_correlation":0.99827789074844,"rms_gain_error_db":-0.08572504616139405,"reference_mean_power":0.004999205734592124,"output_mean_power":0.004901494337180965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8665599913970183e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008900977663526511,"output_power":0.0008692634877687748,"reference_active":true,"delta_db":-0.10286273142827643},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.407197015766335e-05,"output_power":7.653290440371178e-05,"reference_active":true,"delta_db":0.14194298884798592},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8378799869222e-06,"output_power":2.3376920844346843e-06,"reference_active":true,"delta_db":1.0447015763549348}],"active_band_count":3,"active_band_mae_db":0.4298357655437324}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019850808969518194,"log_rms_mae_db":1.503535097929047,"rms_envelope_nrmse":0.10111875855724385,"envelope_correlation":0.9909997534726841,"rms_gain_error_db":-0.15626095585416835,"reference_mean_power":0.003121339390070273,"output_mean_power":0.003011028716450389},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3811041144776226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007780170156378303,"output_power":0.0007234942501148458,"reference_active":true,"delta_db":-0.31554011375700086},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012043207732421152,"output_power":0.00011949148762420211,"reference_active":true,"delta_db":-0.034052095658790235},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313510568503667e-05,"output_power":2.991579933823752e-05,"reference_active":true,"delta_db":1.1162912373823501}],"active_band_count":3,"active_band_mae_db":0.4886278155993804}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016710142967257498,"log_rms_mae_db":1.0196121553399973,"rms_envelope_nrmse":0.056966017777189744,"envelope_correlation":0.9971057091078648,"rms_gain_error_db":-0.15667764115720878,"reference_mean_power":0.003133268891618282,"output_mean_power":0.0030222466353254512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.268336261468391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007382464262949856,"output_power":0.0006934063809108333,"reference_active":true,"delta_db":-0.2721351947581208},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8108189032543396e-05,"output_power":5.881079949540728e-05,"reference_active":true,"delta_db":0.05219742729644219},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9161702335314512e-06,"output_power":2.3730917443526926e-06,"reference_active":true,"delta_db":0.9288043908346993}],"active_band_count":3,"active_band_mae_db":0.41771233762975407}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001248333872410341,"log_rms_mae_db":1.796065577341643,"rms_envelope_nrmse":0.17182887101058447,"envelope_correlation":0.9779177409598574,"rms_gain_error_db":-0.2845106232760861,"reference_mean_power":0.000373657483692677,"output_mean_power":0.00034996339622398034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.54909855359075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.931499876625758e-05,"output_power":8.51068595117297e-05,"reference_active":true,"delta_db":-0.20959831449960248},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4977670974468346e-05,"output_power":1.2099030640362208e-05,"reference_active":true,"delta_db":-0.9269370932455354},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3989606403687803e-06,"output_power":4.079643725956061e-06,"reference_active":true,"delta_db":0.7927610241640708}],"active_band_count":3,"active_band_mae_db":0.643098810636403}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010596065945345545,"log_rms_mae_db":1.2264724306655006,"rms_envelope_nrmse":0.09359826439090392,"envelope_correlation":0.9937238079220914,"rms_gain_error_db":-0.2840948922761304,"reference_mean_power":0.0003754054621930237,"output_mean_power":0.0003516341920087536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4355698091451528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.577569295489108e-05,"output_power":8.182756329757716e-05,"reference_active":true,"delta_db":-0.20464616542466452},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.833740893743027e-06,"output_power":6.582934704058577e-06,"reference_active":true,"delta_db":-0.755496555079918},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5275289003204504e-07,"output_power":2.4166753184382117e-07,"reference_active":true,"delta_db":-0.1947782346825826}],"active_band_count":3,"active_band_mae_db":0.384973651729055}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":1.915265359618906,"rms_envelope_nrmse":0.19458299725696793,"envelope_correlation":0.9732877266673842,"rms_gain_error_db":-0.4584364930662279,"reference_mean_power":5.7125880238454025e-05,"output_mean_power":5.140309345409915e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.2793193332580217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4391078150208962e-05,"output_power":1.1962058482822738e-05,"reference_active":true,"delta_db":-0.8028741025324739},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.344791688657988e-06,"output_power":3.386139368600187e-06,"reference_active":true,"delta_db":-1.082641299307782},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.199993664659816e-07,"output_power":6.802143876075175e-07,"reference_active":true,"delta_db":-0.811677031392397}],"active_band_count":3,"active_band_mae_db":0.8990641444108843}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":1.4794213728744692,"rms_envelope_nrmse":0.1293072835079365,"envelope_correlation":0.9890564316555315,"rms_gain_error_db":-0.45844468278649764,"reference_mean_power":5.7411509461641456e-05,"output_mean_power":5.166001134309517e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.057417179151536e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3970684782124768e-05,"output_power":1.1456458421206924e-05,"reference_active":true,"delta_db":-0.8616731059515366},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2411889958356618e-06,"output_power":1.835439701665411e-06,"reference_active":true,"delta_db":-0.8673835990817963},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.113666747131121e-08,"output_power":4.386785238782591e-08,"reference_active":true,"delta_db":-0.6658604861763272}],"active_band_count":3,"active_band_mae_db":0.7983057304032201}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.7588658332824707,"predicted_mos":3.7588658332824707,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a4731a841dd8dccc45b6904c888c93d635c24777aa094a84e04c7981e24dfafa","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.621424618000674},"started_at":"2026-10-13T04:43:24.645400+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3251999664353207e-05,"native_input_samples":96000,"native_input_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","output_sha256":"472d0c71af497112ca275b18eb4b412b7b927a3c187fcee1692793bb28ff067e","output_samples":96000,"length_delta_samples":0,"analysis_s":1.8740208050003275,"synthesis_s":0.08507949999966513,"total_s":1.9591003049999927,"rtf":0.3265167174999988,"output_peak":0.8516773440135988,"output_rms":0.08033933555306597,"finished_at":"2026-10-13T04:43:26.611756+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"472d0c71af497112ca275b18eb4b412b7b927a3c187fcee1692793bb28ff067e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.670805330059578,"spectral_convergence":0.6782589574738328,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.369538644330747,"spectral_convergence":0.34553677681450207,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.780112140233851,"spectral_convergence":0.3192076540345841,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.000742053205599016,"delta_db":0.951821990550682,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.004025421789537156,"delta_db":1.1958731075843485,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.001147048298454048,"delta_db":1.2353336653365496,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00038079191038011516,"delta_db":0.8092616878409075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":0.00012115708450161859,"delta_db":0.25098921277448055,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.7936548636173714e-05,"delta_db":0.449500781378785,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8154634075776256,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1261006779290403,"pitch":null,"output_rms_dbfs":-21.9014352884898,"output_peak":0.8516773440135988,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.940152038208058},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.670805330059578,"spectral_convergence":0.6782589574738328,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.369538644330747,"spectral_convergence":0.34553677681450207,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.780112140233851,"spectral_convergence":0.3192076540345841,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.000742053205599016,"delta_db":0.951821990550682,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.004025421789537156,"delta_db":1.1958731075843485,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.001147048298454048,"delta_db":1.2353336653365496,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00038079191038011516,"delta_db":0.8092616878409075,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":0.00012115708450161859,"delta_db":0.25098921277448055,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.7936548636173714e-05,"delta_db":0.449500781378785,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8154634075776256,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1261006779290403,"pitch":null,"output_rms_dbfs":-21.9014352884898,"output_peak":0.8516773440135988,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.940152038208058},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9048022475197548,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705705011713163,"rms_dbfs":-23.02753596641884,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.096170390751162,"mean":1.1076291402180989e-05,"zero_energy":false},"output":{"rms":0.08033933555306597,"rms_dbfs":-21.9014352884898,"peak":0.8516773440135988,"peak_dbfs":-1.394498110643123,"crest_factor_db":20.506937177846677,"mean":2.0325223866173728e-07,"zero_energy":false},"rms_difference_db":1.1261006779290383},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":8.350408670295169e-07,"output_rms_dbfs":-60.78292269556996,"output_zero_energy":false,"output_peak":0.008684820836323102,"output_energy_fraction":2.2425046290782113e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":118,"output_estimated_voiced_frames":159,"both_estimated_voiced_frames":115,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":44,"paired_voiced_fraction":0.9745762711864406,"absolute_difference_cents_median":4.686026141860963,"absolute_difference_cents_p95":12.685214816708536,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023539858364162597,"log_rms_mae_db":1.6601090358108006,"rms_envelope_nrmse":0.20466116694976066,"envelope_correlation":0.9791224739676611,"rms_gain_error_db":1.1269433190063525,"reference_mean_power":0.004983368761309487,"output_mean_power":0.006459774493998966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0209735395978665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009552928696150586,"output_power":0.0012384886149440669,"reference_active":true,"delta_db":1.127554824528047},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014624552237472608,"output_power":0.00017334075826978156,"reference_active":true,"delta_db":0.7381811394367737},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.510798757145359e-05,"output_power":1.680690370246555e-05,"reference_active":true,"delta_db":-1.7432419327466966}],"active_band_count":3,"active_band_mae_db":1.2029926322371722}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002083747898533542,"log_rms_mae_db":1.308871878980901,"rms_envelope_nrmse":0.16855055362631235,"envelope_correlation":0.9909245831258998,"rms_gain_error_db":1.1300140783338428,"reference_mean_power":0.004999205734592124,"output_mean_power":0.0064848870720371},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8665599913970183e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008900977663526511,"output_power":0.0011489192503736166,"reference_active":true,"delta_db":1.108517949891617},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.407197015766335e-05,"output_power":8.808749890123338e-05,"reference_active":true,"delta_db":0.7526038296868621},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8378799869222e-06,"output_power":1.020744530987707e-06,"reference_active":true,"delta_db":-2.5540008695131733}],"active_band_count":3,"active_band_mae_db":1.4717075496972172}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019850808969518194,"log_rms_mae_db":1.9620395429373012,"rms_envelope_nrmse":0.244302521046825,"envelope_correlation":0.9749023616811492,"rms_gain_error_db":1.319689632315776,"reference_mean_power":0.003121339390070273,"output_mean_power":0.004229703808038221},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3811041144776226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007780170156378303,"output_power":0.0010493111598848435,"reference_active":true,"delta_db":1.2991519645923417},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012043207732421152,"output_power":0.0001509285805936986,"reference_active":true,"delta_db":0.9802931021464636},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313510568503667e-05,"output_power":1.5082678661604493e-05,"reference_active":true,"delta_db":-1.8579300932381275}],"active_band_count":3,"active_band_mae_db":1.3791250533256445}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016710142967257498,"log_rms_mae_db":1.6686776741558105,"rms_envelope_nrmse":0.20477775732589865,"envelope_correlation":0.9875703865491741,"rms_gain_error_db":1.3213346612286951,"reference_mean_power":0.003133268891618282,"output_mean_power":0.004247477950043421},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.268336261468391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007382464262949856,"output_power":0.000987480372703936,"reference_active":true,"delta_db":1.2632711897880227},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8108189032543396e-05,"output_power":7.26720559145294e-05,"reference_active":true,"delta_db":0.9713010602009233},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9161702335314512e-06,"output_power":1.109282157227035e-06,"reference_active":true,"delta_db":-2.37392061967281}],"active_band_count":3,"active_band_mae_db":1.536164289887252}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001248333872410341,"log_rms_mae_db":1.5707796311505222,"rms_envelope_nrmse":0.16833384546262806,"envelope_correlation":0.9852101888608006,"rms_gain_error_db":0.6953922193121971,"reference_mean_power":0.000373657483692677,"output_mean_power":0.00043854373061838806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.54909855359075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.931499876625758e-05,"output_power":9.790543330296503e-05,"reference_active":true,"delta_db":0.3988239726146992},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4977670974468346e-05,"output_power":1.720920278762129e-05,"reference_active":true,"delta_db":0.6031646631264352},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3989606403687803e-06,"output_power":2.3290513516698956e-06,"reference_active":true,"delta_db":-1.6416707142059592}],"active_band_count":3,"active_band_mae_db":0.881219783315698}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010596065945345545,"log_rms_mae_db":1.1064393946376325,"rms_envelope_nrmse":0.11721404620262181,"envelope_correlation":0.9950705169439658,"rms_gain_error_db":0.695905424429301,"reference_mean_power":0.0003754054621930237,"output_mean_power":0.00044064731665724183},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4355698091451528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.577569295489108e-05,"output_power":9.298866627341515e-05,"reference_active":true,"delta_db":0.3506578344551029},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.833740893743027e-06,"output_power":9.24197986712924e-06,"reference_active":true,"delta_db":0.7179581524308764},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5275289003204504e-07,"output_power":1.5983196919313602e-07,"reference_active":true,"delta_db":-1.9903247994074196}],"active_band_count":3,"active_band_mae_db":1.0196469287644663}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":1.796242061380453,"rms_envelope_nrmse":0.21787040759162962,"envelope_correlation":0.9687621173306142,"rms_gain_error_db":0.40986621134498813,"reference_mean_power":5.7125880238454025e-05,"output_mean_power":6.277974194375409e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.2793193332580217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4391078150208962e-05,"output_power":1.5030031144334519e-05,"reference_active":true,"delta_db":0.18866548895152435},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.344791688657988e-06,"output_power":4.613616096387906e-06,"reference_active":true,"delta_db":0.2607249460564856},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.199993664659816e-07,"output_power":1.184683718594065e-06,"reference_active":true,"delta_db":1.5978890303246018}],"active_band_count":3,"active_band_mae_db":0.6824264884442038}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":1.095480108364869,"rms_envelope_nrmse":0.13867517212920477,"envelope_correlation":0.9875907254044554,"rms_gain_error_db":0.4098895264497992,"reference_mean_power":5.7411509461641456e-05,"output_mean_power":6.309397917704916e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.057417179151536e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3970684782124768e-05,"output_power":1.4885245510949606e-05,"reference_active":true,"delta_db":0.27538308251462923},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2411889958356618e-06,"output_power":2.403684890901029e-06,"reference_active":true,"delta_db":0.3039905215422927},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.113666747131121e-08,"output_power":8.069689947215883e-08,"reference_active":true,"delta_db":1.981244265772306}],"active_band_count":3,"active_band_mae_db":0.8535392899430759}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9857053756713867,"predicted_mos":3.9857053756713867,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"8fe85c12e9c50c5f7b72de5bbcd63d3bcddb475a68cef33bfb0289bd2b759543","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3509575440002664},"started_at":"2026-10-13T06:59:47.916998+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7690000933944248e-05,"native_input_samples":96000,"native_input_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","output_sha256":"9a5d179552f2567f29c94771b22e98ef947d01b8bb9be6d54da1426ce39f2337","output_samples":96000,"length_delta_samples":0,"analysis_s":2.03340111199941,"synthesis_s":12.345292341000459,"total_s":14.378693452999869,"rtf":2.3964489088333116,"output_peak":0.555993378162384,"output_rms":0.06774018491056871,"finished_at":"2026-10-13T07:00:02.305122+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"9a5d179552f2567f29c94771b22e98ef947d01b8bb9be6d54da1426ce39f2337","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.390584540566989,"spectral_convergence":0.23000237637234425,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.752218651442966,"spectral_convergence":0.24193496001196763,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.921932256689458,"spectral_convergence":0.2678423581402598,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0005624038954159603,"delta_db":-0.2520452637426247,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.0029554544327702412,"delta_db":-0.1459981455650465,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.0006935187267868618,"delta_db":-0.949901458394139,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00024420124776047613,"delta_db":-1.1201366611963508,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":9.868944940135942e-05,"delta_db":-0.639791667327032,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.4464899583141e-05,"delta_db":0.03269278189212541,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.523427663019553,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3555364774797789,"pitch":null,"output_rms_dbfs":-23.38307244389862,"output_peak":0.555993378162384,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.6882451495664705},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.390584540566989,"spectral_convergence":0.23000237637234425,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.752218651442966,"spectral_convergence":0.24193496001196763,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.921932256689458,"spectral_convergence":0.2678423581402598,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0005624038954159603,"delta_db":-0.2520452637426247,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.0029554544327702412,"delta_db":-0.1459981455650465,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.0006935187267868618,"delta_db":-0.949901458394139,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00024420124776047613,"delta_db":-1.1201366611963508,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":9.868944940135942e-05,"delta_db":-0.639791667327032,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.4464899583141e-05,"delta_db":0.03269278189212541,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.523427663019553,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3555364774797789,"pitch":null,"output_rms_dbfs":-23.38307244389862,"output_peak":0.555993378162384,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.6882451495664705},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.87803284056209,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705705011713163,"rms_dbfs":-23.02753596641884,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.096170390751162,"mean":1.1076291402180989e-05,"zero_energy":false},"output":{"rms":0.06774018491056871,"rms_dbfs":-23.38307244389862,"peak":0.555993378162384,"peak_dbfs":-5.0986076160085645,"crest_factor_db":18.284464827890055,"mean":-9.69008986057555e-06,"zero_energy":false},"rms_difference_db":-0.3555364774797809},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":6.94335211198632e-07,"output_rms_dbfs":-61.5843081019856,"output_zero_energy":false,"output_peak":0.009110725484788418,"output_energy_fraction":2.6227598280911573e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":118,"output_estimated_voiced_frames":119,"both_estimated_voiced_frames":104,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.8813559322033898,"absolute_difference_cents_median":2.675655557434202,"absolute_difference_cents_p95":10.823154561844905,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023539858364162597,"log_rms_mae_db":1.1115468550692618,"rms_envelope_nrmse":0.08549201667068798,"envelope_correlation":0.9938108478577129,"rms_gain_error_db":-0.3552618520567684,"reference_mean_power":0.004983368761309487,"output_mean_power":0.00459194671911906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0209735395978665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009552928696150586,"output_power":0.0009100025023370302,"reference_active":true,"delta_db":-0.21093949596441308},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014624552237472608,"output_power":0.00013190659865699593,"reference_active":true,"delta_db":-0.44816056318016856},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.510798757145359e-05,"output_power":2.5503976369546017e-05,"reference_active":true,"delta_db":0.06795992323643556}],"active_band_count":3,"active_band_mae_db":0.24235332746033908}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002083747898533542,"log_rms_mae_db":0.8673528910683658,"rms_envelope_nrmse":0.0715162235598221,"envelope_correlation":0.9959478051373414,"rms_gain_error_db":-0.3569054367310315,"reference_mean_power":0.004999205734592124,"output_mean_power":0.004604796754540803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8665599913970183e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008900977663526511,"output_power":0.0008458889255129881,"reference_active":true,"delta_db":-0.22124372044273138},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.407197015766335e-05,"output_power":6.531598978429398e-05,"reference_active":true,"delta_db":-0.5463438368028033},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8378799869222e-06,"output_power":1.689348878553325e-06,"reference_active":true,"delta_db":-0.36597800816338133}],"active_band_count":3,"active_band_mae_db":0.37785518846963867}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019850808969518194,"log_rms_mae_db":1.3508032135571102,"rms_envelope_nrmse":0.10789025685903005,"envelope_correlation":0.9916078996893668,"rms_gain_error_db":-0.4664404179671023,"reference_mean_power":0.003121339390070273,"output_mean_power":0.002803476701586457},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3811041144776226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007780170156378303,"output_power":0.0007317843762533868,"reference_active":true,"delta_db":-0.2660596236708585},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012043207732421152,"output_power":0.00010654552951150359,"reference_active":true,"delta_db":-0.5320694543587433},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313510568503667e-05,"output_power":2.1294760531327497e-05,"reference_active":true,"delta_db":-0.3599872715997521}],"active_band_count":3,"active_band_mae_db":0.38603878320978463}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016710142967257498,"log_rms_mae_db":1.0668442921599939,"rms_envelope_nrmse":0.08977285817087358,"envelope_correlation":0.9945503902795473,"rms_gain_error_db":-0.4674231509908051,"reference_mean_power":0.003133268891618282,"output_mean_power":0.002813554627562188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.268336261468391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007382464262949856,"output_power":0.0006936085111375739,"reference_active":true,"delta_db":-0.2708693966297231},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8108189032543396e-05,"output_power":5.025297813136366e-05,"reference_active":true,"delta_db":-0.6307553631467917},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9161702335314512e-06,"output_power":1.6149564193419713e-06,"reference_active":true,"delta_db":-0.7427328229377996}],"active_band_count":3,"active_band_mae_db":0.5481191942381048}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001248333872410341,"log_rms_mae_db":1.6218482903522922,"rms_envelope_nrmse":0.18708468198398562,"envelope_correlation":0.9784100158692591,"rms_gain_error_db":-0.8580231191470249,"reference_mean_power":0.000373657483692677,"output_mean_power":0.000306670056289963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.54909855359075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.931499876625758e-05,"output_power":8.724359197644098e-05,"reference_active":true,"delta_db":-0.10190858600443184},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4977670974468346e-05,"output_power":1.388545987142749e-05,"reference_active":true,"delta_db":-0.32884018200022935},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3989606403687803e-06,"output_power":4.570615254011017e-06,"reference_active":true,"delta_db":1.2862852919228276}],"active_band_count":3,"active_band_mae_db":0.5723446866424963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010596065945345545,"log_rms_mae_db":1.3575215910706684,"rms_envelope_nrmse":0.16452696850161652,"envelope_correlation":0.9840694172386639,"rms_gain_error_db":-0.8588956567391299,"reference_mean_power":0.0003754054621930237,"output_mean_power":0.00030804277118304013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4355698091451528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.577569295489108e-05,"output_power":8.446561733047518e-05,"reference_active":true,"delta_db":-0.06684274752161484},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.833740893743027e-06,"output_power":6.925834566544081e-06,"reference_active":true,"delta_db":-0.5349708927385415},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5275289003204504e-07,"output_power":3.336079982971652e-07,"reference_active":true,"delta_db":1.2054032424993546}],"active_band_count":3,"active_band_mae_db":0.6024056275865036}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":1.922117617622979,"rms_envelope_nrmse":0.23340233464261056,"envelope_correlation":0.9629305150549672,"rms_gain_error_db":-0.7640965184265331,"reference_mean_power":5.7125880238454025e-05,"output_mean_power":4.7909678136676186e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.2793193332580217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4391078150208962e-05,"output_power":1.3731311130467858e-05,"reference_active":true,"delta_db":-0.2038132390594323},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.344791688657988e-06,"output_power":4.125875439875121e-06,"reference_active":true,"delta_db":-0.2245284651407194},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.199993664659816e-07,"output_power":1.1284230377236466e-06,"reference_active":true,"delta_db":1.3865842722138049}],"active_band_count":3,"active_band_mae_db":0.6049753254713188}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":1.4803628295148,"rms_envelope_nrmse":0.19365116399278365,"envelope_correlation":0.9746829850227183,"rms_gain_error_db":-0.7643172831592299,"reference_mean_power":5.7411509461641456e-05,"output_mean_power":4.814677887266544e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.057417179151536e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3970684782124768e-05,"output_power":1.3499780045882107e-05,"reference_active":true,"delta_db":-0.148910013381937},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2411889958356618e-06,"output_power":2.17012785811364e-06,"reference_active":true,"delta_db":-0.1399315936059559},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.113666747131121e-08,"output_power":8.191105334535159e-08,"reference_active":true,"delta_db":2.0461008878411735}],"active_band_count":3,"active_band_mae_db":0.7783141649430222}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5365333557128906,"predicted_mos":3.5365333557128906,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"40e8ae8d97ca5ef8bd128b2fb19c4ae772aa67f535869a73a66efcb5d2510aa5","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.140109009000298},"started_at":"2026-10-13T07:29:29.493392+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.010783059002278605,"native_input_samples":264600,"native_input_sha256":"5345a488d4af8fc47652cbdcdfcbfc000f6d81d3e7dfce33876efdaf0c5b351a","output_sha256":"50ca49a0f45d55ca71b35d80e994f3783bda3fce1782808a0f98366493c7ad8b","output_samples":264704,"length_delta_samples":104,"analysis_s":0.2328612029996293,"synthesis_s":42.247371386998566,"total_s":42.480232589998195,"rtf":7.080038764999699,"output_peak":0.5404625535011292,"output_rms":0.061484404766449834,"finished_at":"2026-10-13T07:30:12.008197+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"50ca49a0f45d55ca71b35d80e994f3783bda3fce1782808a0f98366493c7ad8b","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.21954493240642,"output_peak":0.5320832733782036},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.14370589149262,"spectral_convergence":0.5045233261090859,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.848528281882446,"spectral_convergence":0.40618080362153697,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.665200726918469,"spectral_convergence":0.4875629244285967,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0005114901117664714,"delta_db":-0.664156042217303,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.0023036411748057705,"delta_db":-1.228092499188493,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.0006136905082047832,"delta_db":-1.4809893226007347,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00022677869626673227,"delta_db":-1.4415929054664245,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":0.00010211442074365605,"delta_db":-0.4916281463673277,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":2.8094151039235927e-05,"delta_db":-0.854918263457856,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0268961965496899,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1908769680795275,"pitch":null,"output_rms_dbfs":-24.21841293449837,"output_peak":0.5320832733782036,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.88581163343118},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.739508783915165,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705705011713163,"rms_dbfs":-23.02753596641884,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.096170390751162,"mean":1.1076291402180989e-05,"zero_energy":false},"output":{"rms":0.06152892866633264,"rms_dbfs":-24.21841293449837,"peak":0.5320832733782036,"peak_dbfs":-5.480407867689583,"crest_factor_db":18.738005066808785,"mean":-8.546327321842601e-05,"zero_energy":false},"rms_difference_db":-1.1908769680795324},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":7.337158542904646e-07,"output_rms_dbfs":-61.3447209649498,"output_zero_energy":false,"output_peak":0.007451492268115702,"output_energy_fraction":3.35931930620728e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":118,"output_estimated_voiced_frames":97,"both_estimated_voiced_frames":69,"reference_voiced_output_unvoiced_frames":49,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.5847457627118644,"absolute_difference_cents_median":8.088533612294755,"absolute_difference_cents_p95":32.45605928087655,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023539858364162597,"log_rms_mae_db":1.6735511870442064,"rms_envelope_nrmse":0.16835969579381016,"envelope_correlation":0.9860865932802491,"rms_gain_error_db":-1.1906751166553198,"reference_mean_power":0.004983368761309487,"output_mean_power":0.003788397259644541},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0209735395978665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009552928696150586,"output_power":0.0007674638724572417,"reference_active":true,"delta_db":-0.9507859546680026},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014624552237472608,"output_power":0.00010394614739338904,"reference_active":true,"delta_db":-1.4827418058787356},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.510798757145359e-05,"output_power":2.177499707099346e-05,"reference_active":true,"delta_db":-0.6185379973963356}],"active_band_count":3,"active_band_mae_db":1.0173552526476912}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002083747898533542,"log_rms_mae_db":1.4261236684594607,"rms_envelope_nrmse":0.14808135230669356,"envelope_correlation":0.9930642835328074,"rms_gain_error_db":-1.1891285165863539,"reference_mean_power":0.004999205734592124,"output_mean_power":0.003801790299147642},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8665599913970183e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008900977663526511,"output_power":0.0007121087874310249,"reference_active":true,"delta_db":-0.9689136619971722},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.407197015766335e-05,"output_power":5.2830403470106284e-05,"reference_active":true,"delta_db":-1.467699686773965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8378799869222e-06,"output_power":1.4960004722267448e-06,"reference_active":true,"delta_db":-0.8938541804336144}],"active_band_count":3,"active_band_mae_db":1.1101558430682505}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019850808969518194,"log_rms_mae_db":2.592119992313787,"rms_envelope_nrmse":0.22203744903413156,"envelope_correlation":0.9719614644664442,"rms_gain_error_db":-1.372208179435569,"reference_mean_power":0.003121339390070273,"output_mean_power":0.002275727067707592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3811041144776226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007780170156378303,"output_power":0.0006089893696167332,"reference_active":true,"delta_db":-1.063793835893804},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012043207732421152,"output_power":8.743735708977463e-05,"reference_active":true,"delta_db":-1.3904515551930756},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313510568503667e-05,"output_power":1.6518150336195607e-05,"reference_active":true,"delta_db":-1.4631007334276025}],"active_band_count":3,"active_band_mae_db":1.3057820415048274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016710142967257498,"log_rms_mae_db":2.2674895948393408,"rms_envelope_nrmse":0.19949711638653658,"envelope_correlation":0.9799449102881626,"rms_gain_error_db":-1.369663512890058,"reference_mean_power":0.003133268891618282,"output_mean_power":0.0022857636165031447},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.268336261468391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007382464262949856,"output_power":0.0005767654732315501,"reference_active":true,"delta_db":-1.072020988818066},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8108189032543396e-05,"output_power":4.2789742926674335e-05,"reference_active":true,"delta_db":-1.328976633435984},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9161702335314512e-06,"output_power":1.3275332809651574e-06,"reference_active":true,"delta_db":-1.593886718412347}],"active_band_count":3,"active_band_mae_db":1.3316281135554657}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001248333872410341,"log_rms_mae_db":2.80489662394546,"rms_envelope_nrmse":0.2531520956908945,"envelope_correlation":0.9593323895453739,"rms_gain_error_db":-1.1813232632358381,"reference_mean_power":0.000373657483692677,"output_mean_power":0.0002846697752660437},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.54909855359075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.931499876625758e-05,"output_power":7.848439886077307e-05,"reference_active":true,"delta_db":-0.5614106035475191},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4977670974468346e-05,"output_power":1.1928867789582797e-05,"reference_active":true,"delta_db":-0.9884506066496979},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3989606403687803e-06,"output_power":1.8954564674226413e-06,"reference_active":true,"delta_db":-2.5363232100391935}],"active_band_count":3,"active_band_mae_db":1.3620614734121368}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010596065945345545,"log_rms_mae_db":2.3959687548736373,"rms_envelope_nrmse":0.2159717745065587,"envelope_correlation":0.9726890775162861,"rms_gain_error_db":-1.1805343984548604,"reference_mean_power":0.0003754054621930237,"output_mean_power":0.0002860534220065386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4355698091451528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.577569295489108e-05,"output_power":7.580548321236163e-05,"reference_active":true,"delta_db":-0.5366361482713764},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.833740893743027e-06,"output_power":5.953253650662832e-06,"reference_active":true,"delta_db":-1.1921481590785723},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5275289003204504e-07,"output_power":1.355842967047414e-07,"reference_active":true,"delta_db":-2.7048673740333147}],"active_band_count":3,"active_band_mae_db":1.4778838937944212}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":3.770802375237312,"rms_envelope_nrmse":0.2413416370027422,"envelope_correlation":0.9684727761002409,"rms_gain_error_db":-1.320121292827134,"reference_mean_power":5.7125880238454025e-05,"output_mean_power":4.215225140354551e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.2793193332580217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4391078150208962e-05,"output_power":1.1953810108000885e-05,"reference_active":true,"delta_db":-0.805869790352242},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.344791688657988e-06,"output_power":3.389154685982739e-06,"reference_active":true,"delta_db":-1.0787756787861114},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.199993664659816e-07,"output_power":6.55842699283778e-07,"reference_active":true,"delta_db":-0.9701382842352001}],"active_band_count":3,"active_band_mae_db":0.9515945844578512}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":3.165489971206559,"rms_envelope_nrmse":0.2011208970270065,"envelope_correlation":0.9822024526227896,"rms_gain_error_db":-1.3202776228413673,"reference_mean_power":5.7411509461641456e-05,"output_mean_power":4.2361487644274424e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.057417179151536e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3970684782124768e-05,"output_power":1.162054744325725e-05,"reference_active":true,"delta_db":-0.799911057381169},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2411889958356618e-06,"output_power":1.737162822088039e-06,"reference_active":true,"delta_db":-1.1063795519828938},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.113666747131121e-08,"output_power":4.352108219882935e-08,"reference_active":true,"delta_db":-0.7003273602128244}],"active_band_count":3,"active_band_mae_db":0.8688726565256291}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"golf","status":"measured","mos":{"status":"measured","value":3.752755641937256,"predicted_mos":3.752755641937256,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"7be85fe5a28ca589fbdb1fe35d5ef0b2df6893ce6129701272aee86c3b6eeaf3","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.704303415999675},"started_at":"2026-10-13T05:03:00.583698+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.005834225999933551,"native_input_samples":144000,"native_input_sha256":"c275225b1eac9667eb04e32ba05cac4202173d169123215f69342e905bc61f99","output_sha256":"b04da7dec72a7219e09cc74770c254ce23e998c9a9bcd5e6d0ac7d7529220770","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5665349279997827,"synthesis_s":0.11219098399942595,"total_s":0.6787259119992086,"rtf":0.11312098533320143,"output_peak":0.666064977645874,"output_rms":0.06317807381155284,"finished_at":"2026-10-13T05:03:01.275766+00:00","analysis_stages":{"external_f0_s":0.0924368809992302},"evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"b04da7dec72a7219e09cc74770c254ce23e998c9a9bcd5e6d0ac7d7529220770","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.984392261454577,"output_peak":0.6362414014720222},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.104503080849662,"spectral_convergence":0.6514724668305275,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.485633123581045,"spectral_convergence":0.3545121458727186,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.716659372029517,"spectral_convergence":0.34498182732711136,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005949565861031008,"output_power":0.0005019694334171879,"delta_db":-0.7380800415906206,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003057105357576049,"output_power":0.002564116365106887,"delta_db":-0.7637267564323146,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008644726977840134,"output_power":0.0006243683794118825,"delta_db":-1.4131038123277229,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00031657191446191464,"output_power":0.00019328165554857027,"delta_db":-2.142817457255389,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011448821057333967,"output_power":8.412320452520566e-05,"delta_db":-1.3384495947980488,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.426267206244987e-05,"output_power":2.754564078594645e-05,"delta_db":-0.9476834983188108,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2239768601204843,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.958305216999792,"pitch":null,"output_rms_dbfs":-23.984392261454577,"output_peak":0.6362414014720222,"reference_rms_dbfs":-23.026087044454787,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.102265192153408},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8545405505537682,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07058227425183229,"rms_dbfs":-23.026087044454787,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.094721468787107,"mean":9.127252288971202e-06,"zero_energy":false},"output":{"rms":0.06320921355938294,"rms_dbfs":-23.984392261454577,"peak":0.6362414014720222,"peak_dbfs":-3.9275614783606496,"crest_factor_db":20.056830783093925,"mean":0.0009682244310078094,"zero_energy":false},"rms_difference_db":-0.9583052169997899},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":6.563343010578502e-07,"output_rms_dbfs":-61.828748982405976,"output_zero_energy":false,"output_peak":0.009381513145397676,"output_energy_fraction":2.852140035163116e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":117,"output_estimated_voiced_frames":136,"both_estimated_voiced_frames":107,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.9145299145299145,"absolute_difference_cents_median":4.4884744838624275,"absolute_difference_cents_p95":20.842485610265705,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00023539858364162597,"log_rms_mae_db":2.1380109698734366,"rms_envelope_nrmse":0.16487752785494547,"envelope_correlation":0.981716556879749,"rms_gain_error_db":-0.9594464502199955,"reference_mean_power":0.004985093898636872,"output_mean_power":0.0039969498394158065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.020162626545387e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009565770265115851,"output_power":0.000743126925587554,"reference_active":true,"delta_db":-1.096569491866556},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014623687860471706,"output_power":0.00011371774860403345,"reference_active":true,"delta_db":-1.0922865566909692},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5183124557968394e-05,"output_power":1.7115086093487267e-05,"reference_active":true,"delta_db":-1.6773052544634213}],"active_band_count":3,"active_band_mae_db":1.2887204343403156}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002083747898533542,"log_rms_mae_db":1.7680131180135241,"rms_envelope_nrmse":0.1371914894494408,"envelope_correlation":0.9905695662786023,"rms_gain_error_db":-0.9609827685721053,"reference_mean_power":0.005001422363398112,"output_mean_power":0.0040086233771324834},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.8656456649484845e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008911215375804883,"output_power":0.0006900618419625028,"reference_active":true,"delta_db":-1.11048927204854},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.416396884631306e-05,"output_power":5.614003834324842e-05,"reference_active":true,"delta_db":-1.2092025788362006},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8347868908187917e-06,"output_power":1.3493476436781893e-06,"reference_active":true,"delta_db":-1.3346177353017574}],"active_band_count":3,"active_band_mae_db":1.2181031953954993}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019850808969518194,"log_rms_mae_db":2.374994472425991,"rms_envelope_nrmse":0.18044346442246567,"envelope_correlation":0.9781710659584187,"rms_gain_error_db":-0.8800866862954857,"reference_mean_power":0.0031239752520412093,"output_mean_power":0.002550932201606422},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3818834531420018e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007783251418047421,"output_power":0.0006236467862693807,"reference_active":true,"delta_db":-0.9622237073757298},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012047407479685384,"output_power":9.571551740223694e-05,"reference_active":true,"delta_db":-0.9991124842278719},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3213701031939312e-05,"output_power":1.3830891944901083e-05,"reference_active":true,"delta_db":-2.248941985137181}],"active_band_count":3,"active_band_mae_db":1.4034260589135943}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00016710142967257498,"log_rms_mae_db":1.9678606230199776,"rms_envelope_nrmse":0.14887625074200228,"envelope_correlation":0.9871504010620286,"rms_gain_error_db":-0.8792796451832913,"reference_mean_power":0.003135605963723456,"output_mean_power":0.0025609052900711316},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2689552811718852e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007384240260612024,"output_power":0.0005924644528868532,"reference_active":true,"delta_db":-0.9564352051537475},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.810722525917767e-05,"output_power":4.449263549379799e-05,"reference_active":true,"delta_db":-1.159420057938909},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9109340417680606e-06,"output_power":1.1158130206309783e-06,"reference_active":true,"delta_db":-2.3365427213955314}],"active_band_count":3,"active_band_mae_db":1.4841326614960628}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001248333872410341,"log_rms_mae_db":2.426607354126992,"rms_envelope_nrmse":0.2626399002250018,"envelope_correlation":0.9744152953808355,"rms_gain_error_db":-1.9706790047702627,"reference_mean_power":0.00037421418257522647,"output_mean_power":0.00023771267684711772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.548049488633641e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.919511816868308e-05,"output_power":4.830359989471225e-05,"reference_active":true,"delta_db":-2.663615868273345},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4993014759630538e-05,"output_power":8.16566751452666e-06,"reference_active":true,"delta_db":-2.638972759608818},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3936331873719565e-06,"output_power":1.7541241777965517e-06,"reference_active":true,"delta_db":-2.866045636380915}],"active_band_count":3,"active_band_mae_db":2.722878088087692}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00010596065945345545,"log_rms_mae_db":2.217967404149201,"rms_envelope_nrmse":0.2290846644891094,"envelope_correlation":0.9891893327228488,"rms_gain_error_db":-1.9701105928761153,"reference_mean_power":0.0003759560776610683,"output_mean_power":0.0002388504427635175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.4344682333073332e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.562384960196187e-05,"output_power":4.607532025455053e-05,"reference_active":true,"delta_db":-2.691263870410654},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.84757128513147e-06,"output_power":4.021421022703227e-06,"reference_active":true,"delta_db":-2.9035572545180384},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5272608860511797e-07,"output_power":1.2926602874647223e-07,"reference_active":true,"delta_db":-2.911656691116174}],"active_band_count":3,"active_band_mae_db":2.835492605348289}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":2.815556193753625,"rms_envelope_nrmse":0.2155878881912084,"envelope_correlation":0.972271790391548,"rms_gain_error_db":-0.86095147196174,"reference_mean_power":5.722037989713482e-05,"output_mean_power":4.693054413385838e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.2761567475693474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4320262656482825e-05,"output_power":1.2351184291798658e-05,"reference_active":true,"delta_db":-0.6424238183349501},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.355776646803811e-06,"output_power":3.896571687485709e-06,"reference_active":true,"delta_db":-0.4838293151765521},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.16936764921555e-07,"output_power":1.0684921091939152e-06,"reference_active":true,"delta_db":1.1658287802833094}],"active_band_count":3,"active_band_mae_db":0.7640273045982706}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":2.4924972635506815,"rms_envelope_nrmse":0.17135807396430203,"envelope_correlation":0.9845672909581509,"rms_gain_error_db":-0.8609474841940195,"reference_mean_power":5.7506895609965925e-05,"output_mean_power":4.7165579560818004e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.0539966960869697e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.38965314970311e-05,"output_power":1.2437417661843253e-05,"reference_active":true,"delta_db":-0.48176197456740144},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2455647185873016e-06,"output_power":2.0020273666278035e-06,"reference_active":true,"delta_db":-0.4985556646538324},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.115826361649839e-08,"output_power":7.820915884670332e-08,"reference_active":true,"delta_db":1.8434181950882744}],"active_band_count":3,"active_band_mae_db":0.9412452781031694}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4794297218322754,"predicted_mos":3.4794297218322754,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"5b272201cff564a7ace258183c3fcccd8fe1575ce2c093508a3637ff8e969406","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.316332046997559},"started_at":"2026-10-13T07:45:20.449947+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00320961699981126,"native_input_samples":144000,"native_input_sha256":"c275225b1eac9667eb04e32ba05cac4202173d169123215f69342e905bc61f99","output_sha256":"175acd8c8fba7088ac88063474aaa0b39c05108d59d5cc2d3c0d4a63b0ca26d5","output_samples":144120,"length_delta_samples":120,"analysis_s":2.36027824300254,"synthesis_s":3.2642784519994166,"total_s":5.6245566950019565,"rtf":0.9374261158336594,"output_peak":0.6529192924499512,"output_rms":0.06787775565349019,"finished_at":"2026-10-13T07:45:26.087244+00:00","analysis_stages":{"external_f0_s":1.521621067000524},"evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"175acd8c8fba7088ac88063474aaa0b39c05108d59d5cc2d3c0d4a63b0ca26d5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.361343797604878,"output_peak":0.6508016010586518},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.713063521529804,"spectral_convergence":0.6734564553115479,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.318376645575763,"spectral_convergence":0.37022815235377726,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.5575053330755475,"spectral_convergence":0.3359897748548017,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0006721485541231894,"delta_db":0.5221242226764976,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.0028551103906242284,"delta_db":-0.2960117777071459,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.000652853430448287,"delta_db":-1.2123264799840698,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.0002997241923198025,"delta_db":-0.2303974622928098,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":0.00010634725614838877,"delta_db":-0.31523603530505606,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":2.8736976285757526e-05,"delta_db":-0.7566666969163526,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5554604458136553,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.33082151179279035,"pitch":null,"output_rms_dbfs":-23.35835747821163,"output_peak":0.6508016010586518,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.862981833393705},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.869454166281905,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705705011713163,"rms_dbfs":-23.02753596641884,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.096170390751162,"mean":1.1076291402180989e-05,"zero_energy":false},"output":{"rms":0.06793320837373762,"rms_dbfs":-23.35835747821163,"peak":0.6508016010586518,"peak_dbfs":-3.7310277462314687,"crest_factor_db":19.62732973198016,"mean":-7.143849751230856e-05,"zero_energy":false},"rms_difference_db":-0.3308215117927915},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":6.702494593462484e-07,"output_rms_dbfs":-61.73763527666658,"output_zero_energy":false,"output_peak":0.02108876627645768,"output_energy_fraction":2.5174120638130375e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":118,"output_estimated_voiced_frames":123,"both_estimated_voiced_frames":101,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.8559322033898306,"absolute_difference_cents_median":4.733217285353112,"absolute_difference_cents_p95":23.12559676583044,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023539858364162597,"log_rms_mae_db":1.8875593658503496,"rms_envelope_nrmse":0.16013147432172692,"envelope_correlation":0.97348983282941,"rms_gain_error_db":-0.33109872899799364,"reference_mean_power":0.004983368761309487,"output_mean_power":0.004617566435105135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0209735395978665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009552928696150586,"output_power":0.0008913450834333665,"reference_active":true,"delta_db":-0.30090662844438093},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014624552237472608,"output_power":0.000122360106763941,"reference_active":true,"delta_db":-0.7744273083447151},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.510798757145359e-05,"output_power":1.917446828744385e-05,"reference_active":true,"delta_db":-1.170885753024136}],"active_band_count":3,"active_band_mae_db":0.7487398966044106}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002083747898533542,"log_rms_mae_db":1.5072856434209263,"rms_envelope_nrmse":0.10819189050346797,"envelope_correlation":0.9877550254451186,"rms_gain_error_db":-0.3296498593591624,"reference_mean_power":0.004999205734592124,"output_mean_power":0.004633786542102482},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8665599913970183e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008900977663526511,"output_power":0.00082635274232674,"reference_active":true,"delta_db":-0.32272238524591446},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.407197015766335e-05,"output_power":6.195759190432444e-05,"reference_active":true,"delta_db":-0.7755936635743084},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8378799869222e-06,"output_power":1.4527952313703879e-06,"reference_active":true,"delta_db":-1.0211274300224835}],"active_band_count":3,"active_band_mae_db":0.7064811596142354}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019850808969518194,"log_rms_mae_db":2.2848070075745737,"rms_envelope_nrmse":0.1887153575675218,"envelope_correlation":0.9699634201197194,"rms_gain_error_db":-0.5663449737839471,"reference_mean_power":0.003121339390070273,"output_mean_power":0.00273972199063015},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3811041144776226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007780170156378303,"output_power":0.0006854749740754026,"reference_active":true,"delta_db":-0.5499749153115384},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012043207732421152,"output_power":9.72875837687916e-05,"reference_active":true,"delta_db":-0.9268476012872501},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313510568503667e-05,"output_power":1.5721043030820898e-05,"reference_active":true,"delta_db":-1.677901313403308}],"active_band_count":3,"active_band_mae_db":1.0515746100006989}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016710142967257498,"log_rms_mae_db":1.8941535428399168,"rms_envelope_nrmse":0.13319580759903554,"envelope_correlation":0.9857858215453341,"rms_gain_error_db":-0.5648039749771827,"reference_mean_power":0.003133268891618282,"output_mean_power":0.0027511690008388315},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.268336261468391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007382464262949856,"output_power":0.0006477554716108902,"reference_active":true,"delta_db":-0.5679026313593127},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8108189032543396e-05,"output_power":4.661661785971185e-05,"reference_active":true,"delta_db":-0.956965793819342},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9161702335314512e-06,"output_power":1.3947972269862363e-06,"reference_active":true,"delta_db":-1.3792301411911512}],"active_band_count":3,"active_band_mae_db":0.9680328554566019}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001248333872410341,"log_rms_mae_db":1.9621638798819285,"rms_envelope_nrmse":0.19403083140223437,"envelope_correlation":0.9715770591468693,"rms_gain_error_db":-0.22138124210423604,"reference_mean_power":0.000373657483692677,"output_mean_power":0.0003550876441730306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.54909855359075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.931499876625758e-05,"output_power":8.723668720012199e-05,"reference_active":true,"delta_db":-0.10225231611190594},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4977670974468346e-05,"output_power":1.3926690466864284e-05,"reference_active":true,"delta_db":-0.31596362770036307},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3989606403687803e-06,"output_power":2.8894893864618827e-06,"reference_active":true,"delta_db":-0.7052503194959924}],"active_band_count":3,"active_band_mae_db":0.37448875443608715}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010596065945345545,"log_rms_mae_db":1.3823738999804023,"rms_envelope_nrmse":0.11756077540521663,"envelope_correlation":0.9892925181900086,"rms_gain_error_db":-0.22097707552037385,"reference_mean_power":0.0003754054621930237,"output_mean_power":0.000356781954108136},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4355698091451528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.577569295489108e-05,"output_power":8.36785194198694e-05,"reference_active":true,"delta_db":-0.10750247943320067},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.833740893743027e-06,"output_power":7.23534243513054e-06,"reference_active":true,"delta_db":-0.3451011258550868},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5275289003204504e-07,"output_power":2.0980772994319243e-07,"reference_active":true,"delta_db":-0.808746452522129}],"active_band_count":3,"active_band_mae_db":0.4204500192701388}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":1.8926939358744923,"rms_envelope_nrmse":0.21610304966004396,"envelope_correlation":0.966263534821691,"rms_gain_error_db":-0.46473045139418445,"reference_mean_power":5.7125880238454025e-05,"output_mean_power":5.1328652120007205e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.2793193332580217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4391078150208962e-05,"output_power":1.361884582185886e-05,"reference_active":true,"delta_db":-0.23953028325620798},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.344791688657988e-06,"output_power":3.6033393860899395e-06,"reference_active":true,"delta_db":-0.8126379025834618},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.199993664659816e-07,"output_power":5.364795073687852e-07,"reference_active":true,"delta_db":-1.8426037955979209}],"active_band_count":3,"active_band_mae_db":0.9649239938125302}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":1.260129762796868,"rms_envelope_nrmse":0.13237833126409285,"envelope_correlation":0.9879757650867048,"rms_gain_error_db":-0.4647212049708589,"reference_mean_power":5.7411509461641456e-05,"output_mean_power":5.158540504936688e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.057417179151536e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3970684782124768e-05,"output_power":1.3134824895626666e-05,"reference_active":true,"delta_db":-0.26793406412087933},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2411889958356618e-06,"output_power":1.8526164447166862e-06,"reference_active":true,"delta_db":-0.8269296668946982},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.113666747131121e-08,"output_power":3.539048232730663e-08,"reference_active":true,"delta_db":-1.59845940488362}],"active_band_count":3,"active_band_mae_db":0.8977743786330659}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.4404892921447754,"predicted_mos":3.4404892921447754,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"5bf582c57c013142c0268e38a679cc49a20466630ac48fa3c508507976827479","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3386527719994774},"started_at":"2026-10-13T05:11:24.753101+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002657822000401211,"native_input_samples":144000,"native_input_sha256":"c275225b1eac9667eb04e32ba05cac4202173d169123215f69342e905bc61f99","output_sha256":"0d4a8674a2e211ece7d508208da0e428d431381fe48f8c9ffcc142de34f4d280","output_samples":144240,"length_delta_samples":240,"analysis_s":1.2375563590012462,"synthesis_s":0.14058346699857793,"total_s":1.378139825999824,"rtf":0.22968997099997068,"output_peak":0.673958420753479,"output_rms":0.06868061830606774,"finished_at":"2026-10-13T05:11:26.142578+00:00","analysis_stages":{"external_f0_s":0.17035088700140477},"evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"0d4a8674a2e211ece7d508208da0e428d431381fe48f8c9ffcc142de34f4d280","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.258943215601555,"output_peak":0.6851511317479195},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.696227987324949,"spectral_convergence":0.1838941186739194,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.390426477987259,"spectral_convergence":0.18420824741747205,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.831478841679076,"spectral_convergence":0.19680222057205007,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0005791642634481879,"delta_db":-0.12451090020556785,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.0029493801578686146,"delta_db":-0.15493328155023964,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.0007934876425991689,"delta_db":-0.36508170826598785,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.00027841336290239014,"delta_db":-0.5507146867795245,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":9.863199291634068e-05,"delta_db":-0.6423208435590044,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.037941648988208e-05,"delta_db":-0.5152831135164344,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3921407556461265,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.22435137160624324,"pitch":null,"output_rms_dbfs":-23.251887338025085,"output_peak":0.6851511317479195,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.639377768997094},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9270093423924823,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705705011713163,"rms_dbfs":-23.02753596641884,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.096170390751162,"mean":1.1076291402180989e-05,"zero_energy":false},"output":{"rms":0.06877104649650596,"rms_dbfs":-23.251887338025085,"peak":0.6851511317479195,"peak_dbfs":-3.2842724111845776,"crest_factor_db":19.967614926840508,"mean":0.00011228364198794166,"zero_energy":false},"rms_difference_db":-0.22435137160624663},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":7.247407185303846e-07,"output_rms_dbfs":-61.39817337776137,"output_zero_energy":false,"output_peak":0.0095716284103531,"output_energy_fraction":2.656155429585178e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":118,"output_estimated_voiced_frames":121,"both_estimated_voiced_frames":107,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9067796610169492,"absolute_difference_cents_median":0.8014438743132826,"absolute_difference_cents_p95":3.835265283062354,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023539858364162597,"log_rms_mae_db":0.8840510905707543,"rms_envelope_nrmse":0.08871048641013411,"envelope_correlation":0.9922183865743529,"rms_gain_error_db":-0.22446042659348728,"reference_mean_power":0.004983368761309487,"output_mean_power":0.004732351369230523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0209735395978665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009552928696150586,"output_power":0.0008832496276261842,"reference_active":true,"delta_db":-0.34053073137494755},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014624552237472608,"output_power":0.00011897560414629589,"reference_active":true,"delta_db":-0.8962465930544504},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.510798757145359e-05,"output_power":1.9237543307809226e-05,"reference_active":true,"delta_db":-1.1566229447130052}],"active_band_count":3,"active_band_mae_db":0.7978000897141344}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002083747898533542,"log_rms_mae_db":0.5949690613060414,"rms_envelope_nrmse":0.05313683098624632,"envelope_correlation":0.9974478088189622,"rms_gain_error_db":-0.22190159081940425,"reference_mean_power":0.004999205734592124,"output_mean_power":0.004750188574820775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8665599913970183e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008900977663526511,"output_power":0.000821221544734159,"reference_active":true,"delta_db":-0.34977376644890873},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.407197015766335e-05,"output_power":5.8835389152182255e-05,"reference_active":true,"delta_db":-1.0001526578636049},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8378799869222e-06,"output_power":1.1632846125910001e-06,"reference_active":true,"delta_db":-1.986311651830621}],"active_band_count":3,"active_band_mae_db":1.112079358714378}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019850808969518194,"log_rms_mae_db":1.3401254490738226,"rms_envelope_nrmse":0.11288409737379501,"envelope_correlation":0.9890103262433215,"rms_gain_error_db":-0.24737778225918017,"reference_mean_power":0.003121339390070273,"output_mean_power":0.002948514139730928},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3811041144776226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007780170156378303,"output_power":0.0007132883688748216,"reference_active":true,"delta_db":-0.3772395301693253},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012043207732421152,"output_power":9.805897018560555e-05,"reference_active":true,"delta_db":-0.8925484955443868},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313510568503667e-05,"output_power":1.9078085447660004e-05,"reference_active":true,"delta_db":-0.8373669817985967}],"active_band_count":3,"active_band_mae_db":0.7023850025041029}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016710142967257498,"log_rms_mae_db":0.8987866813935983,"rms_envelope_nrmse":0.06844204671203942,"envelope_correlation":0.9960930004011918,"rms_gain_error_db":-0.24561325674210327,"reference_mean_power":0.003133268891618282,"output_mean_power":0.002960985912810967},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.268336261468391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007382464262949856,"output_power":0.0006746253859015275,"reference_active":true,"delta_db":-0.3913867380850754},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8108189032543396e-05,"output_power":4.5678471803308714e-05,"reference_active":true,"delta_db":-1.045257747334343},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9161702335314512e-06,"output_power":1.2208012735790636e-06,"reference_active":true,"delta_db":-1.957891157166547}],"active_band_count":3,"active_band_mae_db":1.1315118808619884}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001248333872410341,"log_rms_mae_db":1.2185179422863974,"rms_envelope_nrmse":0.13174706028901717,"envelope_correlation":0.9894737512491545,"rms_gain_error_db":-0.5771840755970598,"reference_mean_power":0.000373657483692677,"output_mean_power":0.0003271562920181824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.54909855359075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.931499876625758e-05,"output_power":7.440929332408832e-05,"reference_active":true,"delta_db":-0.792972165652439},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4977670974468346e-05,"output_power":1.080073475956883e-05,"reference_active":true,"delta_db":-1.4199098492262743},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3989606403687803e-06,"output_power":2.5244847278492055e-06,"reference_active":true,"delta_db":-1.2917338779788488}],"active_band_count":3,"active_band_mae_db":1.1682052976191875}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010596065945345545,"log_rms_mae_db":0.9319013506666087,"rms_envelope_nrmse":0.08448466037953505,"envelope_correlation":0.9975085455146236,"rms_gain_error_db":-0.5770687127651745,"reference_mean_power":0.0003754054621930237,"output_mean_power":0.00032869546796148347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4355698091451528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.577569295489108e-05,"output_power":7.126237777054417e-05,"reference_active":true,"delta_db":-0.8050392616158274},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.833740893743027e-06,"output_power":5.583940216936654e-06,"reference_active":true,"delta_db":-1.4702844289243182},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5275289003204504e-07,"output_power":1.7585945098044358e-07,"reference_active":true,"delta_db":-1.575304171193678}],"active_band_count":3,"active_band_mae_db":1.2835426205779412}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":1.3571679835167554,"rms_envelope_nrmse":0.12911929020388968,"envelope_correlation":0.9892298370851137,"rms_gain_error_db":-0.44770246649872647,"reference_mean_power":5.7125880238454025e-05,"output_mean_power":5.1530298524363286e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.2793193332580217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4391078150208962e-05,"output_power":1.3441462915846168e-05,"reference_active":true,"delta_db":-0.2964679342195722},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.344791688657988e-06,"output_power":3.818048947092136e-06,"reference_active":true,"delta_db":-0.5612746728946293},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.199993664659816e-07,"output_power":6.615937428511432e-07,"reference_active":true,"delta_db":-0.9322212770548368}],"active_band_count":3,"active_band_mae_db":0.5966546280563461}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":1.018765148527337,"rms_envelope_nrmse":0.08232296065513939,"envelope_correlation":0.9965915055718929,"rms_gain_error_db":-0.447692515827972,"reference_mean_power":5.7411509461641456e-05,"output_mean_power":5.178806851448111e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.057417179151536e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3970684782124768e-05,"output_power":1.3097074810483409e-05,"reference_active":true,"delta_db":-0.280433856488543},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2411889958356618e-06,"output_power":1.9182565930870474e-06,"reference_active":true,"delta_db":-0.6757178187479727},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.113666747131121e-08,"output_power":3.74226015778656e-08,"reference_active":true,"delta_db":-1.3559844616272012}],"active_band_count":3,"active_band_mae_db":0.770712045621239}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0028","method":"magphase","status":"measured","mos":{"status":"measured","value":3.4610788822174072,"predicted_mos":3.4610788822174072,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289649,"input_sample_rate":48000,"input_seconds":6.034354166666667,"input_float32_sha256":"9c7e8e2d6eaef5d9d9de13a9c9907aa45906d729fe1df7fab49fd853438b1626","internal_samples":96550,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.660977535000711},"started_at":"2026-10-13T05:18:09.112625+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008573758001148235,"native_input_samples":288000,"native_input_sha256":"681434adde2a112d320a2d2ab556a85c99dfdd21721eaf568059e339a0f04e33","output_sha256":"7c9af17f39343230c50e83c36f80e020f84b3d705948bde9eecf563f1e7be144","output_samples":289649,"length_delta_samples":1649,"analysis_s":2.800526480999906,"synthesis_s":0.8341106130010303,"total_s":3.6346370940009365,"rtf":0.605772849000156,"output_peak":0.7577938950331582,"output_rms":0.0674402555014359,"finished_at":"2026-10-13T05:18:12.773642+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"7c9af17f39343230c50e83c36f80e020f84b3d705948bde9eecf563f1e7be144","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96550,"length_delta_samples":550,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.417313967139552,"output_peak":0.726995573400368},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9943034697048162,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.783741032448631,"spectral_convergence":0.8142204592490415,"spectral_convergence_status":"ok","amplitude_floor":1.7851227833840954e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.93250693025268,"spectral_convergence":0.6271794038361138,"spectral_convergence_status":"ok","amplitude_floor":1.2048994709482952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.478534826200837,"spectral_convergence":0.6624367613397311,"spectral_convergence_status":"ok","amplitude_floor":9.468117719284587e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005960090393190935,"output_power":0.0005445141134248346,"delta_db":-0.3924370563277116,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030564977634421338,"output_power":0.002766168894359464,"delta_db":-0.4334538897429653,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0008630746056010172,"output_power":0.0007925848658914341,"delta_db":-0.37002563085214824,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003160541845765363,"output_power":0.0003094661505758528,"delta_db":-0.09148391986904625,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001143536135409032,"output_power":0.0001166793163992578,"delta_db":0.08743983273953582,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.420642909107181e-05,"output_power":3.280525455627203e-05,"delta_db":-0.18164327217299897,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.259413933617401,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3807031303172928,"pitch":null,"output_rms_dbfs":-23.408239096736132,"output_peak":0.726995573400368,"reference_rms_dbfs":-23.02753596641884,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.064927596300715},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.4600388383504853,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705705011713163,"rms_dbfs":-23.02753596641884,"peak":0.71356201171875,"peak_dbfs":-2.9313655756676793,"crest_factor_db":20.096170390751162,"mean":1.1076291402180989e-05,"zero_energy":false},"output":{"rms":0.06754419734667898,"rms_dbfs":-23.408239096736132,"peak":0.726995573400368,"peak_dbfs":-2.769364670127264,"crest_factor_db":20.63887442660887,"mean":4.01656994713828e-05,"zero_energy":false},"rms_difference_db":-0.3807031303172934},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.541249318048358e-06,"threshold_dbfs":-52.56392309140451,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":104,"quiet_samples":16640,"reference_mean_power":6.842399530936606e-07,"output_mean_power":3.341080445440585e-06,"output_rms_dbfs":-54.76113067435698,"output_zero_energy":false,"output_peak":0.01996239967077188,"output_energy_fraction":0.00012693837405148355,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":118,"output_estimated_voiced_frames":128,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":40,"paired_voiced_fraction":0.7457627118644068,"absolute_difference_cents_median":12.724467181888533,"absolute_difference_cents_p95":65.35908391470234,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023539858364162597,"log_rms_mae_db":5.047636468705572,"rms_envelope_nrmse":0.47978509778479117,"envelope_correlation":0.7527642696092973,"rms_gain_error_db":-0.3802735867174062,"reference_mean_power":0.004983368761309487,"output_mean_power":0.004565576948308819},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.0209735395978665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009552928696150586,"output_power":0.0008855877430700961,"reference_active":true,"delta_db":-0.3290493906528236},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014624552237472608,"output_power":0.0001403804545043823,"reference_active":true,"delta_db":-0.17775933909103397},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.510798757145359e-05,"output_power":2.3669211507252335e-05,"reference_active":true,"delta_db":-0.25628114488965537}],"active_band_count":3,"active_band_mae_db":0.2543632915445043}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002083747898533542,"log_rms_mae_db":4.234837680516223,"rms_envelope_nrmse":0.40548559473136625,"envelope_correlation":0.8085900293864008,"rms_gain_error_db":-0.3744146484490465,"reference_mean_power":0.004999205734592124,"output_mean_power":0.00458626922209155},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8665599913970183e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008900977663526511,"output_power":0.0008249537535900159,"reference_active":true,"delta_db":-0.3300810825618684},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.407197015766335e-05,"output_power":7.297759437900701e-05,"reference_active":true,"delta_db":-0.0646435286419215},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8378799869222e-06,"output_power":1.619208566965528e-06,"reference_active":true,"delta_db":-0.5501435571472555}],"active_band_count":3,"active_band_mae_db":0.3149560561170151}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019850808969518194,"log_rms_mae_db":6.823954185781103,"rms_envelope_nrmse":0.5542124156873454,"envelope_correlation":0.7160834127943688,"rms_gain_error_db":-0.4705858853718598,"reference_mean_power":0.003121339390070273,"output_mean_power":0.002800801978121387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3811041144776226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007780170156378303,"output_power":0.000694641264932672,"reference_active":true,"delta_db":-0.49228516480416207},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012043207732421152,"output_power":0.00011209323709922254,"reference_active":true,"delta_db":-0.31162766356128074},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.313510568503667e-05,"output_power":2.067858781205644e-05,"reference_active":true,"delta_db":-0.48750611296853347}],"active_band_count":3,"active_band_mae_db":0.43047298044465876}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00016710142967257498,"log_rms_mae_db":6.109909589699247,"rms_envelope_nrmse":0.47256088674048624,"envelope_correlation":0.7758124425407268,"rms_gain_error_db":-0.4657914622465026,"reference_mean_power":0.003133268891618282,"output_mean_power":0.0028146119073873734},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.268336261468391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007382464262949856,"output_power":0.0006618169730966774,"reference_active":true,"delta_db":-0.47463452334582945},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8108189032543396e-05,"output_power":5.474751913374069e-05,"reference_active":true,"delta_db":-0.25872896626765113},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9161702335314512e-06,"output_power":1.6630960543105126e-06,"reference_active":true,"delta_db":-0.6151675620197844}],"active_band_count":3,"active_band_mae_db":0.44951035054442173}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001248333872410341,"log_rms_mae_db":6.5069619297456915,"rms_envelope_nrmse":0.5095480315542577,"envelope_correlation":0.8060317331217026,"rms_gain_error_db":-0.06719663499143252,"reference_mean_power":0.000373657483692677,"output_mean_power":0.000367920529274575},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.54909855359075e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.931499876625758e-05,"output_power":8.67962642758105e-05,"reference_active":true,"delta_db":-0.12423363085031953},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4977670974468346e-05,"output_power":1.4832136927002414e-05,"reference_active":true,"delta_db":-0.042405597590859614},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.3989606403687803e-06,"output_power":3.972113373161169e-06,"reference_active":true,"delta_db":0.6767550022913453}],"active_band_count":3,"active_band_mae_db":0.2811314102441748}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00010596065945345545,"log_rms_mae_db":5.577483357924363,"rms_envelope_nrmse":0.42590685907308434,"envelope_correlation":0.856557577601443,"rms_gain_error_db":-0.06603453722795774,"reference_mean_power":0.0003754054621930237,"output_mean_power":0.00036974059319121134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4355698091451528e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":8.577569295489108e-05,"output_power":8.371645249401954e-05,"reference_active":true,"delta_db":-0.10553418553779048},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.833740893743027e-06,"output_power":7.589406660739487e-06,"reference_active":true,"delta_db":-0.13761378755223244},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5275289003204504e-07,"output_power":3.008843899609935e-07,"reference_active":true,"delta_db":0.7570352679050727}],"active_band_count":3,"active_band_mae_db":0.3333944136650319}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.3292043050207136e-05,"log_rms_mae_db":6.279170983985525,"rms_envelope_nrmse":0.6666524149092342,"envelope_correlation":0.6866272785138046,"rms_gain_error_db":0.16725232469634338,"reference_mean_power":5.7125880238454025e-05,"output_mean_power":5.936878178538363e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.2793193332580217e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4391078150208962e-05,"output_power":1.427186328306019e-05,"reference_active":true,"delta_db":-0.03612654872441448},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.344791688657988e-06,"output_power":4.775999410945956e-06,"reference_active":true,"delta_db":0.410953055525859},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.199993664659816e-07,"output_power":1.3566597380339835e-06,"reference_active":true,"delta_db":2.1865741966783196}],"active_band_count":3,"active_band_mae_db":0.877884600309531}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.870517765122227e-05,"log_rms_mae_db":5.32929715239229,"rms_envelope_nrmse":0.5573896820078009,"envelope_correlation":0.7670689659215649,"rms_gain_error_db":0.16737284164132715,"reference_mean_power":5.7411509461641456e-05,"output_mean_power":5.966728125650653e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.057417179151536e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3970684782124768e-05,"output_power":1.4194887558906572e-05,"reference_active":true,"delta_db":0.06914262865196417},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2411889958356618e-06,"output_power":2.601922234456047e-06,"reference_active":true,"delta_db":0.6481583091507265},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.113666747131121e-08,"output_power":7.757920458231078e-08,"reference_active":true,"delta_db":1.8101290042797196}],"active_band_count":3,"active_band_mae_db":0.8424766473608033}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"hnm","status":"measured","mos":{"status":"measured","value":3.4814248085021973,"predicted_mos":3.4814248085021973,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b0e18d8e6b7bad0d11e4dca03d2a5211202dfb8a4bdd6f204ed31dac725cf6f9","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.462128871000459},"started_at":"2026-10-13T04:40:12.234064+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.672100006544497e-05,"native_input_samples":96000,"native_input_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","output_sha256":"77a6144359ef7ed7f5e5e2d3ebf7e1f886434af4e0356d352482106eba90e317","output_samples":96000,"length_delta_samples":0,"analysis_s":1.2162070530002893,"synthesis_s":0.49970462800047244,"total_s":1.7159116810007617,"rtf":0.28598528016679364,"output_peak":0.5531156538892668,"output_rms":0.06512387879857215,"finished_at":"2026-10-13T04:40:13.956469+00:00","evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"77a6144359ef7ed7f5e5e2d3ebf7e1f886434af4e0356d352482106eba90e317","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.796677355064137,"spectral_convergence":0.22308577506136787,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.576366471767581,"spectral_convergence":0.2184513596521425,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.977044345000253,"spectral_convergence":0.2252311395127544,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0006931013195788181,"delta_db":0.0001879043667899488,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0025246795063790066,"delta_db":-0.059099427687188275,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0007257758369128023,"delta_db":0.013839905287702847,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00019165605243009015,"delta_db":-0.7287986685621737,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":5.924320703579346e-05,"delta_db":-0.3944172572758504,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":4.666366743460426e-05,"delta_db":0.22313345898151582,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23657943702687012,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07166509429280052,"pitch":null,"output_rms_dbfs":-23.725194812669056,"output_peak":0.5531156538892668,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.78336272394399},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.796677355064137,"spectral_convergence":0.22308577506136787,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.576366471767581,"spectral_convergence":0.2184513596521425,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.977044345000253,"spectral_convergence":0.2252311395127544,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0006931013195788181,"delta_db":0.0001879043667899488,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0025246795063790066,"delta_db":-0.059099427687188275,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0007257758369128023,"delta_db":0.013839905287702847,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00019165605243009015,"delta_db":-0.7287986685621737,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":5.924320703579346e-05,"delta_db":-0.3944172572758504,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":4.666366743460426e-05,"delta_db":0.22313345898151582,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23657943702687012,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07166509429280052,"pitch":null,"output_rms_dbfs":-23.725194812669056,"output_peak":0.5531156538892668,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.78336272394399},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8642460730425842,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06566342232761468,"rms_dbfs":-23.65352971837625,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.42146394315191,"mean":4.2438507080078126e-07,"zero_energy":false},"output":{"rms":0.06512387879857215,"rms_dbfs":-23.725194812669052,"peak":0.5531156538892668,"peak_dbfs":-5.143681005213462,"crest_factor_db":18.581513807455593,"mean":-1.7794465885105043e-05,"zero_energy":false},"rms_difference_db":-0.07166509429280055},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":148,"quiet_samples":23680,"reference_mean_power":6.957889323726901e-07,"output_mean_power":8.182367250217683e-07,"output_rms_dbfs":-60.87121031906375,"output_zero_energy":false,"output_peak":0.010414025180922864,"output_energy_fraction":4.758925591066909e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":139,"both_estimated_voiced_frames":104,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9369369369369369,"absolute_difference_cents_median":1.2307263427560426,"absolute_difference_cents_p95":6.50398630140461,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019088238002709922,"log_rms_mae_db":0.8849456620787711,"rms_envelope_nrmse":0.07744855360401819,"envelope_correlation":0.9940124369203944,"rms_gain_error_db":-0.07166508162936622,"reference_mean_power":0.004315280923443156,"output_mean_power":0.004244656643115801},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.876167788850743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007125355736933842,"output_power":0.0006984941615485211,"reference_active":true,"delta_db":-0.08643771229908302},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011942531791187452,"output_power":0.0001109282901980516,"reference_active":true,"delta_db":-0.3205408704663779},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5392584995669358e-05,"output_power":2.807209094278901e-05,"reference_active":true,"delta_db":0.43567847332419063}],"active_band_count":3,"active_band_mae_db":0.2808856853632172}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018007320980187037,"log_rms_mae_db":0.5107686789945195,"rms_envelope_nrmse":0.043208943509798106,"envelope_correlation":0.9980350160705832,"rms_gain_error_db":-0.07166518145525774,"reference_mean_power":0.004336982504671315,"output_mean_power":0.004266002956233191},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7431428852385876e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439047386291531,"output_power":0.0006286026948083954,"reference_active":true,"delta_db":-0.1044538266932895},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.115354721039442e-05,"output_power":4.743333912160706e-05,"reference_active":true,"delta_db":-0.32792056607288966},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.965415989281464e-06,"output_power":1.7982320273484156e-06,"reference_active":true,"delta_db":-0.38608756468716177}],"active_band_count":3,"active_band_mae_db":0.272820652484447}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017279669295308916,"log_rms_mae_db":1.454399354000608,"rms_envelope_nrmse":0.09933661723674718,"envelope_correlation":0.99189901986865,"rms_gain_error_db":-0.09268112603586812,"reference_mean_power":0.0026069520032741206,"output_mean_power":0.002551907467857856},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2087875823984064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005529275902426358,"output_power":0.0005415473749119597,"reference_active":true,"delta_db":-0.09031806127787702},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648535673845477e-05,"output_power":8.440057230359387e-05,"reference_active":true,"delta_db":-0.5811601541576945},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7058636340876773e-05,"output_power":2.0504656827785994e-05,"reference_active":true,"delta_db":0.7990819427029319}],"active_band_count":3,"active_band_mae_db":0.49018671937950115}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014230358058689408,"log_rms_mae_db":0.9902401731353059,"rms_envelope_nrmse":0.05706375342892736,"envelope_correlation":0.9971924295425334,"rms_gain_error_db":-0.09268115114972614,"reference_mean_power":0.0026200630509491097,"output_mean_power":0.0025647416672475313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1215946456987494e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005050917134712818,"output_power":0.0004957072658012767,"reference_active":true,"delta_db":-0.0814495894196122},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.798176008445663e-05,"output_power":3.2364971051381364e-05,"reference_active":true,"delta_db":-0.6949986328195225},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3776930382412419e-06,"output_power":1.3897713613858116e-06,"reference_active":true,"delta_db":0.03790894220098738}],"active_band_count":3,"active_band_mae_db":0.27145238814670736}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":1.8410496877845295,"rms_envelope_nrmse":0.19464143215326055,"envelope_correlation":0.975648269920408,"rms_gain_error_db":-0.7053475964466328,"reference_mean_power":0.0002448429483315043,"output_mean_power":0.00020813870151748843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7588735706755046e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.377351846027071e-05,"output_power":8.041360489562802e-05,"reference_active":true,"delta_db":-0.6675067967910471},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.692404819271502e-06,"output_power":8.478253463211778e-06,"reference_active":true,"delta_db":-0.5812514891136014},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.811201191302258e-06,"output_power":2.3333743361141606e-06,"reference_active":true,"delta_db":-2.130774586345375}],"active_band_count":3,"active_band_mae_db":1.1265109574166745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":1.3327852351967744,"rms_envelope_nrmse":0.12476732239539744,"envelope_correlation":0.9925971655892889,"rms_gain_error_db":-0.7053530289991562,"reference_mean_power":0.00024607384660115493,"output_mean_power":0.00020918481496208407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6710797473991248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.068314304092383e-05,"output_power":7.721453608694332e-05,"reference_active":true,"delta_db":-0.698274976831172},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.419207768284794e-06,"output_power":3.9326637131520285e-06,"reference_active":true,"delta_db":-0.5065760937270322},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6860314835877605e-07,"output_power":1.288906909590016e-07,"reference_active":true,"delta_db":-3.188895469597055}],"active_band_count":3,"active_band_mae_db":1.4645821800517531}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":1.9508989926183278,"rms_envelope_nrmse":0.22545742576309138,"envelope_correlation":0.9658909846015498,"rms_gain_error_db":-0.13176606055559403,"reference_mean_power":6.169396290931911e-05,"output_mean_power":5.9850262835575954e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.220333375078007e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9572001812371347e-05,"output_power":1.703149674002248e-05,"reference_active":true,"delta_db":-0.6038243155659603},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.501529531199527e-06,"output_power":7.116567033313811e-06,"reference_active":true,"delta_db":-0.7722652326309577},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.287934963899871e-07,"output_power":1.5857836993470105e-06,"reference_active":true,"delta_db":2.323247836019651}],"active_band_count":3,"active_band_mae_db":1.233112461405523}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":1.4531329016446026,"rms_envelope_nrmse":0.1351462811947433,"envelope_correlation":0.9871335498650777,"rms_gain_error_db":-0.1317659621058023,"reference_mean_power":6.200418008997281e-05,"output_mean_power":6.01512106600667e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.99689742660592e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0058245625049688e-05,"output_power":1.7724240481149725e-05,"reference_active":true,"delta_db":-0.5372531132676818},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3812703887943674e-06,"output_power":3.834399202464626e-06,"reference_active":true,"delta_db":-0.5790273070687055},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0248947956398812e-07,"output_power":1.374194429099534e-07,"reference_active":true,"delta_db":1.2736889579184991}],"active_band_count":3,"active_band_mae_db":0.7966564594182954}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"world_harvest","status":"measured","mos":{"status":"measured","value":4.03670072555542,"predicted_mos":4.03670072555542,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ff29067801eacc2afb72aa0844dd28785771c17352a38343975aec7ab08f23e4","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3560792809985287},"started_at":"2026-10-13T04:43:26.613172+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.264099930471275e-05,"native_input_samples":96000,"native_input_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","output_sha256":"f523cf0efd7dc793bcadf34212b1e25e86c3966684918eaef66d34ab7e856db9","output_samples":96000,"length_delta_samples":0,"analysis_s":1.8835193009999784,"synthesis_s":0.0798129559998415,"total_s":1.96333225699982,"rtf":0.3272220428333033,"output_peak":0.8232670418770927,"output_rms":0.0742541323246826,"finished_at":"2026-10-13T04:43:28.584238+00:00","evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"f523cf0efd7dc793bcadf34212b1e25e86c3966684918eaef66d34ab7e856db9","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.066528782617265,"spectral_convergence":0.8188246778939154,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.549659087903281,"spectral_convergence":0.4041573943195752,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.747153853982129,"spectral_convergence":0.31667224748065564,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0007883543747727669,"delta_db":0.5594354671850936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0034739304253139074,"delta_db":1.3270491877508384,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0008798195862255237,"delta_db":0.8497511114455601,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.0002619519164754427,"delta_db":0.6281917462371779,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":6.187330013194108e-05,"delta_db":-0.20577005474271953,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":4.77465643718905e-05,"delta_db":0.3227660478649181,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6488272692043846,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0679422709879574,"pitch":null,"output_rms_dbfs":-22.585587447388296,"output_peak":0.8232670418770927,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.121113908167558},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.066528782617265,"spectral_convergence":0.8188246778939154,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.549659087903281,"spectral_convergence":0.4041573943195752,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.747153853982129,"spectral_convergence":0.31667224748065564,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0007883543747727669,"delta_db":0.5594354671850936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0034739304253139074,"delta_db":1.3270491877508384,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0008798195862255237,"delta_db":0.8497511114455601,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.0002619519164754427,"delta_db":0.6281917462371779,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":6.187330013194108e-05,"delta_db":-0.20577005474271953,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":4.77465643718905e-05,"delta_db":0.3227660478649181,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6488272692043846,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0679422709879574,"pitch":null,"output_rms_dbfs":-22.585587447388296,"output_peak":0.8232670418770927,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.121113908167558},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8957996802797982,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06566342232761468,"rms_dbfs":-23.65352971837625,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.42146394315191,"mean":4.2438507080078126e-07,"zero_energy":false},"output":{"rms":0.0742541323246826,"rms_dbfs":-22.585587447388296,"peak":0.8232670418770927,"peak_dbfs":-1.6891854099140806,"crest_factor_db":20.896402037474214,"mean":-6.039990787035099e-11,"zero_energy":false},"rms_difference_db":1.067942270987956},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":148,"quiet_samples":23680,"reference_mean_power":6.957889323726901e-07,"output_mean_power":6.828926946327419e-07,"output_rms_dbfs":-61.6564753319997,"output_zero_energy":false,"output_peak":0.0073722343648674,"output_energy_fraction":3.0550735945528664e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":133,"both_estimated_voiced_frames":94,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.8468468468468469,"absolute_difference_cents_median":5.793739604527724,"absolute_difference_cents_p95":19.98465151065361,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019088238002709922,"log_rms_mae_db":1.9166411178239644,"rms_envelope_nrmse":0.22319637384142035,"envelope_correlation":0.971640256269582,"rms_gain_error_db":1.0679423981838143,"reference_mean_power":0.004315280923443156,"output_mean_power":0.00551827466623005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.876167788850743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007125355736933842,"output_power":0.0009572704859828033,"reference_active":true,"delta_db":1.2822811760990072},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011942531791187452,"output_power":0.00014022609826411345,"reference_active":true,"delta_db":0.6973244407554096},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5392584995669358e-05,"output_power":1.996711992845515e-05,"reference_active":true,"delta_db":-1.0439148838056531}],"active_band_count":3,"active_band_mae_db":1.0078401668866899}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018007320980187037,"log_rms_mae_db":1.3050683288036622,"rms_envelope_nrmse":0.16703624049086685,"envelope_correlation":0.9901045886862897,"rms_gain_error_db":1.0679432185325852,"reference_mean_power":0.004336982504671315,"output_mean_power":0.005546027159921312},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7431428852385876e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439047386291531,"output_power":0.0008613521734685056,"reference_active":true,"delta_db":1.263591327516098},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.115354721039442e-05,"output_power":5.951366941880371e-05,"reference_active":true,"delta_db":0.6574097330328417},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.965415989281464e-06,"output_power":1.3198545433071382e-06,"reference_active":true,"delta_db":-1.7292841313241682}],"active_band_count":3,"active_band_mae_db":1.2167617306243692}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017279669295308916,"log_rms_mae_db":2.170792246005397,"rms_envelope_nrmse":0.2632068035172689,"envelope_correlation":0.9719936887827284,"rms_gain_error_db":1.3665406149792125,"reference_mean_power":0.0026069520032741206,"output_mean_power":0.0035709773612169495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2087875823984064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005529275902426358,"output_power":0.0008039330819201097,"reference_active":true,"delta_db":1.6255163917676376},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648535673845477e-05,"output_power":0.00013518379117077788,"reference_active":true,"delta_db":1.4646321496217678},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7058636340876773e-05,"output_power":1.0508503965878904e-05,"reference_active":true,"delta_db":-2.104034184598255}],"active_band_count":3,"active_band_mae_db":1.731394241995887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014230358058689408,"log_rms_mae_db":1.706222352542332,"rms_envelope_nrmse":0.21176688715949543,"envelope_correlation":0.9882781419066393,"rms_gain_error_db":1.3665406595950638,"reference_mean_power":0.0026200630509491097,"output_mean_power":0.0035889367829815993},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1215946456987494e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005050917134712818,"output_power":0.0007290149116194546,"reference_active":true,"delta_db":1.5936616813511595},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.798176008445663e-05,"output_power":5.274876978062374e-05,"reference_active":true,"delta_db":1.4263724918400433},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3776930382412419e-06,"output_power":8.725344001001297e-07,"reference_active":true,"delta_db":-1.983699055920494}],"active_band_count":3,"active_band_mae_db":1.6679110763705653}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":1.7854187102096146,"rms_envelope_nrmse":0.1777100351085631,"envelope_correlation":0.9819905168396088,"rms_gain_error_db":0.5570522258332792,"reference_mean_power":0.0002448429483315043,"output_mean_power":0.00027835102384927117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7588735706755046e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.377351846027071e-05,"output_power":0.00010480306845780809,"reference_active":true,"delta_db":0.48293786843527037},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.692404819271502e-06,"output_power":9.358974124970742e-06,"reference_active":true,"delta_db":-0.15203298298279908},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.811201191302258e-06,"output_power":2.2374840005597013e-06,"reference_active":true,"delta_db":-2.3130193703812676}],"active_band_count":3,"active_band_mae_db":0.9826634072664456}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":1.1174054171754946,"rms_envelope_nrmse":0.10430099198694907,"envelope_correlation":0.9955386488114757,"rms_gain_error_db":0.5570548027775932,"reference_mean_power":0.00024607384660115493,"output_mean_power":0.0002797505431669672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6710797473991248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.068314304092383e-05,"output_power":0.00010022149380585603,"reference_active":true,"delta_db":0.43434307447635556},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.419207768284794e-06,"output_power":4.236025145521542e-06,"reference_active":true,"delta_db":-0.18385890857865222},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6860314835877605e-07,"output_power":1.3997994742966072e-07,"reference_active":true,"delta_db":-2.8304527275178937}],"active_band_count":3,"active_band_mae_db":1.149551570190967}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":1.9456348535602237,"rms_envelope_nrmse":0.21630975284151957,"envelope_correlation":0.9687294863370973,"rms_gain_error_db":-0.25644339959193235,"reference_mean_power":6.169396290931911e-05,"output_mean_power":5.8156508882759814e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.220333375078007e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9572001812371347e-05,"output_power":1.778487630839877e-05,"reference_active":true,"delta_db":-0.4158439823542681},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.501529531199527e-06,"output_power":5.998025984720855e-06,"reference_active":true,"delta_db":-1.5148872485457567},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.287934963899871e-07,"output_power":4.850891471533276e-07,"reference_active":true,"delta_db":-2.820976074070452}],"active_band_count":3,"active_band_mae_db":1.583902434990159}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":1.2481970493561059,"rms_envelope_nrmse":0.12116096209764102,"envelope_correlation":0.9902388350657403,"rms_gain_error_db":-0.2564425323120266,"reference_mean_power":6.200418008997281e-05,"output_mean_power":5.8448950273682964e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.99689742660592e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0058245625049688e-05,"output_power":1.7739188512324573e-05,"reference_active":true,"delta_db":-0.5335919625987993},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3812703887943674e-06,"output_power":2.949075730254116e-06,"reference_active":true,"delta_db":-1.7191413127839295},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0248947956398812e-07,"output_power":5.9346914818077696e-08,"reference_active":true,"delta_db":-2.372811409726964}],"active_band_count":3,"active_band_mae_db":1.5418482283698978}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"facodec","status":"measured","mos":{"status":"measured","value":4.163525104522705,"predicted_mos":4.163525104522705,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4edf5a26302b2467f556e7680b5c6c0819c9ac885abbd730fc21ab09c6417eea","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.856265689999418},"started_at":"2026-10-13T07:00:02.306152+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.0220999178709462e-05,"native_input_samples":96000,"native_input_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","output_sha256":"948aa635663be12e5b2ded6369249482799ccdd2781034089cf946fa0f4d7fbf","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5020885439989797,"synthesis_s":5.529162021000957,"total_s":7.031250564999937,"rtf":1.171875094166656,"output_peak":0.5994961857795715,"output_rms":0.0625700880512365,"finished_at":"2026-10-13T07:00:09.344844+00:00","evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"948aa635663be12e5b2ded6369249482799ccdd2781034089cf946fa0f4d7fbf","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.558130532566438,"spectral_convergence":0.25217113384372836,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.950415933125039,"spectral_convergence":0.2645817453437749,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.102656203054355,"spectral_convergence":0.28657986814682973,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0006590583154356387,"delta_db":-0.2185409129512639,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0023773717509130935,"delta_db":-0.32019099619694424,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0005900621783997618,"delta_db":-0.885207364361424,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00020094834647421136,"delta_db":-0.523179686263202,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":4.860200295421487e-05,"delta_db":-1.2542611855332697,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":3.8973324562567944e-05,"delta_db":-0.5589806814934227,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6267268044665878,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4191349589187027,"pitch":null,"output_rms_dbfs":-24.072664677294956,"output_peak":0.5994961857795715,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.870400889581944},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.558130532566438,"spectral_convergence":0.25217113384372836,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.950415933125039,"spectral_convergence":0.2645817453437749,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.102656203054355,"spectral_convergence":0.28657986814682973,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0006590583154356387,"delta_db":-0.2185409129512639,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0023773717509130935,"delta_db":-0.32019099619694424,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0005900621783997618,"delta_db":-0.885207364361424,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00020094834647421136,"delta_db":-0.523179686263202,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":4.860200295421487e-05,"delta_db":-1.2542611855332697,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":3.8973324562567944e-05,"delta_db":-0.5589806814934227,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6267268044665878,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4191349589187027,"pitch":null,"output_rms_dbfs":-24.072664677294956,"output_peak":0.5994961857795715,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.870400889581944},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8619753145389155,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06566342232761468,"rms_dbfs":-23.65352971837625,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.42146394315191,"mean":4.2438507080078126e-07,"zero_energy":false},"output":{"rms":0.0625700880512365,"rms_dbfs":-24.072664677294956,"peak":0.5994961857795715,"peak_dbfs":-4.4442715140267355,"crest_factor_db":19.628393163268218,"mean":-2.8729830608559345e-05,"zero_energy":false},"rms_difference_db":-0.4191349589187041},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":148,"quiet_samples":23680,"reference_mean_power":6.957889323726901e-07,"output_mean_power":6.951267894589255e-07,"output_rms_dbfs":-61.57935973913048,"output_zero_energy":false,"output_peak":0.005377057008445263,"output_energy_fraction":4.379665667406562e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":110,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.7927927927927928,"absolute_difference_cents_median":2.4853206652654123,"absolute_difference_cents_p95":10.790116068869146,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019088238002709922,"log_rms_mae_db":1.1563135942005636,"rms_envelope_nrmse":0.09456897981389988,"envelope_correlation":0.9929593258484528,"rms_gain_error_db":-0.41913515367307697,"reference_mean_power":0.004315280923443156,"output_mean_power":0.003918280817592422},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.876167788850743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007125355736933842,"output_power":0.0006336540546335755,"reference_active":true,"delta_db":-0.5095433340681109},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011942531791187452,"output_power":0.00010891704895195939,"reference_active":true,"delta_db":-0.400005401997435},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5392584995669358e-05,"output_power":2.9206125748092685e-05,"reference_active":true,"delta_db":0.6076703592457362}],"active_band_count":3,"active_band_mae_db":0.505739698437094}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018007320980187037,"log_rms_mae_db":0.8991625611553328,"rms_envelope_nrmse":0.07864861076752941,"envelope_correlation":0.9955772703674403,"rms_gain_error_db":-0.41913648233034284,"reference_mean_power":0.004336982504671315,"output_mean_power":0.003937984677443485},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7431428852385876e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439047386291531,"output_power":0.0005776037870757029,"reference_active":true,"delta_db":-0.47191589132404377},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.115354721039442e-05,"output_power":4.6222720018101726e-05,"reference_active":true,"delta_db":-0.44020256618907494},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.965415989281464e-06,"output_power":1.9639572864727905e-06,"reference_active":true,"delta_db":-0.003224466423458806}],"active_band_count":3,"active_band_mae_db":0.3051143079788592}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017279669295308916,"log_rms_mae_db":1.4791947245432941,"rms_envelope_nrmse":0.12469491234276944,"envelope_correlation":0.9905220376784432,"rms_gain_error_db":-0.6249893492868479,"reference_mean_power":0.0026069520032741206,"output_mean_power":0.0022575329639536235},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2087875823984064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005529275902426358,"output_power":0.0004839145085432104,"reference_active":true,"delta_db":-0.5789961795304663},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648535673845477e-05,"output_power":8.317639370523803e-05,"reference_active":true,"delta_db":-0.6446132026798269},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7058636340876773e-05,"output_power":1.934498526747083e-05,"reference_active":true,"delta_db":0.5462409238381432}],"active_band_count":3,"active_band_mae_db":0.5899501020161455}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014230358058689408,"log_rms_mae_db":1.1684688058498338,"rms_envelope_nrmse":0.10672826864853567,"envelope_correlation":0.9937891366018681,"rms_gain_error_db":-0.6249897215374451,"reference_mean_power":0.0026200630509491097,"output_mean_power":0.0022688864968510355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1215946456987494e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005050917134712818,"output_power":0.00044538157548961176,"reference_active":true,"delta_db":-0.5463799633137518},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.798176008445663e-05,"output_power":3.525882340830379e-05,"reference_active":true,"delta_db":-0.3230727044483297},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3776930382412419e-06,"output_power":1.2903144494516547e-06,"reference_active":true,"delta_db":-0.2845690313307805}],"active_band_count":3,"active_band_mae_db":0.38467389969762067}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":1.7898593237808123,"rms_envelope_nrmse":0.1483743651263636,"envelope_correlation":0.9858037754842355,"rms_gain_error_db":-0.46996223826887623,"reference_mean_power":0.0002448429483315043,"output_mean_power":0.00021973102250727813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7588735706755046e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.377351846027071e-05,"output_power":7.89228358290467e-05,"reference_active":true,"delta_db":-0.748775296181088},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.692404819271502e-06,"output_power":9.41124188964042e-06,"reference_active":true,"delta_db":-0.12784608984162144},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.811201191302258e-06,"output_power":5.3369883949671676e-06,"reference_active":true,"delta_db":1.4623438289734851}],"active_band_count":3,"active_band_mae_db":0.7796550716653982}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":1.4158921814913052,"rms_envelope_nrmse":0.10445608489634822,"envelope_correlation":0.9937009483025202,"rms_gain_error_db":-0.4699755400341082,"reference_mean_power":0.00024607384660115493,"output_mean_power":0.00022083499927810869},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6710797473991248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.068314304092383e-05,"output_power":7.711048360112049e-05,"reference_active":true,"delta_db":-0.7041313735917571},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.419207768284794e-06,"output_power":4.34828651638199e-06,"reference_active":true,"delta_db":-0.07026267523947308},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6860314835877605e-07,"output_power":3.141249881309655e-07,"reference_active":true,"delta_db":0.679913863638377}],"active_band_count":3,"active_band_mae_db":0.48476930415653574}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":2.255968956331581,"rms_envelope_nrmse":0.2776162541664072,"envelope_correlation":0.9592328957610483,"rms_gain_error_db":-1.5135398584985926,"reference_mean_power":6.169396290931911e-05,"output_mean_power":4.3540025812978294e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.220333375078007e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9572001812371347e-05,"output_power":1.0590878236228082e-05,"reference_active":true,"delta_db":-2.6670327233941316},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.501529531199527e-06,"output_power":5.585961253253182e-06,"reference_active":true,"delta_db":-1.8239914882495902},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.287934963899871e-07,"output_power":7.357592150106814e-07,"reference_active":true,"delta_db":-1.0118345559500193}],"active_band_count":3,"active_band_mae_db":1.8342862558645805}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":1.9795406002664775,"rms_envelope_nrmse":0.24797830135418475,"envelope_correlation":0.9699792221311443,"rms_gain_error_db":-1.5135531755405274,"reference_mean_power":6.200418008997281e-05,"output_mean_power":4.375882495122361e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.99689742660592e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0058245625049688e-05,"output_power":1.0780002243966033e-05,"reference_active":true,"delta_db":-2.6967409394738877},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3812703887943674e-06,"output_power":2.9656930562447663e-06,"reference_active":true,"delta_db":-1.6947385601829232},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0248947956398812e-07,"output_power":7.943315404269475e-08,"reference_active":true,"delta_db":-1.1067748044272068}],"active_band_count":3,"active_band_mae_db":1.8327514346946725}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.553372621536255,"predicted_mos":3.553372621536255,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"0595b2e2c827d791296ea793cc809c782a536fb7fb3598dadcb41b06868c8765","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9247703010005353},"started_at":"2026-10-13T07:30:12.010350+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00713980500222533,"native_input_samples":264600,"native_input_sha256":"4bdb70ff933a146ef65d80f02ed71e0c1508eb170987095ccaadebcbbe395794","output_sha256":"c62db1a5b96000d223f7eb612c6fbd44c42d68a02a564056301938ee3a721815","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1406617489992641,"synthesis_s":6.019722856999579,"total_s":6.160384605998843,"rtf":1.0267307676664739,"output_peak":0.6594476699829102,"output_rms":0.05664865356191933,"finished_at":"2026-10-13T07:30:18.194580+00:00","evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"c62db1a5b96000d223f7eb612c6fbd44c42d68a02a564056301938ee3a721815","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.932135569515744,"output_peak":0.6381424696259529},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.591514985486945,"spectral_convergence":0.5662225283646374,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.339305059668163,"spectral_convergence":0.45481029230741277,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.250187777413071,"spectral_convergence":0.5411177911995294,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.000672985681573575,"delta_db":-0.12772110970145797,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0017389047962223716,"delta_db":-1.6784039217690896,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0005320707169495577,"delta_db":-1.3344915696459085,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00017326858611872752,"delta_db":-1.1668257398589708,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":6.35560154693264e-05,"delta_db":-0.08923623973565345,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":3.256625193104653e-05,"delta_db":-1.3389774414373683,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9559426703580747,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.2768871959091441,"pitch":null,"output_rms_dbfs":-24.930416914285395,"output_peak":0.6381424696259529,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.393669274189394},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6789869090889294,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06566342232761468,"rms_dbfs":-23.65352971837625,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.42146394315191,"mean":4.2438507080078126e-07,"zero_energy":false},"output":{"rms":0.056686436193013626,"rms_dbfs":-24.930416914285395,"peak":0.6381424696259529,"peak_dbfs":-3.9016470254197992,"crest_factor_db":21.0287698888656,"mean":-0.00011164661496032539,"zero_energy":false},"rms_difference_db":-1.2768871959091435},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":148,"quiet_samples":23680,"reference_mean_power":6.957889323726901e-07,"output_mean_power":5.01580982761896e-07,"output_rms_dbfs":-62.996589379631004,"output_zero_energy":false,"output_peak":0.005250859564574694,"output_energy_fraction":3.850288023937041e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":96,"both_estimated_voiced_frames":49,"reference_voiced_output_unvoiced_frames":62,"reference_unvoiced_output_voiced_frames":47,"paired_voiced_fraction":0.44144144144144143,"absolute_difference_cents_median":12.305872141664677,"absolute_difference_cents_p95":34.63623212171221,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019088238002709922,"log_rms_mae_db":2.02614006445555,"rms_envelope_nrmse":0.18867411028991643,"envelope_correlation":0.9815062160543377,"rms_gain_error_db":-1.2768871616004827,"reference_mean_power":0.004315280923443156,"output_mean_power":0.0032160319692970127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.876167788850743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007125355736933842,"output_power":0.0006179524367315282,"reference_active":true,"delta_db":-0.6185150248831627},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011942531791187452,"output_power":9.306055727962157e-05,"reference_active":true,"delta_db":-1.0833075714052518},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5392584995669358e-05,"output_power":2.196808414408269e-05,"reference_active":true,"delta_db":-0.6291473145578792}],"active_band_count":3,"active_band_mae_db":0.776989970282098}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018007320980187037,"log_rms_mae_db":1.8375773459175178,"rms_envelope_nrmse":0.16060699024671232,"envelope_correlation":0.9919004473294566,"rms_gain_error_db":-1.2768870116761943,"reference_mean_power":0.004336982504671315,"output_mean_power":0.003232205530590384},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7431428852385876e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439047386291531,"output_power":0.000561149122227669,"reference_active":true,"delta_db":-0.5974333325419234},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.115354721039442e-05,"output_power":3.9505234015910195e-05,"reference_active":true,"delta_db":-1.1222111621211812},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.965415989281464e-06,"output_power":1.725943500216669e-06,"reference_active":true,"delta_db":-0.5642791012887358}],"active_band_count":3,"active_band_mae_db":0.7613078653172801}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017279669295308916,"log_rms_mae_db":3.1989995682510477,"rms_envelope_nrmse":0.23959147566550504,"envelope_correlation":0.975315834597858,"rms_gain_error_db":-1.7066030599738147,"reference_mean_power":0.0026069520032741206,"output_mean_power":0.0017598381551395064},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2087875823984064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005529275902426358,"output_power":0.00046481918167599504,"reference_active":true,"delta_db":-0.7538421934464334},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648535673845477e-05,"output_power":7.658740413562335e-05,"reference_active":true,"delta_db":-1.0030405718204098},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7058636340876773e-05,"output_power":1.528478493156551e-05,"reference_active":true,"delta_db":-0.47684978673943434}],"active_band_count":3,"active_band_mae_db":0.744577517335426}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014230358058689408,"log_rms_mae_db":2.8893576860399106,"rms_envelope_nrmse":0.21502878565914987,"envelope_correlation":0.985118115823018,"rms_gain_error_db":-1.706604107046161,"reference_mean_power":0.0026200630509491097,"output_mean_power":0.0017686884179180257},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1215946456987494e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005050917134712818,"output_power":0.0004293249101761493,"reference_active":true,"delta_db":-0.7058415581080584},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.798176008445663e-05,"output_power":2.8969887144613535e-05,"reference_active":true,"delta_db":-1.1762828281939388},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3776930382412419e-06,"output_power":1.0930344115996593e-06,"reference_active":true,"delta_db":-1.0051862893738435}],"active_band_count":3,"active_band_mae_db":0.9624368918919469}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":3.1679028448707545,"rms_envelope_nrmse":0.3023843747958804,"envelope_correlation":0.9391318180526411,"rms_gain_error_db":-0.837144814512918,"reference_mean_power":0.0002448429483315043,"output_mean_power":0.00020191710880298306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7588735706755046e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.377351846027071e-05,"output_power":5.736947639027225e-05,"reference_active":true,"delta_db":-2.1339932529725685},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.692404819271502e-06,"output_power":6.897029825426764e-06,"reference_active":true,"delta_db":-1.4776944081908074},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.811201191302258e-06,"output_power":1.5758677067755237e-06,"reference_active":true,"delta_db":-3.835421196524602}],"active_band_count":3,"active_band_mae_db":2.482369619229326}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":2.6117557554525295,"rms_envelope_nrmse":0.2670578351483989,"envelope_correlation":0.9520905296452347,"rms_gain_error_db":-0.8371426109389596,"reference_mean_power":0.00024607384660115493,"output_mean_power":0.00020293230908571177},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6710797473991248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.068314304092383e-05,"output_power":5.524983856793734e-05,"reference_active":true,"delta_db":-2.151955507726372},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.419207768284794e-06,"output_power":3.338485322512448e-06,"reference_active":true,"delta_db":-1.2179494910260382},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6860314835877605e-07,"output_power":1.089724566479478e-07,"reference_active":true,"delta_db":-3.917943571930351}],"active_band_count":3,"active_band_mae_db":2.429282856894254}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":4.211060854286361,"rms_envelope_nrmse":0.2612080785772723,"envelope_correlation":0.9621372061166555,"rms_gain_error_db":-1.2621361347997972,"reference_mean_power":6.169396290931911e-05,"output_mean_power":4.613484380023623e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.220333375078007e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9572001812371347e-05,"output_power":1.3679793846106266e-05,"reference_active":true,"delta_db":-1.5555569466905697},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.501529531199527e-06,"output_power":6.051922051801276e-06,"reference_active":true,"delta_db":-1.4760374202373896},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.287934963899871e-07,"output_power":8.81002096339314e-07,"reference_active":true,"delta_db":-0.22942223904195644}],"active_band_count":3,"active_band_mae_db":1.0870055353233052}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":3.8898156984241594,"rms_envelope_nrmse":0.2098373740454905,"envelope_correlation":0.9793485566934303,"rms_gain_error_db":-1.2621331620184462,"reference_mean_power":6.200418008997281e-05,"output_mean_power":4.6366856440880186e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.99689742660592e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0058245625049688e-05,"output_power":1.4072468641177733e-05,"reference_active":true,"delta_db":-1.539226556411582},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3812703887943674e-06,"output_power":3.2610354344858837e-06,"reference_active":true,"delta_db":-1.282445384741811},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0248947956398812e-07,"output_power":6.335597776413194e-08,"reference_active":true,"delta_db":-2.0889169010130555}],"active_band_count":3,"active_band_mae_db":1.636862947388816}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"golf","status":"measured","mos":{"status":"measured","value":3.5015134811401367,"predicted_mos":3.5015134811401367,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"9a5a087ea237c106b44cd314d4e64004dcca984eb52ff05800a337c08d597821","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7366000519996305},"started_at":"2026-10-13T05:03:01.276423+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004523108000284992,"native_input_samples":144000,"native_input_sha256":"2a1a05cb3e4dd3648fad327e836589428ac7cc9900f82d942e3a9e257d58e2b5","output_sha256":"0d3d9c0ac8d27dfb8e392b3cc06f77fa2a84b5e1fac1d56cfc9cc35cd5d25499","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6214858759994968,"synthesis_s":0.1319276020003599,"total_s":0.7534134779998567,"rtf":0.1255689129999761,"output_peak":0.542223334312439,"output_rms":0.05835303523914188,"finished_at":"2026-10-13T05:03:02.043522+00:00","analysis_stages":{"external_f0_s":0.08799208199980058},"evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"0d3d9c0ac8d27dfb8e392b3cc06f77fa2a84b5e1fac1d56cfc9cc35cd5d25499","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.67610741837095,"output_peak":0.5430689632010242},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.389403509228043,"spectral_convergence":0.6785063461285453,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.740377391248508,"spectral_convergence":0.38001575791749886,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.947319087620366,"spectral_convergence":0.3605182429793775,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006942246983839805,"output_power":0.0006386275612416932,"delta_db":-0.3625240302551299,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002563546344654314,"output_power":0.001960875453515035,"delta_db":-1.1639116316747173,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007246745909179994,"output_power":0.0005886942497641184,"delta_db":-0.9025324028763261,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00022705270010420664,"output_power":0.00014311430621174808,"delta_db":-2.0044362134441966,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.498367694439679e-05,"output_power":4.4922040211286044e-05,"delta_db":-1.6034480894445482,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.440073497649298e-05,"output_power":3.090073256719784e-05,"delta_db":-1.5742138374372863,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2685110341887007,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0298215531628783,"pitch":null,"output_rms_dbfs":-24.67610741837095,"output_peak":0.5430689632010242,"reference_rms_dbfs":-23.64628586520807,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.35903332936564},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8397009783939192,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06571820711173876,"rms_dbfs":-23.64628586520807,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.41422008998373,"mean":7.323709274572204e-07,"zero_energy":false},"output":{"rms":0.058370663380769267,"rms_dbfs":-24.67610741837095,"peak":0.5430689632010242,"peak_dbfs":-5.3029003350770925,"crest_factor_db":19.373207083293856,"mean":0.000976637526785737,"zero_energy":false},"rms_difference_db":-1.0298215531628792},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":147,"quiet_samples":23520,"reference_mean_power":6.987927893341399e-07,"output_mean_power":7.325994531108452e-07,"output_rms_dbfs":-61.35133409874356,"output_zero_energy":false,"output_peak":0.005537848387396332,"output_energy_fraction":5.276766105164651e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":154,"both_estimated_voiced_frames":100,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":54,"paired_voiced_fraction":0.9009009009009009,"absolute_difference_cents_median":5.391664465478517,"absolute_difference_cents_p95":18.753163587356845,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019088238002709922,"log_rms_mae_db":2.238548195669968,"rms_envelope_nrmse":0.18394676737946075,"envelope_correlation":0.9762376627172662,"rms_gain_error_db":-1.029822157623907,"reference_mean_power":0.004322490636266125,"output_mean_power":0.0034099801066029382},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.8727532306731334e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007100711186340043,"output_power":0.0006111462586863601,"reference_active":true,"delta_db":-0.6515669111256605},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011892725740633036,"output_power":8.628279745216334e-05,"reference_active":true,"delta_db":-1.3935718637853614},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5339497770939233e-05,"output_power":2.054717187695367e-05,"reference_active":true,"delta_db":-0.9104594909725249}],"active_band_count":3,"active_band_mae_db":0.9851994219611822}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00018007320980187037,"log_rms_mae_db":1.704708606588228,"rms_envelope_nrmse":0.14122702666920978,"envelope_correlation":0.9910892502272917,"rms_gain_error_db":-1.0298258132976466,"reference_mean_power":0.004344265020300591,"output_mean_power":0.0034271548683500845},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.7402278973159239e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006417971756031907,"output_power":0.0005547827760792051,"reference_active":true,"delta_db":-0.632748321383201},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.083722808879368e-05,"output_power":3.5236120654113354e-05,"reference_active":true,"delta_db":-1.5919377411348528},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9418237781970595e-06,"output_power":1.4937051393677003e-06,"reference_active":true,"delta_db":-1.1394493957849243}],"active_band_count":3,"active_band_mae_db":1.1213784861009928}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017279669295308916,"log_rms_mae_db":2.2014038280113697,"rms_envelope_nrmse":0.20899646625133256,"envelope_correlation":0.9735077845003742,"rms_gain_error_db":-1.1292842946301764,"reference_mean_power":0.002611307730002218,"output_mean_power":0.0020133979629654874},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.2065926633240273e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005512299830440861,"output_power":0.00045281433392692854,"reference_active":true,"delta_db":-0.8541266604001708},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.61260979503435e-05,"output_power":6.83022713581977e-05,"reference_active":true,"delta_db":-1.4840616717106858},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6989579916228942e-05,"output_power":1.3909823930638784e-05,"reference_active":true,"delta_db":-0.868610078868745}],"active_band_count":3,"active_band_mae_db":1.068932803659867}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014230358058689408,"log_rms_mae_db":1.728938871431035,"rms_envelope_nrmse":0.16408850993910234,"envelope_correlation":0.9881689957477106,"rms_gain_error_db":-1.1292843015669398,"reference_mean_power":0.00262446276058592,"output_mean_power":0.002023540889835293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.1197362865671443e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005036812378562802,"output_power":0.00041772818670041595,"reference_active":true,"delta_db":-0.8126199288987299},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.774777353191208e-05,"output_power":2.639514942790139e-05,"reference_active":true,"delta_db":-1.553672158642692},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.358209136911847e-06,"output_power":9.146057086704267e-07,"reference_active":true,"delta_db":-1.7173273993836704}],"active_band_count":3,"active_band_mae_db":1.3612064956416974}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":2.5227781835824463,"rms_envelope_nrmse":0.26356824355102076,"envelope_correlation":0.9725453549411464,"rms_gain_error_db":-1.8722803091062665,"reference_mean_power":0.0002452518823864371,"output_mean_power":0.0001593618337648308},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.7585374060412627e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.351408410254075e-05,"output_power":6.680461550962721e-05,"reference_active":true,"delta_db":-1.4607055586161724},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.684594077811937e-06,"output_power":6.300304337423041e-06,"reference_active":true,"delta_db":-1.8671989368351127},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.816543511739522e-06,"output_power":2.309384492389649e-06,"reference_active":true,"delta_db":-2.1817397287764044}],"active_band_count":3,"active_band_mae_db":1.836548074742563}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":2.0945332127041265,"rms_envelope_nrmse":0.22237673445274156,"envelope_correlation":0.9894123917219567,"rms_gain_error_db":-1.872286171083803,"reference_mean_power":0.0002464868636925138,"output_mean_power":0.00016016409413462338},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.670723329960625e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.041062588392774e-05,"output_power":6.471511190455316e-05,"reference_active":true,"delta_db":-1.4521376924935812},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4044468711034795e-06,"output_power":2.8724946070975545e-06,"reference_active":true,"delta_db":-1.8563215364701535},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6809639810422294e-07,"output_power":1.4042834933209257e-07,"reference_active":true,"delta_db":-2.808361882885011}],"active_band_count":3,"active_band_mae_db":2.0389403706162486}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":3.1726114956395453,"rms_envelope_nrmse":0.2619327537434858,"envelope_correlation":0.9653626108016234,"rms_gain_error_db":-1.4130067026493045,"reference_mean_power":6.179702393591486e-05,"output_mean_power":4.463411111683625e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.203614209738633e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9510041435145518e-05,"output_power":1.1095963298231e-05,"reference_active":true,"delta_db":-2.450931801784762},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.426280468593937e-06,"output_power":5.329275903059153e-06,"reference_active":true,"delta_db":-1.9896770574189229},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.146839611693521e-07,"output_power":9.375681879058248e-07,"reference_active":true,"delta_db":0.1073179925384535}],"active_band_count":3,"active_band_mae_db":1.5159756172473795}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":2.81787769419497,"rms_envelope_nrmse":0.212133479364876,"envelope_correlation":0.9824001620542929,"rms_gain_error_db":-1.413005574352084,"reference_mean_power":6.21082834994606e-05,"output_mean_power":4.4858936088409915e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.981536363450079e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.997127307679342e-05,"output_power":1.1395543911091856e-05,"reference_active":true,"delta_db":-2.4367069107403263},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.336514368152637e-06,"output_power":2.830601596698117e-06,"reference_active":true,"delta_db":-1.8526204236228283},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.010298291572669e-07,"output_power":9.093979402630364e-08,"reference_active":true,"delta_db":-0.45695652375542894}],"active_band_count":3,"active_band_mae_db":1.582094619372861}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.845113515853882,"predicted_mos":3.845113515853882,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"009295ef2e58c6a846c904653eadc0241b6ac592538fae817e86497cc72e89c9","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.822026816000289},"started_at":"2026-10-13T07:45:26.088045+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002616667999973288,"native_input_samples":144000,"native_input_sha256":"2a1a05cb3e4dd3648fad327e836589428ac7cc9900f82d942e3a9e257d58e2b5","output_sha256":"0777b0b721ab223ec54b3cb52379486c77946db3750208e6264a782023e76a29","output_samples":144120,"length_delta_samples":120,"analysis_s":2.474853406998591,"synthesis_s":4.194636551001167,"total_s":6.669489957999758,"rtf":1.1115816596666264,"output_peak":0.6544311046600342,"output_rms":0.06335092585298825,"finished_at":"2026-10-13T07:45:32.771454+00:00","analysis_stages":{"external_f0_s":1.6280902329999662},"evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"0777b0b721ab223ec54b3cb52379486c77946db3750208e6264a782023e76a29","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.961775943728533,"output_peak":0.6302974067700122},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.891517556184852,"spectral_convergence":0.6349274957916408,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.43294209649452,"spectral_convergence":0.35395164405952534,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.66969662273645,"spectral_convergence":0.3354261942132713,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.00086048469491809,"delta_db":0.9396521486374837,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.002287659230645456,"delta_db":-0.48724865020327085,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0005662994436723484,"delta_db":-1.0637237973102347,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00021319078180775792,"delta_db":-0.26633983300682884,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":5.684133597399634e-05,"delta_db":-0.5741600920318369,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":3.5136682735163086e-05,"delta_db":-1.0090476515778264,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7233620287945802,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3046287573998715,"pitch":null,"output_rms_dbfs":-23.958158475776123,"output_peak":0.6302974067700122,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.998052091805274},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8624328325205352,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06566342232761468,"rms_dbfs":-23.65352971837625,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.42146394315191,"mean":4.2438507080078126e-07,"zero_energy":false},"output":{"rms":0.06340041143204682,"rms_dbfs":-23.958158475776123,"peak":0.6302974067700122,"peak_dbfs":-4.00908959373364,"crest_factor_db":19.949068882042486,"mean":-3.612133705364129e-05,"zero_energy":false},"rms_difference_db":-0.30462875739987183},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":148,"quiet_samples":23680,"reference_mean_power":6.957889323726901e-07,"output_mean_power":5.053569928110161e-07,"output_rms_dbfs":-62.96401720428766,"output_zero_energy":false,"output_peak":0.009584145423686576,"output_energy_fraction":3.101162988593739e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":126,"both_estimated_voiced_frames":87,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.7837837837837838,"absolute_difference_cents_median":6.849466034335114,"absolute_difference_cents_p95":16.641071945743008,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019088238002709922,"log_rms_mae_db":2.1648053311031616,"rms_envelope_nrmse":0.1668463050939859,"envelope_correlation":0.9722129543194346,"rms_gain_error_db":-0.3046287335237814,"reference_mean_power":0.004315280923443156,"output_mean_power":0.004022964498405408},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.876167788850743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007125355736933842,"output_power":0.0006707080178903311,"reference_active":true,"delta_db":-0.2627305350246709},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011942531791187452,"output_power":9.964568131698978e-05,"reference_active":true,"delta_db":-0.7863792510417806},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5392584995669358e-05,"output_power":1.6491537097975756e-05,"reference_active":true,"delta_db":-1.8744577876404338}],"active_band_count":3,"active_band_mae_db":0.9745225245689618}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018007320980187037,"log_rms_mae_db":1.7545106764760638,"rms_envelope_nrmse":0.10600875545358265,"envelope_correlation":0.9886987173966048,"rms_gain_error_db":-0.3046281453114701,"reference_mean_power":0.004336982504671315,"output_mean_power":0.004043196565679299},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7431428852385876e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439047386291531,"output_power":0.000603522825424485,"reference_active":true,"delta_db":-0.2812792126858752},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.115354721039442e-05,"output_power":4.274148094519734e-05,"reference_active":true,"delta_db":-0.7802618905492559},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.965415989281464e-06,"output_power":1.2903986597896938e-06,"reference_active":true,"delta_db":-1.8273058153138555}],"active_band_count":3,"active_band_mae_db":0.9629489728496622}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017279669295308916,"log_rms_mae_db":2.3372968504319007,"rms_envelope_nrmse":0.19299319459556022,"envelope_correlation":0.9731352667181516,"rms_gain_error_db":-0.7553791426887722,"reference_mean_power":0.0026069520032741206,"output_mean_power":0.0021907616061089797},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2087875823984064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005529275902426358,"output_power":0.0004752205491483975,"reference_active":true,"delta_db":-0.6577304927932913},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648535673845477e-05,"output_power":8.149649814963322e-05,"reference_active":true,"delta_db":-0.7332245912080376},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7058636340876773e-05,"output_power":1.0991679222010291e-05,"reference_active":true,"delta_db":-1.908802653528594}],"active_band_count":3,"active_band_mae_db":1.0999192458433076}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014230358058689408,"log_rms_mae_db":2.0114950572448773,"rms_envelope_nrmse":0.14174897842293727,"envelope_correlation":0.9875689858822085,"rms_gain_error_db":-0.755379050014489,"reference_mean_power":0.0026200630509491097,"output_mean_power":0.0022017795697350235},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1215946456987494e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005050917134712818,"output_power":0.000433771300947187,"reference_active":true,"delta_db":-0.6610942857932174},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.798176008445663e-05,"output_power":3.295514205833437e-05,"reference_active":true,"delta_db":-0.6165189810941227},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3776930382412419e-06,"output_power":9.10432020430783e-07,"reference_active":true,"delta_db":-1.7990494012296319}],"active_band_count":3,"active_band_mae_db":1.0255542227056573}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":2.0379091401480154,"rms_envelope_nrmse":0.21266411443416272,"envelope_correlation":0.9677953988253734,"rms_gain_error_db":-0.2445674602716075,"reference_mean_power":0.0002448429483315043,"output_mean_power":0.00023143596795396397},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7588735706755046e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.377351846027071e-05,"output_power":9.125267217829692e-05,"reference_active":true,"delta_db":-0.1183462051808645},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.692404819271502e-06,"output_power":9.334146686011581e-06,"reference_active":true,"delta_db":-0.16356923333807621},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.811201191302258e-06,"output_power":2.4440389388134733e-06,"reference_active":true,"delta_db":-1.929537546803781}],"active_band_count":3,"active_band_mae_db":0.7371509951075739}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":1.412267532161146,"rms_envelope_nrmse":0.12307111373106937,"envelope_correlation":0.9891116825986412,"rms_gain_error_db":-0.24456566490860088,"reference_mean_power":0.00024607384660115493,"output_mean_power":0.00023259956150472672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6710797473991248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.068314304092383e-05,"output_power":8.794275094681155e-05,"reference_active":true,"delta_db":-0.13326517476413796},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.419207768284794e-06,"output_power":4.170441094965653e-06,"reference_active":true,"delta_db":-0.2516242890237644},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6860314835877605e-07,"output_power":1.3897008574253017e-07,"reference_active":true,"delta_db":-2.8618977335812144}],"active_band_count":3,"active_band_mae_db":1.082262399123039}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":2.155060493871161,"rms_envelope_nrmse":0.26198191273962046,"envelope_correlation":0.9558131405902479,"rms_gain_error_db":-0.823175994980174,"reference_mean_power":6.169396290931911e-05,"output_mean_power":5.104169270404407e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.220333375078007e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9572001812371347e-05,"output_power":1.4724803560765772e-05,"reference_active":true,"delta_db":-1.2358573759966598},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.501529531199527e-06,"output_power":5.790209590983615e-06,"reference_active":true,"delta_db":-1.6680278336111096},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.287934963899871e-07,"output_power":5.800700014894264e-07,"reference_active":true,"delta_db":-2.0443875935202787}],"active_band_count":3,"active_band_mae_db":1.6494242677093494}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":1.428271157947906,"rms_envelope_nrmse":0.1830882207939838,"envelope_correlation":0.9805808085976655,"rms_gain_error_db":-0.8231758858110216,"reference_mean_power":6.200418008997281e-05,"output_mean_power":5.1298348117534855e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.99689742660592e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0058245625049688e-05,"output_power":1.4856843832900552e-05,"reference_active":true,"delta_db":-1.3036638688941207},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3812703887943674e-06,"output_power":3.0357679319815016e-06,"reference_active":true,"delta_db":-1.5933148732851095},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0248947956398812e-07,"output_power":5.46753434019178e-08,"reference_active":true,"delta_db":-2.7288776840188684}],"active_band_count":3,"active_band_mae_db":1.875285475399366}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.632578134536743,"predicted_mos":3.632578134536743,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"f028b6ccf31b8421a48293cb3640b9d4ad9269f16da648e98f8bd5f7d9806fd0","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8993349760003184},"started_at":"2026-10-13T05:11:26.145153+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003317827999126166,"native_input_samples":144000,"native_input_sha256":"2a1a05cb3e4dd3648fad327e836589428ac7cc9900f82d942e3a9e257d58e2b5","output_sha256":"2806f3535e90c14b661c24d48f1a59c59a1371d7d6e0ee02f292b4c8246e38b9","output_samples":144240,"length_delta_samples":240,"analysis_s":1.387740816000587,"synthesis_s":0.13367810099953203,"total_s":1.521418917000119,"rtf":0.25356981950001983,"output_peak":0.5823395848274231,"output_rms":0.06413913841819012,"finished_at":"2026-10-13T05:11:27.680290+00:00","analysis_stages":{"external_f0_s":0.17115241099963896},"evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"2806f3535e90c14b661c24d48f1a59c59a1371d7d6e0ee02f292b4c8246e38b9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.853968516669816,"output_peak":0.5749661216618835},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.751402645459245,"spectral_convergence":0.19053519980143035,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.428346209833547,"spectral_convergence":0.18726612768588552,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.88169158234661,"spectral_convergence":0.19291743037932269,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0006703669067600937,"delta_db":-0.1446536842382628,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0024855005147450555,"delta_db":-0.12702340105239668,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.0006867715983890682,"delta_db":-0.22606188207129885,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00018905673902157628,"delta_db":-0.7881024291122358,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":5.671503197338794e-05,"delta_db":-0.5838210479285423,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":3.566240874708643e-05,"delta_db":-0.9445484898459706,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4690351557081178,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.19320663035866903,"pitch":null,"output_rms_dbfs":-23.84673634873492,"output_peak":0.5749661216618835,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.687146812546468},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9357211219873208,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06566342232761468,"rms_dbfs":-23.65352971837625,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.42146394315191,"mean":4.2438507080078126e-07,"zero_energy":false},"output":{"rms":0.0642189473569621,"rms_dbfs":-23.84673634873492,"peak":0.5749661216618835,"peak_dbfs":-4.807154883903346,"crest_factor_db":19.039581464831574,"mean":0.00010513434287197203,"zero_energy":false},"rms_difference_db":-0.19320663035867014},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":148,"quiet_samples":23680,"reference_mean_power":6.957889323726901e-07,"output_mean_power":7.928639767684334e-07,"output_rms_dbfs":-61.008013135726245,"output_zero_energy":false,"output_peak":0.011198421140153953,"output_energy_fraction":4.742231885864584e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":123,"both_estimated_voiced_frames":102,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.918918918918919,"absolute_difference_cents_median":1.1423789303594414,"absolute_difference_cents_p95":7.296021948759794,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019088238002709922,"log_rms_mae_db":0.9288500485683214,"rms_envelope_nrmse":0.08689162904321243,"envelope_correlation":0.9927613825244296,"rms_gain_error_db":-0.1932066568728286,"reference_mean_power":0.004315280923443156,"output_mean_power":0.004127512600172168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.876167788850743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007125355736933842,"output_power":0.0006804542858752998,"reference_active":true,"delta_db":-0.20007597743610775},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011942531791187452,"output_power":8.859752637009751e-05,"reference_active":true,"delta_db":-1.2967480943667096},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5392584995669358e-05,"output_power":1.8728454145752608e-05,"reference_active":true,"delta_db":-1.322049828136587}],"active_band_count":3,"active_band_mae_db":0.9396246333131347}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018007320980187037,"log_rms_mae_db":0.6343105324061689,"rms_envelope_nrmse":0.05469458409860569,"envelope_correlation":0.9972299948570981,"rms_gain_error_db":-0.19320644835091977,"reference_mean_power":0.004336982504671315,"output_mean_power":0.004148270092221293},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7431428852385876e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439047386291531,"output_power":0.000612187398099056,"reference_active":true,"delta_db":-0.21937235782626627},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.115354721039442e-05,"output_power":3.800376944131126e-05,"reference_active":true,"delta_db":-1.2904908028947122},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.965415989281464e-06,"output_power":1.0605957888744015e-06,"reference_active":true,"delta_db":-2.6790458644608126}],"active_band_count":3,"active_band_mae_db":1.3963030083939305}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017279669295308916,"log_rms_mae_db":1.2130673209354441,"rms_envelope_nrmse":0.10399560378101225,"envelope_correlation":0.9912688059755744,"rms_gain_error_db":-0.1803692777372538,"reference_mean_power":0.0026069520032741206,"output_mean_power":0.002500898739459459},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2087875823984064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005529275902426358,"output_power":0.0005359137286039323,"reference_active":true,"delta_db":-0.1357337852058513},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648535673845477e-05,"output_power":6.764865963710403e-05,"reference_active":true,"delta_db":-1.5420221083091854},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7058636340876773e-05,"output_power":1.3369072855434866e-05,"reference_active":true,"delta_db":-1.0584302092395546}],"active_band_count":3,"active_band_mae_db":0.9120620342515305}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014230358058689408,"log_rms_mae_db":0.8108165852116431,"rms_envelope_nrmse":0.06742920611531707,"envelope_correlation":0.9962746127833341,"rms_gain_error_db":-0.18036925248597582,"reference_mean_power":0.0026200630509491097,"output_mean_power":0.0025134764319757028},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1215946456987494e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005050917134712818,"output_power":0.0004857174638405023,"reference_active":true,"delta_db":-0.16986524845822135},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.798176008445663e-05,"output_power":2.6062465678523987e-05,"reference_active":true,"delta_db":-1.6355958578556884},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3776930382412419e-06,"output_power":8.929191876194917e-07,"reference_active":true,"delta_db":-1.883403083725207}],"active_band_count":3,"active_band_mae_db":1.2296213966797056}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":1.3247677600843941,"rms_envelope_nrmse":0.17449147732952908,"envelope_correlation":0.982254099733513,"rms_gain_error_db":-0.78647710640004,"reference_mean_power":0.0002448429483315043,"output_mean_power":0.0002042866044553327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7588735706755046e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.377351846027071e-05,"output_power":7.238873330823198e-05,"reference_active":true,"delta_db":-1.1240923419544373},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.692404819271502e-06,"output_power":5.982013681348278e-06,"reference_active":true,"delta_db":-2.095841429065551},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.811201191302258e-06,"output_power":2.1646200769199603e-06,"reference_active":true,"delta_db":-2.456801933295084}],"active_band_count":3,"active_band_mae_db":1.892245234771691}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":0.9388261340953601,"rms_envelope_nrmse":0.11743005965248553,"envelope_correlation":0.9949211264506977,"rms_gain_error_db":-0.7864783176731402,"reference_mean_power":0.00024607384660115493,"output_mean_power":0.000205313556667618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6710797473991248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.068314304092383e-05,"output_power":6.939517661033646e-05,"reference_active":true,"delta_db":-1.1619727880747857},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.419207768284794e-06,"output_power":2.6065026117481624e-06,"reference_active":true,"delta_db":-2.2928625589008944},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6860314835877605e-07,"output_power":1.260853294487965e-07,"reference_active":true,"delta_db":-3.284465412794057}],"active_band_count":3,"active_band_mae_db":2.2464335865899123}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":1.443097387463085,"rms_envelope_nrmse":0.15905706519796864,"envelope_correlation":0.9853904073787758,"rms_gain_error_db":-0.6211835143299698,"reference_mean_power":6.169396290931911e-05,"output_mean_power":5.3471740010251126e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.220333375078007e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9572001812371347e-05,"output_power":1.526322510842712e-05,"reference_active":true,"delta_db":-1.0798893787842936},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.501529531199527e-06,"output_power":7.206668095600091e-06,"reference_active":true,"delta_db":-0.7176254672871403},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.287934963899871e-07,"output_power":5.929048813139197e-07,"reference_active":true,"delta_db":-1.9493413986520607}],"active_band_count":3,"active_band_mae_db":1.248952081574498}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":1.083499348955956,"rms_envelope_nrmse":0.10967684138103422,"envelope_correlation":0.994546655359277,"rms_gain_error_db":-0.6211823887529926,"reference_mean_power":6.200418008997281e-05,"output_mean_power":5.374062712547883e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.99689742660592e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0058245625049688e-05,"output_power":1.5625572897284015e-05,"reference_active":true,"delta_db":-1.0845699588442566},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3812703887943674e-06,"output_power":3.7321627646964797e-06,"reference_active":true,"delta_db":-0.6963948069881288},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0248947956398812e-07,"output_power":6.014430887696593e-08,"reference_active":true,"delta_db":-2.3148474907610055}],"active_band_count":3,"active_band_mae_db":1.365270752197797}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0038","method":"magphase","status":"measured","mos":{"status":"measured","value":4.062680721282959,"predicted_mos":4.062680721282959,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289387,"input_sample_rate":48000,"input_seconds":6.028895833333333,"input_float32_sha256":"f5d517325fe399bd22ec98b6e089f7f5a7ee8174f25d237c905b3999a37ef265","internal_samples":96463,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.90065331400001},"started_at":"2026-10-13T05:18:12.776371+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010559968999587,"native_input_samples":288000,"native_input_sha256":"15f6286416580731ab54a392a52160bd3ade236673813e94f5d8a8caf1450bf8","output_sha256":"195321ac1f7dd83df9964f7a09816015c2ffa6deb988b129ce513419c5b2bcd7","output_samples":289387,"length_delta_samples":1387,"analysis_s":2.7414067390000127,"synthesis_s":0.831195802998991,"total_s":3.5726025419990037,"rtf":0.5954337569998339,"output_peak":0.5892653390953592,"output_rms":0.06278736961159212,"finished_at":"2026-10-13T05:18:16.372547+00:00","evaluation_input":{"reference":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"195321ac1f7dd83df9964f7a09816015c2ffa6deb988b129ce513419c5b2bcd7","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96463,"length_delta_samples":463,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.039291216399036,"output_peak":0.58794938884543},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9952002322133875,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.868282100509276,"spectral_convergence":0.8428991886716111,"spectral_convergence_status":"ok","amplitude_floor":1.6152428194526273e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.709395819205514,"spectral_convergence":0.5974882164929111,"spectral_convergence_status":"ok","amplitude_floor":1.0855275118818902e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.06757367272947,"spectral_convergence":0.5332721607951276,"spectral_convergence_status":"ok","amplitude_floor":7.682277294566263e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006930713321042831,"output_power":0.0006310630471818194,"delta_db":-0.4070518508216236,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002559270540127681,"output_power":0.0023275170292242162,"delta_db":-0.4122333035180563,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007234666478039212,"output_power":0.000676336629020271,"delta_db":-0.29255606139935086,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002266743638240336,"output_power":0.00021778715056242786,"delta_db":-0.17370152718914733,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.48754354732022e-05,"output_power":6.947915364064893e-05,"delta_db":0.2977423340818131,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.432671244156477e-05,"output_power":4.2060262693694394e-05,"delta_db":-0.2279354156888131,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3018700821164674,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3648673033458197,"pitch":null,"output_rms_dbfs":-24.018397021722073,"output_peak":0.58794938884543,"reference_rms_dbfs":-23.65352971837625,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.881750530814753},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.5509598518266697,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06566342232761468,"rms_dbfs":-23.65352971837625,"peak":0.547515869140625,"peak_dbfs":-5.232065775224342,"crest_factor_db":18.42146394315191,"mean":4.2438507080078126e-07,"zero_energy":false},"output":{"rms":0.06296223687515459,"rms_dbfs":-24.018397021722073,"peak":0.58794938884543,"peak_dbfs":-4.613201134639121,"crest_factor_db":19.40519588708295,"mean":4.928892771694488e-05,"zero_energy":false},"rms_difference_db":-0.3648673033458216},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.501139397849329e-06,"threshold_dbfs":-54.557905974578915,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":148,"quiet_samples":23680,"reference_mean_power":6.957889323726901e-07,"output_mean_power":1.0492201732302499e-06,"output_rms_dbfs":-59.79133367873569,"output_zero_energy":false,"output_peak":0.022240193740530538,"output_energy_fraction":6.528550973070454e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":111,"output_estimated_voiced_frames":116,"both_estimated_voiced_frames":75,"reference_voiced_output_unvoiced_frames":36,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.6756756756756757,"absolute_difference_cents_median":16.82723522673381,"absolute_difference_cents_p95":38.78344445732987,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019088238002709922,"log_rms_mae_db":3.8427875123215336,"rms_envelope_nrmse":0.4098950758545497,"envelope_correlation":0.8270603279540629,"rms_gain_error_db":-0.3648671929272839,"reference_mean_power":0.004315280923443156,"output_mean_power":0.003967549502761379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.876167788850743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007125355736933842,"output_power":0.000655201892512003,"reference_active":true,"delta_db":-0.36431408401963206},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011942531791187452,"output_power":0.00010039411960240543,"reference_active":true,"delta_db":-0.753881304871586},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5392584995669358e-05,"output_power":2.4901765368796604e-05,"reference_active":true,"delta_db":-0.08476778060072758}],"active_band_count":3,"active_band_mae_db":0.40098772316398185}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018007320980187037,"log_rms_mae_db":3.054742352151861,"rms_envelope_nrmse":0.3129853711299449,"envelope_correlation":0.8916399961994401,"rms_gain_error_db":-0.36486695723843243,"reference_mean_power":0.004336982504671315,"output_mean_power":0.003987502556377561},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7431428852385876e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006439047386291531,"output_power":0.0005906272024393031,"reference_active":true,"delta_db":-0.37508175834008645},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.115354721039442e-05,"output_power":4.404443125601012e-05,"reference_active":true,"delta_db":-0.6498474863765509},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.965415989281464e-06,"output_power":1.617050468182139e-06,"reference_active":true,"delta_db":-0.8473091039487017}],"active_band_count":3,"active_band_mae_db":0.6240794495551131}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017279669295308916,"log_rms_mae_db":4.615346495020107,"rms_envelope_nrmse":0.4865136085962053,"envelope_correlation":0.7979595253064786,"rms_gain_error_db":-0.4301345244877994,"reference_mean_power":0.0026069520032741206,"output_mean_power":0.0023611282802976264},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.2087875823984064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005529275902426358,"output_power":0.000501134271027388,"reference_active":true,"delta_db":-0.42714157292656396},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.648535673845477e-05,"output_power":8.009173603860934e-05,"reference_active":true,"delta_db":-0.8087369954607772},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7058636340876773e-05,"output_power":1.5455011763952002e-05,"reference_active":true,"delta_db":-0.428749709555426}],"active_band_count":3,"active_band_mae_db":0.5548760926475891}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014230358058689408,"log_rms_mae_db":3.915710184494579,"rms_envelope_nrmse":0.37487082750759065,"envelope_correlation":0.8717984129833527,"rms_gain_error_db":-0.43013461777720924,"reference_mean_power":0.0026200630509491097,"output_mean_power":0.002373002964804602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.1215946456987494e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005050917134712818,"output_power":0.00045660568850855985,"reference_active":true,"delta_db":-0.4382892580277309},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.798176008445663e-05,"output_power":3.1627063456778285e-05,"reference_active":true,"delta_db":-0.7951621623274348},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3776930382412419e-06,"output_power":1.2409750970217404e-06,"reference_active":true,"delta_db":-0.4538939735250812}],"active_band_count":3,"active_band_mae_db":0.5624484646267489}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.058549471038883e-05,"log_rms_mae_db":4.622548915399526,"rms_envelope_nrmse":0.41710838334934214,"envelope_correlation":0.8766729671902038,"rms_gain_error_db":-0.12076129417751887,"reference_mean_power":0.0002448429483315043,"output_mean_power":0.00023812855197548574},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7588735706755046e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.377351846027071e-05,"output_power":9.007341747269985e-05,"reference_active":true,"delta_db":-0.1748357079364539},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.692404819271502e-06,"output_power":9.379664934811758e-06,"reference_active":true,"delta_db":-0.1424422030298876},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.811201191302258e-06,"output_power":3.647018323521345e-06,"reference_active":true,"delta_db":-0.1912393023206377}],"active_band_count":3,"active_band_mae_db":0.16950573776232639}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.399057463688353e-05,"log_rms_mae_db":3.73861177245799,"rms_envelope_nrmse":0.3152230410372808,"envelope_correlation":0.9263979866376829,"rms_gain_error_db":-0.12076121086086868,"reference_mean_power":0.00024607384660115493,"output_mean_power":0.00023932569957041904},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.6710797473991248e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.068314304092383e-05,"output_power":8.710268534286606e-05,"reference_active":true,"delta_db":-0.17495019836996573},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.419207768284794e-06,"output_power":4.24450274711866e-06,"reference_active":true,"delta_db":-0.17517601320604217},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6860314835877605e-07,"output_power":2.1481110522961093e-07,"reference_active":true,"delta_db":-0.9705436921464663}],"active_band_count":3,"active_band_mae_db":0.44022330124082476}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.748604525835054e-05,"log_rms_mae_db":4.726239817656285,"rms_envelope_nrmse":0.6555091600259664,"envelope_correlation":0.7262508759416322,"rms_gain_error_db":0.2719892202116203,"reference_mean_power":6.169396290931911e-05,"output_mean_power":6.568127732377699e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.220333375078007e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9572001812371347e-05,"output_power":2.005812107453881e-05,"reference_active":true,"delta_db":0.1065500116250183},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.501529531199527e-06,"output_power":9.99909054933786e-06,"reference_active":true,"delta_db":0.704634335254753},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.287934963899871e-07,"output_power":1.805124599828933e-06,"reference_active":true,"delta_db":2.885880189997752}],"active_band_count":3,"active_band_mae_db":1.232354845625841}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.636845752844561e-05,"log_rms_mae_db":3.5639159922481003,"rms_envelope_nrmse":0.4772923306775539,"envelope_correlation":0.8461718907667102,"rms_gain_error_db":0.2719876201134925,"reference_mean_power":6.200418008997281e-05,"output_mean_power":6.601151968868732e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.99689742660592e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0058245625049688e-05,"output_power":2.0942380320401708e-05,"reference_active":true,"delta_db":0.18733097054284037},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.3812703887943674e-06,"output_power":5.728905981172019e-06,"reference_active":true,"delta_db":1.1647163860895877},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0248947956398812e-07,"output_power":1.2326019764447273e-07,"reference_active":true,"delta_db":0.8014357196170421}],"active_band_count":3,"active_band_mae_db":0.7178276920831568}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"hnm","status":"measured","mos":{"status":"measured","value":2.724296808242798,"predicted_mos":2.724296808242798,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1a12274df4bcdae7e3acb4c2cc171e7437d40bcdffd2dd83cc63e4b1930dcb8e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4615547110006446},"started_at":"2026-10-13T04:40:13.958079+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4358000044012442e-05,"native_input_samples":96000,"native_input_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","output_sha256":"867e81e1a0cce6dc813a8da5251bbb1543d56b68ae5941eb428c49cb22ce08cb","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1435910609998245,"synthesis_s":0.5324662330003775,"total_s":1.676057294000202,"rtf":0.279342882333367,"output_peak":0.6266237296756909,"output_rms":0.06946404783340415,"finished_at":"2026-10-13T04:40:15.642030+00:00","evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"867e81e1a0cce6dc813a8da5251bbb1543d56b68ae5941eb428c49cb22ce08cb","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1105417433313285,"spectral_convergence":0.3155609006831057,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.954020298628184,"spectral_convergence":0.30753336521361063,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.469448528101313,"spectral_convergence":0.338366586934079,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0006605123769877925,"delta_db":-0.05377761654537173,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0030765871538702095,"delta_db":-0.1019600458789657,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0007203863678858077,"delta_db":-0.16823201249541025,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.0002221714099125647,"delta_db":-0.4670090535133165,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":9.548886026554306e-05,"delta_db":-0.7475064359942519,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":5.010777247954302e-05,"delta_db":0.3523686915744661,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3151423093336304,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13209457235786862,"pitch":null,"output_rms_dbfs":-23.16479825854426,"output_peak":0.6266237296756909,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.178003523353609},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.1105417433313285,"spectral_convergence":0.3155609006831057,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.954020298628184,"spectral_convergence":0.30753336521361063,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.469448528101313,"spectral_convergence":0.338366586934079,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0006605123769877925,"delta_db":-0.05377761654537173,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0030765871538702095,"delta_db":-0.1019600458789657,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0007203863678858077,"delta_db":-0.16823201249541025,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.0002221714099125647,"delta_db":-0.4670090535133165,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":9.548886026554306e-05,"delta_db":-0.7475064359942519,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":5.010777247954302e-05,"delta_db":0.3523686915744661,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3151423093336304,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13209457235786862,"pitch":null,"output_rms_dbfs":-23.16479825854426,"output_peak":0.6266237296756909,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.178003523353609},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7948908241288073,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705285273351481,"rms_dbfs":-23.03270368618639,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.894584102403467,"mean":-1.7856280008951823e-05,"zero_energy":false},"output":{"rms":0.06946404783340415,"rms_dbfs":-23.16479825854426,"peak":0.6266237296756909,"peak_dbfs":-4.059863256048473,"crest_factor_db":19.104935002495786,"mean":-0.00016808368405378768,"zero_energy":false},"rms_difference_db":-0.13209457235786815},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":2.041216440870799e-06,"output_rms_dbfs":-56.90110942323585,"output_zero_energy":false,"output_peak":0.048114073274167966,"output_energy_fraction":0.00010505189757308862,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":108,"both_estimated_voiced_frames":86,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.9555555555555556,"absolute_difference_cents_median":1.4798926199900337,"absolute_difference_cents_p95":6.368128448826868,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023158079660483383,"log_rms_mae_db":1.0281604844318166,"rms_envelope_nrmse":0.10787437544210614,"envelope_correlation":0.9888755151064336,"rms_gain_error_db":-0.13166134070048086,"reference_mean_power":0.004973550834679421,"output_mean_power":0.004825034546317168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9210188012408756e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006213961937331268,"output_power":0.0006022748474091455,"reference_active":true,"delta_db":-0.13573862460131123},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026574930428793645,"output_power":0.00030478735296643304,"reference_active":true,"delta_db":0.5952480591361895},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.273924181429359e-05,"output_power":3.4464417490450875e-05,"reference_active":true,"delta_db":0.22302324884614322}],"active_band_count":3,"active_band_mae_db":0.31800331086121464}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000211279420280134,"log_rms_mae_db":0.5790922814181483,"rms_envelope_nrmse":0.06663441227019447,"envelope_correlation":0.9955754418356878,"rms_gain_error_db":-0.13155985932982814,"reference_mean_power":0.0049805093532051875,"output_mean_power":0.0048318981804004076},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7395423077013874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005644300449164409,"output_power":0.0005572089957579223,"reference_active":true,"delta_db":-0.05592004231949146},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013986675109572898,"output_power":0.00016425958383402376,"reference_active":true,"delta_db":0.6981623121061911},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4974533271774746e-06,"output_power":2.5668227594277696e-06,"reference_active":true,"delta_db":0.11898500619175063}],"active_band_count":3,"active_band_mae_db":0.2910224535391444}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020019856693346264,"log_rms_mae_db":1.3788432118840088,"rms_envelope_nrmse":0.1663146103932428,"envelope_correlation":0.9776435361315802,"rms_gain_error_db":-0.32369167043265595,"reference_mean_power":0.0028106313477950454,"output_mean_power":0.0026087635554237383},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3651110338311905e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046335294017358356,"output_power":0.00042538383320323193,"reference_active":true,"delta_db":-0.3713094271843672},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024122957535154208,"output_power":0.0002350353107017139,"reference_active":true,"delta_db":-0.11297438603808352},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9488783219041956e-05,"output_power":3.236973812166202e-05,"reference_active":true,"delta_db":0.40482333033563656}],"active_band_count":3,"active_band_mae_db":0.29636904785269574}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018155803316515472,"log_rms_mae_db":0.8234993136713389,"rms_envelope_nrmse":0.1144336639079734,"envelope_correlation":0.9892754766163724,"rms_gain_error_db":-0.3248115312687462,"reference_mean_power":0.002819303754025505,"output_mean_power":0.0026161384061299236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2479023732657914e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004464610508434936,"output_power":0.00041540756826112616,"reference_active":true,"delta_db":-0.31309172498499316},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012910816269432612,"output_power":0.00013229865452728395,"reference_active":true,"delta_db":0.10601726616475371},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.446292305046822e-06,"output_power":2.377697726765896e-06,"reference_active":true,"delta_db":-0.12351706649148365}],"active_band_count":3,"active_band_mae_db":0.18087535254707685}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":1.9635798220596188,"rms_envelope_nrmse":0.19737487149212118,"envelope_correlation":0.9718093654093163,"rms_gain_error_db":-0.5154697979878754,"reference_mean_power":0.0002950078039052426,"output_mean_power":0.00026199109185573736},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.385626430084388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.894665895613549e-05,"output_power":8.851258053437011e-05,"reference_active":true,"delta_db":-0.48396131810675913},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2790160745942646e-05,"output_power":1.9205692305831908e-05,"reference_active":true,"delta_db":-0.7431742164560889},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.304247232293747e-06,"output_power":5.558938642556644e-06,"reference_active":true,"delta_db":1.1109467160845685}],"active_band_count":3,"active_band_mae_db":0.7793607502158055}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":1.3065677253561883,"rms_envelope_nrmse":0.12044309131511136,"envelope_correlation":0.9906765422883781,"rms_gain_error_db":-0.5118298590785277,"reference_mean_power":0.00029237164364699544,"output_mean_power":0.0002598676764818746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2489747353423385e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.236418785147835e-05,"output_power":8.413560524912174e-05,"reference_active":true,"delta_db":-0.40523792606532705},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2735133445304275e-05,"output_power":1.103807828048725e-05,"reference_active":true,"delta_db":-0.6211003039021522},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.686566631139435e-07,"output_power":3.383571555310558e-07,"reference_active":true,"delta_db":1.0017774908273003}],"active_band_count":3,"active_band_mae_db":0.6760385735982598}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":2.0634747700730376,"rms_envelope_nrmse":0.2504184556664843,"envelope_correlation":0.9575756626115168,"rms_gain_error_db":-0.2783042585024863,"reference_mean_power":6.736853190004855e-05,"output_mean_power":6.318684302162469e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6302348178615176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0381998821210498e-05,"output_power":1.9945394942781227e-05,"reference_active":true,"delta_db":-0.09404131852842652},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.955680541944289e-06,"output_power":4.399549420572021e-06,"reference_active":true,"delta_db":0.46186989520407074},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.59401259091449e-07,"output_power":1.9356044615153778e-06,"reference_active":true,"delta_db":4.6766684240463094}],"active_band_count":3,"active_band_mae_db":1.7441932125929356}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":1.7788345038084563,"rms_envelope_nrmse":0.18764000050611487,"envelope_correlation":0.975778559993489,"rms_gain_error_db":-0.2791548648210409,"reference_mean_power":6.76432550540454e-05,"output_mean_power":6.343208861943853e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4309026400574305e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.987521517112667e-05,"output_power":2.0880189635524077e-05,"reference_active":true,"delta_db":0.21422599574015153},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.820398831999747e-06,"output_power":2.300866404562064e-06,"reference_active":true,"delta_db":1.0172485473303212},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.685144573008776e-08,"output_power":2.2074058144970064e-07,"reference_active":true,"delta_db":7.774276517193587}],"active_band_count":3,"active_band_mae_db":3.0019170200880203}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.29901385307312,"predicted_mos":3.29901385307312,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d18d22b09af6f0fa9fdf3c969dbc270d18ad9b8ed682f5bf52de3c8531a29697","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":9.23189635000017},"started_at":"2026-10-13T04:43:28.585690+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0110000150452834e-05,"native_input_samples":96000,"native_input_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","output_sha256":"c3d63397654a2d101342069d49f4a369be2a13f94b020f5313d45479e5dd2326","output_samples":96000,"length_delta_samples":0,"analysis_s":2.0900243990008676,"synthesis_s":0.09269888199924026,"total_s":2.182723281000108,"rtf":0.36378721350001797,"output_peak":0.6658339559930181,"output_rms":0.07739179236796671,"finished_at":"2026-10-13T04:43:30.774997+00:00","evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"c3d63397654a2d101342069d49f4a369be2a13f94b020f5313d45479e5dd2326","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.761480094715172,"spectral_convergence":0.6527374437521901,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.474020161411841,"spectral_convergence":0.31784884631547045,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.772663626092746,"spectral_convergence":0.3115171888248068,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0007595642641763825,"delta_db":0.5530580225485866,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0038598216515607093,"delta_db":0.8830201102954336,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0009092181046358632,"delta_db":0.8427938777302564,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.0002900100725393835,"delta_db":0.6902400549594256,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":0.00011982758762315595,"delta_db":0.23853462834478706,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":5.104784539097687e-05,"delta_db":0.4330918863388377,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6067897633695544,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.806601784002024,"pitch":null,"output_rms_dbfs":-22.226101902184364,"output_peak":0.6658339559930181,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.002721294073253},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.761480094715172,"spectral_convergence":0.6527374437521901,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.474020161411841,"spectral_convergence":0.31784884631547045,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.772663626092746,"spectral_convergence":0.3115171888248068,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0007595642641763825,"delta_db":0.5530580225485866,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0038598216515607093,"delta_db":0.8830201102954336,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0009092181046358632,"delta_db":0.8427938777302564,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.0002900100725393835,"delta_db":0.6902400549594256,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":0.00011982758762315595,"delta_db":0.23853462834478706,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":5.104784539097687e-05,"delta_db":0.4330918863388377,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6067897633695544,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.806601784002024,"pitch":null,"output_rms_dbfs":-22.226101902184364,"output_peak":0.6658339559930181,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.002721294073253},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8823078886515644,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705285273351481,"rms_dbfs":-23.03270368618639,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.894584102403467,"mean":-1.7856280008951823e-05,"zero_energy":false},"output":{"rms":0.07739179236796671,"rms_dbfs":-22.226101902184364,"peak":0.6658339559930181,"peak_dbfs":-3.5326812119931565,"crest_factor_db":18.69342069019121,"mean":3.1316183140935895e-06,"zero_energy":false},"rms_difference_db":0.8066017840020265},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":1.0015349162424713e-06,"output_rms_dbfs":-59.993339054147356,"output_zero_energy":false,"output_peak":0.03300357361540854,"output_energy_fraction":4.1525158884344456e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":123,"both_estimated_voiced_frames":82,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.9111111111111111,"absolute_difference_cents_median":4.230996755143632,"absolute_difference_cents_p95":15.360202074825263,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023158079660483383,"log_rms_mae_db":1.6837232034672411,"rms_envelope_nrmse":0.16775747271498365,"envelope_correlation":0.983768006056897,"rms_gain_error_db":0.8102887325784742,"reference_mean_power":0.004973550834679421,"output_mean_power":0.005993705974161114},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9210188012408756e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006213961937331268,"output_power":0.0007454231276297302,"reference_active":true,"delta_db":0.7903427440121714},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026574930428793645,"output_power":0.00029637918749629355,"reference_active":true,"delta_db":0.4737556686410189},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.273924181429359e-05,"output_power":2.634487214968613e-05,"reference_active":true,"delta_db":-0.9437252235847831}],"active_band_count":3,"active_band_mae_db":0.7359412120793246}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000211279420280134,"log_rms_mae_db":1.1578350486105422,"rms_envelope_nrmse":0.12340131684372517,"envelope_correlation":0.9947410507100272,"rms_gain_error_db":0.812136175812272,"reference_mean_power":0.0049805093532051875,"output_mean_power":0.006004645566687166},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7395423077013874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005644300449164409,"output_power":0.0006732835486336141,"reference_active":true,"delta_db":0.7658787938880036},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013986675109572898,"output_power":0.00015728153812712725,"reference_active":true,"delta_db":0.5096326071138433},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4974533271774746e-06,"output_power":1.49214357126246e-06,"reference_active":true,"delta_db":-2.2368676865690897}],"active_band_count":3,"active_band_mae_db":1.1707930291903121}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020019856693346264,"log_rms_mae_db":1.6824051327987397,"rms_envelope_nrmse":0.19446051932908182,"envelope_correlation":0.9828384417658633,"rms_gain_error_db":0.98357312217821,"reference_mean_power":0.0028106313477950454,"output_mean_power":0.003525016855445479},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3651110338311905e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046335294017358356,"output_power":0.0005719680313851502,"reference_active":true,"delta_db":0.9145983259824211},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024122957535154208,"output_power":0.00027918970872323884,"reference_active":true,"delta_db":0.6346885329012525},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9488783219041956e-05,"output_power":1.710769858201637e-05,"reference_active":true,"delta_db":-2.364652629622131}],"active_band_count":3,"active_band_mae_db":1.3046464961686015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018155803316515472,"log_rms_mae_db":1.1668549116832132,"rms_envelope_nrmse":0.14999933083105196,"envelope_correlation":0.9936359153335597,"rms_gain_error_db":0.9846258920466865,"reference_mean_power":0.002819303754025505,"output_mean_power":0.003536750786905656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2479023732657914e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004464610508434936,"output_power":0.0005408617452981971,"reference_active":true,"delta_db":0.8330268811128044},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012910816269432612,"output_power":0.00014988814499973418,"reference_active":true,"delta_db":0.6481358401725001},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.446292305046822e-06,"output_power":1.2531207501363338e-06,"reference_active":true,"delta_db":-2.9051542779266364}],"active_band_count":3,"active_band_mae_db":1.4621056664039802}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":1.7256042888644185,"rms_envelope_nrmse":0.17427868873697633,"envelope_correlation":0.9830361662677916,"rms_gain_error_db":0.6389363937572902,"reference_mean_power":0.0002950078039052426,"output_mean_power":0.0003417646531412098},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.385626430084388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.894665895613549e-05,"output_power":0.00011516993648439981,"reference_active":true,"delta_db":0.6593799297173503},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2790160745942646e-05,"output_power":2.4138859321945847e-05,"reference_active":true,"delta_db":0.24969355325422404},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.304247232293747e-06,"output_power":2.918160189335831e-06,"reference_active":true,"delta_db":-1.6878808052897885}],"active_band_count":3,"active_band_mae_db":0.8656514294204541}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":1.1116472076653692,"rms_envelope_nrmse":0.11412189888399968,"envelope_correlation":0.9946762711254631,"rms_gain_error_db":0.635756383612171,"reference_mean_power":0.00029237164364699544,"output_mean_power":0.00033846275690659007},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2489747353423385e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.236418785147835e-05,"output_power":0.00010654974271891544,"reference_active":true,"delta_db":0.6204878949753615},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2735133445304275e-05,"output_power":1.3292228638342004e-05,"reference_active":true,"delta_db":0.18594302843077615},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.686566631139435e-07,"output_power":1.4664422778326243e-07,"reference_active":true,"delta_db":-2.6293264330051414}],"active_band_count":3,"active_band_mae_db":1.145252452137093}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":1.7840749781984515,"rms_envelope_nrmse":0.2014907304677014,"envelope_correlation":0.9747618135008143,"rms_gain_error_db":0.3075387007296709,"reference_mean_power":6.736853190004855e-05,"output_mean_power":7.231209656717205e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6302348178615176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0381998821210498e-05,"output_power":2.4233235423569636e-05,"reference_active":true,"delta_db":0.7516462932261188},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.955680541944289e-06,"output_power":4.591928827935041e-06,"reference_active":true,"delta_db":0.6477393712952257},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.59401259091449e-07,"output_power":9.378481722783374e-07,"reference_active":true,"delta_db":1.5298276430610436}],"active_band_count":3,"active_band_mae_db":0.976404435860796}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":1.1367567144968656,"rms_envelope_nrmse":0.12247871072942361,"envelope_correlation":0.9907892938549427,"rms_gain_error_db":0.30885446390609606,"reference_mean_power":6.76432550540454e-05,"output_mean_power":7.26289799126903e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4309026400574305e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.987521517112667e-05,"output_power":2.3583568842217394e-05,"reference_active":true,"delta_db":0.7429768735424057},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.820398831999747e-06,"output_power":2.1100348444455426e-06,"reference_active":true,"delta_db":0.6412307897472282},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.685144573008776e-08,"output_power":6.056430891376471e-08,"reference_active":true,"delta_db":2.1576223547947193}],"active_band_count":3,"active_band_mae_db":1.1806100060281177}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"facodec","status":"measured","mos":{"status":"measured","value":4.208195686340332,"predicted_mos":4.208195686340332,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4983196281d16a546102c365386526d5875b2fbe24d2a61e949e03f20853d2a5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2892590449991985},"started_at":"2026-10-13T07:00:09.345446+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.5146000552922487e-05,"native_input_samples":96000,"native_input_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","output_sha256":"7feff17a2d9156a0c39ea55e2774c77dae9c291753de2242ba9a85dd3b3b7fec","output_samples":96000,"length_delta_samples":0,"analysis_s":1.418329176000043,"synthesis_s":7.004642485000659,"total_s":8.422971661000702,"rtf":1.4038286101667836,"output_peak":0.6101133227348328,"output_rms":0.06595069429750655,"finished_at":"2026-10-13T07:00:17.783504+00:00","evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"7feff17a2d9156a0c39ea55e2774c77dae9c291753de2242ba9a85dd3b3b7fec","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.503075132138998,"spectral_convergence":0.24657048553569602,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.890451848176383,"spectral_convergence":0.24881499656305187,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.010573357311187,"spectral_convergence":0.27465713321618596,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0006200215420365461,"delta_db":-0.32851942799571104,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0028124089863094262,"delta_db":-0.49186749985210304,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0005897289349403452,"delta_db":-1.0373625029541247,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.00020640123313992316,"delta_db":-0.786767887990721,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":8.414295857209871e-05,"delta_db":-1.296855750368254,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":3.6790423324822074e-05,"delta_db":-0.989334426112817,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8217845825456217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5829088747743546,"pitch":null,"output_rms_dbfs":-23.615612560960745,"output_peak":0.6101133227348328,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.801366779208856},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.503075132138998,"spectral_convergence":0.24657048553569602,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.890451848176383,"spectral_convergence":0.24881499656305187,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.010573357311187,"spectral_convergence":0.27465713321618596,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0006200215420365461,"delta_db":-0.32851942799571104,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0028124089863094262,"delta_db":-0.49186749985210304,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0005897289349403452,"delta_db":-1.0373625029541247,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.00020640123313992316,"delta_db":-0.786767887990721,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":8.414295857209871e-05,"delta_db":-1.296855750368254,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":3.6790423324822074e-05,"delta_db":-0.989334426112817,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8217845825456217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5829088747743546,"pitch":null,"output_rms_dbfs":-23.615612560960745,"output_peak":0.6101133227348328,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.801366779208856},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8823339033098802,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705285273351481,"rms_dbfs":-23.03270368618639,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.894584102403467,"mean":-1.7856280008951823e-05,"zero_energy":false},"output":{"rms":0.06595069429750655,"rms_dbfs":-23.615612560960745,"peak":0.6101133227348328,"peak_dbfs":-4.29178982871975,"crest_factor_db":19.323822732240995,"mean":-1.837429725957899e-05,"zero_energy":false},"rms_difference_db":-0.582908874774354},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":5.103081871795643e-07,"output_rms_dbfs":-62.92167463965663,"output_zero_energy":false,"output_peak":0.00671922555193305,"output_energy_fraction":2.9135924976002725e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":95,"both_estimated_voiced_frames":84,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9333333333333333,"absolute_difference_cents_median":2.690009594876095,"absolute_difference_cents_p95":8.462115499621522,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023158079660483383,"log_rms_mae_db":1.2429944446230814,"rms_envelope_nrmse":0.11649110453301287,"envelope_correlation":0.9904397265945483,"rms_gain_error_db":-0.5806886362315407,"reference_mean_power":0.004973550834679421,"output_mean_power":0.004351086303730759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9210188012408756e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006213961937331268,"output_power":0.0005375007099866918,"reference_active":true,"delta_db":-0.6298954647378499},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026574930428793645,"output_power":0.0002238304994949564,"reference_active":true,"delta_db":-0.7455287237407919},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.273924181429359e-05,"output_power":2.9803313904849074e-05,"reference_active":true,"delta_db":-0.4080406061508023}],"active_band_count":3,"active_band_mae_db":0.5944882648764813}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000211279420280134,"log_rms_mae_db":1.0543626060760718,"rms_envelope_nrmse":0.10314509708292798,"envelope_correlation":0.9929549900095872,"rms_gain_error_db":-0.5754432751652614,"reference_mean_power":0.0049805093532051875,"output_mean_power":0.004362439655315772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7395423077013874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005644300449164409,"output_power":0.0004875243974350871,"reference_active":true,"delta_db":-0.636137692263493},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013986675109572898,"output_power":0.00011603061804063896,"reference_active":true,"delta_db":-0.8114188142609793},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4974533271774746e-06,"output_power":2.1944809983458798e-06,"reference_active":true,"delta_db":-0.5616555609448239}],"active_band_count":3,"active_band_mae_db":0.6697373558230987}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020019856693346264,"log_rms_mae_db":1.2607811731831549,"rms_envelope_nrmse":0.13653883907106837,"envelope_correlation":0.9884886684458777,"rms_gain_error_db":-0.6788313724403752,"reference_mean_power":0.0028106313477950454,"output_mean_power":0.002403924084039818},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3651110338311905e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046335294017358356,"output_power":0.00040755431278648597,"reference_active":true,"delta_db":-0.5572642989913859},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024122957535154208,"output_power":0.00019219684021631205,"reference_active":true,"delta_db":-0.9868430888906948},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9488783219041956e-05,"output_power":2.4647517243280948e-05,"reference_active":true,"delta_db":-0.7788367372844807}],"active_band_count":3,"active_band_mae_db":0.7743147083888537}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018155803316515472,"log_rms_mae_db":0.9809149822904446,"rms_envelope_nrmse":0.11698411811701497,"envelope_correlation":0.9922949157394094,"rms_gain_error_db":-0.6752495162477069,"reference_mean_power":0.002819303754025505,"output_mean_power":0.0024133311468175433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2479023732657914e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004464610508434936,"output_power":0.0003924139676298575,"reference_active":true,"delta_db":-0.5603911981628033},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012910816269432612,"output_power":0.00010164138319678128,"reference_active":true,"delta_db":-1.038831342758969},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.446292305046822e-06,"output_power":1.8965494547021914e-06,"reference_active":true,"delta_db":-1.105441773055762}],"active_band_count":3,"active_band_mae_db":0.9015547713258448}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":1.5902750714827578,"rms_envelope_nrmse":0.18047252073450606,"envelope_correlation":0.9794215779310178,"rms_gain_error_db":-0.7698975109784159,"reference_mean_power":0.0002950078039052426,"output_mean_power":0.00024708350507384277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.385626430084388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.894665895613549e-05,"output_power":8.778197874844947e-05,"reference_active":true,"delta_db":-0.5199576803167043},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2790160745942646e-05,"output_power":2.6374931345771858e-05,"reference_active":true,"delta_db":0.6344394943165432},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.304247232293747e-06,"output_power":5.131122626144054e-06,"reference_active":true,"delta_db":0.7631518488712736}],"active_band_count":3,"active_band_mae_db":0.6391830078348404}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":1.1831524964806523,"rms_envelope_nrmse":0.15046736006667613,"envelope_correlation":0.9864629253070253,"rms_gain_error_db":-0.7593564381089052,"reference_mean_power":0.00029237164364699544,"output_mean_power":0.00024547066855826855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2489747353423385e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.236418785147835e-05,"output_power":8.257928399253719e-05,"reference_active":true,"delta_db":-0.48632502896911756},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2735133445304275e-05,"output_power":1.4569033351319316e-05,"reference_active":true,"delta_db":0.5842723740088984},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.686566631139435e-07,"output_power":3.6719572820884573e-07,"reference_active":true,"delta_db":1.357000039360512}],"active_band_count":3,"active_band_mae_db":0.809199147446176}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":1.9200023670540387,"rms_envelope_nrmse":0.26499216809300297,"envelope_correlation":0.9645273072974536,"rms_gain_error_db":-1.5102208247162396,"reference_mean_power":6.736853190004855e-05,"output_mean_power":4.7581157276260263e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6302348178615176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0381998821210498e-05,"output_power":1.6796721187463864e-05,"reference_active":true,"delta_db":-0.8402225881319925},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.955680541944289e-06,"output_power":4.3775809686177265e-06,"reference_active":true,"delta_db":0.4401297661190448},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.59401259091449e-07,"output_power":1.5405759574764256e-06,"reference_active":true,"delta_db":3.685333438251535}],"active_band_count":3,"active_band_mae_db":1.6552285975008576}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":1.5879747432477134,"rms_envelope_nrmse":0.2357085664618178,"envelope_correlation":0.9750328387750405,"rms_gain_error_db":-1.5117721720294275,"reference_mean_power":6.76432550540454e-05,"output_mean_power":4.77581264247555e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4309026400574305e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.987521517112667e-05,"output_power":1.7388581198224032e-05,"reference_active":true,"delta_db":-0.5804769139349865},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.820398831999747e-06,"output_power":2.074340302940512e-06,"reference_active":true,"delta_db":0.5671345727643596},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.685144573008776e-08,"output_power":1.0156931117613839e-07,"reference_active":true,"delta_db":4.403079766929455}],"active_band_count":3,"active_band_mae_db":1.850230417876267}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.9567956924438477,"predicted_mos":3.9567956924438477,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"1cd3405f8afcc28a03c954ccb160e9c622ce8dac86a9e9f65372072aa21da73f","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.845470380001643},"started_at":"2026-10-13T07:30:18.195327+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00733616199795506,"native_input_samples":264600,"native_input_sha256":"ae7be62e91534d15bd1dfa3e5fa044c7744aa43459896805e9e394def91b170f","output_sha256":"7ab790aba8180f8ca33ac178c6a8447b1cc618b1e3c131de8068a9e3476fd52d","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13904909299890278,"synthesis_s":4.594570550001663,"total_s":4.733619643000566,"rtf":0.788936607166761,"output_peak":0.5160439014434814,"output_rms":0.06254729026557097,"finished_at":"2026-10-13T07:30:22.951565+00:00","evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"7ab790aba8180f8ca33ac178c6a8447b1cc618b1e3c131de8068a9e3476fd52d","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.071737034992125,"output_peak":0.5155635959156059},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.445620502382656,"spectral_convergence":0.5271294731392144,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.285821527214045,"spectral_convergence":0.40569062639663,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.877106795495983,"spectral_convergence":0.4885310095385677,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0005946808917207199,"delta_db":-0.5097473771471916,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.002453368705047694,"delta_db":-1.0850240577230084,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0005678859234547737,"delta_db":-1.201275834871882,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.0001574258897119495,"delta_db":-1.963129199952593,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":0.0001148906322764431,"delta_db":0.055812661676223264,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":2.650995399684711e-05,"delta_db":-2.4125925353573234,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2045969444547036,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0402426776526954,"pitch":null,"output_rms_dbfs":-24.072946363839087,"output_peak":0.5155635959156059,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.202849608364227},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7140912366538184,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705285273351481,"rms_dbfs":-23.03270368618639,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.894584102403467,"mean":-1.7856280008951823e-05,"zero_energy":false},"output":{"rms":0.06256805891354172,"rms_dbfs":-24.072946363839087,"peak":0.5155635959156059,"peak_dbfs":-5.75435511772783,"crest_factor_db":18.318591246111257,"mean":-0.0001592431200561887,"zero_energy":false},"rms_difference_db":-1.0402426776526958},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":9.959949460415264e-07,"output_rms_dbfs":-60.01742865303054,"output_zero_energy":false,"output_peak":0.0464353483561567,"output_energy_fraction":6.318104272321057e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":84,"both_estimated_voiced_frames":58,"reference_voiced_output_unvoiced_frames":32,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.6444444444444445,"absolute_difference_cents_median":10.26319539888541,"absolute_difference_cents_p95":78.2454519731181,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023158079660483383,"log_rms_mae_db":1.7845563153439516,"rms_envelope_nrmse":0.15681964519074443,"envelope_correlation":0.987598029246576,"rms_gain_error_db":-1.0364983887888048,"reference_mean_power":0.004973550834679421,"output_mean_power":0.003917569612729027},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9210188012408756e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006213961937331268,"output_power":0.0005054083654588008,"reference_active":true,"delta_db":-0.8972616268931211},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026574930428793645,"output_power":0.00021046433001678064,"reference_active":true,"delta_db":-1.0129363507450515},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.273924181429359e-05,"output_power":2.420339799096789e-05,"reference_active":true,"delta_db":-1.3119227543803187}],"active_band_count":3,"active_band_mae_db":1.0740402440061638}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000211279420280134,"log_rms_mae_db":1.4277347481478393,"rms_envelope_nrmse":0.13174872455744935,"envelope_correlation":0.9948155474998917,"rms_gain_error_db":-1.0330588272071901,"reference_mean_power":0.0049805093532051875,"output_mean_power":0.0039261589439542445},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7395423077013874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005644300449164409,"output_power":0.00045932893577744333,"reference_active":true,"delta_db":-0.8948631843509759},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013986675109572898,"output_power":0.00011324490789182398,"reference_active":true,"delta_db":-0.9169580408178496},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4974533271774746e-06,"output_power":1.5560751907461494e-06,"reference_active":true,"delta_db":-2.0546680220344573}],"active_band_count":3,"active_band_mae_db":1.288829749067761}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020019856693346264,"log_rms_mae_db":2.5560140996379954,"rms_envelope_nrmse":0.1970051808340957,"envelope_correlation":0.975743386567282,"rms_gain_error_db":-0.9744691677145477,"reference_mean_power":0.0028106313477950454,"output_mean_power":0.002245727044052704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3651110338311905e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046335294017358356,"output_power":0.000407936319994385,"reference_active":true,"delta_db":-0.5531954935899777},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024122957535154208,"output_power":0.00018124633696344438,"reference_active":true,"delta_db":-1.2416131420114958},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9488783219041956e-05,"output_power":2.0515304059516673e-05,"reference_active":true,"delta_db":-1.5757889476922706}],"active_band_count":3,"active_band_mae_db":1.1235325277645813}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018155803316515472,"log_rms_mae_db":2.3062574434299656,"rms_envelope_nrmse":0.16762090893305234,"envelope_correlation":0.9844115461394455,"rms_gain_error_db":-0.9791327295833174,"reference_mean_power":0.002819303754025505,"output_mean_power":0.0022502387372016546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2479023732657914e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004464610508434936,"output_power":0.0003787115161547067,"reference_active":true,"delta_db":-0.7147506544277948},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012910816269432612,"output_power":0.00010041052542687544,"reference_active":true,"delta_db":-1.0917446117987315},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.446292305046822e-06,"output_power":1.3797140777654438e-06,"reference_active":true,"delta_db":-2.487192536250152}],"active_band_count":3,"active_band_mae_db":1.431229267492226}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":3.117247944780639,"rms_envelope_nrmse":0.2888210431842001,"envelope_correlation":0.9491574418364389,"rms_gain_error_db":-1.5124316488128209,"reference_mean_power":0.0002950078039052426,"output_mean_power":0.0002082525556568003},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.385626430084388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.894665895613549e-05,"output_power":7.394029429444198e-05,"reference_active":true,"delta_db":-1.2651995955086461},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2790160745942646e-05,"output_power":1.7587676496081822e-05,"reference_active":true,"delta_db":-1.1253891970003829},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.304247232293747e-06,"output_power":2.4262190850670247e-06,"reference_active":true,"delta_db":-2.489671942372554}],"active_band_count":3,"active_band_mae_db":1.626753578293861}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":2.5092520738561044,"rms_envelope_nrmse":0.2445385139514737,"envelope_correlation":0.9670261541025514,"rms_gain_error_db":-1.4790960559640114,"reference_mean_power":0.00029237164364699544,"output_mean_power":0.00020798194899647498},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2489747353423385e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.236418785147835e-05,"output_power":6.929917855291933e-05,"reference_active":true,"delta_db":-1.2477552920713377},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2735133445304275e-05,"output_power":9.695149218270526e-06,"reference_active":true,"delta_db":-1.1844900238252625},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.686566631139435e-07,"output_power":2.009819725996679e-07,"reference_active":true,"delta_db":-1.2604051180700746}],"active_band_count":3,"active_band_mae_db":1.2308834779888917}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":4.393766844422459,"rms_envelope_nrmse":0.26371158393238436,"envelope_correlation":0.9592468811192547,"rms_gain_error_db":-1.0946409073052035,"reference_mean_power":6.736853190004855e-05,"output_mean_power":5.235919882279777e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6302348178615176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0381998821210498e-05,"output_power":1.4687087903538966e-05,"reference_active":true,"delta_db":-1.423110779659478},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.955680541944289e-06,"output_power":2.7180054491631456e-06,"reference_active":true,"delta_db":-1.6297088792808818},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.59401259091449e-07,"output_power":6.311265440697848e-07,"reference_active":true,"delta_db":-0.1903332585086255}],"active_band_count":3,"active_band_mae_db":1.081050972482995}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":4.1473584813241,"rms_envelope_nrmse":0.22408545854129336,"envelope_correlation":0.972316559651528,"rms_gain_error_db":-1.092586311410806,"reference_mean_power":6.76432550540454e-05,"output_mean_power":5.259759260538908e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4309026400574305e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.987521517112667e-05,"output_power":1.4006681946727059e-05,"reference_active":true,"delta_db":-1.5197657194307002},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.820398831999747e-06,"output_power":1.406800037630201e-06,"reference_active":true,"delta_db":-1.1193417692162666},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.685144573008776e-08,"output_power":4.400839174844019e-08,"reference_active":true,"delta_db":0.770809674416558}],"active_band_count":3,"active_band_mae_db":1.1366390543545082}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"golf","status":"measured","mos":{"status":"measured","value":3.774672031402588,"predicted_mos":3.774672031402588,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"7c12459e65ef177e6dd48643fd867626c28d4bf9437af7a3f0bad79f4d6672fa","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1539895870009786},"started_at":"2026-10-13T05:03:02.044430+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00341974000002665,"native_input_samples":144000,"native_input_sha256":"1c17891f2d343adbc9ace47da0135345c36d49b149e8951d684a86507b692b75","output_sha256":"545c34e2b0b19f16cdf227e1cdf40f6bceda08b03999f44de3f6d5829dd42134","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5992416219996812,"synthesis_s":0.14449903999957314,"total_s":0.7437406619992544,"rtf":0.12395677699987573,"output_peak":0.5362288951873779,"output_rms":0.06456996277257955,"finished_at":"2026-10-13T05:03:02.804882+00:00","analysis_stages":{"external_f0_s":0.09499916099957773},"evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"545c34e2b0b19f16cdf227e1cdf40f6bceda08b03999f44de3f6d5829dd42134","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.796408885255694,"output_peak":0.5365930691909417},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.06786472466046,"spectral_convergence":0.6538212217729638,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.58641885454368,"spectral_convergence":0.3455478400238722,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.740728541699835,"spectral_convergence":0.34264767906187094,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006655604329500953,"output_power":0.0005903949867865001,"delta_db":-0.5204483446964254,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031507984352875298,"output_power":0.0026992095043254748,"delta_db":-0.6718402665872403,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007490549704583301,"output_power":0.0006262400444307604,"delta_db":-0.7777285539329515,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002457881862878136,"output_power":0.0001399619876206288,"delta_db":-2.4455090352514155,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011357576773544126,"output_power":8.080511238313995e-05,"delta_db":-1.4784684257122171,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.625686079048844e-05,"output_power":3.553065396424137e-05,"delta_db":-1.1457295512702816,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.173287362908422,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7608767917927037,"pitch":null,"output_rms_dbfs":-23.796408885255694,"output_peak":0.5365930691909417,"reference_rms_dbfs":-23.03553209346299,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.131670706967991},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8542128569377966,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07050556469889237,"rms_dbfs":-23.03553209346299,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.897412509680066,"mean":-2.4188619623200126e-05,"zero_energy":false},"output":{"rms":0.06459212250352782,"rms_dbfs":-23.796408885255694,"peak":0.5365930691909417,"peak_dbfs":-5.407098822307619,"crest_factor_db":18.389310062948073,"mean":0.0010253239849031877,"zero_energy":false},"rms_difference_db":-0.760876791792704},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":4.536649495725646e-07,"output_rms_dbfs":-63.43264773291283,"output_zero_energy":false,"output_peak":0.009102981236795987,"output_energy_fraction":2.704802513009256e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":141,"both_estimated_voiced_frames":85,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":56,"paired_voiced_fraction":0.9444444444444444,"absolute_difference_cents_median":5.143594627358358,"absolute_difference_cents_p95":15.952246170310032,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00023158079660483383,"log_rms_mae_db":2.2566904670195838,"rms_envelope_nrmse":0.15644562253123026,"envelope_correlation":0.9824834197374848,"rms_gain_error_db":-0.7574777275243176,"reference_mean_power":0.00496931452775427,"output_mean_power":0.004173964135447828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.925521820579624e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006232265795729316,"output_power":0.0005517038887264019,"reference_active":true,"delta_db":-0.529399218680504},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026470349102031585,"output_power":0.0001999550797290988,"reference_active":true,"delta_db":-1.2182722741703864},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.274421410769972e-05,"output_power":2.0032082096839933e-05,"reference_active":true,"delta_db":-2.1340848001989747}],"active_band_count":3,"active_band_mae_db":1.293918764349955}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.000211279420280134,"log_rms_mae_db":1.9272742074374059,"rms_envelope_nrmse":0.12207904611956812,"envelope_correlation":0.9917606099689078,"rms_gain_error_db":-0.7569161592929021,"reference_mean_power":0.004977536158493307,"output_mean_power":0.0041814105209359385},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.7438941904455022e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005658888785527509,"output_power":0.0005014017813146191,"reference_active":true,"delta_db":-0.5254528627480038},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013874837766516965,"output_power":0.00010204290918428736,"reference_active":true,"delta_db":-1.3344508217677378},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5026055768516805e-06,"output_power":1.4260785657882123e-06,"reference_active":true,"delta_db":-2.4424895561214015}],"active_band_count":3,"active_band_mae_db":1.4341310802123808}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00020019856693346264,"log_rms_mae_db":2.0961431647900937,"rms_envelope_nrmse":0.17890687568570512,"envelope_correlation":0.9781741847475257,"rms_gain_error_db":-0.7246266418796031,"reference_mean_power":0.002811803837547388,"output_mean_power":0.0023797007919350017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3678785675053516e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004650978767557788,"output_power":0.0004174582532537866,"reference_active":true,"delta_db":-0.46931305309287963},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024027627325205717,"output_power":0.00017727348074892803,"reference_active":true,"delta_db":-1.3206711519004193},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.949657708850432e-05,"output_power":1.456970348674876e-05,"reference_active":true,"delta_db":-3.063209080981872}],"active_band_count":3,"active_band_mae_db":1.617731095325057}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00018155803316515472,"log_rms_mae_db":1.6711084613222777,"rms_envelope_nrmse":0.13872602233701206,"envelope_correlation":0.9885996407336456,"rms_gain_error_db":-0.7231291094661743,"reference_mean_power":0.0028201129257798456,"output_mean_power":0.002387556117895844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2504945768169677e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004477722947969885,"output_power":0.0003966591272133665,"reference_active":true,"delta_db":-0.5263976684135576},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001281507086603547,"output_power":9.236813148614888e-05,"reference_active":true,"delta_db":-1.4219885375738894},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4515779049880845e-06,"output_power":1.1739698669991298e-06,"reference_active":true,"delta_db":-3.1978874883607205}],"active_band_count":3,"active_band_mae_db":1.7154245647827224}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":2.268479572161607,"rms_envelope_nrmse":0.2888614458346414,"envelope_correlation":0.9700849414550873,"rms_gain_error_db":-2.2603959455401683,"reference_mean_power":0.0002925794289999848,"output_mean_power":0.00017386180875280624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.3929765596275995e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.930672076966323e-05,"output_power":5.8446473080610895e-05,"reference_active":true,"delta_db":-2.3022033215485047},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2728039295660763e-05,"output_power":1.2578335235698056e-05,"reference_active":true,"delta_db":-2.569388062982479},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.308603875566491e-06,"output_power":2.1384426429097926e-06,"reference_active":true,"delta_db":-3.04238961609789}],"active_band_count":3,"active_band_mae_db":2.6379936668762913}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":1.910605389742015,"rms_envelope_nrmse":0.2550610401983494,"envelope_correlation":0.9865765698045099,"rms_gain_error_db":-2.2464075685134817,"reference_mean_power":0.00029059530832726155,"output_mean_power":0.00017323986693020568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.2559296428814565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.266528488395731e-05,"output_power":5.457736340409547e-05,"reference_active":true,"delta_db":-2.2990451387947877},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.266704999177978e-05,"output_power":6.88505271647087e-06,"reference_active":true,"delta_db":-2.647682146903856},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6921628223358674e-07,"output_power":1.251232215126118e-07,"reference_active":true,"delta_db":-3.3276340501114103}],"active_band_count":3,"active_band_mae_db":2.758120445270018}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":2.8194785414641665,"rms_envelope_nrmse":0.22238998395731333,"envelope_correlation":0.9720701973601522,"rms_gain_error_db":-0.9054794350769001,"reference_mean_power":6.743460551287761e-05,"output_mean_power":5.4743792109646905e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.6268938196154006e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.034874125383778e-05,"output_power":2.0699284865292413e-05,"reference_active":true,"delta_db":0.07417791867431367},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9442294917340445e-06,"output_power":3.5386016594226433e-06,"reference_active":true,"delta_db":-0.4713049923389178},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.572577199206543e-07,"output_power":3.6150262874933914e-07,"reference_active":true,"delta_db":-2.596242361663026}],"active_band_count":3,"active_band_mae_db":1.0472417575587525}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":2.5331748030268644,"rms_envelope_nrmse":0.18914559767668385,"envelope_correlation":0.981158498990102,"rms_gain_error_db":-0.9064622509113529,"reference_mean_power":6.772743327770519e-05,"output_mean_power":5.4969070334227166e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.428239468753105e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9843168121646968e-05,"output_power":2.02457298654232e-05,"reference_active":true,"delta_db":0.08722425883219675},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.8087564644441711e-06,"output_power":1.577153356693557e-06,"reference_active":true,"delta_db":-0.595061716777008},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.669775565204855e-08,"output_power":2.4037330959965353e-08,"reference_active":true,"delta_db":-1.8375326151434974}],"active_band_count":3,"active_band_mae_db":0.8399395302509007}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.8699467182159424,"predicted_mos":3.8699467182159424,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"aafa4f93a859a268d8e2a0cf16372167fe176c5d73bd6a923ba941a87e700b69","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4507782540022163},"started_at":"2026-10-13T07:45:32.772219+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030012220013304614,"native_input_samples":144000,"native_input_sha256":"1c17891f2d343adbc9ace47da0135345c36d49b149e8951d684a86507b692b75","output_sha256":"95edd0b6d7592a545d8e8a3005cbe89b684b7e0e4bd750eef3f5a3657e52d774","output_samples":144120,"length_delta_samples":120,"analysis_s":2.6801836289996572,"synthesis_s":3.296043375001318,"total_s":5.976227004000975,"rtf":0.9960378340001625,"output_peak":0.5651189088821411,"output_rms":0.0683695009327803,"finished_at":"2026-10-13T07:45:38.762670+00:00","analysis_stages":{"external_f0_s":1.8315230980006163},"evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"95edd0b6d7592a545d8e8a3005cbe89b684b7e0e4bd750eef3f5a3657e52d774","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.300048500083864,"output_peak":0.5781404561283372},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.666568481370519,"spectral_convergence":0.6479158111066015,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.4602235398796175,"spectral_convergence":0.37058605691547725,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.664838685859188,"spectral_convergence":0.3525898641370116,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0008448178126062007,"delta_db":1.0150434049832053,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.002867697259842157,"delta_db":-0.4073192557664796,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.000604661869612626,"delta_db":-0.9287610460884221,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.0002168474314218025,"delta_db":-0.5723479418914006,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":0.00010784513023874565,"delta_db":-0.21902814069688864,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":3.770680012117019e-05,"delta_db":-0.8824854881040718,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6708308795884114,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.26523098228838016,"pitch":null,"output_rms_dbfs":-23.29793466847477,"output_peak":0.5781404561283372,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.930543569036441},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8559071336739472,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705285273351481,"rms_dbfs":-23.03270368618639,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.894584102403467,"mean":-1.7856280008951823e-05,"zero_energy":false},"output":{"rms":0.06840742871825181,"rms_dbfs":-23.29793466847477,"peak":0.5781404561283372,"peak_dbfs":-4.759332784836584,"crest_factor_db":18.538601883638187,"mean":1.5024820959195571e-05,"zero_energy":false},"rms_difference_db":-0.2652309822883794},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":3.616393147334118e-06,"output_rms_dbfs":-54.41724362367979,"output_zero_energy":false,"output_peak":0.059312349880025116,"output_energy_fraction":0.0001919128798442376,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":105,"both_estimated_voiced_frames":81,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9,"absolute_difference_cents_median":8.747002420445146,"absolute_difference_cents_p95":18.5910926694529,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023158079660483383,"log_rms_mae_db":2.0542221762937847,"rms_envelope_nrmse":0.16820009218438348,"envelope_correlation":0.9727537398365779,"rms_gain_error_db":-0.26463657611204483,"reference_mean_power":0.004973550834679421,"output_mean_power":0.004679537195289372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9210188012408756e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006213961937331268,"output_power":0.0006008099897107233,"reference_active":true,"delta_db":-0.14631443621411389},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026574930428793645,"output_power":0.0002236876897473165,"reference_active":true,"delta_db":-0.748300521131746},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.273924181429359e-05,"output_power":2.1715296490465434e-05,"reference_active":true,"delta_db":-1.7830285427181158}],"active_band_count":3,"active_band_mae_db":0.8925478333546586}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000211279420280134,"log_rms_mae_db":1.660813916054148,"rms_envelope_nrmse":0.10527218204904651,"envelope_correlation":0.9891172972135053,"rms_gain_error_db":-0.2653364376038959,"reference_mean_power":0.0049805093532051875,"output_mean_power":0.004685329260675857},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7395423077013874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005644300449164409,"output_power":0.0005466124494774789,"reference_active":true,"delta_db":-0.1393060460049547},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013986675109572898,"output_power":0.00011840225003494965,"reference_active":true,"delta_db":-0.7235453146009214},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4974533271774746e-06,"output_power":1.5783744069055175e-06,"reference_active":true,"delta_db":-1.9928735052242197}],"active_band_count":3,"active_band_mae_db":0.951908288610032}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020019856693346264,"log_rms_mae_db":2.0157524164098883,"rms_envelope_nrmse":0.20306514840867618,"envelope_correlation":0.9674091678146656,"rms_gain_error_db":-0.5516768042070913,"reference_mean_power":0.0028106313477950454,"output_mean_power":0.0024753476690232367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3651110338311905e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046335294017358356,"output_power":0.00040706615977711936,"reference_active":true,"delta_db":-0.5624692303872557},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024122957535154208,"output_power":0.0002015888237347181,"reference_active":true,"delta_db":-0.7796410156105499},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9488783219041956e-05,"output_power":2.0401950432900513e-05,"reference_active":true,"delta_db":-1.5998516474468167}],"active_band_count":3,"active_band_mae_db":0.9806539644815407}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018155803316515472,"log_rms_mae_db":1.5074828762822978,"rms_envelope_nrmse":0.1329419424758643,"envelope_correlation":0.9870251747613958,"rms_gain_error_db":-0.5518369819408686,"reference_mean_power":0.002819303754025505,"output_mean_power":0.002482893956727764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2479023732657914e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004464610508434936,"output_power":0.00039189702231790777,"reference_active":true,"delta_db":-0.5661161346181944},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012910816269432612,"output_power":0.00010803679570540193,"reference_active":true,"delta_db":-0.7738200599303477},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.446292305046822e-06,"output_power":1.646085614806104e-06,"reference_active":true,"delta_db":-1.7205592959954483}],"active_band_count":3,"active_band_mae_db":1.0201651635146636}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":2.2078844306176806,"rms_envelope_nrmse":0.21948370477288806,"envelope_correlation":0.9645615797359868,"rms_gain_error_db":-0.5144240666362988,"reference_mean_power":0.0002950078039052426,"output_mean_power":0.0002620541839044158},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.385626430084388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.894665895613549e-05,"output_power":9.080047536366744e-05,"reference_active":true,"delta_db":-0.37313012184815686},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2790160745942646e-05,"output_power":2.066997581134615e-05,"reference_active":true,"delta_db":-0.42407419999824114},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.304247232293747e-06,"output_power":2.1181602689261303e-06,"reference_active":true,"delta_db":-3.0793839131564766}],"active_band_count":3,"active_band_mae_db":1.2921960783342916}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":1.5003055369017557,"rms_envelope_nrmse":0.13874102593972776,"envelope_correlation":0.9863875545800718,"rms_gain_error_db":-0.4881700932770312,"reference_mean_power":0.00029237164364699544,"output_mean_power":0.00026128726317201355},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2489747353423385e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.236418785147835e-05,"output_power":8.47877289815256e-05,"reference_active":true,"delta_db":-0.37170612932003944},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2735133445304275e-05,"output_power":1.141238409125492e-05,"reference_active":true,"delta_db":-0.47627120389313654},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.686566631139435e-07,"output_power":1.300533162766845e-07,"reference_active":true,"delta_db":-3.150761853288383}],"active_band_count":3,"active_band_mae_db":1.3329130621671863}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":2.120796533500648,"rms_envelope_nrmse":0.210035402297537,"envelope_correlation":0.9704234404872559,"rms_gain_error_db":-0.3051121824908771,"reference_mean_power":6.736853190004855e-05,"output_mean_power":6.279800759821388e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6302348178615176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0381998821210498e-05,"output_power":1.999231399378777e-05,"reference_active":true,"delta_db":-0.08383708054618799},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.955680541944289e-06,"output_power":3.4788018845095108e-06,"reference_active":true,"delta_db":-0.5579151435060602},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.59401259091449e-07,"output_power":4.983145422671547e-07,"reference_active":true,"delta_db":-1.216462107383672}],"active_band_count":3,"active_band_mae_db":0.6194047771453067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":1.3869206341452234,"rms_envelope_nrmse":0.1363582457737373,"envelope_correlation":0.9874074827451332,"rms_gain_error_db":-0.30465840082466755,"reference_mean_power":6.76432550540454e-05,"output_mean_power":6.306068122496469e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4309026400574305e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.987521517112667e-05,"output_power":1.9345895786381184e-05,"reference_active":true,"delta_db":-0.11722995110399242},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.820398831999747e-06,"output_power":1.5943311334258715e-06,"reference_active":true,"delta_db":-0.5758802128364207},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.685144573008776e-08,"output_power":3.192377075844537e-08,"reference_active":true,"delta_db":-0.6234034694563173}],"active_band_count":3,"active_band_mae_db":0.43883787779891015}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.82858943939209,"predicted_mos":3.82858943939209,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"e18695ba9cd9c274d2601b65e3a870ab1cb30c12b37a198dbe154bee46384baf","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9027270690003206},"started_at":"2026-10-13T05:11:27.682287+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003064418000576552,"native_input_samples":144000,"native_input_sha256":"1c17891f2d343adbc9ace47da0135345c36d49b149e8951d684a86507b692b75","output_sha256":"862ec950d9ef78113ea5e41e6f1ec70195f3f1c282405242362dd9689d25a85f","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0991981610004586,"synthesis_s":0.12485757300055411,"total_s":1.2240557340010128,"rtf":0.2040092890001688,"output_peak":0.6452924013137817,"output_rms":0.06815005464949885,"finished_at":"2026-10-13T05:11:28.918233+00:00","analysis_stages":{"external_f0_s":0.15642469900012657},"evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"862ec950d9ef78113ea5e41e6f1ec70195f3f1c282405242362dd9689d25a85f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.32791473879172,"output_peak":0.6317274568410011},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.888759714375463,"spectral_convergence":0.22217825113773612,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.553751303580769,"spectral_convergence":0.21517395879937934,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.928275679289293,"spectral_convergence":0.23670766149574765,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0006370166856127529,"delta_db":-0.21107913573390197,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0030081444311094276,"delta_db":-0.19966542568181853,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.0006671084072075725,"delta_db":-0.5019227330757724,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.0002114369353651953,"delta_db":-0.682082210797889,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":9.636615676735395e-05,"delta_db":-0.7077881433778731,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":3.4864533808013556e-05,"delta_db":-1.2228436523099795,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5875635501628724,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.28815809790911817,"pitch":null,"output_rms_dbfs":-23.32086178409551,"output_peak":0.6317274568410011,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.7902622324151745},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9203169256185616,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705285273351481,"rms_dbfs":-23.03270368618639,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.894584102403467,"mean":-1.7856280008951823e-05,"zero_energy":false},"output":{"rms":0.06822709982016176,"rms_dbfs":-23.32086178409551,"peak":0.6317274568410011,"peak_dbfs":-3.9894049381171133,"crest_factor_db":19.331456845978394,"mean":5.692933492842726e-05,"zero_energy":false},"rms_difference_db":-0.2881580979091183},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":1.895755771103978e-06,"output_rms_dbfs":-57.22217613249987,"output_zero_energy":false,"output_peak":0.04827973172183464,"output_energy_fraction":0.00010113549005431149,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":90,"both_estimated_voiced_frames":77,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.8555555555555555,"absolute_difference_cents_median":1.0493614604470498,"absolute_difference_cents_p95":5.9306650016965845,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023158079660483383,"log_rms_mae_db":1.010163025050855,"rms_envelope_nrmse":0.0970783958036587,"envelope_correlation":0.9916864842848453,"rms_gain_error_db":-0.28845979268338373,"reference_mean_power":0.004973550834679421,"output_mean_power":0.004653937878495242},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9210188012408756e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006213961937331268,"output_power":0.0005718695262672074,"reference_active":true,"delta_db":-0.3607163420572807},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026574930428793645,"output_power":0.0001958837411454005,"reference_active":true,"delta_db":-1.3247374631485596},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.273924181429359e-05,"output_power":2.2384447050157324e-05,"reference_active":true,"delta_db":-1.651222469329784}],"active_band_count":3,"active_band_mae_db":1.1122254248452081}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000211279420280134,"log_rms_mae_db":0.7238776223471696,"rms_envelope_nrmse":0.06035847330984327,"envelope_correlation":0.997239556055542,"rms_gain_error_db":-0.28769145678858454,"reference_mean_power":0.0049805093532051875,"output_mean_power":0.004661273805424453},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7395423077013874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005644300449164409,"output_power":0.0005188634516255474,"reference_active":true,"delta_db":-0.3655704305144316},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013986675109572898,"output_power":0.00010173926482696224,"reference_active":true,"delta_db":-1.3822589188620267},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4974533271774746e-06,"output_power":1.5164197979149168e-06,"reference_active":true,"delta_db":-2.166779349985951}],"active_band_count":3,"active_band_mae_db":1.3048695664541363}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020019856693346264,"log_rms_mae_db":1.1354373427213191,"rms_envelope_nrmse":0.12815196476648152,"envelope_correlation":0.9874073725200371,"rms_gain_error_db":-0.36471637767258114,"reference_mean_power":0.0028106313477950454,"output_mean_power":0.0025842364511573005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3651110338311905e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046335294017358356,"output_power":0.0004157989895211408,"reference_active":true,"delta_db":-0.4702849358110669},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024122957535154208,"output_power":0.00018395985701597558,"reference_active":true,"delta_db":-1.1770748898232288},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9488783219041956e-05,"output_power":1.8882978037481125e-05,"reference_active":true,"delta_db":-1.935863648291544}],"active_band_count":3,"active_band_mae_db":1.1944078246419465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018155803316515472,"log_rms_mae_db":0.7146541105674069,"rms_envelope_nrmse":0.08020840161692537,"envelope_correlation":0.9956420532023224,"rms_gain_error_db":-0.3647374583962836,"reference_mean_power":0.002819303754025505,"output_mean_power":0.0025921977169662096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2479023732657914e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004464610508434936,"output_power":0.00039449397789507204,"reference_active":true,"delta_db":-0.537431992083729},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012910816269432612,"output_power":9.603868708871114e-05,"reference_active":true,"delta_db":-1.2850748649604407},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.446292305046822e-06,"output_power":1.46007223323062e-06,"reference_active":true,"delta_db":-2.2413400732645767}],"active_band_count":3,"active_band_mae_db":1.3546156434362489}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":1.3714141748205995,"rms_envelope_nrmse":0.16865446669661122,"envelope_correlation":0.9820400451626514,"rms_gain_error_db":-0.7142294856405791,"reference_mean_power":0.0002950078039052426,"output_mean_power":0.0002502710156384915},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.385626430084388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.894665895613549e-05,"output_power":8.400182986596024e-05,"reference_active":true,"delta_db":-0.7111238767300243},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2790160745942646e-05,"output_power":1.898460354876454e-05,"reference_active":true,"delta_db":-0.7934585610653865},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.304247232293747e-06,"output_power":4.5585012105699486e-06,"reference_active":true,"delta_db":0.24924865588450504}],"active_band_count":3,"active_band_mae_db":0.584610364559972}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":0.961950694562972,"rms_envelope_nrmse":0.10707321650601939,"envelope_correlation":0.9955131114026416,"rms_gain_error_db":-0.7147091570877824,"reference_mean_power":0.00029237164364699544,"output_mean_power":0.0002480072253280392},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2489747353423385e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.236418785147835e-05,"output_power":7.970274058537418e-05,"reference_active":true,"delta_db":-0.6403036097693465},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2735133445304275e-05,"output_power":1.0277724810237269e-05,"reference_active":true,"delta_db":-0.9310651497078026},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.686566631139435e-07,"output_power":2.8088105448746775e-07,"reference_active":true,"delta_db":0.1932483066014894}],"active_band_count":3,"active_band_mae_db":0.5882056886928796}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":1.5095042930088527,"rms_envelope_nrmse":0.16534199210436,"envelope_correlation":0.9854796474283894,"rms_gain_error_db":-0.7734678793289478,"reference_mean_power":6.736853190004855e-05,"output_mean_power":5.637808179142287e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6302348178615176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0381998821210498e-05,"output_power":1.6583524208270145e-05,"reference_active":true,"delta_db":-0.8956994305335351},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.955680541944289e-06,"output_power":2.3003721769912262e-06,"reference_active":true,"delta_db":-2.354231048516044},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.59401259091449e-07,"output_power":4.116175718908477e-07,"reference_active":true,"delta_db":-2.0465586561922957}],"active_band_count":3,"active_band_mae_db":1.7654963784139583}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":1.1269610948492963,"rms_envelope_nrmse":0.11713348846042106,"envelope_correlation":0.9951692016716495,"rms_gain_error_db":-0.7744327750919343,"reference_mean_power":6.76432550540454e-05,"output_mean_power":5.6595411314408466e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4309026400574305e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.987521517112667e-05,"output_power":1.584886476344341e-05,"reference_active":true,"delta_db":-0.9831367943365527},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.820398831999747e-06,"output_power":1.0476610570175217e-06,"reference_active":true,"delta_db":-2.399457472874949},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.685144573008776e-08,"output_power":2.0321883955315903e-08,"reference_active":true,"delta_db":-2.584905633958754}],"active_band_count":3,"active_band_mae_db":1.9891666337234184}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_7729-102255-0042","method":"magphase","status":"measured","mos":{"status":"measured","value":3.59747052192688,"predicted_mos":3.59747052192688,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289089,"input_sample_rate":48000,"input_seconds":6.0226875,"input_float32_sha256":"2a45311e9cfd473f3a6b429ec935478ab43bc2270b28212a5b0aea71948da47e","internal_samples":96363,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.537221237000267},"started_at":"2026-10-13T05:18:16.374293+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008988445999420946,"native_input_samples":288000,"native_input_sha256":"2db222a55e4362675630a6f9513e5a39d859d07be008cdaee9b77a6862e3c9bd","output_sha256":"f1d655c4d660eeb1f0b120c30f39973555f3812dc6239ca15fc3a8889d03a35a","output_samples":289089,"length_delta_samples":1089,"analysis_s":2.6435199810002814,"synthesis_s":0.8921909719992982,"total_s":3.5357109529995796,"rtf":0.5892851588332633,"output_peak":0.7194503340412233,"output_rms":0.06750512167711775,"finished_at":"2026-10-13T05:18:19.932622+00:00","evaluation_input":{"reference":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"output_sha256":"f1d655c4d660eeb1f0b120c30f39973555f3812dc6239ca15fc3a8889d03a35a","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96363,"length_delta_samples":363,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.410516454572438,"output_peak":0.6972392614859076},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9962329939914698,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.301650466792994,"spectral_convergence":0.7480678270681574,"spectral_convergence_status":"ok","amplitude_floor":2.172633843844252e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.939390649009293,"spectral_convergence":0.4617720033971023,"spectral_convergence_status":"ok","amplitude_floor":1.076658885105748e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.457759014068272,"spectral_convergence":0.4167742543267685,"spectral_convergence_status":"ok","amplitude_floor":9.276051433709775e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0006687421878497317,"output_power":0.0006084868928599046,"delta_db":-0.41007493966667674,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0031496712549154435,"output_power":0.0028768689002075065,"delta_db":-0.3934515551458815,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0007488394009410867,"output_power":0.000681894181470537,"delta_db":-0.4067170250134914,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002473939509459554,"output_power":0.00022883279866229466,"delta_db":-0.3387080432629044,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011342358064728732,"output_power":0.00011577005264476546,"delta_db":0.0889287745458346,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.620279276522665e-05,"output_power":3.915274569738869e-05,"delta_db":-0.7190600403643954,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3928233963331973,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3862226419212083,"pitch":null,"output_rms_dbfs":-23.4189263281076,"output_peak":0.6972392614859076,"reference_rms_dbfs":-23.03270368618639,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.23293337662352},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7205097849363524,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0705285273351481,"rms_dbfs":-23.03270368618639,"peak":0.69677734375,"peak_dbfs":-3.138119583782924,"crest_factor_db":19.894584102403467,"mean":-1.7856280008951823e-05,"zero_energy":false},"output":{"rms":0.06746114119656142,"rms_dbfs":-23.4189263281076,"peak":0.6972392614859076,"peak_dbfs":-3.13236331573313,"crest_factor_db":20.28656301237447,"mean":-8.535566255721825e-05,"zero_energy":false},"rms_difference_db":-0.3862226419212078},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.362966535612941e-06,"threshold_dbfs":-52.70594912983601,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":149,"quiet_samples":23840,"reference_mean_power":4.1423067794300167e-07,"output_mean_power":1.0756102172947783e-06,"output_rms_dbfs":-59.68345081036253,"output_zero_energy":false,"output_peak":0.015788335324207654,"output_energy_fraction":5.869249475290713e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":93,"both_estimated_voiced_frames":66,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.7333333333333333,"absolute_difference_cents_median":9.310173540220767,"absolute_difference_cents_p95":38.85504744512912,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023158079660483383,"log_rms_mae_db":3.061804972799712,"rms_envelope_nrmse":0.3079109405770997,"envelope_correlation":0.9078874648522587,"rms_gain_error_db":-0.3832277957614272,"reference_mean_power":0.004973550834679421,"output_mean_power":0.004553483651443446},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.9210188012408756e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006213961937331268,"output_power":0.0005923774697056385,"reference_active":true,"delta_db":-0.20770056376520563},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00026574930428793645,"output_power":0.0002342594520558438,"reference_active":true,"delta_db":-0.5477501313059371},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.273924181429359e-05,"output_power":3.204487591260193e-05,"reference_active":true,"delta_db":-0.09310023459064945}],"active_band_count":3,"active_band_mae_db":0.28285030988726406}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000211279420280134,"log_rms_mae_db":2.2772662490930617,"rms_envelope_nrmse":0.22879642827652574,"envelope_correlation":0.9461177454013501,"rms_gain_error_db":-0.37925026291183506,"reference_mean_power":0.0049805093532051875,"output_mean_power":0.004564032556849231},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.7395423077013874e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005644300449164409,"output_power":0.0005368708155485223,"reference_active":true,"delta_db":-0.21740327195356793},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013986675109572898,"output_power":0.00012477653093466055,"reference_active":true,"delta_db":-0.4958157985037899},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4974533271774746e-06,"output_power":2.270970329435583e-06,"reference_active":true,"delta_db":-0.41285920671520476}],"active_band_count":3,"active_band_mae_db":0.3753594257241875}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020019856693346264,"log_rms_mae_db":3.3491271620998346,"rms_envelope_nrmse":0.3491035209368456,"envelope_correlation":0.89822926245327,"rms_gain_error_db":-0.46340833999930997,"reference_mean_power":0.0028106313477950454,"output_mean_power":0.0025261727831630124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3651110338311905e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00046335294017358356,"output_power":0.0004249125564592766,"reference_active":true,"delta_db":-0.37612358170261423},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00024122957535154208,"output_power":0.00021242645004949335,"reference_active":true,"delta_db":-0.5522196084254939},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9488783219041956e-05,"output_power":2.7494809963294722e-05,"reference_active":true,"delta_db":-0.30406130542811993}],"active_band_count":3,"active_band_mae_db":0.41080149851874265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018155803316515472,"log_rms_mae_db":2.6042871018596765,"rms_envelope_nrmse":0.2584853961722457,"envelope_correlation":0.9411160407833163,"rms_gain_error_db":-0.46445766732739335,"reference_mean_power":0.002819303754025505,"output_mean_power":0.002533355297415769},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2479023732657914e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004464610508434936,"output_power":0.0004083117520746154,"reference_active":true,"delta_db":-0.38791697113874035},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012910816269432612,"output_power":0.00011403496323534617,"reference_active":true,"delta_db":-0.5391567395183529},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.446292305046822e-06,"output_power":2.1616480138649953e-06,"reference_active":true,"delta_db":-0.5372337101736806}],"active_band_count":3,"active_band_mae_db":0.48810247361025794}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.07685175719321e-05,"log_rms_mae_db":3.5343975896061797,"rms_envelope_nrmse":0.34490079230926274,"envelope_correlation":0.9102701521042807,"rms_gain_error_db":-0.30341678174826575,"reference_mean_power":0.0002950078039052426,"output_mean_power":0.0002751007829669904},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.385626430084388e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.894665895613549e-05,"output_power":9.65990481011044e-05,"reference_active":true,"delta_db":-0.10428287508965378},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2790160745942646e-05,"output_power":2.2483521682212286e-05,"reference_active":true,"delta_db":-0.05883050928679107},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.304247232293747e-06,"output_power":4.017302382303028e-06,"reference_active":true,"delta_db":-0.29962686577263353}],"active_band_count":3,"active_band_mae_db":0.1542467500496928}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.03385602019198e-05,"log_rms_mae_db":2.5723586208466775,"rms_envelope_nrmse":0.25364416151370206,"envelope_correlation":0.9487585154428829,"rms_gain_error_db":-0.2608637234268094,"reference_mean_power":0.00029237164364699544,"output_mean_power":0.0002753270470720983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2489747353423385e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.236418785147835e-05,"output_power":9.009749715030735e-05,"reference_active":true,"delta_db":-0.10790889152479682},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2735133445304275e-05,"output_power":1.2350223696697225e-05,"reference_active":true,"delta_db":-0.1332867614457819},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.686566631139435e-07,"output_power":2.4284568354940615e-07,"reference_active":true,"delta_db":-0.4386722770615809}],"active_band_count":3,"active_band_mae_db":0.22662264334405322}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":4.524569706134418e-05,"log_rms_mae_db":3.793942107101849,"rms_envelope_nrmse":0.5207064491094707,"envelope_correlation":0.8262250404945541,"rms_gain_error_db":0.24557227162698553,"reference_mean_power":6.736853190004855e-05,"output_mean_power":7.128765197403786e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.6302348178615176e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0381998821210498e-05,"output_power":2.0230179060931865e-05,"reference_active":true,"delta_db":-0.03247045330398814},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.955680541944289e-06,"output_power":4.138401931977923e-06,"reference_active":true,"delta_db":0.19611457131854995},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.59401259091449e-07,"output_power":6.679713018508691e-07,"reference_active":true,"delta_db":0.056080321487314164}],"active_band_count":3,"active_band_mae_db":0.0948884487032841}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.8717196486707e-05,"log_rms_mae_db":2.9077793424711893,"rms_envelope_nrmse":0.4005073952002322,"envelope_correlation":0.8919771953719708,"rms_gain_error_db":0.24954976639139162,"reference_mean_power":6.76432550540454e-05,"output_mean_power":7.164394217805438e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4309026400574305e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.987521517112667e-05,"output_power":1.9741207386797066e-05,"reference_active":true,"delta_db":-0.029381281541574066},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.820398831999747e-06,"output_power":1.878443644721556e-06,"reference_active":true,"delta_db":0.13631622121749534},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.685144573008776e-08,"output_power":4.285790662945643e-08,"reference_active":true,"delta_db":0.6557642388615205}],"active_band_count":3,"active_band_mae_db":0.2738205805401966}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"hnm","status":"measured","mos":{"status":"measured","value":2.9525396823883057,"predicted_mos":2.9525396823883057,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1ae12bcfbc645f8ed1c399f46f538a19faf30e8e4d9640c9a4542e037a9f43ef","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7343765099994926},"started_at":"2026-10-13T04:40:15.647191+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.460799921915168e-05,"native_input_samples":96000,"native_input_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","output_sha256":"746aff98c6e2b320125dd5bb9adef2755fdbb9502798d4fa1b70eb0154a41f34","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0865216509992024,"synthesis_s":0.49386958600007347,"total_s":1.5803912369992759,"rtf":0.2633985394998793,"output_peak":0.42881423954006015,"output_rms":0.07077392087327442,"finished_at":"2026-10-13T04:40:17.234847+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"746aff98c6e2b320125dd5bb9adef2755fdbb9502798d4fa1b70eb0154a41f34","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.992023528168302,"spectral_convergence":0.21671429975007575,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.604492240127555,"spectral_convergence":0.2076792737518895,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.729073289628451,"spectral_convergence":0.20310532390369915,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0012590117574724593,"delta_db":0.015040440346391936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.002923751340582295,"delta_db":-0.0888651316871662,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.0004936144578314487,"delta_db":0.0515911314361038,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00012644309851649347,"delta_db":-0.2657442105603295,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":0.00010173280263892152,"delta_db":0.02091864876056894,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":0.00010439441873488812,"delta_db":-0.06831594355189922,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.08507925105707659,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.051203434757911226,"pitch":null,"output_rms_dbfs":-23.002534876517554,"output_peak":0.42881423954006015,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7751963526414345},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.992023528168302,"spectral_convergence":0.21671429975007575,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.604492240127555,"spectral_convergence":0.2076792737518895,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.729073289628451,"spectral_convergence":0.20310532390369915,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0012590117574724593,"delta_db":0.015040440346391936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.002923751340582295,"delta_db":-0.0888651316871662,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.0004936144578314487,"delta_db":0.0515911314361038,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00012644309851649347,"delta_db":-0.2657442105603295,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":0.00010173280263892152,"delta_db":0.02091864876056894,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":0.00010439441873488812,"delta_db":-0.06831594355189922,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.08507925105707659,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.051203434757911226,"pitch":null,"output_rms_dbfs":-23.002534876517554,"output_peak":0.42881423954006015,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7751963526414345},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8488053622642499,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07119236623533667,"rms_dbfs":-22.951331441759642,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.982768658370414,"mean":-4.647572835286458e-06,"zero_energy":false},"output":{"rms":0.07077392087327442,"rms_dbfs":-23.002534876517554,"peak":0.42881423954006015,"peak_dbfs":-7.35461603109146,"crest_factor_db":15.647918845426094,"mean":9.168448313016187e-05,"zero_energy":false},"rms_difference_db":-0.05120343475791245},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":4.535169089148928e-06,"output_rms_dbfs":-53.43406516075695,"output_zero_energy":false,"output_peak":0.008448734390314275,"output_energy_fraction":4.979774317665918e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":170,"both_estimated_voiced_frames":154,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9746835443037974,"absolute_difference_cents_median":1.5784384682177066,"absolute_difference_cents_p95":5.191179397649113,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024147266762860616,"log_rms_mae_db":1.2305636811357954,"rms_envelope_nrmse":0.07286801385188159,"envelope_correlation":0.9949853277655383,"rms_gain_error_db":-0.05120229214449586,"reference_mean_power":0.005072567967533502,"output_mean_power":0.005013114749398806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.4164286663529534e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010865583816620668,"output_power":0.001087603090611422,"reference_active":true,"delta_db":0.00417366761949565},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020269542962956985,"output_power":0.00018642564940778353,"reference_active":true,"delta_db":-0.3633829172828248},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.69603101954121e-05,"output_power":1.9540642781553013e-05,"reference_active":true,"delta_db":-1.3978603916526593}],"active_band_count":3,"active_band_mae_db":0.5884723255183265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000225157536036158,"log_rms_mae_db":0.6879817023397671,"rms_envelope_nrmse":0.038790073707767214,"envelope_correlation":0.9985036512612364,"rms_gain_error_db":-0.051192558065739274,"reference_mean_power":0.005098028148503288,"output_mean_power":0.005038287815963855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2603245746946088e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001023229039263937,"output_power":0.0010201925774826468,"reference_active":true,"delta_db":-0.012906974810692737},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010544090587602268,"output_power":9.731104395510742e-05,"reference_active":true,"delta_db":-0.34846996723713974},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0049541105469597e-06,"output_power":1.2894688631566328e-06,"reference_active":true,"delta_db":-1.9169357724545877}],"active_band_count":3,"active_band_mae_db":0.7594375715008067}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020992712290116374,"log_rms_mae_db":1.7337331343757885,"rms_envelope_nrmse":0.1043504743116917,"envelope_correlation":0.9918208536016551,"rms_gain_error_db":-0.0674038419873647,"reference_mean_power":0.0023388664575994473,"output_mean_power":0.0023028467707448953},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4073975659228857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007410848450397675,"output_power":0.0007460433508134493,"reference_active":true,"delta_db":0.028961318632033977},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011531333502346333,"output_power":0.00010356778079929803,"reference_active":true,"delta_db":-0.4665486223336399},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.654328508679614e-05,"output_power":1.738470698984793e-05,"reference_active":true,"delta_db":-1.8378729633532807}],"active_band_count":3,"active_band_mae_db":0.7777943014396516}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002036893389343094,"log_rms_mae_db":1.0544738099413729,"rms_envelope_nrmse":0.060174398165052656,"envelope_correlation":0.9971563186639885,"rms_gain_error_db":-0.06740525290007704,"reference_mean_power":0.0023506239289763434,"output_mean_power":0.002314422419417369},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3312664615669961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215176046332657,"output_power":0.0007207191868169576,"reference_active":true,"delta_db":-0.004808482065553498},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.516212457663081e-05,"output_power":4.90505325833929e-05,"reference_active":true,"delta_db":-0.5099725748968539},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1332065410684056e-06,"output_power":1.0931316058524228e-06,"reference_active":true,"delta_db":-2.903604553982201}],"active_band_count":3,"active_band_mae_db":1.1394618703148696}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":1.8233878289084828,"rms_envelope_nrmse":0.15206396731327326,"envelope_correlation":0.9818149074772345,"rms_gain_error_db":-0.17669062819097217,"reference_mean_power":0.00021265544148097977,"output_mean_power":0.00020417729093394785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1812635588892241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.569393841260955e-05,"output_power":5.4873844075428e-05,"reference_active":true,"delta_db":-0.0644254535532626},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4402699100691383e-05,"output_power":1.2698427333972188e-05,"reference_active":true,"delta_db":-0.5469394947569508},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9279580393512075e-06,"output_power":2.329286015095421e-06,"reference_active":true,"delta_db":-0.9934202933097218}],"active_band_count":3,"active_band_mae_db":0.5349284138733118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":1.128260354058603,"rms_envelope_nrmse":0.08260887595185717,"envelope_correlation":0.9945191662205798,"rms_gain_error_db":-0.17669560707121906,"reference_mean_power":0.00021371687213348283,"output_mean_power":0.0002051961692091107},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0829999638899275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5369039282181205e-05,"output_power":5.418002235379759e-05,"reference_active":true,"delta_db":-0.09427808178271738},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.495407348960159e-06,"output_power":5.77012908350019e-06,"reference_active":true,"delta_db":-0.5142086347678427},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1097616536448917e-07,"output_power":1.4117949444525244e-07,"reference_active":true,"delta_db":-1.7446177208491542}],"active_band_count":3,"active_band_mae_db":0.7843681457999048}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":2.139092520871333,"rms_envelope_nrmse":0.15636746738092855,"envelope_correlation":0.9846902498550603,"rms_gain_error_db":-0.030817728942803876,"reference_mean_power":6.940550281225069e-05,"output_mean_power":6.891474157884226e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.994304097313728e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6831637620284676e-05,"output_power":3.4823296452906353e-05,"reference_active":true,"delta_db":-0.24351148990646387},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.349662193280123e-06,"output_power":4.891695876913683e-06,"reference_active":true,"delta_db":-0.38866910623683715},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1665818120524943e-06,"output_power":1.3214415138467427e-06,"reference_active":true,"delta_db":0.5413274515110086}],"active_band_count":3,"active_band_mae_db":0.3911693492181032}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":1.6066510023137845,"rms_envelope_nrmse":0.10275668089445651,"envelope_correlation":0.9931591802624897,"rms_gain_error_db":-0.030824217865396342,"reference_mean_power":6.975388624239838e-05,"output_mean_power":6.926055813090914e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.725701790778612e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6073966740876396e-05,"output_power":3.3905750674723006e-05,"reference_active":true,"delta_db":-0.26920536419922114},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3003381054168375e-06,"output_power":2.182638167489815e-06,"reference_active":true,"delta_db":-0.22809928256600667},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.80849015690718e-08,"output_power":8.45341008275977e-08,"reference_active":true,"delta_db":0.3446487033565146}],"active_band_count":3,"active_band_mae_db":0.28065111670724746}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9537789821624756,"predicted_mos":2.9537789821624756,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c7b57a60a65e8e3e57c0073420eb00332d2bd8b773618f68b5fb2f369fd97863","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2298051660000056},"started_at":"2026-10-13T04:43:30.776504+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2251000700634904e-05,"native_input_samples":96000,"native_input_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","output_sha256":"f8eff0c09ab8fa3768f71e940eb48c55fb6eff676b37333471b821b6132d6bc2","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6125142159999086,"synthesis_s":0.09162578300038149,"total_s":1.7041399990002901,"rtf":0.284023333166715,"output_peak":0.6652808282001412,"output_rms":0.07901642216773214,"finished_at":"2026-10-13T04:43:32.486633+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"f8eff0c09ab8fa3768f71e940eb48c55fb6eff676b37333471b821b6132d6bc2","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.795219317245865,"spectral_convergence":0.6647203384168616,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.405950839786843,"spectral_convergence":0.2916263607773636,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.68748680854901,"spectral_convergence":0.30656088231097706,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0012112104807195738,"delta_db":-0.1530612165337619,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.003981049094355811,"delta_db":1.2517058622676538,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.000659000646810477,"delta_db":1.3065708242157095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00017052070167406498,"delta_db":1.0330756021545806,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":0.00012003651336545842,"delta_db":0.7394422804974699,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":0.00010177753526388462,"delta_db":-0.17856945447006595,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7770708733565402,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9056786639329235,"pitch":null,"output_rms_dbfs":-22.04565277782672,"output_peak":0.6652808282001412,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.962885655193905},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.795219317245865,"spectral_convergence":0.6647203384168616,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.405950839786843,"spectral_convergence":0.2916263607773636,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.68748680854901,"spectral_convergence":0.30656088231097706,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0012112104807195738,"delta_db":-0.1530612165337619,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.003981049094355811,"delta_db":1.2517058622676538,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.000659000646810477,"delta_db":1.3065708242157095,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00017052070167406498,"delta_db":1.0330756021545806,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":0.00012003651336545842,"delta_db":0.7394422804974699,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":0.00010177753526388462,"delta_db":-0.17856945447006595,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7770708733565402,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9056786639329235,"pitch":null,"output_rms_dbfs":-22.04565277782672,"output_peak":0.6652808282001412,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.962885655193905},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8835289695767616,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07119236623533667,"rms_dbfs":-22.951331441759642,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.982768658370414,"mean":-4.647572835286458e-06,"zero_energy":false},"output":{"rms":0.07901642216773214,"rms_dbfs":-22.04565277782672,"peak":0.6652808282001412,"peak_dbfs":-3.5398998340057046,"crest_factor_db":18.505752943821015,"mean":1.1801176778592863e-08,"zero_energy":false},"rms_difference_db":0.9056786639329211},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":2.0694367694781945e-06,"output_rms_dbfs":-56.841478386936274,"output_zero_energy":false,"output_peak":0.005611614485031693,"output_energy_fraction":1.82297254751922e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":158,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":6.038220450856699,"absolute_difference_cents_p95":16.20013858154098,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024147266762860616,"log_rms_mae_db":2.831916097133332,"rms_envelope_nrmse":0.20105293707074223,"envelope_correlation":0.9776522642885171,"rms_gain_error_db":0.9056878798826475,"reference_mean_power":0.005072567967533502,"output_mean_power":0.006248800547672782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.4164286663529534e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010865583816620668,"output_power":0.001441207277361176,"reference_active":true,"delta_db":1.2267338014373843},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020269542962956985,"output_power":0.0002529485333269464,"reference_active":true,"delta_db":0.9618820923495625},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.69603101954121e-05,"output_power":2.985301403373752e-05,"reference_active":true,"delta_db":0.44263300393712485}],"active_band_count":3,"active_band_mae_db":0.8770829659080238}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000225157536036158,"log_rms_mae_db":2.3235164652305462,"rms_envelope_nrmse":0.15532638044516933,"envelope_correlation":0.9904694000429306,"rms_gain_error_db":0.905718022541775,"reference_mean_power":0.005098028148503288,"output_mean_power":0.006280208051395236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2603245746946088e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001023229039263937,"output_power":0.001368720089079356,"reference_active":true,"delta_db":1.263417846285701},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010544090587602268,"output_power":0.00012792479686757224,"reference_active":true,"delta_db":0.8394560759970898},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0049541105469597e-06,"output_power":2.332794890552289e-06,"reference_active":true,"delta_db":0.6577211849316861}],"active_band_count":3,"active_band_mae_db":0.9201983690714922}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020992712290116374,"log_rms_mae_db":2.204859512772215,"rms_envelope_nrmse":0.2633305990012236,"envelope_correlation":0.9803890625079894,"rms_gain_error_db":1.5340662875048094,"reference_mean_power":0.0023388664575994473,"output_mean_power":0.003329753271812124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4073975659228857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007410848450397675,"output_power":0.0010773550214508132,"reference_active":true,"delta_db":1.624909080389707},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011531333502346333,"output_power":0.00017740867584028556,"reference_active":true,"delta_db":1.8709532165033005},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.654328508679614e-05,"output_power":2.0744992593356236e-05,"reference_active":true,"delta_db":-1.0704138750211083}],"active_band_count":3,"active_band_mae_db":1.5220920573047054}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002036893389343094,"log_rms_mae_db":1.6505714963663978,"rms_envelope_nrmse":0.22944855816463033,"envelope_correlation":0.9902903972819862,"rms_gain_error_db":1.5340645972780151,"reference_mean_power":0.0023506239289763434,"output_mean_power":0.0033464906244201112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3312664615669961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215176046332657,"output_power":0.0010479159996920474,"reference_active":true,"delta_db":1.6207953915628548},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.516212457663081e-05,"output_power":8.096439859065865e-05,"reference_active":true,"delta_db":1.666531095194105},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1332065410684056e-06,"output_power":1.5929252649080415e-06,"reference_active":true,"delta_db":-1.268375061588477}],"active_band_count":3,"active_band_mae_db":1.5185671827818121}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":1.6906153417232548,"rms_envelope_nrmse":0.20926854187008373,"envelope_correlation":0.9784319744154762,"rms_gain_error_db":0.9377355814229951,"reference_mean_power":0.00021265544148097977,"output_mean_power":0.0002639064824835911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1812635588892241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.569393841260955e-05,"output_power":7.107408240254779e-05,"reference_active":true,"delta_db":1.0590333123436355},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4402699100691383e-05,"output_power":1.8018271491393602e-05,"reference_active":true,"delta_db":0.9726923883024413},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9279580393512075e-06,"output_power":3.8023173818588116e-06,"reference_active":true,"delta_db":1.1348351633550644}],"active_band_count":3,"active_band_mae_db":1.0555202880003804}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":1.1656805787088458,"rms_envelope_nrmse":0.14446226471235638,"envelope_correlation":0.9941183603322894,"rms_gain_error_db":0.937820127944182,"reference_mean_power":0.00021371687213348283,"output_mean_power":0.0002652288866199285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0829999638899275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5369039282181205e-05,"output_power":7.091490461931786e-05,"reference_active":true,"delta_db":1.0747053504637227},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.495407348960159e-06,"output_power":8.099326291946843e-06,"reference_active":true,"delta_db":0.9584250310062664},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1097616536448917e-07,"output_power":2.9501444878044857e-07,"reference_active":true,"delta_db":1.4560989226396426}],"active_band_count":3,"active_band_mae_db":1.1630764347032105}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":1.7469228945304776,"rms_envelope_nrmse":0.21218847636378588,"envelope_correlation":0.9737186564441783,"rms_gain_error_db":0.3192064759447552,"reference_mean_power":6.940550281225069e-05,"output_mean_power":7.469895949474906e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.994304097313728e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6831637620284676e-05,"output_power":3.840432794011842e-05,"reference_active":true,"delta_db":0.1815914051223658},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.349662193280123e-06,"output_power":5.230058328762463e-06,"reference_active":true,"delta_db":-0.09819826767119831},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1665818120524943e-06,"output_power":1.1945070585529776e-06,"reference_active":true,"delta_db":0.10273519269243953}],"active_band_count":3,"active_band_mae_db":0.12750828849533455}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":1.0844278967413672,"rms_envelope_nrmse":0.1141999870992478,"envelope_correlation":0.9926929170917774,"rms_gain_error_db":0.3192168082810666,"reference_mean_power":6.975388624239838e-05,"output_mean_power":7.507409223124339e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.725701790778612e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6073966740876396e-05,"output_power":3.7784584934546406e-05,"reference_active":true,"delta_db":0.20120755297366857},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3003381054168375e-06,"output_power":2.2189618916358314e-06,"reference_active":true,"delta_db":-0.15641829897515025},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.80849015690718e-08,"output_power":8.592029304079807e-08,"reference_active":true,"delta_db":0.41528682389099825}],"active_band_count":3,"active_band_mae_db":0.25763755861327237}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"facodec","status":"measured","mos":{"status":"measured","value":4.2057414054870605,"predicted_mos":4.2057414054870605,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"41c9161b4834f512215f961bfe8966361c7320bc89869f0aa38640a998495467","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7705139569989115},"started_at":"2026-10-13T07:00:17.786047+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6230000003124587e-05,"native_input_samples":96000,"native_input_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","output_sha256":"7d3040c4603cff14d6c80b3fc4a2e39a3ff42d37249cf59b7766dce9e237aeb9","output_samples":96000,"length_delta_samples":0,"analysis_s":2.282633115999488,"synthesis_s":6.321228349999728,"total_s":8.603861465999216,"rtf":1.4339769109998695,"output_peak":0.37501826882362366,"output_rms":0.06549734745327432,"finished_at":"2026-10-13T07:00:26.400579+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"7d3040c4603cff14d6c80b3fc4a2e39a3ff42d37249cf59b7766dce9e237aeb9","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.539368622736568,"spectral_convergence":0.24315899401631835,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.8742865002316,"spectral_convergence":0.2431835728621045,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.956688303504563,"spectral_convergence":0.24947275177708997,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0011222261412981807,"delta_db":-0.48445360813326677,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.002467591474278504,"delta_db":-0.8255368606313627,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.00045881051985497594,"delta_db":-0.2659539135213711,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00011475953654623285,"delta_db":-0.6868076514852077,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":5.984221570856756e-05,"delta_db":-2.283614788994548,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":6.667263572847986e-05,"delta_db":-2.015612506840408,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0936632216010274,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7241943171016497,"pitch":null,"output_rms_dbfs":-23.675525758861294,"output_peak":0.37501826882362366,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.79011447549091},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.539368622736568,"spectral_convergence":0.24315899401631835,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.8742865002316,"spectral_convergence":0.2431835728621045,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.956688303504563,"spectral_convergence":0.24947275177708997,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0011222261412981807,"delta_db":-0.48445360813326677,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.002467591474278504,"delta_db":-0.8255368606313627,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.00045881051985497594,"delta_db":-0.2659539135213711,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00011475953654623285,"delta_db":-0.6868076514852077,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":5.984221570856756e-05,"delta_db":-2.283614788994548,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":6.667263572847986e-05,"delta_db":-2.015612506840408,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0936632216010274,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7241943171016497,"pitch":null,"output_rms_dbfs":-23.675525758861294,"output_peak":0.37501826882362366,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.79011447549091},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8463617062813369,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07119236623533667,"rms_dbfs":-22.951331441759642,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.982768658370414,"mean":-4.647572835286458e-06,"zero_energy":false},"output":{"rms":0.06549734745327432,"rms_dbfs":-23.67552575886129,"peak":0.37501826882362366,"peak_dbfs":-8.518951506457043,"crest_factor_db":15.15657425240425,"mean":-5.204683721948792e-05,"zero_energy":false},"rms_difference_db":-0.7241943171016487},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":4.127601047310582e-06,"output_rms_dbfs":-53.84302286023441,"output_zero_energy":false,"output_peak":0.0080906692892313,"output_energy_fraction":5.291916456445873e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":163,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9683544303797469,"absolute_difference_cents_median":2.6924973900165985,"absolute_difference_cents_p95":8.943925779629456,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024147266762860616,"log_rms_mae_db":1.3746951034184458,"rms_envelope_nrmse":0.13549734809636285,"envelope_correlation":0.9876994499365742,"rms_gain_error_db":-0.7241910432657643,"reference_mean_power":0.005072567967533502,"output_mean_power":0.004293473340218559},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.4164286663529534e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010865583816620668,"output_power":0.0010028328408769114,"reference_active":true,"delta_db":-0.34824518436694885},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020269542962956985,"output_power":0.00018211699845909073,"reference_active":true,"delta_db":-0.46493472425252674},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.69603101954121e-05,"output_power":3.012100013539134e-05,"reference_active":true,"delta_db":0.48144503324341825}],"active_band_count":3,"active_band_mae_db":0.4315416472876312}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000225157536036158,"log_rms_mae_db":1.0575510454194927,"rms_envelope_nrmse":0.12559908208695408,"envelope_correlation":0.9896638666772803,"rms_gain_error_db":-0.7241916875165577,"reference_mean_power":0.005098028148503288,"output_mean_power":0.004315022457339424},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2603245746946088e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001023229039263937,"output_power":0.0009542742261263596,"reference_active":true,"delta_db":-0.30299662753103296},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010544090587602268,"output_power":9.594390655888009e-05,"reference_active":true,"delta_db":-0.40991730719237834},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0049541105469597e-06,"output_power":2.1691952121309173e-06,"reference_active":true,"delta_db":0.3419420026127957}],"active_band_count":3,"active_band_mae_db":0.35161864577873564}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020992712290116374,"log_rms_mae_db":1.9980599873505684,"rms_envelope_nrmse":0.14257659209077184,"envelope_correlation":0.9874834617467044,"rms_gain_error_db":-0.6193004704475364,"reference_mean_power":0.0023388664575994473,"output_mean_power":0.0020280346858062917},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4073975659228857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007410848450397675,"output_power":0.000697254722958522,"reference_active":true,"delta_db":-0.26476467460898734},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011531333502346333,"output_power":0.00010379636928638906,"reference_active":true,"delta_db":-0.4569737021999964},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.654328508679614e-05,"output_power":3.45372761708531e-05,"reference_active":true,"delta_db":1.1433341174868341}],"active_band_count":3,"active_band_mae_db":0.6216908314319393}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002036893389343094,"log_rms_mae_db":1.4482426745603696,"rms_envelope_nrmse":0.12265449240359982,"envelope_correlation":0.9912240482310799,"rms_gain_error_db":-0.6193055689182094,"reference_mean_power":0.0023506239289763434,"output_mean_power":0.002038227214424585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3312664615669961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215176046332657,"output_power":0.0006856104366041746,"reference_active":true,"delta_db":-0.2216951215619108},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.516212457663081e-05,"output_power":4.952532815969158e-05,"reference_active":true,"delta_db":-0.4681362283815591},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1332065410684056e-06,"output_power":2.52853246091131e-06,"reference_active":true,"delta_db":0.738356267166458}],"active_band_count":3,"active_band_mae_db":0.47606253903664264}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":1.9667891140690823,"rms_envelope_nrmse":0.23499052693356676,"envelope_correlation":0.9694780401754319,"rms_gain_error_db":-1.4018925659276613,"reference_mean_power":0.00021265544148097977,"output_mean_power":0.0001539881294347858},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1812635588892241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.569393841260955e-05,"output_power":4.009223142378894e-05,"reference_active":true,"delta_db":-1.4274770166828297},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4402699100691383e-05,"output_power":1.0183453284216391e-05,"reference_active":true,"delta_db":-1.5054881211682893},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9279580393512075e-06,"output_power":2.520311449448524e-06,"reference_active":true,"delta_db":-0.6511063616813821}],"active_band_count":3,"active_band_mae_db":1.1946904998441672}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":1.566140194034798,"rms_envelope_nrmse":0.21397366409224103,"envelope_correlation":0.9766833457117262,"rms_gain_error_db":-1.4019315420099,"reference_mean_power":0.00021371687213348283,"output_mean_power":0.00015475534408439476},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0829999638899275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5369039282181205e-05,"output_power":3.97423412097934e-05,"reference_active":true,"delta_db":-1.4401354036894496},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.495407348960159e-06,"output_power":4.401354221887693e-06,"reference_active":true,"delta_db":-1.690200702605173},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1097616536448917e-07,"output_power":1.7901561361423896e-07,"reference_active":true,"delta_db":-0.7134248297950654}],"active_band_count":3,"active_band_mae_db":1.2812536453632293}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":1.8471333317811014,"rms_envelope_nrmse":0.21033962778577403,"envelope_correlation":0.978933886264851,"rms_gain_error_db":-1.1176110382394524,"reference_mean_power":6.940550281225069e-05,"output_mean_power":5.3657792417909323e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.994304097313728e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6831637620284676e-05,"output_power":3.009137579965913e-05,"reference_active":true,"delta_db":-0.877789845724737},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.349662193280123e-06,"output_power":4.964653803408741e-06,"reference_active":true,"delta_db":-0.32437389629053537},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1665818120524943e-06,"output_power":1.1472731571530662e-06,"reference_active":true,"delta_db":-0.07248368667998625}],"active_band_count":3,"active_band_mae_db":0.4248824762317529}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":1.515354402504182,"rms_envelope_nrmse":0.1730420030178052,"envelope_correlation":0.9887613843398979,"rms_gain_error_db":-1.1176429564764165,"reference_mean_power":6.975388624239838e-05,"output_mean_power":5.392673331345179e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.725701790778612e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6073966740876396e-05,"output_power":2.9795260071892545e-05,"reference_active":true,"delta_db":-0.8304671996266572},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3003381054168375e-06,"output_power":2.2784828534877413e-06,"reference_active":true,"delta_db":-0.04145908949479009},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.80849015690718e-08,"output_power":7.638889291652359e-08,"reference_active":true,"delta_db":-0.095368512176318}],"active_band_count":3,"active_band_mae_db":0.3224316004325884}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.3946008682250977,"predicted_mos":3.3946008682250977,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"528b873fe2d1676aec041e588b1c024e42fbbca4a32877dbeb115a2c65638251","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7653507819995866},"started_at":"2026-10-13T07:30:22.952487+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00689207899995381,"native_input_samples":264600,"native_input_sha256":"88337eae28bf5b9db1148de9738c3c86b12b589ce30617d34373e7511960119b","output_sha256":"555c2a21ea7b72b96aa13977ade4058f7171c9d09048a8747bb083c71c791089","output_samples":264704,"length_delta_samples":104,"analysis_s":0.15461117399900104,"synthesis_s":3.9025208489983925,"total_s":4.0571320229973935,"rtf":0.6761886704995655,"output_peak":0.5147438049316406,"output_rms":0.06427556774005934,"finished_at":"2026-10-13T07:30:27.037007+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"555c2a21ea7b72b96aa13977ade4058f7171c9d09048a8747bb083c71c791089","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.83583491536607,"output_peak":0.5045665620098037},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.365908610957268,"spectral_convergence":0.45799039923587315,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.928813609665571,"spectral_convergence":0.320222190091923,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.642938506164922,"spectral_convergence":0.37903541794574674,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.001140883599491818,"delta_db":-0.4128440474083194,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.002385729940380647,"delta_db":-0.9720566622304654,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.00037712989790659855,"delta_db":-1.117377948925617,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":8.436829752634527e-05,"delta_db":-2.022902653001495,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":5.9907963505162774e-05,"delta_db":-2.278845876193709,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":8.805355302364605e-05,"delta_db":-0.8076199009792346,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.26860784812314,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8827863410718312,"pitch":null,"output_rms_dbfs":-23.834117782831473,"output_peak":0.5045665620098037,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.979220242262587},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6886807769189675,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07119236623533667,"rms_dbfs":-22.951331441759642,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.982768658370414,"mean":-4.647572835286458e-06,"zero_energy":false},"output":{"rms":0.06431231026665282,"rms_dbfs":-23.834117782831473,"peak":0.5045665620098037,"peak_dbfs":-5.9416306774710295,"crest_factor_db":17.892487105360445,"mean":-0.00020711298173682884,"zero_energy":false},"rms_difference_db":-0.8827863410718315},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":4.908236444691715e-06,"output_rms_dbfs":-53.090745241537775,"output_zero_energy":false,"output_peak":0.009069524647819793,"output_energy_fraction":6.526794571114732e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":166,"both_estimated_voiced_frames":138,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.8734177215189873,"absolute_difference_cents_median":10.15021768900089,"absolute_difference_cents_p95":25.680460288194055,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024147266762860616,"log_rms_mae_db":2.012519725399462,"rms_envelope_nrmse":0.13229446753655277,"envelope_correlation":0.9917386553988173,"rms_gain_error_db":-0.8827788075970576,"reference_mean_power":0.005072567967533502,"output_mean_power":0.004139520084782577},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.4164286663529534e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010865583816620668,"output_power":0.0009130288485113164,"reference_active":true,"delta_db":-0.7556856634335024},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020269542962956985,"output_power":0.00017293884323482713,"reference_active":true,"delta_db":-0.6895140663387607},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.69603101954121e-05,"output_power":1.9438331371778e-05,"reference_active":true,"delta_db":-1.4206590330779603}],"active_band_count":3,"active_band_mae_db":0.9552862542834077}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000225157536036158,"log_rms_mae_db":1.5591612126738554,"rms_envelope_nrmse":0.11327799946065596,"envelope_correlation":0.9962901418600215,"rms_gain_error_db":-0.882752744737919,"reference_mean_power":0.005098028148503288,"output_mean_power":0.004160322088245519},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2603245746946088e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001023229039263937,"output_power":0.0008651541428675433,"reference_active":true,"delta_db":-0.7287936515274147},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010544090587602268,"output_power":8.759087517526583e-05,"reference_active":true,"delta_db":-0.805502627544104},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0049541105469597e-06,"output_power":1.5619991973233e-06,"reference_active":true,"delta_db":-1.0842363055707969}],"active_band_count":3,"active_band_mae_db":0.8728441948807718}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020992712290116374,"log_rms_mae_db":3.0551636909430058,"rms_envelope_nrmse":0.1645093148166607,"envelope_correlation":0.9861726014439256,"rms_gain_error_db":-0.9370615000427779,"reference_mean_power":0.0023388664575994473,"output_mean_power":0.0018849475735057534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4073975659228857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007410848450397675,"output_power":0.000562056382386126,"reference_active":true,"delta_db":-1.2008804839358425},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011531333502346333,"output_power":0.00010339546042288761,"reference_active":true,"delta_db":-0.4737806124635052},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.654328508679614e-05,"output_power":1.577101630428389e-05,"reference_active":true,"delta_db":-2.260949908920732}],"active_band_count":3,"active_band_mae_db":1.3118703351066932}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002036893389343094,"log_rms_mae_db":2.4099952060445387,"rms_envelope_nrmse":0.13483494084648792,"envelope_correlation":0.9931924205162495,"rms_gain_error_db":-0.9370699103175479,"reference_mean_power":0.0023506239289763434,"output_mean_power":0.0018944195280346673},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3312664615669961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215176046332657,"output_power":0.0005524498700373213,"reference_active":true,"delta_db":-1.1595405636188012},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.516212457663081e-05,"output_power":4.7828888429645136e-05,"reference_active":true,"delta_db":-0.6195069699575688},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1332065410684056e-06,"output_power":1.1901098961123127e-06,"reference_active":true,"delta_db":-2.5344584019892125}],"active_band_count":3,"active_band_mae_db":1.437835311855194}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":4.3782613230314285,"rms_envelope_nrmse":0.322214624276619,"envelope_correlation":0.9576863671589954,"rms_gain_error_db":-2.523795013376162,"reference_mean_power":0.00021265544148097977,"output_mean_power":0.00011893152802036174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1812635588892241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.569393841260955e-05,"output_power":3.2091546735139585e-05,"reference_active":true,"delta_db":-2.39417280686476},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4402699100691383e-05,"output_power":7.629492229395025e-06,"reference_active":true,"delta_db":-2.7594825252910176},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9279580393512075e-06,"output_power":1.8962679015055784e-06,"reference_active":true,"delta_db":-1.8866515482180255}],"active_band_count":3,"active_band_mae_db":2.3467689601246007}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":3.991703905413214,"rms_envelope_nrmse":0.29026879695535884,"envelope_correlation":0.9765652791822386,"rms_gain_error_db":-2.5236991347525812,"reference_mean_power":0.00021371687213348283,"output_mean_power":0.00011952779171040556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0829999638899275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5369039282181205e-05,"output_power":3.097581987345693e-05,"reference_active":true,"delta_db":-2.5224417787523135},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.495407348960159e-06,"output_power":3.3117789780122553e-06,"reference_active":true,"delta_db":-2.9254504726712782},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1097616536448917e-07,"output_power":1.418970240533639e-07,"reference_active":true,"delta_db":-1.7226010718919822}],"active_band_count":3,"active_band_mae_db":2.3901644411051914}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":3.917703218483461,"rms_envelope_nrmse":0.2075858839854761,"envelope_correlation":0.9778171414581514,"rms_gain_error_db":-0.9195733450379255,"reference_mean_power":6.940550281225069e-05,"output_mean_power":5.616122476606348e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.994304097313728e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6831637620284676e-05,"output_power":3.045833059435781e-05,"reference_active":true,"delta_db":-0.8251493283492496},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.349662193280123e-06,"output_power":4.708642914887116e-06,"reference_active":true,"delta_db":-0.5543060268327},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1665818120524943e-06,"output_power":1.3212551293614943e-06,"reference_active":true,"delta_db":0.5407148518484781}],"active_band_count":3,"active_band_mae_db":0.6400567356768093}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":3.683962898008455,"rms_envelope_nrmse":0.1703270259022406,"envelope_correlation":0.9869154118512211,"rms_gain_error_db":-0.9195397201643398,"reference_mean_power":6.975388624239838e-05,"output_mean_power":5.644356507153637e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.725701790778612e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6073966740876396e-05,"output_power":3.0197774761789564e-05,"reference_active":true,"delta_db":-0.7721895904632197},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3003381054168375e-06,"output_power":2.0437164337237636e-06,"reference_active":true,"delta_db":-0.5137103648584433},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.80849015690718e-08,"output_power":8.116109657917279e-08,"reference_active":true,"delta_db":0.16780839036792738}],"active_band_count":3,"active_band_mae_db":0.48456944856319684}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"golf","status":"measured","mos":{"status":"measured","value":3.324934482574463,"predicted_mos":3.324934482574463,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"96e814d6b93e2d0a9d7c336ef1bba97e933143b03914945343a225535c280e95","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7574176870002702},"started_at":"2026-10-13T05:03:02.806338+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0039738510004099226,"native_input_samples":144000,"native_input_sha256":"c7fd7744daac3e05a3f47fc6cc29a811ba5e567efeaca252dc5e8d28d9090cf4","output_sha256":"71feb8a99d2c6257c63eb3ef00b46c4a041eb4f6ac7d291d2f6ef78f759adac8","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5999056199998449,"synthesis_s":0.1280387119995794,"total_s":0.7279443319994243,"rtf":0.12132405533323738,"output_peak":0.4875497817993164,"output_rms":0.06832771242816826,"finished_at":"2026-10-13T05:03:03.548722+00:00","analysis_stages":{"external_f0_s":0.09626127799947426},"evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"71feb8a99d2c6257c63eb3ef00b46c4a041eb4f6ac7d291d2f6ef78f759adac8","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.306773529746753,"output_peak":0.4808726666761176},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.633157217631578,"spectral_convergence":0.5832091334342367,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.3422623979068495,"spectral_convergence":0.25602261031279494,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.489808497334406,"spectral_convergence":0.2581255028754814,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012567423285070982,"output_power":0.001110311404759416,"delta_db":-0.5380144208792627,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029891692667967213,"output_power":0.0028497135067131747,"delta_db":-0.20749307586907834,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004885991462843761,"output_power":0.00043453436074573345,"delta_db":-0.5092857988758934,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013463945946263906,"output_power":0.00010416616508741174,"delta_db":-1.1144568365055754,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010141064913289753,"output_power":6.895528597076326e-05,"delta_db":-1.6751599874621095,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00010622591895982053,"output_power":0.00010238128428847428,"delta_db":-0.1600992374367259,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7007515595047743,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3626627589689767,"pitch":null,"output_rms_dbfs":-23.306773529746753,"output_peak":0.4808726666761176,"reference_rms_dbfs":-22.944110770777776,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.821742704290944},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8579661679606367,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07125157380116985,"rms_dbfs":-22.944110770777776,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.975547987388548,"mean":-8.362402302991966e-06,"zero_energy":false},"output":{"rms":0.06833785193847532,"rms_dbfs":-23.306773529746753,"peak":0.4808726666761176,"peak_dbfs":-6.35939815987964,"crest_factor_db":16.94737536986711,"mean":0.0004588572218249146,"zero_energy":false},"rms_difference_db":-0.3626627589689768},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":6.969195342803645e-06,"output_rms_dbfs":-51.56817362268345,"output_zero_energy":false,"output_peak":0.011882668676425704,"output_energy_fraction":8.221425168539069e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":194,"both_estimated_voiced_frames":155,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.9810126582278481,"absolute_difference_cents_median":5.327514578978582,"absolute_difference_cents_p95":14.291484810356454,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00024147266762860616,"log_rms_mae_db":2.1368482599313596,"rms_envelope_nrmse":0.12404372510718928,"envelope_correlation":0.9869080410490173,"rms_gain_error_db":-0.3626732293086128,"reference_mean_power":0.005081021283987993,"output_mean_power":0.004673946007570586},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.414896544930504e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010866167661999122,"output_power":0.0010632551387061843,"reference_active":true,"delta_db":-0.09438911460682367},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002007320631107036,"output_power":0.00019710715420227272,"reference_active":true,"delta_db":-0.07914360551389862},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6899265468956982e-05,"output_power":2.588608189811317e-05,"reference_active":true,"delta_db":-0.1667410014524619}],"active_band_count":3,"active_band_mae_db":0.11342457385772807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.000225157536036158,"log_rms_mae_db":1.323298714681611,"rms_envelope_nrmse":0.07526406855298655,"envelope_correlation":0.996049950138154,"rms_gain_error_db":-0.3627436045151778,"reference_mean_power":0.005106562161586245,"output_mean_power":0.004697364512217734},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.2588743538177738e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010229292215177734,"output_power":0.0010022814254015326,"reference_active":true,"delta_db":-0.08855903114633618},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010354847932821216,"output_power":0.00010272761083543987,"reference_active":true,"delta_db":-0.03456537675973597},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.981806175895444e-06,"output_power":1.9735008628264282e-06,"reference_active":true,"delta_db":-0.018238568597644247}],"active_band_count":3,"active_band_mae_db":0.04712099216790547}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00020992712290116374,"log_rms_mae_db":2.14336803869881,"rms_envelope_nrmse":0.14545489068266684,"envelope_correlation":0.9846151641169619,"rms_gain_error_db":-0.3053747309455929,"reference_mean_power":0.0023427715092846457,"output_mean_power":0.002183697418991121},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4052113317333496e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000740561499580294,"output_power":0.0007204378917182813,"reference_active":true,"delta_db":-0.11964583495456124},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011452025875620422,"output_power":0.0001014910023566463,"reference_active":true,"delta_db":-0.5245477887430134},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.658147612650053e-05,"output_power":1.4416005938558382e-05,"reference_active":true,"delta_db":-2.6573414204322052}],"active_band_count":3,"active_band_mae_db":1.1005116813765934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002036893389343094,"log_rms_mae_db":1.4041808501837234,"rms_envelope_nrmse":0.08742450855864377,"envelope_correlation":0.9948078823947535,"rms_gain_error_db":-0.30538928607845967,"reference_mean_power":0.002354568257036714,"output_mean_power":0.002194685812672965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.329144537148569e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007208168998153304,"output_power":0.000699004815086238,"reference_active":true,"delta_db":-0.13344792894740504},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.456440806684919e-05,"output_power":4.7546345521759325e-05,"reference_active":true,"delta_db":-0.5979230611292808},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1324812964220933e-06,"output_power":1.1114429710131212e-06,"reference_active":true,"delta_db":-2.82998047118534}],"active_band_count":3,"active_band_mae_db":1.1871171537540086}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":2.0263453352509306,"rms_envelope_nrmse":0.23171346500454423,"envelope_correlation":0.9723704158329706,"rms_gain_error_db":-1.4635169824775063,"reference_mean_power":0.0002130039051232873,"output_mean_power":0.00015206731136252794},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1785525738381704e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5734596128251816e-05,"output_power":3.891259778229508e-05,"reference_active":true,"delta_db":-1.5603463341345003},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.430917551116228e-05,"output_power":1.0725503477042277e-05,"reference_active":true,"delta_db":-1.2519692260495496},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.923060214196923e-06,"output_power":1.927202771736616e-06,"reference_active":true,"delta_db":-1.8091035018995991}],"active_band_count":3,"active_band_mae_db":1.5404730206945498}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":1.4300261251080542,"rms_envelope_nrmse":0.18267602730826565,"envelope_correlation":0.9907717264673216,"rms_gain_error_db":-1.4636073682832804,"reference_mean_power":0.0002140715748790745,"output_mean_power":0.0001528263592955242},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0802554426983461e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5403700906444176e-05,"output_power":3.852141259550145e-05,"reference_active":true,"delta_db":-1.5783657159671713},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.412473899747754e-06,"output_power":4.669013001895554e-06,"reference_active":true,"delta_db":-1.3780052712328381},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1084789824227454e-07,"output_power":1.1425768936253656e-07,"reference_active":true,"delta_db":-2.6608383921174465}],"active_band_count":3,"active_band_mae_db":1.872403126439152}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":2.11569395131415,"rms_envelope_nrmse":0.23782442225988576,"envelope_correlation":0.9679261812451954,"rms_gain_error_db":0.24325682541579355,"reference_mean_power":6.952095737798394e-05,"output_mean_power":7.352608241943354e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.987103682306738e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6796678867126466e-05,"output_power":3.5609576352732003e-05,"reference_active":true,"delta_db":-0.1424181571815244},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.301464820026982e-06,"output_power":5.368998081635048e-06,"reference_active":true,"delta_db":0.054973649685792915},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1551804722504114e-06,"output_power":1.8876703240315059e-06,"reference_active":true,"delta_db":2.132763095980064}],"active_band_count":3,"active_band_mae_db":0.7767183009491272}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":1.7230521705964266,"rms_envelope_nrmse":0.1756722717622602,"envelope_correlation":0.9825555604147439,"rms_gain_error_db":0.24325760486312922,"reference_mean_power":6.987073424448129e-05,"output_mean_power":7.389602330779505e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.718894212424245e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.603255644873508e-05,"output_power":3.545990656155976e-05,"reference_active":true,"delta_db":-0.06957488791169936},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2471246618467463e-06,"output_power":2.380902222697008e-06,"reference_active":true,"delta_db":0.25114394377737514},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.761253840149195e-08,"output_power":1.0564193139007769e-07,"reference_active":true,"delta_db":1.3390444482492239}],"active_band_count":3,"active_band_mae_db":0.5532544266460995}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.264751434326172,"predicted_mos":3.264751434326172,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"18311c9ca0ca68f501d018e4f4b2dff73f16af21abfc8a7b7f418f983de5b04a","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5972863110000617},"started_at":"2026-10-13T07:45:38.763466+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002605371999379713,"native_input_samples":144000,"native_input_sha256":"c7fd7744daac3e05a3f47fc6cc29a811ba5e567efeaca252dc5e8d28d9090cf4","output_sha256":"ac6de93ef7793c2a671b2c9903a19b32371f350888ac087964d36ad1958156cb","output_samples":144120,"length_delta_samples":120,"analysis_s":2.185613050998654,"synthesis_s":3.53416840600039,"total_s":5.719781456999044,"rtf":0.9532969094998407,"output_peak":0.4909937381744385,"output_rms":0.07095224248721965,"finished_at":"2026-10-13T07:45:44.495140+00:00","analysis_stages":{"external_f0_s":1.4618427450004674},"evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"ac6de93ef7793c2a671b2c9903a19b32371f350888ac087964d36ad1958156cb","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.979044135209573,"output_peak":0.4878633507457682},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.806475406920662,"spectral_convergence":0.6001554184384171,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.539166265062819,"spectral_convergence":0.2868799982907658,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.469047321237845,"spectral_convergence":0.27398730508031194,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0015465912446872765,"delta_db":0.9084980547404495,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.002761123360433333,"delta_db":-0.33741136797669047,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.00043468434027628665,"delta_db":-0.5005476363978242,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00012229078531183048,"delta_db":-0.41075816932979076,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":9.886912246499342e-05,"delta_db":-0.10308464678016757,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":7.674598763531717e-05,"delta_db":-1.4045319549744228,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6108053050332242,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.024100521805725788,"pitch":null,"output_rms_dbfs":-22.97543196356537,"output_peak":0.4878633507457682,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.938229664407108},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8544138892007186,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07119236623533667,"rms_dbfs":-22.951331441759642,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.982768658370414,"mean":-4.647572835286458e-06,"zero_energy":false},"output":{"rms":0.07099510434395485,"rms_dbfs":-22.97543196356537,"peak":0.4878633507457682,"peak_dbfs":-6.234036114356673,"crest_factor_db":16.741395849208697,"mean":7.475941186786907e-05,"zero_energy":false},"rms_difference_db":-0.024100521805728192},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":1.1398501243217167e-06,"output_rms_dbfs":-59.43152249066269,"output_zero_energy":false,"output_peak":0.0058500964656503385,"output_energy_fraction":1.2438088333488878e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":197,"both_estimated_voiced_frames":156,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.9873417721518988,"absolute_difference_cents_median":5.597198916512374,"absolute_difference_cents_p95":18.776873472456593,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024147266762860616,"log_rms_mae_db":4.052268124324978,"rms_envelope_nrmse":0.15281837132039475,"envelope_correlation":0.9785291503443788,"rms_gain_error_db":-0.024092058515077334,"reference_mean_power":0.005072567967533502,"output_mean_power":0.005044506303107547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.4164286663529534e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010865583816620668,"output_power":0.0011202928517967418,"reference_active":true,"delta_db":0.13278498619647275},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020269542962956985,"output_power":0.00019006093071990214,"reference_active":true,"delta_db":-0.27951104689939077},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.69603101954121e-05,"output_power":2.574368789760073e-05,"reference_active":true,"delta_db":-0.2005412308660867}],"active_band_count":3,"active_band_mae_db":0.20427908798731673}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000225157536036158,"log_rms_mae_db":3.561965211579307,"rms_envelope_nrmse":0.100918710843978,"envelope_correlation":0.9902008195663698,"rms_gain_error_db":-0.024060302616611177,"reference_mean_power":0.005098028148503288,"output_mean_power":0.005069862708294269},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2603245746946088e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001023229039263937,"output_power":0.0010608517739564031,"reference_active":true,"delta_db":0.1568185001276911},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010544090587602268,"output_power":9.69639724352848e-05,"reference_active":true,"delta_db":-0.3639872899523234},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0049541105469597e-06,"output_power":1.9315289222258467e-06,"reference_active":true,"delta_db":-0.16203221378806915}],"active_band_count":3,"active_band_mae_db":0.2276126679560279}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020992712290116374,"log_rms_mae_db":2.746600075038398,"rms_envelope_nrmse":0.18541102268528617,"envelope_correlation":0.9745452135900426,"rms_gain_error_db":-0.36814091724968645,"reference_mean_power":0.0023388664575994473,"output_mean_power":0.0021487769361213543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4073975659228857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007410848450397675,"output_power":0.0006987840255847544,"reference_active":true,"delta_db":-0.2552496384669069},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011531333502346333,"output_power":0.00010796330642507146,"reference_active":true,"delta_db":-0.286033561730328},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.654328508679614e-05,"output_power":2.2065156303101148e-05,"reference_active":true,"delta_db":-0.8024766340078583}],"active_band_count":3,"active_band_mae_db":0.4479199447350311}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002036893389343094,"log_rms_mae_db":1.9948097803277964,"rms_envelope_nrmse":0.1160725219315468,"envelope_correlation":0.9903186103242324,"rms_gain_error_db":-0.36813719796706446,"reference_mean_power":0.0023506239289763434,"output_mean_power":0.0021595806777678296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3312664615669961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215176046332657,"output_power":0.0006817808133229986,"reference_active":true,"delta_db":-0.24602156956943794},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.516212457663081e-05,"output_power":5.019611968521719e-05,"reference_active":true,"delta_db":-0.40970838595796705},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1332065410684056e-06,"output_power":1.6417365160536341e-06,"reference_active":true,"delta_db":-1.137294486762354}],"active_band_count":3,"active_band_mae_db":0.5976748140965863}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":2.4197692751079356,"rms_envelope_nrmse":0.2277230264482524,"envelope_correlation":0.9588953765533008,"rms_gain_error_db":-0.26456617785248066,"reference_mean_power":0.00021265544148097977,"output_mean_power":0.00020008746512127862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1812635588892241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.569393841260955e-05,"output_power":5.40892562406256e-05,"reference_active":true,"delta_db":-0.12696920556579816},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4402699100691383e-05,"output_power":1.334742452004947e-05,"reference_active":true,"delta_db":-0.33046413967030924},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9279580393512075e-06,"output_power":2.688818402476669e-06,"reference_active":true,"delta_db":-0.3700337672864946}],"active_band_count":3,"active_band_mae_db":0.27582237084086736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":1.6297731003878673,"rms_envelope_nrmse":0.1241105260189373,"envelope_correlation":0.9874230192415414,"rms_gain_error_db":-0.26454251774019755,"reference_mean_power":0.00021371687213348283,"output_mean_power":0.00020108726053250843},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0829999638899275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5369039282181205e-05,"output_power":5.3844577918079774e-05,"reference_active":true,"delta_db":-0.12125011135844375},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.495407348960159e-06,"output_power":5.9242042292304336e-06,"reference_active":true,"delta_db":-0.39976370494444696},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1097616536448917e-07,"output_power":1.8358660872791436e-07,"reference_active":true,"delta_db":-0.603923949852073}],"active_band_count":3,"active_band_mae_db":0.37497925538498783}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":2.3540187051600134,"rms_envelope_nrmse":0.22399271947085087,"envelope_correlation":0.9719702660398736,"rms_gain_error_db":-0.905130516804523,"reference_mean_power":6.940550281225069e-05,"output_mean_power":5.634830454448162e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.994304097313728e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6831637620284676e-05,"output_power":2.9681413666931912e-05,"reference_active":true,"delta_db":-0.9373644734171316},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.349662193280123e-06,"output_power":4.218764672405143e-06,"reference_active":true,"delta_db":-1.031410585684533},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1665818120524943e-06,"output_power":6.595556723570983e-07,"reference_active":true,"delta_db":-2.476637414055512}],"active_band_count":3,"active_band_mae_db":1.4818041577190588}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":1.573820249063369,"rms_envelope_nrmse":0.16202883988781058,"envelope_correlation":0.9882248787648253,"rms_gain_error_db":-0.9051380845499429,"reference_mean_power":6.975388624239838e-05,"output_mean_power":5.663104821412233e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.725701790778612e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6073966740876396e-05,"output_power":2.881577650953936e-05,"reference_active":true,"delta_db":-0.9756357336581095},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3003381054168375e-06,"output_power":1.8509927351944912e-06,"reference_active":true,"delta_db":-0.9438695940682431},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.80849015690718e-08,"output_power":4.7693014505753175e-08,"reference_active":true,"delta_db":-2.141122936685748}],"active_band_count":3,"active_band_mae_db":1.3535427548040335}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.4868662357330322,"predicted_mos":3.4868662357330322,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"31450e0af5cd203be0eec0764bf172928246df4c4e90d9a827c14c4d550062b0","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.662103392998688},"started_at":"2026-10-13T05:11:28.921631+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031305630000133533,"native_input_samples":144000,"native_input_sha256":"c7fd7744daac3e05a3f47fc6cc29a811ba5e567efeaca252dc5e8d28d9090cf4","output_sha256":"5ef57877c7b83788fbd55a2897f35903a26c4b909da1f7adc31ef043131a130c","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7325526919994445,"synthesis_s":0.10752989899992826,"total_s":0.8400825909993728,"rtf":0.14001376516656214,"output_peak":0.45528674125671387,"output_rms":0.07010503189556735,"finished_at":"2026-10-13T05:11:29.772695+00:00","analysis_stages":{"external_f0_s":0.1457438050001656},"evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"5ef57877c7b83788fbd55a2897f35903a26c4b909da1f7adc31ef043131a130c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.083296045659758,"output_peak":0.44332153262265},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.08446890685989,"spectral_convergence":0.18409771908577396,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.616832103146134,"spectral_convergence":0.16455231046552213,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.036523156397095,"spectral_convergence":0.18598225047496186,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.001220224601388483,"delta_db":-0.12085965104729038,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.0029360559172957844,"delta_db":-0.07062624638676629,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.0004646110326416118,"delta_db":-0.21139240101325893,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.00012236288176717234,"delta_db":-0.40819854338999184,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":8.676520568646024e-05,"delta_db":-0.670235435113301,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":9.483374815501632e-05,"delta_db":-0.48545959140721706,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3277953113929709,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12473549760573188,"pitch":null,"output_rms_dbfs":-23.076066939365376,"output_peak":0.44332153262265,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.9126080554677065},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9217184450255876,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07119236623533667,"rms_dbfs":-22.951331441759642,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.982768658370414,"mean":-4.647572835286458e-06,"zero_energy":false},"output":{"rms":0.07017729965547641,"rms_dbfs":-23.076066939365376,"peak":0.44332153262265,"peak_dbfs":-7.06562348039329,"crest_factor_db":16.010443458972084,"mean":-6.530331611134899e-05,"zero_energy":false},"rms_difference_db":-0.12473549760573377},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":3.7120736729383065e-06,"output_rms_dbfs":-54.30383412993837,"output_zero_energy":false,"output_peak":0.008027094170765121,"output_energy_fraction":4.145586395591942e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":166,"both_estimated_voiced_frames":156,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9873417721518988,"absolute_difference_cents_median":1.047120000821983,"absolute_difference_cents_p95":4.360099927979728,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024147266762860616,"log_rms_mae_db":1.1445678612376355,"rms_envelope_nrmse":0.0589259423703011,"envelope_correlation":0.9969211902865267,"rms_gain_error_db":-0.12473890604103499,"reference_mean_power":0.005072567967533502,"output_mean_power":0.004928945138397191},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.4164286663529534e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010865583816620668,"output_power":0.00108425621092927,"reference_active":true,"delta_db":-0.009211476870403712},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020269542962956985,"output_power":0.00019003792320181192,"reference_active":true,"delta_db":-0.2800368068725126},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.69603101954121e-05,"output_power":2.113656625526617e-05,"reference_active":true,"delta_db":-1.0569044941234609}],"active_band_count":3,"active_band_mae_db":0.4487175926221257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000225157536036158,"log_rms_mae_db":0.6721549084793669,"rms_envelope_nrmse":0.03178973499229467,"envelope_correlation":0.9992335129536172,"rms_gain_error_db":-0.12473870101958004,"reference_mean_power":0.005098028148503288,"output_mean_power":0.004953684682993396},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2603245746946088e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001023229039263937,"output_power":0.0010254334136615315,"reference_active":true,"delta_db":0.009346078511142969},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010544090587602268,"output_power":9.758593261950066e-05,"reference_active":true,"delta_db":-0.336219114366538},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0049541105469597e-06,"output_power":1.5445130684997655e-06,"reference_active":true,"delta_db":-1.1331284960014352}],"active_band_count":3,"active_band_mae_db":0.49289789629303876}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020992712290116374,"log_rms_mae_db":1.3657550063377775,"rms_envelope_nrmse":0.07493410438137667,"envelope_correlation":0.995861337565811,"rms_gain_error_db":-0.09762086330179019,"reference_mean_power":0.0023388664575994473,"output_mean_power":0.0022868798045450945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4073975659228857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007410848450397675,"output_power":0.0007334024152990467,"reference_active":true,"delta_db":-0.04525596287187899},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011531333502346333,"output_power":0.00010894649656031425,"reference_active":true,"delta_db":-0.24666263715013334},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.654328508679614e-05,"output_power":1.9147471835126005e-05,"reference_active":true,"delta_db":-1.4184323223652169}],"active_band_count":3,"active_band_mae_db":0.5701169741290765}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002036893389343094,"log_rms_mae_db":0.766639621392192,"rms_envelope_nrmse":0.03937995112139628,"envelope_correlation":0.9988923258262693,"rms_gain_error_db":-0.09761857700513998,"reference_mean_power":0.0023506239289763434,"output_mean_power":0.0022983771492017923},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3312664615669961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215176046332657,"output_power":0.0007147958101485589,"reference_active":true,"delta_db":-0.04064933927708667},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.516212457663081e-05,"output_power":5.1778343852016106e-05,"reference_active":true,"delta_db":-0.27492829472281954},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1332065410684056e-06,"output_power":1.6070295733517656e-06,"reference_active":true,"delta_db":-1.2300903775515515}],"active_band_count":3,"active_band_mae_db":0.5152226705171525}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":1.1836248953435444,"rms_envelope_nrmse":0.11635689667301567,"envelope_correlation":0.9913979709246968,"rms_gain_error_db":-0.5049733339781179,"reference_mean_power":0.00021265544148097977,"output_mean_power":0.0001893124459273605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1812635588892241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.569393841260955e-05,"output_power":4.971173150226613e-05,"reference_active":true,"delta_db":-0.4934903998090795},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4402699100691383e-05,"output_power":1.2008355176545953e-05,"reference_active":true,"delta_db":-0.7896036281073632},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9279580393512075e-06,"output_power":1.9731596078315557e-06,"reference_active":true,"delta_db":-1.7140263202173527}],"active_band_count":3,"active_band_mae_db":0.9990401160445984}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":0.7407269622249911,"rms_envelope_nrmse":0.07362280412007498,"envelope_correlation":0.9980050824987108,"rms_gain_error_db":-0.5049631432271453,"reference_mean_power":0.00021371687213348283,"output_mean_power":0.00019025781073754277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0829999638899275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5369039282181205e-05,"output_power":4.918045908934597e-05,"reference_active":true,"delta_db":-0.5147440958812394},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.495407348960159e-06,"output_power":5.4736687864317054e-06,"reference_active":true,"delta_db":-0.7432787777703004},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1097616536448917e-07,"output_power":1.4395231488305e-07,"reference_active":true,"delta_db":-1.6601474134937662}],"active_band_count":3,"active_band_mae_db":0.9727234290484352}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":1.284610860630583,"rms_envelope_nrmse":0.11268575429009238,"envelope_correlation":0.9924424417114743,"rms_gain_error_db":-0.25866426405916954,"reference_mean_power":6.940550281225069e-05,"output_mean_power":6.53924303421769e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.994304097313728e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6831637620284676e-05,"output_power":3.504991987954038e-05,"reference_active":true,"delta_db":-0.21533999475162968},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.349662193280123e-06,"output_power":4.235608607555918e-06,"reference_active":true,"delta_db":-1.0141053697298774},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1665818120524943e-06,"output_power":6.951111452446509e-07,"reference_active":true,"delta_db":-2.248609492083946}],"active_band_count":3,"active_band_mae_db":1.1593516188551511}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":0.9262672636466495,"rms_envelope_nrmse":0.06713363725541753,"envelope_correlation":0.9976375922188145,"rms_gain_error_db":-0.2586552195230541,"reference_mean_power":6.975388624239838e-05,"output_mean_power":6.572080687878933e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.725701790778612e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6073966740876396e-05,"output_power":3.395545290122447e-05,"reference_active":true,"delta_db":-0.26284373032593467},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3003381054168375e-06,"output_power":1.7591177364805415e-06,"reference_active":true,"delta_db":-1.1649676619077076},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.80849015690718e-08,"output_power":4.437563098661137e-08,"reference_active":true,"delta_db":-2.454225257455484}],"active_band_count":3,"active_band_mae_db":1.294012216563042}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0010","method":"magphase","status":"measured","mos":{"status":"measured","value":3.6782734394073486,"predicted_mos":3.6782734394073486,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288884,"input_sample_rate":48000,"input_seconds":6.018416666666667,"input_float32_sha256":"d2c8599c19de0001116d6cbd639e72c74eafd52adb3277fafb18e43aa14ffbe5","internal_samples":96295,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.507018276000963},"started_at":"2026-10-13T05:18:19.934328+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00851890899866703,"native_input_samples":288000,"native_input_sha256":"ccc029ec489d590481767a412b8a6f2b67126a1fb4fd0a1bd13a2f3ebaa41897","output_sha256":"8a5e43826e3906d87140d145279ba272b3c3f44dff5ce639868082c77fbf19b3","output_samples":288884,"length_delta_samples":884,"analysis_s":3.2572991460001504,"synthesis_s":0.9791997199990874,"total_s":4.236498865999238,"rtf":0.7060831443332063,"output_peak":0.4661667631599869,"output_rms":0.06780279822402978,"finished_at":"2026-10-13T05:18:24.192476+00:00","evaluation_input":{"reference":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"8a5e43826e3906d87140d145279ba272b3c3f44dff5ce639868082c77fbf19b3","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96295,"length_delta_samples":295,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.373506867308965,"output_peak":0.45462210285255855},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.996936497222078,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.536159515745612,"spectral_convergence":0.5832539179762966,"spectral_convergence_status":"ok","amplitude_floor":1.642409573690628e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.554385348260885,"spectral_convergence":0.3343974663271138,"spectral_convergence_status":"ok","amplitude_floor":1.497573454010957e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.935643486304176,"spectral_convergence":0.28255214007993323,"spectral_convergence_status":"ok","amplitude_floor":1.2731241516974033e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0012546591027172457,"output_power":0.0011259067374774295,"delta_db":-0.4702332383855127,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0029841932734943235,"output_power":0.002703604095134199,"delta_db":-0.42883851520867405,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004877853564370709,"output_power":0.0004453695345515938,"delta_db":-0.3950825258029994,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013442175178747268,"output_power":0.0001301865002115458,"delta_db":-0.13903598768331538,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00010124396480278944,"output_power":0.0001063267793623312,"delta_db":0.21273515124894932,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001060495609474016,"output_power":0.00010156655544029891,"delta_db":-0.18758150942954108,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30558448795983195,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4088714802897209,"pitch":null,"output_rms_dbfs":-23.36020292204936,"output_peak":0.45462210285255855,"reference_rms_dbfs":-22.951331441759642,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.675396116770224},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7794834559409113,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07119236623533667,"rms_dbfs":-22.951331441759642,"peak":0.44830322265625,"peak_dbfs":-6.9685627833892285,"crest_factor_db":15.982768658370414,"mean":-4.647572835286458e-06,"zero_energy":false},"output":{"rms":0.06791877650677608,"rms_dbfs":-23.36020292204936,"peak":0.45462210285255855,"peak_dbfs":-6.846989070700912,"crest_factor_db":16.513213851348453,"mean":-0.0005063596028056221,"zero_energy":false},"rms_difference_db":-0.40887148028971865},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.83090492116753e-06,"threshold_dbfs":-52.342640401342535,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":33,"quiet_samples":5280,"reference_mean_power":3.5773805751832147e-06,"output_mean_power":6.249885546997066e-06,"output_rms_dbfs":-52.041279357379636,"output_zero_energy":false,"output_peak":0.011484193721683073,"output_energy_fraction":7.451694573965444e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":158,"output_estimated_voiced_frames":152,"both_estimated_voiced_frames":143,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9050632911392406,"absolute_difference_cents_median":5.367921146106325,"absolute_difference_cents_p95":22.6867096560306,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024147266762860616,"log_rms_mae_db":2.837733123550525,"rms_envelope_nrmse":0.1987245420688618,"envelope_correlation":0.9634392026149163,"rms_gain_error_db":-0.4088629283144928,"reference_mean_power":0.005072567967533502,"output_mean_power":0.004616805535777354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.4164286663529534e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010865583816620668,"output_power":0.001013173761643553,"reference_active":true,"delta_db":-0.3036913202167973},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020269542962956985,"output_power":0.00018337105387541505,"reference_active":true,"delta_db":-0.4351317536134762},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.69603101954121e-05,"output_power":2.7293525908435935e-05,"reference_active":true,"delta_db":0.053347588113389695}],"active_band_count":3,"active_band_mae_db":0.2640568873145544}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.000225157536036158,"log_rms_mae_db":1.8312734588307282,"rms_envelope_nrmse":0.131367237743895,"envelope_correlation":0.984148179517326,"rms_gain_error_db":-0.40885320064120984,"reference_mean_power":0.005098028148503288,"output_mean_power":0.004639988551641145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.2603245746946088e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001023229039263937,"output_power":0.0009530600198433908,"reference_active":true,"delta_db":-0.30852605333496663},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010544090587602268,"output_power":9.421318659842499e-05,"reference_active":true,"delta_db":-0.48897435122787913},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0049541105469597e-06,"output_power":2.1761804095525035e-06,"reference_active":true,"delta_db":0.3559045944453467}],"active_band_count":3,"active_band_mae_db":0.3844683330027308}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00020992712290116374,"log_rms_mae_db":3.1341280172145405,"rms_envelope_nrmse":0.24018217116634957,"envelope_correlation":0.9565970299731485,"rms_gain_error_db":-0.43654505777423774,"reference_mean_power":0.0023388664575994473,"output_mean_power":0.0021151974770983427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4073975659228857e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007410848450397675,"output_power":0.0006795771072420676,"reference_active":true,"delta_db":-0.3762919170223231},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011531333502346333,"output_power":0.00010029328117106204,"reference_active":true,"delta_db":-0.606076929067199},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.654328508679614e-05,"output_power":2.2501780468884802e-05,"reference_active":true,"delta_db":-0.7173778828863127}],"active_band_count":3,"active_band_mae_db":0.5665822429919449}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002036893389343094,"log_rms_mae_db":1.9787471644194128,"rms_envelope_nrmse":0.15629230073005582,"envelope_correlation":0.9817120042150606,"rms_gain_error_db":-0.43654568870702604,"reference_mean_power":0.0023506239289763434,"output_mean_power":0.0021258302566601523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.3312664615669961e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007215176046332657,"output_power":0.0006619157208687573,"reference_active":true,"delta_db":-0.37444236157957633},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.516212457663081e-05,"output_power":4.661851411927728e-05,"reference_active":true,"delta_db":-0.7308255768124817},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1332065410684056e-06,"output_power":1.834892374303168e-06,"reference_active":true,"delta_db":-0.6542231092125834}],"active_band_count":3,"active_band_mae_db":0.5864970158682138}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":8.160365671559211e-05,"log_rms_mae_db":2.8357199123308154,"rms_envelope_nrmse":0.2916741034790374,"envelope_correlation":0.9334848677222775,"rms_gain_error_db":-0.03918984659144803,"reference_mean_power":0.00021265544148097977,"output_mean_power":0.00021074511439555136},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1812635588892241e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.569393841260955e-05,"output_power":5.643854500305559e-05,"reference_active":true,"delta_db":0.057678787954541844},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4402699100691383e-05,"output_power":1.4296681305632535e-05,"reference_active":true,"delta_db":-0.032086513482092835},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9279580393512075e-06,"output_power":3.12890546637511e-06,"reference_active":true,"delta_db":0.2882759346011625}],"active_band_count":3,"active_band_mae_db":0.1260137453459324}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.19953792129777e-05,"log_rms_mae_db":1.8361925035009043,"rms_envelope_nrmse":0.17315446608139423,"envelope_correlation":0.9747783404087841,"rms_gain_error_db":-0.03912524553253361,"reference_mean_power":0.00021371687213348283,"output_mean_power":0.00021180016049259965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0829999638899275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.5369039282181205e-05,"output_power":5.6312670721995426e-05,"reference_active":true,"delta_db":0.07339136908701922},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.495407348960159e-06,"output_power":6.6127752590561335e-06,"reference_active":true,"delta_db":0.07777370738722678},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1097616536448917e-07,"output_power":2.5736895281402e-07,"reference_active":true,"delta_db":0.8632276102019699}],"active_band_count":3,"active_band_mae_db":0.3381308955587386}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.380252572576745e-05,"log_rms_mae_db":2.907165481359,"rms_envelope_nrmse":0.3077091577640619,"envelope_correlation":0.9420955689679301,"rms_gain_error_db":0.16229338092307427,"reference_mean_power":6.940550281225069e-05,"output_mean_power":7.204821787445201e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.994304097313728e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6831637620284676e-05,"output_power":3.768862794544378e-05,"reference_active":true,"delta_db":0.09989298219025793},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.349662193280123e-06,"output_power":5.7591524912914695e-06,"reference_active":true,"delta_db":0.32032218790310646},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1665818120524943e-06,"output_power":1.3804224980448005e-06,"reference_active":true,"delta_db":0.7309682763612967}],"active_band_count":3,"active_band_mae_db":0.38372781548488705}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.8084369731997936e-05,"log_rms_mae_db":1.943159754766888,"rms_envelope_nrmse":0.18932596456881934,"envelope_correlation":0.9772770313233318,"rms_gain_error_db":0.16230878049731248,"reference_mean_power":6.975388624239838e-05,"output_mean_power":7.241012326575038e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.725701790778612e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.6073966740876396e-05,"output_power":3.729073623249828e-05,"reference_active":true,"delta_db":0.1440705720209705},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3003381054168375e-06,"output_power":2.574068982289251e-06,"reference_active":true,"delta_db":0.488285077136265},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.80849015690718e-08,"output_power":9.240041614047382e-08,"reference_active":true,"delta_db":0.731068599621944}],"active_band_count":3,"active_band_mae_db":0.4544747495930599}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"hnm","status":"measured","mos":{"status":"measured","value":3.1025078296661377,"predicted_mos":3.1025078296661377,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"dc2ac341a8cb7ed83cf2ddbd85d7facd0cef662a110117eedd297c21b6e499a5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7332746219999535},"started_at":"2026-10-13T04:40:17.236413+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6992999917420093e-05,"native_input_samples":96000,"native_input_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","output_sha256":"74f16809b4de75af06f45ed6283fc23c49612746a8dc08d84f71ba7fd439d6b8","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1556581680006275,"synthesis_s":0.49404414299988275,"total_s":1.6497023110005102,"rtf":0.2749503851667517,"output_peak":0.49832939478939065,"output_rms":0.09038583595609355,"finished_at":"2026-10-13T04:40:18.892123+00:00","evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"74f16809b4de75af06f45ed6283fc23c49612746a8dc08d84f71ba7fd439d6b8","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.897452634623246,"spectral_convergence":0.2560876462750986,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6229387019057295,"spectral_convergence":0.23874729659394286,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.694228123175252,"spectral_convergence":0.2474732429886755,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.0018601819240623544,"delta_db":0.22663054483115172,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.005338634234627717,"delta_db":-0.1646888224109801,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0006007387946894677,"delta_db":0.020633437081245917,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":7.297671505345873e-05,"delta_db":-0.6248708479036348,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":6.199012090608164e-05,"delta_db":-0.13957849739310968,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00023507755214277528,"delta_db":-0.10833510875328399,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.21412287639556768,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06760085189731199,"pitch":null,"output_rms_dbfs":-20.877992418825528,"output_peak":0.49832939478939065,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7382064865680755},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.897452634623246,"spectral_convergence":0.2560876462750986,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.6229387019057295,"spectral_convergence":0.23874729659394286,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.694228123175252,"spectral_convergence":0.2474732429886755,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.0018601819240623544,"delta_db":0.22663054483115172,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.005338634234627717,"delta_db":-0.1646888224109801,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0006007387946894677,"delta_db":0.020633437081245917,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":7.297671505345873e-05,"delta_db":-0.6248708479036348,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":6.199012090608164e-05,"delta_db":-0.13957849739310968,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00023507755214277528,"delta_db":-0.10833510875328399,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.21412287639556768,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06760085189731199,"pitch":null,"output_rms_dbfs":-20.877992418825528,"output_peak":0.49832939478939065,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7382064865680755},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8501369785973285,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09109203863127006,"rms_dbfs":-20.810391566928214,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.407697194788534,"mean":-4.648844401041666e-06,"zero_energy":false},"output":{"rms":0.09038583595609355,"rms_dbfs":-20.877992418825528,"peak":0.49832939478939065,"peak_dbfs":-6.049669889799415,"crest_factor_db":14.828322529026112,"mean":-7.110955360023217e-05,"zero_energy":false},"rms_difference_db":-0.06760085189731413},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":7.061471390422708e-06,"output_rms_dbfs":-51.511047960925154,"output_zero_energy":false,"output_peak":0.015583923793945582,"output_energy_fraction":3.1693184307600075e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":157,"both_estimated_voiced_frames":144,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9411764705882353,"absolute_difference_cents_median":1.7540219350696646,"absolute_difference_cents_p95":9.08485326670841,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026421160622171636,"log_rms_mae_db":1.2852213000675554,"rms_envelope_nrmse":0.09327104876977818,"envelope_correlation":0.9912517715582767,"rms_gain_error_db":-0.06759155802535627,"reference_mean_power":0.008303652784602253,"output_mean_power":0.00817541909675684},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.437173268209281e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017543654513432086,"output_power":0.0017769634666372847,"reference_active":true,"delta_db":0.055584328406419165},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001336868373880785,"output_power":0.0001351629196726144,"reference_active":true,"delta_db":0.04768915095650258},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.23574194293894e-05,"output_power":4.4934034813986485e-05,"reference_active":true,"delta_db":-2.069076522390015}],"active_band_count":3,"active_band_mae_db":0.724116667250979}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022872947944797367,"log_rms_mae_db":0.712631215765194,"rms_envelope_nrmse":0.05539071864182745,"envelope_correlation":0.9967662741589524,"rms_gain_error_db":-0.06783077581850744,"reference_mean_power":0.008338532782066458,"output_mean_power":0.008209308243816877},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.164744003117204e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015799393891106448,"output_power":0.0016020086468033832,"reference_active":true,"delta_db":0.060244293243956505},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.374442374979845e-05,"output_power":6.067038175437069e-05,"reference_active":true,"delta_db":-0.21465471798896318},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2933120282860937e-06,"output_power":1.9952227669818816e-06,"reference_active":true,"delta_db":-2.176414885582523}],"active_band_count":3,"active_band_mae_db":0.8171046322718144}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024243770094671305,"log_rms_mae_db":1.781646970686544,"rms_envelope_nrmse":0.09414606133368893,"envelope_correlation":0.9934564250422638,"rms_gain_error_db":-0.18223102763991722,"reference_mean_power":0.003959461759548422,"output_mean_power":0.0037967591665875908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.015177817247118e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001136879031892713,"output_power":0.0010725631943484697,"reference_active":true,"delta_db":-0.25291366810814986},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001232393146611943,"output_power":0.00010719707315038002,"reference_active":true,"delta_db":-0.6056634674026757},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0364107060899584e-05,"output_power":2.256949642366696e-05,"reference_active":true,"delta_db":-1.2883864493953314}],"active_band_count":3,"active_band_mae_db":0.7156545283020522}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020477863430336455,"log_rms_mae_db":1.1859567272299925,"rms_envelope_nrmse":0.056747303972893126,"envelope_correlation":0.9977136980491162,"rms_gain_error_db":-0.1822318137970239,"reference_mean_power":0.003979363820398493,"output_mean_power":0.003815842719249714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8942412803353574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010558685473009428,"output_power":0.0009921708627062377,"reference_active":true,"delta_db":-0.2702338423887149},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.091449836890925e-05,"output_power":4.5662856788099226e-05,"reference_active":true,"delta_db":-0.47278390666284115},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8755256559270582e-06,"output_power":1.2710422492651745e-06,"reference_active":true,"delta_db":-1.689630227019285}],"active_band_count":3,"active_band_mae_db":0.8108826586902803}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":1.8827638754337268,"rms_envelope_nrmse":0.17360042267821735,"envelope_correlation":0.9798162707941732,"rms_gain_error_db":-0.6041697553676508,"reference_mean_power":0.00010911551831198982,"output_mean_power":9.494444155906038e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.836309277153808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5256828161206523e-05,"output_power":2.0598622526562948e-05,"reference_active":true,"delta_db":-0.885406303085007},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.883426022529869e-06,"output_power":8.585786511232228e-06,"reference_active":true,"delta_db":-0.611274305358098},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.436572483663533e-06,"output_power":4.031760249899743e-06,"reference_active":true,"delta_db":-1.298304839576102}],"active_band_count":3,"active_band_mae_db":0.9316618160064024}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":1.2424948573043557,"rms_envelope_nrmse":0.11009011676804101,"envelope_correlation":0.9940211922856681,"rms_gain_error_db":-0.6050839522036754,"reference_mean_power":0.00010963296530545586,"output_mean_power":9.53746079316917e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.313909407445063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3642314296732627e-05,"output_power":1.9441159085387668e-05,"reference_active":true,"delta_db":-0.8496783239223121},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.770670858321524e-06,"output_power":5.529730018705977e-06,"reference_active":true,"delta_db":-0.879277740086805},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6513846355224545e-07,"output_power":3.0001650588589983e-07,"reference_active":true,"delta_db":-1.9043710519723902}],"active_band_count":3,"active_band_mae_db":1.2111090386605026}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010577651466639849,"log_rms_mae_db":1.8974046806354994,"rms_envelope_nrmse":0.11781674323072264,"envelope_correlation":0.9926093734203877,"rms_gain_error_db":-0.2936116209761675,"reference_mean_power":0.00018020158938811213,"output_mean_power":0.00016842147056705606},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.705326923064481e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.613613486254777e-05,"output_power":1.6361393950109887e-05,"reference_active":true,"delta_db":0.060207867570922044},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.942687703713093e-06,"output_power":9.098343311448394e-06,"reference_active":true,"delta_db":0.07494255614019522},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0438582736284589e-06,"output_power":9.077741860630032e-07,"reference_active":true,"delta_db":-0.6066370902607812}],"active_band_count":3,"active_band_mae_db":0.2472625046572995}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":1.3352080957123484,"rms_envelope_nrmse":0.06526596841991542,"envelope_correlation":0.9983111808295716,"rms_gain_error_db":-0.3105777301236686,"reference_mean_power":0.000177581626259849,"output_mean_power":0.00016532565646413633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4440347897892334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7728340440048443e-05,"output_power":1.7761529924255605e-05,"reference_active":true,"delta_db":0.008122889096197832},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.955564320473049e-06,"output_power":5.097983282080155e-06,"reference_active":true,"delta_db":0.12305289533792091},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.260796671426968e-08,"output_power":6.490325665681727e-08,"reference_active":true,"delta_db":-1.5438186002393621}],"active_band_count":3,"active_band_mae_db":0.558331461557827}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.8264060020446777,"predicted_mos":2.8264060020446777,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"39460582a6d89816b3769e0144d36ba1ea227d2f7a163f90f67a712cb8a7739c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.562312578000274},"started_at":"2026-10-13T04:43:32.488502+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.7164000130142085e-05,"native_input_samples":96000,"native_input_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","output_sha256":"692bbdae8760864b978d0f88e5b35ab1979d408558a0c474fc926ea4056bec73","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6015065880001202,"synthesis_s":0.09615046299950336,"total_s":1.6976570509996236,"rtf":0.2829428418332706,"output_peak":0.6800981461494345,"output_rms":0.10794431618880328,"finished_at":"2026-10-13T04:43:34.196071+00:00","evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"692bbdae8760864b978d0f88e5b35ab1979d408558a0c474fc926ea4056bec73","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.606386363889669,"spectral_convergence":0.5696102621085066,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.364926537566838,"spectral_convergence":0.3766990561202855,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.597554060414715,"spectral_convergence":0.37636650023799306,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.0015268761734887144,"delta_db":-0.6308854731139836,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.008829270722786501,"delta_db":2.0202578400640894,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0008902213413312198,"delta_db":1.7287566629601925,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":9.785048924550132e-05,"delta_db":0.6489160635134684,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":6.795333876355488e-05,"delta_db":0.259304667177225,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00023980333185284573,"delta_db":-0.021894577044371352,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8850025473122215,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4743871565952027,"pitch":null,"output_rms_dbfs":-19.33600441033301,"output_peak":0.6800981461494345,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.856288987290408},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.606386363889669,"spectral_convergence":0.5696102621085066,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.364926537566838,"spectral_convergence":0.3766990561202855,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.597554060414715,"spectral_convergence":0.37636650023799306,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.0015268761734887144,"delta_db":-0.6308854731139836,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.008829270722786501,"delta_db":2.0202578400640894,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0008902213413312198,"delta_db":1.7287566629601925,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":9.785048924550132e-05,"delta_db":0.6489160635134684,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":6.795333876355488e-05,"delta_db":0.259304667177225,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00023980333185284573,"delta_db":-0.021894577044371352,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8850025473122215,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4743871565952027,"pitch":null,"output_rms_dbfs":-19.33600441033301,"output_peak":0.6800981461494345,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.856288987290408},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8909324659603957,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09109203863127006,"rms_dbfs":-20.810391566928214,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.407697194788534,"mean":-4.648844401041666e-06,"zero_energy":false},"output":{"rms":0.10794431618880328,"rms_dbfs":-19.33600441033301,"peak":0.6800981461494345,"peak_dbfs":-3.3485681795406927,"crest_factor_db":15.987436230792317,"mean":-2.8341585629754937e-06,"zero_energy":false},"rms_difference_db":1.4743871565952027},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":3.4263167640469205e-06,"output_rms_dbfs":-54.651724888549495,"output_zero_energy":false,"output_peak":0.020688974020070645,"output_energy_fraction":1.078200136854192e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":161,"both_estimated_voiced_frames":141,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9215686274509803,"absolute_difference_cents_median":4.6544226257045365,"absolute_difference_cents_p95":24.6342988059905,"difference_over_600_cents_frames":7,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026421160622171636,"log_rms_mae_db":2.821771049386147,"rms_envelope_nrmse":0.27366195945868443,"envelope_correlation":0.9708885250162855,"rms_gain_error_db":1.4746989513382993,"reference_mean_power":0.008303652784602253,"output_mean_power":0.011661088089650261},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.437173268209281e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017543654513432086,"output_power":0.00299417988670755,"reference_active":true,"delta_db":2.321578224704443},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001336868373880785,"output_power":0.0002067690101624115,"reference_active":true,"delta_db":1.89396799353278},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.23574194293894e-05,"output_power":8.941454179976579e-05,"reference_active":true,"delta_db":0.9192508567915186}],"active_band_count":3,"active_band_mae_db":1.7115990250095807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022872947944797367,"log_rms_mae_db":2.501116010806593,"rms_envelope_nrmse":0.25067734855434987,"envelope_correlation":0.9798392126180439,"rms_gain_error_db":1.4760677480616968,"reference_mean_power":0.008338532782066458,"output_mean_power":0.011713762525753709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.164744003117204e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015799393891106448,"output_power":0.002756522844986961,"reference_active":true,"delta_db":2.417211695693628},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.374442374979845e-05,"output_power":8.985630528411429e-05,"reference_active":true,"delta_db":1.4910635783340809},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2933120282860937e-06,"output_power":3.7720065427857284e-06,"reference_active":true,"delta_db":0.5893955703201916}],"active_band_count":3,"active_band_mae_db":1.4992236147826334}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024243770094671305,"log_rms_mae_db":2.380433692933585,"rms_envelope_nrmse":0.31655151720728747,"envelope_correlation":0.983930424412724,"rms_gain_error_db":2.072285313627756,"reference_mean_power":0.003959461759548422,"output_mean_power":0.006380646495981565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.015177817247118e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001136879031892713,"output_power":0.0019544763714119816,"reference_active":true,"delta_db":2.3531616779184117},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001232393146611943,"output_power":0.0002244790085350402,"reference_active":true,"delta_db":2.604264610011276},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0364107060899584e-05,"output_power":5.6470665479526906e-05,"reference_active":true,"delta_db":2.6946239168113433}],"active_band_count":3,"active_band_mae_db":2.5506834015803435}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020477863430336455,"log_rms_mae_db":1.9894365605577662,"rms_envelope_nrmse":0.2996734093406644,"envelope_correlation":0.9895961235510077,"rms_gain_error_db":2.072278919205692,"reference_mean_power":0.003979363820398493,"output_mean_power":0.006412709093814482},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8942412803353574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010558685473009428,"output_power":0.0018293054229234977,"reference_active":true,"delta_db":2.3867636872575084},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.091449836890925e-05,"output_power":9.20625890448795e-05,"reference_active":true,"delta_db":2.5724171497324293},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8755256559270582e-06,"output_power":3.4719353414716046e-06,"reference_active":true,"delta_db":2.674486192457861}],"active_band_count":3,"active_band_mae_db":2.5445556764825996}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":1.5679447877199104,"rms_envelope_nrmse":0.18138271771296594,"envelope_correlation":0.9808362162221974,"rms_gain_error_db":0.6180527640132532,"reference_mean_power":0.00010911551831198982,"output_mean_power":0.00012580323132868317},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.836309277153808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5256828161206523e-05,"output_power":2.9057618049384115e-05,"reference_active":true,"delta_db":0.6088120138420543},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.883426022529869e-06,"output_power":1.2557329243464137e-05,"reference_active":true,"delta_db":1.0398976534711648},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.436572483663533e-06,"output_power":4.090483557488689e-06,"reference_active":true,"delta_db":-1.2355053141457444}],"active_band_count":3,"active_band_mae_db":0.9614049938196545}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":0.9536830633460593,"rms_envelope_nrmse":0.11230806160254417,"envelope_correlation":0.994621827918612,"rms_gain_error_db":0.6178101907265244,"reference_mean_power":0.00010963296530545586,"output_mean_power":0.00012639275489108242},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.313909407445063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3642314296732627e-05,"output_power":2.707528794544926e-05,"reference_active":true,"delta_db":0.5888309757071573},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.770670858321524e-06,"output_power":7.687880881795218e-06,"reference_active":true,"delta_db":0.5517494361738497},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6513846355224545e-07,"output_power":3.119978069033181e-07,"reference_active":true,"delta_db":-1.7343071269276713}],"active_band_count":3,"active_band_mae_db":0.9582958462695594}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010577651466639849,"log_rms_mae_db":1.6954304018438102,"rms_envelope_nrmse":0.15548487228440144,"envelope_correlation":0.9862618746529024,"rms_gain_error_db":-0.030405061430519707,"reference_mean_power":0.00018020158938811213,"output_mean_power":0.00017894439965715467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.705326923064481e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.613613486254777e-05,"output_power":1.7500184838892326e-05,"reference_active":true,"delta_db":0.35243120811765827},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.942687703713093e-06,"output_power":8.976856963623426e-06,"reference_active":true,"delta_db":0.01656240634516453},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0438582736284589e-06,"output_power":9.900670169426105e-07,"reference_active":true,"delta_db":-0.2297694510395176}],"active_band_count":3,"active_band_mae_db":0.19958768850078012}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":1.0522443969033743,"rms_envelope_nrmse":0.08383491340875991,"envelope_correlation":0.9959813799298044,"rms_gain_error_db":-0.02898311218130154,"reference_mean_power":0.000177581626259849,"output_mean_power":0.00017640046175999965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4440347897892334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7728340440048443e-05,"output_power":1.907264317149844e-05,"reference_active":true,"delta_db":0.3174280072055906},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.955564320473049e-06,"output_power":4.943477773866513e-06,"reference_active":true,"delta_db":-0.01060531525155211},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.260796671426968e-08,"output_power":9.27552531956288e-08,"reference_active":true,"delta_db":0.006901662541916634}],"active_band_count":3,"active_band_mae_db":0.11164499499968646}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"facodec","status":"measured","mos":{"status":"measured","value":4.0528388023376465,"predicted_mos":4.0528388023376465,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d3294803068891e299ed5e0c43cc4e0d8c53b90b6d6bf299574aa985298d35b8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.743761469002493},"started_at":"2026-10-13T07:00:26.401937+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00012480100122047588,"native_input_samples":96000,"native_input_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","output_sha256":"b6524276d2559eae51e68a0bd6446bba17109eb85e06f111206ce1a1ec6d86b0","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5030318480003189,"synthesis_s":6.904881436001233,"total_s":8.407913284001552,"rtf":1.4013188806669252,"output_peak":0.4635929465293884,"output_rms":0.08445907164637541,"finished_at":"2026-10-13T07:00:34.830019+00:00","evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"b6524276d2559eae51e68a0bd6446bba17109eb85e06f111206ce1a1ec6d86b0","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.602814895821852,"spectral_convergence":0.2318467182610925,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.001165722409421,"spectral_convergence":0.21950938973220543,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.979023667202124,"spectral_convergence":0.2269682997624185,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.001595244856362236,"delta_db":-0.4406501243179873,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.004770264564070344,"delta_db":-0.6535658318582159,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0005000841707737155,"delta_db":-0.7757922715560626,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":7.362051177773695e-05,"delta_db":-0.5867256294243823,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":5.45257704531998e-05,"delta_db":-0.696785222360331,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00013959490993034214,"delta_db":-2.371750878101257,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.920878326269706,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6566823639608871,"pitch":null,"output_rms_dbfs":-21.4670739308891,"output_peak":0.4635929465293884,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.861001428477799},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.602814895821852,"spectral_convergence":0.2318467182610925,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.001165722409421,"spectral_convergence":0.21950938973220543,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.979023667202124,"spectral_convergence":0.2269682997624185,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.001595244856362236,"delta_db":-0.4406501243179873,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.004770264564070344,"delta_db":-0.6535658318582159,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0005000841707737155,"delta_db":-0.7757922715560626,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":7.362051177773695e-05,"delta_db":-0.5867256294243823,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":5.45257704531998e-05,"delta_db":-0.696785222360331,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00013959490993034214,"delta_db":-2.371750878101257,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.920878326269706,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6566823639608871,"pitch":null,"output_rms_dbfs":-21.4670739308891,"output_peak":0.4635929465293884,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.861001428477799},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8460053808706117,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09109203863127006,"rms_dbfs":-20.810391566928214,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.407697194788534,"mean":-4.648844401041666e-06,"zero_energy":false},"output":{"rms":0.08445907164637541,"rms_dbfs":-21.4670739308891,"peak":0.4635929465293884,"peak_dbfs":-6.677263607174369,"crest_factor_db":14.789810323714729,"mean":6.60554841279198e-05,"zero_energy":false},"rms_difference_db":-0.6566823639608863},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":5.496015205542229e-06,"output_rms_dbfs":-52.599520744066666,"output_zero_energy":false,"output_peak":0.007517825812101364,"output_energy_fraction":2.8250539706396984e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":168,"both_estimated_voiced_frames":146,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.954248366013072,"absolute_difference_cents_median":3.6008799935203135,"absolute_difference_cents_p95":14.338323612399241,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026421160622171636,"log_rms_mae_db":1.265909416957543,"rms_envelope_nrmse":0.1205839957493293,"envelope_correlation":0.9900710225056502,"rms_gain_error_db":-0.6566278249464177,"reference_mean_power":0.008303652784602253,"output_mean_power":0.007138490706649454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.437173268209281e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017543654513432086,"output_power":0.0014816045087300217,"reference_active":true,"delta_db":-0.7338777524900195},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001336868373880785,"output_power":0.00011168424017149193,"reference_active":true,"delta_db":-0.7809675547220117},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.23574194293894e-05,"output_power":5.1438128785118276e-05,"reference_active":true,"delta_db":-1.4819790824150485}],"active_band_count":3,"active_band_mae_db":0.9989414632090265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022872947944797367,"log_rms_mae_db":0.999568792430731,"rms_envelope_nrmse":0.11100904118260058,"envelope_correlation":0.9920825728286953,"rms_gain_error_db":-0.6573365266105107,"reference_mean_power":0.008338532782066458,"output_mean_power":0.007167306680664497},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.164744003117204e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015799393891106448,"output_power":0.001327117874461625,"reference_active":true,"delta_db":-0.7573092795917503},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.374442374979845e-05,"output_power":5.4633639318252394e-05,"reference_active":true,"delta_db":-0.6698206825210516},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2933120282860937e-06,"output_power":2.494507524573995e-06,"reference_active":true,"delta_db":-1.2064806203769392}],"active_band_count":3,"active_band_mae_db":0.877870194163247}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024243770094671305,"log_rms_mae_db":1.818568424774663,"rms_envelope_nrmse":0.11652857525205855,"envelope_correlation":0.9930100602130016,"rms_gain_error_db":-0.6367901663733364,"reference_mean_power":0.003959461759548422,"output_mean_power":0.0034194569170365634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.015177817247118e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001136879031892713,"output_power":0.0009424043326227995,"reference_active":true,"delta_db":-0.8147698253792447},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001232393146611943,"output_power":0.0001016045614562476,"reference_active":true,"delta_db":-0.8383606883038195},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0364107060899584e-05,"output_power":3.358772354610794e-05,"reference_active":true,"delta_db":0.43820055900058164}],"active_band_count":3,"active_band_mae_db":0.6971103575612152}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020477863430336455,"log_rms_mae_db":1.3194081732865919,"rms_envelope_nrmse":0.1019640742831749,"envelope_correlation":0.9954132511638069,"rms_gain_error_db":-0.6367961090952367,"reference_mean_power":0.003979363820398493,"output_mean_power":0.003436639964661668},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8942412803353574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010558685473009428,"output_power":0.0008658976001505395,"reference_active":true,"delta_db":-0.8614331711671043},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.091449836890925e-05,"output_power":4.27972831923329e-05,"reference_active":true,"delta_db":-0.7542526876460659},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8755256559270582e-06,"output_power":2.116805882821921e-06,"reference_active":true,"delta_db":0.5255802441105338}],"active_band_count":3,"active_band_mae_db":0.7137553676412347}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":1.8148441658096681,"rms_envelope_nrmse":0.15895534641031342,"envelope_correlation":0.9824614989165926,"rms_gain_error_db":-0.47560794585605093,"reference_mean_power":0.00010911551831198982,"output_mean_power":9.779704339415234e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.836309277153808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5256828161206523e-05,"output_power":2.232273166749466e-05,"reference_active":true,"delta_db":-0.5363147064776086},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.883426022529869e-06,"output_power":8.853591443255099e-06,"reference_active":true,"delta_db":-0.4778803871470928},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.436572483663533e-06,"output_power":4.820483713812168e-06,"reference_active":true,"delta_db":-0.522345627632846}],"active_band_count":3,"active_band_mae_db":0.5121802404191824}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":1.4187474534112683,"rms_envelope_nrmse":0.1236959774557918,"envelope_correlation":0.9894771437082863,"rms_gain_error_db":-0.4757240316132101,"reference_mean_power":0.00010963296530545586,"output_mean_power":9.825818952960111e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.313909407445063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3642314296732627e-05,"output_power":2.046488651661727e-05,"reference_active":true,"delta_db":-0.6268064576269992},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.770670858321524e-06,"output_power":5.593853249190428e-06,"reference_active":true,"delta_db":-0.8292063323910108},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6513846355224545e-07,"output_power":4.1338717509403013e-07,"reference_active":true,"delta_db":-0.5122525500989618}],"active_band_count":3,"active_band_mae_db":0.6560884467056572}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010577651466639849,"log_rms_mae_db":1.8975419306233203,"rms_envelope_nrmse":0.3067774152101829,"envelope_correlation":0.9822268774910762,"rms_gain_error_db":-2.6675342012713528,"reference_mean_power":0.00018020158938811213,"output_mean_power":9.750013051119149e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.705326923064481e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.613613486254777e-05,"output_power":1.1591928060931322e-05,"reference_active":true,"delta_db":-1.4364383770892277},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.942687703713093e-06,"output_power":6.451308211227904e-06,"reference_active":true,"delta_db":-1.418202737904086},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0438582736284589e-06,"output_power":5.597608151884744e-07,"reference_active":true,"delta_db":-2.7063904441121336}],"active_band_count":3,"active_band_mae_db":1.8536771863684824}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":1.5780285984927898,"rms_envelope_nrmse":0.2898224469245573,"envelope_correlation":0.9907427726114207,"rms_gain_error_db":-2.6811000306496178,"reference_mean_power":0.000177581626259849,"output_mean_power":9.578290966892179e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4440347897892334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7728340440048443e-05,"output_power":1.2676725409081997e-05,"reference_active":true,"delta_db":-1.4566099978736338},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.955564320473049e-06,"output_power":3.4833093693554893e-06,"reference_active":true,"delta_db":-1.5310106947100302},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.260796671426968e-08,"output_power":5.5261708016911286e-08,"reference_active":true,"delta_db":-2.2422404513834433}],"active_band_count":3,"active_band_mae_db":1.7432870479890357}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.4829790592193604,"predicted_mos":3.4829790592193604,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"524a3b14ce7a2f5bb54b7dd1b0485df6c22acef2c0fb7db50d2ba877af392a41","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.983604684999591},"started_at":"2026-10-13T07:30:27.037939+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0059593270016193856,"native_input_samples":264600,"native_input_sha256":"cb969f9dea3eae0fbfc3f093c1e0f7becff96690ab5f1b6f0b00bc5b8bd70fa4","output_sha256":"75ddce573d6713175f6e53cc18bdb3beb99362d639e53a580f3cd58912d63510","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13773970200054464,"synthesis_s":3.8305515110005217,"total_s":3.9682912130010664,"rtf":0.661381868833511,"output_peak":0.5656314492225647,"output_rms":0.08089862625308593,"finished_at":"2026-10-13T07:30:31.027436+00:00","evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"75ddce573d6713175f6e53cc18bdb3beb99362d639e53a580f3cd58912d63510","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.83815916941917,"output_peak":0.5655439113290495},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.35362982633687,"spectral_convergence":0.432869601596982,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.0864167856841,"spectral_convergence":0.34435129809737186,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.708226923208693,"spectral_convergence":0.39484128161838045,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.0014799063871787396,"delta_db":-0.7665812083604704,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.00443813193209984,"delta_db":-0.966988410033891,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.00035835549637602387,"delta_db":-2.2230826517597984,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":4.9355440793918636e-05,"delta_db":-2.3233636014013275,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":4.335098353337022e-05,"delta_db":-1.6928137808400199,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00018238762791303754,"delta_db":-1.2104929559395232,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.5305537680558385,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0262090217999726,"pitch":null,"output_rms_dbfs":-21.83660058872819,"output_peak":0.5655439113290495,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.04942451174322},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.655300541238038,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09109203863127006,"rms_dbfs":-20.810391566928214,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.407697194788534,"mean":-4.648844401041666e-06,"zero_energy":false},"output":{"rms":0.08094126183779773,"rms_dbfs":-21.83660058872819,"peak":0.5655439113290495,"peak_dbfs":-4.950673377722839,"crest_factor_db":16.88592721100535,"mean":0.000252721216011786,"zero_energy":false},"rms_difference_db":-1.0262090217999749},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":5.224266184651216e-06,"output_rms_dbfs":-52.819747031763065,"output_zero_energy":false,"output_peak":0.006387352477479902,"output_energy_fraction":2.923861428626806e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":173,"both_estimated_voiced_frames":124,"reference_voiced_output_unvoiced_frames":29,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.8104575163398693,"absolute_difference_cents_median":11.848119851528715,"absolute_difference_cents_p95":44.722069388641,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026421160622171636,"log_rms_mae_db":2.083562548097433,"rms_envelope_nrmse":0.16076912404745766,"envelope_correlation":0.9855866406796069,"rms_gain_error_db":-1.0259105114832165,"reference_mean_power":0.008303652784602253,"output_mean_power":0.006556591552206912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.437173268209281e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017543654513432086,"output_power":0.0014490276463584066,"reference_active":true,"delta_db":-0.8304339464528332},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001336868373880785,"output_power":0.00010847275688213832,"reference_active":true,"delta_db":-0.9076797128312465},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.23574194293894e-05,"output_power":5.0970749253320826e-05,"reference_active":true,"delta_db":-1.5216205206962774}],"active_band_count":3,"active_band_mae_db":1.0865780599934522}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022872947944797367,"log_rms_mae_db":1.6924490951789344,"rms_envelope_nrmse":0.14232060841123328,"envelope_correlation":0.9912696853608118,"rms_gain_error_db":-1.0250725690818376,"reference_mean_power":0.008338532782066458,"output_mean_power":0.0065854034002797435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.164744003117204e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015799393891106448,"output_power":0.001311672723230529,"reference_active":true,"delta_db":-0.8081493921753766},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.374442374979845e-05,"output_power":5.170173237603606e-05,"reference_active":true,"delta_db":-0.9093710415520667},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2933120282860937e-06,"output_power":2.4386731650786068e-06,"reference_active":true,"delta_db":-1.3047928098044757}],"active_band_count":3,"active_band_mae_db":1.0074377478439729}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024243770094671305,"log_rms_mae_db":2.9628849028485535,"rms_envelope_nrmse":0.1747924004666187,"envelope_correlation":0.984522707384884,"rms_gain_error_db":-1.0119121036230185,"reference_mean_power":0.003959461759548422,"output_mean_power":0.0031364974771435875},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.015177817247118e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001136879031892713,"output_power":0.0008165294785257179,"reference_active":true,"delta_db":-1.4374238826700358},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001232393146611943,"output_power":9.800661545811605e-05,"reference_active":true,"delta_db":-0.99493882911736},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0364107060899584e-05,"output_power":2.3981786986296256e-05,"reference_active":true,"delta_db":-1.02478972914696}],"active_band_count":3,"active_band_mae_db":1.1523841469781184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020477863430336455,"log_rms_mae_db":2.477278812091736,"rms_envelope_nrmse":0.15148610950278027,"envelope_correlation":0.9904849409395791,"rms_gain_error_db":-1.0119446024355443,"reference_mean_power":0.003979363820398493,"output_mean_power":0.003152239355428761},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8942412803353574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010558685473009428,"output_power":0.000766910827000507,"reference_active":true,"delta_db":-1.3886498406465742},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.091449836890925e-05,"output_power":4.178818948370179e-05,"reference_active":true,"delta_db":-0.8578791397642682},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8755256559270582e-06,"output_power":1.8358839589078204e-06,"reference_active":true,"delta_db":-0.0927778219047147}],"active_band_count":3,"active_band_mae_db":0.7797689341051858}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":3.9469582153248557,"rms_envelope_nrmse":0.34412762506742006,"envelope_correlation":0.9316864505589934,"rms_gain_error_db":-2.1385419828899774,"reference_mean_power":0.00010911551831198982,"output_mean_power":6.668563970794776e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.836309277153808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5256828161206523e-05,"output_power":1.4442128014292472e-05,"reference_active":true,"delta_db":-2.4274761921843306},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.883426022529869e-06,"output_power":5.289901337078989e-06,"reference_active":true,"delta_db":-2.7145994391541928},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.436572483663533e-06,"output_power":2.5991934612640485e-06,"reference_active":true,"delta_db":-3.2048657688957465}],"active_band_count":3,"active_band_mae_db":2.78231380007809}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":3.551858796934231,"rms_envelope_nrmse":0.3068042708293548,"envelope_correlation":0.9518079874637211,"rms_gain_error_db":-2.140697098368233,"reference_mean_power":0.00010963296530545586,"output_mean_power":6.696863563206855e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.313909407445063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3642314296732627e-05,"output_power":1.3634167965375663e-05,"reference_active":true,"delta_db":-2.3906134650703543},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.770670858321524e-06,"output_power":3.0688066888538133e-06,"reference_active":true,"delta_db":-3.436621699219073},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6513846355224545e-07,"output_power":1.7969514882658163e-07,"reference_active":true,"delta_db":-4.130459012211128}],"active_band_count":3,"active_band_mae_db":3.319231392166852}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010577651466639849,"log_rms_mae_db":3.9017374597513856,"rms_envelope_nrmse":0.2601667642317976,"envelope_correlation":0.9831997744597197,"rms_gain_error_db":-2.0530422589055144,"reference_mean_power":0.00018020158938811213,"output_mean_power":0.00011231930080201973},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.705326923064481e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.613613486254777e-05,"output_power":1.3015550765886965e-05,"reference_active":true,"delta_db":-0.9333696453113545},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.942687703713093e-06,"output_power":5.942498736289702e-06,"reference_active":true,"delta_db":-1.7749896655276816},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0438582736284589e-06,"output_power":5.666719794731246e-07,"reference_active":true,"delta_db":-2.6530979940047708}],"active_band_count":3,"active_band_mae_db":1.7871524349479355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":3.6904336443544725,"rms_envelope_nrmse":0.23313935601344096,"envelope_correlation":0.9922197938944319,"rms_gain_error_db":-2.035192426136506,"reference_mean_power":0.000177581626259849,"output_mean_power":0.00011114214807140137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4440347897892334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7728340440048443e-05,"output_power":1.3886796710646927e-05,"reference_active":true,"delta_db":-1.0606600505682788},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.955564320473049e-06,"output_power":3.2225314313528886e-06,"reference_active":true,"delta_db":-1.868959554208657},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.260796671426968e-08,"output_power":4.183806963522995e-08,"reference_active":true,"delta_db":-3.4507671057778646}],"active_band_count":3,"active_band_mae_db":2.1267955701849335}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"golf","status":"measured","mos":{"status":"measured","value":3.2763843536376953,"predicted_mos":3.2763843536376953,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"e51ac3150b278506ee32421764d83395c51d463e02a5e2c02fc7ae78ffcf9696","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.826518293000845},"started_at":"2026-10-13T05:03:03.549386+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004975005000233068,"native_input_samples":144000,"native_input_sha256":"9758df2fc622e8c12614a34945c40169fa2a435b070c20149225ddb3102e7090","output_sha256":"6db67196e32d4991990a3f3605670e9325d2b316ba5de5dcd9fb43393efe52de","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.5633584440001869,"synthesis_s":0.11969523700008722,"total_s":0.6830536810002741,"rtf":0.11384228016671234,"output_peak":0.5427258014678955,"output_rms":0.08570024538599696,"finished_at":"2026-10-13T05:03:04.248818+00:00","analysis_stages":{"external_f0_s":0.08843585600061488},"evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"6db67196e32d4991990a3f3605670e9325d2b316ba5de5dcd9fb43393efe52de","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.340467751675977,"output_peak":0.5430473761923558},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.522628337626682,"spectral_convergence":0.5159206191610316,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.479289304734047,"spectral_convergence":0.28625917884212454,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.599172953553376,"spectral_convergence":0.28243522085359,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001768481548760848,"output_power":0.0015080312018428907,"delta_db":-0.6919020564065187,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0055542068844593945,"output_power":0.005009250911702768,"delta_db":-0.44849266523139447,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005988871466027565,"output_power":0.0004954389204342904,"delta_db":-0.8235487162600255,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.439571971134211e-05,"output_power":6.476753124351786e-05,"delta_db":-1.1496307783471953,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.350018538781735e-05,"output_power":6.171931166302416e-05,"delta_db":-0.12353919347915117,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00023829174140229103,"output_power":0.00020513974791112064,"delta_db":-0.6505917355155179,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6479508575399671,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5353088568760751,"pitch":null,"output_rms_dbfs":-21.340467751675977,"output_peak":0.5430473761923558,"reference_rms_dbfs":-20.805158894799902,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.867030198638036},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8330634320625274,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09114693207302399,"rms_dbfs":-20.805158894799902,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.402464522660221,"mean":4.090769263062135e-06,"zero_energy":false},"output":{"rms":0.08569916933551697,"rms_dbfs":-21.340467751675977,"peak":0.5430473761923558,"peak_dbfs":-5.303245606330518,"crest_factor_db":16.037222145345456,"mean":0.0002602704171568382,"zero_energy":false},"rms_difference_db":-0.5353088568760747},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":6.509788758366536e-06,"output_rms_dbfs":-51.86433103991598,"output_zero_energy":false,"output_peak":0.010227726907336409,"output_energy_fraction":3.2554387571638466e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":185,"both_estimated_voiced_frames":145,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":40,"paired_voiced_fraction":0.9477124183006536,"absolute_difference_cents_median":7.3104433618579545,"absolute_difference_cents_p95":20.19007330300392,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00026421160622171636,"log_rms_mae_db":2.069015848850474,"rms_envelope_nrmse":0.15721502439269594,"envelope_correlation":0.9777552321432635,"rms_gain_error_db":-0.5359983827637593,"reference_mean_power":0.008313928827730537,"output_mean_power":0.007348631398926933},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.440570828100632e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017571701849000453,"output_power":0.0015333724011156276,"reference_active":true,"delta_db":-0.5916618348132634},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001329694665213552,"output_power":0.0001527663284967284,"reference_active":true,"delta_db":0.6027571480067657},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.220710673672786e-05,"output_power":4.7699570465888515e-05,"reference_active":true,"delta_db":-1.8006547537684399}],"active_band_count":3,"active_band_mae_db":0.9983579121961563}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00022872947944797367,"log_rms_mae_db":1.3826228020224856,"rms_envelope_nrmse":0.11539146021400368,"envelope_correlation":0.9891862138036402,"rms_gain_error_db":-0.5370407144029953,"reference_mean_power":0.008347831421505005,"output_mean_power":0.007376827003257853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.167788670851285e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015820846695058606,"output_power":0.0013844432487312875,"reference_active":true,"delta_db":-0.5795456442463796},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.317487466984407e-05,"output_power":8.063276653887503e-05,"reference_active":true,"delta_db":1.0596717183997852},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.283088116679328e-06,"output_power":2.645711088755509e-06,"reference_active":true,"delta_db":-0.9374012161373345}],"active_band_count":3,"active_band_mae_db":0.8588728595944998}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00024243770094671305,"log_rms_mae_db":2.0860720779971116,"rms_envelope_nrmse":0.16186500454560754,"envelope_correlation":0.9827738679451257,"rms_gain_error_db":-0.6475065915891209,"reference_mean_power":0.003966072058043115,"output_mean_power":0.003416724338512653},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0180955623988663e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0011391041185374276,"output_power":0.0009292951472018117,"reference_active":true,"delta_db":-0.8840975283333584},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012238623196372665,"output_power":8.084002415102396e-05,"reference_active":true,"delta_db":-1.8010612931112697},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.040954006354703e-05,"output_power":2.3849037884050665e-05,"reference_active":true,"delta_db":-1.0553898814493343}],"active_band_count":3,"active_band_mae_db":1.246849567631321}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00020477863430336455,"log_rms_mae_db":1.4323808784366336,"rms_envelope_nrmse":0.1238063399241186,"envelope_correlation":0.9912063221616165,"rms_gain_error_db":-0.64751087988979,"reference_mean_power":0.003986037294790924,"output_mean_power":0.0034339207639680533},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.896910970390449e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010574441534476522,"output_power":0.0008457263279538601,"reference_active":true,"delta_db":-0.9702758971693766},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.033593314860797e-05,"output_power":3.794535103011287e-05,"reference_active":true,"delta_db":-1.2271954934238443},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8757392155499253e-06,"output_power":1.6576575035180618e-06,"reference_active":true,"delta_db":-0.5367765433991218}],"active_band_count":3,"active_band_mae_db":0.9114159779974477}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":1.880276714418399,"rms_envelope_nrmse":0.22055704814910196,"envelope_correlation":0.9701045768772302,"rms_gain_error_db":-1.0697818091435232,"reference_mean_power":0.00010927473455748256,"output_mean_power":8.541646211013715e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":6.853788264353264e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5345804303216638e-05,"output_power":1.933727166874982e-05,"reference_active":true,"delta_db":-1.1751087866574996},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.913748150082845e-06,"output_power":7.322229248299976e-06,"reference_active":true,"delta_db":-1.3159455996175975},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.429401743454898e-06,"output_power":2.51291010126232e-06,"reference_active":true,"delta_db":-3.345750259619707}],"active_band_count":3,"active_band_mae_db":1.9456015486316014}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":1.385392902378307,"rms_envelope_nrmse":0.1530667928901372,"envelope_correlation":0.9909154568732153,"rms_gain_error_db":-1.0707616348014983,"reference_mean_power":0.00010980099508020556,"output_mean_power":8.580846092840561e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.330001286373085e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3721175891384054e-05,"output_power":1.786894489847932e-05,"reference_active":true,"delta_db":-1.230373041779789},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.775967139101014e-06,"output_power":4.393227099763338e-06,"reference_active":true,"delta_db":-1.8818763726545744},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.651417139019633e-07,"output_power":1.4516944715326363e-07,"reference_active":true,"delta_db":-5.057100658667348}],"active_band_count":3,"active_band_mae_db":2.7231166910339035}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00010577651466639849,"log_rms_mae_db":2.2079683013916838,"rms_envelope_nrmse":0.16295491221954572,"envelope_correlation":0.9850682177995572,"rms_gain_error_db":-0.20073807379052788,"reference_mean_power":0.0001790128989440823,"output_mean_power":0.00017092694017284125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.702842374358609e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.6228045908598298e-05,"output_power":1.550356119859754e-05,"reference_active":true,"delta_db":-0.19834759709779448},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.952937005049124e-06,"output_power":9.49132669280408e-06,"reference_active":true,"delta_db":0.2536139342505383},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0415647339226764e-06,"output_power":1.418533765496468e-06,"reference_active":true,"delta_db":1.341534110026266}],"active_band_count":3,"active_band_mae_db":0.5978318804581996}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":1.8614791097914036,"rms_envelope_nrmse":0.10598743554168272,"envelope_correlation":0.9938763988803582,"rms_gain_error_db":-0.24381523459601126,"reference_mean_power":0.00017522329916629047,"output_mean_power":0.00016565720503070494},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.43611899320361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7830993982677495e-05,"output_power":1.7588041299916684e-05,"reference_active":true,"delta_db":-0.05958076696695464},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.939802719334917e-06,"output_power":5.234713796262298e-06,"reference_active":true,"delta_db":0.2518333709283119},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.275660928943433e-08,"output_power":1.2582776731011952e-07,"reference_active":true,"delta_db":1.3243162588839383}],"active_band_count":3,"active_band_mae_db":0.5452434655930682}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.534252882003784,"predicted_mos":3.534252882003784,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"fcc3522a4d899d8a93fec05e72aa648fec3ed0c270d01cfeaee7bfbc75a8959d","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6412508149987843},"started_at":"2026-10-13T07:45:44.495785+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035531159992387984,"native_input_samples":144000,"native_input_sha256":"9758df2fc622e8c12614a34945c40169fa2a435b070c20149225ddb3102e7090","output_sha256":"2455ad51304a44769468caeec2e6868153f96784a08f3c150f9afde75da3a648","output_samples":144120,"length_delta_samples":120,"analysis_s":2.267752615000063,"synthesis_s":3.425249107000127,"total_s":5.69300172200019,"rtf":0.948833620333365,"output_peak":0.4923818111419678,"output_rms":0.08858255571969473,"finished_at":"2026-10-13T07:45:50.201305+00:00","analysis_stages":{"external_f0_s":1.4739663500004099},"evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"2455ad51304a44769468caeec2e6868153f96784a08f3c150f9afde75da3a648","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.05201948463879,"output_peak":0.4869067336523911},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.400005527268199,"spectral_convergence":0.5404500167857,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.528129877563437,"spectral_convergence":0.31305725604824786,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4831899940996465,"spectral_convergence":0.2983021108298124,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.001901657886896511,"delta_db":0.3224002353532136,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.005089898523708552,"delta_db":-0.3718992535843551,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0005444595438735823,"delta_db":-0.406567201524182,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":7.705287153747405e-05,"delta_db":-0.3888256683548932,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":6.627619621084205e-05,"delta_db":0.15077241859568036,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00017053385242321242,"delta_db":-1.502340675998922,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5238009089018744,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2409788776917351,"pitch":null,"output_rms_dbfs":-21.05137044461995,"output_peak":0.4869067336523911,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.803775132977094},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.853793826460194,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09109203863127006,"rms_dbfs":-20.810391566928214,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.407697194788534,"mean":-4.648844401041666e-06,"zero_energy":false},"output":{"rms":0.08859954218081588,"rms_dbfs":-21.05137044461995,"peak":0.4869067336523911,"peak_dbfs":-6.251084387173915,"crest_factor_db":14.800286057446035,"mean":0.00032965312342420815,"zero_energy":false},"rms_difference_db":-0.24097887769173454},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":1.79524566370034e-05,"output_rms_dbfs":-47.45876113609726,"output_zero_energy":false,"output_peak":0.1501404584855999,"output_energy_fraction":8.385565610222569e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":182,"both_estimated_voiced_frames":145,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9477124183006536,"absolute_difference_cents_median":7.394608739270823,"absolute_difference_cents_p95":20.305752413908262,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026421160622171636,"log_rms_mae_db":3.5194256896181018,"rms_envelope_nrmse":0.15860552745637022,"envelope_correlation":0.9750181941495495,"rms_gain_error_db":-0.2415661384911791,"reference_mean_power":0.008303652784602253,"output_mean_power":0.007854391904096347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.437173268209281e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017543654513432086,"output_power":0.0016241531288356493,"reference_active":true,"delta_db":-0.3349309310485396},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001336868373880785,"output_power":0.0001217299899035387,"reference_active":true,"delta_db":-0.4069106336675577},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.23574194293894e-05,"output_power":4.650468295943678e-05,"reference_active":true,"delta_db":-1.9198638172583986}],"active_band_count":3,"active_band_mae_db":0.887235127324832}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022872947944797367,"log_rms_mae_db":3.0459811242935646,"rms_envelope_nrmse":0.10428466442761643,"envelope_correlation":0.9890395946161292,"rms_gain_error_db":-0.2407346834598471,"reference_mean_power":0.008338532782066458,"output_mean_power":0.007888894935149004},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.164744003117204e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015799393891106448,"output_power":0.0014687552995982164,"reference_active":true,"delta_db":-0.3169097987379665},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.374442374979845e-05,"output_power":5.720324024850897e-05,"reference_active":true,"delta_db":-0.4702156956902782},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2933120282860937e-06,"output_power":2.132582366441322e-06,"reference_active":true,"delta_db":-1.8872706648464428}],"active_band_count":3,"active_band_mae_db":0.8914653864248958}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024243770094671305,"log_rms_mae_db":2.479181212844576,"rms_envelope_nrmse":0.17235377786939923,"envelope_correlation":0.978115430331904,"rms_gain_error_db":-0.35426486796216466,"reference_mean_power":0.003959461759548422,"output_mean_power":0.0036493008318516863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.015177817247118e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001136879031892713,"output_power":0.0010438996459628885,"reference_active":true,"delta_db":-0.37055506321864384},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001232393146611943,"output_power":0.00010810139144467549,"reference_active":true,"delta_db":-0.5691799044098478},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0364107060899584e-05,"output_power":2.8544983411570076e-05,"reference_active":true,"delta_db":-0.2683071907673081}],"active_band_count":3,"active_band_mae_db":0.40268071946526657}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020477863430336455,"log_rms_mae_db":1.872927428895331,"rms_envelope_nrmse":0.11230083340630855,"envelope_correlation":0.9910595380712424,"rms_gain_error_db":-0.35439380789122205,"reference_mean_power":0.003979363820398493,"output_mean_power":0.0036675349934493402},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8942412803353574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010558685473009428,"output_power":0.0009654059984534325,"reference_active":true,"delta_db":-0.3889986017114522},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.091449836890925e-05,"output_power":4.577061901996803e-05,"reference_active":true,"delta_db":-0.46254683234967975},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8755256559270582e-06,"output_power":2.3498133346682275e-06,"reference_active":true,"delta_db":0.9791035461472765}],"active_band_count":3,"active_band_mae_db":0.6102163267361361}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":2.154779068205023,"rms_envelope_nrmse":0.25790641789572055,"envelope_correlation":0.9493214068053499,"rms_gain_error_db":-0.1782115686868507,"reference_mean_power":0.00010911551831198982,"output_mean_power":0.00010472861570195188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.836309277153808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5256828161206523e-05,"output_power":2.4275499211614277e-05,"reference_active":true,"delta_db":-0.17210639771547911},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.883426022529869e-06,"output_power":9.49430703837142e-06,"reference_active":true,"delta_db":-0.17444243570941795},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.436572483663533e-06,"output_power":4.4270781961569235e-06,"reference_active":true,"delta_db":-0.8920798962622161}],"active_band_count":3,"active_band_mae_db":0.41287624322903776}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":1.3471812640276182,"rms_envelope_nrmse":0.13818457364990114,"envelope_correlation":0.9846282829485616,"rms_gain_error_db":-0.1805096359036236,"reference_mean_power":0.00010963296530545586,"output_mean_power":0.00010516959398299068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.313909407445063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3642314296732627e-05,"output_power":2.2377261417582696e-05,"reference_active":true,"delta_db":-0.23883051015336204},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.770670858321524e-06,"output_power":6.281159572732497e-06,"reference_active":true,"delta_db":-0.32591875226677613},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6513846355224545e-07,"output_power":3.761444549271481e-07,"reference_active":true,"delta_db":-0.9222759008742726}],"active_band_count":3,"active_band_mae_db":0.4956750544314703}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010577651466639849,"log_rms_mae_db":2.1791971572764353,"rms_envelope_nrmse":0.19581713547723215,"envelope_correlation":0.9826212440652766,"rms_gain_error_db":-0.9552949715636184,"reference_mean_power":0.00018020158938811213,"output_mean_power":0.00014462025401496863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.705326923064481e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.613613486254777e-05,"output_power":1.3618101679422612e-05,"reference_active":true,"delta_db":-0.7368294245529766},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.942687703713093e-06,"output_power":7.816123522241327e-06,"reference_active":true,"delta_db":-0.5847665045230311},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0438582736284589e-06,"output_power":5.670822411366858e-07,"reference_active":true,"delta_db":-2.64995490741406}],"active_band_count":3,"active_band_mae_db":1.3238502788300226}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":1.455459278466713,"rms_envelope_nrmse":0.15054344233927125,"envelope_correlation":0.9922873598063721,"rms_gain_error_db":-0.9407808766603726,"reference_mean_power":0.000177581626259849,"output_mean_power":0.00014299470008420393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4440347897892334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7728340440048443e-05,"output_power":1.4890123356062852e-05,"reference_active":true,"delta_db":-0.7576978730638926},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.955564320473049e-06,"output_power":4.242975777396786e-06,"reference_active":true,"delta_db":-0.6742256496876898},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.260796671426968e-08,"output_power":5.301188710096447e-08,"reference_active":true,"delta_db":-2.4227508462222174}],"active_band_count":3,"active_band_mae_db":1.2848914563246}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.269472599029541,"predicted_mos":3.269472599029541,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"2457eb1f1fb6337f9dd8fe08a7cd2e30ecdb99b46668b8665d1c5241b0f5d26f","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.5462959589985985},"started_at":"2026-10-13T05:11:29.774668+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002777522000542376,"native_input_samples":144000,"native_input_sha256":"9758df2fc622e8c12614a34945c40169fa2a435b070c20149225ddb3102e7090","output_sha256":"699e98ed91f1d0300ad848cec2ff08d8deb04ad63911039693b98bd94bb198e1","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9011080899999797,"synthesis_s":0.12339756400069746,"total_s":1.0245056540006772,"rtf":0.1707509423334462,"output_peak":0.48539745807647705,"output_rms":0.08918430274922373,"finished_at":"2026-10-13T05:11:30.811564+00:00","analysis_stages":{"external_f0_s":0.14796459399985906},"evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"699e98ed91f1d0300ad848cec2ff08d8deb04ad63911039693b98bd94bb198e1","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.993716560255336,"output_peak":0.4838048116366921},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.942155213902515,"spectral_convergence":0.20044452742360672,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.62938074668301,"spectral_convergence":0.17888584610415015,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.965270697011969,"spectral_convergence":0.19202452670032777,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.0016713871665248014,"delta_db":-0.23815302224110746,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.005378654554633084,"delta_db":-0.13225396827847966,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0005747235842300891,"delta_db":-0.17163316031905312,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":7.507087886997884e-05,"delta_db":-0.5019989482010655,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":5.893517859629714e-05,"delta_db":-0.35905729194437486,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00020897617912224448,"delta_db":-0.6194788627972357,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3370958756302194,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1762527829107059,"pitch":null,"output_rms_dbfs":-20.98664434983892,"output_peak":0.4838048116366921,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.845602219199165},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9154437885395381,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09109203863127006,"rms_dbfs":-20.810391566928214,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.407697194788534,"mean":-4.648844401041666e-06,"zero_energy":false},"output":{"rms":0.08926224029216664,"rms_dbfs":-20.98664434983892,"peak":0.4838048116366921,"peak_dbfs":-6.306596334328878,"crest_factor_db":14.680048015510042,"mean":0.00021550366203445067,"zero_energy":false},"rms_difference_db":-0.176252782910705},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":7.458903758631636e-06,"output_rms_dbfs":-51.273249964631475,"output_zero_energy":false,"output_peak":0.026881609546633555,"output_energy_fraction":3.432502553270532e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":156,"both_estimated_voiced_frames":142,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9281045751633987,"absolute_difference_cents_median":1.4595094521659457,"absolute_difference_cents_p95":6.037222746098375,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026421160622171636,"log_rms_mae_db":1.136074255347879,"rms_envelope_nrmse":0.0711761349720581,"envelope_correlation":0.9953005497634668,"rms_gain_error_db":-0.17611463620040294,"reference_mean_power":0.008303652784602253,"output_mean_power":0.007973660075332706},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.437173268209281e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017543654513432086,"output_power":0.0017059808861367232,"reference_active":true,"delta_db":-0.12145905198018322},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001336868373880785,"output_power":0.00012058281192572508,"reference_active":true,"delta_db":-0.44803242203699956},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.23574194293894e-05,"output_power":6.430821293233644e-05,"reference_active":true,"delta_db":-0.5121662859629144}],"active_band_count":3,"active_band_mae_db":0.36055258666003237}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022872947944797367,"log_rms_mae_db":0.6745530893374728,"rms_envelope_nrmse":0.0432277642109528,"envelope_correlation":0.9984697566840041,"rms_gain_error_db":-0.17559575701674945,"reference_mean_power":0.008338532782066458,"output_mean_power":0.008008110641049055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.164744003117204e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015799393891106448,"output_power":0.0015400714657055272,"reference_active":true,"delta_db":-0.1109955215438433},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.374442374979845e-05,"output_power":5.787570863835595e-05,"reference_active":true,"delta_db":-0.4194587782772907},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2933120282860937e-06,"output_power":3.0273205700336095e-06,"reference_active":true,"delta_db":-0.36574468498440693}],"active_band_count":3,"active_band_mae_db":0.2987329949351803}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024243770094671305,"log_rms_mae_db":1.1687084441386655,"rms_envelope_nrmse":0.06291456447843065,"envelope_correlation":0.9972112516925431,"rms_gain_error_db":-0.16530424348243714,"reference_mean_power":0.003959461759548422,"output_mean_power":0.0038115860478373205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.015177817247118e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001136879031892713,"output_power":0.0010996772791808018,"reference_active":true,"delta_db":-0.14449004581212702},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001232393146611943,"output_power":0.00011080264794852054,"reference_active":true,"delta_db":-0.46199135288091453},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0364107060899584e-05,"output_power":2.436714840661621e-05,"reference_active":true,"delta_db":-0.95555805686293}],"active_band_count":3,"active_band_mae_db":0.5206798185186572}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020477863430336455,"log_rms_mae_db":0.5951320557957113,"rms_envelope_nrmse":0.03698512859570678,"envelope_correlation":0.9991758391993654,"rms_gain_error_db":-0.16530564781287416,"reference_mean_power":0.003979363820398493,"output_mean_power":0.0038307435792033977},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8942412803353574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010558685473009428,"output_power":0.0010155537207327775,"reference_active":true,"delta_db":-0.16906951461424294},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.091449836890925e-05,"output_power":4.801957652999026e-05,"reference_active":true,"delta_db":-0.254231434452302},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8755256559270582e-06,"output_power":1.3059252805702095e-06,"reference_active":true,"delta_db":-1.5720468019635407}],"active_band_count":3,"active_band_mae_db":0.6651159170100286}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":1.190438980607767,"rms_envelope_nrmse":0.13235695306012948,"envelope_correlation":0.9887558722682439,"rms_gain_error_db":-0.5166878454500565,"reference_mean_power":0.00010911551831198982,"output_mean_power":9.687634282506613e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.836309277153808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5256828161206523e-05,"output_power":2.243679010048864e-05,"reference_active":true,"delta_db":-0.5141808438592119},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.883426022529869e-06,"output_power":9.391618119619639e-06,"reference_active":true,"delta_db":-0.2216709086590826},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.436572483663533e-06,"output_power":3.7616700557390738e-06,"reference_active":true,"delta_db":-1.5994448271102366}],"active_band_count":3,"active_band_mae_db":0.7784321932095103}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":0.7486089325651044,"rms_envelope_nrmse":0.08957972883552952,"envelope_correlation":0.9960313184931369,"rms_gain_error_db":-0.5166378696847783,"reference_mean_power":0.00010963296530545586,"output_mean_power":9.733686935231136e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.313909407445063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3642314296732627e-05,"output_power":2.1092293080856327e-05,"reference_active":true,"delta_db":-0.49566189175108494},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.770670858321524e-06,"output_power":6.128343495484253e-06,"reference_active":true,"delta_db":-0.4328860237121113},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6513846355224545e-07,"output_power":2.839547420345794e-07,"reference_active":true,"delta_db":-2.1433312819361934}],"active_band_count":3,"active_band_mae_db":1.0239597324664633}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010577651466639849,"log_rms_mae_db":1.1820839207057607,"rms_envelope_nrmse":0.10824314994954176,"envelope_correlation":0.9938945211011816,"rms_gain_error_db":-0.3220802055532737,"reference_mean_power":0.00018020158938811213,"output_mean_power":0.00016732105590791692},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.705326923064481e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.613613486254777e-05,"output_power":1.45659403655212e-05,"reference_active":true,"delta_db":-0.44460987390222806},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.942687703713093e-06,"output_power":7.914785453220467e-06,"reference_active":true,"delta_db":-0.5302891769796702},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0438582736284589e-06,"output_power":1.092484591690465e-06,"reference_active":true,"delta_db":0.19773782647670599}],"active_band_count":3,"active_band_mae_db":0.39087895911953474}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":0.8087311958014805,"rms_envelope_nrmse":0.061100083934902454,"envelope_correlation":0.9985692818304011,"rms_gain_error_db":-0.3218126028355462,"reference_mean_power":0.000177581626259849,"output_mean_power":0.00016489852410111813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4440347897892334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7728340440048443e-05,"output_power":1.59288061092883e-05,"reference_active":true,"delta_db":-0.46484857030419025},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.955564320473049e-06,"output_power":4.466007964443519e-06,"reference_active":true,"delta_db":-0.4517362417707182},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.260796671426968e-08,"output_power":8.575988504104195e-08,"reference_active":true,"delta_db":-0.3336415881782372}],"active_band_count":3,"active_band_mae_db":0.41674213341771527}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274381-0013","method":"magphase","status":"measured","mos":{"status":"measured","value":3.523186206817627,"predicted_mos":3.523186206817627,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289733,"input_sample_rate":48000,"input_seconds":6.036104166666667,"input_float32_sha256":"cdce8500669df18d8d931ef231aedb07d9667564246ba212d894ab33a69a1cee","internal_samples":96578,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.924130948000311},"started_at":"2026-10-13T05:18:24.195250+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.01083993899919733,"native_input_samples":288000,"native_input_sha256":"d10bff330ccb0d8a3a147cf3c98994bdc6a3779c5186e18873c61dc7b2781d0a","output_sha256":"d8c019f67549c75eafd029fdb47898f2d05e900abd10cc72090664b574ef4b05","output_samples":289733,"length_delta_samples":1733,"analysis_s":3.1314570249996905,"synthesis_s":0.941540112000439,"total_s":4.072997137000129,"rtf":0.6788328561666882,"output_peak":0.5298121486515008,"output_rms":0.08630575414992864,"finished_at":"2026-10-13T05:18:28.291957+00:00","evaluation_input":{"reference":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"d8c019f67549c75eafd029fdb47898f2d05e900abd10cc72090664b574ef4b05","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96578,"length_delta_samples":578,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.27918267848792,"output_peak":0.5291954967387928},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9940152001491023,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.629771196974284,"spectral_convergence":0.5992668615505091,"spectral_convergence_status":"ok","amplitude_floor":2.0096490980240517e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.9692120795368595,"spectral_convergence":0.4745177477710871,"spectral_convergence_status":"ok","amplitude_floor":1.509105853911261e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.310973572273285,"spectral_convergence":0.4942666223501181,"spectral_convergence_status":"ok","amplitude_floor":1.2442618586634314e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017656001839806658,"output_power":0.0015487644201910742,"delta_db":-0.5690700238887179,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.005544968065872631,"output_power":0.005020656523454213,"delta_db":-0.4313853819592871,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005978914392629673,"output_power":0.0005437481986899605,"delta_db":-0.4122450389424403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.426968976514186e-05,"output_power":8.297236783395484e-05,"delta_db":-0.0673791109034421,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":6.40147911812626e-05,"output_power":6.397925586134443e-05,"delta_db":-0.0024114861755263244,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00024101533193812626,"output_power":0.00021757456324035318,"delta_db":-0.44436550251072837,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.32114275739669035,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.45193076994455544,"pitch":null,"output_rms_dbfs":-21.26232233687277,"output_peak":0.5291954967387928,"reference_rms_dbfs":-20.810391566928214,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.969985616261475},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6081369445921204,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09109203863127006,"rms_dbfs":-20.810391566928214,"peak":0.536865234375,"peak_dbfs":-5.402694372139679,"crest_factor_db":15.407697194788534,"mean":-4.648844401041666e-06,"zero_energy":false},"output":{"rms":0.08647366841571427,"rms_dbfs":-21.26232233687277,"peak":0.5291954967387928,"peak_dbfs":-5.52767720313798,"crest_factor_db":15.73464513373479,"mean":-0.0005283259016656799,"zero_energy":false},"rms_difference_db":-0.45193076994455694},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.98077728622593e-06,"threshold_dbfs":-51.5609621744498,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":22,"quiet_samples":3520,"reference_mean_power":4.3761180925437e-06,"output_mean_power":9.654973796006332e-06,"output_rms_dbfs":-50.15248900575969,"output_zero_energy":false,"output_peak":0.013904510664249354,"output_energy_fraction":4.7342889789558076e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":153,"output_estimated_voiced_frames":132,"both_estimated_voiced_frames":109,"reference_voiced_output_unvoiced_frames":44,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.7124183006535948,"absolute_difference_cents_median":10.659699627658558,"absolute_difference_cents_p95":41.31499209393033,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00026421160622171636,"log_rms_mae_db":4.054755514358511,"rms_envelope_nrmse":0.3237615080554693,"envelope_correlation":0.8938060009187386,"rms_gain_error_db":-0.452361402536225,"reference_mean_power":0.008303652784602253,"output_mean_power":0.007482264226865739},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.437173268209281e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017543654513432086,"output_power":0.0016228849760539941,"reference_active":true,"delta_db":-0.33832326440278415},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001336868373880785,"output_power":0.0001425410056150344,"reference_active":true,"delta_db":0.2785116902076692},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.23574194293894e-05,"output_power":7.611983084915549e-05,"reference_active":true,"delta_db":0.22014744850032578}],"active_band_count":3,"active_band_mae_db":0.2789941343702597}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022872947944797367,"log_rms_mae_db":3.1384352217101577,"rms_envelope_nrmse":0.2550920495405172,"envelope_correlation":0.9310323828578416,"rms_gain_error_db":-0.45243110778427365,"reference_mean_power":0.008338532782066458,"output_mean_power":0.007513573336322005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.164744003117204e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015799393891106448,"output_power":0.0014640805617224535,"reference_active":true,"delta_db":-0.33075451885951285},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.374442374979845e-05,"output_power":6.759805517348187e-05,"reference_active":true,"delta_db":0.25492001824057653},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2933120282860937e-06,"output_power":3.875553574636475e-06,"reference_active":true,"delta_db":0.7070086449411546}],"active_band_count":3,"active_band_mae_db":0.43089439401374796}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00024243770094671305,"log_rms_mae_db":5.088549111941858,"rms_envelope_nrmse":0.38533993824720936,"envelope_correlation":0.8837258137931655,"rms_gain_error_db":-0.4612929681819632,"reference_mean_power":0.003959461759548422,"output_mean_power":0.003560466108041842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.015177817247118e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001136879031892713,"output_power":0.0010251217152416209,"reference_active":true,"delta_db":-0.4493882330374258},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001232393146611943,"output_power":0.0001037225621664312,"reference_active":true,"delta_db":-0.7487603830136473},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0364107060899584e-05,"output_power":3.33061125498721e-05,"reference_active":true,"delta_db":0.4016343127342602}],"active_band_count":3,"active_band_mae_db":0.5332609762617778}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00020477863430336455,"log_rms_mae_db":4.14555406919542,"rms_envelope_nrmse":0.30625256788432836,"envelope_correlation":0.9231942647570749,"rms_gain_error_db":-0.4613341149862722,"reference_mean_power":0.003979363820398493,"output_mean_power":0.0035783287320669474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.8942412803353574e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010558685473009428,"output_power":0.0009577156965597352,"reference_active":true,"delta_db":-0.4237324772418538},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.091449836890925e-05,"output_power":4.100341699130399e-05,"reference_active":true,"delta_db":-0.9402141941532933},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8755256559270582e-06,"output_power":2.3680686512912236e-06,"reference_active":true,"delta_db":1.012712792271147}],"active_band_count":3,"active_band_mae_db":0.792219821222098}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.508111273101461e-05,"log_rms_mae_db":4.234037621734513,"rms_envelope_nrmse":0.5230803583110951,"envelope_correlation":0.7969438807981045,"rms_gain_error_db":-0.01939392210195031,"reference_mean_power":0.00010911551831198982,"output_mean_power":0.00010862933671002678},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":6.836309277153808e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5256828161206523e-05,"output_power":2.5605144868652518e-05,"reference_active":true,"delta_db":0.05948427867880857},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.883426022529869e-06,"output_power":1.1417002718056789e-05,"reference_active":true,"delta_db":0.6264458852742052},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.436572483663533e-06,"output_power":5.416833596859344e-06,"reference_active":true,"delta_db":-0.015796882574652275}],"active_band_count":3,"active_band_mae_db":0.23390901550922202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.4434083199144075e-05,"log_rms_mae_db":3.3345201784982827,"rms_envelope_nrmse":0.378452837996786,"envelope_correlation":0.8855730901504794,"rms_gain_error_db":-0.018817441953329767,"reference_mean_power":0.00010963296530545586,"output_mean_power":0.00010915896687649662},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.313909407445063e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.3642314296732627e-05,"output_power":2.4399906420490756e-05,"reference_active":true,"delta_db":0.13698174248939235},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.770670858321524e-06,"output_power":7.575256115436467e-06,"reference_active":true,"delta_db":0.48765618677762645},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.6513846355224545e-07,"output_power":4.391821083724279e-07,"reference_active":true,"delta_db":-0.24937614687315254}],"active_band_count":3,"active_band_mae_db":0.2913380253800571}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010577651466639849,"log_rms_mae_db":4.579871203370751,"rms_envelope_nrmse":0.35641381685958823,"envelope_correlation":0.9276880707314809,"rms_gain_error_db":-0.04923692627328906,"reference_mean_power":0.00018020158938811213,"output_mean_power":0.000178170141347145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.705326923064481e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.613613486254777e-05,"output_power":1.8156527161147874e-05,"reference_active":true,"delta_db":0.5123326870319153},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.942687703713093e-06,"output_power":1.1048026195774511e-05,"reference_active":true,"delta_db":0.9181663072206258},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0438582736284589e-06,"output_power":2.8280709801087864e-06,"reference_active":true,"delta_db":4.32848767580638}],"active_band_count":3,"active_band_mae_db":1.9196622233529734}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.602886913996205e-05,"log_rms_mae_db":3.673370094721287,"rms_envelope_nrmse":0.26436690790338035,"envelope_correlation":0.9593512710415214,"rms_gain_error_db":-0.08617944750167335,"reference_mean_power":0.000177581626259849,"output_mean_power":0.0001740925089121884},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.4440347897892334e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7728340440048443e-05,"output_power":2.1192472165685998e-05,"reference_active":true,"delta_db":0.7751353847271656},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.955564320473049e-06,"output_power":6.4767295650940475e-06,"reference_active":true,"delta_db":1.1626264611811286},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.260796671426968e-08,"output_power":2.561488146268301e-07,"reference_active":true,"delta_db":4.418440014376614}],"active_band_count":3,"active_band_mae_db":2.1187339534283027}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"hnm","status":"measured","mos":{"status":"measured","value":2.8167290687561035,"predicted_mos":2.8167290687561035,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"17c5c1dbad9e26b6209b4c9711c8931396c4eec5868910672a1e4a7fbdcde149","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6294783579996874},"started_at":"2026-10-13T04:40:18.893374+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.797400000214111e-05,"native_input_samples":96000,"native_input_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","output_sha256":"15d125cdc8883ea04e29a7ad01ce2926d682240036fb14a5fc9d4c58cba7d960","output_samples":96000,"length_delta_samples":0,"analysis_s":1.074232036000467,"synthesis_s":0.49547665499994764,"total_s":1.5697086910004145,"rtf":0.2616181151667358,"output_peak":0.42138736108830266,"output_rms":0.06107612180905193,"finished_at":"2026-10-13T04:40:20.471151+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"15d125cdc8883ea04e29a7ad01ce2926d682240036fb14a5fc9d4c58cba7d960","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.865341007859446,"spectral_convergence":0.2511212384008162,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.605891678937489,"spectral_convergence":0.23130916788970907,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.59600539991141,"spectral_convergence":0.23815003832618,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0007480138296383804,"delta_db":0.02686214082777586,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.0023268593013818606,"delta_db":-0.22063653476607412,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.0003872502005489783,"delta_db":-0.036261920134713604,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":0.00011531615783683981,"delta_db":-0.20559324241014168,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":6.101756141736471e-05,"delta_db":0.10387371682343617,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":9.183560441072356e-05,"delta_db":0.5223727681762778,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18593338718973654,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13033083421521555,"pitch":null,"output_rms_dbfs":-24.282570948488235,"output_peak":0.42138736108830266,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.689079362236114},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.865341007859446,"spectral_convergence":0.2511212384008162,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.605891678937489,"spectral_convergence":0.23130916788970907,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.59600539991141,"spectral_convergence":0.23815003832618,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0007480138296383804,"delta_db":0.02686214082777586,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.0023268593013818606,"delta_db":-0.22063653476607412,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.0003872502005489783,"delta_db":-0.036261920134713604,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":0.00011531615783683981,"delta_db":-0.20559324241014168,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":6.101756141736471e-05,"delta_db":0.10387371682343617,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":9.183560441072356e-05,"delta_db":0.5223727681762778,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18593338718973654,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.13033083421521555,"pitch":null,"output_rms_dbfs":-24.282570948488235,"output_peak":0.42138736108830266,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.689079362236114},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.895827215077065,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06199947247277372,"rms_dbfs":-24.152240114273017,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.061093431123016,"mean":-4.431947072347005e-05,"zero_energy":false},"output":{"rms":0.06107612180905193,"rms_dbfs":-24.282570948488235,"peak":0.42138736108830266,"peak_dbfs":-7.5063698914982915,"crest_factor_db":16.77620105698994,"mean":3.250750542458059e-05,"zero_energy":false},"rms_difference_db":-0.13033083421521852},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":1.8580523624312156e-06,"output_rms_dbfs":-57.30942051163421,"output_zero_energy":false,"output_peak":0.023802438325311585,"output_energy_fraction":7.305441467374792e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":171,"both_estimated_voiced_frames":166,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9120879120879121,"absolute_difference_cents_median":1.6784988611515281,"absolute_difference_cents_p95":8.943396523404783,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019565573818943504,"log_rms_mae_db":1.0923800722549861,"rms_envelope_nrmse":0.07650110134622677,"envelope_correlation":0.9941274609590961,"rms_gain_error_db":-0.1303052772086,"reference_mean_power":0.0038470996951946897,"output_mean_power":0.0037333861602499526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.743468430119212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007829242029289011,"output_power":0.0007367206585559738,"reference_active":true,"delta_db":-0.2641687065697604},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011656588263872551,"output_power":0.00011114438861857587,"reference_active":true,"delta_db":-0.2068391528556836},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.15404555717803e-05,"output_power":2.9135853525673524e-05,"reference_active":true,"delta_db":-0.3444021687461254}],"active_band_count":3,"active_band_mae_db":0.2718033427238565}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018472557440126405,"log_rms_mae_db":0.6353917374411323,"rms_envelope_nrmse":0.04396086875288516,"envelope_correlation":0.9980690751421075,"rms_gain_error_db":-0.13008744785032242,"reference_mean_power":0.0038659686805529734,"output_mean_power":0.0037518855902637363},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.614770154461247e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106100759473788,"output_power":0.0006692731095440574,"reference_active":true,"delta_db":-0.2602798517483903},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7305685337334634e-05,"output_power":5.277943885774082e-05,"reference_active":true,"delta_db":-0.35732942204249274},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9649709160478004e-06,"output_power":2.2750623330056362e-06,"reference_active":true,"delta_db":0.6363717344412478}],"active_band_count":3,"active_band_mae_db":0.4179936694107103}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016965929826439602,"log_rms_mae_db":1.601845362513654,"rms_envelope_nrmse":0.09217693656285968,"envelope_correlation":0.9939267650494571,"rms_gain_error_db":-0.2194315232444896,"reference_mean_power":0.0021633278035333714,"output_mean_power":0.002056738982830617},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3458386409803657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007177587670080827,"output_power":0.000653784922029378,"reference_active":true,"delta_db":-0.40543605450526315},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.048912255781678e-05,"output_power":7.923215018972969e-05,"reference_active":true,"delta_db":-0.5769493512560867},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.109541579796594e-05,"output_power":2.2499042444400936e-05,"reference_active":true,"delta_db":-1.4053233343350258}],"active_band_count":3,"active_band_mae_db":0.7959029133654586}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015157137475553504,"log_rms_mae_db":0.9808505873315359,"rms_envelope_nrmse":0.05776171384334819,"envelope_correlation":0.9977666033500124,"rms_gain_error_db":-0.21945764729570066,"reference_mean_power":0.002174161455071234,"output_mean_power":0.002067026418254455},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2731603714689838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006867133707128568,"output_power":0.0006255121318617337,"reference_active":true,"delta_db":-0.4053976615524362},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067232594302272e-05,"output_power":3.4122868782942014e-05,"reference_active":true,"delta_db":-0.7625347316926339},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.085572616795342e-06,"output_power":1.697204063105428e-06,"reference_active":true,"delta_db":-0.89491253190385}],"active_band_count":3,"active_band_mae_db":0.6876149750496401}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":1.7557557368978516,"rms_envelope_nrmse":0.1655333801617481,"envelope_correlation":0.9793469343087413,"rms_gain_error_db":-0.09392776788109819,"reference_mean_power":0.00014077271984820847,"output_mean_power":0.0001377608217932233},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.622290255773313e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707421130487803e-05,"output_power":3.656779439814701e-05,"reference_active":true,"delta_db":-0.05973154397290496},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.087726823495623e-06,"output_power":4.875558343228016e-06,"reference_active":true,"delta_db":-0.18499426813631828},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8011808387081921e-06,"output_power":2.8988141829491204e-06,"reference_active":true,"delta_db":2.0666305930101805}],"active_band_count":3,"active_band_mae_db":0.7704521350398013}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":1.080632342068956,"rms_envelope_nrmse":0.0839598434931209,"envelope_correlation":0.9944831449326793,"rms_gain_error_db":-0.09433676100644747,"reference_mean_power":0.00014139647469838982,"output_mean_power":0.00013835820073807416},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.069193023932186e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.382591725297048e-05,"output_power":3.389116411672041e-05,"reference_active":true,"delta_db":0.008369043994173842},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1244897853414308e-06,"output_power":2.079082022262368e-06,"reference_active":true,"delta_db":-0.09383024215768587},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4002254483560297e-07,"output_power":1.280986606119464e-07,"reference_active":true,"delta_db":-0.38653377635427233}],"active_band_count":3,"active_band_mae_db":0.16291102083537734}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":1.893881559200816,"rms_envelope_nrmse":0.21283971059637494,"envelope_correlation":0.9738175367038503,"rms_gain_error_db":0.30856743727483366,"reference_mean_power":8.952225188868398e-05,"output_mean_power":9.611424124752701e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.314971416941346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.877588546016399e-05,"output_power":6.027378814420975e-05,"reference_active":true,"delta_db":0.10929306773883404},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.0495630695158e-06,"output_power":8.888156792279915e-06,"reference_active":true,"delta_db":-0.07815903921323403},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5294615410314874e-06,"output_power":3.1526700219940324e-06,"reference_active":true,"delta_db":0.9565043656769774}],"active_band_count":3,"active_band_mae_db":0.38131882420968183}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":1.37216579234825,"rms_envelope_nrmse":0.1532539466276058,"envelope_correlation":0.9862459167732355,"rms_gain_error_db":0.3085637509432118,"reference_mean_power":8.996899751083777e-05,"output_mean_power":9.65938010853517e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.951283845645361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.884212028117589e-05,"output_power":6.074298703027496e-05,"reference_active":true,"delta_db":0.13807831054131348},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.43607146326297e-06,"output_power":4.435736269587676e-06,"reference_active":true,"delta_db":-0.0003281693382047269},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7711408251558815e-07,"output_power":1.831978978568775e-07,"reference_active":true,"delta_db":0.14667392210264435}],"active_band_count":3,"active_band_mae_db":0.09502680066072085}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.28086256980896,"predicted_mos":3.28086256980896,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"332a260eb15c1711460babbb253a643f3d60128414b6c4a96c515f5e8986e097","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5505822459999763},"started_at":"2026-10-13T04:43:34.197910+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.6933000299322885e-05,"native_input_samples":96000,"native_input_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","output_sha256":"c8b698b2fa9d230cf474d273a0a5cab554e36a29e13876b68c4f8ae433c0fb2c","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5609584550002182,"synthesis_s":0.26823733000037464,"total_s":1.8291957850005929,"rtf":0.3048659641667655,"output_peak":0.46455676852500566,"output_rms":0.0732080374939261,"finished_at":"2026-10-13T04:43:36.036277+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"c8b698b2fa9d230cf474d273a0a5cab554e36a29e13876b68c4f8ae433c0fb2c","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.377063783256825,"spectral_convergence":0.557910975349616,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.348589426346837,"spectral_convergence":0.33304075586831877,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.474388476486447,"spectral_convergence":0.35031704105737943,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0007021194995700551,"delta_db":-0.24812378681946418,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.003755144009140253,"delta_db":1.8579281978475364,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.0005937127470622799,"delta_db":1.8195853000794093,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":0.0001499054780424891,"delta_db":0.9336801554560716,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":7.839875831650378e-05,"delta_db":1.1924170908146632,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":8.013626158050928e-05,"delta_db":-0.06944733414109883,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.020196977526374,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.443415409658113,"pitch":null,"output_rms_dbfs":-22.7088247046149,"output_peak":0.46455676852500566,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7333472286967035},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.377063783256825,"spectral_convergence":0.557910975349616,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.348589426346837,"spectral_convergence":0.33304075586831877,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.474388476486447,"spectral_convergence":0.35031704105737943,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0007021194995700551,"delta_db":-0.24812378681946418,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.003755144009140253,"delta_db":1.8579281978475364,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.0005937127470622799,"delta_db":1.8195853000794093,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":0.0001499054780424891,"delta_db":0.9336801554560716,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":7.839875831650378e-05,"delta_db":1.1924170908146632,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":8.013626158050928e-05,"delta_db":-0.06944733414109883,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.020196977526374,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.443415409658113,"pitch":null,"output_rms_dbfs":-22.7088247046149,"output_peak":0.46455676852500566,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.7333472286967035},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9237629648743559,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06199947247277372,"rms_dbfs":-24.152240114273017,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.061093431123016,"mean":-4.431947072347005e-05,"zero_energy":false},"output":{"rms":0.0732080374939261,"rms_dbfs":-22.7088247046149,"peak":0.46455676852500566,"peak_dbfs":-6.6592241586714795,"crest_factor_db":16.049600545943424,"mean":8.435822547056086e-08,"zero_energy":false},"rms_difference_db":1.4434154096581153},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":9.69611538025272e-07,"output_rms_dbfs":-60.13402225200382,"output_zero_energy":false,"output_peak":0.013321904550101854,"output_energy_fraction":2.6534546346896415e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":166,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9120879120879121,"absolute_difference_cents_median":7.001879316321276,"absolute_difference_cents_p95":19.865150985708837,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019565573818943504,"log_rms_mae_db":2.204700262155273,"rms_envelope_nrmse":0.23665546560795173,"envelope_correlation":0.9818564062873005,"rms_gain_error_db":1.4433728437531255,"reference_mean_power":0.0038470996951946897,"output_mean_power":0.005363777143464568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.743468430119212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007829242029289011,"output_power":0.0011599216500901943,"reference_active":true,"delta_db":1.7070893583732327},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011656588263872551,"output_power":0.0001797751748689984,"reference_active":true,"delta_db":1.8815826335481711},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.15404555717803e-05,"output_power":5.1500520607600905e-05,"reference_active":true,"delta_db":2.1294365724465916}],"active_band_count":3,"active_band_mae_db":1.906036188122665}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018472557440126405,"log_rms_mae_db":1.8313495215960705,"rms_envelope_nrmse":0.21101425688220596,"envelope_correlation":0.9907189922883362,"rms_gain_error_db":1.443709469521416,"reference_mean_power":0.0038659686805529734,"output_mean_power":0.005390502829021208},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.614770154461247e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106100759473788,"output_power":0.001057499697005697,"reference_active":true,"delta_db":1.7264889042824796},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7305685337334634e-05,"output_power":9.319198441806819e-05,"reference_active":true,"delta_db":2.111808488767735},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9649709160478004e-06,"output_power":3.067606039941258e-06,"reference_active":true,"delta_db":1.9344345752541328}],"active_band_count":3,"active_band_mae_db":1.9242439894347825}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016965929826439602,"log_rms_mae_db":2.1008896434807256,"rms_envelope_nrmse":0.31096559695444714,"envelope_correlation":0.985176143351553,"rms_gain_error_db":2.0536547772583895,"reference_mean_power":0.0021633278035333714,"output_mean_power":0.0034712653217319985},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3458386409803657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007177587670080827,"output_power":0.001141297194930415,"reference_active":true,"delta_db":2.0142024400567533},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.048912255781678e-05,"output_power":0.00014379318201164707,"reference_active":true,"delta_db":2.0114191732958684},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.109541579796594e-05,"output_power":4.7137210402657e-05,"reference_active":true,"delta_db":1.8066750880715843}],"active_band_count":3,"active_band_mae_db":1.944098900474735}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015157137475553504,"log_rms_mae_db":1.7237641811748095,"rms_envelope_nrmse":0.29126595174434683,"envelope_correlation":0.9916361378212012,"rms_gain_error_db":2.053617505363941,"reference_mean_power":0.002174161455071234,"output_mean_power":0.003488619006495379},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2731603714689838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006867133707128568,"output_power":0.0010874890099030338,"reference_active":true,"delta_db":1.9964937325026975},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067232594302272e-05,"output_power":6.414156759919118e-05,"reference_active":true,"delta_db":1.9784056046848781},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.085572616795342e-06,"output_power":2.833627457638089e-06,"reference_active":true,"delta_db":1.3311743616051366}],"active_band_count":3,"active_band_mae_db":1.7686912329309041}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":1.523535407495437,"rms_envelope_nrmse":0.18801519790196608,"envelope_correlation":0.9841806945830204,"rms_gain_error_db":0.896588091761986,"reference_mean_power":0.00014077271984820847,"output_mean_power":0.00017305227410947853},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.622290255773313e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707421130487803e-05,"output_power":4.4961998776863385e-05,"reference_active":true,"delta_db":0.837736885949312},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.087726823495623e-06,"output_power":5.753474434692941e-06,"reference_active":true,"delta_db":0.5340640307464393},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8011808387081921e-06,"output_power":1.8699813945416678e-06,"reference_active":true,"delta_db":0.16279967315636465}],"active_band_count":3,"active_band_mae_db":0.5115335299507052}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":1.055715698306489,"rms_envelope_nrmse":0.13224573094216155,"envelope_correlation":0.9959947818448657,"rms_gain_error_db":0.8969506296323186,"reference_mean_power":0.00014139647469838982,"output_mean_power":0.00017383356815826534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.069193023932186e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.382591725297048e-05,"output_power":4.126225427159771e-05,"reference_active":true,"delta_db":0.863033688635591},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1244897853414308e-06,"output_power":2.3450911930171443e-06,"reference_active":true,"delta_db":0.4290508836838085},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4002254483560297e-07,"output_power":1.1337251190678566e-07,"reference_active":true,"delta_db":-0.9169019739071897}],"active_band_count":3,"active_band_mae_db":0.7363288487421964}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":1.7622687513257513,"rms_envelope_nrmse":0.26703571899366896,"envelope_correlation":0.9604474165890645,"rms_gain_error_db":0.4813809639504201,"reference_mean_power":8.952225188868398e-05,"output_mean_power":0.00010001591086661256},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.314971416941346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.877588546016399e-05,"output_power":6.563641347410896e-05,"reference_active":true,"delta_db":0.47945661668108475},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.0495630695158e-06,"output_power":1.3783168639073094e-05,"reference_active":true,"delta_db":1.827214587108721},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5294615410314874e-06,"output_power":5.354052611691673e-06,"reference_active":true,"delta_db":3.256545538704712}],"active_band_count":3,"active_band_mae_db":1.854405580831506}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":1.1244893262941806,"rms_envelope_nrmse":0.1736886175642312,"envelope_correlation":0.983662605989547,"rms_gain_error_db":0.4813586330400861,"reference_mean_power":8.996899751083777e-05,"output_mean_power":0.00010051450647992782},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.951283845645361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.884212028117589e-05,"output_power":6.859759475784899e-05,"reference_active":true,"delta_db":0.666205748359458},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.43607146326297e-06,"output_power":6.891702654480643e-06,"reference_active":true,"delta_db":1.9132799751857734},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7711408251558815e-07,"output_power":3.8295734111028975e-07,"reference_active":true,"delta_db":3.3489730541257456}],"active_band_count":3,"active_band_mae_db":1.9761529258903259}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"facodec","status":"measured","mos":{"status":"measured","value":4.318931579589844,"predicted_mos":4.318931579589844,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f39feaaa501251fc2be64240884fadf1122a20f66aab2e0c9b3a998d85d1e81c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.879618181003025},"started_at":"2026-10-13T07:00:34.831563+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.085399996256456e-05,"native_input_samples":96000,"native_input_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","output_sha256":"d0593e480aa6d801222dcad4973567b07ee23b7896e0f47a0c0f456f73d85b0d","output_samples":96000,"length_delta_samples":0,"analysis_s":2.1931129999993573,"synthesis_s":6.741790088999551,"total_s":8.934903088998908,"rtf":1.4891505148331514,"output_peak":0.39301085472106934,"output_rms":0.057773413947922896,"finished_at":"2026-10-13T07:00:43.794967+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"d0593e480aa6d801222dcad4973567b07ee23b7896e0f47a0c0f456f73d85b0d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.48993569324207,"spectral_convergence":0.2452015517835771,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.890007682622654,"spectral_convergence":0.23965923786916368,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.794149338834072,"spectral_convergence":0.2525243798112341,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0007011470543730104,"delta_db":-0.2541429946377945,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.002065123466947017,"delta_db":-0.7388775518622421,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.00035590667407564607,"delta_db":-0.40281711247959173,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":9.522215929725245e-05,"delta_db":-1.037114626586417,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":5.295683327141724e-05,"delta_db":-0.5114546814294497,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":6.741117123370813e-05,"delta_db":-0.8204193896200002,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.627471059435916,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6131992529403079,"pitch":null,"output_rms_dbfs":-24.765439367213325,"output_peak":0.39301085472106934,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.7246975715662645},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.48993569324207,"spectral_convergence":0.2452015517835771,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.890007682622654,"spectral_convergence":0.23965923786916368,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.794149338834072,"spectral_convergence":0.2525243798112341,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0007011470543730104,"delta_db":-0.2541429946377945,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.002065123466947017,"delta_db":-0.7388775518622421,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.00035590667407564607,"delta_db":-0.40281711247959173,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":9.522215929725245e-05,"delta_db":-1.037114626586417,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":5.295683327141724e-05,"delta_db":-0.5114546814294497,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":6.741117123370813e-05,"delta_db":-0.8204193896200002,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.627471059435916,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6131992529403079,"pitch":null,"output_rms_dbfs":-24.765439367213325,"output_peak":0.39301085472106934,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.7246975715662645},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8894521725361749,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06199947247277372,"rms_dbfs":-24.152240114273017,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.061093431123016,"mean":-4.431947072347005e-05,"zero_energy":false},"output":{"rms":0.057773413947922896,"rms_dbfs":-24.765439367213325,"peak":0.39301085472106934,"peak_dbfs":-8.111909090182992,"crest_factor_db":16.65353027703033,"mean":-1.5892689311461558e-05,"zero_energy":false},"rms_difference_db":-0.6131992529403085},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":1.437591726492988e-06,"output_rms_dbfs":-58.423644353069704,"output_zero_energy":false,"output_peak":0.005238670855760574,"output_energy_fraction":6.317000673257348e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":191,"both_estimated_voiced_frames":174,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9560439560439561,"absolute_difference_cents_median":3.3503765063922164,"absolute_difference_cents_p95":11.97907125032836,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019565573818943504,"log_rms_mae_db":1.6185503924367057,"rms_envelope_nrmse":0.12227293407009636,"envelope_correlation":0.9886344162772425,"rms_gain_error_db":-0.6131686415578388,"reference_mean_power":0.0038470996951946897,"output_mean_power":0.003340539233554519},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.743468430119212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007829242029289011,"output_power":0.0006702225453332545,"reference_active":true,"delta_db":-0.6750068604185897},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011656588263872551,"output_power":9.493342535009635e-05,"reference_active":true,"delta_db":-0.8915230531614826},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.15404555717803e-05,"output_power":2.6336698852763477e-05,"reference_active":true,"delta_db":-0.7830662417886207}],"active_band_count":3,"active_band_mae_db":0.783198718456231}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018472557440126405,"log_rms_mae_db":1.4295879924280634,"rms_envelope_nrmse":0.11177331775136451,"envelope_correlation":0.9906204643220735,"rms_gain_error_db":-0.6130456556990515,"reference_mean_power":0.0038659686805529734,"output_mean_power":0.00335701874113569},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.614770154461247e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106100759473788,"output_power":0.0006062420597983062,"reference_active":true,"delta_db":-0.6898529767095345},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7305685337334634e-05,"output_power":4.677081054144504e-05,"reference_active":true,"delta_db":-0.8822281443804807},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9649709160478004e-06,"output_power":1.698674808120865e-06,"reference_active":true,"delta_db":-0.6324588056321179}],"active_band_count":3,"active_band_mae_db":0.7348466422407111}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016965929826439602,"log_rms_mae_db":1.7335232087524992,"rms_envelope_nrmse":0.1510592495359756,"envelope_correlation":0.9857526861870285,"rms_gain_error_db":-0.65252311490461,"reference_mean_power":0.0021633278035333714,"output_mean_power":0.0018615299179794119},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3458386409803657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007177587670080827,"output_power":0.0006008603861092199,"reference_active":true,"delta_db":-0.7720493333177513},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.048912255781678e-05,"output_power":7.084537118331457e-05,"reference_active":true,"delta_db":-1.0628489697120755},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.109541579796594e-05,"output_power":2.387828993210319e-05,"reference_active":true,"delta_db":-1.146931473032414}],"active_band_count":3,"active_band_mae_db":0.9939432586874136}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015157137475553504,"log_rms_mae_db":1.314843920078494,"rms_envelope_nrmse":0.1381007473515219,"envelope_correlation":0.9880686914357757,"rms_gain_error_db":-0.6525124511191776,"reference_mean_power":0.002174161455071234,"output_mean_power":0.0018708568004828304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2731603714689838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006867133707128568,"output_power":0.0005722367882463428,"reference_active":true,"delta_db":-0.7919972893653755},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067232594302272e-05,"output_power":3.182930298016084e-05,"reference_active":true,"delta_db":-1.0647188102284415},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.085572616795342e-06,"output_power":1.7548246386007373e-06,"reference_active":true,"delta_db":-0.7499159250314126}],"active_band_count":3,"active_band_mae_db":0.8688773415417432}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":1.677158923636663,"rms_envelope_nrmse":0.19592800572171568,"envelope_correlation":0.9750772677326149,"rms_gain_error_db":-0.816412679358503,"reference_mean_power":0.00014077271984820847,"output_mean_power":0.00011664798300876268},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.622290255773313e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707421130487803e-05,"output_power":3.2439736394825596e-05,"reference_active":true,"delta_db":-0.5799460439333549},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.087726823495623e-06,"output_power":4.956826870994618e-06,"reference_active":true,"delta_db":-0.11320034192285841},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8011808387081921e-06,"output_power":2.495954053240391e-06,"reference_active":true,"delta_db":1.416792681646497}],"active_band_count":3,"active_band_mae_db":0.7033130225009034}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":1.330011275829172,"rms_envelope_nrmse":0.15954834048119068,"envelope_correlation":0.984895960012624,"rms_gain_error_db":-0.818535066488092,"reference_mean_power":0.00014139647469838982,"output_mean_power":0.00011710759850455842},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.069193023932186e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.382591725297048e-05,"output_power":2.994303236926583e-05,"reference_active":true,"delta_db":-0.5295380234292248},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1244897853414308e-06,"output_power":2.1817024369351195e-06,"reference_active":true,"delta_db":0.11540869413883242},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4002254483560297e-07,"output_power":1.3723741241809246e-07,"reference_active":true,"delta_db":-0.08725445527640013}],"active_band_count":3,"active_band_mae_db":0.2440670576148191}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":1.8346196267472958,"rms_envelope_nrmse":0.20789532308155206,"envelope_correlation":0.9806281155892888,"rms_gain_error_db":-1.1618651667926483,"reference_mean_power":8.952225188868398e-05,"output_mean_power":6.850850358708256e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.314971416941346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.877588546016399e-05,"output_power":4.65438496168206e-05,"reference_active":true,"delta_db":-1.0133687980357275},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.0495630695158e-06,"output_power":6.726317884738913e-06,"reference_active":true,"delta_db":-1.2885022297674953},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5294615410314874e-06,"output_power":1.1927700777502127e-06,"reference_active":true,"delta_db":-3.264713448863021}],"active_band_count":3,"active_band_mae_db":1.855528158888748}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":1.5343299850888084,"rms_envelope_nrmse":0.1703477945401044,"envelope_correlation":0.9902771782186943,"rms_gain_error_db":-1.1619136954372011,"reference_mean_power":8.996899751083777e-05,"output_mean_power":6.884961432153757e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.951283845645361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.884212028117589e-05,"output_power":4.604908904522728e-05,"reference_active":true,"delta_db":-1.0646727014655493},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.43607146326297e-06,"output_power":3.062862371572799e-06,"reference_active":true,"delta_db":-1.608710514307641},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7711408251558815e-07,"output_power":9.106292564204455e-08,"reference_active":true,"delta_db":-2.889114946683893}],"active_band_count":3,"active_band_mae_db":1.854166054152361}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.6684489250183105,"predicted_mos":3.6684489250183105,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"68e7a8b68146f21c2b8099ec132e75c9a69588d1b073d9666d1cc023e69c6ff3","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3233833749982296},"started_at":"2026-10-13T07:30:31.028580+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00648591299977852,"native_input_samples":264600,"native_input_sha256":"ca6962a73ec6531952b15a05ce20f0740de90c5c86588e85bdad1e7828ede58e","output_sha256":"d41ff0d34aafc1e277d3f48cf2a557a5dbfc67e104ee9e28d9c19962c3193c7e","output_samples":264704,"length_delta_samples":104,"analysis_s":0.14934363399879658,"synthesis_s":3.8517165910016047,"total_s":4.001060225000401,"rtf":0.6668433708334002,"output_peak":0.37986424565315247,"output_rms":0.056520584398810504,"finished_at":"2026-10-13T07:30:35.047916+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"d41ff0d34aafc1e277d3f48cf2a557a5dbfc67e104ee9e28d9c19962c3193c7e","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.95185806633092,"output_peak":0.38024906770348676},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.605132228912867,"spectral_convergence":0.4408701467715247,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.407156781832912,"spectral_convergence":0.3419441773440098,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.10527170002735,"spectral_convergence":0.40215889382646497,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0005458773434835792,"delta_db":-1.3412834391536856,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.0020998500548580143,"delta_db":-0.6664549305765612,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.00037950026562575816,"delta_db":-0.12405759147236953,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":6.71047921927042e-05,"delta_db":-2.556959523555312,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":4.769289673548241e-05,"delta_db":-0.9661377419118878,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":5.868503062757604e-05,"delta_db":-1.4224647622893738,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1795596648265316,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7980106844223891,"pitch":null,"output_rms_dbfs":-24.950250798695404,"output_peak":0.38024906770348676,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.372520236924375},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7434492228003156,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06199947247277372,"rms_dbfs":-24.152240114273017,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.061093431123016,"mean":-4.431947072347005e-05,"zero_energy":false},"output":{"rms":0.05655714263930874,"rms_dbfs":-24.950250798695404,"peak":0.38024906770348676,"peak_dbfs":-8.398636841579624,"crest_factor_db":16.55161395711578,"mean":-0.00011481494753541934,"zero_energy":false},"rms_difference_db":-0.7980106844223869},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":1.0889837361153828e-06,"output_rms_dbfs":-59.629786063489895,"output_zero_energy":false,"output_peak":0.006670942400968733,"output_energy_fraction":4.993187737563799e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":191,"both_estimated_voiced_frames":156,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.8571428571428571,"absolute_difference_cents_median":12.450144985831754,"absolute_difference_cents_p95":43.49063470501675,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019565573818943504,"log_rms_mae_db":1.6168156649391887,"rms_envelope_nrmse":0.12832971431255605,"envelope_correlation":0.9904687348781083,"rms_gain_error_db":-0.7980112123980697,"reference_mean_power":0.0038470996951946897,"output_mean_power":0.0032013438229090765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.743468430119212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007829242029289011,"output_power":0.0006579385171839187,"reference_active":true,"delta_db":-0.7553440712938785},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011656588263872551,"output_power":0.00010122926625111426,"reference_active":true,"delta_db":-0.6126536758755828},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.15404555717803e-05,"output_power":2.6532812161194118e-05,"reference_active":true,"delta_db":-0.750846795576145}],"active_band_count":3,"active_band_mae_db":0.7062815142485355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018472557440126405,"log_rms_mae_db":1.3707241413614573,"rms_envelope_nrmse":0.10563092957679038,"envelope_correlation":0.9960160054606322,"rms_gain_error_db":-0.7977807052366543,"reference_mean_power":0.0038659686805529734,"output_mean_power":0.00321721630325433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.614770154461247e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106100759473788,"output_power":0.0005974634509485083,"reference_active":true,"delta_db":-0.7532001815374691},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7305685337334634e-05,"output_power":4.8983511252738334e-05,"reference_active":true,"delta_db":-0.6814779761625799},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9649709160478004e-06,"output_power":1.650304018033358e-06,"reference_active":true,"delta_db":-0.7579216961654285}],"active_band_count":3,"active_band_mae_db":0.7308666179551592}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016965929826439602,"log_rms_mae_db":2.5499583364816596,"rms_envelope_nrmse":0.16371027483222483,"envelope_correlation":0.981562318416841,"rms_gain_error_db":-0.557415532697851,"reference_mean_power":0.0021633278035333714,"output_mean_power":0.0019027458293897075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3458386409803657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007177587670080827,"output_power":0.0005753218470057529,"reference_active":true,"delta_db":-0.9606763982033595},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.048912255781678e-05,"output_power":8.632102004624764e-05,"reference_active":true,"delta_db":-0.20479813311458078},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.109541579796594e-05,"output_power":2.6216518711094962e-05,"reference_active":true,"delta_db":-0.7412134717365706}],"active_band_count":3,"active_band_mae_db":0.6355626676848369}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015157137475553504,"log_rms_mae_db":1.9117729816330364,"rms_envelope_nrmse":0.13371711846984075,"envelope_correlation":0.987981877863491,"rms_gain_error_db":-0.5575546888570546,"reference_mean_power":0.002174161455071234,"output_mean_power":0.0019122132499191007},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2731603714689838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006867133707128568,"output_power":0.0005504863952372869,"reference_active":true,"delta_db":-0.9602891319180773},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067232594302272e-05,"output_power":3.9962988344312676e-05,"reference_active":true,"delta_db":-0.07641052916739417},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.085572616795342e-06,"output_power":1.8623950599958174e-06,"reference_active":true,"delta_db":-0.4915350498200711}],"active_band_count":3,"active_band_mae_db":0.5094115703018476}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":3.6661218314221196,"rms_envelope_nrmse":0.34442361768674085,"envelope_correlation":0.9385799135749469,"rms_gain_error_db":-2.35569415511483,"reference_mean_power":0.00014077271984820847,"output_mean_power":8.183688422295841e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.622290255773313e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707421130487803e-05,"output_power":2.219735357388896e-05,"reference_active":true,"delta_db":-2.227707210499927},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.087726823495623e-06,"output_power":3.99991731796887e-06,"reference_active":true,"delta_db":-1.0447277044362044},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8011808387081921e-06,"output_power":1.4536642265409116e-06,"reference_active":true,"delta_db":-0.930932152650224}],"active_band_count":3,"active_band_mae_db":1.4011223558621184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":3.2045427810382683,"rms_envelope_nrmse":0.31158134060675635,"envelope_correlation":0.9575319561188262,"rms_gain_error_db":-2.3534237831755846,"reference_mean_power":0.00014139647469838982,"output_mean_power":8.224248105122569e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.069193023932186e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.382591725297048e-05,"output_power":2.0423331679325224e-05,"reference_active":true,"delta_db":-2.1912299169552427},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1244897853414308e-06,"output_power":1.6133079506468576e-06,"reference_active":true,"delta_db":-1.1953737323439542},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4002254483560297e-07,"output_power":1.1791889781905104e-07,"reference_active":true,"delta_db":-0.7461455525071536}],"active_band_count":3,"active_band_mae_db":1.3775830672687832}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":3.8054705889624345,"rms_envelope_nrmse":0.20519575410479748,"envelope_correlation":0.9847722087489382,"rms_gain_error_db":-1.3390682275983972,"reference_mean_power":8.952225188868398e-05,"output_mean_power":6.576944474458954e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.314971416941346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.877588546016399e-05,"output_power":4.362696908580053e-05,"reference_active":true,"delta_db":-1.2944413838599416},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.0495630695158e-06,"output_power":6.027111395099608e-06,"reference_active":true,"delta_db":-1.7651839281141624},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5294615410314874e-06,"output_power":1.5377656141115126e-06,"reference_active":true,"delta_db":-2.161379351612152}],"active_band_count":3,"active_band_mae_db":1.7403348878620852}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":3.4685002785507715,"rms_envelope_nrmse":0.16435698253120312,"envelope_correlation":0.9953220061896647,"rms_gain_error_db":-1.3389225931978608,"reference_mean_power":8.996899751083777e-05,"output_mean_power":6.609987253616451e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.951283845645361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.884212028117589e-05,"output_power":4.32760690065772e-05,"reference_active":true,"delta_db":-1.334405088072057},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.43607146326297e-06,"output_power":2.817504366172912e-06,"reference_active":true,"delta_db":-1.9713393610931698},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7711408251558815e-07,"output_power":1.1898647894490503e-07,"reference_active":true,"delta_db":-1.7275548070010331}],"active_band_count":3,"active_band_mae_db":1.6777664187220864}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"golf","status":"measured","mos":{"status":"measured","value":3.690859079360962,"predicted_mos":3.690859079360962,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"b91cd9747248bf4b15632c1ffcc4ef44c5ee54e7812394b0629b1e396a033004","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.53896185099984},"started_at":"2026-10-13T05:03:04.249927+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003004999000040698,"native_input_samples":144000,"native_input_sha256":"cf968e2cc5e461341a924a8754600d2f33bad7b3650ba829ac87c95215370b1e","output_sha256":"5c0d74f0223e0fab6eee34eb20b994ffa0a387cff5de4684e99d1bc40c93adcf","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.559571261000201,"synthesis_s":0.12701070200000686,"total_s":0.6865819630002079,"rtf":0.11443032716670132,"output_peak":0.41180849075317383,"output_rms":0.05818816834267773,"finished_at":"2026-10-13T05:03:04.949138+00:00","analysis_stages":{"external_f0_s":0.08729209199918841},"evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"5c0d74f0223e0fab6eee34eb20b994ffa0a387cff5de4684e99d1bc40c93adcf","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.700491434456367,"output_peak":0.411948562061518},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.297467940892668,"spectral_convergence":0.515718685994018,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.334520541615819,"spectral_convergence":0.2418004990524547,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.361369325915818,"spectral_convergence":0.2646279018699348,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007445918498616591,"output_power":0.0006647060805269374,"delta_db":-0.49288626631915655,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002452216810959737,"output_power":0.002170816699186084,"delta_db":-0.5293571160728268,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039111894396448067,"output_power":0.00034115211513586016,"delta_db":-0.5936078340547827,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0001210305070879805,"output_power":9.591997284204457e-05,"delta_db":-1.0098580575037184,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.967081528244016e-05,"output_power":4.311109115945722e-05,"delta_db":-1.4117295648575776,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.156304695861827e-05,"output_power":7.235219847014898e-05,"delta_db":-0.5204170940493679,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7596426554762382,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5553152781652865,"pitch":null,"output_rms_dbfs":-24.700491434456367,"output_peak":0.411948562061518,"reference_rms_dbfs":-24.145176156291082,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.664452602808101},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8973513267903702,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06204991518217342,"rms_dbfs":-24.14517615629108,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.05402947314108,"mean":-5.0277264169938176e-05,"zero_energy":false},"output":{"rms":0.05820702841857272,"rms_dbfs":-24.700491434456367,"peak":0.411948562061518,"peak_dbfs":-7.703140174845928,"crest_factor_db":16.99735125961044,"mean":0.0006830512686507999,"zero_energy":false},"rms_difference_db":-0.5553152781652884},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":2.3138421439266323e-06,"output_rms_dbfs":-56.35666273024752,"output_zero_energy":false,"output_peak":0.012983764406708386,"output_energy_fraction":0.00010033182689490556,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":206,"both_estimated_voiced_frames":172,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.945054945054945,"absolute_difference_cents_median":4.62599007408198,"absolute_difference_cents_p95":17.04508507549913,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00019565573818943504,"log_rms_mae_db":2.0899103428291967,"rms_envelope_nrmse":0.13119075716111384,"envelope_correlation":0.9857261847685017,"rms_gain_error_db":-0.5555894894443364,"reference_mean_power":0.0038533409468832743,"output_mean_power":0.003390615074320168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.74581018024284e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007822976777930797,"output_power":0.0007221190519294348,"reference_active":true,"delta_db":-0.34763237774021183},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011644807695954092,"output_power":0.00011627668049969698,"reference_active":true,"delta_db":-0.0063969601686325725},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.136430779641612e-05,"output_power":2.740594948483481e-05,"reference_active":true,"delta_db":-0.585908543007738}],"active_band_count":3,"active_band_mae_db":0.31331262697219414}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00018472557440126405,"log_rms_mae_db":1.5898164350859691,"rms_envelope_nrmse":0.09314440880557871,"envelope_correlation":0.994492182118945,"rms_gain_error_db":-0.5568487918077272,"reference_mean_power":0.003871939242603398,"output_mean_power":0.003405992244095551},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.6172269455918545e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007098939559829144,"output_power":0.0006615156244621708,"reference_active":true,"delta_db":-0.30653372291097153},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.70683305842977e-05,"output_power":5.798250124556859e-05,"reference_active":true,"delta_db":0.06901777641648266},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.947017281156079e-06,"output_power":1.6390513521666557e-06,"reference_active":true,"delta_db":-0.7477724579595728}],"active_band_count":3,"active_band_mae_db":0.3744413190956757}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00016965929826439602,"log_rms_mae_db":2.1233320113593823,"rms_envelope_nrmse":0.14450935576050433,"envelope_correlation":0.986312134366767,"rms_gain_error_db":-0.5444831097586689,"reference_mean_power":0.002166915479915282,"output_mean_power":0.0019115852073520714},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.347332682319234e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007175355561455209,"output_power":0.0006422892278763715,"reference_active":true,"delta_db":-0.48112788272272344},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.030553325097111e-05,"output_power":9.492243522213074e-05,"reference_active":true,"delta_db":0.2165450999157253},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.101780029946935e-05,"output_power":1.897102552433416e-05,"reference_active":true,"delta_db":-2.13520187197028}],"active_band_count":3,"active_band_mae_db":0.9442916182029095}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015157137475553504,"log_rms_mae_db":1.574485300209343,"rms_envelope_nrmse":0.10510483522633632,"envelope_correlation":0.9937701668916873,"rms_gain_error_db":-0.5445511600725579,"reference_mean_power":0.0021777836057724985,"output_mean_power":0.0019211426258567254},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2746901935567971e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006862343694890643,"output_power":0.0006160893567283018,"reference_active":true,"delta_db":-0.46828759360231237},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.052701942479925e-05,"output_power":4.1827294599382925e-05,"reference_active":true,"delta_db":0.13715110643846667},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.060339393233507e-06,"output_power":1.481998190192664e-06,"reference_active":true,"delta_db":-1.4309109294147837}],"active_band_count":3,"active_band_mae_db":0.6787832098185209}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":1.8719079825804705,"rms_envelope_nrmse":0.20506397300762275,"envelope_correlation":0.975778865434963,"rms_gain_error_db":-1.0630496683171429,"reference_mean_power":0.00014094495381468524,"output_mean_power":0.00011034294345731708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.638608472209241e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.704025898138967e-05,"output_power":2.7865220984467323e-05,"reference_active":true,"delta_db":-1.2361152299158602},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.054160782631721e-06,"output_power":3.6165840742076864e-06,"reference_active":true,"delta_db":-1.4535048801967219},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8004938063931302e-06,"output_power":1.304487345024104e-06,"reference_active":true,"delta_db":-1.3995176144109305}],"active_band_count":3,"active_band_mae_db":1.363045908174504}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":1.4079010360877997,"rms_envelope_nrmse":0.14594695569195212,"envelope_correlation":0.9928679466813707,"rms_gain_error_db":-1.063415524163457,"reference_mean_power":0.00014158770337118802,"output_mean_power":0.00011083680159560034},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.085506421475651e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.3775972288692986e-05,"output_power":2.553566191921523e-05,"reference_active":true,"delta_db":-1.2146073980597056},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.098707206251446e-06,"output_power":1.5743971271204953e-06,"reference_active":true,"delta_db":-1.248375651329014},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3987340323803757e-07,"output_power":7.507193238129683e-08,"reference_active":true,"delta_db":-2.7025754683839365}],"active_band_count":3,"active_band_mae_db":1.7218528392575518}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":2.2203431003615055,"rms_envelope_nrmse":0.20353951794799655,"envelope_correlation":0.976302099789465,"rms_gain_error_db":-0.5611617843205383,"reference_mean_power":8.967000833340151e-05,"output_mean_power":7.880087351078339e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.312930899464215e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8743196477054884e-05,"output_power":5.2396971808589375e-05,"reference_active":true,"delta_db":-0.49651386294341504},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.056455640222234e-06,"output_power":9.236401685222862e-06,"reference_active":true,"delta_db":0.08544547620226757},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.518651438531268e-06,"output_power":3.432459906756228e-06,"reference_active":true,"delta_db":1.3443740438159966}],"active_band_count":3,"active_band_mae_db":0.6421111276538931}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":1.8840891453288697,"rms_envelope_nrmse":0.14463687433884453,"envelope_correlation":0.9891049640789714,"rms_gain_error_db":-0.5611207388857199,"reference_mean_power":9.011745456535459e-05,"output_mean_power":7.919483209337156e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.949538450922873e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8806866696317456e-05,"output_power":5.3325232504973196e-05,"reference_active":true,"delta_db":-0.4249528255705757},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.4323405512174886e-06,"output_power":4.854613173599731e-06,"reference_active":true,"delta_db":0.3952150880272176},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7708815116012427e-07,"output_power":2.703247138514576e-07,"reference_active":true,"delta_db":1.8369624811306178}],"active_band_count":3,"active_band_mae_db":0.8857101315761371}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4511935710906982,"predicted_mos":3.4511935710906982,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"4202c1be4b36d46ba10dd09428bfc8b34ff85cc3473f628a38bcc4302cfb52d3","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6144172240019543},"started_at":"2026-10-13T07:45:50.201964+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026492829965718556,"native_input_samples":144000,"native_input_sha256":"cf968e2cc5e461341a924a8754600d2f33bad7b3650ba829ac87c95215370b1e","output_sha256":"1db2c6a63ab28bd026a05e186060501738dfc68ff4e74ed3d825d526010aa108","output_samples":144120,"length_delta_samples":120,"analysis_s":2.1097061529981147,"synthesis_s":3.525509385002806,"total_s":5.635215538000921,"rtf":0.9392025896668201,"output_peak":0.35812127590179443,"output_rms":0.06070853734720009,"finished_at":"2026-10-13T07:45:55.849228+00:00","analysis_stages":{"external_f0_s":1.3535010170016903},"evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"1db2c6a63ab28bd026a05e186060501738dfc68ff4e74ed3d825d526010aa108","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.332237373990502,"output_peak":0.35842639178124697},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.006191051573966,"spectral_convergence":0.4336615668592461,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.465443477427022,"spectral_convergence":0.2846331096688277,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.161917415306212,"spectral_convergence":0.27149224362575225,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0008166787258495103,"delta_db":0.40827829239704433,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.0022854545968222905,"delta_db":-0.2986117904214125,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.00036284779404847066,"delta_db":-0.31893356246722604,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":0.0001119685529843251,"delta_db":-0.3335342247237125,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":5.1153918653334504e-05,"delta_db":-0.6618856676537137,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":6.28364993832543e-05,"delta_db":-1.1256182927665808,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5244769717382817,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.17638992896027536,"pitch":null,"output_rms_dbfs":-24.32863004323329,"output_peak":0.35842639178124697,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.544517314769067},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8981391032724007,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06199947247277372,"rms_dbfs":-24.152240114273017,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.061093431123016,"mean":-4.431947072347005e-05,"zero_energy":false},"output":{"rms":0.06075310763854953,"rms_dbfs":-24.32863004323329,"peak":0.35842639178124697,"peak_dbfs":-8.912000394094848,"crest_factor_db":15.416629649138443,"mean":0.0001562796042865338,"zero_energy":false},"rms_difference_db":-0.1763899289602726},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":3.734697442466304e-06,"output_rms_dbfs":-54.27744575746758,"output_zero_energy":false,"output_peak":0.0883491384725757,"output_energy_fraction":0.00014840545006794686,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":194,"both_estimated_voiced_frames":166,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9120879120879121,"absolute_difference_cents_median":6.550422873036547,"absolute_difference_cents_p95":28.03650087559135,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019565573818943504,"log_rms_mae_db":2.6470728933809133,"rms_envelope_nrmse":0.14298950289278603,"envelope_correlation":0.9791322170716728,"rms_gain_error_db":-0.17634945651814177,"reference_mean_power":0.0038470996951946897,"output_mean_power":0.0036940136446749084},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.743468430119212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007829242029289011,"output_power":0.0007432506899633162,"reference_active":true,"delta_db":-0.22584397800669143},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011656588263872551,"output_power":0.00011406765939320516,"reference_active":true,"delta_db":-0.09408926362691197},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.15404555717803e-05,"output_power":2.990252817830526e-05,"reference_active":true,"delta_db":-0.23160053700790117}],"active_band_count":3,"active_band_mae_db":0.1838445928805015}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018472557440126405,"log_rms_mae_db":2.0455434572217275,"rms_envelope_nrmse":0.08792256072498542,"envelope_correlation":0.9917164784732785,"rms_gain_error_db":-0.17592215811945072,"reference_mean_power":0.0038659686805529734,"output_mean_power":0.0037124970355306903},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.614770154461247e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106100759473788,"output_power":0.0006772041470131391,"reference_active":true,"delta_db":-0.209117522392346},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7305685337334634e-05,"output_power":5.7709534059464423e-05,"reference_active":true,"delta_db":0.03049857110380961},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9649709160478004e-06,"output_power":2.0449666974399846e-06,"reference_active":true,"delta_db":0.17330113182332718}],"active_band_count":3,"active_band_mae_db":0.13763907510649426}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016965929826439602,"log_rms_mae_db":2.308245626697696,"rms_envelope_nrmse":0.17619328373358287,"envelope_correlation":0.9767187981130824,"rms_gain_error_db":-0.3077914173123691,"reference_mean_power":0.0021633278035333714,"output_mean_power":0.002015316173411513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3458386409803657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007177587670080827,"output_power":0.0006550570854234863,"reference_active":true,"delta_db":-0.39699357309067207},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.048912255781678e-05,"output_power":8.459034323808028e-05,"reference_active":true,"delta_db":-0.29275589849392186},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.109541579796594e-05,"output_power":2.9141691671397787e-05,"reference_active":true,"delta_db":-0.28181609531708324}],"active_band_count":3,"active_band_mae_db":0.3238551889672257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015157137475553504,"log_rms_mae_db":1.625630545277154,"rms_envelope_nrmse":0.11170029202251838,"envelope_correlation":0.9905979713878864,"rms_gain_error_db":-0.30776331901390835,"reference_mean_power":0.002174161455071234,"output_mean_power":0.0020254217070051995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2731603714689838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006867133707128568,"output_power":0.000626329823719787,"reference_active":true,"delta_db":-0.39972411608974806},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067232594302272e-05,"output_power":3.944730293435391e-05,"reference_active":true,"delta_db":-0.13281694013222506},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.085572616795342e-06,"output_power":2.2105497024311655e-06,"reference_active":true,"delta_db":0.2527496812669082}],"active_band_count":3,"active_band_mae_db":0.26176357916296045}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":2.0658324398953254,"rms_envelope_nrmse":0.1808657322199462,"envelope_correlation":0.9753628986528229,"rms_gain_error_db":-0.21409751538971344,"reference_mean_power":0.00014077271984820847,"output_mean_power":0.00013400121929626482},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.622290255773313e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707421130487803e-05,"output_power":3.58777395234909e-05,"reference_active":true,"delta_db":-0.1424684835562249},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.087726823495623e-06,"output_power":5.087018943953349e-06,"reference_active":true,"delta_db":-0.0006042965329735585},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8011808387081921e-06,"output_power":1.988423666756814e-06,"reference_active":true,"delta_db":0.42951605377033486}],"active_band_count":3,"active_band_mae_db":0.19086294461984443}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":1.2174200507658153,"rms_envelope_nrmse":0.09662744828173156,"envelope_correlation":0.9929205514400132,"rms_gain_error_db":-0.21273042929885183,"reference_mean_power":0.00014139647469838982,"output_mean_power":0.0001346373449690666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.069193023932186e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.382591725297048e-05,"output_power":3.282045732713463e-05,"reference_active":true,"delta_db":-0.13104953826339477},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1244897853414308e-06,"output_power":2.1807596045248334e-06,"reference_active":true,"delta_db":0.11353146556638495},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4002254483560297e-07,"output_power":1.3425912486524454e-07,"reference_active":true,"delta_db":-0.1825415444245444}],"active_band_count":3,"active_band_mae_db":0.14237418275144137}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":2.1484378413693164,"rms_envelope_nrmse":0.26028736564493354,"envelope_correlation":0.9622349857544163,"rms_gain_error_db":-1.080841925864395,"reference_mean_power":8.952225188868398e-05,"output_mean_power":6.979861507196387e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.314971416941346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.877588546016399e-05,"output_power":4.238329547794763e-05,"reference_active":true,"delta_db":-1.4200445860885607},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.0495630695158e-06,"output_power":6.9271216554069015e-06,"reference_active":true,"delta_db":-1.160747962653827},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5294615410314874e-06,"output_power":1.352185685051201e-06,"reference_active":true,"delta_db":-2.7199174664908874}],"active_band_count":3,"active_band_mae_db":1.7669033384110915}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":1.4511993449195648,"rms_envelope_nrmse":0.1884283711146582,"envelope_correlation":0.984308168933855,"rms_gain_error_db":-1.08082670542789,"reference_mean_power":8.996899751083777e-05,"output_mean_power":7.014717906798293e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.951283845645361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.884212028117589e-05,"output_power":4.212027021865926e-05,"reference_active":true,"delta_db":-1.4519716474260949},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.43607146326297e-06,"output_power":3.4591700817716598e-06,"reference_active":true,"delta_db":-1.0802661774335878},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7711408251558815e-07,"output_power":1.0698210377844349e-07,"reference_active":true,"delta_db":-2.1894195980205944}],"active_band_count":3,"active_band_mae_db":1.573885807626759}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.56746506690979,"predicted_mos":3.56746506690979,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"c0d83953dfbbf191d876519f34fe2b5648bcebd0c7071caee5ded26054e5d9c6","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.351841005000097},"started_at":"2026-10-13T05:11:30.814017+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0028495410006144084,"native_input_samples":144000,"native_input_sha256":"cf968e2cc5e461341a924a8754600d2f33bad7b3650ba829ac87c95215370b1e","output_sha256":"e48045df0858bd6c335b2a5fae3c895063d66a5557fc4abbd55be5ed309e113d","output_samples":144240,"length_delta_samples":240,"analysis_s":0.7912808480014064,"synthesis_s":0.12811418699857313,"total_s":0.9193950349999795,"rtf":0.15323250583332992,"output_peak":0.396406352519989,"output_rms":0.0608331700851532,"finished_at":"2026-10-13T05:11:31.745419+00:00","analysis_stages":{"external_f0_s":0.1482518720004009},"evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"e48045df0858bd6c335b2a5fae3c895063d66a5557fc4abbd55be5ed309e113d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.314356607045987,"output_peak":0.3966231192721894},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.846357808627619,"spectral_convergence":0.18783124143333735,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.490029095128012,"spectral_convergence":0.1666276774209641,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.73523947006686,"spectral_convergence":0.18570927539797402,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0007367175794762307,"delta_db":-0.03922380233461041,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.0023556136313119015,"delta_db":-0.16729718193041504,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.00037368744514293915,"delta_db":-0.19109336346927422,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":0.00010886449590415097,"delta_db":-0.4556322223339943,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":5.153180548466908e-05,"delta_db":-0.6299211647274071,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":8.283212656382165e-05,"delta_db":0.07424999416072145,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2595696214927371,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1549022464469184,"pitch":null,"output_rms_dbfs":-24.307142360719936,"output_peak":0.3966231192721894,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.690542124607497},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9531165966453029,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06199947247277372,"rms_dbfs":-24.152240114273017,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.061093431123016,"mean":-4.431947072347005e-05,"zero_energy":false},"output":{"rms":0.060903588431911904,"rms_dbfs":-24.307142360719936,"peak":0.3966231192721894,"peak_dbfs":-8.032439484929103,"crest_factor_db":16.274702875790833,"mean":0.00011935053304517344,"zero_energy":false},"rms_difference_db":-0.15490224644691963},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":1.1222374803098472e-06,"output_rms_dbfs":-59.499152308919385,"output_zero_energy":false,"output_peak":0.01920847165513831,"output_energy_fraction":4.437418881060564e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":181,"both_estimated_voiced_frames":169,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9285714285714286,"absolute_difference_cents_median":1.4798489053553878,"absolute_difference_cents_p95":5.848571982488003,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019565573818943504,"log_rms_mae_db":0.9222944299684023,"rms_envelope_nrmse":0.06128697885120627,"envelope_correlation":0.9964060541246158,"rms_gain_error_db":-0.15489157037749313,"reference_mean_power":0.0038470996951946897,"output_mean_power":0.0037123104158438354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.743468430119212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007829242029289011,"output_power":0.0007652524705745564,"reference_active":true,"delta_db":-0.09914978468619923},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011656588263872551,"output_power":9.565071084182729e-05,"reference_active":true,"delta_db":-0.8588325461669537},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.15404555717803e-05,"output_power":2.4773678613223457e-05,"reference_active":true,"delta_db":-1.0487746277713847}],"active_band_count":3,"active_band_mae_db":0.6689189862081792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018472557440126405,"log_rms_mae_db":0.5653470756366229,"rms_envelope_nrmse":0.0376686416524786,"envelope_correlation":0.9987644853832439,"rms_gain_error_db":-0.1547664042881122,"reference_mean_power":0.0038659686805529734,"output_mean_power":0.0037306258133426667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.614770154461247e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106100759473788,"output_power":0.00069458279964418,"reference_active":true,"delta_db":-0.09907336762680724},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7305685337334634e-05,"output_power":4.745471002584794e-05,"reference_active":true,"delta_db":-0.8191838680079031},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9649709160478004e-06,"output_power":1.4273739240226983e-06,"reference_active":true,"delta_db":-1.3881836808845796}],"active_band_count":3,"active_band_mae_db":0.7688136388397634}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016965929826439602,"log_rms_mae_db":1.1295548144879701,"rms_envelope_nrmse":0.06866707778387199,"envelope_correlation":0.9967984165349797,"rms_gain_error_db":-0.20969480071688879,"reference_mean_power":0.0021633278035333714,"output_mean_power":0.002061355288832802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3458386409803657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007177587670080827,"output_power":0.0006913939265592665,"reference_active":true,"delta_db":-0.1625294562490776},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.048912255781678e-05,"output_power":7.549907001852082e-05,"reference_active":true,"delta_db":-0.786547749128882},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.109541579796594e-05,"output_power":2.29701474145414e-05,"reference_active":true,"delta_db":-1.3153258607817873}],"active_band_count":3,"active_band_mae_db":0.754801022053249}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015157137475553504,"log_rms_mae_db":0.6033743713882952,"rms_envelope_nrmse":0.04475223714069848,"envelope_correlation":0.9988149709025113,"rms_gain_error_db":-0.2096701494888228,"reference_mean_power":0.002174161455071234,"output_mean_power":0.0020716900350805376},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2731603714689838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006867133707128568,"output_power":0.0006566936880738762,"reference_active":true,"delta_db":-0.19412661434932102},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067232594302272e-05,"output_power":3.3088981769714244e-05,"reference_active":true,"delta_db":-0.8961560622661964},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.085572616795342e-06,"output_power":1.387109388168548e-06,"reference_active":true,"delta_db":-1.7711460487416781}],"active_band_count":3,"active_band_mae_db":0.9538095751190653}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":1.1156869262894349,"rms_envelope_nrmse":0.11079416019776718,"envelope_correlation":0.9926445538151135,"rms_gain_error_db":-0.48630384278789496,"reference_mean_power":0.00014077271984820847,"output_mean_power":0.00012586011300571873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.622290255773313e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707421130487803e-05,"output_power":3.238762525761473e-05,"reference_active":true,"delta_db":-0.5869281526091051},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.087726823495623e-06,"output_power":4.651771624095028e-06,"reference_active":true,"delta_db":-0.38905399435596877},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8011808387081921e-06,"output_power":1.417022769321863e-06,"reference_active":true,"delta_db":-1.0418048948003749}],"active_band_count":3,"active_band_mae_db":0.6725956805884828}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":0.7188458725870067,"rms_envelope_nrmse":0.07645174508765752,"envelope_correlation":0.9975909316115309,"rms_gain_error_db":-0.48561561054881375,"reference_mean_power":0.00014139647469838982,"output_mean_power":0.00012643782623882743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.069193023932186e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.382591725297048e-05,"output_power":2.983434427494269e-05,"reference_active":true,"delta_db":-0.5453308515119858},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1244897853414308e-06,"output_power":1.9869052946131798e-06,"reference_active":true,"delta_db":-0.2907748026394592},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4002254483560297e-07,"output_power":9.281451983252318e-08,"reference_active":true,"delta_db":-1.7858204422945074}],"active_band_count":3,"active_band_mae_db":0.8739753654819841}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":1.1570624184463858,"rms_envelope_nrmse":0.13931445331059936,"envelope_correlation":0.9882133294016049,"rms_gain_error_db":0.004948400694486257,"reference_mean_power":8.952225188868398e-05,"output_mean_power":8.962431269369373e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.314971416941346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.877588546016399e-05,"output_power":5.57488061235875e-05,"reference_active":true,"delta_db":-0.2296360922242173},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.0495630695158e-06,"output_power":7.753988223691235e-06,"reference_active":true,"delta_db":-0.671024740398093},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5294615410314874e-06,"output_power":1.7891674090655119e-06,"reference_active":true,"delta_db":-1.5037710200850498}],"active_band_count":3,"active_band_mae_db":0.8014772842357867}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":0.8048750179751177,"rms_envelope_nrmse":0.0900405447259222,"envelope_correlation":0.9949677305476983,"rms_gain_error_db":0.0049549262590402095,"reference_mean_power":8.996899751083777e-05,"output_mean_power":9.007170297163927e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.951283845645361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.884212028117589e-05,"output_power":5.5302393612662524e-05,"reference_active":true,"delta_db":-0.2694438448183598},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.43607146326297e-06,"output_power":3.6581667772311354e-06,"reference_active":true,"delta_db":-0.8373503259232319},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7711408251558815e-07,"output_power":1.4976016165662152e-07,"reference_active":true,"delta_db":-0.7285679355833702}],"active_band_count":3,"active_band_mae_db":0.6117873687749874}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8224-274384-0010","method":"magphase","status":"measured","mos":{"status":"measured","value":3.660029411315918,"predicted_mos":3.660029411315918,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288756,"input_sample_rate":48000,"input_seconds":6.01575,"input_float32_sha256":"116cf3e3104136018d6b6c48f6cf6bf3fb53b8435c4819f858646dd078dcf864","internal_samples":96252,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.302344433001053},"started_at":"2026-10-13T05:18:28.293628+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00935234999997192,"native_input_samples":288000,"native_input_sha256":"d56256da764b1337bdfef076fd0cc29e86d38ef126189ea4d06240a9314b7146","output_sha256":"3631580221d1a0fe94f2ecd68ab7b82d71db828d8d002fa36ae96daeea70680b","output_samples":288756,"length_delta_samples":756,"analysis_s":3.761475523999252,"synthesis_s":1.25694836600087,"total_s":5.018423890000122,"rtf":0.836403981666687,"output_peak":0.3962531275581433,"output_rms":0.05930031874419776,"finished_at":"2026-10-13T05:18:33.336219+00:00","evaluation_input":{"reference":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"output_sha256":"3631580221d1a0fe94f2ecd68ab7b82d71db828d8d002fa36ae96daeea70680b","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96252,"length_delta_samples":252,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.536262333395964,"output_peak":0.3949374543808486},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9973818725844658,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.292603564940296,"spectral_convergence":0.5348373995154693,"spectral_convergence_status":"ok","amplitude_floor":1.597965137573199e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.644173787745322,"spectral_convergence":0.3252898548953123,"spectral_convergence_status":"ok","amplitude_floor":1.057274363712373e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.825274283143657,"spectral_convergence":0.27798464279816876,"spectral_convergence_status":"ok","amplitude_floor":9.008040625007043e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007434014661781762,"output_power":0.0006710297457642792,"delta_db":-0.444816410675429,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.002448126060340392,"output_power":0.002237994963731912,"delta_db":-0.3897467211543665,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00039049712732524606,"output_power":0.0003615319061984007,"delta_db":-0.33471212336598855,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012090645510899839,"output_power":0.00011721490989586223,"delta_db":-0.13466630106671157,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.957546973001399e-05,"output_power":5.9290954733199105e-05,"delta_db":-0.020790316469273675,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.142800821939945e-05,"output_power":8.066615407877715e-05,"delta_db":-0.04082460231778165,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2275927458415918,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3728081832445499,"pitch":null,"output_rms_dbfs":-24.525048297517564,"output_peak":0.3949374543808486,"reference_rms_dbfs":-24.152240114273017,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.587350545276425},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.821374141147598,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06199947247277372,"rms_dbfs":-24.152240114273017,"peak":0.393951416015625,"peak_dbfs":-8.09114668315,"crest_factor_db":16.061093431123016,"mean":-4.431947072347005e-05,"zero_energy":false},"output":{"rms":0.05939468523700106,"rms_dbfs":-24.525048297517564,"peak":0.3949374543808486,"peak_dbfs":-8.069433549144161,"crest_factor_db":16.455614748373403,"mean":-0.0006116645014936466,"zero_energy":false},"rms_difference_db":-0.37280818324454756},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.828116788645275e-06,"threshold_dbfs":-54.17014821185035,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":9.899294210216878e-07,"output_mean_power":1.3463908110702287e-06,"output_rms_dbfs":-58.70828861021472,"output_zero_energy":false,"output_peak":0.013400920461257536,"output_energy_fraction":5.59767127109963e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":182,"output_estimated_voiced_frames":180,"both_estimated_voiced_frames":158,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":22,"paired_voiced_fraction":0.8681318681318682,"absolute_difference_cents_median":8.115625435348058,"absolute_difference_cents_p95":25.753620377821104,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00019565573818943504,"log_rms_mae_db":2.41215866929555,"rms_envelope_nrmse":0.19445317406188534,"envelope_correlation":0.9618508033840593,"rms_gain_error_db":-0.3727981917689247,"reference_mean_power":0.0038470996951946897,"output_mean_power":0.0035306415003441},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.743468430119212e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007829242029289011,"output_power":0.000714043240611593,"reference_active":true,"delta_db":-0.39995206512624143},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011656588263872551,"output_power":0.00011026303312556017,"reference_active":true,"delta_db":-0.24141521593317408},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.15404555717803e-05,"output_power":2.8617952302727558e-05,"reference_active":true,"delta_db":-0.42229406498549005}],"active_band_count":3,"active_band_mae_db":0.3545537820149685}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00018472557440126405,"log_rms_mae_db":1.5919847757579841,"rms_envelope_nrmse":0.12292374753722997,"envelope_correlation":0.9846758332569208,"rms_gain_error_db":-0.3730167367994075,"reference_mean_power":0.0038659686805529734,"output_mean_power":0.0035477798083487395},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.614770154461247e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007106100759473788,"output_power":0.0006458717202628277,"reference_active":true,"delta_db":-0.4148509197345382},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7305685337334634e-05,"output_power":5.338984255550087e-05,"reference_active":true,"delta_db":-0.30739070622936426},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.9649709160478004e-06,"output_power":1.8378717335389165e-06,"reference_active":true,"delta_db":-0.290409283073787}],"active_band_count":3,"active_band_mae_db":0.3375503030125631}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00016965929826439602,"log_rms_mae_db":2.8957139790223962,"rms_envelope_nrmse":0.22262770494186887,"envelope_correlation":0.9626187839087388,"rms_gain_error_db":-0.39513464928874764,"reference_mean_power":0.0021633278035333714,"output_mean_power":0.0019751899492798583},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3458386409803657e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007177587670080827,"output_power":0.0006523714694962182,"reference_active":true,"delta_db":-0.4148354630168728},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.048912255781678e-05,"output_power":8.158688739178936e-05,"reference_active":true,"delta_db":-0.4497601228140471},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.109541579796594e-05,"output_power":2.8420449292908425e-05,"reference_active":true,"delta_db":-0.3906542911954688}],"active_band_count":3,"active_band_mae_db":0.4184166256754629}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015157137475553504,"log_rms_mae_db":1.9726558891365602,"rms_envelope_nrmse":0.1428024806744692,"envelope_correlation":0.9845537594562022,"rms_gain_error_db":-0.39514854641861796,"reference_mean_power":0.002174161455071234,"output_mean_power":0.0019850750799045285},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2731603714689838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006867133707128568,"output_power":0.0006233997905556991,"reference_active":true,"delta_db":-0.4200885141483268},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.067232594302272e-05,"output_power":3.7756178016456545e-05,"reference_active":true,"delta_db":-0.32310984194564546},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.085572616795342e-06,"output_power":1.9042613092512993e-06,"reference_active":true,"delta_db":-0.39498772474625576}],"active_band_count":3,"active_band_mae_db":0.379395360280076}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":5.465903254113391e-05,"log_rms_mae_db":2.7851700299066624,"rms_envelope_nrmse":0.25298702823113545,"envelope_correlation":0.9520609830197209,"rms_gain_error_db":-0.10765028178019764,"reference_mean_power":0.00014077271984820847,"output_mean_power":0.00013732622236776448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.622290255773313e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.707421130487803e-05,"output_power":3.789310190626054e-05,"reference_active":true,"delta_db":0.09488236940660749},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.087726823495623e-06,"output_power":4.840848317738841e-06,"reference_active":true,"delta_db":-0.21602309841595374},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8011808387081921e-06,"output_power":2.1096256348365256e-06,"reference_active":true,"delta_db":0.6864807587813291}],"active_band_count":3,"active_band_mae_db":0.3324620755346301}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.107059424790357e-05,"log_rms_mae_db":1.7980913255870068,"rms_envelope_nrmse":0.15307993814173945,"envelope_correlation":0.9815164335510539,"rms_gain_error_db":-0.10689273367992026,"reference_mean_power":0.00014139647469838982,"output_mean_power":0.00013795876832059411},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.069193023932186e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.382591725297048e-05,"output_power":3.4669684030369696e-05,"reference_active":true,"delta_db":0.10700301502206368},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.1244897853414308e-06,"output_power":2.138916737827142e-06,"reference_active":true,"delta_db":0.02939231685510009},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4002254483560297e-07,"output_power":1.253359928731227e-07,"reference_active":true,"delta_db":-0.4812216073698472}],"active_band_count":3,"active_band_mae_db":0.205872313082337}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.428054041295642e-05,"log_rms_mae_db":2.8502254017626254,"rms_envelope_nrmse":0.3484265411999831,"envelope_correlation":0.9252523245797322,"rms_gain_error_db":0.014145893470844727,"reference_mean_power":8.952225188868398e-05,"output_mean_power":8.98143202786955e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.314971416941346e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.877588546016399e-05,"output_power":5.752615679599887e-05,"reference_active":true,"delta_db":-0.09333819828149881},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.0495630695158e-06,"output_power":8.963201090924901e-06,"reference_active":true,"delta_db":-0.04164471128042704},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5294615410314874e-06,"output_power":3.557527193257893e-06,"reference_active":true,"delta_db":1.4812014785369918}],"active_band_count":3,"active_band_mae_db":0.5387281293663059}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.545251643358877e-05,"log_rms_mae_db":2.077603803012675,"rms_envelope_nrmse":0.20888791448024452,"envelope_correlation":0.9719106017863157,"rms_gain_error_db":0.014068185972664692,"reference_mean_power":8.996899751083777e-05,"output_mean_power":9.026090838266208e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.951283845645361e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.884212028117589e-05,"output_power":5.8407990536828225e-05,"reference_active":true,"delta_db":-0.03216048371657591},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.43607146326297e-06,"output_power":4.340364090548148e-06,"reference_active":true,"delta_db":-0.09472372092840395},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7711408251558815e-07,"output_power":2.0068224504185567e-07,"reference_active":true,"delta_db":0.5425585710994971}],"active_band_count":3,"active_band_mae_db":0.22314759191482567}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"hnm","status":"measured","mos":{"status":"measured","value":2.356114387512207,"predicted_mos":2.356114387512207,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"96d5d3ff8eff424c508f90a93a37b5e2cb87e1f0ca015050c2d4d01cc9a9a0f8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8305596230002266},"started_at":"2026-10-13T04:40:20.473282+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5982000149961095e-05,"native_input_samples":96000,"native_input_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","output_sha256":"1d59fe18e24651b70715a24225342dbe1fde912b9993e47fbaed946e5c414e92","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1382367969999905,"synthesis_s":0.5022017179999239,"total_s":1.6404385149999143,"rtf":0.27340641916665237,"output_peak":0.5151427906979105,"output_rms":0.0711700094535144,"finished_at":"2026-10-13T04:40:22.120095+00:00","evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"1d59fe18e24651b70715a24225342dbe1fde912b9993e47fbaed946e5c414e92","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.15900270066238,"spectral_convergence":0.23812004774488346,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.825229627626069,"spectral_convergence":0.22252489523810057,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.187362343805727,"spectral_convergence":0.23307432826958033,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.0013917327385738493,"delta_db":0.058677856802941936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.0032105626409920157,"delta_db":-0.23583129664320493,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.0003110911752994814,"delta_db":0.39394958313768014,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":8.626741630209187e-05,"delta_db":-0.02240364047781882,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.9470838913872557e-05,"delta_db":0.4417324026002184,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":3.60454355320171e-05,"delta_db":1.5312918832602875,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4473144438203587,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10201806503973633,"pitch":null,"output_rms_dbfs":-22.954059529714446,"output_peak":0.5151427906979105,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.057198224031392},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.15900270066238,"spectral_convergence":0.23812004774488346,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.825229627626069,"spectral_convergence":0.22252489523810057,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.187362343805727,"spectral_convergence":0.23307432826958033,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.0013917327385738493,"delta_db":0.058677856802941936,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.0032105626409920157,"delta_db":-0.23583129664320493,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.0003110911752994814,"delta_db":0.39394958313768014,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":8.626741630209187e-05,"delta_db":-0.02240364047781882,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.9470838913872557e-05,"delta_db":0.4417324026002184,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":3.60454355320171e-05,"delta_db":1.5312918832602875,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4473144438203587,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10201806503973633,"pitch":null,"output_rms_dbfs":-22.954059529714446,"output_peak":0.5151427906979105,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.057198224031392},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8614364547558147,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07201084825898034,"rms_dbfs":-22.85204146467471,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.29140388640507,"mean":-1.0033607482910157e-05,"zero_energy":false},"output":{"rms":0.0711700094535144,"rms_dbfs":-22.954059529714446,"peak":0.5151427906979105,"peak_dbfs":-5.761447472894295,"crest_factor_db":17.19261205682015,"mean":-6.342694798217632e-05,"zero_energy":false},"rms_difference_db":-0.10201806503973643},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":1.9898855291893853e-06,"output_rms_dbfs":-57.011719062391855,"output_zero_energy":false,"output_peak":0.014664263718040232,"output_energy_fraction":6.547609894957688e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":225,"both_estimated_voiced_frames":198,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9801980198019802,"absolute_difference_cents_median":1.436964827760761,"absolute_difference_cents_p95":6.250387860969969,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023735521522065317,"log_rms_mae_db":1.1392546639266419,"rms_envelope_nrmse":0.08100796376709929,"envelope_correlation":0.9934864907421433,"rms_gain_error_db":-0.1019850235160244,"reference_mean_power":0.0051897925018650195,"output_mean_power":0.005069340835962608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6580095170174236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00099786392963893,"output_power":0.0009464995195009977,"reference_active":true,"delta_db":-0.2295092639096425},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025981544310873085,"output_power":0.0002372422770184696,"reference_active":true,"delta_db":-0.3947287775180152},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167033379867012e-05,"output_power":4.172629932220573e-05,"reference_active":true,"delta_db":0.005828897591519465}],"active_band_count":3,"active_band_mae_db":0.2100223130063924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022053657198928817,"log_rms_mae_db":0.5828826265317245,"rms_envelope_nrmse":0.038711079330424225,"envelope_correlation":0.9984998114949745,"rms_gain_error_db":-0.1019465148520037,"reference_mean_power":0.0052155458817555326,"output_mean_power":0.005094541669700996},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3916979728964525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008985624617991317,"output_power":0.000850215901732052,"reference_active":true,"delta_db":-0.240190482292099},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001435836821660598,"output_power":0.00013204094905586874,"reference_active":true,"delta_db":-0.36396449309485335},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3776224720397374e-06,"output_power":2.06235121878409e-06,"reference_active":true,"delta_db":-0.6178026914893164}],"active_band_count":3,"active_band_mae_db":0.4073192222920896}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018677846007938983,"log_rms_mae_db":1.847570615546294,"rms_envelope_nrmse":0.10630201524056047,"envelope_correlation":0.9919044287133348,"rms_gain_error_db":-0.22862456292980354,"reference_mean_power":0.0023293656178310037,"output_mean_power":0.0022099131418598585},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5191576623571145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006490500325829311,"output_power":0.0006115840021455614,"reference_active":true,"delta_db":-0.25822059471082515},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013228128531435308,"output_power":0.00010520470443118445,"reference_active":true,"delta_db":-0.9946324552816822},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.716163060824064e-05,"output_power":1.1793338440196397e-05,"reference_active":true,"delta_db":-1.6292178794343841}],"active_band_count":3,"active_band_mae_db":0.9606903098089639}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017927232752442356,"log_rms_mae_db":1.115608639692262,"rms_envelope_nrmse":0.059129423591000256,"envelope_correlation":0.9977432169730214,"rms_gain_error_db":-0.22870178222829549,"reference_mean_power":0.002341010543733861,"output_mean_power":0.0022209214137056885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4246814808712245e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006148042140063018,"output_power":0.000571162168887673,"reference_active":true,"delta_db":-0.31977401505690217},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.176867672883114e-05,"output_power":4.922079552541278e-05,"reference_active":true,"delta_db":-0.9861966888316571},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0947792769720576e-06,"output_power":7.403894777313764e-07,"reference_active":true,"delta_db":-1.6986633005751726}],"active_band_count":3,"active_band_mae_db":1.0015446681545772}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":2.1305265538117486,"rms_envelope_nrmse":0.2526766837804495,"envelope_correlation":0.9496109796046067,"rms_gain_error_db":0.09019127364435375,"reference_mean_power":9.196537764000537e-05,"output_mean_power":9.389522044507948e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3593984658429965e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0478584079151855e-05,"output_power":2.09679972331778e-05,"reference_active":true,"delta_db":0.10257025079054616},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.015542002555996e-06,"output_power":6.909464543933178e-06,"reference_active":true,"delta_db":0.6016962904834503},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6245635384567155e-06,"output_power":2.257635324001121e-06,"reference_active":true,"delta_db":1.4291708988940113}],"active_band_count":3,"active_band_mae_db":0.7111458133893359}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":1.6225262032257486,"rms_envelope_nrmse":0.1724383502481658,"envelope_correlation":0.974521610813919,"rms_gain_error_db":0.09010948487317574,"reference_mean_power":9.242233001846998e-05,"output_mean_power":9.435998466070978e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.902316413249451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.934413521065947e-05,"output_power":2.064998378242246e-05,"reference_active":true,"delta_db":0.28370395780065805},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2438877541621994e-06,"output_power":4.0357421783547125e-06,"reference_active":true,"delta_db":0.9485759550591684},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3065359453687824e-07,"output_power":1.8690169563379811e-07,"reference_active":true,"delta_db":1.5549187893792726}],"active_band_count":3,"active_band_mae_db":0.9290662340796997}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":2.51142591663646,"rms_envelope_nrmse":0.46212429783276915,"envelope_correlation":0.8847438884694591,"rms_gain_error_db":0.7770250256459628,"reference_mean_power":2.938359073792447e-05,"output_mean_power":3.514045407134287e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1132929328909646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.650428912112274e-06,"output_power":9.662359757227055e-06,"reference_active":true,"delta_db":0.005365875374768065},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.01875786446687e-06,"output_power":5.176970992393931e-06,"reference_active":true,"delta_db":1.0998389139239895},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.253460098772304e-07,"output_power":1.6776066774690354e-06,"reference_active":true,"delta_db":3.6414491976577095}],"active_band_count":3,"active_band_mae_db":1.5822179956521556}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":2.422479171317437,"rms_envelope_nrmse":0.39728927390782226,"envelope_correlation":0.9118322211528975,"rms_gain_error_db":0.7769886744836667,"reference_mean_power":2.9530847534057218e-05,"output_mean_power":3.531626596631211e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9758537203809157e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.950101722919108e-06,"output_power":1.0560192261168265e-05,"reference_active":true,"delta_db":0.2584430443689964},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0870738801319313e-06,"output_power":3.1092862961436846e-06,"reference_active":true,"delta_db":1.7312288986402984},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.171675766600783e-08,"output_power":1.4037341444148538e-07,"reference_active":true,"delta_db":3.568817617955368}],"active_band_count":3,"active_band_mae_db":1.8528298536548877}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.122087001800537,"predicted_mos":3.122087001800537,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2f13b34f0fce249c95321dba252f099157e1e12fa3302334e6c7db642a2520ca","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.259060080999916},"started_at":"2026-10-13T04:43:36.037239+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3300000066228677e-05,"native_input_samples":96000,"native_input_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","output_sha256":"82ebb678b6ee02cbd7425ac06f7685f3b1cdb5b175e3482d03067bd24e3290ee","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7123288540005888,"synthesis_s":0.09319355599927803,"total_s":1.8055224099998668,"rtf":0.30092040166664447,"output_peak":0.6624789915006216,"output_rms":0.07870303376989161,"finished_at":"2026-10-13T04:43:37.849259+00:00","evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"82ebb678b6ee02cbd7425ac06f7685f3b1cdb5b175e3482d03067bd24e3290ee","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.699345403155438,"spectral_convergence":0.6019827557407377,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.318020854757015,"spectral_convergence":0.2722516158795313,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.471817491447664,"spectral_convergence":0.26785100006734586,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.0013580590062820886,"delta_db":-0.04769417843159157,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.0043063398842641985,"delta_db":1.039440266973943,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.000372862647499369,"delta_db":1.180561455734415,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":0.00010121440972981607,"delta_db":0.6715519148277029,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":3.0087555103492244e-05,"delta_db":0.5316764031927612,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":2.5604021705735632e-05,"delta_db":0.045870976103313144,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5861325325439546,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7718709337439635,"pitch":null,"output_rms_dbfs":-22.080170530930744,"output_peak":0.6624789915006216,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.829727916453372},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.699345403155438,"spectral_convergence":0.6019827557407377,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.318020854757015,"spectral_convergence":0.2722516158795313,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.471817491447664,"spectral_convergence":0.26785100006734586,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.0013580590062820886,"delta_db":-0.04769417843159157,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.0043063398842641985,"delta_db":1.039440266973943,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.000372862647499369,"delta_db":1.180561455734415,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":0.00010121440972981607,"delta_db":0.6715519148277029,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":3.0087555103492244e-05,"delta_db":0.5316764031927612,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":2.5604021705735632e-05,"delta_db":0.045870976103313144,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5861325325439546,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7718709337439635,"pitch":null,"output_rms_dbfs":-22.080170530930744,"output_peak":0.6624789915006216,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.829727916453372},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9265732292065288,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07201084825898034,"rms_dbfs":-22.85204146467471,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.29140388640507,"mean":-1.0033607482910157e-05,"zero_energy":false},"output":{"rms":0.07870303376989161,"rms_dbfs":-22.080170530930744,"peak":0.6624789915006216,"peak_dbfs":-3.576557789936887,"crest_factor_db":18.50361274099386,"mean":2.858178749531627e-07,"zero_energy":false},"rms_difference_db":0.7718709337439655},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":2.451424393018789e-06,"output_rms_dbfs":-56.10581496748249,"output_zero_energy":false,"output_peak":0.014515812122688154,"output_energy_fraction":6.596055572410299e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":237,"both_estimated_voiced_frames":201,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":36,"paired_voiced_fraction":0.995049504950495,"absolute_difference_cents_median":4.60007591132734,"absolute_difference_cents_p95":17.787718281637844,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023735521522065317,"log_rms_mae_db":2.3186777074781677,"rms_envelope_nrmse":0.1638259740268637,"envelope_correlation":0.9838834840691045,"rms_gain_error_db":0.7719177631346963,"reference_mean_power":0.0051897925018650195,"output_mean_power":0.006199287396809055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6580095170174236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00099786392963893,"output_power":0.0012879688016466577,"reference_active":true,"delta_db":1.1083401905739199},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025981544310873085,"output_power":0.000261545662615949,"reference_active":true,"delta_db":0.02882560800327772},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167033379867012e-05,"output_power":3.192289088774243e-05,"reference_active":true,"delta_db":-1.1572476587336027}],"active_band_count":3,"active_band_mae_db":0.7648044857702668}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022053657198928817,"log_rms_mae_db":1.7408358182829315,"rms_envelope_nrmse":0.12439106857872159,"envelope_correlation":0.9937682986395853,"rms_gain_error_db":0.7717709249898583,"reference_mean_power":0.0052155458817555326,"output_mean_power":0.006229839568331433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3916979728964525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008985624617991317,"output_power":0.001171096722577875,"reference_active":true,"delta_db":1.1504449394566585},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001435836821660598,"output_power":0.00013566441687676907,"reference_active":true,"delta_db":-0.24639134062739942},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3776224720397374e-06,"output_power":1.676195036500905e-06,"reference_active":true,"delta_db":-1.5181834652195494}],"active_band_count":3,"active_band_mae_db":0.9716732484345357}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018677846007938983,"log_rms_mae_db":2.060453580460949,"rms_envelope_nrmse":0.22061886463093688,"envelope_correlation":0.9870059237005556,"rms_gain_error_db":1.3435509149942186,"reference_mean_power":0.0023293656178310037,"output_mean_power":0.0031738964411509546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5191576623571145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006490500325829311,"output_power":0.0008895813613160844,"reference_active":true,"delta_db":1.3690749885336992},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013228128531435308,"output_power":0.00015858357589115034,"reference_active":true,"delta_db":0.7875980040976767},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.716163060824064e-05,"output_power":1.4993125094883064e-05,"reference_active":true,"delta_db":-0.586663853232245}],"active_band_count":3,"active_band_mae_db":0.9144456152878736}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017927232752442356,"log_rms_mae_db":1.5539693585386485,"rms_envelope_nrmse":0.19196493904651576,"envelope_correlation":0.9942707307162051,"rms_gain_error_db":1.3429245035659614,"reference_mean_power":0.002341010543733861,"output_mean_power":0.0031893032842001063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4246814808712245e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006148042140063018,"output_power":0.0008332785325416064,"reference_active":true,"delta_db":1.3205335780561374},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.176867672883114e-05,"output_power":7.260589674042047e-05,"reference_active":true,"delta_db":0.7020359613196197},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0947792769720576e-06,"output_power":7.280143198174958e-07,"reference_active":true,"delta_db":-1.7718664623929679}],"active_band_count":3,"active_band_mae_db":1.2648120005895749}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":1.6225144648823069,"rms_envelope_nrmse":0.19789702891504787,"envelope_correlation":0.9756762547816223,"rms_gain_error_db":0.6573239822963773,"reference_mean_power":9.196537764000537e-05,"output_mean_power":0.00010699334288619127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3593984658429965e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0478584079151855e-05,"output_power":2.337516271566252e-05,"reference_active":true,"delta_db":0.5745471714805315},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.015542002555996e-06,"output_power":6.587381038672997e-06,"reference_active":true,"delta_db":0.3943802184771735},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6245635384567155e-06,"output_power":1.5002147884506367e-06,"reference_active":true,"delta_db":-0.3458325951569794}],"active_band_count":3,"active_band_mae_db":0.4382533283715615}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":0.9876171712070455,"rms_envelope_nrmse":0.11063618823990153,"envelope_correlation":0.995142941998651,"rms_gain_error_db":0.6574110532873111,"reference_mean_power":9.242233001846998e-05,"output_mean_power":0.00010752712114235517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.902316413249451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.934413521065947e-05,"output_power":2.1948453614479167e-05,"reference_active":true,"delta_db":0.548546081584406},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2438877541621994e-06,"output_power":3.509074024516663e-06,"reference_active":true,"delta_db":0.3412671159858457},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3065359453687824e-07,"output_power":1.05910130660253e-07,"reference_active":true,"delta_db":-0.9118385869902261}],"active_band_count":3,"active_band_mae_db":0.6005505948534926}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":1.8402740606268653,"rms_envelope_nrmse":0.20709424316683991,"envelope_correlation":0.9752478170418349,"rms_gain_error_db":0.286887057047613,"reference_mean_power":2.938359073792447e-05,"output_mean_power":3.139016309033683e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1132929328909646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.650428912112274e-06,"output_power":1.0688481865363664e-05,"reference_active":true,"delta_db":0.4436940879323256},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.01875786446687e-06,"output_power":4.87799747491015e-06,"reference_active":true,"delta_db":0.8414973100847642},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.253460098772304e-07,"output_power":1.0278967980545082e-06,"reference_active":true,"delta_db":1.5140428687346408}],"active_band_count":3,"active_band_mae_db":0.9330780889172435}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":1.1603679540858347,"rms_envelope_nrmse":0.13199225537665035,"envelope_correlation":0.9898983225392052,"rms_gain_error_db":0.2868811984481379,"reference_mean_power":2.9530847534057218e-05,"output_mean_power":3.154743333012493e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9758537203809157e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.950101722919108e-06,"output_power":1.132386778710345e-05,"reference_active":true,"delta_db":0.5616726938296637},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0870738801319313e-06,"output_power":2.5863471097211793e-06,"reference_active":true,"delta_db":0.9314898757232857},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.171675766600783e-08,"output_power":8.72494130061708e-08,"reference_active":true,"delta_db":1.5035941178914287}],"active_band_count":3,"active_band_mae_db":0.9989188958147928}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9381134510040283,"predicted_mos":3.9381134510040283,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1ec35e1a5f85b120bd83252697130dc8fb8273be8de056b2fb27600497fe0f75","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.675906273001601},"started_at":"2026-10-13T07:00:43.798003+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.8016000214847736e-05,"native_input_samples":96000,"native_input_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","output_sha256":"f595fce1aed9d0a9c631d7b20eb0637c6a85e58b75a09370e5636194ec1ff225","output_samples":96000,"length_delta_samples":0,"analysis_s":2.1241392140000244,"synthesis_s":7.497619760000816,"total_s":9.62175897400084,"rtf":1.6036264956668067,"output_peak":0.5401490926742554,"output_rms":0.0683135086291175,"finished_at":"2026-10-13T07:00:53.431474+00:00","evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"f595fce1aed9d0a9c631d7b20eb0637c6a85e58b75a09370e5636194ec1ff225","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.412490830559427,"spectral_convergence":0.185375534023853,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.6019758737390095,"spectral_convergence":0.17996807292718592,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6288439509944315,"spectral_convergence":0.19302503291071724,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.001271149213933212,"delta_db":-0.33491524618566526,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.003074557472701255,"delta_db":-0.42381661756066025,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.00021281487690744596,"delta_db":-1.2549074915968457,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":5.798614716177023e-05,"delta_db":-1.7476290182279575,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.5340992209734487e-05,"delta_db":-0.21395642955376995,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":2.4886758307093783e-05,"delta_db":-0.07752759010783256,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6754587322054553,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4578267038647448,"pitch":null,"output_rms_dbfs":-23.309868168539452,"output_peak":0.5401490926742554,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.547770218430955},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.412490830559427,"spectral_convergence":0.185375534023853,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.6019758737390095,"spectral_convergence":0.17996807292718592,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.6288439509944315,"spectral_convergence":0.19302503291071724,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.001271149213933212,"delta_db":-0.33491524618566526,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.003074557472701255,"delta_db":-0.42381661756066025,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.00021281487690744596,"delta_db":-1.2549074915968457,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":5.798614716177023e-05,"delta_db":-1.7476290182279575,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.5340992209734487e-05,"delta_db":-0.21395642955376995,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":2.4886758307093783e-05,"delta_db":-0.07752759010783256,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6754587322054553,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4578267038647448,"pitch":null,"output_rms_dbfs":-23.309868168539452,"output_peak":0.5401490926742554,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.547770218430955},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9109502355946397,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07201084825898034,"rms_dbfs":-22.85204146467471,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.29140388640507,"mean":-1.0033607482910157e-05,"zero_energy":false},"output":{"rms":0.0683135086291175,"rms_dbfs":-23.309868168539452,"peak":0.5401490926742554,"peak_dbfs":-5.349726981737264,"crest_factor_db":17.96014118680219,"mean":7.862521853333722e-05,"zero_energy":false},"rms_difference_db":-0.4578267038647432},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":1.910871930577135e-06,"output_rms_dbfs":-57.18768419019017,"output_zero_energy":false,"output_peak":0.01040443405508995,"output_energy_fraction":6.824442005394839e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":205,"both_estimated_voiced_frames":197,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.9752475247524752,"absolute_difference_cents_median":2.423175928482704,"absolute_difference_cents_p95":8.567317613184729,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023735521522065317,"log_rms_mae_db":1.2613411455185068,"rms_envelope_nrmse":0.10404441390229043,"envelope_correlation":0.9915139723896107,"rms_gain_error_db":-0.45779491608179723,"reference_mean_power":0.0051897925018650195,"output_mean_power":0.004670576637616823},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6580095170174236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00099786392963893,"output_power":0.000860722892775201,"reference_active":true,"delta_db":-0.642079701029239},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025981544310873085,"output_power":0.00023645235912597804,"reference_active":true,"delta_db":-0.4092131002727514},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167033379867012e-05,"output_power":3.62662651825531e-05,"reference_active":true,"delta_db":-0.6032414654021022}],"active_band_count":3,"active_band_mae_db":0.5515114222346975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022053657198928817,"log_rms_mae_db":0.9025658419693755,"rms_envelope_nrmse":0.09398182700997024,"envelope_correlation":0.9931273451901206,"rms_gain_error_db":-0.45784178708177475,"reference_mean_power":0.0052155458817555326,"output_mean_power":0.004693702848554388},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3916979728964525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008985624617991317,"output_power":0.0007775184432246259,"reference_active":true,"delta_db":-0.6283757206127121},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001435836821660598,"output_power":0.0001283909798213252,"reference_active":true,"delta_db":-0.4857057330087603},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3776224720397374e-06,"output_power":2.1714357732769664e-06,"reference_active":true,"delta_db":-0.3939590849650252}],"active_band_count":3,"active_band_mae_db":0.5026801795288326}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018677846007938983,"log_rms_mae_db":1.4404142388632346,"rms_envelope_nrmse":0.12757424960613195,"envelope_correlation":0.9910615530182193,"rms_gain_error_db":-0.6203022928357886,"reference_mean_power":0.0023293656178310037,"output_mean_power":0.0020193306239564145},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5191576623571145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006490500325829311,"output_power":0.0005289416158571412,"reference_active":true,"delta_db":-0.8887043844542406},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013228128531435308,"output_power":0.00012212611275267145,"reference_active":true,"delta_db":-0.34689872274006134},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.716163060824064e-05,"output_power":1.674781746053506e-05,"reference_active":true,"delta_db":-0.1060033112011908}],"active_band_count":3,"active_band_mae_db":0.44720213946516424}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017927232752442356,"log_rms_mae_db":1.0867961260885988,"rms_envelope_nrmse":0.11604278552452185,"envelope_correlation":0.993063002621821,"rms_gain_error_db":-0.6205155069965018,"reference_mean_power":0.002341010543733861,"output_mean_power":0.002029325997131231},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4246814808712245e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006148042140063018,"output_power":0.000502351817950777,"reference_active":true,"delta_db":-0.8772885733206561},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.176867672883114e-05,"output_power":5.751043197922084e-05,"reference_active":true,"delta_db":-0.3102166780646503},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0947792769720576e-06,"output_power":1.211244198198738e-06,"reference_active":true,"delta_db":0.43905141739115183}],"active_band_count":3,"active_band_mae_db":0.5421855562588194}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":1.7105931095721694,"rms_envelope_nrmse":0.24086925618329644,"envelope_correlation":0.9723905327985305,"rms_gain_error_db":-1.6187177221592475,"reference_mean_power":9.196537764000537e-05,"output_mean_power":6.335087042048072e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3593984658429965e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0478584079151855e-05,"output_power":1.2570229717378952e-05,"reference_active":true,"delta_db":-2.1195671118554595},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.015542002555996e-06,"output_power":4.95398370985616e-06,"reference_active":true,"delta_db":-0.8432018936679618},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6245635384567155e-06,"output_power":1.7448411481881357e-06,"reference_active":true,"delta_db":0.31019192813117774}],"active_band_count":3,"active_band_mae_db":1.0909869778848664}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":1.3561517577748778,"rms_envelope_nrmse":0.21752247179127485,"envelope_correlation":0.9818599824566032,"rms_gain_error_db":-1.6204198457688253,"reference_mean_power":9.242233001846998e-05,"output_mean_power":6.364069716232519e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.902316413249451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.934413521065947e-05,"output_power":1.2126707773743638e-05,"reference_active":true,"delta_db":-2.0280640695653904},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2438877541621994e-06,"output_power":2.6555176901312103e-06,"reference_active":true,"delta_db":-0.869166193469023},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3065359453687824e-07,"output_power":1.4980572091647432e-07,"reference_active":true,"delta_db":0.5940703638944935}],"active_band_count":3,"active_band_mae_db":1.1637668756429689}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":1.9151257013947802,"rms_envelope_nrmse":0.22759928206348629,"envelope_correlation":0.9694329580034532,"rms_gain_error_db":0.18657400451241202,"reference_mean_power":2.938359073792447e-05,"output_mean_power":3.067342463128244e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1132929328909646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.650428912112274e-06,"output_power":1.0850651390612817e-05,"reference_active":true,"delta_db":0.5090919476898091},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.01875786446687e-06,"output_power":5.135872301338754e-06,"reference_active":true,"delta_db":1.0652237671168976},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.253460098772304e-07,"output_power":1.427503578527804e-06,"reference_active":true,"delta_db":2.9403197928658242}],"active_band_count":3,"active_band_mae_db":1.5048785025575102}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":1.5202907283923275,"rms_envelope_nrmse":0.1391115270396312,"envelope_correlation":0.9882013605627209,"rms_gain_error_db":0.18659726170115054,"reference_mean_power":2.9530847534057218e-05,"output_mean_power":3.0827310555757164e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9758537203809157e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.950101722919108e-06,"output_power":1.1774772606850947e-05,"reference_active":true,"delta_db":0.7312500814021355},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0870738801319313e-06,"output_power":2.7645959802754953e-06,"reference_active":true,"delta_db":1.2209384933321932},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.171675766600783e-08,"output_power":1.1375323292659467e-07,"reference_active":true,"delta_db":2.6556064636242387}],"active_band_count":3,"active_band_mae_db":1.5359316794528557}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5572397708892822,"predicted_mos":3.5572397708892822,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"d74bc637b07f63b5136aae4c12a7d46e09b812e907a5d00786cb9e235a746ca0","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.500408592000895},"started_at":"2026-10-13T07:30:35.048884+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008911207001801813,"native_input_samples":264600,"native_input_sha256":"588e7884f28ea4c744929777cd453c734b99a1100bc2b236362f893e89ada93b","output_sha256":"d529a09044d658fc6d206c730cee1f8f2a243a4a8a497ae986d11dc240560e15","output_samples":264704,"length_delta_samples":104,"analysis_s":0.13455262600109563,"synthesis_s":3.736445788999845,"total_s":3.8709984150009404,"rtf":0.6451664025001568,"output_peak":0.5137478709220886,"output_rms":0.06709357704465553,"finished_at":"2026-10-13T07:30:38.944322+00:00","evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"d529a09044d658fc6d206c730cee1f8f2a243a4a8a497ae986d11dc240560e15","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.46241080548416,"output_peak":0.512316188111683},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.587334655557962,"spectral_convergence":0.38929198257173364,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.183739571248648,"spectral_convergence":0.28988562997582173,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.556616655652872,"spectral_convergence":0.3223460130872102,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.0012245131221744947,"delta_db":-0.49724614362103375,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.002900120544932599,"delta_db":-0.6774822754190086,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.0002491110260185367,"delta_db":-0.5709978289879843,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":9.322288006224653e-05,"delta_db":0.3143536084155637,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.068261322840831e-05,"delta_db":-1.096138477666181,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":1.97945133915523e-05,"delta_db":-1.0717625912562776,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7046634875610082,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6086544630572776,"pitch":null,"output_rms_dbfs":-23.460695927731983,"output_peak":0.512316188111683,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.109230294153162},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7872498720939216,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07201084825898034,"rms_dbfs":-22.85204146467471,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.29140388640507,"mean":-1.0033607482910157e-05,"zero_energy":false},"output":{"rms":0.06713750590994454,"rms_dbfs":-23.460695927731983,"peak":0.512316188111683,"peak_dbfs":-5.809238422332551,"crest_factor_db":17.651457505399435,"mean":-0.00026721008338156043,"zero_energy":false},"rms_difference_db":-0.6086544630572739},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":1.5018785071547608e-06,"output_rms_dbfs":-58.233651976956,"output_zero_energy":false,"output_peak":0.010148513288140969,"output_energy_fraction":5.5533256910831285e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":221,"both_estimated_voiced_frames":182,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.900990099009901,"absolute_difference_cents_median":5.559771852290146,"absolute_difference_cents_p95":27.402707600473843,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023735521522065317,"log_rms_mae_db":1.8424729592934181,"rms_envelope_nrmse":0.11783516524502466,"envelope_correlation":0.9906648081534745,"rms_gain_error_db":-0.6085939276616632,"reference_mean_power":0.0051897925018650195,"output_mean_power":0.0045111846255987955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6580095170174236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00099786392963893,"output_power":0.0009230631700207839,"reference_active":true,"delta_db":-0.33839901140837625},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025981544310873085,"output_power":0.00021941657054603249,"reference_active":true,"delta_db":-0.7339553862581937},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167033379867012e-05,"output_power":2.9073444273351622e-05,"reference_active":true,"delta_db":-1.5633049454588386}],"active_band_count":3,"active_band_mae_db":0.8785531143751362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022053657198928817,"log_rms_mae_db":1.387361586017784,"rms_envelope_nrmse":0.09456387795478638,"envelope_correlation":0.9955896675071839,"rms_gain_error_db":-0.6084501208685218,"reference_mean_power":0.0052155458817555326,"output_mean_power":0.004533720661555411},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3916979728964525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008985624617991317,"output_power":0.0008385115215753336,"reference_active":true,"delta_db":-0.30039237209090847},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001435836821660598,"output_power":0.00011388612589006084,"reference_active":true,"delta_db":-1.0063426687541983},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3776224720397374e-06,"output_power":1.7022752215695932e-06,"reference_active":true,"delta_db":-1.451131192499331}],"active_band_count":3,"active_band_mae_db":0.919288744448146}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018677846007938983,"log_rms_mae_db":2.5861702880734216,"rms_envelope_nrmse":0.13907862533654583,"envelope_correlation":0.9858871429409641,"rms_gain_error_db":-0.26698837944485404,"reference_mean_power":0.0023293656178310037,"output_mean_power":0.002190477632826348},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5191576623571145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006490500325829311,"output_power":0.0006188467206520245,"reference_active":true,"delta_db":-0.20695082154774957},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013228128531435308,"output_power":0.00013963597994955083,"reference_active":true,"delta_db":0.23498931104721177},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.716163060824064e-05,"output_power":1.3673711058529956e-05,"reference_active":true,"delta_db":-0.9867215134801515}],"active_band_count":3,"active_band_mae_db":0.4762205486917043}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017927232752442356,"log_rms_mae_db":1.858309874654726,"rms_envelope_nrmse":0.10857057360085974,"envelope_correlation":0.9911608151424443,"rms_gain_error_db":-0.2668831829473468,"reference_mean_power":0.002341010543733861,"output_mean_power":0.0022014815567178363},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4246814808712245e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006148042140063018,"output_power":0.000589830036780718,"reference_active":true,"delta_db":-0.18009950595637542},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.176867672883114e-05,"output_power":6.384539305826913e-05,"reference_active":true,"delta_db":0.1436126746398962},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0947792769720576e-06,"output_power":1.0137257936471457e-06,"reference_active":true,"delta_db":-0.33406071076317884}],"active_band_count":3,"active_band_mae_db":0.21925763045315017}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":3.143726381876939,"rms_envelope_nrmse":0.3443729223161283,"envelope_correlation":0.9142760598515481,"rms_gain_error_db":0.1989312694325686,"reference_mean_power":9.196537764000537e-05,"output_mean_power":9.627587746116511e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3593984658429965e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0478584079151855e-05,"output_power":1.866153111801122e-05,"reference_active":true,"delta_db":-0.40352652228494607},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.015542002555996e-06,"output_power":5.5031323757216095e-06,"reference_active":true,"delta_db":-0.38664803809338505},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6245635384567155e-06,"output_power":1.4756595880098445e-06,"reference_active":true,"delta_db":-0.4175051776758424}],"active_band_count":3,"active_band_mae_db":0.4025599126847245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":2.452820059082042,"rms_envelope_nrmse":0.2932151501704479,"envelope_correlation":0.9334792475231215,"rms_gain_error_db":0.19913741087804623,"reference_mean_power":9.242233001846998e-05,"output_mean_power":9.675884023676787e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.902316413249451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.934413521065947e-05,"output_power":1.762666469390954e-05,"reference_active":true,"delta_db":-0.40379175984786386},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2438877541621994e-06,"output_power":3.0657796167943086e-06,"reference_active":true,"delta_db":-0.24524885796968157},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3065359453687824e-07,"output_power":1.2231874819348248e-07,"reference_active":true,"delta_db":-0.2862833465160196}],"active_band_count":3,"active_band_mae_db":0.311774654777855}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":4.251830096074453,"rms_envelope_nrmse":0.2363627367872256,"envelope_correlation":0.9683098429736894,"rms_gain_error_db":-0.8405967611688291,"reference_mean_power":2.938359073792447e-05,"output_mean_power":2.4212809788327768e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1132929328909646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.650428912112274e-06,"output_power":7.985658498074555e-06,"reference_active":true,"delta_db":-0.8223588202617081},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.01875786446687e-06,"output_power":2.4581949330359385e-06,"reference_active":true,"delta_db":-2.1347552101076435},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.253460098772304e-07,"output_power":5.064266430748466e-07,"reference_active":true,"delta_db":-1.5602868052430494}],"active_band_count":3,"active_band_mae_db":1.505800278537467}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":3.9441600087782076,"rms_envelope_nrmse":0.18980515929889333,"envelope_correlation":0.9808906284047513,"rms_gain_error_db":-0.8405870378033414,"reference_mean_power":2.9530847534057218e-05,"output_mean_power":2.433420753336799e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9758537203809157e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.950101722919108e-06,"output_power":8.16429633505422e-06,"reference_active":true,"delta_db":-0.8590876101077839},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0870738801319313e-06,"output_power":1.2797932429127695e-06,"reference_active":true,"delta_db":-2.123980100477822},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.171675766600783e-08,"output_power":3.806952061773261e-08,"reference_active":true,"delta_db":-2.0982569393053434}],"active_band_count":3,"active_band_mae_db":1.6937748832969832}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"golf","status":"measured","mos":{"status":"measured","value":2.7804577350616455,"predicted_mos":2.7804577350616455,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"8060ebe25a3898b5e70a4ff108b5020a3d09c806b0ae039f20858ca3c1a2ffe6","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5888927429987234},"started_at":"2026-10-13T05:03:04.949877+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027042729998356663,"native_input_samples":144000,"native_input_sha256":"1e07856bf1742f99029a8a600de8ad3f233da8dde596f249edfc8c5dab37d2be","output_sha256":"9f3f54c1da0df391437d71f8d9f27f518fb9a2a45681314046b61d86419f8472","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6556786259998262,"synthesis_s":0.129969450000317,"total_s":0.7856480760001432,"rtf":0.13094134600002386,"output_peak":0.561370849609375,"output_rms":0.0659274787149384,"finished_at":"2026-10-13T05:03:05.746208+00:00","analysis_stages":{"external_f0_s":0.09399752799981798},"evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"9f3f54c1da0df391437d71f8d9f27f518fb9a2a45681314046b61d86419f8472","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.615380747381668,"output_peak":0.5618685189761602},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.997553799617655,"spectral_convergence":0.6722183817400411,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.270371186991527,"spectral_convergence":0.27095519940222496,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.32280047160895,"spectral_convergence":0.261470930668201,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0013752013792961728,"output_power":0.001187827689032055,"delta_db":-0.636128545224522,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033950816841826935,"output_power":0.002816612463991574,"delta_db":-0.8112313089718843,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00028463042464631834,"output_power":0.0002461415625952517,"delta_db":-0.6309636740212152,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.685741101810409e-05,"output_power":5.990600864261389e-05,"delta_db":-1.6133649527063145,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6666236220195306e-05,"output_power":2.1642115830371452e-05,"delta_db":-0.9066200485816583,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.537946093815255e-05,"output_power":1.7596407586712152e-05,"delta_db":-1.5905838039382794,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.031482055573979,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.770246845250082,"pitch":null,"output_rms_dbfs":-23.615380747381668,"output_peak":0.5618685189761602,"reference_rms_dbfs":-22.845133902131586,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.863575152739377},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8868722149167646,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07206813856553836,"rms_dbfs":-22.845133902131586,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.284496323861948,"mean":-1.1161332934448039e-05,"zero_energy":false},"output":{"rms":0.06595245444771998,"rms_dbfs":-23.615380747381668,"peak":0.5618685189761602,"peak_dbfs":-5.007306007610723,"crest_factor_db":18.60807473977094,"mean":0.0009228007691605587,"zero_energy":false},"rms_difference_db":-0.7702468452500817},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":2.167130354341904e-06,"output_rms_dbfs":-56.64114964790603,"output_zero_energy":false,"output_peak":0.013191496976234432,"output_energy_fraction":8.317566610391564e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":237,"both_estimated_voiced_frames":200,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.9900990099009901,"absolute_difference_cents_median":3.723644226154935,"absolute_difference_cents_p95":24.549455611112315,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00023735521522065317,"log_rms_mae_db":2.0501457892609447,"rms_envelope_nrmse":0.1645192837504481,"envelope_correlation":0.9784941083562705,"rms_gain_error_db":-0.7702092707187236,"reference_mean_power":0.005198065650214829,"output_mean_power":0.00435332241664129},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.6596322191782467e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000998768011532967,"output_power":0.0008390520967397981,"reference_active":true,"delta_db":-0.756756973432039},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025889270585104996,"output_power":0.0001800027104675269,"reference_active":true,"delta_db":-1.5784076989014053},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1716945868278556e-05,"output_power":3.1032607901072766e-05,"reference_active":true,"delta_db":-1.2849423174779016}],"active_band_count":3,"active_band_mae_db":1.2067023299371151}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00022053657198928817,"log_rms_mae_db":1.4413441186941243,"rms_envelope_nrmse":0.11340947768670259,"envelope_correlation":0.993382339414764,"rms_gain_error_db":-0.7701917498028553,"reference_mean_power":0.005224157316415409,"output_mean_power":0.004375191549109486},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.3931106567679236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008994525087713219,"output_power":0.0007549639870095265,"reference_active":true,"delta_db":-0.7605200189109024},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014254519386015983,"output_power":9.677205226985666e-05,"reference_active":true,"delta_db":-1.682026276145433},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.376601086748361e-06,"output_power":1.7263567106999929e-06,"reference_active":true,"delta_db":-1.3882575406436923}],"active_band_count":3,"active_band_mae_db":1.2769346119000093}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00018677846007938983,"log_rms_mae_db":1.9160353255360487,"rms_envelope_nrmse":0.1669506755185622,"envelope_correlation":0.9821440280116575,"rms_gain_error_db":-0.6706434536305783,"reference_mean_power":0.002333162154603517,"output_mean_power":0.001999312024192725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.519778320229568e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006501410350917948,"output_power":0.0005396126112933554,"reference_active":true,"delta_db":-0.8092548734251483},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013136153795911056,"output_power":9.321626118956201e-05,"reference_active":true,"delta_db":-1.4897654462170555},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7156268449993968e-05,"output_power":1.4202718238385586e-05,"reference_active":true,"delta_db":-0.8205136182258578}],"active_band_count":3,"active_band_mae_db":1.0398446459560204}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00017927232752442356,"log_rms_mae_db":1.3850485497950924,"rms_envelope_nrmse":0.11821456632260526,"envelope_correlation":0.992886037204728,"rms_gain_error_db":-0.6706513558601186,"reference_mean_power":0.002344895212263434,"output_mean_power":0.0020093625527522053},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.4252558909178735e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006157272183118258,"output_power":0.0005049967943070755,"reference_active":true,"delta_db":-0.8609973081794782},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.093953215634015e-05,"output_power":4.3748069399646296e-05,"reference_active":true,"delta_db":-1.439402233581741},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0911311599615197e-06,"output_power":8.498776450290195e-07,"reference_active":true,"delta_db":-1.085205524936333}],"active_band_count":3,"active_band_mae_db":1.1285350222325174}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":2.107497762125597,"rms_envelope_nrmse":0.26880785717862904,"envelope_correlation":0.9587864145428517,"rms_gain_error_db":-1.637868057390227,"reference_mean_power":9.21132040920897e-05,"output_mean_power":6.317352104593977e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.3622748435208365e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0516773615816242e-05,"output_power":1.5090098212613802e-05,"reference_active":true,"delta_db":-1.3341700007600128},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.971369167982581e-06,"output_power":4.1560367468919e-06,"reference_active":true,"delta_db":-1.5739454254955876},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6265997160950674e-06,"output_power":1.2831328167903962e-06,"reference_active":true,"delta_db":-1.0300907973666507}],"active_band_count":3,"active_band_mae_db":1.3127354078740836}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":1.583258517481501,"rms_envelope_nrmse":0.20736550580467839,"envelope_correlation":0.9859394432561495,"rms_gain_error_db":-1.6378674950267058,"reference_mean_power":9.257538631469577e-05,"output_mean_power":6.349050530402569e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.9053844826504204e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.9369737394673202e-05,"output_power":1.4346568432804642e-05,"reference_active":true,"delta_db":-1.3037569856256772},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2122069871690372e-06,"output_power":2.3511559135126007e-06,"reference_active":true,"delta_db":-1.3552209274566112},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3047684272450028e-07,"output_power":1.0618232256886915e-07,"reference_active":true,"delta_db":-0.8948121850413344}],"active_band_count":3,"active_band_mae_db":1.1845966993745407}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":2.392036963418284,"rms_envelope_nrmse":0.2159736948519991,"envelope_correlation":0.9748471922048116,"rms_gain_error_db":-0.8340811907423258,"reference_mean_power":2.9432281596669863e-05,"output_mean_power":2.4289345413525142e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.115275032070489e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.65661578988092e-06,"output_power":7.443732275476035e-06,"reference_active":true,"delta_db":-1.130342076541976},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.9968245045602635e-06,"output_power":2.9905638033767364e-06,"reference_active":true,"delta_db":-1.2596200688896708},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.276283905759771e-07,"output_power":7.178398123856157e-07,"reference_active":true,"delta_db":-0.05882094771212652}],"active_band_count":3,"active_band_mae_db":0.8162610310479245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":1.9926464344351587,"rms_envelope_nrmse":0.1701315486388949,"envelope_correlation":0.9859989229329779,"rms_gain_error_db":-0.8340915981299925,"reference_mean_power":2.9580175405581994e-05,"output_mean_power":2.4411338064147792e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.9775963964193683e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.945669471776276e-06,"output_power":7.636235728074917e-06,"reference_active":true,"delta_db":-1.1475469560622602},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0714332838586525e-06,"output_power":1.621677910077889e-06,"reference_active":true,"delta_db":-1.0630634928085694},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.168791412725568e-08,"output_power":5.957217489104325e-08,"reference_active":true,"delta_db":-0.15156629726493748}],"active_band_count":3,"active_band_mae_db":0.7873922487119223}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.2076005935668945,"predicted_mos":3.2076005935668945,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"55653bf59e356cb0554428ae18c5de8f0afc5cea2698291322b9d206a180d293","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.372685588001332},"started_at":"2026-10-13T07:45:55.850335+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027313339996908326,"native_input_samples":144000,"native_input_sha256":"1e07856bf1742f99029a8a600de8ad3f233da8dde596f249edfc8c5dab37d2be","output_sha256":"a37d105776d4c7a9f7a2dd88d0e51a1ac4aa78f2af1cd08499e33dff602a38c1","output_samples":144120,"length_delta_samples":120,"analysis_s":2.358244987000944,"synthesis_s":3.595080377999693,"total_s":5.953325365000637,"rtf":0.9922208941667728,"output_peak":0.4550347626209259,"output_rms":0.07216455218824051,"finished_at":"2026-10-13T07:46:01.816191+00:00","analysis_stages":{"external_f0_s":1.4722466269995493},"evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"a37d105776d4c7a9f7a2dd88d0e51a1ac4aa78f2af1cd08499e33dff602a38c1","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.83063021495358,"output_peak":0.45525062197045985},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.650653949620955,"spectral_convergence":0.5505101820123172,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.463625660534043,"spectral_convergence":0.27984720681350944,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.379877686394428,"spectral_convergence":0.2842899877501069,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.0017341212054460154,"delta_db":1.0139139112357713,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.0031041050295545344,"delta_db":-0.3822786998825944,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.0002546091236035691,"delta_db":-0.4761877136700682,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":7.595584352251302e-05,"delta_db":-0.5752596408931692,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.608949763147595e-05,"delta_db":-0.08753541761288405,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":2.0509076035083855e-05,"delta_db":-0.9177499390011004,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5754875537159313,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.024909177440126877,"pitch":null,"output_rms_dbfs":-22.827132287234583,"output_peak":0.45525062197045985,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.831385765516475},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8896288065740544,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07201084825898034,"rms_dbfs":-22.85204146467471,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.29140388640507,"mean":-1.0033607482910157e-05,"zero_energy":false},"output":{"rms":0.07221765556838017,"rms_dbfs":-22.827132287234583,"peak":0.45525062197045985,"peak_dbfs":-6.8349890438556145,"crest_factor_db":15.992143243378967,"mean":0.0001550530175213908,"zero_energy":false},"rms_difference_db":0.024909177440125774},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":4.74080535546356e-06,"output_rms_dbfs":-53.24147875261814,"output_zero_energy":false,"output_peak":0.037069698044113746,"output_energy_fraction":0.0001515005128816905,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":225,"both_estimated_voiced_frames":186,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.9207920792079208,"absolute_difference_cents_median":7.0836666819809535,"absolute_difference_cents_p95":25.000961026732956,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023735521522065317,"log_rms_mae_db":3.1554624077601985,"rms_envelope_nrmse":0.15107277501444763,"envelope_correlation":0.9771899098915126,"rms_gain_error_db":0.024941193266512555,"reference_mean_power":0.0051897925018650195,"output_mean_power":0.00521968282204142},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6580095170174236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00099786392963893,"output_power":0.0010750884102800097,"reference_active":true,"delta_db":0.3237285586494323},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025981544310873085,"output_power":0.00025561853564872713,"reference_active":true,"delta_db":-0.07072618838166017},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167033379867012e-05,"output_power":2.9193299916108895e-05,"reference_active":true,"delta_db":-1.5454379036613952}],"active_band_count":3,"active_band_mae_db":0.6466308835641625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022053657198928817,"log_rms_mae_db":2.3531743319885297,"rms_envelope_nrmse":0.09488804732625099,"envelope_correlation":0.990207101741955,"rms_gain_error_db":0.025091966431331682,"reference_mean_power":0.0052155458817555326,"output_mean_power":0.005245766640177617},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3916979728964525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008985624617991317,"output_power":0.000975256973657293,"reference_active":true,"delta_db":0.35570792834728204},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001435836821660598,"output_power":0.00013403801327612302,"reference_active":true,"delta_db":-0.2987710445073816},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3776224720397374e-06,"output_power":1.6532900158533622e-06,"reference_active":true,"delta_db":-1.5779385375585917}],"active_band_count":3,"active_band_mae_db":0.7441391701377519}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018677846007938983,"log_rms_mae_db":2.594642688823939,"rms_envelope_nrmse":0.16786029860998075,"envelope_correlation":0.9799418499889971,"rms_gain_error_db":-0.4423299996427238,"reference_mean_power":0.0023293656178310037,"output_mean_power":0.0021038010178643742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5191576623571145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006490500325829311,"output_power":0.0006025441911525685,"reference_active":true,"delta_db":-0.32289272079350473},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013228128531435308,"output_power":0.00011961387542483467,"reference_active":true,"delta_db":-0.4371684457894963},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.716163060824064e-05,"output_power":1.3611624966650497e-05,"reference_active":true,"delta_db":-1.006485751397011}],"active_band_count":3,"active_band_mae_db":0.588848972660004}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017927232752442356,"log_rms_mae_db":1.7664604848856624,"rms_envelope_nrmse":0.11362435426446872,"envelope_correlation":0.9913925469164475,"rms_gain_error_db":-0.44233891332109343,"reference_mean_power":0.002341010543733861,"output_mean_power":0.002114313965437228},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4246814808712245e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006148042140063018,"output_power":0.0005690142007261221,"reference_active":true,"delta_db":-0.3361373049531912},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.176867672883114e-05,"output_power":5.5546946688871975e-05,"reference_active":true,"delta_db":-0.461081059584075},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0947792769720576e-06,"output_power":8.744879058028039e-07,"reference_active":true,"delta_db":-0.9757276052616295}],"active_band_count":3,"active_band_mae_db":0.5909819899329652}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":2.3770498314256345,"rms_envelope_nrmse":0.24219674773172145,"envelope_correlation":0.9531339588901467,"rms_gain_error_db":-0.45681294834491115,"reference_mean_power":9.196537764000537e-05,"output_mean_power":8.278336270646943e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3593984658429965e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0478584079151855e-05,"output_power":1.8580397531601555e-05,"reference_active":true,"delta_db":-0.42244923971494164},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.015542002555996e-06,"output_power":5.5208092339480455e-06,"reference_active":true,"delta_db":-0.37272022744881644},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6245635384567155e-06,"output_power":1.4456647208620118e-06,"reference_active":true,"delta_db":-0.5066911885817218}],"active_band_count":3,"active_band_mae_db":0.43395355191516}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":1.5097185437156855,"rms_envelope_nrmse":0.1229899421950121,"envelope_correlation":0.9886075975532813,"rms_gain_error_db":-0.45684323495644047,"reference_mean_power":9.242233001846998e-05,"output_mean_power":8.319411183034575e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.902316413249451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.934413521065947e-05,"output_power":1.7665086824293236e-05,"reference_active":true,"delta_db":-0.3943354279383132},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2438877541621994e-06,"output_power":2.950674475464907e-06,"reference_active":true,"delta_db":-0.41144518356206256},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3065359453687824e-07,"output_power":1.1186493366028142e-07,"reference_active":true,"delta_db":-0.6742739310242857}],"active_band_count":3,"active_band_mae_db":0.4933515141748872}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":2.2196869894456173,"rms_envelope_nrmse":0.24741658221593305,"envelope_correlation":0.9619990336772564,"rms_gain_error_db":-0.3976201099187459,"reference_mean_power":2.938359073792447e-05,"output_mean_power":2.6812842395025965e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1132929328909646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.650428912112274e-06,"output_power":8.67565626334412e-06,"reference_active":true,"delta_db":-0.462442793594636},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.01875786446687e-06,"output_power":3.8229663709718715e-06,"reference_active":true,"delta_db":-0.21691362427413308},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.253460098772304e-07,"output_power":6.538915860759078e-07,"reference_active":true,"delta_db":-0.4503947720921907}],"active_band_count":3,"active_band_mae_db":0.3765837299869866}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":1.3481713117505916,"rms_envelope_nrmse":0.14696936965804774,"envelope_correlation":0.9868493264137942,"rms_gain_error_db":-0.3976741307767895,"reference_mean_power":2.9530847534057218e-05,"output_mean_power":2.6946880616446825e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9758537203809157e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.950101722919108e-06,"output_power":9.028694706084709e-06,"reference_active":true,"delta_db":-0.4220255254133026},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0870738801319313e-06,"output_power":1.981186554718653e-06,"reference_active":true,"delta_db":-0.22612450897119524},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.171675766600783e-08,"output_power":5.616929303445146e-08,"reference_active":true,"delta_db":-0.40904144284655364}],"active_band_count":3,"active_band_mae_db":0.3523971590770172}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8167688846588135,"predicted_mos":2.8167688846588135,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"5aa16216369daedef0bcb0a14e2d76379a460ef7adf0426dfd01bdc93bae439a","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.664032849999785},"started_at":"2026-10-13T05:11:31.748120+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003498488000332145,"native_input_samples":144000,"native_input_sha256":"1e07856bf1742f99029a8a600de8ad3f233da8dde596f249edfc8c5dab37d2be","output_sha256":"cbeacab7c26ce2864d02b9dbf1cd7b65fc8a38c145baef7c7ab4a2af511e06a7","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9144291760003398,"synthesis_s":0.12519390000124986,"total_s":1.0396230760015897,"rtf":0.1732705126669316,"output_peak":0.5186994075775146,"output_rms":0.07049651066032894,"finished_at":"2026-10-13T05:11:32.798579+00:00","analysis_stages":{"external_f0_s":0.15142725400073687},"evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"cbeacab7c26ce2864d02b9dbf1cd7b65fc8a38c145baef7c7ab4a2af511e06a7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.033406453051548,"output_peak":0.5184452357062097},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.959351742541693,"spectral_convergence":0.17626800440266455,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.454540311383452,"spectral_convergence":0.15706609067026367,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.6893574125652195,"spectral_convergence":0.181290059834374,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.0013291551951499668,"delta_db":-0.14112364747066367,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.0032563879454023904,"delta_db":-0.17428139090239902,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.0002785808934382407,"delta_db":-0.08541406489843464,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":7.446259257494863e-05,"delta_db":-0.6614900217719417,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.2137650323926236e-05,"delta_db":-0.8008773509822248,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":2.0937849429942484e-05,"delta_db":-0.8278901674135243,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4485127739065314,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1742158469796713,"pitch":null,"output_rms_dbfs":-23.02625731165438,"output_peak":0.5184452357062097,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.7010831554967885},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9535508757058111,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07201084825898034,"rms_dbfs":-22.85204146467471,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.29140388640507,"mean":-1.0033607482910157e-05,"zero_energy":false},"output":{"rms":0.07058089065972047,"rms_dbfs":-23.02625731165438,"peak":0.5184452357062097,"peak_dbfs":-5.70594224303857,"crest_factor_db":17.32031506861581,"mean":0.00027369515273368026,"zero_energy":false},"rms_difference_db":-0.17421584697967063},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":1.8926185893236628e-06,"output_rms_dbfs":-57.22936898578063,"output_zero_energy":false,"output_peak":0.01702567469193201,"output_energy_fraction":6.331951536564696e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":212,"both_estimated_voiced_frames":202,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":0.9481202624024758,"absolute_difference_cents_p95":5.644138042775678,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023735521522065317,"log_rms_mae_db":1.0728045365814654,"rms_envelope_nrmse":0.06644245313599595,"envelope_correlation":0.9959048151674103,"rms_gain_error_db":-0.17419234801921343,"reference_mean_power":0.0051897925018650195,"output_mean_power":0.004985753002261362},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6580095170174236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00099786392963893,"output_power":0.0009645690098030593,"reference_active":true,"delta_db":-0.1473801965627558},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025981544310873085,"output_power":0.00023838080975008622,"reference_active":true,"delta_db":-0.37393670745225494},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167033379867012e-05,"output_power":3.0218741940060507e-05,"reference_active":true,"delta_db":-1.3955059944298667}],"active_band_count":3,"active_band_mae_db":0.6389409661482924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022053657198928817,"log_rms_mae_db":0.5917803811358798,"rms_envelope_nrmse":0.04134785251570547,"envelope_correlation":0.9985645819391953,"rms_gain_error_db":-0.1741580268766738,"reference_mean_power":0.0052155458817555326,"output_mean_power":0.0050105334708526765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3916979728964525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008985624617991317,"output_power":0.0008722371826710289,"reference_active":true,"delta_db":-0.12913675312117426},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001435836821660598,"output_power":0.00013065974005983176,"reference_active":true,"delta_db":-0.40963296658219744},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3776224720397374e-06,"output_power":1.721514234167482e-06,"reference_active":true,"delta_db":-1.4023227884756027}],"active_band_count":3,"active_band_mae_db":0.647030836059658}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018677846007938983,"log_rms_mae_db":1.2258685484392422,"rms_envelope_nrmse":0.07887581209022067,"envelope_correlation":0.9956856807961287,"rms_gain_error_db":-0.20643367834836782,"reference_mean_power":0.0023293656178310037,"output_mean_power":0.00222123390032391},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5191576623571145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006490500325829311,"output_power":0.0006133882444367836,"reference_active":true,"delta_db":-0.24542727568004058},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013228128531435308,"output_power":0.00011514230120680038,"reference_active":true,"delta_db":-0.602635011824965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.716163060824064e-05,"output_power":1.0010451575942064e-05,"reference_active":true,"delta_db":-2.3410488072054343}],"active_band_count":3,"active_band_mae_db":1.0630370315701467}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017927232752442356,"log_rms_mae_db":0.7110140698900326,"rms_envelope_nrmse":0.04819883035773732,"envelope_correlation":0.9985697664375297,"rms_gain_error_db":-0.2064085433695713,"reference_mean_power":0.002341010543733861,"output_mean_power":0.002232351175731288},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4246814808712245e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006148042140063018,"output_power":0.0005755674255375545,"reference_active":true,"delta_db":-0.2864062869507351},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.176867672883114e-05,"output_power":5.3626770684891906e-05,"reference_active":true,"delta_db":-0.613866522832772},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0947792769720576e-06,"output_power":6.207285475056273e-07,"reference_active":true,"delta_db":-2.4642484888741283}],"active_band_count":3,"active_band_mae_db":1.1215070995525451}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":1.2856673771796583,"rms_envelope_nrmse":0.16229046742284003,"envelope_correlation":0.9822687318129285,"rms_gain_error_db":-0.6804451762408192,"reference_mean_power":9.196537764000537e-05,"output_mean_power":7.862847288456044e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3593984658429965e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0478584079151855e-05,"output_power":1.7763662897939096e-05,"reference_active":true,"delta_db":-0.6176740260409832},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.015542002555996e-06,"output_power":4.907829361798567e-06,"reference_active":true,"delta_db":-0.8838530901198429},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6245635384567155e-06,"output_power":1.2693763473152836e-06,"reference_active":true,"delta_db":-1.0714630003630923}],"active_band_count":3,"active_band_mae_db":0.8576633721746395}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":0.8174050040900588,"rms_envelope_nrmse":0.11051952838149587,"envelope_correlation":0.9939310417865473,"rms_gain_error_db":-0.68048863725733,"reference_mean_power":9.242233001846998e-05,"output_mean_power":7.901836684132835e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.902316413249451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.934413521065947e-05,"output_power":1.6802667668311107e-05,"reference_active":true,"delta_db":-0.6117108136027474},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2438877541621994e-06,"output_power":2.710142362977358e-06,"reference_active":true,"delta_db":-0.7807371339542581},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3065359453687824e-07,"output_power":8.698615756646796e-08,"reference_active":true,"delta_db":-1.7667121531442667}],"active_band_count":3,"active_band_mae_db":1.053053366900424}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":1.4892732061536853,"rms_envelope_nrmse":0.13417594343558528,"envelope_correlation":0.9913494571217926,"rms_gain_error_db":-0.624636836089585,"reference_mean_power":2.938359073792447e-05,"output_mean_power":2.5447269123154574e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1132929328909646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.650428912112274e-06,"output_power":8.391992658056536e-06,"reference_active":true,"delta_db":-0.6068152070461847},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.01875786446687e-06,"output_power":3.202456555859561e-06,"reference_active":true,"delta_db":-0.9860859328726647},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.253460098772304e-07,"output_power":4.500222237332148e-07,"reference_active":true,"delta_db":-2.0731126499580554}],"active_band_count":3,"active_band_mae_db":1.222004596625635}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":1.0249259770043289,"rms_envelope_nrmse":0.08823169131093583,"envelope_correlation":0.9979720171346467,"rms_gain_error_db":-0.624683989572367,"reference_mean_power":2.9530847534057218e-05,"output_mean_power":2.5574521242142272e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9758537203809157e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.950101722919108e-06,"output_power":8.513154613793476e-06,"reference_active":true,"delta_db":-0.6773699968092957},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0870738801319313e-06,"output_power":1.5826059333039859e-06,"reference_active":true,"delta_db":-1.201650332911378},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.171675766600783e-08,"output_power":3.719638387597145e-08,"reference_active":true,"delta_db":-2.1990238092897014}],"active_band_count":3,"active_band_mae_db":1.3593480463367917}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0018","method":"magphase","status":"measured","mos":{"status":"measured","value":3.232407331466675,"predicted_mos":3.232407331466675,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289992,"input_sample_rate":48000,"input_seconds":6.0415,"input_float32_sha256":"cd3c74939e2fb4c11e7a4bd6221b4da71df7730166fb40a79abe5e2d35e86527","internal_samples":96664,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3797816969999985},"started_at":"2026-10-13T05:18:33.339664+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008667398999023135,"native_input_samples":288000,"native_input_sha256":"022747e1a9d1a30323e4ae85df8715468ae1df84cde9d52a142ac8ce1abc984f","output_sha256":"3cb9c8d9a5e2b1ccef6e3dadcf5547f77846dc2d0eaa893827b0b644b6986a01","output_samples":289992,"length_delta_samples":1992,"analysis_s":2.9398416670010192,"synthesis_s":1.155251407999458,"total_s":4.095093075000477,"rtf":0.6825155125000796,"output_peak":0.5390477143733479,"output_rms":0.06916597330193014,"finished_at":"2026-10-13T05:18:37.458997+00:00","evaluation_input":{"reference":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"3cb9c8d9a5e2b1ccef6e3dadcf5547f77846dc2d0eaa893827b0b644b6986a01","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96664,"length_delta_samples":664,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.198963115103307,"output_peak":0.5358993769147209},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9931308449888273,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.445678760892166,"spectral_convergence":0.6477102671594817,"spectral_convergence_status":"ok","amplitude_floor":1.9267471569597447e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.475337236456624,"spectral_convergence":0.3655340290686944,"spectral_convergence_status":"ok","amplitude_floor":1.271916808390449e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.796302206367003,"spectral_convergence":0.33457809178023284,"spectral_convergence_status":"ok","amplitude_floor":1.1644398478092437e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001373055391651405,"output_power":0.001242032089118616,"delta_db":-0.4355524144106285,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0033897235254818038,"output_power":0.003158308482699021,"delta_db":-0.3070973072042577,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0002841140646523882,"output_power":0.0002819735788469661,"delta_db":-0.032843168894298475,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.671358771236327e-05,"output_power":8.489299676067022e-05,"delta_db":-0.09215290575610803,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6620686952686656e-05,"output_power":2.9482565677702745e-05,"delta_db":0.44346016333752797,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.533501052724457e-05,"output_power":2.3683716991642775e-05,"delta_db":-0.2927122624240065,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.26730303700447117,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.31715169578453445,"pitch":null,"output_rms_dbfs":-23.169193160459244,"output_peak":0.5358993769147209,"reference_rms_dbfs":-22.85204146467471,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.572439401238597},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7086546022014427,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07201084825898034,"rms_dbfs":-22.85204146467471,"peak":0.527191162109375,"peak_dbfs":-5.560637578269638,"crest_factor_db":17.29140388640507,"mean":-1.0033607482910157e-05,"zero_energy":false},"output":{"rms":0.06942890918122378,"rms_dbfs":-23.169193160459244,"peak":0.5358993769147209,"peak_dbfs":-5.418334958125927,"crest_factor_db":17.75085820233332,"mean":0.0015431685312714727,"zero_energy":false},"rms_difference_db":-0.31715169578453484},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.610765325254761e-06,"threshold_dbfs":-52.50977895630577,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":100,"quiet_samples":16000,"reference_mean_power":1.495228207204491e-06,"output_mean_power":3.700968953586198e-05,"output_rms_dbfs":-44.316845580669394,"output_zero_energy":false,"output_peak":0.09211717953890546,"output_energy_fraction":0.0012796273315259005,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":202,"output_estimated_voiced_frames":213,"both_estimated_voiced_frames":179,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":34,"paired_voiced_fraction":0.8861386138613861,"absolute_difference_cents_median":11.131829376903976,"absolute_difference_cents_p95":46.19866614912433,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023735521522065317,"log_rms_mae_db":3.876640916299123,"rms_envelope_nrmse":0.2673848034547635,"envelope_correlation":0.9279216642446411,"rms_gain_error_db":-0.3170802958551231,"reference_mean_power":0.0051897925018650195,"output_mean_power":0.004824385068635204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6580095170174236e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00099786392963893,"output_power":0.0009615933928370253,"reference_active":true,"delta_db":-0.1607985361053863},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00025981544310873085,"output_power":0.00024624111309045797,"reference_active":true,"delta_db":-0.23304395786203416},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.167033379867012e-05,"output_power":3.311770125396567e-05,"reference_active":true,"delta_db":-0.9976679520046051}],"active_band_count":3,"active_band_mae_db":0.4638368153240085}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022053657198928817,"log_rms_mae_db":2.8908468987798686,"rms_envelope_nrmse":0.19824649870452096,"envelope_correlation":0.9574836145152035,"rms_gain_error_db":-0.3171269925414053,"reference_mean_power":0.0052155458817555326,"output_mean_power":0.004848273051699003},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.3916979728964525e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008985624617991317,"output_power":0.0008774687009742964,"reference_active":true,"delta_db":-0.10316637349580657},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001435836821660598,"output_power":0.00013064302634189243,"reference_active":true,"delta_db":-0.4101885424440982},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.3776224720397374e-06,"output_power":1.7591712105680616e-06,"reference_active":true,"delta_db":-1.308347877886028}],"active_band_count":3,"active_band_mae_db":0.6072342646086443}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00018677846007938983,"log_rms_mae_db":4.754208672925828,"rms_envelope_nrmse":0.27311478935595146,"envelope_correlation":0.9433508918134488,"rms_gain_error_db":-0.3397219337672009,"reference_mean_power":0.0023293656178310037,"output_mean_power":0.002154098049513803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.5191576623571145e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006490500325829311,"output_power":0.0006083163183531081,"reference_active":true,"delta_db":-0.28148709296501473},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013228128531435308,"output_power":0.00011996118024267318,"reference_active":true,"delta_db":-0.42457676163724045},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.716163060824064e-05,"output_power":1.55527555402593e-05,"reference_active":true,"delta_db":-0.4275120409120565}],"active_band_count":3,"active_band_mae_db":0.37785863183810386}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00017927232752442356,"log_rms_mae_db":3.723608709077931,"rms_envelope_nrmse":0.2111833699022685,"envelope_correlation":0.9645644208599252,"rms_gain_error_db":-0.33966683910402723,"reference_mean_power":0.002341010543733861,"output_mean_power":0.0021648942444072346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.4246814808712245e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006148042140063018,"output_power":0.0005776827827658143,"reference_active":true,"delta_db":-0.2704741155072262},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.176867672883114e-05,"output_power":5.524371661965813e-05,"reference_active":true,"delta_db":-0.48485408714923023},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0947792769720576e-06,"output_power":9.893222836878766e-07,"reference_active":true,"delta_db":-0.43988776753970277}],"active_band_count":3,"active_band_mae_db":0.39840532339871976}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.042588377582581e-05,"log_rms_mae_db":4.414991769124124,"rms_envelope_nrmse":0.3838413728305452,"envelope_correlation":0.8836747828274854,"rms_gain_error_db":-0.006895594434547258,"reference_mean_power":9.196537764000537e-05,"output_mean_power":9.181947369871529e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.3593984658429965e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.0478584079151855e-05,"output_power":2.064359267223515e-05,"reference_active":true,"delta_db":0.03485355675527514},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.015542002555996e-06,"output_power":5.815485626968394e-06,"reference_active":true,"delta_db":-0.1468877683098505},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6245635384567155e-06,"output_power":1.5788318034486857e-06,"reference_active":true,"delta_db":-0.12400835654010726}],"active_band_count":3,"active_band_mae_db":0.10191656053507764}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.620294109949224e-05,"log_rms_mae_db":3.383242190654532,"rms_envelope_nrmse":0.27880550943212173,"envelope_correlation":0.9331417825853745,"rms_gain_error_db":-0.006807267349348327,"reference_mean_power":9.242233001846998e-05,"output_mean_power":9.227757784601458e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.902316413249451e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.934413521065947e-05,"output_power":1.957446630644094e-05,"reference_active":true,"delta_db":0.05140610810026421},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2438877541621994e-06,"output_power":3.1667226848349244e-06,"reference_active":true,"delta_db":-0.10455785081844066},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3065359453687824e-07,"output_power":1.133037102519816e-07,"reference_active":true,"delta_db":-0.6187723098484286}],"active_band_count":3,"active_band_mae_db":0.2582454229223778}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.535256840646241e-05,"log_rms_mae_db":4.334554403993571,"rms_envelope_nrmse":0.5420607824481365,"envelope_correlation":0.8235325399976247,"rms_gain_error_db":0.1524940761648304,"reference_mean_power":2.938359073792447e-05,"output_mean_power":3.04336662783656e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.1132929328909646e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.650428912112274e-06,"output_power":9.974382661200972e-06,"reference_active":true,"delta_db":0.14339409711992063},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.01875786446687e-06,"output_power":4.291839802699101e-06,"reference_active":true,"delta_db":0.28551662954414614},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.253460098772304e-07,"output_power":9.721919334816425e-07,"reference_active":true,"delta_db":1.2720678696321515}],"active_band_count":3,"active_band_mae_db":0.5669928654320727}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.903489952134744e-05,"log_rms_mae_db":3.3724030406590373,"rms_envelope_nrmse":0.3984540995081977,"envelope_correlation":0.9001689327327038,"rms_gain_error_db":0.1520619204177037,"reference_mean_power":2.9530847534057218e-05,"output_mean_power":3.0583142156743295e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9758537203809157e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.950101722919108e-06,"output_power":1.0431875882833202e-05,"reference_active":true,"delta_db":0.2053489054351976},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.0870738801319313e-06,"output_power":2.226904932841088e-06,"reference_active":true,"delta_db":0.28163854396724053},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.171675766600783e-08,"output_power":8.748865278715367e-08,"reference_active":true,"delta_db":1.5154862698598512}],"active_band_count":3,"active_band_mae_db":0.6674912397540965}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"hnm","status":"measured","mos":{"status":"measured","value":2.358293294906616,"predicted_mos":2.358293294906616,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":62080,"input_sample_rate":16000,"input_seconds":3.88,"input_float32_sha256":"210848545d26868a786d0b8eaac9090569d0795ba9668cd1bf56c483be6f3172","internal_samples":62080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.162550889999693},"started_at":"2026-10-13T04:40:22.121624+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.4151999494060874e-05,"native_input_samples":62080,"native_input_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","output_sha256":"156dc4e995df2c415d551cdf677cd605a2a79b440ebbce0ca19f56b931ff697a","output_samples":62080,"length_delta_samples":0,"analysis_s":0.6800224089993208,"synthesis_s":0.3470191980004529,"total_s":1.0270416069997736,"rtf":0.26470144510303445,"output_peak":0.8101435138165535,"output_rms":0.06710801028433051,"finished_at":"2026-10-13T04:40:23.152367+00:00","evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"156dc4e995df2c415d551cdf677cd605a2a79b440ebbce0ca19f56b931ff697a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.097322531583025,"spectral_convergence":0.32256909525912736,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.442574233641787,"spectral_convergence":0.28842898940366085,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.521325127595958,"spectral_convergence":0.30772561546784094,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0007803015451113369,"delta_db":-0.06805236636643247,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.0030607777602367505,"delta_db":-0.36470670796224675,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.00042668749516632334,"delta_db":-0.32177109844094615,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":0.00013153198479589685,"delta_db":0.19275920523101206,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":3.563524587888569e-05,"delta_db":0.09273652413791555,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":6.855095225260797e-05,"delta_db":0.09385192635670062,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18897963808254228,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2849062468053716,"pitch":null,"output_rms_dbfs":-23.464512751716494,"output_peak":0.8101435138165535,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.68707396427359},"common_prefix":{"samples":62080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.097322531583025,"spectral_convergence":0.32256909525912736,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.442574233641787,"spectral_convergence":0.28842898940366085,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.521325127595958,"spectral_convergence":0.30772561546784094,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0007803015451113369,"delta_db":-0.06805236636643247,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.0030607777602367505,"delta_db":-0.36470670796224675,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.00042668749516632334,"delta_db":-0.32177109844094615,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":0.00013153198479589685,"delta_db":0.19275920523101206,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":3.563524587888569e-05,"delta_db":0.09273652413791555,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":6.855095225260797e-05,"delta_db":0.09385192635670062,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.18897963808254228,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2849062468053716,"pitch":null,"output_rms_dbfs":-23.464512751716494,"output_peak":0.8101435138165535,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.68707396427359},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":62080,"output_samples":62080,"intelligibility":{"estoi":{"value":0.8259247924806483,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06934572208561864,"rms_dbfs":-23.179606504911124,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.836701063786602,"mean":-2.484469069648035e-06,"zero_energy":false},"output":{"rms":0.06710801028433051,"rms_dbfs":-23.464512751716494,"peak":0.8101435138165535,"peak_dbfs":-1.8287608140871872,"crest_factor_db":21.63575193762931,"mean":-7.700003340161038e-05,"zero_energy":false},"rms_difference_db":-0.28490624680537024},"full_scale":{"reference":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":388,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":89,"quiet_samples":14240,"reference_mean_power":2.407837656504485e-06,"output_mean_power":2.8842849310787403e-06,"output_rms_dbfs":-55.39961839000166,"output_zero_energy":false,"output_peak":0.01269955073853524,"output_energy_fraction":0.00014690876817954242,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":381,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":100,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9565217391304348,"absolute_difference_cents_median":1.6115020886818194,"absolute_difference_cents_p95":7.383678851249605,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":62080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.00032928880743725145,"log_rms_mae_db":1.19705829218281,"rms_envelope_nrmse":0.1135958621305445,"envelope_correlation":0.9900763132951886,"rms_gain_error_db":-0.28491131433245964,"reference_mean_power":0.004815032927514685,"output_mean_power":0.004509289621449082},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.902918899290217e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008077922828075312,"output_power":0.0007696029436847313,"reference_active":true,"delta_db":-0.21032979665860424},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023728667544934464,"output_power":0.00021131059394316,"reference_active":true,"delta_db":-0.5035208089386843},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1018322593991373e-05,"output_power":1.834518929587746e-05,"reference_active":true,"delta_db":-0.5907585599552727}],"active_band_count":3,"active_band_mae_db":0.43486972185085376}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0003041508818572066,"log_rms_mae_db":0.6827991868699673,"rms_envelope_nrmse":0.06316382173793345,"envelope_correlation":0.9974056313464128,"rms_gain_error_db":-0.2849176478470814,"reference_mean_power":0.004852582666148238,"output_mean_power":0.00454444841230213},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.7242336092637012e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007426996521093885,"output_power":0.0007072910333262728,"reference_active":true,"delta_db":-0.2121506813202921},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010890208813537604,"output_power":9.599439841517531e-05,"reference_active":true,"delta_db":-0.5479031589854662},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4730800018618485e-06,"output_power":1.0581537861822224e-06,"reference_active":true,"delta_db":-1.4367754345910086}],"active_band_count":3,"active_band_mae_db":0.732276424965589}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.0002965570632263201,"log_rms_mae_db":1.6510030410115297,"rms_envelope_nrmse":0.16894195415681978,"envelope_correlation":0.9823174214256012,"rms_gain_error_db":-0.505814355254141,"reference_mean_power":0.002818219366063788,"output_mean_power":0.002508380173509724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.2520173458640518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044507178368220307,"output_power":0.0004166545772879053,"reference_active":true,"delta_db":-0.28653904824096665},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017793142616898364,"output_power":0.0001485297954336259,"reference_active":true,"delta_db":-0.7843907684722444},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6359305994808858e-05,"output_power":1.4273343091015991e-05,"reference_active":true,"delta_db":-0.5923917066568996}],"active_band_count":3,"active_band_mae_db":0.5544405077900368}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0002772405007925501,"log_rms_mae_db":1.0479007204940733,"rms_envelope_nrmse":0.10836244739758905,"envelope_correlation":0.9937792377075876,"rms_gain_error_db":-0.5058146599753585,"reference_mean_power":0.0028402077631144488,"output_mean_power":0.0025279509563940557},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.1541839932451843e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343143035723023,"output_power":0.0003872647022732037,"reference_active":true,"delta_db":-0.2839557550341209},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.103778868784843e-05,"output_power":6.72877108231983e-05,"reference_active":true,"delta_db":-0.8075182819791215},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3787553243482724e-06,"output_power":7.196018033897177e-07,"reference_active":true,"delta_db":-2.8239495951162317}],"active_band_count":3,"active_band_mae_db":1.3051412107098246}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":1.844385920063975,"rms_envelope_nrmse":0.2144925766034742,"envelope_correlation":0.9694414293132896,"rms_gain_error_db":0.21201595179308722,"reference_mean_power":0.00014490237542137563,"output_mean_power":0.00015215180198080327},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":6.946266094979906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.029476681448174e-05,"output_power":2.883196704885429e-05,"reference_active":true,"delta_db":-0.2149334102697749},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.606273878647887e-06,"output_power":7.753266932958244e-06,"reference_active":true,"delta_db":-0.4533042626751692},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.077953085521738e-06,"output_power":2.027953274936488e-06,"reference_active":true,"delta_db":2.744580844367479}],"active_band_count":3,"active_band_mae_db":1.1376061724374744}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":1.2694907015285417,"rms_envelope_nrmse":0.14396247655918912,"envelope_correlation":0.9857732170101979,"rms_gain_error_db":0.21201285168483752,"reference_mean_power":0.00014603285132237453,"output_mean_power":0.0001533387258274031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":6.454018573776752e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.924628617930061e-05,"output_power":2.7860812069901217e-05,"reference_active":true,"delta_db":-0.21076954493809755},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6077467608232813e-06,"output_power":3.5073395671880044e-06,"reference_active":true,"delta_db":-0.12258230788009265},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.60530589260614e-08,"output_power":1.4624292598016592e-07,"reference_active":true,"delta_db":2.303085549035477}],"active_band_count":3,"active_band_mae_db":0.8788124672845558}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":1.9590992817077002,"rms_envelope_nrmse":0.18851215382984693,"envelope_correlation":0.976065919829722,"rms_gain_error_db":0.10062112385661684,"reference_mean_power":5.235421681368183e-05,"output_mean_power":5.35813657900359e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":3.452458222743894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.744222722260822e-05,"output_power":1.7338610260150887e-05,"reference_active":true,"delta_db":-0.025876552385892152},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.482818619951916e-06,"output_power":7.0925714473403835e-06,"reference_active":true,"delta_db":-0.23261498878977815},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0100258421135403e-06,"output_power":1.2231791640750977e-06,"reference_active":true,"delta_db":0.8315758898361546}],"active_band_count":3,"active_band_mae_db":0.36335581033727493}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":1.3066596654951905,"rms_envelope_nrmse":0.09970858621348151,"envelope_correlation":0.9929259919282352,"rms_gain_error_db":0.10061977225899092,"reference_mean_power":5.276267962923642e-05,"output_mean_power":5.3999385904247904e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":3.130411452597277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7491368881094e-05,"output_power":1.7597712510733534e-05,"reference_active":true,"delta_db":0.026324196079341798},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.132236890482276e-06,"output_power":3.8968949650404375e-06,"reference_active":true,"delta_db":-0.25466510349682303},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.718357196867295e-08,"output_power":7.892533278405774e-08,"reference_active":true,"delta_db":0.6995333136387274}],"active_band_count":3,"active_band_mae_db":0.32684087107163073}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.9454026222229004,"predicted_mos":2.9454026222229004,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":62080,"input_sample_rate":16000,"input_seconds":3.88,"input_float32_sha256":"5f0713b4affdd986ed4f4407ba151a11fffd5d8b80ff0f6bf3371b207c4ebf52","internal_samples":62080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6121998259986867},"started_at":"2026-10-13T04:43:37.850109+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.1116000425536186e-05,"native_input_samples":62080,"native_input_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","output_sha256":"f0a8c3d6cc9d96a72211f0e6bf6fc55b7a6e33420f302074ac15bd71f3263379","output_samples":62080,"length_delta_samples":0,"analysis_s":1.0657044800000222,"synthesis_s":0.0755218269996476,"total_s":1.1412263069996698,"rtf":0.2941304914947603,"output_peak":0.8028650976410046,"output_rms":0.07860958054778902,"finished_at":"2026-10-13T04:43:38.995023+00:00","evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"f0a8c3d6cc9d96a72211f0e6bf6fc55b7a6e33420f302074ac15bd71f3263379","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.632969247034883,"spectral_convergence":0.5563457152823919,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2415417781587585,"spectral_convergence":0.28807034241646123,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.424296599244287,"spectral_convergence":0.2964318295901655,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0007781591287639915,"delta_db":-0.07999286893161568,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.004631399762741994,"delta_db":1.4340980117992308,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.000503975886285932,"delta_db":0.40122732151935403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":0.0001463139205185047,"delta_db":0.6553019442458873,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":4.055372518784383e-05,"delta_db":0.6542464653338614,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":7.906340911569346e-05,"delta_db":0.7134723725945524,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6563898307374169,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0891160835823537,"pitch":null,"output_rms_dbfs":-22.09049042132877,"output_peak":0.8028650976410046,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.766269208145975},"common_prefix":{"samples":62080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.632969247034883,"spectral_convergence":0.5563457152823919,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.2415417781587585,"spectral_convergence":0.28807034241646123,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.424296599244287,"spectral_convergence":0.2964318295901655,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0007781591287639915,"delta_db":-0.07999286893161568,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.004631399762741994,"delta_db":1.4340980117992308,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.000503975886285932,"delta_db":0.40122732151935403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":0.0001463139205185047,"delta_db":0.6553019442458873,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":4.055372518784383e-05,"delta_db":0.6542464653338614,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":7.906340911569346e-05,"delta_db":0.7134723725945524,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6563898307374169,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.0891160835823537,"pitch":null,"output_rms_dbfs":-22.09049042132877,"output_peak":0.8028650976410046,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.766269208145975},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":62080,"output_samples":62080,"intelligibility":{"estoi":{"value":0.9082183611663174,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06934572208561864,"rms_dbfs":-23.179606504911124,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.836701063786602,"mean":-2.484469069648035e-06,"zero_energy":false},"output":{"rms":0.07860958054778902,"rms_dbfs":-22.09049042132877,"peak":0.8028650976410046,"peak_dbfs":-1.9071484287211817,"crest_factor_db":20.183341992607588,"mean":-1.928524022344309e-08,"zero_energy":false},"rms_difference_db":1.0891160835823541},"full_scale":{"reference":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":388,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":89,"quiet_samples":14240,"reference_mean_power":2.407837656504485e-06,"output_mean_power":1.7428198463911777e-06,"output_rms_dbfs":-57.58747503172192,"output_zero_energy":false,"output_peak":0.011850587324337658,"output_energy_fraction":6.469337671232895e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":381,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":89,"both_estimated_voiced_frames":81,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.8804347826086957,"absolute_difference_cents_median":6.608053710050316,"absolute_difference_cents_p95":22.740336884975644,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":62080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.00032928880743725145,"log_rms_mae_db":2.5831390668177154,"rms_envelope_nrmse":0.20230689727337328,"envelope_correlation":0.9848468648475817,"rms_gain_error_db":1.0891169730195238,"reference_mean_power":0.004815032927514685,"output_mean_power":0.006187439403142594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.902918899290217e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008077922828075312,"output_power":0.0009635333753931849,"reference_active":true,"delta_db":0.7656706273159151},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023728667544934464,"output_power":0.00026434504789632886,"reference_active":true,"delta_db":0.46897827356057326},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1018322593991373e-05,"output_power":2.0133584099470412e-05,"reference_active":true,"delta_db":-0.18676960279330052}],"active_band_count":3,"active_band_mae_db":0.47380616788992963}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0003041508818572066,"log_rms_mae_db":2.067808856070134,"rms_envelope_nrmse":0.17423834957834924,"envelope_correlation":0.9919559243051312,"rms_gain_error_db":1.0891318399106995,"reference_mean_power":0.004852582666148238,"output_mean_power":0.006235713115640557},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.7242336092637012e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007426996521093885,"output_power":0.000876886811675298,"reference_active":true,"delta_db":0.7213031787722446},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010890208813537604,"output_power":0.00013297533952693008,"reference_active":true,"delta_db":0.8673490069698928},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4730800018618485e-06,"output_power":1.2985293675714002e-06,"reference_active":true,"delta_db":-0.5477455759745382}],"active_band_count":3,"active_band_mae_db":0.7121325872388918}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.0002965570632263201,"log_rms_mae_db":2.043938237397623,"rms_envelope_nrmse":0.26645762382462285,"envelope_correlation":0.9863117363905703,"rms_gain_error_db":1.6649051629422276,"reference_mean_power":0.002818219366063788,"output_mean_power":0.004134902859687916},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.2520173458640518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044507178368220307,"output_power":0.000593954069576146,"reference_active":true,"delta_db":1.2532280023127151},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017793142616898364,"output_power":0.00022084664051853873,"reference_active":true,"delta_db":0.9383813747515628},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6359305994808858e-05,"output_power":1.1247030016205932e-05,"reference_active":true,"delta_db":-1.6272702160821875}],"active_band_count":3,"active_band_mae_db":1.2729598643821551}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0002772405007925501,"log_rms_mae_db":1.4452506910215106,"rms_envelope_nrmse":0.2443698039927994,"envelope_correlation":0.9923133403731907,"rms_gain_error_db":1.6649043825195928,"reference_mean_power":0.0028402077631144488,"output_mean_power":0.004167163575996322},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.1541839932451843e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343143035723023,"output_power":0.0005400678002147772,"reference_active":true,"delta_db":1.1604479467942073},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.103778868784843e-05,"output_power":0.00010943206314563345,"reference_active":true,"delta_db":1.3045700540610683},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3787553243482724e-06,"output_power":9.677409545592209e-07,"reference_active":true,"delta_db":-1.537280818946042}],"active_band_count":3,"active_band_mae_db":1.3340996066004394}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":1.5170412958336972,"rms_envelope_nrmse":0.135252066899737,"envelope_correlation":0.9923122602175332,"rms_gain_error_db":0.6605621202601679,"reference_mean_power":0.00014490237542137563,"output_mean_power":0.00016870646171726478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":6.946266094979906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.029476681448174e-05,"output_power":3.496304048637804e-05,"reference_active":true,"delta_db":0.6224157911486673},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.606273878647887e-06,"output_power":8.559126766311832e-06,"reference_active":true,"delta_db":-0.023857039679616752},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.077953085521738e-06,"output_power":8.677989093880983e-07,"reference_active":true,"delta_db":-0.9418075998537984}],"active_band_count":3,"active_band_mae_db":0.5293601435606942}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":0.8840590929392834,"rms_envelope_nrmse":0.10050466559804448,"envelope_correlation":0.9974494825392478,"rms_gain_error_db":0.6605622732231136,"reference_mean_power":0.00014603285132237453,"output_mean_power":0.0001700226544745523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":6.454018573776752e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.924628617930061e-05,"output_power":3.332416888303624e-05,"reference_active":true,"delta_db":0.5668860144165887},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6077467608232813e-06,"output_power":3.7115246507894527e-06,"reference_active":true,"delta_db":0.12316303912858004},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.60530589260614e-08,"output_power":6.440560774610528e-08,"reference_active":true,"delta_db":-1.2584263010529395}],"active_band_count":3,"active_band_mae_db":0.6494917848660361}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":1.9702044023832566,"rms_envelope_nrmse":0.32377937037964166,"envelope_correlation":0.9404751444794685,"rms_gain_error_db":0.7682853610252128,"reference_mean_power":5.235421681368183e-05,"output_mean_power":6.248563725562405e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":3.452458222743894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.744222722260822e-05,"output_power":2.1720561620335834e-05,"reference_active":true,"delta_db":0.952691106372124},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.482818619951916e-06,"output_power":1.0193518769107359e-05,"reference_active":true,"delta_db":1.3425890865856684},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0100258421135403e-06,"output_power":1.978881175569312e-06,"reference_active":true,"delta_db":2.9208723161804584}],"active_band_count":3,"active_band_mae_db":1.7387175030460835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":1.1921278109506501,"rms_envelope_nrmse":0.21327582462102906,"envelope_correlation":0.9757494327629047,"rms_gain_error_db":0.7682856761120103,"reference_mean_power":5.276267962923642e-05,"output_mean_power":6.297314905646831e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":3.130411452597277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7491368881094e-05,"output_power":2.2612131887259458e-05,"reference_active":true,"delta_db":1.115177109985804},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.132236890482276e-06,"output_power":5.719328839520448e-06,"reference_active":true,"delta_db":1.4115985699153306},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.718357196867295e-08,"output_power":1.3597889991018503e-07,"reference_active":true,"delta_db":3.0620843296285525}],"active_band_count":3,"active_band_mae_db":1.8629533365098958}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"facodec","status":"measured","mos":{"status":"measured","value":3.7268738746643066,"predicted_mos":3.7268738746643066,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":62200,"input_sample_rate":16000,"input_seconds":3.8875,"input_float32_sha256":"1289f4402e0f5a324e6f548bf5308e702b776fb77e1acdff06fe5f5ffb00fdba","internal_samples":62200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.7525892369994835},"started_at":"2026-10-13T07:00:53.433232+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.9714001357206143e-05,"native_input_samples":62080,"native_input_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","output_sha256":"2d4696b9685cb02ffef3bccc3bb8c4c35e355f5adc32d99c93292415efff7692","output_samples":62200,"length_delta_samples":120,"analysis_s":0.9722069529998407,"synthesis_s":7.536612010000681,"total_s":8.508818963000522,"rtf":2.192994578092918,"output_peak":0.6030581593513489,"output_rms":0.06453069781921734,"finished_at":"2026-10-13T07:01:01.957652+00:00","evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"2d4696b9685cb02ffef3bccc3bb8c4c35e355f5adc32d99c93292415efff7692","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":62080,"output_samples":62200,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.80467277025891,"output_peak":0.6030581593513489},"common_prefix":{"samples":62080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980707395498393,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.318919439857137,"spectral_convergence":0.20931870339094766,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.3799802955860985,"spectral_convergence":0.21792266849222744,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.380535325722487,"spectral_convergence":0.22877661842561334,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0008260897063476203,"delta_db":0.17959507065056235,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.002761915562844098,"delta_db":-0.8109207067104711,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.00040464733088842674,"delta_db":-0.5521034505039152,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":8.695507460017331e-05,"delta_db":-1.6046052105811692,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":3.107719510223841e-05,"delta_db":-0.501642930162959,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":6.157544724706381e-05,"delta_db":-0.3722072441133601,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6701791021204061,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6166795226561383,"pitch":null,"output_rms_dbfs":-23.796286027567263,"output_peak":0.6030581593513489,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.359811687055241},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":62080,"output_samples":62080,"intelligibility":{"estoi":{"value":0.8919904857268909,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06934572208561864,"rms_dbfs":-23.179606504911124,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.836701063786602,"mean":-2.484469069648035e-06,"zero_energy":false},"output":{"rms":0.06459303613417654,"rms_dbfs":-23.796286027567263,"peak":0.6030581593513489,"peak_dbfs":-4.392816043520046,"crest_factor_db":19.403469984047216,"mean":2.308583855181299e-05,"zero_energy":false},"rms_difference_db":-0.6166795226561383},"full_scale":{"reference":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":388,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":89,"quiet_samples":14240,"reference_mean_power":2.407837656504485e-06,"output_mean_power":2.2680930663719753e-06,"output_rms_dbfs":-56.44339129063721,"output_zero_energy":false,"output_peak":0.01288641057908535,"output_energy_fraction":0.0001246946311036992,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":381,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":100,"both_estimated_voiced_frames":90,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9782608695652174,"absolute_difference_cents_median":2.561730189214172,"absolute_difference_cents_p95":10.293947167361997,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":62080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.00032928880743725145,"log_rms_mae_db":1.1414675459249242,"rms_envelope_nrmse":0.13728621740298141,"envelope_correlation":0.9884336756718998,"rms_gain_error_db":-0.616680457132165,"reference_mean_power":0.004815032927514685,"output_mean_power":0.004177641951749809},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.902918899290217e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008077922828075312,"output_power":0.0007157105052183076,"reference_active":true,"delta_db":-0.5256230800195959},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023728667544934464,"output_power":0.00022593603151288774,"reference_active":true,"delta_db":-0.21287855356342064},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1018322593991373e-05,"output_power":1.891419994980613e-05,"reference_active":true,"delta_db":-0.45810077572319785}],"active_band_count":3,"active_band_mae_db":0.39886746976873816}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0003041508818572066,"log_rms_mae_db":0.792472238112807,"rms_envelope_nrmse":0.1284403805514586,"envelope_correlation":0.990081969151222,"rms_gain_error_db":-0.616677094457072,"reference_mean_power":0.004852582666148238,"output_mean_power":0.0042102242959061904},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.7242336092637012e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007426996521093885,"output_power":0.0006574580629681989,"reference_active":true,"delta_db":-0.5294516462077643},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010890208813537604,"output_power":0.00010780325904258882,"reference_active":true,"delta_db":-0.04404316808464565},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4730800018618485e-06,"output_power":1.335610627766864e-06,"reference_active":true,"delta_db":-0.4254646750920596}],"active_band_count":3,"active_band_mae_db":0.33298649646148987}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.0002965570632263201,"log_rms_mae_db":1.2197891079807044,"rms_envelope_nrmse":0.18155247708241742,"envelope_correlation":0.986000765295227,"rms_gain_error_db":-1.0410902280111873,"reference_mean_power":0.002818219366063788,"output_mean_power":0.002217510945363926},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.2520173458640518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044507178368220307,"output_power":0.0003954288306803878,"reference_active":true,"delta_db":-0.5136173158497114},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017793142616898364,"output_power":0.00016243480289105994,"reference_active":true,"delta_db":-0.39573573987356675},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6359305994808858e-05,"output_power":1.4440374726378076e-05,"reference_active":true,"delta_db":-0.5418641252002605}],"active_band_count":3,"active_band_mae_db":0.4837390603078462}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0002772405007925501,"log_rms_mae_db":0.9493668192577451,"rms_envelope_nrmse":0.173214298847198,"envelope_correlation":0.9878457852103443,"rms_gain_error_db":-1.0410902033129146,"reference_mean_power":0.0028402077631144488,"output_mean_power":0.002234812489571497},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.1541839932451843e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343143035723023,"output_power":0.0003661912126142449,"reference_active":true,"delta_db":-0.5269557152804942},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.103778868784843e-05,"output_power":7.751532301772946e-05,"reference_active":true,"delta_db":-0.19300020596172143},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3787553243482724e-06,"output_power":1.3179477070609697e-06,"reference_active":true,"delta_db":-0.1958902371143515}],"active_band_count":3,"active_band_mae_db":0.3052820527855224}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":1.6441910792298544,"rms_envelope_nrmse":0.22484855908493742,"envelope_correlation":0.9807095684596956,"rms_gain_error_db":-1.5418233059848894,"reference_mean_power":0.00014490237542137563,"output_mean_power":0.00010159987518762459},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":6.946266094979906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.029476681448174e-05,"output_power":2.3867299607430696e-05,"reference_active":true,"delta_db":-1.035643291185138},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.606273878647887e-06,"output_power":7.701224737614096e-06,"reference_active":true,"delta_db":-0.4825536539569091},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.077953085521738e-06,"output_power":1.3824740413030064e-06,"reference_active":true,"delta_db":1.080571253349766}],"active_band_count":3,"active_band_mae_db":0.8662560661639377}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":1.2199704015057555,"rms_envelope_nrmse":0.2077670127710513,"envelope_correlation":0.986694328302312,"rms_gain_error_db":-1.541826718026553,"reference_mean_power":0.00014603285132237453,"output_mean_power":0.0001023924402057259},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":6.454018573776752e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.924628617930061e-05,"output_power":2.3328968137882533e-05,"reference_active":true,"delta_db":-0.9817519534929853},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6077467608232813e-06,"output_power":3.2576126989983876e-06,"reference_active":true,"delta_db":-0.4433659612025647},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.60530589260614e-08,"output_power":9.971040282536361e-08,"reference_active":true,"delta_db":0.6397415799505295}],"active_band_count":3,"active_band_mae_db":0.6882864982153598}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":1.8430134915605416,"rms_envelope_nrmse":0.2612782023538139,"envelope_correlation":0.9529953853192618,"rms_gain_error_db":-0.2874710824719243,"reference_mean_power":5.235421681368183e-05,"output_mean_power":4.900095665838163e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":3.452458222743894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.744222722260822e-05,"output_power":1.340947281345366e-05,"reference_active":true,"delta_db":-1.141902356903197},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.482818619951916e-06,"output_power":6.230671215876543e-06,"reference_active":true,"delta_db":-0.7953038352922857},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0100258421135403e-06,"output_power":1.8173775582929437e-06,"reference_active":true,"delta_db":2.5511267529902524}],"active_band_count":3,"active_band_mae_db":1.4961109817285783}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":1.4060746607775088,"rms_envelope_nrmse":0.184523279981547,"envelope_correlation":0.9753175162651851,"rms_gain_error_db":-0.28747221815205365,"reference_mean_power":5.276267962923642e-05,"output_mean_power":4.9383244730898054e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":3.130411452597277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7491368881094e-05,"output_power":1.3863592987390532e-05,"reference_active":true,"delta_db":-1.0094799914660544},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.132236890482276e-06,"output_power":3.6170975875132197e-06,"reference_active":true,"delta_db":-0.5782498457256021},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.718357196867295e-08,"output_power":1.244527227459738e-07,"reference_active":true,"delta_db":2.677413120545407}],"active_band_count":3,"active_band_mae_db":1.421714319245688}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.1690971851348877,"predicted_mos":3.1690971851348877,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":171520,"input_sample_rate":44100,"input_seconds":3.889342403628118,"input_float32_sha256":"a5bb421e39e468e3797203229e63aa74290a032a5fe3fbe72708da3fc04b4bad","internal_samples":62230,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.60887994299992},"started_at":"2026-10-13T07:30:38.945251+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004867307001404697,"native_input_samples":171108,"native_input_sha256":"4dc4bcd31ce04b677a692dbf69a7ca225e840b75f52708b2d6a3bcb5c763aec2","output_sha256":"d3ea0ec77084974ad4fcccbb82f33191542f8368120923f3ee7c0f76fe0e967f","output_samples":171520,"length_delta_samples":412,"analysis_s":0.08849775999988196,"synthesis_s":20.597511865998968,"total_s":20.68600962599885,"rtf":5.331445779896611,"output_peak":0.6363667249679565,"output_rms":0.06373753134241317,"finished_at":"2026-10-13T07:30:59.647519+00:00","evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"d3ea0ec77084974ad4fcccbb82f33191542f8368120923f3ee7c0f76fe0e967f","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":62080,"output_samples":62230,"length_delta_samples":150,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.907396337136817,"output_peak":0.6346950531486597},"common_prefix":{"samples":62080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975895870159087,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.914597844420275,"spectral_convergence":0.359121145985799,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.423018910136792,"spectral_convergence":0.2903402324858519,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.933528685459637,"spectral_convergence":0.3400768583709473,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0007884091666464782,"delta_db":-0.0231603844396153,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.0026378955175493865,"delta_db":-1.0104487798323736,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.0005081565137741134,"delta_db":0.4371047180842542,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":5.8890302406896e-05,"delta_db":-3.2971166860717522,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":2.9423187137145965e-05,"delta_db":-0.7391639333253639,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":5.392228731234184e-05,"delta_db":-0.9485999327080508,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0759324057435684,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7173091734267587,"pitch":null,"output_rms_dbfs":-23.896915678337884,"output_peak":0.6346950531486597,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.423715146672235},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":62080,"output_samples":62080,"intelligibility":{"estoi":{"value":0.7167926306161608,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06934572208561864,"rms_dbfs":-23.179606504911124,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.836701063786602,"mean":-2.484469069648035e-06,"zero_energy":false},"output":{"rms":0.06384901710256577,"rms_dbfs":-23.896915678337884,"peak":0.6346950531486597,"peak_dbfs":-3.948697731802238,"crest_factor_db":19.948217946535646,"mean":-0.0003283064926358218,"zero_energy":false},"rms_difference_db":-0.7173091734267594},"full_scale":{"reference":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":388,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":89,"quiet_samples":14240,"reference_mean_power":2.407837656504485e-06,"output_mean_power":1.295721598417416e-06,"output_rms_dbfs":-58.874883019063574,"output_zero_energy":false,"output_peak":0.00818474207995618,"output_energy_fraction":7.290571054333044e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":381,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":102,"both_estimated_voiced_frames":82,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.8913043478260869,"absolute_difference_cents_median":10.385394981107714,"absolute_difference_cents_p95":34.83776502293476,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":62080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.00032928880743725145,"log_rms_mae_db":2.081339105091227,"rms_envelope_nrmse":0.11285759078153514,"envelope_correlation":0.9945666964471446,"rms_gain_error_db":-0.7173112880330388,"reference_mean_power":0.004815032927514685,"output_mean_power":0.004081954247083573},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.902918899290217e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008077922828075312,"output_power":0.000642203677391242,"reference_active":true,"delta_db":-0.9962691175119133},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023728667544934464,"output_power":0.00020881952594931828,"reference_active":true,"delta_db":-0.5550224609676783},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1018322593991373e-05,"output_power":1.4605520987407996e-05,"reference_active":true,"delta_db":-1.58081000280861}],"active_band_count":3,"active_band_mae_db":1.0440338604294006}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0003041508818572066,"log_rms_mae_db":1.7351921596828213,"rms_envelope_nrmse":0.09262354281415677,"envelope_correlation":0.9980660080421299,"rms_gain_error_db":-0.7173057724495304,"reference_mean_power":0.004852582666148238,"output_mean_power":0.0041137923411943755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.7242336092637012e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007426996521093885,"output_power":0.0005908812423956965,"reference_active":true,"delta_db":-0.9931301702430068},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010890208813537604,"output_power":9.555005764991241e-05,"reference_active":true,"delta_db":-0.56805253762687},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4730800018618485e-06,"output_power":1.2011450573733369e-06,"reference_active":true,"delta_db":-0.8863087515077573}],"active_band_count":3,"active_band_mae_db":0.8158304864592113}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.0002965570632263201,"log_rms_mae_db":2.2332040406241496,"rms_envelope_nrmse":0.10972412227356763,"envelope_correlation":0.9933101627351223,"rms_gain_error_db":-0.42859526454801516,"reference_mean_power":0.002818219366063788,"output_mean_power":0.002553378919033507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.2520173458640518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044507178368220307,"output_power":0.0004364993256908332,"reference_active":true,"delta_db":-0.08446484949509878},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017793142616898364,"output_power":0.0001814104431327541,"reference_active":true,"delta_db":0.0840962442568714},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6359305994808858e-05,"output_power":1.3347491752529389e-05,"reference_active":true,"delta_db":-0.8836521462353296}],"active_band_count":3,"active_band_mae_db":0.3507377466624333}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0002772405007925501,"log_rms_mae_db":1.6635483839359404,"rms_envelope_nrmse":0.0780820323605741,"envelope_correlation":0.9972857569107442,"rms_gain_error_db":-0.42859551209170776,"reference_mean_power":0.0028402077631144488,"output_mean_power":0.002573300823199064},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.1541839932451843e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343143035723023,"output_power":0.0004152762265629673,"reference_active":true,"delta_db":0.019335796626112},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.103778868784843e-05,"output_power":7.932787811037397e-05,"reference_active":true,"delta_db":-0.09261743956157387},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3787553243482724e-06,"output_power":1.3834755538441303e-06,"reference_active":true,"delta_db":0.01484286947168202}],"active_band_count":3,"active_band_mae_db":0.04226536855312263}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":3.3115033351626755,"rms_envelope_nrmse":0.3893609061063958,"envelope_correlation":0.9389800355293921,"rms_gain_error_db":-3.070383321516308,"reference_mean_power":0.00014490237542137563,"output_mean_power":7.145574850866956e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":6.946266094979906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.029476681448174e-05,"output_power":1.3812126854568598e-05,"reference_active":true,"delta_db":-3.4110705560287773},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.606273878647887e-06,"output_power":4.641574672974908e-06,"reference_active":true,"delta_db":-2.6814982086833083},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.077953085521738e-06,"output_power":6.48406100326823e-07,"reference_active":true,"delta_db":-2.207527678564593}],"active_band_count":3,"active_band_mae_db":2.7666988144255598}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":2.6899073930447712,"rms_envelope_nrmse":0.36864500175047965,"envelope_correlation":0.9532444734255561,"rms_gain_error_db":-3.070385888815994,"reference_mean_power":0.00014603285132237453,"output_mean_power":7.201317783263878e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":6.454018573776752e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.924628617930061e-05,"output_power":1.338318139825524e-05,"reference_active":true,"delta_db":-3.395113608026131},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6077467608232813e-06,"output_power":1.6216583761378485e-06,"reference_active":true,"delta_db":-3.4727667582064967},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.60530589260614e-08,"output_power":7.040845520394743e-08,"reference_active":true,"delta_db":-0.8714149700493372}],"active_band_count":3,"active_band_mae_db":2.5797651120939884}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":3.7049300068179547,"rms_envelope_nrmse":0.31249190957433637,"envelope_correlation":0.932638787166055,"rms_gain_error_db":-0.043683476636285204,"reference_mean_power":5.235421681368183e-05,"output_mean_power":5.183025189774241e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":3.452458222743894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.744222722260822e-05,"output_power":1.6257343410407906e-05,"reference_active":true,"delta_db":-0.30552360206643653},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.482818619951916e-06,"output_power":4.738893462287602e-06,"reference_active":true,"delta_db":-1.983882730618938},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0100258421135403e-06,"output_power":7.950672213001125e-07,"reference_active":true,"delta_db":-1.0392863669312347}],"active_band_count":3,"active_band_mae_db":1.1095642332055364}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":3.34398349040775,"rms_envelope_nrmse":0.26008204777700683,"envelope_correlation":0.9501569158696908,"rms_gain_error_db":-0.04368206098398792,"reference_mean_power":5.276267962923642e-05,"output_mean_power":5.223464381361622e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":3.130411452597277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7491368881094e-05,"output_power":1.5709539317628833e-05,"reference_active":true,"delta_db":-0.4666034932045801},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.132236890482276e-06,"output_power":2.4585659240773232e-06,"reference_active":true,"delta_db":-2.255033525731239},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.718357196867295e-08,"output_power":5.5863727299922083e-08,"reference_active":true,"delta_db":-0.8013318136200175}],"active_band_count":3,"active_band_mae_db":1.1743229441852787}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"golf","status":"measured","mos":{"status":"measured","value":2.3408029079437256,"predicted_mos":2.3408029079437256,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":92880,"input_sample_rate":24000,"input_seconds":3.87,"input_float32_sha256":"eaebe1f14d6f9e94493d9574f6b33119b869fa79faf71c8fedeb92ee2cd9bc70","internal_samples":61920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.044382702999428},"started_at":"2026-10-13T05:03:05.747159+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0023195990006570355,"native_input_samples":93120,"native_input_sha256":"cb1b0ca70ffd6df43fb26504c94600dde285ad9b23b0d5ea84d60e92241b1939","output_sha256":"6dbe3b41b3d40f485aef5a1abba6b62e57a0579e594e954e6538c1e13302d076","output_samples":92880,"length_delta_samples":-240,"analysis_s":0.3820246479999696,"synthesis_s":0.1022701430001689,"total_s":0.4842947910001385,"rtf":0.12481824510312849,"output_peak":0.7079958915710449,"output_rms":0.06640322839360947,"finished_at":"2026-10-13T05:03:06.239384+00:00","analysis_stages":{"external_f0_s":0.03946682300011162},"evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"6dbe3b41b3d40f485aef5a1abba6b62e57a0579e594e954e6538c1e13302d076","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":62080,"output_samples":61920,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.552103074621744,"output_peak":0.7083843234457385},"common_prefix":{"samples":61920,"sample_rate":16000,"reference_coverage":0.9974226804123711,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":61920,"output_samples":61920,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.8983879956950265,"spectral_convergence":0.6736256053040471,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.143320173536978,"spectral_convergence":0.26594927008055025,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.158674957023966,"spectral_convergence":0.25555743776698164,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007952563845350455,"output_power":0.0006918354660535164,"delta_db":-0.6050434263198231,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003340578941928566,"output_power":0.0030677011568849104,"delta_db":-0.3700868900100728,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0004571240511090399,"output_power":0.0004904917677725601,"delta_db":0.3059765089824629,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012606559664849712,"output_power":8.976989455320996e-05,"delta_db":-1.4746586855854273,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.4973819073334e-05,"output_power":2.3653307494825916e-05,"delta_db":-1.6985118132769106,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.725500628197847e-05,"output_power":5.0115089884744616e-05,"delta_db":-1.2775610396161488,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.955306393965141,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3837030159905549,"pitch":null,"output_rms_dbfs":-23.552103074621744,"output_peak":0.7083843234457385,"reference_rms_dbfs":-23.16840005863119,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.733461042085324},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":61920,"output_samples":61920,"intelligibility":{"estoi":{"value":0.8709270109556978,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.0694352489703642,"rms_dbfs":-23.16840005863119,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.825494617506667,"mean":-2.135540471828569e-06,"zero_energy":false},"output":{"rms":0.06643467981893018,"rms_dbfs":-23.552103074621744,"peak":0.7083843234457385,"peak_dbfs":-2.9946211663626814,"crest_factor_db":20.557481908259064,"mean":0.001001576899471031,"zero_energy":false},"rms_difference_db":-0.38370301599055523},"full_scale":{"reference":{"sample_count":61920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":61920,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":387,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":88,"quiet_samples":14080,"reference_mean_power":2.4295388191769067e-06,"output_mean_power":2.3872999452496444e-06,"output_rms_dbfs":-56.220930119177964,"output_zero_energy":false,"output_peak":0.009439204858073182,"output_energy_fraction":0.0001229954377932962,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":380,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":99,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9565217391304348,"absolute_difference_cents_median":4.876701663007349,"absolute_difference_cents_p95":22.956397075409964,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":61920,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":773,"rms_floor":0.00032928880743725145,"log_rms_mae_db":2.201461648982558,"rms_envelope_nrmse":0.12068489662671349,"envelope_correlation":0.9897114101543507,"rms_gain_error_db":-0.3837022534376394,"reference_mean_power":0.004827489044723905,"output_mean_power":0.004419275448965252},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":773,"frequency_resolution_hz":0.258732212160414,"power_floor":1.91232251889731e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008158215245456514,"output_power":0.0007626132100401023,"reference_active":true,"delta_db":-0.2929083562760785},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023432505290094745,"output_power":0.00020509533079908748,"reference_active":true,"delta_db":-0.5786295041268424},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.105914326853773e-05,"output_power":2.4669374661531502e-05,"reference_active":true,"delta_db":0.6871744162473847}],"active_band_count":3,"active_band_mae_db":0.5195707588834352}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":767,"rms_floor":0.0003041508818572066,"log_rms_mae_db":1.4814122125109284,"rms_envelope_nrmse":0.07530076456413441,"envelope_correlation":0.9969450272601468,"rms_gain_error_db":-0.38371004933500674,"reference_mean_power":0.004865233584769988,"output_mean_power":0.004453820306997427},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":767,"frequency_resolution_hz":0.2607561929595828,"power_floor":1.7329734937090006e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007485877193993708,"output_power":0.0007027911841089553,"reference_active":true,"delta_db":-0.274163932058119},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010705252754063112,"output_power":9.161675808985064e-05,"reference_active":true,"delta_db":-0.6762200560860089},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4775450757847758e-06,"output_power":1.9037089004934024e-06,"reference_active":true,"delta_db":1.1005980151490882}],"active_band_count":3,"active_band_mae_db":0.6836606677644053}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":773,"rms_floor":0.0002965570632263201,"log_rms_mae_db":1.5027959843626857,"rms_envelope_nrmse":0.12354362356631375,"envelope_correlation":0.9901247208308419,"rms_gain_error_db":-0.155172332687871,"reference_mean_power":0.0028255109197897156,"output_mean_power":0.002726338376276599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":773,"frequency_resolution_hz":0.258732212160414,"power_floor":1.2596344426742667e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045102350889243687,"output_power":0.0004444283539158355,"reference_active":true,"delta_db":-0.06397420844606051},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017576726622237752,"output_power":0.00016242601008094597,"reference_active":true,"delta_db":-0.34282421827231097},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6409564951145312e-05,"output_power":1.6057455763827397e-05,"reference_active":true,"delta_db":-0.09420332967168735}],"active_band_count":3,"active_band_mae_db":0.16700058546335295}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":767,"rms_floor":0.0002772405007925501,"log_rms_mae_db":1.082499106312694,"rms_envelope_nrmse":0.07351518163891593,"envelope_correlation":0.9965627113350698,"rms_gain_error_db":-0.15517223563416355,"reference_mean_power":0.002847613688188943,"output_mean_power":0.0027476654215137578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":767,"frequency_resolution_hz":0.2607561929595828,"power_floor":1.1613297687033284e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041773131996964764,"output_power":0.00041280623696689456,"reference_active":true,"delta_db":-0.05150788005805999},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.965886580301933e-05,"output_power":7.44061677843939e-05,"reference_active":true,"delta_db":-0.2962518134733201},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3809234791659568e-06,"output_power":1.6176094569167996e-06,"reference_active":true,"delta_db":0.6870406346683924}],"active_band_count":3,"active_band_mae_db":0.3449334427332575}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":773,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":1.961070924642256,"rms_envelope_nrmse":0.23639634610854768,"envelope_correlation":0.9769626805134229,"rms_gain_error_db":-1.5945032835721853,"reference_mean_power":0.00014527725576422777,"output_mean_power":0.00010063459409510118},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":773,"frequency_resolution_hz":0.258732212160414,"power_floor":6.988738951047689e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0499937254931578e-05,"output_power":2.19467130648648e-05,"reference_active":true,"delta_db":-1.429294602682},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.554119354761996e-06,"output_power":5.9334250228900946e-06,"reference_active":true,"delta_db":-1.588698468114465},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0745330415127532e-06,"output_power":9.68328658941013e-07,"reference_active":true,"delta_db":-0.45196988883979083}],"active_band_count":3,"active_band_mae_db":1.1566543198787518}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":767,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":1.4993984481149412,"rms_envelope_nrmse":0.19605364613507306,"envelope_correlation":0.991005352012177,"rms_gain_error_db":-1.5945052805373114,"reference_mean_power":0.00014641361516758592,"output_mean_power":0.00010142171174552239},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":767,"frequency_resolution_hz":0.2607561929595828,"power_floor":6.495057562843701e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.9391681014583505e-05,"output_power":2.136157857218655e-05,"reference_active":true,"delta_db":-1.3859108270702272},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.571152825334589e-06,"output_power":2.32918987949148e-06,"reference_active":true,"delta_db":-1.8560354156400427},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.506685850998368e-08,"output_power":7.670951202074742e-08,"reference_active":true,"delta_db":-0.44911174716751906}],"active_band_count":3,"active_band_mae_db":1.2303526632925965}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":773,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":2.504809132535698,"rms_envelope_nrmse":0.2651550340370213,"envelope_correlation":0.959293540951612,"rms_gain_error_db":-1.2050615121930872,"reference_mean_power":5.248966775594703e-05,"output_mean_power":3.977110637499506e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":773,"frequency_resolution_hz":0.258732212160414,"power_floor":3.4650106346011364e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7542503521421568e-05,"output_power":1.4972196995061429e-05,"reference_active":true,"delta_db":-0.688060397170204},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.425862244761668e-06,"output_power":6.892629626812338e-06,"reference_active":true,"delta_db":-0.3236194603102461},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.006516422845161e-06,"output_power":1.75553928060977e-06,"reference_active":true,"delta_db":2.4158968589561103}],"active_band_count":3,"active_band_mae_db":1.14252557214552}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":61920,"uncovered_tail_samples":0,"metrics":{"frame_count":767,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":2.125846329152971,"rms_envelope_nrmse":0.18315486653648677,"envelope_correlation":0.9865246300984212,"rms_gain_error_db":-1.2050616245717702,"reference_mean_power":5.290025449461735e-05,"output_mean_power":4.0082204446616964e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":767,"frequency_resolution_hz":0.2607561929595828,"power_floor":3.1417031127535694e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.758446217626081e-05,"output_power":1.588672022336024e-05,"reference_active":true,"delta_db":-0.4409484251625146},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.062573203707985e-06,"output_power":3.8113696343243833e-06,"reference_active":true,"delta_db":-0.2772013006062454},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.695514445914984e-08,"output_power":1.5425025582598489e-07,"reference_active":true,"delta_db":3.6244194159757743}],"active_band_count":3,"active_band_mae_db":1.4475230472481782}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.8586432933807373,"predicted_mos":2.8586432933807373,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":93240,"input_sample_rate":24000,"input_seconds":3.885,"input_float32_sha256":"4d0c22d59dcce9478d0a44403b0fb70380bd3e1d16d77defb4ad88250db21689","internal_samples":62160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1463619469977857},"started_at":"2026-10-13T07:46:07.740636+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002150182997866068,"native_input_samples":93120,"native_input_sha256":"cb1b0ca70ffd6df43fb26504c94600dde285ad9b23b0d5ea84d60e92241b1939","output_sha256":"3208591b3a9f0f7deec8a47141b1fce27e6a7fc580d25f910b2b644fbf76ca58","output_samples":93240,"length_delta_samples":120,"analysis_s":1.566158618999907,"synthesis_s":9.040532708000683,"total_s":10.60669132700059,"rtf":2.733683331701183,"output_peak":0.6063663363456726,"output_rms":0.07024041986754778,"finished_at":"2026-10-13T07:46:18.357602+00:00","analysis_stages":{"external_f0_s":0.9843782329990063},"evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"3208591b3a9f0f7deec8a47141b1fce27e6a7fc580d25f910b2b644fbf76ca58","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":62080,"output_samples":62160,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.06496569076337,"output_peak":0.6068611361693583},"common_prefix":{"samples":62080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987129987129987,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.685583737269269,"spectral_convergence":0.5282952379202622,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.168964560478275,"spectral_convergence":0.28331544419613874,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.071157031015701,"spectral_convergence":0.3024307199780374,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0011961577492903214,"delta_db":1.787207546682912,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.0031424336554509596,"delta_db":-0.2503635046183037,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.00041140467190546473,"delta_db":-0.4801780589147356,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":0.00011636784333583379,"delta_db":-0.3392246752899387,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":3.1944659731178275e-05,"delta_db":-0.38207840843523977,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":4.5509511558557275e-05,"delta_db":-1.685261250235632,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8207189073627936,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.12023368984703615,"pitch":null,"output_rms_dbfs":-23.05937281506409,"output_peak":0.6068611361693583,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.641901776254415},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":62080,"output_samples":62080,"intelligibility":{"estoi":{"value":0.8771076215059516,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06934572208561864,"rms_dbfs":-23.179606504911124,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.836701063786602,"mean":-2.484469069648035e-06,"zero_energy":false},"output":{"rms":0.07031230886947196,"rms_dbfs":-23.05937281506409,"peak":0.6068611361693583,"peak_dbfs":-4.338213483099954,"crest_factor_db":18.721159331964135,"mean":-9.212017130527063e-05,"zero_energy":false},"rms_difference_db":0.12023368984703353},"full_scale":{"reference":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":388,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":89,"quiet_samples":14240,"reference_mean_power":2.407837656504485e-06,"output_mean_power":1.9032620410935123e-06,"output_rms_dbfs":-57.20501413937684,"output_zero_energy":false,"output_peak":0.03425888118486796,"output_energy_fraction":8.830680024968148e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":381,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":116,"both_estimated_voiced_frames":86,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":30,"paired_voiced_fraction":0.9347826086956522,"absolute_difference_cents_median":7.253907118822955,"absolute_difference_cents_p95":17.652540542247785,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":62080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.00032928880743725145,"log_rms_mae_db":3.4502015368195993,"rms_envelope_nrmse":0.14489868852317642,"envelope_correlation":0.9833994589776107,"rms_gain_error_db":0.12023461591331638,"reference_mean_power":0.004815032927514685,"output_mean_power":0.004950199739504654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.902918899290217e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008077922828075312,"output_power":0.0008748519898490679,"reference_active":true,"delta_db":0.34634884111608555},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023728667544934464,"output_power":0.0002210052159160616,"reference_active":true,"delta_db":-0.30870828077607504},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1018322593991373e-05,"output_power":1.4872099322529821e-05,"reference_active":true,"delta_db":-1.5022577621660036}],"active_band_count":3,"active_band_mae_db":0.7191049613527213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0003041508818572066,"log_rms_mae_db":2.815594238859339,"rms_envelope_nrmse":0.09605836401854201,"envelope_correlation":0.992366826858135,"rms_gain_error_db":0.12025028470173146,"reference_mean_power":0.004852582666148238,"output_mean_power":0.004988821567218519},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.7242336092637012e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007426996521093885,"output_power":0.0008008048592221316,"reference_active":true,"delta_db":0.32713479258475975},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010890208813537604,"output_power":0.00010323163256974687,"reference_active":true,"delta_db":-0.2322341158432585},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4730800018618485e-06,"output_power":1.110529226228428e-06,"reference_active":true,"delta_db":-1.2269634115705133}],"active_band_count":3,"active_band_mae_db":0.5954441066661772}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.0002965570632263201,"log_rms_mae_db":2.282766169998267,"rms_envelope_nrmse":0.16305694396227816,"envelope_correlation":0.9828870162722668,"rms_gain_error_db":-0.32599403692289636,"reference_mean_power":0.002818219366063788,"output_mean_power":0.0026144202051393258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.2520173458640518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044507178368220307,"output_power":0.00040579443197229985,"reference_active":true,"delta_db":-0.4012397843578244},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017793142616898364,"output_power":0.00014378462686121543,"reference_active":true,"delta_db":-0.9254020504589666},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6359305994808858e-05,"output_power":1.3102839589604196e-05,"reference_active":true,"delta_db":-0.9639945156088943}],"active_band_count":3,"active_band_mae_db":0.763545450141895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0002772405007925501,"log_rms_mae_db":1.5144159920455493,"rms_envelope_nrmse":0.10802664697695093,"envelope_correlation":0.9927410431057213,"rms_gain_error_db":-0.32599384008397414,"reference_mean_power":0.0028402077631144488,"output_mean_power":0.002634818633583939},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.1541839932451843e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343143035723023,"output_power":0.0003767941701448698,"reference_active":true,"delta_db":-0.4029931532008535},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.103778868784843e-05,"output_power":6.631350342091098e-05,"reference_active":true,"delta_db":-0.8708560892824129},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3787553243482724e-06,"output_power":8.078126736406345e-07,"reference_active":true,"delta_db":-2.3217654010690616}],"active_band_count":3,"active_band_mae_db":1.1985382145174428}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":2.340989495488154,"rms_envelope_nrmse":0.1861502670147987,"envelope_correlation":0.976190142021732,"rms_gain_error_db":-0.3435279138976865,"reference_mean_power":0.00014490237542137563,"output_mean_power":0.00013388216097647265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":6.946266094979906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.029476681448174e-05,"output_power":2.7518273055686517e-05,"reference_active":true,"delta_db":-0.41746438270095687},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.606273878647887e-06,"output_power":7.778268489826657e-06,"reference_active":true,"delta_db":-0.43932232557760154},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.077953085521738e-06,"output_power":1.1701050105334317e-06,"reference_active":true,"delta_db":0.3562497907625456}],"active_band_count":3,"active_band_mae_db":0.404345499680368}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":1.3635565858621639,"rms_envelope_nrmse":0.111082042026421,"envelope_correlation":0.9918200941921475,"rms_gain_error_db":-0.3435284056472976,"reference_mean_power":0.00014603285132237453,"output_mean_power":0.00013492664587428776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":6.454018573776752e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.924628617930061e-05,"output_power":2.6730597459102265e-05,"reference_active":true,"delta_db":-0.39062059477889044},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6077467608232813e-06,"output_power":3.313910636975195e-06,"reference_active":true,"delta_db":-0.36895252392229155},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.60530589260614e-08,"output_power":7.431010740039339e-08,"reference_active":true,"delta_db":-0.6371842380196286}],"active_band_count":3,"active_band_mae_db":0.4655857855736036}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":2.308475414842099,"rms_envelope_nrmse":0.27373447885958707,"envelope_correlation":0.9558838047912769,"rms_gain_error_db":-1.308203606050975,"reference_mean_power":5.235421681368183e-05,"output_mean_power":3.873747478863161e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":3.452458222743894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.744222722260822e-05,"output_power":1.442906666893174e-05,"reference_active":true,"delta_db":-0.8236369975470659},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.482818619951916e-06,"output_power":6.191484823116391e-06,"reference_active":true,"delta_db":-0.8227040555757453},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0100258421135403e-06,"output_power":6.862558662593698e-07,"reference_active":true,"delta_db":-1.6784641569360375}],"active_band_count":3,"active_band_mae_db":1.1082684033529495}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":1.5180797742529106,"rms_envelope_nrmse":0.20987912168294162,"envelope_correlation":0.9794986692270473,"rms_gain_error_db":-1.308204795675506,"reference_mean_power":5.276267962923642e-05,"output_mean_power":3.903969033337881e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":3.130411452597277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7491368881094e-05,"output_power":1.4468073334417975e-05,"reference_active":true,"delta_db":-0.8241309745142735},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.132236890482276e-06,"output_power":3.3754043255562425e-06,"reference_active":true,"delta_db":-0.8785940794239491},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.718357196867295e-08,"output_power":4.855159267471779e-08,"reference_active":true,"delta_db":-1.4105960941259756}],"active_band_count":3,"active_band_mae_db":1.0377737160213993}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.782327651977539,"predicted_mos":2.782327651977539,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":93360,"input_sample_rate":24000,"input_seconds":3.89,"input_float32_sha256":"1868b523dd172493d22e334cdd406c946e51f56a3a315fdceb448f4d3d07d2aa","internal_samples":62240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.427336804001243},"started_at":"2026-10-13T05:11:32.801445+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002181014000598225,"native_input_samples":93120,"native_input_sha256":"cb1b0ca70ffd6df43fb26504c94600dde285ad9b23b0d5ea84d60e92241b1939","output_sha256":"5b04f67a326448c3c6f277619019a602d7111010dee8a395873274d6d54d2300","output_samples":93360,"length_delta_samples":240,"analysis_s":0.5376576110011229,"synthesis_s":0.0734260989993345,"total_s":0.6110837100004574,"rtf":0.15749580154650963,"output_peak":0.7518944144248962,"output_rms":0.06677642462142842,"finished_at":"2026-10-13T05:11:33.420054+00:00","analysis_stages":{"external_f0_s":0.07546688199909113},"evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"5b04f67a326448c3c6f277619019a602d7111010dee8a395873274d6d54d2300","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":62080,"output_samples":62240,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.50386092848369,"output_peak":0.7504863124534064},"common_prefix":{"samples":62080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9974293059125964,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.008148655787908,"spectral_convergence":0.19008821504493054,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.395744848052943,"spectral_convergence":0.17401658120978777,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.682210177015957,"spectral_convergence":0.18387244828433874,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0007779943954396298,"delta_db":-0.08091235114092674,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.0031061572338501693,"delta_db":-0.3007903192607467,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.0003908303912045661,"delta_db":-0.7029869776195662,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":0.00011292055400005084,"delta_db":-0.4698245289876255,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":3.2591730038254566e-05,"delta_db":-0.29498692958350825,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":5.387451186926892e-05,"delta_db":-0.9524495113250587,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.46699176965290534,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.31307568836087435,"pitch":null,"output_rms_dbfs":-23.492682193272,"output_peak":0.7504863124534064,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.695367893618936},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":62080,"output_samples":62080,"intelligibility":{"estoi":{"value":0.9291922541755271,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06934572208561864,"rms_dbfs":-23.179606504911124,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.836701063786602,"mean":-2.484469069648035e-06,"zero_energy":false},"output":{"rms":0.0668907230325052,"rms_dbfs":-23.492682193272,"peak":0.7504863124534064,"peak_dbfs":-2.493144482276015,"crest_factor_db":20.999537710995988,"mean":0.0005060026398493613,"zero_energy":false},"rms_difference_db":-0.3130756883608754},"full_scale":{"reference":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":388,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":89,"quiet_samples":14240,"reference_mean_power":2.407837656504485e-06,"output_mean_power":2.008043680485637e-06,"output_rms_dbfs":-56.97226844322063,"output_zero_energy":false,"output_peak":0.00867559628314677,"output_energy_fraction":0.00010294367213855274,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":381,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":85,"both_estimated_voiced_frames":79,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.8586956521739131,"absolute_difference_cents_median":1.0898376719867908,"absolute_difference_cents_p95":5.425270532847908,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":62080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.00032928880743725145,"log_rms_mae_db":1.1660645273444454,"rms_envelope_nrmse":0.0839283022927566,"envelope_correlation":0.9951472181566776,"rms_gain_error_db":-0.31307732579625097,"reference_mean_power":0.004815032927514685,"output_mean_power":0.004480139415266887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.902918899290217e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008077922828075312,"output_power":0.000765668885221047,"reference_active":true,"delta_db":-0.2325870098715532},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023728667544934464,"output_power":0.00019474275007605154,"reference_active":true,"delta_db":-0.8581205297316019},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1018322593991373e-05,"output_power":1.1494070961251236e-05,"reference_active":true,"delta_db":-2.6212417938155137}],"active_band_count":3,"active_band_mae_db":1.2373164444728897}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0003041508818572066,"log_rms_mae_db":0.7069581899468903,"rms_envelope_nrmse":0.054422623612415386,"envelope_correlation":0.9984896172844903,"rms_gain_error_db":-0.31307873065571284,"reference_mean_power":0.004852582666148238,"output_mean_power":0.004515076046867386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.7242336092637012e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007426996521093885,"output_power":0.0007004492177598272,"reference_active":true,"delta_db":-0.2543656586960271},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010890208813537604,"output_power":9.208746013940566e-05,"reference_active":true,"delta_db":-0.7283571230831553},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4730800018618485e-06,"output_power":7.756971051079556e-07,"reference_active":true,"delta_db":-2.785341630227387}],"active_band_count":3,"active_band_mae_db":1.2560214706688564}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.0002965570632263201,"log_rms_mae_db":1.1718924174391725,"rms_envelope_nrmse":0.11387112258013532,"envelope_correlation":0.9927104823582679,"rms_gain_error_db":-0.43901486839052734,"reference_mean_power":0.002818219366063788,"output_mean_power":0.002547260189050624},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.2520173458640518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044507178368220307,"output_power":0.0004143791810508878,"reference_active":true,"delta_db":-0.3103213432297337},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017793142616898364,"output_power":0.00014346327756669153,"reference_active":true,"delta_db":-0.9351191123727489},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6359305994808858e-05,"output_power":8.195417380078518e-06,"reference_active":true,"delta_db":-3.001937993774031}],"active_band_count":3,"active_band_mae_db":1.415792816458838}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0002772405007925501,"log_rms_mae_db":0.792792831094917,"rms_envelope_nrmse":0.0748582116712454,"envelope_correlation":0.9977070760604287,"rms_gain_error_db":-0.4390150117362436,"reference_mean_power":0.0028402077631144488,"output_mean_power":0.00256713441541796},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.1541839932451843e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343143035723023,"output_power":0.0003807135570472487,"reference_active":true,"delta_db":-0.3580514784750921},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.103778868784843e-05,"output_power":6.452920178626912e-05,"reference_active":true,"delta_db":-0.9893128873640299},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3787553243482724e-06,"output_power":5.461626017557609e-07,"reference_active":true,"delta_db":-4.02165243871322}],"active_band_count":3,"active_band_mae_db":1.7896722681841137}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":1.2539836875549437,"rms_envelope_nrmse":0.13361250085754833,"envelope_correlation":0.9889622249856675,"rms_gain_error_db":-0.46012577837326807,"reference_mean_power":0.00014490237542137563,"output_mean_power":0.00013033556147615087},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":6.946266094979906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.029476681448174e-05,"output_power":2.764256390682598e-05,"reference_active":true,"delta_db":-0.39789291673859095},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.606273878647887e-06,"output_power":7.144476900255597e-06,"reference_active":true,"delta_db":-0.8084472618486445},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.077953085521738e-06,"output_power":7.518802762092457e-07,"reference_active":true,"delta_db":-1.5645116768425176}],"active_band_count":3,"active_band_mae_db":0.923617285143251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":0.8177832858546609,"rms_envelope_nrmse":0.08548403387574545,"envelope_correlation":0.996463489372044,"rms_gain_error_db":-0.4601263171954241,"reference_mean_power":0.00014603285132237453,"output_mean_power":0.00013135237606898333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":6.454018573776752e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.924628617930061e-05,"output_power":2.6485321725945555e-05,"reference_active":true,"delta_db":-0.4306547249395187},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6077467608232813e-06,"output_power":3.1746476504631114e-06,"reference_active":true,"delta_db":-0.5554051484570576},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.60530589260614e-08,"output_power":5.735298200284601e-08,"reference_active":true,"delta_db":-1.7621030932341348}],"active_band_count":3,"active_band_mae_db":0.9160543222102371}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":1.5604331251112167,"rms_envelope_nrmse":0.1696671573334121,"envelope_correlation":0.9827596740298624,"rms_gain_error_db":-0.6642718209550893,"reference_mean_power":5.235421681368183e-05,"output_mean_power":4.4928765436368925e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":3.452458222743894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.744222722260822e-05,"output_power":1.5157776135964004e-05,"reference_active":true,"delta_db":-0.6096645109749795},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.482818619951916e-06,"output_power":6.3756916041800015e-06,"reference_active":true,"delta_db":-0.6953791642845314},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0100258421135403e-06,"output_power":1.0249204626637992e-06,"reference_active":true,"delta_db":0.06357678352922497}],"active_band_count":3,"active_band_mae_db":0.45620681959624526}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":1.081856730687362,"rms_envelope_nrmse":0.11626002554668698,"envelope_correlation":0.993620839655549,"rms_gain_error_db":-0.6642714237494506,"reference_mean_power":5.276267962923642e-05,"output_mean_power":4.52792997000445e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":3.130411452597277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7491368881094e-05,"output_power":1.5321580180864087e-05,"reference_active":true,"delta_db":-0.5752024058218694},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.132236890482276e-06,"output_power":3.494435533453388e-06,"reference_active":true,"delta_db":-0.7280817767366237},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.718357196867295e-08,"output_power":5.947085519386061e-08,"reference_active":true,"delta_db":-0.5295890666110229}],"active_band_count":3,"active_band_mae_db":0.610957749723172}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0032","method":"magphase","status":"measured","mos":{"status":"measured","value":3.366032600402832,"predicted_mos":3.366032600402832,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":187619,"input_sample_rate":48000,"input_seconds":3.908729166666667,"input_float32_sha256":"a63630f7ed02780e81801c8fe0ad8c3d767e453bb827f5e925e604aa4786565f","internal_samples":62540,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.896883584000534},"started_at":"2026-10-13T05:18:37.461843+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.01133677899997565,"native_input_samples":186240,"native_input_sha256":"34a74ff0a3303d031cc36937f1f5862355accc94451f95e15ad2af66404a0147","output_sha256":"d0d99ebda66a05a45c3c4e5c81ad3518c923a9c9f2bbe871ae812b9e6e7ef85a","output_samples":187619,"length_delta_samples":1379,"analysis_s":1.8493581319999066,"synthesis_s":1.0108937000004516,"total_s":2.8602518320003583,"rtf":0.7371783072165872,"output_peak":0.7866519762230403,"output_rms":0.06666572421122365,"finished_at":"2026-10-13T05:18:40.343092+00:00","evaluation_input":{"reference":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"d0d99ebda66a05a45c3c4e5c81ad3518c923a9c9f2bbe871ae812b9e6e7ef85a","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":62080,"output_samples":62540,"length_delta_samples":460,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.518343456161496,"output_peak":0.7838141843795339},"common_prefix":{"samples":62080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9926447073872722,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":62080,"output_samples":62080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.107134193041537,"spectral_convergence":0.6348561573582265,"spectral_convergence_status":"ok","amplitude_floor":2.580715109285093e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.828893921798399,"spectral_convergence":0.30983274661971727,"spectral_convergence_status":"ok","amplitude_floor":1.9335078654426385e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.019192088789634,"spectral_convergence":0.2565207227066412,"spectral_convergence_status":"ok","amplitude_floor":1.5577782358148932e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0007926248856121452,"output_power":0.0007191417026557079,"delta_db":-0.4225322952356858,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.003328913174136213,"output_power":0.0030703978366320387,"delta_db":-0.3510781649315916,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00045950159700284816,"output_power":0.00046664534078960754,"delta_db":0.06699908862100096,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00012582167228939335,"output_power":0.00012484668454675616,"delta_db":-0.03378439442674136,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":3.488237987418478e-05,"output_power":3.4266616032779394e-05,"delta_db":-0.07734889101824402,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.708544414240077e-05,"output_power":6.566340953778705e-05,"delta_db":-0.09304869898272425,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.17413192220266463,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3066812004044642,"pitch":null,"output_rms_dbfs":-23.486287705315586,"output_peak":0.7838141843795339,"reference_rms_dbfs":-23.179606504911124,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.985073401209856},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":62080,"output_samples":62080,"intelligibility":{"estoi":{"value":0.7584833379351008,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06934572208561864,"rms_dbfs":-23.179606504911124,"peak":0.763580322265625,"peak_dbfs":-2.3429054411245227,"crest_factor_db":20.836701063786602,"mean":-2.484469069648035e-06,"zero_energy":false},"output":{"rms":0.0669399856211226,"rms_dbfs":-23.486287705315586,"peak":0.7838141843795339,"peak_dbfs":-2.115737630579577,"crest_factor_db":21.370550074736016,"mean":0.0012562297844937547,"zero_energy":false},"rms_difference_db":-0.3066812004044621},"full_scale":{"reference":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":62080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":388,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0250761726638303e-05,"threshold_dbfs":-49.892438613024545,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":89,"quiet_samples":14240,"reference_mean_power":2.407837656504485e-06,"output_mean_power":1.1544676853535393e-05,"output_rms_dbfs":-49.376182188915024,"output_zero_energy":false,"output_peak":0.07346838016119477,"output_energy_fraction":0.0005909746235689668,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":381,"reference_estimated_voiced_frames":92,"output_estimated_voiced_frames":97,"both_estimated_voiced_frames":84,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9130434782608695,"absolute_difference_cents_median":6.586660380681783,"absolute_difference_cents_p95":21.37079631059252,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":62080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.00032928880743725145,"log_rms_mae_db":3.2787413081768135,"rms_envelope_nrmse":0.20092860239799246,"envelope_correlation":0.9673572609190871,"rms_gain_error_db":-0.30668048146180976,"reference_mean_power":0.004815032927514685,"output_mean_power":0.0044867431995966895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.902918899290217e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008077922828075312,"output_power":0.000780881544811345,"reference_active":true,"delta_db":-0.14714540922333566},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00023728667544934464,"output_power":0.0002252088136799282,"reference_active":true,"delta_db":-0.22687968729703545},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.1018322593991373e-05,"output_power":1.7619362434325592e-05,"reference_active":true,"delta_db":-0.7660786421130216}],"active_band_count":3,"active_band_mae_db":0.3800345795444642}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0003041508818572066,"log_rms_mae_db":2.3150572406600327,"rms_envelope_nrmse":0.13616644271401995,"envelope_correlation":0.984692448398778,"rms_gain_error_db":-0.30666622902968005,"reference_mean_power":0.004852582666148238,"output_mean_power":0.004521747630137446},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.7242336092637012e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007426996521093885,"output_power":0.0007178798590547932,"reference_active":true,"delta_db":-0.14761451613432508},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010890208813537604,"output_power":0.00010749237539776918,"reference_active":true,"delta_db":-0.05658547026832317},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4730800018618485e-06,"output_power":1.4352349847776975e-06,"reference_active":true,"delta_db":-0.11303321651368659}],"active_band_count":3,"active_band_mae_db":0.10574440097211162}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":0.0002965570632263201,"log_rms_mae_db":3.5715861996246145,"rms_envelope_nrmse":0.2118888666993106,"envelope_correlation":0.9703871355142103,"rms_gain_error_db":-0.35011872148526957,"reference_mean_power":0.002818219366063788,"output_mean_power":0.002599937588075556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":1.2520173458640518e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00044507178368220307,"output_power":0.0004317573023798705,"reference_active":true,"delta_db":-0.13190370495313647},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017793142616898364,"output_power":0.00016949458241904538,"reference_active":true,"delta_db":-0.21096838415069485},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6359305994808858e-05,"output_power":1.3782661471579174e-05,"reference_active":true,"delta_db":-0.7443178661760109}],"active_band_count":3,"active_band_mae_db":0.36239665175994745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":0.0002772405007925501,"log_rms_mae_db":2.6683579499086942,"rms_envelope_nrmse":0.13719666242441275,"envelope_correlation":0.9876863912692563,"rms_gain_error_db":-0.3501186306374707,"reference_mean_power":0.0028402077631144488,"output_mean_power":0.002620222955194806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":1.1541839932451843e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041343143035723023,"output_power":0.00040179701471042186,"reference_active":true,"delta_db":-0.12396784294852996},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.103778868784843e-05,"output_power":7.920211704864955e-05,"reference_active":true,"delta_db":-0.09950791433231193},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3787553243482724e-06,"output_power":1.2387113136092273e-06,"reference_active":true,"delta_db":-0.46517098384431266}],"active_band_count":3,"active_band_mae_db":0.22954891370838484}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":6.247605344635174e-05,"log_rms_mae_db":3.5352796922199876,"rms_envelope_nrmse":0.2671879172865702,"envelope_correlation":0.9509808482129556,"rms_gain_error_db":-0.002164752743216835,"reference_mean_power":0.00014490237542137563,"output_mean_power":0.0001448301664332434},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":6.946266094979906e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.029476681448174e-05,"output_power":3.0894047755093146e-05,"reference_active":true,"delta_db":0.08507199546026434},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.606273878647887e-06,"output_power":9.20056810662302e-06,"reference_active":true,"delta_db":0.2899948193446603},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.077953085521738e-06,"output_power":9.980701846253628e-07,"reference_active":true,"delta_db":-0.33438777883755555}],"active_band_count":3,"active_band_mae_db":0.2364848645474934}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":5.585752960561643e-05,"log_rms_mae_db":2.4629030784326438,"rms_envelope_nrmse":0.1654517621508198,"envelope_correlation":0.98024032119598,"rms_gain_error_db":-0.0021644874931597427,"reference_mean_power":0.00014603285132237453,"output_mean_power":0.0001459600879005154},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":6.454018573776752e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.924628617930061e-05,"output_power":2.9941676600763237e-05,"reference_active":true,"delta_db":0.10205389931916532},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6077467608232813e-06,"output_power":3.9830378376371345e-06,"reference_active":true,"delta_db":0.4297838665926791},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.60530589260614e-08,"output_power":8.203673906341527e-08,"reference_active":true,"delta_db":-0.2075792383397913}],"active_band_count":3,"active_band_mae_db":0.24647233475054522}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":775,"rms_floor":3.703669227124252e-05,"log_rms_mae_db":3.832885034565015,"rms_envelope_nrmse":0.3677220204276432,"envelope_correlation":0.9071214855293321,"rms_gain_error_db":-0.044762185460709314,"reference_mean_power":5.235421681368183e-05,"output_mean_power":5.181737980069446e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":775,"frequency_resolution_hz":0.25806451612903225,"power_floor":3.452458222743894e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.744222722260822e-05,"output_power":1.664830991848441e-05,"reference_active":true,"delta_db":-0.20231787876158977},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.482818619951916e-06,"output_power":7.11466289293155e-06,"reference_active":true,"delta_db":-0.21910891094215418},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0100258421135403e-06,"output_power":1.6957244459793475e-06,"reference_active":true,"delta_db":2.2502279550043554}],"active_band_count":3,"active_band_mae_db":0.8905515815693664}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":62080,"uncovered_tail_samples":0,"metrics":{"frame_count":769,"rms_floor":2.9701530962488805e-05,"log_rms_mae_db":2.745536513921462,"rms_envelope_nrmse":0.2471776654976998,"envelope_correlation":0.9550171574683023,"rms_gain_error_db":-0.044762226502508606,"reference_mean_power":5.276267962923642e-05,"output_mean_power":5.222165376942254e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":769,"frequency_resolution_hz":0.2600780234070221,"power_floor":3.130411452597277e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.7491368881094e-05,"output_power":1.7055364741476888e-05,"reference_active":true,"delta_db":-0.10962787309299533},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.132236890482276e-06,"output_power":3.936535294904275e-06,"reference_active":true,"delta_db":-0.21071060871715558},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.718357196867295e-08,"output_power":1.1996501802096067e-07,"reference_active":true,"delta_db":2.517915332050331}],"active_band_count":3,"active_band_mae_db":0.9460846046201606}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"hnm","status":"measured","mos":{"status":"measured","value":2.673272132873535,"predicted_mos":2.673272132873535,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"50df2e96b6146c85aff4f4922fd4fb7aa308b1054bfdc04eb68fff7874ccd5ff","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.315751281001212},"started_at":"2026-10-13T04:40:23.153847+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":7.005400038906373e-05,"native_input_samples":96000,"native_input_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","output_sha256":"856d99cf8f8252be503cbff64151151e6691aa89be6f704601823fcba30573c6","output_samples":96000,"length_delta_samples":0,"analysis_s":1.13723131200004,"synthesis_s":0.5035303920003571,"total_s":1.640761704000397,"rtf":0.27346028400006617,"output_peak":0.9658179514113898,"output_rms":0.09377645555590701,"finished_at":"2026-10-13T04:40:24.799923+00:00","evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"856d99cf8f8252be503cbff64151151e6691aa89be6f704601823fcba30573c6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.990135014512973,"spectral_convergence":0.3065165426705528,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.507901607214369,"spectral_convergence":0.2892657106999899,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.810623469121124,"spectral_convergence":0.31448447559653886,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.001886621046646034,"delta_db":0.22158417879982845,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.003826828371002896,"delta_db":-0.07299628362595018,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.0025685685863025528,"delta_db":-0.4257783178589948,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.00042388554071307884,"delta_db":-0.22252701633070585,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.403573829714696e-05,"delta_db":0.19127309850106713,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":3.408433366729357e-05,"delta_db":0.271420427361303,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2342632204129749,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12341657414756428,"pitch":null,"output_rms_dbfs":-20.558123724031574,"output_peak":0.9658179514113898,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.769553363616154},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.990135014512973,"spectral_convergence":0.3065165426705528,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.507901607214369,"spectral_convergence":0.2892657106999899,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.810623469121124,"spectral_convergence":0.31448447559653886,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.001886621046646034,"delta_db":0.22158417879982845,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.003826828371002896,"delta_db":-0.07299628362595018,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.0025685685863025528,"delta_db":-0.4257783178589948,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.00042388554071307884,"delta_db":-0.22252701633070585,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.403573829714696e-05,"delta_db":0.19127309850106713,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":3.408433366729357e-05,"delta_db":0.271420427361303,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2342632204129749,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12341657414756428,"pitch":null,"output_rms_dbfs":-20.558123724031574,"output_peak":0.9658179514113898,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.769553363616154},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8603299370347411,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.095118423251492,"rms_dbfs":-20.434707149884012,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.09612477983555,"mean":-3.6449432373046876e-06,"zero_energy":false},"output":{"rms":0.09377645555590701,"rms_dbfs":-20.558123724031574,"peak":0.9658179514113898,"peak_dbfs":-0.30209453480258547,"crest_factor_db":20.256029189228993,"mean":-0.000695082637953008,"zero_energy":false},"rms_difference_db":-0.12341657414756213},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":96,"quiet_samples":15360,"reference_mean_power":1.9068745435409559e-06,"output_mean_power":3.8142494743628324e-06,"output_rms_dbfs":-54.18590904946052,"output_zero_energy":false,"output_peak":0.024413308812468133,"output_energy_fraction":6.939712041983243e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":191,"both_estimated_voiced_frames":179,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9421052631578948,"absolute_difference_cents_median":0.9068122992972697,"absolute_difference_cents_p95":5.099229715645451,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003179454955178215,"log_rms_mae_db":1.1048439767547746,"rms_envelope_nrmse":0.11101402023226509,"envelope_correlation":0.9868604654425986,"rms_gain_error_db":-0.123416096659984,"reference_mean_power":0.009055057750136907,"output_mean_power":0.00880135654620421},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4928013782154505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203466496672066,"output_power":0.0020477704331425105,"reference_active":true,"delta_db":-0.31825183403110224},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002920043469136947,"output_power":0.00030007601603027474,"reference_active":true,"delta_db":0.11841968647122453},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.410659541472349e-05,"output_power":7.846388441863666e-05,"reference_active":true,"delta_db":0.24812942943718813}],"active_band_count":3,"active_band_mae_db":0.22826698331317163}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002752533875730688,"log_rms_mae_db":0.6066650993154313,"rms_envelope_nrmse":0.05614815782014008,"envelope_correlation":0.996517253689585,"rms_gain_error_db":-0.12341389548063535,"reference_mean_power":0.009100511091100074,"output_mean_power":0.008845540875822998},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.115339465742665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019704973258355524,"output_power":0.0018491311486283455,"reference_active":true,"delta_db":-0.27608135467240097},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014191960918623396,"output_power":0.00013736295093945075,"reference_active":true,"delta_db":-0.1417279443248323},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9761972447933005e-06,"output_power":3.79962106874641e-06,"reference_active":true,"delta_db":-0.19727632887203042}],"active_band_count":3,"active_band_mae_db":0.2050285426230879}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028904791382782896,"log_rms_mae_db":1.4344151570442891,"rms_envelope_nrmse":0.14847768526869048,"envelope_correlation":0.9817993245736957,"rms_gain_error_db":-0.37906826918777925,"reference_mean_power":0.00500094100859156,"output_mean_power":0.004582947737286726},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3504006393160654e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001701857204603173,"output_power":0.0013846151758818363,"reference_active":true,"delta_db":-0.8959403020046953},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029604441935142547,"output_power":0.0002501009147410514,"reference_active":true,"delta_db":-0.7324159857407786},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422561443844994e-05,"output_power":7.983528044231455e-05,"reference_active":true,"delta_db":0.31641053836879396}],"active_band_count":3,"active_band_mae_db":0.6482556087047558}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002500254339817177,"log_rms_mae_db":0.9185817467045656,"rms_envelope_nrmse":0.09256105930124875,"envelope_correlation":0.9934595587214328,"rms_gain_error_db":-0.3790418473200677,"reference_mean_power":0.00502600851498449,"output_mean_power":0.0046059480500599565},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0757929049004627e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015562264230646286,"output_power":0.0012744453705969315,"reference_active":true,"delta_db":-0.8675156083063142},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014399857449561677,"output_power":0.00011552073100345226,"reference_active":true,"delta_db":-0.9569826444134394},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7762361828128385e-06,"output_power":4.621493619749165e-06,"reference_active":true,"delta_db":-0.14303436523252433}],"active_band_count":3,"active_band_mae_db":0.6558442059840927}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010495143398294126,"log_rms_mae_db":1.5648167740464622,"rms_envelope_nrmse":0.1590959195511037,"envelope_correlation":0.9801484409041943,"rms_gain_error_db":-0.1345151464904844,"reference_mean_power":0.00041462711687219214,"output_mean_power":0.00040198161174917593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.072345528907365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001011743074744314,"output_power":0.00010153940142335597,"reference_active":true,"delta_db":0.01564358520162252},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2721317302684436e-05,"output_power":1.6778056936015975e-05,"reference_active":true,"delta_db":1.2020957864762989},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.690399523324186e-06,"output_power":5.317467303367643e-06,"reference_active":true,"delta_db":2.958880507851575}],"active_band_count":3,"active_band_mae_db":1.3922066265098323}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":1.0860695320004674,"rms_envelope_nrmse":0.09354146234981282,"envelope_correlation":0.9928139443437426,"rms_gain_error_db":-0.13455607596803454,"reference_mean_power":0.00041670963521221266,"output_mean_power":0.00040399680897410325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9169450940862239e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.131046807277538e-05,"output_power":9.433778496776685e-05,"reference_active":true,"delta_db":0.14165105829459237},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.750575155696516e-06,"output_power":8.177218836405326e-06,"reference_active":true,"delta_db":1.5289433664806777},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0115145688994783e-07,"output_power":3.746169772801987e-07,"reference_active":true,"delta_db":2.7006427288684454}],"active_band_count":3,"active_band_mae_db":1.4570790512145717}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":2.347871249963625,"rms_envelope_nrmse":0.36644630478374524,"envelope_correlation":0.904813145983991,"rms_gain_error_db":0.35679583999155495,"reference_mean_power":2.7148047549645084e-05,"output_mean_power":2.947258203368451e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3627974020623764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.843668843513599e-06,"output_power":4.282761815970265e-06,"reference_active":true,"delta_db":-0.534505212743158},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2177208449183313e-06,"output_power":2.079776114886348e-06,"reference_active":true,"delta_db":-0.27890292279711504},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.650596314175211e-07,"output_power":1.144601179495253e-06,"reference_active":true,"delta_db":3.911455460337772}],"active_band_count":3,"active_band_mae_db":1.5749545319593483}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":1.878931288636907,"rms_envelope_nrmse":0.2979743900585799,"envelope_correlation":0.9311452690677294,"rms_gain_error_db":0.3567787138344032,"reference_mean_power":2.7284479980131427e-05,"output_mean_power":2.9620579597848676e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2818152654967464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.740914469824879e-06,"output_power":4.4005239044803625e-06,"reference_active":true,"delta_db":-0.32357735839209417},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1748833712187603e-06,"output_power":1.191523438851216e-06,"reference_active":true,"delta_db":0.06107832883513186},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4248385933815314e-08,"output_power":7.667717524427975e-08,"reference_active":true,"delta_db":3.5002599649420456}],"active_band_count":3,"active_band_mae_db":1.294971884056424}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.416494846343994,"predicted_mos":3.416494846343994,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"247e4a61028767d01e375682f1c8bfe02bf1b024dfe81f57e10c61d55cd69047","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.107593302998794},"started_at":"2026-10-13T04:43:38.998464+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3924999368318822e-05,"native_input_samples":96000,"native_input_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","output_sha256":"7095a9b78f8b1e747f6eed5dd5fa8f6407afa127ac512132db6b7e8fdf2bec7a","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6992809539997324,"synthesis_s":0.08273463800014724,"total_s":1.7820155919998797,"rtf":0.29700259866664663,"output_peak":1.0635258516409574,"output_rms":0.09959271068540815,"finished_at":"2026-10-13T04:43:40.786324+00:00","evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"7095a9b78f8b1e747f6eed5dd5fa8f6407afa127ac512132db6b7e8fdf2bec7a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.171231274074423,"spectral_convergence":0.7519602971098978,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.17330573035546,"spectral_convergence":0.26621110628412065,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.521177705779178,"spectral_convergence":0.27664069593589924,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.0018116457934690764,"delta_db":0.04547032806489382,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.004430411827133914,"delta_db":0.5630548436755788,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.003077175679177336,"delta_db":0.3588329268407215,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.0005060571867068377,"delta_db":0.5469829271133723,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.6822512213947695e-05,"delta_db":0.40966650848612834,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":3.659502296629896e-05,"delta_db":0.580092586204578,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4173500200642121,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.3992582105362257,"pitch":null,"output_rms_dbfs":-20.035448939347784,"output_peak":1.0635258516409574,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.955238236736354},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.171231274074423,"spectral_convergence":0.7519602971098978,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.17330573035546,"spectral_convergence":0.26621110628412065,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.521177705779178,"spectral_convergence":0.27664069593589924,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.0018116457934690764,"delta_db":0.04547032806489382,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.004430411827133914,"delta_db":0.5630548436755788,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.003077175679177336,"delta_db":0.3588329268407215,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.0005060571867068377,"delta_db":0.5469829271133723,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.6822512213947695e-05,"delta_db":0.40966650848612834,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":3.659502296629896e-05,"delta_db":0.580092586204578,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4173500200642121,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.3992582105362257,"pitch":null,"output_rms_dbfs":-20.035448939347784,"output_peak":1.0635258516409574,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.955238236736354},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9016164563729676,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.095118423251492,"rms_dbfs":-20.434707149884012,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.09612477983555,"mean":-3.6449432373046876e-06,"zero_energy":false},"output":{"rms":0.09959271068540815,"rms_dbfs":-20.035448939347784,"peak":1.0635258516409574,"peak_dbfs":0.5349610194916836,"crest_factor_db":20.57040995883947,"mean":1.9538429809059948e-08,"zero_energy":false},"rms_difference_db":0.3992582105362281},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":6,"at_or_above_full_scale_fraction":6.25e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":6.25e-05,"longest_full_scale_run_samples":3,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":96,"quiet_samples":15360,"reference_mean_power":1.9068745435409559e-06,"output_mean_power":1.958020198715749e-06,"output_rms_dbfs":-57.08182832376831,"output_zero_energy":false,"output_peak":0.01673290611453905,"output_energy_fraction":3.15850845805878e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":191,"both_estimated_voiced_frames":170,"reference_voiced_output_unvoiced_frames":20,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.8947368421052632,"absolute_difference_cents_median":4.520214626958121,"absolute_difference_cents_p95":23.427535135241268,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003179454955178215,"log_rms_mae_db":2.0944505667082796,"rms_envelope_nrmse":0.1579114717883366,"envelope_correlation":0.976802374787975,"rms_gain_error_db":0.3992593011339408,"reference_mean_power":0.009055057750136907,"output_mean_power":0.009926980174981621},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4928013782154505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203466496672066,"output_power":0.0023110141836607642,"reference_active":true,"delta_db":0.20696159410726653},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002920043469136947,"output_power":0.0002919263513970999,"reference_active":true,"delta_db":-0.0011601726975391021},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.410659541472349e-05,"output_power":7.60142172629052e-05,"reference_active":true,"delta_db":0.11037966340375954}],"active_band_count":3,"active_band_mae_db":0.10616714340285505}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002752533875730688,"log_rms_mae_db":1.4242347150424735,"rms_envelope_nrmse":0.07863643574353912,"envelope_correlation":0.9955717616441542,"rms_gain_error_db":0.39919658321586776,"reference_mean_power":0.009100511091100074,"output_mean_power":0.00997666619551888},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.115339465742665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019704973258355524,"output_power":0.0020821950251582043,"reference_active":true,"delta_db":0.23945554692881083},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014191960918623396,"output_power":0.00014190759252941492,"reference_active":true,"delta_db":-0.0003677426066995322},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9761972447933005e-06,"output_power":3.487670212151743e-06,"reference_active":true,"delta_db":-0.5693250794765521}],"active_band_count":3,"active_band_mae_db":0.2697161230040208}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028904791382782896,"log_rms_mae_db":1.9816943517000156,"rms_envelope_nrmse":0.1846585466951524,"envelope_correlation":0.9754691158661299,"rms_gain_error_db":0.5271343220176141,"reference_mean_power":0.00500094100859156,"output_mean_power":0.005646315788491429},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3504006393160654e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001701857204603173,"output_power":0.0018402593538099927,"reference_active":true,"delta_db":0.3395591634809094},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029604441935142547,"output_power":0.0002804069315946562,"reference_active":true,"delta_db":-0.23568133647555403},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422561443844994e-05,"output_power":6.260361560072092e-05,"reference_active":true,"delta_db":-0.7395438524640806}],"active_band_count":3,"active_band_mae_db":0.43826145080684803}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002500254339817177,"log_rms_mae_db":1.578371998661696,"rms_envelope_nrmse":0.10738720957889597,"envelope_correlation":0.9933709333698869,"rms_gain_error_db":0.5270900076725284,"reference_mean_power":0.00502600851498449,"output_mean_power":0.005674560371220654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0757929049004627e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015562264230646286,"output_power":0.0016838877522312521,"reference_active":true,"delta_db":0.3424035320809327},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014399857449561677,"output_power":0.00013414985906162096,"reference_active":true,"delta_db":-0.30767972110296793},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7762361828128385e-06,"output_power":3.2555713542208247e-06,"reference_active":true,"delta_db":-1.6645857585857196}],"active_band_count":3,"active_band_mae_db":0.77155633725654}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010495143398294126,"log_rms_mae_db":1.4995352133346826,"rms_envelope_nrmse":0.1876759447873684,"envelope_correlation":0.976432094341681,"rms_gain_error_db":0.48374797289325655,"reference_mean_power":0.00041462711687219214,"output_mean_power":0.0004634816016740525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.072345528907365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001011743074744314,"output_power":0.00011239931083002977,"reference_active":true,"delta_db":0.4569340800734484},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2721317302684436e-05,"output_power":1.3491654181543763e-05,"reference_active":true,"delta_db":0.25533115626105735},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.690399523324186e-06,"output_power":4.066691401639691e-06,"reference_active":true,"delta_db":1.794244401271904}],"active_band_count":3,"active_band_mae_db":0.8355032125354699}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":1.030528218391128,"rms_envelope_nrmse":0.09410540979267347,"envelope_correlation":0.9956284517023039,"rms_gain_error_db":0.4837242417269682,"reference_mean_power":0.00041670963521221266,"output_mean_power":0.0004658069526676393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9169450940862239e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.131046807277538e-05,"output_power":0.00010145527918662386,"reference_active":true,"delta_db":0.45754081251712486},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.750575155696516e-06,"output_power":6.402769231287299e-06,"reference_active":true,"delta_db":0.4665656553247205},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0115145688994783e-07,"output_power":3.2307246089435204e-07,"reference_active":true,"delta_db":2.0577675737827765}],"active_band_count":3,"active_band_mae_db":0.9939580138748739}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":1.8172974159298123,"rms_envelope_nrmse":0.22584046092992105,"envelope_correlation":0.9662856823378368,"rms_gain_error_db":0.4752830302571129,"reference_mean_power":2.7148047549645084e-05,"output_mean_power":3.0287742489868016e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3627974020623764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.843668843513599e-06,"output_power":5.052219425484309e-06,"reference_active":true,"delta_db":0.18307761120870325},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2177208449183313e-06,"output_power":1.9903322296125524e-06,"reference_active":true,"delta_db":-0.46981302911153133},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.650596314175211e-07,"output_power":4.404524081693184e-07,"reference_active":true,"delta_db":-0.23609654243475373}],"active_band_count":3,"active_band_mae_db":0.29632906091832945}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":1.109897556020603,"rms_envelope_nrmse":0.14128813905717,"envelope_correlation":0.9870504941066707,"rms_gain_error_db":0.4752740756709516,"reference_mean_power":2.7284479980131427e-05,"output_mean_power":3.043989068458643e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2818152654967464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.740914469824879e-06,"output_power":4.8816835566951e-06,"reference_active":true,"delta_db":0.12707503562777225},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1748833712187603e-06,"output_power":1.0761099026541823e-06,"reference_active":true,"delta_db":-0.3813812912400031},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4248385933815314e-08,"output_power":2.774031651067224e-08,"reference_active":true,"delta_db":-0.9152869680940939}],"active_band_count":3,"active_band_mae_db":0.47458109832062306}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"facodec","status":"measured","mos":{"status":"measured","value":3.5452756881713867,"predicted_mos":3.5452756881713867,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4c5c52b68051ef0579bc7daea402e3f38008e52cbc55978069c3bf3c46c5459e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6867587369997636},"started_at":"2026-10-13T07:01:01.961514+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.260099947510753e-05,"native_input_samples":96000,"native_input_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","output_sha256":"43b75c9194d3572bcfa530553417bbaa942084bbab03e73f3a02036f24e16469","output_samples":96000,"length_delta_samples":0,"analysis_s":2.072742954000205,"synthesis_s":10.114097880001282,"total_s":12.186840834001487,"rtf":2.031140139000248,"output_peak":0.7007619738578796,"output_rms":0.08964465719629723,"finished_at":"2026-10-13T07:01:14.168979+00:00","evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"43b75c9194d3572bcfa530553417bbaa942084bbab03e73f3a02036f24e16469","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.308148567543953,"spectral_convergence":0.18647703698147272,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.539304236733893,"spectral_convergence":0.191580301057273,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.746237771688437,"spectral_convergence":0.21795459730488467,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.001669446396305497,"delta_db":-0.3095377812588914,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.003726557082714325,"delta_db":-0.18830835339244203,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.0022120594500990255,"delta_db":-1.0747220394700085,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.00034493838643018774,"delta_db":-1.1175977663431222,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.067585987116282e-05,"delta_db":-0.08752653251375289,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":3.2487388421445956e-05,"delta_db":0.06302035490669573,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4734521379808188,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5148046340171717,"pitch":null,"output_rms_dbfs":-20.949511783901183,"output_peak":0.7007619738578796,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.531230191988762},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.308148567543953,"spectral_convergence":0.18647703698147272,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.539304236733893,"spectral_convergence":0.191580301057273,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.746237771688437,"spectral_convergence":0.21795459730488467,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.001669446396305497,"delta_db":-0.3095377812588914,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.003726557082714325,"delta_db":-0.18830835339244203,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.0022120594500990255,"delta_db":-1.0747220394700085,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.00034493838643018774,"delta_db":-1.1175977663431222,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.067585987116282e-05,"delta_db":-0.08752653251375289,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":3.2487388421445956e-05,"delta_db":0.06302035490669573,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4734521379808188,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5148046340171717,"pitch":null,"output_rms_dbfs":-20.949511783901183,"output_peak":0.7007619738578796,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.531230191988762},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9087829322739788,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.095118423251492,"rms_dbfs":-20.434707149884012,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.09612477983555,"mean":-3.6449432373046876e-06,"zero_energy":false},"output":{"rms":0.08964465719629723,"rms_dbfs":-20.949511783901183,"peak":0.7007619738578796,"peak_dbfs":-3.088589455057372,"crest_factor_db":17.860922328843813,"mean":6.341672516659003e-05,"zero_energy":false},"rms_difference_db":-0.514804634017171},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":96,"quiet_samples":15360,"reference_mean_power":1.9068745435409559e-06,"output_mean_power":2.52135561646752e-06,"output_rms_dbfs":-55.9836589633958,"output_zero_energy":false,"output_peak":0.023961296305060387,"output_energy_fraction":5.020017888259271e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":195,"both_estimated_voiced_frames":186,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9789473684210527,"absolute_difference_cents_median":2.5164688311047216,"absolute_difference_cents_p95":9.213589025116416,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003179454955178215,"log_rms_mae_db":1.1350998868257118,"rms_envelope_nrmse":0.1046508780522477,"envelope_correlation":0.9915595797264768,"rms_gain_error_db":-0.5148085472149729,"reference_mean_power":0.009055057750136907,"output_mean_power":0.008042857418529281},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4928013782154505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203466496672066,"output_power":0.0018464797046111757,"reference_active":true,"delta_db":-0.7676191301267556},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002920043469136947,"output_power":0.00021996329632392552,"reference_active":true,"delta_db":-1.230390973031105},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.410659541472349e-05,"output_power":6.230788393172522e-05,"reference_active":true,"delta_db":-0.7531385926517074}],"active_band_count":3,"active_band_mae_db":0.9170495652698559}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002752533875730688,"log_rms_mae_db":0.8439451150955894,"rms_envelope_nrmse":0.09683251035176103,"envelope_correlation":0.9927660644288591,"rms_gain_error_db":-0.5148324939046025,"reference_mean_power":0.009100511091100074,"output_mean_power":0.00808318528441423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.115339465742665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019704973258355524,"output_power":0.0016519725796606101,"reference_active":true,"delta_db":-0.7657301543225883},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014191960918623396,"output_power":0.00010033022555673808,"reference_active":true,"delta_db":-1.5061061793899786},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9761972447933005e-06,"output_power":3.3933611090540956e-06,"reference_active":true,"delta_db":-0.6883784190968285}],"active_band_count":3,"active_band_mae_db":0.9867382509364652}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028904791382782896,"log_rms_mae_db":1.1980872563533034,"rms_envelope_nrmse":0.13011836652329678,"envelope_correlation":0.9904161672903612,"rms_gain_error_db":-0.7308173070656744,"reference_mean_power":0.00500094100859156,"output_mean_power":0.0042263941924298755},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3504006393160654e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001701857204603173,"output_power":0.001435684226114875,"reference_active":true,"delta_db":-0.7386418871762211},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029604441935142547,"output_power":0.00021487957838321468,"reference_active":true,"delta_db":-1.391617355411555},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422561443844994e-05,"output_power":6.196962364485024e-05,"reference_active":true,"delta_db":-0.7837494277393163}],"active_band_count":3,"active_band_mae_db":0.9713362234423641}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002500254339817177,"log_rms_mae_db":0.9471244231235443,"rms_envelope_nrmse":0.1204626357753032,"envelope_correlation":0.9920516117877243,"rms_gain_error_db":-0.7308483197533333,"reference_mean_power":0.00502600851498449,"output_mean_power":0.004247548906475518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0757929049004627e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015562264230646286,"output_power":0.0013141163010399741,"reference_active":true,"delta_db":-0.7343898233515258},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014399857449561677,"output_power":9.872720582603625e-05,"reference_active":true,"delta_db":-1.6392134704640695},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7762361828128385e-06,"output_power":3.691522093529244e-06,"reference_active":true,"delta_db":-1.118803223349273}],"active_band_count":3,"active_band_mae_db":1.1641355057216227}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010495143398294126,"log_rms_mae_db":1.3347194559178899,"rms_envelope_nrmse":0.17811044804354825,"envelope_correlation":0.9826876118402544,"rms_gain_error_db":-0.9900408819348254,"reference_mean_power":0.00041462711687219214,"output_mean_power":0.00033010614859411167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.072345528907365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001011743074744314,"output_power":7.9661712532815e-05,"reference_active":true,"delta_db":-1.038206019346254},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2721317302684436e-05,"output_power":9.00971003918638e-06,"reference_active":true,"delta_db":-1.4982127094419901},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.690399523324186e-06,"output_power":3.040720568676345e-06,"reference_active":true,"delta_db":0.5315973451148086}],"active_band_count":3,"active_band_mae_db":1.0226720246343508}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":1.0772494738673988,"rms_envelope_nrmse":0.1632666452094127,"envelope_correlation":0.9864118816072188,"rms_gain_error_db":-0.9900498309591808,"reference_mean_power":0.00041670963521221266,"output_mean_power":0.00033176346580655864},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9169450940862239e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.131046807277538e-05,"output_power":7.146366601802491e-05,"reference_active":true,"delta_db":-1.0643527772146912},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.750575155696516e-06,"output_power":4.166607515435753e-06,"reference_active":true,"delta_db":-1.3992869085312651},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0115145688994783e-07,"output_power":2.137689533658686e-07,"reference_active":true,"delta_db":0.26421448490780525}],"active_band_count":3,"active_band_mae_db":0.909284723551254}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":2.017215258316301,"rms_envelope_nrmse":0.22103512480946183,"envelope_correlation":0.9637921750573717,"rms_gain_error_db":-0.05830935613749324,"reference_mean_power":2.7148047549645084e-05,"output_mean_power":2.678598773066981e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3627974020623764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.843668843513599e-06,"output_power":4.899376383791156e-06,"reference_active":true,"delta_db":0.049663612199817256},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2177208449183313e-06,"output_power":2.4658312353820886e-06,"reference_active":true,"delta_db":0.4605647107228952},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.650596314175211e-07,"output_power":6.31372791714922e-07,"reference_active":true,"delta_db":1.327772194645912}],"active_band_count":3,"active_band_mae_db":0.6126668391895415}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":1.5986604442865462,"rms_envelope_nrmse":0.16470844958179492,"envelope_correlation":0.9778443130183281,"rms_gain_error_db":-0.05830444009024231,"reference_mean_power":2.7284479980131427e-05,"output_mean_power":2.6920631103742226e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2818152654967464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.740914469824879e-06,"output_power":5.000232544524846e-06,"reference_active":true,"delta_db":0.23128082083512147},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1748833712187603e-06,"output_power":1.2712311566401618e-06,"reference_active":true,"delta_db":0.34229771432207834},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4248385933815314e-08,"output_power":4.391224116341375e-08,"reference_active":true,"delta_db":1.0794549412380379}],"active_band_count":3,"active_band_mae_db":0.5510111587984126}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.1376709938049316,"predicted_mos":3.1376709938049316,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"b52f018bcd55e5cc897bf664355b478ffdc8b581bd4cf05ccbab506aa3adad69","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.551648977001605},"started_at":"2026-10-13T07:30:59.648490+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006801847997849109,"native_input_samples":264600,"native_input_sha256":"094088416895256e81567d0ebf6d45742e348d9570b929eb99a7156274364f8d","output_sha256":"9b56a7c870d90aaf1b8198cb9a2775d9e12ab0622ca38ce9b31430a8fa9641f7","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1323284580030304,"synthesis_s":3.89251607499682,"total_s":4.02484453299985,"rtf":0.6708074221666417,"output_peak":0.7010907530784607,"output_rms":0.0866850738450057,"finished_at":"2026-10-13T07:31:03.697405+00:00","evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"9b56a7c870d90aaf1b8198cb9a2775d9e12ab0622ca38ce9b31430a8fa9641f7","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.23541496491676,"output_peak":0.7018106803909403},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.063525515566042,"spectral_convergence":0.4142731545510643,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.970397423510818,"spectral_convergence":0.31056742398887743,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.914042886043132,"spectral_convergence":0.3818258993909562,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.0016350312649696037,"delta_db":-0.4000019549557057,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.00298424004135064,"delta_db":-1.1530486013442363,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.002498602977013011,"delta_db":-0.5457174550833891,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.00032445009961921153,"delta_db":-1.3835339226387378,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.8315067478997765e-05,"delta_db":0.5222700268899831,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":2.6506963285882992e-05,"delta_db":-0.8205278887308222,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8041833082738123,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7989892124900261,"pitch":null,"output_rms_dbfs":-21.23369636237404,"output_peak":0.7018106803909403,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.982655275039997},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7830947443570537,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.095118423251492,"rms_dbfs":-20.434707149884012,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.09612477983555,"mean":-3.6449432373046876e-06,"zero_energy":false},"output":{"rms":0.08675912870538377,"rms_dbfs":-21.23369636237404,"peak":0.7018106803909403,"peak_dbfs":-3.0756005366170083,"crest_factor_db":18.15809582575703,"mean":-0.00015315640433534523,"zero_energy":false},"rms_difference_db":-0.798989212490028},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":96,"quiet_samples":15360,"reference_mean_power":1.9068745435409559e-06,"output_mean_power":1.92999008029665e-06,"output_rms_dbfs":-57.14444923159842,"output_zero_energy":false,"output_peak":0.015964155866152782,"output_energy_fraction":4.10246321613613e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":178,"both_estimated_voiced_frames":159,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.8368421052631579,"absolute_difference_cents_median":7.171546312285733,"absolute_difference_cents_p95":36.75913353290462,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003179454955178215,"log_rms_mae_db":1.750608310831782,"rms_envelope_nrmse":0.13114204758945147,"envelope_correlation":0.9891039789394528,"rms_gain_error_db":-0.7989882538739681,"reference_mean_power":0.009055057750136907,"output_mean_power":0.00753342378756763},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4928013782154505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203466496672066,"output_power":0.0017652968598615107,"reference_active":true,"delta_db":-0.9628870284274798},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002920043469136947,"output_power":0.00025212946258326073,"reference_active":true,"delta_db":-0.6376571849317867},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.410659541472349e-05,"output_power":8.610573943251719e-05,"reference_active":true,"delta_db":0.6517523910704324}],"active_band_count":3,"active_band_mae_db":0.7507655348098995}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002752533875730688,"log_rms_mae_db":1.2854458500705561,"rms_envelope_nrmse":0.1070731650855005,"envelope_correlation":0.9953149911596609,"rms_gain_error_db":-0.7989700206078891,"reference_mean_power":0.009100511091100074,"output_mean_power":0.007571270825279618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.115339465742665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019704973258355524,"output_power":0.0015842308610426862,"reference_active":true,"delta_db":-0.9475738066588484},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014191960918623396,"output_power":0.00012292376220765394,"reference_active":true,"delta_db":-0.6240656272756067},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9761972447933005e-06,"output_power":3.991468736649754e-06,"reference_active":true,"delta_db":0.016648119488995715}],"active_band_count":3,"active_band_mae_db":0.5294291844744836}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028904791382782896,"log_rms_mae_db":2.010731291884297,"rms_envelope_nrmse":0.15641239710786903,"envelope_correlation":0.9833494141758193,"rms_gain_error_db":-0.7193897694666049,"reference_mean_power":0.00500094100859156,"output_mean_power":0.004237529695722229},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3504006393160654e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001701857204603173,"output_power":0.0013936238693414134,"reference_active":true,"delta_db":-0.8677754140499845},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029604441935142547,"output_power":0.00029953254666791946,"reference_active":true,"delta_db":0.050871402125023835},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422561443844994e-05,"output_power":9.106766014842856e-05,"reference_active":true,"delta_db":0.8881037678602586}],"active_band_count":3,"active_band_mae_db":0.6022501946784223}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002500254339817177,"log_rms_mae_db":1.444790353385605,"rms_envelope_nrmse":0.1256608019512576,"envelope_correlation":0.9905234845802297,"rms_gain_error_db":-0.7193403712127912,"reference_mean_power":0.00502600851498449,"output_mean_power":0.004258818999803137},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0757929049004627e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015562264230646286,"output_power":0.001279089948915303,"reference_active":true,"delta_db":-0.8517169863451888},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014399857449561677,"output_power":0.00015404249816606465,"reference_active":true,"delta_db":0.2928236028042329},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7762361828128385e-06,"output_power":4.763065467927686e-06,"reference_active":true,"delta_db":-0.01199243433404559}],"active_band_count":3,"active_band_mae_db":0.3855110078278225}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010495143398294126,"log_rms_mae_db":2.5640460690409403,"rms_envelope_nrmse":0.24412006538750602,"envelope_correlation":0.9638665325521801,"rms_gain_error_db":-1.3178876843540446,"reference_mean_power":0.00041462711687219214,"output_mean_power":0.0003061039496918402},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.072345528907365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001011743074744314,"output_power":8.031779105801301e-05,"reference_active":true,"delta_db":-1.0025848461127087},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2721317302684436e-05,"output_power":9.867723862925806e-06,"reference_active":true,"delta_db":-1.1031509743681662},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.690399523324186e-06,"output_power":2.3773059319310308e-06,"reference_active":true,"delta_db":-0.5373170333034992}],"active_band_count":3,"active_band_mae_db":0.8810176179281247}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":2.0693162821946043,"rms_envelope_nrmse":0.2202412172703115,"envelope_correlation":0.9720996810257555,"rms_gain_error_db":-1.3178623117679786,"reference_mean_power":0.00041670963521221266,"output_mean_power":0.00030764319370807006},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9169450940862239e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.131046807277538e-05,"output_power":7.277434374859933e-05,"reference_active":true,"delta_db":-0.9854227122701207},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.750575155696516e-06,"output_power":4.153708591387739e-06,"reference_active":true,"delta_db":-1.4127525892879402},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0115145688994783e-07,"output_power":1.682251862962651e-07,"reference_active":true,"delta_db":-0.7763216448335468}],"active_band_count":3,"active_band_mae_db":1.0581656487972027}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":3.632012469375252,"rms_envelope_nrmse":0.2608439221003644,"envelope_correlation":0.9489861723937788,"rms_gain_error_db":-0.17912288725958803,"reference_mean_power":2.7148047549645084e-05,"output_mean_power":2.6051114785044444e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3627974020623764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.843668843513599e-06,"output_power":4.197453753148022e-06,"reference_active":true,"delta_db":-0.6218852338263741},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2177208449183313e-06,"output_power":2.41489664770919e-06,"reference_active":true,"delta_db":0.3699167009276311},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.650596314175211e-07,"output_power":4.36410933987852e-07,"reference_active":true,"delta_db":-0.2761302005896032}],"active_band_count":3,"active_band_mae_db":0.4226440451145361}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":3.0886982204552322,"rms_envelope_nrmse":0.20090103154371924,"envelope_correlation":0.9668135484256404,"rms_gain_error_db":-0.17910994265639194,"reference_mean_power":2.7284479980131427e-05,"output_mean_power":2.618211262167513e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2818152654967464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.740914469824879e-06,"output_power":4.047263292950725e-06,"reference_active":true,"delta_db":-0.6870066219464568},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1748833712187603e-06,"output_power":1.2635075172619694e-06,"reference_active":true,"delta_db":0.31583073073663415},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4248385933815314e-08,"output_power":3.691397005319188e-08,"reference_active":true,"delta_db":0.32550646782270987}],"active_band_count":3,"active_band_mae_db":0.44278127350193364}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"golf","status":"measured","mos":{"status":"measured","value":2.8396975994110107,"predicted_mos":2.8396975994110107,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"74686a3e4e418c786c496ecdcbeb1a54cc08e0d1c0f8bec8a83e6fd3499915ce","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.747921371999837},"started_at":"2026-10-13T05:03:06.240206+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027816339998025796,"native_input_samples":144000,"native_input_sha256":"d88e8d947292eda84d7e3dfae7e3d1a4802647d00efbf740296ddd38ac8a53fe","output_sha256":"0e2cb5aaf882415c5f026a913daf13d850aebe4c330c3825a81fdecb4f8f730b","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6121565229996122,"synthesis_s":0.12305669499983196,"total_s":0.7352132179994442,"rtf":0.1225355363332407,"output_peak":0.7441802024841309,"output_rms":0.08637378636838976,"finished_at":"2026-10-13T05:03:06.987253+00:00","analysis_stages":{"external_f0_s":0.08896058899972559},"evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"0e2cb5aaf882415c5f026a913daf13d850aebe4c330c3825a81fdecb4f8f730b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.26759679289732,"output_peak":0.7442329720548438},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.850662283933409,"spectral_convergence":0.6774780865008679,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.0380685336315,"spectral_convergence":0.26893729362809843,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.2773971087663245,"spectral_convergence":0.2558660218402985,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0017957580075624278,"output_power":0.0016077631946354953,"delta_db":-0.4802572924644633,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038981945150312358,"output_power":0.003191828691331106,"delta_db":-0.8682393179301742,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002837874103009799,"output_power":0.0022878621397996205,"delta_db":-0.935632733508224,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0004469161584167408,"output_power":0.00032320512904908426,"delta_db":-1.4074781289964613,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.17938603437769e-05,"output_power":3.539472203421057e-05,"delta_db":-1.6533977539937572,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2072806997182255e-05,"output_power":2.2565390824271974e-05,"delta_db":-1.5269411110125493,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1453243896509382,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8401294398162573,"pitch":null,"output_rms_dbfs":-21.26759679289732,"output_peak":0.7442329720548438,"reference_rms_dbfs":-20.427467353081063,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.722042642110412},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8978247171247423,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.09519773868827539,"rms_dbfs":-20.427467353081063,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.088884983032596,"mean":-7.004650287914754e-06,"zero_energy":false},"output":{"rms":0.08642117372307431,"rms_dbfs":-21.26759679289732,"peak":0.7442329720548438,"peak_dbfs":-2.565821863277926,"crest_factor_db":18.70177492961939,"mean":0.00143215926723703,"zero_energy":false},"rms_difference_db":-0.8401294398162555},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":95,"quiet_samples":15200,"reference_mean_power":1.8679601213845767e-06,"output_mean_power":2.6070041445868443e-06,"output_rms_dbfs":-55.838582783944766,"output_zero_energy":false,"output_peak":0.018721692871234544,"output_energy_fraction":5.5360267989039893e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":212,"both_estimated_voiced_frames":185,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9736842105263158,"absolute_difference_cents_median":2.79191335590703,"absolute_difference_cents_p95":12.821327520701598,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0003179454955178215,"log_rms_mae_db":1.9054756858108068,"rms_envelope_nrmse":0.17370906614341586,"envelope_correlation":0.9745105008436316,"rms_gain_error_db":-0.8401308107200909,"reference_mean_power":0.009070178941288051,"output_mean_power":0.007474855027376105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.498046401185465e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203917471895362,"output_power":0.0018251060980551902,"reference_active":true,"delta_db":-0.8190721168333011},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002867611419247014,"output_power":0.00026562402212804614,"reference_active":true,"delta_db":-0.33252952561816823},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.415825769912891e-05,"output_power":9.155903381332369e-05,"reference_active":true,"delta_db":0.915416833018651}],"active_band_count":3,"active_band_mae_db":0.6890061584900401}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002752533875730688,"log_rms_mae_db":1.2775347043728549,"rms_envelope_nrmse":0.11780745283612452,"envelope_correlation":0.9930785790871338,"rms_gain_error_db":-0.840010625841108,"reference_mean_power":0.00911539909094091,"output_mean_power":0.007512329444032958},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.1207643600984446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019685913962686455,"output_power":0.0016545233856638797,"reference_active":true,"delta_db":-0.7548267266629927},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013889434123617994,"output_power":0.00012897821492416985,"reference_active":true,"delta_db":-0.3216819034044041},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.958789878733436e-06,"output_power":5.012793718879693e-06,"reference_active":true,"delta_db":1.0251738210670456}],"active_band_count":3,"active_band_mae_db":0.700560817044814}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00028904791382782896,"log_rms_mae_db":1.7427130605378411,"rms_envelope_nrmse":0.19731757036264536,"envelope_correlation":0.97261154161016,"rms_gain_error_db":-0.918496654671388,"reference_mean_power":0.0050092956313014395,"output_mean_power":0.004054403774698043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":3.3538475226675194e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017039787156526217,"output_power":0.0013830936608614137,"reference_active":true,"delta_db":-0.9061257488143861},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029158108370376464,"output_power":0.0002780197384972674,"reference_active":true,"delta_db":-0.20683715236392622},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422841901483912e-05,"output_power":8.31690438851187e-05,"reference_active":true,"delta_db":0.49391498196419276}],"active_band_count":3,"active_band_mae_db":0.5356259610475017}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002500254339817177,"log_rms_mae_db":1.3329106512382993,"rms_envelope_nrmse":0.1337086293621361,"envelope_correlation":0.9920075218285495,"rms_gain_error_db":-0.9182712726102591,"reference_mean_power":0.005034157956760917,"output_mean_power":0.004074738202480772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.0792829830748667e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015563495402200382,"output_power":0.0012872603838529869,"reference_active":true,"delta_db":-0.8244073790962808},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014123425772546237,"output_power":0.00014206205994483155,"reference_active":true,"delta_db":0.02538055998341259},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.759298876102514e-06,"output_power":5.029824349735457e-06,"reference_active":true,"delta_db":0.24009840367018886}],"active_band_count":3,"active_band_mae_db":0.3632954475832941}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00010495143398294126,"log_rms_mae_db":1.8439682075783719,"rms_envelope_nrmse":0.24201371077947811,"envelope_correlation":0.966597966583708,"rms_gain_error_db":-1.3911842790964204,"reference_mean_power":0.0004153198392961769,"output_mean_power":0.0003014839865731653},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0774482357837378e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010131606852424694,"output_power":7.26117980361703e-05,"reference_active":true,"delta_db":-1.446711380254235},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2636009977731404e-05,"output_power":1.1011749918672066e-05,"reference_active":true,"delta_db":-0.5975362033515539},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6784746306883353e-06,"output_power":4.422939667367601e-06,"reference_active":true,"delta_db":2.1782347798089488}],"active_band_count":3,"active_band_mae_db":1.407494121138246}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":1.4958777716346492,"rms_envelope_nrmse":0.18802588945632884,"envelope_correlation":0.9868765902809168,"rms_gain_error_db":-1.39136352289184,"reference_mean_power":0.0004173928552646867,"output_mean_power":0.000302976300592358},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.921812883910512e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.140284746797202e-05,"output_power":6.719281809746908e-05,"reference_active":true,"delta_db":-1.3363686951656029},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.669686637705557e-06,"output_power":5.563931750413742e-06,"reference_active":true,"delta_db":-0.08177262089425846},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0070727686441146e-07,"output_power":3.018655179870858e-07,"reference_active":true,"delta_db":1.7725038792029448}],"active_band_count":3,"active_band_mae_db":1.0635483984209353}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":2.1992256667540393,"rms_envelope_nrmse":0.20229897605836317,"envelope_correlation":0.9762009189093622,"rms_gain_error_db":-0.9317438023037856,"reference_mean_power":2.7193389133231448e-05,"output_mean_power":2.194264400333813e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3622255294090722e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.829341682749909e-06,"output_power":3.759675493476302e-06,"reference_active":true,"delta_db":-1.0873757187902604},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2138298232860013e-06,"output_power":1.8048801746365626e-06,"reference_active":true,"delta_db":-0.8869585914436857},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.640155030485364e-07,"output_power":3.4463519416566596e-07,"reference_active":true,"delta_db":-1.2917286529396521}],"active_band_count":3,"active_band_mae_db":1.0886876543911994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":1.8292562410045452,"rms_envelope_nrmse":0.15217470584156687,"envelope_correlation":0.9895888514726567,"rms_gain_error_db":-0.9320235776532284,"reference_mean_power":2.7328788607487915e-05,"output_mean_power":2.205047877240981e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2814075069006191e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.726270135666308e-06,"output_power":3.789763559799234e-06,"reference_active":true,"delta_db":-0.9590642508908455},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1707393253695032e-06,"output_power":9.71228645287856e-07,"reference_active":true,"delta_db":-0.8113872369491189},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.417121998629131e-08,"output_power":2.25705016559141e-08,"reference_active":true,"delta_db":-1.8011927243563466}],"active_band_count":3,"active_band_mae_db":1.1905480707321037}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.413331985473633,"predicted_mos":3.413331985473633,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"aff3f28d0982b162da998f22ec23ac623fd1c41fedac710b74032a4697c1d513","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.51155767399905},"started_at":"2026-10-13T07:46:18.358387+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004076485001860419,"native_input_samples":144000,"native_input_sha256":"d88e8d947292eda84d7e3dfae7e3d1a4802647d00efbf740296ddd38ac8a53fe","output_sha256":"973f78ef0c8221eed5a437b30c30450ac17281a5042c3b2c9755d4c07d5c41ce","output_samples":144120,"length_delta_samples":120,"analysis_s":2.7177792729999055,"synthesis_s":19.514023132000148,"total_s":22.231802405000053,"rtf":3.7053004008333423,"output_peak":0.8495595455169678,"output_rms":0.09383718896002338,"finished_at":"2026-10-13T07:46:40.606138+00:00","analysis_stages":{"external_f0_s":1.805124758000602},"evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"973f78ef0c8221eed5a437b30c30450ac17281a5042c3b2c9755d4c07d5c41ce","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.54828019606028,"output_peak":0.8483380270451241},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.160612498654366,"spectral_convergence":0.7123185973582298,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.178324196173266,"spectral_convergence":0.31901946497703126,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.259304291781616,"spectral_convergence":0.28441070977017835,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.002435724762685242,"delta_db":1.33101954258396,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.0035759567536883443,"delta_db":-0.3674635544207711,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.0023555319619746567,"delta_db":-0.8017999708521659,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.00038052707453065316,"delta_db":-0.691157418127723,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.0651649965756353e-05,"delta_db":-0.08960182846100753,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":2.2930891379163812e-05,"delta_db":-1.4499182829417492,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7884934328978961,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10995555953957845,"pitch":null,"output_rms_dbfs":-20.54466270942359,"output_peak":0.8483380270451241,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.866080328869749},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8732212284977032,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.095118423251492,"rms_dbfs":-20.434707149884012,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.09612477983555,"mean":-3.6449432373046876e-06,"zero_energy":false},"output":{"rms":0.09392189890661186,"rms_dbfs":-20.54466270942359,"peak":0.8483380270451241,"peak_dbfs":-1.4286213032365314,"crest_factor_db":19.11604140618706,"mean":-0.00023613034602762236,"zero_energy":false},"rms_difference_db":-0.10995555953957847},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":96,"quiet_samples":15360,"reference_mean_power":1.9068745435409559e-06,"output_mean_power":3.1304964124340456e-06,"output_rms_dbfs":-55.043867895824235,"output_zero_energy":false,"output_peak":0.039084868233708873,"output_energy_fraction":5.678053287918023e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":217,"both_estimated_voiced_frames":182,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":35,"paired_voiced_fraction":0.9578947368421052,"absolute_difference_cents_median":4.86189905660099,"absolute_difference_cents_p95":21.20779800328408,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003179454955178215,"log_rms_mae_db":2.832894108594146,"rms_envelope_nrmse":0.16446654167498562,"envelope_correlation":0.9707954787217286,"rms_gain_error_db":-0.10995441185448784,"reference_mean_power":0.009055057750136907,"output_mean_power":0.008828680150000965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4928013782154505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203466496672066,"output_power":0.001974602475604744,"reference_active":true,"delta_db":-0.4762677436008288},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002920043469136947,"output_power":0.00025263304167331567,"reference_active":true,"delta_db":-0.6289916564613636},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.410659541472349e-05,"output_power":6.502519898286538e-05,"reference_active":true,"delta_db":-0.5677517163527801}],"active_band_count":3,"active_band_mae_db":0.5576703721383242}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002752533875730688,"log_rms_mae_db":2.2808422206020604,"rms_envelope_nrmse":0.09844793787293564,"envelope_correlation":0.9887091546375435,"rms_gain_error_db":-0.10991591871473122,"reference_mean_power":0.009100511091100074,"output_mean_power":0.008873075796667124},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.115339465742665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019704973258355524,"output_power":0.0017571926751554418,"reference_active":true,"delta_db":-0.4975646557667556},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014191960918623396,"output_power":0.00012601179664905438,"reference_active":true,"delta_db":-0.5163120285557242},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9761972447933005e-06,"output_power":3.4718167879485303e-06,"reference_active":true,"delta_db":-0.5891112118731909}],"active_band_count":3,"active_band_mae_db":0.5343292987318903}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028904791382782896,"log_rms_mae_db":2.177059775869076,"rms_envelope_nrmse":0.20137800344269244,"envelope_correlation":0.9669517933573403,"rms_gain_error_db":-0.63384851946163,"reference_mean_power":0.00500094100859156,"output_mean_power":0.004321822037305879},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3504006393160654e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001701857204603173,"output_power":0.001396222529335268,"reference_active":true,"delta_db":-0.8596847605816802},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029604441935142547,"output_power":0.00025485721381036843,"reference_active":true,"delta_db":-0.6505994779915315},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422561443844994e-05,"output_power":6.786921640957273e-05,"reference_active":true,"delta_db":-0.3888096633315497}],"active_band_count":3,"active_band_mae_db":0.6330313006349205}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002500254339817177,"log_rms_mae_db":1.3766483842637938,"rms_envelope_nrmse":0.12457712553969359,"envelope_correlation":0.9893304135005233,"rms_gain_error_db":-0.6337815235301052,"reference_mean_power":0.00502600851498449,"output_mean_power":0.004343552425317657},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0757929049004627e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015562264230646286,"output_power":0.0012833017987306063,"reference_active":true,"delta_db":-0.8374398190662771},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014399857449561677,"output_power":0.00012841302440096076,"reference_active":true,"delta_db":-0.4974911819043301},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7762361828128385e-06,"output_power":4.374801291205937e-06,"reference_active":true,"delta_db":-0.381274627036716}],"active_band_count":3,"active_band_mae_db":0.5720685426691078}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010495143398294126,"log_rms_mae_db":2.0690503311471193,"rms_envelope_nrmse":0.21780960256668314,"envelope_correlation":0.9636198626367813,"rms_gain_error_db":-0.5457645397480877,"reference_mean_power":0.00041462711687219214,"output_mean_power":0.00036566319351038774},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.072345528907365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001011743074744314,"output_power":8.862397086233272e-05,"reference_active":true,"delta_db":-0.5751903536610051},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2721317302684436e-05,"output_power":1.1145977488146582e-05,"reference_active":true,"delta_db":-0.5741392362152093},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.690399523324186e-06,"output_power":2.1118526709357475e-06,"reference_active":true,"delta_db":-1.0515316010379134}],"active_band_count":3,"active_band_mae_db":0.7336203969713759}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":1.329101762739951,"rms_envelope_nrmse":0.13105172943957946,"envelope_correlation":0.9880734666399852,"rms_gain_error_db":-0.545738442971986,"reference_mean_power":0.00041670963521221266,"output_mean_power":0.0003675019925248329},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9169450940862239e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.131046807277538e-05,"output_power":7.981832927269109e-05,"reference_active":true,"delta_db":-0.5841793599871747},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.750575155696516e-06,"output_power":5.240702979918876e-06,"reference_active":true,"delta_db":-0.40321737233159305},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0115145688994783e-07,"output_power":1.7332964769901783e-07,"reference_active":true,"delta_db":-0.646503281115318}],"active_band_count":3,"active_band_mae_db":0.5446333378113619}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":2.243665498289177,"rms_envelope_nrmse":0.2598597428197816,"envelope_correlation":0.9512555493153093,"rms_gain_error_db":-0.8297424565690557,"reference_mean_power":2.7148047549645084e-05,"output_mean_power":2.242664742882336e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3627974020623764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.843668843513599e-06,"output_power":4.122240129504277e-06,"reference_active":true,"delta_db":-0.700411564506412},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2177208449183313e-06,"output_power":1.8064196364024525e-06,"reference_active":true,"delta_db":-0.8908823300238016},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.650596314175211e-07,"output_power":2.777458133992562e-07,"reference_active":true,"delta_db":-2.2386112174753583}],"active_band_count":3,"active_band_mae_db":1.2766350373351907}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":1.4061693529064017,"rms_envelope_nrmse":0.1784683624173403,"envelope_correlation":0.9788863351615921,"rms_gain_error_db":-0.8297330096529205,"reference_mean_power":2.7284479980131427e-05,"output_mean_power":2.2539401505738615e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2818152654967464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.740914469824879e-06,"output_power":3.944522748454325e-06,"reference_active":true,"delta_db":-0.7986765532060143},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1748833712187603e-06,"output_power":9.407882127952489e-07,"reference_active":true,"delta_db":-0.965028895745977},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4248385933815314e-08,"output_power":2.084909717150807e-08,"reference_active":true,"delta_db":-2.155528553135139}],"active_band_count":3,"active_band_mae_db":1.3064113340290433}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0767710208892822,"predicted_mos":3.0767710208892822,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"a2d595ec358c858ba5f3f2ba78dcbe88db93d5e898d706166450daf83e54873c","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.934557808999671},"started_at":"2026-10-13T05:11:33.422008+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003141348001008737,"native_input_samples":144000,"native_input_sha256":"d88e8d947292eda84d7e3dfae7e3d1a4802647d00efbf740296ddd38ac8a53fe","output_sha256":"a8d9c4cc89d901f794946356c5fa7fd04e98eef6ced2142e2615e3b6dcc810de","output_samples":144240,"length_delta_samples":240,"analysis_s":1.0152758179992816,"synthesis_s":0.13687960900097096,"total_s":1.1521554270002525,"rtf":0.1920259045000421,"output_peak":0.810525119304657,"output_rms":0.09244525482305416,"finished_at":"2026-10-13T05:11:34.585875+00:00","analysis_stages":{"external_f0_s":0.1550981310010684},"evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"a8d9c4cc89d901f794946356c5fa7fd04e98eef6ced2142e2615e3b6dcc810de","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.677604918028763,"output_peak":0.8112778129805619},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.6987247459996455,"spectral_convergence":0.19103107637073283,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.218774873483306,"spectral_convergence":0.17184119468056322,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.6319462092381585,"spectral_convergence":0.18333084189926235,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.0017468592955886392,"delta_db":-0.11268331883958854,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.0037293538322110418,"delta_db":-0.18505023256853242,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.002616732066910463,"delta_db":-0.3450974207104757,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.0004045724102261589,"delta_db":-0.4250504132738032,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":4.646610911974446e-05,"delta_db":-0.4641747184153352,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":2.5659004801178887e-05,"delta_db":-0.9617295741658817,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4156309463289361,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2356656909345124,"pitch":null,"output_rms_dbfs":-20.670372840818523,"output_peak":0.8112778129805619,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.516481942907037},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.937093443684269,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.095118423251492,"rms_dbfs":-20.434707149884012,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.09612477983555,"mean":-3.6449432373046876e-06,"zero_energy":false},"output":{"rms":0.09257236476863508,"rms_dbfs":-20.670372840818523,"peak":0.8112778129805619,"peak_dbfs":-1.816608020979622,"crest_factor_db":18.853764819838904,"mean":-0.00014384576658660702,"zero_energy":false},"rms_difference_db":-0.2356656909345105},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":96,"quiet_samples":15360,"reference_mean_power":1.9068745435409559e-06,"output_mean_power":2.5357613468836893e-06,"output_rms_dbfs":-55.95891622481416,"output_zero_energy":false,"output_peak":0.034767340822502414,"output_energy_fraction":4.734407592146312e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":191,"both_estimated_voiced_frames":182,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9578947368421052,"absolute_difference_cents_median":0.7494583477192867,"absolute_difference_cents_p95":4.992039096147302,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003179454955178215,"log_rms_mae_db":1.0322658659042816,"rms_envelope_nrmse":0.09058066635827981,"envelope_correlation":0.9918037158390196,"rms_gain_error_db":-0.23566571935455802,"reference_mean_power":0.009055057750136907,"output_mean_power":0.008576787548473217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4928013782154505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203466496672066,"output_power":0.001988454321055013,"reference_active":true,"delta_db":-0.44590832737352654},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002920043469136947,"output_power":0.0002654558815310158,"reference_active":true,"delta_db":-0.41396964447550916},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.410659541472349e-05,"output_power":6.673820495238724e-05,"reference_active":true,"delta_db":-0.45482340163267193}],"active_band_count":3,"active_band_mae_db":0.4382337911605692}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002752533875730688,"log_rms_mae_db":0.5318045734314949,"rms_envelope_nrmse":0.056239511169410686,"envelope_correlation":0.9970689688317859,"rms_gain_error_db":-0.2356830398723413,"reference_mean_power":0.009100511091100074,"output_mean_power":0.008619805756514526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.115339465742665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019704973258355524,"output_power":0.0017893131936985432,"reference_active":true,"delta_db":-0.4188948555806986},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014191960918623396,"output_power":0.00013241119526962754,"reference_active":true,"delta_db":-0.3011770060620317},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9761972447933005e-06,"output_power":3.3934023606134213e-06,"reference_active":true,"delta_db":-0.6883256241978599}],"active_band_count":3,"active_band_mae_db":0.4694658286135301}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028904791382782896,"log_rms_mae_db":1.0602959013968607,"rms_envelope_nrmse":0.11758667125573599,"envelope_correlation":0.9887484045521147,"rms_gain_error_db":-0.3081014153067264,"reference_mean_power":0.00500094100859156,"output_mean_power":0.004658451626083353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3504006393160654e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001701857204603173,"output_power":0.001482831091501854,"reference_active":true,"delta_db":-0.5983143393305727},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029604441935142547,"output_power":0.0002799410471022098,"reference_active":true,"delta_db":-0.24290296013039037},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422561443844994e-05,"output_power":7.125942022552331e-05,"reference_active":true,"delta_db":-0.17711516804018845}],"active_band_count":3,"active_band_mae_db":0.3394441558337172}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002500254339817177,"log_rms_mae_db":0.6514957834273416,"rms_envelope_nrmse":0.07559201767619379,"envelope_correlation":0.9956345667328328,"rms_gain_error_db":-0.3081216462969528,"reference_mean_power":0.00502600851498449,"output_mean_power":0.004681780575154912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0757929049004627e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015562264230646286,"output_power":0.0013621091639309884,"reference_active":true,"delta_db":-0.578608701319802},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014399857449561677,"output_power":0.0001406376962611708,"reference_active":true,"delta_db":-0.10256449090936905},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7762361828128385e-06,"output_power":4.043516109831707e-06,"reference_active":true,"delta_db":-0.7232661659746485}],"active_band_count":3,"active_band_mae_db":0.4681464527346065}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010495143398294126,"log_rms_mae_db":1.054458500644894,"rms_envelope_nrmse":0.1233741241295463,"envelope_correlation":0.9895086759762616,"rms_gain_error_db":-0.43081675786558227,"reference_mean_power":0.00041462711687219214,"output_mean_power":0.0003754706772658663},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.072345528907365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001011743074744314,"output_power":8.609278345869214e-05,"reference_active":true,"delta_db":-0.7010349119801571},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2721317302684436e-05,"output_power":1.1551146421276405e-05,"reference_active":true,"delta_db":-0.41906996219053555},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.690399523324186e-06,"output_power":2.758221721003236e-06,"reference_active":true,"delta_db":0.1081239685782538}],"active_band_count":3,"active_band_mae_db":0.40940961424964883}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":0.7111453104039095,"rms_envelope_nrmse":0.08598601561780658,"envelope_correlation":0.9956401935299369,"rms_gain_error_db":-0.4308127710294258,"reference_mean_power":0.00041670963521221266,"output_mean_power":0.00037735687373636686},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9169450940862239e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.131046807277538e-05,"output_power":7.826527979495608e-05,"reference_active":true,"delta_db":-0.6695142688767195},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.750575155696516e-06,"output_power":5.476953004500012e-06,"reference_active":true,"delta_db":-0.21172269284045117},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0115145688994783e-07,"output_power":1.7865569097825942e-07,"reference_active":true,"delta_db":-0.5150632743612545}],"active_band_count":3,"active_band_mae_db":0.4654334120261418}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":1.5073703380898422,"rms_envelope_nrmse":0.13689875968491677,"envelope_correlation":0.9884446942073424,"rms_gain_error_db":-0.5803893835355818,"reference_mean_power":2.7148047549645084e-05,"output_mean_power":2.375197146459195e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3627974020623764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.843668843513599e-06,"output_power":4.236331740729769e-06,"reference_active":true,"delta_db":-0.5818448152325748},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2177208449183313e-06,"output_power":1.6597625730284343e-06,"reference_active":true,"delta_db":-1.2586091134158321},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.650596314175211e-07,"output_power":2.587003684067705e-07,"reference_active":true,"delta_db":-2.547115958915034}],"active_band_count":3,"active_band_mae_db":1.4625232958544803}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":1.0167751097390865,"rms_envelope_nrmse":0.0911229648685843,"envelope_correlation":0.9964120657238957,"rms_gain_error_db":-0.580415230806105,"reference_mean_power":2.7284479980131427e-05,"output_mean_power":2.3871194853902452e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2818152654967464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.740914469824879e-06,"output_power":4.1459922868512215e-06,"reference_active":true,"delta_db":-0.5823363054067296},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1748833712187603e-06,"output_power":8.517909025001008e-07,"reference_active":true,"delta_db":-1.3966175972599957},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4248385933815314e-08,"output_power":1.9532750953939943e-08,"reference_active":true,"delta_db":-2.4387669600601845}],"active_band_count":3,"active_band_mae_db":1.47257362090897}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8230-279154-0033","method":"magphase","status":"measured","mos":{"status":"measured","value":2.9280378818511963,"predicted_mos":2.9280378818511963,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":290869,"input_sample_rate":48000,"input_seconds":6.059770833333333,"input_float32_sha256":"62db02508c04f36d453c8960227aaf9236027009f4ae3b3bcded94882ba299c9","internal_samples":96957,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.974716824999632},"started_at":"2026-10-13T05:18:40.343774+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.010706213000958087,"native_input_samples":288000,"native_input_sha256":"94d32e642c46c306f62a559c56513d899446d5559f012ec81578cdce3237d821","output_sha256":"2fb169e1a972b21be11a068cbe0374246ed7f35b1cb29b6eadf69194b669e72f","output_samples":290869,"length_delta_samples":2869,"analysis_s":2.714112786001351,"synthesis_s":0.9461007969985076,"total_s":3.6602135829998588,"rtf":0.6100355971666431,"output_peak":0.8598324665943163,"output_rms":0.09254896610198134,"finished_at":"2026-10-13T05:18:44.028273+00:00","evaluation_input":{"reference":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"output_sha256":"2fb169e1a972b21be11a068cbe0374246ed7f35b1cb29b6eadf69194b669e72f","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96957,"length_delta_samples":957,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-20.667778597823954,"output_peak":0.8604896061200799},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9901296451004054,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.584291958392194,"spectral_convergence":0.8832934242065733,"spectral_convergence_status":"ok","amplitude_floor":2.5091999427961322e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.41571610776879,"spectral_convergence":0.6194465948901258,"spectral_convergence_status":"ok","amplitude_floor":1.1904358876245626e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.593736471410057,"spectral_convergence":0.620934830186647,"spectral_convergence_status":"ok","amplitude_floor":8.234850781853941e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.001792776939061309,"output_power":0.0016552732425461386,"delta_db":-0.3465656237506786,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0038916933601920887,"output_power":0.003698479603024105,"delta_db":-0.22115385472468024,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002833146281043118,"output_power":0.0027729754732353333,"delta_db":-0.09322970305283085,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00044617096434630785,"output_power":0.0004492683362258982,"delta_db":0.030045071282903263,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":5.1707528478122786e-05,"output_power":5.313144616928334e-05,"delta_db":0.11797857190113786,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":3.2019368729027054e-05,"output_power":3.052572939596971e-05,"delta_db":-0.20746715453137532,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1694066632072677,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1902875355620508,"pitch":null,"output_rms_dbfs":-20.624994685446065,"output_peak":0.8604896061200799,"reference_rms_dbfs":-20.434707149884012,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.531248179190348},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.4135992437003061,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.095118423251492,"rms_dbfs":-20.434707149884012,"peak":0.857177734375,"peak_dbfs":-1.3385823700484625,"crest_factor_db":19.09612477983555,"mean":-3.6449432373046876e-06,"zero_energy":false},"output":{"rms":0.09305726103102717,"rms_dbfs":-20.62499468544606,"peak":0.8604896061200799,"peak_dbfs":-1.3050874232178358,"crest_factor_db":19.319907262228227,"mean":0.0017851408136004784,"zero_energy":false},"rms_difference_db":-0.19028753556204947},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.0108933812007308e-05,"threshold_dbfs":-49.952946469786724,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":96,"quiet_samples":15360,"reference_mean_power":1.9068745435409559e-06,"output_mean_power":9.525714102522107e-05,"output_rms_dbfs":-40.211024572103504,"output_zero_energy":false,"output_peak":0.17798744774221778,"output_energy_fraction":0.0017600175321309717,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":190,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":141,"reference_voiced_output_unvoiced_frames":49,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.7421052631578947,"absolute_difference_cents_median":8.5466902390247,"absolute_difference_cents_p95":58.244378699197036,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0003179454955178215,"log_rms_mae_db":5.77412346519612,"rms_envelope_nrmse":0.4719096287154072,"envelope_correlation":0.7613905189830529,"rms_gain_error_db":-0.19031294622319395,"reference_mean_power":0.009055057750136907,"output_mean_power":0.008666823052625743},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.4928013782154505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.002203466496672066,"output_power":0.002232423602393973,"reference_active":true,"delta_db":0.0567015394240446},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002920043469136947,"output_power":0.00025369026557634823,"reference_active":true,"delta_db":-0.6108551350754875},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.410659541472349e-05,"output_power":8.44139016420447e-05,"reference_active":true,"delta_db":0.5655711250188101}],"active_band_count":3,"active_band_mae_db":0.4110425998394474}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002752533875730688,"log_rms_mae_db":4.486808254513442,"rms_envelope_nrmse":0.38664629750006446,"envelope_correlation":0.8249179031610798,"rms_gain_error_db":-0.19038231310760617,"reference_mean_power":0.009100511091100074,"output_mean_power":0.008710188463142133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.115339465742665e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019704973258355524,"output_power":0.0020110180972097037,"reference_active":true,"delta_db":0.08840129022438767},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014191960918623396,"output_power":0.00012316117391724432,"reference_active":true,"delta_db":-0.6156858679106605},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.9761972447933005e-06,"output_power":4.900841153156978e-06,"reference_active":true,"delta_db":0.908027062776017}],"active_band_count":3,"active_band_mae_db":0.537371406970355}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028904791382782896,"log_rms_mae_db":7.148600643765107,"rms_envelope_nrmse":0.5415548979430299,"envelope_correlation":0.7449562991512999,"rms_gain_error_db":-0.16992472012330187,"reference_mean_power":0.00500094100859156,"output_mean_power":0.004809049635870944},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":3.3504006393160654e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001701857204603173,"output_power":0.0017014549204411999,"reference_active":true,"delta_db":-0.0010267044931318566},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00029604441935142547,"output_power":0.00026863141782860214,"reference_active":true,"delta_db":-0.42200074453462544},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.422561443844994e-05,"output_power":7.27622052542563e-05,"reference_active":true,"delta_db":-0.08647948312926161}],"active_band_count":3,"active_band_mae_db":0.16983564405233964}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002500254339817177,"log_rms_mae_db":6.124002269085695,"rms_envelope_nrmse":0.4419978112166409,"envelope_correlation":0.8154743149334458,"rms_gain_error_db":-0.16990228713655486,"reference_mean_power":0.00502600851498449,"output_mean_power":0.004833180240823837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.0757929049004627e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0015562264230646286,"output_power":0.0015695388640817273,"reference_active":true,"delta_db":0.03699289037906581},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014399857449561677,"output_power":0.00013148866579415256,"reference_active":true,"delta_db":-0.39469874205247246},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.7762361828128385e-06,"output_power":4.7942772355365925e-06,"reference_active":true,"delta_db":0.01637349760420513}],"active_band_count":3,"active_band_mae_db":0.14935504334524782}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00010495143398294126,"log_rms_mae_db":6.664765041448763,"rms_envelope_nrmse":0.48000270991031696,"envelope_correlation":0.8223366169719787,"rms_gain_error_db":0.03626017753337602,"reference_mean_power":0.00041462711687219214,"output_mean_power":0.00041810341958184227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.072345528907365e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001011743074744314,"output_power":0.00010239012202049837,"reference_active":true,"delta_db":0.05187820174642851},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2721317302684436e-05,"output_power":1.2104077070076134e-05,"reference_active":true,"delta_db":-0.21600404885932195},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.690399523324186e-06,"output_power":2.769266057487106e-06,"reference_active":true,"delta_db":0.12547905271948015}],"active_band_count":3,"active_band_mae_db":0.1311204344417435}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.780703381807897e-05,"log_rms_mae_db":5.394901699422043,"rms_envelope_nrmse":0.37756254009935064,"envelope_correlation":0.8823768479748605,"rms_gain_error_db":0.036206941276153946,"reference_mean_power":0.00041670963521221266,"output_mean_power":0.00042019824723777446},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.9169450940862239e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":9.131046807277538e-05,"output_power":9.263529883467373e-05,"reference_active":true,"delta_db":0.06255937801593169},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.750575155696516e-06,"output_power":5.331980721651791e-06,"reference_active":true,"delta_db":-0.3282271321861491},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0115145688994783e-07,"output_power":2.0723749037045856e-07,"reference_active":true,"delta_db":0.12945141969726953}],"active_band_count":3,"active_band_mae_db":0.17341264329978343}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":2.4613391249073974e-05,"log_rms_mae_db":6.956232598932184,"rms_envelope_nrmse":0.8800630556425828,"envelope_correlation":0.4304706134765313,"rms_gain_error_db":0.08948687817990478,"reference_mean_power":2.7148047549645084e-05,"output_mean_power":2.7713239117924263e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3627974020623764e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.843668843513599e-06,"output_power":4.413121458747854e-06,"reference_active":true,"delta_db":-0.40428562889219843},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2177208449183313e-06,"output_power":2.372475845132818e-06,"reference_active":true,"delta_db":0.2929492090404988},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.650596314175211e-07,"output_power":7.53305232915587e-07,"reference_active":true,"delta_db":2.0946234124314023}],"active_band_count":3,"active_band_mae_db":0.9306194167880332}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":2.226832740151301e-05,"log_rms_mae_db":5.565729479269113,"rms_envelope_nrmse":0.7521289458560708,"envelope_correlation":0.5431487729153469,"rms_gain_error_db":0.0712428920858144,"reference_mean_power":2.7284479980131427e-05,"output_mean_power":2.773575357741296e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.2818152654967464e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.740914469824879e-06,"output_power":4.449017340695145e-06,"reference_active":true,"delta_db":-0.27598021866059597},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1748833712187603e-06,"output_power":1.3019834800701144e-06,"reference_active":true,"delta_db":0.44610716784273985},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4248385933815314e-08,"output_power":5.312127748517422e-08,"reference_active":true,"delta_db":1.9062840182538763}],"active_band_count":3,"active_band_mae_db":0.8761238015857374}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"hnm","status":"measured","mos":{"status":"measured","value":2.568856954574585,"predicted_mos":2.568856954574585,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49600,"input_sample_rate":16000,"input_seconds":3.1,"input_float32_sha256":"c34e6afae94eaded96537deff3e2734269ccb3678344c9564770a1e776349414","internal_samples":49600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7437253950010927},"started_at":"2026-10-13T04:40:24.801122+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2127000445616432e-05,"native_input_samples":49600,"native_input_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","output_sha256":"4bef3ab45921a7e87366f85778785187cf1999f4a89d966a56b4eafc1eb816a9","output_samples":49600,"length_delta_samples":0,"analysis_s":0.5424621379997916,"synthesis_s":0.2541199510005754,"total_s":0.796582089000367,"rtf":0.25696196419366674,"output_peak":0.778308374912595,"output_rms":0.06100732047035006,"finished_at":"2026-10-13T04:40:25.600928+00:00","evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"4bef3ab45921a7e87366f85778785187cf1999f4a89d966a56b4eafc1eb816a9","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.870240221955103,"spectral_convergence":0.4679213975171723,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.551347716587651,"spectral_convergence":0.44137757810320133,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9096181243527734,"spectral_convergence":0.5037411109549662,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0007662843806170188,"delta_db":-0.28630003657458053,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0005304489626018898,"delta_db":0.3182642139303917,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.0011586585255105265,"delta_db":-0.5139587840298641,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.0008292264432879042,"delta_db":0.23980453792382384,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.0003225275804795018,"delta_db":0.14020549492111023,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":0.00011474725847515312,"delta_db":0.33744881688825534,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30599698071133763,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11165377592152953,"pitch":null,"output_rms_dbfs":-24.292360988595618,"output_peak":0.778308374912595,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.777068687631842},"common_prefix":{"samples":49600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.870240221955103,"spectral_convergence":0.4679213975171723,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.551347716587651,"spectral_convergence":0.44137757810320133,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.9096181243527734,"spectral_convergence":0.5037411109549662,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0007662843806170188,"delta_db":-0.28630003657458053,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0005304489626018898,"delta_db":0.3182642139303917,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.0011586585255105265,"delta_db":-0.5139587840298641,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.0008292264432879042,"delta_db":0.23980453792382384,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.0003225275804795018,"delta_db":0.14020549492111023,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":0.00011474725847515312,"delta_db":0.33744881688825534,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.30599698071133763,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11165377592152953,"pitch":null,"output_rms_dbfs":-24.292360988595618,"output_peak":0.778308374912595,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.777068687631842},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49600,"output_samples":49600,"intelligibility":{"estoi":{"value":0.856406892273357,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06179660827551199,"rms_dbfs":-24.180707212674086,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.977421996362374,"mean":2.1208486249369958e-06,"zero_energy":false},"output":{"rms":0.06100732047035006,"rms_dbfs":-24.292360988595618,"peak":0.778308374912595,"peak_dbfs":-2.1769659268396224,"crest_factor_db":22.115395061755994,"mean":-5.090645491873521e-05,"zero_energy":false},"rms_difference_db":-0.11165377592153192},"full_scale":{"reference":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":310,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":91,"quiet_samples":14560,"reference_mean_power":1.9840398486089575e-06,"output_mean_power":1.967874810573712e-06,"output_rms_dbfs":-57.06002533346943,"output_zero_energy":false,"output_peak":0.009481728818641939,"output_energy_fraction":0.00015520770028054737,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":303,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":47,"both_estimated_voiced_frames":39,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.975,"absolute_difference_cents_median":2.4061804531300215,"absolute_difference_cents_p95":12.618169383742249,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.0002690918479937382,"log_rms_mae_db":1.0389921316662658,"rms_envelope_nrmse":0.09115183534871364,"envelope_correlation":0.9932176066400141,"rms_gain_error_db":-0.1116533967605811,"reference_mean_power":0.0038249880205223005,"output_mean_power":0.003727904168758024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":2.1583633605855604e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008154057878800692,"output_power":0.0007260007917571295,"reference_active":true,"delta_db":-0.5043669548808891},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.087495849252008e-05,"output_power":9.493881558003033e-05,"reference_active":true,"delta_db":0.18999583824537475},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5994291162280967e-05,"output_power":2.594064915539326e-05,"reference_active":true,"delta_db":-0.00897139218083674}],"active_band_count":3,"active_band_mae_db":0.23444472843570022}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.0002456731948415248,"log_rms_mae_db":0.5292015620534017,"rms_envelope_nrmse":0.05377702060306882,"envelope_correlation":0.9976385901223986,"rms_gain_error_db":-0.11165355914738792,"reference_mean_power":0.0038624153279842384,"output_mean_power":0.0037643813750610614},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.0216648233438394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007562450561775056,"output_power":0.0006703872159068503,"reference_active":true,"delta_db":-0.5233682471484571},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1399930780791805e-05,"output_power":4.368909622744246e-05,"reference_active":true,"delta_db":0.23373445792275346},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0921358976760214e-06,"output_power":1.210867472234133e-06,"reference_active":true,"delta_db":0.44819930546624176}],"active_band_count":3,"active_band_mae_db":0.40176733684581745}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00017000794064843354,"log_rms_mae_db":1.7826195715075508,"rms_envelope_nrmse":0.18721683779204346,"envelope_correlation":0.9767621693730925,"rms_gain_error_db":-0.283662253174723,"reference_mean_power":0.001461911046007658,"output_mean_power":0.0013694769352820013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.2056298552980859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006038941034712422,"output_power":0.0005048166814909671,"reference_active":true,"delta_db":-0.7782709151823787},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14042207191201e-05,"output_power":7.71524605746576e-05,"reference_active":true,"delta_db":0.33625897522602954},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.020128231954835e-05,"output_power":3.0020308336824864e-05,"reference_active":true,"delta_db":1.72036210451709}],"active_band_count":3,"active_band_mae_db":0.9449639983084994}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00015858554889566586,"log_rms_mae_db":1.0510537968912188,"rms_envelope_nrmse":0.09507517105066413,"envelope_correlation":0.9943321659279132,"rms_gain_error_db":-0.2836609741352402,"reference_mean_power":0.0014762175799508475,"output_mean_power":0.0013828792990626204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.148107626412446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000571116387197927,"output_power":0.00048692047250780145,"reference_active":true,"delta_db":-0.6926658688865815},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.637477969141665e-05,"output_power":3.727843816718877e-05,"reference_active":true,"delta_db":0.10657337041544063},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1727904534817564e-06,"output_power":1.5180412102207092e-06,"reference_active":true,"delta_db":1.1206313934630363}],"active_band_count":3,"active_band_mae_db":0.6399568775883528}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00019136572750236332,"log_rms_mae_db":1.4852577944391088,"rms_envelope_nrmse":0.13411751259634166,"envelope_correlation":0.9893800596204557,"rms_gain_error_db":0.15469157516927218,"reference_mean_power":0.0009651549113833476,"output_mean_power":0.0010001524025159655},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":7.999024043731724e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033597435990848984,"output_power":0.00034911797159032244,"reference_active":true,"delta_db":0.16666070420113144},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.996653108292046e-05,"output_power":3.348331918041128e-05,"reference_active":true,"delta_db":0.4819203074448042},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.661292025781762e-06,"output_power":1.0176555889171109e-05,"reference_active":true,"delta_db":1.8404234886538342}],"active_band_count":3,"active_band_mae_db":0.82966816676659}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00016758862552795114,"log_rms_mae_db":0.9836875423724737,"rms_envelope_nrmse":0.0642137137602752,"envelope_correlation":0.9976924580447908,"rms_gain_error_db":0.15469038498180454,"reference_mean_power":0.0009746009072787486,"output_mean_power":0.0010099406429804744},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":7.731649944214165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003128075285039233,"output_power":0.00033016353093315884,"reference_active":true,"delta_db":0.23451903547422157},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.538593181542651e-05,"output_power":1.7513020483714358e-05,"reference_active":true,"delta_db":0.5623725215821328},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119257970392287e-07,"output_power":4.388521465634224e-07,"reference_active":true,"delta_db":1.4826693349199345}],"active_band_count":3,"active_band_mae_db":0.759853630658763}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":1.6521577469844944,"rms_envelope_nrmse":0.21426747641695,"envelope_correlation":0.9710809380364768,"rms_gain_error_db":0.3430135737082073,"reference_mean_power":0.00014982208037105638,"output_mean_power":0.000162135151084258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.5994961899965224e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.519717012759742e-05,"output_power":6.028247004265363e-05,"reference_active":true,"delta_db":0.3827422629371005},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.568381171011351e-05,"output_power":1.506754562478211e-05,"reference_active":true,"delta_db":-0.17409104602677836},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1546072478197707e-06,"output_power":3.5184027973320225e-06,"reference_active":true,"delta_db":0.47400260515665726}],"active_band_count":3,"active_band_mae_db":0.343611971373512}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":1.1859526986399014,"rms_envelope_nrmse":0.10676681803125813,"envelope_correlation":0.9933440452396315,"rms_gain_error_db":0.3430142273749065,"reference_mean_power":0.00015128837405213323,"output_mean_power":0.00016372197619620444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.4928738695410593e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.969547295575367e-05,"output_power":5.426417595785679e-05,"reference_active":true,"delta_db":0.3819638413695309},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.547268075821191e-06,"output_power":7.976888398839697e-06,"reference_active":true,"delta_db":-0.29993809280191136},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.762826939609959e-07,"output_power":1.925580129793635e-07,"reference_active":true,"delta_db":0.3835191697392112}],"active_band_count":3,"active_band_mae_db":0.35514036797021786}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.358591318130493,"predicted_mos":3.358591318130493,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49600,"input_sample_rate":16000,"input_seconds":3.1,"input_float32_sha256":"fa0c2af10f85b47a0d35379fb9eb6f24f821409491344a8c13b3412da54b9c9e","internal_samples":49600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4997022390016355},"started_at":"2026-10-13T04:43:40.788350+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0389999917824753e-05,"native_input_samples":49600,"native_input_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","output_sha256":"a720536d4a3c25d1d239df9477a31e93d4bacc8eabb2078c9eb6ab40e7ecb33d","output_samples":49600,"length_delta_samples":0,"analysis_s":0.8414823409993915,"synthesis_s":0.04138024600069912,"total_s":0.8828625870000906,"rtf":0.28479438290325504,"output_peak":1.2275139940345958,"output_rms":0.0675485879449491,"finished_at":"2026-10-13T04:43:41.674770+00:00","evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"a720536d4a3c25d1d239df9477a31e93d4bacc8eabb2078c9eb6ab40e7ecb33d","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5446109162826955,"spectral_convergence":0.5364611218881071,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.580113017591977,"spectral_convergence":0.2996003132374404,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.840256344015634,"spectral_convergence":0.337760258217345,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0009398905828345733,"delta_db":0.600573212793518,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0006121777095339445,"delta_db":0.9406032911883115,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.001686818801905952,"delta_db":1.1171709321693615,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.000857497577395828,"delta_db":0.385402126347858,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.000334225468865338,"delta_db":0.29493232075791664,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":0.00013220159282088651,"delta_db":0.9523925120165445,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7151790658789183,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7730327094053326,"pitch":null,"output_rms_dbfs":-23.407674503268755,"output_peak":1.2275139940345958,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.988326759296768},"common_prefix":{"samples":49600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.5446109162826955,"spectral_convergence":0.5364611218881071,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.580113017591977,"spectral_convergence":0.2996003132374404,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.840256344015634,"spectral_convergence":0.337760258217345,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0009398905828345733,"delta_db":0.600573212793518,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0006121777095339445,"delta_db":0.9406032911883115,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.001686818801905952,"delta_db":1.1171709321693615,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.000857497577395828,"delta_db":0.385402126347858,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.000334225468865338,"delta_db":0.29493232075791664,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":0.00013220159282088651,"delta_db":0.9523925120165445,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7151790658789183,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7730327094053326,"pitch":null,"output_rms_dbfs":-23.407674503268755,"output_peak":1.2275139940345958,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.988326759296768},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49600,"output_samples":49600,"intelligibility":{"estoi":{"value":0.8881489188067593,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06179660827551199,"rms_dbfs":-24.180707212674086,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.977421996362374,"mean":2.1208486249369958e-06,"zero_energy":false},"output":{"rms":0.0675485879449491,"rms_dbfs":-23.407674503268755,"peak":1.2275139940345958,"peak_dbfs":1.780529038260391,"crest_factor_db":25.18820354152915,"mean":2.948555688569461e-09,"zero_energy":false},"rms_difference_db":0.7730327094053315},"full_scale":{"reference":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":10,"at_or_above_full_scale_fraction":0.00020161290322580645,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.00020161290322580645,"longest_full_scale_run_samples":2,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":310,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":91,"quiet_samples":14560,"reference_mean_power":1.9840398486089575e-06,"output_mean_power":9.517298444232843e-07,"output_rms_dbfs":-60.214863118326775,"output_zero_energy":false,"output_peak":0.009538662489035761,"output_energy_fraction":6.122951747930232e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":303,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":47,"both_estimated_voiced_frames":31,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.775,"absolute_difference_cents_median":13.468994276551053,"absolute_difference_cents_p95":24.477625511952105,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.0002690918479937382,"log_rms_mae_db":2.4486314554121447,"rms_envelope_nrmse":0.1538553018244436,"envelope_correlation":0.9887346747889804,"rms_gain_error_db":0.7730344030066463,"reference_mean_power":0.0038249880205223005,"output_mean_power":0.00457018225409778},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":2.1583633605855604e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008154057878800692,"output_power":0.0009336782835510558,"reference_active":true,"delta_db":0.588234678385577},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.087495849252008e-05,"output_power":0.00010282098730483151,"reference_active":true,"delta_db":0.5363754421996363},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5994291162280967e-05,"output_power":3.059162176491055e-05,"reference_active":true,"delta_db":0.7072452184330655}],"active_band_count":3,"active_band_mae_db":0.6106184463394263}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.0002456731948415248,"log_rms_mae_db":1.8948864583257161,"rms_envelope_nrmse":0.1258150504486902,"envelope_correlation":0.994407753795342,"rms_gain_error_db":0.7730399444672318,"reference_mean_power":0.0038624153279842384,"output_mean_power":0.004614907136580771},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.0216648233438394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007562450561775056,"output_power":0.0008575416053960732,"reference_active":true,"delta_db":0.5459265137263776},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1399930780791805e-05,"output_power":4.806493012985827e-05,"reference_active":true,"delta_db":0.6482870033473852},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0921358976760214e-06,"output_power":1.1565924514438596e-06,"reference_active":true,"delta_db":0.24903671264460814}],"active_band_count":3,"active_band_mae_db":0.48108340990612364}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00017000794064843354,"log_rms_mae_db":2.2581260042628246,"rms_envelope_nrmse":0.20055710753711495,"envelope_correlation":0.9860917232526516,"rms_gain_error_db":1.0490467076061851,"reference_mean_power":0.001461911046007658,"output_mean_power":0.0018613396055149258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.2056298552980859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006038941034712422,"output_power":0.0007621637151505017,"reference_active":true,"delta_db":1.0108748007278736},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14042207191201e-05,"output_power":8.090560054887917e-05,"reference_active":true,"delta_db":0.5425470215295007},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.020128231954835e-05,"output_power":1.4653795472952398e-05,"reference_active":true,"delta_db":-1.3942881241364338}],"active_band_count":3,"active_band_mae_db":0.9825699821312694}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00015858554889566586,"log_rms_mae_db":1.6494754361090027,"rms_envelope_nrmse":0.17582222400796907,"envelope_correlation":0.9912342432437352,"rms_gain_error_db":1.0490448429913302,"reference_mean_power":0.0014762175799508475,"output_mean_power":0.0018795542148059452},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.148107626412446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000571116387197927,"output_power":0.0007121626656227986,"reference_active":true,"delta_db":0.9585458083116422},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.637477969141665e-05,"output_power":4.123717336359369e-05,"reference_active":true,"delta_db":0.544885169136724},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1727904534817564e-06,"output_power":7.457744530229775e-07,"reference_active":true,"delta_db":-1.9661291990580352}],"active_band_count":3,"active_band_mae_db":1.156520058835467}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00019136572750236332,"log_rms_mae_db":1.4051770968493051,"rms_envelope_nrmse":0.09781953207443508,"envelope_correlation":0.9953925726463547,"rms_gain_error_db":0.3540507736424163,"reference_mean_power":0.0009651549113833476,"output_mean_power":0.0010471336171072872},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":7.999024043731724e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033597435990848984,"output_power":0.0003454990911077824,"reference_active":true,"delta_db":0.1214077402671064},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.996653108292046e-05,"output_power":3.585942305595461e-05,"reference_active":true,"delta_db":0.7796682584672886},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.661292025781762e-06,"output_power":5.7564593697853225e-06,"reference_active":true,"delta_db":-0.6340302952768986}],"active_band_count":3,"active_band_mae_db":0.5117020980037645}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00016758862552795114,"log_rms_mae_db":0.9571018131684818,"rms_envelope_nrmse":0.07036480618817328,"envelope_correlation":0.9980760030770129,"rms_gain_error_db":0.35404974303421427,"reference_mean_power":0.0009746009072787486,"output_mean_power":0.0010573816897768108},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":7.731649944214165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003128075285039233,"output_power":0.0003194806848539491,"reference_active":true,"delta_db":0.09167409827522503},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.538593181542651e-05,"output_power":1.950585020118899e-05,"reference_active":true,"delta_db":1.0304108116846564},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119257970392287e-07,"output_power":2.546390717880745e-07,"reference_active":true,"delta_db":-0.8812625095337691}],"active_band_count":3,"active_band_mae_db":0.6677824731645502}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":1.5926057356710481,"rms_envelope_nrmse":0.23864891530227383,"envelope_correlation":0.9700846975594072,"rms_gain_error_db":0.7711149222112619,"reference_mean_power":0.00014982208037105638,"output_mean_power":0.00017893171126066894},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.5994961899965224e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.519717012759742e-05,"output_power":6.904041505783838e-05,"reference_active":true,"delta_db":0.9718658090827142},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.568381171011351e-05,"output_power":2.1068630083270422e-05,"reference_active":true,"delta_db":1.2818467819414525},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1546072478197707e-06,"output_power":9.219273670328094e-06,"reference_active":true,"delta_db":4.657514102293589}],"active_band_count":3,"active_band_mae_db":2.3037422311059186}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":1.071969428389309,"rms_envelope_nrmse":0.17034759059872473,"envelope_correlation":0.9867347119860517,"rms_gain_error_db":0.7711154446231776,"reference_mean_power":0.00015128837405213323,"output_mean_power":0.00018068291971535982},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.4928738695410593e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.969547295575367e-05,"output_power":6.51490869947736e-05,"reference_active":true,"delta_db":1.1759150567660388},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.547268075821191e-06,"output_power":1.1176834612368667e-05,"reference_active":true,"delta_db":1.1649149921794064},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.762826939609959e-07,"output_power":4.826919398232351e-07,"reference_active":true,"delta_db":4.374603680632408}],"active_band_count":3,"active_band_mae_db":2.2384779098592844}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"facodec","status":"measured","mos":{"status":"measured","value":4.904283046722412,"predicted_mos":4.904283046722412,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49600,"input_sample_rate":16000,"input_seconds":3.1,"input_float32_sha256":"964bdc05e3b6d454b48a74025678e9bf2d9948b84286300453d91fb4e42d08a9","internal_samples":49600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.87482812700182},"started_at":"2026-10-13T07:01:14.171382+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":4.041399915877264e-05,"native_input_samples":49600,"native_input_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","output_sha256":"6d0a198bdcd43ce67ab336ee1e003f55bae2d4fac3ab98264fa9866b39b74711","output_samples":49600,"length_delta_samples":0,"analysis_s":0.8295641009990504,"synthesis_s":5.454784265000853,"total_s":6.284348365999904,"rtf":2.0272091503225496,"output_peak":0.7730695009231567,"output_rms":0.057178339697037915,"finished_at":"2026-10-13T07:01:20.465213+00:00","evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"6d0a198bdcd43ce67ab336ee1e003f55bae2d4fac3ab98264fa9866b39b74711","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.468111763845887,"spectral_convergence":0.2923068291240618,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.719582967950171,"spectral_convergence":0.32635229564762674,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.836280751662258,"spectral_convergence":0.35223480073058644,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0007957142663285611,"delta_db":-0.12262832382899179,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0005255455130027294,"delta_db":0.27793149518974136,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.0009734227330112939,"delta_db":-1.2704985555574286,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.0006521691847702384,"delta_db":-0.8033241472852105,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.00023901748265940515,"delta_db":-1.1611664105027695,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":8.349335073763435e-05,"delta_db":-1.0434554612521505,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7798340656027154,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.674661977758646,"pitch":null,"output_rms_dbfs":-24.855369190432732,"output_peak":0.7730695009231567,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.674658494486105},"common_prefix":{"samples":49600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.468111763845887,"spectral_convergence":0.2923068291240618,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.719582967950171,"spectral_convergence":0.32635229564762674,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.836280751662258,"spectral_convergence":0.35223480073058644,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0007957142663285611,"delta_db":-0.12262832382899179,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0005255455130027294,"delta_db":0.27793149518974136,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.0009734227330112939,"delta_db":-1.2704985555574286,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.0006521691847702384,"delta_db":-0.8033241472852105,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.00023901748265940515,"delta_db":-1.1611664105027695,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":8.349335073763435e-05,"delta_db":-1.0434554612521505,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7798340656027154,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.674661977758646,"pitch":null,"output_rms_dbfs":-24.855369190432732,"output_peak":0.7730695009231567,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.674658494486105},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49600,"output_samples":49600,"intelligibility":{"estoi":{"value":0.8847673718523311,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06179660827551199,"rms_dbfs":-24.180707212674086,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.977421996362374,"mean":2.1208486249369958e-06,"zero_energy":false},"output":{"rms":0.057178339697037915,"rms_dbfs":-24.855369190432732,"peak":0.7730695009231567,"peak_dbfs":-2.2356292028606664,"crest_factor_db":22.619739987572068,"mean":-7.763370606605878e-05,"zero_energy":false},"rms_difference_db":-0.674661977758646},"full_scale":{"reference":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":310,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":91,"quiet_samples":14560,"reference_mean_power":1.9840398486089575e-06,"output_mean_power":1.9426816567136752e-06,"output_rms_dbfs":-57.11598360522062,"output_zero_energy":false,"output_peak":0.007942034862935543,"output_energy_fraction":0.00017442882569583185,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":303,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":36,"both_estimated_voiced_frames":32,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.8,"absolute_difference_cents_median":6.053959461915281,"absolute_difference_cents_p95":17.026481807100875,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.0002690918479937382,"log_rms_mae_db":1.1753043462742718,"rms_envelope_nrmse":0.13025552823008915,"envelope_correlation":0.9901646762976323,"rms_gain_error_db":-0.6746627362666995,"reference_mean_power":0.0038249880205223005,"output_mean_power":0.0032746418340292133},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":2.1583633605855604e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008154057878800692,"output_power":0.0006745575278602428,"reference_active":true,"delta_db":-0.8235479657125516},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.087495849252008e-05,"output_power":8.452188703205307e-05,"reference_active":true,"delta_db":-0.31475040972861756},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5994291162280967e-05,"output_power":3.1805078234907835e-05,"reference_active":true,"delta_db":0.8761848903705909}],"active_band_count":3,"active_band_mae_db":0.6714944219372533}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.0002456731948415248,"log_rms_mae_db":0.9076191578175953,"rms_envelope_nrmse":0.12284758922113684,"envelope_correlation":0.991622209648822,"rms_gain_error_db":-0.6746708959947121,"reference_mean_power":0.0038624153279842384,"output_mean_power":0.0033066778201316523},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.0216648233438394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007562450561775056,"output_power":0.000626747241110367,"reference_active":true,"delta_db":-0.8157011763804096},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1399930780791805e-05,"output_power":3.821701179090899e-05,"reference_active":true,"delta_db":-0.3474288867359486},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0921358976760214e-06,"output_power":1.3734479817023122e-06,"reference_active":true,"delta_db":0.9953553322628962}],"active_band_count":3,"active_band_mae_db":0.7194951317930848}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00017000794064843354,"log_rms_mae_db":1.700962064094349,"rms_envelope_nrmse":0.16671341349645696,"envelope_correlation":0.9866855788633602,"rms_gain_error_db":-0.8870753967903748,"reference_mean_power":0.001461911046007658,"output_mean_power":0.0011918275161200728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.2056298552980859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006038941034712422,"output_power":0.0004973565250285047,"reference_active":true,"delta_db":-0.8429296909466493},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14042207191201e-05,"output_power":5.390355347120044e-05,"reference_active":true,"delta_db":-1.2210648775400572},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.020128231954835e-05,"output_power":2.5646483156361394e-05,"reference_active":true,"delta_db":1.0364888162908314}],"active_band_count":3,"active_band_mae_db":1.0334944615925128}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00015858554889566586,"log_rms_mae_db":1.1422076967885093,"rms_envelope_nrmse":0.15525435747851266,"envelope_correlation":0.9890723124684258,"rms_gain_error_db":-0.887076645486051,"reference_mean_power":0.0014762175799508475,"output_mean_power":0.0012034906162750048},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.148107626412446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000571116387197927,"output_power":0.0004746311160186602,"reference_active":true,"delta_db":-0.8036841524199207},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.637477969141665e-05,"output_power":2.7039706601831784e-05,"reference_active":true,"delta_db":-1.287983967195787},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1727904534817564e-06,"output_power":1.5469458869444362e-06,"reference_active":true,"delta_db":1.20254699941777}],"active_band_count":3,"active_band_mae_db":1.0980717063444925}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00019136572750236332,"log_rms_mae_db":1.3071655040379797,"rms_envelope_nrmse":0.16291416621202529,"envelope_correlation":0.9890661612895432,"rms_gain_error_db":-0.9281171392938837,"reference_mean_power":0.0009651549113833476,"output_mean_power":0.0007794447049636499},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":7.999024043731724e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033597435990848984,"output_power":0.00025785955008310015,"reference_active":true,"delta_db":-1.1492291463626132},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.996653108292046e-05,"output_power":2.8193014359443067e-05,"reference_active":true,"delta_db":-0.26494959509955435},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.661292025781762e-06,"output_power":5.5322145002118995e-06,"reference_active":true,"delta_db":-0.8065946256439928}],"active_band_count":3,"active_band_mae_db":0.7402577890353869}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00016758862552795114,"log_rms_mae_db":1.1300593864990467,"rms_envelope_nrmse":0.1486704959285409,"envelope_correlation":0.9920234796994115,"rms_gain_error_db":-0.9281173861108659,"reference_mean_power":0.0009746009072787486,"output_mean_power":0.0007870731055705115},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":7.731649944214165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003128075285039233,"output_power":0.00024136921097116479,"reference_active":true,"delta_db":-1.1259532616805448},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.538593181542651e-05,"output_power":1.3593466897437611e-05,"reference_active":true,"delta_db":-0.5379356952709148},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119257970392287e-07,"output_power":2.492808207890714e-07,"reference_active":true,"delta_db":-0.9736242751116475}],"active_band_count":3,"active_band_mae_db":0.879171077354369}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":1.6012861552414377,"rms_envelope_nrmse":0.21556268449883095,"envelope_correlation":0.9698546116724319,"rms_gain_error_db":-0.664897319886117,"reference_mean_power":0.00014982208037105638,"output_mean_power":0.0001285541470642617},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.5994961899965224e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.519717012759742e-05,"output_power":5.3587346325038845e-05,"reference_active":true,"delta_db":-0.1285456162881753},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.568381171011351e-05,"output_power":1.3940534061489682e-05,"reference_active":true,"delta_db":-0.5117220790486505},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1546072478197707e-06,"output_power":3.6777136979226463e-06,"reference_active":true,"delta_db":0.6663262053925494}],"active_band_count":3,"active_band_mae_db":0.4355313002431251}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":1.2597976027738018,"rms_envelope_nrmse":0.1809927093440668,"envelope_correlation":0.9789369178535985,"rms_gain_error_db":-0.6648985406841358,"reference_mean_power":0.00015128837405213323,"output_mean_power":0.00012981225712403618},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.4928738695410593e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.969547295575367e-05,"output_power":4.930340527027933e-05,"reference_active":true,"delta_db":-0.03439912160413169},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.547268075821191e-06,"output_power":7.300057321941163e-06,"reference_active":true,"delta_db":-0.685010550229604},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.762826939609959e-07,"output_power":1.9272078615501906e-07,"reference_active":true,"delta_db":0.38718879788142124}],"active_band_count":3,"active_band_mae_db":0.368866156571719}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.715036392211914,"predicted_mos":3.715036392211914,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":137216,"input_sample_rate":44100,"input_seconds":3.111473922902494,"input_float32_sha256":"4e17af16f8d889fb3d5f384fa858a8a670e254d44d53c306f88dbc030a468736","internal_samples":49784,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.874068144003104},"started_at":"2026-10-13T07:31:03.699175+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003973217997554457,"native_input_samples":136710,"native_input_sha256":"07387845c99dfa8112eb5bce12570b1cf212c57415740f796d097fecb162a782","output_sha256":"992594ff133b52dbc73a06b2bf4dd28bf16aa02fbfcb28682162feddf5cd28de","output_samples":137216,"length_delta_samples":506,"analysis_s":0.0767809130011301,"synthesis_s":16.221037035000336,"total_s":16.297817948001466,"rtf":5.25736062838757,"output_peak":0.8709110021591187,"output_rms":0.0557749934338951,"finished_at":"2026-10-13T07:31:20.018082+00:00","evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"992594ff133b52dbc73a06b2bf4dd28bf16aa02fbfcb28682162feddf5cd28de","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":49600,"output_samples":49784,"length_delta_samples":184,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.06731240657727,"output_peak":0.8584123769629373},"common_prefix":{"samples":49600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9963040334243933,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.837987243906626,"spectral_convergence":0.49266737258966936,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.791683658336114,"spectral_convergence":0.4693052843770135,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.67644980895429,"spectral_convergence":0.5889894098884694,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0004981885957737191,"delta_db":-2.156261952491301,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0004927435213775517,"delta_db":-0.0019626014701934896,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.001236983658105975,"delta_db":-0.22987377793228964,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.0006079864897704411,"delta_db":-1.1079876030524378,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.0002189583907872506,"delta_db":-1.5418471652154815,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":7.033160531643363e-05,"delta_db":-1.7884690626077826,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1377336937949145,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8705253853089765,"pitch":null,"output_rms_dbfs":-25.051232597983066,"output_peak":0.8584123769629373,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.768706903732344},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49600,"output_samples":49600,"intelligibility":{"estoi":{"value":0.7078553506757894,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06179660827551199,"rms_dbfs":-24.180707212674086,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.977421996362374,"mean":2.1208486249369958e-06,"zero_energy":false},"output":{"rms":0.05590341904688273,"rms_dbfs":-25.051232597983063,"peak":0.8584123769629373,"peak_dbfs":-1.3260805830053204,"crest_factor_db":23.725152014977745,"mean":-1.1540955289691545e-05,"zero_energy":false},"rms_difference_db":-0.8705253853089765},"full_scale":{"reference":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":310,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":91,"quiet_samples":14560,"reference_mean_power":1.9840398486089575e-06,"output_mean_power":1.104068782857744e-06,"output_rms_dbfs":-59.5700386946457,"output_zero_energy":false,"output_peak":0.005240445631150171,"output_energy_fraction":0.00010370485505248907,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":303,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":23,"both_estimated_voiced_frames":18,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.45,"absolute_difference_cents_median":25.358473167028535,"absolute_difference_cents_p95":55.16418537174225,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.0002690918479937382,"log_rms_mae_db":2.4082324706170724,"rms_envelope_nrmse":0.15018116077451546,"envelope_correlation":0.9897322308158739,"rms_gain_error_db":-0.8705254787297756,"reference_mean_power":0.0038249880205223005,"output_mean_power":0.0031302392406531094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":2.1583633605855604e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008154057878800692,"output_power":0.0007170205893175451,"reference_active":true,"delta_db":-0.5584216316233575},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.087495849252008e-05,"output_power":6.406749346473776e-05,"reference_active":true,"delta_db":-1.5180649223426859},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5994291162280967e-05,"output_power":2.1839984004752126e-05,"reference_active":true,"delta_db":-0.7562566320234929}],"active_band_count":3,"active_band_mae_db":0.9442477286631789}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.0002456731948415248,"log_rms_mae_db":2.260664533780156,"rms_envelope_nrmse":0.13334955869266366,"envelope_correlation":0.993815735514032,"rms_gain_error_db":-0.8705222510320418,"reference_mean_power":0.0038624153279842384,"output_mean_power":0.0031608708168625005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.0216648233438394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007562450561775056,"output_power":0.0006681378659495785,"reference_active":true,"delta_db":-0.5379646294967262},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1399930780791805e-05,"output_power":3.0230942053835646e-05,"reference_active":true,"delta_db":-1.3654793416323232},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0921358976760214e-06,"output_power":9.062379402404366e-07,"reference_active":true,"delta_db":-0.8103444202402453}],"active_band_count":3,"active_band_mae_db":0.9045961304564316}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00017000794064843354,"log_rms_mae_db":2.970554202702645,"rms_envelope_nrmse":0.17805642590015047,"envelope_correlation":0.980519556568186,"rms_gain_error_db":0.15850814314702139,"reference_mean_power":0.001461911046007658,"output_mean_power":0.0015162533098397328},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.2056298552980859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006038941034712422,"output_power":0.0005980668581274944,"reference_active":true,"delta_db":-0.04211052448983215},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14042207191201e-05,"output_power":4.407235128589991e-05,"reference_active":true,"delta_db":-2.0955766278241645},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.020128231954835e-05,"output_power":1.163815652797687e-05,"reference_active":true,"delta_db":-2.3949474412923504}],"active_band_count":3,"active_band_mae_db":1.5108781978687824}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00015858554889566586,"log_rms_mae_db":2.1441133112056474,"rms_envelope_nrmse":0.1506757351187327,"envelope_correlation":0.9860147518498673,"rms_gain_error_db":0.15850910956539208,"reference_mean_power":0.0014762175799508475,"output_mean_power":0.0015310919880120208},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.148107626412446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000571116387197927,"output_power":0.0005628956657126224,"reference_active":true,"delta_db":-0.06296717060614036},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.637477969141665e-05,"output_power":2.2375654868678147e-05,"reference_active":true,"delta_db":-2.110246169628565},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1727904534817564e-06,"output_power":7.213841459296186e-07,"reference_active":true,"delta_db":-2.1105382868253004}],"active_band_count":3,"active_band_mae_db":1.427917209020002}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00019136572750236332,"log_rms_mae_db":2.793485851614173,"rms_envelope_nrmse":0.21258968137716494,"envelope_correlation":0.9845859421721883,"rms_gain_error_db":-1.4637975422924878,"reference_mean_power":0.0009651549113833476,"output_mean_power":0.0006889969045837164},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":7.999024043731724e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033597435990848984,"output_power":0.00022367652709324236,"reference_active":true,"delta_db":-1.7668572414396044},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.996653108292046e-05,"output_power":2.1719054454598504e-05,"reference_active":true,"delta_db":-1.3979555785927211},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.661292025781762e-06,"output_power":5.21537533816567e-06,"reference_active":true,"delta_db":-1.0627290417252275}],"active_band_count":3,"active_band_mae_db":1.409180620585851}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00016758862552795114,"log_rms_mae_db":2.756411121374082,"rms_envelope_nrmse":0.1926716348051133,"envelope_correlation":0.9902574768132145,"rms_gain_error_db":-1.463796533620293,"reference_mean_power":0.0009746009072787486,"output_mean_power":0.0006957402965663654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":7.731649944214165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003128075285039233,"output_power":0.0002066471426199864,"reference_active":true,"delta_db":-1.8004779238577244},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.538593181542651e-05,"output_power":1.1334195111977316e-05,"reference_active":true,"delta_db":-1.32733118966628},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119257970392287e-07,"output_power":2.6803136027788844e-07,"reference_active":true,"delta_db":-0.6586568303387745}],"active_band_count":3,"active_band_mae_db":1.2621553146209263}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":4.161124597276907,"rms_envelope_nrmse":0.24344459165212365,"envelope_correlation":0.9649348576405804,"rms_gain_error_db":-1.098254240294048,"reference_mean_power":0.00014982208037105638,"output_mean_power":0.00011634571668576711},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.5994961899965224e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.519717012759742e-05,"output_power":4.395674228414256e-05,"reference_active":true,"delta_db":-0.9889131412955046},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.568381171011351e-05,"output_power":1.0325137355041312e-05,"reference_active":true,"delta_db":-1.815557820202439},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1546072478197707e-06,"output_power":2.035633652452319e-06,"reference_active":true,"delta_db":-1.9024567491045987}],"active_band_count":3,"active_band_mae_db":1.5689759035341808}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":4.035690198378383,"rms_envelope_nrmse":0.20161721328172208,"envelope_correlation":0.978442077160591,"rms_gain_error_db":-1.0982507717101493,"reference_mean_power":0.00015128837405213323,"output_mean_power":0.00011748447438282846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.4928738695410593e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.969547295575367e-05,"output_power":3.9932816954600015e-05,"reference_active":true,"delta_db":-0.9498688074562972},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.547268075821191e-06,"output_power":5.1296437746895455e-06,"reference_active":true,"delta_db":-2.2174011853372555},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.762826939609959e-07,"output_power":8.925452193320674e-08,"reference_active":true,"delta_db":-2.955794506431557}],"active_band_count":3,"active_band_mae_db":2.0410214997417033}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"golf","status":"measured","mos":{"status":"measured","value":3.614272117614746,"predicted_mos":3.614272117614746,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":74160,"input_sample_rate":24000,"input_seconds":3.09,"input_float32_sha256":"508cf4cf1c25b2366a2edf2c5d71ef056a1b28c882cc4a36783d60fb18c35830","internal_samples":49440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4863394720014185},"started_at":"2026-10-13T05:03:06.988494+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002464678999785974,"native_input_samples":74400,"native_input_sha256":"d65be13a8abb4ee178a2e53c070bddeaf7db83549cb810fe523797381e8242ac","output_sha256":"f6f077cd8ffae4cf3b793ce29446bf13df0aa211772865c669fec2dff4011437","output_samples":74160,"length_delta_samples":-240,"analysis_s":0.3195899479997024,"synthesis_s":0.12768064900046738,"total_s":0.4472705970001698,"rtf":0.14428083774199024,"output_peak":0.9735348224639893,"output_rms":0.055102103201157916,"finished_at":"2026-10-13T05:03:07.449139+00:00","analysis_stages":{"external_f0_s":0.03760333900027035},"evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"f6f077cd8ffae4cf3b793ce29446bf13df0aa211772865c669fec2dff4011437","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":49600,"output_samples":49440,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.17721941431429,"output_peak":0.9742955586103921},"common_prefix":{"samples":49440,"sample_rate":16000,"reference_coverage":0.9967741935483871,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49440,"output_samples":49440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.325266334798219,"spectral_convergence":0.7742442147350405,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.584307029767305,"spectral_convergence":0.3887102196029253,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.734684772878546,"spectral_convergence":0.42633906352167394,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008211463271324471,"output_power":0.0007051843875699867,"delta_db":-0.6611786601697178,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004945643818342728,"output_power":0.0004809238447271316,"delta_db":-0.12146525179727598,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00130844193176701,"output_power":0.0010884820023737006,"delta_db":-0.799332010076125,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007872187269302723,"output_power":0.00046068148683758224,"delta_db":-2.326946570169275,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003132921834035494,"output_power":0.00021214501375993886,"delta_db":-1.6931673083134366,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001065122792965915,"output_power":8.841710148505615e-05,"delta_db":-0.8086340478031518,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.068453974721497,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0105405382156676,"pitch":null,"output_rms_dbfs":-25.17721941431429,"output_peak":0.9742955586103921,"reference_rms_dbfs":-24.166678876098622,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.214752712481356},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49440,"output_samples":49440,"intelligibility":{"estoi":{"value":0.876224440396545,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06189649488346454,"rms_dbfs":-24.166678876098622,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.96339365978691,"mean":1.4900775403266587e-06,"zero_energy":false},"output":{"rms":0.05509840529139641,"rms_dbfs":-25.17721941431429,"peak":0.9742955586103921,"peak_dbfs":-0.2261855441056895,"crest_factor_db":24.9510338702086,"mean":0.0006447822684668259,"zero_energy":false},"rms_difference_db":-1.0105405382156682},"full_scale":{"reference":{"sample_count":49440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":309,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":90,"quiet_samples":14400,"reference_mean_power":1.9945526987107264e-06,"output_mean_power":2.4458354226287914e-06,"output_rms_dbfs":-56.11572769478329,"output_zero_energy":false,"output_peak":0.009354536216182764,"output_energy_fraction":0.0002346568313588954,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":302,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":64,"both_estimated_voiced_frames":37,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.925,"absolute_difference_cents_median":5.014735514157994,"absolute_difference_cents_p95":12.155966302785469,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":617,"rms_floor":0.0002690918479937382,"log_rms_mae_db":1.9177981705433678,"rms_envelope_nrmse":0.16658348852543164,"envelope_correlation":0.9854477218295007,"rms_gain_error_db":-1.010542686077894,"reference_mean_power":0.00383738298500236,"output_mean_power":0.0030407511309912404},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":617,"frequency_resolution_hz":0.3241491085899514,"power_floor":2.175170752869687e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008210315779412783,"output_power":0.0006441193363921289,"reference_active":true,"delta_db":-1.0539352414621381},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.072770465284996e-05,"output_power":8.755120303227146e-05,"reference_active":true,"delta_db":-0.15477805613918624},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.60932841678556e-05,"output_power":2.4230865273732185e-05,"reference_active":true,"delta_db":-0.32159821078128004}],"active_band_count":3,"active_band_mae_db":0.5101038361275348}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":611,"rms_floor":0.0002456731948415248,"log_rms_mae_db":1.3336239777717571,"rms_envelope_nrmse":0.125959468303409,"envelope_correlation":0.9963431932582869,"rms_gain_error_db":-1.0105565046054874,"reference_mean_power":0.0038750544835644916,"output_mean_power":0.003070592343238873},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":611,"frequency_resolution_hz":0.3273322422258592,"power_floor":2.0380313521344335e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000761397882478137,"output_power":0.0005979688472821108,"reference_active":true,"delta_db":-1.0493310576972354},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127435982373673e-05,"output_power":4.055538327627685e-05,"reference_active":true,"delta_db":-0.07631835826521352},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0953397301352335e-06,"output_power":1.125464901956628e-06,"reference_active":true,"delta_db":0.11783115238423554}],"active_band_count":3,"active_band_mae_db":0.4144935227822282}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":617,"rms_floor":0.00017000794064843354,"log_rms_mae_db":1.9961698307218032,"rms_envelope_nrmse":0.16722341306247573,"envelope_correlation":0.9824320900193011,"rms_gain_error_db":-0.45423210043093626,"reference_mean_power":0.0014666488115299247,"output_mean_power":0.0013210003254465934},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":617,"frequency_resolution_hz":0.3241491085899514,"power_floor":1.2142410719227126e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006078997239411175,"output_power":0.0005510823475349748,"reference_active":true,"delta_db":-0.42615446389517975},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.1494191657855e-05,"output_power":7.488053464343508e-05,"reference_active":true,"delta_db":0.200981763679845},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0343777560143512e-05,"output_power":1.7203157130731686e-05,"reference_active":true,"delta_db":-0.7282344246615648}],"active_band_count":3,"active_band_mae_db":0.45179021741219655}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":611,"rms_floor":0.00015858554889566586,"log_rms_mae_db":1.2219967581468227,"rms_envelope_nrmse":0.0962178399326488,"envelope_correlation":0.995229321567246,"rms_gain_error_db":-0.45423233925877554,"reference_mean_power":0.001481048748856656,"output_mean_power":0.0013339701749124193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":611,"frequency_resolution_hz":0.3273322422258592,"power_floor":1.1564128413579505e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005748578167219461,"output_power":0.000525042045061368,"reference_active":true,"delta_db":-0.3936635827848324},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6324695094482264e-05,"output_power":3.673681739189777e-05,"reference_active":true,"delta_db":0.04899551949545999},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.166915155074346e-06,"output_power":8.860694297506102e-07,"reference_active":true,"delta_db":-1.1957152695339992}],"active_band_count":3,"active_band_mae_db":0.5461247906047638}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":617,"rms_floor":0.00019136572750236332,"log_rms_mae_db":2.1073401092028243,"rms_envelope_nrmse":0.284623044460286,"envelope_correlation":0.9776437688233002,"rms_gain_error_db":-2.2753472210590675,"reference_mean_power":0.0009682828151026371,"output_mean_power":0.0005734129568408602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":617,"frequency_resolution_hz":0.3241491085899514,"power_floor":8.083944477379101e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003394560096560892,"output_power":0.00018771229798386586,"reference_active":true,"delta_db":-2.572907753487352},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.0093658279697474e-05,"output_power":1.635147178948403e-05,"reference_active":true,"delta_db":-2.6491813572207104},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.750912344451537e-06,"output_power":5.031551409893024e-06,"reference_active":true,"delta_db":-1.276605545825359}],"active_band_count":3,"active_band_mae_db":2.166231552177807}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":611,"rms_floor":0.00016758862552795114,"log_rms_mae_db":1.7546352782012509,"rms_envelope_nrmse":0.2410772659172813,"envelope_correlation":0.9959022792950858,"rms_gain_error_db":-2.2753511788267398,"reference_mean_power":0.0009777906768826752,"output_mean_power":0.000579042944382292},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":611,"frequency_resolution_hz":0.3273322422258592,"power_floor":7.814790794358098e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031601705695843464,"output_power":0.00017437105861071216,"reference_active":true,"delta_db":-2.582361200396306},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.537276404123417e-05,"output_power":8.383534973810669e-06,"reference_active":true,"delta_db":-2.6332478075138015},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.110860252209602e-07,"output_power":2.8553738859209636e-07,"reference_active":true,"delta_db":-0.37217518607938993}],"active_band_count":3,"active_band_mae_db":1.862594731329832}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":617,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":2.614088744868463,"rms_envelope_nrmse":0.23233739910339557,"envelope_correlation":0.9665929850669265,"rms_gain_error_db":-0.8658013524214165,"reference_mean_power":0.0001503076679236543,"output_mean_power":0.00012314052515291252},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":617,"frequency_resolution_hz":0.3241491085899514,"power_floor":1.6069645601479932e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.528322092235623e-05,"output_power":4.9767388938686216e-05,"reference_active":true,"delta_db":-0.4564848210882969},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.542748851536888e-05,"output_power":1.4787298927274953e-05,"reference_active":true,"delta_db":-0.18406379500614786},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1568330454852714e-06,"output_power":3.9276724977939724e-06,"reference_active":true,"delta_db":0.9488365363416793}],"active_band_count":3,"active_band_mae_db":0.5297950508120414}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49440,"uncovered_tail_samples":0,"metrics":{"frame_count":611,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":2.172253890611952,"rms_envelope_nrmse":0.16570789866671287,"envelope_correlation":0.9857027718623297,"rms_gain_error_db":-0.8658010089769741,"reference_mean_power":0.00015178353742719073,"output_mean_power":0.00012434965058418938},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":611,"frequency_resolution_hz":0.3273322422258592,"power_floor":1.4995768603391344e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.9554065383336685e-05,"output_power":4.5411893648626415e-05,"reference_active":true,"delta_db":-0.37909677423073285},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.404662939119468e-06,"output_power":7.975675448419514e-06,"reference_active":true,"delta_db":-0.22752828106130576},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7662891807663936e-07,"output_power":2.0193939744950313e-07,"reference_active":true,"delta_db":0.5815924735500857}],"active_band_count":3,"active_band_mae_db":0.3960725096140414}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.8133389949798584,"predicted_mos":3.8133389949798584,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":74520,"input_sample_rate":24000,"input_seconds":3.105,"input_float32_sha256":"8cc25c96d4ad65141d72d2a8fdc512bda2af6bff271082b244352743fc53d2bd","internal_samples":49680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":18.581119408998347},"started_at":"2026-10-13T07:46:40.609209+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002688676002435386,"native_input_samples":74400,"native_input_sha256":"d65be13a8abb4ee178a2e53c070bddeaf7db83549cb810fe523797381e8242ac","output_sha256":"e61abbfe30acfa2074d0aacc962ecf21211c5c067d383a25253ca1f04a36e946","output_samples":74520,"length_delta_samples":120,"analysis_s":1.53934597999978,"synthesis_s":8.064805183999852,"total_s":9.604151163999632,"rtf":3.0981132787095587,"output_peak":0.8817752599716187,"output_rms":0.05974154386783258,"finished_at":"2026-10-13T07:46:50.226289+00:00","analysis_stages":{"external_f0_s":1.0241240069990454},"evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"e61abbfe30acfa2074d0aacc962ecf21211c5c067d383a25253ca1f04a36e946","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":49600,"output_samples":49680,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.47485183067952,"output_peak":0.8522221728571199},"common_prefix":{"samples":49600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998389694041868,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.4469373147879745,"spectral_convergence":0.5465305306636035,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.44767312884323,"spectral_convergence":0.3709584765960456,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.759178341235213,"spectral_convergence":0.45838827405692634,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0010924952668101545,"delta_db":1.253995871221401,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0004586022156930534,"delta_db":-0.31381034353659576,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.0010925905698370045,"delta_db":-0.768938921661284,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.0006055961432148518,"delta_db":-1.1250958844861576,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.0002568666412377156,"delta_db":-0.8483860216797467,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":6.834375451298338e-05,"delta_db":-1.9129860314793325,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0372021790107528,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2871463447630324,"pitch":null,"output_rms_dbfs":-24.46785355743712,"output_peak":0.8522221728571199,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.884596261622139},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49600,"output_samples":49600,"intelligibility":{"estoi":{"value":0.8666904596795598,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06179660827551199,"rms_dbfs":-24.180707212674086,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.977421996362374,"mean":2.1208486249369958e-06,"zero_energy":false},"output":{"rms":0.05978707712629681,"rms_dbfs":-24.46785355743712,"peak":0.8522221728571199,"peak_dbfs":-1.3889434129383766,"crest_factor_db":23.078910144498742,"mean":-8.770895070922138e-05,"zero_energy":false},"rms_difference_db":-0.287146344763034},"full_scale":{"reference":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":310,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":91,"quiet_samples":14560,"reference_mean_power":1.9840398486089575e-06,"output_mean_power":4.6170754892800826e-07,"output_rms_dbfs":-63.35633024667017,"output_zero_energy":false,"output_peak":0.0032291240809293677,"output_energy_fraction":3.791683071164234e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":303,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":62,"both_estimated_voiced_frames":34,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.85,"absolute_difference_cents_median":13.453464109758304,"absolute_difference_cents_p95":29.580526207305237,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.0002690918479937382,"log_rms_mae_db":3.032117151999012,"rms_envelope_nrmse":0.14226610180823593,"envelope_correlation":0.9836958090722759,"rms_gain_error_db":-0.2871447434362689,"reference_mean_power":0.0038249880205223005,"output_mean_power":0.0035802685616055205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":2.1583633605855604e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008154057878800692,"output_power":0.0007258563910979484,"reference_active":true,"delta_db":-0.505230847148157},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.087495849252008e-05,"output_power":7.858361285605124e-05,"reference_active":true,"delta_db":-0.631122340191146},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5994291162280967e-05,"output_power":2.0381987368110585e-05,"reference_active":true,"delta_db":-1.0563145106847895}],"active_band_count":3,"active_band_mae_db":0.7308892326746975}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.0002456731948415248,"log_rms_mae_db":2.6665118839764896,"rms_envelope_nrmse":0.10007521269884558,"envelope_correlation":0.9921772468639579,"rms_gain_error_db":-0.2871362574218413,"reference_mean_power":0.0038624153279842384,"output_mean_power":0.0036153083662007974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.0216648233438394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007562450561775056,"output_power":0.0006708954294658198,"reference_active":true,"delta_db":-0.5200771532748031},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1399930780791805e-05,"output_power":3.604910942093866e-05,"reference_active":true,"delta_db":-0.6010507488290364},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0921358976760214e-06,"output_power":9.919143384167463e-07,"reference_active":true,"delta_db":-0.4180251410560403}],"active_band_count":3,"active_band_mae_db":0.5130510143866266}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00017000794064843354,"log_rms_mae_db":2.2633137688783354,"rms_envelope_nrmse":0.1678500066825346,"envelope_correlation":0.9836532289395915,"rms_gain_error_db":-0.6527980661384585,"reference_mean_power":0.001461911046007658,"output_mean_power":0.0012578855870483188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.2056298552980859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006038941034712422,"output_power":0.0005004235161331642,"reference_active":true,"delta_db":-0.8162307903299666},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14042207191201e-05,"output_power":6.652086002251515e-05,"reference_active":true,"delta_db":-0.30766028367380455},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.020128231954835e-05,"output_power":1.4589783203250134e-05,"reference_active":true,"delta_db":-1.4133009954505995}],"active_band_count":3,"active_band_mae_db":0.8457306898181235}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00015858554889566586,"log_rms_mae_db":1.6566476264751624,"rms_envelope_nrmse":0.1219702226839922,"envelope_correlation":0.9928810870165753,"rms_gain_error_db":-0.6527944520932483,"reference_mean_power":0.0014762175799508475,"output_mean_power":0.0012701965468394608},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.148107626412446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000571116387197927,"output_power":0.0004747189153637374,"reference_active":true,"delta_db":-0.8028808497406144},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.637477969141665e-05,"output_power":3.379372817183852e-05,"reference_active":true,"delta_db":-0.3196426524212373},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1727904534817564e-06,"output_power":8.541545834436624e-07,"reference_active":true,"delta_db":-1.3768394647414177}],"active_band_count":3,"active_band_mae_db":0.8331209889677563}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00019136572750236332,"log_rms_mae_db":1.699619687304958,"rms_envelope_nrmse":0.19345790404801444,"envelope_correlation":0.9829288827331649,"rms_gain_error_db":-1.0435103797652296,"reference_mean_power":0.0009651549113833476,"output_mean_power":0.0007590073596849031},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":7.999024043731724e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033597435990848984,"output_power":0.00024837347241273776,"reference_active":true,"delta_db":-1.3120092614496275},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.996653108292046e-05,"output_power":1.9325613787830886e-05,"reference_active":true,"delta_db":-1.905031759055122},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.661292025781762e-06,"output_power":6.79632190114024e-06,"reference_active":true,"delta_db":0.08715467416257769}],"active_band_count":3,"active_band_mae_db":1.1013985648891091}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00016758862552795114,"log_rms_mae_db":1.3201775505653703,"rms_envelope_nrmse":0.1463262028326808,"envelope_correlation":0.9939652733071742,"rms_gain_error_db":-1.043510323053062,"reference_mean_power":0.0009746009072787486,"output_mean_power":0.0007664357942080317},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":7.731649944214165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003128075285039233,"output_power":0.00022951401674680845,"reference_active":true,"delta_db":-1.3446798324443732},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.538593181542651e-05,"output_power":1.0174197007894252e-05,"reference_active":true,"delta_db":-1.7962366064799875},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119257970392287e-07,"output_power":4.057676526105682e-07,"reference_active":true,"delta_db":1.1422612908139227}],"active_band_count":3,"active_band_mae_db":1.427725909912761}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":1.737304649639829,"rms_envelope_nrmse":0.22598040581093173,"envelope_correlation":0.9734996714125858,"rms_gain_error_db":-1.228583879967886,"reference_mean_power":0.00014982208037105638,"output_mean_power":0.00011290610752686254},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.5994961899965224e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.519717012759742e-05,"output_power":3.8266631478278924e-05,"reference_active":true,"delta_db":-1.5909657871901437},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.568381171011351e-05,"output_power":9.474298231047028e-06,"reference_active":true,"delta_db":-2.189045685086602},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1546072478197707e-06,"output_power":2.226906160824614e-06,"reference_active":true,"delta_db":-1.5124338002321311}],"active_band_count":3,"active_band_mae_db":1.7641484241696255}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":1.1980640350826397,"rms_envelope_nrmse":0.1702323112643025,"envelope_correlation":0.9905287829012374,"rms_gain_error_db":-1.228585819030725,"reference_mean_power":0.00015128837405213323,"output_mean_power":0.0001140110573781826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.4928738695410593e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.969547295575367e-05,"output_power":3.417306973261949e-05,"reference_active":true,"delta_db":-1.6263283538298843},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.547268075821191e-06,"output_power":4.882464754986006e-06,"reference_active":true,"delta_db":-2.431882084168959},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.762826939609959e-07,"output_power":1.0766382918032137e-07,"reference_active":true,"delta_db":-2.141398568922323}],"active_band_count":3,"active_band_mae_db":2.066536335640389}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"llsm2","status":"measured","mos":{"status":"measured","value":4.336883544921875,"predicted_mos":4.336883544921875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":74640,"input_sample_rate":24000,"input_seconds":3.11,"input_float32_sha256":"f97765f497b7c4282bf27e668a92264a1841a7ca74aca8ea314df9e56a2f06b3","internal_samples":49760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.811764186999426},"started_at":"2026-10-13T05:11:34.588470+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026216919995931676,"native_input_samples":74400,"native_input_sha256":"d65be13a8abb4ee178a2e53c070bddeaf7db83549cb810fe523797381e8242ac","output_sha256":"c7f1482a93b63cfeba3fc8ebb6fe2cff3a0a3f81a45afe9386fb98800c7da392","output_samples":74640,"length_delta_samples":240,"analysis_s":0.5381637339996814,"synthesis_s":0.06303807800031791,"total_s":0.6012018119999993,"rtf":0.19393606838709657,"output_peak":1.0540187358856201,"output_rms":0.058490105248730134,"finished_at":"2026-10-13T05:11:35.199505+00:00","analysis_stages":{"external_f0_s":0.07380428899887193},"evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"c7f1482a93b63cfeba3fc8ebb6fe2cff3a0a3f81a45afe9386fb98800c7da392","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":49600,"output_samples":49760,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.65901687722088,"output_peak":0.9929983476399977},"common_prefix":{"samples":49600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9967845659163987,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.9285803521441585,"spectral_convergence":0.2507324869487011,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.617636544903138,"spectral_convergence":0.249621961139332,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.134089089950195,"spectral_convergence":0.2573559868871676,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0007891768974071932,"delta_db":-0.15845613857862656,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.0004588152135558177,"delta_db":-0.3117937304841592,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.0011824183246517865,"delta_db":-0.4258018833260571,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.000657226683548824,"delta_db":-0.769775015388966,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.0002588444414488113,"delta_db":-0.8150746596408988,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":8.512108388723859e-05,"delta_db":-0.9596029147934948,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5734173903687004,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4643228489739129,"pitch":null,"output_rms_dbfs":-24.645030061648,"output_peak":0.9929983476399977,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.893435328999164},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49600,"output_samples":49600,"intelligibility":{"estoi":{"value":0.9258522815815965,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06179660827551199,"rms_dbfs":-24.180707212674086,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.977421996362374,"mean":2.1208486249369958e-06,"zero_energy":false},"output":{"rms":0.05857988259206117,"rms_dbfs":-24.645030061648,"peak":0.9929983476399977,"peak_dbfs":-0.06102948349475637,"crest_factor_db":24.584000578153248,"mean":2.72089526425721e-05,"zero_energy":false},"rms_difference_db":-0.4643228489739144},"full_scale":{"reference":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":310,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":91,"quiet_samples":14560,"reference_mean_power":1.9840398486089575e-06,"output_mean_power":1.8296133580598568e-06,"output_rms_dbfs":-57.37640677593204,"output_zero_energy":false,"output_peak":0.009506844276560292,"output_energy_fraction":0.0001565099767975876,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":303,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":39,"both_estimated_voiced_frames":36,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9,"absolute_difference_cents_median":1.8987403693202063,"absolute_difference_cents_p95":8.102090864457743,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.0002690918479937382,"log_rms_mae_db":1.023403440229361,"rms_envelope_nrmse":0.10069692656608158,"envelope_correlation":0.9935456656392336,"rms_gain_error_db":-0.46432293431561744,"reference_mean_power":0.0038249880205223005,"output_mean_power":0.0034371444629892342},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":2.1583633605855604e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008154057878800692,"output_power":0.000711997219006316,"reference_active":true,"delta_db":-0.5889549249136998},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.087495849252008e-05,"output_power":7.951348298141114e-05,"reference_active":true,"delta_db":-0.5800344794015816},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5994291162280967e-05,"output_power":2.1300209148970237e-05,"reference_active":true,"delta_db":-0.8649411132258749}],"active_band_count":3,"active_band_mae_db":0.6779768391803854}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.0002456731948415248,"log_rms_mae_db":0.6258097641529039,"rms_envelope_nrmse":0.0694247223024688,"envelope_correlation":0.998095459763895,"rms_gain_error_db":-0.46432173036889235,"reference_mean_power":0.0038624153279842384,"output_mean_power":0.0034707777037178244},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.0216648233438394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007562450561775056,"output_power":0.0006630849089676053,"reference_active":true,"delta_db":-0.5709340455100345},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1399930780791805e-05,"output_power":3.8055917512234585e-05,"reference_active":true,"delta_db":-0.36577417934891276},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0921358976760214e-06,"output_power":9.59286189731078e-07,"reference_active":true,"delta_db":-0.5632849003123955}],"active_band_count":3,"active_band_mae_db":0.4999977083904476}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00017000794064843354,"log_rms_mae_db":1.3470696089339507,"rms_envelope_nrmse":0.13305243091654978,"envelope_correlation":0.9892793837996793,"rms_gain_error_db":-0.4219279096678783,"reference_mean_power":0.001461911046007658,"output_mean_power":0.0013265638450858865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.2056298552980859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006038941034712422,"output_power":0.0005174238769152358,"reference_active":true,"delta_db":-0.6711432341142711},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14042207191201e-05,"output_power":6.319278118506312e-05,"reference_active":true,"delta_db":-0.5305641420575059},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.020128231954835e-05,"output_power":1.756526098533255e-05,"reference_active":true,"delta_db":-0.6072433115125566}],"active_band_count":3,"active_band_mae_db":0.6029835625614446}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00015858554889566586,"log_rms_mae_db":0.7544859019193902,"rms_envelope_nrmse":0.07686512803450425,"envelope_correlation":0.9973684799976383,"rms_gain_error_db":-0.4219278127140954,"reference_mean_power":0.0014762175799508475,"output_mean_power":0.0013395458760016709},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.148107626412446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000571116387197927,"output_power":0.0004921959954085796,"reference_active":true,"delta_db":-0.6458654578688785},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.637477969141665e-05,"output_power":3.21365771268498e-05,"reference_active":true,"delta_db":-0.5380075354104528},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1727904534817564e-06,"output_power":1.056666806356621e-06,"reference_active":true,"delta_db":-0.4528235724164547}],"active_band_count":3,"active_band_mae_db":0.5455655218985953}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00019136572750236332,"log_rms_mae_db":1.0497787745060905,"rms_envelope_nrmse":0.15598591569997047,"envelope_correlation":0.9891262917214041,"rms_gain_error_db":-0.827893387843663,"reference_mean_power":0.0009651549113833476,"output_mean_power":0.0007976413984245983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":7.999024043731724e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033597435990848984,"output_power":0.00025694278133647276,"reference_active":true,"delta_db":-1.1646971426954036},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.996653108292046e-05,"output_power":3.331690901674542e-05,"reference_active":true,"delta_db":0.4602823079152667},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.661292025781762e-06,"output_power":7.608131476679717e-06,"reference_active":true,"delta_db":0.5771953586778461}],"active_band_count":3,"active_band_mae_db":0.7340582697628388}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00016758862552795114,"log_rms_mae_db":0.7773075677991343,"rms_envelope_nrmse":0.11917411901627337,"envelope_correlation":0.9959100513831517,"rms_gain_error_db":-0.8278934841713853,"reference_mean_power":0.0009746009072787486,"output_mean_power":0.0008054479174031656},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":7.731649944214165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003128075285039233,"output_power":0.00024164092788779044,"reference_active":true,"delta_db":-1.121067021807523},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.538593181542651e-05,"output_power":1.7253412674295044e-05,"reference_active":true,"delta_db":0.497512065988144},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119257970392287e-07,"output_power":4.231552559375392e-07,"reference_active":true,"delta_db":1.324484460642012}],"active_band_count":3,"active_band_mae_db":0.9810211828125596}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":1.366564921367396,"rms_envelope_nrmse":0.14428234813261806,"envelope_correlation":0.9873510164852893,"rms_gain_error_db":-0.5144056843159341,"reference_mean_power":0.00014982208037105638,"output_mean_power":0.00013308688334584992},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.5994961899965224e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.519717012759742e-05,"output_power":4.9729833912774105e-05,"reference_active":true,"delta_db":-0.4529980392044721},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.568381171011351e-05,"output_power":1.1468488868142395e-05,"reference_active":true,"delta_db":-1.3594542242287213},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1546072478197707e-06,"output_power":2.9886655187367534e-06,"reference_active":true,"delta_db":-0.234679837901772}],"active_band_count":3,"active_band_mae_db":0.6823773671116551}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":1.1204327552351465,"rms_envelope_nrmse":0.09046165829965401,"envelope_correlation":0.9963199481685362,"rms_gain_error_db":-0.5144049282220382,"reference_mean_power":0.00015128837405213323,"output_mean_power":0.000134389414728127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.4928738695410593e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.969547295575367e-05,"output_power":4.48580353915987e-05,"reference_active":true,"delta_db":-0.4447657892694596},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.547268075821191e-06,"output_power":6.109128822554689e-06,"reference_active":true,"delta_db":-1.458480421991117},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.762826939609959e-07,"output_power":1.6811106997263817e-07,"reference_active":true,"delta_db":-0.2061336649107295}],"active_band_count":3,"active_band_mae_db":0.7031266253904355}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0010","method":"magphase","status":"measured","mos":{"status":"measured","value":3.276120185852051,"predicted_mos":3.276120185852051,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":149920,"input_sample_rate":48000,"input_seconds":3.1233333333333335,"input_float32_sha256":"2b82cd080600048cd3cedb801c253c2569b2bf756e7e0b7348f40373d8bbea1f","internal_samples":49974,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.6541108380006335},"started_at":"2026-10-13T05:18:44.029780+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00511165600073582,"native_input_samples":148800,"native_input_sha256":"7c775fb8e6fb5266216fccfa976273b54b0b8cdf18453c28162f5341a4fe1b71","output_sha256":"365a1861901582713b97238abdf6ecbf61f7ebabd9afd3ce384b7bd03bbcee6c","output_samples":149920,"length_delta_samples":1120,"analysis_s":1.6782627510001475,"synthesis_s":1.0865543540003273,"total_s":2.764817105000475,"rtf":0.8918764854840241,"output_peak":1.037708879678799,"output_rms":0.0601558565852963,"finished_at":"2026-10-13T05:18:46.808470+00:00","evaluation_input":{"reference":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"365a1861901582713b97238abdf6ecbf61f7ebabd9afd3ce384b7bd03bbcee6c","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":49600,"output_samples":49974,"length_delta_samples":374,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.414514149233515,"output_peak":1.0105550854407874},"common_prefix":{"samples":49600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9925161083763557,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":49600,"output_samples":49600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.314085957926181,"spectral_convergence":0.5320945355877025,"spectral_convergence_status":"ok","amplitude_floor":1.4924605057201793e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.292717436542223,"spectral_convergence":0.30204844258157176,"spectral_convergence_status":"ok","amplitude_floor":8.127484488587521e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.032539650641882,"spectral_convergence":0.26593103428127507,"spectral_convergence_status":"ok","amplitude_floor":5.495951303547364e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008185024372796483,"output_power":0.0007377589858368873,"delta_db":-0.45105469090567624,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0004929662452993321,"output_power":0.00047737481227047145,"delta_db":-0.1395768241180519,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.001304221454098428,"output_power":0.0012267624102198114,"delta_db":-0.2659087985844562,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007846801909808332,"output_power":0.0007728271786349215,"delta_db":-0.06610302252641093,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003122815376510157,"output_power":0.00033037481937728524,"delta_db":0.24460629570524264,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0001061689290478197,"output_power":0.00010084716495659986,"delta_db":-0.22333742829916733,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23176451002316756,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20119121578321156,"pitch":null,"output_rms_dbfs":-24.3818984284573,"output_peak":1.0105550854407874,"reference_rms_dbfs":-24.180707212674086,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.546447681703429},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":49600,"output_samples":49600,"intelligibility":{"estoi":{"value":0.7773082794548009,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06179660827551199,"rms_dbfs":-24.180707212674086,"peak":0.97686767578125,"peak_dbfs":-0.20328521631171198,"crest_factor_db":23.977421996362374,"mean":2.1208486249369958e-06,"zero_energy":false},"output":{"rms":0.06038166419780078,"rms_dbfs":-24.3818984284573,"peak":1.0105550854407874,"peak_dbfs":0.09119983848958771,"crest_factor_db":24.473098266946888,"mean":-0.00024036446484962618,"zero_energy":false},"rms_difference_db":-0.20119121578321497},"full_scale":{"reference":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":49600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":1,"at_or_above_full_scale_fraction":2.0161290322580645e-05,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":2.0161290322580645e-05,"longest_full_scale_run_samples":1,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":310,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.153410165337845e-06,"threshold_dbfs":-51.454868724760686,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":91,"quiet_samples":14560,"reference_mean_power":1.9840398486089575e-06,"output_mean_power":1.7221998403672018e-06,"output_rms_dbfs":-57.639164553602356,"output_zero_energy":false,"output_peak":0.016872545850873756,"output_energy_fraction":0.0001386606034687824,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":303,"reference_estimated_voiced_frames":40,"output_estimated_voiced_frames":30,"both_estimated_voiced_frames":22,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.55,"absolute_difference_cents_median":4.746607673288354,"absolute_difference_cents_p95":68.12538351746959,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":49600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.0002690918479937382,"log_rms_mae_db":2.4243131831172104,"rms_envelope_nrmse":0.16239218981341227,"envelope_correlation":0.9783926197294129,"rms_gain_error_db":-0.20118988263066223,"reference_mean_power":0.0038249880205223005,"output_mean_power":0.0036518345323248258},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":2.1583633605855604e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008154057878800692,"output_power":0.0007944955895233551,"reference_active":true,"delta_db":-0.11282299163630427},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.087495849252008e-05,"output_power":9.15213004098455e-05,"reference_active":true,"delta_db":0.030779568041800223},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5994291162280967e-05,"output_power":2.8814942906243016e-05,"reference_active":true,"delta_db":0.4473978424474245}],"active_band_count":3,"active_band_mae_db":0.197000134041843}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.0002456731948415248,"log_rms_mae_db":1.6448675667824366,"rms_envelope_nrmse":0.10672340605153628,"envelope_correlation":0.9905566553988349,"rms_gain_error_db":-0.2011836065940539,"reference_mean_power":0.0038624153279842384,"output_mean_power":0.0036875728709206473},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":2.0216648233438394e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007562450561775056,"output_power":0.0007369047348552371,"reference_active":true,"delta_db":-0.1125120152742248},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1399930780791805e-05,"output_power":4.0002816024040536e-05,"reference_active":true,"delta_db":-0.14909050152295678},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0921358976760214e-06,"output_power":1.2171788630653967e-06,"reference_active":true,"delta_db":0.4707771973682798}],"active_band_count":3,"active_band_mae_db":0.24412657138848712}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00017000794064843354,"log_rms_mae_db":2.9957186191243617,"rms_envelope_nrmse":0.19851718646436184,"envelope_correlation":0.9738675339642033,"rms_gain_error_db":-0.26411063223810877,"reference_mean_power":0.001461911046007658,"output_mean_power":0.0013756561192794095},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.2056298552980859e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006038941034712422,"output_power":0.000572801974465721,"reference_active":true,"delta_db":-0.22956282772354977},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14042207191201e-05,"output_power":6.701451746760774e-05,"reference_active":true,"delta_db":-0.27554988940062425},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.020128231954835e-05,"output_power":1.6793003056242857e-05,"reference_active":true,"delta_db":-0.8025057107005673}],"active_band_count":3,"active_band_mae_db":0.4358728092749138}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00015858554889566586,"log_rms_mae_db":2.0453190290682635,"rms_envelope_nrmse":0.133674959771614,"envelope_correlation":0.9880582056676649,"rms_gain_error_db":-0.2641094501209887,"reference_mean_power":0.0014762175799508475,"output_mean_power":0.0013891189245425978},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.148107626412446e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000571116387197927,"output_power":0.0005411435601766369,"reference_active":true,"delta_db":-0.23412127136576935},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.637477969141665e-05,"output_power":3.438018821765564e-05,"reference_active":true,"delta_db":-0.24492121696362779},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1727904534817564e-06,"output_power":1.0618058212444118e-06,"reference_active":true,"delta_db":-0.4317532016508945}],"active_band_count":3,"active_band_mae_db":0.3035985633267639}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":0.00019136572750236332,"log_rms_mae_db":2.4597783826124977,"rms_envelope_nrmse":0.1287960992527958,"envelope_correlation":0.9898154896763247,"rms_gain_error_db":-0.0040315039685233055,"reference_mean_power":0.0009651549113833476,"output_mean_power":0.0009642593852876328},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":7.999024043731724e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00033597435990848984,"output_power":0.00034505660330732266,"reference_active":true,"delta_db":0.11584207661668165},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.996653108292046e-05,"output_power":3.16135882938023e-05,"reference_active":true,"delta_db":0.23237321092665947},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.661292025781762e-06,"output_power":4.861934992679704e-06,"reference_active":true,"delta_db":-1.3674932546265488}],"active_band_count":3,"active_band_mae_db":0.5719028473899633}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":0.00016758862552795114,"log_rms_mae_db":1.7290438327452333,"rms_envelope_nrmse":0.0784612784234208,"envelope_correlation":0.9961295234470157,"rms_gain_error_db":-0.004030095610798581,"reference_mean_power":0.0009746009072787486,"output_mean_power":0.0009736969324027182},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":7.731649944214165e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003128075285039233,"output_power":0.0003179153369289603,"reference_active":true,"delta_db":0.07034282824637686},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.538593181542651e-05,"output_power":1.61082048881487e-05,"reference_active":true,"delta_db":0.19923341496064806},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.119257970392287e-07,"output_power":2.8158121730090404e-07,"reference_active":true,"delta_db":-0.44447611382001384}],"active_band_count":3,"active_band_mae_db":0.23801745234234625}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":619,"rms_floor":5.611897603565048e-05,"log_rms_mae_db":2.2876834009798666,"rms_envelope_nrmse":0.22514477334361946,"envelope_correlation":0.9672773988629279,"rms_gain_error_db":0.2713912594367377,"reference_mean_power":0.00014982208037105638,"output_mean_power":0.00015948320295644723},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":619,"frequency_resolution_hz":0.32310177705977383,"power_floor":1.5994961899965224e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.519717012759742e-05,"output_power":6.193559794277825e-05,"reference_active":true,"delta_db":0.5002352204284182},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.568381171011351e-05,"output_power":1.6865071323605027e-05,"reference_active":true,"delta_db":0.3153656239621394},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1546072478197707e-06,"output_power":6.080390242869969e-06,"reference_active":true,"delta_db":2.849861566414474}],"active_band_count":3,"active_band_mae_db":1.221820803601677}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":49600,"uncovered_tail_samples":0,"metrics":{"frame_count":613,"rms_floor":4.517690856022871e-05,"log_rms_mae_db":1.608272620311883,"rms_envelope_nrmse":0.1454622062252403,"envelope_correlation":0.9860660214678774,"rms_gain_error_db":0.27139076250500943,"reference_mean_power":0.00015128837405213323,"output_mean_power":0.000161044030648646},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":613,"frequency_resolution_hz":0.3262642740619902,"power_floor":1.4928738695410593e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.969547295575367e-05,"output_power":5.611552621020416e-05,"reference_active":true,"delta_db":0.5276621159546397},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.547268075821191e-06,"output_power":8.73775168940938e-06,"reference_active":true,"delta_db":0.09572373360934591},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.762826939609959e-07,"output_power":3.200800198020276e-07,"reference_active":true,"delta_db":2.5904888644023427}],"active_band_count":3,"active_band_mae_db":1.0712915713221094}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"hnm","status":"measured","mos":{"status":"measured","value":3.193633794784546,"predicted_mos":3.193633794784546,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65760,"input_sample_rate":16000,"input_seconds":4.11,"input_float32_sha256":"e5ff6d4d6d5e72832a136ffd7a9b56f3e4aae3237bb5c35a323fb7cb956573d5","internal_samples":65760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.466359095000371},"started_at":"2026-10-13T04:40:25.602187+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5798000276845414e-05,"native_input_samples":65760,"native_input_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","output_sha256":"a47097429a474a66e994fe51212f946155eaeaefbca39691ec51aa09341cafbf","output_samples":65760,"length_delta_samples":0,"analysis_s":0.6690315910000209,"synthesis_s":0.3706527790000109,"total_s":1.0396843700000318,"rtf":0.25296456690998337,"output_peak":0.4701780392019266,"output_rms":0.05783542476392312,"finished_at":"2026-10-13T04:40:26.646636+00:00","evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"a47097429a474a66e994fe51212f946155eaeaefbca39691ec51aa09341cafbf","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.7997072022228195,"spectral_convergence":0.22419277362714507,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.449695245903603,"spectral_convergence":0.21873088524159082,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.700494101300302,"spectral_convergence":0.22628893514153167,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0008067536471112012,"delta_db":-0.007681756746432826,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0008300746155847151,"delta_db":-0.03680354805458131,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.001189896126438322,"delta_db":-0.307248750188978,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.00036234268942938173,"delta_db":-0.28767409943961775,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00013792351791272814,"delta_db":0.18275493299716075,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.7945761147064415e-05,"delta_db":-0.3840014699153959,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20102742622369443,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14870058853217286,"pitch":null,"output_rms_dbfs":-24.756121409143663,"output_peak":0.4701780392019266,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.649965516475575},"common_prefix":{"samples":65760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.7997072022228195,"spectral_convergence":0.22419277362714507,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.449695245903603,"spectral_convergence":0.21873088524159082,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.700494101300302,"spectral_convergence":0.22628893514153167,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0008067536471112012,"delta_db":-0.007681756746432826,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0008300746155847151,"delta_db":-0.03680354805458131,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.001189896126438322,"delta_db":-0.307248750188978,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.00036234268942938173,"delta_db":-0.28767409943961775,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00013792351791272814,"delta_db":0.18275493299716075,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.7945761147064415e-05,"delta_db":-0.3840014699153959,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20102742622369443,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14870058853217286,"pitch":null,"output_rms_dbfs":-24.756121409143663,"output_peak":0.4701780392019266,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.649965516475575},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65760,"output_samples":65760,"intelligibility":{"estoi":{"value":0.8850524516905915,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05883407895338475,"rms_dbfs":-24.60742082061149,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.980889206210556,"mean":-1.1628793683945407e-05,"zero_energy":false},"output":{"rms":0.05783542476392312,"rms_dbfs":-24.756121409143663,"peak":0.4701780392019266,"peak_dbfs":-6.554753190148976,"crest_factor_db":18.201368218994688,"mean":-9.657257811889996e-05,"zero_energy":false},"rms_difference_db":-0.14870058853217216},"full_scale":{"reference":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":411,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.508459147065878e-06,"output_mean_power":2.8323327337850242e-06,"output_rms_dbfs":-55.478557283980614,"output_zero_energy":false,"output_peak":0.00815952598116265,"output_energy_fraction":0.00014009527993044,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":404,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":99,"both_estimated_voiced_frames":87,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9666666666666667,"absolute_difference_cents_median":1.61964019471221,"absolute_difference_cents_p95":10.337729731762575,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00020667206862955213,"log_rms_mae_db":1.1424951926310563,"rms_envelope_nrmse":0.07043035764922918,"envelope_correlation":0.9962154457916561,"rms_gain_error_db":-0.14870066836835588,"reference_mean_power":0.003465660170380573,"output_mean_power":0.0033490058668308384},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":2.3006803943986064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009732960931433551,"output_power":0.0009269732856036664,"reference_active":true,"delta_db":-0.21177761686158347},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.170752162168666e-05,"output_power":7.324276901155845e-05,"reference_active":true,"delta_db":-0.47497282691272347},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0779113359754193e-05,"output_power":1.8417514023480603e-05,"reference_active":true,"delta_db":-0.5239600315750871}],"active_band_count":3,"active_band_mae_db":0.403570158449798}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00018941444655877122,"log_rms_mae_db":0.6325559799203415,"rms_envelope_nrmse":0.04455185214120554,"envelope_correlation":0.9985786731169287,"rms_gain_error_db":-0.14870677090054266,"reference_mean_power":0.003491154294575285,"output_mean_power":0.003373637116961605},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":2.1871865789273095e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008996086963166702,"output_power":0.0008550995102003822,"reference_active":true,"delta_db":-0.22036987340707764},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8292682982878605e-05,"output_power":3.5410611399604204e-05,"reference_active":true,"delta_db":-0.33982370973957654},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1507764875014338e-06,"output_power":7.444676373588296e-07,"reference_active":true,"delta_db":-1.8914515651881938}],"active_band_count":3,"active_band_mae_db":0.8172150494449494}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.0001875778806757321,"log_rms_mae_db":1.6190297246401988,"rms_envelope_nrmse":0.09065138783869653,"envelope_correlation":0.9950180061635417,"rms_gain_error_db":-0.2492112925565922,"reference_mean_power":0.0016622862050937872,"output_mean_power":0.0015695843924224348},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":1.477365390744301e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007590270471846978,"output_power":0.0007189755786055407,"reference_active":true,"delta_db":-0.235431128469553},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5779188180805434e-05,"output_power":4.3533568155648015e-05,"reference_active":true,"delta_db":-1.0764792435234953},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.466208519389042e-05,"output_power":9.415242400478147e-06,"reference_active":true,"delta_db":-1.9236423299930319}],"active_band_count":3,"active_band_mae_db":1.0785175673286933}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016891518288067464,"log_rms_mae_db":0.8962995974627926,"rms_envelope_nrmse":0.056353533664897815,"envelope_correlation":0.9983672074323201,"rms_gain_error_db":-0.24921439218687746,"reference_mean_power":0.001674521427963224,"output_mean_power":0.001581136157054279},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.422523294383883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007062170888212691,"output_power":0.0006678297811345796,"reference_active":true,"delta_db":-0.24272440241952314},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.54990695717255e-05,"output_power":2.0358369095892017e-05,"reference_active":true,"delta_db":-0.9778135004510617},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.502016839919999e-07,"output_power":4.834593148043863e-07,"reference_active":true,"delta_db":-2.934558635661129}],"active_band_count":3,"active_band_mae_db":1.3850321795105713}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00010800393678736649,"log_rms_mae_db":1.5607089120348068,"rms_envelope_nrmse":0.15026138973968262,"envelope_correlation":0.9850244215311231,"rms_gain_error_db":-0.2171357551486042,"reference_mean_power":0.00040899310436077964,"output_mean_power":0.00038904730419941994},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":4.3909884369057597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019183671841687986,"output_power":0.00017986347990813626,"reference_active":true,"delta_db":-0.27988745052179903},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4527862162762088e-05,"output_power":1.5408467980660143e-05,"reference_active":true,"delta_db":0.2555774951883149},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.245905579917374e-06,"output_power":5.5957187634788295e-06,"reference_active":true,"delta_db":-1.6838247852524177}],"active_band_count":3,"active_band_mae_db":0.7397632436541772}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":0.9903012094506037,"rms_envelope_nrmse":0.09825952495238698,"envelope_correlation":0.9936051541401144,"rms_gain_error_db":-0.21713716640557604,"reference_mean_power":0.0004120037826983826,"output_mean_power":0.00039191103024284013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":4.2438330741066267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018201492488703654,"output_power":0.00016916339194374223,"reference_active":true,"delta_db":-0.31800616063554715},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.209318154228541e-06,"output_power":8.379996140286518e-06,"reference_active":true,"delta_db":0.6534962681536902},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163481758524234e-07,"output_power":2.6591673137149085e-07,"reference_active":true,"delta_db":-0.754196707234529}],"active_band_count":3,"active_band_mae_db":0.5752330453412554}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":1.690425720635897,"rms_envelope_nrmse":0.15579496067532336,"envelope_correlation":0.9853698355851639,"rms_gain_error_db":-0.054858046100563065,"reference_mean_power":8.457687952311121e-05,"output_mean_power":8.351526300686278e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":3.6268766197283325e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4506351813371544e-05,"output_power":1.4159297127298325e-05,"reference_active":true,"delta_db":-0.10516510571237661},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.817293180171656e-06,"output_power":8.130075983310957e-06,"reference_active":true,"delta_db":0.17038204443524496},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650085525655338e-06,"output_power":1.7960597939650498e-06,"reference_active":true,"delta_db":0.598040757421301}],"active_band_count":3,"active_band_mae_db":0.29119596918964086}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":1.128425130484666,"rms_envelope_nrmse":0.0882556654285829,"envelope_correlation":0.9951637444988418,"rms_gain_error_db":-0.05485944135061618,"reference_mean_power":8.519942751057572e-05,"output_mean_power":8.412996968831492e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":3.3424099706380933e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5109841043230118e-05,"output_power":1.5177759866773277e-05,"reference_active":true,"delta_db":0.01947781853890109},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127479817162532e-06,"output_power":4.4504317646268275e-06,"reference_active":true,"delta_db":0.32717188391802243},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2069317938633797e-07,"output_power":1.225202983384845e-07,"reference_active":true,"delta_db":0.06525317709913712}],"active_band_count":3,"active_band_mae_db":0.1373009598520202}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.4075944423675537,"predicted_mos":3.4075944423675537,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65760,"input_sample_rate":16000,"input_seconds":4.11,"input_float32_sha256":"36117caa98757fecf4bc0aaa6a81b5e4ab4d98ef0239938fd07332e0c768c6c7","internal_samples":65760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4127873009983887},"started_at":"2026-10-13T04:43:41.675687+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2049000108381733e-05,"native_input_samples":65760,"native_input_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","output_sha256":"b50ade00ed76fe3b484b40217b2b9e345af99807b5ac7bbfe642da1647a073d1","output_samples":65760,"length_delta_samples":0,"analysis_s":1.1598161769998114,"synthesis_s":0.06296087700047792,"total_s":1.2227770540002894,"rtf":0.297512665206883,"output_peak":0.6862236482245594,"output_rms":0.0654114242074163,"finished_at":"2026-10-13T04:43:42.902543+00:00","evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"b50ade00ed76fe3b484b40217b2b9e345af99807b5ac7bbfe642da1647a073d1","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.673380664828492,"spectral_convergence":0.7041350811389632,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.543999591953038,"spectral_convergence":0.3064734500666249,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.859853447674146,"spectral_convergence":0.30461977350025493,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0009240137407829088,"delta_db":0.5816931647456471,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0010507835926430154,"delta_db":0.9871579515843416,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.001624046002931913,"delta_db":1.0436440120675348,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.0005206197847574998,"delta_db":1.2863375485521331,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00013938321773398455,"delta_db":0.22847653664349185,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.9808077993245716e-05,"delta_db":0.0448030724470998,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6953520476733748,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9204929237792636,"pitch":null,"output_rms_dbfs":-23.68692789683223,"output_peak":0.6862236482245594,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.025744568151892},"common_prefix":{"samples":65760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.673380664828492,"spectral_convergence":0.7041350811389632,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.543999591953038,"spectral_convergence":0.3064734500666249,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.859853447674146,"spectral_convergence":0.30461977350025493,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0009240137407829088,"delta_db":0.5816931647456471,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0010507835926430154,"delta_db":0.9871579515843416,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.001624046002931913,"delta_db":1.0436440120675348,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.0005206197847574998,"delta_db":1.2863375485521331,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00013938321773398455,"delta_db":0.22847653664349185,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.9808077993245716e-05,"delta_db":0.0448030724470998,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6953520476733748,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.9204929237792636,"pitch":null,"output_rms_dbfs":-23.68692789683223,"output_peak":0.6862236482245594,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.025744568151892},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65760,"output_samples":65760,"intelligibility":{"estoi":{"value":0.8942489045974495,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05883407895338475,"rms_dbfs":-24.60742082061149,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.980889206210556,"mean":-1.1628793683945407e-05,"zero_energy":false},"output":{"rms":0.0654114242074163,"rms_dbfs":-23.68692789683223,"peak":0.6862236482245594,"peak_dbfs":-3.270686392560867,"crest_factor_db":20.416241504271362,"mean":-3.552531480734729e-10,"zero_energy":false},"rms_difference_db":0.9204929237792605},"full_scale":{"reference":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":411,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.508459147065878e-06,"output_mean_power":1.6538017505690077e-06,"output_rms_dbfs":-57.81516552700189,"output_zero_energy":false,"output_peak":0.006783331589679866,"output_energy_fraction":6.395040920972265e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":404,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":106,"both_estimated_voiced_frames":77,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.8555555555555555,"absolute_difference_cents_median":4.768323935790997,"absolute_difference_cents_p95":18.448403286422028,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00020667206862955213,"log_rms_mae_db":2.1280770205317388,"rms_envelope_nrmse":0.18076374796561956,"envelope_correlation":0.9857243837753337,"rms_gain_error_db":0.9204984868693742,"reference_mean_power":0.003465660170380573,"output_mean_power":0.004283865470246507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":2.3006803943986064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009732960931433551,"output_power":0.001286458600054379,"reference_active":true,"delta_db":1.2115083449801558},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.170752162168666e-05,"output_power":9.82399389700478e-05,"reference_active":true,"delta_db":0.8002604645743718},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0779113359754193e-05,"output_power":2.1628772923706497e-05,"reference_active":true,"delta_db":0.17404868720421635}],"active_band_count":3,"active_band_mae_db":0.7286058322529145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00018941444655877122,"log_rms_mae_db":1.6271977153823398,"rms_envelope_nrmse":0.13424458020810726,"envelope_correlation":0.9960237656652622,"rms_gain_error_db":0.9205151017450884,"reference_mean_power":0.003491154294575285,"output_mean_power":0.004315394993941966},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":2.1871865789273095e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008996086963166702,"output_power":0.0011949832787162605,"reference_active":true,"delta_db":1.2330818327480881},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8292682982878605e-05,"output_power":4.590490707692729e-05,"reference_active":true,"delta_db":0.7874331628230509},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1507764875014338e-06,"output_power":1.1030798196955416e-06,"reference_active":true,"delta_db":-0.18384040410376928}],"active_band_count":3,"active_band_mae_db":0.7347851332249693}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.0001875778806757321,"log_rms_mae_db":2.1541954509246186,"rms_envelope_nrmse":0.21862498030456282,"envelope_correlation":0.9836925476794793,"rms_gain_error_db":1.1355221293358284,"reference_mean_power":0.0016622862050937872,"output_mean_power":0.002159026702856846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":1.477365390744301e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007590270471846978,"output_power":0.001014318444448972,"reference_active":true,"delta_db":1.2591707097405513},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5779188180805434e-05,"output_power":8.524584188948906e-05,"reference_active":true,"delta_db":1.8420101503689212},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.466208519389042e-05,"output_power":1.0248716039442968e-05,"reference_active":true,"delta_db":-1.5552627831374857}],"active_band_count":3,"active_band_mae_db":1.5521478810823195}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016891518288067464,"log_rms_mae_db":1.6590289894921544,"rms_envelope_nrmse":0.1734121477099556,"envelope_correlation":0.9937048570535515,"rms_gain_error_db":1.135520741619541,"reference_mean_power":0.001674521427963224,"output_mean_power":0.002174917478818711},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.422523294383883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007062170888212691,"output_power":0.0009501200866951421,"reference_active":true,"delta_db":1.2884027742570827},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.54990695717255e-05,"output_power":4.014558224927062e-05,"reference_active":true,"delta_db":1.9711342718801763},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.502016839919999e-07,"output_power":5.671358280434544e-07,"reference_active":true,"delta_db":-2.2412871163826287}],"active_band_count":3,"active_band_mae_db":1.833608054173296}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00010800393678736649,"log_rms_mae_db":1.6628069091199835,"rms_envelope_nrmse":0.23410042908700773,"envelope_correlation":0.9822062870753606,"rms_gain_error_db":1.2415110259363153,"reference_mean_power":0.00040899310436077964,"output_mean_power":0.0005443360386525512},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":4.3909884369057597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019183671841687986,"output_power":0.00026396370677215917,"reference_active":true,"delta_db":1.3861248169277092},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4527862162762088e-05,"output_power":2.2475189832663533e-05,"reference_active":true,"delta_db":1.8950165790754636},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.245905579917374e-06,"output_power":1.1982370581110096e-05,"reference_active":true,"delta_db":1.6230438938523015}],"active_band_count":3,"active_band_mae_db":1.634728429951825}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":1.1528844686047077,"rms_envelope_nrmse":0.17946141072412702,"envelope_correlation":0.9949304189027427,"rms_gain_error_db":1.2415122017196047,"reference_mean_power":0.0004120037826983826,"output_mean_power":0.0005483431512899429},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":4.2438330741066267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018201492488703654,"output_power":0.00025210376981122095,"reference_active":true,"delta_db":1.4147233911969939},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.209318154228541e-06,"output_power":1.0710223540798789e-05,"reference_active":true,"delta_db":1.719043436150012},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163481758524234e-07,"output_power":5.013421495491823e-07,"reference_active":true,"delta_db":1.9996888422536578}],"active_band_count":3,"active_band_mae_db":1.7111518898668878}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":1.577891215298149,"rms_envelope_nrmse":0.24943713019559338,"envelope_correlation":0.962239623790909,"rms_gain_error_db":-0.051374542141870067,"reference_mean_power":8.457687952311121e-05,"output_mean_power":8.358227800920153e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":3.6268766197283325e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4506351813371544e-05,"output_power":1.2797904763614406e-05,"reference_active":true,"delta_db":-0.544193319063849},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.817293180171656e-06,"output_power":5.111878952250895e-06,"reference_active":true,"delta_db":-1.8447583875549638},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650085525655338e-06,"output_power":7.86942211272222e-07,"reference_active":true,"delta_db":-2.985738739302205}],"active_band_count":3,"active_band_mae_db":1.7915634819736725}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":0.9915136038018089,"rms_envelope_nrmse":0.14930728840778387,"envelope_correlation":0.9859480380475916,"rms_gain_error_db":-0.0513756011083009,"reference_mean_power":8.519942751057572e-05,"output_mean_power":8.419748446822417e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":3.3424099706380933e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5109841043230118e-05,"output_power":1.2730416171766088e-05,"reference_active":true,"delta_db":-0.744172940837175},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127479817162532e-06,"output_power":2.5866364632220054e-06,"reference_active":true,"delta_db":-2.029495628661415},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2069317938633797e-07,"output_power":5.79755617854846e-08,"reference_active":true,"delta_db":-3.184377622773898}],"active_band_count":3,"active_band_mae_db":1.9860153974241628}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9155561923980713,"predicted_mos":3.9155561923980713,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65800,"input_sample_rate":16000,"input_seconds":4.1125,"input_float32_sha256":"a5a31ea20f99257057e9513a40a20085ed769ed28a70620f3cbe9980d62eb159","internal_samples":65800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.058083220999833},"started_at":"2026-10-13T07:01:28.566870+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00012079399857611861,"native_input_samples":65760,"native_input_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","output_sha256":"16f43b1db44329c611c1ec63b2593b1be35d3957c21ebbd09eaca268a0bee240","output_samples":65800,"length_delta_samples":40,"analysis_s":1.2275981880011386,"synthesis_s":8.952533580999443,"total_s":10.180131769000582,"rtf":2.4769177053529394,"output_peak":0.44299784302711487,"output_rms":0.05506922801864195,"finished_at":"2026-10-13T07:01:38.765115+00:00","evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"16f43b1db44329c611c1ec63b2593b1be35d3957c21ebbd09eaca268a0bee240","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":65760,"output_samples":65800,"length_delta_samples":40,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.181820231621444,"output_peak":0.44299784302711487},"common_prefix":{"samples":65760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9993920972644377,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.455334660610792,"spectral_convergence":0.261438107843683,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.808473419740109,"spectral_convergence":0.26627179478434937,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.9620427260249516,"spectral_convergence":0.2797935907442024,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0007628240937683416,"delta_db":-0.2508471131649231,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0007486404709234329,"delta_db":-0.4852418672410035,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.0011197780563280194,"delta_db":-0.5710197313960417,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.00028634937890397675,"delta_db":-1.309906679684476,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00010193547628831107,"delta_db":-1.1303747607062538,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.493682147604429e-05,"delta_db":-1.1810383974426577,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8214047582725593,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5717588582014286,"pitch":null,"output_rms_dbfs":-25.179179678812922,"output_peak":0.44299784302711487,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.741950268791951},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65760,"output_samples":65760,"intelligibility":{"estoi":{"value":0.850371310744833,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05883407895338475,"rms_dbfs":-24.60742082061149,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.980889206210556,"mean":-1.1628793683945407e-05,"zero_energy":false},"output":{"rms":0.055085971877494595,"rms_dbfs":-25.179179678812922,"peak":0.44299784302711487,"peak_dbfs":-7.071967767353606,"crest_factor_db":18.107211911459316,"mean":-9.20071309161046e-06,"zero_energy":false},"rms_difference_db":-0.5717588582014308},"full_scale":{"reference":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":411,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.508459147065878e-06,"output_mean_power":2.428036740252908e-06,"output_rms_dbfs":-56.14744745945882,"output_zero_energy":false,"output_peak":0.006073422729969025,"output_energy_fraction":0.0001323854673006042,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":404,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":84,"both_estimated_voiced_frames":79,"reference_voiced_output_unvoiced_frames":11,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.8777777777777778,"absolute_difference_cents_median":2.7179737266187973,"absolute_difference_cents_p95":14.01425312137889,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00020667206862955213,"log_rms_mae_db":1.448766256913984,"rms_envelope_nrmse":0.10220961091311831,"envelope_correlation":0.9945610633200456,"rms_gain_error_db":-0.5717549584650852,"reference_mean_power":0.003465660170380573,"output_mean_power":0.003038158865008393},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":2.3006803943986064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009732960931433551,"output_power":0.0008735554073217433,"reference_active":true,"delta_db":-0.46954523751020394},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.170752162168666e-05,"output_power":7.394296021243095e-05,"reference_active":true,"delta_db":-0.43365204528693324},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0779113359754193e-05,"output_power":1.898446182289094e-05,"reference_active":true,"delta_db":-0.3922872222288973}],"active_band_count":3,"active_band_mae_db":0.4318281683420115}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00018941444655877122,"log_rms_mae_db":1.1198953239717733,"rms_envelope_nrmse":0.09109753943132091,"envelope_correlation":0.9962167493050841,"rms_gain_error_db":-0.5717530304862474,"reference_mean_power":0.003491154294575285,"output_mean_power":0.0030605095584642556},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":2.1871865789273095e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008996086963166702,"output_power":0.0008074865385401892,"reference_active":true,"delta_db":-0.4691835399187556},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8292682982878605e-05,"output_power":3.4915864828485965e-05,"reference_active":true,"delta_db":-0.40092992806865435},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1507764875014338e-06,"output_power":9.871106988497192e-07,"reference_active":true,"delta_db":-0.666251208105359}],"active_band_count":3,"active_band_mae_db":0.5121215586975897}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.0001875778806757321,"log_rms_mae_db":2.006486810011345,"rms_envelope_nrmse":0.10776713278824851,"envelope_correlation":0.994595132730448,"rms_gain_error_db":-0.5597358832166373,"reference_mean_power":0.0016622862050937872,"output_mean_power":0.0014612758686844878},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":1.477365390744301e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007590270471846978,"output_power":0.0006864869808387331,"reference_active":true,"delta_db":-0.4362494653259105},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5779188180805434e-05,"output_power":4.5912905061273895e-05,"reference_active":true,"delta_db":-0.845374163020305},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.466208519389042e-05,"output_power":1.3312166475656337e-05,"reference_active":true,"delta_db":-0.4194699857893037}],"active_band_count":3,"active_band_mae_db":0.5670312047118397}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016891518288067464,"log_rms_mae_db":1.443368860064687,"rms_envelope_nrmse":0.09277608315496551,"envelope_correlation":0.9965879798659135,"rms_gain_error_db":-0.5597347827097006,"reference_mean_power":0.001674521427963224,"output_mean_power":0.0014720319321885101},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.422523294383883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007062170888212691,"output_power":0.0006382176093922979,"reference_active":true,"delta_db":-0.43969439410325784},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.54990695717255e-05,"output_power":2.1456018701296702e-05,"reference_active":true,"delta_db":-0.7497519483219068},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.502016839919999e-07,"output_power":9.272345975009668e-07,"reference_active":true,"delta_db":-0.10626167862076696}],"active_band_count":3,"active_band_mae_db":0.43190267368197716}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00010800393678736649,"log_rms_mae_db":1.8012600646521477,"rms_envelope_nrmse":0.18422656082752709,"envelope_correlation":0.9845253526432279,"rms_gain_error_db":-1.0638387204204915,"reference_mean_power":0.00040899310436077964,"output_mean_power":0.00032013423053037286},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":4.3909884369057597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019183671841687986,"output_power":0.00015255506899992832,"reference_active":true,"delta_db":-0.9950509405127699},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4527862162762088e-05,"output_power":1.569652639809572e-05,"reference_active":true,"delta_db":0.33601844059134073},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.245905579917374e-06,"output_power":1.1856281890069072e-05,"reference_active":true,"delta_db":1.5771015877761903}],"active_band_count":3,"active_band_mae_db":0.9693903229601002}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":1.4277124391822076,"rms_envelope_nrmse":0.16225811678368157,"envelope_correlation":0.9900177540764462,"rms_gain_error_db":-1.0638400147667584,"reference_mean_power":0.0004120037826983826,"output_mean_power":0.00032249070518247636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":4.2438330741066267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018201492488703654,"output_power":0.00014828375924721198,"reference_active":true,"delta_db":-0.8901341318319991},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.209318154228541e-06,"output_power":6.885075716928045e-06,"reference_active":true,"delta_db":-0.19985471118725917},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163481758524234e-07,"output_power":4.0844881672483e-07,"reference_active":true,"delta_db":1.109723077023787}],"active_band_count":3,"active_band_mae_db":0.733237306681015}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":1.9121710561285166,"rms_envelope_nrmse":0.2970643913685528,"envelope_correlation":0.9660956276878524,"rms_gain_error_db":-2.0339092698618786,"reference_mean_power":8.457687952311121e-05,"output_mean_power":5.294936190319121e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":3.6268766197283325e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4506351813371544e-05,"output_power":1.0311443198742136e-05,"reference_active":true,"delta_db":-1.4823875214394577},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.817293180171656e-06,"output_power":5.500150732771762e-06,"reference_active":true,"delta_db":-1.5268180847778425},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650085525655338e-06,"output_power":1.2231197628365243e-06,"reference_active":true,"delta_db":-1.0704773181042162}],"active_band_count":3,"active_band_mae_db":1.3598943081071722}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":1.5116268631506924,"rms_envelope_nrmse":0.2689708601840068,"envelope_correlation":0.9776970695302668,"rms_gain_error_db":-2.0339160152434133,"reference_mean_power":8.519942751057572e-05,"output_mean_power":5.33390252731729e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":3.3424099706380933e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5109841043230118e-05,"output_power":1.0789438643863471e-05,"reference_active":true,"delta_db":-1.462610458817374},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127479817162532e-06,"output_power":2.8816781378775893e-06,"reference_active":true,"delta_db":-1.5603948658964428},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2069317938633797e-07,"output_power":9.997625587115219e-08,"reference_active":true,"delta_db":-0.8178585962420188}],"active_band_count":3,"active_band_mae_db":1.2802879736519452}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.8525147438049316,"predicted_mos":3.8525147438049316,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":181760,"input_sample_rate":44100,"input_seconds":4.121541950113379,"input_float32_sha256":"4760284421c45fdb172ce552ce692b24a26a6fb528076ba952b140f92a4281a0","internal_samples":65945,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6518837619987607},"started_at":"2026-10-13T07:31:27.673343+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.01185031400018488,"native_input_samples":181251,"native_input_sha256":"f9a1d663dfbdc0955f26a0a96acc1bc9e0eec90553ec2ea8278ec2a45cf15d7e","output_sha256":"5d5f3ec7697a2b2e71ffeafd99aa1a56df930fcab0166b918d120d03999842e8","output_samples":181760,"length_delta_samples":509,"analysis_s":0.22902293500010273,"synthesis_s":29.612133460999758,"total_s":29.84115639599986,"rtf":7.260621994160549,"output_peak":0.5245469808578491,"output_rms":0.053019275432309755,"finished_at":"2026-10-13T07:31:57.552836+00:00","evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"5d5f3ec7697a2b2e71ffeafd99aa1a56df930fcab0166b918d120d03999842e8","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":65760,"output_samples":65945,"length_delta_samples":185,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.505710236748023,"output_peak":0.5255196243026825},"common_prefix":{"samples":65760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9971946318902115,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.625468370588983,"spectral_convergence":0.4654025252585317,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.540567455411301,"spectral_convergence":0.4205407216527011,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.255480182666016,"spectral_convergence":0.4561446752903601,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0005419945448040596,"delta_db":-1.7351419771112833,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0008215900965761254,"delta_db":-0.08142292112773364,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.0010538675765937977,"delta_db":-0.8344788258765733,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.00032195654134103545,"delta_db":-0.8008966014659548,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":6.541018681388496e-05,"delta_db":-3.057174430543142,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.7778876303231457e-05,"delta_db":-0.4245772177314184,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1556153289760174,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8860891578957002,"pitch":null,"output_rms_dbfs":-25.493509978507195,"output_peak":0.5255196243026825,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.473838669555432},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65760,"output_samples":65760,"intelligibility":{"estoi":{"value":0.6751393280694558,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05883407895338475,"rms_dbfs":-24.60742082061149,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.980889206210556,"mean":-1.1628793683945407e-05,"zero_energy":false},"output":{"rms":0.05312812647206877,"rms_dbfs":-25.493509978507195,"peak":0.5255196243026825,"peak_dbfs":-5.588221232401418,"crest_factor_db":19.905288746105775,"mean":-0.00041221542439561073,"zero_energy":false},"rms_difference_db":-0.8860891578957037},"full_scale":{"reference":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":411,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.508459147065878e-06,"output_mean_power":1.4362851359946489e-06,"output_rms_dbfs":-58.42759333982026,"output_zero_energy":false,"output_peak":0.005367146050891246,"output_energy_fraction":8.41896598205783e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":404,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":109,"both_estimated_voiced_frames":81,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.9,"absolute_difference_cents_median":8.251456675920346,"absolute_difference_cents_p95":78.96055417280718,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00020667206862955213,"log_rms_mae_db":2.257811082271695,"rms_envelope_nrmse":0.13636316476295107,"envelope_correlation":0.9923418402929678,"rms_gain_error_db":-0.8860875119678259,"reference_mean_power":0.003465660170380573,"output_mean_power":0.0028260329682509577},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":2.3006803943986064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009732960931433551,"output_power":0.0008789482601636834,"reference_active":true,"delta_db":-0.4428166928899676},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.170752162168666e-05,"output_power":5.352753169827116e-05,"reference_active":true,"delta_db":-1.8368482020946009},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0779113359754193e-05,"output_power":1.692758164672785e-05,"reference_active":true,"delta_db":-0.8903209516034293}],"active_band_count":3,"active_band_mae_db":1.056661948862666}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00018941444655877122,"log_rms_mae_db":2.0791704785394476,"rms_envelope_nrmse":0.11818810169728604,"envelope_correlation":0.996236001850946,"rms_gain_error_db":-0.8860770415614858,"reference_mean_power":0.003491154294575285,"output_mean_power":0.002846828723801056},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":2.1871865789273095e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008996086963166702,"output_power":0.0008099544979642145,"reference_active":true,"delta_db":-0.4559302346417326},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8292682982878605e-05,"output_power":2.3957077814458623e-05,"reference_active":true,"delta_db":-2.0368195282709105},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1507764875014338e-06,"output_power":8.582737581143663e-07,"reference_active":true,"delta_db":-1.2736514577656588}],"active_band_count":3,"active_band_mae_db":1.255467073559434}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.0001875778806757321,"log_rms_mae_db":3.3351701837401984,"rms_envelope_nrmse":0.1954259860086985,"envelope_correlation":0.974665867860878,"rms_gain_error_db":-0.2274231384029079,"reference_mean_power":0.0016622862050937872,"output_mean_power":0.0015774786385490292},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":1.477365390744301e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007590270471846978,"output_power":0.000680381001170499,"reference_active":true,"delta_db":-0.47505073875866133},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5779188180805434e-05,"output_power":3.868973296234185e-05,"reference_active":true,"delta_db":-1.5887645696643318},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.466208519389042e-05,"output_power":1.1333868601573905e-05,"reference_active":true,"delta_db":-1.1181756526941913}],"active_band_count":3,"active_band_mae_db":1.060663653705728}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016891518288067464,"log_rms_mae_db":2.5724436910117197,"rms_envelope_nrmse":0.16784556014681679,"envelope_correlation":0.980795770688361,"rms_gain_error_db":-0.22742302520031998,"reference_mean_power":0.001674521427963224,"output_mean_power":0.0015890896785184488},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.422523294383883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007062170888212691,"output_power":0.0006299029987647942,"reference_active":true,"delta_db":-0.49664546386305364},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.54990695717255e-05,"output_power":1.634314717175329e-05,"reference_active":true,"delta_db":-1.9318864228408958},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.502016839919999e-07,"output_power":6.865407529560682e-07,"reference_active":true,"delta_db":-1.4114947376770812}],"active_band_count":3,"active_band_mae_db":1.2800088747936769}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00010800393678736649,"log_rms_mae_db":4.249971709923993,"rms_envelope_nrmse":0.37405626748777504,"envelope_correlation":0.9071191497425316,"rms_gain_error_db":-1.195598980940359,"reference_mean_power":0.00040899310436077964,"output_mean_power":0.0003105675583941645},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":4.3909884369057597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019183671841687986,"output_power":0.00015932473035577482,"reference_active":true,"delta_db":-0.8064854457786099},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4527862162762088e-05,"output_power":1.3862653442744923e-05,"reference_active":true,"delta_db":-0.2035534440721022},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.245905579917374e-06,"output_power":4.967943188329839e-06,"reference_active":true,"delta_db":-2.2006173683059855}],"active_band_count":3,"active_band_mae_db":1.0702187527188993}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":4.040679324413719,"rms_envelope_nrmse":0.3439683663036823,"envelope_correlation":0.9205993575559105,"rms_gain_error_db":-1.1955970871614827,"reference_mean_power":0.0004120037826983826,"output_mean_power":0.0003128538434350531},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":4.2438330741066267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018201492488703654,"output_power":0.00015011531101663826,"reference_active":true,"delta_db":-0.8368201040510336},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.209318154228541e-06,"output_power":7.080944597904318e-06,"reference_active":true,"delta_db":-0.07802995352946056},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163481758524234e-07,"output_power":2.5700729668616687e-07,"reference_active":true,"delta_db":-0.9021988099763691}],"active_band_count":3,"active_band_mae_db":0.6056829558522877}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":3.62365774517211,"rms_envelope_nrmse":0.37667504171500604,"envelope_correlation":0.9410817914638314,"rms_gain_error_db":-2.6760337270369376,"reference_mean_power":8.457687952311121e-05,"output_mean_power":4.5671816494282325e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":3.6268766197283325e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4506351813371544e-05,"output_power":1.0029777890921985e-05,"reference_active":true,"delta_db":-1.6026689023181246},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.817293180171656e-06,"output_power":4.519948842713863e-06,"reference_active":true,"delta_db":-2.3792288061976667},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650085525655338e-06,"output_power":7.03354985832292e-07,"reference_active":true,"delta_db":-3.473421448871568}],"active_band_count":3,"active_band_mae_db":2.4851063857957865}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":3.9329904855704174,"rms_envelope_nrmse":0.33166207344033743,"envelope_correlation":0.9660171440721377,"rms_gain_error_db":-2.6760295636066047,"reference_mean_power":8.519942751057572e-05,"output_mean_power":4.600803873344341e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":3.3424099706380933e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5109841043230118e-05,"output_power":1.0029270956726553e-05,"reference_active":true,"delta_db":-1.7799053092163324},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127479817162532e-06,"output_power":2.3077841957536816e-06,"reference_active":true,"delta_db":-2.52489763421016},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2069317938633797e-07,"output_power":5.980728287108167e-08,"reference_active":true,"delta_db":-3.049286556869088}],"active_band_count":3,"active_band_mae_db":2.451363166765194}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"golf","status":"measured","mos":{"status":"measured","value":3.189939260482788,"predicted_mos":3.189939260482788,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":98400,"input_sample_rate":24000,"input_seconds":4.1,"input_float32_sha256":"2ee9dca237e7276f22bccb4e1543295cf25957d2a89bf0891b33ece0f2c92f52","internal_samples":65600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3056230369984405},"started_at":"2026-10-13T05:03:07.450878+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0026540720000411966,"native_input_samples":98640,"native_input_sha256":"ac87ed57131704b5a69dde0d07d419bcdf07b4a5049e019ac208073e37907744","output_sha256":"4a6ddd91f468cb5b90396d4a14713849d168b7fa998b5cfa762212cc00a87bea","output_samples":98400,"length_delta_samples":-240,"analysis_s":0.6136583099996642,"synthesis_s":0.19950778600014019,"total_s":0.8131660959998044,"rtf":0.1978506316301227,"output_peak":0.49662572145462036,"output_rms":0.05400049449747651,"finished_at":"2026-10-13T05:03:08.275581+00:00","analysis_stages":{"external_f0_s":0.05744086499998957},"evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"4a6ddd91f468cb5b90396d4a14713849d168b7fa998b5cfa762212cc00a87bea","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":65760,"output_samples":65600,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.347792816210454,"output_peak":0.49705794775920875},"common_prefix":{"samples":65600,"sample_rate":16000,"reference_coverage":0.9975669099756691,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65600,"output_samples":65600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.8656898228045415,"spectral_convergence":0.6675830293334791,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.701204378979701,"spectral_convergence":0.3341605161921998,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.917577066520022,"spectral_convergence":0.3355095389357611,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008079297463505974,"output_power":0.0007442686129639456,"delta_db":-0.35643893813308125,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008409804000831533,"output_power":0.0007466091693497898,"delta_db":-0.5169255489971079,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012804528734860071,"output_power":0.001043419430918493,"delta_db":-0.889046790339701,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00038829918073555097,"output_power":0.00027369179466472606,"delta_db":-1.5190469686900143,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013256419026499408,"output_power":9.57859694052828e-05,"delta_db":-1.4112432449563665,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.9659231241904317e-05,"output_power":1.5135105320627833e-05,"delta_db":-1.1358108367176696,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9714187213056568,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7509452533839046,"pitch":null,"output_rms_dbfs":-25.347792816210454,"output_peak":0.49705794775920875,"reference_rms_dbfs":-24.59684756282655,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.161490422768089},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65600,"output_samples":65600,"intelligibility":{"estoi":{"value":0.8429221027939328,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05890574077293748,"rms_dbfs":-24.59684756282655,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.97031594842561,"mean":-1.184044814691311e-05,"zero_energy":false},"output":{"rms":0.05402693859461203,"rms_dbfs":-25.347792816210454,"peak":0.49705794775920875,"peak_dbfs":-6.071859552293591,"crest_factor_db":19.275933263916862,"mean":0.00034815390416597893,"zero_energy":false},"rms_difference_db":-0.7509452533839038},"full_scale":{"reference":{"sample_count":65600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":410,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":67,"quiet_samples":10720,"reference_mean_power":2.514638194341713e-06,"output_mean_power":2.6230958349106187e-06,"output_rms_dbfs":-55.811858421536265,"output_zero_energy":false,"output_peak":0.006893010690339325,"output_energy_fraction":0.00014685352833841624,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":403,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":106,"both_estimated_voiced_frames":83,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.9222222222222223,"absolute_difference_cents_median":4.258223952533648,"absolute_difference_cents_p95":14.251837654425117,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":819,"rms_floor":0.00020667206862955213,"log_rms_mae_db":1.9895799238381568,"rms_envelope_nrmse":0.14799395085924605,"envelope_correlation":0.9870320350140508,"rms_gain_error_db":-0.7509495642260815,"reference_mean_power":0.0034741183073741474,"output_mean_power":0.002922467211910549},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":819,"frequency_resolution_hz":0.24420024420024422,"power_floor":2.3029905573079003e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009699406156445315,"output_power":0.0008409228421282952,"reference_active":true,"delta_db":-0.6198899606229967},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.152020835102618e-05,"output_power":7.587761261262948e-05,"reference_active":true,"delta_db":-0.3115162304069703},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.102559127358178e-05,"output_power":1.6661605572199674e-05,"reference_active":true,"delta_db":-1.010313684392971}],"active_band_count":3,"active_band_mae_db":0.6472399584743127}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":813,"rms_floor":0.00018941444655877122,"log_rms_mae_db":1.4024022175632669,"rms_envelope_nrmse":0.10966917734450257,"envelope_correlation":0.9953844662393085,"rms_gain_error_db":-0.7509644890820099,"reference_mean_power":0.003499733570972882,"output_mean_power":0.002944004941093862},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":813,"frequency_resolution_hz":0.24600246002460022,"power_floor":2.1894772289264065e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008962254016149879,"output_power":0.0007832503823930036,"reference_active":true,"delta_db":-0.5851663332874992},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.833688817176334e-05,"output_power":3.430535579937905e-05,"reference_active":true,"delta_db":-0.48254930040264365},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1490427618428703e-06,"output_power":1.185385121567745e-06,"reference_active":true,"delta_db":0.13523280524039133}],"active_band_count":3,"active_band_mae_db":0.4009828129768447}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":819,"rms_floor":0.0001875778806757321,"log_rms_mae_db":2.2072900279572707,"rms_envelope_nrmse":0.17486573432273897,"envelope_correlation":0.9818981134456498,"rms_gain_error_db":-0.6251859471085942,"reference_mean_power":0.0016663448984069868,"output_mean_power":0.0014429334506107423},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":819,"frequency_resolution_hz":0.24420024420024422,"power_floor":1.4808619205412186e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007575294616803145,"output_power":0.0006934587078973627,"reference_active":true,"delta_db":-0.38378921945421746},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5496897207075e-05,"output_power":5.695391554110389e-05,"reference_active":true,"delta_db":0.11254884178638164},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4774905680233563e-05,"output_power":9.366059105842479e-06,"reference_active":true,"delta_db":-1.9796782298960955}],"active_band_count":3,"active_band_mae_db":0.8253387637122316}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":813,"rms_floor":0.00016891518288067464,"log_rms_mae_db":1.4686471280813704,"rms_envelope_nrmse":0.12535457437376135,"envelope_correlation":0.9920954110723087,"rms_gain_error_db":-0.6251842339886269,"reference_mean_power":0.0016786399942727045,"output_mean_power":0.0014535806825066546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":813,"frequency_resolution_hz":0.24600246002460022,"power_floor":1.4259897649469757e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007042461298481542,"output_power":0.0006496303487043618,"reference_active":true,"delta_db":-0.3505816342811984},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.542523011915813e-05,"output_power":2.4787518502839773e-05,"reference_active":true,"delta_db":-0.11031840937317487},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.499289181256377e-07,"output_power":7.870422617382992e-07,"reference_active":true,"delta_db":-0.8169305558858242}],"active_band_count":3,"active_band_mae_db":0.42594353318006584}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":819,"rms_floor":0.00010800393678736649,"log_rms_mae_db":2.1765329237314948,"rms_envelope_nrmse":0.21504465415702703,"envelope_correlation":0.98367897191404,"rms_gain_error_db":-1.5123272146757658,"reference_mean_power":0.00040999176063991175,"output_mean_power":0.0002894292422433707},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":819,"frequency_resolution_hz":0.24420024420024422,"power_floor":4.384683212907659e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019095778904557896,"output_power":0.00013140142433256657,"reference_active":true,"delta_db":-1.6233730485675744},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4536107517009606e-05,"output_power":9.920727253652588e-06,"reference_active":true,"delta_db":-1.6590461666815228},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.270856653468972e-06,"output_power":3.6628903127827856e-06,"reference_active":true,"delta_db":-3.537265802286449}],"active_band_count":3,"active_band_mae_db":2.2732283391785155}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":813,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":1.7328118724118187,"rms_envelope_nrmse":0.17966175790895553,"envelope_correlation":0.9945005932871143,"rms_gain_error_db":-1.5123275931800348,"reference_mean_power":0.0004130172218500612,"output_mean_power":0.00029156500840076785},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":813,"frequency_resolution_hz":0.24600246002460022,"power_floor":4.2371602148709084e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018114540128217124,"output_power":0.00012439023674192245,"reference_active":true,"delta_db":-1.6324101874751098},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.255805553616058e-06,"output_power":4.94126226022589e-06,"reference_active":true,"delta_db":-1.668477307459855},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.166415583965881e-07,"output_power":1.723698685383961e-07,"reference_active":true,"delta_db":-2.641065639069384}],"active_band_count":3,"active_band_mae_db":1.9806510446681163}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":819,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":2.4762244239824907,"rms_envelope_nrmse":0.2210906507301403,"envelope_correlation":0.9783661122928597,"rms_gain_error_db":-1.2360282273898835,"reference_mean_power":8.478338210170061e-05,"output_mean_power":6.378343650409637e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":819,"frequency_resolution_hz":0.24420024420024422,"power_floor":3.603922523861187e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4435240624319984e-05,"output_power":1.1651809481498229e-05,"reference_active":true,"delta_db":-0.9303065275752677},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.754369033675257e-06,"output_power":6.529479175651242e-06,"reference_active":true,"delta_db":-0.746679242917679},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5594317165798471e-06,"output_power":1.0539439047315773e-06,"reference_active":true,"delta_db":-1.7014886635747482}],"active_band_count":3,"active_band_mae_db":1.1261581446892315}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65600,"uncovered_tail_samples":0,"metrics":{"frame_count":813,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":2.179331997808004,"rms_envelope_nrmse":0.17664494827105878,"envelope_correlation":0.9901914306504134,"rms_gain_error_db":-1.2360288807069162,"reference_mean_power":8.540898797677399e-05,"output_mean_power":6.425407676538455e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":813,"frequency_resolution_hz":0.24600246002460022,"power_floor":3.3201722083793795e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5018104617694129e-05,"output_power":1.2193859665272867e-05,"reference_active":true,"delta_db":-0.9047393276706649},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.107837057754053e-06,"output_power":3.3441572340125024e-06,"reference_active":true,"delta_db":-0.8932651975859336},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1928977043669034e-07,"output_power":7.674021216885528e-08,"reference_active":true,"delta_db":-1.9158020723327098}],"active_band_count":3,"active_band_mae_db":1.2379355325297694}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.390996217727661,"predicted_mos":3.390996217727661,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":98760,"input_sample_rate":24000,"input_seconds":4.115,"input_float32_sha256":"13bddbb8e5258e0125df110c56462f792cc73eafdafb58d1071e7b2ae3498cb9","internal_samples":65840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.055044026001269},"started_at":"2026-10-13T07:46:50.227078+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002962085000035586,"native_input_samples":98640,"native_input_sha256":"ac87ed57131704b5a69dde0d07d419bcdf07b4a5049e019ac208073e37907744","output_sha256":"ddb65ea7c9215614b8b9c1c3b52167ec56acadf94f2b6c882615c7c0c39a27b4","output_samples":98760,"length_delta_samples":120,"analysis_s":1.8655513590019837,"synthesis_s":13.225420824997855,"total_s":15.090972183999838,"rtf":3.67176938783451,"output_peak":0.5653651356697083,"output_rms":0.05775166602301534,"finished_at":"2026-10-13T07:47:05.335978+00:00","analysis_stages":{"external_f0_s":1.280293617001007},"evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"ddb65ea7c9215614b8b9c1c3b52167ec56acadf94f2b6c882615c7c0c39a27b4","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":65760,"output_samples":65840,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.764984404075314,"output_peak":0.5518488927337246},"common_prefix":{"samples":65760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9987849331713244,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.419839089107542,"spectral_convergence":0.6002785405630919,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.564837871290934,"spectral_convergence":0.31672312655993995,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.768168481924748,"spectral_convergence":0.31247589752945565,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0010491826886803188,"delta_db":1.1334200311821836,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0007591239522617224,"delta_db":-0.4248479303967606,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.0010828795880167797,"delta_db":-0.7165375829399524,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.00032437683533755144,"delta_db":-0.7683708097205812,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00011355159819832368,"delta_db":-0.6616958140996181,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.3062813369907659e-05,"delta_db":-1.7632530803744353,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.911354208118922,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1522840992122351,"pitch":null,"output_rms_dbfs":-24.75970491982373,"output_peak":0.5518488927337246,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.917615147441076},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65760,"output_samples":65760,"intelligibility":{"estoi":{"value":0.8704074373200251,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05883407895338475,"rms_dbfs":-24.60742082061149,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.980889206210556,"mean":-1.1628793683945407e-05,"zero_energy":false},"output":{"rms":0.05781156870267925,"rms_dbfs":-24.75970491982373,"peak":0.5518488927337246,"peak_dbfs":-5.163596490192289,"crest_factor_db":19.59610842963144,"mean":-1.6163971062901944e-05,"zero_energy":false},"rms_difference_db":-0.1522840992122383},"full_scale":{"reference":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":411,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.508459147065878e-06,"output_mean_power":9.272301643230801e-07,"output_rms_dbfs":-60.32812448500483,"output_zero_energy":false,"output_peak":0.004759241632189868,"output_energy_fraction":4.590131571313111e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":404,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":121,"both_estimated_voiced_frames":89,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9888888888888889,"absolute_difference_cents_median":7.918831288732174,"absolute_difference_cents_p95":24.35974739327662,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00020667206862955213,"log_rms_mae_db":3.0792271361154455,"rms_envelope_nrmse":0.15759039979635187,"envelope_correlation":0.9799675597447481,"rms_gain_error_db":-0.1522785354473792,"reference_mean_power":0.003465660170380573,"output_mean_power":0.003346247976970758},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":2.3006803943986064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009732960931433551,"output_power":0.0008567156260272887,"reference_active":true,"delta_db":-0.5540829183189931},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.170752162168666e-05,"output_power":6.720141891681557e-05,"reference_active":true,"delta_db":-0.8488359452869099},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0779113359754193e-05,"output_power":1.518601493923006e-05,"reference_active":true,"delta_db":-1.3618318958650588}],"active_band_count":3,"active_band_mae_db":0.9215835864903207}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00018941444655877122,"log_rms_mae_db":2.648312545560729,"rms_envelope_nrmse":0.10446768920645222,"envelope_correlation":0.9909552944711492,"rms_gain_error_db":-0.15225897276579786,"reference_mean_power":0.003491154294575285,"output_mean_power":0.003370878863989206},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":2.1871865789273095e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008996086963166702,"output_power":0.0007898499804075332,"reference_active":true,"delta_db":-0.5650903330130828},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8292682982878605e-05,"output_power":3.100486160348911e-05,"reference_active":true,"delta_db":-0.9168599919970455},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1507764875014338e-06,"output_power":9.290028120613456e-07,"reference_active":true,"delta_db":-0.9297395127604122}],"active_band_count":3,"active_band_mae_db":0.8038966125901802}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.0001875778806757321,"log_rms_mae_db":2.6729684820301984,"rms_envelope_nrmse":0.1951304165145184,"envelope_correlation":0.9770227503594924,"rms_gain_error_db":-0.6858217584578801,"reference_mean_power":0.0016622862050937872,"output_mean_power":0.0014194615237816727},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":1.477365390744301e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007590270471846978,"output_power":0.0006219355817667022,"reference_active":true,"delta_db":-0.8651184773666686},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5779188180805434e-05,"output_power":4.225281912978623e-05,"reference_active":true,"delta_db":-1.2061649858169707},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.466208519389042e-05,"output_power":1.191715337874412e-05,"reference_active":true,"delta_db":-0.9002320970850191}],"active_band_count":3,"active_band_mae_db":0.9905051867562195}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016891518288067464,"log_rms_mae_db":2.0224413460516284,"rms_envelope_nrmse":0.1341767273223545,"envelope_correlation":0.9910004227205347,"rms_gain_error_db":-0.6858183778177955,"reference_mean_power":0.001674521427963224,"output_mean_power":0.0014299105537317275},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.422523294383883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007062170888212691,"output_power":0.0005782474527271399,"reference_active":true,"delta_db":-0.8682449396694575},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.54990695717255e-05,"output_power":2.012405334196441e-05,"reference_active":true,"delta_db":-1.0280887405464327},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.502016839919999e-07,"output_power":7.517290760475484e-07,"reference_active":true,"delta_db":-1.017544471607072}],"active_band_count":3,"active_band_mae_db":0.9712927172743208}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00010800393678736649,"log_rms_mae_db":2.1745306871926093,"rms_envelope_nrmse":0.20026142545592185,"envelope_correlation":0.9757305859959972,"rms_gain_error_db":-0.7356123696880792,"reference_mean_power":0.00040899310436077964,"output_mean_power":0.00034526674371597666},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":4.3909884369057597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019183671841687986,"output_power":0.00014668636371822872,"reference_active":true,"delta_db":-1.165419939571066},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4527862162762088e-05,"output_power":1.1057238686013451e-05,"reference_active":true,"delta_db":-1.1855502615414464},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.245905579917374e-06,"output_power":4.677276439801744e-06,"reference_active":true,"delta_db":-2.4624531893078174}],"active_band_count":3,"active_band_mae_db":1.6044744634734434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":1.5156071992244875,"rms_envelope_nrmse":0.13291514945853755,"envelope_correlation":0.991688685614545,"rms_gain_error_db":-0.735612492541992,"reference_mean_power":0.0004120037826983826,"output_mean_power":0.00034780830999397153},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":4.2438330741066267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018201492488703654,"output_power":0.000138331431404314,"reference_active":true,"delta_db":-1.1918612990235393},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.209318154228541e-06,"output_power":5.75036364914463e-06,"reference_active":true,"delta_db":-0.9819888173459941},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163481758524234e-07,"output_power":2.2636389970546542e-07,"reference_active":true,"delta_db":-1.4535816736814364}],"active_band_count":3,"active_band_mae_db":1.20914393001699}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":2.0060895125376743,"rms_envelope_nrmse":0.2819595509403254,"envelope_correlation":0.9532526751587179,"rms_gain_error_db":-0.9044982274811618,"reference_mean_power":8.457687952311121e-05,"output_mean_power":6.867550078453412e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":3.6268766197283325e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4506351813371544e-05,"output_power":1.1507673202463595e-05,"reference_active":true,"delta_db":-1.0057068572337},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.817293180171656e-06,"output_power":5.813374870553383e-06,"reference_active":true,"delta_db":-1.286280710943863},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650085525655338e-06,"output_power":1.0165993033619696e-06,"reference_active":true,"delta_db":-1.8736690749028666}],"active_band_count":3,"active_band_mae_db":1.3885522143601434}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":1.233201162230774,"rms_envelope_nrmse":0.16975866797218198,"envelope_correlation":0.9867766083967526,"rms_gain_error_db":-0.9044991865189699,"reference_mean_power":8.519942751057572e-05,"output_mean_power":6.918098766181278e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":3.3424099706380933e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5109841043230118e-05,"output_power":1.1713548224491038e-05,"reference_active":true,"delta_db":-1.1057142567769784},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127479817162532e-06,"output_power":3.089904315167067e-06,"reference_active":true,"delta_db":-1.2573992742797229},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2069317938633797e-07,"output_power":6.471182243165917e-08,"reference_active":true,"delta_db":-2.7069909722665244}],"active_band_count":3,"active_band_mae_db":1.6900348344410752}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.53513765335083,"predicted_mos":3.53513765335083,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":98880,"input_sample_rate":24000,"input_seconds":4.12,"input_float32_sha256":"72edc508dc512f69f02d4155888f0dc3cc81f683485dc6fba3f3d9e6e91b2e1b","internal_samples":65920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.039400797000781},"started_at":"2026-10-13T05:11:35.200824+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.001946471000337624,"native_input_samples":98640,"native_input_sha256":"ac87ed57131704b5a69dde0d07d419bcdf07b4a5049e019ac208073e37907744","output_sha256":"7efe80cb52377ee0d79e1f62da79119c1653f75df82cd3c1de69c5a8d395901d","output_samples":98880,"length_delta_samples":240,"analysis_s":0.512410307999744,"synthesis_s":0.07278711199978716,"total_s":0.5851974199995311,"rtf":0.1423838004865039,"output_peak":0.4790019094944,"output_rms":0.05745671500900372,"finished_at":"2026-10-13T05:11:35.793874+00:00","analysis_stages":{"external_f0_s":0.07711649100019713},"evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"7efe80cb52377ee0d79e1f62da79119c1653f75df82cd3c1de69c5a8d395901d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":65760,"output_samples":65920,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.809248747304014,"output_peak":0.4796540143179132},"common_prefix":{"samples":65760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9975728155339806,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.067092016346755,"spectral_convergence":0.1820943003866187,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.6834204525644685,"spectral_convergence":0.16852179749830057,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.252427033464789,"spectral_convergence":0.18067654367180672,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0007969294591117532,"delta_db":-0.06089231928573742,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0007985395295158203,"delta_db":-0.20501068192706887,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.0012319058645519966,"delta_db":-0.1565640304429704,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.00035531718944503986,"delta_db":-0.37270695241627194,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00011433710672142788,"delta_db":-0.6317563427030022,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.5276978717120583e-05,"delta_db":-1.0832455651305761,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4183626486509378,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1912744985478423,"pitch":null,"output_rms_dbfs":-24.798695319159336,"output_peak":0.4796540143179132,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.000979834125338},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65760,"output_samples":65760,"intelligibility":{"estoi":{"value":0.9324811559832048,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05883407895338475,"rms_dbfs":-24.60742082061149,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.980889206210556,"mean":-1.1628793683945407e-05,"zero_energy":false},"output":{"rms":0.057552637889701955,"rms_dbfs":-24.798695319159336,"peak":0.4796540143179132,"peak_dbfs":-6.381438329673007,"crest_factor_db":18.41725698948633,"mean":0.0002071432266862468,"zero_energy":false},"rms_difference_db":-0.19127449854784473},"full_scale":{"reference":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":411,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.508459147065878e-06,"output_mean_power":2.7144115495423604e-06,"output_rms_dbfs":-55.66324305480591,"output_zero_energy":false,"output_peak":0.006346249436315086,"output_energy_fraction":0.00013558520973868255,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":404,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":94,"both_estimated_voiced_frames":88,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9777777777777777,"absolute_difference_cents_median":1.2404523397893437,"absolute_difference_cents_p95":13.93921723600387,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00020667206862955213,"log_rms_mae_db":0.9778149994077566,"rms_envelope_nrmse":0.06674919039779147,"envelope_correlation":0.9967517092403411,"rms_gain_error_db":-0.19127344211688518,"reference_mean_power":0.003465660170380573,"output_mean_power":0.003316336806421814},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":2.3006803943986064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009732960931433551,"output_power":0.000953364744688634,"reference_active":true,"delta_db":-0.08985892368969033},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.170752162168666e-05,"output_power":6.68137120464519e-05,"reference_active":true,"delta_db":-0.8739643651803987},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0779113359754193e-05,"output_power":1.4323315638194077e-05,"reference_active":true,"delta_db":-1.6158344993178457}],"active_band_count":3,"active_band_mae_db":0.8598859293959782}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00018941444655877122,"log_rms_mae_db":0.5781025466059568,"rms_envelope_nrmse":0.042020197950026636,"envelope_correlation":0.998910949842477,"rms_gain_error_db":-0.1912712935564156,"reference_mean_power":0.003491154294575285,"output_mean_power":0.0033407341293270813},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":2.1871865789273095e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008996086963166702,"output_power":0.0008817098095002504,"reference_active":true,"delta_db":-0.08727972430789804},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8292682982878605e-05,"output_power":3.110901269922929e-05,"reference_active":true,"delta_db":-0.9022956811845778},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1507764875014338e-06,"output_power":7.224590908585214e-07,"reference_active":true,"delta_db":-2.0217771960434696}],"active_band_count":3,"active_band_mae_db":1.003784200511982}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.0001875778806757321,"log_rms_mae_db":1.2718101183242267,"rms_envelope_nrmse":0.08829118214491533,"envelope_correlation":0.9950666241773596,"rms_gain_error_db":-0.18272267943998155,"reference_mean_power":0.0016622862050937872,"output_mean_power":0.0015937989370537},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":1.477365390744301e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007590270471846978,"output_power":0.0007494382360867308,"reference_active":true,"delta_db":-0.055214049478841884},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5779188180805434e-05,"output_power":4.5471818445417395e-05,"reference_active":true,"delta_db":-0.8872986693452936},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.466208519389042e-05,"output_power":8.380991148147264e-06,"reference_active":true,"delta_db":-2.4290035667918377}],"active_band_count":3,"active_band_mae_db":1.1238387618719912}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016891518288067464,"log_rms_mae_db":0.6851124232781878,"rms_envelope_nrmse":0.05204371503993325,"envelope_correlation":0.9984324459233545,"rms_gain_error_db":-0.1827240906022275,"reference_mean_power":0.001674521427963224,"output_mean_power":0.0016055295391299747},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.422523294383883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007062170888212691,"output_power":0.0006964758734666956,"reference_active":true,"delta_db":-0.0603214559698289},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.54990695717255e-05,"output_power":2.028787796183969e-05,"reference_active":true,"delta_db":-0.9928771008816906},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.502016839919999e-07,"output_power":5.559039697288725e-07,"reference_active":true,"delta_db":-2.3281602039984404}],"active_band_count":3,"active_band_mae_db":1.1271195869499866}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00010800393678736649,"log_rms_mae_db":1.1634993052720493,"rms_envelope_nrmse":0.09048870091251277,"envelope_correlation":0.9955721933906818,"rms_gain_error_db":-0.3724636317524818,"reference_mean_power":0.00040899310436077964,"output_mean_power":0.0003753786971721737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":4.3909884369057597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019183671841687986,"output_power":0.00018356306370260775,"reference_active":true,"delta_db":-0.1914643918108947},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4527862162762088e-05,"output_power":1.2783648239783938e-05,"reference_active":true,"delta_db":-0.5554689879815854},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.245905579917374e-06,"output_power":5.447291709714246e-06,"reference_active":true,"delta_db":-1.8005772464403864}],"active_band_count":3,"active_band_mae_db":0.8491702087442888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":0.8122352759872344,"rms_envelope_nrmse":0.057855405602254814,"envelope_correlation":0.9988996545503518,"rms_gain_error_db":-0.3724641295238329,"reference_mean_power":0.0004120037826983826,"output_mean_power":0.000378141889934335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":4.2438330741066267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018201492488703654,"output_power":0.00017356132175366036,"reference_active":true,"delta_db":-0.20654051956149855},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.209318154228541e-06,"output_power":6.354556447295683e-06,"reference_active":true,"delta_db":-0.5480894991042049},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163481758524234e-07,"output_power":2.0888855983727957e-07,"reference_active":true,"delta_db":-1.8025067884238632}],"active_band_count":3,"active_band_mae_db":0.8523789356965222}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":1.2592907439361605,"rms_envelope_nrmse":0.14914256648213292,"envelope_correlation":0.989319491884219,"rms_gain_error_db":-0.6842098587638328,"reference_mean_power":8.457687952311121e-05,"output_mean_power":7.224880560558367e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":3.6268766197283325e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4506351813371544e-05,"output_power":1.3620693326745274e-05,"reference_active":true,"delta_db":-0.2735899112745422},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.817293180171656e-06,"output_power":6.709048061721174e-06,"reference_active":true,"delta_db":-0.6639549701573838},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650085525655338e-06,"output_power":1.0004026316654362e-06,"reference_active":true,"delta_db":-1.9434188973331812}],"active_band_count":3,"active_band_mae_db":0.960321259588369}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":0.9382443229261139,"rms_envelope_nrmse":0.10473927487192332,"envelope_correlation":0.9965124680865298,"rms_gain_error_db":-0.6842107599872547,"reference_mean_power":8.519942751057572e-05,"output_mean_power":7.278059480629374e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":3.3424099706380933e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5109841043230118e-05,"output_power":1.3872677557625503e-05,"reference_active":true,"delta_db":-0.37099603454193075},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127479817162532e-06,"output_power":3.5529219271397663e-06,"reference_active":true,"delta_db":-0.6509929400605831},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2069317938633797e-07,"output_power":7.047336137246654e-08,"reference_active":true,"delta_db":-2.3365774137526127}],"active_band_count":3,"active_band_mae_db":1.1195221294517088}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0049","method":"magphase","status":"measured","mos":{"status":"measured","value":3.8055384159088135,"predicted_mos":3.8055384159088135,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":198828,"input_sample_rate":48000,"input_seconds":4.14225,"input_float32_sha256":"2eed87f3cac108407d53f06f93425e5c4a312f5305bc313a85251b455d45594e","internal_samples":66276,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.056337122001423},"started_at":"2026-10-13T05:18:46.811327+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008151575999363558,"native_input_samples":197280,"native_input_sha256":"60c2bf46780cf4b6341d7061816f130d370a998b49673ca75c8374daa1edc95c","output_sha256":"28b2e6fa8c106fec3cc13bd9e6f9196655b0a741678e2b671e9ba38c8de2f423","output_samples":198828,"length_delta_samples":1548,"analysis_s":2.738252131999616,"synthesis_s":0.7681075960008457,"total_s":3.5063597280004615,"rtf":0.8531288875913531,"output_peak":0.48424229065757385,"output_rms":0.05662295126515424,"finished_at":"2026-10-13T05:18:50.333746+00:00","evaluation_input":{"reference":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"28b2e6fa8c106fec3cc13bd9e6f9196655b0a741678e2b671e9ba38c8de2f423","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":65760,"output_samples":66276,"length_delta_samples":516,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.936161000177968,"output_peak":0.48357051158610465},"common_prefix":{"samples":65760,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9922143762447945,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":65760,"output_samples":65760,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.045692100817208,"spectral_convergence":0.7252630056025461,"spectral_convergence_status":"ok","amplitude_floor":1.6529112798953643e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.275175522540859,"spectral_convergence":0.49659045264251295,"spectral_convergence_status":"ok","amplitude_floor":1.0394789070711673e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.792404931948438,"spectral_convergence":0.45110018435255533,"spectral_convergence_status":"ok","amplitude_floor":8.202976886987361e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0008081818875357184,"output_power":0.0007330085562252525,"delta_db":-0.4240006899944012,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0008371388318692298,"output_power":0.0007928066121762331,"delta_db":-0.23630224159170307,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0012771265272884344,"output_power":0.0011694141506935975,"delta_db":-0.3826558096605998,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003871568194673652,"output_power":0.00037567562291894674,"delta_db":-0.1307389822924977,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001322399975596578,"output_power":0.00014907707566759855,"delta_db":0.5204803238954321,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.960478257270517e-05,"output_power":1.4286855164777779e-05,"delta_db":-1.374253880853253,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5114053213813144,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2948179688309339,"pitch":null,"output_rms_dbfs":-24.902238789442425,"output_peak":0.48357051158610465,"reference_rms_dbfs":-24.60742082061149,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.371090851768835},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":65760,"output_samples":65760,"intelligibility":{"estoi":{"value":0.6392876485108948,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05883407895338475,"rms_dbfs":-24.60742082061149,"peak":0.46630859375,"peak_dbfs":-6.626531614400935,"crest_factor_db":17.980889206210556,"mean":-1.1628793683945407e-05,"zero_energy":false},"output":{"rms":0.056870632780429003,"rms_dbfs":-24.902238789442425,"peak":0.48357051158610465,"peak_dbfs":-6.310803810729268,"crest_factor_db":18.59143497871316,"mean":-0.0003479331444180136,"zero_energy":false},"rms_difference_db":-0.2948179688309338},"full_scale":{"reference":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":65760,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":411,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.27133439516183e-06,"threshold_dbfs":-53.69436427100765,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":68,"quiet_samples":10880,"reference_mean_power":2.508459147065878e-06,"output_mean_power":3.3158698742548062e-06,"output_rms_dbfs":-54.79402520945031,"output_zero_energy":false,"output_peak":0.015402785875735691,"output_energy_fraction":0.00016962444857070074,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":404,"reference_estimated_voiced_frames":90,"output_estimated_voiced_frames":89,"both_estimated_voiced_frames":71,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.7888888888888889,"absolute_difference_cents_median":8.822191531135871,"absolute_difference_cents_p95":69.53750644143213,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":65760,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00020667206862955213,"log_rms_mae_db":3.5584619316044734,"rms_envelope_nrmse":0.31208021626694715,"envelope_correlation":0.9208987332683481,"rms_gain_error_db":-0.29481405516914494,"reference_mean_power":0.003465660170380573,"output_mean_power":0.0032382067196679997},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":2.3006803943986064e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009732960931433551,"output_power":0.0008963851629331153,"reference_active":true,"delta_db":-0.35750320668282237},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.170752162168666e-05,"output_power":8.121138148030579e-05,"reference_active":true,"delta_db":-0.02645139239756168},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0779113359754193e-05,"output_power":2.310134471841598e-05,"reference_active":true,"delta_db":0.4601024834010741}],"active_band_count":3,"active_band_mae_db":0.2813523608271527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00018941444655877122,"log_rms_mae_db":2.674042483728618,"rms_envelope_nrmse":0.23150835362950425,"envelope_correlation":0.9548142450785122,"rms_gain_error_db":-0.2948031895284038,"reference_mean_power":0.003491154294575285,"output_mean_power":0.0032620358097989314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":2.1871865789273095e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008996086963166702,"output_power":0.0008258815853999232,"reference_active":true,"delta_db":-0.37135862211992493},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.8292682982878605e-05,"output_power":3.8855074944060445e-05,"reference_active":true,"delta_db":0.06331954643395521},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1507764875014338e-06,"output_power":1.4321130108351342e-06,"reference_active":true,"delta_db":0.9498631048304331}],"active_band_count":3,"active_band_mae_db":0.46151375779477105}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.0001875778806757321,"log_rms_mae_db":3.9938594546169437,"rms_envelope_nrmse":0.3631012352763408,"envelope_correlation":0.9113863822528722,"rms_gain_error_db":-0.33769845647785557,"reference_mean_power":0.0016622862050937872,"output_mean_power":0.0015379278273119623},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":1.477365390744301e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007590270471846978,"output_power":0.0006961628940627221,"reference_active":true,"delta_db":-0.3754638044390318},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5779188180805434e-05,"output_power":5.242855346948155e-05,"reference_active":true,"delta_db":-0.26904313712062067},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.466208519389042e-05,"output_power":1.693007387031087e-05,"reference_active":true,"delta_db":0.6246311441185027}],"active_band_count":3,"active_band_mae_db":0.42304602855938506}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":0.00016891518288067464,"log_rms_mae_db":3.0507099437292644,"rms_envelope_nrmse":0.2718862454573792,"envelope_correlation":0.949029708362266,"rms_gain_error_db":-0.33769727647463343,"reference_mean_power":0.001674521427963224,"output_mean_power":0.0015492481338805488},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":1.422523294383883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007062170888212691,"output_power":0.0006469957896144647,"reference_active":true,"delta_db":-0.3803676781091408},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.54990695717255e-05,"output_power":2.4199520267690708e-05,"reference_active":true,"delta_db":-0.22717577281959753},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.502016839919999e-07,"output_power":1.1011414069127479e-06,"reference_active":true,"delta_db":0.6402729823217939}],"active_band_count":3,"active_band_mae_db":0.4159388110835107}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":0.00010800393678736649,"log_rms_mae_db":3.947735852558848,"rms_envelope_nrmse":0.4007530661628018,"envelope_correlation":0.8929090594985378,"rms_gain_error_db":-0.062715603621527,"reference_mean_power":0.00040899310436077964,"output_mean_power":0.0004031293566749683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":4.3909884369057597e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019183671841687986,"output_power":0.00018277664083753844,"reference_active":true,"delta_db":-0.21011045343912169},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4527862162762088e-05,"output_power":1.5672564666714962e-05,"reference_active":true,"delta_db":0.3293835975746742},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.245905579917374e-06,"output_power":1.009460037755437e-05,"reference_active":true,"delta_db":0.8785077329471833}],"active_band_count":3,"active_band_mae_db":0.4726672613203264}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":8.815867833009976e-05,"log_rms_mae_db":3.0700096074354315,"rms_envelope_nrmse":0.28529807634122284,"envelope_correlation":0.9435360594103268,"rms_gain_error_db":-0.06271495796945471,"reference_mean_power":0.0004120037826983826,"output_mean_power":0.000406096931190672},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":4.2438330741066267e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00018201492488703654,"output_power":0.00017308653675313914,"reference_active":true,"delta_db":-0.21843712470090287},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.209318154228541e-06,"output_power":8.121172404314304e-06,"reference_active":true,"delta_db":0.5172453843790691},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.163481758524234e-07,"output_power":5.236409691584888e-07,"reference_active":true,"delta_db":2.188682834394316}],"active_band_count":3,"active_band_mae_db":0.9747884478247627}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":821,"rms_floor":8.95305280668978e-05,"log_rms_mae_db":3.8354578565402315,"rms_envelope_nrmse":0.8066694179131756,"envelope_correlation":0.6268561483096021,"rms_gain_error_db":0.3503191266374886,"reference_mean_power":8.457687952311121e-05,"output_mean_power":9.168189270586614e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":821,"frequency_resolution_hz":0.24360535931790497,"power_floor":3.6268766197283325e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4506351813371544e-05,"output_power":1.660687824080374e-05,"reference_active":true,"delta_db":0.5872979557064825},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.817293180171656e-06,"output_power":8.745443683349785e-06,"reference_active":true,"delta_db":0.4872544740759742},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5650085525655338e-06,"output_power":2.139562264684565e-06,"reference_active":true,"delta_db":1.3580821441694992}],"active_band_count":3,"active_band_mae_db":0.8108781913173186}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":65760,"uncovered_tail_samples":0,"metrics":{"frame_count":815,"rms_floor":5.618561264778465e-05,"log_rms_mae_db":2.956916217103302,"rms_envelope_nrmse":0.5628352971720588,"envelope_correlation":0.8109180060007891,"rms_gain_error_db":0.3503179341874567,"reference_mean_power":8.519942751057572e-05,"output_mean_power":9.235671345316499e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":815,"frequency_resolution_hz":0.24539877300613497,"power_floor":3.3424099706380933e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.5109841043230118e-05,"output_power":1.7566496361410335e-05,"reference_active":true,"delta_db":0.6542525452649895},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.127479817162532e-06,"output_power":4.722008596267295e-06,"reference_active":true,"delta_db":0.5844181506609214},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2069317938633797e-07,"output_power":1.4095040308667382e-07,"reference_active":true,"delta_db":0.6738359425612749}],"active_band_count":3,"active_band_mae_db":0.6375022128290619}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"hnm","status":"measured","mos":{"status":"measured","value":2.7700517177581787,"predicted_mos":2.7700517177581787,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1280cee06630df722c9994055b5eaacf6e22e55d70b6ff336b9bbc89ef53d194","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1724743750000926},"started_at":"2026-10-13T04:40:26.647882+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.6001000403775834e-05,"native_input_samples":96000,"native_input_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","output_sha256":"e95c108f9271689e289c4087f43fcb1e15cda670bfcaf13f28ee05e6ab60c3ce","output_samples":96000,"length_delta_samples":0,"analysis_s":1.1600403780003035,"synthesis_s":0.4904163189994506,"total_s":1.650456696999754,"rtf":0.2750761161666257,"output_peak":0.7028618643703606,"output_rms":0.08418244400999339,"finished_at":"2026-10-13T04:40:28.303450+00:00","evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"e95c108f9271689e289c4087f43fcb1e15cda670bfcaf13f28ee05e6ab60c3ce","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.8203312656048185,"spectral_convergence":0.26519626431158316,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.588609421625493,"spectral_convergence":0.27133237423706463,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.17941216230582,"spectral_convergence":0.278258533855832,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.0016822669791708147,"delta_db":0.05569969386728607,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.001873118797468141,"delta_db":-0.06078547797871306,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.002488122911668912,"delta_db":-0.39890424308253025,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.0007280957615201692,"delta_db":-0.221432210013134,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00024392321495227414,"delta_db":-0.06289224802057297,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":7.115621471536057e-05,"delta_db":0.4673155921769633,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.21117157752319993,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16752888077049427,"pitch":null,"output_rms_dbfs":-21.49556939674307,"output_peak":0.7028618643703606,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.862784283178709},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.8203312656048185,"spectral_convergence":0.26519626431158316,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.588609421625493,"spectral_convergence":0.27133237423706463,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.17941216230582,"spectral_convergence":0.278258533855832,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.0016822669791708147,"delta_db":0.05569969386728607,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.001873118797468141,"delta_db":-0.06078547797871306,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.002488122911668912,"delta_db":-0.39890424308253025,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.0007280957615201692,"delta_db":-0.221432210013134,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00024392321495227414,"delta_db":-0.06289224802057297,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":7.115621471536057e-05,"delta_db":0.4673155921769633,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.21117157752319993,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.16752888077049427,"pitch":null,"output_rms_dbfs":-21.49556939674307,"output_peak":0.7028618643703606,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.862784283178709},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.866043559559662,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08582187018202725,"rms_dbfs":-21.328040515972575,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.582274007876475,"mean":1.6474088033040366e-05,"zero_energy":false},"output":{"rms":0.08418244400999339,"rms_dbfs":-21.49556939674307,"peak":0.7028618643703606,"peak_dbfs":-3.062600396800752,"crest_factor_db":18.432968999942318,"mean":0.00022284793059690948,"zero_energy":false},"rms_difference_db":-0.16752888077049377},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":6.7786142524337345e-06,"output_rms_dbfs":-51.688590795890185,"output_zero_energy":false,"output_peak":0.010334437042171272,"output_energy_fraction":9.565283802268514e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":158,"both_estimated_voiced_frames":139,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":19,"paired_voiced_fraction":0.9652777777777778,"absolute_difference_cents_median":1.8049486926866298,"absolute_difference_cents_p95":9.650028319199023,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031196729436307625,"log_rms_mae_db":0.9067840456866973,"rms_envelope_nrmse":0.08413839589013437,"envelope_correlation":0.9927348935976253,"rms_gain_error_db":-0.16750645406268838,"reference_mean_power":0.00737113406357093,"output_mean_power":0.007092243936338337},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.772697507541185e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010414087648688783,"output_power":0.0009951921934848091,"reference_active":true,"delta_db":-0.19714267877991137},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001272580988246754,"output_power":0.00012857654800572658,"reference_active":true,"delta_db":0.04476330945147672},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4550125163182702e-05,"output_power":2.7581309376842784e-05,"reference_active":true,"delta_db":0.5056116912284062}],"active_band_count":3,"active_band_mae_db":0.24917255981993144}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002933828481800726,"log_rms_mae_db":0.4853391649810217,"rms_envelope_nrmse":0.042828980755003174,"envelope_correlation":0.9983231510487769,"rms_gain_error_db":-0.16764610548249703,"reference_mean_power":0.007404046443755548,"output_mean_power":0.0071236819897319325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.623838097183376e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009238145152881927,"output_power":0.0008830897785116183,"reference_active":true,"delta_db":-0.19579923775033484},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0904498001922555e-05,"output_power":5.844598332432386e-05,"reference_active":true,"delta_db":-0.17894698107394666},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8155579307174762e-06,"output_power":2.06544283723877e-06,"reference_active":true,"delta_db":0.5600306912304027}],"active_band_count":3,"active_band_mae_db":0.3115923033515614}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027755394478816403,"log_rms_mae_db":1.5685562285301293,"rms_envelope_nrmse":0.13155829340603806,"envelope_correlation":0.9873112759865412,"rms_gain_error_db":-0.33901770646115875,"reference_mean_power":0.003209968762055962,"output_mean_power":0.0029689236921890505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3498130551666851e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006624323096451465,"output_power":0.0006265366207981193,"reference_active":true,"delta_db":-0.2419504628584326},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.187213693276335e-05,"output_power":5.941532460356997e-05,"reference_active":true,"delta_db":-0.8266208352963798},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2558954262897387e-05,"output_power":1.191176756498708e-05,"reference_active":true,"delta_db":-0.22977268417731164}],"active_band_count":3,"active_band_mae_db":0.4327813274440413}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025197665775423135,"log_rms_mae_db":0.9089995389649627,"rms_envelope_nrmse":0.07524060170936282,"envelope_correlation":0.9965101150318064,"rms_gain_error_db":-0.3394218427212834,"reference_mean_power":0.0032255924683410116,"output_mean_power":0.0029830965655301216},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.287517318914353e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006130037756394711,"output_power":0.0005768175267150123,"reference_active":true,"delta_db":-0.26424701416208884},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4809964114174026e-05,"output_power":2.755594059041821e-05,"reference_active":true,"delta_db":-1.0148833580902998},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.24341074108939e-07,"output_power":7.566447706728467e-07,"reference_active":true,"delta_db":-0.8694021689427218}],"active_band_count":3,"active_band_mae_db":0.7161775137317035}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017139073768520646,"log_rms_mae_db":1.6542265992813352,"rms_envelope_nrmse":0.1704584722232639,"envelope_correlation":0.9790038307812716,"rms_gain_error_db":-0.20181711008661107,"reference_mean_power":0.0008077036407054461,"output_mean_power":0.0007710283189277532},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.727013621661233e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024375238055045323,"output_power":0.00023393945803721128,"reference_active":true,"delta_db":-0.17845386374887473},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.980069758207981e-05,"output_power":2.9660178514874013e-05,"reference_active":true,"delta_db":-0.020526696915074204},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.347390229447954e-06,"output_power":8.100715888249296e-06,"reference_active":true,"delta_db":0.4239029427688113}],"active_band_count":3,"active_band_mae_db":0.20762783447758673}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001528919021429252,"log_rms_mae_db":1.0174818345377328,"rms_envelope_nrmse":0.08768518812884402,"envelope_correlation":0.9944724614493112,"rms_gain_error_db":-0.20141807212317653,"reference_mean_power":0.0008114918980720448,"output_mean_power":0.0007747157423292042},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4774424074646424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022848364185803236,"output_power":0.0002208159572957837,"reference_active":true,"delta_db":-0.14824657937653885},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508904456720678e-05,"output_power":1.295786185621204e-05,"reference_active":true,"delta_db":0.15314070237711563},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.746475959138819e-07,"output_power":6.038625777594085e-07,"reference_active":true,"delta_db":0.2153652244731119}],"active_band_count":3,"active_band_mae_db":0.17225083540892214}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":1.9831736377407172,"rms_envelope_nrmse":0.2519275433279852,"envelope_correlation":0.9574698231772957,"rms_gain_error_db":0.24523881504891565,"reference_mean_power":0.0001388420753372345,"output_mean_power":0.00014690784163833495},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.164366134779654e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.59650448810242e-05,"output_power":4.901644834888029e-05,"reference_active":true,"delta_db":0.27914151287982786},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.502002650645408e-05,"output_power":1.5649281541491342e-05,"reference_active":true,"delta_db":0.1782370479463687},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429741089402253e-06,"output_power":3.906520978857019e-06,"reference_active":true,"delta_db":2.0623016279572486}],"active_band_count":3,"active_band_mae_db":0.8398933962611483}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":1.587225995637388,"rms_envelope_nrmse":0.17728080500889165,"envelope_correlation":0.9776708375703893,"rms_gain_error_db":0.24511923136841782,"reference_mean_power":0.00013952517981136642,"output_mean_power":0.00014762656479063837},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0998934517742824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3887505804801683e-05,"output_power":4.856998496066444e-05,"reference_active":true,"delta_db":0.4402706908141593},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188432498267155e-06,"output_power":7.98542209434761e-06,"reference_active":true,"delta_db":0.45663678141584313},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0513859982119923e-07,"output_power":3.2446368981627816e-07,"reference_active":true,"delta_db":1.9911871579121578}],"active_band_count":3,"active_band_mae_db":0.9626982100473868}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.48962140083313,"predicted_mos":3.48962140083313,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"61814d8a6a118c76f47fa5cb9db0be1b7623c2960fd20494b3e21c5299f62f30","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0954092940010014},"started_at":"2026-10-13T04:43:42.903147+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3138000869948883e-05,"native_input_samples":96000,"native_input_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","output_sha256":"1025ca7ad75e487f9a2e42a0cbe1f0918fb382aee5463e3fa566b41457ff042f","output_samples":96000,"length_delta_samples":0,"analysis_s":1.6250385040002584,"synthesis_s":0.07897078600035456,"total_s":1.704009290000613,"rtf":0.2840015483334355,"output_peak":1.2815085810219413,"output_rms":0.09504863397737295,"finished_at":"2026-10-13T04:43:44.613020+00:00","evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"1025ca7ad75e487f9a2e42a0cbe1f0918fb382aee5463e3fa566b41457ff042f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.660290897600125,"spectral_convergence":0.673033414528214,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.129840556650823,"spectral_convergence":0.28117637590557115,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.641323096483302,"spectral_convergence":0.3031293584592337,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.0017608666383487374,"delta_db":0.2540151436698449,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.002251839210383455,"delta_db":0.7389350706661468,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.0037516360971979043,"delta_db":1.3845845114356155,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.0009303935625612676,"delta_db":0.8433497293367437,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00026881605159110623,"delta_db":0.35912837546904863,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":7.069126088214819e-05,"delta_db":0.43884450286883936,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6698095555743732,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8869581087862121,"pitch":null,"output_rms_dbfs":-20.44108240718636,"output_peak":1.2815085810219413,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.810484850244749},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.660290897600125,"spectral_convergence":0.673033414528214,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.129840556650823,"spectral_convergence":0.28117637590557115,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.641323096483302,"spectral_convergence":0.3031293584592337,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.0017608666383487374,"delta_db":0.2540151436698449,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.002251839210383455,"delta_db":0.7389350706661468,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.0037516360971979043,"delta_db":1.3845845114356155,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.0009303935625612676,"delta_db":0.8433497293367437,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00026881605159110623,"delta_db":0.35912837546904863,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":7.069126088214819e-05,"delta_db":0.43884450286883936,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.6698095555743732,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.8869581087862121,"pitch":null,"output_rms_dbfs":-20.44108240718636,"output_peak":1.2815085810219413,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.810484850244749},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9205848416190443,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08582187018202725,"rms_dbfs":-21.328040515972575,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.582274007876475,"mean":1.6474088033040366e-05,"zero_energy":false},"output":{"rms":0.09504863397737295,"rms_dbfs":-20.44108240718636,"peak":1.2815085810219413,"peak_dbfs":2.1544303715913844,"crest_factor_db":22.595512778777746,"mean":1.7356316616497456e-06,"zero_energy":false},"rms_difference_db":0.8869581087862137},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":17,"at_or_above_full_scale_fraction":0.00017708333333333335,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0001875,"longest_full_scale_run_samples":3,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":4.633965694872092e-06,"output_rms_dbfs":-53.340471856049334,"output_zero_energy":false,"output_peak":0.010784327216125043,"output_energy_fraction":5.1293348946949235e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":175,"both_estimated_voiced_frames":126,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.875,"absolute_difference_cents_median":4.863760651088702,"absolute_difference_cents_p95":19.692191573694576,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031196729436307625,"log_rms_mae_db":1.7944937932630838,"rms_envelope_nrmse":0.1775633744214608,"envelope_correlation":0.9815914479795098,"rms_gain_error_db":0.8871918196462836,"reference_mean_power":0.00737113406357093,"output_mean_power":0.009041770761077211},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.772697507541185e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010414087648688783,"output_power":0.0012445912345270132,"reference_active":true,"delta_db":0.7740550945422473},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001272580988246754,"output_power":0.00013746426879433508,"reference_active":true,"delta_db":0.33504395500564216},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4550125163182702e-05,"output_power":2.5887440407983023e-05,"reference_active":true,"delta_db":0.2303540155592748}],"active_band_count":3,"active_band_mae_db":0.44648435503572137}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002933828481800726,"log_rms_mae_db":1.3198559266355578,"rms_envelope_nrmse":0.13288163990623444,"envelope_correlation":0.9943149328513301,"rms_gain_error_db":0.8879134343076882,"reference_mean_power":0.007404046443755548,"output_mean_power":0.009083651789805131},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.623838097183376e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009238145152881927,"output_power":0.0011110635218493244,"reference_active":true,"delta_db":0.8015410743700808},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0904498001922555e-05,"output_power":6.156765580321348e-05,"reference_active":true,"delta_db":0.047032501605892285},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8155579307174762e-06,"output_power":1.7318952119578807e-06,"reference_active":true,"delta_db":-0.20488499354974}],"active_band_count":3,"active_band_mae_db":0.3511528565085711}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027755394478816403,"log_rms_mae_db":1.8149504000942662,"rms_envelope_nrmse":0.23295023657230954,"envelope_correlation":0.9813080741293141,"rms_gain_error_db":1.2828315210196326,"reference_mean_power":0.003209968762055962,"output_mean_power":0.004313044690219459},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3498130551666851e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006624323096451465,"output_power":0.0008897923647979758,"reference_active":true,"delta_db":1.2814716811389801},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.187213693276335e-05,"output_power":8.654785253795219e-05,"reference_active":true,"delta_db":0.80695738770228},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2558954262897387e-05,"output_power":1.1945758720435496e-05,"reference_active":true,"delta_db":-0.2173974017973282}],"active_band_count":3,"active_band_mae_db":0.7686088235461962}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025197665775423135,"log_rms_mae_db":1.2785302512443215,"rms_envelope_nrmse":0.19294281208066794,"envelope_correlation":0.992137366530912,"rms_gain_error_db":1.2829147331398332,"reference_mean_power":0.0032255924683410116,"output_mean_power":0.00433412038019152},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.287517318914353e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006130037756394711,"output_power":0.0008238090004874405,"reference_active":true,"delta_db":1.2836338305867394},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4809964114174026e-05,"output_power":3.960066405483524e-05,"reference_active":true,"delta_db":0.5599889300790655},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.24341074108939e-07,"output_power":8.752420000148072e-07,"reference_active":true,"delta_db":-0.23704101948757555}],"active_band_count":3,"active_band_mae_db":0.6935545933844601}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017139073768520646,"log_rms_mae_db":1.548593626316626,"rms_envelope_nrmse":0.1780887450196805,"envelope_correlation":0.9850006241868652,"rms_gain_error_db":0.7810023762143545,"reference_mean_power":0.0008077036407054461,"output_mean_power":0.0009668348093057461},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.727013621661233e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024375238055045323,"output_power":0.00029632785165003113,"reference_active":true,"delta_db":0.8482360661024851},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.980069758207981e-05,"output_power":3.76443225276199e-05,"reference_active":true,"delta_db":1.0147305542574383},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.347390229447954e-06,"output_power":7.27836526056969e-06,"reference_active":true,"delta_db":-0.04099259785899713}],"active_band_count":3,"active_band_mae_db":0.6346530727396402}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001528919021429252,"log_rms_mae_db":0.9976438677077589,"rms_envelope_nrmse":0.12130727453100106,"envelope_correlation":0.995954654441229,"rms_gain_error_db":0.7810971843199388,"reference_mean_power":0.0008114918980720448,"output_mean_power":0.0009713906225198929},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4774424074646424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022848364185803236,"output_power":0.0002784993037860906,"reference_active":true,"delta_db":0.8596900134862238},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508904456720678e-05,"output_power":1.577416887516073e-05,"reference_active":true,"delta_db":1.0072721060446603},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.746475959138819e-07,"output_power":6.039560013557343e-07,"reference_active":true,"delta_db":0.2160370696185054}],"active_band_count":3,"active_band_mae_db":0.6943330630497965}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":1.8144983304763092,"rms_envelope_nrmse":0.21442535672598237,"envelope_correlation":0.9696866860944731,"rms_gain_error_db":0.29820498256410305,"reference_mean_power":0.0001388420753372345,"output_mean_power":0.0001487104866723869},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.164366134779654e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.59650448810242e-05,"output_power":5.267767537165797e-05,"reference_active":true,"delta_db":0.5919891310085958},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.502002650645408e-05,"output_power":1.6373958543940714e-05,"reference_active":true,"delta_db":0.37482987409637747},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429741089402253e-06,"output_power":3.5798389137504407e-06,"reference_active":true,"delta_db":1.6830348650979763}],"active_band_count":3,"active_band_mae_db":0.8832846234009831}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":1.0682921227781206,"rms_envelope_nrmse":0.12240725143918933,"envelope_correlation":0.9900201446601731,"rms_gain_error_db":0.2981030539582307,"reference_mean_power":0.00013952517981136642,"output_mean_power":0.00014943863647898323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0998934517742824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3887505804801683e-05,"output_power":5.09860473984748e-05,"reference_active":true,"delta_db":0.6511044545908061},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188432498267155e-06,"output_power":8.534569598965152e-06,"reference_active":true,"delta_db":0.7454742546870895},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0513859982119923e-07,"output_power":2.5288679444821077e-07,"reference_active":true,"delta_db":0.9087876451415202}],"active_band_count":3,"active_band_mae_db":0.7684554514731387}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"facodec","status":"measured","mos":{"status":"measured","value":3.604623317718506,"predicted_mos":3.604623317718506,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"9f40803130417cc8362221a236513e1f997efeef82256d51dd57b94cefea8b50","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4664438029976736},"started_at":"2026-10-13T07:01:38.766512+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.9099000282003544e-05,"native_input_samples":96000,"native_input_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","output_sha256":"b27b6babcca496755548a8022b69570c004ff56a924d5af4b56516f74eb9a5c4","output_samples":96000,"length_delta_samples":0,"analysis_s":1.7490553839998029,"synthesis_s":13.384035703000336,"total_s":15.133091087000139,"rtf":2.5221818478333566,"output_peak":0.6875807642936707,"output_rms":0.08075308698781196,"finished_at":"2026-10-13T07:01:53.912768+00:00","evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"b27b6babcca496755548a8022b69570c004ff56a924d5af4b56516f74eb9a5c4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.468897110327321,"spectral_convergence":0.26643435179084746,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.813610609480325,"spectral_convergence":0.2724455754046809,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.01558439932311,"spectral_convergence":0.2898055507359233,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.001505363488019553,"delta_db":-0.42683572767626476,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.0018878927771190058,"delta_db":-0.026665450886888917,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.002269748466868162,"delta_db":-0.7978452330409505,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.000592509018869131,"delta_db":-1.116367585514207,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00020574421883756008,"delta_db":-0.802147193684995,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":5.980308834771437e-05,"delta_db":-0.2875766534283337,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5762396407052733,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5287768178600842,"pitch":null,"output_rms_dbfs":-21.85681733383266,"output_peak":0.6875807642936707,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.766030706376919},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.468897110327321,"spectral_convergence":0.26643435179084746,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.813610609480325,"spectral_convergence":0.2724455754046809,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.01558439932311,"spectral_convergence":0.2898055507359233,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.001505363488019553,"delta_db":-0.42683572767626476,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.0018878927771190058,"delta_db":-0.026665450886888917,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.002269748466868162,"delta_db":-0.7978452330409505,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.000592509018869131,"delta_db":-1.116367585514207,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00020574421883756008,"delta_db":-0.802147193684995,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":5.980308834771437e-05,"delta_db":-0.2875766534283337,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5762396407052733,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5287768178600842,"pitch":null,"output_rms_dbfs":-21.85681733383266,"output_peak":0.6875807642936707,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.766030706376919},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8897273084927358,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08582187018202725,"rms_dbfs":-21.328040515972575,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.582274007876475,"mean":1.6474088033040366e-05,"zero_energy":false},"output":{"rms":0.08075308698781196,"rms_dbfs":-21.85681733383266,"peak":0.6875807642936707,"peak_dbfs":-3.2535256320780674,"crest_factor_db":18.603291701754593,"mean":-1.0435357453710973e-05,"zero_energy":false},"rms_difference_db":-0.5287768178600842},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":5.3684724615748706e-06,"output_rms_dbfs":-52.701492703305554,"output_zero_energy":false,"output_peak":0.011856643483042717,"output_energy_fraction":8.232513717899539e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":153,"both_estimated_voiced_frames":135,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.9375,"absolute_difference_cents_median":2.7676652169014493,"absolute_difference_cents_p95":15.869060651422238,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031196729436307625,"log_rms_mae_db":1.166259386391118,"rms_envelope_nrmse":0.11452944216433865,"envelope_correlation":0.9894212545130724,"rms_gain_error_db":-0.5287304198296751,"reference_mean_power":0.00737113406357093,"output_mean_power":0.0065262133616301125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.772697507541185e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010414087648688783,"output_power":0.001002321484016644,"reference_active":true,"delta_db":-0.1661418929559309},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001272580988246754,"output_power":0.0001288318031771372,"reference_active":true,"delta_db":0.05337654583233791},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4550125163182702e-05,"output_power":2.7536700501635832e-05,"reference_active":true,"delta_db":0.4985819045510573}],"active_band_count":3,"active_band_mae_db":0.2393667811131087}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002933828481800726,"log_rms_mae_db":0.9255372861103303,"rms_envelope_nrmse":0.10478289902245412,"envelope_correlation":0.9913793218335881,"rms_gain_error_db":-0.5283068034321984,"reference_mean_power":0.007404046443755548,"output_mean_power":0.006555992589045708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.623838097183376e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009238145152881927,"output_power":0.000895256968610518,"reference_active":true,"delta_db":-0.13637071535379697},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0904498001922555e-05,"output_power":6.345438876777297e-05,"reference_active":true,"delta_db":0.1781229714596677},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8155579307174762e-06,"output_power":1.8104680019890502e-06,"reference_active":true,"delta_db":-0.012192573649283391}],"active_band_count":3,"active_band_mae_db":0.10889542015424936}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027755394478816403,"log_rms_mae_db":1.6705375216630214,"rms_envelope_nrmse":0.15116288354661103,"envelope_correlation":0.9857089254711504,"rms_gain_error_db":-0.6513218230453655,"reference_mean_power":0.003209968762055962,"output_mean_power":0.00276292199514007},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3498130551666851e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006624323096451465,"output_power":0.0006002695791666286,"reference_active":true,"delta_db":-0.4279517221610507},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.187213693276335e-05,"output_power":7.707365583865984e-05,"reference_active":true,"delta_db":0.30345401711085024},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2558954262897387e-05,"output_power":1.4288019147147309e-05,"reference_active":true,"delta_db":0.5601854470493196}],"active_band_count":3,"active_band_mae_db":0.4305303954404069}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025197665775423135,"log_rms_mae_db":1.2089668239621871,"rms_envelope_nrmse":0.1359336723146389,"envelope_correlation":0.988951383230995,"rms_gain_error_db":-0.6514319470270481,"reference_mean_power":0.0032255924683410116,"output_mean_power":0.0027762994155505405},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.287517318914353e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006130037756394711,"output_power":0.000555077971204305,"reference_active":true,"delta_db":-0.4310915716660679},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4809964114174026e-05,"output_power":3.7999137606160604e-05,"reference_active":true,"delta_db":0.3807016480897893},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.24341074108939e-07,"output_power":1.0240762910508617e-06,"reference_active":true,"delta_db":0.4450005985058694}],"active_band_count":3,"active_band_mae_db":0.4189312727539089}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017139073768520646,"log_rms_mae_db":1.6406134558738732,"rms_envelope_nrmse":0.16835590539364328,"envelope_correlation":0.9858875039413506,"rms_gain_error_db":-0.9429050956229966,"reference_mean_power":0.0008077036407054461,"output_mean_power":0.0006500721054548716},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.727013621661233e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024375238055045323,"output_power":0.0002126477666282796,"reference_active":true,"delta_db":-0.5928804000901677},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.980069758207981e-05,"output_power":2.7396571611633315e-05,"reference_active":true,"delta_db":-0.36530211363542003},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.347390229447954e-06,"output_power":8.065633069935937e-06,"reference_active":true,"delta_db":0.40505354348829703}],"active_band_count":3,"active_band_mae_db":0.45441201907129497}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001528919021429252,"log_rms_mae_db":1.285466467584992,"rms_envelope_nrmse":0.14877534202612255,"envelope_correlation":0.9906173658477172,"rms_gain_error_db":-0.9432146297659034,"reference_mean_power":0.0008114918980720448,"output_mean_power":0.0006530744979260919},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4774424074646424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022848364185803236,"output_power":0.00019871768299864858,"reference_active":true,"delta_db":-0.6061859772514395},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508904456720678e-05,"output_power":1.2036009063132079e-05,"reference_active":true,"delta_db":-0.16736769289510176},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.746475959138819e-07,"output_power":5.771877247097641e-07,"reference_active":true,"delta_db":0.019154920637797265}],"active_band_count":3,"active_band_mae_db":0.26423619692811284}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":1.8727225540501438,"rms_envelope_nrmse":0.21974702153631007,"envelope_correlation":0.9669674143260003,"rms_gain_error_db":-0.5773093116756994,"reference_mean_power":0.0001388420753372345,"output_mean_power":0.00012155985279236851},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.164366134779654e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.59650448810242e-05,"output_power":4.316739000347744e-05,"reference_active":true,"delta_db":-0.27271897466577316},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.502002650645408e-05,"output_power":1.2834276580945742e-05,"reference_active":true,"delta_db":-0.6829930490659228},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429741089402253e-06,"output_power":2.6486676876269915e-06,"reference_active":true,"delta_db":0.3746747529285098}],"active_band_count":3,"active_band_mae_db":0.44346225888673524}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":1.499900898977881,"rms_envelope_nrmse":0.17976611243340346,"envelope_correlation":0.9773740876906302,"rms_gain_error_db":-0.5778220355279006,"reference_mean_power":0.00013952517981136642,"output_mean_power":0.00012214350755415863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0998934517742824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3887505804801683e-05,"output_power":4.1253883556400807e-05,"reference_active":true,"delta_db":-0.2687606150901017},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188432498267155e-06,"output_power":6.299956948097453e-06,"reference_active":true,"delta_db":-0.572966172933435},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0513859982119923e-07,"output_power":1.9081100401118522e-07,"reference_active":true,"delta_db":-0.31443970184345504}],"active_band_count":3,"active_band_mae_db":0.38538882995566387}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.320286273956299,"predicted_mos":3.320286273956299,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"730ee212ed3aaee28dde598e4cbd5546e70d3d038ea4719a57046e454e0ad9d7","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.213534088998131},"started_at":"2026-10-13T07:31:57.556294+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.008278789999167202,"native_input_samples":264600,"native_input_sha256":"220015306695d45bf0399c98ea92729ae0ccb1efdc6aa54399c1c68e21563b15","output_sha256":"1fd31e9f643603e36b5d7d6a269def279dc6d765367bfb15d0c61a56edba7518","output_samples":264704,"length_delta_samples":104,"analysis_s":0.17318720600087545,"synthesis_s":39.68830784799866,"total_s":39.86149505399953,"rtf":6.6435825089999225,"output_peak":0.671639084815979,"output_rms":0.07836038366562179,"finished_at":"2026-10-13T07:32:37.450375+00:00","evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"1fd31e9f643603e36b5d7d6a269def279dc6d765367bfb15d0c61a56edba7518","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.114087649579005,"output_peak":0.6675915452314967},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.127068804956588,"spectral_convergence":0.45004046338430903,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.94752994802893,"spectral_convergence":0.41316293024742634,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.752339925243557,"spectral_convergence":0.45014863325717286,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.0014366262872733692,"delta_db":-0.6298114210140299,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.0017879121576248325,"delta_db":-0.2629769244418799,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.0020947480996338277,"delta_db":-1.1463044962586326,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.0006692084495318444,"delta_db":-0.5877030809640507,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00012750592297483998,"delta_db":-2.880120009539975,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":3.235819431501826e-05,"delta_db":-2.954969990489931,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.4103143204514168,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7843672270376076,"pitch":null,"output_rms_dbfs":-22.11240774301018,"output_peak":0.6675915452314967,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.942312892743024},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7071388349509571,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08582187018202725,"rms_dbfs":-21.328040515972575,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.582274007876475,"mean":1.6474088033040366e-05,"zero_energy":false},"output":{"rms":0.07841147308496209,"rms_dbfs":-22.11240774301018,"peak":0.6675915452314967,"peak_dbfs":-3.509783442329094,"crest_factor_db":18.602624300681086,"mean":-0.00042096653066435983,"zero_energy":false},"rms_difference_db":-0.7843672270376061},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":3.7150724000485406e-06,"output_rms_dbfs":-54.30032718206182,"output_zero_energy":false,"output_peak":0.006925248693627434,"output_energy_fraction":6.0423803046705306e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":137,"both_estimated_voiced_frames":108,"reference_voiced_output_unvoiced_frames":36,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.75,"absolute_difference_cents_median":8.478992393146168,"absolute_difference_cents_p95":65.69915535466602,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031196729436307625,"log_rms_mae_db":1.983067742402839,"rms_envelope_nrmse":0.14190111805929312,"envelope_correlation":0.9871900324139973,"rms_gain_error_db":-0.7845194282026822,"reference_mean_power":0.00737113406357093,"output_mean_power":0.006152935567382765},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.772697507541185e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010414087648688783,"output_power":0.0009757623348306634,"reference_active":true,"delta_db":-0.28277178530845404},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001272580988246754,"output_power":0.00010868385861233188,"reference_active":true,"delta_db":-0.685203819824469},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4550125163182702e-05,"output_power":1.7432791331860826e-05,"reference_active":true,"delta_db":-1.4868677887172586}],"active_band_count":3,"active_band_mae_db":0.8182811312833939}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002933828481800726,"log_rms_mae_db":1.7347001754694842,"rms_envelope_nrmse":0.12265100970054328,"envelope_correlation":0.9924427351440576,"rms_gain_error_db":-0.7843707944816026,"reference_mean_power":0.007404046443755548,"output_mean_power":0.006180620170255197},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.623838097183376e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009238145152881927,"output_power":0.000878285671593188,"reference_active":true,"delta_db":-0.2194898404403858},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0904498001922555e-05,"output_power":5.121900229054039e-05,"reference_active":true,"delta_db":-0.7521825345527275},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8155579307174762e-06,"output_power":1.2703165122947127e-06,"reference_active":true,"delta_db":-1.55098167581743}],"active_band_count":3,"active_band_mae_db":0.8408846836035145}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027755394478816403,"log_rms_mae_db":2.857646210526804,"rms_envelope_nrmse":0.19882417648625636,"envelope_correlation":0.9708786715083428,"rms_gain_error_db":-0.560343760286031,"reference_mean_power":0.003209968762055962,"output_mean_power":0.002821411486086343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3498130551666851e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006624323096451465,"output_power":0.0007266502814555411,"reference_active":true,"delta_db":0.4018393941962953},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.187213693276335e-05,"output_power":6.865552989871024e-05,"reference_active":true,"delta_db":-0.19885034198168472},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2558954262897387e-05,"output_power":1.1540709251227676e-05,"reference_active":true,"delta_db":-0.36720978986283404}],"active_band_count":3,"active_band_mae_db":0.322633175346938}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025197665775423135,"log_rms_mae_db":2.1867544480655434,"rms_envelope_nrmse":0.17121724099680655,"envelope_correlation":0.9780732215109006,"rms_gain_error_db":-0.5609987964460204,"reference_mean_power":0.0032255924683410116,"output_mean_power":0.0028347164028276347},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.287517318914353e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006130037756394711,"output_power":0.0006677780059977743,"reference_active":true,"delta_db":0.37168961535960043},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4809964114174026e-05,"output_power":3.2601478709105805e-05,"reference_active":true,"delta_db":-0.28466276706819926},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.24341074108939e-07,"output_power":9.124801621699437e-07,"reference_active":true,"delta_db":-0.05608820391045457}],"active_band_count":3,"active_band_mae_db":0.23748019544608476}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017139073768520646,"log_rms_mae_db":3.5341729784033027,"rms_envelope_nrmse":0.3088771259022052,"envelope_correlation":0.932113708209844,"rms_gain_error_db":-0.9216411556632326,"reference_mean_power":0.0008077036407054461,"output_mean_power":0.0006532627952824466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.727013621661233e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024375238055045323,"output_power":0.00017522918431865997,"reference_active":true,"delta_db":-1.4334242647717679},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.980069758207981e-05,"output_power":2.450536950982006e-05,"reference_active":true,"delta_db":-0.8496517476088311},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.347390229447954e-06,"output_power":4.101664913327436e-06,"reference_active":true,"delta_db":-2.5317292873024044}],"active_band_count":3,"active_band_mae_db":1.6049350998943346}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001528919021429252,"log_rms_mae_db":3.0590542163420826,"rms_envelope_nrmse":0.28187773083718093,"envelope_correlation":0.9418345770567176,"rms_gain_error_db":-0.9210006980434544,"reference_mean_power":0.0008114918980720448,"output_mean_power":0.0006564234969163699},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4774424074646424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022848364185803236,"output_power":0.0001640607552871208,"reference_active":true,"delta_db":-1.4385040589713693},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508904456720678e-05,"output_power":1.1706958969126882e-05,"reference_active":true,"delta_db":-0.28775179152319186},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.746475959138819e-07,"output_power":3.074658864597714e-07,"reference_active":true,"delta_db":-2.7160465660650694}],"active_band_count":3,"active_band_mae_db":1.4807674721865436}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":4.3367032813026185,"rms_envelope_nrmse":0.38904817215913223,"envelope_correlation":0.9230813990902449,"rms_gain_error_db":-2.6946632532895745,"reference_mean_power":0.0001388420753372345,"output_mean_power":7.465429031999625e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.164366134779654e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.59650448810242e-05,"output_power":2.7302112487474316e-05,"reference_active":true,"delta_db":-2.262314368432294},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.502002650645408e-05,"output_power":9.75108467518782e-06,"reference_active":true,"delta_db":-1.8761777136208067},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429741089402253e-06,"output_power":1.3143132187141902e-06,"reference_active":true,"delta_db":-2.668611222712873}],"active_band_count":3,"active_band_mae_db":2.2690344349219913}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":3.7872333711745316,"rms_envelope_nrmse":0.35741279831707634,"envelope_correlation":0.9416961924365744,"rms_gain_error_db":-2.6946326125787468,"reference_mean_power":0.00013952517981136642,"output_mean_power":7.502211951579062e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0998934517742824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3887505804801683e-05,"output_power":2.7002609765943914e-05,"reference_active":true,"delta_db":-2.1093515975008486},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188432498267155e-06,"output_power":4.762049315995691e-06,"reference_active":true,"delta_db":-1.7884031026144425},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0513859982119923e-07,"output_power":1.0485437674526004e-07,"reference_active":true,"delta_db":-2.9146082375895306}],"active_band_count":3,"active_band_mae_db":2.270787645901607}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"golf","status":"measured","mos":{"status":"measured","value":3.2465476989746094,"predicted_mos":3.2465476989746094,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"c12d72c9aa0d588c4d92123c3a1b343edadec99549ae3ce241373376eb690626","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0235454899993783},"started_at":"2026-10-13T05:03:08.276384+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035595749995991355,"native_input_samples":144000,"native_input_sha256":"ec69e9de668d1340af2e55fa2adae7e7852f7d1420298fe63d06ba4efe4e1f63","output_sha256":"a3ae57a19c3210fd021101bb600c1e60215c1d290b698626eee46552dac28a88","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6650555010000971,"synthesis_s":0.19336477799970453,"total_s":0.8584202789998017,"rtf":0.14307004649996694,"output_peak":0.8593546748161316,"output_rms":0.08082238501411246,"finished_at":"2026-10-13T05:03:09.151878+00:00","analysis_stages":{"external_f0_s":0.09350036599971645},"evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"a3ae57a19c3210fd021101bb600c1e60215c1d290b698626eee46552dac28a88","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.847686503709426,"output_peak":0.8602614980484192},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.027785333889252,"spectral_convergence":0.6653381534223005,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.391595666554267,"spectral_convergence":0.31878686320729965,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.668983696908867,"spectral_convergence":0.3511595918816473,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016621065209557941,"output_power":0.0015741821663251168,"delta_db":-0.23603865340912095,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001902700848556098,"output_power":0.0019167762819973179,"delta_db":0.03200914961498303,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0027320234410405536,"output_power":0.002296538272426442,"delta_db":-0.7541073386824699,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007674053283724871,"output_power":0.0005010957816261812,"delta_db":-1.8510406359895955,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0002478897618313943,"output_power":0.00018969690706423275,"delta_db":-1.16198340272571,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.400594714522244e-05,"output_power":5.6496284769094335e-05,"delta_db":-0.54200439269401,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7628639288526483,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5259732147295831,"pitch":null,"output_rms_dbfs":-21.847686503709426,"output_peak":0.8602614980484192,"reference_rms_dbfs":-21.32171328897984,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.029454899117462},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.863150067104282,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08588440980702813,"rms_dbfs":-21.32171328897984,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.57594678088374,"mean":9.982534163384287e-07,"zero_energy":false},"output":{"rms":0.08083802134026034,"rms_dbfs":-21.847686503709426,"peak":0.8602614980484192,"peak_dbfs":-1.307390279851549,"crest_factor_db":20.540296223857876,"mean":0.000983864950473924,"zero_energy":false},"rms_difference_db":-0.5259732147295857},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":5.041383185376877e-06,"output_rms_dbfs":-52.9745029145887,"output_zero_energy":false,"output_peak":0.009221633286842383,"output_energy_fraction":7.727567137178241e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":200,"both_estimated_voiced_frames":139,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":61,"paired_voiced_fraction":0.9652777777777778,"absolute_difference_cents_median":6.989017676313869,"absolute_difference_cents_p95":24.924532642707245,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00031196729436307625,"log_rms_mae_db":1.5759279542430742,"rms_envelope_nrmse":0.13804419045179542,"envelope_correlation":0.9825115522755417,"rms_gain_error_db":-0.5256640443942988,"reference_mean_power":0.007381401879575031,"output_mean_power":0.0065399201629877625},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.7754107328633566e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010403915419402035,"output_power":0.0008985949522601054,"reference_active":true,"delta_db":-0.6363283818778963},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012705508861081277,"output_power":0.0001293899846168633,"reference_active":true,"delta_db":0.07908597817428614},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4506483914390247e-05,"output_power":2.374666917743593e-05,"reference_active":true,"delta_db":-0.1367830301479624}],"active_band_count":3,"active_band_mae_db":0.28406579673338167}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0002933828481800726,"log_rms_mae_db":1.0471454684808095,"rms_envelope_nrmse":0.08836898995393257,"envelope_correlation":0.9947460342223303,"rms_gain_error_db":-0.5254044999319383,"reference_mean_power":0.007412900254006227,"output_mean_power":0.006568220234973123},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.6263330419653008e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009232402833092753,"output_power":0.0007949627568476786,"reference_active":true,"delta_db":-0.6496796268955456},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0313229107986296e-05,"output_power":6.294907585802918e-05,"reference_active":true,"delta_db":0.18576777934553124},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8080941638903686e-06,"output_power":1.940241382404426e-06,"reference_active":true,"delta_db":0.3063471880284044}],"active_band_count":3,"active_band_mae_db":0.38059819808982703}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00027755394478816403,"log_rms_mae_db":1.8163450029701664,"rms_envelope_nrmse":0.15760971710258997,"envelope_correlation":0.9815975175959843,"rms_gain_error_db":-0.3971770603565628,"reference_mean_power":0.0032152590225572497,"output_mean_power":0.0029342577449406847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.3503712882849838e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006610764552259905,"output_power":0.0005953011004136462,"reference_active":true,"delta_db":-0.4551500434074501},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.189672855744185e-05,"output_power":7.564018891540827e-05,"reference_active":true,"delta_db":0.22043475268654772},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.25845410071126e-05,"output_power":1.2507884752495585e-05,"reference_active":true,"delta_db":-0.026535092517446512}],"active_band_count":3,"active_band_mae_db":0.23403996287048145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00025197665775423135,"log_rms_mae_db":1.091291646464285,"rms_envelope_nrmse":0.09327293396763262,"envelope_correlation":0.9942929815048496,"rms_gain_error_db":-0.3969004939103607,"reference_mean_power":0.0032301743375202597,"output_mean_power":0.0029480572502659868},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.2877123186083307e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006118500927774464,"output_power":0.0005485356278675578,"reference_active":true,"delta_db":-0.4744018959441364},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.457112687325986e-05,"output_power":3.7980538246985135e-05,"reference_active":true,"delta_db":0.40847579169321757},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.237519361202438e-07,"output_power":9.258959551379801e-07,"reference_active":true,"delta_db":0.010068251908192138}],"active_band_count":3,"active_band_mae_db":0.297648646515182}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017139073768520646,"log_rms_mae_db":2.1509905666243547,"rms_envelope_nrmse":0.24655024943403825,"envelope_correlation":0.9739579240718986,"rms_gain_error_db":-1.6813368873592132,"reference_mean_power":0.0008089776844047222,"output_mean_power":0.0005492914677036063},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.733563162358671e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024373860945454444,"output_power":0.0001761284549550843,"reference_active":true,"delta_db":-1.410948037537354},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.991588879104226e-05,"output_power":1.6049340586364855e-05,"reference_active":true,"delta_db":-2.704447167112333},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.373253590611891e-06,"output_power":5.773856089334719e-06,"reference_active":true,"delta_db":-1.0619321589307034}],"active_band_count":3,"active_band_mae_db":1.7257757878601303}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.0001528919021429252,"log_rms_mae_db":1.711696523713607,"rms_envelope_nrmse":0.20023847847008813,"envelope_correlation":0.9916400192055566,"rms_gain_error_db":-1.6808969497148492,"reference_mean_power":0.0008126260604625569,"output_mean_power":0.0005518245919549023},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.483084830870638e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022850112943947353,"output_power":0.00016434356062494215,"reference_active":true,"delta_db":-1.4313565899862024},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508102777185679e-05,"output_power":7.182217708847529e-06,"reference_active":true,"delta_db":-2.4093287570506288},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.740056104104209e-07,"output_power":3.723367863178718e-07,"reference_active":true,"delta_db":-1.879801912894878}],"active_band_count":3,"active_band_mae_db":1.9068290866439028}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":2.3879657878808764,"rms_envelope_nrmse":0.22526029548774817,"envelope_correlation":0.966418572497012,"rms_gain_error_db":-0.5141503622759238,"reference_mean_power":0.00013907137411945498,"output_mean_power":0.00012354429901676467},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1617864330402146e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5849978492799476e-05,"output_power":4.595227472929202e-05,"reference_active":true,"delta_db":0.009678784094492889},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4937671024255798e-05,"output_power":1.4758569557310582e-05,"reference_active":true,"delta_db":-0.0523862417544569},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4155128431643715e-06,"output_power":2.651218948198851e-06,"reference_active":true,"delta_db":0.40436244027373064}],"active_band_count":3,"active_band_mae_db":0.15547582204089347}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":1.997238786392807,"rms_envelope_nrmse":0.15478211553257074,"envelope_correlation":0.9849536022722171,"rms_gain_error_db":-0.5140248957881601,"reference_mean_power":0.00013974109062057894,"output_mean_power":0.0001241428292452456},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0976131366875603e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3750477705123606e-05,"output_power":4.491168987207633e-05,"reference_active":true,"delta_db":0.11376596991753087},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.14691427015292e-06,"output_power":7.3380412453497716e-06,"reference_active":true,"delta_db":0.11461575879245184},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0394932066565738e-07,"output_power":2.461849784367673e-07,"reference_active":true,"delta_db":0.8173928682580073}],"active_band_count":3,"active_band_mae_db":0.3485915323226633}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.099731683731079,"predicted_mos":3.099731683731079,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"aecdccf4f3b04e9a8e0f543a9aaf9d4e48f9e5675a88e643872501b284767204","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.474932028002513},"started_at":"2026-10-13T07:47:05.338535+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004572612000629306,"native_input_samples":144000,"native_input_sha256":"ec69e9de668d1340af2e55fa2adae7e7852f7d1420298fe63d06ba4efe4e1f63","output_sha256":"3589b48d4e7f035d42fa37afd36a224fcb31e6489ed75d81dd834cc01672c83e","output_samples":144120,"length_delta_samples":120,"analysis_s":2.714362261998758,"synthesis_s":3.911550361000991,"total_s":6.625912622999749,"rtf":1.1043187704999582,"output_peak":0.800434947013855,"output_rms":0.08512922884313773,"finished_at":"2026-10-13T07:47:11.990741+00:00","analysis_stages":{"external_f0_s":1.7020587250008248},"evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"3589b48d4e7f035d42fa37afd36a224fcb31e6489ed75d81dd834cc01672c83e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.395802810428428,"output_peak":0.7815328079137887},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.5814721830357765,"spectral_convergence":0.6282732039674271,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.221037958593958,"spectral_convergence":0.30460891179190863,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4561599285773985,"spectral_convergence":0.309912285547599,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.0022070171417154882,"delta_db":1.234807553460425,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.0017554232350788558,"delta_db":-0.3426202753975761,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.0023608050541909014,"delta_db":-0.6270212846104016,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.0006747794945499487,"delta_db":-0.5516984727677914,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00022175374786146507,"delta_db":-0.4767139169773361,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":3.681019945840019e-05,"delta_db":-2.39513107016269,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9379987622293701,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.06463429883451598,"pitch":null,"output_rms_dbfs":-21.392674814807094,"output_peak":0.7815328079137887,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.7528900234023785},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.887747559498439,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08582187018202725,"rms_dbfs":-21.328040515972575,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.582274007876475,"mean":1.6474088033040366e-05,"zero_energy":false},"output":{"rms":0.08518561423653093,"rms_dbfs":-21.392674814807094,"peak":0.7815328079137887,"peak_dbfs":-2.141055720858705,"crest_factor_db":19.251619093948385,"mean":-0.00017078124012589628,"zero_energy":false},"rms_difference_db":-0.06463429883451965},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":3.615854126239751e-06,"output_rms_dbfs":-54.41789098508349,"output_zero_energy":false,"output_peak":0.01510873343099412,"output_energy_fraction":4.9828565316214696e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":188,"both_estimated_voiced_frames":137,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":51,"paired_voiced_fraction":0.9513888888888888,"absolute_difference_cents_median":9.115352780464976,"absolute_difference_cents_p95":25.563785215971023,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031196729436307625,"log_rms_mae_db":2.12362227978267,"rms_envelope_nrmse":0.1425128744206664,"envelope_correlation":0.9782478784173344,"rms_gain_error_db":-0.06465275023032534,"reference_mean_power":0.00737113406357093,"output_mean_power":0.007262213877295047},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.772697507541185e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010414087648688783,"output_power":0.0010481711996583203,"reference_active":true,"delta_db":0.028109939952396124},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001272580988246754,"output_power":0.0001195661316553237,"reference_active":true,"delta_db":-0.27077252132254825},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4550125163182702e-05,"output_power":2.1292265594729527e-05,"reference_active":true,"delta_db":-0.6183183579338326}],"active_band_count":3,"active_band_mae_db":0.3057336064029257}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002933828481800726,"log_rms_mae_db":1.6057041118561466,"rms_envelope_nrmse":0.08814650407599638,"envelope_correlation":0.9911486002601452,"rms_gain_error_db":-0.06490601790617376,"reference_mean_power":0.007404046443755548,"output_mean_power":0.0072942145355224764},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.623838097183376e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009238145152881927,"output_power":0.0009274689500663019,"reference_active":true,"delta_db":0.017145974128636183},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0904498001922555e-05,"output_power":5.7173226845183984e-05,"reference_active":true,"delta_db":-0.27456663527400915},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8155579307174762e-06,"output_power":1.6550439465122567e-06,"reference_active":true,"delta_db":-0.40200580792789603}],"active_band_count":3,"active_band_mae_db":0.23123947244351375}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027755394478816403,"log_rms_mae_db":2.221876795341187,"rms_envelope_nrmse":0.16930477003356842,"envelope_correlation":0.9797422807073689,"rms_gain_error_db":-0.572567236570901,"reference_mean_power":0.003209968762055962,"output_mean_power":0.0028134816205621946},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3498130551666851e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006624323096451465,"output_power":0.0005803780833593962,"reference_active":true,"delta_db":-0.5743050273978633},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.187213693276335e-05,"output_power":6.433146529320256e-05,"reference_active":true,"delta_db":-0.48137114133349335},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2558954262897387e-05,"output_power":1.0867577412719227e-05,"reference_active":true,"delta_db":-0.628207363590713}],"active_band_count":3,"active_band_mae_db":0.5612945107740231}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025197665775423135,"log_rms_mae_db":1.4617699563540443,"rms_envelope_nrmse":0.11331120493923,"envelope_correlation":0.992374904893788,"rms_gain_error_db":-0.5724728149037301,"reference_mean_power":0.0032255924683410116,"output_mean_power":0.0028272369939588014},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.287517318914353e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006130037756394711,"output_power":0.0005353299222861044,"reference_active":true,"delta_db":-0.5884163048383122},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4809964114174026e-05,"output_power":3.123581168677161e-05,"reference_active":true,"delta_db":-0.470507796600935},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.24341074108939e-07,"output_power":8.341091056599525e-07,"reference_active":true,"delta_db":-0.44609389564066454}],"active_band_count":3,"active_band_mae_db":0.5016726656933039}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017139073768520646,"log_rms_mae_db":2.1907557328037406,"rms_envelope_nrmse":0.2022412691899706,"envelope_correlation":0.9713512146401182,"rms_gain_error_db":-0.5137092891369284,"reference_mean_power":0.0008077036407054461,"output_mean_power":0.0007175978214132848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.727013621661233e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024375238055045323,"output_power":0.00023530906378893104,"reference_active":true,"delta_db":-0.15310209888430934},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.980069758207981e-05,"output_power":2.452227560357035e-05,"reference_active":true,"delta_db":-0.8466566114972265},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.347390229447954e-06,"output_power":5.989101829486229e-06,"reference_active":true,"delta_db":-0.887714091783899}],"active_band_count":3,"active_band_mae_db":0.6291576007218116}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001528919021429252,"log_rms_mae_db":1.3979045818401559,"rms_envelope_nrmse":0.12456542351394319,"envelope_correlation":0.9902010553735547,"rms_gain_error_db":-0.5133402517678772,"reference_mean_power":0.0008114918980720448,"output_mean_power":0.000721024734035725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4774424074646424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022848364185803236,"output_power":0.00022036039753733734,"reference_active":true,"delta_db":-0.1572156529645662},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508904456720678e-05,"output_power":1.0277808266033442e-05,"reference_active":true,"delta_db":-0.8531876378229241},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.746475959138819e-07,"output_power":4.796989646752081e-07,"reference_active":true,"delta_db":-0.7843281299767896}],"active_band_count":3,"active_band_mae_db":0.5982438069214266}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":2.1777358576200165,"rms_envelope_nrmse":0.26068767819115235,"envelope_correlation":0.9532552548137346,"rms_gain_error_db":-0.7579359308982463,"reference_mean_power":0.0001388420753372345,"output_mean_power":0.00011660777363855688},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.164366134779654e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.59650448810242e-05,"output_power":3.7026864745677965e-05,"reference_active":true,"delta_db":-0.9391074883753893},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.502002650645408e-05,"output_power":1.1037208141116834e-05,"reference_active":true,"delta_db":-1.3381146647579123},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429741089402253e-06,"output_power":2.0240983167170257e-06,"reference_active":true,"delta_db":-0.7932839440418391}],"active_band_count":3,"active_band_mae_db":1.0235020323917137}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":1.3759257834709722,"rms_envelope_nrmse":0.16501801294619725,"envelope_correlation":0.9834463853087386,"rms_gain_error_db":-0.7579641322142425,"reference_mean_power":0.00013952517981136642,"output_mean_power":0.00011718072418757067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0998934517742824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3887505804801683e-05,"output_power":3.50924085702837e-05,"reference_active":true,"delta_db":-0.9712772284050828},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188432498267155e-06,"output_power":5.220623610400142e-06,"reference_active":true,"delta_db":-1.3891181577508753},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0513859982119923e-07,"output_power":1.7655801884067742e-07,"reference_active":true,"delta_db":-0.6515993994402652}],"active_band_count":3,"active_band_mae_db":1.003998261865408}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.404278039932251,"predicted_mos":3.404278039932251,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"2454829097d19f32effa96cb050a7494b6860226d892ef50785ce44cc173abbb","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.856546854000044},"started_at":"2026-10-13T05:11:35.795431+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00258385199958866,"native_input_samples":144000,"native_input_sha256":"ec69e9de668d1340af2e55fa2adae7e7852f7d1420298fe63d06ba4efe4e1f63","output_sha256":"839b47c21412ef425c73ef22e1cd3d5ed145c5843c79cbd47fdbaac017740f47","output_samples":144240,"length_delta_samples":240,"analysis_s":1.084100527001283,"synthesis_s":0.14298660599888535,"total_s":1.2270871330001683,"rtf":0.20451452216669472,"output_peak":0.693705141544342,"output_rms":0.08362421032951683,"finished_at":"2026-10-13T05:11:37.035396+00:00","analysis_stages":{"external_f0_s":0.1596399019999808},"evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"839b47c21412ef425c73ef22e1cd3d5ed145c5843c79cbd47fdbaac017740f47","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.551147961276907,"output_peak":0.6936839188092576},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.662976835215172,"spectral_convergence":0.1720451816537808,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.349465880089135,"spectral_convergence":0.16332786157606988,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.8414967209969495,"spectral_convergence":0.1702152778401526,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.001634299657240165,"delta_db":-0.06993261334716372,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.0018123936685634505,"delta_db":-0.2039133365305221,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.002622454844149967,"delta_db":-0.17054235878172996,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.0006810331882929747,"delta_db":-0.5116344730798171,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00021383135977226327,"delta_db":-0.634709625976009,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":4.4192898553866634e-05,"delta_db":-1.6012879042337613,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5320033853248339,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21589113114733563,"pitch":null,"output_rms_dbfs":-21.54393164711991,"output_peak":0.6936839188092576,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.617979812100419},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9349210248883036,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08582187018202725,"rms_dbfs":-21.328040515972575,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.582274007876475,"mean":1.6474088033040366e-05,"zero_energy":false},"output":{"rms":0.0837150262293018,"rms_dbfs":-21.54393164711991,"peak":0.6936839188092576,"peak_dbfs":-3.1767674666002033,"crest_factor_db":18.36716418051971,"mean":2.281202494929627e-05,"zero_energy":false},"rms_difference_db":-0.21589113114733394},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":6.052780056332951e-06,"output_rms_dbfs":-52.180451070318384,"output_zero_energy":false,"output_peak":0.012055295309851183,"output_energy_fraction":8.636704439749327e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":146,"both_estimated_voiced_frames":132,"reference_voiced_output_unvoiced_frames":12,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9166666666666666,"absolute_difference_cents_median":1.7422270129239972,"absolute_difference_cents_p95":6.140974655725715,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031196729436307625,"log_rms_mae_db":0.7902815105577716,"rms_envelope_nrmse":0.061977855246968305,"envelope_correlation":0.9964844832777149,"rms_gain_error_db":-0.21584590143475013,"reference_mean_power":0.00737113406357093,"output_mean_power":0.0070137409272428425},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.772697507541185e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010414087648688783,"output_power":0.0010088969915326566,"reference_active":true,"delta_db":-0.13774401535434022},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001272580988246754,"output_power":0.00011299966005987138,"reference_active":true,"delta_db":-0.5160829381136617},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4550125163182702e-05,"output_power":1.713651132424763e-05,"reference_active":true,"delta_db":-1.5613129832828532}],"active_band_count":3,"active_band_mae_db":0.7383799789169517}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002933828481800726,"log_rms_mae_db":0.4737090987789375,"rms_envelope_nrmse":0.04132393994414392,"envelope_correlation":0.9987210711414083,"rms_gain_error_db":-0.21570353540425205,"reference_mean_power":0.007404046443755548,"output_mean_power":0.00704528848189357},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.623838097183376e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009238145152881927,"output_power":0.0009002522977958098,"reference_active":true,"delta_db":-0.1122054321286087},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0904498001922555e-05,"output_power":5.4841043489918724e-05,"reference_active":true,"delta_db":-0.4554365817489889},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8155579307174762e-06,"output_power":1.230489884216239e-06,"reference_active":true,"delta_db":-1.6893206316868448}],"active_band_count":3,"active_band_mae_db":0.7523208818548142}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027755394478816403,"log_rms_mae_db":1.0824812220541986,"rms_envelope_nrmse":0.07237572946119243,"envelope_correlation":0.9962833902939185,"rms_gain_error_db":-0.19656576556053587,"reference_mean_power":0.003209968762055962,"output_mean_power":0.0030679214049384204},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3498130551666851e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006624323096451465,"output_power":0.0006397039096396213,"reference_active":true,"delta_db":-0.15162501708087975},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.187213693276335e-05,"output_power":6.3936885442352e-05,"reference_active":true,"delta_db":-0.5080908101303743},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2558954262897387e-05,"output_power":8.098474332614517e-06,"reference_active":true,"delta_db":-1.9055026875922616}],"active_band_count":3,"active_band_mae_db":0.8550728382678385}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025197665775423135,"log_rms_mae_db":0.6158697767371493,"rms_envelope_nrmse":0.04474384079569498,"envelope_correlation":0.9987769050929334,"rms_gain_error_db":-0.19649938348278873,"reference_mean_power":0.0032255924683410116,"output_mean_power":0.003082900853749071},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.287517318914353e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006130037756394711,"output_power":0.000594210611306891,"reference_active":true,"delta_db":-0.13522746361953222},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4809964114174026e-05,"output_power":3.0636411573843904e-05,"reference_active":true,"delta_db":-0.5546568030345779},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.24341074108939e-07,"output_power":6.309786862878049e-07,"reference_active":true,"delta_db":-1.6581756227777014}],"active_band_count":3,"active_band_mae_db":0.7826866298106038}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017139073768520646,"log_rms_mae_db":1.1612376654485101,"rms_envelope_nrmse":0.12041377050630825,"envelope_correlation":0.9916922084927723,"rms_gain_error_db":-0.5385474685581387,"reference_mean_power":0.0008077036407054461,"output_mean_power":0.0007135054481078644},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.727013621661233e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024375238055045323,"output_power":0.0002200515282123034,"reference_active":true,"delta_db":-0.44424476831784643},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.980069758207981e-05,"output_power":2.4062945478628028e-05,"reference_active":true,"delta_db":-0.9287764322306163},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.347390229447954e-06,"output_power":4.118567687424377e-06,"reference_active":true,"delta_db":-2.513868985236435}],"active_band_count":3,"active_band_mae_db":1.2956300619282992}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001528919021429252,"log_rms_mae_db":0.7734977580726737,"rms_envelope_nrmse":0.08237477928266004,"envelope_correlation":0.9974514189018492,"rms_gain_error_db":-0.5382523156247082,"reference_mean_power":0.0008114918980720448,"output_mean_power":0.000716900621073521},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4774424074646424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022848364185803236,"output_power":0.00020704346895101858,"reference_active":true,"delta_db":-0.4279357695692282},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508904456720678e-05,"output_power":9.863485941022425e-06,"reference_active":true,"delta_db":-1.0318884549214946},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.746475959138819e-07,"output_power":3.105051452058767e-07,"reference_active":true,"delta_db":-2.673327931139156}],"active_band_count":3,"active_band_mae_db":1.3777173852099596}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":1.2577554456287146,"rms_envelope_nrmse":0.1528889346897267,"envelope_correlation":0.9865570718582772,"rms_gain_error_db":-0.6744923383490055,"reference_mean_power":0.0001388420753372345,"output_mean_power":0.00011886989054656335},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.164366134779654e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.59650448810242e-05,"output_power":3.920508680443734e-05,"reference_active":true,"delta_db":-0.6908526873542348},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.502002650645408e-05,"output_power":1.1575025341747423e-05,"reference_active":true,"delta_db":-1.131487485736631},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429741089402253e-06,"output_power":2.3005868449389332e-06,"reference_active":true,"delta_db":-0.2372136600190297}],"active_band_count":3,"active_band_mae_db":0.6865179443699652}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":0.9122098164730961,"rms_envelope_nrmse":0.10491508091413554,"envelope_correlation":0.9956062895225217,"rms_gain_error_db":-0.6744370767035418,"reference_mean_power":0.00013952517981136642,"output_mean_power":0.00011945625166812314},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0998934517742824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3887505804801683e-05,"output_power":3.7548835748934393e-05,"reference_active":true,"delta_db":-0.6774442423856049},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188432498267155e-06,"output_power":5.625694138586999e-06,"reference_active":true,"delta_db":-1.064580824616611},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0513859982119923e-07,"output_power":1.7661438033280152e-07,"reference_active":true,"delta_db":-0.6502132500533703}],"active_band_count":3,"active_band_mae_db":0.797412772351862}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8455-210777-0060","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1144604682922363,"predicted_mos":3.1144604682922363,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":290687,"input_sample_rate":48000,"input_seconds":6.055979166666667,"input_float32_sha256":"a29b29728620c2e1c29f6e1b21684fcad5c54aaaa82db94e71f000f36d2ad1bf","internal_samples":96896,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.726554255999872},"started_at":"2026-10-13T05:18:50.334388+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008825694998449762,"native_input_samples":288000,"native_input_sha256":"552833b8b829c4bc2e2ddbca72b3f374be5742c813416b6c0e7b363fa5c7118a","output_sha256":"154f87c0869dd5cab2ba06c74e6408d1e753ddaae9d472afa225c8864471fa37","output_samples":290687,"length_delta_samples":2687,"analysis_s":3.1003614110013586,"synthesis_s":1.0443926299994928,"total_s":4.144754041000851,"rtf":0.6907923401668086,"output_peak":0.7524943203549325,"output_rms":0.08288788245598744,"finished_at":"2026-10-13T05:18:54.499978+00:00","evaluation_input":{"reference":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"output_sha256":"154f87c0869dd5cab2ba06c74e6408d1e753ddaae9d472afa225c8864471fa37","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96896,"length_delta_samples":896,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-21.628297466018523,"output_peak":0.7485325609509714},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9907529722589168,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.643404080426558,"spectral_convergence":0.7509108328171341,"spectral_convergence_status":"ok","amplitude_floor":2.2239575932173076e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.902167651611789,"spectral_convergence":0.6076068293455378,"spectral_convergence_status":"ok","amplitude_floor":1.640120571472942e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.222733455240473,"spectral_convergence":0.6390691107712806,"spectral_convergence_status":"ok","amplitude_floor":1.3411038650227006e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0016608291202231148,"output_power":0.0014906182887128208,"delta_db":-0.4695850557893613,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.001899519997387598,"output_power":0.0017903453127416317,"delta_db":-0.2570706641324566,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.002727484270807038,"output_power":0.0025500078963377956,"delta_db":-0.29220729455082156,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0007661816102082716,"output_power":0.000745007474330533,"delta_db":-0.12171094037640634,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.000247481283821024,"output_power":0.00026505891680199314,"delta_db":0.29800058557835585,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":6.389711909369124e-05,"output_power":5.511680757320745e-05,"delta_db":-0.6419722291285525,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.34675779492599235,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.2858895325982536,"pitch":null,"output_rms_dbfs":-21.61393004857083,"output_peak":0.7485325609509714,"reference_rms_dbfs":-21.328040515972575,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.922768395759606},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.4241428494527948,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.08582187018202725,"rms_dbfs":-21.328040515972575,"peak":0.728973388671875,"peak_dbfs":-2.7457665080961013,"crest_factor_db":18.582274007876475,"mean":1.6474088033040366e-05,"zero_energy":false},"output":{"rms":0.08304308939639699,"rms_dbfs":-21.613930048570825,"peak":0.7485325609509714,"peak_dbfs":-2.5157860645010794,"crest_factor_db":19.09814398406975,"mean":-0.0007952951208513007,"zero_energy":false},"rms_difference_db":-0.2858895325982509},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":9.732359275221824e-06,"threshold_dbfs":-50.1178186723338,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":5.692408594768495e-06,"output_mean_power":0.00011574492462161661,"output_rms_dbfs":-39.36498043568339,"output_zero_energy":false,"output_peak":0.10565912168590519,"output_energy_fraction":0.0016783980307227505,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":144,"output_estimated_voiced_frames":139,"both_estimated_voiced_frames":90,"reference_voiced_output_unvoiced_frames":54,"reference_unvoiced_output_voiced_frames":49,"paired_voiced_fraction":0.625,"absolute_difference_cents_median":13.94557934916945,"absolute_difference_cents_p95":135.72357636009917,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00031196729436307625,"log_rms_mae_db":5.8282627238419735,"rms_envelope_nrmse":0.4388453734985491,"envelope_correlation":0.7948755642725008,"rms_gain_error_db":-0.28595726445191144,"reference_mean_power":0.00737113406357093,"output_mean_power":0.006901421994858175},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.772697507541185e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010414087648688783,"output_power":0.001048727467205009,"reference_active":true,"delta_db":0.03041414220569059},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001272580988246754,"output_power":0.00012541743574359158,"reference_active":true,"delta_db":-0.06327513752408695},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4550125163182702e-05,"output_power":2.5461860857519183e-05,"reference_active":true,"delta_db":0.15836429889137946}],"active_band_count":3,"active_band_mae_db":0.08401785954038567}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0002933828481800726,"log_rms_mae_db":4.79637874816021,"rms_envelope_nrmse":0.37482985062626945,"envelope_correlation":0.8392058214408298,"rms_gain_error_db":-0.29554239675768734,"reference_mean_power":0.007404046443755548,"output_mean_power":0.006916954113835371},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.623838097183376e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009238145152881927,"output_power":0.0009347631061840451,"reference_active":true,"delta_db":0.05116781304825491},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0904498001922555e-05,"output_power":5.7776740331844295e-05,"reference_active":true,"delta_db":-0.2289633190202281},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8155579307174762e-06,"output_power":2.069138192445632e-06,"reference_active":true,"delta_db":0.5677938613839232}],"active_band_count":3,"active_band_mae_db":0.2826416644841354}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00027755394478816403,"log_rms_mae_db":6.662848786257228,"rms_envelope_nrmse":0.44819132798513206,"envelope_correlation":0.8425435376958439,"rms_gain_error_db":-0.2801479084106293,"reference_mean_power":0.003209968762055962,"output_mean_power":0.0030094422598901543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.3498130551666851e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006624323096451465,"output_power":0.0006427777842403549,"reference_active":true,"delta_db":-0.13080648470129697},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.187213693276335e-05,"output_power":6.811552120808617e-05,"reference_active":true,"delta_db":-0.23314473560893442},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2558954262897387e-05,"output_power":1.1536853631411984e-05,"reference_active":true,"delta_db":-0.3686609607983157}],"active_band_count":3,"active_band_mae_db":0.2442040603695157}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025197665775423135,"log_rms_mae_db":5.665051660615191,"rms_envelope_nrmse":0.37858006317843795,"envelope_correlation":0.8816609461385134,"rms_gain_error_db":-0.2865055702934977,"reference_mean_power":0.0032255924683410116,"output_mean_power":0.003019666210538834},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.287517318914353e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006130037756394711,"output_power":0.0005962230077892262,"reference_active":true,"delta_db":-0.12054418340669495},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4809964114174026e-05,"output_power":3.137652870518361e-05,"reference_active":true,"delta_db":-0.4509868117227793},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.24341074108939e-07,"output_power":9.340475722060653e-07,"reference_active":true,"delta_db":0.04536744135413857}],"active_band_count":3,"active_band_mae_db":0.2056328121612043}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017139073768520646,"log_rms_mae_db":6.84280726899024,"rms_envelope_nrmse":0.4719435918934679,"envelope_correlation":0.8399339617534617,"rms_gain_error_db":-0.06901636227868925,"reference_mean_power":0.0008077036407054461,"output_mean_power":0.0007949693856048273},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.727013621661233e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00024375238055045323,"output_power":0.0002392916053131956,"reference_active":true,"delta_db":-0.08021402608384319},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.980069758207981e-05,"output_power":2.690137250713741e-05,"reference_active":true,"delta_db":-0.44451992009798097},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.347390229447954e-06,"output_power":7.187920987011348e-06,"reference_active":true,"delta_db":-0.09529811906372465}],"active_band_count":3,"active_band_mae_db":0.2066773550818496}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.0001528919021429252,"log_rms_mae_db":5.687216453196896,"rms_envelope_nrmse":0.38128460456551416,"envelope_correlation":0.8899881384872921,"rms_gain_error_db":-0.06802895732262111,"reference_mean_power":0.0008114918980720448,"output_mean_power":0.0007988795286279487},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.4774424074646424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022848364185803236,"output_power":0.00022400145971120857,"reference_active":true,"delta_db":-0.08604264045423977},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2508904456720678e-05,"output_power":1.1044318291478079e-05,"reference_active":true,"delta_db":-0.5408036109458476},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.746475959138819e-07,"output_power":5.314443978444566e-07,"reference_active":true,"delta_db":-0.3394376078409753}],"active_band_count":3,"active_band_mae_db":0.3220946197470209}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.106734387114863e-05,"log_rms_mae_db":6.904556463103875,"rms_envelope_nrmse":0.7676923356418807,"envelope_correlation":0.6084290901095387,"rms_gain_error_db":0.30941071702106077,"reference_mean_power":0.0001388420753372345,"output_mean_power":0.00014909468725362462},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.164366134779654e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.59650448810242e-05,"output_power":5.002540420467338e-05,"reference_active":true,"delta_db":0.3676291791150178},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.502002650645408e-05,"output_power":1.6302491938907447e-05,"reference_active":true,"delta_db":0.3558329504794334},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.429741089402253e-06,"output_power":5.360344360312786e-06,"reference_active":true,"delta_db":3.4363269251100377}],"active_band_count":3,"active_band_mae_db":1.386596351568163}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":4.915795071779818e-05,"log_rms_mae_db":5.744754926436884,"rms_envelope_nrmse":0.6390698249627923,"envelope_correlation":0.7087419291881116,"rms_gain_error_db":0.30723129473198185,"reference_mean_power":0.00013952517981136642,"output_mean_power":0.000149753065167176},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0998934517742824e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.3887505804801683e-05,"output_power":4.9425763696083744e-05,"reference_active":true,"delta_db":0.516124885695295},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.188432498267155e-06,"output_power":8.200760572228529e-06,"reference_active":true,"delta_db":0.5721993357950887},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0513859982119923e-07,"output_power":3.8798143908766176e-07,"reference_active":true,"delta_db":2.7676356267497133}],"active_band_count":3,"active_band_mae_db":1.2853199494133658}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},{"case":"librispeech_8463-294828-0003","method":"hnm","status":"measured","mos":{"status":"measured","value":2.5285608768463135,"predicted_mos":2.5285608768463135,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d1e65d792ce57ca4039f0dc9ea415d4181c9fa4355a120f94d5c02d24b77c7aa","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.928085966999788},"started_at":"2026-10-13T04:40:28.305339+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.944900021859212e-05,"native_input_samples":96000,"native_input_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","output_sha256":"713e5ee3e4042d0fc99aef42cb0fb4d283b01477d52d7aedb71b8acdb5e080c3","output_samples":96000,"length_delta_samples":0,"analysis_s":1.0031502290003118,"synthesis_s":0.5360062279996782,"total_s":1.53915645699999,"rtf":0.256526076166665,"output_peak":0.48447264011670993,"output_rms":0.05703995357602581,"finished_at":"2026-10-13T04:40:29.851720+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"713e5ee3e4042d0fc99aef42cb0fb4d283b01477d52d7aedb71b8acdb5e080c3","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.729577203039873,"spectral_convergence":0.309237797397153,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.476992507596821,"spectral_convergence":0.30252977321265606,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.846527243980887,"spectral_convergence":0.3191559994776914,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00014842124093899625,"delta_db":0.25881102988734456,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.00041894784673977565,"delta_db":-0.027384034424323425,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0017241055186331278,"delta_db":-0.03128238226631099,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0006402159856346902,"delta_db":-0.130769192483207,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.00017770885912645823,"delta_db":-0.19020443541090282,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.0001441568528821305,"delta_db":-0.04704076038398051,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11424863914267823,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0472928028158083,"pitch":null,"output_rms_dbfs":-24.87641673243799,"output_peak":0.48447264011670993,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.684365651539193},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.729577203039873,"spectral_convergence":0.309237797397153,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.476992507596821,"spectral_convergence":0.30252977321265606,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.846527243980887,"spectral_convergence":0.3191559994776914,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00014842124093899625,"delta_db":0.25881102988734456,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.00041894784673977565,"delta_db":-0.027384034424323425,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0017241055186331278,"delta_db":-0.03128238226631099,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0006402159856346902,"delta_db":-0.130769192483207,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.00017770885912645823,"delta_db":-0.19020443541090282,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.0001441568528821305,"delta_db":-0.04704076038398051,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11424863914267823,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.0472928028158083,"pitch":null,"output_rms_dbfs":-24.87641673243799,"output_peak":0.48447264011670993,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.684365651539193},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8848421962962835,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05735137089841256,"rms_dbfs":-24.82912392962218,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.807777249691203,"mean":-2.4424870808919272e-05,"zero_energy":false},"output":{"rms":0.05703995357602581,"rms_dbfs":-24.87641673243799,"peak":0.48447264011670993,"peak_dbfs":-6.294614881318942,"crest_factor_db":18.58180185111905,"mean":4.284439477251115e-06,"zero_energy":false},"rms_difference_db":-0.04729280281580728},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":7.850483510382759e-06,"output_rms_dbfs":-51.05103594283469,"output_zero_energy":false,"output_peak":0.055366797071038164,"output_energy_fraction":0.0002412893085894759,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":156,"both_estimated_voiced_frames":147,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9363057324840764,"absolute_difference_cents_median":1.2112525129312988,"absolute_difference_cents_p95":5.755458423235052,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017345649672524565,"log_rms_mae_db":0.9421548689077769,"rms_envelope_nrmse":0.0818602654309942,"envelope_correlation":0.9931134654454775,"rms_gain_error_db":-0.047469804374223865,"reference_mean_power":0.003291322552458033,"output_mean_power":0.00325554321881435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4377088226614287e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006890430591645026,"output_power":0.0006825091808939043,"reference_active":true,"delta_db":-0.04137864620507214},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716540294963627e-05,"output_power":9.090171367688018e-05,"reference_active":true,"delta_db":-0.28939585442399707},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.855953967229281e-05,"output_power":1.8034922791415886e-05,"reference_active":true,"delta_db":-1.9963691570766064}],"active_band_count":3,"active_band_mae_db":0.7757145525685586}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015649665907765232,"log_rms_mae_db":0.4804512384901468,"rms_envelope_nrmse":0.04026722225211564,"envelope_correlation":0.9981646322956463,"rms_gain_error_db":-0.047607603660935985,"reference_mean_power":0.003302205038848876,"output_mean_power":0.003266203767270802},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.335462563779621e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377316706455328,"output_power":0.0006341511250975865,"reference_active":true,"delta_db":-0.024452178876361613},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.741968368909027e-05,"output_power":4.346669165860435e-05,"reference_active":true,"delta_db":-0.37802066328084216},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7622617238454128e-06,"output_power":1.024546986670959e-06,"reference_active":true,"delta_db":-2.3553852815202676}],"active_band_count":3,"active_band_mae_db":0.9192860412258238}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014746435244218596,"log_rms_mae_db":1.5364359674204742,"rms_envelope_nrmse":0.08579264796653248,"envelope_correlation":0.9937261122505927,"rms_gain_error_db":-0.03844228245901865,"reference_mean_power":0.001906578877331925,"output_mean_power":0.0018897769564611333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.936060821906154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045087668804334914,"output_power":0.0004390666347361733,"reference_active":true,"delta_db":-0.11527345553702392},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.461610811295139e-05,"output_power":5.3560956660075155e-05,"reference_active":true,"delta_db":-0.08472424126841639},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.197452814329269e-06,"output_power":8.479636376353961e-06,"reference_active":true,"delta_db":1.3616400039038992}],"active_band_count":3,"active_band_mae_db":0.5205459002364465}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013246473438636124,"log_rms_mae_db":0.8323093228998776,"rms_envelope_nrmse":0.042996460875543034,"envelope_correlation":0.9983145685008717,"rms_gain_error_db":-0.038505658248650904,"reference_mean_power":0.0019129188838057705,"output_mean_power":0.0018960334223145478},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.520895123968316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004268128119473447,"output_power":0.0004166435947282792,"reference_active":true,"delta_db":-0.10472737892816553},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5267927477999584e-05,"output_power":2.4717154318630527e-05,"reference_active":true,"delta_db":-0.09571152694492031},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.022115629444589e-07,"output_power":5.959486622854824e-07,"reference_active":true,"delta_db":0.7432214154949275}],"active_band_count":3,"active_band_mae_db":0.3145534404560044}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":1.4757849347048477,"rms_envelope_nrmse":0.14921531175703207,"envelope_correlation":0.9815203632064907,"rms_gain_error_db":-0.16674484982390603,"reference_mean_power":0.0007310287661738598,"output_mean_power":0.0007034933264043328},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.375697567971487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020778711952139394,"output_power":0.00019691371710390848,"reference_active":true,"delta_db":-0.23342652298890826},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.101230284766395e-05,"output_power":3.7046586041129596e-05,"reference_active":true,"delta_db":-0.44165963116618584},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4179813437956632e-05,"output_power":7.611187825187747e-06,"reference_active":true,"delta_db":-2.702180775446773}],"active_band_count":3,"active_band_mae_db":1.1257556432006224}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":0.889886532580589,"rms_envelope_nrmse":0.08353778160251496,"envelope_correlation":0.9938695293497449,"rms_gain_error_db":-0.16693070470820473,"reference_mean_power":0.0007338466075139545,"output_mean_power":0.0007061748078021154},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9980296656839775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019557456921819077,"output_power":0.00018344378295782386,"reference_active":true,"delta_db":-0.27809384533523746},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6985486820992017e-05,"output_power":1.5771878475819495e-05,"reference_active":true,"delta_db":-0.3219457657343819},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.922014736318651e-07,"output_power":4.283408302813008e-07,"reference_active":true,"delta_db":-3.648103935563313}],"active_band_count":3,"active_band_mae_db":1.4160478488776442}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011116745339746554,"log_rms_mae_db":1.5543209476161999,"rms_envelope_nrmse":0.18875756157656637,"envelope_correlation":0.9678424705026429,"rms_gain_error_db":-0.2289317241845967,"reference_mean_power":0.0001764719915242474,"output_mean_power":0.00016741046951261184},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1170311171266687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.030239739448222e-05,"output_power":3.9111189611597194e-05,"reference_active":true,"delta_db":-0.13029855268674467},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8894394185439035e-05,"output_power":2.277102535731713e-05,"reference_active":true,"delta_db":-1.0343100634167444},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.877164672472967e-06,"output_power":5.4511267065713986e-06,"reference_active":true,"delta_db":-1.5988364757706295}],"active_band_count":3,"active_band_mae_db":0.9211483639580395}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":0.9238260023236662,"rms_envelope_nrmse":0.11830012612432182,"envelope_correlation":0.9860654960313445,"rms_gain_error_db":-0.2289101106501618,"reference_mean_power":0.00017721561821746633,"output_mean_power":0.0001681167489660364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0001638026368733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2510198337172984e-05,"output_power":4.072721030290329e-05,"reference_active":true,"delta_db":-0.1860846805408746},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6080590241633715e-05,"output_power":1.118411065707204e-05,"reference_active":true,"delta_db":-1.5770053019035635},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.447224316204932e-07,"output_power":4.029859810510003e-07,"reference_active":true,"delta_db":-2.040828424377439}],"active_band_count":3,"active_band_mae_db":1.267972802273959}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0003","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0123772621154785,"predicted_mos":3.0123772621154785,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4fc7b869ea1a634d1550ed594e350010ece8bdfe750b87c3f95109da7f38d307","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.554876680000234},"started_at":"2026-10-13T04:43:44.613851+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.9166000129189342e-05,"native_input_samples":96000,"native_input_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","output_sha256":"62798de55094de97f1aa0673b795d95e26bcd283aedeab8662d66b97bf7983f6","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3734363990006386,"synthesis_s":0.08800387999963277,"total_s":1.4614402790002714,"rtf":0.24357337983337857,"output_peak":0.5944408725667626,"output_rms":0.062488632180394466,"finished_at":"2026-10-13T04:43:46.080536+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"62798de55094de97f1aa0673b795d95e26bcd283aedeab8662d66b97bf7983f6","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.809123098067134,"spectral_convergence":0.7011570319205637,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.553311402242754,"spectral_convergence":0.28193252057061136,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.010213378972877,"spectral_convergence":0.33892091964807836,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00014393419852163836,"delta_db":0.1254903844671173,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.0004902618559918363,"delta_db":0.6552973893310645,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.002117729396469519,"delta_db":0.8617838495254362,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0007917743192533826,"delta_db":0.7919797840561684,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.0002017876470478793,"delta_db":0.36165053959467036,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.00015934173449237398,"delta_db":0.38790171702864507,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.530683944000517,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7451442987878123,"pitch":null,"output_rms_dbfs":-24.08397963083437,"output_peak":0.5944408725667626,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.124215959760922},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.809123098067134,"spectral_convergence":0.7011570319205637,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.553311402242754,"spectral_convergence":0.28193252057061136,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.010213378972877,"spectral_convergence":0.33892091964807836,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00014393419852163836,"delta_db":0.1254903844671173,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.0004902618559918363,"delta_db":0.6552973893310645,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.002117729396469519,"delta_db":0.8617838495254362,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0007917743192533826,"delta_db":0.7919797840561684,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.0002017876470478793,"delta_db":0.36165053959467036,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.00015934173449237398,"delta_db":0.38790171702864507,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.530683944000517,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":0.7451442987878123,"pitch":null,"output_rms_dbfs":-24.08397963083437,"output_peak":0.5944408725667626,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.124215959760922},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9225730788048265,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05735137089841256,"rms_dbfs":-24.82912392962218,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.807777249691203,"mean":-2.4424870808919272e-05,"zero_energy":false},"output":{"rms":0.062488632180394466,"rms_dbfs":-24.08397963083437,"peak":0.5944408725667626,"peak_dbfs":-4.517826739998752,"crest_factor_db":19.566152890835617,"mean":6.20893163417634e-06,"zero_energy":false},"rms_difference_db":0.7451442987878103},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":2.1658973378778377e-06,"output_rms_dbfs":-56.64362132498234,"output_zero_energy":false,"output_peak":0.037343432588426054,"output_energy_fraction":5.5467147311487156e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":190,"both_estimated_voiced_frames":151,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":39,"paired_voiced_fraction":0.9617834394904459,"absolute_difference_cents_median":5.3034989456575,"absolute_difference_cents_p95":16.237220932767567,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017345649672524565,"log_rms_mae_db":1.523644661059814,"rms_envelope_nrmse":0.17142823187404743,"envelope_correlation":0.9799318008680353,"rms_gain_error_db":0.7453353490476431,"reference_mean_power":0.003291322552458033,"output_mean_power":0.003907544931488132},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4377088226614287e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006890430591645026,"output_power":0.0008152606923119685,"reference_active":true,"delta_db":0.7305014101473499},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716540294963627e-05,"output_power":0.00010424601146595886,"reference_active":true,"delta_db":0.30547791466182506},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.855953967229281e-05,"output_power":2.8067194040218827e-05,"reference_active":true,"delta_db":-0.07552206015009011}],"active_band_count":3,"active_band_mae_db":0.3705004616530883}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015649665907765232,"log_rms_mae_db":1.022168567113156,"rms_envelope_nrmse":0.1082221042305799,"envelope_correlation":0.9962101271429589,"rms_gain_error_db":0.7463559545561809,"reference_mean_power":0.003302205038848876,"output_mean_power":0.003921386336002433},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.335462563779621e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377316706455328,"output_power":0.0007611332778821497,"reference_active":true,"delta_db":0.7682272516560538},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.741968368909027e-05,"output_power":5.049469842652481e-05,"reference_active":true,"delta_db":0.2728713006306046},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7622617238454128e-06,"output_power":1.3917509457119661e-06,"reference_active":true,"delta_db":-1.0250888340442335}],"active_band_count":3,"active_band_mae_db":0.688729128776964}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014746435244218596,"log_rms_mae_db":1.7665813446610708,"rms_envelope_nrmse":0.16694283541185256,"envelope_correlation":0.9861527401280721,"rms_gain_error_db":0.826838509018681,"reference_mean_power":0.001906578877331925,"output_mean_power":0.0023064212362345167},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.936060821906154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045087668804334914,"output_power":0.0005415867852636592,"reference_active":true,"delta_db":0.7961027771548275},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.461610811295139e-05,"output_power":6.486621308149593e-05,"reference_active":true,"delta_db":0.7469779483108084},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.197452814329269e-06,"output_power":8.03555809498183e-06,"reference_active":true,"delta_db":1.128028164194806}],"active_band_count":3,"active_band_mae_db":0.8903696298868139}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013246473438636124,"log_rms_mae_db":1.2088129326150523,"rms_envelope_nrmse":0.1158595928531352,"envelope_correlation":0.9971251408349182,"rms_gain_error_db":0.8277146104264048,"reference_mean_power":0.0019129188838057705,"output_mean_power":0.0023145577192146345},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.520895123968316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004268128119473447,"output_power":0.0005154477544171608,"reference_active":true,"delta_db":0.8194720445332024},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5267927477999584e-05,"output_power":3.178424556439595e-05,"reference_active":true,"delta_db":0.9964228566351301},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.022115629444589e-07,"output_power":5.174085144870677e-07,"reference_active":true,"delta_db":0.12946863489697807}],"active_band_count":3,"active_band_mae_db":0.6484545120217702}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":1.7028328502819898,"rms_envelope_nrmse":0.18930315268454231,"envelope_correlation":0.9795471583313523,"rms_gain_error_db":0.7747847854122153,"reference_mean_power":0.0007310287661738598,"output_mean_power":0.0008738018209889361},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.375697567971487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020778711952139394,"output_power":0.00025208204576927573,"reference_active":true,"delta_db":0.8392329204015383},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.101230284766395e-05,"output_power":4.4868990752382565e-05,"reference_active":true,"delta_db":0.3903214528615661},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4179813437956632e-05,"output_power":1.263204549456172e-05,"reference_active":true,"delta_db":-0.5019683580288328}],"active_band_count":3,"active_band_mae_db":0.5771742437639791}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":1.0679009353574342,"rms_envelope_nrmse":0.12265998686936974,"envelope_correlation":0.9948166938814258,"rms_gain_error_db":0.7758246187414,"reference_mean_power":0.0007338466075139545,"output_mean_power":0.0008773800452057191},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9980296656839775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019557456921819077,"output_power":0.00023709129821785718,"reference_active":true,"delta_db":0.8360323236810631},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6985486820992017e-05,"output_power":1.801858004010906e-05,"reference_active":true,"delta_db":0.2564256458030447},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.922014736318651e-07,"output_power":6.579357707556265e-07,"reference_active":true,"delta_db":-1.784163688911635}],"active_band_count":3,"active_band_mae_db":0.9588738861319142}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011116745339746554,"log_rms_mae_db":1.702029406917848,"rms_envelope_nrmse":0.25863970993571633,"envelope_correlation":0.9442519113089085,"rms_gain_error_db":0.3788185010173294,"reference_mean_power":0.0001764719915242474,"output_mean_power":0.0001925562577093304},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1170311171266687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.030239739448222e-05,"output_power":4.3023904924799304e-05,"reference_active":true,"delta_db":0.2837894418104774},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8894394185439035e-05,"output_power":3.1220592237273426e-05,"reference_active":true,"delta_db":0.33627543870438986},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.877164672472967e-06,"output_power":7.014965306259679e-06,"reference_active":true,"delta_db":-0.5034439701918962}],"active_band_count":3,"active_band_mae_db":0.37450295023558783}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":1.006073112712163,"rms_envelope_nrmse":0.14394050861748692,"envelope_correlation":0.9814429805565862,"rms_gain_error_db":0.37906769753596686,"reference_mean_power":0.00017721561821746633,"output_mean_power":0.00019337875678612117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0001638026368733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2510198337172984e-05,"output_power":4.5265220690839445e-05,"reference_active":true,"delta_db":0.27271510884432837},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6080590241633715e-05,"output_power":1.7103893455996753e-05,"reference_active":true,"delta_db":0.26792996974679384},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.447224316204932e-07,"output_power":5.713352398266353e-07,"reference_active":true,"delta_db":-0.5248176867160383}],"active_band_count":3,"active_band_mae_db":0.3551542551023868}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0003","method":"facodec","status":"measured","mos":{"status":"measured","value":4.280514717102051,"predicted_mos":4.280514717102051,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"290bfefb813d49f28b21426d78c15b72777218feb201ce0054a49ba1a71037a2","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.455675812998379},"started_at":"2026-10-13T07:01:53.915047+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":9.641500037105288e-05,"native_input_samples":96000,"native_input_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","output_sha256":"7dc7ff1acabafaca403820f6ca37831833a45808cdc4eb0c9291ffb0075e1ab7","output_samples":96000,"length_delta_samples":0,"analysis_s":1.520738106000863,"synthesis_s":6.839986398999827,"total_s":8.36072450500069,"rtf":1.3934540841667815,"output_peak":0.4248102605342865,"output_rms":0.05443514906010086,"finished_at":"2026-10-13T07:02:02.289848+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"7dc7ff1acabafaca403820f6ca37831833a45808cdc4eb0c9291ffb0075e1ab7","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.108511473668677,"spectral_convergence":0.2794050258647137,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.240449174859713,"spectral_convergence":0.3045193312459835,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4147185797282695,"spectral_convergence":0.32655848424006784,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.0001338493522861425,"delta_db":-0.18998681460932204,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.0004304568273343373,"delta_db":0.09031234342868533,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0015757582526718288,"delta_db":-0.422024898800782,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0005345044274667994,"delta_db":-0.9145212618991241,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.00016265110362824443,"delta_db":-0.5747250787240119,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.0001259654898080458,"delta_db":-0.632877888744688,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4707413810344356,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4532877405958964,"pitch":null,"output_rms_dbfs":-25.282411670218078,"output_peak":0.4248102605342865,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.25455974275222},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.108511473668677,"spectral_convergence":0.2794050258647137,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.240449174859713,"spectral_convergence":0.3045193312459835,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.4147185797282695,"spectral_convergence":0.32655848424006784,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.0001338493522861425,"delta_db":-0.18998681460932204,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.0004304568273343373,"delta_db":0.09031234342868533,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0015757582526718288,"delta_db":-0.422024898800782,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0005345044274667994,"delta_db":-0.9145212618991241,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.00016265110362824443,"delta_db":-0.5747250787240119,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.0001259654898080458,"delta_db":-0.632877888744688,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4707413810344356,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4532877405958964,"pitch":null,"output_rms_dbfs":-25.282411670218078,"output_peak":0.4248102605342865,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.25455974275222},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9171601557611722,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05735137089841256,"rms_dbfs":-24.82912392962218,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.807777249691203,"mean":-2.4424870808919272e-05,"zero_energy":false},"output":{"rms":0.05443514906010086,"rms_dbfs":-25.282411670218078,"peak":0.4248102605342865,"peak_dbfs":-7.436100043823316,"crest_factor_db":17.846311626394765,"mean":-1.1799161979567429e-05,"zero_energy":false},"rms_difference_db":-0.45328774059589705},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":1.4286436137155056e-06,"output_rms_dbfs":-58.450760958418435,"output_zero_energy":false,"output_peak":0.006226417142897844,"output_energy_fraction":4.821310161923564e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":169,"both_estimated_voiced_frames":153,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9745222929936306,"absolute_difference_cents_median":2.5322879199735255,"absolute_difference_cents_p95":11.27348134419257,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017345649672524565,"log_rms_mae_db":1.055543855154027,"rms_envelope_nrmse":0.11093189780759254,"envelope_correlation":0.9896087375414832,"rms_gain_error_db":-0.45370200864013555,"reference_mean_power":0.003291322552458033,"output_mean_power":0.002964833060385374},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4377088226614287e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006890430591645026,"output_power":0.0006149248047241788,"reference_active":true,"delta_db":-0.4942435047295266},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716540294963627e-05,"output_power":9.642020993537895e-05,"reference_active":true,"delta_db":-0.03343583314636205},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.855953967229281e-05,"output_power":2.375046933595678e-05,"reference_active":true,"delta_db":-0.8007900693278887}],"active_band_count":3,"active_band_mae_db":0.4428231357345924}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015649665907765232,"log_rms_mae_db":0.8612396599045508,"rms_envelope_nrmse":0.10109455036230704,"envelope_correlation":0.9910480459599279,"rms_gain_error_db":-0.45740297506792194,"reference_mean_power":0.003302205038848876,"output_mean_power":0.0029721021936122724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.335462563779621e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377316706455328,"output_power":0.0005735141362673644,"reference_active":true,"delta_db":-0.46093858113480574},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.741968368909027e-05,"output_power":4.6977325026263554e-05,"reference_active":true,"delta_db":-0.04070369076215157},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7622617238454128e-06,"output_power":1.3763192336348966e-06,"reference_active":true,"delta_db":-1.0735122944271054}],"active_band_count":3,"active_band_mae_db":0.5250515221080209}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014746435244218596,"log_rms_mae_db":1.5188402431688799,"rms_envelope_nrmse":0.11962026934897227,"envelope_correlation":0.9886043119332159,"rms_gain_error_db":-0.33241242769432805,"reference_mean_power":0.001906578877331925,"output_mean_power":0.0017660928588111753},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.936060821906154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045087668804334914,"output_power":0.00043028787578086694,"reference_active":true,"delta_db":-0.20298672148763036},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.461610811295139e-05,"output_power":5.693933402608525e-05,"reference_active":true,"delta_db":0.1809163371163811},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.197452814329269e-06,"output_power":5.855676843753691e-06,"reference_active":true,"delta_db":-0.2463612748021632}],"active_band_count":3,"active_band_mae_db":0.21008811113539153}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013246473438636124,"log_rms_mae_db":1.157051583452355,"rms_envelope_nrmse":0.11002441400041209,"envelope_correlation":0.989809757077729,"rms_gain_error_db":-0.33722748672518976,"reference_mean_power":0.0019129188838057705,"output_mean_power":0.0017700021982963287},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.520895123968316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004268128119473447,"output_power":0.00040827464336170725,"reference_active":true,"delta_db":-0.1928503936518704},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5267927477999584e-05,"output_power":2.6492610931750193e-05,"reference_active":true,"delta_db":0.20555139773567893},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.022115629444589e-07,"output_power":5.034844801881399e-07,"reference_active":true,"delta_db":0.010993803422707966}],"active_band_count":3,"active_band_mae_db":0.13646519827008577}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":1.5666893658451726,"rms_envelope_nrmse":0.1577623293620242,"envelope_correlation":0.9849250904744159,"rms_gain_error_db":-0.8296280835774275,"reference_mean_power":0.0007310287661738598,"output_mean_power":0.0006039092178038178},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.375697567971487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020778711952139394,"output_power":0.0001719803979827113,"reference_active":true,"delta_db":-0.8213967303030889},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.101230284766395e-05,"output_power":4.336316641945689e-05,"reference_active":true,"delta_db":0.24206831602441004},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4179813437956632e-05,"output_power":1.3199116859868977e-05,"reference_active":true,"delta_db":-0.31125642971513107}],"active_band_count":3,"active_band_mae_db":0.45824049201421}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":1.2369874743942724,"rms_envelope_nrmse":0.13692848861413331,"envelope_correlation":0.9896216624675868,"rms_gain_error_db":-0.8334450973405025,"reference_mean_power":0.0007338466075139545,"output_mean_power":0.0006057044734155531},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9980296656839775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019557456921819077,"output_power":0.00016586645767648945,"reference_active":true,"delta_db":-0.7155381256504},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6985486820992017e-05,"output_power":1.8290739066667772e-05,"reference_active":true,"delta_db":0.3215325549912285},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.922014736318651e-07,"output_power":9.134879770366681e-07,"reference_active":true,"delta_db":-0.35897032005109886}],"active_band_count":3,"active_band_mae_db":0.4653470002309092}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011116745339746554,"log_rms_mae_db":1.595879136241855,"rms_envelope_nrmse":0.1942396013876122,"envelope_correlation":0.9681113722440178,"rms_gain_error_db":-0.6129270210750375,"reference_mean_power":0.0001764719915242474,"output_mean_power":0.00015324386070342824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1170311171266687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.030239739448222e-05,"output_power":2.9146195154295036e-05,"reference_active":true,"delta_db":-1.407490125026256},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8894394185439035e-05,"output_power":2.1535925416642686e-05,"reference_active":true,"delta_db":-1.276500547639099},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.877164672472967e-06,"output_power":6.433299113329939e-06,"reference_active":true,"delta_db":-0.8793618031427628}],"active_band_count":3,"active_band_mae_db":1.187784158602706}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":1.1884535707113477,"rms_envelope_nrmse":0.15891249441280136,"envelope_correlation":0.9771259628906422,"rms_gain_error_db":-0.6130790248316547,"reference_mean_power":0.00017721561821746633,"output_mean_power":0.0001538842214459679},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0001638026368733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2510198337172984e-05,"output_power":2.9697512034166688e-05,"reference_active":true,"delta_db":-1.5577306424156911},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6080590241633715e-05,"output_power":1.2041236560016505e-05,"reference_active":true,"delta_db":-1.2563089702930883},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.447224316204932e-07,"output_power":5.163560781973788e-07,"reference_active":true,"delta_db":-0.9642348713206021}],"active_band_count":3,"active_band_mae_db":1.259424828009794}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0003","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.9928462505340576,"predicted_mos":2.9928462505340576,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"4856ccb4e1298d9f02dac51cb2176878cffbba6cf6566491153a57c9bf29f77f","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8599163649996626},"started_at":"2026-10-13T07:32:37.452235+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00788473300053738,"native_input_samples":264600,"native_input_sha256":"27294ffe1ba9331599c50e6c904e043a6370ce4abee3b90c125cc4dd280194df","output_sha256":"4f41d6a49d72f41bcd5f39c5d9ab1929a1bcaa2be523b46958fbab685647d063","output_samples":264704,"length_delta_samples":104,"analysis_s":0.16856367199943634,"synthesis_s":4.318019090002053,"total_s":4.48658276200149,"rtf":0.7477637936669149,"output_peak":0.46928805112838745,"output_rms":0.05296821716051699,"finished_at":"2026-10-13T07:32:41.970878+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"4f41d6a49d72f41bcd5f39c5d9ab1929a1bcaa2be523b46958fbab685647d063","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.513269135981087,"output_peak":0.46500398832523504},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.215756233238787,"spectral_convergence":0.5563650786009021,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.058944630399932,"spectral_convergence":0.4422517985497425,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.011566971503871,"spectral_convergence":0.5121056525317982,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00018112189552252003,"delta_db":1.1235599923672916,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.00036986693149159734,"delta_db":-0.5685286176698015,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0014514286009277286,"delta_db":-0.7789640344091862,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0005589908967951413,"delta_db":-0.719986977174205,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.00017873873180686831,"delta_db":-0.1651085010785296,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":7.059487694246759e-05,"delta_db":-3.1476618308199757,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0839683255864982,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6826663407631386,"pitch":null,"output_rms_dbfs":-25.511790270385323,"output_peak":0.46500398832523504,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.095422611714197},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7538805446457136,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05735137089841256,"rms_dbfs":-24.82912392962218,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.807777249691203,"mean":-2.4424870808919272e-05,"zero_energy":false},"output":{"rms":0.05301643078788239,"rms_dbfs":-25.511790270385323,"peak":0.46500398832523504,"peak_dbfs":-6.650866443267009,"crest_factor_db":18.860923827118313,"mean":-7.463095031074124e-05,"zero_energy":false},"rms_difference_db":-0.6826663407631415},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":9.723507448397549e-07,"output_rms_dbfs":-60.12177048786009,"output_zero_energy":false,"output_peak":0.009772691711702699,"output_energy_fraction":3.4594095361636164e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":163,"both_estimated_voiced_frames":136,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.8662420382165605,"absolute_difference_cents_median":7.209048441041052,"absolute_difference_cents_p95":26.13928875608294,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017345649672524565,"log_rms_mae_db":1.6177251254206448,"rms_envelope_nrmse":0.1350553694500443,"envelope_correlation":0.9867241472311784,"rms_gain_error_db":-0.6828598259273501,"reference_mean_power":0.003291322552458033,"output_mean_power":0.002812447751322413},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4377088226614287e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006890430591645026,"output_power":0.0006317432899455556,"reference_active":true,"delta_db":-0.3770572459986953},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716540294963627e-05,"output_power":7.999799331327629e-05,"reference_active":true,"delta_db":-0.8443256285730156},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.855953967229281e-05,"output_power":1.9562499191804575e-05,"reference_active":true,"delta_db":-1.6432686617705026}],"active_band_count":3,"active_band_mae_db":0.9548838454474046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015649665907765232,"log_rms_mae_db":1.2757966917289463,"rms_envelope_nrmse":0.10129541940335181,"envelope_correlation":0.9949220262415726,"rms_gain_error_db":-0.6811433800255984,"reference_mean_power":0.003302205038848876,"output_mean_power":0.0028228623266235236},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.335462563779621e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377316706455328,"output_power":0.0005892131371730061,"reference_active":true,"delta_db":-0.3436556382840473},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.741968368909027e-05,"output_power":3.6078008689972464e-05,"reference_active":true,"delta_db":-1.1871609387362176},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7622617238454128e-06,"output_power":1.2102643302095423e-06,"reference_active":true,"delta_db":-1.6319017485796254}],"active_band_count":3,"active_band_mae_db":1.05423944186663}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014746435244218596,"log_rms_mae_db":2.6943254184158167,"rms_envelope_nrmse":0.16150233439730077,"envelope_correlation":0.9832735175941291,"rms_gain_error_db":-0.783047328200991,"reference_mean_power":0.001906578877331925,"output_mean_power":0.00159202559062111},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.936060821906154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045087668804334914,"output_power":0.0003571693814450188,"reference_active":true,"delta_db":-1.0118355962777172},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.461610811295139e-05,"output_power":4.124162756536038e-05,"reference_active":true,"delta_db":-1.2198495352543104},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.197452814329269e-06,"output_power":6.131582171329654e-06,"reference_active":true,"delta_db":-0.046406760962838176}],"active_band_count":3,"active_band_mae_db":0.7593639641649553}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013246473438636124,"log_rms_mae_db":2.105561107553101,"rms_envelope_nrmse":0.13116288883281774,"envelope_correlation":0.9906075268791131,"rms_gain_error_db":-0.7790250928686573,"reference_mean_power":0.0019129188838057705,"output_mean_power":0.0015987996522652296},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.520895123968316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004268128119473447,"output_power":0.00033820451846522793,"reference_active":true,"delta_db":-1.0105804193547478},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5267927477999584e-05,"output_power":1.8631457360302373e-05,"reference_active":true,"delta_db":-1.3232279482760823},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.022115629444589e-07,"output_power":4.4412200929036397e-07,"reference_active":true,"delta_db":-0.5338441173417401}],"active_band_count":3,"active_band_mae_db":0.9558841616575234}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":2.751688073118309,"rms_envelope_nrmse":0.21213687486282806,"envelope_correlation":0.9661720265559061,"rms_gain_error_db":-0.7507198942702209,"reference_mean_power":0.0007310287661738598,"output_mean_power":0.0006149821032831387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.375697567971487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020778711952139394,"output_power":0.00019735876350294838,"reference_active":true,"delta_db":-0.2236220711496265},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.101230284766395e-05,"output_power":3.9452838118963745e-05,"reference_active":true,"delta_db":-0.16835905166944098},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4179813437956632e-05,"output_power":1.1748082086396903e-05,"reference_active":true,"delta_db":-0.817035445550267}],"active_band_count":3,"active_band_mae_db":0.40300552278977814}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":2.088248416721706,"rms_envelope_nrmse":0.1697414281757419,"envelope_correlation":0.9784684705805033,"rms_gain_error_db":-0.7467404484141361,"reference_mean_power":0.0007338466075139545,"output_mean_power":0.0006179185685656599},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9980296656839775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019557456921819077,"output_power":0.00018852851828057016,"reference_active":true,"delta_db":-0.15935328071023216},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6985486820992017e-05,"output_power":1.6410888491085863e-05,"reference_active":true,"delta_db":-0.14945904156792647},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.922014736318651e-07,"output_power":8.491126782522568e-07,"reference_active":true,"delta_db":-0.6763454226373227}],"active_band_count":3,"active_band_mae_db":0.3283859149718271}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011116745339746554,"log_rms_mae_db":4.153264994640318,"rms_envelope_nrmse":0.36307344532001834,"envelope_correlation":0.9191237539383349,"rms_gain_error_db":-2.4442511531430573,"reference_mean_power":0.0001764719915242474,"output_mean_power":0.00010051958143375061},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1170311171266687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.030239739448222e-05,"output_power":2.0107071789014023e-05,"reference_active":true,"delta_db":-3.0198205248061756},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8894394185439035e-05,"output_power":1.339195993863346e-05,"reference_active":true,"delta_db":-3.3396945177762607},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.877164672472967e-06,"output_power":2.872944149596509e-06,"reference_active":true,"delta_db":-4.380427411759399}],"active_band_count":3,"active_band_mae_db":3.5799808181139454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":3.8430603924661444,"rms_envelope_nrmse":0.33450801568154226,"envelope_correlation":0.9330929525483384,"rms_gain_error_db":-2.442332460052166,"reference_mean_power":0.00017721561821746633,"output_mean_power":0.0001009877620569181},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0001638026368733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2510198337172984e-05,"output_power":1.9944639033501995e-05,"reference_active":true,"delta_db":-3.2866695056035606},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6080590241633715e-05,"output_power":7.008742081071776e-06,"reference_active":true,"delta_db":-3.606619071673908},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.447224316204932e-07,"output_power":2.5924008865996253e-07,"reference_active":true,"delta_db":-3.956706194327946}],"active_band_count":3,"active_band_mae_db":3.6166649238684716}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0003","method":"golf","status":"measured","mos":{"status":"measured","value":3.2881484031677246,"predicted_mos":3.2881484031677246,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"d8524e78615c4046bd654a2edcc5cca1a7ef036bd053233491c1f0968945afe5","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4899974960007967},"started_at":"2026-10-13T05:03:09.153608+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003585972000109905,"native_input_samples":144000,"native_input_sha256":"88a15d2128c187d0c44f39591961f4e3867d319b3ade1ff8abf99dba25e476cf","output_sha256":"dc9c32336234dcd966cfbaf49ee3943883fd187af38e638099231adc6b19fa34","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6623754610000105,"synthesis_s":0.11282864399981918,"total_s":0.7752041049998297,"rtf":0.1292006841666383,"output_peak":0.46344834566116333,"output_rms":0.04765755512467775,"finished_at":"2026-10-13T05:03:09.942720+00:00","analysis_stages":{"external_f0_s":0.10735306999958993},"evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"dc9c32336234dcd966cfbaf49ee3943883fd187af38e638099231adc6b19fa34","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.430566091815848,"output_peak":0.4638045649269442},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.735802424001981,"spectral_convergence":0.5596543926014456,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.750370669670354,"spectral_convergence":0.3495647734260863,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.900254052698416,"spectral_convergence":0.38486527106210494,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013974133284308112,"output_power":0.0001392582668276788,"delta_db":-0.015038954519844445,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042180904081497476,"output_power":0.0003247254079656061,"delta_db":-1.1359961282944544,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017377037270426502,"output_power":0.0011536785728572315,"delta_db":-1.7789090599430462,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006601198232599226,"output_power":0.00048541883199766525,"delta_db":-1.3350615378674986,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00018595065659488544,"output_power":0.00010468075961118434,"delta_db":-2.495308507813062,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00014585177156966952,"output_power":6.70390576733488e-05,"delta_db":-3.375838075358359,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.689358710632711,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.6040776303562483,"pitch":null,"output_rms_dbfs":-26.430566091815848,"output_peak":0.4638045649269442,"reference_rms_dbfs":-24.826488461459597,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.12880904879025},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8807849853758294,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05736877506209439,"rms_dbfs":-24.826488461459597,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.80514178152862,"mean":-2.5322202450046953e-05,"zero_energy":false},"output":{"rms":0.04769487285791539,"rms_dbfs":-26.430566091815848,"peak":0.4638045649269442,"peak_dbfs":-6.673299623972454,"crest_factor_db":19.757266467843394,"mean":0.0004672395864992753,"zero_energy":false},"rms_difference_db":-1.604077630356251},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":1.1415910210732088e-06,"output_rms_dbfs":-59.42489455725307,"output_zero_energy":false,"output_peak":0.0056920144450984264,"output_energy_fraction":5.026799711117562e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":175,"both_estimated_voiced_frames":151,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":24,"paired_voiced_fraction":0.9617834394904459,"absolute_difference_cents_median":4.816125810767859,"absolute_difference_cents_p95":15.634352291267803,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00017345649672524565,"log_rms_mae_db":2.3679036452467015,"rms_envelope_nrmse":0.21377210079785983,"envelope_correlation":0.9789594334937808,"rms_gain_error_db":-1.6032864042502308,"reference_mean_power":0.0032926040313614314,"output_mean_power":0.0022762023383709504},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.4358342106082128e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000690110508395343,"output_power":0.0005208076233707939,"reference_active":true,"delta_db":-1.2224130787725156},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.60511447582864e-05,"output_power":6.501531339863956e-05,"reference_active":true,"delta_db":-1.694868848895833},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8619211540291917e-05,"output_power":1.5419036111814723e-05,"reference_active":true,"delta_db":-2.6860043919965726}],"active_band_count":3,"active_band_mae_db":1.8677621065549737}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00015649665907765232,"log_rms_mae_db":2.0823297077614145,"rms_envelope_nrmse":0.18687004016670777,"envelope_correlation":0.9912855400101727,"rms_gain_error_db":-1.6037020667505677,"reference_mean_power":0.0033042545076913186,"output_mean_power":0.002284037786399395},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.333660252023872e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006385868785805495,"output_power":0.0004835134550005543,"reference_active":true,"delta_db":-1.2081142664773576},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.653236427545713e-05,"output_power":3.03791830862925e-05,"reference_active":true,"delta_db":-1.8517902802407449},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7600272457803608e-06,"output_power":9.634656517520733e-07,"reference_active":true,"delta_db":-2.616831545278104}],"active_band_count":3,"active_band_mae_db":1.8922453639987353}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00014746435244218596,"log_rms_mae_db":2.281047303767634,"rms_envelope_nrmse":0.22076657394865845,"envelope_correlation":0.9775339368427483,"rms_gain_error_db":-1.5499171111359513,"reference_mean_power":0.0019072801898495325,"output_mean_power":0.0013348202509575846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":8.92399965631592e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045141602632981794,"output_power":0.00033821528091711504,"reference_active":true,"delta_db":-1.253837479101462},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.401175741879334e-05,"output_power":3.598450025481922e-05,"reference_active":true,"delta_db":-1.7637283278329188},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.161227144628521e-06,"output_power":4.778316434524405e-06,"reference_active":true,"delta_db":-1.103923134727808}],"active_band_count":3,"active_band_mae_db":1.3738296472207299}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00013246473438636124,"log_rms_mae_db":1.815218875594165,"rms_envelope_nrmse":0.1942402362953257,"envelope_correlation":0.9877971483357886,"rms_gain_error_db":-1.5504674126466949,"reference_mean_power":0.0019141441157284542,"output_mean_power":0.0013394542710733564},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":8.510262165557862e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00042727986489670804,"output_power":0.0003201550559837764,"reference_active":true,"delta_db":-1.2535206291798227},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.4827181193092968e-05,"output_power":1.6244650065634286e-05,"reference_active":true,"delta_db":-1.8421705338346681},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.941983271266956e-07,"output_power":4.076963529245421e-07,"reference_active":true,"delta_db":-0.8356444451250755}],"active_band_count":3,"active_band_mae_db":1.3104452027131888}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":2.3025953005012667,"rms_envelope_nrmse":0.23659878580250002,"envelope_correlation":0.9729497626086294,"rms_gain_error_db":-1.6119335184040795,"reference_mean_power":0.0007314912355923217,"output_mean_power":0.0005046796289961567},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.3678163678891514e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020802554418307686,"output_power":0.00015562053216025602,"reference_active":true,"delta_db":-1.2604977067735255},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.05089261950812e-05,"output_power":2.89213563318794e-05,"reference_active":true,"delta_db":-1.4633207480511938},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4190255691691866e-05,"output_power":9.043526189998626e-06,"reference_active":true,"delta_db":-1.9565242036583683}],"active_band_count":3,"active_band_mae_db":1.5601142194943627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":1.8661855911030008,"rms_envelope_nrmse":0.19648394384468099,"envelope_correlation":0.9891532097513074,"rms_gain_error_db":-1.6114303896246156,"reference_mean_power":0.0007346462512322926,"output_mean_power":0.0005069150996007841},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.989780391849273e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001957264929131279,"output_power":0.00014794064944056643,"reference_active":true,"delta_db":-1.2156209348377645},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6552769881791507e-05,"output_power":1.2547275647344868e-05,"reference_active":true,"delta_db":-1.2032123850760892},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.931967968367397e-07,"output_power":5.582865847911609e-07,"reference_active":true,"delta_db":-2.5017811803537104}],"active_band_count":3,"active_band_mae_db":1.6402048334225217}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00011116745339746554,"log_rms_mae_db":4.598733374641076,"rms_envelope_nrmse":0.38076695276450256,"envelope_correlation":0.9614841850547802,"rms_gain_error_db":-3.5946540686506427,"reference_mean_power":0.00017666641635322318,"output_mean_power":7.721267393617326e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.1192203314371691e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.039379999814441e-05,"output_power":2.114912638653283e-05,"reference_active":true,"delta_db":-2.810222782769452},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8953303347585365e-05,"output_power":1.303186111959325e-05,"reference_active":true,"delta_db":-3.4669167747241705},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.883108107789692e-06,"output_power":2.5744342388760326e-06,"reference_active":true,"delta_db":-4.86015679730104}],"active_band_count":3,"active_band_mae_db":3.712432118264888}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":4.538360566974682,"rms_envelope_nrmse":0.36112855275776706,"envelope_correlation":0.9780829956795134,"rms_gain_error_db":-3.5959581839489396,"reference_mean_power":0.00017744052700853215,"output_mean_power":7.752771789879348e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.0020859995340736e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.262082662158869e-05,"output_power":2.210389704316168e-05,"reference_active":true,"delta_db":-2.851530194754783},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.607018076083558e-05,"output_power":6.6838099682532154e-06,"reference_active":true,"delta_db":-3.809966681400758},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.465400111909579e-07,"output_power":1.7572065788081224e-07,"reference_active":true,"delta_db":-5.657725858519413}],"active_band_count":3,"active_band_mae_db":4.106407578224984}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0003","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.9437084197998047,"predicted_mos":2.9437084197998047,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"64ba3a879e6a4f20c6f5d7d39125dc8275cfcee48b19c612a1ac087efe54fec3","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.891019001999666},"started_at":"2026-10-13T07:47:11.995061+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004185202000371646,"native_input_samples":144000,"native_input_sha256":"88a15d2128c187d0c44f39591961f4e3867d319b3ade1ff8abf99dba25e476cf","output_sha256":"f1de4a633b7eabdb37036f78a90a6f21dabf90846bfe30c413b531523274e09f","output_samples":144120,"length_delta_samples":120,"analysis_s":2.725845887001924,"synthesis_s":3.9311842729985074,"total_s":6.657030160000431,"rtf":1.1095050266667386,"output_peak":0.5300824642181396,"output_rms":0.054931119230095714,"finished_at":"2026-10-13T07:47:18.669870+00:00","analysis_stages":{"external_f0_s":1.8324518900008115},"evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"f1de4a633b7eabdb37036f78a90a6f21dabf90846bfe30c413b531523274e09f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.1969917456575,"output_peak":0.5301039765411074},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.394612909558538,"spectral_convergence":0.6194263952900227,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.233331701077181,"spectral_convergence":0.3247991170944867,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.142479028780375,"spectral_convergence":0.3400268482424606,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00015805873745652446,"delta_db":0.5320355317107349,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.0004087661990267894,"delta_db":-0.1342338936932409,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0015762592844587749,"delta_db":-0.42064422537750173,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0005977560750873062,"delta_db":-0.4287943453405736,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.0001678880789673043,"delta_db":-0.4370966250460242,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.00011564401658436187,"delta_db":-1.0041620129168645,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4928277723474899,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.36452342149228945,"pitch":null,"output_rms_dbfs":-25.19364735111447,"output_peak":0.5301039765411074,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.590141213138698},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8993378950686075,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05735137089841256,"rms_dbfs":-24.82912392962218,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.807777249691203,"mean":-2.4424870808919272e-05,"zero_energy":false},"output":{"rms":0.054994294172950896,"rms_dbfs":-25.19364735111447,"peak":0.5301039765411074,"peak_dbfs":-5.512778758580977,"crest_factor_db":19.680868592533496,"mean":-0.0002544476233445067,"zero_energy":false},"rms_difference_db":-0.3645234214922901},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":1.563589439870732e-05,"output_rms_dbfs":-48.05877271356544,"output_zero_energy":false,"output_peak":0.11839711860548092,"output_energy_fraction":0.0005169963342554285,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":153,"both_estimated_voiced_frames":143,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.910828025477707,"absolute_difference_cents_median":6.112223343513017,"absolute_difference_cents_p95":16.56923014555824,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017345649672524565,"log_rms_mae_db":2.171021992240267,"rms_envelope_nrmse":0.1830671513288996,"envelope_correlation":0.9659357579572297,"rms_gain_error_db":-0.36379793957264783,"reference_mean_power":0.003291322552458033,"output_mean_power":0.003026848273421131},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4377088226614287e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006890430591645026,"output_power":0.0006465055648117202,"reference_active":true,"delta_db":-0.27674094929851617},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716540294963627e-05,"output_power":8.110272530799518e-05,"reference_active":true,"delta_db":-0.7847620792640759},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.855953967229281e-05,"output_power":1.964951362207836e-05,"reference_active":true,"delta_db":-1.6239939822874776}],"active_band_count":3,"active_band_mae_db":0.8951656702833567}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015649665907765232,"log_rms_mae_db":1.4937981573662795,"rms_envelope_nrmse":0.10424292799949703,"envelope_correlation":0.9890619436999695,"rms_gain_error_db":-0.36370858417255725,"reference_mean_power":0.003302205038848876,"output_mean_power":0.0030369187809140075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.335462563779621e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377316706455328,"output_power":0.0005982657799274748,"reference_active":true,"delta_db":-0.2774382275531536},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.741968368909027e-05,"output_power":3.945587504637211e-05,"reference_active":true,"delta_db":-0.7984697316800833},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7622617238454128e-06,"output_power":1.1917902202447962e-06,"reference_active":true,"delta_db":-1.6987059118163979}],"active_band_count":3,"active_band_mae_db":0.9248712903498783}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014746435244218596,"log_rms_mae_db":2.4566506304052527,"rms_envelope_nrmse":0.1832619420912127,"envelope_correlation":0.9717337863997326,"rms_gain_error_db":-0.38150791332436473,"reference_mean_power":0.001906578877331925,"output_mean_power":0.0017462402163285241},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.936060821906154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045087668804334914,"output_power":0.00041918038347023496,"reference_active":true,"delta_db":-0.3165683066873037},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.461610811295139e-05,"output_power":4.7140717556035734e-05,"reference_active":true,"delta_db":-0.6392456064803259},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.197452814329269e-06,"output_power":4.588216223093358e-06,"reference_active":true,"delta_db":-1.3056935269470191}],"active_band_count":3,"active_band_mae_db":0.7538358133715496}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013246473438636124,"log_rms_mae_db":1.6665404224858438,"rms_envelope_nrmse":0.10855369101202743,"envelope_correlation":0.9903745613225955,"rms_gain_error_db":-0.3809751994441208,"reference_mean_power":0.0019129188838057705,"output_mean_power":0.001752261966236847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.520895123968316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004268128119473447,"output_power":0.00039550929598088525,"reference_active":true,"delta_db":-0.3308075194339078},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5267927477999584e-05,"output_power":2.1842507307247746e-05,"reference_active":true,"delta_db":-0.6326713207275924},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.022115629444589e-07,"output_power":3.983158920607862e-07,"reference_active":true,"delta_db":-1.0065907341577627}],"active_band_count":3,"active_band_mae_db":0.656689858106421}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":2.340874678434865,"rms_envelope_nrmse":0.2330372936950979,"envelope_correlation":0.9545425870012169,"rms_gain_error_db":-0.33072876581548816,"reference_mean_power":0.0007310287661738598,"output_mean_power":0.0006774255776191444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.375697567971487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020778711952139394,"output_power":0.00018749817411994608,"reference_active":true,"delta_db":-0.4462157978076215},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.101230284766395e-05,"output_power":3.65890954659694e-05,"reference_active":true,"delta_db":-0.4956248243256722},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4179813437956632e-05,"output_power":1.2128408878857795e-05,"reference_active":true,"delta_db":-0.6786668724313605}],"active_band_count":3,"active_band_mae_db":0.5401691648548846}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":1.502826333002186,"rms_envelope_nrmse":0.12749918450047104,"envelope_correlation":0.9859662206650837,"rms_gain_error_db":-0.33203278756483184,"reference_mean_power":0.0007338466075139545,"output_mean_power":0.000679832640171198},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9980296656839775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019557456921819077,"output_power":0.00017685927592091317,"reference_active":true,"delta_db":-0.43684539682639606},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6985486820992017e-05,"output_power":1.537852781159956e-05,"reference_active":true,"delta_db":-0.43163236307466},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.922014736318651e-07,"output_power":8.193051935242225e-07,"reference_active":true,"delta_db":-0.8315415988744526}],"active_band_count":3,"active_band_mae_db":0.5666731195918362}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011116745339746554,"log_rms_mae_db":2.42121262195268,"rms_envelope_nrmse":0.2810102845455647,"envelope_correlation":0.9303809335085252,"rms_gain_error_db":-0.8930204615898374,"reference_mean_power":0.0001764719915242474,"output_mean_power":0.00014367253037438093},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1170311171266687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.030239739448222e-05,"output_power":3.2103831165673526e-05,"reference_active":true,"delta_db":-0.9877401821210314},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8894394185439035e-05,"output_power":1.9355741324288438e-05,"reference_active":true,"delta_db":-1.7400378382030488},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.877164672472967e-06,"output_power":4.1127882574004525e-06,"reference_active":true,"delta_db":-2.8223357362160972}],"active_band_count":3,"active_band_mae_db":1.850037918846726}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":1.5114588685971277,"rms_envelope_nrmse":0.1730155337059696,"envelope_correlation":0.9766258997711068,"rms_gain_error_db":-0.8930778556607248,"reference_mean_power":0.00017721561821746633,"output_mean_power":0.0001442760383338797},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0001638026368733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2510198337172984e-05,"output_power":3.2118962536130654e-05,"reference_active":true,"delta_db":-1.2173162240009563},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6080590241633715e-05,"output_power":1.0191403458208285e-05,"reference_active":true,"delta_db":-1.980679907608219},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.447224316204932e-07,"output_power":3.0093914018799075e-07,"reference_active":true,"delta_db":-3.3089410492584066}],"active_band_count":3,"active_band_mae_db":2.168979060289194}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0003","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.340547561645508,"predicted_mos":3.340547561645508,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"4a9c8ea812f85f03bf730c639646e419f16ad4b76044140c6b8f4232bb8d2a0e","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.308640485000069},"started_at":"2026-10-13T05:11:37.038428+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031362849986180663,"native_input_samples":144000,"native_input_sha256":"88a15d2128c187d0c44f39591961f4e3867d319b3ade1ff8abf99dba25e476cf","output_sha256":"f01efbe35f1b67953590994cf636ff4b3ca6dd80fd3ddd115453c114fe96c454","output_samples":144240,"length_delta_samples":240,"analysis_s":0.9057573970003432,"synthesis_s":0.13548049999917566,"total_s":1.0412378969995189,"rtf":0.17353964949991982,"output_peak":0.4744175672531128,"output_rms":0.05575714732921033,"finished_at":"2026-10-13T05:11:38.091775+00:00","analysis_stages":{"external_f0_s":0.17308132699872658},"evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"f01efbe35f1b67953590994cf636ff4b3ca6dd80fd3ddd115453c114fe96c454","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.06705000692218,"output_peak":0.47460876296022997},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.536506883397776,"spectral_convergence":0.20819920291193722,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.191592759639419,"spectral_convergence":0.19776877239327018,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.640638374480312,"spectral_convergence":0.22854073694278507,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00013604012225479521,"delta_db":-0.11947941836260742,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.0004219378748579062,"delta_db":0.00350145101075321,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0016752103866248392,"delta_db":-0.15622723149283288,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0005972072121031456,"delta_db":-0.43278389354945823,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.00015400020501457913,"delta_db":-0.8120822330001594,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.00013449568947989914,"delta_db":-0.34831002965971025,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.31206404284592026,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23087341600429465,"pitch":null,"output_rms_dbfs":-25.059997345626478,"output_peak":0.47460876296022997,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.456246005839169},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9528575843240744,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05735137089841256,"rms_dbfs":-24.82912392962218,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.807777249691203,"mean":-2.4424870808919272e-05,"zero_energy":false},"output":{"rms":0.05584703654031398,"rms_dbfs":-25.059997345626478,"peak":0.47460876296022997,"peak_dbfs":-6.473284948579269,"crest_factor_db":18.586712397047208,"mean":-2.4947115549761894e-05,"zero_energy":false},"rms_difference_db":-0.23087341600429667},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":3.820452610464427e-06,"output_rms_dbfs":-54.17885183008957,"output_zero_energy":false,"output_peak":0.042903978887089765,"output_energy_fraction":0.00012249392523924809,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":167,"both_estimated_voiced_frames":151,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":0.9617834394904459,"absolute_difference_cents_median":0.9973707519295819,"absolute_difference_cents_p95":3.5528314679948645,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017345649672524565,"log_rms_mae_db":0.7788310434405475,"rms_envelope_nrmse":0.07300030822466119,"envelope_correlation":0.9951059311583238,"rms_gain_error_db":-0.23094976926875008,"reference_mean_power":0.003291322552458033,"output_mean_power":0.003120868492398998},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4377088226614287e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006890430591645026,"output_power":0.0006454273954799494,"reference_active":true,"delta_db":-0.2839896706244793},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716540294963627e-05,"output_power":8.477047594567009e-05,"reference_active":true,"delta_db":-0.592670345305661},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.855953967229281e-05,"output_power":1.8641382313079598e-05,"reference_active":true,"delta_db":-1.852730897058802}],"active_band_count":3,"active_band_mae_db":0.9097969709963141}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015649665907765232,"log_rms_mae_db":0.4808147943621633,"rms_envelope_nrmse":0.0475213534561536,"envelope_correlation":0.99815642248868,"rms_gain_error_db":-0.23046905812220822,"reference_mean_power":0.003302205038848876,"output_mean_power":0.003131533990033083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.335462563779621e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377316706455328,"output_power":0.0005975406144919293,"reference_active":true,"delta_db":-0.2827055582381199},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.741968368909027e-05,"output_power":4.081477302839978e-05,"reference_active":true,"delta_db":-0.6514126698598246},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7622617238454128e-06,"output_power":9.372425759916372e-07,"reference_active":true,"delta_db":-2.742183994822785}],"active_band_count":3,"active_band_mae_db":1.2254340743069099}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014746435244218596,"log_rms_mae_db":0.9944763140300638,"rms_envelope_nrmse":0.06979193888191801,"envelope_correlation":0.9959866369616021,"rms_gain_error_db":-0.1319205439379011,"reference_mean_power":0.001906578877331925,"output_mean_power":0.0018495357183007965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.936060821906154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045087668804334914,"output_power":0.0004362794999308252,"reference_active":true,"delta_db":-0.14292974620602783},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.461610811295139e-05,"output_power":5.005318171478665e-05,"reference_active":true,"delta_db":-0.37889060180391515},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.197452814329269e-06,"output_power":5.393301169315529e-06,"reference_active":true,"delta_db":-0.603585562925346}],"active_band_count":3,"active_band_mae_db":0.3751353036450964}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013246473438636124,"log_rms_mae_db":0.5734981676384667,"rms_envelope_nrmse":0.04404573457462316,"envelope_correlation":0.9983946213810397,"rms_gain_error_db":-0.13144041202791998,"reference_mean_power":0.0019129188838057705,"output_mean_power":0.0018558912030753568},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.520895123968316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004268128119473447,"output_power":0.00041327537396180976,"reference_active":true,"delta_db":-0.1399791993929041},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5267927477999584e-05,"output_power":2.372919004158261e-05,"reference_active":true,"delta_db":-0.2728670723232003},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.022115629444589e-07,"output_power":4.117465911332964e-07,"reference_active":true,"delta_db":-0.8625669539509091}],"active_band_count":3,"active_band_mae_db":0.4251377418890045}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":1.1212501244464084,"rms_envelope_nrmse":0.12529214576367456,"envelope_correlation":0.9885975523428514,"rms_gain_error_db":-0.45421780674538703,"reference_mean_power":0.0007310287661738598,"output_mean_power":0.0006584346632661479},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.375697567971487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020778711952139394,"output_power":0.00018110257655585936,"reference_active":true,"delta_db":-0.596939935641435},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.101230284766395e-05,"output_power":3.2960385008826255e-05,"reference_active":true,"delta_db":-0.9492187965023858},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4179813437956632e-05,"output_power":1.0493441158267805e-05,"reference_active":true,"delta_db":-1.3075258536025278}],"active_band_count":3,"active_band_mae_db":0.9512281952487829}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":0.7192027451439928,"rms_envelope_nrmse":0.08427805208571602,"envelope_correlation":0.9956988470211895,"rms_gain_error_db":-0.4542080153111245,"reference_mean_power":0.0007338466075139545,"output_mean_power":0.0006609741718087508},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9980296656839775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019557456921819077,"output_power":0.00016995992547607464,"reference_active":true,"delta_db":-0.6096585033318382},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6985486820992017e-05,"output_power":1.414051816211477e-05,"reference_active":true,"delta_db":-0.79612674751086},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.922014736318651e-07,"output_power":5.652751607581734e-07,"reference_active":true,"delta_db":-2.443399655488567}],"active_band_count":3,"active_band_mae_db":1.2830616354437552}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011116745339746554,"log_rms_mae_db":1.0979808594707114,"rms_envelope_nrmse":0.15012093392366369,"envelope_correlation":0.9827254342702133,"rms_gain_error_db":-0.6024513005068919,"reference_mean_power":0.0001764719915242474,"output_mean_power":0.0001536139500385245},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1170311171266687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.030239739448222e-05,"output_power":3.4361052547455386e-05,"reference_active":true,"delta_db":-0.6926442231778385},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8894394185439035e-05,"output_power":2.241255839818474e-05,"reference_active":true,"delta_db":-1.1032215908906282},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.877164672472967e-06,"output_power":3.729180160399732e-06,"reference_active":true,"delta_db":-3.247565594432359}],"active_band_count":3,"active_band_mae_db":1.6811438028336088}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":0.7461694205581862,"rms_envelope_nrmse":0.10070046148688701,"envelope_correlation":0.993767610022905,"rms_gain_error_db":-0.6025138852542421,"reference_mean_power":0.00017721561821746633,"output_mean_power":0.00015425903335665513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0001638026368733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2510198337172984e-05,"output_power":3.476291702174252e-05,"reference_active":true,"delta_db":-0.8737691946004078},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6080590241633715e-05,"output_power":1.1890000867884506e-05,"reference_active":true,"delta_db":-1.3112009926127621},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.447224316204932e-07,"output_power":2.7767253498909573e-07,"reference_active":true,"delta_db":-3.6583985559266368}],"active_band_count":3,"active_band_mae_db":1.9477895810466024}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0003","method":"magphase","status":"measured","mos":{"status":"measured","value":3.258652448654175,"predicted_mos":3.258652448654175,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":288081,"input_sample_rate":48000,"input_seconds":6.0016875,"input_float32_sha256":"f2888f400cd9a8978b34cc727487f3a98d6fba14bf5ea10620cb1e754432120c","internal_samples":96027,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.953771357999358},"started_at":"2026-10-13T05:18:54.503233+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008564683999793488,"native_input_samples":288000,"native_input_sha256":"5889ec6f30434862e1ce424737b305ca55faff8b1d62d93273b818847a162aa1","output_sha256":"672f5de74feb417bafb88feb3e9eb6cb5c08169c7b621b9c9cd14e51e92ca73e","output_samples":288081,"length_delta_samples":81,"analysis_s":3.427293283000836,"synthesis_s":1.0835271089999878,"total_s":4.510820392000824,"rtf":0.7518033986668039,"output_peak":0.48032701537122163,"output_rms":0.05671995984209435,"finished_at":"2026-10-13T05:18:59.038672+00:00","evaluation_input":{"reference":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"672f5de74feb417bafb88feb3e9eb6cb5c08169c7b621b9c9cd14e51e92ca73e","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96027,"length_delta_samples":27,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.918523941833854,"output_peak":0.4770789428426105},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9997188290793214,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.095863463782942,"spectral_convergence":0.7091602325153433,"spectral_convergence_status":"ok","amplitude_floor":1.2198198453862723e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.153513139001757,"spectral_convergence":0.49275565638246216,"spectral_convergence_status":"ok","amplitude_floor":7.927151445605076e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.364793676061657,"spectral_convergence":0.46552707299958834,"spectral_convergence_status":"ok","amplitude_floor":6.454878545778745e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00013983470011737857,"output_power":0.00012780989071201228,"delta_db":-0.3905049189396856,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00042159782923140474,"output_power":0.00040412963129192166,"delta_db":-0.18377671441022103,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0017365691446010552,"output_power":0.0016639020074216224,"delta_db":-0.1856433450209,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0006597865121791135,"output_power":0.0006695163984307051,"delta_db":0.06357785484512236,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0001856647778551521,"output_power":0.00018717350181830282,"delta_db":0.035148433200828264,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.000145726779943178,"output_power":0.0001703245058852726,"delta_db":0.677377691116398,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.25600482625552584,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08846714429501318,"pitch":null,"output_rms_dbfs":-24.917591073917194,"output_peak":0.4770789428426105,"reference_rms_dbfs":-24.82912392962218,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.204723426282119},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7592735025754063,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05735137089841256,"rms_dbfs":-24.82912392962218,"peak":0.445587158203125,"peak_dbfs":-7.021346679930981,"crest_factor_db":17.807777249691203,"mean":-2.4424870808919272e-05,"zero_energy":false},"output":{"rms":0.05677020288461049,"rms_dbfs":-24.917591073917194,"peak":0.4770789428426105,"peak_dbfs":-6.428195035372539,"crest_factor_db":18.489396038544655,"mean":-0.00019624236325160675,"zero_energy":false},"rms_difference_db":-0.0884671442950129},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.008715625619516e-06,"threshold_dbfs":-55.21618858474801,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":60,"quiet_samples":9600,"reference_mean_power":1.2587817036546768e-06,"output_mean_power":3.31203967337521e-06,"output_rms_dbfs":-54.79904469645389,"output_zero_energy":false,"output_peak":0.03640131310278449,"output_energy_fraction":0.00010276722694400798,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":157,"output_estimated_voiced_frames":162,"both_estimated_voiced_frames":134,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.8535031847133758,"absolute_difference_cents_median":5.479166952223565,"absolute_difference_cents_p95":23.635523360694016,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00017345649672524565,"log_rms_mae_db":3.44631222155947,"rms_envelope_nrmse":0.32251249009747573,"envelope_correlation":0.8931665880995419,"rms_gain_error_db":-0.08865558293708241,"reference_mean_power":0.003291322552458033,"output_mean_power":0.0032248156094883454},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.4377088226614287e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006890430591645026,"output_power":0.0007045793163235445,"reference_active":true,"delta_db":0.09683527492930216},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.716540294963627e-05,"output_power":8.946087337263906e-05,"reference_active":true,"delta_db":-0.35878522340531155},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.855953967229281e-05,"output_power":2.5562210754544713e-05,"reference_active":true,"delta_db":-0.4815279194178679}],"active_band_count":3,"active_band_mae_db":0.31238280591749384}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00015649665907765232,"log_rms_mae_db":2.4882392789673506,"rms_envelope_nrmse":0.2193561763515267,"envelope_correlation":0.9448645350694465,"rms_gain_error_db":-0.08722049045988918,"reference_mean_power":0.003302205038848876,"output_mean_power":0.0032365475115343},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.335462563779621e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006377316706455328,"output_power":0.0006536494207112421,"reference_active":true,"delta_db":0.10706895434355705},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.741968368909027e-05,"output_power":4.040996410575189e-05,"reference_active":true,"delta_db":-0.694701880299359},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7622617238454128e-06,"output_power":1.2994195404227052e-06,"reference_active":true,"delta_db":-1.3232101517609485}],"active_band_count":3,"active_band_mae_db":0.7083269954679547}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00014746435244218596,"log_rms_mae_db":4.181802521904483,"rms_envelope_nrmse":0.32575174507797056,"envelope_correlation":0.9088852747861127,"rms_gain_error_db":-0.18500107913025698,"reference_mean_power":0.001906578877331925,"output_mean_power":0.0018270678235383746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":8.936060821906154e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045087668804334914,"output_power":0.00043988443256190283,"reference_active":true,"delta_db":-0.10719188652303294},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.461610811295139e-05,"output_power":5.206652638803325e-05,"reference_active":true,"delta_db":-0.2076214479462668},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.197452814329269e-06,"output_power":7.174704387981982e-06,"reference_active":true,"delta_db":0.6359078312677398}],"active_band_count":3,"active_band_mae_db":0.3169070552456798}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00013246473438636124,"log_rms_mae_db":3.119815373056399,"rms_envelope_nrmse":0.22785523444812833,"envelope_correlation":0.9517465784891821,"rms_gain_error_db":-0.1832461354421588,"reference_mean_power":0.0019129188838057705,"output_mean_power":0.0018338843353442043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":8.520895123968316e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004268128119473447,"output_power":0.00041707556827815336,"reference_active":true,"delta_db":-0.10022697269665173},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5267927477999584e-05,"output_power":2.4500830813411485e-05,"reference_active":true,"delta_db":-0.13388810379477858},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.022115629444589e-07,"output_power":4.770331287593307e-07,"reference_active":true,"delta_db":-0.22338166955655517}],"active_band_count":3,"active_band_mae_db":0.1524989153493285}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.508036547105044e-05,"log_rms_mae_db":3.938598781807063,"rms_envelope_nrmse":0.38596931853656985,"envelope_correlation":0.8784250312256616,"rms_gain_error_db":0.04427846701649432,"reference_mean_power":0.0007310287661738598,"output_mean_power":0.0007385200894113519},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.375697567971487e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020778711952139394,"output_power":0.00021466162123148864,"reference_active":true,"delta_db":0.14135782392546248},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.101230284766395e-05,"output_power":3.9461814999450184e-05,"reference_active":true,"delta_db":-0.16737099443983333},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4179813437956632e-05,"output_power":1.425703402308726e-05,"reference_active":true,"delta_db":0.02358669081765133}],"active_band_count":3,"active_band_mae_db":0.11077183639431572}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.582207074769948e-05,"log_rms_mae_db":2.820095524083641,"rms_envelope_nrmse":0.2664007839082237,"envelope_correlation":0.9363203286402015,"rms_gain_error_db":0.04643583012188715,"reference_mean_power":0.0007338466075139545,"output_mean_power":0.0007417351733350415},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.9980296656839775e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00019557456921819077,"output_power":0.00020290238596134917,"reference_active":true,"delta_db":0.15974771679570496},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6985486820992017e-05,"output_power":1.4762609442422464e-05,"reference_active":true,"delta_db":-0.6091486842712791},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.922014736318651e-07,"output_power":9.342831076021928e-07,"reference_active":true,"delta_db":-0.2612137110658802}],"active_band_count":3,"active_band_mae_db":0.34337003737762145}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00011116745339746554,"log_rms_mae_db":4.011230890853036,"rms_envelope_nrmse":0.5432400880694536,"envelope_correlation":0.7436079090755113,"rms_gain_error_db":0.29351657053676394,"reference_mean_power":0.0001764719915242474,"output_mean_power":0.0001888110658560722},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.1170311171266687e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.030239739448222e-05,"output_power":4.7099683110755845e-05,"reference_active":true,"delta_db":0.6768710419625203},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.8894394185439035e-05,"output_power":2.6865490842539195e-05,"reference_active":true,"delta_db":-0.31618813559529946},{"low_hz":20.0,"high_hz":50.0,"reference_power":7.877164672472967e-06,"output_power":5.310217345998973e-06,"reference_active":true,"delta_db":-1.7125762748890545}],"active_band_count":3,"active_band_mae_db":0.9018784841489581}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":6.81572461842462e-05,"log_rms_mae_db":2.8463862397029853,"rms_envelope_nrmse":0.3441527658275825,"envelope_correlation":0.8813238097388668,"rms_gain_error_db":0.2933151925678713,"reference_mean_power":0.00017721561821746633,"output_mean_power":0.00018959789590412753},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.0001638026368733e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2510198337172984e-05,"output_power":4.7510129129361606e-05,"reference_active":true,"delta_db":0.4829307957179502},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6080590241633715e-05,"output_power":1.3573479876561164e-05,"reference_active":true,"delta_db":-0.7361078217672956},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.447224316204932e-07,"output_power":3.8149992306536967e-07,"reference_active":true,"delta_db":-2.2787832603285922}],"active_band_count":3,"active_band_mae_db":1.1659406259379461}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"hnm","status":"measured","mos":{"status":"measured","value":2.8032479286193848,"predicted_mos":2.8032479286193848,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":91600,"input_sample_rate":16000,"input_seconds":5.725,"input_float32_sha256":"f0b741a4c406cdc45e1e9b48482c6c9574add341147d484418b56e8336fe4b30","internal_samples":91600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4875762259998737},"started_at":"2026-10-13T04:40:29.853941+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.9375000192667358e-05,"native_input_samples":91600,"native_input_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","output_sha256":"c8e0c270ca97825a554867bf36c6ed65ca3742a91a2d3c5f65dab765d952536f","output_samples":91600,"length_delta_samples":0,"analysis_s":0.9529187360003561,"synthesis_s":0.4777681220002705,"total_s":1.4306868580006267,"rtf":0.24990163458526232,"output_peak":0.5358252616616784,"output_rms":0.06264882151336942,"finished_at":"2026-10-13T04:40:31.291721+00:00","evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"c8e0c270ca97825a554867bf36c6ed65ca3742a91a2d3c5f65dab765d952536f","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.588885182208207,"spectral_convergence":0.2597228845660181,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.326468803804366,"spectral_convergence":0.25296619711491797,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.595811112501038,"spectral_convergence":0.26104685642068376,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":6.265227784331508e-05,"delta_db":0.5154382537329449,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.0009781228059635454,"delta_db":-0.09612257410784476,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.001803637445220007,"delta_db":-0.05337073284318584,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.0004889667639150053,"delta_db":-0.2060260568609627,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.00031249575903285874,"delta_db":-0.27869048997280194,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.0002789997850392881,"delta_db":-0.042918383216078076,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1987610817889697,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09241860283307396,"pitch":null,"output_rms_dbfs":-24.061741882081208,"output_peak":0.5358252616616784,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5037216995045375},"common_prefix":{"samples":91600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.588885182208207,"spectral_convergence":0.2597228845660181,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.326468803804366,"spectral_convergence":0.25296619711491797,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.595811112501038,"spectral_convergence":0.26104685642068376,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":6.265227784331508e-05,"delta_db":0.5154382537329449,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.0009781228059635454,"delta_db":-0.09612257410784476,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.001803637445220007,"delta_db":-0.05337073284318584,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.0004889667639150053,"delta_db":-0.2060260568609627,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.00031249575903285874,"delta_db":-0.27869048997280194,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.0002789997850392881,"delta_db":-0.042918383216078076,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.1987610817889697,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.09241860283307396,"pitch":null,"output_rms_dbfs":-24.061741882081208,"output_peak":0.5358252616616784,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.5037216995045375},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91600,"output_samples":91600,"intelligibility":{"estoi":{"value":0.8771834304744682,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06331896918232865,"rms_dbfs":-23.96932327924813,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.14269480218138,"mean":-2.0079300393183678e-05,"zero_energy":false},"output":{"rms":0.06264882151336942,"rms_dbfs":-24.061741882081208,"peak":0.5358252616616784,"peak_dbfs":-5.419536305722648,"crest_factor_db":18.64220557635856,"mean":-5.134808095212769e-05,"zero_energy":false},"rms_difference_db":-0.09241860283307801},"full_scale":{"reference":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":5.024525490249213e-06,"output_rms_dbfs":-52.98904946185462,"output_zero_energy":false,"output_peak":0.058475331325461295,"output_energy_fraction":0.00012298621628930619,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":194,"both_estimated_voiced_frames":184,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9246231155778895,"absolute_difference_cents_median":1.1030218876325442,"absolute_difference_cents_p95":6.454096191679759,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00020825955187282386,"log_rms_mae_db":0.820454528428378,"rms_envelope_nrmse":0.0688790831220482,"envelope_correlation":0.9957091102864255,"rms_gain_error_db":-0.09240263367595224,"reference_mean_power":0.004012756910556895,"output_mean_power":0.003928281375788354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":2.2120307480210933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008889321025104637,"output_power":0.0008513487667279834,"reference_active":true,"delta_db":-0.18761079163458277},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022980725896418495,"output_power":0.0002112646691336636,"reference_active":true,"delta_db":-0.36536868861491933},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113193274777301e-05,"output_power":3.416847077413894e-05,"reference_active":true,"delta_db":-0.8055357486414263}],"active_band_count":3,"active_band_mae_db":0.45283840963030947}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.0001984160011662782,"log_rms_mae_db":0.42871756727140836,"rms_envelope_nrmse":0.03599188185672366,"envelope_correlation":0.9988122246491989,"rms_gain_error_db":-0.09242648128998297,"reference_mean_power":0.0040337408532874226,"output_mean_power":0.003948801886520849},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":2.0115882799069226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008024963383362849,"output_power":0.0007706861778192536,"reference_active":true,"delta_db":-0.1756548943741404},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001146343816948982,"output_power":0.00010312642285956394,"reference_active":true,"delta_db":-0.45944939195099127},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4945953602736586e-06,"output_power":2.6842328915479524e-06,"reference_active":true,"delta_db":-1.1457670223420833}],"active_band_count":3,"active_band_mae_db":0.5936237695557384}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00019054595180557928,"log_rms_mae_db":1.4171390099155734,"rms_envelope_nrmse":0.08294554433695024,"envelope_correlation":0.9948947128772927,"rms_gain_error_db":-0.08378081163894951,"reference_mean_power":0.0025757888112863028,"output_mean_power":0.0025265748641825543},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.7595886306573375e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000727930551645222,"output_power":0.0007091478499143572,"reference_active":true,"delta_db":-0.11353156857212875},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019010480993234813,"output_power":0.00016904378910070054,"reference_active":true,"delta_db":-0.5099388647251171},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6294378423488105e-05,"output_power":2.333788996530508e-05,"reference_active":true,"delta_db":-0.5180132104553723}],"active_band_count":3,"active_band_mae_db":0.3804945479175394}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.00018019065505981728,"log_rms_mae_db":0.7481023810809885,"rms_envelope_nrmse":0.046105148212996355,"envelope_correlation":0.9983845647512296,"rms_gain_error_db":-0.08377691071559451,"reference_mean_power":0.0025893611436675294,"output_mean_power":0.0025398901600827725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.6269545971301712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006780414214466183,"output_power":0.0006614422346955063,"reference_active":true,"delta_db":-0.10764303438825236},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.548437806715931e-05,"output_power":8.331019399995842e-05,"reference_active":true,"delta_db":-0.5923417789714354},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.932751011760402e-06,"output_power":2.246713700740134e-06,"reference_active":true,"delta_db":-1.1572745959175637}],"active_band_count":3,"active_band_mae_db":0.6190864697590838}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":1.4593138317215937,"rms_envelope_nrmse":0.11385699362727902,"envelope_correlation":0.9901664313305489,"rms_gain_error_db":-0.2763602093268191,"reference_mean_power":0.0007122971869413385,"output_mean_power":0.0006683827121678166},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":4.689927289528062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002336243034349734,"output_power":0.0002175710296854953,"reference_active":true,"delta_db":-0.3091695240892401},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7979021289898206e-05,"output_power":3.573265127295968e-05,"reference_active":true,"delta_db":-0.2647852792497343},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0421847785855855e-05,"output_power":9.45283484423397e-06,"reference_active":true,"delta_db":-0.4238265572067097}],"active_band_count":3,"active_band_mae_db":0.33259378684856133}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":0.8957063338518879,"rms_envelope_nrmse":0.0680329828011218,"envelope_correlation":0.9967685373699923,"rms_gain_error_db":-0.2764985795212644,"reference_mean_power":0.0007159674763911087,"output_mean_power":0.0006718053166274388},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":4.328159646639027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002170097016572237,"output_power":0.00020283053046329344,"reference_active":true,"delta_db":-0.2934582342575461},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7025087336003105e-05,"output_power":1.6778389070273894e-05,"reference_active":true,"delta_db":-0.06339087518540776},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.679451250427729e-07,"output_power":6.794503534951024e-07,"reference_active":true,"delta_db":-1.063345392867945}],"active_band_count":3,"active_band_mae_db":0.4733981674369663}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":1.4261409981429252,"rms_envelope_nrmse":0.1513619591434767,"envelope_correlation":0.9798506583357065,"rms_gain_error_db":-0.048080986872873384,"reference_mean_power":0.00032373528856046693,"output_mean_power":0.00032017096376052754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.248023532990147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.708460616175106e-05,"output_power":5.673419554347536e-05,"reference_active":true,"delta_db":-0.026741079055973626},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9595191589145463e-05,"output_power":1.9930211326793007e-05,"reference_active":true,"delta_db":0.07362389610422915},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.074204859780516e-06,"output_power":4.101836066468126e-06,"reference_active":true,"delta_db":-0.9238969750390441}],"active_band_count":3,"active_band_mae_db":0.34142065006641564}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":0.8442726553411004,"rms_envelope_nrmse":0.0855459178444499,"envelope_correlation":0.9928774070080326,"rms_gain_error_db":-0.0481573776369498,"reference_mean_power":0.0003253818281247374,"output_mean_power":0.00032179371464784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.1005868407421186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.449763632243476e-05,"output_power":5.443389998366674e-05,"reference_active":true,"delta_db":-0.005082154226604739},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.38575211763481e-06,"output_power":9.509890977832623e-06,"reference_active":true,"delta_db":0.057064580830939536},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7591737704275045e-07,"output_power":3.129854577612288e-07,"reference_active":true,"delta_db":-0.7956824234592084}],"active_band_count":3,"active_band_mae_db":0.2859430528389176}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.506066083908081,"predicted_mos":2.506066083908081,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":91600,"input_sample_rate":16000,"input_seconds":5.725,"input_float32_sha256":"816ee23babebcf081b3be06ac729f28b4c3c6550f082b13f567a25b89ac30cc5","internal_samples":91600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.466216605000227},"started_at":"2026-10-13T04:43:46.081591+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.0454999432549812e-05,"native_input_samples":91600,"native_input_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","output_sha256":"3e3e3006626d94a401718104198b3b191f87439ceab2d3ba18afffdecfde0b91","output_samples":91600,"length_delta_samples":0,"analysis_s":1.463399537999976,"synthesis_s":0.08676194399959059,"total_s":1.5501614819995666,"rtf":0.270770564541409,"output_peak":0.8770002141119679,"output_rms":0.07279527145199786,"finished_at":"2026-10-13T04:43:47.637052+00:00","evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"3e3e3006626d94a401718104198b3b191f87439ceab2d3ba18afffdecfde0b91","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.448113969743061,"spectral_convergence":0.580488288830341,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5530460502060155,"spectral_convergence":0.3449605559581496,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.973204980705509,"spectral_convergence":0.3890950595452272,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":5.484394689573235e-05,"delta_db":-0.06264338200770493,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.001565126974090547,"delta_db":1.9454393386134416,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.0023706945356810286,"delta_db":1.1338928230211434,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.0006307478321512201,"delta_db":0.899738206274132,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.00037274522262121997,"delta_db":0.4869890913526678,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.0003049930343303063,"delta_db":0.34394213717810407,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8121074964078656,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.2113866762230614,"pitch":null,"output_rms_dbfs":-22.757936603025072,"output_peak":0.8770002141119679,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.991455000218195},"common_prefix":{"samples":91600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.448113969743061,"spectral_convergence":0.580488288830341,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.5530460502060155,"spectral_convergence":0.3449605559581496,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.973204980705509,"spectral_convergence":0.3890950595452272,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":5.484394689573235e-05,"delta_db":-0.06264338200770493,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.001565126974090547,"delta_db":1.9454393386134416,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.0023706945356810286,"delta_db":1.1338928230211434,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.0006307478321512201,"delta_db":0.899738206274132,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.00037274522262121997,"delta_db":0.4869890913526678,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.0003049930343303063,"delta_db":0.34394213717810407,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8121074964078656,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.2113866762230614,"pitch":null,"output_rms_dbfs":-22.757936603025072,"output_peak":0.8770002141119679,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.991455000218195},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91600,"output_samples":91600,"intelligibility":{"estoi":{"value":0.876715713608104,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06331896918232865,"rms_dbfs":-23.96932327924813,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.14269480218138,"mean":-2.0079300393183678e-05,"zero_energy":false},"output":{"rms":0.07279527145199786,"rms_dbfs":-22.757936603025072,"peak":0.8770002141119679,"peak_dbfs":-1.1400060120945867,"crest_factor_db":21.617930590930484,"mean":1.4851763982820824e-08,"zero_energy":false},"rms_difference_db":1.2113866762230572},"full_scale":{"reference":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":2.5880085545769768e-06,"output_rms_dbfs":-55.870342924548694,"output_zero_energy":false,"output_peak":0.032882283987859084,"output_energy_fraction":4.691876438148895e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":198,"both_estimated_voiced_frames":167,"reference_voiced_output_unvoiced_frames":32,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.8391959798994975,"absolute_difference_cents_median":4.527678847855812,"absolute_difference_cents_p95":1190.9970964838378,"difference_over_600_cents_frames":11,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00020825955187282386,"log_rms_mae_db":1.4640708430748788,"rms_envelope_nrmse":0.21469430070934176,"envelope_correlation":0.9818651986913013,"rms_gain_error_db":1.2114199753048835,"reference_mean_power":0.004012756910556895,"output_mean_power":0.005303772032217885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":2.2120307480210933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008889321025104637,"output_power":0.0011695786340234463,"reference_active":true,"delta_db":1.1916083554581407},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022980725896418495,"output_power":0.00028175394408946783,"reference_active":true,"delta_db":0.8850626145687615},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113193274777301e-05,"output_power":5.527396711201694e-05,"reference_active":true,"delta_db":1.283415187507774}],"active_band_count":3,"active_band_mae_db":1.1200287191782254}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.0001984160011662782,"log_rms_mae_db":1.0157828873174002,"rms_envelope_nrmse":0.18395366255337287,"envelope_correlation":0.9908041432222556,"rms_gain_error_db":1.211471085244398,"reference_mean_power":0.0040337408532874226,"output_mean_power":0.005331569835149283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":2.0115882799069226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008024963383362849,"output_power":0.0010581679938402406,"reference_active":true,"delta_db":1.2011156192661965},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001146343816948982,"output_power":0.00013600366106278998,"reference_active":true,"delta_db":0.7423570637987178},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4945953602736586e-06,"output_power":4.117096708458502e-06,"reference_active":true,"delta_db":0.7119417249963191}],"active_band_count":3,"active_band_mae_db":0.885138136020411}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00019054595180557928,"log_rms_mae_db":2.0104092202666513,"rms_envelope_nrmse":0.2601349629397888,"envelope_correlation":0.9810028851758142,"rms_gain_error_db":1.518715138755139,"reference_mean_power":0.0025757888112863028,"output_mean_power":0.0036541112573314776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.7595886306573375e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000727930551645222,"output_power":0.0009759231139519594,"reference_active":true,"delta_db":1.2732565669735996},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019010480993234813,"output_power":0.0002514840647094024,"reference_active":true,"delta_db":1.2151736589302877},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6294378423488105e-05,"output_power":3.7701153931589735e-05,"reference_active":true,"delta_db":1.5649173408049726}],"active_band_count":3,"active_band_mae_db":1.3511158555696199}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.00018019065505981728,"log_rms_mae_db":1.4192049392613277,"rms_envelope_nrmse":0.2332102741303438,"envelope_correlation":0.988510260957299,"rms_gain_error_db":1.518720661717328,"reference_mean_power":0.0025893611436675294,"output_mean_power":0.003673370151929824},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.6269545971301712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006780414214466183,"output_power":0.0009074465443340632,"reference_active":true,"delta_db":1.2656482547479446},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.548437806715931e-05,"output_power":0.000123116266586989,"reference_active":true,"delta_db":1.1038311366616347},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.932751011760402e-06,"output_power":3.3301767025121698e-06,"reference_active":true,"delta_db":0.5519208493015184}],"active_band_count":3,"active_band_mae_db":0.9738000802370327}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":1.5926380478493638,"rms_envelope_nrmse":0.17865200434272085,"envelope_correlation":0.9835958523225729,"rms_gain_error_db":0.7911111921500525,"reference_mean_power":0.0007122971869413385,"output_mean_power":0.0008546186153563725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":4.689927289528062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002336243034349734,"output_power":0.00030673637015947013,"reference_active":true,"delta_db":1.1824725442494166},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7979021289898206e-05,"output_power":4.772553877559628e-05,"reference_active":true,"delta_db":0.9920707097940196},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0421847785855855e-05,"output_power":1.1807933135142443e-05,"reference_active":true,"delta_db":0.5422915932506853}],"active_band_count":3,"active_band_mae_db":0.9056116157647072}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":1.0291795803801065,"rms_envelope_nrmse":0.13245377008904188,"envelope_correlation":0.9935242501545197,"rms_gain_error_db":0.7911800955029944,"reference_mean_power":0.0007159674763911087,"output_mean_power":0.0008590358805675705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":4.328159646639027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002170097016572237,"output_power":0.00028649047559917986,"reference_active":true,"delta_db":1.2063103849083743},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7025087336003105e-05,"output_power":2.0985883232954056e-05,"reference_active":true,"delta_db":0.908379036333205},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.679451250427729e-07,"output_power":9.338883025124229e-07,"reference_active":true,"delta_db":0.31802667435515747}],"active_band_count":3,"active_band_mae_db":0.8109053651989123}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":1.5195076604578526,"rms_envelope_nrmse":0.202939026106979,"envelope_correlation":0.9668025600305048,"rms_gain_error_db":0.3618209918866082,"reference_mean_power":0.00032373528856046693,"output_mean_power":0.0003518618170996525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.248023532990147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.708460616175106e-05,"output_power":5.956298315061632e-05,"reference_active":true,"delta_db":0.18457431700947563},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9595191589145463e-05,"output_power":1.9501439850161602e-05,"reference_active":true,"delta_db":-0.020828362675653873},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.074204859780516e-06,"output_power":6.092652440073573e-06,"reference_active":true,"delta_db":0.7943840725945948}],"active_band_count":3,"active_band_mae_db":0.33326225075990806}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":0.9173751775205841,"rms_envelope_nrmse":0.11152158244363859,"envelope_correlation":0.9899284229033637,"rms_gain_error_db":0.3619656838219016,"reference_mean_power":0.0003253818281247374,"output_mean_power":0.00035366319273053974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.1005868407421186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.449763632243476e-05,"output_power":5.714769160192256e-05,"reference_active":true,"delta_db":0.20621025973815604},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.38575211763481e-06,"output_power":9.229132769460781e-06,"reference_active":true,"delta_db":-0.0730818634282293},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7591737704275045e-07,"output_power":4.4554373102564857e-07,"reference_active":true,"delta_db":0.7379793561095993}],"active_band_count":3,"active_band_mae_db":0.3390904930919949}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"facodec","status":"measured","mos":{"status":"measured","value":3.741208553314209,"predicted_mos":3.741208553314209,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":91600,"input_sample_rate":16000,"input_seconds":5.725,"input_float32_sha256":"843b5cda25fcbaf74b9060b438734cecb38319b29e6b5c8172ca4a8832615d14","internal_samples":91600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6317251569998916},"started_at":"2026-10-13T07:02:02.291072+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":7.166700015659444e-05,"native_input_samples":91600,"native_input_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","output_sha256":"e60307d621e064808d5cb69bf9b6cc106c44711c533c3cf9435c35c303e38aa5","output_samples":91600,"length_delta_samples":0,"analysis_s":1.5525958289999835,"synthesis_s":12.778243568998732,"total_s":14.330839397998716,"rtf":2.5032033882967193,"output_peak":0.5302847027778625,"output_rms":0.05722269085917937,"finished_at":"2026-10-13T07:02:16.628782+00:00","evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"e60307d621e064808d5cb69bf9b6cc106c44711c533c3cf9435c35c303e38aa5","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.162673175915214,"spectral_convergence":0.2827496130993018,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.281965369521509,"spectral_convergence":0.3126084322272854,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.375992093633122,"spectral_convergence":0.33930772090298045,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":5.9305744041122085e-05,"delta_db":0.27703719021127726,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.0008793117307750601,"delta_db":-0.5586277540958021,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.0013940477128980945,"delta_db":-1.1720867812865225,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.00043355669913630775,"delta_db":-0.7283604476886506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.0002858467372878528,"delta_db":-0.6657993733175039,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.00022236772502677357,"delta_db":-1.028239540206353,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7383585144676849,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8793111935495019,"pitch":null,"output_rms_dbfs":-24.848634472797634,"output_peak":0.5302847027778625,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.273543546356614},"common_prefix":{"samples":91600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.162673175915214,"spectral_convergence":0.2827496130993018,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.281965369521509,"spectral_convergence":0.3126084322272854,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.375992093633122,"spectral_convergence":0.33930772090298045,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":5.9305744041122085e-05,"delta_db":0.27703719021127726,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.0008793117307750601,"delta_db":-0.5586277540958021,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.0013940477128980945,"delta_db":-1.1720867812865225,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.00043355669913630775,"delta_db":-0.7283604476886506,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.0002858467372878528,"delta_db":-0.6657993733175039,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.00022236772502677357,"delta_db":-1.028239540206353,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7383585144676849,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8793111935495019,"pitch":null,"output_rms_dbfs":-24.848634472797634,"output_peak":0.5302847027778625,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.273543546356614},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91600,"output_samples":91600,"intelligibility":{"estoi":{"value":0.9096080014466251,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06331896918232865,"rms_dbfs":-23.96932327924813,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.14269480218138,"mean":-2.0079300393183678e-05,"zero_energy":false},"output":{"rms":0.05722269085917937,"rms_dbfs":-24.848634472797634,"peak":0.5302847027778625,"peak_dbfs":-5.50981801749995,"crest_factor_db":19.338816455297685,"mean":2.884307491875435e-06,"zero_energy":false},"rms_difference_db":-0.8793111935495048},"full_scale":{"reference":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":1.7540713230164838e-06,"output_rms_dbfs":-57.55952751584154,"output_zero_energy":false,"output_peak":0.01872188225388527,"output_energy_fraction":5.146333116359145e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":221,"both_estimated_voiced_frames":190,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":31,"paired_voiced_fraction":0.9547738693467337,"absolute_difference_cents_median":3.3176823357568415,"absolute_difference_cents_p95":11.424377579228617,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00020825955187282386,"log_rms_mae_db":1.2633214643497799,"rms_envelope_nrmse":0.13217939703648166,"envelope_correlation":0.991852877059855,"rms_gain_error_db":-0.879294194857395,"reference_mean_power":0.004012756910556895,"output_mean_power":0.003277279126099921},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":2.2120307480210933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008889321025104637,"output_power":0.0006887583121811522,"reference_active":true,"delta_db":-1.1080173729419593},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022980725896418495,"output_power":0.00018015298415849604,"reference_active":true,"delta_db":-1.0572228228485623},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113193274777301e-05,"output_power":3.78800029934723e-05,"reference_active":true,"delta_db":-0.3576911239414483}],"active_band_count":3,"active_band_mae_db":0.8409771065773232}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.0001984160011662782,"log_rms_mae_db":1.118369256510381,"rms_envelope_nrmse":0.12426685564111063,"envelope_correlation":0.9934631511117675,"rms_gain_error_db":-0.8792338966403141,"reference_mean_power":0.0040337408532874226,"output_mean_power":0.0032944627693890388},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":2.0115882799069226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008024963383362849,"output_power":0.0006197504362822879,"reference_active":true,"delta_db":-1.1222621829662915},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001146343816948982,"output_power":9.033366928186233e-05,"reference_active":true,"delta_db":-1.0346524157030292},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4945953602736586e-06,"output_power":2.98291798501217e-06,"reference_active":true,"delta_db":-0.6875558331971966}],"active_band_count":3,"active_band_mae_db":0.9481568106221725}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00019054595180557928,"log_rms_mae_db":1.4178426491383542,"rms_envelope_nrmse":0.14539409827639796,"envelope_correlation":0.9922917000155231,"rms_gain_error_db":-1.0006014838771742,"reference_mean_power":0.0025757888112863028,"output_mean_power":0.0020457384316832734},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.7595886306573375e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000727930551645222,"output_power":0.000559623084678957,"reference_active":true,"delta_db":-1.1419432632793252},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019010480993234813,"output_power":0.00014165614255874323,"reference_active":true,"delta_db":-1.2775769395435954},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6294378423488105e-05,"output_power":2.291965491487057e-05,"reference_active":true,"delta_db":-0.5965483444453086}],"active_band_count":3,"active_band_mae_db":1.005356182422743}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.00018019065505981728,"log_rms_mae_db":1.1610558545331569,"rms_envelope_nrmse":0.13886137423125644,"envelope_correlation":0.9934877572920073,"rms_gain_error_db":-1.0006008490236014,"reference_mean_power":0.0025893611436675294,"output_mean_power":0.002056518126112712},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.6269545971301712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006780414214466183,"output_power":0.0005211803590261929,"reference_active":true,"delta_db":-1.1426818494280604},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.548437806715931e-05,"output_power":7.02953299747965e-05,"reference_active":true,"delta_db":-1.3300584978384777},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.932751011760402e-06,"output_power":2.477010868662668e-06,"reference_active":true,"delta_db":-0.7334728109449558}],"active_band_count":3,"active_band_mae_db":1.0687377194038312}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":1.618538648404721,"rms_envelope_nrmse":0.1647221061613278,"envelope_correlation":0.9831562897028582,"rms_gain_error_db":-0.8002978946740622,"reference_mean_power":0.0007122971869413385,"output_mean_power":0.0005924223570639306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":4.689927289528062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002336243034349734,"output_power":0.00019947191735813702,"reference_active":true,"delta_db":-0.6863625730963886},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7979021289898206e-05,"output_power":3.615536821535966e-05,"reference_active":true,"delta_db":-0.21370980025875105},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0421847785855855e-05,"output_power":1.1579177308893236e-05,"reference_active":true,"delta_db":0.4573297833992807}],"active_band_count":3,"active_band_mae_db":0.45246738558480676}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":1.3493544704195166,"rms_envelope_nrmse":0.14382600258379785,"envelope_correlation":0.9878043369356072,"rms_gain_error_db":-0.8000760683859904,"reference_mean_power":0.0007159674763911087,"output_mean_power":0.0005955053775477649},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":4.328159646639027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002170097016572237,"output_power":0.00018762185572585957,"reference_active":true,"delta_db":-0.6319572273716113},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7025087336003105e-05,"output_power":1.6477421605891043e-05,"reference_active":true,"delta_db":-0.14200094364514285},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.679451250427729e-07,"output_power":8.459252074363977e-07,"reference_active":true,"delta_db":-0.1116030166736428}],"active_band_count":3,"active_band_mae_db":0.29518706256346566}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":1.8705603949346126,"rms_envelope_nrmse":0.20752038790783972,"envelope_correlation":0.9650490060276893,"rms_gain_error_db":-0.7143760508266582,"reference_mean_power":0.00032373528856046693,"output_mean_power":0.0002746328203312248},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.248023532990147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.708460616175106e-05,"output_power":5.2713311602896414e-05,"reference_active":true,"delta_db":-0.34598708406264633},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9595191589145463e-05,"output_power":2.0534542741717024e-05,"reference_active":true,"delta_db":0.20335522445538667},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.074204859780516e-06,"output_power":5.652272945128948e-06,"reference_active":true,"delta_db":0.46855128605518637}],"active_band_count":3,"active_band_mae_db":0.33929786485773983}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":1.620167173613123,"rms_envelope_nrmse":0.1770539816601998,"envelope_correlation":0.9730383754145042,"rms_gain_error_db":-0.7141767467694751,"reference_mean_power":0.0003253818281247374,"output_mean_power":0.00027604228911407507},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.1005868407421186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.449763632243476e-05,"output_power":5.0912612353706475e-05,"reference_active":true,"delta_db":-0.29552284922300937},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.38575211763481e-06,"output_power":1.030834484435249e-05,"reference_active":true,"delta_db":0.4071985844994935},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7591737704275045e-07,"output_power":3.969520461550751e-07,"reference_active":true,"delta_db":0.23645643186081444}],"active_band_count":3,"active_band_mae_db":0.31305928852777243}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.9292290210723877,"predicted_mos":3.9292290210723877,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":252928,"input_sample_rate":44100,"input_seconds":5.735328798185941,"input_float32_sha256":"d8b2b8b82ddc77876be6a69bd6af6f383458639cc4df486aec79739357c1169a","internal_samples":91766,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7867396080000617},"started_at":"2026-10-13T07:32:41.972139+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.012249879000592045,"native_input_samples":252473,"native_input_sha256":"0a84ba19b28acff202591bdae398f32f98a5291c0cc05d2cc7dffcdd903408f5","output_sha256":"e9ae0963548c80a85eacd3a515437b08087cf74e3f9194f0bfd25a9973e63e6c","output_samples":252928,"length_delta_samples":455,"analysis_s":0.31983961399964755,"synthesis_s":35.70182249999925,"total_s":36.0216621139989,"rtf":6.291980921632616,"output_peak":0.597723662853241,"output_rms":0.057395749053466776,"finished_at":"2026-10-13T07:33:18.033532+00:00","evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"e9ae0963548c80a85eacd3a515437b08087cf74e3f9194f0bfd25a9973e63e6c","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":91600,"output_samples":91766,"length_delta_samples":166,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.816001533380682,"output_peak":0.5698937459557482},"common_prefix":{"samples":91600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9981910511518427,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.937131403751989,"spectral_convergence":0.46435084283049166,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.930843438811435,"spectral_convergence":0.44478840815392345,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.88441016823359,"spectral_convergence":0.48305426339711116,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":0.00011080297210709488,"delta_db":2.991583699323815,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.000694401284021229,"delta_db":-1.5839512724289713,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.0015564143243492925,"delta_db":-0.6936109722814633,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.00046690780738579187,"delta_db":-0.4065080988182396,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.00029186299759305714,"delta_db":-0.5753413818619119,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.0001847083860588403,"delta_db":-1.8340909342658052,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3475143931633677,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8388336074928342,"pitch":null,"output_rms_dbfs":-24.808156886740967,"output_peak":0.5698937459557482,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.917461670265672},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91600,"output_samples":91600,"intelligibility":{"estoi":{"value":0.7073410954470516,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06331896918232865,"rms_dbfs":-23.96932327924813,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.14269480218138,"mean":-2.0079300393183678e-05,"zero_energy":false},"output":{"rms":0.057489979748781485,"rms_dbfs":-24.808156886740967,"peak":0.5698937459557482,"peak_dbfs":-4.884122179414822,"crest_factor_db":19.924034707326143,"mean":-0.00014959741052638346,"zero_energy":false},"rms_difference_db":-0.8388336074928375},"full_scale":{"reference":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":1.329355298461965e-06,"output_rms_dbfs":-58.76358929110624,"output_zero_energy":false,"output_peak":0.008199728374933204,"output_energy_fraction":3.8640608599405975e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":199,"both_estimated_voiced_frames":173,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8693467336683417,"absolute_difference_cents_median":5.932813521755986,"absolute_difference_cents_p95":40.0432660611578,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00020825955187282386,"log_rms_mae_db":1.568162793605866,"rms_envelope_nrmse":0.1273552701318411,"envelope_correlation":0.992565766042294,"rms_gain_error_db":-0.8388141790807,"reference_mean_power":0.004012756910556895,"output_mean_power":0.0033079690185543963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":2.2120307480210933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008889321025104637,"output_power":0.0007853754918768342,"reference_active":true,"delta_db":-0.5379124568589242},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022980725896418495,"output_power":0.0002155135619320423,"reference_active":true,"delta_db":-0.27889137878148695},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113193274777301e-05,"output_power":3.624993159555372e-05,"reference_active":true,"delta_db":-0.5487192566135142}],"active_band_count":3,"active_band_mae_db":0.4551743640846418}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.0001984160011662782,"log_rms_mae_db":1.3275686272955978,"rms_envelope_nrmse":0.10903436463134407,"envelope_correlation":0.9966101054797878,"rms_gain_error_db":-0.838799174657851,"reference_mean_power":0.0040337408532874226,"output_mean_power":0.0033252788966351117},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":2.0115882799069226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008024963383362849,"output_power":0.0007135096046548758,"reference_active":true,"delta_db":-0.5104323587430833},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001146343816948982,"output_power":0.00011037228540877298,"reference_active":true,"delta_db":-0.16454857554013116},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4945953602736586e-06,"output_power":3.018851277107598e-06,"reference_active":true,"delta_db":-0.6355517781916807}],"active_band_count":3,"active_band_mae_db":0.43684423749163176}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00019054595180557928,"log_rms_mae_db":2.5590185054808416,"rms_envelope_nrmse":0.17299413704068578,"envelope_correlation":0.9858288348503748,"rms_gain_error_db":-1.0589668995707116,"reference_mean_power":0.0025757888112863028,"output_mean_power":0.0020184293954806},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.7595886306573375e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000727930551645222,"output_power":0.0006193453104025793,"reference_active":true,"delta_db":-0.7015709381574058},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019010480993234813,"output_power":0.00017482054666760076,"reference_active":true,"delta_db":-0.3640063134566884},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6294378423488105e-05,"output_power":2.3712349509097912e-05,"reference_active":true,"delta_db":-0.4488832127562401}],"active_band_count":3,"active_band_mae_db":0.5048201547901114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.00018019065505981728,"log_rms_mae_db":1.9332219045002443,"rms_envelope_nrmse":0.15977445424082154,"envelope_correlation":0.9888458405026694,"rms_gain_error_db":-1.0589824668871688,"reference_mean_power":0.0025893611436675294,"output_mean_power":0.002029057619478019},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.6269545971301712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006780414214466183,"output_power":0.0005781281778388393,"reference_active":true,"delta_db":-0.6923208835342494},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.548437806715931e-05,"output_power":8.891154291454328e-05,"reference_active":true,"delta_db":-0.30974176904368633},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.932751011760402e-06,"output_power":2.4916575362181444e-06,"reference_active":true,"delta_db":-0.7078684244009749}],"active_band_count":3,"active_band_mae_db":0.5699770256596368}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":2.838231548313039,"rms_envelope_nrmse":0.2608700999843699,"envelope_correlation":0.9480619999920055,"rms_gain_error_db":-0.55983439387004,"reference_mean_power":0.0007122971869413385,"output_mean_power":0.0006261491419987297},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":4.689927289528062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002336243034349734,"output_power":0.00016861697170261633,"reference_active":true,"delta_db":-1.4161673424349357},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7979021289898206e-05,"output_power":3.1734579508176504e-05,"reference_active":true,"delta_db":-0.7801102071457726},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0421847785855855e-05,"output_power":1.0857777235846485e-05,"reference_active":true,"delta_db":0.1779620130689733}],"active_band_count":3,"active_band_mae_db":0.7914131875498939}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":2.395304747808788,"rms_envelope_nrmse":0.2265173658442988,"envelope_correlation":0.9587389951338506,"rms_gain_error_db":-0.5598519114078311,"reference_mean_power":0.0007159674763911087,"output_mean_power":0.0006293729934752407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":4.328159646639027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002170097016572237,"output_power":0.00015758116901203715,"reference_active":true,"delta_db":-1.3897483193650815},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7025087336003105e-05,"output_power":1.555189095478231e-05,"reference_active":true,"delta_db":-0.39306146002894626},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.679451250427729e-07,"output_power":8.125908132658719e-07,"reference_active":true,"delta_db":-0.2862036010488238}],"active_band_count":3,"active_band_mae_db":0.6896711268142838}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":3.594623171350773,"rms_envelope_nrmse":0.3251290409015357,"envelope_correlation":0.919013134249755,"rms_gain_error_db":-1.5337343702410011,"reference_mean_power":0.00032373528856046693,"output_mean_power":0.00022741368993118262},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.248023532990147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.708460616175106e-05,"output_power":3.92488055715989e-05,"reference_active":true,"delta_db":-1.6269256437949866},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9595191589145463e-05,"output_power":1.8059278161058024e-05,"reference_active":true,"delta_db":-0.3544912674298272},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.074204859780516e-06,"output_power":4.50306655408379e-06,"reference_active":true,"delta_db":-0.5185963115406312}],"active_band_count":3,"active_band_mae_db":0.8333377409218149}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":3.27761937604475,"rms_envelope_nrmse":0.29367655070956683,"envelope_correlation":0.9314140873513663,"rms_gain_error_db":-1.5334523796237978,"reference_mean_power":0.0003253818281247374,"output_mean_power":0.00022858517305524468},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.1005868407421186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.449763632243476e-05,"output_power":3.901715790949883e-05,"reference_active":true,"delta_db":-1.451220351142875},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.38575211763481e-06,"output_power":9.133453043432917e-06,"reference_active":true,"delta_db":-0.11834079758023668},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7591737704275045e-07,"output_power":3.010205238895093e-07,"reference_active":true,"delta_db":-0.9649629449430701}],"active_band_count":3,"active_band_mae_db":0.8448413645553939}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"golf","status":"measured","mos":{"status":"measured","value":3.130446195602417,"predicted_mos":3.130446195602417,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":137280,"input_sample_rate":24000,"input_seconds":5.72,"input_float32_sha256":"eec2e3c7efb6c4006afc833288f87d4cd6cae9920a51338d47f38770980b38fa","internal_samples":91520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5067090519987687},"started_at":"2026-10-13T05:03:09.943386+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0030258150000008754,"native_input_samples":137400,"native_input_sha256":"ae5e893bde4e89f71a715d9e2925da3800e4a6a4bedbbfe6c3d457aed76127fb","output_sha256":"7b56bc4cb50d4f072fcc70b210b40ae577dbd2304517b4108e72df0fe59a601c","output_samples":137280,"length_delta_samples":-120,"analysis_s":0.5915494739992937,"synthesis_s":0.14847688600002584,"total_s":0.7400263599993195,"rtf":0.12926224628809077,"output_peak":0.5108828544616699,"output_rms":0.05230820061986079,"finished_at":"2026-10-13T05:03:10.695391+00:00","analysis_stages":{"external_f0_s":0.10528410900042218},"evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"7b56bc4cb50d4f072fcc70b210b40ae577dbd2304517b4108e72df0fe59a601c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":91600,"output_samples":91520,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.6217868483802,"output_peak":0.5188796032397238},"common_prefix":{"samples":91520,"sample_rate":16000,"reference_coverage":0.9991266375545852,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91520,"output_samples":91520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.688318057782564,"spectral_convergence":0.5463274880747944,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.727922355810512,"spectral_convergence":0.3751834716324937,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.815505380702471,"spectral_convergence":0.41319622912816506,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.5688863782586656e-05,"output_power":8.763046543745072e-05,"delta_db":1.968867613070298,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010008097905811735,"output_power":0.0006206068363557519,"delta_db":-2.07534989972932,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018276094726123104,"output_power":0.0013613011488473928,"delta_db":-1.2792918933081932,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005131385726906514,"output_power":0.00033756182397931236,"delta_db":-1.818813369698606,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003334859835763636,"output_power":0.00020474442298682856,"delta_db":-2.1186550434003926,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00028200277539059966,"output_power":0.0001286017210996655,"delta_db":-3.410066016751268,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.111840639326346,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.65619214469938,"pitch":null,"output_rms_dbfs":-25.6217868483802,"output_peak":0.5188796032397238,"reference_rms_dbfs":-23.96559470368082,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.0772485980985165},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91520,"output_samples":91520,"intelligibility":{"estoi":{"value":0.8385411336738611,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06334615583233157,"rms_dbfs":-23.965594703680818,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.13896622661407,"mean":-1.967236712262347e-05,"zero_energy":false},"output":{"rms":0.05234927333563751,"rms_dbfs":-25.6217868483802,"peak":0.5188796032397238,"peak_dbfs":-5.6986680150680735,"crest_factor_db":19.923118833312124,"mean":0.00045112753836562856,"zero_energy":false},"rms_difference_db":-1.6561921446993821},"full_scale":{"reference":{"sample_count":91520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":9.96758983883339e-07,"output_rms_dbfs":-60.014098413105366,"output_zero_energy":false,"output_peak":0.005018512980892671,"output_energy_fraction":3.4973210690185485e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":212,"both_estimated_voiced_frames":171,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":41,"paired_voiced_fraction":0.8592964824120602,"absolute_difference_cents_median":5.249250878156233,"absolute_difference_cents_p95":18.94420059051953,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1143,"rms_floor":0.00020825955187282386,"log_rms_mae_db":2.538472948026387,"rms_envelope_nrmse":0.2409496889548505,"envelope_correlation":0.9694410575563288,"rms_gain_error_db":-1.656200607167826,"reference_mean_power":0.004016221996345041,"output_mean_power":0.0027428221605440848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1143,"frequency_resolution_hz":0.17497812773403326,"power_floor":2.2109728098334867e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008880247805074585,"output_power":0.0005892348296472878,"reference_active":true,"delta_db":-1.781366749624249},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022958135929501623,"output_power":0.00019111683435590627,"reference_active":true,"delta_db":-0.7963767972526274},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.117615282395463e-05,"output_power":3.452366559141796e-05,"reference_active":true,"delta_db":-0.7652886588845164}],"active_band_count":3,"active_band_mae_db":1.1143440685871309}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1137,"rms_floor":0.0001984160011662782,"log_rms_mae_db":2.2762251723918157,"rms_envelope_nrmse":0.21653851700331342,"envelope_correlation":0.9800530771067698,"rms_gain_error_db":-1.656149893040212,"reference_mean_power":0.004037253394471679,"output_mean_power":0.0027572174540620737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1137,"frequency_resolution_hz":0.17590149516270887,"power_floor":2.0106173143931753e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008016832980131587,"output_power":0.0005352116706186562,"reference_active":true,"delta_db":-1.7547726083321038},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001144583993185857,"output_power":9.79157698420375e-05,"reference_active":true,"delta_db":-0.6779502526706247},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.491886267800856e-06,"output_power":2.3422584839448617e-06,"reference_active":true,"delta_db":-1.7342526941846461}],"active_band_count":3,"active_band_mae_db":1.388991851729125}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1143,"rms_floor":0.00019054595180557928,"log_rms_mae_db":2.2606333749546312,"rms_envelope_nrmse":0.26091433850700124,"envelope_correlation":0.9627843873796446,"rms_gain_error_db":-1.5465577764168679,"reference_mean_power":0.0025780395305431353,"output_mean_power":0.001805650925025815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1143,"frequency_resolution_hz":0.17497812773403326,"power_floor":1.7578172846113887e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000726263230589433,"output_power":0.0004681426949018385,"reference_active":true,"delta_db":-1.9071580630614795},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001900460848169589,"output_power":0.00014341712303784582,"reference_active":true,"delta_db":-1.2225792076955402},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6315383566406435e-05,"output_power":2.2770178732372243e-05,"reference_active":true,"delta_db":-0.6284326498434001}],"active_band_count":3,"active_band_mae_db":1.2527233068668067}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1137,"rms_floor":0.00018019065505981728,"log_rms_mae_db":1.8045300665709614,"rms_envelope_nrmse":0.23526426266633824,"envelope_correlation":0.9724553190287591,"rms_gain_error_db":-1.5465577994018522,"reference_mean_power":0.002591636184683341,"output_mean_power":0.0018151739699861342},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1137,"frequency_resolution_hz":0.17590149516270887,"power_floor":1.6252267768772051e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006765085572318484,"output_power":0.0004364973849130412,"reference_active":true,"delta_db":-1.9029164824733111},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.543780616512371e-05,"output_power":7.240567824209062e-05,"reference_active":true,"delta_db":-1.1994782158392918},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9291934321489077e-06,"output_power":1.9535684886446745e-06,"reference_active":true,"delta_db":-1.7591941026720839}],"active_band_count":3,"active_band_mae_db":1.620529600328229}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1143,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":2.4448646518893944,"rms_envelope_nrmse":0.25683145368169813,"envelope_correlation":0.9734162819871052,"rms_gain_error_db":-1.910681485479646,"reference_mean_power":0.0007128988798691495,"output_mean_power":0.0004591554924948525},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1143,"frequency_resolution_hz":0.17497812773403326,"power_floor":4.690023876861038e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00023356888080030056,"output_power":0.00014584044821708094,"reference_active":true,"delta_db":-2.0453698933740188},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.795965494247885e-05,"output_power":2.92251408862066e-05,"reference_active":true,"delta_db":-1.1356564253684267},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0432976762534725e-05,"output_power":9.662676887706645e-06,"reference_active":true,"delta_db":-0.333107827577962}],"active_band_count":3,"active_band_mae_db":1.1713780487734693}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1137,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":2.104289362834387,"rms_envelope_nrmse":0.2253553438857744,"envelope_correlation":0.9874327958218836,"rms_gain_error_db":-1.9106356936645577,"reference_mean_power":0.0007165801593001434,"output_mean_power":0.0004615313538216974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1137,"frequency_resolution_hz":0.17590149516270887,"power_floor":4.3282262693304325e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021694782225768785,"output_power":0.000137485329259319,"reference_active":true,"delta_db":-1.9809893691737266},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.701963850529053e-05,"output_power":1.4329073162579894e-05,"reference_active":true,"delta_db":-0.7473223135378821},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.672437963065236e-07,"output_power":6.41821987330149e-07,"reference_active":true,"delta_db":-1.307266109329559}],"active_band_count":3,"active_band_mae_db":1.345192597347056}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1143,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":4.70645831637814,"rms_envelope_nrmse":0.35550612123162295,"envelope_correlation":0.969677253808161,"rms_gain_error_db":-3.3224525532169262,"reference_mean_power":0.0003240005565082866,"output_mean_power":0.00015076498916640138},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1143,"frequency_resolution_hz":0.17497812773403326,"power_floor":1.247472722309492e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.704454049685671e-05,"output_power":3.1281455842495086e-05,"reference_active":true,"delta_db":-2.6092712919546894},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.956236168957676e-05,"output_power":1.2266562875186366e-05,"reference_active":true,"delta_db":-2.0269839507998917},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.078311558070462e-06,"output_power":4.000780910532989e-06,"reference_active":true,"delta_db":-1.0357457229765468}],"active_band_count":3,"active_band_mae_db":1.8906669885770426}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91520,"uncovered_tail_samples":0,"metrics":{"frame_count":1137,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":4.4192509977490415,"rms_envelope_nrmse":0.3365277922965512,"envelope_correlation":0.9840444569235502,"rms_gain_error_db":-3.32220589192864,"reference_mean_power":0.0003256551018249911,"output_mean_power":0.00015154349435494988},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1137,"frequency_resolution_hz":0.17590149516270887,"power_floor":1.100227095995457e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.445635789021501e-05,"output_power":3.1192533459577045e-05,"reference_active":true,"delta_db":-2.4199794219795407},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.355260517851848e-06,"output_power":6.014894415527954e-06,"reference_active":true,"delta_db":-1.9182787768310496},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.760182643251983e-07,"output_power":3.060641881710873e-07,"reference_active":true,"delta_db":-0.8939642358917278}],"active_band_count":3,"active_band_mae_db":1.7440741449007728}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.0566208362579346,"predicted_mos":3.0566208362579346,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":137520,"input_sample_rate":24000,"input_seconds":5.73,"input_float32_sha256":"d2a6accd0f7f7c585288569ed41fd9ff544c78aec45e55b62f472040ca6dc568","internal_samples":91680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0280763459995796},"started_at":"2026-10-13T07:47:18.671522+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0035578450006141793,"native_input_samples":137400,"native_input_sha256":"ae5e893bde4e89f71a715d9e2925da3800e4a6a4bedbbfe6c3d457aed76127fb","output_sha256":"59ccd753cc4bace4cde466c8622c13c79dd1f65d2482b7b23df50e5c35011968","output_samples":137520,"length_delta_samples":120,"analysis_s":2.432524135001586,"synthesis_s":18.462766815999203,"total_s":20.89529095100079,"rtf":3.6498324805241555,"output_peak":0.5649623870849609,"output_rms":0.06049804830502786,"finished_at":"2026-10-13T07:47:39.590115+00:00","analysis_stages":{"external_f0_s":1.6514439420025155},"evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"59ccd753cc4bace4cde466c8622c13c79dd1f65d2482b7b23df50e5c35011968","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":91600,"output_samples":91680,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.358496635130095,"output_peak":0.5653406271150989},"common_prefix":{"samples":91600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991273996509599,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.383484881933025,"spectral_convergence":0.5835192739283898,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.219583113568066,"spectral_convergence":0.32387473637599934,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.2148128804344225,"spectral_convergence":0.35387677981527843,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":6.331103103262623e-05,"delta_db":0.5608634630043231,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.0009180941782427754,"delta_db":-0.3711840898690438,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.0016687353539314755,"delta_db":-0.3909884934437823,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.0004925251327281418,"delta_db":-0.1745354920116793,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.00030652881755368374,"delta_db":-0.3624186688327884,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.00021955751037059417,"delta_db":-1.083474094446152,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4905773836012948,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3854931390579833,"pitch":null,"output_rms_dbfs":-24.354816418306115,"output_peak":0.5653406271150989,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.605960291978505},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91600,"output_samples":91600,"intelligibility":{"estoi":{"value":0.8791499301273848,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06331896918232865,"rms_dbfs":-23.96932327924813,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.14269480218138,"mean":-2.0079300393183678e-05,"zero_energy":false},"output":{"rms":0.06057022390464886,"rms_dbfs":-24.354816418306115,"peak":0.5653406271150989,"peak_dbfs":-4.953796073641881,"crest_factor_db":19.401020344664236,"mean":-6.728618442002963e-05,"zero_energy":false},"rms_difference_db":-0.385493139057985},"full_scale":{"reference":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":1.4662472525832425e-05,"output_rms_dbfs":-48.3379278864559,"output_zero_energy":false,"output_peak":0.17019596662732406,"output_energy_fraction":0.000383951219801171,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":175,"both_estimated_voiced_frames":158,"reference_voiced_output_unvoiced_frames":41,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.7939698492462312,"absolute_difference_cents_median":4.445017524252253,"absolute_difference_cents_p95":16.44311334321137,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00020825955187282386,"log_rms_mae_db":1.925141667909748,"rms_envelope_nrmse":0.1755975791275194,"envelope_correlation":0.9722634667329282,"rms_gain_error_db":-0.3854809540990195,"reference_mean_power":0.004012756910556895,"output_mean_power":0.0036719330649969207},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":2.2120307480210933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008889321025104637,"output_power":0.0008176269866660162,"reference_active":true,"delta_db":-0.3631337301859378},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022980725896418495,"output_power":0.00020775278923475612,"reference_active":true,"delta_db":-0.4381687948894294},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113193274777301e-05,"output_power":3.209375128018282e-05,"reference_active":true,"delta_db":-1.0775863444468166}],"active_band_count":3,"active_band_mae_db":0.6262962898407279}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.0001984160011662782,"log_rms_mae_db":1.2831970116772573,"rms_envelope_nrmse":0.10762596289197322,"envelope_correlation":0.9899046636459348,"rms_gain_error_db":-0.3854526811366181,"reference_mean_power":0.0040337408532874226,"output_mean_power":0.003691158764485655},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":2.0115882799069226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008024963383362849,"output_power":0.000736884230880128,"reference_active":true,"delta_db":-0.3704379662471553},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001146343816948982,"output_power":0.00010469268594600196,"reference_active":true,"delta_db":-0.3939855085652142},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4945953602736586e-06,"output_power":2.622866475808031e-06,"reference_active":true,"delta_db":-1.2462071376435664}],"active_band_count":3,"active_band_mae_db":0.6702102041519785}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00019054595180557928,"log_rms_mae_db":2.3153897866188973,"rms_envelope_nrmse":0.17548442503697087,"envelope_correlation":0.9773701899252166,"rms_gain_error_db":-0.3779507982613227,"reference_mean_power":0.0025757888112863028,"output_mean_power":0.0023611043022776505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.7595886306573375e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000727930551645222,"output_power":0.0006541425447178642,"reference_active":true,"delta_db":-0.46417551250428046},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019010480993234813,"output_power":0.00017133437247053197,"reference_active":true,"delta_db":-0.4514860703100137},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6294378423488105e-05,"output_power":2.093618148997213e-05,"reference_active":true,"delta_db":-0.9896543449544377}],"active_band_count":3,"active_band_mae_db":0.635105309256244}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.00018019065505981728,"log_rms_mae_db":1.6190285852486364,"rms_envelope_nrmse":0.11222793352231356,"envelope_correlation":0.9910163343747851,"rms_gain_error_db":-0.3779488255621321,"reference_mean_power":0.0025893611436675294,"output_mean_power":0.0023735464983504224},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.6269545971301712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006780414214466183,"output_power":0.0006066122561404195,"reference_active":true,"delta_db":-0.48345045007075504},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.548437806715931e-05,"output_power":8.600745585504434e-05,"reference_active":true,"delta_db":-0.45396222485274773},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.932751011760402e-06,"output_power":2.2756532982546417e-06,"reference_active":true,"delta_db":-1.1016909642784458}],"active_band_count":3,"active_band_mae_db":0.6797012130673162}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":2.238054886467025,"rms_envelope_nrmse":0.20479653517676105,"envelope_correlation":0.9663225761778561,"rms_gain_error_db":-0.24079123184521334,"reference_mean_power":0.0007122971869413385,"output_mean_power":0.0006738792844074696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":4.689927289528062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002336243034349734,"output_power":0.00021818258387168197,"reference_active":true,"delta_db":-0.2969793884456844},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7979021289898206e-05,"output_power":3.50564570807461e-05,"reference_active":true,"delta_db":-0.34775746115883993},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0421847785855855e-05,"output_power":1.004963335328127e-05,"reference_active":true,"delta_db":-0.1579450845572029}],"active_band_count":3,"active_band_mae_db":0.2675606447205757}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":1.4351067324665123,"rms_envelope_nrmse":0.11570348201967444,"envelope_correlation":0.9888348107785234,"rms_gain_error_db":-0.24066001560727968,"reference_mean_power":0.0007159674763911087,"output_mean_power":0.0006773720815809394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":4.328159646639027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002170097016572237,"output_power":0.00020246899552471293,"reference_active":true,"delta_db":-0.301206216202647},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7025087336003105e-05,"output_power":1.6203179184827668e-05,"reference_active":true,"delta_db":-0.21489113726724682},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.679451250427729e-07,"output_power":8.477684785391012e-07,"reference_active":true,"delta_db":-0.10215003481230073}],"active_band_count":3,"active_band_mae_db":0.20608246276073153}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":2.2598749620148992,"rms_envelope_nrmse":0.24432055846920883,"envelope_correlation":0.9504088487437333,"rms_gain_error_db":-0.8599172572584695,"reference_mean_power":0.00032373528856046693,"output_mean_power":0.00026558180379947634},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.248023532990147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.708460616175106e-05,"output_power":4.660949215355116e-05,"reference_active":true,"delta_db":-0.8804463813112785},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9595191589145463e-05,"output_power":1.484430866333516e-05,"reference_active":true,"delta_db":-1.2058953786585846},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.074204859780516e-06,"output_power":4.206012810247704e-06,"reference_active":true,"delta_db":-0.8149740583484248}],"active_band_count":3,"active_band_mae_db":0.9671052727727627}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":1.4574464839963215,"rms_envelope_nrmse":0.1576256085896771,"envelope_correlation":0.9825358708482046,"rms_gain_error_db":-0.8599435781765641,"reference_mean_power":0.0003253818281247374,"output_mean_power":0.00026693095303795386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.1005868407421186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.449763632243476e-05,"output_power":4.412487893611475e-05,"reference_active":true,"delta_db":-0.9169413957742184},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.38575211763481e-06,"output_power":7.1220684011517316e-06,"reference_active":true,"delta_db":-1.1986293972375797},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7591737704275045e-07,"output_power":3.052482060929828e-07,"reference_active":true,"delta_db":-0.9043928141498823}],"active_band_count":3,"active_band_mae_db":1.00665453572056}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.9436428546905518,"predicted_mos":2.9436428546905518,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":137640,"input_sample_rate":24000,"input_seconds":5.735,"input_float32_sha256":"2ec5a38e89afe71990359d8e44dc96aa37e97ac0ddc89dc6f2f12921cf5dd9d3","internal_samples":91760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6188256929999625},"started_at":"2026-10-13T05:11:38.093980+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002624901000672253,"native_input_samples":137400,"native_input_sha256":"ae5e893bde4e89f71a715d9e2925da3800e4a6a4bedbbfe6c3d457aed76127fb","output_sha256":"5ab261d304a2d1176a02229bbc9cf2164bdc54bdffa530794a600c01b8ba99a9","output_samples":137640,"length_delta_samples":240,"analysis_s":0.694569520001096,"synthesis_s":0.12632861799829698,"total_s":0.8208981379993929,"rtf":0.1433883210479289,"output_peak":0.5126457214355469,"output_rms":0.06060611261128528,"finished_at":"2026-10-13T05:11:38.929175+00:00","analysis_stages":{"external_f0_s":0.16000428100051067},"evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"5ab261d304a2d1176a02229bbc9cf2164bdc54bdffa530794a600c01b8ba99a9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":91600,"output_samples":91760,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.342804922303863,"output_peak":0.5217926438964049},"common_prefix":{"samples":91600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.998256320836966,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.603036430675416,"spectral_convergence":0.2458736163042546,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.2293200011659255,"spectral_convergence":0.2527198769726499,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.728101172600788,"spectral_convergence":0.3051247853981502,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":5.753390560041559e-05,"delta_db":0.14530816587132514,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.0008189015736884922,"delta_db":-0.8677393676748807,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.001789729077670786,"delta_db":-0.0869902225594238,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.00047196969472016457,"delta_db":-0.3596783253297461,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.00029702803977708155,"delta_db":-0.49915727651535613,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.0002501685886171012,"delta_db":-0.5166292859253873,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.41258377397935325,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.36591185866750564,"pitch":null,"output_rms_dbfs":-24.335235137915635,"output_peak":0.5217926438964049,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.520152534814042},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91600,"output_samples":91600,"intelligibility":{"estoi":{"value":0.9324034297166373,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06331896918232865,"rms_dbfs":-23.96932327924813,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.14269480218138,"mean":-2.0079300393183678e-05,"zero_energy":false},"output":{"rms":0.060706926129347395,"rms_dbfs":-24.335235137915635,"peak":0.5217926438964049,"peak_dbfs":-5.6500409553759345,"crest_factor_db":18.685194182539703,"mean":7.5176956346288e-06,"zero_energy":false},"rms_difference_db":-0.36591185866750564},"full_scale":{"reference":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":3.888598278097511e-06,"output_rms_dbfs":-54.10206920472241,"output_zero_energy":false,"output_peak":0.05635079404579103,"output_energy_fraction":0.00010136868012946701,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":183,"both_estimated_voiced_frames":172,"reference_voiced_output_unvoiced_frames":27,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.864321608040201,"absolute_difference_cents_median":0.7458797700851687,"absolute_difference_cents_p95":4.577077511332592,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00020825955187282386,"log_rms_mae_db":0.7096638431743292,"rms_envelope_nrmse":0.1195283018314085,"envelope_correlation":0.9878980573594828,"rms_gain_error_db":-0.36591006765680273,"reference_mean_power":0.004012756910556895,"output_mean_power":0.0036885174684124188},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":2.2120307480210933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008889321025104637,"output_power":0.0008512729677164256,"reference_active":true,"delta_db":-0.18799747872789396},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022980725896418495,"output_power":0.0002001607608564121,"reference_active":true,"delta_db":-0.5998479950443434},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113193274777301e-05,"output_power":3.139493970074512e-05,"reference_active":true,"delta_db":-1.173194637858906}],"active_band_count":3,"active_band_mae_db":0.6536800372103811}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.0001984160011662782,"log_rms_mae_db":0.4602680525846593,"rms_envelope_nrmse":0.10631827366557323,"envelope_correlation":0.9899386519675605,"rms_gain_error_db":-0.3659031659506463,"reference_mean_power":0.0040337408532874226,"output_mean_power":0.0037078117555230925},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":2.0115882799069226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008024963383362849,"output_power":0.0007687152518437062,"reference_active":true,"delta_db":-0.1867756160602439},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001146343816948982,"output_power":0.00010015934108169484,"reference_active":true,"delta_db":-0.5862343407531527},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4945953602736586e-06,"output_power":2.41517762348298e-06,"reference_active":true,"delta_db":-1.60447819655027}],"active_band_count":3,"active_band_mae_db":0.7924960511212221}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00019054595180557928,"log_rms_mae_db":0.9788949864358351,"rms_envelope_nrmse":0.15435451284743992,"envelope_correlation":0.9827281907704295,"rms_gain_error_db":-0.3695128987261266,"reference_mean_power":0.0025757888112863028,"output_mean_power":0.0023656961468010974},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.7595886306573375e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000727930551645222,"output_power":0.0006983630491515275,"reference_active":true,"delta_db":-0.1800869484196036},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019010480993234813,"output_power":0.00017080058692426654,"reference_active":true,"delta_db":-0.4650374656871108},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6294378423488105e-05,"output_power":2.1044390664740978e-05,"reference_active":true,"delta_db":-0.9672655354839246}],"active_band_count":3,"active_band_mae_db":0.537463316530213}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.00018019065505981728,"log_rms_mae_db":0.5948568353929814,"rms_envelope_nrmse":0.13728391788969405,"envelope_correlation":0.9857305741139194,"rms_gain_error_db":-0.369512961068897,"reference_mean_power":0.0025893611436675294,"output_mean_power":0.002378161425920016},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.6269545971301712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006780414214466183,"output_power":0.0006496141748092242,"reference_active":true,"delta_db":-0.1860073286639986},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.548437806715931e-05,"output_power":8.610726651513824e-05,"reference_active":true,"delta_db":-0.4489252096368387},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.932751011760402e-06,"output_power":2.052370361465828e-06,"reference_active":true,"delta_db":-1.5501945895955167}],"active_band_count":3,"active_band_mae_db":0.7283757092987847}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":1.0481869910500978,"rms_envelope_nrmse":0.10683423712464751,"envelope_correlation":0.992271688370945,"rms_gain_error_db":-0.4108194574659452,"reference_mean_power":0.0007122971869413385,"output_mean_power":0.0006480063825047524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":4.689927289528062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002336243034349734,"output_power":0.00021958793873365286,"reference_active":true,"delta_db":-0.2690953749351103},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7979021289898206e-05,"output_power":2.898025754614708e-05,"reference_active":true,"delta_db":-1.1744152817612241},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0421847785855855e-05,"output_power":7.905906738888407e-06,"reference_active":true,"delta_db":-1.1999303894605449}],"active_band_count":3,"active_band_mae_db":0.8811470153856265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":0.6879419635383291,"rms_envelope_nrmse":0.08396643583413865,"envelope_correlation":0.9955974726414754,"rms_gain_error_db":-0.4108399810367157,"reference_mean_power":0.0007159674763911087,"output_mean_power":0.0006513423194280176},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":4.328159646639027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002170097016572237,"output_power":0.00020357924340306727,"reference_active":true,"delta_db":-0.2774565390699935},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7025087336003105e-05,"output_power":1.2552743862080537e-05,"reference_active":true,"delta_db":-1.3235068128782195},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.679451250427729e-07,"output_power":6.013266525665531e-07,"reference_active":true,"delta_db":-1.5938181473142432}],"active_band_count":3,"active_band_mae_db":1.0649271664208186}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":1.0642954935573268,"rms_envelope_nrmse":0.12726887043204635,"envelope_correlation":0.987593534208433,"rms_gain_error_db":-0.4734614612449125,"reference_mean_power":0.00032373528856046693,"output_mean_power":0.000290297901119387},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.248023532990147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.708460616175106e-05,"output_power":5.174905593123553e-05,"reference_active":true,"delta_db":-0.4261657783849596},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9595191589145463e-05,"output_power":1.658228192391946e-05,"reference_active":true,"delta_db":-0.7250521955022695},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.074204859780516e-06,"output_power":3.290572663868714e-06,"reference_active":true,"delta_db":-1.8809651139417043}],"active_band_count":3,"active_band_mae_db":1.0107276959429778}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":0.681974163319615,"rms_envelope_nrmse":0.08294655147188329,"envelope_correlation":0.9957571629630719,"rms_gain_error_db":-0.4733821764329696,"reference_mean_power":0.0003253818281247374,"output_mean_power":0.0002917797025529318},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.1005868407421186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.449763632243476e-05,"output_power":4.8504682073706456e-05,"reference_active":true,"delta_db":-0.5059400408975508},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.38575211763481e-06,"output_power":7.970813439711019e-06,"reference_active":true,"delta_db":-0.709664356941533},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7591737704275045e-07,"output_power":2.2734628740947085e-07,"reference_active":true,"delta_db":-2.1840453522180443}],"active_band_count":3,"active_band_mae_db":1.133216583352376}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0014","method":"magphase","status":"measured","mos":{"status":"measured","value":3.174623966217041,"predicted_mos":3.174623966217041,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":275752,"input_sample_rate":48000,"input_seconds":5.744833333333333,"input_float32_sha256":"f9beff3113576866c70a40d356f4de3c72ed7b2a5c7b1f51b42b5e4573086913","internal_samples":91918,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3096313149999332},"started_at":"2026-10-13T05:18:59.043480+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.01013036199947237,"native_input_samples":274800,"native_input_sha256":"fd9a7cd8ebcd4f6f665edc1d482724d5fa8e3fd77398375c71904f1d86649fe7","output_sha256":"e98bc98a88003b44f6a53dcbabe4cb1013a7a6f5e574a08d39009243a85e764f","output_samples":275752,"length_delta_samples":952,"analysis_s":3.7453178220002883,"synthesis_s":1.0283511810011987,"total_s":4.773669003001487,"rtf":0.8338286468124869,"output_peak":0.5791402765155271,"output_rms":0.06263076834732738,"finished_at":"2026-10-13T05:19:03.839430+00:00","evaluation_input":{"reference":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"e98bc98a88003b44f6a53dcbabe4cb1013a7a6f5e574a08d39009243a85e764f","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":91600,"output_samples":91918,"length_delta_samples":318,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.057560741511246,"output_peak":0.5611616124965104},"common_prefix":{"samples":91600,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9965403946996236,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":91600,"output_samples":91600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.3631076300020295,"spectral_convergence":0.6303721591007407,"spectral_convergence_status":"ok","amplitude_floor":1.4112644553101518e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.287731831604663,"spectral_convergence":0.3545041091670991,"spectral_convergence_status":"ok","amplitude_floor":1.1975531593072457e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.8152877283784905,"spectral_convergence":0.328330822806051,"spectral_convergence_status":"ok","amplitude_floor":1.114415465542996e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":5.564075826256522e-05,"output_power":5.320682738609464e-05,"delta_db":-0.19425676241686993,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0010000129923479963,"output_power":0.0009243909101068733,"delta_db":-0.34149976225828377,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0018259392170112694,"output_power":0.0017524936003466866,"delta_db":-0.17829875666348582,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0005127219920465177,"output_power":0.0005205397093123727,"delta_db":0.06571919298983907,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00033320627903786135,"output_power":0.0003572901886444987,"delta_db":0.3030791365103996,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002817706196064759,"output_power":0.00033426047297095425,"delta_db":0.7418931637546136,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3041244624322486,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07331033538087327,"pitch":null,"output_rms_dbfs":-24.042633614629008,"output_peak":0.5611616124965104,"reference_rms_dbfs":-23.96932327924813,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.488709063328394},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":91600,"output_samples":91600,"intelligibility":{"estoi":{"value":0.8366512438586843,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06331896918232865,"rms_dbfs":-23.96932327924813,"peak":0.51129150390625,"peak_dbfs":-5.82662847706675,"crest_factor_db":18.14269480218138,"mean":-2.0079300393183678e-05,"zero_energy":false},"output":{"rms":0.06278679565615274,"rms_dbfs":-24.042633614629004,"peak":0.5611616124965104,"peak_dbfs":-5.018240910077808,"crest_factor_db":19.024392704551197,"mean":-0.00018076543741173397,"zero_energy":false},"rms_difference_db":-0.07331033538087439},"full_scale":{"reference":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":91600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":572,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":4.337204094626941e-06,"threshold_dbfs":-53.62790140893722,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":55,"quiet_samples":8800,"reference_mean_power":1.4632275666702877e-06,"output_mean_power":4.7113509875683274e-06,"output_rms_dbfs":-53.26854540362479,"output_zero_energy":false,"output_peak":0.034333626886362476,"output_energy_fraction":0.00011481430984307599,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":565,"reference_estimated_voiced_frames":199,"output_estimated_voiced_frames":210,"both_estimated_voiced_frames":183,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":27,"paired_voiced_fraction":0.9195979899497487,"absolute_difference_cents_median":3.258599561198696,"absolute_difference_cents_p95":22.953024727900054,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":91600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00020825955187282386,"log_rms_mae_db":2.3657096541932416,"rms_envelope_nrmse":0.2215639477592548,"envelope_correlation":0.9545268169811502,"rms_gain_error_db":-0.0732905277957142,"reference_mean_power":0.004012756910556895,"output_mean_power":0.003945606756039628},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":2.2120307480210933e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008889321025104637,"output_power":0.0008759443012268099,"reference_active":true,"delta_db":-0.06392098894688976},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00022980725896418495,"output_power":0.0002242484154181855,"reference_active":true,"delta_db":-0.10634359788828493},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.113193274777301e-05,"output_power":4.03031110540781e-05,"reference_active":true,"delta_db":-0.08840545805928593}],"active_band_count":3,"active_band_mae_db":0.08622334829815355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.0001984160011662782,"log_rms_mae_db":1.573891440799419,"rms_envelope_nrmse":0.14380349217487642,"envelope_correlation":0.9791445092862208,"rms_gain_error_db":-0.07329728150237434,"reference_mean_power":0.0040337408532874226,"output_mean_power":0.003966233382026515},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":2.0115882799069226e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008024963383362849,"output_power":0.0007927005267607241,"reference_active":true,"delta_db":-0.05333912676153101},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001146343816948982,"output_power":0.00011014818134669157,"reference_active":true,"delta_db":-0.17337561934893816},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.4945953602736586e-06,"output_power":3.0657562905757423e-06,"reference_active":true,"delta_db":-0.5685926791145313}],"active_band_count":3,"active_band_mae_db":0.2651024750750001}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":0.00019054595180557928,"log_rms_mae_db":2.952382152525571,"rms_envelope_nrmse":0.23006151867693447,"envelope_correlation":0.9599514671855294,"rms_gain_error_db":-0.2565425051929562,"reference_mean_power":0.0025757888112863028,"output_mean_power":0.0024280409210557727},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.7595886306573375e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000727930551645222,"output_power":0.0006877082993057427,"reference_active":true,"delta_db":-0.24685681874578735},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019010480993234813,"output_power":0.00017860229976783825,"reference_active":true,"delta_db":-0.2710605852594748},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6294378423488105e-05,"output_power":2.4065363288196405e-05,"reference_active":true,"delta_db":-0.38470486795699466}],"active_band_count":3,"active_band_mae_db":0.3008740906540856}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":0.00018019065505981728,"log_rms_mae_db":2.1951787528583497,"rms_envelope_nrmse":0.16124790411868464,"envelope_correlation":0.9793598241693732,"rms_gain_error_db":-0.25653933322506595,"reference_mean_power":0.0025893611436675294,"output_mean_power":0.002440836523777633},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.6269545971301712e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0006780414214466183,"output_power":0.0006418421274057303,"reference_active":true,"delta_db":-0.23828006981469418},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.548437806715931e-05,"output_power":8.914543961367169e-05,"reference_active":true,"delta_db":-0.29833192678396786},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.932751011760402e-06,"output_power":2.4513766550412845e-06,"reference_active":true,"delta_db":-0.7786514738121666}],"active_band_count":3,"active_band_mae_db":0.43842115680360955}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.327720993461302e-05,"log_rms_mae_db":2.849585715537523,"rms_envelope_nrmse":0.2501942470441079,"envelope_correlation":0.9513397179038423,"rms_gain_error_db":0.11997563852381042,"reference_mean_power":0.0007122971869413385,"output_mean_power":0.000732249009768986},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":4.689927289528062e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002336243034349734,"output_power":0.00023757037441719803,"reference_active":true,"delta_db":0.07274262631842497},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.7979021289898206e-05,"output_power":3.8314539245557556e-05,"reference_active":true,"delta_db":0.038198383998860594},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0421847785855855e-05,"output_power":1.0581576816264074e-05,"reference_active":true,"delta_db":0.06605663141481635}],"active_band_count":3,"active_band_mae_db":0.05899921391070064}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":8.747046012456963e-05,"log_rms_mae_db":1.9336823138579748,"rms_envelope_nrmse":0.15784239799569846,"envelope_correlation":0.9792003980072591,"rms_gain_error_db":0.11990798040559172,"reference_mean_power":0.0007159674763911087,"output_mean_power":0.000736010639683038},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":4.328159646639027e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002170097016572237,"output_power":0.00022053764861546684,"reference_active":true,"delta_db":0.07003589896667956},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7025087336003105e-05,"output_power":1.71411864936651e-05,"reference_active":true,"delta_db":0.029515315511093827},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.679451250427729e-07,"output_power":8.435713285464219e-07,"reference_active":true,"delta_db":-0.12370457773244498}],"active_band_count":3,"active_band_mae_db":0.07441859740340612}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1144,"rms_floor":9.039773028293343e-05,"log_rms_mae_db":2.9821687476075356,"rms_envelope_nrmse":0.3335120634137572,"envelope_correlation":0.9111719222160649,"rms_gain_error_db":0.528970173625015,"reference_mean_power":0.00032373528856046693,"output_mean_power":0.000365668086654394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1144,"frequency_resolution_hz":0.17482517482517484,"power_floor":1.248023532990147e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.708460616175106e-05,"output_power":6.478360826082358e-05,"reference_active":true,"delta_db":0.5494612418712437},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9595191589145463e-05,"output_power":2.4028738107868954e-05,"reference_active":true,"delta_db":0.8858144992702656},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.074204859780516e-06,"output_power":7.194129181689534e-06,"reference_active":true,"delta_db":1.516102350627269}],"active_band_count":3,"active_band_mae_db":0.9837926972562595}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":91600,"uncovered_tail_samples":0,"metrics":{"frame_count":1138,"rms_floor":6.399293414479691e-05,"log_rms_mae_db":2.0530587904515736,"rms_envelope_nrmse":0.21430601823994055,"envelope_correlation":0.9607344416037724,"rms_gain_error_db":0.5291734306983438,"reference_mean_power":0.0003253818281247374,"output_mean_power":0.00036754510058694534},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1138,"frequency_resolution_hz":0.17574692442882248,"power_floor":1.1005868407421186e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.449763632243476e-05,"output_power":6.368840821270948e-05,"reference_active":true,"delta_db":0.676827281267609},{"low_hz":8.0,"high_hz":20.0,"reference_power":9.38575211763481e-06,"output_power":1.158082979145381e-05,"reference_active":true,"delta_db":0.9127059854801609},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.7591737704275045e-07,"output_power":5.219861734429759e-07,"reference_active":true,"delta_db":1.4256659766363888}],"active_band_count":3,"active_band_mae_db":1.0050664144613863}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"hnm","status":"measured","mos":{"status":"measured","value":1.8468387126922607,"predicted_mos":1.8468387126922607,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37680,"input_sample_rate":16000,"input_seconds":2.355,"input_float32_sha256":"d48c43b1e19d34d802039930cb90ec770785f4bec52633c7851fe50843622c80","internal_samples":37680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6139106800001173},"started_at":"2026-10-13T04:40:31.293135+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.1588999768719077e-05,"native_input_samples":37680,"native_input_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","output_sha256":"1e76b6044ba8ebf426e9dd5075000ea86f739dfcd39bc3454416d776ceb79f28","output_samples":37680,"length_delta_samples":0,"analysis_s":0.36459148500034644,"synthesis_s":0.18686100099967007,"total_s":0.5514524860000165,"rtf":0.23416241443737432,"output_peak":0.4406077154287218,"output_rms":0.05617439282863003,"finished_at":"2026-10-13T04:40:31.847490+00:00","evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"1e76b6044ba8ebf426e9dd5075000ea86f739dfcd39bc3454416d776ceb79f28","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.711581176150119,"spectral_convergence":0.3485257011270288,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.412851650069931,"spectral_convergence":0.3159964797261885,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.901812597623686,"spectral_convergence":0.33536373326628804,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":8.222316155944787e-05,"delta_db":0.25597680684405655,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.0015388482862075386,"delta_db":-0.17392332056788753,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0006690579952975963,"delta_db":-0.09468928350205358,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.0003425757078319997,"delta_db":0.08247325023577443,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.00032756961032724723,"delta_db":-0.3314005155868788,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.00019528764844141077,"delta_db":-0.31968793184801947,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2096918514307784,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14530873858433813,"pitch":null,"output_rms_dbfs":-25.009232260722506,"output_peak":0.4406077154287218,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.6754151412812455},"common_prefix":{"samples":37680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.711581176150119,"spectral_convergence":0.3485257011270288,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.412851650069931,"spectral_convergence":0.3159964797261885,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.901812597623686,"spectral_convergence":0.33536373326628804,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":8.222316155944787e-05,"delta_db":0.25597680684405655,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.0015388482862075386,"delta_db":-0.17392332056788753,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0006690579952975963,"delta_db":-0.09468928350205358,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.0003425757078319997,"delta_db":0.08247325023577443,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.00032756961032724723,"delta_db":-0.3314005155868788,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.00019528764844141077,"delta_db":-0.31968793184801947,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2096918514307784,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14530873858433813,"pitch":null,"output_rms_dbfs":-25.009232260722506,"output_peak":0.4406077154287218,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.6754151412812455},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37680,"output_samples":37680,"intelligibility":{"estoi":{"value":0.8814744056035165,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05712205511060183,"rms_dbfs":-24.863923522138165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.476244606491736,"mean":-4.511304721710788e-05,"zero_energy":false},"output":{"rms":0.05617439282863003,"rms_dbfs":-25.009232260722506,"peak":0.4406077154287218,"peak_dbfs":-7.118958044804353,"crest_factor_db":17.890274215918154,"mean":-2.7627990982745094e-05,"zero_energy":false},"rms_difference_db":-0.14530873858434035},"full_scale":{"reference":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":8.586330347126326e-06,"output_rms_dbfs":-50.66192406658563,"output_zero_energy":false,"output_peak":0.06820785834433107,"output_energy_fraction":0.00045061382818181554,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":72,"both_estimated_voiced_frames":67,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9054054054054054,"absolute_difference_cents_median":2.197202493090288,"absolute_difference_cents_p95":1203.0235640319704,"difference_over_600_cents_frames":6,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00018834951637393047,"log_rms_mae_db":0.8750656469663899,"rms_envelope_nrmse":0.08904470475777665,"envelope_correlation":0.9928150327400733,"rms_gain_error_db":-0.14530372159391225,"reference_mean_power":0.00326985830085679,"output_mean_power":0.003162267180625409},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.479011996203747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005558024994761568,"output_power":0.0005586452427242069,"reference_active":true,"delta_db":0.022156092870459217},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011456565813557757,"output_power":0.0001036659626197315,"reference_active":true,"delta_db":-0.434182694955706},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.821714301732997e-05,"output_power":3.593775162753157e-05,"reference_active":true,"delta_db":-0.26707314694689005}],"active_band_count":3,"active_band_mae_db":0.2411373115910184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00017471448177334785,"log_rms_mae_db":0.45941262383474546,"rms_envelope_nrmse":0.047536591009040334,"envelope_correlation":0.9979941580804034,"rms_gain_error_db":-0.1453009555772169,"reference_mean_power":0.0033120735889998944,"output_mean_power":0.0032030954608129898},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":7.763101583949748e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004881183543297566,"output_power":0.0004898896122119223,"reference_active":true,"delta_db":0.015730922275153873},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.883974713754685e-05,"output_power":5.160100263130861e-05,"reference_active":true,"delta_db":-0.5701265743748647},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7824879244994614e-06,"output_power":1.493263527881436e-06,"reference_active":true,"delta_db":-0.7689013856805449}],"active_band_count":3,"active_band_mae_db":0.4515862941101878}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00017587844894922502,"log_rms_mae_db":1.4520588265819006,"rms_envelope_nrmse":0.1108308086477907,"envelope_correlation":0.9902975826314668,"rms_gain_error_db":-0.19031123299324137,"reference_mean_power":0.002090015296145697,"output_mean_power":0.00200040688799992},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":6.927253047980857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041075223634869497,"output_power":0.00040780314279658975,"reference_active":true,"delta_db":-0.03129368227334989},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.156045953139752e-05,"output_power":7.48435941907591e-05,"reference_active":true,"delta_db":-0.37325028715115194},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2217147797459886e-05,"output_power":2.9126747603919925e-05,"reference_active":true,"delta_db":1.1760368833092474}],"active_band_count":3,"active_band_mae_db":0.5268602842445831}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00016316895420677254,"log_rms_mae_db":0.8385406225247296,"rms_envelope_nrmse":0.062086027914702004,"envelope_correlation":0.9970422980305516,"rms_gain_error_db":-0.1903108680040395,"reference_mean_power":0.002117036355483854,"output_mean_power":0.0020262696026090332},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":5.84659163450443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037072755253225205,"output_power":0.0003688764708451188,"reference_active":true,"delta_db":-0.021739098637848127},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.09979708848843e-05,"output_power":3.661238252303658e-05,"reference_active":true,"delta_db":-0.49134371502622876},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.223097583568823e-06,"output_power":1.230063416583998e-06,"reference_active":true,"delta_db":0.02466394151009146}],"active_band_count":3,"active_band_mae_db":0.17924891839138946}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00010383940879839973,"log_rms_mae_db":1.5009171290963828,"rms_envelope_nrmse":0.18057461728174096,"envelope_correlation":0.9732271325766215,"rms_gain_error_db":-0.08430343452795805,"reference_mean_power":0.0005521302842545546,"output_mean_power":0.0005415159154735135},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":2.374888225952349e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001363070736426414,"output_power":0.00014484632709447934,"reference_active":true,"delta_db":0.2638909307229182},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2931511209828857e-05,"output_power":2.9091487007347156e-05,"reference_active":true,"delta_db":-0.53845739998887},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.586128694071041e-05,"output_power":1.3544095548405269e-05,"reference_active":true,"delta_db":-0.6858841296476331}],"active_band_count":3,"active_band_mae_db":0.49607748678647373}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":0.8767495699952442,"rms_envelope_nrmse":0.1038786074616908,"envelope_correlation":0.9902263439494389,"rms_gain_error_db":-0.08428669511899298,"reference_mean_power":0.0005592400075632056,"output_mean_power":0.0005484910727725139},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.9436382041482275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261572491396227,"output_power":0.00013495745569109716,"reference_active":true,"delta_db":0.2928467141846016},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5195698098116377e-05,"output_power":1.281073805176798e-05,"reference_active":true,"delta_db":-0.7414650556102086},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.080674445150943e-07,"output_power":8.373118501616048e-07,"reference_active":true,"delta_db":-0.3523086823295004}],"active_band_count":3,"active_band_mae_db":0.4622068173747702}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":1.4840375048742702,"rms_envelope_nrmse":0.1528323854377688,"envelope_correlation":0.9832786792512422,"rms_gain_error_db":-0.3771513867776839,"reference_mean_power":0.00026303793349419816,"output_mean_power":0.0002411588727905225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":1.3356760090821463e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8507771208283054e-05,"output_power":5.323986158951681e-05,"reference_active":true,"delta_db":-0.4097663669069189},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1981263949052123e-05,"output_power":1.1830102742673404e-05,"reference_active":true,"delta_db":-0.05514119417964079},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.2561065186681e-06,"output_power":5.072988584429744e-06,"reference_active":true,"delta_db":-2.6116445876423384}],"active_band_count":3,"active_band_mae_db":1.0255173829096327}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":0.7632782527462092,"rms_envelope_nrmse":0.07360152624400063,"envelope_correlation":0.9969840939896161,"rms_gain_error_db":-0.3771319462323691,"reference_mean_power":0.00026641634082957393,"output_mean_power":0.00024425736316495536},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.157398705126345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.177699204483556e-05,"output_power":4.68404172437647e-05,"reference_active":true,"delta_db":-0.43516061724737715},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.036512028004997e-06,"output_power":6.364037024886722e-06,"reference_active":true,"delta_db":-0.43624736740141545},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.58854722653168e-07,"output_power":2.513940856977195e-07,"reference_active":true,"delta_db":-3.469438690786919}],"active_band_count":3,"active_band_mae_db":1.446948891811904}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"world_harvest","status":"measured","mos":{"status":"measured","value":2.3326570987701416,"predicted_mos":2.3326570987701416,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37680,"input_sample_rate":16000,"input_seconds":2.355,"input_float32_sha256":"897a2640561a43b542f6d07bbe57f5bfa822ddacf5d0cbdaec302ea67c791d01","internal_samples":37680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.760533992999626},"started_at":"2026-10-13T04:43:47.637576+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.2274999537330586e-05,"native_input_samples":37680,"native_input_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","output_sha256":"3b76bbf5e59fe732328127e080e3be802db09769835efed4def8e8736e465746","output_samples":37680,"length_delta_samples":0,"analysis_s":0.5785123759997077,"synthesis_s":0.03620820000014646,"total_s":0.6147205759998542,"rtf":0.26102784543518226,"output_peak":0.6762865567355033,"output_rms":0.07002487025634588,"finished_at":"2026-10-13T04:43:48.255430+00:00","evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"3b76bbf5e59fe732328127e080e3be802db09769835efed4def8e8736e465746","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.7581617782469,"spectral_convergence":0.6045270392334015,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.642823128805037,"spectral_convergence":0.4673221363990955,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.131604547825738,"spectral_convergence":0.5390908020786879,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":8.498961687175215e-05,"delta_db":0.39969380349779565,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.002885079890116447,"delta_db":2.555697063503915,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0008239137687342491,"delta_db":0.809490673054125,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.0005188300358974094,"delta_db":1.8851587142120554,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.0003569093737985179,"delta_db":0.041142986934461626,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.0002337597689996979,"delta_db":0.46126200350798313,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.025407540785056,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.7689697786535703,"pitch":null,"output_rms_dbfs":-23.0949537434846,"output_peak":0.6762865567355033,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.177529818292558},"common_prefix":{"samples":37680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.7581617782469,"spectral_convergence":0.6045270392334015,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.642823128805037,"spectral_convergence":0.4673221363990955,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.131604547825738,"spectral_convergence":0.5390908020786879,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":8.498961687175215e-05,"delta_db":0.39969380349779565,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.002885079890116447,"delta_db":2.555697063503915,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0008239137687342491,"delta_db":0.809490673054125,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.0005188300358974094,"delta_db":1.8851587142120554,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.0003569093737985179,"delta_db":0.041142986934461626,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.0002337597689996979,"delta_db":0.46126200350798313,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.025407540785056,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.7689697786535703,"pitch":null,"output_rms_dbfs":-23.0949537434846,"output_peak":0.6762865567355033,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.177529818292558},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37680,"output_samples":37680,"intelligibility":{"estoi":{"value":0.8457275385286426,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05712205511060183,"rms_dbfs":-24.863923522138165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.476244606491736,"mean":-4.511304721710788e-05,"zero_energy":false},"output":{"rms":0.07002487025634588,"rms_dbfs":-23.0949537434846,"peak":0.6762865567355033,"peak_dbfs":-3.397384908409209,"crest_factor_db":19.69756883507539,"mean":-2.8150365755241814e-09,"zero_energy":false},"rms_difference_db":1.7689697786535667},"full_scale":{"reference":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":1.1406807481196378e-05,"output_rms_dbfs":-49.42835888211622,"output_zero_energy":false,"output_peak":0.12760285892291262,"output_energy_fraction":0.00038524160335148567,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":68,"both_estimated_voiced_frames":53,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.7162162162162162,"absolute_difference_cents_median":10.906519057563923,"absolute_difference_cents_p95":1202.4579235707852,"difference_over_600_cents_frames":16,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00018834951637393047,"log_rms_mae_db":1.6200256682504464,"rms_envelope_nrmse":0.30425582768663334,"envelope_correlation":0.9724610855436384,"rms_gain_error_db":1.7689824241067167,"reference_mean_power":0.00326985830085679,"output_mean_power":0.004913909744170298},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.479011996203747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005558024994761568,"output_power":0.0006936621539285349,"reference_active":true,"delta_db":0.9622750463062356},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011456565813557757,"output_power":0.00015485717789772297,"reference_active":true,"delta_db":1.3087688615519701},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.821714301732997e-05,"output_power":2.8211321597803568e-05,"reference_active":true,"delta_db":-1.3183478581626784}],"active_band_count":3,"active_band_mae_db":1.1964639220069613}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00017471448177334785,"log_rms_mae_db":1.0743947760844355,"rms_envelope_nrmse":0.27732217885664623,"envelope_correlation":0.9831352422940605,"rms_gain_error_db":1.7690189590928276,"reference_mean_power":0.0033120735889998944,"output_mean_power":0.0049773923212347694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":7.763101583949748e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004881183543297566,"output_power":0.0006143226533750847,"reference_active":true,"delta_db":0.9987139198785068},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.883974713754685e-05,"output_power":8.093874183861094e-05,"reference_active":true,"delta_db":1.3848565147952066},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7824879244994614e-06,"output_power":1.2748590106777289e-06,"reference_active":true,"delta_db":-1.4556443850585512}],"active_band_count":3,"active_band_mae_db":1.279738273244088}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00017587844894922502,"log_rms_mae_db":1.911801421295465,"rms_envelope_nrmse":0.36521730065278035,"envelope_correlation":0.9752150843868903,"rms_gain_error_db":2.2420459592259205,"reference_mean_power":0.002090015296145697,"output_mean_power":0.0035023057770603855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":6.927253047980857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041075223634869497,"output_power":0.0005750345781731857,"reference_active":true,"delta_db":1.4611402397864244},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.156045953139752e-05,"output_power":0.0001260951543448672,"reference_active":true,"delta_db":1.8921873354906726},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2217147797459886e-05,"output_power":2.3198986266828945e-05,"reference_active":true,"delta_db":0.1878070366736147}],"active_band_count":3,"active_band_mae_db":1.1803782039835706}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00016316895420677254,"log_rms_mae_db":1.2989529287065404,"rms_envelope_nrmse":0.3400160396618442,"envelope_correlation":0.9852055724576549,"rms_gain_error_db":2.2420504730032946,"reference_mean_power":0.002117036355483854,"output_mean_power":0.0035475895214253333},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":5.84659163450443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037072755253225205,"output_power":0.0005257623831423826,"reference_active":true,"delta_db":1.5173464623605282},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.09979708848843e-05,"output_power":6.447386843532803e-05,"reference_active":true,"delta_db":1.966213659955793},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.223097583568823e-06,"output_power":1.4632856833982314e-06,"reference_active":true,"delta_db":0.7786801538334402}],"active_band_count":3,"active_band_mae_db":1.420746758716587}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00010383940879839973,"log_rms_mae_db":1.6875921021031446,"rms_envelope_nrmse":0.29666019830372165,"envelope_correlation":0.9552423159950129,"rms_gain_error_db":1.2066678822733596,"reference_mean_power":0.0005521302842545546,"output_mean_power":0.000728967820034326},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":2.374888225952349e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001363070736426414,"output_power":0.00015387103386905494,"reference_active":true,"delta_db":0.5263847775649616},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2931511209828857e-05,"output_power":3.6283012014236094e-05,"reference_active":true,"delta_db":0.4209167188566309},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.586128694071041e-05,"output_power":1.335671813734075e-05,"reference_active":true,"delta_db":-0.7463866056935754}],"active_band_count":3,"active_band_mae_db":0.5645627007050559}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":1.1707058669511476,"rms_envelope_nrmse":0.23989021817164619,"envelope_correlation":0.9741926506791779,"rms_gain_error_db":1.2067532204694702,"reference_mean_power":0.0005592400075632056,"output_mean_power":0.0007383691702968068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.9436382041482275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261572491396227,"output_power":0.00014233614208281777,"reference_active":true,"delta_db":0.5240297978812303},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5195698098116377e-05,"output_power":1.617481909052523e-05,"reference_active":true,"delta_db":0.2711877535776861},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.080674445150943e-07,"output_power":6.008794392634523e-07,"reference_active":true,"delta_db":-1.7933076220046285}],"active_band_count":3,"active_band_mae_db":0.8628417244878482}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":1.5970829781096996,"rms_envelope_nrmse":0.2026377238829749,"envelope_correlation":0.9741481469131337,"rms_gain_error_db":0.5375850091379798,"reference_mean_power":0.00026303793349419816,"output_mean_power":0.0002976986634489013},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":1.3356760090821463e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8507771208283054e-05,"output_power":6.873663355583383e-05,"reference_active":true,"delta_db":0.699747038866086},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1981263949052123e-05,"output_power":1.412340123240849e-05,"reference_active":true,"delta_db":0.7143666133748766},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.2561065186681e-06,"output_power":5.681393055156846e-06,"reference_active":true,"delta_db":-2.1197350774352115}],"active_band_count":3,"active_band_mae_db":1.1779495765587247}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":0.8333132794519349,"rms_envelope_nrmse":0.11590051689311222,"envelope_correlation":0.9930092031855671,"rms_gain_error_db":0.537661633900977,"reference_mean_power":0.00026641634082957393,"output_mean_power":0.0003015275663190261},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.157398705126345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.177699204483556e-05,"output_power":6.041350127151867e-05,"reference_active":true,"delta_db":0.6699718930765576},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.036512028004997e-06,"output_power":8.006933407016853e-06,"reference_active":true,"delta_db":0.5610878215839314},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.58854722653168e-07,"output_power":2.9433117961290315e-07,"reference_active":true,"delta_db":-2.7846265433770743}],"active_band_count":3,"active_band_mae_db":1.3385620860125211}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9695706367492676,"predicted_mos":3.9695706367492676,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37800,"input_sample_rate":16000,"input_seconds":2.3625,"input_float32_sha256":"8394f95b4a871fd7711d8d58b82e0cfc0fc5e3f4cc269f385ee8acf92cdaba12","internal_samples":37800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.168776362999779},"started_at":"2026-10-13T07:02:16.629598+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00012665299982472789,"native_input_samples":37680,"native_input_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","output_sha256":"9eaba59e2e343e293a7e46a9911c4cd538904022be5020ed55004705d26b8b60","output_samples":37800,"length_delta_samples":120,"analysis_s":0.6218175770009111,"synthesis_s":3.8279720659993473,"total_s":4.449789643000258,"rtf":1.8895072794056298,"output_peak":0.46052518486976624,"output_rms":0.05128558037624248,"finished_at":"2026-10-13T07:02:21.085026+00:00","evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"9eaba59e2e343e293a7e46a9911c4cd538904022be5020ed55004705d26b8b60","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":37680,"output_samples":37800,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.800094508024838,"output_peak":0.46052518486976624},"common_prefix":{"samples":37680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9968253968253968,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.040030713658352,"spectral_convergence":0.2639910730765135,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.142931004247995,"spectral_convergence":0.30393894347260747,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.360508600979068,"spectral_convergence":0.34089693220405554,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":7.423394277006692e-05,"delta_db":-0.18793963373806682,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.0013659061298322687,"delta_db":-0.691672832282463,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0005286963706937827,"delta_db":-1.1172636412869035,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.00022169562644810714,"delta_db":-1.8075211316425048,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.0003101294773320001,"delta_db":-0.5690060841319121,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.0001379256186419139,"delta_db":-1.8299862848366397,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0338982679864148,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9223620211931216,"pitch":null,"output_rms_dbfs":-25.78628554333129,"output_peak":0.46052518486976624,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.181156772961805},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37680,"output_samples":37680,"intelligibility":{"estoi":{"value":0.9242718853200421,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05712205511060183,"rms_dbfs":-24.863923522138165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.476244606491736,"mean":-4.511304721710788e-05,"zero_energy":false},"output":{"rms":0.051367179849765354,"rms_dbfs":-25.78628554333129,"peak":0.46052518486976624,"peak_dbfs":-6.734932288674401,"crest_factor_db":19.05135325465689,"mean":1.1609946752495762e-06,"zero_energy":false},"rms_difference_db":-0.9223620211931234},"full_scale":{"reference":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":1.7330399841335904e-06,"output_rms_dbfs":-57.61191417269775,"output_zero_energy":false,"output_peak":0.04218657314777374,"output_energy_fraction":0.00010877042679444679,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":77,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.972972972972973,"absolute_difference_cents_median":4.188103196314713,"absolute_difference_cents_p95":15.544213416113529,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00018834951637393047,"log_rms_mae_db":1.1727426832819885,"rms_envelope_nrmse":0.12322152049232227,"envelope_correlation":0.9948412113920714,"rms_gain_error_db":-0.9223695465653682,"reference_mean_power":0.00326985830085679,"output_mean_power":0.002644185858829165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.479011996203747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005558024994761568,"output_power":0.00038721985451079996,"reference_active":true,"delta_db":-1.5696287804645261},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011456565813557757,"output_power":9.834346776415273e-05,"reference_active":true,"delta_db":-0.663089360899905},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.821714301732997e-05,"output_power":3.686588839351058e-05,"reference_active":true,"delta_db":-0.15633513579234484}],"active_band_count":3,"active_band_mae_db":0.796351092385592}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00017471448177334785,"log_rms_mae_db":1.0015868791843565,"rms_envelope_nrmse":0.11598721603649749,"envelope_correlation":0.9963298338207575,"rms_gain_error_db":-0.9223672173911095,"reference_mean_power":0.0033120735889998944,"output_mean_power":0.0026783248809354606},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":7.763101583949748e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004881183543297566,"output_power":0.0003374582890540802,"reference_active":true,"delta_db":-1.603050381277769},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.883974713754685e-05,"output_power":5.182807236869525e-05,"reference_active":true,"delta_db":-0.5510574117837264},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7824879244994614e-06,"output_power":1.8203229128458133e-06,"reference_active":true,"delta_db":0.09121839295436361}],"active_band_count":3,"active_band_mae_db":0.7484420620052864}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00017587844894922502,"log_rms_mae_db":1.423504619632462,"rms_envelope_nrmse":0.1256410823644858,"envelope_correlation":0.9930904294057453,"rms_gain_error_db":-0.8101829819208795,"reference_mean_power":0.002090015296145697,"output_mean_power":0.0017343277234036908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":6.927253047980857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041075223634869497,"output_power":0.00029452379295621126,"reference_active":true,"delta_db":-1.444595518372338},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.156045953139752e-05,"output_power":7.140112695985934e-05,"reference_active":true,"delta_db":-0.5777459751509897},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2217147797459886e-05,"output_power":1.4600665406163301e-05,"reference_active":true,"delta_db":-1.823156555033826}],"active_band_count":3,"active_band_mae_db":1.2818326828523847}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00016316895420677254,"log_rms_mae_db":1.0413111143208638,"rms_envelope_nrmse":0.11729283954319972,"envelope_correlation":0.994494229706467,"rms_gain_error_db":-0.8101804972360892,"reference_mean_power":0.002117036355483854,"output_mean_power":0.001756751230555997},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":5.84659163450443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037072755253225205,"output_power":0.00026330146068464665,"reference_active":true,"delta_db":-1.4860159567182063},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.09979708848843e-05,"output_power":3.633225691832094e-05,"reference_active":true,"delta_db":-0.5246998610983694},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.223097583568823e-06,"output_power":9.049201563428009e-07,"reference_active":true,"delta_db":-1.308508462834409}],"active_band_count":3,"active_band_mae_db":1.1064080935503282}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00010383940879839973,"log_rms_mae_db":1.7372442743929961,"rms_envelope_nrmse":0.1880963537654837,"envelope_correlation":0.9828263679450315,"rms_gain_error_db":-1.2163376768663607,"reference_mean_power":0.0005521302842545546,"output_mean_power":0.0004172610062849742},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":2.374888225952349e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001363070736426414,"output_power":8.84699224377512e-05,"reference_active":true,"delta_db":-1.8772274756800473},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2931511209828857e-05,"output_power":2.6274526124530616e-05,"reference_active":true,"delta_db":-0.9807676868380724},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.586128694071041e-05,"output_power":2.1749469282123325e-05,"reference_active":true,"delta_db":1.3711024216141443}],"active_band_count":3,"active_band_mae_db":1.4096991947107547}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":1.3922714963635816,"rms_envelope_nrmse":0.16724252752943852,"envelope_correlation":0.9888352800388889,"rms_gain_error_db":-1.2163868804039455,"reference_mean_power":0.0005592400075632056,"output_mean_power":0.00042262924393808425},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.9436382041482275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261572491396227,"output_power":8.453405480146294e-05,"reference_active":true,"delta_db":-1.738805095725589},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5195698098116377e-05,"output_power":1.3110233694497935e-05,"reference_active":true,"delta_db":-0.6411022337412642},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.080674445150943e-07,"output_power":1.159991297571526e-06,"reference_active":true,"delta_db":1.0633662520469318}],"active_band_count":3,"active_band_mae_db":1.147757860504595}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":1.6413048166018278,"rms_envelope_nrmse":0.2197054722992053,"envelope_correlation":0.9819701851133575,"rms_gain_error_db":-1.571225461830443,"reference_mean_power":0.00026303793349419816,"output_mean_power":0.00018318750082614886},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":1.3356760090821463e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8507771208283054e-05,"output_power":4.2791656812894896e-05,"reference_active":true,"delta_db":-1.3585445255716606},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1981263949052123e-05,"output_power":9.413880742381918e-06,"reference_active":true,"delta_db":-1.0473394360085608},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.2561065186681e-06,"output_power":1.0776900406813126e-05,"reference_active":true,"delta_db":0.6606552560026111}],"active_band_count":3,"active_band_mae_db":1.0221797391942775}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":1.3211948398211608,"rms_envelope_nrmse":0.2049392174157379,"envelope_correlation":0.986961889474163,"rms_gain_error_db":-1.571242660602031,"reference_mean_power":0.00026641634082957393,"output_mean_power":0.00018553959018571584},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.157398705126345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.177699204483556e-05,"output_power":3.9274263415864476e-05,"reference_active":true,"delta_db":-1.2002876809151206},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.036512028004997e-06,"output_power":5.57216285005304e-06,"reference_active":true,"delta_db":-1.0133363389984913},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.58854722653168e-07,"output_power":5.695626472187182e-07,"reference_active":true,"delta_db":0.08242574544463753}],"active_band_count":3,"active_band_mae_db":0.7653499217860831}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.6687114238739014,"predicted_mos":2.6687114238739014,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":103936,"input_sample_rate":44100,"input_seconds":2.3568253968253967,"input_float32_sha256":"eae07f65717c72ffa138531563909ed4aef190650f69cd84c729fa9ad78e053b","internal_samples":37710,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.715912546002073},"started_at":"2026-10-13T07:33:18.035024+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.014260143998399144,"native_input_samples":103856,"native_input_sha256":"d150d3d5cb017b3e1bb5957351abbe718bb4419c15dd6579dcefcf6bba52c038","output_sha256":"f465d0594b1d6014e7ce740a56c0b40363e61e52baf87b2feec8fddfa87876ff","output_samples":103936,"length_delta_samples":80,"analysis_s":0.10124592600186588,"synthesis_s":13.042561502999888,"total_s":13.143807429001754,"rtf":5.581207707007561,"output_peak":0.38139334321022034,"output_rms":0.051561538711692625,"finished_at":"2026-10-13T07:33:31.238118+00:00","evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"f465d0594b1d6014e7ce740a56c0b40363e61e52baf87b2feec8fddfa87876ff","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":37680,"output_samples":37710,"length_delta_samples":30,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.748522165750785,"output_peak":0.3725621484745423},"common_prefix":{"samples":37680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9992044550517104,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.676576849042066,"spectral_convergence":0.46615847445073794,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.767755410517418,"spectral_convergence":0.5014192052016336,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.615928219660308,"spectral_convergence":0.5878155055611358,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":5.1199499989825356e-05,"delta_db":-1.801307714087106,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.0013286205472341737,"delta_db":-0.8118717270140308,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0008070510996563471,"delta_db":0.7196834033003495,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.00012426766554214988,"delta_db":-4.321510988408701,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.00024119318696538182,"delta_db":-1.6607861960406576,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.00011141458472690601,"delta_db":-2.757015231909265,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.0120292101266846,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8811474198595238,"pitch":null,"output_rms_dbfs":-25.745070941997696,"output_peak":0.3725621484745423,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.686753493073263},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37680,"output_samples":37680,"intelligibility":{"estoi":{"value":0.729458733237965,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05712205511060183,"rms_dbfs":-24.863923522138165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.476244606491736,"mean":-4.511304721710788e-05,"zero_energy":false},"output":{"rms":0.0516114966273483,"rms_dbfs":-25.745070941997696,"peak":0.3725621484745423,"peak_dbfs":-8.576025411966679,"crest_factor_db":17.169045530031013,"mean":-4.317562587283911e-05,"zero_energy":false},"rms_difference_db":-0.8811474198595306},"full_scale":{"reference":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":1.3753353733572873e-06,"output_rms_dbfs":-58.615913868965215,"output_zero_energy":false,"output_peak":0.04636823467177596,"output_energy_fraction":8.550458488227113e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":73,"both_estimated_voiced_frames":61,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.8243243243243243,"absolute_difference_cents_median":14.988727510065008,"absolute_difference_cents_p95":1230.98850191413,"difference_over_600_cents_frames":4,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00018834951637393047,"log_rms_mae_db":1.7999203420172898,"rms_envelope_nrmse":0.13813987166139383,"envelope_correlation":0.9903008331331293,"rms_gain_error_db":-0.8811442520462853,"reference_mean_power":0.00326985830085679,"output_mean_power":0.00266940523439718},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.479011996203747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005558024994761568,"output_power":0.0004814981366311413,"reference_active":true,"delta_db":-0.6232588472302703},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011456565813557757,"output_power":0.0001050663594547952,"reference_active":true,"delta_db":-0.37590770072082735},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.821714301732997e-05,"output_power":2.218968534228044e-05,"reference_active":true,"delta_db":-2.3610707374130264}],"active_band_count":3,"active_band_mae_db":1.1200790951213746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00017471448177334785,"log_rms_mae_db":1.4792546809036788,"rms_envelope_nrmse":0.11618435671725441,"envelope_correlation":0.9955175955724016,"rms_gain_error_db":-0.8811260253605023,"reference_mean_power":0.0033120735889998944,"output_mean_power":0.002703879760909323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":7.763101583949748e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004881183543297566,"output_power":0.0004276290081634586,"reference_active":true,"delta_db":-0.5745798054557374},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.883974713754685e-05,"output_power":5.383701417492824e-05,"reference_active":true,"delta_db":-0.3858983197290882},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7824879244994614e-06,"output_power":1.152653410706927e-06,"reference_active":true,"delta_db":-1.8932785670454906}],"active_band_count":3,"active_band_mae_db":0.9512522307434388}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00017587844894922502,"log_rms_mae_db":2.160953236440188,"rms_envelope_nrmse":0.14672578072503228,"envelope_correlation":0.9827651302547918,"rms_gain_error_db":-0.21164948197645503,"reference_mean_power":0.002090015296145697,"output_mean_power":0.0019906023680912694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":6.927253047980857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041075223634869497,"output_power":0.00037230916400003797,"reference_active":true,"delta_db":-0.42676210584799645},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.156045953139752e-05,"output_power":9.059286869778028e-05,"reference_active":true,"delta_db":0.4561434813463461},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2217147797459886e-05,"output_power":1.2611532563520154e-05,"reference_active":true,"delta_db":-2.459204385471803}],"active_band_count":3,"active_band_mae_db":1.114036657555382}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00016316895420677254,"log_rms_mae_db":1.6299335646399917,"rms_envelope_nrmse":0.1154774906552951,"envelope_correlation":0.988670401448694,"rms_gain_error_db":-0.2116592091033156,"reference_mean_power":0.002117036355483854,"output_mean_power":0.0020163336371861193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":5.84659163450443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037072755253225205,"output_power":0.00033572445276264755,"reference_active":true,"delta_db":-0.4307188947519259},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.09979708848843e-05,"output_power":4.4407645223057434e-05,"reference_active":true,"delta_db":0.3469538202943239},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.223097583568823e-06,"output_power":8.84234819545544e-07,"reference_active":true,"delta_db":-1.4089349552542676}],"active_band_count":3,"active_band_mae_db":0.7288692234335058}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00010383940879839973,"log_rms_mae_db":3.3614943716626318,"rms_envelope_nrmse":0.40330019904067815,"envelope_correlation":0.9213965545580152,"rms_gain_error_db":-3.247646577806536,"reference_mean_power":0.0005521302842545546,"output_mean_power":0.00026138274286359995},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":2.374888225952349e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001363070736426414,"output_power":6.799576086448916e-05,"reference_active":true,"delta_db":-3.020365562451453},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2931511209828857e-05,"output_power":1.7313304510190387e-05,"reference_active":true,"delta_db":-2.792316931539393},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.586128694071041e-05,"output_power":9.695607459056933e-06,"reference_active":true,"delta_db":-2.137633977310214}],"active_band_count":3,"active_band_mae_db":2.650105490433687}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":3.0053484092000167,"rms_envelope_nrmse":0.36641383372708153,"envelope_correlation":0.9506028926762548,"rms_gain_error_db":-3.2477109674925564,"reference_mean_power":0.0005592400075632056,"output_mean_power":0.0002647446156095018},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.9436382041482275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261572491396227,"output_power":6.389120518565295e-05,"reference_active":true,"delta_db":-2.9547113028156273},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5195698098116377e-05,"output_power":8.62273906538545e-06,"reference_active":true,"delta_db":-2.4607541256606367},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.080674445150943e-07,"output_power":4.949867809920748e-07,"reference_active":true,"delta_db":-2.6352450497651025}],"active_band_count":3,"active_band_mae_db":2.683570159413789}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":4.629883080784468,"rms_envelope_nrmse":0.40326021302274134,"envelope_correlation":0.9024084207680904,"rms_gain_error_db":-2.417579990385899,"reference_mean_power":0.00026303793349419816,"output_mean_power":0.00015075106358867284},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":1.3356760090821463e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8507771208283054e-05,"output_power":3.8157596297609184e-05,"reference_active":true,"delta_db":-1.8563254552086892},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1981263949052123e-05,"output_power":8.243024073074047e-06,"reference_active":true,"delta_db":-1.624160676780294},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.2561065186681e-06,"output_power":4.380556888458415e-06,"reference_active":true,"delta_db":-3.2489901909745567}],"active_band_count":3,"active_band_mae_db":2.2431587743211803}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":4.342340476853415,"rms_envelope_nrmse":0.3790690032531978,"envelope_correlation":0.9138448194550867,"rms_gain_error_db":-2.4175062054292753,"reference_mean_power":0.00026641634082957393,"output_mean_power":0.00015268987463480863},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.157398705126345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.177699204483556e-05,"output_power":3.517745518180549e-05,"reference_active":true,"delta_db":-1.67872398344683},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.036512028004997e-06,"output_power":4.4016492394056575e-06,"reference_active":true,"delta_db":-2.0374200303040997},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.58854722653168e-07,"output_power":2.4158554710569047e-07,"reference_active":true,"delta_db":-3.642279763393227}],"active_band_count":3,"active_band_mae_db":2.452807925714719}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"golf","status":"measured","mos":{"status":"measured","value":1.7278876304626465,"predicted_mos":1.7278876304626465,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":56400,"input_sample_rate":24000,"input_seconds":2.35,"input_float32_sha256":"2f4ca09f4a0d82a7e1421caf401a93316e892816e239c63c8bca914fbf7eed9c","internal_samples":37600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.218672576000245},"started_at":"2026-10-13T05:03:10.696125+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0021006899996791617,"native_input_samples":56520,"native_input_sha256":"254566f4635b29fcd4a3669efb7624e30f0f7a8246fbfe7d49e8fd3681757664","output_sha256":"0518dc189168412d8a1d894d6c2922aabd756b1c419d5afb6a25dca13f021eda","output_samples":56400,"length_delta_samples":-120,"analysis_s":0.2235945869997522,"synthesis_s":0.07151360300031229,"total_s":0.2951081900000645,"rtf":0.12531133333336072,"output_peak":0.3545818328857422,"output_rms":0.047509767972538856,"finished_at":"2026-10-13T05:03:11.000235+00:00","analysis_stages":{"external_f0_s":0.021609970999634243},"evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"0518dc189168412d8a1d894d6c2922aabd756b1c419d5afb6a25dca13f021eda","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":37680,"output_samples":37600,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.458434524039028,"output_peak":0.34858634721147697},"common_prefix":{"samples":37600,"sample_rate":16000,"reference_coverage":0.9978768577494692,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37600,"output_samples":37600,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.499049576487363,"spectral_convergence":0.4731062810707116,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.775833691239294,"spectral_convergence":0.383000470211012,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.913115180076978,"spectral_convergence":0.421256700039113,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.768297668669443e-05,"output_power":9.164163632116122e-05,"delta_db":0.7176697651302966,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016050902953686857,"output_power":0.001271033468145961,"delta_db":-1.0134248263341075,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006853023201788927,"output_power":0.00041887007926278563,"delta_db":-2.138028632438825,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003368413220822589,"output_power":0.00019113218212968877,"delta_db":-2.460915451555115,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.0003542878292201315,"output_power":0.00021392106819995024,"delta_db":-2.1910267481425167,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.0002106494645413349,"output_power":7.365192858780904e-05,"delta_db":-4.56376236113344,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":2.18080463078905,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.6037184174958665,"pitch":null,"output_rms_dbfs":-26.458434524039028,"output_peak":0.34858634721147697,"reference_rms_dbfs":-24.854716106543165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.062666149267879},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37600,"output_samples":37600,"intelligibility":{"estoi":{"value":0.8879564925543229,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.057182639044363784,"rms_dbfs":-24.854716106543165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.46703719089673,"mean":-4.490466828041888e-05,"zero_energy":false},"output":{"rms":0.04754209043203039,"rms_dbfs":-26.458434524039028,"peak":0.34858634721147697,"peak_dbfs":-9.153792530798837,"crest_factor_db":17.304641993240192,"mean":0.00044117636106942213,"zero_energy":false},"rms_difference_db":-1.6037184174958625},"full_scale":{"reference":{"sample_count":37600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37600,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":7.215710995156571e-07,"output_rms_dbfs":-61.417208695517914,"output_zero_energy":false,"output_peak":0.00431445177049796,"output_energy_fraction":5.298089952573613e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":90,"both_estimated_voiced_frames":74,"reference_voiced_output_unvoiced_frames":0,"reference_unvoiced_output_voiced_frames":16,"paired_voiced_fraction":1.0,"absolute_difference_cents_median":7.589582888961751,"absolute_difference_cents_p95":1190.0158943003812,"difference_over_600_cents_frames":6,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37600,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":0.00018834951637393047,"log_rms_mae_db":2.2491889660149926,"rms_envelope_nrmse":0.22907973216299485,"envelope_correlation":0.9735093155529522,"rms_gain_error_db":-1.6037090770380338,"reference_mean_power":0.0032768122163781923,"output_mean_power":0.0022650648800311315},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":9.503913506079563e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005574871994857403,"output_power":0.000347216938422647,"reference_active":true,"delta_db":-2.0563399657711803},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011495840341296832,"output_power":7.421042093954087e-05,"reference_active":true,"delta_db":-1.9007582849758244},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.812243145660464e-05,"output_power":1.6968425977191066e-05,"reference_active":true,"delta_db":-3.51539034021267}],"active_band_count":3,"active_band_mae_db":2.4908295303198913}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":0.00017471448177334785,"log_rms_mae_db":2.100024226466246,"rms_envelope_nrmse":0.21538424816277815,"envelope_correlation":0.9784763537142058,"rms_gain_error_db":-1.6036787990762778,"reference_mean_power":0.0033192171875491023,"output_mean_power":0.00229439290377587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":7.783924735619665e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004894853516222588,"output_power":0.00030171204863121227,"reference_active":true,"delta_db":-2.1014704587550503},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.895982789205265e-05,"output_power":3.8181722014837694e-05,"reference_active":true,"delta_db":-1.8870069571765715},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7769258048532913e-06,"output_power":8.19621917041753e-07,"reference_active":true,"delta_db":-3.36055731207944}],"active_band_count":3,"active_band_mae_db":2.449678242670354}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":0.00017587844894922502,"log_rms_mae_db":2.242934445761668,"rms_envelope_nrmse":0.26026254304823004,"envelope_correlation":0.9563052588402636,"rms_gain_error_db":-1.3713001723477782,"reference_mean_power":0.00209447099604406,"output_mean_power":0.0015273702720063906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":6.944856689284085e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041189394951405325,"output_power":0.000259797867261524,"reference_active":true,"delta_db":-2.001498309347033},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.182834395957036e-05,"output_power":5.576884483820313e-05,"reference_active":true,"delta_db":-1.665121131477484},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.230303343595139e-05,"output_power":8.957878582994767e-06,"reference_active":true,"delta_db":-3.9615876383711197}],"active_band_count":3,"active_band_mae_db":2.5427356930652123}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":0.00016316895420677254,"log_rms_mae_db":1.9490584832828144,"rms_envelope_nrmse":0.2432652588506894,"envelope_correlation":0.9616713015877093,"rms_gain_error_db":-1.3712963115007604,"reference_mean_power":0.002121608310668317,"output_mean_power":0.0015471612399709468},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":5.860764347710188e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003717646647029237,"output_power":0.00022961420736774632,"reference_active":true,"delta_db":-2.092693522914652},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.1030045705526297e-05,"output_power":2.8753470375455438e-05,"reference_active":true,"delta_db":-1.544117317613925},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.2213168068844765e-06,"output_power":6.065274542680215e-07,"reference_active":true,"delta_db":-3.0397786974279724}],"active_band_count":3,"active_band_mae_db":2.2255298459855166}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":0.00010383940879839973,"log_rms_mae_db":2.345364639125196,"rms_envelope_nrmse":0.2906890256043424,"envelope_correlation":0.9670919838888387,"rms_gain_error_db":-2.2895932260487513,"reference_mean_power":0.0005532983814241691,"output_mean_power":0.00032658789019638035},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":2.382134345969998e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013686314014611198,"output_power":8.162340832370047e-05,"reference_active":true,"delta_db":-2.2447177457687735},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.300379381010628e-05,"output_power":1.9267767931598733e-05,"reference_active":true,"delta_db":-2.337324583497665},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5710274287380966e-05,"output_power":6.252584824865195e-06,"reference_active":true,"delta_db":-4.001241752594156}],"active_band_count":3,"active_band_mae_db":2.861094693953531}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":2.2197547256855357,"rms_envelope_nrmse":0.2659358819999583,"envelope_correlation":0.9795953720668802,"rms_gain_error_db":-2.289531869464065,"reference_mean_power":0.0005604430217518075,"output_mean_power":0.00033080973321590867},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":1.950018749794791e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012653329963244582,"output_power":7.437305630652842e-05,"reference_active":true,"delta_db":-2.3078920449626934},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5196128683512429e-05,"output_power":8.63766664230971e-06,"reference_active":true,"delta_db":-2.4533652348523463},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.018359466107086e-07,"output_power":3.373242688284389e-07,"reference_active":true,"delta_db":-4.270799543456542}],"active_band_count":3,"active_band_mae_db":3.010685607757194}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":469,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":5.218988777111204,"rms_envelope_nrmse":0.3930151847204657,"envelope_correlation":0.9775290784107069,"rms_gain_error_db":-3.962936147571141,"reference_mean_power":0.0002635919406372261,"output_mean_power":0.00010583724167111209},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":469,"frequency_resolution_hz":0.42643923240938164,"power_floor":1.3417377199499141e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.871496167924587e-05,"output_power":2.2114553440045835e-05,"reference_active":true,"delta_db":-4.240706077016769},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2057378071386843e-05,"output_power":5.74058951166907e-06,"reference_active":true,"delta_db":-3.2229638568640437},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.224405551055847e-06,"output_power":3.139354470627699e-06,"reference_active":true,"delta_db":-4.680980329618805}],"active_band_count":3,"active_band_mae_db":4.048216754499872}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37600,"uncovered_tail_samples":0,"metrics":{"frame_count":463,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":5.1914009328109385,"rms_envelope_nrmse":0.3813369090102498,"envelope_correlation":0.9869215299606457,"rms_gain_error_db":-3.962812142629924,"reference_mean_power":0.00026698848180002426,"output_mean_power":0.00010720407934975177},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":463,"frequency_resolution_hz":0.4319654427645788,"power_floor":1.1627885738276739e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.193421891748857e-05,"output_power":1.986503341563994e-05,"reference_active":true,"delta_db":-4.173643043517044},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.0472513236065476e-06,"output_power":3.1848745110039228e-06,"reference_active":true,"delta_db":-3.4492743473020058},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.571862927997389e-07,"output_power":1.5612413099429484e-07,"reference_active":true,"delta_db":-5.525303898601424}],"active_band_count":3,"active_band_mae_db":4.382740429806824}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.463343620300293,"predicted_mos":2.463343620300293,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":56640,"input_sample_rate":24000,"input_seconds":2.36,"input_float32_sha256":"885a6acf57c37ac18db12ab33fe34177c898a10eb0110ba913d551dfc43cc0b8","internal_samples":37760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7657898739998927},"started_at":"2026-10-13T07:47:39.593676+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031519139993179124,"native_input_samples":56520,"native_input_sha256":"254566f4635b29fcd4a3669efb7624e30f0f7a8246fbfe7d49e8fd3681757664","output_sha256":"4492ed5d0f40650ca70ef62836ec1e17398e689aa6dbebb3d4e5acca0930407e","output_samples":56640,"length_delta_samples":120,"analysis_s":0.98975443299787,"synthesis_s":6.617121944000246,"total_s":7.606876376998116,"rtf":3.2300961261138497,"output_peak":0.4585331082344055,"output_rms":0.05505552085658157,"finished_at":"2026-10-13T07:47:47.214945+00:00","analysis_stages":{"external_f0_s":0.6400017059968377},"evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"4492ed5d0f40650ca70ef62836ec1e17398e689aa6dbebb3d4e5acca0930407e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":37680,"output_samples":37760,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.178345719651762,"output_peak":0.43897030988978064},"common_prefix":{"samples":37680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9978813559322034,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.172771629809032,"spectral_convergence":0.5177119304331915,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.264617039461214,"spectral_convergence":0.3623905347350273,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.292578321555503,"spectral_convergence":0.4253184810180258,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":9.27780857970767e-05,"delta_db":0.7804891676540243,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.0015276249829348002,"delta_db":-0.20571385005793946,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0006065815608978498,"delta_db":-0.5204348894931116,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.00031935622014315015,"delta_db":-0.2223385933093147,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.0003235061011167658,"delta_db":-0.3856118004173289,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.00017163137341857088,"delta_db":-0.8804689128665422,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.49917620229971016,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3052292206101333,"pitch":null,"output_rms_dbfs":-25.1691527427483,"output_peak":0.43897030988978064,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.576655663608583},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37680,"output_samples":37680,"intelligibility":{"estoi":{"value":0.8961561438633373,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05712205511060183,"rms_dbfs":-24.863923522138165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.476244606491736,"mean":-4.511304721710788e-05,"zero_energy":false},"output":{"rms":0.0551495994936338,"rms_dbfs":-25.1691527427483,"peak":0.43897030988978064,"peak_dbfs":-7.151297052427826,"crest_factor_db":18.017855690320477,"mean":-0.00014688751807764522,"zero_energy":false},"rms_difference_db":-0.30522922061013347},"full_scale":{"reference":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":3.310379558176361e-05,"output_rms_dbfs":-44.80122208468643,"output_zero_energy":false,"output_peak":0.20914589251613644,"output_energy_fraction":0.001802464671961093,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":61,"both_estimated_voiced_frames":53,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":8,"paired_voiced_fraction":0.7162162162162162,"absolute_difference_cents_median":8.024451936127301,"absolute_difference_cents_p95":22.280655634911867,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00018834951637393047,"log_rms_mae_db":1.945384321229053,"rms_envelope_nrmse":0.16119592945742411,"envelope_correlation":0.9761141703641382,"rms_gain_error_db":-0.3052216931610189,"reference_mean_power":0.00326985830085679,"output_mean_power":0.0030479424573555737},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.479011996203747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005558024994761568,"output_power":0.0004981164971686459,"reference_active":true,"delta_db":-0.4758957008158537},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011456565813557757,"output_power":9.881050119258762e-05,"reference_active":true,"delta_db":-0.6425135214880011},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.821714301732997e-05,"output_power":2.9411502614770445e-05,"reference_active":true,"delta_db":-1.1374100466511312}],"active_band_count":3,"active_band_mae_db":0.7519397563183287}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00017471448177334785,"log_rms_mae_db":1.2766865806080858,"rms_envelope_nrmse":0.097920009956375,"envelope_correlation":0.9911282368486657,"rms_gain_error_db":-0.30519835085554825,"reference_mean_power":0.0033120735889998944,"output_mean_power":0.00308730930927463},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":7.763101583949748e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004881183543297566,"output_power":0.0004354900957981115,"reference_active":true,"delta_db":-0.49546855951120117},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.883974713754685e-05,"output_power":5.044525614253492e-05,"reference_active":true,"delta_db":-0.6685046612477376},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7824879244994614e-06,"output_power":1.654454592734247e-06,"reference_active":true,"delta_db":-0.3237174414908434}],"active_band_count":3,"active_band_mae_db":0.49589688741659405}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00017587844894922502,"log_rms_mae_db":2.197354961980978,"rms_envelope_nrmse":0.16386057658404757,"envelope_correlation":0.9785499192415047,"rms_gain_error_db":-0.2876917571106575,"reference_mean_power":0.002090015296145697,"output_mean_power":0.0019560515176837603},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":6.927253047980857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041075223634869497,"output_power":0.0003651242060693868,"reference_active":true,"delta_db":-0.5113931099114202},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.156045953139752e-05,"output_power":7.166181491569145e-05,"reference_active":true,"delta_db":-0.5619186107002854},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2217147797459886e-05,"output_power":1.899474705381359e-05,"reference_active":true,"delta_db":-0.6805478955338486}],"active_band_count":3,"active_band_mae_db":0.5846198720485181}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00016316895420677254,"log_rms_mae_db":1.4862008262192499,"rms_envelope_nrmse":0.10572756844756015,"envelope_correlation":0.9909359250447768,"rms_gain_error_db":-0.2876888336359345,"reference_mean_power":0.002117036355483854,"output_mean_power":0.0019813419410517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":5.84659163450443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037072755253225205,"output_power":0.0003273940913010472,"reference_active":true,"delta_db":-0.5398402692062079},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.09979708848843e-05,"output_power":3.6129278475920464e-05,"reference_active":true,"delta_db":-0.5490307416320569},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.223097583568823e-06,"output_power":1.2979237669072554e-06,"reference_active":true,"delta_db":0.25788076928625114}],"active_band_count":3,"active_band_mae_db":0.4489172600415053}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00010383940879839973,"log_rms_mae_db":2.2923442569381183,"rms_envelope_nrmse":0.23718546311741773,"envelope_correlation":0.9535024832863648,"rms_gain_error_db":-0.3530017427723946,"reference_mean_power":0.0005521302842545546,"output_mean_power":0.0005090276745769428},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":2.374888225952349e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001363070736426414,"output_power":0.00011749577329030028,"reference_active":true,"delta_db":-0.6449615020361166},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2931511209828857e-05,"output_power":3.0041971534701507e-05,"reference_active":true,"delta_db":-0.39883230434815464},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.586128694071041e-05,"output_power":1.3006852796310637e-05,"reference_active":true,"delta_db":-0.8616619666752662}],"active_band_count":3,"active_band_mae_db":0.6351519243531792}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":1.4877919975139882,"rms_envelope_nrmse":0.13098826225505408,"envelope_correlation":0.9851320489398437,"rms_gain_error_db":-0.3529565113174038,"reference_mean_power":0.0005592400075632056,"output_mean_power":0.0005155877399177564},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.9436382041482275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261572491396227,"output_power":0.00010897455482993076,"reference_active":true,"delta_db":-0.6358710706307363},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5195698098116377e-05,"output_power":1.3782439984548538e-05,"reference_active":true,"delta_db":-0.42394546558261137},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.080674445150943e-07,"output_power":7.888761676112905e-07,"reference_active":true,"delta_db":-0.6110926991199956}],"active_band_count":3,"active_band_mae_db":0.5569697451111144}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":2.2233330494986117,"rms_envelope_nrmse":0.2371741659263008,"envelope_correlation":0.9584749255281975,"rms_gain_error_db":-0.5788666700295746,"reference_mean_power":0.00026303793349419816,"output_mean_power":0.00023021399262199168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":1.3356760090821463e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8507771208283054e-05,"output_power":5.339850720914259e-05,"reference_active":true,"delta_db":-0.39684438229896235},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1981263949052123e-05,"output_power":9.613676258113423e-06,"reference_active":true,"delta_db":-0.9561314273318735},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.2561065186681e-06,"output_power":6.606873844176942e-06,"reference_active":true,"delta_db":-1.4643232961064057}],"active_band_count":3,"active_band_mae_db":0.9390997019124138}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":1.3756932578026002,"rms_envelope_nrmse":0.130540710572497,"envelope_correlation":0.9888896433824127,"rms_gain_error_db":-0.5787440206189983,"reference_mean_power":0.00026641634082957393,"output_mean_power":0.00023317740082151906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.157398705126345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.177699204483556e-05,"output_power":4.6940839099033266e-05,"reference_active":true,"delta_db":-0.42585968101008564},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.036512028004997e-06,"output_power":5.3792738142667124e-06,"reference_active":true,"delta_db":-1.166337831948173},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.58854722653168e-07,"output_power":4.1513883334894036e-07,"reference_active":true,"delta_db":-1.2910556482285853}],"active_band_count":3,"active_band_mae_db":0.9610843870622814}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.660928726196289,"predicted_mos":2.660928726196289,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":56760,"input_sample_rate":24000,"input_seconds":2.365,"input_float32_sha256":"61d947b0dda5acf795753be7ae2fb89b20c1c1c236bce2a5c7cb4d4f2db9e156","internal_samples":37840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.827470952999647},"started_at":"2026-10-13T05:11:38.931978+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002416309000182082,"native_input_samples":56520,"native_input_sha256":"254566f4635b29fcd4a3669efb7624e30f0f7a8246fbfe7d49e8fd3681757664","output_sha256":"9abcaa542348d2fc00b5601f46c76b0b5af521313d7856e5da2f3b308df6177b","output_samples":56760,"length_delta_samples":240,"analysis_s":0.2856286149999505,"synthesis_s":0.06102282000028936,"total_s":0.34665143500023987,"rtf":0.1471980615712271,"output_peak":0.41856512427330017,"output_rms":0.05534196119528552,"finished_at":"2026-10-13T05:11:39.287522+00:00","analysis_stages":{"external_f0_s":0.02774495399899024},"evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"9abcaa542348d2fc00b5601f46c76b0b5af521313d7856e5da2f3b308df6177b","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":37680,"output_samples":37840,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.133107141986688,"output_peak":0.4170266897269728},"common_prefix":{"samples":37680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9957716701902748,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.406272981441878,"spectral_convergence":0.22172410510467527,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.062275981111873,"spectral_convergence":0.19773701381487077,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.492946509917653,"spectral_convergence":0.23229008860526668,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":8.464744578487031e-05,"delta_db":0.38217366796746227,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.0015627492866306316,"delta_db":-0.10698827934818683,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0006369702928573381,"delta_db":-0.30813515173746964,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.0002872849807439612,"delta_db":-0.6819631704403852,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.000318171581876988,"delta_db":-0.4578226846431583,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.0001900229813154402,"delta_db":-0.43837441514153963,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.395909561546367,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.25078566287536735,"pitch":null,"output_rms_dbfs":-25.114709185013538,"output_peak":0.4170266897269728,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.320498490823801},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37680,"output_samples":37680,"intelligibility":{"estoi":{"value":0.9527797843000423,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05712205511060183,"rms_dbfs":-24.863923522138165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.476244606491736,"mean":-4.511304721710788e-05,"zero_energy":false},"output":{"rms":0.055496365369357564,"rms_dbfs":-25.11470918501353,"peak":0.4170266897269728,"peak_dbfs":-7.596722985406328,"crest_factor_db":17.517986199607208,"mean":-8.136155974468077e-06,"zero_energy":false},"rms_difference_db":-0.25078566287536574},"full_scale":{"reference":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":9.248674279053599e-06,"output_rms_dbfs":-50.339205153208944,"output_zero_energy":false,"output_peak":0.07766898457398558,"output_energy_fraction":0.0004973064577068289,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":73,"both_estimated_voiced_frames":71,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9594594594594594,"absolute_difference_cents_median":0.9464951899408819,"absolute_difference_cents_p95":5.326181755387951,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00018834951637393047,"log_rms_mae_db":0.752083130009788,"rms_envelope_nrmse":0.07223515710787218,"envelope_correlation":0.9957894392111513,"rms_gain_error_db":-0.25078177875602525,"reference_mean_power":0.00326985830085679,"output_mean_power":0.00308638965808906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.479011996203747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005558024994761568,"output_power":0.0005005931348296243,"reference_active":true,"delta_db":-0.4543560612610234},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011456565813557757,"output_power":9.80675538068023e-05,"reference_active":true,"delta_db":-0.6752911189297693},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.821714301732997e-05,"output_power":2.6715449797253582e-05,"reference_active":true,"delta_db":-1.554957269434942}],"active_band_count":3,"active_band_mae_db":0.894868149875245}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00017471448177334785,"log_rms_mae_db":0.4695670526124999,"rms_envelope_nrmse":0.0471735420538387,"envelope_correlation":0.9985340064596959,"rms_gain_error_db":-0.25077127932510745,"reference_mean_power":0.0033120735889998944,"output_mean_power":0.0031262438444664407},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":7.763101583949748e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004881183543297566,"output_power":0.00044234005895014755,"reference_active":true,"delta_db":-0.42768866914895315},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.883974713754685e-05,"output_power":4.978239662381264e-05,"reference_active":true,"delta_db":-0.7259499711103691},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7824879244994614e-06,"output_power":1.2708444279055725e-06,"reference_active":true,"delta_db":-1.4693420732024565}],"active_band_count":3,"active_band_mae_db":0.8743269044872596}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00017587844894922502,"log_rms_mae_db":0.9093651855534252,"rms_envelope_nrmse":0.06905509610936304,"envelope_correlation":0.9963408005372758,"rms_gain_error_db":-0.14748841908438243,"reference_mean_power":0.002090015296145697,"output_mean_power":0.0020202291006849254},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":6.927253047980857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041075223634869497,"output_power":0.00039298620929587556,"reference_active":true,"delta_db":-0.19202626293024108},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.156045953139752e-05,"output_power":7.4482697367444e-05,"reference_active":true,"delta_db":-0.39424267937306856},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2217147797459886e-05,"output_power":1.6369196256248827e-05,"reference_active":true,"delta_db":-1.3266094849368297}],"active_band_count":3,"active_band_mae_db":0.6376261424133798}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00016316895420677254,"log_rms_mae_db":0.5190726118667627,"rms_envelope_nrmse":0.04042248410115526,"envelope_correlation":0.9988260501779205,"rms_gain_error_db":-0.14748746305076468,"reference_mean_power":0.002117036355483854,"output_mean_power":0.002046348369764708},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":5.84659163450443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037072755253225205,"output_power":0.000357649422591711,"reference_active":true,"delta_db":-0.15597335833000164},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.09979708848843e-05,"output_power":3.6311545098242855e-05,"reference_active":true,"delta_db":-0.5271763364931379},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.223097583568823e-06,"output_power":9.453516449117685e-07,"reference_active":true,"delta_db":-1.1186772394148794}],"active_band_count":3,"active_band_mae_db":0.6006089780793397}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00010383940879839973,"log_rms_mae_db":1.0683161765312799,"rms_envelope_nrmse":0.14552373455113016,"envelope_correlation":0.985864593185797,"rms_gain_error_db":-0.6428015583502764,"reference_mean_power":0.0005521302842545546,"output_mean_power":0.0004761693232196865},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":2.374888225952349e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001363070736426414,"output_power":0.00010212394314271463,"reference_active":true,"delta_db":-1.253908189407266},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2931511209828857e-05,"output_power":2.4161928336695325e-05,"reference_active":true,"delta_db":-1.3448006882280006},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.586128694071041e-05,"output_power":1.0038195892268736e-05,"reference_active":true,"delta_db":-1.9868275530725152}],"active_band_count":3,"active_band_mae_db":1.5285121435692606}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":0.7880749411226783,"rms_envelope_nrmse":0.10580386671477199,"envelope_correlation":0.9939811205703737,"rms_gain_error_db":-0.6427978244779243,"reference_mean_power":0.0005592400075632056,"output_mean_power":0.000482301319923593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.9436382041482275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261572491396227,"output_power":9.275545780481609e-05,"reference_active":true,"delta_db":-1.335727373563231},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5195698098116377e-05,"output_power":1.087288279166178e-05,"reference_active":true,"delta_db":-1.4537595016813312},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.080674445150943e-07,"output_power":5.852350205282955e-07,"reference_active":true,"delta_db":-1.9078779943152968}],"active_band_count":3,"active_band_mae_db":1.5657882898532864}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":1.1110712585372224,"rms_envelope_nrmse":0.11229056923044517,"envelope_correlation":0.9914177105475436,"rms_gain_error_db":-0.34532104483732107,"reference_mean_power":0.00026303793349419816,"output_mean_power":0.00024293286916196003},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":1.3356760090821463e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8507771208283054e-05,"output_power":5.459580162530462e-05,"reference_active":true,"delta_db":-0.30054307355683424},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1981263949052123e-05,"output_power":1.1172124659549496e-05,"reference_active":true,"delta_db":-0.3036686289288325},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.2561065186681e-06,"output_power":5.407706065453676e-06,"reference_active":true,"delta_db":-2.3341526620062982}],"active_band_count":3,"active_band_mae_db":0.9794547881639883}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":0.7544380551846168,"rms_envelope_nrmse":0.06281088958703666,"envelope_correlation":0.9980016709685483,"rms_gain_error_db":-0.34526959002960034,"reference_mean_power":0.00026641634082957393,"output_mean_power":0.00024605596624765265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.157398705126345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.177699204483556e-05,"output_power":4.811398847098877e-05,"reference_active":true,"delta_db":-0.3186545694381107},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.036512028004997e-06,"output_power":6.271940930138541e-06,"reference_active":true,"delta_db":-0.4995547491334909},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.58854722653168e-07,"output_power":3.1348129285458573e-07,"reference_active":true,"delta_db":-2.5108729611815455}],"active_band_count":3,"active_band_mae_db":1.109694093251049}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8463-294828-0022","method":"magphase","status":"measured","mos":{"status":"measured","value":3.531409502029419,"predicted_mos":3.531409502029419,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":113061,"input_sample_rate":48000,"input_seconds":2.3554375,"input_float32_sha256":"d6590f00d9038e345fa6a8f1611972545031ba5d197b2189b77837c05507644e","internal_samples":37687,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4758618859996204},"started_at":"2026-10-13T05:19:03.842209+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.004004730000815471,"native_input_samples":113040,"native_input_sha256":"cb1d91293e5ae3e0d2d7c673596adc0241c911d625034b022ae57b3cc89544d4","output_sha256":"4645a425774452a4d9eda8763c2e59996eebac5c9a558139348c89bb529f8b55","output_samples":113061,"length_delta_samples":21,"analysis_s":1.5405134159991576,"synthesis_s":0.7926580010007456,"total_s":2.333171416999903,"rtf":0.9907309626326553,"output_peak":0.5250921802473726,"output_rms":0.0566992125310682,"finished_at":"2026-10-13T05:19:06.185725+00:00","evaluation_input":{"reference":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"output_sha256":"4645a425774452a4d9eda8763c2e59996eebac5c9a558139348c89bb529f8b55","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":37680,"output_samples":37687,"length_delta_samples":7,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.922839369908665,"output_peak":0.5251912701644331},"common_prefix":{"samples":37680,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9998142595589992,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":37680,"output_samples":37680,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.666150691388558,"spectral_convergence":0.5448918179570635,"spectral_convergence_status":"ok","amplitude_floor":1.2997261015401668e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.013022063587432,"spectral_convergence":0.24128242813997963,"spectral_convergence_status":"ok","amplitude_floor":1.1046780177944747e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.080044237973593,"spectral_convergence":0.25552736607838006,"spectral_convergence_status":"ok","amplitude_floor":9.696590055390529e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":7.751691690693756e-05,"output_power":6.91987430407274e-05,"delta_db":-0.49298285326059704,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0016017256752476519,"output_power":0.0014691369164891862,"delta_db":-0.3752586542972498,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006838056632308089,"output_power":0.0006573696442216603,"delta_db":-0.17123047757397508,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0003361315194748372,"output_power":0.0003312943113800295,"delta_db":-0.06295259997264575,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00035354415012418854,"output_power":0.00037741327463933433,"delta_db":0.2837351576386363,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00021020525507420865,"output_power":0.00031514869181114945,"delta_db":1.758719396279684,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5241465231704647,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.05810911389408338,"pitch":null,"output_rms_dbfs":-24.922032636032256,"output_peak":0.5251912701644331,"reference_rms_dbfs":-24.863923522138165,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.586405664316527},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":37680,"output_samples":37680,"intelligibility":{"estoi":{"value":0.9504816064256298,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05712205511060183,"rms_dbfs":-24.863923522138165,"peak":0.42718505859375,"peak_dbfs":-7.387678915646433,"crest_factor_db":17.476244606491736,"mean":-4.511304721710788e-05,"zero_energy":false},"output":{"rms":0.056741180650230455,"rms_dbfs":-24.92203263603225,"peak":0.5251912701644331,"peak_dbfs":-5.593650029066416,"crest_factor_db":19.328382606965835,"mean":-0.0002953673811766249,"zero_energy":false},"rms_difference_db":-0.058109113894083464},"full_scale":{"reference":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":37680,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":235,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.469329969375394e-06,"threshold_dbfs":-54.59754392269826,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":39,"quiet_samples":6240,"reference_mean_power":9.852916730615574e-07,"output_mean_power":2.8269968975129575e-05,"output_rms_dbfs":-45.4867466812659,"output_zero_energy":false,"output_peak":0.1319207311434552,"output_energy_fraction":0.0014541268413254664,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":228,"reference_estimated_voiced_frames":74,"output_estimated_voiced_frames":77,"both_estimated_voiced_frames":71,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9594594594594594,"absolute_difference_cents_median":3.940063563511456,"absolute_difference_cents_p95":1200.7226695717488,"difference_over_600_cents_frames":6,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":37680,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00018834951637393047,"log_rms_mae_db":1.4727291225336263,"rms_envelope_nrmse":0.12798310572732707,"envelope_correlation":0.9847560958030551,"rms_gain_error_db":-0.05809816886809579,"reference_mean_power":0.00326985830085679,"output_mean_power":0.003226406738554202},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":9.479011996203747e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0005558024994761568,"output_power":0.000570383677646526,"reference_active":true,"delta_db":0.11246593563937077},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00011456565813557757,"output_power":0.00011623389207356471,"reference_active":true,"delta_db":0.06278326025851105},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.821714301732997e-05,"output_power":3.785804724166068e-05,"reference_active":true,"delta_db":-0.04100008730764246}],"active_band_count":3,"active_band_mae_db":0.07208309440184142}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00017471448177334785,"log_rms_mae_db":0.876784593974663,"rms_envelope_nrmse":0.059011264938566836,"envelope_correlation":0.9964707496533517,"rms_gain_error_db":-0.058110164006314136,"reference_mean_power":0.0033120735889998944,"output_mean_power":0.0032680520218763554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":7.763101583949748e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004881183543297566,"output_power":0.0005013146994229553,"reference_active":true,"delta_db":0.11585300676160067},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.883974713754685e-05,"output_power":5.9887640896089174e-05,"reference_active":true,"delta_db":0.076664079316864},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7824879244994614e-06,"output_power":1.9519054671861613e-06,"reference_active":true,"delta_db":0.3943218410899471}],"active_band_count":3,"active_band_mae_db":0.19561297572280392}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00017587844894922502,"log_rms_mae_db":1.5533695546618809,"rms_envelope_nrmse":0.11476646863619631,"envelope_correlation":0.9903011487919989,"rms_gain_error_db":-0.35358894992469136,"reference_mean_power":0.002090015296145697,"output_mean_power":0.0019265956243575745},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":6.927253047980857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041075223634869497,"output_power":0.00038966164352080974,"reference_active":true,"delta_db":-0.22892278669984828},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.156045953139752e-05,"output_power":7.611482514822851e-05,"reference_active":true,"delta_db":-0.3001040998282673},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2217147797459886e-05,"output_power":1.9982793118744182e-05,"reference_active":true,"delta_db":-0.46032111977770473}],"active_band_count":3,"active_band_mae_db":0.3297826687686068}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":0.00016316895420677254,"log_rms_mae_db":0.9406883992734032,"rms_envelope_nrmse":0.06163616975666403,"envelope_correlation":0.9980225198382238,"rms_gain_error_db":-0.353586121074436,"reference_mean_power":0.002117036355483854,"output_mean_power":0.0019515051604244227},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":5.84659163450443e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037072755253225205,"output_power":0.0003527714097814165,"reference_active":true,"delta_db":-0.2156148340341625},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.09979708848843e-05,"output_power":3.824323291792768e-05,"reference_active":true,"delta_db":-0.3020776415002848},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.223097583568823e-06,"output_power":1.2226279077040107e-06,"reference_active":true,"delta_db":-0.001668033796978198}],"active_band_count":3,"active_band_mae_db":0.17312016977714184}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":0.00010383940879839973,"log_rms_mae_db":1.4917489525694394,"rms_envelope_nrmse":0.1520787198243727,"envelope_correlation":0.9812894969400351,"rms_gain_error_db":0.040038099969133205,"reference_mean_power":0.0005521302842545546,"output_mean_power":0.0005572439714661772},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":2.374888225952349e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001363070736426414,"output_power":0.00013626526367801036,"reference_active":true,"delta_db":-0.0013323316113973429},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.2931511209828857e-05,"output_power":3.4518787615065165e-05,"reference_active":true,"delta_db":0.20443873102108145},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.586128694071041e-05,"output_power":1.579080007104158e-05,"reference_active":true,"delta_db":-0.019342869325728172}],"active_band_count":3,"active_band_mae_db":0.07503797731940233}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":8.776193320418381e-05,"log_rms_mae_db":0.8123048473736495,"rms_envelope_nrmse":0.06364368779982381,"envelope_correlation":0.9963307075640717,"rms_gain_error_db":0.0400017759988415,"reference_mean_power":0.0005592400075632056,"output_mean_power":0.000564414822453811},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.9436382041482275e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001261572491396227,"output_power":0.0001262866625639983,"reference_active":true,"delta_db":0.004452754723590717},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5195698098116377e-05,"output_power":1.610316601642805e-05,"reference_active":true,"delta_db":0.25190613751361346},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.080674445150943e-07,"output_power":9.854385201009412e-07,"reference_active":true,"delta_db":0.3551142864535489}],"active_band_count":3,"active_band_mae_db":0.20382439289691767}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":470,"rms_floor":6.310171076800205e-05,"log_rms_mae_db":1.784651043246924,"rms_envelope_nrmse":0.26409288475984183,"envelope_correlation":0.9695790961010103,"rms_gain_error_db":1.2157199092658975,"reference_mean_power":0.00026303793349419816,"output_mean_power":0.0003480089193034952},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":470,"frequency_resolution_hz":0.425531914893617,"power_floor":1.3356760090821463e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.8507771208283054e-05,"output_power":8.152548994272161e-05,"reference_active":true,"delta_db":1.4407986302350704},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1981263949052123e-05,"output_power":1.8733505479894634e-05,"reference_active":true,"delta_db":1.9411641585634705},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.2561065186681e-06,"output_power":1.4001105791793017e-05,"reference_active":true,"delta_db":1.797339933239496}],"active_band_count":3,"active_band_mae_db":1.7264342406793454}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":37680,"uncovered_tail_samples":0,"metrics":{"frame_count":464,"rms_floor":5.861628173421478e-05,"log_rms_mae_db":1.167161901968294,"rms_envelope_nrmse":0.18375524936776597,"envelope_correlation":0.9921526259259749,"rms_gain_error_db":1.2157886590038687,"reference_mean_power":0.00026641634082957393,"output_mean_power":0.0003524842570770773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":464,"frequency_resolution_hz":0.4310344827586207,"power_floor":1.157398705126345e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":5.177699204483556e-05,"output_power":7.383054792713528e-05,"reference_active":true,"delta_db":1.5409927477541512},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.036512028004997e-06,"output_power":1.0903741987531482e-05,"reference_active":true,"delta_db":1.9021813197494617},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.58854722653168e-07,"output_power":9.027165532003878e-07,"reference_active":true,"delta_db":2.0825248090434636}],"active_band_count":3,"active_band_mae_db":1.8418996255156923}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"hnm","status":"measured","mos":{"status":"measured","value":3.0624163150787354,"predicted_mos":3.0624163150787354,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"dd4bd490550a7f6b3db510a20e3e9b96643386f42831a15e1e98bfa02f46658a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.035170451999875},"started_at":"2026-10-13T04:40:31.848208+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.7901999854075257e-05,"native_input_samples":96000,"native_input_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","output_sha256":"8f5dac053feb7169e27fea1cdf637a8782621f3ac753f87e4926bc61ef44f519","output_samples":96000,"length_delta_samples":0,"analysis_s":0.994699386999855,"synthesis_s":0.48619363699981477,"total_s":1.4808930239996698,"rtf":0.24681550399994498,"output_peak":0.25683363910366236,"output_rms":0.046117504942521134,"finished_at":"2026-10-13T04:40:33.334193+00:00","evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"8f5dac053feb7169e27fea1cdf637a8782621f3ac753f87e4926bc61ef44f519","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.296519068503457,"spectral_convergence":0.13896561255739787,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.352074537839787,"spectral_convergence":0.13429910566837777,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.595903856097296,"spectral_convergence":0.13687483914625126,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.00017368346498190023,"delta_db":0.1900629596428187,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.001333639247474865,"delta_db":-0.09783376051728607,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0005681845856653844,"delta_db":-0.21916793894904388,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":1.8305176048163967e-05,"delta_db":0.03262319113236861,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":2.294660300489875e-05,"delta_db":-0.30840731594194376,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":1.0065184948248949e-05,"delta_db":-0.11846757489119641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16109379017910957,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10926176821494371,"pitch":null,"output_rms_dbfs":-26.722683940385824,"output_peak":0.25683363910366236,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.41483248748018},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.296519068503457,"spectral_convergence":0.13896561255739787,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.352074537839787,"spectral_convergence":0.13429910566837777,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.595903856097296,"spectral_convergence":0.13687483914625126,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.00017368346498190023,"delta_db":0.1900629596428187,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.001333639247474865,"delta_db":-0.09783376051728607,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0005681845856653844,"delta_db":-0.21916793894904388,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":1.8305176048163967e-05,"delta_db":0.03262319113236861,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":2.294660300489875e-05,"delta_db":-0.30840731594194376,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":1.0065184948248949e-05,"delta_db":-0.11846757489119641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.16109379017910957,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.10926176821494371,"pitch":null,"output_rms_dbfs":-26.722683940385824,"output_peak":0.25683363910366236,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.41483248748018},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9434937798487315,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04670129155060261,"rms_dbfs":-26.613422172170882,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.26355967311556,"mean":-1.4899571736653646e-06,"zero_energy":false},"output":{"rms":0.046117504942521134,"rms_dbfs":-26.722683940385824,"peak":0.25683363910366236,"peak_dbfs":-11.806961892412527,"crest_factor_db":14.915722047973297,"mean":4.676720822412616e-06,"zero_energy":false},"rms_difference_db":-0.10926176821494238},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":6.335461223724808e-07,"output_rms_dbfs":-61.9822176284412,"output_zero_energy":false,"output_peak":0.015122119448650813,"output_energy_fraction":5.361905262129583e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":397,"output_estimated_voiced_frames":394,"both_estimated_voiced_frames":387,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9748110831234257,"absolute_difference_cents_median":1.307108486436528,"absolute_difference_cents_p95":9.554893462464976,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001446111949915804,"log_rms_mae_db":0.6477581392168703,"rms_envelope_nrmse":0.05120262169267331,"envelope_correlation":0.9971657381874004,"rms_gain_error_db":-0.10926361509647331,"reference_mean_power":0.002182828397481346,"output_mean_power":0.0021285959602311017},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.790920693994897e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003230276115117931,"output_power":0.00030444798950819197,"reference_active":true,"delta_db":-0.257265357783558},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.477236494661316e-05,"output_power":4.188539439102781e-05,"reference_active":true,"delta_db":-0.28947426010374955},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.548345752003966e-06,"output_power":7.849465316205091e-06,"reference_active":true,"delta_db":-0.8508806189610106}],"active_band_count":3,"active_band_mae_db":0.46587341228277274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014111493703772434,"log_rms_mae_db":0.3685794232381211,"rms_envelope_nrmse":0.030498244270001078,"envelope_correlation":0.9990429854267997,"rms_gain_error_db":-0.1092689230058072,"reference_mean_power":0.002193693032291648,"output_mean_power":0.0021391880483638587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.190420923463323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028912450266098784,"output_power":0.00027294881120595263,"reference_active":true,"delta_db":-0.25003691594037136},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2978791346744758e-05,"output_power":2.1654739469566305e-05,"reference_active":true,"delta_db":-0.25774218597367055},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.072854493661426e-07,"output_power":5.082828535752343e-07,"reference_active":true,"delta_db":-0.7728741604779417}],"active_band_count":3,"active_band_mae_db":0.42688442079732786}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012712872255296568,"log_rms_mae_db":1.0476679477111013,"rms_envelope_nrmse":0.06326034114132933,"envelope_correlation":0.9964867084615352,"rms_gain_error_db":-0.16568798833365422,"reference_mean_power":0.0014813690704446063,"output_mean_power":0.0014259177458243641},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.409297669270013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003163260619212235,"output_power":0.0002973003113495906,"reference_active":true,"delta_db":-0.269396106381359},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6701975061022124e-05,"output_power":3.244730155983793e-05,"reference_active":true,"delta_db":-0.5351085071006328},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.3015260364548945e-06,"output_power":4.800249585157695e-06,"reference_active":true,"delta_db":-1.1818191624353171}],"active_band_count":3,"active_band_mae_db":0.6621079253057697}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012343588785799389,"log_rms_mae_db":0.5846218341946076,"rms_envelope_nrmse":0.03981228892954564,"envelope_correlation":0.9987303654591165,"rms_gain_error_db":-0.1656989234780792,"reference_mean_power":0.0014887936512772218,"output_mean_power":0.0014330607978280513},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.99328908394483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002875949770086079,"output_power":0.000271385418269453,"reference_active":true,"delta_db":-0.2519478760402261},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9768070879384797e-05,"output_power":1.776299471949646e-05,"reference_active":true,"delta_db":-0.4644810288435385},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.520482053574086e-07,"output_power":3.2261389852700215e-07,"reference_active":true,"delta_db":-1.465016761295916}],"active_band_count":3,"active_band_mae_db":0.7271485553932268}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":1.4011142548503643,"rms_envelope_nrmse":0.1315976531514819,"envelope_correlation":0.986553284508621,"rms_gain_error_db":-0.21196286316037982,"reference_mean_power":3.669443825843959e-05,"output_mean_power":3.49465218060251e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6589829493109392e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.317934906149565e-05,"output_power":1.161951355952694e-05,"reference_active":true,"delta_db":-0.5470601355131443},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9890642401085547e-06,"output_power":2.950882937086121e-06,"reference_active":true,"delta_db":-0.055832676962747387},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.136200949313618e-07,"output_power":5.689548580419193e-07,"reference_active":true,"delta_db":-1.5534385631453547}],"active_band_count":3,"active_band_mae_db":0.7187771252070821}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":0.8188975672153813,"rms_envelope_nrmse":0.08000226962854114,"envelope_correlation":0.995016728655165,"rms_gain_error_db":-0.21211566550855582,"reference_mean_power":3.687647047180045e-05,"output_mean_power":3.511864739190188e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4709890047082166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2743108138238891e-05,"output_power":1.1162747091007613e-05,"reference_active":true,"delta_db":-0.5750428330156511},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4131557808728359e-06,"output_power":1.5051203510189744e-06,"reference_active":true,"delta_db":0.27381188520143407},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.713126831346598e-08,"output_power":4.119432670275484e-08,"reference_active":true,"delta_db":-1.42036456538123}],"active_band_count":3,"active_band_mae_db":0.756406427866105}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":1.5093369237714451,"rms_envelope_nrmse":0.17248960342687114,"envelope_correlation":0.97871591234077,"rms_gain_error_db":-0.030391843553276256,"reference_mean_power":1.1971789360385637e-05,"output_mean_power":1.1888303467828547e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0435686252011218e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2483829658988356e-06,"output_power":4.150491792886264e-06,"reference_active":true,"delta_db":-0.10124099414781454},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.589400168019186e-06,"output_power":2.534493937149447e-06,"reference_active":true,"delta_db":-0.0930791517601684},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.486231293405909e-07,"output_power":6.592698423257633e-07,"reference_active":true,"delta_db":0.07070778629307177}],"active_band_count":3,"active_band_mae_db":0.08834264406701824}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":0.9358909154116244,"rms_envelope_nrmse":0.10778103231282465,"envelope_correlation":0.9907942838251249,"rms_gain_error_db":-0.03038723505026354,"reference_mean_power":1.203192810180695e-05,"output_mean_power":1.1948035507240531e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.457061965350649e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5073827107587115e-06,"output_power":4.439504060249095e-06,"reference_active":true,"delta_db":-0.0658997699236248},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3712782786962116e-06,"output_power":1.3854223597690574e-06,"reference_active":true,"delta_db":0.044565958153471134},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.274779912953661e-08,"output_power":4.921977759645339e-08,"reference_active":true,"delta_db":0.6122588698836122}],"active_band_count":3,"active_band_mae_db":0.24090819932023608}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.57151198387146,"predicted_mos":3.57151198387146,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"77ee513812fe909bd54663848aa37b2a18c4817fd3d00e1beaa831ecceba5c95","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6085141229996225},"started_at":"2026-10-13T04:43:48.256061+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.9934000192733947e-05,"native_input_samples":96000,"native_input_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","output_sha256":"c9da0ed008b38e33d0460128fbaa54661551feb9b285d18050a2a706f6b07092","output_samples":96000,"length_delta_samples":0,"analysis_s":1.483857426999748,"synthesis_s":0.1032665390002876,"total_s":1.5871239660000356,"rtf":0.26452066100000593,"output_peak":0.3212816262344611,"output_rms":0.05458470772260165,"finished_at":"2026-10-13T04:43:49.847984+00:00","evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"c9da0ed008b38e33d0460128fbaa54661551feb9b285d18050a2a706f6b07092","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.929902809484164,"spectral_convergence":0.4641074963902323,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.33261652988972,"spectral_convergence":0.24922524971595666,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.738843518934298,"spectral_convergence":0.2797985034849887,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.0001580555318292859,"delta_db":-0.2194247830103245,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.001865393680677293,"delta_db":1.3594875733785239,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0008897353705737766,"delta_db":1.7285461444620467,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":2.3908278232699374e-05,"delta_db":1.1923671038781984,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":3.1300619610454146e-05,"delta_db":1.039938010451366,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":1.1096836238338734e-05,"delta_db":0.3053066000886314,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.974178369211515,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.3548419559678562,"pitch":null,"output_rms_dbfs":-25.258580216203022,"output_peak":0.3212816262344611,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.667120952769395},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.929902809484164,"spectral_convergence":0.4641074963902323,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.33261652988972,"spectral_convergence":0.24922524971595666,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.738843518934298,"spectral_convergence":0.2797985034849887,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.0001580555318292859,"delta_db":-0.2194247830103245,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.001865393680677293,"delta_db":1.3594875733785239,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0008897353705737766,"delta_db":1.7285461444620467,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":2.3908278232699374e-05,"delta_db":1.1923671038781984,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":3.1300619610454146e-05,"delta_db":1.039938010451366,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":1.1096836238338734e-05,"delta_db":0.3053066000886314,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.974178369211515,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.3548419559678562,"pitch":null,"output_rms_dbfs":-25.258580216203022,"output_peak":0.3212816262344611,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.667120952769395},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9433942706671327,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04670129155060261,"rms_dbfs":-26.613422172170882,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.26355967311556,"mean":-1.4899571736653646e-06,"zero_energy":false},"output":{"rms":0.054584707722601655,"rms_dbfs":-25.258580216203022,"peak":0.3212816262344611,"peak_dbfs":-9.862282211855057,"crest_factor_db":15.396298004347969,"mean":1.0039352926290504e-06,"zero_energy":false},"rms_difference_db":1.35484195596786},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":5.920153098893106e-07,"output_rms_dbfs":-62.276670620026344,"output_zero_energy":false,"output_peak":0.0081703443744978,"output_energy_fraction":3.576543114313042e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":397,"output_estimated_voiced_frames":411,"both_estimated_voiced_frames":382,"reference_voiced_output_unvoiced_frames":15,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.9622166246851386,"absolute_difference_cents_median":5.7232423661465095,"absolute_difference_cents_p95":19.34558229763095,"difference_over_600_cents_frames":3,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001446111949915804,"log_rms_mae_db":1.5649914197549037,"rms_envelope_nrmse":0.19638867173788546,"envelope_correlation":0.9904047708760945,"rms_gain_error_db":1.3548434933617206,"reference_mean_power":0.002182828397481346,"output_mean_power":0.002981974631268259},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.790920693994897e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003230276115117931,"output_power":0.00044598546594810105,"reference_active":true,"delta_db":1.400810596683149},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.477236494661316e-05,"output_power":7.092060164985215e-05,"reference_active":true,"delta_db":1.997623761449375},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.548345752003966e-06,"output_power":9.651256926872152e-06,"reference_active":true,"delta_db":0.04655740449262039}],"active_band_count":3,"active_band_mae_db":1.1483305875417147}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014111493703772434,"log_rms_mae_db":1.2509287066657473,"rms_envelope_nrmse":0.18642707199474656,"envelope_correlation":0.993693836982374,"rms_gain_error_db":1.3548456186993496,"reference_mean_power":0.002193693032291648,"output_mean_power":0.0029968183389544466},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.190420923463323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028912450266098784,"output_power":0.0004034368265998585,"reference_active":true,"delta_db":1.4469064031523624},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2978791346744758e-05,"output_power":3.858051816341158e-05,"reference_active":true,"delta_db":2.250408745740954},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.072854493661426e-07,"output_power":5.18903054182817e-07,"reference_active":true,"delta_db":-0.6830664860581934}],"active_band_count":3,"active_band_mae_db":1.4601272116505033}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012712872255296568,"log_rms_mae_db":1.775545304496052,"rms_envelope_nrmse":0.21651530221651527,"envelope_correlation":0.9906709102700176,"rms_gain_error_db":1.4942071024822563,"reference_mean_power":0.0014813690704446063,"output_mean_power":0.0020897001938805127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.409297669270013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003163260619212235,"output_power":0.00044048559507722443,"reference_active":true,"delta_db":1.4379673586068977},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6701975061022124e-05,"output_power":6.0342232914937506e-05,"reference_active":true,"delta_db":2.1593194106952787},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.3015260364548945e-06,"output_power":6.90137427455553e-06,"reference_active":true,"delta_db":0.39489845652101113}],"active_band_count":3,"active_band_mae_db":1.3307284086077291}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012343588785799389,"log_rms_mae_db":1.4874337204683894,"rms_envelope_nrmse":0.20742650602076912,"envelope_correlation":0.9932923046334166,"rms_gain_error_db":1.4941743970792398,"reference_mean_power":0.0014887936512772218,"output_mean_power":0.00210015789777506},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.99328908394483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002875949770086079,"output_power":0.00040252666910639833,"reference_active":true,"delta_db":1.4601336291162157},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9768070879384797e-05,"output_power":3.3991331332701435e-05,"reference_active":true,"delta_db":2.354038852734794},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.520482053574086e-07,"output_power":4.1696727247373887e-07,"reference_active":true,"delta_db":-0.3508278064068518}],"active_band_count":3,"active_band_mae_db":1.3883334294192873}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":1.562590203385112,"rms_envelope_nrmse":0.21581750122648147,"envelope_correlation":0.9881939209088706,"rms_gain_error_db":1.3592577394330534,"reference_mean_power":3.669443825843959e-05,"output_mean_power":5.0179463949089374e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6589829493109392e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.317934906149565e-05,"output_power":1.9089691379993752e-05,"reference_active":true,"delta_db":1.6090494664182442},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9890642401085547e-06,"output_power":4.481479343934703e-06,"reference_active":true,"delta_db":1.758861501389444},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.136200949313618e-07,"output_power":7.755802466069443e-07,"reference_active":true,"delta_db":-0.20794926905090913}],"active_band_count":3,"active_band_mae_db":1.1919534122861992}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":1.2167126864609212,"rms_envelope_nrmse":0.18783565582971642,"envelope_correlation":0.9954937397246465,"rms_gain_error_db":1.359314149938856,"reference_mean_power":3.687647047180045e-05,"output_mean_power":5.042904712467923e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4709890047082166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2743108138238891e-05,"output_power":1.8556797271103252e-05,"reference_active":true,"delta_db":1.632276546493085},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4131557808728359e-06,"output_power":2.128151901623322e-06,"reference_active":true,"delta_db":1.7781258402835776},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.713126831346598e-08,"output_power":5.873456587012436e-08,"reference_active":true,"delta_db":0.12019897562177795}],"active_band_count":3,"active_band_mae_db":1.17686712079948}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":1.6905223314477733,"rms_envelope_nrmse":0.23167233709969007,"envelope_correlation":0.9656670580108144,"rms_gain_error_db":0.46296989854506526,"reference_mean_power":1.1971789360385637e-05,"output_mean_power":1.3318522748178686e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0435686252011218e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2483829658988356e-06,"output_power":4.809509334735269e-06,"reference_active":true,"delta_db":0.5387711314441108},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.589400168019186e-06,"output_power":2.584635756793476e-06,"reference_active":true,"delta_db":-0.007998236269502083},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.486231293405909e-07,"output_power":6.101087182582425e-07,"reference_active":true,"delta_db":-0.26585200213388893}],"active_band_count":3,"active_band_mae_db":0.2708737899491673}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":1.275409887418891,"rms_envelope_nrmse":0.15081407255455972,"envelope_correlation":0.9852105739151245,"rms_gain_error_db":0.46297565070500196,"reference_mean_power":1.203192810180695e-05,"output_mean_power":1.3385444360054208e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.457061965350649e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5073827107587115e-06,"output_power":4.992463551726762e-06,"reference_active":true,"delta_db":0.44390468407661615},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3712782786962116e-06,"output_power":1.3078264160031485e-06,"reference_active":true,"delta_db":-0.20575491588165973},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.274779912953661e-08,"output_power":4.2358844218533626e-08,"reference_active":true,"delta_db":-0.03969658325112081}],"active_band_count":3,"active_band_mae_db":0.22978539440313225}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"facodec","status":"measured","mos":{"status":"measured","value":4.210933208465576,"predicted_mos":4.210933208465576,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"cce5190dc0c7b090f37d365b2c1f80160b2eda6ea3339243565672e25af13910","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9259560649989},"started_at":"2026-10-13T07:02:21.085820+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":5.426099960459396e-05,"native_input_samples":96000,"native_input_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","output_sha256":"9dfa30002cf8da93684908922df3ecac762a225d2c772e73ebc9a278c0ff6003","output_samples":96000,"length_delta_samples":0,"analysis_s":2.3458827509984985,"synthesis_s":6.847583149001366,"total_s":9.193465899999865,"rtf":1.5322443166666442,"output_peak":0.20844441652297974,"output_rms":0.041170113469273004,"finished_at":"2026-10-13T07:02:30.289244+00:00","evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"9dfa30002cf8da93684908922df3ecac762a225d2c772e73ebc9a278c0ff6003","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.485728612480038,"spectral_convergence":0.22167428961422866,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.037976194111315,"spectral_convergence":0.22197074692785349,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.91056279202945,"spectral_convergence":0.23269606863542427,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.0001422186052603031,"delta_db":-0.6779576354180955,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.0009888601430286775,"delta_db":-1.3968687126662482,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0005242619965108872,"delta_db":-0.5685786435900053,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":1.1534577617089198e-05,"delta_db":-1.9730989512135504,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":1.8429199663079855e-05,"delta_db":-1.2605265841170556,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":9.673720992777144e-06,"delta_db":-0.2907495955664149,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0279633537618948,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.094936548597775,"pitch":null,"output_rms_dbfs":-27.708358720768658,"output_peak":0.20844441652297974,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.811422532873601},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.485728612480038,"spectral_convergence":0.22167428961422866,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.037976194111315,"spectral_convergence":0.22197074692785349,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.91056279202945,"spectral_convergence":0.23269606863542427,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.0001422186052603031,"delta_db":-0.6779576354180955,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.0009888601430286775,"delta_db":-1.3968687126662482,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0005242619965108872,"delta_db":-0.5685786435900053,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":1.1534577617089198e-05,"delta_db":-1.9730989512135504,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":1.8429199663079855e-05,"delta_db":-1.2605265841170556,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":9.673720992777144e-06,"delta_db":-0.2907495955664149,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0279633537618948,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.094936548597775,"pitch":null,"output_rms_dbfs":-27.708358720768658,"output_peak":0.20844441652297974,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.811422532873601},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9283638072341375,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04670129155060261,"rms_dbfs":-26.613422172170882,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.26355967311556,"mean":-1.4899571736653646e-06,"zero_energy":false},"output":{"rms":0.041170113469273004,"rms_dbfs":-27.708358720768658,"peak":0.20844441652297974,"peak_dbfs":-13.620194671414387,"crest_factor_db":14.088164049354269,"mean":1.4856248336900062e-05,"zero_energy":false},"rms_difference_db":-1.0949365485977758},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":2.564339330647454e-07,"output_rms_dbfs":-65.91024506580348,"output_zero_energy":false,"output_peak":0.017476387321949005,"output_energy_fraction":2.7232271647318883e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":397,"output_estimated_voiced_frames":384,"both_estimated_voiced_frames":381,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.9596977329974811,"absolute_difference_cents_median":3.1162445050565335,"absolute_difference_cents_p95":13.69491324807989,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001446111949915804,"log_rms_mae_db":1.2707983638445923,"rms_envelope_nrmse":0.15274009716779918,"envelope_correlation":0.9883501186847725,"rms_gain_error_db":-1.0949352542530937,"reference_mean_power":0.002182828397481346,"output_mean_power":0.0016963914295735113},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.790920693994897e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003230276115117931,"output_power":0.00022577546535640386,"reference_active":true,"delta_db":-1.5556290012024323},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.477236494661316e-05,"output_power":4.063370963789422e-05,"reference_active":true,"delta_db":-0.4212356223235588},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.548345752003966e-06,"output_power":8.590828762998303e-06,"reference_active":true,"delta_db":-0.4589307418243206}],"active_band_count":3,"active_band_mae_db":0.8119317884501039}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014111493703772434,"log_rms_mae_db":1.0766277477781223,"rms_envelope_nrmse":0.1486159143469601,"envelope_correlation":0.989315489947532,"rms_gain_error_db":-1.0948878349605413,"reference_mean_power":0.002193693032291648,"output_mean_power":0.0017048535267499912},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.190420923463323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028912450266098784,"output_power":0.00019982731196295902,"reference_active":true,"delta_db":-1.6043005235212426},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2978791346744758e-05,"output_power":2.0466240743795725e-05,"reference_active":true,"delta_db":-0.502891032378895},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.072854493661426e-07,"output_power":5.2901289433308e-07,"reference_active":true,"delta_db":-0.5992661769841138}],"active_band_count":3,"active_band_mae_db":0.9021525776280838}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012712872255296568,"log_rms_mae_db":1.284407707505783,"rms_envelope_nrmse":0.1585603508615999,"envelope_correlation":0.9875003316119254,"rms_gain_error_db":-1.067110381109934,"reference_mean_power":0.0014813690704446063,"output_mean_power":0.0011586499164460804},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.409297669270013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003163260619212235,"output_power":0.00023191560040604754,"reference_active":true,"delta_db":-1.3480501109241168},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6701975061022124e-05,"output_power":3.2763004106927266e-05,"reference_active":true,"delta_db":-0.49305719514227486},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.3015260364548945e-06,"output_power":5.693434341012648e-06,"reference_active":true,"delta_db":-0.4407141844698537}],"active_band_count":3,"active_band_mae_db":0.7606071635120818}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012343588785799389,"log_rms_mae_db":1.0703916645570948,"rms_envelope_nrmse":0.1538367438362475,"envelope_correlation":0.9883505249865042,"rms_gain_error_db":-1.0671338494559504,"reference_mean_power":0.0014887936512772218,"output_mean_power":0.0011644507453394887},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.99328908394483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002875949770086079,"output_power":0.00021006234194179845,"reference_active":true,"delta_db":-1.3643309356939062},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9768070879384797e-05,"output_power":1.770414680927084e-05,"reference_active":true,"delta_db":-0.4788928731784565},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.520482053574086e-07,"output_power":4.2636765695036985e-07,"reference_active":true,"delta_db":-0.2540049651122277}],"active_band_count":3,"active_band_mae_db":0.6990762579948635}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":1.9802998495910178,"rms_envelope_nrmse":0.23747984923487414,"envelope_correlation":0.9780920788017488,"rms_gain_error_db":-1.7636218520550686,"reference_mean_power":3.669443825843959e-05,"output_mean_power":2.4447702857843956e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6589829493109392e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.317934906149565e-05,"output_power":8.518124446095853e-06,"reference_active":true,"delta_db":-1.8954998000104086},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9890642401085547e-06,"output_power":2.29373334628968e-06,"reference_active":true,"delta_db":-1.1499232047033272},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.136200949313618e-07,"output_power":6.724113406938512e-07,"reference_active":true,"delta_db":-0.827866368147405}],"active_band_count":3,"active_band_mae_db":1.2910964576203803}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":1.7036032374278651,"rms_envelope_nrmse":0.21924724033395462,"envelope_correlation":0.9850406736528304,"rms_gain_error_db":-1.763617536520231,"reference_mean_power":3.687647047180045e-05,"output_mean_power":2.4569006399456877e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4709890047082166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2743108138238891e-05,"output_power":8.445709566225312e-06,"reference_active":true,"delta_db":-1.7863922586554961},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4131557808728359e-06,"output_power":1.0579436102236477e-06,"reference_active":true,"delta_db":-1.2572751959249944},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.713126831346598e-08,"output_power":5.189848835973843e-08,"reference_active":true,"delta_db":-0.4171915710386457}],"active_band_count":3,"active_band_mae_db":1.1536196752063788}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":1.9771297042718567,"rms_envelope_nrmse":0.20887032305905992,"envelope_correlation":0.9704492278895909,"rms_gain_error_db":-0.19718019407029203,"reference_mean_power":1.1971789360385637e-05,"output_mean_power":1.1440395739088244e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0435686252011218e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2483829658988356e-06,"output_power":4.114542004054229e-06,"reference_active":true,"delta_db":-0.13902158588094127},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.589400168019186e-06,"output_power":2.5842363779135626e-06,"reference_active":true,"delta_db":-0.008669361573281636},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.486231293405909e-07,"output_power":9.723335794476424e-07,"reference_active":true,"delta_db":1.7582285315025432}],"active_band_count":3,"active_band_mae_db":0.6353064929855887}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":1.6271894526006447,"rms_envelope_nrmse":0.14334672126946782,"envelope_correlation":0.9851799884601571,"rms_gain_error_db":-0.19719407375950265,"reference_mean_power":1.203192810180695e-05,"output_mean_power":1.149782834695848e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.457061965350649e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5073827107587115e-06,"output_power":4.6087728801685855e-06,"reference_active":true,"delta_db":0.09660872155731628},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3712782786962116e-06,"output_power":1.4349797764841208e-06,"reference_active":true,"delta_db":0.19720183738470953},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.274779912953661e-08,"output_power":7.869601618846605e-08,"reference_active":true,"delta_db":2.650389877310855}],"active_band_count":3,"active_band_mae_db":0.981400145417627}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.848743438720703,"predicted_mos":3.848743438720703,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"64447ca341e203b0846748048621a3a9e1f5aaace78b4d2d01e3c1e2c617c6bd","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6772754340017855},"started_at":"2026-10-13T07:33:31.239181+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006674878000922035,"native_input_samples":264600,"native_input_sha256":"5a4ba89605692d3966d8736f204f460d57ced405fd7829deee38367ca43ef7c1","output_sha256":"d11e61617ce9131087e8c8c4d7da776205a9dc9c43c39af244f4a0909610c538","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1367598539982282,"synthesis_s":5.802414671001316,"total_s":5.939174524999544,"rtf":0.9898624208332573,"output_peak":0.23894472420215607,"output_rms":0.0436943319480993,"finished_at":"2026-10-13T07:33:37.210162+00:00","evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"d11e61617ce9131087e8c8c4d7da776205a9dc9c43c39af244f4a0909610c538","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.18532077250929,"output_peak":0.23931627374426948},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.371376782661304,"spectral_convergence":0.39874708241327755,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.92798543147232,"spectral_convergence":0.2642645686218038,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.088283479714828,"spectral_convergence":0.31188454786576575,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.00011769851442488372,"delta_db":-1.4998119746496275,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.0011647937356216837,"delta_db":-0.6857271755302757,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0005874173032104435,"delta_db":-0.0745950628900428,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":1.7186517820392194e-05,"delta_db":-0.24123698198239493,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":1.827635285791101e-05,"delta_db":-1.2966959934765385,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":7.2928492860959495e-06,"delta_db":-1.5177127843565705,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8859633288142416,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5701880999158724,"pitch":null,"output_rms_dbfs":-27.183610272086753,"output_peak":0.23931627374426948,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.795881897949483},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7991002254784157,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04670129155060261,"rms_dbfs":-26.613422172170882,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.26355967311556,"mean":-1.4899571736653646e-06,"zero_energy":false},"output":{"rms":0.04373402877876002,"rms_dbfs":-27.183610272086753,"peak":0.23931627374426948,"peak_dbfs":-12.420555358527,"crest_factor_db":14.763054913559753,"mean":-6.133268987831439e-06,"zero_energy":false},"rms_difference_db":-0.570188099915871},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":4.634954875721705e-07,"output_rms_dbfs":-63.33954489636088,"output_zero_energy":false,"output_peak":0.010136688945634267,"output_energy_fraction":4.361933524441259e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":397,"output_estimated_voiced_frames":378,"both_estimated_voiced_frames":372,"reference_voiced_output_unvoiced_frames":25,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9370277078085643,"absolute_difference_cents_median":7.339138116133164,"absolute_difference_cents_p95":38.338934918764465,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001446111949915804,"log_rms_mae_db":1.4785935130860437,"rms_envelope_nrmse":0.10129333785698617,"envelope_correlation":0.9927835495045791,"rms_gain_error_db":-0.5701951181998781,"reference_mean_power":0.002182828397481346,"output_mean_power":0.0019142562920843459},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.790920693994897e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003230276115117931,"output_power":0.0002990956570529154,"reference_active":true,"delta_db":-0.3342953920784461},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.477236494661316e-05,"output_power":4.267779307675436e-05,"reference_active":true,"delta_db":-0.20808081695598557},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.548345752003966e-06,"output_power":7.745484929174903e-06,"reference_active":true,"delta_db":-0.908795234667431}],"active_band_count":3,"active_band_mae_db":0.4837238145672876}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014111493703772434,"log_rms_mae_db":1.202852959360901,"rms_envelope_nrmse":0.08535235980675002,"envelope_correlation":0.9960703435245695,"rms_gain_error_db":-0.5701684944531453,"reference_mean_power":0.002193693032291648,"output_mean_power":0.0019237959511556797},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.190420923463323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028912450266098784,"output_power":0.00027091428426984044,"reference_active":true,"delta_db":-0.2825299440829805},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2978791346744758e-05,"output_power":2.1196672514237826e-05,"reference_active":true,"delta_db":-0.3505949157864894},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.072854493661426e-07,"output_power":5.236174875345802e-07,"reference_active":true,"delta_db":-0.643787330110704}],"active_band_count":3,"active_band_mae_db":0.42563739666005795}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012712872255296568,"log_rms_mae_db":2.047333995522451,"rms_envelope_nrmse":0.1294530963885913,"envelope_correlation":0.9848271266838468,"rms_gain_error_db":-0.3010930438496546,"reference_mean_power":0.0014813690704446063,"output_mean_power":0.0013821461500787833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.409297669270013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003163260619212235,"output_power":0.0002881343680717993,"reference_active":true,"delta_db":-0.40539911542795376},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6701975061022124e-05,"output_power":3.482035072929575e-05,"reference_active":true,"delta_db":-0.22856294527724547},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.3015260364548945e-06,"output_power":4.9115508776330285e-06,"reference_active":true,"delta_db":-1.0822708783882642}],"active_band_count":3,"active_band_mae_db":0.5720776463644878}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012343588785799389,"log_rms_mae_db":1.582004636931442,"rms_envelope_nrmse":0.11291541832607653,"envelope_correlation":0.9879545586153954,"rms_gain_error_db":-0.301149416945273,"reference_mean_power":0.0014887936512772218,"output_mean_power":0.0013890553977864571},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.99328908394483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002875949770086079,"output_power":0.00026575180213115257,"reference_active":true,"delta_db":-0.34305078337188843},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9768070879384797e-05,"output_power":1.7665534486201725e-05,"reference_active":true,"delta_db":-0.48837507651746936},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.520482053574086e-07,"output_power":4.0202642462395804e-07,"reference_active":true,"delta_db":-0.5093014984294456}],"active_band_count":3,"active_band_mae_db":0.4469091194396011}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":3.7789987820964894,"rms_envelope_nrmse":0.27899435193205385,"envelope_correlation":0.9410164491125576,"rms_gain_error_db":-0.7690799848223311,"reference_mean_power":3.669443825843959e-05,"output_mean_power":3.073917767102799e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6589829493109392e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.317934906149565e-05,"output_power":1.3380832207239592e-05,"reference_active":true,"delta_db":0.06589164136305824},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9890642401085547e-06,"output_power":2.5710372526236865e-06,"reference_active":true,"delta_db":-0.6542687954431904},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.136200949313618e-07,"output_power":5.89365920015463e-07,"reference_active":true,"delta_db":-1.4003664709312627}],"active_band_count":3,"active_band_mae_db":0.7068423025791705}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":3.2482025027502783,"rms_envelope_nrmse":0.2348122008966844,"envelope_correlation":0.9567920715013191,"rms_gain_error_db":-0.7688435202597022,"reference_mean_power":3.687647047180045e-05,"output_mean_power":3.0893349314789554e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4709890047082166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2743108138238891e-05,"output_power":1.2843975078511404e-05,"reference_active":true,"delta_db":0.034240856697397036},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4131557808728359e-06,"output_power":1.1909736962800272e-06,"reference_active":true,"delta_db":-0.7428786963900587},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.713126831346598e-08,"output_power":4.1373870576160965e-08,"reference_active":true,"delta_db":-1.401477139241596}],"active_band_count":3,"active_band_mae_db":0.7261988974430172}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":5.387859349383799,"rms_envelope_nrmse":0.29645809321576205,"envelope_correlation":0.9564250846301737,"rms_gain_error_db":-1.7899227866374825,"reference_mean_power":1.1971789360385637e-05,"output_mean_power":7.928057445236432e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0435686252011218e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2483829658988356e-06,"output_power":2.8201822791322573e-06,"reference_active":true,"delta_db":-1.7794647951095486},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.589400168019186e-06,"output_power":1.7152992671723374e-06,"reference_active":true,"delta_db":-1.7885926977433597},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.486231293405909e-07,"output_power":3.568185727772912e-07,"reference_active":true,"delta_db":-2.5954497940796357}],"active_band_count":3,"active_band_mae_db":2.0545024289775147}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":5.130332798762785,"rms_envelope_nrmse":0.2568082490824896,"envelope_correlation":0.9727358598432447,"rms_gain_error_db":-1.7899172928657394,"reference_mean_power":1.203192810180695e-05,"output_mean_power":7.967893099651755e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.457061965350649e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5073827107587115e-06,"output_power":2.978706212896313e-06,"reference_active":true,"delta_db":-1.798967632560854},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3712782786962116e-06,"output_power":8.961851723910395e-07,"reference_active":true,"delta_db":-1.847278426327624},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.274779912953661e-08,"output_power":2.8831702292639693e-08,"reference_active":true,"delta_db":-1.7104347505828876}],"active_band_count":3,"active_band_mae_db":1.7855602698237885}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"golf","status":"measured","mos":{"status":"measured","value":3.942218542098999,"predicted_mos":3.942218542098999,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"bb612ea5eb0426c54d82fa01fa21e0ca851d9de095a5d31edce237394b1a2335","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.243872670000201},"started_at":"2026-10-13T05:03:11.000973+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00333268299982592,"native_input_samples":144000,"native_input_sha256":"f862f06c343ae56f5d6da883d8058b41cadd89e855a153bc4acbb64b919b4eda","output_sha256":"806f235172212f4c8a9f9b9d6d1d7a05c6290eabeaacb48ba939a0103d6f282d","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6610784750000676,"synthesis_s":0.12295231299958687,"total_s":0.7840307879996544,"rtf":0.1306717979999424,"output_peak":0.2391219139099121,"output_rms":0.038421365677889646,"finished_at":"2026-10-13T05:03:11.799141+00:00","analysis_stages":{"external_f0_s":0.09764063400052692},"evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"806f235172212f4c8a9f9b9d6d1d7a05c6290eabeaacb48ba939a0103d6f282d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.303666828550913,"output_peak":0.2390126571692293},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.539406330060398,"spectral_convergence":0.42184949876445366,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.841219888953511,"spectral_convergence":0.27246924154333213,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.777519248262454,"spectral_convergence":0.2835484537449122,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016650763111812042,"output_power":0.0001412602337670333,"delta_db":-0.7141422159020917,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013663002185277922,"output_power":0.0009925834475977763,"delta_db":-1.3877910920507328,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000598592143060892,"output_power":0.00030915059750612363,"delta_db":-2.8696092191202514,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.819864545669523e-05,"output_power":1.5541901141098e-05,"delta_db":-0.685349220439838,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.4676579351935897e-05,"output_power":1.4201187807966824e-05,"delta_db":-2.399602870498146,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0360873265236357e-05,"output_power":5.122708868467593e-06,"delta_db":-3.0589668647668833,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.852576913796324,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.6974578689456066,"pitch":null,"output_rms_dbfs":-28.303666828550913,"output_peak":0.2390126571692293,"reference_rms_dbfs":-26.606208959605308,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.052715155758787},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9069826596491474,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04674009082982908,"rms_dbfs":-26.606208959605308,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.256346460549985,"mean":-1.3166755587111332e-06,"zero_energy":false},"output":{"rms":0.038442945733755444,"rms_dbfs":-28.303666828550913,"peak":0.2390126571692293,"peak_dbfs":-12.431581998342509,"crest_factor_db":15.872084830208404,"mean":0.00026294066038685964,"zero_energy":false},"rms_difference_db":-1.6974578689456052},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":2.854185059427481e-07,"output_rms_dbfs":-65.4451787155683,"output_zero_energy":false,"output_peak":0.004238186739595876,"output_energy_fraction":3.482136122405337e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":396,"output_estimated_voiced_frames":409,"both_estimated_voiced_frames":389,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9823232323232324,"absolute_difference_cents_median":5.341853497300785,"absolute_difference_cents_p95":16.40247415031407,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001446111949915804,"log_rms_mae_db":2.5918374561902287,"rms_envelope_nrmse":0.21240928908341192,"envelope_correlation":0.9810615230378799,"rms_gain_error_db":-1.6974775620394302,"reference_mean_power":0.0021864525140702735,"output_mean_power":0.0014790821416817647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.773568142242067e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032246698134311266,"output_power":0.0002176339865358945,"reference_active":true,"delta_db":-1.707585347652063},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.465956949470895e-05,"output_power":2.922150209006642e-05,"reference_active":true,"delta_db":-1.8421199581593863},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.553825573469442e-06,"output_power":6.413965526787847e-06,"reference_active":true,"delta_db":-1.730506866078296}],"active_band_count":3,"active_band_mae_db":1.7600707239632485}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00014111493703772434,"log_rms_mae_db":2.5255868735246816,"rms_envelope_nrmse":0.20247928056368464,"envelope_correlation":0.9856809380294276,"rms_gain_error_db":-1.6977030488035223,"reference_mean_power":0.0021971513908893054,"output_mean_power":0.001486242504505047},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.172558164545107e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028852159873825074,"output_power":0.00019454717912075208,"reference_active":true,"delta_db":-1.7115339197286952},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2878825691649602e-05,"output_power":1.5536009138573336e-05,"reference_active":true,"delta_db":-1.6809426150566675},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.05853919751672e-07,"output_power":4.4507953396396776e-07,"reference_active":true,"delta_db":-1.3393029739844609}],"active_band_count":3,"active_band_mae_db":1.577259836256608}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00012712872255296568,"log_rms_mae_db":2.066129117302528,"rms_envelope_nrmse":0.24047379803215815,"envelope_correlation":0.9780553907007562,"rms_gain_error_db":-1.923280837553179,"reference_mean_power":0.001483842742257426,"output_mean_power":0.0009529273527293364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":7.404277038668941e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00031620011837055293,"output_power":0.0002007373958643307,"reference_active":true,"delta_db":-1.9733374236164472},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.666559346998004e-05,"output_power":2.39683271258084e-05,"reference_active":true,"delta_db":-1.8462099546603348},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.304341128538832e-06,"output_power":4.3584571628295766e-06,"reference_active":true,"delta_db":-1.6030692296388593}],"active_band_count":3,"active_band_mae_db":1.8075388693052137}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00012343588785799389,"log_rms_mae_db":1.927968580112372,"rms_envelope_nrmse":0.22979674862397154,"envelope_correlation":0.9828394160978171,"rms_gain_error_db":-1.9234272122567366,"reference_mean_power":0.0014911979213371963,"output_mean_power":0.0009576185901586719},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":6.987631567695023e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028744034397032024,"output_power":0.00018228045093626093,"reference_active":true,"delta_db":-1.9780762960734237},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9708700273482822e-05,"output_power":1.3289992471327123e-05,"reference_active":true,"delta_db":-1.711332499562583},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.5018061696582766e-07,"output_power":3.553303425851941e-07,"reference_active":true,"delta_db":-1.0275449741639038}],"active_band_count":3,"active_band_mae_db":1.5723179232666367}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":2.398905704560611,"rms_envelope_nrmse":0.2478963095911909,"envelope_correlation":0.9721648338061087,"rms_gain_error_db":-1.7086972569204306,"reference_mean_power":3.675560263300603e-05,"output_mean_power":2.4800122270365412e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.6516523532276848e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3141845353177194e-05,"output_power":9.041964085164959e-06,"reference_active":true,"delta_db":-1.6239357464240416},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.985694997525994e-06,"output_power":1.3984984822758793e-06,"reference_active":true,"delta_db":-3.2938344108784987},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.112282969704011e-07,"output_power":4.4698472400515105e-07,"reference_active":true,"delta_db":-2.5885041006762}],"active_band_count":3,"active_band_mae_db":2.5020914193262467}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":2.1204721457964166,"rms_envelope_nrmse":0.21633648362071634,"envelope_correlation":0.9850780497868079,"rms_gain_error_db":-1.7085269623402388,"reference_mean_power":3.693386511985315e-05,"output_mean_power":2.49213785808281e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.463575307724032e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.270528198971661e-05,"output_power":8.599236802477116e-06,"reference_active":true,"delta_db":-1.695243999739484},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4092903751685283e-06,"output_power":7.092466774305906e-07,"reference_active":true,"delta_db":-2.9820317594210772},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.686575953471239e-08,"output_power":2.6206151763080708e-08,"reference_active":true,"delta_db":-3.3644759248458254}],"active_band_count":3,"active_band_mae_db":2.680583894668795}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":3.246229048323586,"rms_envelope_nrmse":0.3252734133682888,"envelope_correlation":0.9654239246239941,"rms_gain_error_db":-2.6434267711008497,"reference_mean_power":1.1991768557911005e-05,"output_mean_power":6.524399727236214e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.0447229351147145e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.250117012599323e-06,"output_power":2.4471011351924855e-06,"reference_active":true,"delta_db":-2.397489685518807},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.595690385892664e-06,"output_power":1.4321985956676997e-06,"reference_active":true,"delta_db":-2.582496449991491},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.496335980362266e-07,"output_power":4.878859520650811e-07,"reference_active":true,"delta_db":-1.243501644046908}],"active_band_count":3,"active_band_mae_db":2.074495926519069}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":3.198024429521823,"rms_envelope_nrmse":0.2939572818327874,"envelope_correlation":0.982197198220758,"rms_gain_error_db":-2.6434622298478274,"reference_mean_power":1.2052043528538895e-05,"output_mean_power":6.557140185394418e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.466306024908561e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.510708744194807e-06,"output_power":2.6948125912061115e-06,"reference_active":true,"delta_db":-2.2371621779463458},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3735303853381898e-06,"output_power":7.830820213595081e-07,"reference_active":true,"delta_db":-2.440310181643311},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.270130237895283e-08,"output_power":4.0934646397847076e-08,"reference_active":true,"delta_db":-0.18350077885205807}],"active_band_count":3,"active_band_mae_db":1.6203243794805715}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"sifigan","status":"measured","mos":{"status":"measured","value":4.063271522521973,"predicted_mos":4.063271522521973,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"1f4d3670b2ca3c6468f7b617dc6d9419ab87fe55b89ad62d4052ae56a53ca134","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.047457054999541},"started_at":"2026-10-13T07:47:47.216597+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0032475369989697356,"native_input_samples":144000,"native_input_sha256":"f862f06c343ae56f5d6da883d8058b41cadd89e855a153bc4acbb64b919b4eda","output_sha256":"6f6c73cf6f82b7a6e4a38a0015225f5e61a9bfed73ec5f47e768f15f2378d42d","output_samples":144120,"length_delta_samples":120,"analysis_s":2.6320554499980062,"synthesis_s":4.946324040000036,"total_s":7.578379489998042,"rtf":1.263063248333007,"output_peak":0.22535257041454315,"output_rms":0.04284189555418069,"finished_at":"2026-10-13T07:47:54.810981+00:00","analysis_stages":{"external_f0_s":1.5133726839994779},"evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"6f6c73cf6f82b7a6e4a38a0015225f5e61a9bfed73ec5f47e768f15f2378d42d","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.357630560223424,"output_peak":0.22588379179832135},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.020695267099313,"spectral_convergence":0.3947255330637484,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.610990468869533,"spectral_convergence":0.21013388315344508,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.3655829093813265,"spectral_convergence":0.21822267893847164,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.00014162027283566098,"delta_db":-0.6962675200265145,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.0011182000457905636,"delta_db":-0.8630223814157802,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0005284067954453965,"delta_db":-0.5343784730902403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":1.756670698122752e-05,"delta_db":-0.1462123373293221,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":2.500896731931639e-05,"delta_db":0.06536625382823619,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":8.268920504558761e-06,"delta_db":-0.9721970022318337,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5462406613203211,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7405911925320895,"pitch":null,"output_rms_dbfs":-27.35401336470297,"output_peak":0.22588379179832135,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.665756215116724},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9154862985645287,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04670129155060261,"rms_dbfs":-26.613422172170882,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.26355967311556,"mean":-1.4899571736653646e-06,"zero_energy":false},"output":{"rms":0.042884399364765784,"rms_dbfs":-27.35401336470297,"peak":0.22588379179832135,"peak_dbfs":-12.922298611791025,"crest_factor_db":14.431714752911946,"mean":4.8035174965342545e-05,"zero_energy":false},"rms_difference_db":-0.7405911925320865},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":2.447537752893697e-06,"output_rms_dbfs":-56.11270600547519,"output_zero_energy":false,"output_peak":0.07360437658107966,"output_energy_fraction":0.00023955389743336903,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":397,"output_estimated_voiced_frames":378,"both_estimated_voiced_frames":369,"reference_voiced_output_unvoiced_frames":28,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.929471032745592,"absolute_difference_cents_median":5.894901487266643,"absolute_difference_cents_p95":22.697276175177567,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001446111949915804,"log_rms_mae_db":1.9976196937835822,"rms_envelope_nrmse":0.14680186294261144,"envelope_correlation":0.9815043379616443,"rms_gain_error_db":-0.7405891080720386,"reference_mean_power":0.002182828397481346,"output_mean_power":0.0018406053680304645},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.790920693994897e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003230276115117931,"output_power":0.00028386713321845193,"reference_active":true,"delta_db":-0.5612453438069951},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.477236494661316e-05,"output_power":3.771570870700671e-05,"reference_active":true,"delta_db":-0.744877622517046},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.548345752003966e-06,"output_power":8.681615309442907e-06,"reference_active":true,"delta_db":-0.41327598782733405}],"active_band_count":3,"active_band_mae_db":0.573132984717125}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014111493703772434,"log_rms_mae_db":1.4519571523108947,"rms_envelope_nrmse":0.11924408135977756,"envelope_correlation":0.9898075743907492,"rms_gain_error_db":-0.7404576903971547,"reference_mean_power":0.002193693032291648,"output_mean_power":0.0018498226242662826},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.190420923463323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028912450266098784,"output_power":0.00025463592301439003,"reference_active":true,"delta_db":-0.551652265625449},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2978791346744758e-05,"output_power":1.9328785618178424e-05,"reference_active":true,"delta_db":-0.7512261246039796},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.072854493661426e-07,"output_power":5.598961418661131e-07,"reference_active":true,"delta_db":-0.35285400630490915}],"active_band_count":3,"active_band_mae_db":0.5519107988447792}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012712872255296568,"log_rms_mae_db":2.1258616628479086,"rms_envelope_nrmse":0.15820502699838696,"envelope_correlation":0.9814265411122213,"rms_gain_error_db":-0.7661026810080804,"reference_mean_power":0.0014813690704446063,"output_mean_power":0.0012418038576987012},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.409297669270013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003163260619212235,"output_power":0.0002695472652621127,"reference_active":true,"delta_db":-0.6950004468284299},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6701975061022124e-05,"output_power":2.987364333750044e-05,"reference_active":true,"delta_db":-0.8940124414595874},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.3015260364548945e-06,"output_power":5.365908743358723e-06,"reference_active":true,"delta_db":-0.6980245253894981}],"active_band_count":3,"active_band_mae_db":0.7623458045591717}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012343588785799389,"log_rms_mae_db":1.6007949971878004,"rms_envelope_nrmse":0.12767852820265208,"envelope_correlation":0.9897542378128095,"rms_gain_error_db":-0.7660537692925642,"reference_mean_power":0.0014887936512772218,"output_mean_power":0.0012480418000974847},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.99328908394483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002875949770086079,"output_power":0.0002460036314555645,"reference_active":true,"delta_db":-0.6783977848732075},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9768070879384797e-05,"output_power":1.6143733596634504e-05,"reference_active":true,"delta_db":-0.8796030734233101},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.520482053574086e-07,"output_power":4.4125123965764136e-07,"reference_active":true,"delta_db":-0.10498810912144615}],"active_band_count":3,"active_band_mae_db":0.5543296558059879}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":2.1896786192302296,"rms_envelope_nrmse":0.21980697382419648,"envelope_correlation":0.9619173149577717,"rms_gain_error_db":0.014122723285379035,"reference_mean_power":3.669443825843959e-05,"output_mean_power":3.6813958292894495e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6589829493109392e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.317934906149565e-05,"output_power":1.2823758777685598e-05,"reference_active":true,"delta_db":-0.11878620546031671},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9890642401085547e-06,"output_power":3.0987845971540166e-06,"reference_active":true,"delta_db":0.15656139720459147},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.136200949313618e-07,"output_power":7.098882572122681e-07,"reference_active":true,"delta_db":-0.5923167408148362}],"active_band_count":3,"active_band_mae_db":0.2892214478265815}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":1.3932271172621793,"rms_envelope_nrmse":0.12611903710287506,"envelope_correlation":0.9864043708842678,"rms_gain_error_db":0.014377758583914662,"reference_mean_power":3.687647047180045e-05,"output_mean_power":3.699875606909492e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4709890047082166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2743108138238891e-05,"output_power":1.244895785416086e-05,"reference_active":true,"delta_db":-0.1014237191089889},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4131557808728359e-06,"output_power":1.4422592982343834e-06,"reference_active":true,"delta_db":0.08853308086935613},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.713126831346598e-08,"output_power":5.27835039221374e-08,"reference_active":true,"delta_db":-0.3437564893106538}],"active_band_count":3,"active_band_mae_db":0.1779044297629996}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":2.364322178896534,"rms_envelope_nrmse":0.3181327009845156,"envelope_correlation":0.9263125075851147,"rms_gain_error_db":-0.7322967602489379,"reference_mean_power":1.1971789360385637e-05,"output_mean_power":1.0114150016445946e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0435686252011218e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2483829658988356e-06,"output_power":3.5874848301602017e-06,"reference_active":true,"delta_db":-0.7343358553322554},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.589400168019186e-06,"output_power":2.2535959234959946e-06,"reference_active":true,"delta_db":-0.6032312345775399},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.486231293405909e-07,"output_power":5.593362412653314e-07,"reference_active":true,"delta_db":-0.6431947154766462}],"active_band_count":3,"active_band_mae_db":0.6602539351288138}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":1.616128665835707,"rms_envelope_nrmse":0.19218453129938276,"envelope_correlation":0.9737693136773543,"rms_gain_error_db":-0.732284923465653,"reference_mean_power":1.203192810180695e-05,"output_mean_power":1.016498485079145e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.457061965350649e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5073827107587115e-06,"output_power":3.82028871240501e-06,"reference_active":true,"delta_db":-0.7182824922571187},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3712782786962116e-06,"output_power":1.2022182886058767e-06,"reference_active":true,"delta_db":-0.5714226642071317},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.274779912953661e-08,"output_power":3.860377361358204e-08,"reference_active":true,"delta_db":-0.44283999890145054}],"active_band_count":3,"active_band_mae_db":0.5775150517885671}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.777099132537842,"predicted_mos":3.777099132537842,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"8f81fae75a419e4191dcf14ca4b443dfa828b3b20bacad5009195caff08d9b22","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.219069562001096},"started_at":"2026-10-13T05:11:39.288832+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003920375000234344,"native_input_samples":144000,"native_input_sha256":"f862f06c343ae56f5d6da883d8058b41cadd89e855a153bc4acbb64b919b4eda","output_sha256":"f03ff708bf0dca9a2d782d9cff63ea7c0b2d39212c496f28de58f5062541fcb5","output_samples":144240,"length_delta_samples":240,"analysis_s":0.6965356779983267,"synthesis_s":0.16286304800087237,"total_s":0.8593987259991991,"rtf":0.14323312099986651,"output_peak":0.24552017450332642,"output_rms":0.04625829516535734,"finished_at":"2026-10-13T05:11:40.160204+00:00","analysis_stages":{"external_f0_s":0.14600859199890692},"evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"f03ff708bf0dca9a2d782d9cff63ea7c0b2d39212c496f28de58f5062541fcb5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.691091491158545,"output_peak":0.24614866410913092},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.52477868778279,"spectral_convergence":0.103077834701911,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.528844926300646,"spectral_convergence":0.09603758093780239,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.849385610176489,"spectral_convergence":0.11384992730863167,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016624637287256398,"output_power":0.00016482217775237647,"delta_db":-0.03736530673705977,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013640231584289317,"output_power":0.0013379020251067662,"delta_db":-0.08397432743821175,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005975940179232537,"output_power":0.0005931488393518744,"delta_db":-0.03242556494689835,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8168186983374496e-05,"output_power":1.7480903111080672e-05,"delta_db":-0.16747725193160828,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.463537251271816e-05,"output_power":2.3710150478845213e-05,"delta_db":-0.16624823369112018,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0343523773544458e-05,"output_power":8.858322514497077e-06,"delta_db":-0.6731702890477589,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.19344349563210952,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07043765948204095,"pitch":null,"output_rms_dbfs":-26.68385983165292,"output_peak":0.24614866410913092,"reference_rms_dbfs":-26.613422172170882,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.634336408086642},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9719997056860392,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04670129155060261,"rms_dbfs":-26.613422172170882,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.26355967311556,"mean":-1.4899571736653646e-06,"zero_energy":false},"output":{"rms":0.0463241019159081,"rms_dbfs":-26.68385983165292,"peak":0.24614866410913092,"peak_dbfs":-12.176050337504154,"crest_factor_db":14.507809494148766,"mean":1.1095859162167677e-06,"zero_energy":false},"rms_difference_db":-0.07043765948203884},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":3.2147114021592086e-07,"output_rms_dbfs":-64.92858009388456,"output_zero_energy":false,"output_peak":0.01175436914361138,"output_energy_fraction":2.6965003368709815e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":397,"output_estimated_voiced_frames":372,"both_estimated_voiced_frames":366,"reference_voiced_output_unvoiced_frames":31,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9219143576826196,"absolute_difference_cents_median":0.5372159223834256,"absolute_difference_cents_p95":5.080679846034558,"difference_over_600_cents_frames":1,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001446111949915804,"log_rms_mae_db":0.6114233923347708,"rms_envelope_nrmse":0.03963275151691549,"envelope_correlation":0.9982694452741935,"rms_gain_error_db":-0.07044683526108905,"reference_mean_power":0.002182828397481346,"output_mean_power":0.0021477064013218725},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.790920693994897e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003230276115117931,"output_power":0.00031923800154735855,"reference_active":true,"delta_db":-0.05125062785070435},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.477236494661316e-05,"output_power":4.329258392382214e-05,"reference_active":true,"delta_db":-0.14596527593167385},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.548345752003966e-06,"output_power":8.871265477235852e-06,"reference_active":true,"delta_db":-0.31942560785313423}],"active_band_count":3,"active_band_mae_db":0.17221383721183747}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00014111493703772434,"log_rms_mae_db":0.44824519720280526,"rms_envelope_nrmse":0.023205937082156913,"envelope_correlation":0.9994268503909393,"rms_gain_error_db":-0.07044738063539188,"reference_mean_power":0.002193693032291648,"output_mean_power":0.002158395951683307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.190420923463323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028912450266098784,"output_power":0.000286177971821691,"reference_active":true,"delta_db":-0.0444869726975308},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2978791346744758e-05,"output_power":2.2997897044941082e-05,"reference_active":true,"delta_db":0.0036094377683306912},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.072854493661426e-07,"output_power":6.091617182222809e-07,"reference_active":true,"delta_db":0.013397274928041204}],"active_band_count":3,"active_band_mae_db":0.0204978951313009}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00012712872255296568,"log_rms_mae_db":0.6781511642876669,"rms_envelope_nrmse":0.046892153330080164,"envelope_correlation":0.9979610816908637,"rms_gain_error_db":-0.07733347828478986,"reference_mean_power":0.0014813690704446063,"output_mean_power":0.001455224255665372},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":7.409297669270013e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003163260619212235,"output_power":0.00031302673678573344,"reference_active":true,"delta_db":-0.04553540812219945},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.6701975061022124e-05,"output_power":3.534839962062301e-05,"reference_active":true,"delta_db":-0.16319679644728674},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.3015260364548945e-06,"output_power":7.2636566990046605e-06,"reference_active":true,"delta_db":0.6170957500819304}],"active_band_count":3,"active_band_mae_db":0.2752759848838055}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00012343588785799389,"log_rms_mae_db":0.41195117304981393,"rms_envelope_nrmse":0.028264141692110566,"envelope_correlation":0.9992596945007343,"rms_gain_error_db":-0.07733499991935616,"reference_mean_power":0.0014887936512772218,"output_mean_power":0.001462517286987759},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":6.99328908394483e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002875949770086079,"output_power":0.0002848612311518897,"reference_active":true,"delta_db":-0.04147949655593913},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9768070879384797e-05,"output_power":1.9903715342969537e-05,"reference_active":true,"delta_db":0.02969862323418272},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.520482053574086e-07,"output_power":4.942638226400822e-07,"reference_active":true,"delta_db":0.3877407426304187}],"active_band_count":3,"active_band_mae_db":0.15297295414018017}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":1.0883053104165457,"rms_envelope_nrmse":0.06787298865468694,"envelope_correlation":0.9965900059935366,"rms_gain_error_db":-0.1649362243704958,"reference_mean_power":3.669443825843959e-05,"output_mean_power":3.532698884504528e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.6589829493109392e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.317934906149565e-05,"output_power":1.2709509787255639e-05,"reference_active":true,"delta_db":-0.1576516073740972},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9890642401085547e-06,"output_power":2.8024725144514418e-06,"reference_active":true,"delta_db":-0.2799388697324603},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.136200949313618e-07,"output_power":6.328806751412992e-07,"reference_active":true,"delta_db":-1.0909983152840326}],"active_band_count":3,"active_band_mae_db":0.50952959746353}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":0.7608788802935234,"rms_envelope_nrmse":0.03622541855759578,"envelope_correlation":0.9991870724550909,"rms_gain_error_db":-0.16501189177205755,"reference_mean_power":3.687647047180045e-05,"output_mean_power":3.550161892100539e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.4709890047082166e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.2743108138238891e-05,"output_power":1.2280896297027743e-05,"reference_active":true,"delta_db":-0.16045304463793736},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4131557808728359e-06,"output_power":1.334025618260927e-06,"reference_active":true,"delta_db":-0.25025869713038873},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.713126831346598e-08,"output_power":4.4894561057511904e-08,"reference_active":true,"delta_db":-1.0468013576125381}],"active_band_count":3,"active_band_mae_db":0.4858376997936214}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":1.3325183674321572,"rms_envelope_nrmse":0.15569944722617898,"envelope_correlation":0.9828548795851386,"rms_gain_error_db":-0.23788217639733805,"reference_mean_power":1.1971789360385637e-05,"output_mean_power":1.1333677466234786e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.0435686252011218e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.2483829658988356e-06,"output_power":4.292961155532713e-06,"reference_active":true,"delta_db":0.04533299977679282},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.589400168019186e-06,"output_power":2.597922396432419e-06,"reference_active":true,"delta_db":0.01427002102661855},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.486231293405909e-07,"output_power":5.079942373294664e-07,"reference_active":true,"delta_db":-1.0613364549019375}],"active_band_count":3,"active_band_mae_db":0.37364649190178295}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":1.0297940937857273,"rms_envelope_nrmse":0.10231375382052485,"envelope_correlation":0.992188132771474,"rms_gain_error_db":-0.23788324793851834,"reference_mean_power":1.203192810180695e-05,"output_mean_power":1.1390607924347545e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.457061965350649e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.5073827107587115e-06,"output_power":4.534903841978425e-06,"reference_active":true,"delta_db":0.0264364821036492},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3712782786962116e-06,"output_power":1.3542617511129315e-06,"reference_active":true,"delta_db":-0.05422984050127894},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.274779912953661e-08,"output_power":3.55792439428413e-08,"reference_active":true,"delta_db":-0.7971704487822124}],"active_band_count":3,"active_band_mae_db":0.29261225712904687}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284447-0007","method":"magphase","status":"measured","mos":{"status":"measured","value":3.4701130390167236,"predicted_mos":3.4701130390167236,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287078,"input_sample_rate":48000,"input_seconds":5.980791666666667,"input_float32_sha256":"4b68a0c6c51a9fb015d4960966968518316444abc523db2b7b548f70c3af04b4","internal_samples":95693,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.684937350000837},"started_at":"2026-10-13T05:19:06.186480+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.012354951000816072,"native_input_samples":288000,"native_input_sha256":"17116c303553d968bf859abd27c48d17bdf87aef0e9797ae04659dd3b96472de","output_sha256":"1d7bed68e8e97e57a2ab033b52bb81290a1b8f4e51174ec9cb312e842d618bdd","output_samples":287078,"length_delta_samples":-922,"analysis_s":4.825567227000647,"synthesis_s":1.152602696998656,"total_s":5.978169923999303,"rtf":0.9963616539998839,"output_peak":0.2467388972856359,"output_rms":0.04439201644682059,"finished_at":"2026-10-13T05:19:12.191453+00:00","evaluation_input":{"reference":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"1d7bed68e8e97e57a2ab033b52bb81290a1b8f4e51174ec9cb312e842d618bdd","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95693,"length_delta_samples":-307,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.04896797153438,"output_peak":0.24322904608807383},"common_prefix":{"samples":95693,"sample_rate":16000,"reference_coverage":0.9968020833333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95693,"output_samples":95693,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.189708515795188,"spectral_convergence":0.4018391435605218,"spectral_convergence_status":"ok","amplitude_floor":9.671435620644006e-07},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.390706410156038,"spectral_convergence":0.15554802639471696,"spectral_convergence_status":"ok","amplitude_floor":9.322400769888516e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.460880370080442,"spectral_convergence":0.1844687114294457,"spectral_convergence_status":"ok","amplitude_floor":9.453157582642776e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00016662371950193944,"output_power":0.00015079311651964257,"delta_db":-0.43355307802930065,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013684959565734719,"output_power":0.0012196947522062187,"delta_db":-0.49992363311025584,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005995057413014197,"output_power":0.0005470721346081901,"delta_db":-0.3974875223850408,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":1.8227135487459837e-05,"output_power":1.8507285222207517e-05,"delta_db":0.06624296255048834,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.4710765664355952e-05,"output_power":2.4515583758107406e-05,"delta_db":-0.03443963245614868,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":1.0376445549016264e-05,"output_power":1.2308632890743088e-05,"delta_db":0.741612074653814,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.36220981719750805,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.44932158467122885,"pitch":null,"output_rms_dbfs":-27.04896797153438,"output_peak":0.24322904608807383,"reference_rms_dbfs":-26.599646386863153,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.68043176534389},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95693,"output_samples":95693,"intelligibility":{"estoi":{"value":0.9534568764866311,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04677541837415955,"rms_dbfs":-26.59964638686315,"peak":0.24127197265625,"peak_dbfs":-12.349862499055321,"crest_factor_db":14.24978388780783,"mean":-2.7844144776459616e-06,"zero_energy":false},"output":{"rms":0.04441724333189881,"rms_dbfs":-27.04896797153438,"peak":0.24322904608807383,"peak_dbfs":-12.279691268698665,"crest_factor_db":14.769276702835716,"mean":0.0007715310717527769,"zero_energy":false},"rms_difference_db":-0.44932158467123173},"full_scale":{"reference":{"sample_count":95693,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95693,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":13,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":2.0879258244531233e-06,"threshold_dbfs":-56.80284934119445,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":108,"quiet_samples":17280,"reference_mean_power":1.9636090318844827e-07,"output_mean_power":2.684196990441815e-07,"output_rms_dbfs":-65.71185614910456,"output_zero_energy":false,"output_peak":0.00734880360051677,"output_energy_fraction":2.4568279847984042e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":395,"output_estimated_voiced_frames":392,"both_estimated_voiced_frames":388,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":4,"paired_voiced_fraction":0.9822784810126582,"absolute_difference_cents_median":4.0223377266112905,"absolute_difference_cents_p95":15.052800827160134,"difference_over_600_cents_frames":2,"uncovered_tail_samples":13,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95693,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1195,"rms_floor":0.0001446111949915804,"log_rms_mae_db":1.0798771397988425,"rms_envelope_nrmse":0.09004203667481153,"envelope_correlation":0.9932918266230293,"rms_gain_error_db":-0.44934783877604717,"reference_mean_power":0.0021900327902708095,"output_mean_power":0.0019747668740478916},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":9.756108911110938e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00032187372792171924,"output_power":0.00029508883055268364,"reference_active":true,"delta_db":-0.3773275874015465},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.455150979619231e-05,"output_power":4.0519311299008675e-05,"reference_active":true,"delta_db":-0.41200371222028154},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.55356576148721e-06,"output_power":9.550661869199244e-06,"reference_active":true,"delta_db":-0.0013202778262462497}],"active_band_count":3,"active_band_mae_db":0.2635505258160247}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1189,"rms_floor":0.00014111493703772434,"log_rms_mae_db":0.7832659111554822,"rms_envelope_nrmse":0.06692294187370419,"envelope_correlation":0.9974697160118784,"rms_gain_error_db":-0.4494230407818039,"reference_mean_power":0.0022003745540808708,"output_mean_power":0.0019840577539895965},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":9.154458854596019e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002878864483870087,"output_power":0.000263772149874371,"reference_active":true,"delta_db":-0.3799228284343939},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.2782584782682672e-05,"output_power":2.0299850859126652e-05,"reference_active":true,"delta_db":-0.5011014791489601},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.072734522289522e-07,"output_power":5.624801697929746e-07,"reference_active":true,"delta_db":-0.33277079747960786}],"active_band_count":3,"active_band_mae_db":0.4045983683543206}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1195,"rms_floor":0.00012712872255296568,"log_rms_mae_db":1.2172184545753926,"rms_envelope_nrmse":0.09482187394545057,"envelope_correlation":0.9934502864918145,"rms_gain_error_db":-0.44068780227051263,"reference_mean_power":0.0014863157536772875,"output_mean_power":0.0013428957542301377},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":7.399185537034209e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000316066153615613,"output_power":0.0002907351907531859,"reference_active":true,"delta_db":-0.36280389096231463},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.662361583406197e-05,"output_power":3.165028130189502e-05,"reference_active":true,"delta_db":-0.6338364549228721},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.299632435044405e-06,"output_power":5.189528780876922e-06,"reference_active":true,"delta_db":-0.8418728552140263}],"active_band_count":3,"active_band_mae_db":0.6128377336997377}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1189,"rms_floor":0.00012343588785799389,"log_rms_mae_db":0.8911480051447318,"rms_envelope_nrmse":0.07100444824688278,"envelope_correlation":0.9972233807679874,"rms_gain_error_db":-0.4408908449041815,"reference_mean_power":0.0014934575898990494,"output_mean_power":0.001349285365103251},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":6.981673052060581e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00028727881200070953,"output_power":0.0002648557609942171,"reference_active":true,"delta_db":-0.3529417225599218},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9639651246564478e-05,"output_power":1.678213057747352e-05,"reference_active":true,"delta_db":-0.6828667558084598},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.5315184413287207e-07,"output_power":3.8368416211472814e-07,"reference_active":true,"delta_db":-0.7226987915631307}],"active_band_count":3,"active_band_mae_db":0.5861690899771708}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1195,"rms_floor":3.4084091617788044e-05,"log_rms_mae_db":1.2665971419646154,"rms_envelope_nrmse":0.1354192893975237,"envelope_correlation":0.9855049030571659,"rms_gain_error_db":-0.03887117361230788,"reference_mean_power":3.6816852408551724e-05,"output_mean_power":3.6488796480949005e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":2.6442721908941114e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3103742815746707e-05,"output_power":1.3033598770288104e-05,"reference_active":true,"delta_db":-0.023310132680449454},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.9820464170889407e-06,"output_power":2.9446873004704654e-06,"reference_active":true,"delta_db":-0.05475215816129559},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.085086055728538e-07,"output_power":6.549770664518426e-07,"reference_active":true,"delta_db":-0.914585530430593}],"active_band_count":3,"active_band_mae_db":0.3308826070907793}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1189,"rms_floor":3.1514621722116854e-05,"log_rms_mae_db":0.8308200196031159,"rms_envelope_nrmse":0.07275618119278326,"envelope_correlation":0.995486147808339,"rms_gain_error_db":-0.03907780545780512,"reference_mean_power":3.6991157780707836e-05,"output_mean_power":3.6659804439271654e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":2.456107073560694e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.266695326847903e-05,"output_power":1.2562520941933347e-05,"reference_active":true,"delta_db":-0.03595369590850942},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.4048540307391891e-06,"output_power":1.4010569225398332e-06,"reference_active":true,"delta_db":-0.011754215812756987},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.705082874923385e-08,"output_power":4.594302206702758e-08,"reference_active":true,"delta_db":-0.9404239850613636}],"active_band_count":3,"active_band_mae_db":0.32937729892754336}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1195,"rms_floor":1.703026630874025e-05,"log_rms_mae_db":1.5060150414408515,"rms_envelope_nrmse":0.2433238596711766,"envelope_correlation":0.9657272386189018,"rms_gain_error_db":0.6845888394073825,"reference_mean_power":1.201181485163395e-05,"output_mean_power":1.4062661186222948e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":1.0458778679611077e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.251923887639918e-06,"output_power":5.176556196519422e-06,"reference_active":true,"delta_db":0.8545545130789457},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6019433185450587e-06,"output_power":3.5288801522898507e-06,"reference_active":true,"delta_db":1.3233907757853878},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.495145431682379e-07,"output_power":1.0300070433253023e-06,"reference_active":true,"delta_db":2.0025131474797826}],"active_band_count":3,"active_band_mae_db":1.3934861454480387}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":13,"metrics":{"frame_count":1189,"rms_floor":1.3691778537856404e-05,"log_rms_mae_db":1.0581463199052532,"rms_envelope_nrmse":0.15948088681250194,"envelope_correlation":0.9864995329983839,"rms_gain_error_db":0.6845631845719004,"reference_mean_power":1.2072185235818177e-05,"output_mean_power":1.4133255464979257e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":9.475559544013283e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.514099821124786e-06,"output_power":5.6594647568578346e-06,"reference_active":true,"delta_db":0.9820420130416508},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3751816027508144e-06,"output_power":1.9783819890174037e-06,"reference_active":true,"delta_db":1.5795009589545557},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.336157377840836e-08,"output_power":7.499666311850012e-08,"reference_active":true,"delta_db":2.37936904203865}],"active_band_count":3,"active_band_mae_db":1.646970671344952}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"hnm","status":"measured","mos":{"status":"measured","value":2.839185953140259,"predicted_mos":2.839185953140259,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":81520,"input_sample_rate":16000,"input_seconds":5.095,"input_float32_sha256":"10b932dfedb8e85e5b4591e158e284a11c21339d60195c2d40c4d7a6d0d82be3","internal_samples":81520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3081794819991046},"started_at":"2026-10-13T04:40:33.335560+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.2606000382220373e-05,"native_input_samples":81520,"native_input_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","output_sha256":"02ed0e95029e11998370d20304c09d7728985bd33143e0a5c7ea8f24c04719bf","output_samples":81520,"length_delta_samples":0,"analysis_s":0.8884461759998885,"synthesis_s":0.4197408219997669,"total_s":1.3081869979996554,"rtf":0.256758978998951,"output_peak":0.31659714076906226,"output_rms":0.043322700878710686,"finished_at":"2026-10-13T04:40:34.648980+00:00","evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"02ed0e95029e11998370d20304c09d7728985bd33143e0a5c7ea8f24c04719bf","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.499865264792755,"spectral_convergence":0.18782523189050057,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.621486080705515,"spectral_convergence":0.18881293426571485,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.989428008911196,"spectral_convergence":0.19031843359634726,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":9.65029861694397e-05,"delta_db":0.33575541108455065,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0011309368847536323,"delta_db":-0.06522241474339892,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0005155070985053108,"delta_db":-0.29590062635714687,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":8.829030188342162e-05,"delta_db":-0.12415633517314709,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":1.9347251878491344e-05,"delta_db":-0.29564355168098694,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.6271888235942754e-05,"delta_db":-0.3018953474448641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23642894774734904,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11890284314867833,"pitch":null,"output_rms_dbfs":-27.265689517316627,"output_peak":0.31659714076906226,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.703593118136489},"common_prefix":{"samples":81520,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.499865264792755,"spectral_convergence":0.18782523189050057,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.621486080705515,"spectral_convergence":0.18881293426571485,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.989428008911196,"spectral_convergence":0.19031843359634726,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":9.65029861694397e-05,"delta_db":0.33575541108455065,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0011309368847536323,"delta_db":-0.06522241474339892,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0005155070985053108,"delta_db":-0.29590062635714687,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":8.829030188342162e-05,"delta_db":-0.12415633517314709,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":1.9347251878491344e-05,"delta_db":-0.29564355168098694,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.6271888235942754e-05,"delta_db":-0.3018953474448641,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23642894774734904,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11890284314867833,"pitch":null,"output_rms_dbfs":-27.265689517316627,"output_peak":0.31659714076906226,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.703593118136489},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81520,"output_samples":81520,"intelligibility":{"estoi":{"value":0.9228744736495437,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04391983160443005,"rms_dbfs":-27.14678667416795,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.47416011838686,"mean":-6.37155519734421e-07,"zero_energy":false},"output":{"rms":0.043322700878710686,"rms_dbfs":-27.265689517316627,"peak":0.31659714076906226,"peak_dbfs":-9.989860232545016,"crest_factor_db":17.27582928477161,"mean":4.5942273002449665e-06,"zero_energy":false},"rms_difference_db":-0.11890284314867827},"full_scale":{"reference":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":3.911217419329144e-06,"output_rms_dbfs":-54.07688041529228,"output_zero_energy":false,"output_peak":0.05455627780178173,"output_energy_fraction":0.0003067594745915873,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":260,"both_estimated_voiced_frames":245,"reference_voiced_output_unvoiced_frames":18,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9315589353612167,"absolute_difference_cents_median":1.1974514376313907,"absolute_difference_cents_p95":9.909687885015261,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.00013747273046535536,"log_rms_mae_db":0.8999334529007895,"rms_envelope_nrmse":0.07575331962161784,"envelope_correlation":0.9935964238146168,"rms_gain_error_db":-0.11890278968980317,"reference_mean_power":0.0019308462257719779,"output_mean_power":0.0018786998842238718},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":6.283614364116833e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041565087617635905,"output_power":0.0003974166013080525,"reference_active":true,"delta_db":-0.19482695352508125},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151711693867131e-05,"output_power":7.818209541781435e-05,"reference_active":true,"delta_db":-0.18141505225877858},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.519660659304176e-05,"output_power":2.0203387991334576e-05,"reference_active":true,"delta_db":-0.9591785095331382}],"active_band_count":3,"active_band_mae_db":0.44514017177233267}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012452330782545057,"log_rms_mae_db":0.45581618588002554,"rms_envelope_nrmse":0.04639357396711105,"envelope_correlation":0.9973928423232157,"rms_gain_error_db":-0.1189028679513863,"reference_mean_power":0.0019422928952156904,"output_mean_power":0.0018898373795435178},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.257129910319434e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037070296778873537,"output_power":0.00035277852378597877,"reference_active":true,"delta_db":-0.21523924358979346},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.263528086715046e-05,"output_power":4.1386584805883305e-05,"reference_active":true,"delta_db":-0.1290953813681024},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8692589849180216e-06,"output_power":1.4709568580375407e-06,"reference_active":true,"delta_db":-1.0406954142349942}],"active_band_count":3,"active_band_mae_db":0.4616766797309633}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.0001332544033329483,"log_rms_mae_db":1.6057559202771394,"rms_envelope_nrmse":0.10057288658599084,"envelope_correlation":0.9911569563706436,"rms_gain_error_db":-0.18054027409515036,"reference_mean_power":0.00116782876738011,"output_mean_power":0.001120276282854846},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":5.939192180731074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041349134282540233,"output_power":0.00039090901894321277,"reference_active":true,"delta_db":-0.24390730790261853},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.995066727507786e-05,"output_power":6.436975047464734e-05,"reference_active":true,"delta_db":-0.3611003606335261},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.967138917853307e-05,"output_power":1.583618468531602e-05,"reference_active":true,"delta_db":-0.9418447259386699}],"active_band_count":3,"active_band_mae_db":0.5156174648249382}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012167319573637772,"log_rms_mae_db":0.9706178993437617,"rms_envelope_nrmse":0.06560323136682791,"envelope_correlation":0.996023693670105,"rms_gain_error_db":-0.1805400777632587,"reference_mean_power":0.0011747523056415098,"output_mean_power":0.0011269179544928431},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.189128511363634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037925439414056246,"output_power":0.00035567730138929475,"reference_active":true,"delta_db":-0.2787447106839991},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.625302583005344e-05,"output_power":3.386277860473077e-05,"reference_active":true,"delta_db":-0.29621669266999817},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7116807182635842e-06,"output_power":1.3467749895802605e-06,"reference_active":true,"delta_db":-1.0412771577387894}],"active_band_count":3,"active_band_mae_db":0.5387461870309288}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":1.5251863332439424,"rms_envelope_nrmse":0.1133410892277543,"envelope_correlation":0.9901229325092282,"rms_gain_error_db":-0.1599796903823127,"reference_mean_power":0.0001030241533277173,"output_mean_power":9.929813347443118e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":3.1359809483594144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4569632781045499e-05,"output_power":1.3944289240687615e-05,"reference_active":true,"delta_db":-0.1905222318490138},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.842890886644966e-06,"output_power":3.414638222987338e-06,"reference_active":true,"delta_db":-0.5131335602128646},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1134901533138973e-06,"output_power":9.440807877458472e-07,"reference_active":true,"delta_db":-0.7167722115134533}],"active_band_count":3,"active_band_mae_db":0.4734760011917773}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":0.8410833191226977,"rms_envelope_nrmse":0.05762060017689198,"envelope_correlation":0.9974552164585355,"rms_gain_error_db":-0.15998299313311412,"reference_mean_power":0.00010363449577271816,"output_mean_power":9.988632602605274e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":2.7762525532534862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3720292251105873e-05,"output_power":1.3126885287158418e-05,"reference_active":true,"delta_db":-0.19201672163056252},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7263230443470644e-06,"output_power":1.490225093157379e-06,"reference_active":true,"delta_db":-0.6387019589404743},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.698769643844656e-08,"output_power":7.342923405298648e-08,"reference_active":true,"delta_db":-1.2084764642890058}],"active_band_count":3,"active_band_mae_db":0.6797317149533475}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":1.8009474426715537,"rms_envelope_nrmse":0.26108639417477886,"envelope_correlation":0.9455316163568266,"rms_gain_error_db":0.1843127082148197,"reference_mean_power":8.51790564150615e-06,"output_mean_power":8.887182439370646e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":4.824507951602351e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4329635827636274e-06,"output_power":2.3367557717405503e-06,"reference_active":true,"delta_db":-0.17522284284009548},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1930715407234241e-06,"output_power":1.0268994803441409e-06,"reference_active":true,"delta_db":-0.6513855217466913},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.146506834057095e-07,"output_power":1.79851596829187e-07,"reference_active":true,"delta_db":-2.4291438125491087}],"active_band_count":3,"active_band_mae_db":1.0852507257119652}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":1.328310330807306,"rms_envelope_nrmse":0.20575819297319847,"envelope_correlation":0.9622667550315543,"rms_gain_error_db":0.18432043619856026,"reference_mean_power":8.568334809688712e-06,"output_mean_power":8.93981377128077e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":4.130413403311083e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5185957053959056e-06,"output_power":2.3360190655644706e-06,"reference_active":true,"delta_db":-0.32682075507455133},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.297939167002703e-07,"output_power":5.482357808667781e-07,"reference_active":true,"delta_db":-0.6023108482479101},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9684693476821212e-08,"output_power":1.3659858783859202e-08,"reference_active":true,"delta_db":-3.370863591587315}],"active_band_count":3,"active_band_mae_db":1.433331731636592}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.3007524013519287,"predicted_mos":3.3007524013519287,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":81520,"input_sample_rate":16000,"input_seconds":5.095,"input_float32_sha256":"15c953a74195900a44c3e2f2eea21e23507957642b604140458f9888c6d4242f","internal_samples":81520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1702884349997476},"started_at":"2026-10-13T04:43:49.848865+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3152000721893273e-05,"native_input_samples":81520,"native_input_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","output_sha256":"b72033d65b88915ff213f5cb30860a10f03a6a0b3e283ec492ef6b68de05afe4","output_samples":81520,"length_delta_samples":0,"analysis_s":1.3803842439992877,"synthesis_s":0.08472488400002476,"total_s":1.4651091279993125,"rtf":0.2875582194306796,"output_peak":0.3824975216512168,"output_rms":0.0527532335980359,"finished_at":"2026-10-13T04:43:51.321611+00:00","evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"b72033d65b88915ff213f5cb30860a10f03a6a0b3e283ec492ef6b68de05afe4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.003218820537746,"spectral_convergence":0.5066001314845633,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.430578231180643,"spectral_convergence":0.3261518894717556,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.97704134001575,"spectral_convergence":0.35945736242236953,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":8.89976086169218e-05,"delta_db":-0.015868739419941835,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0017071251566834593,"delta_db":1.7230475227718887,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0008242334089871777,"delta_db":1.7422550147145497,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":0.00010717823814549396,"delta_db":0.7177797831118754,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":2.5446545022771685e-05,"delta_db":0.8944518394428427,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.992269759311742e-05,"delta_db":0.26319920861657303,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8927670180129451,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5917683860714842,"pitch":null,"output_rms_dbfs":-25.555018288096463,"output_peak":0.3824975216512168,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.803612797244713},"common_prefix":{"samples":81520,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.003218820537746,"spectral_convergence":0.5066001314845633,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.430578231180643,"spectral_convergence":0.3261518894717556,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.97704134001575,"spectral_convergence":0.35945736242236953,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":8.89976086169218e-05,"delta_db":-0.015868739419941835,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0017071251566834593,"delta_db":1.7230475227718887,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0008242334089871777,"delta_db":1.7422550147145497,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":0.00010717823814549396,"delta_db":0.7177797831118754,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":2.5446545022771685e-05,"delta_db":0.8944518394428427,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.992269759311742e-05,"delta_db":0.26319920861657303,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8927670180129451,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5917683860714842,"pitch":null,"output_rms_dbfs":-25.555018288096463,"output_peak":0.3824975216512168,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.803612797244713},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81520,"output_samples":81520,"intelligibility":{"estoi":{"value":0.9433514509163495,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04391983160443005,"rms_dbfs":-27.14678667416795,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.47416011838686,"mean":-6.37155519734421e-07,"zero_energy":false},"output":{"rms":0.0527532335980359,"rms_dbfs":-25.555018288096463,"peak":0.3824975216512168,"peak_dbfs":-8.34742748924977,"crest_factor_db":17.207590798846695,"mean":9.45422040556563e-10,"zero_energy":false},"rms_difference_db":1.591768386071486},"full_scale":{"reference":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":1.6446462110302788e-06,"output_rms_dbfs":-57.83927511162168,"output_zero_energy":false,"output_peak":0.04182848295726023,"output_energy_fraction":8.699441914285575e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":276,"both_estimated_voiced_frames":250,"reference_voiced_output_unvoiced_frames":13,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.9505703422053232,"absolute_difference_cents_median":7.932256517509261,"absolute_difference_cents_p95":1201.2555477071496,"difference_over_600_cents_frames":24,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.00013747273046535536,"log_rms_mae_db":1.7839841692772958,"rms_envelope_nrmse":0.24775304556080904,"envelope_correlation":0.9809448851822573,"rms_gain_error_db":1.5917686262406037,"reference_mean_power":0.0019308462257719779,"output_mean_power":0.0027856371790019956},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":6.283614364116833e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041565087617635905,"output_power":0.000605445512494363,"reference_active":true,"delta_db":1.6334636427725822},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151711693867131e-05,"output_power":0.0001227205762981668,"reference_active":true,"delta_db":1.7766857472998296},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.519660659304176e-05,"output_power":2.9883368680514952e-05,"reference_active":true,"delta_db":0.740874976985406}],"active_band_count":3,"active_band_mae_db":1.3836747890192724}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012452330782545057,"log_rms_mae_db":1.3465986551607199,"rms_envelope_nrmse":0.22924964866794262,"envelope_correlation":0.9879650014936759,"rms_gain_error_db":1.5917690528155346,"reference_mean_power":0.0019422928952156904,"output_mean_power":0.0028021515958352654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.257129910319434e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037070296778873537,"output_power":0.0005443210706890175,"reference_active":true,"delta_db":1.6682908326239883},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.263528086715046e-05,"output_power":6.522343784311411e-05,"reference_active":true,"delta_db":1.846345578567317},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8692589849180216e-06,"output_power":1.8200209009664654e-06,"reference_active":true,"delta_db":-0.11593101420703353}],"active_band_count":3,"active_band_mae_db":1.2101891417994464}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.0001332544033329483,"log_rms_mae_db":2.168722578402443,"rms_envelope_nrmse":0.3062361137059719,"envelope_correlation":0.9806161866605553,"rms_gain_error_db":1.982189930636988,"reference_mean_power":0.00116782876738011,"output_mean_power":0.0018433090790952085},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":5.939192180731074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041349134282540233,"output_power":0.000624158595308195,"reference_active":true,"delta_db":1.7882853431461874},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.995066727507786e-05,"output_power":0.00011514246304245029,"reference_active":true,"delta_db":2.164436536450722},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.967138917853307e-05,"output_power":2.5303082240965e-05,"reference_active":true,"delta_db":1.0933839650558195}],"active_band_count":3,"active_band_mae_db":1.6820352815509096}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012167319573637772,"log_rms_mae_db":1.7733782653177386,"rms_envelope_nrmse":0.28745420775834857,"envelope_correlation":0.987221893778185,"rms_gain_error_db":1.9821904839655133,"reference_mean_power":0.0011747523056415098,"output_mean_power":0.0018542374764663064},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.189128511363634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037925439414056246,"output_power":0.0005737460069923253,"reference_active":true,"delta_db":1.7978905472915159},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.625302583005344e-05,"output_power":5.990949829590065e-05,"reference_active":true,"delta_db":2.1815142220005104},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7116807182635842e-06,"output_power":1.6078928948682926e-06,"reference_active":true,"delta_db":-0.2716564234976114}],"active_band_count":3,"active_band_mae_db":1.4170203975965459}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":1.5755559192516806,"rms_envelope_nrmse":0.1567612957602302,"envelope_correlation":0.9889503630547019,"rms_gain_error_db":0.7877509657643103,"reference_mean_power":0.0001030241533277173,"output_mean_power":0.00012351342098940022},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":3.1359809483594144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4569632781045499e-05,"output_power":1.7002726603040164e-05,"reference_active":true,"delta_db":0.6706996582304385},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.842890886644966e-06,"output_power":4.246974047827317e-06,"reference_active":true,"delta_db":0.4342155366349015},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1134901533138973e-06,"output_power":1.2986902150969126e-06,"reference_active":true,"delta_db":0.6681918749478156}],"active_band_count":3,"active_band_mae_db":0.5910356899377186}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":1.13474793869275,"rms_envelope_nrmse":0.11564980110385412,"envelope_correlation":0.9966604254347211,"rms_gain_error_db":0.787750959781321,"reference_mean_power":0.00010363449577271816,"output_mean_power":0.00012424514712631134},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":2.7762525532534862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3720292251105873e-05,"output_power":1.5956593315932617e-05,"reference_active":true,"delta_db":0.6557681413926857},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7263230443470644e-06,"output_power":1.9394775863942414e-06,"reference_active":true,"delta_db":0.5056269714805596},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.698769643844656e-08,"output_power":9.32366362893272e-08,"reference_active":true,"delta_db":-0.17130047464431192}],"active_band_count":3,"active_band_mae_db":0.44423186250585234}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":1.5494038876083729,"rms_envelope_nrmse":0.2004392412404062,"envelope_correlation":0.9692854914141932,"rms_gain_error_db":0.30587998386402426,"reference_mean_power":8.51790564150615e-06,"output_mean_power":9.13946603432916e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":4.824507951602351e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4329635827636274e-06,"output_power":2.6708989638450304e-06,"reference_active":true,"delta_db":0.4052185126871751},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1930715407234241e-06,"output_power":1.1838487934371572e-06,"reference_active":true,"delta_db":-0.033702504016604316},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.146506834057095e-07,"output_power":2.5562636657303824e-07,"reference_active":true,"delta_db":-0.902230322640969}],"active_band_count":3,"active_band_mae_db":0.4470504464482495}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":1.0914030685858984,"rms_envelope_nrmse":0.12344620690302431,"envelope_correlation":0.9877654848871658,"rms_gain_error_db":0.305885927914459,"reference_mean_power":8.568334809688712e-06,"output_mean_power":9.193587654083188e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":4.130413403311083e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5185957053959056e-06,"output_power":2.6928128039168728e-06,"reference_active":true,"delta_db":0.29047705157487946},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.297939167002703e-07,"output_power":6.248840900215154e-07,"reference_active":true,"delta_db":-0.03398994015312967},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9684693476821212e-08,"output_power":2.0858710868235266e-08,"reference_active":true,"delta_db":-1.5324510457435467}],"active_band_count":3,"active_band_mae_db":0.6189726791571853}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"facodec","status":"measured","mos":{"status":"measured","value":3.6958675384521484,"predicted_mos":3.6958675384521484,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":81600,"input_sample_rate":16000,"input_seconds":5.1,"input_float32_sha256":"d25b12f8cd3434b608ae35639b490fae613a13a132a9a559bf3e553235152a83","internal_samples":81600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":8.465240399000322},"started_at":"2026-10-13T07:02:30.290011+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":8.225800047512166e-05,"native_input_samples":81520,"native_input_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","output_sha256":"9152f4636d39691c87296054dd8a5b52563f0c8e77c7968879a2c71203f19846","output_samples":81600,"length_delta_samples":80,"analysis_s":1.5489842500010127,"synthesis_s":10.427946122999856,"total_s":11.976930373000869,"rtf":2.350722349951103,"output_peak":0.2560080289840698,"output_rms":0.03871529523604018,"finished_at":"2026-10-13T07:02:42.276450+00:00","evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"9152f4636d39691c87296054dd8a5b52563f0c8e77c7968879a2c71203f19846","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":81520,"output_samples":81600,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.24234849068795,"output_peak":0.2560080289840698},"common_prefix":{"samples":81520,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990196078431373,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.419245221734597,"spectral_convergence":0.25126794925568385,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.052859060131315,"spectral_convergence":0.24429891996975794,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.214403018961221,"spectral_convergence":0.2615067684889591,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":9.054203780578738e-05,"delta_db":0.05885053654333964,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0008815237414284518,"delta_db":-1.1472659667441938,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.00043201010504019055,"delta_db":-1.063308071011275,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":5.855607184745541e-05,"delta_db":-1.907567002979207,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":1.5671094825166e-05,"delta_db":-1.210843023475848,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.2041940292692092e-05,"delta_db":-1.06431001005001,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0753574351339787,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0913019981778491,"pitch":null,"output_rms_dbfs":-28.238088672345796,"output_peak":0.2560080289840698,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.895502433609043},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81520,"output_samples":81520,"intelligibility":{"estoi":{"value":0.9224442293581219,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04391983160443005,"rms_dbfs":-27.14678667416795,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.47416011838686,"mean":-6.37155519734421e-07,"zero_energy":false},"output":{"rms":0.03873428702376932,"rms_dbfs":-28.238088672345793,"peak":0.2560080289840698,"peak_dbfs":-11.834928280575738,"crest_factor_db":16.403160391770058,"mean":5.45457944950415e-06,"zero_energy":false},"rms_difference_db":-1.091301998177844},"full_scale":{"reference":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":4.862237155871573e-07,"output_rms_dbfs":-63.13163862240003,"output_zero_energy":false,"output_peak":0.0059566861018538475,"output_energy_fraction":4.7704800067342085e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":264,"both_estimated_voiced_frames":254,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.9657794676806084,"absolute_difference_cents_median":3.219007062503429,"absolute_difference_cents_p95":15.145027414970277,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.00013747273046535536,"log_rms_mae_db":1.3482563018024183,"rms_envelope_nrmse":0.1527052717981997,"envelope_correlation":0.9880785165652172,"rms_gain_error_db":-1.0913024551409993,"reference_mean_power":0.0019308462257719779,"output_mean_power":0.0015018184731123067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":6.283614364116833e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041565087617635905,"output_power":0.0003420446389220397,"reference_active":true,"delta_db":-0.8464591246835756},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151711693867131e-05,"output_power":7.117476803697963e-05,"reference_active":true,"delta_db":-0.5892275090645992},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.519660659304176e-05,"output_power":1.952362064120281e-05,"reference_active":true,"delta_db":-1.107816947858897}],"active_band_count":3,"active_band_mae_db":0.8478345272023574}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012452330782545057,"log_rms_mae_db":1.1553665335129186,"rms_envelope_nrmse":0.14686314386239402,"envelope_correlation":0.989073054127424,"rms_gain_error_db":-1.0913048240513725,"reference_mean_power":0.0019422928952156904,"output_mean_power":0.0015107209057870683},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.257129910319434e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037070296778873537,"output_power":0.00030318200488156606,"reference_active":true,"delta_db":-0.8732264240236652},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.263528086715046e-05,"output_power":3.614098146540942e-05,"reference_active":true,"delta_db":-0.7176918598000988},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8692589849180216e-06,"output_power":1.4714667323427998e-06,"reference_active":true,"delta_db":-1.0391902903624246}],"active_band_count":3,"active_band_mae_db":0.8767028580620627}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.0001332544033329483,"log_rms_mae_db":1.594201089852094,"rms_envelope_nrmse":0.15911224812572047,"envelope_correlation":0.9862331642189026,"rms_gain_error_db":-0.9659883619982134,"reference_mean_power":0.00116782876738011,"output_mean_power":0.0009349326635821746},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":5.939192180731074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041349134282540233,"output_power":0.00032959148022479803,"reference_active":true,"delta_db":-0.9849044438388667},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.995066727507786e-05,"output_power":6.2124767233256e-05,"reference_active":true,"delta_db":-0.5152708713372502},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.967138917853307e-05,"output_power":1.5019376761201306e-05,"reference_active":true,"delta_db":-1.171831188378141}],"active_band_count":3,"active_band_mae_db":0.8906688345180859}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012167319573637772,"log_rms_mae_db":1.2594187711436837,"rms_envelope_nrmse":0.1511827315580739,"envelope_correlation":0.98748232133631,"rms_gain_error_db":-0.9659890887717251,"reference_mean_power":0.0011747523056415098,"output_mean_power":0.0009404753068623283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.189128511363634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037925439414056246,"output_power":0.0003013630590497998,"reference_active":true,"delta_db":-0.9984060572127352},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.625302583005344e-05,"output_power":3.0945039546883064e-05,"reference_active":true,"delta_db":-0.6875321841913409},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7116807182635842e-06,"output_power":1.3356441322866349e-06,"reference_active":true,"delta_db":-1.077319978997033}],"active_band_count":3,"active_band_mae_db":0.9210860734670363}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":1.8440096044800995,"rms_envelope_nrmse":0.2331794169331056,"envelope_correlation":0.9808338715948594,"rms_gain_error_db":-1.7747431073922224,"reference_mean_power":0.0001030241533277173,"output_mean_power":6.846439004699342e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":3.1359809483594144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4569632781045499e-05,"output_power":1.0849690013160825e-05,"reference_active":true,"delta_db":-1.2803127565273362},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.842890886644966e-06,"output_power":2.4882902428006163e-06,"reference_active":true,"delta_db":-1.8875701688526183},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1134901533138973e-06,"output_power":9.301763208801855e-07,"reference_active":true,"delta_db":-0.7812110122634515}],"active_band_count":3,"active_band_mae_db":1.3163646458811353}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":1.5129350906295913,"rms_envelope_nrmse":0.21889951415142908,"envelope_correlation":0.9861666964066759,"rms_gain_error_db":-1.7747712498016186,"reference_mean_power":0.00010363449577271816,"output_mean_power":6.886954499798307e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":2.7762525532534862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3720292251105873e-05,"output_power":1.0101661186402092e-05,"reference_active":true,"delta_db":-1.3297056420603477},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7263230443470644e-06,"output_power":1.1099667369310713e-06,"reference_active":true,"delta_db":-1.9181210367873212},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.698769643844656e-08,"output_power":7.338991514017313e-08,"reference_active":true,"delta_db":-1.2108025896440888}],"active_band_count":3,"active_band_mae_db":1.486209756163919}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":2.012486915303768,"rms_envelope_nrmse":0.29488956516056414,"envelope_correlation":0.9299854568533754,"rms_gain_error_db":-0.36166663146719746,"reference_mean_power":8.51790564150615e-06,"output_mean_power":7.837294413430375e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":4.824507951602351e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4329635827636274e-06,"output_power":2.0687661918791603e-06,"reference_active":true,"delta_db":-0.7042419806414578},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1930715407234241e-06,"output_power":1.062815004538218e-06,"reference_active":true,"delta_db":-0.5020880922538361},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.146506834057095e-07,"output_power":3.744118570969552e-07,"reference_active":true,"delta_db":0.7552091442866892}],"active_band_count":3,"active_band_mae_db":0.6538464057273278}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":1.5997454806937834,"rms_envelope_nrmse":0.2380443863607808,"envelope_correlation":0.9492753354546738,"rms_gain_error_db":-0.3617218152341322,"reference_mean_power":8.568334809688712e-06,"output_mean_power":7.883593936330894e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":4.130413403311083e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5185957053959056e-06,"output_power":2.1293565689738307e-06,"reference_active":true,"delta_db":-0.7291006666887685},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.297939167002703e-07,"output_power":5.774777868419276e-07,"reference_active":true,"delta_db":-0.37663177996439695},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9684693476821212e-08,"output_power":3.655226642799339e-08,"reference_active":true,"delta_db":0.9038174183957555}],"active_band_count":3,"active_band_mae_db":0.6698499550163071}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.523700714111328,"predicted_mos":3.523700714111328,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":224768,"input_sample_rate":44100,"input_seconds":5.0967800453514736,"input_float32_sha256":"49610d809ac243c43efb3a76df0e5cbbc873ad1ef7d4177e5921482b9a2140f0","internal_samples":81549,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.30243864500153},"started_at":"2026-10-13T07:33:37.212926+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.010092780001286883,"native_input_samples":224690,"native_input_sha256":"981654c1812799b81ed6579133eb21e199c6836bbd91ee2245e23bd226f68cf1","output_sha256":"5d077505c7f23b3f68767a79a7a47622d6b536fe31ca81467ed774aacd22ed3f","output_samples":224768,"length_delta_samples":78,"analysis_s":0.11120996400131844,"synthesis_s":33.269561397999496,"total_s":33.380771362000814,"rtf":6.551657915635924,"output_peak":0.2981015145778656,"output_rms":0.04075576099965925,"finished_at":"2026-10-13T07:34:10.626001+00:00","evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"5d077505c7f23b3f68767a79a7a47622d6b536fe31ca81467ed774aacd22ed3f","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":81520,"output_samples":81549,"length_delta_samples":29,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.795366248977118,"output_peak":0.2985348743788363},"common_prefix":{"samples":81520,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996443855841274,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.336394578915316,"spectral_convergence":0.43078145726928113,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.375258656054593,"spectral_convergence":0.3448944391856237,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.961775521955628,"spectral_convergence":0.3926201365893428,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":8.323727696395011e-05,"delta_db":-0.306473469654866,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0009652374031728108,"delta_db":-0.7532646742555836,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0005208336063366818,"delta_db":-0.2512571356934792,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":5.8699972772404134e-05,"delta_db":-1.8969073548180202,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":1.6202289503985063e-05,"delta_db":-1.0660725294723463,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":1.7739003395727195e-05,"delta_db":-2.007516057875545,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0469152036283067,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.6470349578091744,"pitch":null,"output_rms_dbfs":-27.793821631977124,"output_peak":0.2985348743788363,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.224476252308513},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81520,"output_samples":81520,"intelligibility":{"estoi":{"value":0.7819960508301556,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04391983160443005,"rms_dbfs":-27.14678667416795,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.47416011838686,"mean":-6.37155519734421e-07,"zero_energy":false},"output":{"rms":0.04076701549225256,"rms_dbfs":-27.793821631977124,"peak":0.2985348743788363,"peak_dbfs":-10.500098559328592,"crest_factor_db":17.29372307264853,"mean":-3.2067556722864683e-06,"zero_energy":false},"rms_difference_db":-0.6470349578091756},"full_scale":{"reference":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":5.932584311400118e-07,"output_rms_dbfs":-62.26756081055732,"output_zero_energy":false,"output_peak":0.007349611970465498,"output_energy_fraction":5.25464229527839e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":239,"both_estimated_voiced_frames":227,"reference_voiced_output_unvoiced_frames":36,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.8631178707224335,"absolute_difference_cents_median":9.205059366633826,"absolute_difference_cents_p95":24.751872169171072,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.00013747273046535536,"log_rms_mae_db":1.4946255645101936,"rms_envelope_nrmse":0.11920947283721045,"envelope_correlation":0.9892097000792089,"rms_gain_error_db":-0.6470347660583364,"reference_mean_power":0.0019308462257719779,"output_mean_power":0.0016635819936019218},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":6.283614364116833e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041565087617635905,"output_power":0.00036271638882996455,"reference_active":true,"delta_db":-0.5916152130604359},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151711693867131e-05,"output_power":8.188988065764705e-05,"reference_active":true,"delta_db":0.019814268593593434},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.519660659304176e-05,"output_power":2.0700042928128976e-05,"reference_active":true,"delta_db":-0.8537080907456384}],"active_band_count":3,"active_band_mae_db":0.4883791907998892}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012452330782545057,"log_rms_mae_db":1.1082318847174641,"rms_envelope_nrmse":0.0984930417114357,"envelope_correlation":0.993874536185466,"rms_gain_error_db":-0.6470342311241063,"reference_mean_power":0.0019422928952156904,"output_mean_power":0.0016734444419480073},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.257129910319434e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037070296778873537,"output_power":0.0003266151550937363,"reference_active":true,"delta_db":-0.5498973060340615},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.263528086715046e-05,"output_power":4.267969212023514e-05,"reference_active":true,"delta_db":0.0045214956333896415},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8692589849180216e-06,"output_power":1.498849012050007e-06,"reference_active":true,"delta_db":-0.9591159084269074}],"active_band_count":3,"active_band_mae_db":0.5045115700314529}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.0001332544033329483,"log_rms_mae_db":2.4905516591436347,"rms_envelope_nrmse":0.14968618142263249,"envelope_correlation":0.9809841535057272,"rms_gain_error_db":-0.38978328846663757,"reference_mean_power":0.00116782876738011,"output_mean_power":0.0010675810106479968},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":5.939192180731074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041349134282540233,"output_power":0.0003609915797502325,"reference_active":true,"delta_db":-0.5896934929930654},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.995066727507786e-05,"output_power":7.27450095441385e-05,"reference_active":true,"delta_db":0.17011343513665844},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.967138917853307e-05,"output_power":1.785546126412034e-05,"reference_active":true,"delta_db":-0.4206395664800972}],"active_band_count":3,"active_band_mae_db":0.3934821648699403}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012167319573637772,"log_rms_mae_db":1.802209624829452,"rms_envelope_nrmse":0.1272715228610701,"envelope_correlation":0.9850231399597361,"rms_gain_error_db":-0.38978357938958696,"reference_mean_power":0.0011747523056415098,"output_mean_power":0.0010739101525298566},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.189128511363634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037925439414056246,"output_power":0.00033314522297139723,"reference_active":true,"delta_db":-0.5629703111434073},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.625302583005344e-05,"output_power":3.739297200895419e-05,"reference_active":true,"delta_db":0.1344572396758241},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7116807182635842e-06,"output_power":1.4144383887182273e-06,"reference_active":true,"delta_db":-0.8283872361031187}],"active_band_count":3,"active_band_mae_db":0.5086049289741167}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":3.025012305773237,"rms_envelope_nrmse":0.33534225949514,"envelope_correlation":0.9304093070206177,"rms_gain_error_db":-1.9934681034100925,"reference_mean_power":0.0001030241533277173,"output_mean_power":6.510168715508374e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":3.1359809483594144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4569632781045499e-05,"output_power":9.893822163914331e-06,"reference_active":true,"delta_db":-1.6808450588242738},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.842890886644966e-06,"output_power":3.987345115393284e-06,"reference_active":true,"delta_db":0.16025773166425392},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1134901533138973e-06,"output_power":1.100737658542974e-06,"reference_active":true,"delta_db":-0.05002556030291702}],"active_band_count":3,"active_band_mae_db":0.6303761169304816}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":2.5166761868716265,"rms_envelope_nrmse":0.30363728768841963,"envelope_correlation":0.9463292316041958,"rms_gain_error_db":-1.993462860691795,"reference_mean_power":0.00010363449577271816,"output_mean_power":6.548744589432127e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":2.7762525532534862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3720292251105873e-05,"output_power":9.668667583739216e-06,"reference_active":true,"delta_db":-1.5199673311395667},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7263230443470644e-06,"output_power":1.6854255037292787e-06,"reference_active":true,"delta_db":-0.10412506539754068},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.698769643844656e-08,"output_power":1.129691311926915e-07,"reference_active":true,"delta_db":0.6624314426151291}],"active_band_count":3,"active_band_mae_db":0.7621746130507455}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":4.695374437310372,"rms_envelope_nrmse":0.5045322352001111,"envelope_correlation":0.857175715399952,"rms_gain_error_db":0.9175171622955308,"reference_mean_power":8.51790564150615e-06,"output_mean_power":1.0521666717107751e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":4.824507951602351e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4329635827636274e-06,"output_power":2.98799780996193e-06,"reference_active":true,"delta_db":0.8924466648363251},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1930715407234241e-06,"output_power":1.0678276944876562e-06,"reference_active":true,"delta_db":-0.4816530601191452},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.146506834057095e-07,"output_power":2.078770415738569e-07,"reference_active":true,"delta_db":-1.8002215186795802}],"active_band_count":3,"active_band_mae_db":1.0581070812116835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":3.9610939095876616,"rms_envelope_nrmse":0.4509763562952996,"envelope_correlation":0.8828561893251499,"rms_gain_error_db":0.9175492469904973,"reference_mean_power":8.568334809688712e-06,"output_mean_power":1.0584037088001214e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":4.130413403311083e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5185957053959056e-06,"output_power":2.9461180541715785e-06,"reference_active":true,"delta_db":0.6809168702639219},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.297939167002703e-07,"output_power":5.771811097196563e-07,"reference_active":true,"delta_db":-0.3788635254665263},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9684693476821212e-08,"output_power":1.8294896046281496e-08,"reference_active":true,"delta_db":-2.1020262265645933}],"active_band_count":3,"active_band_mae_db":1.053935540765014}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"golf","status":"measured","mos":{"status":"measured","value":2.8525333404541016,"predicted_mos":2.8525333404541016,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":122160,"input_sample_rate":24000,"input_seconds":5.09,"input_float32_sha256":"5c638e9ceac6d41dc9e77c5aba55ea3feab7ebe3f73b7b78e6966e6c9e301027","internal_samples":81440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.551513974000045},"started_at":"2026-10-13T05:03:11.800025+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0031389929999932065,"native_input_samples":122280,"native_input_sha256":"8186c766503a3b0cbcd20e838c5d46554d5d7f2c868fd8b5a1d17a5f324b8131","output_sha256":"b1d9e17781b010dfb1fe011fbda0c7a4937c95216d41731a2265aea66cc29c6c","output_samples":122160,"length_delta_samples":-120,"analysis_s":0.4636447680004494,"synthesis_s":0.11253774899978453,"total_s":0.5761825170002339,"rtf":0.11308783454371618,"output_peak":0.23554491996765137,"output_rms":0.03756863620861521,"finished_at":"2026-10-13T05:03:12.386806+00:00","analysis_stages":{"external_f0_s":0.04271096299999044},"evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"b1d9e17781b010dfb1fe011fbda0c7a4937c95216d41731a2265aea66cc29c6c","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":81520,"output_samples":81440,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-28.500236372664496,"output_peak":0.23567638205047706},"common_prefix":{"samples":81440,"sample_rate":16000,"reference_coverage":0.9990186457311089,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81440,"output_samples":81440,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.800534147905065,"spectral_convergence":0.4678840542724059,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.031339651292889,"spectral_convergence":0.28279776695335074,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.13514067936212,"spectral_convergence":0.302932257557106,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.94111421072453e-05,"output_power":9.754944270720959e-05,"delta_db":0.378331502008002,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011490740368505987,"output_power":0.0008694271814391507,"delta_db":-1.2111479825932183,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000552499791387475,"output_power":0.0003506216636039946,"delta_db":-1.9749337235461042,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.093995377621044e-05,"output_power":6.690121460430532e-05,"delta_db":-1.333207265834714,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.073048607547547e-05,"output_power":1.5903901025007782e-05,"delta_db":-1.151058209685993,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.818967497207309e-05,"output_power":1.2057254548773566e-05,"delta_db":-3.688416387800149,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.6228491785780301,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.3577131746717659,"pitch":null,"output_rms_dbfs":-28.500236372664496,"output_peak":0.23567638205047706,"reference_rms_dbfs":-27.14252319799273,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.322338159520025},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81440,"output_samples":81440,"intelligibility":{"estoi":{"value":0.9001643245624069,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.043941394982182094,"rms_dbfs":-27.14252319799273,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.46989664221164,"mean":-6.617637889090128e-07,"zero_energy":false},"output":{"rms":0.03758271766107286,"rms_dbfs":-28.500236372664496,"peak":0.23567638205047706,"peak_dbfs":-12.553678749096607,"crest_factor_db":15.94655762356789,"mean":0.0002352328015098138,"zero_energy":false},"rms_difference_db":-1.3577131746717654},"full_scale":{"reference":{"sample_count":81440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81440,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":4.889729157205555e-07,"output_rms_dbfs":-63.107151958433896,"output_zero_energy":false,"output_peak":0.008210297157714617,"output_energy_fraction":5.100960053131935e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":270,"both_estimated_voiced_frames":253,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":17,"paired_voiced_fraction":0.9619771863117871,"absolute_difference_cents_median":4.119062276302349,"absolute_difference_cents_p95":18.36225469995873,"difference_over_600_cents_frames":2,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81440,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1017,"rms_floor":0.00013747273046535536,"log_rms_mae_db":2.243787293378359,"rms_envelope_nrmse":0.18595401336253642,"envelope_correlation":0.982299774504817,"rms_gain_error_db":-1.3577129649134976,"reference_mean_power":0.0019327445415587181,"output_mean_power":0.0014138494229551562},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1017,"frequency_resolution_hz":0.19665683382497542,"power_floor":6.293987750957364e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041627844490517923,"output_power":0.00031859106702696185,"reference_active":true,"delta_db":-1.161503290398102},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.147222930757328e-05,"output_power":6.0653697946287885e-05,"reference_active":true,"delta_db":-1.281523158232746},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.5185454901328168e-05,"output_power":1.6145483396944715e-05,"reference_active":true,"delta_db":-1.9309874716366937}],"active_band_count":3,"active_band_mae_db":1.4580046400891806}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1011,"rms_floor":0.00012452330782545057,"log_rms_mae_db":2.0363994886370302,"rms_envelope_nrmse":0.1664756889926199,"envelope_correlation":0.990068438901743,"rms_gain_error_db":-1.3577155509741414,"reference_mean_power":0.0019442138376853126,"output_mean_power":0.0014222386438478526},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1011,"frequency_resolution_hz":0.19782393669634027,"power_floor":5.266402807782247e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003712536580412157,"output_power":0.00028381116391679325,"reference_active":true,"delta_db":-1.1664126679812776},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.254309700707861e-05,"output_power":3.174034745201417e-05,"reference_active":true,"delta_db":-1.272174254891894},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8663581667586551e-06,"output_power":1.345963329514355e-06,"reference_active":true,"delta_db":-1.4196176366982807}],"active_band_count":3,"active_band_mae_db":1.2860681865238173}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1017,"rms_floor":0.0001332544033329483,"log_rms_mae_db":2.111525143455035,"rms_envelope_nrmse":0.21569875867479346,"envelope_correlation":0.9807007505919065,"rms_gain_error_db":-1.61208925037502,"reference_mean_power":0.0011689769900546888,"output_mean_power":0.00080648638044364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1017,"frequency_resolution_hz":0.19665683382497542,"power_floor":5.949048663259079e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004140305171753987,"output_power":0.0002918540664669089,"reference_active":true,"delta_db":-1.518666042743989},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.004430639373476e-05,"output_power":5.074354746825706e-05,"reference_active":true,"delta_db":-1.3999201372787335},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.966261634880877e-05,"output_power":1.2618276759765276e-05,"reference_active":true,"delta_db":-1.926412568101806}],"active_band_count":3,"active_band_mae_db":1.614999582708176}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1011,"rms_floor":0.00012167319573637772,"log_rms_mae_db":1.853434705530126,"rms_envelope_nrmse":0.19461990179229038,"envelope_correlation":0.9887829337400964,"rms_gain_error_db":-1.6120886373999355,"reference_mean_power":0.0011759141928426105,"output_mean_power":0.0008112725254819001},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1011,"frequency_resolution_hz":0.19782393669634027,"power_floor":5.197805749980336e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0003797483286653897,"output_power":0.00026732016507784336,"reference_active":true,"delta_db":-1.5246415049686735},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.624937452660341e-05,"output_power":2.60820084522252e-05,"reference_active":true,"delta_db":-1.4295948602145896},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7068382094557657e-06,"output_power":1.066544258335432e-06,"reference_active":true,"delta_db":-2.0421347440058515}],"active_band_count":3,"active_band_mae_db":1.6654570363963714}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1017,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":2.569825367778723,"rms_envelope_nrmse":0.22533837450804786,"envelope_correlation":0.973109696919344,"rms_gain_error_db":-1.3383566457726777,"reference_mean_power":0.00010312531708171344,"output_mean_power":7.577564335305024e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1017,"frequency_resolution_hz":0.19665683382497542,"power_floor":3.136580265626093e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4560046439049703e-05,"output_power":1.078873636900712e-05,"reference_active":true,"delta_db":-1.3019217924448159},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.838963945155652e-06,"output_power":2.700691507137602e-06,"reference_active":true,"delta_db":-1.5273905461434376},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1130192673390267e-06,"output_power":9.555899466428831e-07,"reference_active":true,"delta_db":-0.6623111034433831}],"active_band_count":3,"active_band_mae_db":1.163874480677212}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1011,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":2.158452428806913,"rms_envelope_nrmse":0.18197910241478096,"envelope_correlation":0.9879122306781658,"rms_gain_error_db":-1.3383475476369977,"reference_mean_power":0.00010373689999668918,"output_mean_power":7.622518918525908e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1011,"frequency_resolution_hz":0.19782393669634027,"power_floor":2.777126469079474e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.370716420685013e-05,"output_power":1.0166144286109057e-05,"reference_active":true,"delta_db":-1.2979134622917186},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.72371618004043e-06,"output_power":1.2575174593192384e-06,"reference_active":true,"delta_db":-1.3695173475041942},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.690858765158156e-08,"output_power":8.479099779140829e-08,"reference_active":true,"delta_db":-0.5801251827310204}],"active_band_count":3,"active_band_mae_db":1.0825186641756444}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1017,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":3.25049591932543,"rms_envelope_nrmse":0.2986412718455733,"envelope_correlation":0.9412904966816537,"rms_gain_error_db":-1.1692969727070235,"reference_mean_power":8.52627721668211e-06,"output_mean_power":6.513729983154646e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1017,"frequency_resolution_hz":0.19665683382497542,"power_floor":4.8239712414862815e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.43284060455123e-06,"output_power":2.0011507729270815e-06,"reference_active":true,"delta_db":-0.8483384468351203},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1940054389612163e-06,"output_power":9.470601466019122e-07,"reference_active":true,"delta_db":-1.0062874372922637},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.1506279867490797e-07,"output_power":2.3596613798096334e-07,"reference_active":true,"delta_db":-1.2554744192029597}],"active_band_count":3,"active_band_mae_db":1.0367001011101145}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81440,"uncovered_tail_samples":0,"metrics":{"frame_count":1011,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":3.0112645618509153,"rms_envelope_nrmse":0.25047679032827613,"envelope_correlation":0.9612074904697451,"rms_gain_error_db":-1.1692764048846314,"reference_mean_power":8.576804701963417e-06,"output_mean_power":6.552361963969295e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1011,"frequency_resolution_hz":0.19782393669634027,"power_floor":4.1295871274683695e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.518537095230936e-06,"output_power":2.1007327105664308e-06,"reference_active":true,"delta_db":-0.7877755400919413},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.304354989310523e-07,"output_power":5.139117079287795e-07,"reference_active":true,"delta_db":-0.8875214778967799},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9669134411341264e-08,"output_power":2.027184239887674e-08,"reference_active":true,"delta_db":-1.6541165490255687}],"active_band_count":3,"active_band_mae_db":1.1098045223380966}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.657539129257202,"predicted_mos":3.657539129257202,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":122400,"input_sample_rate":24000,"input_seconds":5.1,"input_float32_sha256":"5e0b7bbb21735fe11dcfe39770afb066ff4d0eb7f46459f0487b094bd6e8000b","internal_samples":81600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.023305218001042},"started_at":"2026-10-13T07:47:54.812708+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.003634846001659753,"native_input_samples":122280,"native_input_sha256":"8186c766503a3b0cbcd20e838c5d46554d5d7f2c868fd8b5a1d17a5f324b8131","output_sha256":"1d0e172bd509ac2e74852ff900c6e6a6457cc52d3ba3b2cfad5dbe892bd84685","output_samples":122400,"length_delta_samples":120,"analysis_s":2.1163425049999205,"synthesis_s":16.820987348000926,"total_s":18.937329853000847,"rtf":3.716845898528135,"output_peak":0.29227396845817566,"output_rms":0.04153659682310788,"finished_at":"2026-10-13T07:48:13.770227+00:00","analysis_stages":{"external_f0_s":1.3900591519995942},"evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"1d0e172bd509ac2e74852ff900c6e6a6457cc52d3ba3b2cfad5dbe892bd84685","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":81520,"output_samples":81600,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.63218149924443,"output_peak":0.2925579873068504},"common_prefix":{"samples":81520,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9990196078431373,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.124642124448552,"spectral_convergence":0.425534790826923,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.744327837450368,"spectral_convergence":0.2805706309374684,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.657742022608795,"spectral_convergence":0.284873323403877,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":8.138078475601346e-05,"delta_db":-0.4044333772131663,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0010396436580282307,"delta_db":-0.4307610134633123,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0004807118616549663,"delta_db":-0.5993987395001221,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":8.256525105925332e-05,"delta_db":-0.4153132947144064,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":2.175298092347576e-05,"delta_db":0.2133513812177662,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.0609412261075076e-05,"delta_db":-1.3561521589012446,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5699016608350029,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.48113511190586755,"pitch":null,"output_rms_dbfs":-27.627921786073813,"output_peak":0.2925579873068504,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.8422373281692375},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81520,"output_samples":81520,"intelligibility":{"estoi":{"value":0.9051423708789731,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04391983160443005,"rms_dbfs":-27.14678667416795,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.47416011838686,"mean":-6.37155519734421e-07,"zero_energy":false},"output":{"rms":0.04155314607443119,"rms_dbfs":-27.627921786073813,"peak":0.2925579873068504,"peak_dbfs":-10.675760808974166,"crest_factor_db":16.952160977099645,"mean":-3.5455917186043625e-06,"zero_energy":false},"rms_difference_db":-0.4811351119058642},"full_scale":{"reference":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":1.5063171252447795e-05,"output_rms_dbfs":-48.22083586405472,"output_zero_energy":false,"output_peak":0.09216312234129735,"output_energy_fraction":0.0012841793062484893,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":252,"both_estimated_voiced_frames":240,"reference_voiced_output_unvoiced_frames":23,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.9125475285171103,"absolute_difference_cents_median":5.7760487503932465,"absolute_difference_cents_p95":26.380322472014235,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.00013747273046535536,"log_rms_mae_db":2.3466466148071765,"rms_envelope_nrmse":0.1791793338857207,"envelope_correlation":0.965576959699028,"rms_gain_error_db":-0.48113483400642415,"reference_mean_power":0.0019308462257719779,"output_mean_power":0.0017283599898092628},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":6.283614364116833e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041565087617635905,"output_power":0.0003791657862966951,"reference_active":true,"delta_db":-0.3989955807569504},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151711693867131e-05,"output_power":8.00803524302062e-05,"reference_active":true,"delta_db":-0.07722835547972161},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.519660659304176e-05,"output_power":2.2152452089306252e-05,"reference_active":true,"delta_db":-0.5592024923007328}],"active_band_count":3,"active_band_mae_db":0.34514214284580164}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012452330782545057,"log_rms_mae_db":1.5901727485396981,"rms_envelope_nrmse":0.11691768730248511,"envelope_correlation":0.9855579962230406,"rms_gain_error_db":-0.4811342384694407,"reference_mean_power":0.0019422928952156904,"output_mean_power":0.0017386064949702436},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.257129910319434e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037070296778873537,"output_power":0.00033979507319495975,"reference_active":true,"delta_db":-0.37808985341345547},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.263528086715046e-05,"output_power":4.1710598146540625e-05,"reference_active":true,"delta_db":-0.09522710471552369},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8692589849180216e-06,"output_power":1.7999869738492132e-06,"reference_active":true,"delta_db":-0.1640011462445094}],"active_band_count":3,"active_band_mae_db":0.21243936812449618}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.0001332544033329483,"log_rms_mae_db":2.7963718467937935,"rms_envelope_nrmse":0.2108428319757878,"envelope_correlation":0.9617827326510182,"rms_gain_error_db":-0.5763766952613706,"reference_mean_power":0.00116782876738011,"output_mean_power":0.001022684089890516},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":5.939192180731074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041349134282540233,"output_power":0.0003719162306100101,"reference_active":true,"delta_db":-0.4602128965795774},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.995066727507786e-05,"output_power":6.67256078864658e-05,"reference_active":true,"delta_db":-0.20499322674317655},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.967138917853307e-05,"output_power":1.9139110206028095e-05,"reference_active":true,"delta_db":-0.11913287368244757}],"active_band_count":3,"active_band_mae_db":0.26144633233506714}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012167319573637772,"log_rms_mae_db":2.1519941360812957,"rms_envelope_nrmse":0.14440916753998878,"envelope_correlation":0.9820999044395382,"rms_gain_error_db":-0.5763757741267943,"reference_mean_power":0.0011747523056415098,"output_mean_power":0.001028747347997718},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.189128511363634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037925439414056246,"output_power":0.0003428227582291293,"reference_active":true,"delta_db":-0.43860976565585774},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.625302583005344e-05,"output_power":3.4610712849858964e-05,"reference_active":true,"delta_db":-0.2013371625338219},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7116807182635842e-06,"output_power":1.744834280892984e-06,"reference_active":true,"delta_db":0.08331426815799026}],"active_band_count":3,"active_band_mae_db":0.24108706544922331}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":2.443872959705526,"rms_envelope_nrmse":0.22131676969377534,"envelope_correlation":0.9613842569830607,"rms_gain_error_db":-0.29705064338654225,"reference_mean_power":0.0001030241533277173,"output_mean_power":9.6213051671886e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":3.1359809483594144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4569632781045499e-05,"output_power":1.3779249731655333e-05,"reference_active":true,"delta_db":-0.24223034520215658},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.842890886644966e-06,"output_power":3.698875713509861e-06,"reference_active":true,"delta_db":-0.16588314707682034},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1134901533138973e-06,"output_power":1.0701835578000255e-06,"reference_active":true,"delta_db":-0.17228106655049039}],"active_band_count":3,"active_band_mae_db":0.19346485294315577}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":1.5263526215912655,"rms_envelope_nrmse":0.11672295899036791,"envelope_correlation":0.9890959805235221,"rms_gain_error_db":-0.29704703145595324,"reference_mean_power":0.00010363449577271816,"output_mean_power":9.678312383415373e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":2.7762525532534862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3720292251105873e-05,"output_power":1.3012102295585624e-05,"reference_active":true,"delta_db":-0.23015893362040016},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7263230443470644e-06,"output_power":1.639765731366466e-06,"reference_active":true,"delta_db":-0.2234026181472358},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.698769643844656e-08,"output_power":8.237685716238651e-08,"reference_active":true,"delta_db":-0.7091142573938479}],"active_band_count":3,"active_band_mae_db":0.387558603053828}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":2.23429059404581,"rms_envelope_nrmse":0.27668080481529345,"envelope_correlation":0.9371261344469956,"rms_gain_error_db":-0.6484390387789092,"reference_mean_power":8.51790564150615e-06,"output_mean_power":7.336499983805086e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":4.824507951602351e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4329635827636274e-06,"output_power":2.125776766248377e-06,"reference_active":true,"delta_db":-0.5861795223986541},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1930715407234241e-06,"output_power":1.0724434113432444e-06,"reference_active":true,"delta_db":-0.46292100825258586},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.146506834057095e-07,"output_power":2.421679973350215e-07,"reference_active":true,"delta_db":-1.137119290618596}],"active_band_count":3,"active_band_mae_db":0.7287399404232787}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":1.4632372819107862,"rms_envelope_nrmse":0.16868948731159414,"envelope_correlation":0.9770043690739874,"rms_gain_error_db":-0.6484398698296261,"reference_mean_power":8.568334809688712e-06,"output_mean_power":7.379933379073004e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":4.130413403311083e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5185957053959056e-06,"output_power":2.2075160608472564e-06,"reference_active":true,"delta_db":-0.5725458650040277},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.297939167002703e-07,"output_power":5.586766697465021e-07,"reference_active":true,"delta_db":-0.5203792572068823},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9684693476821212e-08,"output_power":2.0277862045738317e-08,"reference_active":true,"delta_db":-1.655104046361264}],"active_band_count":3,"active_band_mae_db":0.9160097228573912}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.2322330474853516,"predicted_mos":3.2322330474853516,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":122520,"input_sample_rate":24000,"input_seconds":5.105,"input_float32_sha256":"7e4c3507274f19f0c5a66a2ceb343e611d2daa53ce8bc22f821190fafad7aed0","internal_samples":81680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0540819939997164},"started_at":"2026-10-13T05:11:40.162075+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0022572779998881742,"native_input_samples":122280,"native_input_sha256":"8186c766503a3b0cbcd20e838c5d46554d5d7f2c868fd8b5a1d17a5f324b8131","output_sha256":"ee3dc73a0a8871e42aece14d7ab3c6af3c06b8e13bb80b179d27b75ab3cee032","output_samples":122520,"length_delta_samples":240,"analysis_s":0.6211625249998178,"synthesis_s":0.12036717899900395,"total_s":0.7415297039988218,"rtf":0.14554066810575503,"output_peak":0.2839542031288147,"output_rms":0.04312768302602999,"finished_at":"2026-10-13T05:11:40.912405+00:00","analysis_stages":{"external_f0_s":0.06033065900010115},"evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"ee3dc73a0a8871e42aece14d7ab3c6af3c06b8e13bb80b179d27b75ab3cee032","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":81520,"output_samples":81680,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.30479862886751,"output_peak":0.28363154792127065},"common_prefix":{"samples":81520,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9980411361410382,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.587085037332105,"spectral_convergence":0.1521881308915762,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.553661916234417,"spectral_convergence":0.1502364626391786,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.09343654287039,"spectral_convergence":0.1763912575386148,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":8.992652360696981e-05,"delta_db":0.029225935504156465,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0011283865491913197,"delta_db":-0.07502709445863441,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0005177966563151371,"delta_db":-0.2766547086790077,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":8.715706566679815e-05,"delta_db":-0.18026034937887056,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":1.9463542707551275e-05,"delta_db":-0.26961748494290994,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.0951129616853475e-05,"delta_db":-1.2847337896075897,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3525865604285281,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.14949646078170284,"pitch":null,"output_rms_dbfs":-27.29628313494965,"output_peak":0.28363154792127065,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.744727832145638},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81520,"output_samples":81520,"intelligibility":{"estoi":{"value":0.9653542824261169,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04391983160443005,"rms_dbfs":-27.14678667416795,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.47416011838686,"mean":-6.37155519734421e-07,"zero_energy":false},"output":{"rms":0.043170377194375206,"rms_dbfs":-27.29628313494965,"peak":0.28363154792127065,"peak_dbfs":-10.944909297274545,"crest_factor_db":16.351373837675105,"mean":-2.133458227949129e-05,"zero_energy":false},"rms_difference_db":-0.14949646078170176},"full_scale":{"reference":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":2.2897459535591235e-06,"output_rms_dbfs":-56.402126997945146,"output_zero_energy":false,"output_peak":0.03950621150723297,"output_energy_fraction":0.00018085590315701288,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":248,"both_estimated_voiced_frames":241,"reference_voiced_output_unvoiced_frames":22,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9163498098859315,"absolute_difference_cents_median":0.6342717065886745,"absolute_difference_cents_p95":4.726241824707578,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.00013747273046535536,"log_rms_mae_db":0.782601349350525,"rms_envelope_nrmse":0.0627682435831343,"envelope_correlation":0.9957785083928017,"rms_gain_error_db":-0.14949631589225867,"reference_mean_power":0.0019308462257719779,"output_mean_power":0.0018655120385744593},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":6.283614364116833e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041565087617635905,"output_power":0.0004039912333816114,"reference_active":true,"delta_db":-0.12356759301926439},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151711693867131e-05,"output_power":7.18286781962676e-05,"reference_active":true,"delta_db":-0.549509367900553},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.519660659304176e-05,"output_power":1.8139838819223113e-05,"reference_active":true,"delta_db":-1.4270863134446405}],"active_band_count":3,"active_band_mae_db":0.7000544247881527}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012452330782545057,"log_rms_mae_db":0.4712395831057923,"rms_envelope_nrmse":0.0368049842782917,"envelope_correlation":0.9986329528757942,"rms_gain_error_db":-0.14949577861694055,"reference_mean_power":0.0019422928952156904,"output_mean_power":0.0018765716183685667},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.257129910319434e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037070296778873537,"output_power":0.0003614585971829924,"reference_active":true,"delta_db":-0.10967504167489478},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.263528086715046e-05,"output_power":3.743259893836233e-05,"reference_active":true,"delta_db":-0.5651914717126595},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8692589849180216e-06,"output_power":1.2243394938874774e-06,"reference_active":true,"delta_db":-1.837676179489271}],"active_band_count":3,"active_band_mae_db":0.8375142309589417}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.0001332544033329483,"log_rms_mae_db":1.0313275244620652,"rms_envelope_nrmse":0.07198464348015096,"envelope_correlation":0.995545502705191,"rms_gain_error_db":-0.14902086788266858,"reference_mean_power":0.00116782876738011,"output_mean_power":0.001128436388495777},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":5.939192180731074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041349134282540233,"output_power":0.0003974038141816015,"reference_active":true,"delta_db":-0.1723439021581371},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.995066727507786e-05,"output_power":6.203408213763056e-05,"reference_active":true,"delta_db":-0.5216150093224304},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.967138917853307e-05,"output_power":1.4687176778557342e-05,"reference_active":true,"delta_db":-1.2689670836777969}],"active_band_count":3,"active_band_mae_db":0.6543086650527882}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012167319573637772,"log_rms_mae_db":0.5663678849755587,"rms_envelope_nrmse":0.039956733675975686,"envelope_correlation":0.9986554489893666,"rms_gain_error_db":-0.14902060204916956,"reference_mean_power":0.0011747523056415098,"output_mean_power":0.0011351264563140332},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.189128511363634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037925439414056246,"output_power":0.000364337818807047,"reference_active":true,"delta_db":-0.17426367342901292},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.625302583005344e-05,"output_power":3.1501045025818464e-05,"reference_active":true,"delta_db":-0.6101929901068239},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7116807182635842e-06,"output_power":1.0935815056851407e-06,"reference_active":true,"delta_db":-1.9457160148881678}],"active_band_count":3,"active_band_mae_db":0.9100575594746682}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":1.1339506732288678,"rms_envelope_nrmse":0.08875603711983597,"envelope_correlation":0.9941198928474145,"rms_gain_error_db":-0.19123942224195,"reference_mean_power":0.0001030241533277173,"output_mean_power":9.858596954522112e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":3.1359809483594144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4569632781045499e-05,"output_power":1.3961811657871465e-05,"reference_active":true,"delta_db":-0.18506830611404768},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.842890886644966e-06,"output_power":3.4746264602983627e-06,"reference_active":true,"delta_db":-0.4374993086343087},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1134901533138973e-06,"output_power":8.03321671694249e-07,"reference_active":true,"delta_db":-1.4179689753862492}],"active_band_count":3,"active_band_mae_db":0.6801788633782019}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":0.670516106144105,"rms_envelope_nrmse":0.04660649488986532,"envelope_correlation":0.9985514635518109,"rms_gain_error_db":-0.1912343299405592,"reference_mean_power":0.00010363449577271816,"output_mean_power":9.917013529227197e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":2.7762525532534862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3720292251105873e-05,"output_power":1.3126398517624363e-05,"reference_active":true,"delta_db":-0.19217776915832702},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7263230443470644e-06,"output_power":1.5176675681752752e-06,"reference_active":true,"delta_db":-0.5594541432188977},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.698769643844656e-08,"output_power":6.368062229935568e-08,"reference_active":true,"delta_db":-1.8270934573512325}],"active_band_count":3,"active_band_mae_db":0.859575123242819}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":1.1298036366142716,"rms_envelope_nrmse":0.145070028837673,"envelope_correlation":0.9835787020488092,"rms_gain_error_db":-0.36611407716435107,"reference_mean_power":8.51790564150615e-06,"output_mean_power":7.829272644438953e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":4.824507951602351e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4329635827636274e-06,"output_power":2.228272053212666e-06,"reference_active":true,"delta_db":-0.38167394915942815},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1930715407234241e-06,"output_power":1.0116689047972908e-06,"reference_active":true,"delta_db":-0.7162808477132707},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.146506834057095e-07,"output_power":1.7823585989734543e-07,"reference_active":true,"delta_db":-2.46833593629788}],"active_band_count":3,"active_band_mae_db":1.1887635777235264}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":0.7852218845382772,"rms_envelope_nrmse":0.0791762296478488,"envelope_correlation":0.9956917757204414,"rms_gain_error_db":-0.36610494233711405,"reference_mean_power":8.568334809688712e-06,"output_mean_power":7.87564141464294e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":4.130413403311083e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5185957053959056e-06,"output_power":2.2342790408447897e-06,"reference_active":true,"delta_db":-0.5202104694159277},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.297939167002703e-07,"output_power":5.313012118933678e-07,"reference_active":true,"delta_db":-0.738576548546943},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9684693476821212e-08,"output_power":1.4003318134122941e-08,"reference_active":true,"delta_db":-3.263016133394653}],"active_band_count":3,"active_band_mae_db":1.5072677171191746}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-284449-0020","method":"magphase","status":"measured","mos":{"status":"measured","value":3.1389031410217285,"predicted_mos":3.1389031410217285,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":245866,"input_sample_rate":48000,"input_seconds":5.122208333333333,"input_float32_sha256":"0f002d456cccba6a8d8e9fa2ee6ed32d312fc773b5266336bfeb46bf63413790","internal_samples":81956,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1443178139998054},"started_at":"2026-10-13T05:19:12.192494+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.008235602999775438,"native_input_samples":244560,"native_input_sha256":"9dbd5f86ab8368fecab621ab1d1856781dd4bd230bc4fe627754935ce64f3e04","output_sha256":"512341271f19fe0e5c3db643074a8a497d7865c10d5db74c3439ac8e766c6384","output_samples":245866,"length_delta_samples":1306,"analysis_s":3.256985486001213,"synthesis_s":0.9314204039983451,"total_s":4.188405889999558,"rtf":0.8220619999999133,"output_peak":0.2973524993089798,"output_rms":0.04158372973066789,"finished_at":"2026-10-13T05:19:16.400958+00:00","evaluation_input":{"reference":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"512341271f19fe0e5c3db643074a8a497d7865c10d5db74c3439ac8e766c6384","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":81520,"output_samples":81956,"length_delta_samples":436,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-27.622123481888327,"output_peak":0.2966970953225752},"common_prefix":{"samples":81520,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9946800722338816,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":81520,"output_samples":81520,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.212254183757937,"spectral_convergence":0.4566149660252748,"spectral_convergence_status":"ok","amplitude_floor":1.0222158521079025e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.8551369760271905,"spectral_convergence":0.29228130774566385,"spectral_convergence_status":"ok","amplitude_floor":7.564059658380168e-07},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.545135960982993,"spectral_convergence":0.28530180584519066,"spectral_convergence_status":"ok","amplitude_floor":6.35834815641291e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":8.932339290337003e-05,"output_power":8.350900371131563e-05,"delta_db":-0.2923190872064989,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0011480494900330098,"output_power":0.0010172201172314564,"delta_db":-0.5254566951074789,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0005518546578574125,"output_power":0.0005039721827623459,"delta_db":-0.3941814671005628,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":9.085077476738271e-05,"output_power":9.09674701119425e-05,"delta_db":0.005574814694618107,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.0710168905088768e-05,"output_power":2.0307642456932083e-05,"delta_db":-0.08524132398247446,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":2.8163123695227053e-05,"output_power":2.224047609787311e-05,"delta_db":-1.0253674276867386,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3880234692963953,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.4521736640654659,"pitch":null,"output_rms_dbfs":-27.598960338233415,"output_peak":0.2966970953225752,"reference_rms_dbfs":-27.14678667416795,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.870842373589373},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":81520,"output_samples":81520,"intelligibility":{"estoi":{"value":0.843694843470413,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.04391983160443005,"rms_dbfs":-27.14678667416795,"peak":0.29266357421875,"peak_dbfs":-10.672626555781086,"crest_factor_db":16.47416011838686,"mean":-6.37155519734421e-07,"zero_energy":false},"output":{"rms":0.041691928383943415,"rms_dbfs":-27.598960338233415,"peak":0.2966970953225752,"peak_dbfs":-10.553734108482999,"crest_factor_db":17.045226229750416,"mean":0.000580118025129126,"zero_energy":false},"rms_difference_db":-0.4521736640654659},"full_scale":{"reference":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":81520,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":509,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":1.857199690421112e-06,"threshold_dbfs":-57.31141397398936,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":75,"quiet_samples":12000,"reference_mean_power":4.2324885725975035e-07,"output_mean_power":3.4040972559493914e-06,"output_rms_dbfs":-54.67998040503203,"output_zero_energy":false,"output_peak":0.03798120859292249,"output_energy_fraction":0.00028828036839130246,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":502,"reference_estimated_voiced_frames":263,"output_estimated_voiced_frames":260,"both_estimated_voiced_frames":234,"reference_voiced_output_unvoiced_frames":29,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8897338403041825,"absolute_difference_cents_median":4.980716276935147,"absolute_difference_cents_p95":22.560152767706786,"difference_over_600_cents_frames":1,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":81520,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.00013747273046535536,"log_rms_mae_db":2.7327063301269217,"rms_envelope_nrmse":0.1851297456076011,"envelope_correlation":0.9621710289357868,"rms_gain_error_db":-0.45217342938499594,"reference_mean_power":0.0019308462257719779,"output_mean_power":0.0017399242642325001},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":6.283614364116833e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041565087617635905,"output_power":0.0003838922888607392,"reference_active":true,"delta_db":-0.3451931166454416},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.151711693867131e-05,"output_power":7.0499813279125e-05,"reference_active":true,"delta_db":-0.6306084458395151},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.519660659304176e-05,"output_power":2.1301001829351007e-05,"reference_active":true,"delta_db":-0.7294202551022035}],"active_band_count":3,"active_band_mae_db":0.5684072725290534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012452330782545057,"log_rms_mae_db":1.834464839185428,"rms_envelope_nrmse":0.12930044000344873,"envelope_correlation":0.9806486806122386,"rms_gain_error_db":-0.4521734823682003,"reference_mean_power":0.0019422928952156904,"output_mean_power":0.0017502390663204628},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.257129910319434e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037070296778873537,"output_power":0.00034223466097134847,"reference_active":true,"delta_db":-0.34702070862452106},{"low_hz":8.0,"high_hz":20.0,"reference_power":4.263528086715046e-05,"output_power":3.6561097170101444e-05,"reference_active":true,"delta_db":-0.6674990822026026},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.8692589849180216e-06,"output_power":1.6367979316933337e-06,"reference_active":true,"delta_db":-0.5767440925462686}],"active_band_count":3,"active_band_mae_db":0.5304212944577974}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":0.0001332544033329483,"log_rms_mae_db":3.5749972703652833,"rms_envelope_nrmse":0.20528396264254178,"envelope_correlation":0.9628044568799443,"rms_gain_error_db":-0.43621976854604283,"reference_mean_power":0.00116782876738011,"output_mean_power":0.0010562268230309693},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":5.939192180731074e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00041349134282540233,"output_power":0.00037887906463255163,"reference_active":true,"delta_db":-0.3796581272609773},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.995066727507786e-05,"output_power":5.9012819466974194e-05,"reference_active":true,"delta_db":-0.7384549717314087},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.967138917853307e-05,"output_power":1.748662287958954e-05,"reference_active":true,"delta_db":-0.5112908649970478}],"active_band_count":3,"active_band_mae_db":0.5431346546631446}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":0.00012167319573637772,"log_rms_mae_db":2.901959859676859,"rms_envelope_nrmse":0.1513726209708292,"envelope_correlation":0.9783567537416258,"rms_gain_error_db":-0.4362194226872249,"reference_mean_power":0.0011747523056415098,"output_mean_power":0.0010624888075274474},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":5.189128511363634e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00037925439414056246,"output_power":0.00034754662137199233,"reference_active":true,"delta_db":-0.3791755040688361},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.625302583005344e-05,"output_power":3.026577262885915e-05,"reference_active":true,"delta_db":-0.7839249534424602},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7116807182635842e-06,"output_power":1.4532055100010486e-06,"reference_active":true,"delta_db":-0.7109572258283575}],"active_band_count":3,"active_band_mae_db":0.6246858944465513}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":4.762512740172365e-05,"log_rms_mae_db":3.1474388119050953,"rms_envelope_nrmse":0.23611171021488647,"envelope_correlation":0.9567281504607078,"rms_gain_error_db":-0.028557615154981807,"reference_mean_power":0.0001030241533277173,"output_mean_power":0.00010234892666778185},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":3.1359809483594144e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.4569632781045499e-05,"output_power":1.4759906935399567e-05,"reference_active":true,"delta_db":0.056350133962302616},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.842890886644966e-06,"output_power":4.012262410688771e-06,"reference_active":true,"delta_db":0.18731275608266384},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.1134901533138973e-06,"output_power":9.409343823782522e-07,"reference_active":true,"delta_db":-0.7312704264846938}],"active_band_count":3,"active_band_mae_db":0.3249777721765534}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":4.327669210129087e-05,"log_rms_mae_db":2.188546024413854,"rms_envelope_nrmse":0.15372420162643466,"envelope_correlation":0.9801618482232903,"rms_gain_error_db":-0.028559458113957305,"reference_mean_power":0.00010363449577271816,"output_mean_power":0.00010295522520094724},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":2.7762525532534862e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":1.3720292251105873e-05,"output_power":1.3954270599747947e-05,"reference_active":true,"delta_db":0.07343778294868943},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.7263230443470644e-06,"output_power":1.8162491997149707e-06,"reference_active":true,"delta_db":0.22053368059616882},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.698769643844656e-08,"output_power":7.364804750918476e-08,"reference_active":true,"delta_db":-1.195554067886775}],"active_band_count":3,"active_band_mae_db":0.4965085104772111}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1018,"rms_floor":1.6513968336769973e-05,"log_rms_mae_db":3.074688526378117,"rms_envelope_nrmse":0.41332097160032705,"envelope_correlation":0.8695417563531326,"rms_gain_error_db":0.44435254195853113,"reference_mean_power":8.51790564150615e-06,"output_mean_power":9.435568655084633e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1018,"frequency_resolution_hz":0.19646365422396858,"power_floor":4.824507951602351e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.4329635827636274e-06,"output_power":2.849978094480737e-06,"reference_active":true,"delta_db":0.687059135985308},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.1930715407234241e-06,"output_power":1.5117475891637836e-06,"reference_active":true,"delta_db":1.0281279843374356},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.146506834057095e-07,"output_power":3.9601880867102155e-07,"reference_active":true,"delta_db":0.998871336154077}],"active_band_count":3,"active_band_mae_db":0.9046861521589401}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":81520,"uncovered_tail_samples":0,"metrics":{"frame_count":1012,"rms_floor":1.2321016663379126e-05,"log_rms_mae_db":2.1342431823443646,"rms_envelope_nrmse":0.2773219018622136,"envelope_correlation":0.9353982393987976,"rms_gain_error_db":0.4443287733358173,"reference_mean_power":8.568334809688712e-06,"output_mean_power":9.491378783704727e-06},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1012,"frequency_resolution_hz":0.19762845849802368,"power_floor":4.130413403311083e-12,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.5185957053959056e-06,"output_power":2.966372775127646e-06,"reference_active":true,"delta_db":0.7106726811707108},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.297939167002703e-07,"output_power":8.350617102511413e-07,"reference_active":true,"delta_db":1.2252010922342471},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9684693476821212e-08,"output_power":3.1979691303835823e-08,"reference_active":true,"delta_db":0.32341698462821705}],"active_band_count":3,"active_band_mae_db":0.7530969193443916}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"hnm","status":"measured","mos":{"status":"measured","value":2.7219724655151367,"predicted_mos":2.7219724655151367,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a8978095215ffc8250826e3773bc2553676d62331d65972fb88b20a47e8ad614","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.218538777000504},"started_at":"2026-10-13T04:40:34.650243+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.575700025568949e-05,"native_input_samples":96000,"native_input_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","output_sha256":"37f88c9389e50fd5b3c7ed56931f0bcf172e4f5f67d200831e3156ec4ec701ce","output_samples":96000,"length_delta_samples":0,"analysis_s":1.074030550000316,"synthesis_s":0.4999214409999695,"total_s":1.5739519910002855,"rtf":0.2623253318333809,"output_peak":0.35889604180693985,"output_rms":0.07322685780816551,"finished_at":"2026-10-13T04:40:36.230837+00:00","evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"37f88c9389e50fd5b3c7ed56931f0bcf172e4f5f67d200831e3156ec4ec701ce","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.298571567755366,"spectral_convergence":0.11765552637920834,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1027264845798435,"spectral_convergence":0.11300679169951228,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.981193321111768,"spectral_convergence":0.11341217223240341,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.0004673732832288897,"delta_db":-0.024695504406582365,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.004423077420750369,"delta_db":-0.08749156621152228,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.000316714802466378,"delta_db":-0.11392042374973196,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":8.129403404765495e-05,"delta_db":-0.1994906896442079,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":2.617366991161892e-05,"delta_db":-0.018979986412191518,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":4.7539494052381945e-05,"delta_db":0.24992295853594557,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11575018816003026,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08213013422134707,"pitch":null,"output_rms_dbfs":-22.706592024036954,"output_peak":0.35889604180693985,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.127497124482326},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.298571567755366,"spectral_convergence":0.11765552637920834,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.1027264845798435,"spectral_convergence":0.11300679169951228,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.981193321111768,"spectral_convergence":0.11341217223240341,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.0004673732832288897,"delta_db":-0.024695504406582365,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.004423077420750369,"delta_db":-0.08749156621152228,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.000316714802466378,"delta_db":-0.11392042374973196,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":8.129403404765495e-05,"delta_db":-0.1994906896442079,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":2.617366991161892e-05,"delta_db":-0.018979986412191518,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":4.7539494052381945e-05,"delta_db":0.24992295853594557,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.11575018816003026,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.08213013422134707,"pitch":null,"output_rms_dbfs":-22.706592024036954,"output_peak":0.35889604180693985,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.127497124482326},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9554428547673377,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07392254417973747,"rms_dbfs":-22.624461889815606,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.89474408065646,"mean":3.481292724609375e-05,"zero_energy":false},"output":{"rms":0.07322685780816551,"rms_dbfs":-22.706592024036954,"peak":0.35889604180693985,"peak_dbfs":-8.900626627756138,"crest_factor_db":13.805965396280818,"mean":-4.360689432052099e-05,"zero_energy":false},"rms_difference_db":-0.08213013422134807},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":1.6450223412763697e-06,"output_rms_dbfs":-57.83828199460415,"output_zero_energy":false,"output_peak":0.03839315031258739,"output_energy_fraction":0.0001043061510437817,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":329,"output_estimated_voiced_frames":333,"both_estimated_voiced_frames":320,"reference_voiced_output_unvoiced_frames":9,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9726443768996961,"absolute_difference_cents_median":0.9026187207909293,"absolute_difference_cents_p95":8.180171577474644,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023914198234423542,"log_rms_mae_db":0.6374014985585941,"rms_envelope_nrmse":0.03828067769287551,"envelope_correlation":0.998788447262883,"rms_gain_error_db":-0.07701348656229282,"reference_mean_power":0.005456771858924342,"output_mean_power":0.005360859783925908},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.071388012833875e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045567084702756786,"output_power":0.0004467090894379745,"reference_active":true,"delta_db":-0.08626454726045907},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.44246106516196e-05,"output_power":6.384581277247412e-05,"reference_active":true,"delta_db":-0.039193825732711275},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7099940640562705e-05,"output_power":1.3826171477102514e-05,"reference_active":true,"delta_db":-0.9229266396319337}],"active_band_count":3,"active_band_mae_db":0.34946167087503466}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022747909354312384,"log_rms_mae_db":0.3634884871550325,"rms_envelope_nrmse":0.020360118299348216,"envelope_correlation":0.9996877993317237,"rms_gain_error_db":-0.07143801784950585,"reference_mean_power":0.005416920568028677,"output_mean_power":0.005328545340028945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.96418119219355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000410225191407457,"output_power":0.00040263742054069294,"reference_active":true,"delta_db":-0.08108191192576297},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7757195967845804e-05,"output_power":2.7784774454598504e-05,"reference_active":true,"delta_db":0.004312841445542569},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0255605769228842e-06,"output_power":8.432246783819767e-07,"reference_active":true,"delta_db":-0.85018009582304}],"active_band_count":3,"active_band_mae_db":0.31185828306478186}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":0.9662243976400919,"rms_envelope_nrmse":0.0789177212801556,"envelope_correlation":0.994873594857182,"rms_gain_error_db":-0.032295804904986104,"reference_mean_power":0.0007990709247977912,"output_mean_power":0.0007931507662103521},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.944736222458957e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022092122499658314,"output_power":0.0002146598372079975,"reference_active":true,"delta_db":-0.12486647030587326},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3665816077667355e-05,"output_power":2.1555280023392623e-05,"reference_active":true,"delta_db":-0.40567815889113257},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.5979447501439136e-06,"output_power":3.358362153351645e-06,"reference_active":true,"delta_db":-2.2190108077118413}],"active_band_count":3,"active_band_mae_db":0.9165184789696158}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":0.6766808359360506,"rms_envelope_nrmse":0.0521126830379262,"envelope_correlation":0.9976875407180542,"rms_gain_error_db":-0.034139724570249,"reference_mean_power":0.0008012273487091395,"output_mean_power":0.000794953621968448},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.73006991688152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002092887788100005,"output_power":0.00020224130980756247,"reference_active":true,"delta_db":-0.14876074605237882},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0995901908054441e-05,"output_power":9.972022836305902e-06,"reference_active":true,"delta_db":-0.42447592562686015},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1412698290887734e-07,"output_power":1.8443843748277026e-07,"reference_active":true,"delta_db":-3.5128209404264643}],"active_band_count":3,"active_band_mae_db":1.3620192040352344}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":1.5153902824274772,"rms_envelope_nrmse":0.12835739828430123,"envelope_correlation":0.9885620594301348,"rms_gain_error_db":-0.23728697268943122,"reference_mean_power":7.57617606122513e-05,"output_mean_power":7.173339098677758e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.472428410826207e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0441893088758165e-05,"output_power":2.795584001510077e-05,"reference_active":true,"delta_db":-0.3699910995089086},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3472076435688955e-06,"output_power":2.8697013177029326e-06,"reference_active":true,"delta_db":-0.6684595744847812},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0111506850143122e-06,"output_power":7.758296613836145e-07,"reference_active":true,"delta_db":-1.1504950095983848}],"active_band_count":3,"active_band_mae_db":0.7296485611973583}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":0.9863665454518984,"rms_envelope_nrmse":0.07892803378195022,"envelope_correlation":0.9957790830453702,"rms_gain_error_db":-0.23834389889340962,"reference_mean_power":7.611087404608644e-05,"output_mean_power":7.20464057235859e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1644437480579494e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.88730260062745e-05,"output_power":2.6531933274584452e-05,"reference_active":true,"delta_db":-0.36723405613320387},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6174980608440689e-06,"output_power":1.3201410233494088e-06,"reference_active":true,"delta_db":-0.8822344172550421},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.602390292027576e-08,"output_power":4.8647755562887074e-08,"reference_active":true,"delta_db":-2.4755633444572958}],"active_band_count":3,"active_band_mae_db":1.2416772726151806}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":1.4171530634140992,"rms_envelope_nrmse":0.18921947339162554,"envelope_correlation":0.9791094123646624,"rms_gain_error_db":0.12124585960479493,"reference_mean_power":5.064740774931043e-05,"output_mean_power":5.2081299031912525e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.700608322524161e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.136344717386396e-05,"output_power":4.239388403230055e-05,"reference_active":true,"delta_db":0.10686482364486621},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.648317567219929e-06,"output_power":6.588728417751608e-06,"reference_active":true,"delta_db":-0.039101493858106344},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8884156339998014e-06,"output_power":2.6667942157659265e-06,"reference_active":true,"delta_db":-1.6378317613576951}],"active_band_count":3,"active_band_mae_db":0.5945993596202226}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":1.04012920972739,"rms_envelope_nrmse":0.10057116741540774,"envelope_correlation":0.9939537314811275,"rms_gain_error_db":0.12074169527669251,"reference_mean_power":5.089623620868824e-05,"output_mean_power":5.233109676370262e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.379392951970341e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.821321246572874e-05,"output_power":3.911274490839523e-05,"reference_active":true,"delta_db":0.10104746534153863},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5895785128125065e-06,"output_power":3.4045004934834036e-06,"reference_active":true,"delta_db":-0.22990055384110833},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2481135158983686e-07,"output_power":1.7007503230143727e-07,"reference_active":true,"delta_db":-1.2117767453368038}],"active_band_count":3,"active_band_mae_db":0.5142415881731502}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.007436513900757,"predicted_mos":3.007436513900757,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2d2b4943832f0e54ababc3a5d3ad57d11ec66c3b485f43699ff33573e76150ab","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3540461789998517},"started_at":"2026-10-13T04:43:51.323203+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.0561999917845242e-05,"native_input_samples":96000,"native_input_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","output_sha256":"0b52a180f29bc57b845d6d8157c7204893f66770a1d5db0344bc06e432a6951b","output_samples":96000,"length_delta_samples":0,"analysis_s":1.3444105209991903,"synthesis_s":0.10884784000063519,"total_s":1.4532583609998255,"rtf":0.24220972683330425,"output_peak":0.46420657539977594,"output_rms":0.08685383520603518,"finished_at":"2026-10-13T04:43:52.783215+00:00","evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"0b52a180f29bc57b845d6d8157c7204893f66770a1d5db0344bc06e432a6951b","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.926474390480137,"spectral_convergence":0.39858580276221023,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.981010317075759,"spectral_convergence":0.2321779370479927,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.020196386694417,"spectral_convergence":0.2470217283811482,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.0005307297345386357,"delta_db":0.5273998665007305,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.006442145210208376,"delta_db":1.5455681200883502,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.0003768891619997264,"delta_db":0.6415324478381773,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":0.00010607362369737315,"delta_db":0.9559966150094691,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":3.766848115394032e-05,"delta_db":1.5621548953112687,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":5.008247839906395e-05,"delta_db":0.4762355287696843,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.95148124558628,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4002418968984158,"pitch":null,"output_rms_dbfs":-21.22421999291719,"output_peak":0.46420657539977594,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.30922703141677},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.926474390480137,"spectral_convergence":0.39858580276221023,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.981010317075759,"spectral_convergence":0.2321779370479927,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.020196386694417,"spectral_convergence":0.2470217283811482,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.0005307297345386357,"delta_db":0.5273998665007305,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.006442145210208376,"delta_db":1.5455681200883502,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.0003768891619997264,"delta_db":0.6415324478381773,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":0.00010607362369737315,"delta_db":0.9559966150094691,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":3.766848115394032e-05,"delta_db":1.5621548953112687,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":5.008247839906395e-05,"delta_db":0.4762355287696843,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.95148124558628,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.4002418968984158,"pitch":null,"output_rms_dbfs":-21.22421999291719,"output_peak":0.46420657539977594,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.30922703141677},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9584127086010163,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07392254417973747,"rms_dbfs":-22.624461889815606,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.89474408065646,"mean":3.481292724609375e-05,"zero_energy":false},"output":{"rms":0.08685383520603518,"rms_dbfs":-21.22421999291719,"peak":0.46420657539977594,"peak_dbfs":-6.665774242719904,"crest_factor_db":14.558445750197288,"mean":5.980970383396016e-05,"zero_energy":false},"rms_difference_db":1.4002418968984145},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":2.9017469917035235e-06,"output_rms_dbfs":-55.37340457128198,"output_zero_energy":false,"output_peak":0.09402016023481145,"output_energy_fraction":0.0001307857596328698,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":329,"output_estimated_voiced_frames":359,"both_estimated_voiced_frames":321,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":38,"paired_voiced_fraction":0.9756838905775076,"absolute_difference_cents_median":4.478987107657678,"absolute_difference_cents_p95":20.258065939423382,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023914198234423542,"log_rms_mae_db":1.5205177182638887,"rms_envelope_nrmse":0.20175403568561306,"envelope_correlation":0.9930198256134717,"rms_gain_error_db":1.3994633383339483,"reference_mean_power":0.005456771858924342,"output_mean_power":0.007531511276340969},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.071388012833875e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045567084702756786,"output_power":0.0005811284181890447,"reference_active":true,"delta_db":1.0562086959373898},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.44246106516196e-05,"output_power":7.894295807777958e-05,"reference_active":true,"delta_db":0.8826159326302616},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7099940640562705e-05,"output_power":1.8228721085917714e-05,"reference_active":true,"delta_db":0.27761597114569914}],"active_band_count":3,"active_band_mae_db":0.7388135332377835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022747909354312384,"log_rms_mae_db":1.1776774025494932,"rms_envelope_nrmse":0.19494923017702398,"envelope_correlation":0.9949141433092593,"rms_gain_error_db":1.3985318244684783,"reference_mean_power":0.005416920568028677,"output_mean_power":0.0074749045219849345},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.96418119219355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000410225191407457,"output_power":0.0005315188343466937,"reference_active":true,"delta_db":1.1249633201475306},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7757195967845804e-05,"output_power":3.413516305300203e-05,"reference_active":true,"delta_db":0.8982639011491539},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0255605769228842e-06,"output_power":1.1681471037837195e-06,"reference_active":true,"delta_db":0.5653621856768929}],"active_band_count":3,"active_band_mae_db":0.8628631356578591}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":1.3465436722732895,"rms_envelope_nrmse":0.18214148592746762,"envelope_correlation":0.9850168012138509,"rms_gain_error_db":0.9377810907032893,"reference_mean_power":0.0007990709247977912,"output_mean_power":0.0009916614658503026},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.944736222458957e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022092122499658314,"output_power":0.00024564929459429944,"reference_active":true,"delta_db":0.4607807865563979},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3665816077667355e-05,"output_power":3.3879919587850094e-05,"reference_active":true,"delta_db":1.5582088587542278},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.5979447501439136e-06,"output_power":6.307431245898573e-06,"reference_active":true,"delta_db":0.5182391738310246}],"active_band_count":3,"active_band_mae_db":0.8457429397138835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":1.1147557181394667,"rms_envelope_nrmse":0.16733742896031,"envelope_correlation":0.9882528844150247,"rms_gain_error_db":0.9308463999671813,"reference_mean_power":0.0008012273487091395,"output_mean_power":0.0009927511638266094},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.73006991688152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002092887788100005,"output_power":0.00023415704690526618,"reference_active":true,"delta_db":0.48761288405434944},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0995901908054441e-05,"output_power":1.5655977532734316e-05,"reference_active":true,"delta_db":1.5344933246559724},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1412698290887734e-07,"output_power":3.9705341560851723e-07,"reference_active":true,"delta_db":-0.18284592025402308}],"active_band_count":3,"active_band_mae_db":0.734984042988115}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":1.6359374243731983,"rms_envelope_nrmse":0.20737929995394158,"envelope_correlation":0.9832303308949051,"rms_gain_error_db":1.042158278112921,"reference_mean_power":7.57617606122513e-05,"output_mean_power":9.630878110309418e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.472428410826207e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0441893088758165e-05,"output_power":3.997201844983554e-05,"reference_active":true,"delta_db":1.1828442280414262},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3472076435688955e-06,"output_power":4.328432497038974e-06,"reference_active":true,"delta_db":1.116479943853786},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0111506850143122e-06,"output_power":7.489285470740464e-07,"reference_active":true,"delta_db":-1.3037549541155018}],"active_band_count":3,"active_band_mae_db":1.2010263753369046}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":1.2765229014845443,"rms_envelope_nrmse":0.16809939546495759,"envelope_correlation":0.992063549942549,"rms_gain_error_db":1.04238829935462,"reference_mean_power":7.611087404608644e-05,"output_mean_power":9.675770065915055e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1644437480579494e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.88730260062745e-05,"output_power":3.796039539286719e-05,"reference_active":true,"delta_db":1.1883842533062268},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6174980608440689e-06,"output_power":1.8461952355381554e-06,"reference_active":true,"delta_db":0.5743385716994385},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.602390292027576e-08,"output_power":5.7757177900609646e-08,"reference_active":true,"delta_db":-1.7301317786308221}],"active_band_count":3,"active_band_mae_db":1.1642848678788291}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":1.7311947311147686,"rms_envelope_nrmse":0.3061399759847506,"envelope_correlation":0.9594670767009718,"rms_gain_error_db":1.0472759151368725,"reference_mean_power":5.064740774931043e-05,"output_mean_power":6.445918551206757e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.700608322524161e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.136344717386396e-05,"output_power":5.313241197910041e-05,"reference_active":true,"delta_db":1.0874280631696194},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.648317567219929e-06,"output_power":8.869807652789389e-06,"reference_active":true,"delta_db":1.2520244596971546},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8884156339998014e-06,"output_power":6.6173861238844135e-06,"reference_active":true,"delta_db":2.309137958143372}],"active_band_count":3,"active_band_mae_db":1.5495301603367153}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":1.3340242986420052,"rms_envelope_nrmse":0.1831267333709251,"envelope_correlation":0.990786993495629,"rms_gain_error_db":1.0475415719873273,"reference_mean_power":5.089623620868824e-05,"output_mean_power":6.477983306569942e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.379392951970341e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.821321246572874e-05,"output_power":5.2189176041863065e-05,"reference_active":true,"delta_db":1.3536689129908177},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5895785128125065e-06,"output_power":4.142789271640451e-06,"reference_active":true,"delta_db":0.6224938605736453},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2481135158983686e-07,"output_power":3.356305936278956e-07,"reference_active":true,"delta_db":1.7404330441716775}],"active_band_count":3,"active_band_mae_db":1.2388652725787137}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"facodec","status":"measured","mos":{"status":"measured","value":3.5710179805755615,"predicted_mos":3.5710179805755615,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b10556e2ad695bdfcb4505d229fbba43f351c00743f8f1f2f69c7184b808d74a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.962620905000222},"started_at":"2026-10-13T07:02:42.277723+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.600000051548705e-05,"native_input_samples":96000,"native_input_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","output_sha256":"1400da435a14cf24b410d9f6c43ee92ea6c13266c7d730dec9963f7fad35af0e","output_samples":96000,"length_delta_samples":0,"analysis_s":1.4910572549997596,"synthesis_s":6.223621278000792,"total_s":7.714678533000551,"rtf":1.2857797555000918,"output_peak":0.3235718011856079,"output_rms":0.06651576546817467,"finished_at":"2026-10-13T07:02:50.007575+00:00","evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"1400da435a14cf24b410d9f6c43ee92ea6c13266c7d730dec9963f7fad35af0e","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.4930385769906245,"spectral_convergence":0.20247978046294232,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7171988213808556,"spectral_convergence":0.19155554496565735,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.280935566304321,"spectral_convergence":0.1979489249076337,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.0005016486479440075,"delta_db":0.28266212721727624,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.0035402198754385505,"delta_db":-1.0544346191047078,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.0002650177289036896,"delta_db":-0.8878547614359635,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":6.08132407564843e-05,"delta_db":-1.4600959629700911,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":2.1511602497005613e-05,"delta_db":-0.8708985550006916,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":3.5135960277479504e-05,"delta_db":-1.0631043205794555,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9365083910513642,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9170462434146452,"pitch":null,"output_rms_dbfs":-23.541508133230252,"output_peak":0.3235718011856079,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.497057654891933},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.4930385769906245,"spectral_convergence":0.20247978046294232,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.7171988213808556,"spectral_convergence":0.19155554496565735,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.280935566304321,"spectral_convergence":0.1979489249076337,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.0005016486479440075,"delta_db":0.28266212721727624,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.0035402198754385505,"delta_db":-1.0544346191047078,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.0002650177289036896,"delta_db":-0.8878547614359635,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":6.08132407564843e-05,"delta_db":-1.4600959629700911,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":2.1511602497005613e-05,"delta_db":-0.8708985550006916,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":3.5135960277479504e-05,"delta_db":-1.0631043205794555,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.9365083910513642,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9170462434146452,"pitch":null,"output_rms_dbfs":-23.541508133230252,"output_peak":0.3235718011856079,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.497057654891933},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9287621521428514,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07392254417973747,"rms_dbfs":-22.624461889815606,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.89474408065646,"mean":3.481292724609375e-05,"zero_energy":false},"output":{"rms":0.06651576546817467,"rms_dbfs":-23.541508133230252,"peak":0.3235718011856079,"peak_dbfs":-9.800586670930008,"crest_factor_db":13.740921462300244,"mean":3.0409969654832854e-05,"zero_energy":false},"rms_difference_db":-0.9170462434146458},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":9.475074738743166e-07,"output_rms_dbfs":-60.234173556611324,"output_zero_energy":false,"output_peak":0.05155205726623535,"output_energy_fraction":7.281357837733257e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":329,"output_estimated_voiced_frames":327,"both_estimated_voiced_frames":313,"reference_voiced_output_unvoiced_frames":16,"reference_unvoiced_output_voiced_frames":14,"paired_voiced_fraction":0.9513677811550152,"absolute_difference_cents_median":2.7789149317338553,"absolute_difference_cents_p95":12.765221469694806,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023914198234423542,"log_rms_mae_db":1.3197406304605874,"rms_envelope_nrmse":0.14368885666901646,"envelope_correlation":0.9899963395024485,"rms_gain_error_db":-0.9175883915416865,"reference_mean_power":0.005456771858924342,"output_mean_power":0.004417504063575817},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.071388012833875e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045567084702756786,"output_power":0.00038190164362794557,"reference_active":true,"delta_db":-0.766997166253136},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.44246106516196e-05,"output_power":6.172877662557693e-05,"reference_active":true,"delta_db":-0.18564132615526777},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7099940640562705e-05,"output_power":1.660746775286548e-05,"reference_active":true,"delta_db":-0.12691184983863557}],"active_band_count":3,"active_band_mae_db":0.35985011408234646}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022747909354312384,"log_rms_mae_db":1.2563684831403041,"rms_envelope_nrmse":0.1408120896532049,"envelope_correlation":0.9907838122980053,"rms_gain_error_db":-0.9272999934930719,"reference_mean_power":0.005416920568028677,"output_mean_power":0.004375447406534246},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.96418119219355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000410225191407457,"output_power":0.00034439825831581513,"reference_active":true,"delta_db":-0.7596137984168622},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7757195967845804e-05,"output_power":2.73974188284938e-05,"reference_active":true,"delta_db":-0.05665942664039136},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0255605769228842e-06,"output_power":9.677311041375129e-07,"reference_active":true,"delta_db":-0.25206617927467573}],"active_band_count":3,"active_band_mae_db":0.3561131347773097}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":1.5453113510775078,"rms_envelope_nrmse":0.13865492626143663,"envelope_correlation":0.985771659121778,"rms_gain_error_db":-0.4579670889034315,"reference_mean_power":0.0007990709247977912,"output_mean_power":0.0007190988922094086},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.944736222458957e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022092122499658314,"output_power":0.00019398641286213716,"reference_active":true,"delta_db":-0.564661304261367},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3665816077667355e-05,"output_power":2.3281506548950788e-05,"reference_active":true,"delta_db":-0.07110404877340093},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.5979447501439136e-06,"output_power":6.866104994743842e-06,"reference_active":true,"delta_db":0.8868183245311262}],"active_band_count":3,"active_band_mae_db":0.5075278925219647}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":1.4660258429816801,"rms_envelope_nrmse":0.12470241799766306,"envelope_correlation":0.9886413709584186,"rms_gain_error_db":-0.4601874354809918,"reference_mean_power":0.0008012273487091395,"output_mean_power":0.0007206709585883569},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.73006991688152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002092887788100005,"output_power":0.00018499762594349902,"reference_active":true,"delta_db":-0.5357978872704658},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0995901908054441e-05,"output_power":1.0576288757121172e-05,"reference_active":true,"delta_db":-0.1689755739990369},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1412698290887734e-07,"output_power":4.866832372787304e-07,"reference_active":true,"delta_db":0.7011285982473912}],"active_band_count":3,"active_band_mae_db":0.46863401983896463}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":1.9857915505996808,"rms_envelope_nrmse":0.22218011314347597,"envelope_correlation":0.9772636520962203,"rms_gain_error_db":-1.3954697495371724,"reference_mean_power":7.57617606122513e-05,"output_mean_power":5.4941825300725595e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.472428410826207e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0441893088758165e-05,"output_power":2.2060988543183227e-05,"reference_active":true,"delta_db":-1.398466873067674},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3472076435688955e-06,"output_power":2.745200545606713e-06,"reference_active":true,"delta_db":-0.8610857805794838},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0111506850143122e-06,"output_power":1.2936636589586053e-06,"reference_active":true,"delta_db":1.0700549789519556}],"active_band_count":3,"active_band_mae_db":1.1098692108663712}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":1.7131783206134275,"rms_envelope_nrmse":0.19983277322007412,"envelope_correlation":0.9843446685323762,"rms_gain_error_db":-1.3959504927064772,"reference_mean_power":7.611087404608644e-05,"output_mean_power":5.518889009833989e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1644437480579494e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.88730260062745e-05,"output_power":2.096570350296599e-05,"reference_active":true,"delta_db":-1.3898286231938348},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6174980608440689e-06,"output_power":1.3136666474285443e-06,"reference_active":true,"delta_db":-0.9035859489630347},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.602390292027576e-08,"output_power":9.173249905556997e-08,"reference_active":true,"delta_db":0.27904082406763275}],"active_band_count":3,"active_band_mae_db":0.8574851320748342}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":1.9169583994812187,"rms_envelope_nrmse":0.20058565642720713,"envelope_correlation":0.9812364902573673,"rms_gain_error_db":-1.0130254944337775,"reference_mean_power":5.064740774931043e-05,"output_mean_power":4.011018569747284e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.700608322524161e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.136344717386396e-05,"output_power":3.0341735931626623e-05,"reference_active":true,"delta_db":-1.345763008031424},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.648317567219929e-06,"output_power":5.815683193918286e-06,"reference_active":true,"delta_db":-0.5811101547543587},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8884156339998014e-06,"output_power":3.174760439998535e-06,"reference_active":true,"delta_db":-0.8806172059457767}],"active_band_count":3,"active_band_mae_db":0.9358301229105198}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":1.7555477922516023,"rms_envelope_nrmse":0.1666080811858545,"envelope_correlation":0.9892245524460354,"rms_gain_error_db":-1.012950485746433,"reference_mean_power":5.089623620868824e-05,"output_mean_power":4.0307941422147404e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.379392951970341e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.821321246572874e-05,"output_power":2.795446782332259e-05,"reference_active":true,"delta_db":-1.3576232011405907},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5895785128125065e-06,"output_power":3.0003524210387147e-06,"reference_active":true,"delta_db":-0.7787118695553942},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2481135158983686e-07,"output_power":1.7591734393120616e-07,"reference_active":true,"delta_db":-1.0650957739505376}],"active_band_count":3,"active_band_mae_db":1.0671436148821742}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"hifiglot","status":"measured","mos":{"status":"measured","value":2.981276273727417,"predicted_mos":2.981276273727417,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"afc521aa4c31fe0f2b13368e6c48c7a273cd2752c12ff88a8884286237ac64dd","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.028768964999472},"started_at":"2026-10-13T07:34:10.627713+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.006725774001097307,"native_input_samples":264600,"native_input_sha256":"687ea82b437ab3d542d1b545c18deb20b158f9413af9e5fd2b7667e8408f2045","output_sha256":"1317c04e1db53b106c15e849c2dd64d7c60b737d311de747a363de24ecf3dd2c","output_samples":264704,"length_delta_samples":104,"analysis_s":0.1821544949998497,"synthesis_s":5.076478949999,"total_s":5.25863344499885,"rtf":0.8764389074998084,"output_peak":0.31346818804740906,"output_rms":0.06671623333839784,"finished_at":"2026-10-13T07:34:15.916685+00:00","evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"1317c04e1db53b106c15e849c2dd64d7c60b737d311de747a363de24ecf3dd2c","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.513182216933053,"output_peak":0.31343298632437816},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.240516698875831,"spectral_convergence":0.4061932664032348,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":8.54520415698716,"spectral_convergence":0.2788113448693156,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.729130267944706,"spectral_convergence":0.3385357030155285,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.0004429496890630787,"delta_db":-0.25779032181276645,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.0036303784252463634,"delta_db":-0.9452180026321416,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.0002769932844956309,"delta_db":-0.6959116403844365,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":5.088843775745356e-05,"delta_db":-2.2338862549952427,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":2.615051155389168e-05,"delta_db":-0.02282430746831121,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":2.1401256472735153e-05,"delta_db":-3.2162298787846315,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2286434010129217,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.8931467714843938,"pitch":null,"output_rms_dbfs":-23.517608661300002,"output_peak":0.31343298632437816,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":8.504950374602565},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8052475808553291,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07392254417973747,"rms_dbfs":-22.624461889815606,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.89474408065646,"mean":3.481292724609375e-05,"zero_energy":false},"output":{"rms":0.06669903750871638,"rms_dbfs":-23.517608661300002,"peak":0.31343298632437816,"peak_dbfs":-10.077105988567958,"crest_factor_db":13.440502672732045,"mean":-7.475402210418388e-05,"zero_energy":false},"rms_difference_db":-0.8931467714843961},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":5.366190330610933e-07,"output_rms_dbfs":-62.70333927626035,"output_zero_energy":false,"output_peak":0.01188938491622468,"output_energy_fraction":4.1011519037694346e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":329,"output_estimated_voiced_frames":316,"both_estimated_voiced_frames":303,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":13,"paired_voiced_fraction":0.9209726443768997,"absolute_difference_cents_median":8.37552481580816,"absolute_difference_cents_p95":36.52259970726326,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023914198234423542,"log_rms_mae_db":1.3163921051464063,"rms_envelope_nrmse":0.12121729424219618,"envelope_correlation":0.9950334980954659,"rms_gain_error_db":-0.8916579671468285,"reference_mean_power":0.005456771858924342,"output_mean_power":0.004443958556257649},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.071388012833875e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045567084702756786,"output_power":0.0003758939765312622,"reference_active":true,"delta_db":-0.8358587765003285},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.44246106516196e-05,"output_power":5.654419468413746e-05,"reference_active":true,"delta_db":-0.5666377945984009},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7099940640562705e-05,"output_power":1.0592562021741439e-05,"reference_active":true,"delta_db":-2.0799358724925128}],"active_band_count":3,"active_band_mae_db":1.1608108145304141}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022747909354312384,"log_rms_mae_db":1.0679531237712625,"rms_envelope_nrmse":0.10822778226338195,"envelope_correlation":0.9978280116726657,"rms_gain_error_db":-0.891545023498604,"reference_mean_power":0.005416920568028677,"output_mean_power":0.004411618659222545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.96418119219355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000410225191407457,"output_power":0.0003400383725920432,"reference_active":true,"delta_db":-0.8149439732615553},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7757195967845804e-05,"output_power":2.464483906239331e-05,"reference_active":true,"delta_db":-0.5164960512962788},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0255605769228842e-06,"output_power":7.992125207452389e-07,"reference_active":true,"delta_db":-1.0829903891014712}],"active_band_count":3,"active_band_mae_db":0.8048101378864351}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":2.5080251204269044,"rms_envelope_nrmse":0.31981693948762835,"envelope_correlation":0.9320361569015089,"rms_gain_error_db":0.6194589928022247,"reference_mean_power":0.0007990709247977912,"output_mean_power":0.000921576152153626},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.944736222458957e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022092122499658314,"output_power":0.00026243419376779237,"reference_active":true,"delta_db":0.7478297793349387},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3665816077667355e-05,"output_power":2.555642409891386e-05,"reference_active":true,"delta_db":0.33378601457750007},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.5979447501439136e-06,"output_power":5.616440970955861e-06,"reference_active":true,"delta_db":0.01432590980608536}],"active_band_count":3,"active_band_mae_db":0.36531390123950797}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":2.2828827079605283,"rms_envelope_nrmse":0.3000684129330988,"envelope_correlation":0.9389239280631846,"rms_gain_error_db":0.6193042640952655,"reference_mean_power":0.0008012273487091395,"output_mean_power":0.0009240302549396661},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.73006991688152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002092887788100005,"output_power":0.00024850656145148357,"reference_active":true,"delta_db":0.7459191622285676},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0995901908054441e-05,"output_power":1.2683569148456756e-05,"reference_active":true,"delta_db":0.6201062399433214},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1412698290887734e-07,"output_power":3.461128190035831e-07,"reference_active":true,"delta_db":-0.7791584378348764}],"active_band_count":3,"active_band_mae_db":0.7150612800022551}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":4.636560923757973,"rms_envelope_nrmse":0.4154949022448036,"envelope_correlation":0.8864952801593577,"rms_gain_error_db":-2.054430749211849,"reference_mean_power":7.57617606122513e-05,"output_mean_power":4.720706323350394e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.472428410826207e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0441893088758165e-05,"output_power":2.007652675428867e-05,"reference_active":true,"delta_db":-1.8078307450798898},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3472076435688955e-06,"output_power":5.357307699669935e-06,"reference_active":true,"delta_db":2.0426393641952556},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0111506850143122e-06,"output_power":1.076369666138128e-06,"reference_active":true,"delta_db":0.271455697218739}],"active_band_count":3,"active_band_mae_db":1.3739752688312947}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":4.263660718962802,"rms_envelope_nrmse":0.37720734706586856,"envelope_correlation":0.9078612262688448,"rms_gain_error_db":-2.0530394702458317,"reference_mean_power":7.611087404608644e-05,"output_mean_power":4.743979048441095e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1644437480579494e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.88730260062745e-05,"output_power":1.9341962797254662e-05,"reference_active":true,"delta_db":-1.7399175840657255},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6174980608440689e-06,"output_power":2.7973701442162227e-06,"reference_active":true,"delta_db":2.379061667472458},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.602390292027576e-08,"output_power":7.501535837977857e-08,"reference_active":true,"delta_db":-0.5946895427058655}],"active_band_count":3,"active_band_mae_db":1.571222931414683}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":4.686233371699487,"rms_envelope_nrmse":0.29827764328721085,"envelope_correlation":0.9589445707086496,"rms_gain_error_db":-1.5712005307795023,"reference_mean_power":5.064740774931043e-05,"output_mean_power":3.527257527773603e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.700608322524161e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.136344717386396e-05,"output_power":2.8032673206397965e-05,"reference_active":true,"delta_db":-1.6895221088780221},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.648317567219929e-06,"output_power":5.800112954414962e-06,"reference_active":true,"delta_db":-0.5927530470052568},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8884156339998014e-06,"output_power":2.7578886226745054e-06,"reference_active":true,"delta_db":-1.4919595740016622}],"active_band_count":3,"active_band_mae_db":1.2580782432949804}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":4.765657083268201,"rms_envelope_nrmse":0.260924375695474,"envelope_correlation":0.9728831471625166,"rms_gain_error_db":-1.5707920179904657,"reference_mean_power":5.089623620868824e-05,"output_mean_power":3.5449202190523905e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.379392951970341e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.821321246572874e-05,"output_power":2.5925549785516226e-05,"reference_active":true,"delta_db":-1.6848557406061015},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5895785128125065e-06,"output_power":3.191740096921791e-06,"reference_active":true,"delta_db":-0.5101593728752498},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2481135158983686e-07,"output_power":2.0168724957679148e-07,"reference_active":true,"delta_db":-0.471397931391393}],"active_band_count":3,"active_band_mae_db":0.8888043482909147}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"golf","status":"measured","mos":{"status":"measured","value":2.8910446166992188,"predicted_mos":2.8910446166992188,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"6659fc824109e32645307cc56460a169bb3963102428bd2bf5532cdc39d04428","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6045713140010776},"started_at":"2026-10-13T05:03:12.387428+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004198014999929001,"native_input_samples":144000,"native_input_sha256":"1a47e4b438169611fff209aa4e9ae7bca097925ba020f52904b4c664bd14f8e5","output_sha256":"685c205ae6a61bae0169d31e3dff11eecfddee132f33dba4800b09b7905498bf","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.6943582489993787,"synthesis_s":0.14137608500004717,"total_s":0.8357343339994259,"rtf":0.139289055666571,"output_peak":0.34786343574523926,"output_rms":0.06922854654839794,"finished_at":"2026-10-13T05:03:13.237325+00:00","analysis_stages":{"external_f0_s":0.09073919700040278},"evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"685c205ae6a61bae0169d31e3dff11eecfddee132f33dba4800b09b7905498bf","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.191937312875034,"output_peak":0.3480503595753579},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.906609312830701,"spectral_convergence":0.43019708655531147,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.0969302045612395,"spectral_convergence":0.15995700056181913,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.735488524558686,"spectral_convergence":0.1654184580218623,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004745845141331796,"output_power":0.00042893698237084444,"delta_db":-0.43920070783607645,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004469242507355049,"output_power":0.003964666034081185,"delta_db":-0.5202731057690009,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00032577925907867166,"output_power":0.0002771484269369779,"delta_db":-0.7021101331879458,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.539530954786978e-05,"output_power":6.96411113718988e-05,"delta_db":-0.8856832392065351,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6382961823164746e-05,"output_power":1.798780823537785e-05,"delta_db":-1.663453000396153,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.500188672709416e-05,"output_power":3.6814592331444625e-05,"delta_db":-0.8721072652125914,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8471379086013838,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5370444998026738,"pitch":null,"output_rms_dbfs":-23.191937312875034,"output_peak":0.3480503595753579,"reference_rms_dbfs":-22.65489281307236,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.2463426806502085},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.9231994353255468,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07366401047095542,"rms_dbfs":-22.65489281307236,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.925175003913212,"mean":1.897827810755556e-05,"zero_energy":false},"output":{"rms":0.06924734619700404,"rms_dbfs":-23.191937312875034,"peak":0.3480503595753579,"peak_dbfs":-9.16715826455355,"crest_factor_db":14.024779048321484,"mean":0.0002341652517608268,"zero_energy":false},"rms_difference_db":-0.537044499802672},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":4.829898930110927e-07,"output_rms_dbfs":-63.16061957147944,"output_zero_energy":false,"output_peak":0.007231610865113677,"output_energy_fraction":3.4303238209819125e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":328,"output_estimated_voiced_frames":313,"both_estimated_voiced_frames":304,"reference_voiced_output_unvoiced_frames":24,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.926829268292683,"absolute_difference_cents_median":2.4794385769337883,"absolute_difference_cents_p95":15.959760455957598,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00023914198234423542,"log_rms_mae_db":1.9156725871220748,"rms_envelope_nrmse":0.09850770761593367,"envelope_correlation":0.9944390918805956,"rms_gain_error_db":-0.5380903489108436,"reference_mean_power":0.00541891515998593,"output_mean_power":0.004787439698553895},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.0737328883020127e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045369401245065377,"output_power":0.0004317876778965987,"reference_active":true,"delta_db":-0.2149280324253941},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.428199077839398e-05,"output_power":5.618465410694985e-05,"reference_active":true,"delta_db":-0.5847160662989579},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.70440787664134e-05,"output_power":1.0857723081856558e-05,"reference_active":true,"delta_db":-1.958347714502571}],"active_band_count":3,"active_band_mae_db":0.9193306044089743}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00022747909354312384,"log_rms_mae_db":1.8564812601693945,"rms_envelope_nrmse":0.08228758213265132,"envelope_correlation":0.9970272944806088,"rms_gain_error_db":-0.5366620342334122,"reference_mean_power":0.0053834705276696836,"output_mean_power":0.0047576899493748815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.9667357267880705e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004082249840050515,"output_power":0.0003913088448330028,"reference_active":true,"delta_db":-0.183799162346992},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.763274540478699e-05,"output_power":2.4465226437974827e-05,"reference_active":true,"delta_db":-0.5287479592706447},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0215743714734416e-06,"output_power":8.313411899438813e-07,"reference_active":true,"delta_db":-0.8949069036990669}],"active_band_count":3,"active_band_mae_db":0.5358180084389011}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":1.5245937511188419,"rms_envelope_nrmse":0.16838183010141045,"envelope_correlation":0.9771483456400398,"rms_gain_error_db":-0.04002161421606586,"reference_mean_power":0.0007994321559523929,"output_mean_power":0.0007920989757188505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.942114097601657e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022008289762911844,"output_power":0.00022325869847745073,"reference_active":true,"delta_db":0.06222093247588877},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.359242018264139e-05,"output_power":2.3417590739892e-05,"reference_active":true,"delta_db":-0.03230282691704658},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.557515468015537e-06,"output_power":3.946039949547888e-06,"reference_active":true,"delta_db":-1.487192024094709}],"active_band_count":3,"active_band_mae_db":0.5272385944958814}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":1.281680806179867,"rms_envelope_nrmse":0.14282121072102696,"envelope_correlation":0.98301927611524,"rms_gain_error_db":-0.03409671086434654,"reference_mean_power":0.0008012638248927989,"output_mean_power":0.0007949976863724554},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.727733229670872e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020836293871437728,"output_power":0.00021294673624273365,"reference_active":true,"delta_db":0.09450514169896154},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0949113566083186e-05,"output_power":1.0802776131087116e-05,"reference_active":true,"delta_db":-0.058435841889386976},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1113981634704473e-07,"output_power":3.266343369488129e-07,"reference_active":true,"delta_db":-0.9992770018825456}],"active_band_count":3,"active_band_mae_db":0.38407266182363137}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":2.598591363911705,"rms_envelope_nrmse":0.2148851116198181,"envelope_correlation":0.9717012890100755,"rms_gain_error_db":-0.9070680429410805,"reference_mean_power":7.586181968425925e-05,"output_mean_power":6.156252887535973e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.4864198068658206e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.043918227820822e-05,"output_power":2.491032991380041e-05,"reference_active":true,"delta_db":-0.8705350192196601},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.340335508616212e-06,"output_power":1.7486714063454674e-06,"reference_active":true,"delta_db":-2.8108188156369636},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.006074937669575e-06,"output_power":6.25678175086548e-07,"reference_active":true,"delta_db":-2.0627932424945303}],"active_band_count":3,"active_band_mae_db":1.9147156924503845}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":2.4321254489473403,"rms_envelope_nrmse":0.17025209894388268,"envelope_correlation":0.9842790649575222,"rms_gain_error_db":-0.908021989988908,"reference_mean_power":7.622558274062687e-05,"output_mean_power":6.184413993985902e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":5.17780637044169e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.885736007529938e-05,"output_power":2.3497772863722712e-05,"reference_active":true,"delta_db":-0.8922989704139705},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.607347117932334e-06,"output_power":8.345929558996931e-07,"reference_active":true,"delta_db":-2.846349609694629},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.524190070357533e-08,"output_power":4.0491028876955475e-08,"reference_active":true,"delta_db":-3.232943126144672}],"active_band_count":3,"active_band_mae_db":2.323863902084424}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":2.9577350838378833,"rms_envelope_nrmse":0.2792245566170875,"envelope_correlation":0.9573980527337952,"rms_gain_error_db":-0.9613838278991476,"reference_mean_power":5.072536666472632e-05,"output_mean_power":4.065245822441255e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":9.709497486562244e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.1398043321491535e-05,"output_power":3.288040897314301e-05,"reference_active":true,"delta_db":-1.0004260393363253},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.626948567260806e-06,"output_power":7.300588791869524e-06,"reference_active":true,"delta_db":0.420442873665007},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.87454881331478e-06,"output_power":3.4481375592795226e-06,"reference_active":true,"delta_db":-0.5063655354975185}],"active_band_count":3,"active_band_mae_db":0.6424114828329502}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":2.993125693748918,"rms_envelope_nrmse":0.20435664090365596,"envelope_correlation":0.980324082568464,"rms_gain_error_db":-0.9615352930793325,"reference_mean_power":5.0979554771806064e-05,"output_mean_power":4.085474545210186e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":9.38766692650384e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8221320135458574e-05,"output_power":3.0908398746672043e-05,"reference_active":true,"delta_db":-0.9222917677464206},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5671113653322416e-06,"output_power":4.119310254200676e-06,"reference_active":true,"delta_db":0.6250783471961223},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.24297703348347e-07,"output_power":2.719641734456808e-07,"reference_active":true,"delta_db":0.8368687031635336}],"active_band_count":3,"active_band_mae_db":0.7947462727020255}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.105867862701416,"predicted_mos":3.105867862701416,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"abfb7fdc05659aea61480b2f514c6a1e3f1d3ad9fa92992474f8aa95f914081b","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.90915657000005},"started_at":"2026-10-13T07:48:21.780290+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.004242982999130618,"native_input_samples":144000,"native_input_sha256":"1a47e4b438169611fff209aa4e9ae7bca097925ba020f52904b4c664bd14f8e5","output_sha256":"aa891eaf75941075bc2e99fc12e6ad1e6256db586fca7bccc8c0833768569cf9","output_samples":144120,"length_delta_samples":120,"analysis_s":2.115393555999617,"synthesis_s":14.397544507999555,"total_s":16.512938063999172,"rtf":2.752156343999862,"output_peak":0.3495749831199646,"output_rms":0.07106647839501685,"finished_at":"2026-10-13T07:48:38.308370+00:00","analysis_stages":{"external_f0_s":1.3175045319985657},"evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"aa891eaf75941075bc2e99fc12e6ad1e6256db586fca7bccc8c0833768569cf9","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.96456845203595,"output_peak":0.34875493498226184},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.624614637857156,"spectral_convergence":0.4144437944569803,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.015765721505556,"spectral_convergence":0.1530325078644255,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.5066671225444574,"spectral_convergence":0.16435216661072885,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.00045481128770060006,"delta_db":-0.1430219868569694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.004157370845299928,"delta_db":-0.3565493175076151,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.0002990456444406141,"delta_db":-0.36322922717192196,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":7.749073003901304e-05,"delta_db":-0.4075799162007637,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":2.5839100490921883e-05,"delta_db":-0.07485228706697931,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":3.2997944767886756e-05,"delta_db":-1.3357536822695795,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4468310695123048,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3447270358795871,"pitch":null,"output_rms_dbfs":-22.969188925695192,"output_peak":0.34875493498226184,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.04901582730239},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9372751274638871,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07392254417973747,"rms_dbfs":-22.624461889815606,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.89474408065646,"mean":3.481292724609375e-05,"zero_energy":false},"output":{"rms":0.07104615086504662,"rms_dbfs":-22.969188925695192,"peak":0.34875493498226184,"peak_dbfs":-9.149592765670098,"crest_factor_db":13.819596160025096,"mean":-4.2612275273064974e-05,"zero_energy":false},"rms_difference_db":-0.3447270358795862},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":4.714782541323701e-06,"output_rms_dbfs":-53.26538333315001,"output_zero_energy":false,"output_peak":0.12134264602057851,"output_energy_fraction":0.00031758463028310906,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":329,"output_estimated_voiced_frames":317,"both_estimated_voiced_frames":308,"reference_voiced_output_unvoiced_frames":21,"reference_unvoiced_output_voiced_frames":9,"paired_voiced_fraction":0.9361702127659575,"absolute_difference_cents_median":4.546934468545079,"absolute_difference_cents_p95":16.622630026465046,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023914198234423542,"log_rms_mae_db":1.5450541864538097,"rms_envelope_nrmse":0.09817445614219468,"envelope_correlation":0.9926300068405541,"rms_gain_error_db":-0.3401012984346047,"reference_mean_power":0.005456771858924342,"output_mean_power":0.00504574928094547},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.071388012833875e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045567084702756786,"output_power":0.0004195833165866442,"reference_active":true,"delta_db":-0.3583303261068738},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.44246106516196e-05,"output_power":6.227111900780199e-05,"reference_active":true,"delta_db":-0.14765132549972615},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7099940640562705e-05,"output_power":1.4539730232187104e-05,"reference_active":true,"delta_db":-0.7043825404993536}],"active_band_count":3,"active_band_mae_db":0.4034547307019845}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022747909354312384,"log_rms_mae_db":1.1020420440355472,"rms_envelope_nrmse":0.06727590699867451,"envelope_correlation":0.9971192327444147,"rms_gain_error_db":-0.33433238178439684,"reference_mean_power":0.005416920568028677,"output_mean_power":0.005015557678205911},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.96418119219355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000410225191407457,"output_power":0.00037929274101447253,"reference_active":true,"delta_db":-0.34047795161509026},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7757195967845804e-05,"output_power":2.710034861318402e-05,"reference_active":true,"delta_db":-0.10400713983498894},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0255605769228842e-06,"output_power":9.295804707778515e-07,"reference_active":true,"delta_db":-0.4267432678655203}],"active_band_count":3,"active_band_mae_db":0.29040945310519983}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":1.4857731982807791,"rms_envelope_nrmse":0.14468179023768074,"envelope_correlation":0.9828251806963094,"rms_gain_error_db":-0.21563411836308244,"reference_mean_power":0.0007990709247977912,"output_mean_power":0.000760364647940022},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.944736222458957e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022092122499658314,"output_power":0.00020991197073273542,"reference_active":true,"delta_db":-0.2220023672709198},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3665816077667355e-05,"output_power":2.441681915907115e-05,"reference_active":true,"delta_db":0.13567601640541088},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.5979447501439136e-06,"output_power":5.7954304692262865e-06,"reference_active":true,"delta_db":0.15057091932655273}],"active_band_count":3,"active_band_mae_db":0.16941643433429446}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":1.1274075195220612,"rms_envelope_nrmse":0.10311945574296166,"envelope_correlation":0.9911702013870011,"rms_gain_error_db":-0.2149429890905618,"reference_mean_power":0.0008012273487091395,"output_mean_power":0.0007625379560288346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.73006991688152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002092887788100005,"output_power":0.00019925771159538157,"reference_active":true,"delta_db":-0.21330805615518805},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0995901908054441e-05,"output_power":1.1670041825057955e-05,"reference_active":true,"delta_db":0.2584155559954554},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1412698290887734e-07,"output_power":4.2383711851074545e-07,"reference_active":true,"delta_db":0.10065460009536488}],"active_band_count":3,"active_band_mae_db":0.19079273741533612}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":2.0310394923201485,"rms_envelope_nrmse":0.21619611484409226,"envelope_correlation":0.9667138995163209,"rms_gain_error_db":-0.36516366912491294,"reference_mean_power":7.57617606122513e-05,"output_mean_power":6.965201609140477e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.472428410826207e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0441893088758165e-05,"output_power":2.838296665275582e-05,"reference_active":true,"delta_db":-0.30413869442469776},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3472076435688955e-06,"output_power":3.186436746962434e-06,"reference_active":true,"delta_db":-0.2137735274546965},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0111506850143122e-06,"output_power":9.55829728691495e-07,"reference_active":true,"delta_db":-0.24435346374021294}],"active_band_count":3,"active_band_mae_db":0.2540885618732024}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":1.3292975331066874,"rms_envelope_nrmse":0.11866324252967278,"envelope_correlation":0.9903423621705001,"rms_gain_error_db":-0.36514159522430745,"reference_mean_power":7.611087404608644e-05,"output_mean_power":6.99733312139134e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1644437480579494e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.88730260062745e-05,"output_power":2.6825018649697256e-05,"reference_active":true,"delta_db":-0.31952269454735016},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6174980608440689e-06,"output_power":1.5732420967259107e-06,"reference_active":true,"delta_db":-0.12048209961013566},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.602390292027576e-08,"output_power":6.376972146517512e-08,"reference_active":true,"delta_db":-1.3000462261530252}],"active_band_count":3,"active_band_mae_db":0.5800170067701703}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":2.029162378251728,"rms_envelope_nrmse":0.3085979363310558,"envelope_correlation":0.9432963592264156,"rms_gain_error_db":-0.8766432578379173,"reference_mean_power":5.064740774931043e-05,"output_mean_power":4.138975887122187e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.700608322524161e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.136344717386396e-05,"output_power":3.230856819119507e-05,"reference_active":true,"delta_db":-1.0729901316000938},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.648317567219929e-06,"output_power":5.513692035742754e-06,"reference_active":true,"delta_db":-0.8126925085625465},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8884156339998014e-06,"output_power":3.5947002421194125e-06,"reference_active":true,"delta_db":-0.3410999969645853}],"active_band_count":3,"active_band_mae_db":0.7422608790424086}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":1.4991622619787968,"rms_envelope_nrmse":0.1672418395737857,"envelope_correlation":0.9871792039663125,"rms_gain_error_db":-0.8768486406174052,"reference_mean_power":5.089623620868824e-05,"output_mean_power":4.1591137975293634e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.379392951970341e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.821321246572874e-05,"output_power":3.0420789604564737e-05,"reference_active":true,"delta_db":-0.9904306656003821},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5895785128125065e-06,"output_power":2.868842467445144e-06,"reference_active":true,"delta_db":-0.9733675572721692},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2481135158983686e-07,"output_power":1.942801177976183e-07,"reference_active":true,"delta_db":-0.6338987852320849}],"active_band_count":3,"active_band_mae_db":0.8658990027015454}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0485494136810303,"predicted_mos":3.0485494136810303,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"3b6c1e469bc22c7f566871d05fe304d73b7e421f86c87fe21ff746e7a1a3c09f","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4917070150004292},"started_at":"2026-10-13T05:11:40.914209+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027042769997933647,"native_input_samples":144000,"native_input_sha256":"1a47e4b438169611fff209aa4e9ae7bca097925ba020f52904b4c664bd14f8e5","output_sha256":"4e8b66eb4621307fdbce1089c8cd0503f512a8dd72fd4a9b45d8a4cfcbdf1801","output_samples":144240,"length_delta_samples":240,"analysis_s":1.1262227139995957,"synthesis_s":0.1770322180000221,"total_s":1.3032549319996178,"rtf":0.21720915533326965,"output_peak":0.36164388060569763,"output_rms":0.07319777225791536,"finished_at":"2026-10-13T05:11:42.229462+00:00","analysis_stages":{"external_f0_s":0.16138802600107738},"evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"4e8b66eb4621307fdbce1089c8cd0503f512a8dd72fd4a9b45d8a4cfcbdf1801","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-22.707814155190892,"output_peak":0.3616614654689116},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.174346515021749,"spectral_convergence":0.09102039068920703,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":3.9917718846673695,"spectral_convergence":0.08166056360242219,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":3.909490167184463,"spectral_convergence":0.08361109153586663,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00047003850181382925,"output_power":0.00046516507258774974,"delta_db":-0.045263355215469196,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004513086919531444,"output_power":0.004456607334011644,"delta_db":-0.05469327230750481,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0003251325154635926,"output_power":0.0003104290635894683,"delta_db":-0.2009802922435996,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.51153215194187e-05,"output_power":7.898785566585643e-05,"delta_db":-0.3244742010016063,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6288307106843834e-05,"output_power":2.0394293581900925e-05,"delta_db":-1.1025395285227018,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.4880972570108e-05,"output_power":3.644613706978416e-05,"delta_db":-0.904107553343369,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4386763671057084,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.07738896433748974,"pitch":null,"output_rms_dbfs":-22.701850854153097,"output_peak":0.3616614654689116,"reference_rms_dbfs":-22.624461889815606,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.025202855624527},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9804932200770412,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07392254417973747,"rms_dbfs":-22.624461889815606,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.89474408065646,"mean":3.481292724609375e-05,"zero_energy":false},"output":{"rms":0.07326683940573939,"rms_dbfs":-22.701850854153097,"peak":0.3616614654689116,"peak_dbfs":-8.833955245046631,"crest_factor_db":13.867895609106466,"mean":-2.9749585120116542e-05,"zero_energy":false},"rms_difference_db":-0.07738896433749076},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":1.1632814860839828e-06,"output_rms_dbfs":-59.3431518375711,"output_zero_energy":false,"output_peak":0.030855667711714517,"output_energy_fraction":7.367986453300917e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":329,"output_estimated_voiced_frames":345,"both_estimated_voiced_frames":325,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9878419452887538,"absolute_difference_cents_median":0.644997535671905,"absolute_difference_cents_p95":7.160391410999167,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00023914198234423542,"log_rms_mae_db":0.6529520955086127,"rms_envelope_nrmse":0.03230592790120639,"envelope_correlation":0.9991640136099925,"rms_gain_error_db":-0.07485810588154837,"reference_mean_power":0.005456771858924342,"output_mean_power":0.005363521010773885},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.071388012833875e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00045567084702756786,"output_power":0.00044473003306942513,"reference_active":true,"delta_db":-0.10554785333494292},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.44246106516196e-05,"output_power":5.598378865070679e-05,"reference_active":true,"delta_db":-0.6098951698512387},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.7099940640562705e-05,"output_power":1.4353201828844607e-05,"reference_active":true,"delta_db":-0.7604581105157087}],"active_band_count":3,"active_band_mae_db":0.4919670445672968}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00022747909354312384,"log_rms_mae_db":0.46333828347763023,"rms_envelope_nrmse":0.021826272528349143,"envelope_correlation":0.9996603724461536,"rms_gain_error_db":-0.07295831164968516,"reference_mean_power":0.005416920568028677,"output_mean_power":0.005326680352782978},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.96418119219355e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000410225191407457,"output_power":0.0004040855022317582,"reference_active":true,"delta_db":-0.06549057206237738},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7757195967845804e-05,"output_power":2.413517067916163e-05,"reference_active":true,"delta_db":-0.607252171563031},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0255605769228842e-06,"output_power":8.599011713476917e-07,"reference_active":true,"delta_db":-0.7651277745147639}],"active_band_count":3,"active_band_mae_db":0.47929017271339075}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":0.6237778638621365,"rms_envelope_nrmse":0.05719193416681199,"envelope_correlation":0.9974038169738102,"rms_gain_error_db":-0.11799674415883447,"reference_mean_power":0.0007990709247977912,"output_mean_power":0.0007776526462540489},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.944736222458957e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00022092122499658314,"output_power":0.0002191056727642554,"reference_active":true,"delta_db":-0.03583820873334765},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3665816077667355e-05,"output_power":2.1863440762343186e-05,"reference_active":true,"delta_db":-0.34402974857075097},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.5979447501439136e-06,"output_power":4.02465893811763e-06,"reference_active":true,"delta_db":-1.4329952498567253}],"active_band_count":3,"active_band_mae_db":0.6042877357202746}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":0.43137277032790056,"rms_envelope_nrmse":0.037719760134349564,"envelope_correlation":0.9989131355864004,"rms_gain_error_db":-0.11799374819511368,"reference_mean_power":0.0008012273487091395,"output_mean_power":0.000779751807337458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.73006991688152e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0002092887788100005,"output_power":0.00020740918270366142,"reference_active":true,"delta_db":-0.039179637477446884},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0995901908054441e-05,"output_power":1.009493890787183e-05,"reference_active":true,"delta_db":-0.37127161903727274},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1412698290887734e-07,"output_power":2.8555412131802623e-07,"reference_active":true,"delta_db":-1.614450957796083}],"active_band_count":3,"active_band_mae_db":0.6749674047702676}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":1.1077347784455034,"rms_envelope_nrmse":0.10829718497235158,"envelope_correlation":0.9928435949135732,"rms_gain_error_db":-0.40318753619011577,"reference_mean_power":7.57617606122513e-05,"output_mean_power":6.904485230995486e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.472428410826207e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0441893088758165e-05,"output_power":2.953026689195758e-05,"reference_active":true,"delta_db":-0.13204284293291363},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.3472076435688955e-06,"output_power":2.7131928980571243e-06,"reference_active":true,"delta_db":-0.912019829893384},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0111506850143122e-06,"output_power":6.390786838668911e-07,"reference_active":true,"delta_db":-1.9926154828549791}],"active_band_count":3,"active_band_mae_db":1.012226051893759}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":0.8459760268565246,"rms_envelope_nrmse":0.07909555633452281,"envelope_correlation":0.9967367148096324,"rms_gain_error_db":-0.40276811655826383,"reference_mean_power":7.611087404608644e-05,"output_mean_power":6.936971299532678e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":5.1644437480579494e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.88730260062745e-05,"output_power":2.8062357498314667e-05,"reference_active":true,"delta_db":-0.12368149040023324},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.6174980608440689e-06,"output_power":1.166997787915663e-06,"reference_active":true,"delta_db":-1.4177373585611228},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.602390292027576e-08,"output_power":4.21801600582839e-08,"reference_active":true,"delta_db":-3.095109192901363}],"active_band_count":3,"active_band_mae_db":1.5455093472875732}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":1.1582965228757556,"rms_envelope_nrmse":0.16238470665710728,"envelope_correlation":0.9899819354595963,"rms_gain_error_db":-0.9632494947093571,"reference_mean_power":5.064740774931043e-05,"output_mean_power":4.057254705454849e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":9.700608322524161e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.136344717386396e-05,"output_power":3.167471673283814e-05,"reference_active":true,"delta_db":-1.1590398543517202},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.648317567219929e-06,"output_power":4.466669084570248e-06,"reference_active":true,"delta_db":-1.7272797728422697},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.8884156339998014e-06,"output_power":1.9758930714800093e-06,"reference_active":true,"delta_db":-2.940092422448748}],"active_band_count":3,"active_band_mae_db":1.9421373498809125}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":0.9045016303151814,"rms_envelope_nrmse":0.1216115636302024,"envelope_correlation":0.997519360716346,"rms_gain_error_db":-0.9634073066992772,"reference_mean_power":5.089623620868824e-05,"output_mean_power":4.0770396648491326e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":9.379392951970341e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.821321246572874e-05,"output_power":2.8943014429254795e-05,"reference_active":true,"delta_db":-1.2066978788374685},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5895785128125065e-06,"output_power":2.363288682469215e-06,"reference_active":true,"delta_db":-1.815266816642266},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2481135158983686e-07,"output_power":9.629811332249392e-08,"reference_active":true,"delta_db":-3.6820045816107823}],"active_band_count":3,"active_band_mae_db":2.234656425696839}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_8555-292519-0001","method":"magphase","status":"measured","mos":{"status":"measured","value":3.095327377319336,"predicted_mos":3.095327377319336,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":287046,"input_sample_rate":48000,"input_seconds":5.980125,"input_float32_sha256":"19ede94aaba1fc54c49424570cff01618f340dca41ef5a268c20557c1c60589a","internal_samples":95682,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6335132339991105},"started_at":"2026-10-13T05:19:16.402730+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.009968320999178104,"native_input_samples":288000,"native_input_sha256":"35ebe1d00edcf8cf04b5d8af278cf4900ac406567cca026c2959930c9df31957","output_sha256":"d84c4679904a34467a6a07de823fac567043cc207dd011e131d61e63110575ea","output_samples":287046,"length_delta_samples":-954,"analysis_s":3.8936559220001072,"synthesis_s":1.083999025000594,"total_s":4.977654947000701,"rtf":0.8296091578334502,"output_peak":0.31624926765380207,"output_rms":0.06921336967816775,"finished_at":"2026-10-13T05:19:21.403346+00:00","evaluation_input":{"reference":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"output_sha256":"d84c4679904a34467a6a07de823fac567043cc207dd011e131d61e63110575ea","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95682,"length_delta_samples":-318,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.194147989768524,"output_peak":0.3161311628182663},"common_prefix":{"samples":95682,"sample_rate":16000,"reference_coverage":0.9966875,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95682,"output_samples":95682,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.186134104187803,"spectral_convergence":0.263090534494903,"spectral_convergence_status":"ok","amplitude_floor":1.7050603882585003e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.667156752246884,"spectral_convergence":0.12596752734612185,"spectral_convergence_status":"ok","amplitude_floor":1.5194442893884604e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.681068733424819,"spectral_convergence":0.1440144902187891,"spectral_convergence_status":"ok","amplitude_floor":1.5327487413055003e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004798371358968865,"output_power":0.00044307756236292717,"delta_db":-0.34614098429597756,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.004427519406708295,"output_power":0.0038858484497345655,"delta_db":-0.5667461460030397,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00032591808771255073,"output_power":0.00030555808053314845,"delta_db":-0.28014690293967587,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":8.544453453256584e-05,"output_power":8.23143409842326e-05,"delta_db":-0.1620878272191635,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":2.6431369001707484e-05,"output_power":3.0777356585732215e-05,"delta_db":0.661116584750874,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":4.5060094163598636e-05,"output_power":4.5178901761279824e-05,"delta_db":0.011435746456489515,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.33794569861087004,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5102053499803562,"pitch":null,"output_rms_dbfs":-23.194147989768524,"output_peak":0.3161311628182663,"reference_rms_dbfs":-22.683942639788167,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.8447865299531685},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95682,"output_samples":95682,"intelligibility":{"estoi":{"value":0.9755266257947364,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.07341805382884788,"rms_dbfs":-22.683942639788167,"peak":0.36602783203125,"peak_dbfs":-8.729717809159148,"crest_factor_db":13.95422483062902,"mean":1.0562599389285602e-05,"zero_energy":false},"output":{"rms":0.06922972404944198,"rms_dbfs":-23.194147989768524,"peak":0.3161311628182663,"peak_dbfs":-10.002653824243614,"crest_factor_db":13.191494165524908,"mean":0.0018324577265006964,"zero_energy":false},"rms_difference_db":-0.5102053499803567},"full_scale":{"reference":{"sample_count":95682,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95682,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":598,"uncovered_tail_samples":2,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":5.718888771953062e-06,"threshold_dbfs":-52.426883500693044,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":204,"quiet_samples":32640,"reference_mean_power":5.475062518563195e-07,"output_mean_power":1.3427080995976327e-06,"output_rms_dbfs":-58.72018391287196,"output_zero_energy":false,"output_peak":0.03806303266085432,"output_energy_fraction":9.556883967421403e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":591,"reference_estimated_voiced_frames":327,"output_estimated_voiced_frames":303,"both_estimated_voiced_frames":301,"reference_voiced_output_unvoiced_frames":26,"reference_unvoiced_output_voiced_frames":2,"paired_voiced_fraction":0.9204892966360856,"absolute_difference_cents_median":3.0690080022493587,"absolute_difference_cents_p95":15.43431645117902,"difference_over_600_cents_frames":0,"uncovered_tail_samples":2,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95682,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1195,"rms_floor":0.00023914198234423542,"log_rms_mae_db":0.8511801331673917,"rms_envelope_nrmse":0.07229805103348759,"envelope_correlation":0.9981809652825726,"rms_gain_error_db":-0.5087267119773018,"reference_mean_power":0.005384255509451421,"output_mean_power":0.004789089903276186},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":2.076222600109883e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.000451696694839285,"output_power":0.0004253350528411505,"reference_active":true,"delta_db":-0.2611573735074285},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.417244314417332e-05,"output_power":6.575158693182706e-05,"reference_active":true,"delta_db":0.10557665231081675},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.6955905134141692e-05,"output_power":1.834872070056477e-05,"reference_active":true,"delta_db":0.34284811914477326}],"active_band_count":3,"active_band_mae_db":0.2365273816543395}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1189,"rms_floor":0.00022747909354312384,"log_rms_mae_db":0.6906081373328262,"rms_envelope_nrmse":0.06408391522935285,"envelope_correlation":0.9991737503977601,"rms_gain_error_db":-0.507015930141448,"reference_mean_power":0.005352039458833999,"output_mean_power":0.004762310564887578},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":1.9694227998108177e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004062162876123142,"output_power":0.0003860735194147783,"reference_active":true,"delta_db":-0.22087318241806825},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.7513677844927834e-05,"output_power":2.9281560007932157e-05,"reference_active":true,"delta_db":0.27045562668481754},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0270541215592872e-06,"output_power":1.165957165562242e-06,"reference_active":true,"delta_db":0.5508926604386568}],"active_band_count":3,"active_band_mae_db":0.3474071565138475}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1195,"rms_floor":9.442623236865579e-05,"log_rms_mae_db":1.0166848549346876,"rms_envelope_nrmse":0.08277504763758296,"envelope_correlation":0.9952931168585442,"rms_gain_error_db":-0.27918214850635703,"reference_mean_power":0.0007991995957000102,"output_mean_power":0.0007494402961382551},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":4.939843941105831e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00021920818041865553,"output_power":0.00021593656086797155,"reference_active":true,"delta_db":-0.06530576868053643},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.3547848676834422e-05,"output_power":2.152600029796932e-05,"reference_active":true,"delta_db":-0.3898789433503983},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.505236203928357e-06,"output_power":4.419452322796609e-06,"reference_active":true,"delta_db":-0.9540750417525755}],"active_band_count":3,"active_band_mae_db":0.4697532512611701}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1189,"rms_floor":8.226666373654229e-05,"log_rms_mae_db":0.8528299713751747,"rms_envelope_nrmse":0.06978170240844823,"envelope_correlation":0.9968093345411955,"rms_gain_error_db":-0.27523036107694593,"reference_mean_power":0.0008011726158427179,"output_mean_power":0.0007519744081263153},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":4.725477390528092e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00020740750412184927,"output_power":0.00020437911813151618,"reference_active":true,"delta_db":-0.06387944462004788},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0908198707461246e-05,"output_power":1.005477061000413e-05,"reference_active":true,"delta_db":-0.35380873566147136},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.1340311950632206e-07,"output_power":3.794524818695994e-07,"reference_active":true,"delta_db":-0.3721635171027085}],"active_band_count":3,"active_band_mae_db":0.2632838991280759}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1195,"rms_floor":3.9464514079665034e-05,"log_rms_mae_db":1.0262286938498033,"rms_envelope_nrmse":0.08328956294198772,"envelope_correlation":0.995176418055078,"rms_gain_error_db":-0.12118382994231526,"reference_mean_power":7.596545004965832e-05,"output_mean_power":7.387504052433418e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":5.50043923034189e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.0433224013748897e-05,"output_power":2.9368586759823628e-05,"reference_active":true,"delta_db":-0.15464914359045542},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.333766451514268e-06,"output_power":3.596054963188986e-06,"reference_active":true,"delta_db":0.3289114913449981},{"low_hz":20.0,"high_hz":50.0,"reference_power":9.953936669799076e-07,"output_power":1.072573423122103e-06,"reference_active":true,"delta_db":0.3243215825379523}],"active_band_count":3,"active_band_mae_db":0.2692940724911353}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1189,"rms_floor":3.608084029396229e-05,"log_rms_mae_db":0.687987637151973,"rms_envelope_nrmse":0.04685971192435195,"envelope_correlation":0.9984885674261943,"rms_gain_error_db":-0.121450292590798,"reference_mean_power":7.634717747822228e-05,"output_mean_power":7.42417083549034e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":5.191096401276866e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":2.8838947912882997e-05,"output_power":2.7890779465289564e-05,"reference_active":true,"delta_db":-0.14518760947085657},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.5954324929511188e-06,"output_power":1.6670701435235543e-06,"reference_active":true,"delta_db":0.1907544080763461},{"low_hz":20.0,"high_hz":50.0,"reference_power":8.719818917090136e-08,"output_power":7.386275634627956e-08,"reference_active":true,"delta_db":-0.7208195586871297}],"active_band_count":3,"active_band_mae_db":0.35225385874477744}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1195,"rms_floor":6.198647979957049e-05,"log_rms_mae_db":0.9229310284184459,"rms_envelope_nrmse":0.22863391308156458,"envelope_correlation":0.9716639253225205,"rms_gain_error_db":0.38272818517529394,"reference_mean_power":5.080858508072561e-05,"output_mean_power":5.548938600685435e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1195,"frequency_resolution_hz":0.16736401673640167,"power_floor":9.717581411601351e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.1432532279266865e-05,"output_power":4.63064828887591e-05,"reference_active":true,"delta_db":0.4830031894515815},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.607157307702109e-06,"output_power":6.664580999493254e-06,"reference_active":true,"delta_db":0.03758203714601174},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.857274566415924e-06,"output_power":7.628219084162458e-06,"reference_active":true,"delta_db":2.961426039085306}],"active_band_count":3,"active_band_mae_db":1.1606704218942998}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95680,"uncovered_tail_samples":2,"metrics":{"frame_count":1189,"rms_floor":5.0010568231240365e-05,"log_rms_mae_db":0.7060517026478241,"rms_envelope_nrmse":0.135437477718423,"envelope_correlation":0.9905341404783327,"rms_gain_error_db":0.3826640862192191,"reference_mean_power":5.106479187052956e-05,"output_mean_power":5.576837303995362e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1189,"frequency_resolution_hz":0.16820857863751051,"power_floor":9.395357115540106e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.8229242037948375e-05,"output_power":4.5296464921059104e-05,"reference_active":true,"delta_db":0.7366862217324611},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5444118614514788e-06,"output_power":3.4109946051423895e-06,"reference_active":true,"delta_db":-0.1666314873910028},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.2411464962053866e-07,"output_power":4.1787820090276556e-07,"reference_active":true,"delta_db":2.705794704312215}],"active_band_count":3,"active_band_mae_db":1.2030374711452263}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},{"case":"librispeech_908-157963-0018","method":"hnm","status":"measured","mos":{"status":"measured","value":2.2057104110717773,"predicted_mos":2.2057104110717773,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68080,"input_sample_rate":16000,"input_seconds":4.255,"input_float32_sha256":"b0c83e6673802f5bc7248822e2e7313ee110fab69b107d6b0faf27768fcc60e0","internal_samples":68080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6223218710001674},"started_at":"2026-10-13T04:40:36.232429+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.409699962096056e-05,"native_input_samples":68080,"native_input_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","output_sha256":"54bd59a355c4c53a54c6d919f31188d1c28554b2566d658bd991e0a63407024a","output_samples":68080,"length_delta_samples":0,"analysis_s":0.7208783300002324,"synthesis_s":0.34908152200023324,"total_s":1.0699598520004656,"rtf":0.2514594246769602,"output_peak":0.506013110814265,"output_rms":0.06569296682773537,"finished_at":"2026-10-13T04:40:37.307057+00:00","evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"54bd59a355c4c53a54c6d919f31188d1c28554b2566d658bd991e0a63407024a","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.804479117841189,"spectral_convergence":0.2970584457240224,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.47056228912991,"spectral_convergence":0.27034541455278216,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.880766991600836,"spectral_convergence":0.2851154362342357,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005891590783391547,"delta_db":0.35234684544527906,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.002877217648007006,"delta_db":-0.22277490975022446,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.0003885192962278006,"delta_db":0.07572034390999513,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.0002401499856143222,"delta_db":-0.23149926125716902,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00013681105883265487,"delta_db":-0.011444035630998215,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":8.370882360900176e-05,"delta_db":0.04182187209041764,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15593454468068058,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11080268170198138,"pitch":null,"output_rms_dbfs":-23.64962248137868,"output_peak":0.506013110814265,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.718602799523978},"common_prefix":{"samples":68080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.804479117841189,"spectral_convergence":0.2970584457240224,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.47056228912991,"spectral_convergence":0.27034541455278216,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.880766991600836,"spectral_convergence":0.2851154362342357,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005891590783391547,"delta_db":0.35234684544527906,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.002877217648007006,"delta_db":-0.22277490975022446,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.0003885192962278006,"delta_db":0.07572034390999513,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.0002401499856143222,"delta_db":-0.23149926125716902,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00013681105883265487,"delta_db":-0.011444035630998215,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":8.370882360900176e-05,"delta_db":0.04182187209041764,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.15593454468068058,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.11080268170198138,"pitch":null,"output_rms_dbfs":-23.64962248137868,"output_peak":0.506013110814265,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.718602799523978},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68080,"output_samples":68080,"intelligibility":{"estoi":{"value":0.8228520690229705,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06653635567215388,"rms_dbfs":-23.538819799676702,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74685731684548,"mean":-0.000349972250879021,"zero_energy":false},"output":{"rms":0.06569296682773537,"rms_dbfs":-23.64962248137868,"peak":0.506013110814265,"peak_dbfs":-5.916764608637904,"crest_factor_db":17.73285787274078,"mean":-0.00014707298323956842,"zero_energy":false},"rms_difference_db":-0.11080268170197982},"full_scale":{"reference":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":5.205620713023158e-06,"output_rms_dbfs":-52.835274782248376,"output_zero_energy":false,"output_peak":0.008748864811723376,"output_energy_fraction":7.087208575377727e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":80,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.9420289855072463,"absolute_difference_cents_median":2.0313070090409107,"absolute_difference_cents_p95":8.45226256183356,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00026228408676881105,"log_rms_mae_db":1.0433985026962824,"rms_envelope_nrmse":0.09834228021237355,"envelope_correlation":0.9899635252165004,"rms_gain_error_db":-0.11080196876063415,"reference_mean_power":0.004432288013980724,"output_mean_power":0.004320636961951346},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":1.6030570640082068e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008005952323264084,"output_power":0.0007786170798075116,"reference_active":true,"delta_db":-0.1208907327751081},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020547999455019126,"output_power":0.00019216835581724591,"reference_active":true,"delta_db":-0.2908767117841441},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.989786705158196e-05,"output_power":4.387088496837015e-05,"reference_active":true,"delta_db":0.41226716297491356}],"active_band_count":3,"active_band_mae_db":0.2746782025113886}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00021831174911177267,"log_rms_mae_db":0.512862527560923,"rms_envelope_nrmse":0.046031471213047945,"envelope_correlation":0.9977428305400463,"rms_gain_error_db":-0.11081174361184701,"reference_mean_power":0.004463740704660867,"output_mean_power":0.00435128755352969},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":1.4245155353761532e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007405152195784611,"output_power":0.0007253684755731569,"reference_active":true,"delta_db":-0.08975311599543408},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001021109367695941,"output_power":9.624386132865962e-05,"reference_active":true,"delta_db":-0.25699221755995627},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.434051823082556e-06,"output_power":2.1600770784506757e-06,"reference_active":true,"delta_db":-0.5186057205819797}],"active_band_count":3,"active_band_mae_db":0.2884503513791234}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00023386465365005232,"log_rms_mae_db":1.6370787415239976,"rms_envelope_nrmse":0.12924924292266143,"envelope_correlation":0.9864334631175399,"rms_gain_error_db":-0.20969099581785516,"reference_mean_power":0.002112021838569365,"output_mean_power":0.002012469489928595},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":9.681983603336735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048709120906038954,"output_power":0.00047587850272749223,"reference_active":true,"delta_db":-0.10114205088857015},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014674692890756196,"output_power":0.00012612597968621883,"reference_active":true,"delta_db":-0.6576446853094691},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.836155879757368e-05,"output_power":2.8341898279527322e-05,"reference_active":true,"delta_db":-0.0030116170269624985}],"active_band_count":3,"active_band_mae_db":0.2539327844083339}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00020000043006212289,"log_rms_mae_db":0.9635793689433428,"rms_envelope_nrmse":0.06606672139386427,"envelope_correlation":0.9965667333471763,"rms_gain_error_db":-0.20969571164059025,"reference_mean_power":0.002127032769351675,"output_mean_power":0.0020267706640704057},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.787606965551091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004705886462934124,"output_power":0.00045992868957920756,"reference_active":true,"delta_db":-0.09950943935406485},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.123456963821095e-05,"output_power":6.137963799976097e-05,"reference_active":true,"delta_db":-0.6466648211303561},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5465229502268847e-06,"output_power":1.3764636281984724e-06,"reference_active":true,"delta_db":-0.5059162922472624}],"active_band_count":3,"active_band_mae_db":0.41736351757722784}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":1.6269788678123351,"rms_envelope_nrmse":0.17787352914383361,"envelope_correlation":0.9751768133648258,"rms_gain_error_db":-0.22801923870798918,"reference_mean_power":0.0003312880907167699,"output_mean_power":0.00031434307705408313},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":2.2485323271997548e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014154142624946066,"output_power":0.0001340974724815565,"reference_active":true,"delta_db":-0.2346297529522971},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6917189688069e-05,"output_power":2.5575105535884937e-05,"reference_active":true,"delta_db":-0.22212280528824538},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.319943953798976e-06,"output_power":1.0077650806435087e-05,"reference_active":true,"delta_db":2.0264607893737265}],"active_band_count":3,"active_band_mae_db":0.8277377825380897}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":1.1731673946672736,"rms_envelope_nrmse":0.10225832564136414,"envelope_correlation":0.9916317180215851,"rms_gain_error_db":-0.22802228599264007,"reference_mean_power":0.0003336427865109947,"output_mean_power":0.0003165771106731654},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":2.0170856876139243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013578457159768304,"output_power":0.00013062769789801777,"reference_active":true,"delta_db":-0.1681515324130627},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3100830411180002e-05,"output_power":1.1929580443912116e-05,"reference_active":true,"delta_db":-0.4067365471655638},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732875286933278e-07,"output_power":4.851172800881365e-07,"reference_active":true,"delta_db":1.1380326400235856}],"active_band_count":3,"active_band_mae_db":0.5709735732007374}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":1.5607898735746344,"rms_envelope_nrmse":0.22381507623018182,"envelope_correlation":0.9664720517715056,"rms_gain_error_db":0.09177227945736591,"reference_mean_power":8.138813397574016e-05,"output_mean_power":8.312627493243485e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":8.567729341398194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0300307346839755e-05,"output_power":4.107447912481357e-05,"reference_active":true,"delta_db":0.08263706487575648},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8095151025280505e-06,"output_power":5.192979580494817e-06,"reference_active":true,"delta_db":-0.48723270150406395},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.571956353556119e-06,"output_power":4.208096110863654e-06,"reference_active":true,"delta_db":0.7117950681476974}],"active_band_count":3,"active_band_mae_db":0.42722161150917265}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":1.2872431721173514,"rms_envelope_nrmse":0.1516251610192266,"envelope_correlation":0.9839140509643797,"rms_gain_error_db":0.09176896226389537,"reference_mean_power":8.196654666810836e-05,"output_mean_power":8.37169763756823e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.099518469769931e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9096129724567364e-05,"output_power":4.022494824119908e-05,"reference_active":true,"delta_db":0.12361726898346637},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.153274768568064e-06,"output_power":2.718447369936689e-06,"reference_active":true,"delta_db":-0.64440886413697},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4027458558124966e-07,"output_power":1.7562754666651118e-07,"reference_active":true,"delta_db":0.9761364040664484}],"active_band_count":3,"active_band_mae_db":0.5813875123956282}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-157963-0018","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.6532111167907715,"predicted_mos":3.6532111167907715,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68080,"input_sample_rate":16000,"input_seconds":4.255,"input_float32_sha256":"7d43bc910bf1b881641d65ac43efffe557013552041ec51ee198ad4e3eaffa4c","internal_samples":68080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6023232389998157},"started_at":"2026-10-13T04:43:52.784281+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.3474000045098364e-05,"native_input_samples":68080,"native_input_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","output_sha256":"da2c27a46c09e0614bc5d54f3647137718fa89206d676501a974fb52fc9b5077","output_samples":68080,"length_delta_samples":0,"analysis_s":1.2766432479993455,"synthesis_s":0.05930698000065604,"total_s":1.3359502280000015,"rtf":0.31397185146886053,"output_peak":0.5799483117918864,"output_rms":0.07553625443818679,"finished_at":"2026-10-13T04:43:54.124963+00:00","evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"da2c27a46c09e0614bc5d54f3647137718fa89206d676501a974fb52fc9b5077","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.740078264417843,"spectral_convergence":0.6888821176952689,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.357202134535691,"spectral_convergence":0.33788467692168644,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.811042200539304,"spectral_convergence":0.33854397195167735,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005715919958868672,"delta_db":0.2208824909361694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.004162907184475151,"delta_db":1.381465215865648,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.00046262422341705663,"delta_db":0.8338781009336831,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.0002831982809663936,"delta_db":0.4845812174228896,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00014719153688400033,"delta_db":0.3061723231684152,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":7.821251292102452e-05,"delta_db":-0.2531281172917879,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5800179109364322,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1019287192349538,"pitch":null,"output_rms_dbfs":-22.436891080441747,"output_peak":0.5799483117918864,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.969440866497613},"common_prefix":{"samples":68080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.740078264417843,"spectral_convergence":0.6888821176952689,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.357202134535691,"spectral_convergence":0.33788467692168644,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.811042200539304,"spectral_convergence":0.33854397195167735,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005715919958868672,"delta_db":0.2208824909361694,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.004162907184475151,"delta_db":1.381465215865648,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.00046262422341705663,"delta_db":0.8338781009336831,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.0002831982809663936,"delta_db":0.4845812174228896,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00014719153688400033,"delta_db":0.3061723231684152,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":7.821251292102452e-05,"delta_db":-0.2531281172917879,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5800179109364322,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1019287192349538,"pitch":null,"output_rms_dbfs":-22.436891080441747,"output_peak":0.5799483117918864,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.969440866497613},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68080,"output_samples":68080,"intelligibility":{"estoi":{"value":0.8911805860455363,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06653635567215388,"rms_dbfs":-23.538819799676702,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74685731684548,"mean":-0.000349972250879021,"zero_energy":false},"output":{"rms":0.07553625443818679,"rms_dbfs":-22.436891080441747,"peak":0.5799483117918864,"peak_dbfs":-4.73221422887485,"crest_factor_db":17.704676851566898,"mean":7.212613063899043e-08,"zero_energy":false},"rms_difference_db":1.1019287192349552},"full_scale":{"reference":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":3.2349837475127717e-06,"output_rms_dbfs":-54.901278968757296,"output_zero_energy":false,"output_peak":0.00827777414266122,"output_energy_fraction":3.331207268370884e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":84,"both_estimated_voiced_frames":63,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":21,"paired_voiced_fraction":0.9130434782608695,"absolute_difference_cents_median":8.293976791669918,"absolute_difference_cents_p95":1205.500069564144,"difference_over_600_cents_frames":7,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00026228408676881105,"log_rms_mae_db":1.782297436777928,"rms_envelope_nrmse":0.22190546542621048,"envelope_correlation":0.9733331899620661,"rms_gain_error_db":1.1019348726305689,"reference_mean_power":0.004432288013980724,"output_mean_power":0.005712437490363156},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":1.6030570640082068e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008005952323264084,"output_power":0.0009878481412039515,"reference_active":true,"delta_db":0.9127718750713525},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020547999455019126,"output_power":0.00024993318923993046,"reference_active":true,"delta_db":0.850543854493525},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.989786705158196e-05,"output_power":3.908672050299037e-05,"reference_active":true,"delta_db":-0.08920445524819422}],"active_band_count":3,"active_band_mae_db":0.6175067282710239}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00021831174911177267,"log_rms_mae_db":1.2093509246908989,"rms_envelope_nrmse":0.1832365743857777,"envelope_correlation":0.9862196843337115,"rms_gain_error_db":1.101978085407562,"reference_mean_power":0.004463740704660867,"output_mean_power":0.0057530317040873705},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":1.4245155353761532e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007405152195784611,"output_power":0.0009173689374225543,"reference_active":true,"delta_db":0.9301004174546946},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001021109367695941,"output_power":0.0001262282709120225,"reference_active":true,"delta_db":0.9208437280013515},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.434051823082556e-06,"output_power":1.9874980955100197e-06,"reference_active":true,"delta_db":-0.8802309932753458}],"active_band_count":3,"active_band_mae_db":0.9103917129104641}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00023386465365005232,"log_rms_mae_db":2.2932703807924804,"rms_envelope_nrmse":0.26870644972307384,"envelope_correlation":0.9753541382732931,"rms_gain_error_db":1.5011378128677133,"reference_mean_power":0.002112021838569365,"output_mean_power":0.002984091845127744},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":9.681983603336735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048709120906038954,"output_power":0.0006561575577477238,"reference_active":true,"delta_db":1.2939784391067932},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014674692890756196,"output_power":0.00019555766660576398,"reference_active":true,"delta_db":1.2470582544000162},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.836155879757368e-05,"output_power":2.438418230610503e-05,"reference_active":true,"delta_db":-0.6562190006619784}],"active_band_count":3,"active_band_mae_db":1.0657518980562626}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00020000043006212289,"log_rms_mae_db":1.8071051910753908,"rms_envelope_nrmse":0.233748041981604,"envelope_correlation":0.9864202736908187,"rms_gain_error_db":1.5011351125676677,"reference_mean_power":0.002127032769351675,"output_mean_power":0.00300529903569776},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.787606965551091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004705886462934124,"output_power":0.0006303786594246891,"reference_active":true,"delta_db":1.2696005739193184},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.123456963821095e-05,"output_power":9.420976595912155e-05,"reference_active":true,"delta_db":1.2140512003129849},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5465229502268847e-06,"output_power":1.3123985840136428e-06,"reference_active":true,"delta_db":-0.7129061612748601}],"active_band_count":3,"active_band_mae_db":1.0655193118357211}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":1.496666667517833,"rms_envelope_nrmse":0.16540447043077244,"envelope_correlation":0.9818107865136507,"rms_gain_error_db":0.4475692513343302,"reference_mean_power":0.0003312880907167699,"output_mean_power":0.00036725079880600486},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":2.2485323271997548e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014154142624946066,"output_power":0.00015729804953705303,"reference_active":true,"delta_db":0.4583977001387476},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6917189688069e-05,"output_power":2.8394607257190634e-05,"reference_active":true,"delta_db":0.23206151037583855},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.319943953798976e-06,"output_power":4.6841408466443895e-06,"reference_active":true,"delta_db":-1.3008328158704239}],"active_band_count":3,"active_band_mae_db":0.6637640087950033}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":0.8642887687698052,"rms_envelope_nrmse":0.09219564836867314,"envelope_correlation":0.9954433381837302,"rms_gain_error_db":0.44756873188083157,"reference_mean_power":0.0003336427865109947,"output_mean_power":0.000369861062440277},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":2.0170856876139243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013578457159768304,"output_power":0.00014942610311877527,"reference_active":true,"delta_db":0.4157604422023268},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3100830411180002e-05,"output_power":1.382065316474478e-05,"reference_active":true,"delta_db":0.23229743518944423},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732875286933278e-07,"output_power":2.458342113867624e-07,"reference_active":true,"delta_db":-1.8140115954129068}],"active_band_count":3,"active_band_mae_db":0.8206898242682259}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":1.3806541804090482,"rms_envelope_nrmse":0.2252810592162257,"envelope_correlation":0.9650620090976693,"rms_gain_error_db":-0.2605691692829371,"reference_mean_power":8.138813397574016e-05,"output_mean_power":7.664859075075877e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":8.567729341398194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0300307346839755e-05,"output_power":3.840466628042467e-05,"reference_active":true,"delta_db":-0.20924362627121929},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8095151025280505e-06,"output_power":4.883204728461904e-06,"reference_active":true,"delta_db":-0.7543495257239462},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.571956353556119e-06,"output_power":2.060605663552436e-06,"reference_active":true,"delta_db":-2.389112543733313}],"active_band_count":3,"active_band_mae_db":1.1175685652428262}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":0.7793188887864189,"rms_envelope_nrmse":0.12591457901075911,"envelope_correlation":0.9891191990548109,"rms_gain_error_db":-0.26057038538745414,"reference_mean_power":8.196654666810836e-05,"output_mean_power":7.719329863786268e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.099518469769931e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9096129724567364e-05,"output_power":3.688451153874255e-05,"reference_active":true,"delta_db":-0.25289730577157615},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.153274768568064e-06,"output_power":2.534295654147701e-06,"reference_active":true,"delta_db":-0.9490453689582689},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4027458558124966e-07,"output_power":8.072792858157844e-08,"reference_active":true,"delta_db":-2.3995518537006664}],"active_band_count":3,"active_band_mae_db":1.2004981761435038}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-157963-0018","method":"facodec","status":"measured","mos":{"status":"measured","value":3.8612873554229736,"predicted_mos":3.8612873554229736,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68200,"input_sample_rate":16000,"input_seconds":4.2625,"input_float32_sha256":"eed533c37957d9c99cda3bb293655a3647dfb94794e0c8f55c4a389ca453b9a0","internal_samples":68200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.633944468001573},"started_at":"2026-10-13T07:02:50.009570+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":8.992599941848312e-05,"native_input_samples":68080,"native_input_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","output_sha256":"ba9a8eebd0dff24de3fa7bf96fa9d48934fab73d27788d8ab18b1295412cf4c0","output_samples":68200,"length_delta_samples":120,"analysis_s":1.0826185219993931,"synthesis_s":8.697521660998973,"total_s":9.780140182998366,"rtf":2.298505330904434,"output_peak":0.48882973194122314,"output_rms":0.06264809653238645,"finished_at":"2026-10-13T07:02:59.802438+00:00","evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"ba9a8eebd0dff24de3fa7bf96fa9d48934fab73d27788d8ab18b1295412cf4c0","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":68080,"output_samples":68200,"length_delta_samples":120,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.061842397000586,"output_peak":0.48882973194122314},"common_prefix":{"samples":68080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9982404692082112,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.620184270809427,"spectral_convergence":0.2607808179220246,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.995018418402875,"spectral_convergence":0.2713049543947034,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":7.243400245410916,"spectral_convergence":0.29406746511719284,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005491678896018767,"delta_db":0.04707246114122218,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.002736308537672631,"delta_db":-0.4408514093728877,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.00029711488563587764,"delta_db":-1.0891614821757054,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.00017612035919410029,"delta_db":-1.5782292856153326,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00010657892680299009,"delta_db":-1.0959426484567545,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":6.641129088692673e-05,"delta_db":-0.9634712942791221,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8691214301735041,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5153743877798586,"pitch":null,"output_rms_dbfs":-24.05419418745656,"output_peak":0.48882973194122314,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.952867644874407},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68080,"output_samples":68080,"intelligibility":{"estoi":{"value":0.8472327467584168,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06653635567215388,"rms_dbfs":-23.538819799676702,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74685731684548,"mean":-0.000349972250879021,"zero_energy":false},"output":{"rms":0.0627032845215815,"rms_dbfs":-24.05419418745656,"peak":0.48882973194122314,"peak_dbfs":-6.216847740035187,"crest_factor_db":17.837346447421375,"mean":8.547777782521919e-06,"zero_energy":false},"rms_difference_db":-0.5153743877798576},"full_scale":{"reference":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":4.3490043993768345e-06,"output_rms_dbfs":-53.61610153009735,"output_zero_energy":false,"output_peak":0.006997246295213699,"output_energy_fraction":6.499047482837096e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":71,"both_estimated_voiced_frames":61,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":10,"paired_voiced_fraction":0.8840579710144928,"absolute_difference_cents_median":2.9868565766937762,"absolute_difference_cents_p95":7.954142650473,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00026228408676881105,"log_rms_mae_db":1.3147254364005923,"rms_envelope_nrmse":0.12143908405992841,"envelope_correlation":0.9874849188294854,"rms_gain_error_db":-0.5153722629085978,"reference_mean_power":0.004432288013980724,"output_mean_power":0.003936323175618067},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":1.6030570640082068e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008005952323264084,"output_power":0.0007000431020857935,"reference_active":true,"delta_db":-0.5828821886644091},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020547999455019126,"output_power":0.00018241064813482756,"reference_active":true,"delta_db":-0.5171935906759927},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.989786705158196e-05,"output_power":3.463573143055011e-05,"reference_active":true,"delta_db":-0.6142531541742551}],"active_band_count":3,"active_band_mae_db":0.571442977838219}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00021831174911177267,"log_rms_mae_db":1.0496753088115993,"rms_envelope_nrmse":0.10959736982517082,"envelope_correlation":0.9896816324658885,"rms_gain_error_db":-0.5153591523242672,"reference_mean_power":0.004463740704660867,"output_mean_power":0.003964268335741125},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":1.4245155353761532e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007405152195784611,"output_power":0.0006472724568659343,"reference_active":true,"delta_db":-0.5844686182175884},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001021109367695941,"output_power":8.700006872549936e-05,"reference_active":true,"delta_db":-0.6955266475583233},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.434051823082556e-06,"output_power":2.295690178908547e-06,"reference_active":true,"delta_db":-0.254165442086445}],"active_band_count":3,"active_band_mae_db":0.5113869026207856}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00023386465365005232,"log_rms_mae_db":1.9162422920334885,"rms_envelope_nrmse":0.16095185386256328,"envelope_correlation":0.984087168793258,"rms_gain_error_db":-0.7974992568012073,"reference_mean_power":0.002112021838569365,"output_mean_power":0.0017577150806862027},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":9.681983603336735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048709120906038954,"output_power":0.0003930662425173354,"reference_active":true,"delta_db":-0.9314454441154614},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014674692890756196,"output_power":0.0001240624194747143,"reference_active":true,"delta_db":-0.7292877462514049},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.836155879757368e-05,"output_power":2.4357708856117644e-05,"reference_active":true,"delta_db":-0.660936615880784}],"active_band_count":3,"active_band_mae_db":0.7738899354158835}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00020000043006212289,"log_rms_mae_db":1.4868810691498042,"rms_envelope_nrmse":0.14809921529075215,"envelope_correlation":0.9868199426067585,"rms_gain_error_db":-0.7975032267211618,"reference_mean_power":0.002127032769351675,"output_mean_power":0.0017702062023876464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.787606965551091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004705886462934124,"output_power":0.00038091565005789755,"reference_active":true,"delta_db":-0.9181262874069358},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.123456963821095e-05,"output_power":5.864125436618442e-05,"reference_active":true,"delta_db":-0.8448755335512286},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5465229502268847e-06,"output_power":1.4596761795291811e-06,"reference_active":true,"delta_db":-0.25099848444050477}],"active_band_count":3,"active_band_mae_db":0.6713334351328898}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":1.7727135850062428,"rms_envelope_nrmse":0.21077387555854213,"envelope_correlation":0.9801076019124126,"rms_gain_error_db":-1.4210616327931633,"reference_mean_power":0.0003312880907167699,"output_mean_power":0.00023883592940437718},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":2.2485323271997548e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014154142624946066,"output_power":0.00011105251937761959,"reference_active":true,"delta_db":-1.0535515192402507},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6917189688069e-05,"output_power":2.0024694881247578e-05,"reference_active":true,"delta_db":-1.2846380762854002},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.319943953798976e-06,"output_power":6.206004823452836e-06,"reference_active":true,"delta_db":-0.0790111811937768}],"active_band_count":3,"active_band_mae_db":0.8057335922398092}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":1.445587102015797,"rms_envelope_nrmse":0.1936318511937257,"envelope_correlation":0.9854284133304697,"rms_gain_error_db":-1.421066064786104,"reference_mean_power":0.0003336427865109947,"output_mean_power":0.0002405332576649005},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":2.0170856876139243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013578457159768304,"output_power":0.00010744137596870824,"reference_active":true,"delta_db":-1.016788648475214},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3100830411180002e-05,"output_power":1.002920975367535e-05,"reference_active":true,"delta_db":-1.1603211042021369},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732875286933278e-07,"output_power":3.5822183323782183e-07,"reference_active":true,"delta_db":-0.17891428553091349}],"active_band_count":3,"active_band_mae_db":0.7853413460694214}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":1.9181325676243413,"rms_envelope_nrmse":0.2318647747875131,"envelope_correlation":0.9653566667186105,"rms_gain_error_db":-0.7641728078300135,"reference_mean_power":8.138813397574016e-05,"output_mean_power":6.825646784608107e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":8.567729341398194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0300307346839755e-05,"output_power":3.105026293164615e-05,"reference_active":true,"delta_db":-1.1324307615723743},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8095151025280505e-06,"output_power":7.430547589136569e-06,"reference_active":true,"delta_db":1.0688093493197737},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.571956353556119e-06,"output_power":3.409550749449162e-06,"reference_active":true,"delta_db":-0.20208984434594127}],"active_band_count":3,"active_band_mae_db":0.8011099850793632}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":1.6039739989901007,"rms_envelope_nrmse":0.20081303486458707,"envelope_correlation":0.9742173126150686,"rms_gain_error_db":-0.7641827706336036,"reference_mean_power":8.196654666810836e-05,"output_mean_power":6.874139815424823e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.099518469769931e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9096129724567364e-05,"output_power":3.085900890535072e-05,"reference_active":true,"delta_db":-1.0275179327819075},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.153274768568064e-06,"output_power":4.132422002031854e-06,"reference_active":true,"delta_db":1.1744284942552472},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4027458558124966e-07,"output_power":1.5758460808047157e-07,"reference_active":true,"delta_db":0.5053480162033349}],"active_band_count":3,"active_band_mae_db":0.9024314810801631}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-157963-0018","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.6938207149505615,"predicted_mos":3.6938207149505615,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":187904,"input_sample_rate":44100,"input_seconds":4.260861678004535,"input_float32_sha256":"8a490bcd07947da6f1563d205818a034b5dd308a6861e979a92d5c48d811cd59","internal_samples":68174,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5798457419987244},"started_at":"2026-10-13T07:34:15.918409+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.009780290998605778,"native_input_samples":187646,"native_input_sha256":"23cc1180c409e015deb3c4d84324c3d09add56d117779c6183799cc575cf14d7","output_sha256":"a5d476ea05008ed94006694bef6ccaeeb578a598282d6a666ef644bd01134516","output_samples":187904,"length_delta_samples":258,"analysis_s":0.2010806420003064,"synthesis_s":26.392825854996772,"total_s":26.59390649699708,"rtf":6.250020125755792,"output_peak":0.5524000525474548,"output_rms":0.05979509485964713,"finished_at":"2026-10-13T07:34:42.542070+00:00","evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"a5d476ea05008ed94006694bef6ccaeeb578a598282d6a666ef644bd01134516","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":68080,"output_samples":68174,"length_delta_samples":94,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.46281033384766,"output_peak":0.5528392531439571},"common_prefix":{"samples":68080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9986211752280928,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.29578614267153,"spectral_convergence":0.4186745392531684,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.098355530264868,"spectral_convergence":0.36464978204492576,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.07429490966192,"spectral_convergence":0.4243814956251302,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005951501351987959,"delta_db":0.3962864703320313,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.002184183385042577,"delta_db":-1.4196110731841014,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.00036218598304625573,"delta_db":-0.22908920303414806,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.00025961526559319835,"delta_db":0.10697734670120049,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00012520553743572603,"delta_db":-0.3964207090361694,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":5.724758918738287e-05,"delta_db":-1.6083184909043045,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.692783882198659,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.9179995905313991,"pitch":null,"output_rms_dbfs":-24.4568193902081,"output_peak":0.5528392531439571,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.156145527532773},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68080,"output_samples":68080,"intelligibility":{"estoi":{"value":0.7250701815285014,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06653635567215388,"rms_dbfs":-23.538819799676702,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74685731684548,"mean":-0.000349972250879021,"zero_energy":false},"output":{"rms":0.05986307622820544,"rms_dbfs":-24.4568193902081,"peak":0.5528392531439571,"peak_dbfs":-5.148022568154872,"crest_factor_db":19.30879682205323,"mean":-0.00011938234962830922,"zero_energy":false},"rms_difference_db":-0.917999590531398},"full_scale":{"reference":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":2.5213911933229383e-06,"output_rms_dbfs":-55.98359768396784,"output_zero_energy":false,"output_peak":0.004735751304125664,"output_energy_fraction":4.133925206825138e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":89,"both_estimated_voiced_frames":52,"reference_voiced_output_unvoiced_frames":17,"reference_unvoiced_output_voiced_frames":37,"paired_voiced_fraction":0.7536231884057971,"absolute_difference_cents_median":9.508306113050523,"absolute_difference_cents_p95":42.5239771623261,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00026228408676881105,"log_rms_mae_db":1.9508531571103687,"rms_envelope_nrmse":0.14426537165614797,"envelope_correlation":0.9882724010039214,"rms_gain_error_db":-0.9179941779941367,"reference_mean_power":0.004432288013980724,"output_mean_power":0.0035878027273898412},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":1.6030570640082068e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008005952323264084,"output_power":0.0006722736838044877,"reference_active":true,"delta_db":-0.7586688838045088},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020547999455019126,"output_power":0.00015685812742617804,"reference_active":true,"delta_db":-1.1726251944044754},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.989786705158196e-05,"output_power":2.187220435935089e-05,"reference_active":true,"delta_db":-2.610571238304974}],"active_band_count":3,"active_band_mae_db":1.513955105504653}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00021831174911177267,"log_rms_mae_db":1.7688890571119145,"rms_envelope_nrmse":0.12457478471564229,"envelope_correlation":0.9938195806665003,"rms_gain_error_db":-0.9179578638215923,"reference_mean_power":0.004463740704660867,"output_mean_power":0.0036132929395705855},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":1.4245155353761532e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007405152195784611,"output_power":0.0006213276169896849,"reference_active":true,"delta_db":-0.7621333115117611},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001021109367695941,"output_power":7.904772876528353e-05,"reference_active":true,"delta_db":-1.1118286434039761},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.434051823082556e-06,"output_power":1.5184173133139633e-06,"reference_active":true,"delta_db":-2.0493867347153727}],"active_band_count":3,"active_band_mae_db":1.3077828965437033}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00023386465365005232,"log_rms_mae_db":3.2823605104581675,"rms_envelope_nrmse":0.1833220856605568,"envelope_correlation":0.9821788552982453,"rms_gain_error_db":-1.0926679205551157,"reference_mean_power":0.002112021838569365,"output_mean_power":0.001642221041944602},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":9.681983603336735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048709120906038954,"output_power":0.00041931473298630035,"reference_active":true,"delta_db":-0.6507016958487365},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014674692890756196,"output_power":0.00010357156424978279,"reference_active":true,"delta_db":-1.513284856993781},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.836155879757368e-05,"output_power":1.6026658485456478e-05,"reference_active":true,"delta_db":-2.478871141221922}],"active_band_count":3,"active_band_mae_db":1.547619231354813}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00020000043006212289,"log_rms_mae_db":2.8374914898179617,"rms_envelope_nrmse":0.15602271740147544,"envelope_correlation":0.9899863623089639,"rms_gain_error_db":-1.0926631337631767,"reference_mean_power":0.002127032769351675,"output_mean_power":0.0016538947453073102},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.787606965551091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004705886462934124,"output_power":0.0004020893054842311,"reference_active":true,"delta_db":-0.6831892274685452},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.123456963821095e-05,"output_power":4.8958529784357315e-05,"reference_active":true,"delta_db":-1.6286243731694736},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5465229502268847e-06,"output_power":1.2851162803436611e-06,"reference_active":true,"delta_db":-0.8041394383694498}],"active_band_count":3,"active_band_mae_db":1.0386510130024895}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":3.4301889733305706,"rms_envelope_nrmse":0.24146355600041747,"envelope_correlation":0.9552063080830379,"rms_gain_error_db":-0.22805265854754506,"reference_mean_power":0.0003312880907167699,"output_mean_power":0.00031434065812977754},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":2.2485323271997548e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014154142624946066,"output_power":0.00011039443215432884,"reference_active":true,"delta_db":-1.079363975818073},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6917189688069e-05,"output_power":2.5367228105023622e-05,"reference_active":true,"delta_db":-0.25756700887740674},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.319943953798976e-06,"output_power":5.530175496647452e-06,"reference_active":true,"delta_db":-0.5797431332542479}],"active_band_count":3,"active_band_mae_db":0.6388913726499093}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":3.0139620685754522,"rms_envelope_nrmse":0.20123056781667578,"envelope_correlation":0.9668576493853058,"rms_gain_error_db":-0.22804866371696247,"reference_mean_power":0.0003336427865109947,"output_mean_power":0.0003165751878860381},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":2.0170856876139243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013578457159768304,"output_power":0.00010558768889204106,"reference_active":true,"delta_db":-1.0923714229346235},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3100830411180002e-05,"output_power":1.3335062728361825e-05,"reference_active":true,"delta_db":0.07696238176612309},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732875286933278e-07,"output_power":2.811976168651442e-07,"reference_active":true,"delta_db":-1.2303184491649735}],"active_band_count":3,"active_band_mae_db":0.7998840846219067}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":3.4105937663761368,"rms_envelope_nrmse":0.24555368035943534,"envelope_correlation":0.9634958169345341,"rms_gain_error_db":-1.0321549585574872,"reference_mean_power":8.138813397574016e-05,"output_mean_power":6.417200304921761e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":8.567729341398194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0300307346839755e-05,"output_power":2.9391221595567952e-05,"reference_active":true,"delta_db":-1.3709072111255163},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8095151025280505e-06,"output_power":5.80668087540607e-06,"reference_active":true,"delta_db":-0.0021192638731301104},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.571956353556119e-06,"output_power":2.4860151314997638e-06,"reference_active":true,"delta_db":-1.574023758069899}],"active_band_count":3,"active_band_mae_db":0.9823500776895152}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":3.0825486364596393,"rms_envelope_nrmse":0.1994083439230626,"envelope_correlation":0.9786124637766794,"rms_gain_error_db":-1.0321469307591864,"reference_mean_power":8.196654666810836e-05,"output_mean_power":6.462818286464468e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.099518469769931e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9096129724567364e-05,"output_power":2.8701216434638514e-05,"reference_active":true,"delta_db":-1.342334633610478},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.153274768568064e-06,"output_power":3.2979654276586567e-06,"reference_active":true,"delta_db":0.19484282909313472},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4027458558124966e-07,"output_power":1.5626355659918087e-07,"reference_active":true,"delta_db":0.4687871041734055}],"active_band_count":3,"active_band_mae_db":0.6686548556256727}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-157963-0018","method":"golf","status":"measured","mos":{"status":"measured","value":3.080813407897949,"predicted_mos":3.080813407897949,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":102000,"input_sample_rate":24000,"input_seconds":4.25,"input_float32_sha256":"af283c467a624aeed8f37f4a44e9c9ffd90473509a400488516eb452cc5558c9","internal_samples":68000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.606104086000414},"started_at":"2026-10-13T05:03:13.238367+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0019668370005092584,"native_input_samples":102120,"native_input_sha256":"9c266bdfe6845ba8d6283df2e6adbab2a3aee6c68239697a665cf6e43107e2ac","output_sha256":"15a0d41cf755e4ebe869be57f5e39f139e1d1e166de141e87e1bade4b89b62c2","output_samples":102000,"length_delta_samples":-120,"analysis_s":0.4163587679995544,"synthesis_s":0.10246739700050966,"total_s":0.5188261650000641,"rtf":0.12193329377204797,"output_peak":0.4630403518676758,"output_rms":0.05777684541342641,"finished_at":"2026-10-13T05:03:13.766135+00:00","analysis_stages":{"external_f0_s":0.04613485100071557},"evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"15a0d41cf755e4ebe869be57f5e39f139e1d1e166de141e87e1bade4b89b62c2","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":68080,"output_samples":68000,"length_delta_samples":-80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.762484155450313,"output_peak":0.4634203893921101},"common_prefix":{"samples":68000,"sample_rate":16000,"reference_coverage":0.9988249118683902,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68000,"output_samples":68000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.059946943480824,"spectral_convergence":0.6196756268275146,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.469149130139179,"spectral_convergence":0.3223384443657294,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.790441020217034,"spectral_convergence":0.34008570210810607,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005438854969147737,"output_power":0.0004901851547994921,"delta_db":-0.45147323647769055,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030315663433911344,"output_power":0.0022155933113518823,"delta_db":-1.36177031474122,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038290368985406924,"output_power":0.00031634819049201196,"delta_db":-0.8292419680678409,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002535964473055393,"output_power":0.00017016734772155317,"delta_db":-1.7326693524650825,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013733323501891157,"output_power":8.569183353443652e-05,"delta_db":-2.0483621495021085,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.30040204126504e-05,"output_power":6.205352025596153e-05,"delta_db":-1.263327047516818,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.28114067812846,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.2287650845507703,"pitch":null,"output_rms_dbfs":-24.762484155450313,"output_peak":0.4634203893921101,"reference_rms_dbfs":-23.53371907089954,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.106512364612346},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68000,"output_samples":68000,"intelligibility":{"estoi":{"value":0.853333814970128,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06657544016299914,"rms_dbfs":-23.53371907089954,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74175658806832,"mean":-0.00034814767276539525,"zero_energy":false},"output":{"rms":0.057793073617478916,"rms_dbfs":-24.762484155450313,"peak":0.4634203893921101,"peak_dbfs":-6.680497244080965,"crest_factor_db":18.081986911369345,"mean":0.0010350671797129812,"zero_energy":false},"rms_difference_db":-1.2287650845507727},"full_scale":{"reference":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":4.775576492788896e-06,"output_rms_dbfs":-53.20974194232137,"output_zero_energy":false,"output_peak":0.009352257336020181,"output_energy_fraction":8.41056748615221e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":85,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9420289855072463,"absolute_difference_cents_median":7.350767573731209,"absolute_difference_cents_p95":32.71204480015266,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00026228408676881105,"log_rms_mae_db":1.9452085275906275,"rms_envelope_nrmse":0.19745850607643986,"envelope_correlation":0.9740285416573758,"rms_gain_error_db":-1.228765102097442,"reference_mean_power":0.004437503233987748,"output_mean_power":0.0033439684584596306},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":1.6041806754964725e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008012042328921627,"output_power":0.0006080029076497755,"reference_active":true,"delta_db":-1.198375788735279},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0002055238993766061,"output_power":0.00015521400245143652,"reference_active":true,"delta_db":-1.2193143308263386},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.980344056384173e-05,"output_power":2.2810294057687583e-05,"reference_active":true,"delta_db":-2.4178972961728507}],"active_band_count":3,"active_band_mae_db":1.6118624719114896}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00021831174911177267,"log_rms_mae_db":1.4641589815882965,"rms_envelope_nrmse":0.15617319013956493,"envelope_correlation":0.990784255325588,"rms_gain_error_db":-1.2287703965117602,"reference_mean_power":0.004469027175667612,"output_mean_power":0.0033677198435918265},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":1.4254476041586354e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007408866379032469,"output_power":0.0005615002706623012,"reference_active":true,"delta_db":-1.204017922888718},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00010194186386509333,"output_power":7.6212642466168e-05,"reference_active":true,"delta_db":-1.2632554992731326},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.428475268148623e-06,"output_power":1.5610180850451293e-06,"reference_active":true,"delta_db":-1.9192575035606536}],"active_band_count":3,"active_band_mae_db":1.462176975240835}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":0.00023386465365005232,"log_rms_mae_db":2.0635877846135324,"rms_envelope_nrmse":0.21139354717908596,"envelope_correlation":0.9721673584544211,"rms_gain_error_db":-1.1333555579194037,"reference_mean_power":0.0021145087116098225,"output_mean_power":0.0016288231115550325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":9.68888675079529e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048786308639233274,"output_power":0.0003973088284314112,"reference_active":true,"delta_db":-0.8916974351891893},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001462062812176992,"output_power":0.00011600506635502999,"reference_active":true,"delta_db":-1.0048907406512384},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.8308307240601417e-05,"output_power":1.6059502783571816e-05,"reference_active":true,"delta_db":-2.4618180550951}],"active_band_count":3,"active_band_mae_db":1.4528020769785093}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":0.00020000043006212289,"log_rms_mae_db":1.4139575457909141,"rms_envelope_nrmse":0.16407424378551186,"envelope_correlation":0.9878819768210869,"rms_gain_error_db":-1.1333571684835868,"reference_mean_power":0.0021295552807363348,"output_mean_power":0.0016404129967450105},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":8.793551773152994e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00047116604915037,"output_power":0.000382314302750947,"reference_active":true,"delta_db":-0.9075344321120598},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.062785881121919e-05,"output_power":5.7605408024123753e-05,"reference_active":true,"delta_db":-0.8851278304900932},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5464332159215628e-06,"output_power":9.926610198213367e-07,"reference_active":true,"delta_db":-1.9253020117640818}],"active_band_count":3,"active_band_mae_db":1.239321424788745}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":2.1367380646997964,"rms_envelope_nrmse":0.2714841614011301,"envelope_correlation":0.9647606085421399,"rms_gain_error_db":-1.865889351897429,"reference_mean_power":0.00033167822412518246,"output_mean_power":0.0002158380577394677},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":2.2520658626731323e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014183157833093393,"output_power":8.681744985179215e-05,"reference_active":true,"delta_db":-2.1316591097845268},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6782950855174703e-05,"output_power":1.73377134548199e-05,"reference_active":true,"delta_db":-1.888666034431985},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.322147229373675e-06,"output_power":3.771225401804962e-06,"reference_active":true,"delta_db":-2.2438211495381566}],"active_band_count":3,"active_band_mae_db":2.0880487645848897}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":1.6889506293180394,"rms_envelope_nrmse":0.21945494207902713,"envelope_correlation":0.9887499271614403,"rms_gain_error_db":-1.865892704079789,"reference_mean_power":0.0003340384768502799,"output_mean_power":0.0002173738134104391},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":2.020124357938857e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013602429067654945,"output_power":8.2895266883726e-05,"reference_active":true,"delta_db":-2.150867357669058},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2971128220707267e-05,"output_power":8.460867019067395e-06,"reference_active":true,"delta_db":-1.8556288317088843},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.72810741850643e-07,"output_power":2.3769114180239017e-07,"reference_active":true,"delta_db":-1.954754203712702}],"active_band_count":3,"active_band_mae_db":1.987083464363548}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":849,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":2.6345489966171693,"rms_envelope_nrmse":0.22114616361768483,"envelope_correlation":0.9738403107243697,"rms_gain_error_db":-1.049950107571847,"reference_mean_power":8.148396414730941e-05,"output_mean_power":6.398484736145114e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":849,"frequency_resolution_hz":0.23557126030624262,"power_floor":8.546968728480114e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0215603768046545e-05,"output_power":3.17984930412392e-05,"reference_active":true,"delta_db":-1.0198805439957852},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.80808446115822e-06,"output_power":6.632303355500077e-06,"reference_active":true,"delta_db":0.5763145898173718},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.554920809430742e-06,"output_power":2.6643619012058625e-06,"reference_active":true,"delta_db":-1.2523671581292073}],"active_band_count":3,"active_band_mae_db":0.9495207639807881}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68000,"uncovered_tail_samples":0,"metrics":{"frame_count":843,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":2.2928748588566146,"rms_envelope_nrmse":0.15534694971639854,"envelope_correlation":0.9919177737880733,"rms_gain_error_db":-1.0499488252189022,"reference_mean_power":8.20637474953365e-05,"output_mean_power":6.444013816428806e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":843,"frequency_resolution_hz":0.2372479240806643,"power_floor":8.078579898158075e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.900632097728974e-05,"output_power":3.098203838129296e-05,"reference_active":true,"delta_db":-1.000250025631114},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.150421006895197e-06,"output_power":3.5075188846170366e-06,"reference_active":true,"delta_db":0.46631423301089997},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.3864016189443074e-07,"output_power":1.0726106568295375e-07,"reference_active":true,"delta_db":-1.1144694930194219}],"active_band_count":3,"active_band_mae_db":0.8603445838871453}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-157963-0018","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.4696478843688965,"predicted_mos":3.4696478843688965,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":102240,"input_sample_rate":24000,"input_seconds":4.26,"input_float32_sha256":"7a34ede7207dceb45da311535baf00654fbbc7f41126d9cab400cb3d6be710fd","internal_samples":68160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.742357885999809},"started_at":"2026-10-13T07:48:38.309010+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0019671119989652652,"native_input_samples":102120,"native_input_sha256":"9c266bdfe6845ba8d6283df2e6adbab2a3aee6c68239697a665cf6e43107e2ac","output_sha256":"967174d48befb7c738be89599491e5bf591e657471390244e47a53197e86674f","output_samples":102240,"length_delta_samples":120,"analysis_s":1.8303976110000804,"synthesis_s":9.231013054999494,"total_s":11.061410665999574,"rtf":2.599626478495787,"output_peak":0.4779982566833496,"output_rms":0.06537987833307753,"finished_at":"2026-10-13T07:48:49.379776+00:00","analysis_stages":{"external_f0_s":1.2563422819985135},"evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"967174d48befb7c738be89599491e5bf591e657471390244e47a53197e86674f","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":68080,"output_samples":68160,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.68843674742637,"output_peak":0.4759343271497732},"common_prefix":{"samples":68080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9988262910798122,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.545495016408079,"spectral_convergence":0.6171798151567348,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.216927215330679,"spectral_convergence":0.32158793913111877,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.510297006932153,"spectral_convergence":0.3123184681571039,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0007200846423285136,"delta_db":1.223856590057644,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.0027935975923513937,"delta_db":-0.350863587155551,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.00035347983283133793,"delta_db":-0.33475918049378917,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.00022731818244264309,"delta_db":-0.4699831616670149,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00013392528904866973,"delta_db":-0.10403015265946722,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":5.378817668867203e-05,"delta_db":-1.8790222864265826,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.7270858264100082,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.1445171023765999,"pitch":null,"output_rms_dbfs":-23.6833369020533,"output_peak":0.4759343271497732,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.75757307955697},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68080,"output_samples":68080,"intelligibility":{"estoi":{"value":0.852278895347492,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06653635567215388,"rms_dbfs":-23.538819799676702,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74685731684548,"mean":-0.000349972250879021,"zero_energy":false},"output":{"rms":0.06543847274876784,"rms_dbfs":-23.6833369020533,"peak":0.4759343271497732,"peak_dbfs":-6.449059404588669,"crest_factor_db":17.234277497464635,"mean":4.130045012456247e-05,"zero_energy":false},"rms_difference_db":-0.14451710237659654},"full_scale":{"reference":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":1.9859884402185427e-06,"output_rms_dbfs":-57.020232837177325,"output_zero_energy":false,"output_peak":0.007071426097469914,"output_energy_fraction":2.7249017682966988e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":88,"both_estimated_voiced_frames":62,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.8985507246376812,"absolute_difference_cents_median":7.918188341324804,"absolute_difference_cents_p95":16.722634821633186,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00026228408676881105,"log_rms_mae_db":2.1961745078476915,"rms_envelope_nrmse":0.174458032668581,"envelope_correlation":0.9682772374168602,"rms_gain_error_db":-0.14451091017407633,"reference_mean_power":0.004432288013980724,"output_mean_power":0.0042872309814976205},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":1.6030570640082068e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008005952323264084,"output_power":0.0008312372411762203,"reference_active":true,"delta_db":0.16311992807785647},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020547999455019126,"output_power":0.00019519580672182582,"reference_active":true,"delta_db":-0.22299061799092287},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.989786705158196e-05,"output_power":3.457172724772921e-05,"reference_active":true,"delta_db":-0.6222860083993336}],"active_band_count":3,"active_band_mae_db":0.33613218482270435}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00021831174911177267,"log_rms_mae_db":1.6720773265062097,"rms_envelope_nrmse":0.10640873568453647,"envelope_correlation":0.9873164469916639,"rms_gain_error_db":-0.144465049110479,"reference_mean_power":0.004463740704660867,"output_mean_power":0.004317699903261896},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":1.4245155353761532e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007405152195784611,"output_power":0.0007712294825926923,"reference_active":true,"delta_db":0.17649634571554965},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001021109367695941,"output_power":9.805867149945697e-05,"reference_active":true,"delta_db":-0.17586255323472913},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.434051823082556e-06,"output_power":1.986225804867378e-06,"reference_active":true,"delta_db":-0.8830120059385809}],"active_band_count":3,"active_band_mae_db":0.4117903016296199}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00023386465365005232,"log_rms_mae_db":2.173967849616012,"rms_envelope_nrmse":0.19618988334710502,"envelope_correlation":0.9688335219966181,"rms_gain_error_db":-0.3952787719313924,"reference_mean_power":0.002112021838569365,"output_mean_power":0.0019282819116593962},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":9.681983603336735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048709120906038954,"output_power":0.00046174463713689385,"reference_active":true,"delta_db":-0.23208431475142877},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014674692890756196,"output_power":0.00013703494192787805,"reference_active":true,"delta_db":-0.29737701092909397},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.836155879757368e-05,"output_power":2.5271947844952448e-05,"reference_active":true,"delta_db":-0.5009138007908767}],"active_band_count":3,"active_band_mae_db":0.3434583754904665}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00020000043006212289,"log_rms_mae_db":1.4438569304351294,"rms_envelope_nrmse":0.1285398338243509,"envelope_correlation":0.986571526357161,"rms_gain_error_db":-0.3952781134992151,"reference_mean_power":0.002127032769351675,"output_mean_power":0.001941987228340226},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.787606965551091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004705886462934124,"output_power":0.00044649814258927625,"reference_active":true,"delta_db":-0.22821788330557616},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.123456963821095e-05,"output_power":6.605259014204668e-05,"reference_active":true,"delta_db":-0.3280095244936771},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5465229502268847e-06,"output_power":1.4431521622957571e-06,"reference_active":true,"delta_db":-0.30044244854377794}],"active_band_count":3,"active_band_mae_db":0.2855566187810104}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":2.0089283014156445,"rms_envelope_nrmse":0.24474866213310828,"envelope_correlation":0.9525375342820879,"rms_gain_error_db":-0.31518621958722154,"reference_mean_power":0.0003312880907167699,"output_mean_power":0.0003080968102128704},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":2.2485323271997548e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014154142624946066,"output_power":0.0001287470631551765,"reference_active":true,"delta_db":-0.41146236332221064},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6917189688069e-05,"output_power":2.6926187091832176e-05,"reference_active":true,"delta_db":0.0014514404838118903},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.319943953798976e-06,"output_power":5.318627564001223e-06,"reference_active":true,"delta_db":-0.749136469241293}],"active_band_count":3,"active_band_mae_db":0.38735009101577184}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":1.272930139889035,"rms_envelope_nrmse":0.13536036384966474,"envelope_correlation":0.985089989963707,"rms_gain_error_db":-0.31518528867094936,"reference_mean_power":0.0003336427865109947,"output_mean_power":0.000310286735892818},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":2.0170856876139243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013578457159768304,"output_power":0.00012314102293261458,"reference_active":true,"delta_db":-0.4244766946466687},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3100830411180002e-05,"output_power":1.3668741323650836e-05,"reference_active":true,"delta_db":0.18429699926988852},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732875286933278e-07,"output_power":2.8750923856717024e-07,"reference_active":true,"delta_db":-1.1339167617322057}],"active_band_count":3,"active_band_mae_db":0.5808968185495876}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":1.7787187017467523,"rms_envelope_nrmse":0.3098657129696506,"envelope_correlation":0.9443164732299353,"rms_gain_error_db":-1.5669418779906792,"reference_mean_power":8.138813397574016e-05,"output_mean_power":5.673706989512848e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":8.567729341398194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0300307346839755e-05,"output_power":2.6545034877910633e-05,"reference_active":true,"delta_db":-1.8132505795994147},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8095151025280505e-06,"output_power":4.171020847549839e-06,"reference_active":true,"delta_db":-1.4389752450649584},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.571956353556119e-06,"output_power":1.1145330990589338e-06,"reference_active":true,"delta_db":-5.0581317296758215}],"active_band_count":3,"active_band_mae_db":2.770119184780065}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":1.1435294520964145,"rms_envelope_nrmse":0.22871624175774155,"envelope_correlation":0.9799773135473788,"rms_gain_error_db":-1.5669477251970054,"reference_mean_power":8.196654666810836e-05,"output_mean_power":5.714021441277894e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.099518469769931e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9096129724567364e-05,"output_power":2.5224562186025855e-05,"reference_active":true,"delta_db":-1.903101299759658},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.153274768568064e-06,"output_power":2.2187347713163093e-06,"reference_active":true,"delta_db":-1.5265642619291508},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4027458558124966e-07,"output_power":6.182913663886702e-08,"reference_active":true,"delta_db":-3.557858121239517}],"active_band_count":3,"active_band_mae_db":2.3291745609761088}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-157963-0018","method":"llsm2","status":"measured","mos":{"status":"measured","value":2.8035202026367188,"predicted_mos":2.8035202026367188,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":102360,"input_sample_rate":24000,"input_seconds":4.265,"input_float32_sha256":"00c3cc796d94e5911997e3ca3facaff1d56a179b7763276c6753b6829e1e42a9","internal_samples":68240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2185198469996976},"started_at":"2026-10-13T05:11:42.231783+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0021412349997262936,"native_input_samples":102120,"native_input_sha256":"9c266bdfe6845ba8d6283df2e6adbab2a3aee6c68239697a665cf6e43107e2ac","output_sha256":"61ac0d061e3845c6b01ec6042ce9ac70fcedb5b1861337bd6d2141881401c6e7","output_samples":102360,"length_delta_samples":240,"analysis_s":0.7616059510000923,"synthesis_s":0.10038895000070625,"total_s":0.8619949010007986,"rtf":0.2025839955348528,"output_peak":0.513477087020874,"output_rms":0.06436843887388137,"finished_at":"2026-10-13T05:11:43.101926+00:00","analysis_stages":{"external_f0_s":0.0772999990003882},"evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"61ac0d061e3845c6b01ec6042ce9ac70fcedb5b1861337bd6d2141881401c6e7","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":68080,"output_samples":68240,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.82370669702667,"output_peak":0.5150821779832716},"common_prefix":{"samples":68080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9976553341148886,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.674917537427752,"spectral_convergence":0.18285066878149847,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.349139011022407,"spectral_convergence":0.1758910865411284,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":4.8616081662201855,"spectral_convergence":0.1916138340438025,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005330792576847026,"delta_db":-0.08206105267006636,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.0028428540626382567,"delta_db":-0.27495640606893734,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.00036756685037276835,"delta_db":-0.16504222072521998,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.0002274754139169131,"delta_db":-0.4669802713482373,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00012142433941493558,"delta_db":-0.5295985818383102,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":6.334326319344326e-05,"delta_db":-1.1688861872223977,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.44792078664552815,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.27469318447216345,"pitch":null,"output_rms_dbfs":-23.813512984148865,"output_peak":0.5150821779832716,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.628554904890115},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68080,"output_samples":68080,"intelligibility":{"estoi":{"value":0.9221054224263437,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06653635567215388,"rms_dbfs":-23.538819799676702,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74685731684548,"mean":-0.000349972250879021,"zero_energy":false},"output":{"rms":0.0644650539999853,"rms_dbfs":-23.813512984148865,"peak":0.5150821779832716,"peak_dbfs":-5.762469531894279,"crest_factor_db":18.051043452254586,"mean":7.571610434475025e-05,"zero_energy":false},"rms_difference_db":-0.27469318447216295},"full_scale":{"reference":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":5.958401314384697e-06,"output_rms_dbfs":-52.248702492322124,"output_zero_energy":false,"output_peak":0.007826661202994767,"output_energy_fraction":8.424060814720568e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":76,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.9420289855072463,"absolute_difference_cents_median":1.0557750552898713,"absolute_difference_cents_p95":4.8440674188866994,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00026228408676881105,"log_rms_mae_db":0.924449619657202,"rms_envelope_nrmse":0.07535361910144095,"envelope_correlation":0.994942519191063,"rms_gain_error_db":-0.27469391263799653,"reference_mean_power":0.004432288013980724,"output_mean_power":0.004160625075835793},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":1.6030570640082068e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008005952323264084,"output_power":0.0007712202313736339,"reference_active":true,"delta_db":-0.16234585618119968},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020547999455019126,"output_power":0.0001885051849049092,"reference_active":true,"delta_db":-0.3744624539716837},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.989786705158196e-05,"output_power":2.574833844972819e-05,"reference_active":true,"delta_db":-1.9020046974243519}],"active_band_count":3,"active_band_mae_db":0.8129376691924118}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00021831174911177267,"log_rms_mae_db":0.5622487582779463,"rms_envelope_nrmse":0.04783827264140336,"envelope_correlation":0.9984538233951502,"rms_gain_error_db":-0.2746998404181876,"reference_mean_power":0.004463740704660867,"output_mean_power":0.004190144255069829},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":1.4245155353761532e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007405152195784611,"output_power":0.0007180449830815851,"reference_active":true,"delta_db":-0.1338233669794153},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001021109367695941,"output_power":9.425104750780184e-05,"reference_active":true,"delta_db":-0.34786074783316195},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.434051823082556e-06,"output_power":1.589219313039469e-06,"reference_active":true,"delta_db":-1.8514598630689245}],"active_band_count":3,"active_band_mae_db":0.7777146592938339}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00023386465365005232,"log_rms_mae_db":1.3094092242106725,"rms_envelope_nrmse":0.09967509457160598,"envelope_correlation":0.9929465976931094,"rms_gain_error_db":-0.3623043238957014,"reference_mean_power":0.002112021838569365,"output_mean_power":0.0019429783980287386},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":9.681983603336735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048709120906038954,"output_power":0.0004594732423043504,"reference_active":true,"delta_db":-0.25350066506409646},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014674692890756196,"output_power":0.00012677543857706687,"reference_active":true,"delta_db":-0.6353389952511274},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.836155879757368e-05,"output_power":1.3280209551155572e-05,"reference_active":true,"delta_db":-3.295251687928056}],"active_band_count":3,"active_band_mae_db":1.3946971160810933}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00020000043006212289,"log_rms_mae_db":0.6903911183150411,"rms_envelope_nrmse":0.06285497711549375,"envelope_correlation":0.9978906369931055,"rms_gain_error_db":-0.3623049835938041,"reference_mean_power":0.002127032769351675,"output_mean_power":0.001956787576484249},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.787606965551091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004705886462934124,"output_power":0.0004390086539564567,"reference_active":true,"delta_db":-0.3016836355993656},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.123456963821095e-05,"output_power":6.155819684943723e-05,"reference_active":true,"delta_db":-0.6340491478944223},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5465229502268847e-06,"output_power":8.155659060931797e-07,"reference_active":true,"delta_db":-2.7789730702380733}],"active_band_count":3,"active_band_mae_db":1.238235284577287}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":1.1758123787517134,"rms_envelope_nrmse":0.12252307764264977,"envelope_correlation":0.9902163007138305,"rms_gain_error_db":-0.5021852030920917,"reference_mean_power":0.0003312880907167699,"output_mean_power":0.00029511229510406303},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":2.2485323271997548e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014154142624946066,"output_power":0.0001213320343463389,"reference_active":true,"delta_db":-0.6690808810672616},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6917189688069e-05,"output_power":2.2302957751711582e-05,"reference_active":true,"delta_db":-0.8166725333058904},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.319943953798976e-06,"output_power":5.223955713679253e-06,"reference_active":true,"delta_db":-0.8271374038001832}],"active_band_count":3,"active_band_mae_db":0.7709636060577784}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":0.7442305927546883,"rms_envelope_nrmse":0.08159124472528366,"envelope_correlation":0.9968791169674841,"rms_gain_error_db":-0.5021852821031704,"reference_mean_power":0.0003336427865109947,"output_mean_power":0.00029720985885072225},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":2.0170856876139243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013578457159768304,"output_power":0.00011674395696464472,"reference_active":true,"delta_db":-0.6561601705179625},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3100830411180002e-05,"output_power":1.0595468438985883e-05,"reference_active":true,"delta_db":-0.9217866260617736},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732875286933278e-07,"output_power":2.388947204475063e-07,"reference_active":true,"delta_db":-1.938369286714845}],"active_band_count":3,"active_band_mae_db":1.1721053610981937}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":1.0635627834114219,"rms_envelope_nrmse":0.18740641149922532,"envelope_correlation":0.9816204396831808,"rms_gain_error_db":-0.9480666378327346,"reference_mean_power":8.138813397574016e-05,"output_mean_power":6.542661137010823e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":8.567729341398194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0300307346839755e-05,"output_power":3.141401883246578e-05,"reference_active":true,"delta_db":-1.081848585055626},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8095151025280505e-06,"output_power":4.023718326574872e-06,"reference_active":true,"delta_db":-1.5951231396386705},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.571956353556119e-06,"output_power":1.2992258584702517e-06,"reference_active":true,"delta_db":-4.392214877795931}],"active_band_count":3,"active_band_mae_db":2.356395534163409}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":0.8074450663400186,"rms_envelope_nrmse":0.1399815636610946,"envelope_correlation":0.9930772422660333,"rms_gain_error_db":-0.9480674614396728,"reference_mean_power":8.196654666810836e-05,"output_mean_power":6.589157553101295e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.099518469769931e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9096129724567364e-05,"output_power":2.9780011314368926e-05,"reference_active":true,"delta_db":-1.1820890863327742},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.153274768568064e-06,"output_power":2.12993620263943e-06,"reference_active":true,"delta_db":-1.7039522040726585},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4027458558124966e-07,"output_power":5.173760363171207e-08,"reference_active":true,"delta_db":-4.331726850116634}],"active_band_count":3,"active_band_mae_db":2.4059227135073553}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-157963-0018","method":"magphase","status":"measured","mos":{"status":"measured","value":3.3190221786499023,"predicted_mos":3.3190221786499023,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":205947,"input_sample_rate":48000,"input_seconds":4.2905625,"input_float32_sha256":"f9c55c015b16c68a7af4b740f229ce20cc5fbbb09843613c6b9a6b8e25451592","internal_samples":68649,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.156955585000105},"started_at":"2026-10-13T05:19:21.404102+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.007231382000099984,"native_input_samples":204240,"native_input_sha256":"fc5399473b2f85ecbaf9979c31fa97ebaffdfc5b0da147b68b64a783e68dc3e5","output_sha256":"eeeec76f5624f61a268841792169b2bcfc4ad3e16fe395065760b7625d0a24d4","output_samples":205947,"length_delta_samples":1707,"analysis_s":2.5596300729994255,"synthesis_s":0.771141720999367,"total_s":3.3307717939987924,"rtf":0.7827900808457797,"output_peak":0.5036829189051306,"output_rms":0.06378933907517191,"finished_at":"2026-10-13T05:19:24.750113+00:00","evaluation_input":{"reference":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"eeeec76f5624f61a268841792169b2bcfc4ad3e16fe395065760b7625d0a24d4","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":68080,"output_samples":68649,"length_delta_samples":569,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.90266120113317,"output_peak":0.4971972094339347},"common_prefix":{"samples":68080,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9917114597444974,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":68080,"output_samples":68080,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.099060838619819,"spectral_convergence":0.6882368252833468,"spectral_convergence_status":"ok","amplitude_floor":2.0850976744763412e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.310585616762612,"spectral_convergence":0.48523366184871336,"spectral_convergence_status":"ok","amplitude_floor":1.4381271492089952e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.933226352093255,"spectral_convergence":0.49559649744495593,"spectral_convergence_status":"ok","amplitude_floor":9.411979934974602e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0005432476915893821,"output_power":0.0005110914848882827,"delta_db":-0.2649924416728675,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0030286578008489277,"output_power":0.0027350259388380536,"delta_db":-0.44288756932233864,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00038180407470981095,"output_power":0.00037068448343535374,"delta_db":-0.12836151802867682,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.0002532984259990213,"output_power":0.0002519314620217425,"delta_db":-0.023500840904346013,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00013717204322211298,"output_power":0.00014760350354146197,"delta_db":0.3183105845363311,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":8.29065897621083e-05,"output_power":8.898776831283953e-05,"delta_db":0.307412639547516,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.24757759900201268,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.32770514061269623,"pitch":null,"output_rms_dbfs":-23.8665249402894,"output_peak":0.4971972094339347,"reference_rms_dbfs":-23.538819799676702,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.447624269158562},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":68080,"output_samples":68080,"intelligibility":{"estoi":{"value":0.6090835007738907,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06653635567215388,"rms_dbfs":-23.538819799676702,"peak":0.513336181640625,"peak_dbfs":-5.79196248283122,"crest_factor_db":17.74685731684548,"mean":-0.000349972250879021,"zero_energy":false},"output":{"rms":0.06407280734475221,"rms_dbfs":-23.8665249402894,"peak":0.4971972094339347,"peak_dbfs":-6.069426350842956,"crest_factor_db":17.797098589446442,"mean":0.00044297347506569017,"zero_energy":false},"rms_difference_db":-0.3277051406126965},"full_scale":{"reference":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":68080,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":425,"uncovered_tail_samples":80,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":6.879294217214919e-06,"threshold_dbfs":-51.624561160220566,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":25,"quiet_samples":4000,"reference_mean_power":4.6952294651418924e-06,"output_mean_power":5.734245636832903e-06,"output_rms_dbfs":-52.415237071976684,"output_zero_energy":false,"output_peak":0.010372784824958172,"output_energy_fraction":8.206712721597036e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":418,"reference_estimated_voiced_frames":69,"output_estimated_voiced_frames":76,"both_estimated_voiced_frames":50,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":26,"paired_voiced_fraction":0.7246376811594203,"absolute_difference_cents_median":16.25344357689174,"absolute_difference_cents_p95":48.36724140041905,"difference_over_600_cents_frames":0,"uncovered_tail_samples":80,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":68080,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00026228408676881105,"log_rms_mae_db":2.9385724706614393,"rms_envelope_nrmse":0.31167620885617525,"envelope_correlation":0.8971606562704288,"rms_gain_error_db":-0.32770010406348976,"reference_mean_power":0.004432288013980724,"output_mean_power":0.004110152757032187},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":1.6030570640082068e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0008005952323264084,"output_power":0.000773387778579714,"reference_active":true,"delta_db":-0.1501569460794783},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00020547999455019126,"output_power":0.00019642795896537339,"reference_active":true,"delta_db":-0.19566241511673316},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.989786705158196e-05,"output_power":3.9779785701796864e-05,"reference_active":true,"delta_db":-0.01287239631156397}],"active_band_count":3,"active_band_mae_db":0.11956391916925847}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00021831174911177267,"log_rms_mae_db":2.092255608980041,"rms_envelope_nrmse":0.2179588354247781,"envelope_correlation":0.9454940140352163,"rms_gain_error_db":-0.3276820179657139,"reference_mean_power":0.004463740704660867,"output_mean_power":0.004139336728676294},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":1.4245155353761532e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0007405152195784611,"output_power":0.0007142287272810894,"reference_active":true,"delta_db":-0.15696674726846369},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001021109367695941,"output_power":9.60219768177288e-05,"reference_active":true,"delta_db":-0.26701617838690483},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.434051823082556e-06,"output_power":2.8323667236334005e-06,"reference_active":true,"delta_db":0.6581966288884751}],"active_band_count":3,"active_band_mae_db":0.36072651818128126}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":0.00023386465365005232,"log_rms_mae_db":4.055095058823205,"rms_envelope_nrmse":0.39279851441504293,"envelope_correlation":0.8699282718213865,"rms_gain_error_db":-0.41903570550167385,"reference_mean_power":0.002112021838569365,"output_mean_power":0.0019177625522382859},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":9.681983603336735e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00048709120906038954,"output_power":0.0004600031458753515,"reference_active":true,"delta_db":-0.2484948981893077},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00014674692890756196,"output_power":0.00013604150283872367,"reference_active":true,"delta_db":-0.3289760027981356},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.836155879757368e-05,"output_power":2.4008352160384276e-05,"reference_active":true,"delta_db":-0.7236774389949889}],"active_band_count":3,"active_band_mae_db":0.4337161133274774}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":0.00020000043006212289,"log_rms_mae_db":3.2092492180604917,"rms_envelope_nrmse":0.29349783371257915,"envelope_correlation":0.9221734999073433,"rms_gain_error_db":-0.41903474673360286,"reference_mean_power":0.002127032769351675,"output_mean_power":0.0019313932358482255},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.787606965551091e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0004705886462934124,"output_power":0.0004447379875546665,"reference_active":true,"delta_db":-0.24537218364965274},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.123456963821095e-05,"output_power":6.432249775177625e-05,"reference_active":true,"delta_db":-0.4432790440707911},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5465229502268847e-06,"output_power":1.5695453181679379e-06,"reference_active":true,"delta_db":0.06417490523125635}],"active_band_count":3,"active_band_mae_db":0.2509420443172334}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":7.720600674415526e-05,"log_rms_mae_db":4.123304061692444,"rms_envelope_nrmse":0.46377393192692457,"envelope_correlation":0.8352377292028086,"rms_gain_error_db":0.09476855785813376,"reference_mean_power":0.0003312880907167699,"output_mean_power":0.00033859666775032203},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":2.2485323271997548e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00014154142624946066,"output_power":0.0001492104633640577,"reference_active":true,"delta_db":0.2291571156579768},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.6917189688069e-05,"output_power":2.8676048385828057e-05,"reference_active":true,"delta_db":0.27489589474277665},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.319943953798976e-06,"output_power":7.3819390694338585e-06,"reference_active":true,"delta_db":0.6745722929382652}],"active_band_count":3,"active_band_mae_db":0.3928751011130062}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":6.301020513193742e-05,"log_rms_mae_db":3.1576016850067337,"rms_envelope_nrmse":0.3359993267779927,"envelope_correlation":0.9064493436061077,"rms_gain_error_db":0.09476886781704988,"reference_mean_power":0.0003336427865109947,"output_mean_power":0.00034100333504263235},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":2.0170856876139243e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00013578457159768304,"output_power":0.00014413851837907085,"reference_active":true,"delta_db":0.25929627149957485},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3100830411180002e-05,"output_power":1.3304823697057864e-05,"reference_active":true,"delta_db":0.06710299283030538},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.732875286933278e-07,"output_power":3.939432372064118e-07,"reference_active":true,"delta_db":0.23390168721721574}],"active_band_count":3,"active_band_mae_db":0.18676698384903198}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":850,"rms_floor":5.974138539374114e-05,"log_rms_mae_db":3.7089021567241747,"rms_envelope_nrmse":0.5970200084755691,"envelope_correlation":0.7768628136575694,"rms_gain_error_db":0.49290074179483057,"reference_mean_power":8.138813397574016e-05,"output_mean_power":9.116983430505983e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":850,"frequency_resolution_hz":0.23529411764705882,"power_floor":8.567729341398194e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":4.0300307346839755e-05,"output_power":4.337783612439408e-05,"reference_active":true,"delta_db":0.3195952545272706},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.8095151025280505e-06,"output_power":5.636282016367149e-06,"reference_active":true,"delta_db":-0.13147169753202737},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.571956353556119e-06,"output_power":8.278531509143939e-06,"reference_active":true,"delta_db":3.650471625857013}],"active_band_count":3,"active_band_mae_db":1.3671795259721036}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":68080,"uncovered_tail_samples":0,"metrics":{"frame_count":844,"rms_floor":4.362120322552387e-05,"log_rms_mae_db":2.9409016557056162,"rms_envelope_nrmse":0.4147036538460985,"envelope_correlation":0.8881487075510459,"rms_gain_error_db":0.4929013345050601,"reference_mean_power":8.196654666810836e-05,"output_mean_power":9.181777653717344e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":844,"frequency_resolution_hz":0.23696682464454977,"power_floor":8.099518469769931e-11,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":3.9096129724567364e-05,"output_power":4.38747650957283e-05,"reference_active":true,"delta_db":0.5008103722784469},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.153274768568064e-06,"output_power":2.9245701348054214e-06,"reference_active":true,"delta_db":-0.32699775001800757},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.4027458558124966e-07,"output_power":4.0490314184935354e-07,"reference_active":true,"delta_db":4.603721523691497}],"active_band_count":3,"active_band_mae_db":1.8105098819959837}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"hnm","status":"measured","mos":{"status":"measured","value":2.6298842430114746,"predicted_mos":2.6298842430114746,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76640,"input_sample_rate":16000,"input_seconds":4.79,"input_float32_sha256":"bc8c27a1dde3a7758678b61dcab4a8141c8ab4353d73c46ee444fdf30acaea01","internal_samples":76640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.203561644999354},"started_at":"2026-10-13T04:40:37.308063+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.1893000191776082e-05,"native_input_samples":76640,"native_input_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","output_sha256":"171115d2189e5657c47d2c833f4b77af7620f07443a9f8977cac63389adb1cc4","output_samples":76640,"length_delta_samples":0,"analysis_s":0.8349359040003037,"synthesis_s":0.3902884109993465,"total_s":1.2252243149996502,"rtf":0.2557879572024322,"output_peak":0.40244264626301435,"output_rms":0.05195160132512759,"finished_at":"2026-10-13T04:40:38.537304+00:00","evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"171115d2189e5657c47d2c833f4b77af7620f07443a9f8977cac63389adb1cc4","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.935770104949943,"spectral_convergence":0.29967767952362273,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.619384355607423,"spectral_convergence":0.29022985924907635,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.861254377267218,"spectral_convergence":0.3126958666114974,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.00041535645345757517,"delta_db":0.28367386783576953,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0012925460111910926,"delta_db":-0.29262367342773093,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0006717527428349002,"delta_db":-0.029860198011106127,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":0.00010004814675145573,"delta_db":-0.05324751937628619,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":7.141703763680438e-05,"delta_db":-0.058240270612616364,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.00014784848837317159,"delta_db":-0.6651706245469517,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23046935896841017,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15044225481546847,"pitch":null,"output_rms_dbfs":-25.688021229335874,"output_peak":0.40244264626301435,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.805469612608195},"common_prefix":{"samples":76640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.935770104949943,"spectral_convergence":0.29967767952362273,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.619384355607423,"spectral_convergence":0.29022985924907635,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.861254377267218,"spectral_convergence":0.3126958666114974,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.00041535645345757517,"delta_db":0.28367386783576953,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0012925460111910926,"delta_db":-0.29262367342773093,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0006717527428349002,"delta_db":-0.029860198011106127,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":0.00010004814675145573,"delta_db":-0.05324751937628619,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":7.141703763680438e-05,"delta_db":-0.058240270612616364,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.00014784848837317159,"delta_db":-0.6651706245469517,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.23046935896841017,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.15044225481546847,"pitch":null,"output_rms_dbfs":-25.688021229335874,"output_peak":0.40244264626301435,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.805469612608195},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76640,"output_samples":76640,"intelligibility":{"estoi":{"value":0.8509299066988453,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05285925663037164,"rms_dbfs":-25.537578974520407,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.031195842748737,"mean":-0.00030972872994886806,"zero_energy":false},"output":{"rms":0.05195160132512759,"rms_dbfs":-25.688021229335874,"peak":0.40244264626301435,"peak_dbfs":-7.905920079154778,"crest_factor_db":17.7821011501811,"mean":-0.00027563771658142616,"zero_energy":false},"rms_difference_db":-0.15044225481546647},"full_scale":{"reference":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":479,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":2.5423033030541847e-06,"output_rms_dbfs":-55.94772638288291,"output_zero_energy":false,"output_peak":0.006777898074504038,"output_energy_fraction":0.0001553535073277732,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":472,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":85,"both_estimated_voiced_frames":70,"reference_voiced_output_unvoiced_frames":3,"reference_unvoiced_output_voiced_frames":15,"paired_voiced_fraction":0.958904109589041,"absolute_difference_cents_median":1.3260348814222223,"absolute_difference_cents_p95":6.0378475296942975,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00018226521943209611,"log_rms_mae_db":1.2315243958393338,"rms_envelope_nrmse":0.10070968410137918,"envelope_correlation":0.9902654174401705,"rms_gain_error_db":-0.15043842059073334,"reference_mean_power":0.0027970153309965396,"output_mean_power":0.0027017863597632364},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.029329446119484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010840604762391139,"output_power":0.0010545990209273241,"reference_active":true,"delta_db":-0.11966146963574455},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017643244749495566,"output_power":0.00015774919461126368,"reference_active":true,"delta_db":-0.48611308124985353},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.741521666729805e-05,"output_power":4.3465468163801574e-05,"reference_active":true,"delta_db":-0.3777337769545646}],"active_band_count":3,"active_band_mae_db":0.3278361092800542}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001676463927845064,"log_rms_mae_db":0.6728577880583884,"rms_envelope_nrmse":0.06292818272123223,"envelope_correlation":0.9960716554558857,"rms_gain_error_db":-0.1504472090380429,"reference_mean_power":0.0028146416719843875,"output_mean_power":0.0027188070812589464},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.889944424347358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010453379933036454,"output_power":0.001015443923713281,"reference_active":true,"delta_db":-0.12600790144749158},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.750212405766423e-05,"output_power":6.156459976734225e-05,"reference_active":true,"delta_db":-0.39986378326657734},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9228813706071634e-06,"output_power":2.795782035298892e-06,"reference_active":true,"delta_db":-0.1930787930572332}],"active_band_count":3,"active_band_mae_db":0.23965015925710073}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00017586243391937327,"log_rms_mae_db":1.8971741379708333,"rms_envelope_nrmse":0.12398074671886526,"envelope_correlation":0.9886991613787923,"rms_gain_error_db":-0.18017762260029405,"reference_mean_power":0.0014437040494073058,"output_mean_power":0.0013850339289029033},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":1.5100208608044743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242573938618173,"output_power":0.0008884465189249259,"reference_active":true,"delta_db":-0.17161643457217932},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001371496531392223,"output_power":0.00011475465759311155,"reference_active":true,"delta_db":-0.774243921311451},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2803310723594984e-05,"output_power":2.8810053443844507e-05,"reference_active":true,"delta_db":-0.5637361383368786}],"active_band_count":3,"active_band_mae_db":0.5031988314068364}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001539617967365963,"log_rms_mae_db":1.231769135292498,"rms_envelope_nrmse":0.06720617201473936,"envelope_correlation":0.996719904300163,"rms_gain_error_db":-0.18017947507273188,"reference_mean_power":0.0014528118102568712,"output_mean_power":0.0013937709685351517},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.4301118553576922e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009152950896245062,"output_power":0.0008702824423713039,"reference_active":true,"delta_db":-0.21900910586080535},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.331274162853222e-05,"output_power":4.7947036594987884e-05,"reference_active":true,"delta_db":-0.46069246444639117},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0543911334413043e-06,"output_power":2.311140391195438e-06,"reference_active":true,"delta_db":0.511431956627298}],"active_band_count":3,"active_band_mae_db":0.39704450897816485}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":1.8599482168261119,"rms_envelope_nrmse":0.18509464309669924,"envelope_correlation":0.9764970846995525,"rms_gain_error_db":-0.05523638833704192,"reference_mean_power":0.0001541240756859989,"output_mean_power":0.0001521762389344325},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1034128857485058e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011785014167056142,"output_power":0.00011027876993821265,"reference_active":true,"delta_db":-0.28838195645445286},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.031054512020665e-06,"output_power":9.560797297666089e-06,"reference_active":true,"delta_db":0.7572153688052768},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6793453202028483e-06,"output_power":3.1734607396568234e-06,"reference_active":true,"delta_db":0.7350443973662981}],"active_band_count":3,"active_band_mae_db":0.5935472408753426}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":1.206074516501808,"rms_envelope_nrmse":0.10464969259233828,"envelope_correlation":0.9921452715395634,"rms_gain_error_db":-0.05524392123304539,"reference_mean_power":0.00015509566252518503,"output_mean_power":0.00015313528113854811},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.037312640211567e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011306875179582826,"output_power":0.00010683161490616144,"reference_active":true,"delta_db":-0.2464280447275915},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4410710139131097e-06,"output_power":4.618672381257251e-06,"reference_active":true,"delta_db":1.2782352198218057},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5029543761566635e-07,"output_power":1.935658509608236e-07,"reference_active":true,"delta_db":1.098829438597705}],"active_band_count":3,"active_band_mae_db":0.8744975677157006}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00013414824565394508,"log_rms_mae_db":1.577627446235891,"rms_envelope_nrmse":0.24172433142195732,"envelope_correlation":0.9677568907636498,"rms_gain_error_db":-0.6742199779914426,"reference_mean_power":0.0001730034207748909,"output_mean_power":0.00014812647660711963},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1522167661145565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012025367220488917,"output_power":9.355084331857767e-05,"reference_active":true,"delta_db":-1.090506405436023},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5558902161138395e-05,"output_power":2.0058401286097986e-05,"reference_active":true,"delta_db":-1.0524588000222928},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5309687441279477e-05,"output_power":8.924853737651437e-06,"reference_active":true,"delta_db":-2.3436521680729787}],"active_band_count":3,"active_band_mae_db":1.4955391245104313}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":1.4089729327008003,"rms_envelope_nrmse":0.18643767623866817,"envelope_correlation":0.9819388696229866,"rms_gain_error_db":-0.674224421321691,"reference_mean_power":0.00017409446216015185,"output_mean_power":0.00014906047971204354},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.085538754133612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012977220791713165,"output_power":9.90399870580976e-05,"reference_active":true,"delta_db":-1.173711187671975},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3304937813307312e-05,"output_power":9.048415461635291e-06,"reference_active":true,"delta_db":-1.6744031583786976},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0464863211082628e-06,"output_power":6.035301528724772e-07,"reference_active":true,"delta_db":-2.3903458324512212}],"active_band_count":3,"active_band_mae_db":1.7461533928339648}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.475158929824829,"predicted_mos":3.475158929824829,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76640,"input_sample_rate":16000,"input_seconds":4.79,"input_float32_sha256":"1ea0da32afc39f961d53940afe18f4bd79ed53862e40ca55c060e3289ac26af8","internal_samples":76640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1850835839995852},"started_at":"2026-10-13T04:43:54.125973+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.407200008747168e-05,"native_input_samples":76640,"native_input_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","output_sha256":"cb5e613b7086c98ac50b69b1c8a7bbed4286d3ff91d2ceb194e16f422670a343","output_samples":76640,"length_delta_samples":0,"analysis_s":1.3482449419998375,"synthesis_s":0.0621444819998942,"total_s":1.4103894239997317,"rtf":0.2944445561586079,"output_peak":0.5856033752636008,"output_rms":0.060045515973337966,"finished_at":"2026-10-13T04:43:55.541536+00:00","evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"cb5e613b7086c98ac50b69b1c8a7bbed4286d3ff91d2ceb194e16f422670a343","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.764509582158898,"spectral_convergence":0.6794509613883367,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.371765803647086,"spectral_convergence":0.3797123055683177,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.734845332514561,"spectral_convergence":0.37583556003240454,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.00041645717575067683,"delta_db":0.295167738102911,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0017818475022709159,"delta_db":1.1016215350049596,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0010111829848706408,"delta_db":1.7463428496805464,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":0.00011099527928503357,"delta_db":0.39770707998194266,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":8.098709404836773e-05,"delta_db":0.4878995718631051,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.0002039939522787502,"delta_db":0.7328334123183737,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.793595364491973,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1071905962274993,"pitch":null,"output_rms_dbfs":-24.43038837829291,"output_peak":0.5856033752636008,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.957040239440182},"common_prefix":{"samples":76640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.764509582158898,"spectral_convergence":0.6794509613883367,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.371765803647086,"spectral_convergence":0.3797123055683177,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.734845332514561,"spectral_convergence":0.37583556003240454,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.00041645717575067683,"delta_db":0.295167738102911,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0017818475022709159,"delta_db":1.1016215350049596,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0010111829848706408,"delta_db":1.7463428496805464,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":0.00011099527928503357,"delta_db":0.39770707998194266,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":8.098709404836773e-05,"delta_db":0.4878995718631051,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.0002039939522787502,"delta_db":0.7328334123183737,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.793595364491973,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.1071905962274993,"pitch":null,"output_rms_dbfs":-24.43038837829291,"output_peak":0.5856033752636008,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.957040239440182},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76640,"output_samples":76640,"intelligibility":{"estoi":{"value":0.9092308236787718,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05285925663037164,"rms_dbfs":-25.537578974520407,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.031195842748737,"mean":-0.00030972872994886806,"zero_energy":false},"output":{"rms":0.06004551597333796,"rms_dbfs":-24.43038837829291,"peak":0.5856033752636008,"peak_dbfs":-4.647928575669097,"crest_factor_db":19.782459802623812,"mean":-2.720732003406928e-08,"zero_energy":false},"rms_difference_db":1.1071905962274968},"full_scale":{"reference":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":479,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":1.3705409104465564e-06,"output_rms_dbfs":-58.63107996303968,"output_zero_energy":false,"output_peak":0.007601951622867624,"output_energy_fraction":6.269348606351931e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":472,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":97,"both_estimated_voiced_frames":72,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":25,"paired_voiced_fraction":0.9863013698630136,"absolute_difference_cents_median":7.305777577839531,"absolute_difference_cents_p95":19.253837887733674,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00018226521943209611,"log_rms_mae_db":2.599287864857397,"rms_envelope_nrmse":0.23670133658083214,"envelope_correlation":0.9689693385754905,"rms_gain_error_db":1.1071983968393293,"reference_mean_power":0.0027970153309965396,"output_mean_power":0.003609231063255428},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.029329446119484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010840604762391139,"output_power":0.0015098544749346519,"reference_active":true,"delta_db":1.4388157971897209},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017643244749495566,"output_power":0.00021526002412064775,"reference_active":true,"delta_db":0.86384925848098},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.741521666729805e-05,"output_power":4.761569434493803e-05,"reference_active":true,"delta_db":0.018323822717839333}],"active_band_count":3,"active_band_mae_db":0.7736629594628468}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001676463927845064,"log_rms_mae_db":1.9104514215770156,"rms_envelope_nrmse":0.1729709895523488,"envelope_correlation":0.9901672984910974,"rms_gain_error_db":1.107219914405163,"reference_mean_power":0.0028146416719843875,"output_mean_power":0.003631993852263728},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.889944424347358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010453379933036454,"output_power":0.001459850730405363,"reference_active":true,"delta_db":1.4505171616519292},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.750212405766423e-05,"output_power":8.654976454772559e-05,"reference_active":true,"delta_db":1.0794845196007785},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9228813706071634e-06,"output_power":2.129915264248758e-06,"reference_active":true,"delta_db":-1.3744886326834882}],"active_band_count":3,"active_band_mae_db":1.3014967713120653}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00017586243391937327,"log_rms_mae_db":2.858751507536635,"rms_envelope_nrmse":0.26880550700122396,"envelope_correlation":0.9733195769774511,"rms_gain_error_db":1.3775349210058985,"reference_mean_power":0.0014437040494073058,"output_mean_power":0.00198258431835652},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":1.5100208608044743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242573938618173,"output_power":0.0013771008106218652,"reference_active":true,"delta_db":1.7317280041248746},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001371496531392223,"output_power":0.00018437385874519013,"reference_active":true,"delta_db":1.2850463151162719},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2803310723594984e-05,"output_power":2.435444755845039e-05,"reference_active":true,"delta_db":-1.2933939500281575}],"active_band_count":3,"active_band_mae_db":1.4367227564231013}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001539617967365963,"log_rms_mae_db":2.3302656167675457,"rms_envelope_nrmse":0.207497263029036,"envelope_correlation":0.9911817476772461,"rms_gain_error_db":1.3775306368135467,"reference_mean_power":0.0014528118102568712,"output_mean_power":0.0019950896947100193},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.4301118553576922e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009152950896245062,"output_power":0.0013510177059748553,"reference_active":true,"delta_db":1.6909990830881452},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.331274162853222e-05,"output_power":7.601641512009791e-05,"reference_active":true,"delta_db":1.5407636786972385},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0543911334413043e-06,"output_power":1.3664517748126042e-06,"reference_active":true,"delta_db":-1.770888229584556}],"active_band_count":3,"active_band_mae_db":1.6675503304566466}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":1.7129682306155274,"rms_envelope_nrmse":0.20316135877279975,"envelope_correlation":0.9746184485186885,"rms_gain_error_db":0.40750592631884575,"reference_mean_power":0.0001541240756859989,"output_mean_power":0.0001692860133979775},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1034128857485058e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011785014167056142,"output_power":0.00012510251154462603,"reference_active":true,"delta_db":0.25935919693693177},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.031054512020665e-06,"output_power":8.009992688906125e-06,"reference_active":true,"delta_db":-0.011404540737601032},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6793453202028483e-06,"output_power":2.0344691009793767e-06,"reference_active":true,"delta_db":-1.195775918089914}],"active_band_count":3,"active_band_mae_db":0.48884655192148224}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":0.92438017214773,"rms_envelope_nrmse":0.0887905917406996,"envelope_correlation":0.9962566031241915,"rms_gain_error_db":0.4075073225791588,"reference_mean_power":0.00015509566252518503,"output_mean_power":0.00017035323474622066},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.037312640211567e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011306875179582826,"output_power":0.00012010639528123913,"reference_active":true,"delta_db":0.26223534937979265},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4410710139131097e-06,"output_power":3.6872660379066634e-06,"reference_active":true,"delta_db":0.3001083796359664},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5029543761566635e-07,"output_power":1.1546364991414335e-07,"reference_active":true,"delta_db":-1.14500515462359}],"active_band_count":3,"active_band_mae_db":0.5691162945464497}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00013414824565394508,"log_rms_mae_db":1.4259964008942354,"rms_envelope_nrmse":0.3906845364066246,"envelope_correlation":0.9255674027812314,"rms_gain_error_db":0.7672294051188565,"reference_mean_power":0.0001730034207748909,"output_mean_power":0.00020643229031193784},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1522167661145565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012025367220488917,"output_power":0.0001472528536642117,"reference_active":true,"delta_db":0.8796537251856782},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5558902161138395e-05,"output_power":3.529261829747541e-05,"reference_active":true,"delta_db":1.4014168358994603},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5309687441279477e-05,"output_power":2.0033749845217093e-05,"reference_active":true,"delta_db":1.1679592223879194}],"active_band_count":3,"active_band_mae_db":1.1496765944910192}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":0.9829236163551938,"rms_envelope_nrmse":0.23581750338844482,"envelope_correlation":0.9750886291470556,"rms_gain_error_db":0.7672305450900979,"reference_mean_power":0.00017409446216015185,"output_mean_power":0.0002077342044848606},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.085538754133612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012977220791713165,"output_power":0.00016220935452218585,"reference_active":true,"delta_db":0.9689420242324007},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3304937813307312e-05,"output_power":1.8571911026688434e-05,"reference_active":true,"delta_db":1.4484374538560347},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0464863211082628e-06,"output_power":1.237487900644551e-06,"reference_active":true,"delta_db":0.7280740543913238}],"active_band_count":3,"active_band_mae_db":1.0484845108265863}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"facodec","status":"measured","mos":{"status":"measured","value":4.2577409744262695,"predicted_mos":4.2577409744262695,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76800,"input_sample_rate":16000,"input_seconds":4.8,"input_float32_sha256":"c295b0952111d51b9ff5c513c686b8fc8a15267344b0242537b10c3b5887efd8","internal_samples":76800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6154950070013},"started_at":"2026-10-13T07:02:59.805056+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":6.050799856893718e-05,"native_input_samples":76640,"native_input_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","output_sha256":"2880383a422c730c1963d065c6bbe4771e70450a74f1780e4141a8a7a097f42a","output_samples":76800,"length_delta_samples":160,"analysis_s":1.5900900390006427,"synthesis_s":10.130781883000964,"total_s":11.720871922001606,"rtf":2.446946121503467,"output_peak":0.4061483144760132,"output_rms":0.0495450998735917,"finished_at":"2026-10-13T07:03:11.536828+00:00","evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"2880383a422c730c1963d065c6bbe4771e70450a74f1780e4141a8a7a097f42a","native_sample_rate":16000},"measurement":{"status":"length_mismatch","reference_samples":76640,"output_samples":76800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.09998583592254,"output_peak":0.4061483144760132},"common_prefix":{"samples":76640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9979166666666667,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.517184450169443,"spectral_convergence":0.2789384695829309,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.734737733862231,"spectral_convergence":0.2943802757473796,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.8715861188217655,"spectral_convergence":0.32009305081357237,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.00038294716826806107,"delta_db":-0.06914713322503668,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0012690915692200508,"delta_db":-0.37215420078249745,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.000532230819220905,"delta_db":-1.0409544909050197,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":8.500933921480877e-05,"delta_db":-0.7606715992066801,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":4.747709814101176e-05,"delta_db":-1.8314169237334021,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.00014308553507665723,"delta_db":-0.8073821912706927,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.8136210898538879,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.5533497318289057,"pitch":null,"output_rms_dbfs":-26.090928706349313,"output_peak":0.4061483144760132,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.707836100951146},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76640,"output_samples":76640,"intelligibility":{"estoi":{"value":0.8890407506226273,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05285925663037164,"rms_dbfs":-25.537578974520407,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.031195842748737,"mean":-0.00030972872994886806,"zero_energy":false},"output":{"rms":0.04959678950437713,"rms_dbfs":-26.090928706349313,"peak":0.4061483144760132,"peak_dbfs":-7.826306895141323,"crest_factor_db":18.26462181120799,"mean":-1.5756441801899853e-05,"zero_energy":false},"rms_difference_db":-0.5533497318289058},"full_scale":{"reference":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":479,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":1.8146353091946925e-06,"output_rms_dbfs":-57.412106428695054,"output_zero_energy":false,"output_peak":0.005246927496045828,"output_energy_fraction":0.0001216672818461506,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":472,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":66,"both_estimated_voiced_frames":59,"reference_voiced_output_unvoiced_frames":14,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.8082191780821918,"absolute_difference_cents_median":3.8884610061171268,"absolute_difference_cents_p95":14.192777823034207,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00018226521943209611,"log_rms_mae_db":1.4113102003827263,"rms_envelope_nrmse":0.12354854723731332,"envelope_correlation":0.9882240571838147,"rms_gain_error_db":-0.5533442608181667,"reference_mean_power":0.0027970153309965396,"output_mean_power":0.002462410309365458},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.029329446119484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010840604762391139,"output_power":0.000872299576814395,"reference_active":true,"delta_db":-0.9438784898407647},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017643244749495566,"output_power":0.0001433779283648414,"reference_active":true,"delta_db":-0.9009615768281558},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.741521666729805e-05,"output_power":4.055202993145766e-05,"reference_active":true,"delta_db":-0.6790514067190446}],"active_band_count":3,"active_band_mae_db":0.8412971577959883}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001676463927845064,"log_rms_mae_db":1.1353156064559695,"rms_envelope_nrmse":0.10858323034467841,"envelope_correlation":0.991220994392175,"rms_gain_error_db":-0.5533524272697946,"reference_mean_power":0.0028146416719843875,"output_mean_power":0.0024779233638166564},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.889944424347358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010453379933036454,"output_power":0.0008302122798691678,"reference_active":true,"delta_db":-1.0006758247703926},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.750212405766423e-05,"output_power":5.6031445580199995e-05,"reference_active":true,"delta_db":-0.8088561161799994},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9228813706071634e-06,"output_power":2.9243962304687196e-06,"reference_active":true,"delta_db":0.0022502618768376337}],"active_band_count":3,"active_band_mae_db":0.6039274009424099}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00017586243391937327,"log_rms_mae_db":1.8479964695747262,"rms_envelope_nrmse":0.1641003706267886,"envelope_correlation":0.9849019820277491,"rms_gain_error_db":-0.8142713723844243,"reference_mean_power":0.0014437040494073058,"output_mean_power":0.0011968811780139932},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":1.5100208608044743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242573938618173,"output_power":0.0007120142136505099,"reference_active":true,"delta_db":-1.133042701544591},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001371496531392223,"output_power":9.633705418454163e-05,"reference_active":true,"delta_db":-1.5340135135168511},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2803310723594984e-05,"output_power":2.190536187835772e-05,"reference_active":true,"delta_db":-1.7536724559555323}],"active_band_count":3,"active_band_mae_db":1.4735762236723247}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001539617967365963,"log_rms_mae_db":1.4139127542931171,"rms_envelope_nrmse":0.14820603792306766,"envelope_correlation":0.9883637487992174,"rms_gain_error_db":-0.8142720470651109,"reference_mean_power":0.0014528118102568712,"output_mean_power":0.0012044316433661069},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.4301118553576922e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009152950896245062,"output_power":0.0006972886057236239,"reference_active":true,"delta_db":-1.1814856393561877},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.331274162853222e-05,"output_power":3.6252134844651664e-05,"reference_active":true,"delta_db":-1.6749743012219707},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0543911334413043e-06,"output_power":1.4056984148678125e-06,"reference_active":true,"delta_db":-1.6479097691663718}],"active_band_count":3,"active_band_mae_db":1.5014565699148434}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":1.6396263610688282,"rms_envelope_nrmse":0.17216723734189118,"envelope_correlation":0.9873208148928665,"rms_gain_error_db":-1.052106405442324,"reference_mean_power":0.0001541240756859989,"output_mean_power":0.00012096503195873169},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1034128857485058e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011785014167056142,"output_power":8.990846799084849e-05,"reference_active":true,"delta_db":-1.1752951143594543},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.031054512020665e-06,"output_power":8.107933385749666e-06,"reference_active":true,"delta_db":0.04137598143222291},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6793453202028483e-06,"output_power":2.1306411195569057e-06,"reference_active":true,"delta_db":-0.9951838581035494}],"active_band_count":3,"active_band_mae_db":0.7372849846317422}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":1.2829222804159104,"rms_envelope_nrmse":0.1500979889883226,"envelope_correlation":0.9925639735794198,"rms_gain_error_db":-1.052114093947121,"reference_mean_power":0.00015509566252518503,"output_mean_power":0.00012172737112523307},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.037312640211567e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011306875179582826,"output_power":8.688642235174347e-05,"reference_active":true,"delta_db":-1.143906828553072},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4410710139131097e-06,"output_power":3.3198503826396708e-06,"reference_active":true,"delta_db":-0.15575123737518362},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5029543761566635e-07,"output_power":9.173187667889843e-08,"reference_active":true,"delta_db":-2.1442551876905265}],"active_band_count":3,"active_band_mae_db":1.147971084539594}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00013414824565394508,"log_rms_mae_db":1.5732599436008057,"rms_envelope_nrmse":0.21173580278807458,"envelope_correlation":0.9785588899766279,"rms_gain_error_db":-0.9502153041963153,"reference_mean_power":0.0001730034207748909,"output_mean_power":0.00013900587634631083},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1522167661145565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012025367220488917,"output_power":9.088665845282865e-05,"reference_active":true,"delta_db":-1.2159821095973449},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5558902161138395e-05,"output_power":2.9134690745574467e-05,"reference_active":true,"delta_db":0.5686821705706281},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5309687441279477e-05,"output_power":2.010099849468519e-05,"reference_active":true,"delta_db":1.182513067083404}],"active_band_count":3,"active_band_mae_db":0.9890591157504591}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":1.4526121813485695,"rms_envelope_nrmse":0.1796643889454336,"envelope_correlation":0.9860596179092397,"rms_gain_error_db":-0.9502197011465946,"reference_mean_power":0.00017409446216015185,"output_mean_power":0.0001398823715169501},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.085538754133612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012977220791713165,"output_power":0.00010505364726226256,"reference_active":true,"delta_db":-0.917705588126708},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3304937813307312e-05,"output_power":1.6470440026866286e-05,"reference_active":true,"delta_db":0.9269235298554026},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0464863211082628e-06,"output_power":1.5362933807848662e-06,"reference_active":true,"delta_db":1.6674060343644617}],"active_band_count":3,"active_band_mae_db":1.1706783841155242}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.5033111572265625,"predicted_mos":3.5033111572265625,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":211456,"input_sample_rate":44100,"input_seconds":4.794920634920635,"input_float32_sha256":"7b8564fa04c26386b79b3d0e3eb59a293044dbbfaee1e261ea04d1f4b18b68ab","internal_samples":76719,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6642507719989226},"started_at":"2026-10-13T07:34:42.546002+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.007194157999037998,"native_input_samples":211239,"native_input_sha256":"9a37a9b22737d1265fa60598c61e47e0aecb264e43bbd95dfd3c3ca92092fa0f","output_sha256":"5501d4b8e297e1eeb29f9c2830305c4de515126d0b70722e38b5eedfdc4b7160","output_samples":211456,"length_delta_samples":217,"analysis_s":0.11235858899817686,"synthesis_s":26.09639457799858,"total_s":26.208753166996757,"rtf":5.47155598475924,"output_peak":0.44412511587142944,"output_rms":0.04631953645166484,"finished_at":"2026-10-13T07:35:08.782298+00:00","evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"5501d4b8e297e1eeb29f9c2830305c4de515126d0b70722e38b5eedfdc4b7160","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":76640,"output_samples":76719,"length_delta_samples":79,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.67946421357545,"output_peak":0.43136216916385084},"common_prefix":{"samples":76640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9989702681213258,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":10.330484942219169,"spectral_convergence":0.5016004005298499,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.829778970744037,"spectral_convergence":0.4520134931565819,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":9.642382111262112,"spectral_convergence":0.5247126829711396,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.0004361179348570534,"delta_db":0.4955037113952515,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0008626808317356094,"delta_db":-2.0485823073405305,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0005502564582746026,"delta_db":-0.8963031967636886,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":0.0001081444956039158,"delta_db":0.2847061908454911,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":9.018906184087641e-05,"delta_db":0.9552801050695037,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.00010291931672707592,"delta_db":-2.238370567402488,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.1531243464694922,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1374141187598839,"pitch":null,"output_rms_dbfs":-26.674993093280293,"output_peak":0.43136216916385084,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.93421534140844},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76640,"output_samples":76640,"intelligibility":{"estoi":{"value":0.7143607083160431,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05285925663037164,"rms_dbfs":-25.537578974520407,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.031195842748737,"mean":-0.00030972872994886806,"zero_energy":false},"output":{"rms":0.04637141467584457,"rms_dbfs":-26.674993093280293,"peak":0.43136216916385084,"peak_dbfs":-7.303158911318284,"crest_factor_db":19.371834181962008,"mean":-6.758997661510688e-05,"zero_energy":false},"rms_difference_db":-1.1374141187598852},"full_scale":{"reference":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":479,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":9.366313941644699e-07,"output_rms_dbfs":-60.28431289556403,"output_zero_energy":false,"output_peak":0.003521434180730071,"output_energy_fraction":7.183888740713471e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":472,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":66,"both_estimated_voiced_frames":54,"reference_voiced_output_unvoiced_frames":19,"reference_unvoiced_output_voiced_frames":12,"paired_voiced_fraction":0.7397260273972602,"absolute_difference_cents_median":9.285934723347125,"absolute_difference_cents_p95":27.660320160790718,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00018226521943209611,"log_rms_mae_db":2.784304104800037,"rms_envelope_nrmse":0.18673721473144092,"envelope_correlation":0.9793307033130763,"rms_gain_error_db":-1.137411384869582,"reference_mean_power":0.0027970153309965396,"output_mean_power":0.00215255228092828},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.029329446119484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010840604762391139,"output_power":0.0009230755014874955,"reference_active":true,"delta_db":-0.6981628586444129},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017643244749495566,"output_power":0.0001395654333849691,"reference_active":true,"delta_db":-1.0180059024694001},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.741521666729805e-05,"output_power":2.720376578612474e-05,"reference_active":true,"delta_db":-2.412887123570269}],"active_band_count":3,"active_band_mae_db":1.3763519615613606}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001676463927845064,"log_rms_mae_db":2.5227120133233534,"rms_envelope_nrmse":0.15881904154596305,"envelope_correlation":0.9889419313164014,"rms_gain_error_db":-1.1373967640480769,"reference_mean_power":0.0028146416719843875,"output_mean_power":0.0021661246117685978},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.889944424347358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010453379933036454,"output_power":0.0008908022042918494,"reference_active":true,"delta_db":-0.6947545229510438},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.750212405766423e-05,"output_power":5.70430290054124e-05,"reference_active":true,"delta_db":-0.7311486013699322},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9228813706071634e-06,"output_power":2.098621896610982e-06,"reference_active":true,"delta_db":-1.4387698934777673}],"active_band_count":3,"active_band_mae_db":0.9548910059329145}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00017586243391937327,"log_rms_mae_db":2.8749779233522164,"rms_envelope_nrmse":0.21505237953462455,"envelope_correlation":0.9758573741013633,"rms_gain_error_db":-1.2411511920072793,"reference_mean_power":0.0014437040494073058,"output_mean_power":0.0010848334189073197},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":1.5100208608044743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242573938618173,"output_power":0.0007060488306282409,"reference_active":true,"delta_db":-1.1695819544623551},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001371496531392223,"output_power":0.00011598961963635921,"reference_active":true,"delta_db":-0.7277558932350415},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2803310723594984e-05,"output_power":2.5329069487474694e-05,"reference_active":true,"delta_db":-1.1229844231618287}],"active_band_count":3,"active_band_mae_db":1.0067740902864084}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001539617967365963,"log_rms_mae_db":2.2497079136801608,"rms_envelope_nrmse":0.18429697890315228,"envelope_correlation":0.9853703144501631,"rms_gain_error_db":-1.2411514516483968,"reference_mean_power":0.0014528118102568712,"output_mean_power":0.0010916771408744882},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.4301118553576922e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009152950896245062,"output_power":0.0007021225478247511,"reference_active":true,"delta_db":-1.1514821235930255},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.331274162853222e-05,"output_power":4.828756585484931e-05,"reference_active":true,"delta_db":-0.42995703415871495},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0543911334413043e-06,"output_power":1.5871956711095579e-06,"reference_active":true,"delta_db":-1.1205266169436707}],"active_band_count":3,"active_band_mae_db":0.9006552582318038}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":3.471982300459974,"rms_envelope_nrmse":0.2828067618179847,"envelope_correlation":0.9532136886417687,"rms_gain_error_db":0.4645919053726859,"reference_mean_power":0.0001541240756859989,"output_mean_power":0.00017152588769256283},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1034128857485058e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011785014167056142,"output_power":0.0001535170841152812,"reference_active":true,"delta_db":1.1482660393272108},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.031054512020665e-06,"output_power":1.2871667105708896e-05,"reference_active":true,"delta_db":2.0486222551081257},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6793453202028483e-06,"output_power":1.5445113398984338e-06,"reference_active":true,"delta_db":-2.3923758882338593}],"active_band_count":3,"active_band_mae_db":1.8630880608897318}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":3.1104499396194805,"rms_envelope_nrmse":0.22680371633576268,"envelope_correlation":0.9696457735191725,"rms_gain_error_db":0.46460747972927163,"reference_mean_power":0.00015509566252518503,"output_mean_power":0.00017260779326750127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.037312640211567e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011306875179582826,"output_power":0.0001487825138409745,"reference_active":true,"delta_db":1.1920929443570139},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4410710139131097e-06,"output_power":6.106701844426205e-06,"reference_active":true,"delta_db":2.4911308106450134},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5029543761566635e-07,"output_power":1.0678694664326955e-07,"reference_active":true,"delta_db":-1.4842762838075512}],"active_band_count":3,"active_band_mae_db":1.722500012936526}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00013414824565394508,"log_rms_mae_db":2.7682960980166187,"rms_envelope_nrmse":0.28468245058947494,"envelope_correlation":0.9682265002445075,"rms_gain_error_db":-1.8132944094684245,"reference_mean_power":0.0001730034207748909,"output_mean_power":0.00011395286522111217},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1522167661145565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012025367220488917,"output_power":7.565377846951963e-05,"reference_active":true,"delta_db":-2.012677239625099},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5558902161138395e-05,"output_power":1.4460470305910112e-05,"reference_active":true,"delta_db":-2.4735977752248854},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5309687441279477e-05,"output_power":6.344092393041543e-06,"reference_active":true,"delta_db":-3.8259682508045856}],"active_band_count":3,"active_band_mae_db":2.77074775521819}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":2.9891383364009125,"rms_envelope_nrmse":0.2327210361691262,"envelope_correlation":0.9867606783758082,"rms_gain_error_db":-1.8132851109554942,"reference_mean_power":0.00017409446216015185,"output_mean_power":0.00011467175139689518},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.085538754133612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012977220791713165,"output_power":7.894716099961536e-05,"reference_active":true,"delta_db":-2.1584517662959253},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3304937813307312e-05,"output_power":7.1971626675392115e-06,"reference_active":true,"delta_db":-2.6685153046141084},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0464863211082628e-06,"output_power":5.00447314751467e-07,"reference_active":true,"delta_db":-3.203751926372952}],"active_band_count":3,"active_band_mae_db":2.6769063324276616}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"golf","status":"measured","mos":{"status":"measured","value":3.1683907508850098,"predicted_mos":3.1683907508850098,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":114720,"input_sample_rate":24000,"input_seconds":4.78,"input_float32_sha256":"9107dee0704ff0208e8072da9b7f3441b3beb6c0eed1a5ab073d0ea411a3aaec","internal_samples":76480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3019941640013712},"started_at":"2026-10-13T05:03:13.767099+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002782597999612335,"native_input_samples":114960,"native_input_sha256":"3a7796dcdee481595881139d73d70685f0c2436c483ed3b8b03763eef328cd2e","output_sha256":"6424e7f09f377728a7e62c0863bc2424b4a69774fa9be6da361cbb0beb9ebfde","output_samples":114720,"length_delta_samples":-240,"analysis_s":0.48831340100059606,"synthesis_s":0.13238441099929332,"total_s":0.6206978119998894,"rtf":0.12958200668056147,"output_peak":0.3932121992111206,"output_rms":0.046229369540011105,"finished_at":"2026-10-13T05:03:14.398488+00:00","analysis_stages":{"external_f0_s":0.04873031399984029},"evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"6424e7f09f377728a7e62c0863bc2424b4a69774fa9be6da361cbb0beb9ebfde","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":76640,"output_samples":76480,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-26.697734056774788,"output_peak":0.3883201165001124},"common_prefix":{"samples":76480,"sample_rate":16000,"reference_coverage":0.9979123173277662,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76480,"output_samples":76480,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.099362599507586,"spectral_convergence":0.6343754976549502,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.474004328367047,"spectral_convergence":0.351880373777685,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.753343930015956,"spectral_convergence":0.3833384991254251,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003898982007711312,"output_power":0.0003129429143797363,"delta_db":-0.9548610834107967,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013855265888015845,"output_power":0.0010666827686067824,"delta_db":-1.1357958500756302,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.000677802882927138,"output_power":0.0005283444877528469,"delta_db":-1.0818623094325257,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010149411634562094,"output_power":7.296824678880239e-05,"delta_db":-1.4330695518885381,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.253244065385576e-05,"output_power":5.3406820193741785e-05,"delta_db":-1.3293557020440503,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017267956408843942,"output_power":0.00010473263034760769,"delta_db":-2.17158932729172,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.3510889706905436,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.1692116794132716,"pitch":null,"output_rms_dbfs":-26.697734056774788,"output_peak":0.3883201165001124,"reference_rms_dbfs":-25.52852237736152,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.108903619296863},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76480,"output_samples":76480,"intelligibility":{"estoi":{"value":0.870648232171891,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05291440062580101,"rms_dbfs":-25.52852237736152,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.022139245589848,"mean":-0.00030733731002488394,"zero_energy":false},"output":{"rms":0.04625016614099367,"rms_dbfs":-26.697734056774788,"peak":0.3883201165001124,"peak_dbfs":-8.216202214384325,"crest_factor_db":18.481531842390467,"mean":0.0006845278804263216,"zero_energy":false},"rms_difference_db":-1.1692116794132694},"full_scale":{"reference":{"sample_count":76480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76480,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":478,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":2.5543736759698866e-06,"output_rms_dbfs":-55.92715570055805,"output_zero_energy":false,"output_peak":0.008416587315812586,"output_energy_fraction":0.00019735904287292055,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":471,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":94,"both_estimated_voiced_frames":65,"reference_voiced_output_unvoiced_frames":8,"reference_unvoiced_output_voiced_frames":29,"paired_voiced_fraction":0.8904109589041096,"absolute_difference_cents_median":5.631423682265124,"absolute_difference_cents_p95":20.693254375518485,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76480,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":955,"rms_floor":0.00018226521943209611,"log_rms_mae_db":2.3207779730002676,"rms_envelope_nrmse":0.18722092104161836,"envelope_correlation":0.9796113812721056,"rms_gain_error_db":-1.1692102463677654,"reference_mean_power":0.0028028632407271696,"output_mean_power":0.0021413165973962647},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":955,"frequency_resolution_hz":0.20942408376963348,"power_floor":2.026199853247066e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010802413614656484,"output_power":0.000926828844891923,"reference_active":true,"delta_db":-0.6652126050730073},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017611763083258479,"output_power":0.00013978328186145487,"reference_active":true,"delta_db":-1.0034760190072343},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.749152157235806e-05,"output_power":2.326361887426262e-05,"reference_active":true,"delta_db":-3.0993880985757913}],"active_band_count":3,"active_band_mae_db":1.5893589075520111}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":0.0001676463927845064,"log_rms_mae_db":1.756402304019073,"rms_envelope_nrmse":0.14925601248132173,"envelope_correlation":0.9930262513076924,"rms_gain_error_db":-1.1692208137873186,"reference_mean_power":0.0028205662041198903,"output_mean_power":0.002154835970269198},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":1.886893768036947e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010352118271483802,"output_power":0.0008880976084863083,"reference_active":true,"delta_db":-0.6656852453292222},{"low_hz":8.0,"high_hz":20.0,"reference_power":7.338823053935517e-05,"output_power":5.9677983094419515e-05,"reference_active":true,"delta_db":-0.8981227948007187},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9782609222667115e-06,"output_power":1.5131850049873e-06,"reference_active":true,"delta_db":-2.940707141893742}],"active_band_count":3,"active_band_mae_db":1.501505060674561}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":955,"rms_floor":0.00017586243391937327,"log_rms_mae_db":2.3378823241869124,"rms_envelope_nrmse":0.18030291484911692,"envelope_correlation":0.9820420682890676,"rms_gain_error_db":-0.8969866441740599,"reference_mean_power":0.0014467272335078722,"output_mean_power":0.0011767602568328315},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":955,"frequency_resolution_hz":0.20942408376963348,"power_floor":1.5076896162063482e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009218630030747751,"output_power":0.0008191216183496533,"reference_active":true,"delta_db":-0.5131799784824685},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00013693792933404243,"output_power":0.00011469042913845858,"reference_active":true,"delta_db":-0.7699657878054484},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.255143774832526e-05,"output_power":1.5512186536051762e-05,"reference_active":true,"delta_db":-3.218971569162016}],"active_band_count":3,"active_band_mae_db":1.5007057784833109}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":0.0001539617967365963,"log_rms_mae_db":1.5738840478478318,"rms_envelope_nrmse":0.1305626163684088,"envelope_correlation":0.9942394642410959,"rms_gain_error_db":-0.8969876773888099,"reference_mean_power":0.0014558733124644778,"output_mean_power":0.0011841993472073524},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":1.4278365698641378e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009077393461892654,"output_power":0.0008007665874525676,"reference_active":true,"delta_db":-0.5445521684897573},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.7913402260113924e-05,"output_power":5.0524441239439584e-05,"reference_active":true,"delta_db":-0.5927756008308263},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.106081994357459e-06,"output_power":1.0797467490793453e-06,"reference_active":true,"delta_db":-2.9015337006836868}],"active_band_count":3,"active_band_mae_db":1.3462871566680903}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":955,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":1.9956453327343069,"rms_envelope_nrmse":0.23467656006101084,"envelope_correlation":0.9731170128570004,"rms_gain_error_db":-1.365893369829729,"reference_mean_power":0.00015444652835375333,"output_mean_power":0.00011276876226496168},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":955,"frequency_resolution_hz":0.20942408376963348,"power_floor":2.100419289066715e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001175020824645962,"output_power":8.810878218507675e-05,"reference_active":true,"delta_db":-1.250263649981715},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.029699331836202e-06,"output_power":6.419963419866976e-06,"reference_active":true,"delta_db":-0.9716673011737387},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6655742055696503e-06,"output_power":1.8160016663394086e-06,"reference_active":true,"delta_db":-1.6667453444510074}],"active_band_count":3,"active_band_mae_db":1.2962254318688202}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":1.5110347049896757,"rms_envelope_nrmse":0.17042087463312033,"envelope_correlation":0.9934507192801906,"rms_gain_error_db":-1.36589380759483,"reference_mean_power":0.00015542223832233455,"output_mean_power":0.00011348116311965024},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":2.0344342834319836e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011261567857477966,"output_power":8.521400526283617e-05,"reference_active":true,"delta_db":-1.210878793814958},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.5271793407535564e-06,"output_power":2.616407513187919e-06,"reference_active":true,"delta_db":-1.2972215452718452},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.52434631252955e-07,"output_power":9.818207166131971e-08,"reference_active":true,"delta_db":-1.9105145295890311}],"active_band_count":3,"active_band_mae_db":1.4728716228919447}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":955,"rms_floor":0.00013414824565394508,"log_rms_mae_db":2.01150429362792,"rms_envelope_nrmse":0.298418952858785,"envelope_correlation":0.9699578391108493,"rms_gain_error_db":-2.1355433844415197,"reference_mean_power":0.00017336544307927846,"output_mean_power":0.0001060249790790332},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":955,"frequency_resolution_hz":0.20942408376963348,"power_floor":2.1634350548658424e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012087511002091672,"output_power":7.477651097177656e-05,"reference_active":true,"delta_db":-2.085716850228092},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5729860656796774e-05,"output_power":1.4558992731400076e-05,"reference_active":true,"delta_db":-2.473061050290987},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5359449681051777e-05,"output_power":6.594470920219306e-06,"reference_active":true,"delta_db":-3.6719569800993126}],"active_band_count":3,"active_band_mae_db":2.743578293539464}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76480,"uncovered_tail_samples":0,"metrics":{"frame_count":949,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":2.1662698895073054,"rms_envelope_nrmse":0.2488748124772964,"envelope_correlation":0.9892962762763653,"rms_gain_error_db":-2.13554017753291,"reference_mean_power":0.00017446111512375764,"output_mean_power":0.00010669513722720174},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":949,"frequency_resolution_hz":0.2107481559536354,"power_floor":2.0964990472810277e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0001302361540056499,"output_power":7.865001951366426e-05,"reference_active":true,"delta_db":-2.190327279048093},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3563092093752421e-05,"output_power":6.557740932134017e-06,"reference_active":true,"delta_db":-3.156044550516694},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0627470575574723e-06,"output_power":5.608243155060837e-07,"reference_active":true,"delta_db":-2.776030762500776}],"active_band_count":3,"active_band_mae_db":2.7074675306885205}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"sifigan","status":"measured","mos":{"status":"measured","value":2.901871681213379,"predicted_mos":2.901871681213379,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":115080,"input_sample_rate":24000,"input_seconds":4.795,"input_float32_sha256":"0fc360a8a4d921d6f5c67a7035baac18be97ea971a41a8460306f8582b8a33d1","internal_samples":76720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.574485630997515},"started_at":"2026-10-13T07:48:49.380418+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00247558200135245,"native_input_samples":114960,"native_input_sha256":"3a7796dcdee481595881139d73d70685f0c2436c483ed3b8b03763eef328cd2e","output_sha256":"f318c6eb6c1db8c7ec0b4ffc3c8ca47d3ab33edaebdc0e4257dc719721b99569","output_samples":115080,"length_delta_samples":120,"analysis_s":2.006824780997704,"synthesis_s":9.949690845001896,"total_s":11.9565156259996,"rtf":2.4961410492692275,"output_peak":0.468325138092041,"output_rms":0.052061895839525946,"finished_at":"2026-10-13T07:49:01.346458+00:00","analysis_stages":{"external_f0_s":1.3327548740016937},"evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"f318c6eb6c1db8c7ec0b4ffc3c8ca47d3ab33edaebdc0e4257dc719721b99569","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":76640,"output_samples":76720,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.66569571572721,"output_peak":0.46845817330268397},"common_prefix":{"samples":76640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9989572471324296,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.797540106566134,"spectral_convergence":0.6726390372346901,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.327694715406396,"spectral_convergence":0.36813905222974447,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.5330374360583106,"spectral_convergence":0.35067140578167477,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.0005452839284299932,"delta_db":1.4656912181148192,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0012963632824554566,"delta_db":-0.27981657314412167,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0006286487034334374,"delta_db":-0.3178744400570472,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":9.036497823837068e-05,"delta_db":-0.49533651958623626,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":7.512498469257869e-05,"delta_db":0.16158537668289547,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":7.992455021507629e-05,"delta_db":-3.3365375013590204,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.0094736048240234,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.12358642704833044,"pitch":null,"output_rms_dbfs":-25.661165401568734,"output_peak":0.46845817330268397,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.886090752676947},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76640,"output_samples":76640,"intelligibility":{"estoi":{"value":0.8839374294727246,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05285925663037164,"rms_dbfs":-25.537578974520407,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.031195842748737,"mean":-0.00030972872994886806,"zero_energy":false},"output":{"rms":0.05211247861563401,"rms_dbfs":-25.661165401568734,"peak":0.46845817330268397,"peak_dbfs":-6.586583588149888,"crest_factor_db":19.07458181341885,"mean":2.9916358102566205e-05,"zero_energy":false},"rms_difference_db":-0.1235864270483269},"full_scale":{"reference":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":479,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":7.200153622007816e-07,"output_rms_dbfs":-61.42658237391154,"output_zero_energy":false,"output_peak":0.00728910114003253,"output_energy_fraction":4.3727019948935234e-05,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":472,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":98,"both_estimated_voiced_frames":66,"reference_voiced_output_unvoiced_frames":7,"reference_unvoiced_output_voiced_frames":32,"paired_voiced_fraction":0.9041095890410958,"absolute_difference_cents_median":4.013060363711638,"absolute_difference_cents_p95":17.814340636316324,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00018226521943209611,"log_rms_mae_db":3.2613438837659543,"rms_envelope_nrmse":0.19986696176672555,"envelope_correlation":0.9610198911231366,"rms_gain_error_db":-0.12357867944958026,"reference_mean_power":0.0027970153309965396,"output_mean_power":0.0027185478332937955},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.029329446119484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010840604762391139,"output_power":0.001082193933406812,"reference_active":true,"delta_db":-0.007484157440674808},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017643244749495566,"output_power":0.00013034575832046518,"reference_active":true,"delta_db":-1.314815557254005},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.741521666729805e-05,"output_power":2.357795164466679e-05,"reference_active":true,"delta_db":-3.0341166675824347}],"active_band_count":3,"active_band_mae_db":1.4521387940923713}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001676463927845064,"log_rms_mae_db":2.7328185292836413,"rms_envelope_nrmse":0.13083811794526887,"envelope_correlation":0.9821775618079028,"rms_gain_error_db":-0.12355278785779206,"reference_mean_power":0.0028146416719843875,"output_mean_power":0.002735695994237489},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.889944424347358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010453379933036454,"output_power":0.0010302764671240464,"reference_active":true,"delta_db":-0.06302955241796826},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.750212405766423e-05,"output_power":5.3086885112106004e-05,"reference_active":true,"delta_db":-1.0433019504869994},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9228813706071634e-06,"output_power":1.445875020917913e-06,"reference_active":true,"delta_db":-3.0568043436861387}],"active_band_count":3,"active_band_mae_db":1.387711948863702}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00017586243391937327,"log_rms_mae_db":2.3521680799390756,"rms_envelope_nrmse":0.21631915103445432,"envelope_correlation":0.965381329333887,"rms_gain_error_db":-0.40355639054444503,"reference_mean_power":0.0014437040494073058,"output_mean_power":0.0013155959733404815},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":1.5100208608044743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242573938618173,"output_power":0.0008585011667571799,"reference_active":true,"delta_db":-0.32052043791270113},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001371496531392223,"output_power":0.00011183316625088795,"reference_active":true,"delta_db":-0.8862409264619979},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2803310723594984e-05,"output_power":1.901528914716575e-05,"reference_active":true,"delta_db":-2.3681474405392096}],"active_band_count":3,"active_band_mae_db":1.1916362683046362}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001539617967365963,"log_rms_mae_db":1.4120648255369617,"rms_envelope_nrmse":0.12894435933391435,"envelope_correlation":0.9880247832150212,"rms_gain_error_db":-0.4035565759878694,"reference_mean_power":0.0014528118102568712,"output_mean_power":0.0013238954942096075},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.4301118553576922e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009152950896245062,"output_power":0.0008395249781200993,"reference_active":true,"delta_db":-0.3752751037298101},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.331274162853222e-05,"output_power":4.441657600503431e-05,"reference_active":true,"delta_db":-0.7928594034991588},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0543911334413043e-06,"output_power":1.5389997907948361e-06,"reference_active":true,"delta_db":-1.2544457122339625}],"active_band_count":3,"active_band_mae_db":0.8075267398209771}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":2.0470313768273574,"rms_envelope_nrmse":0.236225069668348,"envelope_correlation":0.9615889261712839,"rms_gain_error_db":-0.16686582606351527,"reference_mean_power":0.0001541240756859989,"output_mean_power":0.00014831459919097592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1034128857485058e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011785014167056142,"output_power":0.00011348476706467866,"reference_active":true,"delta_db":-0.16392538365633122},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.031054512020665e-06,"output_power":7.472424806707539e-06,"reference_active":true,"delta_db":-0.31311020229545716},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6793453202028483e-06,"output_power":2.32390470899551e-06,"reference_active":true,"delta_db":-0.6181037411919863}],"active_band_count":3,"active_band_mae_db":0.3650464423812583}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":1.190897283866684,"rms_envelope_nrmse":0.11065357991195274,"envelope_correlation":0.9913405383249055,"rms_gain_error_db":-0.16685857274038002,"reference_mean_power":0.00015509566252518503,"output_mean_power":0.0001492498127853338},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.037312640211567e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011306875179582826,"output_power":0.00010927064205434519,"reference_active":true,"delta_db":-0.14839102952873895},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4410710139131097e-06,"output_power":3.1653757466197525e-06,"reference_active":true,"delta_db":-0.3626836488079582},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5029543761566635e-07,"output_power":1.1410471859314835e-07,"reference_active":true,"delta_db":-1.196421930492533}],"active_band_count":3,"active_band_mae_db":0.56916553627641}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00013414824565394508,"log_rms_mae_db":1.6738666375925564,"rms_envelope_nrmse":0.43039943488621113,"envelope_correlation":0.9256235228223401,"rms_gain_error_db":-3.161747358522817,"reference_mean_power":0.0001730034207748909,"output_mean_power":8.353720771746578e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1522167661145565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012025367220488917,"output_power":5.5232584763268355e-05,"reference_active":true,"delta_db":-3.3790297964168974},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5558902161138395e-05,"output_power":5.600700792202807e-06,"reference_active":true,"delta_db":-6.592998236579436},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5309687441279477e-05,"output_power":2.3071414805898484e-06,"reference_active":true,"delta_db":-8.21892096810378}],"active_band_count":3,"active_band_mae_db":6.063649667033371}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":1.0929079188040858,"rms_envelope_nrmse":0.37407090385843395,"envelope_correlation":0.9623116640690519,"rms_gain_error_db":-3.1617560507853666,"reference_mean_power":0.00017409446216015185,"output_mean_power":8.406386460308253e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.085538754133612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012977220791713165,"output_power":5.3814403803037154e-05,"reference_active":true,"delta_db":-3.8228316049931443},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3304937813307312e-05,"output_power":2.481430223004671e-06,"reference_active":true,"delta_db":-7.2931078153072315},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0464863211082628e-06,"output_power":1.522303166060959e-07,"reference_active":true,"delta_db":-8.372324053183604}],"active_band_count":3,"active_band_mae_db":6.496087824494659}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.322415590286255,"predicted_mos":3.322415590286255,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":115200,"input_sample_rate":24000,"input_seconds":4.8,"input_float32_sha256":"371a8c75e6946231050dd96705e5f5a62c8f45ddb966f7931a6181137809b560","internal_samples":76800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0401917069993942},"started_at":"2026-10-13T05:11:43.103537+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0023770789994159713,"native_input_samples":114960,"native_input_sha256":"3a7796dcdee481595881139d73d70685f0c2436c483ed3b8b03763eef328cd2e","output_sha256":"c44a35ae04bc96eef32f0fdc62d0179001f62f3645e85ecad236d2c9d3c050b5","output_samples":115200,"length_delta_samples":240,"analysis_s":0.7334403700006078,"synthesis_s":0.09646959199926641,"total_s":0.8299099619998742,"rtf":0.1732588647181366,"output_peak":0.40891802310943604,"output_rms":0.05094792288380316,"finished_at":"2026-10-13T05:11:43.943745+00:00","analysis_stages":{"external_f0_s":0.06743095899946638},"evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"c44a35ae04bc96eef32f0fdc62d0179001f62f3645e85ecad236d2c9d3c050b5","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":76640,"output_samples":76800,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.853320509619845,"output_peak":0.41292555085055854},"common_prefix":{"samples":76640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9979166666666667,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":4.961332486722377,"spectral_convergence":0.21946779742262232,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.524562742660927,"spectral_convergence":0.22790637545770645,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.001364776030008,"spectral_convergence":0.23854266877818966,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.0003754508875498408,"delta_db":-0.1550044045205766,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.0012782596365484003,"delta_db":-0.3408930363758358,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0006614334412981695,"delta_db":-0.09709319626709588,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":9.202964162218116e-05,"delta_db":-0.41606069482244984,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":6.245009635170957e-05,"delta_db":-0.640927460850001,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.00013397144818032088,"delta_db":-1.0932169908797909,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.4571992972859584,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3066864842629769,"pitch":null,"output_rms_dbfs":-25.844265458783386,"output_peak":0.41292555085055854,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.829086668471104},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76640,"output_samples":76640,"intelligibility":{"estoi":{"value":0.9313261382948103,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05285925663037164,"rms_dbfs":-25.537578974520407,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.031195842748737,"mean":-0.00030972872994886806,"zero_energy":false},"output":{"rms":0.0510254363190617,"rms_dbfs":-25.844265458783383,"peak":0.41292555085055854,"peak_dbfs":-7.682564863693507,"crest_factor_db":18.161700595089876,"mean":5.657480424109797e-05,"zero_energy":false},"rms_difference_db":-0.3066864842629755},"full_scale":{"reference":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":479,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":2.2479177325629564e-06,"output_rms_dbfs":-56.482195867629926,"output_zero_energy":false,"output_peak":0.012617922889000965,"output_energy_fraction":0.0001423962449732726,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":472,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":73,"both_estimated_voiced_frames":68,"reference_voiced_output_unvoiced_frames":5,"reference_unvoiced_output_voiced_frames":5,"paired_voiced_fraction":0.9315068493150684,"absolute_difference_cents_median":1.5416929162526865,"absolute_difference_cents_p95":9.329474648833372,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00018226521943209611,"log_rms_mae_db":1.1982841779805449,"rms_envelope_nrmse":0.10014026800117332,"envelope_correlation":0.9911426001231793,"rms_gain_error_db":-0.30668294949215386,"reference_mean_power":0.0027970153309965396,"output_mean_power":0.002606312889808096},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.029329446119484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010840604762391139,"output_power":0.0010520813320653432,"reference_active":true,"delta_db":-0.13004196137874802},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017643244749495566,"output_power":0.00014936133257556085,"reference_active":true,"delta_db":-0.7234027909837317},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.741521666729805e-05,"output_power":2.7876441613803782e-05,"reference_active":true,"delta_db":-2.3068040351916816}],"active_band_count":3,"active_band_mae_db":1.0534162625180539}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001676463927845064,"log_rms_mae_db":0.6767755208183928,"rms_envelope_nrmse":0.06395140088316112,"envelope_correlation":0.9969000828732612,"rms_gain_error_db":-0.3066794674705503,"reference_mean_power":0.0028146416719843875,"output_mean_power":0.0026227395574877394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.889944424347358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010453379933036454,"output_power":0.0010097724826955926,"reference_active":true,"delta_db":-0.15033203756836244},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.750212405766423e-05,"output_power":6.180472528471809e-05,"reference_active":true,"delta_db":-0.3829575841267081},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9228813706071634e-06,"output_power":1.9100034813428016e-06,"reference_active":true,"delta_db":-1.847770304094051}],"active_band_count":3,"active_band_mae_db":0.7936866419297072}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00017586243391937327,"log_rms_mae_db":1.330249566483747,"rms_envelope_nrmse":0.1292377740284453,"envelope_correlation":0.9880382912432347,"rms_gain_error_db":-0.2864395818280123,"reference_mean_power":0.0014437040494073058,"output_mean_power":0.0013515565339958435},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":1.5100208608044743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242573938618173,"output_power":0.0008965157119770201,"reference_active":true,"delta_db":-0.132350282852923},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001371496531392223,"output_power":0.00011943118063326032,"reference_active":true,"delta_db":-0.6007698805201279},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2803310723594984e-05,"output_power":2.0923518556887283e-05,"reference_active":true,"delta_db":-1.9528295924847923}],"active_band_count":3,"active_band_mae_db":0.8953165852859476}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001539617967365963,"log_rms_mae_db":0.7532475549126962,"rms_envelope_nrmse":0.07672036751071204,"envelope_correlation":0.996092301384418,"rms_gain_error_db":-0.2864396066923613,"reference_mean_power":0.0014528118102568712,"output_mean_power":0.001360082964637596},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.4301118553576922e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009152950896245062,"output_power":0.000880412718031019,"reference_active":true,"delta_db":-0.16874824923177395},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.331274162853222e-05,"output_power":4.843346272342422e-05,"reference_active":true,"delta_db":-0.4168549706389682},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0543911334413043e-06,"output_power":1.5885670238248778e-06,"reference_active":true,"delta_db":-1.1167758898822648}],"active_band_count":3,"active_band_mae_db":0.5674597032510024}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":1.0979604150540838,"rms_envelope_nrmse":0.1070052504228142,"envelope_correlation":0.9937176325233138,"rms_gain_error_db":-0.46344428672932103,"reference_mean_power":0.0001541240756859989,"output_mean_power":0.00013852432939328224},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1034128857485058e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011785014167056142,"output_power":0.00010741501280767029,"reference_active":true,"delta_db":-0.40265124371391925},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.031054512020665e-06,"output_power":6.872872436134315e-06,"reference_active":true,"delta_db":-0.6763429049096331},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6793453202028483e-06,"output_power":1.4247977477463035e-06,"reference_active":true,"delta_db":-2.7427547014638254}],"active_band_count":3,"active_band_mae_db":1.2739162833624593}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":0.7312647484638866,"rms_envelope_nrmse":0.07123546163127187,"envelope_correlation":0.9982227329767585,"rms_gain_error_db":-0.463443125832098,"reference_mean_power":0.00015509566252518503,"output_mean_power":0.00013939761384088766},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.037312640211567e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011306875179582826,"output_power":0.00010315184654786793,"reference_active":true,"delta_db":-0.39865591075604945},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4410710139131097e-06,"output_power":3.1636240528596297e-06,"reference_active":true,"delta_db":-0.3650876651692593},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5029543761566635e-07,"output_power":7.925518114529656e-08,"reference_active":true,"delta_db":-2.7791813436645016}],"active_band_count":3,"active_band_mae_db":1.1809749731966035}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00013414824565394508,"log_rms_mae_db":0.9182610669214784,"rms_envelope_nrmse":0.19334867981337592,"envelope_correlation":0.9849211860948632,"rms_gain_error_db":-1.0838136207182485,"reference_mean_power":0.0001730034207748909,"output_mean_power":0.00013479485888324638},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1522167661145565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012025367220488917,"output_power":9.740467967049514e-05,"reference_active":true,"delta_db":-0.9151852491989722},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5558902161138395e-05,"output_power":1.3908675903980195e-05,"reference_active":true,"delta_db":-2.6425640807625896},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5309687441279477e-05,"output_power":6.901650463742689e-06,"reference_active":true,"delta_db":-3.460133638102122}],"active_band_count":3,"active_band_mae_db":2.339294322687895}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":0.9537462496748783,"rms_envelope_nrmse":0.14505727869096505,"envelope_correlation":0.9952806722121497,"rms_gain_error_db":-1.0838126824488277,"reference_mean_power":0.00017409446216015185,"output_mean_power":0.0001356449683136198},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.085538754133612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012977220791713165,"output_power":9.995250536862045e-05,"reference_active":true,"delta_db":-1.133880092540297},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3304937813307312e-05,"output_power":6.511223299229994e-06,"reference_active":true,"delta_db":-3.1035025949898367},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0464863211082628e-06,"output_power":4.0090951889414434e-07,"reference_active":true,"delta_db":-4.166871880024434}],"active_band_count":3,"active_band_mae_db":2.8014181891848557}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0002","method":"magphase","status":"measured","mos":{"status":"measured","value":3.9954278469085693,"predicted_mos":3.9954278469085693,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":231088,"input_sample_rate":48000,"input_seconds":4.814333333333333,"input_float32_sha256":"43c020f05fea3e4eaa63813036d1776e834759837e19a79f232cb5fba6f8b5a3","internal_samples":77030,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3294198389994563},"started_at":"2026-10-13T05:19:24.750777+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.007780996000292362,"native_input_samples":229920,"native_input_sha256":"9376e92e766da59de10016eeaefebe2c3c62319cf64a6068dcaa31d0b599cca8","output_sha256":"07820f4b59b01e7c60cf0a096efc020a2ec1ca0fee4de34a203adf0262891030","output_samples":231088,"length_delta_samples":1168,"analysis_s":2.3997550359999877,"synthesis_s":0.8082466260002548,"total_s":3.2080016620002425,"rtf":0.6697289482255203,"output_peak":0.43060737271449606,"output_rms":0.050471713945496086,"finished_at":"2026-10-13T05:19:27.976967+00:00","evaluation_input":{"reference":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"07820f4b59b01e7c60cf0a096efc020a2ec1ca0fee4de34a203adf0262891030","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":76640,"output_samples":77030,"length_delta_samples":390,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-25.935100859672513,"output_peak":0.4309110802244212},"common_prefix":{"samples":76640,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9949370375178502,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":76640,"output_samples":76640,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.875607758544913,"spectral_convergence":0.6839730048939443,"spectral_convergence_status":"ok","amplitude_floor":1.4500777065259659e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.742682607347823,"spectral_convergence":0.4943947186341352,"spectral_convergence_status":"ok","amplitude_floor":1.1426990309647766e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.216397580700477,"spectral_convergence":0.4395446033812615,"spectral_convergence_status":"ok","amplitude_floor":7.641152847432081e-07}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0003890931416418852,"output_power":0.0003704885153751876,"delta_db":-0.21278825767118215,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0013826376510160613,"output_power":0.001240644422552385,"delta_db":-0.4706105164037958,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.0006763873367355198,"output_power":0.0006123871514081354,"delta_db":-0.4316939805809317,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00010128235718512909,"output_power":0.00010062135842188904,"delta_db":-0.028436241518358458,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":7.238121331616846e-05,"output_power":7.51062836237802e-05,"delta_db":0.16050414091648757,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00017231931162072514,"output_power":0.00016341991949356222,"delta_db":-0.23028958828582718,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.2557204542294305,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.3754981581363397,"pitch":null,"output_rms_dbfs":-25.913077132656746,"output_peak":0.4309110802244212,"reference_rms_dbfs":-25.537578974520407,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.944895982197738},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":76640,"output_samples":76640,"intelligibility":{"estoi":{"value":0.7647694748425676,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.05285925663037164,"rms_dbfs":-25.537578974520407,"peak":0.42138671875,"peak_dbfs":-7.506383131771671,"crest_factor_db":18.031195842748737,"mean":-0.00030972872994886806,"zero_energy":false},"output":{"rms":0.050622797738518355,"rms_dbfs":-25.913077132656746,"peak":0.4309110802244212,"peak_dbfs":-7.3122467708919086,"crest_factor_db":18.60083036176484,"mean":0.0003610539369156991,"zero_energy":false},"rms_difference_db":-0.37549815813633813},"full_scale":{"reference":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":76640,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":479,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":3.3220610214630148e-06,"threshold_dbfs":-54.785923944489625,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":79,"quiet_samples":12640,"reference_mean_power":2.026790753006935e-06,"output_mean_power":1.7329129478730043e-06,"output_rms_dbfs":-57.61223253332226,"output_zero_energy":false,"output_peak":0.005785028436314572,"output_energy_fraction":0.00011152597734231575,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":472,"reference_estimated_voiced_frames":73,"output_estimated_voiced_frames":61,"both_estimated_voiced_frames":43,"reference_voiced_output_unvoiced_frames":30,"reference_unvoiced_output_voiced_frames":18,"paired_voiced_fraction":0.589041095890411,"absolute_difference_cents_median":8.837640564143573,"absolute_difference_cents_p95":29.760293722750816,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":76640,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00018226521943209611,"log_rms_mae_db":3.1717088547694825,"rms_envelope_nrmse":0.3197020225474263,"envelope_correlation":0.899105719035058,"rms_gain_error_db":-0.3754959878499732,"reference_mean_power":0.0027970153309965396,"output_mean_power":0.0025653418613567195},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.029329446119484e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010840604762391139,"output_power":0.001008021451412293,"reference_active":true,"delta_db":-0.3158373646441426},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00017643244749495566,"output_power":0.00015910068693757504,"reference_active":true,"delta_db":-0.44906403983096244},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.741521666729805e-05,"output_power":3.972554475513301e-05,"reference_active":true,"delta_db":-0.7684787805348964}],"active_band_count":3,"active_band_mae_db":0.5111267283366671}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001676463927845064,"log_rms_mae_db":2.020232501872836,"rms_envelope_nrmse":0.20964454535154986,"envelope_correlation":0.9540035388743036,"rms_gain_error_db":-0.37548760836426887,"reference_mean_power":0.0028146416719843875,"output_mean_power":0.0025815132142714713},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.889944424347358e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0010453379933036454,"output_power":0.0009634601107653202,"reference_active":true,"delta_db":-0.3542299664413815},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.750212405766423e-05,"output_power":6.343220916034334e-05,"reference_active":true,"delta_db":-0.2700760190542065},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.9228813706071634e-06,"output_power":3.0603941588623733e-06,"reference_active":true,"delta_db":0.19966175146690712}],"active_band_count":3,"active_band_mae_db":0.2746559123208317}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00017586243391937327,"log_rms_mae_db":3.493121037598008,"rms_envelope_nrmse":0.38937795026253524,"envelope_correlation":0.8838986011385804,"rms_gain_error_db":-0.46185183567292726,"reference_mean_power":0.0014437040494073058,"output_mean_power":0.001298054691149043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":1.5100208608044743e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009242573938618173,"output_power":0.0008262073118850479,"reference_active":true,"delta_db":-0.4870389938453771},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001371496531392223,"output_power":0.00011935287375097292,"reference_active":true,"delta_db":-0.6036183327067114},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.2803310723594984e-05,"output_power":2.866208124055825e-05,"reference_active":true,"delta_db":-0.5860995510466724}],"active_band_count":3,"active_band_mae_db":0.558918959199587}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":0.0001539617967365963,"log_rms_mae_db":2.329965752427233,"rms_envelope_nrmse":0.2618214204121493,"envelope_correlation":0.9453662847013649,"rms_gain_error_db":-0.4618522291744693,"reference_mean_power":0.0014528118102568712,"output_mean_power":0.00130624348914544},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":1.4301118553576922e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0009152950896245062,"output_power":0.0008161813097098446,"reference_active":true,"delta_db":-0.49774487118283633},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.331274162853222e-05,"output_power":4.712724338877985e-05,"reference_active":true,"delta_db":-0.5355897952272588},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.0543911334413043e-06,"output_power":1.7554255300654672e-06,"reference_active":true,"delta_db":-0.6830072178193403}],"active_band_count":3,"active_band_mae_db":0.5721139614098117}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":8.126163402159473e-05,"log_rms_mae_db":3.357378713945374,"rms_envelope_nrmse":0.36202969974107113,"envelope_correlation":0.9117815948658754,"rms_gain_error_db":0.04989881953383977,"reference_mean_power":0.0001541240756859989,"output_mean_power":0.00015590511610353954},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1034128857485058e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011785014167056142,"output_power":0.00011961940431576836,"reference_active":true,"delta_db":0.06471526432262488},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.031054512020665e-06,"output_power":8.228752098965031e-06,"reference_active":true,"delta_db":0.10561405121311862},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.6793453202028483e-06,"output_power":2.993947891094175e-06,"reference_active":true,"delta_db":0.482155471975044}],"active_band_count":3,"active_band_mae_db":0.2174949291702625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":6.921465297386047e-05,"log_rms_mae_db":2.1675739276995643,"rms_envelope_nrmse":0.24204177368533317,"envelope_correlation":0.9585196547478868,"rms_gain_error_db":0.049900796073521346,"reference_mean_power":0.00015509566252518503,"output_mean_power":0.00015688800189262353},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.037312640211567e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00011306875179582826,"output_power":0.00011487948205986292,"reference_active":true,"delta_db":0.06899870991174402},{"low_hz":8.0,"high_hz":20.0,"reference_power":3.4410710139131097e-06,"output_power":3.676588863161786e-06,"reference_active":true,"delta_db":0.2875143201770093},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5029543761566635e-07,"output_power":1.8628051846984565e-07,"reference_active":true,"delta_db":0.9322164072580731}],"active_band_count":3,"active_band_mae_db":0.42957647911560876}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":957,"rms_floor":0.00013414824565394508,"log_rms_mae_db":2.70812975259694,"rms_envelope_nrmse":0.46599887929070083,"envelope_correlation":0.8747199520717955,"rms_gain_error_db":-0.18284834781675977,"reference_mean_power":0.0001730034207748909,"output_mean_power":0.00016587076807168043},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":957,"frequency_resolution_hz":0.2089864158829676,"power_floor":2.1522167661145565e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012025367220488917,"output_power":0.00012094763216780564,"reference_active":true,"delta_db":0.024990230872095894},{"low_hz":8.0,"high_hz":20.0,"reference_power":2.5558902161138395e-05,"output_power":2.2255625761737945e-05,"reference_active":true,"delta_db":-0.6010238565725261},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.5309687441279477e-05,"output_power":1.6045872041754555e-05,"reference_active":true,"delta_db":0.20397000244987734}],"active_band_count":3,"active_band_mae_db":0.27666136329816643}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":76640,"uncovered_tail_samples":0,"metrics":{"frame_count":951,"rms_floor":8.410768197622124e-05,"log_rms_mae_db":2.031068464383991,"rms_envelope_nrmse":0.24082413878508233,"envelope_correlation":0.9658650796986418,"rms_gain_error_db":-0.1828505045072709,"reference_mean_power":0.00017409446216015185,"output_mean_power":0.00016691674468567592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":951,"frequency_resolution_hz":0.2103049421661409,"power_floor":2.085538754133612e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00012977220791713165,"output_power":0.00013045399234886903,"reference_active":true,"delta_db":0.022756808787141435},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.3304937813307312e-05,"output_power":1.1707467553958513e-05,"reference_active":true,"delta_db":-0.555498861655693},{"low_hz":20.0,"high_hz":50.0,"reference_power":1.0464863211082628e-06,"output_power":1.1091986427427861e-06,"reference_active":true,"delta_db":0.25275773550089065}],"active_band_count":3,"active_band_mae_db":0.27700446864790834}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"hnm","status":"measured","mos":{"status":"measured","value":2.6089630126953125,"predicted_mos":2.6089630126953125,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b7349f3e55b7eb6b42ac098e91be6f4075072d98bc03b0e2cf99530c4de84c5e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.603478712000651},"started_at":"2026-10-13T04:40:38.538810+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":1.5253999663400464e-05,"native_input_samples":96000,"native_input_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","output_sha256":"a4d10f5065a5c321e1a6d56c5ebb74c26e1b54fe9fa69773ad265873677bfa19","output_samples":96000,"length_delta_samples":0,"analysis_s":0.8496574359996885,"synthesis_s":0.4917810050001208,"total_s":1.3414384409998092,"rtf":0.22357307349996822,"output_peak":0.6373299929456127,"output_rms":0.0635858439959604,"finished_at":"2026-10-13T04:40:39.886312+00:00","evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"a4d10f5065a5c321e1a6d56c5ebb74c26e1b54fe9fa69773ad265873677bfa19","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.181625679835015,"spectral_convergence":0.3434306454148203,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.748977814352,"spectral_convergence":0.3065352913337704,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.83299948129214,"spectral_convergence":0.3400716084111398,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.0004811559178580691,"delta_db":0.33148950831091434,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.0026461717974164923,"delta_db":-0.3175104495299103,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.00046600226420920516,"delta_db":-0.07288653252702787,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.0001333011276841209,"delta_db":-0.11324874684299109,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":0.00010823460289557788,"delta_db":-0.11785353325373879,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00020829384661514774,"delta_db":-0.26192218579356163,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20248515937635736,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20218937869785292,"pitch":null,"output_rms_dbfs":-23.93279119606777,"output_peak":0.6373299929456127,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.921200991826385},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.181625679835015,"spectral_convergence":0.3434306454148203,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.748977814352,"spectral_convergence":0.3065352913337704,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.83299948129214,"spectral_convergence":0.3400716084111398,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.0004811559178580691,"delta_db":0.33148950831091434,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.0026461717974164923,"delta_db":-0.3175104495299103,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.00046600226420920516,"delta_db":-0.07288653252702787,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.0001333011276841209,"delta_db":-0.11324874684299109,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":0.00010823460289557788,"delta_db":-0.11785353325373879,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00020829384661514774,"delta_db":-0.26192218579356163,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.20248515937635736,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.20218937869785292,"pitch":null,"output_rms_dbfs":-23.93279119606777,"output_peak":0.6373299929456127,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":5.921200991826385},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.7823003671340751,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06508335151161738,"rms_dbfs":-23.73060181736992,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.754232294373892,"mean":-0.00029132080078125,"zero_energy":false},"output":{"rms":0.0635858439959604,"rms_dbfs":-23.932791196067768,"peak":0.6373299929456127,"peak_dbfs":-3.9127128601446435,"crest_factor_db":20.020078335923124,"mean":-0.00015272954199921297,"zero_energy":false},"rms_difference_db":-0.2021893786978488},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":5.220017273228122e-06,"output_rms_dbfs":-52.82328059899031,"output_zero_energy":false,"output_peak":0.06339795986738882,"output_energy_fraction":0.0004066882392387273,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":42,"both_estimated_voiced_frames":35,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":7,"paired_voiced_fraction":0.9459459459459459,"absolute_difference_cents_median":1.4139917054097955,"absolute_difference_cents_p95":4.580517719245065,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028051035499693775,"log_rms_mae_db":1.3730181260682932,"rms_envelope_nrmse":0.09832368028717002,"envelope_correlation":0.9932202814366208,"rms_gain_error_db":-0.20218733977093184,"reference_mean_power":0.004239368434074771,"output_mean_power":0.004046526862831812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.0558356273882855e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018964431225846456,"output_power":0.0017833316510625643,"reference_active":true,"delta_db":-0.26707704417264844},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001936564415781366,"output_power":0.0001790710868102255,"reference_active":true,"delta_db":-0.34006478043687177},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.246230567335058e-05,"output_power":5.935813969760968e-05,"reference_active":true,"delta_db":-0.22137729756645133}],"active_band_count":3,"active_band_mae_db":0.27617304072532384}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026381677718961605,"log_rms_mae_db":0.7020748848882262,"rms_envelope_nrmse":0.05003537316253766,"envelope_correlation":0.9984869068359127,"rms_gain_error_db":-0.20218341922930633,"reference_mean_power":0.004260663575263385,"output_mean_power":0.0040668569960160906},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.853150694908015e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017547087963691286,"output_power":0.0016589166603901634,"reference_active":true,"delta_db":-0.24380484450266607},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.102196929549433e-05,"output_power":7.260442263285986e-05,"reference_active":true,"delta_db":-0.4763971864409734},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.037808910542995e-06,"output_power":2.526023563989882e-06,"reference_active":true,"delta_db":-0.8012305410023356}],"active_band_count":3,"active_band_mae_db":0.5071441906486583}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002736722962808348,"log_rms_mae_db":2.006922038779444,"rms_envelope_nrmse":0.12351245581201073,"envelope_correlation":0.9913330696685968,"rms_gain_error_db":-0.37540506928746437,"reference_mean_power":0.0024581880191069405,"output_mean_power":0.002254626414311165},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.077905340514435e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019022535341174436,"output_power":0.0017634391592368244,"reference_active":true,"delta_db":-0.32907919048757117},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012849688146439824,"output_power":0.00011250619979252459,"reference_active":true,"delta_db":-0.5771613231263113},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.095154000220669e-05,"output_power":3.9246954644374006e-05,"reference_active":true,"delta_db":-0.184642748156931}],"active_band_count":3,"active_band_mae_db":0.3636277539236045}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025339024948019116,"log_rms_mae_db":1.2625004780502864,"rms_envelope_nrmse":0.06990370681943822,"envelope_correlation":0.997966183073165,"rms_gain_error_db":-0.37540684418861614,"reference_mean_power":0.002470549817529662,"output_mean_power":0.0022659636108910697},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.981046082661891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017987268874748686,"output_power":0.001675775247850905,"reference_active":true,"delta_db":-0.3074945520905887},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5079379851283876e-05,"output_power":4.705215048452991e-05,"reference_active":true,"delta_db":-0.6840956447987967},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4721075346991194e-06,"output_power":1.6633890834420126e-06,"reference_active":true,"delta_db":-1.7207351133891935}],"active_band_count":3,"active_band_mae_db":0.9041084367595262}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":1.9192842884753585,"rms_envelope_nrmse":0.20380033554195987,"envelope_correlation":0.9712471701453432,"rms_gain_error_db":0.002690140658525679,"reference_mean_power":0.00019492290635367453,"output_mean_power":0.00019504368441918945},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7570632979474254e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.555963644612227e-05,"output_power":7.113136473904857e-05,"reference_active":true,"delta_db":-0.2622871842530821},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.293127801504828e-05,"output_power":1.1874755280907561e-05,"reference_active":true,"delta_db":-0.3701678059363867},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505165714035146e-06,"output_power":7.766537677518533e-06,"reference_active":true,"delta_db":0.7696908911169766}],"active_band_count":3,"active_band_mae_db":0.46738196043548186}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":1.209679401964608,"rms_envelope_nrmse":0.1139486480222374,"envelope_correlation":0.9905701776889434,"rms_gain_error_db":0.0026774587771605666,"reference_mean_power":0.00019590175284412936,"output_mean_power":0.00019602256501367334},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.622675888346401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.225076970752045e-05,"output_power":6.741225470368476e-05,"reference_active":true,"delta_db":-0.30103625102770804},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.071619770568576e-06,"output_power":5.845416248604605e-06,"reference_active":true,"delta_db":-0.1648912400878912},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6400879277670163e-07,"output_power":4.0334196635110166e-07,"reference_active":true,"delta_db":0.44561536881808583}],"active_band_count":3,"active_band_mae_db":0.303847619977895}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001240823686589812,"log_rms_mae_db":1.739542663168951,"rms_envelope_nrmse":0.17733657900236993,"envelope_correlation":0.9821456712708468,"rms_gain_error_db":-0.38315849049404277,"reference_mean_power":0.00018911407706020164,"output_mean_power":0.00017314422409131773},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8583468534278454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010082037892943071,"output_power":9.326012831680998e-05,"reference_active":true,"delta_db":-0.3385231671834559},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.94400784876693e-05,"output_power":1.6940715514631835e-05,"reference_active":true,"delta_db":-0.59766264608686},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0091360826952136e-06,"output_power":3.493747869954408e-06,"reference_active":true,"delta_db":-1.5647126983147888}],"active_band_count":3,"active_band_mae_db":0.8336328371950349}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":1.361563169056475,"rms_envelope_nrmse":0.10767989039856082,"envelope_correlation":0.9940724163696585,"rms_gain_error_db":-0.3831607838658874,"reference_mean_power":0.000190064830092334,"output_mean_power":0.00017401459831521812},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7984096251410155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010012152579842306,"output_power":9.170865144495582e-05,"reference_active":true,"delta_db":-0.381171521116329},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0693666398121261e-05,"output_power":9.404880845228183e-06,"reference_active":true,"delta_db":-0.557733339889145},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.760055744312888e-07,"output_power":2.1830503414977344e-07,"reference_active":true,"delta_db":-1.0185410278922264}],"active_band_count":3,"active_band_mae_db":0.6524819629659001}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"world_harvest","status":"measured","mos":{"status":"measured","value":3.0806827545166016,"predicted_mos":3.0806827545166016,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4074edf5806c0dfc2889e535fbbc8997e03c7ea1f3aae74d14fb5856b73722aa","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.365594216000318},"started_at":"2026-10-13T04:43:55.542221+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":2.315399979124777e-05,"native_input_samples":96000,"native_input_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","output_sha256":"23a263934fed10da9c0ec8cff112b81b23d64a990c204649e018089c6d1d0164","output_samples":96000,"length_delta_samples":0,"analysis_s":1.5453468100004102,"synthesis_s":0.08743614599916327,"total_s":1.6327829559995735,"rtf":0.27213049266659556,"output_peak":0.798552726076276,"output_rms":0.07753952795482191,"finished_at":"2026-10-13T04:43:57.180578+00:00","evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"23a263934fed10da9c0ec8cff112b81b23d64a990c204649e018089c6d1d0164","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.692053780852633,"spectral_convergence":0.6456037075547757,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.623119861286212,"spectral_convergence":0.3765955020940614,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.721128580598781,"spectral_convergence":0.3812850073201917,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.00044995728365064926,"delta_db":0.04034405489639417,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.004502438901487092,"delta_db":1.9907874669397543,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.0005647192171806282,"delta_db":0.7615588676929403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.00015287484511066965,"delta_db":0.48177331923677036,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":0.00011358231604615105,"delta_db":0.09159238678101309,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00022880583198141638,"delta_db":0.14598431042714785,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5853400676623366,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5210648729912948,"pitch":null,"output_rms_dbfs":-22.20953694437862,"output_peak":0.798552726076276,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.012100740912541},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.692053780852633,"spectral_convergence":0.6456037075547757,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":5.623119861286212,"spectral_convergence":0.3765955020940614,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.721128580598781,"spectral_convergence":0.3812850073201917,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.00044995728365064926,"delta_db":0.04034405489639417,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.004502438901487092,"delta_db":1.9907874669397543,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.0005647192171806282,"delta_db":0.7615588676929403,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.00015287484511066965,"delta_db":0.48177331923677036,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":0.00011358231604615105,"delta_db":0.09159238678101309,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00022880583198141638,"delta_db":0.14598431042714785,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5853400676623366,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":1.5210648729912948,"pitch":null,"output_rms_dbfs":-22.20953694437862,"output_peak":0.798552726076276,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.012100740912541},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8688847109847203,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06508335151161738,"rms_dbfs":-23.73060181736992,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.754232294373892,"mean":-0.00029132080078125,"zero_energy":false},"output":{"rms":0.07753952795482189,"rms_dbfs":-22.20953694437862,"peak":0.798552726076276,"peak_dbfs":-1.9539280679552362,"crest_factor_db":20.255608876423384,"mean":1.0062512082091753e-07,"zero_energy":false},"rms_difference_db":1.5210648729912997},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":2.601434804493776e-06,"output_rms_dbfs":-55.84787053643935,"output_zero_energy":false,"output_peak":0.05194299158411708,"output_energy_fraction":0.00013629414343494707,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":51,"both_estimated_voiced_frames":31,"reference_voiced_output_unvoiced_frames":6,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.8378378378378378,"absolute_difference_cents_median":6.0810358749213345,"absolute_difference_cents_p95":1813.5520892268155,"difference_over_600_cents_frames":11,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028051035499693775,"log_rms_mae_db":2.431871285625653,"rms_envelope_nrmse":0.26624763270196894,"envelope_correlation":0.9803825720413656,"rms_gain_error_db":1.5210720070864023,"reference_mean_power":0.004239368434074771,"output_mean_power":0.0060173928060228195},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.0558356273882855e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018964431225846456,"output_power":0.0030552034541027257,"reference_active":true,"delta_db":2.0710031432731806},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001936564415781366,"output_power":0.0002009291038165641,"reference_active":true,"delta_db":0.16010899744977625},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.246230567335058e-05,"output_power":6.043552848700889e-05,"reference_active":true,"delta_db":-0.14325687107940033}],"active_band_count":3,"active_band_mae_db":0.7914563372674523}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026381677718961605,"log_rms_mae_db":1.8149628330754608,"rms_envelope_nrmse":0.23405594711931918,"envelope_correlation":0.9897727538420811,"rms_gain_error_db":1.5210928916816244,"reference_mean_power":0.004260663575263385,"output_mean_power":0.006047648378649247},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.853150694908015e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017547087963691286,"output_power":0.002854818925931922,"reference_active":true,"delta_db":2.113735140258609},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.102196929549433e-05,"output_power":7.931480167197386e-05,"reference_active":true,"delta_db":-0.09248552197313623},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.037808910542995e-06,"output_power":2.67946533303978e-06,"reference_active":true,"delta_db":-0.5451230914306205}],"active_band_count":3,"active_band_mae_db":0.9171145845541219}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002736722962808348,"log_rms_mae_db":2.6191756891201385,"rms_envelope_nrmse":0.29881512954155404,"envelope_correlation":0.9846259242993184,"rms_gain_error_db":1.8724703623380536,"reference_mean_power":0.0024581880191069405,"output_mean_power":0.0037832246776317263},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.077905340514435e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019022535341174436,"output_power":0.0030523078624637395,"reference_active":true,"delta_db":2.0535993575485336},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012849688146439824,"output_power":0.0001192985009570663,"reference_active":true,"delta_db":-0.32257601165084354},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.095154000220669e-05,"output_power":3.615677391383738e-05,"reference_active":true,"delta_db":-0.5408056462409104}],"active_band_count":3,"active_band_mae_db":0.9723270051467625}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025339024948019116,"log_rms_mae_db":2.100655806338635,"rms_envelope_nrmse":0.27246859791139455,"envelope_correlation":0.9919425505533597,"rms_gain_error_db":1.8724704729029562,"reference_mean_power":0.002470549817529662,"output_mean_power":0.0038022499508984127},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.981046082661891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017987268874748686,"output_power":0.002879842295448051,"reference_active":true,"delta_db":2.0440347921713204},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5079379851283876e-05,"output_power":4.547259816340858e-05,"reference_active":true,"delta_db":-0.8323927265664632},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4721075346991194e-06,"output_power":2.00224897244235e-06,"reference_active":true,"delta_db":-0.9154927883556464}],"active_band_count":3,"active_band_mae_db":1.26397343569781}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":1.678441554574472,"rms_envelope_nrmse":0.18470808604303116,"envelope_correlation":0.9791697480613636,"rms_gain_error_db":0.42420331543731976,"reference_mean_power":0.00019492290635367453,"output_mean_power":0.0002149231562417816},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7570632979474254e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.555963644612227e-05,"output_power":8.398123117128979e-05,"reference_active":true,"delta_db":0.4589237748453217},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.293127801504828e-05,"output_power":1.413114429374786e-05,"reference_active":true,"delta_db":0.38535882125365306},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505165714035146e-06,"output_power":4.17238713491382e-06,"reference_active":true,"delta_db":-1.928737667177535}],"active_band_count":3,"active_band_mae_db":0.9243400877588366}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":1.0105306103104656,"rms_envelope_nrmse":0.10562912771268504,"envelope_correlation":0.9940205135900707,"rms_gain_error_db":0.42420563215476503,"reference_mean_power":0.00019590175284412936,"output_mean_power":0.00021600255343103122},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.622675888346401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.225076970752045e-05,"output_power":7.919718667737632e-05,"reference_active":true,"delta_db":0.39867276302359506},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.071619770568576e-06,"output_power":6.937233473194045e-06,"reference_active":true,"delta_db":0.57881744505107},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6400879277670163e-07,"output_power":2.499245390567189e-07,"reference_active":true,"delta_db":-1.6330297453236844}],"active_band_count":3,"active_band_mae_db":0.8701733177994498}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001240823686589812,"log_rms_mae_db":1.5939442546588303,"rms_envelope_nrmse":0.259614041348177,"envelope_correlation":0.9606797236753224,"rms_gain_error_db":-0.053893294510406003,"reference_mean_power":0.00018911407706020164,"output_mean_power":0.0001867817878610068},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8583468534278454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010082037892943071,"output_power":9.214536819451851e-05,"reference_active":true,"delta_db":-0.3907481565106858},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.94400784876693e-05,"output_power":2.524116132345672e-05,"reference_active":true,"delta_db":1.1341131851493707},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0091360826952136e-06,"output_power":8.851341252416423e-06,"reference_active":true,"delta_db":2.472462544687325}],"active_band_count":3,"active_band_mae_db":1.3324412954491274}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":1.0686448248232077,"rms_envelope_nrmse":0.13200318092770236,"envelope_correlation":0.9896682697725792,"rms_gain_error_db":-0.053891170749870744,"reference_mean_power":0.000190064830092334,"output_mean_power":0.00018772090732917636},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7984096251410155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010012152579842306,"output_power":9.45867014077246e-05,"reference_active":true,"delta_db":-0.24697379144201637},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0693666398121261e-05,"output_power":1.3804962672063564e-05,"reference_active":true,"delta_db":1.109086050329334},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.760055744312888e-07,"output_power":4.919395931633442e-07,"reference_active":true,"delta_db":2.509939241166903}],"active_band_count":3,"active_band_mae_db":1.2886663609794178}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"facodec","status":"measured","mos":{"status":"measured","value":3.9403457641601562,"predicted_mos":3.9403457641601562,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"db2116de6388f25f6c70ac6b7b041d3e5677ca9d0dd04ae10798b080282e8ad1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.024228687998402},"started_at":"2026-10-13T07:03:23.345349+00:00","native_sample_rate":16000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":3.802500032179523e-05,"native_input_samples":96000,"native_input_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","output_sha256":"4219fcc7ab3ece3c7354fdc43d5ff3b674d826ac077632b8213cd3ab594b3fee","output_samples":96000,"length_delta_samples":0,"analysis_s":2.3913950529986323,"synthesis_s":13.396351852001317,"total_s":15.787746904999949,"rtf":2.631291150833325,"output_peak":0.5582762360572815,"output_rms":0.060768449843807304,"finished_at":"2026-10-13T07:03:39.148547+00:00","evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"4219fcc7ab3ece3c7354fdc43d5ff3b674d826ac077632b8213cd3ab594b3fee","native_sample_rate":16000},"measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.583007532262961,"spectral_convergence":0.27257012809418113,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.806595409981408,"spectral_convergence":0.2716360145800304,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.784524871857812,"spectral_convergence":0.2860960002523088,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.000407079219806,"delta_db":-0.3945794733700254,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.0026103825433551188,"delta_db":-0.37664924928078825,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.0003809811092630302,"delta_db":-0.9477323813693375,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.00011231329398474096,"delta_db":-0.8572753344930145,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":6.787681287457356e-05,"delta_db":-2.144300393959157,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00011417151713585968,"delta_db":-2.8731088676360947,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2656076166847363,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.595835023189834,"pitch":null,"output_rms_dbfs":-24.32643684055975,"output_peak":0.5582762360572815,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.724709271367394},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":6.583007532262961,"spectral_convergence":0.27257012809418113,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.806595409981408,"spectral_convergence":0.2716360145800304,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.784524871857812,"spectral_convergence":0.2860960002523088,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.000407079219806,"delta_db":-0.3945794733700254,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.0026103825433551188,"delta_db":-0.37664924928078825,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.0003809811092630302,"delta_db":-0.9477323813693375,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.00011231329398474096,"delta_db":-0.8572753344930145,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":6.787681287457356e-05,"delta_db":-2.144300393959157,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00011417151713585968,"delta_db":-2.8731088676360947,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2656076166847363,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.595835023189834,"pitch":null,"output_rms_dbfs":-24.32643684055975,"output_peak":0.5582762360572815,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.724709271367394},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8485828586837805,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06508335151161738,"rms_dbfs":-23.73060181736992,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.754232294373892,"mean":-0.00029132080078125,"zero_energy":false},"output":{"rms":0.060768449843807304,"rms_dbfs":-24.32643684055975,"peak":0.5582762360572815,"peak_dbfs":-5.06301716424677,"crest_factor_db":19.26341967631298,"mean":-7.680030492792867e-06,"zero_energy":false},"rms_difference_db":-0.5958350231898315},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":2.8003147376935354e-06,"output_rms_dbfs":-55.527931539571455,"output_zero_energy":false,"output_peak":0.007768076378852129,"output_energy_fraction":0.00023886971087388423,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":41,"both_estimated_voiced_frames":35,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":6,"paired_voiced_fraction":0.9459459459459459,"absolute_difference_cents_median":3.3474710210952234,"absolute_difference_cents_p95":7.717159486250624,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028051035499693775,"log_rms_mae_db":1.5560578457479068,"rms_envelope_nrmse":0.131959335077626,"envelope_correlation":0.9899254041491957,"rms_gain_error_db":-0.5958309610059819,"reference_mean_power":0.004239368434074771,"output_mean_power":0.0036958817345725324},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.0558356273882855e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018964431225846456,"output_power":0.0019459100774096283,"reference_active":true,"delta_db":0.11182945146644732},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001936564415781366,"output_power":0.0001821530718867354,"reference_active":true,"delta_db":-0.26595447745128264},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.246230567335058e-05,"output_power":6.737954802998041e-05,"reference_active":true,"delta_db":0.3291008216676106}],"active_band_count":3,"active_band_mae_db":0.23562825019511355}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026381677718961605,"log_rms_mae_db":1.2242208778110568,"rms_envelope_nrmse":0.1193049112827081,"envelope_correlation":0.9921016725443028,"rms_gain_error_db":-0.5958270534469822,"reference_mean_power":0.004260663575263385,"output_mean_power":0.003714450182407493},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.853150694908015e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017547087963691286,"output_power":0.0018237539319059754,"reference_active":true,"delta_db":0.1676118800569293},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.102196929549433e-05,"output_power":8.05214450852334e-05,"reference_active":true,"delta_db":-0.026912344153798222},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.037808910542995e-06,"output_power":3.2720187134928522e-06,"reference_active":true,"delta_db":0.32255327194163574}],"active_band_count":3,"active_band_mae_db":0.1723591653841211}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002736722962808348,"log_rms_mae_db":1.9349428340191694,"rms_envelope_nrmse":0.12499084066005314,"envelope_correlation":0.9922339046950435,"rms_gain_error_db":-0.5490982474886207,"reference_mean_power":0.0024581880191069405,"output_mean_power":0.0021662335256568323},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.077905340514435e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019022535341174436,"output_power":0.0017369671907039393,"reference_active":true,"delta_db":-0.39476784425515454},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012849688146439824,"output_power":0.00013602160999505827,"reference_active":true,"delta_db":0.24715323237200829},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.095154000220669e-05,"output_power":4.9854227734921084e-05,"reference_active":true,"delta_db":0.8543175496646596}],"active_band_count":3,"active_band_mae_db":0.49874620876394077}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025339024948019116,"log_rms_mae_db":1.520805288246821,"rms_envelope_nrmse":0.10624263179052514,"envelope_correlation":0.9949504634689393,"rms_gain_error_db":-0.5490981204213443,"reference_mean_power":0.002470549817529662,"output_mean_power":0.002177127199597515},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.981046082661891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017987268874748686,"output_power":0.0016535932265416109,"reference_active":true,"delta_db":-0.36536541902482067},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5079379851283876e-05,"output_power":5.856085038259602e-05,"reference_active":true,"delta_db":0.2661833282607269},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4721075346991194e-06,"output_power":2.7356581148466355e-06,"reference_active":true,"delta_db":0.43994462786872124}],"active_band_count":3,"active_band_mae_db":0.35716445838475624}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":1.891166804221874,"rms_envelope_nrmse":0.1820722415612289,"envelope_correlation":0.9831424477539819,"rms_gain_error_db":-0.972103881716563,"reference_mean_power":0.00019492290635367453,"output_mean_power":0.00015583050928651788},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7570632979474254e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.555963644612227e-05,"output_power":6.179788974370808e-05,"reference_active":true,"delta_db":-0.8731621447359171},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.293127801504828e-05,"output_power":1.2116157751951302e-05,"reference_active":true,"delta_db":-0.2827652969419579},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505165714035146e-06,"output_power":8.104900153031561e-06,"reference_active":true,"delta_db":0.9548930464636334}],"active_band_count":3,"active_band_mae_db":0.7036068293805028}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":1.4916501965015638,"rms_envelope_nrmse":0.15179060607160277,"envelope_correlation":0.9903886084480935,"rms_gain_error_db":-0.9720898877648875,"reference_mean_power":0.00019590175284412936,"output_mean_power":0.00015661354970262738},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.622675888346401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.225076970752045e-05,"output_power":6.0660836155410426e-05,"reference_active":true,"delta_db":-0.7593408573580669},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.071619770568576e-06,"output_power":5.615365761917514e-06,"reference_active":true,"delta_db":-0.3392651677693925},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6400879277670163e-07,"output_power":3.907879640156131e-07,"reference_active":true,"delta_db":0.308293049681495}],"active_band_count":3,"active_band_mae_db":0.4689663582696515}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001240823686589812,"log_rms_mae_db":1.7458357210474158,"rms_envelope_nrmse":0.33448891509875117,"envelope_correlation":0.9728904687599966,"rms_gain_error_db":-2.792664920132992,"reference_mean_power":0.00018911407706020164,"output_mean_power":9.941624727970718e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8583468534278454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010082037892943071,"output_power":4.714859832172378e-05,"reference_active":true,"delta_db":-3.3007953923719127},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.94400784876693e-05,"output_power":9.307517596868477e-06,"reference_active":true,"delta_db":-3.198641480490725},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0091360826952136e-06,"output_power":3.512972631788056e-06,"reference_active":true,"delta_db":-1.540880638994494}],"active_band_count":3,"active_band_mae_db":2.6801058372857103}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":1.6311718578150247,"rms_envelope_nrmse":0.32598229233661935,"envelope_correlation":0.9777028325298693,"rms_gain_error_db":-2.7926661988956236,"reference_mean_power":0.000190064830092334,"output_mean_power":9.991602358563338e-05},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7984096251410155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010012152579842306,"output_power":4.759095193069394e-05,"reference_active":true,"delta_db":-3.2300306760769666},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0693666398121261e-05,"output_power":5.247165551850962e-06,"reference_active":true,"delta_db":-3.0920186491504698},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.760055744312888e-07,"output_power":2.2397183338472402e-07,"reference_active":true,"delta_db":-0.9072444846150871}],"active_band_count":3,"active_band_mae_db":2.409764603280841}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"hifiglot","status":"measured","mos":{"status":"measured","value":3.1422512531280518,"predicted_mos":3.1422512531280518,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":264704,"input_sample_rate":44100,"input_seconds":6.002358276643991,"input_float32_sha256":"d1d50ca92d97ba872be435b904e532314e597fd0fbbf50c72b56e76ad281123e","internal_samples":96038,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3884017290001793},"started_at":"2026-10-13T07:35:23.221006+00:00","native_sample_rate":44100,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.00717715399878216,"native_input_samples":264600,"native_input_sha256":"c192fb3af62bec5393c1478d79a185b03daf95001f1b0b64a3596d12b183bfa9","output_sha256":"829a9200468581a80cad90cca325c426fb21f968074559318dfec15f7d66b297","output_samples":264704,"length_delta_samples":104,"analysis_s":0.2417982020015188,"synthesis_s":36.56335514199964,"total_s":36.805153344001155,"rtf":6.134192224000192,"output_peak":0.5762966275215149,"output_rms":0.05989245650306664,"finished_at":"2026-10-13T07:36:00.050009+00:00","evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"829a9200468581a80cad90cca325c426fb21f968074559318dfec15f7d66b297","native_sample_rate":44100},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96038,"length_delta_samples":38,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.447458686695892,"output_peak":0.5737169203377953},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9996043232886982,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":9.509385434775679,"spectral_convergence":0.4213378449490149,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":9.079999725374655,"spectral_convergence":0.36564815750260854,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":8.45641450536532,"spectral_convergence":0.4242983628652371,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.0004309350679606018,"delta_db":-0.1472504396517047,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.0022914380758415264,"delta_db":-0.9426095611959646,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.00046506727713104077,"delta_db":-0.08160897155360922,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.0001446416784312902,"delta_db":0.24134754637925054,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":0.00010798008615578758,"delta_db":-0.12807811764098365,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.0001526783307338385,"delta_db":-1.6109125579085433,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.5253011990550093,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.7151396661039586,"pitch":null,"output_rms_dbfs":-24.445741483473874,"output_peak":0.5737169203377953,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":9.015266555171886},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6827444225839915,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06508335151161738,"rms_dbfs":-23.73060181736992,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.754232294373892,"mean":-0.00029132080078125,"zero_energy":false},"output":{"rms":0.05993947377358334,"rms_dbfs":-24.445741483473874,"peak":0.5737169203377953,"peak_dbfs":-4.826046830101783,"crest_factor_db":19.619694653372093,"mean":4.185014912249189e-06,"zero_energy":false},"rms_difference_db":-0.7151396661039549},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":1.7904794050520046e-06,"output_rms_dbfs":-57.470306700855936,"output_zero_energy":false,"output_peak":0.009244738547533158,"output_energy_fraction":0.000156983508839494,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":55,"both_estimated_voiced_frames":35,"reference_voiced_output_unvoiced_frames":2,"reference_unvoiced_output_voiced_frames":20,"paired_voiced_fraction":0.9459459459459459,"absolute_difference_cents_median":6.920778697036693,"absolute_difference_cents_p95":20.44192127747846,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028051035499693775,"log_rms_mae_db":2.465703391696344,"rms_envelope_nrmse":0.12097101082546227,"envelope_correlation":0.9937615722918105,"rms_gain_error_db":-0.7151378214102879,"reference_mean_power":0.004239368434074771,"output_mean_power":0.0035957325344707276},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.0558356273882855e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018964431225846456,"output_power":0.0017001692026338221,"reference_active":true,"delta_db":-0.474476770614446},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001936564415781366,"output_power":0.0001621859534002954,"reference_active":true,"delta_db":-0.7701870933268257},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.246230567335058e-05,"output_power":4.5873502419432236e-05,"reference_active":true,"delta_db":-1.3405611174925258}],"active_band_count":3,"active_band_mae_db":0.8617416604779325}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026381677718961605,"log_rms_mae_db":2.1902389087733836,"rms_envelope_nrmse":0.09817654023016595,"envelope_correlation":0.9975942405918488,"rms_gain_error_db":-0.7151243429217545,"reference_mean_power":0.004260663575263385,"output_mean_power":0.0036138057876183394},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.853150694908015e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017547087963691286,"output_power":0.0015784167845256745,"reference_active":true,"delta_db":-0.4598336275253341},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.102196929549433e-05,"output_power":7.159555678346789e-05,"reference_active":true,"delta_db":-0.5371672395507363},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.037808910542995e-06,"output_power":2.669698178618765e-06,"reference_active":true,"delta_db":-0.5609828645475651}],"active_band_count":3,"active_band_mae_db":0.5193279105412119}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002736722962808348,"log_rms_mae_db":2.670189309772061,"rms_envelope_nrmse":0.12731610927085427,"envelope_correlation":0.9920056378652671,"rms_gain_error_db":-0.5693490802566996,"reference_mean_power":0.0024581880191069405,"output_mean_power":0.002156156051324341},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.077905340514435e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019022535341174436,"output_power":0.0016852747331246058,"reference_active":true,"delta_db":-0.5259769005910657},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012849688146439824,"output_power":0.00012681207011971035,"reference_active":true,"delta_db":-0.05731995586193388},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.095154000220669e-05,"output_power":2.89122934748794e-05,"reference_active":true,"delta_db":-1.5118769469974624}],"active_band_count":3,"active_band_mae_db":0.6983912678168207}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025339024948019116,"log_rms_mae_db":2.2111606193525084,"rms_envelope_nrmse":0.09922247508865563,"envelope_correlation":0.9960995036533845,"rms_gain_error_db":-0.5693502999133474,"reference_mean_power":0.002470549817529662,"output_mean_power":0.002166998375134803},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.981046082661891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017987268874748686,"output_power":0.0015919466583697655,"reference_active":true,"delta_db":-0.5303671490011255},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5079379851283876e-05,"output_power":6.0817327145740115e-05,"reference_active":true,"delta_db":0.4303828765323886},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4721075346991194e-06,"output_power":2.2357595406287526e-06,"reference_active":true,"delta_db":-0.4364226560233023}],"active_band_count":3,"active_band_mae_db":0.46572422718560547}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":3.841972376349459,"rms_envelope_nrmse":0.26278507341617846,"envelope_correlation":0.9553307791297836,"rms_gain_error_db":0.0865384400909172,"reference_mean_power":0.00019492290635367453,"output_mean_power":0.00019884593732081112},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7570632979474254e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.555963644612227e-05,"output_power":9.014563791071774e-05,"reference_active":true,"delta_db":0.7665485668249437},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.293127801504828e-05,"output_power":1.1038915564423907e-05,"reference_active":true,"delta_db":-0.6871503741061269},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505165714035146e-06,"output_power":5.747470346348248e-06,"reference_active":true,"delta_db":-0.5378162511597346}],"active_band_count":3,"active_band_mae_db":0.6638383973636017}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":3.6690951471586124,"rms_envelope_nrmse":0.20886383006192147,"envelope_correlation":0.9712454336264013,"rms_gain_error_db":0.08656638338607506,"reference_mean_power":0.00019590175284412936,"output_mean_power":0.00019984576997859172},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.622675888346401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.225076970752045e-05,"output_power":8.524877089542067e-05,"reference_active":true,"delta_db":0.7184564800099358},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.071619770568576e-06,"output_power":4.915117575812478e-06,"reference_active":true,"delta_db":-0.917706552945784},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6400879277670163e-07,"output_power":3.608798044640511e-07,"reference_active":true,"delta_db":-0.03749295555743351}],"active_band_count":3,"active_band_mae_db":0.5578853295043844}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001240823686589812,"log_rms_mae_db":2.808371641995067,"rms_envelope_nrmse":0.2371208129894532,"envelope_correlation":0.9779245747768875,"rms_gain_error_db":-1.461507696825506,"reference_mean_power":0.00018911407706020164,"output_mean_power":0.00013507441269185736},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8583468534278454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010082037892943071,"output_power":6.53845148168217e-05,"reference_active":true,"delta_db":-1.880734199634553},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.94400784876693e-05,"output_power":1.4914706842398558e-05,"reference_active":true,"delta_db":-1.1508329256021332},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0091360826952136e-06,"output_power":2.459722446913757e-06,"reference_active":true,"delta_db":-3.088767258533257}],"active_band_count":3,"active_band_mae_db":2.040111461256648}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":2.8870440878176247,"rms_envelope_nrmse":0.1975628184484979,"envelope_correlation":0.9895482245782758,"rms_gain_error_db":-1.4615016462946495,"reference_mean_power":0.000190064830092334,"output_mean_power":0.00013575367558215054},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7984096251410155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010012152579842306,"output_power":6.309631160631671e-05,"reference_active":true,"delta_db":-2.005234867796603},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0693666398121261e-05,"output_power":8.38212417589078e-06,"reference_active":true,"delta_db":-1.0577254128841},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.760055744312888e-07,"output_power":1.6058524481774826e-07,"reference_active":true,"delta_db":-2.352122153663349}],"active_band_count":3,"active_band_mae_db":1.805027478114684}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"golf","status":"measured","mos":{"status":"measured","value":3.677403450012207,"predicted_mos":3.677403450012207,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":143760,"input_sample_rate":24000,"input_seconds":5.99,"input_float32_sha256":"b0dfae488957a7717bf3061011ca6f492a60c3099b4aeff5def4f8971a19d551","internal_samples":95840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.188032918000317},"started_at":"2026-10-13T05:03:14.400121+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0077981829999771435,"native_input_samples":144000,"native_input_sha256":"2cfc53645536e84fc32dde6a2840095cdebf0177caad5c1142acd4688838de26","output_sha256":"f4531de15ef9eaf8d3fbb273b70dee530a20d4e293cc60b32b42e452115e551e","output_samples":143760,"length_delta_samples":-240,"analysis_s":0.7094410310000967,"synthesis_s":0.12751207600013004,"total_s":0.8369531070002267,"rtf":0.13949218450003778,"output_peak":0.5175380706787109,"output_rms":0.05778430086060241,"finished_at":"2026-10-13T05:03:15.260449+00:00","analysis_stages":{"external_f0_s":0.09042246699937095},"evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"f4531de15ef9eaf8d3fbb273b70dee530a20d4e293cc60b32b42e452115e551e","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":95840,"length_delta_samples":-160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.764594894458654,"output_peak":0.5147525548461762},"common_prefix":{"samples":95840,"sample_rate":16000,"reference_coverage":0.9983333333333333,"output_coverage":1.0,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":95840,"output_samples":95840,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.772613663958781,"spectral_convergence":0.5883198583870989,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.536332677480329,"spectral_convergence":0.3112517865506357,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.484606250129385,"spectral_convergence":0.3405316188768335,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.0004465194024741137,"output_power":0.0003852899464693106,"delta_db":-0.6405265796842635,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.0028516300786191793,"output_power":0.0022455756413025437,"delta_db":-1.0376549826802492,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047468018133869943,"output_power":0.00038040189670531974,"delta_db":-0.9615842591594368,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013705114149495807,"output_power":9.358059576756791e-05,"delta_db":-1.6569685137756616,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011139748449483312,"output_power":6.998589071504133e-05,"delta_db":-2.0186488972619907,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022161183167656228,"output_power":0.00016358247031757507,"delta_db":-1.3185618096556453,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":1.2723241737028745,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-1.0412127425210653,"pitch":null,"output_rms_dbfs":-24.764594894458654,"output_peak":0.5147525548461762,"reference_rms_dbfs":-23.72338215193759,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.9311841971894985},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":95840,"output_samples":95840,"intelligibility":{"estoi":{"value":0.8563694093837791,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06513747093722895,"rms_dbfs":-23.72338215193759,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.747012628941565,"mean":-0.0002858284518794345,"zero_energy":false},"output":{"rms":0.057779031155578905,"rms_dbfs":-24.764594894458654,"peak":0.5147525548461762,"peak_dbfs":-5.768029783840468,"crest_factor_db":18.996565110618185,"mean":0.0005363638448010722,"zero_energy":false},"rms_difference_db":-1.0412127425210649},"full_scale":{"reference":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":95840,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":599,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":3.3823550194278833e-06,"output_rms_dbfs":-54.707808098410204,"output_zero_energy":false,"output_peak":0.011928126191964575,"output_energy_fraction":0.00031967867124393766,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":592,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":56,"both_estimated_voiced_frames":33,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":23,"paired_voiced_fraction":0.8918918918918919,"absolute_difference_cents_median":12.00559503301735,"absolute_difference_cents_p95":22.483829624882368,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":95840,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.00028051035499693775,"log_rms_mae_db":2.123910091887323,"rms_envelope_nrmse":0.1608085220479982,"envelope_correlation":0.9892927407380289,"rms_gain_error_db":-1.0412256095816939,"reference_mean_power":0.00424643160123981,"output_mean_power":0.003341193071732811},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":5.068086443171425e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.001898947178055285,"output_power":0.0014822184911965112,"reference_active":true,"delta_db":-1.0760065743908755},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00019351452386156493,"output_power":0.00013925504473231268,"reference_active":true,"delta_db":-1.429026285886227},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.262481907558987e-05,"output_power":3.677609409848694e-05,"reference_active":true,"delta_db":-2.3118088222428503}],"active_band_count":3,"active_band_mae_db":1.6056138941733178}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00026381677718961605,"log_rms_mae_db":1.5104437680079479,"rms_envelope_nrmse":0.130259572714278,"envelope_correlation":0.9964422890860222,"rms_gain_error_db":-1.0412577271334973,"reference_mean_power":0.004267808371366708,"output_mean_power":0.003357987987783594},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":4.86457863200391e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017568296337401336,"output_power":0.001368949507861223,"reference_active":true,"delta_db":-1.0834221842331684},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.07149942778965e-05,"output_power":5.803613500129752e-05,"reference_active":true,"delta_db":-1.4325573804197653},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.0388587596504733e-06,"output_power":1.7670361911290646e-06,"reference_active":true,"delta_db":-2.3546507087479616}],"active_band_count":3,"active_band_mae_db":1.623543424466965}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0002736722962808348,"log_rms_mae_db":2.2834103185685657,"rms_envelope_nrmse":0.18521828041554012,"envelope_correlation":0.9867630790078706,"rms_gain_error_db":-1.1543244496147382,"reference_mean_power":0.0024622947639266115,"output_mean_power":0.0018875896901573844},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":4.089177072166294e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019032249684491319,"output_power":0.0014536236158628672,"reference_active":true,"delta_db":-1.1703815659203034},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012890119346337483,"output_power":9.847094519052236e-05,"reference_active":true,"delta_db":-1.1694883180347282},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.084184245399792e-05,"output_power":2.1201894735338797e-05,"reference_active":true,"delta_db":-2.8473065175214147}],"active_band_count":3,"active_band_mae_db":1.7290588004921486}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":0.00025339024948019116,"log_rms_mae_db":1.6712754689882394,"rms_envelope_nrmse":0.14795651770311988,"envelope_correlation":0.9953968670367689,"rms_gain_error_db":-1.1543246719354696,"reference_mean_power":0.002474697610342589,"output_mean_power":0.001897097587522546},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":3.992022728509576e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017993885335837196,"output_power":0.0013715868402839235,"reference_active":true,"delta_db":-1.1790163903935922},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5061745616237636e-05,"output_power":4.3267957830120575e-05,"reference_active":true,"delta_db":-1.0468357790086462},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.473623395085469e-06,"output_power":1.2056817886728503e-06,"reference_active":true,"delta_db":-3.1210087861495746}],"active_band_count":3,"active_band_mae_db":1.7822869851839378}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":2.0944644162306814,"rms_envelope_nrmse":0.2595184205595733,"envelope_correlation":0.9733480017323642,"rms_gain_error_db":-1.8331817739674254,"reference_mean_power":0.00019524821099892506,"output_mean_power":0.00012801736560566542},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":1.7536363187807986e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.528722452275137e-05,"output_power":4.840570493898277e-05,"reference_active":true,"delta_db":-1.9182473789426164},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.2907305749596809e-05,"output_power":7.286057809556744e-06,"reference_active":true,"delta_db":-2.483429850598738},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.521066443018085e-06,"output_power":3.878666790993905e-06,"reference_active":true,"delta_db":-2.2563615347167545}],"active_band_count":3,"active_band_mae_db":2.2193462547527028}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":1.5683546914471866,"rms_envelope_nrmse":0.21366030840596273,"envelope_correlation":0.9915367986112751,"rms_gain_error_db":-1.8332514170381162,"reference_mean_power":0.00019622979041898744,"output_mean_power":0.00012865888943251833},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":1.6190209913041136e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.19516354261681e-05,"output_power":4.5825340445483114e-05,"reference_active":true,"delta_db":-1.9593496958459988},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.0719830509885795e-06,"output_power":3.4159205921039343e-06,"reference_active":true,"delta_db":-2.498227842765042},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.633380349557141e-07,"output_power":1.9813704808929576e-07,"reference_active":true,"delta_db":-2.633451746184861}],"active_band_count":3,"active_band_mae_db":2.3636764282653004}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1197,"rms_floor":0.0001240823686589812,"log_rms_mae_db":2.234014068086881,"rms_envelope_nrmse":0.21872557309898968,"envelope_correlation":0.9798800407184022,"rms_gain_error_db":-1.2265906847930181,"reference_mean_power":0.0001894299260380246,"output_mean_power":0.00014282016195415592},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1197,"frequency_resolution_hz":0.1670843776106934,"power_floor":2.8708152286442503e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010116026746216715,"output_power":7.727118355291381e-05,"reference_active":true,"delta_db":-1.169924045950929},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.9494836234029315e-05,"output_power":1.429331885705816e-05,"reference_active":true,"delta_db":-1.3478650914570145},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.031418572673627e-06,"output_power":4.721522889512622e-06,"reference_active":true,"delta_db":-0.2760834911289934}],"active_band_count":3,"active_band_mae_db":0.9312908761789789}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":95840,"uncovered_tail_samples":0,"metrics":{"frame_count":1191,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":2.088842997913848,"rms_envelope_nrmse":0.168134726633793,"envelope_correlation":0.992638059269812,"rms_gain_error_db":-1.2265866242741361,"reference_mean_power":0.00019038366089045735,"output_mean_power":0.0001435393618785792},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1191,"frequency_resolution_hz":0.16792611251049538,"power_floor":2.8107156981092694e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010045167255304249,"output_power":7.661703480441226e-05,"reference_active":true,"delta_db":-1.17631832172563},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0716608306994712e-05,"output_power":7.903664118799004e-06,"reference_active":true,"delta_db":-1.3222888169610454},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.763339934658089e-07,"output_power":2.979809277966727e-07,"reference_active":true,"delta_db":0.32754154728492496}],"active_band_count":3,"active_band_mae_db":0.9420495619905335}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"sifigan","status":"measured","mos":{"status":"measured","value":3.3964645862579346,"predicted_mos":3.3964645862579346,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144120,"input_sample_rate":24000,"input_seconds":6.005,"input_float32_sha256":"4d7e1958992e1f620e990de3bd30d2be73607744c2f540367c63470d8ac15dbf","internal_samples":96080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9859436940023443},"started_at":"2026-10-13T07:49:01.347059+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.002820703997713281,"native_input_samples":144000,"native_input_sha256":"2cfc53645536e84fc32dde6a2840095cdebf0177caad5c1142acd4688838de26","output_sha256":"61a29b65f0a0e2c02d8d66cd517bf1437d6d8d1802bbb6c858e61e413bea96df","output_samples":144120,"length_delta_samples":120,"analysis_s":2.355293251999683,"synthesis_s":3.458270987997821,"total_s":5.813564239997504,"rtf":0.9689273733329173,"output_peak":0.5137442350387573,"output_rms":0.0634612725000268,"finished_at":"2026-10-13T07:49:07.174035+00:00","analysis_stages":{"external_f0_s":1.5967887789993256},"evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"61a29b65f0a0e2c02d8d66cd517bf1437d6d8d1802bbb6c858e61e413bea96df","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96080,"length_delta_samples":80,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.949112523674586,"output_peak":0.5140608821340311},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9991673605328892,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":7.70725415715143,"spectral_convergence":0.5848303576774487,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":6.559644832246497,"spectral_convergence":0.31811639809072506,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.363339076249295,"spectral_convergence":0.3267211553150886,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.0006507599742926087,"delta_db":1.6428395249753234,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.00258916843150306,"delta_db":-0.41208778188740863,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.00042138669469009787,"delta_db":-0.5099586148091897,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.00012733756766715457,"delta_db":-0.31202148153098563,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":0.00010631187570930278,"delta_db":-0.19569700741747564,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00013638484129004644,"delta_db":-2.101025562033186,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.862271662108928,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.21489379457949648,"pitch":null,"output_rms_dbfs":-23.94549561194941,"output_peak":0.5140608821340311,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":6.876746021882407},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.8473537165657603,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06508335151161738,"rms_dbfs":-23.73060181736992,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.754232294373892,"mean":-0.00029132080078125,"zero_energy":false},"output":{"rms":0.0634929081484875,"rms_dbfs":-23.94549561194941,"peak":0.5140608821340311,"peak_dbfs":-5.779708857096239,"crest_factor_db":18.165786754853176,"mean":9.010329255416945e-05,"zero_energy":false},"rms_difference_db":-0.21489379457949198},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":6.829207560012061e-06,"output_rms_dbfs":-51.6562968757475,"output_zero_energy":false,"output_peak":0.10803588038856335,"output_energy_fraction":0.0005336179467170009,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":64,"both_estimated_voiced_frames":36,"reference_voiced_output_unvoiced_frames":1,"reference_unvoiced_output_voiced_frames":28,"paired_voiced_fraction":0.972972972972973,"absolute_difference_cents_median":6.141377522852744,"absolute_difference_cents_p95":18.014716514926544,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028051035499693775,"log_rms_mae_db":3.5688112225339905,"rms_envelope_nrmse":0.16403828331792566,"envelope_correlation":0.9805853722075145,"rms_gain_error_db":-0.21488710851428472,"reference_mean_power":0.004239368434074771,"output_mean_power":0.004034711172631545},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.0558356273882855e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018964431225846456,"output_power":0.0019028395903712006,"reference_active":true,"delta_db":0.014623567120230458},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001936564415781366,"output_power":0.00017565367791434,"reference_active":true,"delta_db":-0.4237469979060496},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.246230567335058e-05,"output_power":4.359845658699216e-05,"reference_active":true,"delta_db":-1.5614689607732344}],"active_band_count":3,"active_band_mae_db":0.6666131752665048}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026381677718961605,"log_rms_mae_db":2.9285835724600195,"rms_envelope_nrmse":0.10475961745070582,"envelope_correlation":0.9921162866319497,"rms_gain_error_db":-0.2148641492002596,"reference_mean_power":0.004260663575263385,"output_mean_power":0.00405499971910055},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.853150694908015e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017547087963691286,"output_power":0.0017655239744126948,"reference_active":true,"delta_db":0.02668566103005083},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.102196929549433e-05,"output_power":7.700240752180286e-05,"reference_active":true,"delta_db":-0.2209849097028224},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.037808910542995e-06,"output_power":2.0763291868155734e-06,"reference_active":true,"delta_db":-1.6526424280013763}],"active_band_count":3,"active_band_mae_db":0.6334376662447498}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002736722962808348,"log_rms_mae_db":2.5967834792154045,"rms_envelope_nrmse":0.16442287439221626,"envelope_correlation":0.9840722873103883,"rms_gain_error_db":-0.43214292339337407,"reference_mean_power":0.0024581880191069405,"output_mean_power":0.0022253627045058888},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.077905340514435e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019022535341174436,"output_power":0.0017352455268128946,"reference_active":true,"delta_db":-0.399074660972215},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012849688146439824,"output_power":0.0001164729656819724,"reference_active":true,"delta_db":-0.42667453956491597},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.095154000220669e-05,"output_power":3.229502246536883e-05,"reference_active":true,"delta_db":-1.0313464726748924}],"active_band_count":3,"active_band_mae_db":0.6190318910706745}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025339024948019116,"log_rms_mae_db":1.9655617475032126,"rms_envelope_nrmse":0.1054439750418768,"envelope_correlation":0.9942332740779274,"rms_gain_error_db":-0.432142096592365,"reference_mean_power":0.002470549817529662,"output_mean_power":0.0022365540908563302},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.981046082661891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017987268874748686,"output_power":0.0016386640255072597,"reference_active":true,"delta_db":-0.40475307055843235},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5079379851283876e-05,"output_power":5.139629170154268e-05,"reference_active":true,"delta_db":-0.3005725650667944},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4721075346991194e-06,"output_power":1.8513943275111524e-06,"reference_active":true,"delta_db":-1.2556842953000513}],"active_band_count":3,"active_band_mae_db":0.6536699769750927}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":2.330871035510926,"rms_envelope_nrmse":0.24329423706913156,"envelope_correlation":0.95882132098845,"rms_gain_error_db":-0.13697740578865425,"reference_mean_power":0.00019492290635367453,"output_mean_power":0.00018887093819266367},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7570632979474254e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.555963644612227e-05,"output_power":7.591097066149187e-05,"reference_active":true,"delta_db":0.020146850611590057},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.293127801504828e-05,"output_power":1.2658176244796406e-05,"reference_active":true,"delta_db":-0.09270310656294085},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505165714035146e-06,"output_power":5.061156158732371e-06,"reference_active":true,"delta_db":-1.0900862695888744}],"active_band_count":3,"active_band_mae_db":0.40097874225446845}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":1.4897360213543547,"rms_envelope_nrmse":0.12518459825786538,"envelope_correlation":0.9887174426492449,"rms_gain_error_db":-0.13696075752012563,"reference_mean_power":0.00019590175284412936,"output_mean_power":0.0001898201211045222},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.622675888346401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.225076970752045e-05,"output_power":7.223628627016604e-05,"reference_active":true,"delta_db":-0.0008706768783567821},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.071619770568576e-06,"output_power":6.345662535672478e-06,"reference_active":true,"delta_db":0.19172406002657788},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6400879277670163e-07,"output_power":3.1544837255524143e-07,"reference_active":true,"delta_db":-0.6218358317276019}],"active_band_count":3,"active_band_mae_db":0.27147685621084555}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001240823686589812,"log_rms_mae_db":2.0435557412307914,"rms_envelope_nrmse":0.2488946107127547,"envelope_correlation":0.9737570815130713,"rms_gain_error_db":-1.4389486538670402,"reference_mean_power":0.00018911407706020164,"output_mean_power":0.00013577787023067288},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8583468534278454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010082037892943071,"output_power":6.645819336921961e-05,"reference_active":true,"delta_db":-1.809997943461983},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.94400784876693e-05,"output_power":1.5580579157689817e-05,"reference_active":true,"delta_db":-0.9611441689019168},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0091360826952136e-06,"output_power":1.8219558647351984e-06,"reference_active":true,"delta_db":-4.3922497791842545}],"active_band_count":3,"active_band_mae_db":2.387797297182718}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":1.4231600555315893,"rms_envelope_nrmse":0.19443354393896056,"envelope_correlation":0.9901792061210529,"rms_gain_error_db":-1.438946980405996,"reference_mean_power":0.000190064830092334,"output_mean_power":0.00013646053313894752},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7984096251410155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010012152579842306,"output_power":6.529971393803277e-05,"reference_active":true,"delta_db":-1.8561618065983352},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0693666398121261e-05,"output_power":8.53747103759999e-06,"reference_active":true,"delta_db":-0.9779738823570625},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.760055744312888e-07,"output_power":1.3403622743008183e-07,"reference_active":true,"delta_db":-3.1369565778830073}],"active_band_count":3,"active_band_mae_db":1.9903640889461347}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"llsm2","status":"measured","mos":{"status":"measured","value":3.0518527030944824,"predicted_mos":3.0518527030944824,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":144240,"input_sample_rate":24000,"input_seconds":6.01,"input_float32_sha256":"9690e9be67122ec78a58c3f611422cf8aa7dfe0195c3a0b3b58df28b9a8f2b18","internal_samples":96160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4903019329995004},"started_at":"2026-10-13T05:11:43.945664+00:00","native_sample_rate":24000,"seed":0,"metrics_include_timing":false,"input_quantization":"none","resampling_and_input_io_s":0.0027765160011767875,"native_input_samples":144000,"native_input_sha256":"2cfc53645536e84fc32dde6a2840095cdebf0177caad5c1142acd4688838de26","output_sha256":"3aff74649a4caa83874a7add2b9b554316d42eb7bfb308b2dd8a056524e99f41","output_samples":144240,"length_delta_samples":240,"analysis_s":0.8032308780002495,"synthesis_s":0.09860889599985967,"total_s":0.9018397740001092,"rtf":0.1503066290000182,"output_peak":0.5286650657653809,"output_rms":0.06325663002309642,"finished_at":"2026-10-13T05:11:44.858696+00:00","analysis_stages":{"external_f0_s":0.15014554100162059},"evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"3aff74649a4caa83874a7add2b9b554316d42eb7bfb308b2dd8a056524e99f41","native_sample_rate":24000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96160,"length_delta_samples":160,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-23.97764294651472,"output_peak":0.5292276775841801},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9983361064891847,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":5.192709040586625,"spectral_convergence":0.20728320443069193,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":4.692649082156448,"spectral_convergence":0.19264266105624722,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":5.028151213441517,"spectral_convergence":0.2032704551848179,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.00043837949001154986,"delta_db":-0.07286653745297329,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.002724359758197317,"delta_db":-0.19104624772465323,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.0004319893698448814,"delta_db":-0.40203620027079423,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.00011899162425415652,"delta_db":-0.6064230535153535,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":9.864596176709117e-05,"delta_db":-0.5207217001587241,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00019592076962318686,"delta_db":-0.5278818072407304,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.3868292577272048,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.23981010934762545,"pitch":null,"output_rms_dbfs":-23.970411926717542,"output_peak":0.5292276775841801,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":4.971169778728197},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.9125768164249807,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06508335151161738,"rms_dbfs":-23.73060181736992,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.754232294373892,"mean":-0.00029132080078125,"zero_energy":false},"output":{"rms":0.06331103358576753,"rms_dbfs":-23.970411926717542,"peak":0.5292276775841801,"peak_dbfs":-5.5271490225796525,"crest_factor_db":18.44326290413789,"mean":6.244935213386421e-05,"zero_energy":false},"rms_difference_db":-0.2398101093476228},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":4.294347558765968e-06,"output_rms_dbfs":-53.67102809283602,"output_zero_energy":false,"output_peak":0.06216665429368174,"output_energy_fraction":0.00033748069683823417,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":36,"both_estimated_voiced_frames":33,"reference_voiced_output_unvoiced_frames":4,"reference_unvoiced_output_voiced_frames":3,"paired_voiced_fraction":0.8918918918918919,"absolute_difference_cents_median":1.2488797732095551,"absolute_difference_cents_p95":3.837966590195541,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028051035499693775,"log_rms_mae_db":1.1830073913167636,"rms_envelope_nrmse":0.08090005598534646,"envelope_correlation":0.9956917311353038,"rms_gain_error_db":-0.2398136991548422,"reference_mean_power":0.004239368434074771,"output_mean_power":0.004011620037240064},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.0558356273882855e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018964431225846456,"output_power":0.0018898031113814041,"reference_active":true,"delta_db":-0.015232623572137047},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001936564415781366,"output_power":0.00016483169121352752,"reference_active":true,"delta_db":-0.6999123284540318},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.246230567335058e-05,"output_power":3.434218165281912e-05,"reference_active":true,"delta_db":-2.597901302285436}],"active_band_count":3,"active_band_mae_db":1.1043487514372017}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026381677718961605,"log_rms_mae_db":0.6589349164117976,"rms_envelope_nrmse":0.048370061991598115,"envelope_correlation":0.9987930939060067,"rms_gain_error_db":-0.239809867424293,"reference_mean_power":0.004260663575263385,"output_mean_power":0.0040317747127875505},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.853150694908015e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017547087963691286,"output_power":0.0017512864790167647,"reference_active":true,"delta_db":-0.008478585376042631},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.102196929549433e-05,"output_power":7.116894495094545e-05,"reference_active":true,"delta_db":-0.5631226716084867},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.037808910542995e-06,"output_power":1.7845815856479512e-06,"reference_active":true,"delta_db":-2.3102404428119407}],"active_band_count":3,"active_band_mae_db":0.9606138999321567}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002736722962808348,"log_rms_mae_db":1.4904746955584285,"rms_envelope_nrmse":0.10059689173596617,"envelope_correlation":0.9940323618557633,"rms_gain_error_db":-0.22556657149891302,"reference_mean_power":0.0024581880191069405,"output_mean_power":0.0023337720833631983},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.077905340514435e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019022535341174436,"output_power":0.001887741159057393,"reference_active":true,"delta_db":-0.03325954636876638},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012849688146439824,"output_power":0.00010475271474128946,"reference_active":true,"delta_db":-0.8872730091574367},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.095154000220669e-05,"output_power":2.0855806402975627e-05,"reference_active":true,"delta_db":-2.9304325147684156}],"active_band_count":3,"active_band_mae_db":1.2836550234315396}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025339024948019116,"log_rms_mae_db":0.849587706752702,"rms_envelope_nrmse":0.06000408952583971,"envelope_correlation":0.9981220865265179,"rms_gain_error_db":-0.2255668373448358,"reference_mean_power":0.002470549817529662,"output_mean_power":0.0023455080721880676},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.981046082661891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017987268874748686,"output_power":0.0017818192076703748,"reference_active":true,"delta_db":-0.04101590329716701},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5079379851283876e-05,"output_power":4.771698408491057e-05,"reference_active":true,"delta_db":-0.6231605519342018},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4721075346991194e-06,"output_power":1.3026118791122988e-06,"reference_active":true,"delta_db":-2.7825232390429595}],"active_band_count":3,"active_band_mae_db":1.1488998980914429}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":1.2788323442501546,"rms_envelope_nrmse":0.1640700768500239,"envelope_correlation":0.9834709015369328,"rms_gain_error_db":-0.6103820037162291,"reference_mean_power":0.00019492290635367453,"output_mean_power":0.00016936539441421818},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7570632979474254e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.555963644612227e-05,"output_power":6.161350203599803e-05,"reference_active":true,"delta_db":-0.8861396539303462},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.293127801504828e-05,"output_power":1.1539387256303103e-05,"reference_active":true,"delta_db":-0.49458700579916526},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505165714035146e-06,"output_power":4.1525431942929136e-06,"reference_active":true,"delta_db":-1.9494420572266749}],"active_band_count":3,"active_band_mae_db":1.1100562389853954}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":0.8555362550981818,"rms_envelope_nrmse":0.10459507931754244,"envelope_correlation":0.9950805304616945,"rms_gain_error_db":-0.6103820992403654,"reference_mean_power":0.00019590175284412936,"output_mean_power":0.00017021589472353694},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.622675888346401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.225076970752045e-05,"output_power":5.908951212702911e-05,"reference_active":true,"delta_db":-0.8733207387395331},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.071619770568576e-06,"output_power":4.961734141144258e-06,"reference_active":true,"delta_db":-0.8767107619762318},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6400879277670163e-07,"output_power":1.8940093079635756e-07,"reference_active":true,"delta_db":-2.837297653517038}],"active_band_count":3,"active_band_mae_db":1.529109718077601}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001240823686589812,"log_rms_mae_db":0.9652673326079259,"rms_envelope_nrmse":0.09171690771369109,"envelope_correlation":0.9956182705473694,"rms_gain_error_db":-0.29166668798729817,"reference_mean_power":0.00018911407706020164,"output_mean_power":0.00017683050563745492},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8583468534278454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010082037892943071,"output_power":9.752490676316329e-05,"reference_active":true,"delta_db":-0.1443278159890809},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.94400784876693e-05,"output_power":1.8692504395365126e-05,"reference_active":true,"delta_db":-0.1703052258200095},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0091360826952136e-06,"output_power":3.3623696846088273e-06,"reference_active":true,"delta_db":-1.7311736890259521}],"active_band_count":3,"active_band_mae_db":0.6819355769450142}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":0.6910318901251598,"rms_envelope_nrmse":0.04943354768615002,"envelope_correlation":0.9991706521561993,"rms_gain_error_db":-0.2916661971244445,"reference_mean_power":0.000190064830092334,"output_mean_power":0.00017771952426959696},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7984096251410155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010012152579842306,"output_power":9.664863893056298e-05,"reference_active":true,"delta_db":-0.15331717013606042},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0693666398121261e-05,"output_power":1.017993949129014e-05,"reference_active":true,"delta_db":-0.21381435034838026},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.760055744312888e-07,"output_power":2.0072200590804453e-07,"reference_active":true,"delta_db":-1.3832286508127805}],"active_band_count":3,"active_band_mae_db":0.583453390432407}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},{"case":"librispeech_908-31957-0016","method":"magphase","status":"measured","mos":{"status":"measured","value":3.412137746810913,"predicted_mos":3.412137746810913,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":289364,"input_sample_rate":48000,"input_seconds":6.028416666666667,"input_float32_sha256":"540d8816402a2a588a3a150c7e227bfb0485c2cd1f06d332d8d6e9d142fa514e","internal_samples":96455,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4643947669992485},"started_at":"2026-10-13T05:19:27.977578+00:00","native_sample_rate":48000,"seed":0,"metrics_include_timing":false,"input_quantization":"PCM16 (MagPhase/REAPER file API only)","resampling_and_input_io_s":0.00881588899937924,"native_input_samples":288000,"native_input_sha256":"dbbc298a955e1e20913374b96f9222d8424bbe0f6bf1261075c95f93aa2bd9f8","output_sha256":"00c5a9eae857371e1c70f5ec05aa042483b2e212b178a0288cfd4e3b51aaf75a","output_samples":289364,"length_delta_samples":1364,"analysis_s":3.500715668000339,"synthesis_s":1.0806744740002614,"total_s":4.5813901420006005,"rtf":0.7635650236667667,"output_peak":0.568192476729069,"output_rms":0.06195803016178597,"finished_at":"2026-10-13T05:19:32.582588+00:00","evaluation_input":{"reference":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"output_sha256":"00c5a9eae857371e1c70f5ec05aa042483b2e212b178a0288cfd4e3b51aaf75a","native_sample_rate":48000},"measurement":{"status":"length_mismatch","reference_samples":96000,"output_samples":96455,"length_delta_samples":455,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":null,"band_energy":null,"rms_gain_error_db":null,"pitch":null,"output_rms_dbfs":-24.158515327974843,"output_peak":0.5677393871586619},"common_prefix":{"samples":96000,"sample_rate":16000,"reference_coverage":1.0,"output_coverage":0.9952827743507335,"alignment":"none","gain_normalization":"none","measurement":{"status":"ok","reference_samples":96000,"output_samples":96000,"length_delta_samples":0,"sample_rate":16000,"reference_finite":true,"output_finite":true,"alignment":"none","gain_normalization":"none","mr_stft":[{"window_ms":5,"window_samples":80,"hop_samples":20,"log_magnitude_mae_db":8.26819553345101,"spectral_convergence":0.6238430411282252,"spectral_convergence_status":"ok","amplitude_floor":2.313542282753755e-06},{"window_ms":20,"window_samples":320,"hop_samples":80,"log_magnitude_mae_db":7.554462555977465,"spectral_convergence":0.4595191292494894,"spectral_convergence_status":"ok","amplitude_floor":1.7511633455315343e-06},{"window_ms":80,"window_samples":1280,"hop_samples":320,"log_magnitude_mae_db":6.654419100604982,"spectral_convergence":0.44980989879091793,"spectral_convergence_status":"ok","amplitude_floor":1.6157176909573571e-06}],"band_energy":{"bands":[{"low_hz":0.0,"high_hz":200.0,"reference_power":0.00044579673236332844,"output_power":0.0004096296125335473,"delta_db":-0.36745536145122276,"reference_active":true},{"low_hz":200.0,"high_hz":500.0,"reference_power":0.00284687947566611,"output_power":0.002555800291267023,"delta_db":-0.4684216612471986,"reference_active":true},{"low_hz":500.0,"high_hz":1000.0,"reference_power":0.00047388905696308834,"output_power":0.0004406299468399048,"delta_db":-0.3160267001933754,"reference_active":true},{"low_hz":1000.0,"high_hz":2000.0,"reference_power":0.00013682287067652582,"output_power":0.00013616604254196193,"delta_db":-0.020898825633749504,"reference_active":true},{"low_hz":2000.0,"high_hz":4000.0,"reference_power":0.00011121195656671286,"output_power":0.00011790317828387498,"delta_db":0.2537403099408548,"reference_active":true},{"low_hz":4000.0,"high_hz":8000.0,"reference_power":0.00022124255174898303,"output_power":0.00019641741029764514,"delta_db":-0.516886781487494,"reference_active":true}],"power_floor":1e-12,"active_band_count":6,"band_count":6,"active_band_mae_db":0.32390493999231584,"active_policy":"reference power > max(1e-12, total reference power * 1e-6)","status":"ok"},"rms_gain_error_db":-0.40741248115628803,"pitch":null,"output_rms_dbfs":-24.138014298526205,"output_peak":0.5677393871586619,"reference_rms_dbfs":-23.730601817369916,"rms_gain_status":"ok","mr_stft_log_magnitude_mae_db":7.4923590633444865},"extended":{"schema":"extended-diagnostics-v1","status":"measured","alignment":"none","gain_normalization":"none","sample_rate":16000,"reference_samples":96000,"output_samples":96000,"intelligibility":{"estoi":{"value":0.6184507532294365,"library":"pystoi","version":"0.4.1","extended":true,"reference_required":true,"internal_sample_rate":10000,"evaluation_seed":0,"internal_silence_removal_db":40,"warnings":[],"normalization":"公式実装の行・列正規化","scope":"明瞭性の推定。MOS・自然性・話者類似度ではない","status":"measured"}},"level":{"reference":{"rms":0.06508335151161738,"rms_dbfs":-23.73060181736992,"peak":0.563873291015625,"peak_dbfs":-4.976369522996023,"crest_factor_db":18.754232294373892,"mean":-0.00029132080078125,"zero_energy":false},"output":{"rms":0.06210109887726591,"rms_dbfs":-24.138014298526205,"peak":0.5677393871586619,"peak_dbfs":-4.9170195073432135,"crest_factor_db":19.22099479118299,"mean":0.00036280640005942333,"zero_energy":false},"rms_difference_db":-0.40741248115628537},"full_scale":{"reference":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"},"output":{"sample_count":96000,"full_scale_amplitude":1.0,"at_or_above_full_scale_samples":0,"at_or_above_full_scale_fraction":0.0,"near_full_scale_amplitude":0.999,"near_full_scale_fraction":0.0,"longest_full_scale_run_samples":0,"meaning":"固定振幅尺度の閾値超過。録音時のクリッピング発生の証明ではない"}},"quiet_regions":{"block_ms":10.0,"block_samples":160,"total_blocks":600,"uncovered_tail_samples":0,"meaning":"参照低エネルギー区間。人が付けた無音・無声ラベルではない","status":"measured","threshold_power":7.456894731149078e-06,"threshold_dbfs":-51.27441987793588,"threshold_policy":"max(-60 dBFS, 最大参照ブロック電力より40 dB下)","quiet_blocks":189,"quiet_samples":30240,"reference_mean_power":3.2891175357339085e-06,"output_mean_power":5.4439817089948165e-06,"output_rms_dbfs":-52.64083342634016,"output_zero_energy":false,"output_peak":0.06222522126899646,"output_energy_fraction":0.0004446605911382675,"energy_fraction_denominator":"出力波形全体のエネルギー"},"pitch_consistency":{"status":"estimated_consistency_not_ground_truth","estimator":"independent-yin-difference-v2-prefiltered","fmin_hz":65.0,"fmax_hz":700.0,"frame_ms":80.0,"hop_ms":10.0,"prefilter":{"kind":"butterworth-forward-backward","order_per_pass":6,"cutoff_hz":2000.0,"pad_type":"even","max_pad_samples":1280},"total_frames":593,"reference_estimated_voiced_frames":37,"output_estimated_voiced_frames":38,"both_estimated_voiced_frames":27,"reference_voiced_output_unvoiced_frames":10,"reference_unvoiced_output_voiced_frames":11,"paired_voiced_fraction":0.7297297297297297,"absolute_difference_cents_median":6.514315048097743,"absolute_difference_cents_p95":20.209280029072865,"difference_over_600_cents_frames":0,"uncovered_tail_samples":0,"scope":"自然音声の参照F0は未知。同じ独立測定器の入出力推定値の整合性のみ"},"temporal":{"schema":"temporal-diagnostics-v2","status":"measured","sample_rate":16000,"samples":96000,"alignment":"none","normalization":"none","bands":[{"name":"全帯域","status":"measured","filter":{"kind":"none"},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.00028051035499693775,"log_rms_mae_db":3.796190850535257,"rms_envelope_nrmse":0.282554061125062,"envelope_correlation":0.9412708710946033,"rms_gain_error_db":-0.4074059146616804,"reference_mean_power":0.004239368434074771,"output_mean_power":0.0038597623928734587},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":5.0558356273882855e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0018964431225846456,"output_power":0.0017351764569700015,"reference_active":true,"delta_db":-0.38596175536304395},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.0001936564415781366,"output_power":0.00017974760076719309,"reference_active":true,"delta_db":-0.32368845240572197},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.246230567335058e-05,"output_power":5.3209578485320036e-05,"reference_active":true,"delta_db":-0.6962819276584975}],"active_band_count":3,"active_band_mae_db":0.46864404514242114}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00026381677718961605,"log_rms_mae_db":2.6126807290288636,"rms_envelope_nrmse":0.1917535966398391,"envelope_correlation":0.9728935513133152,"rms_gain_error_db":-0.40738904616094196,"reference_mean_power":0.004260663575263385,"output_mean_power":0.003879165768960659},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":4.853150694908015e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017547087963691286,"output_power":0.0016098825547893548,"reference_active":true,"delta_db":-0.3741085895925229},{"low_hz":8.0,"high_hz":20.0,"reference_power":8.102196929549433e-05,"output_power":7.545936996876722e-05,"reference_active":true,"delta_db":-0.3088961997135976},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.037808910542995e-06,"output_power":3.1701257581649486e-06,"reference_active":true,"delta_db":0.18516039271494786}],"active_band_count":3,"active_band_mae_db":0.2893883940070228}}]},{"name":"300–1000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[300.0,1000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0002736722962808348,"log_rms_mae_db":3.9828523942033423,"rms_envelope_nrmse":0.31025591072522885,"envelope_correlation":0.9393466078922904,"rms_gain_error_db":-0.46348062972648046,"reference_mean_power":0.0024581880191069405,"output_mean_power":0.002209362786489751},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":4.077905340514435e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0019022535341174436,"output_power":0.0017301161480419856,"reference_active":true,"delta_db":-0.41193139993173766},{"low_hz":8.0,"high_hz":20.0,"reference_power":0.00012849688146439824,"output_power":0.00011121996505020127,"reference_active":true,"delta_db":-0.6270983348300284},{"low_hz":20.0,"high_hz":50.0,"reference_power":4.095154000220669e-05,"output_power":4.0546888628445596e-05,"reference_active":true,"delta_db":-0.043127040809907764}],"active_band_count":3,"active_band_mae_db":0.360718925190558}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":0.00025339024948019116,"log_rms_mae_db":2.870048726365636,"rms_envelope_nrmse":0.2170639894613938,"envelope_correlation":0.9705639437517278,"rms_gain_error_db":-0.4634816326418041,"reference_mean_power":0.002470549817529662,"output_mean_power":0.0022204727734816444},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":3.981046082661891e-09,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.0017987268874748686,"output_power":0.0016394813951489163,"reference_active":true,"delta_db":-0.4025873391045056},{"low_hz":8.0,"high_hz":20.0,"reference_power":5.5079379851283876e-05,"output_power":4.8028946994782425e-05,"reference_active":true,"delta_db":-0.594859767090858},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.4721075346991194e-06,"output_power":2.6459178594146595e-06,"reference_active":true,"delta_db":0.29508999416071924}],"active_band_count":3,"active_band_mae_db":0.4308457001186943}}]},{"name":"1000–3000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[1000.0,3000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":9.470771529403075e-05,"log_rms_mae_db":4.108006567264492,"rms_envelope_nrmse":0.4241110629836373,"envelope_correlation":0.8762546173072813,"rms_gain_error_db":-0.01294899158271143,"reference_mean_power":0.00019492290635367453,"output_mean_power":0.00019434258676988012},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":1.7570632979474254e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.555963644612227e-05,"output_power":7.675425042206601e-05,"reference_active":true,"delta_db":0.06812575097188978},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.293127801504828e-05,"output_power":1.2616908259328544e-05,"reference_active":true,"delta_db":-0.10688503682826356},{"low_hz":20.0,"high_hz":50.0,"reference_power":6.505165714035146e-06,"output_power":6.621033019251804e-06,"reference_active":true,"delta_db":0.07667389371262422}],"active_band_count":3,"active_band_mae_db":0.08389489383759252}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":7.885515861839925e-05,"log_rms_mae_db":2.857487558048214,"rms_envelope_nrmse":0.27317545398518017,"envelope_correlation":0.9460694220920439,"rms_gain_error_db":-0.01294175660241469,"reference_mean_power":0.00019590175284412936,"output_mean_power":0.00019531884444760848},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":1.622675888346401e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":7.225076970752045e-05,"output_power":7.355747026201599e-05,"reference_active":true,"delta_db":0.0778430675241198},{"low_hz":8.0,"high_hz":20.0,"reference_power":6.071619770568576e-06,"output_power":6.019029068457933e-06,"reference_active":true,"delta_db":-0.03778125735736161},{"low_hz":20.0,"high_hz":50.0,"reference_power":3.6400879277670163e-07,"output_power":3.6597445488579873e-07,"reference_active":true,"delta_db":0.02338898250930598}],"active_band_count":3,"active_band_mae_db":0.04633776913026247}}]},{"name":"3000–6000 Hz","status":"measured","filter":{"kind":"butterworth-forward-backward","prototype_order":4,"edges_hz":[3000.0,6000.0],"pad_type":"even","pad_samples":27},"windows":[{"window_ms":10.0,"window_samples":160,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1199,"rms_floor":0.0001240823686589812,"log_rms_mae_db":4.09150994098517,"rms_envelope_nrmse":0.3779582320855748,"envelope_correlation":0.917200243368539,"rms_gain_error_db":0.0029576944851062437,"reference_mean_power":0.00018911407706020164,"output_mean_power":0.0001892429141039065},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1199,"frequency_resolution_hz":0.16680567139282734,"power_floor":2.8583468534278454e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010082037892943071,"output_power":9.73306790181263e-05,"reference_active":true,"delta_db":-0.15298572184282555},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.94400784876693e-05,"output_power":2.622802769915169e-05,"reference_active":true,"delta_db":1.3006761962771998},{"low_hz":20.0,"high_hz":50.0,"reference_power":5.0091360826952136e-06,"output_power":4.496568143421294e-06,"reference_active":true,"delta_db":-0.46881650951810155}],"active_band_count":3,"active_band_mae_db":0.6408261425460423}},{"window_ms":40.0,"window_samples":640,"hop_samples":80,"hop_ms":5.0,"status":"measured","covered_start_sample":0,"covered_end_sample":96000,"uncovered_tail_samples":0,"metrics":{"frame_count":1193,"rms_floor":9.970743476594609e-05,"log_rms_mae_db":3.1395706184733707,"rms_envelope_nrmse":0.2517300599336862,"envelope_correlation":0.9623553368057393,"rms_gain_error_db":0.002939345334853428,"reference_mean_power":0.000190064830092334,"output_mean_power":0.00019019351127361092},"modulation_bands":{"status":"measured","envelope_rate_hz":200.0,"frames":1193,"frequency_resolution_hz":0.16764459346186086,"power_floor":2.7984096251410155e-10,"window":"hann","detrend":"constant","gain_normalization":"none","bands":[{"low_hz":2.0,"high_hz":8.0,"reference_power":0.00010012152579842306,"output_power":9.79084728336339e-05,"reference_active":true,"delta_db":-0.09707182862583862},{"low_hz":8.0,"high_hz":20.0,"reference_power":1.0693666398121261e-05,"output_power":1.4668764563969094e-05,"reference_active":true,"delta_db":1.3726690650157802},{"low_hz":20.0,"high_hz":50.0,"reference_power":2.760055744312888e-07,"output_power":2.8834616600915666e-07,"reference_active":true,"delta_db":0.18996327679099578}],"active_band_count":3,"active_band_mae_db":0.5532347234775382}}]}],"event":null,"silent_intervals":null,"limitations":["有限長雑音の包絡差は独立seedでもゼロにならない","帯域フィルタの端点過渡は測定に含むがイベント測定には使わない","RMS包絡は有声・無声や知覚品質の正解測定ではない"]}}},"kind":"natural_speech","language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"}],"calibration":[{"case":"librispeech_1089-134686-0008","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"8f2c75e70d2dfde4cea5c2020edbeb245b488fa365bd4fe467c3c025be3f6232","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1089","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.073605537414551,"predicted_mos":4.073605537414551,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"10cabd75d8a85c14e77bf1489974a434f5ed836d2c398b22362a8eaa3c9d988f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":46.3957767819993}},{"case":"librispeech_1089-134686-0011","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"268fc1c32f96268178577f4a1e99eea58e49783318f4f0bcffc51b66ffc40254","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1089","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.2903971672058105,"predicted_mos":4.2903971672058105,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f4a352d9828bf71d7120a3cdb583d04065a14d4100cc6ad72a997590d4df31d6","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1210077140003705}},{"case":"librispeech_1089-134686-0032","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"688fca4389961f6b0bd8deda2592f5b06da201e63eda73f6a6bf6edcf15ab69d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1089","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1089","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.574302673339844,"predicted_mos":4.574302673339844,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65440,"input_sample_rate":16000,"input_seconds":4.09,"input_float32_sha256":"bf9b5af891930f0a9556716b887be2e7c74c73daff10b5442892fda8a4088950","internal_samples":65440,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2397291819997918}},{"case":"librispeech_1188-133604-0003","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"03c32db10fd28dbd49c60828b42dce30fddefc66a7c0d86e8a9e08fbf28728d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1188","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.435666084289551,"predicted_mos":3.435666084289551,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7cc851975e60146506e013525ba5cbeee0496db0d607e9aeb5884216aa9c765e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.529655193999133}},{"case":"librispeech_1188-133604-0031","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"b778edbfb1ce196dc35a34be6520c0d9190db4cc0da3e26aad18e4c58b91d02b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1188","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.388341903686523,"predicted_mos":4.388341903686523,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68000,"input_sample_rate":16000,"input_seconds":4.25,"input_float32_sha256":"78197fab1270c63f807e907dcfb4198a85410ce5134c8067a179838fa8706354","internal_samples":68000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5737685250005597}},{"case":"librispeech_1188-133604-0037","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"df50bd0b59a3f15ff3a79f723c41554190031530b7e84b00b8a2f274a4b60bb9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1188","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1188","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.968539237976074,"predicted_mos":3.968539237976074,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"fbb58b0c21de840cb4e115a0c73146db05aa36a4ff3b9fdd97c318dbb5e3b5df","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.060849331999634}},{"case":"librispeech_121-121726-0012","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"2c0d314811a5b914d67dbb04053af5984d0640286aeb77c5c42c29d3053a40a8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"speaker":"librispeech_121","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.6820969581604004,"predicted_mos":3.6820969581604004,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64720,"input_sample_rate":16000,"input_seconds":4.045,"input_float32_sha256":"26fa69114e5ce7a8b5b9b707fdc1678969c8e4bee09c2438054694023a7a465b","internal_samples":64720,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3752159209998354}},{"case":"librispeech_121-123852-0000","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"190d75b03584c6b306515c84e82755dd87db3fa972a2712a9d84ef5e8bb1ad6a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"speaker":"librispeech_121","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.144017219543457,"predicted_mos":4.144017219543457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"24f639a3432ec89f4bec813c254fcc2928b4405206f72116e3136e6265f94a31","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.157900823998716}},{"case":"librispeech_121-127105-0035","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"247582f0c196dd9fd92b943bfe48912bc02639fed23d30bca172503ca6b2fbb0","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_121","speaker_gender_as_distributed":"F"},"speaker":"librispeech_121","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.892571210861206,"predicted_mos":3.892571210861206,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"cddb4be4134f4baa46869e07aa6333d20588e11279af60c4637072acc842238d","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8922604600011255}},{"case":"librispeech_1221-135766-0005","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"1f503547f2d1465a905ed163ab028761aa313d9d0cc8afd85b4bee61486855da","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1221","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.516181468963623,"predicted_mos":4.516181468963623,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7ccd5e4ed85dd1100d19d3f4fd9a2c974dca00abcc6817ea21e805aa15923f99","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0183007100004033}},{"case":"librispeech_1221-135767-0006","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ee7427579155c79e64bde953469fd2dde4fea448d792c6424ab0138fcfd2bfe6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1221","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.328368186950684,"predicted_mos":4.328368186950684,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"56179cfd36d6aead071e45cb2b7effc63e172033c952d5287bd0ef4531eda47f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.2330141380007262}},{"case":"librispeech_1221-135767-0019","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"f06bb7369221d9fddb3c3ff4bd28ae7542d38e76e7463e3e40050b7dd69cd44d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1221","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1221","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.184542655944824,"predicted_mos":4.184542655944824,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60480,"input_sample_rate":16000,"input_seconds":3.78,"input_float32_sha256":"70239a0fa36ebe9bbdd465391645ea523ac00560058f3ddec57c33181feb5125","internal_samples":60480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.9554910530005145}},{"case":"librispeech_1284-1180-0020","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"9d3708aa3f9421b4a1ed1e34b2deee7b18cdd9bf1e184129e207efa961681d78","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1284","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.140274524688721,"predicted_mos":4.140274524688721,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":93920,"input_sample_rate":16000,"input_seconds":5.87,"input_float32_sha256":"c397b5490b9d7fa2f0b21a1024965580556c040b06869432306f99625585032f","internal_samples":93920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5382934999997815}},{"case":"librispeech_1284-1180-0027","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"68a5d8b806eb86b34e801ba4f0969a7a01ecb9ca8c6d1ffc3d90095df5902581","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1284","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.25638484954834,"predicted_mos":4.25638484954834,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":52320,"input_sample_rate":16000,"input_seconds":3.27,"input_float32_sha256":"30486f3b2de234fe157905dd9aa6db4ce32c3a1f10728b8d678fb3b2da754d3e","internal_samples":52320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3248816799987253}},{"case":"librispeech_1284-1180-0028","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"c6bceb07f11a06baea3de8390a53b6801aaef593980c4190e744c8e02a3c132b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1284","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1284","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.210791110992432,"predicted_mos":4.210791110992432,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"abcca764b7cf0ddd8946994bf5b3236d18ad615ea60ec5a099546dcdb7026454","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.327272569998968}},{"case":"librispeech_1320-122617-0019","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"bf028356b65e83cb2ab15fe67bd09746deca0dacc4887b6fd35768fda85d6ff2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1320","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.940176010131836,"predicted_mos":3.940176010131836,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1ea5ba5d2e83ffcaa7657ab2853ba086d5b370841572dfec823bddde3e610925","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.179403412999818}},{"case":"librispeech_1320-122617-0023","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"494d7ccd491f694d116342af5f8f115d162cd05da817d56d28c76c0517ee54c2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1320","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.9725403785705566,"predicted_mos":3.9725403785705566,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bd37eebbe4307f494c9d2c5714d3b9fe2029626edf1ed65d9be6cf7f295fe0d8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.865536761999465}},{"case":"librispeech_1320-122617-0039","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"8dcb2c6673a862ccb02473bbe93b2916a2aa38837da679645bb034918140997e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1320","speaker_gender_as_distributed":"M"},"speaker":"librispeech_1320","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.414013624191284,"predicted_mos":3.414013624191284,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"8700e4d8c918e770f9ed571fbd6d154bae096a64bce40dbf4087cb0c57606c23","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.1736249519999546}},{"case":"librispeech_1580-141083-0010","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"145adf5df07a9a7eec9a7d49645c669e04ebdedcda539f6ef834990e093919e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1580","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.066134929656982,"predicted_mos":4.066134929656982,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85120,"input_sample_rate":16000,"input_seconds":5.32,"input_float32_sha256":"60e50d988c75b342274e7cc9b831c4110606ab222f94130695a684705250abb8","internal_samples":85120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.57721879500059}},{"case":"librispeech_1580-141084-0016","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"a161a72f7b85b3e942b3fe9d69244e1d6507a7b4f7e213241e952cd94a83a6ba","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1580","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.472730875015259,"predicted_mos":3.472730875015259,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":95360,"input_sample_rate":16000,"input_seconds":5.96,"input_float32_sha256":"7d1d126930807769a7c7787f7748c65428a500e0cf7bb3728b161e47fa96e2b6","internal_samples":95360,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1130812789997435}},{"case":"librispeech_1580-141084-0047","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"52bad41cded3b6dbdcd685dfab41e4ec5b1b143787635e562fdb722852bd0723","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1580","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1580","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.520827293395996,"predicted_mos":4.520827293395996,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84000,"input_sample_rate":16000,"input_seconds":5.25,"input_float32_sha256":"c351b90320dd37e220a70447d155c8859fc05d3b743f250c74e7393cf8c02148","internal_samples":84000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7150671529998363}},{"case":"librispeech_1995-1836-0000","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"f0e345b89e3e794d8e6fec53760bc595d338684cd6decd613706bce6896c4a26","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1995","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.015103340148926,"predicted_mos":4.015103340148926,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d82b4a3d4d2f4241d960b592dc1760e20d827057b4dbb01a317f3058a1918c05","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.003123417000097}},{"case":"librispeech_1995-1836-0014","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ed1c5db71bb405ad1d7009f1705d001ebfd5d9f07a94807d7bcc92061b409bf6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1995","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.416754722595215,"predicted_mos":3.416754722595215,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2fa12ac683c7bb10767e025867eac9e5d76e063b603b6d2cf688cd556c972a24","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.006715322000673}},{"case":"librispeech_1995-1837-0023","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"508e303526c2c6002ade422af2aaeb10b37da39845c58f9287f8868c2896f1cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_1995","speaker_gender_as_distributed":"F"},"speaker":"librispeech_1995","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.16876220703125,"predicted_mos":4.16876220703125,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7c529ea9767e4426d960d28c18b5ebc84a9715f189f8e459579735d1d433c3e1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.423594609999782}},{"case":"librispeech_2094-142345-0032","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"9664f2e88981f1ccee68cd933947e9a810f82f0ba4c2d9dcce3cc5644a812b53","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"speaker":"librispeech_2094","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.497434139251709,"predicted_mos":4.497434139251709,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":51840,"input_sample_rate":16000,"input_seconds":3.24,"input_float32_sha256":"80ff88c720d36465c43098db2f00c593a7018953709ca1eca7721afef2c062ba","internal_samples":51840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2921932930003095}},{"case":"librispeech_2094-142345-0039","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"de89944b671e1501f635e242f4e3c2d80de56a7b0aaec440812403f591926c0b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"speaker":"librispeech_2094","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.157635688781738,"predicted_mos":4.157635688781738,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"862eea5ada3154f6046f407df3d570df2b5f37f26eaecb34ab0287c9c6e8325a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9525997979999374}},{"case":"librispeech_2094-142345-0040","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"d297f859114f0799ff2f2920d326f83624c57bcdc03fd9cf7873dff16000d3dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2094","speaker_gender_as_distributed":"F"},"speaker":"librispeech_2094","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.176393985748291,"predicted_mos":4.176393985748291,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":38160,"input_sample_rate":16000,"input_seconds":2.385,"input_float32_sha256":"2ccefe8d31a8f360231ebcccab41b3acc0778bbe0819c93933477e6e89aced02","internal_samples":38160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0330231269999786}},{"case":"librispeech_2300-131720-0022","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"95fda183c946e2ebca88647269dda5af01f680a6926dfbafc2468e344c75a80f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"speaker":"librispeech_2300","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.09295129776001,"predicted_mos":4.09295129776001,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"da124c3f2b63d741f237fa8475795c16d790bec46f1fb7b67c7b8f66670c478c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4800141129999247}},{"case":"librispeech_2300-131720-0028","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"cb6c18fe24eedbb13418b7a83cfcc32eeca63ad3a8fa11972ae570c559caece1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"speaker":"librispeech_2300","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.251725196838379,"predicted_mos":4.251725196838379,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1ea34655dd23b722e2110567cb5f5b1feb53239efe3125fb4b9384ca828f89dc","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.197362624001471}},{"case":"librispeech_2300-131720-0041","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"6fc2f951bd4d3cac06540f04d5cabf9b61a4ce73592101896ce63a45786cf2c1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2300","speaker_gender_as_distributed":"M"},"speaker":"librispeech_2300","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.769174337387085,"predicted_mos":3.769174337387085,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"38e132aca81796256d7fc76b4e832530b7cd5e1090a913b11748397e1dfb1333","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.567875829998229}},{"case":"librispeech_237-126133-0003","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"01f9d20c7a4f345112deb645eb73b9b5b90463fab27653f6bcb4cb53416e25aa","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"speaker":"librispeech_237","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.541549205780029,"predicted_mos":4.541549205780029,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"780d0e36696d3b72761adc5acb2a0730fca079902a3a71a337a87dae2388a789","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8978855290006322}},{"case":"librispeech_237-134500-0029","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"e4a4e8f1b2dfd2faa1f6c4800180b132be5c1d9414c0de798eb1cb3dd6407dcd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"speaker":"librispeech_237","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":2.923694610595703,"predicted_mos":2.923694610595703,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":33040,"input_sample_rate":16000,"input_seconds":2.065,"input_float32_sha256":"533cd620dbfe671a4659bf680292208f5744c534963802a1e31ce0e69d1c8a5c","internal_samples":33040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9785569899995608}},{"case":"librispeech_237-134500-0031","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"f7458f5cd57372c576c6e2c3ff0ecbbb0163c7d6e605635773b976aa0d3d6b94","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_237","speaker_gender_as_distributed":"F"},"speaker":"librispeech_237","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.7653515338897705,"predicted_mos":3.7653515338897705,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"49402d92c425e74fcf31e0293bc450adca73cd447e7dddd8a7ee60679fa3221f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5521389189998445}},{"case":"librispeech_260-123286-0008","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"e160d989cf387b6a6f312b2fd14b06a460bbc0fedf01c7349698111a2b27db0a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"speaker":"librispeech_260","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.942025899887085,"predicted_mos":3.942025899887085,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59760,"input_sample_rate":16000,"input_seconds":3.735,"input_float32_sha256":"1f879e0baae5d4d51551094678fdeb23ba773ef4d866a1b2f2aa67a9fcaca2d0","internal_samples":59760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.57841237699904}},{"case":"librispeech_260-123288-0003","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"12922fedcb52971c14fc130c535bfe582451512b78d5d8df0e9c5fa22fae44fc","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"speaker":"librispeech_260","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.6410369873046875,"predicted_mos":3.6410369873046875,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"8333faa6f2e05fbbee30ea66662dbc72bef5a6df1ed83ed9a3d7de66ccf556b9","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.286923793999449}},{"case":"librispeech_260-123440-0020","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"f103e14578925f427325206f9def059a06db63f2d71d01fb65141b9ea89adfa2","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_260","speaker_gender_as_distributed":"M"},"speaker":"librispeech_260","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.3061916828155518,"predicted_mos":3.3061916828155518,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":79920,"input_sample_rate":16000,"input_seconds":4.995,"input_float32_sha256":"8e022374cf85d7c8f80cd89ab4e054fbbad65546d5e9832f785e1e3ce8bf180a","internal_samples":79920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.285462075000396}},{"case":"librispeech_2830-3980-0000","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"564123165cf754844d6fc8341aa085218f84b8376484e40f5ed6e9e68a841496","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"speaker":"librispeech_2830","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.712026834487915,"predicted_mos":3.712026834487915,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59680,"input_sample_rate":16000,"input_seconds":3.73,"input_float32_sha256":"8f63ce651ea69baf52fe115eb6d0f6053551b53ce699899b6417e104cd136b1d","internal_samples":59680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1629573889986204}},{"case":"librispeech_2830-3980-0038","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"3dc586a374f17fd7598ee21da92af6f6f9af270bfd925d48a76374d017d128d9","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"speaker":"librispeech_2830","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.049131870269775,"predicted_mos":4.049131870269775,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88640,"input_sample_rate":16000,"input_seconds":5.54,"input_float32_sha256":"9fae48abb3c4bca97d7bc79471e9a2e3bfa5c9f304abfc7de9d09035f1ab13a3","internal_samples":88640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6526382610009023}},{"case":"librispeech_2830-3980-0042","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"c39f61fa5b8ff311a5d162de55deb83bdaf963eba163a085aafc78c7698fbe70","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2830","speaker_gender_as_distributed":"M"},"speaker":"librispeech_2830","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.767185211181641,"predicted_mos":4.767185211181641,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48320,"input_sample_rate":16000,"input_seconds":3.02,"input_float32_sha256":"de6b2b3874f099d5354f5235e16a8b2d0e4f870086a1d0af14e9607a95e00862","internal_samples":48320,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.523100094000256}},{"case":"librispeech_2961-960-0010","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"5405a9c97fb9c09ee12ee530c53a60040bf84b2bccd8b84197ec0d56ca37df19","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"speaker":"librispeech_2961","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.5710227489471436,"predicted_mos":3.5710227489471436,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"bc8aec5ef2edfa25d6eb1815bef006ce8ce4386f213753706e16be3e1f3c0ee6","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.418293818998791}},{"case":"librispeech_2961-960-0019","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"c077b3902731393c8c82656194a5bd4884787aa3bb90928b6d1fcc10bd7c6e09","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"speaker":"librispeech_2961","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.478992462158203,"predicted_mos":3.478992462158203,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"5da349fbe2278a97551abe4e062b607a6d6f3ba031b4eb6cbb5164ca7a7666c0","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.645138885000051}},{"case":"librispeech_2961-961-0011","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"01ec611a2f745ad9383737d7b767815c1664f64d6fb59d64700fe0c7ede3d269","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_2961","speaker_gender_as_distributed":"F"},"speaker":"librispeech_2961","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.4711978435516357,"predicted_mos":3.4711978435516357,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"7df238b38f176de7b24ec58cec1306675398b016a2e82f11b1512b374bffcffd","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.144838559999698}},{"case":"librispeech_3570-5694-0001","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"856ea95ba064455e98cd718bf598eeb9be917a75edd8200e6076eb8a628bf7de","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3570","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.253516674041748,"predicted_mos":4.253516674041748,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":90800,"input_sample_rate":16000,"input_seconds":5.675,"input_float32_sha256":"3309caddeb328bf21c4b7ef7b9f165e484371842cb10c51ada50dff7ee8991bc","internal_samples":90800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.974019273000522}},{"case":"librispeech_3570-5694-0004","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"d59a94d7e2fa05c6de936923f7a9da45bf688c0fd19235644f1d2fb52f22ad88","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3570","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.2625603675842285,"predicted_mos":4.2625603675842285,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":85280,"input_sample_rate":16000,"input_seconds":5.33,"input_float32_sha256":"6f2bf49f577cb0bba65458633c45d569de3c13d910c149cbde33dce48bcdec04","internal_samples":85280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0055695440005366}},{"case":"librispeech_3570-5696-0010","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ff7e0fb14fae0fd0e700111ffc2ecf199144b0052e3341e02a2608cc4748a83f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3570","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3570","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.8221404552459717,"predicted_mos":3.8221404552459717,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"27cd80506539ced9831512ca48eba0acd53a1b411666ae1418d04a9c06b4fde3","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.063520792999043}},{"case":"librispeech_3575-170457-0034","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"79f84b3f8b9b4a2012a8f872f10a2930f5eba33f3712fd20c7bb4237140514e3","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3575","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.7620348930358887,"predicted_mos":3.7620348930358887,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":55920,"input_sample_rate":16000,"input_seconds":3.495,"input_float32_sha256":"ac8ff148d90e17b8dd51e29bb3180d903d9284bfe0d3b6105fcd645c44734f25","internal_samples":55920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0907051200010756}},{"case":"librispeech_3575-170457-0041","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"9d6add8daf4676dfd49f4be224a85f9dffa35392a55e0b65c7952213af638630","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3575","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.7349026203155518,"predicted_mos":3.7349026203155518,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d6912402c8c323aee6a7b6a1d12627253745e0f237fc5cb375d9b1184a1e8f0a","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.30072038400067}},{"case":"librispeech_3575-170457-0054","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"158f379fab74f94c27027aefdac522817287aca1719dbfe910ca56925d1b10c8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3575","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3575","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.1323416233062744,"predicted_mos":3.1323416233062744,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"3d1434c601d1d2f47ecd28e2a73e2f1f94013f7309d07bc8576f743d4f9d11bd","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.253767541000343}},{"case":"librispeech_3729-6852-0000","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"b2cd755ecc008f15735a2f15bdedce12fe19f0e5c63dae8f5cd17489ecdd442a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3729","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.2270827293396,"predicted_mos":4.2270827293396,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"b69e34f93bbf40e5b9b1ae373611ffd5a03d9a210266ecf32682443739b5e402","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3913800190002803}},{"case":"librispeech_3729-6852-0007","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"27ded72cf3f10c21bb0ea9cd2c595691a9fe1797015a956a0fb305d7d765459a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3729","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.101188659667969,"predicted_mos":4.101188659667969,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"27dac7420766d0f0fd74833f8ada599c85fe69b6ef7dc16f938fbfed928f23bb","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3938145539996185}},{"case":"librispeech_3729-6852-0017","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ee017a9e9cdf581e2a249279df5f87a8ca5e58438a5c4dc57a1e52bd9064bb9c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_3729","speaker_gender_as_distributed":"F"},"speaker":"librispeech_3729","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.028924465179443,"predicted_mos":4.028924465179443,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a606fe2f723e6d1773dd1b66dda6e2ab056e9def5a98f8cebfb3877194ede967","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.4826362139992852}},{"case":"librispeech_4077-13751-0001","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ca692eb8edf8235874710b3ce4eacf8a6a019b5667a56687941b93591bf94727","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"speaker":"librispeech_4077","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.011976718902588,"predicted_mos":4.011976718902588,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"935015400615ea7003cc3bbd6d27f3f289fa242297b7b2ed9e3a5b447bbd4f4c","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9904246460009745}},{"case":"librispeech_4077-13754-0002","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"560d222c23bd21d55da20d00071cd38d37163381bde3dc54b587b941e2106d81","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"speaker":"librispeech_4077","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.4701428413391113,"predicted_mos":3.4701428413391113,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d5268ab53285e3bc6ed8483a2a804f99f23eece1836d96e4a04451974aa8bb05","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.912747890000901}},{"case":"librispeech_4077-13754-0012","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"fa071bd34babebdc20dfcf3b697f8164f1169a0c0db76b023de174d146616bed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4077","speaker_gender_as_distributed":"M"},"speaker":"librispeech_4077","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":2.7227189540863037,"predicted_mos":2.7227189540863037,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"8b4ad8956f301639d9d0d59dd594d8020fe66ec2a233200c1a9cacf1c7e627c3","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.1362336140009575}},{"case":"librispeech_4446-2271-0003","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"91cdbf5d08dc8408492fd848cb1436130a5e8ea85b420b839c9970a81f2d5e68","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4446","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.4791417121887207,"predicted_mos":3.4791417121887207,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59200,"input_sample_rate":16000,"input_seconds":3.7,"input_float32_sha256":"44460e50fae0ac58d10f8bde235e44f7351a8d9e2530826d2c629443558482d0","internal_samples":59200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.259858252999038}},{"case":"librispeech_4446-2271-0022","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"04376c5c35675defb078fbba4a6b5891fde1a4c0e252f74fcf9be7c56155af1a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4446","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.719693183898926,"predicted_mos":3.719693183898926,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37200,"input_sample_rate":16000,"input_seconds":2.325,"input_float32_sha256":"0fd3f224cd1563164ea2a06e6f6a8357cc009c3410083ad514c7e2f6e55ae0b6","internal_samples":37200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.427238934000343}},{"case":"librispeech_4446-2273-0010","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"43333cefc7220f5a37b58caf0464944a75495317308df87fed74cb1ad5eb184f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4446","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4446","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.523749351501465,"predicted_mos":3.523749351501465,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ff17b39b1cb5e6e39d69b8e58d4e54f6176da9af1035a4bd5e02a7efc5cc19be","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.315867404000528}},{"case":"librispeech_4507-16021-0016","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"200cb4b53bf073d51a6a0f2d807b496d7693350f48909dcadc98509801a4386a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4507","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.265434503555298,"predicted_mos":3.265434503555298,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ced7aaa55f66f55c9b8baa0204ce5bc103752bffba5ba59c179c646a6faf5360","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.956015296998885}},{"case":"librispeech_4507-16021-0032","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"864bce3e777b305b7905ed520fdbc57fef87cdbb8737ca82417258e3318b27ad","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4507","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":2.98887300491333,"predicted_mos":2.98887300491333,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"939abfd897c6303be34824829576e19f920f3abb7f09fa57d03d4b550dc06020","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.654606571999466}},{"case":"librispeech_4507-16021-0045","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"cc56cdadd88c64dc1ed516d187d9104c5ac72742d91e60c01fccf78455717463","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4507","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4507","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.7273411750793457,"predicted_mos":3.7273411750793457,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":78240,"input_sample_rate":16000,"input_seconds":4.89,"input_float32_sha256":"6c835bb3f8878645ff8533ebdcae1eabf5cefa5ea8d153ae2496b238459c99da","internal_samples":78240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":15.330569276999086}},{"case":"librispeech_4970-29093-0004","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"4deba560ea194983e9fdfe0bbf06e0ec3f730e860b9a19cdf13f613f4f310f25","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4970","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.0058486461639404,"predicted_mos":3.0058486461639404,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":60000,"input_sample_rate":16000,"input_seconds":3.75,"input_float32_sha256":"6053e5c91a19f7740d1b7237aee5109b050af7e2b5c2b9624935c889dea23518","internal_samples":60000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.6530782709996856}},{"case":"librispeech_4970-29093-0005","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"c7c4b1b2b4030500d33128c0f0abed5fb602b3c08a2f44819fc29ff7e54a83c7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4970","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":2.6727232933044434,"predicted_mos":2.6727232933044434,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"cbbe34022bfc8eb1a67cc4d76e1352743a0d14f0a9697cad60fed51a41dab133","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.24131848900106}},{"case":"librispeech_4970-29095-0008","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"d9e8326a4b1dbb30af8a974533637f1a73e6c5f12220266a50e3448cde3b30e4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4970","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4970","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.036919593811035,"predicted_mos":4.036919593811035,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":48640,"input_sample_rate":16000,"input_seconds":3.04,"input_float32_sha256":"04cec3b427dbbf48dfd02f834c89e555fcc8e973100feb0211cffc0c22b254cb","internal_samples":48640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.01402096899983}},{"case":"librispeech_4992-23283-0016","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"9a10df052cdbf286199ded20bd9ae4a7df44553f2bbe9d5449e739e5feed1788","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4992","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.6870737075805664,"predicted_mos":3.6870737075805664,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":71920,"input_sample_rate":16000,"input_seconds":4.495,"input_float32_sha256":"3cdb94e1da84305f01e63847deb8e69762ad8006b40fa32d05ff8533b0693535","internal_samples":71920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0730975860005856}},{"case":"librispeech_4992-41797-0000","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"07206995f08d467bfbad7be6f15b2565a069488d70ff3fbcb988195866148ea1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4992","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.561612367630005,"predicted_mos":3.561612367630005,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87760,"input_sample_rate":16000,"input_seconds":5.485,"input_float32_sha256":"9a82559561ea973f3dc9b0d5fd2463947b7f738edf3be97771c3e1a07abb4e55","internal_samples":87760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3630814450007165}},{"case":"librispeech_4992-41806-0000","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"d9b2dd38a3a5b9fc8944a5461f5b883e2c7043a3515d1d4cbf107c1cc74f0149","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_4992","speaker_gender_as_distributed":"F"},"speaker":"librispeech_4992","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.2910542488098145,"predicted_mos":4.2910542488098145,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d3117e8022d7a4b21392ff93373a9fc410b97f793c7b132c978c4013e925a9bd","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.337952190999204}},{"case":"librispeech_5105-28240-0011","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"b64140ffcc51e3dd38f417dd2636ab868d8a641c045f329808082f63f507ad16","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"speaker":"librispeech_5105","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.913151979446411,"predicted_mos":3.913151979446411,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"c1978e86ebbcccd9c43b5680cc6f89db524854c938b34fb6986e5c7e6bd3fc76","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5730945389987028}},{"case":"librispeech_5105-28241-0004","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"1f0a54bb7ef12d0bc36217dd6de373865abdf943d539c8642dbeb130fac8a696","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"speaker":"librispeech_5105","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.6913208961486816,"predicted_mos":3.6913208961486816,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ad077fa222d7666f411b813a32a7331436a48171bff1e7e354b8f3bca4c35557","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.513764629000434}},{"case":"librispeech_5105-28241-0019","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"c3615be3dffc2131482b9c1da56141e30810e6953baf706b08aa08d9b8f40864","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5105","speaker_gender_as_distributed":"M"},"speaker":"librispeech_5105","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.247162103652954,"predicted_mos":3.247162103652954,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":84640,"input_sample_rate":16000,"input_seconds":5.29,"input_float32_sha256":"9a975ab933be439cf9354ae2e4bb0361afc3e57c4e37a574497cd3764b94d55f","internal_samples":84640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.711798799000462}},{"case":"librispeech_5142-33396-0019","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"9ad2d3ebc1745e731a34f461a5b6c863050913d20e2506d2e8862e13ffb153eb","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"speaker":"librispeech_5142","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":2.779564619064331,"predicted_mos":2.779564619064331,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":79760,"input_sample_rate":16000,"input_seconds":4.985,"input_float32_sha256":"f9fb1a92a39a8c50e49a864ca20c1761a0d76de0b668e405071c19a50398c553","internal_samples":79760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5278099269999075}},{"case":"librispeech_5142-33396-0030","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"186dd69950795798e98229733531431639ffa9e88a434b87eb869c3ce6c1e4e7","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"speaker":"librispeech_5142","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.437170505523682,"predicted_mos":4.437170505523682,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44240,"input_sample_rate":16000,"input_seconds":2.765,"input_float32_sha256":"38f2b978cd8d9e0037d87bf71013ec19c8e3acb556672c6fc5de03fd60ac87ad","internal_samples":44240,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.3716731799995614}},{"case":"librispeech_5142-33396-0065","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"dd65f086482ec20aa3a480ca0c1d0039d607a4911958471d47d22f46cd2c0291","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5142","speaker_gender_as_distributed":"F"},"speaker":"librispeech_5142","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.326632499694824,"predicted_mos":4.326632499694824,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83120,"input_sample_rate":16000,"input_seconds":5.195,"input_float32_sha256":"5aad7ecb1f54d37f9f8c6a166eccb362abb255b16dc9eaee2147a04cc1e03d0f","internal_samples":83120,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7330033399994136}},{"case":"librispeech_5639-40744-0013","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"b13c3c891baee028aeb44a1fb87d0941daa65ade2ad7e1ae86f613f3734c46cd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"speaker":"librispeech_5639","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.079738616943359,"predicted_mos":4.079738616943359,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"af23c5afae81553f107efba31683fcf029d531d83ad635a7b328e5995e91ed40","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.848505137000757}},{"case":"librispeech_5639-40744-0023","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"a90ebe89fb50ba87e055b92ce6c6b27cfeeb11bd3f6724a1c055c5e6c736193d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"speaker":"librispeech_5639","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.042727947235107,"predicted_mos":4.042727947235107,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"689f52cca314c6a0cfca585009a65fc6c0481fff0b5471b87373963fdac9d14f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.8631768640007067}},{"case":"librispeech_5639-40744-0030","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"c239ad16f285608bbaed1e01e2ffa336d3f7c8b38553a941605d1564b7b657d1","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5639","speaker_gender_as_distributed":"M"},"speaker":"librispeech_5639","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.415323257446289,"predicted_mos":4.415323257446289,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4a3fdb4ec69b25e35e202c4a0e41f6c2cf1817cd33274c0a0ab218548852a9e5","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.7978171949998796}},{"case":"librispeech_5683-32865-0017","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"baeb7c7cdf1a83c2be5e17ee5e79e97cbde2658ed8cc73bc791f8604d7e0cdff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"speaker":"librispeech_5683","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":2.5191690921783447,"predicted_mos":2.5191690921783447,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":87280,"input_sample_rate":16000,"input_seconds":5.455,"input_float32_sha256":"a2cb55f48070d07f4f7a09d098a56a574637adb9ebd5453423db08b1cff77bb5","internal_samples":87280,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.451029841999116}},{"case":"librispeech_5683-32879-0009","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"257f958f39474b4ff1d24b7b6df61b299e9c632dfff49a8100cd5bfc96f70473","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"speaker":"librispeech_5683","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.7318708896636963,"predicted_mos":3.7318708896636963,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":82160,"input_sample_rate":16000,"input_seconds":5.135,"input_float32_sha256":"c3dbf96155daa7a77fd55580c3af7272825ffcce906d44f210d49fe8d41cf34b","internal_samples":82160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.368942381001034}},{"case":"librispeech_5683-32879-0019","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"b2db7540eb9160dc5644d563e9e48e36c6b0b780fb593a617608c87feafb575e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_5683","speaker_gender_as_distributed":"F"},"speaker":"librispeech_5683","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.4352827072143555,"predicted_mos":3.4352827072143555,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"94acd51133660839055d5b2df725cf443deb1a9a8b39c3a85c4f6c4a7158a4c8","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.0676958359999844}},{"case":"librispeech_61-70968-0042","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"a2e937fb3036e30f7dc72eb8d40dd1701c2782f863356bbed3ce9625cc97b24f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"speaker":"librispeech_61","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.277327060699463,"predicted_mos":4.277327060699463,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44560,"input_sample_rate":16000,"input_seconds":2.785,"input_float32_sha256":"f1b413a3333e8660deda8c827a58772b2c14c5003bd4ad6cf9e98a5a5592cd40","internal_samples":44560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3992010640013177}},{"case":"librispeech_61-70968-0043","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"a882dd67b14ef32c0a17850f88908d8f4c65acbd6198f1d243a122acb813b9dd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"speaker":"librispeech_61","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.179370880126953,"predicted_mos":4.179370880126953,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"a9abc1caa4fa3514a44723d671d8741bec80f3d35c8bd852c7d5b2bf6c44cca1","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.210993728000176}},{"case":"librispeech_61-70968-0054","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"07687018f47d68a7613a69e2cc1fa78cc96bbd25c5d94e9a150f44d76ab5d734","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_61","speaker_gender_as_distributed":"M"},"speaker":"librispeech_61","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.157613277435303,"predicted_mos":4.157613277435303,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"5915f0a66f2b1aa7ffa48d1b641ac26c677d0e3db4129aa851a7c04743878710","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.44403069100008}},{"case":"librispeech_672-122797-0003","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"7760598110a83b94850a11d0a66b5b75c73c6dded5661bcfe38f735b703b03ff","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"speaker":"librispeech_672","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.5692389011383057,"predicted_mos":3.5692389011383057,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76160,"input_sample_rate":16000,"input_seconds":4.76,"input_float32_sha256":"84117bdcd6486d7427c042e289684276fa8699885b4a95eb49400d64454c0426","internal_samples":76160,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.185259328000029}},{"case":"librispeech_672-122797-0044","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"59b5e2e520738994bb37f6108fc19774e01edfd82781f0c4ea1c7b47cfdb4250","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"speaker":"librispeech_672","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.247802734375,"predicted_mos":4.247802734375,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":59840,"input_sample_rate":16000,"input_seconds":3.74,"input_float32_sha256":"91ccd5d317df4e0ef2e9c356c4e56d5cd86286976b50858f9ffa08eb513b65c5","internal_samples":59840,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2306577729996206}},{"case":"librispeech_672-122797-0056","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"e632b94e60e856f03c5e1d4ff531317fd06cd3e3cbd79f1a9cd17c9e89092c41","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_672","speaker_gender_as_distributed":"M"},"speaker":"librispeech_672","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.4085872173309326,"predicted_mos":3.4085872173309326,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":83600,"input_sample_rate":16000,"input_seconds":5.225,"input_float32_sha256":"6fe1ac355644383e56a818b8e2903dd662eda2633844085342aa222e77256ed5","internal_samples":83600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2940390429994295}},{"case":"librispeech_6829-68769-0014","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"de9366a70fcd551627daf6b402a30dab0d733d07436a89e852c6ef0fec68bc0f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"speaker":"librispeech_6829","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.764028310775757,"predicted_mos":3.764028310775757,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":58480,"input_sample_rate":16000,"input_seconds":3.655,"input_float32_sha256":"6c2326dfad893a99b40c2c32714bcb701c5566c86e84cebc17cbca7a7608aa1c","internal_samples":58480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.844933241000035}},{"case":"librispeech_6829-68769-0038","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"6904de2537a637ca4737379aaa742046999dcb209503282889f73e54b9570b57","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"speaker":"librispeech_6829","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.2226064205169678,"predicted_mos":3.2226064205169678,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":40080,"input_sample_rate":16000,"input_seconds":2.505,"input_float32_sha256":"6934a0611fddcb9e50971df510c2041882a867c4b3899e450e1acf3ed2970bc2","internal_samples":40080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2030148429985275}},{"case":"librispeech_6829-68769-0041","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"4ac99fa1efe3570746b8152189c0d7d6769b29024ba23291f9ee1927362c5821","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6829","speaker_gender_as_distributed":"F"},"speaker":"librispeech_6829","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.9850542545318604,"predicted_mos":3.9850542545318604,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":42560,"input_sample_rate":16000,"input_seconds":2.66,"input_float32_sha256":"00d848fcbd4f0e18f07397e09d5c2748448fdd93e0274aa0188591272042e61a","internal_samples":42560,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.08358921800027}},{"case":"librispeech_6930-75918-0013","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"04dec7e13a36b5e5b94b637b84308463a973a752ad37c79bd0c674469224e064","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"speaker":"librispeech_6930","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.062915325164795,"predicted_mos":4.062915325164795,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":47040,"input_sample_rate":16000,"input_seconds":2.94,"input_float32_sha256":"f38f497c4ae6fc9108020c809f36529449591a97f64f030035a5b56ed251477a","internal_samples":47040,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.277591288000622}},{"case":"librispeech_6930-76324-0024","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"9499f92cc4b97907f14b7c877a693db9926ef5a22e240aec2b789f0ee1d06eef","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"speaker":"librispeech_6930","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.9916317462921143,"predicted_mos":3.9916317462921143,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":64800,"input_sample_rate":16000,"input_seconds":4.05,"input_float32_sha256":"04931a10cacc15e36e68c2278ab832d51edf639fd1cddad7eb239ab5af58a2b6","internal_samples":64800,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.50147933000153}},{"case":"librispeech_6930-81414-0026","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"2703ae6a45806eed0c2614dd1227be4c1b4214eacbaa4d65cbf8eb3a6ee76b50","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_6930","speaker_gender_as_distributed":"M"},"speaker":"librispeech_6930","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.508355140686035,"predicted_mos":4.508355140686035,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49200,"input_sample_rate":16000,"input_seconds":3.075,"input_float32_sha256":"c3e7e496a47ea45c5ebb3ea20ddf9db01aa5c6faab9b6e6155057327712135a5","internal_samples":49200,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.29455391800002}},{"case":"librispeech_7021-79740-0002","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"cfe56e2b7a2bea590f2c2e1e816d7ea9bbcd982c147429aae00573e719f163f8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7021","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.963559150695801,"predicted_mos":3.963559150695801,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"4f5ecd7381a554666ca2e944f50c4eb5316c4bb7ec21321405fca3a208837baf","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.6135127839988854}},{"case":"librispeech_7021-79740-0011","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ae04bc4914e2f5e83750f86cb59d9675928b1e44a0a6149b56880d5e627a2317","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7021","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.028006076812744,"predicted_mos":4.028006076812744,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6f9f5b456608d7f44ee223d639bfb84e148501c5efa2b76e94910a9cfb3a5904","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4252490760009096}},{"case":"librispeech_7021-85628-0025","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"226fd879b1a24e6b684bb87b8ee51f6a6cf489d4395c4f5b90d14b2608031fb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7021","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7021","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.9930195808410645,"predicted_mos":3.9930195808410645,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":44400,"input_sample_rate":16000,"input_seconds":2.775,"input_float32_sha256":"9344b5161e42054fd8435a7b6f4d9974066d7a4113abea38945b3ea07bec26b2","internal_samples":44400,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.493004955998913}},{"case":"librispeech_7127-75946-0002","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"884df05f05d892747191a80f7b00e60cdf5b49113a7ca3c8f9c69c6054101abd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7127","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.093369722366333,"predicted_mos":3.093369722366333,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"3d3ac76699a876710096c0ed4bc097114e9e5718acc096decce5c4f6b7632133","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5059247839999443}},{"case":"librispeech_7127-75946-0029","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"af1dc648c3b99ae062977caf3a6bfa24167adf682d81f36b97b884bb40ffc6c6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7127","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":2.5369904041290283,"predicted_mos":2.5369904041290283,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"5a57ed2ebacc92b40830bf9a348d6aabd69d4af04f5e2c4fdc7112113f875372","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.627983731999848}},{"case":"librispeech_7127-75947-0011","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"36d422042dc43d194d3ef019a4c8a4de6ecc475aa6f63d6db75af2def6aa9792","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7127","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7127","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.5510833263397217,"predicted_mos":3.5510833263397217,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":57920,"input_sample_rate":16000,"input_seconds":3.62,"input_float32_sha256":"58c9722bc309337ffb7846c1a829a7be5b0377c076fa205290110ce9beca5085","internal_samples":57920,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":7.7522006390008755}},{"case":"librispeech_7176-88083-0002","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ebdf5f7472633541d54cd08467b25220e802fd6140944ce9729f1598d1dd5e5a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7176","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.166065216064453,"predicted_mos":4.166065216064453,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"825749ad1963b2c54ac9985616177c60eaf18eac859b4852d8b4ba9d6141470f","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.378674064999359}},{"case":"librispeech_7176-88083-0004","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"ad159815493738ed2a35cd10b2831934d60eaf50a5befee706caf056a6ec5eb4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7176","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.508129119873047,"predicted_mos":4.508129119873047,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"44210103bf3e3cc157184e7ee954d25869da0ee1188ff6ed4d1c8f815fd7acf6","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.4512061179993907}},{"case":"librispeech_7176-88083-0026","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"8371ac2373026a5b01d73875b84c2d318473c1b18281baf9c1053d17f540d325","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7176","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7176","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.9027793407440186,"predicted_mos":3.9027793407440186,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":88480,"input_sample_rate":16000,"input_seconds":5.53,"input_float32_sha256":"803ff98e7c871bad36ff22d335d9a777e29a57ddf3910411406ff69435a764b2","internal_samples":88480,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.715736423000635}},{"case":"librispeech_7729-102255-0028","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"d2bdce79b92eb3e74d27d2c7fd9dae73f39b586a14aa7afbf1b34d77c531a34b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7729","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.1194868087768555,"predicted_mos":4.1194868087768555,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"2dd107c270a40dd1c3df5531d50e44b8fed6e67e55658ed3cf75430f90f39a37","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.688656049000201}},{"case":"librispeech_7729-102255-0038","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"242240025bd08ef5c731f41c61f37c04877a98bd88c7eaedaab7bee3e376c82e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7729","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.430004119873047,"predicted_mos":4.430004119873047,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"681e09ee06b146f3f30633f5c64fa9efb0061deaf08f51d8cfda52461d584029","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2609648480010947}},{"case":"librispeech_7729-102255-0042","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"c20c0c4dabed17be81ee6c87286d301405edd4a5005c0ca76c082564c631307d","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_7729","speaker_gender_as_distributed":"M"},"speaker":"librispeech_7729","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.228044509887695,"predicted_mos":4.228044509887695,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"771bae28a9ddc1de4e6b039e3c3bf8346c4942f076859b624472c4dc44a30465","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.885859210999115}},{"case":"librispeech_8224-274381-0010","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"cfbbf8fadec3486c7a03aaf9dad19856e26728a499788f1aa731f0aad6ba6845","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8224","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.5045270919799805,"predicted_mos":3.5045270919799805,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"ecf849f4bd61cc84bae983ed3f28234ecf5f50711e044834178faf84607794b2","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.474378921000607}},{"case":"librispeech_8224-274381-0013","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"2d077c9086bdf2d4b2576bbf269dcf2f0df4a0cd9812e00707e5f3d4b00e3945","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8224","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.8276994228363037,"predicted_mos":3.8276994228363037,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"1150f8c0f52c5a33501feffd6a3d20ec73f26560f9162de27a11499813385b21","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":5.3990938879996975}},{"case":"librispeech_8224-274384-0010","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"d2b99a32b4a5d4da05a8abf9ffcd92946daaa67fb01d2935faabd5c737b55878","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8224","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8224","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.9012155532836914,"predicted_mos":3.9012155532836914,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"0b685c0750383611c038234e972881dd66164719a4c1d40086e12c28397b39de","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.9013266160000057}},{"case":"librispeech_8230-279154-0018","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"7acc60bc75b135f3a1dacbf6cc189f9afa6e4e2415e3b6dac5f2f92d55c034e8","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8230","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.586350679397583,"predicted_mos":3.586350679397583,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"6ed97aea1b2ea4e52b1a3c908f5d690233bad4edd48c4b417c1ed84cd12b1074","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.758262288998594}},{"case":"librispeech_8230-279154-0032","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"eeef365e58757cbc9681c598a7303877c88125cd3a6d473712a502a78be730e6","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8230","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.94101619720459,"predicted_mos":3.94101619720459,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":62080,"input_sample_rate":16000,"input_seconds":3.88,"input_float32_sha256":"6bea6e571970db55e3f2e4de99b31d4d51feef0f6cb5f219b681b9d92d4dccb6","internal_samples":62080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.5862650380004197}},{"case":"librispeech_8230-279154-0033","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"f863a8423f890d3c6c9060541100e4245d487141538814acc424d3b45a1938ed","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8230","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8230","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.739861011505127,"predicted_mos":3.739861011505127,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"07d541eb914552fe1bce57ecc08bb98484f854cd0d6939f87f8b557592a63ec7","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.166523263000272}},{"case":"librispeech_8455-210777-0010","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"6795520c658502bca8975557eedc286083bc835a7b9e2cd209f1900ddb0f986e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8455","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.212153911590576,"predicted_mos":4.212153911590576,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":49600,"input_sample_rate":16000,"input_seconds":3.1,"input_float32_sha256":"f34cb77e42a0793b339c70dd0c299eed04c5bb174b9dfd9f23bc25ae24fb4676","internal_samples":49600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.2921235489993705}},{"case":"librispeech_8455-210777-0049","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"5fd6cf9b661d57f63420ab4d4bab8d1b472c80f649c971f56d0f8462e006017e","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8455","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.054943084716797,"predicted_mos":4.054943084716797,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":65760,"input_sample_rate":16000,"input_seconds":4.11,"input_float32_sha256":"a6a2a45322093cf40be0045ad24596c6bdb50624a3eab13bb2c8e0ba7bf60176","internal_samples":65760,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.440155733998836}},{"case":"librispeech_8455-210777-0060","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"a71fe124c4f06bae3a34a87bb2f344391e1bf4ceeb26a938a9bbbed915ceef8a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8455","speaker_gender_as_distributed":"M"},"speaker":"librispeech_8455","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.6533191204071045,"predicted_mos":3.6533191204071045,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"71c3ff8ad7d01f0df67fe46cab8bf2c04b190bcf82e266a5d48e996f38896307","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":4.006155760000183}},{"case":"librispeech_8463-294828-0003","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"e4d0c619f42a79391341f91588707d57c4f96d434694c5084f5e9fd76c6e023b","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"speaker":"librispeech_8463","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":4.3283371925354,"predicted_mos":4.3283371925354,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"cc028d09fdf0b9dcddce85841840fe9d1e3bddfcaccd48cb8cb2ea05dd1a0fdb","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":2.583535777999714}},{"case":"librispeech_8463-294828-0014","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"e040fa7fffe54e73458e71e66250aab290c6896ebe0be7a8432e3208c9ddc412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"speaker":"librispeech_8463","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.6556360721588135,"predicted_mos":3.6556360721588135,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":91600,"input_sample_rate":16000,"input_seconds":5.725,"input_float32_sha256":"1703405f5a666ca2ab8bcb794a4ce99e1c5f5977c71ded9ac683e4496e61c71b","internal_samples":91600,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7309301650002453}},{"case":"librispeech_8463-294828-0022","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"7a78f3729cdec17debc172ec595fcb13dcbc1acfac9b44a59f4092217cc8069f","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8463","speaker_gender_as_distributed":"F"},"speaker":"librispeech_8463","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.719768762588501,"predicted_mos":3.719768762588501,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":37680,"input_sample_rate":16000,"input_seconds":2.355,"input_float32_sha256":"c3bed8029386dc2619bf5e69fe8c860e5459e1bcf14b2f0b07c54ce70672f5c3","internal_samples":37680,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":2,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.7726752959988517}},{"case":"librispeech_8555-284447-0007","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"9720b5392d525c6de0117cd351e1fe3fc056ac103b3d1d86d67855cc3befcb2a","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"speaker":"librispeech_8555","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.8323888778686523,"predicted_mos":3.8323888778686523,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"29dbbe93727c4ac865cf172a2782845313b7658e5ea40b5c03de76939b47ae36","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.9946629040005064}},{"case":"librispeech_8555-284449-0020","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"e40cd4211d007bc9394208bf2e4486d5ec9a6a5cbbc4b2f86777b251d9c7badd","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"speaker":"librispeech_8555","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.6273300647735596,"predicted_mos":3.6273300647735596,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":81520,"input_sample_rate":16000,"input_seconds":5.095,"input_float32_sha256":"224f98950a81464aa9e795adc9b1038197065447a29b71942cef31644455cd7f","internal_samples":81520,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.128459130999545}},{"case":"librispeech_8555-292519-0001","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"e61c13f859de80e6c97056a3b8b397fd7d074a8bcad3f0d24b2b332d9860d8af","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_8555","speaker_gender_as_distributed":"F"},"speaker":"librispeech_8555","speaker_gender_as_distributed":"F","language":"en","style":"normal","mos":{"status":"measured","value":3.6484482288360596,"predicted_mos":3.6484482288360596,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"d2fa49628c1e044eb66043373ed06871614341f1e9ee81f1f1b06bb2e0453f4b","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":6.033064874000047}},{"case":"librispeech_908-157963-0018","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"a3feefddd101b98726e864acec0bd14cb9b9fe021d3e89399be584d25fc7f412","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"speaker":"librispeech_908","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.8767919540405273,"predicted_mos":3.8767919540405273,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":68080,"input_sample_rate":16000,"input_seconds":4.255,"input_float32_sha256":"4f7ac2354f36190dd460ce2e3d3db121238d43d090b0d66df1811f478e19b90d","internal_samples":68080,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.5909527290004917}},{"case":"librispeech_908-31957-0002","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"03489c63af406dadd390321956e2c3cd034b9969a4c9344230717108a5d4b4d4","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"speaker":"librispeech_908","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":4.012836933135986,"predicted_mos":4.012836933135986,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":76640,"input_sample_rate":16000,"input_seconds":4.79,"input_float32_sha256":"8e4d04598117f3f4c4532ddfe11f136397793be0b75d1a2e03ad35405419f2f4","internal_samples":76640,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.3191363389996695}},{"case":"librispeech_908-31957-0016","method":"identity","status":"measured","evaluation_input":{"pcm_sha256":"88305126e4c150b81c537a015518a730ff0bc9438c9dd2e020c566706104c78c","sample_rate":16000,"language":"en","style":"normal","speaker":"librispeech_908","speaker_gender_as_distributed":"M"},"speaker":"librispeech_908","speaker_gender_as_distributed":"M","language":"en","style":"normal","mos":{"status":"measured","value":3.588137626647949,"predicted_mos":3.588137626647949,"model":"UTMOSv2","language":"en","style":"normal","domain_validation":"未校正の探索的予測","extraction_seed":20261012,"repetitions":1,"input_samples":96000,"input_sample_rate":16000,"input_seconds":6.0,"input_float32_sha256":"f9cf81fa1b43b7486fc8af45ddc81003fab67c8d6aa84878c1f0b7b56caf6d9e","internal_samples":96000,"ssl_processed_samples":48000,"spectrogram_processed_samples_per_frame":22400,"ssl_tile_repetitions":1,"spectrogram_tile_repetitions":1,"reason":null,"seconds":3.8901302000012947}}],"processes":[{"method":"hnm","returncode":0,"log":"logs/hnm-1791866269155712759.log"},{"method":"world_harvest","returncode":0,"log":"logs/world_harvest-1791866440628045620.log"},{"method":"facodec","returncode":0,"completed":120,"planned":120,"initial_attempt":{"method":"facodec","returncode":-9,"log":"logs/facodec-1791866637646380252.log"},"recovery_attempts":[{"attempt":"facodec-001","method":"facodec","started_at":"2026-10-13T06:56:06.615266+00:00","input_ids":["librispeech_61-70968-0043","librispeech_61-70968-0054","librispeech_672-122797-0003","librispeech_672-122797-0044","librispeech_672-122797-0056","librispeech_6829-68769-0014","librispeech_6829-68769-0038","librispeech_6829-68769-0041","librispeech_6930-75918-0013","librispeech_6930-76324-0024"],"input_specifications":[{"id":"librispeech_61-70968-0043","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0043.npy","sha256":"11395567141c5a03808579bb9837eae0b644cb0d8d9e75b3474c4e55832698e7","sample_rate":16000},{"id":"librispeech_61-70968-0054","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0054.npy","sha256":"e07ca663384d7b0ea8ac49ce7905c02e33e1c111573b428d6692890acf3f6ecb","sample_rate":16000},{"id":"librispeech_672-122797-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0003.npy","sha256":"cdb8e35cf9303bf26bc3f9cba71892434576732b577ad73f2aca4b6dcec6f2ac","sample_rate":16000},{"id":"librispeech_672-122797-0044","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0044.npy","sha256":"8287d82eeca7b7a411e5e89ca1c7537c5badcb42acb25504ce71209865ab8804","sample_rate":16000},{"id":"librispeech_672-122797-0056","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0056.npy","sha256":"22780bba4503a457a8d49959a4601aca0ceb4e387b94d1e3c1ba07185cf8e7cb","sample_rate":16000},{"id":"librispeech_6829-68769-0014","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0014.npy","sha256":"3e63bfef2739048aec1a1dd1664bc947427270be279a01e2399186b1cc779f6b","sample_rate":16000},{"id":"librispeech_6829-68769-0038","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0038.npy","sha256":"0d340ad2987da76914c8379223b9ebe6ae6f6ad2c3070d2beb52a071272d0c7a","sample_rate":16000},{"id":"librispeech_6829-68769-0041","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0041.npy","sha256":"0d776c98d75836f78dcbf836cd371840458d509810ee9e94758fbd26584e9d1f","sample_rate":16000},{"id":"librispeech_6930-75918-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-75918-0013.npy","sha256":"321a0414ac42f02e98ce930c29201d21e63bf4d633ddb31c75f7bbcc7654bddb","sample_rate":16000},{"id":"librispeech_6930-76324-0024","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-76324-0024.npy","sha256":"22e1e852f27049dc0fb233b38864a80ef568f82169bbc07c6fb7e15c8d9f1003","sample_rate":16000}],"input_sha256":"f8a3e388071bf9727ecff1873a4026e30588327f12de4dc68828bd3fc0abba48","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","facodec","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/facodec-001-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/facodec","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-facodec-001.log","cold_first_case":"librispeech_61-70968-0043","returncode":0,"finished_at":"2026-10-13T06:57:56.236437+00:00","implementation":{"method":"facodec","prepared":{"schema_version":1,"method":"facodec","prepared_at":"2026-10-13T04:26:02.301643+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/naturalspeech3_facodec/ns3_facodec_encoder.bin":"c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","models/naturalspeech3_facodec/ns3_facodec_decoder.bin":"4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536","src/facodec/ns3_codec/__init__.py":"6b86c557f111bb6c0c1c43fe77bf34ac16ad973411d652fda63449f112cb24c5","src/facodec/ns3_codec/alias_free_torch/__init__.py":"7e8bdb5ab07a79de67d1490600964f60b0a995c6691217f4fcfb76e175a0b79a","src/facodec/ns3_codec/alias_free_torch/act.py":"2add4414a8dec8b5ffe55e55dc26c17aeba31224b5a5fb3959cc82804454b558","src/facodec/ns3_codec/alias_free_torch/filter.py":"674e5fafffd4bc46b58a7de3590b3bc6c2228930e70b50f4ceea014367438dce","src/facodec/ns3_codec/alias_free_torch/resample.py":"9341fe6756d3c8222ecbd4155366ee12aefab8c98ab0cd9badc34a0c6d7eb755","src/facodec/ns3_codec/facodec.py":"086962bbe3b0913491bf1d5874b9005afc8f3a6b7f2707f385e5711f0df84140","src/facodec/ns3_codec/gradient_reversal.py":"8d727bbb012ee5275ad4c823ac747a86db88a69faf393b04b9a46c7408c85ebf","src/facodec/ns3_codec/melspec.py":"6a14da2a4d531f1e8fdb3f83398bfb80c379f5fa41162a28742d6b9dd31b0c3f","src/facodec/ns3_codec/quantize/__init__.py":"69964f5c0ecf93a00b2b4767cc2952cad19e7eebbc7bbd154214c01fc34c2943","src/facodec/ns3_codec/quantize/fvq.py":"40a1efbbcedd5b09fff3db1d91d831114dba49a63d4f58215892d3fe5dbf1c10","src/facodec/ns3_codec/quantize/rvq.py":"e761c3b3f6130e91546830ff4308111ebd23a9e9a7c7443204a358399d98cfd4","src/facodec/ns3_codec/transformer.py":"1cf20b2262cd3dd754b4884f9a134ced9b0ce416332382c00e75b41ef6401f12","src/facodec/setup.py":"1cc676afa26edc2e6d2d0babf2864f787f8940e7d510f74e537f21b45bd1b0e9","src/facodec/test.py":"a4e50e05663a335615c07a93e0586d36a64eaa7fe8cccd317ac95caea882d90a"},"sources":{"facodec":{"repository":"https://github.com/lifeiteng/naturalspeech3_facodec.git","revision":"aac16b7268b1b96e70b9e3526b13e216984e80d0"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":6.125313962000291,"details":{"variant":"FACodec v1 all six codebooks incl. residual","parameter_count":102331180,"checkpoint_sha256":["c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536"]}},"successful_ids":["librispeech_61-70968-0043","librispeech_61-70968-0054","librispeech_672-122797-0003","librispeech_672-122797-0044","librispeech_672-122797-0056","librispeech_6829-68769-0014","librispeech_6829-68769-0038","librispeech_6829-68769-0041","librispeech_6930-75918-0013","librispeech_6930-76324-0024"]},{"attempt":"facodec-002","method":"facodec","started_at":"2026-10-13T06:57:56.238598+00:00","input_ids":["librispeech_6930-81414-0026","librispeech_7021-79740-0002","librispeech_7021-79740-0011","librispeech_7021-85628-0025","librispeech_7127-75946-0002","librispeech_7127-75946-0029","librispeech_7127-75947-0011","librispeech_7176-88083-0002","librispeech_7176-88083-0004","librispeech_7176-88083-0026"],"input_specifications":[{"id":"librispeech_6930-81414-0026","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-81414-0026.npy","sha256":"c88868dff5818474fe738bb4157cb999f7d1139e9923ec83577117e882de4b31","sample_rate":16000},{"id":"librispeech_7021-79740-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-79740-0002.npy","sha256":"6bdd332e5cfad3622e452101f852fa3f8dd6eb5b9be2cdc45f5708ddd2557256","sample_rate":16000},{"id":"librispeech_7021-79740-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-79740-0011.npy","sha256":"15c66be27acb18c9df694af8bc9d44c5bd348f8389a0837a1e4616376dda091f","sample_rate":16000},{"id":"librispeech_7021-85628-0025","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-85628-0025.npy","sha256":"58e5faaa9c6edaa7abbb651251a2000da6e90e3c1943e767c58f215528a210cc","sample_rate":16000},{"id":"librispeech_7127-75946-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75946-0002.npy","sha256":"0981fb6efc50f918e211f5056d645e7f31a403d249c14e3bd928b63982ab99f0","sample_rate":16000},{"id":"librispeech_7127-75946-0029","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75946-0029.npy","sha256":"fb56543512bf946cb8e8680e6e9940fdf458e66eefce85babf56618639407acb","sample_rate":16000},{"id":"librispeech_7127-75947-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75947-0011.npy","sha256":"0a958123a4ddc0a331da3b5e89c6e5b2f141b1c7b4279ba5e6848c7973c3be80","sample_rate":16000},{"id":"librispeech_7176-88083-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0002.npy","sha256":"6ce789c9d4e40824a7597a188ba764b0912cc3d4dceaf7c8ac7aeb28469d4407","sample_rate":16000},{"id":"librispeech_7176-88083-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0004.npy","sha256":"45dd31e9af00a32cf192a87f3fc4d575cf62b88c59a697486955eb85310466db","sample_rate":16000},{"id":"librispeech_7176-88083-0026","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0026.npy","sha256":"d8ac2a315ecbb8b2cd771de10cee21a68a53a6c060b81e6b96a589b954912c21","sample_rate":16000}],"input_sha256":"53915c03d3a84747f4c300887f65e5597df497c7384ed936d74efebe12c1b9f8","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","facodec","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/facodec-002-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/facodec","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-facodec-002.log","cold_first_case":"librispeech_6930-81414-0026","returncode":0,"finished_at":"2026-10-13T06:59:40.724659+00:00","implementation":{"method":"facodec","prepared":{"schema_version":1,"method":"facodec","prepared_at":"2026-10-13T04:26:02.301643+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/naturalspeech3_facodec/ns3_facodec_encoder.bin":"c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","models/naturalspeech3_facodec/ns3_facodec_decoder.bin":"4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536","src/facodec/ns3_codec/__init__.py":"6b86c557f111bb6c0c1c43fe77bf34ac16ad973411d652fda63449f112cb24c5","src/facodec/ns3_codec/alias_free_torch/__init__.py":"7e8bdb5ab07a79de67d1490600964f60b0a995c6691217f4fcfb76e175a0b79a","src/facodec/ns3_codec/alias_free_torch/act.py":"2add4414a8dec8b5ffe55e55dc26c17aeba31224b5a5fb3959cc82804454b558","src/facodec/ns3_codec/alias_free_torch/filter.py":"674e5fafffd4bc46b58a7de3590b3bc6c2228930e70b50f4ceea014367438dce","src/facodec/ns3_codec/alias_free_torch/resample.py":"9341fe6756d3c8222ecbd4155366ee12aefab8c98ab0cd9badc34a0c6d7eb755","src/facodec/ns3_codec/facodec.py":"086962bbe3b0913491bf1d5874b9005afc8f3a6b7f2707f385e5711f0df84140","src/facodec/ns3_codec/gradient_reversal.py":"8d727bbb012ee5275ad4c823ac747a86db88a69faf393b04b9a46c7408c85ebf","src/facodec/ns3_codec/melspec.py":"6a14da2a4d531f1e8fdb3f83398bfb80c379f5fa41162a28742d6b9dd31b0c3f","src/facodec/ns3_codec/quantize/__init__.py":"69964f5c0ecf93a00b2b4767cc2952cad19e7eebbc7bbd154214c01fc34c2943","src/facodec/ns3_codec/quantize/fvq.py":"40a1efbbcedd5b09fff3db1d91d831114dba49a63d4f58215892d3fe5dbf1c10","src/facodec/ns3_codec/quantize/rvq.py":"e761c3b3f6130e91546830ff4308111ebd23a9e9a7c7443204a358399d98cfd4","src/facodec/ns3_codec/transformer.py":"1cf20b2262cd3dd754b4884f9a134ced9b0ce416332382c00e75b41ef6401f12","src/facodec/setup.py":"1cc676afa26edc2e6d2d0babf2864f787f8940e7d510f74e537f21b45bd1b0e9","src/facodec/test.py":"a4e50e05663a335615c07a93e0586d36a64eaa7fe8cccd317ac95caea882d90a"},"sources":{"facodec":{"repository":"https://github.com/lifeiteng/naturalspeech3_facodec.git","revision":"aac16b7268b1b96e70b9e3526b13e216984e80d0"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.814107044001503,"details":{"variant":"FACodec v1 all six codebooks incl. residual","parameter_count":102331180,"checkpoint_sha256":["c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536"]}},"successful_ids":["librispeech_6930-81414-0026","librispeech_7021-79740-0002","librispeech_7021-79740-0011","librispeech_7021-85628-0025","librispeech_7127-75946-0002","librispeech_7127-75946-0029","librispeech_7127-75947-0011","librispeech_7176-88083-0002","librispeech_7176-88083-0004","librispeech_7176-88083-0026"]},{"attempt":"facodec-003","method":"facodec","started_at":"2026-10-13T06:59:40.744011+00:00","input_ids":["librispeech_7729-102255-0028","librispeech_7729-102255-0038","librispeech_7729-102255-0042","librispeech_8224-274381-0010","librispeech_8224-274381-0013","librispeech_8224-274384-0010","librispeech_8230-279154-0018","librispeech_8230-279154-0032","librispeech_8230-279154-0033","librispeech_8455-210777-0010"],"input_specifications":[{"id":"librispeech_7729-102255-0028","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0028.npy","sha256":"bdefd8b41a55fe29ad275ad0cdc0930d498f0e9c563afc84f3f746edbb5622c1","sample_rate":16000},{"id":"librispeech_7729-102255-0038","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0038.npy","sha256":"9e25b2715b44c695b1208e22e4cec275bbd491d99983858945470dba564f7b1d","sample_rate":16000},{"id":"librispeech_7729-102255-0042","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0042.npy","sha256":"8e0eb42509778a5dd3e14ba6fb8dbb8055bbeca1cb9858080ff50cf3162ecbc6","sample_rate":16000},{"id":"librispeech_8224-274381-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274381-0010.npy","sha256":"47f2be273a089762a811311e58170df296cedada30080030c701877e87b9cb90","sample_rate":16000},{"id":"librispeech_8224-274381-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274381-0013.npy","sha256":"8f62ebb1c8f2288d8c4b16ccb1481b74a43e4039ba77e57303a5b3e57b940d29","sample_rate":16000},{"id":"librispeech_8224-274384-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274384-0010.npy","sha256":"742ea1f463a5d6275ca4dea75eb76e3d478fb9b550110c351f3d7dd4d220e2be","sample_rate":16000},{"id":"librispeech_8230-279154-0018","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0018.npy","sha256":"812b9900ccbe6b84e9444fed1c7770731b7dfca39437dccad215d79db129df01","sample_rate":16000},{"id":"librispeech_8230-279154-0032","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0032.npy","sha256":"13541ca079afe56ea25803a0e8910f0e6fb5f0e24063e867981a992039b868f4","sample_rate":16000},{"id":"librispeech_8230-279154-0033","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0033.npy","sha256":"88f96bf577b1f1b95250b099c6716d41e7e2bed368df2791fdb21adff8013cf5","sample_rate":16000},{"id":"librispeech_8455-210777-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0010.npy","sha256":"153385cc5de0b72725a7f9cb203194181b7d234e40b71aae096375c5bfee23b6","sample_rate":16000}],"input_sha256":"a83c279dcfae53a9c983691a33798638f8fa0a19ebb8042903a2ba39d755de0e","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","facodec","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/facodec-003-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/facodec","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-facodec-003.log","cold_first_case":"librispeech_7729-102255-0028","returncode":0,"finished_at":"2026-10-13T07:01:21.794703+00:00","implementation":{"method":"facodec","prepared":{"schema_version":1,"method":"facodec","prepared_at":"2026-10-13T04:26:02.301643+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/naturalspeech3_facodec/ns3_facodec_encoder.bin":"c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","models/naturalspeech3_facodec/ns3_facodec_decoder.bin":"4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536","src/facodec/ns3_codec/__init__.py":"6b86c557f111bb6c0c1c43fe77bf34ac16ad973411d652fda63449f112cb24c5","src/facodec/ns3_codec/alias_free_torch/__init__.py":"7e8bdb5ab07a79de67d1490600964f60b0a995c6691217f4fcfb76e175a0b79a","src/facodec/ns3_codec/alias_free_torch/act.py":"2add4414a8dec8b5ffe55e55dc26c17aeba31224b5a5fb3959cc82804454b558","src/facodec/ns3_codec/alias_free_torch/filter.py":"674e5fafffd4bc46b58a7de3590b3bc6c2228930e70b50f4ceea014367438dce","src/facodec/ns3_codec/alias_free_torch/resample.py":"9341fe6756d3c8222ecbd4155366ee12aefab8c98ab0cd9badc34a0c6d7eb755","src/facodec/ns3_codec/facodec.py":"086962bbe3b0913491bf1d5874b9005afc8f3a6b7f2707f385e5711f0df84140","src/facodec/ns3_codec/gradient_reversal.py":"8d727bbb012ee5275ad4c823ac747a86db88a69faf393b04b9a46c7408c85ebf","src/facodec/ns3_codec/melspec.py":"6a14da2a4d531f1e8fdb3f83398bfb80c379f5fa41162a28742d6b9dd31b0c3f","src/facodec/ns3_codec/quantize/__init__.py":"69964f5c0ecf93a00b2b4767cc2952cad19e7eebbc7bbd154214c01fc34c2943","src/facodec/ns3_codec/quantize/fvq.py":"40a1efbbcedd5b09fff3db1d91d831114dba49a63d4f58215892d3fe5dbf1c10","src/facodec/ns3_codec/quantize/rvq.py":"e761c3b3f6130e91546830ff4308111ebd23a9e9a7c7443204a358399d98cfd4","src/facodec/ns3_codec/transformer.py":"1cf20b2262cd3dd754b4884f9a134ced9b0ce416332382c00e75b41ef6401f12","src/facodec/setup.py":"1cc676afa26edc2e6d2d0babf2864f787f8940e7d510f74e537f21b45bd1b0e9","src/facodec/test.py":"a4e50e05663a335615c07a93e0586d36a64eaa7fe8cccd317ac95caea882d90a"},"sources":{"facodec":{"repository":"https://github.com/lifeiteng/naturalspeech3_facodec.git","revision":"aac16b7268b1b96e70b9e3526b13e216984e80d0"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":4.530719553000381,"details":{"variant":"FACodec v1 all six codebooks incl. residual","parameter_count":102331180,"checkpoint_sha256":["c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536"]}},"successful_ids":["librispeech_7729-102255-0028","librispeech_7729-102255-0038","librispeech_7729-102255-0042","librispeech_8224-274381-0010","librispeech_8224-274381-0013","librispeech_8224-274384-0010","librispeech_8230-279154-0018","librispeech_8230-279154-0032","librispeech_8230-279154-0033","librispeech_8455-210777-0010"]},{"attempt":"facodec-004","method":"facodec","started_at":"2026-10-13T07:01:21.801689+00:00","input_ids":["librispeech_8455-210777-0049","librispeech_8455-210777-0060","librispeech_8463-294828-0003","librispeech_8463-294828-0014","librispeech_8463-294828-0022","librispeech_8555-284447-0007","librispeech_8555-284449-0020","librispeech_8555-292519-0001","librispeech_908-157963-0018","librispeech_908-31957-0002"],"input_specifications":[{"id":"librispeech_8455-210777-0049","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0049.npy","sha256":"2d76f8265f7a52a2814eb8e00d8edc511804ac1369740fd998f7cd19435e2fe8","sample_rate":16000},{"id":"librispeech_8455-210777-0060","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0060.npy","sha256":"9b6304a33d078fd527310e09782e4424d1a63e37c4d42866c9aa87d2920ab687","sample_rate":16000},{"id":"librispeech_8463-294828-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0003.npy","sha256":"299a537dc8c17a869309de103cb4a613416f5a2ddbabf2579d3eae9635449db9","sample_rate":16000},{"id":"librispeech_8463-294828-0014","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0014.npy","sha256":"57c0273c8992ab87674c0b415b90f57ef6597f45937ac6ef94c308d8b0bb1aaf","sample_rate":16000},{"id":"librispeech_8463-294828-0022","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0022.npy","sha256":"908baa31161703eb40e83b7043052a4a50ebf6e6dc303d833ee950422de3dd86","sample_rate":16000},{"id":"librispeech_8555-284447-0007","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-284447-0007.npy","sha256":"93bf4a5143f192f1bec4014f5ed1882d65fcf3ffef7d30a7c6262721decfe4b4","sample_rate":16000},{"id":"librispeech_8555-284449-0020","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-284449-0020.npy","sha256":"0ad90c10b2c7d8a7b3190141694c0f9202a9fa41f2cc5c03ef81a8fffa9a0703","sample_rate":16000},{"id":"librispeech_8555-292519-0001","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-292519-0001.npy","sha256":"1d71c61f160ff40a72fef1912367b5ac430b341d3c951361c0a870cdcd7e39fe","sample_rate":16000},{"id":"librispeech_908-157963-0018","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-157963-0018.npy","sha256":"068cbe41c56c5de3cb2928395586f7d207582cd51ef6612040fe2872c397e9dd","sample_rate":16000},{"id":"librispeech_908-31957-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-31957-0002.npy","sha256":"b69e3ce913f862c4a8715e3c55451386e805eade08afccb7a2d3eb6e55696791","sample_rate":16000}],"input_sha256":"888dee89260758d087a5c3871c0277765b7042929ea9a43cd60155dcb249f253","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","facodec","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/facodec-004-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/facodec","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-facodec-004.log","cold_first_case":"librispeech_8455-210777-0049","returncode":0,"finished_at":"2026-10-13T07:03:14.581774+00:00","implementation":{"method":"facodec","prepared":{"schema_version":1,"method":"facodec","prepared_at":"2026-10-13T04:26:02.301643+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/naturalspeech3_facodec/ns3_facodec_encoder.bin":"c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","models/naturalspeech3_facodec/ns3_facodec_decoder.bin":"4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536","src/facodec/ns3_codec/__init__.py":"6b86c557f111bb6c0c1c43fe77bf34ac16ad973411d652fda63449f112cb24c5","src/facodec/ns3_codec/alias_free_torch/__init__.py":"7e8bdb5ab07a79de67d1490600964f60b0a995c6691217f4fcfb76e175a0b79a","src/facodec/ns3_codec/alias_free_torch/act.py":"2add4414a8dec8b5ffe55e55dc26c17aeba31224b5a5fb3959cc82804454b558","src/facodec/ns3_codec/alias_free_torch/filter.py":"674e5fafffd4bc46b58a7de3590b3bc6c2228930e70b50f4ceea014367438dce","src/facodec/ns3_codec/alias_free_torch/resample.py":"9341fe6756d3c8222ecbd4155366ee12aefab8c98ab0cd9badc34a0c6d7eb755","src/facodec/ns3_codec/facodec.py":"086962bbe3b0913491bf1d5874b9005afc8f3a6b7f2707f385e5711f0df84140","src/facodec/ns3_codec/gradient_reversal.py":"8d727bbb012ee5275ad4c823ac747a86db88a69faf393b04b9a46c7408c85ebf","src/facodec/ns3_codec/melspec.py":"6a14da2a4d531f1e8fdb3f83398bfb80c379f5fa41162a28742d6b9dd31b0c3f","src/facodec/ns3_codec/quantize/__init__.py":"69964f5c0ecf93a00b2b4767cc2952cad19e7eebbc7bbd154214c01fc34c2943","src/facodec/ns3_codec/quantize/fvq.py":"40a1efbbcedd5b09fff3db1d91d831114dba49a63d4f58215892d3fe5dbf1c10","src/facodec/ns3_codec/quantize/rvq.py":"e761c3b3f6130e91546830ff4308111ebd23a9e9a7c7443204a358399d98cfd4","src/facodec/ns3_codec/transformer.py":"1cf20b2262cd3dd754b4884f9a134ced9b0ce416332382c00e75b41ef6401f12","src/facodec/setup.py":"1cc676afa26edc2e6d2d0babf2864f787f8940e7d510f74e537f21b45bd1b0e9","src/facodec/test.py":"a4e50e05663a335615c07a93e0586d36a64eaa7fe8cccd317ac95caea882d90a"},"sources":{"facodec":{"repository":"https://github.com/lifeiteng/naturalspeech3_facodec.git","revision":"aac16b7268b1b96e70b9e3526b13e216984e80d0"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":4.454492849999951,"details":{"variant":"FACodec v1 all six codebooks incl. residual","parameter_count":102331180,"checkpoint_sha256":["c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536"]}},"successful_ids":["librispeech_8455-210777-0049","librispeech_8455-210777-0060","librispeech_8463-294828-0003","librispeech_8463-294828-0014","librispeech_8463-294828-0022","librispeech_8555-284447-0007","librispeech_8555-284449-0020","librispeech_8555-292519-0001","librispeech_908-157963-0018","librispeech_908-31957-0002"]},{"attempt":"facodec-005","method":"facodec","started_at":"2026-10-13T07:03:14.597165+00:00","input_ids":["librispeech_908-31957-0016"],"input_specifications":[{"id":"librispeech_908-31957-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-31957-0016.npy","sha256":"e82f7cc1bcf52dc8247694fec36c8608076490306e8962865090852777303b82","sample_rate":16000}],"input_sha256":"27d9d0e7957b81f424033b79cf0545533c0b737c57cd741377e208cb83dbe4d2","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","facodec","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/facodec-005-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/facodec","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-facodec-005.log","cold_first_case":"librispeech_908-31957-0016","returncode":0,"finished_at":"2026-10-13T07:03:40.341223+00:00","implementation":{"method":"facodec","prepared":{"schema_version":1,"method":"facodec","prepared_at":"2026-10-13T04:26:02.301643+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/naturalspeech3_facodec/ns3_facodec_encoder.bin":"c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","models/naturalspeech3_facodec/ns3_facodec_decoder.bin":"4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536","src/facodec/ns3_codec/__init__.py":"6b86c557f111bb6c0c1c43fe77bf34ac16ad973411d652fda63449f112cb24c5","src/facodec/ns3_codec/alias_free_torch/__init__.py":"7e8bdb5ab07a79de67d1490600964f60b0a995c6691217f4fcfb76e175a0b79a","src/facodec/ns3_codec/alias_free_torch/act.py":"2add4414a8dec8b5ffe55e55dc26c17aeba31224b5a5fb3959cc82804454b558","src/facodec/ns3_codec/alias_free_torch/filter.py":"674e5fafffd4bc46b58a7de3590b3bc6c2228930e70b50f4ceea014367438dce","src/facodec/ns3_codec/alias_free_torch/resample.py":"9341fe6756d3c8222ecbd4155366ee12aefab8c98ab0cd9badc34a0c6d7eb755","src/facodec/ns3_codec/facodec.py":"086962bbe3b0913491bf1d5874b9005afc8f3a6b7f2707f385e5711f0df84140","src/facodec/ns3_codec/gradient_reversal.py":"8d727bbb012ee5275ad4c823ac747a86db88a69faf393b04b9a46c7408c85ebf","src/facodec/ns3_codec/melspec.py":"6a14da2a4d531f1e8fdb3f83398bfb80c379f5fa41162a28742d6b9dd31b0c3f","src/facodec/ns3_codec/quantize/__init__.py":"69964f5c0ecf93a00b2b4767cc2952cad19e7eebbc7bbd154214c01fc34c2943","src/facodec/ns3_codec/quantize/fvq.py":"40a1efbbcedd5b09fff3db1d91d831114dba49a63d4f58215892d3fe5dbf1c10","src/facodec/ns3_codec/quantize/rvq.py":"e761c3b3f6130e91546830ff4308111ebd23a9e9a7c7443204a358399d98cfd4","src/facodec/ns3_codec/transformer.py":"1cf20b2262cd3dd754b4884f9a134ced9b0ce416332382c00e75b41ef6401f12","src/facodec/setup.py":"1cc676afa26edc2e6d2d0babf2864f787f8940e7d510f74e537f21b45bd1b0e9","src/facodec/test.py":"a4e50e05663a335615c07a93e0586d36a64eaa7fe8cccd317ac95caea882d90a"},"sources":{"facodec":{"repository":"https://github.com/lifeiteng/naturalspeech3_facodec.git","revision":"aac16b7268b1b96e70b9e3526b13e216984e80d0"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":6.250011363001249,"details":{"variant":"FACodec v1 all six codebooks incl. residual","parameter_count":102331180,"checkpoint_sha256":["c8fa7b8d79ccd915e6c3b19cd3c9d5f4cba3fbeb794a2e598d7e31db8efee47e","4b8f2c9a691385c1a61542585116f52263a029758dcb4a3a955ff2c190afd536"]}},"successful_ids":["librispeech_908-31957-0016"]}],"log":"logs/recovery-facodec-005.log","history":"preparation.json / recovery_execution and recovery/initial-processes.json"},{"method":"hifiglot","returncode":0,"completed":120,"planned":120,"initial_attempt":{"method":"hifiglot","returncode":-9,"log":"logs/hifiglot-1791867402620840354.log"},"recovery_attempts":[{"attempt":"hifiglot-001","method":"hifiglot","started_at":"2026-10-13T07:03:40.362049+00:00","input_ids":["librispeech_1580-141084-0016","librispeech_1580-141084-0047","librispeech_1995-1836-0000","librispeech_1995-1836-0014","librispeech_1995-1837-0023","librispeech_2094-142345-0032","librispeech_2094-142345-0039","librispeech_2094-142345-0040","librispeech_2300-131720-0022","librispeech_2300-131720-0028"],"input_specifications":[{"id":"librispeech_1580-141084-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_1580-141084-0016.npy","sha256":"93e1881b9bee5339d8df1b2c0d3aaebe7308511149012f5220080d4003f1ce2c","sample_rate":16000},{"id":"librispeech_1580-141084-0047","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_1580-141084-0047.npy","sha256":"47811c228da75f3becc352fe4eec0fd525c13a036f7e91e4f7d68d6b945e3c79","sample_rate":16000},{"id":"librispeech_1995-1836-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_1995-1836-0000.npy","sha256":"9b4757ae23cce8e56532839de5fdf9b43c346388e66b2494c20bf772f7621084","sample_rate":16000},{"id":"librispeech_1995-1836-0014","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_1995-1836-0014.npy","sha256":"9a99a3a90e1f4c343f336f0da5022029b2f18c4ff64d633641d0471804c64565","sample_rate":16000},{"id":"librispeech_1995-1837-0023","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_1995-1837-0023.npy","sha256":"44084159233cef136382433cc3d4cb8938a7407571790316e3a868a2292e6e56","sample_rate":16000},{"id":"librispeech_2094-142345-0032","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2094-142345-0032.npy","sha256":"8096cbcdd3e093fefedfc7d4466fc50ab9467c5c11071c10b83dc920b31f9ec9","sample_rate":16000},{"id":"librispeech_2094-142345-0039","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2094-142345-0039.npy","sha256":"e88159d23ff9307495753f2c3452423ef53b095a3206159f6a200988e22b9758","sample_rate":16000},{"id":"librispeech_2094-142345-0040","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2094-142345-0040.npy","sha256":"a413a9c3cb359ef7f53b985134e93c63b5eacc911b858d84965efaf6c1fb910d","sample_rate":16000},{"id":"librispeech_2300-131720-0022","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2300-131720-0022.npy","sha256":"ba50eb8b1e006b7341f273bf21fa396079a09996fabc4f27cde2993eeda0f3d7","sample_rate":16000},{"id":"librispeech_2300-131720-0028","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2300-131720-0028.npy","sha256":"cf9bc8e7c16c92de5a6a34b62b32a7e149958c948fb1e3cc5e20a5dd2b9cedfe","sample_rate":16000}],"input_sha256":"d8e495d14f23d9c0b797e31747d3e3ecd150bc9e69791cec6a1ba645619c675d","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-001-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-001.log","cold_first_case":"librispeech_1580-141084-0016","returncode":0,"finished_at":"2026-10-13T07:06:39.868445+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.8181539160013926,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_1580-141084-0016","librispeech_1580-141084-0047","librispeech_1995-1836-0000","librispeech_1995-1836-0014","librispeech_1995-1837-0023","librispeech_2094-142345-0032","librispeech_2094-142345-0039","librispeech_2094-142345-0040","librispeech_2300-131720-0022","librispeech_2300-131720-0028"]},{"attempt":"hifiglot-002","method":"hifiglot","started_at":"2026-10-13T07:06:39.882790+00:00","input_ids":["librispeech_2300-131720-0041","librispeech_237-126133-0003","librispeech_237-134500-0029","librispeech_237-134500-0031","librispeech_260-123286-0008","librispeech_260-123288-0003","librispeech_260-123440-0020","librispeech_2830-3980-0000","librispeech_2830-3980-0038","librispeech_2830-3980-0042"],"input_specifications":[{"id":"librispeech_2300-131720-0041","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2300-131720-0041.npy","sha256":"1fb0d0ccb72963d69461dc23b437371942e1244437c77215b6a1e1be34891b3d","sample_rate":16000},{"id":"librispeech_237-126133-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_237-126133-0003.npy","sha256":"3090a528faff51d26d0ea1fe3914063151c1e8eaf9354b24e51173247b7a69fc","sample_rate":16000},{"id":"librispeech_237-134500-0029","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_237-134500-0029.npy","sha256":"71d86eaff9e13fc4a8e0e41c271e36be5ba1a6cd179c3ba3a6931b89dbd78206","sample_rate":16000},{"id":"librispeech_237-134500-0031","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_237-134500-0031.npy","sha256":"66b8f0fa3a5988c5f43c9ed60894944381007f86fe6b413cf7370344f5a1b084","sample_rate":16000},{"id":"librispeech_260-123286-0008","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_260-123286-0008.npy","sha256":"fde9082de1536b3cb7b9b64030ee3c897ee296663d8ee9fc2fea3b658c7a179e","sample_rate":16000},{"id":"librispeech_260-123288-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_260-123288-0003.npy","sha256":"380f18150746b92f548e2cc642146570a9fdb278b3fbf7470f430b6e5ea125e7","sample_rate":16000},{"id":"librispeech_260-123440-0020","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_260-123440-0020.npy","sha256":"66ad247eff03a910dee5b9feed83094cbef18c21f694b77ba10c47f4d65c570a","sample_rate":16000},{"id":"librispeech_2830-3980-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2830-3980-0000.npy","sha256":"fd97b4748ed6a2437861ffee06416e2dbd09229870db6e19899ec2d39491809a","sample_rate":16000},{"id":"librispeech_2830-3980-0038","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2830-3980-0038.npy","sha256":"34283f63866e01407d3dd8060fcc0517a3925172939da95984cc41b8a93e3ecd","sample_rate":16000},{"id":"librispeech_2830-3980-0042","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2830-3980-0042.npy","sha256":"0042d8afd9bfdec43aeb47b3c8d03f059e19efc4a4103d4bd165e674531c4a9b","sample_rate":16000}],"input_sha256":"f0ff4912377d86be97bd00a37c33dbff9565f2138de1eebe6f014ae8544d1846","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-002-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-002.log","cold_first_case":"librispeech_2300-131720-0041","returncode":0,"finished_at":"2026-10-13T07:09:58.576992+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.5163310499992804,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_2300-131720-0041","librispeech_237-126133-0003","librispeech_237-134500-0029","librispeech_237-134500-0031","librispeech_260-123286-0008","librispeech_260-123288-0003","librispeech_260-123440-0020","librispeech_2830-3980-0000","librispeech_2830-3980-0038","librispeech_2830-3980-0042"]},{"attempt":"hifiglot-003","method":"hifiglot","started_at":"2026-10-13T07:09:58.603277+00:00","input_ids":["librispeech_2961-960-0010","librispeech_2961-960-0019","librispeech_2961-961-0011","librispeech_3570-5694-0001","librispeech_3570-5694-0004","librispeech_3570-5696-0010","librispeech_3575-170457-0034","librispeech_3575-170457-0041","librispeech_3575-170457-0054","librispeech_3729-6852-0000"],"input_specifications":[{"id":"librispeech_2961-960-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2961-960-0010.npy","sha256":"78b0dff6046aa18d21883b8e7e1a87010826b65f7bd7e20d432dff9e773dfeb2","sample_rate":16000},{"id":"librispeech_2961-960-0019","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2961-960-0019.npy","sha256":"3aaa036a73b4b15888e1926e1d8bc85c1d6623633692de1b8c03961c9680dd7d","sample_rate":16000},{"id":"librispeech_2961-961-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_2961-961-0011.npy","sha256":"a0ddd73876a677f0ddbd97b071a56f85b42d159a0706fb433464b949cb2d2079","sample_rate":16000},{"id":"librispeech_3570-5694-0001","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3570-5694-0001.npy","sha256":"04fce52520379ca7b0df597b54831a312b6b920b3c63943448968dd0e39c4e24","sample_rate":16000},{"id":"librispeech_3570-5694-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3570-5694-0004.npy","sha256":"2a7d7164eedcdf7707de438f724ea4169c4a77e4b474dae06130777a0dc42808","sample_rate":16000},{"id":"librispeech_3570-5696-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3570-5696-0010.npy","sha256":"091793e4f5b935128df697e4a23a90f93fc6a9b25f5cc0dd77806c81e18fe35a","sample_rate":16000},{"id":"librispeech_3575-170457-0034","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3575-170457-0034.npy","sha256":"f057399e7bc52496d756298eb590948a7652ca6ab292cde23ed3a5d6e4f624f5","sample_rate":16000},{"id":"librispeech_3575-170457-0041","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3575-170457-0041.npy","sha256":"6be0695d1829e2c748580d3856118cd6eabdb573ad8adcd1eb7f3b1824b606fa","sample_rate":16000},{"id":"librispeech_3575-170457-0054","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3575-170457-0054.npy","sha256":"bf86c0c00709c2e5eee207cba7f35d17746c67a2a9f9c9ec302b8a3636061627","sample_rate":16000},{"id":"librispeech_3729-6852-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3729-6852-0000.npy","sha256":"1b783a7c2c383d87de40953142aeba194d79926cf87591d0c55408731d8625ce","sample_rate":16000}],"input_sha256":"a35bdbc384898cf8c788cd31572039d017c8e19ff84192d0f29514bfba94d7bb","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-003-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-003.log","cold_first_case":"librispeech_2961-960-0010","returncode":0,"finished_at":"2026-10-13T07:12:43.522238+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.526378943999589,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_2961-960-0010","librispeech_2961-960-0019","librispeech_2961-961-0011","librispeech_3570-5694-0001","librispeech_3570-5694-0004","librispeech_3570-5696-0010","librispeech_3575-170457-0034","librispeech_3575-170457-0041","librispeech_3575-170457-0054","librispeech_3729-6852-0000"]},{"attempt":"hifiglot-004","method":"hifiglot","started_at":"2026-10-13T07:12:43.538183+00:00","input_ids":["librispeech_3729-6852-0007","librispeech_3729-6852-0017","librispeech_4077-13751-0001","librispeech_4077-13754-0002","librispeech_4077-13754-0012","librispeech_4446-2271-0003","librispeech_4446-2271-0022","librispeech_4446-2273-0010","librispeech_4507-16021-0016","librispeech_4507-16021-0032"],"input_specifications":[{"id":"librispeech_3729-6852-0007","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3729-6852-0007.npy","sha256":"5ab238d03e046aeef6b327b51b75630087f055df4633f0a25c71d45f3af209b7","sample_rate":16000},{"id":"librispeech_3729-6852-0017","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3729-6852-0017.npy","sha256":"fa4d86d7b8c74df263aaabe4276525584eed7ed36d8b9421d2158a8fad60fd8c","sample_rate":16000},{"id":"librispeech_4077-13751-0001","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4077-13751-0001.npy","sha256":"e251b8c76c67e0e5339a7ad31f65d12566654050606a6b0e7f395fe9bfdd03bf","sample_rate":16000},{"id":"librispeech_4077-13754-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4077-13754-0002.npy","sha256":"5ebee7ab93321e8f9c22488b77a9d83b935967cc2b1755b26ca51d20e4dbcdd6","sample_rate":16000},{"id":"librispeech_4077-13754-0012","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4077-13754-0012.npy","sha256":"da6d5022702fb704c7663cfd01a247727508d36f33ba6cdf57c29ec0bca2540e","sample_rate":16000},{"id":"librispeech_4446-2271-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4446-2271-0003.npy","sha256":"8fd653ce4da1222d377403b051d957dbbd98eca247e0a76171fa28bf4a523f5f","sample_rate":16000},{"id":"librispeech_4446-2271-0022","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4446-2271-0022.npy","sha256":"0e102c1dba2f0649f1df1e8dfcc4595e3cf32e9e0dcf6815b2f748a1174aa58c","sample_rate":16000},{"id":"librispeech_4446-2273-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4446-2273-0010.npy","sha256":"66fab68f8236f2b9617cfd6928803d223cdcd7756c14f365c225f0ed88f58739","sample_rate":16000},{"id":"librispeech_4507-16021-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4507-16021-0016.npy","sha256":"bf03704b4a0939f52cf691a192ebddece3d996e02ed4c4bccd90a88cb5b78db0","sample_rate":16000},{"id":"librispeech_4507-16021-0032","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4507-16021-0032.npy","sha256":"ef0e7a805e8d327859d4b390159adeb5aea5e9adfcd0a91a422cd585becc0a79","sample_rate":16000}],"input_sha256":"90ac4a3682785457b1bef05445e10a9c947d23152dc5bb7dde1e5b5aaf026dc8","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-004-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-004.log","cold_first_case":"librispeech_3729-6852-0007","returncode":0,"finished_at":"2026-10-13T07:14:37.572827+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.6595868929998687,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_3729-6852-0007","librispeech_3729-6852-0017","librispeech_4077-13751-0001","librispeech_4077-13754-0002","librispeech_4077-13754-0012","librispeech_4446-2271-0003","librispeech_4446-2271-0022","librispeech_4446-2273-0010","librispeech_4507-16021-0016","librispeech_4507-16021-0032"]},{"attempt":"hifiglot-005","method":"hifiglot","started_at":"2026-10-13T07:14:37.598234+00:00","input_ids":["librispeech_4507-16021-0045","librispeech_4970-29093-0004","librispeech_4970-29093-0005","librispeech_4970-29095-0008","librispeech_4992-23283-0016","librispeech_4992-41797-0000","librispeech_4992-41806-0000","librispeech_5105-28240-0011","librispeech_5105-28241-0004","librispeech_5105-28241-0019"],"input_specifications":[{"id":"librispeech_4507-16021-0045","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4507-16021-0045.npy","sha256":"364c398f7e47ebcead99b59ac008405f9502702cf89d4940bfeb13f7b50e73fa","sample_rate":16000},{"id":"librispeech_4970-29093-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4970-29093-0004.npy","sha256":"bfe001d3c4144b13a19f4dadaababd475cde8489e0343573d7c93e2ed3b27180","sample_rate":16000},{"id":"librispeech_4970-29093-0005","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4970-29093-0005.npy","sha256":"05fbfba7cad8726c9fa6d847225778b1f0f993516f8e315e0732959dacfa0d1d","sample_rate":16000},{"id":"librispeech_4970-29095-0008","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4970-29095-0008.npy","sha256":"98cb7d4b93856ed5577d455b6af75b342eb49cf42be0c4bd39ba08000ad1229e","sample_rate":16000},{"id":"librispeech_4992-23283-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4992-23283-0016.npy","sha256":"421415266b6d6a08914ccb879fcef3847317f4a6ef5cf11d1a65a92b0b747247","sample_rate":16000},{"id":"librispeech_4992-41797-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4992-41797-0000.npy","sha256":"b07633293b756654df0f8a661b93eaeafa136a1adfd7b75d4b0518561159c264","sample_rate":16000},{"id":"librispeech_4992-41806-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4992-41806-0000.npy","sha256":"31ad5018e5e57c365e8a2ba88f046ad939b81e19a335769d1abb4bd9cc8ca7c3","sample_rate":16000},{"id":"librispeech_5105-28240-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5105-28240-0011.npy","sha256":"3ac3db42bbe992b387196f6c806baaddd2efa79e26f56a4565441675620504a3","sample_rate":16000},{"id":"librispeech_5105-28241-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5105-28241-0004.npy","sha256":"eda03c395cdc9d3f1e9518a18019741a17681b4698ba9987763b4c461fbd7fcb","sample_rate":16000},{"id":"librispeech_5105-28241-0019","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5105-28241-0019.npy","sha256":"9b33cdb5987a1a118cb734b090e58c46dfc7460fe25e94984a7d20e8f55b8af7","sample_rate":16000}],"input_sha256":"1dc957a9036d42c4b53e59d45263e37a8192ca86c71f5edf181034d1d6f56d10","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-005-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-005.log","cold_first_case":"librispeech_4507-16021-0045","returncode":0,"finished_at":"2026-10-13T07:18:27.068043+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.641107256999021,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_4507-16021-0045","librispeech_4970-29093-0004","librispeech_4970-29093-0005","librispeech_4970-29095-0008","librispeech_4992-23283-0016","librispeech_4992-41797-0000","librispeech_4992-41806-0000","librispeech_5105-28240-0011","librispeech_5105-28241-0004","librispeech_5105-28241-0019"]},{"attempt":"hifiglot-006","method":"hifiglot","started_at":"2026-10-13T07:18:27.085684+00:00","input_ids":["librispeech_5142-33396-0019","librispeech_5142-33396-0030","librispeech_5142-33396-0065","librispeech_5639-40744-0013","librispeech_5639-40744-0023","librispeech_5639-40744-0030","librispeech_5683-32865-0017","librispeech_5683-32879-0009","librispeech_5683-32879-0019","librispeech_61-70968-0042"],"input_specifications":[{"id":"librispeech_5142-33396-0019","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5142-33396-0019.npy","sha256":"4114fe2feba025e8426b7975a60af7eba2625ecf55c140d4d942ca0dc0361c57","sample_rate":16000},{"id":"librispeech_5142-33396-0030","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5142-33396-0030.npy","sha256":"bf44bc8f7b5de1f328031c764965206870e3b2a80609be308d2d2082012fb5c3","sample_rate":16000},{"id":"librispeech_5142-33396-0065","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5142-33396-0065.npy","sha256":"fef7806050fadfa47764bf831e8f4678f2d63580b079ce66cfa32ee7dd9c929f","sample_rate":16000},{"id":"librispeech_5639-40744-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5639-40744-0013.npy","sha256":"7ddaea57f9e5ab7533b5a1b89a5446fc88a177a7e3314f61f6d0dd3f1f5f3321","sample_rate":16000},{"id":"librispeech_5639-40744-0023","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5639-40744-0023.npy","sha256":"7c6399830af60d82e6872b227aa2254101ab88790d9e2ca6eee36fdb0cf762a8","sample_rate":16000},{"id":"librispeech_5639-40744-0030","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5639-40744-0030.npy","sha256":"275d5d566a2394fbcd42dc2039695c9ffa82180022ac0926fe455c46d70e2ac4","sample_rate":16000},{"id":"librispeech_5683-32865-0017","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5683-32865-0017.npy","sha256":"6b805ad37580bd662b327382203a4d8dcbce78d1e6ea6d584ccb141d9858eb07","sample_rate":16000},{"id":"librispeech_5683-32879-0009","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5683-32879-0009.npy","sha256":"b74b0dd3e4e57aa921833f8a2e77bec268aa0de7494ce4760b2cad80bb0ffb8c","sample_rate":16000},{"id":"librispeech_5683-32879-0019","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5683-32879-0019.npy","sha256":"ba9e262f1a4e5d33d19cd97573eda08b9fced12e76deaf30d03c322a36da8b82","sample_rate":16000},{"id":"librispeech_61-70968-0042","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0042.npy","sha256":"961c69b03b318b9af140f68a462f11556afbc167c633fa7b9f84f744bc866984","sample_rate":16000}],"input_sha256":"10728f81ea726ebf62d65040c82d79472439ceb21ba79df713dc3b5599b14fa3","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-006-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-006.log","cold_first_case":"librispeech_5142-33396-0019","returncode":0,"finished_at":"2026-10-13T07:22:25.648861+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":2.025881445999403,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_5142-33396-0019","librispeech_5142-33396-0030","librispeech_5142-33396-0065","librispeech_5639-40744-0013","librispeech_5639-40744-0023","librispeech_5639-40744-0030","librispeech_5683-32865-0017","librispeech_5683-32879-0009","librispeech_5683-32879-0019","librispeech_61-70968-0042"]},{"attempt":"hifiglot-007","method":"hifiglot","started_at":"2026-10-13T07:22:25.667038+00:00","input_ids":["librispeech_61-70968-0043","librispeech_61-70968-0054","librispeech_672-122797-0003","librispeech_672-122797-0044","librispeech_672-122797-0056","librispeech_6829-68769-0014","librispeech_6829-68769-0038","librispeech_6829-68769-0041","librispeech_6930-75918-0013","librispeech_6930-76324-0024"],"input_specifications":[{"id":"librispeech_61-70968-0043","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0043.npy","sha256":"11395567141c5a03808579bb9837eae0b644cb0d8d9e75b3474c4e55832698e7","sample_rate":16000},{"id":"librispeech_61-70968-0054","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0054.npy","sha256":"e07ca663384d7b0ea8ac49ce7905c02e33e1c111573b428d6692890acf3f6ecb","sample_rate":16000},{"id":"librispeech_672-122797-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0003.npy","sha256":"cdb8e35cf9303bf26bc3f9cba71892434576732b577ad73f2aca4b6dcec6f2ac","sample_rate":16000},{"id":"librispeech_672-122797-0044","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0044.npy","sha256":"8287d82eeca7b7a411e5e89ca1c7537c5badcb42acb25504ce71209865ab8804","sample_rate":16000},{"id":"librispeech_672-122797-0056","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0056.npy","sha256":"22780bba4503a457a8d49959a4601aca0ceb4e387b94d1e3c1ba07185cf8e7cb","sample_rate":16000},{"id":"librispeech_6829-68769-0014","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0014.npy","sha256":"3e63bfef2739048aec1a1dd1664bc947427270be279a01e2399186b1cc779f6b","sample_rate":16000},{"id":"librispeech_6829-68769-0038","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0038.npy","sha256":"0d340ad2987da76914c8379223b9ebe6ae6f6ad2c3070d2beb52a071272d0c7a","sample_rate":16000},{"id":"librispeech_6829-68769-0041","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0041.npy","sha256":"0d776c98d75836f78dcbf836cd371840458d509810ee9e94758fbd26584e9d1f","sample_rate":16000},{"id":"librispeech_6930-75918-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-75918-0013.npy","sha256":"321a0414ac42f02e98ce930c29201d21e63bf4d633ddb31c75f7bbcc7654bddb","sample_rate":16000},{"id":"librispeech_6930-76324-0024","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-76324-0024.npy","sha256":"22e1e852f27049dc0fb233b38864a80ef568f82169bbc07c6fb7e15c8d9f1003","sample_rate":16000}],"input_sha256":"f8a3e388071bf9727ecff1873a4026e30588327f12de4dc68828bd3fc0abba48","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-007-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-007.log","cold_first_case":"librispeech_61-70968-0043","returncode":0,"finished_at":"2026-10-13T07:26:06.783361+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.5852556970003207,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_61-70968-0043","librispeech_61-70968-0054","librispeech_672-122797-0003","librispeech_672-122797-0044","librispeech_672-122797-0056","librispeech_6829-68769-0014","librispeech_6829-68769-0038","librispeech_6829-68769-0041","librispeech_6930-75918-0013","librispeech_6930-76324-0024"]},{"attempt":"hifiglot-008","method":"hifiglot","started_at":"2026-10-13T07:26:06.801378+00:00","input_ids":["librispeech_6930-81414-0026","librispeech_7021-79740-0002","librispeech_7021-79740-0011","librispeech_7021-85628-0025","librispeech_7127-75946-0002","librispeech_7127-75946-0029","librispeech_7127-75947-0011","librispeech_7176-88083-0002","librispeech_7176-88083-0004","librispeech_7176-88083-0026"],"input_specifications":[{"id":"librispeech_6930-81414-0026","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-81414-0026.npy","sha256":"c88868dff5818474fe738bb4157cb999f7d1139e9923ec83577117e882de4b31","sample_rate":16000},{"id":"librispeech_7021-79740-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-79740-0002.npy","sha256":"6bdd332e5cfad3622e452101f852fa3f8dd6eb5b9be2cdc45f5708ddd2557256","sample_rate":16000},{"id":"librispeech_7021-79740-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-79740-0011.npy","sha256":"15c66be27acb18c9df694af8bc9d44c5bd348f8389a0837a1e4616376dda091f","sample_rate":16000},{"id":"librispeech_7021-85628-0025","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-85628-0025.npy","sha256":"58e5faaa9c6edaa7abbb651251a2000da6e90e3c1943e767c58f215528a210cc","sample_rate":16000},{"id":"librispeech_7127-75946-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75946-0002.npy","sha256":"0981fb6efc50f918e211f5056d645e7f31a403d249c14e3bd928b63982ab99f0","sample_rate":16000},{"id":"librispeech_7127-75946-0029","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75946-0029.npy","sha256":"fb56543512bf946cb8e8680e6e9940fdf458e66eefce85babf56618639407acb","sample_rate":16000},{"id":"librispeech_7127-75947-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75947-0011.npy","sha256":"0a958123a4ddc0a331da3b5e89c6e5b2f141b1c7b4279ba5e6848c7973c3be80","sample_rate":16000},{"id":"librispeech_7176-88083-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0002.npy","sha256":"6ce789c9d4e40824a7597a188ba764b0912cc3d4dceaf7c8ac7aeb28469d4407","sample_rate":16000},{"id":"librispeech_7176-88083-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0004.npy","sha256":"45dd31e9af00a32cf192a87f3fc4d575cf62b88c59a697486955eb85310466db","sample_rate":16000},{"id":"librispeech_7176-88083-0026","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0026.npy","sha256":"d8ac2a315ecbb8b2cd771de10cee21a68a53a6c060b81e6b96a589b954912c21","sample_rate":16000}],"input_sha256":"53915c03d3a84747f4c300887f65e5597df497c7384ed936d74efebe12c1b9f8","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-008-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-008.log","cold_first_case":"librispeech_6930-81414-0026","returncode":0,"finished_at":"2026-10-13T07:29:25.368288+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":2.085220355002093,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_6930-81414-0026","librispeech_7021-79740-0002","librispeech_7021-79740-0011","librispeech_7021-85628-0025","librispeech_7127-75946-0002","librispeech_7127-75946-0029","librispeech_7127-75947-0011","librispeech_7176-88083-0002","librispeech_7176-88083-0004","librispeech_7176-88083-0026"]},{"attempt":"hifiglot-009","method":"hifiglot","started_at":"2026-10-13T07:29:25.385928+00:00","input_ids":["librispeech_7729-102255-0028","librispeech_7729-102255-0038","librispeech_7729-102255-0042","librispeech_8224-274381-0010","librispeech_8224-274381-0013","librispeech_8224-274384-0010","librispeech_8230-279154-0018","librispeech_8230-279154-0032","librispeech_8230-279154-0033","librispeech_8455-210777-0010"],"input_specifications":[{"id":"librispeech_7729-102255-0028","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0028.npy","sha256":"bdefd8b41a55fe29ad275ad0cdc0930d498f0e9c563afc84f3f746edbb5622c1","sample_rate":16000},{"id":"librispeech_7729-102255-0038","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0038.npy","sha256":"9e25b2715b44c695b1208e22e4cec275bbd491d99983858945470dba564f7b1d","sample_rate":16000},{"id":"librispeech_7729-102255-0042","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0042.npy","sha256":"8e0eb42509778a5dd3e14ba6fb8dbb8055bbeca1cb9858080ff50cf3162ecbc6","sample_rate":16000},{"id":"librispeech_8224-274381-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274381-0010.npy","sha256":"47f2be273a089762a811311e58170df296cedada30080030c701877e87b9cb90","sample_rate":16000},{"id":"librispeech_8224-274381-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274381-0013.npy","sha256":"8f62ebb1c8f2288d8c4b16ccb1481b74a43e4039ba77e57303a5b3e57b940d29","sample_rate":16000},{"id":"librispeech_8224-274384-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274384-0010.npy","sha256":"742ea1f463a5d6275ca4dea75eb76e3d478fb9b550110c351f3d7dd4d220e2be","sample_rate":16000},{"id":"librispeech_8230-279154-0018","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0018.npy","sha256":"812b9900ccbe6b84e9444fed1c7770731b7dfca39437dccad215d79db129df01","sample_rate":16000},{"id":"librispeech_8230-279154-0032","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0032.npy","sha256":"13541ca079afe56ea25803a0e8910f0e6fb5f0e24063e867981a992039b868f4","sample_rate":16000},{"id":"librispeech_8230-279154-0033","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0033.npy","sha256":"88f96bf577b1f1b95250b099c6716d41e7e2bed368df2791fdb21adff8013cf5","sample_rate":16000},{"id":"librispeech_8455-210777-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0010.npy","sha256":"153385cc5de0b72725a7f9cb203194181b7d234e40b71aae096375c5bfee23b6","sample_rate":16000}],"input_sha256":"a83c279dcfae53a9c983691a33798638f8fa0a19ebb8042903a2ba39d755de0e","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-009-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-009.log","cold_first_case":"librispeech_7729-102255-0028","returncode":0,"finished_at":"2026-10-13T07:31:22.745249+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.7768203639971034,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_7729-102255-0028","librispeech_7729-102255-0038","librispeech_7729-102255-0042","librispeech_8224-274381-0010","librispeech_8224-274381-0013","librispeech_8224-274384-0010","librispeech_8230-279154-0018","librispeech_8230-279154-0032","librispeech_8230-279154-0033","librispeech_8455-210777-0010"]},{"attempt":"hifiglot-010","method":"hifiglot","started_at":"2026-10-13T07:31:22.764010+00:00","input_ids":["librispeech_8455-210777-0049","librispeech_8455-210777-0060","librispeech_8463-294828-0003","librispeech_8463-294828-0014","librispeech_8463-294828-0022","librispeech_8555-284447-0007","librispeech_8555-284449-0020","librispeech_8555-292519-0001","librispeech_908-157963-0018","librispeech_908-31957-0002"],"input_specifications":[{"id":"librispeech_8455-210777-0049","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0049.npy","sha256":"2d76f8265f7a52a2814eb8e00d8edc511804ac1369740fd998f7cd19435e2fe8","sample_rate":16000},{"id":"librispeech_8455-210777-0060","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0060.npy","sha256":"9b6304a33d078fd527310e09782e4424d1a63e37c4d42866c9aa87d2920ab687","sample_rate":16000},{"id":"librispeech_8463-294828-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0003.npy","sha256":"299a537dc8c17a869309de103cb4a613416f5a2ddbabf2579d3eae9635449db9","sample_rate":16000},{"id":"librispeech_8463-294828-0014","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0014.npy","sha256":"57c0273c8992ab87674c0b415b90f57ef6597f45937ac6ef94c308d8b0bb1aaf","sample_rate":16000},{"id":"librispeech_8463-294828-0022","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0022.npy","sha256":"908baa31161703eb40e83b7043052a4a50ebf6e6dc303d833ee950422de3dd86","sample_rate":16000},{"id":"librispeech_8555-284447-0007","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-284447-0007.npy","sha256":"93bf4a5143f192f1bec4014f5ed1882d65fcf3ffef7d30a7c6262721decfe4b4","sample_rate":16000},{"id":"librispeech_8555-284449-0020","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-284449-0020.npy","sha256":"0ad90c10b2c7d8a7b3190141694c0f9202a9fa41f2cc5c03ef81a8fffa9a0703","sample_rate":16000},{"id":"librispeech_8555-292519-0001","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-292519-0001.npy","sha256":"1d71c61f160ff40a72fef1912367b5ac430b341d3c951361c0a870cdcd7e39fe","sample_rate":16000},{"id":"librispeech_908-157963-0018","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-157963-0018.npy","sha256":"068cbe41c56c5de3cb2928395586f7d207582cd51ef6612040fe2872c397e9dd","sample_rate":16000},{"id":"librispeech_908-31957-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-31957-0002.npy","sha256":"b69e3ce913f862c4a8715e3c55451386e805eade08afccb7a2d3eb6e55696791","sample_rate":16000}],"input_sha256":"888dee89260758d087a5c3871c0277765b7042929ea9a43cd60155dcb249f253","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-010-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-010.log","cold_first_case":"librispeech_8455-210777-0049","returncode":0,"finished_at":"2026-10-13T07:35:18.825754+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":2.0264481650010566,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_8455-210777-0049","librispeech_8455-210777-0060","librispeech_8463-294828-0003","librispeech_8463-294828-0014","librispeech_8463-294828-0022","librispeech_8555-284447-0007","librispeech_8555-284449-0020","librispeech_8555-292519-0001","librispeech_908-157963-0018","librispeech_908-31957-0002"]},{"attempt":"hifiglot-011","method":"hifiglot","started_at":"2026-10-13T07:35:18.849191+00:00","input_ids":["librispeech_908-31957-0016"],"input_specifications":[{"id":"librispeech_908-31957-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-31957-0016.npy","sha256":"e82f7cc1bcf52dc8247694fec36c8608076490306e8962865090852777303b82","sample_rate":16000}],"input_sha256":"27d9d0e7957b81f424033b79cf0545533c0b737c57cd741377e208cb83dbe4d2","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","hifiglot","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/hifiglot-011-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/hifiglot","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-hifiglot-011.log","cold_first_case":"librispeech_908-31957-0016","returncode":0,"finished_at":"2026-10-13T07:36:01.203944+00:00","implementation":{"method":"hifiglot","prepared":{"schema_version":1,"method":"hifiglot","prepared_at":"2026-10-13T04:26:09.468488+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/HiFi-Glot/hifiglot/config.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4","src/hifiglot/configs/baseline/config_feature_map.json":"43187c3e97d1a45be976fa43b7d2fe21630e5a01f1670b0859b2bcd3449f4cbd","src/hifiglot/configs/baseline/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/configs/hifiglot/config_feature_map.json":"d293e790bd06fd217bd24f9a80251cd5a7ad08c209bcf9a3e0076267096d9f87","src/hifiglot/configs/hifiglot/config_hifigan.json":"d8af1f8f4207d3d56a7959ebff54d0645b9557256d518002869bdfeefa691c87","src/hifiglot/dataset/filelist/AIShell.json":"6ef6e4bec7077e063a24854613f23732d405df2b688cf8398b156cb25e6c8910","src/hifiglot/dataset/train.json":"d44f55a104301b14fe8f914ee035f5472d96ba8fd6079b7613fd73c75e7a4f06","src/hifiglot/inference_baseline.py":"438ddf835f2a75001a0d7f154aef966c96a86cc3dd5b8e52902a8a3fc1c04fba","src/hifiglot/inference_hifiglot.py":"fa317b8301d26cc1f5d88f4c72f50d2366a7bf31113549ee935fad18a5940d99","src/hifiglot/src/neural_formant_synthesis/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/dataset.py":"2006f4a492e3af8750b4d2540c3446521c76240e2d8904b79f7c2ed4b896cea1","src/hifiglot/src/neural_formant_synthesis/feature_extraction.py":"1c4a4870802bcd37a8ea0b5d5b58aee8aa46cf3a9e261c0ed954a6acfc603973","src/hifiglot/src/neural_formant_synthesis/functions.py":"2bd16d715b16e606c1d2f682611eba332d485344f4de603a499cf0ebc1bf292e","src/hifiglot/src/neural_formant_synthesis/glotnet/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/activations.py":"bb5a2695d386174f1d97f42dd924f4f034b22ecb915267a24f2f5b26a94524e9","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution.py":"64933a35dbc1ba6abefc2ec55b62615efe703d4df7ecc4ab7556e8bab292772b","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_layer.py":"b0e1d8f20284dcb7a1441725443fff3be7deaee897ee81d79e04ac5311f1443d","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/convolution_stack.py":"01a717165dc3681ac95506a73924c353e1a5d89bb11da286426c5ec5756089af","src/hifiglot/src/neural_formant_synthesis/glotnet/model/feedforward/wavenet.py":"dd094bc413310531fd1f39c36a3f0f363996cbda540f70492a30f3886224aefd","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/allpole.py":"a5ccef9b233b3b00e6eb4ebfbb6056321aab2c5eef20454e9d36e49ee8b5080e","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/biquad.py":"92cc1be986c988b33bff148f87374efe3e251db7c1f05477a7da3c45d430a7f0","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/emphasis.py":"07d217c53c8df9d2c60ccca688265849975c39bd316af8e218903b985e519f62","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/levinson.py":"c9d6955346e72f6f31fd86432a14aa6d4bbc9c52ee21e3df7d35a9fabc8821be","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lfilter.py":"7b5c2a8f7adbb9fbf29b52ff726c7f12523af0ec3ece332529a2f60bd58b994b","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lpc.py":"35dc64c1ec4ff7dd5f475318f0573b4084ea4d99e868363634c374f34e427b05","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/lsf.py":"b8f363001862c59c43726e82c4963adc4a9e6023a9cf3214f87b1d6c43cde923","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/melspec.py":"4d4308c727e78403f162a3c4841ca3c65a3399259a1b302375d81add16206155","src/hifiglot/src/neural_formant_synthesis/glotnet/sigproc/oscillator.py":"98d7908bf1481f5d703629087c706478d9e71bc5beec47cd4adf4aea8ca516be","src/hifiglot/src/neural_formant_synthesis/models.py":"5413d55019c4936e19d71b409cc87dd9a5daf9c563b3903dad6893861681f8c1","src/hifiglot/src/neural_formant_synthesis/sigproc/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/sigproc/levinson.py":"310e4e5b2059affc3e3d10fa71f02e7a3e7796565260e13d4035dc8ec041a349","src/hifiglot/src/neural_formant_synthesis/sigproc/lpc.py":"b3f90c50610ac5ad2ebc0de26c747954776be53002bf177498f92fdf03956e51","src/hifiglot/src/neural_formant_synthesis/third_party/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/env.py":"f5212e8402468ba2b68f79fac906ae48b0762625db368232bbff3bc31fa2b42b","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/models.py":"9d7c786bd53e691c1d3551250108ef7ed5a9a87fb7da4c0854a259d0bdcf1a7f","src/hifiglot/src/neural_formant_synthesis/third_party/hifi_gan/utils.py":"b496698a1c6d9e0fdf7f84c4231746f491ef477631089503da195c3477d7a0a0","src/hifiglot/train_e2e_baseline.py":"98ec299765142efe6f1ce80d62b2ec754acda656a2ef6a83b4d8960f854ae87d","src/hifiglot/train_e2e_hifiglot.py":"af5e4096acf0880028d5bec8bb2df9b37b2eedc13d26ea4ca925810405c81a79"},"sources":{"hifiglot":{"repository":"https://github.com/PupuAI/HiFi-Glot.git","revision":"78951625e8f0a113f401fef7032e98d58cfa7dac"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":1.856291901000077,"details":{"variant":"official HiFi-Glot44.1k inference recipe","parameter_count":20903793,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/HiFi-Glot/hifiglot/epoch-0175_step-1070000/model.safetensors","checkpoint_sha256":"8955e6f43593099b912c25e2e5f9abd2120950e2927b34c5aaebbe3df9eb14c4"}},"successful_ids":["librispeech_908-31957-0016"]}],"log":"logs/recovery-hifiglot-011.log","history":"preparation.json / recovery_execution and recovery/initial-processes.json"},{"method":"golf","returncode":0,"log":"logs/golf-1791867700880177159.log"},{"method":"sifigan","returncode":0,"completed":120,"planned":120,"initial_attempt":{"method":"sifigan","returncode":-9,"log":"logs/sifigan-1791867796517666669.log"},"recovery_attempts":[{"attempt":"sifigan-001","method":"sifigan","started_at":"2026-10-13T07:36:01.283286+00:00","input_ids":["librispeech_3575-170457-0041","librispeech_3575-170457-0054","librispeech_3729-6852-0000","librispeech_3729-6852-0007","librispeech_3729-6852-0017","librispeech_4077-13751-0001","librispeech_4077-13754-0002","librispeech_4077-13754-0012","librispeech_4446-2271-0003","librispeech_4446-2271-0022"],"input_specifications":[{"id":"librispeech_3575-170457-0041","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3575-170457-0041.npy","sha256":"6be0695d1829e2c748580d3856118cd6eabdb573ad8adcd1eb7f3b1824b606fa","sample_rate":16000},{"id":"librispeech_3575-170457-0054","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3575-170457-0054.npy","sha256":"bf86c0c00709c2e5eee207cba7f35d17746c67a2a9f9c9ec302b8a3636061627","sample_rate":16000},{"id":"librispeech_3729-6852-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3729-6852-0000.npy","sha256":"1b783a7c2c383d87de40953142aeba194d79926cf87591d0c55408731d8625ce","sample_rate":16000},{"id":"librispeech_3729-6852-0007","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3729-6852-0007.npy","sha256":"5ab238d03e046aeef6b327b51b75630087f055df4633f0a25c71d45f3af209b7","sample_rate":16000},{"id":"librispeech_3729-6852-0017","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_3729-6852-0017.npy","sha256":"fa4d86d7b8c74df263aaabe4276525584eed7ed36d8b9421d2158a8fad60fd8c","sample_rate":16000},{"id":"librispeech_4077-13751-0001","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4077-13751-0001.npy","sha256":"e251b8c76c67e0e5339a7ad31f65d12566654050606a6b0e7f395fe9bfdd03bf","sample_rate":16000},{"id":"librispeech_4077-13754-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4077-13754-0002.npy","sha256":"5ebee7ab93321e8f9c22488b77a9d83b935967cc2b1755b26ca51d20e4dbcdd6","sample_rate":16000},{"id":"librispeech_4077-13754-0012","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4077-13754-0012.npy","sha256":"da6d5022702fb704c7663cfd01a247727508d36f33ba6cdf57c29ec0bca2540e","sample_rate":16000},{"id":"librispeech_4446-2271-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4446-2271-0003.npy","sha256":"8fd653ce4da1222d377403b051d957dbbd98eca247e0a76171fa28bf4a523f5f","sample_rate":16000},{"id":"librispeech_4446-2271-0022","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4446-2271-0022.npy","sha256":"0e102c1dba2f0649f1df1e8dfcc4595e3cf32e9e0dcf6815b2f748a1174aa58c","sample_rate":16000}],"input_sha256":"0608ab994bf79be1eff4ff980b9a6ec4a1535600ad48680fd94c9554e873e1df","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-001-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-001.log","cold_first_case":"librispeech_3575-170457-0041","returncode":0,"finished_at":"2026-10-13T07:37:22.284884+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.4922552520001773,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_3575-170457-0041","librispeech_3575-170457-0054","librispeech_3729-6852-0000","librispeech_3729-6852-0007","librispeech_3729-6852-0017","librispeech_4077-13751-0001","librispeech_4077-13754-0002","librispeech_4077-13754-0012","librispeech_4446-2271-0003","librispeech_4446-2271-0022"]},{"attempt":"sifigan-002","method":"sifigan","started_at":"2026-10-13T07:37:22.323645+00:00","input_ids":["librispeech_4446-2273-0010","librispeech_4507-16021-0016","librispeech_4507-16021-0032","librispeech_4507-16021-0045","librispeech_4970-29093-0004","librispeech_4970-29093-0005","librispeech_4970-29095-0008","librispeech_4992-23283-0016","librispeech_4992-41797-0000","librispeech_4992-41806-0000"],"input_specifications":[{"id":"librispeech_4446-2273-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4446-2273-0010.npy","sha256":"66fab68f8236f2b9617cfd6928803d223cdcd7756c14f365c225f0ed88f58739","sample_rate":16000},{"id":"librispeech_4507-16021-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4507-16021-0016.npy","sha256":"bf03704b4a0939f52cf691a192ebddece3d996e02ed4c4bccd90a88cb5b78db0","sample_rate":16000},{"id":"librispeech_4507-16021-0032","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4507-16021-0032.npy","sha256":"ef0e7a805e8d327859d4b390159adeb5aea5e9adfcd0a91a422cd585becc0a79","sample_rate":16000},{"id":"librispeech_4507-16021-0045","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4507-16021-0045.npy","sha256":"364c398f7e47ebcead99b59ac008405f9502702cf89d4940bfeb13f7b50e73fa","sample_rate":16000},{"id":"librispeech_4970-29093-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4970-29093-0004.npy","sha256":"bfe001d3c4144b13a19f4dadaababd475cde8489e0343573d7c93e2ed3b27180","sample_rate":16000},{"id":"librispeech_4970-29093-0005","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4970-29093-0005.npy","sha256":"05fbfba7cad8726c9fa6d847225778b1f0f993516f8e315e0732959dacfa0d1d","sample_rate":16000},{"id":"librispeech_4970-29095-0008","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4970-29095-0008.npy","sha256":"98cb7d4b93856ed5577d455b6af75b342eb49cf42be0c4bd39ba08000ad1229e","sample_rate":16000},{"id":"librispeech_4992-23283-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4992-23283-0016.npy","sha256":"421415266b6d6a08914ccb879fcef3847317f4a6ef5cf11d1a65a92b0b747247","sample_rate":16000},{"id":"librispeech_4992-41797-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4992-41797-0000.npy","sha256":"b07633293b756654df0f8a661b93eaeafa136a1adfd7b75d4b0518561159c264","sample_rate":16000},{"id":"librispeech_4992-41806-0000","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_4992-41806-0000.npy","sha256":"31ad5018e5e57c365e8a2ba88f046ad939b81e19a335769d1abb4bd9cc8ca7c3","sample_rate":16000}],"input_sha256":"96f7f6b6ca9be310ad4a0f3c73d7a870d559c6f1ad8209392c669b193c945574","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-002-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-002.log","cold_first_case":"librispeech_4446-2273-0010","returncode":0,"finished_at":"2026-10-13T07:39:15.926612+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.2018654700004845,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_4446-2273-0010","librispeech_4507-16021-0016","librispeech_4507-16021-0032","librispeech_4507-16021-0045","librispeech_4970-29093-0004","librispeech_4970-29093-0005","librispeech_4970-29095-0008","librispeech_4992-23283-0016","librispeech_4992-41797-0000","librispeech_4992-41806-0000"]},{"attempt":"sifigan-003","method":"sifigan","started_at":"2026-10-13T07:39:15.959440+00:00","input_ids":["librispeech_5105-28240-0011","librispeech_5105-28241-0004","librispeech_5105-28241-0019","librispeech_5142-33396-0019","librispeech_5142-33396-0030","librispeech_5142-33396-0065","librispeech_5639-40744-0013","librispeech_5639-40744-0023","librispeech_5639-40744-0030","librispeech_5683-32865-0017"],"input_specifications":[{"id":"librispeech_5105-28240-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5105-28240-0011.npy","sha256":"3ac3db42bbe992b387196f6c806baaddd2efa79e26f56a4565441675620504a3","sample_rate":16000},{"id":"librispeech_5105-28241-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5105-28241-0004.npy","sha256":"eda03c395cdc9d3f1e9518a18019741a17681b4698ba9987763b4c461fbd7fcb","sample_rate":16000},{"id":"librispeech_5105-28241-0019","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5105-28241-0019.npy","sha256":"9b33cdb5987a1a118cb734b090e58c46dfc7460fe25e94984a7d20e8f55b8af7","sample_rate":16000},{"id":"librispeech_5142-33396-0019","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5142-33396-0019.npy","sha256":"4114fe2feba025e8426b7975a60af7eba2625ecf55c140d4d942ca0dc0361c57","sample_rate":16000},{"id":"librispeech_5142-33396-0030","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5142-33396-0030.npy","sha256":"bf44bc8f7b5de1f328031c764965206870e3b2a80609be308d2d2082012fb5c3","sample_rate":16000},{"id":"librispeech_5142-33396-0065","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5142-33396-0065.npy","sha256":"fef7806050fadfa47764bf831e8f4678f2d63580b079ce66cfa32ee7dd9c929f","sample_rate":16000},{"id":"librispeech_5639-40744-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5639-40744-0013.npy","sha256":"7ddaea57f9e5ab7533b5a1b89a5446fc88a177a7e3314f61f6d0dd3f1f5f3321","sample_rate":16000},{"id":"librispeech_5639-40744-0023","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5639-40744-0023.npy","sha256":"7c6399830af60d82e6872b227aa2254101ab88790d9e2ca6eee36fdb0cf762a8","sample_rate":16000},{"id":"librispeech_5639-40744-0030","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5639-40744-0030.npy","sha256":"275d5d566a2394fbcd42dc2039695c9ffa82180022ac0926fe455c46d70e2ac4","sample_rate":16000},{"id":"librispeech_5683-32865-0017","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5683-32865-0017.npy","sha256":"6b805ad37580bd662b327382203a4d8dcbce78d1e6ea6d584ccb141d9858eb07","sample_rate":16000}],"input_sha256":"c1b3c8bdc56d5001de3b4beb503c429e206539653df434c93edfe69640170198","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-003-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-003.log","cold_first_case":"librispeech_5105-28240-0011","returncode":0,"finished_at":"2026-10-13T07:41:16.336885+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.4525942699983716,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_5105-28240-0011","librispeech_5105-28241-0004","librispeech_5105-28241-0019","librispeech_5142-33396-0019","librispeech_5142-33396-0030","librispeech_5142-33396-0065","librispeech_5639-40744-0013","librispeech_5639-40744-0023","librispeech_5639-40744-0030","librispeech_5683-32865-0017"]},{"attempt":"sifigan-004","method":"sifigan","started_at":"2026-10-13T07:41:16.367522+00:00","input_ids":["librispeech_5683-32879-0009","librispeech_5683-32879-0019","librispeech_61-70968-0042","librispeech_61-70968-0043","librispeech_61-70968-0054","librispeech_672-122797-0003","librispeech_672-122797-0044","librispeech_672-122797-0056","librispeech_6829-68769-0014","librispeech_6829-68769-0038"],"input_specifications":[{"id":"librispeech_5683-32879-0009","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5683-32879-0009.npy","sha256":"b74b0dd3e4e57aa921833f8a2e77bec268aa0de7494ce4760b2cad80bb0ffb8c","sample_rate":16000},{"id":"librispeech_5683-32879-0019","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_5683-32879-0019.npy","sha256":"ba9e262f1a4e5d33d19cd97573eda08b9fced12e76deaf30d03c322a36da8b82","sample_rate":16000},{"id":"librispeech_61-70968-0042","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0042.npy","sha256":"961c69b03b318b9af140f68a462f11556afbc167c633fa7b9f84f744bc866984","sample_rate":16000},{"id":"librispeech_61-70968-0043","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0043.npy","sha256":"11395567141c5a03808579bb9837eae0b644cb0d8d9e75b3474c4e55832698e7","sample_rate":16000},{"id":"librispeech_61-70968-0054","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_61-70968-0054.npy","sha256":"e07ca663384d7b0ea8ac49ce7905c02e33e1c111573b428d6692890acf3f6ecb","sample_rate":16000},{"id":"librispeech_672-122797-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0003.npy","sha256":"cdb8e35cf9303bf26bc3f9cba71892434576732b577ad73f2aca4b6dcec6f2ac","sample_rate":16000},{"id":"librispeech_672-122797-0044","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0044.npy","sha256":"8287d82eeca7b7a411e5e89ca1c7537c5badcb42acb25504ce71209865ab8804","sample_rate":16000},{"id":"librispeech_672-122797-0056","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_672-122797-0056.npy","sha256":"22780bba4503a457a8d49959a4601aca0ceb4e387b94d1e3c1ba07185cf8e7cb","sample_rate":16000},{"id":"librispeech_6829-68769-0014","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0014.npy","sha256":"3e63bfef2739048aec1a1dd1664bc947427270be279a01e2399186b1cc779f6b","sample_rate":16000},{"id":"librispeech_6829-68769-0038","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0038.npy","sha256":"0d340ad2987da76914c8379223b9ebe6ae6f6ad2c3070d2beb52a071272d0c7a","sample_rate":16000}],"input_sha256":"f30966ba6daf1b9abbb3b4f10f55639557115a35deb9d4c6fa5ccaf145b7d253","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-004-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-004.log","cold_first_case":"librispeech_5683-32879-0009","returncode":0,"finished_at":"2026-10-13T07:43:07.304193+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.232147143000475,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_5683-32879-0009","librispeech_5683-32879-0019","librispeech_61-70968-0042","librispeech_61-70968-0043","librispeech_61-70968-0054","librispeech_672-122797-0003","librispeech_672-122797-0044","librispeech_672-122797-0056","librispeech_6829-68769-0014","librispeech_6829-68769-0038"]},{"attempt":"sifigan-005","method":"sifigan","started_at":"2026-10-13T07:43:07.340606+00:00","input_ids":["librispeech_6829-68769-0041","librispeech_6930-75918-0013","librispeech_6930-76324-0024","librispeech_6930-81414-0026","librispeech_7021-79740-0002","librispeech_7021-79740-0011","librispeech_7021-85628-0025","librispeech_7127-75946-0002","librispeech_7127-75946-0029","librispeech_7127-75947-0011"],"input_specifications":[{"id":"librispeech_6829-68769-0041","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6829-68769-0041.npy","sha256":"0d776c98d75836f78dcbf836cd371840458d509810ee9e94758fbd26584e9d1f","sample_rate":16000},{"id":"librispeech_6930-75918-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-75918-0013.npy","sha256":"321a0414ac42f02e98ce930c29201d21e63bf4d633ddb31c75f7bbcc7654bddb","sample_rate":16000},{"id":"librispeech_6930-76324-0024","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-76324-0024.npy","sha256":"22e1e852f27049dc0fb233b38864a80ef568f82169bbc07c6fb7e15c8d9f1003","sample_rate":16000},{"id":"librispeech_6930-81414-0026","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_6930-81414-0026.npy","sha256":"c88868dff5818474fe738bb4157cb999f7d1139e9923ec83577117e882de4b31","sample_rate":16000},{"id":"librispeech_7021-79740-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-79740-0002.npy","sha256":"6bdd332e5cfad3622e452101f852fa3f8dd6eb5b9be2cdc45f5708ddd2557256","sample_rate":16000},{"id":"librispeech_7021-79740-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-79740-0011.npy","sha256":"15c66be27acb18c9df694af8bc9d44c5bd348f8389a0837a1e4616376dda091f","sample_rate":16000},{"id":"librispeech_7021-85628-0025","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7021-85628-0025.npy","sha256":"58e5faaa9c6edaa7abbb651251a2000da6e90e3c1943e767c58f215528a210cc","sample_rate":16000},{"id":"librispeech_7127-75946-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75946-0002.npy","sha256":"0981fb6efc50f918e211f5056d645e7f31a403d249c14e3bd928b63982ab99f0","sample_rate":16000},{"id":"librispeech_7127-75946-0029","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75946-0029.npy","sha256":"fb56543512bf946cb8e8680e6e9940fdf458e66eefce85babf56618639407acb","sample_rate":16000},{"id":"librispeech_7127-75947-0011","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7127-75947-0011.npy","sha256":"0a958123a4ddc0a331da3b5e89c6e5b2f141b1c7b4279ba5e6848c7973c3be80","sample_rate":16000}],"input_sha256":"84fda5e4b8cd19101bba36382c94cf94ee4594361132e532037d2757b7df0168","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-005-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-005.log","cold_first_case":"librispeech_6829-68769-0041","returncode":0,"finished_at":"2026-10-13T07:44:40.392117+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.3345166690014594,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_6829-68769-0041","librispeech_6930-75918-0013","librispeech_6930-76324-0024","librispeech_6930-81414-0026","librispeech_7021-79740-0002","librispeech_7021-79740-0011","librispeech_7021-85628-0025","librispeech_7127-75946-0002","librispeech_7127-75946-0029","librispeech_7127-75947-0011"]},{"attempt":"sifigan-006","method":"sifigan","started_at":"2026-10-13T07:44:40.434329+00:00","input_ids":["librispeech_7176-88083-0002","librispeech_7176-88083-0004","librispeech_7176-88083-0026","librispeech_7729-102255-0028","librispeech_7729-102255-0038","librispeech_7729-102255-0042","librispeech_8224-274381-0010","librispeech_8224-274381-0013","librispeech_8224-274384-0010","librispeech_8230-279154-0018"],"input_specifications":[{"id":"librispeech_7176-88083-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0002.npy","sha256":"6ce789c9d4e40824a7597a188ba764b0912cc3d4dceaf7c8ac7aeb28469d4407","sample_rate":16000},{"id":"librispeech_7176-88083-0004","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0004.npy","sha256":"45dd31e9af00a32cf192a87f3fc4d575cf62b88c59a697486955eb85310466db","sample_rate":16000},{"id":"librispeech_7176-88083-0026","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7176-88083-0026.npy","sha256":"d8ac2a315ecbb8b2cd771de10cee21a68a53a6c060b81e6b96a589b954912c21","sample_rate":16000},{"id":"librispeech_7729-102255-0028","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0028.npy","sha256":"bdefd8b41a55fe29ad275ad0cdc0930d498f0e9c563afc84f3f746edbb5622c1","sample_rate":16000},{"id":"librispeech_7729-102255-0038","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0038.npy","sha256":"9e25b2715b44c695b1208e22e4cec275bbd491d99983858945470dba564f7b1d","sample_rate":16000},{"id":"librispeech_7729-102255-0042","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_7729-102255-0042.npy","sha256":"8e0eb42509778a5dd3e14ba6fb8dbb8055bbeca1cb9858080ff50cf3162ecbc6","sample_rate":16000},{"id":"librispeech_8224-274381-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274381-0010.npy","sha256":"47f2be273a089762a811311e58170df296cedada30080030c701877e87b9cb90","sample_rate":16000},{"id":"librispeech_8224-274381-0013","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274381-0013.npy","sha256":"8f62ebb1c8f2288d8c4b16ccb1481b74a43e4039ba77e57303a5b3e57b940d29","sample_rate":16000},{"id":"librispeech_8224-274384-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8224-274384-0010.npy","sha256":"742ea1f463a5d6275ca4dea75eb76e3d478fb9b550110c351f3d7dd4d220e2be","sample_rate":16000},{"id":"librispeech_8230-279154-0018","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0018.npy","sha256":"812b9900ccbe6b84e9444fed1c7770731b7dfca39437dccad215d79db129df01","sample_rate":16000}],"input_sha256":"b58f0558ba7bbf9afed1cc7cfac406aabf907dd608a4f4dbbf1678080961b1f8","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-006-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-006.log","cold_first_case":"librispeech_7176-88083-0002","returncode":0,"finished_at":"2026-10-13T07:46:02.571939+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.2248152289976133,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_7176-88083-0002","librispeech_7176-88083-0004","librispeech_7176-88083-0026","librispeech_7729-102255-0028","librispeech_7729-102255-0038","librispeech_7729-102255-0042","librispeech_8224-274381-0010","librispeech_8224-274381-0013","librispeech_8224-274384-0010","librispeech_8230-279154-0018"]},{"attempt":"sifigan-007","method":"sifigan","started_at":"2026-10-13T07:46:02.581863+00:00","input_ids":["librispeech_8230-279154-0032","librispeech_8230-279154-0033","librispeech_8455-210777-0010","librispeech_8455-210777-0049","librispeech_8455-210777-0060","librispeech_8463-294828-0003","librispeech_8463-294828-0014","librispeech_8463-294828-0022","librispeech_8555-284447-0007","librispeech_8555-284449-0020"],"input_specifications":[{"id":"librispeech_8230-279154-0032","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0032.npy","sha256":"13541ca079afe56ea25803a0e8910f0e6fb5f0e24063e867981a992039b868f4","sample_rate":16000},{"id":"librispeech_8230-279154-0033","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8230-279154-0033.npy","sha256":"88f96bf577b1f1b95250b099c6716d41e7e2bed368df2791fdb21adff8013cf5","sample_rate":16000},{"id":"librispeech_8455-210777-0010","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0010.npy","sha256":"153385cc5de0b72725a7f9cb203194181b7d234e40b71aae096375c5bfee23b6","sample_rate":16000},{"id":"librispeech_8455-210777-0049","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0049.npy","sha256":"2d76f8265f7a52a2814eb8e00d8edc511804ac1369740fd998f7cd19435e2fe8","sample_rate":16000},{"id":"librispeech_8455-210777-0060","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8455-210777-0060.npy","sha256":"9b6304a33d078fd527310e09782e4424d1a63e37c4d42866c9aa87d2920ab687","sample_rate":16000},{"id":"librispeech_8463-294828-0003","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0003.npy","sha256":"299a537dc8c17a869309de103cb4a613416f5a2ddbabf2579d3eae9635449db9","sample_rate":16000},{"id":"librispeech_8463-294828-0014","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0014.npy","sha256":"57c0273c8992ab87674c0b415b90f57ef6597f45937ac6ef94c308d8b0bb1aaf","sample_rate":16000},{"id":"librispeech_8463-294828-0022","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8463-294828-0022.npy","sha256":"908baa31161703eb40e83b7043052a4a50ebf6e6dc303d833ee950422de3dd86","sample_rate":16000},{"id":"librispeech_8555-284447-0007","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-284447-0007.npy","sha256":"93bf4a5143f192f1bec4014f5ed1882d65fcf3ffef7d30a7c6262721decfe4b4","sample_rate":16000},{"id":"librispeech_8555-284449-0020","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-284449-0020.npy","sha256":"0ad90c10b2c7d8a7b3190141694c0f9202a9fa41f2cc5c03ef81a8fffa9a0703","sample_rate":16000}],"input_sha256":"1377b0d9306e1231ace8f9403f818194328207773b67a892526940888e76f515","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-007-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-007.log","cold_first_case":"librispeech_8230-279154-0032","returncode":0,"finished_at":"2026-10-13T07:48:16.054771+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.2285226480016718,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_8230-279154-0032","librispeech_8230-279154-0033","librispeech_8455-210777-0010","librispeech_8455-210777-0049","librispeech_8455-210777-0060","librispeech_8463-294828-0003","librispeech_8463-294828-0014","librispeech_8463-294828-0022","librispeech_8555-284447-0007","librispeech_8555-284449-0020"]},{"attempt":"sifigan-008","method":"sifigan","started_at":"2026-10-13T07:48:16.089367+00:00","input_ids":["librispeech_8555-292519-0001","librispeech_908-157963-0018","librispeech_908-31957-0002","librispeech_908-31957-0016"],"input_specifications":[{"id":"librispeech_8555-292519-0001","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_8555-292519-0001.npy","sha256":"1d71c61f160ff40a72fef1912367b5ac430b341d3c951361c0a870cdcd7e39fe","sample_rate":16000},{"id":"librispeech_908-157963-0018","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-157963-0018.npy","sha256":"068cbe41c56c5de3cb2928395586f7d207582cd51ef6612040fe2872c397e9dd","sample_rate":16000},{"id":"librispeech_908-31957-0002","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-31957-0002.npy","sha256":"b69e3ce913f862c4a8715e3c55451386e805eade08afccb7a2d3eb6e55696791","sample_rate":16000},{"id":"librispeech_908-31957-0016","path":"/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/inputs/librispeech_908-31957-0016.npy","sha256":"e82f7cc1bcf52dc8247694fec36c8608076490306e8962865090852777303b82","sample_rate":16000}],"input_sha256":"4d1f7b4de95c46ff18fe67c2167f635250494185db3d2200ef24177659217f62","command":["/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/env/bin/python","-m","vocoderbench.references.worker","--method","sifigan","--runtime-cache","/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references","--inputs","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/recovery/sifigan-008-inputs.json","--output","/workspace/scratch/71b268329986/formal-vocoder-report/runs/formal-hnm-v3-reference-120/native/sifigan","--threads","1","--seed","0","--retry-failed"],"log":"logs/recovery-sifigan-008.log","cold_first_case":"librispeech_8555-292519-0001","returncode":0,"finished_at":"2026-10-13T07:49:07.952840+00:00","implementation":{"method":"sifigan","prepared":{"schema_version":1,"method":"sifigan","prepared_at":"2026-10-13T04:27:37.329020+00:00","environment":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","platform":"Linux-6.18.44-x86_64-with-glibc2.41","dependencies":"absl-py==2.5.1\naiohappyeyeballs==2.7.1\naiohttp==3.14.5\naiosignal==1.4.0\nantlr4-python3-runtime==4.9.3\napprise==2.0.1\nargbind==0.3.9\nasttokens==3.0.2\nasync-timeout==5.0.1\nattrs==26.1.0\naudioread==3.1.0\nbeautifulsoup4==4.15.0\ncertifi==2026.7.22\ncffi==2.1.1\ncharset-normalizer==3.5.2\nclick==8.5.0\ncloudpickle==3.1.2\ncontourpy==1.3.2\ncycler==0.12.1\ncython==3.3.0\ndecorator==5.3.1\ndescript-audiotools==0.7.2\ndiffsptk==4.0.1\ndocstring-parser==0.18.0\neinops==0.8.2\neinx==0.4.3\nexceptiongroup==1.3.1\nexecuting==2.3.0\nffmpy==1.0.0\nfilelock==3.32.3\nfire==0.7.1\nflatten-dict==0.5.0\nfonttools==4.65.0\nfrozendict==2.4.7\nfrozenlist==1.8.0\nfsspec==2024.12.0\nfuture==1.0.0\ngdown==6.4.2\ngrpcio==1.84.0\nh5py==3.16.0\nhf-xet==1.7.0\nhuggingface-hub==0.36.2\nhydra-core==1.3.7\nidna==3.20\nimportlib-resources==7.1.0\nipython==8.39.0\njedi==0.20.0\njinja2==3.1.6\njoblib==1.6.0\njsonargparse==4.52.0\njulius==0.2.8\nkazane==1.0.4\nkiwisolver==1.5.1\nlazy-loader==0.6\nlibrosa==0.10.2.post1\nlightning==2.1.3\nlightning-utilities==0.15.3\nllvmlite==0.50.0\nloguru==0.7.3\nmarkdown==3.10.3\nmarkdown-it-py==4.2.0\nmarkdown2==2.5.5\nmarkupsafe==3.0.3\nmatplotlib==3.10.9\nmatplotlib-inline==0.2.2\nmdurl==0.1.2\nmpmath==1.3.0\nmsgpack==1.2.3\nmultidict==7.1.0\nnetworkx==3.4.2\nnnaudio==0.3.4\nnumba==0.68.0\nnumpy==1.26.4\noauthlib==4.0.0\nomegaconf==2.3.1\npackaging==24.2\nparso==0.8.7\npenn==0.0.14\npexpect==4.9.0\npillow==12.3.0\nplatformdirs==4.12.4\npooch==1.9.0\npraat-parselmouth==0.4.7\nprompt-toolkit==3.0.53\npropcache==0.5.4\nprotobuf==3.19.6\npsutil==7.2.2\nptyprocess==0.7.0\npure-eval==0.2.4\npycparser==3.11\npygments==2.21.0\npyloudnorm==0.2.0\npylstraight==0.1.3\npyparsing==3.3.3\npysocks==1.7.1\npysptk==1.0.1\npystoi==0.4.1\npython-dateutil==2.9.0.post0\npytorch-lightning==2.6.6\npyworld==0.3.5\npyyaml==6.0.3\nrandomname==0.2.1\nrequests==2.34.2\nrequests-oauthlib==2.0.0\nrich==15.0.0\nsafetensors==0.8.0\nscikit-learn==1.7.2\nscipy==1.13.1\nsetuptools==80.10.2\nsix==1.17.0\nsoundfile==0.14.0\nsoupsieve==2.10\nsoxr==1.1.0\nstack-data==0.6.3\nsympy==1.13.1\ntensorboard==2.20.0\ntensorboard-data-server==0.7.2\ntensorboardx==2.6\ntermcolor==3.3.0\nthreadpoolctl==3.7.0\ntorch==2.5.1+cpu\ntorch-einops-utils==0.1.36\ntorch-fftconv==1.0.0\ntorch-stoi==0.2.3\ntorchaudio==2.5.1+cpu\ntorchcomp==0.2.1\ntorchlpc==0.6\ntorchmetrics==1.9.0\ntorchutil==0.1.2\ntqdm==4.70.1\ntraitlets==5.16.1\ntyping-extensions==4.16.0\nurllib3==2.8.0\nvector-quantize-pytorch==1.31.6\nwcwidth==0.9.2\nwerkzeug==3.1.9\nyapecs==0.5.0\nyarl==1.25.1\n"},"lock_sha256":"a9545775f1381f6d56e390286384993ab8f1b6c3b3010058d5fc54bccef494c1","requirements_sha256":"52fe25d537c36ec24ca7ba6a00e6856cfb540c3d776d0af928ac03ecd06042d0","files":{"models/sifigan.tar.gz":"5f9ceaafd8d2fd14a683b6cf2e208e8da149fc59d17e34d4027ba02573eeedfe","models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","src/sifigan/setup.py":"955660e15fab96d91c0c195ecec4d7da25790dd3ff9e42fe9480935db9d0a1f6","src/sifigan/sifigan/__init__.py":"f32c41fc1efa7937f87ee36aeedcec4fbb8aeb4b0f1dca5e33d1321a396a6cc5","src/sifigan/sifigan/bin/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/anasyn.py":"3d12a912478076e73bd558ec4b28da4cdd7e27be4cd514b511d67ac28b88d490","src/sifigan/sifigan/bin/compute_statistics.py":"ffb76f39a1930dd3222d9ef6758f36ee914fafed0b8d2a929f1204e9ab9ace39","src/sifigan/sifigan/bin/config/__init__.py":"e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855","src/sifigan/sifigan/bin/config/anasyn.yaml":"0cae3f3f1258392019b08c0fcad782eaad45ed4c1d2f4b3689906b7f3bc655eb","src/sifigan/sifigan/bin/config/compute_statistics.yaml":"c6d2c147339f08984c226d05760748d08054e45194056bc08799b2c08027fe2b","src/sifigan/sifigan/bin/config/data/libritts-r-clean+nus-48e.yaml":"2a21098aabf5ec7c217d1aede1dbd594095babc6c1bb0684828b352406cebd3e","src/sifigan/sifigan/bin/config/data/namine_ritsu.yaml":"333443ecbea006f6687123474ea5107264b2ac9815528b12747fab67835ba2db","src/sifigan/sifigan/bin/config/decode.yaml":"904304f93df7e35de6c1bcd73c0aa088610c9ee14b4c875ed5864e4a6aa90220","src/sifigan/sifigan/bin/config/discriminator/hifigan.yaml":"7e2b0892c5329ad2f578d0a2286099176003cd9fe05d5de9a09cc1449b6539db","src/sifigan/sifigan/bin/config/discriminator/univnet.yaml":"aee026484dc245ffa67ecdac9516125549bebb4c5784db284d5a257788760b9e","src/sifigan/sifigan/bin/config/extract_features.yaml":"fefe173d53b0740b0f9f2925adf9b7b17ced1d11dce72451e0eb888ce1670056","src/sifigan/sifigan/bin/config/generator/hifigan.yaml":"18956fc64c7d9691439038708bdd1879a50e2c15d9a535e22eff893f7482e576","src/sifigan/sifigan/bin/config/generator/sifigan.direct.yaml":"a85a2b28cc899370f89185a9833f073081a9aa61a6edb8e27fcb8eaabbbeb803","src/sifigan/sifigan/bin/config/generator/sifigan.yaml":"16219976449841293da7fa90f2623a47b3d1de8aae0487b15fc2d500e21891b9","src/sifigan/sifigan/bin/config/param_count.yaml":"65c71910367e44f436adc4282f25b453f090e6fe5bce236187b4e1e510a30e16","src/sifigan/sifigan/bin/config/train/hifigan.yaml":"05aeb83f4595c1acedd6d42c73d349807808bcc80641f270274f8122470e7356","src/sifigan/sifigan/bin/config/train/sifigan.yaml":"357f35d8131b9260f328f8670a5bf0ab71d7ac2bb9499067b836f34cbffd7253","src/sifigan/sifigan/bin/config/train/sifigan_1000k.yaml":"a7e0aab88afdc9aed56da82cbc5c2446988326291e39b24a246d32b0a02b3e0f","src/sifigan/sifigan/bin/config/train.yaml":"0b1fd45eb554954c42b46189cf9cee8ab120d717575f2baa10ac418a331f9aa6","src/sifigan/sifigan/bin/decode.py":"75b3f5bdf31d3d294d460cacf8cfa289e1638c7a8c84b23cbedba18203ab53ab","src/sifigan/sifigan/bin/extract_features.py":"d93d46084f5ccdeeb7154b865e763440ec4683a6d88fab6362f4ead3705a553a","src/sifigan/sifigan/bin/param_count.py":"40c28ae974a9359f244c638cf50b7d063eb7bc579fb64abed1c6c3107dc20274","src/sifigan/sifigan/bin/train.py":"e3340565559c6e00343a4cfb5f966435c3b4a86bf1ebb7fc3b800c8068445e7a","src/sifigan/sifigan/datasets/__init__.py":"f3ecd1bd5729339ee21e2058351f71fd6ab2963bf0d22122381b5a5b03dcd4be","src/sifigan/sifigan/datasets/audio_feat_dataset.py":"efe5a8ce9f41ea4a7b2320cf77e41fc5262f658cda866aed5fde2a1d268762b0","src/sifigan/sifigan/layers/__init__.py":"5ce2179c6c21bff9c91f22e5d4d1a0d625b7112208b08fc29fce810805dc0959","src/sifigan/sifigan/layers/cheaptrick.py":"c0e6655c0530146147eb516267d7e163212c9dc2841fcca19e5222c875552d86","src/sifigan/sifigan/layers/residual_block.py":"d52cc8d6a434e49e0306e7da578de9e4d1dd390962250665a100337dfa27ded9","src/sifigan/sifigan/layers/snake.py":"b3ef5f036e96d50813fe72d459d829516dd394d8ad1fa01562ac8d0d6f7eb563","src/sifigan/sifigan/losses/__init__.py":"54ccd9347f5f1ab1c67b3cf44b2ee6087883d7acc86af919efedf83bd16e3cd8","src/sifigan/sifigan/losses/adv.py":"79269d3b8c438ea8ff29dcfe5ee1877dcbd68f67de59d107a1e07b26b7553488","src/sifigan/sifigan/losses/mel.py":"20a73ebde9716386e61ae5ca9e6552b70985113129332475e3092ee9b8cff6ac","src/sifigan/sifigan/losses/reg.py":"6c14086c8dd74a03f12f0452d2eac0a62557c38e0fa9efb763afe8ad6a2fc2ad","src/sifigan/sifigan/models/__init__.py":"e495ee47a097d992bd332b85d70f84f6adfc96d402e278c158365567798ea49a","src/sifigan/sifigan/models/discriminator.py":"7e8caa445914a8ba91d7122eee8d8d91d0d6cec041004af01e1f7345ad0afb1d","src/sifigan/sifigan/models/generator.py":"9d7b361049587d26c21f1101278892536efe7b36503ba31e7705490a507070ba","src/sifigan/sifigan/utils/__init__.py":"09b9e8bc81ab3565c80b755892e300b786e9a14712159378a809126eaba4bc81","src/sifigan/sifigan/utils/features.py":"cf8a7d0a56e54fc073fba6698c11b6ee2e35114797c2a99fca0b149c0199a609","src/sifigan/sifigan/utils/index.py":"be3134981743d4738b4ffc42956df0040d5dbe5d66f5ce323ff442a5119ecf37","src/sifigan/sifigan/utils/utils.py":"eef463c90a5b3a45ac8e1199abe2295339dbf991ef9d67ee3a48c53f6e47125f","models/model+stats/libritts_r_clean+nus-48e_train_no_dev.joblib":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708"},"sources":{"sifigan":{"repository":"https://github.com/chomeyama/SiFiGAN.git","revision":"cdb20daeb4da4a126077819091706ee90522c7a8"}},"note":"上流コードと重みをキャッシュへ取得。第三者コードを本リポジトリに複製しない。"},"runtime":{"python":"3.10.22 (main, Oct  3 2026, 01:32:06) [Clang 22.1.3 ]","torch":"2.5.1+cpu","threads":1,"cpu_count":9,"processor":"","device":"cpu"},"initialization_s":3.4441552450007293,"details":{"variant":"published24k multispeaker speech+singing; WORLD Harvest included","parameter_count":11291618,"checkpoint":"/workspace/scratch/71b268329986/formal-vocoder-report/.cache/references/models/model+stats/sifigan_libritts-r-clean+nus-48e_checkpoint-1000000steps.pkl","checkpoint_sha256":"2d673e3cee10b4e88dd69b344443b38e0ef0d48f15f1ee5b417c66c0544e5df6","stats_sha256":"c2ae18f7be617c3d11fb1f5ceff6f1fdb57dc35bc9737bee51416c0886dc4708","config":{"hydra":{"run":{"dir":"./"},"output_subdir":null,"job_logging":{"formatters":{"simple":{"format":"[%(asctime)s][%(levelname)s][%(module)s | %(lineno)s] %(message)s"}},"disable_existing_loggers":false}},"in_dir":null,"out_dir":null,"stats":null,"checkpoint_path":null,"f0_factors":[1.0],"seed":100,"sample_rate":24000,"frame_period":5,"f0_floor":60,"f0_ceil":1000,"mcep_dim":39,"mcap_dim":19,"aux_feats":["mcep","bap"],"dense_factors":[0.5,1,4,8],"df_f0_type":"cf0","sine_amp":0.1,"noise_amp":0.003,"sine_f0_type":"cf0","signal_types":["sine"],"generator":{"_target_":"sifigan.models.SiFiGANGenerator","in_channels":43,"out_channels":1,"channels":512,"kernel_size":7,"upsample_scales":[5,4,3,2],"upsample_kernel_sizes":[10,8,6,4],"source_network_params":{"resblock_kernel_size":3,"resblock_dilations":[[1],[1,2],[1,2,4],[1,2,4,8]],"use_additional_convs":true},"filter_network_params":{"resblock_kernel_sizes":[3,5,7],"resblock_dilations":[[1,3,5],[1,3,5],[1,3,5]],"use_additional_convs":false},"share_upsamples":false,"share_downsamples":false,"bias":true,"nonlinear_activation":"LeakyReLU","nonlinear_activation_params":{"negative_slope":0.1},"use_weight_norm":true}}}},"successful_ids":["librispeech_8555-292519-0001","librispeech_908-157963-0018","librispeech_908-31957-0002","librispeech_908-31957-0016"]}],"log":"logs/recovery-sifigan-008.log","history":"preparation.json / recovery_execution and recovery/initial-processes.json"},{"method":"llsm2","returncode":0,"log":"logs/llsm2-1791868189294193768.log"},{"method":"magphase","returncode":0,"log":"logs/magphase-1791868305481874296.log"}],"failure_count":0,"process_failure_count":0,"operation_evaluation":{"status":"not_implemented","reason":"本PRは再合成比較の追加。属性操作・下流学習は未評価"},"mos_provenance":{"name":"UTMOSv2","quantity":"predicted_mos_not_human_rating","source_repository":"https://github.com/sarulab-speech/UTMOSv2","source_commit":"cc2700db57bb83ee13dc31ebe1b868c254e15d09","checkpoint_revision":"506474f2b33dc77c234d668cc419be1861899cad","checkpoint_sha256":"c8149d988e4bbf3f347e6966b5d769de347a5f8c59ffca1dc4bd4bf5b8585e57","fold":0,"device":"cpu","configuration":"fusion_stage3","predict_dataset":"bvcc","repetitions":1,"extraction_seed":20261012,"cpu_threads":4,"remove_silent_section":false,"nnpack_enabled":false,"internal_sample_rate":16000,"ssl_crop_seconds":3.0,"spectrogram_crop_seconds":1.4,"spectrogram_frames_per_repetition":2,"spectrogram_inner_mixup":true,"short_input_extension":"tile","code_license":"MIT","weights_license":"MIT","independent_human_rating_calibration":"not_performed_for_this_project","scope":"JVS/LibriSpeechの再合成への適用は未校正。言語・発声別に集計する。","paper_protocol_difference":"論文の5fold×5抽出平均ではなく、fold0と上記反復数。","initialization":"official_architecture_then_complete_strict_state_dict","packages":{"torch":"2.8.0+cpu","torchaudio":"2.8.0+cpu","torchvision":"0.23.0+cpu","transformers":"4.51.3","timm":"1.0.15","librosa":"0.11.0","numpy":"2.3.5","utmosv2":"1.3.1.dev0"},"checkpoint":{"url":"https://huggingface.co/sarulab-speech/UTMOSv2/resolve/506474f2b33dc77c234d668cc419be1861899cad/fold0_s42_best_model.pth","sha256":"c8149d988e4bbf3f347e6966b5d769de347a5f8c59ffca1dc4bd4bf5b8585e57","bytes":818531314},"feature_configuration":{"config.json":{"url":"https://huggingface.co/facebook/wav2vec2-base/resolve/0b5b8e868dd84f03fd87d01f9c4ff0f080fecfe8/config.json","sha256":"4937977e24d12d1bba70cdce8709c3c04807a8e4ae8ddac4229c48c436ae99ae","bytes":1842},"preprocessor_config.json":{"url":"https://huggingface.co/facebook/wav2vec2-base/resolve/0b5b8e868dd84f03fd87d01f9c4ff0f080fecfe8/preprocessor_config.json","sha256":"b225d617c025463b9e157e06afea8b90dc7078fc70b013c533328423e0486b4a","bytes":159}},"initialization_status":"ready"}}
